Liquid Rocket Engine Design

d _ '" _/ :_'_ _'
,-7{ "_';""_"'-7
i <._.,:>;,
_, 77:7>
. /
-I
i Declasslficd by _ut._or!ty df NA_A-,,._
i
I
.!
!ii i :7
d <:_!
• ['i
.-]_ % % ,,
i.__s_/'"7,7
k.: __-_
............. >....................... • .......... .. • .

NASA SP-125
DESIGN OF
LIQUID
ROCKET PROPELLANT
ENGINES
Dieter K. Huzel and David H. Huang
Rocketdyne Division, North American Aviation, Inc.
_,,x :e_ ,
31
Scientific and Technical Information Division
_:_ '2'/
OFFICE OF TECHNOLOGY UTILIZATION 1967
NATIONAL AERONAUTICS AND SPACE ADMINISTRATION
Washington, D, C.
FORE WORD
Success in space demands perfection. Many of the brilliant achievements

made in this vast, austere environment seem almost miraculous. Behind each
apparent miracle, however, stands the flawless performance of numerous highly
complex systems. All are important. The failure of only one portion of a launch
vehicle or spacecraft may cause failure of an entire mission. But the first to
feel this awesome imperative for perfection are the propulsion systems, especially
the engines. Unless they operate flawlessly first, none of the other systems will
get a chance to perform in space.
Perfection begins in the design of space hardware. This book emphasizes

quality and reliability in the design of propulsion and engine systems. It draws
deeply from the vast know-how and experience which have been the essence of
several well-designed, reliable systems of the past and present. And, wi_h a
thoroughness and completeness not previously available, it tells how the present
high state of reliability, gained through years of research and testing, can be
maintained, and perhaps improved, in engines of the future.
As man ventures deeper into space to explore the planets, the search for
perfection in the design of propulsion systems will continue. This book will aid
materially in achieving this goal.
WERNHER VON BRAUN

Director
Matshall_pace Flight Center, NASA
iii
PREFACE
This book intends to build a bridge for the student and the young engineer: to link the rocket
propulsion fundamentals and elements (which are well covered in the literature) with the actual rocket
engine design and development work as it is carried out in industry (which is very little, if at all
covered in literature). The book attempts to further the understanding of the realistic application of
liquid rocket propulsion theories, and to help avoid or at least reduce time and money consuming
errors and disappointments. In so doing, it also attempts to digest and consolidate numerous closely
related subjects, hitherto often treated as separate, bringing them up to date at the same time.
The book was written "on the job" for use by those active in all phases of engine systems
design, development, and application, in industry as well as government agencies. Since it addresses
itself to human beings set out to create new machines, rather than describing machines about to
dominate man, the language chosen may not always be "functional" in the strict sense of the word.
The book presents sufficient detail to familiarize and educate thoroughly those responsible for
various aspects of liquid propellant rocketry, including engine systems design, engine development,
and flight vehicle application. It should enable the rocket engineer to conduct, independently, com-
plete or partial engine systems preliminary detail designs and to understand and judge the activities
in, and the problems, limitations, and "facts of life" of the various subsystems making up a complete
engine system. It also attempts to educate those ultimately interested in specialized subsystems and
component design (thrust chamber, turbopump, control valves, etc.) about their own as well as neigh-
boring subsystems and about the complete engine system. This should enable the student to prepare
realistic analytical calculations and design layouts with a long headstart toward the final specialized
designs for subsystem production release.
Special emphasis has been placed on engine flight application to stimulate engine systems and
subsystem designers to think in these terms from the outset. The book is intended as a textbook,
with specific consideration of the teacher without industry experience. We hope it will stimulate
those desiring to specialize in the area of a rocket engine subsystem by supplying adequate informa-
tion to enable them to benefit fully from the specialized literature. Thus it provides a realistic expert
introduction for those joining the liquid propellant rocket engine field.
We gratefully acknowledge the most valuable assistance by members of the Rocketdyne and the
Space Divisions of North American Aviation, Inc., Los Angeles. We are particularly indebted to R. E.
Grate, C. A. MacGregor, H. M. Alexander, S. B. Macaluso and T. Holwager of Rocketdyne Division,
and to R. E. G. Epple, R. W. Westrup, R. D. Hammond, and D. A. Engels of Space Division, who
reviewed the various chapters of the manuscript and contributed valuable ideas.
Special recognition goes to R. F. Strauss of Astrosystems International, New Jersey, who

inspired the manuscript and rendered valuable assistance during the various phases of its preparation.
In particular, the authors are indebted to the manifold support they received from North American
Aviation, Inc., and its divisions. Rocketdyne's engine technology has provided a major foundation
for the book.
Dieter K. Huzel
David H. Huang
CONTENTS
Chapter I. INTRODUCTION TO LIQUID PROPELLANT ROCKET ENGINES

1.1 The Generation of Thrust by a Rocket Engine ...................................... i
1.2 The Gas-Flow Processes in the Combustion Chamber and the Nozzle .................. 4
1.3 Performance Parameters of a Liquid PropellantRocket Engine ....................... 10
1.4 Liquid Rocket Propellants....................................................... 18
1.5 The Basic Elements of a Liquid PropellantRocket Engine System .................... 21
Chapter II. ROCKET ENGINE DESIGN IMPLEMENTS

2.1 The Major Rocket Engine Design Parameters ....................................... 31
2.2 The Importance of Design Quality................................................. 50
2.3 Systems Analysis and Design Layout ............................................. 51
2.4 Stress Analysis ................................................................ 56
2.5 Selectionof Materials........................................................... 59
Chapter III.INTRODUCTION TO SAMPLE CALCULATIONS

3.1 Approach ...................................................................... 63
3.2 A-I Stage Engine ............................................................... 64
3.3 A-2 Stage Engine ............................................................... 68
3.4 A-3 Stage Engine ............................................................... 72
3.5 A-4 Stage Engine ............................................................... 74
Chapter IV. DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES

4.1 The Basic Thrust Chamber Elements .............................................. 81
4.2 Thrust Chamber Performance Parameters ........................................... 83
4.3 Thrust Chamber Configuration Layout ............................................. 86
4.4 Thrust Chamber Cooling ......................................................... 98
4.5 Injector Design ................................................................. 121
4.6 Gas-Generating Devices ......................................................... 131
4.7 Ignition Devices ................................................................ 136
4.8 Combustion Instability .......................................................... 143
Chapter V. DESIGN OF PRESSURIZED-GAS PROPELLANT-FEED SYSTEMS

5.1 Determinationof Pressurant Requirements ......................................... 151
5.2 Stored Gas Systems ............................................................. 156
5.3 PropellantEvaporation Systems .................................................. 165
5.4 Systems Evaporating Nonpropellants.............................................. 167
5.5 Systems Using Products of Chemical Reactions..................................... 167
5.6 Selectionof the PressurizationSystem ............................................ 173
vii
viii DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
Page
Chapter YI. DESIGN OF TURBOPUMP PROPELLANT-FEED SYSTEMS
6.1 Elements of Turbopump Propellant Feed Systems ................................... 176
6.2 Turbopump System Performance and Desi_ou Parameters .............................. 186
6.3 Design of Centrifugal Pumps ..................................................... 204
6.4 Design of Axial-Flow Pumps ..................................................... 225
6.5 Design of Turbines ............................................................. 238
6.6 Design of Turbopump Bearings, Seals, and Gears ................................... 257
6.7 Design Layout of Turbopump Assemblies .......................................... 261
Chapter VII. DESIGN OF CONTROLS AND VALVES

7.1 Control Methods ............................................................... 263
7.2 Basic Liquid Propellant Rocket Engine Control Systems ............................ 265
7.3 Engine Thrust Level Control .................................................... 267
7.4 Propellant-Mixture-Ratio and Propellant-Utilization Control ......................... 268
7.5 Thrust-Vector Control .......................................................... 272
7.6 Design Considerations for Fluid-Flow Control Components .......................... 280
7.7 Design of Dynamic Seals for Fluid-Flow-Control Components ........................ 289
7.8 Design of Propellant Valves .................................................... 294
7.9 Design of Control Pilot Valves .................................................. 301
7.10 Design of Fixed-Area-Type Regulating Devices .................................... 307
7.11 Design of Servovalves .......................................................... 309
7.12 Design of Gas Pressure Regnalators .............................................. 315
7.13 Design of Liquid Flow and Pressure Regulators ................................... 321
7.14 Design of Pressure Relief Valves ................................................ 322
7.15 Design of Miscellaneous Valves ................................................. 325
Chapter VIII. DESIGN OF PROPELLANT TANKS

8.1 Basic Design Configurations of Propellant Tanks ................................... 329
8.2 General Design Considerations for Propellant Tanks ................................ 332
8.3 Structural Design of Propellant Tanks ............................................. 336
8.4 Design of Storable Liquid Propellant Tanks ........................................ 343
8.5 Design of Cryogenic Liquid Propellant Tanks and Their Insulation .................... 345
8.6 Design of Fiber-Glass Filament-Wound Liquid Propellant Tanks ...................... 348
8.7 Design of Propellant Tank Pressurant Diffusers .................................... 349
8.8 Propellant Expulsion Under Zero Gravity or Oscillatory G-Loading Conditions .......... 349
Chapter IX. DESIGN OF INTERCONNECTING COMPONENTS AND MOUNTS

9.1 The Principal Interconnecting Components and Mounts ............................... 353
9.2 Design of Tubings and Tube Fittings .............................................. 358
9.3 Design of Flange Joints ......................................................... 360
9.4 Design of Brazed Joints for Rocket Engines ........................................ 370
9.5 Design of Ducts for Rocket Engines ............................................... 372
9.6 Design of Gimbal Mounts ........................................................ 379
Chapter X. ENGINE SYSTEMS DESIGN INTEGRATION

10.1 Systems Engineering ........................................................... 383
10.2 Engine System Design Integration by Dynamic Analyses ............................ 384
10.3 Design Integration for Engine System Calibration .................................. 390
10.4 Engine System Integrated Performance Characteristics .............................. 394
10.5 Mechanical Integration of Engine Systems ......................................... 399
10.6 Electrical System .............................................................. 403
CONTENTS iX
Page
10.7 Engine Instrumentation ......................................................... 411

10.8 Clustering of Liquid Propellant Rocket Engines ................................... 415
10.9 Engine-to-Vehicle Interface ..................................................... 419
Chapter Xl. DESIGN OF LIQUID PROPELLANT SPACE ENGINES

11.1 PrincipalSpace Engine Applications............................................. 429
11.2 General Design Considerations.................................................. 430
11.3 Design of SpacecraftMain Propulsion Systems..................................... 435
11.4 Design of Reaction Control Engine Systems ....................................... 442
INDEX .............................................................................. 451

Chapter I
Introduction To Liquid Propellant Rocket Engines
In order that the reader may better understand Combining these two fundamental relations, we
the basic laws and the operation of liquid pro- obtain
pellant rocket engine systems, a brief review of
the fundamentals is presented.
F =t v (1-3)
This expression, known as the momentum theo-

1.1 THE GENERATION OF THRUST
rem, is the basic thrust equation for rocket en-
BY A ROCKET ENGINE
gines. When applied to rocket engines, the term
for mass, m, and the term for velocity, v, may
The function of a chemical rocket engine apply either to the vehicle or to the ejected
system is to generate thrust through combustion; gases. The products of v and m, in opposite
i.e., release of thermal energy derived from the directions, must be equal, as prescribed by the
chemical energy of the propellants. The gener- law of action and reaction. This condition exists
ated force (pressure) imparts a momentum to the even in a "tiedown" static rocket firing. In this
combustion products. In accordance with the case, however, the "vehicle mass" (the earth) is
basic laws of motion, a momentum in the oppo- so large that reaction effects are undetectable.
site direction is also imparted to the vehicle. In The vehicle designer is primarily interested
practice, high temperature, high-pressure gases in the utilization of the engine thrust available
are produced in combustion chambers through for the acceleration of the vehicle, which at any
chemical reactions of either solid or liquid pro-
point of the trajectory may be expressed as
pellants. These gases are ejected through a
nozzle at high velocity. The operation of a
rocket engine system is independent of its en- F = Wmf = Wma (1-3a)
vironment except for slight effects on perform-
ance caused by ambient air pressure. The rocket The vehicle designer uses this equation for ve-
or, in a more general sense, the "reaction motor" hicle design and trajectory calculations, properly
presently is the only practical device able to considering that thrust F and vehicle mass Wm
propel a vehicle in space. change during flight.
Let us examine briefly the process of thrust In contrast, the engine designer and builder
generation and summarize the most frequently is primarily concerned with the generation of
used laws and formulae needed to design the thrust. His attention, therefore, is focused on
shape and to predict the behavior of a rocket the efficient conversion of the chemical energy
engine. These laws are mere adaptations of of the propellants into thermal energy, and thus
basis physical laws. We know that into kinetic energy of the gaseous combustion
products. His particular concern is to do this in
F= ma (1-1) the most efficient way. For the designer, the
basic equation (1-3) may be rewritten as
Force equals mass times acceleration. We also
know that the velocity increase experienced by
_F =-_Ve = fnVe (1-3b)
the accelerated mass, during the time the force
is imparted, is
where m is the mass flow rate of the gases, and
v = at (1-2) v e is their velocity at the nozzle exit. Even in
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
this simple form, it becomes clear that, for a where Pe (lb/in2 or psia) is the static abso-
given mass flow rate, thrust will increase with lute gas pressure in the exit plane Ae (in2). The
increased gas velocities obtained. equation assumes that the injection flow velocity
It should be remembered that the vehicle, in of the propellants can be neglected and that the
particular the thrust chamber of the engine, is flow of gases through the exit plane is one-
also subject to the pressure environment, which dimensional; i.e., that all molecules of the gas
is a function of altitude. Equation (1-3b) states

move on parallel lines. The expression--re
that if a mass is flowing out of a container, the g
sum of all internal and external forces acting on is often referred to as the momentum thrust, and
all surfaces of this container is equal to the AeP e as the pressure thrust. The pressure thrust
total momentum flowing out of the surface. The is not a desirable form of thrust generation in
liquid propellant rocket thrust chamber, with the rocketry. The presence of the term AePe indi-
inclusion of tile exit plane, is such a container cates that not all of the pressure forces avail-
able have been converted into gas kinetic energy
(fig. 1-1).
inside the chamber nozzle. In other words, a
portion of the gas pressure generated by the re-
lease of chemical energy has not been used for
the generation of gas momentum. It is the spe-
cific function of the thrust chamber nozzle to
convert, at maximum efficiency, the available
F_ ...........
chamber pressure into gas momentum, and thus
_ Pe obtain maximum thrust for a given propellant flow
rate.
We now assume that the rocket is operated at
Figure I-1 a finite ambient pressure Pa > 0 (lb/in 2) (low-

altitude condition). The resulting pressure
forces acting on the outside of the chamber walls
Let us first assume that the chamber is oper-
have no effect on the gas on the inside. How-
ated at an ambient pressure Pa = 0 (high-altitude
ever. these pressure forces cancel part of the
condition). Then, the net force acting on the
pressure thrust by an amount AePa. Since gases
gas in the chamber is the sum of the reactions
are flowing with supersonic velocity in the exit
from the chamber walls and of the reaction of the
plane A e, the ambient pressure Pa does not have
absolute gas pressure at the exit. These two
access to it. The ambient pressure Pa thus cre-
reaction forces are opposed (fig. 1-1). Accord-
ates a net unbalanced force onto the projected
ing to the momentum theorem, the net force on
thrust chamber area (opposing the thrust) of mag-
the gas must be equal to the momentum flux out
nitude Aep a. Including this term in equation
of the chamber:
(1-5), the general rocket thrust equation is ob-
tained:
A Ptc dA- AePe --_-V e F =/2: Ve. Ae(Pe - Pc) (1-6)

tC g
Tile integral describes that force F (Ib) which

acts on the thrust chamber and thus on the ve- The following model may extend the under-
hicle. We can write: standing of the nature of the terms which com-
pose equation (1-6). Let us assume we have a
(1-4) movable cylinder (representing the thrust cham-

F-AePe =_Ve ber and vehicle ,nass), a spring (representing gas
or pressure), a piston (representing the gas mass),
and a stationary rack (representing ambient con-
F = W---ve+ AePe (1-5)
g ditions) (fig. 1-2).
INTRODUCTIONTO LIQUID PROPELLANT ROCKET ENGINES
MOVEABLE CYLINCER (REPRESENTING

(g/_/)indicatesthatoptimum ve has not been
THRUST CHAMBER AND VEHICLE)
obtained.
MASS)
Sample Calculation (I-1)
-l_:,Ig_YI'iY'_LI
_U-- -T.
The following data are given for a liquid pro-
pellant rocket engine: thrust, F : 100 000 lb at
(REPRESENTING '_- STATIONA RY RACK sea level; propellant consumption rate, g/= 369.3
GAS PRESSURE) (REPRESENTING AMBIENT
CONBITION$)
Ib/sec; thrust chamber exit area, A e = 760.8 in 2"
Figure I-2 gas exit static pressure, Pe : 10.7 psia; ambient
pressure, Pa : 14.7 psia (sea level); gravitational
The spring is so made that its end slips side- constant, g: 32.2 ft/sec _.
ways upon reaching the end of the cylinder and From what we have just learned, we will de-
engages the stationary rack. The cylinder is termine (a_) gas exhaust velocity, (_b_)engine
suspended in a suitable manner to move freely. thrust in space, and (c__)the effective exhaust
When releasing the spring force ("Pc"), the velocities at sea level and in space.
"gas" is expelled to the rear. If, upon reaching
the chamber exit, some spring force remains, the Solution
spring engages the rack and continues to act (a) From equation (1-6) the gas exhaust
upon the cylinder, but ceases to act upon the velocity
"gas." We find that the model works for all
cases: underexpanded (as assumed above, where ve : IF - Ae(P e - pa)](g/g/)
spring free length is longer than cylinder length);
overexpanded (spring free length is less than = [100 000- 760.8(10.7- 14.7)](32.2/369.3)
cylinder length and the spring force is exhausted : 9040 ft/sec
prior to the "gas" reaching the exit, the "gas"
therefore being subject to deceleration within Our calculation assumes a nozzle somewhat too
the cylinder); and ideal expansion (where spring long for sea-level conditions, as indicated by
free length equals cylinder length). the fact that Pe is smaller than Pa; a pressure
The model can also illustrate the case of the "undershoot" and an exhaust velocity "over-
overexpanded nozzle without jet separation, shoot" occurred. If no jet separation occurred,
which will be further explained below. This i.e., if the nozzle remained _filled" to the exit
situation is equivalent to that of the inertia of plane, the calculation is valid. The "penalty"
piston ("gas") and spring pulling the spring be- of incorrect nozzle length simply appears as the
yond its null point. The negative-loaded spring, negative thrust term Ae(Pe-Pa). If jet separa-
in engaging the rack ("ambient"), will pull the tion does occur within the nozzle, or if it is
cylinder backward. combined with decelerating shock waves, the
Equation (1-6) is often expressed as situation becomes considerably more complicated
and requires elaborate mathematical treatment.
However, there should be no concern at this point.
F:c-- (1-7)
g From equation (1-6), we know that the dif-
ference in thrust between space and sea level is
Where c is defined as the effective exhaust AeP a. Since the nozzle selected was too long at
velocity (ft/sec) and comprises sea level, this thrust increase AeP a during rocket
ascent will be obtained in two distinct steps.
c= v e + Ae(Pe - Pa) (g./W) (1-8) First, by reduction of the negative thrust term
Ae(Pe-Pa) to zero. This will occur when Pe = Pa;
The effective exhaust velocity is not the that is, when the rising vehicle reaches an alti-
actual gas velocity except when Pe : Pa where c tude where Pa = 10.7 psia, in our specific case.
becomes equal to Ve. As explained with equa- As we have learned, this represents ideal expan-
tion (1-6), the presence of a term Ae(Pe- Pa) sion. As the vehicle continues to ascend farther
and eventually reaches "empty space" where ible, it is additionally called an isen-
Pa---0, the increase of the positive term tropic process.
Ae(Pe- Pa) raises the thrust level farther. The (4) No friction
combined effect of the two phases, however, is (5) Steady flow rate
simply the elimination of paAe, provided the (6) One-dimensional flow (all gas molecules
nozzle is filled at all times. move on parallel lines)
Thus, we obtain engine thrust in space: (7) Velocity uniformity across any section
normal to chamber axis
F= 100000+760.8× 14.7= 111 183.8 lb (8) Chemical equilibrium established within
the combustion chamber and not shifting
(c) From equation (1-8) the effective exhaust in the nozzle.
velocity at sea level results Certain correction factors, usually empirically
obtained, will be applied to the results derived
c = v e + Ae(Pe - pa)(g/W) from these ideal assumptions in the actual de-
sign of a rocket and for the prediction of its
=9040-_760.8×(10._- 14.,)_ X (32.2,' 369.3)
behavior.
= 8772, ft/sec
and in space
c= v e + AePe(g/_l)
= 9040 + 760.8 × 10.7 × (32.2,/369.3)
= 9750 ft/sec
1.2 THE GAS-FLOW PROCESSES IN THE

COMBUSTION CHAMBER AND THE
NOZZLE
Since the analytical treatment of compressi-

ble fluids flowing through cylindrical ducts and
nozzles can be found in standard aerodynamics
and thermodynamics textbooks, no attempt will
be mad'_ here to derive basic equations governing
gas flows. Rather, significant applications of
those equations used in actual rocket design are
presented.
The parameters and terms applicable to gas
flows in a liquid propellant rocket thrust chamber
are shown in figure I-3 and table 1-1. These
parameters serve to define the characteristics of The Perfect Gas Law
gas flow at various points within the thrust cham-
At any section X the peifect gas law states:
ber. Gas-flow calculations for rocket thrust-
chamber design usually assume the following 144pxVx = RTx (1-9)
ideal conditions:
(1) Homogeneous gas composition
(2) Perfect gas
The Principle of Conservation of Energy
(3) No heat transfer through the motor walls
in either direction; i.e., adiabatic proc- In an adiabatic process, the increase in
esses. If no increase in entropy occurs, kinetic energy of the flowing gases between any
i.e., if the process is considered revers- two points is equal to the decrease in enthalpy.
ix _ • / •
INTRODUCTION TO LIQUID PROPELLANT ROCKET ENGINES
TABLE l-l.-Terms Used in the Calculation el The Principle of Conservation of Matter

Gas Flows
V¢ : AIvi Axvx
144 Vi = 144 Vx =constant (1-11)
aC, a¢ Local velocity of sound in chamber
and at nozzle throat (ft/sec);
(at =v gy-F/_).
mc Cylindrical cross-sectional area of The IsentropicFlow Process
chamber (in2).
Aj, At, Ae, Ax Flow areas at nozzle inlet, throat and For any isentropicflow process the following
exit; and at any section X normal to relationshold between any two points:
axis (in2).
Cp, Cv Specific heats for constant pressure
piViY: PxVxY= constant (1-12)
and for constant volume (Btu/lb ° F).
g Gravitational constant (32.2 ft/sec -_ at
sea level). and
J Energy conversion factor (778 ft-lb/
Btu).
TI/Tx:(pI/px)(Y-9"Y=(Vx/VI)Y -1 (1-13)
Mc, M i, M_, M e , Mx Flow Mach number (v/a) at chamber;
nozzle inlet, throat and exit; and at
any section X normal to axis. Gas Flow Through Liquid PropellantRocket
Molecular weight of combustion Combustion Chambers
products.
(Pc)tnj Chamber total pressure at injector end The functionof a liquidrocket combustion
(lb/in2). Because of the relatively chamber is to convertpropellantsintohigh-
low propellant injection flow veloc-
temperature,high-pressuregas through combus-
ities vtaj, the measurable static
pressure at this station is generally tionwhich releases the chemical energy of the
treated as equivalent to the total propellant,resultingin an increase of internal
pressure. energy of the gas. Combustion chambers are
(Pc)ns Nozzle stagnation pressure or chamber
generallytubular,as shown in figure1-3. The
total pressure at nozzle inlet (lb/in2 );
(Pc)ha = pi[l + _ (y- I)Mi] Y/y- '. liquidpropellantsare injectedat the injection
Pi, Pt, Pc, Px Flow static pressures at nozzle inlet, plane with a small axial velocitywhich is as-
throat and exit; and at any section X sumed to be zero in gas-flow calculation.The
normal to axis (lb/in2). combustion process proceeds throughoutthe
R Gas constant (1544Dli)(ft/°R)
lengthof the chamber and is assumed to be com-
(Tc)ns Nozzle stagnation :emperature or
chamber total temperature (° R). pleted at the nozzle inlet.As heat is liberated
(Te)ns = Ti[1 + 'z(y- 1)Mi] between injectionplane and nozzle inlet,the
Ti, T¢, Te, Tx Flow temperature at nozzle inlet, specificvolume of the gas is increased. To
throat, and exit; and at any section
satisfythe conditionsof constantmass flow, the
normal to axis (OR).
gas must be accelerated toward the nozzle inlet
Vin/ Injector flow velocity =0 (by assump-
tion). with some drop of pressure. In brief, the follow-
Vi, Vt, re, V x Flow velocities at nozzle inlet, throat, ing takes place:
and exit; and at any section X nor-
The gas-flow process within the combustion
mal to axis (ft/sec).
V_, Vt, Ve, Vx chamber, that is, within the volume upstream of
Flow specific volumes at nozzle inlet.
throat, exit; and at any section X the nozzle entrance, is not entirely isentropic
normal to axis (fta/lb). but is a partly irreversible, adiabatic expansion.
Steady weight flow rate (lb/sec). Although the stagnation temperature or total
E Nozzle expansion area ratio (Ae/At).
temperature remains constant, the stagnation
(c Nozzle contraction area ratio (Ac/At).
Y Specific heat ratio (Cp/Cv). pressure or total pressure will decrease. This
causes permanent energy losses, which are a
function of the gas properties as expressed by y,
Applied to a nozzle, this yields for unit weight and of the nozzle contraction area ratio ec or
of gas flowing (Ac/At). Wherever the acceleration of gases is
largely effected by expansion due to heat re-
1 2 lease, rather than by a change of area as in a
(i-10)
_-2(Vx - vi:): Cp(Ti - Tx) nozzle, the stated losses occur. The greater the
contribution of the nozzle, the more efficient is of the converging-diverging De Laval type, with
the gas acceleration. Conversely, with no noz- the cross-sectional area decreasing to a minimum
zle attached, the losses are maximum. The great at the throat and then increasing to the exit area,
importance of ec to the thrust chamber design be- as shown in figure 1-3. The flow velocity
comes apparent. It will be discussed further in through a nozzle increases to sonic velocity at
chapter IV. the throat and then increases further superson-
Figure 1-4 shows the loss of total pressure ically in the diverging section.
for two typical y values as a function of the In practice it is assumed that the gas flow
nozzle contraction area ratio ec. These data are through a rocket nozzle is an isentropic expan-
generally used in rocket design, and are calcu- sion process, and that both the total temperature
lated from the Rayleigh flow process. and the total pressure remain constant throughout
the nozzle. The pressure ratio Pt/(Pc)ns be-
tween throat and chamber is called the critical
pressure ratio and is solely a function of spe-
cific heat ratio
09
1.0
08 Pt/(Pc)ns = [2/(y+ 1)] y/(F-0 (1-16)

o.
, l l I L L 3 The static pressure Pt at a nozzle throat with

i.o 2.0 30 40 Ao/At sonic flow, where the maximum weight flow per
( CYLINDER )
unit area occurs, is defined as critical pressure.
Figure 1-4 The velocity of sound is equal to the velocity of
propagation of a pressure wave within a medium.
Neglecting the flow velocity at the injecting It is, therefore, impossible for a pressure dis-
turbance downstream of the nozzle throat to in-
end, i.e., assuming Vin j = 0 and (Pc)inj = Pinj, the
total pressure ratio (Pc)inj/(Pc)ns can also be fluence the flow at the throat or upstream of the
expressed in terms of flow Mach number Mi at throat, provided that this disturbance will not
create a higher throat pressure than the critical
the nozzle inlet and of the specific heat ratio y:
pressure.
It is one of the characteristic features of an
(Pc)inj/(Pc)ns =
attached diverging or De Laval nozzle, however,
that sonic velocity in the nozzle thrcat is main-
(1 + yMi_)/(1 +_'_Mi2) )''(y-I) (1-14)
tained even if the back pressure (ambient pres-
sure) at the nozzle exit is greater than the pres-
For the reasons mentioned above, it is desirable sure required at the throat for sonic velocity. As
that the Mach number at the nozzle entrance be
a result, a pressure adjustment (recovery) must
small. A typical value for a thrust chamber with take place between the throat and the nozzle
a contraction area ratio of Ac/At= 2 is Mi=
exit (ambient pressure). This adjustment may
0.31(),: 1.2). For the static pressure ratio, the take place through subsonic deceleration (isen-
expression simplifies to tropic), or by way of nonisentropic discontinui-
ties called shock waves, or by a combination of
Pinj/Pi = 1 + y Mi 2 (1-15) both. Figure 1-5a represents several of the pos-
sible conditions that may occur in a overex-
panded nozzle. The situations shown represent
Gas Flow Tl_ough Rocket Nozzles
cases of an overexpanded nozzle which was
The prime function of a rocket nozzle is to mentioned earlier.
convert efficiently the enthalpy of the combus- We see that pressures lower tba_ ambient may
tion gases into kinetic energy and thus high gas be obtained in a supersonic nozzle. The higher
exhaust velocity. The nozzle is the most effi- ambient pressure cannot advance upstream within
cient device for accelerating gases to supersonic the nozzle, however, since the gases are flowing
velocities. Rocket nozzles are conventionally with supersonic velocity. An exception is along
Or
EXF&NS_ON
Pe ' Pa JET SEPaRaTION
lit
(al
,>%
,
iPe < F==

*'-
Ve = V _
297
R(Tc)ns I< 1-
Pe
(P"_-ns (1-
lS)
Theoretical gas weight flow rate:

TYPICAL PRESSURE RECOVERY
NOZZLE (I_NT ROP!C) ONLY

[ y+t
W - A_(pc)ns _gT[2/(Y + 1)]Y:-I (i-19)

TYPtCAL
THROUGH
P_ESSURE
S_CX
RECOVERY
WAVES (NON- R-- L
DECELERATION (ISEN TROPPC],
(P,)-4_
,
i TYPlCAL _ESSURE RECOVERY
Theoretical nozzle expansion ratio:
THROUGH SHOC_ WAVES ONLY NO
FURTHER SUBSON*C PRESSURE RE-
'' - -"--k/ / _/CO_ERY, o_ETo_eTSEOAR,T

0_ 1 1
PRESSURE I \\ 1
( 2 _Y-IF!Pc)ns]Y
_@ *DEAL EXPANSION (P I : PQ)
(1-20)
OIST£NCE ALONG NOZZLE
(b)
AXiS
,/ ( po y
I _ - \(Pc)ns]
Figure 1-5.-a, Effect of incorrect nozzle length
and of jet separation on thrust F; b, pressure
distribution in overexpanded De Laval nozzle. At the throat:
the nozzle walls, where, due to friction, a bound- __L_
ary layer of slow-moving gases may exist. In [- 2 ]Y -1 (1-21)

Pt = (Pe)ns LF_-J
this subsonic boundary layer, ambient pressure
may advance for a distance, forcing the low-
pressure center jet away from the walls. It might 2gy
vt=V_--_R(Tc)ns (1-22)
be expected that the point of separation will be
at the point of optimum expansion. Actually, in-
fluenced by wall friction, nozzle divergence At any section X between nozzle inlet and
angle, and other factors, separation usually oc- nozzle exit:
curs further downstream. In fact, it rarely occurs
at all in conventional rocket nozzles within their
[ +Y__t.L
design region of operation, unless an extreme IF- y-1 -1 y-_
case of overexpansion exists or unless exces-
(1-23)
sive nozzle divergence angles are chosen. Thus,
in many cases it is correct to base all nozzle
calculations on the assumption that no separa-
tion occurs; i.e., that the nozzle is "filled" at
At any section X between nozzle inlet and the
all stations (see fig. 1-Sb).
nozzle throat:
Following are some significant and useful
relations for an ideal gas flow through a rocket
nozzle: .y_=2_3__E22
. y q2(y-_)
Theoretical exit velocity: 2 /
(1-24)
At
2 (Pc)ns_ y
/ -
At any section X between the nozzle throat and divergent sections of a De Laval nozzle are
the nozzle exit: shown in figure I-6.
1 1
Useful values of functions of the specific

[ 2 _y-1/(,.___c)_s,
heat ratio y are listed in table 1-2.
Ax ky+l] p_
(1-25)
Sample Calculation (1-2)

p, 5)'?";
The following data are given for the thrust
chamber of an ideal liquid propellant rocket

engine: thrust chamber propellant flow rate,
l_tc = 360.7 ib/sec; nozzle stagnation pressure,

J2gy t r
)'-P 11(1.26
)
Vx= _ Y -1R(Tc)ns -_ s/' (Pc)ns = 1000 psia; chamber total temperature,
(Tc)ns=6540 ° R; product gas molecular weight,
_{ = °_2.67; gas specific heat ratio, y= 1.20; noz-
./ F zle expansion area ratio, _ = 12. The following
Vx
: Ii T y+ 1
(-,
,c),sp.._ _ ) Y (1-27) conditions
at injecting
are
plane,
also assumed:
Minj : O; flow
flow
Mach
Mach
number
number
at
the nozzle inlet, Mi = 0.4. (In practice, thrust

Variations of isentropic pressure ratio and Mach chamber design values for Mi range from 0.15 to
number with the area ratio in the convergent and o.45.)
MACH NUMBER Mx
0.1 0.2 0.5 I Z
IOO
5o
2o
rr
',' e
CONVERGING
--_)-- 1.40
, yy. 1.10 ,#1.
MACH NUMBER _;t

VS r-.
i,/ PRESSURE RATIO
AREA RATIO ._
ill /
20 50 IOO 200
PRESSURE RATIO(Pc)nr,/Pa
Figure 1-6.-Variations of isentropic pressure ratio and Mach number with area ratio in converging and
diverging sections of De Laval nozzle.
TABLE 1-2.-Useful Values o[ Functions o[ the From equation (1-21):

Speci[ic Heat Ratio y
4/( 2 _(y+l} (y-O

Pt = (Pc)ns : 1000 × 0.564 = 564 psia
1.10 26.61 0.0909 0.5847 0.6590

1.15 22.21 .1304 .5744 .6848 From figure 1-6 or equation (1-25) at
1.20 1965 .1667 .5645 .7104 Ax/At = 4
1.21 19.26 .1736 .5626 .7155
1.22 18,89 .1803 .5607
.7205 1 _ 1000
1.23 18,55 .1870 .5588
.7257 Px=(Pc)nsX23 23 =43.5 psia
1.24 18.23 .1936 .5569 •7307
1.25 17.94 .2000 .5549 .7356
1.26 17.66 2064 .5532 .7408 From figure 1-6 or equation (1-20):
127 17.40 .2126 .5513 .7457
1.28 17.15 .2188 •5494 .7508
1.29 16.92 .2248 .5475 ,7558 (Pc)ns 1000
- - -- = 9.85 psia
1.30 16.70 .2308 .5457 .7608 Pe 101.5 101.5
.33 16.10 .2481 .5405 .7757
.36 15.59 .2647 .5352 .7906
1.40 15.01 2857 .5283 .8102 (b) Flow temperatures:
.50 13.89 .3333 .5120 .8586 Since (Tc)in j = (Tc)ns = constant and Minj = 0
1.60 113.10 ,3750 .4968 .9062
Tinj = (Tc)inj = (Tc)ns = 6540 c R
Determine the following: a(_ flow static pres-

sures: Pin j, Pi, Pt, Px at Ax/At : 4, and Pc; (_ By definition:
flow temperatures: Tinj, Ti, Tt, Tx, Te; c(c(c(c(_
flow specific volumes: Vinj, Vi, Vt, Vx, Ve; (Tc)ns
d(.d) flow velocities: vi, vt, Vx, re; flow Mach Ti: + ;(×_1)M,2]
numbers: Mx, Me; (f) flow areas: Ac, Ai, At,
Ax, Ae. 6540 6540
- = 6440 ° F
1 +0.1 x 0.16 1.016
Solution
From equation (1-13):
(a) Flow static pressures:
Y
Tt ( c)ns =6540x0-909=5945° R
(pc)inj=(Pc)ns(1 + yMi2)/ t 1 +-_ Mi 2) }'-1
y-I
= 1000 x (1 + 0.16 × 1.20) _ 1.192 = 1082 psia 1
o. °-
o) Tx=( Tc )ns-PL(p-_-_
_I
- × = 6540 × I._ = 38800 R
y-i
Since, by assumption, Minj = 0
T Pe Y
e = (Tc)ns [(p-p_n s ] : 6540 x 2-_: 3025° R
Pinj = (Pc)inj = 1082 psia
From equation (1-15) (c) Flow specific volumes:
Pi = Pinj/(1 + yMi 2) = 1+0.16×1.20-

1082 1082 R = 1544 __ 1544 = 68
1._ = 909 psia 22.67
10 DESIGNOF LIQUID PROPELLANT ROCKET ENGINES

Mx = Vx _ 8360_ 2.59
ax 3226
RTinj 68×6540
= 2.846 cu ft/lb
Vinj - 144 Pinj 144 x 10S2 ae -- \ gy---_e = \'32.2 ), 1.2 × 68 x 3025 = 2820 ft,/sec
ve 9620
RTi _68 x 6440=3.34 cu ft/lb
Vi - 144 Pi 144 × 909 Me = _-_e= 2--820= 3.43
RTt 68 x 5945 (f) Flow areas:

t.97 cu ft/lb From equation (1-11):
Vt -144p_ 144×564
RTx 68 x 3880
= 144px "= 144x43.5 - 42.1 cu ft/lb Ai= 144 WtcVi _ 144 × 360.7 × 3.34 = 105.4 in 2
Vx vi 1646
Ac:Ai:105.4 in 2
RTx 68 x 3025
Ve =144px=144x9.85-145.1 cu ft/lb
At = 144WtcVt _ 144 x 360.7 x 4.97.= 65.4 in 2
vt 3958
(d) Flow velocities:
Since the sonic velocity a i = \,_ y RTi 144 WtcVx 144 x 360.7 x 42.1
An - = 261.8 in 2
vx 8360
vi=Miai=O.4 ×32.2x l.2x68x6440=O.4 x 4110
= 1646 ft/sec
or
vt = M_at = 1 x 32.2 x 1.2 x 68 x 5945 = 3958 ft/sec

Ax=4×At=261.8 in 2

Ae-
144WtcVe 144 x 360.7 x145.1= 782 in 2
Ve 9670
2gy R(Tc)n _( Px _ or
vx: \(Pc)ns/J
Ae= 12xAt=782 in2
= ¥J6540x 64.4x 6 × 68 x [I- 0.593]
= \/64.4× 6 x 68 x 0.407x 6540

1.3 PERFORMANCE PARAMETERS OF A
= 8360 ft/sec LIQUID PROPELLANT ROCKET ENGINE
The performance of a rocket engine is ex-

From equation(I-18):
pressed by a quantitycommonly called "specific
impulse,"Is. Ifimpulse impartedto the vehicle
and propellantweight consumption were meas-
ured during a given time interval,Is would have
the dimension ib-sec/Ib. In practice,thrustis
usuallymeasured, in conjunctionwith propellant
weight flow-ratemeasurements. This yieldsthe
= \/64.4 x 6 x 68 x 6540 x 0.543 = 9760 ft/sec same dimension: ]b/(Ib/sec).Is may thus be
expressed as
(e) Flow Mach numbers:
Since
Is = F/'ff (1-28)
ax = vg),_--RTx= \32.2 x 1.2 x 68 x 3880 = 3226 ft/sec Since weight is the force exerted by a mass
INTRODUCTION TO LIQUID PROPELLANT ROCKET ENGINES 11
on its rigid support under the influence of gravi- would result. The situation would be further
tation (by convention at sea level on Earth), it has complicated by the fact that it is nearly impossi-
become accepted practice to measure Is in "sec- ble to improve the specific impulse once an en-
onds," by canceling out the terms for the forces. gine and thrust chamber have been designed, for
Obviously, the expression does not denote a a given propellant combination.
time, but rather a magnitude akin to efficiency. Another important performance parameter is
is directly contributes to the final velozity of the propellant mass fraction Rp of the complete
the vehicle at burnout and thus has a pronounced vehicle, of which the engine system is a part.
effect on range or size of payload, or both. This The propellant mass fraction is defined as
will be shown further below in connection with
equation (1-30). Usable propellant mass (1-29)
Rp- Initial rocket mass
It is important to state whether a specific im-
pulse quoted refers to the thrust chamber assem-
bly only (ls)tc, or to the overall engine system where the initial rocket mass is equal to the sum
(Is)on- Often, the distinction may not be self- of the masses of the engine system at burnout,
evident. It is important, therefore, to state the structure and guidance system, the payload,
accurately to what system the quoted specific and the propellant. The significance of the pro-
impulse refers. For instance, in a turbopump fed pellant mass fraction can be illustrated by the
system, overall engine specific impulse may in- basic equation for the rocket burnout velocity
clude turbine power requirements, vernier, and Vbo (ft/sec)
attitude control devices. All of these may be 1
fed from one or all of a given vehicle's propel- Vbo=Cvc" g(ls)oa In 1-Rp (1-30)
lant tanks. If they are properly considered, the
user, in this case the vehicle builder, will obtain where the coefficient Cvc corrects for the effects
the correct value for his own optimization stud- of aerodynamic and gravitational forces. It is
ies, which include propellant tank sizes, payload composed of several parameters which vary with
weight, and range, among other parameters. type of trajectory and with elapsed time during
In many instances, statement of the specific flight. Although they are of no concern here,
impulse (ls)tc for the thrust chamber only may be they are of great importance to the vehicle
desirable, such as during the component develop- builder. Also, they cannot be neglected for
ment period of this subassembly. Since, in that rigorous engine design analyses which must in-
case, those propellant demands which are inade- clude trajectory information.
quately or not at all contributing to the genera-
tion of thrust are not included, the specific Thrust Chamber Specific Impulse (ls)tc
impulse stated will be higher than for a complete The overall performance of the liquid propel-
system, by 1 to 2 percent, as a rule. The spe- lant thrust chamber is a direct function of the
cific impulse thus stated would be too high for
propellant combination, the combustion effi-
the vehicle builder, who must consider the sup-
ciency of propellants in the chamber, and the
ply of propellants to the auxiliary devices men- product gas expansion performance in the noz-
tioned above as well. If, due to improper identi-
zle. The expression for (Is)tc may be obtained
fication of Is, a thrust chamber value were used
in several ways:
as an engine value, the consequences would be
serious. This becomes clear, if one realizes From equation 1-28:
that when relying on a better-than-actual value,
propellant tank sizes would be designed too
(1-31)
small, resulting in premature propellant deple-
tion. This would eliminate the last seconds of
required burning time, when the vehicle mass Combine equations 1-31 and 1-7:
being accelerated is near empty weight and
acceleration, therefore, is near maximum. A (/s)tc c (1-31a)
substantial loss of range for a given payload g
12 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
The effective exhaust velocity c may be further This form shows that c* is a function of the
defined as the product of two convenient param- properties of the combustion product gas at the
eters, c* and Ct exit of the combustion chamber, i.e., at the noz-
zle inlet; namely, specific heat ratio y, gas
c=c*C! (1-31b) constant R, and temperature (Tc)ns.
where the characteristic velocity c* in feet per

Thrust Coefficient Cf
second (commonly pronounced "cee-star') is a
parameter primarily used to rate the propellant The quantity Cf reflects the product gas ex-
combustion performance. The thrust coeHicient pansion properties and design quality of the
Ct is a dimensionless parameter used to measure nozzle. Combining equations 1-31, 1-31c, and
the gas expansion performance through the noz- 1-32, the expression for theoretical Cf may be
zle. Combine equations 1-31a and 1-31b: written as:
c*Cf F
(Is)tc =- (1-31C)
g CI: At (Pc)n s (1-33)
While I s and Rp are of ultimate importance to the This form shows that Cf measures the force aug-
missile or space vehicle builder, both c* and Cf mented by the gas expansion through the nozzle
are of great and early importance to the engine as compared with the force which would be gen-
and thrust chamber designer and developer. erated if the chamber pressure acted over the
throat area only. By combining equations 1-6,
1-18, 1-19, and 1-33, the equation for theoretical
Characteristic Velocity c*
Ct at any altitude may be rewritten in the fol-
In a system with sonic velocity at the throat, lowing form:
the quantity c* reflects the effective energy
level of the propellants and the design quality of
injector and combustion chamber. It may be de-
2 I y-' y
fined by the following expression: -
(Pc)nsA_ Pe - Pal
c* = (1-32)
g'tc +'
This form shows that c* measures combustion Equation 1-33a shows that CI is a function of
performance in a given combustion chamber by specific heat ratio y, chamber pressure (Pc)ns,
indicating how many pounds per second of pro- ambient pressure Pc, and the nozzle area expan-
sion ratio _.
pellant must be burned to maintain the required
nozzle stagnation pressure. A lower value of As will be noted, the throat stagnation pres-
propellant consumption l_ under the given condi- sure (Pc)ns has been used in equations (1-32)
tion indicates a combustion process of higher and (1-33). This has been the practice in indus-
energy and efficiency and gives a corresponding try and in most of the literature. Briefly, the
higher value of c*. By substituting lYtc with reason is that (Pc)ns reflects the true theoretical
equation 1-19 in equation 1-32, the equation for gas property at combustion chamber exit, i.e., at
theoretical c* may be rewritten in the following the nozzle inlet, and gives a more logical value
form: to c* and C_. In actual operation the true value
of (Pc)ns cannot be measured. (Pc)ns is mathe-
matically converted from the measured value of
c* = \/gyR(Tc)ns (1-3_) the gas static pressure at the injector, Pinj. The
)'+ 1 accuracy of this calculated value has to be veri-
fied by the test results. Likewise, the gas prop-
erties, and thus the specific heat ratio y which
INTRODUCTIONTO LIQUID PROPELLANT ROCKET ENGINES 13
additionally changes along the chamber axis, coefficient CfL is introduced to correct for this
affect the true values of c* and Cf. This will fact. For instance, if the ratio (Pc)ns/Plnj was
have to be verified by actual test results. found to be 0.8 from figure 1-4, CII would have
To understand better the nature of Cf and the to be 1.25 to offset the introduction of (Pc)ns.
design parameters which influence it, let us first As will be seen, the use of a thrust coefficient
rearrange equation (1-33): of 1.25, for instance, in a straight cylinder
thrust chamber for which (Pc)ns/Pinj is 0.8, is
F =(Pc)ns" At" CI (1-34) merely part of a mathematical rearrangement, but
does not signify an increase in thrust for a given
The formula expresses that the thrust gener- Pinj.
ated by a thrust chamber (the effect) is produced It is noted that the combustion chamber in-
by pressure (the cause) as a function of the cluding injector will have to produce the required
physical properties of the chamber itself. The pressure (Pc)ns with a flow rate, the magnitude
relationships and effects of the principal design of which is determined by c* and by the throat
parameters become clearer if we proceed in steps area At. Transformation of equation (1-32)
as follows: shows the relationship:
Assume we wish to generate a certain thrust F.

(Pc)ns =Wtc" c*. I-L-- (1-37)
Our chamber is a straight cylinder (fig. 1-7). At'g
The pressure in this chamber has a very small
effect on the cylindrical wall (the forces normal In actual practice, the value of c* for a given
to the chamber axis will cancel each other), ex- propellant combination and thrust chamber design
cept for effects of friction, which we will ne- is arrived at tentatively from existing experience
glect. There is no part of the chamber for the and is subsequently refined during development
pressure to act upon at the exit. The only cham- testing.
ber area upon which the pressure can act is the Let us now redesign our cylindrical thrust
injector plate. Since, for the cylindrical cham- chamber, as shown in figure 1-8. Maintaining
ber, the injector area Ai is equal to At, we can the same throat area At=Ae, we enlarge the
write: combustion chamber including injector to a diam-
eter somewhat larger than that of At. The flow
Fcyl : Pinj "At (1-35) rate remains V/tc.
In the straight cylindrical chamber (fig. 1-7),
For the reasons explained above, we rewrite to the gas velocity was sonic at the end of the
include (Pc)ns: cylindrical chamber portion, which coincided
with At and Ae. From earlier discussions (see
Fcyl = (Pc)ns "At" C[ 1 (1-36) sec. 1-2), we know that expansion (acceleration)
is nonisentropic in that case. In the redesigned
Since (Pc)ns is smaller than Pinj (fig. 1-4), the chamber (fig. 1-8) by contrast, gas velocities are
still well below sonic velocity at the end of the
cylindrical portion. Most of the acceleration to
I sonic velocity will now occur in the added, con-
vergent nozzle. Since we can make the assump-
tion, which is essentially correct, that the com-
bustion process is complete at the end of the
INJECTOR -----F_ --- cylindrical chamber portion, the subsequent
Pa
expansion (acceleration) in the convergent noz-
zle is assumed to be isentropic; i.e., to occur
Ain j =A¢ J without further total pressure losses. Since we
A¢ = A i =A t keep lYtc and At constant, and assume that c*
= Ae
remains unchanged, the nozzle stagnation pres-
Figure i-7 sure (Pc)ns, too, will retain the same value as in
]4 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
INJECTOR
go
1 I
Ainj "A¢ At=A+ At<A e
A I >A t
Ainj =A¢
Ac=Ai At =A e Figure 1-9
Ai>A t
expansion to atmospheric pressure, flowing
Figure 1-8 freely in all directions. By attaching a diver-
gent nozzle, we prevent the gases from dissipat-
the case of the straight cylindrical chamber. ing at random, and further accelerate the gases
However, because of the reduced pressure losses in one preferred direction only. Since this proc-
in the combustion chamber, the required total ess takes place in the divergent part of the
pressure at the injector end will definitely be thrust chamber, the static pressures of the ex-
lower. panding gases produce a force on the chamber,
The redesign, then, has the favorable result as indicated by the arrows in figure 1-9. The
that, for instance, in pressurized systems, the expression of the thrust for the complete thrust
same propellant flow rate can be sustained with chamber with convergent-divergent nozzle can
lower tank pressures, thus, slightly lighter tanks now be written as:
can be used. In turbopump-fed systems, required
turbopump horsepower will decrease. However,
the forces acting upon the thrust chamber, and Ftc pdA pdA Ai pdA
roAr + fAinj
thus the developed thrust, can be assumed to ,]A t •*A t
have remained unchanged, since a lower pres-
sure acts upon the larger injector, and since
opposing forces are present at the converging + pdA (1-38)
e
nozzle. $
In short, it may be stated that our redesign
(fig. 1-8) results in reduced demands on the pro- The last expression in the equation represents
pellant feed system for the same Wtc and the the gain realized from attaching the divergent
same thrust level. nozzle to the throat. By combining all gains
We now proceed to further redesign the cham- into a single coefficient C[ (see eq. 1-33), we
ber to include a divergent nozzle section, as in arrive again at equation (1-34):
figure 1-9. Up to the throat area, nothing
F = (Pc)ns "At" Ct
changes over the preceding configuration, which
includes a convergent nozzle only. Since the
gas velocity in the throat area is always sonic In brief, it may be stated that the redesign
(fig. 1-9) results in an increased thrust level, for
(except for very low, subcritical chamber pres-
sures), the attachment of the divergent nozzle the same Wtc and the same feed system config-
uration.
section, likewise, will have no effect on the
previously described gas processes and the
pressures upstream of the throat. However, con- Summary of the Influences of Pa, e, y, R, and
ditions downstream from the throat are now (Pc)ns on Engine Performance
different.
With the cylindrical chamber, and the chamber The Effect of Pa
with convergent nozzle, the static pressure An ambient pressure Pa reduces the vacuum
energy available at the throat Pt is dissipated by thrust F of an engine by the amount Pa "Ae. (See
eq. (1-6).) C{ is similarly affected by the amount

E. Pa/(Pc)ns, as shown in equation 1-33a). This
may be rewritten as C{=(CDvac-e-Pal(pc)as.
The lower the ambient pressure, the higher
thrust and performance. Maximum values are ob-
tained in vacuum. ___ (Pc)NS Pe>P=
The Effect of e
PO
Optimum thrust for a given ambient pressure

is obtained when the nozzle expansion area ratio
e=Ae/At is such that Pe--Pa. This may be seen OPTIM_
from figure 1-10. If the divergent nozzle section EXPANSION

UNDER - I = OVER
is extended in the region where Pe > Pa, thrust
EXPANSION I EXPANSION
will increase. Where Pe <Pa, however, lengthen-
(VARIABLE NOZZLE LENGHT)
ing of the nozzle will decrease thrust. Hence it
would be beneficial to design the nozzle to yield Figure I-10
Pe = Pa, to reach an optimum value for the thrust
coefficient. The e for this condition is called constant R increases; i.e., the gas molecular
optimum nozzle expansion area ratio. Unfortu- weight decreases. Thus, a higher value of R
nately, because of changing ambient pressure will yield a higher engine performance.
during flight, no one single • is optimum. Opti-
mization studies are usually made to determine The Effect of (Pc)ns
the best compromise. Such a study is not re-
The effective chamber pressure or nozzle
quired (except for weight and size considera-
stagnation pressure (Pc)as appears in equation
tions) for rockets which start and stop at the
(1-33a) for Ct in the form of two pressure ratios
same ambient pressure, such as upper stages,
Pe/(Pc)ns and Pa/(Pc)ns. As is evident from
where ambient pressure is zero or near zero at
equation (1-20), the ratio Pe/(Pc)ns has a singu-
all times. For the special case of Pa = 0 (vacuum
lar value, for a given E and y. (Pc)ns in equation
conditions), e would become infinity, to satisfy
(1-33), therefore, influences C{ only through the
"ideal expansion." Even for this case, however,
negative term -Pal(pc)as. An increase in (Pc)as
expansion ratios over 25 contribute little. The
decreases this negative term and hence in-
nozzle design is usually "cut" at this point,
creases C{. This effect is more pronounced
mainly for weight considerations. This leaves a
when Pa is high. Since the thrust is proportional
small positive pressure at the exit which is un-
to both (Pc)as and Ct, we see now clearly how
available for final gas acceleration. However, it
an increase in (Pc)ns in a given thrust chamber
still contributes as a positive term Pe "Ae.
will increase the thrust.
The Effect of y (Pc)ns also has some effect on the combus-
tion process. Increasing (Pc)as tends to in-
The specific heat ratio is an indication of the
crease (Tc)ns and to reduce y and R. The over-
energy storing capacity of the gas molecule. A
all result is usually an increase in c*. However,
smaller value of y indicates a higher energy-
these effects are slight, especially at (Pc)as
storing capability, and in turn gives higher en-
above 300 psi.
gine performance. As shown in equations (1-32a)
and (1-33a), a smaller y will yield a higher value
for both, c* and C[. The influence of the prop- Correction Factors and Magnitudes of Engine
erties of the selected propellants and of the Performance Parameters
combustion products is apparent.
The actual performance of a liquid propellant
rocket engine differs from that of an ideal one
The Effect of R(R= 1544/_)
because of friction effects, heat transfer, non-
It can be seen from equation (1-32) that for perfect gases, nonaxial flow, nonuniformity of
constant (Tc)as, c* will increase if the gas working substance and of flow distribution, and
TABLE I-3
shifting gas composition. The latter refers to
the fact that the gas properties (y, _, R) are not Gas temperature, T ............... 4000 ° R to 7000 o R
truly constant along the nozzle axis, as the Nozzle stagnation pressure (pc)ns .. 10 to 2500 psia
isentropic treatment of the processes assumes, Molecular weight, E ............. 2 to 30
Gas constant, R .................. 51.5 to 772
Therefore, correction factors have to be applied
Gas flow Math number, M .......... 0 to 4.5
to the performance parameters which are derived Specific heat ratio, y . ........... 1.13 to 1.66
from theoretical assumptions. Following are Nozzle expansion area ratio, e ..... 3.5 to 100
some important correction factors: Nozzle contraction area ratio, _c • .. 1.3 to 6
Thrust coefficient, Ct ........... 1.3 to 2.0
Characteristic velomty, e* ....... 3000 to 8000 R/see
Correction factor for thrust and thrust coefficient
Effective exhaust velocity, c ...... 4000 to 12000 ft/sec
Specific impulse, Is .............. 150 to 480 sec
Actual thrust fvaeuum)
coefficient _ Actual thrust
(1-39)
_?t= Ideal thrust Ideal thrust
Values of the vacuum or altitude thrust coeffi-
coefficient
cient (Ct)vac plotted as functions of nozzle ex-
pansion area ratio e and gas specific heat ratio y
The values for _f range from 0.92 to 1.00. are shown in figure 1-11.
Correction factor for effective exhaust velocity

and specific impulse
Assume a thrust chamber for the same ideal
Actual effective Actual specific liquid propellant rocket engine as given in sam-
exhaust velocity impulse ple calculation (1-2), in which Wtc = 360.71b/sec;
(1-40)
_?v- Ideal effective - Ideal specific (Pc)ns = 1000 psia; (Tc)ns =6540 ° R; 111_= 22.67;
exhaust velocity impulse y= 1.20; e= 12.
Determine the following: (a) Theoretical c*;
The values for _?vrange from 0.85 to 0.98. (b) theoretical Cf at sea level and in space;
(c.) theoretical (ls)tc at sea level and in space;
Correction factor for characteristic velocity (d) actual c*, if c* correction factor ?v*= 0.97;
(e._)actual Cr at sea level and in space, if sea
Actual characteristic velocity level Ct correction factor: W= 0.983; (f) actual
(1-41)
_v*- Ideal characteristic velocity (ls)tc at sea level and in space; (g)
m (Is)tc cor-
rection factor at sea level; (h) thrust at sea
The values for _?v* range from 0.87 to 1.03. level and in space; (i._.)actual At and Ae.
Correction factor for propella_t mass flow rate Solution
(a) From equation (1-32a):

Actual propellant mass flow rate
(1-42)
_?w= Ideal propellant mass flow rate
Theoretical c* :- v'gyR(Tc)ns
The values for ?w range from 0.98 to 1.15. [ Y+_"
,tI{ 2 \k-_
The relation between correction factors may be
expressed as:
_/3 2.2 x 1.2 x _1544 x 6540

yv = _v*'W (1-43)
0.7104
_v = 1/rlw (1-44)
= 5830 ft/sec
Actual ranges of liquid propellant rocket engine
parameters are listed in table 1-3. (b) From equation (1-33a):
2.06
2.00 = LIO"
1.94
1.88
_1.82
i-. I.76
z
tJ.i
G 1.70
LI-
_ 1.64
1.58
n,,,
p-_ 1.52
hi
a
F-_ 1.46
g
,_ 1.4o
1.54
1.28
1.22
I Z 3 4 5 6 7 8 9 I0 20 _,0 40 50 60 SO I00
NOZZLE EXPANSION AREA RATIO,( =As/At
Figure 1-11.-Altitude thrust coefficient as function of area ratio and specific heat ratio.
_/ 2y
y__
z ×+__!I Y-__L = 1.65 - 0.0582
Theoretical Ct: _ Pe = 1.5918
In space:
Pe - Pa
+e(pc)ns
14 7
Theoretical Ct= 1.5918+12× _ = 1.5918+0.1764
From sample calculation (1-2)
Pe = 9.85 psia:
= 1.7682
At sea level Pa = 14.7 psia:

(c) From equation (1-31):
Theoretical Ct=2.247× c* • C!
T- \ Theoretical Is --
g
9.85 - 14.7
+12×
1000 At sea level:
4.85 x 12
= 2.247 x _/1 - 0.4625 Theoretical (Is)to -- 5830 × 1.5918: 288.4 lb sec/lb
1000 32.2
= 2.247 x 0.734- 0.0582 In space:

5830 × 1.7682 = 319.6 lb sec/'lb Thrust F in space : lt'tc • (Is)to in space

Theoretical (ls)tc- 32.2
= 360.7 × 306 = 108 500 lb
(d) From equation (1-41):

(i) From equation (1-33):
Actual c* =_Tv*"theoretical c*
F = 99200 -63.4 in 2
At =Cf.(Pc)n s 1.566x1000
= 0.97 x 5830= 5650 ft/sec
Ae=e'At=12×63.4=760.8 in 2
(e) From equation (1-39):
Actual Cr= W" theoretical Cf

1.4 LIQUID ROCKET PROPELLANTS
At sea level: The term "liquidpropellant"is used to de-

fineboth liquidoxidizers(liquidoxygen, liquid
Actual Cf=0.983 × 1.5918 = 1.566 fluorine,nitricacid, etc.)and liquidfuels (RP-1,
alcohol,liquidhydrogen,etc.). In some cases
In space: additivesare used (water,ferricchloride,etc.).
The propellantsfurnishthe energy and the work-
14.7 ing substance forthe rocketengines. The
Actual C[= 1.566 + 12 x --
1000 selectionof the propellantsis one of the most
importantsteps in the design of an engine. It
= 1 •566 + 0.1764 = 1.7424
greatlyaffectsoverallengine system perform-
ance as well as the design criteriaforeach
(f) At sea level:
engine component. The propellantselectionin
turnis influencedby price,supply,handling,and
5650 x 1.566
Actual (/s)tc = -275 lb sec/lb storageconsiderations.
32.2
In space:
Monopropellants
5650 x 1.7424
=306 lb sec/lb Liquid monopropellants may be either a mix-
Actual (ls)tc- 32.2
ture of oxidizer and combustible matter, or a
single compound which can be decomposed with
(g) From equation (1-40):
attendant heat release and gasification. A
rocket monopropellant must be stable in a natural
(Is)tc correction factor Actual (ls)tc at sealevel
or controlled environment, yet should produce hot
_v at sea level Theoretical (/s)tc at
combustion or decomposition gases when pres-
sea level
surized, heated, or fed through a catalyst. A
liquid monopropellant engine system usually
275
does have the advantage of simplicity of tank-
= 288.----_-0.954
age, feed plumbing, flow control, and injection.
Unfortunately, most of the practical monopropel-
Or from equation (1-43):
lants, such as hydrogen peroxide (H202), have a
relatively low performance. Thus, they are
_Tv= _?v*" W =0.97 × 0•983 = 0.954
mainly used as secondary power sources in
rocket engine systems, such as for turbopump
(h) From equation (1-31):
gas generators and auxiliary power drives, and
for attitude and roll control jets. Certain high-
Thrust F at sea level = Wtc • (ls)tc at sea level
performance monopropellants, such as methyl
= 360.7 × 275 = 99 200 lb nitrate (CH3NOs), are rather unstable and are
considered unsafe for rocket applications. How- under development which should greatly reduce
ever, some monopropellants promising relatively these losses. Adequate venting systems are
high-performance and safer operational charac- needed for the developed gases. Storage and
teristics have been under development recently. handling equipment and their components are
If successful, these may effect wider application extremely sensitive to atmospheric or other
of liquid monopropellant engines. moisture; even minute quantities may cause a
jamming of, for instance, a valve. Likewise, the
design criteria, including materials selection for
Bipropellants
engine systems using cryogenic propellants,
In a liquid bipropellant system, two different must consider the very low temperatures in-
propellants are used, usually an oxidizer and a volved. The mechanical design of engine com-
fuel. Separate tanks hold oxidizer and fuel ponents for cryogenic propellant applications
which are not mixed until they reach the combus- will be discussed in subsequent chapters.
tion chamber. Present-day liquid propellant
rocket engines use bipropellants almost exclu-
StorableLiquid Propellants
sively because they offer higher performance,
combined with safer operation. In contrastto the cryogenic propellants,cer-
The combustion of many bipropellant com- tainother liquidpropellantsare stable over a
binations is initiated by ignition devices such reasonable range of temperatureand pressure,
as: (a) chemical pyrotechnic igniters, _ elec- and are sufficiently nonreactivewith construc-
tric spark plugs, (_ injection of a spontaneously tionmaterialsto permit storage in closed con-
ignitable liquid fuel or oxidizer ("pyrophoric tainersforperiods of a year or more. These
fluid") ahead of the propellant proper, (d) a small propellantsare defined as storables. Storable
combustor wherein ignition is started by devices liquidpropellantspermit almost instantreadiness
(a) or (b), in turn starting the main chamber by of the rocket engine and may result in greater
the hot gas produced. reliability due to the absence of extremely low
Other bipropellant combinations ignite spon- temperatures and the need to dispose of boiloff
taneously upon mixing. Those combinations are vapors. Their applic_ttion to military vehicles
defined as hypergolics and permit greatly simpli- as well as to the upper stages of space vehicles
fied ignition, but pose certain hazards. For has increased significantly during recent years.
instance, accidental mixing of the fuel and oxi- The mechanical design of storable liquid engine
dizer due to tank and other hardware failures components will be further discussed in subse-
could cause a violent explosion. These hazards quent chapters.
must be considered when designing an engine
system using hypergolic propellants.
Additives for Liquid Rocket Propellants
Cryogenic Propellants
Sometimes, additivesare mixed intoliquid
Some liquid propellants are liquefied gases propellants for one of the following reasons: (a)
with a very low boiling point (-230 ° F to -430°F) to improve cooling characteristics; (b_) to de-
at ambient pressure and a low critical temper- press freezing point; (c_) to reduce corrosive
ature (10 ° F to -400 ° F). These propellants are effects; (d._)to facilitate ignition; and (.e.) to
defined as cryogenics. The most common cryo- stabilize combustion.
genic propellants for rocket applications are
liquid oxygen (02), liquid hydrogen (H2), liquid Optimum Mixture Ratio
fluorine (F 2), and oxygen difluoride (OF 2), or
mixtures of some of them. Cryogenic propellants A certain ratio of oxidizer weight to fuel
pose storage and handling problems. Elaborate weight in a bipropellant combustion chamber will
insulation must be provided in order to minimize usually yield a maximum performance value.
losses due to boiloff, the complexity depending This is defined as the optimum mixture ratio. As
on storage period and type of cryogenic. Re- a rule, the optimum mixture ratio is richer in fuel
cently, novel insulating techniques have been than the stoichiometric mixture ratio, at which
2O DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
theoretically all the fuel is completely oxidized tion gases, for high specific impulse.
and the flame temperature is at a maximum. This (2) Ease of ignition.
is because a gas wt_ich is slightly richer in fuel (3) Stable combustion.
tends to have a lower molecular weight. This (4) High density or high density impulse to
results in a higher overall engine systems per- minimize the size and weight of propel-
formance. The optimum mixture ratio of some lant tanks and feed system.
propellant combinations shifts slightly with (5) Ability to serve as an effective coolant
changes in chamber pressure. Also, in actual for the thrust chamber (optimum com-
application the mixture ratio may be shifted away bination of high specific heat, high
from the optimum value for one of the following thermal conductivity and high critical
reasons: (a) lower chamber temperature to stay temperature).
within the temperature limitations of chamber (6) Reasonably low vapor pressure at 160 ° F
construction material; (.b.) required coolant flow; (a frequent specification value) for low
(c) improved combustion stability. tank weight and low net positive pump
suction head requirement.
(7) Low freezing point (preferably less than
Density Impulse
-65 ° F') to facilitate engine operation at
In addition to the overall system-oriented low temperature.
specific impulse which we thoroughly discussed (8) Absence of corrosive effects; compatibil-
in paragraph 1-3, a quantity called "density im- ity with engine construction materials.
pulse" is an important propellant performance (9) For storables: good storability as as-
parameter. It is the expression for the total sisted by a high boiling point (prefer-
impulse delivered per unit volume of the propel- ably above 160 ° F), by items 6, 7, 8
lant. It is defined as: and by the resistance to deterioration
during storage.
Density impulse = Is • d (sec) (1-45) (10) Low viscosity (preferably less than 10 cp
down to -65'= F) to minimize pressure
wherein drops through feed system and injector.
d :bulk density or propellant combination, (11) High thermal and shock stability to mini-
spec. weight mize explosion and fire hazard.
(rw -1) (12) Low toxicity of raw propellants, their
d - (1-46) fumes, and their combustion products.
rw 1
do+dr (13) Low cost.
(14) Availability.
rw = (oxidizer/fuel) weight mixture ratio
d o : bulk density of the oxidizer, spec. weight
di= bulk density of the fuel, spec. weight
Liquid Rocket Propellant Performance and
Physical Properties
The Selection of Liquid Rocket Propellants
Detailed methods to calculate the performance
When selecting a propellant or propellant for any given liquid propellant or propellant com-
combination for a specific application, it is well bination can be found in the standard combustion
to realize that most propellants, in addition to engineering or rocket propellant textbooks. For
their advantages, may have certain disadvan- the theoretical calculations, it is generally as-
tages. Thus, propellant selection usually in- sumed that the ideal conditions exist as de-
cludes some compromises. The more important scribed in section 1.2 (Gas Flow Processes) of
and desirable propellant features are listed this chapter. The prime objective of propellant-
below. Order of importance may vary as a func- performance calculations is to derive the quan-
tion of application. tities c*, Ct, and Is through evaluation of the
(1) High energy release per unit of propellant flame or chamber temperature (Tc)ns; of the gas
mass, combined with low molecular mean molecular weight _Ii; and of the specific
weight of the combustion or decomposi- heat ratio y for a given (Pc)ns, Pe and Pa. The
INTRODUCTIONTO LIQUID PROPELLANT ROCKET ENGINES 21
chamber temperature can be calculated from the the theoreticaldata based on a shiftingcomposi-
heat of the chemical reaction of the propellants tionusually give values several percenthigher
and from the specific heat of the gases. In than those based on a frozen one. Therefore, in
practice it has been found that actual test re- presentingperformance data, the assumption of
sults are usually 5 to 12 percent lower than the the type of composition assumed must be speci-
theoretical values obtained from calculations. fied. As a rule,the thrustchamber designer will
In addition to the assumption of certain ideal- be supplied with the basic parameters by depart-
ized gas conditions, the performance equations ments specializingin thisfield. We need not,
discussed assumed and employed certain singular therefore,concern ourselves furtherwith this
values for the most important gas properties: matter.
y, _, R, (Tc)ns. For basic design information Performance and physicalpropertiesofnumer-
requiring greater accuracy, more rigorous calcu- ous importantliquidmonopropellantsand bipro-
lations frequently employing electronic comput- pellantsare given in tables 1-4 through 1-10.
ers are usually conducted by specialists in the
field. These consider that the gas properties are
not necessarily constant along the path of flow.
1.5 THE BASIC ELEMENTS OF A LIQUID
Two basic approaches can be taken: Calcula-
PROPELLANT ROCKET ENGINE SYSTEM
tions based on the assumption of unchanging or
"frozen" gas composition along the nozzle axis, A vehicle system has occasionallybeen de-
or based on the assumption of shifting composi- fined as a purposefulconglomerationof subsys-
tion. The applicable literature frequently uses tems. One of these is the engine system. The
the term "equilibrium" instead of "composition." definition of the scope of the various vehicle
In calculations based on frozen composition, subsystems has not always been uniform and
it is assumed that no further chemical reactions probably,by necessity,never willbe. For in-
take place in the gases after leaving the com- stance, forvehicle systems in which the propel-
bustion chamber and entering the nozzle, and lanttanks simultaneouslyserve as the vehicle
that the combustion products at Ae are in the airframe,itmay be a matter of opinion whether
same relative proportion as they were at Ai. The they are part of the structure or of the engine
remaining principal variables then are pressure system. The decision to which subsystem they
and temperature at the various stations. Assum- belong may well depend on the fact whether the
ing different initial sets of mixture ratios, cham- tanks will be supplied by the engine manufac-
ber pressures, and gas compositions, a typical turer, or by a separate contractor. Similarly,
set of calculations, probably involving succes- some, notably the engine system supplier, may
sive approximations, may be conducted to deter- consider the guidance system a part of the pay-
mine the optimum values of, for instance mixture load, while the vehicle user will hold that any-
ratio, chamber length, expansion area ratio, and thing without which the vehicle cannot fly
nozzle contour, for a given propellant combina- reliably and accurately to its destination is not
tion and vehicle trajectory. payload. Whatever the definitions may be, it is
Calculations based on shifting composition important that they are used uniformly and con-
take into account additional variations, mainly sistently in a given project.
those of gas composition, as they result from, For the purpose of this book, we will define a
for instance, incomplete combustion, dissocia- vehicle as being composed of the following major
tion, and reassociation. These calculations are subsystems:
an attempt to consider more nearly the true (1) Engine system
physical processes. Due to their extreme com- (2) Vehicle structure
plexity and unpredictability, however, the results (3) Guidance sys.tem
are frequently no more reliable predictions of (4) Payload
test results than those obtained from calcula- (5) Accessories
tions assuming frozen composition. In the following, we will concern ourselves
Thus, it is probably a matter of preference with the engine system only, except for brief
which approach should be taken. It is noted that references to the other systems, as required. We
o_
L_.
c_
..4
_._. _._., _ _
0 0 _
8 8 o o _
i
_ :._ ._ _
E- ;>
¢)
-=
8 8 O
I
L_r.
r. r..
O
b-
o0
O
c_
Q
o ._
-_._
<
r. 0 0 _" r.
A
:E
z
±
r. z o z
L_ _ _ L_ z o
• _
.. -,'- 0
=_
o
o
LO
¢kl ,:5 ¢5
o . N
G_
;>
= ,-;
E ._ = r.,..
• O
¢3
,_r,,
,-..1 ,..4
¢...
r._
o
¢3
0,3
o_ o
c_
o
I
f == =_. ._ ._ .
..'E "-- .O ""_ .O
r.=.. _ r,,. _ r,..
r. r,,, r,,,
,-.1
o_ "= _ z _ _ z
z
r,. o"
co
e._ - _ _-_ _ -,
o
0=
O
° _._. _= _;_ _; ._ _
-_=
_=
- _. _ _'_ _
:E
=¢
_r
E_
.-_ ¢5 _ _ ..4
;=
,-_ .-
r.
I
.4
E_
•_ _ ....
r_ _ _ c_ o c7
-_ _ .= o
.o
O.
=
TABLE 1-6.-Perlormance of Some Liquid Rocket Monopropellants
Specific Density
Propellant impulse Is, impulse Id, Applications Remarks
lb-sec/lb a sec gm/cc
Hydrogen peroxide (H=O2) (95%) . 140 198 Gas generators for turbopump Difficult handling
and auxiliary drive; small
control rockets
Hydrazine (N2H ,) .............. 205 207 Gas generators: small control Difficult handling (can de-
rockets compose at high temper-
ature)
Nitromethane (CH3NO =) ......... 180 204.8 Small ordnance rockets Dangerous handling (can
detonate unexpectedly)
Methylacetylene ............... 160 108.6 Gas generators; small rockets Safe handling; dangerous and
very smoky exhaust fumes
a Theoretical value at 300 psia (Pc)ns, sea-level optimum expansion, frozen gas composition or frozen equilibrium.
TABLE 1-7.-Theoretical Performance of Some Medium-Energy Storable Liquid Rocket Bipropellant

Combinations
Oxidizer Fuel rw rv d Tc }_ c* Ct Is ilsd Applications
IRFNA (15% NO 2) . UDMH .................... 2.99 1.51 1.26 !5340 23.7 5490 1619 276 348 Small air-to-air,
3.24 1.63 1.27 !5315 24.2 5435 1630 275 350 air-to-surface
Hydrazine ................ 1.47 95 1.28 5090 20.8 5690 1.602 283 362 rockets and
1.54 ,99 1.29 5100 211 5665 1.608 283 365 upper stages of
space vehicles
50% UDMH-50% hydrazine... 2.20 1.26 1.27 5250 22.4 5580 1.610 279 354
242 1.39 1 29 5220 23.0 5510 1.618 277 358
Hydyne ................... 3.11 1,70 1.31 5295 24,1 5425 1.620 273 358
3.33 1.82 1.32 5270 24.5 5375 1.630 272 359
268 ]362
RP-1 ...................... 4.80 2A8 1.35 5355 25.8 5275 II.636
i
5.14 2.65 136 5330 262 5225 !1.646 267 363
TMB-1, 3-D ............... 4,09 2.08 1.32!5325 25.1 5335 1.632 270 356
4.37 2.23 1.33 5300 25.5 5280 1.640 269 358
JP-X (60% JP-4, 40% UDMH) 4.13 2.16 1.33,5310 246 5320 1628 269 358
2,89 1.47 1.26!4935 .... 5130_1.626 259 326

92.5% E.A ................
MMH ..................... 2.47 138 1.28 5290 .... 5550:1.618 279 357
TMA ..................... 4.01 1,61 1.21 5285 .... 5375 1625 271 328
95% hydrogen UDMH .................... 4.54 2,53 1.24 4800 217 5530 ;1.620 278 345 Manned aircraft,
peroxide 4.74 2.64 1.25 4780 21,3 5505 !1.620 277 346 small air-to-air.
Hydrazine ................ 2.17 1.54 1.26 4675 19.5 5655 11.604 282 355 air-to-surface
2,20 1,57 1.26 4675 19.5 5655 1.604 282 '355 rockets, and
upper stages of
50% UDMH-50% Hydrazine .. 3.35 212 1.25 4760 20.5 5580:1.610 279 349 space vehicles
3,47 2,20 1,26 4740 20,6 5560 1.615 279 351
Hydyne ................... 4,68 2.83 1.27 4765 21,3 5485 1.'622 276 350
4.87 2.95 1.28 4745 21.4 5465 '1.619 275 352
I
RP-1 ..................... 7,35 4.18 1.30 4785 22,1 5405 1.627 273:355
7.58 432 1.31 4765 22.2 5390 1.620 271 355
J
TMB-1, 3-D ............... 6,20q3.49 1.28 4770 218 544011622 274 351
6,45/3.63 1.29 4745 21.9 5415 1.618 272 351
i
TABLE 1-7.-Theoretical Performance of Some Medium-Energy Storable Liquid Rocket Bipropellant
Combinations (Continued)
Oxidizer Fuel I rye IV

, T.l,lc.lc, 1 Is , I sfl Applications
Nitrogen tetroxide. UDMH ................... I 2.95 1.61 1.20 5685 24.5 55551632 282 339 Manned aircraft,
I
2.71 ] 1.61 1.22 5650 124.1 55801 626 ! 282 344 ICBM, IRBM,
Hydyne .................. I
1.24 5655 !24,7 5525 1631 280 347 ALBM. smallair-
295 1.75 to air, surface-
4.04 2.26 1.25 5745 25.7 544011 636 276 345 to-air rockets,
' RP-1 ....................
4.50 2.51 1.27 5755265 538511639 274 _348 upper stages of
J 3.55 1 23 5715 25.2 5495 1631
T_IB-I, 3-D .......... i 196 278 342 space vehicles
3.90 2.15 1.24 5710 25.9 5425 1.645
J 277 344
2.59 1.45 1.19 5290 .... 5260 1.635 267 rr
318
% E.A .......... !
Chlorine trifluoride t UDMH .................... 303 131 1.38 6305 258 5630 1.602 280 386 ICBM, IRBM
I 328 1.42 1 40 6330 26.2 5605 1.589 277 388 I ALBM. and small
Hydyne ................... 298 1 40 1.43 6220 26.1 5555 1,599

320 150 144 6250 265 5535 1.595 274 395 rockets, upper
stages of space
RP-1 .................. 1 41 5890 29.1 5140 1.618 276 395iL air-launched
320 1 42 258 364 vehicles
230 386
12.80 566 168 5735 370 4535 1636
TMB-I, 3-D .............. 3 17 1.39 1.40 6035 27.6 5330 1608 266 373
3.60 1.57 261 374
1 43 6040 28,1 5280 i1 592 _

Bromine Hydrazine .............. 3.35 1.37 186 5570 243 453 Small air-launchec
.... 5000 1 565
pentafluoride rockets
TABLE 1-8.-Theoretical Performance o[ Some High-Energy Storable Liquid Rocket Bipropellant

Combinations
Oxidizer Fuel rw rv C! Is Isd Applications

Tc] _
c*
95% Hydrogen peroxide Hydrazine ............. 201 1.41 1 26 4775 19.5 I1 601 285 359 ]CBM, IRBM.
1 037 6067 11600 302 313 ALBM

5390 19 01 5735
Pemaborane ............ 270 1 188 i
Nitrogen tetroxide . . UDMH ................ _2.61 1.42 1.18 5685 236 15650 1.624 285 336 !FBM, ICBM.
Hydrazine ............. 1.34 .93 1.22 5390 20.9 5845 1 610 292 357 IRBM, ALBM.
1.42 .99 1 23 5415 213 5815 1605 290 357 upper stages
50% UDMH-50% Hydra_ine 2.00 1.24 1.21 5590 22.6 5725 1.620 288 348 of space
2.15 1 33 1.21 5570 23 0 5665 1 636 288 348 vehicles
MMH ................. 2.16 1 31 1 20 5635 I ..... 5720 1.621 288:346
1 582 t294,444 FBM, ICBM,

2hlorine trifluoride ... Hydrazine .............. 2.77 1.53 1.51 6550 [232 5995
2.94 1.62 1.52 6600'236 59£0 1 572 292:444 IRBM, ALBM
50% UDMH-50% Hydrazine 2.89 1.42 1.45 6385 24.5 5795 1 596 287 416 upper stages
311 1.53 1.46 6420124.9 5770 1.598 286 417 of space
MMH ................... 3.00 1.44 1,44 6400 .... 5763 1,591 285 410 vehicles
Hydrazine ........... Pentaborane ............ 1.4 85 .796 4430 147 6402 1644 327 261 ICBMIRBM
A
TABLE 1-9.-Theoretical Performance of Some High-Energy Cryogenic Liquid Rocket Bipropellant

Combinations
Oxidizer Fuel rw rv d Tc _ c* Ct Is Isd Applications
Liquid oxygen.. RP-1 ..................... 2.00 1.421 0.998 5760 211 5898 1,605 294 293 ICBM, [RBM, large
2.40 1.708 1.012 6100 22.8 5953 1.620 300:303 space-probe and
2.56 1.82 1.02 6150 23.3 5920 1.632 300306 space craft
2.73 1.94 1.03 6200 239 5865 1.642 299 308 boosters
Ammonia ................. 1.30 .78 .88 5055 19.3 5920 1.608 296 260
1.40 .84 .89 5100 198 5865 1.612 294 261
95% E.A .................. 1,73 1.23 .99 5640 24.1 5605 1.648 287 2S4
1.80 1.28 1.00 5675 24.4 5585 1.644 285 285
Hydrazine ................. 90 .80 1.07 5660 19,3 6235 1618 313 335
50% UDMH-50% Hydrazine., 1.30 1.03 1.02 5980 20.6 6160 11,628 312 318
1.37il.08 1.03 5905 20.9 6155 '1.622 310 319
Hydyne .................. 1.73_1.31 1.02 5990 21.81 6035 1,632 306 312
1.80 1.36 102 6030 22.2 6010 1.639 306 312
UDMH .................... 1.65 1.14 .98 6010 21.3 6115 1631 310 30,t
1.83 1.27 .99 6065 22.1 6040 :1.638 307 304
TMB-1.3-D ................ 2.28 ! 1.60 1.01 6100 22.9 5945 11.642 303 308
2.37 1.66 1,01 6120 23.2 5915 1.650 303 306
TABLE 1-10.-Theoretical Performance of Some Very-High-Energy Cryogenic Liquid Rocket Bipro-
pellant Combinations
Oxidizer Fuel r., rv d Tc ,_ c* CI Is Isd Applications
Liquid oxygen .......... Liquid hydrogen ..... 4.02 0.25 0.28 4935 10.0 7980 1.578 391 109 Space probe and
19.50 1.20 .65 4960 23,4 5300 1.610 265 172 space craft upper
stage and booster
Liquid fluorine .......... Hydrazine .......... 2.30 1.54 1.31 7955 19,4'7245 1615 363 476 Space probe upper
2.40 1.61 1.32 7980 19,6 7225 1.614 362 478 stage
Liquid hydrogen ..... 7.60 .35 .45 6505 11.818365 1.578 410 185
23.70 1.10 .82 8230 18.5 7515 1.592 372 305
Ammonia ........... 3.29 1.48 1.18 7715 19,3 7155 1.605 357 421
3.40 1.53 1.18 7745 19.5 7140 1612 357 422
NOTES FOR TABLES I-7 THROUGH I-i0
(1) Conditions upon which the performance calculations tw = Propellant weight mixture ratio (wt. oxidizer/wt.
are based = fuel)
(a) Combustion chamber pressure = 1000 psia tv = Propellant volume mixture ratio (vol. oxidizer/vol.
fuel)
(b) Nozzle exit pressure = ambient pressure = 14,7 psia
(optimum nozzle expansion ratio at sea-level oper- d = Bulk density of propellant combination (gm/cc).
ation) (The density at boiling point was used for those
oxidizers or foels which boil below 68 ° F at one
(c) Chamber contraction ratio (chamber area/nozzle
atmosphere pressure)
throat area) = infinity
(d} Adiabatic combustion Tc =Theoretical chamber temperature, _F
(e) Isentropic expansion of ideal gas with shifting = Average molecular weight of combustion products
composition or shifting equilibrium in the nozzle at Tc
(2) Symbols: c* =Theoretical characteristic velocity (ft/sec)

NOTES FOR TABLES i-7 THROUGH I-i0 (Continued)
C[ :Theoretica] thrust coefficient 900..................................... 99

800 ................................. 98
Is : Theoretical maximum specific impulse, lb-sec/lb
700................................ 97
lsd= Theoretical maximumdensity impulse, sec-gm/cc
600 .................................. 95
(3) To approximate Is and lsd at other chamber pressures. 500 ................................. 93
Pressure (psia): Multiply by- 400 ............................... 91
I000 ............................... i00 300................................ 88
further define that the engine system shall com- pulse control, propellant utilization control
prise all parts without which the propulsive (sometimes called propellant management), stor-
force cannot be generated. Thus, we will in- ability, ease of handling, etc. Thus, modern
clude the propellant tanks and their accessories. rocket engines contain more subsystems than
A system thus defined frequently is called a their basic principle of operation may suggest,
propulsion system. We know, from the above, to meet the often stringent vehicle requirements.
that by including the tanks, we may be "infring- This is true for both liquid as well as solid
ing" on the vehicle structure by other defini- propellant systems. In general however, the
tions. liquid propellant engine is the more flexible one,
Thus prepared, we may now proceed to sub- particularly where large systems are considered.
divide the engine system further into major
components or subassemblies as follows:
(I) Thrust chamber assembly
(2) Propellant feed system: One of the fol-
lowing two is generally used: Pres-
surized gas propellant feed system and
turbopump propellant feed system. The
latter includes some type of tank pres-
surization system
(3) Valves and control systems
(4) Propellant tankage
(5) Interconnect components and mounts
Depending on the engine system selected, one A Check valve K High-pressure
B Pressurizing gas helium bottle
or another subsystem may not be required or may
be integrated with another one. Typical liquid diffuser L Pressure regulator
propellant rocker engine systems are shown in C Fuel tank M Heat exchanger
figures 1-12 and 1-13. D Pressurizing gas N Fuel tank vent and
diffuser relief valve
The rocket has occasionally been called the E Pressurizing gas O Oxidizer tank vent
simplest propulsion system known. The simplest line and relief valve
form of a solid propellant rocket or of a pressur- F Check valve P Oxidizer tank fill
ized gas-fed storable liquid propellant rocket G Oxidizer tank and drain valve
appears to come close to this ideal. Unfortu- H Fuel duct Q Oxidizer duct
nately, simplicity frequently is synonymous with I Fuel tank fill and R Main oxidizer valve
inflexibility. Due to vehicle requirements, sub- drain valve S Thrust chamber
stantial departures from the basic simplicity may J Main fuel valve assembly
become necessary to meet requirements such as:
light weight, high performance, thrust control, Figure 1-12.-Typical pressurized gas feed liquid
thrust vector control, restartability, cutoff ira- propellant rocket engine system.
A Check valve L Pressure regulator cryogenic propellant combi-

B Fuel tank M Heat exchanger nation)
C Check valve N Turbine exhaust duct U Fuel tank fill and drain
D Pressurizing gas line O Thrust chamber assembly valve
E Oxidizer tank P Fuel tank vent and relief V Oxidizer tank fill and drain
lr Fuel duct valve valve
G High pressure helium bottle Q Pressurizing gas diffuser W Oxidizer duct
H Gas generator and valve R Pressurizing gas diffuser X Oxidizer pump
assembly S Oxidizer tank vent and relief Y Fuel pump
I Turbine starting spinner valve Z Gear box
J Gas turbine T Inter-tank insulation (re- AA Main oxidizer valve
K Main fuel valve quired for cryogenic and non-
Figure 1-13.-Typical turbopump feed liquid propellant rocket engine system.

Chapter II
Rocket Engine Design Implements

2,1 THE MAJOR ROCKET ENGINE DESIGN results from the decision whether a single- or a
PARAMETERS multiple-engine system is to be used. This de-
cision is often strongly influenced by the avail-
To fit the engine system properly into a
ability of already existing engines, which would
vehicle system, engine systems design and de-
velopment specifications will have to cover the eliminate, or at least drastically reduce, the
following parameters above all: design and development cost for the propulsion
(1) Thrust level system. The selection of individual engine
thrust level also is-or at least should be-
(2) Performance (specific impulse)
(3) Run duration influenced by the general state of the art, par-
(4) Propellant mixture ratio ticularly if sizes substantially larger than pre-
viously developed are considered.
(5) Weight of engine system at burnout
More recently, largely as a result of the
(6) Envelope (size)
advent of manned rocket flight and of the high
(7) Reliability
(8) Cost cost of very large vehicle systems, the decision
(9) Availability (time table-schedule) to use a multiple (clustered) propulsion system
consisting of several engines rather than a single
As the design progresses, numerous addi-
tional parameters will have to be considered. one has been additionally affected by safety
considerations, to permit mission completion, or
Before turning to the latter, let us briefly review
at least safe return of the crew, in case of an
and discuss those listed above. It should be
engine failure. This "engine out" principle is
noted that the last five items are closely inter-
analogous to the consideration of multiple-
dependent. For instance, making an engine
versus single-engine airplanes. Extensive stud-
available in the shortest possible time ("crash
ies have been conducted in this field for rocket
program _) will raise the cost and will unfavor-
ably affect reliability. A longer design and vehicles to establish the "break-even" point
regarding the minimum and maximum number of
development period may not necessarily reduce
engines profitably employed in a cluster. Fail-
cost, but it will offer higher values in exchange
ure of single-engined rocket vehicles not only
for the dollar; higher reliability, refined (lower)
might destroy the vehicles themselves but also
weight, and an optimized (smaller) envelope.
could cause severe damage to expensive ground
facilities. This explains the great emphasis
Thrust Level placed on thrust subdivision.
This engine parameter is a basic one, similar Thrust levels for first-stage booster engines,
to the power rating of a gasoline engine or elec- which start at or near sea-level altitude and stop
tric motor. It will affect most of the other engine at a specified higher altitude, are usually quoted
parameters and many of the development con- for sea-level conditions. Additionally, the spec-
siderations. ifications may contain information on thrust level
The total thrust requirement of a rocket- at altitudes above sea level, frequently in the
propelled vehicle is predominantly governed by- form of a graph (see fig. 2-1).
1. The total takeoff weight of the vehicle The nominal thrust of engines in stages start-
(including engine!) ing and operating at or near-vacuum conditions
2. Minimum and maximum accelerations per- is quoted for that environment. Most engines are
missible designed for a single nominal thrust (sea level
Selection of the proper engine thrust level or altitude), for which they are calibrated by
31
I SlEC its true significance. In June 1959, Dr. von

S_C_FE q_PUL_ m CUTO=r
,ooT Kgrmgn observed:

]t is my personal belief that the length of the
zr_* F_ period of attaining reasonable reliability in the devel-
IZSO'
opment process could be essentially reduced if simple
1
design were emphasized as a leading principle, even
if we had to make some sacrifice in the quantitative
measure of "efficiency." Essential elements have to
be designed as simply as possible, even if this means
¢L_Crr a reduction in quantitative efficiency and a certain
increase of bulkiness andsor weight.
Undoubtedly, these observations were
so s¢ go L2o _r_ prompted by a noticeable trend on the part of

ALTITUDE (FT X I0 $) both the engine builder as well as the customer,
to sacrifice, or at least to compromise, nearly
Figure 2-I.-Typical graph o[ rocket engine per-
all other capabilities of a rocket propulsion sys-
lormance as function of altitude.
tem for Is increases, which sometimes amounted
to less than 1 percent.
means of propellant line orifices or, less fre-
Frequently, increasing emphasis on Is during
quently, with the aid of regulators. Engines
the life of a project can be traced to marginal
designed for variable thrust (throttling) always
engineering reserves in the initial vehicle design
require some type of regulator. This will be
especially with weight assumptions and tank
discussed in section 7.3, "Engine Thrust Level
Control." capacities. The need for competitive bidding
may have contributed to this situation.
On the other hand, the highest Is which can

Performance
be obtained without compromise will pay off sub-
Although the general term _performance" of a stantially. For instance, in the case of a typical
rocket engine in the strict sense covers a num- medium-range ballistic missile, an increase of
ber of parameters (Is, c*, Cf, etc.), specific 1 second in Is will effect a range increase of
impulse (Is) is considered the prime performance approximately 15 nautical miles. In other terms,
parameter• As was seen in chapter I, the spe- an Is increase of less than one-half percent
cific impulse, also referred to as specific thrust, results in a range increase of I percent. As
is measured in seconds, which obviously is not impressive as these figures for increased flight
the dimension of time, but an abbreviation of the range are, it should be kept in mind that those
dimension lb-sec/lb (specific impulse), lb/(lb/ engine properties which will determine whether
sec) (specific thrust), respectively. It is impor- the vehicle will fly at all should not be corn-
tant to state whether a specified value of Is
refers to the complete engine system, or to the
Duration
thrust chamber only. Frequently, by stating a
percentage an "actual" or "practical" value of Is Because, by definition, a rocket vehicle
is linked to the maximum value theoretically carries its own complete propellant supply, in-
possible. The theoretical values for the better cluding the oxidizer, its run duration is limited,
known propellant combinations are well estab- as a result of an optimized balance between
lished and as a result practical values have takeoff weight, trajectory, thrust level, and mini-
become quite predictable. With less well-known mum and maximum accelerations. Consequently,
combinations, disappointments have often re- run-duration times of most large liquid-propellant
sulted. Therefore, great caution is advisable in rocket engines fall into a relatively narrow band,
the use of theoretical values which have not about 50 to 400 seconds.
been verified in an actual test. User specifications include a formal demon-
In recent years, the performance of a rocket stration (such as preliminary flight rating tests
engine, as expressed by its specific impulse, (PFRT) and qualification tests) requiring accu-
has received considerable attention, far beyond mulated duration times, without breakdown, of
ROCKET ENGINE DESIGN IMPLEMENTS 33
many times the comparatively short rated flight This is analogous to a cannon, where muzzle
duration (typical: six full duration tests for exit velocity of the projectile, gun-barrel atti-
PFRT of an ICBM). tude, and location of the gun emplacement will
determine the point of impact (neglecting envi-
These specifications, therefore, govern most
ronmental influences such as wind). With a
engine design considerations, with the exception
ballistic rocket, the gun barrel is literally re-
of the following areas, which for weight consid-
placed by the guidance system, the intricate
erations are tailored to the flight-run duration:
components of which not only predetermine the
(1) Auxiliary tank capacity, for systems
three basic parameters mentioned but also have
which employ a separate turbine power
the capability to compensate for deviations of
supply
any or all of them. If, for instance, the trajec-
(2) Propellant-tank pressurization supply, if
tory angle near the point of cutoff is too steep,
it is part of the engine system
the guidance system will compensate accord-
(3) Lube oil tank capacity, if applicable
ingly, by calling for a higher final velocity, by
(4) Temperature nonequilibria, such as those
slightly delaying the cutoff signal, simultane-
of uncooled nozzles
ously considering the distance over ground
Closely related to the run duration are the already covered.
start and shutdown characteristics of an engine It is obvious that a prompt and repeatable
system, the requirements for both of which may execution of the cutoff signal is imperative.
be very stringent in a given vehicle system. However, for several reasons, it is impossible to
The characteristics and the quality of the effect a truly instantaneous thrust cessation:
"start," or "thrust buildup," of a liquid rocket time is required to sense and then transmit the
engine are judged by- cutoff signal; closing of valves requires a finite
(l) Compliance with specified thrust versus time; structural (hydraulic hammer) considera-
time characteristics tions are superimposed; residual propellants
(2) Maximum rate of increase at any time below the valves have an effect. Figure 2-2
during buildup shows a typical thrust decay diagram.
(3) Freedom from surges and thrust over-
Let us recall:
shoots
(4) Smoothness (freedom from damaging oscil-
Ft = may
lations)
(5) Repeatability from run to run and from
Thrust multiplied by time equals mass times
engine to engine
velocity increase, or
These characteristics will be discussed in
greater detail in chapter X, "Engine Systems Ft
Av=
Design Integration." Suffice it to state, at this m
point that a rocket engine is not easy to adapt
with special thrust buildup requirements. Diffi-
cu_rr ro ca.as[
culties in this area can arise from inadequate

F.Y.
IOO
communication between the vehicle contractor
and the engine contractor. Thorough understand-
ing of the problems by both contractors is vital,
The characteristics of engine "shutdown" or

"thrust decay" are predominantly influenced by
guidance considerations. To understand this
better, let us consider the case of a single-stage,
ground-to-ground, ballistic missile. As the term
"ballistic" implies, the missile is designed to 0
impart a desired speed to a known payload, in a

TIWE F_OI, I CUTOFF S_ONAL {SEC )
desired direction from a desired point, after
which the payload coasts freely to the target. Figure 2-2.-Typical thrust decay diagram.
The velocity increase following cutoff signal amount of oxidizer. That mixture ratio which
is a function of the residual thrust acting on the effects complete combustion, with no leftover of
vehicle mass m, and is integrated over the time either fuel or oxidizer, is called the stoichio-
from cutoff signal to final thrust cessation; this metric mixture ratio. This ratio depends on the
integral is commonly referred to as the "cutoff type of propellants used. Tbeoretical temper-
impulse." A typical value for a well-known ature and heat release are maximum at this ratio.
earlier rocket (Redstone) was 16 000 lb-sec In rocket engines, however, where the highest
z2500. Note the tolerance. This deviation will possible exhaust velocity is desired, optimum
obviously influence missile accuracy. Reduction conditions often prevail at other than stoichio-
of the tolerance is thus an important design and metric ratios. Equation 1-18 indicates that the
development goal. gas properties strongly affect exhaust velocity.
It might be concluded that a substantial re- The expression for the specific gas constant, R,
duction of the tolerance is the principal task, in equation 1-18 may be rewritten as-
zero deviation being the optimum. This is un-
fortunately not so because the final vehicle mass
m, on which the decaying thrust force acts, is
unpredictable within certain limits, due to weigh-
ing tolerances of the initial vehicle mass, and to where R' is the universal gas constant and is
flow rate and mixture ratio tolerances. The the molecular weight of the gas (see table 1-1).
engine designer and developer will have to con- The lower tile molecular weight, the higher
centrate on reducing both: base value and the exhaust velocity, other things being equal.
tolerance. Analytical and experimental investigations will
A glance at figure 2-2 shows that the area determine the optimum point of balance between
under the thrust curve is a function of not only energy release (heat) and composition (molecular
decay time but also of main-stage thrust level. weight) of the gas, a portion of which will con-
In fact, the major portion of the shaded area is sist of gasified but unburnt propellants. The
accumulated prior to the beginning of thrust optimum point may also be affected by-
decay. This observation has led to the utiliza- (l) Stay time of the burning gas in the com-
tion of vernier thrust systems. bustion chamber.-Stay time is a function
A vernier cutoff system is characterized by a of combustion chamber volume and of
substantial thrust reduction before final cutoff. gas volumetric flow rate. Complete com-
This can be accomplished by thrust reduction, bustion, even though desirable, requires
for a few seconds, of the main engine itself (V-2 a finite time which is not available un-
fashion) or by shutdown of the main engines, less the chamber is relatively large, and
while much smaller engines continue for a brief correspondingly heavy. A compromise
period (typical: 0-25 seconds, depending on in chamber size, therefore, is often
final __v required). made. This leaves unburned a small
It should be emphasized that any components percentage even of those propellants
that must be added to improve cutoff character- entering the nozzle, which could have
istics are basically undesirable, since engine burned given sufficient time (chamber
complexity is drastically increased. The addi- volume). This percentage must be con-
tion of such components should be avoided at all sidered for accurate determination and
costs. Here again, close coordination between optimization of the composition of the
the vehicle (guidance) designer and engine de- combustion gases and when optimizing
signer, and thorough understanding of their com- the gas properties with energy release
mon problems, is vital. and system weight.
(2) Cooling conslderations.-The temperatures
Mixture Ratio resulting from stoiehiometric or near-
stoichiometric mixture ratios, dependent
As is well known, complete combustion of a on propellant type, may impose severe
given amount of fuel requires a corresponding demands on the chamber-wall cooling
system. A lower temperature, therefore, Weight

may be desired and obtained by select-
The parameter of weight, as no other, domi-
ing a suitable ratio.
nates the thinking of those employed in rocketry.
Once the optimum mixture ratio has been deter-
Weight of payload flown over a distance, or
mined for a given engine system, based on the
placed into orbit, is the ultimate accomplish-
major factors just discussed, it is obvious that
ment. Success is often gaged directly in pounds
deviations from it would result in engine per-
of payload flown per dollar spent.
formance penalties. Since the vehicle powered The importance weight rightfully carries does
by an engine will have been sized and tanked to
not necessarily mean that it is all important. For
conform with the specified engine mixture ratio,
instance, a somewhat smaller payload placed
it is important to know that deviations will also
into orbit more reliably, or at a lower cost per
result in reduced vehicle performance, namely:
pound, may be preferred. By and large however,
(1) Reduced engine duration, due to premature
weight is a most important consideration.
exhaustion of one of the propellants As we have seen earlier, a vehicle's final
(2) Reduced mass ratio, due to excessive
velocity is a function of, among other param-
residual amounts of the other propellant
eters, its mass ratio. The smaller the final
(increased burnout weight)
mass, the higher the final velocity. However,
Since the relationship between engine per-
since payload mass should be as high as possi-
formance (/s) and mixture ratio for many systems
ble, the weight squeeze is applied to all those
is usually relatively flat near the optimum point
vehicle components which are not payload. This
(fig. 2-3), the effects from duration and burnout
includes the engine.
weight may well be the most influential ones for To isolate the influence of vehicle-structures
vehicle range. weight, a parameter called "propellant fraction"
The effects of even minor discrepancies in has come into increased usage. This factor ex-
mixture ratio (propellant utilization) are sub- presses the ratio of the total propellant weight
stantial. For instance, in a typical single-stage to the fueled vehicle weight without payload.
medium-range ballistic missile, each pound of Typical values are 0.94 for turbopump-fed sys-
excess burnout weight will result in a range de- tems, and 0.89 for pressure-fed systems. For
crease of approximately 0.2 nautical miles. For turbopump-fed engines, the ratio of thrust to
long-range vehicles, the penalty is still higher. engine weight is a useful additional yardstick.
The close target tolerances that have occasion- Larger modern liquid rocket engines may fall
ally been reported for test flights illustrate the into a range from 75 to 125 pounds of thrust/lb
remarkable degree of accuracy which can be of engine weight. These figures represent asub-
achieved from all contributing subsystems. stantial progress over the past (see fig. 2-4).
As was seen with residual propellants, exces-
sive dead weight at burnout imposes penalties.
Therefore, whenever rocket engines can be made
lighter without compromising reliability and
320
structural integrity, the payoff in range and pay-
load will be sizable.
Engine and vehicle builders usually distin-
guish several types of engine weight:
,.-
0 zao / (1) Dry weight.-The net weight of the engine
as it leaves the factory.
Z60 (2) Burnout weight.-The engine dry weight
LO I 2 L4 1.6 L Jr 2D
O/F MIXTURE RATIO

plus residual, measurable propellants
remaining in the engine at cutoff. In a
Figure 2-3.-Theoretical thrust chamber perform- typical engine design, burnout weight
ance vs mixture ratio for N204/N2H4 at may be 4 percent higher than dry weight.
Pc = 1000 psia shifting equilibrium and optimum Burnout weight is significant for vehicle
sea level expansion. mass ratios (eq. 1-30).
(a)
(b) I (c) l
_arly Navaho Engine Early Redstone Engine German V-2 Engine

(Rocketdyne 1953) (Rocketdyme 1952) (1942)
ThrustsL: 120,000 ib ThrustsL: 75,(>00 ib Tb_ustsL: 56,000 lb
Dry Weight: 1230 lbs _ry Wei___ht: 1475 lb Dry _elght: 2484 lb
Is3 L = 230 sec Iss L = 215 sec Iss L = 199 sec
Figure 2-4.-Substantial progress has been made in ratio o[ thrust to engine weight as demonstrated
by: (a) postwar engine (1953), thrustsL: 120000 lb, dry weight: 1230 Ib, IssL =230 sec; (b) post-
war engine (1952), thrustSL: 75 000 lb, dr}, weight: 1475 Ib, lss L =215 sec; (c) German V-2 engine
(I942), thrust: 56 000 Ib, dry weight: 2484 lb, lss L --199 sec.
(3) Wet weight.-The engine dry weight plus ing and routing of lines, avoidance of traps, and
all propellant within it, during main location of valves.
stage. In a typical design, engine wet Because of the importance of weight control,
weight may be 6 percent higher than dry rocket engine manufacturers employ engineers
weight. Wet weight is significant for specifically in charge of this area. Table 2-i
vehicle in-flight center-of-gravity loca- shows a typical weight progress form, as it is
tion and moments of inertia. used by the Rocketdyne Division of North Amer-
can Aviation. Itis revised and reissued period-
(4) Wet gimbaled weight.-That portion of wet
ically. Thus it becomes a useful tool to raise
weight representing engine mass which
early danger warnings. In our arbitrary example
is gimbaled for steering purposes. In
a slight underweight is shown. However, the
earlier designs this meant essentially
table also shows that the data are based almost
the thrust chamber and injector wet
entirely on estimated and calculated figures,
weight. In later designs it often refers
rather than on actual weighing results. This is
to the entire engine less a relatively
characteristic for the earlier phases of design
small amount of stationary parts. This
and development of a rocket engine. More often
weight is significant for gimbal actuator
than not, the weight advantage will disappear
loads and guidance control loop re-
gradually as the design firms up; then the
sponse characteristics.
squeeze will be on. For convenient display of
Ideally, dry weight and burnout weight should be the weight tendencies over time, a graph such as
equal: that is, no propellants should be trapped shown in figure 2-5 will be useful.
in the engine at shutdown. In practice, this will The weight changes of the various compo-
not always be possible. However, the engine nents as well as of the entire engine affect cen-
designer can do much through proper design, siz- ters of gravity and moments of inertia. Through
2600
2450
REV SPEC BURNOUt
2300
REV SPEC DRY
//
1
2150
BURNOUT
2000
_" ......_ /
_ )85o
/
ENGINE 1700 /
ACCESS ORIGINAL
BURNOUT ..................... IS PEc 'r...........................

ORIGINAL
ENGINE
ACCESS ................ iSPECI.......................
DRY 1400
I 2 3 4 5 6 7 8 9 I0 II 12 13 14 15 16
GO MONTH
AHEAD ,
Figure 2-5.-A-2 stage rocket engine and accessory weight history.

z
_=
E; 4. + _- 4-
• + -_-
oj
b-
• ¢¢
I
"7,
c_
L_.
E_
+
=
ROCKET ENGINE DESIGN IMPLEMENTS
A-2 STAGE ROCKET ENGINE
CENTER OF GRAVITY AND MOMENT OF INERTIA DATA
ISSUE ENCLOSURE
DATE PAGE I OF r
, Z
( LOX Pump I
( • y
( _"'_ F uet Pump I

_ GIHBAL ( Y 0 )
I
NOTE: (_erns (I) thru (31 Tepcesent the rnorr_t o( iner',ia
about specified C.G,'s. items (41 and (S) represent
U_e moment of inerti_ about the referenced gJmbal axis.
MOMENT OF INERTIA - SLUG FT 2

WEIGHT
DESCRIPTION
LBS. t, Y - CENTER(OF
ARH X GRAVITY
- ARM - INCHES
Z - ARM
Y-Y ( X-X IZ-Z
.233 . IS 07 t76 391 362

(I) RocketEnsine _ Ace;- DrX 2181
(2) Rocket Engine _ Act. - Wet 2317 -225 .15 --0 I 185 4tl 379
(3) Rocket En|ine . Ace. - Burnout 2292 -227 .I 6 -0 2 184 408 375
{4) Gimb_lled Mass - Dry 2061 -25.2 -I 5 0.2 I_ 672 649
(5) Gimballed Mass - Wet 2086 -246 . I S -02 177 688 662
Figure 2-6.-Typical data sheet for center of gravity and moment of inertia.
issue of a data sheet as shown in figure 2-6, all Vbo -- Cvc • g" (Is)on in
parties concerned can be kept informed on
changes as they occur. Stage usable Stage Stage \
propellant +payload+ inert
Note that the data presented in table 2-1 and
weight weight weight.)
figures 2-5 and 2-6 are for the 150K A-2 stage (2-1)
engine system which is a part of an assumed weight + weight /
multistage space vehicle configuration treated in
later chapters. where
Let us now explore the influence structural
weight has on the performance and gross takeoff Stage inert_ Stage residual propellant
weight of a rocket vehicle system, and how its weight weight at burnout
magnitude varies with the design parameters of
Stage structure,
different vehicle systems. The quantitative
Stage engine . guidance and other
relationships will be evaluated individually for (2-2)
system weight T weights, which are
each case.
not payload
Equation (1-30) can be rewritten for the stage
burnout velocity of a single-stage vehicle, or the It can be concluded from these equations that
stage velocity increment, of any individual stage for a given burnout velocity, there is an even
of a multistage vehicle system as: weight trade off between stage engine system
weight and stage payload weight. If the weight vehicle trajectory. Therefore, if one part of the
of all other items were kept constant, a pound vehicle system exceeds its weight allotment by
decrease in the stage engine system weight will 1 pound, an increase of the total vehicle system
increase the stage payload capacity by 1 pound. weight at takeoff by a certain number of addi-
For a fixed payload, and assuming other items tional pounds will result. Growth factor is de-
except engine weight to be constant, the relation fined as the total vehicle system (including pay-
between the stage velocity increment, Vbo, and load) weight increase at takeoff, divided by the
stage engine system weight for a given system causal increment of added inert and/or payload
can be written as weight. It is emphasized that the growth factor,
for a given vehicle system, is not a precise
value, but varies within a band. For instance, a
Vbo = k 1 In k(__.
++Stagestage
engineengine
system
_] weight._ (2-3) small weight increase of a component in an ex-
isting system may only require the addition of a
where corresponding small amount of propellants, but
k_ = Cvcg(Is)oa = constant not require enlargement of the tanks, valves, etc.

Stage residual Accordingly, the growth factor will be small. In
k2 = Stage payload + propellant weight another case, the weight increase may be "the
weight at burnout straw that breaks the camel's back," requiring
the use of the next larger valve size, duet size,
, Stage structure, guidance = constant
and other weight or the like. The growth factor will then be large.
lca = Stage usable propellant weight+k: = constant In general, however, the growth factor of a
Since k 2 < k3, the denominator will decrease vehicle system is a useful tool during the pre-
more rapidly than the numerator, with decreasing liminary design of an engine system, because it
engine weight. Thus for fixed payloads, an in- attaches a tangible value to the importance of
crease in burnout velocity is realized which will the engine-system weight. A systems weight
pay off in longer range or higher orbit. increase may be considered "uninvited payload."
For a given burnout velocity and for a fixed For single-stage vehicles, and relatively small
weight changes, the value of the growth factor
payload, the required stage average overall spe-
cific impulse (ls)oa in terms of stage engine then can be expressed with sufficient accuracy as
system weight can be established as

Total vehicle system
weight at takeoff
Growth facto[- (2-5)
Payload weight
k3
, + system
Stage weight
engine ]\ (2-4)
('s_oa:k4.'/in k 2 / For any stage of a multistage vehicle, the
+ system weight /
approximate value of the growth factors against
where total vehicle system weight at takeoff can be

expressed as
Vbo
k 4 = C--vcg = constant
Total vehicle system
weight at takeoff
Equation (2-4) shows that with decreasing Growth factor = (2-6)
Stage payload weight
engine-system weight, the overall specific im-
pulse requirements decrease.
The growth factors of any stage against the
Another parameter illustrating the importance
vehicle system weight at ignition of the same or
of weight is the growth factor of a rocket vehicle
system. For instance, if the weight of a compo- lower stage can be expressed as
nent increases, it is possible to adjust for this

Vehicle system weight
by increasing the weight of the propellants
at same or lower
loaded and thus possibly that of other compo-
stage ignition
nents, such as a pump, to maintain the same Growth factor = (2-7)
required vehicle performance; i.e., payload and Stage payload weight
ROCKET ENGINE DESIGN IMPLEMENT5 41
Sample Calculation (2-I) For each pound increase of third-stage engine-

system weight, the increase on vehicle sys-
A three-stage rocket vehicle system has the
tem takeoff weight = 62.9 pounds
following weight data: Total vehicle system
(b) Note that the weight growth of lower
weight at takeoff, 40000 pounds. Vehicle sys-
stages will not affect the upper stage weight
tem weight at second-stage ignition, 7500
growth. For an increase of first-stage vehicle
pounds; vehicle system at third-stage ignition,
system weight, there will be no weight changes
2200 pounds; payload weight, 700 pounds. For
on second and third stages, and for an increase
each pound increase of engine system weight of
on second-stage vehicle system weight, no
first, second, and third stages, respectively,
weight change is required for third stage.
determine (at a constant vehicle performance):
(a) increases of total vehicle system weight at
(1) Growth factor of second stage against
takeoff; (b.) increases of vehicle system weight
vehicle system weight at second-stage
at second- and third-stage ignition.
ignition =
Solution Vehicle system weight
(a)Payload weight of first stage = vehicle at second-stage ignition 7500 3

Sec'-ond-stag------_
payl----oa--d
wei--_ht =2_ = .41
system weight at second-stage ignition =
7500 pounds
Payload weight of second stage =vehicle (2) Growth factor of third stage against
system weight at third stage ignition = vehicle system weight at second-stage
2200 pounds ignition =
Payload weight of third stage = actual
system payload weight =700 pounds Vehicle system weight
From equation (2-6): at second-stage ignition 7500
-10.72
(1) Growth factor of first stage against ve- Third-stage payload weight- 700
hicle system takeoff weight =
(3) Growth factor of third stage against
Vehicle system takeoff weight 44000 vehicle system weight at third-stage
=_=5.86 ignition =
First-stage payload weight 7500
(2) Growth factor of second stage against Vehicle system weight

vehicle system takeoff weight = at third-stage ignition 2200
-_=3.14
Third-stage payload weight- 700
Vehicle system takeoff weight 44000 _^
Therefore:
Second-stage payload weight = _ =gu
For each pound increase of second-stage en-
gine system weight, the increase on vehicle
(3) Growth factor of third stage against
system weight at second-stage ignition = 3.41
vehicle system takeoff weight =
pounds
For each pound increase of third-stage engine
Vehicle system takeoff weight =44 000 = 62.9 system weight, the increase on vehicle sys-
Third-stage payload weight 700
tem weight at second-stage ignition = 10.72
pounds, and the increase on vehicle system
Therefore:
weight at third-stage ignition = 3.14 pounds
For each pound increase of first-stage engine- The correctness of results can be checked by
system weight, the increase on vehicle sys-
recombining the individual stage growth factors
tem takeoff weight = 5.86 pounds to obtain the growth factor for the entire vehicle
For each pound increase of second-stage system:
engine-system weight, the increase on vehicle
system takeoff weight = 20 pounds 3.14 × 3.41 × 5.86 = 62.9
Envelope (Size) the reliability of their products. This emphasis

on reliability is well justified and is of particu-
The linear dimensions of liquid propellant
lar significance to rocket engines. The advent
rocket engines require relatively elaborate de-
of manned space flight has placed even greater
scription and frequently cannot be made clear
emphasis on rocket-engine reliability.
without a drawing. In those cases where only
Reliability may be defined as the capability
approximate values are required for comparison
of the engine to perform according to specifica-
or for overall estimates, the term "envelope" is
tions, whenever "the button is pushed." The
preferred. For instance, definition of a hypo-
degree to which this is met can be expressed in
thetical smallest cylinder, cube, or sphere into
figures and graphs. If the evaluation is made
which the engine would fit conveys a good feel-
following a test series, reliability can be simply
ing of engine size or bulkiness.
expressed as the ratio of success to failure, say
Obviously, engine size directly affects engine
98 percent (2 failures and 98 successes in 100
weight, the importance of which was emphasized
runs). As there is no guarantee, however, that
above (fig. 2-4). Aside from the engine itself,
the system under test will perform identically in
mnnerous other areas are directly affected by
subsequent tests, reliability predictions are
increasing engine size:
made, the accuracy ("confidence level") of which
(1) The vehicle structure, which becomes increases with the amount of previous informa-
heavier, especially with upper stages. tion available. The interrelation of reliability
Engine size directly affects the size and and its confidence level is something the statis-
thus weight of the aft end and/or inter- ticians have much to say and write about.
stage structure. What can the rocket engine designer do to
(2} Handling equipment and procedures be- achieve the highest possible reliability, as early
come more costly as possible? Below are compiled a few pointers
(3) Servicing becomes more difficult and thoughts which have proven valuable, not
(4) Manufacturing machinery becomes larger only in rocket engine design. They will be fol-
(5) Storage and transportation means become lowed by specific details for the implementation
more bulky of a reliability-assurance program.
In several of these areas, there is a definite First of all, painstaking execution of all cal-
upper limit, such as railroad tunnel sizes, clear- culations and drawings that are part of a given
ances on bridges and underpasses, and available design is an obvious requirement. This includes
machine tools. the thorough study of previous experience, one's
The selection of the thrust-chamber expansion- own as well as that of others; familiarity with
area ratio has a very pronounced effect on engine and correct application of accepted and proven
envelope. When optimizing the thrust chamber design standards and procedures; clearly written
expansion area ratio, which is also influenced by statements and instructions; clear line draw-
performance, weight, pressure drop, heat trans- ings. It cannot be overemphasized: it pays to
fer, and other considerations, its effect on en- spend that extra hour in carefully checking re-
velope, and thus on other vehicle systems, must peatedly every detail of a design and its contem-
be considered (section 10.9). plated mode of operation, before its commitment
to manufacture and subsequent use. Neglectmay
have to be paid for by many months of toilsome,
Reliability
tearful, embarrassed "corrective action," often
The subject of reliability has become almost causing losses of hundreds of thousands, even
a branch of science by itself. In addition to the millions of dollars. When making these checks,
designer, to the development engineer, and to the most pessimistic assumptions of what some-
the user, mathematicians, statisticians, and one else may do wrong during manufacture,
"human factor" and "man rating" specialists are assembly and use, are not out of place.
involved. Numerous books have been written on The designer should not rely solely on his
the subject and manufacturers maintain entire own judgement. Careful and independent check-
g['oups to predict, monitor, tabulate, and evaluate ing of all calculations and designs by superiors
--..__ _. T -- ¢
and by independent checkers is important. Early several years, may elapse between delivery and
availability of a wooden (or "soft") mockup of final use. Much can happen during this period.
the engine under design will be an invaluable For instance, changes of plans for the mission
tool to avoid costly errors that subsequently may may have made another thrust level more desir-
seriously affect schedules and reliability. Spe- able. In this case, the adjustment by means of
cific recommendations for design and checking orifices, in particular its verification, becomes a
techniques will be made in section 2.2. major operation. While the omission of a stra-
"Reliability" is sometimes treated as being tegic regulator was indeed an engine simplifica-
synonymous with "simplicity." Undeniably, tion, for the vehicle system it turned out to be a
simplicity of a design contributes significantly complication. The point here again is; the care-
to increased reliability. Parts which do not ful- ful evaluation of a planned omission must con-
fill a truly useful purpose should be omitted. sider all aspects, including changes of plans.
This may include many of the so-called safety
features and interlocking devices, which often
Reliability Assurance
cause more trouble than they prevent. Early
designs of liquid-propellant rocket engines have The emphasis on reliability must not remain
indeed frequently suffered from such an overdose an empty slogan. Fortunately, implements are
of sophistication and safety devices. Many of available to the rocket engine designer which
the more recent designs have been substantially can assist him effectively to achieve the highest
improved in this area, to a point where caution degree of reliability. One of these, an effective
must be exercised not to overshoot the target and failure reporting and correction system, will be
not to lose that flexibility which only liquid- discussed in section 2.2. Equally, if not more
propellant systems can provide, as compared to important, is a most effective failure prevention
solid-propellant systems. Simplifications, like system. The numerous activities contributing to
all other design features, must be carefully the latter may all be considered part of a reli-
planned and evaluated. Simplification by elimi- ability assurance program. The quality of de-
nation of a useful component must not become an sign, without question, is the program's founda-
excuse for failure to improve that component if tion upon which all subsequent phases rest. The
its absence could severely penalize other sub- characteristics of a reliability-assurance pro-
systems, or maintenance and servicing proce- gram, then is that its most significant steps
dures. (analyses, design reviews, design improvements)
For instance, to avoid a troublesome sealed are taken before the design of a component is
connection it may be decided to omit flanges and finalized; before the development test program
seals and to weld it. However, if one of the is initiated; and again before the first vehicle is
lines thus connected were inadvertently pinched committed to launch.
in the field, removal of the entire engine from a
vehicle under preparation for launch would be-
Definitions
come necessary. Thus, a simple replacement
may be magmfied into a major operation. To be The definitions used in rocket vehicle reli-
sure, welding or preferably brazing may indeed ability assurance programs vary widely with
be the best solution for many problem connec- individual preferences, with the object under
tions. The point is, this will not be true for all design and development, and with the missions
connections. Careful analysis of all aspects contemplated. The definitions given below are
including handling and in particular mishandling typical, have been used in actual rocket engine
by the user, is necessary. and vehicle programs, and can be readily adapted
In another example, tests may have shown to others. For the sake of clarity, irrelevant
that an engine could readily be set up and cali- jargon and detail have been omitted.
brated to specifications by means of orifices,
eliminating previously-used regulators. Engines
Reliability
are delivered accordingly. With rocket engines,
it is entirely normal that many months, if not The probability that a part or system will
function properly and if necessary repeatedly Man Rating

under rated operating conditions, within the
Design and operationalprovisionsto assure
specified load and time limits.
crew survivaleven in case of mission failure.
Thus, man-ratedreliability must be higherthan
Mission Success mission reliability. For instance, overall vehicle
reliability to achieve mission success may be
Completion of the rocket vehicle mission
95 percent. By the addition of an escape mecha-
objectives within specified tolerances. All sub-
nism, man-rated reliability may be increased to
systems, including the engine, contribute to the
99.5 percent. Caution is advised not to become
success. It is an inherent characteristic of
entirely "wrapped up" in man rating, at the ex-
mission-success analysis and assurance that
pense of mission reliability. A single launch of
they anticipate the probability of certain part and
a man-carrying space vehicle costs several
subsystem malfunctions, offsetting them with
hundred million dollars, all told. Investment in
appropriate countermeasures (such as redundan-
means to save the mission as well as the man,
cies, emergency power sources, power and pro-
therefore, appears to be prudent. Table 2-2
pellant reserves, and others).
illustrates this clearly. For optimum reliability
of spacecraft and launch vehicle including the
Mission Failure engines, the need for a crew escape system is
minimized. Both, mission and crew survival are
Failure of the rocket vehicle to complete the
assured with high reliability.
mission objectives. Mission failures can be
classified as: a) Catastrophic, b) Critical, and
c) Deferred. TABLE 2-2.-Relationship of Vehicle Reliability
to Flight Safety
Catastrophic Failure Reliability Flight safety
A failure in which the time between the fail-

Spacecraft and Escape system Probability of
ure event and a subsequent crew hazard is less launch vehicle crew survival
than 500 milliseconds. Abort sequence must be
0.50 0.998
automatically initiated. 090 0.99 0.999
0.999 0.00
Critical Failure
Engine Out
ure event and the hazard ranges from 500 milli- Design and operational provisions to permit
seconds to five seconds. Abort sequence may be limited or complete mission continuance in case
initiated automatically or manually. one engine fails to fire, or malfunctions and is
shut down. Possible only with vehicles having
Deferred Failure engine clusters. See airplane analogy under
"Deferred Failure," above.
ure event and the hazard is five seconds or
Failure Mode
greater. Action to cope with the failure is de-
ferred to allow analysis by the pilot or an auto- The manner in which a part or system mal-
matic logic, to decide whether corrective action functions. This may be a "short" or "open"
can be taken or an abort sequence should be circuit, an incorrectly "closed" or "open" valve,
initiated. Typical example: shutting off an an engine out, or similar malfunction.
engine with a feathered propeller in a four engine
airplane and reaching destination safely though
Order of Failure
with a delay. Analogous provisions are antici-
pated for manned rocket flight. The number of components in a system which
would have to fail, regardless of their failure Failure Modes of Engine Components
mode, to cause systems or mission failure.
The failures of rocket engine components may
First-order failures are failures caused by a mal- be attributed to one or a combination of several
function of a single component or part. Second- of the following principal modes:
and higher-order failures are defined in a like
(1) Functional failures
manner. Typical example: a stuck pressurizing (2) Fatigue failures
valve causing overpressure in a vessel would
(3) Over-stress and over-strain
rupture it only if the safety valve failed to open;
(4) Failures pertaining to combustion devices
this would be second-order failure. However,
(5) Failures pertaining to electrical devices
continuous venting of a properly opening vent (6) Manufacturing and material defects
valve may prematurely deplete gas supply. A (7) Unexplained failures
thorough failure-effect analysis will reveal all (8) Human errors
ramifications. In the example, depletion would
not occur instantaneously, this would be deferred
failure. The designer can do something about it Functional Failures
in advance: provide an overriding closing valve These are malfunctionsof parts or compo-
for the pilot, which remains completely inactive nents due to reasons other than structuralfail-
when not needed, but adds weight.
ures. For instance,an "0" ringmay failto seal
due to impropergroove depth specifiedin the
Failure-Mode-Effect Analysis design. Or, a plunger may freeze in the bore of
a guiding bushing,because of improper surface
An orderly and qualitative listing of the finishand/or noncompatibilityof materials. To
modes in which components or parts of a system minimize possible functionfailuresin the design
can fail; the effects of the failures on the en- of engine components the followingprecautions
gine's or vehicle's ability to complete the mis- are recommended:
sion; and the order of the failures. Such an (I) Choose proven designs with an estab-
analysis should distinguish between the pre- lished servicerecord.
launch, launch, and cutoff phases. Also, all (2) Use standardmechanical elements (bolts,
identified failures should be classified as cata-
nuts, threads,gears, pins, rivets,
strophic, critical, or deferred. springs,seals, tube fittings,pistons,
keys, shafts,bearings)wherever pos-
sible.
Failure Mode Cause Analysis (3) Select simple designs, but without impair-
An analysis listing all the conceivable rea- ing flexibility.In particular,minimize
sons why each mode of failure could occur. the number of moving parts and sealing
surfaces.
Likewise, reasons for each potential cause not
(4) Allow adequate functional margins in the
occurring should be explained in detail.
design of components (spring forces,
actuating powers, supply of lubricants,
supply of coolants).
Emergency Detection System (EDS)
(5) Subject newly-designed parts to extensive
The EDS comprises the electromechanical functional testing, under simulated work-
devices, including sensors and discriminators, to ing and environmental conditions, before
detect an imminent malfunction. Depending on "freezing" the final configuration.
the type of failure (catastrophic, critical or de- (6) Provide redundancy. This is a "buddy
ferred) it may initiate immediate action, or defer plan": where one component would be
but store and/or display it in a suitable manner sufficient, two of the same type are
(timer; visual gage or light). Inputs to the EDS actually provided. If one fails, the other
must be analyzed, selected and provided by the takes over. This can be achieved in two
designer, in particular the engine designer, at ways: by noncomplex and by complex
the outset. redundancy. Intelligently applied,
4G DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
redundancy can significantly increase A typical example is an electric power emer-

reliability. gency battery with voltage sensor and switchover
(7) At all times, pursue a rigorous program of circuitry.
product improvement.
Fatigue Failures
Noncomplex Redundancy
Fatigue failures are fractures caused by re-
The simultaneous function of identical equip- peated load applications at stresses consider-
ment. Application depends upon the particular ably lower than those causing failures in a single
failure mode wh - is to be eliminated. For a
load application. They are the most common
typical example, _ve figures 2-7 and 2-8. Other type of mechanical failure. The ability of a part
examples are: dual (series) seals, parallel to resist fatigue failure cannot be checked with-
valves.
out destroying the part. Checking is possible,
however, through destructive endurance tests
with representative samples selected at random.
PRESSURE Most fatigue failures start with a crack at or

SWITCH :IS I SOLENOID near an outside surface because stresses are apt
POWER VALVE_
to be greatest there. The actual failure will
result from gradual propagation of these cracks.
PRESSURE
SWITCH = 2 1
The point at which the crack will start will
depend upon the geometry of the part and on
Figure 2-7.-Noncomplex paraI1el redundancy.
surface conditions. Any notch or other stress
(This type of redundancy guards against fail-
raiser, being a point of highest stress concentra-
ure to close when called upon to close.)
tion, is a potential starting point for fatigue
cracks. Fillet radii that are too small, threads,
SOLENOID oil holes, keyways and similar surface irregulari-
POWE R VALVE ties are all potential sources of fatigue failure.
Although a part may be designed to be free of
geometric irregularities, having no shoulders,
SWITCH =1_"I SWITCH _ 2
grooves or the like, it may still contain a great
Figure 2-8.-Noncomplex s e tie s redundancy. number of minute stress raisers. These may be
tool marks, scratches, identification stamp
(This type of redundancy guards against inad-
marks, or various inherent discontinuities in the
vertent closing, i.e., closing when not called
material itself, such as inclusions of foreign
upon to close.)
matter and quenching cracks. The design engi-
neer should make every effort to avoid stress
Complex Redundancy
concentrations in a highly-stressed part subject
The original function carried out by one com- to repeated load applications. In the design,
ponent. Failure sensors, logic circuits and rigid specifications should be called out for
switching devices energize an identical standby surface finishes. For repeated load services,
component, when needed. The advantages ob- forgings are generally preferred to castings.
tained can be completely offset by the additional Ductile materials are preferred to material prone
complexity of sensing and switching circuitry. to become brittle.
The potential problem area may be merely shifted In welded constructions, the joints are sub-
from the equipment to the failure-detection com- ject to almost all types of stress concentration
ponents. However, this standby redundancy may and fatigue failure. Wherever possible, welded
be advantageous when long mission times are joints should be minimized in the design of parts
involved (e.g., days or weeks) and where it may subject to repeated loads. Rigid procedures for
be undesirable to also subject the backup equip- welding and inspection must be called out in the
ment to prolonged operation. design.
tli_. 7, -,
Over-Stress and Over-Strain Failures of Combustion Devices
Stress analysis in mechanical design to pre- Under steady-state operating conditions, com-
vent over-stress and over-strain will be dis- bustion devices in liquid-propellant rocket en-
cussed in section 2-4. The interrelationship of gines are exposed to hot gases with temperatures
stress and reliability of mechanical parts is ranging from 1000 ° F to 6000 ° F. The walls of
illustrated in figure 2-9. these devices are either made from high-
temperature-resisting (refractory) materials, or
are provided with effective cooling, through heat-
absorbing effects, ablative cooling, propellant
z _z film and/or regenerative cooling. Structural
m failure may occur because of erosion, from wall
temperatures exceeding the values assumed dur-
ing design. Or failure may occur from a combina-
-ID RELIABILITY
tion of excessive temperatures and pressures.
WORKING _-_ MARGIN /-.- DAMAGING
STRESS _ . I_ STRESS
Under certain transient or unstable conditions,
such as during engine start or stop, combustion
y f O,ST.,BOT,ON
iV\
instability or abrupt pressure surges may occur
and cause a failure. See chapter IV, "Design of
Thrust Chambers and Combustion Devices."
STRESS Electflcal Failures

Figure 2-9.-Interrelationship of stress and reli-
Although predominantly an assembly of me-
ability as related to mechanical parts.
chanical parts, a modern rocket engine employs
a number of electrical devices without which it
Two stress levels exist for every part in a cannot function reliably. Among the electrical
given engine component: the working stress, and components used most widely are: power sources
the damaging stress at which failure occurs. The (batteries), converters (DC to AC), wires and
failure may be either a fracture, or a deformation
harnesses, connectors, switches, relays (elec-
beyond allowable tolerances. Each of the two tromechanical and solid-state), timers, pressure
stresses are mean values of a distribution about switches, diodes, solenoid valves, servomotors
a mean. The difference between the working and and position indicators.
the damaging stress mean values is indicative of All of these devices are, to various degrees,
the stress reliability margin of the part. potential sources of failure, the consequences of
The deviations from the mean working stress which are just as detrimental as failure of me-
result mainly from variations in the dimensions chanical parts. By proper design and assembly
of the part, and from operational and environ- instructions, and by careful selection of the
mental conditions. The distribution about the elements, the designer can forestall electrical
mean damaging stress results from variations in failures and thus assure overall systems relia-
material properties, fabrication processes, quality bility. Most common potential electrical failures
control and maintenance practices. The area Pt, which must be prevented are:
where the two distributions overlap represents
the probability of failure, or unreliability. Cold solder spots in connectors, wirings
Close control of functional and environmental and electrical elements. They result in often
loads may decrease the variation of the working sporadic discontinuities, particularly under
stress about the mean. Likewise, better mate- vibration.
rials and strict quality control should increase Short circuits in wirings, connectors and
the damaging-stress mean value, and decrease other electrical devices. This may be due to
the variation about the mean. Thus, the area of poor design, leaving insufficient separation
overlap may be substantially reduced or elimi- between connector pins, lugs and the like;
nated, and reliability increased. excessive solder; damaged insulation due to
poor harness installation, chafing under vibra- cal properties, fabrication and heat-
tion and poor handling; overload and/or over- treating characteristics, and possibly of
heating in solenoids; moisture in connectors. other properties as the application may
Fused relay contacts, due to overload demand.
and/or incorrect current rating of the ele- (2) Traceability. By the time a materials
ments. defect is detected, many more parts may
Relay and switch contact loss under vibra- have been made from the same lot or
tion. This is really an electromechanical batch of raw material. If accurate rec-
malfunction. It can be prevented by proper ords have been kept, for what parts
relay selection, shockmounting, orientation of which numbered material lots have been
installation, replacement by solid-state cir- used, it will be possible, without a
cuitry where possible. "panic," to withdraw and replace all
Power failure resulting from one or more of parts made from the faulty lot. The im-
the causes listed above. Prevention means portance of accurate and complete rec-
include emergency batteries and overload ords is obvious.
switches, combined with subcircuit isolation (3) Process inspection. This refers to all
through diodes. shop inspections made of the parts as
A liquid-propellant rocket engine usually in- they are being processed. Its purpose
cludes additional electrical elements as required is to check the performance of the oper-
for instrumentation and telemetry. These may ators and tools or equipment, and to sort
include instrumentation power supplies, end out faulty parts as soon as errors are
organs (sensors, pickups, thermocouples, accel- detected.
erometers, position indicators), signal condition- (4) Salvage. This includes the decision
ers (analog-to-digital), and wiring. Although, as whether defective materials or parts can
a rule, instrumentation is not required directly be corrected or will be scrapped.
for proper function of the engine system, its fail- (5) Final inspection. This is the inspection
ure may indirectly cause engine malfunction, by of the completed component parts to in-
interference with engine operation. For instance, sure they are within the limits required
a pressure pickup may rupture and cause pre- by the design.
mature depletion of a gas supply; the same (6) Checking of actual service performance.
event in the fuel system may cause an engine- This includes investigation of com-
compartment fire; improper installation of a plaints and studies of actual service
thermocouple may block a vital lubricant or other performance of the part in the engine
line. system (see "Failure Reporting System").
The engine designer's task includes engine
instrumentation. It is obvious, therefore, that
Unexplained Failures
this peripheral system, too, will require his full
attention. From time to time failures occur which cannot
readily be isolated as having originated in a
Manufacturing and Material Defects
given component or part. This may be because
Manufacturing and material defects of engine several causes were present simultaneously and
component parts directly affect the reliability of could not be separated. Or instrumentation was
the components. These defects can be prevented inadequate, recording only the ultimate effects
only by strict quality control. The areas of but not the cause. Or that an incomplete inves-
quality control may be subdivided as follows: tigation was conducted, maybe in the hope it
(1) Materials inspection. The extent of in- would not happen again.
spection and testing conducted with raw Classification of a failure as _unexplained"
materials depends upon the nature of the should never be done as a matter of convenience,
part for which they will be used. In- but only as a last resort if the most thorough
spection of materials includes testing of investigation did not establish a clear cause.
their composition, uniformity, mechani- Complete and accurate records must be kept of
the details associated with the failure. Through mostly known methods and common sense, the
special statistical methods it may be possible to rocket-engine designer can do much to prevent
isolate the cause or causes at a later date. human error by the design of his parts and by
Also, special instrumentation may immediately mandatory actions during their building, checking
be added in the areas of suspected but unproven and handling. The following are only typical
causes. But the most sophisticated explanation examples of a probably infinite number of possi-
of a failure cannot transform it into a success. bilities of preventing human error by design.
Clear marking of bolts, lines, connectors,
wires, etc.
Human Error
Use of dissimilar connectors, dissimilar
Experience with early manned rocket flights threads, a variety of keyways, and the like,
has shown that certain functions could be per- to make incorrect electrical and mechanical
formed better by trained men than by automatic connections impossible.
devices. It appears certain that future space- Safety wiring; electrical and mechanical
flight efforts will continue to employ combina- interlocks.
tions of automatic and manual systems, the ratio Painting bright red, and/or attaching bright
of the two depending on mission requirements. red streamers to, auxiliary devices which
Even in those missions which are unmanned and must be removed before operation.
therefore appear to be fully automatic, man is Storing of components and systems which
still directly involved: during launch prepara- have successfully passed all inspections and
tions, and through postlaunch commands for checkouts, in locked rooms requiring two dif-
trajectory-correction maneuvers, information ferent keys from two different individuals.
playback, and similar actions. Intentional attempts at improper handling,
We must realistically recognize that even a assembly, checkout and operation of compo-
fully-trained, and very alert individual is subject nents and systems, to assist "foolproofing,"
to error. He may commit an error at a desk while during development.
using a slide rule, or while connecting a number None of these and numerous similar steps will
of hydraulic lines, or under stress while pushing eliminate human errors completely. However,
buttons. The history of rocketry is full of glar- they will substantially reduce their possibility
ing examples. Whenever it happens, the most and thus contribute to overall reliability.
eloquent regrets are probably voiced by the de-
signers of the system involved: if someone had
Design Reviews for Reliability
not done something to their perfect creation, it
would have worked faultlessly. The reaction is The reliability of rocket engines and their
understandable, but wrong. components depends on many factors. Design
For whatever reason a system failed, it obvi- reviews are among the most potent methods for
ously was not perfect. Its designer did not con- increasing reliability. In practice, a design re-
sider human nature, nonideal operating condi- view is the progressive evaluation of a design.
tions, nor emergency situations. Clear and It starts with the preliminary schematics, layouts
complete drawings, specifications and other and specifications and extends through the re-
written instructions are important but they will lease of all final drawings. Those conducting
not prevent, by themselves, human errors, be- reliability design reviews should include: relia-
cause they can be misplaced, misread, or not bility specialists, design engineers, develop-
read at all. ment, test and service engineers, and various
It is not surprising that the elimination of specialists for structures, materials, manufactur-
human error, or more positively, the perfection of ing, and quality control.
the integration of man with the machine, is be- A minimum of three design reviews is recom-
coming an independent branch of engineering. It mended for each design:
would be beyond the scope of this book to go Preliminary design review.-This is pre-
into the details of "Human Engineering" or liminary review of work statements, basic
"Human Factors Analysis." Simply by applying concepts, schematics, layouts, and analyses.
It determines development needs and results the program, is very costly indeed. The rocket-
in decisions for the next design phases. This engine design can contribute substantially to
review should uncover misapplications, criti- prevent it.
cal areas and marginal designs at an early
stage, when changes can be effected without Cost
difficulty.
Critical design review.-This is the most Cost considerations should enter a design at
important review for decisions and approvals. the very beginning. Not only from the viewpoint
It includes formal review of all reliability of competitive bidding and narrow profit margins
aspects of layouts, analyses, planned devel- but of available and/or developed national re-
opment tests and procurement specifications. sources as well. A major rocket vehicle program
Final design review.-This is the final quite possibly may tax the resources of certain
review of overall design layouts, detail and materials, facilities, and services to the limit.
assembly drawings, analyses, process speci- Moreover, it should be ascertained that the pro-
fications, and R&D test results, before program does not rely on facilities, and on supply
duction release. of materials and of propellants that may not be-
The reviews should emphasize: come available for several years. The rocket
(1) Structural integrity engine under design may well be for a project
(2) Function and performance costing a total of hundreds of millions of dollars.
(3) Customer (vehicle) connections and en- Savings of even fractions of a percent can there-
velope fore amount to millions of dollars.
(4) Materials compatibility and component Cost, as far as the designer is concerned,
interfaces will be affected by the selection of materials; by
(5) Produceability and cost the machinery required to make the parts; by the
(6) Reliability and repeatability: malfunction time and skill required to make them; by the
effects difficulty of assembling, testing, and servicing
(7) Environment and servicing the parts; and last, but not least, by the experi-
(S) Special requirements ence of the people making the design.
It is highly recommended that the agreed-to
layout bear the approval signatures of the groups
Availability(Scheduling)
involved.
A reliability checklist tailored to the individ- The best design,the most perfectdevice,may
ual designs may be used as a guide during the be useless ifitis not availablewhen needed. In
various design review phases. The "Design a way, the design of (inour case) a rocketengine
Check-Off Sheet" shown in section 2-2 may be is a productby itself.Itmust be planned,pre-
used for this purpose. pared,and made. Itresultsin physicalproducts:
Design reviews are valuable and cost-saving drawings,writteninstructions, and procedures,
tools for improving reliability. A well organized which allmust be faultlessand availablewhen
and staffed review group will assure a meaning- needed. Delays in the releaseof drawings can
ful review and instigate effective recommenda- become very costlyindeed. In a manufacturer's
tions and actions. Complete documentation of overall plan of action, the design will be timed
all review details will provide valuable data for and synchronized with other activities. If the
future reference. drawings are not available on schedule, sizable
Under "Manrating" the cost of a single launch portions of a machine shop or an assembly plant
was mentioned. Press releases give the cost of may be forced to stand idle.
a major space program as about 3.5 billion dol-
lars a year-10 millions a day! To save a mis- 2.2 THE IMPORTANCE OF DESIGN QUALITY
sion, or just a single day of preparation, by
improved reliability, quite clearly, is worth the Even today many people still regard the
effort. design of technical objects as not much more
Likewise, the loss of a launch, or a day in than the act of putting pencil lines on large
pieces of paper. Of course, the general concept knowledge, analytical capabllltms, and
of design is much broader. Many a designer may judgment of the designer himself.
not even occupy a drawing board, since he may A number of time-proven tools are available
be completely devoted to the creation of basic to the designer to optimize his design and the
ideas and solutions, to the numerical checking of end product it describes. Among these are:
their validity, or to the planning, direction, and Design checkoff sheets.-Checkoff sheets
integration of a design team. By contrast, many force the designer to check his design sys-
of the most successful designers always retain tematically from all imaginable angles, before
their board, their most valuable tool of creation, release. See sample sheet on next page.
however advanced their position in the organiza- Design change checkoff sheets.-Used for
tion may be. systematic evaluation of design changes for
It is a commonplace to state that a technical
effects on all potentially affected systems,
project will stand or fall with the quality of its including the user and the vehicle. A sample
design. Obviously, no device can be built, sheet is shown below.
tested, or used until and unless it has been Failure reports.-The feedback to the de-
specified first how to build, test, and operate it. signer of failures of his product in the field,
With this in mind, a completed design can be for immediate corrective action.
considered a set of instructions for shop actions Change control.-A projectwide effort to
to follow. For instance, final drawings are only scrutinize and minimize proposed changes;
the end product of an extensive, diversified and to establish and monitor implementation
effort which preceded their release. All this is by engine serial number, if the change must
well known, yet sometimes forgotten in tim daily be made.
grind of a project. May it suffice to state that Specifications.-Detailed specifications,
the design and the designer are principal links of often based on applicable Government speci-
the chain in a project. fications, to establish the exact execution of
This book attempts to supply some of the all manufacturing processes, materials used,
necessary special tools required to arrive at a parts purchased, tests to be performed, and
detailed "set of instructions _ to the shop fore- SO on.
man, from which an engine could be actually

built and prepared for test. The following chap-
2.3 SYSTEMS ANALYSIS AND DESIGN
ters are devoted to the necessary technical de-
LAYOUT
tail, based on the latest state of the art in the
field of liquid propellant rocket engines. So far, numerous important basic considera-
There are numerous areas which the designer tions for rocket engine design have been dis-
should consider early and keep in mind at all cussed. The detailed engine design procedures
times, for overall increase of quality and relia- and the treatment of typical examples as they
bility, and for reduction of cost. Among these may occur in practice will now follow. The
are: questions are where to start, how to proceed,
(1) Experience, reliability, and reputation of and what are the expected results.
subcontractors and their products It would be beyond the scope of this book to
(2) Adequate receiving inspection describe the detailed mechanics of the genera-
(3) Clear and complete instructions for in- tion of design drawings, their breakdown, their
spection and quality control execution, and the system for their release. The
(4) Full use of the experience of others: young engineer, who uses this book and already
in-plant and out-of-plant works in industry, knows. The student reader
(5) Use of suitable existing designs, from can be assured that the future employer has a
small parts to subsystems manual (usually voluminous) covering this sub-
(6) Availability and application of an effec- ject exhaustively, and which is slanted to his
tive failure analysis and correction specific needs. The following are specific de-
system sign techniques as they are used in liquid rocket
(7) Last, but not least, the experience, engine design. Many of these techniques will be
LIQUID ROCKET ENGINE DESIGN CHECK-OFF SHEET
Project:
Subsystem: Supervisor:
Answer & initials Answer & initials

Item of demgner Date of supervisor Date
1. Have designs for similar earlier projects been

thoroughly reviewed and understood?
2. Have principal participants m those projects been

contacted? Has their advice been solicited?
3.. Has a list been prepared of all problems which

were encountered in previous, related projects,
including their solutions?
4 Has the new engine system schematic diagram

been reviewed for hidden "won't works?"
5. Has the number of components and their complexity

been reduced to a minimum, without loss in flexi-
bility and serviceability, for maximum reliability?
(In particular: has minimum of moving parts been
achieved?)
6 tlas a thorough malfunction analysis been made?

(Assessment of malfunction effects of each com-
ponent on all other parts and on the complete
system. Include external systems, such as the
vehicle and GSE.)
7. Have all existing detail designs been reviewed for

possible inclusion and'or adaptation to the new
design?
8. Have standard parts been used wherever possible;

permissible: or as prescribed by customer
specifications?
9. Has the number of external connections ("customer

connections") been held to a minimum?
10. tlas the locatmn and type of customer connections

been chosen in the best interest of the customer?
Has he been consulted?
11. Has the need for ground support equipment (GSE)

been reduced to a minimum, by number as well as
by complexity '_
12 Are there good reasons it support eqmpment used

during R&D is different from the one supplied to
the field?
13. Have all environmental conditions been consid-

ered, including those not likely but possible to
occur? (Salt spray, sand, fungus, humidity,
temperature, etc.)
14 Has resistance to vibration and shock effects been

considered? In all planes? Including improper
handling?
15. Have acceleration effects been considered? In all

planes?
B r
LIQUID ROCKET ENGINE DESIGN CHECK-OFF SHEET--Continued
Answer & initials Answer & initials

Item of designer Date of supervisor Date
16. Has the use of criticalmaterials been held to a

minimum? (Chromium, Molybdenum, Tungsten,
Cobalt, etc.?)
17. Has it been made absolutely certain that no

cheaper materials will do?
18. Has it been assured that no electrolytic action can

occur due to attachment of dissimilar metals?
19. Can the engines be gimbaled according to specifi-

cations?
20. Can the engine be clustered, if necessary?
21. Can the engine be attached to air frames other than

the one presently contemplated?
22. Has it been confirmed that all parts can be made?
23. Has it been confirmed that there is no cheaper way

of making these parts?
24. Can the parts readily be assembled, with a mini-

mum of special tools?
25. Has it been made impossible to incorrectly assem-

ble and install any part? (Or incorrectly reas-
semble and reinstallthem in the field?)
26. Will all parts requiring service be readily

accessible, prior to and following mating of the
engine with the vehicle?
27. Have all markings been called out completely and

correctly?
28. Do instructions for inspection and quality control

leave no gap?
29. Has it been ascertained that (without penalty) the

design cannot save further weight?
30. Is the envelope the smallest possible?
31. Can the system be drained, readily and completely?

(Avoidance of traps, low spots, etc.)
32. Have engine propellant feed system components,

including pumps and thrust chamber been de-
signed for minimum trapped propellants after
cutoff? (Minimum wet weight)
33. Has the purchasing department been appraised of

the significance of mandatory (proven) sources,
where applicable?
34. Can it be transported? In one or several pieces?

LIQUID ROCKET ENGINE DESIGN CHANGE CHECK-OFF SHEET
Project:
Change No.: Supervisor:
Item I Initials of designer I Date l Initials of supervisor I Date
Assuming that it has been ascertained beyond doubt (has itg) that a design change must be made, check that the effect
of this change is fully understood and implemented in the following areas:
1. Interchangeability of changed parts•
2,. "Chain Reaction" to other parts and systems, in-

cluding vehicle, test facilities, and GSE.
3. Engine start and stop sequence,
4. Engine performance, including gain factors.
5 Conformance with all specifications.
6. Weight, envelope, e.g., moments of inertia, loads
7. Reliability.
8. Instrumentation, data recording, telemetry.
9. Packaging, shipping, storage.
10. Development and qualification.
11. Human factors: skills, training, notification of

persons concerned.
12. Field servme equipment and procedures•
13 Handbooks and other user documents.
14. Logistics (spare parts maintenance).
15. Manufacturing tooling and processes.
16. Purchased materials.
t7. Subcontractor-supplied parrs.
18. Funding.
19 Agreed-to dehvery dates.
20. Customer notification and approval.
applied in subsequent chapters, in connection nent in relation to the other as a function of time
with the discussion and demonstration of various (fig. 2-11).
component designs. Concurrently, analyses will have been con-
The activities discussed below are not clearly ducted to establish preliminary engine perform-
separated phases, following one another in a ance parameters and operating characteristics,
rigid sequence. Rather, they overlap, frequently as well as individual component configurations
occur in parallel, and are tightly interwoven. and operating principles. A typical example of a
One of the first drawings the liquid-rocket de- performance diagram is shown in figure 3-1.
signer will most likely prepare is a schematic Engine and component starting and operating
diagram of the engine system. A typical example characteristics can be analytically predicted
is shown in figure 2-10. This diagram shows with a high degree of accuracy by computers.
how the principal components are linked together. Thus, important knowledge required for optimum
The schematic may be accompanied by a sequence design is obtained long before the part is actually
diagram, showing the operation of each compo- built and tested. Savings in time and cost are
_am.-
- ---- _ • m,,L_
:-_uuto _ecK ¸ _ _,_c_ AOOBTI_
_,Jeu _hJEct1_m - ', o
Figure 2-10.-Typical engine system schematic diagram.
START CUTOFF
I: ....
_dECONDS SECONDS
Figure 2-I l.-Typical engine system sequence diagram.
substantial. These analyses and computer pro- Consultations with specialists in their fields and
grams will draw heavily upon experience with rigorous calculations doublechecked for accuracy
earlier systems and on advanced design studies. and presented in a readily understood form, will
Once _he prototype schematic diagram is con- contribute materially. Layout drawings should
sidered completed, and performance parameters be made accurately and to true scale. In this
have been established by the analysis, the first manner, all components of the engine systemwill
actual "engine picture" will be drawn. This very have been designed and optimized for one
likely will be a general, in-scale preliminary another, rather than "hung on a mounting frame
layout of the engine system and components. A in Christmas-tree fashion."
typical example of an engine system preliminary Before working drawings are made from the
layout is shown in figure 3-2. layouts, they should again be reviewed, and
Through continued analyses, calculations, revised as necessary, taking into consideration
consultations, and joint reviews by all partici- all design aspects, the basic considerations, and
pants, the layout will gradually take final shape. the reliability assurance aspects, which were
In this process, nothing should be left to chance. discussed earlier.
2.4 STRESS ANALYSIS Analysis of Working Loads and Environmental

Effects
Stringent reliability and weight requirements

In stress analyses for mechanical design,
call for rigorous and complete stress analyses working loads and environmental effects should
during the mechanical design of engine compo-
be considered jointly. Stresses and strains in-
nents. Stress analyses predict the manner in
duced in mechanical parts by external forces,
which a mechanical part is likely to fail under
which we will call loads, are affected signifi-
anticipated working conditions. They also gen-
cantly by environmental effects such as tempera-
erate means to prevent failure. The goal is to ture, chemical reactions, corrosion, etc. Further-
design a part with sufficient, but not excessive,
more, the mechanical properties of most materials
strength in every detail.
are affected by temperature. Sometimes, thermal
In conjunction with the engine design layouts, stresses are induced as a result of temperature
the approximate shape of the parts will be estab- gradients within the part. Chemical reactions or
lished, based on functional requirements and on corrosion can change the mechanical properties
similar satisfactory designs of the past. First, a of the material, as well as the size and shape of
preliminary design sketch is made. Subsequently, the part.
the determination of probable working loads, For the analysis of working loads and envi-
environmental effects, deflections, stresses, and ronmental effects, determination of the following
determination of the final dimensions is made is essential:
step by step, together with the selection of ma- (1) The type of load: constant, impact, or
terials. The following steps are typical for repeated
stress analyses: (2) The maximum value and duration of a con-
(1) Analyze and determine the loads and en- stant load; the maximum and the mini-
vironmental effects to be expected dur- mum value for repeated or varying loads
ing the useful life of the part. (3) The nature of load application: concen-
(2) Evaluate the various possible modes of trated or uniform; rate per unit of time
part failure from stress and strain in- and, for repeated loads, total number of
duced within the part by the working working cycles
loads and from other effects.
(4) Vibration load effects
(3) Select the materials and establish their (5) Load effects with respect to the nature of
mechanical properties under anticipated material: ductile or brittle
working conditions. Applying a likely (6) Load effects with respect to the shape of
theory of failure, arrive at the final a part: effect of geometry on stress
dimensions of the part. concentration
(4) Apply experimental stress analyses to (7) Temperature effects: thermal stresses,
refine the above procedure. Redesign high-temperature creep and reduction of
the part if necessary, depending on the strength, low-temperature embrittlement
results. (8) Chemical reaction or corrosion effects:
embrittlement, stress concentration
In certain cases, such as with highly stressed
lightweight members, further stress analysis re- A part will have a proper margin of safety if
finement must be obtained. The greater the it is designed with a design limit load larger
refinement desired, the more nearly the methods than the maximum expected working load. The
of stress analysis must indicate the true strength design limit load in turn should be smaller than
of the member. This requires consideration of the calculated damaging loads because of the
complex states of stress, residual stresses, uncertainty and inaccuracy involved in stress
stress concentrations, dynamic effects, inelastic analyses. Damaging loads include: endurance
effects, and other influences which determine the limit load, yield load, and ultimate load, which
true stresses within the member. The following are defined below. The more accurate the analy-
is a discussion of the four steps of stress analysis, the smaller the allowable margin between
sis enumerated above. the design limit loads and damaging loads.
Below are given typical recommended criteria When a part is subjected to an indefinite num-
for the working loads, the design limit loads, and ber of cycles during service life, such as in
the damaging loads (yield and ultimated loads). rotating machinery, the endurance limit of a mate-
The proof-testing loads applied to component rial should be applied instead of the ultimate
design are also defined. strength. The endurance limit is the stress
(1) Design limit load: Select the largest of which can be repeated an infinite number of times
the following: without causing failure of the material from pro-
gressive fracture or fatigue. The endurance limit
1.2 x load (A) of metals, depending largely on range of stress
1.2 ×load (B) variation, is as low as between 20 to 60 percent
(2-8)
1.1 ×load (C) of their ultimate strength in tension. An addi-
1.0 x load (D) tional design margin of safety should also be
allowed for dynamic impact loads. When the
where shape of a part changes abruptly, as with a
Load (A)= Working load under normal steady groove, a notch, a hole, or where a small section
operating conditions joins a large one, the value of unit stress at
Load (B)= Working load under normal transient points close to the abrupt change or discontinuity
operation conditions, such as during normal increases steeply. The amount of stress in-
engine start and stop crease generally ranges from 100 to 300 percent
Load (C)= Working load under occasional tran- of the mean stress in the section.
sient operation condition, such as load during
irregular starts
Sample Calculation 2-2
Load (D)= Mandatory malfunction load which
must be taken into account. For example, in Tim hydraulic accumulator of a large liquid-
a clustered engine configuration, certain propellant rocket engine has the following design
mount members may carry the greatest load parameters: (a) Required volume (fluid capacity),
when one engine ceases to fire while the 7238 cu in.; (b) working pressure (load) under
others are still operating (engine-out capabil- normal steady and transient operating conditions,
ity). In certain instances it is mandatory that 2000 psia; (c) occasional surge pressure, 2200
an individual rocket engine continue to oper- psia; _) mandatory malfunction pressure, 2450
ate when a given component fails. If this ps!a; (e) maximum ambient temperature, 300 ° F;
causes significant structural loads, these are (f_) material selected, AISI 4340 H.T.-180.
considered mandatory malfunction loads. (Strength at room temperature: Ultimate, 185000
(2) Yield load = 1.1 × design limit load (2-9) psi; yield, 170000 psi. Strength at 300 ° F:
Yield load is the load which will induce a Ultimate, 178000 psi; yield, 150000 psia.)
stress equal to the yield strength of the Determine the following: (a_) Lightest possi-
material used under rated ambient con- ble configuration and resulting dimensions; _)
ditions. required proof test pressure at room temperature.
(3) Ultimate load = 1.5 × design limit load (2-10)
Ultimate load is the load which will in-
Solution
duce a stress equal to the ultimate
strength of the material used under rated
ambient conditions. (a) Since a sphere is the lightest pressure
(4) Proof test load = vessel for a given volume and pressure, we will
1.0 x design limit load (2-11) use this configuration. For a 7238-cu-in. volume
Proof test load is the load which is ap-
plied to test the part during the accept- Required inside diameter of the sphere
ance inspection. Its value can be ad-
justed for material properties if the rated =_ =_/_ × 7238
ambient conditions cannot be duplicated
for the test. = 24 inch
From equation (2-8), design limit pressure : a material failure,it may cause a part to perform
largest of the following: improperly with resulting malfunction of a com-
ponent or system. The other two-plastic defor-
1.2 x 2000 = 2400 psia; mation and fracture-are material failures in-
1.1 x 2200=2420 psia; fluenced by material properties, load and
1.0 × 2450 = 2450 psia environmental conditions, and by the shape of
Selected: 2450 psia the part.
Each of the three failure modes is character-
From equation (2-9), yield pressure = 1.1
ized by certain criteria. For elastic deflection,
× 2450= 2695 psia
strain is the criterion. For plastic deformation
and fracture, the criterion is stress. In the proc-
Thickness of sphere wall
ess of stress analyses, following load determina-
Yield pressure × diameter of sphere tion, tilepossible modes of failure of the part can
Z
4 x yield strength at 300 ° F be established in relation to the criteriainduced

by the loads. Failure cause can thus be deter-
2695 x 24
mined, and the failure prevented through design
- 4"_ _--_0---_-0 = 0.108 inch
changes. Some of the possible combinations of
failure modes and criteriaare listed in table 2-3
or from equation (2-10):
together with suggested design remedies.
Ultimate pressure = 1.5 x 2450 = 3675 psia
Selection of Materials and Dimensions

Thickness of sphere wall
For the process of finalizing the dimensions
Ultimate pressure x diameter of sphere
of a part to endure all working loads and envi-
4 × ultimate strength at 300 ° F
ronmental conditions without failure, the strength
3675x24 or ability of the selected material to withstand
- 4"_. _ 8-'_'0 : 0.124 inch these loads must be known. Material properties
are determined through materials tests conducted
We will use the higher value 0.124 inch. with specimens. In these tests, all conceivable
Therefore, the sphere dimensions = 24-inch loads such as tension, compression, torsion, and
inside diameter x 0.124-inch wall thickness. shear are applied, often with simultaneous appli-
(b) From equation (2-11), nominal proof test cation of temperature, vibration, or chemical
pressure at 300°F environment. The results are compiled in graphs

and tables. From these tables, materials with
: Design limit pressure properties most suitable for a particular applica-

tion can be selected.
= 2450 psia
Proof test pressure corrected for room temper- Experimental Stress Analyses
ature conditions:
A rocket engine part may be of such shape, or
may be loaded in such a way, that design based
Yield strength at room temperature
=2450x on theoretical analysis alone is difficult and
Yield strength at 300 ° F
unreliable. In such cases experimental stress
analyses can supplement the theoretical meth-
= 2450 x 170000= 2780 psig
150 000 ods. Many recent advances in stress analysis
can be attributed to the development of effective
experimental methods.
Evaluation of Failure Modes
Applying loads simulating as closely as pos-
There are three basic types of failure modes: sible those expected to occur in actual use,
elastic deflection, permanent plastic deforma- measurements of strains and stresses are made.
tion, and fracture. Although the first type is not These loads can be applied to full-size prototype
(
TABLE 2-3.-Failure Modes and Their Criteria
Failure mode Conditions Criteria Design remedies
1. Elastic deflection:
a. Stable equilibrium... Loads within elastic limits Strain; linear or angular dis- Change of shape or dimen-
placement (stretch or bend- sions (stiffening); material
ing) selection
b. Unstable equilibrium. Loads within elastic limits Buckling: ratio of applied Change of shape or dimen-
vs. critical load sions
c. Vibration ........... Within elastic limits: abrupt Amplitude, frequency trans- Stiffening; change of natural
changes of loads; repeated missibility, resonance frequency; damping
application of load at or
near natural frequency
2. Plastic deformation:
a. Yield .............. Loads exceed elastic limits Stress;permanentset Change of dimensions and 'or
material
b. Creep .............. Loads may or may not exceed Stress;slow permanentset Change of dimensions and or
elastic limits; elevated material
temperatures
3. Fracture:
a. Overload ........... Load increase beyond yield Stress:elongation;area Change of dimensions and 'or
point to.ultimate strength reduction;rupture material
b. Brittleness ......... Load above ultimate strength Stress;rupturewith little or Change of dimensions and. or
no yield material: change of heat
treatment; change of con-
tour
c. Impact or shock .... Abrupt load application to Stress; behavior like brittle Selectionof most duclile
ductile materials materials material:increasedmar,'.4'In
of safety
d. Fatigue ............ Many repeated load applica- Stress; number of load appli- Change of shape and dimen-
tions within elastic limits cations sions: change of material;
increase of endurance lilni[s
parts, to scale models made from the real mate- of the extreme temperature conditions in
rial or from special plastic material, or to por- liquid rocket engines
tions of full-scale parts. Not infrequently, (3) Required physical and chemical proper-
applied loads are intentionally increased beyond ties; density, thermal conductivity,
rated levels, until failure of the part occurs. specific heat, coefficient of expansion,
These "tests to failure" will establish the actual Poisson's ratio, strength-to-weight ratio,
margin of safety achieved in the design. corrosion resistance, compatibility with
The tools used in experimental stress analy- propellants as a function of temperature
ses include electrical, mechanical, and optical (4) Considerations related to fabrication,
strain gages; photoelastic plastic models, lac- such as forgeability, castability, weld-
quers, and paints. ability, machinability, and formability
(5) Cost and availability
(6) Existing industry and Government
2.5 SELECTION OF MATERIALS
standards
The selection of the most suitable material Extreme temperature and corrosion conditions
for a given rocket engine part will be governed combined with the need for very-high-strength-to-
by- weight ratios drastically narrow the choice of
(l) The function, size, and shape of the part available materials. In particular, the extremely
(2) Required mechanical properties; strength, low temperatures encountered with cryogenic
stiffness or rigidity, hardness, and liquids have introduced serious materials prob-
ductility, with particular consideration lems. As a rule, tensile strength and yield
strength increase with decreasing temperature. Resistance to Chemical Reactions

However, ductility is seriously affected. Apart While low-temperature hydrogen is considered
from selecting the best alloys for extremely low essentially noncorrosive, gaseous hydrogen forms
temperatures, highest purity of the metals is hydrides with some metals, such as uranium and
mandatory. titanium.
The increased usage of liquid hydrogen has For up-to-date detail on material properties,
introduced additional problems which further the reader is referred to material handbooks,
narrow the selection of available materials. industry (material supplier) information, and
Specifically for the LH: application, metals must Government publications (Bureau of Standards).
exhibit: The principal groups of materials used for
liquid rocket engines are as follows:
Resistance to Low-Temperature Embrittlement
This can also be referred to as toughness or
resistance to brittle fracture. Toughness, in Low-Alloy Steels
general, describes resistance to fracture under Uses for rocket engine components include
shock-type loads and stresses. Most rocket pins, bolts, shafts, brackets, mounts, thrust
engine parts are exposed to these loads; how- chamber structure members, ducts, injector bod-
ever, conditions are much more severe at the
ies, and certain pressure vessels. The standard
extremely low liquid-hydrogen temperatures. The grades AISI 4130, AISI 4140, AISI 4340, MAS
tendency of various alloys to brittle failure is 6434 are prominent in this group of steels.
measured experimentally by the notched/unnotched
The low-alloy steels are normally used in the
tensile strength ratio. Typically, for 2014-T6
temperature range from -60 ° F to 600 ° F. They
aluminum at -423 ° F, the ratio is 0.94 (longitudi- are not suitable for corrosive environments.
nal) and 0.83 (transverse). Elevated temperatures produce excessive creep,
thus reducing the strength. Also, very low tem-
Resistance to Thermal Shock
peratures tend to induce brittleness in most of
This is a measure of a material's ability to
them. An exception is AMS (SAE 9310).
resist weakening or fracture as a result of sud-
den heating or cooling. The following properties
appear to be requirements for high thermal shock Austenitic Stainless Steels (3{}0 Series)
resistance: High tensile strength (Ftu), high
thermal conductivity (k), low modulus of elastic- Steels in this group possess the highest cor-
ity (E), and a low coefficient of thermal expan- rosion resistance in the family of stainless
steels and are highly qualified for storable liquid
sion (a). The ratio Ftuk/Ea provides a relative
measure of thermal shock resistance for com- propellant application. They are inherently
parison of different metals. Typical values are tough and well adapted for fabrication by deep
5 to 8 for stainless steel and 40 to 48 for drawing and other similar means. They can be
2014-T6 aluminum. welded easily, can be soldered by proper tech-
nique, and are well suited for machining and
Resistance to Hydrogen Embrittlement forming under normal conditions. Ordinary sand
Certain metals, such as steels and titanium castings, precision investment castings, and
alloys, have a tendency to embrittlement in a forgings can also be produced from these steels.
hydrogen atmosphere. This tendency is greatest They are widely employed in rocket engines
in the intermediate-temperature range, but dis- using cryogenic and storable propellants. Parts
appears at low and high temperatures, The such as regenerative-cooled thrust-chamber tubes
effect is often delayed until a critical hydrogen and manifolds,injectorbodies and domes, valve
concentration in the metal is reached when poppets and bodies,propellantducts and tanks
cracks start to appear, as a result of a marked are made from these steels.
decrease in ductility. Also, the embrittlement
effect decreases with increasing strain in the
Martensitic-TypeStainlessSteels(400 Series)
metal. Heat-treated steels are more susceptible
to it than annealed ones. The steelsin thisgroup are hardenable,in
which condition they exhibit their best mechani- spinning, stamping, and shape bending. Most of
cal as well as corrosion-resisting properties. them are also adaptable for forging, welding,
The thermal conductivity of these steels is low brazing, and soldering. Aluminum alloys can be
but still the best of the stainless-steel family. cast by all three common casting methods: sand,
They are specially suitable for hot working or permanent mold, and pressure die casting.
forging. Their cold-forming characteristics are Mechanical properties and workability of alumi-
fair. They are well suited for most applications num castings are excellent.
requiring high strength, hardness, and resistance Aluminum alloys are the most widely used
to abrasion, wet and dry erosion, and moderate materials in rocket engine construction except
corrosion. They are not suitable for cryogenic where elevated temperatures are encountered.
applications, because of brittleness and shock Typical applications are valve bodies and pop-
sensitivity under these conditions. They are pets, injector domes, propellant tanks and ducts,
used for turbopump ball bearings and shafts, pump housings, impellers and inducers, and
gears, valve actuators, and cams. structure mounts.
SemiausteniticStainlessSteels Magnesium Alloys
The steelsin thisgroup can be formed in the Magnesium alloys have found many applica-
soft stateand then precipitation hardened. They tions in rocket engines and vehicles because of
are intended foruse in parts requiringcorrosion their excellent strength-to-weight, fatigue and
resistance and high strength at operating temper- stiffness characteristics. These alloys are used
atures up to 800 ° F, and where such parts may to make pump housings, valve bodies, and struc-
require welding and soldering during fabrication. ture mounts and are available in sheets, rods,
However, the corrosion resistance of this type of and castings.
steel is not as good as that of the austenitic Magnesium sheet alloys can be formed at
stainless steels. Rocket engine component elevated temperatures. They are also suitable
parts, such as thrust chambers, pump shafts, for various machining processes. They can be
levers, brackets, bellows, ducts, springs, clamp joined by fusion and resistance welding as well
rings, valve poppets, housings, and pressure as by adhesive bonding. Magnesium alloys can
vessels, have been made from the steels of this be cast by all three common casting methods:
group. sand, permanent mold, and pressure die casting.
Certain cast alloys can also be welded and heat
treated.
Aluminum Alloys
Pure metallic aluminum has a relatively low

High-Temperature Nickel-Base Alloys
strength. However, the strength can be greatly
increased by alloying aluminum with one or more The metals included in thisgroup are used
metals or metalloids. This can be accomplished primarilyfortheirstrengthat temperaturesup to
without affecting appreciably the other desirable 1700° F. The majorityof them contain aluminum
properties of aluminum, such as low weight, cor- or titaniumas precipitation-hardeningagents and
rosion resistance, ductility, good thermal and are vacuum melted. Their resistanceto oxida-
electric conductivity. tionand corrosionis excellent. These alloys
Wrought alloys of aluminum are generally of have found wide applicationin rocket engine
two types: one group that can be hardened by components such as: turbinehousings, wheels,
cold-working only (non-heat-treatable), such as and blades; thrustchamber tubes and injectors;
1100, 3003, 3004, 5050, and 5052 and a second gas generators;high-temperaturegas ducts,
group that will respond to both cold-working and bolts,and fasteners.
heat-treatment, such as 2011, 2014, 2017, 2024,
6061, 6066, and 7075. Wrought-aluminum alloys
Special Alloys
are suitable for fabrication processes such as
machining, shearing, drawing, stretch forming, The ever-present extreme temperature condi-
tions in liquid rockets calls for continued and Columbium.-This metal has been consid-
intensive materials research, particularly with ered for cryogenic application, but is liable
the advent of liquid hydrogen systems. In addi- to become embrittled at very low temperatures.
tion to the metals discussed in the preceding Titanium-base alloys.-These alloys have
paragraphs, other metals and alloys are receiving attracted considerable attention because of
increasing attention. Among these are: their high strength-to-density ratios, particu-
Copper base a11oys.-These metals exhibit larly at very low temperatures.
excellent ductility and toughness at very low
temperatures Typical representatives are Nonmetallic Materials
Berylco-10, -25 alloys, and Fe-Si bronze.
For gaskets, seals, lubricants, thread com-
Cobalt base a11oys.-The properties of
pounds, and the like, liquid rocket engines re-
these metals, such as Haynes-25, are similar
quire compatible nonmetallic materials. A great
to those of the nickel alloys.
variety of commercial products is available.
Tantalum.-Tantalum, when pure, has good In advanced LOX-pump designs, as well as in
properties at both low and elevated tempera- liquid-hydrogen pumps, the pumped fluid is used
tures. as the lubricant.
Chapter III
Introduction to Sample Calculations
3.1 APPROACH lant combinations, feed systems, and thrust

levels. In practice for logistics reasons, or to
It is a major goal of this book to familiarize
permit multiple use of parts, fewer combinations
the student with the detailed techniques used by
would be chosen. In fact, the student and the
industry in liquid propellant rocket engine sys-
teacher using this book may find it interesting
tems and component design. The authors feel
and instructive to modify the designs chosen by
that to convey a feeling for this subject effec-
the authors.l For instance, the student may wish
tively, nothing serves better than a set of real-
to determine what engine-design parameters
istic sample calculations. To promote a good
would result if stages A-2 and A-3 were to use
feeling for the interrelationship between major
the same propellant combination; or, what design
subsystems, the principal calculations were made
parameters would be obtained if stages A-3 and
for the engines of a hypothetical multistage
A-4 were combined into one, capable of restart
space vehicle. These calculations and their
and throttling to 30 percent nominal thrust.
associated designs were especially prepared for
It is not intended to suggest a specific mis-
this book and are not related to existing or
sion for the Alpha vehicle• However, a "primary
planned engines. As the various subsystems of
mission" for it could be the landing of an un-
liquid rocket engines are discussed in subse-
manned scientific payload on the Moon to gather
quent chapters, most of the supporting sample
samples and return them to Earth. The staging
calculations will be for the engines of that as-
sequence may then be as follows:
sumed vehicle, which thus will appear through-
Stage A-I: Boost to 250000-foot altitude.
out the entire book. For simplicity of reference,
Stage A-2: Boost to 300-nautical-mite alti-
the space vehicle will be called the "Alpha"
tude and inject into Earth parking orbit.
vehicle: it is assumed to be composed of four
Stage A-3: Accelerate to escape velocity and
stages: A-l, A-2, A-3, and A-4. Table 3-1 lists
inject into a translunar trajectory.
the major parameters of the Alpha vehicle.
Stage A-4:
The Alpha vehicle combination is realistic,
First start: Deceleration for lunar orbit
though not necessarily optimized. For instance,
and soft Moon landing of scientific payload
a different propellant combination has been
Second start: Moon takeoff for return to
chosen for each stage to permit sample calcula-
Earth
tions and designs for a number of typical propel-
In addition to its main powerplant, stage A-4
will require very-low-thrust attitude-control
TABLE 3.1.-4-Stage Space Alpha Vehicle
jets.
Takeoff weight, 2 100 000 lb; Even if designed for a given "primary mission,"
Payload a for 300-n.mi. orbit. 109 500 lb. a vehicle combination retains a certain degree of
Stage Number , Engine t Several good books on astronautics and space

Stage thrust, of thrust, Propellant dynamics are available from which student and teacher
lb engines lb may gather mission data for their own engine selec-
tion _tnd design. Among them is a little book by Dr.
A-1 ............. 3000000 4 750000 LO2/RP-1 Wernher yon Braun, "The Mars Project" (University of
A-2 ............. 600000 4 150000 LO2/LH 2 Illinois, Urbana, 1962, $0.95). Most of the calcula-
A-3 ............. 48000 3 16000 LF2/LH 2 tions in this book were made as early as 1948. yet are
A-4 ............. 15000 2 7500 N204/N2H 4 still fundamentally applicable. They appear ideally
suited for the reader to design his own up-to-date en-
a Consisting of stages 3 and 4, and of the mission pay- gine system for the manned Mars mission described,
load. for which all necessary vehicle data are presented.
63
flexibility. Within the limits of existing propel- ters, and through calculations and layouts, this
lant tank configurations, the following principal summarizing description can serve as a guide
possibilities of modification exist: and reference, throughout the book.
Omission of the upper two stages, for
Earth-orbital tankers, shuttle vehicles, space- 3.2 A-I STAGE ENGINE
station assembly, and supply ships.
Omission of stage A-4, for unmanned deep Four engines of 750 000-pound thrust each
space probe assignments, with no return in- were selected for a combined thrust of 3 million
tended. pounds (3000K), as a compromise between num-
Off-nominal tanking of one or more stages. ber of engine systems, and thus complexity on
This may yield some overall performance one hand, and flexibility on the other. Flexi-
gains for special missions. bility is offered through the possibility of includ-
It is emphatically not intended to say that the ing engine-out capabilities; of using existing
stated modifications can be made a few days smaller systems or designs; and for guidance
before launch. Rather, the stages and certain of and packaging considerations.
their subsystems, in particular the engines, The propellant combination of liquid oxygen and
should be regarded as building blocks. Their kerosene type RP-I fuel was selected for the A-I
availability can permit meeting a new require- engine. The selection was guided by the con-
ment within, for example, a year, as compared to sideration that high performance is not as criti-
several years when "starting from scratch." In cal for first booster stages as it is for upper
such ways, substantial gains have been obtained stages. Both propellants are abundantly avail-
in practice. The earlier Thor, and the Redstone able and comparatively inexpensive. The fluids
and Atlas Mercury boosters are well-known such and their combustion products are "docile"; their
cases. corrosivity is nil. Both fluids are relatively
Brief mention should be made here of a spe- dense. Liquid propellant rocket engine systems
cial type of system: experimental engine sys- using these propellants are well developed and
tems, sometimes referred to as breadboard en- reliable, and many "off the shelf" components
gines. Because of time and fund limitations, the and designs are available for them.
design and development of liquid rocket engines
for a given mission rarely permit the investiga-
General Engine System Description
tion of novel ideas and principles. New ideas
must then be tried out independently, detached The A-1 engine is a single-start, fixed-thrust,
from rigid schedules. Here the test effort can be gimbaled, bipropellant system. The fuel, RP-1,
conducted with full awareness that many of the is also used as the turbopump lubricant and as
principles under investigation will not "make the the engine control system actuating fluid. The
grade." However, while those that succeed can major components of the A-1 engine are a regen-
be applied to advanced operational systems, the eratively fuel-cooled, double-pass, tubular-wall
eliminated marginal ones are just as valuable, as thrust chamber with bolt-on injector; a direct-
they were early prevented from finding their way drive turbopump consisting of two centrifugal
into operational engines. If experimentally veri- pumps and a single-stage, two-wheel turbine; an
fied advances are selected for operational use uncooled gas generator with dual-ball valve;
with strong emphasis on vehicle application, butterfly main valves; and required controls.
true progress will have been made. The major The gas generator uses the same propellant com-
U.S. liquid propellant engine manufacturers have bination as the thrust chamber. Table 3-2 pre-
been conducting experimental engine programs sents all necessary operating parameters on
with excellent results for a number of years. which engine component designs will be based
The reader will now be acquainted with some for the A-1 engine system.
of the characteristics of the engines which have The A-1 engine system schematic diagram is
been selected for the different stages of the shown in figure 2-7. This diagram identifies
Alpha vehicle. While discussing and implement- clearly all major engine components and their
ing these in greater detail in subsequent chap- interconnecting plumbing. For the various
INTRODUCTION TO SAMPLE CALCULATIONS 65
TABLE 3-2.-750K A-1 Stage Engine Operating Parameters

[Sea-level conditions]
Engine (turbopump feed): Line pressure drop ................ psi ........... 10

Main valve pressure ............... psi ........... 15
Thrust .......................... lb ........ 750000
Nominal single-firing duration ...... sec .......... 165 Calibration orifice pressure drop .... psi .......... 110
Pump:
Specific impulse .................. sec ........ 252.4
Oxidizer LO2: Inlet pressure ................. psia .......... 45
Discharge pressure ............ psia ........ 1720
Flow rate ..................... lb/sec ..... 1967.7
Developed pump head ............. ft .......... 4790
Density ....................... lb/ft 3 ...... 71.38
Fuel RP-I: Pump:
Flow rate ..................... lb/sec ....... 892
Flow rate ..................... lb/see... 892.3
Density ....................... lb/ft 3 ...... 50.45 Shaft power ................... bhp ........ 11 790
Efficiency .................... Percent ..... 65.8
Mixture ratio ..................... O/F ........ 2.20
Shaft speed ................... rpm ......... 7000
Thrust chamber (tubular wall construction regeneratively

Turbine:
cooled by fuel):
Inlet pressure .................... psia ......... 640

Thrust .......................... lb ........ 747 000
Inlet temperature ................. °F ......... 1400
Specific impulse .................. sec .......... 270
Pressure ratio ................................ 23.7
Injector end pressure .............. psia ........ 1095
Gas flow rate .................... lb/sec ........ 92
Nozzle stagnation pressure ........ psia ........ 1000
Shaft power ...................... bhp ........ 27 140
Oxidizer flow rate ................ Ib/sec ...... 1941
Efficiency ....................... Percent ..... 58.2
Fuel flow rate .................... lb/sec ....... 827
Shaft speed ...................... rpm ......... 7000
Mixture ratio ..................... O/F ........ 2.35
Shaft torque ...................... in-lb ....... 20380
c* efficiency ..................... Percent ..... 97.5
c* .............................. ft/sec ...... 5660
Auxiliary drive:
Ct efficiency ..................... Percent ....... 98
Ct ......................................... 1.532 Shaft power ...................... bhp .......... 500
Contraction ratio ................. Ac/Ae ....... 1.6
Expansion ratio .................. Ae/A t ........ 14 Gas generator system:
Throat area A_ ................... in s .......... 487
Oxidizer side:
L* .............................. in ............ 45
Nozzle contour .......................... 80 percent Flow rate ..................... lb/sec ...... 26.7
bell Entrance loss ................. psi ........... 25
Line pressure drop ............. psi ........... 25
Oxidizer side:
Control-orifice pressure drop .... psi .......... 615
Valve pressure drop ............ psi ........... 10
Injector pressure drop ............. psi .......... 200
Injector pressure drop .......... psi ......... 120
Torus dome pressure drop .......... psi .......... 150
Fuel side:
Line pressure drop ................ psi ........... 25
Flow rate ..................... lb/sec ...... 65.3
Main valve pressure drop .......... psi ........... 35
Pump inlet pressure ............... psia .......... 55 Entrance loss ................ psi ........... 25
Line pressure drop ............. psi ........... 25
Pump discharge pressure .......... psia ........ 1505
Developed pump head ............. ft .......... 2930 Control-orifice pressure drop .... psi .......... 800
Pump: Valve pressure drop ............ psi ........... 20
Injector pressure drop .......... psi .......... 140
Flow rate ..................... lb/sec ...... 1971
Gas generator:
Shaft power ................... bhp ........ 14 850
Mixture ratio .................. O/F ....... 0.408
Efficiency .................... Percent ..... 70.7
Injector end pressure ........... psia ......... 710
Shaft speed ................... rpm ......... 7000
Heat exchanger ................... lb/sec ......... 3 Combustor pressure drop ........ psi ........... 70
Thrust vector control:

Fuel side:
Minimum acceleration ............. rad/sec: ....... 1
Injector pressure drop ............. psi .......... 200 Maximum velocity ................. deg/sec ....... 10
Jacket and manifold pressure drop., psi .......... 290 Displacement ................... deg .......... _q4
6G DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
MAIN OXIDIZER VALVE _P-_OXIDIZER INLET

[ PiN = 55 PStA
P= 1505 PSIA
LDX-DISCHARGE W= 1971 :¢/'SEC OXIDIZER PUMF
DUCT
WC,GO=27 #/SEC
INJECTOR ORIFICE_
_P I10 PSi /
P =1720 PSIA
FUEL- VI,/=892-'#/SEC FUEL PUMP
DISCHARGE
PINJ END = 109,5 PSIA
PIN=45 PSIA
PC = t000 PSIA MAIN FUEL =65 #/SEC
4-- FUEL INLET
VALV E
W= 92.'fk/SEC-
GENERATOR == 6"1.0 PSIA TURBINE
= 1400 ° F
TURBINE EXHAUST P=27PSIA
A-I
THRUST CHAMBER 750K First Stage Engine
Fuel pump
Developed head 4790 ft

Flow rate 892 lb/sec
Efficiency 65.8%
Nominal engine parameters Horsepower 11 790 bhp
Speed 7000 rpm
Propellants:
Liquid oxygen density 71.38 lb/ft 3 Oxidizer pump
RP-1 density 50.45 lb/ft 3
Thrust (sea level) 750000 lb Developed head 2930 ft
Specific impulse 262.4 sec Flow rate 1971 lb/see

Mixture ratio 2.20 Efficiency 70.7%
Horsepower 14 850 bhp
Thrust chamber Speed 7000 rpm
Expansion area ratio 14 Turbine

Throat area 487 in 2
Thrust 747 000 lb Pin 640 psia
Specific impulse 270 sec Pressure ratio 23.7
Mixture ratio 2.35 TemPin 1400 ° F
Flow rates: Efficiency 58.2%

oxidizer 1941 lb ' sec Horsepower 27 140 bhp
fuel 827 lb sec Exhaust thrust 3000 lb
Figure 3-1.-A-I engine performance diagram.

phases of engine design and development, it has chamber. All other components are either mounted
been found useful to work from an "engine per- on these two assemblies, or are located in the
formance diagram." This is a combination of the plumbing system between them. This arrange-
basic engine schematic and the principal per- ment permits engine gimbaling without punlp-
formance parameters. The A-1 engine diagram is discharge high-pressure propellant-duct flexure.
shown in figure 3-1. It is suggested that the Rather, thrust-vector control is achieved by
reader prepare his own performance diagrams for gimbaling the entire engine. The engine weighs
the other three stages. approximately 7500 pounds dry, 7900 pounds wet,
and 7830 pounds at burnout. The preliminary
For simplicity of mounting and compactness, design layout of the A-1 engine system and its
the turbopump is attached directly to the thrust overall dimensions are shown in figure 3-2,
1555
42,0 _ CHAMBER
FOX f_FUEL
._LCI"u ATOR ATTACH
: / DOME/" MANIFOLD
95.9
\ i G_MBAL_-..
I II_.-,BEAR,NG
/ _ | VEHICLE
I#/ / ATTACH/
OX INLET --,//
GENERATOR
FLEXBILE PROPELLANT i 6 VALVE
DISCHARGE DUCTS _ ,3.1 --4o.5--_,\\ iNLET8 OUTLET

' \LFUEL INLET
OX PUMP _FUEL PUMP
Figure 3-2.-A-1 750K first stage engine system preliminary layout.
System Operation spark plugs. At this point, the main oxidizer

valve remains closed, since the valve spring
The starting method of the A-1 engine is the
prevents opening until substantially higher
"main-tank-head start," combined with pressure
actuating fuel pressures are developed by the
ladder sequence (figs. 2-7 and 2-8). Propellants
pump later in the sequence.
are used under vehicle-tank-head pressures to
The low initial gas generator power level is
initiategas generator operation. As the turbo-
sufficient to start and accelerate the turbopump.
pump starts to accelerate, main propellant pres-
This, in turn, increases the propellant pressures
sures "bootstrap" the system to mainstage level.
available to the gas generator, which is con-
nected upstream of the still closed main valves.
Starting Sequence
As a result, the gas generator-turbine system
After all vehicle launch preparations are com- "bootstraps" itself.
pleted and the main propellant tanks are pres- At a predetermined fuel pump outlet pressure
surized, the gas generator dual sparkplugs are (valve spring rate selection), the main oxidizer
activated upon a given signal and the engine valve opens. During the oxidizer valve stroke,
four-way control valve is energized to open. The an integral and mechanically linked igniter fuel
closing side is vented and fuel pump outlet pres- sequence valve is actuated. The sequence valve
sure (50 psia from tank head) is admitted to the in turn admits fuel pressure to the hypergol
opening side of the main oxidizer and gas gener- igniter cartridge, rupturing its diaphragms.
ator valve actuators. The gas generator valves Hypergol (such as triethylaluminum) enters the
open and admit both propellants to the gas gen- combustion chamber igniter elements and ignites
erator combustor, where they are ignited by the with the oxidizer just being admitted by the main
oxidizer valve. Fuel following the hypergol equally true for liquid hydrogen. Both elements
sustains the igniter flame. are abundantly available. Their combustion
The chamber pressure resulting from igniter product-water vapor-is the most harmless of all
combustion is sufficient to actuate the ignition propellant combinations, solid or liquid. Telem-
monitor valve. This valve admits fuel pressure etry engineers like it because of the low attenua-
to the main fuel valve opening actuator, opening tion effects of the exhaust gases on ItF signals,
the valve, which is the last step in the sequence. an important aspect for vehicle guidance and
Main fuel enters the combustion chamber; cham- telemetry. Most important, the yield in specific
ber pressure and thrust climb to the rated level. impulse of this combination is close to the theo-
retical maximum for chemical reactions. Only
Cutoff Sequence certain fluorine hydrogen combinations are
slightly higher (approximately 4 percent). How-
Upon receipt of a cutoff signal, the engine
ever, the extreme toxicity of fluorine and fluorine
four-way control valve is deenergized to close.
compounds, both as a liquid and as components
The opening pressures of all valves vent; the of combustion products, makes fluorine less
valves close. Turbine power and main chamber attractive for operational use. By building the
pressure decay. While the main propellant tanks
A-1 and A-2 stage a little larger, the same pay-
vent, all valves are held in their closed position load can be obtained as with fluorine. However,
by springs. Note that the gas generator valve all advantages of liquid oxygen and hydrogen are
and the main oxidizer valve are closed by ad- retained, including the great amount of available
mitting actuation pressure to their closing port, experience.
while the main fuel valve is closed by spring Hydrogen does have its drawbacks. The
force only. Experience has shown that cutoff principal one is low density, resulting in
precision is largely influenced by turbopump
rather bulky tanks. For a typical oxygen-to-
decay characteristics and by cessation of that
hydrogen mixture ratio of 5:1, the corresponding
propellant flow which has the smallest duct vol- volume ratio is inversed: 1:3. This disadvan-
time below the valve, in this case the oxidizer
tage, however, is successfully offset for upper
(no cooling jacket). The fast closing valves
stages by the high energy yield (Is). It can be
need some timing, however, from water-hammer
expected that the improved state of the art of
considerations.
ultralightweight vehicle and tank constructions
will further increase the superiority of hydrogen
3.3 A-2 STAGE ENGINE systems for upper stages.
A secondary effect of low density is a high
For the same reasons as with the A-1 stage, boiloff rate, unless tanks and lines are properly
a four-engine cluster has been chosen, having a insulated. Because of the low density, and of
total vacuum thrust of 600 000 pounds (150 000 the resulting large surface area of the contain-
pounds per engine). Because of the substantial
ers, the heat input per unit hydrogen mass is
vehicle performance gains obtainable through the high. Furthermore, the temperature of liquid
use of high-energy propellants in upper booster hydrogen is sufficiently low to liquefy air on the
stages, liquid oxygen (LO2) and liquid hydrogen
tank surfaces. This sharply increases heat
(LH2) have been selected as propellants. For transfer rates, resulting in extreme boiloff rates.
decades this combination has attracted the atten-
Tank and line insulation, therefore, is vital.
tion of rocket experimenters and developers.
Although techniques of rocket vehicle insulation
However, only relatively recently has the art of
are highly developed, some weight penalties are
hydrogen engines reached maturity. Several incurred.
major systems are in active development or in
Overall, however, a substantial net perform-
early flight application.
ance gain can be obtained for upper stages.
Through the years, the production and handling
of the cryogenic propellant liquid oxygen (LO 2 or
"LOX') has become a routine matter; its price has
come down considerably. It can be reasonably
expected that in the near future this will be The A-2 engine is a single-start, fixed-thrust,
[Vacuum conditions]
Engine (turbopump feed): Shaft speed ................... rpm ......... 8600

Turbine:
Thrust .......................... lb ........ 150 000
lnlet pressure ................. psia ......... 700
Nominal single-firing duration ...... see .......... 250
Inlet temperature ............... °F ......... 1200
Specific impulse .................. see .......... 434
Pressure ratio ............................... 16
Oxidizer LO2:
Gas flow rate .................. ib/sec ...... 1.58
Flow rate ..................... lb/sec ....... 288
Shaft power ................... bhp ......... 1940
Density ....................... lb/fi 3 ...... 71.38
Efficiency .................... Percent ..... 543
Fuel LH2:
Shaft speed ................... rpm ......... 8600
Flow rate ..................... lb/sec ...... 57.6
Shaft torque ................... in-lb ....... 14200
Density ....................... lb/fi 3 ....... 4.42
Mixture ratio ..................... O/F ........... 5
Fuel side:
Thrust chamber (tubular wall construction regeneratively Injector pressure drop ............. psi .......... 100
cooled by fuel. Nozzle extension film Jacket and manifold pressure drop .. psi .......... 325
cooled by turbine exhaust gas): Line pressure drop ................ psi ........... 20
Thrust .......................... Ib ........ 149 500
Calibration orifice pressure drop .... psi ........... 60
Pump inlet pressure ............... psia .......... 25
Injector end pressure .............. psia ......... 875
Nozzle stagnation pressure ........ psia ......... 800
Developed pump head ............. ft ......... 44 800
Oxidizer flow rate ................ lb/sec ..... 285.2
Pump weight flow rate ............. Ib/sec ...... 59.8
Fuel flow rate .................... lb/sec ...... 54.5
Pump volumetric flow rate ......... gpm ........ 6080
Mixture ratio ..................... O/F ........ 5.22
Heat exchanger bleed ............. Ib/sec ....... 2.2
c* efficiency ..................... Percent ..... 97.5
Pump:
c* .............................. fi/sec ...... 7480
Shaft power ................... bhp ......... 6100
Ct efficiency ..................... Percent ...... 101
Efficiency .................... Percent ....... 80
Ct ......................................... 1.895
Shaft speed ................... rpm ........ 27 000
Contraction ratio ................. At At ...... 1.60
Chamber coolant passage bleed for
Expansion ratio .................. Ae/A t ........ 40
fuel tank pressurization .......... Ib/sec ....... 2.2
Throat area, At ................... in 2 ......... 98.6
Turbine;
L* .............................. in ............ 26
Inlet pressure ................. psia ......... 700
Nozzle contour .......................... 75 percent
Inlet temperature ............... °F ......... 1200
bell
Pressure ratio ............................... 16
Gas flow rate .................. lb/sec ...... 4.32
Oxidizer side:
Shaft power ................... bhp ......... 6100
Injector pressure drop ............. psi .......... 160 Efficiency .................... Percent ..... 62.5
Torus dome pressure drop .......... psi ........... 40 Shaft speed ................... rpm ........ 27000
Line pressure drop ................ psi ........... 20 Shaft torque ................... in-lb ....... 14 250
Calibration orifice pressure drop .... psi ........... 60 Tapoff gas from thrust chamber for turbine drive:
Pump inlet pressure ............... psia .......... 35
Pressure ........................ lb/in 2 ........ 750
Temperature ..................... °F ......... 1200
Developed pump head ............. ft .......... 2305
Weight flow rate .................. lb/sec ....... 5.9
Pump weight flow rate ............. lb/sec ..... 290,5
Mixture ratio ..................... O/F ........ 0.90
Pump volumetric flow rate ......... gpm ........ 1830
Heat exchanger bleed (oxidizer
Thrust vector control:
tank pressurization) ............. lb/sec ....... 2,5
Pump: Minimum acceleration ............. rad/sec 2 ....... 2
Shaft power ................... bhp ......... 1910 Maximum velocity ................. deg/sec ....... 15
Efficiency .................... Percent ....... 64 Displacement .................... deg ........... ±6
gimbaled, bipropellant system. The thrust cham- passage to pressurize the vehicle main fuel tank.
ber features a combination of fuel regenerative A small portion of liquid oxygen bled from the
cooling, and film cooling with turbine exhaust oxidizer pump discharge is heated in a heat ex-
gas. The chamber assembly is fed by two inde- changer and used for vehicle main oxidizer tank
pendent, direct-drive centrifugal turbopumps. pressurization.
For the fuel, an alternative axial pump may be A listing of the A-2 engine operating param-
chosen. Each operates at optimum speed. Hot eters is presented in table 3-3. The engine
gases are tapped off from the main combustion schematic diagram is shown in figure 3-3. Note
chamber to power the turbines. A hot gas orifice that engine parameters are based on vacuum con-
in the tapoff duct controls the engine thrust ditions. This is justified and customary, if a
level. The turbines are gas coupled in parallel. stage starts and performs in vacuum for its en-
Their exhaust gases are routed to the thrust tire duration. In the case of the A-2 stage, the
chamber and injected in the 30:1 expansion area starting altitude of 250 000 feet is not an abso-
ratio plane. Thus the gases provide film cooling lute vacuum, but for atl practical purposes may
for the nozzle portion from there to the 40:1 area be considered absolute.
ratio plane. The remainder of the chamber, above Thrust-vector control is achieved by gimbaling
the 30:1 plane, is regeneratively cooled (1!4 the entire engine. It weighs apploximately 2181
pass). Helium gas is used to actuate the con- pounds dry, 2317 pounds wet, and 2292 pounds at
trols. No lubricants or any other fluids are used burnout. The overall dimensions and the prelimi-
which could freeze at low temperatures. The nary design layout of the A-2 engine are shown
engine is started by the hot gases generated by in figure 3-4.
a solid-grain turbine spinner. Chlorine trifluoride
(C1F3), which is hypergolic with LH 2, is used to
System Operation
ignite the combustion chamber. Both turbine
spinner and igniter fluid are insulated and elec- The starting method of the A-2 engine (figs.
trically temperature conditioned from a ground 3-3 and 3-5) is a "turbine spin start," for very
source until first-stage takeoff. Gaseous hydro- fast buildup (less than 2 seconds from start sig-
gen is bled from the thrust chamber coolant nal to main stage). Starting power is furnished
FUEL.PUMP
OXIDIZER-PUMP INLET
INLET / /- OXIDIZER _ .._ FUEL PUMP
OXIDIZERTANK _._ PuMP

PRESSURANT-HEAT
EXCHANGER -----1 _ \_-// " II FUEL
TURBINE
HYPERGDL CARTRIDGE
AND MONITORINGVALVE
"r" I _ H TURBINESPINNER
MAINOXIDIZER
VALVE
PU CONTROL _'Jl HYPERGOL SEQUENCE

CHAMBER.GASTAPOFF ==_ VALVE
MANIFOLD
_ ,1 d_ '=='_ FUEL BLEEDVALVE
-\ I/ "_ ENGINE-CONTROL
4.WAYVALVE
_ \ _ RETURN TO
A,N
THRU.
CN. ER \ FUEL
TANKAIH
FOEVALVE
ANIFoLDTOR
EXHAUS. IHE ( .EL,U
TANK
Figure 3-3.-A-2 st3ge engine system schematic diagram.
THRUST CHAMBER-
FUEL .u
I
II 7?..5"
D
I
L i
EXHAUST
GAS MANIFOLD
Figure 3-4.-150K A-2 stage engine system preliminary layout.
START CUT OFF
E,._._.__'_'_TURBINE SPINNER IGNITES
i
ENGINE CONTROL 4-WAYVALVE SOLENOID ENGINE CONTROL 4-WAY VALVE SOLENOID
I'
ENERGIZED FOR OPEN CONTROL DE-ENERGIZED FOR CLOSE CONTROL
I
TU_BOPUMP BUILD UP MAIN OXIDIZER VALVE CLOSES
MAIN FUEL VALVE AND HYPERGOL
_\-_,,_\,__ MAIN FUEL VALVE AND HYPERGOL
SEQUENCE VALVE OPEN; FUEL BLEED _,'X_, SEQUENCE VALVE CLOSE FUEL
VALVE CLOSES. 1 BLEED VALVE OPEN
I I
_MAIN OX. VALVE OPENS =_--_-_:* _'*_ THRUST DECAYS
I
_ _
1_-__ THRUST BUILD UP
iii'
I
0 .4 .8 L2 1.6 2.0 0 .2 .4 .6 .8 I.O
Figure 3-5.-A-2 engine system sequence diagram.

by a turbine spinner. Chamber tapoff gases then 3.4 A-3 STAGE ENGINE
bootstrap the turbine and main stage operational
The totalthrustof 48000 pounds is sub-
level is established.
dividedintothree engines of 16 000-pound thrust
each. This effectsa shorteroverallpropulsion
Starting Sequence system and a shorter,lightervehicle interstage.
Furthermore, the liquid fluorine/liquid hydrogen
As part of the separation and staging se-
propellant combination was chosen because of
quence, a vehicle programer furnishes a start
the relatively stringent performance requirements
signal to the engine, which ignites the turbine
for upper stages.
spinner and supplies hot gases at 2000 ° F to
Fluorine is the most reactive and energetic
turbines and combustion chamber. This signal
chemical element. It has vigorous and reliable
also energizes the solenoid of the engine control
hypergolic ignition characteristics, and superior
valve, which vents the closing side of both main
specific impulse capabilities with most fuels.
propellant valve actuators and pressurizes the
The high density of liquid fluorine, combined
opening side of the fuel valve actuator with
with high performance with liquid hydrogen, re-
helium gas. Simultaneously, the hypergol se-
sults in maximum payloads. As mission require-
quence valve, which is mechanically linked to
ments become more ambitious, payload advan-
the main fuel valve, is opened and the actuator
tages from the fluorine-oxidized propellant
of the normally open fuel bleed valve is pres-
combination should compensate for handling
surized to close. The fuel flows through the
problems caused by fluorine toxicity and corro-
chamber cooling jacket under increasing pump
siveness. Past experience has indicated that
discharge pressure and injects into the combus-
the operation of a fluorine-oxidized engine is
tion chamber. Ignition is achieved by the hyper-
practical at this thrust level (50000 pounds or
golic reaction between hydrogen and the slug of less). Fluorine when used for gaseous passiva-
chlorine trifluoride forced into the chamber by tion of metals renders a metallic surface resist-
increased oxidizer-pump discharge pressure. ant to future chemical reaction. Thus, once a
When the main fuel valve reaches the 90-
metallic fluoride film is formed, further action by
percent open position, ports integral with the
the liquid fluorine is either prevented or signifi-
actuating piston open and permit helium gas to cantly retarded, making handling or storing of
flow through the hypergol monitor valve and to
liquid fluorine less of a problem. No known
pressurize the open side of the main oxidizer elastomer is completely compatible with fluorine;
valve actuator. The main oxidizer valve opens,
however, flow tests of liquid fluorine with Teflon
admitting oxidizer to the chamber where pressure
have given satisfactory results.
builds up rapidly. Chamber tapoff gases boot-
strap the turbines to main stage operation. The
spinner will burn for about 1.2 seconds. After
the main stage is achieved, the propellant-
utilization servo system will begin to function. The A-3 engine is a multiple-start, gimbaled,
bipropellant system. The basic system includes
a thrust chamber assembly using a combination
Cutoff Sequence
of fuel-film (LH2) and radiation cooling; propel-
The cutoff signal, received from the vehicle lant ducts; valves; and a control subsystem.
programer, deenergizes the engine control valve Ignition is achieved by the hypergolicity of the
causing it to close. This vents the open side propellants. These are fed directly from pres-
and pressurizes the closing side of the main pro- surized propellant tanks, through main propellant
pellant valve actuators. The main oxidizer valve valves, to the tl_rust chamber inlets. The pro-
is made to close faster than the main fuel valve, pellant tanks and their gas pressurization sys-
by proper orificing of the helium lines, to assure tem are considered part of the engine propellant
a fuel-rich cutoff. Engine thrust decays. The feed system. Gaseous helium supplied from a
fuel bleed valve opens after the helium pressure high-pressure helium bottle located inside the
in its actuator is vented. main fuel tank is used for main oxidizer tank
pressurization. The main fuel tank is pressur- highly reactive and toxic nature of fluorine.
ized by gaseous hydrogen supplied from a liquid Welded joints are used extensively. No rotating
hydrogen bottle which is pressurized by helium seals are employed. Sliding seals are of the
and which is also located inside the main fuel metal-bellows type.
tank. Both pressurants are heated in heat ex- A preliminary design layout of the A-3 propul-
changers located at the thrust chamber nozzle sion system and dimensions are shown in figure 3-7.
extensions before they are expanded through The fuel tank is pressure stabilized rather
pressure regulators and transferred to the propel- than mechanically stabilized. The thrust loads
lant tanks. Helium gas is used to operate the are transmitted to the payload through the fuel
main valves and the gimbal actuators, and to tank. Both tanks are insulated, as are the ducts
purge the propellant manifolds during start. between tanks and engine systems.
A-3 engine operating parameters, for vacuum Thrust-vector control is achieved by gimbal-
conditions, are presented in table 3-4. The pro- ing the thrust chambers. Each basic engine
pulsion system schematic diagram is shown in weighs approximately 330 pounds dry and 365
figure 3-6. pounds at burnout. It has a cylindrical space
The design of the entire propulsion system is envelope of 5 feet 4 inches diameter by 7 feet
governed by simplicity and minimum number of 6 inches length. The propulsion system (includ-
components. This is essential because of the ing the three engines and the tankage) weighs
OXIDIZER TANK
THRUST
FUEL TANK
CHAMBER
\
HELIUM SUPPLY TO PURGE AND ACTUATING SYSTEMS (NOT SHOWN)
TANK FILL AND DRAIN

,-_PRESSURE REGULATOR =_ PRESSURANT SHUT-OFF
VALVE VALVE
V TANK RELIEF VALVE MAIN PROPELLANT =O PRESSURANT FILL VENT

VALVE AND RELIEF VALVE
CHECK VALVE
Figure 3-6.-A-3 stage propulsion and engine system schematic diagram.

[Vacuum conditions]
Engine (pressurized gas-feed): Calibration orifice pressure drop .... psi ........... 17
Oxidizer tank pressure ............ psia ......... 170
Thrust .......................... lb ......... 16 000
Total oxidizer weight (300 sec
Nominal total multiple-firing
duration for 3 engines, plus
duration ........................ sec .......... 300
1 percent residual) .............. lb ......... 27950
Oxidizer tank volume (including
Oxidizer LF2:
3 percent ullage volume) ......... ft 3........... 305
Density ....................... lb 'ft 3 .... 94.16
Pressurant (helium) flow rate
Flow rate ..................... lb'sec... 80.78
(assuming tank gas temperature
Fuel LH2:
400 _ R) ........................ lb/sec .... 0.1555
Density ....................... lb/ft * .... 4.42
Total pressurant weight (including
Flow rate ..................... lb/sec ...... 5.13
other requirements in the system)
Mxxture ratio ..................... O/F ........... 6
(assume storage bottle final pres-
sure 350 psi, plus 2 percent
Thrust chamber (solid wall film cooled by fuel and radia-
reserve) ....................... lb ............ 60
tion cooled on nozzle extension):
Pressurant storage tank, volume
Thrust ......................... lb ......... 16 000 (assume 200 _ R storage temper-
Specific impulse .................. see .......... 446 ature, including 3 percent
Injector end pressure .............. psia ......... 110 ullage volume) .................. ft s .......... 7.35
Nozzle stagnation pressure ........ psia ......... 100 Pressurant storage tank, initial
Oxidizer flow rate ................ lb/sec ..... 30.78 pressure ....................... psia ........ 4500
Fuel flow rate .................... lb/sec ...... 5,13
Mixture ratio ..................... O/F ........... 6
c* efficiency ..................... Percent ....... 98 Fuel side (pressurized by heated hydrogen):
c* .............................. ft/sec ...... 7910
Injector pressure drop ............. psi ........... 25
CI efficiency ..................... Percent ...... 102
Inlet manifold pressure drop ........ psi ........... 10
Ct ......................................... 1.817
Contraction ratio ................. Ac/At ......... 2
Line pressure drop ................ psi ............ 5
Expansion ratio ................. Ae/At ........ 35
Fuel tank pressure ................ psia ......... 160
Throat area At ................... in 2 ........... 88
Total fuel weight (300 sec duration
L* .............................. in ............ 28
for 3 engines, plus 1 percent
Nozzle contour .......................... 70 percent
residual) ....................... lb .......... 4660
bell
Fuel tank volume (including 3 per-
Thrust vector control: cent ullage volume) .............. ft 3.......... 1087
Pressural_t (hydrogen) flow rate
Minimum acceleration ............. tad/sec 2 ....... 2
(assuming tank vapor temperature
Maximum velocity ................. deg/sec ....... 15
300 ° R) ........................ lb/sec ..... 0.346
Displacement .................... deg ........... ±7
Total pressurant weight (assuming
storage bottle final pressure 350
Oxidizer side (pressurized by heated helium):
psia, plus 4 percent reserve) ...... lb ........... 108
Injector pressure drop ............. psi ........... 25 Pressurant storage tank, volume
Oxidizer dome pressure drop ....... psi ............ 5 (liquid hydrogen including 3
Main valve pressure drop .......... psi ............ 8 percent ullage volume) ........... ft 3.......... 25.2
Line pressure drop ................ psi ............ 5 Pressurant storage tank, pressure.., psia ......... 350
approximately 5130 pounds dry, 37 900 pounds Figure 3-8 shows the operational sequence of the
wet, and 5530 pounds at burnout. A-3 stage engine. In conjunction with figure3-6,
this illustrates the system starting and shutdown
System Operation
operations.
The A-3 propulsion system is designed for
3.5 A-4 STAGE ENGINE
automatic start, on receipt, of a signal from the
guidance system. A similar signal effects auto- For the A-4 stage, two engines of 7500-
marie engine shutdown. One or more restarts can pound thrust each were selected, for a combined
be made by merely sending additional start and thrust of 15 000 pounds. It is assumed that the
shutdown signals to the propulsion system. mission assigned to this fourth and last stage of
(L H2)TANK
CHAMBERS (3)
[L F2) TANK
l
12'D
160 PSIA
170 PSIA
VECTOR
CONTROL GIMBAL I
ACTUATORS (6) J
l
i
-HIGH-PRESSURE
ITERNAL HELIUM BOTTLE
TANK INTERSTAGE CONNECTING SKIRT
INSULATION
LIQUID-HYDROGEN
PRESSURANT BOTTLE
Figure 3-7.-48K A-3 stage propulsion system preliminary design layout.
our space vehicle may require prolonged cruising lems. Although cryogenic propellants could
periods prior to ignition and possibly even longer probably be used with refined insulation tech-
waiting periods prior to reignition. While it niques, they were not selected because of the
would be desirable to utilize the high-energy systems complication for a vehicle of this size.
propellants of the second and third stages, the Solid propellants were also ruled out because of
fact that they are cryogenics poses some prob- the need for repeated starts and throttling.
START CUTOFF
I I .,'-CUTOFF SIGNAL
I IF" j
|
I
I_ PRESSURANT SHUTOFF VALVES OPEN k_._,_N]j PRESSURANT SHUTOFF
VALVES CLOSE
I ' I l
I_ THRUST CHAMBER PURGE I i_._-_= THRUST CHAMBER PURGE
! I I
I _'_'%."_\X_,'X_\_X_ PROPELLANT TANKS PRESSURIZED _TANK PRESSURES DECAY
I I I
] _ MAIN PROPELLANT VALVE CONTROL MAIN PROPELLANT VALVE CONTROL
_'_\'_"'"-'"_ SOLENOID VALVE OPEN & PURGE STOPS
_I SOLENOID VALVE CLOSES PURGE STARTS
I I_ MAIN PROPELLANT VALVES OPEN
, j-- START _l MAIN I_ROPELLANT VALVES CLOSE
I
_' SIGNAL
I _ TH_ST BU_LD-UP
I i THRUST DECAYS
I I I I I I I 1 I
0 I 2 3 4 5 0 I 2 3
TIME, SECONDS TIME, SECONDS
Figure 3-8.-A-3 stage engine and propulsion system operational sequence.

A hypergolic, storable propellant combination before expansion through a pressure regulator

possesses certain characteristics which contrib- and transfer to the propellant tanks. Helium gas
ute to high reliability. Among these are simplic- is also used to operate the main valves and the
ity of ignition and ease of propellant mainte- gimbal actuators.
nance, since tile propellants can be contained in A-4 engine operating parameters at vacuum
closed vessels over reasonable temperature condition are presented in table 3-5. The engine
ranges for considerable periods of time without and propulsion system schematic diagram is
developing excessively high pressures, or under- shown in figure 3-9.
going unacceptable changes in composition. The engine gimbal blocks are fastened to
Among the applicable storable propellant thrust mounts which are attached to the aft end
combinations with high performance are chlorine of the oxidizer tank. The fuel tank is attached
trifluoride (C1F3)/hydrazine (N2H4) and nitrogen forward of the oxidizer tank to form an integral
tetroxide (N20,)/hydrazine. Hydrazine, as a vehicle structure. As in the A-3 system, the
monopropellant, is prone to explosive thermal thrust loads are transmitted to the payload
decomposition. However, the condition can be through the pressure-stabilized tank assembly.
remedied by certain additives. The C1F3/N2H 4 The propellant ducts between fuel tank and en-
combination has slightly higher performance than gine systems are routed outboard and covered by
the N204/N2H4 combination. Handling of C1F3, fairings for protection against aerodynamic heat-
however, requires special design provisions being and for lower air resistance during first-stage
cause of its thermal characteristics. For this boost.
reason, the N204/N2H4 propellant combination Both throttling and propellant-utilization con-
was chosen for the A-4 engine. It is worthy of trol are achieved by varying the degree of open-
note that the performance of N204/N2114 is coming of both propellant valves. The positions of
parable to that of LO2/RP-1. the valves are controlled by the vehicle guidance
Teflon and Teflon 100X can be used as seal system in conjunction with a vehicle propellant
material in the A-4 engine system. Kel-F, while quantity measuring system. Thrust vector con-
a satisfactory material for use with N2ll4, de- trol is accomplished by gimbaling the thrust
grades after short-term service in N204. Most chambers. The basic single engine weighs ap-
series 300 stainless steels, aluminum alloys, proximately 150 pounds dry and 170 pounds at
nickel, and nickel-base brazing alloys can be burnout. It has a cylindrical space envelope of
used as construction materials. 3 feet 6 inches diameter by 5 feet 9 inches
length. The complete propulsion system (includ-
General Engine System Description ing the two engines and the tanks) weighs approx-
imately 725 pounds dry, 19 649 pounds wet, and
The A-4 engine is a multiple-start, variable- 795 pounds at burnout. The preliminary design
thrust, gimbaled, bipropellant system. The basic layout of the A-4 propulsion system is shown in
system includes a thrust chamber assembly uti- figure 3-10.
lizing combined ablative and radiation cooling,
Note that for the A-3 and A-4 engines a
propellant ducts, valves, and control subsys-
slightly smaller nozzle expansion area ratio has
tems. Thrust chamber ignition is achieved by been specified than for the A-2. While all three
the hypergolicity of the propellants. One signif-
upper stages operate in the vacuum and can use
icant feature of this engine system is the clus-
the largest practical expansion area ratio for
tering of two thrust chambers to one propellant
best performance, other considerations will in-
feed system and one set of propellant controls.
fluence the ratio actually chosen.
The propellants are fed by pressurants directly
from the propellant tanks through the main pro-
pellant valves to the thrust chamber inlets.
System Operation
Gaseous helium supplied from high-pressure
bottles is used for pressurization of both tanks. The propulsion system is designed to start
The pressurant is heated in heat exchangers automatically upon a signal from the guidance
located at the thrust chamber nozzle extensions system. During main-stage operation, engine
--A. --
,lUswn.
TABLE 3-5.-7.5K A-4 Stage Engine Operating Parameters
[Vacuum conditions]
Engine (pressurized gas-feed and throttlable): Main valve pressure drop .......... psi ............ 4
Calibration orifice pressure drop .... psi ............ 8
Thrust .......................... lb .......... 7500 Mixture ratio control reserve ....... psi ........... 10
Nominal total multiple-firing Oxidizer tank pressure ............ psia ......... 165
duration at full thrust ............ sec .......... 410 Total oxidizer weight (410 sec full
Specific impulse .................. sec .......... 320 thrust duration for 2 engines, plus
Oxidizer N:O,: 08 percent residual) ............. lb ......... 10 560
Density ....................... lb/ft 3 ...... 90.88 Oradizer tank volume (inch,ding
Flow rate ..................... lb/sec .... 12.78 2.5 percent ullage votume) ........ ft 3........... 120
Fuel N_-I,: Nominal pressurant (helium) flow
Density ....................... lb/ft 3 ..... 63.25 rate (assuming tank ullage
Flow rate ..................... lb/sec .... 10.65 temperature 7GO° R) .............. lb/sec .... 00225
Mixture ratio ..................... O/F ......... 1.2 Total pressurant weight (assuming
storage bottle final temperature
Thrust chamber (ablatively cooled and radiation cooled on 191 ° R, pressure 400 psia, plus
nozzle extension): 2 percent reserve) ............... lb ......... 12.95
Pressurant storage tank:
Thrust .......................... lb .......... 7500
Volume ....................... ft 3........... 4.3
Pressure ...................... psia ........ 4500
Injector end pressure .............. psia ......... 110
Temperature ................... °R ..... 560 max.
Nozzle stagnation pressure ........ psia ......... 100
Oxidizer flow .................... lb/sec ..... 12.78
Fuel side (pressurized by heated helium):
Fuel flow ........................ lb/sec ..... 10.65
Mixture ratio ..................... O/F ......... 1.2 Injector pressure drop ............. psi ........... 25
c* efficiency ..................... Percent ....... 98 Inlet manifold pressure drop ........ psi ............ 4
c* .............................. ft/sec ...... 5540 Line pressure drop ................ psi ............ 4
C[ efficiency ..................... Percent ...... 101 Main valve pressure drop .......... psi ............ 4
C[ ......................................... 1.858 Calibration orifice pressure drop .... psi ............ 8
Contraction ratio ................. At At ......... 2 Fuel tank pressure ................ psi .......... 155
Expansion ratio .................. Ae/.A_ ........ 35 Total fuel weight (410 sec full
Throat area, A_ ................... in 2 ......... 40.4 thrust duration for 2 engines,
L* .............................. in ............ 32 plus 1.2 percent residu'M) ........ lb .......... 8840
Nozzle contour .......................... 70 percent Fuel tank volume (including
bell 2.5 ullage volume) ............... ft 3......... 143.5
Nominal pressurant (helium) flow
Thrust vector control: rate (assuming tank ullage
temperature 700 ° R) .............. lb/sec ..... 0.025
Minimum acceleration ............. tad/set 2 ....... 2
Total pressurant weight (assuming
Maximum velocity ................. deg/sec ....... 15
storage bottle final temperature
Displacement .................... deg ........... ±7
191 '_ R, pressure 400 psia. plus
2 percent reserve) ............... lb .......... 14.4
Oxidizer side (pressurized by heated helium):
Pressurant storage tank:
Injector pressure drop ............. psi ........... 25 Volume ....................... ft a.......... 4.77
Oxidizer dome pressure drop ....... psi ............ 3 Pressure ...................... psia ........ 4500
Line pressure drop ................ psi ............ 5 Temperature ................... °R ..... 560 max.
thrust level and mixture ratio are controlled con- restart an indefinite number of times. It can be
tinuously through the engine control package by operated at any thrust level between 10 percent
the guidance and propellant utilization systems. and full thrust. Figure 3-11 shows the opera-
Upon a shutdown signal, engine shutdown is tional sequence of the A-4 stage engine.
effected. The propulsion system is capabIe of

P
CHECK
TANK VENT TO VALVE AND.,__

AND RELIEF REGULATOR
GIMBALACTUATORS
VALVES MAIN
FUEL
FUEL OXIDIZER ACTUATOR

TANK TANK
MIXTURE
THRUST
CHAMBERS
EXCHANGERS
]MAIN OXE
VALVE
START
VALVE
BOTTLES
TANK FILL
DRAIN VALVES HELIUM RELIEF
VALVE
,IUM FILL AND

VENT VALVE
Figure 3-9.-A-4 stage engine and propulsion system schematic diagram.
5'-- 9"
FUEL (N2H4) OXIDIZER (N204)
TANK TANK
3'--6" D
'- HIGH PRESSURE
HELIUM BOTTLES (2)
}SS PSIA 165 PSIA
/
PRESSURANT IHTERSTAGE _-THRUST CHAMBERS (2)
DIFFUSER CONNECTING 7'D
SKIRT THRUST VECTOR CONTROL
GIMBAL ACTUATORS (4]
Figure 3-I0.-15K A-4 stage propulsion system preliminary design layouL.

START SHUTDOWN
PRESSURANT SHUTOFF PRESSURANT SHUTOFF

_.LV E OPENS VALVE CLOSES
PROPELLANT TANK
_7_PROPELLANT TANK
PRESSURIZATION PRESSURIZATION
TERMINATES
_ THRUST CHAMBER
THRUST CHAMBER
"/////////////////;'_ PURGING
I
_/7_ OXIDIZER VALVE
OXIDIZER VALVE CLOSES
"/t OPENS
I I I
_FUEL VALVE OPENS _. FUEL VALVE CLOSES
I I I
THRUST LEVEL AND THRUST LEVEL AND MIXTURE RATIO
MIXTURE RATIO CONTROL DEACTIVATES
CONTROL ACTIVATES
I I
THRUST BUILD-UP AND DECAYS
MAIN STAGE
0,5
THiUST
II
1.5 2.0 ?_,5
1.0 15 2.0 2.5 3.0 O 0.5 I.O
SECONDS SECONDS
Figure 3-II.-A-4 stage engine operational sequence.

_L_=41_o '__ •_._rT _ • =
Chapter IV
Design Of Thrust Chambers

And Other Combustion Devices
While the proud designers of the various sub- 2. The droplets are subsequently vaporized
systems of a rocket engine each consider their by heat transfer from the surrounding gas. The
product as "the heart of the engine," the thrust size and velocity of the droplets change contin-
chamber assembly undeniably embodies the es- uously during their entrainment in the combustion
sence of rocket propulsion: the acceleration and gas flow.
ejection of matter, the reaction of which imparts 3. The vaporized propellants are mixed rap-
the propulsive force to the vehicle. The design- idly, further heated and promptly reacted at their
er's goal is essentially to accomplish this with stoichiometric mixture ratio where ever they are
a device of maximum performance, stability and formed, thus effecting a continuous increase of
durability, and of minimum size, weight, and the gaseous mass flow rate within the combus-
COSt. tion chamber. This gas reaction is further aided
The design of thrust chambers is one of the by the high-speed diffusion of active molecules
more complex subjects in the field of liquid pro- or atoms. The combustion is essentially com-
pellant rocket engineering. This is primarily due plete upstream of the chamber throat, when all
to the fact that the basic processes, especially liquid droplets have been vaporized. Under cer-
the combustion within the thrust chamber, are tain conditions, shock and detonation waves may
comparatively difficult to define and to study be generated by local disturbances in the com-
analytically. Thus, during most engine develop- bustion front, possibly caused by instability of
ment programs, a major effort must be expended mixing process and propellant flow prior to reac-
toward the design and development of the thrust tion. These effects may trigger sustained pres-
chamber. A rational approach to this effort is sure oscillations at certain frequencies within
attempted here. the thrust chamber, resulting in destructive com-
bustion instability. A major portion of the de-
sign and development effort, therefore, is directed
4.1 THE BASIC THRUST CHAMBER
toward achievement of stable combustion.
ELEMENTS
4. As the gaseous products of the combustion
The thermodynamic processes governing the process pass toward and through the throat, they
generation of thrust within a thrust chamber have are accelerated to sonic, and then to supersonic,
been treated in chapter I. The primary function velocities within the diverging nozzle section,
of the thrust chamber is to convert the energy of and are finally ejected to the rear.
propellants into thrust. In a liquid bipropellant The basic elements of a thrust chamber re-
rocket engine, this process is characterized by quired for its function, include a combustion
the following basic functional steps: chamber section, an expansion nozzle section,
1. The liquid propellants, at their proper mix- an injector, an ignition device (for nonhypergolic
ture ratio, are injected into the combustion cham- propellant combinations), propellant inlets and
ber through orifices in an injector, as jets at distributing manifolds, and interconnecting sur-
velocities ranging from 20 to 150 feet per second. faces for, component and thrust mounts. The con-
These jets either impinge to form a mixed drop- struction of the various thrust chamber elements
let spray, or run straight into the chamber hot depends largely on their specific operational
gas as a series of droplets. Part of the combus- function. However, low weight and simplicity,
tion reaction may already take place in the liquid which make manufacturing easier, are two impor-
state. tant factors to be considered at all times.
81
Figures 4-1 and 4-2 illustrate a typical liquid into the return tubes. The fuel then flows
bipropellant rocket engine thrust chamber assem- through an injector fuel screen into the radial
bly. The illustrated thrust chamber assembly is injector passages, and finally through the fuel
composed of four major subassemblies or basic injector orifices into the thrust chamber combus-
elements; namely, thrust chamber body, injector, tion zone. The fuel manifolds of this chamber
liquid oxygen dome, and igniter. are made of 4130 steel or 347 stainless steel.
The thrust chamber body subassembly is of a Other structural members, such as tension bands,
venturi shape, consisting of a cylindrical section stiffening rings, and outriggers were all made of
in which the combustion occurs; a section nar- 4130 steel. The oxidizer (liquid oxygen) enters
rowing toward a throat; and a bell-shaped, ex- the LOX dome under pressure through a screened
panding nozzle section through which the com- central port and is distributed within the dome
bustion gases are expelled (fig.4-1). The body directly to the liquid oxygen passages and
wall of this chamber is constructed of nickel orifices (fig. 4-2).
tubes running longitudinally, joined by silver The following are the operating characteris-
brazing, and retained by external tension bands. tics and principal dimensions of a hypothetical
The tubes, of 0.012-inch wall thickness, are of thrust chamber similar to the one shown in figure
rectangular cross section of varying area, to 4-1:

1. Propellants ........................ LOX:'RP-1
conform to the thrust chamber shape. This con-
2. O;F mixture ratio ................... 2.80
struction permits simple thrust chamber cooling
3. Characteristic velocity, c*, ft/sec .... 5400
during operation, by flowing fuel through the 4. Thrust coefficient, CI (sea level) ..... 1.489
tubes which form the chamber wall. The fuel, 5. Specific impulse (ls)tc (sea level),
under pressure, enters the thrust chamber body sec .............................. 249
at the fuel manifold inlet and is distributed to 6. Total propellant flow rate, lb/sec ..... 402
7. Thrust (sea level), lb ............... 100000
alternate thrust-chamber tubes. It then flows
8. Chamber pressure (injector end), psia . 520
down toward the thrust chamber nozzle exit 9. Chamber pressure (nozzle stagnation),
where the fuel return manifold reverses the flow psia ............................. 480
"x\ ..&..C C_,GE,. SC_E_'r'
Ja.<j" ;C
Figure 4-1.-Thrust chamber assembly. Figure 4-2.-Thrust chamber injector.

DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 83
I0. Average gas specific heat ratio (y) .... 1.233 the injector are sealed by a spiral-wound gasket
11. Combustion chamber cross-section
made of 304 stainless-steel strips with asbestos
area, in 2 ......................... 244
fillers. This type of gasket is designed specifi-
(at injector)
12. Throat area, in 2 .................... 140 cally for cryogenic and elevated-temperature
13, Nozzle exit area, in 2 ................ 1120 applications.
14. Combustion chamber volume, in 3 ...... 5820 The electrically fired pyrotechnic igniter is
(above the throat) secured centrally to the injector surface by
15. Combustion chamber length, in ....... 28.5
means of a threaded joint. It is designed for one
(injector to throat)
16. Characteristic chamber length, L*, in . 38 start only and must be replaced after each firing.
17. Overall thrust chamber length, in ..... 73 It receives an electrical firing signal from wires
18. Design contraction area ratio, ec ...... 1.60:1 connected through the nozzle exit.
19. Design expansion area ratio, e ........ 8:1
The thrust chamber injector (fig. 4-2) is a 4.2 THRUST CHAMBER PERFORMANCE
round plate, honeycombed with circular and radial PARAMETERS
inner passages, leading to drilled orifices. It is
constructed of 4130 steel with nickel-plated sur- The significance of the parameters which
faces, and held in position at the fuel manifold express or influence the efficiency of thrust
below the liquid oxygen dome with high-strength chamber operation has been discussed in section
bolts. The seals between injector and thrust 1.3 of chapter I. Before discussing the details
chamber body are of the 0-ring type, made of of actual thrust chamber design, the following
rubber selected for compatibility with the fuel summarizes these parameters and illustrates
(RP-1). A threaded hole is provided in the cen- their use for design calculations by applying
ter of the injector face to permit pyrotechnic them to the engine systems of the Alpha vehicle
thrust chamber igniter installation. The injector which was discussed in chapter III.
has 20 circular concentric copper rings which
contain the injection orifices and are fed from
Specific Impulse, ts (sec)
the main propellant systems. Fuel and oxidizer
are kept separate by an elaborate distribution From equations 1-31 and 1-31c:
system, feeding alternate rings. Fuel flows
through the outermost ring, through each alter- F c*Ct
(ls)tc- . -
nate inner ring, and through a central fuel disk Wtc g
which is separately fed from an igniter fuel valve
through an ignition fuel inlet port. Liquid oxy- The specific impulse figure indicates the
gen emerges from the remaining rings. The in- overall quality of the thrust-chamber design. As
jection orifices are so angled that the propel- was learned earlier, it shows how much thrust is
lants impinge in the thrust-chamber combustion generated for what "running propellant expendi-
zone in a like-on-like pattern (liquid oxygen on tures."
liquid oxygen and fuel on fuel). The primary
orifices are arranged in pairs, with a 0.416-inch
Characteristic Velocity, c* (ft/sec)
distance between centerlines and a 40 ° included
impingement angle, for both propellants. Inother From equation 1-32a:
designs, impingement angles or orifice separa-
tions are made different for the two propellants, c* = [(y, R, (Tc)ns) (4-1)
so as to effect their impingement in different
planes (multiplanar impingement as opposed to Assuming that the propellant and mixture ratio
uniplanar). selection has been made, it can be expected that
The liquid oxygen dome is a single-piece, the gas properties (y, R) will fall into a known
2014-T6 aluminum-alloy die forging. It provides band. From there on, c* almost entirely depends
the inlet for the liquid oxygen. It also serves as on the temperature of the gases. Obviously, this
the thrust-chamber-to-vehicle attachment inter- temperature has a theoretical maximum for a
face. The flanges of the liquid oxygen dome and selected propellant combination. How close
to this maximum the chamber will operate de- well as from the chosen design configurations.
pends on influences discussed for mixture ratio Typical performance calculation methods have
in chapter II. Figures 4-3, 4-4, 4-5, and 4-6
illustrate this. It is seen that c* peaks at com-
*- w
bustion temperatures somewhat lower than maxi- g =g _o
mum. Other considerations, such as bulk densi- _m
a:_J
25
_m
ties, which affect vehicle tank sizes, may cause g_d 5 zo
further adjustments of the mixture ratio for opti-
mum overall vehicle performance. Within these
boundaries, the quality of the combustion proc-
ess greatly depends on the design efficiency of _ _ 6ooo
-u
2 z _ _ 5o00
a thrust chamber assembly, in particular the _o=E
injector.
u 3000 I _oo 5
/ 1 ..
Thrust Coefficient, CI (dimensionless) o i30
\ ,.°o-, I
From equation 1-33a:
1 25 , -4-'
C(= f(),, e, Pa) (4-2)
",o
K-120
Let us assume that the performance of energy I
15 2: Z.5 3 35
generation through the combustion process, the
MIXTURE R&TIO, r w • WIp/ Wf
effects of which were just summarized for c*,
has been determined. Then, with a given gas
Figure 4-3.-Theoretical 02/RP-1 combustion data
property (y), the performance of the remaining
ffrozen composition), (Pc)ns = I000 psia.
thrust generating functions of the thrust chamber,
essentially those of the divergent nozzle, will
depend on the nozzle geometry (mainly _ which
determines the pressure ratio (Pe/(Pc)ns) and the
ambient pressure (Pa).
Performance Calculation
In actual design practice, the calculation of

the thrust chamber performance is based on theo-
retical propellant combustion data and the appli- ._

cation of certain correction factors as explained _w
in chapter I. The theoretical propellant combus- _w
tion data are derived from thermochemical compu- -_ _
rations which equate the heat of reaction of the _°
propellant combination and the rise in enthalpy
of the combustion gases. Typical propellant

combustion data at frozen composition are pre-
sented in figures 4-3 through 4-6. For given
propellant combinations and chamber-nozzle
stagnation pressures (Pc)ns, the values for the
combustion gas temperature (Tc)ns, molecular
weight _, and specific heat ratio ), are plotted M¢XTOR£ Rn7,O, r w • wo/ wf
against the O/F mixture ratio rw. Performance

correction factors are determined by the theoreti- Figure 4-4.-Theoretical 02/H 2 combustion data
cal assumptions and from earlier test data, as (Irozen composition), (Pc )ns = 800 psia.
83O0 been demonstrated earlier by sample calculation

(1-3). The following sample calculations illus-
trate the more specific approaches.
_:li 8100
Determine the design values of c*, Ct, and

/ _ moo (ls)_c for the engine thrust chambers of the
- I 7eoo
"too _, 7-too stages of the hypothetical Alpha vehicle, with

the following assumed design parameters:
(a) 750K A-1 Stage Engine:
Propellants, LOa/I{P-1; thrust chamber
_z w
O/F mixture ratio, 2.35; (Pc)us, 1000
psia; propellant combustion data, figure
8 /
4-3; nozzle expansion area ratio, _= 14.
2ooo ":
(b) 150K A-2 Stage Engine:
o I 3S
Propellants, LO_/LHa; thrust chamber
O/F mixture ratio, 5.22; (Pc)us, 800
psia; propellant combustion data, figure
4-4; nozzle expansion area ratio, (--40.
] I t s Solutions
25 35 45 _5 65 75 S5 95 _05 1_5
MIXTUR Ir RATIO, r w • Wo/ Wf

(a) A-1 Stage Engine:
From figure 4.3 for LOJRP-1 at (Pc)us = 1000
Figure 4-5.-Theoretical F_/H 2 combustion data
(frozen composition), (Pc)as : 1 O0 psia. psia and a mixture ratio of 2.35, the following
values are derived for the chamber product gases:
(Tc)ns : 6000 ° F or 6460 ° R,

_ :22.5 lb,/tool, y: 1.222
<m .
Substitute into equation (1-32a):
Theoretical c* - _r32'2 x 1.222 X 6460 x 1544/22.5

0.7215
: 5810 ft/sec
This value for c* can also be derived from fig-

, .: ure 4-3.
For a good combustion chamber and injector
design, the c* correction factor for LO_/RP-1
8 _,oo
and frozen composition will be about 0.975.
/-'Y
o_ \ w// Fd _
"_ t25 8Q
Design c* = 5810 x 0.975 = 5660 ft/sec
For y= 1.222, e= 14, a theoretical vacuum CI

value of 1.768 can be derived from figure 1-11:
•---.. - ,u_
05 06 07 09 09 kO _l t2 k3 _4 L5 ePa
MIXTUR| RATIO, t w • W4/ Wf Theoretical Ct at sea level: (C/)vac (Pc)as
Figure 4-6.-Theoretical N_O_/NaH _ combustion 14 x 14.7

= 1.768 - 1.562
data (frozen composition), (Pc)us = 1 O0 psia. 1000
Sea level Ct can also be calculated using equa- 4.3 THRUST CHAMBER CONFIGURATION
tion (1-33a), with the aid of equation (1-20). LAYOUT
With effective nozzle contour design, an over-
After major thrust chamber operating param-
all Cf correction factor of 0.98 for LO2/RP-1
eters such as type of propellants, thrust level,
frozen composition can be used.
chamber pressure, C f, c*, and Is have been es-
tablished from engine system requirements and
Design sea level Cf= 1.562 × 0.98= 1.531
performance calculations, one of the fundamental
dimensions of the thrust chamber, the throat
From equation (1-31c):
area At, can be readily derived (eq. (1-33)). The
throat area At or throat diameter D t usually is
5660x 1.531
= 270 sec the starting point of a thrust chamber configura-
Design sealevel (Is)tc- 82.2
tion layout. The combustion chamber and nozzle
section are commonly designed as an integral
(b) A-2 Stage Engine:
thrust chamber body. For light weight and ease
From figure 4-4 for LO2/LH: at (Pc)ns = 800
of manufacture, thrust chambers will have the
psia and a O/F mixture ratio of 5.22, the follow-
general shape of a pressure vessel with wall
ing values are derived for the chamber product
surfaces of rotation and smooth contours.
gases:
(Tc)ns = 5580 ° F or 6040 ° R,

Combustion Chamber Volume
= 12 lb/mole, y= 1.213
The combustion chamber serves as an enve-
Substitute into equation (1-32a): lope to retain the propellants for a sufficient
period (stay time) to assure complete mixing and
X/32.2 x 1.213 × 6040 x 1544/12 combustion before entering the nozzle. The vol-
Theoretical c* =
0.717 ume of the combustion chamber thus has a defi-
nite effect on combustion efficiency. The theo-
= 7670 ft/sec
retical required chamber volume is a function of
the mass flow rate of the propellants and their
Based on experimental data, a c* correction average density, and of the stay time needed for
factor can be assumed for the LO:/LH2 frozen- efficient combustion. The relationship can be
composition data of about 0.975. expressed by the following equation:
Design c* = 7670 x 0.975 = 7480 ft/sec

Vc = WtcVts (4-3)
For ),= 1.213, _=40, a theoretical vacuum Cf

value of 1.876 can be derived from figure 1-11. where:
Cf can also be calculated using equations (1-33a) Vc =chamber volume, ft 3
and (1-20). With effective nozzle contour design, ¢/tc = propellant mass flow rate, lb/sec
an overall Cf correction factor value of 1.01 can V =average specific volume, fta/lb
be used for LO2/LH2 frozen-composition data: ts =propellant stay time, sec
Design vacuum Cf= 1.876 x 1.01 = 1.895 A useful parameter, the characteristic length L*
(commonly pronounced "L-star"), can be used to
From equation (1-31c): specify the propellant stay time in the combus-
tion chamber. L* is defined as the ratio of
7480× 1.895 chamber volume to nozzle throat area, and can be
Design vacuum (ls)tc = - 440 sec
32.2 expressed by the following equation:
The reader should perform his own calcula-

Vc VitcVts
tions for the A-3 and the A-4 engines, with the L* _ _ (4-4)
aid of tables 3-4 and 3-5, and figures 4-5 and 4-6. At At
DESIGN OF THRUST CHAMBERSAND OTHER COMBUSTIONDEVICES 87
Since the value of At is in nearly direct propor- TABLE 4-1.-Recommended Combustion Chamber
tion to the product of ')/tc and V, L* is essen- Characteristic Length (L*) for Various Propel-
tially a function of is. The effect of L* on c* in lant Combinations
an experimental combustion chamber is shown in
Combustion
figure 4-7. The c* value increases with L* to an
chamber
asymptotic maximum. Increasing L* beyond a Propellant
combination characteristic
certain point tends to decrease overall engine length (L*), m.
system performance because of the following:
(1) Larger L* results in higher thrust chamber Chlorine trifluoride/hydrazine-base fuel.. 30-35
volume and weight. Liquid fluorine/hydrazine ............. 24-28
Liquid fluorine/liquid hydrogen (GH_
(2) Larger L* creates more surface area in
injection) .......................... 22-26
need of cooling. Liquid fluorine/liquid hydrogen (LH 2
(3) Larger L* increases friction losses at the injection_ .......................... 25-30
chamber walls. Hydrogenperoxide/RP-I (including
catalyst bed) ....................... 60-70
In actual design practice, optimization analyses
Nitric acid/hydrazine-base fuel ........ 30-35
will determine the minimum possible combustion
Nitrogen tetroxide/hydrazine-hase fuel.. 30-35
chamber L* consistent with efficient combustion. 30-40
Liquid oxygen/ammonia ...............
Liquid oxygen/liquid hydrogen (GH2
Under a given set of operating conditions,
injection) .......................... 22-28
such as type of propellants, mixture ratio, cham- Liquid oxygen/liquid hydrogen (LH 2
ber pressure, injector design, and chamber geom- injection) .......................... 30-40
etry, the value of the minimum required L* can Liquid oxygen/RP-1 .................. 4@50
only be evaluated by actual firings of experi-
mental thrust chambers. L* values of 15 to 120
inches for corresponding propellant stay-time Combustion Chamber Shape
values of 0.002-0.040 second have been used in
As can be seen from equation (4-3), the stay
various thrust chamber designs. Typical L*
time ts is independent c; the combustion chamber
values for different propellants are given in
geometry. Theoretic:ally, for a given required
table 4-1. With At and minimum required L*
volume, the chamber can be of any shape. In
established, the required combustion chamber
actual design, however, the choice of the com-
volume Vc can be calculated by equation (4-4).
bustion chamber configuration is limited. In a
long chamber with a small cross section, higher
1 nonisentropic gas flow pressure losses will re-
NZ041 50-50 NzH4-UDMH |
sult as explained in chapter I. This approach
O/F MIXTURE
{Pc]n== 275
RATIO
I::'SIA.
• 190
EC=2.O0
/ also dictates a longer thrust chamber space en-
velope and imposes certain space limitation on
5800 the injector design to accommodate the neces-
¢.)
to
sary number of injector holes. With a short
chamber of large cross section, the propellant
_
rr
u. 5600
/ atomization or vaporization zone occupies a
/
relatively large portion of the chamber volume,
while the mixing and combustion zone becomes
too short for efficient combustion. Other factors,
_5._oo
such as heat transfer, combustion stability,
>
weight, and ease of manufacturing, are to be
uJ 5200
considered in determining the final combustion
I0 20 30 40 50
chamber configuration.
CHAMBER CHARACTERISTICLENGTH( Le) IN. Three geometrical shapes which have been
used m combustion chamber design are shown in
Figure 4-7.-Effect 6[ L* on c* value of experi- figure 4-8. While the spherical and the near-
mental thrust chamber. spherical chambers were used in earlier European
v v __ --
designs, the cylindrical chamber has been used novel thrust chamber designs will also be dis-
most frequently in the United States. cussed.
The spherical or nearly-spherical chamber, as In the design layout of the cylindrical com-
compared to the cylindrical one of the same vol- bustion chamber of a given At and L*, the value
ume, offers the advantage of less cooling surface of the contraction area ratio, (ec =(Ac/At)) can
and weight. A sphere has the smallest surface- be optimized through careful studies of the fol-
to-volume ratio. For equal strength of material lowing factors:
and chamber pressure, the structural walls of the (1) Combustion performance in conjunction
spherical chamber are about half the thickness of with the injector design
the walls of a cylindrical chamber. However, the (2) Chamber gas flow pressure drop
spherical chamber is more difficult to manufac- (3) Chamber wall cooling requirements
ture and has poorer performance under most cir- (4) Combustion stability
cumstances. For these practical reasons, the (5) Weight
design details of the cylindrical combustion (6) Space envelope
chamber will be treated in this book. Several (7) Ease of manufacturing
For pressurized-gas propellant feed, low-
thrust en_ne systems contraction area ratio
THROAT /NOZZLE values of 2 to 5 have been used. For most turbo-
pump propellant feed, high thrust and high cham-
IN_ THRUST--CHAMBER ber pressure engine systems lower ratio values
of 1.3 to 2.5 are employed. The reader is also
referred to section 1.2 chapter I, "The Gas-flow
Processes in the Combustion Chamber and the
AXIS Nozzle."
COMBUSTION The basic elements of a cylindrical combus-

CHAMBER tion chamber are identified in figure 4-9. In
design practice, it has been arbitrarily defined
that the combustion chamber volume includes the
space between injector face I-I and the nozzle
throat plane II-II. The approximate value of the
IN_ NOZZLE combustion chamber volume can be expressed by
the following equation
Vc = A_ [Lcec + ½_-A-_Cot O(ec_/3 -1)] (4-5)

_-NEAR SPHERICAL
COMBUSTION INJECTOR THROAT
CHAMBER FACE [-CHAMBER n"
I [ DIA " Dc I ,--- NOZZLE
T_T NOZZLE I AREA Ac , /
INJECTOR
? HRUSTC.___.HHAMBER
AXIS
[
CHAMBE:"
CYLINDRICAL
SECTION
_
[
[ OIA
AREA
Dt
A t
z_LENGTH Lc
CYLINDRICAL
COMBUSTION CHAMBER CONTRACTION Ac
CHAMBER AREA RATIO (¢ = A'-'_-

T
Figure 4-8.-Frequently used geometrical shapes Figure 4-9.-Elements of basic cylindrical com-
[or combustion chambers. bustion chamber.
DESIGNOF THRUST CHAMBERSAND OTHER COMBUSTIONDEVICES 89
The total surface area of the combustion The selection of an optimum nozzle shape for
chamber walls excluding injector face can be a given expansion area ratio is generally influ-
approximated by the following expression: enced by the following design considerations
and goals:
(1) Uniform parallel axial gas flow at the
Total area =2Lc_cAt + csc 9(ec - 1)At (4-6) nozzle exit for maximum momentum
vector
(2) Minimum separation and turbulence losses
Nozzle Expansion Area Ratio within the nozzle
It was learned earlier that with all other (3) Shortest possible nozzle length for mini-
parameters fixed, in particular chamber pressure, mum space envelope, weight, wall fric-
there is only one optimum nozzle expansion area tion losses, and cooling requirements
ratio for a given altitude or, more specifically, (4) Ease of manufacturing
ambient pressure. Except for those systems In actual design practice, any abrupt change or
which start in vacuum, ambient pressure will discontinuity in the nozzle wall contour should
have to be considered. This is especially true be avoided to eliminate the possibility of shock
for boosters which start at or near sea-level waves or turbulence losses. Theoretically, the
conditions. nozzle throat is simply the unique plane of mini-
mum cross-section area. In practice, a well-
It is the ultimate purpose of a rocket engine
rounded throat section is employed. Only at the
to lift vehicles to altitudes. Inherently, then,
nozzle exit plane is a sharp edge used because
ambient pressure will not be a constant (except
a rounded one would permit overexpansion and
for high-altitude starts, as mentioned). It is,
flow separation.
therefore, extremely important for the designer to
know the trajectory of the vehicle to be propelled
1. Conical Nozzle
or, more specifically, its altitude-versus-time
characteristics. With this information, the de- In early rocket engine applications, the coni-
signer is in a position to make a first, optimizing cal nozzle, which had proved satisfactory in
selection of a nozzle expansion area ratio, for most respects, was used almost exclusively.
best results throughout the entire trajectory. As The advantages of a conical nozzle are ease of
shown earlier, area ratio will be truly optimum manufacturing and flexibility of converting an
for only one specific altitude. The optimization existing design to higher or lower expansion area
for ambient pressure then is essentially an ratios without major redesign of the nozzle con-
averaging process. tour.
Other considerations usually cause the de- The configuration of a typical conical nozzle
signer to deviate from the "paper optimum" for is shown in figure 4-10. The nozzle throat sec-
the nozzle expansion area ratio. Some of the tion has the contour of a circular arc with a
most common are: weight, size, ease of manu- radius R ranging from 0.5 to 1.5 times the throat
facturing, handling, and cooling (heat transfer) radius Rt. The half angle of the nozzle conver-
considerations. gent cone section can range from 20 ° to 45 °. The
Nozzle Shape
Most rocket nozzles are of the converging-

RI
diverging De Laval type. Since the flow velocity
of the gases in the converging section of rocket
nozzle is relatively low, any smooth and well-
rounded convergent
very low energy losses.
nozzle section will have
By contrast, the contour
,. :- y-os
of the diverging nozzle section is very important
to performance, because of the very high flow
velocities involved. Figure 4-I O.-Conical nozzle contour.
9O DESIGNOF LIQUID PROPELLANT ROCKET ENGINES
EXIT
divergent cone half angle a varies from approxi- [ PLANE
mately 12 ° to 18 °. The length of the conical [: L_ . E. L_E
nozzle section can be expressed by the equation
/ / / / _-c:,,-_,. 1
Rt(_e - 1) + R(sec a - 1)
Ln= (4-7)
tan a
-,---__ / _/ .....,
_, C_CrERIS_r ic
The conical nozzle with a 15 ° divergent half
angle has become almost a standard, as it is a 1 ._
good compromise on the basis of weight, length,
and performance. Figure 4-11.-Bell nozzle contour.
Since in a conical nozzle certain performance
chosen for the nozzle contour MT upstream of the
losses occur as a result of the nonaxial compo-
throat. Contour TNE is the diverging portion of
nent of the exhaust gas velocity, a correction
the nozzle. The initial expansion occurs along
factor h is applied for the calculation of the exit
contour TN; contour NE turns the flow over to a
gas momentum. This factor or thrust efficiency
direction nearer to axial. For design conven-
is the ratio between the exit gas momentum of
the conical nozzle and that of an ideal nozzle ience, the contour TN is also a circular arc, with
a smaller radius R2.
with uniformly parallel axial gas flow. The
For those familiar with compressible flow
value of h can be expressed by the following
theories, it is noted that, using transonic flow
equation:
analyses, a constant-Mach-number line TO can
be defined at the throat. Given the flow condi-
;_= _ (1 + cos a) (4-8)
tion along TO and the solid boundary TN, a
kernel flow field TNKO can be generated by the
where a = half angle of the conical nozzle.
method of characteristics developed in gas dy-
For an ideal nozzle, A would be unity. For a
namics. The kernel of the rocket nozzle contour
conical nozzle with a = 15 ° and h =0.983, the exit
is defined as that portion of the supersonic flow
gas momentum or the exit velocity will be 98.3
field determined entirely by throat conditions.
percent of the ideal nozzle exit velocity calcu-
The last right characteristic line NK of kernel
lated by equation (1-18). The value of the vac-
uum thrust coefficient of a nozzle is in direct TNKO, and thus the location of the point N along
contour TN, is to be determined by specific
proportion to the thrust generated by the nozzle,
design criteria.
or to the nozzle exit gas velocity. Therefore,
The location of the end point E along contour
the theoretical vacuum thrust coefficient (neglect-
NE is defined by the given nozzle expansion
ing friction and other flow losses) of a conical
area ratio and nozzle length (distance between
nozzle with 15 ° half angle will be 98.3 percent
throat and exit plane). Then the right character-
of the ideal nozzle thrust coefficient calculated
istic line NK can be determined by satisfying the
by equation (1-33a).
following conditions concurrently:
(1) A control surface PE can be generated
2. Bell Nozzle
between the point E and a selected point
For increased performance and shorter length, P along the line NK
bell-shaped nozzles have been developed. This (2) Mass flow across PE equals the mass
nozzle design employs a fast expansion or radial flow across NP
flow section in the initial divergent region, (3) Maximum thrust by the nozzle is attained.
which then leads over to a uniform, axially By selecting points P', P", etc., along line
directed flow at the nozzle exit. The wall con- NK, a series of control surfaces P'E', P"E",
tour is changed gradually enough so that oblique etc., can be generated to define points E I, E",
shocks will not form. etc., along the contour NE. Calculations for the
Figure 4-11 shows the contour of a bell noz- nozzle contour can be effectively performed by a
zle. A circular arc of selected radius R, is computer.
8,\
I
IO0 I
99
.J
!
-_ 98
j / ,jo POINT
!
i NOZZLE
p-
Z
_J
//// i
I, Ln
AX*S
{J THaT
nr
bJ
O. 97
Figure 4-13.-Parabolic approximation of bell
nozzle contour.
,,,o//
96
/
/
/ section with a radius of 0.382 Rt from the throat
T to the point N and a parabola from there to the
9_
60 70 80 90 I_ Lf % exit E.
For the design of a specific nozzle, the fol-
FRACTIONAL NOZZLE LENGTH (Lf) BASED ON A 15 = HALF ANGLE
CONICAL NOZZLE WITH ANY AREA RATIO lowing data are required:
(1) Throat diameter, Dr, inches
Figure 4-12.-Thrust efficiency versus bell nozzle (2) Axial length of the nozzle from throat to
length. (Shown for comparison: effect of short- exit plane, Ln, inches (or the desired
ening conical nozzle, increasing half angle.) fractional length Lf based on a 15 °
conical nozzle)
Commonly, an equivalent 15 ° half-angle coni- (3) Expansion area ratio e
cal nozzle is used as a standard to specify bell (4) Initial wall angle of the parabola, On,
nozzles. For instance, the length of an 80- degrees
percent bell nozzle (distance between throat and (5) Nozzle exit wall angle, 0e, degrees
exit plane) is 80 percent, or 0.8 of that of a 15 ° The wall angles, On and Oe are shown in figure
half-angle conical nozzle having the same throat 4-14 as a function of the expansion area ratio E.
area, radius below the throat, and area expansion

ratio.
Figure 4-12 shows the thrust efficiency h
40
! L/= toO%
versus fractional nozzle length LI for conical
Lf= 70%
and bell nozzles. Lf = 80%
L_= 90%
As may be seen, bell nozzle lengths beyond _5o Lp 100%
approximately 80 percent do not significantly

contribute to performance, especially when con-
sidering weight penalties.
3. Parabolic Approximation of Bell Nozzles 2¢
One convenient way to design a near-optimum- __

Lf" 60=/=
thrust bell nozzle contour is through the use of _ _'5.,c i LI" 70%
z
the parabolic approximation procedures as sug- E Lf = 80%
.f, 90°/°
gested by G.VR.
of a parabolic
Rao.
approximation
The design configuration
bell nozzle is
_ "_
0
i
30
t
40
Lf, 100%
5O
_0 20
shown in figure 4-13. The nozzle contour imme- EXPANSION AREA RATIO •
diately upstream of the throat T is a circular arc

with a radius of 1.5 Rt. The divergent section Figure 4-14.-0n and Oe as function of expansion
nozzle contour is made up of a circular entrance area ratio _.
Optimum nozzle contours can be approximated zles. For comparison of the effect of nozzle
quite accurately by selecting the proper inputs. type on size, all nozzles shown are scaled to
Although no allowance is made for different pro- the same thrust level, nozzle expansion area
pellant combinations, experience has shown that ratio, and theoretical nozzle efficiency. These
the effect of specific heat ratio y upon the con- nozzles show potential of adapting their geom-
tour is small. A computer program can be readily etry to space vehicle application, because short-
set up to perform the calculation. ened nozzles reduce interstage structure weight
and will permit an increase in payload through
4. Annular Nozzles increased performance for a given length.
The nozzle expansion area ratio ( for an
Based on the momentum theorem, for ideal ex-
annular nozzle is defined by equation (4-9):
pansion the thrust generated by a thrust chamber
depends only upon the mass flow conditions
(velocity and direction) at the nozzle exit. In Projected area of the
some nozzle designs, such as annular nozzles, contoured nozzle wall Ae-Ap
e: - (4-9)
the gas flow at the throat is not necessarily Throat area At
parallel to the axis, but the exit flow is similar
to that of a conical or bell nozzle and thus pro- where the projected area of the contoured nozzle
duces the same thrust results. wall equals nozzle exit plane area Ae, less the
There are two basic types of annular nozzles: centerbody projected area Ap. Another conve-
the radial in-flow type (spike nozzle) and the lent design parameter for annular nozzles is the
radial out-flow type (expansion-deflection or annular diameter ratio, Dp/Dt, where Dt is the
E-D; reverse-flow or R-F; and horizontal-flow or throat diameter of an equivalent circular throat,
H-F nozzles). They are shown in figure 4o15, and Dp the centerbody diameter. The parameter
together with conventional conical and bell noz- Dp/Dt is an index of the annular nozzle design
AREA RATIO = 36:1

OF EFFICIENCY = 9B.3% (ALT}
JECTOR
OMBUSTION
HAMBER
HROAT INJECTOR
I_ COMBUSTION
INJECTOR / INJECTOR
: rl ..--"- COMBUSTION i;;rl_ _../-- cOM BUSTION
[_ JJ'- CHAMBER _l.J Z" CHAMBER
_ j/"-/pD'_ TH R OAT
'tl X / i "'THROAT / j.._D,I._, _
:i' CONE w.,, ' , T

' BELL
T,'
,
i':44--..
"i
sPIKE
5f 'i,! E-D ' "
Dp/D I = 1.3 Dp/D T = 1.3
NOZZLE LENGTH = 100% NOZZLE LENGTH = 74.2% NOZZLE LENGTH = 41.4% NOZZLE LENGTH = 41.4%
OVERALL LENGTH = I00% OVERALL LENGTH = 70% OVERALL LENGTH = 51% OVERALL LENGTH = 519=,
OVERALL DIAMETER = 100% OVERALL DIAMETER = tO0% OVERALL DIAMETER = 105"/. OVERALL DIAMETER = 102.5 %
_ THROAT INJECTOR-,, /- COMBUSTION

COMBUSTION
X_ / CHAMBER
CHAMBER
{NJECTOR
Ii- I p---Dp"---'4 ;_
R-F Y H-F i1_
Dp/D T = ,5 Dp/D t = t0
NOZZLE LENGTH = 24,9 NOZZLE LENGTH = 14.5 %
OVERALL LENGTH = 21% OVERALL LENGTH = 12 %
OVERALL DIAMETER = 150 % OVERALL DIAMETER = 194 %
Figure 4-I 5.-Comparison of nozzle shapes.

geometry as compared to a conventional nozzle.

CONSTANT_ _NOZZLE WALL
The contour-calculating methods for annular
nozzles are similar to those for bell nozzles. MACHLINE_ __/I
In a conical or bell nozzle, the gases may ex-
THROAT----x _,,/f_-_'_ __!
pand to pressures well below the ambient (sea- \ /_I_ FREESTREAMSURFACE
level or low-altitude operation) before flow sepa-
ration from the nozzle wall occurs. As explained °"--_B-_]_ _--SNOCK .OZZkE
in chapter
this overexpansion
I, for nozzles
results
with large
in thrust
area
losses
ratios,
at '------CENTER
....
BODY
low altitudes. Annular nozzles, because of their WALL
special characteristics, are not subject to these
losses. As shown in figure 4-16 for an E-D noz-
zle (and equally applicable to other annular noz-
zles), the back pressure Pb at the back face of
, Essu,
the centerbody plays an important role in regu-
lating the nozzle flow. The value of Pb is a
Figure 4-17.-E-D nozzle at high altitude
function of the ambient pressure and generally is
operation.
lower than ambient. Downstream of the throat,
the expansion of the gases around the centerbody pressive turning at the nozzle wall, which is
shoulder C will continue unaffected until this also typical for the spike nozzle, is responsible
base pressure is reached. After the initial gas for improved nozzle performance at low altitude.
expansion through the constant-Mach line CD, Because of the self-adjusting nature of the inner
the downstream flow of the gases is controlled jet boundary, there is no flow separation from
by the following two boundary conditions: the nozzle wall, as is the case for a conven-
(i) The nozzle wall contour DE which turns tional nozzle.
the gases to near-axial flow. At high-altitude operation the base pressure
(2) The base pressure Pb which influences Pb becomes so low that the nozzle flow con-
the free stream surface of the inner jet verges behind the centerbody, as shown in figure
boundary. 4-17. Since the flow at the closure point must
Because of the curved-wall contour, the gases be axial, a shock wave may occur depending on
are deflected, which leads to some compression the flow conditions. However, the expansion of
and local increases in wall pressure. A typical the gases may continue unaffected up to the end
nozzle wall pressure distribution for low-altitude of the nozzle. The nozzle wall pressure distri-
operation is shown in figure 4-16. This corn- bution under this condition is also shown in
figure 4-17.
An improved spike nozzle concept is the
NOZZLE WALL _ E aerodynamic spike nozzle.' This nozzle concept
CONSTANTI_ACH---k '"_""_'"_ is a truncated annular spike nozzle (radial in-
L.NE \ flow type), which utilizes a small amount of sec-
TN"OAT F,,EE ondary flow introduced into the nozzle base
,NJECTO, " SURFACE region.
NOZZLE
Performance of the aerodynamic spike nozzle
AXIS is a function of various nozzle geometric param-
WALL BODY eters, the amount of secondary flow, the manner
PRESSURE_ in which this secondary flow is introduced, and
I. the relative energy between the primary and sec-
1 ondary streams. To describe the flow field and

interrelated effect of truncating the spike nozzle,
_Source: AIAA Paper No. 66-828, "Liquid Rocket

Engines: Their Status and Their Future" By S. F.
Figure 4-16.-E-D nozzle at low altitude operation. Iacobellis.
the base pressure and the base pressure increase

achieved through the secondary flow addition
requires a lengthy, detailed discussion; only the IDEAL NOZZLE (NO LOSSE!
basic operation can be presented here.

The primary flow (high-pressure gases) which
HIGH-AREA-RATIO
produces the major portion of the engine thrust AEROSPIK E NOZZLE _
i,
is exhausted from an annular-type combustion LL
chamber and expands against the metal surface

L)
/" /
/
of the center tru-_ated-spike nozzle (fig. 4-18).
/ /_ HICa-I-AR EA-RATIO
The characteriF of the primary flow field up- i / /_ BELL NOZZLE
stream of the basv, shown as region 1 in figure f I

I
4-18, are determined by the annular throat geom-
etry, the nozzle wall contour, and the ambient
LU
,'
//SEA'LEVEL / (VACUUM)
pressure. The annular primary flow continues to

expand beyond the nozzle surface and encloses / _ / OPERATING RANI
a subsonic, recirculating flow field in the base

/
region (region 2). The pressure acting on the SEA LEVEL IVACUUMI
nozzle base contributes additional thrust to the _,t_-- OPERATING RANGE
nozzle.
IJ iI I ] K JlJli{
When a small amount of secondary flow is 50 100 200 300 500 1000 2000 4000
introduced into the base (added to the recirculat-

PRESSURE RATIO (pc,/pa)
ing flow), the base pressure is increased further.
As the secondary flow is increased, the overall
Figure 4-19.-Nozzle perlormance comparison.
nozzle efficiency (considering the additional
flow) increases because of this increase in base
presses the outer free-jet boundary of the primary
pressure. There is a limit to this gain in effi-
flow field. This compression increases the static
ciency, and an optimum secondary flow exists
pressure on the nozzle wall and partially offsets
for each configuration.
the negative effect of the higher ambient pres-
The outer surface of the annular primary flow
sure on the back side of the nozzle. The base
is a free-jet boundary, which is influenced by
pressure also is increased with the higher am-
ambient pressure. This ambient pressure influ-
bient, because the compressed primary flow field,
ence on the primary nozzle flow endows this type
which influences the base pressure, has higher
of nozzle with altitude compensation. In oper-
static pressures. This combination of flow field
ation at high-pressure ratios (i.e., altitude con-
effects provides the altitude compensation in-
ditions), the outer free-jet boundary of the pri-
herent to the aerodynamic spike nozzle.
mary flow expands outward, governed by the
Figure 4-19 presents the performance com-
Prandtl-Meyer turning angle at the throat. At
parison of various nozzle designs. The nozzle
low-pressure ratios (i.e., sea level operation),
thrust coefficient Cf for an ideal nozzle (i.e., a
the relatively higher ambient pressure corn-
variable-area-ratio nozzle having the optimum
expansion for each chamber pressure to ambient
TOR_DJU._Ew
pressure ratio, pc/Pa) is shown together with
those of the high-area-ratio aerodynamic spike
and bell nozzle. As is evident, the CI curve of
the aerodynamic spike follows the ideal nozzle
performance (altitude-compensation), rather than
dropping off rapidly like the bell design at low
pc/Pa (i.e., sea level) operating points. M1
nozzles have a higher CI at a high Pc/Pa (i.e.,
vacuum).
Figure 4-18.-Aerodynamic spike flow [ield i11us- The development of the annular-nozzle con-
Srated under altitude conditions. cept may influence the design of rocket vehicles,
especially in the areas of boattail structure and 4-23. The reader is urged to conduct his own
mission staging optimization. The advantages calculations using the first stage as a guide, and
and disadvantages of annular nozzles are sum- to compare his results with those shown.
marized as follows:
Advantages Solution
(1) Shortened nozzle length for the same per-
A-1 Stage Engine:
formance, or increased performance (higher
From sample calculation (4-1):
expansion area ratios) for a given length.
(2) Improved performance at sea level or low
Design sea level C/= 1.531;
altitudes. (Annular nozzles with high
(Pc)as = 1000 psia; e= 14
expansion area ratios can be used for a
single-stage sea level to vacuum vehicle
Substitute into equation (1-33):
mission.)
(3) The relatively stagnant region in the cen-
747 000
ter of the nozzle can possibly be used Throat area:
At - 1.531 x 1000 = 487 in:
for installation of gas generators, turbo-
pumps, tanks, auxiliary equipment, and
turbine gas discharges. Throat diameter: Dt = _= 24.9 in
(4) A segmented combustion chamber design
approach can be used, easing develop-
ment effort (individual segments can be Rt=_= 12.45 in
built and tested during the early phases)
and improving combustion stability. Exit diameter: De --'-_V/_ -x 24.9 =93.4 in
Disadvantages R 93.4
(1) Relatively high cooling requirements, e =---_- = 46.7 in
because of higher heat fluxes and greater
surface areas to be cooled. Use a combustion chamber L* of 45 in for
(2) Heavier structural construction in some LO2/RP-1 application. Substitute into equation
applications. (4-4):
(3) Manufacturing difficulties.
Chamber volume: Vc=487x45=21915cuin
Sample Calculation (4-2) Use a nozzle convergent half angle of 20 °, a

Lay out the thrust chamber internal configura- contraction area ratio ec = 1.6, and a circular arc
tion (cylindrical combustion chamber with bell of radius R= 1.5Rt, or 18.68 in, for nozzle con-
nozzle) for the engines on the Alpha vehicle with tour upstream of the throat.
the data derived from sample calculation (4-1)
Chamber diameter: Dc = _'.'.6x 24.9 =31.5 in
and the following required chamber thrusts Ftc:
(a) A-1 stage engine: Ftc =747000 lb at sea R 31,5
level C
=_----=--=15.75 in
(b) A-2 stage engine: Ft%= 149500 lb at
altitude Use equation (4-7) to calculate the chamber
(c) A-3 stage engine: Ftc3= 16000 Ib at convergent cone length
altitude
(d) A-4 stage engine: Ftc4= 7500 lb at Convergent cone length
altitude
_ 12.45 (x/1.6 - 1) + 18.68 (sec 20 °- 1)
The detailed calculations and their results
tan 20 °
are presented in the following for the first-stage
engine only. For the other stages, the calcula- 4.515
tion results are summarized in figures 4-21 to = 0.---3"64= 12.4 in
INJECTOR
FACE
THRUST
CHAMBER AXrS
L n = 102.4" -"-
Figure 4-20.-A-I stage engine thrust chamber internal configuration layout:

(--14, 80% bell, L*=45", (o=1.6.
INJECTOR THROAT EXIT

FACE ,d
=9.25°
I
b
I
8n=33 °
THRUST N J
__ _L_
I
CHAMBER AXIS 14"R N0=5.94.
De=71"
I
L
17.3" - ! _ Ln =83.6"
Figure 4-21.-A-2 stage engine thrust chamber, internal configuration layout:

_=40, 75% hell, L*=26", _c=1.6.
k.,
4mmmal
INJECTOR
FACE
THRUST
CHAMBER AXIS
_=35, 70% bell, L*=28", _c=2.
INdECTOR
FACE THROAT EXIT
I _1
V Et:46.1, `
THROAT
CHAMBER AXIS
_=35, 70% bell, L*=32", ec=2.

Using the frustrum cone volume equation and Since the calculations for the thrust chamber
neglecting the slight rounding of the throat, the configuration are based on the calculated design
approximate convergent cone volume is obtained: C[ value which has to be verified by later actual
testing, a slight change of chamber pressure is
usually allowed to compensate for C! deviations
Volume =3 x12.4 [(15.75)2 + (12.45)2 +15.75×12.45 ] in order to meet the required thrust value.
= 7760 cu in
4.4 THRUST CHAMBER COOLING

Required volume for cylindrical chamber section
=21 915- 7760:14 155 cu in
Techniques and Their Selection
Required length for cylindrical chamber section Because of the high combustion temperatures
= 14 155/1.6At= 18.17 inches (4000 ° to 6000 ° F) and the high heat transfer
rates from the hot gases to the chamber wall (0.5
Distance from injector face to throat to 50 Btu/in2-sec), thrust chamber cooling be-
= 18.17+ 12.40= 30.57, say 31 inches comes a major design consideration. For short-
duration operation (up to a few seconds), un-
Design an "80-percent bell" nozzle configura- cooled chamber walls can be used. In this case,
tion using the parabolic approximation procedure. the heat can be absorbed by the sufficiently
The nozzle contour downstream of the throat will heavy chamber wall material which acts as a
be a circular arc of radius 0.382 Rt, or 4.75 heat sink, before the wall temperature rises to
inches. By definition, the nozzle length Ln will the failure level. For most longer durationappli-
be 80 percent of the length for an equivalent 15 ° cations, a steady-state chamber cooling system
half-angle conical nozzle. Substitute into equa- has to be employed. One or a combination of the
tion (4-7) following chamber cooling techniques is often
used:
1. Regenerative coollng.-Regenerative cool-
Ln=0.8× [12.45 (VIT-1)+4.75
tan 15 ° (sec 15 °- 1!]
ing is the most widely applied method and uti-
lizes one or possibly both of the propellants, fed
= 0.8 x 128 = 102.4 inches
througt_ passages in the thrust chamber wall for
cooling, before they are injected into the com-
The parabolic contour wall angles 0, and 0e can
bustion chamber. (See par. 4.1 and fig. 4-1.)
be derived from figure 4-14, for e= 14 and L/=0.8;
2. Dump cooJing.-With this principle, a small
On = 27.4 ° and 0e = 9.8 °. The location of N and E
percentage of the propellant, such as the hydro-
along the nozzle contour, with respect to throat
gen in a LO2/LH2 engine, is fed through pas-
and nozzle axis, can be calculated
sages in the thrust chamber wall for cooling and
subsequently dumped overboard through openings
Nt =0.382 R: sin On = 2.19 inches
at the rear end of the nozzle skirt. Because of
Na = Rt + 0.382 Rt(1 - cos 0n) : 12.99 inches inherent problems, this method has only limited
application.
Et : Ln = 102.4 inches
3. Film cooling.-Here, exposed chamber wall
Ea= Re=46.7 inches surfaces are protected from excessive heat with
a thin film of coolant or propellant which is in-
With the aid of the established coordinates for troduced through manifolded orifices in the cham-
points N and E, and the angles 0_ and 0e, a ber wall near the injector, and usually in several
parabola can be fitted to complete the contour. more planes toward the throat. The method has
The general layout of the A-1 stage engine thrust been widely used, particularly for high heat
chamber is shown in figure 4-20. With the aid of fluxes, either alone or in combination with
a computer program, more accurate calculations regenerative cooling.
of the divergent nozzle contour can be made by 4. Transpiration cooling.-Transpiration cool-
the method of characteristics. ing is accomplished by introducing a coolant
(either gaseous or liquid propellants) through availability of this pressure drop permits the use
porous chamber walls at a rate sufficient to of regenerative cooling which requires propellant
maintain the desired combustion gas side cham- pressure sufficient to force the coolant through
ber wall temperature. This method is essentially the cooling passage before entering the injector.
a special type of film cooling and has been A pressurized-gas-fed engine system usually has
widely used. more stringent pressure limitations and operates
on relatively low chamber pressures. This sug-
5. Ablative cooling.-In this process a sacri-
gests the application of film, ablative, or radia-
fice of combustion-chamber gas-side wall mate-
tion cooling.
rial is made by melting and subsequently vapor-
izing it to dissipate heat. As a result, relatively 4. Thrust chamber configuration.-The geo-
cool gases flow over the wail surface, thus metric shape of the chamber affects local
creating a cooler boundary layer, assisting the combustion gas mass flow rates and wall surface
cooling process. Ablative cooling has been used areas to be cooled. This influences the choice
in numerous designs, initially mainly for solid of cooling method. It can also limit the design
propellant systems, but later equally success- arrangements for regeneratively cooled tubular
fully for low Pc, pressure-fed liquid systems. wall thrust chambers.
6. Radiation cooling.-_tith this method, beat 5. Thrust chamber construction material.-The

is radiated away from the surface of the outer properties of the thrust chamber materials will
thrust chamber wall. It has been successfully affect the cooling system design profoundly.
applied to low heat flux regions, such as nozzle Strength at elevated temperature, combined with
extensions. heat conductivity properties of a metal, will
The selection of the best cooling method for Jetermine suitability for regenerative cooling
a given thrust chamber depends on many design systems. For film-cooled chambers higher allow-
considerations. There are no simple-and-fast able material working temperatures are desired
rules. However, the following are the main fac- to reduce heat transfer rates and thus film cool-
tors which influence the selected design ant flow rates. The application of radiation-
approaches: cooling to a chamber depends largely on the
1. Propellants.-The properties of the com- availability of high temperature (3000 ° F and up)
bustion products, such as temperature, specific refractory alloys. The success of ablative cool-
heat, specific weight, viscosity, etc., have a ing depends entirely on the availability of suit-
direct bearing on the heat transfer rate and in able composite plastic materials.
turn affect the chamber cooling requirements and In practice, the design of thrust chamber
methods. The cooling properties of the propel- cooling systems is a major link in the complete
lants and their relative flow rate decide whether engine system design. It cannot be treated inde-
they are suitable or sufficient for regenerative or pendently, without due consideration of other
film cooling. Therefore, in evaluating a chamber engine system aspects. For instance, optimiza-
cooling system, the propellants involved will be tion of the chamber pressure value for a high-
one of the primary design considerations. performance engine system is largely limited by
2. Chamber pressure.-High chamber pressure the capacity and efficiency of the chamber cool-
is linked with higher combustion gas mass flow ing system. In turn, chamber pressure will affect
rates per unit area of chamber cross section and other design parameters such as nozzle expan-
thus raises the heat transfer rate. Combined sion area ratio, propellant feed pressure, and
regenerative and film-cooling methods are usually weight. Because of the complex interrelation
employed for the stringent requirement of higher between these factors, the complete analysis of
chamber pressure applications. chamber cooling systems is a specialized field
3. Propellant feed system.-The type of pro- and requires thorough knowledge of heat transfer,
pellant feed used in an engine system deter- fluid mechanics, thermodynamics, and thermal
mines the pressure budget for the system. In a stresses. The engine system designer, therefore,
turbopump-fed engine system, more pressure drop will enlist the services of heat transfer spe-
is usually available for chamber cooling. The cialists.
Gas-Side Heat Transfer dominantly influenced by the mass wlocity or

the mass flow rate per unit area of the gas,
One of the primary steps in the design of a
subject to the exponent 0.8. In comparison, all
thrustchamber cooling system is the analysisof
other factors are relatively minor. A rough
the heat transferfrom the combustion gases to
approximation of hg can thus be expressed by
the chamber walls (gas-sideheat transfer).Be-
the following equation:
cause of the very high surface velocityof the
gases along the chamber walls, the heat transfer
hg=(p'V) °8 (4-11)
occurs mainly through forcedconvection; i.e.,
throughthe transferof heat energy resultingfrom
where
the relativemotion of different
parts of a fluid.
p' = Free stream value of local gas density,
Before the gases can transferheat to the wall, lb/cu in
the heat energy must pass througha layer of
= Free stream value of local gas velocity,
stagnantgas along the wall, calledthe boundary in/set
layer.
Thus, under normal circumstances, the heat-
The basic correlationforthiscomplicated
transfer coefficient varies with the chamber pres-
convective heat transfercan be expressed by the
sure to the 0.8 power and throughout a given
followingequation:
chamber inversely with the local chamber diam-
eter to an exponent of 1.8.
q : hg (Taw - Twg) (4-10)
Based on experience with turbulent boundary
layers, some relatively simple correlations for
where
the calculation of the gas-side heat-transfer
q = Heat flux or heat transferred across the
coefficient have been developed. A much-used
stagnant gas film per unit surface area
form is that credited to Colburn
per unit time, Btu/in2-sec
hg =Gas-side heat transfer coefficient, Btu/
Nu = C Re °8 Pr °_ (4-12)
in2-sec-deg F
Taw = Adiabatic wall temperature of the gas, where
deg R= (Tc)ns x turbulent boundary layer
Nu = Nusselt number = hg D/k
recovery factor (ranging from 0.90 to
C =Dimensionless constant
o.9s)
Re = Reynolds number = p'VD/_
Twg = Hot-gas-sidelocal chamber-wall tempera- = Free stream velocity,in/set
ture,deg R
Pr = Prandtlnumber = _Cp/k
The determination of the gas-side heat trans-
D =Hydraulic diameter,in
fer coefficient hg is a rather complex problem.
k =Gas thermalconductivity,Btu/sec-in2-
The convection phenomenon as it occurs in
deg F/in
rocket thrust chambers eludes complete under-
_t =Viscosity, Ib/in sec
standing. Attempts to compare analytical results
Cp=Specific heat at constantpressure,Btu/Ib-
with experimental heat-transfer data obtained on
deg F
rocket thrust chambers have often shown disa-
greement. The differences are largely attributed or as Bartz has shown
to the initial assumptions for analytical calcula-
tions. For example, there is good evidence
oxidizing and reducing atmospheres
wide range of temperature
covering a
exist locally in the
that
'++L+.
+ ,.'+c+++"
It+
] i+)o.
combustion product gases within the thrust cham-
ber, because of the imperfect mixing of the pro- (-+)°'o
+,++
pellants at the injector face. This results in where
deviations from calculations based on the as- R = l_dius of curvature of nozzle contour at
sumption of homogeneous product gases. throat, in
However, it has been established by experi- o = Correction factor for property variations
ment _hat the beat-transfer coefficien_ is pre- across the boundary layer
A = Area under consideration along chamber axis (1) A substantial fraction of the combustion
The value of a can be evaluated in terms of noz- gases are strong radiators.
zle stagnation temperature, local gas-side cham- (2) There is substantial dissociation, with
ber wall temperature, and local Mach number. subsequent recombination near the wall.
(3) There are strong high-frequency flow
1 instabilities.
The calculated values may be higher than the
actual ones, because of the following:
(1) The combustion reactions may not be
values of a for various Twg/(Tc)n s and y, as completed in the chamber.
computed by Bartz, are shown in figure 4-24. (2) The combustion gases may deposit solids
If Pr and # data are not available for particu- on the chamber walls, which act as
lar combustion gas mixtures, the following equa- insulators.
tions can be used for approximate results: In certain propellant combinations, the com-
bustion products contain small amounts of solid
particles. These solids tend to deposit on the
Pr=9__5 (4-15)
chamber wails, and form a rather effective insu-
lating layer. A quantitative evaluation of _he
/_=(46.6 × 10-_°) _ °ST °'6 (4-16) insulation effectiveness of this layer, necessary
where T= temperature of gas mixture, °R for correct heat transfer calculations, has been
Equations (4-13), (4-14), (4-15), and (4-16) accomplished only experimentally.
can be used to calculate the approximate hg val- In the case of the LO2/RP-1 combination,
ues along the thrust chamber walls. However, carbon solids are deposited on the chamber
the calculated values can be expected to be walls. After a firing, the carbon gives the in-
lower than the actual ones if the following con- terior of the thrust chamber the appearance of
ditions exist: being freshly painted black. The outer surface
of the carbon appears sooty and can easily be
removed by light rubbing. Underneath the exte-
rior soot layer is a harder, graphitelike layer
which can also be removed, but is more tena-
cious. This carbon deposit significantly in-
creases the gas-side thermal resistance. The
temperature of the carbon deposit at the hot
gas-side interface approaches the gas tempera-
ture as the carbon thickness increases.
The values of the thermal resistance of the
carbon deposit based on actual experimental
testing results of a thrust chamber burning
LO2/RP-1 are shown in figure 4-25.
0.6 For the heat transfer calculation of the gas-
side heat transfer with solid deposit on chamber
"'l I JI walls, the following equations can be used
TI/
O -I'z J I I
q = hgc (Taw - Twg) (4-17)
where hgc = overall gas-side thermal conductance,

o.6[ 1 I I I I _1 I
4 3 2 I Z 3 4 5 6 78510 20 30 40 Btu/in2-sec-deg F
CONTRACTION I[XP&NSION
1
(4-18)
Figure 4-24.-Values of correction factor a for hgc- i
Rd
property variation across boundary layer. hg +
2400
!,! ,; = 6460 × (0.975) 2 : 6140 ° R
(See eq. 1-32a and 1-41).

22(311 _
Design c* = 5660 ft/sec
I_ u? ,8o0 From sample calculation (4-2):

i
Dc=24.9 in
Mean radius of the throat contour

fzoo I 18.68+4.75
- - 11.71 in
2 I 2 4 6
I ,
8 _
,
12
i
_4
2
CONTRI.CllON _ _¢ t _ EXPANSION
Figure 4-25.-Thermal
AREA
resistance
RATIO
el carbon de-
yR 1.222 x e 44)
Cp - (),_ I),I - (i. 222 - I) × 778 = 0.485 Btu/lb-deg F
posit on chamber walls LO2/RP-1, mixture
ratio = 2.35, (Pc)n s = 1000 psia.
where Rd thermal resistance caused by the solid

4 × 1.222
deposit, in2-sec-deg F/Btu
Pr : (9 × 1.222) - 5 =0.816
When there is no solid deposit, Rd =0 and
hgc = hg, and equation (4-10) is used for heat
transfer calculations. From equation (4-16):
: (46.6 × 10 -1°) × (22.5) 05 x (6140) 0.6

= 46.6 × 10-1° × 4.76 × 188
Determine the approximate design gas-side
overall thermal conductance hgc in the combus- - 4.18 x 10 -6 lb/in-sec
tion chamber, at the throat, and at the exit noz-
zle point of e=5, for the regeneratively cooled From equation (4-13):
thrust chambers on the A-1 and A-2 stage
engines.
h -V0"026 x((4"18×10-_)°2x0'485)
g-L 0.8160.6
Solution 0.9
(7
(_.) A-I Stage Engine ×\-g-6-65 7
First, let us consider equation (4-13). The
combustion reactions are assumed to be homoge- At\ 0.9
neous and complete. From figure 4-3 the follow- : 0.01366 × 0.046 × 4.02 × 1.078 × k-_} o
ing values are derived for the chamber product
gases, for LO2/RP-1 at (Pc)ns = 1000 psia and a "At_O.9
mixture ratio of 2.35: = 0.0027 x ("_') a
(Tc)ns = 6000° F or 6460 ° R, Since the carbon deposit temperature ap-

=22.5 lb/mol, y= 1.222
proaches the gas temperature, a (Twg/(Tc)ns)
value of 0.8 is used to determine the a values
The design (Tc)ns
from figure 4-24 (),_1.2).
=Theoretical (Tz)ns × (c* correction factor) 2 At the combustion chamber:
J
b(b_)A-2 Stage Engine

(__) °9:(1_16)
" 1" 0.9 :0,655, a=l,05 Again, the combustion reactions are assumed
to be homegeneous and complete. From figure
4-4, the following values are derived for the
hg : 0.0027 x 0.655 × 1.05
= 0.00185 Btu/in2-sec-deg F chamber product gases, for LO;/LH2 at (Pc)as
= 800 psia and a mixture ratio of 5.22:
At the throat:
(Tc)ns = 5580 ° F or 6040 ° R,
=12 lb/mol, y= 1.213
-_)°'9=l,a= 1
The design (Tc)ns =
hg = 0.0027 x 1 x 1 : 0.0027 Btu/in 2-sec-deg F

Theoretical (Tc)ns x (c* correction factor) 2
At the exit nozzle point of = 6040 × (0.975) 2 = 5740 ° R

e=5,(_-)°'9=(1)°9:O.235, a=0.8
Design c* = 74S0 ft/sec
hg = 0.0027 × 0.235 × 0.8

From figure 4-21:
: 0.000507 Btu/in2-sec-deg F
Dt = 11.2 in
The experimental data of figure 4-25 can be
used to determine the values of thermal resist-
Mean radius of the throat contour =
ance Rd, for the carbon deposit. The thermal
resistances are
8.4+2.14
- 5.27 in
2
in2-sec-deg F in:-sec-deg F
1670
Btu , 1125 Btu '
1544
in2-sec-deg F yR 1.213 x i----_
and 1645
Btu Cp = (y_ 1)g = (1.213 - 1) × 778- 0.943 Btu/lb-deg F
for points at the combustion chamber, the throat, From equation (4-15):
and the exit nozzle area ratio of e = 5.
Substitute into the equation (4-18); at the 4x 1.21a
combustion chamber Pr = (9 x 1.213) - 5 = 0.820
1 From equation (4-16):

= 0.00045 Btu/in2-sec-deg F
hgc - 1
--+ 1670
0.00185 = (46.6 x 10-lo) (12) o.s (5740)o6
= 46.6 x 10- ,o × 3.47 x 180

at the throat
=2.92x 10 -6 lb/in-sec
hg c - = 0.00067 Btu/in2-sec-deg F
1
--+ 1125
0.0027 From equation (4-13):
At the exit nozzle of e= 5.

hg=L
[ 0.026 x ((2.92x, _10-6)°2×0.943) .
hg c- = 0.000276 Btu/in2-sec-deg F
1
-- e1645
0.000507 x\ 7480 x\5--._ ] j\-_--] a
(A_ 09 t.,4_CHAMBER
:0.01605
×0.0828
×2.69×1.078×\--X-J
°
GAS SIDE I INNER WALL
BOUNDARY LAYER ['_qI.--COOLANT SIDE
I A_ x°9
_At BOUNDARY LAYER
= 0.00385 x _---_-) a Z
\
Since there is no solid deposit on the chamber To.
wails, an average gas-side wall temperature of COM__
COOLANT
°1 It
1500 ° R is assumed, and a (Twg/(Tc)ns) valueof
(1500/5740) or 0.26 is used to determine the a
values from figure 4-24. Tw_
--Twc Tco
At the combustion chamber:
-- = 0.655, a= 1.38 RADIAL DISTANCE FROM

(__) 0.9 G.__16 ) 0.9
CENTER OF CHAMBER
hgc = hg : 0.00385 x 0.655 x 1.38 Figure 4-26.-Heat trans[er schematic for regen-
= 0.00348 Btu/in2-sec-deg F erative cooling.
At the throat:
=hc(Twc-Teo) (4-20)
_)°9 = 1, o=1.35 = H (Taw -Tco) (4-21)
1
H- (4-22)
fig c :hg=0.00385 × 1 x 1.35 1 t 1
where
At the exit nozzle point of e = 5:
q = Heat flux, Btu/in2-sec
hgc = Overall gas-side thermal conductance,

Btu/in2-sec-deg F (see eq. 4-18; with-
(-_)°9=(1)°9=0.235, a=1.16
out deposits, hg c = hg)
hc =Coolant side heat-transfer coefficient,
hgc = hg = 0.00385 × 0.235 × 1.16 Btu/in2-sec-deg F
=0.00105 Btu/in2-sec-deg F k = Thermal conductivity of chamber wall,
Btu/in2-sec-deg F/in
Regenerative Cooling t = Chamber wall thickness, in

Taw =Adiabatic wall temperature of the gas,
The heat transfer in a regeneratively cooled
deg R
chamber can be described as the heat flow be-
Twg= Gas-side wall temperature, deg R
tween two moving fluids, through a multilayer Two = Coolant side wall temperature, deg R
partition. Figure 4-26 shows this process sche- Teo = Coolant bulk temperature, deg R
matically. The general steady-state correlation H = Overall heat-transfer coefficient, Btu/in:-
of heat transfer from the combustion gases sec-deg F
through the layers, which include the metal The bulk temperature Too of the coolant in-
chamber walls, to the coolant can be expressed creases from the point of entry until it leaves the
by the following equations: cooling passages, as a function of the heat
absorbed, and of the coolant flow rate. Proper
balance of these parameters, to maintain the
chamber walls at temperatures below those at
which failure might occur because of melting or The characteristics of coolant side heat
stress, is one of the major criteria for the design transfer depend largely on the coolant pressure
of regeneratively cooled thrust chambers. For and coolant side wall temperature. In figure
metals commonly used in thrust-chamber walls, 4-27, the heat flux is plotted versus wall temper-
such as stainless steel, nickel, and Inconel, the ature for a constant coolant pressure, bulk tem-
limiting hot-gas-side wall temperature is around perature, and flow velocity. Curve A indicates
1500°-1800 ° F. The resultant differences be- the behavior of heat transfer at coolant pres-
tween combustion gas temperature and wall tem- st, res below critical. Line segment At-A2 repre-
perature range from 2500 ° to 6000 ° F. sents the heat transfer without boiling when the
Assume a station in the thrust chamber with wall temperature is below the saturation temper-
gas temperature Taw and coolant bulk tempera- ature of the coolant corresponding to the fluid
ture Tco. Referring to equation 4-21, it is seen pressure. As the wall temperature at A: exceeds
that the heat flux q, which must be the same the saturation temperature by a certain margin
through all layers, is a function of the tempera- (50 ° to 100 ° F), bubbles will form within the
tures, and of overall heat transfer coefficient H. coolant layer close to the wall. The bubbles
The value of H is composed of the individual grow continuously out into the colder liquid
coefficients for the boundary layers and the stream until condensation at the vapor to liquid
chamber metal wall (eq. 4-22). The smaller H, surface begins to exceed the rate of vaporization
the smaller is q. However, it is one of the major at the base of the vapor bubble, whereupon the
design goals to keep coefficient hgc low, but bubbles start to collapse. This process, which
heat transfer coefficient hc and conductivity t/k occurs at high frequencies, is described as
high, in relation to hgc. Since the temperature "nucleate boiling." It substantially increases
differentials are inversely proportional to the the heat-transfer coefficient, resulting in little
heat-transfer coefficients of the heat flow paths, increase in wall temperature for a wide range of
the temperature drop will then be steepest be- heat fluxes. The heat transfer with nucleate
tween hot gas and inner chamber wall. The boiling is represented by line A2-A 3. At A 3,
effect is analogous to voltage drops along resis- further increase in the heat flux abruptly leads
tors in electrical circuits. to such'a dense bubble population that the bub-
It is noted that the heat absorbed by the pro- bles combine into a vapor film with an attendant
pellant used for regenerative cooling raises large decrease in heat-transfer coefficient. The
temperature of the propellant, and thus the energy region of heat transfer with film boiling is repre-
level before it is injected into the combustion sented by line A3-A 4. The resulting increase in
chamber. However, this effect on overall engine
performance is slight, the gain usually being
less than 1 percent. On the other hand, regener-
ative cooling with attendant pressure losses
_6 m .A 3 -w,. .... ° .........
requiring additional turbopump power or higher ¢o
gas pressurization levels imposes a performance P"_ C U RVE A
x
N (Pco =I/2 Pc_,r,cAO
penalty.
_ z4
Coolant Side Heat Transfer
The coolant side heat-transfer coefficient hc Mr I ,"--CURVE

is influenced by many factors. At the high heat At _ _" I I ( P"OI > I PCmTICALI I)
fluxes and temperatures encountered in thrust O B_2 4 6 e I0 12 14 16 is 20

chamber operation, the propellants used for cool- COOLANT SIDE WALL TEMP.
ing may become corrosive, may decompose or TWC (°F) X I0 -z
deposit impurities upon the heated surface,
thereby reducing cooling effectiveness. It is Figure 4-27.-Heat flux versus coolant side wall
impossible to calculate the hc values under temperature of typical propellant in various
these conditions without experimental data. heat trans/er regions.
wall temperature is so high that failure of the gw =coolant viscosity at coolant sidewall
wall material often occurs. The heat flux at A_ temperature
is defined as the upper limit of nucleate boiling d = coolant passage hydraulic diameter, in
of the coolant qul, which therefore should be k = coolant thermal conductivity, Btu/sec-in 2-
used as the design limit for a regenerative cool- deg F/in
ing system. p =coolant density, ib/in 3
Curve B indicates the heat transfer behavior Vco = coolant velocity, in/sec
of a coolant above critical pressure. Since no Cp = coolant specific heat at constant pres-
boiling can occur, the wall temperature continu- sure, Btu/ib-deg F
ally increases with increasing heat flux. Line The heat flux at the upper limit of nucleate
B1-B2 represents the heat-transfer region, when boiling qul can be estimated from
the wall temperature is below the coolant critical
temperature. The heat-transfer coefficient re- qul C 2 x 10*
(4-24)
mains essentially constant. As the wall temper- qnonboiling PcoG
ature reaches the critical temperature B2 and
higher, a gradual transition to a stable super- where
critical vapor-film boundary layer begins, which C2 = constant, its value depending on
results in somewhat lower heat-transfer coeffi- coolant used
cients. Line B2-B 3 represents the heat transfer qnonboiling -- heat flux without nucleate boiling,
in this region. Wall failure temperatures are Btu/in2-sec
usually reached at lower temperatures when the Pco = coolant pressure, psia
coolant is above the critical presstire than when G = coolant maximum flow rate per
it is below it. Where possible, a coolant oper- unit area, lb/in2-sec
ating pressure between 0.3 to 0.7 of critical
pressure should be used to take advantage of the When the heat is transferred through a vapor-
high heat-transfer coefficients available with film boundary layer (coolant at supercritical
nucleate boiling. However, in most systems, pressure and temperature, region B2-B 3 in fig.
particularly those which are fed from a turbo- 4-27), the coolant-side heat-transfer coefficient
pump, the cooling jacket pressure, which is hc can be estimated from
equal to or larger than the sum of chamber pres-
sure and injection pressure, is supercritical.
hc: 0.029 Cpp °2 (.GO*,_.2._, _ oss
For the nonboiling subcritical temperature re- pr2/3 \--d'_] \Twc! (4-25)
gions of both, subcritical and supercritical cool-
ant pressures (AI-A2 and B_-B 2 in fig. 4-27), the where
relationship between wall temperature and heat Cp = coolant specific heat at constant pres-
flux, which depends on the heat transfer coeffi- sure, Btu/Ib-deg F
cient hc, can be predicted with sufficient accu- coolant viscosity, Ib/in-sec
racy for design purposes with the help of the Pr = Prandtl number
Sieder-Tate equation (eq. 4-23) for turbulent heat G = coolant weight flow rate per unit area,
transfer to liquids flowing in channels: Ib/in2-sec
d z coolant passage hydraulic diameter, in
014
Tco = coolant bulk temperature, deg R
Nu=C 1 Re °SPr°4 (4-23)
(ju+) Twc--- coolant side wall temperature, deg R
The bulk temperature of most coolants should
where be kept below the critical temperature, since the
C I =a constant (different values for various vapor-film heat-transfer coefficient would be too
coolants) low to cool the wall effectively. The cooling
Nu =Nusselt number= hcd/k capacity of the liquid-state regenerative coolant
Re = Reynolds number = pVcod/_ system can be estimated by
Pr = Prandtl number = i_Cp/k
/_ =coolant viscosity at bulk temperature Qc = #cCp (Tcc - Tci) (4-26)
transfer calculation. There are several basic

design approaches for regenerative-cooled thrust
chambers. Axial-flow cooling jackets, made up
of tubes, are used in the design of large thrust
chambers (3000 pounds of thrust and up); coaxial
k._.._, _f.._ L
shells separated by helical ribs or wires are
typical of the smaller thrust chamber designs.
Figure 4-1 shows a large regenerative cooled
tubular wall thrust chamber. Figure 4-28 repre-
i sents a typical coaxial shell design for a smaller
thrust chamber.
In this design, the coolant passage is defined
as the rectangular area between inner and outer
i shell and two adjacent ribs, which are wrapped
helically around the inner shell or liner.
/ Tubular Wall Thrust Chamber Design
In the design of tubular wall thrust chambers,

the number of coolant tubes required is a func-
tion of the chamber geometry, the coolant weight
flow rate per unit tube area, the maximum allow-
able tube wall stress, and fabrication considera-
tions. The critical cooling region is near the
Figure 4-28.-Coaxial shell thrust chamber cuta- throat where the heat flux is highest. It is this
way. Note overheated and burnt-through spot region, then, which determines the number of
on chamber wall. tubes required for a given coolant flow rate. For
easier fabrication and lower stress, tube cross
where sections of circular shape are preferred. How-
Qc =coolant capacity, Btu/sec ever, other shapes are often used to meet certain
We = coolant mass flow rate, lb/sec flow-area requirements. The stress analysis of
Cp =coolant specific heat at constant pres- the tubes is based upon three primary considera-
sure, Btu/lb-deg F tions: the hoop stress caused by coolant pres-
Tcc =coolant critical temperature, deg R sure, the thermal stress caused by temperature
Tci = coolant inlet temperature, deg R gradient across the tube section and the wall,
The allowed value of the total chamber wall gas and the bending stress caused by distortion in-
heat-transfer rate Q should be kept below Qc by duced by the pressure differential between adja-
a safe margin (Q<Qc). cent tubes (if any) or by other effects such as
However, there is no such limitation for hy- discontinuities. The tube design stress is based
drogen when used as a coolant. Hydrogen has on the combined stress from the above three con-
excellent heat-transfer characteristics with a siderations. It has been found that the maximum
reasonably high heat-transfer coefficient even in combined stress will occur at the throat region.
the supercritical pressure and temperature region. As shown in figure 4-29, the maximum com-
Usually liquid hydrogen enters the chamber cool- bined tangential stresses of the circular-tube
ant passage under supercritical pressure and will be at section A-A and can be expressed by
reaches supercritical temperature a short dis-
tance from the inlet. (Pco - Pg) r Eaqt 6 MA
The coolant passage areas at various points St- t + 2 (1 - v) k + t---7-- (4-27)
along the chamber walls are designed to maintain
the proper coolant velocity dictated by the heat- where
transfer coefficients determined by the heat- St =combined tangential tensile stress, lb/in-'
q =heat flux, Btu/inLsec and (4-29) can be also applied to calculate the
r = tube radius, in stresses for this design. Here again the maxi-
t =tube wall thickness, in mum combined stress is at section A. The bend-
Pco = coolant pressure, lb/in 2 ing moment at section A, M:4, should take into
pg =combustion gas pressure, lb/in2 consideration the pressure differential (if any)
E =modulus of elasticity of tube wall mate- between adjacent tubes.
rial, lb/in 2
a =thermal expansion coefficient of tube wall F-CHAMBER
material, in/in-deg F
k =thermal conductivity of tube wall material, / JACKET _ .__ ZONEII
Btu/in2-sec-deg F/in
v = Poisson's ratio of tube wall material
MA =bending moment caused by discontinuity,
in-lb/in (no effect of pressure differen-
tial between adjacent tubes for circular
tube design)
Since the combustion-gas-side portion of the tube
(zone I) has a much higher mean temperature than
that of the back side tube portion and chamber
outer shell (zone II), the thermal expansion of
EQUIV. CHAMBER
zone I will be restrained by zone II. Because of
INTERNAL RAD.
/
the considerably greater mass of zone II, thermal
inelastic buckling is induced under certain con-
ditions, in zone I, in the longitudinal direction.
The longitudinal thermal stress can be estimated
THRUST CHAMBER
by
Figure 4-29.-Circular tube wall of regeneratively

81 = Ea AT (4-28) cooled thrust chamber.
where
S_ =longitudinal thermal stress, lb/in 2 CHAMBER
h T: mean temperature difference between zone JACKET / ZONE II

I and zone II, deg F /
S 1 should be kept at a level not higher than
0.9 Sc, below.
The critical stress for longitudinal inelastic
buckling on zone I can be estimated by
4 ErEct (4-29)
)r Z_P= Pco1 . pco2 /
EQUIVALENT CHAMBER
where INTERNAL RADIUS
Sc = critical stress for longitudinal inelastic

buckling in zone I, lb/in 2
Et = tangential modulus of elasticity at wall
temperature, lb/in 2
Ec = tangential modulus of elasticity from com-
pression stress-strain curve, at wall
THRUST CHAMBER
temperature, lb/in 2
An elongated cross section tube design is Figure 4-30,-Elongated tube wall of regenera-
shown in figure 4-30. Equations (4-27), (4-28), lively cooled thrust chamber.
,4
lap k =thermal conductivity of inner shell mate-

M'A: MA+ KA-_- (4-30) rial, Btu/in2-sec-deg F/in
v =Poisson's ratio of inner shell material
where
M_ = combined bending moment at section A, Pressure Drop in Cooling Passages
in-lb/in
MA = bending moment due to discontinuity It is desirable to design a cooling passage
KA = dimensionless design constant based on with minimum pressure drop. Abrupt change of
test results (range 0.3-0.5) flow direction and sudden expansion or contrac-
1 =length of flat portion on tube wall, in tion of flow areas should be avoided. The inner
Ap = pressure differential between adjacent surface of cooling passages should be smooth
tubes, lb/in 2 and clean. The pressure drop in a cooling pas-
Substituting equation (4-30) into equation (4-27), sage can be treated as that in a hydraulic con-
the maximum stress of the elongated tube can be duit and be calculated accordingly.
calculated.
The working loads induced in a regenerative L pVco2
tubular wall chamber by the chamber pressure are Ap:[ _ 2g (4-32)
designed to be absorbed by a chamber jacket or
tension bands wrapped around the tube bundle. where
Ap :coolant pressure drop through the portion
of cooling passage under consideration,
Coaxial Shell Thrust Chamber Design lb/in 2
In a coaxial-shell-type thrust chamber as L =length of that portion, in
shown in figure 4-28, the outer shell is sub- d = equivalent average diameter of that por-
jected only to the hoop stress induced by the tion, in
coolant pressure• The inner shell, however, is 9 = average density of the coolant, lb/in3
subjected to the combination of compressive Vco = average coolant flow velocity, in/sec
stress caused by the pressure differential be- g =mass conversion factor, equal to gravita-
tween the coolant and combustion gases, and of tional constant, 32.2 × 12 in/sec 2
the thermal stress caused by the temperature I = friction loss coefficient, which is a func-
gradient across the wall. The maximum stress tion of the Reynolds number, and of
occurs at the inner-wall surface of the inner cooling passage conditions such as
shell and can be calculated from the following surface smoothness, geometric shape,
equation. etc. This can be determined only ex-
perimentally (also see fig. 7-20 and
table 7-3)
Sc _ (Pco-Pg) R _ Eaqt (4-31)
t 2(1 -v)k
where Determine the cooling passage and the tube

Sc =combined maximum compressive stress, design at the throat for the thrust chambers of
lb/in 2 the A-1 and A-2 stage engines.
q =heat flux, Btu/inLsec
R =radius of the inner shell, in
Solution
t =thickness of the inner shell, in
Pco =coolant pressure, lb/in 2 (a_) A-1 Stage Engine
pg =combustion gas pressure, lb/in 2 The fuel, RP-1, is used as the coolant. Since
E =modulus of elasticity of inner shell mate- the cooling requirement is most stringent at the
rial, lb/in 2 throat, the tube design for this station will serve
a =thermal expansion coefficient of inner as the starting point for the entire chamber. For
shell material, in/in-cleg F high strength, Inconel X is chosen as the tube
material. Based on experimental test results 3.00

h c
which showed good solid carbon deposits, design Twc - Tco 1000 - 600
values not exceeding 1000 ° F or 1460 ° R may be
permitted for gas-side tube-wall temperature.
Specifically for the throat region, a Twg value of
1188 ° R is taken. Using the results of sample The relationship between required hc and correct
calculation 4-3, the value for the adiabatic wall tube diameter is established by equation (4-23),
temperature Taw can be calculated by multiply- and experimental data for RP-1 (C 1 =0.0214):
ing (Tc)ns by the estimated stagnation recovery
factor of 0.923, r- Taw -- 6140 x 0.923= 5667 ° R.
From the same caiculation, the overall gas-side Nu = 0.0214 Re° SPr°4 (#)0.,4
thermal conductance at the throat region is
hgc =0.00067 Btu/in2-sec-deg F. Substitute into or, substituting corresponding terms:
equation (4-10), to obtain the heat flux at the
throat:
----_ = 0.0214 (#/°14 (a)
q = (5667 - 1188) x 0.00067 =3.00 Btu/in2-sec
The following additional relationships exist:

From supplier's specifications, the following
average data are obtained for Inconel X at 1000 °-
1200 ° R: Coefficient for thermal expansion,
Number of tubes N = rr[Dt + 0.8 (d + 0.04)]
a = S x 10 -6 in/in-deg F; modulus of elasticity, (d+ 0.04)
E = 28 x 106 psi; thermal conductivity, k = 3.19
× 10 -4 Btu/in2-deg F/in; Poisson's ratio, _ (0.8 d + 24.93)
v=0.35. - (d+ 0.04) (b)
A circular tube configuration with an internal
diameter d and a wall thickness t of 0.020 inch
From sample calculation (4-2), Dr= 24.9
is used. The assumption for thickness is to be inches. The factor 0.8 considers the fact that
verified by heat transfer and stress calculations. the tube centers are located on a circle, rather
From equation (4-19), the coolant side wall tem- than a straight line.
perature then is:
For our double-pass design, the coolant
velocity then is
qt 3.00
Twc=Twg--_=l188 x 002 -'rl000o R
3.19 x 10-4 wf
P 827×8 2106
A double pass design is used; i.e., the coolant (c)
V o:N 5d 5= =
passes down through alternating tubes and up 2 4
through adjacent tubes.
For an "up" tube, assume a coolant bulk tem-
perature Tco = 600 ° R at the throat (the more From table3-2: w/= 8271b/sec; p=local
severe case, since the coolant has passed the value of fluid density.
throat region before, on the way down). This RP-1 at 600°R has the following properties:
temperature is well below the critical one and
can be expected to remain so for the remainder # =4.16× 10 -s lb/in-sec
of the passage. Total temperature increase for a
k =1.78× 10 -6 Btu/in2-sec-deg F/in
typical thrust chamber design is of the order of
100 ° P between cooling jacket inlet and outlet. Cp = 0.5 Btu/lb-deg F
The heat-transfer coefficient required to permit
the calculated heat flux for the temperature dif- For RP-1 at 1000 ° R, #w = 0.416 × 10 -s lb/in-sec.
ferential assumed can now be calculated from Now substitute known values and equation (c)
equation (4-20): into equation (a)
/ ., 2106 \0.8 as an interpolation between fuel pump outlet

pressure and injector manifold pressure.
0.0075 d _ 0.0_.14 ×|_.___ n_.___._
s] Combustion gas pressure at the throat
1.78× 10 -6 \4.16 × I0-_/
/0 o.4 ..L.Y
\- I \0.416×i0-V Pg= pt=(Pc)ns(._?_y)y--,
=.-L 1000×0.562=562 psia
\y_ll
11 \o.s
4220 d = 115 000 | ..'-F--.
]
\_al (y= 1.222 from sample calculation (4-2); use
table 1-2).
N : 62.4d-22s (d) Thus maximum tensile stress at the inner tube
wall face can be determined using equation
Substitute equation (d) into equation (b) (4-27):
(1500 - 562) × 0.427

62.4 d -22s - n(0.8 d + 24.93) St-:
0. 020
(d + 0.04)
28×106×8x 10-6×3.0×0.02
+ +6 MA
d= 0.85 in 2 x (1- 0.35) × 3.19 × 10-4 t2
Substitute (d) into equation (b) = 20 000 + 32 500 + 15 000 MA
N= 94.5 = 52 500 + 15 000 MA
Since for two-pass design a whole, even tube

number is needed, the design value of N = 94 is Based on suppliers' recommendation, Fry
:82000 was used for Inconel X at 1000 ° R.
used. Substituting this into equation (b)
d = 0.855 in 82 000- 53 700

Maximum allowable M,4 = 15 000
For the determination of a new tube design, re- = 1.88 in-lb/in

peated calculations, with varied assumptions,
will be required. An experienced designer will From experience it can be assumed that the
require fewer approaches, particularly if test bending moment due to discontinuity in this case
results of prior, comparable designs are avail- will be less than 1.88 in/lb/in. Thus the as-
able. However, even for complicated conditions, sumption of 0.020 in thickness for the tube wall
great amounts of data can be generated in a rela- is sufficient. Summarizing the tube configuration
tively short time if an electronic computer is at the throat:
available.
From table 3-2: d=0.855 in, t=0.020 in, N=94
50.45
p=_ = 0.0292 lb/in 3 A-2 Stage Engine
The fuel, hydrogen, is used as the coolant.
Again, Inconel X is chosen as the tube material.
Substitute into equation (c)
To avoid the "hot shortness" or low-ductility
2106 properties of Inconel X in the range 1200 °-
Vc 0 - 1400 ° F, the mean temperature of the tube wall
94 × (0.855) 2 × 0.0292
must be kept under 1000 ° F (or 1460 ° R). The
= 1051 in/sec or 87.6 ft/sec value for adiabatic wall temperature Taw of the
gas can be calculated using an assumed stagna-
At the throat Pco = 1500 psia is established, tion recovery factor of 0.92.
e
Taw = (Tc)ns x 0.92 = 5740 x 0.92 = 5270 ° R [Dr + 0.8 (d × 0.016)]

Number of tubes N =
(d+0.016)
From sample calculation (4-3) the overall
gas-side conductance at the throat region hgc _ _r(0.S d+ 11.213) (b)
= 0.00520 Btu/in2-sec-deg F. Substitute into (d + 0.016)
equation (4-19), to obtain the heat flux at the
throat: A l_A-pass design is used (i.e., the coolant
enters the fuel manifold at the e = 8 nozzle plane,
q = (5270-1600) x 0.00520 = 3670 x 0.00520 flows to e = 30 and back, then passes through the
= 19.10 Btu/in2-sec throat and combustion chamber zone before it
enters the injector). From table 3-3:
A value of 1600° R will be used for the gas-
side wall temperature Twg at the throat region. _/[= 54.5 lb/sec
From supplier's specifications the following data
is obtained for tnconel X at 1600 ° R or 1140 ° F: Coolant weight flow rate per unit area
a=8.2x10 -6 in/in-deg F; E=24x106 psi; G- _'[ _54.5×4_69.3 (c)

k=3.86 x 10 -4 Btu/in:-sec-deg F/in
N (--_-) _Nd2 Nd2
v=0.35.
For hydrogen at 135 ° l_:

Use a circular tube configuration with an in-
ternal diameter d and a wall thickness t of 0.008 Pr:0.82; Cp=3.5 Btu/lb-deg F;
inch. : 0.367 x 10-6 lb/in-sec
From equation (4-19) the coolant side wall
temperature Substitute these values and equation (e) into
equation (a):
19.I0 x 0.008
Twc : 1600 : 1600 - 396 : 1204 ° R
3.86 x 10-4
0.029 x 3.5 x (0.367 x 10-6) °_
0.0179-
A mean value will be used for the wall tem- (0.82) z_3
perature
r(0 . o.1
1600 + 1204 / (-135.V
55
Twc- 2 - 1402 ° R < 1460 ° R
L do.
or
Assume a coolant bulk temperature, Tco
= 135 ° R at the throat; then, from equation (4-20),
N:3.91d -2-25
19.10 Substitute equation (d) into equation (b)

hc : 1204- 135- 0.0179 Btu/in:-sec-deg F
3.91 d -:as - 7r(0.8 d+ 11.213)

(d+0.016)
Substitute into equation (4-25),
d =0.185 in
0.029 Cp# °a (GO.S) [, '_Tc
° o ss
(a)
0.0179- pr2/a (d°'2) _wJ
Substitute into equation (b)
From figure 4-21, Dr= 11.2 in. N- 7r(0.8 × 0.17 + 11.213)_ 178
The following relationships exist: (0.17 + 0.016)
DESIGN OF THRUST CHAMBERS AHD OTHER COMBUSTION DEVICES 113
Maximum tensile stress is now checked at the From experience, it can be assumed that the
inner wall surface using equation (4-27): bending moment due to discontinuity will be less
At the throat, estimated Pro = 1200 psia; than 0.131 in-lb/in. Thus the selection of 0.008-
Pg = Pt = (Pc)n s (2/(y + 1))Y / (Y- 1) = 800 x 0.554 : 443 inch tube thickness is valid. Summarizing the
psia (¥= 1.213 from sample calculation (4-1)) tube configuration at the throat:
St = (1200- 443) x 0.0925 d=0.185 inch, t=0.008 inch, N= 178.

0.008
24 x 106 x 8.2 x 10 -6 x 19.10 x 0.008

+ As a general design aid, figures 4-31 and4-32
2 (1 - 035) x 3.86 × 10 -4 present construction detail for a typical regener-
MA atively cooled thrust chamber. The chamber

+6
shown is very similar to the one presented in
: 68 750 + 93 900 MA (0"008)2
figures 4-1 and 4-2. Figure 4-31 shows clearly
how the tube shape changes along the longitudi-
Based on suppliers' recommendations,
nal axis. In a typical manufacturing process,
tubes of uniform circular cross-section area are
Fry = 81000 psi for Inconel X at 1200 ° R
first cut to length, then swaged. The latter
operation can best be accomplished by internal
81 000- 68 750
Maximum allowable MA- 93 900 hydraulic pressure in a die. The tubes are filled
with wax, and are then bent (preshaped), in a
=0.131 in-lb/in special fixture, to the thrust-chamber contour.
_ _KnoN_
Figure 4-31.-Typical regeneratively cooled tube wall thrust chamber.
.... ......
Figure 4-32.-Detail of injector manifolding and return manifold of typical regeneratively cooled tube
wall thrust chamber.
The tube is then placed in a die of varying b

cross-sectional area. Hydraulic pressure applied
to the inside of the tube forces it to aline with
the die and to assume final shape. In prepara-
tion for assembly, a trimming process usually
follows.
i
In preparation for assembly into a chamber,
the tubes are arranged on a brazing fixture (core).
For proper brazing great care is required to as-
sure even distribution of the gaps between tubes.
Earlier chamber models were then hand brazed, a
process requiring many weeks and considerable
A. LONGITUDINAL PASSAGE
skill. More recently, furnace brazing has been DOUBLE-WALL CHAMBER B. SPIRAL PASSAGE
successfully applied, drastically cutting chamber- DOUBLE-WALLCHAMBER
assembly time.
Dump Cooling
The dump cooling technique may be particu-

larly effective for applications in hydrogen-
fueled, low-pressure systems (Pc < 100 psia), or
for nozzle extensions of high-pressure hydrogen
systems. A small amount of the total hydrogen
flow is diverted from the main fuel feed line, fed
through cooling passages and ejected. The heat
transfer mechanism is similar to that of regener-
ative cooling. The coolant, however, becomes
superheated as it flows toward the nozzle exit,
where it is expanded overboard at reasonably C. LONGITUDINAL TUBEWALL D. SPIRAL TUBEWALL
high temperatures and velocities, thus contribut- CHAMBER CHAMBER
ing some thrust. Application of dump cooling is

Figure 4-33.-Typical dump-cooled chamber iab-
often limited, however, by various technical dif-
rication methods.
ficulties, such as discharge nozzle design at
low coolant flow rates.
The type of coolant path for a dump-cooled The various constructions differ considerably
thrust chamber is selected to assure maximum
in complexity and fabrication cost. Selection
overall engine system performance. Two possi-
will depend on an optimum tradeoff between reli-
ble paths are shown in figure 4-33:
ability, performance, cost, and weight. The
(1) Axial flow: a one-pass longitudinal pas-
longitudinal passage designs are often employed
sage, using double-wall (fig. 4-33A) or for larger coolant flow rates, as related to the
tube-wall designs (fig. 4-33C). Both are physical size of the thrust chamber. The spiral
open ended, with provision for expansion passage designs are used for smaller coolant
of the dumped superheated hydrogen gas flow rates to alleviate the difficulties in main-
at the nozzle exit.
taining proper flow velocities and dimensional
(2) Circumferential flow: a double-wall de- clearances in the coolant passage.
sign having a spiral flow path for the
coolant and provision for expansion of Because the hydrogen coolant gas can be dis-
the dumped superheated hydrogen gas in charged at relatively high temperatures (1000 c R
the axial direction (fig. 4-33B). A and up), overall engine system performance will
spiral-wound tube design may also prove not be affected appreciably by the dump-coolant
advantageous (fig. 4-33D). flow.
Film Cooling An important advantage of film cooling is the

fact that it reduces heat transfer through the
Porous wall materials, or slots and holes pro-
wall. Consequently, thermal stresses become
vided in thrust chamber walls, can serve to in-
less critical. This is an important considera-
troduce a coolant, a process commonly referred
tion, as thermal stresses may establish the
to as film cooling. Because of interaction be-
feasibility limits of conventional regenerative
tween coolant film and combustion gases, as a
cooling.
result of heat and mass transfer, the effective
thickness of the coolant film decreases in the
direction of flow. In most cases, therefore, Liquid Film Cooling
additional coolant must be injected at one or
It would appear, and has been verified experi-
more downstream chamber stations. Figure 4-34
mentally, that for simply reducing the heat trans-
shows a model of the film-cooling process. The
fer to the wall, film cooling would be more effec-
coolant is introduced through rows of holes. The
tive with the coolant injected as a liquid rather
fluid introduced through row "A" will cover the
than a gas. When the coolant film is liquid, it
wall surface between "A" and "B." Fluid from
should behave essentially as an isothermal heat
row "B" will cover the surface between "B" and
sink, as it evaporates and diffuses into the free
"C," etc. In an optimum design a flow rate from
stream. However, this process results in two-
each row is provided which is just sufficient to
phase flow, consisting of an annular liquid-
cover the area to be cooled.
coolant film and a combustion gas core. This
Although heat protection exclusively by film
effect introduces coolant losses which reduce
cooling has not been applied in the past for the
the theoretical cooling potential. Disturbances
major operational rocket engines, it is signifi-
in the form of capillary waves appear on the
cant that in practice regenerative cooling is
surface of the liquid film adjacent to the com-
nearly always supplemented by some form of film
bustion gases and cause accelerated coolant
cooling. In most instances, a fuel-rich gas
loss.
boundary layer is created by the injection of fuel
The theoretical equation by Zucrow and
from the outermost circle of injector orifices,
Sellers can be used for design calculations of
toward the chamber wall.
liquid-film-cooled thrust chambers.
k|
h| ITaw
Taw
Gc 1 H
HEAT TRANSFER • -- _-_" _o (4-33)
I wg OMBUSTIOH CHAMBER
Gg tic a(l +bCpvc/Cpg)
where
_'FUEL CHAaRER WALL Gc : film coolant weight flow rate per unit
Tc,o area of cooled chamber wall surface,
lb/in _-sec
Gg = combustion gas weight flow rate per

unit area of chamber cross section
perpendicular to flow, lh/in_-sec
_c :-film cooling efficiency
H = film coolant enthalpy
_000 00_ 00_ = Cpvc (Taw - Twg) Btu/lb

Cplc (Twg- Tco)+AHvc '
Cplc =average specific heat at constant pres-
sure of the coolant in the liquid phase,
Btu/lb-deg F
I
I Cpvc = average specific heat at constant pres-
sure of the coolant in the vapor phase,
Figure 4-34..Film-cooling model. Btu/lb-deg F
]16 DESIGNOF LIQUID PROPELLANT ROCKET ENGINES
hg
Cpg --average specific heat at constant pres-
Taw-Twg e-(GcCpvcrlc) (4-34)
sure of the combustion gases, Btu/lb-
"Taw- Tco
deg F
Taw --adiabatic wall temperature of the gas, where
deg R Taw = adiabatic wall temperature of the gas,
Twg =gas-side wall temperature and coolant deg R
film temperature, deg R Twg = maximum allowable gas side wall tem-
Tco = coolant bulk temperature at manifold, perature, deg R
deg R Too = initial film-coolant temperature, deg R
AHvc = heat of vaporization of coolant, Btu/lb e =base of natural logarithms, 2.718
a = 2 Vd/Vm f hg = gas-side heat-transfer coefficient,
b =(Vg/Va) - 1 Btu/in2-sec-deg F
I =applicable friction coefficient for the Gc = film-coolant weight flow rate per unit
two-phase flow between combustion area of cooled chamber wall surface,
gases and liquid film coolant lb/inLsec
Vd --axial stream velocity of combustion Cpvc =average specific heat at constant pres-
gases at edge of boundary layer, ft/sec sure of the gaseous film coolant,
Vm =average axial stream velocity of com- Btu/lb-deg F
bustion gases, ft/sec ?c = film-cooling efficiency
Vg = axial stream velocity of combustion The film-cooling efficiency r/c corrects for the
gases at the center line of the thrust amount of gaseous-film coolant lost into the com-
chamber, ft/sec bustion gas stream without producing the desired
In practice the theoretically determined film cooling effect. Values range from about 25 to 65
coolant flow would be inadequate because of percent, depending upon coolant injection geom-
losses. Therefore, the film-cooling efficiency _?c etry and on flow conditions.
is introduced to correct for this. Liquid-film- The above equation assumes that a balance
cooling efficiency values range from about 30 to exists between heat input and coolant tempera-
70 percent. They are determined experimentally ture rise. The heat input is based upon the gas-
in actual hot firings of a specific design or test side heat-transfer coefficient hg and the differ-
model. ence between the adiabatic gas temperature at
Hydrocarbon fuels have been found to be very the wall and the coolant film temperature. The
effective liquid film coolants. Their effective- heat absorbed is proportional to the heat capac-
ness is attributed to their action as both film and ity of the coolant film from initial to final tem-
deposition cooling agents. As was mentioned perature values. Once equilibrium is reached, no
earlier, these fuels deposit carbon on the wall, heat is transferred to the wall (adiabatic condi-
which serves as an effective heat insulator. tion) and the chamber wall surface will have
achieved the film-coolant temperature corre-
sponding to the various axial locations. Accord-
Gaseous Film Cooling
ingly, the wall-surface temperature will range
With the increasing use of hydrogen, gaseous- axially from the value of initial coolant tempera-
film cooling has become important. Even if hy- ture to a maximum allowable design wall temper-
drogen were injected as a liquid for film-cooling ature, at which point the next film-coolant injec-
purposes, the film between the combustion gases tion station must be provided. It is the specific
and the chamber wall would be heated within a aim of film-cooled thrust chamber design to
very short distance to temperatures above the accomplish cooling with an optimum number of
critical, after which the film would behave as a coolant injection stations.
gas. Figure 4-35 shows an experimental hydrogen
For design calculation of gaseous-film-cooled film-cooled thrust chamber. Cooling is provided
thrust chambers, the theoretical equation of by four film-coolant injecting rings upstream, and
Hatch and Papell can be used. This equation one downstream of the throat. Axial coolant
can be written as follows: injection, in the direction of combustion gas
Figure 4-35.-Experimental hydrogen oxygen, film-cooled thrust chamber.
flow, greatly improves film-cooling efficiency, 0.00061 0.00061= 0.001392 lb/in:-sec

Gc=_- 0.439
whereas normal injection results in the escape,
without any benefit, of large portions of the
coolant into the combustion gas stream. In a
To calculate the heat flux for a regenerative
typical case, the film coolant flow was approxi-
cooling system, with added film cooling, a cor-
mately 3 percent of the propellant.
rected value (Taw) must be used in equation
(4-10) or (4-17). This corrected adiabatic wall
Sample Calculation (4-5) temperature can be determined experimentally
under the specific thrust chamber operation and
For the design of the hydrogen film-cooling
film-cooling conditions. As to the gas-side heat-
system for the thrust chamber of the A-3 stage
transfer coefficient, it was found that there is
engine, the following data are given at the throat
practically no difference with and without film
section:
cooling. Thus, the normal gas-side heat-transfer
hg =0.0011 Btu/in2-sec-deg F coefficient hg can be used in equation (4-10).
Taw = 5240 ° R Note that if a hydrocarbon fuel is used as the
Twg = 1900 ° R max film coolant, the effect of carbon deposition must
Tco =50 ° R be taken into account (eq. 4-17).
Cpvc =3.6 Btu/lb-deg F average
Assuming a value of 0.5 for film-cooling

Transpiration Cooling
efficiency, determine the film-coolant weight
flow rate per unit area of cooled chamber surface Figure 4-36 shows the principle of transpira-
in the throat section. tion cooling. The coolant is introduced through
numerous drilled holes in the inner chamber wall.
Solution
In other designs, the wall may be made of porous
Substitute the given data into equation (4-34): material. In both cases, the permeable chamber
inner liner is enclosed by an outer shell (similar
[ o.oou _ to fig. 4-28), forming a jacket from which the
5240- 1900 e-_Oc×a6×as) coolant emerges. For adequate design, the total
5240- 50 =
coolant flow requirement and coolant weight flow
rate per unit area of cooled chamber wall (lb/in:-
0,00061 sec) must be determined and then implemented
1.554=e\ Oc / by a practical method.
an efficiency factor of approximately 0.85 is

used for calculations.
The porous material used for the transpiration-
cooled chamber walls must be selected and di-
mensioned for correct hydraulic resistance to
render the required coolant flow rate per unit
surface area. It must also be able to withstand
the stresses caused by the pressure differential
between coolant and combustion gases, and
thermal stresses. These requirements impose
Figure 4-36.-Transpiration cooling model. certain limitations on the selection of materials
and on construction methods. The mechanical
Transpiration coolant flow requirements de- design of the coolant distribution system, there-
termined from theoretical equations turn out to fore, is an important factor for successful appli-
cation of transpiration cooling.
be significantly lower than those for film cool-
ing. This is due to the more efficient coolant
distribution. The Rannie equation for transpira-
Ablative Cooling
tion cooling can be used to calculate the theo-
retical coolant flow requirements: Ablatively cooled thrust chambers have many
advantages for upper-stage applications. They
are designed to meet accumulated duration re-
Twg_Tc o- + 1.18(Reb)°l-1 quirements varying from a few seconds to many
minutes. Most designs are limited to lower cham-
ber pressure applications, 300 psia or less.
When assisted by film cooling, or by throat in-
serts made from refractory materials, successful
firings have been made up to a chamber pressure
level of 1000 psia. In general, ablative chamber
construction is rugged, exterior wall tempera-
\Og/ (Reb)°"Prm 1 (4-35) tures are held to a minimum and the cost is low.
Ablative cooling is accomplished by the
pyrolysis of resins contained in the chamber wall
where
material. The thrust chamber construction will
Taw = adiabatic wall temperature of the gas,
vary with mission requirements. As shown in
deg R
figure 4-37, chamber and nozzle are composed of
Twg = gas-side wall temperature, deg R
an ablative liner, a thifi layer of insulation, and
Tco : coolant bulk temperature, deg R (entering)
a high-strength outer shell. The ablative liner
Gc =transpiration coolant weight flow rate
is fabricated from a phenolic-resin-impregnated
per unit area of cooled chamber wall
high-silica fabric which is wrapped in tape form
surface, lb/in2-sec
on a mandrel at optimum orientation. The thick-
Gg :combustion gas weight flow rate per unit
ness is programed as a function of chamber sta-
area of chamber cross section perpen-
tion to provide adequate strength, char thickness,
dicular to flow, lb/in2-sec
and minimum weight for a particular mission. A
Prm = mean film Prandtl number
wrap of oriented phenolic-impregnated asbestos
e =base of natural logarithms, 2.718
is placed on the outer (far) surface of the abla-
Reb = bulk combustion-gas Reynolds number tive liner as an insulator.
Test data from various transpiration-cooling The high-strength outer shell is composed of
experiments have been in good agreement with layers of unidirectional glass cloth for longitudi-
the Rannie equation. However, the equation pre- nal stren_h, and of circumferential-wound glass
dicts coolant flows slightly lower than those filaments for hoop strength. The glass wrap is
required in experiments. It is recommended that bonded with epoxy resin. Aluminum alloy and
P'_ANGI_ FOR RAO_ATIC_ --

C_LED MOZZLE EXI"E_$10N
,,Lgr_ _"_ ....
Figure 4-37.-Ablatively cooled thrust chamber. Figure 4-38.-Ablatively cooled thrust chamber
with throat insert for high chamber pressure
stainless steel also are sometimes used for the applications.
outer shell. The combined thermal resistance of
the ablative liner and of the insulation layer
protects the outer shell and keeps it at moderate a = c [RrRvCpp
F2ktL In (1 ÷ RrRvCp(Taw-Td)-)I°SLP
temperatures.
Figure 4-38 shows an ablative-cooled thrust s] °4 (4-36)
chamber fitted with a throat insert. Both 98 per-
cent tungsten/2-percent molybdenum alloy and where
pyrolytic graphite have been successfully em- a -- char depth, in
ployed as insert materials. The tungsten- c = correction factor based on experimental
molybdenum alloy has given the best results. data for the specific design at the
Although pyrolytic graphite has a much lower throat section, and op a nozzle stag-
density than tungsten alloy, and therefore has a nation pressure of 100 psia
substantial weight advantage, it is vulnerable to Rr = weight fraction of resin content in the
fracture from thermal shock, making design and ablative material
installation critical. The throat insert is in- Rv =weight fraction of pyrolyzed resin
stalled with heavy graphite backups for better versus total resin content Rr
structural results. Insert and backups are bonded Cp = heat capacity at constant pressure of
to the thrust chamber main ablative liner with pyrolysis gases, Btu/lb-deg F
epoxy adhesives. These adhesives have per- p =density of ablative material, lb/in 3
formed satisfactorily up to 500 ° F. Certain k :heat conductivity of char, Btu/sec-in 2-
ceramic materials, such as silicon carbide, also deg F/in
have been used successfully as throat inserts in t =thrust chamber firing duration, sec
space engine applications. Lp :latent heat of pyrolysis, Btu/lb
The design of an ablative thrust chamber for Taw :adiabatic wall temperature of the gas,
a given mission depends on the accuracy of pre- deg F
dicting the depth of char during exposure, and on T d = decomposition temperature of resin,
the soak-back temperature variation in the insu- deg F
lation surrounding the charred portion of the (Pc)as = nozzle stagnation chamber pressure,
thrust chamber wall during and after the hot fir- psia
ing. Test data from hot firings with various Results predicted by equation (4-36) have
ablative thrust chambers indicate that the charring been compared with char depth data obtained
process in the combustion chamber (including from firings of Refrasil-filled phenolic chambers.
throat), that is, the relation between mass pyro- They were found to agree very closely with the
lyzed and heat absorbed, can be expressed by experimental data. However, for areas down-
the following equation: stream of the throat, char depths were found to
be somewhat greater than predicted, and when C=1.05; Rr=0.3; Rv=0.41; Cp=0.38Btu/lb-
using the equilibrium gas temperature. Tempera- deg F; p=0.061 lb/in3; k=9.8x10 -6 Btu/
ture recovery in the boundary layer may be one in2-sec-deg F/in; Lp=686 Btu/lh; Taw
cause for the discrepancy. A modified equation =5060°R; Td=1460°R; b=0.0335
is used, therefore, to predict char depths in the Determine the char thickness at the throat and
nozzle areas: combustion chamber section, and in the nozzle
at station e = 5, after firing for the design dura-
a =bt°_e -°°247_ (4-37) tion of 410 seconds.
where
b = a constant depending upon the nature of the Solution
ablative shield (to be determined experi-
mentally) From table 3-5: (Pc)ns = 100 psia
e = nozzle expansion area ratio at the investi- Substitute this and given data into equation
(4-36). The char thickness at the throat and
gated section
combustion section results as:
e =base of natural logarithms, 2.718
The char-rate analysis is characterized by
physical dimensions and the formation of a char
a=l.05 IO .:x xO.41xO.38xO.061
9.8x I0-6 x 410
layer that progresses from the heated surface
toward the supporting wall. During the pyrolysis
of the resin, the formation of a hard carbona- xln (14 0'3 x 0"41x 0"38
686 (5060-1460))l°_ × (1)°"4
ceous surface of increasing thickness is vital
because it resists thermal and mechanical abla- = 1.05 x [2.82 xln 1.245] °s =0.828 in
tion and chemical attack. At the charring inter-
face, which slowly travels away from the hot Char thickness at nozzle station e = 5, using
chamber gases, a large amount of heat energy is equation (4-37):
absorbed by pyrolysis; i.e., melting and vapori-
zation of the bonding material. As gaseous a = bt°_e - o._47e = 0.0335 x (410) °-s x (2.718) -°.°247xs
pyrolysis products flow through and out of this l
char layer, they control the heat flux to the walls

= 00335 x 20.248 x (2.7182) 8"1
by their own endothermic decompositions, and by
migrationinto the boundary layer. = 0.599 in
No gross dimensional change occurs due to
energy considerationsthroughoutmost of the
thrustchamber; however, mechanical erosionis Radiation Cooling
evidenced in some designs in the throatregion,
Cooling by radiatiofi heat transferis practical
due to the high prevailing shear stresses. At
only forthrustchamber nozzle extensions,where
chamber pressures below 150 psia, throat erosion
pressure stressesare lowest. High metal-wall
is generally reduced. Throat erosion rates vary
temperatures are required to attain the heat
from 0.0005 to 0.00005 in/see
fluxes needed. Assuming negligible temperature
drop through the metal and coatings, if any, the
The adaptation of ablative thrust chamber
steady-state heat transfer for a radiation-cooled
technology to the special field of space engines
nozzle, as shown schematically in figure 4-39,
has been significantly advanced during recent
can be expressed by the following correlation:
years. Approaches typical for this type of engine
will be discussed in chapter XI.
hgc (Taw - Twg) = eaTwg 4 (4-38)

Sample Calculation(4-6)
The followingdesign data are given forthe where

ablativelycooled thrustchamber of the A-4 stage hgc = overall gas-side thermal conductance,
engine: Btu/in2-sec-deg R
Twg
hgc =7.1 × 10 -s Btu/in2-sec-deg R; Taw=4900°R
COMBUSTION
GASES Tow Assuming a total emissivity of 0.95 of outer
wall surface, determine the bulk temperature and
q = hgc(Taw.Twg ) : RADIATION heat-radiated flux.
q = E'O Tw4
Solution
Substitute data into equation (4-37):
7.1 x 10 -s (4900- Twg )

Figure 4-39.-Schematic of radiation cooling.
: 0.95 × 0.3337 x 10 -14 x (Twg) 4
Taw = adiabatic wall temperature of the gas, (Twg) 4 = 22.4 × 109 × (4900- Twg)
deg R
Twg = 2660 ° R
Twg =gas-side wall temperature = bulk wall tem-
perature, deg R Heat flux=7.1 × 10 -s (4900-2660)
= total emissivity of outer wall surface =0.159 Btu/in2-sec
a =Stefan-Boltzmann radiation heat transfer
constant, 0.3337× 10 -'4 Btu/in2-sec -
4.5 INJECTOR DESIGN
(deg R) 4
The design approach to radiation cooling is The function of an injector, which is located
to determine a Twg value which will satisfy both in general, at the forward end of the combustion
equation (4-38) and the structural capability of chamber as shown in figures 4-1 and 4-2, is sim-
the wall material used under operational condi- ilar to that of the carburetor of an internal com-
tions. bustion engine. The injector introduces and
Only alloys which possess short-time strength meters the propellant flow to the combustion
in the temperature range of 2600 ° R to 3500 ° R chamber, and atomizes and mixes the propellants
have been successfully applied to radiation cool- for satisfactory combustion.
ing. A molybdenum alloy containing 0.5 percent
titanium, and a 90 percent tantalum-10 percent
Design Objectives
tungsten alloy appear to have sufficient short-
time strength for use at 3500 ° R. Because of the A great number of injectors have been devel-
low emissivity of molybdenum and also for re- oped and many details of successful injector
sistance against oxidation, a coating of MoSi 2 is designs are now available. However, there still
required on both sides of the metal. Titanium are no hard-and-fast rules to assure a successful
alloys and other commercial alloys, such as design. In the past, most injectors were de-
Haynes 25, have been operated successfully at signed by a trial-and-error approach, with the
2600 ° R. If a temperature capability higher than help of previous test data. While good results
the working range of bare metals is required, have eventually been obtained, it was usually at
insulating coatings of ceramic materials on the the expense of large amounts of time and money.
gas-side wall surface may be needed. Because A more rational approach toward the design of
of their brittleness and coefficient of thermal injectors is through understanding and prediction
expansion relative to that of the alloys, experi- of the chemical and physical processes that are
enced judgment is advised before using these encountered within the combustion chamber, and
coatings for a specific application. using this information as a basis for initial in-
jector design. For a given propellant combina-
tion, the chemical reactions and the kinetics of
stream breakup, mixing, droplet formation, and
The following design data are given for the heat transfer should be studied and clearly un-
A-4 stage chamber nozzle extension at station of derstood, before the approach to the design of an
area ratio = 8: injector is established.
17) DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
There are numerous requirements to qualify a To prevent chamber-pressure fluctuations from

given injector for operational use. The following affecting the propellant flows and thus from in-
are the most important objectives for injector ducing combustion instability, sufficient pres-
design: sure drop through injector orifices must be main-
1. Combustion stability.-In combination with tained. Effective and even mixing of the
a given combustion-chamber configuration and propellants will be achieved through the choice
for a given propellant combination, an injector of a suitable injector impingement pattern. This
should give smooth combustion, during engine will help to minimize accumulation within the
start and stop transients as well as during combustion chamber of unburned propellants
steady-state operation. which could cause local detonations and thus
Depending upon the propellants and their trigger combustion instability. Under certain
ignition characteristics, the arrival sequence of conditions, combustion instabilities of the tan-
oxidizer and fuel streams during start is of great gential oscillation mode can be prevented by
importance. Any accumulation of unburned pro- isolating local detonations by partitioning the
pellants in the combustion chamber prior to igni- injector face into several compartments, as
tion must be prevented to avoid destructive shown in figure 4-40.
chamber-pressure surges. Similarly, during 2. Performance.-Combustion performance of
engine shutoff, chamber overheating and burnout an injector is influenced by: propellant mass
may be prevented by maintaining a fuel-rich mix- distribution; local mixture ratios; degree of
ture. Arrival sequences are best controlled by mixing of injected propellants, in either the liq-
propellant valve timing. Furthermore, minimum uid or the gaseous phase, or both; droplet atomi-
feed-line and injector-manifold volumes between zation and vaporization; rate of heat input; and
propellant valves and injector face will mate- chemical reaction rates. These are predominantly
rially improve propellant sequencing during start a function of suitable manifolding and proper
and stop. selection of injector-hole patterns. The more
L J "!iii',ll_{Ik={l!iI_ijll
[iiJlJ
Figure 4-40.-Baffled injector.

thorough the mixing and uniform the distribution streaks of high mixture ratio (O/F) from occa-
of the oxidizer relative to the fuel, produced by sionally reaching the chamber wall. To offset
the injector, the more rapidly will the combustion this, a special set of fuel holes is often provided
products reach the equilibrium composition nec- at the periphery of the injector, close to the
essary for optimum performance. Although turbu- chamber wall. Excess fuel along the chamber
lence induced by the combustion probably con- wall is thus provided which tends to lower the
tributes a major portion of the energy required for O/F mixture ratio of any errant streak. It also
gas-phase mixing, thorough premixing ot the assists in cooling the chamber wall.
liquid propellants must be accomplished by the 6. Special requirements.-Certain engine sys-
injector if maximum performance is to be achieved. tems are required to operate at off-nominal con-
Furthermore, reaction between certain specific ditions, such as at lower thrust levels during
propellant combinations such as hypergolic pro- throttling, or other than nominal mixture ratios
pellants cannot reliably be initiated and main- as a result of propellant-utilization control. In
tained without it, since the energy released by these cases, injectors must be capable of oper-
liquid-phase reactions supplements the kinetic ating reliably under modified as well as rated
energy available for the process of atomization conditions.
through combustion-gas evolution. In addition,
the heat release from liquid-phase reaction ac-
Injector Configurations
celerates the process of vaporization.
Experience has shown that for a given injec- A typical injector design construction and
tion velocity, propellant-droplet size is reduced propellant-distribution method is illustrated in
with decreasing injector-orifice size. Smaller figure 4-2. Different distribution methods are
droplet size, in turn, results in a higher overall shown in figures 4-41 and 4-42. The injector in
vaporization rate, as a function of increased figure 4-42 uses an integral faceplate. This plate
total droplet surface area. This is true whether is secured to the main injector body by brazing it
the heat of vaporization is supplied internally at the periphery and at posts which are an inte-
via liquid phase reaction or externally by heat gral part of the main body. A fuel compartment
transfer from the hot gaseous combustion prod- is located immediately behind the faceplate, and
ucts. Consequently, injector designs with the fed from an inlet passage. The oxidizer com-
largest practical number of injection elements partment is separated from the fuel by a parti-
can be expected to be the most efficient ones in tion. The fuel is injected through orifices
a given combustion chamber volume. drilled in the faceplate, while the oxidizer is
3. Structural integrity.-An injector should be injected through orifices drilled in the posts.
able to withstand the maximum loads incurred The injector construction for a typical liquid-
during all phases of engine operation. Sufficient bipropollant gas generator is illustrated in figure
cooling must be provided to prevent the injector 4-43. The copper injector body is secured to the
face or any other portion from overheating. stainless-steel outer shell by brazing. The
4. Hydraulic qualities.-The holes or orifices oxidizer inlet forms an integral part of the injec-
of the injector must be designed to effect pre- tor body. Fuel is supplied through a manifold in
determined pressure drops at specific flow rates, the outer shell. In this injector, 2 fuel streams
and to atomize the propellants properly. A low impinge on each oxidizer stream, producing a
injector pressure drop is desirable from the total of 44 impingement points.
standpoint of overall engine-system performance. A variety of injector patterns have been de-
However, minimum pressure drop is determined signed to satisfy the needs of various propellant
from combustion-stability considerations. combinations. In most cases, for good mixing
5. Combustion chamber heat protection.-An the injected streams are made to impinge at a
injector should be designed to avoid formation of predetermined point. The impingement point
hot spots or streaks on the combustion chamber should be as close to the injector face as heat-
wall. Complete mixing of the propellants wilt transfer conditions permit. The arrangement in
prevent oxidizer-rich peak temperature zones which all impinging points are the same dis-
from forming, although this may not prevent tance from the injector face is called uniplanar
!
=_
o_
=i
t.
|._.
'-o
_f!:i
251 ._
N
ii
GIMBAL BEARING ploys nonimpinging oxidizer and fuel streams

- MOUNTING SURFACE which emerge normal to the injector face. It
relies entirely on combustion chamber turbulence
for mixing. While being the simplest to fabri-
cate, the showerhead injector exhibits poor per-
formance in most applications, with the excep-
tion of certain cryogenic propellant combinations.
2. Doublet (fig. 4-44b).-In this design, oxi-

dizer and fuel jets are made to impinge in pairs.
0- Thus good liquid-phase mixing and atomization
is obtained. One of the disadvantages of this
doublet arrangement is that even if the injector
holes have been accurately drilled, the resultant
angle of momentum vector, or beta angle, /_,will
vary with mixture ratio, particularly if a large
impinging angle is used. This variation can
adversely affect combustion performance and
chamber-wall heat transfer. The doublet design
is frequently used in systems using liquid
POSTS oxygen.
| INTEGRAL 3. Triplet (fig. 4-44c).-Two streams of one
propellant impinging symmetrically on one stream
of the other propellant will eliminate the change
of vector angle fl, as a result of mixture-ratio
_i FACE PLATE variations. This arrangement also provides
_T_ m _-- FUEL INLET intimate mixing. Application and propellant
combination will determine whether two oxidizer
SECTION
' A-A_ PASSAGE
OXIDIZER INLET jets will impinge on one fuel jet, or vice versa.
Injectors using this triplet pattern have given
Figure 4-42.-Integral lace plate injector. high combustion performance. These injectors
have been widely used for various propellant
impingement. If two or more different impinging- combinations.
point distances are used (fig. 4-44), the arrange-
4. Quintuplet (fig. 4-44d).-Four streams of
ment is called biplanar or multiplanar impingement.
one propellant impinging on one stream of the
Numerous tests have been conducted to deter-
other propellant in a symmetrical quintuplet pat-
mine impingement-angle and distance effects.
tern provide excellent mixing and performance.
Large included angles will enhance stability,
This design has been applied for various propel-
but can result in some of the propellants splash-
lant combinations.
ing back on the injector face, which can cause
burnouts. Close spacing of the impinging holes 5. Self-impinging (fig. 4-45a).-This pattern,
in a pair has similar advantages and disadvan- also known as a like-on-like impingement, gener-
tages, as has increased spacing between pairs. ally employs self-impinging pairs of fuel and
The satisfactory design, value for the included oxidizer. Mixing is accomplished in the combus-
angle is usually found to be between 20 ° and tion chamber by volatilization of the propellants
45 °. The injector face can be further protected and by turbulence. This design usually provides
against overheating by circulating the propel- good inherent combustion stability, at a moderate
lants on the back side of the faceplate or by performance level. Applications have been suc-
introducing film coolant (propellant) on the cessful for both cryogenic and storable hyper-
surface. Some of the impingement patterns used golic propellant combinations. A modification of
are described below: this design provides for secondary impingement
1. Showerhead (fig. 4-44a).-This pattern era- of the two propellants following self-impingement.
1_ DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
I
7. Ring slot (fig. 4-45c).-The ring-slot injec-

FU(L FUEL
_ANIFaI.D - tor employs concentric pairs of annular slots
• ILrrA
." • A_,_e.. ( which eject the propellants as conical sheets.

¢ The slots are so arranged that fuel and oxidizer
OX_Z[R sheets impinge much in the same manner as in
U,UlIFO_.0 _/
P F pNJI_CTO_
_cz the doublet-type injector.
8. Splash plates (fig. 4-45d).-This injector
(
is designed for good propellant mixing while the
O_ IJ(C_°R_Ac_[JI
p _
_ClE
propellants are still in the liquid state. For this
{al _OWER NEAO {hi OOU_UET purpose, injected propellants are deflected by
the splash plates. The plates will be kept cool
FUEL
C- M,I.mFOLD
by the impinging liquid propellants which do not
_L_L
ignite until they have left the plate.
i 9. Premix.-Figure 4-46 shows a typical
premix-type injector. Fuel and oxidizer are in-
OXOZ_
0 F!
2:2 "-- jected
they are
radially
intimately
into the
mixed
premixing
before
chamber,
entering the
where
I
combustion chamber by a gas jet introduced tan-
gentially at the chamber end. The length and
I_I_FOI.I:I
diameter of the premixing chamber with relation

INJI[CT _ to the mass flow of propellants is critical, as is
(el TRIPLET
the reaction time of the propellants.
10. Throttleable injector.-Certain require-
Figure 4-44.-Injector impinging patterns. ments for space vehicle missions, such as orbit
corrections, rendezvous and docking maneuvers,
= O_ltlt and lunar soft landings, demand engine systems
u#*lrO_.o
capable of thrust control. Figure 4-47a shows a

typical throttlable injector with variable injection-
N
_., ,O*T'"_'"_ slot areas. This is a very effective means of
_J,J
controlling the propellant flows and injector

Ira.==
pressure drops at various engine thrust levels.

N The addition of moving parts, however, causes
o
--injI
design complications. Another approach to a
PUlL throttlable injection system is the "aeration
tb) O0 - AXUU,
method," as shown in figure 4-47b. _ An inert gas
is introduced into the injector propellant mani-
OX_Z_
fold for reduced thrust levels. Through variation

ii,l._C,r live
of the propellant/gas mixture ratio, the effective
I/.OT
density of the propellants can be varied over a

wide range to achieve any desired thrust level
without affecting combustion stability. The gas
can be supplied by the same source used to
pressurize the propellant tanks. This method
m_cro_ _L_L
has increased the range of rocket-engine throt-
O_FCI tling up to a ratio of 100 to 1.
(¢1RI_ SLOT (d) $P{.ASH F_.ATI[
Design Calculations
Figure 4-45.-Injector impinging patterns.
For the design of injectors, various parame-
6. Coaxial (fig. 4-45b).-This injector em-
ters, such as injector pressure drop, impingement
ploys two concentric tubes for the two propel-
lants, which are injected coaxially. _Astron_utics magazine, December 1962, pp. 36-37.
and resultant vector angle relationship, and Injection Pressure Drop

structural loads, can be calculated with reason- The injection pressure drop APi (lb/in _) can
able accuracy. be calculated as
Injection Velocity
The propellant (oxidizer or fuel) injection -XPi='_g :2-_p (4-40)
velocity V (in/sec) can be calculated from the
basic relation: where g is the gravitational constant, in/sec2;
Cd is a dimensionless discharge (velocity and
jet contraction) coefficient which is a function
V= _---- (4-39)
Ap of injector orifice configuration.
The value of this coefficient ranges from 0.5
where g, is the propellant weight flow rate, lb/ to 0.92 and can be determined accurately by ex-
sec; A is the calculated injector orifice area, periment (water flow tests). Injector orifices
in2; and p is the propellant density, lb/in z. with well-rounded entrance and smooth bore give
high values of the discharge coefficient. For a
given injection velocity, a higher value of dis-
-oFUEL _OXIDIZER
INJECTING _NJECTING charge coefficient gives a lower injection pres-
RIFICE ", ORIFICE
sure drop. The rule-of-thumb design value for
injector pressure drop varies from 15 to 20 per-
GAS i_ cent of the chamber-nozzle stagnation pressure.
_ANIFOLD _ ,i..-, ._. nc
N o t_ ;"')';RE_'_'_,._ co,,8os_',o,, Resultant Angle of Impinging Streams
G,sJE'r ...--l,_f'Z_, ( £- /-'-;;,:;:" " -- The angle between the thrust chamber axis
TANGENTIAL [} [ _ %._'%J l_J
ORIFICE " i " x , - and the resultant momentum vector of a pair of
- L,,__F impinging streams is defined as the beta angle,

FUEL OXIDIZER _. By definition when the vector is directed
ItANIFOLD IAANIFOLD
toward the chamber wall, the beta angle is posi-
Figure 4-46.-Prernixing type injector. tive, and when the vector is directed toward the
//// F THRUST
F VARIABLE AREA
OXIDIZER INJECTING
___ ANNULARSLOT FUEL ] CHAMBER
I VALVE
"_ 0XIDIZER THROTTLE__

0
/ MANIFOLD
._ _- CHAMBER
THRUST
THROTTLE/
0 VALVE
OXIDIZER/_'-__'='7_)_ _
/_,,._-- INJECTOR
FACE
INERT GAS
CONTROL
VALVES
MOVABLE / [ /
PINTLE RING-" b"
L/ ANNULARSLOT
FUEL
/_ANIFOLD
(A) VARIABLE AREA INJECTOR (B) AERATION THROTTLEABLE INJECTINGSYSTEM
Figure 4-47.-Throttleable injecting methods.

central axis of the thrust chamber, the beta angle (YoVo

is negative. With a hypergolic-type propellant Rm- _vtV t (4-42)
combination, a positive beta angle (2 ° to 5 °)
tends to increase the combustion performance by where ti, o and wt are weight flow rates, and Vo
causing recirculation and better mixing of the and V! are injection velocities for oxidizer and
liquid propellants along the chamber wall. How- fuel. The injection momentum ratio is a useful
ever, in a cryogenic propellant combination, injector design parameter for the prediction of
where gaseous mixing is predominant, the com- combustion stability and performance of certain
bustion performance will not be noticeably propellant combinations. In the design of oxygen-
affected by the beta angle. A negative beta hydrogen injectors, the value of the momentum
angle should be used in this case to avoid the ratio varies from 1.5 _o 3.5 for liquid hydrogen
possibility of hot streaks on the chamber wall injection, and from 0.5 to 0.9 for gaseous hydro-
caused by excessive heat transfer. gen injection.
The beta angle may be readily calculated,
StructuralLoads
from the principle of conservation of momentum:
The main loads to be considered in the struc-
turaldesign of injectorsresultfrom propellant
wzVt sin a, -w_V2 sin a2 (4-41) pressures behind the injectorface, and in the
tan fl=Ce_V I cos at +_i'2V2 cos a2
manifolds. During steady-statemain-stage oper-
ation,the pressureload on the injectorface is
For the impinging streams shown in figure equal to the injectorpressuredrop:
4-48, al and a_ are the respective angles be-
tween the thrust chamber axis and the streams; Pt = APi (4-42)
¢¢1 and _i'2 are the weight flow rates; and V_ and
V 2 are the injection velocities. The pressure load in the injector manifolds is
equal to the sum of the injector-end chamber
Injection Momentum Ratio pressure and the injector pressure drop:
The injection momentum ratio can be defined
by the expression Pra = (Pc)i + APi (4-44)
where P! is the pressure load on the injector

-/I _ THRUST
face, Pm is the pressure load in the manifold;
APi is the injector pressure drop; and (Pc)i is
CH the injector-end chamber pressure.
During start transients, however, maximum
pressure loads on the injector may be substan-
tially higher than during steady state. When the
wl,v / propellant valves are opened rapidly, propellants
rushing into the empty injector passages can
cause severe hydraulic ram. This pressure load
can be estimated empirically as
Pi =Pm = 4 Pp (4-45)
where Pp is the propellant pressure at time of

valve opening.
,_ "--T-- POINT
¢/ //dHA;ABEk
W2,V2 WALL Using data given in tables 3-2 and 3-3, deter-
mine the injector orifice sizes, injection veloci-
Figure 4-48.-Resultant ang]e of impinging ties and momentum ratios for the A-1 and A-2
streams. engines.
Solution In support of the injector development engi-

neer, who may wish to compare with earlier test
(a) A-1 Engine
data, the injection momentum ratio is determined,
Thrust chamber propellant flow rates are 1941
using equation (4-42):
lb/sec (oxidizer) and 827 lb/sec (fuel); propel-
lant densities are 71.38 lb/ft 3 (oxidizer) and
121×1941
50.45 lb/ft 3 (fuel); injector pressure drops are Rm- 144 x 827 - 1.97
200 psi for both propellants. Based on compo-
nent test results an injector orifice discharge
(b.._)A-2 Engine
coefficient Cd of 0.75 is used for both sides.
From table 3-3, the propel]ant flow rates for
Substituting it into equation (4-40), and convert-
the thrust chamber are 285.2 lb/sec (oxidizer)
ing feet to inches:
and 54.5 lb/sec (fuel); the propellant densities
For the oxidizer side: are 71.38 lb/ft 3 (oxidizer) and 0.72 lb/ft 3 (fuel;
gaseous hydrogen at 180 ° R); the injector pres-
2o0= ! ( 1941 h sure drops are 160 psi (oxidizer) and 60 psi
2x32.2x12x71 38× --_1 _'0"75×A°] (fuel). With the coaxial injecting pattern (fig.
" 1728
4-45b), experimental tests give a value of 0.62
for the oxidizer-side discharge coefficient and a
Total oxidizer injector orifice area: Ao = 32.4 in 2 value of 0.9 for the fuel side. Substitute these
into equation (4-40).
For the fuel side: For the oxidizer side:
200 = ( 227 , 160= 1 ( 285.2 ,_2

2 × 32.2 × 12 × 50.45 × 1-_28 \0''5 ×At) 4
\0.62 A oJ
X
2 x 32.2 x 12× 71.38x. 1^^

lr[_5
Total fuel injector orifice area: At= 16.4 in 2 Total injector oxidizer orifice area: A o =6.43 in 2
An injector pattern of 700 pairs of self- For the fuel side:

impinging streams is used for both oxidizer and
fuel. The following orifice areas and diameters 60: 1 ( 54.5 V
result: 2_32.2_ 12×o.72×----- 1 ko-Y_AJ
32.4 1728
ao = 1-i-_-_= 0.0232 in 2 do= 0.172 in
Total injector fuel orifice area: At= 13.79 in:
16 4
Use a total of 300 coaxial elements for the
a[ =_=0.0117 in2 d[ =0.122 in
injector. The individual orifice areas and diam-
eters will be
From equation (4-39), and using orifice areas
obtained above for available 5P and Cd, mean
6.43
injection velocities are determined: ao = 3--_-= 0.0214 in: do =0.165
For the oxidizer: 13.79

at =--= 0.0459 in 2
300
Wo 1941
= 1452 in/sec or 121 ft/sec
V° = A°P° 32.4 x 71.3_.___8 Use a tube inner wall thickness of 0.025inch.
1728
The diameters for the annular fuel orifice will be
For the fuel:

dr=do +2 × 0.025=0.215 in
wt 827
16.4×-- 50.45 1728 in/secor 144 ft/sec
Vt=-Atpt
1728 dt2=V(4at+dt12)=0.323 in
From equation (4-39), the injecting velocities It is often beneficial if during the hot-firing
are tests of a given injector configuration, certain
operational parameters such as injector pressure
For the oxidizer: drops, thrust chamber shape, and L* can be
changed to determine the effects on performance
wo 285.2 and stability. In such an experimental evalua-
Vo = A-_p ° - 71.38-1076 in/sec or 89.6 ft/sec tion program injector, orifice hole patterns can
6.41×
be redrilled or holes plugged, until an optimual
configuration is obtained.
For the fuel:
Heat-transfer characteristics are an important
flit 54.5 factor when evaluating an injector design.
Vt=_'°r=.r. 1:3 79 × 0.72 =9500 in/sec or 790 ft/sec Temperature-measuring instruments embedded in
1728 chamber walls and injector face are required to
measure heat-transfer rates and to detect local
hot spots. Instrumentation for measuring propel-
From equation (4-42), the injection momentum
ratio lant flows, chamber pressure, and combustion
vibration characteristics are similarly important
for determining the true levels of injector per-
Rm 285'2 x 89"6 = 0.593
- 54.5 × 790 formance and stability.
Experimental Evaluation of Injector Designs 4.6 GAS-GENERATING DEVICES
The design of an injector can be improved In liquid propellant rocket engine systems,
through experimental testing. Three types of gases are required to power the propellant feed
tests are usually employed: hydrostatic pres- systems and other subsystems. Bottled com-
sure, water flow, and hot firing. The hydrostatic pressed gases such as helium can be employed
pressure tests are used to determine whether the for these purposes; however, the use of higher
injector structure will withstand the required temperature gases generated by suitable devices
pressure loads. The water flow tests are used gives much higher overall system performance.
to evaluate the following design characteristics: Gases at temperatures ranging from 400 ° to
1. Effective injector pressure drop.-The data 1000 ° F have been generated for pressurizing
from the water-flow tests can be used to deter- propellant tanks. Gases in the range from 1200 °
mine the orifice-discharge coefficient and to pre- to 1700 ° F are used to drive gas turbines for
dict the injector pressure drop for the design pump-fed systems. Where ever possible, the
propellant, with corrections for density and engine system primary propellants are used for
viscosity. gas-generating purposes in the interests of over-
2. Injection pattern.-Injection pattern and all system simplicity. However, for certain
impingement can be observed, and faulty opera- applications, such as for a pressurized gas-fed
tion can be detected and corrected. system or for starting a turbopump-fed system,
3. Atomization.-Water-flow tests at veloci- high-pressure propellants other than those tapped
ties corresponding to those employed in actual off the primary system are required to supply the
service indicate the quality of atomization to be gas-generating systems.
expected with the actual propellants. Most operational engine systems use special
The true injector operational characteristics, devices for gas generation. However, tapping
such as performance, combustion stability, and hot gases from the main chamber has shown
heat-transfer characteristics for main-stage con- promise for certain applications. The following
ditions, as well as start-and-stop transients, can is a list of design objectives for operational gas
only be fully evaluated by hot firing tests, in a generators:
thrust chamber of representative design or a (1) Ability to produce gases safely, with re-
"workhorse" equivalent. quired properties (temperature, pressure,
nonexplosive) in a compact unit, at the short-duration pressure-fed systems. The tem-

required flow rate• perature of the gases generated by solid propel-
(2) Ability to start and stop smoothly, with- lants is generally in excess of 2000 ° F and is
out abrupt temperature surges, pressure not suitable for uncooled components over ex-
oscillations, or overflow of unburned tended durations• Diluents can be used with a
propellants. loss of basic simplicity.
(3) Ability to operate over a wide range of Figure 4-49 shows the typical design of an
propellant flow rates and (in the case of operational solid-propellant gas generator used
bipropellants) mixture ratios, and to to supply power to the turbine for engine start.
respond closely to the control system. It is built in the form of a cartridge that bolts to
(4) Ability to maintain safe shutdown without a flange at the liquid bipropellant gas generator
complicated purging and draining systems. (shown in fig. 4-51)• The particular cartridge
(5) Ability to restart safely (restartable shown is a disposable type that cannot be re-
engine systems only)• loaded and reused. Upon an engine start signal,
Additional design requirements depend on the two initiators or igniters set off solid propellant
particular engine system involved. combustion.
Gas generators can be classified according to Twenty milliseconds after start, the solid
the propellants employed: grain produces a gas-flow rate of approximately
(1) Solid propellant systems 4.68 pounds per second. It will maintain this
(2) Liquid monopropellant systems flow rate for approximately 1.0 second. This
(3) Liquid bipropellant systems. gas generator operates at 2500 ° F and at a cham-
ber pressure of 1000 psia. The product gas
renders an approximate characteristic velocity of
Solid Propellant Gas Generators
4000 ft/sec. A burst diaphragm located just
Solid propellant gas generators are applied to upstream of the gas generator orifice is used to
liquid propellant engine systems for limited- seal the unit during storage. It is ruptured by
duration applications only, such as for turbine the increased gas pressure at start. Body and
spinners for engine start, or as pressurants for end cap are made of 4130 steel.
|GNrrER MI L-T- 6841

pELL£TS $ PLACES ON TABS
35 GRAMS
Figure 4-49.-Disposable solid propellant gas generator (SPGG).

Solid propellants burn uniformly at the ex- these systems is that they are relatively easy to
posed surfaces at a rate which is primarily a control and that the gases are generated at pre-
function of the temperature and the pressure of dictable temperatures. However, unless the
the surrounding gases. The correlation for a monopropellant is also employed as one of the
given propellant can be expressed as engine system main propellants, the monopropel-
lant gas generator system introduces a third
propellant often requiring special handling and
n: k 1 _1---_0] (4-46) tankage.
Figure 4-50 shows the schematic of a typical
where monopropellant gas generator using 90 percent
R :propellant linear burning rate, in/sec hydrogen peroxide. The catalytic screen pack or
k_ =constant representing the linear burning bed consists of alternate layers of stainless-
rate of a given propellant, at a given steel mesh and silver-plated brass wire screens
initial temperature and a chamber pres- secured by perforated end plates or grids, which
sure of 1000 psia are applied with a preload of approximately 800-

1000 pounds per square inch of bed cross-
Pc =chamber pressure, psia
n :constant allowing for the sensitivity of the sectional area. The length of the catalytic bed
generally ranges from 2 to 3 inches. The allow-
propellant burning rate to changes in
pressure, at a given temperature. able design throughput, that is, propellant flow
The weight flow rate of a solid propellant gas rate per unit cross-sectional area of catalytic
generator can be calculated by bed, is about 0.4 lb/in2-sec. The propellant

pressure drop across the bed can be approxi-
mated by
wg: AbRpp (4-47)
where C1Gb 19s

APb : + C2t (4-49)
wg = weight flow rate through gas generator, Pc 0.9
lb/sec
Ab :burning area, in 2 where
pp = propellant density, lb/in 2 APb =pressure drop through the catalytic
It can be seen that a solid propellant gas gener- bed, psi
ator of constant flow requires a propellant grain Gb : throughput, lb/in2-sec
design which assures constant burning area. Pc = gas generator chamber pressure at the
The correlation between Pc, Ab, and gas gener- end of the catalytic bed, psia
ator orifice area can be expressed as = accumulated running time, sec
C,, C2 :design constants determined experi-
1
mentally for a given bed configuration.
(4-48)
Pc = k2 _--_o]
where
Ao:gas generator orifice area, in 2
k 2 : constant characteristic for a given propeb
lant at a given temperature.
Liquid Monopropellant Gas Generators
Monopropellants such as hydrogen peroxide

(H202) and hydrazine (N2H4) have been used as
gas generants in many applications. They per-
mit a simple generator system and do not require
mixture-ratio adjustments. The advantage of
Sample Calculation (4-9) dizer rich, depending on the propellant combina-

tion. Oxidizer-rich gases tend to accelerate
A hydrogen peroxide monopropellant gas gen- erosion of structural members, while fuel-rich
erator attached directly to the inlet flange of a
gases continue to burn with ambient air after
turbine has the following design data: Turbine
discharge, requiring special measures to prevent
inlet pressure, 340 psia; total turbine nozzle
damage.
throat area, 0.776 in 2. Assume a c* value of
Figure 4-51 and table 4-2 describe a typical
3080 ft/sec for 90 percent H202 and catalytic liquid bipropellant gas generator system. It is
bed design constants C1 =7.2× 104 and C 2 =0.021 designed to produce hot gases using primary
psi/sec. Determine the gas generator propellant engine propellants (LO2/gP-1) for driving the
flow rate wg, catalytic bed area, and pressure turbine of a pump-fed system. The control sys-
drop after 480 seconds of accumulated running tem consists of two normally closed, linked
time.
poppet valves that control the flow of propellants
to the gas generator injector. The valve assem-
Solution bly includes an oxidizer strainer, oxidizer pop-
pet, fuel poppet, timing orifice, actuating piston,
Gas generator pressure = turbine inlet pressure and valve main body. The valve assembly is
=340 psia. Equivalent gas generator throat actuated by gas pressure which forces the piston
area = total turbine nozzle throat area = 0.776 in 2. on the fuel side down to open the fuel poppet. A
Substitute into equation (1-32): yoke integral with the piston actuates the oxi-
dizer poppet. The valve design, through a com-
340 x 0.776 × 32.2
3080= bination of manifold volumes and LOX poppet
#g adjustment, effects a slight oxidizer lead to
prevent detonations, and a fuel-rich cutoff to
Propellant flow rate wg : 2.76 lb/sec eliminate the possibility of turbine burning. The
propellants flow through the poppets to the in-
Use a design value for catalytic bed throughput, jector and into the gas generator combustor, are
Gb, of 0.4 lb/inLsec mixed, and burn within the inner chamber and
combustor body. Ignition of the propellants is
2.76
accomplished by two pyrotechnic igniters. A
Catalytic bed area =-_-.-.-ff=6.9 in 2
gas duct with two flanges opposite to each other
is located at the end of the combustor body.
Substitute given data into equation (4-49): These flanges connect with the solid propellant
gas generator turbine spinner and the turbine
7.2x 104 × (0.4) 19s
Bed pressure drop APb- inlet.
(340) 0.9
The basic design parameters for bipropellant
+0.021x480 gas generators are similar to those for thrust
chambers. The total throat area of the turbine
:63.5+ 10=73.5 psi nozzles may be considered to be the equivalent
throat area of the gas generator combustion
chamber. In calculating combustion chamber
Liquid Bipropellant Gas Generators characteristic length L*, the volume between
This gas-generating system is used more injector and turbine nozzle throats is used, with
widely in liquid rocket engine systems than any a correction factor allowing for the specific
other, since it makes use of the primary rocket design configuration. Because of temperature
limitations of the turbine construction materials,
engine propellants. Bipropellant gas generators
react in the same manner as the main thrust gas generators are rarely operated at gas temper-
chamber, except that the oxidizer-fuel-mixture atures higher than 1S00 ° F. In most designs no
ratio will be adjusted to yield the desired gas cooling is provided for the combustion-chamber
temperatures and chemical properties. The wall and other surfaces exposed to the hot
gases generated may be either fuel rich or oxi- gases.
.4
PORT
PORT
IGNITER (2)
TURBINE
COMB UFIt'OR
Figure 4-51.-Liquid bipropellant gas generator.
The maximum available energy per pound of pressure. This quantity of energy is termed the
gas generator propellants is obtained when the available energy content AHt, and is expressed
products of combustion are expanded isentrop- by the correlations in equations (6-16), (6-17),
ically through a supersonic nozzle to ambient and (6-18).
TABLE 4-2.--Operating Characteristics o[ a OXIDIZER F'U_ P_ v
Typical Liquid Bipropellant Gas Generator

System, as Shown in Fig. 4-51 [
i i
i I T_RUST
Oxidizer .............................. Liquid oxygen
Fuel .................................. RP-1 TURBINE __ _ $
EX.,us'r
DUCT
---._ ___---- _ '-,-._,_
Total propellant flow rate ............... 17.34 lb/see
O/F mixture ratio ...................... 0.342
; t
i /- _OT C,AS \
r_c'r
Oxidizer flow rate ...................... 4.42 lb/sec \ /
Fuel flow rate ......................... 12.92 lb/see
GG chamber pressure (injector end) .......
GG chamber temperature .................
612.1 psia
1200: F
/ \,
Oxidizer-side pressure drop of line, valve,
and in]eetor .......................... 114 psi
Oxidizer side pressure drop across orifice. 121 psi Figure 4-52.-Schematic diagram o[ thrust chamber
GG oxidizer supply line takeoff pressure
gas tapoH system.
(total at main oxidizer pump discharge) .. 846 psia
Fuel side pressure drop of line, valve,
and injector ......................... 216 psi same primary propellants. However, a tapoff
Fuel side pressure drop across orifice .... 80 psi turbine must be designed to operate at an inlet
GG fuel supply line takeoff pressure (total
pressure lower than the thrust chamber pressure.
at main fuel pump discharge ............ 907 psi
Furthermore, a tapoff engine system will require
some sort of simple starting device, such as gas
Thrust Chamber Gas Tapoff Systems
spinner. With the aid of a hot gas regulating
With these systems, combustion product gases valve placed at the turbine inlet, hydrogen tapoff
are bled from the main thrust chamber and ducted engine systems have been successfully throttled
to the turbine, where they are used as the work- to thrust ratios of 10:1 or better.
ing fluid. This arrangement eliminates the need
for a separate gas generator system and contrib-
utes significantly to the simplicity and its relia- 4.7 IGNITION DEVICES
bility potential of the engine system. The tech-
The initiationof the release of the chemical
nique has been successfully developed by North
American Aviation's R, ocketdyne Division and energy stored in liquid rocket propellants is
accomplished by a number of methods from which
appears to be particularly promising for hydrogen
the engine designer will select the most suitable
engines requiring throttling. Figure 4-52 shows
a schematic of a tapoff system. for a particular system:
(I) Igniters
In a separate gas generator system, gases are
(2) Hypergolic ignition
produced and "tailormade" for turbine-power
(3) Catalysts
purposes only, with the benefit of relatively high
All of these methods have been in use for
liquid supply pressures. In a tapoff system, the
bulk of the extremely hot gases of the main both thrust chambers and gas generators. Note
chamber would not be suitable as the turbine that in liquid propellant rocketry, gas generators
drive fluid, in view of the limitation of the tur- have been used not only for turbine power but for
bine construction materials. It has been suc- propellant-tank pressurization as well. The
cessfully demonstrated, however, that by with- selection of the preferred ignition method de-
pends on the chosen type of propellants, whether
drawing chamber gases from the boundary zones
bipropellants or monopropellants are used, on
only, and through proper shaping and location of
the size of the combustion chambers, the heat
the bleedports, any desired turbine inlet temper-
release per unit time required in relation to the
ature (usually less than 1700 ° F) can be reliably
and repeatably produced, for a given thrust cham- amount of propellants entering, and on a number
of other considerations which will be discussed.
ber and injector design. The bleed gases thus
withdrawn have been found to possess thermo- All ignition methods, particularly those for
dynamic properties comparable to the products of bipropellant systems, have one overriding re-
a liquid bipropellant gas generator utilizing the quirement in common: minimum ignition delay.
If the propellants entering the combustion cham- difficult. It is one of the inherent shortcomings
ber are not promptly ignited, explosive mixtures of solid propellants that they cannot be switched
can form and detonate with damaging results. on briefly for checkout and then stopped again.
The assurance of faultless ignition is dependent
on the selection of the ignition method, the
quality of design, and on adequate heat release.
Igniters
These are defined as devices which release

heat and thus initiate the reaction of the main
propellants which subsequently sustains itself.
Igniters derive their power from an outside
source or from a limited amount of energy stored
as solid propellants within themselves. Follow-
ing ignition, igniters do not participate further in
the combustion process. Some of the principal
igniter types are discussed below.
Pyrotechnic Igniters
These are literally slow-burning fire crackers,

somewhat modified for rocket engine application.
Burning time is in the range of a few seconds.
For thrust chamber use they can be mounted to
the injector face or inserted from below at the
end of a wooden or plastic stick (fig. 4-2). For
better heat distribution, multiple units firing in
different directions, have been used, as a rule
radially outward from the center across the in-
jector face. In other designs they have been
mounted to pinwheels, achieving distribution
through rapid rotation.
Although pyrotechnic igniters are used, they
should be considered obsolescent• To achieve
adequate heat release for modern large engines,
their size becomes impractical. Also, ejection
of their inert parts can cause damage to the
delicate thin-walled tubes of modern chamber
walls. Under cryogenic conditions, they have
exhibited a tendency to cause ignition delays,
complete duds, or explosive popping.
Pyrotechnic igniters for gas generators and
small thrust chambers have been mounted in
recesses as screw-in-type plugs. (See figs. 4-53
and 4-54.) The igniters are initiated by electri-
cally triggered squibs, of which there are a
variety of types in use. The need to connect
wires to the pyrotechnic igniters is another in-
convenience. Furthermore, checkout of the Figure 4-53.-Radially outward tiring pyrotechnic
integrity and readiness of pyrotechnic igniters is igniter [or center ot injector mounting.
Main chamber: centrally mounted unit;

propellant weight, 33 gms; 4.5 sec
burning time; heat release, 45 Btu/sec;
perchlorate-type fuel. Electrically initi-
ated. (Later models of this engine used
hypergolic slugs.)
Gas generator: Pyrotechnic: propellant
weight, 2 gins; 8 sec burning time; heat
release, 1.6 Btu/sec; perchlorate-type
fuel.
Vernier engines (earlier models only):
Similar to gas generator units.
It is vital that certain of the specifications
fall within a stated band. For instance, it can
be specified that in a family of samples no
igniter will fire at currents below 1 amp, and
that all must fire below 4 amp. The first condi-
tion concerns the avoidance of accidental firing
due to stray currents introduced by extraneous
RF signals or other sources; the latter is simply
a reliability requirement. Similarly, minimum
and maximum burning times, i.e., tolerances of
the nominal times, will be dictated by start-
sequence conditions.
In operational designs, the pyrotechnic ignit-
ers frequently do not ignite the main propellants
directly, but ignite a pilot flame fed by a small
portion of the main fuel. The pilot flame then
ignites the main propellants.
Hypergolic Igniters
The term "hypergolic" was coined by the

German chemist Noeggerath about 1942 and is
composed of elements of the Greek words for
"high energy liquid." The term now denotes a
bipropellant combination which ignites spontane-
ously when the two components meet. Such a
Figure 4-54.-Gas generator igniter with built-in
[usible link. system was used as an ignition source for the
earlier German A-4 engine (later called V-2
Assurance of their reliability, therefore, is by engine), utilizing hydrazine-hydrate (N2H4- H20)
statistical and sampling methods. For opera- and 80 percent hydrogen peroxide (H202). Thin
tional application, additional safety margins are tubes supported by a wooden stick were inserted
secured by redundancy. However, this tends to into the thrust chamber from below. Upon an
increase bulkiness further. Also, pyrotechnic "ignition" signal, a ground-mounted supply unit,
igniters are unsuitable for repeated starts. including remotely operated valves, fed the two
Engine manufacturers have been procuring components to near the injector elements where
pyrotechnic igniters from sources specializing in they burned with a spontaneously igniting hot
this field. Typical examples of selected param- flame. This method may also have been used
eters are as follows: temporarily on other systems. However, its
For a 150000-pound thrust, LOX/RP-1 engine: clumsiness, the frequent clogging of feed lines,
and the need to eject a considerable amount of Typical of fluid hypergolic with oxygen is
inert solid material made it undesirable. Mso, triethyl aluminum. For optimum behavior, i.e.,
adaptation to repeated starts would be complex for minimum ignition delays and avoidance of
and would require vehicle mounting, thus adding undesirable deposits in fuel and sensing lines,
inert flight weight. optimum mixtures of the two have been success-
fully established experimentally. In a typical
application, an amount of 6 cubic inches has
HypergoUc Slugs
been found adequate, although 9 cubic inches
A more elegant way of using the hypergolic are actually used for maximum safety margin. By
effect for main-propellant ignition is through use comparison, a 1.5-million-pound thrust engine
of a hypergolic slug. In this design a small uses 35 cubic inches. The hypergolic slug
amount of fluid is used which is hypergolic with method, first explored at the German Peenemtinde
one of the main propellants but not with the installation, is well developed and has found
other. The fluid is stored in a cylindrical car- wide application. Limited design and develop-
tridge which has burst diaphragms at either end. ment work has been done more recently toward
The cartridge, in turn, is loaded into a housing repeated-start units, using a device resembling
which is part of a bypass line paralleling a high- an automobile brake master cylinder with its
pressure main propellant feed line (fig. 4-55). If replenishing features. However, the hypergolic
a fluid is chosen which is hypergolic with the slug is truly a single-start device. Because of
oxidizer but neutral to the fuel, it is installed in relative bulkiness, the hypergolic slug is not
the fuel system, and vice versa. The former recommended for small units such as gas gener-
type is the more common one. Here a fuel by- ators, where pyrotechnic igniters or spark plugs
pass line feeds an injection element in the cen- are preferred.
ter of the injector, or a set of elements evenly
Spark Plugs
distributed over the injector face. When the
pumps start and outlet pressures rise, the oxi- Spark plugs and their accessories have been
dizer valve is opened. As pressures rise further, developed to high levels of efficiency and relia-
the burst diaphragms in the hypergolic-fluid con- bility for liquid rocket engine use. They are
tainer rupture and the fluid meets with the oxi- eminently suitable for repeated starts. For
dizer in the chamber, igniting spontaneously. direct ignition, however, they are confined to
The fuel following the slug sustains the ignition relatively small combustion devices. (See fig.
flame. The main fuel valve is then opened and 4-56.) In a typical 200000-pound thrust engine,
all parameters reach main-stage level. Since the the gas-generator spark plugs fire at the rate of
igniter elements carry fuel fed from the main 50 sparks per second, releasing approximately
source following ignition, they continue to par- 1/10 joule per spark. This corresponds to 5
ticipate in the combustion, undistinguished from joules/sec or 0.005 Btu/sec per plug. The
the remainder of the main injector. (See sche- efficiency of spark generation from the electrical
matic, fig. 2-13.)
Figure 4-56.-Spark igniter assembly. At right,

Figure 4-55.-Hypergol slug cartridge and housing, screw-in spark plug; at le[t, cable connector.
storage device is approximately 20 percent. The main stage, which prevents any of the main
overall efficiency of the spark ignition system is chamber gases from backing up into the igniter.
approximately 10 percent. Thus, 100 watts of Approximately one-half percent of the main pro-
electrical power is required. If a 28-V source is pellant flow rates has been found adequate for
used, the nominal current is 4 amp when two the augmented spark igniters.
redundant systems are used in parallel. A Typi-
cal voltage at the spark plug is 15 000 V. Special Designs
For the engine used in the Saturn S-I booster,

Augmented Spark Igniters gas-generator ignition by the solid-propellant
turbine spinner has been successfully developed.
The limitationof directspark ignitionto The turbine spinner will be discussed in a later
small units has led to the design and develop- chapter.
ment of augmented spark igniters(ASI). In this
design a spark plug,similarto the above, fires
into a small chamber about the size of a gas Hypergolic Main Propellant Ignition
generator. A small amount of the main propel- In preceding paragraphs it was learned that
lants is fed intothischamber where they ignite,
hypergolic fluids are being used as ignition
The hot flame generated in turnignitesthe main sources for main propellant combinations which
propellants, Figure 4-57 shows the principles of by themselves are not hypergolic (i.e,, "anergol"
an earlier augmented spark igniter design. Per-
propellants). Hypergolic liquid main propellants
manently located at the injector end of the thrust
have attracted attention since the early days of
chamber, this igniter directs its products of com- modern rocketry. Their use permits a substantial
bustion across the face of the main injector. It
simplification of the engine system through elim-
is film cooled by the tangential injection of fuel. ination of the entireignitionsystem, leaving
The oxidizer is injected through two copper
entry timingthe only functionalrequirement.
tubes which impinge at the centerline of the
This gain however, is not entirelywithoutpenal-
igniter, resulting in a hot-core type of combus- ties. The practicalhypergolicpropellantcom-
tion, The igniter is made of 4130 steel and has
binationshave a somewhat limitedspecificim-
a convergent, throatless nozzle. This results in
pulse. Furthermore,some are highlycorrosive
a wide dispersion angle of gases emanating from and/or pose handlingand storageproblemswhich
the nozzle. The igniter is capable of an unlim- the engine designerhas to consider.
ited number of starts, since the spark plug is so
A number of hypergolicmain propellantcom-
located that the combustion does not seriously binationshave been in successful operational
affect its life. The igniter has proven operable use for many years. During World War II, several
over a wide range of mixture-ratio and pressure guided-missile systems using hypergolic propel-
conditions. It continues to operate throughout lants, were under development in Germany. Hydra-
zinc hydrate (N_H4 •H20) and high-percentage
FUEL I_ILET SPARK MDNITOR TERmINALised. hydrogen peroxide were used in certain Messer-
_;P J, RK PLUG
schmitt rocket fighters and the antiaircraft mis-
sile Enzian. Amines and nitric acid were used
forthe AA rocketSchmetterling.Optolinesand
nitricacid with sulfuricacid additives(approxi-
inatelyI0 percentwere appliedinthePeenemiinde
developments of the antiaircraft rocketWasserfall
(17000-pound thrust)and the small 1300-pound
Taifun. "Optoline"was a generic term forvari-
ous mixturesof aniline,hydrocarbons,and other
substances.
In the United States,severalpropulsionsys-
tems utilizing hypergolicpropellantshave been
Figure 4-57.-Augmented spark igniter. developed.
To make available to the high performing but TURBINE

FUEL PUMP
anergol propellant combinations the simplicity of (RP-II
hypergolic behavior, the effectiveness of small

amounts of additives ("sweeteners") has been
successfully demonstrated. GAS GENERATOR
(SOLID CATALYST}
GAS
GENERATOR
Catalysts VALVE
In a general sense, catalysts are not igniters THRUST

CONTROL MAIN
but initiators and sustainers of reactions, which VALVE OXIDIZER
VALVE
themselves remain unchanged during these reac-
tions. In rocketry, catalysts have been used
predominantly to initiate and sustain the com-
position of monopropellants ("Monergols"), nota-
bly that of hydrogen peroxide. Several opera-
MAIN FUEL
FUEL
tional or near-operational systems existed during VALVE INJECTOR
World War II, such as the earlier Messerschmitt

Me-163 which used hydrogen peroxide with SCREEN PACK
(SOLID CATALYST) THRUST CHAMBER
potassium permanganate solution as the catalyst.
Probably the most widely used application of
this principle during that period was for the tur- Figure 4-58.-Schematic of a Rocketdyne AR-1
bine steam-generating system of the German A-4 superperiormance rocket engine.
(later called V-2) ballistic missile, which em-
ployed 80 percent hydrogen peroxide with either the decomposition gases (1400 ° F), the RP
potassium permanganate or sodium permanganate ignites and burns spontaneously with the free
as catalysts. oxygen of the decomposed H202. (See fig. 4-58.)
Because of the need for relatively elaborate In this manner, the solid catalyst indirectly
timing, valving, and interlocking devices, the serves as an ignition system. While the specific
use of liquid catalysts was soon found to be impulse with RP afterburner is still moderate
cumbersome and undesirable. Application of (approximately 245 seconds for the AR), these
solid catalysts, therefore, as they were being systems offer great versatility, storability, and
used for underwater torpedoes, led to design and extreme simplicity, including throttling to low
development work toward use of these systems levels and restartability.
for rocket application. They never reached More recently, it has been successfully dem-
maturity for the German World War II systems, onstrated that catalytic operation offers a simple
but were perfected after the war by the British alternative to augmented spark ignition for hydro-
and to a limited degree by the United States. gen systems.
The Redstone rocket steam plant, using solid In a process patented for Engelhard Indus-
catalysts, has consistently and successfully tries, Newark, N. J., a gaseous mixture of oxygen
operated in many flights, among them the first and hydrogen is fed through a catalyst bed of
U. S. manned rocket flight by Commander Shepard. palladium-impregnated alumina (A1203) pellets,
Another successful development are the AR air- by which the mixture is ignited. Installation of
plane superperformance rockets. Analogous to this igniter is similar to an augmented spark
certain British systems', the AR systems decom- igniter.
pose hydrogen peroxide fed through a solid cata-
lyst bed consisting of impregnated wire screens.
Ignition Detection
Since the specific impulse of decomposed hydro-
gen peroxide alone is low (below 200, depending The reader familiar with the news stories
on concentration and design parameters), RP fuel about rocket launchings over the past years is
is injected below the decomposition chamber. well aware of the consequences of rocket stages
Because of the sufficiently high temperature of failing to ignite: loss of mission. This is
equally true for liquid- and for solid-propellant problem increased because of the large amount
systems. With the former, however, an additional of oxidizer present in full flow ignitions which
hazard exists in case of ignition failure: that of shrouds the ignition flame. Thus, means had to
accumulation of explosive propellant mixtures be found to detect ignition by other means.
which can be accidentally set off with cata-
strophic consequences. This consideration has
Optical Detection
always been a concern with unmanned vehicles,
but has become even more important for manned Ground-mounted optical devices can be moved
ones.
up close to the chamber exit. A number of types
Recognition of these potential dangers has have been investigated, such as simple light or
prompted extensive investigation of means to infrared-sensitive cells. They were found, how-
detect reliably absence or presence of ignition ever, to be subject to the limitations mentioned
in liquid propellant rocket engines. Only upon for human observers. It is possible to mount the
an "ignition OK" signal should the engine-start optical devices into the chamber wall facing
sequence be permitted to proceed. This refers toward the inside near the injector face; how-
mainly to the thrust chamber. For gas gener- ever, the devices used thus become vehicle
ators, redundancy appears to be adequate pro- mounted and require interfaces to ground-support
tection for most applications. Desirable detec- equipment. Also, "windows" in the chamber wall
tion systems must judge ignition both qualitatively represent undesirable surface discontinuities. It
(absence or presence) and quantitatively (ade- is unlikely, therefore, that optical devices will
quate heat release). Not all methods are equally find wide application for ignition detection.
good in both respects.
In some form or another, the engine designer
Pyrometers
will have to provide means for ignition detection.
A survey follows of several which have found Heat-sensitive pyrometers are closely related
operational application. to the optical devices and subject to the same
limitations.
Visual Detection
Fusible Wire Links
For the German A-4 (V-2) and the early U. S.
Redstone missiles, visual observation by the For many applications these are simple and
test conductor was used. Man in this case was reliable devices. A wire is strung across the
the interlocking device and would initiate the chamber exit which, when fused by the ignition
next sequence step only if, in his judgment, flame, interrupts a circuit and signals "ignition
ignition was adequate. This simple procedure OK." Through proper selection of wire gage,
was satisfactory because these early systems material and distance from the chamber exit
employed a prestage, during which the main pro- and/or center, some quantitative judgment is
pellants were admitted under tank head only. obtained. The wire can be ground mounted or
The resulting relatively low flow rates were then chamber mounted. It must be isolated and should
increased by starting the turbopump upon a have spring loading, like the well-known electric
"prestage OK" signal. fuses, to assure positive separation.
With the increasing size of modern rocket Wire links have a number of shortcomings.
engines, visual observation became unreliable. The fused wire ends may touch other metal parts
The type of installation of these systems in and thus reconnect the circuit before the relay
static firing stands and on launch tables made drops out. Suitab)e circuitry and mounting must
direct observation difficult. Furthermore, im- therefore be applied. If a pyrotechnic igniter is
proved igniters, developed to keep the ignition used, the wire can be broken by inert particles,
flame concentrated where it should be, i.e., at or even by a dud igniter coming out of the cham-
the injector face, resulted in little or no visible ber, giving an incorrect "ignition OK" signal.
fire emerging at the chamber exit. With the dis- This has been overcome by providing redundancy
appearance of the prestage step, the visual using several wires in parallel, all of which
must be broken before the sequence can proceed. senses pressure buildup in the igniter injection
In another design the wire has been mounted as line upon rupture of the hypergolic fluid cartridge
a loop placed in a groove on a wooden or plastic burst diaphragms. The switch signal then initi-
stick. It is thus supported against all reason- ates the next sequential step. A modification of
ably expected mechanical damage and adequate the system substitutes a pressure-actuated valve
insulation is maintained after fusion. for the switch with similar effects. This method
does not assure, however, that the cartridge is
properly filled with the right amount of the cor-
Pressure-Sensing Devices
rect fluid.
Because of the need to mount the fusible Spark igniters use electric devices which
wires at the exit of the thrust chamber, they are ascertain that the plug is sparking based on
subject to some of the limitations noted for vis- conductivity effects due to ionization near the
ual and optical methods. It has been attempted, electrodes.
therefore, to sense the pressure rise in the com- The methods in the above list, which un-
bustion chamber resulting from the burning igniter doubtedly is not complete, are described as in-
flame. However, since the pressure rise is small direct because none of them directly and reliably
(a few psi at best), reliable discrimination is detects ignition; i.e., the release of adequate
difficult. Furthermore, the sensing-pressure heat. This is a drawback and cannot entirely be
switches must be able to withstand the much offset by weighing, certifying, and inspecting.
higher pressures during subsequent main stage.
Pressure-sensing devices have potential for 4.8 COMBUSTION INSTABILITY
multistart engines.
"Combustion instability" is defined in terms
of amplitude of pressure fluctuations in the com-
Resistance Wires
bustion chamber. Chamber-pressure fluctuations
Another method designed to overcome the are always present during normal, stable opera-
shortcomings of fusible wire links is the appli- tion of a rocket engine system. These fluctua-
cation of resistance wires. Constructed like a tions are generally quite random, showing fre-
glow plug and connected to a bridge circuit, the quency spectra which are essentially continuous
resistance wire will signal by a distinctly differ- in nature, with few, if any, recognizable peaks.
ent resistance in the presence or absence of However, in case of instability, large concentra-
ignition. The art is to find that spot in the tions of vibratory energy appear at one or more
thrust chamber or gas generator which experi- frequencies in the spectrum. They can easily be
ences a clear temperature rise as a function of recognized against the normal random-noise
ignition, yet remains cool enough to prevent backgmund.
fusion of the wire. Resistance-wire sensors are It has been found experimentally that the
ideally suited for repeatable start engines. amplitude of the chamber pressure oscillations
which will cause detrimental physical or opera-
tional effects varies widely for different thrust
Indirect Methods
chambers and engine systems. Also, in a given
In conjunction with hypergolic slug ignition, chamber or system the effects of various types
other approaches to ignition sensing have been of instability can be quite different at the same
developed. In one design an electric contact amplitude. Thus itis difficult to assign a quan-
assures that a cartridge is actually installed. titative value to the amplitudeat which the com-
This does not assure, however, that the cartridge bustionchamber should be considered as running
is loaded or completely loaded, nor that the unstable.
downstream lines are not clogged or that the Itis an interestingobservation thatthe first
diaphragms will burst. Weighing of the cartridge large liquid-propellantrocket-propulsionsystem,
and purging of all lines must be included in the the German A-4 (V-2)rocket,never experienced
firing preparation. combustion instability in over 4000 launchings
In another arrangement a pressure switch and in severaltimes as many staticchamber and
engine firings. The term "combustion inst.abil-

ity" was literally unknown. It has been sug-
gested that the Germans may have had instabil-
ity, but did not know it because of their poor
Lc
high-speed recorders. To this it can be said
that for those instabilities which have caused
today's real difficulties, the high-speed recorder
chart is merely a postmortem confirmation of the
cause of often very costly failures, which did
not require any special instrumentation but'were
unmistakable even to the untrained visual ob- i
server. It is much more likely that there is a LONGITUDINAL TANGENTIAL RADIAL
relationship between the low performance level MODE MODE M__

of the A-4 or the chamber geometry, or both, and Ac 0.59A c L22A c
N'E'g c N= dC N=- dc
the inherent stability.
Experimentally it has been found that as long Figure 4-59.-Three modes o[ instability. Lc =
as the ratio of the peak-to-peak amplitude of combustion chamber length (injector [ace to
pressure oscillation to average chamber pressure throat); dc = combustion chamber d i a m e t e r;
is kept below 0.10, there is usually no physical N = normal acoustic [requency; Ac = velocity o[
damage to the chamber. However, while a 10- sound in chamber.
percent variation in pressure for any class of
instability may not appear to have detrimental combined, of both feed system and chamber.
physical effects instantly or within a short This will be further discussed in connection
period, it would be unacceptable for longer range with methods to improve stability.
rocket vehicle missions. One practical way to The effects of the oscillations on an engine
detect combustion instability and to prevent it system are very much dependent on frequency.
from causing damage during engine operation is These effects may range from simple shaking
by monitoring the vibratory acceleration of the (usuaLly at the lower frequencies), possibly re-
system. Accelerometers are mounted on the sulting in an eventual mechanical failure after
system to monitor in all three coordinates. They sometimes prolonged exposure, to "acoustic"
are connected to discriminator circuits which are vibrations (usually at the higher frequencies)
set to specific g-load limits. When these limits capable of destroying the entire system in a few
are exceeded, a counting instrument begins to hundred milliseconds. But, how low is "low _
record. As soon as an allowable cumulative and how high is "high"? As pointed out earlier,
number of oscillations is exceeded, an engine the general field of combustion stability is ex-
cutoff signal is automatically triggered to pre- tremely complex and it would be far beyond the
vent damage. scope of this book to attempt to present a gen-
eralized theory of the subject. Each system,
because of configuration and dimensions, be-
Types of Instability haves somewhat differently and requires special
For proper remedial action, it will often be treatment. The thrust chamber designer must
important to know whether the observed oscilla- have a basic understanding of the stability prob-
tions are of a longitudinal, radial, or tangential lem, and it is felt that this can best be conveyed
mode, or a combination of these. These three by describing it in terms of a typical system, for
modes and their normal acoustic frequency are which a substantial amount of experimental data
indicated in figure 4-59. Furthermore, in certain exists.
cases, it will be extremely valuable to know The frequency of the chamber pressure oscil-
whether the oscillations originated in the thrust lations in a given chamber is determined by the
chamber, or in the feed system, or whether they geometry of the system as well as by complex
originate in an interaction of characteristics, interactions between the fluid flow in the pro-
harmless when separate but destructive when pellant lines, the physical and chemical process
_d
i
eliminating them once they are initiated. They

do not occur at, nor can they be generally damped
to, low-amplitude levels. They are either pres-
ent at high amplitudes or not at all. It is be-
lieved that these oscillations are predominantly
of the radial and tangential types.
In many systems extremely unpredictable
high-frequency instability has occurred. It is
often, but by no means always, connected with
the buildup phase to main stage. Systems which
ran stably during numerous successive tests can
become unstable without warning or subsequent
clear indication of a cause. Figure 4-61 shows
Figure 4-60.-Approximate vibration characteris-
the starting of a typical high-frequency instabil-
tics at 150000 lb thrust level.
ity, indicated by the sudden shift in the acceler-
ometer trace due to high-amplitude chamber pres-
of combustion, and the dynamics of the combus-
sure oscillations. All types of rocket propulsion
tion gases in the chamber. It has been found
that each of the frequency components in the systems, including solid systems, have been
plagued by high-frequency instability. A consid-
instability spectrum of a thrust chamber is pre-
erable amount of research and engineering has
dominantly influenced by only one physical
been devoted to the explanation and elimination
process so that it is possible to group the ob-
of this phenomenon.
served instabilities in broad general classes,
Unless a run in which instability is encoun-
which are: high frequency or gas dynamics, low
tered is terminated within fractions of a second,
frequency or hydrodynamics, and intermediate
serious damage to the engine hardware almost
frequency or combustion dynamics, listed in the
always occurs. It is assumed that the rapid gas
order of their relative importance. Figure 4-60
pulsations directly interface with propellant
presents vibration amplitudes of various fre-
injection, their mixing, and with the combustion
quencies versus mixture ratio of a typical L02/
RP-1 engine at the 150000-pound thrust level. process, upsetting the condition in the boundary
The graph indicates the large difference in vibra- zones, in particular at the injector face, to such
tion amplitude between a stable and an unstable an extent that the heat transfer to the metal parts
increases at a high rate. Within seconds, or
7egion.
even fractions thereof, the injector can burn
through, permitting propellant mixing behind the
_ligh-Frequency Instabilities injector face. This, in turn, leads to explosions
which often completely destroy the system.
High-frequency instabilities at frequencies of
It has been observed that the degree and
approximately 1000 cps and over are sometimes
speed of damage is somewhat related to the
referred to as "damaging acoustic," or "scream-
level of energy release occurring in the combus-
ing" modes of instability. They are gas-dynamic
instabilities which are both sustained and initi- tion chamber. This may explain why "bursts of
ated by the combustion process and are believed

to be concentrated in the uppermost portion of I HIGH-FREQUENCY
the combustion chamber where they cause in-
J"-- ,NSTABILITY--
creased heat-transfer rates to the injector suf- i T _ I _y, : !I
ficient to melt and burn it through within a few
hundred milliseconds. They also frequently have
serious damaging effects upon other parts of the
rocket engine system. High-frequency instabili- , ENG!N,E,ArCCELEROMETERI
I , )I !
ties are further characterized by instantaneous
initiation (a few milliseconds from absence to Figure 4-61.-High #equency combustion insmbil'-
full amplitude), and by extreme difficulty in ity shown on oscillograph for engine accelerometer.
instability" during shutdown cause damage less

frequently than they do during buildup and main
stage.
tRtSSaRt - --'-- " -- - - - -- - "
Low-Frequency Instabilities
This type of instability, at frequencies below

approximately 180 cps, is a hydrodynamic oscil-
lation, characterized by a cause-and-effect-type
coupling between combustion process and pro-
!
pellant feed system flow processes. The phe-
nomenon is sometimes referred to as "chugging." _o_ P_[ss_[ II_ _'_1" l_q1_1tl_lllT_l_[IZXl"l_l_11111_111_11'l_ll_11
Its secondary effects can be serious indeed.

"Chugging" may trigger destructive high-frequency
_tssun[
instability. Also, prolonged chugging can lead
to loosening of bolts and other vital connections
and to ruptures in general. Low-frequency in-
stability is self-sustaining but may damp out.
As a rule, it is predictable from analytical and Figure 4-62.-High-speed pressure measurements
from test result studies. In figure 4-62 the high- of low-frequency instability.
speed-pressure instrumentation measurements
indicating chugging clearly show that the oscil- bustion chamber and of the propellant ducts, and
lations of propellant feed system pressures are the magnitude of the propellant flow rates and
at the same frequency as the "rough combustion their ratio to one another (in a bipropellant sys-
cutoff accelerometer" reading. tem), are critical to the phenomenon of low-
Chugging occurs most frequently during frequency system oscillations.
buildup and shutdown of an engine system, or
when operating at off-rated operating levels,
Intermediate-Frequency Instabilities
such as at incorrect mixture-ratio values. Like-
wise, too high, and particularly, too low a thrust This instability, with frequencies ranging
level can lead to chugging. This is especially from 200 to 1000 cps, is sometimes referred to
important for systems requiring throttling to a as combustion dynamics or "buzzing." It is
lower-than-rated thrust level during flight. If characterized by a spring-and-mass-type coupling
sustained, chugging will cause measurable per- between combustion process and propellant feed
formance losses, which are attributable to widely system flow processes. It is often present in
fluctuating mixture ratios. only a portion of the feed system, or is confined
The chugging phenomenon is frequently asso- to the combustion chamber, or, in a bipropellant
ciated with the quality and promptness of igni- system, to one of the two propellant systems
tion of the entering propellants. This can be only. It appears to be initiated by the combus-
described as "flame holding characteristics," tion process and to be sustained by acoustic
"combustion timelags," "flame propagation resonance of a critical portion of the system.
velocity," or other terms, which ultimately are Some researchers have shown that in a system
all traceable to the excessive accumulation of having a pump, the pump may be the prime source
unburned fuel, with subsequent detonation or of these oscillations.
cyclic higher-than-rated combustion. The re- This type of instability has not appeared to
sulting excessive chamber-pressure spikes effect be a problem in the development of large engines.
a reduction, or even reversal, of the propellant While it can occur occasionally in large engines,
flows. This will cause rapid collapse of the it appears to be much more prevalent in small-
chamber pressure, allowing propellants to rush scale low-thrust systems. Systems oscillations
in again, thus repeating the cycle. It is readily of the buzzing type are undesirable because of
evident that the physical dimensions of the corn- their adverse effects on engine reliability and
rl
!;I I:!'p t
....... i iI:,
i- "q %== \'\

:°,r,'? !!ii,
:!211 "
" ,\ \ \, \ \,
\ \" \/ .................
............
ii!ii:! Ioo e._
i_ Tiiii!
tll I!
CMA_a£R _ESSU_E
Figure 4-64.-Field olstability and sale operaling

region of a typical engine system.
Figure 4-63.-High-speed pressure measurements
o[ intermediate lrequency instability.
bility in liquid propellant rocket engine systems
performance. When exposed to prolonged buzz- can be attacked in several ways, including sys-
ing, critical parts of the engine may fail be- tems design, choice of propellant combination,
cause of material fatigue, and thus cause sec- and operating conditions, and through the use of
ondary major failures. In addition, measurements special control devices. Two basic approaches
have shown that a performance (Is) loss of up to toward eliminating combustion instability are
approximately 7 percent can be incurred, which employed. The first is to eliminate triggering
is largely attributed to widely fluctuating mixture processes and/or to limit the driving, energy per
ratios. cycle to a value below the damping per cycle
Figure 4-63 is the record of high-speed-pressure inherent to the system. This approach is typi-
measurements of a typical test afflicted by buzz- fied by design investigations in which injector
ing. The oscillation is attenuated in the fuel configurations are varied to give different atomi-
pump outlet pressure, and nonexistent in the zation and propellant distribution characteristics
pump inlet pressure. This indicates that the with varying resistance to initiation of instabil-
buzzing is limited to only a part of the feed ity, or in which propellant additives are used
system. which modify the physical properties controlling
spray formation, chemical kinetics, etc. Experi-
Field of Stability mental studies, both with full-scale engines and

with research model thrust chambers, have shown
For a given engine system, combustion stabil- that injector modification yields relative stabil-
ity limits can be defined experimentally in terms ity ratings, differing by a factor from 5 to 6 from
of certain operational parameters such as cham- the most stable to the least stable configuration.
ber pressure, injection AP, and mixture ratio. The second approach is to introduce addi-
Figure 4-64 presents the stability field for a tional damping in the system through the use of
typical LO2/RP-I, 150000-pound nominal thrust mechanical or other devices so that any opera-
engine system. The stability field, together tional disturbance, or oscillation triggered by
with heat transfer limits, can serve as a guide to the disturbance, are sufficiently and rapidly
the design of a safely operating engine system. damped out. Experience over the past several
years has shown that the destructive transverse
Design Approaches Toward Control of Combustion acoustic modes of instability can be most effec-
Instability tively combated through the use of this second

approach. Mechanical devices, such as baffles
The problem of controlling combustion insta- or a divergent wall gap, have been found to
introduce sufficient damping into the system so teristics will not trigger the interact._on with the
that it will recover from an instability triggered combustion process.
by an explosive charge as large as can be used 2. Combustion chamber design.-Analytical
without damaging the thrust chamber in some studies and experimental results have indicated
other manner. This ability of a system to re- that the geometrical configuration of the combus-
cover from a triggered instability has been des- tion chamber will determine the type of frequency
ignated as "dynamic stability." A prerequisite of the acoustic modes of instability. Chambers
for any propulsion system to operate reliably is having large length-to-diameter ratios appear to
that it should exhibit dynamic stability with be quite prone to large-ampiitude longitudinal
respect to all modes of instability. As a mini- instability. On the other hand, chambers having
mum requirement it should be "dynamically small length-to-diameter ratios appear to be
stable" at least with respect to the destructive sensitive to the transverse modes. Also, small-
transverse (radial and tangential) acoustic diameter chambers are much more stable than
modes. large-diameter chambers. The requirement is to
The successful application of the above design a chamber geometry which will have least
methods has been based primarily on criteria tendency to trigger instability, in conjunction
established empirically in research model thrust with other considerations.
chambers, together with testing in actual engine 3. Injector design.-The injector design ap-
systems. However, the understanding of the pears to be a most critical factor in triggering
fundamental physical principles of the damping instability. In turn, it offers great potential for
processes is still limited. controlling instability-triggering processes
through variation of parameters. Of the most
common types of injectors, the self-impinging
Prevention of Triggering Processes injector (fig. 4-45a) has been chosen by many
The most desirable design method of con- investigators as the best compromise between
trolling instability is the prevention of those performance and stability. There is some indi-
physical or chemical processes which trigger cation that longitudinal instability may be en-
and/or sustain the resonant modes of the com- hanced if the propellant travel time from the
bustion chamber or engine system. While a great injector face to the point of impingement is close
to the half-period (or an odd multiple of the half-
number of studies, in which different design
period) of the longitudinal mode oscillations.
parameters were varied systematically, have
been made by various investigators, the results There are strong indications that increasing the
have failed to yield truly generalized design injection .AP to too great a value may cause the
criteria. This can be traced to the fact that thrust chamber to operate unstably in the trans-
verse acoustic modes. The effect of injection
basic processes which trigger and sustain the
,AP on the longitudinal acoustic modes and on
various types of instability have not been iso-
lated. Thus, while a parameter which controls hydrodynamic instabilities appears to be just the
one type of instability may have been estab- opposite, with stability improving as AP in-
lished on an engineering basis, this same design creases. The propellant distribution across the
injector face has a significant relation to trig-
criterion may be enhancing another type of insta-
gering transverse modes of instability.
bility. The following is a general discussion of
"the prevention of triggering instability in various 4. Propellant combination and mixture ratio.-
Control tests with various propellants have
component and subsystem designs:
shown that there are certain oxidizers or fuels
1. Propellant feed system design.-Past ex-
perience has indicated that certain combustion which can be triggered into instability more
instabilities, such as buzzing, are sustained easily than others, in general or at certain mix-
through an interaction between feed system and ture ratio conditions. In LO2/LH2 systems, a
lower hydrogen injection temperature tends to
combustion dynamics. It is believed that hy-
draulic resonances are a major factor in sustain- trigger instability.
ing this type of instability. The requirement is 5. Engine system operating characteristics.-
to design a feed system whose hydraulic charac- In some cases, combustion instability can be
avoided by modification of engine-system oper- the most effective method of suppressing trans-
ating characteristics. For instance, if the cham- verse acoustic modes of combustion instability.
ber pressure buildup period during engine start This has been demonstrated in both full-scale
is excessive, "chugging" may have time to de- thrust chamber and small-scale models, operating
velop. This low-frequency instability, in turn, with various propellant combinations. These
can trigger the destructive acoustic modes of baffles are usually designed to be secured to the
instability by contaminating one proPellant sys- injector face as shown in figure 4-40. Adequate
tem with the other propellant by driving it back cooling means should be provided to keep the
through the injector ports. The triggering chug- baffles from burnout. The depth or height of the
ging can be eliminated through a fast chamber baffles is a function of the distance of the
pressure buildup. combustion-flame front from the injector face.
Experimental evaluations should be conducted to
support the design and development effort.
Application of Damping Devices
3. Chamber divergent wall gap.-It was found
In further support of design for stability, experimentally that leaving blank an annular
special damping devices can be provided for portion of the propellant injection area adjacent
"dynamic stability" in the combustion chamber to the combustion chamber wall, as shown in
and engine system. The following are several figure 4-66, improved the capability of the com-
devices which have been applied with good bustion chamber to recover from triggered insta-
results: bilities. It was further determined that by filling
!. Feed system hydraulic capacitor.-Self- this "wall gap" with a contoured filler block,
sustained combustion and feed system instabil- "dynamic stability" could be drasticallyimproved
ties of the buzzing type can be eliminated by in most cases. The exact shape of the contour
introducing hydraulic capacitors in the feed sys- is critical, and experimental evaluations are
tem. A hydraulic capacitor is any device which required to determine the most effective design
will increase the effective compressibility at a configuration.
given point in the liquid system. Figure 4-65 4. Acoustic chamber liners.-The feasibility
shows the schematic of an experimental hydraulic of using resonant and nonresonant acoustic sup-
capacitor. It is an isolation-type capacitor with pressing liners on the combination chamber wails
large capacitance and zero resistance between it has been demonstrated in turbojet engine appli-
and the system. Its function is to isolate the cations. In this case the principle of suppres-
transmission of pressure disturbances through sion is similar to the use of patches of acoustic
• the system above a given frequency. Another tile to reduce the sound level in a room, whereby
type is called absorption capacitor, which has the energy absorbed from the mode will reduce
small capacitance and high resistance between its amplitude. Figure 4-67 shows a typical
it and the system. Its function is to absorb the arrangement. The combined area of the sup-
oscillatory energy of the system by damping or pressor openings must be of the order of from 3
attenuating the resonant frequencies of the
system.
2. Combustion-chamber baffles.-The use of
_/_INJECTOR FACE
combustion-chamber baffles has been found to be
•:/" "' - -'-'_'- --_ " . CO_PAESS_a GAS
i.:-:.u-_--_- _--_--.....
l ::::
°°° --- .. • [ - _k__CONTOURE D
:1. _ _ NYDRAUL_ SYIT_Jt
WALL GAP
FILLER BLOCK
Figure 4-65.-Schematic of experimental isolation Figure 4-66.-Combustion chamber divergent wall

type hydraulic capacitor. gap.
-- _.._..
B_JRST COMTROL
j OtAPHRAGi r EXPLOSIVE e/" VALVE
INJECTOR FACE
i, : suPPLY
CHARGE
I_l DIRECTED EXPLOSIVE (b} NON DIRECTED Ic) DIRECTED NOI¢ EXPLOSIVE
PULSE EXPLOSIVE PULSE C_5 FLOW
Figure 4-68.-Combustion chamber perturbation

methods.
Combustion Chamber Disturbances

Figure 4-67.-Combustion chamber acoustic liner.
The introductionof disturbancein the cham-
to 4 percent of the total chamber wall surface ber proper offersa simplermethod of inducing
area. Furthermore, the suppressor thickness instability.Transverse acousticmodes have
should be maximum in the area of maximum pres- been initiatedmost successfullyby the following
sure variation; i.e., near the injector face. methods:
Rating Stability 1. Directed explosive pulses.-The directed

explosive pulse method of inducing instability,
It is desireable to establish the combustion
as shown in figure 4-68a, uses a high explosive
stability level of a particular engine system
charge mounted in an external fixture which is
without an excessive number of tests. This can
attached to the combustion chamber in such a
be accomplished by perturbing a normally stable
way that the gas pulse resulting from the detona-
system by suitable means until instability is
tion enters the chamber with any desired orien-
initiated. The relative stability of various sys-
tation.
tems is then judged as a function of the magni-
tude of perturbation needed to reduce the insta-
2. Nondirected explosive pulses.-The closest
bility. The larger the perturbation, the more simulation of localized random detonations which
stable the system. The perturbation can be
can occur in a chamber during normal operation
introduced in either the propellant feed system
due to accumulation of unburned propellants is
or in the combustion chamber. Several effective
achieved by the nondirectional explosive pulse
methods exist for inducing different types of
method as shown in figure 4-68b. An explosive
instability.
is placed into a thin-walled Micarta shell which
is designed to be mounted at any desired posi-
Feed System Perturbations tion in the chamber.
Perturbation in the propellant feed system to
induce disturbances in the chamber of the low- 3. Directed nonexplosive gas Hows.-In this
frequency hydrodynamic type, can be introduced method, as shown in figure 4-68c, a flow of gas
by- from a regulated high-pressure source is con-
(l) Explosive charges in the fuel feed system trolled by a fast-acting valve. This valve is
(2) Single-stroke positive displacement placed as near to the chamber as possible. This
pistons method additionally permits a better definition of
(3) Oscillating pistons the parameters associated with the disturbance.
Chapter V
Design Of Pressurized-Gas Propellant-Feed Systems
For the transferof the rocketpropellantsfrom of the art of systems components used.
the tanks to the thrustchamber at the required (3) Molecular weight of the pressurant gases:
flow ratesand pressures,a suitablefeed system Lower molecular weight reduces required
is required.The selectionof :he feed system pressurant weight per unit pressure and
will depend on the mission of the vehicle,its per unit tank volume, and thus results in
size and weight,thrustlevel and duration,space lower vehicle system weight at burnout
availableforthe propulsionsystem, on reliabil- and thus improved mass ratio.
ityconsiderations,and on other factors. There (4) Pressurization system specific weight;
is no simple ruleforthe choice between a pres- i.e., required pressurization system
surizedgas feed system (fig.1-12)or a turbo- gross weight (including system compo-
pump feed system (fig.1-13). The advancement nents and pressurant), per unit of weight
in the state of the art of lightweight pressurized of useful pressurant.
gas feed systems, in conjunction with the avail-
ability of high-strength tank construction mate-
5.1 DETERMINATION OF PRESSURANT
rials, has enlarged their field of applications to
REQUIREMENTS
engine thrust levels of approximately 100 000
pounds, and total impulse values of over 5 million The physical and chemical processes which
pound-seconds. For large vehicle applications, take place during the expulsion of a liquidpro-
turbopump feed systems are predominant. How- pellantfrom a tank by a gas or gas mixture are
ever, some type of pressurized gas feed system numerous and difficult to analyze. Applicable
is always required even in a turbopump feed experimentaldata fora selected system areoften
system to supply propellants to the pump at the limited. Thus, the basis for the analyticalap-
correct inlet conditions. proach is frequentlynarrow and uncertain. As a
Classified according to their power sources, result,the initialdesign calculationsof the
four major types of pressurized feed systems can quantityof pressuringgas requiredmust be con-
be distinguished: sidered approximate until verified experimen-
(1) Stored gas systems tally. The refinement of the analytical approach
(2) Propellant evaporation systems to minimize discrepancies between theoretical
(3) Systems evaporating nonpropellants predictions and actual test results is an art
(4) Systems using products of chemical re- requiring experience and thorough understanding
actions of the physical processes.
Selection depends largely upon engine system Basic considerations and necessary proce-
design, type of propellants,mission require- dures for the calculation of pressurant require-
ments, and availableexperience. ments are described below.
Among the considerationsforselectionof the
type of pressurizedgas propellantfeed system
Required Engine System Data and Assumptions
are-
(l) Compatibility of pressurant gases with Before startingcalculationsof pressurant
propellants and tank materials, consid- requirements,the followingsignificantengine
ering chemical interactions, temperature, oper2tingparameters must be known or assumed:
solubility, etc. (1) Design operatingtemperaturerange of the
(2) Expected pressurizationsystem reliabil- propellantsand the feed system includ-
ityand complexity,consideringthe state ing pressurant
151
(2) Type of propellants, their weights and walls in flight may cause heating of the pressur-
corresponding volumes at the extremes izing gas. It may also heat the propellant and
of the operating temperature range thus increase vaporization effects and raise
(3) Total tank volumes: Nominal value and NPSH requirements in turbopump-fed systems.
tolerances 3. Vapor condensation.-Certain components
(4) Initial tank ullage volume, percent of total of the pressurizing gas, such as water vapor,
tank volume at the temperature limits. may condense. Even if the bulk of the gas re-
(The term "ullage" denotes that portion mains above the dew point for the condensible
of a propellant tank not occupied by the component, local condensation may occur at the
liquid propellants.) tank walls, or at the propellant surface.
(5) Trapped propellant volumes, percent of 4. Solubility ol the pressurizing gas.-If the
total propellant load, at engine burnout pressurizing gas contains components which are
(6) Operating tank pressure: Nominal value soluble in the propellant, diffusion of these com-
and tolerances ponents into the propellant can occur. Solubility
(7) Operating duration of the engine systems: is generally affected by temperature and pressure
Nominal value and variations. conditions.
To avoid later marginal conditions, calcula- 5. Ullage gas compression.-If, before start,
tions should assume that pressurization must be the tank ullage space is filled with low-pressure
supplied for maximum systems operating duration gas, onset of pressurization will cause adiabatic
even though some missions may require shorter compression. This can raise the ullage space
durations. Realistic assumptions for the temper- temperature considerably during the initial few
ature of tanks and propellants at burnout must be seconds of operation.
made. If a mission calls for several system 6. Chemical reaction.-If any components of
restarts and coasting periods, the environmental the pressurizing gas are chemically reactive
conditions during the coasting periods must also with the propellant, the reaction products may
be given or assumed. become a component of the gas.
7. Pressurizing gas turbulence.-The heat
Factors Influencing Pressurant Requirements exchange between pressurizing gas and propel-
lant would be extremely large if the gas were
Several important factors which will influence permitted to agitate the liquid propellant surface.
considerably the final state of a pressurizing gas This effect can be prevented through the use of
or gases, and thus their required quantity, are a diffuser which spreads the gas in a gentle flow
discussed below. toward the top and sides of the tank.
1. Propellant vaporization.-Propellants evap-
orate to various degrees from the gas-liquid in-
terface within the tank. The amount depends Design Calculations of Pressurant Requirements
upon the volatility of the propellant, the temper-
ature of the entering gas, the turbulence of the If the system operating duration is relatively
gas, the sloshing of the liquid, the tank geometry short, or if the pressurant temperature is close
including internal structural members, and the to or lower than the propellant temperature, heat-
rate of propellant expulsion. To whatever degree and mass-transfer effects can be neglected. The
vaporization takes place, it lowers the temper- required pressurant weight can then be calcu-
ature of the gas and adds propellant vapor as a lated by the perfect gas law:
component of the pressurizing gases. Also, as
the liquid propellant level recedes, a film of PTVT
liquid may be left on the tank wall surface, Wg- RgTg (5-1)
further contributing to propellant evaporation.
2. Tank wall temperature.-If the pressurizing where
gas is hotter than the tank walls, cooling of the Wg : required pressurant weight in the tank, lb
gas and heating of the wall may result. On the PT = propellant tank pressure, lb/ft 2
other hand, aerodynamic heating of the tank V T = total volume of the empty propellant tank, ft 3
DESIGN OF PRESSURIZED-GAS PROPELLANT-FEED SYSTEMS 15,.3
Rg =gas constant of the pressurant, ft-lb/Ib- The partial volume occupied by the vaporized
deg R propellant is given by
Tg = mean temperature of entering pressurant, °R
However, in cases where longer systems dura-
WvZRpTu (5-4)
tion and higher pressurant temperatures are in- Vv = PT
volved, the pressurant requirement can best be
determined by the following procedure, keeping where
in mind the limitations set forth at the beginning Vv:total volume occupied by the vaporized
of section 5.1.
propellant, ft _
Considering first a single-start operation (not Z =compressibility factor evaluated at the
requiring coast periods and restarts), and neglect-
total pressure (PT) and the temperature
ing heat transfer from the tank walls, the total (Tu) of the gaseous mixture at burnout
heat transferred from the pressurant gas to the
Rp=gas constant of the propellant vapor, ft-lb/
vaporized propellant can be approximated by lb-deg R
equation (5-2). The remaining tank volume at burnout, ne-
glecting residual propellants, can be assumed to
Q = HAt (Tu - Te) (5-2) be occupied by the pressurant gas
where Vg = VT - Vv (5-5)
Q -total heat transferred, Btu
H =experimentally determined heat transfer where Vg = volume of pressurant gas at burnout,
ft 3
coefficient at the gas-liquid interface,
Btu/sec-ft2-deg R The weight of pressurant is calculated by the
A =area of the gas-liquid interface (in case of perfect gas law
a cylindrical tank, the cross-section area

of the tank), ft 2 iq P TV g
t =operating duration, sec g- (5-6)
To-temperature of the gases at burnout, °R
Te = temperature of the propellant, oR In order to maintain the heat balance, the value
Both Tu and Te are treated as constant values for Q should satisfy the following equation:
at the interface between liquid and gas.
This heat, Q, is assumed to have heated and Q : WgCpg (Tg- Tu) (5-7)
vaporized the propellant according to the equa-

tion where Cpg=specific heat at constant pressure of
pressurant gas, Btu/lb-deg R
From equation (5-7) the required value of Tg
Q = Wv [Cpl (Tv- Te) + by + Cpv (Tu - Tv)](5-3)
for the assumed Tu can thus be calculated. If
however, Tg is a predetermined fixed value, then
where the values of Wg, Wv, and Tu must satisfy the
Wv :total weight of vaporized propellant, lb following as well as other correlated equations:
Cpl : specific heat of the liquid propellant,
Btu/lb-deg R WgCpg(Tg- Tu)
hv =heat of vaporization of the propellant, = Wv [Cpl (Tv - Te) + hv + Cpv (Tu - Tv)] (5-8)
Btu/lb
Cpv = specific heat of the propellant vapor, Thus far, heat transfer from the tank walls
Btu/lb-deg R has been neglected. However, if there is a con-
Tv =vaporization temperature of the liquid siderable temperature differential between pres-
propellant, °R surizing gases, propellant, and tank walls, the
The value of Wv can now be obtained from total heat transferred between them during the
equations (5-2) and (5-3) with an assumed value mission must be taken into consideration for the
for Tu. determination of vaporized propellant at burnout.
Equation (5-3) can be rewritten as on experimental data. However, no set of equa-

tions can be applied directly.
In some cases, the uncertainties in pressuriz-
Q*-Qw,
ing system design can be reduced by providing
=Wv[Cpl(Tv- Te)+hv+Cpv(Tu- Tv)] (5-9)
adjustability of the pressurant temperature at the
propellant tank inlet. In this approach, the tem-
where Qw, = total heat transferred between tank perature of the pressurizing gas at system burn-
walls and liquid and gaseous pro- out is assumed or targeted from the beginning.
pellant during the mission, Btu. Based on this and other given or assumed data,
The positive (+) or negative (-) the values of required pressurant quantity and
sign indicates whether QwI is con- inlet temperature can be calculated by equations
tributed by, or lost to, the tank (5-1) through (5-11). Certain correction factors
wails. such as pressurant solubility, etc., can be ap-
Furthermore, equation (5-7) becomes plied later. If the required pressurant quantity
in experimental engine system evaluation devi-
Q=WgCpg(Tg-Tu)+_Qw2 (5-10) ates from the calculations, because of temper-
ature discrepancies of the pressurizing gas at
where Qw 2 = total heat transferred between pres- system burnout, an adjustment of the pressurant
surizing gases and tank walls dur- temperature at the propellant tank inlet can often
ing a mission, Btu. Again the be made to correct for this difference, such as
positive (+) or negative (-) sign in- through an adjustment of the pressurant supply
dicates whether heat is contributed from a heat exchanger, or from a gas generator.
by, or lost to, the tank wails. Because of the narrow safety margins employed
Combining equations (5-9) and (5-10), the heat in rocket vehicle design, the effect of varied gas
balance considering heat transfer from the tank temperatures on structural members must be
walls can be written as carefully weighed.
The following is a sample calculation to
[VCgCpg (Tg- Tu)] +-Qw 2-- Wv [Cpl (Tv - Te) demonstrate this design approach:
+hv+Cpv(Tu- Tv)]-(+-Qw_) (5-11)
If the vehicle mission includes several pow-
ered flight and coasting periods, the calculation From table 3-5, the following data are obtained
of the heat transfer across the gas-liquid inter- for the oxidizer tank of the A-4 stage propulsion
face should take the total mission time into system:
consideration. Equation (5-2) can be rewritten Oxidizer, N204
as Pressurant, gaseous He
Tank volume, neglecting the volume of resid-
Q= HAtm (Tin - Te) (5-12) ual propellant, (VT) = 119 ft 3
Average tank cross-section area, (A)=20 ft 2
where Tank pressure, (PT)= 165 psia, or 23760 psfa
tm =total mission time including powered flight Propellant temperature, (Te) = 520 ° R
and coasting period, sec Calculate the following:
Tm =mean temperature of the pressurizing (a) The total pressurant weight (Wg) and
gases during the entire mission, °R. required temperature (Tg) at the tank inlet, for a
This is a function of many factors such single operation time (t) of 500 seconds, with an
as length of coasting periods, heat trans- experimentally determined heat transfer coeffi-
fer between gases and tank wall, etc. cient (H) at the gas-liquid interface of 0.002
Other effects such as vapor condensation, Btu/sec-ftLdeg R. It is assumed the ullage gas
solubility of the pressurizing gas in the propel- temperature Tu at burnout is 700°R and that
lant, and chemical reactions of the pressurizing there is no heat transferred at the tank wall
gas with the propellant can be included, based surfaces.
DESIGN OF PRESSURIZED-GASPROPELLANT-FEED SYSTEMS 155
(b) The required pressurant weight (g/g) and Substitute this into equation (5-5), to obtain
temperature (Tg) at the tank inlet, for a mission the volume occupied by pressurant gas
consisting of several powered flight and coast-
ing periods, with a total mission duration (tin) of Vg=VT-Vv = 119-7.13 = 111.87 ft 3
18000 seconds. The mean temperature of the
pressurizing gases during the mission (Tin) is From equation (5-6), the required pressurant
526 ° R. The total heat transferred between pro- weight results:
pellant and tank walls (Qw_) is -2000 Btu. The
total heat transferred between pressurizing gas W PTVg
and tank walls (Qw2) is -600 Btu. The temper-
ature (Tu) of the ullage gases at final burnout is
660 ° R. 23760× 111.87
Solution
=9.79 lb
(a) From standard propellant references, the
following data are obtained for N204 at a pres- Substitute results into equation (5-7),
sure of 165 psia:
Vaporization temperature, Tv = 642 ° R Q =9.79 Cpg(Tg - Tu)
Heat of vaporization, hv= 178 Btu/lb
Mean value of specific heat in liquid state, 3600=9.79× 1.25 (Tg- Tu)
Cpl = 0.42 Btu/lb deg F

The required pressurant temperature at tank
Mean value of specific heat in vapor state,
inlet is:
Cpv=0.18 Btu/lb-deg F
Compressibility factor, Z = 0.95
3600
Molecular weight = 92
Tg = 9.79 x 1.25 _ 700: 995 ° R
The specific heat of helium, Cpg, is 1.25
Btu/lb-deg F, and its molecular weight is 4.
(b) From equation (5-12), the total heat trans-
From equation (5-2), total heat transferred at
ferred at the gas-liquid interface:
the gas-liquid interface:
Q = HAt (Tu - Te) Q= gAtm (Tin - Te)

= 0.002 x 20 x 500 (700- 520) = 3600 Btu =0.002 × 20 × 18 000 (526 - 520)
Substitute this into equation (5-3): =4320 Btu
3600 = Wv [Cpi (Tv - Te) + hv + Cpv (Tu - Tv)] Substitute into equation (5-9):
= g/v [0.42 (642 - 520) + 178 + 0.18 (700 - 642)] 4320 - 2000 = Wv [0.42 (642 - 520)
=WvX239.6 + 178 + 0.18 (660 - 642)]
g/v = 10.0 lb
Total weight of vaporized propellant, Wv = 15.11b
Substitute into equation (5-4), to obtain the
volume occupied by the vaporized propellant:
g/vZRpTu
10.0 x 0.95x _1544_ x 660
Vv- PT \921
Vv - 23 760
1544
15.1 × 0.95 ×--9--_TM 700 = 4.45 ft 3
23 760
=7.13 ft 3 Substitute into equation (5-5):

15_ DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
Vg= 119 - 4.45= 114.5 ft 3
23 760 x 110.76
W'= (_) x 660
= 10.65 lb
4320 = 10.65 × 1.25 (Tg- 660) - 600
4320 + 600
Tg- 10.65 x 1.25 + 660 = 1030 ° R
FF VALVE
5.2 STORED GAS SYSTEMS

REGULATOR
Stored gas pressurization systems are widely
used in numerous combinations. The gas is TO MAIN
PROPELLANT TANKS
usually stored in a vessel at pressures ranging
from 3000 to 5000 psi and supplied to the propel-
lant feed system at a specified pressure level Figure S-t.-Helium pressurization system without
heating.
controlled by a regulator. These systems have
achieved a high level of reliability. In earlier
systems, compressed nitrogen gas was frequently to the main propellant tanks. This has the ad-
used or even air (German V-2), mainly for logis- vantage of great simplicity. However, the weight
tics and supply reasons. As it became more of the system is relatively high because of the
readily available, helium gas found increased lower temperature and thus lower specific volume
usage because of its substantially lower molec- of the gas.
ular weight and thus reduced total pressurant 2. HeJium system using thrust chamber heat
weight, and its superiority as an inert agent, exchangers.-This system, which is used in the
with very low boiling point. For hydrogen-fueled design for the A-3 and A-4 stage propulsion sys-
engine systems, reliable compressed hydrogen tems, consists of a high-pressure helium storage
gas systems have been successfully developed. vessel, a start and shutoff valve, manifolded
In general, the most important design require- thrust chamber heat exchangers, and a pressure
ments for a stored gas system are: low molec- regulator. A typical schematic is shown in fig-
ular weight of the gas, high gas density under ure 5-2. The heat exchangers are part of the
storage conditions, minimum residual gas weight, thrust chamber divergent nozzle section and
and high allowable stress-to-density ratio of the absorb heat from the thrust chamber combustion
storage vessel material. Since helium systems process. From overall performance and weight
are now the most widely used ones, the following considerations it is considered preferable to
discussions will be based on them. locate the heat exchanger in the high-pressure
portion of the system. The volume increase of
the gas due to heating reduces the mass required
Discussion of Commonly Used Configurations
for tank pressurization. However, a considerable
1. Helium system without heating.-This quantity of cold, high-density helium still re-
pressurization system is shown schematically in mains in the storage vessel at the end of the
figure 5-1. It consists of a high-pressure storage system operation.
vessel, a start and shutoff valve, and a pressure 3. Helium cascade system.-The cascade
regulator. Regulated helium is ducted directly system shown schematically in figure 5-3 is an
DESIGN OF PRESSURIZED-GAS PROPELLANT-FEED SYSTEMS 157
all-helium pressurization system designed to

minimize the weight penalty resulting from cool-
ing of the helium during expansiou. System com-
ponents include helium storage vessels of equal
high-pressure level but of different sizes, two of
which are divided internally by a flexible dia-
phragm, three thrust chamber heat exchangers, a
start and shutoff valve, and a pressure regulator.
During operation of the system, helium flows
from the firstand smallest vessel, through a
heat exchanger, and displaces completely the
_ START AND SHUTOFF VALVE helium in the intermediate vessel. The helium
in the intermediate vessel in turn flows through
IIli
a second heat exchanger and completely dis-
places the helium in the last and largest vessel.
THRUST CHAMBER
The latter flows through a third heat exchanger
HEAT EXCHANGERS
and pressurizes the main propellant tanks. At
the end of the operation, only the small storage
vessel contains low-temperature, high-density
[ REGULATOR helium, while the two large storage vessels con-

tain relatively warm low-density helium. This
contrasts with the preceding systems in which
TO MAIN orie large storage vessel remains partially filled
PROPELLANT TANKS with low-temperature, high-density helium. The
Figure 5-2.-Helium pressurization system using disadvantages of cascade systems are high
weight and complexity.
thrust chamber heat exchangers.
4. Helium system with heating inside the
storage vessel.-This system is shown schemat-
ically in figure 5-4 and consists of a high-pressure
(_HELIUM helium storage vessel containing a heat ex-
changer or other heat-generating device mounted
internally, a start and shutoff valve, and a pres-
_THRUST CHAMBER sure regulator. This system provides higher
temperature helium to the main propellant tanks,
simultaneously assuring relatively warm residual
gas in the storage vessel. Disadvantages are
the need for larger and more complex high-pressure
storage vessels and the possibility of control
problems during operation.
For the pressurization of liquid hydrogen pro-
pellant tanks, stored hydrogen gas can be used
in place of helium for the systems described
above.
Calculations for Stored Gas Requirements
t ..... ._' _ TO MAIN

The design calculations discussed in section
/ PROPELLANT
/ REGULATOR TANKS 5.1 apply only to the net or effective quantity of
START AND SHUTOFF
VALVE gas required to pressurize the propellant tanks.

However, the gross weight of the stored gas
Figure 5-3.-Helium cascade system. required for a given system depends also on the
38O
370 "I- MIN. HELIUM TANK --

T PRESS. 3 70 PSIA
LINE Ap
360
--_-__MIN. HEAT EXCH, INLET
350 PRESS. 3 54 PSlA --
340
330
HEAT EXCHANGER Ap (IOOPSI)
320
VALVE
(3.
E_ REGULATOR I 260
hl
n-
25O
(/)
TO MAIN W 240
PROPELLANT TANKS n_
EL
23O LINEIP
Figure 5-4.-Helium pressurization system using
220
heaters in storage vessel, REGULATOR,'%9(75 PSI)
210
system design, on the expansion process during

2O0
operation, and on the environmental temperature
range within which the system must function. 190
Based on conditions at systems burnout, the 180

total or gross stored gas requirement can be
170
expressed by the following correlation: LINE-AP PROPELLANT TANK
PRESS. 165 PSIA
160 T
Stored gas gross_ Net system pressurant PRESSURIZATION SYSTEM
requirement requirement
Residual gas in
Residual gas in Figure 5-5.-Estimated pressure drops for A-4
+various lines, heat (5-13)
storage vessel stage oxidizer tank pressurization system.
exchangers, etc.
A parameter to define these additions is the sure drops. Figure 5-5 shows estimated pressure
pressurant use factor, defined as the ratio of drops for the stored helium pressurization system
gross stored gas requirement or initial gas weight with thrust chamber heating which was selected
in the storage vessel to the net weight of pres- for the A-4 stage engine oxidizer tank. In addi-
surant utilized: tion, a safety margin is usually provided. In
this case, the mission is assumed to be com-
Pressurant_ _ Gross stored gas requirement (5-14) pleted when the storage vessel pressure decays
use factor System pressurant net to 400 psia.
requirement If a heating source is provided inside the
vessel, as shown in figure 5-4, the expansion
The lowest pressure level in a storage vessel process of the gas within would be polytropic.
required to safely operate the pressurization sys- For a system without heating inside the storage
tem is determined by the individual system pres- vessel, the expansion process of the gas can be
DESIGN OF PRESSURIZED-GAS PROPELLANT-FEED SYSTEMS 159
assumed to be isentropic; i.e., no heat is trans- Volume of the helium lines downstream of
ferred between gas and vessel walls. From regulator, 0.4 ft 3
equation (1-13) the following correlation can be Volume of heat exchangers, 1.0 ft 3
derived to calculate the final gas temperature in Volume of lines between storage vessel, heat
the storage vessel: exchangers and regulator, negligible
Pressurant reserve, 2 percent
n-__A Assuming an isentropic expansion process,
calculate the following:
T__£
T1 = Q___) " (5-15) (a) Oxidizer tank pressurant gross weight,
where storage-vessel volume, and use factor, for case
T_ =initial helium temperature in the vessel, °R (a) of sample calculation (5-1).
T 2 = final helium temperature in the vessel, °R (b_) Oxidizer tank pressurant gross weight,
Pt =initial helium pressure in the vessel, psia storage-vessel volume, and use factor, for case
P2 =final helium pressure in the vessel, psia (b) of sample calculation (5-1).
n =exponent for the polytropic expansion (C) Oxidizer tank pressurant gross weight,
process storage-vessel volume, and use factor for single
The exponent n is estimated during analytical start operation but without heat exchangers.
treatment and verified experimentally. For isen- (d) Same as c(_c_),but assuming a polytropic
tropic expansion of helium, n = 1.67. expansion process with n = 1.2 in the storage
For most analyses it has been found adequate vessel.
to assume an adiabatic flow process through

regulator and lines. It is a characteristic of this Solution
process that the total (or "stagnation") temper-
ature remains constant. Since the gas essen- The following calculations establish the re-
tially comes to rest in the propellant tank and no quirements for the oxidizer tank, which in the
further compression takes place following initial A-4 stage has the higher minimum storage pres-
propellant tank pressurization, it is assumed sure requirement. To this required pressurant

that propellant tank temperature is equal to heat mass and volumes for fuel tank pressurization,
exchanger outlet temperature. purges and valve actuations will have to be
If an operating temperature range is specified added.
for a system, the lower temperature limit must be (a_)A net pressurant requirement of 9.79
used to calculate the weight of pressurant re- pounds is obtained from sample calculation (5-i),
quired, while the upper limit determines the vol- case (a). Temperature and pressure of the resid-
ume of the gas storage vessel for a given storage ual pressurant in the lines downstream of the
pressure. Thus, for a stated propellant tank net regulators following shutdown are assumed to be
pressurant gas requirement, a wider environmen- the same as those of the propellant tank ullage
tal temperature range results in a heavier pres- gases at system burnout (700 ° R and 23 760
surization system, regardless of the expansion psfa). Residual gas weight in these lines then
process followed by the gas. is
The following data are given for the stored

helium pressurization system of the A-4 stage
oxidizer tank:
Temperature range in the storage vessel at
system start, 500o-560 ° R It is further assumed that the temperature of the
Storage vessel pressure at system start, 4500 residual pressurant in the heat exchangers is at
psia run tank inlet conditions, or 995 ° R, and that the
Pressure in the storage vessel at system pressure is that of the residual helium in the
burnout, 400 psia storage vessel, or 400 psia. Thus a value of
the lines are the same as those in the propellant

ix400x144 = 015
. Ib
tank at system burnout (660_ R and 23 760 psfa),
--_I x 995 residual gas weight in these lines is
is obtained for these residuals. 0.4 × 23 760

=-0.037 lb
Substituting the lower limit of the system
operating temperature, and the initial and final
helium pressures into equation (5-15), the tem-
perature of the residual helium in the storage The temperature of the residual pressurant in the
vessel at system burnout is obtained: heat exchangers is assumed to be at the run tank
inlet level, or 1030 ° R, and that the pressure is
1.67--I that of the residual helium in the storage vessel,
or 400 psia. The gas weight in the heat ex-
T 2 = 500 x ( 400 '_ 1.67
\4500) = 500 x (0.089) °' = 191°R changers then amounts to:
Using equation (5-13), the pressurant volume VL i x 400 x 144

=0.145 lb
required for oxidizer tank pressurization can now
be calculated, based on the lower storage tem-
perature limit of 500 ° R:
From equation (5-13), pressurant volume VL
4500 x 144 x V L and mass are obtained, using the lower temper-
=9.79+ 400 × 144 x VL --0.035+0.15
ature limit:
_--_) x 191
4500 x144 xVL 400x 144xVL
9.975 {1544\x500 =10.65+ [1544\ +0.037+0.145

VL-0.374 × 6.9
3.78 ft 3
x191
Gross pressurant weight: i, 10.832 _
L =_ - 4.2 ft a
x 1.02 = 12.95 lb
Gross pressurant weight:
4500 x 144 x 4.2

× 1 B2 = 14.45 lb
including 2 percent reserve.
× 500
Using the upper start temperature hmit
(560 ° R), the required volume of the storage ves-
sel for oxidizer pressurization is calculated: including 2 percent reserve.
Required volume of the storage vessel:
x 560x 12.95
x
Vu= 4500x 144 -4.3 ft 3 (_)x560 14.45
Yn = = 4.82 ft 3
4500 x 144
From equation (5-14), LOX tank pressurant

use factor: Pressurant use factor:
12.95 14.45
-- = 1.325 - 1.36
9.79 10.65
(_b) From sample calculation (5-1), case (__, (c_.) Without heat exchangers, the bulk temper-
the net pressurant requirement is given as 10.65 ature of helium in the propellant tank at system
pounds. Assuming temperature and pressure in burnout can be expected to be the average of the
initial and final helium temperatures in the stor- (_ The expansion process of helium in the
age vessel, or storage vessel is assumed to be polytropic, with
n= 1.2. From equation (5-14), the temperature of
500 + 191 residual helium in the vessel at system burnout:
Tg = 2 = 346 ° R
1.2--1
Since this temperature is lower than the propel-

T2=500x(400_ 12 ×2.73
lant temperature, no heat loss from pressurant to _4500] = 500 4.--.'.[_7
= 336° R
propellant is assumed. The net pressurant
weight required may then be calculated by equa- The helium bulk temperature in the propellant
tion (5-1): tank at system burnout:
23 760 × 119
Wg:/l_aa\ =21.2 Ib Tg=500+336=418 ° R
2
x 346
The required net pressurant weight:

The weight of residual helium in the lines at
system burnout: 23760× 119
Big_ [_A_,\ - 17.5 lb
0.4 × 23 760
= 0.071 lb &) X 418
x 346
The residual helium weight in the lines at
system burnout:
From equation (5-13), pressurant volume VL
and mass are obtained, based on the lower ambi- 0.4 x 23 760
- 0.059 lb
ent temperature limit:
4500 x 144 x VL 400 × 144 x VL +

Pressurant volume and mass, based on the
"(__/×500 =21,2+ (_)×191 0.071
lower ambient temperature limit:
21.27 4500x 144 x V L 400x 144x VL

VL =_ = 8.25 ft3 + 0.059
2,58
(% ooo
Gross pressurant weight =
, 17.56
4500 x 144 x 8.25 L=_=6.01 ft 3
x 1.02 = 28.2 lb
× 500
Gross pressurant weight:

including 2 percent reserve.
The required volume of the storage vessel is 4500x 144x 6.01
x 1.02 =20.6 lb
-_)x560 × 28.2
Vu = 4500 x 144 - 9.4 fts including 2 percent reserve.
The required volume of the storage vessel:
Pressurant use factor:
28.2 _1_4--) × 560 × 20.6

_=1.33
21.2 Vu - 4500 x 144 - 6.88 ft 3
Pressurant use factor: Experimental data indicate that gases of low

molecular weight such as helium and hydrogen
20.6 will not leak through homogeneous metals such
--= 1.178
17.5 as good-quality, hot-rolled stock or forgings.
However, they can leak through porous metals
such as may exist in castings and in welded
Design of Stored Gas System Components joints. Good welding workmanship and effective
leakage inspection are most important in the
Since system components can be expected to
fabrication of storage vessels.
require relatively extensive development effort
to achieve satisfactory performance and reliabil-
Pressure Regulators
ity, they require careful design considerations.
For most pressurization systems a regulator
This is especially true for large, high-pressure,
of high accuracy is a necessity. Regulation be-
lightweight, pressurant storage vessels, pres-
comes particularly difficult if gases with temper-
sure regulators, and thrust chamber nozzle skirt-
atures higher than 1200 ° R have to be handled,
type heat exchangers.
or if high flow rates or large pressure differen-
tials are involved. Design detail of pressure
Storage Vessels
regulators will be discussed in chapter VII. For
Because of the combined requirement of high
some applications, a control system combining a
pressure and light weight, pressurant storage
pressure switch, a solenoid valve, and an orifice
vessels are generally spherical in shape and
may be preferred.
made of high-strength-to-weight materials.
PH-15-7-MO stainless steel, 6A1-4V-titanium,
Thrust Chamber Heat Exchangers
and light aluminum liners wound with plastic
For helium systems using thrust-chamber heat
filaments for strength, such as fiber glass, have
been successfully employed as construction
materials for the vessels. Design details for NOZZLE WALL
HEAT
pressurant storage vessels will be further dis- EXCHANGER
cussed in chapter VIII.
For weight estimates in preliminary designs,
it is assumed that the vessel is made of two
hemispherical shells, and that the thickness of
the weld lands can be accounted for by assuming
a 3-inch-wide band of one-half the wall thick-
ness placed over the weld seam. Hence, the
total weight of the vessel can be estimated as:
Wv = rrd2 pm (pd/4s) + 3,_dpm (0.5 pd/ 4s) (5-16)

_.THRUST CHAMBER
where
Wv = weight of the vessel, lb
d =inside diameter of the vessel, in
p =maximum storage pressure, psia
s =allowable working stress of the material,
psi
pm =density of the material, lb/in 3
It is of prime importance in the design of
stored gas systems that the storage vessel be
capable of containing the gas at high pressure
for long periods of time without loss by leakage. f
Frequent checking of the storage pressure or
recharging is undesirable in most applications. Figure 5-6.-Thrust chamber heat exchanger.
DESIGNOF PRESSURIZED-GASPROPELLANT-FEED SYSTEMS 163
exchangers as shown in figure 5-2, the heat ex- compatibility with the main chamber nozzle wall.
changer should be designed as an integral part Firm attachment of the heat exchanger to the
of the thrust chamber expansion nozzle. As a nozzle wall is mandatory, since heat exchanger
rule, the heat exchanger is made of coiled tubing efficiency depends on intimate contact. Selec-
formed to fit the nozzle contour (fig. 5-6). tion of tube thickness must be based on pressure
The combustion-gas side heat-transfer coeffi- and thermal stress conditions.
cient can be determined by methods described in
chapter IV. The heat conducted through the wall
of the heat exchanger is assumed to be totally
absorbed by the pressurant helium, raising its The following data are given for the design of
temperature. The determination of the helium- the pressurant heat exchangers for the A-4 stage
side heat-transfer coefficient and the design of engine thrust chamber nozzle extension, when
the heat exchanger tubing are similar to the re- located at the station of area ratio = 10, and used
generatively cooled tubular wall thrust chamber in parallel:
analyses. The number of tube turns for the heat Helium flow rate through each heat exchanger,
exchanger is a function of the helium temperature I_h = 0.024 lb/sec (considers requirements
rise required and of the heat exchanger location for both tanks and for other uses)
at the nozzle. The various operating parameters Helium specific heat ratio, y= 1.67
of a thrust chamber heat exchanger can be corre- Helium specific heat at constant pressure,
lated by the following equation: Cp = 1.25
Mean temperature of helium at heat exchanger
inlet, Ti = 346 ° R
WhCp (To - Ti)
Mean temperature of helium at heat exchanger
outlet, To : 1030° R (from sample calcula-
A ! (5-17)
tion (5-1), case (_)
Combustion-gas side adiabatic wall tempera-
ture, Taw = 4900 ° R
where Combustion-gas side heat transfer coefficient,
=helium flow rate, lb/sec fig--5.7 x 10 -s (Btu/in:-sec-deg F)
Cp = specific heat at constant pressure of Calculate:
helium, Btu/lb-deg F (.E) Heat exchanger tube dimensions, assum-
Ti =mean helium temperature at heat ex- ing it to be made of 13V-11Cr-3AC titanium alloy
changer inlet, °R with the following physical characteristics, at a
To =mean helium temperature at heat ex- recommended maximum working temperature range
changer outlet, °R of 1400°-1550 ° R:
A =effective area of the heat exchanger, in: Minimum yield strength, Sy : 40 000 psi
hg =combustion-gas side heat transfer coeffi- Modulus of elasticity, E-- 12 × 106 psi
cient, Btu/in2-sec-deg F Thermal conductivity, k= 2.04 x 10 -4 Btu/in 2-
ha =helium-side heat transfer coefficient, sec-deg F/in
Btu/in2-sec-deg F Coefficient of thermal expansion, a : 5.0 × 10 -_
t = heat exchanger tube wall thickness, in in/in-deg F
k = thermal conductivity of the tube material, Poisson's ratio, v: 0.33
Btu/in:-sec-deg F/in (b) Number of turns of the heat exchanger
Taw : combustion-gas side adiabatic wall tem- tubing.
perature, °R
Heat exchanger design must consider that the Solution
temperature of the helium leaving the heat ex-
changer at any given time depends on the storage (_a) The wall temperature at given sections of
vessel exit temperature. the heat exchanger will vary directly with the
The choice of heat exchanger tube material bulk temperature of the helium in these sections.
must be made with consideration of its brazing Maximum wall temperature occurs at the heat
exchanger outlets. A mean combustion-gas side Inside tubing diameter, d = 0.44 inck.
wall temperature (Twg) at the outlets of 1400 ° R Check for combined pressure and thermal
is allowed. From equation (4-I0), the heat flux stresses at the heat exchanger outlet (neglecting
at that section is: bending stress), using equation (4-27), and as-
suming bending stress due to discontinuities to
q = hg (Taw - Twg) be negligible
= 5.7 x 10 .5 × (4900- 1400) = 0.2 Btu/in 2 sec
A tube wall thickness (t) of 0.05 inch isused. St- (Pco -t Pg) r + 2 (1-v)k
Eaqt
This will be checked further below for compati-
bility with pressure and thermal stresses. From
4500 x (--_)
0.44
equation (4-19), the mean helium-side wall tem- 12 x 106 x 5.0 x 10-6 x 0.2 x 0.05
- 4
perature is 0.05 2 (1 - 0.33) x 2.04 x 10 -4
tq 0.05x0.2
- 1350 ° R = 19800+2200= 22000 psia
Twc= Twg --_ = 1400 2.04 x 10 -4
The combined pressure and thermal stresses

Using equation (4-20), the helium-side heat at the heat exchanger inlet are now checked. It
transfer coefficient is calculated as can be assumed that the difference between the
combustion-gas side wall temperature and the
q 0.2 helium bulk temperature remains approximately
hh = hc =(Twc -Tco) - (1350- 1030) constant throughout the heat exchanger. Then:
= 6.25 x 10 -4 Btu/in2-sec-deg F Twg at the inlet = 1400- (1030- 346) = 716 ° R
(mean temperature)
Heat flux q at the inlet = 5.7 × 10 -s (4900- 716)
Equation (4-15) permits determination of the
= 0.239 Btu/in2-sec
Prandtl number of the flowing helium:
Combined stress St at the inlet
47 4xl.67
Pr= 9---_= 9 x 1.67_ 5:0.665 = 19800+ 2200 x (_)
The viscosity of the helium according to = 19800+2630=22430 psi

equation (4-16) is:
Therefore, it is safe to use the selected tube
_=(46.6 x 10-1°)x_°ST °6 size of 0.440-inch inside diameter and 0.050-inch
wall thickness, with sufficient margin for the
#=46.6 x 10-1° x 4 °_ x (1030) °_
fact that the heat exchanger helium inlet temper-
=60.2 x 10 -8 lb/in-sec
ature will be higher than the maximum at the
beginning of the process.
From equation (4-25)
(b_) From equation (5-17), the required effec-
tive area for each heat exchanger element
0.029 Cp# °2 (GO.S._(Tco_ °ss
hh: p2 3 A = 0.024 × 1.25 (1030 - 346)
6.25 x 10-4 = 0.029 x 1.25 x (60.2 x 10-s) °2

1 ÷ 0.05 + 1 )
(0.665) 2`,3 5.Tx10-s 2.04×10 -4 6.25_10 -d =95in 2
1030-346
(4*___hh_
°s 4900
2
x(103o °.ss
× d0_ k1--_-6-) The nozzle diameter at area ratio = I0 of the
A-4 chamber is:
d 1.8= 0.235 D = Dt x V1--0-=7.18 x _/i-G= 22.7 in
DESIGN OF PRESSURIZED-GAS PROPELLANT-FEED SYSTEM_ 165
Assuming 40 percent of the internal tube sur- and mass transfer processes within the main pro-
face as the effective heat exchanger area, the pollant tank, which in turn are influenced by
number of heat exchanger tube turns required is: pressurant and environmental temperatures. For
a given rate of evaporation of the propellant in
A 95 the tank, average flow rate through the tank vent,
N= - 2.42 turns
reD x rrd x 0.4 _2 x 22.7x 0.44 x0. 4 and tank ullage gas or vapor condition, the fol-
lowing steady-state correlation can be estab-
lished:
5.3 PROPELLANT EVAPORATION SYSTEMS
This concept is practical only for thermally tkPv (We-Wv)
stable, low normal boiling point propellants, WP-pRT N (5-18)

such as cryogenics and near cryogenics, for
which it is widely used. It is particularly suit- where
able for cryogenics of low molecular weight, _Pp = required flow rate per engine bled off for
such as hydrogen. propellant tank pressurization, Ib/sec
[Pe =rate of propellant evaporation in the tank,
Ib/sec
Application in Pump-Fed Systems
OA_EOU$ NI'I _OGEN

Propellant evaporation systems for pump-fed *[0 _AIN FUEL
engines usually employ propellants tapped off

TANK -_ MOT G4S FRO_
downstream of the pump and vaporized in a heat _,*$EOU$ OXY6EN TO M,L{N _ I_ TURBINE E_AUST
exchanger, after which they are used to pressur-

ize the main propellant tank from which they
were withdrawn. Figure 5-7 shows a typicalheat
exchanger design used in an LOjRP-1 pump-fed
uoo.° o,,..
engine system. The turbine exhaust gases are
used as the heat source. Sometimes the vapor-
ized propellant is bled directly from the manifold ..---..
downstream of the chamber cooling passage if it
is the coolant in a regeneratively cooled thrust
chamber. As shown schematically in figures 3-3
and 5-8, the pressurant for the A-2 stage main
oxidizer tank is provided by oxygen tapped off
downstream of the oxidizer pump and vaporized EXHAUST OUCT
in a heat exchanger located at the turbine ex-

haust duct of the oxidizer turbopump. The main Figure 5-7.-Typical heat exchanger design.
fuel tank of this stage is pressurized by bleeding
hot hydrogen from the thrust chamber fuel mani-
fold downstream of the thrust chamber cooling
tubes. The pressure of both tanks can be regu-
lated by pressure switch/solenoid valve com-
binations, as shown in figure 5-8, or by regula-
tors. The latter are often preferred, particularly
if a narrow band of regulation is essential. In
both cases, flow limiting orifices may be used in r_ibr" :
series, for increased reliability, with valve or
regulator design-biased to fail open. [ cox._., ft
The required propellant flow rate bled for
vaporization and main tank pressurization is
determined by the main propellant flow rate at Figure 5-8.-A-2 stage propellant tank pressuri-
the pump inlet (or tank outlet), and by the heat zation system schematic.
Wv = average flow rate through the tank vent, (b) The required steady-state flow rate, per
lb/sec engine, bled off for fuel tank pressurization.
= main propellant flow rate (per engine) at

pump inlet, lb/sec Solution
p =density of the liquid propellant, lb/ft 3
PT :propellant tank pressure, lb/ft 2 (a_) The density of liquid oxygen is 71.38
R =gas constant of the propellant vapor, ft-lb/ lb/ft 3, the gas constant of the gaseous oxygen
lb-deg R is 1544/32= 48.2 ft-lb qb-deg R.
T =temperature of the tank ullage gas, °R Substitute this and data given above into
N =number of engines in the system equation (5-18) to obtain the required steady-
state flow rate of evaporated oxidizer pressurant:

Wp= 290.Sx45x144 (1.6-1.64)_2.50 lb/sec
The following data were established for the 71
.---_--8_ 4{_.2 x-2"20 4
A-2 stage engine and vehicle systems during
steady-state operation conditions: b(.b) The density of the liquid hydrogen is 4.42
Main oxidizer flow rate at pump inlet, per lb/ft 3, and the gas constant of the gaseous hy-
engine, 290.5 lb/sec (table 3-3) drogen is 1544/4 = 386 ft-lb/lb-deg R.
Main oxidizer tank pressure, 45 psia Substitute this and data given above into
Rate of oxidizer evaporation in the tank, 1.6 equation (5-18) to obtain the required steady-
lb/sec state flow rate of the evaporated fuel pressurant:
Average flow rate through the oxidizer tank
vent,' 1.64 lb/sec
i. 59.8×38×144 (42-6"6)--2.21b/sec
Temperature of the oxidizer tank ullage gas, P - _-_ x _g _-i-20 4
220 ° R
Main fuel flow rate per engine, at pump inlet, It is noted that some engine specialists pre-
59.8 lb/sec (table 3-3) fer to use slightly lower propellant densities; for
Main fuel tank pressure, 38 psia instance, 71.0 for LOX and 4.4 for LH 2. These
Rate of fuel evaporation in the tank, 4.2 values consider the fact that storage and vehicle
lb/sec containers, even when vented to atmosphere,
Average fuel flow rate through the fuel tank have small positive pressures because of vent
vent, z 6.6 lb/sec valve resistance, resulting in slightly increased
Temperature of the fuel tank ullage gas, propellant temperatures. However, since later
120 z R engine calibration-run evaluations will require
Number of engines in the vehicle system, 4 corrections for a number of run-to-run engine
Calculate: input deviations, consistent usage of design
(a_) The required steady-state flow rate, per parameters is probably more important than their
engine, bled off for oxidizer tank pressurization. absolute value. It is further pointed out that
tank pressure regulation through venting, partic-
IFor fire hazard reasons, vehicle design very ularly if used throughout the systems duration,
likely will require provisions to prevent venting dur- is not an efficient method, since onboard gas
ing first stage boost. Also, during regulated A-2
storage must allow for the maximum vent rate
stage operation, venting should normally not occur, as
it would be a performance loss. However, as a pres- anticipated.
surization system performance margin it is well to lay
out the system on the basis of some vent losses.
Also, in cryogenic systems it may be desirable to in- Applications in Pressurized Gas Propellant Feed
crease the tank pressures toward the end of stage
Systems
operation to improve pump NPSH conditions, when the
upper tank layer, in which somewhat warmer liquid The application of propellant evaporation sys-
may have accumulated, is about to reach the pump.
tems to pressurized propellant feed systems is
This can be simply done by opening an orificed by-
pass around the regulator, using the vent valve liftoff somewhat limited. Evaporation systems can
pressure as the regulating factor. result in lower pressurant storage vessel weight,
, i ¸ _
DESIGNOF PRESSURIZED-GAS PROPELLANT-FEED SYSTEMS 167
as compared to stored gas systems, because of solubility in the propellants (nitrogen in liquid
higher storage densities and lower storage pres- oxygen) and a storage temperature significantly
sures. However, this can be offset by the higher lower than that of the propellants (liquid helium).
required pressurant weight per unit volume, par- The system design for this concept is similar to
ticularly for propellants with higher molecular that of the propellant evaporation systems.
weight. For hydrogen, the principal propellant
with low molecular weight, another limitation 5.5 SYSTEMS USING PRODUCTS OF
exists because of the low critical pressure. To CHEMICAL REACTIONS
obtain reasonable volume increases due to vapor-
ization, the tank pressure must be kept suffi- Pressurizationsystems using hot gaseous
ciently below the critical pressure. products generated from solid or liquidpropel-
It must be further considered that the propellants have been successfullydeveloped forthe
lant evaporation concept, when applied to storableliquid propellantengine systems.
pressure-fed systems, requires a pressurization Another technique used with noncryogenics is
system within a pressurization system, since a the main propellant tank injection pressurization
separate stored gas is required to expel the pres- system. Here a hypergolic fluid is injected into
surant from the storage vessel as a liquid, after the tank and pressurization is provided by the
which it is vaporized in a heat exchanger. This products of the reaction occurring within the
system comprises a relatively complicated array tank.
of components, line assemblies, heat exchangers, These methods are not applicable to cryogenic
and support structures. It is further complicated propellants because the products of reaction,
because of the auxiliary pressurization system such as water, will solidify, and because the
required to initiate the main propulsion system heat of combustion will raise undesirably the
operation. As shown schematically in figure g-6, bulk temperature of the cryogenic propellant.
the main fuel tank of the A-3 stage propulsion EspeciaiIy in the ease of liquid hydrogen, bulk
system is pressurized by evaporated hydrogen heating cannot be tolerated because of its lim-
supplied from a separate liquid hydrogen storage ited liquid range (normal boiling point to critical
vessel which in turn is pressurized by the stored point).
helium gas. The hydrogen pressurant is vapor- Two important considerations for the applica-
ized in the heat exchangers, located at the thrust tion of combustion products for pressurization
chamber nozzle extensions. •are: propellant compatibility and gas tempera-
For the various reasons stated, the propellant ture level. Except for very short operating dura-
vaporization principle will be used only for the tions (few seconds), fuel-rich hot gases are used
fuel tank of the A-3 stage, for which it still for fuel tanks, to prevent reactions. Similarly,
appears attractive because of the relatively low oxidizer-rich hot gases are applied to oxidizer
pressure levels selected, and the low molecular tanks. The temperature of the product gas pres-
weight of hydrogen. The A-3 oxidizer tank will surant should be maintained at, or should be
be pressurized by stored helium gas. This de- cooled to, a level below 1200 ° F.
cision was further influenced by the difficulty in
handling gaseous fluorine and by its toxicity.
Solid Propellant Gas Generator Pressurization
Systems
5.4 SYSTEMS EVAPORATING
Severaleffectivesolid-propellant gas-generator
NONPROPELLANTS
systems have been developed fortank pressuri-
This type of pressurizationsystem has not zationof prepackaged storable liquid propulsion
been employed frequently.Two types of inert systems. Pressurant gas temperatures up to
cryogenics could be considered applicable: liq- 3000 ° F and tank pressures up to 2000 psia have
uid nitrogenand liquidhelium. Both have def- been proven successful for short-duration appli-
initedisadvantages which would generally pre- cations. This pressurization method is primarily
clude their use in nonpropellant evaporation employed for its inherent simplicity, low produc-
systems. The main disadvantages are their tion cost, long-term storability, relatively light
_m
system weight, and compactness. The system is phragms to maintain system reliability even after
usually composed of two electrically fired initi- long storage periods.
ators, or squibs, a charge of igniter pellets, Upon ignition, the propellant grains are
safety and arming devices, a pressure-relief-type ignited by the igniter pellets. Combustion starts
regulator, and the propellant grains. A device to and produces pressurization gases for the dura-
cool the hot gases may be required in specific tion for which the propellant grains were de-
applications. signed. The burning rate of the grain, and in
turn the gas pressure level of a solid propellant
Figure 5-9 shows a typical design. The system, is affected by grain bulk temperature.
solid-propellant gas-generator assembly is en- Within a given service temperature range, the
closed in an insulated steel housing. This hous- grain is designed to produce required gas pres-
ing is installed within an aluminum casing, sures and flow rates at the lower temperature
which in turn is an integral part of the propellant limit. When operating at a higher temperature,
tank constructed of aluminum alloy. The gas the pressure level will be maintained by a regu-
generator unit is completely integrated into a lator which vents all excess gases overboard.
compact package ready for testing, storage, and The required grain charge is sized for full pro-
installation into the propulsion system, with pulsion system operating duration at the upper
minimum effort and maximum safety. All gas temperature limit, at which maximum grain burn-
outlets are hermetically sealed with burst dia- ing rate occurs.
OXIDIZER TANK
O-RING SEALS
PROPELLANT ISOLATION DIAPHRAGM

_URST DIAPHRAGM
PELLETS
INHIBITOR
INHIBITOR
PERFORATED
SUPPORT PLATE
DIAPHRAGM
PREssuRE-
RELIEF
TYPE
REGULATOR
SQUIB
MOLDED
CHOPPED
PHENOLIC
-- PROPELLANT _AS GENERATOR HOUSING

GRAINS _8-7-5 STEEL
/_- O- RING SEAL /--FF-34 FORMICA INSULATION
GAS DIFFUSER
/_OXIDIZER EXPULSION DIAPHRAGM
PROPELLANT TANK WALL
Figure 5-9.-Typical solid propellant gas generator pressurization system.

The selection and design of solid propellant

gas generator pressurization systems must con-
sider potential problem areas, notably the
following:
(1) Chemical and temperature compatibility of
the gases with the propellants _ w_.,--_ WITH
SOLID IGNITER
PROPELLANT
(2) Pressure regulation difficulties
(3) System clogging by solid particles carried
in the gas stream
(4) Lack of restart capability
(5) Requirement to vent gas rapidly in the
event of premature engine cutoff.
A brief discussion of several commonly used

solid propellant systems follow: FILTER_
1. Solid propellant gas generator system with- VENT_--J __PRESS-RELIEF TYPE
out cooling.-This system is shown schemati- '-r' REGULATOR OR
_ORIFICE
cally in figure 5-10 and consists of a solid pro- V
pellant charge, including igniter, filter, and hot TO MAIN PROPELLANT TANK
gas regulator or orifice. This system is suitable
mainly for relatively short durations. Upon igni- Figure 5-10.-Solid propellant gas generator with-
tion, hot gases are fed through the filter, regu- out cooling.
lated, and ducted to the main propellant tanks.
The regulator dumps excess gas overboard, for
which a vent line must be provided. As an alter-
nate, an orifice may be installed in place of the
regulator. In the latter case, the engine thrust
level is a direct function of the solid gas gener- _--- SOLID PROPELLANT
ator pressure which in turn is a function of
environmental temperature.
2. Solid propellant gas generator system with
solid coolant.-This system (fig. 5-11) consists
of a solid propellant charge and igniter, a sub-
limating solid coolant, a filter and a regulator.
In operation, the hot gases generated are cooled
as they pass through a bed of solid material, ID COOLANT
subjecting it to decomposition or sublimation.
Thus the cooling process simultaneously results
in additional gases available for pressurization.
The gases finally pass through a filter and are
regulated and dueted to the main propellant tanks.
In a typical design an ammonium nitrate base
propellant is used with a theoretical flame tem- FILTE
perature of 2320 ° F. Oxalic acid pressed into PRESSURE-RELIEF
VENT
pellets is the solid coolant. This chemical TYPE REGULATOR
decomposes endothermically at temperatures
above 250 ° F, producing a mixture of CO, CO:,
and H20. The desired temperature level of the
TO MAIN
mixed product gases is achieved through selec-
PROPELLANT TANK
tion of the propellant to coolant ratio. Final gas
temperatures as low as 400 ° F have been ob- Figure 5-11.-Solid propellant gas generator sys-
tained. tem with solid coolant.
3. Solid propellant gas generator system with tems is in providing relatively pure nitrogen gas
azide cooling pack.-The principal components at a reasonable temperature level (as low as
of' this system are shown in figure 5-12. Here, 600 ° F).
hot gases are cooled when passing through a bed 4. Helium system with solid propellant gas
of azide material wluch decomposes and yields generator heating.-This system (fig. 5-13) con-
essentially pure nitrogen. However, the gas sists of a high-pressure helium storage vessel
leaving the azide pack is often contaminated with a solid-propellant gas generator mounted
with metal particles resulting from the decompo- internally, a filter, and a pressure regulator.
sition of the azide. These have to be removed The solid-propellant charge provides both heat
in a cyclone separator. The gases are filtered for helium expansion and additional pressurizing
to remove any remaining particles, then regu- gas. A disadvantage of this system is the need
lated, and directed to the main propellant tanks. for a relatively large, high-pressure storage
The advantage of these pressurization sys- vessel.
Liquid Propellant Gas Generator Pressurization

Systems
SOLID Both liquid monopropellant and liquid bi-

PROPELLANT propellant gas generators have been used suc-
CHARGE cessfully for generating pressurant gases in
• -_//
AZI DE
PACK
\\\\\
\\\\\
SEPARATOR
I CYCLONE
_FI LTER
JL
DUMP TYPE
,.-.-..-, REGULATOR
#
#
TO MAIN TO MAIN
PROPELLANT TANKS PROPELLANT TANKS
Figure 5-i2.-Solid propellant gas generator sys- Figure 5-13.-Helium system with solid propellant
tem with azide cooling pack. gas generator heating.
: i :_
engine systems with relatively long operating weight. However, if excess fuel is injected to
durations. The selection and design of a gas lower the temperature to well below 1000 ° F, this
generation system must consider propellant benefit is not obtained, resulting in a gas with
compatibility, operating temperature limits of high molecular weight which readily condenses.
propellants and tank materials, and gas molecular Liquid-propellant gas-generator pressuriza-
weight. tion systems, some of which will now be dis-
Among the monopropellants, hydrazine is con- cussed, require relatively complex components
sidered the most satisfactory with respect to and controls. Their selection and application is
chemical characteristics and molecular weight of determined mainly from vehicle mission require-
the product gases. The decomposition products ments, such as long operating duration, restart,
of hydrazine can be made even lighter by cata- etc.
lytic decomposition of the ammonia component. 1. Single gas generator system with injection
cooling.-Here, a single generator provides pres-
The gases contain no carbon, deposits of which
could lower the heat transfer rate if a heat ex- surant gas for both the fuel and the oxidizer
changer is used for cooling. The theoretical gas tank. The liquids employed consist of either a
decomposition temperature of pure hydrazine is monopropellant and a nonreacting injection cool-
1800 ° F. Additives can be used to reduce the ant, or a bipropellant combination with cooling
temperature level. achieved by injecting an excess of one propel-
To meet the requirement for compatibility lant. As shown in figure 5-14, system compo-
with the propellant, bipropetlant gas generators nents include a small, high-pressure helium gas
possess the flexibility of operation at either storage vessel assembly (including start, relief,
fuel-rich or oxidizer-rich conditions. Thus the and fill valves), a pressure regulator, two liquid
same propellant combination can be used to pro- storage vessels, and a gas generator assembly
duce both a fuel-compatible and an oxidizer- (including controls).
compatible pressurant gas. For instance, nitro- This system is potentially simple and relia-
gen tetroxide in combination with hydrazine, ble. The product gases are cool enough for use
UDMH, and various other amine fuels permits in the propellant tanks. The difficulty is to find
hypergolic starting and stable operation in either a combination of liquids that will produce a
fuel- or oxidizer-rich modes. product gas meeting all requirements, including
Several avenues are open to meet the pres- compatibility with both propellants. It has been
demonstrated, however, that fuel-rich gases can
surant gas temperature requirement. The afore-
mentioned flexibility to operate at either fuel- or
oxidizer-rich conditions has the additional ad-
vantage to combust well off stoichiometric mix-
ture ratios, resulting in lower temperatures. A (_R HELIUM
limitation exists if one of the bipropellants is
also a monopropellant which continues to decom- EGULATOR
pose exothermically. The cooling effect can

also be obtained by injecting into the hot gases
a given amount of noncombustible liquid which
absorbs heat when evaporating. A third method GAS GENERATOR
L ts uSED
of cooling is heat exchange with one of the liq-
uid propellants in a heat exchanger. This can be
applied with essentially any propellant combina-
tion, provided the cooling liquid can safely ab-
sorb the heat transferred from the gas.
The requirement of low molecular weight will
be met by most fuel-rich product gases if they TO MAIN
P_LLA_T TA_KS
are at approximately 1000 ° F, where the complex
organic compounds composing most fuels are Figure 5-14.-Single liquid propellant gas genera-
cracked into simple gases of low molecular tor system with injection cooling.
be used to pressurize storable oxidizers. For

instance, ammonia-rich, ammonia-nitrogen tetrox-
ide gas generator products and hydrazine decom-
HELIUM
position products with either water or ammonia
injection have been successfully applied as the
pressurant for storable oxidizers. AUXILIARY
PROPELLANT
2. Single gas generator-helium system.- TANKS
Figure 5-15 shows the main components compris-

ing this system. The number of propellant tanks
depends on whether a monopropellant or bipro-
pellant gas generator is used. The hot gas gen- FUEL- R ICH
GAS GENERATOR G/LS GENERATOR
erator products are ducted to a heat exchanger OXIDIZER - RICH "_
where heat is transferred to the cold helium.

The heated and expanded helium gas is then TO MAIN TO MAIN
OXIDIZER FUEL
used to pressurize the main oxidizer tanks, while TAN K TAN K
the gas generator gases are used to pressurize

Figure 5-16.-Dual bipropellant gas generator
the main fuel tank. While the system efficiently
system with injection cooling.
uses the available heat for both tanks, it has the
disadvantages of requiring a moderate-size, high-
pressure helium storage vessel and of pressure dizer injection to produce relatively cool,
regulation problems. oxidizer-rich gases which are used directly to
3. Dual bipropellant gas generator system pressurize the main oxidizer tank. A design
with injection cooling.-The major components problem is produced by the need to balance the
for this system are shown in figure 5-16. During output of the two gas generators and to provide
operation, fuel and oxidizer are fed to the gas proper pressure control in both tanks. Gas tem-
generators by pressurizing the auxiliary propel- peratures as low as 600 ° F have been success-
lant tanks with helium. One gas generator oper- fully generated with these systems.
ates with excess fuel injection to produce rela-
tively cool, fuel-rich gases which are used
directly to pressurize the main fuel tank. The Main PropellantTank DirectInjection
PressurizationSystems
second gas generator operates with excess oxi-
The directinjectionsystems employ injection
of small quantitiesof fueland oxidizerintothe
main oxidizerand fueltanks,respectively.
Following hypergolicreaction,the pressurizing
gases are thus produced withinthe main propel-
REGULATOR -]
lanttanks themselves. The fuel system (fig.
5-17) includes a small, high-pressure helium
/ x
PROPELLANT storage vessel assembly, a helium pressure
SECOND PROPELLANT regulator, and two small auxiliary propellant
TANK
(BIPROPELLANT I.... TAINK tanks from which propellants are injected into
GAS GENERATOR)_
the main tanks. The series injection system
L._
ENERATOR (fig. 5-18) consists of only one auxiliary oxidizer
EXCHANGER
HEAT :t tank for main fuel tank pressurization,
versa, instead of two separate auxiliary
or vice
tanks
TO OXIDIZER
1 for each main tank.
advantage
The series system takes
of the situation where one of the main
TO FUEL
TANK
propellant tanks can operate at a slightly lower
TANK
pressure than the other. In the case shown in
Figure 5-15.-Single liquid propellant gas genera- Figure 5-18, a small quantity of the main fuel
tor helium system. supply is fed through a regulator to the main
,a mml
DESIGN OF PRESSURIZED-GAS PROPELLANT. FEED SYSTEMS 173
HELIUM
_ HELIUM
_REGULATOR !REGULATOR
MAIN
FUEL MAIN MAIN
TANK FUEL OXIDIZER
TANK TANK
D ]=.
Figure 5-17.-Main propellant tank dual direct Figure 5-18.-Main propellant tank series direct
injection system. injection system.
oxidizer tank. Although these systems appear to The four principal selection criteria are:
offer the lightest and simplest method of pres- vehicle mission requirements, propellant and
surization by number of components, system material compatibility, system reliability, and
safety, reliability, and ability to produce steady system performance. Mission requirements in-
regulated tank pressures are questionable. clude storability, system instant start and re-
start, and closely controllable pressure levels
(constant or variable). Compatibility includes
5.6 SELECTION OF THE PRESSURIZATION
chemical inertness, freedom from excessively
SYSTEM
condensible and soluble gas products, and proper
At the beginning of the design of a pressur- pressurant temperature level. Reliability is
ization system, a preliminary study of the vari- evaluated on the basis of system complexity,
ous design approaches is conducted. This will number of failure modes, and number of system
provide a technical basis for systems selection, components the reliability level of which must
but can become a relatively complex task in view be attained economically within the limits of
of the great number of tank pressurization tech- development time and funds allotted. Compo-
niques developed. In the course of the study, nents which can be expected to require consider-
all information pertinent to the wide range of able development effort toward satisfactory reli-
systems is gathered. A standard vehicle con- ability levels include gas generator assemblies,
figuration and uniform assumptions for mission thrust chamber heat exchangers, large high-
and performance, structural materials, and other pressure storage vessels, and hot-gas ducting
factors affecting the pressurization system de- and regulators. System performance is deter-
sign are adopted to permit valid weight and size mined entirely on the basis of gross pressur-
comparisons of the various systems studied. ization system weight which is influenced by
J.74 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
TABLE 5-1.-Comparisons o[ Various Tank Pressurization Systems for the A-4 Stage Propulsion System
Total Number of i System System

I
Stored gas quantity complex overall overall

System Gas produced
and generants of system component weight, volume
components designs percent )ercent
Helium with thrust chamber Helium Helium at 995 ° R aver- 11 3 100 100
heating (fig.5-2) age
Helium-no heating (fig.5-I) Helium Helium at 354 ° R aver- 9 2 182 183

age
Helium cascade (fig. 5-3) Helium Helium at 1000° R 17 7 81 75

average
Solid propellant gas generator Helium, ammonium aMixed helium and solid 15 4 79 72
heating helium (fig. 5-12) nitrate base grain propellant products at
I000 _ R average
Solid propellant gas generator Ammonium nitrate base aN)lid propellant prod- 3 53 12
with solid subliming coolant grain, oxalic acid ucts at 1000 c R
(fig. 5-10) pellets
Solid propellant gas generator Viton azide propellant.l Nitrogen at I000 c R 11 4 70 ii

with azide cooling (fig. lithium azide coolant
5-11)
Single liquid propellant gas Helium, hydrazine bHelium at 1000 ° R on 11 4 63 51

generator-helium (fig. 5-14) oxidizer, hydrazine
products at 1200 ° R on
fuel
Dual liquid propellant gas Helium, nitrogen aAmmonia-rich gas 16 6 48 16

generator with injection tetroxide, ammonia generator products at
cooling (fig. 5-15) 1000 ° R on fuel;
N20,-rich gas gener-
ator products at
I000 ° R on oxidizer
aQuestionable compatibility of pressurant with propellants.

bMarginal pressurant temperature level on fuel.
pressurant molecular weight and specific system The results presented in table 5-1, "Total Quan-
weight. The gross weight of the pressurization tity of System Components" include components
system is treated as inert weight, since the of like design; i.e., two identical storage ves-
entire pressurant and system components are sels or check valves are considered a quantity of
part of the vehicle system weight at burnout. two. "Complex Component Designs" refers to
A sample preliminary design study was con- parts requiring relatively long development time
ducted for the tank pressurization system for the and high production cost. Instead of absolute
A-4 stage propulsion system. The vehicle mis- values for overall weight and volume of the
sion was assumed to require single-start, various systems, percentages of the selected
constant-thrust, full-duration operations. Eight system (stored helium with thrust chamber heat-
different pressurization systems were compared. ing) are used for comparison.
Chapter VI
Design of Turbopump Propellant-Feed Systems
In high-thrust, long-duration liquid propellant larly with the introduction of high-porto ........
rocket engine applications, turbopump feed sys- hydrogen-t'ueled engines. As ohown in figure 6-1
tems, as shown in figures 1-13, 2-10, 3-3, and for two propellant combinations, the proportionaI
4-58, generally result in lower systems weight turbopump-assembly weight in an engine system
and higher performance when compared to pres- rises with increasing chamber pressure, as does
surized gas feed systems. Turbopump f_oa sys- the required turbopump power expressed as a
tems require only relatively low putnp-inlet pres- percentage of total engine propellant flow rate.
sures, and thus propellant-tzL_k pressures, while A representative range of pump operating
the major portion of the pressure required at the parameters for various liquid propellant rocket
thrust chamber inlets is supplied by the pumps. engine applications is shown in figure 6-2.
This saves considerable tank weight, particularly Figure 6-3 depicts an envelope of turbine power
in large vehicles. requirements for a number of actual turbine
As the overall trend toward higher chamber designs.
pressure of liquid propellant rocket engines con-
tinues, the role of turbopumps in an engine sys- io_,oo0
tem becomes of even greater importance, particu-
IO
8
L_
Z
*oo0
ZU
ItP-LUOMM. NIN4. SO/S_O
! LOz' _zO_. '*104
_- 2 pooi , _ ,,,,HI , , J,,,,,I d _ ,LJ,,,I _ L ,,,,:L

I0 _ _ IO_OCO '00_00
__.LOX/LH 2 FLOW, GPM
Figure 6-2.-Range of operation for typical pro-

pellant pumps.
7O
_ 40i
,, ..=.30;
m 20 h. a
LOX/RP,
0 ,
0 500 I000 1500 2000 i
CHAMBER PRESSURE, PSI mo i(x_
TURBINE
to,ooo
HORSEPOWER
_o,o(_ i ooo.o(x_
Figure 6-1.-Variation of turbopump/engine weight Figure 6-3.-Envelope of rocket engine turbine

and Ilow ratios with change of chamber pressure. designs.
175
,ummm
OUTLET
6.1 ELEMENTS OF TURBOPUMP
PROPELLANT FEED SYSTEMS
INLET---_ DRIVE
The supply of propellants to the inlet of the
pumps at required minimum pressures is custom-
arily considered the responsibility of the vehicle
propellant system and thus of the vehicle de- CENTRIGUGAL
IMPELLER
signer. The main function of the turbopump feed
SINGLE STAGE CENTRIFUGAL PUMP
system then is to raise the pressure of the pro-
t_ot_,_,_t_ received from the vehicle tanks and
deliver them to the main thrust chamber, through
ducts and valves, at pressures and flow rates CROSS-OVERVOLUTES 7 OUTLET
commensurate with rated engine operation. A
turbopump feed system may consist of the fol-
lowing basic elements:
(1) Propellant pumps
(2) Turbine(s) to drive them
(8) A power source for the turbine(s) (during
engine start as well as main stage)
(4) Speed reduction gear transmissions (if
any) MULTISTAGE CENTRIFUGAL PUMP
(5) Lubrication system for bearings and gears
OUTLET
(if any)
(6) Shaft-speed pickup for instrumentation and
for safety purposes (overspeed cutoff)
(7) Accessory drives (if any)
(S) Propellant inlet and discharge ducts (if
any)
INDUCER_/L..f;:: E_X
G tV: L_::E LLER
(9) Turbopump mounts
MULTISTAGE AXIAL PUMP

Propellant Pumps
Figure 6-4.-Schematics of various pump configu-
The principalrequirementsof a rocket engine rations.
propellantpump are reliability, optimum speed,
lightweight, high deliveryrateat maximum pres- 6-14, 6-43, 6-47, and 6-48. Centrifugal pumps,
sure head, smooth flow fora wide range of oper- like other steady-flow rotating machinery, con-
ating conditions,and high efficiency.The most sist essentially of two basic elements: the rotor
widely used pump types are centrifugal(orradial) and the stator. Their working principle is the
flow, axialflow, and mixed flow pumps. Centrif- acceleration of the fluid flow by imparting kinetic
ugal pumps are generallydesigned with a single energy to it in the rotor and then decelerating or
stage, while axial pumps are primarilyof multi- "diffusing" it in the stator. This results in
stage design. However, multistage centrifugal increased fluid pressure head. The rotor assem-
pumps with crossover-type volutes have also bly usually includes an inducer, an impeller, and
been considered. Various pump configurations a shaft. The stator assembly consists of a
are shown schematically in figure 6-4. casing with stationary diffuser vanes, a volute
1. Centxilugal pumps.-Almost all operational with discharge outlet, shaft bearings, and seals.
rocket propellant pumps (except those for high- An inducer is an axial-flow-type impeller.
flow, high-pressure liquid hydrogen applications) Its main function is the increase of the static
are of this type. They can handle large flows at pressure of the entering fluid sufficiently to
high pressures efficiently as well as economi- permit normal operation of the main impeller.
cally in terms of weight and size. The elements An inducer can reduce the pump inlet pressure
of a centxifugal pump are shown in figures 6-5, net-positive suction head (NPSH) requirements
DESIGN OF TURBOPUMPPROPELLANT-FEED SYSTEMS 177
_- VOLUTE PASSAGE
(TO DISCHARGE)
PUMP CASING
FRONT WEARINGRING
REAR WEARING RING
(OPTIONAL FOR HYDRAULIC
BALANCING OF AXIAL
THRUST IN PLACE OF BALANCE RIBS)
ALANCE RIBS
DRIVE
FLUID
SHAFT__ __
SHAFT BEARINGS
INLET FLANGE
SHAFT SEALS
DIFFUSER VANES
Figure 6-5.-Elements of a centrHuga1-flow pump.
substantially. The impeller of a centrifugal (or 2. Multistage centriiugal pumps.-For higher

radial) pump basically is a rotating wheel with pressure rises, multiple-stage centrifugal pumps
radial vanes. Fluid is admitted axially to the can be designed if a single stage proves limited.
impeller which, when rotating in an enclosure, The basic construction of a multistage pump is
ejects it at the periphery with increased velocity similar to that of a single-stage pump, except
(fig. 6-5). that proper channeling of the fluid between
The primary functions of the pump stator as- stages is added.
sembly are: diffusing (i.e., decelerating) the 3. Multistage axial pumps.-This design is
fluid to convert the velocity head into pressure well suited to liquid hydrogen service which
head, collecting and redirecting the fluid to the entails the problems of extremely low fluid tem-
pump discharge outlet, and providing structural perature and density. Low fluid density results-
support and a pressure enclosure for the pump. in high-volume flow and in high pressure-head
The main function of the wearing rings shown in rise requirements. For applications at either c_
figure 6-5 is to provide axial thrust control and flow rates higher than 60 lb/sec or pressure-head
to minimize internal leakage, or circulating of rises above 1400 psi, a multistage axial flow
the fluid between the high-pressure (discharge) pump is generally superior with respect to con-
and the low-pressure (inlet or suction) zones. struction and performance.
External leakage along the shaft is prevented by Elements of an axial flow pump are shown in
the use of dynamic shaft seals. figures 6-6, 6-51, and 6-53. The rotor assembly
V ROTOR BLADES
PUMP CASING
'\
STATOR BLADES
'\
\ VOLUTE PASSAGE
\, (TO DISCHARGE)
TFLANGE '\
\\
SHAFT
FLUID
FLOW DRIVE
SHAFT SEALS
/ -_--. BEARING
/ INTERNAL SEAL INTERNAL SEAL
INDUCER EARING =
Figure 6-6.-Elements of an axial-flow pump.
consists of an inducer, a cylindrical rotor with Turbines

multiple rows of rotating blades, and a rotor The turbines which provide shaft power to the
shaft. The stator assembly includes a cylindri- propellant pumps derive their energy from the
cal casing with rows of stationary blades spaced expansion of a high-pressure, high-temperature
between inducer and rotating blades, a volute
gas to lower pressures and temperatures. The
casing, bearings, and seals. basic elements of a turbine are shown in figures
An inducer is placed at the pump inlet to 6-7, 6-14, and 6-54. Turbines can be divided
supply the fluid to the main-pump section at the into two major types: impulse turbines and reac-
required pressure and velocity. Both rotor and tion turbines. Impulse turbines can be either
stator blades have a hydrofoil shape. The main single or multiple stage. Reaction turbines are
function of the rotor blades is to accelerate the usually multistage. Various turbines employed
flow relative to the stator and thus to increase in rocket engine applications are described as
the kinetic energy of the fluid, while the stator follows:
blades, acting as diffusers, convert the velocity 1. Single-stage, single-rotor impulse turbine.-
head of the fluid into pressure head. However, This turbine consists of a single-rotor disk or
the velocity vector of the fluid in the axial direc- turbine wheel to which is attached a row of tur-
tion is kept constant throughout the various bine blades or buckets. Gas is fed to the rotat-
stages of the pump. ing blades through stationary nozzles (fig. 6-8).
DESIGN OF TURBOPUMP PROPELLANT-FEED SYSTEMS 179
STATIONARY NOZZLES ,/-- FIRST ROW ROTATING BLADES
INLET GAS MANIFOLD -STATIONARY BLADES OR SECOND ROW NOZZLES
SECOND ROW ROTATING BLADES
DIAPHRAGM
BEARING ROTATING SEAL

HOUSING
TURBINE SHAFT
DRIVE
BEARINGS
'-- TURBINE EXHAUST DUCT
J
---TURBINE WHEELS
/ -- TURBINE HOUSING
SHAFT SEALS- /
Figure 6-7.-Elements of a turbine.
DIRECTION
OF MOTION
Figure 6-8.-Schematic of a single-stage, single- Figure 6-9.-Schematic of a single-stage, two-

rotor impulse turbine. rotor, velocity-compounded impulse turbine.
In the nozzles, the gas pressure is converted

into kinetic energy (velocity head) with attendant
static pressure drop. The gas flow velocity is
maximum upon entering the rotating blades, where
the kinetic energy of the gas is imparted to the
turbine rotor as mechanical energy of rotation.
Ideally, the static pressure of the gas remains
constant when it passes through the rotating
blades (except for the effects of friction).
2. Velocity-compounded impulse turbine.-
Figure 6-9 shows a schematic of this turbine
type. Here, two separate rows of rotating blades
instead of one are used to transfer the kinetic
energy of the gas discharged from the set of sta-
tionary nozzles. A row ot stationary blades is
placed between the wheels to guide the gas into
the second set of moving blades. This principle
is credited to Curtis who originally developed it.
Ideally, the entire pressure drop occurs in the Figure 6-10.-Schematic of a two-stage, pressure-
stationary nozzles. The gas velocity decreases compounded impulse turbine.
during passage through the first row of rotating
blades, remains constant through the stationary
blades, and decreases further as it passes
through the second rotating row. Velocity-
compounded turbines are considered single
stage, since only one pressure step is involved.
3. Pressure-compounded impuIse turbine.-As
shown schematically in figure 6-10, the expan-
sion of the gas is accomplished in steps, through
two or more rows or stages of stationa_' noz-
zles, each set being followed by a row of rotat-
ing blades. A design objective is that the veloc-
ity, and thus the kinetic energy, of the gas flow
is the same at the entrance of each row of rotat- Figure 6-II.-Schematic o[ a reaction turbine.
ing blades. This results in equal energy
transferred to each rotating blade row, while the
impulse and reaction wheels are driven by a
pressure drop in each stationary nozzle row will
change in momentum of the gas. In a pure
vary. Since the pressure is greater in region A
reaction-type turbine, the driving force is de-
than in region B, it is necessary to separate the
stages by a sealing diaphragm to prevent bypass rived entirely from the reaction due to gas ex-
pansion within the rotating blades (similar to the
flows (fig. 6-7). However, because of clear-
gas expansion in a rocket nozzle). In actual
ances required at the rotating seal between dia-
phragm and turbine shaft, some losses do occur reaction turbine designs, however, a portion of
the driving force is derived from impulse due to
at this point due to leakage from stage to stage.
gas impingement on the rotating blades.
4. Reaction turbine.-The main difference
between an impulse turbine and a reaction tur-
Turbine Power Sources
bine (shown schematically in fig. 6-11) is that in
an impulse turbine no static pressure drop occurs In most applications the turbine is driven by
(no expansion) while the gas passes through the gas produced in either bipropeltant or monopro-
rotating blades, whereas in a reaction turbine the pellant gas generators. Other turbine power
pressure does drop (expansion occurs). Both sources have also been used, such as "tapoff"
and "topping" systems. Typical turbine power

sources are shown in figure 6-12 and are de-
scribed briefly as follows:
1. Bipropellant gas generator.-This is the
most widely used system since it has the advan-
tage of using the engine main propellants. A. BIPROPELLANT B. MONOPROPELLANT C. THRUST CHAMBER
BLEED (TAPOFF)
2. Monopropellant gas generator.-This pro-

vides the simplest gas-generating system. How-
ever, it requires a third propellant if neither one
of the main propellants is suitable for monopro- _$T i T. TURaJNE CMA=aER
i c._. _s ¢_XE R_, TOll
pellant application.
D. TOPPING E. DUAL COMBUSTION
3. Thrust chamber bleed.-This is applied in
"tapoff" engine systems. Gases are bled off Figure 6-i2.-Typical turbine power sources.
directly from the main combustion chamber to
drive the turbine.
start tanks pressurized to up to 90 percent of
4. Topping.-In a topping cycle, such as with rated pump outlet pressures supply the propel-
a hydrogen-fueled engine system, the heated
lants to the gas generator during engine start
hydrogen gas emerging from the thrust chamber until main propellant pump discharge pressures
cooling jacket is used as the turbine working
build up. In some applications (first stages), the
fluid before being injected into the main combus- start tanks have been made a part of the ground-
tion chamber. In systems using a monopropellant support equipment.
as one of the main propellants, the monopropel- 3. Main propellant tanks.-Experimental en-
lant can be decomposed and used to drive the
gine systems have been successfully started
turbine prior to injection into the main combus- with the propellants supplied directly from the
tion chamber. Topping gas turbine drives render vehicle main propellant tanks, thus initiating gas
the highest possible theoretical cycle efficiency. generator and main thrust chamber operation until
5. Dual combustion.-The entire fuel flow
pump "bootstrap" occurred. In a hydrogen-fueled
reacts with a portion of the oxidizer in a bipro-
topping cycle engine, hydrogen vapor is used
propellant gas generator and thus provides the under tank head pressure to start the turbine
gas to drive the turbine. The usually fuel-rich directly.
exhaust gas is then ducted into the main com-
4. Stored gas.-Stored gas under high pressure
bustion chamber and reacts with the balance of
has been used to spin the turbine during engine
the oxidizer. Dual combustion cycle efficiency
start. In the case of hydrogen-fueled engines,
equals that of the topping cycle.
hydrogen gas stored in a rechargeable bottle is
For most systems, an auxiliary power source used to drive the turbine during initial as well
is required during engine start to drive the tur- as restarts.
bine until the main power source takes over. 5. Solid propellant gas generator.-Solid pro-
The energy and its rate of delivery required for pellant gas generators or turbine spinners, as
the start transient depend to a large extent on shown in figure 4-49, have also been widely used
the engine system design. Several common tur- to power turbines during engine start.
bine power sources for engine start are as
follows:
Turbopump Drive Arrangements
1. Monopropellant.-In systems using a liquid
monopropellant gas generator, the fluid is sup- The specific type of coupling between turbine
plied by an independently pressurized tank, such and pumps depends not only upon the propellants
as in the German V-2 engine. Thus, no addi- being pumped but also on the design of the over-
tional turbine power source for engine start is all engine system. Various turbopump drive
required. arrangements are shown schematically in figure
2. Bipropellant start tanks.-In a system 6-13. Where a single turbine directly drives both
using a liquid-bipropellant gas generator, fed propellant pumps through a common shaft, the
from the main propellant system, bipropellant turbine can be located either on the shaft end
DIRECT DRIVE GEARED DUAL SHAHS pump shaft. The turbopump gears and bearings
are cooled and lubricated either by a separate
oil supply system, or by a fuel additive sub-
system (fig. 6-17). During mainstage operation,
the turbopump supplies oxidizer and fuel to the
main thrust chamber as well as to the gas gen-
O_A. TURBINES IN SERIES erator at the required pressures and flow rates.
Operating characteristics and materials of con-
struction for this turbopump are listed in table
F FuEa F,u_p 6-1.
*" _'_ g. 1URBINES IN PARALLEL
Both oxidizer and fuel pumps are of single-
C. SINGLE GEARED PUMP entry, centrifugal-flow type. They are mounted
back to back on a common shaft, one on each
Figure 6-I 3.-Principal tllrbopump drives. side of the gearbox. The fuel pump is bolted to
the gearbox, while the oxidizer pump is secured
to it by radially inserted steel pins. These pins
(with back-to-back pump arrangement), or be-
allow the oxidizer pump housing to expand and
tween pumps. In this case both pumps and tur-
bine will operate at the same shaft speed. Gear- contract during extreme temperature changes
without distortion and misalinement. Each pump
dirven turbopump arrangements include: the
pancake type, which uses different reduction has an axial-flow inducer, a radial-flow impeller
with backward curved vanes, stationary diffuser
gears and is applied where there are speed dif-
ferentials between pumps and turbine; the offset vanes, and a volute. The propellants pass from
turbine, with both pumps on one shaft but driven the inducers to the guide vanes in the impeller
inlets through the impeller rotor vanes into sta-
through a gear train: and the single-geared pump
where one pump is mounted with the turbine on tionary diffuser vanes in the pump casing and
into the pump volutes. The diffuser vanes as-
the same shaft, while the other is driven through
sure uniform pressure distribution and reduction
a reduction gear. Dual-shaft turbopump arrange-
of fluid velocity around the impellers. Balance
ments with pump and turbine for each propellant
on separate shafts include: two gas turbines in ribs are provided on the back side of the impel-
lers to neutralize pump shaft axial thrust.
series with the discharge gas from the first tur-
bine driving the second turbine; and two gas The gearbox includes a series of full-depth
turbines in parallel, both receiving gas directly reduction spur gears with integral bearing inner
from the power source. races, gear carrier and main shaft bearings,
accessory drives, pump shaft bearing seals, and
a bearing heater on the oxidizer pump shaft. A
Description of Developed Turbopump Systems drain manifold is provided for horizontal drain-
age. The gears reduce the speed of rotation
Figures 6-14 and 6-15 illustrate a typical between turbine and pump shaft by an overall
liquid bipropellant rocket engine turbopump sys- ratio of 4.88 to 1. Details of typical turbo-
tem. It was developed for a 188000-pound-thrust pump gears and bearings are shown in figure 6-16.
LOX/RP-1 booster engine. Figure 6-14 shows The pump shaft turns c:lockwise as viewed from
the major elements of this turbopump design. the oxidizer pump. The sequence of powertrans-
Figure 6-15 is a cutaway view of an actual as- mission is as follows: turbine to high-speed
sembly with the inlet elbow ducts attached. pinion gear, to intermediate shaft gear, to inter-
This assembly is a dual-pump unit consisting of mediate pinion gear, and to pump shaft gear.
an oxidizer pump, a fuel pump, a reduction gear- Power is also transmitted to a main accessory
box, an accessory drive adapter, and a turbine. drive gear from a drive pinion gear mounted on
The turbine is started by hot gases from a tur- the intermediate gear shaft.
bine spinner (solid propellant gas generator) and The turbine is an impulse-type, two-stage
powered from a liquid propellant gas generator pressure-compounded unit (fig. 6-10). It is
during mainstage. The turbine shaft drives a bolted to the fuel pump housing and consists of
series of reduction gears, which in turn drive the hot gas inlet manifold, stationary nozzles and
FUEL PUMP
OXIDIZER PUMP
.UTE PASSAGE (TO DISCHARGE)
PUMP
GEAR BOX
PELLER BALANCE RIBS
RADLAL- FLOW IMPELLER
WEARING RING
IUCER
_MP INLET
IMP SHAFT AND SHAFT NUT
SHAFT SEAL
PUMP INLET SHAFT BEARING

FLANGE AND ROW ROTATING BLADES
ROW ROTATING BLADES
BLADES OR
SECOND ROW NOZZLES
WHEELS
ACCESSORY
DRIVE PAD SHAFT AND
SHAFT NUT
TURBINE DRIVE SEAL

GEAR TRAINS
SEAL DIAPHRAGM
TURBINE INLET GAS MANIFOLD
SHAFT SEALS
STATIONARY NOZZLES TURBINE
Figure 6-I4.-Major elements of a typical turbopump.
gear. The turbine shaft is supported on the in-

LOX PUMP dry, GEAR _ board side by a ball bearing; on the outboard
=-_V', Box ,_ E_ELpu,,P side, by a roller bearing. Carbon-ring shaft
seals are used to prevent hot-gas leaks. The
turbine inlet manifold distributes the gases to
the first row of stationary nozzles which, in turn,
distribute the gases to the first row of rotating
blades. When leaving these, the gases again
increase their velocity when passing through the
"_= - "" ..... f ' TURBINE
second row of stationary nozzles. They finally
pass through the second row of rotating blades
and leave the turbine through an exhaust duct.
A sealing diaphragm between the first and sec-
ond turbine wheel prevents the hot gas from by-
Figure 6-15.-Cutaway view of a turbopump passing the second row of stationary nozzles.
assembly. In later systems, a fuel additive blender unit
(fig. 6-17) was substituted for the oil lubrication
blades, first- and second-stage turbine wheels, a system to increase reliability and to reduce
turbine shaft, and a splined quill shaft connect- weight by eliminating the separate oil tank, its
ing the turbine shaft to the high-speed pinion pressurizing equipment, plumbing, and controls
Figure G-iG.-Typical turbopump gears and bearings.
f "FULL" INDICATOR
_ / INDICATOR
, ,>. ,.,,., ....,,...,
ADDITIVE CHAMBER
¼.!i_:}!;_ii:.':!:{H _----
i':?,i:!i'%i:._
ADDITIVE -_
.";._ PISTON
ADDITIVE :}ii!
l _ DRAIN iC{i
J ' /L PLUG , !ili Ii /-STRA,NER

t .. _, ,1 __,-=:_#!_____/ ASSEMBLY
t '
OUTLET PORT -J' / " _ L_ INLET PORT
FUEL UNDER
[ i _._ ADDITIVE FILL _ / SPOOL-' L_ METERING TURBOPUMP
_,/ BLENDING CHAMBER -- ORIFICE DISCHARGE
PRESSURE
Figure G-I7.-Fuel additive blender uniL

TABLE 6-1.-Operating Characteristics and Materials o[ Construction [or the Turbopump Shown in
Figure 8-14
Oxidizer Fuel
Pumps-
Fluid ............................................... Liquid oxygen RP-1

inlet density ........................................ 71.38 lb/ft a 50.45 lb/ft _
Inlet pressure (total) ................................. 80.0 psia 57.0 psia
Discharge density .................................... 70.95 lb/ft a 50.55 lb/ft _
Discharge pressure (total) ............................. 915.2 psia 1023.0 psia
Pressure rise in pump ................................ 835.2 psi 966.0 psi
Pump developed head ................................. 1696.2 ft 2751.0 ft
Volume flow ......................................... 3257,4 gpm 2007.6 gpm
Flow rate ........................................... 518.0 lb/sec 225.7 lb/sec
Shaft speed ......................................... 653' rpm
Efficiency .......................................... 75.5 percent 72.1 percent
Shaft power ......................................... 2117 bhp 1565 bhp
NPSH required ....................................... 35.0 ft 35.0 ft
Casing material ...................................... TENS 50.T6 aluminum alloy sand casting
Inducer material ..................................... 7075-T6 aluminum alloy die I 2024-T351 aluminum alloy
forging [ plate
Impeller material ..................................... TENS 50.T6 aluminum alloy ] 9669-48230-3 aluminum alloy
sand casting ' [ sand casting
Shaft material ....................................... 4340 alloy steel
Bearing material ..................................... 9310 alloy steel
Turbine:
Inlet gas pressure (total) .............................. 597,6 psia
Exit gas pressure (static) ............................. 32.86 psia
Pressure ratio: Total inlet/static exhaust .............. 18.21

Inlet gas pressure (static) ............................ 517.8 psia
Inlet gas temperature ................................. 1200 ° F
Exit gas temperature ................................. 938 ° F
Gas flow rate ....................................... 17.34 Ib/sec
Brake horsepower .................................... 3793 hp
Shaft speed ......................................... 31740 rpm
Efficiency .......................................... 66,2 percent
Housing material ..................................... Hastelloy "B"
Nozzle block material ................................ Hastelloy "B"
Wheel material ...................................... Timken alloy 16-25-6 AMS-5727 steel
Shaft material ....................................... 4340 alloy steel
Gearbox:
Reduction speed ratio ................................ 1/4.855
Gearbox material ..................................... TENS 50-T6 aluminum alloy sand casting
Gear and-shaft material ............................... 9310 alloy steel die forging
previously required. The blender works on the to inject a proportionate amount (2.75 percent) of
principle of adding a small amount of a chemical additive into the fuel flowing through the blend-
to the fuel tapped off through metering orifices ing chamber. The mixture then flows to the
at the fuel pump outlet and of using this mixture turbopump where it lubricates and cools the
as the lubricant. The fuel entering the blender gears and bearings.
inlet port is admitted to the top side of the addi- Figure 6-18 and table 6-2 illustrate another
tive chamber and to the blending chamber. The turbopump system. It was developed for a 6000-
fuel pressure causes the piston to displace and pound-thrust aircraft superperformance rocket
NPSH. Pump inducers, impellers, shaft and cast

]URBINE
u_L ' ' .---T:-- housings are fabricated from 300 series stainless
steel. Turbine manifold, housing and nozzle
blocks form an integral Hastelloy-X weldment.
The turbine wheel is made of Timken alloy.
C 6.2 TURBOPUMP
AND
Turbopump
DESIGN
SYSTEM
PARAMETERS
performance
PERFORMANCE
and design param-

eters are closely related. Variation of the turbo-
Figure 6-18.-Cutaway view of an aircraft rocket pump design and operating parameters will con-
turbopump assembly. tribute to the optimization of both turbopump and
engine system performance. The turbopump
system design procedure includes the evaluation
engine (shown schematically in fig. 4-60).
of all possible design approaches and mechanical
Pumped fluids are 90 percent hydrogen peroxide
and JP-4 fuel. The turbopump consists of two configurations which can satisfy engine system
centrifugal pumps mounted back to back on a specifications. From this, the selection of the
best design with respect to overall systems reli-
single shaft which is directly driven by a single-
ability and performance is made.
stage, single-rotor impulse turbine. The shaft is
supported by antifriction bearings located be-
tween the two pumps, and just forward of the Turbopump System Performance
overhung turbine. The bearings are lubricated
The "best performing" turbopump system is
by a pressure-fed lube oil system. Inducers are
used on both pumps to permit operation at a low simply defined as that which affords the heavi-
est payload for a vehicle with a given thrust
level, range or velocity increment: gross stage
TABLE 6-2.-Aircraft Rocket Turbopump Oper-
takeoff weight; and thrust chamber specific
ating Characteristics
impulse (based on propellant combination, mix-
Oxidizer Fuel
ture ratio, and chamber operating efficiency).
Turbopump performance affects the vehicle pay-
Pumps:
Fluid .................... 90% H:O 2 JP-4 load in three ways:
Density .................. 86.7 lb/ft 3 50.5 lbtft s 1. Turbopump component weight.-Since the
Inlet pressure (total) ....... 25 psia 15 psia weight of the turbopump components is part of
Discharge pressure (total)., 862 psia 657 psia
stage burnout weight, it directly affects stage
Pressure rise in pump ...... 837 psi 642 psi
1910 ft payload.
Pump developed head ...... 1382 ft
Volume flow ............. 106.8 gpm 264 gpm 2. Required pump-inlet suction pressure
Flow rate ................. 20521b'sec 284 lb/sec head.-Required suction pressure directly trans-
Shaft speed ............... 30 000 rpm lates into required main propellant tank pressure
Efficiency ................ 60 percent 40 percent level. If it is raised, tank and pressurization
Shaft power ............... 87 bhp 25 bhp
system weights increase and thus reduce the
Turbine: stage payload for a given burnout weight.

Inlet gas pressure (total) ,.. 510 psia 3. Turbine gas [love rate.-Since the turbine
Exit gas pressure (static)... 25.5 psla drive gases, the flow rate of which is a function
Pressure ratio:
of propellant types and flow rates, pump dis-
Total inlet*static exhaust 2O
1364 ° F charge pressures, and of pump and turbine effi-
Inlet gas temperature .......
910 ° F ciencies, are usually ejected at a lower specific
Exhaust gas temperature ....
Gas flow rate ............. 0 54 lb/sec impulse than the thrust chamber gases (except in
Brake horsepower .......... 112 hp the case of the topping cycle), their flow rate
Shaft speed ............... 30000 rpm decreases the overall engine system specific
Efficiency ................ 40 percent
impulse and thus decreases the allowable vehicle
stage burnout weight. For a fixed weight of Solution

engines, tanks, guidance, and other equipment, a
decrease in allowable stage burnout weight re-
Turbopump equivalent weight per engine
sults in decrease of payload weight.
= 1900 + 92 x 55
The relation between turbine gas flow rate
: 6960 lb
and stage payload weight can be expressed by
an equivalent-weight factor, EWF, which is de-
fined as the decrease of payload weight (lb) as a Turbopump System Design Parameters
function of turbine flow rate (lb/sec). EWF is In the design of turbopump systems the follow-
proportional to the ratio between turbine exhaust ing parameters, which are often interdependent,
specific impulse (Is)te and thrust chamber spe- are considered paramount and will be discussed:
cific impulse (Is)tc. (1) Propellant properties
(2) Pump developed heads and flow rates
(3) Pump specific speeds
(6-1)
(Is)tc ] (4) Pump net positive suction head (NPSH)
(5) Pump efficiencies
The total effect of the turbopumo on allow- (6) Turbine overall performance and operating
able vehicle stage payload, i.e., the equivalent efficiency
weight of the turbopump, EW, is the sum of turbo- (7) Turbopump system cycle efficiency
pump component weight, Wtb , and the product of (8) Turbopump system calibration and off-
EWF and the turbine weight flow rate, _i,t: design characteristics
No simple rules are available for optimizing
EW : Wtb + (EWF) (#t) (6-2) the correlations of these parameters when de-
signing a specific type of turbopump for a given
engine systems application. Generally speaking,
Equation (6-2) permits a quantitative defini- however, available pump suction pressure to-
tion of the "best performing" turbopump system gether with the basic pump flow characteristics
possessing the lowest equivalent weight. Values will determine the maximum shaft speed at which
for the turbopump equivalent-weight factor EWF the unit can operate. The higher this shaft
range from 5(lb/lb/sec) for booster stage en- speed, the lower the turbopump weight is likely
gines to 200(lb/lb/sec) for upper stage engines. to be. Once the pump speed is determined, tur-
The equivalent-weight EW helps to establish the bine type, turbine driving arrangement, and tur-
optimized point between turbine and pump effi- bine power source are selected on the basis of
ciencies, and corresponding turbopump component efficiency, weight, simplicity, and other factors.
weights. It is not practical to include the effects
of pump inlet pressure on vehicle tank weights, Propellant Properties
since the required information is usually unavail- General data for some propellants used in
able to the turbopump designer. It is advisable liquid rocket engines are given in tables 1-3 to
however, that the pumps should be designed for 1-5. Table 6-8 lists those properties of com-
the lowest possible inlet pressure. monly used liquid propellants which have spe-
cific significance in the design of pumps• These
propellants include Earth-storable liquids such
as RP-1 and N204, cryogenics such as LO2 and
The following design data are set forth for LH2, and liquids having a wide range of physical
the hypothetical A-1 stage engine: and chemical properties.
Turbopump weight, Wtb , 1900 lb The low temperature of cryogenic liquids
Turbine gas flowrate, wt, 92 lb/sec creates problems with turbopump construction
Turbopump equivalent-weight factor EWF, materials, seals, bearings, lubricants, and through
55 lb/lb/sec the danger of ice formation• The total tempera-
Determine the turbopump equivalent weight (EW) ture range to which the structural elements of a
per engine. cryogenic turbopump may be exposed varies from
TABLE 6-3,-Fluid Properties of Commonly Used Liquid Propellants
aData at normal conditions
Conversion factors
Liquid Temper-I Vapor
Density, Viscosity,
ature, pressure, Head, ft
lb/ft a lb-sec/sqin
°F psia Pressure, gpm
lb/sec
psi
N204 .................................... 60 1t.1 90.7 1.59 4.96 0.637 * 10-r

H202(90 percent) ............................ 60 026 87.8 1.64 5.12 1.868 _10-"
N2H ( ....................................... 60 .158 63.3 2.28 7.12 1.49 × 10-7
RP-I ...................................... 60 031 508 (max) 2.83(min) 8.84(rain) 3.22 ×10 -7
. 49.8 (rain) 2 89(max) 9.00(max)
Ethyl-alcohol (95 percent) ..................... 60 .62 504 2.85 8.91 2.22 ,10-:
UDMH .................................... 60 183 49.66 2 90 9.06 .842 _ 10-:
50 percent UDMH,,'N2H 4 ..................... 60 1.77 5666 2.54 7.94 1.378 _ 10-7
LO: ....................................... -297.6 14.7 71.17 2.02 6.32 .2765_10 -7
LF 2 ....................................... -307 14.7 94.21 1.54 4.79 .353 x 10-:
L]'], ...................................... -4229 147 4.43 32.5 101.5 .0208×10-:
b_ LI 2 ...................................... -320.4 14.7 50.44 286 8.92 .226 ×10-:
bWatet ..................................... 60 .256 62.37 2.31 7.2 1.64 ×10-:
aNormal conditionsdo not necessarilyimply standardconditions,iftank pressureshave been applied.

blncludedhere because thesefluidsare frequentlyused as pump calibrationmedia.
-300 to -430 ° F at the pumps to 1200%1700 ° F hydrogen pump requires more than 10 times the
at the turbine. This induces temperature gradi- volume flow and driving power as compared to
ents between the various turbopump components other propellant pumps. Thus the design of
which must be accommodated and requires struc- liquid hydrogen pumps requires specific con-
tural flexibility or suitable devices to permit the siderations.
required thermal expansion and contraction. The viscosity of the pumped fluids effects a
Radial connecting pins are often employed to boundary layer along the surface of the flow
permit a cryogenic pump to contract independ- passages within the pump. There is a minimum
ently of a turbine and/or of a normal-temperature size of the impeller cross-section below which
pump. pump performance will fall off rapidly. As the
The vapor pressure of the propellants under passage clearances in the impeller are decreased,
normal en_ne operation conditions directly in- a point will be reached where the flow is pre-
fluences total suction pressure requirements at dominantly boundary layer, and viscous rather
the pump inlet. This will be further discussed than turbulent forces become predominant. This
in conjunction with the pump inlet net positive completely changes the performance of a pump.
suction head (NPSH). The drag effect of the boundary layers, together
The density variations of different propel- with the rotation of the fluid in certain pump
lants produce substantially different pump head- passages from an axial to a radial direction,
rise requirements, as well as large differences such as in the impeller of a centrifugal pump,
in volume flow. The power required per unit results in secondary flows being set up. This,
weight flow and pressure rise of a pump is in- together with friction losses which also are in
versely proportional to the density of the fluid. direct proportion to the viscosity of the fluid,
This is shown dramatically in the case of liquid and with leakage losses, constitutes a major
hydrogen which has a density of less than 10 portion of the energy losses in a pump. It is
percent of that of other propellants. For the seen that a high propellant viscosity tends to
same weight flow and pressure rise, a liquid lower pump efficiency.
DESIGNOF TURBOPUMPPROPELLANT-FEED SYSTEMS 189
Some liquid propellants, such as LF 2 and curves is thus obtained experimentally for the
N2H 4, are highly reactive chemically, and ther- range of speeds through which the pump may
mally unstable, beyond certain temperature lim- operate, all of them having similar parabolic
its. In the design of turbopumps for these pro- shape. Typical H-Q curves of a pump at various
pellants, special consideration must be given to speeds are shown in figure 6-19, along with the
the selection of compatible materials as well as system flow resistance curve. Together they
to the construction of mechanical parts. Seals, form the design operating range of a system.
bearings, and the protection (insulation) of the Two dimensionless coefficients are frequently
pump section against heat influx from the turbine used to indicate the head and flow characteris-
section following engine shutdown, are typical tics of a given pump. They are the pump head
problem areas. coefficient 0 and the pump flow coefficient _5.
The pump head coefficientis the ratioof rated
Pump Developed Heads and Flow Rates pump head (ft)to the maximum theoreticalhead
The term "pump developed head," AH (ft), is at zero flow formeridional(axial)inlet(no pre-
defined as the difference between pump dis- rotation)expressed as
charge head and pump suction head. The rela-
tionship between developed fluid pressure Ap AH
(6-4)
(psi) and fluid head AH (ft) is given by
g
144xAp(lb _
\in2/ where
AH (ft)= (6-3)
g, =overall pump head coefficient at rated de-
Fluid density l(f___) sign point (range is 0.2 to 0.7 for single
stage centrifugal pumps and up to 1.5 to
The required pump developed head at the de- 2.0 for multistage axial pumps, depend-
sign propellant flow rate (i.e., engine thrust ing on number of stages)
level) is dictated by the sum of the hydraulic AH=pump rated developed head, ft
resistances within the engine propellant flow u 2 =mean tip velocity of pump impeller at rated
system. These resistances include the pressure design rotating speed, ft/sec
drops across injector, thrust chamber manifold, g =gravitational constant, 32.2 ft/sec 2
cooling jacket, propellant valves and ducts, as The pump flow coefficient can be expressed as
well as the injector end thrust chamber pressure.
An additional pressure margin is usually allowed CrD2
for systems calibration. An engine propellant ¢ =-- (6-5)

U 2
flow system hydraulic resistance curve repre-

senting the resistance head to flow-rate relation-
f-ENGINE PROPELLANT FLOW
ship at various pump operating levels is shown SYSTEM HYORAULIC RESISTANCE
in figure 6-19. CURVE

r-PUMP H-O
Any value for developed fluid head of a given
pump is tied to a corresponding level of fluid \ // ,_IO%RATED
CHARACTERISTIC DESIGN
CURVE
flow at any impeller rotating speed. At the max- .

imum flow rates, the entire pump drive energy is
consumed by internal flow losses and kinetic
energy imparted to the fluid. Thus the measured
head rise at the pump outlet is zero. At the
other extreme, where fluid flow is zero (except DESIGN POINT-" '
I
for secondary flows), head rise is a maximum. A t
O Des
pump developed head versus flow capacity curve,
PUMP FLOW RATE, Q, gDm
commonly called H-Q curve, is derived by con-
necting the pump operating points between the Figure 6-19.-Engine system resistance and pump
two extremes for constant speeds. A family of characteristic curves.
where The relationships established in equations

6 =overall pump flow coefficient at rated de- (6-6a), (6-6b), and (6-6c) permit us to derive a
sign point (range 0.01 to 0.15 in rocket useful pump design parameter, the pump specific
engine application) speed, Ns:
cm2 =velocity of pump flow in meridional
direction at rated design point, ft/sec Ns : N(Q)°3 (6-7)
(AH) o-Ts
Rated pump flow, ft3/sec
A2 where
A2 =discl_,;',:'ge area normal to meridional di- Ns : pump specific speed (dimensionless)
rection, ft 2 N :pump rotating speed, rpm
Q =pump flow rate, gpm
_H :- pump developed head, feet
Pump Specific Speeds Pump specific speed is a characteristic value
For any given pump design, the relation be- defined as the rpm required to produce 1-gpm
tween fluid flow rate Q, developed fluid head flow at 1-foot head rise across the pump impeller
(or across the combination of inducer and im-
,AH, required driving power hp, and rotating
speed N can be defined by three expressions peller). In pump design, this term is very useful
called the affinity laws of a pump. These laws to classify inducers or impellers on the basis of
state: their performance and geometric proportions
(1) Pump volume flow rate varies directly regardless of the actual size or speed at which
with speed: they operate. Ns is a function of design con-
figuration; it does not vary significantly for a
series of geometrically similar impellers (having
QI _N1
(6-6a) the same angles and proportions), or for one
Q_ N_
particular impeller operating at any speed.
Since the H-Q characteristic curve of a pump
(2) Pump developed head varies directly as
ranges from zero flow at shutoff to zero head at
the square of the speed:
maximum flow, the specific speed for one curve
varies from zero to infinity. To make the term
(6-6b) definite, it is necessary to relate it to a defined
condition. The logical point is that of maximum
efficiency, usually the rated design point. It is
(3) Pump driving power varies directly as a generally understood that this point is meant
cube of the speed: when specific speed is stated.
Figure 6-20 indicates typical pump specific
hp I N_3
(6-6c) speeds for various impeller geometries. For a
hp 2 N23 given speed, a low value of specific speed is
characteristic for low-volume flow, high-head-
The affinity laws are based on the assumption rise pumps. Higher specific speed indicates a
that the pump efficiency is independent of the higher volume flow, low-head-rise pump.
rotating speed. Actual pump operation has 1. Radial-type impeller.-The head is largely
shown, however, that pump efficiency does developed by the action of centrifugal force.
change with speed. The degree depends on the This type is used for heads above 200 feet.
individual pump design, but is usually not more Specific speed ranges from 500 to 1200. Geo-
than 2 or 3 percent within a reasonable range metric proportion, r2/r 1, varies from 2 to 3.
from the rated design point. Furthermore, it 2. Francis-type impelier.-This type has an
affects only the power requirements rather than axial inlet, a radial discharge, and is used for
the relationship between developed head and lower heads. Specific speed ranges from 1200 to
flow rate. Thus the pump affinity laws hold 2400. Geometric proportion, r 2/q , varies from
quite well in most cases. 1.3 to 1.8.
Since 1 ft3/sec= 449 gpm
Oxidizer pump volume flow rate,
---.o}
l._w F
11 _°_0 J&$1(: IOI[LL|R TYPt
( .....
mG_ F_W
Nm=IZ1300
Qo = 1971 x 449/71.38 = 12 420 gpm
Figure 6-20.-Relationship between the pump From equation (6-7):

specific speeds and pump impeller geometries.
Oxidizer pump specific speed,
3. Mixed-flow type impeller.-The head devel- 7000 x (12 420) °s

Ns =
oped in this impeller is due partly to change in (2930) °-Ts
tangential velocity and partly to change in fluid
7000 × 111.7
velocity relative to the rotor. The discharge is - - 1980
395
partly radial and partly axial. The impeller
vanes are doubly curved. Specific speed varies
Fuel pump developed pressure,
from 2200 to 3500,
4. Propeller-type impeller.-The head devel- Ap= 1720-45= 1675 psi
oped by this type is through push of the vanes
only. Flow direction is axial. Specific speeds From equation (6-3):
range from 3000 to 6000 for multistage impellers,
and from 6000 to 12 000 for inducers. Fuel pump developed head,
AH= 144 x 167_-=

4790_ ft
Sample Calculation (6-2) 50.45
The following data are given for the propel- Fuel pump volume flow rate,
1ant pumps of the hypothetical A-1 stage engine
at the rated design point: Qi= 892 x 449/50.45 = 7960
Fluid Pump inlet Pump Pump Pump

suction dischargeweight speed, Fuel pump specific speed, Ns- 7000 x (7960) °s
Pump ideasity' pressures, pressure, flowrate, rpm (4790) °.rs
Ib/ft3psia(total)psia Ib/sec
=.7000 x.89:1 = 1083

Oxidizer.71.38 55 1505 1971 7000 576
Fuel... 50.45 45 1720 892 7000
Pump Net Positive Suction Head; Cavitation

Steady flow operation of a pump creates a
Determine the pump specific speeds.
low-pressure area at the pump inlet, thus allow-
ing the static head upstream of the pump to push
Solution fluid into the inlet at a continuous rate. There
are local regions within the pump which are at
Oxidizer pump developed pressure,
static pressures even lower than the inlet static
pressure. If the static pressure of the fluid at
Ap= 1505-55= 1450 psi
the pump inlet or any regions within the pump is
allowed to drop below the local fluid vapor pres-
sure level, these regions will cavitate; i.e., the
fluid will pass from liquid to vapor and form
Oxidizer pump developed head,
bubbles. The formation of vapor alters the effec-
144 × 1450 tive flow passages of the fluid and hence seri-
AH = - 2930 ft
71.38 ously affects normal pump performance. The
192 DESIGN OF LIQUaD PROPELLANT ROCKET ENGINES
subsequent collapse of these vapor regions cre- /--CURVE OF CRrTICAL NET

,/ POSITIVE SUCTION HEAD
ates local pressure forces which can result in
flow instabilities and/or substantial damage.
To avoid cavitation during operation of a pro- laJ
pellant pump, the pump-inlet available net posi- H I:'ERCENT RATED DESIGN FLOW
C_ DES- , ,
tive suction head, (NPSH)a, furnished by the

"r J__'_ RAI'ED DESIGN FLOW
propellant feed system upstream of the pump,
//x.
must be higher than the suction head above the PERCENT RATED DESIGN FLOW
propellant vapor pressure at which cavitation

would set in. (NPSH)a is the difference between (D
the propellant inlet total pressure head and the n
propellant
pressed as:
vapor pressure. This can be ex-
°
Q.
\ { NPSH }C DESIGN
NET POSITIVE. SUCTION HEAD AT THE PUMP INLET, FEEl

Pt × 144 Api× 144 pv × 144
(NPSH)a - + Z- -- (6-8) Figure 6-2I.-Typical cavitation characteristics
P P P
of a pump operated at rated design speed.
where
supplied by the vehicle must be larger than the
(NPSH)a =available net positive suction head,
ft (NPSH)c of the propellant pumps, or
Pt = propellant tank pressure, psia

(NPSH)a > (NPSH)c (6-9)
p =density of propellant, lb/ft 3
Z = height of propellant above the pump
It is useful to compare the suction character-
inlet and within the tank (corrected
istics of various pump designs on the basis of a
in flight for vehicle acceleration
design parameter called suction specific speed,
and gravity effects), ft
Nss, which is defined as follows:
Apt = pressure drop due to friction losses
within the propellant suction ducts,
psi Nss - N(Q)°s (6-10)
Pv =propellant vapor pressure for the pro- (NPSH)c °_
pellant temperature at the pump where
inlet, psia Nss : pump suction specific speed
N = pump rated design speed, rpm
V = pump rated design volume flow rate,
In design practice the term "critical net posi-
gpm
tive suction head," or (NPSH)c, is used to indi-
(NPSH)c : pump critical net positive suction
cate the minimum suction head required above
head, ft
the propellant vapor pressure to assure suppres-
Suction specific speed is related to the criti-
sion of cavitation. This critical net positive
cal net positive suction head in the same manner
suction head is defined by convention, as that
the specific speed is related to overall pump
value which will result in a 2-percent head-
developed head. Design values of suction spe-
generation loss at the rated design speed and
cific speeds for rocket propellant pumps range
flow rate of a given pump. Usually further re-
from 10 000 without inducers approximately to
duction in inlet NPSH below the (NPSH)c point
55 000 with inducers.
results in rapidly increasing cavitation. In turn,
Another coefficient describing pump suction
the developed head is further reduced, and non-
characteristics is the Thoma parameter r. It is
steady flow can result. This fluctuation of pro-
the ratio of critical net positive suction head
pellant flow can cause erratic combustion in the
(NPSH)c and rated pump developed head AH, or
thrust chamber. Figure 6-21 represents the
cavitation characteristics of a typical pump
operated at rated design speed. To insure a (NPSH)c _ 1/ Ns _1.333
(6-11)
margin of safety for pump operation, the (NPSH)a
The Thoma parameter is a function of pump de-

Pump Pump Pump
sign quality and specific speed. Pump Fluid volume developed
Pump
For a given vehicle (NPSH)a, the pump suc- speed,[
pumped flowrate, head, (NPSH)c'
rpm gpm ft ft
tion characteristics (Nss) determine the maximum
permissible rpm at the design flow rate. A high 3xidizer.. 7000 LO_ 12420 2930 58
pump Nss or vehicle (NPSH)a permits a higher _uel ..... 7000 RP-I 7960 4790 70
pump rpm, lower turbopump weight, and possibly
higher turbine performance. It is desirable to
operate a pump at the highest practicable speed. Determine pump suction specific speeds and
Figure 6-22 represents the results of a para- Thoma parameters.
metric turbopump system design study for a typi- (b_..)The following system design data are
cal LO2/RP-1 booster stage rocket engine sys- given for the A-1 stage vehicle:
tem, such as the A-1 stage engine. The study
reflects the effects of N, (NPSH)c, and Nss on
Longitu-apermanent[
the selection of turbopump configuration. dinal static Propel-
lant
In addition to the pump (NPSH)c values during !distance pressure
Fluid Tank !between
steady-state operation, the engine starting tran- loss in temper-
Pump density, pressure, tankand suction ature
sient pump (NPSH)c must be determined and lb/ft _ psia pump duct. atpum_
specified to permit satisfactory engine accelera- inlet,
inlet, including OF
tion from zero to nominal design speed and flow R valves, psi'
rate in the desired time and manner. The start-
ing (NPSH)c depends on the rate of acceleration 3xidizer 71.38 60 3.5 5 -297.6
Fuel .. 50.45 50 25 8 60
and on the control system of the engine, as well
as on vehicle acceleration, gravity effects, and
aThe dynamic bead at the pump inlet is considered a
propellant suction duct geometry. Therefore, part of the available NPSH and thus is not subtracted from
sufficient tank pressure must be provided to the tank pressure.
accelerate the propellant and to overcome the
hydraulic resistance in the suction duct, as well
as to supply the necessary pump (NPSH)c during
all phases of system operation. Determine the available NPSH at the pump inlets.
Sample Calculation (6-3) Solution
(a__)The following test data are given for the (a_) Substitute given data into equation (6-10):
propellant pumps of the A-1 stage:
Oxidizer pump suction specific speed,
7000 × (12 420) os

( 't'_ STAP.d[$I Ns _
(58)o.;s
AIir_
_ 7000 x 111.7
TUPtOINE "
= 37 230
Fuel pump suction specific speed,
Nss - 7000 ×(7960) °s

(70) o._
{ NPe_,I IC,, F'T
_ 7000 x 89.i
Figure 6-22.-Effects or"N, (NPSH) c, and Nss on
24.2
turbopump configuration selection for a typical
LO2/RP-1 booster stage rocket engine system. = 25 790
194 DESIGN OF LIQUID PROPELLAHT ROCKET ENGINES
Substitutegiven data intoequation (6-II): disk, which is relatively minor, and (b) a pump-
ing action on the fluid in contact with the disk,
Thoma parameter of the oxidizer pump, whereby the fluid is circulated locally by cen-
trifugal action. The energy loss due to disk
58
friction is transformed into heat and can appre-
r = 2--9-_ = 0.0198
ciably increase the temperature of the fluid.
3. Mechanical 1osses.-These are losses in
Thoma parameter of the fuel pump, bearings and seals caused by mechanical fric-
7O tion.
r =4_=0.0146 4. Leakage Iosses.-To prevent the pumped
fluid from leaking back to the suction side after
(b) From table 6-3: it has passed the impeller and is at outlet pres-
sure levels, close-clearance labyrinth-type seals
Vapor pressure of liquid oxygen or wearing rings are provided. Leakback lowers
= 14.7 psia at -297.6 ° F the flow capacity of a pump and thus increases
required pumping power.
Vapor pressure of RP-1 = 0.031 psia at 60 c F For a new design, losses are difficult to pre-
dict and are usually estimated from data derived
Assume the vehicleis at sea-levelholddown during actual testing and other past experience
conditionand substitutegiven data intoequation with similar designs.
(6-8): The overall efficiency of a pump, r/p, can be
expressed by the ratio of pump fluid horsepower
output, fhp, to brake horsepower input by the
Available NPSH at oxidizerpump inlet
pump drive, bhp:
(60-5-14.7)× 144 +3.5
(NPSH)a - 71.38 fl_p
(6-12)
_P = bhp
=81.5+3.5=85ft
The pump fluid horsepower flip is the actual

Available NPSH at fuelpump inlet
usable output delivered by the pump. It is the
(50-S-0.031)
× 144 + 25 product of delivered propellant weight flow, _i,p
(NPSH)a = 50.45
(lb/sec), times the actual head AH(ft) developed
=120+25=145 ft by the pump, divided by a conversion factor:
Pump Operational Efficiency; Losses (_p,_H (6-13)

fhp = 550
Several types of energy losses occur during
pump operation which affect efficiency.
1. Hydraulic losses.-These include friction The brake horsepower bhp represents the me-
losses in the passages and flow turbulence chanical horsepower delivered to the pump by the
losses. The friction losses are a function of the drive. This delivered brake horsepower is con-
"wetted areas" in the passages and of the rough- sumed in the pump as fluid horsepower and as
ness of their surfaces. The turbulence losses the various losses.
are caused by disturbances in certain regions of
the pump, such as at the inlet and outlet edge of
the vanes of both impeller and diffuser and in bhp: fhp + (hp)h + (hp)df + (hp)m + (hp)l (6-14)
the return guide vanes.
2. Disk friction losses.-The energy required where
to rotate a disk, such as an impeller or inducer, bhp = brake horsepower
in a fluid is known as disk friction loss. The fhp = fluid horsepower
disk friction losses are due to two actions: (hp)h = horsepower required to overcome hy-
draulic losses
namely, (a) the actual friction of the fluid on the
(hP)dt= horsepower required to overcome disk developed head and rotating speed ranges from
friction losses 60 to 85 percent. This is about 10 percent lower
(hp)m =horsepower required to overcome me- than that of industrial pumps. The variation of
chanical losses pump efficiency with specific speed, at various
(hp)l =horsepower required to overcome leak- pump capacities, is shown in figure 6-23. Figure
age loss 6-24 shows the correlation between pump flow
The efficiency of a pump is related to the rate and the following three parameters: devel-
volume flow rate Q, the developed head AH, and oped head, efficiency, and required brake horse-
the rotating speed N, all of which are included power, for a typical centrifugal pump.
in the pump specific speed Ns. Of these, the
most important one is the design volume flow
rate or capacity of the pump. There is a defi-
nite trend toward increased efficiency with higher The following test data are given for the
pump capacities, because hydraulic, disk fric- propellant pumps of the A-1 stage engine:
tion, and mechanical losses represent a smaller
percentage of the total brake horsepower when
large capacities are being handled. On the other Fluid Pump volumel Pump Drive shaft
hand, increased pump-developed head and rotat- Pump pumped flow rate, developed power, hp
gpm head, ft
ing speed for a pump of given capacity results in
lower efficiency because of increased disk fric-
Oxidizer... LO 2 12 420 2930 14 850
ti6n and mechanical losses. The overall effi- Vuel ....... RP-1 7960 4790 11 790
ciency of rocket engine propellant pumps of high
D5
[
ql. dl, ,wll _ _
..
_
.[
III, ._... I
I
i
b.]
tD
e,-
W
r,
" I I I I -7"
!_:L:
Z
I,U
tL
U.
t_ i
_0 I000 aO00 $000 4000 IO, O00 (5,000

RPM,
VALUES OF SPECIFIC SPEED, NS: H 3/4
Figure 6-23.-Variation _f pump elliciency with specilic speed. •

Determine the overall efficiencies _p of both Turbine Overall Performance and Operating
pumps. Efficiency
The overall performance of a turbine is de-
fined as the ratio of turbine shaft horsepower
Solution
output, thp, and required turbine working-fluid
Substitute given data and flow rates from flow rate, #vt, in pounds per second or
sample calculation (6-2) into equation (6-13):
Overall turbine performance
Oxidizer pump fluid horsepower z-- thp (hp/lb/sec) (6-15)

vvt
1971 ×2930
fhp- 550 - 10500 hp
In general, turbine overall performance depends
upon two variables: the available energy content
Fuel pump fluid horsepower per pound of turbine working-fluid, AHt, and the
892×4790 overall turbine operating efficiency, 7/t.

fhp- 550 -7760 hp The available energy content of the turbine
working fluid is defined as the enthalpy drop per
pound of working fluid in the turbine

AH t=H o- H e (6-16)
Oxidizer pump overall efficiency

where
10500
=70.7% AHt---available energy content of the working
9P=14 850 fluid, Btu/lb
H o = enthalpy per unit weight of the working
7760 . fluid at turbine-inlet, Btu/lb
Fuel pump overall efficiency _p=]-_-_=6a.8_
He = enthalpy per unit weight of the working
fluid at exhaust pressure, Btu/lb, as-
suming isentropic expansion
PUMP DEVELOPED HEAD H, FEET Using equations (1-10) and (1-13), equation
(6-16) can be rewritten as
PUMP EFFICIENCY _p, PERCENT
PUMP REQUIRED POWER, Bhp
I _(Pe_ Y /-I
)'=' (6-17)
AHt=Cp(T°- Te)=CpT° \Po/ J
where
DEVELOPED HEAD
Cp:working fluid specific heat at constant

RATED
DES IGN /'- EFFICIENCY

pressure, Btu/'lb-F
T o = working fluid total temperature at turbine
COND TION /
- I / inlet, °R
Te = working fluid static temperature at turbine
exhaust, °R
Po =working fluid total pressure at turbine
inlet, psia
Pe = working fluid static pressure at turbine
exhaust, psia
y =working fluid specific heat ratio
PUMP FLOWRATE Q, gpm
The ratio of turbine inlet and exhaust pressures,
Figure 6-24.-H-Q, e[ficiency, and required power P_,/P2, can be expressed as the turbine pressure
characteristic curves o[ a typical centrilugal ratio Rt, which is a frequently used parameter in
pump. turbine design.
perature is offset by the turbine efficiencylosses

17 (blade losses; see below) resulting from higher
AH_--CpTo -
(6-18)
gas jet speed (spouting velocity Co) which is
proportional to the turbine-inlet temperature.
It is seen that the available energy in the Figure 6-26 shows the relationship between
turbine working fluid is a function of gas proper- turbine pressure ratio and available energy of
ties and inlet temperature, and of the turbine the working fluid for a turbine-inlet temperature
pressure ratio. Most of the turbine working fluids of 1200 ° F, again for the propellant combinations
for rocket engine application are fuel-rich prod- LO_/RP-1 and LO2/LH 2. Although a large
uct gases generated by bipropellant combustion. amount of energy may be available in a working
Typical working fluid properties are listed in fluid, it may be difficult to convert it efficiently
table 6-4. into turbine shaft horsepower because of the
Figure 6-25 shows the relationship between severe weight limitations on rocket engine turbo-
turbine-inlet temperature and available working pumps. Thus the available pressure ratio of a
fluid energy for a turbine pressure ratio of 20 for turbine often cannot be fully used.
the two propellant combinations LO2/RP-I and The overall efficiency of a turbine, 77t, is
LO2/LH 2. Turbine inlet temperatures are lim- defined as the ratio of turbine shaft horsepower,
ited by the high-temperature properties of the
turbine construction materials. A practical de- 2O00 I
sign limit is around 1700 ° F. Above certain

levels the gain from a higher turbine-inlet tern- _" 1#00
i E_
-r
TABLE 6-4.-Properties of Typical Fuel-Rich <:1 12oo
>.
Combustion Product Gases
ul
.oo
Inlet
Fluid temper- Btu/
Cp, y R, Mixtur_
ratio, /
ature ft/° R ,< 4oo
ib° F O/F
°F .<
LOX/RP-1 ......... 1100 0.635 1.097 43.3 0.308 IOO_ 1100 1200 1300 _4OO 1500 I_00 1700
TURBINE INLET TEMPERATURE, F

1150 .639 i.i00 45.1 .320
1200 .643 1.106 47.1 .337
Figure 6-25.-Effect of turbine-inlet temperature
1250 .646' 1.111 58.6 .354
on working fluid available energy.
1300 .648 1.115 50.4 .372
1350 .651 1.119 51.8 .390
1400 .653 1.124 53.6 .408
1450 .655_1 1.128 55.4 .425
2O00
1500 .6571 1.132 58.0 .443
1550 .659 1.137 59.0 .460
1600 .660 1.140 60.7 .478
1650 .661 1.144 62.4 .497 tOX/U41
1700 .662 1.148 64.0 .516

12OO
N20(/CH,(UDMH) .. 1400 .380 1.420 87.5 .lt0
1500 .398 1.420 91.6 .165 g
1600 .416 1.420 95.7 .220 Z e
1700 .434 1.420 999 ! .274
1800 .452 1.420, 104.0: .328
1900 .470 1.4201108.21 .382
LOX/LH 2 ......... 1000 2.05 1.374 434 .785 .I
1200 1.94 1.3641 403 .903
1400 1.86 1.3541 378 1.025 o _ i ;2 ,; 2'o ;, ==
PRESSURE RATIO
1600 1.80 1.343 358 1.143
1800 1.73 1.333 336 1.273 Figure 6-26.-Effect of turbine pressure ratio on
2000 1.69 1.322 320 1.410
working fluid available energy.
thp, to enthalpy drop rate or the available energy Also, centrifugal action of the rotor flisk causes
delivered rate of the working fluid. some of the gas to flow radially to the casing
and to be dragged along the face of the casing
550 thp thp by the rotor blades.
5. Mechanical losses.-These result from the
mechanical friction in bearings and rotating
seals.
The design of turbines for rocket turbopumps

Combining equations (6-15) and (6-19): tends toward the simpler and lighter impulse
types, with most of the expansion occurring in
Turbine overall performance the stationary elements. Figure 6-27 shows the
typical efficiency curves of various impulse-type
turbines. The design problem becomes one of
balancing efficiency (optimum velocity ratio
_7?tAHt_ 7ItCpT° _ 1 -_-
(6-20) U/Co, weight (number of stages or rows), and
O.7O7 0.7O7
structural considerations. A higher performance
can be achieved by employing a working fluid
In turbine operation efficiency can be affected with high available energy, and by matching its
by- high gas spouting velocity Co with a high rotor
I. Nozzle losses.-Similar to those in thrust
blade pitch speed U. However, the blade speed
chamber nozzles, these losses are due to the gas
is often limited by the required pump rpm, by the
leaving the nozzle at a lower velocity as com- practical size of the rotor wheel, and by the
pared to that of an ideal nozzle. The losses are
strength of materials. The optimum velocity
due to flow turbulence, fluid friction, and loss of
ratio (or optimum value of blade speed for a
heat to and through the turbine nozzle blocks. given gas spouting velocity) is reduced by the
2. Blade losses.-These are caused by resid-
use of velocity or pressure-compounded arrange-
ual velocity of the gas as it leaves the rotor
ments (shown schematically in fig. 6-9 and 6-10).
blades, the obliquity of the nozzle (i.e., the However, these designs increase weight and
nozzle angle 0 in fig. 6-8 cannot be made zero),
complexity.
flow turbulence, and fluid friction. Losses due
In general, for a direct-drive configuration
to residual gas velocity can be reduced by opti- (fig. 6-13) with an rpm lower than ideal, a low
mizing the turbine blade-to-gas velocity ratio
value of U/Co results, and a velocity compounded
U/Co, where U is the pitch speed of the rotor turbine will be used because of its low overall
blades, and C o is the ideal spouting velocity of
weight and simplicity. Where a reduction gear is
the gas based on available energy and isentropic
expansion. These conditions can be analyzed
I00
quantitatively by means of turbine gas flow
velocity diagrams which will be discussed in _-_ PllSSUR_ tOM P OU_,C_D
section 6.5. Flow turbulence can be reduced

through improved blade shape and through full
turbine nozzle admission.
3. Leakage or clearance losses.-The clear-
, /
ante required between rotor blade tips and cas-
ing permits some gas to leak past the blades
without doing work, thus causing energy losses. Z 2(
The gas leakage from stage to stage in a multi- u
stage pressure-compounded turbine due to re-

i
quired clearance between shaft and sealing 01 (1.2 G.3 04 05 O.6
OVERALL ISENTROPIC VELOCITY RATIO- U/Co
diaphragm results in similar losses.

4. Disk friction losses.-Fluid friction occurs Figure 6-27.-Typical efficiency curves of
at the interface of gas and rotor disk surface. impulse-type turbines.
1 I
provided between pumps and turbine, a higher Total shaft horsepower required by the pumps
value of U/Co is possible and a more efficient (sample calculation (6-4)):
pressure-compounded turbine can be used.
Since the turbine pressure ratio has only a 14 850 + 11 790-- 26 640 hp
small effect upon available energy content of the
working fluid (fig. 6-26), the power level of a Thus a margin of 510 hp is available for auxiliary
turbine is usually regulated by controlling the drives and contingencies.
inlet pressure Po and in turn the flow rate wt of Substitute the values for thp, AHt, and ¢¢_
the turbine. into equation (6-19):
0.707 × 27 150
Sample Calculation (6-5) Overall turbine efficiency r/t - 92 x 359
The following test data are given for the
=58.2%
turbine of the A-1 stage engine turbopump:
Turbine gas mixture ratio, LO2/RP-1 = 0.408
Gas total temperature at turbine-inlet,
T O= 1860 ° R (1400 ° F)
27 150 hp
Gas total pressure at turbine-inlet, Po = 640 Overall turbine performance = 9_ =295 lb/se_
psia
Gas static pressure at turbine exhaust, Pe = 27
Turbopump System Cycle Efficiency
psia
Turbine gas flow rate, ti,t= 92 lb/sec Turbopump system cycle efficiency is an
Turbine shaft speed, 7000 rpm indicator of the energy losses and their effect on
Turbine shaft torque, 20 380 ft-lb overall engine systems performance because of
Determine the overall turbine efficiency in per- turbopump operation. It can be defined as the
cent and the performance in horsepower per ratio of the specific impulse of the complete
lb/sec of turbine flow ¢_t. turbopump-fed engine system, (Is)eng, to the
thrust chamber specific impulse, (Is)tc.
Solution
(]s)eng
From table 6-4, for a LO2/RP-1 mixture ratio Turbopump cycle efficiency r/c- (6-21)
(Is)to
of 0.408, we derive a specific heat Cp = 0.653
and a specific heat ratio y = 1.124. Generally, two types of energy cycles are
Substitute the above and other given data into employed in rocket turbopump applications: the
equation (6-17) to obtain the available energy separate-flow cycle and the topping-flow cycle.
content of the turbine gas: Because of certain systems-design limitations
the second type is used less frequently than the
former.
In the separate-flow cycle the turbine exhaust
AHr,:0.B53x1860 I 1-(6-_0) _'J
1.124--1 -]
gas flow is discharged separately or is ducted

= 0.653 × 1860 [1 - 0.705] into the low-pressure region of the main thrust
chamber exit nozzle. This is shown schemati-
= 359 Btu/lb
cally for cases (A), (B), and (C) in figure 6-12,
and for the A-1 and A-2 stage engines in figures
Turbine shaft horsepower
2-10 and 3-3. The separate-flow cycle system
Turbine shaft torque × 2 _rN design affords a lower discharge pressure by the
thp - 33 000 propellant pumps and yields a higher pressure
ratio for the turbine. Generally, this results in a
_ 20 380 x 2 rrx 7000
simpler and lighter system. However, the tur-
33 000
bine exhaust gas of a separate-flow cycle is
= 27 150 hp usually not utilized efficiently to generate thrust.
2OO DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
Thus, it has a relatively low cycle efficiency. topping flow cycle is sometimes considered a
The separate-flow cycle efficiency can be ex- "no loss" cycle. However, that net energy which
pressed as is required to lift the propellants from their re-
spective pump inlet pressures to the thrust cham-
(ls)eng _ Wtc (Is)tc + _'_ (Is)re ber nozzle stagnation pressure must be consid-
ered. The topping flow cycle efficiency may be
r/sc= (Is)re Weng (Is)to
approximated as
Fte+ Fte
(6-22)
IPeng (/s)tc Ep
where _hc = (Is)eng
(/she _ 1 Cp(Tc)ns
_sc = separate-flow cycle efficiency
MRhHo + hHI
(Is)eng -- engine system specific impulse, sec (6-23)
(ls)tc =main thrust chamber specific impulse, 778 (1 + MR) Cp (Tc)n s
see
where
(Is)re = turbine exhaust specific impulse, sec
=engine system total propellant flow the = topping flow cycle efficiency
liPeng
rate, lb/sec Ep =ideal energy required to pump 1 pound
Wtc = thrust chamber propellant flow rate, total propellant flow of an engine
lb/sec system from pump inlet to main cham-
= turbine gas flow rate, lb/sec ber nozzle stagnation pressures
Cp = specific heat of the thrust chamber
gases, Btu/lb-F
The efficiency of the separate-flow cycle can
(Tc)ns--thrust chamber nozzle stagnation tem-
be increased somewhat by better utilization of
perature, °R
the turbine exhaust gases to generate thrust. A
MR = propellant mixture ratio of the engine
higher expansion area nozzle may be attached to
system, O/F
the turbine exhaust duct for upper stage engine
£Ho =pressure head, based on the difference
applications, for a higher (Is)re, or the turbine
between thrust chamber nozzle stag-
exhaust gases may be burned with additional
nation and oxidizer pump inlet pres-
propellant in an afterburner similar to that in a
sures, ft
jet engine. Separate-flow-cycle efficiencies
AHt :pressure head, based on the difference
range from 0.96 to 0.99.
between thrust chamber nozzle stag-
In the topping flow cycle the turbine exhaust
nation and fuel pump inlet pressures,
gases, before being expanded and ejected through
ft
the thrust chamber nozzle, are ducted into the
Topping flow cycle efficiencies may range from
thrust chamber combustion zone for further reac-
0.996 to 0.9996. Typical cycle efficiency versus
tion with one of the main propellants. This is
chamber pressure curves of various propellants
illustrated schematically in cases (D) and (E) of
are illustrated in figure 6-28.
figure 6-12. Because this system requires a
turbine exhaust pressure higher than the thrust
chamber pressure, it results in higher propellant I O0
pump discharge pressures and a much lower tur- 099
bine pressure ratio. Also, for bipropellant com- _-- -- r-
binations, two-stage combustion is required. , 091
Usually combustion products of 100 percent of Z
097
one propellant and a portion of the other are used

as the turbine drive fluid. This method results
in higher cycle efficiency, but tends to be _ 095
heavier and less flexible. 1000 2O0O 3OO0 40O0 5O0O
CH/_.MSER PRESSURE, PSIA

Since the residual kinetic and chemical ener-
gies of the turbine exhaust gas are generating Figure 6-28.-Typical cycle efficiency versus
thrust as efficiently as the main propellants, the chamber pressure curves of various propellants.
Sample Calculation (6-6) Total engine system propellant flow rate

= 2768 + 92 = 2860 lb/sec
(a_) The following component test results are
given for the A-1 stage engine, as schematically
Substitute the above data into equation (6-22)
shown in figure 2-10, for sea level and rated
to obtain the flow cycle efficiency of the turbo-
design conditions:
pump
Thrust chamber oxidizer flow rate. 1941
lb/sec
2768x270 + 92×29.34 _ 747 300 + 2700
Thrust chamber fuel flow rate, 827 lb/sec = 0.971
_sc- 2860x 270 772200
Thrust chamber nozzle stagnation pressure,
1000 psia
From equation (6-21), the engine system spe-
Thrust chamber thrust, 747 300 lb
cific impulse at sea level
Gas generator oxidizer flow rate, 26.7 lb/sec
Gas generator fuel flow rate, 65.3 lb/sec
(Is)eng = 0.971 × 270= 262.2 see
Thrust generated by turbine exhaust, 2700 lb
Determine the cycle efficieucy of the turbopump
Total engine system thrust
system, engine system sea-level specific im-
Feng : 262.2 x 2860= 750 000 Ib
pulse and thrust, and overall engine system pro-
pellant mixture ratio at rated conditions.
1941 + 26.7
(b_) Assume that, as an alternate, a topping Engine system mixture ratio- -2.20
827 + 65.3
flow cycle is used for the turbopump of the A-1
stage engine and that this engine system is to
The calculations can also be made by first com-
be operated at the same engine system thrust and
bining Ftc and Fie to derive Feng. Then Feng
mixture ratio as in case (}). Estimate the ideal
and l),'eng
are used to calculate (Is)eng. rlsc can
turbopump cycle efficiency and engine system
sea-level specific impulse with the following be calculated by forming the ratio (Is)eng to
(Is)to.
assumptions: (1) no change in (Pc)ns, combustion-
produced gas properties, and performance of the
thrust chamber due to the minor changes in mix-
ture ratio and two-stage combustion; (2) oxidizer b(b.)
Use the combustion product gas properties
pump inlet pressure = 55 psia; and (3) fuel pump from sample calculation (4-1)(a__):
inlet pressure = 45 psia.

(Tc)ns = 6460 ° R, y= 1.222
Solution 1544 y
1941 _=22.51b/mol: C9=778_(y - 1)

(a__)
Thrust chamber mixture ratio =--_ = 2.35
= 0.486 Btu/lb-deg F
Thrust chamber propellant flow rate

wtc = 1941 + 827 = 2768 lb/sec Pressure head difference between chamber
pressure and oxidizer pump inlet:
Sea-level thrust chamber specific impulse
AH o = (1000 - 55) × 144/71.38 = 1910 ft
747 300
(Is)tc- 2768 -270 sec
Pressure head difference between chamber
pressure and fuel pump inlet:
Turbine gas flow rate wt=26.7+65.3= 92 lb/sec
AHt= (1000- 45) × 144/50.45 = 2728 ft
Sea-level turbine exhaust specific impulse

Substitute these data into equation (6-23) for
2700 the topping flow cycle efficiency of the turbo-
(Is)re =_= 29.34 sec pump
necessary to calibrate a turbopump prior to its

77re= _/ 1 778 × (i
2.2÷ x2.2)
1910x0.486_6460
+ 2728 integration into an engine system. Here, the
performance of each pump is determined experi-
: ,,,'0.9991 --0.9996
mentally. By modifying the pump geometry (im-
peller trimming) and varying the discharge hy-
Since (Is)to--270 see, substitute 77tc and (Is)re draulic resistances, the desired operating
into equation (6-21); the ideal engine system characteristics required by the engine system
specific impulse
are achieved. For turbopumps with a single
direct or geared turbine drive (as shown in fig.
(/s)eng-- 0.9996 × 270 = 269.9 sec 6-13), the calibration procedure begins with the
experimental determination of that shaft speed at
Turbopump System Calibration and Off-Design which one of the pumps (usually the oxidizer
Characteristics pump) develops the required head and flow.
Simultaneously, the suction characteristics of
Ideally, an engine system is designed to
this pump are determined at this speed. The
operate at a single set of conditions, specifically other pump is then operated at the same speed
at rated thrust and mixture ratio. Each of its
and the discharge adjusted for the required flow.
components in turn is designed for optimum func- Based on the pressure readings, the diameter of
tion at that rated thrust level. In addition, how- the pump impeller is trimmed on a lathe until the
ever, most of these engine components, notably desired head is produced at rated speed and
pumps and turbines, are also required to operate flow. The pump suction characteristics at these
satisfactorily within a certain range away from conditions are also determined. Figure 6-30
the design point. This may be caused by- shows the trimming effects of a typical pump.
(l) System and component calibration charac-
The trimming procedure requires that the pump
teristics
impellers be made sufficiently large initially,
(2) System operational deviations since addition of material is not feasible. The
(3) System start and shutdown transients calibrated pumps are then combined with a tur-
(4) Special system requirements, such as
bine. The amount of turbine working fluid which
throttling
must be bled from the pump discharges to operate
Figure 6-29 illustrates typical operating
ranges for the turbopump of a rocket engine sys-
tem. The elliptically shaped areas represent
ranges of operational probability. For instance, 120
d
the 95-percent ellipse envelopes those values at
which the pumps or turbine will operate with a
95-percent probability.
(z:
Because of engine and component performance
I00
tolerance requirements and to correct for hydro-
dynamic variations within the components, it is
Q:
90
6aoo NO=_A_
80
6000 Q.
q
/
¢3"=¢64OO _
_ 6ooo ,_t_ w
5500 70
$2oo _ i
45oo13 114 115 116 17 4o $1o _o 7o 8o 60
60 70 80 90 IO0 HO 12'O
PUMP FLOW, GPM IIN THOU$=_ND$} TURBINE POWER,B HP IIN J'HOUSANDS)
PUMP FLOWRATE Q, PERCENT OF RATED LEVEL

Figure 6-29.-Typical pump and turbine operating
ranges o[ a rocket engine system. Figure 6-30.-Trimming effects of a typical pump.
DESIGN OF TURBOPUMP PROPELLANT-FEED SYSTEMS 2O3
the gas generator is now established experimen- dome priming. The time in which a rocket turbo-
tally at the required operating point. This pro- pump must attain full-power level is substantially
cedure considers the influence of turbine varia- shorter than that of conventional turbomachinery.
tions. The turbopump is then matched with the Frequently, full-power operation must be achieved
remaining engine system in complete engine in less than 1 second. This is dictated by the
systems calibrations. Adjustments in engine need for propellant consumption economy in
mixture ratio can be made by orificing the dis- flight and for avoiding the possibility of flow
charge of one of the pumps. Orifice installation instability in thrust chamber and other compo-
considerations and type of propellant often make nents. This, together with other transient re-
it desirable to assure that orificing is always quirements such as throttling, must be satisfied
done for the same propellant; in LOX/RP sys- by the off-design characteristics of a turbopump.
tems, usually the fuel. For this reason, the The latter substantially influence selection of
pump for that system is trimmed for a slight ex- type. Basically, the problem is that of coupling
cess head. Adjustments in engine thrust level the pump characteristics with those of the rest
are made by varying the turbopump speed. of the engine system under off-design operating
An engine system is frequently required to conditions. Figure 6-32 shows typical off-design
operate at certain off-design levels. For in- characteristics of various types of pumps.
stance, upper stage vehicles often require an One of the most significant pump off-design
effective propellant utilization system based on characteristics is the pump stalling point which
variation of the engine mixture ratio. (Seech. II.) usually occurs in the low-flow region. The pump
This involves control of the discharge of one operation tends to be erratic at this point, re-
pump and possibly the control of turbine supply sulting in the abrupt loss of developed head and
gas if correction of thrust due to the shifting of the danger of overspeeding.
mixture ratio is desired. Design and develop-
ment of the turbopump must consider these opera-
tional requirements.
During the start and shutdown transients of
14 I I
an engine system, propellant flow and pressure
characteristics are determined by engine system
._'_PUMP STALLINGPOll'
design characteristics. Figure 6-31 shows the
propellant flows and chamber pressure buildup
1.2 BACKWARD / ............ "'-.. .+," \
history for a typical engine start transient. Note
temporary oxidizer flow drop as a result of LOX
J RADI_L
\--.
""
I00 ,,=,_
I \
,- \
_- tD p_
> 80
//'--
"-,- \
O,.uJ
+'_- \
/'
OXIOIZER_ I /
u- _ {z: AXIAL
FUEL-"_
•2 --
STARTING RANGE
0
I
O Q5 1.0 1.5 20 2.5 30
0 .2 .4 .6 .8 1.0 1.2 1.4
"--'_TIME FROM MAINSTAGE SIGNAL, SECONDS
PUMP FLOWRATE, Q
RATED DESIGN PUMP FLOW, Qd,s.
Figure 6-31.-Propellant flow and chamber pres-
sure transient characteristics during engine Figure 6-32.-Typical oH-design characteristics
system start. o[ various types of pumps.
2O4 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
6.3 DESIGN OF CENTRIFUGAL PUMPS

(6-24c)
Because of its specific needs, the rocket
industry has developed its own pump design
where
approaches which may differ from those for con-
ventional applications. In addition, designers gl, Q1, and AH] :rotating speed (rpm), flow
may employ their individual methods of analysis rate (gpm), and developed
and calculation. However, the broad underlying head (ft) of the existing
model at rated conditions
principles are quite similar. The range of
speeds, proportions, design coefficients, and N2, Q2, and AH: : rotating speed (rpm), now
other mechanical detail for rocket engine pumps rate (gpm), and developed
has been well established by earlier designs as head fit) of the new pump
at rated conditions
well as through experiments.
[: D2 /D 1 : scaling factor
D, : impeller diameter of the
General Design Procedures existing model, ft
As a rule, rated pump head-capacity (H-Q) D2 = impeller diameter of the new
requirements and expected available NPSH at pump, ft
the pump inlets will be established by engine This approach assumes that other dimensions of
system design criteria. The first step then is to the pump are in approximately linear proportion
choose a suitable suction specific speed (Nss) to the impeller diameter.
If a suitable model is not available for the
and the type of inducer which will yield the
highest pump speed (N) at design conditions design of a new pump, the designer can use
(eq. (6-10)). The pump specific speed (Ns) or "design factors" established experimentally by
type of impeller can now be established from the other successful designs. These may permit
chosen pump speed and required head-capacity establishing relations between rated pump devel-
characteristics. Owing to its relatively light oped head and flow rate, and such parameters as
weight and simplicity of construction, a single- velocity ratios. However, best results are ob-
stage centrifugal pump may be given first contained through experimental resting of proposed
sideration. design itself. The test results then are used for
With suction specific speed and specific design revisions and refinements.
speed of the proposed pump design established, In the discussions below, the following basic
the designer can now look for a suitable "design symbols are used:
model" among comparable existing pumps which c--flow velocities absolute (relative to ducts
approximate the desired performance. The latter and casing)
v : flow velocities relative to inducer or im-
includes satisfactory suction requirements, suit-
able head-capacity characteristics, and accepta- peller
ble efficiency. If a suitable model is available, u = velocities of points on inducer or impeller
the design calculations of the new pump will Subscript:
0 : inducer inlet
include application of a scaling factor to the
parameters of the existing model. The following 1 = inducer outlet : impeller inlet
correlations are valid for pumps with like spe- 2 = impeller outlet
cific speed, based on the pump affinity laws 3 = pump casing
(eqs. (6-6a) and (6-6b)): prime 1 : actual or design
Operating Principles of the Centrifugal Pump

N1Q1 °3 N2Q2 °'s Impeller
(6-24a)
In its simplest form, the impeller of a centrif-

ugal pump can be regarded as a paddle wheel
(6-24b) with radial vanes, rotating in an enclosure, with
the fluid being admitted axially and ejected at
the periphery. This is shown schematically in

2_ OUTLET _qOTH _ _ ¢ u= 8ACXWA_O
Ve, NE$
figure 6-33. The tangential velocity component
of each fluid element increases as it moves out
radially between the vanes. Therefore, the cen-
trifugal force acting on these fluid elements
increases as the fluid moves out radially. As-
suming constant flow velocity in the radial direc-
tion and no energy losses, the ideal head rise A s_c_r io_ _,-A ,_
due to centrifugal force between the central

Figure 6-34.-Typical shrouded centrifugal im-
entrance (1) and the peripheral exit (2) is
peller with backward curved vanes.
2
Co) 2
AH ie :-_ (r 2 - r l 2) (6-25) (3) The flow is two dimensional (velocities
at similar points on the flow lines are
where uniform)
AHic =ideal head rise due to centrifugal forces, (4) The fluid leaves the impeller passages
ft tangentially to the vane surfaces (com-
= angular velocity of the wheel, rad/sec plete guidance of the fluid at the outlet)
rl =vane radius at the entrance, ft
The ideal inlet (point (1)) and outlet (point (2))
r_ = vane radius at the periphery, ft
flow velocity diagrams of the impeller described
g =gravitational constant, 32.2 ft/sec 2
in figure 6-34 are shown in figure 6-35. At this
For optimum performance, most impellers in
point with corresponding fluid velocities u, v,
high-speed centrifugal rocket engine pumps have
and c (as identified above), a is the angle be-
shrouded, backward curved vanes. The impeller
tween c and u, and fl is the angle enclosed by a
width is tapered toward the periphery to keep the
cross-sectional area of the radial flow path near
constant. A typical impeller design of this type
is shown in figure 6-34.
Velocity diagrams may be constructed to T v:
analyze the fluid flow vector correlations at
various points of an impeller. Let us assume
the following ideal conditions: co--Lb _1
(1) There are no losses, such as fluid-friction
u -_ ut _-',
losses
(2) The impeller passages are completely INLET VELOCITY DIAGRAM
filled with actively flowing fluid at all
times
.._WMEEL
.! !- Fos.
PAOIIL VINES EN CLOSUI_E
Cu=
OUTLET VELOCITY DIAGRAMS
Figure 6-35.-Flow velocity diagrams for the im-

peller shown in figure 6-34 (draw in a plane
Figure 6-33.-Paddle wheel (schematic). normal to the impeller axis).
tangent to the impellervane and a linein the Crn_ ="meridional" or (by definition for radial
directionof vane motion. The latteris equal to flow impellers) radial component of the
the angle between v and u (extended). absolute inlet flow velocity, ft/sec
Based on these velocitydiagrams,the follow- c2 = absolute outlet velocity of the flow,
ing correlationshave been established:' ft/sec
Cu2 =tangential component of the absolute
u22- u,2+ v,2 - v2 outlet flow velocity, ft/sec
AHip - 2g (6-26a) Cm 2 = meridional or radial component of the
absolute outlet flow velocity, ft/sec
l/2 2 -- U, 2 + V] 2 - V2 2 + C2 2 - C12
/9, =impeller inlet vane angle
AHi= /32 =impeller discharge vane angle
2g
=I(u2CU 2 - UlCu])(6-26b)
For pumping low-density propellants (such as
liquid hydrogen), which is associated with very
Qimp=448.8cm,A, =448.Scm2A2 (6-27) high developed heads, straight radial vanes are
frequently used in centrifugal impellers, since
cm2 they permit a higher obtainable head coefficient
Cu 2 = U2 - tan fl-----_ (6-28) _. Figure 6-36 presents a typical radial vane
impeller and its outlet velocity diagram. The
where
vane discharge _2 =90° and Cu2 = u2. The ideal
AHip=ideal static pressure head rise of the
developed head of a radial vane impeller becomes
fluid flowing through the impeller due
to centrifugal forces and to a decrease
U2 2 - tlICUl
of flow velocity relative to the impel- AHi = (6-29)
ler, ft g
_\H i =ideal total pressure head rise of the
fluid flowing through the impeller= the For centrifugal pumps of the noninducer type
ideal developed head of the pump (which are now rarely used in rocketry), proper
impeller, ft selection of the impeller inlet vane angle _ or
the provision of guide vanes at the inlet mini-
Qimp=impeller flow rate at the design point
(rated conditions), gpm mizes the absolute tangential component of fluid
flow at the inlet, cu_, which for best efficiency
A_ = area normal to the radial flow at the
impeller inlet, ft 2 should be zero. This is defined as no prerota-
As =area normal to the radial flow at the tion, where a_ =90 °. Thus, equation (6-26)
impeller outlet, ft 2 becomes
d, = vane diameter at the impeller inlet., in
U2Cu2
d_ =vane diameter at the impeller outlet _H, - (6-3o)
= outside diameter of the impeller, in g
u, = ,7 × rpm × d,,/720 = impeller peripheral
velocity at inlet, ft/see
u2 = r, × rpm × d2/720 = impeller peripheral
velocity at outlet = impeller tip speed,
ft/sec
v_ =inlet flow velocity relative to the impel-
ler, ft/sec c.2,v 2 _fi
v2 = outlet flow velocity relative to the im-

peller, ft/sec
c1 =absolute inlet velocity of the flow, OUTLET VELOCITy DI&V.*_M R_D_.L VANE _£LLER
ft/sec
Cul =tangential component of the absolute Figure 6-36,-Typical radial vane impeller and
inlet flow velocity, ft/see its outlet velocity diagram.
DESIGN OF TURBOPUMP PROPELLANT-FEED SYSTEMS '2O7
The above discussions assumed ideal condi- Cm 2
tions. For most rocket applications, centrifugal Cu2' = u2 - tan

2"----'_p
_ (6-33)
pumps are designed with an inducer upstream of
and in series with the impeller. The flow condi- By definition, the required impeller developed
tions at the impeller inlet thus are affected by head can be determined as
the inducer discharge flow pattern. In addition,
two types of flow usually take place simultane-
ously in the flow channels; namely, the main
AHim p =AH + He - hHind (6-34)
flow through the passages, and local circulatory
flows (eddy currents). The latter are relatively where
small but modify the former. The resultant effect
AH =rated design pump developed head, ft
at the impeller inlet is to make the flow enter at
AHin d =required inducer head-rise at the rated
an angle flz', larger than the impeller inlet vane design point, ft
angle f_L. The fluid is also caused to leave the He = hydraulic head losses in the volute, ft.
impeller at an angle/_2', less than the impeller Typical design values of He vary from
discharge vane angle f_:, and to increaae the 0.10 to 0.30 AH.
absolute angle a2 to a2'. This and the hydraulic The required impeller flow rate can be esti-
losses in the impeller correspondingly change mated as
the relative flow velocities v 1 and v 2 to vl' and
v2', the absolute flow velocities ci and c 2 to c,'
Qirap = Q + Qe (6-35)
and c:', and the absolute tangential components
Cul and cu2 to cu_' and Cur'. Since the radial where
flow areas A_ and A s, and the impeller flow rate
Qimp=required impeller flow rate at the rated
Qimp remain constant, the absolute radial or design point, gpm
meridional components cm_ and cm2 also remain Q = rated delivered pump flow rate, gpm
unchanged. The inlet and outlet flow velocity
Qe =impeller leakage losses, gpm. Most of
diagrams in figure 6-35 may now be redrawn as these occur at the clearance between
represented by the dotted lines.
impeller wearing rings and casing.
The correlation established in equation Typical design values of Qe vary from
(6-26b) may be rewritten as
1 to 5 percent of Qimp.
U2Cu2 s - U 1 Cul t
AHim p = g (6-31) Centrifugal Impeller Design Elements
After general pump design parameters, such

where as developed head Afl, capacity Q, suction spe-
3,Himp = impeller actual developed head, ft cific speed Nss, rotating speed N, and specific
Cut' --tangential component of the design speed Ns have been set forth or chosen, the de-
absolute inlet flow velocity, ft/sec sign of a centrifugal (radial) pump impeller may
Cu2' = tangential component of the design be accomplished in two basic steps. The first
absolute outlet flow velocity, ft/sec is the selection of those velocities and vane
The ratio of the design flow velocity Cu2' to angles which are needed to obtain the desired
the ideal flow velocity Cu2 can be expressed as characteristics with optimum efficiency. Usually
this can be achieved with the help of available
C/j2 _ design or experimental data such as pump head
e v = -- (6- 32 )
CU2 coefficient _,, impeller vane coefficient ev, and
leakage loss rate Qe. The second step is the
where ev=impeller vane coefficient. Typical design layout of the impeller for the selected
design values range from 0.65 to angles and areas. Considerable experience and
0.75. skill are required from the designer to work out
Referring to figure 6-35, equation (6-28) may graphically the best-performing configuration
be rewritten as based on the given design inputs.
The following are considered minimum basic 28 °, with an average value of 22.5 c for most
design elements required for proper layout of a specific speeds.
radial-flow impeller: Figure 6-37 presents the basic layout of a
1. Radial velocity at the impeller entrance or typical radial-flow impeller with backward curved
eye, Crn] ,-This is a function of inlet conditions vanes. The shaft diameter d s may be determined
such as inducer discharge velocity and inlet by the following correlations
duct size. For best performance, the value of
cml should be kept reasonably low. Typical 16 T
Ss = _ (6-37)
design values of Cml range from 10 to 60 ft/sec. rrds3
2. Radial velocity at the impeller discharge,
Cm 2.-Its value is a function of the impeller 32M
St- (6-38)
peripheral velocity u 2 and the flow coefficient (5. rrds 3
Typical design values for Crn2 range from 0.01 to
0.15 u 2. Ssw= ½ \/4Ss 2 +St 2 (6-39)
3. Diameter of the impeller at the vane en-
trance, d 1.-Its value is determined by the in- Stw= { St+ _ \_q Ss2 + St 2 (6-40)
ducer design as well as by impeller shaft and
hub size. where
4. The impeller peripheral velocity at the ds =impeller shaft diameter, in

discharge, u2.-The value of u 2 can be calcu- T = shaft torque corresponding to yield or
lated by equation (6-4) for a given pump devel- ultimate loads as defined by equations
oped head H and a selected overall pump head (2-9) and (°-10), lb-in
coefficient, d,. The maximum design value of u 2 M =shaft bending moment corresponding to
is often limited by the material strength which yield or ultimate loads as defined by
thus determines the maximum developed head equations (2-9) and (2-10), lb-in
that can be obtained from a single-stage impel- Ss = shear stress due to torque, lb/in 2
ler. Typical design values of u 2 range from 200 S_ =tensile stress due to bending moment,
to 1500 ft/sec. With u 2 and N known, the impel- lb/in 2
ler discharge diameter d 2 (in) can be calculated Ssw = allowable working shear stress (yield or
readily. ultimate) of the shaft material, lb/in 2
5. The inlet vane angle _,.-The value of/3, Stw = allowable working tensile stress (yield or
is affected by the inlet flow conditions. Gener- ultimate) of the shaft material, lb/in 2
ally, /31 should be made equal or close to the Impeller hub diameter dh and eye diameter de
inlet flow angle /3]' which can be approximated may be equal to hub diameter and tip diameter of
by the inducer. The maximum tensile stress in-
duced in an impeller by the centrifugal forces
tan /3,'=Cml/(ulcu,') (6-36)
Typical design values for /3_ range from 8 ° to

30 ° . b I ---
6. The discharge vane angle/32.-In the

special case of radial vane impeller designs
i
/32 =90°. For backward curved vane impellers,

/32 is the most important single design element. _2
Usually the selection of/32 is the first step in

determining the other impeller design constants,
since most of them depend on /32. Pump effi-
ciency and head-capacity characteristics are
important considerations for the selection. For
a given u2, head and capacity increase with /32. Figure 6-37.-Basic layout o[ a typical radial-
Typical design values for /32 range from 17 to flow impeller with backward curved vanes.
1 ,._
DESIGN OF TURBOPUMP PROPELLANT-FEED SYSTEMS 2O9
occurs as the tangential stress at the edge of the Qtmp= impeller flow rate at the rated design
shaft hole. It may be checked by point, gpm
After the vane angles and other dimensions at
inlet and discharge have been established, no
set rule is available for designing the backward
S'max= PU22rnax(3+F)
576g I1 +3-_:-:]
1-F[ds_7 JKs (6-41)' curved vanes. However, the number of vanes is
usually between 5 and 12, and may be determined
where
empirically by
Stmax =maximum tensile stress, lb/in 2 (should
be less than the allowable working
z =-- (6-44)
tensile stress of the impeller material) 3
p =density of the impeller material, lb/ft 3
where
/_ = Poisson's ratio of the impeller material
/32 =discharge vane angle
ds =impeller shaft hole diameter, in
z =number of vanes
d: = impeller outside diameter, in
If there is a space limitation at the impeller
u2 max = maximum allowable peripheral impeller
entrance, every other vane may be made a partial
speed, ft/sec = 1.25 x design value of
vane, starting at a larger radius. The contour of
u 2, for most rocket engine applica-
the vanes is designed to afford a gradual change
tions
of flow cross-sectional area (total divergence of
10 ° to 14°), at reasonably short flow passage
Ks =design factor, determined experimen-
length. The flow passage shape should be as
tally. Typical values vary from 0.4
close to a square as possible. The vanes should
to 1.0, depending on impeller shape.
be as thin as material strength and manufactur-
The surface finish and contour of the impeller
ing processes will permit. They may be of con-
shaft hole should be free of stress concentra-
stant thickness; i.e., a contour similar for both
tions. First-class splines are preferred rather
sides may be used. This allows a thinner edge
than ordinary keyways.
(typical value: 0.12 inch) at the inlet and re-
The width of the impeller can be calculated
suits in better efficiency if the angle fit has the
by the following correlations:
correct value. The impeller is usually a high-
quality aluminum-alloy casting, the vanes being
Qimp (6-42) integral with the shrouds. In some high-speed
bt =3.12 rrdlcmlq
applications, forged aluminum alloys or titanium
alloys are used. A typical aluminum forging, the
Qirnp (6-43) 7075 alloy with a T73 heat treat, has a yield
b2 -3.12 rrd2crn_e2 strength of 63 000 psi and an ultimate strength of
74 000 psi. In this case a two-piece construc-
where tion might be preferred to facilitate machining
bl =impeller width at the vane inlet, in operations.
/92 =impeller width at the discharge, in
Mixed-flow-type vanes which extend into the
el =contraction factor at the entrance. It
impeller entrance or eye (shown in fig. 6-38a)are
considers effective flow area reduction
frequently used in radial-flow impellers or cen-
from vane thickness and other effects
trifugal pumps. This is done to match the impel-
such as local circulatory flows. Typi-
ler inlet flow path with the inducer discharge
cal design values range from 0.75 to
flow pattern and to provide more efficient turning
0.9.
of the flow.
42 =contraction factor at the discharge.
The mixed-flow-type impeller as shown in
Typical design values range from 0.85
figure 6-38b is also frequently used in a "cen-
to 0.95•
trifugal flow pump." The velocity correlations
and design constants of a mixed-flow impeller
are essentially the same as those of a radial-
According to S. Timoshenko. flow impeller. Mean effective impeller diameters
are used in the calculations for head rise, flow diagrams and the plane normal to the axis. The
velocities, etc. The), are presented in figure value of y varies along the flow passage. The
6-38 a and b as: layout of a mixed-flow impeller on the drawing
board is a rather complicated drafting problem.
This is due to the three-dimensional vane curva-
dl2 - (d102 + d112) (6-45)
2 ture and other complexities. The method of
"error triangles" suggested by Kaplan may be
used. Details of this method can be found in
(d202 +d2i 2)
d22 - 2 (6-46) standard pump reference books.
where
Design of Cavitating Inducers
d I =mean effective impeller diameter at the
inlet, in The cavitating inducer of a centrifugal pro-
dl o=outer vane diameter at the inlet, in pellant pump is a lightly loaded axial-flow im-
dli =inner vane diameter at the inlet, in peller operating in series with the main pump
d 2 =mean effective impeller diameter at the impeller as shown in figure 6-5. The term
discharge, in "cavitating" refers to the fact that the inducer is
d2o=outer vane diameter at the discharge, in capable of operating over a relatively broad
d2i =inner vane diameter at the discharge, in range of incipient cavitation prior to a notice-
Effective impeller widths at inlet and dis- able pump head dropoff. It produces from 5 to 20
charge, b, and b2, are also presented in figure percent of the total head rise of a pump. The
6-38 a and b. They are equal to the diameter of conditions of pump critical NPSH at the 2-percent
a circle which is tangent to the contours of both dropoff point may correspond to a 10- to 30-
front and back shrouds. 7 is the angle between percent inducer-developed-head reduction, de-
the meridional flow vectors (cm_ and cm2) and pending upon its match to the main pump impel-
the plane normal to the axis of rotation. It is ler. The required inducer head rise for a given
also the angle between the plane of the velocity design is expressed by the correlation
t /c_z
t fco2
T-
!
I
I dZo
d2i I
d2
t _
i
I
d R
dlo
dl
AXIS--
L__ dli
I L.
Figure 6-38.-(a) Radial-flow impeller with mixed-flow vanes at the impeller entrance;
(b) Mixed-flow impeller:
DESIGN
OFTURBOPUMP
PROPELLANT-FEED
SYSTEMS 211
hHind = (NPSH)imp - (NPSH)ind (6-47) Figure 6-39 presents the basic elements of a
or typical inducer design. The primary increase in
static pressure occurs at the leading (upstream)
o,5 l edge of the vane through free stream diffusion;

i.e., through a reduction of relative speed and by
operating at a small angle of attack between rel-
ative inlet flow and inducer inlet vane. The
- (6-48) cavitation performance of an inducer depends
strongly on the inlet flow coefficient qSin d (ratio
where of inlet axial flow velocity Cmo to inlet tip speed
AHind = required inducer head rise at the %0- To obtain high suction specific speeds
design point, ft (for highest pump speed N), the inducer must
(NPSH)ind = inducer critical NPSH = pump have a low flow coefficient. This results in
critical NPSH or (NPSH)c small angles (Or, Oh) between vanes and the
= Thoma parameter r × pump total plane normal to the axis of rotation. As a rule,
developed head H the inducer vane angle 0 varies radially accord-
(NPSH)im p = impeller critical NPSH ing to a constant
N =pump shaft speed, rpm (same for
inducer and impeller c=dtan0=dttan0t=dhtan0h (6-48a)
Q =rated pump flow rate, gpm
(Ns)in d :inducer specific speed It also often varies axially, according to the
(Nss)imp :impeller suction specific speed variation of the axial or meridional component
(Nss)ind :inducer suction specific speed cm of the absolute flow velocity. Inducer inlet
=pump suction specific speed flow coefficient Ctnd, inducer diameter ratio ra
Nss (ratio of hub diameter dh to tip diameter dr), and
AXIAL FLOW
L_ _-VANES
AXIS
ROTATION OF INLET
ATTACK FLOW --- Pi
l_ _ -,
DEVELOPED VIEW OF VANES AT THE TIP DIAMETER
OF VANES AT THE
HUB DIAMETER
Figure 6-39.-Elements of a typical inducer design (three-vanes; cylindrical hub and tip contour).
suction specific speed (Nss)ind are related, element used to describe the proport}ons of an
based on theoretical one-dimensional fluid cavi- inducer.
tation considerations, by the expression Frequently the design of high suction per-
formance inducers dictates a relatively large
(Nss)ind 8150(1 - 2 6 ind 2)075 inlet eye diameter, while the pump main impeller
(6-49)
(1 - rd2) O's 6ind inlet eye diameter must remain small for best
performance. This condition can be accommo-
dated by tapering the inducer tip to lead over
Figure 6-40 shows this relationship graph-
from one diameter to the other (as shown in fig.
ically. The actual performance of a typical
6-41). To minimize the tip taper, tapering of the
inducer is also shown for comparison.
inducer inlet hub diameter may be added to main-
The three vanes shown in figure 6-39 are
tain the desired inducer inlet flow area. In the
equally spaced at a tip distance Pi. This is
calculation of tapered inducers, mean values
defined as "pitch" and can be expressed as
may be used for dt and dh. For structural rea-
sons, the inducer vane elements sometimes are
ndt
p_ = -- (6-50) designed to cant forward instead of being normal
Z
to the axis of rotation. The angle between the
canted vane and the plane normal to the axis is
where
defined as the sweep angle.
Pi = pitch or vane spacing, in
Table 6-5 contains typical values for inducer
d e =inducer tip diameter, in
design parameters and variables. Figure 6-42
z =number of vanes
presents inducer inlet and outlet velocity dia-
The ratio of vane tip chord length C_ to vane
grams based on the mean effective diameters.
pitch Pi is an important design element. It is
For the design calculations of inducers, the
defined as "vane solidity at the tip" of an in-
following correlations may be used (figs. 6-39,
ducer. Vane solidity Sv is a descriptive term
6-41 and 6-42). For inducers with cylindrical
relating the vane area (actual or projected) to
hub and tip contour:
the area of the annuli normal to the axial flow.
It can be expressed as
dt2 + dh 2
Ci d°2 = d12 "= 2 (6-52)
Sv =_/ (6-51)
(assume c. o'= O)
The ratio of inducer length Li to inducer tip
diameter de, (Li,/d_) is another important design
Qind
CO'--Cmo:Cml : (6-53)
3.12× -_
4 (dr2 - dh _)
r.N nN
(6-54)
Uo=U_ =_20do = 7_d_
FLOw
- _,_- '__
i
Figure 6-40.-Relation between inducer inlet _L_
flow coefficient and inducer suction specific

speed. Figure 6-4I.-Taper contour inducer.
°: ¸
TABLE 6-5.-Cavitating Inducer Design Parameters and Variables
Parameter or variable Typical design values Design requirement
Specific speed, (Ns)ind .......................... 6000 to 12 000 Head-capacity characteristics

Suction specific speed, (Nsshn d .................. 20 000 to 50 000 Suction characteristics
Head coefficient. _ind ........................... 0.06 to 0.15 Head rise
Inlet flow coefficient, _lnd ....................... 0.06 to 0.20 Cavitation performance
Inlet vane angle, O .............................. 8 ° to 16 ° (measured from plane Flow coefficient, angle of attack
normal to axis)
Angle of attack, i ............................... 3 ° to 8 ° Performance, flow coefficient,
vane loading
Diameter ratio, rd ............................... 0.2 to 0.5 Performance, shaft critical
speed
Vane solidity, Sv ............................... 1.5 to 8.0 at the tip Desired flow area
Number of vanes, z .............................. 3 to 5 Desired solidity
Hub contour .................................... Cylindrical to 15 ° taper Compatibility with main impeller
and shaft geometry
Tip contour .................................... Cylindrical to 15 ° taper Compatibility with main impeller
and shaft geometry
Vane loading ................................... Leading edge loading, channel Performance
lead
Leading edge ................................... Swept forward, radial, swept back Vane stress, performance
(shown in fig. 6-39)
Sweep angle .................................... Normal to shaft to 15 ° forward Vane stress
Vane thickness ................................. 0.070 to 0.300 chord length Ci Vane stress
Tip clearance (between inducer outside diameter and 0.5 to 1 percent of inducer out- Shaft axial and radial deflec-
casing) side diameter tions
Length to tip diameter ratio (Li/dt) ................ 0.3 to 0.6 Head-capacity characteristics
dlt 2 +d,h 2
412 - 2 (6-56)
uo t r dot+d[t
dt - 2 (6-57)
INLET VELOCITY DIAGRAM
doh +d,h
dh = 2 (6-58)
, Vt _:
Cl ;, C'ml "_
Qind
Co' __-Cmo :
3.12 x4(dot2 -doh 2)
OUTLET VELOCITY DIAGRAM

(no prerotation) (6-59)
Figure 6-42.-Typical flow velocity diagrams of
an inducer based on the mean effective diam- Qind
eters. Cml : (6-60)
/7
3.12x_-(dlt 2-dlh 2)
(6-61)
For inducers with tapered hub and tip contour:
do 2-d°t 2 +do h2
2 (6-55) (6-62)
For all inducers: u1 = inducer peripheral veloci:y at mean

effective outlet diameter, ft/sec
Qind = Q ÷ Qee + ½Qe (6-63) ut = mean tip speed of the inducer, ft/sec
Uot =inducer inlet tip speed, ft/'sec
dh ult =inducer outlet tip speed, ft/sec
rd =_- (6-64) v o' = inlet velocity of the flow relative to
the inducer, ft/sec
v I' = outlet velocity of the flow relative to
7rN (6-65)
ut = _-_ dt the inducer, ft/sec
c o' = absolute inlet velocity of the flow,
nN ft/sec
Uot =-_-_dot (6-65a) Cuo' = tangential component of the absolute
flow velocity, ft/sec
llr 2 UlCII I' Cm o = meridional or axial component of the
_Hind = ¢ind .... (6-66)
g g absolute inlet flow velocity, ft/sec
cl =ideal absolute outlet flow velocity,
ft/sec
Hind UlCul
_Hindt =" - (6-67) c 1' =absolute outlet flow velocity, ft/sec
Tlind g
Cu ]' = tangential componen t of the absolute
outlet flow velocity, ft/sec
Cm o
_ind = -- (6-68) cm 1 = meridional component of the absolute
Uot
outlet flow velocity, ft/sec
rd = hub to tip diameter ratio
¢ind = inducer head coefficient (for range,
where see table 6-5)
Q = rated pump flow rate, gpm _ind =inducer inlet flow coefficient (for
Qind = required inducer flow rate at the rated range, see table 6-5)
design point, gpm The inducer is generally made from a high-
Ve = impeller leakage losses at the rated quality aluminum-alloy forging of single-piece
design point, gpm construction. For manufacture, special machines
Qee = inducer leakage loss rate through the and tooling are required for best results.
tip clearance, gpm. Typical design Experimental results have indicated that a
values vary from 2 to 6 percent of high-pressure fluid-injection system can be de-
Q signed to increase the suction performance of a
AHind =required inducer head rise at rated pump with inducer by imparting an inlet "pre-
conditions, ft whirl" to the fluid entering the inducer. Fluid
AHindt=ideal inducer head rise at rated condi- injection provides a tangential component c u o in
tions, ft the proper direction to the absolute fluid inlet
7lind = inducer efficiency velocity c o' and thereby lowers the fluid inlet
dt = inducer mean tip diameter, in velocity v o relative to the inducer. Jet momen-
dot = inducer tip diameter at the inlet, in tum and directed whirl in the direction of blade
d_t = inducer tip diameter at the outlet, in rotation combined should serve to reduce the
dh = inducer mean hub diameter, in tendency for the blade tips to cavitate as aresult
doh = inducer hub diameter at the inlet, in of high relative velocities and low static pres-
d_h = inducer hub diameter at the outlet, in sure. The suction specific speed of one typical
do =inducer mean effective diameter at the inducer design was increased from 34 000 to
inlet, in 44 000 by applying "prewhirl." Fluid injection
dl =inducer mean effective diameter at the is introduced tangentially (at a small angle with
outlet, in the plane normal to the axis of rotation) several
UO = inducer peripheral velocity at mean inches upstream of the inducer inlet. It is fed
effective inlet diameter, ft/sec from the pump outlet fluid pressure.
Sample Calculation (6-7) Substitute(NPSH)imp intoequation (6-47)to

obtain the requiredinducerhead rise
The following required design data and exper-
imental model test results are given for the oxi-
AHind = (NPSH)imp - (NPSH)iad
dizer pump of the A-1 stage engine:
Required pump developed head, AH :2930 ft = (NPSH)imp - (NPSH)c = 293 - 58--235 ft
Required pump flow rate, Q= 12420 gpm
Pump shaft speed, N = 7000 rpm
Pump specific speed, Ns = 1980 From equation (6-66),inducer mean tipspeed
Pump critical NPSH, (NPSH)c = 58 ft
Pump actual suction specific speed, Nss
ut= ][ 0.06 =355 ft/'sec
= 37 230 (from experimental tests)
Pump overall head coefficient, ¢ = 0.46
Basic inducer configuration (tip and hub taper Inducer mean tip diameter
contours) (similar to fig. 6-41)
d 720 720×355 ,.
Inducer head coefficient, ¢ind = 0.06
t:-_-ut--_ =ii.62 in
Inducer diameter ratio, rd = 0.3
Inducer ratio, Li/dt = 0.4
Angle of attack at inducer inlet tip, i = 4 ° (from eq. (6-65)).
maximum For the given (Li/dt) ratio of 0.4, the axial
Inducer tip contour taper half angle = 7 ° length of the inducer becomes
Inducer hub contour taper half angle = 14 °
Li=11.62×0.4=4.65 in
Inducer solidity based on mean tip diameter,
dt=2.2
Inducer leakage loss rate, Qee : 0.032 Q For a given tip contour taper half angle of 7 °,
the tip diameter at the inducer inlet
Basic impeller configuration = radial-flow type
with mixed flow vanes at the inlet(similar
to fig.6-38a) Li
dot=dt+2×-_×tan 7°
Impellersuctionspecificspeed, (Nss)imp
= 11 000
=11.62+4.65x0.1228= 12.19 in
Impeller discharge vane angle, /3 2 = 24 °
Impeller contraction factor at the entrance,
'rip diameter at the inducer outlet
_,=0.82
Impeller contraction factor at the discharge, Li
% = 0.88 dlt=dt-2x-ff×tan 7°=11.62-0.57=11.05 in
Impeller coefficient, ev = 0.74
Impeller leakage loss rate, Qe = 0.035 Q
Mean hub diameter
Pump volute head loss, He --0.19H
Design and calculate the basic parameters
dh = dtrd = 11.62 x 0.3 = 3.49 in
and dimensions of: (a._)pump inducer, (b_) pump
impeller. (from eq. (6-64)).
Solution For the given hub contour taper half angle of

14 ° , the hub diameter at the inducer inlet
(a__)Oxidizer pump inducer
Modifying equation (6-10), the impeller critical
NPSH can be calculated as doh=dh- 2X-_-tan14°=3.49-4.65×O.2493=2.a3 m
[ NQ°'S I"3"_ The hub diameter at the inducer outlet

(NPSH)imp = L(Ns s)impj
: [7000×(12420)°Q,3. = 293 ft
Li
L 11 ooo j d,h =dh+2x_-tan 14 ° =349+ 1.16=4.65 in
4
Substitute the given leakage loss rates Qee From equation (6-62), the inducer peripheral
= 0.032 for the inducer and ½ Qe = 0.0175 for the velocity at d 1
impeller into equation (6-63) to obtain the re-
quired inducer flow rate x 7000
ul- 720 ×8.45=258.5ft/sec
Qind=Q+Qee + ½Qe
From equation (6-66), the tangential compo-
= 12 420(1 + 0.032 + 0.0175) = 13 040 gpm
nent of the inducer absolute outlet velocity
From equation (6-59), the actual inducer

absolute inlet flow velocity c o' (= its meridional Cul' =AHind g -235x32"2-29.2 ft/sec
u1 258.5
component cm o, assuming its tangential compo-
nent Cuo =0; no' =0)
Refer to figure 6-42 for the flow velocity dia-
grams of the inducer, based on the mean effective
Qind
C O' :- C m 0- diameters d o and d I. Inducer design relative
3.12 ×4(d ot 2 - doh 2) inlet flow velocity
13040 v o' : V'Cm o2 + uo 2 = \/1384 + 71 825 = 270.6 ft/sec

= 37.2 ft/sec
3.12 ×_-× (148.6 - 5.43)

Inducer design relative inlet flow angle
From equation (6-60), the meridional compo-

sin/_o ,=cmo, =37'2 =0.135; f_o'=7 °45.
nent of the inducer absolute outlet flow velocity v o 270.6
Qind Inducer design absolute outlet flow velocity

Cm 1 -
3.12x4(d 1 $2_dlh2) c/= ,v/cu 1' :+ Cm 12 -- \/_5264 +2819.61 = 60.5 ft/sec
13 040
= - 53.1 ft/sec Inducer design absolute outlet flow angle
77
3.12 ×_× (122.1 - 21.6)

, cml 53.1
From equation (6-55), the inducer mean effec- tan al =--=_9--._
1.82;cu
1' = al' =61°13'
tive diameter at the inlet
Inducer design relative outlet flow velocity
2 _ 2 V1'= V'(ul - Cut')2 + Cml 2
= \'52 578 + 2820 = 235 ft/sec

From equation (6-61), the inducer peripheral
velocity at d o Inducer design relative outlet flow angle
× 7000 ^ ,,_
cml 53.1
u o :_ x 5. _o = 268 ft/sec
tan/31'-ul_cul,-229.3 -0.232; fi1'=13°3'
From equation (6-56), the inducer mean effec-

Since the cavitation performance of an inducer
tive diameter at the outlet
depends largely on the angle of attack of the
vane leading edge at the inducer inlet tip, and on
_fd,t2 + dlh 2 _122.1 +21.6 the inducer inlet flow coefficient ¢ind, we now
d:= 2 - 2
proceed to determine the vane angle 0or at the
inducer inlet tip, and to check ¢ind with the help
= _'1.85=845 in of equation (6-49).
From equation (6-65a) the inducer inlet tip Our inducer exhibits characteristics similar
speed to those shown in figure 6-40 for a typical in-

ducer. If we use a vane angle 0, of 14°30 ' at
7rxN . 7rx 7000 the inducer outlet mean effective diameter d,.
u°t=-7-_ °°t: 72----O--× 12.19=372.5 ft/sec the difference between 0_ and the relative outlet
flow angle /3 l'
The relative flow angle at the inducer inlet 0_ - t3,' = (14o30 ') - (13°3 ') = 1_27 '
tip
This allows for the effect of local circulatory
flow (boundary effects).
tanflo_ '=cm°- 37'2 :0.0998; riot':5°42 ' The vane angle Ù_t at the inducer outlet tip
not 372.5
diameter d _t
If we use a vane angle Oot = 9 ° at the inducer

inlet tip. the angle of attack at the inlet tip d, 8.45
tan O_t:--d, t tan 01 = _--_--_ tan (14°30')= 0.198;
0 ot-t3ot I :9°-(5°42 I)_7__ 3°18 ' (<4 °, as desired)

01t=11°12 '
The vane angle 0o at the inducer inlet mean

The vane angle OLh at the inducer outlet hub
effective diameter do
diameter d_ b
dl 845
tan 0o: ___ooot
tan 0ot=_xtan
12.19 . 9 ° :0.220; --- tan 0_ tan 14o30 , =0.471;
tan 01h-d1 h =4..6-5
0o : 12025 '
01h=25°13 '
(see eq. (6-48a)).
We will use three vanes (z : 3). The vane
The vane angle 0oh at the inducer inlet hub
pitch at the mean tip diameter dt can be calcu-
diameter doh
lated from equation (6-50)
tan _ dot 0 12.19 90

Uoh = do---_tan ot : _-3-_ × tan : 0.829: pi:n'dt_rrx11.62 12.18 in
z 3
0oh = 39040 `
The chord length at vane tip can be calcu-
From equation (6-68), the inducer inlet flow lated as
coefficient
Li 4.65 4.65
: 26.57 in
Cmo 37.2 Ci= -. {Oot+O_t_=sin 10o6,=0.175
¢ind =
u0t
- 372.5- 0.0998 s n\- )
Substitute this into equation (6-49), to obtain From equation (6-51) the inducer solidity
the theoretical inducer suction specific speed based on the mean tip diameter dt
8150(1_2 _
(Pind) 2_o._
S _ Ci_26.57 2
(Nss)ind =
¢ind (1 - rd2) 0"s 18
_ 8150 × (0.9601) °_s x (0.91) 05

0.0998 A-1 Stage Engine Oxidizer Pump Inducer Design
Summary
_ 8150 × 0.97 x 0.954 Following completion of calculations it is
=75 700>Nss =37 300
0.098 advisable to compile the results systematically
in a summary, prior to start of layouts. This Substitute the given hydraulic head losses
gives an opportunity for cross checks and re- He = 0.19 AH into equation (6-34), to obtain the
duces the probability of errors. required impeller developed head
Required head rise and capacity, AHind
= 235 It, Qind = 13 040 gpm AHimp = AH + He - AHind
Inlet velocity diagram (at inlet mean effective = 2930 (1 + 0.19) - 235 = 3252 ft
diameter do)
ao' = 90 °, _o' = 7°45' From equation (6-35), the required impeller
u o = 268 ft/sec, v o' = 270.6 flow rate
Co' = cm 0 = 37,2 ft/sec, Cuo = 0
Outlet velocity diagram (at outlet mean effec- Qimp = 12 420 (1 + 0. 035)= 12 855 gpm
tive diameter all)
a1'=61°13 ', fl1'=13°3'
u_ =258.5 ft/sec, v_' =235 ft/sec From equation (6-31), the tangential compo-
c_' =60.5 ft/sec, Cul' =29.2 ft/sec nent of the impeller design absolute outlet flow
cm I = 53.1 ft/sec velocity
Axial length of inducer, Li =4.65 in
Taper half angle at tip: 7°; at hub: 14 ° Cu2, =g-_Himp + UlCul'
Inlet dimensions u2
dot:12.19 in, doh =2.33 in, do :8.76 in
vane angle at dot, 0or=9 ° = 32.2 x 3252 + 258.5 × 29.2 = 248 ft/sec
453
vane angle at doh, 0oh =39°40 '
vane angle at d o, 0o = 12025 .
Outlet dimensions
nent of the impeller ideal absolute outlet flow
dlt=11.05 in, dlh =4.65 in, dl =8.45 in velocity
vane angle at d_t, 01t=11°12 '
vane angle at d,h, O_h =25°13 '
cu 2' 248
vane angle at d_, 0_ = 14030 ' -- = -- = 335 ft/sec
cu2 = ev 0.74
Number of vanes, z = 3
Solidity at vane tip, Sv--2.18
Referring to figures 6-34 and 6-35, and to
Inlet flow coefficient, gSin d = 0.0998
equation (6-28), the meridional component of the
(b) Oxidizer pump impeller
impeller design absolute outlet flow velocity
We will use a radial-flow-type impeller with
mixed-flow-type vanes extending into the impeller
era2 :(%-c.2) tan 132
entrance eye, as shown in figure 6-38a. The
flow path and velocity conditions at the impeller = (453-335) tan 24°= 52.5 ft/sec
inlet can be assumed to be the same as those at
the inducer outlet. The impeller design absolute outlet flow
From equation (6-4), the tip or peripheral velocity
speed at the impeller discharge
C2' = V/Cu2' 2 + Cm22 _-_
\/2482 + 52.52 = 253.4 ft/sec
Impeller design absolute outlet flow angle:

453 ft/sec
u2 = _f_ = × 2930 =
32.20.46
, cm2 52.5
tan a2 - c. 2' - 248 - 0.212; 11°58 '
a2' =
The impeller outlet diameter

Impeller design relative outlet flow velocity
v 2' = V'(U2 - cu2') 2 + cm2 2

720 × u 2 _ 720 x 453
d2 = ---14.8 in = \/2052 + 52.52 -- 211.6 ft/sec
_r× N lr x 7000
Impeller design relative outlet flow angle: Outlet dimensions
outside diameter d 2 = 14.8 in

discharge vane angle fi2 = 24°
Cm2 , 52.5
tan f12' = (u2---c-u2) =-_--_= 0.256; /32' = 14°22 ' impeller outlet width b 2 = 1.91 in
Number of impeller vanes (eq. 6-44) z i =/32/3
Referring to figure 6-38a, and to equation = 24/3 =8
(6-42), the width of the impeller at the vane inlet
Qimp
Design of Casings
bl - 3.12 _rdlCmle 1
The main function of a pump casing is to con-

12 855
=3.12 x 7rx 8.45 x 53.1 x 0.82 =3.56 in vert the kinetic energy of high flow velocity at
the impeller discharge into pressure. It does not
contribute to the generation of head. The con-
From equation (6-43), the width of the impel-
struction of a typical centrifugal pump casing is
ler at the discharge
shown in figure 6-5. The front section of the
casing, which provides the pump inlet and houses
Qimp
the inducer, is called the suction nozzle. The
b2 = 3.12 _d 2Cm2e-2
rear section of the casing, which collects the
12 855 fluid from the impeller and converts the velocity
-3.12x,x14.8x52.5x0.88=l.91in head into pressure prior to discharge, is called
the volute.
From equation (6-5) the overall pump flow Since the flow path in a suction nozzle is
coefficient short and the flow velocities are relatively low,
the head loss in a suction nozzle due to friction
is very small. The contour of the suction nozzle
cm2 52.5=0.116
= _ - 453 is designed to suit the inducer configuration. A
tapered suction nozzle (as shown in fig. 6-14),
also known as an end suction nozzle, together
A-1 Stage Engine Oxidizer Pump Impeller Design with a tapered inducer, yields best results in
Summary
most respects. This nozzle, the area of which
Required impeller developed head aud _Him p gradually decreases toward the impeller eye,
= 3252 ft, Qimp = 12 855 gpm flow capacity greatly steadies the flow and assures uniform
feed to the impeller. In liquid oxygen pumps,
Inlet velocity diagram (at mean inlet effective frequently a liner made of a material such as
diameter d _) Kel-F is inserted between inducer and suction
al' = 61°13', /9,'=13°3 ' nozzle wall. This eliminates the possibility of
u, =258.5 ft/sec, Vl' =235 ft/sec metal-to-metal rubbing in the presence of narrow
c,' =60.5 ft/sec, cu,' =29.2 ft/sec, inducer tip clearances. Rubbing in liquid oxygen
Cm, =53.1 ft/sec pumps may cause dangerous explosions. In
Outlet velocity diagram (at outlet diameter d2) turbopumps of the single-shaft type (fig. 6-18),
a2' = 11°58 ', fi2' = 14°22' the fuel is introduced to the fuel pump in a radial
u 2 =453 ft/sec, v2' =211.6 ft/sec direction. Special guide devices are required in
c 2' =253.4 ft/sec, cu2' =248 ft/sec, the inlet to minimize pressure drops because of
Cm 2 = 52.5 ft/sec the need of turning the flow axially into the
Inlet dimensions inducer.
inlet eye diameter= dl t = 11.05 in Two types of volute casing are used in rocket
inlet hub diameter = d lh = 4.65 in centrifugal pumps, the plain volute and the dif-
inlet mean effective diameter = d, = 8.45 in fusing vane volute (see fig. 6-43). In the first,
inlet vane angle at diameter d, = fl, = 13°3 ' the impeller discharges into a single volute
inlet vane width b, =3.56 in channel of gradually increasing area. Here, the
_lrruss_ / \ I40ZZLE IMPELLER

! STATIONA_'Y -- _k4 PEU-E_
pLA,IN VOLUTE _}MP I_IF'FUS_NG V_.htE V_)L_TE PUM_ =
Figure 6-43.-Plain vo]ute and diffusing vane

volute centrifugal pump casings.
major part of the conversion of velocity to pres-

sure takes place in the conical pump discharge
nozzle. In the latter, the impeller first dis-
charges into a diffuser provided with vanes. A
major portion of the conversion takes place in
the channels between the diffusing vanes before
_ -. e=
the fluid reaches the volute channel. The main
advantage of the plain volute is its simplicity.
However, the diffusing volute is more efficient.
Head losses in pump volutes are relatively high.
Approximately 70 to 90 percent of the flow kinetic
Figure 6-44.-Plain volute casing of a centrHugal
energy is converted into pressure head in either
volute type. pump.
The hydraulic characteristics of a plain volute

The design value of the average volute flow
are determined by several design parameters
velocity c_' may be determined experimentally
which include: volute throat area av and flow
from the correlation
areas a 0. included angle Os between volute side
walls (fig. 6-44), volute tongue angle av, radius
c3' = Kv¥'_ (6-70)
rt at which the volute tongue starts, and volute
width b 3. Their design values are somewhat where
influenced by the pump specific speed Ns and Kv = experimental design factor; typical values
are established experimentally for best per- range from 0.15 to 0.55. Kv is lower for
formance. higher specific speed pumps
All of the pump flow Q passes through the AH = rated design pump developed head, ft
volute throat section av, but only part of it g :gravitational constant, 32.2 ft/sec 2
passes through any other section, the amount In order to avoid impact shocks and separa-
depending on the location away from the volute tion losses at the volute tongue, the volute angle
tongue. One design approach is to keep a con- av is designed to correspond to the direction of
stant average flow velocity c 3' at all sections of the absolute velocity vector at the impeller dis-
the volute. Thus charge: av"_a=', ttigher specific speed pumps
have higher values of %' and thus require higher
, Q 1 0 Q av. The radius rt at which the tongue starts
(6-69)
c3 =3.12av 3.12 360 a0 should be 5 to 10 percent larger than the outside
radius of the impeller to suppress turbulence and
where to provide an opportunity for the flow leaving the
c 3' = average flow velocity in the volute, ft/sec impeller to equalize before coming into contact
Q =rated design pump flow rate, gpm with the tongue.
av =area of the volute throat section, in 2 The dimension b 3 at the bottom of a trap-
a 0 =area of a volute section (in2), at an angular ezoidal volute cross section is chosen to mini-
location 0 (degrees) from the tongue mize losses due to friction between impeller
b,,.
DESIGN OF TURBOPU/_P PROPELLANT-FEED SYSTEMS 221
discharge flow and volute side walls. For small itsinletcan be approximated in a manner similar
pumps of lower specific speeds, b 3 =2.0b 2, to thatused forthe width of a plain volute (i.e.,
where b 2 is the impeller width at the discharge, 1.6 to 2.0 impellerwidth ba). A typicaldiffuser
in. For higher specific speed pumps, b 3 = 1.6 to layout is shown in figure6-46. The vane inlet
1.75 b 2. Tile maximum included angle Os be- angle a 3 should be made equal or close to the
tween the volute side walls should be about 60 ° . absolute impellerdischarge flow angle a2'. The
For higher specific speed pumps, or for higher design value of the average flow velocityat the
impeller discharge flow angles aa', the value of diffuser throat c s' may be approximated by
9s should be made smaller.
The pressure in the volute cannot always be
c 3' =_c2' (6-71)
kept uniform, especially under off-design oper-
ating conditions. This results in a radial thrust
on the impeller shaft. To eliminate or reduce where
the radial thrust, double-volute casings have C3'--average flow velocity at the diffuser throat,
been frequently used (fig. 6-45). Here, the flow ft/see
is divided into two equal streams by two tongues d 2 =impeller discharge diameter, in
set 180 ° apart. Although the volute pressure d_ = pitch diameter of the diffuser throats, in
unbalances may be the same as in a single- Ca'=absolute flow velocity at impeller dis-
volute casing, the resultant of all radial forces charge, ft/sec
may be reduced to a reasonably low value, owing Since each vane passage is assumed to carry
to symmetry. an equal fraction of the total flow Q, the follow-
The diffusing vane volute has essentially the ing correlation may be established:
same shape as a plain volute, except that a num-
ber of passages are used rather than one. This z - Q (6-72)
b3h3 -_
permits the conversion of kinetic energy to pres-
sure in a much smaller space. The radial clear-
where
ance between impeller and diffuser inlet vane
tips should be narrow for best efficiency. Typical b 3 = width of the diffuser at the throat, in
values range from 0.03 to 0.12 inch, depending h 3 =diffuser throat height, in
z =number of diffuser vanes
upon impeller size. The width of the diffuser at
Q =rated design pump flow rate, gpm
The number of diffuser vanes z should be
minimum, consistent with good performance, and
should have no common factor with the number of
impeller vanes to avoid resonances. If possible,
// f _ "_X U_PELLER
VX_l[S
,/
vot_ur¢
_$b,:.E
rid
Figure 6-45.-Typical single discharge, 180 °

opposed double-volute casing o[ a centrifugal Figure 6-46.-Typical layout o[ the diHuser [or a
pump. centrifugal pump volute casing.
the cross section of the passages in the diffuser any section from 0 ° to 180 ° away from the volute
are made nearly square; i.e., b 3 :h 3. The shape tongue may be calculated for both volutes as
of the passage below the throat should be diverg-
ing, with an angle between 10 ° to 12 °. The 0Q 12420
0=0.0760
velocity of the flow leaving the diffuser is kept a0 = 3,12 × 360 x c 3' : 3.12 × 360 x 146
slightly higher than the velocity in the pump
discharge line. At 0=45 °, a4s =3.42 in2; at 0=90 °, ago =6,84in2;
Rocket pump casings are frequently made of at 0=135 °, a13s=10.26 in2; and at 0:180 °,
high-quality aluminum-alloy castings. In low- also = 13.68 in 2.
pressure pumps, the casing wall thickness is Total volute throat area at the entrance to the
held as thin as is consistent with good foundry discharge nozzle
practice. Owing to the intricate shape of the
castings, stress calculations are usually based av : 2 × 13.68 : 27.36 in 2
upon prior experience and test data. For a rough
check, the hoop stress at a casing section may The volute angle av can be approximated as
be estimated as
a av-=a 2' = 11°58 ', say 12 °
St = p_- (6-73)
The radius rL at which the volute tongues
where start can be approximated as (assuming 5 percent
St --hoop tensile stress, lb/in 2 clearance)
p =local casing internal pressure, psia (or
pressure difference across the casing d2 _. 14.8
wall, psi) r_ =-_ × 1. uv --T × 1.05 --7.77 in
a : projected area on which the pressure acts,
in 2 The width at the bottom of the trapezoidal
volute section shall be
a'= area of casing material resisting the force
pa, in 2
b 3 =1.75b 2 = 1.75x 1.91 in=3.34 in
The actual stress will be higher, because of
bending stresses as a result of discontinuities
Allowing for a transition from the shape of the
and deformation of the walls, and thermal stresses
volute to round, we use a diameter of 6.25inches,
from temperature gradients across the wall.
or an area of 30.68 in 2, for the entrance to the
discharge nozzle. With a 10 ° included taper
angle and a nozzle length of 10 inches, the exit
The flow conditions at the outlet of the A-1 diameter of the discharge nozzle can be deter-
stage engine oxidizer pump impeller were de- mined as
rived in sample calculation (6-7). Calculate and
design a double-volute (spaced 180°), single- de:6.25+2×lOxtan 5°
discharge-type casing (as shown in fig. 6-45) for : 6.25 + 2 x 10 x 0.0875 : 6.25 + 1.75
the same pump, assuming a design factor Kv of
0.337. = 8 in (or an area of 50.26 in 2)
Flow velocity at the nozzle inlet:

Solution
From equation (6-70), the average volute flow 12420

:130 R/sec
velocity may be calculated as 3.12×30.68
c3' = Kv _/2 g_/-/= 0.337 x \/2 x 32.2 × 2930 =146 ft/sec Flow velocity at the nozzle exit:
Referring to figures 6-44 and 6-45, and from 12 420 = 79.4 ft/sec
equation (6-69), the required volute flow area at 3.12 x 50.26
DESIGNOF TURBOPUMP PROPELLANT-FEED SYSTEMS 223
Balancing the Axial Thrust of Centrifugal Pumps the turbine; lb. A positive sign indi-
cates a force which tends to pull tile
Unbalanced axial loads acting on the inducer-
impeller away from the suction side, a
impeller assembly of centrifugal pumps are pri-
negative sign indicates the opposite.
marily the result of changes in axial momentum,
The static pressure at the inducer outlet, p;,
and of variations in pressure distribution at the
can be either measured in actual tests, or approx-
periphery of the assembly. These unbalanced
imated by
forces can be reduced by mounting two propellant
pumps back to back, as shown in figures 6-14
Pl = kiPo (6-75)
and 6-18. More subtle balancing of the axial
loads can be accomplished by judicious design where
detail, which is especially important in high-
ki :design factor based on experimental data
pressure and high-speed pump applications.
(ranging from 1.1 to 1.8)
Either one of the following two methods is fre-
Po = static pressure at the inducer inlet, psia
quently used.
The average pressure in the space between
With the first method (as shown in fig. 6-47),
impeller shrouds and casing side walls, Pv, may
a balance chamber is provided at the back shroud
be approximated by
of the impeller, between back wearing ring diam-
eter dbr and shaft seal diameter ds. Balancing
3 u2 2 - Ul 2
of axial loads is effected by proper selection of
Pv:P, +57"----6 2g 'P (6-76)
the projected chamber area and of the admitted
fluid pressure. The pressure level Pc in a bal-
where
ance chamber can be controlled by careful ad-
u 2 : peripheral speed at the impeller outside
justment of the clearances and leakages of the
diameter d 2, ft/sec
back wearing ring and the shaft seals. The
u 1 ---peripheral speed at the impeller inlet mean
required Pc may be determined by the following
correlation: effective diameter d,, ft/sec
p =density of the pumped medium, lb/ft 3
pcrr (dbr 2 - d s 2) : pvrr (dbr 2 - d fr 2)

The main advantage of the balance chamber
+ plrr(dfr 2 -dt_)+ porrdh _ + 4fvicm° + Te method is flexibility. The final balancing of the
(6-74)
g turbopump bearing axial loads can be accom-
plished in component tests by changing the value
where of Pc through adjustment of the clearances at the
wearing ring and shaft seals. However, this
Pc :balance chamber pressure, psia
Pv :average net pressure in the space be- tends to increase leakage losses.
In the second method (as shown in fig. 6-48),
tween impeller shrouds and casing
straight radial ribs are provided at the back
walls, psia
shroud of the impeller to reduce the static pres-
p_ : static pressure at the inducer outlet, psia
sure between the impeller back shroud and cas-
Po = static pressure at the inducer inlet, psia
ing wall through partial conversion into kinetic
ds :effective shaft seal diameter, in
dh =hub diameter at the inducer inlet, in energy. This reduction of axial forces acting on
the back shroud of the impeller may be approxi-
d t =inducer tip diameter =eye diameter at the
mated by the following correlation:
impeller inlet, in
dtr = front wearing ring diameter, in
dbr = back wearing ring diameter, in _ a,_ " =-d =_(u_-2-us2) (s+t)
_ei :inducer weight flow rate, lb/sec Fa -4--_ (at s _ 2g p_ (6-77)
cm o =axial flow velocity at the inducer inlet,
ft/sec (converts to radial) where
g =gravitational constant, 32.2 ft/sec 2 Fa :reduction of the axial forces acting on the
Te = external axial thrust due to unbalanced back shroud of impeller, lb
axial loads of the other propellant and/or dr =outside diameter of the radial ribs, in
,4.--- pv
EXTERNAL
,,L.:E l
CIt&MBER
AXIAL
THRUST T e
It -'----r dE',
AXIAL POSITIVE----_
MOMENTUM ---=" dfr d_ dh NEGATIVE.,4.---
CHANGE dt
w C
i mo $H_F'T /
Figure 6-47.-Balancing axial thrusts of a centrifugal pump by the balance chamber method.
'i ......
Pl \
EXTERNAL
AXIAL
THRUST T a
AXIAL
MOMENTUM = SEA,
CHANGE ----- POSITIVE ---"-
d_, d_ , ._._.___
NEGATIVE-.,"---
Wc
mo
3i i "'" RADIAL ,
!
- _ RI/ RIBS __]L
WEARING
_
RING
Figure G-48.-Balancing axial thrusts of a centrifugal pump by the radial rib metl_od.
ds = effective shaft seal diameterTinside diam- pvr;(dtr2 -ds 2) -4 F a =p, r,(dtr2 -dr 2)
eter of the radial ribs, in
ur =peripheral speed at diameter dr, ft/sec + po.rdh2+4 fVCmo + Te (6-78)
Us =peripheral speed at diameter ds, ft/sec g
g :gravitational constant, 32.2 ft/sec 2

p :density of the pumped medium, lh/ft s The pressures p, and Pv may be approximated
t =height or thickness of the radial ribs, in by equations (6-75) and (6-76). See equation
(6-74) for other terms.
s :average distance between casing wall and
impeller back shroud, in Sample Calculation (6-9)
The required Fa may be determined by the tZadial ribs (similar to those in fig. 6-48) are
following correlation: used on the back shroud of the A-1 stage engine
_'.¢
oxidizer pump impeller, with the following dimen- axial-flow pump is selected. For each stage of
sions: an axial-flow pump, head rises of 5000 to 9000
Outside diameter of the radial ribs, dr= 14.8 feet can be obtained.
in (equal to d 2) The capacity of an axial-flow hydrogen pump
Inside diameter of the radial ribs, ds = 4.8 in is usually limited to about 5500 gpm as a mini-
Height of the radial ribs, t=0.21 in mum. This is due to the minimum practical
Width of the radial ribs, w = 0.25 in (not height h v of the vanes (fig. 6-49). For heights
critical) below 0.5 inch, the tip clearance required for
Average distance between the casing wall efficient performance becomes critical, causing
and impeller back shroud, s = 0.25 in manufacturing problems. A reduction in rotor
Estimate the reduction of the axial forces diameter below certain values is not practical
acting on the back shroud of the impeller, due to either, because of the high rpm required for
the radial ribs. proper blade speed.
Figure 6-50 presents typical operating regions
Solution of various liquid hydrogen pump types, These
include centrifugal pumps of (Ns)l = 500 per
The peripheral speed at diameter dr
_rN . tr×7000
ur:'7ff'd°r= 7"2-0 ×14.8=452 ft/sec STATOR ROTOR
VANES --_ _ VANES

The peripheral speed at the diameter ds \ /TiP
=N = x 7000
× 4.8= 147 ft/sec
Us=_ds= 720
From equation (6-77), the reduction of the

axial forces
F 3zr rolaaA oa aAx (204 304 - 21609)

"= 4-g-6g_...... - ..... ' _ _-5g_.2 PU_P AXIS " 1
x 71.38 × (0.25 +0.21)_ 74 680 lb OF ROTATION

2x0.25
6.4 DESIGN OF AXIAL-FLOW PUMPS
Except when used as inducers, application

of axial-flow pumps in rocket engines is essen-
tially limited to liquid hydrogen systems in a
multistage configuration. Thus, the following I.U
k3
Z
discussions are applicable to axial-flow hydro-
gen pumps only. Multistage axial-flow hydrogen n-
O
I_. _!:¸¸_
pumps are applied in regions which are beyond e_
i,i
the capability of a single-stage centrifugal pump,
O_
since their construction is comparatively simple =E
(fig. 6-4). As can be seen in figure 6-6, the fluid

in an axial-flow pump flows from one stage to
.S
the next with a minimum of connecting passages. VANE HEIGHT hv, INCH
The head rise of a typical single-stage cen-
trifugal hydrogen pump is limited to about 65 000 Figure 6-49.-E[fect of vane height on the per-
ft (2000 psi). Beyond this point, a multistage formance of an axial-flow pump.
I C[_'rmFU_,L I Operation of the Impeller Rotor

! t_,,,, I
The main function of the impeller rotor of an
c E_TR_FUGAL ;
_,u=l,
axial-flow pump is to impart kinetic energy to
zo_
_ TIP U'!IED
the fluid by increasing the tangential component
i u_ • 16_r_¢'f_C of the absolute flow velocity, This is accom-
plished by the action of airfoil-shaped rotor
vanes (figs. 6-51 and 6-52). It is convenient to
,i
] AXlIL FLOe PUlP i
describe the vanes on several developed cylin-
drical sections. Three sections are of particular
..... I, , I,J
interest: at the impeller tip diameter dr, at the
impeller hub dh, and at its mean effective diam-
Figure 6-50.-Typical operating regions of vari- eter dm (inches). The mean effective diameter
ous pump types for liquid-hydrogen-fueled is defined by
rocket engine applications.
dt2+dh 2 dt2(l+rd 2)
stage (1 and 2 stages); centrifugal pumps of - (6-79)
dm2 - 2 2
(Ns) 1 = 1000 per stage (1 to 6 stages; not recom-
mended for rocket engine use); and axial-flow
pumps of (Ns)l -=3000 per stage (1 to 12 stages).
where rd = impeller hub ratio or dh/dt.
For any given operating region, there is usually
For simplicity, vane characteristics and flow
a best-suited design configuration. However,
conditions are discussed here only with respect
overlapping regions occur which could be ful-
to the mean effective diameter din. The vanes
filled by either a multistage axial-flow pump or
are equally spaced at a circumferential distance
by a single-stage centrifugal pump. The best
Pr
solution then is dictated by other considerations,
such as space envelope, mounting and ducting
ri_ m
arrangement, and others. For instance, the re- Pr -=- (6-80)

Zr
quirements for the region above 5500 gpm, and
head rises from 30000 to 65 000 feet, could be
where
met by either a single-stage centrifugal pump or
Pr--pitch or rotor vane spacing at the mean
a multistage axial-flow pump. The centrifugal
effective diameter d m, in
pump also has its dimensional limitations. A
value of less than 0.2 inch for the impeller dis- z r -=number of rotor vanes
The ratio of the rotor vane chord length Cr to
charge width, b 2 (fig. 6-34), would complicate
manufacture of shrouded impellers and make the pitch Pr is called rotor vane solidity Sr
critical the tip clearance of an open-faced impel-

Cr
ler. This establishes the lower capacity limit
for centrifugal hydrogen pumps at about 250 gpm.
s,:_ (6-s_)
where S_-= rotor vane solidity at the mean effec-

Basic Assumptions for Axial-Flow Pumps
tive diameter din.
During operation of an axial flow pump, it is The chord to pitch ratio generally increases
assumed that the meridional or axial component from rotor tip diameter dt to hub diameter dh for
of the absolute flow velocity Cm is constant structural reasons. The profile of the vane can
throughout all stages of the impeller rotor and be represented by the vane mean line (fig. 6-52)
the stator. To satisfy the flow continuity equa- which determines most of the important hydraulic
tion, the cross-sectional areas of the various properties of the vane. The thickness of the
flow passages at right angles to cm must also vane varies along the mean line for better per-
remain constant. This assumption is reasonable, formance and for structural strength. To impart
except for the effects of frictional drag at the effectively the driving action to the fluid, the
casing walls and the vanes. angle of the vane mean line, or rotor vane angle,
pUmP.
impeller st_tor of f_n _xi_l-[loW
impeller rotor, _ncl
o',/\/ \__
is gradually increased from fi2 to f13. The dif- (AH)I

ference between the two, /33-/32, is a measure (d,), - 2 (6-90)
Um
of the vane curvature along any particular vane
g
section. Generally, in axial-flow pump designs,
all vane mean lines can be approximated by a
Cm =Cu2' tan a2' =cu3' tan aj =c 2' sin a2'
circular arc. The following correlations can be
established: = c 3' sin a 3' = v 2' sin /32' = vj sin/33' (6-91)
where
/3c - 2 (6-82) i = angle of attack, deg
ii = angle allowed for circulatory flow at
the outlet, deg
Cr=2Rr sin -sin f_c (6-83) /32', /33' = relative flow angles at the rotor
inlet and outlet, deg
where a2', a3' =absolute flow angles at the rotor
/3c = chord angle of the rotor vane, deg inlet and outlet, deg
cm = meridional or axial component of the
/32 =vane angle at the rotor inlet, deg
absolute flow velocities, ft/sec
/33 =vane angle at the rotor outlet, deg
Cr = chord length of the rotor vane, in Um = rotor peripheral velocity at mean
Rr = radius of the rotor vane curvature, in effective diameter din, ft/sec
Lr = axial length of the rotor vane, in c2', cj =design absolute flow velocities at
the rotor inlet and outlet, ft/sec
(All parameters refer to the mean effective diam-
eter, din) Cu2', cu3' = tangential components of the design
An angle of attack or incidence angle "i" be- absolute rotor inlet and outlet flow
velocities, ft/sec
tween rotor inlet vane angle /32 and the direction
of the relative velocity of the flow entering the v2', v 3' : design relative flow velocities at
the rotor inlet and outlet, ft/sec
rotor, /32', is allowed for more effective driving
of the fluid. Also, an angle "ii" is allowed for Qimp =required impeller flow rate at the
circulatory flow between the rotor outlet vane rated design point, gpm
Q = rated design pump flow rate, gpm
angle /33 and the direction of the relative veloc-
ity of the flow leaving the rotor, fls'- For the Qe = impeller leakage loss rate, gpm (2 to
design of impeller rotors, velocity diagrams of 10 percent of Q)
the flows at the inlet and outlet of rotor vanes = contraction factor of vane passage
(0.85 to 0.95)
can be constructed (fig. 6-52) with the following
correlations: AHimp = required developed head per impeller
stage, ft
/32 =/32' + i (6-84) (AH) 1 = rated design developed head per

axial-flow pump stage, ft
He = hydraulic head losses per stage of
[93=/33' + ii (6-85)
impeller stator, ft
(d,), =head coefficient per axial flow pump
Qimp
Cm - (6-86) stage
3.12 x;(dl 2 - d h 2) (All applicable parameters refer to the mean
effective diameter, din)
Qimp = Q + Qe (6-87) At various cylindrical sections between vane
tip diameter dL and hub diameter dh, the follow-
.N ing correlations between vane angles and flow
um= _-_dm (6-88) velocities are established:
UmCuj - UmCu2 f dm tan fi2 --dr tan/32_

AHirap = (AH)1 + He - (6-89)
=dh tanfl2h=dxtan /32x (6-92)
DESIGNOF TURBOPUMPPROPELLANT. FEED SYSTEMS 229
dm tan fla :d_ tan fi3t Referring to figure 6-52, the velocity diagrams
at the stator inlet and outlet are constructed with
=rib tan /_3h=dx tan /_3x (6-93)
the assumption that the absolute flow velocities
um ut Uh t/x and angles at stator inlets and outlets are equal
(6-94) to the corresponding ones at the rotor outlets
dm dt dh dx
and inlets. This facilitates the design of multi-
stage axial-flow pumps using uniform rotor and
: ca2,__'
Cur__' : (6-95) stator stages.
rlm dt dh
To deflect the fluid effectively, the stator
inlet vane angles a 3 should be greater by a few
C u.___3i Cu3t' _ Cu3h ,
(6-96) degrees than the inlet absolute flow angles aa';
dm dt dh i.e., an angle of attack "i" should be allowed.
Also, an angle "ii" should be allowed between
where outlet vane angle a4 and outlet absolute flow
= rotor inlet vane angles at tip and angle a 4' for the circulatory flow (boundary con-
hub diameter, deg dition). The following correlations can be estab-
= rotor outlet vane angles at tip and lished for the vane and flow velocity diagrams of
_3t, _3h
hub diameter, deg the stator (fig. 6-52):

Ut, Uh = rotor peripheral velocities at tip
and hub diameter, ft/sec rrdm
Ps :-- (6-97)
Cu2t', Cu2 h' :tangential components of the de- gs
sign absolute rotor inlet flow
velocities at tip and hub diam- Cs
eter, ft/sec Ss :P_-s (6-98)

Cu3t', Cu3h' =tangential components of the de-
sign absolute rotor outlet flow (23 + (24
(6-99)
velocities at tip and hub diam- {2C-- 2
eter, ft/sec
• -
Function of the Statot Cs=2R s sm _---_---) - si-_-_ac (6-100)
The purpose of the stator of an axial-flow

aa :as'+ i (6-101)
pump is to convert a major portion of the tangen-
tial component of the absolute flow velocity
a 4 :a 4' +ii (6-102)
leaving the rotor into static pressure. This is
accomplished by "straightening" the flow as it C m : CU31 tan _3': CU4' tan a 4'
leaves the rotor. The stator vane curvature is
= c a' sin a a' = c 4' sin a 4' (6-103)
designed so that the fluid enters the vanes with
minimum loss, and leaves the stator with a re-
dm tan an=dr tan a3t
duced tangential component of the absolute flow
velocity. The cross-sectional areas of the stator :dh tan aah=dx tan a3x (6-104)
flow passages normal to the axial direction are
dm
equal to those of the rotor. Thus, the axial com- tan a 4 =dt tan a 4 t
ponent of the absolute flow velocity is main- =dh tan a4h=dx tan a4x (6-105)
tained. The dimensions dt and dh of the stator
can be treated as equal to the tip and hub diam- whe:e
eters of the rotor. The chord-pitch ratio of the Ps : pitch or stator vane spacing, in
stator vanes generally increases from hub diam- zs = number of stator vanes
eter d h to tip diameter dr. The axial length Ls Ss : stator vane solidity
of the stator vane at the mean effective diameter Cs = stator vane chord length, in
is usually made equal to that of the rotor, Lr. _C = stator vane chord angle, deg
a3, a4 :vane angles at stator inlet and A higher pump specific speed is linked with
outlet, deg lower solidity. Typical design values for vane
Rs : radius of the stator vane curva- solidities for the rotor and stator at the mean
ture, in effective diameter dm range from 1 to 1.3, and
Ls = axial length of the stator vane, 1.5 to 1.8, respectively.
in 3. Number of vanes Zr, zs.-A lower pump
i : angle of attack, deg specific speed generally results in a larger num-
ii : angle allowed for circulatory ber of vanes. Design values of Zr range from 14
flow at the outlet, deg to 20. Design values of Zs vary between 35 and
a3', a 4' = absolute flow angles at stator 45. Their number should have no common factor
inlet and outlet, deg with z r.
Cm = axial component of the absolute 4. Vane curvature and vane setting.-Experi-
flow velocities, ft/sec ments indicate that the head developed by an
%', c,' = design absolute flow velocities impeller rotor is essentially determined by the
of stator inlet and outlet, ft/sec
vane curvature; i.e., fis-/32- Changes in vane
cu3', Cu,' =tangential components of the de- settings, i.e., outlet vane angle /33 and inlet
sign absolute velocities at vane angle/32, by the same amount (/3a- f12
stator inlet and outlet, ft/sec = const) will not affect head rise and efficiency
ast, ash, asx:stator inlet vane angles at tip, materially.
hub and any intermediate diam- The design procedure for the impeller rotors
eter, deg and stators of a multistage axial-flow pump is
a4t, a4h, a4x : stator outlet vane angles at tip, essentially the same as that for a single-stage
hub and any intermediate diam- centrifugal pump, except for the determination of
eter, deg the number of pump stages. Design parameters
(All parameters refer to the mean effective diam- and coefficients established experimentally with
eter din, unless specified differently.) earlier successful designs should be utilized to
the fullest. Special development tests are still
Design of Impeller Rotors and Stators required to verify the characteristics of the new
design. The design procedure includes the
A number of design factors directly affect the following steps:
performance and characteristics of an axial flow 1. To meet a given set of engine system re-
pump. Evaluation of test information, on the quirements such as rated design pump developed
basis of specific speed per stage (Ns)l, shows head H, flow rate Q, and rated pump (NPSH)c,
definitely the following correlations: the pump rot.tting speed N is deternfined first
I. Impeller bob ratio, rd.-The ratio of impel- through selection of a suitable inducer of a given
ler hub diameter dh to tip diameter dt (fig. 6-51) suction specific speed (Nss)ind.
has a direct bearing on the specific speed per 2. With N established, selection of impeller
stage (Ns)_. Higher specific speed pumps have rotor and stator of a given specific speed per
smaller hubs or hub ratios which results in stage (Ns)z, combined with the determination of
greater free flow area, and thus greater capacity, the number of pump stages, can now proceed
but lower head (H/Q characteristics). On the with the aid of the following correlations.
other hand, a higher hub ratio tends to yield a
higher head coefficient per stage (¢)1. Typical (Ns) _ - N(Q)°s (6-106)
values of r d in rocket engine hydrogen pump (H)I o-Ts
designs range from 0.76 to 0.86. Typical design
values for (Ns)_ and (¢), range from 3000 to AH=hHind:Hee+n(hH)l (6-107)
5000, and from 0.25 to 0.35, respectively.
2. Vane solidities St, Ss.-The vane solidi- where
ties or chord-spacing ratios of the rotor and (Ns), = specific speed per axial-flow pump
stator are important design parameters. They stage
are selected on the basis of previous experience. N = rated design pump rotating speed, rpm
Q = rated design pump flow rate, gpm impeller tip diameter. Design values for tip
(AH) 1 = rated design pump developed head per clearances, c, between rotor and stator (fig.
axial-flow pump stage, ft 6-49) range from 0.005 to 0.010 inch.
AH = rated design pump overall developed The vane thickness along the mean line is
head, ft mainly determined by structural considerations,
AHin d = inducer rated head rise, ft since experiments indicate that there is very
Hee = hydraulic head loss at the inducer little effect on Ferforma_Lce from variation of
stator, ft vane thickness. The problems with stressing
n =number of axial-flow pump stages pump rotor vanes are similar to those with tur-
3. For the specific speed per stage (Ns), bine blades. The methods given in section 6-5
thus obtained, various design factors and coeffi- for turbines can be applied here also. Vane
cients such as impeller hub ratio rd, vane solidi- stresses include centrifugal stresses, bendit_g
ties Sr and Ss, number of vanes zr and Zs, head stresses due to lift and drag loadings on the
coefficient per stage (_),, etc., are selected vanes, and vibrational stresses.
based on past designs with comparable (Ns)l
values.
Design of Cavitating Inducers for Axial-Flow
4. The required impeller rotor and stator
Pumps
diameters, velocity diagrams and vane profiles
can now be derived from equations (6-79) through The design procedures and parameters for
(6-105). eavitating inducers in axial-flow pumps are
Impeller rotor and stator vanes are generally essentially the same as those for a centrifugal
machined from forgings (fig. 6-51 and 6-53) using pump (fig. 6-51). Usually, the inducer has a
aluminum alloys or nickel-base alloys such as cylindrical tip contour and the same tip diameter
K-Monel. In view of the relatively low head dt as the impeller. The contour of the inducer
produced by an individual axial-flow impeller hub is highly tapered fl'om a relatively small
stage, reduction of skin friction and flow turbu- diameter at the inlet to a diameter close to that
lence losses are more important than with cen- of the impeller at the outlet.
trifugal pumps. A high degree of vane stream- An inducer stator, which also serves as the
lining and polishing is required for high efficiency. front bearing support, is positioned behind the
The axial distance da between impeller rotor inducer rotor. It is designed to convert into
vanes and stator vanes (fig. 6-51) has some pressure, a portion of the tangential component
bearing on performance. Typical design values of the absolute flow velocity leaving the inducer,
of d a range from 0.02 to 0.05dr, where dr=the and to discharge the fluid at an absolute flow
velocity and angle equal to that at the outlet of
an impeller stator (c 2' = Cu' ). The inducer stator
has the same db and dt as the impeller, and thus
['.." "._I i ROTOR
the same effective passage cross-sectional area

normal to the axial velocity component cm. The
] inlet and outlet velocity diagrams
stator are shown in figure 6-52. The following
for the inducer
correlations can be established for the design of

an inducer stator:
_rdm
Pis = (6-i08)
Zis
Cis
Sis -- _ (6-109)
Figure 6-53.-Inducer, impeller rotor, and turbine
rotor assembly of a typical multi-stage axial- al + a2
flow pump. aic= 2 (6-110)

Inducer stator vane solidity design values

Cis=2Ris sin -sin aic range from 1.5 to 1.8; the number of vanes Zis,
ranges from 15 to 20. The number Zis should
al =al' +i (6-112) have no common factor with the number of impel-
ler rotor vanes, zr.
a2 = a2' + ii (6-113)
Cm=Cul ' tan UII=Cu2 ) tan a 2' Design of Casings for Axial-Flow Pumps
=c 1' sinai'=C2' sina 2' (6-114) As shown in figures 6-6 and 6-54, the casing
of an axial-flow pump consists of a cylindrical
dm tan al=dt tan al_
section which houses the inducer stage and the
=dh tan a_h:d x tan alx (6-115) impeller stages. It also includes a volute sec-
tion with radial guide vanes which is located
d m tan a 2 =d t tan a2$ behind the last 'impeller rotor stage. In addition
=dh tan a2h=dx tan a2x (6-116) to converting the tangential flow velocity com-
ponent into pressure, the volute section also
where serves to reduce the axial velocity component by
Pis = pitch or inducer stator vane spac- gradually increasing the flow area toward the
ing, in volute discharge.
Zis =number of inducer stator vanes The radial guide vanes of the volute section
Sis =inducer stator vane solidity are designed such that the fluid enters them with
Cis =inducer stator vane chord length, in minimum losses and that it leaves them in a
aic = chord angle of the inducer stator radial plane, analogous to a centrifugal pump
vane, deg (fig. 6-46). The number of radial guide vanes
a I , a 2 = vane angles at inducer stator inlet usually ranges from 17 to 23. They should have
and outlet, deg. no common factor with the number of impeller
Ris = radius of the inducer stator vane rotor vanes. Vane angle av can be determined
curvature, in by constructing the flow velocity diagram for
Lis =axial length of the inducer stator that section.
vane, in The calculations of the required areas at the
various sections of an axial flow pump volute
i = angle of attack, deg
are essentially identical to those for a centrifu-
ii = angle allowed for circulatory flow
gal pump (eqs. 6-69 and 6-70). For liquid hydro-
at the outlet, deg
gen, design values for volute flow velocities
al' , a 2' = absolute flow angles at inducer
range from 100 to 150 ft/sec. The section of the
stator inlet and outlet, deg
volute is generally circular in shape to accom-
Cm = axial component of the absolute
modate the high pressures. First-class thermal
flow velocity, ft/sec
insulation should be applied to the pump outside
CI' , C 2' = design absolute flow velocities at
surfaces. This will prevent excessive hydrogen
inducer stator inlet and outlet,
boiloff.
ft/sec
c i u', c2 u' = tangential components of the design
absolute velocities at inducer
Balancing the Axial Thrust of Multistage
stator inlet and outlet, ft/sec
Axial-Flow Pumps
alt, a_h =inducer stator inlet vane angles at
tip and hub diameters, deg Balancing of the combined axial thrust of a
a2t, a2h =inducer stator outlet vane angles at multistage axial-flow pump is an important func-
tip and hub diameters, deg tion, in view of the high pressures involved.
alx, a2x =vane angles at any diameter dx Special balancing devices, such as automatic
(All parameters refer to the mean effective diam- balance pistons, are frequently used. The bal-
eter din, unless specified differently.) ance piston is secured to the rotor assembly, as
shown in figure 6-54. It consists of a disk hav- r7 2

ATa=APc--c(df -dr 2) (6-117)
ing small clearances with a pair of seal rubs,
located on either side of the disk. A forward where
movement (toward the inlet) of the rotor assem- ATa = variation of the forward hydraulic axial
bly, and thus the balance piston, reduces the thrust, lb
clearance at the front seal rub, simultaneously Apc =variation of the fluid pressure in the con-
increasing it at the rear seal rub. As a result, trol chamber, psi
the pressure in the control chamber between front dt =diameter of the front seal rub, in
and rear seal rubs is reduced. This effect coun- dr = diameter of the rear seal rub, in
teracts the forward hydraulic axial thrust of the
rotor assembly and restrains its forward motion.
Similarly, a rearward movement (reverse thrust)
of the rotor assembly is counteracted by a pres- The following design data, based on engine
sure increase in the control chambers. The system requirements and on experimental model
volume to the rear of the control chamber, through test results, are given for the alternative A-2
cavities in the rotor, communicates with the low- stage engine fuel (liquid hydrogen) pump, based
pressure region of the pump inlet. The variation on an axial-flow, multistage configuration (simi-
in forward axial thrust can be expressed as lar to figs. 6-51 and 6-54).
o t "- vo,
(:Iv, - ..... _,:\_ _ _ :\

--4---.. ....... \\
--Jr -- ---- dr df
•"_"_!\_4, ',"-,"R.P.M. _..-_,_

.\",",\
/.L_.__\'< :_<\\ : \:',c', _ _ , ,,\', >_ Zf I
/-- ......... K_;,_' I _?/ I
/ _ _,_ , /V" I
........ _/ jl
_.Y/>
FRONT SEAL RUB'--" X-BALANCE PISTON

CONTROL CHAMBER
Figure 6-54.-Axial-(low pump volute casing and balance piston arrangement.

Rated design pump developed head, AH

[N(q)°s] _.33_
= 44 800 ft (AH), = L(--N_), J
Rated design pump flow rate, Q = 6080 gpm
Rated design pump (NPSH)c = 135 ft _[27000×(60s0)°713
Inducer general configuration = cylindrical tip - [ 325-O ] = 5580 ft
contour; tapered hub contour
Inducer suction specific speed, (Nss)in d
= 53 400
From equation (6-90), the peripheral speed at
Inducer inlet flow coefficient, ¢i=0.09 max
the impeller mean effective diameter dm
Inducer head coefficient, ¢ind = 0.307
Inducer leakage loss rate, Qee = 0.03 Q
132.2 × 5580
Inducer stator head loss, Hee =0.08AHind
Inducer stator vane solidity, Sis--1.53
Number of inducer stator vanes, Zis = 17
Specific speed per axial-flow pump stage, From equation (6-88), the impeller mean effec-
tive diameter
(Ns) 1 = 3200
Head coefficient per axial-flow pump stage,
(,/,), : 0.304 720 Um_ 720 x 768
Impeller hub ratio, rd : 0.857 din- 7rN 7r×270_=6.52 in
Impeller leakage loss rate, Qe = 0.06 Q

Head loss per stage of the impeller stator, From equation (6-79), the impeller rotor tip
He = 0.08 (AH)I diameter (cylindrical tip contour)
Impeller rotor vane solidity Sr at the mean
effective diameter = 1.05 2 _ 2
---7 in
Number of impeller rotor vanes, z_ = 16 dt--dm (1 +rd2 ) 6.52 1+0.734
Impeller stator vane solidity 8s at the mean
effective diameter : 1.61 The impeller rotor hub diameter
Stator and rotor vane passage contraction
factor, e = 0.88 dh--rd×dt:O.S57x7=6 in
Angle of attack at the vane inlet, i = 4°
Angle allowed for circulatory flow at the vane The rotor vane height
outlet, ii : 5°
Calculate and design basic pump dimensions
and vane detail of: (a) inducer stator, and (b) hv =_=_-_= 0.5 in
impeller rotor and stator.

Seven main pump stages and one inducer
stage are used. From eq. (6-107), the required
inducer head rise
Solution (refer to fig. 6-52)
(a) Inducer stator AHind (1 - 0.08) : AH- n(hH),

From equation (6-10), the rated pump rotating
speed can be determined: _ (44 800 - 7 × 5580)
AHind
0.92
Nss (NPSH)c o._ = 6240 ft

N: (N ss : (N ss)ind : 53 400)
QO_
Use an inducer tip diameter of 7 inches, equal
to the impeller rotor tip diameter. Inducer head
N - 53 400 × (135) °Ts = 27 000 rpm
rise:
(6080) _s
From equation (6-106), the pump developed Nrrdt _ 27 000 × 7r× 7

= 826 ft/sec
head per stage ut = 720 720
Substitute this into equation (6-66):
3.12 × _-× 230

ut2 (826)2 = 6500 ft
AHind = _'ind _ = 0.307 × 32.----_
The mean effective diameter at the inducer
This is slightly more than the minimum of 6240 outlet
feet required.
The required impeller and inducer flows are
obtained from equations (6-35) and (6-63): d_ -- dt2 +d,h
2 2 = _49+
' _37.58 =6.57 in
Qimp -- Q + Qe = 6080 (1 + 0.06) = 6450 gpm Tile peripheral speed at d,
Vind=Q+Qee +½Qe
= 6080 (1 + 0.03 + 0.03) = 6450 gpm Nrrd _ _ 27 000 × ,7 × 6.57 : 774 ft/sec
"1- 720 720
We use a hub diameter doh=2.9 inches at the
inducer inlet. Considering that we are using a From equation (6-66), the tangential compo-
cylindrical tip contour, the absolute inlet veloc- nent of the inducer design absolute outlet flow
ity and its meridional component of the inducer velocity,
flow result from equation (6-59) as:
, g 6500×32.2
: 270 ft/sec
Qind cu, = AHind _ - 774
Co' _- Cmo :
3.12 x-_(dt 2 -doh 2)

The inducer design absolute outlet flow
velocity
6450
- = 64.8 ft/sec
3.12 ×4(49- 8.41) cl, =VfCul, 2 + Cm 2 : V/72 900+ 52 900 : 354.7 ft/sec
Check for inducer inlet flow coefficient: The inducer design absolute outlet flow angle
Cmo 64.8 , cm 230

_ind ..... 0.0784 (< 0.09 max. specified)
ut 826 tan a I =--cul' :_-0= 0.852; a 1 =40°26 '
Use identical values cm for the meridional

We use an inducer stator with a meridional
component of the absolute flow velocities through
the inducer outlet, the stators and the rotors. flow area equal to that of the impeller rotors and
From equation (6-86) stators (i.e., dr=7 in, dh=6 in, dm=6.52 in, and
e=0.88). We also assume that the absolute flow
conditions at the inducer stator inlet are identi-
Qimp
C/B-- cal to those at the inducer outlet. From equation
2
3.12x_(dt - dh 2) e (6-112), the stator inlet vane angle at dm
6450 al =a,'+i=40° 26 ' +4 ° =44° 26 '

= 230 ft/sec
3.12 x4× (49- 36) x 0.88

For a design absolute flow angle at the stator
outlet, a2' of 65 °, equation (6-113) yields a vane
From equation (6-60), the required hub diam- angle at the stator outlet
eter at the inducer outlet
a2 =a2'+ii=65 ° +5 ° =70 °
dlh=_dt2 Qind The tangential component of the stator abso-

3.12×4×cm lute outlet flow velocity
cm 230 230 Nominal mean effective diameter, d m =6.52 in

c, 2' = tan a 2' = tan 65 ° - 2.145 = 107.2 ft/sec
Vane elements (fig. 6-52), al =44°26'; a2
=70°; aic=57°13'; Sis=l.53; Zis=17;
The stator absolute outlet flow velocity Pis=l.205in; Cis=l.844in; Lis=l.55in;

Ris =4.16 in
c 2' = \/c u 2' 2 + cm 2 = X/11 490 + 52 900
-- ,,:'65 350 = 253.8 ft/sec (b) Impeller rotor and stator

Assume that the design absolute flow condi-
From equation (6-108), the pitch of the in- tions at the impeller rotor inlets and at the im-
ducer stator vanes at dm peller stator outlets are identical to those at the
inducer stator outlet. Thus
Pis - rrdm _ ,7×6.52 1.205 in

Zis 17 a2' =a4' =65 °
c 2' =c 4' =253.8 ft/sec

From equation (6-109), the chord length of the
inducer stator at dm cm =230 ft/sec
cu2' = cu4' = 107.2 ft/sec

Cis = Sis Pis = 1.53 x 1.205 = 1.844 in
The design relative flow angle at the impeller

From equation (6-110), the chord angle of the
rotor inlets
inducer stator vanes at dm
cm 230
a, +a2_44°26 ' +70 °
=57o13 ' tan /32 ' - um-cu2 -1 - 768-107.2 = 0.344; f12' = 19°
aic - 2 2
The relative flow velocity at the impeller

From equation (6-111), the axial length of the
rotor inlets
inducer stator vanes at d m
v2' =\/ium-Cu2') 2 +cm 2

Lis = Cis sin aic -- 1.844 × sin 57°13 ' = 1.55 in
= ,_/436 700 + 52 900 = 699.6 ft/sec
The radius of the inducer stator vane curva-

ture at dm From equation (6-84), the rotor inlet vane
angle at d m
Cis 1.844
/32 =/32' +i=19 ° +4° :23 °
Ris=2 sin(-_) -2xsin(12°47')=4"16 in
From equation (6-89), the required developed

A-2 Stage Engine Fuel Pump Inducer Stator
head for the impeller rotor
Design Summary
(Unless specified otherwise, all data are at
AHimp = (AH), + He = 5580 (1 + 0.08)
the mean effective diameter din.)
= 6026 ft per stage
Inlet flow velocity diagram (fig. 6-52), aa'
=40°26'; c 1' =354.7 ft/sec; c m =230 ft/sec;
Cu,' = 270 ft/sec
nent of the design absolute flow velocity at the
Outlet flow velocity diagram (fig. 6-52),
impeller rotor outlet
a s' = 65°; c 2' = 253.8 ft/sec; Cm = 230 ft/sec;
cu2' = 107.2 ft/sec
ghHimp 32.2 × 6026
Nominal tip diameter, dt = 7 in '- _ Cu2' - +107.2
Cu_ um 768
Nominal hub diameter, dh = 6 in
Nominal vane height, hv=0.5 in = 359.6 ft/sec
The impeller rotor design absolute outlet flow Cr 1.346

velocity Rr= /33-/32 2sin5°33 '=6'95in
2 sin-
2
C3' = V/Cu3' 2 + Cm2 _- V/129 300+ 52 900 = 426.9 ft/sec
Assume that the design absolute flow condi-

The impeller rotor design absolute outlet flow tions at the impeller stator inlet are identical to
angle
those at the impeller rotor outlet. From equation
(6-101), the impeller stator inlet vane angle at
Cm 230 dm
tan a3'=Cu 3,-359.6=0.641; a3'=32°40'
a3 =a3' +i = 32°40 ' +4 ° -=36o40 ,
The impeller rotor design relative outlet flow

From equation (6-102), the impeller stator
velocity
outlet vane angle at dm
v 3' :V/(Um - Cu3') 2 + cm 2

(Z4 -- a 4' + ii : 65 ° + 5 ° -- 70 °
= X/166 800 + 52 900 : 468.7 ft/sec
The axial length of the stator vanes at d m is

The impeller rotor design relative outlet flow
equal to that of the rotor vanes; thus
angle
Ls-- Lr: 0.645 in

_ cm 230
tanfl3' (um-cu3') 408.4 -0.564; fij=20°26'
stator vanes at dm
From equation (6-85), the rotor outlet vane
angle at d m a3+a4_36°40'+70°=53o20,
ac- 2 2
f13 : fi3' + ii : 29026 ' + 5 ° = 34026 '

From equation (6-100), the chord length of the
From equation (6-80), the rotor vane pitch stator vanes at d m
rfdm 77×6.52 Ls 0.645

Pr= -= - 1.281 in Cs:_n ac -sin- (53o20,)=0.805 in
Zr 16
From equation (6-81), the chord length of the The radius of the stator vane curvatures at dm
rotor vanes at dm
Cs 0,805
Rs: = 1.41 in
Cr = SrPr = 1.05 x 1.281 = 1.346 in

rotor vanes at dr, From equation (6-98), the stator vane pitch
at dm
/92 +/33 _23 ° +34026 '
/3c - 2 2 = 28o43'
Cs 0.805 _
Ps-=_-;: 1-y-AT:0._ in
From equation (6-83), the axial length of the
rotor vanes From equation (6-97), the number of the stator
vanes
Lr = Cr sin tic = 1.346 × sin 28°43 ' = 0.645 in

Z_
m rrx6.a2 r- ..
Z s =--:--"-x--c---. = ql
The radius of the rotor vane curvature at dm Ps u.5
4
A-2 St:age Engine Fuel Pump Impeller Rotor and able energy of the turbine working fluid and thus
Smtor Design Summary the gas spouting velocity Co is relatively low, a
(Unless otherwise specified, data are all at higher turbine velocity ratio U/C o may be achieved
the mean effective diameter din.) with a moderate turbine rotor blade speed U. As
Rotor inlet flow velocity diagram (fig. 6-52), shown in figure 6-27, this suggests the use of a
a2' =65°; fi2' : 19°; Um=768 ft/sec; v2' relatively simple single-stage single-rotor im-
=699,6 ft/sec; c2'= 253.8 ft/sec; ca2' =107.2 pulse turbine. We have selected this type for
cm = 230 ft/sec the A-2 stage oxidizer tnrbopump, at the same
Rotor outlet flow velocity diagram (fig. 6-52), time taking advantage of its overall simplicity.
aj=32°40'; /33':29°26'; um=768; v 3' In most direct-drive turbopump configurations,
=468.7 ft/sec; cj =426.9 ft/sec; cuj=359.6 such as the A-1 stage engine turbopump (fig.
ft/sec; cm : 230 ft/sec 6-63), where turbine rotating speed N and conse-
Stator inlet flow velocity diagram (fig. 6-52), quently turbine velocity ratio U/C o tends to be
a3' =32°40'; c3' =426.9 ft/sec lower than ideal, a single-stage two-rotor
Stator outlet flow velocity diagram (fig. 6-52), velocity-compounded impulse turbine (figs. 6-9
a4' = 65°; c4' = 253.8 ft/sec; ca4' = 107.2 and 6-55) is selected for best results. Figure
ft/sec; cm = 230 ft/see 6-27 indicates that the optimum efficiency of a
Nominal rotor and stator tip diameter, dt= 7 in velocity-compounded turbine can be achieved at
Nominal rotor and stator hub diameter, dh = 6 in a relatively low U/Co value.
Nominal rotor and stator vane height, hv = 0.5 On the other hand, if a reduction gear train is
in provided between pumps and turbine, such as in
Nominal mean rotor and stator effective diam- the turbopump shown in figure 6-14, the turbine
eter, dm=6.52 in can be operated at a much higher rotating speed
Rotor vane elements (fig. 6-52), d2 =23°; (over 25 000 rpm). A higher value of U/Co can be
fl3=34026'; /_c=28°43'; Sr=l.05; zr=16; achieved with reasonable turbine wheel size.
Pr=l.281in; Cr=l.346in; Lr=0.645in; Then a higher performance, two-stage, two-rotor,
Rr=6.95 in pressure-compounded impulse turbine (fig. 6-10)
Stator vane elements (fig. 6-52), a 3 =36°40'; may be used.
a4=70°; ac=53°20'; Ss=l.61; Zs=41; 2. After the type of impulse turbine has been
Ps=0.5; Cs=0.S05in; Ls=0.645in; selected, the next step is the determination of
Rs = 1.41 in the turbine rotor size. Once the characteristics
of the turbine working-fluid (i.e., inlet temper-
ature To, specific heat ratio y, etc.), the turbine
6.5 DESIGN OF TURBINES
pressure ratio R_, and the pump or turbine roSa-
For rocket engine applications, impulse tire speed N have been set forth, a larger diam-
turbines are preferred, for their simplicity and eter for the turbine rotor tends to result in a
light weight. Our discussion will be confined to higher velocity ratio U/C o , or higher efficiency.
these turbines only. Figure 6-55 shows the gen- However, it also results in higher assembly
eral arrangement of a typical single-stage two- weight, larger envelope, and higher working
rotor velocity-compounded impulse turbine, stresses. Thus, the final selection of the tur-
bine rotorsize, and consequentlythe U/C o ratio,
is often a design compromise.
General Design Procedure
3. The requiredpower output from the turbine
The following steps are essential in the de- shaft must be equal to the net input to the pro-
sign of a rocket engine impulse turbine: pellant pumps, plus the mechanical losses in the
1. The first item of importance is the selec- gear train (if any), plus the net power required
tion of the proper type. A single-stage single- for auxiliary drives. The required flow rate of
rotor turbine (fig. 6-8) is used if the required the turbine working fluid can then be calculated
turbine power is low, since in this case the by equation (6-19) after required turbine power,
efficiency of the turbine has less effect on over- available energy of the working fluid (eq. 6-18),
all engine systems performance. When the avail- and overall turbine efficiency (estimated from
DESIGNOF TURBOPUMP PROPELLANT-FEED SYSTEMS 239
FLOW
TURBINE --
TURBINE INLET INLET
GAS MANIFOLD SHAFT FLANGE
ROTATING
SEALS
FtRST
BLADES
ROTATION
SECOND ROTOR .......
BLADES
SECTION A-A
DRIVING
/-'- ROTOR SPL/NES
DISC
CONTINUOUS RING
TYPE ROTOR BLADE
SHROUD
-_ dt
Figure 6-55.-Typical single-stage, two-rotor velocity compounded impulse turbine.
figure 6-27 for a given U/C o ratio and turbine above that of an isentropie process. This effect
type), have been established. is known as reheat. As a result of the above
4. Now the dimensions of the stationary noz- effects, the actual gas spouting velocity at the
zles, as well as those of the rotor blades, can turbine nozzle exit tends to be less than the
be calculated based on the characteristics and ideal velocity calculated for isentropic expan-
the flow rate of the turbine working fluid. sion (from stagnation state at the nozzle inlet to
the static pressure at the rotor blade inlet).
Furthermore, the effective flow area of a nozzle
is usually less than the actual one, because of
Design of Turbine Nozzles
circulatory flow and boundary layer effects. The
The nozzles of most rocket engine turbines following correlations are established for the
are basically similar to those of rocket thrust design calculations of turbine nozzles:
chambers. They are of the conventional
converging-diverging De Laval type. The main
function of the nozzles of an impulse-type tur- Nozzle velocity coefficient kn
bine is to convert efficiently the major portion of
available energy of the working fluid into kinetic
energy or high gas spouting velocity. The gas- Actual gas spoutingvelocity at the nozzle
flow processes in the thrust chamber nozzles are exit,ft/sec
directly applicable to turbine nozzles. However, Idealgas velocitycalculated forisentropic
the gas flow in an actual nozzle deviates from expansion from stagnation state at the
ideal conditions because of fluid viscosity, fric- nozzle inlet to static pressure at the rotor
tion, boundary layer effects, etc. In addition, blade inlet, ft/sec
the energy consumed by friction forces and flow
C1
turbulence will cause an increase in the temper- (6-118)
ature of the gases flowing through a nozzle, Co
Nozzle efficiency _?n AH0__" = isentropic enthalpy drop of the gases

flowing through the nozzles, due to
Actual gas kinetic energy at the nozzle exit expansion, Btu/lb
Ideal gas kinetic energy (isentropic The performance of a turbine nozzle, as ex-
expansion) pressed by its efficiency or velocity coefficient,
is affected by a number of design factors, such
C12
as
2g _ka2 (6-119) (1) Exit velocity of the gas flow
(2) Properties of the turbine gases
2g (3) Angles and curvatures at nozzle inlet and
exit
Nozzle throat area coefficient ent (4) Radial height and width at the throat
(5) Pitch or spacing, and number of nozzles
Effective area of the nozzle throat
Z (6-120) Design values for the efficiency and velocity
Actual area
coefficients of a given turbine nozzle may be
determined experimentally, or estimated from
Actual gas spouting velocity at the nozzle exit,
past designs. Design values of nozzle effi-
ft/sec:
ciency 77_ range from 0.80 to 0.96. Design val-
ues of nozzle velocity coefficient kn vary from
0.89 to 0.98. The nozzle throat area coefficient
ent generally will increase with nozzle radial
C_ = k n C O = kn gJCpTo -\Po/ J
height, with design values ranging from 0.95 to
=kn_/2gJAHo-i" (6-121) 0.99.
The cross-sectional shape (fig. 6-56) of rocket
Amount of nozzle reheat: turbine nozzles is square, or, more frequently,
rectangular. They are closely spaced on a cir-
(i - kn 2)C 2 _ (1 - _n) C 2 cular arc extending over a part of (partial admis-
(6-122)
qnr= ka22gJ 7?n2gJ sion), or all (full admission), the circumference.
Most high-power turbines use full admission for
better performance.
Required total nozzle throat area, in2: While the gases are passing through a nozzle
and expanding, the direction of flow is changing
wt (6-123) from an approximately axial direction to one
Ant- ),+___A_l forming the angle al (fig. 6-56) with the plane of
r 2 7Y-' rotation, at the nozzle exit. Thus the turning
angle is 90 ° - a,. The angle 0a of the nozzle
_ntPo_ RT ° centerline at the exit usually is the result of a
design compromise. Theoretically, better effi-
where
ciency is obtained through the use of a smaller
Cp = turbine gas (working fluid) specific
nozzle exit angle, since the rotor blading work
heat at constant pressure, Btu/lb-
is larger and the absolute flow velocity at the
deg F rotor blade exit is smaller. However, a smaller
Y = turbine gas specific heat ratio
nozzle exit angle means a larger angle of flow
R =turbine gas constant (1544/molecular
deflection within the nozzle, which causes higher
weight), ft/°R
friction losses. Design values of 9n range from
To = turbine gas total (stagnation) temper-
15 ° to 30 ° . The actual effective discharge angle
ature at the nozzle inlet, °R
al of the gas jet leaving the nozzle tends to be
fvt = turbine gas mass flow rate, lb/sec
greater than On, because of the unsymmetrical
Po = turbine gas total pressure at the noz-
nozzle shape at the exit.
zle inlet, psia
= turbine gas static pressure at the rotor A sufficiently large nozzle passage aspect
Pl
blade inlet, °R ratio, hnt/bnt, is desirable for better nozzle
DESIGN OF TURBOPUMP PROPELLANT.FEED SYSTEMS 24l
-°'-1 ( -°.
NOZZLE
t'b BLADE
Jl TURBINE
II SYMMETRICAL
ROTOR.__"___/, _,"__ "'-__ ,__ ROTOR
IIBLADES .....
',_( ROTATION,,_ U \-
-F--i-_f--'--,'_---
----I- Y -7('2
e _ I..-P=--4
_--b.o
".--SHROUD _"_t b b2 " ] U "'-7""--
VELOCITY DIAGRAMS OF A
TYPICAL SINGLE-STAGE
IMPULSE TURBINE
F--"=,---I _.___ __ .
_".._._ l I_ UNSYMMETRICAL _',,._ _'.."_
_X,,_",,_ ) /'_ ROTOR BLADES _-'X."_ "_'_
I hb2> hbl
F b2 7 _--SHROUD b2
Figure 6-5G.-Nozzles, rotor blades, and velocity diagrams o[ a typical single-stage impulse turbine.
efficiency. For a given nozzle height, an in- Pitch or nozzle spacing:

crease in aspect ratio can be secured by de-
creasing the nozzle pitch, Pw However, a small
dm
pitch, and consequently a large number of noz- Pa = rr-- (6-126)
Zn
zles, zn, w_th attendant increase in wall surface,
tends to increase friction losses. The determi-
where
nation of nozzle pitch thus also requires a de-
#/t = turbine gas mass flow rate, lb/sec
sign compromise. The following correlations are
established for the calculation of nozzle flow pt =density of the gases at nozzle exit, lb/ft 3
areas: C 1 =gas spouting velocity at nozzle exit,
ft/sec
_ne =nozzle exit area coefficient
Total nozzle throat area, in2: hat =radial height at nozzle throat, in
hne =radial height at nozzle exit, in
Ant = znbnthnt (6"124) bnt =width normal to flow at nozzle throat, in

bne=width normal to flow at nozzle exit, in
Total nozzle exit area, in2: zn =number of nozzles
0n =angle between nozzle exit centerline and
144 _i,t plane of rotation, deg
Ane-pICtene-Znbnehne tn =thickness of nozzle partition at exit, in
dm= mean diameter of nozzles and rotor blades,
=znhae (Pn sin 0n-tn) (6-125) in
Turbine nozzle block and inlet gas manifold

cos al +C2 cos a2)
assembly can be made of, for instance, welded
sections of forged Hastelloy C. However, the
airfoil surfaces should be blended smoothly =_(VI cos E1 +V2 cos/_2) (6-128)
between the defined contour and the sections.
,TdmN
U- (6-129)
Design of Impulse Turbine Rotor Blades 720
The function of the rotor blades in an impulse

For subsequent calculations, the following
turbine (figs. 6-55 and 6-56) is to transform a
relation will be useful:
maximum of the kinetic energy of the gases
ejected from the nozzles into useful work. Theo-
C 1 sin al
retically, there should be no change of gas pres-
tan/_1-C1 cos al-U (6-130)
sure, temperature, or enthalpy in the rotor blades.
In actual operation however, some gas expan-
sion, i.e., reaction, usually occurs. Furthermore, Axial thrust at blades, lb/lb of gas flow/sec
the actual gas flow through the rotor blades
deviates from ideal flow conditions because of C 1 sin al - V2 sin f12
Fa = (6-131)
friction, eddy currents, boundary layers, and g
reheating.
Blade velocity coefficient:
The velocity vector diagram shown in figure
6-56 describes graphically the flow conditions at V2
the rotor blades of a single-stage, single-rotor kb:V_-, (6-122)

turbine, based on tile mean diameter din. The
gases enter the rotor blades with an absolute Blade efficiency:
velocity C_, and at an angle a, with the plane of
rotation. The tangential or peripheral speed of Work transferred to blades Eb
the rotor blades at the mean diameter is U. V 1 ,?b : Kinetic energy input =C---5 (6-133)
and V2, the relative velocities at the blade inlet 2g
and outlet, differ, i.e., V, > V2, due to friction
losses. Ideally, the gas should leave the blades
Ideally, 71b is a maximum for a single-rotor
at very low absolute velocity C 2 and in a direc- impulse turbine, when the turbine velocity ratio:
tion close to axial for optimum energy conver-
sion in the blades. The forces generated at the
U cos al
rotor blades are a function of the change of
C1 2
momentum of the flowing gases. The following
correlations may be established for design cal-
i.e., when U=½C,t
culations of the rotor blades of a single-stage,
where C_ t is the tangential component of C_.
single-rotor turbine.
Tangential force acting on the blades (lb/lb of

Max. ideal _b = cos22a_(1 ' cos_
+ _b co--_-_-_fll)(6-134)
gas flow/sec):
Ft=I(c, COS a 1 + C 2 cos a2)

If there is some reaction or expansion of the
gas flowing through the blades, the relative gas
=l(v, cos _, + V2 cos f12) (6-127) flow velocity at the rotor blade outlet can be
5
calculated as
Work transferred to the blades (ft-lb/lb of gas

V 2=\/kb2Vl2 + 2gJ_?nAHi-2' (6-135)
flow/sec):
DESIGN OF TURBOPU_P PROPELLANT-FEED SYSTEMS 243
Amount of reheat in the rotor blades, Btu/lb of Design values of kb vary from 0.80 to 0.90. De-
gas flow: sign values of _b range from 0.7 to 0.92.
Referring to figure 6-56, the radial height at
qbr=(1 v?, the rotor inlet, hb, is usually slightly larger (5
-kb )2g j-r(1- r]n)A/-/l-2' (6-136)
to 10 percent) than the nozzle radial height hn.
where This height, together with the blade peripheral
al, a2 =absolute gas flow angles at the inlet speed U, will determine the centrifugal stress in
and outlet of the rotor blades, dog the blades. The mean diameter of the rotor
ill, fi2 = relative gas flow angles at the inlet blades is defined as dm= d_- hb, where d t is the
and outlet of the rotor blades, dog rotor tip diameter. Pitch or blade spacing, Pb,
C,, C: =absolute gas flow velocities at tile is measured at the mean diameter dm. There is
inlet and outlet of the rotor blades, no critical relationship between blade pitch Pb
ft/sec and nozzle pitch Pn. There just should be a
V,, V 2 =relative gas flow velocity at the inlet sufficient number of blades in the rotor to direct
and outlet of the rotor blades, ft/sec the gas flow. The number of blades zb to be
U = peripheral speed of the rotor, ft/sec employed is established by the blade aspect
dm = mean diameter of the rotor, in ratio, hb/Cb and the solidity Cb/Pb, where Cbis
_?n = equivalent nozzle efficiency appli- the chord length of the rotor blades. The magni-
cable to the expansion process in the tude of the blade aspect ratio ranges from 1.3 to
blades 2.5. Design values of blade solidity vary from
AH,_2,=isentropic enthalpy drop of the gases 1.4 to 2. Best results will be determined by
flowing through the rotor blades due experiment. The number of rotor blades should
to expansion or reaction, Btu/lb; have no cdmmon factor with the number of noz-
AH1-2' = 0 if only impulse is ex- zles or of stator blades.
changed The blade face is concave, with radius rt.
All parameters refer to the mean diameter d m, The back is convex, with a circular arc of small
unless specified otherwise, The turbine overall
radius rr concentric with the face of the adjoin-
efficiency Ut defined by equation (6-19) can be ing blade ahead. Two tangents to this arc to
established for a single-stage, single-rotor im-
form the inlet and outlet blade angles 0b_ and
pulse turbine as
Oh2 complete the blade back. The leading and
trailing edges may have a small thickness tb.
y_: _?n_b _rn (6-137)
The inlet blade angle 0b_ should be slightly
where larger than the inlet relative flow angle fl_. If
rl_ =nozzle efficiency Ob_ <fl_, the gas stream will strike the backs of
)7b =rotor blade efficiency the blades at the inlet, exerting a retarding effect
fir. =machine efficiency indicating the mechan- on the blades and causing losses. If Ob_ >fi_,
ical, leakage, and disk-friction losses in the stream will strike the concave faces of the
the machine. blades and tend to increase the impulse. The
Equation (6-134) shows that the blade effi- outlet blade angle Oh2 is generally made equal
ciency ;?b improves when/32 becomes much to the outlet relative flow angle /32.
smaller than I3L. Reduction of/32 without de- The mass flow rate _'t through the various
creasing the flow area at the blade exit can be nozzle and blade sections of a turbine is as-
achieved through an unsymmetrical blade design sumed constant. The required blade flow areas
(fig. 6-56), where the radial blade height in- can be calculated by the following correlations.
creases toward the exit. In actual designs, the Note that the temperature values used in calcu-
amount of decrease of fi2, or the increase of lating the gas densities at various sections must
radial height, is limited considering incipient be corrected for reheating effects from friction
flow separation and centrifugal stresses. Gener- and turbulence.
ally, the _2 of an unsymmetrical blade will be
approximately f3_-(5 ° to 15°). Equation (6-134) p1VlAblebl _p2V2Ab2eb2
also indicates that _b improves as a_ is reduced. Wt = 144 144 (6-138)
1L
Total blade inlet area, in2: of various blade sections at different radii gen-
erally do not fall on a true radial line. Thus the
Abi =Zbbbzhbz =Zbhbi (Pb sin Obl -tb) (6-139) centrifugal forces acting upon the offset cen-
troids will produce bending stresses which also
Total blade exitarea,in2: are a maximum at the root section.
2. Bending due to gas loading.-The tangen-
Ab2=Zbbb2hb2=Zbhb2(Pb sin Ob2-tb) (6-140) tial driving force and the axial thrust produced
by the momentum change of the gases passing
where over the blades may be treated as acting at the
Pb = pitch or rotor blade spacing midheight of the blade to determine the amount
= rrdm/zb, in (6-140a) of bending induced.
P,, P2 =density of the gases at the inlet and 3. Bending due to vibration loads.-The gas
outlet of the rotor blades, lb/ft _ flow in the blade passages is not a uniform flow
V z, V 2 =relative gas flow velocities at the as assumed in theory, but varies cyclically from
inlet and outlet of the rotor blades, minimum to maximum. The resultant loads repre-
ft/sec sent a dynamic force on the blades, having a
ebl, (b2 = area coefficients at inlet and outlet corresponding cyclic variation. If the frequency
of the rotor blades of this force should become equal to the natural
Zb --number of blades frequency of the blades, deflections may result
hb_, bb2 :radial height at the inlet and outlet which will induce bending stresses of consider-
of the rotor blades, in able magnitude.
bbl, bb2 : passage widths (normal to flow) at Detail stress analyses for rotor blades can be
the inlet and outlet of the rotor rather complex. A basic approach is to counter-
blades, in act a major portion of the bending moments from
0b_, 062 :rotor blade angles at inlet and out- gas loading with the bending moments induced
let, deg by the centrifugal forces at nominal operating
tb = thickness of blade edge at inletand speeds. This can be accomplished by careful
outlet,in
Typical constructionsof rocketturbinerotor
blades and disks are shown in figures6-53,6-55,
i- INTEGRAL TYPE SHROUD
6-56,and 6-57. Usually,blades are designed
with a shroud, to preventleakage over the blade
tips and to reduce turbulenceand thus improve
efficiency.Frequentlythe shroud forms an
integralportionof the blade, the shroud sections
fittingcloselytogetherwhen assembled. In _ gLADE
otherdesigns the shroud may form a continuous _ ROOT
A-A
ring (fig.6-55)which is attachedto the blades
by means of tongues at the blade tip,by rivets, S.HROUDED BLADES FABRICATED BY PRECISION
or is welded to the shrouds. The blades may be CASTING PROCESS
eitherwelded to the disk, or attached to itusing

"fir-tree"or otherdovetailshapes.
The main loads to which a rotorblade is
exposed can be dividedintothreetypes:
_ BLADE !_ z
1. Tension and bending due to centri[ugal
[orces.-The radial component of the centrifugal JOINT L.,_ DISK
forces acting on the blade body produces a cen-

BLADES WELDED TO 8LADES ATTACHED TYPICAL "FIR TREE"
trifugal tensile stress which is a maximum at the THE DISK TO THE DiSK BY TYPE TANG
"FIR TREE"TYPE
root section. As a remedy, blades are often TANG
tapered, with the thinner section at the tip, for

lower centrifugal root stresses. The centroids Figure 6-57.-Typical rotor blade constructions.
blade design. Thus the centrifugal tensile The stresses in a turbine rotor disk are in-
stresses become a first consideration in blade duced by (1) the blades, and (2) the centrifugal
design, while other details such as centroid forces acting on the disk material itself. In
location and root configuration are established addition, there will be shear stresses resulting
later to fulfill design requirements. The follow- from the torque. As seen in figure 6-55, turbine
ing correlations are established at the blade root disks are generally held quite thick at the axis,
section where stresses are most critical. but taper off to a thinner disk rim to which the
Centrifugal tensile stress at the root section blades are attached. In single-rotor applica-
of blade of uniform cross section, psi: tions, it is possible to design a disk so that
both radial and tangential stresses are uniform
Sc = 0.0004572 1pbhbdmN2 (6-141) at all points, shear being neglected. In multi-
g
rotor applications, it is difficult to do this be-
cause of the greatly increased axial length and
Centrifugal tensile stress at the root section of
the resulting large gaps between rotor and stator
a tapered blade, psi:
disks.
Equation (6-144) may be used to estimate the
Sct= O.O0045721pbhbdmN 2 stresses in a uniform stress turbine disk, neglect-

ing rotor blade effects:
(6-142) Sd=O.OOOll41pd dd 2N2

log ___o_ (6-144)
Bending moment due to gas loading at the root where
section, in-lb: Sd =centrifugal tensile stress of a constant

stress turbine disk, psi
Pd= density of the disk material, lb/in s
M _ hbcct /F,_+Fa 2 (6-143) d d = diameter of the disk, in
S-2Zb _
N =turbine speed, rpm
to =thickness of the disk at the axis, in
where tr = thickness of the disk rim at d d, in
pb =density of the blade material, lb/in 3 Equation (6-144a) permits estimation of the
hb =average blade height, in stresses in any turbine disk, neglecting effects
din =mean diameter of the rotor, in of the rotor blades:
N =turbine speed, rpm

ar = sectional area at the blade root, in 2 Sd = O.O0044251W dri N2 (6-144a)
at'= sectional area at the blade tip, in 2 g ad
_i,t = turbine gas flow rate, lb/sec
zb =number of blades where
Ft = tangential force acting on the blades, Sd = centrifugal tensile stress of the turbine
lb/lb/sec (eq. (6-127)) disk, psi
Fa = axial thrust acting on the blades, lb/lb/ Wd = weight of the disk, lb
sec (eq. (6-131)) ri =distance of the center of gravity of tile half
The bending stresses at the root can be cal- disk from the axis, in
culated from the resultant bending moment. The ad =disk cross-sectional area, in 2
vibration stresses can be estimated from past N =turbine speed, rpm
design data. If the blade is fitted with a sepa- For good turbine design, it is recommended
rate shroud, its centrifugal force produces addi- that at maximum allowable design rotating speed,
tional stresses at the root. The total stress at the S d calculated by equation (6-144a) should be
the root section is obtained by adding these about 0.75 to 0.8 material yield strength.
stresses to those caused by the centrifugal Turbine rotor blades and disks are made of
forces acting on the blades. high-temperature alloys of three different base
materials: iron, nickel, and cobalt, with chro- blades. Angles ]3,, f12, /33, and /34 represent
mium forming one of the major alloying elements. the flow directions of V,, V 2, V 3, and V 4.
Tensile yield strength of 30 000 psi minimum at As with single-rotor turbines, the exit veloc-
a working temperature of 1800°F is an important ity from any row of blades (rotary or stationary)
criterion for selection. Other required properties is less than the inlet velocity, because of fric-
include low creep rate, oxidation and erosion tion losses. It can be assumed that the blade
resistance, and endurance under fluctuating velocity coefficient k b has the same value for
loads. Haynes Stellite, Vascojet, and Inconel X any row of blades:
are alloys frequently used. The rotor blades are
fabricated either by precision casting or by pre- V2 _Ca _V4
cision forging methods. Rotor disks are best kb- v 1 C: V 3 (6-145)
made of forgings for optimum strength.
In a multirotor turbine, the total work trans-
ferred is the sum of that of the individual rotors:
Design of Single-Stage, Two-Rotor Velocity-
Compounded Impulse Turbines (figs. 6-9, 6-55,
and 6-58)
In most impulse turbines, the number of rotors

is limited to two. It is assumed that in a single-
stage, two-rotor, velocity-compounded impulse
C]
turbine, expansion of the gases is completed in
Vl
the nozzle, and that no further pressure change NOZZLE
_lI
occurs during gas flow through the moving ROTA
blades. As mentioned earlier, the two-rotor, -- -_ .... FIRST ROTOR
velocity-compounded arrangement is best suited

for low-speed turbines. In this ease, the gases C2
ejected from the first rotor blades still possess

considerable kinetic energy. They are, there-
fore, redirected by a row of stationary blades
into a second row of rotor blades, where addi-
tional work is extracted from the gases, which
ROT.T,ONOX
"_'_ _/,,_,= SECOND ROTOR
usually leave the second rotor blade row at a

moderate velocity and in a direction close to the S_a4 _¥4 4
axial. U
The velocity diagrams of a single-stage, two- Figure 6-58.-Velocity diagrams o{ a typical

rotor, velocity-compounded impulse turbine are single-stage, two-rotor, velocity-compounded
shown in figure 6-58, based on the mean rotor impulse turbine.
diameter. The peripheral speed of the rotor
blades at this diameter is represented by U. The
gases leave the nozzles and enter the first rotor
Total work transferred to the blades of a two-
blades with an absolute velocity Cx, at an angle
a_ with the plane of rotation. V_ and V2 are the rotor turbine, ft-lb/lb of gas flow/see
relative flow velocities in ft/sec at the inlet and
outlet of the first rotor blades, The gases leave U
E2b=-_(Cl cos al +C2 cos a2
the first rotor blades and enter the stationary
blades at an absolute flow velocity C 2, and at +C a COS a3+C 4 C0S a4)
an angle a2. After passing over the stationary

blades, the gases depart and enter the second H
rotor blades at an absolute flow velocity C 3, and =g(V, cos/31 +V2 cos/32
at an angle as. V3 and V4 are the relative inlet
and outlet flow velocities at the second rotor +V 3 cos /3a +V4 cos /34) (6-146)
Combined nozzle and blade efficiencyof a two- Design of Two-Stage, Two-Rotor Pressure-
rotorturbine: Compounded Impulse Turbines (figs. 6-10, 6-14
and 6-59)
E2b
T]nb= JAH (6-147) An operational schematic of a typical two-
stage, two-rotor, pressure-compounded impulse
where turbine and its velocity diagrams at the mean
diameter are shown in figures 6-10 and 6-59.
AH =overall isentropicenthalpydrop of the
Each stage of a pressure-compounded impulse
turbinegases, Btu/Ib
turbine may be regarded as a single-stage im-
= totalavailableenergy contentof the tur-
pulse turbine rotating in its own individual hous-
bine gases (eq.6-17)
ing. Most of the design characteristics of a
Equation (6-137)can be rewrittenforthe tur-
single-stage turbine are applicable to the indi-
bine overallefficiencyqt of a two-rotorturbine
as vidual stages. The gas-spouting velocities Cz
and C3, at flow angles a, and a3, of the first-
and second-stage nozzles, are designed to be
7It
= T/nbT/m (6-148)
approximately the same. Vz, V 2, V 3, and V 4
represent the relative flow velocities at inlets
Ideally,tin
b is a maximum forthe single-
and outlets of the rotor blades, fiz, fi2, f13, and
stage,two-rotor,velocity-compoundedimpulse
f14 are the corresponding flow angles for Vz, V2,
turbinevelocityratio
V 3, and V 4. The second-stage nozzles are de-
signed to receive the gas flow discharged from
U cos a I
the first-stage rotor blades at an absolute veloc-
C, 4
ity C2, and to turn it efficiently to a desired
angle a 3. Simultaneously, the gases are accel-
i.e.,when U= ¼C,t. The workload forthe sec-
erated to a desired velocity C3, through expan-
ond rotorof a two-rotor,velocity-compounded sion to a lower pressure. The flow at the outlet
turbineis designed at about one-fourthof the
of the second rotor has an absolute velocity C4
totalwork.
and a flow angle a4. U is the rotor peripheral
speed at the mean effective diameter din.
The totalwork performedin the turbineis the
The design procedures for the gas flow pas-
sum of thatof the separate stages. These may
sages of the rotor and stationary blades of a be designed to divide the load equally (i.e.,the
single-stage, two-rotor turbine are exactly the
same as those for a single-rotor turbine. How-
ever, velocities and angles of flow change with
each row of blades. As a result, the radial FIRST STAGE
height of symmetrical blades increases with each C1
row, roughly as shown in figure 6-55. The V1 01
effects of reheating (increase of gas specific "_l FIRST STAGE
ROTOR
NOZZLE
volume) in the flow passages must be taken into
account when calculating the gas densities at _ _' '= 2 SECOND STAGE
various sections. Equation (6-136) may be used U _ NOZZLE
to estimate the amount of reheat at each row of

V3 a 3
blades. Also see sample calculation (6-11) and
C3 SECOND STAGE
figure 6-60 for additional detail. ROTOR
In the calculations for multirow unsymmetrical

blades, the radial heights at the exit side of
each row are determined first by equation (6-140).
The radial heights at the blade inlets are then Figure 6-59.-Velocity diagrams of a typical two-
made slightly larger, approximately 8 percent, stage, two-rotor, pressure-compounded impulse
than those at the exit of the preceding row. turbine.
velocity diagrams of each stage are identical or where

Cl=C 3, C2=C 4, al=a 3, a2=a 4, etc.). The T2t = turbine gas total (stagnation) temper-
friction losses occurring in the first stage is ature at second-stage nozzle inlet,
passed on in the gas stream as additional en- °R
thalpy and increases the available energy for the T 2 = turbine gas static temperature at
second stage. Also, the kinetic energy of the second-stage nozzle inlet, °R
gases leaving the first stage is largely used and P2 t = turbine gas total pressure at second-
not entirely lost as with a single-stage turbine. stage nozzle inlet, psia
The carryover ratio rc, i.e., the ratio of the P2 = turbine gas static pressure at second-
kinetic energy actually utilized as inlet energy stage nozzle inlet, psia
by the second-stage nozzles to the total kinetic C2 =absolute gas flow velocity at first-
energy of the gases leaving the first stage, can stage rotor blade outlet, ft/sec
vary from 0.4 to close to unity. The axial dis- C3 =gas-spouting velocity at second-stage
tance between the first-stage rotor and the nozzle exit, ft/sec
second-stage nozzle, as well as the leakages rc = second-stage carryover ratio of kinetic
through the sealing diaphragm between stages, energy
should be minimized for optimum carryover. Cp = turbine gas specific heat at constant
The determination of the right enthalpy drop pressure, Btu/lb-deg F
resulting in equal work for each stage may re- y =turbine gas specific heat ratio
quire a trial-and-error approach, in view of the AH2_ 3, =isentropic enthalpy drop of the gases
effects of reheating. Or, the proper enthalpy flowing through the second-stage
drop may be estimated from previous designs and nozzles due to expansion, Btu/lb
test data. With the velocity coefficients for noz- (Ant) 2 =required total second-stage nozzle
zles and blades given by past or concurrent area, in 2
experiments, equations (6-122) and (6-136) can k_ : nozzle velocity coefficient
be used to estimate the amount of reheating. ent :nozzle throat area coefficient
Most equations established for the single-
stage turbines may be employed in the design
calculations for two-stage turbines. The follow-
ing additional correlations are available for the From sample calculation (6-5), the following
design of second stage nozzles: data have been obtained for the turbine of the
A-1 stage engine turbopump.
C2 2
Turbine gas mixture ratio, LO2/RP-1 : 0.408
T2 t = T2 + rc _ gJCp (6-149) Turbine gas specific heat at constant pres-
sure, Cp=0.653 Btu/lb-deg F
Y Turbine gas specific heat ratio, y= 1.124
Turbine gas constant, R = 53.6 ft/°R
p2( T2t_ )'-1 (6-150)
P2t: k_] Gas total temperature at turbine inlet, T o
= 1860OR
Gas total pressure at turbine inlet, Po = 640
psia
C3=kn gJCpT2t _ P3 Y Gas static pressure at turbine exhaust, pc=27
psia
= kn ,(_-cC22 + 2 gJAH2_ s, (6-151) Total available energy content of the turbine

gases, AH = 359 Btu/lb
Turbine gas flow rate, _i,_= 92 lb/sec
(Ant)2 = (6-152) Turbine shaft speed, N= 7000 rpm
l Y+l
Overall turbine efficiency (when using velocity-
/ r27 _-zT
compounded wheels), _t= 58.2 percent
4/ ,,'LT-qj In addition, the following design data are set
forth:
L-
Nozzle aspect ratio = 9.7 Solution

Nozzle velocity coefficient, kn = 0.96
(a._) Single-stage, two-rotor, velocity-
Nozzle throat area coefficient, ent: 0.97
compounded impulse turbine.
Nozzle exit area coefficient, ene = 0.95
A representative velocity diagram for this
Rotor and stator blade velocity coefficient,
turbine is shown in figure 6-58. Figure 6-60
kb=0.89
represents the temperature-entropy-enthalpy dia-
Rotor and stator blade exit area coefficient,
gram for the gas processes involved in the oper-
_b2 =0.95
ation of this turbine. The following subscripts
Chord length of rotor and stator blades,
denote the various points and processes listed:
Cb=l.4 in
Partition thickness at the exit of nozzles and
blades, tn = tb = 0.05 in
0, 1, 2, 3, 4 = Points representing inlet condi-
Solidity of first rotor blades = 1.82 tions at the nozzles; first rotor blades;
Solidity of stator blades = 1.94
stator blades; second rotor blades; and the
Solidity of second rotor blades = 1.67 exit conditions of the second rotor blades.
1', 2', 3', 4' =Points representing exit condi-
(a_) Determine the velocity diagrams and prin- tions at the nozzles; first rotor blades;
cipal dimensions of the single-stage, two-rotor,
stator blades; and second rotor blades, for
velocity-compounded, impulse-type turbine for
an ideal isentropic expansion process.
the A-1 stage engine turbopump, with about 6 0-1', 1-2', 2-3', 3-4' =Path of an ideal
percent reaction in rotor and stator blades down-
isentropic expansion process in the noz-
stream of the nozzles.
zles; first rotor blades; stator blades; and
(b_) Determine the velocity diagrams of an second rotor blades.
alternate two-stage, two-rotor, pressure- 0-1, 1-2, 2-3, 3-4=Path of actual processes
compounded, impulse-type turbine for the A-1 in the nozzles; first rotor blades; stator
stage engine turbopump, with equal work in each blades; and second rotor blades.
stage and about 3 percent reaction in the rotor 1'-1, 2'-2, 3'-3, 4'-4 = Differences along con-
blades downstream of the nozzles of each stage.
stant pressure lines, between ideal isen-
tropic expansion processes and actual
processes, due to friction losses and re-
a_ PO
heating in the nozzles, first rotor blades,
stator blades, and second rotor blades
0 '_k_- CONSTANT
To
A, /_PR_LFE UNES
s I Point " O"-Nozzle Inlet

:2: T O = nozzle inlet total temperature : turbine inlet
>_- total temperature = 1S60°R
/" II _..,"_,'_.,% / p4 o.
Po = nozzle inlet total pressure = turbine inlet

-r
........... -._-_/ b== total pressure = 640 psia
AH = overall isentropic enthalpy drop of the tur-
bine gases = total available energy content
of the turbine gases = 359 Btu/lb
T/n = nozzle efficiency = kn 2 : (0.96) 2 = 0.92
Point "1 "-Nozzle exit= First Rotor Blade Inlet

ENTROPY, S
Since about 6 percent of the overall isentropic
Figure 6-60.-Temperature-entropy-enthalpy dia- enthalpy drop AH is assumed to occur in the
gram o[ the gas processes in a single-stage, rotor and stator blades, the isentropic enthalpy
two-rotor, velocity-compounded impulse turbine drop in the nozzles
with small amount of reactions downstream o[
the nozzles. 'AH o-1' : A H (1 - 0.06) -- 359 x 0.94 : 337.5 Btu/lb
We can write: Ideally, the efficiency r/n b of a two-rotor,

velocity-compounded impulse turbine is a maxi-
mum when the turbine velocity ratio
AHo_p = CpT o I 1-(P-21U y_,]

\Po] U COS a 1
C1- 4
From this, the gas static pressure at the noz-
zle exit
From this, the peripheral speed at the mean diam-
eter of the rotor
COS a I
U=C1_=3940xCOS 425 °
Pl =Po CpToJ
1.124 =3940 × 0.226 = 890 fps

337 5 ]o.124
=640× E1 0.65-3x-_s60.J From equation (1-129), the turbine rotor mean
diameter
= 640 x (0.722) 9._ = 640 x 0.053 = 33.94 psia
dm =720 U_0_720 × 89___- 29.1 in

From equation (6-121), the gas spouting 7r N 7rx7000
velocity at the nozzle exit
From equation (1-130), the relative gas flow
CI = kn V2 gJAHo_l, = 0.96 _/16.9 × 106 = 3940 fps angle/3, at the inlet to the first rotor blade can
be calculated:
From equation (6-122), the amount of reheat
in the nozzles
tanfll=Cl C,cosal-U
sinai _ 3940×0.906-
3940x0.42390=0.622
(1 - kn 2) C12 _ 0.08 × 15 524 000 /31 =31053 '

qnr = = 27 Btu/lb
kn22gJ 0.92× 64.4 x 778
Referring to figure 6-58, the relative gas flow
Referring to figure 6-60, the gas temperature velocity at the first rotor blade inlet
at the nozzle exit, following an isentropic ex-
pansion C 1 sin a1._3940× sin 25 °
sin fll sin 31°53 '
AHo_,, 337.5
_ 3940 x 0.423
TI, ZTo Cp -1860 0.653-1344°R 3156 fps
O. 528
The actual gas static temperature at the

Point "2"-First Rotor Blade Exit= Stator Blade
nozzle exit
Inlet
Assume that the given 6 percent reaction
27 downstream of the nozzles is equally divided
T1 = TI' + Cp -- 1344 +0.---_
qn---sr = 1385° R
between the two rotors and the stator. Then the
isentropic enthalpy drop in the first rotor blade
The gas density at the nozzle exit can be approximated as
P, 144 33.94x 144

Pl =-_-T x--_ -1385.4 × 53.6 = 0.0658 lb/ft 3 AH,_2, =P-_ x 359 = 7.18 Btu/lb
We will use an angle a_ of 25 ° for the Using equation (6-135), the relative gas flow
spouting-gas-flow direction at the nozzle exit. velocity at the exit of the first rotor blades
V 2 = V'kb 2V l2 + 2 gJTlnAH l_2, V 2 sin/32 2866×sin 25 °

tan a 2 = V cos /32 - U 2866 × cos 25 °- 890 = 0.707
= V/(0.89 × 3156) 2 + 64.4 × 778 × 0.92 × 7.18
a 2 =35°15 '
= 2866 fps
The absolute flow velocity at the first rotor

From equation (6-136), the amount of reheat blade exit
in the first rotor blades,
V2 c2-V 2 sin /32 _ 2866 × sin 252 1210 2080fps

2 1 sin a 2 sin 35o15 ' =0.57---_=
qbrl =(1 - kb )_-_+ (1 - r/n) AHI_ 2,
Point "3"-Stator Blade Exit : Second Rotor
(3156)2
= [1- (0-89)2] x 64.4 x 778 _(1-099)×7.18
• , Blade Inlet
The isentropic enthalpy drop in the stator
= 41.975 Btu/lb blades
The static gas pressure at the first rotor AH2_ 3, = AH 1-2' = 7.18 Btu/lb
blade exit
Analogous to equation (6-135), the absolute
gas flow velocity at the stator blade inlets
Y
F A., C a = x/kb2C22 +2gJ_nAH2_ a,
= X/(0.89 × 2080) 2 + 64.4 × 778 × 0.92 × 7.18
=33.94x 1 0.653×1385 = 1938 fps
= 33.94 x 0.93 = 31.6 psia Reheat in the stator blades
, 2_ 622
The gas static temperature at the exit of the qbs =(l-Kb )2--_+ (1- rln)
AH2_ 2
first rotor blade row following an isentropic
(Analogous to eq. (6-136))
expansion
= [i- (o.89) 2] × (2080)2 + (1 - 0.92) × 7.18

T 2, = T 1 - AH,_2,/C p = 1385 - 7.18/0.653= 1374 ° R 64.4 × 778
= 18.53 Btu/lb
The actual static gas temperature at the first
rotor blade row exit The static gas pressure at the stator blade
exits
qbr2 41.975
--- 1374-+ - 1438 ° R
T2 = T2' + Cp - 0.653 Y
7.18 _9.06
Gas density at the first rotor blade exit p3 =p2 [1 -CDT2
_H2-21
j Y-1 =31.6x E1 0.653×1438J
=29.42 psia
144p2 144x31.6
P2 = RT 2 - _ 1--_-_-8
= 0.059 lb/ft 3
Gas static temperature at the stator blade
exits following an isentropic expansion
We use an angle/32 of 25 ° for the relative gas
flow direction at the first rotor blade exits (un- T3, = T2 -AH2_2/Cp= 1438-7.18/0.653= 1427 ° R
symmetrical blades). The absolute flow angle
a 2 at the first rotor blade exits can be calculated Actual staticgas temperatureat the stator
from blade exits
qbs 18.53 }r
T3 = T3'+-_--= 1427 + x-_--_.
= 1456° R
r =v
Gas density at the stator blade exit
= 29.42 x 1 - 0.6537.1s
x 1456] -1_°_
144 P3 _ 144 x 29.42
.= 0.0544 lb/ft 3
P3 = R T 3 53.6x1456 = 27.46 psia > 27 psia (Pc)
We use an angle a3 of 35 ° for the absolute P4 is slightly higher than the turbine exit pres-
gas flow direction at the stator blade exit (a s sure (underexpansion), because of the reheating
Ta2). The relative flow angle t33 at the stator effects.
blade exit can be calculated from
C 3 sin a s 1938x 0.574 The gas static temperature at the second

= 1.596
tan/33-C3 cos a3- U=1938x0.819-890 rotor blade exits following an isentropic expan-
sion
f13 = 57°56'
7.18
The relative flow velocity at the stator blade T4' --T3 -AH3__,/C p = 1456--- - 1445 Btu/lb
0.653 -
exit
The actual gas static temperature at the sec-

V3 = C s sin a 3 _ 1938 × 0.574 - 1312 fps ond rotor blade exit
sin /33 0,847
Point "4"-Second Rotor Blade Exit T4 -- T4 ' qbr2 1445 ' 7.73
+ C--_---= +_= 1457 ° R
The isentropic enthalpy drop in the second
rotor blades
Gas density at the second rotor blade exits
AH___,= AHa_f --7.18 Btu/Ib

144 P4 _ 144 x 27.46
= 0.0506 lb/ft a
P4 - RT 4 53.6 x 1457
The relative gas flow velocity at the second
rotor blade exit
We use an angle/34 of 44 ° for the relative gas
flow direction at the second rotor blade exits
g 4 = \/kb 2 V3 2 + 2 gJ77nAHa.. 4,
(unsymmetrical blades). The absolute flow angle
= V'(0.89 × 1312) 2 + 64.4 x 778x 0.92 x 7.18 u 4 at the second rotor blade exits can be calcu-
lated from
= 1306 fps
The amount of reheat in the second rotor V 4 sin /34 1306 × 0.695
blades tan a 4 =y4 cos /34- U=1306x0-719-890 -18"5
a 4 =86°55 '
,,y 2
qbr2 = (1 - kb2)-_gj+ (1- 9n) AH_-4' The absolute flow velocity at the second rotor
blade exits
(1312) 2 , ,.
= [1 - (0.89):] x 64--A x 778" Ll - 0.92) x 7.18
C4 =V 4 sin/34_ 1306x0'695=908 fps
= 7.73 Btu/sec sin a 4 0.9985
Nozzle Dimensions
Gas static pressure at the second rotor blade From equation (6-123), the required total noz-
exit zle throat area
wt First Rotor Blade Dimensions (at dm)

Ant =
/ y+l The pitch or blade spacing
r 2 i?-_
gyL iJ Blade chord length Cb 1.4
= 0.769 in
Pbrl = Blade solidity - 1.82
92
From equation (6-140a), the number of blades
0.97 x 640 ll/32.2 x 1.124(0.94) lTns
53.6 x 1860
,,dm ,Tx 29.1
= 13.22 in2 Zbrl = Pbr 1 0.769 -119
We use a radial height hnt of 1.5 inches atthe

Allow 2°7 ' between inlet blade angle _)b_ri and
nozzle throat. Thus the nozzle width at the
inlet relative flow angle fl_; thus
throat
Obxrl --_1 +2°7' = 31°53' +2°7' =3't°

hnt 1.5
bnt =Nozzle aspect ratio - 9.7 - 0.1548 in
Make exit bladeangle Ob2rl equal toexit rela-
tive flow angle /_2
The number of nozzles
Ob2rl =_2 =25°

Ant 13.22
Z/l= - _
bnthnt 0.1548× 1.5 57
We select a blade radial height at the inlet
Pitch or nozzle spacing hb,rl =hne (1 x 0.08)= 1.64 x 1.08= 1.77 in
Pn - ,din _ _ x 29.1 = 1.604 in The blade passage width at the inlet

zn 57
bblrl = Pbrl sin Oblrl - tb = 0.769 x 0.559- 0.05
We allow 2 ° between nozzle exit angle On and
= 0.379 in
nozzle spouting-gas flow angle al; thus
On=a1 -2=25-2=23 ° From equation (6-138), the required total

blade exit area
From equation (6-125), the required total noz-

zle exit area, 144 w_ 144 x 92
= --
Ab2rl p2V2Eb2 0.059 x 2866 x 0.95 82.5 in 2
144 w_ 144 x 92
Ane plClene 0.0658×3940×0.95 53'75in2 Combining equations (1-139) and (1-140a), we
obtain the blade radial height at the exit
Combining equations (6-125) and (6-126), we
obtain radial height and width at the nozzle exit: Ab2rl
hb2rl-_dm sin Ob2rl-Zbtb
Ane 53.75
82.5
hne = rrd m sin On - Zntn - rrx29.1x 0.391- 57 x 0.05 = 2.52 in
x 29.1 x 0.423 - 119 x 0.05
= 1.64 in
The blade passage width at the exit
Ane
zn 53.75 bb2rl = Pbrl sin Ob2rl - to = 0.769 x 0,443 - 0.05
bne = hne - 57 x 1.6_- 0.576 in = 0.291 in

The mean blade radial height Using equation (6-138), we obtain the required
total blade exit area
1.77+2.52
hbr, - 2 - 2.145 in 144 w_ 144 x 92
Ab2 s = P3 C3eb2 0.0544 x 1938 x 0.95 = 132.5 in s
Assume a tapered blade with shroud, and that
it is subject to approximately the same tensile Combining equations (6-139) and (6-140a), we
stresses from centrifugal forces, as would be a calculate the blade radial height at the exit
uniform blade without shroud. The blades shall
be made of Timken alloy, with a density pb =0.3 Ab2s
lb/in _. Cheek the centrifugal tensile stresses hb2s=rtdm sin Ob2s-Zbstb
at the root section using equation (6-141).
132.5
- _x 29.1 x 0.574 - 127x 0.05 -2.87 in
Scr I = O.O0045721pbhbrldm N2
The blade passage width at the exit
0.0004572 × _-0_-X,
× 2.145 × 29. i X (7000)2
bb2 s = Pbs x sin 062 s - Ib = 0.721 x 0.574- 0.05
= 13 050 psi = 0.364 in
Stator Blade Dimensions Second Rotor Blade Dimensions
Pitch or blade spacing Pitch or blade spacing
Blade chord length Cb 1.4 Blade chord length Cb 1.4
Pbs - Blade solidity - 1.94 - 0.721 in Pbr2 = Blade solidity 1.67 0.888 in
From equation (6-140a), the number of the

From equation (6-140a), the number of blades blades
ndm_ _× 29.1 vdm _ v×29.1=i09

- 127
Zbs = Pbs 0.721 Zbr2 - Pbr2 0.838
Allowing 2024 ' between inlet blade angle Allow 2o4 ' between the inlet blade angle
0b, s and inlet absolute flow angle a 2 0blr2 and the inlet relative flow angle f13; thus
Obls=a2 +2024 `= 34° 36 '+2o24 ' =37 ° Oblr2 =f13 +2_4' = 57_56' +2° 4' =60o
We hold exit.blade angle 0b2s equal to exit We make the exit blade angle 062r2 equal to
absolute flow angle a3: the exit relative flow angle /94
8b2s=a3 =35 ° Ob2r2 =/_4 =44°
From equation (6-149), blade radial height at From equation (6-149), the blade radial height
the inlet at the inlet is
hb,s= l.08×2.52=2.72 in hbir 2 = 1.08× 2.87= 3.10 in
The blade passage width at the inlet The blade passage width at the inlet
bbls= Pbs sin Ob,s-tb=O.721xO.602-O.05 bb 1r2 = Pbr2 sin Ob l r2 - tb = 0.838 x 0.866- 0.05
=0.384 in = 0.677 in
DESIGN OF TURBOPUMP PROPELLANT-FEED SYSTF.j_,S 255
From equation (6-138), the required total _: 0.582

blade exit area
A-1 Stage Engine Turbine (Single-Stage, Two-

144 #t 144 × 92
Rotor, Velocity-Compounded Impulse Type)
Ab2r2 -P4 V4eb2 - 0.0506 x 1306 x 0.95 = 211 in 2
Design Summary
For velocity diagrams at mean diameter din,
Combining equations (1-139) and (1-140a), we
see figure 6-58.
obtain the blade radial height at the exit
U = 890:
a 1=25°; fl_=31°53'; C_=3940fps; V1
Absr2
=3156fps; a:=35°15'; fl_=25:; C 2=2080
hbsr2-_d m sin Obsr2-Zbtb
fps; V 2=2866 fps; a3=35°; fl3=57°56';
211 C 3 =1938 fps; V 3 =1312 fps; a4 =86°55';
=_x29.1 x 0.695- 119× 0.05 =3"66 in _4=44_; C4:908fps; V4=1306 fps
Isentropic enthalpy drops:
The blade exit passage width Nozzles, AHo_ I,= 337.5 gtu/lb
First rotor blades, AHI_ 2, = 7.18 Btu/lb
bbsr2 = Pbr2 sin Obsr: - tb = 0.838 x 0.695- 0.05 Stator blades, AHs. _, = 7.18 Btu/lb
Second rotor, AH3. _, = 7.18 Btu/lb
= 0.533 in Total 5H = 359 Btu/lb
Working efficiencies:
The mean blade radial height
7/t=58.2%; ;7n=92%; _7nb=68.3%; qm=85.2%
Mean diameter of nozzles and blades:
3.10+ 3.66
dm= 29.1 in
hbr2 - 2 - 3.38 in
Nozzle dimensions (at din):
Aspect ratio = 9.7; z n = 57; Pn = 1.604 in;
Check the centrifugal tensile stress at the 0n=23°; hnt=l.5in; nne=l.64in; bnt
root section using equation (6-141) =0.1548 in; bne=0.576in
First rotor blade dimensions (at din):
Solidity=l.82; Cb=l.4in; Zbr_=llg;
Scr 2 = O.O0045721pbhbrsdmN 2 Pbrl =0.769 in; t)blrl =34°; Obsri =25¢;
bblrl = l.77 in; hb2rl =2.52 in; bblrl=0.379
= 0.0004572 x_x 3.38 × 29.1 x (7000): in; bbsrl =0.291 in
Stator blade dimensions (at din):
= 20 550 psi Solidity = 1.94; Cb = 1.4 in; Zbs = 1.27;
Pbs=0.721 in; 0bls=37°; 0b28=35°;
Turbine Efficiencies hbls=2.72in; bb:s=2.87in; bbls=0.384
From equations (6-146) and (6-147), the com- in; bb28=0.364 in
bined nozzle and blade efficiency Second rotor blade dimensions (at din):
Solidity = 1.67; Cb = 1.4 in; Zbr2 = 109;
U(C 1 cos a 1 +C: cos a2 Pbr2 =0.838 in; Oblr2 =600; 0b2t2 =44°;
÷C 3 cos a 3 +C 4 cos a4) hblr 2 =3.10 in; hbsb2 =3.66 in; bblrs=0.677
tinb -- in; bbsr2 =0.533 in
gJAH
(_b) Two-stage, two-rotor, equal-work, pressure-

890 (3940 × 0.906 + 2080 x 0.817
compounded impulse turbine. (For velocity dia-
+ 1938 x 0.819 + 908 × 0.055)
grams, see fig. 6-59.)
32.2 x 778x 359
From prior trial-and-error calculations, the
= 0. 683 following isentropic enthalpy drops resulting in
(approximately) equal work for each stage were
From equation (6-148), the turbine machine obtained. We assume a stage carryover ratio
efficiency rc = 0.91.
First-stage nozzles: Point "2"-First-Stage Rotor Blade Exit=Second-

Stage Nozzle Inlet
AHo_I, = 50%; AH = 0.5 × 359= 179.5 Btu/lb From equation (6-135), the relative gas flow
velocity at the first-stage rotor blade exit
First-stage rotor blades:
V 2 : \/kb 2V12 + 2 gJT?n'_H1_ 2,
AH__ 2, = 3%; AH : 0.03 × 359 = 10.75 Btu/lb
: V_(0.89 × 1784) 2 + 64.4 × 778 × 0.92 × 10.75
Second-stage nozzles: = 1736 fps
AH2_ 3, = 44%; AH = 0.44 × 359 = 158 Btu/lb We chose a relative exit gas flow angle
/3 2 --38 ° for the first-stage rotor blades. The
Second-stage rotor blades: absolute gas flow angle, a2, can then be calcu-
lated as
AH3_4, = 3%; AH = 0.03 × 359 = 10.75 Btu/lb
V 2 sin f_2 1736 ×0.616
tan a s- i,,2 cos /32- U 1736 x 0.788- 1308 17.25
Point " O'-First-Stage Nozzle Inlet
T O-- 1860oR a2= 86040 '
Po = 640 psia
The absolute gas flow velocity at the first-
stage rotor blade exits
Point "1 "-Firs$-Stage Nozzle Exit = Rotor Blade
Inlet
V 2 sinfl2 1736×0.616
From equation (6-121), the gas-spoutingveloc-
C2 : sin a2 0.998 - 1070 fps
ity at the first-stage nozzle exit
C 1 = kn \/2 gJAHo_ f = 0.96 × 223.8 × vJl-7-9_5.5

Point "3"-Second-Stage Nozzle Exit = Second
= 2880 fps Rotor Blade Inlet
From equation (6-151), the second-stage noz-
We use a value of 25 ° for the spouting-gas zle gas-spouting velocity
flow angle a l. For optimum efficiency, tile

peripheral speed at the rotor mean diameter C a = k n \/rcC22 + 2 gJAH2_ 3'
= 0.96 V'0.91 × (1070) 2 + 64.4 × 778 x 158
-- cosalC1 -- 0.906×2880 _ 1308fps

2 2 : 2880 fps
Since C 3 = C_, the remainder of the second-

Using equation (6-130), the relative gas flow
stage velocity diagram is the same as that of the
/3_ at the first-stage rotor blade inlet can be
calculated as first stage, i.e., a3:a I =25°; J33 =]31 =43°8';
V3=Vl=1784 fps; a4:a2=86°40'; /_4=/_2=38°;
C 4:C 2=1070fps; V 4=V_=1736fps.
C, sin al 2880×0.423
tan /31 = C1 cos al- U - 2880 × 0.906 - 1308 = 0.936
From equation (6-129), the turbine rotor mean
131 =43°8 ' diameter
The relative gas flow velocity at first-stage

rotor blade inlet dm=720U-720x1308 42.7 in
_N _r × 7000
Yl C1 sinai 2880×0.423 1784 fps From equation (6-147), the combined nozzle
sin 131 0.683 and blade efficiency
U(C x cos a_ +C 2 cos a 2 result of forces which act on the shaft or on the
+C 3 cos a3+C 4 cos a4) parts supported by the shaft. These forces may
r/nb = g JAil be divided into two classes: those which act at
right angles to the shaft axis (radial forces), and
1308 (2880 × 0.906 + 1070 x 0.058
those which act parallel to the shaft axis (thrus:
+ 2880 × 0.906 + 1070 × 0.058)
loads).
32.2 × 778 × 359
=0.78 Radial loads on turbopump bearings may re-

sult from one or more of the following sources:
The turbine machine efficiency is assumed to (1) Weights of parts such as shafts, pump
be the same as that used in design (a._): impellers, turbinerotors, gears
(2) Centrifugal forces du_:, to unbalance of
qm = 0.852 these rotating parts
(3) Forces due to inertia, resulting from rapid
From equation (6-148), the overall turbine acceleration
efficiency (4) Resultant radial forces on the impeller
due to nonuniform pressure distribution
r/¢ = rlnbr/m = 0.78 x 0.852 = 0.664
in the discharge volute of the pump
(5) Tangential or torque forces induced by the
A-I Stage Engine Alternate Turbine Design
gears
Summary (Two-Stage, Two-Rotor, Pressure-
Compounded, Impulse Type)
Thrust loads on turbopump bearings may re-
For velocity diagrams at mean diameter din,
sult from one or more of the following sources:
see figure 6-59.
(1) Weight of rotating parts mounted on a
U = 1308 fps:
vertical shaft
a1=25°; _1=43°8 '' C1=2880fps; V 1=1784
(2) Unbalanced axial thrust of the pumps
fps; a_=86°40'; _2=38°; C2=1070 fps;
(3) Resultant axial thrust on the turbine rotor
V 2 = 1736 fps; a3 =25°; f13 =43°8'; C3=2880
blades
fps; Va=1784 fps; a4=86°40'; _4=38°;
C4=1070fps; V 4=1736fps
Isentropic enthalpy drops: For the turbopumps of liquid rocket engines,
First-stage nozzles, AH__,, = 179.5 Btu/lb high-speed ball and roller bearings are used
First-stage rotor blades, AH,_ 2, = 10.75 Btu/lb almost exclusively. A typical two-bearing de-
Second-stage nozzles, AH2_2 -- 158 Btu/lb sign is shown in figure 6-7. A ball bearing
Second-stage rotor blades, AHz._4, = 10.75 carries both radial and thrust loads. It is paired
Btu/lb with a roller bearing which carries only radial
Working efficiencies: loads, however, of a higher magnitude. A typi-
r/t= 66.4%; r/n = 92%; rlnb= 78%; _?m= 85.2% cal three-bearing arrangement is shown in figure
Mean diameter of nozzles and blades: 6-63. The shaft radial loads are carried by a
dm= 42.7 in single roller bearing at the turbine end and by a
Comment: The overall efficiency of the roller and a ball bearing on the pump side. The
pressure compounded turbine is higher than that ball bearing also absorbs the thrust loads. As a
of design (a). However, a relatively large dm is rule, the shaft thrust loads in a turbopump are
required (weight, size). carried by a single or dual bearing located at
one end of the shaft. Thus loads from thermal
expansion or contraction of the shaft are avoided.
6.6 DESIGN OF TURBOPUMP BEARINGS,
Bearing design data with regard to load-
SEALS, AND GEARS
carrying capacity, operating speed, and service
life are usually furnished by the manufacturers.
Turbopump Bearing Design
The useful life of a bearing is dependent upon
A turbopump shaft is supported by two or more its speed and load, and may be expressed by the
bearings. The loads on the bearings are the correlation:
- lira _
proportionally with the increase of shaft design

{ Rated speed, rpm ._
Life, hours = gb \Actu-a--1o--p-er_p-_'-d, rpm/ speed. Thus the required bearing DN value
rapidly increases for high-speed turbopumps. As
[ Rated capacity, ib ,_4 a result, especially for liquid hydrogen applica-
×_Actual working load, lt)] (6-153)
tion, the turbopump rpm is often determined by
the DN limits of the bearings.
where Kb = design factor usually furnished by
A most important bearing design consideration
manufacturer.
is the expected operating life of the rocket en-
If a bearing is subjected to both thrust and
gine. The bearings must have adequate statisti-
radial loads, the two can be combined into a
cal probability of conforming with this require-
single equivalent radial load:
ment. A generally accepted life rating for ball
and roller bearings is the "B-10 life." The term
P = R + xA (6-154)
denotes the operating life (hours) of a _popula-
tion" of bearings at a given load and speed, at
where the expiration of which statistically 10 percent
P = equivalent radial load used for bearing of them will have failed. Of course, in actual
selection, lb rocket engine operation, component reliability
R = actual radial load, lb must be much higher. Bearing life at a given
A = actual thrust load, lb load and speed varies inversely with reliability.
x = design coefficient usually furnished by For instance, the B-1 life (99 percent reliability)
manufacturer is one-tenth of the B-10 life (90 percent relia-
Rocket turbopump bearings quite commonly bility), or one-fiftieth of the B-50 life (50 percent
are cooled and lubricated by the propellants reliability). Therefore, turbopump bearings are
pumped. They are usually operated at very high generally designed for a B-10 life of at least 100
"DN" values, a parameter which is the product hours. This corresponds to a B-1 life of 10
of the bearing bore D (millimeters), and the bear- hours, or a B-0.1 life (99.9 percent reliability) of
ing rotative speed N (rpm). 1 hour, the latter by order of magnitude being the
Propellant lubrication has the advantage of life the bearing most likely will actually see.
eliminating an additional lubricant supply sys- For critical applications, an even higher life
tem, and of simplifying bearing sealing problems. rating may be selected.
The following are important design considera- Figure 6-61 presents the centrifugal load DN
tions for propellant-lubricated bearings: limits in terms of 10, 100, and 1000 hours of
(1) Characteristics of the propellants, such B-10 life for a typical ball bearing design (extra-
as thermal stability, operating temper- light series).
ature, chemical inertness, viscosity. The stress-limiting DN values of roller bear-
(2) Compatibility of the bearing materials ings are much higher than for ball bearings;
with the propellants. The application however, it is extremely difficult to control the
of certain high-strength alloys is some- temperature rise in a roller bearing, if the DN
times limited by the propellants used. value is above 1.5 × 106, due to excessive cage
The "DN" rating is convenient when selecting slip. Generally, rocket turbopump bearings have
high-speed ball or roller bearings. As the rota- been successfully operated at DN values up to
tire speed of a bearing increases, contact fatigue 1.5 × 106. Limited test information indicates
of the outer race caused by centrifugal loads of possible satisfactory operation at 2.0 x 106 DN.
the balls or rollers may cause failures. In addi-
tion, bearing contact speeds will result in non- Dynamic Seal Design
rolling phenomena with attendant failures caused
by overheating. Through proper selection of the The principal dynamic, i.e., rotating seal
bearing geometry, these problems can be mini- types used in liquid rocket turbopumps are the
mized, and the DN rating increased. Note that labyrinth, face-riding, and shaft-riding seals.
for a given horsepower rating, the shaft size Satisfactory seal operation depends upon good
based on allowable stress does not decrease design which considers many factors, including
_.:1• [
I
I
I I II
I [ I [
JBALL BEA
=,
NG C
. TRt
N IFUGAL LOAD
1LIMIT
1
r
¢ol I J I I ] L,_,TSE=,ESS,ZEI !
(EXTRA
&o i
I _.ou_-
e
II - I I
It
l [
II IOUTE_
../---TST'0,
=.cE
_*FO*=,T.
T--SPEC, .58
.55
o $O MRS
i J I '
J I
i i \-T-_ _ --_-\-q----
_.o#-------------J_/--\--4_,_. _ ._ I I I I -_ _ --- ' _.3u-=_.
",., .c__..c- 2:.':. ... 't, , I .....
Io ZO 3o 40 50 60 70 8090 _
o. 200 _0 400 =X>o
BEARING BORE SIZE, D mrn
Figure 6-61.-Bali bearing centrifugal load DN limits.
fluid pressure surges, vibration, expansion and p = density of the fluid lb/in 3
contraction of sealing components, contact pres- Cs = seal leakage coefficient, established
sure between sealing surfaces, rubbing veloci- experimentally
ties of the sealing surfaces, smooth and friction- Labyrinth seals are used for the wearing rings
free operation of internal sealing parts, and on of pump impellers, as well as for the rotating
squareness of the sealing surfaces. Any influ- seals attached to tim sealing diaphragm between
ence which directly or indirectly subsequently two turbine stages.
alters these factors can cause improper operation In a face-riding-type seal the sealing is
of the seal.
accomplished through rubbing contact between
As shown in figure 6-62, the labyrinth seal is the precision-lapped faces of a floating seal
a clearance-type seal. The fluid tending to pass washer and a shoulder ring. The mating faces
through the sealing interface is throttled many are at a right angle to the axis of rotation. As
times and is forced to follow a devious path.
shown in figure 6-62, the floatin£ seal washer is
The function of a labyrinth seal is not to prevent attached to a metal bellows. The bellows, in
fluid leakage entirely, but rather to reduce leak- turn, is welded to a stationary seal housing
age to a reasonable level at a minimum of fric- which is secured and statically sealed to the
tion and wear. The amount of leakage through a casing. The bellows provides flexibility and
labyrinth seal can be estimated by the correla- spring force to the contact face, permitting it to
tion follow axial and angular movement without leak-
age. Sometimes a lip seal is used in conjunction
Qe=CsAcV'24g Aps/p (6-155)
with a spring-loaded floating washer, instead of
where a bellows.
Qe =leakage rate, in3/sec The shaft-riding seal (fig. 6-62) consists of a

Ac =seal clearance area, in 2 seal housing, a seal retaining plate, and several
Aps = pressure differential across the seal, seal segments. The segments form a ring around
lb/in 2 the shaft and are held against it by garter springs.
-PUMP CASING
HOUSING
WEARING RING ....,..\\_ /-- PUMP IMPELLER
___ _SHAFT
-_-_- CC SHAFT
LABYRINTH SEALS
_ ,.AL RETAINING PLATE
/-_ TURBOPUMP
_-TURBOPUMP CASING /CASING
FLOATING .\'..x>,_" P- SE;TL HO_SING
SEAL WASHER_\_\_/ST ICSE L ____GARTER SPR,NGS
_HER_/SEAL HOUSING
SHOULDER RING_ BELLOWS
....._/,/_
_ ..... SEGMENTS
_ ___cu-_L_L - -- _ SHAFT
FACE-RIDING SEAL SHAFT- RIDING SEAL
Figure 6-62.-Principal turbopump dynamic seal types.
Thus self-adjusted dynamic sealing is accom- low-speed inducer. Gear arrangement and geom-
plished between shaft outside diameter and seg- etry depend upon power transmitted, propellants,
ment inside diameter. Axially, the segments are speed ratio, and other factors. During operation,
forced against a flat surface of the seal housing the gears are cooled and lubricated with oil, or
by a retaining plate and a spring washer, thus with the propellant being pumped.
providing a static seal. Shaft-riding seals tend The gears are usually housed in an aluminum
to occupy less space than the face-riding seals. casing. To minimize weight further, webs are
A wide variety of materials is available for held as thin as possible as are cross sections at
floating seal washers and seal ring segments. the rim and hub. The hubs are often internally
Carbon is used most frequently. The rubbing splined for best results. Spur gears are most
seal faces on shoulder ring or shaft must be widely used, since they minimize thrust on bear-
hardened or plated, and lapped to a very smooth ings. Tooth loads and speeds in turbopump
finish. The seal face rubbing speed should not gears are very high. The designer, therefore,
exceed 300 fps. Frequently, vent or purge lines must achieve high tooth strength and high re-
are connected to the cavities between two or sistance to wear. Turbopump gears are usually
more dynamic seals installed in series. This made of high-alloy steel, with the tooth surfaces
assures positive sealing for critical applications hardened by either case carburizing or induction
such as interpropellant seals. hardening. If possible, the tooth surface should
be accurately finished by a grinding process.
Materials and dimensional tolerances of turbo-
Turbopump Gear Design
pump gears must be held under very close control
The gear trains used in liquid rocket turbo- during manufacturing.
pumps (fig. 6-16) afford speed differentials be- To improve gear life and load-carrying stabil-
tween turbine, pumps and accessory drives, and ity, certain modifications to standard design
also sometimes between a pump impeller and a practices can be applied. Pinions are frequently
DESIGN OF TURBOPUMP PROPELLANT.FEED SYSTEMS 261
made with long addendum and gears with short considerations in preparing the layout. For
addendum to adjust tip-sliding velocities and to instance, one of the more important criteria which
strengthen the pinion. Furthermore, pinion tooth influences the selection or arranging of the turbo-
thicknesses are often increased, at the expense pump mechanical elements is the ease of devel-
of gear tooth thickness. High pressure angles opment. Standard or proven mechanical detail
as high as 22_ ° ,25 _j,or 271/:
° may be applied to should be extensively adopted in the layouts.
reduce contact stresses on filetooth surface and The following is a list of important turbopump
to increase the width of the tooth at the base, design layout considerations:
Involute-profile modifications are often also (1) Compatibility with engine systems pack-
made to compensate for bending and to keep the aging and plumbing
tips from cutting the mating part. (2) Structural integrity
(3) Positive interpropellant sealing
(4) Compensation for thermal expansion and
6.7 DESIGN LAYOUT OF TURBOPUMP
contraction
ASSEMBLIES
(5) Ease of development
Figure 6-63 presents the design layout of (6) Ease of assembly
the A-1 stage engine turbopump assembly. Logi- (7) Ease of manufacturing
cal packaging and arranging of the basic mechan- Considerable experience and skill are required
ical elements of tim turbopump are among the in turbopump design layout work for best results.
iNLET
PUMP i _-FUEL PUMP
PUMP VOLUTE AND
VOLUTE AND DISCHARGE
DISCHARGE MANIFOLD
i
FUEL PUMP -TURBINE NOZZLES
ROW ROTATING
BLADES
IMPELLER-_
OXIDIZER
PUMP
ROTATION /
OXIDIZER
PUMP
ROTATING
OXIDIZER
BLADES
PUMP INDUCER
SEAL
IMPELLER
HYDRAULIC
3TATIONARY
PUMP IMPELLER
BLAOES
AUXILIARY POWER
P1CKUP SHAFT
Figure 6-63.-Assembly design layout of the hypothetical A-1 stage engine turbopump.
Chapter VII
Design of Controls and Valves
7.1 CONTROL METHODS opening and closing the propellant valves. Minor
deviations from the design mixture ratio or pro-
The foremost design requirements for any
pellant flow rates, such as from fabrication
control system are accuracy, stability, and reli-
tolerances of engine components, are corrected
ability. Two basic control methods are avail-
beforehand by insertion of accurately sized
able: open-loop (no feedback) and closed-loop
orifices into the propellant flow lines to effect
(feedback) control systems. Both have found
the desired pressure drops (also see ch. II). The
wide application in liquid propellant rocket pro-
extent of correction is determined from systems
pulsion systems. Open-loop control is confined
preflight calibration test data. Open-loop control
to those systems which are designed to operate
has the advantage of simplicity. However, it is
at a fixed, steady-state level over a narrow range
limited to a specific set of operating parameters,
of environmental conditions. Most other applica-
and is unable to compensate for variable condi-
tions require one of the many forms of closed-
tions during operation.
loop control. For these, mathematical models
Accurate sequencing of an open-loop control
can be constructed with which the functions and
system such as is used for engine start and stop
dynamics, such as gain factors and stability of a
is usually accomplished with the aid of inter-
proposed system, can be analyzed.
locks. Mechanical interlocks are preferred for
The selection of the best-suited method then
their high reliability. For instance, the propel-
is an important firststep in control systems
lant valves of many small engines or gas gener-
design. It will be influenced by the required
ators are mechanically linked and are operated
accuracy, the dynamic characteristics of the
by a single actuator. Proper sequencing between
system being controlled, and by allowable time-
fuel and oxidizer valves is achieved by adjusting
lags. Once the method is determined, the basic
the relative positions of the valve gates or pop-
elements for the proposed system must be se-
pets, with respect to the mechanical linkage.
lected, such as type of components of the power
Interlock requirements can also be furnished by
supply or working fluid (electric,hydraulic, or
other means (electric, hydraulic, or pneumatic).
pneumatic), and of the operating mechanism for
In high-thrust engines, sequencing between main
the specific control, which all depend on the
propellant valves and ignition system is often
specific application. Ideally, the basic theories
accomplished by the combination of various
and past experience should permit design without
interlock designs. A typical example is the A-1
experiment or development work. However, some
stage engine, for which the start and stop se-
development work with attendant redesign will
quence and their interlocks were described in
always be required in the process of perfecting a
detail in chapter III.
new system.
Open-Loop Control Closed-Loop Control
With this system, control is accomplished by Closed-loop control is also called automatic
preset control means, such as orifices, and on- or feedback control. This system usually in-
off command devices. A typical example of cludes sensing means, computing means to detect
open-loop control is an engine propellant flow errors, and control means to correct them. An
system, calibrated to a fixed set of conditions. accurately sensed feedback is compared with a
The propellant flows are controlled simply by fixed or variable reference by a computer, which
263
_mm I= mw _
then generates signals to correct for any devia-

tions. The main system thus does not require
precise calibration for a specific set of condi-
tions. Unlike open-loop control, closed-loop
control depends on sensing absence or presence
of an error to maintain a desired condition or to
Figure 7-1.-Schematic of a typical closed-loop
bring about a correction. In general, the objec-
control system.
tive of closed-loop control is to minimize errors
during operation and reduce system sensitivity (electric summing junction and amplifier), and a
to environmental changes and changes in compo- controller (gas generator flow control valve), the
nent characteristics. It is applied to areas such command reference input r is compared with the
as engine-thrust control and/or throttling, pro- sensor feedback b. The controller then manipu-
pellant mixture-ratio control, and thrust-vector lates • g in response to an error signal e from
control. the computer. Ideally, r should be in linear pro-
For rocket engine application, closed-loop portion to Pr and b to Pc, save for constants
control systems usually employ one or a combi- required to convert one physical quantity into the
nation of the following modes of operation: other. However, this ideal condition is difficult
1. Simple "on" and "off" type.-(Example: to attain because of the dynamic characteristics
pressure switch/valve combination for tank pres- of the pressure transducers. These characteris-
sure control.) tics are influenced by physical properties such
2. Proportional type.-Employs a continuous as mass inertia, fluid compressibility and vis-
control signal which is proportional to the error. cosity, and frictional resistance. Instead of r
(Example: transducer output for chamber pres- being directly proportional to Pr, the two param-
sure control.) eters are actually related through a differential
3. Derivative type.-Employs a continuous equation which represents the dynamic behavior
control signal which is a function of the error of the elements involved. The same is true for
and its time derivative(s) (rate of change). This the feedback b and the controlled variable Pc. It
is principally used when systems stability is is also applicable to other systems components.
critical. (Example: thrust vector control system Hence, the analysis of a closed-loop control
with phase lead.) system usually involves the solution of sets of
4. Integral type.-Employs a continuous sig- often complicated differential equations.
nal which is proportional to the cumulative inte- Refer again to figure 7-1, where Pc is the
gral of one or more errors. (Example: two flow- controlled variable, _¢g the manipulated variable,
meter outputs for mixture-ratio control.) e the error signal, b the feedback, r the reference
Closed-loop or feedback control systems are input, and Pr the desired value. A, B, C, and D
essentially dynamic systems. Their design symbolically represent the dynamic relation
characteristics may be analyzed according to the between input and output of the respective com-
basic laws of physics. Figure 7-1 shows a typi- ponents. The following terms representative of
cal example. Its function is to maintain the the differential equations for this closed-loop
controlled variable Pc equal to the desired value control system can be written:
Pr, by manipulating the variable wg. Maintaining
Pc equal to Pr is assumed to maintain the in- r =Apt e =r- b Pc =C_vg
directly controlled quantity F. In a typical (7-1)
turbopump fed engine control system, Pc would b = Dpc ¢¢g = Be
be the combustion chamber pressure, maintained
equal to a fixed reference pressure Pr by means The solution of these equations in combina-
of a valve controlling the gas generator propel- tion with a systematic experimental program will
lant flow wg. F then would be engine thrust, suffice to analyze the dynamic performance of
which is indirectly maintained at a desired value. the system.
In this control system which consists of a The continuous corrective action of a closed-
sensor (chamber pressure transducer), a computer loop control system may promote dangerously
DESIGN OF CONTROLS AND VALVES 265
unstable operation when control elements or Engine System Cutoff Control

components are employed having high gain and
Rapid and safe engine shutdown, during nor-
significant response lags. An unstable control
mal operation as well as in an emergency, is
system is one that is no longer effective in main-
desirable for minimum and repeatable cutoff
taining a variable at its desired value. Instead,
impulse, and to enhance reliable systems opera-
large divergent oscillations may set in. The
tion. The cutoff sequence usually consists of
requirements for control accuracy and for stabil-
shutoff of subsystems power (gas generator,
ity are often difficult to combine. Higher accu-
etc.); shutoff of main chamber power; and, in
racy requires high amplification; i.e., high gain.
case of test firings, postfiring securing (purges,
The high amplification results in overshoot dur-
flushes). As a rule, the propellant valve-closing
ing corrective action, thus promoting various
sequence is adjusted to provide a fuel-rich cut-
degrees of system instability. However, through
off in the main combustion chamber. This pre-
appropriate means of compensation, such as
vents damaging temperature spikes and results
"anticipatory" phase leads (time derivatives), it
in smooth and rapid thrust termination.
is possible to obtain a high gain control system
with satisfactory stability. Additional informa-
tion on compensation will be presented in con-
nection with thrust-vector control. Engine Main Stage Duration Control
Important consideratious governing engine

duration have been discussed in section 2.i.
7.2 BASIC LIQUID PROPELLANT ROCKET The signal for engine in-flightcutoff, unless it
ENGINE CONTROL SYSTEMS is the result of a malfunction, will be supplied
by the vehicle and fed directly into the cutoff
Most engine systems require several or all
control system discussed in a preceding para-
of the basic control systems summarized in the
following paragraphs. Typical applications are graph. For lower stages, where optimum utiliza-
tion of the propellants is desired, a tank low-
found in chapter Ill, for the A-I, A-2, A-3, and
A-4 propulsion systems (figs. 2-10, 3-3, 3-6, and level sensor is often employed. In final stages,
3-9). where precise cutoff velocity is essential, an

integrating aceelerometer or equivalent device
will signal cutoff.
Engine System Start Control
The prime objective of a start-sequence con-

Engine System Safety Controls
trol is to bring the engine system safely from
start signal to main-stage operation. A typical Special monitoring devices, such as combus-
sequence may consist of systems preconditioning tion stability monitors for detecting combustion
(purging, chilldown); application of start energy, instability, gas generator overtemperature sen--
if required (start tanks, turbine spinner); and sors, or turbopump overspeed trips, are frequently
introduction and ignition of the propellants in the employed to prevent undesired or unsafe condi-
main combustion chamber. Secondary sequences tions by effecting prompt, automatic, nonhazard-
may be required for certain subsystems such as ous, fail-safe system shutdown during all phases
the gas generator system. A reliable engine- of engine operation. In addition, most engine
start sequence is maintained through interlocks control systems are designed so that an interrup-
and by monitoring each functional step of engine tion of electrical power supply will cause the
operation during the start transient. The system to shutdown safely. (For certain mis-
propellant-valve opening sequence is set to sions it may be desirable to switch to an emer-
effect either an oxidizer-lead or a fuel-lead start. gency power source, or prevent shutdown by
This is usually dictated by propellant type and mec_Janical latching, for continued operation.)
chamber ignition and cooling methods. Figures Mechanical and electrical interlock devices are
2-11, 3-5, 3-8, and 3-11 present typical engine extensively used in the control system to assure
system start and cutoff sequences. the reliability of the safety control systems.
Propellant Tank Pressurization Control Most propellant tank pressurization control

systems are of the closed-loop type.
Various propellant tank pressurization sys-
tems have been discussed in chapter V. The
Engine System Control Calibration
design requirements for the control of these sys-
tems must consider- The systems described in the preceding para-
(1) Means to maintain the required tank pres- graphs require proper adjustment and calibration
sure level within an allowable range for desired engine operating characteristics and
during all phases of vehicle and engine performance. This includes the setting of timing
systems operation, including steady- devices, pressure switches, position switches,
state engine mainstage; start or dynamic and the sizing of orifices. The correct values
throttle transients; and vehicle coasting for each of these are verified during engine cali-
periods between restarts. bration and checkout firings. Of the orifices,
(2) Effective safety devices such as pressure some are placed in propellant lines for perform-
relief valves to prevent overpressuriza- ance parameter calibration. Others are used in
tion and rupture of the propellant tanks. pneumatic or hydraulic lines as timing and re-
(3) Compatibility with other subsystem con- stricting devices. Specific orifice applications
trols, such as propellant-utilization- for thrust and mixture ratio control will be dis-
control and thrust-control systems. cussed in sections 7.3 and 7.4; orifice design
Figure 7-2.-Control ori[ice locations and sizes o[ a typical engine system.

elements will be presented in section 7.10. lators, and with a minimum of calibration firings.
Following sizing, orifices must be properly However, "thrust" regulators or "controllers" are
identified, by stamping or banding, and their employed in vehicle systems which require a
actual sizes recorded in the engine logbook. higher degree of precision and repeatability,
Control and calibration orifice locations and such as in single-stage vehicles starting at sea
sizes of a typical system are shown in figure 7-2. level, or in final stages of a multistage system.
Thrust regulators are actually chamber pressure
regulators. At altitude (vacuum), their effect is
Engine Systems Checkout and Test Controls
identical to thrust regulation, since at altitude
To verify operational readiness of the engine thrust for a given engine and mixture ratio is
system and its subsystems, suitable control solely a function of chamber pressure. The same
means are required for postassembly and pre- is essentially true for systems starting at sea
firing checkouts. These permit simulation of the level, because the relationship of thrust to cham-
operation of the engine and its critical control ber pressure as a function of altitude is predict-
components, without actually firing the engine able with high accuracy.
system. Utilizing suitable ground-support equip- Occasionally, vehicle missions will require
ment (GSE), an engine checkout control system in-flight thrust control over a wider range.
should include- Usually, in such cases, the need is for a planned
(l) Provisions to conduct leak checks and reduction of thrust, or "throttling," during the
electrical-continuity checks of the entire last portion of propelled flight. Two basic pro-
engine system. cedures are possible:
(2) Provisions for verifying proper operation (a._) Stepwise reduction of chamber pressure,
of all instrumentation pickups, such as Pc
dc bus voltage and spark plug firing (b) Continuous reduction of Pc
monitors; open, closed, and continuous Each of them can be accomplished by control
position signals for valves; propellant of-
flowmeters; and pressure transducers. (1) Turbine power (in the case of turbopump
(3) Provisions for verifying the proper func- fed systems), through regulation of gas
tion and operating range of all control generator propellant flow rate or hot gas
devices and subsystems, such as flow flow rate (preferred method).
control valves, pressure regulators, and (2) Main propellant flow rate
thrust and mixture ratio control devices. (3) Variation of main tank pressures (in the
(4) Provisions to simulate vehicle signals for case of pressure-fed systems).
"cold" checkout of the engine system Additionally, in multiple (clustered) engine sys-
operating sequence, such as for start tems, stepwise thrust reduction can be effected
and cutoff. by shutoff of one or more engines of the sub-
In addition to tile checkout equipment, the engine systems.
ground-support system must include equipment to The example chosen in section 7.1 to illus-
permit control of static test firings. This often trate a closed-loop control system is typical for
requires additional instrumentation. a system effecting thrust control through turbine'
power variation. Figure 7-3 shows the thrust
7.3 ENGINE THRUST LEVEL CONTROL control system proposed for our A-4 stage engine,
which relies on main propellant flow variation.
The significance of the thrust level of a Here, the closed-loop control system operates on
liquid rocket engine (sea level or vacuum) has the principle of variable fluid resistances in the
been explained in section 2.1. It is usually main oxidizer and fuel feed lines to achieve
specified with a tolerance; for instance, "_+3%." propellant flow-rate modulation. Engine reaction
It is possible, with modern "fixed thrust level" is determined by sensing chamber pressure, the
engines, to guarantee this band with simple parameter most indicative of thrust level, and by
orifice calibrations in the various propellant comparing the feedback b with the command
subsystems of the engine, without resort to regu- reference pressure input r. Any resultant error e,
7.4 PROPELLANT-MIXTURE-RATIO AND

PROPELLANT-UTILIZATION CONTROL
The significance of propellant mixture ratio

C_lu_m and its control have been discussed in section
_¢_E
,NPUT
, _ o _mu T
2.1. The principal reasons for mixture-ratio
control are recalled:
Optimum engine performance (important)
Complete propellant utilization; i.e., mini-
mum residuals (most important)
Both goals are closely interrelated and essen-
[_ C_EC_m_ i,,_._,_ tially inseparable.
Figure 7-3.-Main-stage thrust throttle control
loop Ior the ,4-4 stage engine.
Open-Loop Mixture Ratio Control
following amplification and compensation as The simplest form of engine mixture ratio
required, is used to drive the thrust throttle con- control is obtained by the installation of properly
trol actuator of the main propellant control valves sized calibration orifices in the main propellant
in a direction which reduces the error. Ideally, lines. Acceleration effects during flight are
the system operates over the entire thrust throt- usually accurately predictable as a function of
tle range with minimal disturbances to other trajectory and flight time. Thus, simple averag-
critical engine parameters; in particular, the ing of flight mixture ratio and selection of the
propellant mixture ratio. In practice, these dis- corresponding orifice size reduces mixture ratio
turbances are not entirely avoidable, but can be deviations over the duration of flight to a level
minimized by maintaining a given resistance acceptable for optimum total propellant utiliza-
ratio between the two main propellant control tion in many missions.
valves throughout the control range. A most Open-loop mixture-ratio control can often be
reliable method toward this objective would be further refined by the following procedures:
mechanical coupling of the two propellant valves 1. Weighing ot the propellants loaded; i.e.,
(fig. 7-4). accurate determination oI the tanked propellant
Orifices, propellant valves, and servovalves mixture ratio.-The vehicle to be launched rests
required for thrust control will be described in on load cells, thus permitting weighing of the
subsequent chapters. propellants actually loaded. In mixed systems,
the noncryogenic component is loaded and
weighed first. The cryogenic component follows
ENGINE ST_.RT, CUTOFF {I TPiRUST
and is subsequently maintained at level through

FROM THROTTLE CONTROL _II£TL_TO_ 7 FI_:_I
OXIDIZER a topping line. The mass of both propellants is
TANK / FUEL
_'ANK
._, _ / i determined from on-the-spot temperature and

N OXIDIZER
T _ MECHAN ICdI, L
CO_ ROL LVE COOPLING ambient pressure readings while the tanking
procedure is progressing.
2. Use o[ adjustable, rather than [ixed, ori-
[ices in one or both propellant lines.-As close
to vehicle takeoff as possible, and as a function
of tanked weight and temperature readings, a
hand or remotely ground-controlled prestart-ori-
flee adjustment is made. This method is usually
TO NO. I TO NO Z TO NO. I TO NO Z confined to noncryogenic fluids.
THRUST CHAMBER THRUST CH*_*M_R THCtUST CHC_MBER THRUST CHAMBER
For systems where engine operation closely

Figure 7-4.-Schematic of the propellant control follows that obtained during final calibration,
system for A-4 stage engine start, cutoff, remarkable accuracy of targeted mixture ratio
throttle and mixture ratio control. and thus propellant utilization can be obtained
DESIGN OF CONTROLSAND VALVES _9
with the open-loop method, approaching that of a oxidizer flow in a similar manner. For instance,
closed-loop system (single stages; first stages). a propellant-utilization servo control valve,
In certain applications, however, the varia- which regulates the pneumatic pressure to the
tion of mixture ratio as a function of increasing main oxidizer valve actuator, may control the
acceleration may exceed tolerable limits. Accel- oxidizer flow by adjusting the angular position
eration in most vehicle tank arrangements affects of the oxidizer valve gate during engine main-
predominantly the propellant in the forward tank. stage operation. In certain applications it may
Because of the long supply line, acceleration be desirable to integrate the propellant flow
continues to act upon a relatively large fluid rates and to compare the masses consumed to
column, even near the end of powered flight (tank one another and to those tanked for optimum
depletion). By comparison, the effect on the propellant utilization.
fluid in the rear tank is often nearly completely It is readily seen that control systems, based
offset by the simultaneous decrease in fluid head on propellant flow-rate measurements, are a
(short liquid column). refinement of open-loop systems using fixed
To offset excessive acceleration effects on orifices. They are basically still mixture-ratio
the fluid from the forward tank and thus on mix- controls and thus merely "assume," but do not
ture ratio, head-suppression valves are some- measure directly, the amount of propellants actu-
times used at the pump inlet of turbopump fed ally remaining in the tanks and their unbalance.
systems. Here, pump inlet pressure increase is To accomplish this function, usually referred to
sensed as a function of acceleration. Corre- as "propellant utilization" (PU), additional con-
sponding signals are fed through a logic device trol elements must be employed in the form of
to the head-suppression valve which will gradu- vehicle tank-level sensors. Numerous principles
ally close, thus acting as a throttling device. are known: point sensing, sonar, acoustic, radi-
This valve also protects the pump structurally. ation sensing, differential pressure, and capaci-
tance probes.
Figure 7-6 presents the propellant utilization
Closed-Loop Mixture Ratio Control control system for the A-4 stage propulsion sys-
tem. The residual propellant quantities in the
In certain cases, such as in last stages, or in
missions requiring engine restart following ex- main tanks are continuously monitored, summed,
and compared with a PU control reference in the
tensive cruising periods involving propellant
boiloff, a closed-loop system may be required. propellant utilization control computer. Any
error detected is used to modify the command
In figure 7-5 we see the A-4 stage engine mixture
ratio control loop which operates on the basis of reference mixture ratio input, (MR)r, to tl_e
mixture-ratio control computer. This method
continuous propellant mass flow sensing. Both
fuel and oxidizer mass flow rates are monitored isolates the mixture ratio control from the pro-
and integrated to establish the ratio of either the pellant utilization control, and thus prevents
interaction between them. The bandwidth of the
propellants consumed or the propellants remain-
ing. The mixture ratio feedback, (MR)b, is then
compared with a command reference mixture ratio
input, (MR)r, in the propellant utilization control ilXTU_E _*TIO
computer. The resulting error signal, (MR)e, is

fed to the mixture ratio control oxidizer valve mXTU_f _A_
vernier position actuator, which forms a link in

the mechanical coupling between the two main
propellant control valves, as shown in figure 7-4.
The oxidizer flow rate is thus modified to elimi-
nate the error. In high-thrust turbopump-fed
[CEC_e,( SUVltNG
engine systems such as the A-2 stage engine, _UV.C'nON
_' eLecrllc lm_t.,1,1Ei
where the propellant valves are independently

actuated, the system propellant mixture ratio Figure 7-5.-Propellant mixture ratio control
control can be accomplished by varying the main loop for the A-4 stage engine.
mixture ratio to assure simultaneous propellant

depletion. However, by programing a mixture
ratio in favor of the heavier component during the
early portion of flight, and then switching it in
favor of the lighter one, the accelerated vehicle
mass is reduced faster. Also, mixture ratio may
°° ] be programed to provide a higher thrust level
during the steeper portion of a trajectory. This
provides a better thrust-to-weight ratio in the
ELECT't; AMPL=FI(R
presence of gravitation, with attendant velocity

Figure 7-6.-Propellant utilization control system increase benefits. These methods, possibly in
[or the A-4 stage propulsion system. combination, may substantially increase stage
payload capacity, since the effects of mixture
propellant utilization control system is made ratio on performance (Is) are usually small within
a reasonable range (see table 7-1). Optimization
narrow as compared to that of the mixture-ratio
can readily be made with the aid of an electronic
control system, because propellant residual
computer program. In a number of applications,
errors may be expected to develop slowly; i.e.,
initial tanking errors can be corrected over the programed mixture ratio control without PU con-
trol, i.e., open loop mixture ratio control with
entire duration of engine operation.
The sensors used in the vehicle tanks may PMR, may give best results, simultaneously
serve additional purposes. In combination with reducing complexity.
Valves suitable for mixture ratio control will
suitable ground equipment, they may permit an
be discussed in section 7.8.
automatically controlled loading, high-level limit-
ing and topping procedure. In static firings and
flight, they may serve as redundant low-level TABLE 7-1
sensors to initiate engine cutoff. For such a
Flow rates
complete system, the term "propellant manage- Mixture i
ment system" has come into increased usage. ratio, Thrust NPSH Is Oxi-
Apart from throttle valves placed in the main O/F Fuel dizer
propellant lines, bypass lines have been suc-
cessfully applied to vary mixture ratio. Here, a 3hange, percent +10 +11 +12 -1.3 +4 +14
-10 -11 -12 +I.3 -4 -14
line is tapped off the pump outlet and ducted
back to the pump inlet. A servo valve, possibly
supported by an orifice, can be varied so that
the bypass flow is adjusted from no flow to full
bypass flow.
General Design Considerations
The implementation of closed-loop propellant-
utilization control through mixture-ratio control The precision with which a desired mixture
is a major vehicle-to-engine interface area. The ratio is obtained or maintained is affected con-
requirements or criteria will usually be estab- siderably in open-loop systems, and to some
lished by the vehicle builder and/or user. Close extent in closed-loop systems, by the following:
coordination between engine and vehicle designer (1) Instrumentation accuracies (in particular,
is essential. flow and tank-level metering)
A closed-loop mixture-ratio- and propellant- (2) Machining tolerances of orifices
utilization-control system may not only be used (3) Operating tolerances of regulators
for accurate maintenance of a fixed mixture ratio (4) Temperature influences on orifices and
but it also has the potential for programed mix- regulators
ture ratio control (PMR). Here, the mixture ratio (5) Density tolerances of the propellants, as
is varied during flight, either continuously or in a function of temperature and of purity
steps. It must be kept in mind that the average (composition according to specifica-
mixture ratio still must be equal to the tanked tions; contamination and dilution)
DESIGN OF CONTROLSAND VALVES 271
(6) Acceleration effects during flight always a possibility and must be prevented by
(7) Propellant tank pressure deviations proper design and handling procedures. Many of
(8) Turbopump speed deviations these, such as cleaning procedures, will be
(9) Differences between fuel and oxidizer called out in the shop drawings. Furthermore,
pump characteristics as a function of the design, where applicable, will have to in-
speed clude filters, check valves, and suitable line
(10) Line resistance changes as a function of routing in order to prevent contamination and/or
temperature and for miscellaneous contact with incompatible materials. Note that
mechanical reasons some propellants may change their properties
(11) Temperature effects in rotating machinery merely as a function of time, such as hydrogen-
In the following we will discuss important peroxide, which loses its concentration due to
steps toward maintenance of high quality, and (very slow) decomposition (with attendant gas
toward further improvement in the listed areas, development), even if absolute cleanliness has
for highest accuracy of mixture-ratio (and been maintained. This affects design conditions
propellant-utilization) control. in addition to contamination considerations since
First, continued improvement of propellant proper venting devices must be provided. The
flow-metering devices is imperative. Here, latter, in turn, have to be designed in such a
turbine-type flowmeters have achieved a high way that no contaminants, including moisture,
degree of accuracy (conformance with truth) and can enter the propellant system.
precision (repeatability). Since mass flow rates delivered by pumps
The accurate calibration of these meters to and/or regulated by orifices will be a function of
most reliable standards is important. Since the fluid densities, mixture ratio may be affected
engine inlet pressures also affect the mixture accordingly. The densities, in turn, aside from
ratio, pressure measurements of the highest conformance with specifications, will be affected
reliability are equally necessary. Wherever pos- by temperature (noncryogenic fluids) or ambient
sible, the rocket engine design should include pressure; i.e., boiling point (cryogenic fluids).
vital metering and measuring elements from the To overcome these effects, it may be necessary
outset. Dynamic sensing devices, in particular to temperature-condition the propellants. This
flow meters, are drastically influenced by their may be done by heating or cooling. Or, it may
installation configuration. If these end organs, be accomplished by suitable storage, such as
following accurate calibration, remain with the shielding against solar radiation. For cryogenic
engine through its entire life cycle, including propellants, it is usually sufficient to keep the
flight, a maximum degree of accuracy is obtained. containers vented to atmosphere until immedi-
The design and machining of all calibration ately prior to use, since the possible changes of
orifices should closely follow accepted stand- atmospheric pressure at a given altitude can
ards (see section 7.10). Selection of suitable only introduce relatively minor temperature
materials to eliminate or at least to reduce to a changes. The designer, through a suitable oper-
minimum, temperature influences and corrosion, ating sequence (engine schematic) and through
is important. The design of orifice holders must provision of vent valves, recirculators, heaters,
prevent the possibility of incorrect (upside down) and other components, can minimize temperature
installation and of distortion of the orifices. effects.
Regulators, if any are used, must be designed The actuation of mixture ratio control devices
for highest accuracy and precision with particular affects the nominal engine performance param-
consideration of the medium to be controlled. eters. Depending on the type of engine, in par-
More detail will be presented in section 7.12. ticular its turbopump characteristics, these
The purity and composition of the better effects may be significant. In an actual case,
known propellants are regulated by official gov- the effects shown in table 7-1 were observed.
ernment specifications. The designer can expect It is clear that the vehicle thrust structure
that approved sources will deliver the propel- must be capable of absorbing the higher thrust
lants in conformance with these. However, sub- loads. Also, the vehicle tanks and their oper-
sequent contamination, dilution or alteration is ating pressures must be capable of meeting the
NPSH requirements for extreme mixture ratio ary coils by a passing magnet, or others? (The
excursions. Furthermore, chamber cooling may engine designer will be involved in this selec-
be affected. During sea-level testing, nozzles tion only if the generated signals affect engine
with high expansion area ratios may experience components.)
jet separation at the lower thrust levels (low Selection of the best-suimd electronic control
Pc), resulting in vibration, destructive to engine system.-This will be largely influenced by sen-
as well as vehicle structure. sors and control-valve selections.
Since vehicles are tanked for their nominal A propellant utilization system is not a mal-
mixture ratio, and since engines are calibrated function prevention system. It does not add to
to this ratio, mixture-ratio valve excursions vehicle reliability, possibly subtracts from it.
should be small for vehicles which are expended Rather, it is a system required to live with a
within a few minutes after takeoff. For stages, marginal preliminary vehicle design. It is a safe
with long cruising periods prior to operation or assumption, however, that the first flights of a
reignition, and which use one or two cryogenic new vehicle will not be for its ultimate mission.
propellants, boiloff may have altered the ratio of PU, therefore, will not be a vital necessity for
the propellants in the tanks to such a degree these flights.
that the PU system may be called upon to oper- Thus, enough time is available to thoroughly
ate at or near its maximum excursion. It is, investigate, analyze, select, and develop the PU
therefore, vital that the engine designer appraise system. This time should be utilized. Both
the vehicle builder of all performance variations engine and vehicle builder have facilities and
as a result of mixture-ratio adjustments, beyond test programs to permit mutual exposure of their
the standard tolerances of the nominal perform- selected systems to flight and simulated-flight
ance values. Also, engine turbopumps must be environment.
capable of operating for extended periods with
the valve in either extreme position.
7.5 THRUST-VECTOR CONTROL
A propellant utilization system is a complex
system. If required, it must be of the highest To steer a vehicle over its trajectory, thrust-
quality. Otherwise, it will do more harm than vector control is applied. The following methods
good. Only closest cooperation between vehicle have found application:
and engine designer will assure optimum quality. (1) Gimbaled thrust chamber or engine assem-
Areas of particular significance to teamwork are: bly (widely used)
Selection of the mixture ratiocontrol method.- (2) Jet vanes (obsolescent)
For instance, should the PU system be active (3) Jetevator
during the entire flight duration, or only for the (4) Gimbaled thrust chamber nozzle (rare with
last, say, 30 percent. (Both methods have been liquid propellants)
successfully used.) (5) Secondary injection (into the thrust cham-
Selection of the mixture ratio control valve ber)
specilications.-Should it be a variable orifice, (6) Auxiliary jets
or a bypass valve? What should be the permissi- The first method is used most frequently, due
ble pressure drops, required response rates, and to its inherent reliability and performance. The
accuracies? In case of sensor failure, should first four systems require actuators which may
the valve return to the neutral position or remain be operated by hydraulic, pneumatic, or electric
in its last working position? (Self-locking.) means. The remaining systems are controlled by
Selection of the sensors.-Should it be one of flow regulation.
several available continously reading types,
such as capacitance gages or differential pres-
Thrust Vector Control Systems Using Actuators
sure (tank top to bottom) gages? Or should point
sensors be employed, such as hot wires (change Figure 7-7 presents a simplified schematic
of heat loss as a function of being immersed in for a thrust vector control system, employing
fluid or exposed); switches triggered magneti- hydraulic or pneumatic actuators. It may serve
cally by floats; voltage pips induced in station- to explain the fundamentals of closed-loop thrust
DESIGNOF CONTROLSAND VALVES 273
toward the same end. Malfunction safety circuits

are included to effect engine cutoff in the event
V_HOJ
C_OANCE
CO_MANO of erratic operation.
_FE_ENC_
A typical schematic for a thrust vector con-

trol system using electromechanical actuators is
shown in figure 7-8. Here, the actuator is pow-
II =° ered by a continuously
metal bidirectional
operating,
clutches.
constant-
speed, 28 volt de motor, fitted with dry-powder
The control com-
puter consists of summing junctions and an am-
(_ ELeCTRiC
#U_N
9,JMMPNG
plifier as in the case of hydraulic actuators.
_,.ECT RIC _PLIFJER
The dc motor drives the actuator through the
bidirectional clutches which are controlled by
Figure 7-7.-Typical schematic o{ a thrust vector
the error signal generated through comparing
control system using hydraulic or pneumatic
guidance command reference input with systems
actuators.
position feedback. To provide adequate systems
damping, the actuating speed is sensed by a rate
vector control, even though the systems used in generator or through differentiation of the posi-
practice may differ significantly in detail. The tion signal.
actuators are controlled by commands, originating Apart from electrical feedback and compensa-
in the vehicle guidance system, which are a tion systems, mechanical feedback systems
function of the vehicle's deviations from a pre- coupled with hydromechanical compensation
scribed path and of its response to corrective "networks" are coming into increased usage.
steering action. These signals are fed through They are inherently simpler and thus offer higher
an electronic thrust vector control logic to servo reliability. Two basic types of hydromechanical
valves. In the system shown in figure 7-7, each compensating devices may be distinguished:
servovalve modulates the fluid flow to its re- piston-bypass devices and load-pressure-sensing
spective actuator assembly in response to an devices.
electrical error signal which is proportional to Piston-bypass devices utilize leakages past
the difference between desired actuator position the actuator piston to introduce system damping
and its actual position. Feedback of the actual and may make use of dynamic relationships to
position is obtained through a transducer attached control time constants (a hole drilled through the
to the actuator. Additionally, tim actuating piston is an example).
speed is sensed by a rate transducer and applied Load-pressure-sensing devices, commonly
tO the control computer to stabilize the closed- called "pressure feedback" (PQ) valves or
loop control through adequate damping. Instead "derivative pressure feedback" (DPQ) valves,
of a rate transducer, electronic differentiation of are widely used.
the position transducer output may be applied
Figure 7-9 shows a typical servovalve and
actuator schematic with derivative pressure
feedback (DPQ) and mechanical feedback. The
only electrical signal required is the input to the
"torque motor" (an electromagnetic actuator)
resulting in deflection of the flapper of a differ-
ential valve, which drains to the sump. If the
flapper is deflected, as indicated in figure 7-9
by the arrow, nozzle flow on side B decreases,
r.lc _m,¢ lU_W,.G
with an attendant pressure rise. The reverse is
true for side A. The resulting pressure differen-
Figure 7-8.-Typical schematic for a thrust vec- tial forces the power-stage spool to the left,
tor control system using electromechanical blocking the return line on side B, and opening
actuators. it on side A. As a result, pressure Pb increases,
N"I TO_OU[ MOTOe

and transfer functions by substituting the equiv-
alent hydromechanical time constants.
Table 7-2 may be found valuable by those
NOZZLE • B
D4ERJVATIV_ PlES$_RE _TE_ - _IECH

who wish to familiarize themselves with some
FEEDBACK LEAF SPRING FEEDBACX
fundamentals in this field. Detail on the design
DERIV&TIVE LO_D DIRECTNDN O$"
PE'E$ SURE P_STON FEEDBACK of servovalves will be found in section 7.11.
IS,
OL AT _0_ P'_ST_ON i_OT _',J
_ I, IEc:,I _&
FEEDB, ACI(
_WER ST_ _'OOL S_4mNG
Demonstration Example
RETURN -- m[TUml
Two basic types of electrical compensation
O_IFICE ORF_.J[ networks exist: current output for voltage input,
_'N.TER
and voltage output for voltage input networks.

Figure 7-10 shows a simple form of a current
output for voltage input network. Find the anal-
ogous hydromechanical network.
Figure 7-9.-DPQ valve wiLh actuator.
and Pa decreases, forcing the actuator piston to Solution

the left to apply the desired load force. Attached The transfer function for the electrical net-
to the actuator piston rod is a tapered extension work is
which acts upon the mechanical feedback link-
age, including a roller and a spring. The me- i I+RCS
chanical feedback attaches to the torque motor. --:-- Amp/volt
V R
The pivot point of the valve flapper becomes the
where
error torque summing junction. Note that the
i = electrical current (amps)
nozzle jets also have a feedback effect. The
time derivative of the actuator motion, i.e., the V = voltage (volts)
R : resistance (ohms)
hydromechanical compensation, is obtained
C : capacitance (farads)
through action of a derivative load pressure pis-
S : LaPlace transform operator (= j_o for sinus-
ton. This piston is affected by the same pres-
oidal forcing functions)
sure differential that acts upon the actuator
From table 7-2, we obtain the equivalent
piston; i.e., by the load pressure. However, by
hydromechanical parameters for i, V, R, and C.
inserting an isolation piston and permitting flow
The new transfer function then is
through an orifice bypassing the derivative pres-
sure piston, the pressures affecting the latter
A 2
can equalize. The degree of this effect is a
function of the actuator pressure differential and Q _1+ C_xPK)S
its rate of change and of the bypass orifice size AP 1
(shock absorber principle). As seen in figure Cx
7-9, the derivative load pressure piston acts
upon the valve flapper when displaced. Thus it
provides the required time derivative of the
actuator motion for compensation.
As has been seen, it is possible to provide
compensation in thrust-vector-control systems by o
either electrical or hydromechanical means, the V
1
latter now being often preferred for actuators.
Conceivably, other control systems could be
converted from electrical to hydromechanical
G
networks. The analogies between the differen-
tial equations of the two network types often Figure 7-10.-Current output for voltage input
permit the use of existing electrical networks network.
DESIGN OF: CONTROLS AND VALVES 275
TABLE 7-2.-Electrical-Hydromechanical Component Analogies
Electrical quantity Describing Hydromechanical Describing

Analogy Remarks
or component equation component or quantity equation
V, voltage drop, volts ......... ' V = V A - VB AP, pressure drop, psi AP: PA - PB V _AP
i, current coulombs/sec ....... dq Q, flow, in3/sec dV i_Q

_ =-_ Q = d--t
V: volume
q = coulomb
charge
dv
_I_------- V -------_ .... [_---- A p ----_ Q (Ap)2d(,_p) Ap 2 ! Massless piston as-
i=C_- = -g -_ C _ --_ sumed
--.._Q
Capacitor, farads
Ap : piston area, in 2
K :spring constant, lb/in
V Parabolic flow curve

_=_-
Ap Q_Cx_,A-'_-_ R _ 1._
linearized about
CSIIE3 _CxA P CX
operation point
Q ---...,lllb
Resistor. ohms inS/sec

Orifice, --
Piston mass not

V -----_ .... ,, p AP2 y(AP)dt b
Q:"M- "Ap" negligible
t
------D- Q
Inductor. henries Ap= piston area, in 2

M =piston mass
Source: D. h. Engels, "A Method of Synthesizing Electro-Mechanical Compensation Networks for Hydraulic Servo-
mechanisms," Proceedings of the IEEE, PTGAC, October 1964.
The correct hydromechanical network, which The significance of good thrust alinement can
is of the piston-bypass type, is shown in figure be seen from the fact that in an engine cluster,
7-11. at the randomly distributed maximum of these
tolerances, a trim deflection of close to 0.5 °
would be required from all engines to offset the

Engine-to-Vehicle Interfaces With Actuator
misalinement.
Systems
For larger (looser) alinement tolerances, the
trim deflection would be further increased. Even
Engine Installation and Alinement if the trim deflections seem to reduce effective
For minimum demands on the vehicle guid- thrust and guidance capability only slightly, the
ance and engine actuation systems, it is required need to apply them for the full duration of pow-
that the engine thrust vector be properly pre- ered flight results in appreciable payload re-
alined with respect to the vehicle attachment ductions.
point in all three planes. Typical specified It is customary to aline the engine thrust
tolerances are: t0.25 inch laterally, _*0.5 ° verti- vector to the upper face of the gimbal bearing
cally. prior to shipment• Both optical and dynamic
Q
_C-_ /--VECHICLE THRUST MOUNT CENTERLINE
//,4i7/// /
_7 ._// / /-_..,_ cE.TE_ ,
-/- ,'.t-_.._q
EAANCDH
yIA_/ Pt:CAHE ) F_ __ _ GIMBAL "ARING [
P_llel to T_IJ_ v_14m
\ \ \ Actu_ L.el_gl.h
:A • B * COITe_I_
_tCCD
\ \ \
),4 Cx
\.\\ -, \
AP I " "
_--,.,UST
,,c,o, /l\
I,=,
Figure 7-12.-Engine alignment.
"o
"_ x'x- installation of the engine then simply consists

of attaching it, observing the engine logbook
S
specifications. Figure 7-12 shows installation
methods of a prealined engine into the vehicle.
For the first vehicles of a new type produced, it
_L_ is advisable to specify verification of engine
Figure 7-11.-Piston bypass hydromechanical alinement following transportation to the launch-

network. ing site.
methods (load cells) are used. Tile optical or Actuators, Loads

cold alinement establishes the geometrical loca- Actuators are usually of the hydraulic-piston
tion of the thrust vector in the shop, through type. Hydraulic-rotary, electromechanical, and
finding the centers of nozzle throat and nozzle pneumatic turbine-driven types have also been
exit, and alining their connecting line perpen- investigated.
dicular to the gimbal plane. A simple plumb Engine gimbal actuators are attached to the
attached to the injector center may be used in engine at one end, and to the vehicle at the
support of this operation. Subsequently, during other (fig. 7-13). They may be procured by the
engine firing, this measurement may be confirmed vehicle builder or by the engine builder, and
dynamically using side load cells in lieu of gim- must be properly dimensioned. The attachment
bal actuators. As a rule, after a few engines points at either end must be capable of absorb-
have been alined in this manner, experience will ing the forces encountered with an adequate
permit meeting specification by optical means reserve. As a rule, two actuators are required
alone. The vertical alinement can be simply for each engine. Together, they permit deflec-
documented as the eye-to-eye distance of the tion of the engine in all directions. It is impor-
actuator attach points, or as the line through tant to note that if the maximum deflection
two index points (fig. 7-12). Lateral dimensions effected by an individual actuator, for instance,
can be marked in a suitable manner at or on the is 7 ° , the combined maximum angle ("corner
gimbal bearing face. If the mating vehicle face deflection") of a pair is approximately 10 °. Inlet
had been properly alined to the vehicle axes, ducts, flex lines, gimbal bearing, and possibly
where filling of the nozzle takes a longer time

,_-_ ACTUATOR_
-- -- ..... /-" LONG STROKE during buildup than with shorter ones. Unless
\
AOTUA.OR
--\
SHORT
".CON,,OORAT,O
STROKE
these loads _-j_ ACTUATOR
can be eliminated or at least re-
duced, they must be considered for the design of
R ,\,.\\ actuators and attach points at both ends (notifi-
cation of vehicle builder).
Crosstalk and Spring Rate

Since engine and vehicle designer are not
- - I I entirely free regarding actuator installation, a

OlMBAL CENTER_ I I --I [i 1 I situation may exist wherein motion of the actu-
ator in one plane affects the other actuator in its
Figure 7-13.-Engine actuator installations.
plane. This is referred to as "crosstalk." If it
is excessive, control instabilities may develop.
other components affected, must be able to Therefore, close coordination between engine
"take" this deflection. If their capability is designer and vehicle builder to minimize cross-
limited, proper gimbal restriction (stops or snub- talk is essential.
bers) must be provided (circular gimbal pattern The actuator must be able to translate its
instead of square). motion without delay into engine deflection. If
Selection and design of the actuators is based the control loop formed by actuator, engine struc-
on the gimbal forces required. In a typical case, ture, and vehicle thrust structure is "soft," i.e.,
the actuator force may be 25 percent of the en- if it has a low spring rate (lb/in), the engine
gine thrust level. The force is determined by does not react promptly to an actuator motion as
considering the following: called for by the guidance system. A delay with
Inlet duct reactions subsequent overshoot and continued oscillation
Flexible service line reactions may result. The natural frequency of this oscil-
Gimbal bearing friction lation is a function of actuator stroke per degree
Heat shield reaction (if any) engine deflection, feedback gains, and compen-
Correction for misalinements sation network parameters.
Aerodynamic loading (if any) Figure 7-13 shows two typical engine actuator
Vehicle acceleration effects installations. The short-stroke configuration has
Inertia of gimbaled mass the advantage of high spring rate and high stiff-
Miscellaneous minor effects ness, and of compactness. However, the effec-
It is stressed that the engine design should tive gimbaled mass is high, requiring larger
reduce these forces to a minimum for smallest actuators and a more powerful hydraulic system.
size and lightest weight of hydraulic pump, Resolution is low (gain, expressed in degrees
actuators, and associated equipment. Recogni- deflection per inch actuator stroke, is high).
tion of this need and careful design can do much The long-stroke design results in low effective
toward this goal. The system must be capable gimbaled mass and reduces actuator size and
of stable, well-damped response when cold gim- hydraulic system dimensions. However, the
baled, such as during prelaunch checkout, even spring rate is low and so is the stiffness. More-
though the loads encountered here may be quite over, the arrangement is bulky and requires extra
different from those occurring during engine fir- structural members. Only a detailed design
ing. This dual-load situation may pose serious analysis conducted jointly by engine and vehicle
problems. designer will determine which configuration is
During startup of the engine, brief peak side best for the flight system.
loads in excess of those occurring during normal

gimbaling can be generated by the thrust chamber Hydraulic System
itself. This is especially true for high expan- Until other means are available, a hydraulic
sion area nozzles being developed at sea level system is probably required to power the engine
(for engines designed for altitude operation), actuators. Its basic elements are:
Hydraulic pump
Reservoir (low pressure, or "sump")
Accumulator (high pressure)
Servovalve
Actuators
Feedback (electric or mechanical)
Lines, check valves, filters, connectors,
instrumentation
If continuous hydraulic power is required

prior to engine start, such as for recirculation of
the hydraulic fluid or for gimbal tests, an elec-
trically driven auxiliary pump is also provided. Figure 7-I4.-Engine actuation system schematic
In most instances, the auxiliary pump will be (hydraulic).
operated until vehicle liftoff only, and can,
therefore, be ground powered. For upper stages,
the accumulator will then provide, for a limited
time, the hydraulic power required during staging
1 1 ......
PulIP SuCI_N '--'=-'
and turbopump buildup following its unlocking.
Since some of these components will be part
of the engine system, while others are stage
...................o....
mounted, an important vehicle/engine interface , II OUI_-'K Lq$CONN|C_._It|$ $ SUPPI.y
exists. Through an auxiliary drive shaft, the

main hydraulic pump may be driven from the ;._
j ll] : .......................
HIGH-PrESS RELIEF VALVE
13 LON-F'IF|S IIELIEF VALVE

engine turbopump. It is connected to the other _
i
12
14
OUICK D_SCONNECI-[LOW
NITIINEN CHA|GING
PI|SSI
VALVE
hydraulic equipment and to the actuator through 15

I_
I_IIMAL
ILEED VALVE
SWITCH
high-pressure lines, several of which must be _10
11
PIEr $$ TRA NSD_JCEJl
TEIIIIp TI l N _'_DU_EEII
flexible. These other elements may be mounted xt 4 33 IESEIIVOII Po£JrlON TRANSDUCEI
on the vehicle at the expense of longer lines

Figure 7"15.-Accumulator-reservoir schematic.
which also must cross the gimbal plane and must
therefore be flexible. Or, they may be engine
mounted. This, however, increases the engine PRESS. SUPPLY
gimbaled mass and may pose space and envelope RETURN
PUMP SUCTION
problems. To compensate for misalinements and EXTEND _1
RETRACT
thermal expansion and contraction, a certain 9. FILEEI
amount of flexibility must be provided for the 16 BLEED VALVE
17 SERVO-ACTUATOR
lines even in this case. It is possible to con- II. PREFILTRAIION VALVE
nect an electric generator to the main turbopump, 19 SAMPLING VALVE
20 DIFFERENTIAL PRESS. INDICATOR

and drive electrically a stage-mounted hydraulic 21 . CYLINDER BYPASS VALVE
pump. Only electrical wires will then cross the 22. SERVO VALVE
gimbal plane, with the exception of the hydraulic 23 SNUBBER
24. PISTON BYPASS VALVE
lines to the actuators which always must be 2S. LOCKING MECHANI_d4
flexible. Another possible simplification is to 26 MECHANICAL LOCK
26 30. PRESS. TRANSDUCER

combine servo valve and actuator into one single
31. [_:.W. TRANSDUCER
unit. Figure 7-14 shows a typical hydraulic 32. FEEDBACK TRANSDUCER
engine actuation system. Figures 7-15 and 7-16

may serve to identify the major components of Figure 7-16.-Servoactuator schematic.
this system.
From the above it becomes apparent that
numerous hydraulic connections will have to be must be long enough and of the proper pressure
made when installing the engine into the vehicle. rating, and, above all, must have a mating part
All of them must fit, and permit adequate flexure, on the vehicle.
Furthermore, the designer must know and

consider the environment in which the hydraulic
system will perform. The narrow-tolerance com-
ponents and the hydraulic fluid are very temper-
ature and contamination sensitive. Thus, the
designer will have to specify extreme cleanli-
ness requirements and adequate temperature
conditioning for maximum reliability of this vital
system. On cryogenic engine systems, heating
of the more sensitive components of the hydraulic
system is often required. Groin. supplied and
thermostat-controlled electric heaters, which
disconnect at liftoff, are a favored temperature-
conditioning method. Continuous recirculation
of the hydraulic fluid by means of the aforemen-
tioned ground electrically driven auxiliary pump
is another.
Secondary Injection
(C) LIQUID INJECI'O'R
Thrust vector control through secondary injec-

tion of matter into the thrust chamber nozzle Figure 7-17.-Secondary injection systems.
(SITVC) has been successfully applied to solid

motors. It has found only limited, predominantly factors: amplification (K) and axial thrust aug-
experimental application in liquid propulsion mentation (K,). These factors are defined as
systems, where it appears especially promising follows:
for upper stage engines, in which the lateral
forces required are smaller than with boosters.
K=ISs Fs/fvs_Fs/Fp (7-1a)
The principal methods of secondary injection
lsp Fp/_Vp fVs/_V p
are-
(l) Gas injection, using- lsa_AFa/ws_AFa/Fp

(a) Inert stored gas K, (7-15)
=lsp Fp/_'p fVs/fV p
(b) Thrust chamber tapoff (fig. 7-17A)
(c) Gas generator (fig. 7-17B) where
(2) Liquid injection (fig. 7-17C), using- ¢Is =secondary flow rate, lb/sec
(a) Inert fluid ¢¢p =primary flow rate, lb/see
(b) Propellants Fs =side force, pounds
Other methods, such as injection of preheated Fp =undisturbed axial primary thrust, pounds
hydrogen, have been investigated but proven AFa = axial thrust increase, pounds
uneconomical.
Isp = undisturbed axial specific impulse of the
In a gimbaled thrust chamber, the side force primary chamber (seconds) = Fp/fyp
is located approximately at the injector end. Is s = side specific impulse (seconds)= Fs/_'s
With an SITVC system, the applied side force is Is a = secondary axial specific impulse (sec-
located downstream of the nozzle throat and onds) = AF a/fV s
approximately at the point of injection, resulting Essentially, the K factor determines the quan-
in an increased moment arm which decreases the tity of fluid required to obtain the side force,
required side force. and the K_ factor determines the penalty on the
overa_A system Is to obtain the required side
Performance Evaluations force. If both of these factors are known, the
The performance of any type of secondary total effect of a given secondary injection sys-
injection system is based upon two performance tem on a propulsion system may be determined.
The K factor determines the quantity of second- Liquid injection systems (inert fluid or pro-
ary injectant fluid required (for a known duty pellants) offer the simplest arrangement. This
cycle); the maximum flow rate; the additional is offset by their low performance, K-factors
tankage, pressurization fluid, and secondary being in the order of unity, at flow rates from 5
injection hardware weight; and the effect of the to 6 percent of the primary flow. However, in
added inert weight on vehicle trajectory. The systems with low-duty cycles, they may still be
K 1 factor evaluates the Is penalty on the propul- very attractive.
sion system. If K_ is equal to 1, the specific As a rule, four elements are required for a
impulse of the secondary fluid is equal to that of given system, equally spaced on the main cham-
the primary fluid and, therefore, the propulsion ber circumference, of which no more than two
system suffers no specific impulse penalty due adjacent ones would be operating at a given
to the SITVC system. time. The control of the required valves is ac-
Both the amplification factor K and the thrust complished through a logic and a servosystem
augmentation factor K1 are influenced by the analogous to that of a hydraulic gimbal actuator
secondary injection orientation. For each appli- system.
cation, a tradeoff must be made between the two
factors to determine the optimum injection orien-
7.6 DESIGN CONSIDERATIONS FOR FLUID-
tation for maximum propulsion efficiency. Let
FLOW CONTROL COMPONENTS
the force of an external jet of comparable geom-
etry at right angles to the primary nozzle be By theirvery nature,liquidpropellantrocket
unity. Then amplification factors greater than 2 engines use many control elements for regulating
are obtained if secondary injection is made with and measuring of fluid flows, such as valves,
the nozzles pointing upstream, rather than in a pressure regulators, and flowmeters. Some of
normal or downstream direction. Side forces for the design considerations governing these cqm-
a given _Ps are further increased if injection is ponents are discussed below.
made through a series of holes arranged on a
horizontal arc, rather than through a single ori-
Basic Flow Characteristics of an Ideal Fluid
fice. Note that the manifolds required in this
case may adversely affect response, however. Fluids, by definition, include both liquids and
Test experience suggests that overall pressure gases. A liquid is an incompressible fluid which
ratio and injector size appear to have little is characterized by a tendency to retain a fixed
effect, while gas temperature does, optimum val- density or volume; but not shape. A gas is a
ues being a function of propellant combination. compressible fluid which has no tendency to
For an oxygen/hydrogen tapoff system, the either a definite shape or volume. Its density or
range between 3000 ° and 4000 ° F appears most volume will vary according to the basic gas laws
favorable; however, as with turbines drives, (eqs. 1-9, 1-12, and 1-13). In general, the same
material strength and cooling problems will dic- fundamental laws of force, mass, and velocity
tate values substantially lower, say 1800 ° F. In apply to matter in all forms, and thus are also
a typical tapoff SITVC system, the gas flow rate applicable to the flow of fluids.
may be 1.5 to 2.5 percent of the primary flow The analysis of fluid-flow controls may be
rate, the upper value indicating the situation of simplified by initially assuming ideal conditions.
maximum force required between two injection For the calculation of physical dimensions and
stations (two jets operating). The tapoff system functional characteristics of specific control
offers simplicity and good performance. How- components, the results can then be modified by
ever, with low-duty cycles, a continuous bleed additional assumptions and empirical factors,
may be necessary to maintain temperatures at which often are the result of extensive testing.
the valves. A frictionless (zero viscosity), incompressible
The performance of a gas generator SITVC fluid which is nonturbulent and loses no mechan-
system is comparable to that of a tapoff system, ical energy as heat is referred to as an ideal
probably slightly better. This is offset by higher fluid. For steady, ideal-fluid flow in a closed
complexity (valves, injectors, ignition, cooling). conduit, Bernoulli's energy equation applies:
DESIGN OF CONTROLS AND VALVES
144pl V2"I V2 to at least 10 times its diameter for repeatable

z[ +-- + = z2 + __
144 P2 +_-=2 constant (%2)
p 2g p results. For liquid flows, this flow-measuring
method is fairly accurate, if frictionlosses are
Assuming z t =z 2, and rearranging the expres- compensated for by the velocity coefficient Cv.
sions, we obtain For gaseous flows, however, pressure and tem-
perature have a significant influence on the den-
sity of the fluid and must be taken into account
144 (Pl - P2)_ v22 - v,_
(7-3) for calculations.
p 2g
In conformance with the continuity law of

fluid flow OtFFE_ENTIAL _IAPH_ C,kA
PrLOT _¢ALVE
# = Cvpv,A 1 _ Cvpv2A2 (7-4)

144 144
and
vl A2
(7-5)
v2 A_
_ENTURI
where _ATE
z z, z 2 =elevations at sections 1 and 2, ft

P,, P2 = static pressures of the fluid at sec- Figure 7-18.-Schematic of a typical closed-loop,
tions 1 and 2, psia fluid-flow control system.
v,, v 2 =velocities of the fluid flow at sections
1 and 2, fps
Figure 7-18 is the schematic of a typical
p = density of the fluid, lb/ft 3
closed-loop fluid-flow control system, in which
the static pressure differential across a venturi
IP = weight flow rate of the fluid-flow,
is sensed and used to control fluid flow. The
ib/sec
flow is regulated by means of a butterfly gate
Cv =venturi or orifice flow velocity coeffi-
which is positioned by a fluid-powered actuator
cient. This is a function of the de-
diaphragm. The working fluid pressure to the
sign configuration and of the fluid-flow
diaphragm is controlled by a pilot valve. Its
characteristics, and is determined by
position in turn is controlled by the pressure
tests.
differential between venturi inlet (sec. 1) and
A_, A 2 =cross-sectional areas of the fluid-flow
throat (sec. 2). Because of the dynamic charac-
conduit at sections I and 2, in 2
teristics of the venturi sensing ports, diaphragms,
The above basic fluid-flow characteristics springs, butterflies, and other parts, the rela-
can be used to measure or sense the flow rate in tionship between flow rate w and venturi pres-
flow control systems. An accurately sized re- sure differential (p_ -P2) is not exactly linear.
striction, such as an orifice, nozzle, or venturi, However, theoretical analyses usually permit
is inserted in the conduit. Pressure taps are good approximations of these dynamic functions.
provided for reading the static pressure p, and The empirical factors thus obtained will permit
P2 at the inlet (sec. I) and at the minimum area design calculations resulting in a reasonable
of the restriction (sec, 2). If the flow areas A I, degree of control accuracy. Flow-bench calibra-
A 2 and the fluid density p are known, the flow tions, including adjustments of, for instance,
velocities v, and v 2, and the flow rate _¢ can be spring forces, serve to further increase this
calculated with the aid of equations (7-3), (7-4), accuracy. The control of fluid flow and pressure
and (7-5). The venturi or orifice meter should be by means of orifices and regulators will be further
preceded by a straight length of pipe equivalent discussed in subsequent sections.
Sample Calculation (7-1) A =area of the surface in consideration, ft 2

# =viscosity of the fluid, lb/ft-sec
The followingdata are given fora horizontal
(viscosity conversions:
venturimeter,measuring liquidoxygen flow:
Venturiinletdiameter,dj = 6 in
1 lb/ft-sec = 3@.21b-sec/ft2
Venturithroatdiameter,d2 = 3 in
Venturiflow velocitycoefficient,Cv = 0.92
Pressure differential
between inletand throat - 4616.81b-see/in 2 = 14.84 poise)
(Pl - P2) = 22.5 psi
Density of LOX, p=71.38 lb/ft 3 U =velocity of a fluid particle at the surface
Determine flow rate w. in consideration, ft/sec
t =distance from the point where the velocity
of a fluid particle is zero, to the surface
Solution
in consideration, ft
From equation (7-5): U/t =rate of angular deformation of the fluid
When a fluid is forced to flow through a closed

VZ A2 (d2_ 2 1
conduit, its flow is laminar or nonturbulent below
v2-A , \d,/ =-4
certain "critical" velocities. In a laminar flow,
V 1 = ¼V 2 the fluid moves in layers, or laminae, one layer
gliding smoothly over an adjacent layer, with
Substitute this into equation (7-3): only a molecular interchange of momentum. The
velocity of the fluid is greatest at the center of
144 (p_- P2)_ v22_ (,_v2)
2 the conduit and decreases sharply to zero at the
p 2g conduit wall. As the flow velocity is increased
above the "critical" point, the flow becomes
turbulent. In turbulent flow an irregular random
V2 1 motion of the fluid exists, in directions trans-
=_ 2× 32'2 ×i44 × 2i'5-55.9 fps
71._s1- verse to the direction of the main flow. The
turbulent flow velocity distribution is more uni-
Substitute this into equation (7-4): form across the conduit than with laminar flow.
Even in turbulent flow there is always a thin
77
layer at the conduit wall, the boundary layer,
Cvpv2A 2 0.92 x 71.38 × 55.9 ×-_-x 9 which moves as a laminar flow.
Flow rate ¢¢=--
144 144 Experiments and theoretical considerations
have shown that the Reynolds number, R e of a
= 180.2 lb/sec
given fluid flow can be used as a criteria to
indicate whether a flow is laminar or turbulent.
Real Fluid Flows Involving Pressure Drops
All real fluids possess the physical property

mA
of viscosity; i.e., they offer resistance to shear
F u
stresses. The viscosity of the fluid directly ira,
/ /
affects friction. The basic correlation is given / I
by Newton's law of viscosity (see fig. 7-19): / I
/ I
/ I
gU / I
r =-- (7-6)
gt l I
/ I
where t
r = shear stress = F/A, lb/ft: -- _-veuoc_tv OF FWtO PArTtcue_s zeao
F = shear or friction force of the fluid tangent
to the surface in consideration, lb Figure 7-I9.-Angular deformation of a real Iluid.
(R e : Dvp/tz, where D = equivalent diameter of the ciably. Also, there will be a slight change in
conduit, ft; v=flow velocity, fps; p=fluid den- the friction factor. Consequently, itis recom-
sity, lb/ft3; and/_=fluid viscosity, lb/ft-sec.) mended that equation (7-7) be used with com-
For most calculations, it is assumed that the pressible fluids only where the pressure drop Ap
flow is laminar for Reynolds numbers less than is less than 10 percent of the fluid static pres-
1200, and turbulent for Reynolds numbers greater sure at the outlet point. To calculate higher
than 1200. pressure drops of compressible fluids, other
Real fluid flows always involve friction methods should be used.
caused by rubbing of the fluid particles against If the flow is laminar (R e < 1200), the friction
one another and against the conduit wall. Con- factor is a function of the Reynolds number, and
sequently, there is a loss of energy; i.e., drop can be arrived at by Poiseuille's equation for
in pressure in the direction of the flow. This laminar flow
energy loss is converted into heat energy. The
heat thus produced may be entirely absorbed by
f__6_4_4 (7-8)
the fluid, in one extreme case (adiabatic flow), Re
or it may be entirely dissipated through the
conduit wall, in the other extreme (constant- When the flow is turbulent (R e > 1200), the
temperature or isothermal flow). Generally-at friction factor depends not only upon the Reynolds
ambient temperature-the flow of liquids and number but also upon the roughness of the duct
gases through pipes is assumed to be isothermal. or tube walls. Tile friction factors of turbulent
However, adiabatic flow is assumed to take flows may be found by means of the Moody dia-
place in nozzles, orifices, short tubes, and gram' shown in figure 7-20. The dimensionless
valves through which the fluid is moving at high term, "relative roughness" (e/D), is a measure of
velocities. the size of the surface roughness projections
The pressure drop Ap (psi) of a fluid flowing relative to the duct diameter. Average values of
in straight conduits (ducts or tubes) in a hori- surface roughness projections e for rocket engine
zontal position can be estimated by equation flow control components are given in table 7-3.
(7-7). This is essentially the same as equation For a curved-flow passage or for other shapes,
(4-32), except for the dimensions. the friction factor f obtained from figure 7-20 has
to be modified by an empirical correction factor,
which is a function of the Reynolds number and
Ap:f L PV2 (7-7)
288 g
where
TABLE 7-3.-Average Values of Surface Rough-
L = length of the conduit, in
ness Projections for Rocket Engine Fluid-Flow
p = density of the fluid, lb/ft 3
Control Component Designs
v = flow-velocity of the fluid, fps
d = equivalent diameter of the duct or tube, in
IRoughness
4 × duct cross-sectional area (any shape) Surface description projection,
_, ft
Wetted perimeter
I : friction factor, determined experimentally Drawn tubing with very clean surface ........ 0.000005
Smooth machined and clean surface .......... .00001
Machined or commercial cold-rolled surface... •00005
Equation (7-7) is valid for laminar or turbulent .0001
Rough machined surface ....................
flow of any incompressible fluid in ducts or Smooth cast or forged surface ............... .0003
tubes. With suitable restrictions it may also be Commercial cast, forged and welded surface .. .0008
used when compressible fluids are being han-

dled. The density of compressible fluids changes
considerably as a function of pressure; there-
fore, if the pressure drop between two points is 1Moody, L. F., Friction Factors for Pipe Flows,
great, density and velocity will change appre- Trans. ASME, Nov. 1944.
Figure 7-20.-Moody diagram.
of the design configuration. Sometimes the in-

creased resistance of a specific flow passage
can be accounted for by assigning to it a ficti-
tious or equivalent length Le, of straight duct
which is arrived at empirically. The sum of this
equivalent length and the actual passage length,
(L e + L), is then used in equation (7-7) for the
calculation of turbulent flow. Figure 7-21 pre-
sents typical resistance characteristics of 90 °
bends.
/
Because flow-control components
valves and fittings disturb the flow pattern, they
such as
/
produce an additional pressure drop in a duct or /
line of tubing. The loss of pressure produced by
a flow-control component cQnsists of the pres-
sure drop within the component itself, as well as
the pressure drop in the upstream and down-
stream ducting or tubing in excess of that which
would normally occur if there were no component Figure 7-21.-Typical resistance characteristics
in the line. With certain exceptions, the fluid o! 90 _ bends.
flows through rocket engine control components where

are usually treated as being turbulent. The true Ap = pressure drop chargeable to the com-
pressure drops chargeable to the components can ponent as defined by the test method
only be evaluated accurately through actual flow shown in figure 7-22, psi
tests. p =density of the fluid, lb/ft 3
Figure 7-22 shows a typical test setup for v= 144 _= characteristic flow velocity of the
fluid-flow-control components. Pressure taps pA* component, fps
are located 4 diameters upstream and 10 diam-
= flow rate of fluid passing through the
eters downstream of the component to be evalu-
component, lb/see
ated. This minimizes the flow disturbances at
A* = characteristic flow area of the com-
the pressure tapoff points caused by the compo-
ponent. This is the minimum cross-
nents. The combination of pressure gages,
sectional area in the flow path of
U-tube manometer, and weighting tank produces
the component when fully open, in 2.
quite accurate and repeatable test data. The net
Usually this area is designed to
pressure drop caused by the component is ob- have about the same cross-sectional
tained by subtracting from the measured Ap that
area as the duct leading to the
pressure drop which is caused by an uninter-
component.
rupted straight pipe of the same size and length
The coefficient K is essentially constant for
(a+b)=14 diameters, at the same flow condi-
any given component over a large range of Reyn-
tions.
olds numbers, providing the flow is turbulent.
Because of the large number of fluid-flow-
For a given type of component configuration, K
control components used in rocket engines and
may vary with its size. A smaller size tends to
the great variety of service conditions, it is
have a higher K value. In general, the higher
virtually impossible to obtain individual test the flow resistance of the component, the more
data for every type and size of a component for
nearly independent of size is the resistance
the determination of pressure drop. It is desir- coefficient g. If a series of flow-control compo-
able instead to extrapolate from test data which
nents of different sizes were geometrically simi-
may be already available. This can be done by lar (constant ratio in all the linear dimensions),
employing a component resistance coefficient K
the resistance coefficient K would then depend
when calculating pressure drop using the corre- upon the Reynolds number only and would not be
lation
influenced by component size. However, the
design of a component is influenced by design
pv 2 standards, economy of material, structural
Ap = K288----
_ (7-9)
strength, available space, etc. None of these
considerations necessarily require geometric
similarity of the various sizes of a given design.
/.--MANOMETER
Average resistance coefficients for various fluid-
flow-control components of liquid propellant
i-I rocket engines are presented in table 7-4.
For minimum pressure drop across a flow-
control component, the following flow-passage
CONTROL VALVE __ / I /CONTROL VALVE design considerations should he observed:
(1)-Allow sufficient characteristic flow area
for the component
(2) Avoid abrupt changes of flow area
(3) Avoid abrupt changes of wall contour, and
sharp turns in the flow path
(4) Minimize the length of the flow path within
the component
Figure 7-22.-Typical test setup for fluid-How (5) Provide a smooth surface finish for the
control components. flow passages
TABLE 7-4.-Typical Resistance Coefficients for main oxidizer valve (butterfly type) of the A-I
Various Fluid-Flow-Control Components of the stage engine.
Liquid Propellant Rocket Engines
Liquid oxygen flow rate, Q = 12 420 gpm
Resistance Liquid oxygen density, p = 71.38 lb/ft 3
Component description coefficient K Flexible duct inside diameter, d=8 in
Flexible duct actual length, L= 16 in
Butterfly-type valves (fig, 7-33):
0.31 Flexible duct equivalent length considering
90 ° open .............................
80 ° open............................. .41 resistance due to flow passage contour devi-
70 ° open ............................. ,77 ation, Le = 6 d
60 ° open ............................. 198 Main oxidizer valve characteristic flow area
50 ° open ............................. 5.68
= 78 percent of duct area
40 ° open ............................. 15.45
30° open ............................. 44.7
20 ° open ............................. 124.2 Estimate:
Ball-type valves (fig. 7-38): (a_.) The pressure drop chargeable to the duct
90 ° open ............................. 0.81
The pressure drop chargeable to the valve
70 ° open ............................. 1.58
50 ° open ............................. 3.6
18.2 Solution
30 ° open .............................
20 ° open ............................ 63
10 ° open ............................ 362 (a) Oxidizer flexible duct
2.5-3.5 The average flow velocity in the flexible duct
Poppet-type valves (fig. 7-40): Full open
Venturi-type valves (noncavitation)
(fig. 7-41) ............................ 0.8-I.5 Q 12 420
Gate-type valve (fig. 7-42): - 79.4 fps
v= d2 =3.12xTrx16
Full open ............................ 0.18
3.12x _-
_Aopen .............................. 120
½ open ............................. 5.6
_Aopen .............................. 24 From table 6-3, the viscosity of liquid oxygen
Poppet-type cheek valve (fig. 7-60) ....... 2to 4 is 0.277 × 10 -_ lb-sec/in2; thus/1 = 0.277 × 10-7
Swing-gate-type check valve (fig. 7-61) .... 1 to 2.5
x4636.8-0.1282x 10 -3 lb/ft-sec (see eq. 7-6).
Standard tee ........................... 18
.90 The Reynolds number of the flow in the duet
Standard elbow (90 °) ....................
Medium sweep elbow .................... .75
Long sweep elbow ..................... .60
45° elbow ............................. .42 Dvp 8x 79"4 x 71"38
Sudden enlargement: Re --_
/z
_
0.1282 x 10 -s =2"94x 107
d_/d 2= ¼ ........................... 0.92
d_/d2='_ ........................... .56
dl/d2=_A ........................... .50 Use a surface roughness projection size • of
Ordinary entrance ...................... .50 0.00005 or a relative roughness
Sudden contraction:
d_/d I --tA ........................... 0.42
d:/d 1:½ ........................... .33 0.00005 0 000075
.19 D 8
d2/d I :¾ ...........................
12
for the duct. From figure 7-20, friction factor,

f: 0.0112. Substitute the equivalent total length
The characteristics of fluids flowing through (L e + L) into equation (7-7). The pressure drop
orifices will be further discussed in section
chargeable to the oxidizer flexible duct then is
7.10.
f(Le + b)pv 2
Sample Calculation (7-2) Ap =- 288 gd
The following design data are given for the _ 0.0112 (6 x 8 + 16) x 71.38 x (79.4) 2 = 4.34 psi
oxidizer pump discharge flexible duct and the 288 x 32.2 x 8
(b_._)gain oxidizer valve surfaces by the fluid stream. This is especially

The characteristic velocity of the valve acute with the higher density liquids. To obtain
reliable control performance characteristics with
79.4 liquids at high velocities, the control compo-
V:o.--._18= 101.6 fps nents subject to impact loading must be designed
to withstand the stresses involved. Also, they
From table 7-4, the resistance coefficient for should be contoured so as to maintain small
butterfly valves K=0.31. Substitute this into impingement angles with the fluid stream and to
equation (7-9) to obtain the pressure drop charge- keep inpact forces to a minimum.
able to the main oxidizer valve:
Fluid-Flow Temperature
Pv2 0"31×71'38×(101'6)2 = 24.65 psi
Ap = g-_8_g - 288 × 32.2 Temperature is an import_tnt consideration for
the design of fluid-flow controls. This is espe-
Control Fluid Pressure Level cially true if the controls are for fluids at tem-
peratures in excess of, or far below, norn_al
The working pressure level and the temper- ambient.
ature of compressible fluid-flow-control system In liquid propellant rocket engines, fluid-flow
are important factors, since both govern the den- controls may have to handle hot gases at temper-
sity of the fluid. Means of compensation for atures up to about 1700 ° F. Example: the con-
changes of pressure in a compressible fluid control of a turbine working fluid. Hot liquids need
trol system must always be provided. With an not be considered, since none of the liquid pro-
incompressible fluid, the pressure has relatively pellants have sufficiently low vapor pressures to
little influence on density. permit handling at high temperatures. Ability to
The working pressure level of the fluid deter- operate at elevated temperatures without any
mines the selection of the structural design of form of lubrication is a prime objective in the
the components as well as of the sealing meth- mechanical design of fluid-flow control. This
ods, especially for dynamic seals. Special pro- can be accomplished by using bearings of either
visions are often made to meet the stringent extremely hard, wear-resistant alloys, such as
requirements in high-pressure applications. For stellite and sintered carbides (high loading
example, the cutoff events in a high-pressure condition), or relatively soft materials such as
turbopump-feed engine system may be sequenced graphite (low loading condition). Bearings are
so that turbine power is cut first; thus the main usually subject to compression loads only and
propellant valves are not required to shut off are therefore not subject to failure if the mate-
against the high main-stage discharge pressures. rials used are of low ductility. For structural
members not subject to wear or bearing loads,
conventional high-temperature alloys such as
Fluid-Flow Velocity
stainless steels and other nickel-base alloys
The requirements for smooth component-flow- may be used. For static and dynamic seals,
passage contours are more critical with controls metal gaskets and bellows, carbon or graphite
for compressible, or low-density, fluids such as face seals, and labyrinth-type seals are suitable
hydrogen than for incompressible fluids, because at high temperatures.
their design flow velocities usually are much At the other end of the scale, liquid propel-
higher than those of the denser liquids. Also, in lant rocket engine controls may see extremely
general the design trend for high-thrust, high- low-temperature levels, such as in liquid hydro-
pressure engine systems is toward smaller pro- gen service (-425 ° F). Hero, two principal con-
pellant duct and valve sizes, and consequently ditions must be considered: (1) The physical
toward higher flow velocities (over 100 fps). characteristics of the fluids which at these low
An important consideration in the design of temperatures may affect control performance; and
high velocity flow-control components is the (2) the physical characteristics of the materials
high-impact loading imposed upon the control from which the control components are made and
• • [ .
which may affect the operation and, thus, the limiting factors governing response rate are (1)
performance characteristics of the control the speed with which signals can be transmitted,
devices. and (2) the mass/force ratio or its function, the
inertia/force ratio of the main control organ.
Many of the cryogenic fluids, i.e., liquefied
In many fluid-control systems the controlled
gases, experience somewhat unpredictable phase
fluid is used to transmit the sensed signal, In
changes (two-phase conditions) for relatively
others, part of the sensing link employs electri-
small temperature changes. No serious difficul-
cal or mechanical means. However, in most
ties need to be expected, however, if the heat-
cases, part or all of the sensing loop utilizes an
transfer rate from components in critical control
impulse generated by a pressure change. This
areas is low enough to prevent vaporization of
impulse is transmitted at the speed of sound in
the liquid. This is particularly important in
the fluid. As a typical example, the velocity of
liquid hydrogen service, where insulation may
sound in water is five times that in air; accord-
pose difficult design problems. At any rate,
ingly, a control signal would be transmitted five
except for viscosity changes, nearly all liquids
times faster in water.
exhibit more stable physical characteristics with
The actuators for most fluid-flow-control
large temperature variations, within the range
mechanisms use pistons or diaphragms, powered
between their freezing and boiling points, than
by fluid pressure which, in turn, is regulated by
do gases if the temperature range reaches to
some form of pilot valve. If suitable, the con-
their liquefaction temperatures.
trolled fluid may be used as the actuating fluid.
The construction materials for fluid-control
The response and flow capacity of the pilot
components for low-temperature applications
valve, the effective area of actuator piston or
must be especially carefully selected. Practi-
diaphragm, and the actuating fluid pressure level
cally every metal undergoes irregular phase
influence directly the response rate of the con-
changes at low temperatures which may seri-
trol mechanism for given mass inertia and fric-
ously affect its physical properties. While the
tional or other resistances.
strength of metals generally increases with a
decrease in temperature, further temperature To satisfy certain operating conditions and to
decrease beyond certain limits may result in a attain stable control it sometimes becomes nec-
decrease in strength. Many metals also become essary to introduce simple damping devices. In
brittle at very low temperatures. Most of the most control systems, stability is inversely pro-
aluminum alloys and the 300-series stainless portional to sensitivity or response rate. Thus,
steels exhibit much better stability at tempera- the design of a fluid-flow-control system should
tures in the cryogenic range than do others. reflect a realistic balance between sensitivity or
Elastomers such as Teflon, Kel-F, and Mylar, response rate, control accuracy, and system
when used for sealing purposes, exhibit satis- stability.
factory mechanical characteristics at extremely Figure 7-23 illustrates the schematic of a
low temperatures. Teflon-coated surfaces addi- typical piston-type actuator for fluid-flow-control
tionally have good anti-icing characteristics. devices. The piston when actuated moves
For further detail on materials, see chapter II. against the spring in the direction of the arrow.
Fluid-flow-control components for operation The basic correlation between the response rate
at cryogenic temperatures should be designed to or acceleration of the piston, and other opera-
be free of external icing effects. In addition to tional parameters, can be expressed by
insulation, moisture-preventing purges should be
provided internally in critical areas such as Map
-Alp 1-A2p2- Fr- Ft- Fs- Cx (7-10)
bearing interfaces. Also, actuators and/or bear- g
ings may require heating. where
M :effective mass accelerated by the actuator
piston, lb. It consists of piston mass,
Rate of Response in Fluid-Flow Controls
that of moving parts mechanically con-
Response rate is an important design consid- nected to the piston, and of the mass of
eration in any control system. Basically, the all the fluid columns in the system
ACTUATING FLUID
SUPPLY LINE
tons, shafts, and rods; and seals for valve seats.
i Here, too, temperature is one of the most impor-
tant design considerations. Seals can be classi-
SPRING fied into those for medium-temperature service

(-60 ° to 400 a F), low-temperature service (-60 °
to -425 ° F), and high-temperature service (400 _ F
0 FI BODY
and up). The selection of the configurations and
the materials for these seals is based to a large
extent on service conditions and type of fluid
involved. Generally, soft nonmetallic or elas-
VSNT _-__!I!! ///, tomeric seals are used wherever possible. The
outstanding advantage of these seals is that
p,sTo;
N"I
A. they
fections
function
in the
satisfactorily
seal or the
despite
mating part.
minor imper-
Figure 7-23.-Schematic of a typical piston-type

actuator. Design of Dynamic Seals for Medium-Temperature
Services
ap = acceleration of the piston, ft/sec 2
A_ --area of the piston actuating side, in s Elastomeric O-rings have been widely applied
A s =area of the piston vent side, in 2 as dynamic seals for moving cylindrical parts as
p_ =actuating pressure, psia. This is tile well as for valve seats. However, proper design
pressure at the actuating fluid source, techniques must be observed to assure success.
less the supply-line pressure drop which Figure 7-24 and table 7-5 represent recommended
depends on the flow rate (a function of design practices for diametral-squeeze-type
acceleration ap) O-ring seals for typical dynamic and static appli-
P2 =vent pressure, psia. This is the ambient cations. Figure 7-26 shows a typical O-ring seal
pressure, plus the vent-line pressure drop used for a valve seat. Important design consid-
which also depends on flow rate (again a erations for dynamic O-ring seals are summarized
as follows:
function of acceleration ap)
Fr = resistance force of the control function, lb, 1. Design correlations between fluid pres-
which also may be a function of accelera- sure, O-ring compound hardness and diametral
tion ap clearance.-A frequent cause of seal failure is

Ff = friction forces (seals, sliding surfaces, the extrusion of diametral-squeeze-type O-rings
etc.), lb into the clearance gap adjacent to the O-ring
Fs = initial spring force, lb (at x = 0) groove (when under pressure) (fig. 7-25), leaving
C =spring rate, lb/in a permanent deformation after the pressure is
x =distance traveled by the piston from its reduced. Fluid pressure influences O-ring seal
initial position, in design because it affects the choice of compound
Since the relations between p_, P2, Fr, and ap hardness. The combination of fluid pressure and
are not linear, equation (7-10) may become com- chosen hardness will determine the maximum
plex and require a high-speed computer for clearance E that can be tolerated safely. A
solution. proper combination of clearance and O-ring hard-
ness may prevent O-ring extrusion (table 7-5). In
general, the O-ring nominal section diameter is
7.7 DESIGN OF DYNAMIC SEALS FOR
chosen as large as space and installation con-
FLUID-FLOW-CONTROL COMPONENTS
siderations will permit. Past designs usually
Apart from the static seals, which will be will provide a useful guide. In addition, Teflon
treated in chapter IX, two basic types of dynamic backup rings as shown in figure 7-25 are recom-
seals are required for fluid-flow-control compo- mended for dynamic O-ring seals at sealing pres-
nents: seals for moving (reciprocating and rotat- sures over S00 psi and for static O-ring seals at
ing) cylindrical elements such as actuator pis- pressures above 1500 psi.
illtBl #:
STATIC
SEALING\
_ ¢,-_ I'-- "1 D-GROOVE
\BREAKCORNERS, : _b -,J U:NGTH
?o APPROX.
o.oos' D l .-f
"=" . _L__L._4___ a_-_--L "- CENTERLINE OF

THE PISTON ACTUATOR
Figure 7-24.-Diametral-squeeze-type O-ring seals in typical dynamic and static applications.
2. Surface finish requirements.-The finish of ter is recommended for surfaces in contact with
the sliding surfaces in contact with dynamic static diametral-squeeze-type O-ring seals.
O-ring seals should be as smooth as possible. 3. Friction of dynamic O-ring seals.-The
They should be ground, honed, or polished to a breakaway friction of a dynamic O-ring seal is
microinch finish of 8 to 10 rms. It has been usually about three times the running friction.
found that a finish within this range yields a Breakaway and running frictions increase with
longer life than either rougher or smoother fin- fluid pressure, O-ring hardness, diametral
ishes. Codirectional patterns, as produced by squeeze, and decrease of temperature. Accurate
honing, have been proven to be the best surface values of O-ring frictions can only be obtained
finish for any type of dynamic sliding seal. For experimentally for a given design.
still better results, after an initial finishing, the 4. Selection of O-ring compounds,-A great
surface could be hard chrome or nickel plated variety of O-ring elastomer compounds is avail-
and again finished. The plating provides a hard, able, with trade names such as Silicone rubber,
slippery surface that resists corrosion, wear, and Buna N, Butyl, Viton, Teflon, and Kel-F. The
scratching. A microinch finish of 60 rms or bet- selection of an O-ring compound and its physical
TABLE 7-5.-Recommended Design Practice for Diametral-Squeeze-Type O-Ring Seals
[See fig. 7-15 for explanation of dimensions; all dimensions in inches]
O-ring Diametral squeeze, 2 x E-

O-ring C-glandwidth D- R-
diametral
nominal rain
section groove radius,
section clearance
diameter length rain
diameter Dynamic Static aDynamic bstatic max
1/16 0.070± 0.003 0.010 0.015 0.057 0.052 3/32 1/64 0.005
3/32 0.103 *_0.003 .010 .017 .090 .083 9/64 1/64 .005
1/8 0.139 *_0.004 .012 .022 .123 .113 3/16 1/32 .006
3/16 0.210 *_0.005 .017 .032 .188 .173 9,f32 3/64 .007
1/4 0.275 *_0.006 .029 .049 .240 .220 3/8 1/16 .008
Fluid pressure O-ring compound hardness
(3-1000 psi 70 Shore "A" Durometer

1000-2000 psi 80 Shore "A" Durometer
2000 psi and higher 90 Shore "A" Durometer
aTolerance = +0.000, -0.001.

bTolerance : +0.000, -0.005.
_AOIAL -,-x _" _PRESSU_E __ P_ESSURE

7-27, 7-28, and 7-29) are used effectively as
C k F.A,_I* N C _ __]_ _"
dynamic seals for moving cylindrical parts and
valve seats. This type of seal has been applied
EXTRUSION OF (_RING RJNGS TO PREVENT
UNDER PRESSURE EXTRUSION
successfully at sealing pressures over 2000 psi,
and at temperatures as low as -425 ° F. They
t"igure 7-25.-Extrusion of the diamelraI-squeeze-
are also reasonably effective when sealing low-
type O-ring under pressure and the application
molecular-weight gases such as helium and
o[ backup rings.
hydrogen.
properties (furnished by the producer) is based

on operating conditions such as type of fluid or
°"' _° ...........
'Z:,;_ I 1710 t 010
propellant, pressure, temperature, and type of
seal (dynamic or static).
5. Installation o[ diametral-squeeze-type
O-ring seals.-Proper installation of O-rings dur- ....,oo,I 1, ...../ \ 1
ing component assembly is extremely important
to assure an effective seal. Generous chamfers L k t, "-/->' 1 ;
or radii should be provided on all edges and
corners in contact with O-rings to minimize the
possibility of cutting or scratching during the
/ k
assembly process. L _IPSEALS _. --A- _A_ _TE UNLESS aT_E_SE

121 _(_QLIIRIEO "_ _ _ °PIER SPECJFP[I]I F_NISH O_l i'LL
SUtteES _ T_E ._AkS

6. O-ring seals for valve seazs.-O-rings can ANC} t_ CONTACT WiTH THE
SEALS TO BE 12 RM$ OR
be applied effectively as seals for valve seats OETTER. ALL NICENSIOhS
to _ _H _NC_ES
(fig. 7-26). The resiliency of the O-ring absorbs
shock loads and seals tightly at all pressures,
Figure 7-27.-Typical rotating lip-seal design [or
even when some dirt and grit are present in the valve actuator shalt.
system. One design problem is to prevent the
O-ring from being blown out of the groove. This
can be prevented by providing a dovetail O-ring
groove in a two-piece valve poppet (see fig.
7-26).
Design of Dynamic Seals for Low-Temperature

Services
For cryogenic or low-temperature services,

lip-type seals made of elastomer sheets (figs.
VALVE STEM
Figure 7-28.-Typical valve seat lip-seal design

VALVE PRESSURE ' PRESSURE
POPPET I VALVE
[or butterfly-type valves.
_ OUStNG
The basic design principle of lip seals is to

employ the fluid pressure to increase the contact
,_AL RETAINING pressure at the sealing surfaces. Due to their
RING
lip configuration, the resilience of these seals

Figure 7-26.-Typical valve seat O-ring seal is maintained even at very low temperatures.
design for poppet-type valves. Design considerations for lip seals are similar
-Acrvz. ,_ _%ulO
to those for O-ring seals. The design approaches

can best be illustrated by examples. Figure 7-27
shows a typical valve shaft rotating lip seal
.-T_P,Ca_ WEL._EO .N_N_
arrangement, including dimensions and surface /' _ S,{'rUJXOm I,LA'II
_o_ ( CLOS_N& ) }_ KLLDW$ ASEhqI._

finishes, for liquid oxygen and hydrogen service.
Figure 7-28 presents the valve seat lip seal of a
butterfly valve for use with the same liquids.
Valve seat O-ring seals (fig. 7-26) made of Kel-F
have also been successfully applied in poppet-
W: i _.,: . ',_ ../'
type valves for liquid oxygen. The design of

lip seals for piston-type actuators using low- ;b"_ . }
temperature helium gas as the actuating fluid is
shown in figure 7-29.
In liquid hydrogen service, metallic bellows
(as shown in fig. 7-30) have been used to a great
extent to achieve positive dynamic sealing. _L_tE_ _ 001NV_
However, pressure levels and available space _.&LIN* 5t._f_3t5 LBO_H _LVE AN0 ,_v_t ) L_PPED N 8-10
impose limitations on their application. WELI)ED TYPE B£LLOWS USEO AS

RIEC_PRI_C,ATING Ty_P'_ D'yINiAIMI_;
Design of Dynamic Seals for High-Temperature Figure 7-30.-Metallic bellows used as recipro-
Services cating-type dynamic seals in a typical poppet
valve for high- and low-temperature services.
The metallic bellows (fig. 7-30) is most fre-
quently used as reciprocating-type dynamic seals
elastic modulus, and of approximately the square
for high-temperature services. Two types of
of the thickness of the material. It is also a
metallic bellows are distinguished: the hydraulic-
function of the outside-inside diameters and of
formed and the multidisk welded type. The
former is made of one to three plys of sheet the number of convolutions and their curvature.
metal and is designed for all pressure ranges. For maximum flexibility (inches of stroke/lb of
The latter is for relatively low-pressure services load), a minimum inside diameter combined with
a maximum outside diameter should be used.
and for high flexibility.
A metallic bellows of any type behaves, in Also, material thickness (within stress limita-
part, like a helical spring. The spring rate tions) and modulus of elasticity should be mini-
mum.
(lb/in of.movement) is a direct function of the
Generally, bellows design data, such as stock
DOUBLE LIP SEALS
size, allowable working pressure, spring rate,
FOR ACTUATOR PISTON
materials and service temperature, are supplied
DOUBLE LIP SEALS / by the manufacturers. Important design consid-
erations are discussed in the following:
1. Applicalion of pressure.-When a bellows

is subjected to a differential pressure between
interior and exterior, it is preferable to apply the
higher pressure to the exterior. This reduces
stress, and permits higher pressures and longer
VALVE [_ __-t__ _ _j;!
_ i'_.>_AC T UATOR ROD life for a given design.
2. Provision of mechanical stops.-These
should always be provided to prevent extension
ACTUATOR PISTON
of the bellows beyond its permissible extended
Figure 7-29.-Lip seals for piston-type actuators. length and compression beyond its "bottomed"
Double lip seals seal pressures both ways. height.
3. Selection or materials.-Selecting of bel- gas throttle valve which was operated success-
lows material should be governed by fluid com- fully at temperatures ranging from 1200 ° to
patibility or corrosion considerations, operating 1S00 ° F at pressures up to 700 psia. An alter-
temperature ranges, and spring characteristics. nate design is to attach a flat-face graphite seal
Some high-temperature alloys such as stainless ring to the end of a metallic bellows which is
steels, Monel, Inconel, and Hastelloy B have welded to the shaft (fig. 7-31). Here, the shaft
proven suitable. misalinement is compensated by the flexibility
4. EHective area.-This is that area which, of the bellows.
when multiplied by a change in bellows length, The sealing of valve seats for high-temperature
yields the actual displaced volume. This area services is usually achieved by metal-to-metal
can he approximated by contact, as shown in figures 7-30 and 7-32. This

seal design has two basic requirements. Firstly,
Bellow effective area = 0.1963 a finish of 10 nns or better is required for the
× (inside diameter +outside diameter) 2 (7-11) sealing surfaces. Secondly, a high-enough unit
loading must be applied to create a compensating
5. End attachment.-Typical welded joints for deformation of the sealing surfaces and to
the end attachment of bellows are illustrated in achieve the intimate contact required to over-
figure 7-30. Silver brazing and soft soldering come manufacturing tolerances, distortion of the
can also be employed for low-temperature serv-

ices.
A typical design of a rotating-type dynamic
seal for high-temperature services is illustrated PRESSURE
HOT GAS _ HOUSING
VALVE
in figure 7-31. The dynamic sealing is achieved

through the spherical mating surfaces between
the graphite seal ring and the steel shaft collar.
The contact force of the sealing surfaces is
maintained by the shaft thrust spring. Any mis-
alinement between the thrust bearing and the
shaft is compensated by the spherical seal face SEAL "X O
SURFACES ROTATE
and side movement of the seal ring. This seal
FINISH tO RMS TO OPEN \ BUTTERFLY VALVE GATE
arrangement has been applied to a turbine hot
VALVESHAFT
OR BETTER VALVE
{a) CONVOLUTED R_NG TYPE VALVE SEAT SEAL USED _N A
HOT GAS SHUTOFF BUTTERFLY VALVE
_ •. _ /- _.. _ _1_?,_,_
_ _,,_
HOT GAS
PRESSURE VALVE POPPET
SPRING DISK | /
/ VALVE
VALV¢_ ...... _, _ sto,,
/ HOUSING
POSITIVE SEAL SURFACES FINISH

STOP IO RMS OR BETTER
_b} SPRING DISK TYPE VALVE SEAT SEAL USED IN A
S(AL _s, aN HOT GAS SHUTOFF POPPET VALVE
Figure 7-3I.-Turbine hot gas throttle valve with Figure 7-32.-Valve seat seals for high-tempera-
typical rotating-type dynamic seals. ture services.
_1 mi _rx
valve parts due to temperature, internal stress, control purposes, ability for continu,_usly vari-
and mechanical loading of the mating parts. able opening position may be required.
Figure 7-32(a) shows a convoluted-ring-type In addition to propellant compatibility and
valve seat seal used in a hot gas shutoff butter- structural integrity, prime design considerations
fly valve. Depending upon the specific applica- for propellant valves are:
tion, the convoluted ring may be made of high- (1) No leakage of propellant through the valve
when closed
temperature alloys such as Inconel-718. The
rings effect a leakproof seal in the closed posi- (2) Proper actuating time during opening and
tion, since the upstream fluid pressure tends to closing in accordance with the require-
expand the convolute and produces a high con- ments of the control system
tact unit force at the sealing surfaces. The (3) Minimum pressure drop
curvature of the convolute ring tends to maintain A great variety of propellant valve types is
a continuous contact with the valve seat. Figure available. Each design has certain characteris-
tics which make it suitable for a specific appli-
7-32(b) presents a metallic-spring-disk-type
valve seat seal used in a hot gas shutoff poppet cation. Frequently used propellant valves, clas-
valve. Again, the upstream gas pressure pro- sified according to their design configurations,
duces a high contact unit load on the sealing are:
surfaces. The valve seat has a curved contour (1) Butterfly valves
which effects a continuous contact with the flat (2) Ball valves
face of the seal disk. (3) Poppet valves
(4) Venturi valves
(5) Gate valves
Sealing Specifications (6) Needle valves
The degree of sealing (or the allowable leak

Butterfly-Type Propellant Valves
rate) is a very important specification which will
dictate the type of seal to be selected for a The butterfly valve is one of the most widely
specific fluid-flow-control component design. used propellant valve types in large liquid pro-
The basic reference for leak rates is Specifica- pellant rocket engines. It has established a
tion MIL-S-8484. It states that a Grade A seal, reliable operational record in LO2/RP-1, LO2/
the highest quality seal, shall have a leakage LH 2, storable, and other liquid propellant serv-
rate not to exceed 1 standard cubic centimeter of ices. Existing butterfly valve designs range
air/year/inch of seal at a pressure differential from 2 to 17 inches nominal diameter, for use at
of 1 atmosphere. This corresponds to a leakage propellant pressures from 20 to over 1500 psia.
rate of 3.171 × 10 -s cc/sec/inch of seal. It is a With improvements in sealing and structural de-
design assumption that any seal leak rate below tails, successful designs for higher capacities
or equal to this value is considered zero leak- and propellant pressures are certain to be
age. For many applications, higher leak rates achieved.
are permissible. For instance, a check valve Fignre 7-33 presents a typical butterfly valve
may be specified with a leak rate of 5 scim's design. Sealing is provided by a lip seal, which
(standard cubic inches of gas per minute). This engages a spherical surface on the valve gate,
is still a relatively tight specification. similar to figure 7-28. The valve gate pivots on
the valve shaft, the axis of which passes through
the geometric center of the spherical sealing
7.8 DESIGN OF PROPELLANT VALVES
surface. In most designs, the valve gate rotates
Propellant valves are used to initiate and 90 ° from the closed to the fully opened position.
terminate propellant flows to main thrust cham- The valve is operated by a piston-type actuator,
bers and gas generators. They are usually open- through a connecting link and shaft crank arm.
closed, two-position, normally-closed valves. Lip seals are used as dynamic seals for the
To meet specific sequencing requirements, other rotating valve shaft (fig. 7-27). The actuating
designs may provide for an intermediate opening power is furnished either by noncryogenic pro-
position. For thrust-throttle or mixture-ratio- pellant pressure, or by an inert gas supply, and
plsroN _ A
O"Rr
_'_pE _- -_ : i' _-AT FL;LLYOP_NEO PO_ITIO"+, _ i
+ ...... ; .... • •
/ -- \ \ I,o+,,,o.......... + ......,.+o
/ ++ ....... -
/ SEe:tON A--A\ u_ + I
L_.......o X__ ....... r---,_ %A
z,
+, I+.'-C +,.<
+- + "-:"+-._ : , _ + i"
//. r --:-_ ....

---_ _ ,...... Figure 7-34.-Four-inch butterfly-type main liquid
oxygen valve used on Rocketdyne Atlas ICBM
booster engines.
Figure 7-33.-Typical butterfly-type propellant
valve design (shown in the closed position). valve. A butterfly valve maintains a relatively
smooth fluid-flow stream over a wide range of
controlled by a pilot valve. The valve shown is valve-gate angular positions. Thus, when used
designed to be normally closed by a spring which as a throttle valve, it has little tendency toward
is installed on the closing side of the actuator turbulence with attendant adverse effects such
piston. Except for shaft and pins which are as local propellant cavitation. Typical fluid-
made of stainless steels, most of the other parts flow resistance coefficients K at various opening
are made of aluminum-alloy forgings. Figure positions of a butterfly valve are listed in table
7-34 shows a 4-inch, butterfly-type, liquid oxy- 7-4.
gen valve used on the Rocketdyne Atlas ICBM When RP-1 is used as the actuating fluid for
booster engine. the liquid oxygen valve, in a LO2/RP-1 system
Butterfly valves have relatively low resist- such as the A-1 stage engine, a heater may be
ance to fluid flow. They are compact, light, and required at the actuator to keep the RP-1 from
easy to service. They have a high characteris- freezing. The actuator-valve arrangement of the
tic area which can be expressed as butterfly valve shown in figure 7-33 provides
flexibility for specific engine control system
needs: the valve may be normally open or nor-
A*-_-d _-Ag (%12)
-4 s mally closed; position indicators may be added;
closing of the valve may be accomplished by
where (see fig. 7-33 for dimension references) means of a pyrotechnic squib, rather than by
A* = characteristic area of the valve, in _ pneumatic pressure. Figure 7-35 illustrates the
ds =inside diameter of the valve seat lip seal, linkage between the main oxidizer valve and the
in igniter fuel sequence valve of the A-1 stage
Ag=projected valve gate area at the fully open engine. During the opening stroke, the cam
position, in _ attached to the main oxidizer valve shaft actu-
Values for A* range from 65 percent of the ates _he igniter fuel sequence valve to open.
duct area (duct area = v/4 dn 2, where d_ = valve Frequently, a potentiometer is also attached to
nominal diameter, in) on a 2-inch size valve, to the shaft for continuous indication of the angular
about 87 percent of the duct area on a 12-inch position of the valve gate.
The amount of torque required to turn valve

_._tN OXJDtZEk VALV_ shaft and gate is determined by the summation of
OPE*WtNG SIDE ADTIJATING.
P=E_J_E APPUED hydraulic and friction torques. Hydraulic torque
M,&pN OXIDIZ_I VALVE
iN FULLY C_=ENED POSITION
is the unbalance of forces on the valve gate
AC IIJAIIN G FWID caused by the flow of fluid around it. If the axis
_CLOSlNG) ° of the valve shaft is located as shown in figure
FtUIP INLET PORT
7-36(a), the fluid striking the gate portion pro-
AC_UMING
(C_ENI*WG)
truding farthest upstream is deflected more than
FUEL PIIS_U_! r_ow _,
E_G_NE CO_WIItOL VALVI that at a point near the other end of the gate.
POIATE TO J
O_EN VALVt
This produces an unbalanced force which tends
VALVl SH_.= T --_ IGNiIEI FUEL to close the gate. Offsetting the valve gate as
SEOU_NC_ VALV_
shown in figure 7-36(b) would further increase

the closing torque, because the fluid velocity
rises as it approaches the downstream side open-
IGNI_EI C_mrDC_
/,ND IGNmON ing. Consequently, the resulting low-pressure,
MONIIO_ VAtVt
area tends to increase the unbalance in the clos-
ing direction. For this reason, butterfly valves
Figure 7-35.-Mechanically linked arrangement
are usually designed offset as shown in figure
between the main oxidizer valve and the igniter
7-36(c) (also see fig. 7-33). This produces a
fuel sequence valve of the A-I stage engine.
fluid velocity effect tending to ease opening of
the gate, because of the lower net closing hy-
BUTTERFLYVALVE GATE draulic torque acting on the valve gate.
Nevertheless, the net hydraulic torque will
AXISOF VALVE
still be acting in the closing direction for most
./ / SHAFT
angular gate positions (9°-80°), unless the valve
FLOW gate is further offset. Friction torque always
(a)
opposes rotation. For most operational valve
DIRECTION
designs
FARTHESTUPSTREAM
PROTRUDING To = Tt+ Th (7-13)
T c = Tf- Th (7-14)
where
To = required opening torque, in-lb
FLOW
Tc = required closing torque, in-lb
(b)
Tf = friction torque, in-lb
DIRECTION
Th = hydraulic torque, in-lb (assumed to act in
the closing direction)
BUTTERFLY
VALVEGATE
The friction torque Tt varies with the pres-
sure differential across the valve gate, and with
AXISOF VALVESHAFT the valve gate projected area which is a function
./ of gate angular position. Friction torque can be
estimated by
(c)
FLOW
DIRECTION _ ____ T! = Ktrsfmds2hp (7-15)
\
where
BUTTERFLYVALVEGATE
K[ = friction torque coefficient, which is a func-
Figure ?-36.-Various locations of valve shaft tion of gate angular position (to be de-
axis with respect to butterfly valve gate. termined experimentally)
rs =radius of valve shaft at the bearing sec- estimated opening and closing torques. In addi-
tion, in tion, at the start of the opening stroke, the actu-
fm = coefficient of friction between shaft and ator has to overcome the static friction forces of
bearing (0.20 for aluminum journal and all seals. Butterfly-type propellant valves are
steel shaft; 0.05 for needle bearing and relatively fast acting. Opening and closing
steel shaft) times range from 20 to 200 milliseconds.
ds = inside diameter of valve seat lip seal, in
Ap=pressure differential across the valve gate,
psi
Hydraulic torque Th may be estimated by The following design and experimental data
are given for the main oxidizer valve (butterfly
Th = Khds 3Ap (7-16) type) of the A-1 stage engine.
Design Data
where Kh ---hydraulic torque coefficient, which is
Radius of valve shaft at bearing section,
a function of gate angular position
rs=0.8 in
(to be determined experimentally)
Inside diameter of the valve seat lip seal,
Figure 7-37 shows plots of required opening
ds -- 7.7 in
and closing torques versus gate angular posi-
Coefficient of friction between shaft and
tions for a typical butterfly valve. In actual
needle bearing, f m =0.05
design practice, the actuator of a butterfly valve Test Data
will provide two to three times the maximum
Valve gate angular Ap, psi K[ Kh

REQUIRED OPENING TORQUE position, deg
(To=T f +T h )
5 .............. 1058 0.78 1.11 × 10 -a
15 .............. 769 0.78 2.55x 10 -_
/FRICTION TORQUE (Tf) 40 .............. 87.5 1.57 12.50×10 -_

W 85 .............. 25 3.61 -11.64×10 -3
D
0
re"
0
I-- Determine the required opening and closing
torques at the 5° , 15 ° , 40 ° , and 85 ° angular posi-
T, ORQUE (Th) tions of the valve gate.
0° 90 ° Solution
OPENINg"
From equation (7-15), the friction torques
TI= Klrsfmds2.Ap
at 5°: T[=O.78xO.8×O.O5×(7.7)2x1058
= 1960 in-lb
REQUIRED CLOSING TORQUE
at 15°: T[=0.78×0.8×0.05×(7.7) 2×769
' = -- )
= 1425 in-lb
o at 40°: T[=l.57x0.8×0.05×(7.7) 2×87.5
o
-'- 326 in-lb
I--
at85°: T[=3.61×0.Sx0.05×(7.7) 2×25

= 214 in-lb
GATE ANGULAR POSITION / From equation (7-16), the hydraulic torques
0= _ 90 °
Th = KhdsSAp
CLOSING
at 5°: Th = 1.11 x 10-3 x (7.7)3 x 1058
Figure 7-37.-Typical required opening and clos- = 535 in-lb
ing torques versus gate angular position fora at 15°: Th;2.55x10 -3x(7.7) 3×769
butterfly valve. = 895 in-lb
at 40°: Th = 12.50x 10 -3 × (7.7) 3 x 87.5

= 500 in-lb
at 85°: Th=-11.64×10-3×(7.7)3x25 J_L
aCT_TOa
=-133 in-lb
From equation (7-13), the required opening
torques
To = TI+ Th
at 5°: To=1960+535 =2495in-lb
at15°: To=1425+895 =2320in-lb
at 40¢: To=326+500 = 826in-lb
at 85°: To=214+(-133)= 81in-lb
From equation (7-14), the required closing
torques
Tc = (Tt- Th)
Figure 7-38.-Typical ball-type propellant valve
at 5°: Tc=1960-535 =1425in-lb
design.
at l5°: Tc=1425-895 = 530in-lb
at40°: Tc=326-500 =-174in-lb
at 85°: Tc=214-(-133) = 347in-lb
assembly consists of a seal ring and an attached
metallic bellows. The area enclosed within the
Ball-Type Propellant Valves

effective seal diameter, (4ds2) , is designed to
The major advantage of a ball valve is its be less than the effective area of the bellows.
low-pressure drop, since it permits in-line, unre-
During valve closing, this creates an unbalanced
stricted fluid flow. Its use also enhances struc-
force acting on the sealing surface, as affected
tural soundness for high-pressure service. It has
by the fluid pressure within the bellows.
a reliable record in cryogenic and some storable
Each valve ball is trunnion mounted using
propellant applications, for high-capacity gas
two integral axles on antifriction bearings. The
generators as well as for lower thrust main cham-
axis of valve rotation passes through the geo-
bers (up to about 50000-1b thrust). Up to a nom-
metric center of the spherical sealing surface.
inal diameter of 3 inches, ball valves are com-
In most designs, the valve ball rotates 90 ° from
parable to other valve types with respect to
the closed to the fully open position. Typical
space envelope and weight. For larger diam-
fluid-flow resistance coefficients K for ball-type
eters, ball valves are used only infrequently,
valves, as determined at the fully open position,
because it becomes increasingly difficult to meet
are listed in table 7-4. Ball-type valves can
flight weight and envelope requirements. How-
readily be used as flow-regulating devices, such
ever, for ground service applications where as for propellant throttling. Flow characteristics
weight and size are not critical, the ball valves,
for constant pressure drops of a typical ball
in all sizes, are used quite frequently. Many
valve at various angular positions are presented
ball-type propellant valves are designed in a
in figure 7-39.
mechanically linked, dual-valve arrangement,
The activation of the ball valves shown in
operated by a single actuator.
figure 7-38 is provided by a piston-type actuator,
Figure 7-38 illustrates the design of such a which could be powered by either fuel pressure
valve. Here, the valve elements are mechani- or inert gas pressure. The reciprocating motion
cally linked, controlling both oxidizer and fuel of the actuator is translated to a rotary motion of
flows. The valves can be sized either individ- the balls by means of a connecting link and crank
ually as shown, and according to the specific arm arrangement. The opening sequence of the
volumetric propellant flows, or be designed to two valves can be adjusted by varying the rela-
have a common size. The sealing of a ballvalve tive angular positions between the valve axles
is accomplished by lip- or O-ring-type seals, and the crank arm. The actuator shown has two
riding on the spherical sealing surface of the stages. First, the valves are opened to an inter-
valve ball. In our specific case, the valve seal mediate position (partial opening), then to the
*,_l_ vAtvl oo_Lt7
t
tooor_ I _ svtL
700_
i [ ,2"- ' P'
I - X-_-__ .._--n-Jh_'_i_-_,,--_ ' _-\1"'× ....
400---
=J
2_ 30" .t_"
ANGULAR
._" 60"
POSITION
70" gO" _+
= \!; ......
OPEN
Figure 7-40.-Typical large-size poppet-type
NOTE: Zero f_ from 0"- S'due to lea1 deslgn propellant valve design.
_P2 a5 -'%" PI
Figure 7-39.-Flow characteristics for constant

in the balance chamber are dimensioned so as to
pressure drops ol a typical ball-type valve at
various angular positions. result in the proper counteracting force which
varies as a function of the unbalanced force at
various positions of the valve. A small sequence
fully open position using separate pistons for valve is mechanically attached to the main valve.
each stage. The closing of the valve is effected This type of poppet valve is suitable for high-
by venting both opening ports and pressurizing flow attd high-pressure, storable as well as
the closing port. Dual seals with a drain be- cryogenic, propellant services•
tween them are provided for all dynamic seals
sealing to ambient.
Venturi-Type Propellant Valves
Figure 7-41 presents a typical design for a

Poppet-Type Propellant Valves
venturi-type propellant valve. In certain instal-
Figure 7-30 shows a typical poppet valve lations, it may be desirable, for various reasons,
with metal-to-metal seats. This valve is de- to use a valve of a nominal size smaller than
signed to be pneumatically operated and nor- that of the main duct. A valve installed in the
mally closed. All sealings are achieved without throat of a venturi is a possible solution. The
the use of elastomers. Because of the nonwiping smooth contours of the venturi limit pressure
characteristics of all dynamic seals, this design drop penalties to a few psi. Adjacent ducting
is particularly suitable for use with fluorine and permitting, it is conceivable that the venturi may
other highly reactive propellants. A main ad- simultaneously be used for flow measurements.
vantage of poppet valves is their relative sim- Typical resistance coefficients K for venturi
plicity. This is largely due to the reciprocating valves are given in table 7-4.
operation which permits the direct, in-line con- The venturi may be designed to operate as a
nection of an actuator. However, this arrange- cavitating venturi. Based on Bernoulli's energy
ment requires turning of the flow inside the pas- equation (eq. (7-2)), the minimum pressure of a
sage. and consequently results in relatively liquid is made to fall below its vapor pressure.
high-pressure drops. Typical fluid-flow resist- As a result, a gaseous region forms at the throat.
ance coefficients K for poppet valves are given If this gas moves through the throat at the veloc-
in table 7-4. ity of sound, downstream pressure variations and
Figure 7-40 presents the design of a typical, disturbances cannot advance beyond the throat.
large size (6 to 10 inches nominal diameter), Up to minimum pressure differentials across the
poppet-type propellant valve. To reduce the venturi (say _0 percent of upstream pressure),
unbalanced hydrodynamic forces, and thus the flow rate is dependent on upstream pressure
size of the actuator, a balance chamber is pro- only. When used as a throttling device, the
vided. The effective area and the fluid pressure cavitating venturi affords smaller pressure drops,
loaded poppet seated at the throat. Propellant

-- VALVE POPPET
line pressure controlled by a solenoid pilot valve
is used to actuate the valve open or closed. As
shown in the valve schematic (fig. 7-41), the
normally closed pilot valve is inserted in a pas-
,._j_ LET sageway interconnecting the poppet cavity and
an opening in venturi throat. Normally, upstream
propellant pressure fills the poppet cavity and
provides additional seating force on the poppet
in the same direction as the valve spring to as-
sure valve closure. When the pilot valve is ener-
SPRING gized to open, propellant pressure behind the
tNLET
valve poppet is vented out at a greater rate than

_---SOLENOID PILOT VALVE
it can be replaced by leakage through the poppet
clearance area. This results in a reduced pres-
.,,_F--- SOLENO1D PILOT VALVE sure overcoming the valve spring and causes the
main valve poppet to open. The venturi valve
contains no dynamic seals. Since the valvebody
is not pierced by a shaft or actuator rod and
_ --------=_ FLOW
there are no dynamic seals, no pathways exist
POPPET C/_VITY I \ _ _---'- VENTUR) for leakage to ambient. The small number of
I \ "----VALVE
POPP_ moving parts further enhances reliability.
CLEARANCE AREA LVALV E SPRING
VALVE SCHEMATIC
Gate-Type Propellant Valves

Figure ?-41.-Typical venturi-type propellant
valve designed and manufactured by Fox Valve Figure 7-42 shows a typical design of a pro-
Development Co. pellant gate valve. Its major advantage is unre-
stricted fluid flow, resulting in low-pressure
drop. It also provides a relatively short distance
since the gaseous characteristics at the throat
between the valve inlet and outlet in the direc-
effect a near linear relationship between flow
tion of the flow. The design shown in figure
rate and supply pressure, rather than according
7-42 uses elastomer O-rings as the valve-seat
to a square law. The fluid-flow venturi valves
have been applied successfully in cryogenic and
storable propellant services.
In fluid-flow systems which require flow limi- ACTLIATrP,_G F'LUI_ INLET
tation as well as a shutoff control valve, the

venturi valve with a cavitating diffuser will pro-
vide both, at a weight of only the valve and at a f PISTO_ TYPE ACTUATOR
pressure drop of only the venturi.

INLET PORT
Venturi valves occupy a relatively long space (O_ENING) f ACTUATOrl _OD DYP4,_,I_C O-PiNG
in the direction of flow, between 4 to 6 times of

the line nominal diameter. This length imposes ST_TtC O.R G_,TE
VALVE SEAT t _ _SEAT SPRING
limitations on size and application in engine
systems. However, venturi valves up to 10
IhlLE T /_u= _ OCTLET
inches nominal diameter have been successfully
built for rocket vehicle systems.
Figure 7-41 presents a typical venturi-type
V_LVE G_TIE GUIOE _" 0 - RINI_ RET_INEA
propellant valve designed and manufactured by PLATE SEAL \L
_R_ VALVE GATE GUIDE PLATE
Fox Valve Development Co. It is a pilot-operated
shutoff valve which consists of a convergent- Figure 7-42.-Typical gate-type propellant valve
divergent venturi section with a simple, spring- design.
seal. These are suitable only for medium- actuate other fluid-flow-control components, such
temperature services. In cryogenic application as propellant valves, or to control engine se-
other seal types are required. Gate valves are quence events, such as the admission of igniter
designed for propellant line pressures up to 3000 fuel. There is a great variety of control pilot
psi. Because of their relative bulkiness, gate valves available for liquid propellant rocket
valves are limited to low propellant-flow appli- engine services. Basically, they can be grouped
cations such as for gas generator control and into two categories: the on-off type and the pro-
ground-support services. portional type. We confine our discussion here

to the on-off type. Since the proportional-type
pilot valves are used widely in closed-loop con-
Needle-Type Propellant Valves
trol systems, they will be treated as regulating
A typical needle-type propellant valve is devices, and discussed in section 7.11.
shown in figure 7-43. This valve type is used Proportional-type pilot valves are also known as
for extremely low flow applications such as for servo valves.
attitude-control thrust chambers. The assembly
A pilot valve may be operated eleetrieally or
shown is a dual-valve arrangement, positively by fluid pressure, or through a mechanical con-
linked by a mechanical yoke. The valve body is nection with other control components. Impor-
an integral part of the thrust chamber injector tant design considerations for pilot valves are:
assembly. Both valves are normally closed. (1) Fast response
Their actuation is provided by a quick-response (2) No leakage of control fluid through the
electric solenoid. valve when closed
Sealing at the valve seat is achieved by the (3) Required actuating power souree compati-
elastomer tip of the valve needle. Dynamic seal- ble with systems design
ing at the actuator rods is achieved by means of (4) Sufficient output at the design point
metallic bellows. This seal design is compatible The output of a pilot valve can be defined as
with cryogenic as well as storable propellants.
The pintle vanes provide a guide for the recipro- W = Pd_' (7-17)
cating motion of the valve needle.
In chapter XI we will discuss other special where
valve types, as they are needed for very low W = pilot valve output at the design point,
propellant flow service in miniature-size space in-lb/sec
engines. Pd = valve control fluid discharge pressure at
the design point, psig
7.9 DESIGN OF CONTROL PILOT VALVES {/ =valve volumetric flow rate at the design
point, in3/sec
The main function of a pilot valve is to con- The most frequently used on-off pilot valves
trol a fluid which in turn is used to control or
may be classified according to their design con-
figurations into
(I) Two-way types
(2) Three-way types
\ ..........
.
THR_T CHAMSI[R A F iNLET
iNJECTOR _lOOY U(L
(3) Four-way types
4;:: i]___' ,;::" . :: "'.- ACTUATOR Two-Way-Type Pilot Valves

?', .><::.: ":r " " .,,
The term "two-way" refers to the number of
ports. A two-way pilot valve is basically a two-
__ \'\'\_ WETALLIC BELLOWS
port, open-close-type shutoff valve, similar to a
L. ELASTOMER TiP OF
propellant valve. The sequence valves shown in
T_E VALVE NEEDLE
figures 7-35 and 7-40 are typical examples of
Figure 7-43.-Typical needle-type propellant two-way-type pilot valves. Both examples use a
mechanical link, actuated from the main valves.
valve design.
Solenoid or fluid pressure operated two-way pilot developed by General Controls. Figure 7-45
valves are also frequently used. shows a typical design of a pilot-operated sole-
noid, three-way pilot valve developed by Skinner.
It uses a solenoid to control fluid flow to a dia-
Three-Way-Type Pilot Valves
phragm which opens or closes the valve. The
A three-way pilot valve (fig. 7-44) has three valve may be normally closed or normally open.
ports: inlet or pressure port, outlet or cylinder The selection of standard pilot valves is based
(actuator) port, and vent or return port. If the on the design data furnished by their manufac-
valve is designed normally closed (N.C.), the turer. For specific applications, modifications
fluid path between pressure and cylinder ports is can be incorporated into standard designs.
closed, while the path between cylinder and A typical fhid-pressure-actuated, three-way
return ports is open. Actuation of the valve pilot valve design is shown in figure 7-46. This
effects closing of one fluid path and opening of valve may be used as the ignition monitor valve
the other. The reverse is true for a normally in the A-1 stage engine control system. The
open (N.O.) valve; i.e., the fluid path between valve is held normally closed by a spring. The
pressure and cylinder ports is normally open. valve diaphragm is designed as a combined sens-
Most of the standard pilot valves furnished by ing and actuating diaphragm. During engine
specializing manufacturers are solenoid valves, start and when satisfactory main thrust chamber
actuated by electrical energy, A direct-acting ignition has been achieved, the pressure buildup
solenoid valve (fig. 7-44) is one in which open- sensed at the thrust chamber injector fuel mani-
ing and closing is controlled by solenoid only. fold will cause the ignition monitor valve to open
A pilot-operated solenoid valve (fig. 7-45) is one
in which the solenoid controls the flow of a
small portion of the pressure fluid, which in turn
operates the valve. This results in a smaller
CYL. VENT CYL. VENT
electrical current required to operate a smaller
PRESS PRESS
solenoid for a high-capacity valve. The pilot-

operated solenoid valve, necessarily, requires a NORMALLY CLOSED VALVE NORMALLY CLOSED VALVE
IN NORMAL POSITION IN OPEN POSITION
certain minimum actuating fluid pressure to over-

come friction and spring loads before it will open
or close.
Figure 7-44 presents a typical direct-acting CYL VENT CYL VENT
solenoid, three-way normally closed pilot valve RET RET
NORMALL'f O_[N VALVE NC.RMALLY GREN VALVE
IN NORMAL POSITION IN CLOSED PO$fflON
CONNECTION
- ELECR_CAL
5PRgNG
PffS_dRE
DIAPH Rt, GM -_.f
-- SOLENOID CYUNOER RETURN
CLOSED _J AS_E_,_LY
• OffN ELECTRIC
POPPET _ ; VENT OR
OR
CYLINOER • RETURN _-- ELECTRO

PEES_GRE
MAGNET
$OLENO_b
ENE_GIZEbOIt
Figure 7-44.-Typical direct-acting solenoid Figure 7-45.-Typical design ol a pilo_-operalefl

three-way normally closed pilot valve devel- solenoid, _hree-way pilot valve developed by
oped by General Controls. Skinner.
by pressurizing its diaphragm. The opening of Valve characteristic flow area in the fully
the ignition monitor valve, in turn, directs the open position, A* : 0.19 in2; d d = 2.1 in;
fuel pressure to the main fuel valve actuator Ps=20psig; F/:141b
opening port. The valve spring can be calibrated Valve resistance coefficient at the fully open
corresponding to the effective diaphragm area, position, K= 3.5
so that the valve will open at a predetermined Required valve volumetric flow rate, _,: 200
sensed pressure. During engine cutoff, decreas- in_/sec
ing fuel pressure allows the ignition monitor Inlet port fuel pressure :350 psig
valve to close. This, in turn, vents the opening Determine the required preload of the valve
side of the main fuel valve actuator, closing the spring, Sp. Also, calculate the valve output W
valve. in the fully open position.
The valve poppet is balanced by internal fluid
pressure acting on a dynamic O-ring seal which Solution
has the same diameter ds as the poppet. The

From equation (7-18), the required valve
valve diaphragm is made of multilayer, thin (0.01
spring preload
inch) Mylar sheets which are pressure formed
with heat added. The effective diaphragm area
can be determined experimentally. The required _r
Sp:-_dd 2 Ps _Ft=4(2.1)2×20_14:69.2_14
preload of the valve spring may then be esti-
mated by : 55.2 lb
2
The characteristic flow velocity of the valve
-_dd Ps = Ft+ Sp (7-18)
where
- 1052 in/see = $7.6 fps
dd: effective diaphragm diameter, in
Ps : rated sensed threshold pressure to open the
The density of the fuel (RP-1) is 50.451b/ft 3.
valve, psig
Substitute this and other data into equation (7-9).
Ft= static friction of the valve poppet, ib
The pressure drop through the valve at the de-
Sp = required preload of the valve spring, Ib
sign point
7 2
Sample Calculation (7-4) _ 3.5 x 50.45 × (87.6)
Ap = K_p _- 2
288g 288×32.2 = 146.5 psi
The following design data are given for the
ignition monitor valve of the A-1 stage engine The fuel pressure at the valve discharge
(fig. 7-46):
pd=350- 146.5=203.5 psig

DYN_,_C O-RJN_ SaA_
_rUtJ_ _,LrH _Ue_ _tES_UmE INlay p'Olr Substitute into equation (7-17), the valve
rO V_Nr eOl_
output
¢ONT_Ot _lT _ VAL_! FC_r W: Pcl_,: 203.5 × 200 = 40 700 in-lb/s ec

¢ON_EC_[0 rO rH_US_
v[
; ¢tos_D Four-Way-Type Pilot Valves
A four-way-type pilot valve can replace two

three-way valves, for control of double-acting
*ctu*ro* --
CO_ErEo ro M_ _u_ _*LV_ *C_UMO_

(two-directional) actuators, as shown in figures
7-33 and 7-40. Figure 7-47 presents the basic
Figure 7-46.-Typical [luid pressure actuated schematic of a four-way pilot valve. The ports
three-way pilo_ valve used as the ignition mon- are arranged so that one side is venting while
itor valve in the .4-1 stage engine. the other is pressurizing.
304 DESIGN OF LIQUID PROPELLANT RI_CKET ENGINES
PRESS PRESS
from the opening port. The valve can be closed
only by pressurizing the closing port, and vent-
ing the opening port.
CYL. I. RET. CYL.2. CYLI RET CYL2.

NORMAL POSITION ENERGIZED POSITRON
The following data are given for the self-

Figure 7q7.-Basic schematic oi a {our-way
locking-type, four-way pilot valve shown in figure
pilot valve.
7-49.
$1 spring: Preload=35 lb, rate=210 ib/in

The typical design of a pilot-operated four-
S 2 spring: Preload = 25 lb, rate = 250 Ib/in
way solenoid pilot valve, developed by Valvair,
Static friction of the valve poppet, F{=24 lb
is shown in figure 7-48. This valve type is used
Pressure port pressure, Pi = 400 psig
to control pneumatic actuators.
Actuating fluid pressure, Pa = 250 psig
Figure 7-49 presents the schematic of a typi-
l_eturn port pressure = ambient
cal self-locking-type, four-way pilot valve. It is
Poppet guide diameter, d, = 0.5 in
held normally closed (cylinder 1 port) by springs
Valve poppet total travel = 0.05 in
S_ and S 2. When an actuating pressure Pa is
Determine diameters d 2, d 3, and d 4 (using an
applied to the opening port, the pintle moves and
actuating-force contingency factor of 1.5).
connects the pressure port to cylinder 1 port, and
cylinder 2 port to the return port. An unbalanced
Solution
self-locking force, 77/4 (all -dl 2) Pi, acting in the
opening direction, causes the valve to stay open, Calculate diaphragm diameter d 2 for the re-
even after the actuating pressure Pa is removed quired actuating force to open the valve:
A'_
PILOT PASSAGE i TERMINAL BOX /-- COIL
LOACNL%INLGoTvYEPRE'D-_, _ _ _L" _/y' P,LOT
, /\
BODY _ _'. Z_II_ ! III _=-_i__I
// /' ^ ] O-RING _---_ L-SEA L
PLUNGER ASSEMBLY _/ / _ SECTION A-A

SPACER _ PO2T EXHAUST PO;T
SECTION B-B
INLET
Figure 7-48.-TypicaI design of a pilot-operated, lout-way solenoid pilot valve, developed by Valvair.
Design of Solenoid Actuators
........ \ / _, ........... r ............... Solenoid actuators are usually applied to the

....... '\ \ / // r" / .........,_o,,:o, sliding stem of pilot valve poppets, for on-off,
two-position operation (figs. 7-44, 7-45, and
7-48). Solenoids are electromagnets, containing
an armature or phmger which moves in a coil of
wire. When the coil is energized, a magnetic
force is exerted on the plunger. The magnitude
_- AC"UAnNO OEA_,_O_ 'S-- C"n.t_ER _ _O_t _- ,EAt OlUL4N of the force is proportional to the cross-sectional
area of the plunger, the number of the coil turns,
Figure 7-49.-Schematic of a typical self-locking- the electric current applied to the coil, and the
type, four-way pilot valve. gap G between plunger and core. It increases as
the gap narrows.
Figure 7-50 shows a typical direct-current
t/" 2
solenoid actuator as frequently used in two-
-_d2 Pa = (S _ preload + S_ preload + F/)
position-type, pilot valve services. Its magnetic
x contingency f_tctor
current consists of a stationary core, the case, a
base, and the plunger. When the coil is ener-
d22_(35+25+24)× 1.5_0.64
gized, the plunger, along with the valve stem
rrx250
4 connected to it by a bolt, is pulled toward the
core, against a compression spring. This spring
d_ = 0.8 in
pushes the plunger back to the normal position
when the coil is deenergized.
Determine piston diameter d s, based on the
The following are general correlations for
required force to lock the valve in the open posi-
flat-faced, plunger-type magnets of the type
tion, in the event the actuating pressure Pa is
shown in figure 7-50.
lost:
F= B_'A (7-19)
4 (ds _ - dt _) Pi = (spring forces - Ft) C
x contingency factor
B fPNI
(35+25+210x 0.05 :--C- (7-20)
+250x 0.05-24) × 1.5
ds_ : +(0.5) _
where
-_x 400 F : pull force on the plunger in its normal posi-
tion, lb
=0.281 + 0.25=0.531
d s = 0,73 in
c m[sso_ _ss SE_L
Calculate diameter d, for the required force to

close the valve:
G
SOtE_¢*_ east, i_o_ ./_
4 d,,2pa=(4(d3 _ -d,2)pi+ F.e-spring forces)
x contingency factor
0.222 x 400 + 24- (35 + 25
d2= + 210 × 0"05 + 250 x 0"05) x 1.5 = 0.22
_× 250 Figure 7-50.-Typical direct-current solenoid

actuator design frequently used in two-position-
d 4 = 0.47 in type pilot valves.
B = magnetic flux density in the air gap, kilo- Substitute this and other values into equation
maxwells/in 2 (7-19):
A = plunger cross-sectional area, in 2
C = a factor comprising constants and allow- F=B2A. 27 B2x0.246
ances for stray flux. A value of 72 is C' 72
applicable to round, flat-faced, plunger-
type magnets The required flux density:
P=a factor comprising constants and the
permeability of the fluid in gap G between B = V27 × 72/0.246 = 88.8 kilomaxwells/in 2
core and plunger; a value of 0.00319 ap-
plies if an airgap is assumed
Substitute this into equation (7-20):
f = flux leakage factor. It is determined by
analyzing the magnetic circuit
N = number of coil turns
B=f-_; 88.8= 0"7 x 0"00319 x NI
I =electric current applied to the coil, amperes 0.05
G = gap between core and plunger, in
Solenoid actuators, particularly if energized The required ampere-turns for the solenoid
coil
for extended periods, must be designed with
sufficient radiating surface to prevent the tem-
perature from becoming excessive. To give the 88.8 × 0.05
required current, the resistance of the coil should NI = 0.7 × 0.00319 = 1990 ampere-turns
be based on its maximum temperature. Suitable
protection, such as seals, should be provided to Assume a current of 1.4 amperes for the sole-
prevent the solenoid from becoming contaminated noid; then the required number of turns
with propellants. If the plunger is designed to
bottom-out against the core when energized, it is
advisable to provide for a thin wafer of nonmag- N 1990 1990_ 1425 turns
I 1.4
netic material at its face to prevent sticking.
The required electric resistance of the sole-

noid coil, at operating temperature
The following data are given for the solenoid
actuator of the A-2 stage engine four-way pilot R E 28 _^
control valve: =_-= 1.--:_=zu ohms
Required actuating force at start of stroke,
F=27 lb The wire can now be selected for the required
Nominal electric supply: 28 volts de: 2.0 resistance, For instance, we may choose a No.
amperes maximum 26 AWG copper wire (0.017-inch diameter, enam-
Valve poppet travel = Air gap between solenoid eled) and wind it to an average coil diameter of
core and plunger, G=0.05 in 1.05 inches.
Plunger diameter =0.56 in
Resistance = 51 ohm/1000 ft, warm.
Assumed flux leakage factor, f= 0.7
Total length = 1.05 x ,-r× 1425/12 = 391 ft.
Determine the amphere-tum requirements of
R =391 × 51/1000=20 ohms.
the solenoid.
Lay wires 22 deep, 65 high.
Thickness of coil: 22 × 0.017 = 0.374 in.
Solution Inside diameter = 1.05- 0.374 =0.676 in.
Plunger area: This leaves 0.676-0.56= 0.116 for insulation,

sleeves, and plunger clearance. Height of coil:
77 65 x 0.017 = 1.1 inches. Outside diameter = 1.424
A =_-x (0.56) 2 = 0.246 in 2 inches.
7.10 DESIGN OF FIXED-AREA-TYPE number and can only be evaluated by experi-

REGULATING DEVICES ments. Typical designs and flow coefficients of
frequently used orifices and flow nozzles are
Fixed-area-type regulating devices, such presented in figures 7-51(a), 7-51(b). and7-51(c).
as orifices, nozzles, and venturis, are frequently Orifices of the VDI type are preferred because
used in rocket engines for fluid system calibra- their behavior is more predictable. This is at-
tributable to the bevel at the backside, which
tion and for control purposes. Although the
energy loss caused by an orifice is high, as prevents erratic wall reattachment of the con-
compared to that of nozzles or venturis, orifices tracted jet within the orifice. It is important
are extensively used because of their compact- that the leading edge be absolutely sharp. The
value of k is a function of the diameter ratio
ness and simplicity (fig. 7-2).
(do d,). Tile (Apo/Ap) versus (do/d 1) relation-
ship is plotted in figure 7-52.
Orifices and Nozzles for Liquid Flow
The basic correlations for liquid flow through

orifices and nozzles are (assuming uniform flow
distribution):
, AIAIAIAIAIAIAIAIAI_Ap
(7-21)
_'= 0.0438 Cdo2v/_"L'_= 0.525 Cdo-_p
_i, = 0.0438 CdoZp ,_ = 0.525 Cdo 2 _ (7-22)
APo
k: Ap (7-23)
where
:volumetric flow rate, fta/sec
_i, =weight flow rate, lb/sec
p = density of the fluid, lb/ft a
Ah = static pressure head drop across the ori-
fice or nozzle, ft
Ap = static pressure drop across the orifice or
nozzle, psi
FLOW
do =diameter of the orifice or nozzle throat, in
dI =diameter of the duct leading to the orifice
or nozzle, in
C =flow coefficient for orifices or nozzles.
Considers the effects of discharge jet
contraction in orifices, velocity of ap-
proach, diameter ratio do/d z, friction
and flow profile
Apo = total (or permanent) pressure drop charge-
able to the orifice or nozzle, psi
k = factor, correcting nozzle or orifice work-
ing pressure to permanent pressure loss
The correlations are still reasonably accurate
for gas flows, if the pressure drop across the Figure 7-51.-Typical designs and flow coeffi-
control orifice is small (<10 percent). cients C of flow regulating orifices. A, VDI
The value of flow coefficients C is a function type; B, square edged; C, VDI type flow
of the design configurations and flow Reynolds nozzles.

do/d t (ORIFICE)
The following data are given for the fuel pump
0.2 0.4 0.6 Q78
discharge duct of the A-1 engine: 1.0
Required fuel weight flow rate, w = 892 lb/sec
0.9
kl I I
"l
Fuel pump discharge duct diameter, d 1 = 7 in
Selected design value of total pressure drop
for orifice or flow nozzle installed in the 0.8 _
fuel duct, APo = 100 psi

07 \1'%1
Estimate the required diameter d o, (a__)for a
VDI-type orifice, and (b_) for a VDI-type flow
nozzle (figs. 7-51(a) and 7-51(c)).
<3 0.6 \
I!
0.5 \
Solution
0.4 ,\
The average flow velocity in the fuel duct
892
0.3 , \
V--
- _r 7 2=66.3fps i i
P4(dl) 2 50-45 x_x (_2) 0 Q2 0.4 0.6
do/d I ( NOZZLE )
From table 6-3, we obtain the viscosity of K
I-do/d I FOR ORIFICES
RP-1 =3.22× 10 -7 lb-sec/in 2, or #=3.22× 10 -7
× 4636.8= 1.494 x 10 -3 lb/ft-sec (see eq. 7-6).
Figure 7-52.-(Apo/Ap) versus (do d1) [or ori-
The Reynolds number, based on duet diameter
fices and nozzles.
dl
Re _ Dvp _7 x 66.3 × 50.45 value of 0.32 for k is obtained from figure 7-52.
/1 1.494 × 10 -3 = 1.306 × 10 6
Substitute this into equation (a):
From equations (7-22) and (7-23)

Cdo 2 0.67 x (4.47) 2 = 23.9
\/_-- \/0:5:
w= 0.525 Cdo2_/-_ = 0.525 Cdo2V_oop/k
(b_) For the VDI-type flow nozzle, the same

Cd 02 w 892
approach can be used to determine do. A value
•,'_ - 0.525 \/_p 0.525 x \/100 x 50.45 of 3.55 inches is found for do; thus
892
- 0.525 x 71 =23.9 (a) do_3.55
- 0.506
dl 7
(a_) For the VDI-type orifice, we determine by
trial and error that d o , which will simultaneously From figure 7-51(c), a value of C= 1.005 is
satisfy equation (a), figure 7-51(a), and figure obtained, for Re-- 1.306 × i0 6. From figure 7-52,
7-52. A value of 4.47 inches is found for do. the corresponding value for k is 0.28. Substitute
Confirm: this into equation (a):
do 4.47 Cdo 2 1.005_(3.55) 2

=0.638
d1 7 = V"_/-.-.-.-.-.-.-.-_8
:23.9
From figure 7-51(a), a value of C = 0.67 is Thus, for the VDI orifice, do=4.47 in; for the
derived for R e = 1.306 × 106. A corresponding VDI flow nozzle, d0=3.55 in.
DESIGN OF CONTROLS AND VALVES 3O9
Orifices for Gas Flow Values for Z applicable to equation (7-24) for
various pressure ratios (p2/pl) are presented in
The basic correlations for an isentropie, com-
figure 7-53.
pressible gas flow through an orifice are:

CApIZ
w: V_-- (7-24) The following data are given for an orifice:
Orifice diameter, d o = 0.06 in
If (p2/p,)<_ Critical ratio, i.e., Flow coefficient, C = 0.6
Helium gas pressure = 500 psig
y [ y+l Helium gas temperature = 100 ° F

</ -\y-1 Ill / 2 \_-=-i" Calculate the orifice flow rates for down-
= , z: Igy ,yr0 (7-253 stream pressure, P2 = 14.7 (ambient) and P2 = 350
psia.
If (P2/Pl) > Critical ratio,
Solution
I_ [ £ Y+'I
For a downstream pressure of 14.7 psia, the
(7-26) orifice pressure ratio
where
P f) :500+14.7
14.7 =0.0286
=gas weight flow rate, lb/sec
A =orifice area, in 2
Pl =gas pressure upstream of the orifice, psia From figure 7-53, the compressibility factor Z
P2 =gas pressure downstream of the orifice, is derived as 4.11 for Y--1.66. Substitute this
psia and other parameters into equation (7-24):
R =gas constant, R/°R

T = gas temperature upstream of the orifice, °R 77 ,_
CAp_Z 0.6x_(O.O6)-x(500+14.7)x4.11
Z = compressibility factor, ft aS/sec
y =gas specific heat ratio %/-R-T x/386 × (460+ 100)
C = flow coefficient, a function of design con-

_ 0.6 x 0.00283 x 514.7 x 4.11
figuration and flow Reynolds number• = 0.00774 lb/sec
V'386 x 560
May be approximated from figure 7-51(a)
and 7-51(b)
For a downstream pressure of 350 psia, the
orifice pressure ratio
0.3
P{ 2'__ 350 _
514.7 0.682
From figure 7-53, we obtain a compressibility

0.4_
0.6 factor Z=3.88; thus
o.7 _/_o' I I ! /_/I__

_'iZ0.5 • 3.88
x 0.0077
w =-:-:-v_. = 0.00726 lb/sec
4.11
0.9
7.11 DESIGN OF SERVOVALVES

o__ • 1.0 1.5 2.0 2.5 3.0 3,5 4.0 1.5
CO_P_6$SI_ILITY FACTOR Z, F, 0'5 APPUCASLE TO EQUATION ;7-221 Fixed-area-type regulating devices have
sec
definite limitations. For instance, an orifice
Figure 7-53.-Compressibility factors for an isen- regulates flow and pressure only under specific
tropic compressible flow through an orifice. conditions of flow, but does not function under
other conditions. By contrast, variable-area- (4) Minimum bleed of control fluid

type devices will function under both dynamic (5) Simplicity of construction and line con-
and static conditions. Classified according to nections
their function, the most frequently used variable- (6) Compatibility with environmental condi-
area-type pressure and flow regulating devices tions: temperature, humidity, accelera-
for rocket engines can be grouped into- tion, and vibration.
(1) Throttle valves (including valves for The open-loop gain and phase shift versus
frequency characteristics of a typical servovalve
thrust and PU control)
and driving amplifier combination are shown in
(2) Gas pressure regulators
figure 7-54. These characteristics are obtained
(3) Liquid flow regulators
by applying an input signal to the servoamplifier
Throttle valves have been discussed in sec-
from an oscillator. The amplifier drives the
tion 7.8. Detail on gas pressure and liquid flow
valve by means of a current input to the valve
regulators will be found in sections 7.12 and
transducer coil (torque motor). In turn, the valve
7.13. Many of these devices use some form of
controls the flow of working fluid to the actuator
fluid-pressure-operated actuator. The position
which produces the desired load force. The
of the actuator, and thus the area of the con-
voltage output from a potentiometer attached to
trolling valve opening, is effected by applying a
the actuator is then compared to the amplifier
pressure differential across the actuator piston
input. Instead of electrical feedback, mechani-
or diaphragm, by means of various types of servo
cal feedback may be employed. (See sec. 7.5,
pilot valves, which will be discussed. The
function of a servovalve is similar to that of an "Engine Thrust Vector Control.')
on-off pilot valve, except that it can continu- Flapper-Nozzle-Type Servovalves

ously vary the pressure or flow rate of the actu-
ating fluid to control the desired actuator posi- The flapper-nozzle-type servovalve is essen-
tion. In a pneumatic actuating system, the tially a variable orifice or nozzle. Figure 7-55
servovalve operates as a pressure control de- illustrates schematically the operation of a typi-
vice. It functions as a flow control device in cal double-bleed unit. Actuating fluid is sup-
hydraulic systems. In rocket engine applica- plied at a constant pressure or flow rate. _'aen
tions, where usually only low-power electrical an input signal is fed to the torque motor from a
signals or sensing pressure forces are available servoamplifier, the electric signal is converted
to operate the valves, servovalves with very into a mechanical force. The resultant deflec-
high gain characteristic are required. Three tion of the flapper is proportional to the input
basic servovalve types are most frequently used: signal and causes the flow of actuating fluid to
the flapper-nozzle type (fig. 7-55), the spool type increase at one nozzle, and to decrease at the
(fig. 7-56), and the poppet type (fig. 7-57). They
are used independently, or jointly, to form a two-
stage servovalve in which the flapper-nozzle
4o t1_,8o
valve acts as a pilot valve for the spool valve.
A typical example was presented in figure 7-9. ,_ '-_o_
28, _ phose/I I
E" "- /I I"/5o
General Design Considerations for Servovalves:
(1) Type of control fluid (gas or liquid) and

its conditions (pressure and temperature)
::,,,! '
",>/
. \ I I 2oo
(2) Systems gain: In some applications the " I e
ratio between electrical quiescent input
power to the valve coils and the maxi-
mum valve output (as defined by eq.
I I , , -,oo
0608Z0 2 3 4 56 8 /0 20 30-90
(7-17)) is as high as 1600-2000 Frequency in cycles per second
(3) Bandwidth and frequency response; Figure 7-54.-System characteristics of a typical

dynamic stability servovalve and driving amplifier combination.
.r- ACTUATOR the throatarea of the nozzle. Most designs are

PISTON based on the firstcondition. The effective (ring
FIXED /
ORWICE--_ et A / shaped) flow area then is
i.__.N_ I, I / , [ I SIGNAL
IIF,.
i!INPUT An = 7rdnX (7-27)
sGu_PPLy .____fl (_I _-_-[__
where
PRESSORE
---7 F;/:i-F-q A n =effective flow area, in 2
dn =diameter of the nozzle, in
X =displacement of the flapper from the noz-
FIXED / NOZZLES--" zle, in. (Maximum value should be less
ORIFICE J TRANSDUCER J
than dn/5.)
Figure 7-55.-Schematic of a typical flapper-
nozzle-type pneumatic servovalve.
Sample Calcul'ation (7-9)
other. An increased flow reduces the fluid pres- The following dimensions and data are de-
sure (compressible fluid) or fluid volume (incom- fined for a flapper-nozzle pneumatic servovalve
pressible fluid) on the corresponding side of the (schematically shown in fig. 7-55), which is used
actuator piston. Correspondingly, the fluid pres- as a pilot valve of the servo PU control valve
sure or fluid volume on the other side of the attached to the main oxidizer valve of the A-2
piston increases. The resultant pressure differ- stage engine:
ential across the actuator piston causes it to Helium supply pressure, Ps = 500 psia, and
move in the desired direction. The flapper- temperature T = 560 ° R
nozzle valve is also applicable to servo systems d z,A,,d 2'A 2, d 3, A 3, d 4,A 4=corresponding
with single-control nozzle bleed. Here, the actu- diameters and flow areas of fixed orifices
ator position is controlled by regulating the and nozzles
actuating fluid on one side of the piston or dia- d, =d 2 ds =d 4
phragm only. This is analogous to the single- #¢,, 0¢2, _i'3, w4, _i'o, VCc=fl°w rates through
bleed pneumatic poppet servovalve (fig. 7-57). fixed orifices and nozzles, and to and from
Flapper-nozzle valves are particularly suit- the actuator, lb/sec
able as pilot valves for larger servovalves (see Z,, Z 2, Z3, Z 4 =compressibility factors of
fig. 7-9). Because the transducers or torque the flows through the orifices and nozzles
motors for these valves require rather low power Flow coefficient of the orifices and nozzles,
levels, they usually consist of coil relays exert- C=0.7
ing forces of only a few ounces. The effect of Distance between the two nozzles,
the flapper spring rate is often counterbalanced T=(d3/4)+ t
by the gradient of the magnetic force developed t=thickness of the flapper=0.004 in
in a properly designed transducer or by mechani- At the neutral or equilibrium position of the
cal means. To prevent spreading of the jets servovalve:
leaving the nozzles and to ease the rate balanc- The pressures in the actuators, Pc = PD= 450
ing between flapper spring and transducer mag- psia
netic forces, the travel of the flapper should be The bleed through nozzles, w3 = w4 =0.000778
kept reasonable small. lb/sec
Equations (7-22) and (7-24), which describe Determine:
the flow of liquids and gases through orifices (a__)The dimensions of fixed orifices and noz-
and nozzles, are applicable to the design calcu- zles, and of distance T.
lations of flapper-nozzle valves. Two conditions (b) The pressure differential across the actu-
may exist for the flow through the nozzles of a ator piston when the flapper is deflected down-
flapper-nozzle valve: first, when the restriction ward 0.001431 inch from its neutral position, and
is determined by the position of the flapper; and the flow rates _i_o and Wc to and from the actuator
secondly, when the flapper has moved far enough are 0.00021 lb/sec (as governed by the speed of
from the nozzle for the flow to be restricted by the piston).
"
Solution Since A 3 = _d3X = nd32/8
(a) Since the gas flow through the nozzles is

discharged to ambient, it is assumed that the d3 = d4 :_ 8 x 0.000279 = \r/7 .095 X 10 -4 = 0.0266 in
77
pressure ratio across the nozzles will always be
less than critical (sonic flow).
d3 t=0"0266+0.004=0.01065 in
From figure 7-58, Z 3 =Z 4 =4.11. T=_-+ 4
From equation (7-24), the following correla-
tions are established: (b_ When the flapper is deflected downward
0.001431 inch from its neutral position, the dis-
CAtPsZI 0.7×500 . _ placement of the flapper from the upper nozzle
#' - \/RT -_/_ AIzI =O'753AIZ'
T-t
X =2
CA2PsZ2 + 0.001431 = 0.003325 + 0.001431
#2- V_-_- -0.753A2Z2
= 0.004756 in
CA3PoZ3 0.7x4.11
#3 =
VR T \/_ A3Po=O'OO62A3P° With the flapper deflected
CA4PcZ4
A 3 = ,,-rd3X = ,7× 0.0266 × 0.004756 = 0.000397 in s
#4- V_R-T
- -O'O062A4Pc
The flow rate through the upper nozzle, #3, is

The flow areas of fixed orifices and flapper now equal to the fixed orifice flow rate #x, plus
nozzles (eq. (7-27)) are: the flow rate from the actuator #o; thus
2 _ 2 #3 =#I +#o
A, =-_d, , A 2 =-_d2 ,
0.0062 Asp o = 0.753 A,Z 1 +0.00021
A s = vd_X, A4 = ,'zd4
(T - t - X) 0.0062 ×0.000397 p o =0.753 × 0.000427 Z 1+ 0.00021
When the valve is at neutral position, the 0.753 × 0.000427 Z, +0.00021

actuator is at rest, and P°= 0.0062×0.000397 = 130.6 Z 1 +85.4
#c = #o = O; #1 = #2 = #3 = #4 = 0.000778 lb/sec;
We use a trial-and-error method to find these
T-t ds values for Po and Z,, which will satisfy the
X-
2 8 above correlations and figure 7-53. We find that
Since Pc = Po = 450 psia under these condi- Po=436.5 psia

tions,the pressure ratioacross the fixedorifices
is 450/500=0.9. Z, = 2.687
Checking the results for a pressure ratio of

From figure 7-53, Z a =Z 2 =2.418
Po 436.5
fi' 1 0.000778 = 0.875,
Ps 500
A, = A 2 = 0.753 Z, = 0.753 x 2.418 - 0.000427 in 2
a Z_ value of 2.687 is derived from figure7-53;

d /o oo0427
thus
dr= 2=_ 4 -\/5"44×10-4=0"0233in
Po = 130.6 × 2.687 + 85.4 = 436.5 psia

#3 0.000778
= --
As =A4 =0.0062p o 0.0062x450 0.000279 in s With the flapper deflected
A4 = _4 (T- t - X) ._ ..... _::_:_::_;_i;!!i;iiiili iii!i!iiii!i!ii:!;i!i!i:ii

= . ×0.0266 (0.01065 - 0.004 - 0.004756)
iiiiiiiiiii::"_::::::::::::::':'""'"'_i '- ==========================_ _ _r_RN
=0.0001583 in 2
The flow rate through the lower nozzle, w4, is

equal to the fixed orifice flow rate w2, minus the
flow rate to the actuator, We:
0.0062 A4P c = 0.753 A 2 Z 2 - 0.00021 C'_LINC(R I CYLINaER 2 '

PORT PORT \_ V,_LVE 8CgY
(A) SPO(O. TYPE $ERV_ -VALVE SCF_E_ATIC

0.753 ×0.000427 Z2 - 0.00021
Pc= 0.0062×0.0001583 =328Z 2 -214
We use again a trial-and-error method to ob-

tain the values for Pc and Z 2 which will satisfy
the above correlations and figure 7-53. We have x ,I
found that
Pc --464 psia (8)_E3,GN CO*W_ENSaT;ON _0_ _ _XrALLY
UNBALANCED FC)RCEO
Z 2 = 2.0808
Figure 7-56.-$pool-typeservovalve.
Thus the pressure differential across the Although the spool valve theoretically is
actuator piston force balanced, hydrodynamic and friction forces
cause relatively large loads which must be over-
P=Pc- Po =464-436.5=27.5 psi come. Refer to figure 7-56(a). Due to the dif-
ference in flow velocities, the static pressure at
face A will be less than that at face B. Simi-
Spool-Type Servovalves
larly, the pressure is less at face C than at face
The spool-type servovalve (schematically D. This results in two approximately equalaxial
shown in fig. 7-56) is basically a four-way valve. forces, both of which tend to move the valve
A cylindrical valve spool is accurately fitted spool to the right so as to close the valve.
into valve inserts, which in turn are shrunk into These unbalanced axial forces can be compen-
the valve body. Both inserts and spool are made sated by design remedies. One is to increase
of hardened alloy steels. The thickness of the diameter d2 (as shown in fig. 7-56(b)). It is rec-
inserts in the axial direction, and thus the loca- ommended that the maximum control port flow
tion of the ports, is held to very close tolerances area rrd,X (where X= spool displacement, in) be
by lapping their faces. The outside diameter of just less than twice the annular area between
the inserts is accurately ground for a tight seal spool diameters d_ and d2. As a result, the flow
with the valve body. The surfaces of their axial velocity along the spool is substantially in-
passages are also lapped. The diametral clear- creased and the axial forces on faces B and D
ance between insert and spool is of the order of are considerably reduced.
0.0002 inch, at which the spool must still slide The difference between minimum flow rate
freely. The axial location of the spool lands (leak flow in neutral position) and maximum flow
must also be closely controlled. To minimize rate (actuator in motion) is substantial. Various
leak flow in the neutral position, the spool lands means of adjustment may be employed, such as
may be designed for slight overlap. As a rule, simple relief bypass valves, or the pump output
leak flows are less when the spool is displaced, may be adjusted. For instance, in a piston pump
due to better isolation of the drain lines. A operated by a wobble plate, the pitch of the plate
typical leak flow rate in neutral position is 0.5 may be adjusted as a function of pressure to
gpm. give strokes varying from maximum to zero.
. • • ~
The correlation between pressure drop and To find the maximum valve output design
flow in a spool valve is not as predictable as point, this expression is differentiated and set
one might expect from a sharp-edged orifice. equal to zero
Experimental.data are required to verify a de-
sign. Equation (7-24) for gas flow orifices may
_-_=2000- 20 - 22.86 ¢,2= 0
be used to approximate the flow through a pneu-
matic spool valve, where orifice area A= _d,X or
and flow coefficient C = 0.8. For a spool valve 22.86/r2+20 P- 2000= 0
using hydraulic oil or RP-1 as the actuating
fluid, the following empirical equation applies: -20 + X/202 - 4 × 22.86 × (-2000)
2 × 22.86
] p2
Ap = I_ + (I + ___21/z2 (7-28) -20 + V400 +183 000
= 8.95 in3/sec
where 45.72
Ap = valve pressure drop, psi
Ap= 10×8.95+7.62× (8.95) 2 = 89.5+610= 699.5 psi
= valve volumetric flow rate, in3/sec
X = spool displacement, or valve opening, W = (2000-699.5)×8.95=11 620 in-lb/sec=l.764 hp
in
p = density of the liquid, lb/ft 3 Itshould be noted thatstabilizedflow condi-
a, b, c = empirical constants depending on the tions rarelyexist,because of mass inertiaslim-
design itingaccelerationand deceleration,and feedback
effects.
Sample Calculation (7-1 O)

ACTUATED BY REGULATOR
The following design data are given for a CONTROLLER
spool-type hydraulic servovalve (shown sche-

matically in fig. 7-56): VALVE STEM--,_w
BLEED
RP-1 is supplied at pressure Ps=2000 psia PORT
/_____,VALVE POPPET
Valve pressure drop and flow characteristics GAS SUPPLY ___ OUTPUT CONTROL
PRESSURE _ TO
may be obtained from equation (7-28), for the REGULATOR VALVE
following design constants:

PRESSURE _._____%
a =254 _ED
ACTUATOR
b = 1270 ORIFICE l
/-BALANCE PISTON
c=1.79× 10 -7
Maximum spool displacement, X = 0.009 in Io)
Determine the flow rate _,, pressure drop Ap,
and output W of the valve at maximum displace- ACTU,_TEO BY REGULATOR
CONTROLLER
ment.
VALVE ST EM---__
Solution
DYNAMIC SEAL
BLEED PORTI - _ PRESSURE Pc TO

Substitute a, b, c, X, and p into equation
OUTPUT
REGULATORCONTROL
VALVE
(7-28) SUPPLY -.,-..._]'_ h'-_ ACTUATOR
GAS ,_
Ap=[_-_+(12_+ 1"79×10-7''2]
(0._9-_ ]v ] ×(50.45) 2 PRESSURE P=----_"_VALVE POPPET
BALANCE
Ap= 10_,+7.62_ "2 PISTON
(b}
From equation (7-17), the valve output Figure 7-57.-Schematics o[ typical single-bleed,
poppet-type, pneumatic servovalves used in
W = pdi' =(Ps - Ap)/_: 2000/_ - 10 _2 _ 7.62/_3 gas pressure regulators.
Single-Bleed, Poppet-Type Servovalves between actual and desired pressure. Its output,
called the error signal, can be directly applied
The single-bleed, poppet-type servovalve
mechanical force, or control pressure output to a
operates as a variable orifice like the flapper-
servovalve. Figure 7-58 presents the schematic
nozzle valve. Figure 7-57 presents, schemat-
of a typical gas pressure regulator controller.
ically, the principle of operation of typical
Here, the pressure being regulated is sensed
single-bleed, poppet-type, pneumatic servovalves
externally by a bellows which is internally evac-
as used in gas pressure regulating services.
uated, or vented to atmosphere. The vacuum
Two basic configurations are in use. The first
establishes an absolute pressure reference, while
(fig. 7-57(a)) effects output control pressure Pc
a vented bellows uses ambient pressure for ref-
regulation through variation of bleed port flow
erence ("gage pressure"). As the regulated pres-
area. In the second (fig. 7-57(b)), the bleed port
sure (the sensing pressure) changes, the bellows
area is fixed. Here, Pc is regulated by varying
deflects against the calibrated internal spring
the supply gas flow rate.
load, simultaneously positioning a directly con-
The selection of configuration depends on
nected servovalve (fig. 7-57), which in turn regu-
application. To minimize unbalance forces, a
lates a control pressure output.
balance piston is usually attached to the valve
A regulator valve consists of the control
poppet. The area of the balance piston is made
valve and an actuator. If only limited accuracy
equal to the projected area of the poppet seat
is required, or very small capacities are involved,
diameter, less that of the valve stem.
it suffices if the regulator controller develops
the error signal directly as a mechanical force to
7.12 DESIGN OF GAS PRESSURE position the regulator valve. Such a directly
REGULATORS spring-loaded pressure regulator is represented
schematically in figure 7-59(a).
Basically, gas pressure regulators are
Where greater accuracy is required, the regu-
variable-area-type pressure-reducing valves.
lator valve actuator is positioned by controlled
Their prime function is to maintain constant
pressure from a servovalve connected to the
pressure at their outlet, or in a downstream
regulator controller.
region, even though the pressure at their inlet
Since the servovalve amplifies the regulated
may vary (decrease). Gas pressure regulators
pressure errors, small errors in regulated pres-
may be operated independently, or in conjunction
sure will cause large changes in its control
with pressure relief devices, under either dy-
pressure output. This control pressure Pc can
namic or static conditions. For example, the
then be applied to control the regulator valve
gas pressure regulator for the A-4 stage propul-
position in the following ways:
sion system (fig. 3-9) is designed to maintain a
(1) Control pressure Pc is used as the loading
constant main oxidizer tank pressure of 165 psia,
pressure for a simple dome-loaded pres-
while the helium gas pressure at the regulator
sure regulator, as shown schematically
inlet varies from 4500 to 245 psia. Additional
protection is provided by a tank relief valve,
should the tank pressure continue to rise with
/-- SCREW ADJUSTMENT FOR
the regulator completely closed, because of such
effects as aerodynamic heating. j_// SPRING CALIBRATION
METALLIC BELLOWS _ _ CALIBRATING SPRING

INTERNALLY EVA C b___
Elements of Gas Pressure Regulators
Most pressure regulators include two basic SENSING PRESSURE _'_ "_
elements: the regulator controller and the regu-

lator valve. The former controls the regulator
valve opening, which in turn regulates the gas
SENSING INPUT
flow through the regulator.

A regulator controller essentially is a sensing- Figure 7-58.-Schematic of a typical gas pressure
computing unit which measures the difference regulator controller.
• . --- .i:, I
CONTROL PRESSUFE Pc FROM A SINGLE- BLEED TYPE
ou . T LO O . .soo O
SERVO - VALVE ATTACHED TO THE REGULATOR CONTROLLER.
////--- REFERENCE SPRING LOAD / S DOME
_,_11_.\\\\_ ---T----_OuTLET Pr
(A ) DIRECT SPRING - LOADED (B) DOME - LOADED
CONTROL PRESSURE PC FROM

A SIGNAL-BLEED TYPE SERVO-VALVE F- CONNECTIONS TO A FOUR-WAY SERVO-VALVE
ATTACHED TO THE REGULATOR CONTROLLER _ ATTACHED TO THE REGULATOR CONTROLLER
!
._-\\_ DIA PH RAG M _ ._>.\_-_/j_--- ACT UAT OR PIST ON
VENT TO _ CALIBRATING SPRING I _ DYNAMIC SEAL

ATMO. pa-.F-- DYNAM,C SEAL _-i [
_'_----_OUTLET Pr _----_"__--OUTLET Pr
,NLET _-----__ ,NLET pi ___.__ REGULATOR VALVE
REGULATOR VALVE
_,C]CONTROL PRESSURE OPERATED (O) FOUR-WAY SERVO-VALVE OPERATED
Figure 7-59.-Schematics oi various gas pressure regulator designs.
in figure 7-59(b). A large-capacity, introducedby the oftenextreme variationof Pi.

dome-loaded pressure regulator can also A simple,directlyspring-loaded, low-capacity
be loaded by a separate, small-capacity regulatorsimilarto figure7-59(a)is requiredto
precision pressure regulator (precision reduce Pi to a reasonably constant supply pres-
loader). sure Ps for the control servo circuit. This regu-
(2) Control pressure Pc is used as the input lator is commonly known as the bleed regulator
to a mechanism that positions the regu- and can be a part of the main regulator controller.
lator valve as a linear function of con-
trol pressure. In its simplest form, this
Design ConsiderationsforGas Pressure
mechanism consistsof a spring and a
Regulators
diaphragm (fig.7-59(e)).
In some designs,a four-way servovalve (as in The followingare basic considerationsforthe
figs.7-55 and 7-56)operates the double-acting design of gas pressure regulators.
piston-typeactuatorof the regulatorcontroller to 1. The principal causes of error in a pressure
controlthe positionof the regulatorvalve (fig. regulator are-
7-59(d)).This design is known as an integrating- a. Variation in inlet pressure
type pressureregulator. b. Variation in flow demand
When a single-bleed servovalve is used with c. Variation in temperature
the regulator controller, the supply pressure Ps d. Mechanical hysteresis and friction
to the control servo circuit must be isolated from e. Creep of stress members
the inlet pressure Pi to the main regulator valve, f. Variation in effective length of mem-
because of the steady-state error that would be bers because of angular displacement
GAS INLET PRESSURE GAS INLET PRESSURE

g. Variation in vehicle acceleration 3O00 PSIA 12(_3 PS;A
h. Vibration
i. The fact that pressures are not sensed
at the point at which control of pres- ci
.
c_t_ ............
sure is desired (influence of pressure

drops caused by flow through down-
wet.
stream systems) o 2
The basic test of a good regulator design is TIME, SEC TIME, SEC
whether these errors have been held within

Figure 7-60.-Dynamic response characteristics
allowable tolerances, without mechanisms (in-
o[ a typical pneumatic pressure regulator.
cluding diaphragms and springs) of unreasonable
complexity and size.
relief valves are provided for the main propellant
2. Where feasible, proven concepts and con-
trol mechanisms of previous designs should be tanks to prevent possible intermixing of propel-
lant vapors.
utilized. This is especially important for dy-
namic characteristics, stability and transient 6. Other design considerations for gas pres-
sure regulators are-
response, and for mechanical details such as
poppets, seals, diaphragms, and springs. a. Type of regulated gas
b. Gas inlet-pressure and temperature
3. Balance of the regulator valve poppet
range
should be provided to minimize the forces im-
c. Regulated gas outlet pressure level
parted to the valve by inlet pressure Pi. This
with respect to ambient or vacuum, its
may be achieved by attaching a balance piston,
tolerance or accuracy of regulation,
similar to that of the poppet servovalves (fig.
and range of adjustment
7-57). However, this will require some type of
d. Required maximum flow capacity
dynamic seal, which are known trouble sources.
e. Required response time and allowable
In certain applications, therefore, an unbalanced overshoot
valve may be preferred. f. Type of line connections
4. In most cases, the regulator must lend
g. Environmental conditions (temperature,
itself to the alternate mode of operation as a
vibration, humidity, etc.)
shutoff valve, until a start pilot valve is opened.
Figure 7-60 presents the dynamic response
5. The various gas pressure regulator designs
characteristics of a typical pneumatic pressure
shown in figure 7-59 regulate pressure by ad-
regulator upon initiation of a demand. The regu-
mitting additional gas, as required to the regu-
lator was set at a regulated outlet pressure of
lated pressure region. If the flow demand on the
400 psia. The gas used (helium) is discharged
regulator is reduced to zero (dead-end condi-
into a system of approximately 60-cubic-inch
tions), the regulator valve shuts off. However, volume.
the pressure in the regulated region may increase
above the design point, even with the regulator
Sizing of the Gas Pressure Regulator
valve closed, due to-
a. A timelag in closing, following reduc- The required flow area across the regulator
tion of demand to zero valve when fully open. i.e.,the characteristic
b. Inevitable regulator valve leakage flow area of a pressure regulator, may be calcu-
c. Thermodynamic effects, such as in- lated for required flow capacity and regulated
crease in temperature of the gas in the outlet pressure, at minimum allowable gas inlet
regulated pressure region, or from pressure.
vaporization of a liquid. The following correlation derived from equa-
This condition is known as a lockup, and often tion (7-24) for gas flow orifices is applicable to
poses problems. Lockup can be eliminated by gas pressure regulator design:
incorporating a relief valve downstream of the
regulator valve as part of the regulator assembly. w,]'R-T
A* - (7-29)
In the A-4 stage propulsion system, individual CpiZ
where employs a metallic diaphragm for low friction.

A* :characteristic area of the pressure regu- The dome pressure Pc, which determines the
lator, in 2 force to the actuator diaphragm, and thus regu-
: design weight flow rate of the pressure lates valve position and outlet pressure, can be
regulator, lb/sec regulated either internally by a pressure con-
R =gas constant, ft/°R troller, or externally by a low-capacity pressure
T :gas inlet temperature at minimum inlet regulator or loader. There are many variations
pressure condition, °R of dome-loaded regulators. Design approaches
Z :compressibility factor, fto's/sec, a func- may best be illustrated by typical examples.
tion of pressure ratio pr/Pi and specific Fig-ure 7-61 presents the schematic of a typi-
heat ratio (fig. 7-53) cal dome-loaded gas pressure regulator which
C : flow coefficient, a function of design con- has an alternate mode of operation as a shutoff
figuration. Design values range from 0.6 valve. The prime purpose of the regulator is to
to 0.7 maintain a regulated outlet pressure at a pre-
Pi :minimum allowable gas inlet pressure, psia selected value called the set point. The regu-
Pr =required regulated outlet pressure, psia lator is composed of four elements: the start
pilot valve, the bleed regulator, the sensing-
Sample Calculation (7-11) control unit, and the main regulator valve. The
combination of bleed regulator and sensing-
The following design data are given for the
control unit forms the function of main regulator
helium gas pressure regulator of the A-4 stage
controller.
propulsion system:
Design weight flow rate, w: 0.048 lb/sec The start pilot valve is a solenoid-actuated
Minimum allowable gas inlet pressure, pi=245 poppet valve which is normally held closed by a
psia spring. In the closed position, it locks up the
Gas inlet temperature at minimum inlet pres- outlet pressure of the bleed regulator. Thelatter
sure condition, T= 1030 ° R is a normally closed, directly spring-loaded-
Required regulated outlet pressure, Pr: 168 type, pressure-reducing valve and supplies an
psia approximately constant preset pressure Ps to the
Regulator flow coefficient, C =0.65 sensing-control unit. This pressure is greater
Determine the characteristic area A* of the pres- than the main regulator outlet pressure Pr. It is
sure regulator. the sensing-control unit. then, which establishes
the main regulator outlet pressure. It detects
Solution small deviations from the set point and magnifies
the error signals by means of a single-bleed,
At minimum inlet pressure conditions, the
flapper-nozzle servovalve linked to a sensor
pressure ratio across the regulator
Pr_ 168 - 0.685

Pi 245
From figure 7-53, the compressibility factor Z

is 3.9 for helium. Substitute this and other data
into equation (7-29). The characteristic area
A* = _'Vr/_'_- 0.048 × \/386 x 1030 : 0.0487 in s

CpiZ 0.65 x 245x 3.9
Dome-Loaded Pressure Regulators

Figure 7-61.-Schematic of a typical dome-loaded,
This regulator type (fig. 7-57(b)) is one of the negative-gain-type gas regulator with an alter-
most frequently used. The main actuator usually nate mode o[ operation as a shutoff valve.
diaphragm by increasing or decreasing the con- spring preload. Under these steady-state condi-
trol pressure Pc which operates the actuator dia- tions, gas continues to flow from the bleed regu-
phragm of the main regulator valve. This valve lator through the servovalve at a rate determined
is a normally closed, dome-loaded pressure re- by the restrictions and out to the outlet manifold•
ducing valve. The regulator controller circuitry has what is
In operation, the following sequence of events called a negative gain. A sensed increase in
occurs. Gas enters the regulator through a filter pressure Pr causes an amplified decrease in con-
located at the regulator inlet port, but is pre- trol pressure Pc, with attendant decrease of the
vented from entering the main regulator control main regulator valve opening. Similarly, a de-
pressure dome by the closed start pilot valve.
crease in Pr causes an increase of valve opening.
Upon opening of the latter by energizing its The controller gain can be defined as
solenoid, gas flows through the bleed regulator
and is reduced to pressure Ps, as determined by Ape
the preset reference spring force. The gas then G= - Ap---_- (7-30)
flows through a fixed restrictor and passes into
the control dome. This tends to open the main where
regulator valve against the actuator spring. G =controller gain
Dome pressure Pc is controlled by varying the Ape = change in control pressure, psi
flow area of the flapper-nozzle servovalve which Apt = change in sensed pressure, psi
bleeds the loading gas into the main regulator The principal design parameters of the gas
outlet manifold. When the sensed regulated out- pressure regulator shown in figure 7-61 are listed
let pressure Pr and the sensor spring force are in in table 7-6.
equilibrium, the servovalve flapper is positioned Fig-ure 7-62 presents the design layout of a
a sufficient distance off the nozzle seat and
typical dome-loaded, zero-gain-type gas pressure
maintains a steady-state control pressure Pc. regulator. The main regulator controller con-
This pressure is always greater than Pr, as de- sists of a spring-loaded bleed regulator and a
termined by the main regulator valve actuator fixed-area bleed orifice, which gives a constant
TABLE 7-6.-Principal Design Parameters of a Typical Dome-Loaded, Negative-Gain-Type Gas

Pressure Regulator (Fig. 7-61)
Parameter Design data
Regulated gas ................. Helium

Inlet gas pressure, Pi ........... 4500 psia nominal. 5800 psia maximum, 500 psia minimum
Inlet gas temperature, T ........ 110 ° to 360 ° R
Regulated outlet pressure, Pr .... 400 psig t 25 psi
Flow demand .................. a. Maximum: 0.1 lb/sec at 500 psia inlet pressure and 160 ° R inlet temperature
b. Minimum: zero
Modes of operation ............. a. Start pilot valve solenoid deenergized: regulator remains shut off. Outlet pres-
sure : 0
b. Start pilot valve solenoid energized: regulator functions normally
Controller bleed ............... a. None when start pilot valve solenoid deenergized
b. 0.0002 lb/sec at 530 _ R when solenoid energized
Main regulator valve ............ Seating diameter, 0.205 in; characteristic area, 0.0281 in 2
Main regulator valve actuator .... Effective area, 7.55 in-_; bias spring preload, 75.5 lb; rate, 500 lb/m
Start pilot valve ............... Seating diameter, 0.025 in; characteristic area, 0.000437 in 2
Bleed regulator valve ........... Seating diameter, 0.028 in; characteristic area, 0.000301 in 2
Bleed regulator sensor .......... Effective area, 0.185 in2; reference spring preload, 93.5 lb; rate, 1500 lb/in
Bleed flow restrictor ........... Diameter, 0.0116 in; area, 0.000105 in 2
Main regulator controller sensor.. Effective area, 0.315 in2: reference spring preload, 126 lb; rate, 702 lb/in
Flapper-nozzle servovalve ...... Seating diameter, 0.043 in (2 holes); maximum stroke, 0.011 in; maximum flow area,
0.003 in_; bias spring preload, 1.0 lb; rate, 5.0 lb/in
REGULATOR CONTROLLER
SENSO_ ASSEMBLY
SENSOR OUTEB BELLOWS
ORb=ICED _ARTITION
___ SENSOR &AMRER

$ERVO- VALVE
VENT
/ r-- SPOOL TYRE FOUR-WAY

\ \ % & r SEmvO-
VALVE
SENSED PRESSURE
FROM OXIDIZER
TANK INLET
SENSOR INNER ' REGULATOR VALVE

BELLOWS ACTUATOR
DYNAMIC SE_L
_ -..._,
OXIDIZER TANK OX+D,ZER lANK
_r
' [ _ BALANCE TYPE
_ REGLJLATOR INLET,
Pi
Figure 7-63.-integrating-type, gas pressure

regulator with spool-type, lout-way servovalve.
Figure 2-62.-Typical dome-loaded, zero-gain-

type gas pressure regulator loaded by a bleed tank pressure is to be sensed through a rela-
regulator. tively short external line. Fuel tank pressure is

maintained at I0 psi below the oxidizer tank
pressure by different pressure settings of line
bleed from the control dome to the regulator out- check valves and tank relief valves. A spool-
let. The control dome pressure Pc is maintained type, four-way servovalve, which is directly con-
at a constant level by the bleed regulator. This nected to the sensor of the regulator controller,
is known as a zero-gain-type control circuit. controls the regulator valve actuator. Supply
Here, too, control pressure Pc is higher than the pressure Ps of the servovalve is taken from the
regulated outlet pressure Pr by an amount deter- regulator outlet pressure Pr; i.e.,Ps = Pr. The
mined by the main regulator valve actuator spring sensor contains two bellows with an oil-filled
preload. The regulator shown is designed to damper in between. The damper consists of a
maintain an outlet pressure Pr of 2S2 psig + 5 psi, partition having a properly sized orifice. The
at inlet pressures ranging from 5000 to 375 psig. outer bellows, which is exposed to the propellant
The actuator diaphragms are made of a 0.0025- vapors, is hydroformed from thin 321 stainless-
inch 17-7 PH high-strength stainless-steel sheet. steel tubing, because the pressure differential
440C stainless steel is used for all sliding mem- across it is small and because it contains fluid.
bers, such as regulator valve poppets and guides, The inner bellows must withstand a pressure
to eliminate the surface galling problem. Regu- differentialequal to fulltank pressure, since the
lator main body and housing are made of aluminum- interior of this bellows is vented directly to
alloy forgings. atmosphere. For this reason, the bellows is
machined from 17-4 PH stainless-steel stock.
Integrating-Type Pressure Regulators
This regulator is called an integrating type be-
Figure 7-63 presents the design of a typical cause a constant tank pressure error will produce
integrating-type gas pressure regulator, designed a constant actuator piston velocity (neglecting
to regulate the propellant tank pressures of a extraneous forces). This, in turn, gives a con-
pressurized gas propellant feed system such as stant regulator valve opening (or closing) veloc-
the A-4 stage propulsion system. Main oxidizer ity and a constantly increasing (or decreasing)
gas flow rate. Tank pressure is then the integral

of tank pressure error versus time. The integrating-
type pressure regulator provides the constantly
increasing regulator valve flow area required for
decreasing supply pressures, with minimum error
in tank pressure, but may be unstable if the gain
of the servovalve-actuator combination is too
high. A suitable gain value should be deter-
mined by computer studies of the regulator dy-
namic characteristics.
7.13 DESIGN OF LIQUID FLOW AND

PRESSURE REGULATORS
Figure 7-64.-Schematic of a typical sliding-
Liquid flow and pressure regulators basic- piston-type liquid flow regulator.
ally are variable-area-type, pressure-reducing
valves. A liquid flow regulator maintains a con-
stant rate of flow, or limits the maximum rate of that a constant pressure drop is maintained
a flow. A liquid pressure regulator, like a gas across the orifice disk.
pressure regulator, maintains a constant fluid As the fluid inlet pressure increases, or the
pressure at its outlet under variable flow condi- back pressure decreases, the flow rate tends to
tions. Design considerations for liquid flow and increase also. However, the attendant increase
pressure regulators are similar to those for gas is pressure differential, i.e., pressure force on
regulators. the piston, moves it against the reference springs,
simultaneously increasing tlle flow restriction.
Design of Liquid Flow Regulators A new equilibrium is achieved as the inlet pres-
sure increase is absorbed by an increase in
Liquid flow control can be obtained with a pressure drop across the piston throttle port.
venturi. The pressure at the throat is propor- The following design correlations are estab-
tional to tile velocity of the (incompressible) lished for this flow regulator:
fluid and thus the flow rate. The pressure dif-
ferential between throat and inlet can be used to
Fs =(Api +Ap2)Ap (7-31)
control the position of a butterfly valve, and
thus the flow rate, by means of a servocontrol Ap
circuit, A regulator of this type is shown sche- K s = -_ Ap (7-32)
matically in figure 7-18.
Figure 7-64 presents the principle of another where

type flow regulator which is frequently used in Fs =regulator reference spring preload, for
rocket engine systems, because of its simplicity. a given flow rate, Ib
It consists of two restrictors, one of fixed and Ap_ =pressure drop across the fixed orifice
the other of variable area. This combination disk at that flow rate, psi
automatically maintains a constant pressure drop AP2 = pressure drop across the piston throt-
across the fixed restrictor, and thus a constant tle ports at that flow rate, psi
flow. The fixed restrictor consists of a disk Ap = effective piston area, in 2
containing a number of orifices and mounted in a Ks =regulator reference spring rate, ib/in
piston. The variable restrictor is formed by (Ap/A}) = change of pressure drop across the
variable throttle ports located around the piston piston throttle ports per inch of
periphery. The pressure differential across tim movement of piston, lb/in_-in
fixed orifices causes the piston to move against Adjustment of the regulated flow rate is made by
the regulator reference springs. Piston throttle adjusting the reference spring force, using a
ports and reference spring rate are designed so graduated vernier mechanism.
' j
? " !
_RING
calibrated reference spring. The valve opens as
the pressure force acts against the spring force.
In the second, the main relief valve actuator is
MYLAR GASKI[TS
controlled by a pilot valve which is calibrated
[N_ A for the desired relief pressure setting. The
DIIP_N I_ ASSEMSLV
direct-operated relief valves are used for large-

tolerance, low-capacity services. For large flow
BUNA-N PACKING
requirements, pilot-operated relief valves are
used for quick response and to avoid excessive
a'NEUMATIC REFEPENEE
size.
PRESSURE INLET_ 4_ DIAPHRAGM
qSEE INSET A) Most pressure relief valves are used in gas
pressures systems. Important design considera-
tions are-
(1) Type of gas and its conditions
(2) Pressure relief level, its tolerance and
range of adjustment
(3) Required response time, dead band (dif-
ferential between actuation or opening,
....
., and deactuation or closing pressure),
and other dynamic characteristics
(4) Required maximum flow capacity
Figure 7-65.-Typical liquid pressure regulator (5) Simplicity of construction and line con-
nections
design [or liquid oxygen service.
(6) Environmental conditions, such as tem-
perature, humidity, acceleration, and
vibration
Design of Liquid Pressure Regulators
The required flow area for a gas pressure
Figure 7-65 presents a liquid pressure regu- relief valve at its maximum opening position,
lator design frequently used in rocket engine i.e., its characteristic flow area, can be calcu-
systems. The actuator diaphragm of the regulated from equation (7-24) for gas flow orifices.
lator valve (plunger) can be loaded either by a
pneumatic reference pressure (as shown), or by a
reference spring. Direct-OperatedGas Pressure Relief Valves
Regulated outlet pressure is sensed directly. Figure 7-66(a)shows the design of a typical
An increase in outlet pressure tends to reduce
low-capacity,direct-operated gas pressurerelief
the flow area of the regulator valve until an
valve. The valve poppet is loaded directlyby a
equilibrium is reached between outlet pressure coilspring. This is one of the simplestdesigns.
and reference pressure. The actuator diaphragm Itmay, however, become dynamically unstable
is made of Mylar laminations which are compati- under externallyintroducedvibration.
ble with cryogenic propellants such as liquid Figure 7-66(b)is the schematic of an im-
oxygen. proved design of a direct-operatedreliefvalve.
Itassures a predictabledead band and elimi-
nates chatterunder vibration.This design em-
7.14 DESIGN OF PRESSURE RELIEF
ploys coned-disk-type(Belleville) springs,oper-
VALVES
ating between positivestops,in the negative
The prime functionof pressurereliefvalves rateportionof the force-deflection curve of the
is to protectfluidsystems and pressurevessels spring. As shown in figures7-66(b)and 7-67,
from being overpressurized.There are two basic the springremains againstpretravelstop C,
types of pressure reliefvalves; the directoper- untilthe appliedforce (valveinletpressure
ated and the pilotoperated. In the first,the Pi× valve poppet seat area nd2/4) reaches a
valve poppet is loaded directlyby a properly level A. The springthen snaps from C to the
CAP APPLIED FORCE, LB.

(VALVE INLET PRESSURE X
VALVE POPPET SEAT AREA)
SETTI NG __
SPRING _L-------_-_23_ /--8ODY SEAL SPRING
_ O- RING
(o)
__ LOCK NUT
pOPPET_ RELIEF A -ACT'UATION

FO_ /
B 'DE-ACTUATION/-- " -I-- _ I----at_
SOFT o- ,NOo,
DISC
INLET
POST-TRAVEL C D
(b) _ ADJUSTMENT SPRING DEFLECTIC, N, IN.
_ PRE-TRAVEL
Figure 7-G7.-Coned-disk-spring, force-deflection
__ ADJUSTMENT
"; r/Z_.--_',.,%-_./r_ _ STACKOFCONEO-O_SC Curve.
I _',._ _('_/JL_'%('T/_"_ ]/'/J / ." SPRINGS IN SERIES OR
c t PRE,SU E
_'_ CC_NT RO(_ ORIFICE
do, %
P[STONflNG SE_-L"_ \
JRELIEF_ . d
VALVESEAT
---/ Zl _
m
Figure 7-66.-Low-capacity, direct-operated gas

pressure relief valves. (a) Coil-spring-loaded,
direct-operated relief valve; (b) Snap-action,
coned-disk, spring-loaded, direct-operated
relief valve.
PILOT VALVE CONTROLLEF
posttravel stop D, with no further increase in

applied force. Reduction of the applied force to
B will cause the spring to snap back from D to
C. The positions of the pre- and posttravel
stops can be adjusted for constant actuation and
deactuation forces, independent of spring manu- VENT Ldp,_'!=
e_
facturing tolerances. The coned-disk spring

washers are usually made of beryllium copper or Figure 7-68.-Schematic o[ a typical high-capac-
17-7 PH stainless steel. This type of relief ity, pilot-operated tank gas pressure relief
valve is suitable for high-pressure helium stor- valve.
age bottle services, such as in the A-4 stage
propulsion system.
controller also senses the tank pressure p:.
When the latter reaches or exceeds the preset
Pilot-Operated Gas Pressure Relief Valves
level, the pilot valve is actuated to crack. This
Figure 7-68 presents the schematic of a typi- vents the main relief valve actuator control pres-
cal high-capacity, pilot-operated tank gus pressure Pc and, in turn, permits opening of the main
sure relief valve. Normally, both main and pilot valve. Main valve poppet position is a function
valves, are held closed by valve spring forces of the control pressure, which in turn is con-
Fsm, Fs3, and pressure Pc. The pilot valve trolled by the position of the pilot valve poppet
and by the tank pressure. This correlation can Force balance equations of the main relief
best be illustrated by the following sample cal- valve poppet at various conditions:
culation. (1) Basic equation:
Fsm+pcA2- Pt(As-A1)- paA3 = Fseat

The following design data are given for the (2) Condition at cracking:
A-4 stage propulsion system main oxidizer tank
relief valve (schematically shown in fig. 7-68): Fsm+ PccA2- Ptc (A2 - A1)- paA3 = 0
Tank pressurant temperature, T--700 ° R (he-
lium gas)
(3) Condition at any intermediate valve posi-
Relief pressure set point, Pt= 165 psia
tion:
Required maximum flow capacity, w = 3 lb/sec
Main valve flow coefficient, C = 0.75
Fs m + K s mXm + Pc A 2 - Pt (A 2 - A 1) - Pa A 3 = 0
Estimated leakage past the main valve actu-
ator piston seal, _i's = 0.003 lb/sec
(4) Fully open condition:
Control orifice diameter, d o =0.080 in
Flow coefficient for control orifice and pilot
Fsm + KsmXmo+ PcoA2 - Pro(As -A,)- PaAa =0
valve, Co=0.6
d, A = Diameters and areas of various ele-
(5) Condition at start to reseat:
ments, in; in s
Pc = Control pressure, psia
Fsm+KsmXmo+PcrA2- Ptr (A2- A1)- paA3 =0
Pa = Ambient pressure (14.7 psia maximum)
Determine:
(6) Fully reseated:
(a) Dimensions of the main relief valve and
force balance equations for various
Fsm + PcrA2 - Ptr(A2-A1)-PaA3 =0
conditions
(_b) Dimensions of pilot valve poppet and where
actuator, and force balance equations
Fseat = main valve seating force, lb
Fsm, Ksm : preload and rate of main
Solution valve spring, lb; lb/in
Xm, Xmo = main valve poppet travel in-
(a__)Main relief valve.-The valve is sized for termediate and fully open
an isentropic compressible flow through an ori-
position
fice. From equation (7-24), the main relief valve
Pcc, Pco, and Pcr:Control pressures at crack-
port area
ing, full open, and reseat
of the main valve, psia
A3- CptZ
Ptc, Pto, and P_r =tank pressures at cracking,
full open and reseat of the
Since Pa/Pt= 14.7/165 = 0.089, Z = 4.11 (from main valve, psia
fig. 7-53) (b_.) Pilot valve.-The pilot valve flow capac-
3 × V;386 x 700 ity must be larger than the combined flow rate of
A3 =0.75× 165× 4.11=3'07 in 2 control orifice and leakage past the main valve
actuator piston seal for adequate venting of Pc.
d 3 : 1.98 in'_2 in From equation (7-24), the flow rate of the control
orifice
Required minimum travel of the main valve to
the fully opened position CozrdoSPt z 0.6x 0.00503x 165×4.11
_i,o = 4 _/-R-T V386 x 70-----'-O
A3 3.07
Xmo =-----=
-rzd3=---_= 0.487 in_0.5 in = 0.00394 lb/sec
The total flow into the control cavity, _i,0 + Ws 7.15 DESIGN OF MISCELLANEOUS VALVES
= 0.00394 + 0.003 = 0.00694 Ib/sec.
Setting the flow capacity of the pilot valve 50 Check Valves
percent higher, its flow capacity results as
The prime function of check valves is to
allow fluid flow in only one direction. Tilere are
_i,p= I. 5 × 0.00694 = 0.0104 Ib/sec
two basic check valve types: the poppet alid the
swing-gate type. The selection depends to a
It is desirable that this flow be maintained
great extent on application. General design con-
freely at all points, independent of back pres-
siderations for check valves are-
sures. This is achieved by maintaining sonic
(l) Type of fluid and its pressure and tem-
velocity, i.e., critical or supercritical pressure
perature
ratios, at the restrictions. Thus, tile maximmn
(2) Required flow capacity
allowable control pressure
(3) Allowable pressure drop
(4) Allowable rate of leakback (including zero
_/ 1.67
leakage)
ky----r (. 2 = 80.6 psia (5) Space envelope and line connection
= t oT/ (6) Simplicity of construction
Figure 7-69 presents the design of a typical
Based on this pressure, and using equation poppet-type check valve. A light, compression
(7-24), the required minimum pilot valve port area return spring normally holds the poppet in the
results as
closing position. This prevents any possibility
of fluid leaking back. When fluid pressure is
A wp RV_--T 0.0104 _,3/'3_ × 700 introduced upstream, the poppet will open against
P= C--_cZ = ff-.6_8O.-.-.6x4.11 = 0.0272 in 2 the return spring. Because of tlleir relatively
high-pressure drop, check valves of this type are
dp=0.186 in_0.2 in
used only in low-capacity services. Either elas-
tomer O-rings or metal to metal seals are used.
Required travel of the pilot valve to the fully
open position Figure 7-70 presents a swing-gate-type check
valve. The valve consists of two elements: the
inlet and the outlet body. The swing gate is
Xp o Ap 0.0272 secured to the inlet body. A torsional-type re-
- _dp- _x 0.2 =0.0433 in
turn spring holds the gate in the closing posi-
tion. Swing-gate check valves minimize pressure
drop. However, positive sealing against back-
Force balance equations of the pilot valve flow is more difficult. In some applications,
poppet and actuator:
(1) Poppet:
S_ CLOSURE PLUG
(Pc - Pa) Ap + Fs3 = Fp

__ _ UPSTREAM
(2) Actuator:
_PET
(Pt- pa)Ab- Fb- Fsl - Fs 2 = Fa z FLOW -- RETURN _ O-RiNGS

SLEEVE SPR_NG
where
Fp = pilot valve poppet seating
force, lb
Fa =pilot valve actuating force, lb
Fb = sensor bellows force, lb OPEN CLOSED
Fsl, Fs2, Fs3 =forces of the various springs

When F a > Fp, pilot valve starts to open. Figure 7-69.-Typical poppet-type check valve.
" ii.....
UPSTREAM
WING-GATE BURST FLUID P_RESSURE
HINGE
DOWNSTREAM _ UPSTREAM
V-GROOVE COINED IN
A NORSESHOE PATTERN
OUTLET BOD_I __ R_ _ R_T S_ G A
Figure 7-70.-Typical swing-gate-type check

valve.
(A} FLAT DISC BURST DIAPHRAGM
orifice holes are drilled in poppets or swing

gates to allow a controlled backflow for specific
control purpose. O_'EN CLOSE I:- ELASTOI/£R
I _ _,L_E=,L
Burst Diaphragms
When positive, hermetic sealing is essential,

in fluid-flow systems burst diaphragms are used. pROPELLANT MACHINED GROOVE
They are especially useful in storable liquid _VALVE BODY _'-- BURS T D_APHR_GM
(S) BURST _IAPHRAGM ATTACHED TO _l,
propellant engine system applications; they also PROPELLANT VALVE POPPET
serve as safety valves to prevent excessive

pressures, or to initiate flow at a predetermined Figure 7-71.-Typical burst diaphragm designs.
pressure. Burst diaphragms can be ruptured
either by the upstream fluid pressure or mechani-
cal means. General design considerations for
jointed to the duct by welding, brazing, or bolt-
burst diaphragms are-
ing. This design has been satisfactory for many
(l) Type of fluid and its corrosive character- applications.
istics Figure 7-71(b) presents a poppet-type, stor-
(2) Duration of storage (especially with cor- able propellant valve with built-in burst dia-
rosive storable liquid propellants) phragm. The latter assures a positive, hermetic
(3) Method of diaphragm installation in duct seal of zero leakage during long storage periods;
or valve body it also protects the elastomer seal of the valve
(4) Method of rupture (upstream fluid pressure seat. Rupture of the diaphragm occurs during
or other means) valve opening, either from valve actuation or
(5) Burst pressure level (if upstream fluid from upstream propellant pressure. When clos-
pressure is used) and its tolerance; ing, satisfactory valve sealing is provided by
environmental temperature effects the elastomer seal only.
(6) Retention of the diaphragm after rupture Burst diaphragms are made from a wide variety
(no metal particles must be ejected) of metals. Annealed aluminum alloys such as
(7) Allowable pressure drop across the burst 1100-0 and 6061-0 are among the most easily
diaphragm controlled. Burst diaphragms rupture in a com-
Figure 7-71(a) presents a typical flat-disk bination of shearing, bending, and tearing. The
type burst diaphragm. A V-groove has been exact burst pressure of a specific diaphragm
coined into it in a horseshoe pattern. The depth design can only be evaluated experimentally.
of the groove will determine the burst pressure. Variations of material ultimate strength have a
The uncoined section serves as a hinge during pronounced effect. Consistency of the desired
rupture and as a retainer. The diaphragm can be diaphragm burst pressure level will be greatly
enhanced by the following precautions in manu- --'A _ EXPLOSIVE CHARGE
facturing and quality control: --=B ° RAM
"C _ METAL I"O _4ETAL SE_L C

(1) Grinding of raw sheet or diaphragm to
insure uniform and accurate thickness
(2) Close control of groove depth, by preci- b,,d_t ,-'_" _,,o_L_,O_ _ Oo_", _ F
sion coining or machining, and by stress
relieving before and after coining
(3) Continuous testing of the diaphragm mate-
rial for hardness INLET
(4) Close control of the clamping pressure

with bolted diaphragms {o) BEFORE FIRli'qG (b) AFTER FIRING
(5) Close control of the welding or brazing

processes to minimize heat effects Figure 7-72.-Typical explosive-actuated pilot
valve.
(6) Proper dimensioning of diaphragm thick-
ness and groove depth to allow for cor-
rosive effects of the propellant during "D." The seal of the valve is provided by a
storage solid metal diaphragm "E," machined as an inte-
It is possible to hold burst pressure varia- gral part of the valve body. Upon an electrical
tions of a specific diaphragm design to within signal the explosive charge detonates and drives
_+2percent for diaphragms of over 1 inch diam- the ram forward to cut out the seal diaphragm as
eter and design burst pressures greater than 300 a single piece of metal and hold it firmly against
psi. For diaphragms of smaller size and lower the end plug "F." Ram "B" has a tapered head,
burst pressure, this tolerance may increase up to which rams into the guide-hole edge and causes
± 5 percent. a perfect metal-to-metal seal at point "C." This
prevents contamination of the working fluid with
gases from the explosive charge. The input cur-
Explosive-Actuated Valves
rent required for this valve is about 0.5 ampere;
In certain applications, a valve may only be actuation of the valve takes about 0.002 second.
required to operate once (to either open or close). Explosive cartridges may also be used to
In these cases, explosive-actuated valves pro- actuate valves of larger size, such as the main
vide the smallest possible size and weight. propellant valves. In the form of threaded plugs,
Since the power of the explosive actuation is they can be attached to the valve actuator ports
virtually unlimited, a rigid, hermetic zero- in lieu of pneumatic lines. For maximum relia-
leakage-type seal can be used with this type of bility, two separate cartridges are often installed.
valve. The pressures produced by the cartridges range
Figure 7-72 presents a typical explosive- from approximately 2000 to 10 000 psi. Tile valve
actuated pilot valve developed and produced by designer will have to determine the required
the Conax Corp. The actuator of this normally pressure and volume of the gases at the end of
closed valve consists of an explosive charge the actuator stroke, which then serve as the
"A," and a ram "B" including a shearing head specification to tile cartridge supplier.
Chapter VIII
Design of Propellant Tanks
Liquid propellant rocket engines and the dynamic loads. In some designs the tanks are
tanks feeding them, together with certain auxil- further stabilized by the internal pressure against
iary systems such as pressurization, form the buckling; i.e., the wails are always kept under
propulsion system. For reasons set forth in tension loads by a specified pressure level main-
section 1.5, a discussion of tank design is in- rained during storage and handling. In smaller
cluded here. The need for close coordination units, the wails are usually capable of taking
betweeri engit_e and tank designer is reempha- external loads without being pressurized in_er-
sized, particularly if they are members of dif- nally. Operational tank pressures for these range
ferent contractors. from 4OO to 2000 psia.
Prepackaged storable liquid systems are usu-
ally employed to relatively short-duration, low-
8.1 BASIC DESIGN CONFIGURATIONS OF thrust applications. Since the tanks form an
PROPELLANT TANKS integral part of the propulsion system, they are
The configurationof propelhmt tanks de- designed and furnished by the engine builder. In
pends largelyon vehiclemission and size. In one design, the thrust chamber is located inside
most modern designs, the tanks form an integral the aft tank and welded directly to the tank
bulkhead.
part of the vehicle structure. Propellant tanks
can be categorized, according to vehicle appli-
cation, as follows: Propellant Tanks fur Booster Stage Systems
(1) Prepackaged storable liquid systems Figure 8-2 presents the propellant tank design
(2) Booster stage systems configuration of a typical propulsion system as
(3) Upper stage systems used in the booster stages of a large launch
vehicle such as the Saturn V. The system shown
can be used for either storable or cryogenic pro-
Propellant Tanks for Prepackaged Storable
pellants. The tanks are arranged in tandem;
Liquid Systems
their wails form an integral part of the vehicle
Figure 8-1 presents the configuration of a structure. For booster application, overall ve-
typical prepackaged storable liquid propulsion hicle systems optimization usually dictates use
system. The tanks are arranged in tandem, with of a turbopump-fed engine system. This permits
a common bulkhead between. This system is relatively low operational tank pressures, rang-
designed for long storage periods, perhaps 5 to ing from 30 to 100 psia.
10 years. A main characteristic of these sys- Since the tanks represent a large percentage
tems is that the propellants are factory loaded of the vehicle structural (inert) weight, advan-
and are hermetically sealed in the tanks by burst tage is taken of the low-pressure levels by con-
diaphragms. Both tank and diaphragm construc- structing the tanks with extremely thin-wall
tion materials must be compatible with the pro- thicknesses. However, the often huge tank
pellants for the storage duration. In the exam- structures thus become sensitive to external
ple, the propellants are expelled by pressurized buckling loads. To stabilize the tank structures
gases produced by a solid propellant gas gener- of large booster stage systems, two basic design
ator. The tank walls form an integral part of the avenues are open: pressure-stabilization and
vehicle structure and are designed to withstand self-supporting structures. In the pressure-
the internal pressure loads as well as the vehicle stabilized systems, such as the Atlas ICBM, the
329
STRUCTURAL CONNECTION TO PAYLOAD (INTERFACE OR FWD SKIRT)
GRAIN
DIAPHRAGM
fUEL TANK
COOLING
BURST DIAPHRAGM TYPE MAIN FUEL VALVE

REMOVABLE
IGNITER
BURST DIAPHRAGM TYPE

MAIN OXIDIZER VALVE
THRUST CHAMBER
ASSEMBLLES
PRESSURE
RELIEF TYPE
REGULATOR
VENT
SOLID PROPELLANT GAS GENERATOR
Figure 8-1.-Propellant tank design configuration of a typical prepackaged storable liquid propulsion
system.
_-6.0 TYP-_ --_ _---I.25 TYP
,.',--TYPICAL SKIN-STRINGER SECTION
l ,, ,,
[ :
, i!
_ ___ _ __ '_/'_ _ iT _'_ _
IZ_TNTER-STAGE _MAIN TANK WALL LN"' _ /FUEL IOXIDI_Z_T_--THRUST S STEM
D_;TwR
UC T URET } FUEL L__ ER __2111E22D I_IC_2_ON D UCT DUCT SKIRT MOUNTS ! M'PEFE
(INS. OPTIONAL FOR CRYOGENIC PROPELLANTS)
Figure 8-2.-Propellant tank design configuration of a typical booster stage propulsion system.
DESIGN OF PROPELLANT TANKS 331
tank pressures must be constantly maintained are contained within an outer cylindrical shell,
above a specified minimum by elaborate con- through which thrust is transmitted to the pay-
trols. This tank structure basically is a thin- load. The shell is designed to withstand all
wall monocoque, requiring special handling anticipated boost and flight loads. The propel-
procedures. In most booster-stage systems, the lant tankage consists of two individual units.
propellant tanks are self-supporting types, the The main fuel tank is located forward, and the
tank walls being reinforced by skin stringers main oxidizer tank near the aft end. The two
(fig. 8-2), or by other structural means, such as welded aluminum-alloy tanks are modified
waffle grid patterns. spheres, faired into conical sections at the bot-
When cryogenic propellants are used, tank tom, for propellant discharge. The tanks are
insulation may be required. Insulation is manda- bolted to the shell structure around their support
tory in liquid hydrogen system to prevent ambient ring. The thrust chamber assembly is located
air liquefaction which causes high heat transfer just below, and closely coupled to, tile oxidizer
rates with attendant high boiloff rates, and safety tank. Thrust is transmitted to the outer shell
hazards. through the aft half of the oxidizer tank. }5oth
tanks may be insulated independently, for cryo-

genic propellant service.
Propellant Tanks for Upper Stage Systems
Many upper-stage vehicles employ a pressur-
Figure 8-3 presents the propellant tank de- ized gas feed propulsion system. Tank pres-
sign configuration of a typical pressure-fed, sures range from 100 to 400 psia. The system
upper stage propulsion system. Here, the tanks shown in figure 8-3 uses stored helium gas for
HIGH PRESSURE HELL{UM

STORAGE 8OTTLES MAiN OXIDIZER VALVE
MAiN FUEL DUCT
MAIN FUEL VALVE
TO FUEL TANK SUPPORT
P£YLOAO LN2VEN T
OXID*ZER TANK
VENT
SECTION A-A
Figure 8-3.-Propellant tank design coMiguration o[ a typical upper stage propulsion system.
pressurization. The gas is stored at an initial vessel for a given volume is a spherical shell,
pressure level from 4500 to 5500 psia at -300 ° F since it has the smallest surface to volume ratio.
in two liquid-nitrogen-jacketed, high-pressure It also has the smallest shell stress for a given
spherical bottles located between the two main internal pressure. While a sphere may be the
propellant tanks. lightest pressure vessel, the combination of
It is for upper stage propulsion systems that several spheres into the generally cylindrical
the engine designer may most likely also design envelope typical for most rocket vehicles would
the tankage. The following discussions, there- cause a sizable volume penalty. Furthermore, a
fore, will confine themselves to these systems. sphere would preclude the use of the tank wall
By contrast, the design of booster tanks will as a load-carrying member of the vehicle struc-
probably always be made by an independent ture (figs. 8-1 and 8-2), resulting in further
group. However, many of the design principles weight and volume penaIties.
presented here are equally applicable to booster Thus both vehicle configuration and tank
tanks and may thus further the understanding of pressure level will determine the shape of pro-
their design problems. pellant tanks. For vehicles of relatively large
length-to-diameter ratios and of limited space
8.2 GENERAL DESIGN CONSIDERATIONS envelopes, cylindrically shaped tanks are used.
FOR PROPELLANT TANKS For relatively high tank pressures and less
stringent space conditions, spherical tanks may
The tank design is greatly influenced by be empIoyed to best advantage (fig. 8-3). The
systems optimization within the overall vehicle ends of cylindrical tanks can have either spheri-
design. A principal design objective is for the cal or ellipsoidal shapes. The basic cylindrical
vehicle to yield the highest payload and/or ve- tank with spherical ends is lighter than one with
locity increment with maximum possible relia- ellipsoidal ends. However, the overall weight of
bility. Design details depend largely upon type an ellipsoidally ended tank may be less when
of propellants, vehicle mission requirement and the shorter interstage structure required is con-
configuration, propulsion system design, and sidered. In some designs the propellant tank aft
available construction materials and fabrication ends are faired into conical or other special
techniques. Some of the most important consid- shapes to accommodate the thrust loads from
erations follow. engine assemblies as well as to minimize trapped
propellants.
The required size or volume of a propellant
Propellant Properties
tank is the sum of usable propellant volume and
Propellants affect tank design mainly by their other volume requirements:
physical and chemical characteristics. The boil-
ing point or storage temperature of a propellant Vt:V+ T+B+U (8-1)
determines the operating temperature range of where
the tank assembly. Cryogenic propellants cause Vt = propellant tank design volume, ft3
tank design problems from thermal gradients, V = usable propellant volume calculated from
from the need for insulation and from the need propulsion system requirements, ft$ (may
for construction materials capable of remaining include a "usable residual" term repre-
ductile at very low temperatures. The very low senting design reserves, mixture ratio
density of some propellants, such as liquid hy- shift effects, etc.)
drogen, requires tanks of considerable volume.
T =trapped propellant volume. This is a func-
The highly corrosive and reactive properties of
tion of system design configuration and
other propellants severely limit the selection of
tank materials. includes propellants trapped in tank,
ducts, thrust chamber cooling jacket,
etc., ft 3
Shape and Size of Propellant Tanks
B =boiled-off propellant volume (applicable
Propellant tanks are pressure vessels. Dis- only to cryogenic propellants), ft 3
regarding other factors, the lightest pressure U =tank ullage volume, ft 3
The calculation of propellant volume is based The required usable fuel volume
on the propellant density at specific tempera-
tures. A standard temperature of 68° F is used 10.65 × 2 x 410_ 13S.2 ft3
for storable propellants. Boiling point condi- Vi= 63.17
tions at ambient pressure are used for the cryo-

genic propellants. The tank ullage-volume cal- The fuel tank ullage volume
culations should allow for propellant volume
changes due to temperature change of the tanked Ut= (Vi + Ti) × 0.02,5= 140 × 0.025 = 3.5 ft3
propellant, and for tank deformation when pres-
surized. This is especially important for pre- From equation (8-i), the required design vol-
packaged storable liquid systems to prevent ume of the fuel tank
excessive tank ullage pressures when the system
is exposed to a specified upper temperature limit Vii = VI+ TI+ U[= 13S.2 + 1.8 +3.5= 143.5 ft3
during storage. Adequate ullage volume is also
required to maintain tank pressure at starting
Propellant Tank Arrangement
when the relative ullage volume increase is large
and may tax the response of the pressurization In most vehicle systems, the propellant tanks
system. are arranged in tandem. Other arrangements are
used for specific design reasons. Figure 8-4
Sample Calculation (8-i)
The following data are given for the A-4 stage

propulsion system, including two engines:
Oxidizer (N204) density, 90.12 Ib/ft 3
,,L
-_DRYC.G./( )_
Oxidizer weight flow rates, 12.78 Ib/sec/
engine
(ol TANDEM PROPELLANT TANKS WtTH COMMON BULKHEAD
Fuel (N2H4) density, 63.17 Ib/ft s AND SEPA.R-STE HEUUM BOTTLES -

TANKAGE WEIGHT, IOO PERCENT
Fuel weight flow rate, 10.65 Ib/sec/engine
Nominal engine firing duration at full thrust,
410 sec
Trapped oxidizer volume, T o = 0.9 ft 3
Trapped fuel volume, Tt= 1.8 ft 3 .... / , _ DRY CG. /II I_
Tank ullage volume, U = 2.5% of propellant

I
volume [b) TANDEM PROPELLANT TANKS WITH INTEGRATED
HELIUM BOTTLE IN THE MIDDLE -
Determine the volume of the propellant tanks. TANKAGE WEIGHT, 93 PERCENT
Solution ooo4aO_,5 f 0,o86 o isoJ _
The required usable oxidizer volume

.Tp,
YCG\
Vo = 12.78x 2 × 410_ 116.2 ft3 (c) CONCENTRIC PROPELLANT TANKS WITH INTEGRATED
90.12 HELIUM BOTTLE AT AFT END-
TANKAGE WEIGHT, He PERCENT
The oxidizer tank ullage volume

i-o 3TB o,o2s--_ F_
.... DRY CG.-_O.tBB-, _/_'_ J
Uo=(Vo + To) x 0.025= 117.I ×0.025=2.9 ft3

I _'--0.130
.E.... E- O, OBO
to. I-.,,;
0.020-_
FUEL i
OXI'OIZ E R _ _ ._ "_'-'_ I"_
From equation (8-i), the required design vol-

(d) MULTIPLE PROPELLANT TANKS AND HELIUM BOTTLE
ume of the oxidizer tank IN CLUSTER- TANKAGE WEIGHT, 16; _ P_RCENT
Figure 8-4.-Various propellant tank arrange-

Vto=Vo+To+Uo=l16.2+0.9+2.9=120 ft3 merits of a typical vehicle syste_n.
presents various propellant tank configurations Safety Factors for Propellant Tank Designs
for a typical vehicle system using helium for
The recommended criteria for working loads
tank pressurization. A design analysis will de-
presented in chapter II (eqs. 2-8 through 2-11) are
termine the best solution for a given propellant
generally applicable to propellant tank designs.
storagevolume and vehiclespace envelope.
However, when calculating allowable working
General considerationsare:
stresses from tank internal pressure, the follow-
(I) Minimum overallweight
ing correlations are recommended for various
(2) Maximum storagevolume in a given en- situations:
velope
(1) No hazard to personnel or vital equipment:
(3) Least possibilityof propellantmixing
(4) Clean aerodynamic vehicleshape
Sw = Fy (8-2)
(5) Ease of installation of ducts and lines
or
(6) Ease of insulation
Eli
(7) Ease of fabricationand handling Sw - 1.25 (8-3)
(S) Minimum trapped(unusable)propellants
Arrangement (a)is taken as standard. Ar- (2) Special safety devices are provided for
rangement (b),combining tandem propellant personnel (example: the booster for a
tanks with an integratedhelium bottlein be- manned upper stage which has an ejec-
tween, resultsin lowest weight; however, it tion device with an exceptionally high
poses design problems in the routingof pneu- degree of reliability):
matic lines and controls.Arrangement (c)with
concentrictanks eases the installation of pro- Fy
pellantducts,but has the possibilityof simul- Sw =]-_- (8-4)
taneous punctureof both tanks (by bulletsor or
from other causes),and subsequent mixing of the F.
propellants.Arrangement (d)with multipletanks Sw = _ .35 (S-5)
has the highest weight, but is easier to fabricate
and handle especially for very large vehicles.
(3) ttazard to personnel or vital equipment:
Working Loads
f-'y
The propellant tanks are structural members Sw = 1.33 (8-6)
which must be designed to withstand a combina- or
tion of the following probable working loads: Vtt
(1) Internal pressure loads and their dynamic Sw = 1.65 (8-7)

effects
(2) Axial thrust loads and their dynamic where
effects Sw =maximum allowable working stress, psi;
(3) Bending moments due to vehicle trans- i.e., the stress due to maximum tank
verse accelerations, wind loads, and working pressure under normal transient
shifting of the center of gravity and steady operating conditions
(4)Aerodynamic forces Fy = yield strength, psi, of the tank construc-
(5)Thrust vector control forces tion material, at operating temperature
(6)Vibration loads conditions
(7)Loads produced by mounting arrangement F u = ultimate strength, psi, of the tank con-
(8)Loads caused by thermal transients and struction material under operating tem-
gradients perature conditions
(9) Loads produced during ground handling Sw is calculated for both Fy and F,. The
In most vehicle systems, internal tank pres- lower value should then be used. All propellant
sure loads and axial thrust loads are the princi- tanks are subjected to hydrostatic pressure tests
pal ones. Other loads require careful evaluation, prior to acceptance. For case (1), the proof test
including model tests or full-size experiments. pressure equals maximum tank working pressure.
For case (2) and (3), the proof test pressures or

should be II0 percent of the maximum tank work- Sw=IF._5 45000
= 1.-']-_--5-o: 33 360 psi
ing pressure. For high-pressure inert gas, long-
duration storage bottles, the following correla- The maximum allowable working sttess for
tions are recommended: mainstage = 31 S00 psi.
No hazard to personnel: (b._) Tank made of aluminum alloy 6066-T6.-
During start transient
Fy (8-8)
Sw - 1.33
Fy _ 50 000_ 37 600 psi
Sw= 1.33 1.33
Hazard to personnel: or
Fu _57000 34 560 psi

Sw-1.33 1.65 :
Fy (S-9)
Sw - 1.67
The maximum allowable working stress during

start transient = 34 560 psi.
During mainstage operation
The start transient of a prepackaged storable
liquid propulsion system for an aircrafblaunched
missile is programed not to reach main-stage s., _FY
- i = 50000 45 400
i.---/-: psi
level until the missile is at a specific distance or

from the aircraft.
Fu 57 000
Calculate the maximum allowable working Sw=-f_ - 1.35 - 42200 psi
stresses for the propellant tanks, if they are
made of-
The maximum allowable working stress during
(a) Aluminum alloy 6061-T6, Fy = 35 000 psi,
mainstage = 42 200 psi.
Fu = 45 000 psi
(b) Aluminum alloy 6066-T6, Fy = 50 000 psi, Material and Fabrication Considerations in
Fu = 57 000 psi
Propellant Tank Design
In addition to considerations of propellant

Solution
compatibility and operational temperature ranges,
(a) Tank made of aluminum alloy 6061-T6.- selection of construction materials for propellant
Since the system involves personnel safety during tanks is based on their strength-to-density ratio
start transient, equations (8-6) and (8-7) will be at a given temperature and on their ductility.
applied. For a given working pressure, the lightest tank
structure will be the one made of the material
Fy 35 000
with the highest ratio of ultimate strength to
Sw-1.33 1.33 =26300 psi
density. Most frequently used construction ma-
or terials for propellant tanks are:
Fu 45 000 (1) Aluminum alloys, such as 6061-T6,
Sw=1.65- 1.65 =27260 psi 6066-T6, and 2014-T6. Room temper-
ature properties: Average density p
Thus, the maximum allowable working stress =0.1 lb/in 3. Fy up to 60000 psi, Fu up
during start transient = 26 300 psi. to 70000 psi, Fu/p=70× 104
During mainstage operation, personnel are (2) Stainless steels, such as AISI 347 (for
considered safe and equations (8-4) and (8-5)can low-pressure tanks only), 17-7 PH and
be used: PH 15-7 Mo. Room temperature proper-
ties: Average density p =0.285 lb/in 3,
Fyupto200000psi. Fu up to 220000
Fy _ 35 000 = 31 S00 psi
Sw=l.l I.i psi. Fu/p=77.2×lO 4
(3) Fiber glass, filament wound with an

aluminum-alloy liner. Room temperature
GORE SEGMENTS_
properties (fiber glass only): Average
density p= 0.0S Ib/in 3, Fu = 120000 psi.
Fu/p = 150 x 104
CIRCUMFERENTIAL
Aluminum alloys are compatible with most
storable and cryogenic propellants, and may be
used for working temperatures up to 350 ° F.
Stainless steels are suitable for storable (lim-
ited duration) and cryogenic propellants, and are
suitable for higher temperatures (800 ° F maxi- 4...
mum). Fiber glass is limited to moderate tem-
perature conditions (-60 ° to 160 ° F).
Fabrication methods for propellant tanks
depend largely upon the type of material used. WELDS_..: .......................
Most important considerations for tank fabrica-

tion are dimension control, heat treating, and
welding. The lowest recommended wall-stock
size for propellant tanks is about 0.010 inch for
stainless steels and 0.020 for aluminum alloys.
While tank stress calculations must consider the
lower limit of wall-thickness variation, the upper
Figure 8-5.-Typical welded propellant tank
limit is used for tank-weight calculation. The construcLion.
strength of a metal may fall into a band, too, the
width of which depends on the heat-treating subject to aerodynamic heating, the filled por-
process. Stress calculation will be based on the
tion may be at a very low temperature. Addi-
minimum expected strength. The quality of the
tional thermal problems may arise in outer space,
welding process or the efficiency of a welded
from solar heating of one tank side and radiation
joint may require extra stock added to the wall cooling of the other.
thickness as calculated from other working
Also partly the tank designer's responsibility
stresses. An assumed weld efficiency of 85 to
is the solution of certain problems associated
95 percent is a reasonable value for steels (50
with the management of the propellants within
to 65 percent for aluminum). To minimize weight the tanks, such as:
penalties, build-up lands may be used at the (1) Uniform dispersion of the entering tank
welds (fig. 8-9) for an equivalent 100 percent
pressurant
weld efficiency. Figure 8-5 presents the con-
(2) Sensing of propellant quantities (PU)
struction of a welded propellant tank. Note the
(3) Prevention of propellant sloshing
segmented tank end which is typical for large- (4) Expulsion of the propellants under ad-
size tanks.
verse conditions
(5) Fill, drain, vent and pressure relief of the
Other Propellant Tank Design Problems propellant tanks
Many other design and analysis problems will

8.3 STRUCTURAL DESIGN OF PROPELLANT
have to be carefully considered before a suc-
TANKS
cessful propellant tankage can be produced. The
relatively thin, highly stressed shells make it As a rule, the wall thickness of propellant
difficult to attach concentrated loads. The loads tanks is first calculated from stresses caused by
must be spread out in a suitable way to prevent internal pressure loads and discontinuities. Then
localized overstresses. Cryogenic propellants the design is checked for other loads. If the
may create thermal transient and gradient prob- wall thickness of a pressure vessel is small com-
lems. While the empty portion of a tank may be pared to the radii of wall curvature (t/r< 1/15),
I • • III
and offers no resistance to bending, the wall where a = nominal radius of the tank, in
is subjected only to direct or hoop-membrane (2) Wall thickness, in, required to withstand
stresses, assumed to be uniformly distributed membrane stresses from internal task
over the thickness. However, if any disconti- pressure:
nuity is present along the wall, such as an abrupt
change in radius of curvature or wall thickness, pta
discontinuity and bending stresses are added. ts-2Swe w (8-11)
At a sufficient distance from the juncture be-
tween tank ends (sphere or ellipsoid) and cylin- (3) Wall surface area, in2:
drical shell (where interaction does not occur),
the maximum stress in the tank wall due to in- As =4 =a2 (8-12)
ternal pressure is calculated using the hoop-
membrane stress fornmla only. (4) Weight, lb:
Figure 8-6 identifies the major tank elements.
In an optimum tank design, the wall thickness Ws = 4 ;ra _ tsp (8-13)
varies according to a combination of local mem-
brane, bending, and discontinuity stresses. This (5) Criticalpressuredue to externalloading,
is especially true for the spherical and ellipsoi- psi. When the external pressure is
dal tank ends. The structural calculation meth- higher than the internal tank pressure,
ods for volume, wall thickness, wall surface the pressure differential across the tank
area, and weight of various tank shapes will now wall may cause the tank to buckle.
be discussed. The following general terminology
is used: 2 Ets2 ,
pcr=_V3(1-v 2) (s-14)
Pt = maximum tank working pressure, psig
Sw =maximum allowable wo,'king stress of the
tank construction material, psi Ellipsoidal and Spherical Tank Ends (fig. 8-6)
p =density of tile tank construction material,
Ib/in3 Note that the spherical end is a special case
of ellipsoidal end, in which the major half-
E =modulus of elasticity, psi
diameter, a, equals the minor half-diameter b.
v =Poisson's ratio
(1) Volume:
e w = weld efficiency
Ellipsoidal tank end volume, in3:
Spherical Tanks _._a2b

Ve - 3 (8-15)
(1) Volume, in3:
Spherical tank end volume, in3:

4 7ra3
Ys - 3 (8-10)
2 _a a (8-16)
Vs=---ff--
where
a = elliptical tank end major half-diameter, in
= radius of the cylindrical tank section
b = elliptical tank end minor half-diameter, in
(2) Wall thicknesses consideringcombined
membrane, discontinuity, and localbend-
ing stressescaused by internaltank
pressurePt. An equivalentwall thick-
Figure 8-6.-Nomenclature or principal tank ness, which is an average value of
elements. knuckle-and-crownthickness,may be
used to calculate the weight of the tank

,2° ...... _ . :;. ...... _L_ ....3
ends. K
gpta
tk : Swew (8-17)
ptR
tcr=2 Swe W (8-18)
6 !'
o /" _"_/ i /
(tk+ tcr) pta(K+_) .......I/--/- -4-.<-
te = 2 2 Sw (8-19)
p_a (g + _-)
ts-- 28w (8-20)
ioo _2_ ESo 175 200
(8-21) =
ts (g+_) (LkiPS[ RATIO
"K'NOT[ FOR SPH[RICAL T&NK$ USE K* 050 &T k it 0
FOR SPH£RICJL MEADS USE K-0 $7 &T i ,_ 0
Figure 8-7.-E11ipse ratio k versus knuckle stress

(8-22)
tc 2 /actor K, compression stress-K, and parameter
E'. (From ARS paper, "Design Criteria and
where Analyses for Thin-Walled Pressurized Vessels
k = tank end ellipse ratio = a/b; k = 1 for a and lnterstage Structures," by T. J. Hart.)
spherical end
R =tank end crown radius, in=ka; R=a for a
spherical end where
K = stress factor, a function of the ellipse
ratio k. Figure 8-7 presents a K versus e =eccentricitv. - a - k2
k curve for combined membrane, discon-
tinuity, and local bending stresses (4) Weight:
tk = wall thickness at the knuckle, in Ellipsoidal tank end weight, lb:
tcr = wall thickness at the crown, in
te : equivalent wall thickness of an ellipsoidal 2 i
_,a teE p
tank end, in We - 2k (8-25)
ts =equivalent wall thickness of a spherical
tank end, in
Spherical tank end weight, lb:
tc :wall thickness of a cylindrical tank sec-
tion, in
Ws = 2 _a _tsp (8-26)
(3) Wall surface area:
Ellipsoidal tank end surface area, in2: where
k+ 1
r(1 + e)-] E' --design factor =2 k + 1 in
\/k - 1 k- \,/-k_ - 1
.b 2 In [f_-_j (8-23)
Ae=a_÷ 2e " (see fig. 8-7)
(5) Critical pressure due to external loading,
Spherical tank end surface area, in2: psi:
For an ellipsoidal tank end, it may be
As=2 rta 2 (8-24) approximated as
Cb2Ete 2 (5) Weight, lb:

Pcre: R2 (8-27)
Wc = 2 =alctcp (8-32)
For a sphericaltank end
(6) External loading critical pressure, psi:
0.342 Ets _ For a short tank (i.e., lc<4.gav/a,-_):
Pcrs -- a2 (approximately) (8-28)
where
Etc 2 _f 1 _3 tc 2
Cb = buckling coefficient, a function of R/te, (8-33)
Pcrc=0"807 Ica _\1_V2/ a2
ranging from 0.05 to 0.10
For long tanks (i.e.,Ic_ 4.9 a V/a/---7_c):
Cylindrical Tank Section
Etc 3
(1) Volume, in3:
Pcrc-4(l_ v2)a a (8-34)
Vc = .a21c (8-29)
where E = modulus of elasticity.
where
a =radius, in
I c = length, in
(2) Wall thickness, in, required to withstand
membrane stresses due to internal tank The following design data are specified for
pressure: the A-4 stage propulsion system, which employs
a cylindrical propellant tank section with ellip-
Pta soidal ends (preliminary layout shown in fig.
tc = Swew (8-30) 3-10).
Required design volume of the oxidizer tank,
(3) Discontinuity stresses.-The discontinuity Vto = 120 ft 3
at the juncture between the cylindrical Maximum oxidizer tank working pressure,
tank section and the tank ends will Pro = 180 psia
cause bending and shear loads along the Required design volume of the fuel tank,
cylindrical circumference at the junc- Vtf= 143.5 ft 3
ture, and the adjacent areas. These Maximum fuel tank working pressure, ptf= 170
discontinuity stresses are superimposed psia
upon the membrane stresses and com- Internal radius of the cylindrical tank section,
a=41 in
prise: (a._) axial bending stress; (b__)hoop
Tank construction material, aluminum alloy
bending stress; (2) additional hoop
stress due to the shear load; (__)shear 6066-T6: Fy = 50 000 psi; F,, = 57 000 psi;
stress. Discontinuity stresses fade out p=0.1011b/in3; E=10.4×106psi; v=0.36
rapidly,so that they become negligibly Weld efficiency, ew = 100 percent.
small a shortdistance from the juncture. Determine the following:
Detailanalyses ofdiscontinuitystresses _a) Required internal tank dimensions
can be found in standardtextbooks. In (b_._)Required thickness of the tank walls at
general,buildupof wall thickness of various sections, considering internal
less than 0.5 tc, near the juncture, will pressure loads, discontinuity, and local
suffice for most designs, with only small bending stresses
weight penalty. (,_C_)Approximate weight of the tankage
(4) Wall surface area, in 2" (d__)Critical external loading pressures, using
a buckling coefficient, Cb = 0.10 for the
Ac = 2 _alc (8-31) tank ends
Solution quired wall thickness at the knuckle of the oxi-

dizer tank end
(a__) Since the oxidizer tank consists of two
ellipsoidal ends without a cylindrical section,
Kpta _ 0.80 × 180 x 41
equation (8-15) may be applied: =0.171 in
tk° = S w 34 560
2 × 2 _ra2b
Vt° - 3 From equation (8-18), the required wall thick-
ness at the crown of the oxidizer tank end
We rearrange to obtain the minor elliptical

half-diameter of the tank ends: ptR 180× 1.395×41
tcro = =-=--= -0.149 in
z3w 2×34560
3xVto 3x120x1728 29. 4

b=2x2_a2- 4 x ,_x (41)2 in From equation (8-19), the equivalent wall
thickness of the oxidizer tank end
The tank end ellipse ratio:

(tko+tero) (0.171+0149)_0.16in
a 41 re°= 2 - 2
k =-_= 2---_-_.4
= 1.395
NOTE: In some designs, weight has been saved
by taking advantage of the fact that the bulkhead com-
mon to both tanks is subject to a relativelysmall dif-
ferentialpressure in operation. Such systems, how-
Since we use an ellipsoidal end of the same
ever, requiremore elaboratepressurizationand loading
proportion at the fuel tank top, the fuel tank vol-
systems including interlocks. In case of malfunction,
ume may be treated as the volume of a cylindri- the common bulkhead may suffer seriousdamage.
cal tank section with the length lc. From equa- The required wall thickness at the knuckle of
tion (8-29), the volume of the fuel tank: the fuel tank end:
Vtf= ;ra2 Ic
tkoPtf 0.171x170
tkf .... 0.162 in
Pro 1S0
Ic Vtf 143.5x1728
= a 2- rrx(41)2 =46.9in
The required wall thickness at the crown of
the fuel tank end:
To summarize the internal dimensions of the
tankage:
tcroPtf 0.149× 170
tctf: - - 0.141 in
a=41in, b=29.4in, k=1.395, Ic=46.9in Pto 180
(b_.) We assume that certain missions of the The equivalent wall thickness of the fuel tank
A-4 vehicle require it to be man rated. From end:
equations (8-6) and (8-7), the maximum allowable

working stresses are derived:
(tkf + tcrf) _ (0.162 + 0.141) = 0.152
tel= 2 2
Fy _ 50 000 _ 37 600 psi From equation (8-30), the required wall thick-
Sw = 1.33 1.33
ness of the cylindrical tank section:
Fu _ 57 000_ 34 560 psi

Sw=1.65 1.65
ptfa_ 170x 41 _ 0.202 in
tc - Sw 34 560
We use the lower value of 34560 psi.
From figure 8-6, the tank end stress factor K Provide a buildup of 0.4 t c on the cylindrical
of the combined stresses is 0.80 for an ellipse tank section wall near the juncture to allow for
ratio k of 1.395. From equation (8-17), the re- discontinuity stresses:
tcj = tc + 0.4 t = 0.202 + 0.4 × 0.202 -- 0.283 in Cb2 Etef 2 _ 0.10x 2 × 10.4 × 106x(0.152)-*
Pcre[ = R2 (1.395x 41) 2
To summarize:
= 14.7 psi
tko=0.171 in, tcro=0.149in, teo=0.16in

tkf=0.162in, tcrf=0.141, ter=0.152in We find that Ic<4.9av'a/tc. Thus equation
(8-33) applies, yielding an external loading criti-
tc=0.202 in, tcj =0.283 in
cal pressure for the cylindrical tank section
(c_) From equation (8-25), the weight of the

oxidizer tank end: 0.807 × 10.4 x 106 × (0.20.,)-
Pcrc- 46.9× 41
_a2teoE'p
We°- 21: x {(1 - (_-.36_2)
1 3 \_--/
(0.202_ "_
=13.9 psi
From figure 8-7, E' is 4.56, for k=1.395:
Weo- rr(41) 2 x 0.16 x 4.56 x 0.101 _ 139.5 lb Axial Compressive Loading on the Cylindrical
2 x 1.395
Tank Section
The weight of the fuel tank end: In integrated propellant tank designs (figs,
8-2 and 8-3), the cylindrical tank section must
Weotef_ 139.5 x 0.152 withstand large axial compressive loads during

_/ef--- te ° 0.16 - 132.8 lb vehicle handling and operation. It' the tank is
not pressurized, i.e., if tank pressure = ambient
From equation (8-32), the v)eight of the cylin- pressure, the critical axial compressive stress
drical tank section: for an unstiffened cylindrical tank may be calcu-
lated as
14Ic= 2 rralctcp = 2 x 41 x 46.9 x 0.202 x0.101= 246.4 lb

tc 13
Add 4 percent of overall tankage weight for

local wall thickness buildups, to allow for
welded joints, discontinuity stresses, etc., and where S c = critical axial compressive stress, psi.
for tolerances during fabrication. This is the axial compressive stress
that will cause the tank to buckle
Approximate overall weight of the tankage One method of increasing the axial load-
(less accessories): carrying ability of a cylindrical tank section with
minimum weight penalty is to pressurize the tank.
Wt = 1.04 (2 x Weo + Wef + We) This is known as pressure stabilization. Inter-
hal pressure will raise the critical buckling
= 1.04 (2 x 139.5 + 132.8 + 246.4) = 685 lb
stress of a tank: or it may be used to counter-
balance an axial compressive load Fa, lb, where
(d__) From equation (8-27), the critical external
loading pressure for the oxidizer ends
F a = za=pt (8-36)
Cb2Eteo 2_0.10x 2x 10.4x 106×(0.16) 2

Pressurization will also reduce tank failures
Pcreo- R 2 (1.395 x 41)2
from very large bending loads. However, if the
= 16.3 psi pressure is ever permitted to drop below a value
necessary to carry the axial and bending loads,
The critical external loading pressure for the the tank will collapse and the vehicle will prob-
fuel tank end: ably be damaged beyond repair.
An alternative method of increasing the exter- the tank. This effect produces a su:ge of the
nal load-carrying ability of a cylindrical tank is tank internal pressure. For very short impacting
to make it self-supporting. This revolves stiff- times (less than 1.2 x 10 .3 see), the following
ening the cylindrical skin by means of longitudi- correlations are established for cylindrically
nal and circumferential members, or honeycomb shaped propellant tanks:
structures. The members may be either separate cw
(8-37)
stiffeners welded to the tank wall, or may be Ps = ,Ta2g
made integral with the wall by machining or
chemically milling a thicker sheet. C I
(s-38)
Sample Calculation (8-4) c= / 1 + [-25 EPalEtc

(1- 0"8 v)
For the A-4 stage tankage, calculate: where

(a.) Critical axial compressive load of the Ps =pressure surge due to the impact, psi
cylindrical tank section with no internal
= equivalent flow rate of the propellant due
pressure. to the impact, lb/sec
(b__)Required internal tank pressure to offset a =radius of the cylindrical tank, in
an axial compressive load of 100 000 tc = wall thickness of the cylindrical tank, in
pounds with no compressive stress on g =gravitational constant, 32.2 ft/sec 2
the cylindrical tank section. c = acoustic velocity of the restrained propel-
lant, in/sec
Solution c' = free acoustic velocity of the propellant,
in/sec
(a_) From equation (8-35), the critical com-
Ep = compressive modulus of elasticity of the
pressive stress of the cylindrical tank section: propellant, psi
E = modulus of elasticity of the tank construc-
sc: [+(-°2°2Y
°+01+
• (0202
/ j,+7
×10.4×106
tion material,
v = Poisson's
psi
ratio of the tank construction
= 20 360 psi material
In many prepackaged liquid applications, the
The critical axial compressive load of the propellant tankage is required to withstand cer-
cylindrical tank section: tain impact loads, as specified by the height of
the drop tests. The details of estimating tank
Fc=S c x2_ratc=20360×2,_× 41 ×0.202 pressure surges from a free-fall impact are illus-
trated by sample calculation (8-5).
= 1 060 000 lb
From the results it is obvious that the A-4 Sample Calculation (8-5)
stage tankage is capable of withstanding a sub- The following data are given for the cylindri-
stantial axial compressive load without internal cal fuel tank of a prepackaged storable liquid
pressurization. propulsion system:
(b__)From equation (8-36), the required internal Fuel, N2H 4
tank pressure:
Fuel density, pp=63.17 lb/ft 3
Compressive modulus of elasticity of the fuel,
Fa 100 000 Ep= 6.06x l0 s
Pt= _-_a = _x (41) 2"= 18.95 psi Free acoustic velocity of the fuel, c' = 80 100
in/sec
Tank construction material, aluminum alloy
Water Hammer Effects Due to Impact 6066-T6
When a loaded propellant tank is subject to an Modulus of elasticity of the tank construction
impact force, a water-hammer effect occurs within material, E= 10.4x 106
Poisson's ratio of the tank construction mate- Tile equivalent propellant flow rate due to
rial, v=0.36 impact:
Radius of the cylindrical tank, a = 4 in
Length of the cylindrical tank, lc = 50 in
Wall thickness of the cylindrical tank, tc _¢=pp,va2V=63.25x;rx(t-_)2x35.9 = 792.5 lb/see
=0.167 in
For a tank falling in direction of its longitu- The tank pressure surge due to impact after a
dinal axis, estimate: 20-foot free drop:
(_a) Tank pressure surge due to the impact
after a 6-foot free drop cw 43100×792.5
psi
(b__)Tank pressure surge due to the impact Ps = _a_- _,x (4) 2 x386
after a 20-foot free drop
8.4 DESIGN OF STORABLE LIQUID

Solution PROPELLANT TANKS
(a.) For a 6-foot free drop the final velocity at The design of tanks for storable liquid pro-
impact: pellants uses the s'tme general design practices
applied to other propellant tanks, except in the
V = _2gfi = _¢'2× 32.2 × 6 = 19.65 fps area of compatibility. Most storable propellants
will remain stable for long periods of time if
stored in tanks constructed of materials compati-
This yields an equivalentpropellantflow rate
ble with the propellants. To minimize propellant
due to impact of
decomposition and tank material corrosion, the
surface of tank walls in contact with the propel-
9¢= ppTra2V = 63.17x ,7× × 19.65 = 434.51b/see lants must be smooth and clean.
From equation (8-38), the acoustic velocity of Tank Material Compatibility

the restrained fuel
Prior to final selection, a detail study is re-
80 100 quired to determine the compatibility of potential
C = ,,.
tank construction materials with the propellant
combination involved. In this study, two major
considerations should be kept in mind: (1) the
=43100 in/see expected storage life of the propulsion system,
and (2) the credibility of the available data.
Impact time delay in tile tank: For a required storage period of less than 3
years, the compatibility data for many materials
lc 50 with various propellants are generally reliable.
--= --=l.16x10 -3 sec(i.e., <1.2×10 -a sec)
c 43 100 However, for longer periods, such as a 10-year
storage requirement, the design data are usually
Using equation (8-37), we obtain the tank extrapolated from the results of relatively short-
pressure surge due to impact after a 6-foot free term testing. Figure 8-8 illustrates the typical
drop: extrapolated corrosion and passivity rate of
aluminum alloy 6066-T6 exposed to tile liquid
c#,, 43100×434.5 phase of inhibited red fuming nitric acid (IRFNA).
Ps = _a-_-g= 7r×(4) 2 x386 =964 psi Much of the available compatibility test data
tend to contradict each other. This is the result
(b__)After a 20-foot free drop, the final veloc- of differing test methods, conditions, and criteria.
ity at impact is It involves the risk that although a material may
be acceptable for a relatively short exposure, it
V=_=,_2×32.2×20=35.9fps may fail when the exposure time to a particular
A full-penetration "100 percent" weld is one

where complete fusion has taken place to the
root of the joint. Unfused edges, oxide notch, or
I any concavity at the root indicate lack of full
0.32
penetration. All full-penetration welds which
cannot be visually examined on the reverse side
must be X-rayed to assure full penetration. Fig-
ure 8-9 presents the design detail of a typical
full-penetration, single-welded butt joint.
I0 If possible, interpropellant welding (i.e., the
fuel and oxidizer are separated by only one
Figure 8-8.-Typical extrapolated corrosion and welded joint) should be avoided in tank designs.
passivity rate of aluminum alloy 6066-T6 ex- A double-weld design will prevent the mixing of
propellants due to a single failure of welded
posed to the liquid phase of IRFNA.
joints. Figure 8-10 presents the design of a
typical storable propellant tank with a forged
environment is increased. Whenever a discrep-
one-piece common bulkhead. A double-weld de-
ancy of materialcompatibilityis observed, thor-
sign is used on the center-hold joint for the
ough verification of the testconditionsis in-
oxidizer duct.
dicated. For most of the long-termstorable
propellanttankage,the aluminum alloys are used
forbest results.Austenitieand semiaustenitic Tank Wall Surface Requirements
stainlesssteelsshould be used only forrather
The inner surfaces of the storable propellant
short-term storage.
tank wall must be relatively free of rust deposits
Tank Construction
/
One of the prime design objectives of storable

propellant tankage is the elimination of mainte- _° T,. ,L ----
nance during storage and prefiring activities.
This may be accomplished by an all-welded,
hermetically sealed construction. To achieve fft_P orr_L
hermetic sealing, leakproof burst diaphragms are

installed throughout the inlets and outlets of the Figure 8-9.-Design detail of a typical full-
tankage. Propellant fill ports are welded shut penetration, single-welded butt joint for pro-
after tanking. Tandem tanks are favored over pellant tanks.
concentric tanks, since propellants are more
positively separated during storage, transporta- FO_3-F_D ONE - PIECE
tion, and other adverse conditions. In some ap- COMMON BULKHEAD
plications, the common bulkhead between two N_R

-WELD DE95N
-PROPELL3_gr
APPLJED
JOINT
propellant tanks is designed to withstand a dif-

ferential pressure in either direction that may
L T_NK t_
result from inconsistent burst diaphragms. How- ,
ever, the burst diaphragm at the pressurant inlet ---_-_--_

to the oxidizer tank of the A-4 system is de-
signed to a lower burst pressure
pressurant
This insures
always the higher.
than that of the
inlet diaphragm for the fuel tank.
that the oxidizer tank pressure is
FULL PF'N ETR£rlON
WELDED JOINT
OX$1Z_l
/'2
D_T
Full-penetration welds should be used in the Figure 8-IO.-Design o[ a typical storable pro-
construction of all propellant tanks. This is pellant tank with a [orged one-piece common
especially important for storable propellants. bulkhead.
and tool marks. All welded joints should be acute with the hydrogen tank insulation. It often
ground to achieve a smooth dent-free contour. If becomes one of the most critical design factors in
an aluminum alloy is used for tank construction, a hydrogen-fueled vehicle system. The difficul-
anodizing of the inner surfaces is recommended. ties arising in hydrogen systems in connection
The tank should be cleaned and treated before with heat transfer may be dramatically illustrated
use. The process rendering the surfaces inac- as follows:
tive is known as passivation. Basically, passi- Assume two tanks of equal size, subject to
vation involves the cleaning of tank interior with the same heat influx per unit time. One is filled
solvents and, in some cases, the treatment with with liquid oxygen, the other with liquid hydro-
alkaline or acid solution. Passivation differs gen. The ratio of heat of vaporization per unit
from conditioning or stability testing of a tank. weight 0/H=0.48, but the density ratio is
Stability testing of a tank involves the applica- O/H= 14.3. Thus the volume rate of vaporization
tion of a propellant rinse prior to propellant in the hydrogen tank is 6.85 times faster than in
tanking. oxygen. In reality, the heat influxes would not
be equal for two uninsulated tanks, because of
the higher temperature differential across the
8.5 DESIGN OF CRYOGENIC LIQUID wall, and especially because of the greatly in-
PROPELLANT TANKS AND THEIR
creased heat-transfer rate from air liquefaction
INSULATION
on the hydrogen tank surface. This may further
accelerate the hydrogen volume boiloff rate to
In the design of cryogenic propellant tanks,
approximately 70 times that of oxygen. In an
there are several potential problem areas which
may affect proper functioning and reliability: actual oxygen/hydrogen system, operating at a
(1) Properties of the tank construction mate- weight mixture ratio O/H=5, the tanked mixture
rials at the cryogenic propellant service ratio by volume is H/O--3.23. Depending on the
shape of the tanks (surface ratio), this may again
temperature range
(2) Thermal stresses induced in the tank double or triple the relative boiloff rate of hydro-
gen. The absolute necessity for insulation to
structure by temperature gradients
(3) The relief of tank pressure caused by drastically reduce heat influx into a hydrogen
boiloff of the cryogenic propellants system becomes apparent.
(4) Thermal insulation of the tank walls Boiloff rates are not the only problem caused
The knowledge of the precise strength char- by the physical properties of hydrogen. Near the
acteristics, degree of brittleness, and notch sen- ambient boiling temperature, the gradient of
sitivity of the tank construction materials at vapor pressure is 2.4 psi/°R, as compared to
cryogenic temperatures (as low as -423 ° F for 0.78 psi/°R for liquid oxygen. Moreover, this
liquid hydrogen service) is a prerequisite for gradient increases rapidly with increasing tem-
their selection. In general, most of the aluminum perature, which would be experienced following
alloys, the austenitic and semiaustenitic stain- tank pressurization. Figures 8-11 and 8-12 show
less steels, possess good mechanical properties the trend for both liquid oxygen and hydrogen.
at cryogenic temperatures (also see ch. II). The The data have great significance to pump NPSH.
thermal stresses can be analyzed by determining As may be seen, just one degree of liquid-
the temperature profile at various regions of the hydrogen-temperature rise requires a 3-psi in-
tank and may be minimized by discrete design crease in tank pressure to maintain proper NPSH.
approaches. The capacity of the tank relief In a large vehicle, the required increase in tank-
valve should be based on the maximum antici- wall thickness may affect payload capability
pated propellant boiloff rate during ground hold noticeably. The situation is further aggravated
and actual operation of the vehicle systems. by the high heat influx into hydrogen, for reasons
Among the cryogenic propellants, liquid hy- mentioned in connection with boiloff. Even for
drogen imposes the most serious tank design relatively short boost periods, rapid warmup may
problems. This is mainly due to its very low create a problem more severe than boiloff, and
service temperature and its relatively large spe- places further emphasis on adequate insulation.
cific volume. Design problems are especially The following discussion of tank insulations,
to provide an insulating blanket on the tanks,

which is remotely removed just prior to liftoff.
During the boost phase, high temperatures from
aerodynamic heating, and large aerodynamic
forces are encountered. Although of short dura-
tion, this phase dictates the structural elements
of the insulation. During coast flight in space,
8 the principal source of thermal energy is radia-

tion from the Sun and the planets. By the use of
6 radiation shields surrounding the basic tank in-
w sulation, the heat flux across the tank wall can
be effectively controlled. The properties of the
S,&'TURA'r_DN TEM_'EF_TURE AND PRESSLIR£ O_ UOUI'D H_fDROGEN_

materials used in a solar shield are very impor-
tant, such as the absorptivity and emissivity of
Figure 8-I1.-Rate of change of saturation vapor
the surface when subjected to various types of
pressure to temperature [or liquid hydrogen.
radiation and body temperatures. Magnesium
oxide and silver are two materials showing prom-
therefore, is specially slanted to hydrogen ise for use in solar shields. These materials
service. may be applied as coatings onto a lighter base
material such as aluminum. The amount of
shielding required depends on the duration of the
Insulation Requirements for the Cryogenic
coast flight and on optimization of shield weight
Propellant Tanks
versus propellant boiloff weight.
Most vehicle missions include three distinct
phases during which effective insulation for the
cryogenic propellant tanks is essential: ground Basic Insulation Types
hold period, the boost phase, and coast flight in
Major desirable design features of an insula-
space. The propellant tank insulation design
tion include: light weight, uniform and repeat-
will primarily be based on flight performance
able insulation characteristics, ease of applica-
criteria. However, propellant evaporative losses
tion, low cost, low hazard, reasonable ruggedness,
during long hold periods on the ground requiring
ease of repair, good reliability, and above all,
continuous topping may also become a signifi-
low heat conductivity.
cant cost item. If long hold periods with a filled
Excellent results can be achieved with a
tank are anticipated, it may be indeed economical
laminated-type insulation. This employs an
aluminum foil and fiber-glass structure, often in
multiple layers. Tim aluminum foils act as re-
flectors, effectively rejecting radiative heat,
while the evacuated space in between prevents
conductive heat transfer. This insulation can be
applied to single-curved and to large-diameter,
double-curved surfaces. The laminar insulation
is sensitive to damage, however, possibly re-
sulting in loss of vacuum due to cracks and to
o=t-
infiltration of leak gases. Application to air-
borne system thus has been infrequent.
io tures
For
are
the latter,
finding
honeycomb-supp0rted
wide application. Figure
struc-
8-13
,='l ,s _c =z i
shows the typical example of an externally ap-
SATU_ATICI_I TEle_RATUPE AND PRESSURE OF LIOUID OXYGEN
plied hydrogen tank insulation of this type. A

Figure 8-12.-Rate of change of saturation vapor _- to _-inch nominal-size plastic honeycomb is
pressure to temperature [or liquid oxygen. installed between an inner and outer facing
v,
ON[_C_U_ CC_[ wao_ O_ Solution
@_S _ • Z 2 Le _FT 1 -- ATMOSPH£RE
\\ (a) During ground hold, [he temperature dif-

ferential across the insulation XT= 70-(-360)
= 430 _ F. The overall thickness of the insulation
.......... _k. o;,. /// t=2×0.03_0.75=0.81in. From equation (4-19),
the heat transfer rate
_¢o @Y _Euu_ 5AS
LIQUIO HYDROGg N
= 2.02 × 10-4 Btu/in_-sec

Figure 8-13.-Construction elements of a typical
liquid hydrogen tank insulation design (exter-
(b) During the boost phase, AT = 800- (-210)
nal type).
= 1010 ° P. The heat-transfer rate
sheet. The cells may or may not (weight!) be

[k'_,_ /3.8 × 10-?_
filled with an isocyanate-type foam. The foam
bubbles or the properly sealed ceils will form
individual vacuum spaces when cold (cryopump- :4.74× 10 -_ Btu/in_-sec
ing). However, because of the possibility of
vacuum degradation by infiltration of air (outer
Selection of Tank Insulation Designs
insulation), or hydrogen (inside insulation), it is
often preferred to purge the ceils with helium, Many factors will influence the selection of a
for which lateral passage ways must be provided. tank insulation design. The insulation may be
The purge also serves as a leak-detection de- located internal or external to the tank wall; it
vice, in conjunction with gas analyzers, to de- can be integral, or disposable during boost; i.e.,
tect contamination of the helium from leaks. In it can be bonded in place or mechanically re-
figure 8-13, a separate gap is purged with helium, tained. Basically, any insulation applied to a
rather than the honeycomb ceils. tank must be justified with respect to advantages
The thermal conductivity of the insulation of performance and/or economy.
shown is about 3.8× 10 -_ Btu-in/in_-sec-°F (0.2 Locating the insulation inside of the propel-
Btu-in/ft2-hr -° F). lant tank has the obvious adwmtage of protecting
Thermal conductivities of various types of the insulation from handling damage. The tank
insulation vary from 0.05× 10 -? to 6.0× 10 -7 structure is isolated from the severest low-
Btu-in/in_-sec-°F. Their densities range from temperature effects of the propellant and is thus
2.0 to 20 lb/ft a. Since the quality of an insula- subjected to only moderate thermal cycling from
tion will affect cost and weight, an optimization its source. Internal insulation also minimizes
study will have to be made, based on mission propellant loss when chilling the tank during
characteristics. filling. However, if a crack or leak should occur
in internal insulation of a hydrogen tank, gaseous
hydrogen would enter the crack and gradually in-
Determine the heat-transfer rates in Btu/in _- crease the heat transfer. Other undesirable fea-
see across the tank insulation shown in figure tures of internal insulation are apparent, such
8-13 during- as: difficulties in installation; in locating and
(a__) Ground hold.-The temperature of the in- repairing of leaks; and in cleaning the tank.
sulation surface near the tank wall is Internal insulation is also subject to higher
around -360 ° F and the outer surface is pressures and more severe temperature effects
70 ° F. which tend to impair the insulation sealing.
(b_.)
Boost phase.-The inner insulation sur- External insulation has the advantage of iso-
face temperature reaches -210 ° F and lating the tank structure from the extreme tem-
the outer insulation temperature reaches perature of aerodynamic heating during boost.
800 ° F. Installation, repair, and sealing of the external
be produced with mechanical precision and ver-

satility of construction and materials. A strength-
to-density ratio, Fu/p, can be obtained by fila-
II i ............... 14' i (-1 ment winding and impregnating the fiber-glass
// M co,_ • c,.os.,-ouT I _il; '_" __:"_R _'''L
roving with a suitable resin, which is higher than
can be achieved by any other method or material.
For instance, the strength-to-density ratio of a
fiber-glass filament-would resin-bonded structure
(average p =0.08 lb/in 3, Fu = 120 000, F,/p
= 150 x 104) is about twice that of high-strength
i_ _ECr,O_
_-. alloy steels.
SECTION A-A
Good reliability and close dimensional control
Fig.re 8-14.-Design o[ a typical insulated com- of filament-wound pressure vessels is assured
through the use of calibrated winding machines
mon bulkhead separating LH 2 and LO 2 tanks.
which orient the reinforcing fibers precisely.
However, most of the liquid propellants are not
insulation can all be done without special con- compatible with resin-impregnated, fiber-glass
sideration of access, ventilation, and curing structures. Therefore, an aluminum tank liner
techniques in closed areas. However, special which is compatible with both propellants, must
consideration will be required for tank leak de- be used to separate the fiber glass from the liq-
tection, especially with integrally insulated uid. It also provides a positive sealing barrier,
areas. If a crack does occur in the external in- since the resin-bonded fiber-glass material is
sulation, liquefaction of air and cryopumping pervious to both liquid and vapor over extended
will occur resulting in a significant rise in heat periods. Figure 8-15 presents a typicalaluminum-
transfer. lined, fiber-glass filament-wound liquid propel-
lant tank.
Design and fabrication of fiber-glass filament-
Insulation for Common Bulkheads
wound tanks with a thin aluminum liner is
When a common bulkhead is used between basically simple. However, a key problem arises
propellant tanks and cryogenics are involved, from the fact that the modulus of elasticity of
insulation is required to prevent freezing of the fiber-glass resin-bonded materials is about 5×106
propellant with the higher boiling point. Figure psi, while that of aluminum is about 10x 106 psi,
8-14 shows the equatorial area of a typical in- and that the strength of the aluminum is usually
sulated common bulkhead. A fiber-glass honey- no more than a third of that of the fiber glass.
comb is located between a forward and an aft Since the geometry of the tanks usually imposes
facing sheet. In the equatorial area, where the equal strain on aluminum liner and fiber glass as
facing sheets have to be faired into the cylindri- the tank is pressurized, the aluminum becomes
cal tank portion, they are reinforced by waffle loaded to its elastic limit long before the fiber
grid ribs. This design also requires special glass reaches the level of its high-strength capa-
insulation on the inside of the LOX tank in the bility. Consequently, with a plain liner config-
equatorial region. No foam filler was used in uration, the aluminum is stretched far beyond its
tkis particular example. elastic limit and forced back to its original shape
each time the tank is pressure cycled. Thus, the
liner may experience fatigue failure after a small
8.6 DESIGN OF FIBER-GLASS FILAMENT- number of pressure cycles. The pressure-cycle
life of a plain liner depends on amount of stretch
WOUND LIQUID PROPELLANT TANKS
beyond the elastic limit, type of aluminum, bond-
Most of the pressure vessels which possess ing between lines and fiber glass, weld joints,
surfaces of revolution can be fabricated conven- variation in thickness and contours, etc. Awell-
iently by the technique of filament winding. By designed, plain-type tank liner should have a
this method a lightweight structural member can life of about 10 to 20 pressure cycles.
tural members. In these cases, a separation tube

(Hilsch tube) may be combined with the diffuser.
Without moving parts, it operates on the principle
of separating the higher energy molecules from
those with lower energy. Temperature spreads
of 100 ° F or more can be obtained, depending on
construction and available pressure drop. Ad-
vanced configurations have been developed by
Figure 8-15.-A typical aluminum-lined, tiber- the AiResearch Manufacturing Co. In a LOX/
glass, filament-wound liquid propellant tank. hydrogen system with common bulkhead, the de-
vice may be used to direct tim cold stream toward
the bulkhead, thus lowering the temperature dif-
ferential and resultant heat transfer across it.
One design approach to extend cycle life is
to use a corrugated liner. The corrugations have
an effect equivalent to reducing the modulus of i
elasticity of the aluminum liner to a value less

than that of the fiber glass. For instance, if the
fiber glass is stressed to 100 000 psi and has a
modulus of elasticity of 5× 106, its extension
will be 2 percent. The corrugations of the liner
then should be designed to permit the 2-percent
extension so that the aluminum will not be /' ' M _"c._._-_ _," -,"_:--_f_"
stressed beyond its elastic limit.
\\"1
--_ /' :z_ 4 .'
/ )"4'-,. ./
8.7 DESIGN OF PROPELLANT TANK ", ' , d_-< _ l],
PRESSURANT DIFFUSERS
The main function of a pressurant diffuser

is to introduce the pressurant gases evenly into
the propellant tank at a desired direction and *
velocity. The gases are usually injected in a

Figure 8-16.-Typical designs o[ propellant tank
plane near the tank forward end, at a right angle
pressurant diHusers.
to the tank and vehicle axes. This will minimize
disturbances at the pressurant-propellant inter-
face. Figure 8-16 presents some typical propel-
8.8 PROPELLANT EXPULSION UNDER
lant tank pressurant diffuser designs. A radial-
ZERO GRAVITY OR OSCILLATORY
type diffuser, located at the tank axis, is used
G-LOADING CONDITIONS
for the fuel tank. This arrangement permits a
simple, lightweight diffuser design. However, Under zero or oscillatory g-loading condi-
the reversed ellipsoidal common bulkhead at the tions typical for many vehicle trajectories, fluid
forward end of the oxidizer tank requires a ring- location in a propellant tank becomes uncertain
type diffuser, consisting of many individual dif- and requires means to prevent gas pockets from
fusing nozzles located at the circumference of interfering with propellant expulsion. The pro-
the tank. pellants must be either subjected to an accelera-
The pressurant enters the tank at a tempera- tion force for settling immediately prior to usage,
ture level depending on the source, such as a or they must be continuously confined in the
LOX heat exchanger. In some cases, this may tank. The two approaches can be defined as
not be the optimum temperature for critical struc- impulse settling and positive expulsion.
Impulse Settling propellants and higher temperature pressurant

gases. Figure 8-17 presents a concentric-
This method employs a small propulsive force
circular, convoluted metal positive expulsion
directed axially, parallel to the vehicle center-
diaphragm used in a spherical tank. This sys-
line of thrust. This acceleration forces the pro-
tem is unique for bipropellant systems because
pellant to cover the tank outlet prior to initiation
it allows both propellants to be stored in the
of main engine operation. This approach is es-
same tank, as long as the volumetric mixture
pecially suitable for most space vehicles, because
their low-thrust reaction control units can be ratio does not deviate appreciably from unity.
An expulsion efficiency above 97 percent (i.e.,
used to perform the auxiliary function of provid-
the percentage of the propellant which can be
ing the thrust required for propellant settling.
expelled from the tank) may be expected. This
Although this method would eliminate the
design also minimizes shift of center of gravity
need for positive expulsion devices for the main
during the entire process of expulsion.
propellant tanks, it would necessitate separate
Figure 8-18 presents the corrugated metal
positive-expulsion propellant tank or tanks for
positive expulsion diaphragm applied to a cylin-
the sole use of the reaction control system. The
drical tank. A center post is provided to support
possible disadvantages of this impulse-settling
the diaphragm. The diaphragm is formed into a
method are (I) no control of vehicle center-of-
lobed corrugation capable of performing the re-
gravity shifts, and (2) low thrust-to-weight ratio
quired mechanical expansion. Transition sec-
under these conditions, which may increase
tions are provided at both ends of the diaphragm
response times beyond tolerable limits.
for attachment to the tank ends. An expulsion
For single-start upper stages, jettisonable
efficiency of 97 to 98 percent may be achieved
short-duration solid settling rockets may also be
with this design. However, the utilization of
applied to advantage.
_SSVRAmr GAS
INLEt
Positive Expulsion _O_C[_ R_¢ _RCU_t_ ¢VmvoLwr_
The other method of achieving proper fluid

orientation within the propellant tanks is by con-
tinuously confining the propellant to the vicinity <
; \
of the tank outlet. A positive expulsion propel-
lant tank assembly usually consists of an outer
tank shell and an inner movable expulsion de-
vice. Frequently used positive expulsion de-
vices are: (1) metallic diaphragms, (2) elastomer
diaphragms, and (3) moveable pistons. The
design of the expulsion devices depends upon a
number of considerations. Primary among these Figure 8-17.-Concentric-circular convoluted
is the design configuration of the propellant metal positive expulsion diaphragms used in a
tank. Another is the allowable shift of vehicle spherical-shape propellant tank.
center of gravity during expulsion, which in many
vehicle applications must be minimized to avoid
instabilities induced by a pitch or yaw maneuver.
A choice of construction materials capable of
sustained contact with the pressurants and pro-
pellants must also be made.
Metal Expulsion Diaphragms
Metal expulsion diaphragms made of 1100-0 Figure 8-18.-Corrugated metal positive expulsion
aluminum (0.010 to 0.020 inch thick) are desir- diaphragm used in a cylindrically shaped pro-
able for long-term storage contact with storable pellant tank.
tank volume is less efficient due to dimensional disadvantage of pure elastomers in storage con-
limitations (i.e., size of the center post com- tact with many storable propellants is tensile-
pared to tank diameter). This design, too, mini- strength degradation as a function of time. They
mizes shift of center of gravity. are also incompatible with hot pressurant gases
One disadvantage of the metal diaphragms is produced by gas generators.
that only one complete expulsion cycle is possi-
ble, because the metal probably will be cold
Positive Expulsion By Movable Pistons
worked. The cold working not only changes
metal elastic properties and the Ap required for An alternate method of obtaining positive ex-
expulsion, but may cause fatigue failure. pulsion in cylindrical tanks is a movable piston
actuated by pressurant gas. Positive attachment
is required during storage to keep the piston in
Elastomer Expulsion Diaphragms
the proper position. To prevent leakage during
Elastomer-type diaphragms are applicable to operation, seals will be required. The seals
most tank configurations and often offer a more may be piston-type rings or some type of metal-
efficient utilization of tank volume. One design lic wiper seal. In either case, the dimension and
approach takes advantage of the stretching prop- surface finish of the tank inside diameter should
erties of pure elastomer and uses the diaphragm be maintained relatively accurate and smooth. In
as a bladder. Figure 8-19 presents the bladder- some designs a concentric center post is utilized
type elastomer positive expulsion diaphragm to guide the piston, requiring an additional seal.
used in a spherical tank. As the bladder is Figure 8-20 presents the design of a movable
pressurized and inflates, the propellant is dis- piston guided by a center post.
placed and positively expelled and uniformly The pressure differential across the movable
confined during all phases of expulsion. By piston required to overcome friction during oper-
positioning the bladder in the geometric center ation increases the required pressurant pressure
of the tank, center-of-gravity shift during expul- and the tank structural pressure loads for a given
sion is virtually eliminated. propellant pressure at the tank outlet. It may be
The initial unpressurized volume within the estimated by
bladder can be utilized as the ullage volume.
The elastomer diaphragm will produce an expul- ApAp = [sFnLs (8-39)
sion efficiency of up to 99 percent in many tank
configurations. In addition, it is capable of where
many expulsion and refill cycles. An inherent Ap=pressure differential across the piston, psi
TAN_ _D
pR_SS_
I_L_r
ANTI - EXT_u S4ON _TE
Cai.ON _t_eAo_ ¢_e_rm_
I, _ p_pe_LA_r
_r_eT
_'-LAOO_[R _ £LASTOMgR POSITIVE

_*_G SelL kl CY_Na_C_L 11_
Figure 8-19.-Bladder-type elastomer positive ex-

pulsion diaphragms used in a spherically Figure 8-20.-Movable piston used in a cylindrical
shaped propellant tank. propellant tank [or positive expulsion.
r
Ap-- cross-sectional area of the piston, in: Estimate the pressure differential across the
is = coefficient of friction of the piston seals movable piston.
Fn =unit normal force reacted by the piston
seals on tank wall and guide post, lb/in
Solution
of seal
Ls = total length of all piston seals, in The cross-sectional area of the piston
Ap - 7r× (20) _ _ _× (1.5) 2 _ 314.16 - 1.76 = 312.4 in s

4 4
The total length of piston seals:
The following data are given for a cylindrical
positive expulsion tank with a movable piston as
Ls= _× (20+ 1.5)=67.54 in
shown in figure 8-20:
Inside diameter of the tank, 20 in
Substitute this into equation (8-39); the pres-
Diameter of the guide post, 1.5 in
sure differential across the piston
Coefficient of friction of the piston seals,
is =0.1
Unit normal force reacted by the seals, Fn _p- fsFnLs - 0.1>:500×67.54 10.8psi
= 500 lb/in Ap 312.4
Chapter IX
Design of Interconnecting Con ponents
and Mounts
9.1 THE PRINCIPAL INTERCONNECTING increases the duct envelope and may cause in-
COMPONENTS AND MOUNTS terferencewith other vehicle systems. Figure
9-2 shows a typical flexible propellant supply
In section1.4,we brieflyintroducedthe
duct which has restraining linkages for stabiliz-
principalinterconnectingcomponents and struc-
ing the bellows. At the vehicle end, these ducts
tures. Figure 9-I shows these fora typical
will connect to longer or shorter vehicle ducts,
LH2/LO 2 turbopump-fedengine system. A dis-
the length of which depends on whether the for-
cussion of the most frequentlyused components
ward or the rear tank is being connected. It is
follows. Design detailis presented in subse-
important that the engine designer not only m-
quent subehapters.
form the vehicle builder of connecting flange
dimensions and types of gasket being used but
also of the forces transmitted by the engine duct
Propellant Supply Ducts
to the vehicle during gimbaling as well. In most
Among the principal interconnections between pump-feed systems, the working pressure of pro-
engine and vehicle are the propellant supply pellant supply ducts usually does not exceed 50
ducts. Since every psi of propellant tank pres- psig. In upper stages, during lower stage boost,
sure above the minimum required for proper en- however, pressures may temporarily be substan-
gine performance results in additional weight of tially higher as a result of a combination of high
the tank walls and of the gas pressurants, it is accelerations and full tanks (100 psig and over).
desirable to keep the pressure losses between Figure 9-3 shows the propellant ducting and
tank outlets and engine inlets to an absolute mounting for a cluster of four storable propellant
minimum. pressure-feed engines. The thrust mounts are of
The task of designing these ducts, or at least a box-type, beam construction. The propelhmt
the flexible portion thereof immediately upstream duets, which consist of restrained bellows and
of the engine, frequently falls to the engine de- rigid sections including flowmeters, connect the
signer. The designer must find an optimum bal- thrust chambers to the main propellant valves.
ance between low pressure drop, by making the The main valves, in turn, connect directly to the
duct diameter as large as possible, and flexibil- propellant tank ducts. Care has been taken in
ity and structural integrity, which in general the design to keep the flow path and ducting
becomes more difficult with increasing diam- volume constant between main valves and indi-
eters. The designer must further consider the vidual chambers. This arrangement also assures
fact that the ducts, because of their location off uniform pressure drops to all thrust chambers.
the engine gimbal center, are subjected to tor- Furthermore, all interconnecting components on
sional loads, in addition to bending. Furthermore, the fuel side as well as on the oxidizer side are
the ducts are subjected to internal pressure, fre- designed to be interchangeable. The valves
quently in a stringent cryogenic and vibration used in this design include burst diaphragms.
environment. Because of the many forces acting Thus, exposure to the propellants of all ducting
upon the ducts, restrainers against buckling are downstream of the valves will occur only during
frequently required. These may be located in- engine firing. In most pressure-feed systems,
side the ducts, thus adding to undesired pressure the working pressure of the propellant ducts is
drop; or they may be applied externally, which less than 500 psia.
353
ENGINE GIMBAL MOUNT ASSEMBLY
TURBOPUMP
ASSEMBLY
G.G. OXIDIZER
LI NE
NE GIMBAL
ACTUATING LUGS
OXIDIZER
TURBOPUMP ASSEMBLY
FUEL
TURBOPUMP
OXIDIZER MOUNT
TURBOPUMP MOUNT
PUMP DISCHARGE
HIGH PRESSURE
MAIN OXIDIZER
DUCT
TURBINE
EXHAUST
HOT GAS DUCT
PUMP DISCHARGE
TURBINE
HIGH PRESSURE
DRIVE HIGH MAIN FUEL DUCT
PRESSURE HOT
GAS DUCT
CHAMBER
ASSEMBLY
OXIDIZER
TURBINE
EXHAUST HOT
GAS DUCT
Figure 9-1 .-Various interconnecting components and mounts in a typical LH2/LO 2 pump {eed engine
system.
Pump-Discharge, High-Pressure Propellant Ducts rating pressure loads acting on the two compo-
nents connected by a flexible member must be
In turbopump-fed engines the pump-discharge absorbed by restraining links attached to the
high-pressure propellant ducts are designed to bellows, or by other compensating means.
connect the oxidizer and fuel pump discharges to Figure 9-4 presents a typical pump-discharge,
the main oxidizer and main fuel valves attached high-pressure propellant duct with external re-
to the thrust chamber. The ducts contain bel- straining links. It is used for the main oxidizer
]ows sections which permit the degree of move- (LO2) duct of the engine system shown in figure
ment required between thrust chamber and turbo- 9-1. Another typical pump-discharge, high-
pump to accommodate tolerance buildups, pressure propellant duct, used as the main fuel
misalinements, and motion due to temperature (LH2) duct for the same engine system, is shown
change and acceleration loads. However, in in figure 9-5. This duct has a unique end-load
some engine designs rigid, in place welded ducts compensator which incorporates two bellows tied
have been successfully applied. The working together by a restraining rod to limit bellows
levels have been proposed. Therefore, the sepa- movements• One bellows opposes the other in
DESIGN OF INTERCONNECTING COMPONENTS AND MOUNTS 355
BRU- J_TS J
Figure 9-4.-Typical pump-discharge, high-

pressure propellant duct with restraining links.
CON_IECTbNG TO
PUMP DISC_ARC-[--_r
Figure 9-2.-Typical (lexible propellant supply

duct for a large turbopump-[eed engine system.
(Note: Duct is mounted in test fixture with
fluid pressure connections.)
8ELA._fS
RI[ST RAtNING RO0
E_O LOAO
COMPENSAT_ --
BEU.
, 1 J
O_JTCET FL._,_IG_E
CONN£C_r EO TQ
Figure 9-5.-Typical pump-discharge, high-

pressure propellant duct with end-load
compensator.
balancing the duct, thus reducing duct-separating

loads by 90 percent.
The high-pressure propellan_ lines between
pump discharge and gas generator usm_lly can be
made of wire-braided-type flexible hoses or tub-
ing, because of their relatively small size (often
less than 1 inch).
Propellant-Tank-Pressurization Lines
Figure 9-3.-Propellant ducting and mounting These lines are employed to connect the main
arrangements for a cluster o[ [our storable- propellant tanks to the pressurant sources such
propellant pressure-Iced engines. as stored pressurant gas systems (including
storage bottles, heat exchangers, pressure regu- Cryogenic Propellant Bleed Lines
lators), gas generators, cryogenic propellant heat
Cryogenic prope!!ant engine systems with
exchangers, etc. High-pressure hoses and tub-
ings are used. turbopump feed may experience difficultiesdur-
ing start if the metal parts containing cryogenic
fluid are at ambient or insufficiently low temper-
Seal Drain Lines
atures, and if the fluids in the volumes below
It is difficult to achieve perfect dynamic seal- the tank outlet are superheated and form gas
ing at the shafts or rods of turbopumps and other pockets. Since the pressure upon opening of the
components. Therefore, seal drains are often main valves and start of the turbopumps will
provided between two dynamic seals placed in further reduce the static pressure at the pump
series. The required seal drain lines including inlet, accelerated gas production will result.
flexible hoses and tubings are routed away and This, in turn, may lead to pump cavitation and
overboard. For LO2/RP-I systems, this can be starving of the gas generator. To prevent this, a
done by routing the drain lines along the thrust continuous bleed from a point farthest down-
chamber wall to the chamber exit. For propellant stream from the pump inlet is applied until
shortly before engine start. In this manner, fresh
combinations which can form highly explosive
mixtures, routing to sufficiently spaced vent liquid at tank bulk temperature will continuously
replace the warming fluid, and cool the contain-
ports at the vehicle periphery is required. Fig-
ing metal parts. To avoid hazardous conditions
ure 9-6 presents a typical pump seal drain line
schematic of an upper stage system. The seal at the launch site, the bleeds, particularly if
drain lines are routed to the vehicle periphery they can form combustible or explosive mixtures,
during boost flight and to chamber exits during are ducted away. To make this possible, a line
stage operation. connection from the engine to the vehicle and

beyond is required. The lines are generally
made of wire-braided flexible hoses and tubings.
Pneumatic Supply Lines
A special ease is presented by cryogenic
Liquid rocket engines usually are equipped engines in upper stages which will not start
until some time after the bleeds have been closed
with one or more pressure vessels to supply
pneumatic pressure for valve actuation, for tur- at liftoff,or which have to start after prolonged
bine start, for sequenced purges, and possibly cruising periods. Here, a recirculation system
which returns the fluids to the tanks rather than
for other purposes. The vessels must be charged
dumping them overboard is preferred. Lines
prior to test run or flight,requiring high-pressure
flexible line connections to the vehicle, and across the interface between engine and vehicle
are required. Bleed and recirculation flows can
disconnects at the vehicle periphery. Their
design not only must consider the mating coun- be minimized if the vehicle builder provides
terpart on the vehicle side but also the type of means, such as subcooling, insulation, avoid-
ance of temperature stratification, maximum eco-
fluid and its temperature and pressure.
nomic tank pressure to keep the temperature of
the bulk sufficiently below the boiling tempera-
\ ture at operating tank pressures. In some cases,
Z
.... : J_ prechilling of metal parts exposed to cryogenic
\,
fluids may be advantageously accomplished by
\ cooling media supplied from ground through a
vehicle disconnect to the engine. Routing of all
lines mentioned will be similar to drain lines
,,(Lr _'_':{ -_ x (fig. 9-6).
Purge Lines
During the start and shutdown sequences, and

Figure 9-6.-Typical pump seal drain schematic. for prerun and postrun servicing, inert gas purges
DESIGN OF INTERCONNECTING COMPONENTSAND MOUNTS 357
are frequently required to keep a system dry, to Turbine-Exhaust, Hot-Gas Ducts

prevent combustible mixtures from forming, or to
The hot gases from the turbine exhaust may
expel residual propellants. If they are contin-
be ducted to near the main chamber nozzle exit
uous during ground hold, it is advantageous to or into a thrust chamber exhaust manifold. Most
feed them from ground to avoid increased flight
turbine-exhaust duets are of welded stainless
weight. This requires connecting lines from the
steel or nickel-alloy, sheet-metal construction.
englne to the vehicle. It is strongly recom-
Bellows sections permit the degree of movement
mended that entirely separate purge systems for
required by the system for thermal expansion,
each propellant be employed to avoid backup of
misalinements, and tolerances. In most cryo-
one fluid into the system of the other. Check
genic propellant systems, a heat exchanger is
valves are not considered as reliable as strict
incorporated in the exhaust duct assembly to
separation. For engine design we are concerned
vaporize one of the propellants, usually LOX,
here only with those purges which connect di-
for tank pressurization (fig. 5-7).
rectly to an inlet port of an engine part, such as
the thrust chamber cooling jacket or injector
Thrust Mounts
manifold. From the viewpoint of the environment
in which the engine must operate, the engine The thrust mount transmits the thrust loads
designer should additionally specify the purges between engine and vehicle thrust frame. In
to be provided by the vehicle builder to condi- some designs (fig. 2-4), the thrust mount also
tion the engine compartment as needed. All serves as the structural mount for many other
purge lines can be made of wire-braided flexible major components of the engine system. Thrust
hoses and tubings and may be routed as are the mounts usually consist of truss members of tubu-
drain lines.
lar and box shape, made of aluminum alloy or
alloy steels. In engine systems using fixed
thrust mounts, thrust vector control is accom-
Hydraulic Lines plished by swiveIing the thrust chamber, or by
other means such as jet vanes, secondary injec-
Various high-pressure hydraulic lines are
tion, or jetavators.
used in a liquid rocket engine for hydraulic actu-
ating, lubrication, etc. All lines must permit
flexure and must be of the proper size and pres- Gimbal Mount Assemblies
sure rating. Most of the hydraulic lines are made
of wire-braided flexible hoses and tubings. In most large turbopump-fed engine systems
(figs. 3-2, 3-4, and 9-i), the compIete engine
assembly is gimbaied at a universal-joint-type
gimbal mount assembly which connects the thrust
Turbine-Drive, High-Pressure, Hot-Gas Ducts chamber to the vehicle thrust structure.
In most turbopump-feed engines, the gas gen-
erator is connected directly to the turbine inlet;
Turbopump and Other Component Mounts
thus, there is no need for ducting. In some de-
signs, however, high-pressure hot-gas ducts are Turbopump mounts (fig. 9-1) are used to se-
required to connect between gas generator and cure turbopumps to thrust chamber or other struc-
turbines (in systems with two individual turbo- tural members, such as the thrust mount. They
pumps), or between main thrust chamber bleedoff often consist of welded tubular trusses, forged
ports and turbines (tapoff gas-turbine-drive sys- structural members, and flexible end joints. In
tems). These ducts usually consist of rigid and some designs, rigid-type, pump-discharge propel-
flexible sections made of stainless steels and lant ducts are used as members of the turbopump
nickel-based alloys, for service temperatures up mounts. Most propellant valves are fastened
to approximately 1700 ° F. The ducts must be directly to the thrust chamber or gas generator.
capable of absorbing considerable deflections Other control components and small storage ves-
from thermal expansion, in addition to deflec- sels may be secured by aluminum-alloy sheet-
tions due to misalinements and dynamic loads. metal brackets.
9.2 DESIGN OF TUBINGS AND TUBE decreases tubing outside diameter less than 6
FITTINGS percent. Recommended minimum bend radii for
stainless-steel and aluminum-alloy tubing are
To interconnect fluid lines in rocket en-
presented in table 9-3.
gines, high-quality, cold-finished seamless
Tubing assemblies must be properly supported
tubing of either aluminum alloys or austenitic
to prevent stresses and consequent weakening of
stainless steels are frequently used. In rocket
the system under vibrating conditions. In addi-
engine design practice, the term "tube" refers to
tion, proper support minimizes the danger of
lines up to 2 inches. Stated nominal tubing
recoil and live whip in the event of tube failure.
sizes refer to their outside diameters. For sizes
Where tube fittings are employed, support spac-
of more than 3/4 inch, flanged joints should be
ing should be reduced by 20 percent to account
used.
for the added weight. Supports should be placed
Two principal classes of tubing are distin-
as close to each side of fittings, valves, and
guished: common pressure tubing and mechan-
other components as practical. Overhang should
ical tubing. The latter is designed and manufac-
be minimized by placing supports as close to
tured with closer tolerances then the common
bends as conditions allow. Table 9-4 presents
tubing. It is more expensive, but may save much
the recommended maximum bracket spacing for
time when machining and chucking in uutomatic
common-size tubing assemblies in engine sys-
production machines is involved. Unless speci-
tems. Where tubes of different diameters are
fied otherwise, pressure tubing is used in rocket
connected, average spacing may be used.
engines for all purposes.
General design practicerequiresno detail
drawings fortubingassemblies less than 3/8-
Tubing Design Working Pressures inch size. They are to be shop fittedaccording
to an engine mockup duringthe assembly of the
Pressure tubes are supplied fully annealed
engine system.
and pressure tested. They fall into two major
classifications: average wall and minimum wall.
Average-wall pressure tubes are made with a Tube-Fitting Designs
wall thickness which may vary from the nominal
size by plus or minus 10 percent. Minimum-wall The three-piece AN flared tube fitting per
tubes are made with a wall thickness that may AND10056 or MS33656 (fig. 9-7) is widely used
not be less than that specified, but may be heav- as a standard in rocket engine designs.
ier by 20 to 40 percent, depending on the type of The angle of flare of the fitting was estab-
tubing. For lines carrying pressurized fluids, lished at 37 ° to give maximum flare contact and
minimum-wall tubes should be used. provide a nose sufficiently strong to resist crush-
A factor of safety (i.e., the ratio of the ulti- ing. The AN three-piece fitting consists of a
mate strength of the tubing material to the maxi- coupling nut (AN818), a sleeve (ANS19), and a
mum allowable working stress) of 4 may be used connector. The sealing occurs between the nose
for general rocket engine applications. Higher of the fitting connector and the inside of the
tube flare. The AN819 sleeve has an external
values (6 to 8) should be used in applications
involving hazard and excessive vibration. For shoulder against which the AN818 nut acts.
high-pressure application, aluminum tubing This produces a locking effect on the nut, as the
should be avoided where possible. Allowable sleeve is slightly sprung upon proper torque.
working pressures for a factor of safety of 4, for The sleeve is free to turn during initial assembly
various stainless-steel and aluminum tubings, to allow for slight eccentricity of the tubing or
are presented in tables 9-1 and 9-2. the nose of the fitting connector.
Figure 9-8 illustrates a typical flareless tube
fitting per MS33514, the Ermeto, manufactured by
Tubing Installations in Engine Systems
the Weatherhead Co. It is designed for use with
Installations requiring bends must be accom- flareless, heavy-walled tubing in high-pressure
plished with minimum distortion and constriction applications. The assembly of this fitting in-
of the tubing. A satisfactory bend is one which volves a preassembly to check for contact of the
TABLE 9-1.-Corrosion-Resistant Steel (18-8) Annealed (MIL-T-8504 ASG) Tubing

[Allowable working pressures in psi at 100 ° F; safety factor of 4]
Maximum working pressure, 3000 psi
--1/2 5/8 3/4 1
Wall thickness, in ......................... 0.022

_0.035
• _°JXX_LL_:__I
0 042
• I 0.109
0 058
0 083 0 065
_e oD.in.............................. 1/4 _ .2-1---_-1---_"_---I

1-1/4
I 1-1J2
I
Wa.<.io_ne_s,
in......................... 10.020
1°.028
10.035
100_9|0.0_8
10.0_2
10.095
10.10._
1
Tube OD, in .............................. --

1,/4
0 020
318
0 028
1/2
0 032 0 035 0 0-t2
1
0 049
-1,t4
0.058
1-1f2
0.065
2
0.095
TABLE 9-2.-Aluminum Alloy, 5052 Round Seamless Drawn WW-T-7Sa Temper H34
[Allowable working pressures in psi at 100 ° F; safety factor of 4]
Tube OD, in ..............................
Wall ,hickness. in ......................... i 0.025

1/4
[ 0.042
3/8
I
1/2
0.049 i
-[--1
518
0.05S [
3/4
0.072
1
J 0.095
1-1/4
I 0.120 I
]
_beOO,.
.............................. I lJ41 3_81lJ2I 5_8 3'4 _ 111_11'21_
w_ll<.,ek.ess,
in ........................
I _ 1_0 020 0 02S
I 0 032
°k22_L___
0 035 0.0-t2 0.049
I 0 065 10.072 l °°95
SLEEVE SUPPORTS COUPLING

TUBE AND PREVENTS OR B-NUT
WIPING OF NUT
ON FLARE
SPECIAL
OUTER EDGE COUPLING
sE
OF FLARE
MUST PASS NUT
FLARE PER INNER EDGE
AND t0061 C OF SLEEVE
MS 33584 BUT NOT
EXCEEO
OUTER EDGE
*'AN THREAD
SEALING SURFACE
STRAIGHT THREAD
MALE CONNECTOR MALE _
CONNECTOR
Figure 9-7.-3-piece AN flared-tube fitting per Figure 9-S.-Flareless tube fitting (Ermeto) per
AND 10056 or MS3365G. MS33514.
TABLE 9-3.-Minimum Bend Radii [or Stainless-

Steel and Aluminum-Alloy Tubing
[All measurements in inches]
Inside bend Radii to center

Tube OD Wallthickness Figure 9-9.-Typical installations o[ $ube fitting
radii of tube
ends into AND10050 boss.
3/16 Any 5/8 23/32
1/4 Any 3/4 7/8
5/16 Any 3/4 29/32 The tube fittings should be of the same mate-
3/8 Through 0.022 1-1/2 1-11/16 rial as the tubing. Fittings are designed to be
Over 0.022 1 1-3/16 as strong as the strongest tubing of like material
1/2 Through 0.028 1-3/4 2 used with the fitting. Design details and dimen-
Over 0.028 1-1/2 1-3/4
sions of tube fittings can be found from AND and
5/8 Through 0.028 2-1/2 2-13/16
Over 0.028 I-3/4 2-1/16 MS specifications. In some cases, AN and MS
parts are interchangeable.
3/4 Through 0.028 3 3-3/8
The boss design for engine components, to
Over 0.028 2-1/2 2-7/8
7/8 Through 0.035 3-1/4 3-11/16 connect AN tube fittings, is specified by
Over 0.035 2-3/4 3-3/16 AND10050. Figure 9-9 illustrates typical in-
1 Through 0.035 3-1/2 4 stallation of tube fitting ends into the AND10050
Over 0.035 3 3-1/2 boss.
1-1/8 Through 0.035 4 4-9/16
Over 0.035 3-I/4 3-13/16
1-1/4 Through 0.035 4-1/2 5-1/8 9.3 DESIGN OF FLANGE JOINTS
Over 0.035 3-1/2 4-1/8
1-1/2 Through 0.035 6 6-3/4 The interconnecting lines of a rocket engine
Over 0.035 4 4-3/4
system must retain fluids at various pressures
1-3/4 Through 0.035 7 7-7/8 and temperatures, Each of these lines may have
Over 0.035 5 5-7/8 one or more mechanical joints requiring static
2 Through 0.035 7 8
seals to prevent leakage. The joint as an as-
Over 0.035 6 7
10-1/4 sembly must be capable of maintaining intimate
2-1/2 Through 0.049 9
Over 0.049 7 8-1/4 conformance of seal and sealing surfaces through-
out its operating life, regardless of all strains,
3 Through 0.049 11 12-1/2
Over 0.049 9 10-1/2 loads, and thermal gradients. For line sizes of
4 Through 0.065 12 14 3/4 inch or larger, bolted flange joints are gen-
Over 0.065 I0 12 erally used in rocket engine design. The design
of these flange joints is also applicable to other
components, such as thrust chamber injector
TABLE 9-4.-Recommended Support Bracket Spac- dome flanges and turbopump and valve housing
ing [or Tubing Assemblies in Engine Systems flanges.
Maximum support spacing in inches

Flange Joint Design Considerations
Tube size(OD), in
Aluminum ahoy Stainlesssteel
Figure 9-10 illustrates typical flange joints
12 14 used in a propellant duct assembly. Most flange

1/4 through 1/2 ....
5/8 through 3/4 .... 17 20 joints consist of three elements: the flanges,
1 and over ......... 21 24 the fasteners (bolts and nuts), and the gasket.
In addition to an effective gasket, a leaktight
flange joint must have proper flanges and bolt-
special hardened sleeve, which cuts into the ing. In flange joint structural design, the goal
tube wall. Flareless tube fittings may be used of optimum weight within safe stresses must
optionally in engine designs. consider the elastic behavior of the part.
a material so that when properly loaded, a barrier

'A/ASHER BOLT
will be formed in depressed surface regions, pre-

UI _BOLr
venting leakage of the contained fluid. Gaskets
which damage the flange sealing surfaces are
usually unsatisfactory, if the joint has to be
repeatedly disassembled and reassembled. For
these applications, gasket surfaces should
always be softer than those of the flanges.
4. Sealing surface condition.-As a rule, leak-
age will occur when the internal fluid pressure
begins tO exceed the compressive stress which
\r_ \ _-[. x.x_) _ S ,_LLO,.._ $ A$$,MB LY /
holds the gasket in contact with the sealing

surface. Radial scratches on the flange sealing
surface tend to induce a lower gasket local com-
pressive stress along that radial path, resulting
in leakage. Conversely, if the flange sealing
surface consists of concentric serrations, higher
_WASHER
local compressive stress will be induced circum-
Figure 9-I O.-Typical flange joints of a propellant ferentially along the serrations, and will seal
duct assembly. higher internal pressures for a given nominal

compressive stress of the gasket. The required
sealing-surface finish is a function of gasket
design. Generally, it ranges from a mieroinch
In addition to structural needs, other design
finish of 32 to 125 rms, with concentric tool
considerations for flange joints are:
1. Working temperature and pressure of the markings.
fluid.-These greatly affect the type of seal de- 5. Gasket toading.-The minimum requirement
for good sealing of a flange joint is sufficient
sign and material. The problems of maintaining
a leaktight flange joint are greatly compounded gasket precompression to close tip all paths
through which fluid flow could occur. Further-
by temperature effects. Temperature differentials
more, the sealing load must be maintained so
at the flange joints of rocket engines are apt to
that a specific level of gasket compressive
be large, because of steep heating and flow
stress is induced to resist the internal fluid
transients. They introduce thermal stresses and
pressure.
strains which may disturb the sealing. When
designing an elevated temperature (or subzero) 6. Sea/drain.-In sonie applications, positive
flange joint, tim temperatures of the various sealing at the flange joint is required. Dual
(series) seals and overboard drain line are then
parts should be analyzed to assess their effect
on the sealing load. Ideally, thermal effects provided (fig. 9-15).
should neither loosen nor tighten the joint. In

actual design the flange joint must be sufficiently
Flange Joint Structural Design
elastic, so that sealing loads are at least main-
tained within reasonable limits. Figure 9-11 presents the structural design
2. Degree of sealing.-For most applications configuration of a typical flange joint as fie-
it is sufficient to prevent mass flow as evidenced quently used in rocket engines. The flange ring,
by liquid or gas tightness under soap bubble or under various working loads, is subject to bend-
immersion tests. When sealing the lighter gases ing as shown exaggeratedly by the dotted lines.
(H 2 and He), diffusive flow through openings of The bending moments may become quite large;
molecular size may become significant. the resulting stresses reach their maximum at
3. Gaskets.-The function of a gasket is to corner Z, where the flange joins the wall of the
seal effectively between two flange surfaces. It duct.
is usually made of a material which will readily The basic approach to the design of a flange
conform to the surfaces, or which is coated with joint is to prestress the flange bolts in tension
_II I II I I _
D s =outside diameter of the gasket, in

1
p =internal fluid pressure, psi (the mammum

J %
working pressure should be used)
F, =force per unit length of the flange rag from
bolt loading, lb/in
F 2 = force per unit length of the flange rag from
, ;k_ _
longitudinal tension of the duct, Ib/in
F 3 = force per unit length of the flange ring from
/ _ !' internal pressure p, lb/in
F 4 = force per unit length of the flange rag from
....... )2 A_Kg'r
!
F s = force
gasket
per
loading,
unit length
lb/in
of the flange rag from
compression load at the flange outside
diameter, lb/in
Figure 9-11.-Structural design configuration o[ a
Sg = required average gasket compressive
typical flange joint.
stress, for proper seating against an in-
ternal fluid pressure p, psi
so that a gasket compressive stress is main- m =gasket factor, a function of gasket design
tained to seal effectively against fluid leakage, and to be determined experimentally;
under maximum working pressure and other loads. design values range from 0.8 to 10
Defining a flange ring segment of unit length n = flange factor, a function of flange config-
(i.e., 1 inch along the arc of a circle passing uration and its rigidity; design values
through the flange cross-section centroid) as a range from 0.1 to 0.8
free body, the correlations between all forces We = end loads on the duct due to inertia and
acting on that segment, and the minimum required thermal effects (tension or compression),
design flange bolt loading, are established as lb
follows: Wb = minimum required design flange bolt load-
ing, lb
F1=F2+F3+F4+Fs (94)
2._ We The following design data are given for the

F2: \-7-5-] + ;-5 (9-2)
flange joint of the oxidizer-pump-discharge flex-
ible duct of the A-1 stage engine:
p(D22 - Dt 2)
Working pressure under normal steady oper-
F3 - 4 D (9-3)
ating conditions, 1505 psia
Maximum fluid working pressure under occa-
Sg(D32 - D22) (9-4) sional transient conditions, 1750 psia

F4 - 4 D
Inside diameter of the duet, D 1 =8 in
Inside diameter of the gasket, O 2 = 8 in
Sg =mp (9-5) Outside diameter of the gasket, D 3 = 8.5 in
End loads on the duct, due to thermal con-
F s = nF1 (9-6) traction, W e = 2400 lb
Gasket factor, m = 0.8
W b = FtzrD (9-7) Flange factor, n = 0.3
Determine the minimum required design bolt
where (referring to fig. 9-11)
loading of the flange joint.
D =diameter of the circle passing through the
eentroid of the flange ring cross section,
Solution
in
D_ =inside diameter of the flange and duct, in We will use a maximum working pressure
D2 = inside diameter of the gasket, in p= 1750 psia
From equation (9-5), the required average A = design factor, ranging from 4 to 6
gasket compressive stress: B =design factor, ranging from 10 to 14

A taper angle of 25 ° is generally used for the
Sg = mp= 0.8 x 1750 : 1400 psi hub portion of the flange ring.
The stress and strain analysis of the flange
Combine equations (9-1) and (9-6): ring may be treated as the twisting of a thick
circular ring of uniform cross section under the
(F2 + F 3 + F 4) influence of turning couples which are uniformly
Fi - (1 - n) distributed along a circle passing through the

eentroid of the ring cross section. The following
Combine this and equations (9-2), (9-3), (9-4), correlations approximate the maximum working
stress and strain of a flange ring (refer to fig.
and (9-7) to obtain the minimum required design
bolt leading of the flange joint: 9-11):
[prtDi 2 + 4 We + p,'r(D22 - D12) + Sgzr(D32 - D22)] Mtz FlYi + F2Y2 + F3Y3 + F4Y4- FsYs (9-11)
Wb: 4(l-n)
(D 4 - D)
[1750 × zr x 82 + 4 x 2400 + 1400 × r;(S.5 _ - 8:)] (9-12)
Y_: 2
Wb : 4 (1 - 0.3)
= 141 000 lb (D - D _ - t)
(9-13)
Y2-" 2
After the required design bolt loading has
been determined, the number, size, and torque
(2 D - D 2 - D1)
value of the bolts needed to produce that load (9-14)
Y3- 4
must be chosen. Since tensile elasticity is ad-
vantageous, the use of small-diameter, high-
(2 D- D 3 - D2)
strength bolts is desirable. Bolt spacing should (9-15)
Y4-
be sufficiently close to insure a reasonably even
distribution of the load around the gasket cir-
cumference. The following empirical correlation (D5 - D)
(9-16)
is recommended for maximum spacing to produce Ys:- 2
a tight joint:
MtD 2
Ps : 2 d + T (9-8) 0 : -- (9-17)
4 Ely_y
where
MtD2(L1 - L)
Ps = maximum bolt spacing, in
S z- 2Dlly_y (9-18)
d =nominal bolt dia, in
T = flange thickness, in
where
The general proportion of the flange ring may
be determined by the following empirical equa- D4 : diameter of the bolt circle,
in
tion (fig. 9-11):
D s =outside diameter of the
T=At (9-9) flange ring, in
Y_, Y2, Y3, Y4, Y5 :distances between the ring
cross-section centroid and
L_ = B_ (9-10)
forces F,, F 2, F 3, F 4 and
where F s , in
t =thickness of the duct wall, in, as deter- Mt = resultant twisting couple
mined by a hoop stress calculation per unit length of flange
L1 = overall axial length of the flange ring, in ring, in-lb/in

ly_y = the moment of inertiaof the Yield pres sure × D 2117×8

t=
ring cross sectionabout 2 x 170 000 x 0.75 = 0.066 in
2xFy×ew
the Y-Y axis, in4
E = modulus of elasticityof the From equation (2-10), ultimate pressure = 1.5
flangematerial,psi x 1925 = 2887 psi.
0 = angular displacement of the Based on ultimate strength, the thickness of
flange ringunder maximum the duct wall
working pressureand
loads, rad Ultimate pressure x D: _ 2887 x 8
S x = flange ringmaximum tensile t=
2 × Fu × ew 2 × 200 000 x 0.75
stress, which occurs at
the corner Z and is normal = 0.077 in
to the plane of figure 9-11, We will use the higher value and round it to
i.e., in the hoop direction t= 0.08 inch, as the selected wall thickness of
the duct.
Sample Calculation (9-2) Generally, based on given bolt size, gasket
diameters, and duct dimensions, a proposed
With the data given in sample calculation
flange configuration may be derived simply from
(9-1), design the flange for the oxidizer pump
good design layout practice. Then the proposed
discharge flexibleduct of the A-I stage engine,
design is checked for working stresses and
with the followingrecommended materialand
strains. The flange thickness T may be adjusted
boltdata:
accordingly. Assume flange design factors A = 6
Flange and duct material,Inco 718 and B = 12.2.
Minimum yieldstrength,Fy = 170 000 psi From equation (9-9). the flange thickness:
Minimum ultimatestrength,F u = 200 000 psi
Modulus of elasticity,E = 29.6× 106 psi T=At= 6 x 0.08=0.48 in
Duct weld efficiency,ew = 0.75
Bolt size and material,5/16-24 A-286 (200000
From equation (9-10), the flange axial length:
psi),stainless-steelbolt,0.526 head dia
Allowable ultimateboltload, 10262 lb
L_ =Bt=12.2xO.08=0.976 in, say 0.98
Solution Also, from a design layout, we obtained the

Use the flange design and nomenclature shown following additional values:
in figure 9-11. From equation (2-8):
D4 = 9 in
Design limit pressure Ds =9.62 in
= 1.2 × steady operating pressure = 1.2 x 1505 L = 0.37 in (approximately)

=1806 psi D = 8.6 in (approximately)
or ly_y = 0.0273 in 4 (approximately)
Design limitpressure From equation (9-7):
= i.i× maximum transientpressure= I.I× 1750 Wb 140400

= 1925 psi F, =_-_= 8.-------ff
r,× "= 5200 lb/in
From equation (2-9), yield pressure= 1.1 From equation (9-2):

× 1925=2117 psi.
In accordance with standard practice, the fpo,2xwo(1750×82
/ 2400
thickness of the duct wall is determined as F 2 = _--_D-]+,_= _ -_x-83g _/+--_-___.6 = 3347 lb/in
DEMON OF INTERCONNECTING COMPONENTS AND MOUNTS 365
From equation (9-3): stress under maximum transient pressure con-

ditions:
p(D 22 - D L2)
F3 - 4D -0 MtD2(L1 - L) _ 1174 x (8.6) 2 × (0.98 - 0.37)
Sz- 2Dlly_y 2 ×8x 0.0273
From equation (9-,1):
= 121 200 psi
F4 = Sg(D32 - D22) _ 1400 (8.52 - 82) _ 336 lb/in From equations (2-8) and (2-9), the yield
4D 4x8.6
load stress = 121 200 x 1.1 x 1.1 = 146 600 psi.
This is smaller than the minimum yield strength
Fy = 170 000 psi of the material.
From equations (2-8) and (2-10), tile ultimate
F s = aF 1 = 0.3 x 5200= 1560 lb/in
load stress= 121 200x 1.1 x 1.5=200 000 psi.
Thus, the proposed flange configuration is
satisfactory. We will now determine the number
of bolts and, from equation (9-8), the maximum
(D 4 - D) _ (9 - 8.6) _ 0.2 in bolt spacing
Yl=- 2 2
Ps=2d+ T=2×0.312 +0.48= l.104 in

The required number of bolts:
(D - D_ - t) _ (8.6 - 8- 0.08) = 0.26 in

Y2- 2 2 r;D4 n"× 9
P--T- 1.104=256' say 26
The required bolt loading W b calculated in
(2 D- D 3 - D2) _ (2 × 8.6 -8.5 - 8) _ 0.175 in sample calculation (9-1) was based on the maxi-
Y4- 4 4 mum transient pressure. Using equations (2-8)
and (2-10), the required ultimate bolt loading:
W b x 1.1 × 1.5= 140 400× 1.65=231800 lb
(D s - D) _ (9.62 - 8.6)_ 0 51 in
Ys- 2 2 " Therefore, the ultimate loading on each bolt
=231 800,/26=8900 lb. This is smaller than the
From equation (9-11): allowable ultimate bolt load of 10262 lb.
The required preload on each bolt:
Mr= Flyl + F2y 2+F_y 3 + F4y 4-Fsy s
Wb_ 140 400
= 5200 x 0.2 + 3347 x 0.26 + 336 x 0.175-1560 × 0.51 - 5400 Ib
26 26
= 1174 in-lb/in
Sheet Gaskets for Flange Joints

From equation (9-17), the angular displace-
ment of the flange under maximum transient pres- Gasket materials consisting of asbestos fibers
sure condition: bonded with rubber or neoprene are readily avail-
able in flat sheets, from which tim gasket may
be cut to suit most requirements. Frequently
MtD 2 1174 x(8.6) 2
8=
used sheet gaskets range from 1/32 to 1/8 inch
4 Ely_y 4 × 29.6 x 106 ×0.0273
in thickness. At fluid pressures less than 50
:0.0268 rad or 1.54 ° psi, sheet gaskets can be used with flat faces
at both flanges. For higher pressure applica-
From equation (9-18), the maximum working tions, sheet gaskets can be secured in a flat,
concentric groove machined on one of the flanges,

with the groove depth sized for proper preloading
of the gasket. Preferably, the flange facing
should be serrated concentrically. In rocket
_SEAL SURFACE FINISH fl'
engine applications, sheet gaskets should be IT $1"(OULD EE FREE OE
NICKS, SCR,ATCHES AND
limited to pressure levels of less than 200 psi, OTHER Ih_ERFECTION$
EXCEPT ANNUblk:_ TOOL
and to temperatures in the range from -100 ° F to MARKS
900 ° F.
Elastomer and Metal O-Rings for Flange Joints BRIAR COR_ERS 005"
_--i
D3 --
D2
_ G_OOVE SIDE FINISH f2
Elastomer O-rings have been used with good

results for rocket engine flange joints in medium
temperature service (-60 ° to 400 ° F). Satisfac-
tory results have also been experienced at pres-
sure levels as high as 3000 psi. Typical elas-
NIC*:S, SC_,TC._ES J,t'4D
tomer O-ring flange joint design data are OT_E_ IMPER_E£TIO_S
EXCEPT Ar_,'_UL_R TOOL
presented in table 9-5 for use in groove-type

flanges, as shown in figure 9-12. Table 7-3 may
Figure 9-I 2.-Flange seal groove design.
be used to select the hardness of the O-ring
materials with respect to working pressures.
Elastomer O-ring seals for various flange size
may easily be made by cutting straight O-ring sn_t._ Metal O-rings can be used in a groove-
stock to a specific length, and cementing the type flange, as shown in figure 9-12. Table 9-6
ends together. presents typical design data for metal (3-ring
For low- and high-temperature services (-430 ° flange joints,
to 2000 ° F), and for sealing against very high
pressures (up to 4000 psi), metal O-rings may be
used in rocket engine flange joints, with proper
Spiral-Wound Gaskets for Flange Joints
detail design. Hollow stainless-steel O-rings,
such as manufactured under the trade name Spiral-wound gaskets manufactured by the
"Toruseal" by the D.S.D. Co., have been used Flexitallic Gasket Co. have been used exten-
extensively in various services. They can be sively in rocket engine flange joints for liquid
coated with Kel-F and Teflon for cryogenic serv- oxygen and hot gas (up to 1600 ° F) services.
ice. One of the advantages of metal O-rings is Pressures are generally kept below 1000 psi.
that they require considerably less space than The gaskets (detail shown in fig. 9-13) are made
other seal types. However, the O-rings must of spiral-wound 304 stainless-steel ribbon, with
have a good seal surface finish, and high com- a Canadian asbestos filler. It is not recom-
pressive seal loads, to assure satisfactory re- mended that gasket widths of less than 3/16 inch
TABLE 9-5.-Typical Elastomer O-Ring Flange Joint Design Data a

[All dimensions in inches]
Nominal Actual Bottom

Minimum Depth of groove, Seal surface
O-ring O-ring Width of groove, radius, Seal surface
squeeze a b finish, fl finish, f2
crosssection cross section i"
1/16 0.070 *_0.003 0.017 0.045_ 0.005 3/32 1/64

3/'33 •103 * .003 .020 075± .005 9/64 1/64
1/8 •139 -* .004 .025 •1052 .005 3/16 1/32 64 rms 64 rms
3/16 .210 t .005 .030 .170± .005 9/32 1/32
1/4 .275 _* .006 .039 .225± .005 3/8 1/16
TABLE 9-6.-Typical Metal O-Ring Flange Joint Design Data a

Actual O-ring size Flange groove dimensions
Nominal O-ring, OD rube OD [ Tube OD I

Wall D2 [ Ds + 0.005 a+0.005
+0.005 I +0.003
thickness - 0 [l
-0 maximum ] - 0
,
1 .................................. 1.000 ] 0.061 0.009 0.867 I 1.005 0.042

1-1/2 .............................. 1.500 ] .061 .009 1.367 f 1.505 .042
2 .................................. 2.000 ] .061 .009 1.867 ] 2.005 .042
2-1/2 .............................. 2.500 ] .061 .011 2.367 ] 2.505 .042
3 .................................. 3.000 I .061 .011 2.867 ] 3.005 .042
4 .................................. 4.000 I .094 .011 3.801 [ 4.005 .075
5 .................................. 5.000 ] .094 .011 4.801 I 5.005 .075 32 rms
6 .................................. 5.000 I .094 .011 5.801 i 6.005 .075
8 .................................. 8.000 1 .124 .011 7.741 ] 8.005 .105
10 .................................. 10.000 1 .124 .011 9.741 I 10.005 .105
15 .................................. 15.000 ! .124 .011 14.741 I 15.005 ,i05
20 .................................. 20.000 ] .124 .011 19.741 I 20.005 .105
30 .................................. 29.741 I 30.005 .105
30.000 1 .124 .011
aSee fig. 9-12 for flange design. Duct diameter D, may be equal to D,..
be used. These gaskets are made to various centric. Typical design data for spiral-wound
stiffnesses (spring rates) by varying the tension gasket flange joints are presented in table 9-7.
of the wrapping. The harder gaskets are recom- Gaskets of this type require high compressive
mended for the higher pressures. loads. The values range from 6000 to 25 000psi.
The gaskets are used in groove-type flanges, The amount of gasket compression is concrolled
as shown in figure 9-12. For optimum results, by metal-to-metal contact of the flanges, with
tool marks on sealing surfaces should be con- allowance for maximum tolerance buildup of
TABLE 9-7.-Typical Spiral-Wound Gasket Flange Joint Design Data a

Approximate re-
Gasket dimensions Flange groove dimensions
quired compressive
load, lb at deflection t
A diameter B diameter Sealing area, I D2 I D: r [,
_'0.016 t 0.032 sq. in. 0.025 0.035
tO.OlO t ±0010
i ]
1.000 1.375 0.442 82O0 10 600 0.9001 1.425 I

e_ e,
1.500 1.875 .638 8300 11 500 1.400 1.925 I
2.000 2.375 .835 9000 12 700 1.900 i 2.425 l
2.500 2875 1.031 9800 14 100 2.400 2.925 ]
3.000 3.375 1.227 I0 800 15600 2.900 3.425 ]
4.000 4.375 1.620 12950 18 600 3.900 4.425 I
5.000 5.500 3.056 23000 32 400 4.875 [ 5.560 ] 0.010 125 rms
6.000 6.500 3.645 26400 37 000 5.875 6.560 } +0
8.000 8500 4.832 33 600 43 000 7.875
8.560[ _
Ii .000 11.750 11.110 73 600 97 000 10.875 11.810 I c_
c5 c_ c5 c_
15.000 15.750 15.040 98 600 122 500 14.860 15.820 I + , -'-
20.125 21.000 24.150 153000 178500 19.985 _i.005
25.000 25.875 29.900 186 500 206000 24,840 25.955 j _ d
aSee fig. 9-13 for seal design. See fig. 9-12 for flange design. Duct dlameter D, may be equal to D2.
._PIRAL WOUND EJSd

r STAINLESS STEEL RIBBON
A_*BESTO$
Fs, = 8 (H - 0.02)_C, (9-19)
FILLEII
p(H - 0.02)
ENL _t_GED •
Fs2 = Fsl + 2 (9-20)
where
II/ T----T/V- II '_" "="
Fs_ = seal contact load per inch of seal circum-
ference without internal fluid pressure,
lb/in
Fs2 = seal contact load per inch of seal circum-
_¢ .... C-C ference with internal fluid pressure,

lb/in
E =modulus of elasticity of the sealmaterial,
Figure 9-I3.-Spiral-wound gasket for llange
psi
joints.
J =end thickness of the seal leg, in
H =length of the seal leg, in
d = seal deflection in the flange joint, in
gasket thickness and groove depth. One diffi- (i.e., the difference between initial seal
culty with spiral-wound gaskets is the presence width M and flange groove depth)
of erratic compression loads, due to thickness p =internal fluid pressure, psi
tolerances and variations in wrapping. Another

is their tendency to indent the flange sealing
surfaces, particularly when the flanges are of
M.*,TERIAL ¢340 STEEL H'g#.T TREATED TO _SO 0'_ - 18OOOO
aluminum. FINISH - _L_CTROLI;SS NICRFA. PLATE. 00_0 - 0012 INCH £NBRITTLE.M_;_"i" _ELIL',rED
TEFLON COAT Co1 - 0,_ INCH
Pressure-Actuated Seals for Flange Joints
seal
An ingenious
deflection
approach
capabilities
toward
to accommodate
attaining better
de-
_
,,
K : 7 ' -oo3
,)i oo._ooo.
'' -:oz
formations of the flange joint under working

loads is the pressure-actuated type seal, shown
....... - ....... J. I,.L_o? sEc'r_. A -A
in figure 9-14. The spring effect is achieved by
a metal, circular seal having a U-shaped cross THESE SURFACES TO BE FRE:E OF NICKS B_RS _ND TOOL MARKS EXCEPT ANNULAR TOOL
_tARCS
section. When such a seal is compressed be- _-_)FLAT I¢'I%"rllN _3(]_ INCH P£R INCH CIKCUM} £R_t4C£
tween the surfaces of a flange joint, the high @ PARALLEL TO SURFACE X IHTHI_ _J:_ INCH PER INCH CIRCU=FER[;N_CE.
/_J THICKNE&S OF TEFLON COATING ON THESE SURFACES OPTIONAL

points at its open ends make contact with the
flange surfaces. If properly designed, the spring-
like legs flex to follow flange deflections.
The open end of the seal must always be
oriented toward the inside of the flange joint.
Thus the internal fluid pressure is utilized to
generate a considerable portion of the seal con-
tact load. Pressure-actuated seals are installed
in a concentric groove cut into the flange joint.
This limits the amount of total flange preload
OPEN GRO0_I_ CLO_EI) GROOVE
transmitted through the contact points of the L_T #,LLA TION INSTALLS'lION
seal. The following correlations are established

for the seal design shown in figure 9-14 for the Figure 9-I 4.-R o c k e td y n e Naflex pressure-
approximation of seal contact loads: actuated seal design [or [lange joints.
C I =design factor. This is a function of seal tc :thickness of tllecoating, in

leg configuration, its value ranging from Many versions of pressure-actuated seals
0.04 to 0.8 have been applied to rocket engines. Rocketdyne
Generally, the contact surfaces of pressure- has developed the Naflex-type pressure-actuated
actuated seals are coated with a material of a seals, made of alloy steel and coated with a thin
lower elastic modulus than that of the material layer of Teflon. They are most effective for
from which the seal is made. Thus the seal cryogenic (-430 ° F) and high-pressure (up to
material can be utilized to achieve the spring- 4000 psi) services. Figure 4-32 illustrates a
like p,'opertiesrequired for maximum deflection typical Naflex seal installation at tileflange
capability, while the coating material provides joint between thrust chamber and liquid oxygen
the properties needed to establish tileintimate dome. Design details and data for typicaINaflex
contact required to prevent leakage between seal flange joints may be seen in figure 9-14. Also
and flange surfaces. To prevent leakage, suffi- see table 9-S. For hot gas applications, up to
cient compressive stress must be developed in approximately 1700 ° F, the seals may be coated
the coating to make it flow into imperfections on with soft metals such as copper, aluminum, or
the flange surfaces. The required coating com- _old.
pressive stress may be determined by For certain stringent applications, all propel-
lant and turbine gas flanges in a rocket engine
Eoh can be equipped with a Naflex-type dual static
S c : -- (9-2i)
tc seal incorporating an intermediate cavity be-
tween high-pressure and low-pressure sealing
where areas. A seal drain line may be connected to the
Sc = seal coating compressive stress, psi intermediate cavity, as shown in figure 9-15.
Ec = modulus of elasticity of the coating mate- This line can also be used to measure the seal
rial, psi leakage rate, ifany. This arrangement may also
h =depth of the imperfections on the flange be applied to pnenmatic system flange joints, to
sealing surfaces, in monitor leakage.
TABLE 9-8.-TypicaI Design Data o[ NaHex Pressure-Actuated Seals and Flange Joinzs
Seal dimensions I Flange groove dimensions

J
A diameter B diameter!
H D2 D3 a
+0 I *0 +0 +0.004+0 + O. 002 r [,
K + 0J - ooa.M03 R ac
- 0.005 (REF) [ - 0.005 i- 0.005 -0 - 0,003 +00.005f+0 _-0.010 -0
- 0.010 -0
0.934
1.496
2,058
2.621
0.464
1.026
1.548
2.111
235_
I ,o.,o,[ " l
.155 [
' 0,190 0.433
190 '
l
.995
1.507
2.070
0.948
1.510
2.072i 0.14S
2.6351
i
3.080
3.620
4.618
2.570
3.110
4.108 |
I3
o .210 2529
3.070
4.070
3.0941
3.6351
4.6351
1
0.020t0.005 32 rms
I 1
4
5.616 5.116
6.614 6.114
I
6.070 6.635,
5.070 5.635' I
8.610 8.110 8.070 8.635i
10.606 10.106 .250 .017 . .1 4 .205 10070 10.635 .173
15.596
20.870
15.096
20.370 { 15.070
20.320
15.635 i
20.$85 i
L
aDimension increases 0.009-meh maximum after mekel plating and Teflon coating.
NOTES: 1. See fig. 9-14 for Naflex seal design.
2. See fig. 9-1_ for flange design.
3. Duct diameter D_ may be equal to D2.
From figure 9-14, the seal land of contact

=0.019. Thus, the seal coating compressive
stress without imernal pressure
582
Sc = 1 x 0.019 - 30600 psi
From equation (9-20), the seal contact load

with internal pressure
2000 (0.205- 0.02)_ 767 lb/in

Fs2 =582+ 2 -
"--- NAFLEX TYPE DUAL STATIC SEAL

'-- SEAL DRAIN LINE
The seal coating compressive stress with an
internal pressure of 2000 psi
G AREA
767
Sc = 1 × 0.01---_
= 40 400 psi
_ The allowable
flange sealing
depth of imperfections
surfaces should be calculated
on the
from the initial coating compressive stress,

Figure 9-]5.-Typical design of a flange joint using equation (9-21):
with Naflex-type dual static seal and seal
drain line.
h- Sctc 30600x0.002=0.0012 in
Ec 5×104
The following data are given for a Naflex 9.4 DESIGN OF BRAZED JOINTS FOR
seal represented by section A-A of figure 9-14: ROCKET ENGINES
Design factor, C 1 =0.07; H=0.205 in; J=0.017
in; M=0.187 in. The seal material modulus of The best tube fitting or gasketed flange
elasticity, E=30× l0 6 psi. The flange seal joint falls short of a good welded or brazed joint
groove depth, a=0.173 in. The thickness of the in reliability. Hence, every sealed joint on each
Teflon coating, tc = 0.002 in, with a modulus of rocket engine should be studied aiming at its
elasticity, Ec = 5 x 104 psi. Estimate the seal elimination, or its replacement by a hermetically
coating compressive stresses with and without sealed joint, for increased overall system relia-
an internal pressure, p, of 2000 psi. Also, de- bility. Specifically, certain sealed joints origi-
termine the allowable depth of imperfections on nally provided for convenience during the devel-
the flange sealing surfaces. opment phase of an engine system may no longer
be needed as development aDproaches completion.
The brazed joint design shown in figure 9-16
Solution
may be utilized effectively to replace tube fit-
The seal deflection tings and flange joints. A union sleeve with
internal recesses for preplaced braze alloy foils
d = M - a = 0.187 - 0.173 = 0.014 connects the tube ends. The brazing can be
accomplished by induction-heating coils, either
From equation (9-19), the seal contact load in-place on the engine, or on the bench. Coax
without internal pressure cables with low impedance losses have been
developed, permitting a brazing fixture to be
80 × 106 × (0.017) 3 x 0.014 used in-place, even at great distances from the
= 582 Ib/in
Fsl- 8 (0.205- 0.02) a ×0.07 generator.
.__.- -" .... _-- _:_,S,

__-CLEAP, ANCE FOR BRAZING FIXTURE successfully induction brazed. Tubes and
sleeves for brazed joints can be made of alumi-
1 A /------'---INTERNAL RECESSES FOR PRE-PLACEO i num alloys, austenitic and semiaustenitic steels,
and nickel-base alloys such as Monel.
A preplaced braze alloy in the form of 0.001-
to 0.003-inch-thick foil is snapped into recesses
inside of the union sleeve. Diametrical clear-
ances of 0.002-0.008 inch between sleeve and
tubing are found to give satisfactory results.
I :
Holes are provided on the sleeve to permit visual
joint inspection after brazing. The quality of
the brazed joint can be also judged by the de-
gree of external braze-alloy filleting at the outer
edges of the sleeve. The recesses for the braze
alloy are located midway between tube and
sleeve ends. The length of the union sleeve can
be determined by the braze bond shear strength
and the effective bond area. Sleeve lengths
_---- UNION SLEEVE l
range from 0.5 to 2 times the tube outside diam-
J
eter.
Many types of brazing alloys can be used. A
Figure 9-16.-Typical brazed ]aint design. eutectic alloy of composition 71.8 percent silver,
28 percent copper, and 0.2 percent lithium, with
a melting point of 1435 ° F, can braze without
Pertinent design consideration and details of flux and permit in-place brazing.
brazed joints for rocket engines are as follows:
System Applications of Brazed Joints
Working Fluids, Pressures, and Temperatures
After an engine system has reached its pro-
Applicable to Brazed Joints
duction phase, the majority of the small- or
There are virtually no design limitations on medium-size sealed fittings and flange joints
the types of fluids that can be handled by brazed should be replaced by well-located brazed joints.
joints. Chemically active propellants such as These should be designed to facilitate in-place
fluorine, hydrazine, etc., have posed no problems brazing, if possible. This requires a specific
to brazed joints. Similar satisfactory results clearance around the joint for the brazing fixture.
have been obtained with brazed joints used in The minimum clearance A for the brazing fixture
liquid hydrogen, liquid oxygen, hydraulic, and (fig. 9-16) is 5/8 inch for tube diameters of 1/4
pneumatic lines. to 5/8 inch; 3/4 inch for a 3/4-inch diameter;
Structurally, brazed joints are designed to be 7/8 inch for a 7/8-inch diameter; and 1 inch for
as strong as the strongest tubing or duct of like 1- to 1-3/4-inch diameters. For repair of lines
material. Brazed joints have demonstrated relia- or servicing of components with brazed joints, a
bility at working fluid pressures as high as 4000 length of line may be cut out and a new length
to 5000 psi. Generally, brazed joints are recom- inserted and brazed in-place, with two union
mended to be limited to service temperatures of sleeves.
less than 1200 ° F. Allowance must be made for Figure 9-17 presents the in-place brazed
strength degradation of brazed joints used at joints applied to an upper stage propulsion sys-
elevated temperatures. tem. About S0 percent of the original fittings
A typicalbrazed jointis illustrated in figure and flanged joints were replaced by brazed
9-16. The union sleeve can be used to join tube joints, which eliminate considerable weight. In
to tubes, tube to bellows, or tube to components. addition, it also permits the use of high-strength,
Line sizes as large as 6 inches (insidediam- thin-wall tubing, instead of the older heavy-wall
eter)and walls as thinas 0.005 inch have been tubing.
allow fortemperaturechanges, dynamic loads,

and engine gimbal effects,ifany. This often
determines the linegeometry. In general,how-
ever, lines should be as shortand straightas
possible.
5. Structural integrity.-The structural design
of a duct system may involve a number of com-
plex problems. This is especially true with
ducts for elevated or very low temperatures, and
requiring a high degree of flexibility. The de-
sign and fabrication of metal bellows for applica-
tion in ducts is a highly specialized technology.
Duct Design for Minimum Pressure Drop
Figure 9-17.-In-place brazed joints applied in Techniques for fluid-flow pressure drop deter-
an upper stage propulsion system. mination have been discussed in chapter VII.
Equation (7-7) may be used to estimate the pres-
sure drop of a straight duct section. Either
9.5 DESIGN OF DUCTS FOR ROCKET
equation (7-7) or (7-9) can be utilized to predict
ENGINES
the pressure drops of other shapes, for which
some experimental data are available.
Basic Design Considerations Besides avoiding excessive flow velocities
Primary design considerationsforvarious at various sections of a duct system, basic re-
ducts used on rocketengines are: quirements for minimum pressure drop are: con-
1. Fluid flow rate and system pressure drop.- stant flow area, smooth flow surfaces and path,
The size of the duct is determined largely by the minimum length, and few turns. In most duct
required flow rate and permissible pressure drop designs, a certain number of turns is unavoid-
of the flow system. An optimization must be able. The design of these turns affects consid-
made considering duct weight, pressure drop, and erably the overall pressure drop of a duct system.
space, before finalizing the duct size. A turn may offer a large resistance to the flow
2. Working temperature and pressure of the if not carefully designed. Investigations show
fluid.-The selection of construction materials that the flow resistance in a bend with constant
and structural design for ducts depends mainly cross section is affected directly by its turning
on the working temperature and pressure level of radius ratio R/D (where R = radius of curvature
the fluid, but also on chemical compatibility. of the axis of the duct, D=diameter or width of
Temperature also influences line flexibility and the duct) (see fig. 7-21). By increasing the turn-
duct geometry. ing radius ratio, large reductions in pressure
3. Duct/oints.-The number and type of joints drop are possible. Bends with a circular section
in a duct system are usually determined by sys- are better than those with a square section.
tem assembly and servicing requirements. All However, the square section may be much im-
mechanically sealed joints should be reviewed, proved by changing it to a rectangle, so that the
at various engine design and development phases, turn is made on the short side; that is, the loss
for their possible elimination or replacement by can be decreased by increasing the rectangle
in-place, induction-brazed or welded joints. (See aspect ratio W/D (where W= width of the long
secs. 9.8 and 9.4.) side, D = width of the short side). For a duct
4. Line flexibility and geometry.-The re- with a rectangular section, therefore, a very
quired degree of flexibility in a duet system is efficient corner may be produced if the values of
dictated by component tolerance buildups, mis- both the radius ratio and the aspect ratio are
alinements, and required freedom of movement to kept large.
This consideration should be applied in all

cases where a sharp turning corner is unavoid- SMOOTHED
DIAGONAL
able, such as the elbow-type turbopump inlet WALL
ducts shown in figure 6-15. Here, the loss JUNCTION
around the sharp corner is due largely to the

zero radius ratio. To reduce this pressure drop,
guide vanes can be very effectively used. If the
junctions of the duet walls on the diagonal are
smoothed out and vanes added as shown in figure
9-18, the compartments so formed may be madeto
have a good radius ratio and a high aspect ratio.
r * 1¼
Since in most cases only a few vanes need to be
added for high aspect ratios, the remaining prob-
Figure 9-18.-Elbow-type sharp turning duct with
lem is one of securing a good radius ratio.
guide vanes.
The radius and aspect ratios of each compart-
ment are increased as the gap/chord ratio S/C is
reduced (fig.9-18). Theoretically, then, the I
corner pressure drop will decrease as tileratio

S/C is decreased. However, each additional
THRUST CHAMBER/-_p_i -, _Q_
vane adds more surface area and blockage, and
cop sequently tends to increase losses. The S/C
ratio can be optimized in tests and ranges from ---÷--- I _-
0.25 to 0.45. Well-designed elbow-type sharp [l + ") x'x i , 1 ', ] UNIVERSAL
turns with guide vanes may have resistance

coefficients K as low as 0.i to 0.2 (see eq. 7-9).
Duct Design for Flexibility

. _
For most interconnecting ducts between two

BEILvL2226! .Inlet ]INLET CONNECTED
engine components, the required flexibilitysuch
................. _ TO VEHICLE DUCT
as for misalinements and temperature changes
can be achieved by placing one or two bellows
sections in the line (figs. 9-4 and 9-5). These
may be either restrained or unrestrained, depend-
ing on structural loading considerations. In
ducts connecting two components involving large
relative movements, a minimum of three flexible
OUTLET CONN CTED _--r-,k}..__._ CHAMBER
sections is required. The longitudinal axis of at
least two of the bellows sections should be posi-
tioned at a right angle toone another. Restrained
ENGINE GIMBAL
bellows are preferred.
The duct systems shown in figure 9-3 illus-
Figure 9-19.-Typical propellant supply duct de-
trate a basic three-bellows arrangement as ap-
signed for the ilexibility required ior engine
plied to gimbaled clustered engines. Figure 9-19
gimbaling.
presents one of the frequently used design
approaches for a relatively short duct system to
facilitate large movements during gimbaling of iently incorporated into this duct system by
an engine system. The three flexible sections welding. These joints are made of stainless
are kept in the plane of the engine gimbal point. steel and will operate at cryogenic and elevated
This results in minimum displacements of the temperatures (up to 1000 ° F). The internal re-
sections for a given engine movement. Bellows- straining mechanism is designed to give a smooth
type universal joints (fig.9-20) may be convert- flow passage for low-pressure drop.
ally expanding thin-wall tubing, at spaced inter-

vals along its axis, coincident with an axial
--PLAIN END FOR WELDED compression, to a configuration as shown in fig-
INSTALLATION
ure 9-21. Our discussion will be confined to
this type of bellows. The various structural
FORK ASSEMBLY characteristics and design correlations of bel-
lows are presented using the following nomen-
clature (see figs. 9-21 and 9-22):
--.._ GIMB/kL RING
_ ---'_"--_ ASSEMBLY Ct :bellows wall-thinning correction factor

RIVET Cp = ply interreaction factor (1.00 for 1-ply bel-
lows, 0.90 for 2-ply bellows, and 0.85 for
3-ply-or-more bellows)
d o : outside diameter of the bellows, in
di = outside diameter of the convolution root of
the bellows, in
d m = V'(di 2 + do2)/2 = root-mean-square diameter
of the bellows, in
d d =mean duct diameter, in
E --modulus of elasticity of the bellows mate-
rial, psi
e a = axial deflection of the bellows, in
eb = equivalent axial deflection of the bellows
,?+o + due to pure bending, in

ep = equivalent axial deflection
due to parallel offset, in
of the bellows
e s = equivalent axial deflection of the bellows

due to pure shear, in
F s = shear load, lb
Fp = pressure separating load, lb
G = shear modulus of elasticity of the bellows
material, psi
b : (d o - di)/2 : mean convolution height, in
L --pitch of the bellows (axial length of a con-
volution), in
Figure 9-20.-Typical gimbal-ring-type bellows
La = (Nc - 0.5)L + Npt = free axial length of the
universal joint designed and manufactured by
bellows, in
Marman Division of Aeroquip Corp.
Nc = number of bellows convolutions
Np =number of bellows plys
For higher pressure applications (over 200 t = thickness of the bellows wall, in
psi), external wire braids or links may be used Lb = axial length of the rigid duct section, in
to restrain the bellows section. The braids can M =bending moment, in-lb
be made of stranded stainless-steel wires, about p =internal (or external) fluid pressure, psi
0.012-inch diameter. In extreme cases, more Pcr = critical stability pressure of the bellows,
than one layer of braid should be used with re- psi
gard to the high separation loads. R a = axial spring rate of the bellows, lb/in
R b = bending spring rate of the bellows, lb-in/
degree
Structural Design of Bellows For Flexible Ducts
Rp : parallel offset spring rate of the bellows,
Most of the bellows used for flexible ducts in lb/in
rocket engines are hydraulically formed by radi- Rs -- shear spring rate of the bellows, lb/in
,XIAL
"qF--- OEFLECTION
(a)8EL.LO'_$ UNDER PURE

SENDING
{hi _ELLOW$ UNO£R L

PURE SHEAR
i
(¢} BELLOWS UNDER

PARALLEL OFFSET
1.2
I,I
(diPARALLEL 1.0
OFFSET DF .--L°---_
ARTICULATED p.--- L--_I
BELLOWS 0.9
0.8
07
0.6
Figure 9-21.-Elements and various motions of
bellows. 0.5
0.4
0 JO 20 30 40 5O 6O
Rt = torsional spring rate of the bellows, lb-in/ PERCENT THINNING
degree
Sb :bellows bulging stress, psi Figure 9-22.-Bellows wall thinning correction
Sa =bellows hoop stress, psi [actor Ct versus percent thinning.
Sm = bellows motion stress, psi
Ss :bellows shear stress, psi the minimum thicknessat the convolution outside
St =bellows torsion stress, psi
diameter(fig.9-22). Amounts of thinningrange
T =torsional moment, in-lb
from I0 to 40 percent. Effects of thinning,
are
Tcr : critical stability torque of the bellows, psi
considered by applying the thinningcorrection
v = Poisson's ratio of the bellows material, in
factorCt (fig.9-22) to the bellows design.
y = transverse deflection of the bellows, in
2. Bellows axial spring rate:
0 =bending angle of rotation, degrees
6 =torsional angle of rotation, degrees
1.49 CtCpNpEdit _ lb/in
1. Thinning of the bellows wall.-Hydraulic- Ra - Nch_
formed bellows are usually made by starting with
(steel and nickel alloys) (9-22)
tubes of the same diameter as the bellows diam-
eter at the root of convolution, d i. The typical
thinning profile of a bellows wall starts with the 1.23 CtCpNpEdit 3 , lb/in
Ra- Nch _
original material thickness at the root of the
convolution, and tapers approximately linearly to (aluminum alloys) (9-23)
3. Bulging stresses.-These are the radial 1.78 CtEtea

bending stresses induced in the side walls of the Sm : Nch2 , psi
bellows, by internal or external pressure. Bulg-
(aluminum alloys) (9-28)
ing stresses should be kept below those listed
in table 9-9:
All other types of bellows motions, such as
angular and parallel offset, can be converted into
S Ctph 2
an equivalent bellows axial deflection, and ap-
b : 2--'ffp-_'pt2.
psi (9-24)
plied to equations (9-27) and (9-28) to calculate
the corresponding motion stresses.
4. Separating loads.-Bellows are pressure
7. Bellows under pure bending (fig. 9-21a):
loaded so that they experience a separatingload,
in addition to the normal duct axial force: M dm2Ra
Rb =_- 458.4 ' lb-in/degree (9-29)
P(dm 2 - dd2)_
Fp- 4 , lb (9-25)
dm sin 0
, in (9-30)
eb- 2
5. Hoop stresses.-The bellows hoop stresses
are calculated by using the total area per inch of Motion stresses due to pure bending can be cal-
bellows axial length of a tube of equivalent wall culated by substituting eb for ea in equations
thickness. Hoop stresses should always be kept (9-27) and (9-28).
lower than the yield and nltimate strength of the 8. Bellows under pure shear (fig. 9-21b):
bellows materials by a specified margin.
Fs 3dmSRa
Rs : lb/in (9-31)
(9-26) y 8 La 2 '
Sh : 2--'_p_
229.2 FsLa
6. Motion stresses due to axial de[lections of (9-32)
0=' Radm _ , degrees
the bellows.-These are due to bending of the
bellows side wails. Allowable motion stresses
3dmY
for bellows materials, with respect to design --, in (9-33)
es-- 2L a
cycle life, are given in table 9-9:
Motion stresses due to pure shear can be calcu-

1.40 CtEtea
lated by substituting es for ea in equations
Sm: Nch2 , psi
(9-27) and (9-28).
(steel and nickel alloys) (9-27) 9. Bellows under parallel oHset (fig. 9-21c):
TABLE 9-9.-Yield Strength, Limiting Bulging Stresses, and Allowable Motion Stresses of Frequently
Used Bellows Materials
Limiting bulging Allowable motion stresses,

Yield stresses,psi psi
Material strength,
psi 1000 10000 100000
t_0.012in t_0.013 in
cycles cycles cycles
321 and 347 stainless steels ........................ 39 000 140000 120000 208000 150000' 92000
19-9DL ............................................ 88000 140000 120 000 208000 1500001 92000
A-286 ............................................ 180000 190000 150000 160000 150000 138000
lnco 718 .......................................... 170000 190000 150000 160000 150000 138000
lnconel X ......................................... 98 000 190 000 150000 160000 150000 138000
6061-T6 aluminum alloy ............................. 40 000 65000 106000 68000i 28000
Fs 3dm2Ra 13. Bellows buckling due to external pres-

Rp - y - 2 La 2 ' Ib/in (9-34) sure.-When a bellows is pressurized externally,
it buckles in same manner as a thin cylinder:
FsLa 3 dm2YRa
M =_--=-+ + in-lb (9-35) 4 EtNph 3
2 4 La 2 '
Per-(1 _ v2)dmSPb, psi (9-43)
3dmy
14. Bellows squirm due to torsion.-When a
ep----L--_a,
in (9-36)
bellows is loaded by pure torsion, it tends to
buckle in some manner as with internal pressure
Motion stresses due to paralleloffsetcan be
squirm:
calculatedby substitutingep forea in equations
(9-27) and (9-28).
rrdm2Ra
10. Parallel offset of articulated bellows Tcr- 2 , in-lb (9-44)
(fig. 9-21d):
The values for Pcr and Tcr of bellows under
Fs 3dm2Ra angular and offset deflections will be reduced
Rp=-7=- 2 considerably. A correction factor which ranges
from 0.2 to 0.9, as determined by experiments,
[4La 2+6LaLb+3Lb2],Ib/in (9-37)
should be applied.
Fs(2 La + Lb) 15. Allowable working stresses for bellows

M--±
2 materials.-Bellows generally operate in the
plastic range. However, it has been the practice
3dm2(2 La + Lb)yRa
in-lb (9-38) to calculate stresses on the basis of elastic
= 4(4 La 2 +6 LaLb+3 Lb2) '
deformation, and correlate the data on that basis.
In reality, the calculated stresses are only index
3 din(2La + Lb)Y
(9-39) stresses, which define the plastic conditions of
ep_(4 L 2 + 6 LaLb + 3 Lb 2),in the bellows, under the influence of pressure and
motion loading. The yield strength, limiting
ep can be substitutedforea in equations(9-27) bulging stresses, and allowable stresses of fre-
and (9-28)to calculatemotion stresses. quently used bellows materials are presented in
table 9-9.
Ii. Bellows torsion.-The stress due to bel-
lows torsion is given by the twist of a thin tube: 16. Bellows used at elevated temperatures.-
For applications at elevated temperatures allow-
2T
able working stresses for bellows materials must
St- _"prrdi2 t , psi (9-40) be adjusted accordingly. Generally, an internal
liner should be provided to protect the bellows
T 1.37 × i0 -2 GditNp, (9-41) against high-velocity, hot-gas flows.
Rt lb-in/degree
(_ (2 h +0.57 L)Nc
17. Bellows forming lin_,its.-Bellows designs
12. Bellows squirm due to internal pressure.- are limited by how severely the material can be
When a restrained bellows is pressurized inter- worked during forming. Generally, the following
nally beyond a critical level, it experiences a geometric limits should be used for bellows up
stability failure of the same type as a buckling to three plys:
column:
do
5.02 Ra Maximum: _ = 1.35 (9-45)

Per= d/, \' psi (9-42)
La
Minimum: L = (8 + 2 Np)t, in (9-46)
Design a bellows, as shown in figure 9-4, for din= _/(di2+ 2 d°2) - _ (82+ _ '822)
-- - 8.42 in
the oxidizer pump discharge flexible duct of the

A-1 stage engine, with the following data and From equation (9-30), the equivalent axial
requirements, in addition to those given in deflection of a bellows, due to pure bending or
sample calculations (9-1) and (9-2). angulation:
Bellows material, Inco 718
Thinning of bellows wall, 20 percent tim sin 0_ 8.42× sin 3°
Outside diameter of the convolution root of eb = 2 2 = 0.22 in
the bellows, di= 8 in
Free axial length of the bellows, La = 7 in Table 9-9 lists, for Inco 718 and a life
maximum
of 10 000 cycles, an allowable motion stress of
Required angular movement, 0= ±3 ° 150000 psi. It is good practice, however, to use
Life, 10 000 cycles a lower value for high-pressure bellows designed
In addition, determine the following:
for improved stability. In our design, we use a
Bellows axial spring rate, Ra value of 0.36 times the limiting bulging stress.
Bending or angular spring rate of the bellows, Thus, the motion stress = 0.36 × 150 000 = 54 000
Rb psi. Substitute this and eb into equation (9-27):
Bending moments of the duct at 3 ° angular
motion, M I AO CtEteb
Required restraining link load at maximum
Sm- Nch2
working pressure
Solution 1.40 CtEteb

(from fig. 9-22, Ct=l for 20 per-
Nc - Smh 2
We will first use the limiting bulging stress cent thinning)
of the bellows material to establish the ratio of
convolution height h to wall thickness t. Prom _ 1.40× 1× 29.6× 106× 0.022x 0.22 = 22.1,
- 54 000 × (0.410) 2 say 22
figure 9-22, wall-thinning correction factor
Ct= 1.36, for 20 percent thinning. From table
9-9, the limiting bulging stress for Inco 718 at From equation (9-46), the pitch of the bellows
t>0.013 inch is 150000 psi. From sample cal-

culation (9-2) the design limit pressure of the L=(8+2 Np)t=(8+2× 3)×0.022=0.308, say 0.310in
duct is 1925 psi. Substitute all these into equa-

tion (9-24) for a three-ply bellows: Free axial length of the bellows:
La = (Nc - 0.5)L + Npt
= (22 - 0.5) × 0.310+ 3 × 0.022
=6.66+0.066=6.726 in (i.e., lessthan 7in)

h 2N/_-_ 4/2×3x150000 18.5
-CTT :Ip 1.36x1925 From figure 9-22, the correction factor C t for
the axial spring rate at 20 percent thinning is
For a reasonable value of h, in a bellows of this
0.72. Substitute this into equation (9-22) to
size, we arrive at a wall thickness t of 0.022 in
obtain the axial spring rate of the bellows:
(after several tries):
h=18.5×0.022=0.407 in, say 0.410 in 1.49 CtC pN pEdit 3

Ra= Ncha
The bellows root-mean-square diameter and
the equivalent axial deflection due to 3 ° angula- _ 1.49 × 0.72 × 0.85 × 3 × 29.6 × 106 × 8 × (0.022) 3
tion can now be found: 22 × (0.410) 3
do=di+2h=8+2×0.410=8.820 in = 4550 lb/in

From equation (9-42), the critical internal p(dm 2 - dd2)r_

stability pressure for bellows without angulation: Fp= 4
5.02×Ra 502×4550 The required restraining link load at maximum

Per= do working pressure, considering the normal axial
La 6.726 force
Comparing this with the maximum working pdd 2tr pdm 2rr_ 1750× (8.42) 2× ,7
pressure of 1750 psi, a safety factor of F=FP+ 4 - 4 4
3080/1750= 1.76 remains to allow for bellows
= 97 500 Ib
stability under conditions of angulation.
From sample calculation (9-2), the yield
pressure =2117 psi, and the ultimate pressure 9.6 DESIGN OF GIMBAL MOUNTS
= 2887 psi. Substitute these into equation (9-26),
to obtain the yield hoop stress of tile bellows: Gimbal Design Considerations
Primary design considerations for gimbal

mounts are:
Sh =-2 _pt I. Required pivotal movement o[ engine as-
sembly or thrust chamber for thrust vector con-
_ 22117
x 3 xx0.022
8.42 × I0.57 + (2
k_ x 0"41)1J trol.-Geuerally ranging from ±4 ° to +__10 °.
2. Required adjustment for thrust alinement
= 42 000 psi (less than 170000 psi)
and positioning.
3. Thrust level.-This determines the struc-
Tile ultimate hoop stress of the bellows: tural and bearing design of the gimbal mounts.
4. Required operational li[e.-Generally 1000
42 000×2887 =57 300 psi (less than 200 000 psi) cycles minimum.
Sh = 2117 5. Minimum deformation of the bearing sur-
face, prevention of bearing surface galling.
6. Propellant duct installation.-Some designs
The bellows design configuration is now es- require flow of one of the propellants through the
tablished: center of the gimbal mount.
7. Maintenance o[ the gimbal rnount.-This is
di =8 in d o =8.82 in dm =8.42 in largely affected by the lubrication requirement of
t =0.022 in Np=3 h =0.41 in the bearing surfaces. If possible, a gimbal boot
Nc=22 L =0.31 in La=6.726in should be provided to protect the bearing sur-
R a = 4550 lb/in faces from dust, water, and foreign materials.
8. Lightweight.-High strength-to-weight
From equation (9-29), the angular spring rate materials should be used.
of the bellows
Ring-Type Gimbal Mounts

dm2Ra _ (8.42) _ x 4550_ 705 in-lb/degree
Rb = 458.4 458.4 Figure 9-23 presents a typical gimbal mount
designed for low-thrust, upper stage engine ap-
The bending moment on the duct at 3 ° angula- plications up to 20 000 pounds. The design is a
tion closed-yoke, flowthrough-ring-type gimbal, utiliz-
ing plain bearing pivot joints. This configura-
M = RbO = 705 × 3 = 2115 in-lb tion provides for the main oxidizer duct to pass
through the assembly in the longitudinal axis to
From equation (9-25), the bellows pressure the thrust chamber dome. The gimbal mount is
designed to connect the vehicle main oxidizer
separating load
.'
THRUST CHAMBER
Figure 9-24.-Typical cross-type gimbaI mount

designed for medium-thrust engine applications.
incorporating bearing surfaces, upper and lower

gimbal bearing blocks, upper and lower retainers,
and thrust vector alining slides. All parts of
this design are made of 4340 alloy steel. The
bearing surfaces of the cross unit are chromium
or electroless nickel plated. The bearing sur-
Figure 9-23.-Typical ring-type gimbal mount de-
races are phosphate treated. Solid-film dry-
signed [or low-thrust, upper stage engine
lubricant coatings are applied to all bearing
applications.
surfaces. The design bearing pressure of this
type gimbal mount ranges from 15 000 to 20 000
duct and the top of the thrust chamber dome. The psi of bearing projected area. The bearing coef-
lower support of the gimbal mount provides an ficient of friction varies between 0.06 and 0.1.
adjustment mechanism for thrust vector aline-
ment. Except for the steel pivot pins and the
alinement bolts, all other parts can be made of
either aluminum or titanium alloys for minimum
weight. The bearing surfaces of the pivot pins
should be chromium or electroless nickel plated.
Solid film dry lubricant coatings or grease may
be applied to all bearing surfaces. The design
bearing pressure of this gimbal type is about
10000 psi of bearing projected area. The bear-
ing coefficient of friction is around 0.08.
Cross-Type Gimbal Mounts
Figure 9-24 presents the design of a typical

cross-type gimbal mount used on medium thrust
engines (up to 200000 Ib), such as shown in fig-
ure 9-I. Here, the gimbal mount secures the
engine assembly to the vehicle thrust frame, and
is mounted to thrust chamber dome and elbow Figure 9-25.-Typical socket-type gimbal mount
assembly. It consists of a cross-shaped unit designed tot high-thrust engine applications.
Socket-Type Gimbal Mounts position coating, which provides a dry, low-

Figure 9-25 presents a typical heavy-duty friction bearing surface.
gimbal mount designed for high-thrust engine The bearing surface on the sphere is electro-
applications (up to several million pounds). This less nickel plated. No lubrication or maintenance
mount is also designed to secure the engine is required. For these gimbal mounts, design
assembly to the vehicle thrust structure. The bearing pressures range from 24 000 to 2S 000
unit consists of integral upper and lower bearing pounds per square inch of bearing projected area.
blocks, and their retaining bar. All parts are The bearing coefficient of friction is about 0.15.
made of alloy steel. The thrust load is carried Lateral adjustments for thrust alinement are
by a socket-type spherical bearing located be- accomplished using a tongue and groove arrange-
tween two bearing blocks. The surface of the ment between gimbal lower bearing block and
spherical socket has a Teflon-fiber-glass corn- thrust chamber dome.
Chapter X
Engine Systen s Design Integration
10.1 SYSTEMS ENGINEERING superiors he needs adequate authority, commen-

In the foregoing chapters the design of the surate with his responsibilities. He must have a
major liquid rocket engine subsystems and their broad and thorough understanding of the various
components has been discussed. By themselves, disciplines governing the design and develop-
these subsystems cannot fulfill a useful purpose. ment of the subsystems. In view of the latter's
Only when integrated into a whole will they func- high specialization, he cannot be expected to
tion as a system and produce for what they were have a complete education in all these disci-
designed: thrust. While the engine designer plines. However, in the words of Du Pont's
thus must view the complete rocket engine as the Crawford H. Greenewalt, he must have "the abil-
system, he should not lose sight of the fact that, ity to create a harmonious whole out of what the
to the stage builder, the engine is again a sub- academic world calls dissimilar disciplines,"
system. (In fact, to the launch vehicle systems and that in this respect his job may be compared
engineer, the stage in turn becomes a subsystem, to that of a "symphony conductor under whose
in addition to other stages, ground-support equip- hand a hundred or so highly specialized and very
merit, launch facilities, downrange stations, etc.) different talents become a single effort of great
In an earlier chapter the major engine param- effectiveness."
eters requiring optimization during implementa- The job of systems integration does not
tion of rocket engine design were defined. It is merely assure that all parts physically fit to-
apparent that the subsystems discussed will not gether, although this certainly is a first basic
form an optimized complete system unless they requirement. Wherever parts and subsystems
are designed for one another witil this goal in join, an interface exists, a demarcation line be-
mind from the beginning. The subsystems pre- tween these systems. This may be a bolt hole
sented in the foregoing chapters have become pattern; a flow of fluid, heat, or electrical cur-
highly specialized fields during recent years. rent; an exchange of forces, loads, or torques;
Their designers and developers often have only or a complex dynamic interaction.
a rather general feeling for the functions and The importance of systems engineering is
peculiarities of the other subsystems. Their being recognized to an increasing degree. Figure
concentrated effort will not come about unless it 10-1 shows the final configuration of the engine
is guided by an important function: systems for an early postwar ballistic missile. Compar-
engineering. ing this to figure 2-4 we realize the progress
In most rocket engine projects this function made through good systems engineering. How-
is assigned to the "systems engineer," commonly ever, to have designed and developed a good-
called project engineer or project manager. In looking, perfectly functioning rocket engine is
conformance with user requirements, he has the not enough.
job of establishing all ground rules and perform- It may appear superfluous to state that the
ance parameters (ch. II) and to subsequently design of rocket engines must also consider
optimize them through complete integration of all vehicle application. Yet in the daily grind of
subsystems during the entire design and devel- problem solving, engine detail design and devel-
opment period. opment, this is easily forgotten. The rocket
Because of the previously discussed interac- engine and component designer and developer
tion of funding, time, and reliability, the project who can think and converse in terms of vehicle
engineer must be a good manager, too, in addi- application will be sought after indeed, and so
tion to his engineering qualification. From his will be the company he represents.
383
We can easily realize the importance of con-

tinuous and close communication between engine
designer and vehicle builder. Complete and free
exchange of correspondence, documentation and
progress reports, frequent design reviews, and
above all, prompt notification of changes is
vital. During engine development, closest simu-
lation of flight vehicle configuration, when per-
forming static engine development tests, will
materially reduce the possibility of later "sur-
prises." Each vehicle type, according to its
mission, will have somewhat different interface
problems. Many of these occur, however, in
almost any vehicle type.
10.2 ENGINE SYSTEM DESIGN INTEGRATION

BY DYNAMIC ANALYSES
The Scope of Dynamic Analyses
Dynamic analyses are essential for optimum

engine systems design integration, commensurate
with vehicle mission requirements. Dynamic
analyses may be grouped into two basic areas:
1. lnternai operating dynamics of the engine
system.-This refers to system schematic and
control optimization, component optimization
from a functional as well as system transient
point of view, optimization of system start tran-
sients, minimization of cutoff impulse, determi-
nation of cutoff surges, etc.
2. Engine-vehicle operating dynamics.-This
refers to vehicle tank pressurization systems
design for adequate transient and steady-state
engine performance, engine-vehicle structure and
engine-guidance operating compatibility and
Figure IO-I.-Ballistic missile liquid rocket stability, overall vehicle performance during
engine showing greatly improved simplicity. special maneuvers, etc.
General Approach to the Analyses

Rocket engine design for vehicle application
is a consideration not only for vehicle flight or The techniques and equations used, and the
for optimum mission performance. Long before general approach taken by industry and Govern-
the engine will fly, and after it has been de- ment agencies toward the solution of the various
signed and developed, it must be installed. The dynamic problems in rocket engine systems, are
facets of engine installation in a vehicle range the result of many years of effort and experience
from more complex areas, such as guidance loop in the areas of analysis, synthesis, and correla-
compliance and prestart conditioning to simple tion of rocket engine operation. The philosophy
"nuts and bolts" problems, such as matching governing this approach postulates that the rele-
connecting fitting dimensions and attachment vant characteristics of any system depend on the
bolt sizes. characteristics of their components and physical
dlmmma
ENGINE SYSTEMS DESIGN INTEGRATION 385
processes. By describing these components and Some physical processes, such as thrust
processes in detail, as well as their interaction, chamber combustion dynamics, are not always
the system can be described analytically with as quantitatively fully understood. Rate of combus-
much detail as is necessary. The complete set tion is known to be a function of pressure, pro-
of equations then represents a mathematical pellant type, mixture ratio, and combustion
model of tile engine system. chamber geometry, but a specific quantitative
Through the solution of the equations repre- expression for reliable use with rocket engine
senting the mathematical model, the important combustion chambers is not available. Certain
characteristics of an engine system are studied, system-start-transient analyses made with the
problem areas are defined, and improved compo- aid of an engine model (set of equations like eq.
nent designs for the solution of these problems (10-2)) have ideally assumed instantaneous com-
may be evaluated. Also, transient and steady- bustion. Thus, combustion instability or thrust-
state engine systems operation, as affected by chamber-feed-line, system-coupled instability is
various component characteristics, may be simu- not described: This deficiency can be rectified
lated and checked by these dynamic analyses. by experimental systems evaluation.
If necessary, modifications will be incorporated Dynamic analyses can also be effectively
prior to hardware testing. used during the development-redesign phase of
Dynamic analyses, however, have their limita- an engine system. Once test information is
tions, because not all of the physical processes available, the predicted characteristics (with
involved in a given rocket engine system are idealized assumptions) and the actual system
immediately and/or thoroughly understood. As operating characteristics can be compared. Dif-
each of the processes becomes better defined ferences can be noted and evaluated. This test-
functionally and quantitatively, confidence in the analysis cycle defines the limits of component
mathematical analyses increases. performance and thus serves as a guide for the
Let us look at an example. The hydraulic redesign of the components to be integrated into
head developed by a propellant pump is, as we an optimum, final engine system. Similarly, from
know, a function of pump speed, flow rate, and analyses of the engine-vehicle operating dynam-
geometry. We can write: ics, and in conjunction with test results, the
engine system and its components can be modi-
H=[(N, Q, r, A) (10-1) fied and improved.
where
H = pump developed head, ft
Criteria for the Mathematical Model of an Engine
N = pump speed, rpm
System
Q = pump flow rate, gpm
t = pump impeller radius, in The mathematical model of an engine system
A=area normal to the meridional flow, in 2 generally consists of a group of lumped param-
Correlations, such as equation (10-1), can be eters, and of linear or nonlinear algebraic and
used to determine the interdependence of the differential equations, which are formulated and
many processes within an engine system. then programed for an analog or digital computer.
Furthermore, equation (10-1) may be expressed Careful examination of an engine system sche-
as a specific form of function, as shown by equa- matic will be sufficient to determine whether a
tion (10-2), which is valid for a particular pump mathematical model will be possible for the sys-
design only. tem. This simply amounts to an observation of
the many significant physical processes involved
H = aN _ + bNQ + cQ (10-2) in the entire system which may be expressed
mathematically. Some idealized assumptions are
When the numerical values of a, b, c are usually required to obtain a quantitative expres-
known, equation (10-2) becomes a quantitative sion of the various equations. The physical
description of a given pump design and a means significance of these assumptions must be under-
to obtain the numerical solution of the operation stood before the mathematical model can become
of an engine system. meaningful.

There are many ways to describe a rocket K (10-4)

engine system mathematically. The choice will a=2gpA2Ra2
be based on the answers expected from the
model. A model used to predict engine systems
orificing requirements may call for high steady- b=_g (10-5)
where
state accuracy over a stated operating range,
Pl =pressure at the inlet, psi
while the dynamic characteristics are of little
P2 =pressure at the outlet, psi
consequence. A model used for the design inte-
a =design factor for turbulent flow pressure
gration of control components and subsystems
drop, sec2/lb-in 2
may require good dynamic accuracy for inter-
b = design factor for fluid inertia to changes
mediate frequencies (0.01-100 cycles/sec)and
in flow rate, sec2/in 2
good static accuracy. A model needed to de-
_z = weight flow rate, lb/sec
scribe high-frequency effects (100-10000 cycles/
it = change in weight flow rate, lb/sec 2
sec) may require little in the way of static accu-
h1_2 = gravity head of the fluid in the duct, in.
racy, since the phenomena under study occur so
If the system is subject to an accelera-
rapidly that the system as a whole has no time
tion other than I g, this term must be
to shift to new operating levels.
modified
The mathematical model most frequently used
p = fluid density, lb/in s
is of the type employed for system main-stage
K = duct or component resistance coefficient,
operation and control design. This usually
to be obtained from flow tests
serves as the basic mathematical model of an
A = flow area of the duct, or flow area of the
engine system, with some modifications incorpo-
component at the design operating
rated for other, special applications. In general,
point, in 2
all mathematical descriptions of a basic model
Ra = component area ratio=component flow
are for conditions around the systems design
area/A (at the design operating point,
point, with the following assumptions:
Ra = 1)
(1) All liquid propellant flows are incompres-
L :length of the duct or component, in
sible and at constant temperature
g =gravitational constant, 386.4 in/sec 2
(2) All gas equations are based on perfect
2. Combustion process and operating dy-
gas performance, where the gas proper-
namics:
ties are functions of composition
Refer to figure 10-2, which describes sche-
(3) All vehicle-supplied parameters, such as
matically the combustion process. A transporta-
main propellant tank discharge pres-
tion time delay is assumed, representing the time
sures, are constant.
required for the propellants to pass through the
injector and enter into the combustion process,
and expressed as
Examples of Equations for a Mathematical
Engine Model poAo( Lo + Lo')
(10-6)
r° = g'o
The mathematical expressions and equations
for the various physical processes and operating
p[At(L[+ LI')
dynamics of a rocket engine may be derived from (10-7)
equations given in chapter I and in other chap- r[= Wf
ters for the design of the various components.
Here, we will present several typical examples
to illustrate the application of these equations
, _ -, w o
in a mathematical model.
,.=: AI
1. Pressure drop of fluid flow in a duct or

component:
Figure I O-2.-Schematic description of the com-
P, - P2 =aw2 +bit-Ph,-2 (10-3) bustion process.
These can usually be indicated by the Laplace Vc =volume of combustion chamber from
transformation operators p-ros and p-rfs. If we injector to throat, in 3
assume a homogeneous combustion gas, we can R' = universal gas constant, 18 528 in-lb,/°R
define the following correlations: mole
_Ii =molecular weight of the combustion
gas, lb/mole
pc=R,(\T/kV3
(wc (lo-8) Wc = weight of the gas stored in the com-
bustion volume, lb
Wc = wo + wt (10-9) g'c = flow rate of the gas emerging from the

combustion chamber, lb/sec
Wo
Ro =weight fraction of oxidizer stored in
Ro = (w ° + wt) (10-10) the combustion chamber
At =throat area, in 2
C* = ctiaracteristic velocity, ft/sec
wo=_o+ (p-r°Sfvo-Rofvc)dt (10.11)
These equations for the combustion process can
be applied to engine main thrust chambers, gas
t generators, and other types of combustor.
3. Turbopump operating dynamics:

wf= wf÷ fo [P-rfSwl-(1 - R°)fVc]dt(lO'12)
P2 - Pl =aNp 2_y (10-16)

A tPcg
wc- C* (10-13)
4 rr2rp2p
a= (10.17)
g
T¢
=t(Ro) (10.14) ¢ = [(4,) (10-1s)
C*=t(Ro) (10-15) 4,¢

x= (10-19)
r/p
= time required for oxidizer and fuel to
pass through the injector and enter
Tp = bNp2x (10-20)
into the combustion process, sec
Po, Pt = density of oxidizer and fuel, lb/in 3
Ao, At =injector area, oxidizer and fuel, in 2 b = 8 z73rp3App (10-21)
Lo, Lt = travel distance of injected oxidizer g
and fuel prior to impingement, in

Lo', Lt' = travel distance of impinged oxidizer Nt = RgNp (10-22)
and fuel prior to combustion, in
t = time period in consideration, sec 1485 _i'tAHot
Tt - (10-23)
¢¢o. fvf =injected flow rate of oxidizer and Nt
fuel, lb/sec
iI_o, _f = weight of oxidizer and fuel stored in
the combustion chamber volume, lb,
(10.24)
at the beginning of time period t AH=CpT° - -_t
Wo, wt = weight of oxidizer and fuel stored in
the combustion chamber volume, lb, lNp = f (RgTt - Tp)dt (10-25)
at the end of time period t
s = complex operator in a Laplace trans- Equation (10-25) is a torque-balanced equa-
formation tion for turbopumps. Any unbalanced torque be-
Pc = pressure of the combustion gas, psia tween turbine and pump will initiate a change of
Tc = temperature of the combustion gas, °R shaft speed causing the integrand to seek zero.
where This includes pressure drops in various

p, =pump inlet pressure, psia flow passages, turbopump operating per-
P2 =pump discharge pressure, psia formance, and thrust chamber heat trans-
a =pump pressure rise design factor, lb- fer and combustion characteristics
sec2/in 2 (3) Optimization of engine system steady-
b =pump torque design factor, in-lb-sec 2 state operation and performance, by

properly calibrating and matching tile
rp =pump impeller radius, in
Ap =pump flow area, in 2 design operating points of various com-
p =density of the pumped fluid, lb/in 3 ponents. (See sec. 10.3.)
Np = pump shaft speed, rev/sec (4) Determination of engine system mainstage

Nt =turbine shaft speed, rev/sec performance characteristics, including
Rg = speed ratio of the turbopump gear train perforalance variations and engine int2u-
ence coefficients. (See see. 10.4.)
Tp =torque at pump shaft, in-lb
Tt = torque at turbine shaft, in-lb (5) Evaluation of various engine control prob-
_Pt =turbine gas flow rate, lb/sec lems during main-stage operation, such
AH = available energy content of the turbine as thrust and mixture ratio controls
gas, Btu/lb (6) Evaluation of various potential perturba-
Cp = turbine gas specific heat at constant pres- tions and their effects on mainstage
sure, Btu/lb-°F operation
To =turbine gas total temperature at inlet, °R Once the basic mathematical model for the
y =turbine gas specific heat ratio main-stage operation of an engine system is
Rt =turbine pressure ratio established, it can be utilized to study special
I =inertia of the gear train, referred to the problems with additional inputs. Vor example, a
main pump shaft, in-lb-sec 2 basic mathematical description of the 150 000-
_, =nondimensional pump head coefficient pound-thrust LOz/RP-1 pump-fed rocket engine
¢ =nondimensional pump flow coefficient for an intermediate-range ballistic missile had

x =nondimensional pump torque coefficient been established on an analog computer. Fol-
r/p = pump overall efficiency lowing design of an additional engine-thrust-
r/t =turbine overall efficiency control subsystem, its electronics, main valves,
and servovalve drive system were tied to an
analog computer by suitable transducers to allow
Dynamic Analysis of Engine System Mainstage transient performance checkout and controller
Operation gain adjustment. An updated mathematical en-
gine model, including the nonlinear perturbation,
In general, engine design requirements at the
was then used for more detailed investigations
mainstage level, as well as initial component
of the thrust-control-loop dynamics.
and system specifications, can be determined
with the aid of a mathematical model consisting
of linearized descriptions of the complete engine
Dynamic Analysis of Engine System Start and
system and a computer. Based on the static
Cutoff Transients
operating values at main stage (such as given in
tables 3-2 to 3-5), tile static design factors (such The main objectives of dynamic analyses of
as a and b in eqs. 10-2, 10-3, 10-16, and 10-22) engine system start transients are:
may be obtained. The primary dynamic analysis (1) Investigation of the systems schematic
objectives for mainstage operation are: for needed start-transient controls, snch
as type and quantity of control compo-
(1) Evaluation of the engine system sche- nents, and sequencing and timing of
matic, with respect to mainstage opera- their operation
tion (2) Determination of thrust chamber ignition
(2) Evaluation of the dynamic characteristics, requirements
interactions, and the performance of (3) Estimation of start energy, time, and
various components at mainstage level. thrust buildup characteristics
(4) Evaluation of comp¢,uent dynamic charac-

, FUEL PUMP OISCHARGE-_%,
teristics and interactions during start ,ooo, _ _i ]
i rI 7-
I "--_--__,,....
! .._,w,
.,,- "- j iVALVE
900 " _,-- - b -_--_ ..... ,..... T " -_
transients, such as combustion chamber
I Ga,S i
I i
' ! f ;
, , RO$.',O,,
ignition delays, gas generator tempera- [ \ VALVE I i ' / "
800 _---_- f--_. --- "_--_ --_- ....... ; F._ --./_ .... -OPEN
ture surges, and propellant pump stalling r "{l I / ' MA,N ////"_"--OX,D,ZER
' FUEL , , PUMP I
(5) Evaluation of system dynamic stability
7°°!1/ r /M_IN [ I / i II .....
during the start transient. (The aim is I ,/ '/OXiDiZe'..
I .....
I/ I t/ GASGENERATOR
0 600 --_--'_-" ..... _ t1 --- PPESSbRE-'_ . -CLOSED
to avoid prolonged operation at levels
exhibiting system or thrust chamber
instability.)
IC-NITER FUEL IGNITION--,( ,/ _,_"/
(6) Evaluation of various potential perturba-
tions and their effects on start transients,
such as a start where the propellant-
settling effects of gravitation are absent r PRPRESSURE
200 --
For some engine systems, such as the
LO2/LH2 turbopump-feed A-2 stage engine, IO0
dynamic analyses of its start transient become

rather complex. They may include effects such °o_.l 02 0.3 0.4 o._ 0.6 o7 o8 09 _o
as water hammer (wave equation) in the propel- TIME FROM START SIGNAL, SECONDS
lant feed systems, distribution of heat transfers

Figure lO-3.-Graphic presentation of the .start
and pressure drops throughout the high-pressure
fuel feed system, choking of hydrogen gas in the transient model for a typical turbopump-feed
engine system utilizing a gas generator for
chamber coolant passages, stall characteristics
turbine drive.
of the fuel pump, cavitation at the pump inlets,
changes in fuel density in the pump caused by
enthalpy changes from pumping, and many others. (3) Evaluation of possible temperature surges
Because of this complexity, me equations of in main chamber or gas generator during
these mathematical models are usually programed the cutoff transient
for a digital computer. (4) Evaluation and optimization of total cutoff
Figure 10-3 presents graphically the start impulse, by minimizing cutoff time and
transient model of a typical turbopump-feed en- improving repeatability
gine system utilizing a gas generator for turbine (5) Evaluation of engine thrust decay charac-
drive. Valve timing, as well as pressure buildup teristics
characteristics of gas generator, propellant A series of cutoff sequences with modified
pumps, and main chamber are indicated. Other shutoff timing of the control valves is usually
parameters, such as gas-generator combustion- simulated with the mathematical model. The
gas temperature and propellant flow rates, can various simulated cutoff runs are then analyzed
be included in the model. to determine potential problem areas, and whether
Several alternate engine-start methods are these problems are a function of the particular
usually simulated with the start model, in order sequence used.
to evaluate potential problem areas and to opti- Figure 10-4 presents graphically the cutoff
mize systems start transient operation. transient model of the typical engine of figure
Important objectives of dynamic analyses of 10-3. The engine-thrust-decay characteristics
engine cutoff transients are: are represented by the main chamber pressure
(1) Investigation of the systems schematic decay curve. The integrated area under the
for needed cutoff transient controls, chamber pressure versus time curve may be used
including operational sequencing and to assess the engine cutoff impulse.
timing of various control valves
(2) Evaluation of pressure surges and other
Dynamic Analysis of Engine-Vehicle Interactions
adverse effects in propellant ducts and
feed system components during the cut- These analyses may be performed during the
off transient initialdesign phase of an engine system, as
I000 10.3 DESIGN INTEGRATION FOR ENGINE

I ! ; r I I
SYSTEM CALIBRATION
900 ---
- i • i,.___VAI
|POSITION
......... ' E
"--FUEL PUMP 3_SCHARGE I
i I
I
i
[
z
Design Requirements for the Calibration of an
\1_- [ i' _ .... r T OPE N
_ i b I ",. I J Engine System
_'_ /-_ GAS GENERATOR _ ././././_MAINFUEL
L,_"F VALVE! i x',_i'l VALVE 1 Because of unavoidable mechanical toler-
700
ances, it may be expected that the operating

I' _'_' VALVE I I _ /
L_
--_.- _; - ._ -_::-CLOSEI_ characteristics and performance of the various
6oo-d
II I ',I I ! , engine system components will deviate some-
Ld what from their nominal design value. A certain
E 500----
03 \ \ _ OXiDiZER PUMP DISCHARGE, amount of calibration is always required for

03
hi
at"
r_
400 -- -"L-/I I I_\_ :_ GAS GENERATOR
I L
CHAMBER
these components, as well as the engine system
as a whole, to attain the desired engine perform-
l__, / PRESSURE I [ I !1
ance characteristics within design specification.
[ _ \_lr"_" MAIN CHAMBER PRESSURE [ Therefore, provisions must be made in component
and systems design to permit effective calibra-
tion during system integration.
The specific impulse Is of an engine system
is the ratio of thrust F to propellant weight flow
o i rate &. Thus, any deviations affecting F or &
0 0._ o.z 0.3 0.4 0.5 0.6 o.-t o.e 0.9 Lo
TIME FROM CUTOFF SIGNAL, SECONDS will affect system performance. I s also is a
function of propellant mixture ratio. It is desir-
Figure lO-4.-GrapMc presentationof the cutoff able, therefore, and beneficial to calibrate an
transient model of the typical engine of figure engine system by adjusting its propellant feed
10-3.
system. Prior to complete engine system cali-
bration, the pressure or pressure drop versus
well as following development test firings. flow characteristics of each individual compo-
A_eas of general interest to be analyzed may nent should be calibrated and evaluated.
include, but should not be limited to- Hydraulic and pneumatic components, such as
(l) Engine system operation and performance pressure and flow regulators, valves, flowraeters,
requirements from a vehicle mission ducts, and lines, can all be readily calibrated on
point of view flow benches. However, those components which
(o) Matching of engine propellant supply re- operate at temperature extremes, such as thrust
quirements with the vehicle propellant chamber assemblies, gas generators, and turbo-
system, including dynamic evaluation of pumps, are best calibrated by combining the flow
the vehicle propellant tank pressuriza- tests with actual hot firings. The characteristic
tion system, PU control system, vehicle propellant flow curve of an engine system is
acceleration and sloshing effects, and obtained by summing the pressure or pressure
feed system-combustion coupled in- drop versus flow curves of the various compo-
stabilities nents (figs. 10-5 and 10-6).
(3) Matching of the engine controls with the The general design approaches toward cali-
vehicle guidance system, including re- brating an engine system to attain its design
sponse of engin.o start, cutoff, thrust thrust at design mixture ratio are:
level, and vector controls to vehicle
guidance commands (1) The design operating point of each com-
(4) Simulation of interaction between engine ponent should be kept within the rela-
systems operation and vehicle dynamics. tively flat region of its pressure or
(This may involve closed-loop coupling pressure drop versus flow curve.
of an analog simulation of vehicle guid- (2) The mechanical tolerances and built-in
ance and trajectory characteristics with adjustments of each component should
an engine system during hot firing) be designed so that the random deviation
_____._.__._>
_MINIMUM REQUIRED TANK PRESSURE VS r_.OW

_CURVE OR CALIBRATED SYSTEM RESIST/kNCE
REGULATOR OUTLET PRESSURE SYSTEM _r _ CURVE
OE SIGN POINT--\ /
\ / / _ SYSTEM DESIGN PRESSURE VS FLOW CURVE
/ _ OR SYSTEM RESISTANCE CURVE WITHOUT
PRESSURANT LINE PRESSURE DROP_._ _'x= _
CALIBRATION ORIFICE "_1 -- _=/-_/ _ ORIFICE
DESIGN PRESSURE DROP-_._ /I __--*_---'-_PRESSURANT L_NE PRESSURE VS FLOW CURVE
DESIGN PRESSURE DROPS
/ _ {AT ENTRANCE TO TANK)
iN VALVES AND LINES_ _
// _ / _ DESIGN CHAMBER INLET PRESSURE VS FLOW CURVE
DESIGN PRESSURE DROP !N THE // /i /-_OESIGN CHAMBER INJECTOR END PRESSURE VS
CHAMBER ( MANIFOLD COOLING_ /f _ ./ PLOW OURVE
PASSAGE AND INJECTOR -_
DESIGN CHAMBER
_N,JECTOR END PRESSURE J
PROPELLANT FLOW, LB/SEC
Figure 10-5.-Propellant flow design characteristics of a typical pressure feed engine system
(oxidizer or fuel).
v--CALIBRATED FUEL SYSTEM

\RESISTANCE CURVE
x--FUEL SYSTEM RE, STANCE CURVE WITHOUT ORIFICE

_. _- OXIDIZER SYSTEM RESISTANCE
\CURVE WITHOUT ORIFICE
---.... ,,,
,---FUEL PUMP PRESSUI_:
CALIBRATION
"_, _ VS FLOW CURVE AT NZ
ORIFICE DESIGN
PRESSURE DROP [ /--FUEL PUMP PRESSURE
t
Pfd
° °J
° " OX,Q, ZER
PUMP
PRESSURE
i i
FUEL FLOW LB/SEC
OXIDIZER FLOW LB/SEC
Wfd wfo Wfb %a Wod
FUEL SYSTEM FLOW DESIGN CHARACTERISTICS OXIDIZER SYSTE},IFLOW DESIGN CHARACTERISTICS
Figure IO-E-Propellant flow design characteristics of the A-I stage turbopump feed engine system.
of its flow characteristics from its de- ratio, and Is (as verified by actual thrust cham-
sign value will be kept within a reason- ber test firings). Based on these flow rates, the
able limit, in order to facilitate systems pressure drops of the various components at the
calibration, and to keep other system design operating point can be estimated from
components in their design operating previous design data, or as obtained from actual
region. testing. Certain components may have to be
(3) Sufficient pressure head should be set newly designed for the specific design pressure
aside in each engine propellant feed drops allowed by the system.
system to compensate for contingencies The design pressure versus flow curve of
due to component flow resistance devia- either propellant flow system can be obtained by
tions. The propellant feed system can the summation of design chamber pressure versus
then be calibrated by means of orifices flow and component design pressure drop charac-
or other adjusting means. teristics, as shown in figure 10-5. In addition,
an orifice is introduced in each propellant flow
system for calibration. The minimum required
Design for Calibration of a Pressure Feed System
tank _ressure versus flow curve for each propel-
The first design step is the determination of lant is thus derived. In most pressure feed sys-
the design flow rate of each propellant, as calcu- tems, the design orifice pressure drop for
lated from rated systems thrust, design mixture systems calibration determines the maximum
..... _ ----_:::.:=?:
allowable cumulative pressure drop increase of feed system are similar to those of a pressure
the components above their nominal values. A feed system. However, the difference in turbo-
suitable tank pressurization system can then be pump pressure or head versus flow characteris-
designed, compatible with minimum required tank tics from those of a pressurized system dictates
pressure versus flow characteristics. a somewhat different approach to systems cali-
bration. For mechanically coupled turbopump
Sam ple Calcu la tion (I0-I)
feed systems, such as the A-1 stage engine,
The following data are available from analyses systems calibration generally involves adjust-
and component tests for the A-4 stage propulsion ment of the turbopump speed as well as the in-
system, at rated thrust conditions: stallation of an orifice in one of the propellant
Thrust chamber injector end pressure range lines. For turbopump feed systems with dual
required to maintain rated thrust = 110 ¢ 3 psia turbine drive, such as the A-2 stage engine, the
Thrust chamber injector pressure drop range calibration can be accomplished by adjusting the
(both oxidizer and fuel) = 25 ¢ 2 psi speeds of both turbopumps.
Thrust chamber oxidizer dome pressure drop The design principles for the calibration of
=3_ + 1 psi mechanically coupled turbopump feed engine
Oxidizer line pressure drop = 5-+ 1 psi systems are best illustrated by a typical exam-
Main oxidizer valve pressure drop (at the fully ple, as shown in figure 10-6. Here, the propel-
open position) = 4 ¢ 1 psi lant system resistance curves without orifices
Thrust chamber fuel manifold pressure drop (representing conditions downstream of the pump
=4_+1 psi discharges) are constructed based on the designs
Fuel line pressure drop = 4 ¢ 1 psi and test results of the components for the A-1
Main fuel valve pressure drop (at the fully stage engine system. Next, the discharge pres-
open position) = 4 ¢ 1 psi sure versus flow curves of both pumps are con-
Pressure allowance required for mixture ratio structed from test data obtained with the A-1
control by oxidizer valve vernier positioning stage engine turbopump, operated at speed N1.
(fig. 7-4)=¢ 10 psi These pump curves intersect the corresponding
Determine the design pressure drops of the cali- system resistance curves at point A. At this
bration orifices, and the minimum required tank speed, fuel flow rate _/fa is above, and oxidizer
pressures for design flow rates. flow rate _#oa is below the required design flow
rates, Wfd and _i,od.
Solution
To achieve the design oxidizer pump flow
The design pressure drop of a calibration _Pod, at a desired discharge pressure Pod, the
orifice must be equal to the sum of the maximum design operating speed of the turbopump assem-
pressure drop increases of components above bly mr, st be raised to a required level N 2 by in-
their design values. Thus: creasing the turbine gas flow. However, at this
The design pressure drop of the oxidizer cali- speed, the fuel pump, which is mounted on the
bration orifice = 3 + 2 + 1 + 1 + 1 = 8 psi. same shaft as the oxidizer pump, would be de-
The minimum required oxidizer tank pressure livering a flow rate d'fb considerably above the
at the design flow rate = 110 + 25 + 3 + 5 + 4 + 8 + 10 required design flow rate 1//fd (point B in fig.
= 165 psia. 10-6). To reduce the fuel flow to _fd, a calibra-
The design pressure drop of fuel calibration tion orifice is placed in the fuel line. This
orifice=3+2+l + 1 +1 =8 psi. amounts to increasing the fuel pump discharge
The minimum required fuel tank pressure at the pressure at constant speed hr2 to Pfc, where _'fd
design flow rate = 110+ 25 + 4 + 4 + 4 + 8 = 155 psia. is reached at point C. The pressure drop across
the calibrating orifice is represented by Pfc-Pfd,
where Pfd is the desired fuel pressure.
Design for Calibration of a Turbopump Feed
If fuel flow rate _/fa is below and oxidizer
System
flow rate _/o._ is above the required design flow
The propellant flow characteristics down- rates, the calibrating process would be to speed
stream of the pump discharges of a turbopump up the turbopump to obtain the desired fuel flow,
ENGINE SYSTEMSDESIGN INTEGRATION 393
and to place an orifice in the oxidizer line. Determine the location of the calibration orifice,
However, it is generally desirable to place the its nominal design pressure drop, and its ex-
orifice in the system of the propellant with the pected range of adjustment.
higher boiling point. In this situation, therefore,
and also when the pressure drop across a cali- Solution(see sample calculation(6-2))
brating orifice tends to become excessive, it is
customary to trim the pump impeller so as to The required oxidizer pressure head at the
reduce the effective speed, and thus attain the design point = 1095 + 200+ 150 + 25 + 35 = 1505
required flow and pressure levels. In view of psia.
pump efficiency effects, it is desirable to trim The required fuel pressure head at the design
the pump drawing the smaller horsepower, usually point = 1095 + 200 + 290 + 10 + 15 = 1610 psia.
the one with the lower mass flow rate, except in Since the LOX pump discharge pressure is
cases of extreme density differences. The ad- 1505 psia, but the fuel pump discharge pressure
justment of the turbine gas flow rate, and thus is 1720 psia, the calibration orifice must be
the turbopump operating speed, can also be made located in the fuel system.
by means of orifices in the turbine inlet gas line, The nominal orifice design pressure drop
or in the gas generator propellant lines. = 1720- 1610= 110 psi.
In general, turbopump feed systems afford From a detail analysis, we have found that
less stringent requirements for the various com- the change of the fuel pump discharge pressure,
ponents regarding deviations from their design as a function of turbopump speed increase or
steady-state flow values, because the system is decrease, is a fraction of that of the oxidizer
inherently more flexible. However, systems pump discharge pressure. Due to the effects of
dynamic characteristics under transient condi- chamber pressure deviations, therefore, the max-
tions may restrict these deviations. imum value of fuel calibration orifice pressure
drop is required when the following conditions
Sample Calculation (I0-2) exist:
(a._) Thrust chamber injector end pressure is at
The following design values and allowable
its lower limit (1065 psia)
deviations are given for the A-1 stage LOX/RP-1
(_b) All pressure drops in oxidizer passages
engine system components, at rated thrust:
are at their higher limits
Thrust chamber injector end pressure= 1095
(c_) All pressure drops in fuel passages are at
_+30 psia their lower limits
Thrust chamber injector pressure drop (both
(d.._)Oxidizer pump discharge pressure is 25
oxidizer and fuel) = 200 ± 20 psi
psi below its nominal value at the turbo-
Thrust chamber oxidizer dome pressure drop
pump speed commensurate with the
=150+10 psi
stated specific speed
Oxidizer line pressure drop = 25 -+2 psi
(e) Fuel pump discharge pressure is 25 psi
Main oxidizer valve pressure drop = 35 ± 3 psi above its nominal value, at the same
Oxidizer pump specific speed, Ns = 1980 rpm
speed
Oxidizer pump suction pressure = 55 psia rain
Oxidizer pump discharge pressure at 7000 rpm The equivalent required oxidizer pump dis-
and a design flow rate of 1971 lb/sec = 1505 charge pressure under these conditions = 1065
-+25 psia +220+160+27+38+25= 1535 psia.
Thrust chamber fuel jacket and manifold pres- Required oxidizer pump developed head
sure drop = 290 _+20 psi
Fuel line pressure drop = 10 + 2 psi H = 144 × (1535 - 55) _ 2990 ft
Main fuel valve pressure drop = 15 +-2 psi 71.38
Fuel pump specific speed, Ns = 1090 rpm
Fuel pump suction pressure = 45 psia rain Oxidizer pump volumetric flow rate
Fuel pump discharge pressure at 7000 rpmand
a design flow rate of 892 lb/sec = 1720-+ 25 1971 ×449 12 420 gpm
Q= 71.38 =
psi
......LL ::---
Substitute this into equation (6-7) to obtain

F6970 (7950) °'s11.aaa
the required pump speed H = L: _ ] = 4730 ft
or
N- Ns H°'Ts _ 1980 × (2990) °'Ts = 7190 rpm
QOS (12 420) °-s 4730 × 50.45
= 1666 psi
144
Fuel pump volumetric flow
The equivalent fuel pump discharge pressure

892 × 449 = 7950 gpm under these conditions would be 1666 + 45- 25
Q- 50.45
= 1686 psia.
The required pressure drop of fuel line cali-
From equation (6-7),the fuel pump nominal bration orifice under these conditions would be
developed head at 7190 rpm 1686 - 1125- 220- 310- 12- 17=2 psi.
Therefore the required range of adjustment for
the pressure drop of the fuel line calibration
H = (.NQaS_
\"_-s / I333 = f7170 ×
_ (7950)°s_ ] 1333 = 4900 ft orifice is from 2 to 216 psi.
or
10.4 ENGINE SYSTEM INTEGRATED
PERFORMANCE CHARACTERISTICS
4900×50.45
-1715 psi
144 In the process of engine system design
integration, an importanttask is the integration
The equivalent fuel pump discharge pressure of engine system perfonnance characteristics.
under these conditions would be 1715 +45 + 25 These data are preparedand compiled by the
= 1785 psia. rocketengine designer to provide the vehicle
The required pressure drop for the fuel line systems engineer with intbrmationnecessary to
calibration orifice thus would be 1785-1065-180 integratethe propulsionsystem with the vehicle
- 270 - 8 - 13 = 249 psi. system. _ere possible,a briefexplanationof
Similarly, a minimum fuel calibration orifice the data and itsapplicationshould be included
pressure drop is required when the following to provide clearerunderstandingand greateruse-
conditions exist: fulness. The followingare importantaspects of
(a_a_)Thrust chamber injector end pressure is at integratedengine performance characteristics.
its higher limit (1125 psia), and condi-
tions (b), (c), (d), and (2.) above are re-
Nominal Engine Performance Values at Rated
versed
Conditions
The equivalent required oxidizer pump dis-
charge pressure under these conditions = 1125 These are usually prescribed by the engine
+ 180 + 140 + 23 + 32 - 25 = 1475 psia. model specification. These data are for engine
Required oxidizer pump developed head system nondnal steady-state operation, at rated
conditions. Tables 3-2 to 3-5 are typical exam-
H=144(1475-55)=2870 ft ples of nominal engine operating and performance
71.38 parameters, which include nominal thrust, spe-
cific impulse, propellant combination, flow rates,
Substitute this into equation (6-7); the re- mixture ratio, and various component operating
quired pump speed data. Allowable deviations are specified for
important parameters such as: thrust, -+3 per-
N =1980×(2870)°_ =6970 rpm cent, and mixture ratio, *-2 percent. Engine sys-
(12420) °.s tem specific impulse is usually specified at its
minimum value. The performance of all deliver-
From equation (6-7), the fuel pump nominal able engine systems must be above this minimum
developed head during acceptance tests.
In addition to tables for nominal engine per- Required oxidizer flow for vehicle tank pres-
formance parameters, nominal engine performance surization-- 3 lb/sec
graphs such as chamber pressure versus engine Determine the following nominal performance
thrust, and engine specific impulse versus engine values at rated conditions:
thrust, are often included as additional monitor- _.) Thrust generated by the turbine exhaust
ing aid. Figure 10-7 presents a typical perform- gas
ance graph for the A-1 stage engine system, of (b) Thrust generated by the main thrust cham-
chamber pressure versus engine thrust at sea ber
level. _) Engine system propellant flow rates
Engine system mixture ratio
(e_) Engine system specific impulse
Sample CalcuIation (10-3)
The following data were obtained from design Solution

analyses and component tests of the A-1 stage
A trial-and-error
method is used to solve this
LOX/RP-1 engine system at nominal rated con-
ditions; i.e., 750000 pounds thrust at sea level: problem. Our firststep is to approximate engine
Thrust chamber sea level specific impulse at system and gas generatorpropellantflow rates.
We substitutethrustchamber Is intoequation
1000-psia nozzle stagnation pressure, and a
mixture ratio of 2.35 O/F=270 sec (1-28):
Turbine exhaust gas specific impulse = 32.6
sec = F _ 750 000 = 2778 lb/sec
Oxidizer pump developed head--2930 ft Is 270
Oxidizer pump overall efficiency = 70.7 per-
cent The thrust chamber mixture ratio can now be
Fuel pump developed head = 4790 ft used to approximate the corresponding oxidizer
and fuel flow rates:
Fuel pump overall efficiency =65.9 percent
Gas generator O/F mixture ratio = 0.408
Turbine gas available energy content = 359 Engine oxidizer flow rate
Btu/lb 2778x2.35
Turbine overall efficiency = 58.2 percent _i'° = (2.35+ 1) - 1948 lb/sec
Required auxiliary drive shaft power = 500 bhp
Oxidizer flow rate = 1948 + 3 = 1951 lb/sec
1250 l ' "

Engine fuel flow rate _i'f= 2778 - 1948 = 830 lb/sec
n,.,
0 1200
From equations (6-12) and (6-13), the required
oxidizer and fuel pump drive horse power for this
Z .
-- t,t.I 1150 approximation are:
n., rr'
1951 × 2930
-r'
_' ] IOO
Oxidizer pump horsepower - 550 × 0.706 = 14720hp
u
::)W -
(_
-r
1050 --/--
/
--
CONFIDENCE
Fuel pump horsepower - 830x 4790 = 10 96Q hp
550 x 0.659
I.- / LIMITS [
I000 --_ _
The corresponding turbine shaft horsepower:
7OO,OO0 7500(30 8OO,OO0

Thp = 14 720 + 10 960 + 500 = 26 180 hp
ENGINE THRUST, POUNDS
Figure lO-7.-Chamber pressure versus engine From equation (6-19), the corresponding tur-
thrust at sea level [or the A-1 stage engine. bine gas flow rate:
0.707 x 26180 Fuel pump flow rate

_i't= 359x0.582 -88.4 lb/sec
=(91.7- 26.6)+(2768- 1941)= 892.1 lb/sec
We use this value to start a new calculation Turbine shaft horsepower

cycle to separate main chamber and gas gener-
ator data. 1970.6x 2930 892.1 x 4790
ThP = 550 x 0.706 + 550 × 0.659 + 500
The thrust generated by a turbine exhaust gas
flow rate of 88.4 lb/sec = 88.4 x 32.6 = 2880 lb. = 14850+ 11 790+ 500= 27 140 hp
Thus the main chamber thrust:
Turbine gas flow rate w_ = 0.707 ×27140 _ 92 lb/sec

Ft = 750 000- 2880 = 747120 lb 359 × 0.582
Thrust chamber flow rate This value closely confirms the assumptions
for the last trial. Thus:
l_,tc = 747270120 2769 ib/sec (a__)Nominal turbine exhaust gas thrust

= 92 x 32.6 = 3000 lb
(b__)Nominal main thrust chamber thrust
Oxidizer pump flow rate = 750 000 - 3000 = 747 000 lb
(c) Nominal thrust chamber propellant flow
_ 88.4 x 0.408 + 2769 x 2.35 + 3
1+0.408 1+2.35 rate = 747 000/270 = 2768 lb/sec
Nominal engine system propellant flow
= 25.6 + 1942+ 3= 1970.6 lb/sec rate = 2768 + 92 = 2860 lb/sec
Nominal engine system oxidizer flow rate
Fuel pump flow rate 92 x 0.408 2768×2.35
+ - 1967.7 lb/sec
= (88.4- 25.6) + (2769- 1942) = 889.8 lb/sec 1+0.408 1+2.35
Nominal engine system fuel flow rate
Turbine shaft horsepower -: (92 - 26.7) + (2768- 1941) = 892.3 lb/sec
(d__)Nominal engine system O/F mixture ratio
1970.6×2930 889.8x4790 = 1967.7/892.3 = 2.20
Thp- 550x0.706 + 550x0.659 +500=27090 hp
(e_) NominaI engine specific impulse
= 750000/2860= 262.4 sec
Turbine gas flow rate
0.707x27090 Engine Performance Variations Resulting From

wt- 359 x 0.582 = 91.7 lb/sec Off-Nominal Conditions
Engine performance characteristics at various

We use this value for another calculation
off-nominal conditions must be available to the
trial:
vehicle system engineer. They can be summa-
rized in the graphic form such as figure 2-1 (en-
The thrust generated by a turbine exhaust gas gine thrust and specific impulse versus altitude
flow rate of 91.7 lb/sec=91.7 x 32.6=2980 lb. curve), or by means of tabulated engine influence
coefficients which will be discussed, The effects
Main chamber thrust Ft = 750 000-2980 = 747 020 lb of off-nominal conditions of the following engine
system performance parameters are considered
Thrust chamber flow rate vital for the design of a vehicle system:
(1) Atmospheric pressure
747 020
_i'tc = 27-----0_
= 2768 lb/sec (2) Propellant densities
(3) Pressures at _he engine propellant inlets
(4) Propellant mixture ratio and vehicle PU
Oxidizer pump flow rate
control
(5) Vehicle acceleration
_ 91.7 × 0.408 _ 2768 x 2.35 + 3= 1970.6 lb/sec
1 + 0.408 1 + 2.35 (6) Throttling of the engine system
-,qmNlm
Engine Influence Coefficients example, the change of engine thrust for the A-1
stage engine system (without C* correction) can
These are used to convert or correct steady-
be expressed as:
state, main-stage engine system performance
parameters (dependent variables) from one condi-
tion to another of parameters (independent vari- ( F - F n) _ C l(pa- Pan)+ C 2(po - pen)+ C 3([ [- pfn)
Fn Pan pon Pfn
ables) such as atmospheric pressure, fuel tem-
perature, oxidizer density, etc. This may be a C4(Poi- Poin) Cs(Pfi- Pfin)
correction to standard sea-leveI conditions (first- + + (10-26)
Poin Pfin
stage booster engine), or a conversion to other where
specified conditions. The coefficients are de- F, F, : engine system thrust and its nominal
rived from the linearized solution of a set of value, lb
steady-state differential equations which describe
Pa, Pan : atmospheric pressure and its nominal
the performance of an engine system. These •value, psia
equations are solved by a digital computer and
Po, Pon : oxidizer density and its nominal
presented in tabular form, as shown in table 10-1 value, lb/ft 3
for the A-1 stage engine system. Each influence
P[, Pfn = fuel density and its nominal value,
coefficient is expressed as a percentage and lb/ft s
represents the change of a dependent engine Poi, Poin :oxidizer pump inlet suction pressure
variable, such as thrust, as produced by a and its nominal value, psia
1-percent change in an independent variable, Pfi, Pfin = fuel pump inlet suction pressure and
such as atmospheric pressure. A coefficient its nominal value, psia
preceded by a positive sign (+) indicates that an
C_, C 2, C a, C 4, C s :influence coefficients
increase of an independent variable produces an
increase in the dependent variable. Conversely,
a coefficient with a negative sign indicates a
decrease in the dependent variable, as a result Estimate the thrust of the A-1 stage engine
of independent variable increase. These influ- system operated at the following conditions,
ence coefficients are usually sufficiently accu- without considering the effects of C* correction:
rate over the entire design operation range of an Atmospheric pressure, Pa = 10.2 psia
engine system. Oxidizer density, Po = 71.00 lb/ft 3
Because the influence coefficients are linear, Fuel density, p[= 50.90 lb/ft 3
the total effects of several influences acting Oxidizer pump inlet suction pressure, Poi = 65
simultaneously on an engine system can be de- psia
termined by summing the individual effects. For Fuel pump inlet suction pressure, Pfi--49 psia
TABLE lO-1.-Influence Coefficient for the A-I Stage Engine System
[Value of C* correction to be obtained from the C* correction versus mixture ratio curve shown in fig. 10-8]
Independent variables and nominal values

A 1-percent increase of
independent variables
Oxidizer pump Fuel pump
causes the following Atmospheric Oxidizer Fuel C _
inlet suction inlet suction
percentage change of pressure, density, density, correction
pressure, pressure,
dependent variables 14.696 psia 71.38 lb/ft a 50.45 Ib,/ft a 1.0000
56 psia 45 psia
Dependent variables and nominal values:

Engine thrust, 750000 lb .......... -0.1780 1.8750 -0.7420 0,0440 -0.0066 1.1030
Engine specific impulse, 262.4 sec . - .1780 ,2650 - .0640 .0072 - .0150 1.1350
Engine mixture ratio, 2.20 ......... .0000 1.6420 -1.3650 .0270 - .0020 - .0260
Engine oxidizer flow, 1967.7 lb/sec. .0000 2.0430 -1.1120 .0465 .0108 - .0632
Engine fuel flow, 892.3 lb/sec ..... .0000 .6530 .3120 .0207 .0045 .0094
Solution 10-1, which are for the A-1 engine system. The
change of engine mixture ratio is computed for
From equation (10-26) and table 10-1:
changes in atmospheric pressure, propellant
densities, etc., assuming the C* correction first
(F - Fn)_ (-0.178) x (10.2 - 14.696)
to be zero. For the resultant change in engine
F_ 14.696
mixture ratio, the C* correction is read from the
graph. The value of C* correction found is then
+ 1.875 x (71.0 - 71.38) _ (-0.742) x (50.90 - 50.45)
71.38 50.45 used with other independent variables to compute
the changes in the remaining dependent variables.
0.044 × (65- 55) + (-00066) × (49 - 45)
55 45
=0.04531 or 4.531 percent
Estimate the thrust of the A-1 stage engine
Engine system thrust of an altitude, where system operated at the conditions listed for
Pa = 10.2 psia: sample calculation (10-4), adding the effects of
C* correction, Also, for the same conditions,
F = 0.04531 x Fn + Fn = 0.04531 x 750 000 + 750 000 estimate the thrust assuming an additional mix-
: 784 000 lb ture ratio error of +10 percent, due to faulty
calibration.
Nonlinear Corrections
Solution
When the linear approximation is not suffi-
ciently accurate, the usefulness of the engine By analogy with equation (10-26) and using
influence coefficients can be extended by a table 10-1, the engine system mixture ratio
technique which allows nonlinear corrections for change due to the conditions of sample calcula-
certain parameters. An example of this method tion (10-4) are determined as
is the C* correction. For instance, a plot of C*
correction versus engine mixture ratio change (MR- MRn) _ 1.642 x (71.0- 71.38)
may be used in conjunction with a table of influ- MR n 71.38
ence coefficients such as figure 10-8 and table
-_ (-1.365) x (50.90 - 50.45) ÷ 0.027 × (65- 55)
50.45 55
(-0.002) × (49-45)__0.0162 or -1.62 percent

45
-0,1
_-.
Z
L_ From figure 10-8, the C* correction for a mix-
-02 I ture ratio change of-1.62 percent is approxi-
i mately -0.02 percent. From table 10-1, the in-
Z
9 -o.3 fluence coefficient for engine system thrust is
i.-
1.1030, for a i percent C* correction. Thus
e¢
_ -0.4
u (Percent change in F) = 4.531 + (-0.02) x 1.103
-0.5 = 4.509 percent
-O.6 Therefore, engine system thrust considering

-16 -12 -8 -4 8 12 effects of C* correction:
CHANGE IN ENGINE MIXTURE RATIO,
PERCENT
F= 750000 × (1 + 0.04509) = 783 820 lb
Figure 10-8.-C* correction versus change in
engine mixture ratio curve for the A-1 stage If the mixture ratio error of 10 percent is
engine. added, the total mixture ratio change = 10-1.62
ENGINE SYSTEMSDESIGN iNTEGRATiON 399
= 8.38 percent. From figure 10-8, the C* correc- in the vehicle in the field should not require
tion then is approximately -0.11 percent. assembly of additional major components. Integ-
Thus rity of the propellant feed and hot-gas systems,
(Percent changein F)=4.531 +(-0.ii)× 1.103 once verified in a complete system during ac-
ceptance test, is not necessarily mdlified by the
=0.441, or 4.41 percent need to temporarily disassemble the engine for
Engine system thrust: shipment. The integrated engine package con-
cept provides added assurance that static test-
F = 750 000 × (i + 0.0441) = 783 080 Ib
stand firing results have verified structural
soundness of the package to a substantially
greater degree, than is the case for a system
10.5 MECHANICAL INTEGRATION OF where the vehicle provides portions of the engine
ENGINE SYSTEMS structure.
An example of a special case of mechanical
Basic Considerations integration of a liquid propellant rocket engine
is the prepackaged storable liquid rocket propul-
Besides combining allcomponents and sub-
sion system shown in figure 8-1. This system is
systems functionallyand physically,the design
a completely integrated assembly of all-welded
formechanical integrationof an engine system
construction, consisting of thrust chamber as-
must considerti_eoverallenvelope of the system
semblies, propellant tanks, pressurization sys-
and itsweight. This includes the locationof
tem, and necessary controls. This provides
the system's centerof gravity. Also, itshould
maximum assurance of system integrity from the
permit simplifiedmaintenance and checkout
time of manufacture, which includes loading of
practices. Judiciouspackaging design tech-
the propellants, through delivery, vehicle assem-
niques should be applied to minimize the number
bly, and launch. Complete propellant separation
of interconnectinghydraulic,pneumatic, and
until systems start is achieved by hermetically
electrical lines, with their attendant fittings,
sealed burst diaphragms for maximum safety.
connectors, joints, and other potential trouble
Acceptance tests are conducted by taking sample
spots. Welded and brazed joints should be used
units at random from the production line, and hot
as much as possible. Problems introduced by
firing them. In addition, destructive tests of
vibration, high temperatures and pressures, leak-
various types are performed.
age and space restrictions are thus more easily
handled. Engine mechanical integration is a
Packaging of Rocket Engine Components
vital part of the system design concept; there-
fore, all factors related to integration and pack- Most major rocket engine components, such
aging of components and subsystems must as thrust chamber (fig. 4-1) and turbopump (fig.
receive careful consideration early in the prelim- 6-14) assemblies, readily form a logical, inde-
inary design stage. pendent mechanical unit by virtue of their func-
In general, a modular engine packaging ap- tion and their physical shape. However, in the
proach should be selected such as used for the case of minor components such as control valves,
A-1 (fig. 3-2) and A-2 (fig, 3-4) stage engine gas generators and igniters, packaging design
systems, as well as for many advanced opera- principles can best be served by making them an
tional engine systems. This assures engine integral part of a major component assembly, or
integrity from time of manufacture through vehi- to integrate them by grouping. A typical example
cle launch. It also provides a compact package is a gas generator assembly externally attached
for ease of handling, transportation, and installa- to a turbine inlet flange (fig. 3-2). Similarly, gas
tion in the vehicle. Ease of checkout and com- generator propellant valves and combustor can
ponent accessibility is also afforded by the be integrated into one unit (fig. 4-51).
packaging concept. Certain types of hydraulic and pneumatic
The engine should be completely assembled rocket engine control components lend them-
in the manufacturer's plant. Subsequent accept- selves most conveniently to the packaging de-
ance testing, air transportation, and installation sign. Here, one of the main objectives is to
4O0 DESIGNOF LIQUID PROPELLANT ROCKET ENGINES
reduce line runs, by combining all parts and sure regulator to the control system. The helium
passages into one housing. Such a housing (or is routed internally to the main control valves
mounting plate) is relatively leakproof, trouble through a fail-safe check valve. This insures
areas now being limited to external line connec- that the various engine propellant valves remain
tions to other components. Furthermore, if com- pressurized and thus open, should the helium
ponents are packaged in this manner, reductions gas supply system fail.
of weight and size are achieved through the use
of common walls and through the elimination of
extra mounting platforms, clamps, and fasteners. Packaging of Turbopump Feed Engine Systems
Since relatively few packages are required as In earlierhigh-thrustrocket propulsionsys-
compared to the usually large number of individ- tems, some of which may still be in operational
ual components, maintenance of such a system use, allmajor engine components were mounted
is greatly simplified. Integrated packages are into a cage-shaped thrust mount, which was
about as easily removed and replaced as are the bolted to the vehicle thrust frame by way of lugs.
separate components making up each package. Figure 2-4 allows several typical examples. With
However, the packaged design is not necessarily these systems, vehicle steering was accom-
desirable for every control system. Each case plished by means of carbon jet vanes protruding
must be carefully studied. into the jet (V-2 and Redstone), or by swiveling
As a rule, one or a combination of the follow- the thrust chamber (Thor, Jupiter). In the latter
ing methods is used for packaging engine control case, the high-pressure feed lines between
corn ponents: pumps and injector had to be much more flexible
(1) Bank packaging: A group of similar flat- than for misalinements and thermal expansion/
sided component assemblies are bolted contraction alone.
together in a bank or stack, with common Most advanced liquid rocket engines are
porting through the mating surfaces from tightly packaged. All major components are
one unit to the next. attached to the main thrust chamber, directly or
(2) Subplate packaging: Attachment of two or by means of mounting structures, as shown in
more individually housed components to figures 3-2, 3-4, and 9-1. Here, the thrust cham-
a subplate, so that all ports of the in- ber serves as the principal structural member of
dividual component housings lead into the entire engine system. For steering, the com-
the subplate manifold, through their plete engine package is gimbaled from a gimbal
mating surfaces with the subplate, and bearing which attaches directly to the thrust
on to the systems plumbing. chamber dome. The other half of the bearing is
(3) Cartridge packaging: Two or more compo- attached to the vehicle thrust structure. The
nents housed individually in cylindrical low-pressure propellant supply duets must be
cartridges are in turn assembled in a sufficiently flexible to accommodate the gimbal
common body with suitable manifolding motions. It is noted that vehicle steering through
to the systems plumbing. gimbaling of a single engine or chamber is effec-
(4) Multiple-component packaging: Detail tive only for the pitch and yaw planes. For roll
parts for two or more components are control, at least two engines are required. For
assembled in a normal fashion in a com- vehicles with a duster of engines, therefore,
mon housing or body. this poses no difficulties. For single-engine
Figure 10-9 presents a typical pneumatic con- vehicles, special roll-control devices are needed.
trol package for a large liquid propellant rocket These may be small auxiliary nozzles, possibly
engine. This package combines two pressure- simultaneously used as vernier engines after
regulator assemblies, two relief valves, a series main-engine cutoff. The use of the turbine ex-
of solenoid valves, filter units, and check valves. haust for roll control has also been proposed.
It controls the flow of helium gas to various Whether the engine attaches to the vehicle
engine components. When engine start is initi- thrust structure by means of a thrust frame or a
ated, the helium control solenoid is energized gimbal bearing, either device must be designed
allowing helium to flow through the main pres- to be capable of transmitting the full thrust
5 MICRON FILTER- fG.G._ BLEED VALVE CONTROL SOLENOID VALVE

BLEED PRESSURE REGULATOR-_ / HELIUM CONTROL SOLENOID VALVE
\ / /---'--'--MAIN PRESSURE REGULATOR
MA,NSTA_F CC1NT _\ tlJ_.._/_ ,/._t_ j.---,GN,T,ON PHASE CONTROL

.......... _ L
-..-1::;-_:_i'_>_!_ __FAIL-SAFE CHECK VALVE
HIGH PRESSURE __ _ 10MICRON FILTER

RELIEF VALVE t _x__LOW PRESSURE RELIEF VALVE
HELIUM
INLET
PNEUMATIC CONTROL PACKAGE SCHEMATIC
VENT PORT CHECK VALVE PRESSURE RELIEF VALVE
IGNITION PHASE
HELIUM SOLENOID CONT.
IN LET VALVE
VENT
CHECK VALVE
VENT PORT CHECK

VALVE
INSTAGE SOLENOID
CONTROL VALVE
HELIUM
CONTROL
SOLENOID
VA LVE
BLEED PRESSURE
REGULATOR G.G BLEED VALVE CONTROL SOLENOID VALVE
Figure lO-9.-Typical pneumatic control package design used in liquid propellant rocket engine
systems.
J
forces at full gimbal deflection, including an

adequate reserve for normal and for side loads.
The bolt-hole pattern must permit adjustment for
tolerance deviations. In general, engine and
vehicle attachment halves must be designed for
one another. LU|E OIL TANK
Figures 10-10 and 10-11 illustrate the pack-

aging design details of a typical pump feed sys-
tem. It is a LO2/RP-1, fixed-thrust engine with
constant chamber pressure control. The basic
engine package consists of the following sub-
packages:
(I) Gimbaled main thrust chamber assembly
(thrust chamber, injector, dome, oxi-
dizer elbow, and gimbal mount)
(2) Turbopump assembly (propellant pumps,
turbine, gearbox, lube pump, electric
heater, auxiliary drive)
(3) Gas generator assembly (combustor, con-
trol valves, regulator and turbine inlet
duct) L_N OX_OIZE_ V_LVE
(4) Main oxidizer duct assembly (including

main oxidizer valve) Figure 10-I I.-System packaging design detail of
(5) Main fuel duet assembly (including main the engine shown in figure 10-10.
fuel valve)
(6) Turbine exhaust duct assembly (includ-
(10) Electrical control package
ing heat exchanger)
(11) Engine thrust frame assembly
(7) Engine start subsystem (oxidizer and
fuel tanks, control valves)
The majority of the major component and sub-
(8) Turbopump lube subsystem (lube oil tank system packages are installed within or at the
and fittings) periphery of the engine thrust frame assembly.
(9) Pneumatic control package The main thrust chamber assembly is attached to
the thrust frame through a gimbal mount.
Mechanical Protection of Engine System

Packages
It is normal that several years may elapse

between the date a liquid propellant rocket en-
gine is completed and accepted by the user, and
its vehicle flight. Design, therefore, must con-
sider means to protect the engine system pack-
age in transit or storage against moisture, dirt,
and shock. These include simple closures, such
as caps, plugs, and cover plates applied to valve
and regulator vent holes, propellant inlets, and
to other openings. Frequently, these closures
contain desiccant bags and indicators which
Figure lO-lO.-Major component and subsystem warn, through change of color, of undesired in-
packages of turbopump-[ed liquid propellant trusion of moisture. Certain lines, however, may
rocket engine. require communication with the ambient air
ENGINE SYSTEMS DESIGN INTEGRATION 4O3
are many tasks, such as sequencing which can

be accomplished nmch more effectively electri-
cally than would be possible by mechanical
means.
It is not possible nor necessary, in the frame-
FUEL
work of this book, to describe the physical laws
and the general fundamentals of electrical cir-
cuitry. They are covered abundantly in the
_CLOSURE COVER literature. Moreover, in contrast with most other
basic liquid engine subsystems, the rocket en-
gine designer will try to use commercially avail-
able "off the shelf" components for his electrical
system. However, other cognizant members of
the design team will have to provide the basic
circuit diagram (schematic) and other data in
support of the installation of the required elec-
trical components. Characteristically, the elec-
trical system of a rocket engine is one of the
"_'__THRUST CHA,_SE/_ EXIT
last subsystems to be "frozen" before produc-
CLOSURE COVE_ tion. This is because sequencing for start and
stop represents one of the major engine develop-
Figure 10-12.-Various protective closure covers
ment activities, often resulting in repeated modi-
for the engine shown in [igure 10-10.
fication of the electrical system as development
progresses. Emphasis is therefore placed on the
flexibilityof electrical design. More recently,
("breathing"). In this case, the closures may be
this process has been greatly aided by dynamic
equipped with desiccant filters to permit access
analyses (see sec. 10.2).
of dry air only. Some of the covers must be re-
moved for installation of engine to vehicle.
Others will be left in place until the engine is
Electrical Schematic
actually operated. These must be readily acces-
sible and clearly marked, such as with bright The complete electrical schematic of a typi-
colors, to prevent their being left in place inad- cal liquid rocket engine system, including its
vertently. Figure 10-12 shows the location of ground- and vehicle-based elements, fills a
various protective closures for the Rocketdyne sizable drawing. With the aid of figure i0-13,
LR79-NA- 11 engine. which presents a portion of an earlier engine
static-firingschematic, the basic features are
discussed as follows. In ordinary wiring dia-
10.6 ELECTRICAL SYSTEM
grams, such as that of a radio receiver, all the
All rocket engines depend on some type of contacts of, for instance, a multiple switch or a
electrical system for their operation. This is tube are drawn to appear in the same location, as
true for solid systems, where at least ignition is they do in reality. This requires numerous wire
initiated electrically, as well as for liquid sys- crossovers in the diagram. The number of cross-
tems, in which the electrical system assumes overs would become prohibitive in a typical
numerous additional tasks. As with any common engine electrical schematic and may lead to
household device, electrical circuits in rocket confusion and errors. For rocket engine pur-
engines have caused troubles, due to poor de- poses it has long been found preferable to draw
sign, misapplication, abuse, poor maintenance, the basic diagram so as to show each circuit
human errors, and wear. Properly applied, how- separately. In this "functional flow diagram,"
ever, electrical circuits can" substantially sim- the various contacts of a relay, for instance,
plify the operation of a rocket engine, and will appear in different places, and often away from
increase its usefulness and reliability. There the circuit for the corresponding relay coil. The
Figure I 0-13.-Typical liquid rocket engine electrical diagram (partial).
drawing system also will materially simplify K28C, "Hypergol Cartridge Installed," as well
later troubleshooting. The diagram, of which as the normally closed contacts of cutoff relay
figure 10-13 shows a portion, was drawn sequen- KglC. If all contacts are properly closed, the
tially from left to right; i.e., circuits which are "C" contacts of K34C will cause signal light
energized during test preparation and start are DS34C to light up. Following selection of the
shown in the left portion, while those associated ig-nition power source by means of switch S16C,
with the cutoff sequence appear on the right. ignition can now be initiated by means of push-
In the schematic, connectors are shown as button $51C, since the "D" contacts of K34C are
continuous double lines, or portions thereof, now closed, and provided ignition disconnect
running horizontally through the diagram (J16, timer K54C has not picked up (TDPU e time de-
P16-- receptacle 16 and plug 16, etc.). layed pickup : 0.1 seo). In the diagram, several
Each of the contacts is called out by a letter circuit elements appear which are part of other
(T, G, K, etc.). All wires are numbered, as in- circuits not discussed. Note that in places two
dicated. Power buses, like connectors, are relays are used in parallel (e.g., K16C), if the
shown as horizontal lines, or portions thereof number of contacts required is too large for one
(heavy, single line = positive buses, usually relay. The numbers shown in hexagonal frames
shown near the top of the drawing; and double refer to the channels of an inking sequence strip
lines = negative or ground bus). The meaning of chart recorder or equivalent instrument. A spe-
the remaining symbols becomes clear by follow- cial test bus K615 is provided which when ener-
ing the circuit at the left of figure 10-13. Plug gized makes all signal lights go on and thus
P5 is shown connected to the main power bus permits spotting burnt-out bulbs.
K101 at terminal TB1-8. If certain facility sig- In earlier engine designs, many of the ele-
nal contacts are properly closed, such as those ments shown in figure 10-13 were installed in an
verifying "'Cooling water OK," "Firex armed," engine-mounted relay box. The trend has been
"Observer on Station," and many more, power to place as many parts of the electrical system
returns through plug P5, contact "Z," and is on ground as possible. This is easier with first
applied to relay coil K31C. The "B" contact of stages, which start while still connected to
this relay closes a circuit to lamp DS31C, which ground, or even held down mechanically until
lights up. The "A" contacts of the same relay released, for brief periods following start, than
are in the chain to the coil of K34C, together it is with upper stages which must start and
with the normally open contacts of several other stop, and sometimes restart, some time after
relays, such as K29C, "Heater Power On," and. takeoff. Beca_se of individual approach and of
ENGINE SYSTEMS DESIGN INTEGRATION 40,5
preferences for the types of interlocks, safe- must be at a lower potential than the emitter.
guards required, and type of component used, This is accomplished by inserting a bias volt-
two designers of a comparable engine may arrive age, V_, in the emitter leg, which raises the
at substantially different electrical diagrams. emitter potential V 1 volts above ground. Again,
Specifically with respect to the number of inter- as long as no appreciable current flows through
locks and monitoring circuits applied, caution is R I and R 2, the base will essentially be at ground
advised, since these circuit elements by them- potential, and Q1 will remain off.
selves are subject to malfunction and may do When input voltage Vin is applied, current Ix
more harm than good. will flow. This raises the Q, base potential,
The diagram discussed above employed re- until the combined Q, base-emitter voltage,
lays, well-developed types of which continue to (Vbel), plus bias voltage (V_), is overcome. At
be used in several of today's rocket engines. In this point, current will flow into the base (1b,)
others, solid-state (transistorized) switches are causing the transistor to turn on. This occurs
being applied which fulfill a similar function. when
Solid-state switches have the advantage of re-
R3(Icbol)>(Vbel +VI) max (10-27)
quiring no moving parts and thus are much less
sensitive to vibration effecLs. A circuit of this
When Q, turns on, resistor R 4 is switched to
type is shown in figure 10-14. It functions as
follows: ground, and current Ic, begins to flow. As Ic
increases, the potential at the base of Q2 is
Transistor Q2 is held in a "turned off" mode
lowered until it reaches the combined potential
by maintaining the base at a higher voltage
of V 2 and the Q2 base-emitter voltage, Vbe2. At
potential than the emitter. This is achieved by
this point, current Ib2 flows out of the base of
inserting a bias voltage, V 2. As can be seen,
Q2 and the transistor "turns on," thereby supply-
as long as there is no appreciable current flow
ing current to the load. This occurs when
in R,, the base potential will be essentially Vo,
while that at the emitter is Vo-V 2. Thus, Q2
Rs(Icbol+lcbo2)>(V2+Vbe2) max (10-28)
will remain off.
Transistor QI is held in a "turned off" mode
The switch is turned off by either removing
in a similar fashion, except that here the base
the input signal, Vin , or by using another static
switch to ground the base of Q_.
Switch lock-in (to maintain output after input
, _ vo
signal is removed) is accomplished by feeding
V2
the output voltage back to the input through a

diode. In this case, shorting the base of Q,
must be used to turn off the switch.
Figure 10-15 shows a typical solid-state

LOCK-IN METHOD
L'I OUTPUT switch module. Following assembly, a module

DIODE is immersed in a potting compound. Overall
R4
module dimensions after potting are comparable
L to those of a matchbox.
Ichor//
i t IM '4-" "_
Figure 10-16 shows a small portion of an
INPUT engine-mounted sequence-controller diagram.
+( --------_v.
R2 LOAD The modules shown, together with the others
required for engine operation, including connect-
ing wiring and connectors, are housed in a her-
tiN
metically sealed box or can.
Although the two diagrams are not related,
the signal emitted at pin X of receptacle J-5 in
GROUND
figure 10-13 would be suitable to initiate engine
Figure I 0-14.-Typical rocket engine solid-state start at contact A of plug P51 in figure 10-16.
switch circuit or "module" diagram. Similarly, the signal emitted from pin K of plug
Electrical Components
To implement rocket, engine electrical cir-

cuitry such as the one partially shown in figure
10-13, a number of components are required.
These are:
Relays and Switches

Relays and, more recently, the solid state
switches are used to receive usually low current
command signals from the vehicle or ground con-
trol center and to translate them into properly
sequenced actuation signals to igniters, control
valves and other elements. In combination with
interlocking relay contacts, bias voltages, valve
Figure lO-15,-Assembled soJid-state switch position switches, continuity monitors, tempera-
module be[ore and after potting. ture sensors, spark plug monitors, voltage sen-
sors, timers, and other devices, they form an
engine-contained logic which will execute a
P51 in figure 10-16, with the aid of an auxiliary sometimes elaborate starting sequence in re-
relay, could be used to assure engine readiness sponse to only two external signals: start and
in the chain leading to the coil of relay K34C in stop. In practice many more signals are ex-
figure 10-13. changed between engine and vehicle and/or
R * ( GIN[ ST*AT
Jal FO_ 20m$ _l_

P_r
•5, I <D
. I s c
I' _
T($1 ¢O_'L(?( z Z?-','

..... ,
Cm,T g_BI (_smg s'rAx'r
r,_;'(_ ........
, _ e÷ CLOS*J_I[ • _ _Z{)$-II,,.,
3 r •
I_NITION
L. ................. J
zz_'s_ c_2_ c_SO
Figure 10-16.-Portion of a typical engine-mounted sequence controller diagram using

solid-state switches.
ENGINE SYSTEMSDESIGN INTEGRATION 4O7
ground control center, for checkout, monitoring, houses 4 exciter coils for the spark plugs of
confirmation, instrumentation, telemetry and thrust chamber and gas generator.
emergency reaction systems. Also, power supply
connections are required. Position Indicators
The relays and switches, together with sup- To verify the position of a valve, or the proper
porting circuitry elements such as resistors, installation of an igniter and similar mechanical
capacitors, diodes, terminal strips, and connect- conditions, position indicators are used. Most
ing wiring, are best housed in a common box. common types are:
This sequence controller is hermetically sealed Switches.-These are mainly used for "black
and often includes temperature-conditioning pro- and white" indications, such as "open" or
visions by means of electrical heaters (typical "closed," "installed," "connection OK," etc.
power requirement: 200 watts) or inert gas Numerous commercial products are qualified for
purges. The engine systems designer very likely rocket engine application and are available as
will receive from the cognizant department merely compact, miniaturized, sealed units. They are
a "black box" description of the sequence con- part of the individual component designs and are
troller, giving external and mounting bracket described in other chapters. It will be the burden
dimensions, connector descriptions, and installa- of the engine systems designer, however, to as-
tion specifications, with all internal detail sure uniform standards, specifically for connec-
omitted. Because of miniaturization, relay boxes tors. Position indicators of the switch type are
or sequence controllers have become very com- instrumental for sequencing the start and stop of
pact units. Figure 10-17 shows a typical can. a liquid rocket engine.
Note brackets for engine attachment. In addition Continuous indicators.-To know any inter-
to approximately 30 modules, numerous diodes, mediate position of a component, such as the
resistors, timers, etc., the container shown
angular position of a valve gate or the linear
displacement of an actuator, continuously read-
ing devices are required. Most widely used
types are potentiometers and variable reluctance
pickups. Potentiometers require adc power
source (typical: 5 volts), while inductive-type
sensors require an ac power source (such as 400
cps, 28 volts). The output of both, often in com-
bination with a bridge circuit, is fed to a telem-
etry and/or ground recording system. Indicators
of the continuous type are predominantly used
for instrumentation. It should be a goal to de-
crease their number as the development of the
engine progresses.
Timers
During engine start and stop, timing devices
are required for two principal reasons: correct
sequencing of valve actuations and other events,
and for monitoring the correctness of critical
sequence times. In the latter case, they will
inhibit progression of the sequence or initiate
cutoff in case of malfunctions. A degree of tim-
ing can be accomplished by means of orifices
and by sizing of volumes in the pneumatic or
hydraulic activation system. All other timing
Figure lO-17.-Liquid rocket engine sequence must be provided by timers as part of the elec-
controller. trical system. For engine systems starting on
the ground, these timers are mostly located in main battery. Because of the relatively high
the ground control system. Several types of power requirements for heaters airborne heating
timers are commercially available for this appli- should be kept to a minimum. This can be ac-
cation. They may be motor driven or may use complished by judicious placement of compo-
dashpots, springs, or other delaying devices for nents away from areas of extreme cold, and by
their function, usually with the capability for insulation and isolation. For instance, a hydrau-
external adjustment within a specified range. If lic pump attached to the auxiliary drive of a LOX
ground mounted, the timers' weight and size are pump, or the actuator of a cryogenic valve (see
of minor importance. fig. 10-18, near center) may require drastically
For systems starting at altitude, all timers reduced heating, or none at all, if a heat barrier
required must be engine mounted (or at least (gasketlike wafer of suitable material) is placed
stage provided). Here, weight and size do be- between connecting flanges. Furthermore, in the
come important. Modifications of the solid-state hydraulic system of an engine, heating require-
switches described above, through addition of ments may be completely avoided if the hydraulic
suitable capacitance/resistance circuits, have fluid is continuously circulated during standby.
been successfully applied. As a rule, these For cases where electrical heaters must be
timers are not adjustable once assembled, but applied, a number of types are available. A
require replacement of the entire module in case common one is the blanket heater, which is ap-
of timing changes. plied externally as a sleeve or cover, formfitted
for best efficiency, and equipped with a ther-
mostat.
Heaters
In other applications, an immersed heater,
Ideally, all components of an engine system
i.e., a Calrod-type heating element cast or em-
should be capable of operating reliably and un-
bedded into the component metal, is used, for
aided at all temperatures occurring during opera-
instance, to protect a cryogenic pump bearing.
tion. In many applications, however, extreme
temperature variations occur, due to the environ-
ment at high altitudes, or within an engine sys-
tem, particularly within those using cryogenics.
The development of components able to with-
stand these extremes without assistance would
often be very costly or altogether impossible.
Thus it is much simpler to reduce or eliminate
the temperature extremes by suitable means. For
some areas, mere insulation may be adequate. j
Others require heating (external means for cool-
ing are rarely required for rocket engines, but are
used for other vehicle systems). Bulk tempera-
ture conditioning is usually provided on static !II11
firing stands and within vehicles during standby
by blowing preheated inert gases or air into the
engine compartment. However, certain strategic r-::. -_- _:_ ._: __'_.:_;.:,_ig
components will require individual heating at all

times after propellant loading. Of these, some
may require this only through takeoff from a 28-
volt dc ground source, if their heat capacity
keeps them sufficiently warm for their relatively
short period of operation following heater dis-
connect. Other components, particularly those
which have to operate during extended periods
following takeoff, will require supply from an Figure IO-18.-Wire harness installation for main
airborne power source, which is usually the 28-V valve position switches.
Power Sources; Batteries defines the branch points and their rela-
tive distance from the main plug or
For its operation, a rocket engine electrical
receptacle.
system requires power which is almost always
It is usually difficult to determine the exact
supplied by the vehicle or from the ground. The
engine designer need not concern himself with length of individual wires and of some of the
branches on the drawing board. It is customary
the power supply, except for the specification of
to finalize these dimensions on a mockup engine.
requirements and provisions for connection.
For this reason, certain dimensions in the physi-
Frequently used voltages are:
cal routing diagram are left blank and the total
28-volt dc for heaters, control solenoids,
cable length is specified "as required."
relays, switches, igniter spark exciters,
For proper design of a rocket engine wire
certain instrumentation (typical peakload:
harness, the following must be considered:
2000 watts per engine)
5-volt dc for instrumentation (potentiometers) Routing (minimum distance, weight, electrical
resistancel interference)
8-volt dc for spark monitors (see ch. IV, spark
Secure clamping (safety, avoidance of wire
plugs)
chafing)
ll5-volt, 400-cps ac for instrumentation
Selection of attachment points (avoidance of
The required power is supplied directly from
special brackets)
batteries for dc needs, or through converters for
Adequate support (harness weight, vehicle
ac consumptions.
acceleration)
Miscellaneous Components Moisture protection (potting, sheathing)
Heat protection (routing, wrapping)
To complete an engine electrical system, a
Arcing protection (component selection, dry-
number of standard or special components are
heSS)
required: resistors, capacitors, diodes, terminal
Flexibility (installation, stiffness versus
strips, connectors, and receptacles. All are
gimbal load)
commercially available, qualified for rocket
Connectors must have means to secure them
engine use.
to prevent accidental disconnect. This has been
Connecting Wiring done with the connectors shown in figure 10-18
The various subsystems of the engine elec- by means of threaded sleeves, further secured by
trical system must be connected by suitable wir- safety wires. To prevent incorrect connections
ing, which for proper mounting and clean routing and/or damage to the connector pins, "clocking"
is combined into one or several wire harnesses. by means of key and slot, of connectors and
Figure 10-18 shows portion of a harness in- receptacles, is highly recommended.
stalled on a liquid rocket engine. For manufac- In addition to the cabling required to inter-
ture, the engine systems designer must supply connect the various engine electrical subsys-
certain information. This is best combined into tems, other cables are required to link these
a single drawing, which contains the following systems to the vehicle systems. This includes
information (fig. 10-19): wiring for power supply, controls (start and stop,
(1) Wire list.-This list calls out each harness PU, throttling, etc.), instrumentation and check-
wire by number and lists its length. It out circuits. To facilitate engine installation
also specifies the wire routing between and line connections, these wires are combined
plugs and/or receptacles. into trunks, each terminating in connectors which
(2) Harness schematic.-For clarity, the sche- must have, at the proper location, a mating coun-
matic repeats in pictorial form most of terpart on the vehicle. For the stage, a "J-Box"
the information supplied in the wire list, (junction box) is recommended, into which are
together with additional information. It plugged both the engine and the stage systems
also calls out the total length of wire connectors. On the inside, the box permits easy
required for material procurement pur- and environment protected redistribution of
poses. incoming and outgoing wires to assigned con-
(3) Physical routing diagram.-This diagram nectors. Figure 10-20 shows a typical block
WIRE LIST*
__PlO
WIRENO 1LENGTH I FROM I TO
EIOBtB{RD)
EIOAI8 12OO JIIO-A PIO-A
EIOBIe I 1200 I JIIO-B I PiO-e
EIIAI6 19SO I JIIO-C PII-A
i IOAIB(WH]
EIIBIe l 19.50 JIIO-D PII-B
El2 A H_I ! 12,50 JIIO-E PIZ-A
: EIZBI8 ' 2,50 J O-F PI2-B
_- EIIBI8tRD)
EI3AI8 12.00 JIIO-G PI3-A
JIlO
EI3BIS 12.OO J IIO-H PI3-B
EI4A 18 23OO JIIO-J P}4-A
EI4BI8 23.00 g IIO-K PI4-B
-'_ EIOAtBtWH) -_
EIOBIB(RD) -_ EISAle 23,75 JIIO-L PIS-A
EISBI8 23,75 Jll0-M PIS-B
" F'I6AI8 2775 JIIO-N ] PI6-A
_1_ El II AIB(wH) "_
El BI8(RD) "_
EIYAIB 16.00 JIIO-R JIII-A
• EI6BI8 27.75 JIIO-P t PI6-B
EIYBI8 I6.O0 JllO-$ JHI-B
El2 BI8(RD) "_

,.- EI3AI8(WH) :?'-_ PP3
_- E]3BI8 (RD)
EI3B18 (RD) "_
EI4BI8(RD} •
EI5BIB(RD) J
_'J _ PI6 8.'25

EI6BI8(RD} -J
RI_0"_ E,ZAIS<WH)_ EISAIB

EI5BIB(RD) (WH) _._% PI5
E_.3 PI4 _____ P[2
;_____ j/IEIyBIB_RD)
PHYSICA-ROUTING DIAGRAM
(CUMULATIVE DISTANCES, INCHES
-_ RECEPTACLE ._PI6
::::::((:;:'-_--J RECEPTACLE
DENOTES BRANCH POINT
, (WH) B (RD} DENOTES COLOR OF WIRE INSULATION
_" DENOTES TWISTED SHIELDED PAIR

'_LEN_H--GT DENOTES TWISTED PAIR
FROM FACE OF MAIfJ RECEPTACLE (J-riO),
Figure l O-lg.-Typical rocket engine wire harness diagram.
[
GSE STATIC / VEHICLE
TO G ROUNO
CONTAJNE!R
(BATTERIES AND UMBIUCAL
ELECTRICAL 1 CONNECTIONS PATCH pANEL I VEHELE 1
CHECKOUT [ INSTRUMENTATION CONNECT{ONS
CONTROL DEVICES
I
VEHICLE MOUNTE_ JUNCTION BOX
(ONE PER ENGINE]
[] [] ]
=u _ _'IN o c.,....r)_r@ 0]_

TEMPERATURE
MEASUREMENTS
!li
c_ < c_ < (SENSORs)
i E.,NE I i PRIMARY ENGINE

SECONDARY
ENG_
: CONTROL I INSTRUMENTATION --E
INS TRU MENTATI0_N
PACKAGE
PACKAGE
PACKAGE I
Y' LTI PU V_LVE POSITION (POTE.NTIOMETERS AND SWITCHES'I L}.J

MONITOR _ MEASUREMENTS
Figure 10-20.-Typical block diagram, engine to vehicle electrical connections.

diagram for the principalelectrical

connections its significance. The engine designer must be
between engine and vehicle. Flexibility of these fully cognizant of it, and must be sure to consult
trunksusually poses no problem in view of the the development engineer and the instrumentation
relativelysmall engine deflectionduring gimbal- specialist at the very beginning of his design.
ing,but ifheavy sheathing(armor)is used, care
must be taken to avoid excessive loads to the Principal Types of Instrumentation
gimbal actuatorsor damage to the cables. Instrumentation signals to the vehicle are
generated by engine-mounted end organs or
transducers, with typical applications as follows:
10.7 ENGINE INSTRUMENTATION
Temperature gages.-Thermocouples or re-
To evaluate the qualityof his design dur- sistance bulbs are the principally used types.
ing and aftertest,to monitorprogress,and finally Their main application is for fluid-temperature
to "sell"his engine to the customer,the de- measurement at the pump inlets for the determi-
signerneeds recorded evidence of the engine's nation of mass flow rates in combination with
overallperformance. These records may broadly volume-reading flowmeters. Also, temperature
be classifiedby two categories: visual inspec- measurements are made in other critical metal or
tion and measurements. fluid areas, such as at the gas generator, at the
Visual inspectionsare self-explanatory. They turbine inlet, at the gas storage vessels, and at
consist,oftenafterdisassembly of a component, solid propellant gas spinners, if such are em-
of an inspectionforcracks, fractures, deforma- ployed.
tions,signs of overheating,etc. Surfacedis- Pressure switches.-A typical application
colorationsmay oftenbe a sign of fluidleaks. would be a signal from the engine to the vehicle
In fact, minute amounts of a dye have been added when thrust (chamber pressure) reaches a preset
to a number of propellants such as alcohol- and value during decay to initiate start of the next
the kerosene-type fluids, which leave qualitative stage and separation of the spent one.
evidence at the leak spot. Visua}-checks as- Pressure transducers.-A variety of types is
sume a particularly important role in the event of available: strain gages, capacitance types,
serious malfunctions. As a rule, they do not variable reluctance types, potentiometer types,
require special preparations by the designer, piezoelectric types. Table 10-2 lists, among
except perhaps written instructions (drawing others, important pressures which can be ex-
callouts) and forms, pected to require monitoring and transmission to
By contrast, the second category, measure- the vehicle and its telemetry system.
ments, does require consideration during engine Accelerometers.-It is often important to
design. To make the required measurements, measure the vibrations occuring in various loca-
instruments are needed. Some of these may be tions of the engine system, and how these vary
simple devices such as gages, micrometers, and between static (tiedown) firing and flight. Accel-
torque wrenches, which can only be used before erometers are used for this purpose and also,
and/or after test, and which will not be dis- instead of pressure switches, to sense and sig-
cussed here. However, for the majority of the nal thrust decay for staging.
instruments used for engine design and perform- Tachometers and flowmeters.-Small magnets
ance evaluation, particularly during test, the mounted in a suitable manner to rotating parts of
designer must provide adequate provisions for flowmeters or turbine wheels will induce periodic
installation. Furthermore, the correct execution voltage pips in stationary coils. By counting the
of command signals fed to the engine and the pips as a function of time, turbopump speed or
behavior of the major performance parameters volume flow rate can be determined.
must be transmitted back to the vehicle and from Depending on the type of transducer used, a
there to ground, either by hard wire (static tests) certain amount of signal conditioning will have
or by means of a radiating telemetry system to be provided at or near the transducer. Some
(flight). The instrumentation of a liquid rocket of the latter give relatively weak signals (ther-
propulsion system is a large field, and within the moelements, capacitance-type gages, tachom-
framework of this book we can merely point out eters, and flowmeters), which must be amplified
prior to transmission. In other cases, it may be TABLE lO-2.-Basic Liquid Propellant Rocket
desirable to convert a digital signal into an Engine Measurement List
analog one, prior to transmission (tachometers, (A-1 Stage Engine)
flowmeters). Small though the electronic (solid
Parameter Range
state) signal condition packages may be, room in
Thrust ......................... 0-1 x 106 Ibs
a compatible environment must be provided for Thrust chamberpressure
them on the engine. Under adverse conditions, (injector end) .................. 0--1200 psig
it may be necessary to mount the conditioners on Fuel pump inlet pressure ......... 0-50 psig
the vehicle near the engine. Oxidizer pump inlet pressure ...... 0-75 psig
Some transducers generate their own electric Fuel pump outlet pressure ........ 0-2000 psig
Oxidizer pump outlet pressure ..... 0-2000 psig
signal, such as the thermoelements, piezoelec-
Fuel injection pressure .......... 0-1500 psig
tric elements, and many of the flowmeters and Oxidizer injection pressure ....... 0-1500 psig
tachometers. Most others, however, require an Heat exchanger outlet pressure .... 0-1000 psig
external electrical input which they modify as a Gas generator chamber pressure... 0-1000 psig
function of the quantity sensed. Potentiometers, Turbine inlet pressure ........... 0-800 psig
Turbine discharge pressure ....... 0--30 psig
capacitors, inductances, and resistance tempera-
Stored gas container pressure ..... 0-3500 psig
ture bulbs are examples. Resistance types can
Fuel pump inlet temperature ...... Ambient-100 ° F
use ac or de; capacitance and inductance types Oxidizer pump inlet temperature... Ambient to -300 _ F
require ac. Typical instrumentation power sup- Cooling jacket inlet temperature... Ambient-100 ° F
plies are 88 volts and 5 volts de, and 115 volts Cooling jacket outlet temperature.. Ambient-150 ° F
Turbine inlet temperature......... Ambient-l,500 ° F
ac, 400 cps. Certain transducers with built-in
Stored gas container temperature .. -I00°-+100 ° F
electronics receive 28 volts de, chop, use and Fuel flow ...................... 0-10000 gpm
rectify it, and then return 5 volts (maximum) Oxidizer flow ................... 0-15000 gpm
signals (de-to-de transducers). Strain gages and Turbopump speed ................ 0-10000 rpm
variable reluctance gages are typical examples. Electric bus ,,Jltage ............. 20--35 volts dc
Spark igniter OK ................ On/off
Table 10-2 presents the basic static-firing
Miscellaneous valve positions .... On/off
measurement list for an engine of an A-1 stage.
In practice, particularly during early develop-
ment, this list may be expected to be augmented
by special measurements, requiring additional
thermocouple well proposed by B. N. Bose (ISA
instrumentation such as flowmeters, temperature
Journal. Sept. 1962) designed to offset this
gages, high-accuracy pressure gages, vibration effect.
pickups, and strain gages.
Carbon deposits.-In systems using kerosene-
based fuels, carbon deposits may cause erratic
Instrumentation Installation thermoelectric and grounding effects. While
Some basic design considerations for correct certain electrical connections are possible to
installation of instrumentation end organs (pick- minimize this, regular cleaning and sometimes
ups, sensors) will be briefly discussed. replacement of the probe is required. The engine
designer must consider this need for easy acces-
1. Temperature Sensors sibility of the instruments.
The most widely used temperature sensors are Installation of resistance thermometers or
thermocouples. In rocket engine application, "bulbs" is governed by similar consideration.
thermocouples are subject to certain measure-
ment errors. Among these are: 2. Pressure Gages
Velocity errors.-Fluid friction may cause an Two principal types of pressure measurements
excessive temperature reading, particularly at are usually conducted during rocket engine test-
high fluid velocities. Suitable protection which ing: static and dynamic. As a rule static meas-
does not affect the true measurement may have urements are employed to record, with high accu-
to be provided. racy, steady-state conditions or parameters
Conductive errors.-A heat leak may exist varying only very slowly. Dynamic measure-
from probe to support. Figure 10-21 shows a ments are used where rapidly changing conditions
cant levels. A transducer thus mounted, in com-

bination with a good inking recorder or precision
gage, may have a full span response time of
about 1 second. It would be pointless to con-
nect such a transducer to an oscillograph.
For certain parameters it may be desirable to
combine accuracy with at least some frequency
response. By shortening the line length to, for
instance, 18 inches, a given setup may have a
flat frequency response up to 25 cps. In these
cases it will be the burden of the engine de-
signer to provide for means to mount and connect
the transducer. It should be noted that it is tube
length which affects response, while bends
appear to have a negligible effect. Occasionally
attempts have been made to improve response by
filling the connecting line with a liquid. Apart
from operational inconvenier;ces, severe un-
damped oscillations may result and the practice
is not recommended. Figure 10-22 shows a pres-
sure transducer installed and connected tn engine
pressure taps.
INSULATOR For frequency responses in the range from 100
to 2000 cps, close-coupled mounting of the trans-
ducer is required. Typical transducers, resem-
bling a spark plug, are directly screwed into the
instrumentation boss (fig. 10-23). It is importan_
that the resonant frequency of the combined tap-
and-transducer cavity should be well above that
Figure l O-2I.-Properly designed thermocouple

well (installed in a 4-inch pipe).
(transients) must be recorded, with rapid response

but at some sacrifice in accuracy.
Because of the higher accuracy usually de-
sired for static measurements, the sensitive
transducers are often mounted in environment-
controlled boxes to protect them against vibra-
tion, temperature (both ambient and of the meas-
ured medium), and other influences. They are
therefore usually mounted some distance away
from the engine on the test-stand structure or on
the vehicle, at the end of several feet of tubing, Figure I O-22.-Pressure transducers installed and
which reduces frequency response below signifi- connected to engine pressure taps (lower right).
of the frequency range of the measured param-

eter. Figure 10-24 shows two typical installa-
BRAIDED CABLE
tions and associated equations to determine in.
Figure 4-28 shows a typical tap arrangement for
injector-end chamber pressure.
For still higher frequencies, above 2000 ops,
transducers are installed so as to leave essen-
tially no cavity at all (flush mounting). This
poses problems with hot-gas measurements.
Special water-cooled transducers have been de-
veloped for these applications. Needless to say,
an effort toward high-frequency response at the
pickup point is lost if the recording system is
i INNER SURFACE
not capable of using it.
Flowmeters
Turbine-type flowmeters appear best suited to
rocket engine installations. They are well de-
Figure lO-23.-Typical installation of a thrust veloped and, as a rule, can be mounted in a sim-
chamber pressure transducer. ple fashionbetween availableflangesin existing
._-----TRANSDUCER T-T /---TRANSDUCER
F 3 I I1 !l
L/.---DIAPHRAGM
BOSS_
ADAPTER -_/
_.--ENGINE
COMPONENT
_--- DIAPHRAGM
ENGINE COMPONENT WAll
WALL
TRANSDUCER INSTALLATION TRANSDUCER INSTALLATION
JL-s _ ;
fn =
2"/r
WHERE :
WHERE: C = ACOUSTIC VELOCITY
n -" ANY WHOLE NUMBER (I,2,5...etc)
S = CROSS-SECTIONAL AREA OF TUBE IN SQUARE IN.
n = I FOR THE FUNDAMENTAL FREQUENCY
V = VOLUME OF CAVITY AT THE END OF THE TUBE
C = ACOUSTIC VELOCITY OFTHE MEDIUM IN CUBED IN.
03,550 IN./SEC FOR AIR AT 68 ° F}
IJ= EFFECTIVE LENGTH OFTUBE IN INCHES
L= LENGTH QFTUBE iN INCHES
I_= L +l.7a
WHERE: L: ACTUAL LENGTH
o=RADIUS OF TUBE
ACOUSTIC EQUIVALENT ACOUSTIC EQUIVALENT
(A) {B)
Figure 10-24.-Acoustic equivalents of pressure transducer mounting cavities.

For both large or small units, reliability is not a

fixed value, but is a function of development
"_"_ ELECTRIC PICKUP time. A realistic analysis considering all fac-
tors may determine the appropriate choice. The
absolute truth, however, may never be known,
since the cost of going both routes for a given
mission and of comparing final scores would be
prohibitive.
Clustering of liquid rocket engines or, in
other words, subdivision of vehicle thrust into
FL smaller units may be accomplished in a number
of ways. The choice for the most part will be
based on vehicle considerations. Depending on
the method chosen, however, engine design will
be directly affected.
ROTOR
Some of the basic considerations of this topic
Figure I O-25.-Turbine-type [lowmeter. which are specifically related to engine cluster-
ing and as they affect engine systems design and
integration, are briefly reviewed.
ducting (fig. 10-25). Turbine flowmeters are very There are no hard-and-fast rules which can be
sensitive to changes in upstream duet configura- applied to determine optimum cluster configura-
tion. Once calibrated as installed, however, tions. With the aid of liquid engine clusters
they exhibit a high degree of precision. actually used in practice, however, we will pre-
sent a number of considerations which may serve
Accelerometers and Strain Gages as a guide for future designs.
These are almost always applied externally,
in connection with special measurement pro-
grams. At the time of engine systems design, no Earlier Cluster Configurations
special consideration is usually required. Figure 10-26 shows the cluster of two experi-
mental LOX/alcohol engines for a planned but
discontinued ballistic missile. It consisted of
10.8 CLUSTERING OF LIQUID PROPELLANT
two units of 120 000-pound thrust each. Both
ROCKET ENGINES
subunits included a tubular-wall, gimbaled thrust
The idea of obtaining higher thrust levels chamber, and a geared turbopump. The turbines
expeditiously through combination of several were powered m parallel from a single, common
smaller rockets probably is as old as rocketry gas generator. The control system was also
itself. To today's liquid rocket vehicle and common for both units. Thus this propulsion
engine builders the topic of clustering still system was not a true cluster, since it was not
causes spirited debates, which reached a peak possible to develop and fire each unit independ-
in March 1952 when Collier's published an arti- ently. Several successful experimental flights
cle by Dr. yon Braun in which a space vehicle were achieved with this engine.
was described as having a first stage powered Figure 10-27 shows another experimental
by a cluster of 51 turbopump-fed rocket engines. cluster consisting of three units which, although
The debates invariably hinge around the deceiv- never flown, achieved a remarkable reliability
ingly simple question: is higher overall relia- record during static firings at thrust levels up to
bility obtained through combination of a number 500 000 pounds. This cluster, too, used a com-
of well-developed smaller engines, or through mon control and gas generating system for the
application of one large, specially developed three subunits. Most of its components were
unit. The answer is not simple. The failure essentially the same as those used for the two-
probability of the smaller units may be low, but engine cluster, except that propellants were
it increases with the number of units clustered. RP-1 and liquid oxygen.
"
Figure 10-26.-Dual-engine cluster. Figure 10-27.-Three-engine cluster.
Still another approach for ballistic missiles zation of thrust and total impulse (propellant
is the combination (clustering) of a pair of booster load).
engines with a large-expansion-area sustainer To examine some of the considerations which
engine. During flight, the booster engines may govern engine cluster arrangements, we assume
be jettisoned following the boost period, while that payload and mission of a multistage vehicle
the sustainer continues, fed from the same tanks. are defined and that stage propellant loads (total
These configurations are also referred to as impulse) and tank geometry (diameter) have been
"One-and-one-half- stage vehicles." optimized. We further assume that engines are
available and will not be redesigned, and that
the optimum thrust for a typical stage suggests a
Recent Cluster Design Trends
range of from four to six engines. An analysis
The first clustered engine vehicle, signifying has shown that payload performance probably
the modern trend toward multipurpose engines, is increases with the larger number of engines (fig.
NASA's S-1 first-stage booster for the Saturn I. 10-28); however, there are other considerations.
It is powered by 8 LOX/RP-1 engines, the basic One of the most influential considerations is
elements of which were transplanted with rela- "Engine Out Capability" for increased vehicle
tively minor modifications from earlier engines. reliability (also see ch. II, "Thrust Level").
The S-1 is noteworthy for its tank arrangement. Specific engine provisions for engine-out (EO)
Eight tanks from an earlier missile are clustered are summarized in section 10.9. Capability of a
around a central larger tank. This "multicellular" vehicle to complete its mission with one engine
design, which in the case of the S-1 permitted inoperative always entails some performance
early availability of large tank capacity without losses. However, these losses decrease with
major retooling, has been recommended by some increasing total number of engines, because less
vehicle designers for still larger vehicles. The performance reserve is needed, and because less
eight engines of the S-1 are grouped into four correcting thrust vector trim is required from the
fixed inner and four gimbaled outer engines. operating engines. Trim from engine-out is
Except for the outer-engine actuators, all eight further affected by the diameter of the engine-
are identical, independent units, built and tested mounting circle. The contribution to mission
singly, and combined for the first time on the reliability by EO capability is substantial, with
vehicle proper. cluster failure potential reduced by more than
In the S-1, the number of engines was, no half. In addition, absolute reliability values are
doubt, almost entirely governed by the availa- a function of the number of engines in the cluster.
bility of existing major engine components and This is most evident with no engine-out.
tankage. For subsequent vehicles, such as the Figure 10-29 illustrates this at two points of
Saturn V three-stage vehicle, greater freedom time of engine overall development status. As
existed, particularly with respect to the optimi- the latter progresses, the difference between
load for a substantially higher cluster reliability

at any single-engine reliability level.
NO ENGINE D_JT
Another important consideration is cluster
diameter in relation to vehicle diameter. As witl
be discussed in section 10.9, it is mandatory
that all engines fit into the interstages from
which they must separate without interference.
This restriction is not valid for the first stage,
where engines may protrude beyond the tank pro-
J _- 0NE ENG,NE OUT file. The degree of p['otrusion, however, must be
/ balanced against drag losses.
/
Specific considerations for each chosen num-
L I i
4. 5 6 ber of engines follow (fig. 10-30).
NUMBER OF ENGINES
Figure 10-28.-Stage-payload weight as a function &Engine Cluster

of number of engines in cluster. This configuration provides the lowest total
thrust in the established band of 4 to 6 engines,
but it is optimum for engine interehangeability.
..____E_IEONE ENGINE OUT ( ADVANCED ENGINE
All four engines will be gimbaled. Only one
1,0
LOPMENT PHASE ) installation is required: all inlet ducts and other
engine to stage connections can be made id.. _ti-
NO ENGINE OUT _ ADVANCED ENGINE -'-"--
cal. The thrust vector control mode is simplest:
all engines deflect in pitch and yaw, and all
actuators participate equally in roll control (fig.
10-80(a)).
0.9--
F-
08
YAW -- --0_
ACTUAT
I
a. b.
E I i
4 fi 6
NUMBER OF ENGINES
Figure 10-29.-Typical cluster reliability predic-

tion as a lunction ol number ot engines in
cluster and of development time.
engine-out and no engine-out narrows (see also

table 2-2). It may be stated that for a given
number of vehicle flights, the benefits of engine-
out capability are the greater, the lower the
single-engine reliability. Figure 10-28, in com-
¢. _L
bination with figure 10-29, shows that simulta-
neous addition of one engine and inclusion of Figure lO-30.-Typical engine cluster arrange-
engine-out capability retain about the same pay- ments.
5-Engine Cluster and systems thrust vector control response is

This number provides greater thrust, but com- seriously affected, due to roll and yaw coupling,
plicates matters in other respects. Two basic particularly under malfunction conditions. As to
arrangements .are possible, as shown in figure mounting-circle diameter, the two 5-engine clus-
10-30(b) and (c). The arrangement with center ters are about equal.
engine is preferred. The center engine is mounted
fixed; i.e., stiff arms are installed in place of 6-Engine Cluster
gimbal actuators. It is advisable to install the This combination provides the highest thrust,
engine with complete gimbal bearing for stand- at the expense of a larger mounting circle. This
ardization and ease of alinement. With the re- in turn increases the effects of one-engine-out
maining four engines gimbaled and grouped as in on trim required and may compound interstage
a 4-engine cluster, the same benefits are derived and interengine clearance problems. The control
as with the latter. However, because of the mode is comparable to the four engine cluster in
center engine, two installations are required. simplicity. The 6-engine cluster requires two
Furthermore, two sets of inlet ducts are needed, installations, with inlet and other effects similar
their different length affecting pressure drop and to those discussed with the 5-engine configura-
fluid velocity profile (engine turbopump NPSH), tion.
water hammer from closure of valves (valve tim- This cluster offers the potential of removing
ing), trapped propellants, and possibly insula- (not installing) 2 engines and still retaining a
tion requirements (weight penalties and complex- satisfactory 4-engine combination, which pro-
ity). Figure 10-31 shows a typical 5-engine vides added flexibility at minimum scar weight
installation with fixed center engine. (i.e., weight of components that cannot readily
With all five engines arranged on a circle, the
be removed along with the engines).
basic engine package may become more uniform, If at all possible, the engine designer should
since all will be gimbaled. However, three dif- specify that the engine intended for cluster use
ferent installations are required (fig. 10-30(c)), be tested under conditions closely simulating
vehicle installation. It may be expected that the
vehicle builder will conduct a firing program of
his own; however, this must be devoted to clus-
ter behavior and performance evaluation, and
should not deteriorate into continued single en-
gine development. A major difference between
single engine and cluster firing is in the base
conditions, i.e., the heat and pressure environ-
ment of the engines, particularly at altitude.
Flame radiation effects, backwash of combustion
gases, and impingement may create much more
severe conditions than are present during single-
engine firing. The pressure environment pro-
duced by several engines firing together may
create moments on the engines which must be
accounted for in the design of the gimbal sys-
tem. The engine designer should be familiar
with these conditions so that his claim that the
engine can be clustered remains valid after
delivery.
ENGINE Awareness by the engine designer of the con-
{FIXED) siderations governing engine clustering, as pre-
Figure ]O-3].-Typica] five-engine cluster con- sented here, will enable him to complete his
figuration. Center engine fixed, four outer systems integration with a broader view to
engines gimbaled. application.
10.9 ENGINE-TO-VEHICLE INTERFACE vehicle builder is in need of numerous engine

design details in support of his stage design.
Throughout this book, references have been
An excellent source for this infonnation is an
made to the ultimate purpose of rocket engine
"Engine Design Manual." It should be started
design and production: propulsion of a vehicle.
during engine initial design and should be aug-
Some of the principal engine-to-vehicle inter-
mented as the design and, at a later date, the
faces, such as thrust mount, and pneumatic,
development progresses. Necessary revisions
fluid, electrical and propellant lines, have been
must be disseminated promptly.
discussed in preceding chapters. In this section,
Equally important is the early generation of
we will summarize a number of vital engine-to-
an Interface Control Document. This defines
vehicle interfaces, of which the engine designer
each interface, for both mating sides, on one
should also be cognizant.
drawing.
Design Documentation Space Envelope
Adequate mechanical design data, vital for When installing an engine in a vehicle, it is
the physical integration of the engine into a not only necessary that it can be properly bolted
vehicle system, must be properly documented by to the vehicle thrust mount, but it is equally
the engine systems designer. The following data important that no other vehicle parts interfere.
are considered minimum requirements: Space in the vehicle engine compartment must
(1) Engine system general arrangement draw- not only accommodate the engine envelope when
ing.-This drawing defines the engine in the neutral position, i.e., pointing straight aft,
space envelope and the locations and but when fully deflected in all directions as
detail of various agreed-upon vehicle well. Typical maximum engine deflections range
connect points, such as from 4 ° to 7 ° . For clusters, moreover, the possi-
Thrust or gimbal mount bility of faulty deflection of engines must be
Gimbal actuator attach points taken into account. Since space in the engine
Fuel and oxidizer inlet flanges compartment of vehicles is usually limited, the
Hydraulic and pneumatic system con- need for the closest cooperation between engine
nections and vehicle designer becomes apparent. This is
Electrical and instrumentation con- especially true for upper stage clusters where
nections large expansion area nozzles must be housed in
(2) Mechanical dimensions, tolerances, seals minimum interstage structures.
(if any), fasteners, and loads at the

vehicle connect points listed
Connect Panels
(3) Engine system mass properties, as are
shown in figure 2-6, which include en- All engine-to-vehicle lines, which may amount
gine weights, gimbaled mass, center of to a dozen or more per er_ ine, must be reliably
gravity, and moment of inertia for the connected for each engine installation, not only
basic engine, including accessories in the vehicle but in the static development and
(4) Engine performance data (as in fig. 3-1) acceptance firing stands as well. It has been
(5) Engine functional description (as in ch. found beneficial to define this interface clearly
III) by combining all lines in one or several terminal
(6) Engine handling procedures and equip- connecting panels (fig. 10-32). These panels,
ment, needed for installation and main- uniformly designed for all test locations, may be
tenance mounted on the vehicle (or test stand), or on the
(7) Engine servicing needs engine. For certain installations it may be ad-
As a rule, this information is compiled in vantageous to have matching panels on both:
detailed handbooks, which will accompany the vehicle (or stand) and engine. Alined a short
engines when delivered to the vehicle contractor. distance apart, standardized jumper lines be-
However, several years before the engines are tween pairs of panels will permit rapid and reli-
delivered and handbooks become available, the able connections. A possible disadvantage of
TEST STAND
,ul
STRUCTURE TA,
ENGINE CONNECT
Lax IEL_O_--
PANEL .o= pu_ -- - c, _, _,.o_
Lox ,_c,c_,-
Figure 10-33. -Closed-loop coupling of propulsion

system parameters and longitudinal vehicle
*: ,= 'q
mode.
acceleration, as a function of thrust, a powerful

feedback exists. It has been demonstrated that
the system can be detuned or uncoupled, through
manipulation of propellant duct volumes, stiffen-
Figure lO-32.-Typical line connections on an
ing of structural members, and by other means.
experimental liquid rocket engine.
Itis difficultif not impossible for the engine
designer to predict analytically and prevent this
the panel method is that routing a line through problem, particularly since the vehicle configu-
the panels may result in additional line length, ration may not yet be firm. He may be called
as compared to individual routing of each line. upon later for changes, however, if the problem
Through careful design and close coordination shows tipduring flight.
between engine and vehicle builder, considering
optimum location, subdivision and orientation of
Malfunction Reaction Systems
the panels, advantages will, in most cases, far
outweigh disadvantages. In chapter II we have discussed reliability
and failure modes, as well as certain preventive
measures. In the framework of engine-to-vehicle
Dynamic Interactions
interfaces, some additional detail follows.
The close coupling between propellant feed

system, propulsion system, and vehicle structure Engine Failure Sensing and Shutoff System
may lead to undesired interactions. These have (EFSS)
manifested themselves as longitudinal vehicle These systems have been in use since the
oscillations (also referred to as "Pogo Stick early beginnings of liqmd rocket engine applica-
Effect," "Accordion Motion," or "Spring-Mass tion. They are specially important during engine
Effect") at frequencies from approximately 10 to
20 cps. They have led, in at least one instance,
to vehicle destruction. The relationship of the BURNOUT
<..j..1
contributing parameters is illustrated in figure
10-33 for a typical single-engine vehicle. The
analyses and correction of these oscillations are
extremely complex, particularly since they can-
not be reliably reproduced during captive firings.
FLIGHT TIME
It is believed, however, that any one of the con-
tributing factors can modulate thrust at a fre- Figure IO-34.-Typical pump inle$ pressure vari-
quency at which the vehicle is resonant at cer- ation of a vehicle affected by longitudinal
tain tank levels (flight time; fig. 10-34). Through oscilla$ions.
and early cluster development. Part of it may Heat Protection

later become a portion of flight emergency detec-
Rarely in our technology had lowest and high-
tion systems (EDS) and engine-out systems (see
est temperatures to be handled so close together
ch. II). The following list of major engine-
as in a liquid propellant rocket engine. Earlier
oriented malfunctions, which is not necessarily
we discussed the need for insulation, to main-
complete, should be considered:
tain temperatures as low as -423 _ F in certain
ducts, or for heaters to protect sensitive compo-
Pump inlet pressures below safe minimum
nents against these temperatures. At the same
Turbopump overspeed
time, in other areas at or around the engine, pro-
Turbopump bearing overheating
tection must be provided against the very high
Excessive turbopump leaks
temperatures of the combustion process and the
Turbine gas overtemperature
emerging gas jet, such as cooling of the thrust
Combustion instability
chamber. The exhaust jet, at sea level, usually
Abnormal injection pressures
Ignition failure is not a major problem, unless blowback occurs
from the flame deflector. At higher altitudes
Premature propellant depletion
with vacuum or near-vacuum pressures, however,
Electrical power failure
Pneumatic pressure failure which are experienced even by first stages for
the last portion of their flight, a substantial
Improper valve positions
Fires portion of the thrust chamber gas jet expands
sideways from the nozzle exit, forming a plume,
Because of their potential sudden destructive
creating considerable backwash and radiating
effects, many of the malfunctions will be sensed powerfully back into the engine compartment.
and the signals used to initiate immediate auto- This endangers both engine and vehicle compo-
matic cutoff. For others which would not create
nents. Surface temperatures of 1000 ° F or more
an emergency within, say, fractions of a second, may result unless heat protection is provided. In
it is not uncommon to simply mark their allow- some cases it may be too cumbersome to provide
able minimum and/or maximum values on an ink-
individual insulation for each component. Also,
ing recorder chart which is watched by an ob- excessive weight penalties may be incurred. A
server during test. If the recorder needle goes protective heat shield, forming a closeout dia-
outside of the marked margin, the observer initi- phragm, may then be more effective (fig. 10-35).
ates cutoff manually. This shield may be supported from a stationary
For flight, a few selected highly critical (center) engine, if available, or from a supporting
parameters of the EFSS may be retained for auto- structure.
matic engine cutoff or mission abortion. In this
case, if the engine failure sensing and shutoff
system were called upon to shut down an engine,
and the flight is continued with the remaining
engines, a vehicle-mounted electronic logic must
sense the shutdown and take certain actions.
These may include:
A
Closure of emergency shutoff valves in the
FLEX
ducts to the defunct engine, but not those in CURTAIN _
the others.
Disconnection of electric power to the defunct
engine only. SECTION
A-A
Resetting or disarming of backup cutoff tim-
FULl. CLOSE-OUT PARTIAL CLOSE-OUT
ers, since the reduced number of engines (CURTAINS OPTIONAL)
will consume the available propellants over

a longer period of time. Figure l O-35.-Typical base beat protection con-
Locking of the defunct engine in the neutral cepts (center engine [ixed, outer engines
gimbal position. gimbaled).
Design and installation of the base heat

shield requires closest cooperation between
engine and vehicle designer. It must be deter-
mined, through special model tests, whether a CHANh
partial or a full closeout is required. In most

designs the shield connects to the engine noz-
zle. Suitable brackets must therefore be pro-
REFRACTORY
vided. During gimbaling the heat shield will SHI
resist the engine motion. The forcesencountered
must be considered in the power budget for the CHAMBER
hydraulic actuators. Here it must be taken into TUB ES
account that the heat shield flex curtains may be

INSULATI(
quite cold prior to engine start and still remain
relatively cool at the far side, as planned, during
stage operation.
In addition to backwash and radiation, which
are generally not harmful to the internally cooled
engine nozzle itself, mutual gas jet impingement
CHA_
between engines may occur from extreme deflec-
tion during gimbaling. Unless a major control
malfunction occurred, this should affect the noz-
zle only for very brief periods near the exit.
However, the heat may affect chamber structural
elements, such as stiffening bands, which are EXIT RING AND FLANGE
not internally cooled. A heat-protective strip, a
few inches wide, of ablative or other suitable Figure JO-36.-Center-engine flame impingement
material applied to the nozzle should suffice in shield.
most cases (fig. 10-36). Obviously, this type of
protection is not required for single-engine
vehicles. lion mechanisms, can function properly only if
they are within a specified temperature range.
Engine systems not employing cryogenic propel-
Engine Prestart Conditioning and Start
lants may be dependent on heating or cooling
Functional conditions required within the only if prolonged coasting times in space are
engine to assure its readiness to start have been involved. Engine systems which do use cryo-
discussed earlier. When installed in the vehicle, genic propellants almost always need at least
the engine also requires external conditions some heating. The cryogenic propellants within
which must be met by the vehicle builder. Pro- the engine following start of tanking, and the
vided that prelaunch checkouts have ascertained heat absorption of tank surfaces and lines may
readiness with regard to absence of leaks, cor- rapidly lower the air temperature surrounding the
rect valve positions, etc., prestart conditioning engine to several hundred degrees below 0 ° F.
of the engine essentially refers to temperature Since most vehicle systems will specify an
and pressure levels around the engine and at the allowable hold period following tanking to allow
pump inlets. For both, early cooperation be- for adjustments, checkout of other systems, to
tween engine and vehicle designer to arrive at a wait for optimum launch times (rendezvous mis-
mutually feasible solution is essential. Without sions), etc., severe subcooling of engine compo-
it, the optimistic note in engine drawings "to be nents may occur.
supplied by vehicle contractor" will accomplish The temperature environment can be substan-
little. tially improved by the vehicle builder through
Certain engine subsystems, such as hydraulic engine compartment purges with warm gases (fig.
components, control systems, and valve actua- 10-37). If an inert gas is used, this has the
80
80
4C _ _- SKIN BELOW THRUST MOUNT /--SKIN BELOW THRUST MOUNT
40 WL,._--SKIN At3OVE MOUNT
_"
-40
(_ _ /_-SKiN ABOVE THRUST MOUNT
g o \
_:_ j_,,_.NITROGEN AMBIENT E'XTERN_L TO MOUNT
-40
_.-ao
I'-
-120
/_ N_TROGEN AMBIENT EXTERNAL TO MOUNT I.U_''1l_1 _%,_ _--ENGINE TEMPERATURE
F- -80
-160 _ ENGINE TEMPERATURE _. "N_'-THRUST MOUNT TEi'IRERA'FURE
i(
-200 _ I_THRUST MOUNT TEMPERATURE _NITROGEN AMBENT INTERNAL TO MOUNT
m_-NITROGEN AMBIENT INTERNAL TO MOUNT

-16( I I I l l I
i I i
-2,o. , 2 , _, , _ , _, , ,_ ,2 2 4 6 8 I0 12
HOLD TIME FROM PROPELLANT LOADING(HRS) HOLD TIME FROM PROPELLANT LOAI_NG (HOURS)
Figure 10-37.-TypicaI interstage temperature en- Figure 10-38.-Typical interstage environment for
vironment for an uppei space vehicle stage an upper space vehicle stage using cryogenic
using cryogenic propellants. 565 SCFM, propellants. 4170 SCFM, (260 ° F) nitrogen
(-I00 ° F) nitrogen purge. purge.
additional advantage of counteracting the accu- be expected to maintain their temperature within
mulation of combustible gas mixtures from minor an allowable band after vehicle liftoff during the
leaks. However, the vehicle builder's possibili- relatively short boost periods. Only stages with
ties of heating through use of compartment gas prolonged cruising times may require an onboard
purges are definitely limited. Some of the limita- purge supply system which imposes payload
tions stem from the engine designer's own speci- penalties.
fications, which require avoidance of elevated
temperatures around components containing cryo-
genic fluids to minimize boiloff and to prevent
formation of gas bubbles. Also, certain struc-
tural members must be kept below maximum tem- UPPER STAGE, LOWER TANK BOTTOM
peratures because of their strength characteris-

tics. Another limitation is established by the
purge gas requirements and necessary heating _ (LIQUtO OXYGEN ] /
provisions which would become prohibitive if it

were attempted to raise compartment tempera-
tures to above 32 ° F or even to 0 ° F. Figure INTER-
STAGE
10-38 shows the effects of increased purge flow
and temperatures over those of figure 10-37. The
analyses on which the graphs are based have
shown that a further increase of the purge rate to
12 000 SCFM (standard cubic feet per minute)and l ..... _ \ /ENG,
NE\ / ENO,
NE
250 ° F barely raises the interstage temperature
above 0 ° F.
At the same time the propellant boiloff rates
in exposed ducts increased tenfold over those
under conditions of figure 10-37. The locations
quoted are identified in figure 10-39.
For most applications the purges to the vari-
ous areas are supplied from a ground source.
Once preconditioned, affected components can Figure 10-39.-Multistage vehicle interstage.
Literally a few inches away from the compo- conditions, and remoteness from ground stations
nents requiring protection against low tempera- can only crudely be simulated in ground tests.
tures are others which must be protected against Therefore, until first flights of a new vehicle
warmup due to influx of heat. Here again non- have been accomplished, the engine in-flight
cryogenic propellants pose few problems. By start sequence, which is closely interwoven with
contrast, cryogenic systems, in particular those the stage separation sequence, must be planned
including liquid hydrogen, are very sensitive to based largely on analytical work. Neither ve-
heat influx. The effects on pump NPSH, which hicle builder nor engine designer can do this
includes the vapor pressure of the fluid pumped, without consideration of, and consultation with
were discussed in detail in section 8.5 in con- the other. The following principal relationships
nection with tank insulation. For engine start, must be considered:
which involves the cryogenic fluids stored in the Starting of the upper stage engines while the
ducts immediately upstream of the engine and lower stage is still at full-thrust level is diffi-
within the engine itself, insulation may become cult, mainly because of the problems of flame
prohibitively complex. Continuous removal of impingement and exhaust gas dissipation from
superheated cryogenic fluids from the engine the interstage. Therefore, engine start following
system and substitution of fresh liquid, there- stage separation is preferred. However, in this
fore, has been widely used. This can be accom- case propellant settling from acceleration effects
plished through continuous overboard bleeding or of the lower stage is no longer assured. Auxiliary
through recirculation back to the tank. In both propellant settling (or "ullage") rockets (usually
cases, the liquids are tapped off at a point far- solid propellant type) are required instead during
thest away from the pump inlets and routed to the engine starts. These add weight and drag and
stage for overboard dumping or return to the reduce payload. The size of the weight penalty,
tanks. A small auxiliary pump is often required in turn, depends on engine-start characteristics
to obtain the minimum flow rates required for (thrust-buildup time). For engines with relatively
adequate bleeding. protracted buildup times, advanced initiation
During engine development, and even during (with respect to lower stage thrust decay and
vehicle static firing programs, relatively favor- separation) of upper stage engine start may sub-
able prestart conditions exist: the enclosing stantially reduce the penalties. A secondlopped
interstage is not in place, hold periods are under off the ullage rocket burning times may save
better control, cold lower stage tank surfaces hundreds of pounds of stage weight. The same
are absent. There is thus the danger that what amount of time added may not only reduce pay-
worked faultlessly under these conditions will load weight but pose serious control problems to
fail in vehicle flight. A realistic recognition of the space vehicle which floats essentially "rud-
this situation and close coordination between derless" after separation and prior to the start of
the two design teams in the early phases of its own engines. From the foregoing the need
engine and vehicle design and development is for thorough mutual understanding of this impor-
therefore vital. tant vehicle-to-engine interface becomes apparent.
For first- stage applications, the engine builder
is relatively independent regarding start charac- Vibration Environment
teristics and sequence. The vehicle will not
take off until thrust exceeds vehicle weight, Even a normally or "smoothly" operating
gravity holds the propellants at the tank outlets, rocket engine generates a vibration pattern which
and all parameters are closely monitored from the is transmitted mechanically through the thrust
ground control center. If a holddown period prior mount, or acoustically through the ambient at-
to vehicle release and takeoff is included, an mosphere, if present, to the vehicle structure.
additional safety margin exists. Neglect of these vibrations by the vehicle builder
For upper stage engine start, conditions be- may lead to serious structural weakness or mal-
come considerably more complicated. Absence function of vital control organs mounted to the
of the propellant-settling effects of gravitation, vehicle structure. Means of reducing the normal
the surrounding vacuum, altitude temperature engine vibrations are limited and hard to analyze.
Rather, itis important that the engine developer interstage length will cost approximately 40
establish as accurately as he can the existing pounds. Thus, here too, the added interstage
acoustic and vibration pattern and inform the weight required to accommodate a longer, better
vehicle designer of his findings. Since the tie- performing nozzle may void the theoretical gains.
down firing measurements cannot reliably predict
free-flightconditions, a serious handicap exists. Stage Separation
However, through his thorough understanding of When separating the stages of a space vehi-
the potential problems, and through specification cle, three basic possibilities exist:
in cooperation with the vehicle designer of static 1. Leaving the interstage with the lower
test and flight measurements, the engine designer stage.-This requires the engines to pull out of
can materially assist in reducing the incidence the interstage cylinder during separation (fig.
of serious problems. 10-40). If the clearance between engine skirts

and interstage wall is marginal, complicated
Nozzle Expansion Area Ratio means may be required to avoid collisions be-
tween the separating stages. These may include
From his own analyses, the engine designer
a control system to tuck the engines inward, and
may have selected a nozzle expansion area ratio
then swing them out upon separation and start
which appears to be an optimum compromise be-
them. Any such system adds complexity, lowers
tween engine weight, engine performance, engine
reliability,and may add weight. A shorter noz-
size, and producibility. When analyzed in the
zle skirt, of lesser diameter although somewhat
larger framework of vehicle performance, the
lesser performance, may be better overall.
selection may turn out not to be optimum. In
2. Leaving the interstage with the upper
reviewing the various influences, we will disre-
stage.-To avoid the problem of stage collisions,
gard single-engine firststages which are least
it may appear attractive to leave the interstage
affected. For stages with engine clusters, in
attached to tileupper stage. However, this adds
particular, upper stages, the following must be
inert weight to the upper stage. This penalty
considered:
may be considerably larger than the performance
loss from a somewhat shorter nozzle. Also, a
Gimbal Angle
control problem may be incurred in that the en-
Following determination of the maximum gim-
gine nozzles, in their outward gimbal deflection,
bal angle required for safe vehicle guidance,
stillmust clear the interstage wall. In joint
including consideration of actuation malfunc-
optimization studies between engine and vehicle
tions, the nozzle exit diameter will determine
designer, the possibility of leaving part of the
the attachment point distances (mounting pattern)
interstage with either stage may be considered.
of the engines in a cluster.
3. Leaving the interstagewith the upper stage
but dropping it in a second separation maneuver
Vehicle Diameter
several seconds a/ter first separation.-In this
When mounted for proper gimbal capability,
dual-separation sequence, the second separation
the envelope of the engine cluster must be in
may consist of shedding the interstage as a com-
proper relationship to the vehicle diameter. If
plete ring, possibly in combination with guide
the cluster envelope is too large, the increase in
rails, or of blowing it off to the side in segments.
vehicle dimensions may void all gains from a
For either method, some form of actuation is
larger expansion area ratio. If engines are not
required in addition to accurate timing. Also,
permitted to protrude beyond the projected vehicle
while the interstage is still attached, a serious
plane in neutral and/or in gimbaled position, an
base-heating problem may develop. The means
even more stringent situation exists.
required to overcome any difficulties with this
separation method again may increase complexity
Interstage Length
and reduce reliability and partially neutralize
When mounted to the lower stage, the length
nozzle performance gains.
of the upper stage engines will determine the
length of the required interstage. For a vehicle Whichever method is applied, the thrust-decay
of the Saturn V class, each additional inch of characteristics of the spent lower stage engines
Shutdown of lq_er stage engines (from propellant depletmn signal)
I 1
Upper stage ullage rockets ignited
I I
Engine start sequence initiated
I
Separation timer started
Physical separation begins

(from timer or accelerometel
1
Lower stage retro rockets f_e
• Upper
I
stage engine nozzles clear interstate
I
• Tuner allows engines to gimbal
• Engines i_nlte
• Engines reach
full thrust
t
Tm_e (Seconds)
Figure 10-40.-Typical stage-separation sequence.
greatly affect the separation sequence and the stands and for various servicing functions, the
clearances. To optimize these, the vehicle engine builder requires numerous fixtures com-
builder needs from the engine designer accurate monly referred to as engine GSE (ground support
information on thrust-decay characteristics and equipment). Some of these fixtures can be used
tolerances. This may be in the form of a graph, by the vehicle contractor as designed, others
as shown in figure 10-41. could be used if only a few minor changes or
additions were made. The funding required for
GSE in a typical vehicle program is substantial
Flame Deflector Size
and approximates that of the flight hardware.
Although this is not entirely an engine-to- Common use of as many of these fixtures as pos-
vehicle interface, it is pointed out that engine- sible by engine and vehicle builders is strongly
nozzle size and arrangement on the vehicle, advised. To assure this the two must work to-
together with maximum gimbal angles, strongly gether from the earliest design inceptions.
influence the design of required flame deflectors Standardization
on engine and stage static firing test stands, and
of launcher flame deflectors. Here the problem The task of mating engine and vehicle will be
is not so much one of optimizing vehicle per- substantially facilitated if the designers of both
formance, but of minimizing cost and of assuring work to the same standards. This applies to
the adequacy of these indispensable development national standards, Government standards, com-
tools. pany standards, to terminology, mathematical
symbols, and to the measuring system (metric or
English, decimals or fractions, tolerances).
Engine Handling, Installation, and Servicing
Since several avenues are open, it is necessary,
Fixtures
at the very outset, to agree on which one shall
For handling engine parts and assemblies be pursued. This requires close cooperation and
during engine installation into development test full documentation.
too
\
\
\\ \
\ \
\ \
7s \
\ \
\ \
ILl \
I&.
so
o [ t I
0.S 0.6 0.7
CUTOFF TI_E SEC ONDS
SIGNAL
Figure 10-41 .-Typical engine thrust decay deviations.
Delivery Schedules Maintenance and Logistics
For the vehicle builder to deliver his stage Until reliable and economic methods are de-
on time, complete and equipped with engines, the veloped to recover and refurbish rocket vehicles,
latter must be available to him with sufficient rocket engines will be used only once in flight,
leadtime. If the engine builder is directly con- which will be preceded by a reasonable number
tracted by the vehicle contractor, this can be of checkout and acceptance firings. The engine
negotiated and followed up in a straightforward model specification will include the total pre-
manner. In most cases, however, particularly scribed run capability (lifetime) of the engine,
with large liquid engines, these will be supplied usually expressed in multiples of the rated flight
to the vehicle contractor as Government-furnished duration. A factor of, for instance, 6, which
equipment (GFE). The engine supplier, there- includes a reserve for repetition of checkout
fore, may not necessarily be fully cognizant of runs, is typical. Because of the stringent weight
the detail vehicle needs. In either case, the considerations in rocket vehicles, engine design
engine designer and developer, to avoid sudden attempts to assure this life expectancy, but no
unexpected compression of his schedules, must more. Although reasonable safety factors make
make sure that his schedules for design, drawing it most unlikely that the engine would completely
release, manufacture, development, and engine collapse shortly after the allowable maximum run
delivery are in accord with the vehicle needs. time is exceeded, the statistical probability of
some engine part failing does increase. For this ing blocks, such as turbopumps, injectors, com-
reason most engine specifications prescribe bustion chambers, gas generators, valves, gimbal
overhauls of a stated scope whenever this limit blocks, and high-pressure ducts. Peripheral
is exceeded. equipment, such as instrumentation lines, servic-
Similarly, upper limits are set for storage ing lines, inIet ducts, and wire connections,
times. If these are exceeded, routine replace- should be left flexible enough to adapt them to
ment of certain engine parts will be made, nota- each vehicle without compromise. It is tempting
bly of gaskets and seals. Rocket engine preser- to procuring agencies to warehouse just one
vation and packaging methods are well developed, engine model and ship "from the shelf _ to wher-
permitting storage periods from 3 to 5 years ever the need arises. This convenience, how-
without overhaul. ever, may cost substantially more than the ex-
During shipping, handling, installation, servic- pected savings: In mission compromises, reduced
ing, checkout, static firing, and launch prepara- reliability, increased coordination effort between
tion, damage of various types may be incurred by engine builder and vehicle builders, substantially
the engine because of parts failures, mishandling, increased possibility of oversights and communi-
oversights, or accidents. All incidents must be cation gaps, and cost- and time-consuming retro-
corrected by repair and/or parts replacements. fits. As always, a joint thorough and unbiased
Any or all of the described maintenance action analysis considering all aspects, including that
may be required while the engine is still with of the long-range future, will readily yield in-
the engine builder, or following _ts delivery to formation about the point of diminishing returns.
the vehicle contractor. An accurate and effec-
tive maintenance or logistics plan must be Reserves and Safety Margins
worked out between the two contractors, or with
their contracting agency, considering all han- In his negotiations with the engine user, i.e.,
dling needs at all test stations. This includes "the customer," which quite likely is a Govern-
consideration of under what conditions engine ment agency, the engine designer, like the vehicle
return to the factory for overhauls, repairs, and builder, will frequently find himself under pres-
parts replacements should be made. If return is sure to compromise. This may be to cut weight,
not required, the correct handling fixtures and to accelerate schedules, or to maintain relia-
tools must be provided for each location. Above bility with dwindling funds. It is then that he
all, an adequate stock of spare parts must be will be most in need of his top management's
planned. The engine designer will frequently be understanding and support. But it is here also
consulted for his advice in an effort to avoid where he will be most criticized if it is found
both time losses due to lack of parts and costly that his analyses were incomplete, superficial,
overstocking. not optimized, or heavily biased by safety factor
upon safety factor. To strike an optimum bal-
ance between high performance and adequate
Multiple-Engine Use
safety factors and reserves is one of the finest
In view of the cost of rocket engine design, arts of engineering and is directly translatable
development, and procurement, common usage of into the degree of success. Once the designer
a given engine for several vehicles, or for sev- is certain that he has achieved this balance, or
eral stages of the same vehicle, is very desir- if he is concerned that he may lose it, he should
able. Also logistics, handling, checkout, launch so go on record. If he does, he should also
preparation, and instrumentation requirements remember Edmund Berkeley's observation:
can be substantially reduced or simplified. Cau- "Thoughtful and tolerant disagreement is the
tion must be exercised however, not to go "over- finest climate for scientific progress. "_
board," lest engine-to-vehicle interfaces become
considerably more complicated.
Effort should be concentrated on the common _Edmund C, Berkeley: Preface to "Giant Brains,"
usage of the major cost items and engine build- John Wiley, 1949.
Chapter XI
Design of Liquid Propellant Space Engines
Liquid propellant space engines embody the Spacecraft Main Propulsion Systems
same operating principles and general character- The main propulsion systems for most space-
istics as the liquid propellant rocket systems craft are pressure-feed, storable hypergolic pro-
previously discussed. However, their specific pellant systems, such as the A-4 stage propul-
missions for the use in spacecraft require special sion system, with thrust levels upto about 25 000
design considerations, which will now be dis- pounds. These systems include propellant tanks
cussed. and their pressurization system, control valves,
main thrust chamber assembly or assemblies,
and a gimbal mechanism or some other type of
II.I PRINCIPAL SPACE ENGINE
thrust vector control. Most main thrust chambers
APPLICATIONS
are ablatively cooled, while attached nozzle
By definition, space engines supply all extension skirts are radiation cooled.
those forms of rocket propulsionwhich a space- In a few cases, turbopump-feed systems have
craft requires for various maneuvers in space. been used, such as the 50 000-pound thrust (sea
This may include attitude control and stabiliza- level) turbopump-feed liquid oxygen/ammonia
tion, coplanar and interplanar orbit changes, rocket engine produced by the Reaction Motors
trajectory corrections, rendezvous maneuvers, Division of the Thiokol Corp. for the North
lunar and planetary landings and takeoffs, and American X-I5 research plane. As space mis-
retrofiring (reversed thrust for deceleration) dur- sions become more ambitious, requiring increased
ing reentry into the Earth's atmosphere. The total impulse and higher-energy cryogenic pro-
thrust of a space engine may be a fraction of a pellants, turbopump-feed systems will undoubtedly
pound or many thousands of pounds. Besides a play an increasingly important role in spacecraft
few solid propellant systems (used mainly for propulsion. Most space missions require multiple
single-start, retrofiring rockets) and stored gas starts and a certain degree of thrust throttling.
systems (used only in applications for thrust These requirements usually account for the prin-
levels less than 1 pound, and for a total impulse cipal differences between the main propulsion
of less than 5000 lb-sec), the great majority of systems for spacecraft and those for booster
the space engines are of the liquid propellant vehicles.
type. Because of their inherent operational
advantages, liquid systems most likely will con- Reaction Control System
tinue to dominate the space engine field, even These systems deviate more drastically from
when advanced nuclear and electrical propulsion other liquid propellant rockets, because of their
systems become available. design arrangement and their thrust levels, which
Liquid propellant space engine systems may run from 1 pound up to 500 pounds, with most
be divided into two basic groups: vehicle main systems probably failing within the 1- to 100-
propulsion systems and reaction control systems. pound range. The thrusts of these systems may
These differ not only in function and thrust level be used to provide attitude control to properly
but also in the type of propellant they use, the position a spacecraft, to aline a spacecraft for a
degree of required controllability and thrust midcourse corrective or terminal maneuver, and
variation, and system components. Common to to stabilize the vehicle after separation from
virtually all of them is the requirement that they another stage or during Earth reentry.
be able to start and operate reliably in the cold As a rule, attitude-control engines are used
vacuum conditions of space. in opposing pairs to produce pure couples about
429
x._. ii--
an axis. They are mounted in clusters to sim-

plify plumbing and wiring. Parallel pairs of
engines, or individual units, are used for trans-
lational movements along the vehicle axis, as in tNS'_U_NT "9 _
rendezvous and docking maneuvers. -- Off EN,_,kWE _ _EACT,O_
1
The reaction control engine systems are
usually pressure fed, using monopropellants or
storable hypergolic propellants. In all cases
thrust level and duration must be very closely
%t71
......
controllable. The thrust chambers may be abla-
G[ RO_E1 $ :
tively cooled or radiation cooled, depending on

the application.
L___@_A_
Application Example
Figure 11-1.-The propulsion systems of the
The application of various space engine sys- Saturn V [or the U.S. manned Apollo flight to
tems may best be illustrated by typical exam- _he Moon program.
ples, such as the spacecraft for the U.S. Apollo
Moon-landing program. The space vehicle sys- The command module requires thrust only for
tem to carry the Apollo aloft, the Saturn V, con- three-axis attitude control during reentry into the
sists of the S-1C (five F-1 LO2/RP-1 engines- Earth's atmosphere and to assist the service
total thrust, 7500000 pounds), the S-If (five J-2 module's reaction control system during separa-
LO2/LH 2 engines-total thrust, 1.15 million tion of the two modules. The command module
pounds), and the S-IVB (one J-2 LO2/LH 2 reaction control system is made up of two sepa-
engine-thrust 230000 pounds) stages (fig. 11-1). rate, interconnected pressure feed N20_/MMH
The Apollo spacecraft itself consists of the com- engine systems, either of which is capable of
mand, the service, and the lunar excursion completing the Earth reentry mission. These
modules. engine systems are located in the aft equipment
The service module has a 21 900-pound, fixed- compartments. Each system consists of six
thrust restartable pressure feed N_O4/50 (UDMH)- pulse-modulated, 91-pound, thrust, chamber
50 (N2H4) engine as its main propulsion system. assemblies of a 9-to-1 nozzle expansion area
Its thrust chamber has a nozzle expansion area ratio.
ratio of 60 to 1 and is gimbaled for thrust vector The lunar excursion module (LEM) has two
control by a redundant electromechanical device. main engine systems, one for the descent to the
This engine system provides the thrust for mid- Moon's surface, and the other for takeoff fromthe
course correction as well as for entry into, and Moon and injection into lunar orbit for rendez-
escape from, lunar orbit. It also serves as an vous with the command and service modules.
abort system, if necessary, after the launch The LEM descent engine system has a nominal
escape system has been jettisoned. thrust level of about 10 500 pounds, which can
The service module's reaction control system be gimbaled and throttled over a 10:1 range. The
is composed of four separate and interconnected takeoff engine is a rigidly mounted, 3500-pound,
pressure-feed engine systems utilizing the same fixed-thrust engine, which doubles as an abort
propellants as the main propulsion system. The engine in the early portion of the mission. The
systems are mounted external to the forward skin LEM also carries an attitude control system
of the service module at 90 ° intervals. Each which uses 16 thrust-chamber assemblies. All
engine system consists of four pulse-modulated, are pressure fed, using a N20_/50 (UDMH)-
100-pound thrust chamber assemblies with a 50 (N20 _) propellant combination.
40-to-1 nozzle expansion area ratio. These pro-
vide the three-axis attitude control for the space-
11.2 GENERAL DESIGN CONSIDERATIONS
craft during Earth orbit, lunar injection, lunar
orbit, LEM docking and separation from the Prior to the actual design of a space en-
command module. gine system, it is necessary to examine the
DESIGN OF LIQUID PROPELLANT SPACE ENGINES 431
basic engine parameters to determine their in- mass (limited by the payload capability of the
fluence on the spacecraft's ability to perform its launch vehicle) and required mission velocity
intended mission. Here, again, an overriding increment, the mission payload capability is a
consideration is reliability. Typical space mis- function of the main engine system specific im-
sions include the establishment of an orbit pulse and the spacecraft inert weight, as indi-
around a planet following a transfer from Earth cated by equation (2-1). Similarly, for a given
which may last several months or years. The spacecraft initial mass, inert weight, and pay-
ability of a propulsion system to coast in space load, the mission velocity increment capability
for these prolonged periods and then fire reliably is directly proportional to the main engine sys-
is typical for many contemplated missions. It is tem specific impulse. The specific impulse of a
one of the foremost considerations in the design reaction control engine system affects directly
of space engines. Based on mission, trajectory, its system package weight. A higher specific
and payload requirements, the following major impulse will result in a lower system weight for
design areas must be examined for the optimiza- a given total impulse requirement.
tion of one or a group of propulsion systems: 2. Operating temperature.-In an overall
(1) Propellant type vehicle optimization study, the criterion of spe-
(2) Vehicle system operational requirements cific impulse must not be used as the only hard-
(3) Mission environmental effects and-fast rule for propellant selection, because
Several different designs (assuming various the problem of maintaining the propellants at
propellants, feed systems, chamber cooling temperatures that permit effective use of the
methods, etc.) may be conducted simultaneously propulsion system after coast is an additional,
for a given space mission. These designs must major consideration in spacecraft design. The
be carried far enough to evaluate their relative requirement for thermal control of the propellants
merits such as reliability and performance, and during coast affects the selection of the propel-
to establish the advantage of one design over lants and the vehicle configuration. Heat trans-
the other. An optimum system design should fer between propellants, and between propellants
thus emerge. and other vehicle components, must be consid-
ered, as well as heat radiation out to space, in
Selection of Propellants
from the Sun, and between vehicle components.
Various aspects of liquid rocket propellants Studies have indicated the feasibility of insu-
have been discussed in section 1.4. Here we lating the propellants against excessive temper-
will discuss the characteristics which are perti- ature changes, even during long coast periods in
nent to their selection for space engine applica- deep space missions. However, for the cryo-
tion. The reliability and performance of a space- genic propellants more insulation weight is
craft will depend greatly on the propellants used. required than for Earth-storable propellants.
This is of paramount importance in establishing This difference becomes greater the longer the
the optimum design criteria and final operational mission coast periods.
capabilities of a vehicle. For any planned space 3. Density.-The bulk density of a propellant
mission, the projected program time periods, as combination has a pronounced effect on vehicle
related to the anticipated state of the art and payload. This can most easily be seen in the
operational reliability of the propellants to be case of the LF2/N2H _ combination, the specific
used, should be the first criterion for their selec- impulse of which is comparatively low, as com-
tion. Propellants with undesirable characteris- pared to LF2/LH 2 or LO2/LH 2. However, its
tics and those that have no advantages over bulk density advantage may result in a higher
similar, more desirable types should be elimi- ideal velocity increment for certain space
nated early in the studies. In addition, propel- missions.
lants are selected on the basis of the following: 4. Ignition cflaraczeristics.-Hypergolicity of
1. Speci[ic impulse.-The criterion of spe- the propellant combination used in space mis-
cific impulse is, perhaps, the most commonly sions is always considered very desirable to
used basis for comparison in nearly all propel- effect a simpler and more reliable engine sys-
lant evaluations. For a given initial spacecraft tem, particularly for multiple starts.
5. Cooling and other characteristics.-Some

propellants are excellent coolants (for regenera-
tive or film cooling), while others have little
cooling capability. The propellant combination
will determine the combustion temperature and -- FI/_ z PuuP
the gas constituents, which vary widely as to

their compatibility with chamber materials, espe- _ Fem_N , px[ssu_£
cially with ablative or refractory materials which %_1%_4 PUllS
are frequently used in space engines. r WEEK _ kq:_TM $ blONT_tS :AR I0 YEA_
If possible, all propulsion systems in a space- 1.000 i0,O00 t_.0¢¢
craft should use the same propellant combina- COAST I_UI_TIO_, _*OUl_
tion. However, in some applications the reaction

Figure ll-2.-EHec$ of mission coast duration on
control engines are also used for propellant set-
relative payload capabilities o[ various propel-
tling in the main system tanks. Thus they must
1ant engine systems for a given mission veloc-
be supplied by separate tanks equipped with
ity increment.
positive expulsion devices. Different propel-
lants may have to be used in this case. Charac-
teristics of various propellants suitable for Vehicle System Operational Requirements
space engine applications are summarized in
table 11-1. Characteristic trends of relative After the vehicle system operational require-
payload capabilities versus mission coast dura- ments have been analytically established for a
tion, for various propellants and engine systems, given space mission, many design parameters
are shown in figure 11-2. It can be seen that for and special considerations for the engine system
coast periods up to about 4300 hours (6 months), or systems can be derived. Principal areas of
there is no significant change in payload capa- consideration, in addition to propellant selection,
bility. are:
TABLE ll-1.-Comparison of Various Liquid Propellant Combinations [or Space Engine Applications
aSpecific Density Freezing and boiling points C'F, Mixture Combustion

Propellant oxidizer/fuel ratio, temperature, Remarks
impulse, impulse,
combination
seconds sec-gm/cc (FP)(BP)/(FP)(BP) O/F °F
241 (-364)(- 307)/(-435X-423) 9.3 6440 Hypergolic

_2/H2 ............... 474
456 143 (-362)(-298)/(-435)(-423) 4.7 5110 Nonhypergolic
_O2/H 2 ..............
424 7010 Hypergolio
3F2/B2H6 ........... 430 (-371)(-299)/(-_65X- 135) 3.5
357 (- 371X-299)/(-265)(- 135) 2.15 5910 Hypergolic
3F2/B2H 6 ........... 420
419 551 (-364)(-307)/(35)(236) 2.4 7285 Hypergolic
F'2/N2H 4 .............
417 451 (-371)(-299)/(-300)(-260) 5.6 670O Nonhypergolic
3F2/CH 4 ............
408 303 (-362)(-298)/(-265)(-135) 2.0 5960 Nonhypergolic
32/B2H 6 .............
402 254 (35)(9.36)/(- 265)(- 135) 1. 4085 Nonhypergolic
_2H4/B2H_ ..........
375 340 (11)(70)/(-265)(- 135) 2.9 5710 Nonhypergolic
_20,/B2H6 ..........
341 409 (11)(70)/(35)(9.36) 1.23 5513 Hypergolic
_N:O_/N2H _ .........
_MON/EMHF ......... 341 407 (-23)(29)/(-76X144) 2.2 5330 Hypergolic
bMON/MMH ........... 338 401 (-9.3)(29)/(-63)(189) 2.4 5370 Hypergolic
bN_o4/50-50 ......... 339 408 (11)(70)/(18)(170) 2.1 5175 Hypergolic
bN_O4/MMH . 339 407 (11)(70)/(-63)(189) 2.3 5290 Hypergolic
Notes: MON =Mixed oxides of nitrogen, 85% N204-15% NO.

MMH =Monomethyhydrazine, CH_ N_H_, 50-50= 50% UDMH. (H_C)_N_H_-50% N:H_.
EMHF = Eutectic mixture of hydrazine fuels = 87.6% MMH-19..4% N_H,.
abased upon theoretical shifting equilibrium at 150-psia nozzle stagnation chamber pressure and 40:1 nozzle expan-
sion area ratio in the vacuum.
bpropellant technology and application are well established.
.,a======
I. Engine system total impulse, design thrust tion. This may be expected, considering that
level and run time.-These are determined and continuous operation involves a minimum of con-
optimized by analyses of the mission trajectory, trol components and actuations. Multiple starts,
considering spacecraft operating limitations such on the other hand, require that each component
as acceleration loads. Typical liquid propellant responds perfectly each time it is called upon.
rocket engine total impulse, thrust, and run times In addition, multiple starts require much longer
for various space missions are presented in fig- periods of absolutely minimum propellant and
ure 11-3. Required mission total impulse and pressurant leakage. Therefore, more rigorous
engine thrust level influence, to a large extent, approaches toward improved reliability, such as
the choice between a pressure feed and a pump component or subsystem redundancies, must be
feed system. considered in the design.
2. Engine system operating characteristics.- 4. Engine system component design.-Besides
This includes engine thrust throttling range, mission environmental effects (to be discussed
thrust as a function of time, number of starts and below), the detail design of the engine system
repetition rate, cutoff impulse and accuracy, and components are directly affected by the space-
thrust vector control requirements. Most of these craft design configurations. For instance, the
characteristics are determined by various pro- location of the system (internal in or external to
jected spacecraft maneuvers. It may also be the spacecraft) determines the feasibility of
desirable for the same engine system to fulfill radiation cooling for the thrust chamber. Engine
more than one type of maneuver, or to be reused space envelope limitations affect the choice of
on subsequent missions. Typical thrust-time nozzle type, allowable expansion area ratio, or
histories for various spacecraft maneuvers are even the feed system type. (In some applica-
presented in figure 11-4. tions a pump feed system affords a much higher
3. Engine system design.-Experience has
shown that multiple-start operation as required
by most space engine systems has a more severe
effect on a rocket engine than has continuous
steady-state operation of comparable firing dura- TIME
CONSTANT THRUST, VARIABLE IMPULSE
(LUNAR LANDING)
i 10,000 / VARIABLE THRUST,

TIME
VARIABLE
(RENDEZVOUS)
IMPULSE
._ .CIO X PU_LSE ROCKET--

(ATTITUDE
TIME
DISCRETE
CONTROL)
IMPULSE BITS
10
0.i l
THRUST - THOUSAND
|
I0
LB";
100
FI/ T IME
(LUNAR TAKEOFF)
Figure 11-3.-Typical liquid propellant rocket Figure ll-4.-Typical liquid propellant rocket
engine total impulse, thrust, and run time for engine thrust-time histories for various space-
various space missions. (Coulbert, C.D., crait maneuvers. (Coulbert, C. D., "Selecting
"Selecting Cooling Techniques [or Liquid Cooling Techniques [or Liquid Rockets [or
Rockets lot Space Cra/t," J. Spacecra[t and Space CraIt," J. Spacecralt and Rockets 1,
Rockets I, 129-139, 1964.) 129-139, 1964. )
expansion area ratio for a given envelope.) The planetary emitted radiation. Direct solar radia-
structural design of the system components is tion is the largest external heat source. It is
influenced by the maximum vehicle acceleration therefore advantageous to orient the spacecraft
and by vibration loads. for maximum protection from direct solar radia-
tion to the vulnerable subsystems such as pro-
pellant tanks. The quantitative evaluation of
Mission Environmental Effects
thermal radiation in space indicates that damage
Afterthe propellantshave been selected and to some components may result from the exces-
the basic propulsionsystem design established, sive absorption of radiant energy, unless protec-
with preliminaryconsiderationsof mission envi- tive means are applied such as reflective sur-
ronmental effects,itis possible to examine the faces or coatings. A very real problem with the
various engine subsystems in detailto determine prediction of the thermal properties of materials
which components or phase of system operation in the vacuum is the lack of knowledge about the
is stillaffectedby the projectedmission envi- effect of the ever-present oxide-metal surface
ronment in space. The characteristics of this film on heat transfer mechanisms.
environment include the following: 3. Nuclear radiation.-The two most important
effects of nuclear radiation in space on metals
1. High vacuum.-The low ambient pressures
are the production of internal heat, and disloca-
experienced in space missions may cause vapor-
tion of the atoms within their crystalline struc-
izationof metals and partialor complete removal
ture. The nuclear particles of interest are fast
of filmor adsorbed gas layers at the surfaceof
neutrons, lower energy protons, alpha particles,
the material. Principal problems associated
electrons, and gamma rays. The fast neutrons
with high-vacuum conditions are:
present the most severe problem. It is known,
(a) Reduced ability of a component to perform
however, that they do not exist to a great extent
its function, due to mass loss through
in space. Generally, the total radiative flux
material vaporization or bulk property
from all nuclear radiations is not sufficient to
changes.
(b) Changes in the radiative heat transfer cause any appreciable damage to metals over a
period less than 2 years.
properties of the material, which could
increase, to a destructive level, the 4. Meteoric bombardment.-Spacecraft will be
exposed to the impact of meteoroids ranging in
bulk temperature of a component such as
a radiation-cooled nozzle skirt. size from microscopic dust particles to bodies of
asteroid dimensions with a wide distribution of
(c) Possibility of condensation of metallic
kinetic energies• Collision of the spacecraft
vapor on electrical components, result-
with these materials would result in surface
ing in short circuiting.
erosion,punctures,or totaldestructionof the
(d) Changes in tile fatigue, frictional, and
vehicle. The idealprotectionforpropulsion
creep properties of materials, as well as
system components would be to locatethem
the possibility of self-welding of metals.
entirelywithin the vehicle. This, however, is
Design remedies against high-vacuum effects
usually not practical for parts such as the thrust
include material selection and proper location of
the components within the vehicle to take advan- chamber nozzle skirt. Therefore, adequate con-
sideration must be given to the design of these
tage of nonvacuum environments.
2. Thermal environment.-The thermal envi- parts with respect to the cited effects.
ronment encountered in space missions, such as 5. EHects ol gravity.-The absence of accel-
in Earth orbit or during the transfer phase in eration forces, as well as the presence of large
interplanetary flights, is an important considera- acceleration forces, should cause no mechanical
tion in the design of space propulsion systems, design problems if proper provisions are made in
especially for the storage of the propellants. the design phase. The main problem will be
Three external sources of radiant energy must be with the propellant feed system. This can be
considered when evaluating the thermal environ- overcome by positive expulsion or by providing
ment of a spacecraft. These are direct solar, propellant settling rockets (which themselves
albedo (solar energy reflected from a planet), and would require a positive expulsion feed system).
11.3 DESIGN OF SPACECRAFT MAIN tioncould be providedby severalindividual

PROPULSION SYSTEMS pressuranttanks of differentsize,each of which
remains sealed by explosive-actuatedvalves
System Design (fig.7-72) untilneeded. A sealed pressurization
system would not requiregas venting during any
The reliability requirementsforspace mis-
partof the mission, thus eliminatinglosses due
sions have led to simple, storedinertgas pres-
to intentionalventing,or because of leakage
sure feed main propulsion systems (fig.5-1),
past the vent valves.
using conventionalhypergolic Earth-storable
In some applications,the propulsion system
propellantsand redundantcontrolcomponents.
can be operated on a propellanttank ullage gas
For tube connections, the trend is to welded or
blowdown principleduring the latterportionof
brazed joints, with flexibility provided either by
the mission. This could reduce the required
braided metal hoses or metal bellows. Integrated
amount of pressurant (and pressurant tank vol-
engine packages are preferred, consisting of
ume) by as much as one-half of that for a regu-
thrust chamber and injector assembly, integrated
lated, constant pressure system.
propellant control valve package, and flight
instrumentation. Besides meeting thrust and Ignition of the main thrust chamber is critical
performance specifications, principal system with any multiple-start space propulsion system.
design requirements for a spacecraft main engine If hypergolic propellants are used, ignition de-
are: vices such as spark plugs of ASI units are not
needed. However, hypergolics suffer from igni-
(1) High reliability and crew safety (through
tion delays at low temperatures (such as are
redundancy in the control system; use
encountered in space). Under the peculiar con-
of proven design concepts, materials,
ditions of vacuum starting, irregular transients
and fabrication techniques; and ex-
may occur upon ignition, such as violent chamber
tended life and overstress testing)
pressure surges. These may not be destructive,
(2) High combustion stability rating with
but they could place a severe burden on the
respect to perturbations during start
control system.
and throttle transients
For nonhypergolic propellants, spark plugs or
(3) Capability of numerous starts, with re-
augmented spark igniters (ASI) may be used.
peatable cutoff impulse
Spark plugs have the disadvantage of corona-
(4) Capability of deep throttling (optional,
arcing and electrical leakage, even in a partial
not required by all missions), with
vacuum. However, injection of unburned propel-
simple controls lants into the thrust chamber or ASI with attend-
(5) Capability of propellant utilization and
ant vaporization usually builds up sufficient
tank blowdown (ullage expansion)
chamber back pressure to allow the plugs to
(6) One signal each, for start and cutoff
function. In some systems, this process may be
(7) No purge, bleed, or lubrication equipment
augmented by purge gases. Certain propellant
required
combinations tend to ionize the gas around the
(8) Effective isolation of the propellants
plug and thus quench the spark.
from the system during long coast peri- Various schemes have been considered and
ods (use of propellant isolation valves)
developed for the deep throttling often required
(9) System growth potential and flexibility
by spacecraftmain propulsion systems. For
to mission modifications
instance,the Apollo LEM descent engine is
(10) Sufficient protection against the effects
requiredto vary thrustcontinuously over a range
of heat, nuclear radiation, and meteoric
from 1050 to 10500 pounds to permit hovering,
bombardment.
selectionof a landing site,and the landing
For missions involving intermittent propulsion itself.Two alternateapproaches to throttling
system operation, during periods of more than a are being pursued. One calls forvariationof the
few months, the pressurization system design thrustlevel by main propellantline throttle
calls for special provisions, such as a com- valves,combined with aerationof the propellants
pletely sealed pressurant system. Pressuriza- with helium gas, as shown in figure4-47(b).
Such aeration requires a surprisingly small pro- propellants for deep throttling is provided by
portion of gas by weight. The alternative ap- separate helium lines and redundant valves. If
proach is to vary combustion chamber injection the system throttle requirements were removed,
area, as shown in figure 4-47(a). Other potential the main propellant throttling valves could be
space engine throttling methods include use of replaced by simple on-off types. Propellant
throttlable propellant gasifiers or precombustors, filters are provided in both main propellant lines
or the subdivision of the injector manifolds into to minimize system line contamination with
segments, each of which can be shut off by foreign particles.
simple on-off valves. The package design layout of the pressure-
Figure 11-5 presents the schematic of a typi- feed spacecraft main propulsion system sche-
cal pressure-feed spacecraft main propulsion matically illustrated in figure 11-5 is shown in
system using hypergolic Earth-storable propel- figure 11-6. The integrated main valve package
lants. A smaller and a larger pressurant tank is mounted directly to the main thrust chamber
satisfy pressurization requirements before and assembly, immediately upstream of the injector.
after a long coast period. The propellant tanks The integrated engine assembly can be either
are equipped with positive expulsion diaphragms rigidly mounted, or gimbal mounted, depending
and remain pressurized to mission completion. on requirements. Lightweight spherical propel-
An integrated main valve package contains main lant and pressurant tanks are secured by simple
propellant throttling and isolation valves, in- frames. An all-welded and brazed construction
cluding redundant features. This redundancy is is employed for system assembly to prevent
obtained by series-parallel connection of the external leakage of pressurant gas or propellant.
mechanically linked main propellant valves and In a complete space vehicle system, perform-
increases system reliability, virtually eliminating ance gains of the spacecraft stages will effect
critical failures (see also fig. 2-8). Aeration of increasing weight savings for each succeeding
lower booster stage, as indicated by equations
(2-1) through (2-5). When planning future space-
craft missions, high-performance liquid propel-
lants, such as LO2/LH 2 and LF2/LH_, thus
definitely should be considered. For high mis-
sion total impulse requirements, these propel-
lants, when used in turbopump-feed systems,
tend to provide a considerable performance edge
over pressure-feed systems (fig. 11-2). A
turbopump-feed system for these applications
must reliably supply high-pressure propellants to
the combustion chamber, under vacuum and zero
gravity conditions, and following extended coast
=__-'_'_ CHECK VALVE
=_ FtLTE_ _w{} PRES_E

REGULATOR
._,- PROPELLINT FILTER
Figure ll-5.-$chematic of a typical pressure- Figure ll-6.-Package design layou( ol a $ypical

feed spacecraft main propulsion system using pressure-feed spacecraft main propulsion sys-
hypergolic Earth-storable propellants. tem using hypergolic Ear_h-s_orable propellan$s.
periods. All pumps must be fully primed prior to LFz/LH 2 propellant combination. The main
starting of the engine system to prevent delays thrust chamber is regeneratively cooled by the
in pump buildup and overspeeding. This may fuel to a nominal nozzle expansion area ratio of
require ullage settling rockets, before and during 44:1. A radiation-cooled nozzle extension skirt
main engine start, or positive propellant expul- may be added for certain missions to yield an
sion devices. overall expansion area ratio of 75:1. The pro-
The well-proven turbopump subsystem designs pellants are supplied to the main chamber by two
and their operating concepts as established for separate turbopump assemblies which are mounted
large booster engine systems, and as discussed to the thrust chamber body, resulting in an inte-
in preceding chapters, should be effectively gral, compact engine. The turbines are driven
applied to the design of turbopump-feed space- by gases bled from the main combustion chamber
craft main propulsion systems. In addition, the in a parallel arrangement (fig. 6-13). A hot-gas
systems must be able to start an unlimited num- throttle valve (such as shown in fig. 7-31)located
ber of times (if possible from available tank upstream of the'turbine inlets controls tile tur-
pressure energy, without auxiliary starting de- bine power, and thus the thrust level. A Y-type
vices) and to vary the thrust level (for some hot-gas duct is used to route the turbine ex-
applications over a wide range). Major problem hausts into the main chamber nozzle. The entire
areas associated with turbopumps under space engine package can be gimbaled at the gimbal
environmental conditions other than material mount by linear gimbal actuators.
considerations, are as follows:
(1) Vacuum, temperature, and radiations
Main Thrust Chamber Design
effects on exposed high-speed bearings
and dynamic seals In addition to the design elements presented
(2) Micrometeorite penetration of fluid pas- in chapter IV, primary requirements for thrust
sages chambers of spacecraft main propulsion systems
(3) Absence of the effects of gravitation on are:
pump operation
Figure 11-7 presents the package design lay- (1) Optimum steady-state cooling methods,
out of a hypothetical turbopump-feed spacecraft with due consideration of application in
main engine system using a LO2/LH2 or space

(2) Ability to withstand postrun soakback,
intermittent operation, and storage in the
space environment
(3) Ability to enter planetary atmospheres at
OX;D;ZERTUR_O_U_
F_:!__/_ _AO_AT'ON-COOLEO high velocities
NOZZLE EXTENSION SKIRT
GI M_L MOUN_ _- "_ Many of the thrust-chamber-cooling techniques

described in chapter IV are applicable to space
engines. Especially the ablatively cooled thrust
chambers, which offer the inherent advantage of
simplicity, ruggedness, and independence of
propellant coolant flows (including their pressure
EXHAUST I
drops), and thus offset their limited total life and
oucr i. 44 • - 75 relatively heavy weight, are suitable for most
/ -- THRUST
/ C HAMIER pressure-feed space engine systems using Earth-

storable propellants at chamber pressures of
/ L HOT-,GAS ,o::'Z'¢:
THROTTLE VALVE
less than 150 psia. In some applications, abla-
/-'--- FUEL TURI, OPUMP
tive cooling is supplemented by one or a com-

Figure ll-7.-Design layout of a hypothetical bination of the following techniques:
turbopump-feed spacecraft main engine system (i) Refractory throat inserts
using a LOJLtt 2 or LFJLtt 2 propellant com- (2) Radiation-cooled nozzle extension skirts
bination. (3) Propellant film cooling
One of the problems associated with ablative coated with molybdenum disilicide (MoSi2) to
cooling is the dimensional change of the chamber prevent oxidation, are frequently used for nozzle
due to erosion. Nozzle throat erosion, if con- skirt extensions (for skirts starting at _ = 8 to 12)
trolled and pre2ictable, is acceptable for some which operate at sustained temperatures around
engine systems and becomes proportionally less 2600o-2800 ° F. For nozzle skirts of higher expan-
significant in larger thrust chamber units. For sion ratios (starting at _ = 40 to 45)° Hastelloy-X
fixed-area injectors and fixed propellant supply is sometimes used at operating temperatures
pressures, propellant flow and engine thrust. around 2000° -2200 ° F. One potential disadvan-
would increase with time, while specific impulse tage of radiation-cooled devices is their require-
would decrease, due to throat enlargement. How- ment to "see" space. A radiation-cooled skirt
ever, for a 40:I expansion nozzle operated in should face outboard and should not radiate
vacuum, the specific impulse loss would be only undesirable heat to vehicle components. Radia-
0.5 percent, for as much as i0 percent increase tion cooling appears unsuitable for clustered
in throat area, ifthe aerodynamic characteristics engines.
of the nozzle contour did not deteriorate. For exclusively ablative-cooled chambers
One remedy against excessive throat erosion used in LO2/LH 2 or LF2/LH z systems, analyses
is the use of a refractory throat insert. A ceramic, and tests have proven liquid hydrogen to be a
silicon carbide, has been used extensively for very effective throat film coolant in low chamber
throat inserts in space engine applications. It pressure applications (less than 100 psia). As
has a high melting point (4400 ° F), excellent low as 0.1 percent of the total propellant flow
thermal shock characteristics, relatively high used as the throat film coolant greatly reduces
thermal conductivity (115 Btu/hr-ft2-° F/in), low throat erosion for extended firing durations. The
coefficient of thermal expansion (2.4 x 10 -6 effect of such a small film coolant flow on sys-
in/in-° F), excellent oxidation resistance, and tem performance is so slight that it can be ne-
high abrasion resistance. Sometimes, a molyb- glected. Ablative chambers with their throat
denum backup sleeve is used when the silicon film-cooled by liquid hydrogen, therefore, may be
carbide insert cannot conduct heat sufficiently. considered excellent prospects forfuture pressure-
Another potential problem is the danger that feed spacecraft main propulsion systems using
the high-silica glass in the ablative material of LO2/LH 2 or LF2/LH _.
the combustion chamber section becomes suffi- Following the firing of an ablative thrust
ciently fluid to be swept downstream and to be chamber, the heat stored in the charred phenolic
deposited in the throat section. This causes and silica reinforcement or in the throat insert
thrust variances and promotes an unsymmetrical refractory material soaks into the unburned virgin
velocity profil6 in this area. This phenomenon material. This postrun soakback propagates
may be prevented by a liner of JTA (45 percent further thermal degradation, which might also be
graphite, 45 percent zirconium diboride, and I0 affected by the vacuum conditions, for 100 sec
percent silicon) inserted in the combustion zone. or more, until the mean temperature of the char
The liner is usually segmented, to provide a path is reduced to about 500 ° F. The weight of gas
for the gases from the pyrolyzed ablative (and generated and expelled by soakback charring is
reduce the Ap across the liner),and to prevent about 15 percent of the weight of ablative mate-
cracks that would develop in an expanding un- rial charred. It could cause a postrun residual
seg[;ented liner. impulse undesirably exceeding the desired mini-
Exclusively radiation-cooled thrust chambers mum cutoff impulse. However, this effect is
would be subject to large stresses caused by the small for larger systems. For long-duration
high temperatures in combustion zone and throat space missions, the temperature effects from
by the thrust transmission, and at the injector solar radiation may cause vaporization of the
attachment points. By contrast, the radiation- ablative chamber material during coasting and
cooled nozzle skirt is designed to accept and should be prevented ifpossible.
emit only the heat flux transmitted by the ex- Figure 11-8 shows the design of a typical
panding gases, and some of the loads imposed thrust chamber for a pressure-feed spacecraft
by thrust transmission. Molybdenum alloys, main propulsion system using hypergolic
ings should not affect total chamber duration.

Equations (4-36) and (4-37) may be used to cal-
culate the char depth. For Refrasil phenolic
;,, i',.'_ ,_,

.............. _ iS.- j ablative thrust chambers operated at pressures
around 100-125 psia, and using N204 and hydra-
zine type fuels, the following empirical equa-
tions may also be used to estimate char depth,
including the effects of charring due to soakback:
(1) For the combustion zone and the throat

section (with or without liner and/or
insert):
I
a=0.0415t °-s (11-1)

Figure l l-8.-Typical thrust chamber design de-
tails [or a pressure-[eed spacecraIt main pro-
(2) For nozzle sections downstream of the
pulsion system using bypergolic Earth-storable
throat:
propellants.
a = 0.0335 t as e -°024;e (11-2)
Earth-storable propellants at a chamber pressure

of 125psia. Refrasil-filled, phenolic-impregnated where
ablative materials are tape wrapped on a man- a = char depth, in
drel, at the combustion chamber and nozzle sec- t --thrust chamber cumulative firing duration,
tions, up to an expansion ratio of 42. A JTA sec
liner and a silicon carbide insert are provided e = base of natural logarithms. 2.718
for combustion chamber section and throat, =nozzle expansion area ratio at the section
respectively. The chamber outer wall is made of under consideration
fiber-glass filament, wound to accept hoop Once the char depth based on design duration,
stresses, and then wrapped longitudinally (heli- including effects of soakback, is determined, the
cally) to contend with meridional stresses. An thickness of the insulating layer required to keep
insulating layer, consisting of the basic or the outer wall at the required temperature can be
another ablative material such as asbestos or calculated.
alumina silica, is wrapped around the basic For turbopump-feed space engine systems
ablative layer and cured to reduce the rate of operated at relatively high chamber pressures,
heat flux to the structural fiber-glass outer wall. regeneratively cooled tubular-wall thrust cham-
A radiation-cooled skirt made of 0.030-inch bers appear more suitable from heat transfer
Hastelloy-X sheet is tape wrapped to the thrust considerations. A possible disadvantage for
chamber nozzle to provide the nozzle extension space engine application is the response time
from an expansion ratio of 42:1 to 80:1. An inte- from start signal to full thrust, which may be
grated injector is attached to the end of the com- substantially increased if the propellant valve is
bustion chamber and is made of an aluminum- located upstream of the coolant jacket. Relocat-
alloy forging. It is a conventional, fixed-orifice, ing it to a point downstream of the jacket would
single-ring, unlike-impinging doublet type. reduce the response problem, but would introduce
Ablative thrust chambers for space engines, heat-_ransfer problems after shutdown, with the
such as shown in figure 11-8, should be designed propellants trapped in the jacket.
to permit many restarts. At the end of a speci- Other potential problems are the heat-transfer
fied total firing duration (design values: 400- characteristics during throttling. Also, tubular
2000 sec), sufficient insulation material must walls are more susceptible to meteoritic damage
remain between char and fiber-glass shell to than are solid walls. In some applications, an
limit the outer wall temperature to a specified optimum overall design may result from combin-
maximum value (design values: 400°-700 ° F). ing regenerative cooling with film cooling, and
Any heat soakback during ofttimes between fir- adding a radiation-cooled nozzle extension skirt.
Design of Control Components seals, and hermetically sealed solenoids or

torque motors.
The various aspects of control component
The valve is spring loaded normally closed,
design have been discussed in chapter VII.
and additionally pressure unbalanced closed
Pertinent considerations for the design of space
when inlet pressure is applied. Actuation to
main engine control components, in general and
open is achieved by allowing the oxidizer fluid
for propellant valves in particular, are:
to enter the actuation chamber at nominal oxi-
(1) Maximum reliability
dizer valve inlet pressure which, in turn, over-
(2) Use of bellows for dynamic sealing in
comes the closing forces on the poppets. For
view of the need for long-term operating
on-off and isolation valve applications, the full
lifein a vacuum environment
open or closed valve positions are effected by a
(3) Combination hard and softvalve seats,
three-way solenoid pilot valve which, when ener-
forminimum leakage
gized, pressurizes the valve actuation chamber
(4) Provisionof a minimum of two seals and
to open, and which, when deenergized, vents the
a vent between different propellants
same chamber allowing the pressure unbalance
(5) Avoidance of slidingsurfacesin compo-
and the springs to close the poppets.
nents operatingin a vacuum environment
In the applications as a dual-propellant throt-
(6) Where possible,mechanical linkage of the
tling valve, the position of the linked main pop-
propellantvalve actuatorsby a mecha-
pets is controlled by a servo pilot spool valve
nism thatensures positive,consistent,
which meters the control fluid flow (oxidizer) to
and synchronized opening and closing
the main valve actuation chamber. The pilot
forsmooth, repeatableignitionand
spool valve, in turn, is proportionally positioned
thrusttermination
by a servo torque motor, as a function of the
(7) Minimum electrical
energy requirementfor
command signalcurrentand itspolaritysupplied
actuations
from a servo amplifier.By means of properly
The basic design detailsforthe various con- shaped valve poppet contours,the main throt-
tlingvalves controlthe flow of oxidizerand fuel
trolcomponents forspace main engine systems,
to the main thrustchamber at constantmixture
such as pilotvalves,regulators,and ventvalves,
ratio,and thus regulatethe main chamber pres-
are quite similarto those forbooster-stageappli-
sure (thrust).
cations. The mechanically linked,poppet-type
dual-propellantvalve arrangement shown in
figureIi-9 is typicalformain-propellantor
System Design of Propellant Storage for Space
propellant-isolationvalves of space main engines
Missions
using hypergolicEarth-storablepropellants.In
the valve shown, linked actuationis accom- As previously mentioned, the storage of pro-
plished by locatingthe valve poppets in series, pellants for space-system propulsion represents
thus causing the oxidizerpoppet motion to simul- an important design area. Cryogenic propellants
taneously move the fuelpoppet. Complete sepa- require protection against high temperatures,
rationof fuel and oxidizeris achieved by welded particularly for extended storage periods. Earth-
bellows seals locatedat both ends of the fuel storable propellants, by contrast, require protec-
poppet and at the connection end of the oxidizer tion against low temperatures, because of the
poppet. Since valve actuationis provided by danger of propellant freezing. There are two
oxidizerfluidpressure,no bellows is requiredat basic design approaches to propellant storage
the actuationchamber. The center vent chamber for space missions, the vented and the nonvented
between valves sectionsis the criticalarea systems. The selection of the optimum system
where propellantmixing must be prevented. design depends on mission, propellants type,
However, both bellows would have to failto and type of engine feed system.
cause this failuremode (reliability throughre- In a yen-ted storage system, propellant pres-
dundancy). The design also incorporatesall- sure and temperature are maintained at a con-
welded static sealing joints, all-metal construc- stant value. This system is mainly used for
tion except for the Teflon dynamic poppet valve cryogenic propellants. Any heat input to the
propellants is absorbed by allowing a small per- divided into a number of intervals. Equilibrium
centage of boiloff. Since conditions will vary skin temperatures and net heat fluxes (consid-
widely during travel, the net heat input to each ering both internal and external sources) to the
propellant tank may be computed by a numerical propellant tanks are then determined for each
integration process for the specific vehicle path, orbit location. By algebraically adding the net
such as a complete orbit. This path can be heat inputs for each, the net heat input rate for
OUT
i E- PISTON SEALS
SOLENOID_ II] /\ _OXIDIZER OUTLET /_ FUEL OUTLET
I_ _-t_" _Y__NNDSEE_L_ VE %NT_/ TEF(LBONDSEEA_

/--'---WELD
ox,o,z
ACTUATING
o.-O FV W
FLOW
VENT
f SERVO
MOTOR TORQUE-- 7 _
. A
WELD'_ ACTUATINGPISTON_ /__?-_
_',_ I I/ //// t \\\X'I

-- A_..:._ --%'j_ONTOURED(TYPICAL) POPPET
T__ I_ _j \ "IROTTLING VALVE ONLY
OXID
SERVO I,_--':1 /_ I _
I',,, VALVE / v
AS A PROPELLANT THROTTLING VALVE
1 _ WELD THERMAL RELIEF (TYPICAL)
---- A THROTTUNGVALVEONLY
SECTION A- A
Figure 1 l-9.-Typical mechanically linked, poppet-type, dual-propellant on-oil or throttling valve, [or
spacecra[t main propulsion systems using hypergolic Earth-storable propellants.
one complete orbit can then be obtained and the jectory correction maneuvers, for which it must
amount of propellant boiloff per orbit calculated. be oriented properly.
In a system without venting, any heat input Another example is spacecraft separation from
to the propellant results in a propellant tempera- the launch vehicle, during which momentums may
ture rise and a corresponding increase in vapor be imparted which require correction for proper
pressure. This system can be applied to both orientation.
cryogenic and Earth-storable propellants. The Ideally, when a spacecraft is oriented at some
allowable pressure rise, and consequently the attitude, it should remain there indefinitely.
maximum heat input, is dictated by the allowable This, however, is not the case, because the
tank pressure. In this case, the storage analysis spacecraft is continuously subjected to small
assumes an allowable vapor pressure rise during external and internal forces which will cause it
the mission, based on initialconditions. This to driftoff the desired position. An attitude
permits determination of the final propellant control system must function to counteract all
temperature, expected changes in propellant disturbing torques. It could be of either a con-
density and ullage volume requirements, and tinuous, or of an on-off type. Practical consid-
pressure-dependent propellant tank weights. For erations tend, in general, to favor the latter.
cryogenics, it is usually assumed that the pro- The attitude control system itself contains cer-
pellant experiences a uniform temperature rise. tain nonlinearities and nonideal conditions. In
The maximum storage time is then computed by actual operation these inherent nonlinearities
dividing the propellant heat capacity by the cause the vehicle to settle in a periodic motion
average rate of heat input. In this system, the about a reference point. This motion is repre-
storage weight penalties are composed of extra sented by a closed curve in the phase-plane and
tank weight required to accommodate the pres- is termed the "limit cycle of operation."
sure increase, the extra tankage required to The problem of attitude control, then, be-
allow for thermal expansion of the propellants, comes twofold: (i) what is required to rotate the
and the time-dependent insulation weight. For vehicle through a given angle to some new atti-
Earth storables, sufficient insulation is also tude, and (2) what is required to maintain the
required to prevent freezing. vehicle in this attitude. The energy requirement
to accomplish these two basic maneuvers can be
calculated, and total attitude control require-
11.4 DESIGN OF REACTION CONTROL
ments for any space mission can be computed,
ENGINE SYSTEMS
by simply determining the total number of times
that these two maneuvers occur.
Spacecraft Attitude Control Requirements
Of the various attitude control systems, the
Spacecraft attitude control requires the appli- reaction systems are the most versatile and pro-
cation of torques about the three axes (yaw, vide a wide range of torque, up to high levels.
pitch, and roll) passing through the vehicle's They are universally applied to manned space-
center of mass. The problem of attitude control craft and in large unmanned vehicles.
implies rotation and/or stabilization of the vehi-

cle about these axes. Its operation consists of Operational Modes of Reaction Control Systems
two main phases:
If any reaction control system, the amount of
(1) Rotation of the vehicle over a given angle, thrust delivered on command must be very pre-
within a specified time cisely controlled, if overshoot or undershoot and
(2) Stabilization of the vehicle in a required hunting are to be avoided. Three basic opera-
tional modes are available:
position, within acceptable tolerance
limits 1. On-and-off, or "bang-bang" control.-This
"multiple start" system operates intermittently
The position of a vehicle, if left uncontrolled, as long as necessary at its rated thrust level. If
may not be the one required for efficient use of the on-and-off command signals are given as a
its main propulsion system. For example, a function of sensed position error only, the sys-
spacecraft may have to perform a series of tra- tem would tend to be unstable because of the
timelags that are present in all real systems. tolerances, maximum impulse per cycle, space
This situation can be corrected through the use storage and environment, systems integration,
of control systems which utilize both position and logistics.
error and rate of error change to time the on and For monopropellant reaction control systems,
off command signals. the system vacuum specific impulse varies from
2. Proportional control.-The operating thrust 140 to 165 seconds for 90 percent H202, and
of th(sreaction control system is varied accord- from 200 to 230 seconds for N2H4. For bipropel-
ing to the sensed error signals. lant systems using Earth storables or cryogen-
3. Repetitive pulse control.-This type of ics, the system vacuum specific impulse may be
system needs only position error sensing and estimated by applying an efficiency factor rang-
uses a system which delivers thrust in a con- ing from 90 to 94 percent for steady state, and
tinuous series of accurately reproducible impulse from 80 to 86 percent for transient operation
bits to assure orientation and stabilization of the (less than 25 milliseconds pulse width), of the
vehicle. The optimum thrust pulse outline should ideal specific impulse values given in table
be a square-wave. Control may be achieved by 11-1. The 90 percent H202 systems are applied
modulating pulse width, or the frequency of a up to a maximum total impulse of about 50 000
fixed pulse width, or by combining both methods. lb-sec and the N2H4 systems to the same or
somewhat higher values. Beyond this level, the
Selection of Reaction Control Engine Systems hypergolic Earth-storable bipropellant systems,
For reaction control engine systems, both which are in the majority, take over.
liquid monopropellants and bipropellants have
Optimization of Reaction Control System
been used. The selection of an optimum system
Operating Parameters
for a given mission is based primarily on total
systems weight (including propellants) versus For most reaction control engine systems, a
total impulse. As mission time increases, so pressure-feed system using stored inert gas as
does the propellant and tankage portion of the pressurant is employed because of its simplicity
total systems weight, while the weight of noz- and reliability. After the propellants and their
zles, valves, and plumbing remains fixed. A feed system have been selected, parametric
comparison of two reaction control systems of analyses of all engine system design variables
different fixed weight and specific impulse may are performed to establish the minimum weight
show the lower fixed-weight system to have an system, within mission requirements, which still
overall weight advantage, despite a considerably satisfies other important considerations. Major
lower specific impulse. However, sometimes the operating parameters affecting systems weight
lower fixed-weight system cannot be used, simply are: chamber pressure, pressurant storage pres-
because it has not enough thrust or cannot be sure, mixture ratio, and nozzle expansion area
operated more than a specific number of cycles. ratio.
If total impulse is the variable and total sys- Chamber pressure and nozzle expansion area
tems weight the evaluating criteria, a plot of ratio are foremost optimization parameters, since
total systems weight versus total impulse can be the weight of major system components is directly
used to compare competitive systems. If thrust related to and defined by them. Component
and number of restarts (operating cycles) are weights considered in the analyses should in-
also factors, added dimensions in presenting the clude the pressurant and its storage tank, the
comparison are required. Based on the selection fuel and oxidizer tanks, and the thrust chamber
criteria of minimum total systems weight, the assemblies. Optimum values for chamber pres-
three major factors to be considered as inde- sure range from 50 to 200 psia, and for nozzle
pendent variables for different space missions expansion area ratios from 25:1 to 60:1. The
are total impulse, thrust level, and number of optimum mixture ratio depends on propellant type
cycles. Selection of a reaction control engine and thrust chamber cooling method used. Opti-
system is largely dependent on these three re- mum design values for inert gas pressurant stor-
quirements, in addition to reliability considera- age pressure range from 3000 to 5000 psia (using
tions. Other important factors are performance 6 A1-4V titanium alloy forgings as tank material).
i ....
Basic System Design for Reaction Control helium at the required pressure level pressurizes
Engines the positive expulsion bladders in the oxidizer
and fuel tanks. Opening of propellant isolation
For N2H 4 and 90 percent H202 monopropel- solenoid valves allows the pressurized propel-
lant systems, the propellant tanks can be pres-
lants to flow through micronic filters, distribu-
surized by a stored inert gas system, as shown tion lines, and line isolation solenoid valves to
in figure 5-1. The propellant valves are located
the normally closed injector solenoid valves
downstream of the propellant tanks and are used
which control the operation of the individual
to control the propellant flow to the thrust cham-
reaction control thrust chambers. The system is
ber. In an H202 system, each thrust chamber shut down by removal of the start signal. This
has its own silver-mesh catalyst, similar to the
closes the propellant-isolation valves and dis-
one in the H202 gas generator shown in figure connects the reaction-control injector-valve
4-50. When H202 comes in contact with the
solenoids from the command system.
catalyst, a hot-gas mixture of steam and oxygen A relief valve sealed with a burst diaphragm
is produced, the temperature of which is a func-
protects the pressurant tank against overpres-
tion of H_O 2 concentration and its liquid temper-
surization. A high-pressure transducer monitors
ature. For N2H 4 monopropellant systems, some helium-tank pressure. In a hermetically sealed
kind of ignition source is usually provided be-
system, it could also be utilized to indicate the
cause the catalyst life with N2H4 is limited.
amount of propellants remaining in the tanks. A
The ignition source could be either an electrical
low-pressure transducer is located downstream
heating element, or a continuously operated
of the helium pressure regulators to detect regu-
nuclear energy device. Frequently, to simplify lator malfunctions. Relief valves sealed with
ig'nition requirements in a N2H 4 system, a com-
burst diaphragms are also provided for the pro-
mon gas generator feeds several nozzles.
pellant tanks to protect them against over-
Since the liquid H202 is cool until it decom-
pressurization.
poses, the components upstream of the catalyst
do not require high-temperature materials. How-
ever, feedback of heat from the catalyst chamber Systems Redundancy in Reaction Control Systems
and the nozzle to other components must be pre- For reliable reaction control engine operation,
vented, since it would not only damage these
redundancy may be provided in the following
components but could also initiate vaporization three areas:
or even decomposition of the stored propellant.
The H20 2 system has a temperature advantage 1. Redundancy within a subsystem.-Several
(maximum temperature 1500 ° F) over a system typical examples of redundancy within a sub-
using N2H 4, although the latter has a higher system are shown m figure 11-10. Two explosive-
specific impulse. In the N2H4 system, the high actuated pressurant start valves are used in
operating temperature of 1S00 ° F requires suit- parallel. Similarly, two pressure regulators are
able materials all the way from the generator to used in parallel, pilot selected by a three-way
the nozzles. solenoid valve. In each propellant pressurant
Figx_re 11-10 illustrates the basic schematic line, a "quad _ check valve assembly is used to
of a typical reaction control system using hyper- insure that propellants which may have permeated
golic Earth-storable bipropellants and helium gas through the positive-expulsion bladders will not
pressurization. Two redundant subsystems are reverse-flow into the common pressurant line
provided, each of which is physically and func- downstream of the pressure regulators. A similar
tionally independent. Each subsystem is acti- "quad" check valve assembly is installed down-
vated independently by an electrical command stream of each propellant tank to prevent reverse
signal to the explosive-actuated pressurant-start flow of propellants from the other propellant sub-
valves located in the helium pressurization line, system, or from the main propulsion system.
and to the various solenoid isolation valves. 2. Redundancy between subsystems.-Refer
After passing through a micronic filter, different again to figure 11-10. Here, redundancy between
pressurant-isolation valves, pressure regulators, two independent subsystems is provided by nor-
and the "quad" check-valve assemblies, the mally closed solenoid intersubsystem-propellant
-- - -- -- -- --- - I III I ii
_oo__O_T_O_
_,oo_
---_ --_U_NT,_?L_O_
NoR SSU NT,SOLAT,ON
, /T NS OCE
ACTUAT_O
P_ESSURANT
\ F----" _"
N.c.
EXPLOSIVE-..
START VALVE \J
_ \
_ _"
_ , _ I_° _
° 'o
3 WAY
i _ LOW
MODULAR CONTROL------X ! _ _ / ' PRESSURE

i, TRANSDUCER
VALVE ',_ I¢I I

l _ I CONTROL
MODULAR CONTROL__
._ FUEL TANK
PACKAGE _ _ __._
SOLENOID ISOLATION_----_ •.._-----_-- ,,mI.---MODULAR
VALVE I__ I
I
PACKAGE
ouc_
ECHEC_
_ : _r _ I
l_ _ CONTROL
"----- SOLENOID
_SOLATION
J VALV E
OUAD
CHECK
IDENTICAL --_" VALVE
ISOLATION VALVE
f FUEL SOLENOID
OXIDIZE SOLENOID
SUBSYSTEM
--Z_Z iSOLATION VALVE
SUBSYSTEM PROPELLANT
ASSEMBLY (iNCLUDING
r- THRUST CHAMBER
VALVE INJECTOR VALVES) TYP
m
PITCH YAW ROLL
[] MICRONIC FILTER _ BURST DIAPHRAGM ( RELIEF VALVE
@ SERVICE CONNECTION
Figure ll-lO.-Basic schematic ofa typical reaction control system.

valves. If one propellant subsystem should fail, systems are pressure fed and use a N204/MMH
these valves open and connect to the propellant propellant combination. The tasks of the Gemini
lines of the other subsystems. The distribution reaction control engine systems built by Rocket-
line solenoid isolation valves can isolate a dyne are high-altitude abort trim, attitude control
group of thrust chambers (pitch, roll, or yaw) in orbit and during reentry, and rendezvous and
shouid an individual thrust chamber in the group docking control. In addition, four solid propel-
malfunction. lant engines built by Thiokol are used for high-
3. Redundancy outside the system.-Redun- altitude abort and reentry retrofiring.
dancy outside the system may be provided by The reaction control engine systems for the
connecting to the pressurant and propellant sys- Gemini spacecraft (fig. 11-11) consist of-
tem of the spacecraft main propulsion system, (1) Sixteen 25-pound thrust chambers for
which must be capable of supplying one or both reentry control (also see fig. 11-14),
subsystems with pressurant and/or propellants. consisting of two complete propulsion
systems, with a pulsing requirement of 6
cycles/see and a minimum impulse
Packaging and Installation of Reaction Control bit= 0.25 lb-sec
Engine Systems
(2) Eight 25-pound thrust chambers for orbital
As with spacecraftmain propulsionsystems, attitude control-pulsing requirement: 6
the controlcomponents of a reactioncontrol cycles see, minimum impulse bit = 0.25
engine system can be modular packaged. The lb-sec
modular packages will incorporate related con- (3) Four 100-pound thrust chambers for trans-
trol components within the same housing (forging lation control in rendezvous and docking
or casting), as shown in figure 10-9. Typical maneuvers-pulsing requirement: 2
modular control packages for reaction control cycles/see; minimum impulse bit = 26
systems are indicated schematically in figure lb-sec
11-10. The thrust chamber injector propellant (4) Two 100-pound thrust chambers and two
valves are frequently designed as an integral S5-pound thrust chambers for longitudinal
part of the thrust chamber assembly (fig. 11-13). propulsion-pulsing requirement: 2
All-welded and brazed construction is preferred cycles/see; minimum impulse bit = 25
to prevent pressurant and propellant external lb-sec
leakage. The Gemini control thrust chambers are all
Figure 11-11 presents the installations of the located inside the vehicle, with the nozzles
reaction control engine systems used on a typi- trimmed flush with the outer skin. Thus, no
cal manned spacecraft, the Gemini capsule. All radiation cooling can be used. All thrust cham-
10('_._ IM_'LI5 _ _NAMB[I_ 12
bers are ablatively cooled.
Design of Reaction Control Thrust Chambers
For systems using hypergolic Earth-storable

bipropellants, the reaction control thrust cham-
bers may be effectively cooled by one of the
following methods:
(1) Ablative cooling
(2) Radiation cooling
(3) Ablative cooling with radiation-cooled
nozzle skirt
(4) Ablative cooling supplemented by regen-
erative or film cooling
Other methods, such as regenerative cooling and.
Figure 11-11.-Reaction control engine systems film cooling, can probably be used successfully
for the Gemini manned spacecraft. to a certain extent. However, none can compete
with ablative or radiation cooling in overall sys- equations (4-36), (4-37), (11-1), and (11-2) may
tem simplicity, minimum system pressure drop, be used. However, in the case of reaction con-
and minimum performance loss (film or transpira- trol thrust chambers designed for repetitive
tion cooling requires extra propellant flow). pulse, o; for intennittent operation, the heat-
Also, the advancement of the state of the art of transfer conditions may be quite different. Many
materials, and of analytical and design tech- studies and experiments have been conducted in
niques, has made both ablative and radiation this area, and theoretical and empirical correla-
cooling, rather reliable and practical cooling tions have been generated.
methods for reaction control thrust chambers. A study by Lee and Hahn' indicates that for
The design and construction principles for repetitive short-pulse operation, the char-front
ablative-cooled reaction control thrust chambers region in an ablative-cooled thrust chamber does
are not basically different from those for main not experience appreciable temperature fluctua-
propulsion systems. However, the small physi- tions, because of the attenuating effect of the
cal sizes (from 1- to 100-pound thrust), and the low thermal diffusivity of the char layer. The
operational modes (such as the pulse mode) of char front region therefore stays at the pyrolyzing
reaction control thrust chambers, require some temperature during the entire period of cycling.
special considerations. The char regression under pulse operation can,
For example, a typical roll control thrust therefore, be treated as a case of continuous
chamber is designed to produce thrust ranging firing with effectively reduced gas-side heat
from 1.6 to 2.5 pounds at chamber pressures from transfer. The results show that the char depth
S0 to 130 psia. The design throat area is 0.0120 based on equal cumulative firing time is a func-
in 2 (0.124-inch diameter). This throat diameter tion of percentage of the firing time over the
should be produced accurately and remain un- elapsed time. It increases with the decrease of
affected by erosion during firing. Generally, in the percentage of firiug, to a maximum value
small ablative rocket motors, the internal geom- several times that obtained with continuous fir-
etry remains essentially unchanged as ablation ing. At low percentages of firing, i.e., below
progresses. The thermal protection in this case approximately 5 percent burn (pulse width/pulse
is provided by the internal change from the cycle), the char depth drops again, because of
pyrolysis or decomposition of the plastic resin. the increasing p_oportion of radiation or convec-
This yields a porous char layer without any sig- tion losses from the outer skin surface of the
nificant dimensional changes. The transition chamber, to heat influx from the combustion
zone between the virgin ablative and the char is gases. In fact, at some critical pulse mode, the
referred to as the _char front." The ablation char regression ceases after an equilibrium char
rate, in this case, is not governed by a surface depth is attained.
regression, but rather by a regression of the During intermittent, on/off, i.e., multiple-
interface between the virgin ablative and the start firings, with off periods in between fiLings
char. However, if highest accuracy is required, of relatively long durations, the temperature at
throat inserts made of refractory ceramics, such the char front does not remain at the pyrolyzing
as silicon carbide, and combustion chamber temperature all the time. The char-depth pro-
liners made of graphite or ceramics, will further gression for a typical multiple-start system is
reduce throat-size changes due to erosion and shown in figure 11-12. The ascending straight
will prevent ablative fluid from being swept line represents char-depth progression for a con-
downstream from the combustion chamber and tinuous firing (100 percent burn). Assume that
deposited in the throat section. Silicon carbide the first firing cycle was terminated at "a."
can be machined to rather close tolerances by Since the char layer at this time is at a higher
grinding. The diameter of a throat can generally temperature than the pyrolyzing temperature,
be maintained with a tolerance of _+0.001 inch.
The methods of determining the char depth of
ablativereactioncontrolthrustchambers are LLee, J. C., and Hahn, J. R., "Regression Rate of
similarto those of the largersize chambers. Char Front in Ablative-Cooled Rocket Motor Under
Semiempiricalor empiricalequationssuch as Pulse Operation," AIAA Preprint 64-262.
eratively cooled jacket is provided for extra-long

durations (up to 30 minutes). Fuel is used as
the regenerative coolant to supplement the abla-
tive cooling. A segmented 3TA graphite liner is
used in the combustion chamber section; the
throat insert is made of silicon carbide. Abla-
tive materials are oriented at 90 ° and 0 ° ,respec-
tively, at various regions according to structural
requirements. Asbestos-phenolic is used for the
outer insulation. The injector propellant valves
TIME, SECONDS are brazed to the injector, which forms an inte-
gral part of the thrust chamber assembly.
Figure II-I2.-Combustion chamber char depth Figure 11-14 presents the 25-pound thrust
versus cumulative firing time of a typical chamber assembly for a reentry control system,
ablative-cooled reaction control thrust chamber.
also designed and manufactured by Rocketdyne.
Its design details are similar to the thrust cham-
ablation continuous until the temperature at the ber shown in figure 11-13. However, no auxiliary
char front drops below the pyrolyzing temperature regenerative cooling jacket is provided. Rather,
(point "b"). When the next cycle starts at _e," a ceramic liner (silicon carbide) is used in the
the char depth remains constant during warmup, combustion chamber section. The nozzle exit is
until pyrolyzing resumes through shutoff at "e. _ trimmed flush with the vehicle outer skin.
The various events may be projected onto the Total radiation cooling applied to reaction
100-percent burn line for convenience. The line control thrust chambers may pose the following
d'-e' in this case represents the soakback char problems:
depth following the second cycle. Subsequent
(1) Difficulties when inboard-installation of
cycles can be treated similarly. Note that in the
the thrust chamber is required
short second cycle, soakback charring is a mul-
(2) Thermal shielding requirements for com-
tiple of the charring during burning. In general,
ponents surrounding the thrust chamber
it can be seen from figure 11-12 that soakback
(3) Recrystallization of the chamber construc-
usually consumes a large percentage of virgin
tion metals in multiple-start applications
ablative and does not contribute to the useful
(4) Need for a larger thrust chamber, when
firing time. Even though some char preheating
operated at a relatively low chamber
time is gained during the subsequent firing, it
pressure
could hardly compensate for the loss of burn
time which would have been available had the
chamber been fired continuously. In this re-
spect, it is desirable to minimize the number of
soakbacks, particularly toward the last portion
of the virgin ablative.
Combinations of repetitive pulsings and mul-

tiple starts can be treated in the same manner as
described above, except that the "on" cycles
composed of repetitive pulsings will effect a
different gradient of the ablation curve, based on
/' /
the particular percent-bum value, instead of the t
INJIfCTOll J SILICON CARBIOE --f
continuous firing or 100 percent curve.

Figure 11-13 presents the design details of a Figure II-13. -Typical I O0-pound-thrust, ablative-
typical 100-pound-thrust, ablative-cooled, reac- cooled, reaction control thrust chamber assem-
tion control thrust chamber assembly for hyper- bly for hypergolic Earth-storable bipropellants,
golic Earth-storable bipropellants designed and designed and manufactured by Rocketdyne (re-
manufactured by Rocketdyne. An optional regen- generative cooling jacket optional).
INJECTOR _-_GLASS WRAP 10 percent tungsten alloy, electroplated with a

PROPELLANT X (STRUCTURAL) _ PARALLEL
cermet coating of chromium and zirconium
VALVE$--?_ MOUNTING _ ,_---ASBESTOS \WRAP
diboride, could be used for the chamber. An
/_ CAN __WRAP "_BLATIVE
alloy of molybdenum and 5 percent titanium,
coated with MoSi 2, may also be used.
Figure 11-15 presents a typical, totally
radiation-cooled reaction control thrust chamber
,.:
",Z PARALLEL WRAP Lgo"
...... \ ....NSER,
ORIENTED
assembly.
ber andCERAMIC LINER
nozzle,
It has
with
an integral
a thick
combustion
throat section.
cham-
The
ABLATIVE _ L A90/L;TIVE (SEGMENTED)
injector plate can be either bolted as shown, or

Figure ll-14.-25-pound thrust reentry control welded to the chamber. A heat shield is mounted
system thrust chamber assembly designed and aft of the injector to reduce radiant heat transfer
manufactured by Rocketdyne. to the injector propellant valves. The shield is
shaped to minimize reflection back to the thrust
A radiation-cooled thrust chamber, if properly chamber. The shield is attached to the mount
designed and applied, is probably the simplest, structure and is thermally insulated by ceramic
lightest, and relatively most reliable. The re- spacers.
quired working temperature of a radiation-cooled The injectors for reaction control thrust cham-
thrust chamber wall could be as high as 3400 °- bers using hypergolie Earth-storable propellants
3700 ° F for Earth-storable propellants operated are usually designed with a conventional fixed-
at or near their optimum mixture ratio, and at orifice, single-ring, unlike-impinging doublet
chamber pressures of 100 psia. Advanced refrac- pattern. A splash plate is often utilized to
tory materials such as the 90 percent tantalum- improve performance.
FUEL INJECTOR VALVE
OXIDIZER INJECTOR VALVE
INJECTOR
WALL THICKNESS
FUEL INLET
COMB. CHAMBER
DIA
THROAT DIAMETER
OXIDIZER'-_-
INLET
POLISHED REFLECTING INTEGRATED RADIATION-

TYPE HEAT SHIELD COOLED THRUST CHAMBER
-- FITTING
MOUNTI NG
STRUCTURE
Figure 11-15.-Typical total radiation-cooled reaction control thrust chamber assembly.

Design of Control Components for Reaction

Control Engine Systems
The various aspects of control component

design for reaction control engine systems are
similar to those for spacecraft main propulsion
systems, except that propellant and pressurant
flow rates are at much lower levels. For in-
stance, the propellant flow rates of a typical roll
control thrust chamber may be 0.0041 to 0.007
lb/sec (oxidizer), and 0.0021 to 0.0032 lb/sec
(fuel), to produce a thrust range from 1.6 to 2.5
--lr"--*.o o,,.
pounds. Besides the propellant injector valves,
all other control components, such as pressure Figure 11-16.- Typical injector propellant valve
regulators, check valves, and vent valves, are used with reaction control thrust chambers in
similar to those of other systems. pulsing applications, designed and manufac-
In addition to high reliability and suitability tured by Rocketdyne.
for the space environment, pertinent design re-
quirements for the injector propellant valves of thrust chamber assemblies shown in figures
reaction control engine systems are: 11-13 and 11-14 for pulsing applications. The
(1) Fast (as short as 2 milliseconds) and pre- valve is a fast-acting, solenoid-operated, direct-
cise response when opening or closing actuated type. The propellants flow through the
(2) Tight shutoff, with no afterdribble solenoid core at nearly constant velocity, result-
(3) Low power consumption ing iu low-pressure drop. The valve incorporates
Most injector valves are solenoid operated. a hermetically seaied electrical coil, and a
The needle-type propellant valve design shown metal-ball-type seal. The latter is crimped
in figure 7-43 is suitable for integral injector soiidly into tl_e nose of the armature and engages
valve assemblies, with both valves actuated by a conical metal seating surface. Positive seat-
a common solenoid.
ing is obtained by lapped fits. The inlet port
Figure 11-16 shows another typical injector contains a filter to protect against contaminants.
propellant valve, designed and manufactured by The valve is capable of operating reliably in the
Rocketdyne. It is used with the reaction control temperature environment of space.
INDEX
A-1 stage engine, 64, 65, 66, 67, 68, 86, 96, 102, A-2 stage engine (continued)
109, 110, 129, 130, 193, 201, 217, 218, 219, 222, starting sequence, 72
224, 249, 250, 261, 295, 296, 303, 364, 378,391, steady state operation, data for, 166
392, 393, 395, 397, 398, 399, 4I_ system operation, 70
thermal conductance, 103
alternate turbine design, 257
assembly design layout, 261 thrust chamber configuration, 95, 96
bellows, 378 turbopump, 199, 389, 392
C* correction in performance, 398-399 venting, 166
centrifugal pump, 224 A-3 stage engine, 72, 73, 74, 75, 76. 86, 97, 156, 167
chamber pressure vs. engine thrust, 395 chamber product gases, values for, 86
chamber product gases of, values, 85, 86 engine and propulsion system operational sequence,
cutoff sequence, 68 75
design values, 393 general description, 72, 73

engine performance diagram, 66 helium system, 156
general description, 64, 67 preliminary design layout for 48K, 75
ignition monitor valve, 303 system operation, 73, 74
impulse turbine, 249, 250 thrust chamber configuration, 95, 97
influence coefficient, 397-398 thrust load, 76
injection momentum ratio, 129 A-4 stage engine, 74, 75, 76. 77, 78, 95, 97, 121, 144,
injection velocity, 130 154, 156, 158, 159, 163, 268, 269, 270, 324, 325,
main oxidizer valve, 295-297 338, 342, 344, 392
oxidizer pump, 217-219, 222, 364 burst diaphragm on oxidizer tank, 344
chamber nozzle extension, 121
packaging, 399
preliminary layout for 750K, 67 chamber product gases, values for, 86
propellant flow design characteristics, 391 comparison of various tank pressurization systems,
starting sequence, 67 158, 159

static-firing measurement list, 412 engine and propulsion schematic, 78
system operation, 67 engine mixture ratio control loop, 269
thermal conductance, 102 estimated pressure drops in oxidizer tank, 158
thrust chamber, 95, 96, 109, 110 feed system, 320
turbopump, 193, 199, 201, 249, 392 general description, 76
A-2 stage engine, 37, 38, 39, 68-72, 76, 96, 111, 112, helium system, 156
129, 131, 154, 165, 166, 199, 233, 236, 238, 269, internal configuration for thrust chamber, 97
389, 392, 399 main-stage thrust throttle control loop, 268
accessory weight history, 37, 38, 39 omission, 64
coaxial shell chamber design, 111, 112 operating parameters for 7.5K, 77
control of propellant, 269 operational sequence, 79
cutoff sequence, 72 oxidizer tank, 154
fuel pump, 233, 236, 238 propellant control system, 268

general description, 68-70 propellant mixture ratio control loop, 269, 270
injection momentum ratio, 129, 131 propulsion system, 78, 338, 392
injection velocity, 131 relief valve, design data on, 324, 325
injector data, 130 storage pressure requirement, 159
nozzle expansion area ratio, 76 system operation, 76
operating parameters (150K), 69 tankage, 342
packaging, 399 thrust chamber configuration, 95
payload weight of, 38 Ablative cooling, 118, 119, 448

Acceleration, isentropic and nonisentropic, 13
preliminary layout for 150K, 71
schematic for tank pressurization of propellant, 165 Accelerometers, 411, 415
Accordion motion, 420
sequence diagram, 71
451
Acoustic chamber liners, 149, 150 C* correction, 398-399

Actuators for thrust vector control, 272.279 Calculations for 4 stages of Alpha vehicle, samples
Additives for propellants, 19 of, 63-79
Aerodynamic spike nozzle, 93 Catalysts, 141
AirReseaxch Manufacturing Co., 349 Catalytic bed of monopropellant gas generator, 133,
ALBM, 26 134
Alloys, 61-62 Catastrophic failure, definition of, 44
aluminum, 61, 76, I18 Cavitating inducers, 210
cobalt base, 62 Center of gravity, data sheet for typical, 39
copper base. 62 Centrifugal pumps, 176, 177, 196, 220-225
magnesium, 61 Chamber, 4, 13, 14, 15, 149, 200
nickel-base, high temperature, 61 Chamber wall cooling system, demands on, 34
special, 61, 62 Change control, 51
Alpha vehicle, 4-stage, 63, 95 Char front, 447-448
Altitude control requirements, 442 Char-rate analysis, 120
Altitude thrust coefficient, 17 Chemical reaction products, systems using, 167-171
Alumina, palladium-impregnated, 141 Chugging, 146
Aluminum-alloy tubing, minimum bend radii for, 360 Closed-loop mixture ratio control, 269-270
Aluminum alloys, 61, 76 Clustering of liquid propellant, 415-419
Aluminum on ablative liner, 118 Coaxial injector, 127
Ambient air pressure, effect on performance of, 1, 14 Coaxial shell thrust chamber, 109
Apollo LEM, 435 Cobalt base alloys, 62
Apollo moon-landing program, U.S., 430 Colburn calculation of gas-side heat-transfer coef-
AR systems, 141 ficient, 100
Asbestos on ablative liner, I18 Collier's magazine, 415
ASI (augmented spark igniters), 140, 435 CoIumbian, 62
Assurance of reliability, 43 Combustion chambers, 4. 5, 123, 148, 149, 150
Astronautics magazine, 127 baffles, 149
Atlas ICBM, 329 design as factor in preventing triggering, 148
Atlas Mercury, 64 disturbances, 150
Atomization, evaluation of, 131 divergent wall gap, 149
Augmented spark igniters (ASI), 140, 435 function, 5
Axial pumps, multistage, 177-178 gas flow processes, 4
gas flow through, 5
heat protection, 123
shape, 5
velocity of liquid propellants at injection plane, 5
B- 10 life, 258 Combustion devices, failures, definition, 47
Baffled injector, 122 Combustion instability, 144-147
Baffles, combustion-chamber, 149 design approaches toward control, 147
Ball bearing centrifugal load DN limits, 259 German type, 144
Ball-type valves, 298, 299 high-frequency, 145
Bung-bang control. See On-off type control intermediate-frequency, 146
Bartz calculation, 100:101 low-frequency, 146
Bell nozzle, 90, 91 types, 144
Bellows, 292, 293, 375-379 Combustion product gases, 12, 197
Berkeley, Edmund C., 428 Combustion stability, 122, 147, 150
Bernoulli's energy equation, 280 injector, 122
Beta angle, definition of, 128, 129 limits, 147
Bipropellant combinations, 25-27 rating, 150
Bipropellant gas generators, 134, 171, 181 Combustion timelags, 146
Bipropellant injector, 126 Complex redundancy, 46
Bipropellants, 19, 132, 138 Components and mounts, interconnecting, 353-381.
Bleed lines, cryogemc propellant, 356 See also Interconnecting components and mounts
Bootstrap, 67 Conax Corp., 327
Breadboard engines, 64 Confidence level, relation to reliability, 42
British systems, 141 Conical nozzle, 90
Buddy plan, 45 Connectors, terminology for, 404, 405
Burnout weight of engine, 35 Conservation of energy,, principle of, 4, 5
Burst diaphragms, 326 Conservation of matter, principle of, 5
Butterfly valves, 294-298 Control of manufacturing and material defects, 48
INDEX 453
Control systems, 263-327 De Laval nozzle, 6, 8, 89, 239

calibration, 266 Density impulse of propellant additives, 20
checkout and test, 267 Derivative pressure feedback (DPQ) valves, 273-274
closed-loop, 263-265 Derivative type control, 264
cutoff, 265 Design change checkoff sheets, 51
dynamic seals for fluid-flow components, design Design Check-Off Sheet, purpose of, 50
for, 289- 294 Design integration of engine systems, dynamic analy-
engine thrust level, 267-268 sis for, 384-390
fixed-area-type regulators, design of, 307-309 Design quality, importance of, 50, 51
fluid flow components, design for, 280-294 Design reviews, 49-50
gas pressure regulators, designof, 315-321 Diametral-squeeze-type O-ring seals, 290
liquid flow and pressure regulators, design of, Directed nonexplosive gas flows, 150
321-322 Discontinuity stresses, 339
main stage duration, 265 Divergent wall gap, chamber, 149
miscellaneous valves, 325-327 DN value, 258
open-loop, 263 Dome-loaded gas pressure regulator, 3 19-320
orifice locations and sizes, 266 Doublet injector. 125, 127
pilot valves, design for, 301-307 Dry weight of engine, 35
pressure relief valve, design for, 322-827 D.S.D. Co., 367
propellant-mixture ratio and utilization, 268-272 Du Pont, 383
propellant valves, design of, 294-301 Dual bipropellant gas generator system, 172
safety, 265 Dual combustion, 181
servovalves, design of, 309-315 Dual shaft turbopump, 182
start, 265 Duct design, 373, 374
thrust vector actuators, 272-279 Dump cooling, 114
Coolant side heat transfer, 105, 106, 107 Dynamic analysis in engine system design integra-
Cooling system for thrust chamber, 98-99, 100-102, tion, 384-390
104- 121 Dynamic seal, design, 258
ablative, 119 Dynamic stability, 148
circular tube wall of regenerative, 108
coaxial shell, 109
considerations in selecting, 99 E-D (expansion-deflection) nozzle, 92, 93
coolant side-heat transfer, 105
EDS. See Emergency detection system
correction factor, variation across boundary layer,
Eiastomer rings, 366
101 Electrical failures, definition of, 47, 48
dump, 114
Electrical hydromechanical component analogies, 275
film, 115
Emergency detection systems (EDS), 45, 421
gaseous film, 116, 117
Energy losses during pump operation, 194-195
gas-side heat transfer, i00
Engelhard Industries, Newark, N.J., 141
heat transfer schematic for regenerative, 104
Engels, D. A., 275
liquid film, 115, 116
Engine components, failure modes, 45
radiation, 120, 121
Engine design, 1, 3-4, 16, 28, 31-62, 96-97, 181, 416,
regenerative, 104- 114
419-421
techniques and selection, 98
A-I, A-2, A-3, A-4 stage, thrust chamber configura-
thermal resistance of carbon wall deposit, 102
tions for, 96-97
transpiration, 117, 118
areas for increase in quality, 51
Copper base alloys, 62
availability, 50
Cost of vehicle design, 50
burnout weight, 35
Coulbert, C. D., 433
departures from basic simplicity, 28
Critical failure, definition, 44
detailed procedures, 51-55
Critical pressure ratio between throat and chamber, 6
dry weight, 35
Cryogenic propellants, 19, 68, 165, 166, 167, 187,
duration of run, 32-34
271, 279, 288, 291, 331, 332, 336, 345-348, 356,
envelope (size), 4_
422-424, 429, 440
flight run duration, 33
Curtis turbine, 180
gas exhaust velocities, sea level and in space,
Cutoff impulse, 34
sample calculation, 3-4
manual for, 419
parameters, major, 31-62
Damping devices, 149 parameters, ranges of, 16
Decay deviations, 427 performance as function of altitude, 32
Deferred failure, definition of, 44 performance parameters, 32
Engine design (continued) Error, human. 49

quality, importance of, 50, 51 EWF (equivalent-weight factor), 187
shape, most used laws and formulae for, 1 Expansion of gas, isentropic and nonisentropic, 13
simplicity, 32. 48
stress analysis, 56
thrust in space, sample calculation to determine, 3-4
Failure, 44-48, 51, 58, 59
thrust level, 31-32
catastrophic, 44
tools for improving quality. 51
combustion devices, 47
typical sequence diagram, 55
criteria, 59
wet gimbaled weight, 87
critical. 44
wet weight, 37
deferred, 44
Engine failure sensing and shutoff system (EFSS).
electrical, 47
420-421
fatigue, 46
Engine Out (EO) capability, 416
functional, 45, 46
Engine out, definition, 44
mission, 44
Engine out source, 181
modes of, 44, 58
Engine system, 1-30, 42, 43, 55, 59, 60. 63-79, 144,
order of, 44, 45
148
reports, 51
A-l, A-2, A-g, A-4 stages for Alpha, 63-79
unexplained, classification of, 48
basic elements, 21
Failure modes of engine components, definition, 45
components, 9.8
Fatigue failure, definition, 46
function, 1
FBM, 26
introduction, 1-30
Feed system hydraulic capacitor, 149
low alloy steels. 60
Feed system perturbations, 150
model of thrust equation, 3
Fiber-glass filament-wound propellant tanks, 348-349
operating characteristics in prevention of triggering,
148 Film cooling of thrust chamber, 115
Flame holding characteristic, 146
oscillations, effects of, 144
Flame propagation velocity, 146
pressure environment as function of altitude, 2
Flange joints, design, 360-370
reliability of design, 42, 43
Flexitallic Gasket Co., 367
selection of materials, 59
Flow (of propellant), 9- I0
size, effects of, 42
areas, computations for, 10
thrust chamber nozzle, specific function of, 2
computations, 10
thrust equation (general), 2
specific volumes. 9
thrust generation, momentum theorem in basic equa-
static pressures, computation for, 9
tion for, 1
temperatnres, computation for, 9
typical pressurized gas feed, 28
velocities, computation for, 19
typical schematic, 55
Flowmeters, 271. 411, 414-415
typical turbopump feed, 29
Fluid, ideal, basic flow characteristics of, 280
Engine systems design integration, 883-428
Fluorine, in A-3 stage engine, 72
calibration, 390-394
Formula(s), 1, 7, 8, 18
clustering of hquid propellant, 415-419
design of shape and prediction of engine behavior, 1
dynamic analysis, 384-390
effect of pressure on thrust, 13
electrical, 403-411
ideal gas flow, 7, 8
engine-to-vehicle interfaces, 383-384
Fox Valve Development Co., 300
instrumentation, 411-415
Franms-type impeller, specific speeds, 190
mechanical development, 399-403
Frozen vs. shifting composition of gas flow, 21
performance characteristics, 394-399
Fuel additive blender unit. 183, 184
simplicity, 383-384
Fuel additive subsystem, 182, 184, 185
Engine-to-vehicle interfaces, actuator systems. 276-
Fuel characteristics. 34, 35
278
Functional failures, definition, 45
crosstalk and spring rate, 277
Fusible wire links, 142
engine installation and alinement, 276
hydraulic system, 277, 278
loads, 276-277
performance evaluations, 279 Gas constant, effect on performance of engine, 14,
secondary injection of matter, 279 15, 34
Envelope, definition, 42 Gas feed systems, 151, 167
Environmental effects on stress, 56 Gas flow, 4-8
Enzian missile, 140 combustion chamber, 4
Ermeto flareless tube fitting, 358 formulas for ideal, 7, 8
INDEX 455
Gas flow (continued) Hydrogen (continued)

parameters and terms applicable in thrust chamber, 4 drawbacks, 68
terms used in calculation, 5 gas tapoff systems, 136
thrust chamber, ideal conditions, 4 Hydrogen peroxide, 133, 138, 140. See also Catalysts
Gas flow orifices, 309 Hypergolic fluid, 167
Gas-generating devices, 131-136 Hypergolic ignition, 136, 138, 139, 140
classification, 132 Hypergolic slug, 139
design objectives, 131-132 Hypergolic vs. nonhypergolic, 432
solid propellant, 132 Hypergolics, definition of, 19. See also Noeggerath
Gas generator igniter, 138 and hypergolics, 138
Gas generators, 132-136, 171, 172
Gas generators, liquid bipropellant, operating charac-
teristics, 136
Iacobellis, S. F., 93
Gas law, perfect, 4
ICBM, 26, 27, 33
Gas pressure relief valves, 32.2-323
Igniters, 136, 137
Gaseous film cooling of thrust chamber, 116, 117
Ignition detection, 141
Gaskets, 361
Ignition devices, 136-143
Gate-type valves, 300-301
augmented spark, 140
Gears for turbines of turbopump, 260-261
hypergolic, 138- 140
Gemini capsule, 446
hypergolic slugs, 139
General Controls pilot valve, 302
igniters, 137
German A-4 engine (V-2), 138, 141, 142, 143, 156, 181
indirect, ].43
German efforts, Peenemunde, 139
optical, 142
German systems, 141. See also Catalysts
pressure- sensing. 143
German V-2 (1942), ratio of engine weight to thrust, 36
pyrotechnic. 137
GFE. See Government-furnished equipment
resistance wires. 143
Gimbal angle, 425
spark plugs, 139
Gimbal mounts, 379-381
special designs, 140
Government-furnished equipment (GFE), 427
visual, 142
Gravity, effects of, 4.14
Impeller, centrifugal pump, definition, 177
Greenewalt, Crawford H., 383
Impeller rotor and stator, 236
Ground support equipment (GSE), 267, 426
Impeller rotor, ope[ation, 226
Growth factor of single stage or multistage vehicle, 40
Impellers for centrifugal turbopumps, 204-207, 230-231
GSE. See Ground support equipment
Impellers, specific speeds, 190- 191
Impulse, specific, 10
Inducer, centrifugal pump, definition, 176
Hahn, J. R., 447 Inducer, typic_l, 211
Hart, T. J., 338 Influence coefficients for engine. 397-398
Hatch and Papell equation, 116 Inhibited red fuming nitric acid (IRFNA). 343, 344
Heat combustion chamber, protection, 123 Injecting methods, throttleable, 128
Heat exchanger propellant evaporation system, 165 Injection pattern, 131
Heaters, 408 Injection pressure drop. 128
Helium in A-3 stage engine, 73 Injectors. 83, 113. 121-131. 123
Helium in stored gas systems, 156-158 baffled-type, 122
Helium thrust chamber systems, 162 bipropellant. 126
H-F (horizontal-flow) nozzle, 92 coaxial-type, 127
High-frequency instabilities, 145, 146 concentric ring. 124
High vacuum, characteristics of, 434 configurations, 124- 127
Hilsch separation tube, 349 doublet. 125. 127
H-Q curve, 189 experimental evaluation, 131
H-Q, efficiency, and required power curves of cen- heat protection in combustion chamber, 123
trifugal pump, 196 hydraulic qualities, 123
Human error, 49 impingement patterns, 125. 127
Hydraulic capacitor, feed system for, 149 integral face plate, 125
Hydraulic capacitor, schematic of isolation type, 149 manifold, 113
Hydraulic qualities of injector, 123 objectives in design, 121-123
Hydrazine, as a monopropellant, 133, 171 parameters, 127- 131
Hydrazine-hydrate, 188, 141 premixing-type, 127, 128
Hydrogen, 68, 107, 136, 167 quintuplet-type, 125, 127
coolant properties, 107 ting-slot-typ e. 127
critical pressure considerations, 167 self-impingingotype, 125, 127
Injectors (continued) Liquid-hydrogen-fueled engines, operating regions of

showerhead-type, 125, 127 pumps, 226
special requirements, 123 Liquid hydrogen tanks, 345-348
splash plates, 127 Liquid monorepellant gas generators, 133
throttleable-type, 127 Liquid propellant, definition, 18
thrust chamber, 83 Liquid propellant rocket combustion chambers, gas
triplet-type, 125, 127 flow through, 5, 9, 10
Instability, 146, 147 Liquid Rocket Engine Design Check-Off Sheet, 52-54
Instability of combustion. See Combustion instability Lip seals, 291, 292
Instrumentation of engine, additional elements, 48 Loads on rotor blades of turbine, 244
Insulation for cryogenic propellant tanks, 345-348 Low-frequency instabilities, 146
Integral-type control, 264 L-star(L*), 86, 87, 131, 134
Interconnecting components and mounts, 353-381
assemblies, 357
brazed joints, 370-371
Magnesium alloys, 61
cryogenic propellant bleed lines, 356
Main propellant tank direct injection pressurization
ducts, 372-379
system, 172, 173
flange joint, 360-370
Man Rating, definition, 44
gimbal, 379-381
Manufacturing and material engine defects, control
hydraulic lines, 357
areas defined, 48
pneumatic supply lines, 356
Mark III turbopump gears and bearings, 184
propellant supply ducts, 353
Mark V turbopump, 186
propellant-tank-pressurization lines, 355-356
Marman Division of Aeroquip Corp., 374
purge lines, 356-357
Materials for engine part, selection of. 59, 60
seal drain lines, 356
Messerschmitt Me- 163, 141
thrust, 357
Messerschmitt rocket fighters, 140
tubings and tube fittings, 358-360
Metals, criteria for use with liquid hydrogen, 60
turbine-drive, high-pressure, hot-gas ducts, 357
Meteoric bombardment, 434
turbine-exhaust, hot gas ducts, 357
Mission coast duration, effect, 432
turbopump and other, 357
Mission failure, definition, 44
Interstage length, 425
Mission success, definition, 44
IRBM, 26, 27
Mixed-flow type impeller, specific speeds, 190, 210
IRFNA (Inhibited red fuming nitric acid), 343, 344
Mixture ratio of fuel, 34
Isentropic flow process, relationship between any two
Moment of inertia, data sheet for typical, 39
points in, 5
Momentum theorem, 1, 2
Monergols. See Propellants, mono-
Monopropellant gas generator schematic, 133, 181
J-Box, 409 Monopropellants, 18, 19, 132, 133, 171
J. Spacecraft and Rockets, 433 Moody, L. F., diagram, 283, 284
JTA liner. 438, 439, 448 Multiceilular design, 416
Jupiter rocket, 400 Multiple-engine use, 428
Multiple start, 442
Multistage vehicle, growth factor equations, 40
Kaplan method of error triangles, 210

Kinetic energy, conversion from thermal energy, 1
NASA, S-1 first stage booster for Saturn I, 416
Needle-type valves, 301
Newton' s law of viscosity, 282
L*, (characteristic length of combustion table), 87. Nickel-base alloys, 61, 76
See also L-star Noeggerath and hypergolics, 138
L*, effect on c* value of thrust chamber, 87 Noncomplex redundancy, 46
Laws, design and predict rocket behavior, 1-4 Noncryogenics, 167, 271
Lee, J. C., 447 Nonmetallic materials, uses, 62
LEM (lunar excursion module), 430 Nonpropellants, systems evaporating, 167
Li quid bipropellant gas generator, 134-136 North American Aviation, Rocketdyne Division, 136
operating characteristics, 136 North American X-15, 429
valve assembly, 134 Nozzles, 2, 4, 6, 7, 8, 13, 14, 15, 89, 90, 92, 93, 94,
Liquid film cooling of thrust chamber, 115, 116 95, 97, 98. 239, 307
Liquid flow regulators, 321-322 aerodynamic spike, 93, 94
Liquid hydrogen, selection of metals for use with, 60 annular, 92, 94, 95
INDEX 457
Nozzles (continued) Performance. injector, 122

area expansion ratio, effect on performance, 15 Performance parameters for engine, 10, 15-18
bell, 90, 91, 93, 98 correction factors and magnitudes, 15. 16
conical, 89. 93 measurement and dimensions, 10
convergent section. 13, 14 sample calculation, 16, 17, 18
De Laval type. 6, 7, 8, 89 Perturbations, feed system, 150
divergent section, 14 PFRT. See Preliminary flight rating tests
E-D (expansion-deflection), 93 Pilot valves, control, design. 301-306
expansion area ratio, 89 Pneumatic pressure regulator, 317
gas flow through, 4. 6 Pogo stick effect, 420
incorrect length and jet separation, effect on thrust, Poiseuille's equation, 284
7 Poisson's ratio, 108, 109, 110, 163, 209, 343, 375
liquid flow througb, 307 Poppet-type valves, 299
overexpanded, 6, 7 Position indicators. 407
performance comparison, 94 Postwax engine, 36
pressure losses at inlet, 6 Potassium permanganate, 141. See also Catalysts
prime functions, 6 Power sources, turbine, 181
radial in-flow and out-flow, 92 Prandtl-Meyer turning angle, 94
reverse flow (R-F), 92 Prandtl number, 100, 106, 118
specific function in thrust chamber, 2 Preliminary flight rating tests, 32
turbine, 239 Pressurant requirements. 152- 156
NPSH (net-positive suction head). 166, 176, 188, 192- Pressurant use factor, 158
194, 210, 346 Pressure-compounded impulse turbine, 180
Nuclear radiation, 434 Pressure drop, effective injector, 131
Nusselt number, 100, 106 Pressure environment as function of altitude, 2
Pressure feedback (PQ) valves. 273
Pressure, formula demonstrating effect on thrust of, 13
Pressure gages, principal types of. 412-414
Off-design characteristics of turbopump, 203 Pressure losses at nozzle inlet, 6
Off-nominal conditions, engine performance variations, Pressure-sensing devices, 143
396 Pressure switches and transducers, 411
Offset turbopump, 182 Pressurized-gas propellant-feed systems, 166-167,
One-and-one-half stage vehicles, 416 170, 171, 173-174
On-off type control, 264, 442 A-4 stage, 174
Open-loop mixture-ratio control, procedures, 268-269 applications, 166-167
Optical detection, 142 criteria for selecting, 173, 174
Optoline. 140 evaporating nonpropellants, 167
Order of failure, definition, 44, 45 liquid propelled, 170. 171
Orifices, liquid flow through, 307 using products of chemical reactions, 167
O-rings, 45, 83, 289-291, 292, 298, 300, 303 Pressurizing gas turbulence, 152
correlations in design, 289, 290 Pro rained mixture ratio control (PMR), 270
flange joints, 366 Pro Jellant combination, 129, 148
friction, 290 Pro)ellant evaporation systems, 165-167
poppet-type, 291 Pro)ellant feed system, triggering prevention, 148
seal failure in. 45 Pro)ellant feed system, centrifugal pumps, 176. 177
selection, 290-291 Pro)ellant feed system perturbations, 150
squeeze-type, 291 Pro)ellant flow design, pressure feed engine, 391
surface finish, 290 Pro _ellant fraction, definition of, 35
valve seat, seals for, 291 Pro)ellant management system, 270
Over-stress and over-strain, definition, 47 Pro)ellant mass fraction of complete vehicle, defi-
Oxidizer pump inducer, 215 nition of, 11
Propellant storage, system design for, 440-442
Propellant tanks, 329-352
cryogenic, 345-348
Packaging, 399-403 design, basic, 329-331
Pancake turbopump, 182 design, general, 332-336
Parts list, rocket system, symbols, 28-29 expulsion under zero gravity or oscillatory g load-
Payload, specific impulse effect on, 11 ing, 349- 352
Payload weight vs. growth factor, 40 fiber-glass filament-wound, 348-349
Peenemunde, 139, 140 pressurant diffusers. 349
Perchlorate-type fuel, 138 storable liquid, 343-345
Perfect gas law, 4 structure, 336, 343
Propellant utilization (PU), 268, 269, 270, 272, 311, Reliability, 42-49
336, 393 definition, 42
Propellant vaporization, 152 design reviews, 49
Propellants, 18-20, 22-23, 188, 203, 431-432 implements for obtaining highest, 43
additives, 19, 20 pointers for obtaining, 42, 43
bi-, 19, 26, 27 Repetitive pulse control, 443
cooling and other characteristics, 432 Resins, pyrolysis, 118
cryogenic, 19, 24 Resistance coefficients for fluid-flow control com-
density, 431 ponents, 286
flow of, engine start, 203 R-F (reverse-flow) nozzle, 92
fluid properties, 188 RF signals, 68, 138
ignition characteristics, 431 Ring slot injector, 127
mono-, 18, 25 Rocket thrust, general, equation and model for, 2, 3
operating temperature, 431 Rocketdyne, 448, 449, 450
optimum mixture ratio for additives, 19, 20 Rocketdyne AR-1, 141
performance and physical properties, 20 Rocketdyne Atlas ]CBM, 295
selection, 18-20 Rocketdyne Division of North America, 37
specific impulse in evaluation, 431 Rocketdyne LR79-NA-11 engine, 403
storable, 19, 22-23 Rocketdyne Naflex, 368-370
Propeller type impeller, specific speeds, 191 Rotor blades of turbine, main loads, 244
Proportional-type control, 264, 443 Rotor-stator assembly, centrifugal pumps, 176, 177
Propulsion system, definition, 28 RP fuel, 141
Pulses, 150 RP-1 fuel, 64, 109
Pump, cavitation, 192 Run duration, 32-34
Pump developed head, 189
Pump head coefficient, 189
Pyrolytic graphite, 119
Pyrometers, 142 Safety features, disadvantages, 43
Pyrophoric fluid, 19 Safety margins in design, 334-335, 428
Pyrotechnic igniters, 137, 138 Sample calculations. See Calculations for 4 stages
of Alpha vehicle, samples of
Saturn I first stage (S-I) booster for, 416
Saturn V, 329, 416, 425, 430
Scheduling, availability of design in, 50
Qualification tests, 32
Schmetterling AA rocket, 140
Quality control, areas, 48
Sealing, 361
Quintuplet injector, 125, 127
Secondary injection thrust vector control (SITVC),
279, 280
Self-impinging injector, 125, 127
Servovalve flapper, 319
Radial-flow impeller, minimum basic design elements, Servovalves, 309-315
208 Shepard, Cdr., 141
Radial in-flow nozzle, 92 Showerhead injector, 125, 127
Radial out-flow nozzle, 92 Sieder-Tate equation. 106
Radial-type impeller, specific speeds, 190 Simplicity of engine design, 32. See also yon Karman
Radiation cooling of thrust chamber, schematic, 120, Single gas generator-helium system, 172
121
Single gas generator system with injection cooling,
Rannie equation, 118 171
Rao, G. V. R., 91 Single-geared turbopump, 182
Rating stability, 150 Single-stage vehicle, growth factor equation, 40
Rayleigh flow process, 6 SITVC. See Secondary injection thrust vector control
Reaction control system, 429-430 Skinner pilot valve, 302
Reaction Motors Division of the Thiokol Corp., 429 Solid propellant gas generators, 132, 133
Reaction turbine, 180 Solubility of pressurizing gas, 152
Redstone rocket, 33, 34, 64, 141, 142, 400 Space engines, design, 429-450
Redundancy, 46, 142, 444-446 general applications, 430-434
Refrasil-filled chambers, 119 main propulsion systems, 435-442
Regenerative cooling, 105-114 principal applications, 429-430
Regulating devices, fixed-area, 307-309 reaction control, 442-450
Regulators, gas pressure, 315-321 Space system operational requirements, 433
Regulators, liquid flow, 321-322 Spark plugs, 139, 140
Relays and switches, 406-407 Special alloys, 61, 62
INDEX 459
Specific heat ratio, 9, 15 Thiokol, 446

effect on performance of, 15 Thoma parameters, 193, 194, 211
values for functions of, 9 Thor rocket, 64, 400
Specific impulse, 10, 11, 12 Three-stage vehicle system, weight increase calcula-
Specifications improving engine design quality, 51 tions, 41
SPGG. See Solid propellant gas generator Throttleable injector, 127
Spike nozzle, 92 Thrust, 7, 12, 15, 31, 33. 34, 81-83, 1S1
Splash plates injector, 127 coefficient, 12
Spool-type servovalves, 313, 314, 315 decay, 34
Spring mass effect, 420 effect of incorrect nozzle length and jet separation,
Squeeze-type O-rings, 291 7
Stage burnout velocity of single-stage vehicle, equa- optimum, ambient pressure, 15
tions, 39, 40 total requirement, 31
Stage separation, basic possibilities, 425-4°6 Thrust chamber
Stainless steel in pressurant storage vessels, 162 assembly, 82
Stainless steel on ablative liner, 119 basic cylindrical combustion in. 88
Start (thrust buildup), characteristics, 33 basic elements. 81-83
Stator, primary functions of, 177 bleed, 181

Stay time of burning gas in combustion chamber, 34 characteristic velocity, 83
Steels, stainless, 60, 61, 76, 119, 132, 162 coaxial shell, 107, 109, 110
Stefan-Boltzmann radiation heat transfer constant, 121 coefficient, 84
Stoichiometric mixture ratio, 34 combustion data (theoretical), 84, 85
Storable liquid propellants, 19 configuration layout, 86
Storedgas systems, 156- 165 cooling, 98- 121
commonly used configurations, 156-157 deposits on walls, 101. 102., 103, 104
design of components, 162 designs, 13, 14
requirements of calculations, 157-162 film-cooled, I17
Strain gages, 415 gas tapoff systems, 136
Stress analysis, 56-59 geometrical shapes, 88
experimental, 58, 59 heat exchanger, 162
failure modes and criteria, 59 ideal, data, 8
selection of materials and dimensions, 58 injector, 82
typical steps, 56 injector design, 121-131
working loads and environmental effects, 56, 57 L*,effect on c* value, 87
Structural integrity of injector, 123 model of, 3
Suction speclhc speed, 192 nozzle expansion area ratio, 89
Sweeteners, 141 nozzle, specific function of, 2
Swing-gate-type check valve, 326 nozzle shape, 89
Symbols for system parts, 28, 29 performance calculation, 84
System parts, symbols, 28, 29 performance parameters, £,3
Systems using chemical reaction products, 167-171 pressure, 144
recommended combustion characteristic length (L*)
for propellants, 87
selecting cooling method, 99
Tachometers, 411 shape, 87, 131
Taifun rocket, 140 specific impulse, 11, 83
Tank wall temperature, 152 thrust generated, 13, 14
Tanks, liquid hydrogen, 345-348 tubular wall, 107
Tanks, propellant. See Propellant tanks Thrust efficiency vs. bell nozzle length, 91
Tantulum, 62, 121 Thrust vector control, 274-275
Tapoff system, 136, 180 Timers, 407-408
Telemetry of engine, additional elements required, 48 Timosheuko, S., 209
Temperature Titanium alloys, 121, 163
effect on low-alloy steels, 60 Titanium-base alloys, 62, 12I, 162
chamber, calculation, 21 Topping flow cycle, 200
gages, 411 Topping systems, 181
sensors, 412 Transpiration cooling of thrust chamber, 117-118
steels, low-alloy, effect, 60 Triethyl aluminum and oxygen, 139
Thermal energy, conversion from chemical energy, 1 Triggering processes, prevention of, 148-149
Thermal energy, momentum caused by release of, 1 Triplet injector, 125, 127
Thermal environment, 434 Tubing, 358-360
Thermocouples, errors in measuring by 412, 413 Tungsten, 119, 121
Turbine gas flow rate, 186 Valv_ir four-way solenoid pilot valve, 3q4
Turbines, 178-181 Valves, 294-306
Turbopump systems, 175, 181-189, 191- 199, 200-261 Vapor condensation, 152
axial-flow, design, 225-238 Vasco_et, 246
balancing axial thrust, 223-225 VDt orifices, 307, 308
booster engine. 182. 183 Vehicle, definition, 21
calibration and off-design characteristics. 202, 203 Vehicle, design, cost, 50
cavitating inducers, 210-219 Vehicle diameter, 425
cavitation, 192 Vehicle performance, deviations, 35
centrifugal, 204-225 Vehicle range, effect, 35
cycle efficiency, 199, 200. 201 Vehicle reliability and flight safety, 44
description of developed, 182-186 Vehicle system, three-stage, weight increase calcula-
design layout, 261 tions, 41
design parameters, 187-202 Velocity, characteristic, 12
drive arrangements, 181- 182 Velocity-compounded impulse turbine, 180
elements, 176 Venting, 166
envelope, 175 Venturi-type valves, 299-300
gear design, 260 Vernier cutoff system, 34
heads and flow rates developed, 189 Von Braun, Wernher, 63. 415
inlet suction pressure head. 186 Von K_rm_n, 82
operating ranges, 202 V-2 rocket, 400
overall performance and operating efficiency, 196,
197
pump flow coefficient, 189
\%'asserfall antiaircraft rocket, 140
range of operation, 175
Weatherhead Co., 358
resistance and pump characteristic curves, 189
Weight of payload, effect, 35-37
specific speeds, 190
Wet gimbaled weight of engine, 37
trimming effects, 202
Wet weight of engine, 87
turbines, 238-257
variation of efficiency with speed, 195 Wiring, connective, 409
variation of engine weight and flow ratios with Working loads. 56-58
WorldWar II guided missiles, 140, 141
chamber pressure change, 175
Ullage gas compression, 152 Zucrow and Sellers equation, 115
NASA-l.,angley, 1968 _ $8

Liquid Rocket Engine Design

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Liquid Rocket Engine Design

Uploaded by

Copyright:

Available Formats

d _ '" _/ :_'_ _'

............. >....................... • .......... .. • .

Dieter K. Huzel and David H. Huang

Rocketdyne Division, North American Aviation, Inc.

Success in space demands perfection. Many of the brilliant achievements

Perfection begins in the design of space hardware. This book emphasizes

WERNHER VON BRAUN

Special recognition goes to R. F. Strauss of Astrosystems International, New Jersey, who

Chapter I. INTRODUCTION TO LIQUID PROPELLANT ROCKET ENGINES

Chapter II. ROCKET ENGINE DESIGN IMPLEMENTS

Chapter III.INTRODUCTION TO SAMPLE CALCULATIONS

Chapter IV. DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES

Chapter V. DESIGN OF PRESSURIZED-GAS PROPELLANT-FEED SYSTEMS

Chapter VII. DESIGN OF CONTROLS AND VALVES

Chapter VIII. DESIGN OF PROPELLANT TANKS

Chapter IX. DESIGN OF INTERCONNECTING COMPONENTS AND MOUNTS

Chapter X. ENGINE SYSTEMS DESIGN INTEGRATION

10.7 Engine Instrumentation ......................................................... 411

Chapter Xl. DESIGN OF LIQUID PROPELLANT SPACE ENGINES

INDEX .............................................................................. 451

Introduction To Liquid Propellant Rocket Engines

This expression, known as the momentum theo-

is a function of altitude. Equation (1-3b) states

We now assume that the rocket is operated at

Figure I-1 a finite ambient pressure Pa > 0 (lb/in 2) (low-

A Ptc dA- AePe --_-V e F =/2: Ve. Ae(Pe - Pc) (1-6)

Tile integral describes that force F (Ib) which

(1-4) movable cylinder (representing the thrust cham-

MOVEABLE CYLINCER (REPRESENTING

= 9040 + 760.8 × 10.7 × (32.2,/369.3)

1.2 THE GAS-FLOW PROCESSES IN THE

Since the analytical treatment of compressi-

INTRODUCTION TO LIQUID PROPELLANT ROCKET ENGINES

TABLE l-l.-Terms Used in the Calculation el The Principle of Conservation of Matter

08 Pt/(Pc)ns = [2/(y+ 1)] y/(F-0 (1-16)

, l l I L L 3 The static pressure Pt at a nozzle throat with

The prime function of a rocket nozzle is to mentioned earlier.

Pe ' Pa JET SEPaRaTION

iPe < F==

Theoretical gas weight flow rate:

NOZZLE (I_NT ROP!C) ONLY

W - A_(pc)ns _gT[2/(Y + 1)]Y:-I (i-19)

'' - -"--k/ / _/CO_ERY, o_ETo_eTSEOAR,T

the nozzle walls, where, due to friction, a bound- __L_

ary layer of slow-moving gases may exist. In [- 2 ]Y -1 (1-21)

Useful values of functions of the specific

Sample Calculation (1-2)

chamber of an ideal liquid propellant rocket

l_tc = 360.7 ib/sec; nozzle stagnation pressure,

_{ = °_2.67; gas specific heat ratio, y= 1.20; noz-

./ F zle expansion area ratio, _ = 12. The following

the nozzle inlet, Mi = 0.4. (In practice, thrust

MACH NUMBER _;t

TABLE 1-2.-Useful Values o[ Functions o[ the From equation (1-21):

4/( 2 _(y+l} (y-O

1.10 26.61 0.0909 0.5847 0.6590

Tinj = (Tc)inj = (Tc)ns = 6540 c R

Determine the following: a(_ flow static pres-

(pc)inj=(Pc)ns(1 + yMi2)/ t 1 +-_ Mi 2) }'-1

= 1000 x (1 + 0.16 × 1.20) _ 1.192 = 1082 psia 1

From equation (1-15) (c) Flow specific volumes:

Pi = Pinj/(1 + yMi 2) = 1+0.16×1.20-

From equation (1-9):

RTt 68 x 5945 (f) Flow areas: