Professional Documents
Culture Documents
,-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.: __-_
DESIGN OF
LIQUID
ROCKET PROPELLANT
ENGINES
_,,x :e_ ,
31
Scientific and Technical Information Division
_:_ '2'/
OFFICE OF TECHNOLOGY UTILIZATION 1967
NATIONAL AERONAUTICS AND SPACE ADMINISTRATION
Washington, D, C.
FORE WORD
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.
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.
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
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
Page
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 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
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.
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
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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)
= 9750 ft/sec
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
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
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
INTRODUCTION TO LIQUID PROPELLANT ROCKET ENGINES
Or
EXF&NS_ON
lit
(al
,>%
,
Ve = V _
297
R(Tc)ns I< 1-
Pe
(P"_-ns (1-
lS)
(P,)-4_
,
i TYPlCAL _ESSURE RECOVERY
Theoretical nozzle expansion ratio:
THROUGH SHOC_ WAVES ONLY NO
FURTHER SUBSON*C PRESSURE RE-
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:
2 (Pc)ns_ y
/ -
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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
MACH NUMBER Mx
0.1 0.2 0.5 I Z
IOO
5o
2o
rr
',' e
CONVERGING
--_)-- 1.40
, yy. 1.10 ,#1.
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.
INTRODUCTION TO LIQUID PROPELLANT ROCKET ENGINES
Solution
From equation (1-13):
(a) Flow static pressures:
From equation (1-14):
Y
Tt ( c)ns =6540x0-909=5945° R
y-I
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
ve 9620
RTi _68 x 6440=3.34 cu ft/lb
Vi - 144 Pi 144 × 909 Me = _-_e= 2--820= 3.43
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
2gy R(Tc)n _( Px _ or
vx: \(Pc)ns/J
Ae= 12xAt=782 in2
= ¥J6540x 64.4x 6 × 68 x [I- 0.593]
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.
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:
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 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
INTRODUCTION TO LIQUID PROPELLANT ROCKET ENGINES 15
The Effect of e
PO
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.
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
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
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
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):
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
INTRODUCTION TO LIQUID PROPELLANT ROCKET ENGINES 19
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
22 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
=_
INTRODUCTION TO LIQUID PROPELLANT ROCKET ENGINES 23
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
° _._. _= _;_ _; ._ _
24 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
-_=
_=
- _. _ _'_ _
:E
=¢
_r
E_
.-_ ¢5 _ _ ..4
;=
,-_ .-
r.
I
.4
E_
•_ _ ....
r_ _ _ c_ o c7
-_ _ .= o
.o
O.
=
INTRODUCTION TO LIQUID PROPELLANT ROCKET ENGINES 25
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.
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
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
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
26 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
Combinations (Continued)
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
95% Hydrogen peroxide Hydrazine ............. 201 1.41 1 26 4775 19.5 I1 601 285 359 ]CBM, IRBM.
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
Hydrazine ........... Pentaborane ............ 1.4 85 .796 4430 147 6402 1644 327 261 ICBMIRBM
A
INTRODUCTION TO LIQUID PROPELLANT ROCKET ENGINES 27
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
pellant Combinations
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
(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
(e) Isentropic expansion of ideal gas with shifting = Average molecular weight of combustion products
composition or shifting equilibrium in the nozzle at Tc
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.
INTRODUCTION TO LIQUID PROPELLANT ROCKET ENGINES
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
32 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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[
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
ROCKET ENGINE DESIGN IMPLEMENTS 35
(a)
(b) I (c) l
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.
ROCKET ENGINE DESIGN IMPLEMENTS 37
(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
ENGINE
ACCESS ................ iSPECI.......................
DRY 1400
I 2 3 4 5 6 7 8 9 I0 II 12 13 14 15 16
GO MONTH
AHEAD ,
ISSUE ENCLOSURE
DATE PAGE I OF r
, Z
( LOX Pump I
( • y
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.
(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
(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
4O DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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
--..__ _. T -- ¢
ROCKET ENGINE DESIGN IMPLEMENTS 43
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
44 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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.
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, -,
ROCKET ENGINE DESIGN IMPLEMENTS 47
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."
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
ROCKET ENGINE DESIGN IMPLEMENTS 49
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.
5O DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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 pro- gram 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
ROCKET ENGINE DESIGN IMPLEMENTS 51
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.
Project:
Subsystem: Supervisor:
Project:
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:
7. Reliability.
18. Funding.
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_
START CUTOFF
I: ....
_dECONDS SECONDS
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.
56 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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 analy- sis, the smaller the allowable margin between
sis enumerated above. the design limit loads and damaging loads.
ROCKET ENGINE DESIGN IMPLEMENTS 57
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
58 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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
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
(
ROCKET ENGINE DESIGN IMPLEMENTS 59
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
60 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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.
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
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
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
63
64 DESIGNOF LIQUID PROPELLANT ROCKET ENGINES
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
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-
= 1400 ° F
A-I
THRUST CHAMBER 750K First Stage Engine
Fuel pump
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
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
General Engine System Description
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,
INTRODUCTION TO SAMPLE CALCULATIONS 69
[Vacuum conditions]
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
Main valve pressure drop .......... psi ........... 20
Thrust .......................... Ib ........ 149 500
Calibration orifice pressure drop .... psi ........... 60
Specific impulse .................. sec .......... 440
Pump inlet pressure ............... psia .......... 25
Injector end pressure .............. psia ......... 875
Pump discharge pressure .......... psia ........ 1400
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
Main valve pressure drop .......... psi ........... 20
Calibration orifice pressure drop .... psi ........... 60 Tapoff gas from thrust chamber for turbine drive:
Pump inlet pressure ............... psia .......... 35
Pressure ........................ lb/in 2 ........ 750
Pump discharge pressure .......... psia ........ 1175
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
70 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
A,N
THRU.
CN. ER \ FUEL
TANKAIH
FOEVALVE
ANIFoLDTOR
EXHAUS. IHE ( .EL,U
TANK
Figure 3-3.-A-2 st3ge engine system schematic diagram.
INTRODUCTION TO SAMPLE CALCULATIONS 71
THRUST CHAMBER-
FUEL .u
I
II 7?..5"
D
I
L i
EXHAUST
GAS MANIFOLD
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
I
_ _
1_-__ THRUST BUILD UP
iii'
I
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
General Engine System Description
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
INTRODUCTION TO SAMPLE CALCULATIONS 73
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)
[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
Specific impulse .................. sec .......... 446
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
Main valve pressure drop .......... psi ........... 10
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
INTRODUCTION TO SAMPLE CALCULATIONS 75
(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
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
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 be- ing 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 com- ing 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.
INTRODUCTION TO SAMPLE CALCULATIONS 77
[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
Specific impulse .................. sec .......... 320
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.
CHECK
THRUST
CHAMBERS
EXCHANGERS
]MAIN OXE
VALVE
START
VALVE
BOTTLES
TANK FILL
DRAIN VALVES HELIUM RELIEF
VALVE
5'-- 9"
FUEL (N2H4) OXIDIZER (N204)
TANK TANK
3'--6" D
'- HIGH PRESSURE
/
PRESSURANT IHTERSTAGE _-THRUST CHAMBERS (2)
DIFFUSER CONNECTING 7'D
SKIRT THRUST VECTOR CONTROL
GIMBAL ACTUATORS (4]
START SHUTDOWN
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
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
82 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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
Ja.<j" ;C
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
84 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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
] I t s Solutions
25 35 45 _5 65 75 S5 95 _05 1_5
: 5810 ft/sec
o_ \ w// Fd _
"_ t25 8Q
Design c* = 5810 x 0.975 = 5660 ft/sec
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:
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:
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
/
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."
INJECTOR
? HRUSTC.___.HHAMBER
AXIS
[
CHAMBE:"
CYLINDRICAL
SECTION
_
[
[ OIA
AREA
Dt
A t
z_LENGTH Lc
CYLINDRICAL
COMBUSTION CHAMBER CONTRACTION Ac
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
EXIT
divergent cone half angle a varies from approxi- [ PLANE
/ / / / _-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.
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 91
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)
cal nozzle is used as a standard to specify bell (4) Initial wall angle of the parabola, On,
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.
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 •
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
HROAT INJECTOR
I_ COMBUSTION
INJECTOR / INJECTOR
: rl ..--"- COMBUSTION i;;rl_ _../-- cOM BUSTION
[_ JJ'- CHAMBER _l.J Z" CHAMBER
_ j/"-/pD'_ TH R OAT
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 %
{NJECTOR
Ii- I p---Dp"---'4 ;_
R-F Y H-F i1_
Dp/D T = ,5 Dp/D t = t0
,'
//SEA'LEVEL / (VACUUM)
nozzle.
IJ iI I ] K JlJli{
When a small amount of secondary flow is 50 100 200 300 500 1000 2000 4000
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
INJECTOR
FACE
THRUST
CHAMBER AXrS
L n = 102.4" -"-
=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"
INJECTOR
FACE
THRUST
CHAMBER AXIS
INdECTOR
FACE THROAT EXIT
I _1
V Et:46.1, `
THROAT
CHAMBER AXIS
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
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
DESIGN OF THRUST CHAMBERSAND OTHER COMBUSTIONDEVICES 99
(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.
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)
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 +
102 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
2400
!,! ,; = 6460 × (0.975) 2 : 6140 ° R
Dc=24.9 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.
From equation (4-15):
J
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 103
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
hg c - = 0.00067 Btu/in2-sec-deg F
1
--+ 1125
0.0027 From equation (4-13):
hg c- = 0.000276 Btu/in2-sec-deg F
1
-- e1645
0.000507 x\ 7480 x\5--._ ] j\-_--] a
104 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
(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:
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)
1
H- (4-22)
fig c :hg=0.00385 × 1 x 1.35 1 t 1
= 0.00520 Btu/in2-sec-deg F
where
At the exit nozzle point of e = 5:
q = Heat flux, Btu/in2-sec
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
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)
DESIGN OF THRUST CHAMBERSAND OTHER COMBUSTIONDEVICES 107
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
where
S_ =longitudinal thermal stress, lb/in 2 CHAMBER
4 ErEct (4-29)
)r Z_P= Pco1 . pco2 /
EQUIVALENT CHAMBER
where INTERNAL RADIUS
/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):
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
112 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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:
_ _KnoN_
.... ......
Figure 4-32.-Detail of injector manifolding and return manifold of typical regeneratively cooled tube
wall thrust chamber.
114 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
assembly time.
Dump Cooling
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
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
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 117
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.
118 DESIGNOF LIQUID PROPELLANT ROCKET ENGINES
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
120 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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
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
Sample Calculation (4-7)
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
L J "!iii',ll_{Ik={l!iI_ijll
[iiJlJ
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
124 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
!
=_
o_
=i
t.
|._.
'-o
_f!:i
251 ._
N
ii
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 125
I
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 127
_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
(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
--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_
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.
128 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
//// F THRUST
F VARIABLE AREA
OXIDIZER INJECTING
___ ANNULARSLOT FUEL ] CHAMBER
I VALVE
._ _- CHAMBER
THRUST
THROTTLE/
0 VALVE
OXIDIZER/_'-__'='7_)_ _
/_,,._-- INJECTOR
FACE
INERT GAS
CONTROL
VALVES
MOVABLE / [ /
PINTLE RING-" b"
L/ ANNULARSLOT
FUEL
/_ANIFOLD
Pi =Pm = 4 Pp (4-45)
Total fuel injector orifice area: At= 16.4 in 2 Total injector oxidizer orifice area: A o =6.43 in 2
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.
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,
132 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
The weight flow rate of a solid propellant gas rate per unit cross-sectional area of catalytic
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.
PORT
PORT
IGNITER (2)
TURBINE
COMB UFIt'OR
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).
136 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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.
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 137
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
Pyrotechnic Igniters
Hypergolic Igniters
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.)
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
GAS
GENERATOR
Catalysts VALVE
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
DESIGN OF THRUST CHAMBERSAND OTHER COMBUSTIONDEVICES 143
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
144 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
_d
i
DESIGN OF THRUST CHAMBERSAND OTHER COMBUSTIONDEVICES 145
Low-Frequency Instabilities
rl
!;I I:!'p t
....... i iI:,
i_ Tiiii!
tll I!
CMA_a£R _ESSU_E
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-
Design Approaches Toward Control of Combustion acoustic modes of instability can be most effec-
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
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 149
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
i.:-:.u-_--_- _--_--.....
l ::::
°°° --- .. • [ - _k__CONTOURE D
:1. _ _ NYDRAUL_ SYIT_Jt
WALL GAP
FILLER BLOCK
-- _.._..
150 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
B_JRST COMTROL
INJECTOR FACE
i, : suPPLY
CHARGE
I_l DIRECTED EXPLOSIVE (b} NON DIRECTED Ic) DIRECTED NOI¢ EXPLOSIVE
PULSE EXPLOSIVE PULSE C_5 FLOW
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
152 DESIGNOF LIQUID PROPELLANT ROCKET ENGINES
(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
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)
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.
154 DESIGNOF LIQUID PROPELLANT ROCKET ENGINES
(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)
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
Substitute into equation (5-4):
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
23 760 x 110.76
= 10.65 lb
4320 + 600
Tg- 10.65 x 1.25 + 660 = 1030 ° R
FF VALVE
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
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.
38O
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
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 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
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-
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
x 1.02 = 12.95 lb
Gross pressurant weight:
x 560x 12.95
x
Vu= 4500x 144 -4.3 ft 3 (_)x560 14.45
Yn = = 4.82 ft 3
4500 x 144
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
DESIGN OF PRESSURIZED-GASPROPELLANT-FEED SYSTEMS 161
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
23 760 × 119
Wg:/l_aa\ =21.2 Ib Tg=500+336=418 ° R
2
x 346
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:
-_)x560 × 28.2
Vu = 4500 x 144 - 9.4 fts including 2 percent reserve.
The required volume of the storage vessel:
Pressurant use factor:
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
Sample Calculation (5-3)
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
164 DESIGNOF LIQUID PROPELLANT ROCKET ENGINES
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
47 4xl.67
Pr= 9---_= 9 x 1.67_ 5:0.665 = 19800+ 2200 x (_)
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
downstream of the pump and vaporized in a heat _,*$EOU$ OXY6EN TO M,L{N _ I_ TURBINE E_AUST
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.
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:
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 propel- lants 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
168 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
INHIBITOR
PERFORATED
SUPPORT PLATE
DIAPHRAGM
PREssuRE-
RELIEF
TYPE
REGULATOR
SQUIB
MOLDED
CHOPPED
PHENOLIC
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.
• -_//
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
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.
172 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
Helium with thrust chamber Helium Helium at 995 ° R aver- 11 3 100 100
heating (fig.5-2) age
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
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
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
IO
8
L_
Z
*oo0
ZU
ItP-LUOMM. NIN4. SO/S_O
_ 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_
175
,ummm
176 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
_- 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
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 =
DIAPHRAGM
DRIVE
BEARINGS
'-- TURBINE EXHAUST DUCT
J
---TURBINE WHEELS
/ -- TURBINE HOUSING
SHAFT SEALS- /
DIRECTION
OF MOTION
BLEED (TAPOFF)
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
182 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
DESIGN OF TURBOPUMP PROPELLANT-FEED SYSTEMS 183
FUEL PUMP
OXIDIZER PUMP
.UTE PASSAGE (TO DISCHARGE)
PUMP
GEAR BOX
PELLER BALANCE RIBS
WEARING RING
IUCER
_MP INLET
SHAFT SEAL
BLADES OR
SECOND ROW NOZZLES
WHEELS
ACCESSORY
DRIVE PAD SHAFT AND
SHAFT NUT
f "FULL" INDICATOR
_ / INDICATOR
, ,>. ,.,,., ....,,...,
ADDITIVE CHAMBER
¼.!i_:}!;_ii:.':!:{H _----
i':?,i:!i'%i:._
ADDITIVE -_
.";._ PISTON
ADDITIVE :}ii!
l _ DRAIN iC{i
t '
OUTLET PORT -J' / " _ L_ INLET PORT
FUEL UNDER
[ i _._ ADDITIVE FILL _ / SPOOL-' L_ METERING TURBOPUMP
_,/ BLENDING CHAMBER -- ORIFICE DISCHARGE
PRESSURE
TABLE 6-1.-Operating Characteristics and Materials o[ Construction [or the Turbopump Shown in
Figure 8-14
Oxidizer Fuel
Pumps-
Turbine:
Inlet gas pressure (total) .............................. 597,6 psia
Exit gas pressure (static) ............................. 32.86 psia
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
Bearing material ..................................... 9310 alloy steel
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
186 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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
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
-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
---.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
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
pellant pump, the pump-inlet available net posi- H I:'ERCENT RATED DESIGN FLOW
C_ DES- , ,
propellant
pressed as:
vapor pressure. This can be ex-
°
Q.
\ { NPSH }C DESIGN
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
(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,
(58)o.;s
AIir_
_ 7000 x 111.7
TUPtOINE "
= 37 230
_ 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
(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
Sample Calculation (6-4)
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
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
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
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
LOX/RP-1 ......... 1100 0.635 1.097 43.3 0.308 IOO_ 1100 1200 1300 _4OO 1500 I_00 1700
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.
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
From equation (6-15):
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 -]
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
097
pump inlet pressure = 55 psia; and (3) fuel pump from sample calculation (4-1)(a__):
Solution 1544 y
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)
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
.._WMEEL
.! !- Fos.
PAOIIL VINES EN CLOSUI_E
Cu=
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
U2Cu2 s - U 1 Cul t
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
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
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
g =gravitational constant, 32.2 ft/sec 2
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
210 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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)
AXIAL FLOW
L_ _-VANES
AXIS
ROTATION OF INLET
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).
212 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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_
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 :
(6-61)
For inducers with tapered hub and tip contour:
do 2-d°t 2 +do h2
2 (6-55) (6-62)
214 DESIGNOF LIQUID PROPELLANT ROCKET ENGINES
: [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
13 040
= - 53.1 ft/sec Inducer design absolute outlet flow angle
77
× 7000 ^ ,,_
cml 53.1
u o :_ x 5. _o = 268 ft/sec
tan/31'-ul_cul,-229.3 -0.232; fi1'=13°3'
_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).
DESIGN OF TURBOPUMP PROPELLANT-FEED SYSTEMS 217
From equation (6-65a) the inducer inlet tip Our inducer exhibits characteristics similar
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
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)
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
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 .
From equation (6-32), the tangential compo-
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
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
Qimp
Design of Casings
bl - 3.12 _rdlCmle 1
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
22O DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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
Sample Calculation (6-8)
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)
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
,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 ,
!
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
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
_'.¢
DESIGN OF TURBOPUMP PROPELLANT-FEED SYSTEMS 225
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
=N = x 7000
× 4.8= 147 ft/sec
Us=_ds= 720
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
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.
o',/\/ \__
228 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
/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
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
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
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
eter, ft/sec
• -
Function of the Statot Cs=2R s sm _---_---) - si-_-_ac (6-100)
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
230 DESIGNOF LIQUID PROPELLANT ROCKET ENGINES
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
DESIGN OF TURBOPUMP PROPELLANT-FEED SYSTEMS 231
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
_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
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
DESIGN OF TURBOPUMP PROPELLANT-FEED SYSTEMS 233
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.
Sample Calculation (6-10)
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,
--Jr -- ---- dr df
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 ×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
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
a2 =a2'+ii=65 ° +5 ° =70 °
cm 230
a, +a2_44°26 ' +70 °
=57o13 ' tan /32 ' - um-cu2 -1 - 768-107.2 = 0.344; f12' = 19°
aic - 2 2
Cis 1.844
/32 =/32' +i=19 ° +4° :23 °
Ris=2 sin(-_) -2xsin(12°47')=4"16 in
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
4
238 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
240 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
-°'-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.
,TdmN
U- (6-129)
Design of Impulse Turbine Rotor Blades 720
calculated as
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
244 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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
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
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
246 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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)
_lI
occurs during gas flow through the moving ROTA
axial. U
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
ROTOR
NOZZLE
volume) in the flow passages must be taken into
account when calculating the gas densities at _ _' '= 2 SECOND STAGE
L-
DESIGN OF TURBOPUMP PROPELLANT-FEED SYSTEMS 249
a_ PO
heating in the nozzles, first rotor blades,
stator blades, and second rotor blades
0 '_k_- CONSTANT
To
A, /_PR_LFE UNES
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
tanfll=Cl C,cosal-U
sinai _ 3940×0.906-
3940x0.42390=0.622
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
DESIGN OF TURBOPUMPPROPELLANT-FEED SYSTEMS 251
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,
, 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
= 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
252 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
qbs 18.53 }r
T3 = T3'+-_--= 1427 + x-_--_.
= 1456° R
r =v
= 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
7.18
The relative flow velocity at the stator blade T4' --T3 -AH3__,/C p = 1456--- - 1445 Btu/lb
0.653 -
exit
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
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
DESIGN OF TURBOPUMP PROPELLANT-FEED SYSTEMS 253
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
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
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
Pbs - Blade solidity - 1.94 - 0.721 in Pbr2 = Blade solidity 1.67 0.888 in
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
From equation (6-149), blade radial height at From equation (6-149), the blade radial height
the inlet at the inlet is
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
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
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
Yl C1 sinai 2880×0.423 1784 fps From equation (6-147), the combined nozzle
sin 131 0.683 and blade efficiency
DESIGN OF TURBOPUMP PROPELLANT-FEED SYSTEMS 257
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
- lira _
258 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
_.: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----
Io ZO 3o 40 50 60 70 8090 _
o. 200 _0 400 =X>o
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.
-PUMP CASING
HOUSING
___ _SHAFT
-_-_- CC SHAFT
LABYRINTH SEALS
/-_ TURBOPUMP
_-TURBOPUMP CASING /CASING
FLOATING .\'..x>,_" P- SE;TL HO_SING
SEAL WASHER_\_\_/ST ICSE L ____GARTER SPR,NGS
_HER_/SEAL HOUSING
....._/,/_
_ ..... SEGMENTS
_ ___cu-_L_L - -- _ SHAFT
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
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.
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 _
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,
268 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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
(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
272 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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,
274 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
_ 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
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
V, voltage drop, volts ......... ' V = V A - VB AP, pressure drop, psi AP: PA - PB V _AP
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
CSIIE3 _CxA P CX
operation point
Q ---...,lllb
t
------D- Q
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 °
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-
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
276 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
Q
_C-_ /--VECHICLE THRUST MOUNT CENTERLINE
//,4i7/// /
_7 ._// / /-_..,_ cE.TE_ ,
-/- ,'.t-_.._q
EAANCDH
yIA_/ Pt:CAHE ) F_ __ _ GIMBAL "ARING [
\ \ \ 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
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
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.
Hydraulic pump
Reservoir (low pressure, or "sump")
Accumulator (high pressure)
Servovalve
Actuators
Feedback (electric or mechanical)
Lines, check valves, filters, connectors,
instrumentation
11
PIEr $$ TRA NSD_JCEJl
TEIIIIp TI l N _'_DU_EEII
PUMP SUCTION
RETRACT
17 SERVO-ACTUATOR
pump. Only electrical wires will then cross the 22. SERVO VALVE
Secondary Injection
(C) LIQUID INJECI'O'R
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
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
PrLOT _¢ALVE
and
vl A2
(7-5)
v2 A_
_ENTURI
where _ATE
/ /
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.
DESIGN OF CONTROLS AND VALVES 283
(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
/.--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
286 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
f(Le + b)pv 2
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
DESIGN OF CONTROLSAND VALVES 287
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 con- trol 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
DESIGN OF CONTROLS AND VALVES 289
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-
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
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 #:
29O DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
STATIC
SEALING\
_ ¢,-_ I'-- "1 D-GROOVE
\BREAKCORNERS, : _b -,J U:NGTH
?o APPROX.
o.oos' D l .-f
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
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
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).
VALVE STEM
-Acrvz. ,_ _%ulO
However, pressure levels and available space _.&LIN* 5t._f_3t5 LBO_H _LVE AN0 ,_v_t ) L_PPED N 8-10
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:
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
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
_ •. _ /- _.. _ _1_?,_,_
_ _,,_
HOT GAS
SPRING DISK | /
/ VALVE
VALV¢_ ...... _, _ sto,,
/ HOUSING
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
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"
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)
DESIGN OF CONTROLS AND VALVES 297
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,
Sample Calculation (7-3)
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
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--
=-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
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
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
seal. These are suitable only for medium- actuate other fluid-flow-control components, such
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-
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
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.
302 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
CONNECTION
- ELECR_CAL
5PRgNG
PffS_dRE
DIAPH Rt, GM -_.f
CLOSED _J AS_E_,_LY
• OffN ELECTRIC
POPPET _ ; VENT OR
OR
MAGNET
$OLENO_b
ENE_GIZEbOIt
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
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):
_rUtJ_ _,LrH _Ue_ _tES_UmE INlay p'Olr Substitute into equation (7-17), the valve
rO V_Nr eOl_
output
v[
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.
A'_
PILOT PASSAGE i TERMINAL BOX /-- COIL
, /\
SECTION B-B
INLET
Figure 7-48.-TypicaI design of a pilot-operated, lout-way solenoid pilot valve, developed by Valvair.
DESIGN OF CONTROLS AND VALVES 305
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
G
SOtE_¢*_ east, i_o_ ./_
x contingency factor
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.
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
, AIAIAIAIAIAIAIAIAI_Ap
(7-21)
_'= 0.0438 Cdo2v/_"L'_= 0.525 Cdo-_p
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.
3O8 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
"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 _
0.5 \
Solution
0.4 ,\
The average flow velocity in the fuel duct
892
0.3 , \
V--
- _r 7 2=66.3fps i i
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 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:
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):
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)
0.3
P{ 2'__ 350 _
514.7 0.682
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
310 DESIGNOF LIQUID PROPELLANT ROCKET ENGINES
(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.')
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).
"
312 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
CA4PcZ4
A 3 = ,,-rd3X = ,7× 0.0266 × 0.004756 = 0.000397 in s
#4- V_R-T
- -O'O062A4Pc
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
#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
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
=0.0001583 in 2
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.
/_____,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
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
Ap=[_-_+(12_+ 1"79×10-7''2]
(0._9-_ ]v ] ×(50.45) 2 PRESSURE P=----_"_VALVE POPPET
BALANCE
(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.
DESIGN OF CONTROLS AND VALVES 315
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
Most pressure regulators include two basic SENSING PRESSURE _'_ "_
ou . T LO O . .soo O
SERVO - VALVE ATTACHED TO THE REGULATOR CONTROLLER.
_,_11_.\\\\_ ---T----_OuTLET Pr
_'_----_OUTLET Pr _----_"__--OUTLET Pr
REGULATOR VALVE
stream systems) o 2
The basic test of a good regulator design is TIME, SEC TIME, SEC
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
REGULATOR CONTROLLER
SENSO_ ASSEMBLY
ORb=ICED _ARTITION
DYNAMIC SE_L
_ -..._,
_r
_ REGLJLATOR INLET,
Pi
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)
DESIGN OF CONTROLS AND VALVES 321
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.
? " !
_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
_ O- RING
(o)
__ LOCK NUT
SOFT o- ,NOo,
DISC
INLET
POST-TRAVEL C D
_ 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
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:
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
DESIGN OF CONTROLS AND VALVES 325
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
(2) Actuator:
_PET
where
Fp = pilot valve poppet seating
force, lb
Fa =pilot valve actuating force, lb
Fb = sensor bellows force, lb OPEN CLOSED
" ii.....
326 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
UPSTREAM
WING-GATE BURST FLUID P_RESSURE
HINGE
DOWNSTREAM _ UPSTREAM
V-GROOVE COINED IN
A NORSESHOE PATTERN
I _ _,L_E=,L
Burst Diaphragms
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
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
330 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
GRAIN
DIAPHRAGM
fUEL TANK
COOLING
THRUST CHAMBER
ASSEMBLLES
PRESSURE
RELIEF TYPE
REGULATOR
VENT
Figure 8-1.-Propellant tank design configuration of a typical prepackaged storable liquid propulsion
system.
l ,, ,,
[ :
, i!
D_;TwR
UC T URET } FUEL L__ ER __2111E22D I_IC_2_ON D UCT DUCT SKIRT MOUNTS ! M'PEFE
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
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.
332 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
DESIGN OF PROPELLANT TANKS 333
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
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
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
Fy (8-8)
Sw - 1.33
Fy _ 50 000_ 37 600 psi
Sw= 1.33 1.33
Hazard to personnel: or
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
DESIGN OF PROPELLANT TANKS 337
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:
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
338 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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)
(8-21) =
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
Sample Calculation (8-3)
(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
340 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
2 × 2 _ra2b
Vt° - 3 From equation (8-18), the required wall thick-
ness at the crown of the oxidizer 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:
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:
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
_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
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)
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)
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
DESIGN OF PROPELLANT TANKS 343
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
(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.
Tank Construction
/
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.
DESIGN OF PROPELLANT TANKS 345
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
(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,
346 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
pressure to temperature [or liquid oxygen. installed between an inner and outer facing
DESIGN OF PROPELLANT TANKS 347
v,
ON[_C_U_ CC_[ wao_ O_ Solution
@_S _ • Z 2 Le _FT 1 -- ATMOSPH£RE
LIQUIO HYDROGg N
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
Sample Calculation (8-6)
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
348 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
\\"1
--_ /' :z_ 4 .'
/ )"4'-,. ./
8.7 DESIGN OF PROPELLANT TANK ", ' , d_-< _ l],
PRESSURANT DIFFUSERS
_SSVRAmr GAS
INLEt
Positive Expulsion _O_C[_ R_¢ _RCU_t_ ¢VmvoLwr_
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.
DESIGN OF PROPELLANT TANKS 351
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
I, _ p_pe_LA_r
_r_eT
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
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
354 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
CON_IECTbNG TO
PUMP DISC_ARC-[--_r
E_O LOAO
COMPENSAT_ --
BEU.
, 1 J
O_JTCET FL._,_IG_E
CONN£C_r EO TQ
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
356 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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
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
DESIGN OF INTERCONNECTING COMPONENTS AND MOUNTS 359
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]
[ 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
]
Maximum working pressure, 750 psi
_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
SPECIAL
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.
36O DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
Figure 9-I O.-Typical flange joints of a propellant ferentially along the serrations, and will seal
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).
, ;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)
Sample Calculation (9-1)
From equation (9-5), the required average A = design factor, ranging from 4 to 6
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
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
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
From equation (9-6):
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
From equation (9-12):
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
(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:
From equation (9-16):
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
900 ° F.
Elastomer and Metal O-Rings for Flange Joints BRIAR COR_ERS 005"
_--i
D3 --
D2
_ G_OOVE SIDE FINISH f2
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
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 ]
aSee fig. 9-13 for seal design. See fig. 9-12 for flange design. Duct dlameter D, may be equal to D2.
368 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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-
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)
seal
An ingenious
deflection
approach
capabilities
toward
to accommodate
attaining better
de-
_
,,
K : 7 ' -oo3
,)i oo._ooo.
'' -:oz
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.
TABLE 9-8.-TypicaI Design Data o[ NaHex Pressure-Actuated Seals and Flange Joinzs
[All dimensions in inches]
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.
370 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
582
Sc = 1 x 0.019 - 30600 psi
G AREA
767
Sc = 1 × 0.01---_
= 40 400 psi
_ The allowable
flange sealing
depth of imperfections
surfaces should be calculated
on the
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.
__-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.
372 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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.
DESIGN OF INTERCONNECTING COMPONENTS AND MOUNTS 373
,XIAL
"qF--- OEFLECTION
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
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)
376 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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:
TABLE 9-9.-Yield Strength, Limiting Bulging Stresses, and Allowable Motion Stresses of Frequently
Used Bellows Materials
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
DESIGN OF INTERCONNECTING COMPONENTSAND MOUNTS 377
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.
Design a bellows, as shown in figure 9-4, for din= _/(di2+ 2 d°2) - _ (82+ _ '822)
-- - 8.42 in
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
.'
380 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
THRUST CHAMBER
383
384 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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
Wo
Ro =weight fraction of oxidizer stored in
Ro = (w ° + wt) (10-10) the combustion chamber
At =throat area, in 2
g =gravitational constant, 22.2 ft/sec 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
a =pump pressure rise design factor, lb- fer and combustion characteristics
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
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
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-
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
ENGINE SYSTEMS DESIGN INTEGRATION 389
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
_____._.__._>
ENGINE SYSTEMS DESIGN INTEGRATION 391
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
Figure 10-5.-Propellant flow design characteristics of a typical pressure feed engine system
(oxidizer or fuel).
---.... ,,,
,---FUEL PUMP PRESSUI_:
CALIBRATION
"_, _ VS FLOW CURVE AT NZ
ORIFICE DESIGN
t
Pfd
° °J
° " OX,Q, ZER
PUMP
PRESSURE
i i
FUEL FLOW LB/SEC
OXIDIZER FLOW LB/SEC
Wfd wfo Wfb %a Wod
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
..... _ ----_:::.:=?:
392 DESIGNOF LIQUID PROPELLANT ROCKET ENGINES
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 ::---
394 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
or
N- Ns H°'Ts _ 1980 × (2990) °'Ts = 7190 rpm
QOS (12 420) °-s 4730 × 50.45
= 1666 psi
144
Fuel pump volumetric flow
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.
ENGINE SYSTEMSDESIGN INTEGRATION 395
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)
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
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:
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:
396 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
Thrust chamber flow rate This value closely confirms the assumptions
for the last trial. Thus:
-,qmNlm
ENGINE SYSTEMS DESIGN INTEGRATION 397
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,
Sample Calculation (10-4)
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
[Value of C* correction to be obtained from the C* correction versus mixture ratio curve shown in fig. 10-8]
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
Sample Calculation (10-5)
=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
_ -0.4
u (Percent change in F) = 4.531 + (-0.02) x 1.103
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
ENGINE SYSTEMS DESIGN INTEGRATION 401
IGNITION PHASE
HELIUM SOLENOID CONT.
IN LET VALVE
VENT
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
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
4O4 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
Electrical Components
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
•5, I <D
. I s c
I' _
r,_;'(_ ........
, _ e÷ CLOS*J_I[ • _ _Z{)$-II,,.,
3 r •
I_NITION
L. ................. J
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
408 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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
410 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
EI4BI8(RD} •
EI5BIB(RD) J
;_____ j/IEIyBIB_RD)
PHYSICA-ROUTING DIAGRAM
-_ RECEPTACLE ._PI6
::::::((:;:'-_--J RECEPTACLE
[
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]
[] [] ]
MEASUREMENTS
!li
c_ < c_ < (SENSORs)
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
ENGINE SYSTEMS DESIGN INTEGRATION 413
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
F 3 I I1 !l
L/.---DIAPHRAGM
BOSS_
ADAPTER -_/
_.--ENGINE
COMPONENT
_--- DIAPHRAGM
ENGINE COMPONENT WAll
WALL
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
(A) {B)
"
416 DESIGNOF LIQUID PROPELLANT ROCKET ENGINES
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
ENGINE SYSTEMSDESIGN INTEGRATION 417
F-
08
YAW -- --0_
ACTUAT
I
a. b.
E I i
4 fi 6
NUMBER OF ENGINES
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-
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.
TEST STAND
,ul
STRUCTURE TA,
ENGINE CONNECT
Lax IEL_O_--
PANEL .o= pu_ -- - c, _, _,.o_
Lox ,_c,c_,-
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.
ENGINE SYSTEMS DESIGN INTEGRATION 421
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)
80
80
4C _ _- SKIN BELOW THRUST MOUNT /--SKIN BELOW THRUST 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
i(
-200 _ I_THRUST MOUNT TEMPERATURE _NITROGEN AMBENT INTERNAL TO MOUNT
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
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.
ENGINE SYSTEMS DESIGN INTEGRATION 425
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.
I 1
Upper stage ullage rockets ignited
I I
Engine start sequence initiated
I
Separation timer started
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)
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.
ENGINE SYSTEMS DESIGN INTEGRATION 427
too
\
\
\\ \
\ \
\ \
7s \
\ \
\ \
ILl \
I&.
so
o [ t I
0.S 0.6 0.7
SIGNAL
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
428 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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--
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 $ :
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.
432 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
are frequently used in space engines. r WEEK _ kq:_TM $ blONT_tS :AR I0 YEA_
craft should use the same propellant combina- COAST I_UI_TIO_, _*OUl_
TABLE ll-1.-Comparison of Various Liquid Propellant Combinations [or Space Engine Applications
.,a======
DESIGN OF LIQUID PROPELLANT SPACE ENGINES 433
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)
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. )
434 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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).
DESIGN OF LIQUID PROPELLANT SPACE ENGINES 435
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
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,
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
DESIGN OF LIQUID PROPELLANT SPACE ENGINES 439
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
ox,o,z
ACTUATING
o.-O FV W
FLOW
VENT
f SERVO
MOTOR TORQUE-- 7 _
. A
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.
442 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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.
(1) Rotation of the vehicle over a given angle, thrust delivered on command must be very pre-
(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
DESIGN OF LIQUID PROPELLANT SPACE ENGINES 443
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
DESIGN OF LIQUID PROPELLANT SPACE ENGINES 445
_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__
._ FUEL TANK
PACKAGE _ _ __._
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
@ SERVICE CONNECTION
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.
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.
448 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
,.:
",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)
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.
INJECTOR
WALL THICKNESS
FUEL INLET
COMB. CHAMBER
DIA
THROAT DIAMETER
OXIDIZER'-_-
INLET
-- FITTING
MOUNTI NG
STRUCTURE
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
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,
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
451
452 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
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
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
NASA-l.,angley, 1968 _ $8