Mechanical Engineering Design by J. E.

Shigley
Machine Design by E. M. Badawy
Machine Design by V. L. Maleev & J. B.
Hartman
Design of Machine Elements by V. M. Faires
Design of Machine Elements by M. F. Spotts
Machine Design Theory & Practice by
A. D. Deutshmen and others
Machine Design by R. S. Khurmi
 - Design considerations.
 - Factor of safety.
 - Power screws.
 - Design of detachable joints ( threaded joints, keys
and splines).
 - Design of permanent joints (welding, riveting,
adhesion, interference fitting).
 - Theories of failure.
 - Column design.
 - Fatigue.
 - Mechanical springs.
 - Pressure vessels.
 - Rotating disks.
Prerequisite: ME 211
1 Functionality 14 Noise
2 Strength/stress 15 Styling
3 Deflection/stiffness 16 Shape
4 Wear 17 Size
5 Corrosion 18 Control
6 Safety 19 Thermal properties
7 Reliability 20 Surface
8 Manufacturability 21 Lubrication
9 Utility 22 Marketability
10 Cost 23 Maintenance
11 Friction 24 Volume
12 Weight 25 Liability
13 Life recovery 26 Remanufacturing
• Aluminium Association (AA)
• American Gear Manufacturers Association
(AGMA)
• American Institute of Steel Construction (AISC)
• American Iron and Steel Institute (AISI)
• American National Standards Institute (ANSI)
• ASM International
• American Society of Mechanical Engineers
(ASME)
• American Society of Testing and Materials
(ASTM)
• American Welding Society (AWS)
• American Bearing Manufacturers Association
(ABMA)
• British Standards Institution (BSI)
• Industrial Fasteners Institute (IFI)
• Institution of Mechanical Engineers (I. Mech. E.)
• International Bureau of Weights and Measures
(BIPM)
• International Standards Organization (ISO)
• National Institute for Standards and Technology
(NIST)
• Society of Automotive Engineers (SAE)
Stress and Strength
Stress σ or τ : Calculated from load and dimensions
Strength Su or Ss : Material property
Strength Su
• Design Factor nd  
Stress 
• Dimensions = Function of(Load & Stress)
= Fun. of(Load, design factor & Strength)
Actual Stress = function of( Load & Actual Area)
Strength Su
Factor of Safety n 
Actual _ Stress  a
EXAMPLE
A rod with a cross-sectional area of A, loaded in tension
with an axial force of P 8.9 kN undergoes a stress of σ =
P/A. Using a material strength of 165 MPa and a design factor
of 3.0, determine the minimum diameter of a solid circular rod.
Using Standard sizes select a preferred standard diameter
and determine the rod’s factor of safety.
Standard rod diameter in millimeters:
0.05, 0.06, 0.08, 0.10, 0.12, 0.16, 0.20, 0.25, 0.30, 0.40, 0.50, 0.60, 0.70, 0.80, 0.90,
1.0, 1.1, 1.2, 1.4, 1.5, 1.6, 1.8, 2.0, 2.2, 2.5, 2.8, 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, 6.0, 6.5,
7.0, 8.0, 9.0, 10, 11, 12, 14, 16, 18, 20, 22, 25, 28, 30, 32, 35, 40, 45, 50, 60, 80,
100, 120, 140, 160, 180, 200, 250, 300
Solution

factor of safety =3.728

• In the symbolic units equation for Newton’s
second law, F= ma or F = M L T-2
• F stands for force, M for mass, L for length, and T
for time.
• F =ML/T2 =(kilogram)(meter)/(second)2
• F= kg · m/s2 = N
• W =m g (W is the weight)
• W = (1 kg)(9.81 m/s2) = 9.81N
• Stress = F/Area = N/m2=Passcal
• giga G 1 000 000 000 =109
• mega M 1 000 000 =106
• kilo k 1 000 =103
• hecto h 100 =102
• deka da 10 =101
• deci d 0.1 =10−1
• Centi c 0.01 =10−2
• milli m 0.001 =10−3
• micro μ 0.000 001 =10−6
name symbol 10n Decimal English word Since
yotta Y 1024 1000000000000000000000000 septillion 1991
zetta Z 1021 1000000000000000000000 sextillion 1991
exa E 1018 1000000000000000000 quintillion 1975
peta P 1015 1000000000000000 quadrillion 1975
tera T 1012 1000000000000 trillion 1960
giga G 109 1000000000 billion 1960
mega M 106 1000000 million 1960
kilo k 103 1000 thousand 1795
hecto h 102 100 hundred 1795
deca da 101 10 ten 1795
1 1 one –
deci d 10−1 0.1 tenth 1795
centi c 10−2 0.01 hundredth 1795
milli m 10−3 0.001 thousandth 1795
micro μ 10−6 0.000001 millionth 1960
nano n 10−9 0.000000001 billionth 1960
pico p 10−12 0.000000000001 trillionth 1960
femto f 10−15 0.000000000000001 quadrillionth 1964
atto a 10−18 0.000000000000000001 quintillionth 1964
zepto z 10−21 0.000000000000000000001 sextillionth 1991
yocto y 10−24 0.000000000000000000000001 septillionth 1991
 Calculations and Significant
Figures
The accuracy of a real number depends on the number of
significant figures describing the number. Usually, but not
always, three or four significant figures are necessary for
engineering accuracy. Unless otherwise stated, no less than
three significant figures should be used in your calculations. The
number of significant figures is usually inferred by the
number of figures given (except for leading zeros). For
example, 706, 3.14, and 0.002 19 are assumed to be numbers
with three significant figures.
Lecture 2
Thread Standards and Definitions
 Pitch is the distance between adjacent thread forms measured parallel to the
thread axis
 The major diameter d is the largest diameter of a screw thread.
 The minor (or root) diameter dr is the smallest diameter of a screw
thread.
 The pitch diameter dp is a theoretical diameter between the major and
minor diameters.
 The lead l, not shown, is the distance the nut moves parallel to the screw
axis when the nut is given one turn. For a single thread, as in Fig., the lead is
the same as the pitch.
 A multiple-threaded product is one having two or more threads cut beside
each other (imagine two or more strings wound side by side around a pencil).
Standardized products such as screws, bolts, and nuts all have single threads; a
double-threaded screw has a lead equal to twice the pitch, a triple-threaded screw
has a lead equal to 3 times the pitch, and so on.
 All threads are made according to the right-hand rule unless otherwise noted.
 The American National (Unified) thread standard has thread angle 60◦ and the
crests of the thread may be either flat or rounded.
The equations and data used to develop this table
have been obtained from ANSI B1.1-1974 and
B18.3.1-1978. The minor diameter was found
from the equation dr=d-1.226p, and the pitch
diameter from dp=d-0.649 p. The mean of the pitch
diameter and the minor diameter was used to
compute the tensile-stress area.
For Square threads
a -The system is in equilibrium under the
action of the forces F, fN, N and PR hence,
for raising the load, we have

b -In a similar manner, for lowering the

load, we have
From drawing : tan λ = l/πdm
Where :
l= lead of thread
f = the coefficient of friction
dm = mean diameter
where TR is the torque required for overcome thread friction and
to raise the load. TL is the torque required to overcome a part
of the friction in lowering the load. It may turn out if the
lead is large or the friction is low, that the load will lower
itself by causing the screw to spin without any external
effort. In such cases, the torque TL will be negative or zero.
When a positive torque is obtained from this equation, the
screw is said to be self-locking. Thus the condition for self-
locking is π f dm > l or f > tan λ
This relation states that self-locking is obtained whenever
the coefficient of thread friction is equal to or greater than
the tangent of the thread lead angle.
Screw efficiency
If we let f = 0 we obtain T0 = Fl/2π The efficiency is therefore
e =T0 / TR =Fl / 2πTR
Acme threads
The normal thread load is inclined to the
axis because of the thread angle 2α and
the lead angle λ. Since lead angles are
small, this inclination can be neglected
and only the effect of the thread angle
considered. The effect of the angle α is to
increase the frictional force by the
wedging action of the threads. Therefore
the frictional terms must be divided by
cos α.
 Collar friction
. If fc is the coefficient of collar friction
and dc is the mean collar diameter the
torque required to turn the collar is
Tc = F fc dc / 2
Stresses in power screws
Screw Body:
τ is the shear stress in torsion
σ is the axial stress in the body
 
2

 max     2
2
Threads:
σB is the bearing stress where nt is
the number of engaged threads
σb is the bending stress at the root
of the thread
τ is the transverse shear stress τ
Differential Screws
If P1 =P2 no motion between A and B
Equivalent Pitch = P1 –P2
Torque Diagram
Lecture 3
Screw Fasteners
Joints—Member Stiffness
There may be more than two members included in the
grip of the fastener. All together these act like
compressive springs in series, and hence the total
spring rate of the members is
(spring rate , k = force /deflection)
If one of the members is a soft gasket, its stiffness relative
to the other members is usually so small that for all
practical purposes the others can be neglected and
only the gasket stiffness used.
The Figure show the pressure distribution at the member
interface. The results show that the pressure stays high
out to about 1.5 bolt radii. Thus the use of the
pressure-cone method for stiffness calculations with
fixed cone angle α .
With α = 30◦, this becomes:
If the members of the joint have the same Young’s
modulus E with symmetrical shape back to back, then
they act as two identical springs in series. Using the grip
as l = 2t and D as the diameter of the washer face =1.5d ,
we find the spring rate of the members
km’ = 1/(1/k1 + 1/k3 )
kb =F/ δb =Ab Eb /l (if threaded part is small)
kg=Ag Eg /tg .. Ag=π (Dg2-d2)/4 .. Dg=1.5d+l tan(α)
Bolt load
Consider:-
Fa = applied external tensile load per bolt
Fi = preload (clamping force applied by tightening the nut before Fa is applied)
Pb = portion of Fa taken by bolt
Pm = portion of Fa taken by members
Ft = Pb + Fi = total resultant bolt load & Fm = Pm − Fi = resultant load on members
δa = Pb /kb =Pm /km & Pm = Pb . km/kb ........ Since Fa = Pb + Pm , we have
Pb =Fa . kb /(kb + km) = C Fa and Pm = Fa − Pb = (1 − C)Fa
where C =kb/(kb + km) ,is called the stiffness constant of the joint.
The resultant bolt load is
Ft = Pb + Fi = C Fa + Fi and Fm = Pm − Fi = (1 − C)Fa − Fi Fm < 0
 Bolt Torque

dc = (d + 1.5d)/2 = 1.25d

T = K Fi d

Bolt Preload

where Fp is the proof load, obtained from the equation

Fp = Ap Sp
Here Sp is the proof strength obtained from Tables, or from an approximate value
Sp =0.85Sy . Be very careful not to use a soft material in a threaded fastener.
 Strength requirements
Stress in bolt = σb =Ft /Ar = (C Fa + Fi ) / Ar
For factor of safety f.s.
Proof strength Sp = f.s. (C Fa + Fi ) / Ar
f.s.= Sp Ar / (C Fa + Fi )

Gasketed Joints
The gasket must prevent liquid or gas at pressure P from leakage. If a full gasket
is present in the joint, the gasket pressure Pg is found by dividing the force in the
member by the gasket area per bolt. Thus, for N bolts,
Pg = −Fm /(Afg / N) = - Fm N / Afg
From previous relations
Pg = [Fi − Fa (1 − C)] N / Afg > P (= 2P to 4P)
In full-gasketed joints uniformity of pressure on the gasket is important. To
maintain uniformity of pressure, adjacent bolts should not be placed more than
six nominal diameters apart on the bolt circle. To maintain wrench clearance,
bolts should be placed at least three diameters apart. A rough rule for bolt
spacing around a bolt circle is
3.d ≤ πDb/N ≤ 6.d
where Db is the diameter of the bolt circle and N is the number of bolts.
Bolted Joints with Eccentric Loading
 From geometry
1  2  3 Pi .lb
    i 
h1 h2 h3 Ab .E
For the same bolt length and diameter
P1 P2 P3
   M o  0
h1 h2 h3
F . L  P1 .h1 .n1  P2 .h2 .n2  P3 .h3 .n3
h22 h23
F . L  P1 .h1 .n1  P1 . .n2  P1 . .n3
h1 h1
F . L.h1 hi
P1   Pi  F . L.
n1 . h12  n2 . h22  n3 . h23 k
1
 j j
h 2 .n
j 1
 Shear Joints with Eccentric Loading
The rotational pivot point lies at the centroid
of the cross-sectional area pattern of the pins,
rivets, or bolts. Using statics, we learn that the
centroid G is located by the coordinates x and y,
where xi and yi are the distances to the ith area
center:
 Primary shear
Fi′ = F.Ai / Σ Aj ...... Fi′ = F.di 2/ Σ dj 2
If di is the same
F′ = F/n
where n refers to the number of bolts
in the group and the force F′ is
called the direct load, or primary shear

Secondary shear
For the same bolt area
Fi′′α Δi and Δi = θi.hi
θ1 = θ 2= θ 3=…..=constant
Fi′′=S.hi
ΣMc =0
F.L= F1′′.h1+ F2′′.h2+........
F.L=S.h12 +S.h22 + S.h32 …
F.L=S. Σ hi2
S=F.L/ Σ hi2 "

h i 
'

F . Ai
F i
F .L.
n F i n
Fi′′=F.L . hi / Σ hi2
h  Ai
2
i
1 1
Where the diameters of the bolts are assumed equal. If not, then one replaces
F′′ with the shear stresses τ ′′ = 4F′′/πd2 for each bolt.
τ i′′α Δi
Fi′′α Δi di2 ….. Fi′′=S . hi . di2
F.L=S. Σ (hi .di)2 ….. S=F.L/ Σ (hi .di)2

Fi′′=F.L .hi . di2 / Σ (hi . di ) 2

Fb  F '2  F "2 2.F '. F ". cos( )

 F
h

t
From Fig. (a) and (b) if
b = Σ Li F
1-The stress in plates du to bending :
σ = M.y/I = F.h.t/(2.b.t3 /12)
σ = 6.F.h/(b.t2 )
2 – From Fig. (c) Shear stress in rivet
τ = F/A = F/ (Ab .n)
3 - From Fig. (d) : Tension in plates c
σ = F/A = F/(t.Σ Li )= F/b.t
4 - From Fig. (e) : Bearing stress
σb = F/(t.d.n)
5 - From Fig. (f ) :
τ = F/(2.t.c)
Lap joint:
Single, double or triple riveted joints
d = 1.2 t , p > 3 d , pr = 0.6 p
t = plate thickness , d = rivet diameter
Butt joint
t1 = 0.7 t , p > 3 d , pr = 0.6 p , d = 1.2 t
Keys and pins are used on shafts to secure rotating elements, such as gears, pulleys,
or other wheels. Keys are used to enable the transmission of torque from the shaft to
the shaft-supported element. Pins are used for axial positioning and for the transfer
of small torque or thrust or both
(a) Gib-head key;

(b) Woodruff key.

 Key design

F F .d / 2 2.T
 all   
b.L b.L.d / 2 b.L.d
F 4.T
 br  
L.h / 2 h.L.d
b  d / 4  L  1.5.d  h  d / 6
 T   br
D  d  .L. N . D  d 
2 4

T= Torque capacity L = Length of hub N = Number of splines

Butt or groove welds:
(a) Square butt-welded on both sides
(b) Single V with 60° bevel and root
opening of 2 mm
(c) Double V with 60° bevel
(d) Single 45° bevel
Special groove welds:
(a) T joint for thick plates
(b) U and J welds for thick plates
(c) Corner weld (may also have inside weld for greater strength but should
be used for light loads)
(d) Edge weld for sheet metal and light loads.
Butt joint. F

a- Tensile force: h.l
F

b- Shear force: h.l
My M .6
 
c- Bending moment: I h 2 .l
Fillet welds:
a- Load perpendicular to the welding line
(Symmetrical case).
No bending moment
Throat t = 0.707 h

b- Load parallel to the welding line

(Symmetrical case).
Primary shear τ’ :
A = throat area of all the weld.
F = shear force.

Secondary shear τ” :
M = the applied torque
r = the distance from the centroid of
the weld group to the point in the
weld of interest (farthest distance)
J = the second polar moment of area
of the weld group about the centroid
of the group.
J = 0.707 h Ju
Ju = unit second polar moment of area
φ = angle between τ’ and τ”

   '2  "2 2. '. ". cos( )

 Primary shear τ’ :
A = throat area of all the weld.
F = shear force.

Secondary shear τ” :
M = the applied moment.=FL
c = the distance from the neutral axes of the weld group to the point in the
weld of interest (farthest distance).=d/2
I = the second moment of area = 0.707 h Iu
Iu = the unit second moment of area (from table)=d3 /6

Mc Md / 2 1.414 M * 3
 
"

3

I 0.707hd / 6 hd 2
   '2  "2
Ductile materials (yield criteria)
• Maximum shear stress (MSS)
• Distortion energy (DE)
• Ductile Coulomb-Mohr (DCM)

Brittle materials (fracture criteria)

• Maximum normal stress (MNS)
• Brittle Coulomb-Mohr (BCM)
• Modified Mohr (MM)
Maximum-Shear-Stress Theory for Ductile Material
The maximum-shear-stress theory predicts that yielding begins whenever the
maximum shear stress in any element equals or exceeds the maximum shear
stress in a tension test specimen of the same material when that specimen
begins to yield.
For simple tensile stress, σ = P/A, and the maximum shear stress occurs on a
surface 45° from the tensile surface with a magnitude of τmax = σ/2. So the
maximum shear stress at yield is τmax = Sy/2. For a general state of stress, three
principal stresses can be determined and ordered such that σ1 ≥ σ2 ≥ σ3. The
maximum shear stress is then τmax = (σ1 − σ3)/2. Thus, for a general state of
stress, the maximum-shear-stress theory predicts yielding when:
τmax = (σ1 − σ3 )/2 ≥ Sy / 2 or σ1 − σ3 ≥ Sy
Note that this implies that the yield strength in shear is given by Ssy = 0.5Sy
For factor of safety, n.
τmax = Sy / (2 n) or σ1 − σ3 = Sy / n
Plane stress problems are very common where one of the principal stresses is
zero, and the other two, σA and σB, are determined. Assuming that σA ≥ σB, there
are three cases to consider in plane stress:
Case 1: σA ≥ σB ≥ 0 . Then , σ1 = σA and σ3 = 0. ,
then begins to yield if σA ≥ Sy and the factor of safety n = Sy / σA
Case 2: σA ≥ 0 ≥ σB . Then , σ1 = σA and σ3 = σB ,
then begins to yield if σA − σB ≥ Sy and the factor of safety n = Sy / (σA - σB )

Case 3: 0 ≥ σA ≥ σB .( │σB │ ≥ │σA │ ≥ 0) , then σ1 = 0 and σ3 = σB ,

then begins to yield if σB ≤ −Sy and the factor of safety n = -Sy / σB
Case of Pure Shear σA = - σB = τmax = Sy/2.
Ssy = 0.5 Sy

Case of Hydrostatic test

σ1 = σ2 = σ3 = -P .... τmax = 0 …. Py → 
No failure n= 
The distortion-energy theory predicts that yielding occurs when the distortion strain
energy per unit volume reaches or exceeds the distortion strain energy per unit volume
for yield in simple tension or compression of the same material.
The total strain energy per unit volume U=½ 1 1 + ½ 2 2 + ½ 3 3
and 1 = 1 / E – μ (2 + 3 )/ E then
U= (12 + 22 + 32 -2μ(12 + 23 + 31))/(2E) ……(1)
This energy can be divided to two type
a) Ud Distorting energy which give zero volume change and only change the angles
b) Uv Non distorting energy from hydrostatic stress (for volume change only)
For type (a) ∆V= 0 …… ℓ3 - ℓ3 (1+ 1d) (1+ 2d) (1+ 3d) = 0 (ℓ= element length)
Neglect 1d 2d and similar terms we get :
1d + 2d + 3d = 0 …….. 1d + 2d + 3d = 0
For type (b) if the hydrostatic stress = P
1 = P + 1d …… 2 = P + 2d …… 3 = P + 3d then
P = (1 + 2 + 3)/3
The energy stored by the stress P, from equation (1) :
Uv=3P2 (1-2μ)/2E = (1-2μ) (1 + 2 + 3 )2/6E
Ud = Utotal - Uv
From equation (1)
Ud = (12 + 22 + 32 )/2E - μ(12 + 23 + 31)/E - (1-2μ) (1 + 2 + 3 )2/6E
Ud = (1+ μ ) ( (1 - 2 )2 + (2 - 3 )2 + (3 - 1 )2 )/6E
in simple tension test: 1 = Sy …. 2 = 0 …… 3 = 0
Udy = (1+ μ ) 2 Sy2 /6E
So for the general state of stress yield is predicted if Ud equals or exceeds
Udy . This gives the limiting safe stress ’ < Sy

For plane stress 3 = 0

‘ = [12 - 12 +22 ] ½
This equation is a rotated ellipse in the σ1, σ2 plane
Using xyz components of three-dimensional stress, the von Mises stress can be
written as:

Factor of safety n = Sy / σ′
The shear yield strength:
Consider a two dimension case of pure shear stress τxy , σx =0 , σy =0
by analysis σ1 = τxy and σ2 = -τxy then σ′= (3 τxy2 )½
τxy = 0.577 σ′ or Ssy = 0.577 Sy
Not all materials have compressive strengths equal to their corresponding
tensile values ( Sc > St )
The Coulomb-Mohr theory or the internal-friction theory, assumes that
the boundary B1B2B3 in Fig. is straight. With this assumption only the
tensile and compressive strengths are necessary, consider the conventional
ordering of the principal stresses such that σ1 ≥ σ2 ≥ σ3, the largest circle
connects σ1 and σ3. The centre of this circle is C2 , C1 and C3 are the pure
tension and compression circles. Triangles OBiCi are similar, therefore
Case 1: σA ≥ σB ≥ 0. For this case, σ1 = σA and σ3 = 0.
The failure condition is σA ≥ St and the factor of safety n = St / σA
Case 2: σA ≥ 0 ≥ σB . Here, σ1 = σA and σ3 = σB .
The failure condition becomes or

Case 3: 0 ≥ σA ≥ σB . For this case, σ1 = 0 and σ3 = σB .

The failure condition gives σB ≤ −Sc and the factor of safety n = -Sc / σB
Case 4: For pure shear τ, σ1 = − σ3 = τ .
The torsional yield strength occurs when
τmax = Ssy . Substituting σ1 = − σ3 = Ssy and
simplifying gives
The maximum-normal-stress (MNS) theory
states that failure occurs whenever one of the
three principal stresses equals or exceeds the
strength. The principal stresses for a general
stress state are σ1 ≥ σ2 ≥ σ3. This theory then
predicts that failure occurs whenever σ1 ≥ Sut
or σ3 ≤ −Suc where Sut and Suc are the
ultimate tensile and compressive strengths,
respectively, given as positive quantities.
Case 1: σA ≥ σB ≥ 0 then n = Sut /σA

Case 2: σA ≥ 0 ≥ σB &
then n = Sut /σA
Case 3: σA ≥ 0 ≥ σB &
then n = -Suc /σB

Case 4: 0 ≥ σA ≥ σB then n = -Suc /σB

For a brittle material the strength parameter is the ultimate strength
instead of the yield strength:
Brittle-Coulomb-Mohr
Case 1: σA ≥ σB ≥ 0. For this case, σ1 = σA and σ3 = 0.
The failure condition is σA ≥ Sut and the factor of safety n = Sut / σA

Case 2: σA ≥ 0 ≥ σB . Here, σ1 = σA and σ3 = σB .

The failure condition becomes or

Case 3: 0 ≥ σA ≥ σB . For this case, σ1 = 0 and σ3 = σB .

The failure condition gives σB ≤ −Suc and the factor of safety n = -Suc / σB
Modified Mohr
Case 1: σA ≥ σB ≥ 0 then n = Sut /σA

Case 2: σA ≥ 0 ≥ σB &
then n = Sut /σA

Case 3: σA ≥ 0 ≥ σB &

then

Case 4: 0 ≥ σA ≥ σB then n = -Suc /σB

At x
Mx = - P.y
d2y M  Py
 
2 EI EI
dx
P
if  
2
EI
d2y
2y  0
General solution dx 2
y = A sin (α x) + B cos (α x)
The boundary conditions :
at x = 0 , y = 0 then B = 0
at x = l , y = 0 then
A sin (α l) = 0
Either A = 0 or α l = 0 , π , 2 π , …., n π
α = 0 , π/l , 2 π/l , …., n π/l
P = EI α2 = 0 , π2EI/l2 , 4 π2EI/l2, …., n2 π2EI/l2
the critical load is the smallest one
Buckling mode
y = A sin (π x/l)
Pcr = π2EI/l2
which is called the Euler column formula which can be extended to apply
to other end-conditions by writing
l
Pcr = C π2EI/l2 leq 
C
Using the relation I = Ak2, where A is the area and k the radius of gyration,
enables us to rearrange the last equation into the more convenient form
Pcr /A = C π2E/(l/k)2 = π2E/(leq/k)2
where l/k is called the slenderness ratio. I
k
The quantity Pcr /A is the critical unit load. A
The factor C is called the end-condition constant, and it may have any
one of the theoretical values 1/4 , 1, 2, and 4
The parabolic or J. B. Johnson formula
The general form of the parabolic formula is

where a and b are constants that are evaluated

by fitting a parabola tangent to the Euler curve.
If l/k = 0 then a = Sy
At point T the two curves are tangent and
have the same x & y , then point T is
x1 = (2 π2 CE / Sy ) 1/2 and y1 = Sy /2

The parabolic equation, for l/k < x1

Working formula for structural steel columns
The common structural steel St37 has Su = 370 MPa & Sy = 240 MPa &
E = 210 Gpa … if the end-condition constant = C … then
x1 = (l/k)1 = (2 π2 CE / Sy)1/2 =131 C1/2
if l’/k > 131 then use Euler equation and l
if l’/k < 131 then use Johnson curve l ' 
C
Ex. If the factor of safety is 1.5, Find the working stress for the given conditions
1- Both ends are pivoted
2- One end is pivoted and the other is fixed
3- Both ends are pivoted with laterally support in x-x plan as shown
1- l = 150 cm ….. C = 1 …. A =12 cm2 Ixx = 3 . 43 / 12 = 16 cm4
Iyy = 4 . 33 /12 = 9 cm4
kx = (Ixx/A)1/2 = 1.15 cm
ky = (Iyy/A)1/2 = 0.866 cm
From the curve, the smallest load has the largest (l/k) which obtained from ky
l/k =150/0.866=174 > 131
Then use Euler equation
cr = π2 CE/(l/k)2 = π2 . 1. 210. 103 /(174)2 = 69 MPa
w = cr /1.5 = 46.5 MPa

2 – C=2 …… l/(k C1/2)= 150/(0.866 . 1.414) = 123 < 131

Then use Johnson

cr = 240 106 – 6.8 103 1232 =137MPa

w = 137/1.5 = 91 MPa

3- C=1 in x-x plan and y-y plan

lyy = 150/2 =75cm …… lyy/ky = 75/0.866 = 86.8
lxx = 150 cm …………. lxx /kx = 150 / 1.15 = 130 use the largest of x-x
(you can use any method ) for Euler cr = Sy/2 = 120 MPa …. w = 80 MPa
 Fatigue Failure
Stress-time relations:

Completely reversed sinusoidal stress Repeated stress

σmin = minimum stress

σmax = maximum stress
σm = midrange component
σr = range of stress
σa = amplitude component
R= the stress ratio
A = the amplitude ratio

Sinusoidal fluctuating stress

 Fatigue-Life
The fatigue life is the number of stress cycles to failure
An S-N diagram is the fatigue strength, life diagram plotted from the results of completely
reversed axial fatigue tests.

Material: UNS G41300 steel, normalized; Sut = 125 kpsi

The Endurance Limit
In the case of the steels, a knee occurs in the graph, and beyond this
knee failure will not occur. The strength corresponding to the knee is
called the endurance limit S’e , or the fatigue limit. The nonferrous metals
and alloys, do not have an endurance limit. Figure shows scatter bands
indicating the S-N curves for most common aluminium alloys. Since
aluminium does not have an endurance limit, normally the fatigue
strength Sf is reported at a specific number of cycles, normally N = 5(108)
cycles of reversed stress.
S-N bands for representative
aluminium alloys
 The graph appears to suggest that the endurance limit ranges from about 40 to 60
percent of the tensile strength for steels up to about 1450 Mpa (210 kpsi). Beginning at
about Sut = 1450 MPa, the scatter appears to increase, but the trend seems to level off,
as suggested by the dashed horizontal line at S’e = 700 MPa (105 kpsi).

Ratios of S’e /Sut of 0.60, 0.50, and 0.40 are shown by the solid and dashed lines

S′e = 0.5Sut if
Sut ≤ 1400MPa
(200 kpsi)
S′e = 700 MPa
(100 kpsi) if
Sut > 1400 Mpa
(200 kpsi)
Fatigue fracture surface of a 200-mm
diameter piston rod of an alloy steel.
This is an example of a fatigue fracture
caused by pure tension, a crack may
initiate at a forging flake slightly below
centre, grew outward symmetrically and
ultimately produced a brittle fracture
without warning
Fatigue fracture surface of an AISI
8640 pin. Sharp corners of the
mismatched grease holes provided
stress concentrations that initiated
two fatigue cracks indicated by the
arrows.
Endurance Limit Modifying Factors
Se = ka kb kc kd ke kf S′e
ka = surface condition modification factor
kb = size modification factor
kc = load modification factor
kd = temperature modification factor
ke = reliability factor
kf = miscellaneous-effects modification factor
S′e = rotary-beam test specimen endurance limit
Se = endurance limit at the critical location of a machine part in the geometry and
condition of use
When endurance tests of parts are not available estimations are made by
applying the earlier relations to the endurance limit (S′e = 0.5Sut ....) .

Surface Factor ka
ka = aSbut
Size Factor kb
For bending and torsion loading (d in millimetre), rotating bar
kb = 1.24 d−0.107 for 2.79 ≤ d ≤ 51 mm
kb = 1.51 d−0.157 for 51 < d ≤ 254 mm
For axial loading there is no size effect kb = 1
When a round bar in bending is not rotating, or when a noncircular cross
section is used employs an effective dimension de obtained by equating the
volume of material stressed at and above 95 percent of the maximum stress
to the same volume in the rotating-beam specimen
de = 0.37 d
A rectangular section of dimensions h × b has
de = 0.808 (hb)1/2

Loading Factor kc
kc = 1 for bending load
kc = 0.85 for axial load
kc = 0.59 for torsion load
Temperature Factor kd
kd = ST / SRT

Reliability Factor ke

ST = tensile strength at operating temperature

SRT = tensile strength at room temperature

Miscellaneous-Effects Factor kf
The factor kf is proposed to account for the reduction in endurance limit due to
all other effects as Residual stresses, Corrosion, stress-concentration, .....
Stress Concentration and Notch Sensitivity
The factor Kf is commonly called a fatigue stress-concentration factor

Kf = 1 + q(Kt − 1) or Kf s = 1 + qshear(Kts − 1)

q = Notch sensitivity
Kt (or Kts ) = Stress concentration factor
Fatigue Failure Criteria for Fluctuating Stress
The modified Goodman diagram
 The criterion equation for the Soderberg line is
Sa/Se +Sm/Sy = 1
Similarly, we find the modified Goodman relation to be
Sa/Se +Sm/Sut = 1
The Gerber failure criterion is written as

The ASME-elliptic is written as

where n is the design

factor or factor of safety
Lecture 8
 Mechanical Springs
Types of springs
Helical springs
Tension or compression

Torsion springs

Laminated or leaf springs Disc springs ( Belleville springs)

 Mechanical Springs
Stresses in Helical Springs
As shown in the figure, from
equilibrium the cut portion would
contain a direct shear force F and a
torsion T = FD/2.
τmax = T.r / J + F/A
Substitution of τmax = τ , T = FD/2,
r = d/2, J =πd4/32, and A = πd2/4 gives

Now we define the spring index C =D/d

(the preferred range of spring index is 4 ≤ C ≤ 12)

shear-stress correction factor

With this relations the sheer stress equation can be rearranged to give
 The Curvature Effect
Suppose Ks is replaced by another K factor, which corrects for both curvature and direct
shear. Then this factor is given by either of Wahl factor KW or Bergsträsser factor KB ,
the results of these two equations differ by less than 1 percent

Deflection of Helical Springs

Using Castigliano’s theorem, the total strain energy for a helical spring is
composed of a torsional component and a shear component: if the wire
Modulus of rigidity G and length l = πDN where N = Na = number of active
coils then the strain energy is

Then using Castigliano’s theorem, to find total deflection y gives

The spring diameter D after deflection to solid length will increase by

0.05(p2-d2)/D where p is the pitch
The spring rate
Also called the scale of the spring, is k = F / y, and so
Formulas for the Dimensional Characteristics of Compression-Springs
Types of ends for compression springs: (a) both ends plain; (b) both ends squared; (c)
both ends squared and ground; (d) both ends plain and ground.

Type of Spring Ends

Plain and Squared or Squared
Term Plain Ground Closed and Ground
Total coils, N t Na Na +1 Na + 2 Na + 2
Free length, Lo pNa + d p(Na + 1) pNa + 3d pNa + 2d
Solid length, Ls d(Na + 1) d(Na + 1) d(Na + 3) d(Na + 2)
Pitch, p (Lo - d)/Na Lo/(Na +1) (Lo-3d)/Na (Lo-2d)/Na
 Stability
Compression coil springs may buckle when the deflection becomes too large in
relatively long springs. The critical deflection is given by the equation

The quantity λeff is the effective slenderness ratio and is given

by the equation
Where α is the end-condition constant.
C′1 and C′2 are elastic constants defined by the equations

End Condition Constant α

- Spring supported between flat parallel surfaces (fixed ends) ......... 0.5
- One end supported by flat surface perpendicular to spring
axis (fixed); other end pivoted (hinged) ............................................ 0.707
- Both ends pivoted (hinged) ................................................................. 1
- One end clamped; other end free ...................................................... 2
Absolute stability occurs when the term C′2 /λ2eff is greater than unity. This means
that the condition for absolute stability is that

For steels, this turns out to be Lo < 2.63 D / α

If L>4D spring must be guided from inside or outside diameter
Spring Materials
A great variety of spring materials are available to the designer, including plain
carbon steels, alloy steels, and corrosion-resisting steels, as well as nonferrous
materials such as phosphor bronze, spring brass, beryllium copper, and various
nickel alloys. Prehardened wire should not be used if D/d < 4 or if d > 12mm.
Most commonly used steels are:
a- Music wire,0.80–0.95C
b- Oil-tempered wire, 0.60–0.70C
c- Hard-drawn wire, 0.60–0.70C
d- Chrome-vanadium
e- Chrome-silicon
The tensile strength versus wire diameter, for some materials, can be given by the
equation Sut =A/dm
where A and m are constants given in the following table
Then the distortion-energy theory can be employed to obtain the torsional yield
strength (Sys = 0.577Sy)
In general:
Sy=Kt.Sut and
Sys=Ks.Sut
 Critical Frequency of Helical Springs
The harmonic, natural, frequencies for a spring placed between two flat and
parallel plates, in radians per second, are

where k = spring rate

g = acceleration due to gravity
W = weight of spring
γ = specific weight (N/m3)

For m=1

d G.g
 (cycles / sec)  .
 .D2 .N 8.
 Fatigue Loading of Helical Compression Springs
Zimmerli discovered that size, material, and tensile strength have no effect on the
endurance limits (infinite life only) of spring steels in sizes under 10 mm.
Unpeened and peended Surface treatment springs were tested, the corresponding
endurance strength components for infinite life were found to be :
For d<10mm
Sse=310 MPa for unpeened
Sse=465 MPa for peened
For d>10mm
Sse=200 - 300 MPa for heat formed
Ssu = 0.67Sut

If the spring loaded with variable load has a maximum

force Fmax and minimum force Fmin then to calculate the
fatigue safety factor using Goodman line method:

Where KB is the Bergsträsser factor, Wahl factor KW can

be used instead, if desired.
Combination of springs
Series
Keq=P/ δ
P=P1=P2=P3
δ= δ1+ δ2+ δ3 =P1/K1+P2/K2+P3/K3
δ=P(1/K1+1/K2+1/K3)
1/Keq=1/K1+1/K2+1/K3

Parallel
Keq=P/ δ
δ= δ1= δ2= δ3
P=P1+P2+P3
Keq=P1/δ+P2/δ+P3/δ
Keq=K1+K2+K3
 2 2  P1  22   12
U  
1
p.d  P
 1 1 1 
.d 
1
.
2
1to2
P1  P2  22   12 P1  P2
U   1   2 . 2  2 . 2   1 
1to2
if P1  0  1  0
P .  . .d 3
U 2 2 2  we have  P 
2 8. K s . D
8. P . D 3 . N  . . N . D 2
 4

G .d K s .G .d
2    2
U . .d 2 . . D. N   .V
4. K s .G  4
2  4. K s .G
2

where V  volume of spring

In designing a spring with a hook end, bending and torsion in the hook must be
included in the analysis. The maximum tensile stress at A, due to bending and
axial loading, is given by

where (K)A is a bending stress correction

factor for curvature, given by
The maximum torsional stress at point B is given by

where the stress correction factor for

curvature, (K)B, is

When extension springs are made with coils in contact with one another, they are
said to be close-wound. Spring manufacturers prefer some initial tension in close-wound
springs in order to hold the free length more accurately. The corresponding load
deflection curve is shown in Figure, where y is the extension beyond the free length
and Fi is the initial tension in the spring that must
be exceeded before the spring deflects. The load-
deflection relation is then
F = Fi + k.y
where k is the spring rate.
The initial tension in an extension
spring is created in the winding
process by twisting the wire as it is
wound onto the mandrel. When the
spring is completed and removed from
the mandrel, the initial tension is
locked in because the spring cannot
get any shorter. The amount of initial
tension that a spring maker can
routinely incorporate is as shown in
the Figure. Where the uncorrected
torsional stress is given by
The ends of the spring ultimately
connect a force at a distance from the
coil axis to apply a torque. Helical coil
torsion springs are usually used with
a rod or arbor for reactive support to
maintain alignment, and to provide
buckling resistance if necessary.
The wire in a torsion spring is in
bending. The springs are designed to
wind tighter in service. As the applied
torque increases, the inside diameter
of the coil decreases. Care must be
taken so that the coils do not interfere
with the pin, rod, or arbor. The
bending mode in the coil might seem
to invite square- or rectangular-
crosssection .
The number of body turns Nb is the number of
turns in the free spring body by count. The
body-turn count is related to the initial
position angle β by
Nb = integer +β/360◦ = integer + Np
where Np is the number of partial turns
Bending Stress
The bending stress can be obtained from curved-beam theory expressed in the
form
σ = K . Mc / I
where K is a stress-correction factor. The value of K depends on the shape of the wire
cross section and whether the stress sought is at the inner or outer fibre. Wahl
analytically determined the values of K to be, for round wire,

where C is the spring index and the subscripts i and o refer to the inner and outer
fibres, respectively. In view of the fact that Ko is always less than unity, we shall
use Ki to estimate the stresses. When the bending moment is M = Fl and the section
modulus I/c = d3/32, we express the bending equation as
σ = Ki .32.F.l /πd3
which gives the bending stress for a round-wire torsion spring.
 Deflection and Spring Rate
For torsion springs, angular deflection can be expressed in radians, or in
revolutions (turns). The spring rate k is expressed in units of torque/radians
(Nmm/rad) and moment is proportional to angle θ expressed in radians. The
spring rate, if linear, can be expressed as
k = M1 / θ1 = M2 / θ2 = (M2 − M1) / (θ2 − θ1)
where the moment M can be expressed as Fl.
The strain energy in bending is

Applying Castigliano’s theorem gives the deflection (l.θ) by

Substituting I = πd4/64 for round wire and solving for θ gives

The total angular deflection in radians is obtained by adding the effect of the end
deflection of a cantilever for each end of lengths l1, l2:
The spring rate
The spring rate k in torque per radian is

Change of diameter
Torsion springs are frequently used over a round bar or pin. When the load is
applied to a torsion spring, the spring winds up, causing a decrease in the
inside diameter of the coil body. It is necessary to ensure that the inside
diameter of the coil never becomes equal to or less than the diameter of the pin,
in which case loss of spring function would result. The helix diameter of the coil
DT becomes

where T is the angular deflection of the body of the coil in number of turns, given by
T = θ / (2 π)
Leaf springs (laminated springs)

a – Uniform width

Cantilever
6. P . L

b .t 2
4. P . L3

E .b.t 3
Beam
3. P . L

2.b.t 2
P . L3

4. E .b.t 3
b – Uniform strength
Cantilever
x L
  constant
bx b
6. P . x 6. P . L
x  
b x .t 2 b .t 2
6. P . L3

E .b.t 3
Beam

3. P . L
x 
2.b.t 2
3. P . L3

8. E .b.t 3
Lecture 9
Stresses in thick walled cylinders:
Cylindrical pressure vessels, hydraulic cylinders, gun barrels, and pipes
carrying fluids at high pressures develop both radial and tangential
stresses with values that depend upon the radius of the element under
consideration. In determining the radial stress σr and the tangential
stress σt .
Consider the element shown in Figure in a pressure cylinder loaded
with internal and external pressure Pi &Po
 The element force balance will be
a) Fr = 0 in the radial direction for unit length

d d r
 r .r .d  2. t .dr .  ( r  .dr )( r  dr )d  0
2 dr
d r
 t   r  r.  0  (1)
dr

b) Strain

du
r   ( 2)
dr
 t 2. .( r  u)  2. .r u
t     ( 3)
Lt 2. .r r
E
r  ( r   . t )
1  2
from 2, 3
E du u
r  (   . )  (4)
1  2 dr r
d r E d 2u u  du
 (   .  . )  (4)
dr 1   dr
2 2
r 2 r dr
E
t  (   . r )
2 t
1 
from 2,3
E u du
t  (   . )  (5)
1  2 r dr
from 4,5 in 1
d 2u 1 du u
 .   0  (6)
2 r dr r 2
dr
Equation “6” is a second degree differential equation has a general
solution:
 C2
u  C1 .r 
r
du C
 C1  2  (7 )
dr r2
from 4,5
E C2 C2 
r  C 
2 1
  ( C1 
2 
)
1   r 2
r 
E  (1   )  B
r  2 1
C (1   )  C 2 2 
 A   (8)
1   r  r2
E C2 C2 
t  C 
2 1
  ( C1 
2 
)
1   r 2
r 
E (1   )  B
t  C (1   )  C 2
2 1 2 
 A  (9)
1   r  r2
 A & B from boundary conditions:
At r = a (inner radius) …… r = -Pi & at r = b (outer radius) …… r = -Po
- Pi = A – B/a2 ……… - Po = A – B/b2
Solving for A & B:

Pi .a 2  Po .b 2 2 ( Pi  Po )
2
A  B  b .a .
2 2 2 2
b a b a
b 2 .a 2
Pi .a 2  Po .b 2  2
.( Pi  Po )
t  r
b2  a 2
2 2 b 2 .a 2
Pi .a  Po .b  2
.( Pi  Po )
r  r
b2  a 2

b 2 .a 2
Pi .a 2  Po .b 2  2
.( Pi  Po )
or  t or r  r
b2  a 2
Case of internal pressure only:
Po = 0
Pi .a 2 b2
r  2 2
.(1 
2
)
(b  a ) r
Pi .a 2 b2
t  2 2
.(1 
2
)
(b  a ) r
Pi .(b 2  a 2 )
 t max  2 2
 at r  a
(b  a )
( t   r ) b2
 max   Pi .(
2 2
)  at r  a
2 b a

In thin wall cylinder

b – a = t (thickness)
b2 – a2 = (a + b).(a – b)  2 a . t & b2 + a2  2 a2
t = Pi . a/t
Case of external pressure only:
Pi = 0

Po .b 2 a2
r   2 2
.(1 
2
)
b a r
Po .b 2 a2
t   2 2
.(1 
2
)
b a r
Po .2.b 2
 t max   2 2
 at r  a
b a
Case of Solid shaft:
a=0
r = t = - Po
 Deformation
If r and t are known:

r t
r   .
E E
t r
t   .
E E
u   r  increase in radius
2. .( r  u)  2. .r u
t  
2. .r r
r
u   t .r  .( t   . r )
E
Stresses produced by Shrink-Fits
Using Superposition method for both cylinders
t = t (created from shrink fit) + t (created from internal pressure)

  U i (cyl 1 b  c )  U o (cyl 2  a  b)
 b   b 
   .( t  1 . r )  .( t   2 . r )
 E1  b  c  E2 ab
 b  c 2  b2  b  b2  a 2 
  Psh . . 2 2  1   Psh . .   2 
E1  c  b  E 2  b 2  a 2 
b2  a 2
  2
2 2
b a
b  c 2  b2  b  b2  a 2 
  Psh . . 2 2  1   Psh . .
2 2
 2 
E1  c  b  E 2  b  a 
if E &  are the same for cyl . 1 & 2
E (b 2  a 2 ).( c 2  b 2 )
Psh  .
b 2.b 2 .(c 2  a 2 )
for solid shaft a  0
E . (c 2  b 2 )
Psh  .
b 2.c 2
Maximum torque that can be transmitted by shrink-fit connection

T = f . Psh .π . d2 . L /2
f = coefficient of friction
L = length of contact area
d = shaft diameter
Force required to assemble the two members
F = f . Psh .π . d . L
Design considerations:
If shaft radius has maximum tolerance + s1 and minimum + s2 , and the
maximum disc tolerance +d1 and minimum +d2
Where s1 > s2 > d1 > d2
maximum interference δmax = maximum shaft radius – minimum disc radius
δmax = (b + s1) – (b + d2) = s1 – d2
where b = nominal radius of the assembly
Minimum interference δmin = (b + s2) – (b + d1) = s2 – d1
The joint must transmit torque in case of minimum interference, and to
have safe stress in case of maximum interference.
E . min (c 2  b 2 )
Pshmin  .
b 2.c 2
T = f . Psh min .π . 2 . b2 . L
E . max (c 2  b 2 )
Pshmax  .
b 2.c 2
c2
 max  Psh max .( 2 2
)
c b
E . max
 max 
2.b
Lecture 10
 γ = weight / unit volume
∆m = ∆V γ/g ……. For unit thickness
∆m = r dφ . dr . γ/g
The element force balance will be
Fr = 0 in the radial direction for unit length
d d r 
 r .r .d  2. t .dr .  ( r  .dr )( r  dr )d  .r .d .dr . 2 .r  0
2 dr g
d r  2 2
 t   r  r.  . .r  0  (1' )
dr g
 We can have the same relations 2 , 3 , 4 , 5 and substitute in 1’ we get:

d 2u 1 du u  . 2
 .   (1   2 ). .r 2  0  ( 2' )
dr 2 r dr r 2 g.E
Equation (2' ) is a second degree differenti al equation has a general solution :

1   2  . 2 2 C2
u  ( ). .r  C1 .r   ( 3' )
8 g.E r
substitut eq. 3' in eq. 4 & 5
B
r  A  k .r 2  (4' )
r2
B 1  3. 
t  A  .k .r 2  (5' )
r2 3 
3    . 2
where k  ( ).
8 g
 Case of Disc with a hole

A & B from boundary conditions:

At r = a (inner radius) …… r = 0 & at r = b (outer radius) …… r = 0
0 = A – B/a2 – k.a2 ……… 0= A – B/b2 -k.b2
Solving for A & B:
A = k ( a2 + b2 ) ……… & …… B = k ( a2 . B2 )

a 2 .b 2
 r  k .(a 2  b 2   r2)
r2
 2 2 2
2 a .b  1  3.   2 
 t  k .a  b  2    . r 
 r  3   
 r max  k .(b  a )2  at r  a .b
 1  2
 t max  2.k .b 2  .a   at r  a
 3  
 Case of Solid Disc without hole

A & B from boundary conditions:

At r = 0 from symmetry …… r = t
A – B/r2 – k.r2 = A + B/r2 – k.r2 .(1+3.μ)/(3+ μ )
2.B = k . r4 . [(1+3.μ)/(3+ μ ) -1 ] = 0 …… B = 0
at r = b (outer radius) …… r = 0……… 0= A – B/b2 -k.b2
A = k .b2

 r  k .(b 2  r 2 )
 2  1  3.   2 
 t  k .b    . r 
  3   
 t max   r max  k .b 2  at r  0
1  2
 t min  2.k . .b  at r  b
3 
tH =tangential stress in disc with small hole (a  0)
tS = tangential stress in solid disc without hole
Shaft is assembled by Shrink Fit with a rotating disc
If the speed is relatively high the two parts may separates, in this case the radial
stress at the separated surfaces will equal to zero, the disc radial deformation at
inner radius Ud and shaft radius enlargement Us are given by:
 td a .2.k 2 1   2
U d  a . td  a .  .(b  .a )
E E 3 
 ts a .2.k 1   2
U s  a . ts  a .  .( .a )
E E 3 
Consider that the interference at zero speed is δo and will reduce to δω at ω
δω = δo – (Ud – Us) = δo – 2.k.a.b2/E
Where k is function of ω. The shrink-fit pressure can be calculated in this
case to transmit the torque at speed ω
E .  (b 2  a 2 ) E 2.k ( b 2
 a 2
)
P sh  .  .( o  .a .b 2 ).
a 2 a E 2
2.b 2.b
the maximum shear stress at   0 .... k  0

b2 E
 max  P0 sh .  . o
2
b  a2 2.a
2 2
2.k ( b  a )
Torque  2. . f . L.a 2 . Psh   . f . L.a . E .( o  .a .b 2 ).
E b2