EMG 2403: Solids and Structural Mechanics III

Lecture Notes Compiled By:

Jared Amisi

© May 2023
Students are reminded not to treat these lecture notes as a comprehensive and solely sufficient material
for their studies since the purpose of the notes is not meant to be a substitute for regularly attending
classes, reading relevant textbooks and recommended books. The notes are aimed at providing a quick
reference and a brief guidance for the students.
Course Outline

Energy methods: Determinate and indeterminate structures; virtual displacement and virtual
forces. Strain energy in tension, torsion, bending and shear, impact loading, work under several
loads and Castigiliano’s theorem. Unit load method for calculating displacements.
Thick and compound cylinders: Lame’s equations, thick cylinders with internal and external
pressure, effect of end constraints, compound cylinders, stresses produced by shrink-fit.
Rotating discs and cylinders: Stresses and strains, rotation of shrink fit assemblies, discs with
varying values of thickness, thermal effects. Rotation of cylinders/shafts.
Deformation beyond the elastic limit: Bending of beams beyond the elastic limit, torsion of shafts
beyond the elastic limit, plastic deformation of thick cylinders under internal pressure, residual
stresses.
Unsymmetrical bending: Revision of simple bending of straight beams - concepts of plane of
loading, plane of moments - resolution of moments. General flexure formula, applications, concept
of stress variation with distance from the neutral axis.
Bending of curved beams with plane loading: Winkler’s analysis

Prerequisites
ˆ Engineering Mechanics.

ˆ Integral and differential calculus.

ˆ Solids and Structural mechanics I & II.

Reference Textbooks
1. Stephens, R.C., Strength of Materials: Theory and examples, Edward Arnold Publishers,
1970

2. Gere, J.M., and Timoshenko, S.P, Mechanics of Materials, Chapman and Hall Publishers,
4th Ed., 1991

3. Benham, P.P. and Crawford, R.J., Mechanics of Engineering Materials, Longman Publishers,
4th Ed., 1990.

4. Case, J., and Chilver L. and Carl, T. F., Strength of Materials and Structures,
Topics Coverage
1. Thick and compound cylinders

2. Deformation beyond the elastic limit

3. Unsymmetrical bending

4. Energy methods

5. Rotating discs and cylinders

6. Bending of curved beams

Contents

Course Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i

Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i

Reference Textbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i

Topics Coverage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i

1 Thick walled pressure cylinders 1

1.1 Lame’s equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1.1 Longitudinal and Maximum Shear Stress . . . . . . . . . . . . . . . . . . . 4

1.2 Methods of reinforcing thick walled cylinders to accommodate higher pressures . . 7

1.2.1 Compound Cylinder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2 Deformation Beyond Elastic Limit 16

2.1 Bending of Beams Beyond the Elastic Limit (Plastic Bending) . . . . . . . . . . . 16

2.1.1 Plastic bending of of beams with one axis of symmetry . . . . . . . . . . . 21

2.2 Torsion of Shafts Beyond the Elastic Limit (Plastic Torsion) . . . . . . . . . . . . 26

2.3 Tutorial 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

3 Unsymmetrical Bending of beams 32

3.1 Properties of area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

3.1.1 Second Moment of area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

3.1.2 Parallel axes theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

3.1.3 Perpendicular axes theorem . . . . . . . . . . . . . . . . . . . . . . . . . . 34

3.1.4 Centroid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
 3.1.5 Product of inertia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

3.2 Revision of simple bending theory . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

3.2.1 Stress due to simple bending . . . . . . . . . . . . . . . . . . . . . . . . . . 46

3.2.2 Radius of curvature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

3.3 Unsymmetrical bending . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

3.3.1 Stresses due to unsymmetrical bending . . . . . . . . . . . . . . . . . . . . 50

3.4 Tutorial 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

4 Application of Strain Energy Methods to Deflection of Beams and Structures 57

4.1 Strain Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

4.2 Castigliano’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

5 Bending of Curved Beams with Plane Loading 67

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

5.2 Stresses and Strains in Curved Beams - Winkler Bach Analysis . . . . . . . . . . . 67

5.2.1 Assumptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

5.2.2 Case 1 - Slender Beam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

5.2.3 Case 2: Deeply Curved Beams . . . . . . . . . . . . . . . . . . . . . . . . . 69

5.3 Position of the Neutral Axis for a Deeply Curved Beam . . . . . . . . . . . . . . . 70

5.4 Bending Moment on the Cross-section . . . . . . . . . . . . . . . . . . . . . . . . 70

5.4.1 Terminologies used with Curved Beams . . . . . . . . . . . . . . . . . . . . 72

5.4.2 To determine R1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

5.5 Combined Direct and Bending Stresses . . . . . . . . . . . . . . . . . . . . . . . . 75

6 Rotating Discs and Cylinders 79

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

6.2 Circumferential and Radial Strains . . . . . . . . . . . . . . . . . . . . . . . . . . 80

6.2.1 Solid Disc with Unloaded Boundaries . . . . . . . . . . . . . . . . . . . . . 82

6.2.2 Maximum Speed for Initial Yielding . . . . . . . . . . . . . . . . . . . . . . 83

 6.2.3 Increase in Radius . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

6.2.4 Change in Thickness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

6.3 Disc with Central Hole and Unloaded Boundaries . . . . . . . . . . . . . . . . . . 85

6.4 Disc Shrunk onto a Shaft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

6.5 Disc with Loaded Outer Boundary . . . . . . . . . . . . . . . . . . . . . . . . . . 89

6.6 Disc of Uniform Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

Chapter 1

Thick walled pressure cylinders

In thin-walled pressure cylinders only two stresses, longitudinal σl and circumferential σc stresses
were considered since t << r. In thin walled cylinders there is a third stress σr (radial stress)
which is very small compared to the other stresses.
r
To support higher pressures, we must increase the wall thickness. Since t
is now small, then σr
becomes comparable to σl and σc and therefore cant be neglected

1.1 Lame’s equation

The problem of thick-walled cylinders was first solved by Lame. The resulting equations, ie, the
equations showing variation of σr and σc are termed as Lame’s equations.

Consider figure 1.1

σ r + dσ r

dr dθ dθ dθ
r 2 dr 2

σr
σc σc
R1
dθ r

R2

Figure 1.1:
Assuming unit thickness and summing forces in the radial direction, we have;
X
Fr = 0
dθ
(σr + dσr )(r + dr)dθ = σr rdθ + 2(σc dr sin )
2
dθ
(σr + dσr )(rdθ + drdθ) = σr rdθ + 2σc dr
2
σr rdθ + σr drdθ + dσr rdθ + dσr drdθ = σr rdθ + σc drdθ
σr r + σr dr + dσr r + dσr dr = σr rd + σc dr
σr dr + rdσr = σc dr
dσr
σc = σr + r (1.1)
dr
Equation 1.1 is the equilibrium equation and is applicable for stationary or rotating thick cylinder.

Recall the general stress-strain relations

1h i
ϵx = = σx − ν(σy + σz )
E
Therefore the longitudinal strain is,
1h i
ϵl = = σl − ν(σc + σr )
E
Assuming that plane surfaces remain plane, then ϵl is constant, and since σl do not vary with r,
then for ϵl to be constant,
σr + σc = = constant = 2A
σc = 2A − σr (1.2)
Equation 1.2 is the compatibility equation.

Combining equations 1.1 and 1.2 we get,

dσr
σr + r = 2A − σr
dr
dσr
2σr − 2A = −r
dr
dσr dr
= −2
σr − A r
Ln(σr − A) = −2Ln(r) + Ln(B)
B
σr − A = 2
r
B
σr = A + 2 (1.3)
r
Substituting equation 1.3 into 1.2 we get,
 B
σc = 2A − A + 2
r
B
= A− 2 (1.4)
r
Equations 1.3 and 1.4 are called the Lame’s equation. The constants A and B are determined by
using boundary conditions.
Case 1: Cylinder under internal pressure pi only

The boundary conditions are; At r = R1 , σr = −pi and at r = R2 , σr = 0. Then using equation

1.3 we can obtain the constants A and B.

At r = R2 , σr = 0,
B
0 = A+
R22
−B
A =
R22

At r = R1 , σr = −pi ,
B
−pi = A +
R21
−B B hB Bi
= + = B −
R22 R12 R12 R22
h R2 − R2 i
= B 22 21
R1 R2
−pi R12 R22
B =
R22 − R12
pi R12
A = 2
R2 − R12

Substitute A and B back into Lame’s equations

pi R12 −pi R12 R22 1

σr = −
R22 − R12 R22 − R12 r2
pi R 2 h R2 i
= 2 1 2 1 − 22
R2 − R1 r
2
pi R −pi R2 R2 1
σc = 2 1 2 + 2 1 22 2
R2 − R1 R2 − R1 r
2 h
pi R R2 i
= 2 1 2 1 + 22
R2 − R1 r

The stress variation with the cylinder radius can be plotted as shown in figure below,

Case 3: Cylinder under both internal pi and external po only

At r = R1 , σr = −pi and at r = R2 , σr = −po .

B
−pi = A +
R12
B
−po = A + 2
R2
Subtract the two equations to eliminate A,
B B
−pi + po = 2
− 2
R1 R1
h R2 − R2 i
= B 22 21
R2 R1
(po − pi )R22 R12
B =
R22 − R12
B (po − pi )R22 R12
A = −pi − 2 = −pi −
R1 R22 − R12
−pi R22 + pi R12 − po R22 + pi R22 pi R12 − po R22
= =
R22 − R12 R22 − R12

Therefore stresses are,

pi R12 − po R22 (po − pi )R22 R12 1

σr = +
R22 − R12 R22 − R12 r2
pi R12 − po R22 (po − pi )R22 R12 1
σc = −
R22 − R12 R22 − R12 r2

1.1.1 Longitudinal and Maximum Shear Stress

It is assumed in Lame’s theory that the longitudinal stress is uniform across a section and its
value is obtained by considering the equilibrium of forces exerted on the end of the cylinder

These have to be defined based on specific loading condition. For instance in case of a cylinder
under internal pressure pi only and with closed ends, we have,.

σl

R1

pi

R2

σl

Figure 1.2:
 X
Faxial = 0
σl π(R22 − R12 ) = pi πR12
h R2 i pi R12
1
σl = pi 2 = 2 (1.5)
R2 − R12 R2 − R12

Equation 1.5 shows that σl is independent of r and is always positive.

It has been shown that the σc for a cylinder under internal pressure only is given by,

pi R12 h R22 i
σc = 1 +
R22 − R12 r2

which when compared to equation 1.5 indicates that σl is always less than σc .

Therefore for a cylinder under internal pressure only, then always, σc > σl > σr . Hence the
principal stresses are,

σ1 = σc
σ2 = σl
σ3 = σr

Then the maximum shear stress is,

σ1 − σ3 σc − σr
τmax = = (1.6)
2 2
Since it was seen that the bore is the highly stressed area for a cylinder under internal pressure,
then the maximum shear stress occurs at r = R1 where we have,

pi (R12 + R22 )
σ1 = σc |R1 =
R22 − R12
σ3 = σr |R1 = −pi

Therefore the maximum shear stress in the cylinder is,

1 h pi (R12 + R22 ) i
τmax = + p i
2 R22 − R12
pi R 2
= 2 2 2 (1.7)
R2 − R1

And the maximum shear stress at any other radius on the cylinder is,
1h B B i B
τmax = (A − 2 ) − (A + 2 ) = − 2
2 r r r
Example 1.1

The cylinder of a hydraulic jack has a bore of 120mm diameter and is required to operate at
an internal pressure of 80M P a. If the tensile stress on the cylinder material is not to exceed
260M P a, determine the required wall thickness of the cylinder.

Solution:
120
The internal radius is R1 = 2
= 60mm.

For a cylinder under internal pressure only we saw that the maximum tensile stress is σc at r = R1 ,
then
pi R 2 h R2 i pi R 2 h R 2 + R 2 i
σmax = σc |R1 = 2 1 2 1 + 22 = 2 1 2 1 2 2
R2 − R1 R1 R2 − R1 R1
2 2
pi (R1 + R2 )
=
R22 − R12
h 2 2i
6 6 60 + R2
260 × 10 = 80 × 10
R22 − 602
3.25R22 − 11700 = 3600 + R22
2.25R22 = 15300
R2 = 82.5mm

The thickness of the cylinder is therefore,

t = R2 − R1 = 82.5 − 60 = 22.5mm

Example 1.2

A thick walled cylinder, 0.2m internal diameter is to contain fluid at a pressure of 50M P a. Find
the necessary thickness if the maximum shearing stress is not to exceed 100M P a. What will then
be the greatest and least values of the circumferential stress in the material.

Solution: It has been seen that the maximum shearing stress occurs at the bore and is given by,
pi R22
τmax =
R22 − R12
50 × 106 R22
100 × 106 =
R22 − 0.12
100R22 − 1 = 50R22
50R22 = 1
R2 = 0.1414m

The thickness of the cylinder is therefore,

t = R2 − R1 = 0.1414 − 0.1 = 0.0414m = 41.4mm

The greatest value of the circumferential stress occurs at the bore, that is at r = R1 and is given
by,

pi R12 h R22 i pi R12 h R12 + R22 i

σc |R1 = 2 1+ 2 = 2
R2 − R12 R1 R2 − R12 R12
pi (R12 + R22 )
=
R22 − R12
50(0.12 + 0.14142 )
= = 150M P a
0.14142 + 0.12
The least value of the circumferential stress occurs at the outer surface of the cylinder, that is at
r = R2 and is given by,

pi R12 h R22 i 2pi R12

σc |R2 = 1 + =
R22 − R12 R22 R22 − R12
2 × 50 × 0.12
= = 100M P a
0.14142 + 0.12

1.2 Methods of reinforcing thick walled cylinders to ac-

commodate higher pressures

Figure 1.3:

1. Wire winding: Wind high tensile wire of rectangular cross section around the outer surface
with a sufficiently tensile stress to induce hoop compression to the cylinder. Any subsequent
internal pressure has to overcome the hoop compression stresses before tensile stresses are
set up in the cylinder.

2. Autofrettage: Apply internal pressure until yielding occurs to a known depth. When aut-
ofrettaging pressure is withdrawn, the elastic recovery causes the bore to be subjected to to
compressive hoop stresses, while tensile hoop stresses will occur to the outer surface of the
cylinder. The residual stresses in the bore must be nullified first when pressure is applied to
the cylinder, before positive stresses can be developed.
 3. Shrinking two or more cylinders together: The outer cylinder which has a smaller inside
diameter than the outside diameter of the inner cylinder is heated until the inner cylinder
can slide inside it. When the assembly is cooled, the inner cylinder will be under compressive
hoop stresses while the outer one is under tensile hoop stresses. Any subsequent internal
pressure has to overcome the hoop compression stresses before tensile stresses are set up in
the cylinder.

1.2.1 Compound Cylinder

This is a cylinder which consists of two or more thick-walled cylinders shrunk-fit together.

Compound cylinders carry higher pressures than single cylinders of similar dimensions.

When two cylinders are shrunk together, the inner cylinder cylinder behaves as a cylinder under
external pressure, that is, it’s outer surface is under compressive stresses.

When working pressure is applied, the total stress at any point is the algebraic sum of stress due
to shrinking and that due to the internal pressure.

We shall adopt the following nomenclature,

a: the inside radius of the compound cylinder.

b: the radius of the the interface of the compound cylinder.

c: the outside radius of the compound cylinder.

pc : the contact pressure introduced by the shrinkage at the common radius b.

δ: the difference between the outer cylinder’s inside diameter and the inner cylinder’s outside
diameter.

Subscripts i and o to denote the inner and outer cylinders respectively.

b
a

Figure 1.4:
Consider first the inner cylinder

The inner cylinder is under external pressure pc . Therefore at r = a, σri = 0 and at r = b,

σri = −pc .
B
0 = A+
a2
B
A = −
a2
B B B h b 2 − a2 i
−pc = A + = − + = −B
b2 a2 b 2 a2 b 2
2 2
pc a b
B = 2
b − a2
−pc b2
A = 2
b − a2
Therefore the stress distribution in the inner cylinder is,
−pc b2 p c a2 b 2 1 −pc b2 h a2 i
σri = + = 1−
b 2 − a2 b 2 − a2 r 2 b 2 − a2 r2
−pc b2 p c a2 b 2 1 −pc b2 h a2 i
σci = 2 − = 2 1+
b − a2 b 2 − a2 r 2 b − a2 r2
Stresses at the interface where r = b are,
−pc b2 h a2 i
σri |b = 2 1− = −pc
b − a2 b2
−pc b2 h a2 i h b 2 + a2 i
σci |b = 2 1+ = −p c
b − a2 b2 b 2 − a2

Consider then the outer cylinder

The outer cylinder is under internal pressure pc . Therefore at r = c, σro = 0 and at r = b,

σro = −pc .
B
0 = A+
c2
B
A = −
c2
B B B h c2 − b 2 i
−pc = A + 2 = − 2 + 2 = B
b c b b2 c 2
2 2
pc b c
B = − 2
c − b2
p c b2
A = 2
c − b2
Therefore the stress distribution in the outer cylinder is,
p c b2 −pc b2 c2 1 p c b2 h c2 i
σro = 2 + 2 = 2 1− 2
c − b2 c − b2 r 2 c − b2 r
p c b2 −pc a2 b2 1 p c b2 h c2 i
σco = 2 − = 1 +
b − a2 c 2 − b2 r 2 c 2 − b2 r2
Stresses at the interface where r = b are,

p c b2 h c2 i
σro |b = 1− = −pc
c 2 − b2 b2
p c b2 h c2 i h c 2 + b2 i
σco |b = 2 1+ = p c
c − b2 b2 c 2 − b2

Let δo be the increase in diameter of the outer cylinder, and δi be the decrease in diameter of the
inner cylinder. Therefore,

δ = δo − δi

If the ends of the cylinders are free which is the case during shrunk fitting process, then σl = 0,
and from stress-strain relationship we have,
1 h i
ϵco = σco − νo σro
Eo
1h i
ϵci = σci − νi σri
Ei
And therefore,

δo = 2bϵco
2b h i
= σco |b − νo σro |b
Eo
δi = 2bϵci
2b h i
= σc |b − νi σri |b
Ei i

Therefore,

δ = δo − δi
2b h i 2b h i
= σc |b − νo σro |b − σc |b − νi σri |b
Eo o Ei i
2b h  c2 + b2  i 2b h  b 2 + a2  i
= pc 2 + ν p
o c − − p c + ν p
i c
Eo c − b2 Ei b 2 − a2
h 1  c 2 + b2  1  b 2 + a2 i
= 2bpc + ν o + + ν i
Eo c2 − b2 E i b 2 − a2
δ
pc = h 
2 2
  i (1.8)
b2 +a2
2b E1o cc2 +b
−b2
+ ν o + 1
Ei b2 −a2
− ν i

Using this equation 1.8 we can determine the contact pressure which can created on the interface
give the shrinkage allowance δ.

In the case where the two parts of the cylinder are made from the same material, that is, Eo =
Ei = E and νo = νi = ν, then we have,
h 1  c 2 + b2  1  b 2 + a2 i
δ = 2bpc + ν + − ν
E c 2 − b2 E b 2 − a2
2pc b h [c2 (1 + ν) + b2 (1 − ν)][b2 − a2 ] + [a2 (1 + ν) + b2 (1 − ν)][c2 − b2 ]
=
E (c2 − b2 )(b2 − a2 )
2pc b h c2 b2 (1 + ν + 1 − ν) − a2 b2 (1 − ν + 1 + ν) i
=
E (c2 − b2 )(b2 − a2 )
2pc b h 2c2 b2 − 2a2 b2 4pc b3 (c2 − a2 )
= =
E (c2 − b2 )(b2 − a2 ) E(c2 − b2 )(b2 − a2 )
δE(c2 − b2 )(b2 − a2 )
pc = (1.9)
4b3 (c2 − a2 )

Example 1.3

An aluminium cylinder whose E = 70GP a and ν = 0.33 with an outer radius of 150mm and inner
radius of 100mm is to be shrunk-fitted over a steel cylinder whose E = 200GP a and ν = 0.29. If
the outer radius of the steel cylinder is 100.25mm and the inner radius is 50mm, determine the
resulting contact pressure.

Solution:

b
a

The shrinkage allowance (diametral clearance) is (100.25 − 100) × 2 = 0.5mm.

a = 50mm, b = 100mm, c = 150mm, Eo = EAl = 70GP a, Ei = ESt = 200GP a, νo = νAl = 0.33

and νi = νSt = 0.29. Therefore using equation 1.8 we can get the contact pressure as,
δ
pc = h    i
1 c2 +b2 1 b2 +a2
2b Eo c2 −b2
+ νo + Ei b2 −a2
+ νi
0.5
= h 
2 2
  i
1 0.15 +0.1 1 0.12 +0.052
2(100) 70×109 2
0.15 −0.12 + 0.33 + 200×109 0.12 −0.052
− 0.29
0.5
= = 51.3M P a
200(4.186 × 10−11+ 6.883 × 10−12 )
Example 1.4

A steel cylinder 100mm internal diameter and 150mm external diameter is to be strengthened by
shrinking it into another cylinder made of the same material with internal diameter of 149.925mm.
Find the required external diameter of the outer cylinder if contact pressure is to be 17M P a. Take
E = 200GP a.

Solution:
100
The shrinkage allowance (diametral clearance) is (150 − 149.925) = 0.075mm. a = 2
= 50mm
and b = 150
2
= 75mm, therefore,

δE(c2 − b2 )(b2 − a2 )
pc =
4b3 (c2 − a2 )
6 0.075(200 × 109 )(c2 − 752 )(752 − 502 )
17 × 10 =
4 × 753 (c2 − 502 )
143.44c2 − 358593.75 = 237.375c2 − 1318359.375
c = 101.2mm

Therefore the external diameter of the outer cylinder is 2c = 202.4mm.

Example 1.5

A cylinder made of gun metal (E = 105GP a, and ν = 0.35) is 150mm external diameter and
99.85mm internal diameter. It is forced onto a steel cylinder (E = 200GP a, and ν = 0.29) 100mm
external diameter and 50mm internal diameter.

(a) Calculate and sketch the stress distribution on both on both gun metal and steel after shrunk
fitting.

(b) The compound cylinder is subjected to an internal pressure of 80M P a. Calculate the re-
sulting radial and circumferential stresses and sketch their variation with radius.

Solution: a) The shrinkage allowance (diametral clearance) is 100 − 99.85 = 0.15mm.

a = 25mm, b = 50mm, c = 75mm, Eo = Egm = 105GP a, Ei = ESt = 200GP a, νo = νgm = 0.35

and νi = νSt = 0.29. Therefore using equation 1.8 we can get the contact pressure as,
δ
pc = h    i
1 c2 +b2 1 b2 +a2
2b Eo c2 −b2
+ νo + Ei b2 −a2
+ νi
0.15
= h 
2 2
  i
1 0.075 +0.05 1 0.052 +0.0252
2(100) 105×109 2
0.075 −0.052 + 0.35 + 200×109 0.052 −0.0252
− 0.29
0.15
= = 42.88M P a
100(2.8095 × 10−11+ 6.883 × 10−12 )
First determine radial stress distribution

The gun metal cylinder (outer cylinder) is under internal pressure pc = 42.88M P a. Therefore at
r = b = 50mm, σro |b = −42.88M P a and at r = c = 75mm, σro |c = 0.

The steel cylinder (inner cylinder) is under external pressure pc = 42.88M P a. Therefore at
r = b = 50mm, σri |b = −42.88M P a and at r = a = 25mm, σri |a = 0.

The radial stress distribution can be sketched as shown in figure below,

c
b

0 0

− 42.88MPa

Then determine circumferential stress distribution

The circumferential stresses on the gun metal (outer) cylinder,

b 2 + c2 42.88(502 + 752 )
σco |b = pc = = 111.5M P a
c 2 − b2 752 − 502
2pc b2 2 × 42.88 × 502
σco |c = 2 = = 68.6M P a
c − b2 752 − 502

The circumferential stresses on the steel (inner) cylinder,

b 2 + a2 −42.88(502 + 252 )
σci |b = −pc = = −71.5M P a
b 2 − a2 502 − 252
−2pc b2 −2 × 42.88 × 502
σci |a = 2 = = −114.3M P a
b − a2 502 − 252
The hoop stress distribution in the compound cylinder can be sketched as shown in figure below,

b) When the pressure is applied, the whole compound cylinder acts as a single cylinder under
internal pressure pi = 80M P a. Therefore the internal radius is R1 = a = 25mm and external
radius is R2 = c = 75mm.

The radial stress distribution is,

σr |a = −80M P a
p i a2 h c2 i 80 × 252 h 752 i
σr |b = 2 1− 2 = 2 1 − 2 = −12.5M P a
c − a2 b 75 − 252 50
σr |c = 0
 c
111.5
b

a
68.6MPa

− 71.5

− 114.3MPa

The circumferential stresses on the gun metal (outer) cylinder,

a2 + c 2 80(252 + 752 )
σc |a = pc = = 100M P a
c 2 − a2 752 − 252
p i a2 h c2 i 80 × 252 h 752 i
σc |b = 2 1 + = 1 + = 32.5M P a
c − a2 b2 752 − 252 502
2pc a2 2 × 80 × 252
σc |c = 2 = = 20M P a
c − a2 752 − 252
Note that the resultant stress is the algebraic sum of the stresses due to shrink-fitting and due to
internal pressure.

Resultant stresses are:

1. At steel cylinder,

σr |a = 0 + (−80) = −80M P a
σr |b = −42.88 + (−12.5) = −55.38M P a
σc |a = −114.33 + 100 = −14.33M P a
σc |b = −71.5 + 32.5 = −39M P a

2. At gun metal cylinder,

σr |b = −42.88 + (−12.5) = −55.38M P a

σr |c = 0 + 0 = 0M P a
σc |b = 111.5 + 32.5 = 144M P a
σc |c = 68.6 + 20 = 88.6M P a

The resultant radial stress distribution can be sketched as shown in figure below,
 c

− 55 .38

− 80 MPa

The resultant hoop stress distribution can be sketched as shown in figure below,

c
144MPa
b

a 88.6MPa

− 14.33MPa

− 39MPa
Chapter 2

Deformation Beyond Elastic Limit

2.1 Bending of Beams Beyond the Elastic Limit (Plastic

Bending)

If pure bending moment is applied to abeam sufficient to cause yielding of the material, the
moment (M ) versus bending angle (θ) curve is of the form shown in figure 2.1 below,

Figure 2.1:

The onset of yield is less defined than in a simple tension or compression test since it occurs first
in the outer fibres only, and extends towards the axis of the beam.

However the stress-strain relation for any particular fibre which is either in simple tension or
compression, is assumed that after yielding, the stress at a particular point remains constant for
a considerable increase in strain.
σ

σY

Figure 2.2:
Consider a beam of a symmetrical section which is subjected to a bending moment M . The stress
distribution over the section when no yielding has occurred is represented as,
σ

σ
M M

Figure 2.3:

If M is increased to My such that yielding just starts at the outer fibres of the section, we have,
σY

σY
My My

Figure 2.4:

If M is now increased beyond My yielding penetrates towards the neutral axis of the beam. In
this case the stress distribution is as shown in figure below. the maximus stress remains constant
at the yield stress as M increases.

σY

σY
M ep M ep

Figure 2.5:

On continuous application of the bending moment M , the material fully deformed plastically, and
the stress distribution over the section becomes,
 σY

σY
M fp M fp

Figure 2.6:

σY
In design, σw = S.F
, where S.F is the factor of safety which is greater or equal to one. This implies
that σw ≤ σY .

If σw > σY , then the component designed is considered unsafe, and if σw = σY implying that the
design is based on the yield stress, then the design is conservative, that is, the design is in error
but on the safe side.

For beams subjected to pure bending beyond the elastic limit, the following assumptions are made,

ˆ Fibres are on simple tension or compression

ˆ The material is elastic-perfect-plastic

ˆ Any plane cross section will remain plane during the bending, that is, strain distribution is
linear.

Recall for elastic simple bending we have,

M E σ
= =
I R y
My
σ =
I
The bending moment required to initiate yielding at the outer fibres is,
IσY
My = (2.1)
y
where y is the distance from the neutral axis of the section to the extreme fibres.

Therefore using equation 2.1, we can determine the bending moment just required to initiate
yielding of the beam.

Consider now a beam subjected to a bending moment which has caused partial yielding of the
beam and assume the section to be rectangular.
 b
σY
dy

h ho

σY
M ep M ep

Figure 2.7:

bh2
I =
12
h
y =
2
Therefore the bending moment at the onset of yielding is,

IσY bh3 2 bh2 σY

Me = = σY =
y 12 h 6
If the bending moment is increased beyond yield point, the outer fibres at top and bottom surfaces
begin to yield. as M increases plastic deformation penetrates deeper into the beam. In this case
the elastic-plastic bending moment Me/p is the sum of the moment on the elastic part (Mel ) and
the moment on the plastic part (Mpl ), that is,

Me/p = Mel + Mpl

The bending moment at the elastic part is given as,

bh2o σY
Mel =
6
To determine Mpl , lets take an infinitesimal section on the yielded part of thickness dy and at a
distance y from the neutral axis.
Z h Z h Z h
2 2 2
Mpl = 2 σY ydA = 2 σY ybdy = 2σY b ydy
ho ho ho
2 2 2
h y 2 i h2 h h2 h2o i bσY h2 h h2 i
= 2σY b = bσY − = 1 − 2o
2 ho
2
4 4 4 h
Therefore, the total elastic-plastic moment becomes,

Me/p = Mel + Mpl

bh2o σY bσY h2 h h2o i
= + 1− 2
6 4 h
bh2 σY h h2o 3  h2o i
= + 1− 2
6 h2 2 h
2
bh σY ho 3 3 h2o i
h 2
= + −
6 h2 2 2 h2
h 3 1 h2 i 3 h 1 h2o i
o
= My − = M y 1 −
2 2 h2 2 3 h2
For fully plastic case, we have ho = 0 therefore,

3 h 1 02 i 3
Mf p = My 1 − = My
2 3 h2 2
Mf p 3
=
My 2

The ratio of fully plastic moment to the fully elastic moment is called the shape factor (f ). It has
bean shown that the shape factor of a rectangular cross section is equal to 32 .

Assignment 3

For the beam section shown in figure below, show that the shape factor is given as,
b1 d21
Mp 3h1 − bd2
i
f= = b1 d31
My 2 1−
bd3

d d1
b1

Figure 2.8:
2.1.1 Plastic bending of of beams with one axis of symmetry

For symmetrical beams, the centroidal axis coincide with the neutral axis. For asymmetrical
beams, the centroidal axis doesn’t necessarily coincide with the neutral axis.

Consider the beam shown in the figure below which has one axis of symmetry.

σY

F1
C1
_
y1 A1

_
y2
F2 C2 A2

σY

Figure 2.9:

The beam is bend to fully plastic condition about oz− axis perpendicular to oy.

A1 is the area of the section above the plastic neutral axis, and A2 is the area below the plastic
neutral axis.

F1 = σY A1
F2 = σY A2
For static equilibrium,
F1 = F2
σY A1 = σY A2
A1 = A2
Note: For fully plastic condition, the neutral axis divide the cross section into two equal areas.
the neutral axis for fully plastic condition is generally different from the neutral axis for linear
elastic ending.

Let C1 be the centroid of area A1 at a distance y 1 from the plastic neutral axis, and C2 be the
centroid of area A2 at a distance y 2 from the plastic neutral axis.
Mf p = F1 y 1 + F2 y 2 = σY A1 + σY A2
1 1
= AσY y 1 + AσY y 2
2 2
1  
= AσY y 1 + y 2
2
To illustrate how this concept is applied, lets consider a section of rectangular shape, bend to fully
plastic condition. The fully plastic moment is,
A1

C1 h
_ h
y1 2
4
h
_ h
y2
4 h
C2
2

A2

Figure 2.10:

1  
Mf p = AσY y 1 + y 2
2
1  h h  σ bh2
Y
= bhσY + =
2 4 4 4
Bending moment at the onset of yielding is,

σY I bh3 2 σY bh2
My = = σY =
y 12 h 6
Therefore the shape factor is,
σY bh2
Mf p 4 3
f = = σY bh2
=
My 6
2

Example 2.1

Determine the shape factor of the T-section shown in figure below. All units are in mm. First
lets find the bending moment at onset of yielding,

y = 70.6mm
IZ = 2.1 × 106 mm4
σY IZ 2.1 × 10−6 σY
My = = = 2.9745 × 10−5 σY
y 0.0706
 100

A1 x _ C1
y1 PNA
12
PNA

_
y2 C
ENA ENA z

100
C2
A2
_
y

12

Figure 2.11:

To determine the fully plastic bending moment, lets first locate the position of the plastic neutral
axis which divides the section into two equal areas.

A1 = A2
100x = 100(12 − x) + 88 × 12
x = 11.28mm
x 11.28
y1 = = = 5.64mm
2 2
(12 × 88 × 44.72) + (0.72 × 100 × 0.36
y2 = = 41.9mm
12 × 88 + 0.72 × 100
1  
Mf p = AσY y 1 + y 2
2
1 h i
= σY (100 × 12 + 88 × 12) × 10−6 41.9 + 5.64 × 10−3 = 5.36251 × 10−5 σY
2
Shape factor therefore is,

Mf p 5.36251 × 10−5 σY
f = = = 1.8
My 2.9745 × 10−5 σY

Example 2.2

A 8cm × 3cm rectangular bar is used as a simply supported beam on a span of 1.2m and loaded
at mid span. The yield stress is 275M P a and the long edges of the section are vertical. Assuming
that after yielding, the stress remains constant at 275M P a, determine the load required to cause
yielding for a depth 1cm at the top and bottom of the section at mid span.

Solution: For a simply supported beam loaded at the mid-span, the bending moment is;
PL
M =
4
The elastic-plastic moment is,

bh2 σY h h2o 3 3 h2o i bh2 σY h 3 1 h2o i

Me/p = + − = −
6 h2 2 2 h2 6 2 2 h2
0.03(0.08)2 275 × 106 h 3 1 62 i
= − = 10725N m
6 2 2 82
The required load at mid span is,
4Me/p 4 × 10725
P = = = 35750N
L 1.2

Example 2.3

The flange and web of a T-section are each 12mm thick, the flange width is 100mm and the overall
depth is 100mm. The beam is simply supported over a length of 2m and it is subjected to a point
load W at the mid-span. Calculate the maximum value of W if the beam is to be designed such
that yielding is permitted to penetrate the web to a depth of 20mm. The yield stress of the beam
material is 300M P a.

Solution: Yielding starts at the bottom of the web, and the stress distribution is,
100

σ'

12 F4
68 − d σ"
F3

100 d
F2
σY

20
F1

σY
12

Figure 2.12:
Using similarity, we have,

σY (68 − d)
σ” =
d
σ Y (80 − d)
σ′ =
d
The position of the neutral axis is defined by the following condition,

force above the neutral axis=force below the neutral axis, that is

F1 + F2 = F3 + F4

The individual forces are,

F1 = 300 × 106 (0.02 × 0.012) = 72000N

1 
F2 = 300 × 106 × 0.012 × d × 10−3 = 1800dN
2
1  
F3 = (σ” + 0) 0.012(68 − d) × 10−3
2
300 × 106 (68 − d)  1 
= × 0.012(68 − d) × 10−3
d 2
 (68 − d)2 
= 1800
d
1 ′  148 − 2d 
F4 = (σ + σ”)(0.012 × 0.1) = 180000
2 d
Therefore,

F1 + F2 = F3 + F4
 (68 − d)2   148 − 2d 
72000 + 1800d = 1800 + 180000
d d
 (68 − d)2   148 − 2d 
72 + 1.8d = 1.8 + 180
d d
676.8d = 34963.2
d = 51.66mm

The forces are F1 = 72kN , F2 = 93kN , F3 = 9.3kN and F1 = 1555.7kN , Therefore the bending
moment is obtained by summing the moments due to these forces about the neutral axis.
2d 2 
M = F1 (d + 10) + F2 + F3 (68 − d) + F4 (68 − d + 6) = 11.2kN m
3 3
But for a simply supported beam with a point load at mid-span
WL
M =
4
4M 4 × 11.2
W = = = 22.4kN
L 2
2.2 Torsion of Shafts Beyond the Elastic Limit (Plastic
Torsion)
T Gθ τ
Equation J
= L
= r
applies only to elastic torsion

In torsion beyond the elastic limit, the following assumptions are made during analysis,

1. The material is elastic perfect plastic

2. A plane section remain plane as in the case of elastic torsion

3. Radial lines remain radial during twisting as in the case of elastic torsion.

From equation TJ = τr , if applied torque is less that the torque which can initiate yielding at the
surface of the shaft, that is T < TY shear stress increases from zero at the shaft axis to a maximum
at the surface of the shaft. This is shown below,

τ max

O ro r

Figure 2.13:
When the torque is increased such that yielding just starts at the surface of the shaft, then
τmax = τY , and the shear stress distribution is as shown below, In this case,

TY

TY
τ

τY

O ro r

Figure 2.14:

TY τY
=
J ro
πro4 τY πr3
TY = = o τY (2.2)
2 ro 2
Using equation above, we can determine the torque which the solid shaft can accommodate before
yielding starts at the surface of the shaft.

If torque is increased beyond TY , plasticity penetrates towards the center of the shaft, and the
shear stress distribution is as shown below, In this case the elastic-plastic torque Te/p is the sum
of the torque on the elastic part (Tel ) and the torque on the plastic part (Tpl ), that is,

Te/p = Tel + Tpl

The torque at the elastic part is given as,

πre3
Tel = τY
2
The torque on the plastic part is,
Z ro
2πτY  3 
Tpl = τY 2πrdr.r = ro − re3
re 3
Therefore the total elastic-plastic torque is,
πre3 2πτY  3 
Te/p = Tel + Tpl = τY + ro − re3
2 3
2πro3 τY πre3 τY
= −
3 6
2πro3 τY  r3 
= 1 − e3
3 4ro
 Tep
ro re

r
Tep
dr

τ
τY

O re ro r

Figure 2.15:

If the whole section is fully plastic, then re = 0 and,

2πro3 τY   2πr3 τ
o Y
Tf p = 1−0 =
3 3
The ratio of fully plastic torque to the torque at onset of yielding is,
2πro3 τY
Tf p 3 4
= πro3
=
TY τ 3
2 Y
1
This implies that for a solid shaft, the torque at yield point needs to be increased by 3
to yield
the shaft completely.

Angle of twist at plastic torsion

Gθ
From equation L
= τr , the angle of twist at yield point is,
τY L
θY =
Gro
Beyond yield point, (elastic-plastic condition),

The angle of twist at the elastic region is,

τY L
θel =
Gre
Since radial lines remain radial during plastic torsion, the angle of twist at the plastic region is
equal to the angle of twist at the elastic region, that is,
τY L
θpl = θel =
Gre
 Tep
ro re

r
Tep

Figure 2.16:

Therefore,
τY L
θpl Gre ro
= τY L
=
θY Gro
re
r 
o
θpl = θY
re
As re → 0, that is for fully plastic situation, θpl → ∞, implying that a shaft will rotate many
times before failing.

Example 2.4

A mild steel shaft in a power transmission system is 40mm diameter and 250mm length. It is
transmitting a torque of 1800N m which is known to have caused shear yielding on the shaft.
determine the radial depth to which plasticity has penetrated, and the angle of twist if τY =
120M P a and G = 80GP a.

Solution: ro = 402
= 20mm, Tep = 1800N m and L = 250mm. The elastic-plastic torque in a solid
shaft is given by,
2πro3 τY  re3 
Te/p = 1− 3
3 4ro
2π(0.02) × 120 × 106 
3
re3 
1800 = 1−
3 4(0.02)3
re = 15mm

The radial depth to which plasticity has penetrated is,

d = ro − re = 20 − 15 = 5mm

Angle of twist is,

τY L 120 × 106 (0.25)
θ = = = 0.025rad = 1.430
Gre 80 × 109 (0.015)
2.3 Tutorial 1
Q1 A beam has a T-section and is made by welding together two plates, the cross section of
which are rectangular with dimensions 160mm×10mm. The beam is subjected to a bending
moment of such magnitude that yielding occurs at the lower part of the over a depth of 40mm
of the web, although yielding does not occur at the top surface of the flange. The yield stress
of 250M P a may be assumed constant over the yielded area, while over the remainder of the
section, the stress is proportional to the distance from the neutral axis. Determine the
position of the neutral axis, the stress at the top of the flange and the moment of resistance
of the section.(ANS: Neutral axis is 45mm from the top edge, M = 24.5kN m)

Q2 A compound steel cylinder has a bore of 80mm and an outside diameter of 160mm, the
diameter of the common surface being 120mm. Find the radial pressure pressure at the
common surface which must be provided by the shrinkage, if the resultant maximum hoop
tension in the inner cylinder under a superimposed pressure of 60M P a is to be half the value
of the maximum hoop stress which would be produced in the inner cylinder if that cylinder
alone was subjected to an internal pressure of 60M P a. (ANS: 6.11M P a)
Determine the hoop stresses at the inner and outer surfaces of both cylinders under the
internal pressure of 60M P a, and sketch a graph to show how the hoop stresses varies across
the cylinder wall.(ANS: Inner cylinder: -78MPa, -39.6MPa, Outer cylinder: -77.3MPa, -
55.7MPa)

Q3 Show that the shape factor for a beam of elastic-perfect material subjected to pure bending
and of a circular cross-section is 1.7.

Q4 A shaft having a diameter of 90mm is turned down to 87mm diameter for part of its length.
If a torque is applied to the shaft of sufficient magnitude just to produce yielding at the
surface of the shaft in the unturned part, determine the depth of yielding which would occur
in the turned part. Find also the ratio of angle of twist per unit length in the turned part
to that in the unturned part of the shaft.AN S; 5.3mm, 1.18)

Q5 A hollow shaft 50mm outside diameter is made of steel with a yield stress in shear of 150M P a
and a modulus of rigidity 83M P a. Calculate the torque and the angle of twist when the
material first yields if the length of the shaft is 2m. On the assumption that the shear yield
stress remains constant after yielding, calculate the depth of penetration of plastic yield for
an increase in torque of 10% above the initial yield. Determine also the angle of twist of the
shaft at the increased torque.(AN S : 3.45kN m, 8.290 , 2.84mm, 9.340 )

Q6 A solid circular shaft is subjected to pure torsion and the material is elastic-perfect-plastic
with a yield stress in shear of 152M P a. When the shear stress at one-third of the radius
from the center of the shaft reaches the yield stress, determine the shear strain on the outer
surface. Also find the ratio of the torque carried in the above condition to the maximum
elastic torque for the shaft. Take G = 83GP a. (AN S : 5.5 × 10−3 , 1.32)
Q7 A rectangular beam 30mm wide and 50mm deep is simply supported over a length of 2m. If
it is subjected to a uniformly distributed load of 8kN/m, calculate the depth of penetration
of plastic deformation in the beam. The yield stress of the beam material is 240M P a.
Chapter 3

Unsymmetrical Bending of beams

3.1 Properties of area

3.1.1 Second Moment of area

If each element of an area is multiplied by the square of its distance from an axis, then the
summation of these quantities for the whole area is termed as the second moment of area (I)
about the axis.
Y

x dA

r
y
X X

Figure 3.1:
Z
IXX = y 2 dA
Z
IY Y = x2 dA
R
IZZ = r2 dA is the second moment of area about the axis ZZ perpendicular to the paper, and is
termed as the polar second moment of area (J).

The stresses and radius of curvatures in a beam for a given bending moment are proportional to I
of the cross section about an axis through the centroid and perpendicular to the plane of bending.

Values of I for some simple cases,

 Y
b
Y

X X X X
d

d
O O Y
Y
bd 3 db 3 bd 3 πd 4 πr 4 πd 4 πr 4
I XX = I YY = I OO = I XX = I YY = = I ZZ = J = =
12 12 3 64 4 63 2

Figure 3.2:

3.1.2 Parallel axes theorem

The second moment of area of a plane about any axis is equal to the second moment of area of
the same plane about a parallel axis through the centroid plus the product of the area and the
square of the distance between the axes.

dA

y
C
X X
h

O O

Figure 3.3:

The second moment of area of an element of area dA about axis OO is,

dIOO = (y + h)2 dA = y 2 dA + 2yhdA + h2 dA

Therefore for the whole area,

Z Z Z
2 2
IOO = y dA + h dA + 2h ydA
Z
2
= IXX + h A + 2h ydA
R
Since XX passes through the centroid, the total first moment of area about XX is ydA = 0,
therefore,

IOO = IXX + Ah2 (3.1)

which is the parallel axes theorem.

3.1.3 Perpendicular axes theorem

Consider figure 3.4.

r y
X X

Figure 3.4:

Let the second moments of area of the element about axes XX, Y Y and ZZ be dIXX , dIY Y and
dJ respectively, then,

dIXX + dIY Y = y 2 dA + x2 dA = (y 2 + x2 )dA = r2 dA = dJ

Therefore for the whole area,

IXX + IY Y = IZZ = J (3.2)

which is the perpendicular axes theorem.

3.1.4 Centroid

The centroid or center of area is defined as the point where the whole area of the section is assumed
to be concentrated.

Centroids can be taken as quite analogous to center of gravity, when bodies have area only and
not weight.

Positions of centroid for plane geometric figures

The table 3.1 below shows the positions of centroids of the shapes shown in figure 3.5.
 (i ) y (ii ) y (iii ) y

φd
_ _
x x G
h G h _
x G
_ _
y _ y
y
x x x
b b

(iv ) y (v ) y

φd
φd
_
_
x G x G
_
y
x x

Figure 3.5:

Determination of centroids

Example 3.1

Find out the position of the centroid of the L− section shown below. All units are in mm.

Procedure: Divide the figure into two simple areas, that is 120 × 40 and 40 × 120.

The x− coordinate of the centroid is,

P
Ai xi 384000
x = P = = 40mm
Ai 9600

The y− coordinate of the centroid is,

P
A i yi 576000
y = P = = 60mm
Ai 9600

Example 3.2

Determine the position of the centroid of the I−section which is shown in the figure below. All
units are in mm.

Procedure: Divide the figure into three simple areas, that is portions (1), (2) and (3) of area
150 × 30 and 20 × 250 and 100 × 20 respectively.
 S/no. Section Shape x-coordinate of the y-coordinate of the
_ _
centroid ( x ) centroid ( y )
(i) Rectangle _ _
b h
x= y=
2 2
(ii) Triangle _ _
b h
x= y=
3 3
(iii) Circle _ _
d d
x= y=
2 2
(iv) Semi-circle _ _
d 2d
x= y=
2 3π
(v) Quarter-circle _ _
2d 2d
x= y=
3π 3π

Table 3.1:

Portion Area ( Ai ) ( mm 2 ) Centroidal Centroidal Ai xi Ai y i

(i) distance distance 3
( mm ) ( mm 3 )
along x- along y-axis
axis ( xi ) ( yi )
1 120 × 40 = 4800 60 20 4800 × 60 4800 × 20
. = 288000 . = 96000
2 40 × 120 = 4800 20 100 4800 × 20 4800 × 60
. = 96000 . = 288000
∑ 9600 384000 576000

Portion Area ( Ai ) ( mm 2 ) Centroidal Centroidal Ai xi Ai y i

(i) distance distance 3
( mm ) ( mm 3 )
along x- along y-axis
axis ( xi ) ( yi )
1 150 × 30 = 4500 75 15 4500 × 75 4500 × 15
. = 337500 . = 67500
2 20 × 250 = 5000 75 155 5000 × 75 5000 × 155
. = 375000 . = 775000
3 100 × 20 = 2000 75 290 2000 × 75 2000 × 290
. = 150000 . = 580000
∑ 11500 862500 1422500
 y
40

120
_

160 x G

_
y

40

120
x

100
y
(3) 20

(2)
_
G
x
300
20
_
y

30 (1)
x
150

The x− coordinate of the centroid is,

P
Ai xi 862500
x = P = = 75mm
Ai 11500

The y− coordinate of the centroid is,

P
Ai yi 1422500
y = P = = 123.7mm
Ai 11500
3.1.5 Product of inertia

Consider a plane of area A as shown in figure below,

x dA

Figure 3.6:

Further consider an elemental area dA at a distance x and y from the y−axis and x−axis re-
P
spectively. The summation xydA is defined as the product of inertia of the cross section, that
is,
X Z
IXY = xydA = xydA (3.3)

Parallel axes theorem for product of inertia

Consider figure 3.7 below.

When the product of inertia of a plane area is known about its centroidal axes XY , then the
product of inertia of the plane about any other axes, say X ′ Y ′ parallel to centroidal axes XY , is
given as,

IX ′ Y ′ = IXY + A × a × b (3.4)

Principal axes and Principal moments of inertia

The two axes about which the product of inertia of a plane is zero are called the principal axes.
The moment of inertia (second moment of area) about a principal axis is called the principal
moment of inertia.

If a figure has an axes of symmetry, then that axis is a principal axis.

Determination of principal moments of inertia and directions of principal axes.

 Y'
Y

Area ( A)

a G
X

X'

Figure 3.7:

Consider figure 3.8 below. XY is the coordinate axis through the centroid G of the section, while
X ′ Y ′ is another coordinate axis which is inclined at an angle θ to the centroidal XY axes.

dA ia a small elemental area with coordinates (x, y) with respect to XY coordinate axis, and
(x′ , y ′ ) with respect to the X ′ Y ′ coordinate axis.

dA
x'
Y' S θ
θ
y' X'
M

y
L N
θ
G X

Figure 3.8:
From the figure,

x′ = LM + x cos θ = GN + x cos θ = y sin θ + x cos θ

y ′ = GL = M N = y cos θ − x sin θ

Therefore, the product of inertia of the plane about X ′ Y ′ coordinate axis is,
Z Z h i
′ ′
IX ′ Y ′ = x y dA = (y sin θ + x cos θ)(y cos θ − x sin θ) dA
Z h i
= y 2 sin θ cos θ − xy sin2 θ + xy cos2 θ − x2 cos θ sin θ dA
Z h i Z h i Z h i
2 2 2
= xy(cos θ − sin θ) dA − x cos θ sin θ dA + y 2 cos θ sin θ dA

From trigonometric identities, cos θ sin θ = 21 sin 2θ and cos2 θ − sin2 θ = cos 2θ, therefore,
Z Z Z
sin 2θ 2 sin 2θ
IX Y = cos 2θ xydA −
′ ′ x dA + y 2 dA
2 2
Z Z Z
sin 2θ h i
= cos 2θ xydA − x2 dA − y 2 dA
2
R R R
But it was previously seen that, xydA = IXY , x2 dA = IY and y 2 dA = IX , therefore,

(IY − IX )
IX ′ Y ′ = IXY cos 2θ − sin 2θ (3.5)
2
This is the general equation for finding the product of inertia about any two perpendicular axes
if the moments of inertia and product of inertia about two other perpendicular axes are known.

The moments of inertia about the X ′ − and Y ′ − axes are derived as follows,
Z Z
′2
IX ′ = y dA = (y cos θ − x sin θ)2 dA
Z h i
2 2 2 2
= y cos θ − 2xy cos θ sin θ + x sin θ dA
Z Z Z
= (y cos θ)dA − (2xy cos θ sin θ)dA + (x2 sin2 θ)dA
2 2

From trigonometric identities, cos2 θ = 1+cos

2
2θ
, 2 cos θ sin θ = sin 2θ and sin2 θ = 1−cos
2
2θ
, therefore,

1 − cos 2θ
Z Z Z
1 + cos 2θ 2
IX ′ = y dA − sin 2θ xydA + x2 dA
2 2
IX IY
= (1 + cos 2θ) − IXY sin 2θ + (1 − cos 2θ)
2 2
IY + IX h IY − IX i
= − cos 2θ − IXY sin 2θ (3.6)
2 2
Similarly, it can be shown that,
IY + IX h IY − IX i
IY ′ = + cos 2θ + IXY sin 2θ (3.7)
2 2
The product of inertia of a section about the principal axes is zero. Hence, when IX ′ Y ′ is zero,
then the axes X ′ and Y ′ become the principal axes. Hence,
IY − IX
0 = IXY cos 2θ − sin 2θ
2
2IXY
tan 2θ =
IY − IX
Let,

U U be the major principal axis

V V be the minor principal axis

IU be the maximum principal moment of inertia

IV be the minimum principal moment of inertia

Then, it follows that,

IY + IX 1
q
2
IU = + (IY − IX )2 + 4IXY (3.8)
2 2q
IY + IX 1 2
IV = − (IY − IX )2 + 4IXY (3.9)
2 2
Adding equations 3.6 and 3.7

IX ′ + IY ′ = IX + IY

Adding equations 3.8 and 3.9

IU + IV = IX + IY

Hence this shows that the sum of moments of inertia about any two perpendicular axis remains
constant.

Example 3. 3

Find the product of inertia for the section shown in figure below about axes XX and Y Y , and
hence find the product of inertia about a parallel axis X1 X1 and Y1 Y1 passing through the centroid.
All units are in mm.

Procedure: Divide the figure into two simple areas, that is 200 × 20 and 20 × 100.

The product of inertia about XY axis is,

X
IXY = Ai xi yi = −540 × 104 mm4
 200

X Y Y1 X

_
y (1) 20

_
x
120
G X1
100

( 2)

20
Y

Portion Area ( Ai ) Centroida Centroida Ai xi y i Ai xi Ai y i

(i) l distance l distance
( mm 2 ) mm 4 ( mm 3 ) ( mm 3 )
along x- along y-
axis ( xi ) axis ( y i )
1 200 × 20 100 -10 4000 × 100 × −10
= 4000 = −4000000 4000 × 100 4000 × −10
. = 400000 . = −40000

2 20 × 100 10 -70 2000 × 10 × −70 2000 × −70

= 2000 = −1400000 2000 × 10 . = −140000
. = 20000
∑ 6000 -5400000 420000 -180000

The position of the centroid is,

P
Ai xi 420000
x = P = = 70mm
Ai 6000

P
Ai yi −180000
y = P = = −30mm
Ai 6000

Using parallel axis theorem of product of inertias to determine IX1 Y1 . NOTE:X1 Y1 axis is the
centroidal axis.

IX1 Y1 = IXY + A × a × b
IXY = IX1 Y1 − A × a × b = −540 × 104 − (6000 × 70 × −30) = 720 × 104 mm4
Example 3.4

Find the principal moments of inertia and the directions of principal axes of the angle section
shown in figure below. All units are in cm.

Y 10

(2)
50 _
x
60 G

_
y (1) 10

40 X

Procedure: Divide the figure into two simple areas, that is 40 × 10 and 10 × 50 to first locate the
centroid.

Portio Area Centroida Centroida Ai xi Ai y i

n (i) ( Ai ) l distance l distance
( cm 3 ) ( cm 3 )
2 along x- along y-
( cm )
axis ( xi ) axis ( y i )
1 40 × 10 20 5 400 × 20
= 400 . = 8000 400 × 5
. = 2000
2 10 × 50 5 35 500 × 5 500 × 35
= 500 . = 2500 . = 17500
∑ 900 10500 19500

The position of the centroid is,

P
A i xi 10500
x = P = = 11.6667cm
Ai 900

P
Ai yi 19500
y = P = = 21.6667cm
Ai 900

To determine the principal moments of inertia, all the moments of inertia and product of inertia
must be about the centroidal axis.
 Y

Y1

_
x 22.50
G X1

_
y

Using parallel axis theorem of product of inertias to determine IX1 Y1 . NOTE:X1 Y1 axis is the
centroidal axis.

IXY = IX1 Y1 + A × a × b
X X
IX1 Y1 = IXY − A × a × b = A i xi y i − Ai × x × y
 
= (400 × 20 × 5) + (500 × 5 × 35) − 900 × 21.6667 × 11.6667
= −100000cm4

To determine the moments of inertia about the centroidal axis, we need to apply the parallel axis
theorem.
h 40 × 103 i h 10 × 503 i
IX1 = + 400(21.6667 − 5)2 + + 500(35 − 21.6667)2
12 12
4
= 307500cm

and
h 10 × 403 i h 50 × 103 i
2 2
IY1 = + 400(20 − 11.6667) + + 500(11.6667 − 5)
12 12
= 107500cm4

Maximum principal moment of inertia is,

IY1 + IX1 1 q 2
IU = + (IY1 − IX1 )2 + 4IX 1 Y1
2 2
107500 + 307500 1 p
= + (107500 − 307500)2 + 4(−100000)2
2 2
= 348921.4mm4
Minimum principal moment of inertia is,
IY1 + IX1 1q 2
IV = − (IY1 − IX1 )2 + 4IX 1 Y1
2 2
107500 + 307500 1 p
= − (107500 − 307500)2 + 4(−100000)2
2 2
= 66078.6mm4

The direction of the principal axes is,

2IXY
tan 2θ =
IY − IX
1 h 2I
XY
i
θ = tan−1
2 IY − IX
1 −1
h 2 × −100000 i
= tan = 22.50
2 107500 − 307500
Therefore the principal axes (U V ) is inclined at an angle of 22.50 to the centroidal axes as shown
in figure above.

3.2 Revision of simple bending theory

If a beam is subjected to pure bending moment M , the fibres in the upper part are in tension while
those in the lower part are in compression. Hence tensile and compressive stresses are induced in

M M

Figure 3.9:

the beam which produce a moment of resistance.

In simple theory of bending, the following assumptions are made,

(a) Beam is initially straight and the radius of curvature is large compared to the dimensions
of cross-section.

(b) Material is homogeneous,elastic and obeys Hooke’s law.

(c) The young modulus of the material is the same in compression and tension.
 (d) Stresses are uniform across the width if the section

(e) The cross-section is symmetrical about the plane of bending.

(f) Plane sections remain plane even after bending.

(g) Every longitudinal fibre is free to extend or contract without being restrained by the neigh-
bouring fibres.

3.2.1 Stress due to simple bending

Figure 3.10 shows the cross-section of a beam to which is applied a bending moment M acting in
a vertical plane through the centroid G.
a d

dA
y
u
X X
G
M M

c b

Figure 3.10:

Since section ab remains plane after bending, then its new position is cd. abcd then represents the
strain distribution for the cross-section, ac and bd representing the strains at the top and bottom
faces respectively

It is seen that strain varies linearly from a maximum at the top fibres to another maximum at the
bottom fibres (strain changes from tensile to compressive.

Axis XX at which the strain is zero is termed as the neutral axis. The neutral axis must pass
through the centroid of the cross-section.

The strain at any point and hence the stress is directly proportional to the distance of that point
from the neutral axis.

If stress on a point at a distance y from the NA is σ, then the stress on element DA at a distance
u from the NA is uy σ and therefore force on area dA is dF = uy σdA.

Moment of force about XX is,

u
dM = σdAu
y
and the total moment of force about XX which is equal to the applied moment is,
Z
σ σ
M = u2 dA = I
y y
M σ
= (3.10)
I y

3.2.2 Radius of curvature

Due to the action of the bending moment, the beam will bend. Let the radius of curvature of the
NA at a particular section be R,

c
e
a d
y
b

Figure 3.11:

A layer initially at of length ce has been extended to cd, therefore strain is,
ed ed σ
ϵ= = =
ce ab E
from similarity,
ed ab
=
bd ob
ed bd y
= =
ab ob R
Therefore,
σ y
=
E R
σ E
= (3.11)
y R
Combining equations 3.10 and 3.11, we get the general bending formula as,
M σ E
= = (3.12)
I y R
Example 3. 5

The figure below represents the cross-section of an extruded alloy member, which acts as a simply
supported beam with the 60mm wide flange at the bottom. Determine the bending moment
which can be applied on the section, if the permissible stresses in tension and compression are
respectively 60M P a and 45M P a.
140 mm
20 mm

20 mm

10 mm

60mm
_
5mm
y

5mm

60 mm

Figure 3.12:

Let ȳ be the distance of the NA XX above the base of the section. Then,

(20 × 20)70 + (140 × 5)57.5 + (50 × 5)30 + (60 × 5)2.5 = (20 × 20) + (140 × 5)

+(50 × 5) + (60 × 5) ȳ
ȳ = 46.36mm

Therefore, the second moment of area of the section about the NA XX is,
h 20 × 203 i h 140 × 53 i
IXX = + 20 × 20 × 23.642 + + 140 × 5 × 11.142 +
12 12
h 5 × 503 i h 60 × 53 i
+ 5 × 50 × 16.362 + + 60 × 5 × 43.862
12 12
4
= 1021931.84mm

For a maximum tensile stress of 60M P a, bending moment is,

σIXX 1021931.84 × 10−12 × 60 × 106
M1 = =
y 0.04636
= 1322.6N m
For a maximum compressive stress of 45M P a, bending moment is,

σIXX 1021931.84 × 10−12 × 45 × 106

M2 = =
y 0.03364
= 1367N m

Therefore the maximum permissible bending moment is 1322.6N m

3.3 Unsymmetrical bending

In symmetrical bending in which equation MI = σy is used, its assumed that the neutral axis of the
cross-section of the beam is perpendicular to the plane of loading or plane of bending.

In case of unsymmetrical bending, the direction of neutral axis is not perpendicular to the plane
of loading/bending. the following are two reasons for unsymmetrical bending,

(a) The cross section itself is unsymmetrical, such as an angle section.

(b) The cross-section is symmetrical, but the load line is inclined at an angle to the principal
axis.

For any shape of cross-section, there is always at least one pair of mutually perpendicular axis
y and z on which Iyz = 0. These axes are called principal axes of the cross section. Further Iy
and Iz about the two perpendicular principal axis, are the maximum and minimum values of the
moments of inertia.

If the section has an axis of symmetry, this axis is a principal axis.

If the section is symmetrical about the y− axis which lies in the assumed plane of bending, Iyz = 0,
and the beam is in moment equilibrium about this axis, and equation, MI = σy holds.

If the section is not symmetrical, then Iyz ̸= 0, that is equilibrium is not possible without an
externally applied moment about the y−axis for the bending to be confined in xy plane. In
practice, such moment is absent and bending occurs not only in xy plane but also in xz plane.
The neutral axis which still passes through the centroid, is no longer horizontal

Consider the following sections,

(a) In this case, the section is symmetrical about y and z axes through the centroid. Hence these
are the principal axes, Iy , and Iz are principal moments of inertia, and Iyz = 0. Therefore
equation MI = σy holds.
M σ
(b) Same comments as in (a) above and equation I
= y
holds
 (a ) (b) P (c ) P
P

C C C
z z
z

y y
y

(d ) (e) (f) P

C C
C z z
z

y
y
y

Figure 3.13:

(c) In this case, the section is symmetrical about any diameter, and therefore there are infinite
number of pairs of principal axes. Iy , and Iz are the principal moments of inertia and are
the same for each pair of principal axes. Also Iyz = 0, hence equation MI = σy holds.

(d) In this case, the section is only symmetrical about the z−axis, therefore this axis is a principal
axis and an axis perpendicular to it (i.e y-axis) is also a principal axis. This implies that
Iyz = 0, and hence equation MI = σy holds.

(e) In this case, there is no plane of symmetry on the section, hence the y and z axes are not
the principal axes. The principal axes must be inclined to the yz plane. The lateral loading
in y direction causes the beam to deflect in both y and z direction. Also Iyz ̸= 0, and hence
equation MI = σy is not valid.

(f) Since the plane of loading is not perpendicular to the y-axis, the lateral loading in the y
direction causes the beam to deflect in both y and z direction, hence equation MI = σy is not
valid.

3.3.1 Stresses due to unsymmetrical bending

Consider figure 3.14.

 Mz
z

Mz
z

My
y
y
z
x
My
dA

Figure 3.14:

Origin O is at some point on neutral axis. External moments My and Mz are applied about an
arbitrary centroidal axes.

Consider first bending in the xy-plane only, then,

Z
Mz = σx ydA
Z
My = σx zdA

yE
Let the radius of curvature in the xy−plane be Ry , then σx = Ry
, and therefore,
Z Z Z
yE E EIZZ
Mz = σx ydA = ydA = y 2 dA =
Ry Ry Ry
Z Z Z
yE E EIY Z
My = σx zdA = zdA = yzdA =
Ry Ry Ry

Consider now bending in the xz-plane only, and let the radius of curvature in the xz−plane be
Rz , then σx = zE
Rz
, and therefore,
Z Z Z
zE E EIY Z
Mz = σx ydA = ydA = yzdA =
Rz Rz Rz
Z Z Z
zE E EI YY
My = σx zdA = zdA = z 2 dA =
Rz Rz Rz
For simultaneous bending in the xy and xz−planes, we superimpose equations above correspond-
ingly.
EIZZ EIY Z
Mz = + (3.13)
Ry Rz
EIY Y EIY Z
My = + (3.14)
Rz Ry
 1
From equation 3.14, we can make Rz
the subject of formula and substitute it in equation 3.13 to
solve for R1y
1 1 h EIY Z i
= My −
Rz EIY Y Ry
EIZZ EIY Z EIY Z i
Mz = + [My −
Ry EIY Y Ry
2 i
1 h EIY Z My IY Z
= EIZZ − +
Ry IY Y IY Y
Mz IY Y − My IY Z E IY Y IZZ − IY2 Z
h i
=
IY Y Ry IY Y
1 Mz IY Y − My IY Z
=
Ry E(IY Y IZZ − IY2 Z )
1
Therefore Rz
becomes,
1 1 h EIY Z (Mz IY Y − My IY Z ) i
= My −
Rz EIY Y E(IY Y IZZ − IY2 Z )
1 h My IY Y IZZ − My IY2 Z − Mz IY Z IY Y + My IY2 Z i
=
EIY Y IY Y IY Z − IY2 Z
My IZZ − Mz IY Z
=
E(IY Y IZZ − IY2 Z )
The resultant stress is equal to the sum of stress components in each xy and xz planes, that is;
yE zE
σx = +
Ry Rz
yE(Mz IY Y − My IY Z ) zE(My IZZ − Mz IY Z )
= +
E(IY Y IZZ − IY2 Z ) E(IY Y IZZ − IY2 Z )
y(Mz IY Y − My IY Z ) + z(My IZZ − Mz IY Z )
= (3.15)
IY Y IZZ − IY2 Z
Equation 3.16 is the general flexure equation for unsymmetrical bending of beams.

My and Mz are considered positive if they induce tensile stresses on the positive yz quadrant,
otherwise they are considered negative.

If bending is about the principal axes, say yz for which IY Z = 0, then,

yMz IY Y + zMy IZZ
σx =
IY Y IZZ
Mz y My z
= +
IZZ IY Y
which is the equation for symmetrical bending.

At neutral axis, σx = 0, and hence;

y(Mz IY Y − My IY Z ) + z(My IZZ − Mz IY Z )
0 =
IY Y IZZ − IY2 Z
y My IZZ − Mz IY Z
= −
z Mz IY Y − My IY Z
This is the equation of a line passing through the origin of the centroidal xy-plane, and whose
gradient is,
My IZZ − Mz IY Z
tan ϕ = −
Mz IY Y − My IY Z
where ϕ is the angle of inclination of the neutral axis from the horizontal z-axis.

If the gradient is positive, then the orientation of the neutral axis is shown in figure 3.15(a), and
if its negative the orientation of NA is shown in figure 3.15(b).

(a )
(b) N

N
φ
C
C Z Z
φ

N
N
Y Y

Figure 3.15:

Example 3. 6

The angle section shown in figure below is subjected to a positive bending moment of 2kN m about
the Z−axis. Determine the stress at points A, B and C and the position of the neutral axis. All
units are in mm.

First determine the centroid location,

(60 × 10 × 30) + (10 × 90 × 5
z = = 15mm
(60 × 10) + (10 × 90)
(60 × 10 × 5) + (10 × 90 × 55
y = = 35mm
(60 × 10) + (10 × 90)

The moments of inertia about the centroidal axes are,

h 60 × 103 i h 10 × 903 i
IZZ = + 60 × 10(35 − 5)2 + + 10 × 90(55 − 35)2 = 151.25 × 104 mm4
12 12
h 10 × 603 i h 90 × 103 i
2
IY Y = + 60 × 10(30 − 15) + + 10 × 90(15 − 5) = 41.25 × 104 mm4
2
12 12

The product of inertia about the centroidal axes is,

h i
IY Z = (10 × 60 × 5 × 30) + (90 × 10 × 5 × 55) − (60 × 10 + 90 × 10)15 × 35 = −45 × 104 mm4
 10
C

_
z N

Mz
G φ Z
100
N
B
_

10 y

A
Y

60

The general characteristic equation for unsymmetrical bending is,

y(Mz IY Y − My IY Z ) + z(My IZZ − Mz IY Z )
σx =
IY Y IZZ − IY2 Z
But My = o, then,
yMz IY Y − zMz IY Z
σx =
IY Y IZZ − IY2 Z
Mz IY Y = 2000 × 41.25 × 10−8 = 8.25 × 10−4
Mz IY Z = 2000 × −45 × 10−8 = −9 × 10−4
IY Y IZZ − IY2 Z = (41.25 × 10−8 )(151.25 × 10−8 − (−45 × 10−8 )2 = 4.241 × 10−13

Therefore the stress distribution is,

8.25 × 10−4 y + 9 × 10−4 z
σx =
4.241 × 10−13
Stress at point A, whose coordinates are y = 35mm and z = 15mm is,
8.25 × 10−4 (0.035) + 9 × 10−4 (0.015)
σx |A = = 100.6M P a
4.241 × 10−13

Stress at point B, whose coordinates are y = 25mm and z = −45mm is,

8.25 × 10−4 (0.025) + 9 × 10−4 (−0.045)
σx |B = = −47.2M P a
4.241 × 10−13

Stress at point C, whose coordinates are y = −65mm and z = 5mm is,

8.25 × 10−4 (−0.065) + 9 × 10−4 (0.005)
σx |C = = −116.6M P a
4.241 × 10−13
No stress at neutral axis, and therefore,

y −9 × 10−4
= tan ϕ = = −1.09
z 8.25 × 10−4
ϕ = −47.50

The neutral axis is inclined at an angle ϕ = 47.50 from the Z−axis as shown in figure above.

3.4 Tutorial 2
Q1 A Z-section beam is 2m long and is supported as a cantilever with a 1kN point load at
the free end. The direction of the load relative to to the beam section is as shown in
figure 3.16. Calculate the magnitude and position of the maximum tensile and compressive
stresses on the section. All dimensions are in mm. (AN S : 136.9M P a tensile at point P,
and −136.9M P a compressive at point Q)

60

10

10
z
G
100
30 0
1kN

10
P

60
y

Figure 3.16:
Q2 A channel section beam 1m long is built in at one end and subjected to a point load of 1kN
at the free end. If the direction of the load is as shown in figure 3.17, calculate the direction
of the neutral axis, and the maximum values of the tensile and compressive stresses on the
section.(AN S : ϕ = 13.080 , 54.7M P a-tensile and −33.5M P a-compressive).
10 10

1kN
450
60 z

10

100
y

Figure 3.17:

Q3 A cantilever consists of a 60mm × 60mm × 10mm angle with the top face AB horizontal.
It carries a load of 800N at a distance of 1m from the fixed end, the direction of action
of the load passing through the centroid of the section and inclined at 300 to the vertical.
Determine the stresses at corners A, B and C, and also the position of the neutral axis.
(AN S : σA = 136.5M P a, σB = −70.4M P a, σC = −102M P a and ϕ = −4.60 )
800N

30 0

A B

Figure 3.18:
Chapter 4

Application of Strain Energy Methods

to Deflection of Beams and Structures

4.1 Strain Energy

When a piece of material is deformed in simple tension, compression, bending or torsion within
its elastic range, work is done by the applied loading.

On removal of the loading, the material returns to its un deformed state due to the release of the
stored energy. This energy is termed as elastic-strain-energy and has the same magnitude as the
external work done by the load.

It is the release of strain energy in a stretched rubber band that enables a stone to be projected
from a catapult.

Consider a load-extension curve as shown below;

δ
Figure 4.1:
The work done by the load P is the area under the curve, that is,
1
U = Pδ
2
This is the strain energy store on the specimen.

But

P = σA
σLo
δ =
E
Therefore strain energy becomes;
σ2
U = ALo
2E

But ALo = Initial volume of the material, hence,

σ2
U = × V olume
2E
Hence specific strain energy which is the strain energy per unit volume is
σ2
U =
2E

If a piece of material is subjected to pure shear, then the work done by the applied torque is the
area under the T − θ curve. Hence;
1
U = Tθ
2
Gθ T
But from the general torsion formula L
= J
= τr , we have;
τJ
T =
r
τL
θ =
Gr
Therefore;
1 τJ τL τ2 JL
U = × × = × 2
2 r Gr 2G r
d
For a solid circular shaft maximum shear stress occurs at the surface of the shaft, that is r = 2
4
and J = πd32
. Therefore;
2
τmax πd2 L 2
τmax πd2 L
U = × = ×
2G 8 4G 4
2
τmax
= × V olume
4G
Specific strain energy in pure torsion is thus given by;
2
τmax
U =
4G
We can apply this concept of the strain energy stored during elastic deformation to determine
deflection of a beam or structure. Example;
Simple tension:

Figure 4.2:

P
Note that σ = A
, then the total strain energy is;
σ2 P 2L
U = AL =
2E 2EA

Strain energy stored is equal to the work done by load P to produce extension x,
1 P 2L
Px =
2 2EA
PL
x =
EA

Alternatively, if we differentiate the equation of total strain energy with respect to the applied
load P, we get the extension of the specimen.
P 2L
U =
2EA
dU PL
=
dP EA

A Cantilever Beam:

Figure 4.3:

Bending moment produces a direct tension or compression stress on the fibres of the beam. For
direct tension or compression, the total strain energy is given as;
σ2
U = × V olume
2E
Consider a beam under a bending moment about the z-axis;

Figure 4.4:

From the general bending equation, the stress on the element at distance y from the neutral axis
is;
My
σ =
I
Assuming that the beam is of a unit length then the volume of the element is;

dV = dA × 1

Therefore strain energy stored in an element of area dA is,

M 2 y 2 dA
dU =
I 2 2E
Then the total strain energy per unit length is;
Z h/2 2 2
M y dA
U = 2
0 I 2 2E
h/2
M2
Z
= 2 y 2 dA
2EI 2 0
M2 M2
= I =
2EI 2 2EI
Therefore total strain energy in bending is;
M2
Z
U = dx
2EI

For a cantilever;

M = Px

Therefore strain energy is;

L
P 2 x2
Z
U = dx
0 2EI
Z L
P2 P 2 L3
= x2 dx =
2EI 0 6EI
 Figure 4.5:

Therefore deflection is;

dU P L3
δ= =
dP 3EI

Example 4.1

A simply supported beam carries a concentrated load P at a distance a from the left hand support
and a distance b from the other support. Determine the deflection of the beam underneath the
load.

Figure 4.6:

A free body diagram is drawn as shown below; The support reactions are determined by applying

Figure 4.7:

static equilibrium conditions;

X
Fy = 0
RA + RB = P

X
MB = 0
P b = RA L
Pb
RA =
L
Hence;

Pb P (L − b) Pa
RB = P − RA = P − = =
L L L

For section AD,

Figure 4.8:

Pb
Mx = RA x = x
L
Strain energy is;
a
P 2 b2 x 2
Z
UAD = 2
dx
0 2EIL
Z a
P 2 b2 2 P 2 b 2 a3
= x dx =
2EIL2 0 6EIL2

For section DB,

Figure 4.9:

Pa
Mx = RB x = x
L
Strain energy is;
b
P 2 a2 x 2
Z
UDB = 2
dx
0 2EIL
Z b
P 2 a2 2 P 2 a2 b 3
= x dx =
2EIL2 0 6EIL2
The total strain energy on the beam is;

U = UAD + UDB
P 2 b 2 a3 P 2 a2 b 3
= +
6EIL2 6EIL2
P 2 a2 b2 (a + b)
=
6EIL2
The deflection of the beam underneath the load P is;

dU P a2 b2 (a + b)
δ= =
dP 3EIL2

4.2 Castigliano’s Theorem

Castigliano’s theorem states that if the total strain energy of a body or framework is expressed
in terms of the external loads and is partially differentiated with respect to one of the loads, then
the result is the deflection of the point of application of that load and in its direction. This is
expressed mathematically as,
∂U
δn =
∂Pn

where Pn is the load at the point, δn is the deflection of the point towards the direction of the
load, and U is the strain energy.

Advantages of Castigliano’s theorem are,

ˆ It is easy to use.

ˆ It is ideally suited for digital computations.

If the deflection is positive, then it is in the same direction as the applied load.

This theorem is used to find deflection that corresponds to the load acting on a structure. if the
required deflection does not correspond to the direction of applied load, then place an imaginary
(Fictious) load on the structure corresponding to the direction of that deflection. The resulting
expression of strain energy will be in terms of both the actual and fictious load. Then set the
fictious load to zero and obtain deflection due to the actual loads.
Example 4.2

For the structure shown in figure 4.10, determine the vertical and horizontal deflection at point A
due to the load P.

Figure 4.10:

Solution: The vertical deflection is towards the applied load P. Consider figure 4.11

Figure 4.11:
For section AB;

Mθ = P R sin θ
dx = Rdθ
Z π/2 2 Z π/2 2 2 2
Mθ dx P R sin Rdθ
UAB = =
0 2EI 0 2EI
π/2
P 2 R3
Z
= sin2 θdθ
2EI 0
P R3 π/2 2 P R3 π/2 1 1
Z Z
v ∂UAB
δAB = = sin θdθ = ( − cos 2θ)dθ
∂P EI 0 EI 0 2 2
3
PR π
=
4EI

For section BC;

Mx = P R
Z h 2 Z h 2 2
P 2 R2 h
Z
Mx dx P R
UBC = = dx = dx
0 2EI 0 2EI 2EI 0
P R2 h P R2 h
Z
v ∂UBC
δBC = = dx =
∂P EI 0 EI

Therefore the total vertical deflection of point A is;

δAv = δAB
v v
+ δBC
P R3 π P R2 h P R3 π h
= + = ( + )
4EI EI EI 4 R

Since no horizontal load at point A, then to determine the horizontal deflection, we need to apply
a fictious horizontal load at point A as shown in figure 4.12

Figure 4.12:
For section AB;

Mθ = P R sin θ + H(R − R cos θ)

dx = Rdθ
Z π/2 2 Z π/2
Mθ dx [P R sin θ + H(R − R cos θ)]2
UAB = = Rdθ
0 2EI 0 2EI
Z π/2 2 3 2
P R sin θ + 2P HR3 (1 − cos θ) sin θ + H 2 R3 (1 − cos θ)2
= dθ
0 2EI
Z π/2
h ∂UAB 2P R3 (1 − cos θ) sin θ + 2HR3 (1 − cos θ)2
δAB = = dθ
∂H 0 2EI
But H = 0, then,
π/2 π/2
2P R3 (1 − cos θ) sin θ P R3
Z Z
h
δAB = dθ = (sin θ − sin θ cos θ)dθ
0 2EI EI 0
3 Z π/2
PR sin 2θ P R3
= (sin θ − )dθ =
EI 0 2 2EI

For section BC;

Mx = P R + (x + R)H
Z h 2 Z h
Mx dx [P R + (x + R)H]2
UBC = = dx
0 2EI 0 2EI
Z h 2 2
P R + 2P RH(x + R) + H 2 (x + R)2
= dx
0 2EI
Z h
h ∂UBC 2P R(x + R) + 2H(x + R)2
δBC = = dx
∂H 0 2EI
But H = 0, then,
Z h Z h
h P R(x + R) PR
δBC = dx = (x + R)dx
0 EI EI 0
2
PR h
= ( + Rh)
EI 2
Therefore the total horizontal deflection of point A is;

δAh = δAB
h h
+ δBC
P R3 P R h2
= + ( + Rh)
2EI EI 2
Chapter 5

Bending of Curved Beams with Plane

Loading

5.1 Introduction

In theory of bending, the bending equation,

M E σ
= =
I R y

was derived by assuming the beam to be initially straight (besides other fundamental assumptions).
However machine members and structures subjected to bending are not always straight as in the
case of crane hooks, chain links, bridge members, building trusses eg in warehouses.

The problem of curved beams can be classified into two:

1. Initially curved beams where the depth of the cross-section is small in relation to the initial
radius of curvature of the beam ( Rd > 10). Such beams are called slender beams.

2. Beams where the depth of the cross-section is significantly large in relation to the initial
radius of curvature of the beam ( Rd < 10) - Deeply curved beams.

5.2 Stresses and Strains in Curved Beams - Winkler Bach

Analysis

Consider a section of a curved beam OAB as shown in Figure 5.1.

5.2.1 Assumptions

i. Plane sections remain plane during bending.

 (a) Curved beam before bending (b) Curved beam after bending

Figure 5.1: Nomenclature of a curved beam.

ii. Material obeys Hooke’s law (elastic limit is not exceeded)

iii. Radial strain is negligible ⇒ y = y ′

iv. The fibres are free to expand or contract without any constraining effect from the adjacent
fibre.

For a fibre AB on the neutral axis,

AB = A′ B ′ , R1 θ = R2 ϕ
R1
⇒ ϕ= θ (5.1)
R2
Strain on fibre C’D’, a distance y from the neutral axis is given by:

C ′ D′ − CD (R2 + y)ϕ − (R1 + y)θ

εC ′ D ′ = =
CD (R1 + y)θ
R2 ϕ + yϕ − R1 θ − yθ
=
(R1 + y)θ
but R1 θ = R2 ϕ
y(ϕ − θ)
∴ εC ′ D ′ = (5.2)
(R1 + y)θ

Substituting the expression for ϕ, equation 5.1 in equation 5.2, we obtain:

R1


y R 2
θ − θ y(R1 − R2 )
εC ′ D ′ = = (5.3)
(R1 + y)θ R2 (R1 + y)
5.2.2 Case 1 - Slender Beam

y<<R1 ⇒ y can be neglected in relation to R1 , i.e., R1 + y ≈ R1 . The strain is then given by:

y(R1 − R2 ) h 1 1 i
εC ′ D ′ = =y − (5.4)
R2 R1 R2 R1
From equation 5.4, we can deduce that:

1. The strain is directly proportional to the distance y from the neutral axis.

2. The strain distribution and hence stress distribution is linear.

3. The neutral axis and the centroidal axis coincide.1

From Hooke’s Law, σ = Eε

h 1 1 i σ h 1 1 i
∴ σ = Ey − ⇒ =E −
R2 R1 y R2 R1

and from the simple theory of bending,

M σ E
= =
I y R
We can use a modified bending theory to determine the stress distribution, i.e:
M σ h 1 1 i
= =E − (5.5)
I y R2 R1
If the beam was initially straight, R1 → ∞ and equation 5.5 reduces to the simple theory of
bending.

5.2.3 Case 2: Deeply Curved Beams

In this case, the distance y is not negligible when compared to R1 .

The strain distribution is no longer directly proportional to y i.e, the stress and strain distribution
are non-linear.

The neutral axis no longer passes through the centroid.

1
R R h 1 1
i
Transverse forces across the section, in absence of direct forces = 0 ∴ σdA = 0 or y R2 − R1 EdA = 0 ⇒
h iR
E R12 − R11
R
ydA = 0 ⇒ ydA = 0
 (a) slender (linear) (b) deeply curved (non-linear-
hyperbolic)

Figure 5.2: Strain distribution for curved beams.

5.3 Position of the Neutral Axis for a Deeply Curved

Beam

For equilibrium of transverse forces in absence of any applied direct load, the net force = 0.
Z
σdA = 0
Ey(R1 − R2 )
but σ = Eε = (Hooke’s law) (5.6)
R2 (R1 + y)
E(R1 − R2 )
Z Z
y
∴ σdA = dA = 0
R2 R1 + y
R1 −R2
Since E̸=0 and R2
= 0 only when R2 = R1 , i.e, no bending,
Z
y
⇒ dA = 0 (5.7)
R1 + y
Which implies that the neutral axis does not coincide with the centroidal axis.

5.4 Bending Moment on the Cross-section

Consider the cross-section shown in Figure 5.3

The bending moment on the cross-section is given by:

E(R1 − R2 ) y2
Z Z
M = σdA · y = dA
R2 R1 + y
y2 y(y + R1 − R1 ) y[(R1 + y) − R1 ]
Z Z Z
dA = dA = dA
R1 + y R1 + y R1 + y
Z Z
y
= ydA − R1 dA (5.8)
R1 + y
 Figure 5.3: Beam section

y
R
From equation 5.7, R1 +y
dA = 0,
E(R1 − R2 )
Z
∴M = ydA (5.9)
R2

Figure 5.4: Section of a curved beam

From Figure 5.4, y = yc + e, where e is the eccentricity, therefore,
Z Z Z Z
ydA = yc dA + edA, but yc dA = 0
Z
∴ ydA = Ae
E(R1 − R2 )
M = · Ae
R2
M E(R1 − R2 )
rearranging, = (5.10)
Ae R2
From equation 5.6,
E(R1 − R2 σ
= (R1 + y)
R2 y
My
∴σ = (5.11)
Ae(R1 + y)
which is the general bending equation for a deeply curved beam.

5.4.1 Terminologies used with Curved Beams

y is measured from the neutral axis and is considered positive when measured away from the
centre of curvature.

M is considered positive when it tends to decrease R (increase curvature).

From equation 5.11,

M h1 M h2
σi = − ; σo = (5.12)
Aea Aec
which implies that:

ˆ The stress is always greater at the inside radius.

ˆ The neutral axis moves towards the centre of curvature.2

5.4.2 To determine R1

Rectangular section

Figure 5.5: Rectangular section

We need to determine R1 , since Rc can be easily determined. Recall that for no applied load,
R y
R1 +y
dA = 0. Consider an element, a distance r from the center of curvature:

y = r − R1 ⇒ R1 + y = r dy = dr
dA = bdy = bdr
2 y2
R
R1 +y dA
= Ae, for Ae to be positive, it implies that e should be positive, i.e measured from the N.A away
from the centre of curvature.
 r − R1
Z Z Z Z
y R1
dA = dA = dA − dA = 0 (5.13)
R1 + y r r
Z
dA
= A − R1 =0 (5.14)
r
Z
dA
⇒ A = R1 (5.15)
r
A
∴ R1 = R dA (5.16)
r

For a rectangular beam,3 dA = bdr, and A = bd

Z Z c
dA dr c
∴ = b = b ln
r a r a
bh h
⇒ R1 = c =
b ln a ln ac
e = Rc − R1

I-section

Figure 5.6: I - section

The section can be treated as a composite section consisting of three rectangles.

For the top flange, Z Z a+h1

dA1 dr a + h1
= b1 = b1 ln
r a r a
For the web, Z Z a+h1 +h2
dA2 dr a + h1 + h2
= b2 = b2 ln
r a+h1 r a + h1
3 M h2 M h1 h1 a 2ac
To design a curved beam such that maximum stress=minimum stress, Aec = Aea ⇒ h2 = c and R1 = a+c
For the bottom flange,
Z Z c
dA3 dr c
= b3 = b3 ln
r a+h1 +h2 r a + h1 + h2

For the whole section,

Z
dA a + h1 a + h1 + h2 c
= b1 ln + b2 ln + b3 ln
r a a + h1 a + h1 + h2
and,
b1 h1 + b2 h2 + b3 h3
R1 = a+h1
b1 ln a
+ b2 ln a+h 1 +h2
a+h1
+ b3 ln a+h1c +h2
e = Rc − R1

Example 5.4.1. A curved rectangular bar has a mean radius of curvature Rc = 100mm and a
cross-section of width b = 50mm and depth h = 25mm as shown in Figure 5.7. Determine the
largest tensile and compressive stresses given that the bending moment in the bar is M = 500Nm

Figure 5.7: Example

Solution

h
a = Rc − = 100 − 12.5 = 87.5mm
2
h
c = Rc + = 100 + 12.5 = 112.5mm
2
h 25
R1 = c = = 99.477mm
ln a ln 112.5
87.5
e = Rc − R1 = 100 − 99.477 = 0.523mm

Using the bending equation for curved beams,

My
σ=
Ae(R1 + y)
Maximum tensile stress
M h2
σmax =
Ae(R1 + h2 )
h2 = c − R1 = 112.5 − 99.477 = 13.023 A = bh = 25 × 50 = 1250 × 10−6 m2
500 × 13.023 × 10−3
σmax =
1250 × 10−6 × 0.523 × 10−3 × (99.477 + 13.023) × 10−3
= 88.54 Mpa

Maximum compressive stress;

−M h1
σmax =
Ae(R1 − h1 )
h1 = R1 − a = 99.477 − 87.5 = 11.977
−500 × 11.977 × 10−3
σmax =
1250 × 10−6 × 0.523 × 10−3 × (99.477 − 11.977) × 10−3
= −104.688 Mpa

Example 5.4.2. Compare the percentage error4 computed in calculating the maximum tensile and
compressive stresses if the initial curvature of the beam was neglected.

solution

For a straight beam,

M h2 M h 12 6M
σmax,min = ± =± · 3 = 2
I 2 bh bh
6 × 500
= ± = ±96 M P a
50 × 10−3 × (25 × 10−3 )2
96 − 88.54
% error, σt = × 100 = 8.4%
88.54
104.688 − 96
% error, σc = × 100 = 8.3%
104.688

5.5 Combined Direct and Bending Stresses

Most structural members are usually subjected to combined bending and direct stresses5 . In such
cases, the total stress is computed by using the principle of superposition. For a curved beam,
My P
σ= ± , Mxx = P × R
Ae(R1 + y) A

4
Normally in design, % error between predicted value and measured value should not exceed 5%
5
Bending moment may be induced in a number of ways:
ˆ due to externally applied moment
ˆ due to longitudinal eccentric loading
ˆ due to the member buckling
 Figure 5.8: Example of a curved beam subjected to combined loading

Example 5.5.1. A punch press is loaded as shown in Figure 5.9.

(i) Determine the location of

(a) the centroidal axis

(b) the neutral axis

(ii) Determine the limiting value of P if the resultant stress at point A and B are 44N/mm2 and
-31N/mm2 respectively.

(a) Press (b) section AB

Figure 5.9: Punch press

Solution
12.355 × 106
P
A i xi
x̄ = P = = 195.79mm
Ai 63000

The distance from the centre of curvature to the centroidal axis is given by:

Rc = a + x̄ = 195.79 + 300 = 495.79mm

 Table 5.1:
Part Ai xi Ai xi
1 300 × 70 = 21000 35 735000
2 380 × 200 = 76000 70+190=260 19.76×106
3 340 × 100 = −34000 70+170=240 -8.16×106
12.355×106
P
63000

The distance from the centre of curvature to the neutral axis is given by:
A
R= R dA
r

Divide the section into three rectangular areas.

Figure 5.10:

Z Z 370
dA 300dr  370
1 = = 300 ln r 300 = 62.916
r r
Z Z300
750
dA 200dr  750
2 = = 200 ln r 370 = 141.314
r 370 r
Z Z 710
dA 100dr  370
3 = − = − 100 ln r 300 = −62.916
r 370 r
Z
dA
= 62.916 + 141.314 − 65.176 = 139.054
r

A 63000
R= R dA
= = 453.061mm
r
139.054

e = Rc − R = 495.061 − 453.061 = 42.729mm

h1 = R − a = 453.061 − 300 = 153.061mm
h2 = c − R = 750 − 453.061 = 296.939mm

The bending moment is given by

M = P (800 + Rc ) = −P (800 + 495.79) = −1295.79P N mm

and is taken to be -ve since it tends to reduce curvature.

The total direct stress at any point is determined by superposition of the direct stress and bending
stress using the equation:
P My
σmax,min = ±
A Ae(R + y)

At point A,
P M h1
σmax = +
A Aea
P 1295.79 × 153.061P
44 = + = 2.615 × 10−4 P
63000 63000 × 42.729 × 300
⇒ P = 168.28kN

At point B,
P M h2
σmin = −
A Aec
P 1295.79 × 153.061P
44 = − = −1.747 × 10−4 P
63000 63000 × 42.729 × 750
⇒ P = 177.44kN

The limiting value of P is 168.28kN

Chapter 6

Rotating Discs and Cylinders

6.1 Introduction

A rotating disc is a uniformly thin disc which on rotating at constant velocity, is subjected to
stresses induced by centrifugal forces.

Components modeled as uniform discs include:

- gas turbine rotors

- flywheel

- rotating shrink-fit assemblies e.g. shaft-hub assemblies, pulleys

For a thin disc, plane stress is assumed such that we only have

i. σh = Circumferential (hoop stress)

ii. σr = Radial stress

iii. σa = 0 (axial stress)

For a solid disc,

Consider the element of the disc, a distance r from the center. Assuming unit thickness, for
equilibrium of forces in the radial direction, we have:
dθ  dσr 

2σh sin · dr + σr · rdθ − σr + dr r + dr dθ − F = 0 (6.1)
2 dr
dθ dθ
For small angles, sin ≈ (6.2)
2 2
dθ dσr dσr 2
∴ 2σh · dr + σr rdθ − σr rdθ − rdrdθ − σr drdθ − dr dθ − F = 0 (6.3)
2 dr dr
 Figure 6.1: Solid disc

ignoring higher powers of small quantities,

dσr F
∴ σh dr − r dr − σr dr − =0 (6.4)
dr dθ
where, F is the centrifugal force on the element given by:
F = mω 2 r = ρdr · rdθ · rω 2 = ρr2 ω 2 drdθ
where ω is the angular speed of the disc and ρ is the density of the disc

Equation 6.4 becomes,

dσr
r + σr − σh + ρr2 ω 2 = 0 (6.5)
dr
dσr σr − σh
or + + ρrω 2 = 0 (6.6)
dr r

6.2 Circumferential and Radial Strains

New circumference - initial circumference

εh =
initial circumference
2π(r + u) − 2πr u
= =
2πr r
New thickness - Initial thickness
εr =
Initial thickness
du
u + dr dr + dr − u) − dr du
= =
dr dr
From the general stress-strain relationships:
σh σr u
εh = −ν = (6.7)
E E r
σr σh du
εr = −ν = (6.8)
E E dr
 Figure 6.2: Deformed element

Differentiating equation 6.7 with respect to r, we get,

r du
dr
−u 1 h dσh dσr i
= −ν
r2 E dr dr
1 du u 1 h dσh dσr i
− = −ν
r dr r2 E dr dr
du u r h dσh dσr i
= + −ν
dr r E dr dr
1 1 r h dσh dσr i
From equation 6.8, (σr − σh ) = (σh − σr ) + −ν
E E E dr dr
h dσ dσ i
h r
σr (1 + ν) − σh (1 + ν) = r −ν
dr dr
σr − σh dσh dσr
(1 + ν) = −ν
r dr dr
σr − σh 1 h dσh dσr i
or = −ν
r 1 + ν dr dr
From equation 6.6,
dσr 1 h dσh dσr i
+ −ν + ρrω 2 = 0
dr 1 + ν dr dr
dσr dσh dσr
(1 + ν) + −ν + (1 + ν)ρrω 2 = 0
dr dr dr
dσr dσh
(1 + ν − ν) + + (1 + ν)ρrω 2 = 0
dr dr
dσr dσh
+ + (1 + ν)ρrω 2 = 0 (6.9)
dr dr
Integrating equation 6.9,
ρr2 ω 2
σr + σh = −(1 + ν) + 2A (6.10)
2
dσr
from equation 6.6 σh − σr − r = ρrω 2 (6.11)
dr
equation 6.10 − equation 6.11 (6.12)
dσr h 1+ν i
2σr + r = − − 1 ρω 2 r2 + 2A (6.13)
dr 2
dσr 3+ν 2 2
2σr + r = 2A − ρω r (6.14)
dr 2
Rewriting equation 6.14,
1h dσr i 3+ν 2 2
2σr r + r2 = 2A − ρω r
r dr 2
d h 2i 3+ν 2 3
or σr r − 2Ar + ρω r = 0
dr 2
Integrating,

2Ar2 3 + ν 2 r4
σr r2 − + ρω +B = 0 (6.15)
2 2 4
B 3+ν 2 2
or σr = A − 2
− ρω r (6.16)
r 8
From equation 6.10,
1+ν 2 2
σh = 2A − ρω r − σr (6.17)
2
1+ν 2 2 B 3+ν 2 2
= 2A − ρω r − A + 2 + ρω r (6.18)
2 r 8
B 4 + 4ν − 3 − ν 2 2
= A+ 2 − ρω r (6.19)
r 8
B 1 + 3ν 2 2
= A+ 2 − ρω r (6.20)
r 8

6.2.1 Solid Disc with Unloaded Boundaries

(a) Rotating disc Disc (b) Stress distribution

Figure 6.3:

Apply boundary conditions,

When r = 0, both σr and σh → ∞. This is not practical and so we set B = 0

When r = R2 , σr = 0
3+ν 2 2
∴0 = A− ρω R2
8
3+ν 2 2
⇒A = ρω R2
8
3+ν 2 2 3+ν 2 2 3+ν 2 2
σr = ρω R2 − ρω r = ρω (R2 − r2 )
8 8 8
3 + ν 2 2 1 + 3ν 2 2
σh = ρω R2 − ρω r
8 8
at r = R2 , σr = 0 and:
3 + ν − 1 − 3ν 2 2 2 − 2ν 2 2
σh = ρω R2 = ρω R2
8 8
1−ν 2 2
= ρω R2
4

6.2.2 Maximum Speed for Initial Yielding

Tresca’s Criterion (Max shear stress criterion)

Tresca’s Yielding criterion states that:

σmax − σmin σy
= = τmax
2 2
The maximum stress occurs at r = 0 and is:
3+ν 2 2
σr = σh = ρω R2
8
Minimum stress σz = 0

∴ σmax = σy
3+ν 2 2
ρωy R2 = σy
8 s
1 8σy
ωy =
R2 (3 + ν)ρ

von-Mises Criterion (Shear strain energy criterion)

(σ1 − σ2 )2 + (σ2 − σ3 )2 + (σ3 − σ1 )2 = 2σy2
3+ν 2 2
σ1 = σ2 = ρω R2
8
σ3 = 0 ⇒ σ1 = σy
s
1 8σy
ωy =
R2 (3 + ν)ρ

which is the same value as for the Trescas criterion

6.2.3 Increase in Radius

1 u
εh = (σh − νσr ) = ∵ r = R2
E R2
R2
∴u = (σh − νσr but atr = R2 , σr = 0
E
R2 R2 1 − ν 2 2
⇒u = σh = ρω R2
E E 4
ρω 2 R22 1 − ν
u =
4 E

6.2.4 Change in Thickness

Let the original thickness be t. The change in thickness in the z-direction is ∆t

σz ν
εz = − (σr + σh ) but σz = 0
E E
ν ∆t
∴ εz = − (σr + σh ) =
E t
at r = 0, σr = σh
∆t ν h 3 + ν 2 2i
= − 2 ρω R2
t E 8
ν h 3 + ν 2 2i
∆t = − 2 ρω R2 t
E 4
when r = R2 , σr = 0,
∆t νh 1 − ν 2 2i ν h 1 − ν 2 2i
= − 0+ ρω R2 = − ρω R2
t E 4 E 4
ν h 1 − ν 2 2i
∆t = − ρω R2 t
E 4
 Figure 6.4: Change in thickness

6.3 Disc with Central Hole and Unloaded Boundaries

Boundary conditions:
at r = R1 , σr = 0 and at r = R2 , σr = 0
B 3+ν 2 2
0 = A− 2
− ρω R2
R2 8
B 3+ν 2 2
0 = A− 2 − ρω R1
R1 8
3+ν 2 2 2 3+ν 2 2
⇒B= ρω R1 R2 , A= ρω (R1 + R22 )
8 8
3 + ν 2h 2 2R12 R22 i
∴ σr = ρω R1 + R22 − − r 2
8 r2
3+ν 2 2 h
2 2R12 R22 1 + 3ν 2 i
σh = ρω R1 + R2 − − r
8 r2 3+ν
Maximum σh occurs at r = R1
3 + ν 2h 2 1 + 3ν 2 i
σhmax = ρω R1 + R22 − r
8 3+ν
√
Maximum σr occurs at r = R1 R2 . Stresses are either zero or tensile

6.4 Disc Shrunk onto a Shaft

Applied in shaft-hub assemblies.

Use shrink-fit (interference fit) between components instead of mechanical fittings e.g. keys and
keyways, rivetting, bolts etc.
Same principle used for compound cylinders applies.

At the mating surface, we have radial compressive stresses set up by the shrink fit and on rotation,
radial tensile stresses are set-up. It is therefore necessary to ensure that the shrink-fit stress is
always greater than the rotational stresses to avoid the disc running free from the shaft.

Figure 6.5: Free body diagrams of shrink-fit components

Let

δ = shrinkage between the hub and the shaft.

= difference in radius between the mating surfaces
= uhub − ushaf t

when the assembly is stationary, ω = 0 and the contact pressure is Pc . Therefore,

B
σr = A − 2
r
B
σh = A + 2
r
We can determine the contact pressure by using a similar analysis to that of compound cylinders.
Applying the boundary conditions:
Considering the hub,
at r = R1 , σr = −Pc
at r = R2 = σr = 0
B
∴ −Pc = A − 2 (6.21)
R1
B
0 = A− 2 (6.22)
R2
subtracting equation 6.22 form 6.21,we obtain:
h 1 1 i R2 R2
−Pc = −B 2 − 2 ⇒ B = 2 1 2 2
R1 R2 R2 − R1
Eliminating B,
−Pc R12 = AR12 − B
subtracting 0 = AR22 − B
Pc R 2
−Pc R12 = −A(R22 − R12 ) ⇒ A = 2 1 2
R2 − R1
The radial and circumferential strains become:
Pc R12 h R22 i
σr = 2 1−
R2 − R12 r2
Pc R 2 h R22 i
σh = 2 1 2 1 +
R2 − R1 r2
Considering the shaft,

σr and σh → ∞ at r = 0 ⇒ B = 0
∴ σr = σh = = A = −Pc everywhere on the shaft

For the hub,

uh 1 1 h R12 + R22 i
εh = = (σh − νσr ) = Pc 2 + νP c
R1 E E R2 − R12
R1 Pc h R12 + R22 i
uh = Pc 2 +ν
E R2 − R12
For the shaft:
us 1 1
εh = = (σh − νσr ) = (−Pc + νPc )
R1 E E
Pc R1
us = − (1 − ν)
E

δ = uh − us
R1 Pc h R12 + R22 i PR
c 1
= Pc 2 + ν + (1 − ν)
E R2 − R12 E
R1 Pc h R12 + R22 i
= Pc 2 + ν + 1 − ν
E R2 − R12
R1 Pc h R12 + R22 + R22 − R12 i
= Pc
E R22 − R12
R1 Pc h 2R22 i
= Pc 2
E R2 − R12
Eδ(R22 − R12 )
⇒ Pc =
2R1 R22

Loosening speed

When rotating freely, σr = 0 at the interface.

ˆ Hub can be considered as a disc with a central hole and unloaded boundaries
3 + ν 2h 2 2R12 R22 1 + 3ν 2 i
σh = ρω R1 + R22 − − r
8 r2 3+ν
3 + ν 2h 2 1 + 3ν 2 i
at r = R1 σh = ρω R1 + R22 − r
8 3+ν
R1 R1  3 + ν 2 h 2 1 + 3ν 2 i
uh = (σh − νσr ) = ρω R1 + R22 − r
E E 8 3+ν
 ˆ Shaft can be considered as a solid disc with unloaded boundaries.
At the loosening speed,
δ = uh − us

Example 6.4.1. A solid steel disc of small constant thickness has a steel ring of outer diameter
610 mm and same thickness shrunk onto it. The assembly has an interface diameter of 457mm. If
the interface pressure is reduced to zero at a rotational speed of 3000rpm, calculate the difference
in diameters of the mating surfaces of the disc and the ring before assembly and the interface
pressure. Take ν = 0.29 and E =207 GN/m2
6.5 Disc with Loaded Outer Boundary

Rotor discs normally have a large number of blades mounted on the outer boundary layer. These
will themselves have a centrifugal force component acting at the periphery of the disc. Given the
mass of each blade, its effective centre of mass and the number of blades, we can compute the
total force acting on the outer surface. Dividing this force by the area of the outer surface gives
us the required value of σr to be used as a boundary condition.

Fc = N × ω 2 × re

and,
Fc
at r = R2 , σr =
2πR2 t
Example 6.5.1. A steel rotor disc of uniform thickness 50mm has an outer rim of diameter
750mm and a central hole of diameter 150mm. There are 200 blades each of mass 0.22kg at
an effective radius of 430mm pitched evenly around the periphery of the rotor. Determine the
rotational speed at which yielding first occurs according to the maximum shear stress criterion.
Yield stress in simple tension for the steel is 700MN/m2 , ν = 0.29, ρ = 7300kg/m3 and E =
207GN/m2

Solution

centrifugal force caused by each blade, = mω 2 re

= 0.22ω 2 × 0.43 = 0.0946ω 2
Total centrifugal force F = 200 × 0.0946ω 2 = 18.92ω 2

Radial stress on the outer surface,

F 18.92ω 2
σr |r=R2 = = = 160.598ω 2
2πR2 t 2π × 0.375 × 0.05
General expressions for radial and hoop stress are:
B 3+ν 2 2
σr = A −2
− ρω r
r 8
B 1 + 3ν 2 2
σh = A − 2 + ρω r
r 8
Boundary conditions:

at r = R1 = 0.075, σr = 0
at r = R1 = 0.375, σr = 160.598ω 2
Applying Boundary conditions,

A − 177.78B = 16.89ω 2 (6.23)

A − 7.11B = 582.77ω 2 (6.24)

Equation 6.24 - 6.23 gives,

170.67B = 565.88ω 2
⇒ B = 3.316ω 2
A = 606.35ω 2

and the general equations for the radial and hoop stress become:
3.316ω 2
σr = 606.35ω 2 − 2
− 3002.125ω 2 r2
r
2
3.316ω
σh = 606.35ω 2 + + 1706.375ω 2 r2
r2
dσr
(σr )max occurs at dr
=0

dσr r2 (0) − 2r × 3.316ω 2

= − − 2 × 3002.125ω 2 r = 0
dr (r2 )2
6.632ω 2
3
= 6004.25ω 2 r
r
4 6.632
r = ⇒ r = 0.182
6004.25
3.316ω 2
(σr )max = 606.35ω 2 − − 3002.125ω 2 × 0.1822 = 406.78ω 2
0.1822
The maximum circumferential stress occurs at r = R1
3.316ω 2
σh = 606.35ω 2 + + 1706.375ω 2 × 0.0752 = 1186.26ω 2
0.0752
at r = R1 , σr = 0

σmax = 1186.26ω 2
σmin = 0
σmax − σmin σy
= ⇒ σmax=σy
2 2
1186.26ω 2 = 700 × 106 ⇒ ω = 768.173ω 2 rad/s
60ω
N = = 7335.5 rpm
2π

6.6 Disc of Uniform Strength

A disc of uniform strength is one in which the values of radial and circumferential stresses are
equal in magnitude for all values of the radius r. This means that the disc of uniform strength
must have varying thickness.
 Figure 6.6: Disc with varying thickness

In practice, components such as rotor of a steam turbine which have constant strength throughout
are designed by varying their thickness. centrifugal force on the element is given by:

F = ρ · dr · t · rdθω 2 r = ρω 2 tdrdθr2 (6.25)

Considering equilibrium of forces in the radial direction,

dθ
2σ tdr + σrtdθ − σ(r + dr)dθ(t + dt) − F = 0
2
σtdrdθ + σrtdθ − σrtdθ − σtdrdθ − σrdtdθ − σdrdtdθ − ρω 2 tdrdθr2 = 0

Ignoring products of small quantities,

σrdt + ρtω 2 r2 dr = 0
dt ρω 2 r
= − dr
t σ
ρω 2 r2
integrating, ln t = − +A
2σ
Applying the boundary conditions,

at r = 0, t = to
∴ ln to = A
ρω 2 r2
ln t = − + ln to
2σ
t ρω 2 r2
or ln = −
to 2σ
ρω 2 r 2
t = e− 2σ