MoS Lecture Notes

Chapter 1 - Equations of Statics & Free Body Diagrams
1.1 BASIC DEFINITIONS

Equilibrium: A state of no acceleration, in either translational or rotational senses.
Scalar: A quantity which has only magnitude. Examples include mass and area.
Vector: A quantity which has both magnitude and direction, and satisfies the
parallelogram rule of vector addition. Examples include displacement and
force.
Force: The interaction between bodies which gives rise to an acceleration or to
the deformation of the body.
Moment: The product of the magnitude of a force and the perpendicular distance of
its line of action from a particular point. (Also a vector.)
Couple It consists of two forces equal in magnitude but opposite in direction
whose line of action are parallel but no collinear.
1.2 PHYSICAL FUNDAMENTALS

Equilibrium where everything in a body (or part of it) is in equilibrium.
j's
Compatibility where all the deformations in the body are smooth. We are not interested
in what happens to a body in the event of a catastrophic occurrence, such
as an explosion, or two objects colliding.
Constitutive laws where the forces exerted to a body are related to the way the body
ra
deforms. For example if we apply a force to a body, it will deform always
by the same amount if the force is the same.
Energy The work done in deforming the body is retained by that structure as
vi
internal strain energy which will be released when the applied force is
removed. The classical example is the bow and arrow. You apply a force
to the bow and deflect it by pulling on its string. But when you let go of
ith
the string the energy stored in the bow is enough to propel the arrow for
hundreds of meters.
The types of problems that we will be dealing with are all concerned with the bodies all
being in static equilibrium. In the context of this course the state of Equilibrium is when a
Pr
body, although acted upon by many forces or moments experiences no acceleration.
1.3 EQUATIONS OF STATICS (SI & 4th Ed p.4-6; 3rd Ed p.3-5)

For a 3-D body at rest the coordinate system used is the x-y-z Cartesian system, in which the
definition of positive moments is given by the right hand rule that states that moments are
positive is their sense is counterclockwise as shown in Fig. 1.1.
Fy
Fy
y
My y
Mz
Mx x
z Fx
Fz Fx Mz x
Fig. 1.1 3-D positive axis system (Right-Hand) Fig. 1.2 2-D positive axis system
Lecture Notes of Mechanics of Solids, Chapter 1 1

∑ Fx = 0

Force equilibrium equation is given as: ∑ F y = 0 (1.1)

∑ Fz = 0
∑ M x = 0

and Moment equilibrium as ∑ M y = 0 (1.2)

∑ M z = 0
The 2D x & y -axis system looks like Fig 1.2: For a two dimensional body in the xy-axis
system, the 3D equilibrium Eqs. (1.1) and (1.2) simplify to:
∑ Fx = 0

∑ Fy = 0 (1.3)

∑ M z = 0
NOTE : Not only a body/structure, but every part of a body/structure must be in equilibrium.
j's
1.4 FREE BODY DIAGRAMS (SI &4th Ed p.7-15; 3rd Ed p.6-15)
They are a complete diagram or simplified line sketch of the structure (or body), showing the
ra
position, direction and point of application of all externally applied forces (e.g. P) acting on
the structure, including ground reaction forces (e.g. RDX, RDY and RBX) and/or moments. Figs.
1.3 and 1.4 depict the extraction process from an original structure to a FBD.
vi
α P
ith
A
Pr
B
5m
RBX
17
30°
m
5m
32
RDX D
RDY
Fig. 1.3 Original diagram of Fig. 1.4 Free body diagram (FBD) of the structure
structure with applied loads
1.5 TYPES OF SUPPORTS (4th Ed p.5; 3rd Ed p.4)

There are 6 commonly used types of supports which prevent a structure, or part of it, from
accelerating when acted upon by external forces. In general, supports are there to keep a
structure in equilibrium. These six types of the supports are given in Table 1.1:

Table 1.1 Boundary conditions
Original structure Free Body Diagram Unknowns
Cable Support T
α α
A A
Roller support
RAY
RAY
A A
RAX
RAX
External Pin RAY
RAY
A A
RAX
RAX
Pin Joint RAY
RAY
Fixed/
A
j's RAX
A
RAX
ra
Cantilever MAZ RAY
Support RAY MAZ
A
A
vi
Smooth
RA
Support α α
RA
ith
Example 1.1: Solve for the ground reactions of the structure shown in Fig. 1.3.
Step 1: Free Body Diagram
From Table 1.1: B  External pin, two unknown reactions RDX and RDY
Pr
A  Pin joint, there should be two unknown reactions: RBX and RBY. But
there are only two pins connecting bar BC with others. That implies bar BC is a “two-force
member”, i.e. it is acted upon by two equal but opposite forces directed along the bar axis. In
this case, we hence have vertical component RBY = 0. The free body diagram is given as Fig 1.4.
Step 2 Equilibrium Totally, there are three unknowns RDX, RDY and RBX. They can be solved
based on the three equilibrium equations as (1.3).
→ ∑ Fx = 0 = RBX + RDX − P sin α = 0
+
 (1)
+ ↑ ∑ F y = 0 = RDY − P cos α = 0 R DY = P cos α (2)
+ ∑ M D = 0 = − RBX × DC = 0 ∴ RBX = 0 (3)
Substituting (3) into (1) gives ∴ R DX = P sin α
Example 1.2: The beam of Figure 1.5 is subjected to a vertical force, a horizontal force and a
heavy box of 30kN as shown in Fig. 1.5. Determine the reactions at the supports A and B.
Step 1: Free Body Diagram
From Table 1.1: A  External pin, unknown reactions RAX and RAY
B  Roller support, unknown reaction: RBY

The held box can be approximately treated as a “uniform distributed load” (UDL) w on the
beam, which will be further represented by a concentrated resultant force F at the centre of
the box in the equilibrium as shown.
15kN
30kN
A 45kN
C D E B
4m 5m 3m 1.5m
Fig. 1.5 Original structure of Example 1.2
F=w×3=30kN (resultant force)

w=10kN/m
15kN
A 45kN
RAX C D E B
1.5m
RAY RBY
4m 5m 3m 1.5m
Step 2: Equilibrium
→+
∑ Fx = 0 = R AX + 45 = 0
j's ∴ R AX = −45kN (1)
ra
+ ↑ ∑ F y = 0 = R AY − 15 − 3 × 10 + R BY = 0 ∴ R AY + R BY = 45 (2)
+ ∑ M A = 0 = −15 × 4 − 3 × 10 × (4 + 5 + 1.5) + RBY × 13.5 = 0 ∴ R BY = 27.78kN (3)
vi
Substituting (3) into (2) gives ∴ R AY = 17.22kN
ith
Example 1.3: Determine the ground reactions at A and B of structure as shown.

20kN
20kN
D
Pr
3m
A B
A B
C RAX
45º C
30kN
RB
3m 3m RAY
Step 1 FBD, The reaction force at roller support B should be perpendicular to the surface.
Step 2 Equilibrium:
→ ∑ Fx = 0 = R AX − R B cos 45o = 0
+
 (1)
+ ↑ ∑ Fy = 0 = R AY + RB sin 45o − 20 = 0 (2)
+ ∑ M A = 0 = −20 × 3 + RB sin 45o × 6 = 0 ∴ R B = 10 / sin 45 o kN (3)
Substituting (3) into (1) gives R AX = R B cos 45 o = 10kN
Substituting (3) into (2) gives ∴ R AY = 10kN

Chapter 2 Stress and Strain
2.0 INTRODUCTION
In the first Chapter we discuss the equations of statics, and how to determine the ground
reaction for any structure. The method can also be used to determine the internal loads carried
by the members or parts of a body. We now need to define how these internal loads are
distributed and carried by the material and the deformation they create.
2.1 STRESS (SI&4th p. 22-23, 3rd p. 22-23)

Consider an element of continuous (no voids) and cohesive (no cracks, breaks and defects)
material subjected to a number of externally applied loads as shown in Fig. 2.1a). It is
supposed that the member is in equilibrium.
F4 Free Body Diagram
n
F3 F5 ∆F
∆Fn
∆F t t
j's
Cross section: A ∆A
F2 F2
ra
F1 F1
vi
(a) (b)
Fig. 2.1 External and internal forces in a structural member
ith
If we now cut this body, the applied forces can be thought of as being distributed over the cut
area A as in Fig. 2.1b). Now if we look at infinitesimal regions ∆A, we assume the resultant
force in this infinitesimal area is ∆F. In fact, ∆F is also a distributed force. When ∆A is
extremely small, we can say that the distributed force ∆F is uniform. In other words, if we
Pr
look at the whole sectioned area, we can say that the entire area A is subject to an infinite
number of forces, where each one (of magnitude ∆F) acts over a small area of size ∆A. Now,
we can define stress.
Definition: Stress is the intensity of the internal force on a specific plane
passing through a point.
Mathematically, stress can be expressed as
∆F
σ = lim (2.1)
∆A→ 0 ∆A
Dividing the magnitude of internal force ∆F by the acting area ∆A, we obtain the stress. If we
let ∆A approach zero, we obtain the stress at a point. In general, the stress could vary in the
body, which depends on the position that we are concerning. The stress is one of most
important concepts that we introduced in mechanics of solids.
Normal and Shear Stress

As we known, force is a vector that has both magnitude and direction. But in the stress
definition, we only consider the magnitude of the force. Obviously, this may easily confuse
us. Let’s still take patch ∆A as an example. As we can see, force ∆F is not perpendicular to
the sectioned infinitesimal area ∆A. If we only take the magnitude of the force into account,

apparently, the stress may not reflect the real mechanical status at this point. In other words,
we need to consider both magnitude and direction of the force.
Now let’s resolve the force ∆F in normal and tangential direction of the acting area as Fig.
2.1b). The intensity of the force or force per unit area acting normally to section A is called
Normal Stress, σ (sigma), and it is expressed as:
∆Fn
σ = lim (2.2)
∆A→0 ∆A
If this stress “pulls” on the area it is referred as Tensile Stress and defined as Positive. If it
“pushes” on the area it is called Compressive Stress and defined as Negative.
The intensity or force per unit area acting tangentially to A is called Shear Stress, τ (tau), and
it is expressed as:
∆Ft
τ = lim (2.3)
∆A→0 ∆A
Average Normal Stress (SI&4th p. 24-31, 3rd p. 27-34)

To begin, we only look at beams that carry tensile or compressive loads and which are long
j's
and slender. Such beams can then be assumed to carry a constant stress, and Eq. (2.2) can be
simplified to:
F
σ = (2.4)
ra
A
We call this either Average Normal Stress or Uniform Uniaxial Stress.
Units of Stress
vi
The units in the SI system is the Newton per square meter or Pascal, i.e. : Pa = N/m2.
In engineering, Pa seems too small, so we usually use:
ith
Kilo Pascal KPa (=Pa×103) e.g. 20,000Pa=20kPa

Mega Pascal MPa (=Pa×106) e.g. 20,000,000Pa=20MPa
9
Giga Pascal GPa (=Pa×10 ) e.g. 20,000,000,000Pa=20GPa
Pr
Example 2.1: An 80 kg lamp is supported by a single electrical

F
copper cable of diameter d = 3.15 mm. What is the stress carried
by the cable.
To determine the stress in the wire/cable as Eq. (2.4), we need
the cross sectional area A of the cable and the applied internal Section
a-a
force F: A
a a
πd 2 π × 0.00315 2
A= = = 7.793 × 10 −6 m 2
4 4 d
F = mg = 80 × 9.8 = 784 N
F 784 F
so σ= = = 100.6MPa
A 7.793 × 10 −6 80kg
FBD
Allowable Stress (SI&4th p. 48-49, 3rd p. 51-52)
From Example 2.1, we may concern whether or not 80kg would be too heavy, or say
100.6MPa stress would be too high for the wire/cable, from the safety point of view. Indeed,
stress is one of most important indicators of structural strength. When the stress (intensity of
force) of an element exceeds some level, the structure will fail. For convenience, we usually
adopt allowable force or allowable stress to measure the threshold of safety in engineering.

Moreover, there are following several reasons that we must take into account in engineering:
• The load for design may be different from the actual load.
• Size of structural member may not be very precise due to manufacturing and assembly.
• Various defects in material due to manufacturing processing.
One simple method to consider such uncertainties is to use a number called the Factor of
Safety, F.S., which is a ratio of failure load Ffail (found from experimental testing) divided by
the allowable one Fallow
F fail
F .S . = (2.5)
Fallow
If the applied load is linearly related to the stress developed in the member, as in the case of
using σ = F / A , then we can define the factor of safety as a ratio of the failure stress σfail to
the allowable stress σallow
σ fail
F .S . = (2.6)
σ allow
Usually, the factor of safety is chosen to be greater than 1 in order to avoid the potential
j's
failure. This is dependent on the specific design case. For nuclear power plant, the factor of
safety for some of its components may be as high as 3. For an aircraft design, the higher the
F.S. (safer), the heavier the structure, therefore the higher in the operational cost. So we need
ra
to balance the safety and cost.
The value of F.S. can be found in design codes and engineering handbook. More often, we
vi
use Eq. (2.6) to compute the allowable stress:
σ fail
σ allow = (2.7)
F .S .
ith
Example 2.2: In Example 2.1, if the maximum allowable stress for copper is
σCu,allow=50MPa, please determine the minimum size of the wire/cable from the material
strength point of view.
Pr
F mg
Mathematically, σ = = 2 ≤ σ Cu ,allow
A πd 4
4mg
Therefore: d≥ = 4.469 × 10 −3 = 4.469mm
πσ Cu ,allow
Obviously, the lower the allowable stress, the bigger the cable size. Stress is an indication of
structural strength and elemental size.
In engineering, there are two significant problems associated with stress as follows.
Problem (1) Stress Analysis: for a specific structure, we can determine the stress level.
With the stress level, we then justify the safety and reliability of a structural
member, i.e. known size A and load F, to determine stress level: σ = F A
Problem (2) Engineering Design: Inversely, we can design a structural member based on the
allowable stress so that it can satisfy the safety requirements, i.e. known material’s
allowable stress σ allow and load F, to design the element size: A ≥ F σ allow

2.2 DEFORMATION (SI&4th p. 67-68, 3rd p. 70)
Whenever a force is applied to a body, its shape and size will change. These changes are
referred as deformations. These deformations can be thought of being either positive
(elongation) or negative (contraction) in sign as shown in Fig. 2.2.
F F F F
(+) (−)
Deformation
Original length (contraction)
Deformation Original length
(elongation)
Fig. 2.2 Deformation due to applied axial forces
It is however very hard to make a relative comparison between bodies or structures of different
size and length as their individual deformations will be different. This requires the development
of the concept of Strain, which relates the body’s deformation to its initial length.
2.3 STRAIN (SI p.68-69; &4th p. 67-69; 3rd p.71-72)
j's
Normal Strain
The elongation (+ve) or contraction (−ve) of a body per unit length is termed Strain.
F1
ra
n B’
vi
A ∆S B
A’
∆S’ F3
ith
F2
(a) Before deformed (b) After deformed
Fig. 2.3 Generalized deformation due to applied forces
Pr
Let’s take the arbitrarily shaped body in Fig. 2.3 as an example. Consider the infinitesimal
line segment AB that is contained within the undeformed body as shown in Fig. 2.3(a). The
line AB lies along the n-axis and has an original length of ∆S. After deformation, points A
and B are displaced to A’ and B’ and in general the line becomes a curve having a length ∆S’
The change in length of the line is therefore ∆S-∆S’. We consequently define the generalized
strain mathematically as
∆S' −∆S
ε = lim (B→A along n) (2.8)
B→ A ∆S
Average Normal Strain

If the stress in the body is everywhere constant, in other words, the deformation is uniform in
the material (e.g. uniform uniaxial tension or compression) as shown in Fig. 2.2, the strain can
be computed by
LDeformed − LOriginal ∆L
ε= = (2.9)
LOriginal L
i.e. the change in length of the body over its original length,

Unit of Strain
From Eqs. (2.8) and (2.9), we can notice that the normal strain is a dimensionless quantity
since it is a ratio of two lengths. Although this is the case, it is common in practice to state it
in terms of a ratio of length units. i.e. meters per meter (m/m)
Usually, for most engineering applications ε is very small, so measurements of strain are in
micrometers per meter (µm/m) or (µ/m).
Sometimes for experiment work, strain is expressed as a percent, e.g. 0.001m/m = 0.1%.
A normal strain of 480µm for a one-meter length is said:
ε= 480×10-6 = 480(µm/m) = 0.0480% = 480µ (micros) = 480µs (micro strain)
Example 2.3: In Example 2.1, if it is measured that the cable was elongated by 1.35 mm due
to the weight of the light, what would its strain be?
0.00135
ε= = 900 × 10 −6 = 900 µs
1.5
2.4 STRESS-STRAIN RELATIONSHIP, HOOKE'S LAW (SI&4th p.83-93; 3rd
j's
p.85-95)
Material Test and Stress-Strain Diagram
The material strength depends on its ability to sustain a load without undue deformation or
ra
failure. The property is inherent in the material itself and must be determined by experiment.
One of the most important tests to perform in this regard is the tension or compression. To do
so, a bunch of standard specimen is made. The test is performed in universal test machine.
vi
Shown in Fig. 2.4 is the specimen and test result of Stress-Strain Diagram.
F σ Elastic Yielding Hardening Necking
ith
Ultimate
stress σu
Standard
Specimen Yield stress σY
Fracture
Pr
Proportional σpl stress σf

limit
Plastic behavior ε
F
Elastic
behavior
Fig. 2.4 Material test and Stress-Strain Diagram
The Stress-Strain diagram consists of 4 stages during the whole process, elastic, yielding,
hardening and necking stages respectively. From yielding stage, some permanent plastic
deformation occurs. About 90% of engineering problems only concern the elastic deformation
in structural members and mechanical components. Only 10% of engineering work concerns
plastic and other nonlinear stage (e.g. metal forming). In this subject, we are only involved in
the linear elastic region, in which the relationship between the strain and stress is linear.
Hooke’s Law
The stress-strain linear relationship was discovered by Robert Hook in 1676 and is known as
Hooke's law. It is mathematically represented by Eq. (2.10),

σ = Eε (2.10)
where E is terms as the Modulus of Elasticity or Young's Modulus with units of N/m2 or Pa.
For most of engineering metal material, GPa is used, e.g. mild steel is about 200GPa~210GPa.
2.5 POISSON'S RATIO (SI&4th Ed p104-105; 3rd Ed p107-108)

Definition
When a deformable body is stretched by a tensile force, not only does it elongate but it also
contract laterally, i.e. it would contract in other two dimensions as shown in Fig 2.5(a).
Likewise, a compressive force acting on a deformable body cause it to contract in the
direction of force and yet its sides expand laterally as in Fig. 2.5(b).
When the load P is applied to the bar it changes the bar’s length by δ and its radius by δr. The
strain in axial direction and in lateral/radial direction are respectively
δ δr
ε axial = ε lateral =
L r
In early 1800s, French scientist Poisson realized that within elastic range the ratio of these
two strains is a constant. We called the constant as Poisson’s ratio by v (Nu)
Lateral Strain ε
v=− = − Lateral (2.11)
ε Axial
j's
Axial Strain
F Original shape F Original shape

ra
Defrormed
Defrormed shape shape
vi
L r δr
L+ δr L
δ L– r
δ
F
ith
F
(a) (b)
Fig. 2.5 Relationship of the axial strain with the lateral strain
Pr
The negative sign is used here since longitudinal elongation (positive strain) cause lateral
contraction (negative strain), vice versa. So Poisson’s ration is positive, i.e. v≥0.
Remarks
• The lateral strain is caused only by axial force. No force or stress acts in lateral direction;
• Lateral strain is the same in all lateral direction;
• Usually 0 ≤ v ≤ 0.5 . For most linearly elastic material v=0.3;
• Poisson’s ratio is a constant.
Strain in Lateral Direction

For bars subjected to a tensile stress σx, the strains in the y and z planes are:

 σx
ε x = E

 vσ x
ε y = −vε x = − (2.12)
 E
 vσ x
ε z = −vε x = − E

2.6 THERMAL STRAIN (SI&4th Ed p. 148-152; 3rd Ed p. 152-156)
Thermal Deformation
When the temperature of a body is changed, its overall size will also change. In other words,
temperature change may cause the dimension or shape change in the material. More specially,
if the temperature increases, generally a material expands. Whereas if the temperature
decreases, the material will contract. It is supposed that this is a common sense for anyone.
y
L δT
σx
j's
σx
TRef Tref+ ∆T
TRef
x
ra
z Tref+ ∆T
(a) Thermal deformation (b) Thermal and mechanical deformation
Fig. 2.6 Thermal and mechanical deformation
vi
For the majority of engineering materials this relationship is linear. If we assume that the
ith
material is homogeneous and isotropic, from experiment, we can find a linear relation
between thermal deformation and temperature change as:
δT = α ⋅ ∆T ⋅ L (2.13)
where : α : o
Coefficient of thermal expansion , units are strain per C
Pr
∆T : algebraic change in temperature ( oC ) (increase +; decrease −)

δT : algebraic change in length (“+” = elongation; “−” = contraction)
Thermal Strain
δT
ε Thermal = = α ⋅ ∆T (2.14)
L
Coupled Strain Status

If we consider both mechanical strain εσ and thermal strain εT in the structure as shown in Fig.
2.6(b), by referring to Eq. (2.12), the total strains in all directions would be computed as:
 σx
ε x = ε T + ε σx = α∆T +
E

 σx
ε y = ε T + ε σy = α∆T − (2.15)
 E
 σx
ε z = ε T + ε σz = α∆T − E


2.7 ELASTIC DEFORMATION OF AXIALLY LOADED MEMBER
(SI&4th p.120-127;3rd p.122-129)
Now we are going to find the elastic deformation of a member subjected to axial loads. Let’s
consider a generalized bar shown in Fig. 2.7, which has a gradually varying cross-sectional
are along its length L. For a more general case, the bar is subjected to concentrated loads at its
right end and also a variable external load distributed along its length (such as a distributed
load could be for example, to represent the weight of a vertical bar or friction forces acting on
bar surface). Here we wish to find the relative displacement δ of one end with respect to the
other.
L FBD dδ (elongation of dx) View a-a
a
P P(x)
P(x)+dP
δ
a A(x)
x dx
j's
dx (original length)
Fig. 2.7 Thermal and mechanical deformation

ra
We pick a differential element of length dx and cross-sectional area A(x). FBD can be drawn
as middle of Fig. 2.7. Assume that resultant internal axial force is represented as P(x). The
load P(x) will deform the element into the shape indicated by the dashed outline.
vi
P ( x)
The average stress in the cross-sectional area would be σ ( x ) =
A( x)
ith
dδ
The average strain in the cross-sectional area would be ε ( x ) =
dx
Provided these quantities do not exceed the proportional limit, we can relate them using
Hook’s law, i.e.
Pr
σ = Eε
Therefore
P ( x)  dδ 
= E ( x ) 
A( x)  dx 
Re-organize the equation, we have
P ( x)
dδ = dx
A( x) E ( x )
For the entire length L of the bar, we must integrate this expression to find the required end
displacement
L
P( x)
δ =∫ dx (2.16)
0
A ( x ) E ( x )
Where: δ = displacement between two points
L = distance between the points
P(x) = Internal axial force distribution
A(x) = Cross-sectional area
E(x) = Young’s modulus

Constant Load and Cross-Sectional Area
In many engineering cases, the structural members experience a constant load and have a
constant cross-sectional area and made of one homogenous material, i.e.
P(x) = P = constant (no axially distributed load)
A(x) = A = constant (uniform area)
E(x) = E = constant (homogeneous material)
From Eq. (2.16), we have
PL
δ = (2.17)
EA
Multi-Segment Bar
If the bar is subjected to several different axial forces or cross-sectional areas or Young’s
moduli, the above equation can be used for each segment. The total displacement can be
computed from algebraic addition as
PL
δ =∑ i i (2.18)
i Ai E i
Example 2.4: The composite bar shown in the figure is made of two segments, AB and BC,
j's
having cross-sectional areas of AAB=200mm2 and ABC=100mm2. Their Young’s moduli are
EAB=100GPa and EBC=210GPa respectively. Find the total displacement at the right end.
ra
Step 1 FBDs for Segments AB and BC.
F2=40kN Assume the internal forces are in
F1=10kN tension.
vi
E 1A 1 E 2A 2
C Step 2 Equilibriums
A B Internal force in AB
ith
4m 4.2m  +
→ ∑ Fx = 0 = − PAB − F2 + F1 = 0
∴ PAB = −30kN
F2=40kN
(Opposite to our assumption of tension,
Pr
PAB F1=10kN so Segment AB is in compression)

FBD1 E 1A 1 E 2A 2
C
B Internal force in BC
PBC F1=10kN
 +
→ ∑ Fx = 0 = − PBC + F1 = 0
FBD2 ∴ PBC = F1 = 10kN (in tension)
C
Step 3 Compute the total deformation by using Eq. (2.18)

PAB L AB PBC LBC − 30 × 10 3 × 4 10 × 10 3 × 4.2
δ AC = δ AB + δ BC = + = +
E AB AAB E BC ABC 100 × 10 9 × 200 × 10 −6 210 × 10 9 × 100 × 10 −6
δ AC = −0.006 + 0.002 = −0.004m = −4mm (towards left)

2.8 STATICALLY INDETERMINATE MEMBERS LOADED AXIALLY
(SI&4th Ed p. 134-139; 3rd Ed p. 137-142)
Statically Determinate and Indeterminate
When a bar is supported at one end and subjected to an axial force P at the other end as shown
in Fig. 2.8(a), there is only one unknown reaction force FA. By using the equations of statics,
the unknown reaction can easily be determined. So such a system with the same number of
unknown reactions as equations of statics is called statically determinate. – i.e. known
reactions can be determined strictly from equilibrium equations.
FA FA
A A
LAC
L
P LCB
j's B
ra
P FB
(a) Statically determinate (b) Statically indeterminate
Fig. 2.8 Statically determinate and indeterminate structures
vi
If the bar is also restricted at the free end as shown in Fig. 2.8(b), it has 2 unknown reactions
FA and FB, one known force P and one equation of statics as:
ith
+ ↑ ∑ Fy = 0 = FA + FB − P = 0
∴ FA + FB = P (2.19)
It cannot be solved if do not introduce more other conditions. If the system has more
Pr
unknown forces than equations of statics it is called statically indeterminate.
Compatibility Conditions
What we need is an additional equation that specifies how the structure is displaced due to the
applied loading. Such an equation is usually termed the compatibility equation.
Since there are 2 unknown and only 1 equation of statics herein, what we need is an
additional equation that specifies how the structure is displaced due to the applied loading.
Such an equation is usually termed the compatibility equation or kinematic conditions.
In order to determine the compatibility for this example we need to determine how point C is
going to move, and how much point B moves in relation to point A. Now, since both ends of
the bar are fully fixed, then the total change in length between A and B must be zero.
Basically the amount that length AC elongates CB contracts as shown in Fig. 2.9, so the
equation can be written as:
δ AC + δ CB = 0 (2.20)

FBD
FA FA
FA
A FBD
Contracted
C C PCB
C PAC
δAC δCB
δ=0 C’ C’
C’ Elongated
δ =δAC +δCB=0
B B
FB
Fig. 2.9 Compatibility condition
Let’s respectively look at the free body diagram for segment AC and CB as in Fig. 2.9.
j's
(indeed FBD can be in any level of structural system or structural members).
For segment AC,
+ ↑ ∑ F y = 0 = FA − PAC = 0 ∴ PAC = FA Tension (+)
ra
PAC L AC F L
δ AC = = A AC elongation (+) (2.21)
AAC E AC AAC E AC
vi
For segment CB,
+ ↑ ∑ F y = 0 = FB + PCB = 0 ∴ PCB = − FB Compression (−)
ith
PCB LCB F L
δ CB = − = − B CB contraction (−) (2.22)
ACB ECB ACB ECB
Compatibility condition:
Pr
F L  F L 
δ = δ AC + δ CB = A AC +  − B CB  = 0 (2.23)
AAC E AC  ACB E CB 
Combining Compatibility equation (2.23) with the equation of statics (2.19), we now can
solve for the two unknowns FA and FB as,
 FA L AC F L
 − B CB = 0
 AAC E AC ACB ECB (2.24)
F + F = P
 A B
L AC LCB
AAC E AC ACB ECB
i.e. FB = P FA = P
L AC LCB L AC LCB
+ +
AAC E AC ACB ECB AAC E AC ACB ECB
If AAC E AC = ACB ECB = Const , we have
L L
FB = AC P and FA = CB P (2.25)
L L

Example 2.5: Two bars made of Copper and Aluminum are fixed to the rigid abutments.
Originally, there is a gap of 5mm between the ends as shown in the figure. Determine average
normal stress in both bars if increase the temperature from 10°C to 210°C.
0.005m
αcu=17 × 10-6 αal=23×10-6
Copper Aluminum
Ecu=110GPa Eal=69GPa
d = 0.01m 0.4m 0.8m d = 0.01m
δT,cu
∆T Thermally expanded δT,cu due to ∆T
Copper
δF,cu
Copper F Mechanical force push it back by δF,cu
F Aluminum
T0=10oC δF,Al
j's
T=210oC
∆T= 200oC ∆T Aluminum
δT,Al
ra
π 2 3.14
A= d = 0.012 = 7.85 × 10 −5 m2 and ∆T = 210 − 10 = 200 o C
4 4
vi
Let’s firstly look at the copper bar. When the bar system is heated up from 10°C to 210°C, the
copper bar expand towards right by δ T ,Cu . After the copper bar touch to the aluminum bar, a
ith
mechanical force F will develop, which will prevent the copper bar from expanding further.
We assume that due to such a mechanical force, the copper bar is pressed back by δ F ,Cu . The
real total deformation of copper bar will be computed as
δ Cu = δ T ,Cu − δ F ,Cu (elongation +, Contraction −)
Pr
Similarly, we have
δ Al = δ T ,Al − δ F ,Al (elongation +, Contraction −)
Because these two expanding bars should fill the gap, we prescribe a compatibility condition as
δ Al + δ Cu = 0.005 (2.26)
From these two equations, we have
(δT ,Cu − δ F ,Cu ) + (δT ,Al − δ F ,Al ) = 0.005
 F × LCu   F × L Al 
i.e.  α Cu ∆TLCu −  +  α Al ∆TL Al −  = 0.005
 E A
Cu Cu   E A
Al Al 
α ∆TLCu + α Al ∆TL Al − 0.005 17 × 10 × 200 × 0.4 + 23 × 10 −6 × 200 × 0.8 − 0.005

−6
F = Cu =
LCu L Al 0 .4 0 .8
+ −5
+
ECu ACu E Al AAl 110 × 10 × 7.85 × 10
9
69 × 10 × 7.85 × 10 −5
9
= 206.2 N
The average normal stress can be computed as
F 206.2
σ= = = 2.63MPa
A 7.85 × 10 −5

 F×L  206.2 × 0.4
δ Cu = δ T ,Cu − δ F ,Cu = α Cu ∆TLCu − Cu
 = 17 × 10 −6 × 200 × 0.4 −
 ECu ACu  110 × 10 9 × 7.85 × 10 −5
= 1.36 × 10 −3 − 9.55 × 10 −6 = 1.35 × 10 −3 m = 1.35mm →
 F × L Al  206.2 × 0.8
δ Al = δ T , Al − δ F , Al = α Al ∆TL Al −  = 23 × 10 −6 × 200 × 0.8 −
 E Al AAl  69 × 10 9 × 7.85 × 10 −5
= 3.68 × 10 −3 − 3.05 × 10 −6 = 3.65 × 10 −3 m = 3.65mm ←
2.9 AVERAGE SHEAR STRESS (SI&4th Ed p. 32-39; 3rd Ed p. 35-41)

The intensity or force per unit area acting tangentially to A is called Shear Stress, (tau), and it
is expressed as in Eq. (2.3) as:
∆Ft
τ = lim (2.3)
∆A→0 ∆A
In order to show how the shear stress can develop in a structural member, let’s take a block as
an example. The block is supported by two rigid bodies and an external force F is applied
vertically as shown in Fig. 2.10. If the force is large enough, it will cause the material of the
j's
block to deform and fail along the vertical planes as shown.
A FBD of the unsupported center segment indicate that shear force V=F/2 must be applied at
each section to hold the segment in equilibrium.
ra
F
vi
A C
Block
ith
Rigid B D Rigid
F
F
Sectioned area A
Pr
V
V
Fig. 2.10 Average shear stress
+ ↑ ∑ F y = 0 = 2V − F = 0
∴V = F / 2
The average shear stress distributed over each sectioned area that develops the shear force is
defined by
V
τ avg = (2.27)
A
τavg = assume to be the same at each point over the section
V = Internal shear force
A = Area at the section

2.10 STRESS CONCENTRATIONS (SI&4th Ed p. 156-161; 3rd Ed p. 159-163)
For a uniform cross-sectional bar that is applied an axial force, both experiment and theory of
elasticity find that the normal stress will be uniformly distributed over the cross-section
Stress Concentration
However, if we drill a hole for some reasons in the component, the typical example is to build
a connection with other structural elements. For such a case, if we cut from the hole’s center
plane, we find that the stress distribution is no longer uniform, as in Fig. 2.11(a). It may
distribute over such a smaller area in highly uneven pattern. We call this phenomenon as
Stress Concentration.
Stress Concentration Factor

In engineering practice, though, the actual stress distribution does not have to be determined.
Instead, only the maximum stress at these sections must be known, and the member is then
designed to resist this highest stress when the axial load is applied. The specific values of the
maximum normal stress at the critical section can be determined by experimental methods or by
advanced mathematical techniques using the theory of elasticity. The results of these
investigations are usually reported in graphical form (as in Fig. 2.11(c)) in terms of Stress
j's
Concentration Factor K.
σ max
K= (2.28)
σ avg
ra
in which σavg=P/A’ is the assumed average stress as in Fig. 2.11(b). Provided K has been known
from the figures or tables (as in Fig. 2.11(c)), and the average normal stress has been calculated
from σavg=P/A’, where A’ is the smallest cross-sectional area. Then from the above equation the
vi
maximum stress at the cross section can be computed as:
P
σ max = Kσ avg = K (2.29)
A'
ith
P
Stress Concentration Factor K
Pr
(a)
A’
P
w 2r
r/w
(b)
(c)
Fig. 2.11 Stress concentration
Stress concentration occurs in the case that there is a sudden change in cross-sectional area.
By observing Fig. 2.11(c), it is interesting to note that the bigger the ratio of change in the
sectional area, the higher the stress concentration.

Chapter 3 Analysis of Pin Jointed Frames
A pin jointed frame (PJF) consists of rod members sustaining only uniform uniaxial stress.
They are joined together by frictionless pin joints. Since the entire truss is in equilibrium, each
pin must be in equilibrium. For each pin, we can only give two force equilibrium equations as
∑ Fx = 0 and ∑ Fy = 0 . The moment equation ∑ M o = 0 does not work in fact.
3.1 STATICALLY DETERMINATE PJF

The methods used in “engineering mechanics/statics” are subject to a condition of statical
determinacy. It would be helpful to examine the determinacy before solving the problem.
If J denotes the number of pin-joints:

Known 2×J = number of equilibrium equations
Unknown m = number of member forces
r = number of reaction forces
j's
• If m + r = 2J , the frame is statically determinate
• If m + r > 2J , the frame is statically indeterminate
• If m + r < 2J , the frame is a mechanism.
ra
Examples 3.1: To check the determinacy of the following structures.
A B
vi
J = 4, m = 5, r = 3
2J = 8
m+r=8
ith
RDx D ∴m + r = 2J
C This is a statically determinate structure. The internal forces
RDy RCy carried by each member can be fully determined by using
statics method
Pr
A B
J = 4, m = 6, r = 3
2J = 8
RDx D m+r=9
C
∴m + r > 2J
RDy RCy This is a statically indeterminate structure.
A’ B’
A B
J = 4, m = 4, r = 3
2J = 8
m+r=7
RDx D
C
∴m + r < 2J
RDy
This is a mechanism.
RCy
Lecture Notes Chapter 3, Mechanics of Solid I 1

Examples 3.2: Check the determinacy of the following structures
A B
J = 5, m = 7, r = 3
2J = m + r
RDx D 10 = 7 + 3
C This is a statically determinate structure.
A
RDy RCy
J = 6, m = 10, r = 3
2J = m + r
12 < 10 + 3 = 13
This is a statically indeterminate structure.
J = 6, m = 8, r = 3
2J = m + r
12 > 8 + 3 = 11
j's
This is a mechanism.
ra
Now that we know how to determine if the pin-jointed frame is statically determinate, we need
to find out the force carried by each of the beam members. To do this we can apply one of two
vi
analytical methods:
1) Joint Equilibrium
ith
or
2) Method of Sections
You could also use a graphical method known as Maxwell diagram, but that would not give as
Pr
accurate an answer.
3.2 JOINT EQUILIBRIUM

At each pin joint set up the two equilibrium equations and solve them simultaneously. It is best
to start by first evaluating the global equilibrium of the structure (ie; find the required ground
reaction forces).
Notation of Internal Forces

Before we begin, however we need to define the notation for a beam experiencing a tensile and
compressive force. A tensile force in member is denoted in positive sign and a compressive
force in member is represented in negative sign, as shown in Fig. 3.1.
In pin equilibrium, however, we need to look at the reaction from the member. A pin force
pointing away from the pin-joint corresponds to a tensile internal force, which is positive.

Pin B Hole B
Tensile Force Compressive

in Member Force
m m m in Member
em em em
be be be
r r r
Pin A Hole A
(+) (―)
A force pointing away from pin-joint

is equivalent to a tensile internal
force in the member according to
the relationship between action and Pin A
reaction forces. (+) Pin A (―)
Fig. 3.1 Notation of internal forces
j's
Remark
In order to do this analysis we have to start with a pin-joint that has at least ONE known force
and no more than 2 unknown forces.
ra
Example 3.3: Determine the internal forces carried by all the beams in the following pin-
jointed frame (all angles are 45o or 90o and the length of AB is L).
vi
RDY
ith
D E Step 0: Determinacy
RDX
J = 5, m = 7, r = 3
2J = m + r
10 = 7 + 3
Pr
This is a statically determinate structure

A 45º 45º and can be solved through the method of
B
RA C pin equilibrium.
L L
P=10kN
Step 1: Global Equilibrium
+ ↑ ∑ Fy = 0 = RDY − 10kN = 0 ∴ RDY = 10kN

+∑ M = 0 = R L − 10 × 2 L = 0
D A ∴ R A = 20kN
→ ∑ F = 0 = R + R = 0
 +
x A DX ∴ R DX = − R A = −20kN
Step 2: Pin-Joint Equilibrium

Let’s start with Pin-Joint C: Because we don't know the direction of forces FEC and FBC we
assume them to be positive (i.e. in tension), hence the arrow which describes them has to point
away from the pin-joint as shown.

FCE Pin-Joint Equilibrium at C
+ ↑ ∑ Fy = 0 = FCE sin 45° − 10kN = 0 ∴ FCE = 10 2kN
45O +
→ ∑ Fx = 0 = − FCE cos 45° − FBC = 0 ∴ FBC = −10kN
FBC C
Pin-Joint Equilibrium at E:
10kN
FDE E  +
→ ∑ Fx = 0 = − FDE + FCE cos 45° = 0
( )
∴ FDE = 10 2 × cos 45° = 10kN
+ ↑ ∑ Fy = 0 = − FCE sin 45° − FBE = 0
FCE
FBE
( )
∴ FBE = − 10 2 × sin 45° = −10kN
FAD Pin-Joint Equilibrium at A:

→+
∑ Fx = 0 = R A + FAB = 0 ∴ FAB = − R A = −20kN
RA + ↑ ∑ Fy = FAD = 0 ∴ FAD = 0
FAB
j's
A
Pin-Joint Equilibrium at B:
FBE Although we’ve known that FBC and FAB are in compression, we can
FBD
ra
still draw them in the positive directions and then substitute their
values together with the negative signs to the equilibrium equations.
FBC
FAB
B
→+
∑ Fx = 0 = FBC − FBD cos 45° − FAB = 0
vi
∴ FBD = (FBC − FAB ) / cos 45° = ((− 10) − (− 20)) / cos 45°
∴ FBD = 10 2kN
ith
+ ↑ ∑ Fy = FBE + FBD sin 45° = 0

∴ FBE = − FBD sin 45° = −10kN (Checks)
Pr
Step 3: Tabulate the Results
member Fi (kN) Behavior

CE 10 2 Tension
BC -10 Compression
BE -10 Compression
DE 10 Tension
BD 10 2 Tension
AB -20 Compression
AD 0

3.3 METHOD OF SECTIONS
Cut the FBD of the pin jointed frame through 3 or less members, then use the three
equilibrium equations of statics. After you draw the unknown internal forces pointing away
(tensile – positive) from the cut members, you may treat the rest of the free body as a whole
body and solve for unknowns from equilibrium equations.
Example 3.4: Using the structure of example 3.3, determine the forces carried by members
DE, DB, and AB.
I
D E
45º
C
A B
I
j's
P=10kN
Instead of solving for all the forces we just cut through those three beams and draw a new
FBD. We can draw a FBD of the left hand side or of the right hand side, it really makes no
difference which one you decide to analyse, both will give you the same answer.
ra
For this example we will analyse the RHS.
vi
E + ↑ ∑ Fy = 0 = FDB sin 45° − 10 = 0

FDE
ith
∴ FDB = 10 2kN
+ ∑M B = 0 = FDE L − 10 L = 0
FBD ∴ FDE = 10kN
Pr
45º
FAB
B
C
+
→ ∑ Fx = 0 = − FAB − FDE cos 45° − FDE = 0
L ∴ FAB = −20kN
P=10kN
3.4 STRAIN ENERGY AND WORK (4th:705-709, 5th:705-709)

Work Done by Force
The work (W) done on an elastic body by applied external force(s) (P) is fully accumulated
inside of it in the form of strain energy (U).
What actually happens is as follows, as you apply forces onto a structure it is going to deform
elastically. This deformation, combined with the applied forces gives you a measure of the
work done on the structure. Now because the structure is made from an elastic material, this
deformation causes the structure to store the work as energy, to be specific, Strain Energy.

To derive the strain energy term, let’s look at a small element of size dx×dy×dz, with an
applied normal stress σz.
P
σz
δ
d∆z
dz y
dx
dy
x
Fig. 3.2 Strain energy in a differential element
Resultant Force of differential element

j's
ra
The force developed in the top and bottom surfaces due to the applied stress is:
dFz = σ z dA = σ z dxdy (3.1)
vi
Deformation
The deformation due to the resultant differential force is:
d∆ z = ε z dz (3.2)
ith
We are only considering structures made from materials with linear elastic behavior. So
plotting a diagram of applied force vs displacement gives:
dFz
Pr
Work done
by dFz
d ∆z
Fig. 3.3 Diagram of force vs displacement for the differential element
The work done on this differential element by the applied force is defined as the area
underneath this diagram, such that:
1
dW = dFz ⋅ d∆z (3.3)
2
To understand why the work done by force dFz must be computed through dividing the product
of force dFz and deformation d∆z by 2 as in Eq. (3.3), it should be noted that the force
magnitude is gradually increased from zero to dFz, as depicted in Fig. 3.3.

But the loading process is isentropic (reversible), meaning that the Strain Energy must be equal
to the applied external Work, so W=U.
Relationship between Work and Strain Energy

So for our small volume, the elemental work is equal to the elemental strain energy
according to the energy conservation law:
dW = dU (3.4)
Combining the above 4 equations (Eqs. (1) to (4)) together we obtain:
1 1
dU = dFz d∆ z = [σ z dxdy ]ε z dz
2 2
But since the volume of this element is :
dV = dx × dy × dz
then the above equation becomes:
1
dU = σ z ε z dV (3.5)
2
An important relationship can be found from Eq. (3.5), if we compare it with a stress-strain
diagram for the small element.
σz j's
ra
vi
εz
ith
Fig. 3.4 Stress-strain diagram for the differential element
Because the structure is made of a linear elastic material, then the stress strain diagram looks
Pr
like Fig. 3.4.
Strain Energy Density

dU 1
Therefore: = σ z ε z ― Strain-energy per unit volume =Strain Energy Density
dV 2
The term σ z ε z / 2 is the area underneath the stress-strain diagram (Fig. 3.4) and is called the
Strain-Energy Density. It is a measure of the strain energy stored in the structure per unit
volume.
Mathematically can the strain energy density be written like this:
1
U = σε (3.6)
2
Modulus of resilience Ur ―When stress reaches the proportional limit, strain energy density is
referred to Modulus of resilience Ur as
1
U r = σ pl ε pl (3.7)
2

Total area = Modulus of toughness Ut
σ
σy
σpl
1
Modulus of resilience U r = σ pl ε pl
2
Fig. 3.5 Modulus of resilience u r and modulus of toughness
Modulus of toughness Ut ― This property becomes important when designing member is

accidentally overloaded. The higher the modulus of toughness, the more strain energy can be
stored (absorbed) inside the material before fracturing
Total Strain Energy inside deformable body
j's
Equation (3.5) is the elemental strain energy, by integrating it w.r.t. the total volume and
substituting Hooke's law we have that:
σε σ2
U =∫ dV = ∫ dV (3.8)
ra
V 2 V 2E
vi
3.5 STRAIN ENERGY UNDER AXIAL LOADING (4th:710-711, 5th:710-711)
ith
Let’s look at a general case of bar under an axial loading as illustrated in Fig. 3.6. Assume the
internal force F(x) at a cross-sectional area A(x) located at a distance x from end B. Assume
that the normal stress σx is uniformly distributed over any transverse section.
Pr
y
x
A(x)
B F(x) C
z x
Fig. 3.6 Strain energy for a bar under axial loading
General Cases
F (x )
Average normal stress σ x =
A( x )
Substituting for σx into Eq. (3.8), we have the Total Strain Energy as:
F 2 (x )
U =∫ dV (3.9)
V 2 E (x )A 2 (x )

But dV = A(x )dx , therefore
L F 2 (x )
U =∫ dx (3.10)
0 2 E ( x )A( x )
Uniform rod
If F ( x ) = constant, A( x ) = constant, E ( x ) = constant (homogeneous material) as shown in
Fig. 3.7, we can rewrite Eq. (3.10) as
F2 L F 2L
2 EA ∫ 0
U= dx = (3.11)
2 EA
F A
F
L
j's
Fig. 3.7 Strain energy for a uniform bar
Example 3.5: A circular shaft consists of two portions BC and CD of the same material and
ra
same length, but of different cross sections as shown. Determine the total strain energy of the
shaft when it is subjected to a centric axial load P, expressing the result in terms of P, L, E, the
cross-sectional area A of portion CD and the ratio n of the two diameters.
vi
L/2 L/2
ith
P
B C D
Pr
A
Area=n2A
We use Eq. (3.11) to compute the strain energy of each of the two portions and add the
expressions obtained the total strain energy as
P 2 (L 2 ) P 2 ( L 2 ) P 2 L  1 
U n = U CD + U BC = + = 1 + 2 
2 EA 2
(
2 E n A 4 EA  n )
1+ n2  P2L
or U n =  
2 
 2n  2 EA
P2L
We check that, for n =1, we have U 1 = , which is the expression given in Eq. (3.11) for a
2 EA
rod of length and uniform cross-section area A. We also note that, for n>1, we have Un < U1;
for example, when n = 2, we have U 2 = 5 U 1 . Since the maximum stress occurs in portion
8
CD, it follows that for a given allowable stress, increase the diameter of portion BC will result
in a decrease of overall energy-absorbing capability of the rod. So unnecessary changes in
cross-sectional area should avoided in design of structure, where the energy-absorbing
capability is critical.

3.6 DEFLECTION UNDER A SINGLE CONCENTRATED LOAD
BY THE WORK-STRAIN ENERGY METHOD (4th:724-726, 5th:724-726)
External Work
If the deflection ∆1 of a structure or member under a single concentrated load P1 is known, the
corresponding Work W may be obtained by writing
1
W = P1 ∆ 1 (3.12)
2
The total work done by the external load P1 will be fully stored in the structure.
Internal Strain Energy

Conversely, if the strain energy U of a structure or member subjected to a single concentrated
load P1 can be computed as Eqs. (3.10) and (3.11).
Work-Strain Energy Method

by using energy conservation as W = U, we can determine the deflection under such a single
load
W = U (External Work = Internal Strain Energy) (3.13)
j's
1
P1 ∆ 1 = U
2
2U
Therefore, we have: ∆1 = (3.14)
ra
P
Example 3.6: A load P is supported at B by two rods of the same uniform cross-sectional area
vi
A. Determine the vertical deflection of point B.
C F.B.D. of Pin B
ith
3 FBC
4 B Step 0: Geometric parameters
θ1 4 3
B cos θ1 = , cos θ 2 =
L θ2 5 5
Pr
4 4
3 sin θ1 = , sin θ 2 =
FBD 5 5
4
P LBC = L sin θ1 = 0.6 L ,
P
LBD = L cos θ1 = 0.8 L
D
Step 1. Internal Forces carried by Members
Pin-Joint Equilibrium Method

+ → ∑ Fx = 0 = − FBC cos θ1 − FBD cos θ 2 = 0
+ ↑ ∑ Fy = 0 = FBC sin θ1 − F BD sin θ 2 − P = 0
we have
 4 3
 − 5 FBC − 5 FBD = 0
3 4
 FBC − FBD − P = 0
5 5
Thus FBD = −0.8 P , FBC = 0.6 P

Step 2. Total Strain Energy
F2 L F2 L
U = BC BC + BD BD
2 ABC E BC 2 ABD E BD
Therefore
U=
(
P 2 L 0.6 3 + 0.8 3 )
= 0.364
P2L
2 AE AE
Step 3. Determine the deflection from the Work-Energy Method

1
External Work done by P: W = P∆ BY
2
P2L
Total Strain Energy: U = 0.364
AE
From the Conservation of Energy W = U , we have
1 P2L
P∆ BY = 0.364
2 AE
PL
j's
Thus: ∆ BY = 0.728 ANS
AE
ra
Remarks
It should be noted that, once the internal forces in the two rods have been obtained, the
deformations δBD and δBC (elongation or contraction) can be computed. Determining the
vertical deflection at point B from these deformations, however, would require a careful
vi
geometric analysis of various displacements involved. The strain energy method used here
makes such an analysis unnecessary and significantly simplifies the problem.
ith
Example 3.7: Members of the truss shown consist of sections of aluminum pipe with the
cross-sectional areas indicated. Using E=70GPa, determine the vertical deflection of point E
caused by the load P.
Pr
Step 0: Basic Parameters

LDE = L2CE + L2CD = 1.5 2 + 0.8 2 = 1.7 ; L AD = L2AB + L2BD = 0.8 2 + 0.6 2 = 1.0
15 8
cos α = , sin α = ; cos θ = 0.8 , sin θ = 0.6
17 17
I P=40kN
A C E
α
Cross-Sectional Areas
0.8m AAC=ACE=AAB=AAD=500mm2
ABD=ACD=ADE=1000mm2
B θ
D
I
0.6m 1.5m

Step 1: Internal Forces
Method of Section (I-I)
P=40kN + ↑ ∑ Fy = FAD sin θ − P = 0
C E
FAC
α ∴ FAD = 5 P / 4
FAD
+ ∑M D = 0 = FAC × 0.8 − P × 1.5 = 0
∴ FAC = 15P / 8
FBD + → ∑ Fx = − FAC − FBD − FAD cos θ = 0

θ
D ∴ FBD = −21P / 8
Pin-Joint Equilibrium at E
+ ↑ ∑ Fy = − FDE sin α − P = 0
P
∴ FDE = −17 P / 8
FCE α E + → ∑ Fx = − FCE − FDE cos α = 0
FDE ∴ FCE = − FDE cos α = −(− 17 P / 8)(15 / 17 ) = 15P / 8
Pin-Joint Equilibrium at B
j's
ra
FAB + ↑ ∑ Fy = FAB = 0
∴ FAB = 0
FBD
vi
RBX
B
ith
Pin-Joint Equilibrium at C
C
FCE + ↑ ∑ Fy = − FCD = 0
FAC
Pr
∴ FCD = 0
FCD
Step 2: Total Strain Energy
F 2L
U =∑ i i
i 2 E i Ai
Member Fi (N) Li(m) Ai (m2) ( )

U i = Fi 2 Li (2 Ei Ai )
AB 0 0.8 500×10-6 0
AC 15P/8 0.6 500×10-6 4219P2/2E
AD 5P/4 1.0 500×10-6 3125P2/2E
BD -21P/8 0.6 1000×10-6 4134P2/2E
CD 0 0.8 1000×10-6 0
CE 15P/8 1.5 500×10-6 10547P2/2E
DE -17/8 1.7 1000×10-6 7677P2/2E
From the summation of the last column, we have the total strain energy as

Fi 2 Li 29700 P 2
U =∑ =
i 2 Ei Ai 2E
Step 3: Principle of Work ― Strain Energy

1
We recall that the work done by the force P as it is gradually applied is P∆ Ey . Equating the
2
work done by P to the total strain energy U and recalling that E = 70GPa and P=40kN, we have
W=U
1 29700 P 2
P∆ Ey =
2 2E
∆ Ey =
29700 P
=
(29700) × (40 × 10 3 ) = 16.97 × 10 −3 m
E 70 × 10 9
Therefore:
∆ Ey = 16.97 mm ↓ (“+” stands for the same direction as load P)
Remarks
Work―Strain Energy method is very efficient for determining the deflection. However, it is
j's
effective only when
(1) there is a single load; and
(2) the unknown deflection is at the loading point and along the same direction as load P.
ra
vi
3.7 DEFLECTION UNDER MULTIPLE LOADS
― CASTIGLIANO’S SECOND THEOREM (4th:762-767, 5th:762-767)
ith
In order to derive Castigliano’s second theorem, we consider two loading sequences:

Pr
Sequence 1
Pn P1 P1
Pn
∆n ∆1 ∆1
∆n
∆2 ∆2
∆j P2 P2
∆j
Pj Pj
dPj
Firstly, apply a set of forces Secondly, apply a
P1, P2, …Pj, …Pn differential force dPj,
Fig. 3.8 Loading Sequence 1

First step, apply a set of force (P1, P2, …Pj, …Pn) onto the elastic body. Suppose that only
linear elastic deformation is considered. By the end of the first step, the external work done by
the force set will be
1 1 1 1
WS1 = P1∆1 + P2 ∆2 + L + Pj ∆ j + L + Pn ∆n
2 2 2 2
i.e.
(
WS1 = f P1 , P2 ,L , Pj ,L , Pn ) (3.15)
The work done by such n forces will be fully stored inside the body in a form of strain energy,
i.e.
( ) (
U = WS1 = f P1 , P2 ,L , Pj ,L , Pn = U P1 , P2 ,L , Pj ,L , Pn ) (3.16)
Second step, increase one of those external forces, e.g. Pj by a differential amount dPj . The
new total strain energy inside the elastic body will become
(
U S1 = U + dU = U P1 , P2 ,L , Pj + dPj ,L , Pn )
According to integral calculus, we can express the new energy as
∂U
U S1 = U + dPj (3.17)
∂Pj
Sequence 2
We are going to reverse the load sequence.
j's
ra
Pn P1
∆1
∆n
vi
ith
∆2
d∆j dPj ∆j P2
Pr
dPj Pj
Firstly, apply a
differential force dPj, Secondly, apply a set of forces
P1, P2, …Pj, …Pn
Fig. 3.9 Loading Sequence 2
First step Apply a differential force dPj onto the body, which cause the body to deform by a
differential amount d∆j. The work done by the differential force should be
dW1 = (dPj )(d∆ j )
1
(3.18)
2
Second step Apply a set of forces (P1, P2, …Pj, …Pn) onto the elastic body. The work done at
this loading step should be computed as
W2 = P1 ∆ 1 + P2 ∆ 2 + L + Pj ∆ j + L + Pn ∆ n + (dPj )(∆ j )
1
2
1
2
1
2
1
2
[ ]
(3.19)
It is worth noting that since force (P1, P2, …Pj, …Pn) cause the jth loading point move by ∆j, the
work done by dPj needs to be taken into account as (dPj )(∆ j ), where the differential force dPj
does not change.

Adding Eqs. (3.18) and (3.19) gives the total external work done by the forces in Sequence 2 as
1
( )( ) 1 1 1
WS 2 = dPj d∆ j + P1∆1 + P2 ∆2 + L + Pj ∆ j + L + Pn ∆n + dPj ∆ j
2 2 2 2
1
2
[( )( )]
Neglecting the second-order differential (dPj )(d∆ j ) / 2 , and referring to Eq. (3.16), we have
[( )( )]
U S 2 = WS 2 = U + dPj ∆ j (3.20)
However, the strain energy should NOT depend on the loading sequences. Therefore,
U S1 = U S 2 (3.21)
Substituting Eq. (3.17) and (3.20) into Eq. (3.21) gives
U+
∂U
∂Pj
[ ]
dPj = U + (dPj )(∆ j )
Thus we obtain
∂U
∆j = (3.22)
∂Pj
Castigliano’s 2nd theorem, Eq. (3.22), states that displacement ∆j in the direction of Pj is
j's
equal to the first partial derivative of the total strain energy with respect to Pj.
Castigliano’s theorem applied to truss

ra
In a truss structure (consisting of N members), the total strain energy can be often expressed as
N
F 2L
U =∑ i i (3.23)
i =1 2 Ei Ai
vi
The displacement can be thus computed as

ith
∂  F 2L   ∂F  Li 
∆=  ∑ i i  = ∑ (Fi ) i   (3.24)
 
∂P  i 2 Ei Ai  i  ∂P  Ei Ai 
In order to determine the partial derivative ∂Fi ∂P , it will be necessary to treat external load
Pr
P as a variable, not a specific numerical quantity. In other words, each internal force Fi
must be expressed as a function of variable P.
Virtual load in Castigliano’s theorem

It is worth pointing out that the unknown deflection ∆j at a given point j can be obtained by
direct application of Castigliano’s theorem ONLY IF a load P happens at j in the direction in
which ∆j is to be determined. When no load is applied at j, or when a load is applied in a
direction other than the desired one, we may still obtain the deflection ∆j by using Castigliano’s
theorem. But we need a special technique. Firstly we need to apply a virtual (fictitious) load
Qj at j in the direction in which the deflection ∆j is to be determined. And then apply
Castigaliano’s theorem to obtain the deflection as
∂U ∂  F 2L   ∂F  L 
∆= =  ∑ i i  = ∑ (Fi ) i  i  (3.24)
∂Q j ∂Q  i 2 E i Ai  i  ∂Q  E i Ai 
in which the internal forces should be found in terms of both real load P and virtual load
Qj. In other words, internal force Fi could be a function of both P and Qj. After finding out all
the partial derivatives ∂Fi ∂Q j , we then make Q j = 0 .

Example 3.8: As in Example 3.6, a load P is supported at B by two rods of the same uniform
cross-sectional area A. Determine both the vertical and horizontal deflections at point B.
Step 0: Geometrical parameters

4 3 4 4
cos θ1 = , cos θ 2 = ; sin θ1 = , sin θ 2 =
5 5 5 5
LBC = L sin θ1 = 0.6 L ; LBD = L cos θ1 = 0.8 L
C F.B.D. of Pin B
FBC
3
4 B B
θ1
Q Q
L θ2
3
FBD
j's
4
P
P
D
ra
Step 1: Apply a virtual load Q at B in order to find unknown deflection ∆ BX
A “real” load has been applied at point B along vertical direction. So Castigliano’s theorem can be
vi
directly used to determine ∆ BY .
However, since there is no horizontal “real” force applied at point B, we cannot directly compute
ith
the deflection ∆ BX . For this reason, a virtual load Q (horizontal direction as shown) is necessary.
Step 2 Internal forces due to BOTH real load P AND virtual load Q
Pin-Joint Equilibrium at B
Pr
+ → ∑ Fx = 0 = − FBC cos θ1 − FBD cos θ 2 + Q = 0

+ ↑ ∑ F y = 0 = FBC sin θ1 − F BD sin θ 2 − P = 0
we have (assume that Q is known)
FBC = 0.6 P + 0.8Q ; FBD = −0.8 P + 0.6Q
Step 3 Total Strain Energy

F2 L F2 L (0.6 P + 0.8Q )2 LBC (− 0.8P + 0.6Q )2 LBD
U = BC BC + BD BD = +
2 ABC E BC 2 ABD E BD 2 ABC E BC 2 ABD E BD
Step 4 Castigliano’s Theorem

∂U FBC LBC  ∂FBC  FBD LBD  ∂FBD 
∆ BX = =    
∂Q ABC E BC  ∂Q  +
  ABD E BD  ∂Q 
∂U FBC LBC  ∂FBC  FBD LBD  ∂FBD 
∆ BY = =    
∂P ABC E BC  ∂P  + ABD E BD ∂P
   

Computing the partial derivatives
∂FBC ∂ (0.6 P + 0.8Q ) ∂FBD ∂ (− 0.8 P + 0.6Q )
= = 0 .8 ; = = 0 .6
∂Q ∂Q ∂Q ∂Q
∂FBC ∂ (0.6 P + 0.8Q ) ∂FBD ∂ (− 0.8P + 0.6Q )
= = 0.6 ; = = −0.8
∂P ∂P ∂P ∂P
Substituting these partial derivatives into the previous equations, we have

∂U (0.6 P + 0.8Q )LBC (− 0.8P + 0.6Q )LBD ( )
∆ BX = = × (0.8) + × 0.6
∂Q ABC E BC ABD E BD
∂U (0.6 P + 0.8Q )LBC (− 0.8P + 0.6Q )LBD (
∆ BY = = × (0.6 ) + × − 0 .8 )
∂P ABC E BC ABD E BD
In fact, Q is a fictitious load and does not exist at all, so we make Q = 0, thus
∂U (0.6 P + 0.8 × 0 )LBC (− 0.8P + 0.6 × 0)LBD ( )
∆ BX = = × (0.8) + × 0 .6
∂Q ABC E BC ABD E BD
∂U (0.6 P + 0.8 × 0 )LBC (− 0.8P + 0.6 × 0)LBD (
∆ BY = = × (0.6 ) + × − 0 .8 )
j's
∂P ABC E BC ABD E BD
i.e.
(0.6 P )LBC (− 0.8P )LBD ( ) PL
∆ BX = × (0.8) + × 0.6 = −0.096
ra
AE AE EA
(0.6 P )LBC ( )
− 0.8 P LBD PL
∆ BY = × (0.6 ) + × (− 0.8) = 0.728
AE AE EA
vi
Finally we have,
PL
∆ BX = −0.096 (- stands for the opposite direction to the assumed virtual load Q)
ith
EA
PL
∆ BY = 0.728 (+ stands for the same direction as real load P)
EA
Pr
Example 3.9: As in Example 3.7, determine the vertical deflections at point C.
Q
I P=40kN
A C E Q
α P=40kN
C E
FAC
α
0.8m
FAD
B θ
D FBD θ
I D
0.6m 1.5m
Step 1: Apply a virtual load Q at C in order to find unknown vertical deflection ∆ CY

Since there is no external load applied at point C, we cannot directly use Castigliano’s theorem
to compute the deflection ∆ CY . For this reason, we have to apply a virtual load Q at C along
vertical direction, as shown in the illustration (with dashed line).

Step 2 Internal Forces due to both real load P and virtual load Q
Method of sections (I-I)
Q
P=40kN + ↑ ∑ Fy = FAD sin θ − P − Q = 0
C E
FAC ∴ FAD = 5 P / 4 + 5Q / 4
α
+ ∑ M D = 0 = FAC × 0.8 − P × 1.5 = 0
FAD
∴ FAC = 15P / 8
FBD θ + → ∑ Fx = − FAC − FBD − FAD cos θ = 0

D ∴ FBD = −21P / 8 − 3Q / 4
Pin-joint equilibrium at E
P + ↑ ∑ Fy = − FDE sin α − P = 0
α E ∴ FDE = −17 P / 8
FDE + → ∑ Fx = − FCE − FDE cos α = 0

∴ FCE = − FDE cos α = −(− 17 P / 8)(15 / 17 ) = 15P / 8
Q
j's
C FCE
FAC Pin-joint equilibrium at E
+ ↑ ∑ Fy = − FCD − Q = 0
ra
FCD ∴ FCD = −Q
Similarly to Example 3.7, from Pin-Joint equilibrium B we can
have FAB = 0 .
vi
Fi 2 Li
Step 3: Total Strain Energy U =∑
ith
i 2 E i Ai
Step 4: Application of Castigliano’s Theorem

∂U   L  ∂F  1  L  ∂Fi 
∆ CY = = ∑ (F )i  i  i  = ∑ (F )i  i  
Pr
∂Q i   Ei Ai  ∂Q  E i   Ai  ∂Q 
Summarize the computations in a table as below
Member Fi (N) ∂Fi ∂Q Li(m) Ai (m2) Fi (Li Ai )(∂Fi ∂Q )
-6
AB 0 0 0.8 500×10 0
AC 15P/8 0 0.6 500×10-6 0
AD 5P/4+5Q/4 5/4 1.0 500×10-6 3125P+3125Q
BD -21P/8-3Q/4 -3/4 0.6 1000×10-6 1181P+338Q
CD -Q -1 0.8 1000×10-6 800Q
CE 15P/8 0 1.5 500×10-6 0
DE -17/8 0 1.7 1000×10-6 0
From the summation of the last column, we have:

∆ CY = (4306 P + 4263Q ) E
Since Q is not part of the original loading, we set Q = 0 and thus have
1
∆ CY = (4306 P + 4263 × 0) =
(4306 × 40 × 10 3 ) = 2.46 × 10 −3 m = 2.46mm ANS
E (70 × 10 9 )

(“+” stands for the same direction as applied virtual load Q, i.e. ↓ )
j's
ra
vi
ith
Pr

Chapter 4 Torsion of Circular Shafts
In addition to the bars/rods under axial loads as discussed in Chapters 1 to 3, there are other
loading cases in engineering practice. In this chapter we will discuss the effects of applying a
torsional loading to a long straight circular member such as a shaft or tube, as extracted from
the machine showing in Fig. 4.1. We are going to show how to determine both the
• Shear strain and shear stress
• The angle of twist
Wires
A B
Shaft
Turbine Transmit
Generator electrical
Machine
Transmit power
mechanical
power
F.B.D.
Driven Torque TD Resistant Torque TR
j's
Fig. 4.1 Engineering example of torsional shaft
ra
4.1 SHEAR STRESS/STRAIN RELATIONSHIP
(4th: 69-70,106-107; 5th:69-70,106-107)
vi
Shear Stress
Let’s recall the definition of shear stress in Chapter 2. When parts of a deformable body try to
ith
slip past another part, a shear stress is set up.

P Shear force Shearing in torsion
Pr
∆ Fn ∆ F T
∆ Ft
Cross section ∆A τ
Fig. 4.2 Definition of shear stress and shearing in torsion
The shear stress equation was defined in section 2.1 as:

∆Ft
τ = lim (4.1)
∆A→0 ∆A
which is a shear force intensity that acts parallel to the material cross sectional plane as
shown in Fig. 4.2. It is worth pointing out that the shear stress in an element always comes
with pairs to maintain equilibrium as shown in Fig. 4.3.

Shear Deformation
We call the deformation created by shear stress as Shear Strain, given the symbol γ
(gamma). It is defined as the change in angle of the element, it is a non-dimensional quantity.
Unit of shear strain is radian

τ
τ
γ
τ
τ
Fig. 4.3 Element of material with applied shear stress τ and shear strain γ
Hooke’s Law for Shear

By conducting a similar material testing to normal stress (Section 2.4), there is a linear
relationship (for most engineering materials) between the shear stress and shear strain, as
j's
shown in Fig. 4.4.
τ
ra
Gradient = G
1
vi
γ
Fig. 4.4 Relationship of shear stress τ – shear strain γ for linear elastic material
ith
This relationship is called Hooke's law for Shear and is represented by equation Eq. (4.2).
τ = Gγ (4.2)
where: G = Shear Modulus of Elasticity (for short, Shear Modulus) or Modulus of Rigidity.
Pr
As the shear modulus is a material property (determined by material shear testing), it is

related to the Young's Modulus E and Poison's ratio v by the following equation (we are going
to prove this late),
E
G= (4.3)
2(1 + v )
Now that we have these relationships we can examine the effect of an applied torque to the
shaft of circular cross section.
4.2 TORSION OF CIRCULAR SECTIONS (4th: 177-188;5th:177-188)
Assumptions
• This analysis can only be applied to solid or hollow circular sections
• The material must be homogeneous
• Torque is constant and transmitted along bar by each section trying to shear over its neighbor.
• Transverse planes remain parallel to each other.
• For small angle of rotation, the length of shaft and its radius remain unchanged.

Shear Strain/Stress Distribution
Examine the deformation of a length dx between two transverse planes of a shaft with an
applied torque T.
For this differential element, assume the left end is fixed and the right end rotates by dφ due to
the applied torque T, where dφ is termed as Angle of Twist of the element
Twisted
End
γ ρ
Fix
End x
da dφ
T
dx
Fig. 4.5 Small transverse element with applied torque T rotated by an amount dφ
The surface of radius “ρ” rotates through angle γ, which is shear strain.
The arc is defined as length da, which is equal to:
j's
da = ρdϕ = γdx
which gives that:
dϕ
γ=ρ (4.4)
ra
dx
dϕ
where: = Rate of Twist (4.5)
dx
vi
which is constant for the cross-sectional plane. Eq. (4.4) states that the magnitude of shear
strain for any of these elements varies only with its radial distance ρ.
ith
By using Hooke’s law:

τ = Gγ (4.2)
and by substituting for shear strain γ, Eq. (4.4), Eq. (4.2) becomes that:
Pr
dϕ
τ = ρG (4.6)
dx
which relates the shear stress linearly to the distance ρ away from the centre of the section.
As a result, the shear stress distribution then looks as Fig. 4.6.
Distribution of
shear stress
Fig. 4.6 Shear stress distribution in circular section with applied torque T
Torque T and Rate of Twist

We now equate the applied torque T to the torque generated in the section by the shear stress
distribution. To do this, look at a small circumferential section dA, as shown in Fig. 4.7.

Elemental
dρ distribution of
shear stress
dA
Fig. 4.7 Shear stress distribution in elemental section dρ of cross section
The elemental torque of a thin circular strip of thickness dρ is given by

dT = ρ ⋅ (τdA) (4.7)
Integrating over the area of the circular beam gives
T = ∫ ρτdA = ∫ ρτ(2πρdρ)
A ρ
Substituting for shear stress as Eq. (4.6):
 dϕ 
T = ∫ 2πρ ρG ρdρ (4.8)
ρ  dx 
j's
Since the rate of twist dϕ dx is constant through the section, it is not a function of radius ρ. If
we assume a homogeneous material, G is also constant, so:
dϕ dϕ
T =G ∫ 2πρ 3 dρ = G J (4.9a)
ra
dx ρ dx
dϕ T
or = (4.9b)
dx GJ
vi
Polar Moment of Inertia J

We represent the integral term as the geometric rigidity of the cross section. We call this term
ith
the Polar (Second) Moment of Inertia, J.

J = ∫ 2πρ3 dρ (4.10)
ρ
Pr
This term indicates the cross sectional properties to withstand the applied torque.
Since this applies to circular bars, the standard terms for J are:
• Solid Shaft of radius R, diameter D :
R
R πR 4 πD 4
J =∫ 2πρ3 dρ = = (4.11)
0 2 32
D
• Hollow Shaft with Inner Radius Ri and Outer radius Ro:
Ri Ro
J =∫
Ro
2πρ3 dρ =
(
π Ro4 − Ri4
=
) (
π Do4 − Di4 ) (4.12)
Ri 2 32
Di
Do

• Thin Walled Tube with t < R/10:
t Rm = mean radius of the thin walled tube

Rm ≈ Ro ≈ Ri and t = Ro − Ri
Rm From Eq. (4.12),
π π
( )( )( ) ( )
J = Ro2 + Ri2 Ro + Ri Ro − Ri = 2 Rm2 (2 Rm )t = 2πRm3 t
2 2
J = 2πRm t
3
(4.13)
Engineer's Theory of Torsion ( ETT )

When we equate Eq. (4.9) with the shear stress term, Eq. (4.6), gives that:
T dϕ τ
=G = (4.14)
J dx ρ
T
or τ= ρ (4.15)
J
where
T = the internal torque at the analyzed cross-section;
j's
J = the shaft’s polar moment of inertia;
G = shear modulus of elasticity for the material
ρ = radial distance from the axis (centre).
ra
This is called Engineer's Theory of Torsion ( ETT ).
The Maximum Shear Stress

vi
The maximum shear stress can be computed as
T
τ max = Ro (4.16)
J
ith
where Ro is the radius of the outer surface of circular shaft.

Pr
Example 4.1 Compare the weight of equal lengths of hollow and solid shafts to transmit a
torque T for the same maximum shear stress. For hollow shaft, the inner and outer diameters
have relationship Di = 2/3 Do = 2/3 DH.
From ETT (Eq. 4.15):
T J 2J
= cons tan t = =
τ ρ D
For Solid Shaft:
πDS4
J Solid =
32
For Hollow Shaft:
π   π 65 4
4
2
J Hollow =  DH4 −  DH   = × DH
32  3   32 81
If we then equate the RHS of the above equation (due to the same T and τ), we get:
 2J   2J  2 J Solid 2 J Hollow
  =   , i.e. =
 DS  Solid  DH  Hollow DS DH
Substituting for the J's we get:

πDS3 π 65
= × DH3
16 16 81
which gives that:
DH 3 81
= = 1.075 or DH = 1.075 DS, an increase in size by 7.5%
DS 65
Comparing the weight ratio:
π  2  2  
2
DH −  DH 
VH AH 4  3  
= = = 0.642
VS AS π 2
DS
4
which is a reduction in weight of 35.8 % if the hollowed shaft is used!
4.3 ANGLE OF TWIST (4th: 198-212;5th:198-212)
j's
The maximum shear stress is one of major design constraints in relation to strength of shaft.
However, sometime the design may depend on restricting the amount of rotation or twist
when the shaft is subjected to a torque.
ra
Angle of Twist for General Cases
In this section, we will develop a formula for determining the angle of twist φ (phi) of one end
of a shaft with respect to its other end as shown in Fig. 4.8.
vi
dx
From Eq. (4.4), we have dϕ = γ
ρ
ith
y
x dx
J(x)
Pr
B φ
z C x
T3 T2 T1
Fig. 4.8 Rotational shaft under general loading conditions
According to Hooke’s law, γ = τ G , and substituting Eq. (4.15) i.e. τ = T ( x )ρ J ( x ) , we have

T (x )
dϕ = dx
GJ ( x )
Integrating over the entire length L of the shaft, we obtain the angle of twist for the whole
shaft as
L T (x )
ϕ=∫ dx (4.17)
o GJ ( x )
where

φ = the angle of twist of one end with respect to other end, measured in radian
T(x) = the internal torque at arbitrary position x, found from the method of sections and
equation of moment equilibrium
J(x) = the shaft’s polar moment of inertia expressed as a function of position x.
G = shear modulus of elasticity for the material
Single Constant Torque and Uniform Cross-Section Area

T
GJ φ
T
L
Fig. 4.9 Uniform shaft under a constant torque T
Usually in engineering practice, the material is homogeneous and the shaft’s cross-sectional
j's
area and applied torque are constant as shown in Fig. 4.9. Eq. (4.17) becomes
TL
ϕ= (4.18)
GJ
ra
Multiple Torque and Cross-Section Areas
vi
If the shaft is subjected to several different torques, or consists of a number of different the
cross-sectional areas or shear moduli, Eq. (4.18) can be applied to each segment of the shaft
where these quantities are all constant.
ith
TL
ϕ=∑ i i (4.19)
i Gi J i
Pr
Sign Convention of Internal Torque
+T
+φ -T
-φ
-T
+T
Fig. 4.10 Sign conventions for torque and angle of twist
In order to apply the above equation (Eq. (4.18)), we must develop a sign convention for
internal torque and angle of twist of one end with respect to the other end. To do this, we will
use the right-hand rule, whereby both the torque and angle will be positive, provided the
thumb is directed outward from the shaft when the fingers curl to give the tendency for
rotation, as illustrated in Fig. 4.10.

Example 4.2: a) Determine the maximum shear stress and rate of twist of the given shaft if a
T = 10kNm torque is applied to it; b) if the length of the shaft is 15 m, how much would it
rotate by? c) if restrict the maximum shear stress level within τallow = 60MPa, what is the
maximum torque that the shaft can transmit? Let G = 81GPa, D = 75 mm.
T=10kNm
φ
D=75mm
L=15m
a) The maximum shear stress:

πD 4 π(0.075)
4
J= = = 3.106 × 10 −6 m 4
32 32
TR
τ max = o = =
( )
T (D / 2 ) 10 × 10 3 × (0.075 / 2 )
= 120.7 MPa
J J (
3.106 × 10 −6 )
j's
The rate of twist:
dϕ T
= =
(
10 × 10 3 ) = 0.03974rad / m
dx GJ ( )(
81 × 10 9 × 3.1063 × 10 −6 )
ra
which equates to :
dϕ 180 × (0.03974)
= = 2.277 o / m
dx π
vi
b) If the shaft is 15 m long, the angle of rotation at the free end is
dϕ
ϕ= L = 2.277° × 15 = 34.155°
ith
dx
or directly from Eq. (4.18)
ϕ=
TL
=
( )
10 × 10 3 × 15
= 0.596rad = 34.155°(= 0.596 × 180 / π )
( )( )
Pr
GJ 81 × 10 9 × 3.1063 × 10 −6
TRo
c) The maximum shear stress must be less than the allowable stress; τ max = ≤ τ allow , i.e.
J
Tmax ≤
Jτ allow
=
( )(
3.106 × 10 −6 × 60 × 10 6
= 4.97kN ⋅ m
)
Ro 0.0375
4.4 STATICALLY INDETERMINATE TORQUE-LOADED MEMBERS

(4th: 213-220; 5th:213-220)
Compound Shafts
A compound shaft is one made from more than one material property. The aim here is to
determine how much of the applied torque is carried by each material.
Equilibrium:
+ ∑ M x = 0 = T − T1 − T2 = 0 (4.19)
Compatibility:
ϕ = ϕ1 = ϕ 2 (4.20)

z
T1 T2 y
J1 G1
G2
Fully J2
bonded
Distribution of Distribution of
Shear Stress (G1<G2) Shear Strain
Substituting Eq. (4.18) into Eq. (4.20) and equating with Eq. (4.19), we can find T1 and T2,
hence the rate of twist and shear stresses carried by each material.
TL TL GJ
ϕ= 1 = 2 , i.e T1 = 1 1 T2
G1 J 1 G2 J 2 G2 J 2
G2 J 2 G1 J 1
T2 = T ; T1 = T (4.21)
G1 J 1 + G2 J 2 G1 J 1 + G2 J 2
G2 ρ G1ρ
τ2 = T ; τ1 = T (4.22)
(G1 J 1 + G2 J 2 ) (G1 J 1 + G2 J 2 )
j's
Indeterminate Shafts
ra
A
T
C B
T
vi
TA A
LAC Ro
LBC C
L B
TB
ith
Global Equilibrium (for ground reactions):

+ ∑M x = 0 = T − T A − TB = 0 (4.23)
Since only one equilibrium equation is relevant and there are two unknowns, this problem is
Pr
statically indeterminate. However, the angle of twist of one end of the shaft with respect to
other end is zero. We can give compatibility condition as
TL
Compatibility: ϕ A B = ∑ i i = 0
i Gi J i
and note that the internal torque in segment CB is negative by using the right-hand rule.
T A L AC (− TB )LBC
+ =0 (4.24)
JG JG
From Eqs. (4.22) and Eq. (4.23), we have
L  L 
T A =  BC T ; TB =  AC T (4.25)
 L   L 
Therefore,
L R L R
τ max
A = BC o T ; τ max B = AC o T (4.26)
LJ LJ

COMPARISON OF TORSIONAL SHAFTS WITH AXIALLY LOADED BARS/RODS
Axial Loaded Bar/Rod Torsional Shaft
(Chapters 1, 2&3) (Chapters 4)
Load Type Force F (N) Torque T (Nm)
Tension or Compression Right-Hand Rule
+F +T
+φ
+ Tension +F +
Sign Convention
-F +T
-T
- Compre
ssion -F
- -φ
-T
2
Geometric Property A – Area (m ) J – Polar Moment of Inertia(m4)
Material Property E – Young’s Modulus (Pa) G – Shear Modulus (Pa)
F T
Normal stress: σ avg = (Pa) Shear stress: τ = ρ (Pa)
A J
j's
Distribution of
Uniform shear stress
Stress Distribution
T
ra
Strain ∆L Shear strain γ - radian
Normal strain: ε = (ms)
L
vi
Hooke’s Law σ = Eε τ = Gγ
Deflection (m) Angle of Twist (Radian)
(Elongation/Contraction) T (x )
L
ith
General: ϕ = ∫ dx
F (x ) G ( x )J ( x )
L
General: δ = ∫ dx 0
E ( x )A(x ) TL
Deformation 0
Single uniform: ϕ =
FL GJ
Pr
Single uniform: δ =
EA TL
Fi Li Multi-segments: ϕ = ∑ i i
Multi-segments: δ = ∑ i Gi J i
i Ei Ai
1 1
Work W = P∆ p W = Tϕ T
2 2
2
F L T 2L
Strain Energy U = ∑ i i (for a Truss) U = ∑ i i (Multi-segments)
i 2 E i Ai i 2Gi J i
Work-Strain 2U 2U
Energy Method ∆P = ϕT =
P T
 ∂F  L   ∂T  L 
∆ P = ∑ (Fi ) i  i  ϕ T = ∑ (Ti ) i  i 
Castigilinao’s i  ∂P  Ei Ai  i  ∂T  Gi J i 
Method By introducing a virtual force Q: By introducing a virtual torque S:
 ∂F  L   ∂T  L 
∆Q = ∑ (Fi ) i  i  ϕ S = ∑ (Ti ) i  i 
i  ∂Q  Ei Ai  Q =0 i  ∂S  Ei Ai  S =0

j's
ra
vi
ith
Pr

Chapter 5 Bending Moments and Shear Force
Diagrams for Beams
In addition to axially loaded bars/rods (e.g. truss) and torsional shafts, the structural members may
experience some loads perpendicular to the axis of the beam and will case only shear and bending
in the beam. The current chapter together with Chapters 6 to 8 will focus on such an issue.
5.0 SHEAR FORCE AND BENDING MOMENTS DIAGRAMS FOR BEAMS

A Shear Force Diagram (SFD)indicates how a force applied perpendicular to the axis (i.e.
parallel to cross section) of a beam is transmitted along the length of that beam. A Bending
Moment Diagram (BMD) will show how the applied loads to a beam create a moment
variation along the length of the beam. These diagrams are used to determine the normal and
shear stresses as well as deflection and slopes in the following chapters.
5.1 BEAM SIGN CONVENTION (SI&4th:256-257; 5th:256-257)

At any point along its length, a beam can transmit a bending moment M(x) and a shear force
V(x). If a loaded beam is cut, the definitions of a positive distributed load, shear force and
j's
positive bending moment are as Fig. 5.1 below:
Positive internal
bending moment
ra
Positive distributed load Positive internal shear force
vi
Fig. 5.1 Beam shear force and bending moment sign convention
ith
Where distributed load acts downward on the beam; internal shear force causes a clockwise
rotation of the beam segment on which it acts; and the internal moment causes compression
in the top fibers of the segment, or to bend the segment so that it holds water.
Pr
5.2 RELATIONSHIP BETWEEN BEAM LOADINGS (SI&4th:264-268; 5th:264-

268)
A beam (Fig. 5.2) is loaded with vertical forces Fi, bending moments Mi and distributed loads
w(x).
F1 F2 F.B.D. of element dx
w(x) w(x)
M(x) dM(x)
M(x)+ dx
dx
M1 M2 dV(x)
V(x) V(x)+ dx
dx dx
x dx
Fig. 5.2 Transversely loaded beam and free body diagram of element dx

Look at the FBD of an elemental length dx of the above loaded beam (Fig. 5.2). As it has an
infinitesimal length, the distributed load can be considered as a Uniformly Distributed Load
(UDL) with constant magnitude w(x) over the differential length dx.
It is now necessary to equate the equilibrium of the element. Starting with vertical equilibrium
 dV (x ) 
+↑ ∑F y = 0 = V ( x ) − w(x )dx − V (x ) +
 dx
dx  = 0

(5.1)
Dividing by dx in the limit as dx→0,

dV (x )
= − w(x ) (5.2)
dx
Taking moments about the right hand edge of the element:
dx  dM ( x ) 
+ ∑ M R .H .Edge = 0 = − M ( x ) − V ( x )dx + w( x )dx +  M (x ) + dx  = 0 (5.3)
2  dx 
Dividing by dx in the limit as dx→0,
dM ( x )
= V (x ) (5.4)
dx
Eqs. (5.2) & (5.3) are important when we have found one and want to determine the others.
j's
5.3 BENDING MOMENT AND SHEAR FORCE EQUATIONS
ra
Introductionary Example - Simply Supported Beam
By using the free body diagram technique, the bending moment and shear force distributions
vi
can be calculated along the length of the beam. Let’s take a simply supported beam, Fig. (5.3),
as an example to shown the solutions:
F.B.D. (global equilibrium) F.B.D. (method of section I-I)
ith
P
a
I II M(x)
A B A o
Pr
I II
V(x)
L x
RAY=(1-a/L)P RBY=Pa/L RAY=(1-a/L)P
Fig. 5.3 FBD of beam cut before force P
Step A: Cut beam just before the force P (i.e. Section I-I), and draw a free body diagram
including the unknown shear force and bending moment as in Fig. 5.3.
Take moments about the right hand end (O):

 a  a
+ ∑ M o = 0 = − P1 −  x + M ( x ) = 0 → M ( x ) = P 1 −  x (5.5)
 L  L
To determine the shear force, use Eq. (5.4), giving that:
dM ( x )  a
V (x ) = = 1 −  P (5.6)
dx  L
To verify Eq. (5.6), equate vertical equilibrium:

 a  a
+↑ ∑F y = P  1 −  − V ( x ) = 0 → V ( x ) = P 1 − 
 L  L
which is the same equation as (5.6). These then, are the equations for the bending moment and
shear force variation in the range of 0 ≤ x ≤ a .
To find out the rest of the bending moment and shear force distributions, it is necessary to
now carry out a similar analysis, but cutting the beam just before the end (Section II-II).
Step B: Cut beam just before the right hand end (RHE)
F.B.D. (Section II-II)
P
a
II M(x)
A o
II
V(x)
x
RAY=( 1-a/L)P
Fig. 5.4 FBD of beam cut before the right hand end
Equate moments about the right side:
∑M
 a
j's
= 0 = − P 1 −  x + P ( x − a ) + M ( x ) = 0
ra
+ 0
 L
giving:
 a a
M (x ) = P1 −  x − P(x − a ) = − Px + Pa (5.7)
vi
 L L
and using Eq. (5.4), the shear force equation is :
dM ( x )
ith
a
V (x ) = =− P (5.8)
dx L
These expressions for the bending moment and shear force can now be plotted against x to
produce the Shear Force and Bending Moment Diagrams as Fig. 5.5:
Pr

P
a
Loading Diagram
(1-a/L)P L Pa/L
V(x)
(1-a/L)P
Shear Force Diagram +ve

x
-ve
-a/LP
M(x) (1-a/L)Pa
j's
+ve
Bending Moment Diagram
x
Fig. 5.5 Shear Force and Bending Moment Diagrams for simply supported beam
ra
Macauley's Notation (4th:590-599; 5th:590-599)
The two sets of equations for V(x) and M(x), Eqs. (5.5), (5.6), (5.7) and (5.8), can be condensed
vi
to just one set of equations if we use a special type of notation called Macauley's Notation.
The above equations would look like this (to be derived in Example 5.0)
 a 0
ith
V ( x ) = P 1 −  x − P x − a
0
(5.9)
 L
 a 1
M ( x ) = P 1 −  x − P x − a
1
(5.10)
 L
Pr
Where the notation has the following meaning:

 0 for x < a
x−a
n
= (n ≥ 0) (5.11)
(x − a ) for x ≥ a
n
when differentiating:
n x − a n −1
for n ≥ 1
∂ x−a
n
 0
= x−a for n = 1 (5.12)
∂x 

0 for n = 0
Remarks
To derive the bending moment equation by using Macauley's notation, you may need to do
the following:
1) Determine the ground reactions from global equilibrium;
2) Cut the beam just before the right hand end;

3) Equate the cut FBD to equilibrium about the right hand end;
4) All length terms in the bending moment/shear force equations MUST be written using
Macauley's notation;
5) Always indicate the powers, even if they are 0 or 1.
Example 5.0: As in the introductory example, determine the shear force and bending moment
equations and plot them for a simply-supported beam as in the introductory example.
Step 1: Determine the ground reactions;

F.B.D. (global equilibrium)
P
a
I
A B
I
j's
L
RAY=(1-a/L)P RBY=Pa/L
We have
ra
RAY = (1-a/L)P and RBY = a/LP
Step 2: Draw FBD of beam cut just before the RHS (Section I-I).
vi
F.B.D. (Section I-I)
P
a
ith
I M(x)
A o
I
V(x)
Pr
x
RAY=( 1-a/L)P
Step 3: Equilibrium for FBD of beam cut just before the RHS (Section I-I).
Take moments about RHS:
 a 1
+ ∑ M O = 0 = − P1 −  x + P x − a + M ( x ) = 0
1
 L
 a 1
M ( x ) = P 1 −  x − P x − a
1
 L
and differentiating w.r.t. 'x', as Eq. (5.4), gives the shear force equation as:
dM ( x )  a 0
V (x ) =
0
= P 1 −  x − P x − a
dx  L
Step 4: Plotting the Shear Force and Bending Moment Diagrams

According to M(x) and V(x) to depict the diagrams, Look at the equations segment by segment
When 0 ≤ x ≤ a
 a 1  a  a 0  a
M ( x ) = P1 − ( x ) − P × 0 = P1 −  x and V ( x ) = P1 − ( x ) − P × 0 = P1 − 
 L  L  L  L
To plot this segment in the diagram, firstly look at the boundary points as x = 0, M ( x) = 0
and x = a , M ( x) = Pa(1 − a / L ) . Draw two points and then connect them because the equation
gives a line. Likewise, one can plot Shear Force Diagram in this region.
When a ≤ x ≤ L
 a 1  a a
M ( x ) = P1 − ( x ) − P × ( x − a ) = P1 −  x − Px + Pa = Pa − P x
1
 L  L L
 a 0  a a
and V ( x ) = P1 − ( x ) − P × ( x − a ) = P1 −  − P = − P
0
 L  L L
Remarks: Please draw global FBD of the beam firstly and follow by its Sear Force and Bending
Moment Diagrams. The reason for doing this is that when you get sufficient experience, you may
be able to directly plot the Shear Force Diagram by observing the external forces as well as plot
j's
Bending Moment Diagrams by observing the Shear Force Diagram. Nevertheless you MUST still
work out and indicate the locations and values (including +ve or –ve) at all turning points in the
diagrams in detail.
ra
It is also interesting to note that concentrated forces (e.g. reaction forces and external forces)
correspond to inclined line in BMD and horizontal line in SFD.
P
vi
a
Loading Diagram
ith
(1-a/L)P L Pa/L
Pr
V(x)
(1-a/L)P

x
-ve
-a/LP
M(x) (1-a/L)Pa
+ve
x
Example 5.1: Determine the shear force and bending moment equations and plot them for a
simply-supported beam loaded with a UDL.
From global equilibrium the ground reaction forces can be found to be both equal to wL/2 as,

w I
A B
I
L
wL/2 wL/2
L
+ ∑ M = 0 = − R L + (wL ) 2  = 0 →∴ R = wL / 2 (+ upwards)
B AY AY
+ ↑ ∑ F = 0 = R + R − wL = 0 →∴ R = wL / 2 (+ upwards)
y AY BY BY
j's
F.B.D. (Section I-I) Rw

ra
x/2
w
M(x)
A o
vi
V(x)
x
ith
RAY=wL/2
As far as V(x) and M(x) are concerned the UDL can be temporarily replaced by its resultant Rw
Pr
(=wx) applied at the centroid of the UDL distribution in the moment equilibrium equation. So if
we take moments about the RHS of the beam we get:
( )
1 1
x x
∑ M O = 0 = − R AY x + Rw 2 + M (x ) = −(wL / 2) x + w x 2 + M (x ) = 0
1 1 1
+
wL 1 w 2
∴ M (x ) = x − x
2 2
and differentiating w.r.t. 'x', as Eq. (5.4), gives :
dM ( x ) wL
V (x ) =
0 1
= x −w x
dx 2

According to M(x) and V(x) to depict the diagrams

w
A B
Loading Diagram
wL/2 L wL/2
V(x)
wL/2

x
-ve
-wL/2
M(x) wL2/8
Parabola
+ve
j's
x
It is worth pointing out that one should not completely replace such a UDL by its
corresponding resultant concentrated force Rw (=wx) in the beginning of the solution. There is
ra
a significant difference of the Shear Force and Bending Moment Diagrams between a
concentrated force (Example 5.0) and a UDL (Example 5.1). It is also interesting to note that
the UDL corresponds to an inclined line in the Shear Force Diagram and a quadratic curve
vi
(parabola) in the Bending Moment Diagrams.
ith
Pr

beam loaded with a UDL between A and B and two concentrated forces at C and E.
UDL=w=1kN/m 20kN 10kN

B C D I
A E
I
RAY 10m 5m 5m 10m
RDY

+ ∑M A = 0 = −1 × 10 × 5 − 20 × 15 + R DY × 20 − 10 × 30 = 0 →∴ R DY = 32.5kN
+↑∑F y = 0 = R AY − 1 × 10 − 20 + R DY − 10 = 0 →∴ R AY = 7.5kN
j's
Note: The only problem with Macauley's Notation is that it does not work when a UDL stops.
It however does work for a UDL which starts anywhere along a beam and continues to the
end. The problem can be corrected by applying a UDL of equal magnitude but opposite sense
ra
where the first UDL ends.
F.B.D. (Section I-I) and application of equivalent UDL

vi
w=1kN/m 20kN
M(x)
ith
A o
32.5kN V(x)
Pr
7.5kN
x

2 2
x x − 10
∑M + 20 x − 15 − 32.5 x − 20 + M (x )
1 1 1
+ = 0 = −7.5 x +1 −1
O
2 2
1 1
M ( x ) = 7.5 x − x
1 2 2 1 1
+ x − 10 − 20 x − 15 + 32.5 x − 20
2 2
dM ( x )
V (x ) =
0 1 1 0 0
= 7.5 x − x + x − 10 − 20 x − 15 + 32.5 x − 20
dx

UDL=w=1kN/m 20kN 10kN

B C D
A E
Loading
Diagram 10m 5m 5m 10m
7.5kN 32.5kN
V(x) kN
10
Shear 7.5
+ve +ve
Force x
Diagram
-2.5 -ve
-22.5
M(x) kNm
quadratic 28.125
25
j's
+ve 12.5
x
Bending
Moment
ra
Diagram -ve
vi
-100
ith
Again, the UDL segment corresponds to an inclined line in SFD and a quadratic curve in
BMD.
Pr
cantilever beam loaded with a moment MB = 40kNm and a force F= 10kN.
Global F.B.D.
MB=40kNm 10kN
A B I
MA + C
I
RAY
4m 1.5m

The cantilever beam is fully clamped in the left hand end A as shown. The ground reaction for
this point should have reaction force RAY and reaction moment MA.. So the global equilibrium
is given as
+ ↑ ∑ Fy = 0 = R AY + 10 = 0 →∴ R AY = −10kN (- downwards)
+ ∑M A = 0 = M A + 40 + 10 × 5.5 = 0 → M A = −95kNm (- clockwise)

95kNm MB=40kNm
O M(x)
B
A +
4m V(x)
10kN
x
+ ∑ M O = 0 = 10 x − 95 x + 40 x − 4 + M ( x )
1 0 0
M ( x ) = −10 x
1 0 0
+ 95 x − 40 x − 4
j's
dM ( x )
V (x ) =
0 0
= −10 x + 0 − 0 = −10 x
dx
ra
MB=40kNm 10kN
95kNm
B
vi
Loading A +
Diagram C
ith
10kN
V(x) kN
Shear
+ve
Pr
Force x
Diagram -ve
-10
M(x) kNm
95
Bending 55
Moment
Diagram
15
x
It is interesting to observe that due to concentrated bending moments MA (reaction moment)

and MB (external moment), there is respectively a sudden leap and drop in the Bending
Moment Diagram. In addition, such concentrated moments do not affect the Shear Force
Diagram (Note that the drop at point A (the left end) in SFD is due to the concentrated
reaction force RAY. In fact, both MA and MB do not appear in shear force equation V(x) at all).

Chapter 6 Bending
6.0 PURE BENDING
Bending of beams is the next form of internal stress after uniform tensile/compressive stress
and torsional shear stress. It is the most significant type of internal load transfer. What we
want to derive in the present chapter is a relationship between the normal stresses in the beam
cross-section and the applied bending moments. To achieve this we need to do three things:
• Determine some assumptions to simplify the analysis
• Determine the geometric relationships of the deformed beam
• Determine the load equilibrium conditions.
6.1 ASSUMPTIONS (SI&4th : 282-283; 5th : 282-283)

In order to be able to simplify the analysis it is best to define some assumptions about the
beam, how it will deform.
M M
j's
Fig. 6.1 Beam before and after a positive bending moment is applied
ra
• Transverse planes before bending remain transverse after bending, Fig. 6.1, ie. no
warping.
vi
• Beam material is homogeneous and isotropic and obeys Hook's law with E the same in
tension or compression.
• The beam is straight and has constant or slightly tapered cross section.
ith
• Loads do not cause twisting or buckling. This is satisfied if the loading plane
coincides with the section’s symmetry axis.
• Applied load is pure bending moment (recall the B.M.D. in Assignment Q.4.3).
Pr
The definition for beams with applied positive and negative bending moments is as Fig. 6.2:
+ve M
-ve M
Fig. 6.2 Diagrams showing beam experiencing positive (left)
and negative (right) bending moments
6.2 GEOMETRIC RELATIONSHIPS OF BEAM (SI&4th:283-285; 5th : 283-285)

To start the analysis we first must establish the nature of the deformation by observing how a
beam deflects when a bending moment is applied to it.
When a beam is subjected to a pure bending moment, it will deform into a curved shape and
this shape is the arc of a circle with a very large radius compared to the size of the beam. Let’s
now look at a segment of beam before and after the application of a positive bending moment:

M Compression M
Tension
Fig. 6.3 Beam before and after a positive bending moment is applied
indicating regions of +ve and –ve stresses
As you can see from Fig. 6.3, the fibres on the top surface are experiencing a compressive
stresses, and those on the bottom a tensile stress. What this means is that at some point
between these two surfaces, there must be a plane where the normal stresses and strains are
ZERO. We call this plane the Neutral Plane (N.P.) or Neutral Axis (N.A.).
Look now again at a small segment of beam before the application of a bending moment.
O
j's
dθ
R
ra
M M
vi
i j .
y N.A. i1 j1 N.A
m n y
m1 n1
ith
Fig. 6.4 Undeflected segment of beam Fig. 6.5 Beam deformed by positive bending moment
Mark a longitudinal section with a distance y from the Neutral Axis as ij, and another section
Pr
on the Neutral Axis as mn, as in Fig. 6.4. Initially, these sections are of equal length as they
define the length between two transverse planes, i.e. ij = mn.
A pure positive bending moment M is then applied to the beam which makes the above
section deform as Fig.6.5, where the applied bending moment causes the segment ij and mn to
deform into concentric arcs i1j1 and m1n1 with an angle dθ between the segments i1m1 and j1n1.
The distance between these two arcs is still y.
Let R = Radius of curvature of the Neutral Plane, then the strain of segment i1j1 is defined as
length i1j1 minus the original length ij over the original length ij, i.e.:
i j − ij
εx = 1 1 (6.1)
ij
Now length mn and ij are defined as:
ij = mn = m1 n1 = Rdθ
and length i1j1 is defined as:
i1 j1 = (R − y )dθ
so the strain becomes:
εx =
(R − y )dθ − Rdθ = − y (6.2)
Rdθ R

which indicates that the strain is linearly varying with y. And since stress is strain times
Young's Modulus E then the stress can be defined by the following equation and is also
linearly varying with y.
Ey
σ x = Eε x = − (6.3)
R
In effect, this equation has been derived using geometrical compatibility conditions; we need
also to consider the equilibrium conditions of the beam cross-section.
6.3 APPLIED LOAD EQUILIBRIUM (SI&4th:286-294; 5th:286-294)

The stress equation (6.3) defined as above indicates that the normal stresses induces by the
bending moment vary linearly through the depth of the beam. Look now at a drawing of the
right hand end of the beam showing the normal stress distribution and applied bending
moment as shown in Fig. 6.6.
dA
y y
N.A. N.A.
M
j's
Cross-section
ra
Fig. 6.6 Right hand end of beam showing applied bending moment M
and normal stress distribution
Let dFx = σxdA be the component of force acting on the element of area dA. We now use
vi
equilibrium conditions on the stresses generated on the right hand side of the beam:
ith
Force Equilibrium
∑ Fx = ∫ dFx = ∫ σ x dA = 0
A A
Substituting the above equation (6.3) for stress gives:
Pr
 Ey  E
∫ A  − R dA = 0 , or: R ∫ A ydA = 0 (6.4)
For materials with E constant the condition,

∫ ydA = 0
A
(6.5)
gives the origin for the “y-axis” on the centroid of the section, i.e. the location of the Neutral
Plane. We will look at what this means shortly
Moment Equilibrium
When equating moment equilibrium we have that the applied moment M must be equal to the
moment generated internally by the normal stress caused by the external moment, such that:
∫ dM = ∫ y ⋅(dFx ) = ∫ yσ x dA = M
A A A
(6.6)
Substituting for the stress as in equation (6.3) gives :
Ey 2 E 2
M = ∫ yσ x dA = ∫ − dA = − ∫ y dA (6.7)
A A R R A
We now define the term
2
I = ∫ y dA (6.8)
A

as the Second Moment of Area or Moment of Inertia of the beam about the Neutral Axis
(which was actually defined by Eq. (6.5)). It is a measure of the stiffness of the cross
sectional shape from a geometric point of view, without considering the material properties.
Note that the unit of second moment of area is m4. Now substituting I into the equation for
stress gives (from Eqs. (6.3) and (6.7))
σx M E
=− =− (6.9)
y I R
which is called Engineer's Theory of Bending (ETB). The standard form of writing and
using this equation is:
My
σx = − (6.10)
I
So, if you know the applied bending moment, the location of the centroid and the section’s
second moment of area you can then find the stresses along the depth of the beam’s section.
Let’s now define the location of the centroid and an easier equation to use than Eq. (6.5).
6.4 DETERMINATION OF NEUTRAL AXIS (SI&4th: 775-777; 5th: 775-777)

Look at an arbitrary symmetrical beam cross section, Fig. 6.7:
j's
dA
y N.A. to be
determined
ra
s
S
vi
Reference Axis
Fig. 6.7 Determination of Neutral Axis of an arbitrary cross section
ith
Centroid of an Arbitrary Area In order to find the centroid it is often best to find it in
reference to the bottom of the beam cross section. If we do this and because the centroid
equation (6.5) is integrated about the neutral plane we firstly need to change the axis from y to s.
Pr
Changing beam axis from y (distance away from Neutral Plane) to s (distance away from
bottom of beam) has that:
y=s−S
Substituting into Eq. (6.5) gives:
A
( )
∫ s − S dA = ∫ sdA − ∫ S dA = 0 A A
but since S is the distance to the centroid, it is a constant and can be taken out of the integral
equation, giving :
S ∫ dA = ∫ sdA
A A
and dividing by total area gives:
S= A
∫ sdA (6.11)
∫ dA A
Composite Areas However as most engineering beams are made of several simpler regular
shapes, for which you know the areas and the centroids of these areas, then this equation can
be used in finite summation form instead of integral form as in (6.11).

yi dAi
Local N.A. of Ai
Global N.A. of entire cross section
si
S
Reference Axis
Fig. 6.8 Parallel axis method for composite area
From Fig.6.8:
∫ A
sdA = ∑ ∫ (s
Ai i + y i )dAi = ∑  ∫ Ai
s i dAi + ∫
Ai
y i dAi  =

∑  s ∫
i
Ai
dAi + 0  =

∑s A
i i
∴S =
∑ s i Ai (6.12)
∑ Ai
where si represents the reference coordinate for the centroid of each part and Ai is its area.
We can now use Eqs. (6.8), (6.10) and (6.12) to determine the stresses in a beam under an
applied bending moment.
j's
Example 6.1: Determine the second moment of area I for the following T-shaped cross section.
40mm
A2=40×5
5mm 2
ra
N.A.
40mm s2=37.5mm
vi
S A1=35×5
1 s1=17.5mm
ith
For all these type of problems it is best to follow the following methodology when solving them:
Step 1: Determine the location of the Neutral Plane
Pr
Using Eq. (6.12), do this by firstly subdividing the beam’s cross section into regular
geometric shapes. In this example the beam can be divided into two rectangular sections 1 and
2 as shown. Substituting the areas of each of the rectangles that make this shape as well as the
distances to their respective centroids gives:
s A +s A 17.5 × (35 × 5) + 37.5 × (40 × 5)
S = 1 1 2 2 = = 28.167 mm
A1 + A2 (35 × 5) + (40 × 5)
Step 2: Coordinate transformation Once the position of the centroid (neutral plane) for the
section has been found, re-draw the section with all new coordinates about the neutral axis:
11.833
y 2
6.833
N.A.
1
-28.167

Step 3: Determine the section’s Second Moment of Area
I = ∫ y 2 dA = ∫ y 2 dA + ∫ y 2 dA
A A1 A2
11.833 6.833
which for the section becomes I = ∫y 40dy + ∫y
2 2
5dy
6.883 − 28.167
I = 17837 + 37777 = 55613 mm4 or I = 55614×10-12 m4

(Note that, 1 mm =1×10-3 m, ∴(1 mm)4 = 1 mm4 = (1×10-3 m)4 = 1×10-12 m4)
Regular Shapes:
You can use the following standard solutions to the second moments of areas of regular
shapes. These results can also be found inside the Front Cover of your textbook.
a) Rectangle Sections
h
N.A. bh 3
I =∫
h 2
h y bdy =
2
h/2 −
2
12
b A = bh
j's
R
b) Circular Sections
N.A. πR 4 πD 4
I= =
4 64
ra
A = πR 2
c) Semi-Circular Sections
vi
N.A.
I = 0.110 R 4
R
4R/3π A = πR 2 / 2
ith
2R
t d) Thin Tubular Sections for t<R/10

Pr
I = πRm3 t
Rm N.A.
A = 2πR m t
e) Half-Thin Tubular Sections for t<R/10

N.A. I = 0.095πRm3 t
Rm A = πRm t
2Rm/π
2Rm
f) Triangular Sections
bh 3
h I=
N.A. 36
h/3 1
A = bh
b 2

Having said that you can use the above standard solutions to determine the second moment of
area, it is fine if the centroid of the section you are analyzing lies on the neutral plane of the
whole cross section. As you have seen from the examples just done (Example 6.1), the
individual centroids of each of the shapes that the cross section was divided in, did not lie at
the centroid of the cross section. For this reason, and to save the trouble of always having to
determine an integral to find the second moment of area, we need to look at the derivation of
the Parallel Axis theorem for evaluating the second moment of area.
6.5 PARALLEL AXIS THEOREM (SI&4th: 778-781; 5th: 778-781)

If the second moment of an area is known about its centroid, we can use the parallel axis
theorem to find I about a corresponding parallel axis, Fig. 6.8.
dA
y1
local N.A.
y
y
N.A. of entire cross section (global N.A.)
j's
Fig. 6.8 Known section a distance y above neutral axis of section
ra
The equation for the second moment of area can be computed by the integral as:
2
I = ∫ y dA (6.13)
A
vi
But the distance y about the global Neutral Axis of the entire cross section is given by:
y = y1 + y
Substituting this gives:
ith
( y1 + y )
2
I =∫ dA = ∫ y12 dA + 2 y ∫ y1 dA + y 2 ∫ dA
A A A A
The first term is the local second moment of area, the second is zero because y1 passes
through the local centroid (Eq. (6.5)) and the third is the distance between the local and global
Pr
centroids squared times the area of this shape, such that:

I N .P . = I local + y 2 A (6.14)
For a cross-section consisted of a number of regular shapes, the total second moment of area
can be computed as
I N .P . = ∑ (I i local + y i2 Ai ) (6.15)
here y i is the distance between the global and local centroid (N.A.) of area Ai For the
previous example if Eq. (6.15) was used instead of Eq. (6.8), you would have the following:
Example 6.2: Several concentrated external bending moments are applied over a cantilever
beam as shown. The cross-sectional area is the same as in Example 6.1. Please use parallel axis
theorem to determine I, and then further to analyze the maximum normal stresses in the beam.
Global F.B.D.
MB=40Nm MC=30Nm MD=30Nm
MA A C I
B
+ + +D
I
RAY
2m 2m 1m

Step 1: Plot the bending moment diagram to determine the bending moment peaks
MA=100Nm MB=40Nm MC=30Nm
A M(x)
B C
+ + o
2m 2m V(x)
x
Equilibrium for FBD of beam cut just before RHS (Section I-I) and take moments about RHS:
+ ∑ M O = 0 = −100 x + 40 x − 2 + 30 x − 4 + M ( x ) = 0
0 0 0
M ( x ) = 100 x
0 0 0
− 40 x − 2 − 30 x − 4
M(x) Maximum Bending Moment

100Nm
60Nm
30Nm
x
0 2 4 5
From the bending moment diagram, the bending moments at all cross sections are positive
j's
and the peak occurs between sections A and B with Mmax = 100Nm.
Step 2: Determine the location of the Neutral Plane As in Example 6.1:

ra
s A +s A 17.5 × (35 × 5) + 37.5 × (40 × 5)
S = 1 1 2 2 = = 28.167 mm
A1 + A2 (35 × 5) + (40 × 5)
vi
Step 3: Determine the distances between the global and local N.A.
Maximum compressive stress
11.833 (under a positive bending
2
ith
moment M=100Nm)
y 9.333mm
N.A.
-10.667mm
Pr
1
-28.167
Maximum tensile stress
Step 4: Determine I using Parallel Axis Theorem, Eq. (6.15)

( ) (
I = I 1 + y12 A1 + I 2 + y 22 A2 )
 5 × 353   40 × 53 
I = + (− 10.667 )2 × (35 × 5) +  + (9.33)2 × (40 × 5) = 55614.6mm 4
 12   12 
   
or I = 55614×10-12 m4
Step 5 Compute the normal peak stresses

Substituting this value for I and the maximum bending moment into Eq. (6.10) gives:
M y 100 y
σ max = − max = −
I 55614 × 10 −12
and the maximum stresses occur at the furthest points (fibre) away from the neutral plane:

at y = 11.833 mm , σy = 0.011833 = -21.280 MPa, the compressive stress peak and
at y = -28.167 mm, σy = -0.028167 = 50.648 MPa, the tensile stress peak.
Remarks: For a non-symmetrical cross section (w.r.t. N.A.), it is worth noting that the
maximum compressive or tensile stresses may correspond to other bending moment peaks
than the maximum positive one. That is to say that it may be very necessary for an analyst to
carefully examine several sections with both positive and negative moment peaks.
6.6 PRINCIPLE OF SUPERPOSITION (SI&4th: 133; 5th:133)

Used to determine the stress or displacement at a point in a structural component which is
subjected to several types of loads.
The Principle of Superposition states that the resultant stress or displacement at a point can be
determined by finding the stress or displacement caused by each load separately on the
structure, and then add their contributions.
6.7 COMBINED LOADINGS (Bending and Tension) (SI&4th: 409; 5th 409)
j's
We now need to look at how to analyze a beam with both a compressive or tensile load and
a bending moment acting simultaneously. We call such analysis Combined Loadings.
Before we do this though we need to define the Principle of Superposition
ra
If a structural element is applied a tensile or compressive axial force and a bending moment
simultaneously, as Fig. 6.10, we can determine the resultant stresses from the loading
condition by using the principle of superposition as in the Table 6.1.
vi
M M
ith
P P
Fig. 6.10 Beam with applied bending moment M and tensile load P
Pr
Table 6.1 Superposition method to analyze the combined loads:

Stress Stresses
Stresses Produced by Each Load Individually
Distributions
Tensile
Load P P σP = P/A
M M
Bending σM = -My/I
Load
M M
Both Tensile
and Bending P P σM =P/A -My/I
Loads

a
Section a-a
20
N.A.
P P 40
10mm 10
30
a
90mm
M=Pe M=Pe
N.A. N.A.
e P P
P P
Fig. 6.11 Beam with an eccentric tensile load P
The combined loading cases can be created by applying an axial load on a beam a distance
away from the beam's the Neutral Plane/Axis as illustrated in Fig. 6.11. This loading can be
considered as a bending moment of magnitude equal to the applied force multiplied by its
distance from the Neutral Plane (i.e. Pe) and the tensile or compressive load P.
j's
Example 6.3 An open-link chain is obtained by bending low-carbon steel rods of 12-mm
diameter into the shape shown in the figure of this example. Knowing that the chain carries a
load of 800N, determine the maximum tensile and compressive stresses in the straight portion
ra
of a link.
Step 1: Determine the internal force and bending moment from section equilibrium of F.B.D.
vi
+ → ∑ Fx = 0 = − P + 800 = 0 ∴ P = 800 N
+ ∑M = 0 = P×e− M = 0 ∴ M = Pe = 800 × 0.015 = 12 Nm

ith
12mm e=15mm M
P
Pr
800N 800N o 800N

e=15mm
Step 2: Second moment of area

Area of cross section: A = πR 2 = 3.14159 × 0.006 2 = 113.1 × 10 −6 m 2
πR 4 π × 0.006 4
Second moment of area: I = = = 1.018 × 10 −9 m 4
4 4
(Note that if the cross section is not a regular shape, you may need to use the steps in
Examples 6.1 and 6.2 to determine I)
Step 3: Superposition of P and M

The stress distribution due to the centric tensile force P is uniform as
P 800
σa = = = 7.07 MPa
A 113.1 × 10 − 6
The stress distribution due to the bending moment M is linear with a maximum stress as
My 12 × 0.006
σm = − =± = ±70.7 MPa
I 1.018 × 10 −9

It is worth noting that the maximum bending normal stress σm is about 10 times of the average
tensile normal stress σa. This means that the eccentric load can cause a substantial change in
stress of the structural element.
Superposing these two distributions, the largest tensile and compressive stresses in the section
are found to be respectively
P My
σ= ± = 7.07 ± 70.70
A I
σ tensile = 77.8MPa (occurs in the inner fibre)
σ compressive = −63.6MPa (occurs in the outer fibre)
So far we have been looking at beams which are made up of a homogeneous material, this
is not always the case. You can have beams where the material properties vary through the
depth of the material. Such beams are called composite beams, and we are now going to
look at how we can modify them so that they can be analysed by the equations and
procedures that have already been derived.
6.8 COMPOSITE BEAMS (SI&4th: 315-321; 5th:315-321)
j's
Beams made of two or more different materials are referred to as composite beams. Such
beams can be made of wood with straps of steel at the top and bottom surfaces, or concrete
beams reinforced with steel. The reason for manufacturing such beams is to develop
ra
structures that can support loads more efficiently. Look at a beam made from 3 materials as
shown in Fig. 6.12:
vi
ith
A1E1
A2E2
Pr
A3E3
b
Fig. 6.12 Beam made from three materials of different Young’s Modulus of Elasticity
Equivalent Second Moment of Area

Going back to the definition of the Neutral Axis equation, (6.4), if E is not a constant, Eq.
(6.4) becomes:
1
R∫A
EydA = 0 (6.16)
but because the beam is made from 3 materials with different Young's Modulus, we can
divide this into 3 integrals:
∫ E1 ydA1 + ∫ E 2 ydA2 + ∫ E3 ydA3 = 0
A1 A2 A3
(6.17)
For this beam, the term dA is given by:
dA1 = dA2 = dA3 = bdy
If the cross section was made from the same material we would not have this problem. So
let’s pretend that the beam is made from material 1.
We can achieve this by dividing through by E1.
 E  E  E  
Then E1  ∫  1 b  ydy + ∫  2 b  ydy + ∫  3 b  ydy  = 0 (6.18)
 1  E1 
A A2 E A3 E
 1   1  
We define the terms:

E E2 E3
b3e1 = 1 b ; b2 e1 = b; b3e1 = b (6.19)
E1 E1 E1
as the Equivalent Sections of Material 1, where E1/E1, E2/E1 and E3/E1 are termed as material
transformation factors. What has been done is to change each section to property number 1 by
altering their widths. This beam made from an equivalent material 1 would look as Fig. 6.13:
By using the dimensions of the equivalent beam cross section, the position of the Neutral Axis can
be calculated in the way as if the beam was made from one material property, using Eq. (6.12).
b
b1e1
j's
b2e1
b3e1
ra
Fig. 6.13 Equivalent beam cross section
vi
The second moment of area can now be found for this equivalent beam cross section. Using
ith
Eq. (6.13) you have that:

I eq1 = ∫ y 2 b1e1 dy + ∫ y 2 b2e1 dy + ∫ y 2 b3e1 dy (6.20)
A1 A2 A3
Also you can compute the second moment of area by the parallel axis method as Eq. (6.15).
Pr
Actual Stresses
What we need to determine, however is the stresses through the composite beam, to do this
look at the strains through the depth.
Although the cross section has been converted to an equivalent section of a homogeneous
material, the strains in this section must be the same to those of the real material for this
assumption to work. This means that, say looking at the bottom fibre of the beam cross
section, the strains there must be the same in the real beam with material property 3 as in the
equivalent beam of material property 1. Such that:
σ1eq My 3
ε 3bottomFibre = ε 1bottomFibre = = (6.21)
E1 I eq1 E1
But as we are interested in determining the stress in material 3 then:
My  E  My My
σ 3 = E3 ε 3 = E3 =  3  = (6.22)
I eq1 E1  E1  I eq1 I eq 3
which gives that:

E 
I eq 3 = I eq1  1  (6.23)
 E3 
E 
Similarly I eq 2 = I eq1  1  (6.24)
 E2 
So the general term becomes
E 
I eqj = I eqi  i  (6.25)
E 
 j
This gives the stresses in the actual materials as
E   E  My      
σ 3 =  3 σ 3eq =  3  −  and σ 2 =  E 2 σ 2eq =  E 2  − My  (6.26)
  E   E  I 
 E1   E1  I eq1   1  1  eq1 
Example 6.4: Determine the maximum stresses in each material if a moment of 20 Nm is

applied to the composite beam shown, where Eal = ES / 3 and EW = ES / 10:
j'sWood 50mm
ra
Aluminium 50mm
vi
Steel 50mm
ith
100mm
Step 1: Transform this to an equivalent section for one of the three materials,
In this case we select steel.
bweqs=Ew / Es × b=10mm
Pr
50mm
bAeqs=Ea / Es × b=33.33mm
50mm
bSeqs=Es / Es × b=100mm
50mm
Steel
100mm
Step 2: Determine the position of Neutral Plane
S =
∑ si Ai = 25 × (100 × 50) + 75 × (33.33 × 50) + 125 × (10 × 50) = 43.6mm
∑ Ai (100 × 50) + (33.33 × 50) + (10 × 50)
Which gives that: S = 43.6mm
Step 3: Coordinate transformation and re-draw the section with vertical distances about N.P.

106.4
3
56.4
y 81.4
2
31.4
6.4 N.P.
18.6
1
-43.6
Step 4: Determine Ieq S

I eqs = I 1 + I 2 + I 3
From the Parallel Axis Theorem: I i = I iLocal + y i2 Ai
bh 3
whereas I iLocal =
12
100 × 50 3
+ (18.6) × (100 × 50) = 2771467 mm 4
2
So: I1 =
12
j's
And similarly: I2 = 1990325 mm4 , and I3 = 3417147 mm4
So adding these three values together and converting the results to metres gives
Therefore: I eqS = I1 + I2 + I3 = 8.1789×10-6 m4
ra
Step 5: Determine Stresses in each material
My S 20 × (− 0.0436)
In Steel : σS = − =− = 106.6kPa
vi
I eqS 8.1789 × 10 − 6
In Aluminium;
ith
E 
I eqAl = I eqS  S ( )
 = 8.1789 × 10 − 6 × (3) = 24.537 × 10 − 6 m 4
 E Al 
and:
20 × 0.0564
Pr
My Al
σ Al = − =− = −45.97kPa
I eqAl 24.537 × 10 − 6
and in Wood:
E 
I eqW = I eqS  S ( )
 = 8.1789 × 10 − 6 × (10 ) = 24.537 × 10 −5 m 4
 EW 
MyW 20 × 0.1064
σW = − =− = −26.018kPa
I eqW 8.1789 × 10 −5

Chapter 7 Transverse Shear Stresses in Beams
7.0 SHEAR STRESSES IN BEAMS (SI&4th:363-365; 5th: 363-365)
In addition to the pure bending case, beams are often subjected to transverse loads which generate
both bending moments M(x) and shear forces V(x) along the beam as discussed in Chapter 5. The
bending moments cause bending normal stresses σ to arise through the depth of the beam, and the
shear forces cause transverse shear-stress distribution through the beam cross section as shown in
Fig. 7.1.
a P Transverse Force Cross-section a-a
Area A
a τ= 0 at the
top surface
x
Area A
Resultant Shear Shear stress τ

Force V(x)
j's
Fig. 7.1 Transverse shear force and transverse shear stress over cross-section of beam
ra
If we look at a typical beam section with a transverse stress as in Fig. 7.1, the top and bottom
surfaces of the beam carries no longitudinal load, hence the shear stresses must be zero here. In
other words, at top and bottom surfaces of beam section τ = 0. As a consequence of this, the
vi
shear stress distribution is not uniform and the formula of average shear stress is no longer valid
V (x )
τ avg = (7.1)
A
ith
7.1 SHEAR FORMULA (SI&4th: 365-367; 5th: 365-367)

Pr
Recall that in the development of the flexure formula, we assumed that the cross section must
remain plane and perpendicular to the longitudinal axis of the beam after deformation. Although
this is violated when the beam is subjected to both bending and shear, we can generally assume
the cross-sectional warping described above is small enough so that it can be neglected. This
assumption is particularly true for the most common cases of a slender beam, i.e. one that has a
small depth compared with its length.
To determine the shear stress distribution equation, consider a loaded beam as Fig. 7.2:
F1 F2
w(x)
M1 M2
x dx
Fig. 7.2 Beam with applied loads
Look at a FBD of the element dx with the bending moment stress distribution only, Fig. 7.3,
in which we do not need to look transverse forces if only horizontal equilibrium is considered.

M(x) N.A. dM(x)
M(x)+ dx dx
dx
Fig. 7.3 Length of beam dx with normal stress distribution due to bending moment
Summing the forces horizontally on this infinitesimal element, the stresses due to the bending
moments only form a couple, therefore the force resultant is equal to zero horizontally. Consider
now a segment of this element a distance y above the N.A. up to the top of the element. In order
for it to be in equilibrium, a shear stress τxy must be present, as shown in Fig. 7.4.
t(y)
σx1 σx2
dy dy
ytop ytop
τxy y y
M(x) N.A. dM(x) N.A.
M(x)+ dx dx
dx
j's
Fig. 7.4 Segment of length dx cut a distance y from N.A., with equilibrating shear stress τxy
dx
ra
Let the width of the section at a distance y from the N.A. be a function of y and call it “t(y)”.
Applying the horizontal equilibrium equation, gives:
ytop ytop
vi
+ → ∑ Fx = 0 = ∫ σ x1t ( y )dy −
y
∫ σ t ( y )dy + τ t ( y )dx = 0
y
x2 xy (7.2)
ith
Substituting for the magnitude of the stresses using ETB gives:

M (x ) y (M (x ) + dM (x )) y t ( y )dy + τ t ( y )dx = 0
ytop ytop
∫ t ( y )dy − ∫ xy
y
I y
I
Pr
Simplifying and dividing by dx and t(y) gives:

dM (x ) 1
ytop
τ xy =
dx It ( y ) ∫ yt ( y )dy
y
dM ( x )
But since V ( x ) =
dx
then, the Shear Stress Distribution is given by:
V (x ) V ( x )Q( y ) VQ
ytop
τ xy =
It ( y ) ∫ yt ( y )dy =
y
It ( y )
=
It
(7.3)
where:
V(x) the shear force carried by the section, found from the shear force diagram
I the second moment of area
t(y) the sectional width at the distance y from the N.A.
Q( y ) = ∫ yt ( y )dy = y ′A′
ytop
A’ is the top (or bottom) portion of the member’s cross-sectional
y
area, defined from the section where t(y) is measured, and y ′ is the distance to the centroid of
A’, measured from the Neutral Axis.

7.2 SHEAR STRESSES IN BEAMS (SI&4th: 368-376; 5th: 368-376)
Consider the beam to have a rectangular cross section of width b and height h as in Fig. 7.5
Centroid of A’
Parabolic
τ=0 curve
A’ y’
h τmax
τmax
h/2 y
NA
NA
b
Shear Stress distribution
Fig. 7.5 Computation and distribution of shear stress in a rectangular beam
j's
The distribution of the shear stress throughout the cross section due to a shear force V can be
determined by computing the shear stress at an arbitrary height y from the Neutral Axis.
   h   1h 
2
1h
ra
Q = y' A' =  y +  − y   ×   − y b  =  − y 2 b (7.4)
 22   2   2 4 
bh 3
The second moment of entire area: I =
vi
12
With t = b, applying the shear formula, Eq. (7.3), we have
1  h2 
ith
V ×  − y 2 b
2 4  = 6V  h − y 2 
2
VQ
τ= = (7.5)
It bh 3 bh 3  4 

×b
12
Pr
The result indicates that the shear stress distribution over the cross section is parabolic, as
plotted in Fig. 7.5. The shear force intensity varies from zero at the top and bottom, y = ± h/2,
to a maximum value at the neutral axis at y = 0 (Please comparing this with the normal stress
distribution in Chapter 6, Fig. 6.6).
From Eq. (7.5), the maximum shear stress that occurs at the Neutral Axis is computed as
V
τ max = 1.5 (7.6)
A
This same value for τmax can be obtained directly from the shear formula τ = VQ/It, by
realizing that τmax occurs where Q is largest. By inspection, Q will be a maximum when the
area above (or below) the neutral axis is considered, that is A’ = bh/2 and y' = h / 4 .
By comparison, τmax is 50% greater than the average shear stress determined from Eq. (7.1).

Example 7.1: Determine shear stress distribution in the following “I beam”, with a shear
force of V = 25 kN applied.
10 0.05
1
0.04
y1 = 45
N.A.
100 2 N.A.
0 y2 = 0
10 y3 = -45
S
-0.04
10 Ref 3
-0.05
100 Parallel Axis Theorem to find global I
Step 1: Determine the sectional geometric properties
Neutral Axis Location: S =

∑ si Ai = 95 × (100 × 10) + 50 × (80 × 10) + 5 × (100 × 10) = 50mm
∑ Ai (100 × 10) + (80 × 10) + (100 × 10)
j's
Parallel Axis Theorem: I = ∑ (I i local + y i2 Ai ) = (I 1loc + y12 A1 ) + (I 2 loc + y 22 A2 ) + (I 3 loc + y 32 A3 )
 100 × 10 3   10 × 80 3   100 × 10 3 
I =  + (45) × 100 × 10  +  + 0 2 × 10 × 80  +  + (− 45) × 100 × 10 
2 2
ra
 12   12   12 
+6 4 -6 4
I = 4.493×10 mm =4.493×10 m
vi
Step 2: Determine shear stress distribution using Eq. (7.3).
Start the integration from the top and work yourself down through all sub-sections of constant
thickness, ALWAYS integrating about the Neutral Axis.
ith
The shear stress equation is:

V (x )
ytop
VQ
τ xy =
It ( y ) ∫ yt ( y )dy =
y
It
Pr
We need to express shear stress segment by segment as divided in Step1.

i) For the range between 0.04 ≤ y ≤ 0.05, i.e. Area 1, the shear stress is given by:
0.05
25 × 10 3
0.1ydy = 2.782 × 10 9 (0.0025 − y 2 )
4.493 × 10 −6 × 0.1 ∫y
τ xy =
ii) Range -0.04 ≤ y ≤ 0.04, i.e. Area 2, the shear stress is given by:
25 × 10 3  0.05 0.04 
 
4.493 × 10 −6 × 0.01  0.∫04 ∫y
τ xy = 0 .1 ydy + 0.01 ydy


  0.05 2 0.04 2   0.04 2 y 2 
= 5.564 × 1010 0.1 ×  −  + 0.01 ×  −  = 2.782 × 10 9 (0.0106 − y 2 )
  2 2   2 2 
iii) Range -0.05 ≤ y ≤ -0.04, i.e. Area 3, the shear stress is given by:
25 × 10 3  0.05 0.04 −0.04 
 0.1ydy +  = 2.782 × 10 9 (0.0025 − y 2 )
4.493 × 10 −6 × 0.1  0.∫04 ∫ ∫
τ xy = 0 .01 ydy + 0 .1 ydy

− 0.04 y 
Plotting these distributions between their limits, gives the following discontinuous parabolic
distribution of shear stress:

y
2.5MPa
25.04MPa
N.A.
29.49MPa
Shear Stress
Distribution
25.04MPa
2.5MPa
7.3 COMBINED LOADS (SI&4th: 416-427; 5th: 416-427)

In the previous chapters, we developed methods for determining the stress distribution in a
member subjected to different types of load such as an axial force or a transverse shear force
(Chapter 2), a torsional moment (Chapter 4), and a bending moment (Chapter 6). Most often,
the cross section of a member is subjected to several of these loadings simultaneously. As we
shall see presently, we may combine the knowledge that we have acquired in the previous
j's
chapters. As long as the relationship between stress and the loads is linear and the geometry
of the member would not undergo significant change when the loads are applied, the
principle of superposition can be used as shown in Chapter 6. Here we are going to discuss
the situation due to tensile force F, torque T and transverse load P, as shown in Table 7.1.
ra
Table 7.1 Superposition of individual loads
Stresses Produced by Each Load Individually Stress Stresses
vi
Distributions
B B
Torsional Torsional shear
ith
Load A x
A C stress
(Torque T) T τT = Tρ/J
D T
D
Axial B B Tensile average

Pr
Load F
A A σavg normal stress
(Force F) D D σavg=F/A
B
σM Bending normal
B P N.A. A,C stress
Bending
Load y
D σM = -My/I
(Transverse A N.A. x B
Force P) τ Transverse
A
D N.A. C shear stress
τV = VQ/It
D
B
B Total normal
P A,C stress
C σ =F/A -My/I
A D
Combined
Loads y B
D Total shear
N.A.
F
A
N.A.
C stress at N.A.
x τ = VQ/It±Tρ/J
T D

Example 7.2: Two forces P=18kN and F=15kN are applied to the shaft with a radius of
R=20mm as shown. Determine the maximum normal and shear stresses developed in the shaft.
y y
B B
b=1 b=1
C 00m C 00m
m mm m
z
A a= 50 z
A
N.A N.A
D . R D .
F
F x x
P T=Pa
P
Step 0: Determine the geometrical properties of cross section:
Area of cross section: A = πR 2 = 3.1416 × 0.02 2 = 1.257 × 10 −3 m 2
Polar moment of inertia: J = πR 4 / 2 = 3.1416 × 0.02 4 / 2 = 251.3 × 10 −9 m 4
Second moment of area: I = πR 4 / 4 = 3.1416 × 0.02 4 / 4 = 125.7 × 10 −9 m 4
 πR 2   4 R 
First moment of semicircle: Q = A y = 
' '
 ×  −6
 = 5.33 × 10 m
3
 2   3π 
Step 1: Move eccentric force P to the center of the shaft
j's
This causes a uniform torsional moment (Torque) about axis x by T=Pa=18000×0.05=900Nm
as shown. Centric force P also will produce a varying bending moment M(x) along axis x. Axial
force F leads to a constant average compressive normal stress at cross sections along the shaft.
ra
Step 2: Determine the maximum bending moment Mmax and maximum shear force Vmax
P
B
y N.A.
vi
A x
M(x)
D Mmax
Loading Diagram Pb
ith
V(x)
0.1 x
x
0.1
-P Bending Moment Diagram
Shear Force Diagram
Pr
From the shear force and bending moment diagrams, one can identify that the shear force is uniform
along the shaft with V=P=18000N, and the maximum bending moment occurs at the section ABCD
with a magnitude of Mmax = Pb=18000×0.1=1800Nm. So the critical section is ABCD.
Step 3: Apply the superposition for determining the maximum normal stress
The maximum compressive stress occurs at point B, where both the maximum bending
moment Mmax and axial force F will form a highest combined compressive stress as
P M y − 15000 1800 × 0.02
σ max = σ B = − max max = −3
− = −11.93 − 286.40 = −298.33MPa
A I 1.257 × 10 125.7 × 10 −9
Step 4: Apply the superposition for determining the maximum shear stresses
As shown in table 7.1, the maximum shear stress occurs at point C, where both the transverse
shear force V=P and the torsional moment T=Pa give a highest combined shear stress as
TR 900 × 0.02
The max twist shear stress τ Tmax = = = 71.63MPa (at outer surface)
J 251.3 × 10 −9
VQ (18000) × (5.33 × 10 −6 )
The max shear stress in bending τ max = =
(125.7 × 10 −9 )× (2 × 0.02) = 19.08MPa (at N.P.)
V
It
The total combined max shear stress: τ max = τ C = τ Tmax + τVmax = 71.36 + 19.08 = 90.44MPa

Chapter 8 Displacement of Beams
It is often the case that the amount a beam deflects or rotates due to the applied transverse
loads has a limit placed on it. For this reason we now need to determine the deflection and
slope of a transversely loaded beam. This can be done in one of three ways: (1) Double
integration method; (2) Moment area method and (3) Energy method.
8.1 DOUBLE INTEGRATION METHOD (SI&4th: 569-599; 5th: 569-599)

Elastic Curve Equation
Consider a loaded beam as shown in Fig. 8.1:
y, v w
P
M
x
θ
dx
x
j's
Fig. 8.1 Loaded beam indicating its deflection and slope due to the applied loads
Look at a FBD of element dx, and consider only the bending moment that it is experiencing:
O
ra
dθ
vi
R R
dθ
ith
M M dv
. θ
N. A
y
Pr
dx dx
Fig. 8.2 The curvature of deflection Fig. 8.3 Infinitesimal segment dx showing angle
and vertical displacement
From Engineer's Theory of Bending (ETB), Eq. (6.9), we know that a beam under an applied
bending moment deflects with a curvature equal to the radius of a circle (arc), and that this
radius is related to the applied bending moment by:
1 M
= (8.1)
R EI
From Fig. 8.2, we can approximately compute arc dx = Rdθ , therefore,
1 dθ
= (8.2)
R dx
where θ can be considered to be the slope of the beam. But from the a more detailed diagram showing
Fig. 8.3, when dx approximates to zero and the slope is small enough, we have relationship as
dv
= tan θ ≈ θ (8.3)
dx

Combining Eqs. (8.2) and (8.3), we can obtain:
1 dθ d 2 v
= = (8.4)
R dx dx 2
and then substituting the part from ETB, Eq. (8.1), it gives Elastic Curve Equation:
d 2v M
= (8.5)
dx 2 EI
d 2v
or EI = M (x ) (8.6)
dx 2
here EI is referred to as flexural rigidity. Furthermore, from the relationship between distributed
load w(x) and shear force V(x) established in Chapter 5 (Eqs. (5.2) and (5.2)), one can have
d 3v
EI 3 = V ( x ) (8.7)
dx
d 4v
and EI = − w( x ) (8.8)
dx 4
Knowing the material and cross sectional properties of the beam, i.e. flexural rigidity(EI),
j's
Eq. (8.6) can be integrated ONCE to give an equation for the slope (θ) and TWICE to give
an equation for the displacement (v) of the beam as a function of x as:
dv M (x ) dv
=∫ = M ( x )dx + C
dx ∫
θ= dx + C or EIθ = EI
ra
(8.9)
dx EI
M (x )
v = ∫∫ dxdx + Cx + D or EIv = ∫∫ M (x )dxdx + Cx + D (8.10)
EI
vi
Integration of Macaulay’s Function

ith
Recalling Chapter 5, we adopted Macaulay’s function, Eq. (5.11), as

 0 for x < a
x−a
n
= (n ≥ 0) (8.11)
(x − a ) for x ≥ a
n
Pr
in the expression of the shear force and bending moment equations. In order to compute the
deflection v and slope θ = dv/dx from Eq. (8.5) or (8.6), we need to integrate Macaulay’s function as
n +1
n x−a
∫ x − a dx =
n +1
+ C' (8.13)
where C’ is constant of integration and can be determined from the kinematic boundary conditions.
Kinematic Boundary Conditions

The boundary conditions of the Cantilever and Simply Supported Beams can be seen from Fig. 8.4.
At x=0, v=0 and θ=dv/dx=0
Cantilever x
Beam
At x=0, v=0 At x=L, v=0

Simply x
Supported
Beam
L
Fig. 8.4 Kinematic boundary conditions

Example 8.1 Determine the slope and displacement equations of the Simply Supported (SS)
beam with a point load P.
F.B.D. (global equilibrium) F.B.D. (Section I-I)
P P
a L/2
I I
B M(x)
A x A o
θA vmax θB I I
V(x)
L x
RAY=(1-a/L)P RBY=Pa/L RAY=P/2
Step 1: Determine the ground reactions at supports A and B;
RAY = P/2 and RBY = P/2
Step 2: Bending moment equation via equilibrium for FBD of Section I-I.
By cutting the beam just before the RHS (Section I-I), the bending moment can be determined
as discussed in Chapter 5. Take moments about RHS:
1 1
P 1 L P 1 L
+ ∑MO = 0 = − 2 x + P x − 2 + M (x ) = 0 ∴ M (x ) = x −P x−
2 2
Step 3: Double Integration for the elastic curve equation.
j's
Substituting M(x) into the elastic curve equation Eq. (8.6), gives that:
1
d 2v d 2v P 1 L
EI 2 = M ( x ) → EI 2 = x −P x−
dx dx 2 2
ra
2
dv P 2 P L
Integrate once: EIθ = EI = x − x− + C (slope equation)
dx 4 2 2
3
vi
P 3 P L
Integrate again: EIv = x − x− + Cx + D (elastic curve equation)
12 6 2
where C and D are the constants of the integration. To determine them one needs to use Kinematic
ith
Boundary Conditions, which are from the known displacements and rotations of the beam.
Step 4: Determine the integration constants based on Kinematic Boundary Conditions
Our beam is simply supported at both ends, so kinematic boundary conditions are, as Fig. 8.4,
• when x = 0, v = 0, (recall the definition of Macaulay’s function as in Eq. (8.11))
Pr
P P
EI × 0 = × 0 3 − × 0 + C × 0 + D = 0 ∴D = 0
12 6
• when x = L, v = 0
3 3
P P L P P L PL2
+ CL + 0 = (L ) −  L −  + CL = 0 ∴ C = −
3 3
EI × 0 = L − L−
12 6 2 12 6 2 16
Step 5: Express the Slope and Elastic Curve Equations respectively:
2
dv P 2 P L PL2
Slope Equation: EIθ = EI = x − x− −
dx 4 2 2 16
3
P 3 P L PL2
Elastic Equation: EIv = x − x− − x
12 6 2 16
PL3
At x = L/2, the deflection reaches the maximum as v max = − (downwards)
48 EI
PL PL
At x = 0, θA = − (clockwise) and at x = L, θB = (counter-clockwise)
16 EI 16 EI

8.2 MOMENT-AREA METHOD (SI&4th: 600-613, 5th: 600-613)
The Moment-Area Method uses the Elastic Curve Equation derived above, but the integration
is done graphically and by doing so the kinematic boundary conditions are not considered.
Let’s look at elastic curve equation (8.5):
d 2v M
=
dx 2 EI
To integrate this equation graphically, you firstly need to draw the Bending Moment M(x)
diagram and then the M(x)/EI diagram. So look at a beam with arbitrary loadings as illustrated
in Fig. 8.5:
F1 F2
w(x)
Loading
Diagram
x1
x2
j's
M(x)
Bending
Moment
Diagram
ra
x
x1 x2
M(x)/EI
vi
A=θ2 - θ1
M(x)/EI
Diagram
ith
x
x1 x2
Fig. 8.5 Bending moment and M/EI diagrams for beam with arbitrary loading
1st Theorem of Moment Area

Pr
Integrating the elastic curve equation with respect to x, between two points x1 and x2, gives:
2
x2 d v x2 M
∫ x1 dx 2 dx = ∫ x1 EI dx
which can be reduced to:
x
 dv  2 x2 M
 dx  = ∫ x1 EI dx (8.14)
x1
This is
 dv   dv  x2 M
  −   = θ 2 − θ1 = ∫ x dx (8.15)
 dx  x2  dx  x1 1 EI
This equation gives the change in slope of the beam between x1 and x2. It is represented by the
area in the M/EI diagram between x1 and x2, and this equation is called the 1st THEOREM OF
MOMENT AREA. Note that the AREA should be considered in an algebraic sense, i.e. can
be positive or negative.
In this theorem, if dv/dx is known at x1, dv/dx at x2 can be very easily found via Eq. (8.15).
A useful side effect of this is that if I varies along the length of the beam, it can easily be
accommodated for, as to be shown in Example 8.2 below.

Example 8.2 Look at a cantilever beam where I = I0 for the left half of beam and I = I 0 / 2,
for the right half. Find the slope θ = dv/dx at the end C.
L/2 L/2 P
Loading
Diagram A I0 B I0/2 C
M(x)
L/2 L
x
Bending
Moment
Diagram
-PL/2
-PL
M(x)/EI
L/2 L
x
M(x)/EI -PL 1
Diagram 3
2EI0 2
-PL
-PL
EI0
EI0
We know that at x = 0, dv/dx = 0, so use the 1st Theorem of Moment Area Eq. (8.15) gives that:
j's
 dv   dv  M (x )
  −   = θ L − θ 0 = Area under diagram between x=0 and x=L
 dx  L  dx  0 EI
which can be computed by adding those three sub areas as shown,
ra
 dv   PL L 1 L PL 1 L PL  5 PL2
so:   − 0 = A1 + A2 + A3 = − × − × × − × ×  = −
 dx  L  2 EI 0 2 2 2 2 EI 0 2 2 EI 0  8 EI 0
vi
which gives that :
 dv  5 PL2
θC =   = −
ith
 dx  L 8 EI 0
2nd Theorem of Moment Area

Equation (8.15) only gives the change of slope between any two points, to determine the
Pr
displacement at some points along the beam, the second moment theorem must be applied. To
find displacement let’s return to elastic curve equation (8.5):
d 2v M
=
dx 2 EI
Multiply both sides by x and integrate between x1 and x2,
2
x2 d v x2 M x2  dv  x2 M
∫ x1 dx 2 xdx = ∫ x1 EI xdx i.e. ∫ x1 xd  dx  = ∫ x1 x EI dx
Integrating this by parts:
∫ pdq = pq − ∫ qdp (8.16)
where p and q are functions, gives:
x
x2  dv   dv  2 x2 dv x2 M
∫x
1
xd   =  x  − ∫
 dx   dx  x1 x1 dx
dx = ∫ x dx
x1 EI
(8.17)
x
dv  dv  2 x2 M
But, [v ]xx
x2
1
= ∫ x1 dx
2
dx , Eq. (8.17) becomes:  dx x − v 
x1
= ∫ x1
x dx
EI
As a result, we have

 dv   dv   x2 M
  x 2 −   x1  − [v2 − v1 ] = ∫ x x dx (8.18)
 dx  x2  dx  x1  1 EI
M(x)/EI [θ 2 x2 − θ1 x1 ] − [v2 − v1 ] = Ax
x
M(x)/EI A
Diagram
x
x1 x2
Fig. 8.6 The 2nd theorem of moment area for finding deflection
as the 2nd THEOREM OF MOMENT AREA. It can give the change in deflection (v1-v2)
between any two points x1 and x2 in terms of the change in slope and the first moment of the
area in the M/EI diagram as shown in Fig. 8.6.
Example 8.3 To see how this works look at the above example, but this time we require to
determine the displacement v = ? at the tip.
j's
Kinematic conditions: at x = 0 , v = 0 , dv/dx = 0 and
at x = L , dv/dx = -5PL2/8EI0,
Setting x1=0 and x2=L in this example, substituting these values into Eq. (8.18) gives:
ra
M(x)/EI
L/2+L/6=2L/3
L/4 L/2 L
xM(x)/EI x
vi
Diagram -PL 1 3
2EI0 2
-PL -PL
EI0 EI0
L/6
ith
 5 PL2   PL L  L  1 L PL  L  1 PL L  2 L
− × L − 0 − [v − 0] = − ×  × −  ×  × −  ×  ×
 8 EI   2 EI 2  4  2 2 2 EI 0  6  2 EI 2  3
Pr
0 0 0
When this is solved it gives that:
3 PL3
v=−
8 EI 0
8.3 ENERGY METHOD (SI&4th: 712-717, 768-774; 5th: 712-717, 768-774)

Consider a generalized beam loaded as shown in Fig. 8.7.
P1 P2
w(x)
MC
A B C D
MB θC
x dx
Fig. 8.7 Beam under arbitrary loads

Because of the loading condition this beam has a bending moment distribution along its length. In
Section 3.4, an equation, Eq. (3.8), for the stored strain energy of a structure was defined as:
σ2
U =∫ dV (8.19)
V 2E
In the transversely loaded beam, the normal stress distribution is given by Engineer’s Theory
of Bending, Eq. (6.10), such that:
M (x )
σx = − y (8.20)
I
Substituting Eq. (8.20) into Eq. (8.19) gives:
M 2 (x ) y 2
U =∫ dV (8.21)
V 2 EI 2
but for a transversely loaded beam, dV = dA×dx, thus:
1 M 2 (x ) y 2
U= ∫ ∫ dAdx (8.22)
2 L A EI 2
because M(x), E and I are constant for a specific cross section then:
1 M2
U= ∫
2 L EI 2 ∫ A
y 2 dAdx (8.23)
M2 j's
Since I = ∫ y 2 dA , then the total Strain Energy Stored in a Straight Beam is given by:
ra
U =∫ dx (8.24)
L 2 EI
And for a Circular Beam, the equation becomes:

vi
M 2 (θ)
U =∫ Rdθ (8.25)
θ 2 EI
To determine the displacement at the point of application of the load, Castigliano's 2nd
ith
Theorem is used. So differentiating the total Strain Energy with respect to the applied load P
gives the desired deflection as:
∂U L M ( x ) ∂M ( x )
vP = =∫ dx (8.26)
Pr
∂P 0 EI ∂P
In order to determine the slope of tangent θ at a point on elastic curve, the partial derivative of
the internal bending moment M(x) with respect to an external bending moment M’ acting at
the point must be found, as
∂U L M ( x ) ∂M ( x )
θM ′ = =∫ dx (8.27)
∂M ′ 0 EI ∂M ′
For example, at point C in Fig. 8.7, one can find the slope at C by formulating as
∂U L M ( x ) ∂M ( x )
θC = =∫ dx (8.28)
∂M C 0 EI ∂M C
Example 8.4 Determine vertical deflection for a simply supported beam with a central load P.
Step 1: Bending moment equation From Example 8.1, the bending moment equation is:
1
P 1 L
M (x ) = x −P x−
2 2

P
L/2
C I
A B
I
L
RAY=P/2 RBY=P/2
Step 2: Compute the total strain energy U

Because of Macaulay’s notation we have to do the following:
M 2 (x ) M 2 (x ) M 2 (x )
L/2 L
U =∫ dx = ∫ dx + ∫ dx
L 2 EI 0
2 EI L/2
2 EI
and substituting for the bending moments it gives:
2 2
 Px  P 
L/2   L  (L − x )
 2   2  dx
U= ∫ dx + ∫
2 EI 2 EI
j's
0 L/2
doing this integration gives the total bending strain energy as:
P 2 L3 P 2 L3 P 2 L3
U= + =
192 EI 192 EI 96 EI
ra
Step 3: Castigliano's 2nd Theorem
The displacement is then found by Castigliano's Theorem:
vi
∂U ∂  P 2 L3  PL3
vP = =  =
∂P ∂P  96 EI  48EI
ith
which is the displacement at the point of application of P in the direction of P.
Remarks
Pr
We may observe that the deflection vj of a beam at a given point C can be obtained by direct
application of Castigliano’s theorem only if a real load Pj happens to be applied at C in the
direction in which vj is to be determined. When no real load is applied at Cj, or when a real load
P is applied in a direction other than the desired one, we need to apply a fictitious or virtual
load Qj at Cj along the direction in which the deflection vj is to be determined and use
Castigliano’s theorem to obtain deflection vj, similarly to the approach stated in Chapter 3, as
∂U (P ,Q j )
vj = (8.29)
∂Q j
Keep in mind that the internal strain energy here contains the contributions from both actual
load P and virtual load Qj. After computing the partial derivative with respect to Qj, then
make Qj = 0 in Eq. (8.29).
The slope θj of a beam at point Cj may be determined in a similar manner by applying a fictitious
couple Mj at Cj, then computing the partial derivative as
(
∂U P , M j )
θj = (8.30)
∂M j
and making Mj = 0 in the expression obtained.

Example 8.5 The cantilever beam AB supports a uniformly distributed load w as shown.
Determine the deflection vB and slope θB at the free end B.
w w I
MA x
B
A A
B
L I
L QB
RAY
Virtual load to
find deflection vB
Part I: Determine Deflection vB
Since there is no real concentrated load at B along vertical direction, a fictitious load QB must
be applied at the point along the desired direction as shown.
Step 0: Ground reactions due to both w and QB:
∑ M A = 0 = M A − QB L − (wL )(L / 2) = 0
+
∴ M A = wL2 / 2 + QB L
+ ↑ ∑ Fy = 0 = R AY − QB − wL = 0 ∴ R AY = wL + QB
Step 1: Bending moment equation (via Section I-I)
w 2
j's
+ ∑ M O = 0 = M A x − R AY x + x + M (x ) = 0
0 1
2
M (x ) = − M A x
0 1
+ R AY x − w x
2
(
/ 2 = − wL2 / 2 + Q B L x ) 0
+ (wL + QB ) x − w x
1 2
/2
ra
Step 2: Compute the total strain energy U in terms of both real (w) and virtual load (QB)
M 2 (x )
2
U (w,Q B ) = ∫
L 2 EI
dx =
1
∫
2 EI L 

 ( )
− wL2 / 2 + Q B L x + (wL + Q B ) x −
0 1 w 2
2
x  dx

vi
Step 3: Using Castigliano’s Theorem
∂U (w,Q B ) ∂ M 2 ( x , w,Q B ) 1 ∂M ( x , w,Q B )
vB = = ∫ dx = ∫ M dx
ith
∂QB ∂Q B L 2 EI EI L ∂Q B
∂M ( x , w,QB ) 0 1
But = −L x + x
∂QB
Pr
Substituting for M and ∂M ∂QB into the previous equation and setting QB = 0, we have
vB =
1
EI ∫L M
∂M
∂QB
dx =
1
EI
  L2 
∫ L  −  w 2 + 0  x
0
+ (wL + 0) x −
1 w 2 
2
x

(
−L x
0
+ x
1
)dx
 
L
1  L2 w 2 1  wL3 3wL2 3wL 2 w 3 
∫ L  2 ( )
EI ∫0  2
=  − w + wLx − x  x − L dx =  − x + x − x dx
EI 2  2 2 2 
wL4
vB = (“+” means the same direction as QB)
8EI
Part II: Determine slope θB

w I B
MA
x
A I MB
L Virtual moment to
RAY find slope θB
Slope corresponds to a couple in Castigliano’s 2nd theorem. It is hence necessary to apply a
virtual moment MB at the B as shown.

Step 0: Ground reactions due to both w and MB:
+ ∑ M A = 0 = M A − M B − (wL ) L  = 0 ∴ M A = wL2 / 2 + M B
2
+ ↑ ∑ Fy = 0 = R AY − wL = 0 ∴ R AY = wL
w 2
+ ∑ M O = 0 = M A x − R AY x + x + M (x ) = 0
0 1
2
M ( x ) = − M A x + R AY x −
0 1 w 2
2
( 0
x = − wL2 / 2 + M B x + wL x −
1 w 2
2
x)
Step 2: Total strain energy U in terms of both real force (w) and virtual moment (MB)
M 2 (x )
2
U (w, M B ) = ∫
L 2 EI
dx =
1
∫
2 EI L 

− wL 2
(
/ 2 + M B x
0
+ wL )x
1
−
w 2
2
x  dx

Step 3: Castigliano’s Theorem
∂U (w, M B ) ∂ M 2 ( x , w, M B ) 1 ∂M ( x , w, M B )
θB =
∂M B
=
∂M B ∫ L 2 EI
dx =
EI ∫ L
M
∂M B
dx
∂M ( x , w, M B )
j's
0
but =− x
∂M B
Substituting for M and ∂M ∂M B into the previous equation and then setting MB = 0, we have
ra
θB =
1
EI ∫L M
∂M
∂QB
dx =
1
EI ∫L
  L2
  2

− w + M B  x

0
+ wL x −
1 w 2 
2
x

(
− x
0
)dx
   
vi
1   L2  w 2  1
L
 wL2 w 2 wL3
 −  w + 0 x x (− 1)dx =
0 1
=
EI ∫L   2  + wL x −
2  EI ∫ 
 2 − wLx + x dx =
2  6 EI
ith
    0 
3
wL
θB = (the same rotational direction as MB)
6 EI
Pr
So far we have just been looking at beams that are statically determinate, we now need to
look at the cases when the beams are statically indeterminate, that is there are more unknown
reaction forces than equations of statics. In these cases we need to come up with as many
compatibility equations as are necessary to solve the problem.
8.4 STATICALLY INDETERMINATE BEAMS (SI&4th: 614-647; 5th: 614-647)

A beam is statically indeterminate when there are more unknown support loads (forces &
moments) than equations of statics. Because these excess support loads have associated with
them excess boundary conditions, we can use these to solve the problem. For example, for a
propped cantilever as shown in Fig. 8.8:
P
2L/3
I
B
MA
A I
RAY L RBY
Fig. 8.8 Statically indeterminate cantilever beam with load P

Three unknown reactions, RA, RB and MA ;
Two equations of statics (ΣFy=0 and ΣM=0)
Three kinematic boundary conditions, at x = 0, v =0 and dv/dx = 0, at x = L, v = 0.
8.4.1 INTEGRATION METHOD TO SOLVE STATICALLY INDETERMINATE BEAMS

When employ the integration method as given in Section, 8.1, two of these kinematic boundary
conditions are necessary for determining the constants of integration; the third is used with the
two equations of statics to solve for the three reaction loads.
Example 8.6 Determine the reaction loads in the indeterminate beam with a point force P
applied at 2/3L, as shown in Fig. 8.8.
Step 1: Global equilibrium from statics:
+ ↑ ∑ Fy = 0 = R AY + R BY − P = 0 (8.31)
+ ∑ M A = 0 = M A − 2 PL / 3 + RBY L = 0 (8.32)
j's
Step 2: Bending moment equation
Take moments about Section I-I by cutting just before RHS, as shown in Fig. 8.8:
∑MO = 0 = M A − R AY x + P x − 2 L / 3 + M ( x ) = 0
0 1 1
+ x
ra
The Moment equation is given by:
M (x ) = − M A x
0 1 1
+ R AY x − P x − 2 L / 3
vi
Step 3: Determine Elastic Curve equation
We now need to derive the displacement equation and apply the kinematic boundary conditions:
ith
Using the Elastic curve equation gives that:

d 2v
EI 2 = M ( x ) = − M A x + R AY x − P x − 2 L / 3
0 1 1
dx
R
Pr
dv P
= ∫ M ( x )dx = − M A x + AY x −
1 2 2
Integrate once: EIθ = EI x − 2L / 3 + C
dx 2 2
MA R AY P
Integrating again: EIv = ∫∫ M ( x )dxdx = −
2 3 3
x + x − x − 2 L / 3 + Cx + D
2 6 6
Step 4: Determine the integration constants based on Kinematic Boundary Conditions
Since at x = 0 , dv/dx = 0 , then C = 0 Since at x = 0 , v = 0 , then D = 0
MA 2 R 3 P 3
Elastic Curve Equation: EIv = − x + AY x − x − 2L / 3
2 6 6
Step 5: Give an additional equation using other Kinematic Boundary Condition
3
MA 2 R AY 3 P 2
Also, at x = L, v = 0, we have equation: − L + L − L− L =0
2 6 6 3
which gives : R AY L − 3M A − PL / 27 = 0 (8.33)
Solving Eqs. (8.31), (8.32) and (8.33) simultaneously gives that:
RA = 13P/27, RB = 14P/27, MA = 4PL/27
You can then substitute these values into the above equations to obtain the slope and
displacement of the beam.

8.4.2 SUPERPOSITION METHOD TO SOLVE STATICALLY INDETERMINATE BEAMS
An alternative way of solving the above problem is by using the superposition method. Using
the superposition method we can very easily generate the extra equations necessary to solve
the statically indeterminate beam. This can be done by determining the deflection on the beam
due to each of the applied loads and then add all of these displacements together.
Example 8.7 The same as Example 8.6 but using Superposition Method.
The loads and displacements in this beam are equivalent to treating it as two separate statically
determinate beams, then combining the separate displacements in the following way.
P P B
2L/3 2L/3 v2
B
AM
A
= A
B
v1 + A RBY
MA1 MA2
L L L
RAY RBY RAY1 RAY2
j's
To determine the displacements v1 and v2 you can refer to standard solutions given in the
Textbook in Appendix C, 4th:800-801; 5th 800-801 or in other references.
ra
For this example, the equation for v1 is:
P(2 L 3) 
2
2L 
v1 = −  3L − 
6 EI  3 
vi
The equation for v2 is :
R L3
ith
v 2 = BY
3EI
But because at B, v = 0, then kinematic compatibility condition is
v1 + v 2 = 0
Pr
P(2 L 3) 
2
2 L  RBY L3
i.e. v1 + v 2 = −  3L − + =0
6 EI  3  3EI
and when you solve for this you get that:
14
R BY = P
27
8.4.3 CASTIGLIANO’S METHOD FOR STATICALLY INDETERMINATE BEAMS

The reactions at supports of a statically indeterminate elastic structure may be determined by
Castigliano’s 2nd theorem as Eq. (8.26) or Eq. (8.27), in which the redundant reaction is treated as
an unknown load Rj. We firstly calculate the strain energy U of the structure due to the combined
action of the given loads and the redundant reaction Rj. Observing that the partial derivative
∂U ∂R j , represents the deflection (or slope) at the support. We then set this derivative equal to
zero (because of zero deflection at the support) and solve the equation obtained for the redundant
reaction. The remaining reactions may be obtained from the equation of statics.

Example 8.8 Determine the reactions at the supports for the prismatic beam and loading shown.
w w I
A B = A
I B
x
L RAY L
The beam is statically indeterminate to the first degree (i.e. one redundant reaction). We
consider the reaction at A as redundant and release the beam from the support. The reaction
RAY is now considered as an unknown load as shown and will be determined from the
condition that the deflection vA must be zero. Note that as an unknown, RAY in this case is in
effect a real load.

w 2 w
+ ∑ M O = 0 = − R AY x + x + M (x ) = 0 ∴ M ( x ) = R AY x −
1 1 2
x
2 2
Step 2: Total strain energy U in terms of RAY
M 2 (x )
j's
2
1  w 2
U (w, R AY ) = ∫
1
L 2 EI
dx =
2 EI ∫ L 

R AY x −
2
x  dx

ra
Step 3: Castigliano’s Theorem
∂U (w, R AY ) 1 ∂M ( x , w, R AY ) 1 ∂M
vA =
∂R AY
=
EI ∫ L
M
∂R AY
dx =
EI ∫ L
M
∂R AY
dx
vi
∂M ∂M ( x , w, R AY ) 1
but = = x
∂R AY ∂R AY
ith
Substituting for M and ∂M ∂R AY into the previous equation, we have
( )dx = EI1 ∫
L
1 ∂M 1  1 w 2 1  w 3
∫L ∫ L  R AY x − 2 x  x
2
vA = M dx =  R AY x − x dx
EI ∂QB EI 0  2 
Pr
Step 4: Kinematic compatibility condition

vA = 0
L
1  R AY 3 w 4 R AY L3 wL4
∴vA = x − x = − =0
EI  3 2 × 4  0 3 8
∴ R AY = 3wL / 8 ↑
From the conditions of equilibrium for the beam, we find that the reaction at B consists of
following force and bending couple:
R BY = 5wL / 8 ↑ M B = wL2 / 8

Table 8.1 Comparison of bending beams with axially loaded bars and torsional shafts
Axial Loaded Bar/Rod Torsional Shaft Bending Beams
Load Type Force F (N) Torque T (Nm) Transverse Force P (N) or/and
Bending moment M (Nm)
Tension or Compression Right-Hand Rule
+T Bending moment
+F
+φ +ve M
Sign +
Convention
+ Tension +F
-ve M
Of Internal +T
Load -F -T Shear Force
- Compre
ssion -F - -φ +ve V -ve V
-T
I–Second moment of area (m4)
2 4
A – Area (m ) J–Polar moment of inertia (m ) πR 4 πD 4
π 4 4 Circle: I = =
Circle A = πR 2 = D 2 Circle: J = πR = πD 4 64
Geometric 4 2 32
Property Hollow: bh 3
Rectangle A = bh , Rectangle I =
1
Triangle A = bh J=
(
π Ro − Ri4
4
=
) (
π Do4 − Di4 ) 12
2 32 bh 3
2 Triangle I =
36
j's
Material E – Young’s Modulus (Pa) G – Shear Modulus (Pa) E – Young’s Modulus (Pa)
Properties
My
Shear stress: Normal stress: σ = − (Pa)
I
ra
Normal average stress: T
τ= ρ (Pa) y
σ avg =
F
(Pa) J N.A. M
A
vi
Stresses Distribution of VQ
Uniform shear stress Shear stress: τ = (Pa)
Distribution ρ It
Q=y’A’ t
ith
T
A’
y y’
V
N.A.
I
Pr
Strains Normal strain: ε = ∆L / L Shear strain γ - radian Normal strain or Shear strain
Hooke’s Law σ = Eε τ = Gγ σ = Eε or τ = Gγ
Deflection (m) (Elongation) Angle of Twist (Radian) Deflection v (m)
(Elastic Curve Equation)
F (x ) T (x )
L L
General: δ = ∫ E (x )A(x )
dx General: ϕ = ∫ G(x )J (x )dx v= ∫∫
M (x )
dxdx + Cx + D
0 0 EI
Deformation FL TL
Single segment: δ = Single segment: ϕ =
EA GJ Slope of Elastic Curve (Radian)
dv M (x )
Multi-segments: δ = ∑E A
Fi Li
Multi-segments: ϕ = ∑G J
Ti Li θ=
dx
= ∫ EI
dx + C
i i i i i i
External 1 1 1 1
Work W = P∆ p W = Tϕ T W= Pv P or W = M ' θ'
2 2 2 2
Total Strain F2 Fi 2 Li T2 Ti 2 Li M2
Energy U = ∫V 2 EA
dV = ∑ 2E A
i i i
U= ∫ V 2GJ
dV = ∑ 2G J
i i i
U= ∫
V 2 EI
dV
Castigliano’s ∂U ∂U ∂U ∂U
2nd Theorem ∆P = ϕT = vP = or θ M ' =
∂P ∂T ∂P ∂M '
Chapters Chapters 2-3 Chapters 4 Chapters 5-8

Chapter 9 Buckling of Columns
9.0 INTRODUCTION TO BCUKLING (SI&4th:649-652)
In discussing the analysis and design of various structures in the previous chapters, we had
two primary concerns: (1) the strength of the structure, i.e. its ability to support a specified
load without experiencing excessive stresses; (2) the ability of the structure to support a
specified load without undergoing unacceptable deformations. In this chapter, we shall be
concerned with stability of the structure, i.e. with its ability to support a given load without
experiencing a sudden change in its configuration. Our discussion will relate mainly to
column, i.e. to the analysis and design of vertical prismatic members supporting axial loads.
If a beam element is under a compressive load and its length if the orders of magnitude are
larger than either of its other dimensions such a beam is called a columns. Due to its size its
axial displacement is going to be very small compared to its lateral deflection called buckling.
Quite often the buckling of column can lead to sudden and dramatic failure. And as a result,
special attention must be given to design of column so that they can safely support the loads.
In looking at columns under this type of loading we are only going to look at three different
types of supports: pin-ended, doubly built-in and cantilever.
j's
9.1 SLENDER PIN-ENDED COLUMN (SI 649-657; 4th:652-657; 3rd Ed p.653-
661)
ra
Due to imperfections no column is really straight. At some critical compressive load it will
buckle. To determine the maximum compressive load (Buckling Load) we assume that
buckling has occurred as shown in Fig. 9.1,
y,v
vi
P P
x
ith
Fig. 9.1 Deflection column due to applied compressive load P
Look closely at the FBD of the left hand end of the beam as in Fig. 9.2:
Pr
M(x)
y,v
P
V(x) v
P
x
Fig. 9.2 FBD of section of length x of deflected column
Equating moments at the cut end:

∑ M = 0 = Pv + M (x ) = 0 ∴ M ( x ) = − Pv (9.1)
But since the deflection of a beam is related with its bending moment distribution, then:
d 2v
EI 2 = − Pv (9.2)
dx
d 2v  P 
which simplifies to: +  v = 0 (9.3)
dx 2  EI 
where P/EI is a constant. This expression is in the form of a second order differential
equation of the following type:

d 2v
2
+ α 2v = 0 (9.4)
dx
P
where: α2 = (9.5)
EI
The solution of this equation is:
v = A cos(αx ) + B sin(αx ) (9.6)
where A and B are constants, which can be determined using the column’s kinematic
boundary conditions.
Kinematic Boundary Conditions

at x = 0, v = 0: 0 = A + 0, giving that A = 0
at x = L , v = 0, then: 0 = B sin(αL )
If B = 0, No bending moment exists, so the only logical solution is for: sin(αL ) = 0 and the
only way that this can happen is if :
α L = nπ , (9.7)
2
P  nπ 
where n = 1,2 ,3,L . But since: α 2 = =  (9.8)
EI  L 
j's
then we get that buckling load as:
π 2 EI
P = n2 2 (9.9)
L
ra
The values of 'n' define the buckling mode shapes, as in Fig. 9.3:
vi
P1 P1
π 2 EI
First mode of buckling P1 =
L2
ith
P2 P2
4π 2 EI
Second mode of buckling P2 =
Pr
L2
P3 P3
9π 2 EI
Third mode of buckling P3 =
L2
Fig. 9.3 First three modes of buckling loads
Critical Buckling Load

However, since P1 < P2 < P3, the column buckles at P1 and never gets to P2 or P3 unless
bracing is placed at the points where v = 0 to prevent buckling at lower loads.
The critical load for a pin ended column is therefore:
π 2 EI
PCrit = 2 = PE (9.10)
L
which is also called Euler Buckling Load,
PCrit Critical or maximum axial load on the column just before it begins to buckle
E Young’s modulus of elasticity
I least second moment of area for the column’s cross sectional area
L unsupported length of the column, whose ends are pinned
9.2 BUILT-IN COLUMN (SI&4th: 658-668; 3rd Ed p.662-672)
The critical load for other columns can be expressed in terms of the critical buckling load for
a pin- ended column PE. A built-in column looks like Fig. 9.4:
L
P A Zero Bending Moment
B P
L/4 L/2 L/4
P P
LE
Fig. 9.4 Built-in column at both ends showing the effective pin-ended length
j's
From symmetry conditions, at the points of inflection
d 2v
= 0 = M (x )
dx 2
ra
which occurs at 1/4L points. Thus the middle half of the column can be taken out and treated
as a pin-ended column of length LE = L/2 as shown in Fig. 9.4. The critical load for this half
length is then :
vi
π 2 EI 4π 2 EI
PCrit = 2 = = 4 PE (9.11)
LE L2
ith
9.3 CANTILEVER COLUMN

Pr
P L=LE/2
A
B P
LE
Fig. 9.5 Cantilever column and its effective length
This is similar to previous case. However, this span is equivalent to 1/2 of the Euler span LE,
as illustrated in Fig. 9.5, thus:
π 2 EI π 2 EI PE
PCrit = 2 = = (9.12)
LE 4 L2 4
Note: Since PCrit is proportional to I, the column will buckle in the direction corresponding to
the minimum value of I, as shown in Fig. 9.6:

Buckling Direction Cross-section
P y
A
y P z
h
x
z
b
Iy>Iz
bh 3 hb 3
Fig. 9.6 Column cross section showing the direction of buckling (here: I z = < Iy = )
12 12
9.4 CRITICAL COLUMN STRESS

A column can either fail due to the material yielding, or because the column buckles, it is of
interest to the engineer to determine when this point of transition occurs.
Consider the Euler buckling equation 9.10
π 2 EI
PE = 2
L
Because of the large deflection caused by buckling, the least second moment of area term I
can be expressed as follows:
j's
I = Ar 2 (9.13)
where: A is the cross sectional area and r is called radius of gyration of the cross sectional
area, i.e. r = I / A . Note that the smallest radius of gyration of the column, i.e. the least
ra
second moment of area I should be taken in order to find the critical stress.
Dividing the buckling equation by A, gives:
P π2 E
σE = E =
vi
(9.14)
A (L / r )2
where: σE is the compressive stress in the column and must not exceed the yield stress σY of the
ith
material, i.e. σE<σY, L / r is called the slenderness ratio, it is a measure of the column's flexibility.
If this equation is plotted for steel it gives:
σx
Pr
σY
240MPa
π2 E
σ Crit =
( L / r )2
L/r
89
Fig. 9.7 Critical stress vs slenderness ratio for steel
For a column not to fail by either yielding or buckling, its stress must remain underneath this
diagram in Fig. 9.7.
Example 9.1 A 2m long pin ended column of square cross section. Assuming E=12.5GPa,
σallow=12MPa for compression parallel to the grain, and using a factor of safety of 2.5 in
computing Euler’s critical load for buckling, determining the size of the cross section if the
column is to safely support (a) a P = 100kN load and (b) a P = 200kN load.
Section a-a
I y
P A a B P
s z
a
s
Part (a)
1 3 s4
Second moment of area Iz = Iy = ss =
12 12
Buckling criterion
 F fail 
Using given Factor of Safety FS=2.5  FS =  , we make the required critical load as
 Fallow 
PCrit ≥ FS × P = 2.5 × 100kN = 250 × 10 3 N
Based on Euler’s formula, Eq. (9.10), we have
π 2 EI 250 × 10 3 L2
PCrit = 2 ≥ 250 × 10 3 N ∴I ≥
L π2 E
250 × 10 3 L2 250 × 10 3 × 2 2
j's
or: s B1 ≥ 4 × 12 = 4 × 12 = 0.0993m = 99.3mm
π2 E π 2 × 12.5 × 10 9
P P
Stress criterion σ = =≤ σ allow ∴ A = s2 ≥
A σ allow
ra
P 100 × 10 3
i.e. s σ1 ≥ = = 0.0913m = 91.3mm
σ allow 12 × 10 6
vi
Comparing the results from these two criteria, we have s ≥ max{s B 2 , s σ 2 } = 99.3mm . In this
case, the design is taken against the buckling criterion. Finally, one may select a round-up
ith
amount, e.g. s = 100mm, as the design of the size of cross section.
Part (b)
Buckling criterion
Pr
PCrit ≥ FS × P = 2.5 × 200kN = 500 × 10 3 N

π 2 EI 500 × 10 3 L2
Step 2: Euler’s formula PCrit = 2
≥ 500 × 10 3 N ∴I ≥
L π2 E
500 × 10 3 L2 500 × 10 3 × 2 2
or: sB2 ≥ 4 × 12 = 4 × 12 = 0.1181m = 118.1mm
π2 E π 2 × 12.5 × 10 9
P P
Stress criterion σ = =≤ σ allow ∴ A = s2 ≥
A σ allow
P 200 × 10 3
i.e. s σ 2 ≥ = = 0.1291m = 129.1mm
σ allow 12 × 10 6
Comparing the results from these two criteria, we have s ≥ max{s B 2 , sσ 2 } = 129.1mm . In this
case, the design is taken against the stress criterion. One may select s = 130mm as the design
of the size of cross section.

Example 9.2 Determine the largest load P which may be applied to the structure as shown.
Assume that E=200GPa, allowable vertical deflection at point A δallow=0.5mm and allowable
compressive and tensile stress σallow=50MPa.
B Cross section for AB & AC

Pin A
FAB h=50mm
3m y
30º C
30º A z
A FAC
8m
P P
b=100mm
Step 1: Determine the members’ internal forces
+ ↑ ∑ Fy = 0 = FAB sin 30° − P ∴ FAB = 2 P (+ tensile force)
+ → ∑ Fx = 0 = FAB cos 30° + FAC ∴ FAC = − 3P (- compressive force)
Step 2: Buckling criterion FAB is in tension, we do not considered its buckling. But bar AC is
a strut and we need to check for buckling. I about y and z is computed respectively
 bh 3 0.1 × 0.53 −6 4   hb 3 0.5 × 0.13 −6 4 
Iz = = = 1.04267 × 10 m  <  I = = = 41.667 × 10 m 
   y
j's
 12 12   12 12 
∴ PCrit ,AC =
π 2 E AC I AC
=
( )(
π 2 × 200 × 10 9 × 1.04267 × 10 −6 ) = 32.128kN
L2AC 8 2
ra
But FAC = PCrit ,AC = − 3PB , ∴ PB = PCrit / 3 = 18.55kN
Step 3: Strength criterion Consider tensile and compressive stresses in AB and AC respectively.
vi
F 2P 50 × 10 6
σ AB = AB = ≤ σ allow = 50 × 10 6 P= = 125kN
AAB 0.05 × 0.1 400
ith
FAC 3P 6 50 × 10 6
σ AC = = ≤ σ allow = 50 × 10 P= = 144.3kN
AAC 0.05 × 0.1 400
From stress criterion, the maximum allowable load should be the smallest one i.e. Pσ=125kN
Pr
Step 4: Stiffness criterion Consider vertical deflection at point A using Castigliano’s method.
F 2L F2 L F2 L
Total strain energy due to axial forces: U = ∑ i i = AB AB + AC AC
i 2 E i Ai 2 E AB AAB 2 E AC AAC
∂  Fi 2 Li  ∂F  L
 = ∑ (Fi ) i  i

The displacement can be then computed as: ∆ P = ∑
∂P  i 2 Ei Ai  
 ∂P  Ei Ai

 i 
Member Fi (N) ∂Fi ∂P Li(m) Ai (m2) Fi (Li Ei Ai )(∂Fi ∂P )
AB 2P 2 6 0.05 2.4×10-9P
AC - 3P - 3 8 0.05 2.4×10-9P
( ) ( )
Thus we have: ∆P = 2.4 × 10 −9 P + 2.4 × 10 −9 P = 4.8 × 10 −9 P ≤ δ allow
0.0005
∴ Pδ = = 104.17kN
4.8 × 10 −9
Step 5: Determine the maximum allowable load P from the above three criteria
Clearly, for the safety reason, we should pick the lowest level as the allowable load
P = min{PB , Pσ , Pδ } = 18.55kN

Chapter 10 Biaxial Stress Systems
Until now we have only considered structures with individual direct stress constant (axial
stress) or varying (e.g. bending stress) across the section and/or shear stresses. A biaxial stress
system has a stress state in two directions and a shear stress typically showing in Fig. 10.1:
σyy
τxy
σxx σxx
y
τyx
x
σyy
Fig. 10.1 Element of a structure showing a biaxial stress system
The direct strain in any direction is the sum of all factors contributing to that strain:
a) Direct stresses in that direction
j's
b) Poisson's ratio effects of stresses at right angles
c) Thermal strain
Recalling our discussion in Chapter 2, we can give the relationship between stress and strain as:
ra
σ σ yy
ε xx = xx − v + α∆T
E E
σ yy σ
ε yy = − v xx + α∆T (10.1)
vi
E E
σ xy τ xy
γ yy = =
ith
G G
If these equations are rearranged to compute the stresses for a set of given the strains, we get:
)[ ]
E
σ xx = ε xx1 + vε yy1
( 1− v2
Pr
)[ ]
E
σ yy = ε yy1 + vε xx1
( 1− v2
(10.2)
E
τ xy = Gγ xy = γ xy
2(1 + v )
where: ε xx1 = ε xx − α∆T and ε yy1 = ε yy − α∆T . When a Biaxial Stress state occurs in a thin
metal, all the stresses are in the plane of the material. Such a stress system is called PLANE
STRESS. We can see plane stress in pressure vessels, aircraft skins, car bodies, and many
other structures. Some of which we are able to analyze relatively easily.
10.1 THIN WALLED PRESSURE VESSELS (SI&4th: 409-415; 3rd Ed p.413-419)

Cylindrical or spherical pressure vessels are used in industry as tanks, boilers or containers.
When under pressure the material is subjected to loadings in all directions.
In general, thin wall refers to an inner radius to wall thickness ratio greater than 10, e.g.
r/t=20, in most cases it is actually r/t > 50. If the vessels wall is thin, the stress distribution
through the thickness can be assumed to be uniform.

10.1.1 CYLINDRICAL PRESSURE VESSELS
This analysis will look at tubes with an internal pressure and closed ends. Let σxx be the Axial
Stress due to the pressure on the end walls, and σθθ be the Hoop Stress due to the pressure
acting on the curved surface.
t Cut surface
σxx
P
Fig. 10.2 FBD of axial section of a cylindrical pressure vessel
Axial Stress
Look at a FBD of the axial section as shown in Fig. 10.2 and check for the axial equilibrium.
∑ Fx = 0 = − Pπr 2 + 2πrtσ xx
j's
ie: Pπr 2 = (2πrt )σ xx
which gives the equation for Axial Stress( or Longitudinal Stress):
Pr
σ xx = (10.3)
ra
2t
Hoop Stress
vi
Look now at a FBD of the circumferential section as shown in Fig. 10.3:
σθθ σθθ
ith
P
Pr
Fig. 10.3 FBD of circumferential section of the cylindrical pressure vessel
Equating the forces vertically gives:

∑ Fy = 0 = − P × (2r × L ) + σ θθ × 2(L × t ) ∴ 2σ θθ (Lt ) = 2rLP
which simplifies to give the equation for Hoop Stress (or Circumferential Stress):
Pr
σ θθ = (10.4)
t
By substituting into the above biaxial strain equations, we get:
1 Pr  1 
ε xx = [σ xx − vσ θθ ] + α∆T =  − v  + α∆T
E tE  2 
(10.5)
1 Pr  v 
ε θθ = [σ θθ − vσ xx ] + α∆T = 1− + α∆T
E tE  2 
If there is no temperature change, the α∆T term disappears.

Combined Load Conditions (SI&4th: 416-437; 3rd 420-440)
As well as these stresses, a pressure vessel can also have stresses created by the weight of the
pressurised fluid inside, its own weight, externally applied loads and by an applied torque. To
analyse this, each loading condition is considered individually and the stresses are then
combined along their respective axes by the superposition method. Let’s look at a pressure
vessel with several possible loading conditions, such as, bending moment M, Torque T and
axially tensile force F, as shown in Fig. 10.4:
t
E
M M
/2
F
d
F T
r=
P
L T
Fig. 10.4 Pressure vessel with Axial force F, Bending moment M,

Torque T and internal pressure P
Look at a small element E from the vessel with dimensions dθ×dx, the stresses are given in Fig. 10.5:
j's
Pr
Hoop Stress σ θθ = Tr
t Torsional
Shear Stresses τ xθ =
J
ra
Combined Axial Stresses
E Pr My F
σ xx = − +
2t I A
vi
ith
Fig. 10.5 Element of pressure vessel with various possible stresses from various applied loads
9.1.2 SPHERICAL PRESSURE VESSELS

They are commonly in use to store liquid gases as they provide the greatest storage volume
Pr
for the least volume of material. The FBD of hemisphere is shown in Fig. 10.6.
t σ
P
Fig. 10.6 Diagram of spherical pressure vessel with internal pressure P

Equating the forces vertically:
∑ Fy = 0 = − Pπr 2 + 2πrtσ
Pr
which simplifies to: σ= (10.6)
2t
Note that the value for the stress is not dependent on the orientation of our FBD, so this stress
acts on both surfaces of the element simultaneously.
σ
And the strain is given by: ε = [1 − v ] + α∆T (10.7)
E

Example 10.1: A cylindrical steel gas bottle L=2m long with a diameter of d = 250mm and a t
= 3mm wall thickness is pressurized to P = 3MPa as shown in Fig. 10.4. Determine the
stresses and the strains, Young’s modulus E = 210 GPa and Poisson’s ratio v = 0.3.
σ xx = =
(
Pr 3 × 10 6 × (0.125) )
= 62.5MPa
2t 2 × (0.003)
σ θθ = =
( )
Pr 3 × 10 6 × (0.125)
= 130 MPa
t (0.003)
ε xx =
Pr  1 
− v  + α∆T =
) (
3 × 10 6 × (0.125)  1 
− 0.3 + 0 = 119 × 10 −6 ms = 119µs

tE  2  ( ) 9 
(0.003) × 210 × 10  2 
ε θθ =
Pr  v 
1 −  + α∆T =
(3 × 10 )× (0.125) 1 − 0.3  + 0 = 505 × 10
6
−6
ms = 505µs

tE  2  (0.003) × (210 × 10 )  2 
9
Note: Other shaped thin walled pressure vessels can be analyzed by similarly taking
appropriate Free Body Diagrams and using the equilibrium equations.
j's
10.2 PRINCIPAL STRESSES AND MAXIMUM SHEAR STRESS
(SI&4th:439-473; 3rd 441-475)
Having determined both the axial stresses and the shear stresses in a biaxial stress system, this
ra
does not guarantee that these are the maximum stresses experienced by the actual structure.
Just like when you have two or more forces, it is the resultant force the maximum experience,
so to with a stress system. What is now necessary is to find a stress transformation of
vi
determining what these normal and shear maximum stresses are.
Sign Convention
ith
Before the transformation equations are derived, it is necessary for us to review the sign
convention for the normal and shear stress components. As shown in Fig. 10.7, the sign
convention can be remembered by simply noting that positive normal stress acts outwards
from all faces and positive shear stress acts upward on the right-hand face of the element.
Pr
+σyy Upward in the -σyy Compressive or

+τxy right hand face -τxy inward direction
σxx +σxx σxx -σxx

y y Downward in the
+τyx Tensile or right hand face
-τyx
x outward direction x
Fig. 10.7 Sign convention of stress components
To understand how a structure fails in Plane Stress status it is necessary to resolve stresses in
any direction (similar to resolving force vectors). Plane stress lies in the plane of thin
material such as the pressure vessels above. Consider a small element of sheet metal of
thickness t and under plane stress in a biaxial stress system as shown in Fig. 10.1.
Cut a triangular section, leaving the left and bottom sides and a third side inclined at an angle
θ from the vertical. Two of its surfaces have the normals in the x and y directions, the third
has a normal at an angle θ from the x axis, as shown in Fig. 10.8.

σyy
s n
τxy τsn
σnn
θ
θ
σxx σxx σxx
Acosθ
θ A
θ
Asinθ
y y τxy
τyx θ
x σyy x θ σyy
Fig. 10. 8 FBD of triangular element with all normal and shear stresses
It is now necessary to apply the equilibrium equations about the Normal n & Tangent s axes.
( ) ( )
∑ Fn = 0 = Aσ nn − σ yy A sin θ sin θ − (σ xx A cos θ)cos θ − τ xy A cos θ sin θ − τ xy A sin θ cos θ ( )
which simplifies to give:
σ nn = σ xx cos 2 θ + σ yy sin 2 θ + 2τ xy cos θ sin θ (10.8)
j's
Using the following trigonometric functions:
1 1
cos 2 θ = (1 + cos 2θ ) sin 2 θ = (1 − cos 2θ) sin 2θ = 2 cos θ sin θ
2 2
ra
we can obtain:
(
σ xx + σ yy ) (
σ xx − σ yy )
σ nn = + cos 2θ + τ xy sin 2θ (10.9)
2 2
vi
And in a similar way, by applying equilibrium about 's' axis and using the trigonometric
functions we can get:
( )
ith
σ yy − σ xx
τ sn = sin 2θ + τ xy cos 2θ (10.10)
2
These equations can be used to transform the stresses from one coordinate axis to another.
However there is an easier method of determining the stresses in any axis, this requires Eqs
Pr
(10.9) and (10.10) to be squared and then added them together to give,
 (
σ xx + σ yy 
2
)  σ xx − σ yy 
2
 + τ sn =   + τ 2xy
2
σ nn −  (10.11)
 2   2 
Mohr Circle
In fact, Eq. (10.11) has the same format as the equation for a circle of radius R and centre at x = c.
( x − c )2 + y 2 = R 2 (10.12)
From Eq. (10.11), the radius and centre of the circle can be represented as:
(
σ xx + σ yy )
c= (10.13)
2
2
 σ xx − σ yy 
R =   + τ 2xy
 (10.14)
 2 
Equation (10.12) represents a circle of stress, known as MOHR CIRCLE, where σnn is the
horizontal axis positive to the right and τsn is the vertical axis positive downwards.
The Mohr Circle for Stress looks like the left part of Fig. 10.9:

(σ xx + σ yy ) Orientation of the Principal Stresses
2 σ22
θp2= θ p2+90º
180°
σ11
2θp2 θp1= θ
σ22 σxx σ11 σ
nn
σyy 2θ=2θp1 2
R =   + τ 2xy Orientation of the Maximum Shear Stress
2ϕ 2
 
τxy
σyy= c
τxy τmax
τmax
τmax τmax
σxx= c
τsn
ϕ
Fig. 10.9 Mohr Circle for plane stress status

Principal Stresses
There are two points in this circle where the shear stress is zero, τxy = 0. These are points of
the Maximum σ11 and Minimum σ22 normal stresses and are called PRINCIPAL STRESSES.
Thus the principle stresses can be computed as:
(σ xx + σ yy )  σ xx − σ yy 
2
σ11 ,σ 22 = ±   + τ 2xy
 (10.15)
j's
2  2 
Note that σ11 ≥ σ 22 . As shown in Fig. 10.9, these principle stresses occur at angles of θp1=θ
and θp2=θ+π/2. The principal stresses represent the maximum and minimum normal stress at
ra
the point. When the state of stress is represented by the principal stresses, no shear stress will
act on the element.
vi
Orientations of Principal Stresses
When equation (10.10) is set to equal zero, it gives the orientation of principal stresses:
(
σ yy − σ xx )
ith
τ sn = 0 = sin 2θ + τ xy cos 2θ
2
2τ xy
∴ tan 2θ = (10.16)
(
σ xx − σ yy )
Pr
The solution of Eq. (10.16) has two roots, specifically θp1 and θp2 will be 90° apart as shown in the
infinitesimal element, and 2θp1 and 2θp2 will be 180° apart as shown in Mohr circle in Fig. 10.9.
To find the orientation of the principal stresses the line representing the current state of
stresses is rotated until it reaches the horizontal axis. In the real structure, the element is
then rotated in the same direction as in the Mohr Circle but by half that angle.
Maximum Shear Stress
2
 σ xx − σ yy  σ − σ 22
τ max =   + τ 2xy = 11
 (10.17)
 2  2
The orientation of the maximum shear in infinitesimal element can be identified by rotating
the element by ϕ as in Fig. 10.9. Note that at the maximum shear stress, the normal stress may
not be zero (σxx=σyy=c).
Remarks: Principal stress σ11, σ22 are the two points in Mohr circle that cross the
horizontal axis, in which the shear stresses are zero.
The maximum shear stress τmax is the point in Mohr circle that intersect a vertical line
drawn through the centre of the circle.

Example 10.2 Find σ1, σ2, τmax and their orientations for the following stress system: σxx =
40 MPa, σyy = -20 MPa, τxy = 30 MPa.
Step 1: Draw the state of stresses (σ xx + σ yy )
σyy c=
2
= 10 MPa
τxy=30MPa C(10,0) σxx=40

σyy=-20 0 σ
σxx σxx=40MPa τxy=30
R=42.43 A
y
τyx The first point drawn
x σyy= -20MPa τ σxx=40
Step 2: Draw the Mohr Circle It is necessary to first establish σ and τ axis. Since σxx, σyy ,
and τxy are known, the center of the circle can then be plotted at C(10,0). To obtain the radius,
one can either plot point A (40,30) as shown above or compute the value as follows.
2
σ − σ yy  40 − (− 20 ) 
2
Radius of Mohr Circle: R =  xx  + τ 2xy =   + (30 ) = 42.43MPa
2
  2  2 
 
Center of Mohr Circle:
(σ xx + σ yy )
j's
c = 10MPa 40 + (− 20 )
c= = = 10 MPa
2 2
σ22=-32.43MPa 2θp2=225°
ra
σ11=52.43MPa
C σxx=40 Step 3: Determine the orientation of the
σyy=-20 0 σ principle stress:
2θp1=45°
2ϕ R=42.43MPa 2τ xy 2 × 30
tan 2θ =
(σ xx − σ yy ) = 40 − (− 20) = 1.0
vi
τxy=30 A
arctan(1.0 )
∴ θ p1 = = 22.5° and
τmax
ith
τmax=42.43MPa
2
τ
∴ θ p2 = 90° + θ p1 = 112.5°
Step 4: Compute the principle stresses and the maximum shear stress
σ11 (
σ xx + σ yy )
2
52.43MPa
Pr
= ±   + τ 2xy = c ± R = 10 ± 42.43 =
σ 22 2 2  − 32.43MPa
 
2
 σ xx − σ yy 40 − (− 20) 
2

τ max = R =   + τ 2xy =   + (30) = 42.43MPa
2
2   2 
 
Step 5: Draw infinitesimal elements indicating magnitude and orientations of both the Principal
Stresses (the left hand of the figure) and the Maximum Shear Stress (the right hand of the figure).
From the Mohr circle, σ11 rotates from the bold line by 2θp1 anticlockwise. So we rotate the oriented
element showing the principal stress by θ=θp1 in the same direction as given on the left below.
Similarly, in the Mohr circle, τmax rotates by 2ϕ clockwise; hence the oriented element showing the
maximum shear stress should be rotated by ϕ clockwise as given on the right figure below
Orientation of Principal Stresses Orientation of Maximum Shear Stresses
σ22= -32.43MPa 10MPa
τmax=42.43MPa
σ11=52.43MPa
10MPa
θ = 22.5o ϕ = 22.5o

Example 10.3 When the torque of T=10kN/m is applied to a torsional shaft as shown, it produces
a state of pure shear stress in the material. Determine the detailed stress status at element A for: a)
the maximum shear stress and the principal stresses; b) the orientation of principal stresses.
Cross-section
Distribution of
T=10kNm shear stress
R
A τxy
φ
D=75mm
A
T T
L=15m
Step 1: Compute the shear stress in the cross section

πD 4 π(0.075)
4
J= = = 3.106 × 10 −6 m 4
32 32
TR T (D 2 ) 10 × 10 3 × 0.0375
τ= = = = 120.7 MPa
J J 3.106 × 10 −6
τ xy = − τ = −120.7 MPa ; No normal stress at A, so we have σ xx = σ yy = 0
Step 2: Compute the maximum shear stress

τ max = 

 + τ xy = 
2
0−0
2
j's
 + (− 120.7 ) = 120.7 MPa
2
ra
2
2   2 
 
As expected, the maximum shear stress corresponds to a pure shear status. Through
experiment, it has been found that ductile will fail due to shear stress.
vi
Step 3: Compute the principal stresses and their orientation

ith
σ11 (
σ xx + σ yy )  σ xx − σ yy  (
2
0 + 0)  0−0
2
120.7 MPa
±   + τ xy =  + (− 120.7 ) =
2 2
=  ± 
σ 22 2  2  2  2  − 120.7 MPa
2 × (− 120.7 ) arctan(− ∞ )
Pr
Orientation of principle stress: tan 2θ = = −∞ ∴ θ p1 = = −45°

0 − (0) 2
τxy= -120.7MPa
R=120.7 Orientation of Principal Stresses
C=0
σ22 = -120.7MPa
2θ=2θp1
σ22=-120.7MPa σ11=120.7
σ A
θp1= - 45°
σ11 = 120.7MPa
τmax τ
Thus the first principal stress σ1 = τxy acts at θp1=−45° and the second principal stress σ2=−|τxy| acts
at θp2=90°+θp1=45°. Brittle material fails due to normal stress. That is why when a brittle material
such as cast iron and chalk, is subjected to torsion (because usually its allowable tensile stress is
much smaller than its allowable compressive stress) it will fail in tension at a 45°inclination.

Example 10.4 A single horizontal force P=900N is applied to end D of lever ABD. Knowing
that portion AB of lever has a diameter of 36mm. Determine a) the normal and shear stress
status at element at point H; b) the principal planes and principal stresses at H.
y
B 540m
m
σyy=58.9 τxy=53.1
300mm FBD at Section H
D
Mx=270Nm V=P=900N σxx H σxx=0
H
120mm P=900N T=486Nm y
H τyx
A x
z
A x
x
z
Step 1: Internal loads at the section H
Shear force: V = P = 900N
Torque: T = 900×0.540 = 486Nm
Bending moment: Mx = 900×0.300 = 270Nm
Step 2: Compute normal bending stress at H
MR (− 270) × 0.018 = 58.9MPa , ∴ σ = σ = 58.9MPa
j's
σ=− =−
π × (0.018) / 4
4 yy
I
Step 3: Compute combined shear stress at H
Shear stress consists of torsional shear component T and transverse shear component due to V.
ra
However, transverse shear τV is zero at H from Table 7.1 and Example 7.2. We have
TR VQ 486 × 0.018
τ = τ T + τV = + = + 0 = 53.1MPa ∴ τ xy = τ = 53.1MPa
π × (0.018) / 2
4
J It
vi
Step 4: Determine stress status as shown in the right hand side of the top figure.
Step 5: Compute principal stress and their orientation
ith
σ11 (
σ xx + σ yy )
 σ xx − σ yy  (
2
0 + 58.9 )  0 − 58.9 
2
90.2MPa
±   + τ xy =  + (53.1) =
2 2
=  ± 
σ 22 2  2  2  2  − 32.3MPa
2 × 53.1
Pr
Orientation of principle stress: tan 2θ = = −1.8

0 − 58.9
arctan(− 1.8)
∴ θ p2 = θ = = −30.5° ∴ θ p1 = 90 + (− 30.5°) = 59.5°
2
Step 6: Draw Mohr Circle
C = 29.45
Rotate to σ22 Orientation of Principal Stresses
2θ=2θp2
σ11=90.2
σ22=-32.3 σ11=90.2
θp1= θ +90° =59.5°
σxx=0 σyy=58.9 σ
2θp1 H
2ϕ θp2= θ = -30.5°
R=60.7
σ22= -32.3
τxy=53.1
τmax=60.7
τmax
τ

10.3 STRAIN IN ANY DIRECTION (SI&4th:489-502; 3rd 491-506)
In a similar way to stress, strain can be transformed to determine the maximum direct and shear
strains experienced by the structure. And like stress, a MOHR CIRCLE of strain can be drawn.
Although the derivation of the equations for strain transformation is an interesting
mathematical exercise, you can read section 10.2 to see how this is done. What we are more
interested in however, is the application of the derived equations. What you find when you
do the analysis is that the equations for strain transformation are very similar to the
equations for stress transformation.
Define the terms ε xx' ,ε yy' , γ xy' as the strains of an element of size dx by dy at an angle θ with
respect to the horizontal axis, like this:
y
y’ Deformed Element
dy Original Element
x
dx x’
j's
x’ v’ u’
y’
y
θ x
ra
Fig. 10. 10 Element of size dx×dy at angle θ before and after the application of biaxial strains
Then the equation which defines these strains will be:

vi
ε xx' = ε xx cos 2 θ + ε yy sin 2 θ + γ xy cos θ sin θ (10.18)
ith
A similar equation to stress transformation can be derived to give the equation for the Mohr
Circle of strain:
 ( )
ε xx + ε yy 
2
 γ xy' 
2
 ε − ε yy
 =  xx
2
  γ xy 
2
ε xx' −  +   
 +
  2 
 (10.19)
Pr
 2   2   2   
This is similar to the Mohr circle of stress except that the vertically downward positive Shear
Strain Axis is γxy'/2.
c=
(ε xx + ε yy )
2
ε 22 ε xx ε 11
εnn
εyy 2θ
2 2
 ε xx − ε yy   γ xy 
2ϕ R =   +  
 2   2 
γ xy
2
γ xy′ γ max
2 2
Fig. 10. 11 Mohr Circle for strain

10.3.1 STRAIN GAUGE ROSETTES (SI&4th: 508-519; 3rd 513-525)
Using Eq. (10.18), the strain at any angle can be determined. Inversely, if the strain at any
angle θ has been measured, Eq. (10.18) can then be used to determine the direct and shear
strains in the structure about the x and y axes. One of typical approaches is that these
measurements are done using a Strain Gauge Rosette. A normal arrangement is to have three
strain gauges oriented at three different angles w.r.t the horizontal axis of the structure as in
Fig. 10.12. From experiment, we can acquire three sets of data as εθ1, εθ2, εθ3.
y 3
2
θ3 θ2 θ1
x
j's
Fig. 10.12 Strain Gauge rosettes
Based on these three sets of data, we want to determine the normal and shear strains about the
ra
x and y axes. Because we have three unknown terms and we want to find, εxx, εyy γxy, we can
use Eq. (10.18) three times, once for each angle. Then simultaneously solve for the three
unknown strain terms, εxx, εyy γxy. The procedure is as follows.
vi
Step 1: simultaneously solve for the three unknown strains, εxx, εyy, γxy.
 ε θ1 = ε xx cos 2 θ1 + ε yy sin 2 θ1 + γ xy cos θ1 sin θ1

ith
2 2
ε θ 2 = ε xx cos θ 2 + ε yy sin θ 2 + γ xy cos θ 2 sin θ 2 (10.20)
 ε = ε cos 2 θ + ε sin 2 θ + γ cos θ sin θ
 θ3 xx 3 yy 3 xy 3 3
Step 2: Compute stress components, σxx, σyy, τxy

Pr
Once we find the strains εxx, εyy γxy, then use the relationships between stress and strain to find
the stress components:
)[ ]
 E
σ xx = ε xx + vε yy
 (
1 − v 2
)[ ]
 E
σ yy = ε yy + vε xx
 (
1− v2
(10.21)
 E
 τ xy = Gγ xy = 2(1 + v ) γ xy

Step 3: Determine principal stresses, maximum shear stress and their orientations as well as Mohr
circle when necessary
(σ xx + σ yy )  σ xx − σ yy 
2
2τ xy
σ11 ,σ 22 = ±   + τ 2xy , tan 2θ =
2  2 
 (σ xx − σ yy )
2
τ max =   + τ 2xy

 2 

10.4 THEORIES OF FAILURE (SI&4th: 524-532; 3rd 530-538)
The curve in Fig. 10.13 shows the typical stress-strain response of a ductile material such as
mild steel. Such a stress-strain curve is obtained from a uniaxial tensile test, and elements in
the material experience normal stress in one direction and has no shear stress. The material
yields at yield stress σY as shown.
F -τmax
σ Elastic Yielding Hardening Necking
σ11= σY
σY σY
σ22= 0 σ σP
2θ=90
uY
ε
τmax=σY /2
τ
F
Fig. 10. 13 Tensile testing of ductile material
However, if a material is subjected to a combination of normal and shear stresses, some
j's
combination of stresses will cause the material to yield. The combination of stresses that
produces yielding is known as a yield criterion. It is assumed that the material is ductile,
isotropic and the same behavior in tension and compression.
ra
10.4.1 Tresca’s Yield Criterion (Maximum-Shear-Stress Theory)
Yielding can be considered a shear phenomenon, in which layers of crystals or atoms slip
vi
relative to each other in shear. Hence Tresca’s criterion is based on the maximum shear stress
reaching a critical level. From Eq. (10. 17), we have
2
ith
σ − σ 22 σ Y
τ max =   + τ 2xy = 11
 ≤ (10.22)
 2  2 2
But yield stress σY is in fact found via the uniaxial tensile test as illustrated in Fig. 10.13. The
maximum shear stress τmax corresponds to σY/2 at yielding. Therefore the failure criterion
Pr
should be formulated as
σ Tresca = max{σ11 − σ 22 , σ11 , σ 22 } ≤ σ Y (10.23)
10.4.2 von Mises Yield Criterion (Maximum-Distortion-Energy Theory)

It is possible to formulate a failure criterion based on the distortions caused by strain energy.
As known in Chapter 3, the strain energy density can be computed as u=σε/2. Apply Hooke’s
law, we can derive von Mises’ criterion as (refer to the textbook).
2
σ vm = σ11 − σ11σ 22 + σ 222 = σ 2xx − σ xx σ yy + σ 2yy + 3τ 2xy ≤ σ Y (10.24)
Note that σTresca and σvm are usually called Tresca stress and von Mises stress respectively
for convenience.
Brittle material such as gray cast iron, tends to fail suddenly by fracture with no apparent
yielding. The Maximum-Normal-Stress Theory can be applied, in which brittle material will
fail when the maximum principal stress reaches a limiting value that is equal to the ultimate
normal stress the material can sustain under simple tension.

MoS Lecture Notes

Uploaded by

Copyright:

Available Formats

MoS Lecture Notes

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MoS Lecture Notes

Uploaded by

Copyright:

Available Formats

Chapter 1 - Equations of Statics & Free Body Diagrams

1.1 BASIC DEFINITIONS

1.2 PHYSICAL FUNDAMENTALS

body, although acted upon by many forces or moments experiences no acceleration.

1.3 EQUATIONS OF STATICS (SI & 4th Ed p.4-6; 3rd Ed p.3-5)

Lecture Notes of Mechanics of Solids, Chapter 1 1

1.5 TYPES OF SUPPORTS (4th Ed p.5; 3rd Ed p.4)

Lecture Notes of Mechanics of Solids, Chapter 1 2

Lecture Notes of Mechanics of Solids, Chapter 1 3

Fig. 1.5 Original structure of Example 1.2

F=w×3=30kN (resultant force)

Example 1.3: Determine the ground reactions at A and B of structure as shown.

Lecture Notes of Mechanics of Solids, Chapter 1 4

2.1 STRESS (SI&4th p. 22-23, 3rd p. 22-23)

Normal and Shear Stress

Lecture Notes of Mechanics of Solids, Chapter 2 1

Average Normal Stress (SI&4th p. 24-31, 3rd p. 27-34)

Kilo Pascal KPa (=Pa×103) e.g. 20,000Pa=20kPa

Example 2.1: An 80 kg lamp is supported by a single electrical

Lecture Notes of Mechanics of Solids, Chapter 2 2

Lecture Notes of Mechanics of Solids, Chapter 2 3

Fig. 2.2 Deformation due to applied axial forces

2.3 STRAIN (SI p.68-69; &4th p. 67-69; 3rd p.71-72)

Average Normal Strain

Lecture Notes of Mechanics of Solids, Chapter 2 4

2.4 STRESS-STRAIN RELATIONSHIP, HOOKE'S LAW (SI&4th p.83-93; 3rd

Proportional σpl stress σf

Lecture Notes of Mechanics of Solids, Chapter 2 5

2.5 POISSON'S RATIO (SI&4th Ed p104-105; 3rd Ed p107-108)

F Original shape F Original shape

Strain in Lateral Direction

Lecture Notes of Mechanics of Solids, Chapter 2 6

∆T : algebraic change in temperature ( oC ) (increase +; decrease −)

Coupled Strain Status

Lecture Notes of Mechanics of Solids, Chapter 2 7

Fig. 2.7 Thermal and mechanical deformation

Lecture Notes of Mechanics of Solids, Chapter 2 8

PAB F1=10kN so Segment AB is in compression)

Step 3 Compute the total deformation by using Eq. (2.18)

Lecture Notes of Mechanics of Solids, Chapter 2 9

unknown forces than equations of statics it is called statically indeterminate.

Lecture Notes of Mechanics of Solids, Chapter 2 10

Lecture Notes of Mechanics of Solids, Chapter 2 11

α ∆TLCu + α Al ∆TL Al − 0.005 17 × 10 × 200 × 0.4 + 23 × 10 −6 × 200 × 0.8 − 0.005

Lecture Notes of Mechanics of Solids, Chapter 2 12

2.9 AVERAGE SHEAR STRESS (SI&4th Ed p. 32-39; 3rd Ed p. 35-41)

Fig. 2.10 Average shear stress

Lecture Notes of Mechanics of Solids, Chapter 2 13

Stress Concentration Factor

Lecture Notes of Mechanics of Solids, Chapter 2 14

3.1 STATICALLY DETERMINATE PJF

If J denotes the number of pin-joints:

Lecture Notes Chapter 3, Mechanics of Solid I 1

3.2 JOINT EQUILIBRIUM

Notation of Internal Forces

Lecture Notes Chapter 3, Mechanics of Solid I 2

Tensile Force Compressive

A force pointing away from pin-joint

Fig. 3.1 Notation of internal forces

This is a statically determinate structure