1.

Equilibrium of Forces & Moments

Force
There are many kinds of force, such as gravity, gas pressure,
atmospheric pressure and wind pressure, magnetic and
electromagnetic attractions, nuclear. Four fundamental forces are
recognised in physics relating to the interaction of particles (i)
gravitational, (ii) electromagnetic, (iii) strong forces bonding atomic
nuclei, (iv) weak forces bonding atomic nuclei.

Definition: A force is an action on a body that

maintains or alters its position or distorts it or
changes its velocity if it is moving.

The unit of force is the NEWTON - symbol N in the SI system.

Multiples of this basic unit are commonly used.

Kilonewton kN = 103N
Meganewton MN = 106N
Giganewton GN = 109N

Newton’s laws
Newton's laws concern the action and movement of solid bodies at
rest or in motion. The subject of bodies at rest is termed STATICS,
while that of bodies in movement is called DYNAMICS.

The first law states that - if a body is at rest or moving at constant

speed in a straight line, it will remain at rest or keep moving in a
straight line at constant speed unless it is acted upon by a force.

The second law is a quantitative description of the changes that a

force can produce in the motion of a body. Simply stated it is

Force = Mass ⋅ acceleration

In the SI system, the unit of mass is the kilogram (kg), while the
unit of acceleration is 1m/s2. Hence 1N = 1kgm/s2

The third law states - "Reaction is always equal and opposite to

action". Thus if a book is pushing down on a table with its own

DEN-102: Handout No. 1 Page 1

gravitational force (mass x local specific force due to gravity), there
is an equal and opposite force from the table pushing upwards on
the book - the reaction. In turn the gravitational force of the book
and table on the floor is subject to a reaction from the floor on the
table.

Stress Analysis/Mechanics of Materials

The subject of stress analysis is mainly concerned with statics, ie
bodies and structures at rest under the action of several forces.
The structure is simply a series of 'bodies' joined together.
Examples of structures are:

- rods in tension or compression

- beams in bending
- space frames for buildings
- an aircraft in flight under the action of many forces
- a crank-shaft and piston in an engine
- a composite material with fibres and a matrix.

Stress analysis deals with the effects of applied loads which are
external to a body or structure. The external forces result in
internal reacting forces together with deformations and
displacements.

A system must be defined which is an identifiable collection of

matter on which the attention is focused and analysed. A system
may consist of a single component (e.g. a rod) or a number of
components connected together in a specified manner. It is
sometimes useful to draw a boundary around the system.

DEN-102: Handout No. 1 Page 2

 Figure 1.1

The applied load and the reactions are forces external to the
system. Internal forces are found in the cable and the beam. To
find the internal forces, it becomes necessary to consider part of the
whole system called a subsystem. A diagram showing forces on
the subsystem is referred to as a free body diagram (FBD).

Figure 1.2

Conditions of Static Equilibrium

A body is said to be in equilibrium when the combined action of the
forces and moments acting on that body cause no bodily movement
(i.e. dynamic movement). The forces and moments can, and do,
cause distortion of the body.

DEN-102: Handout No. 1 Page 3

1.1.1 Coplanar, concurrent forces
Consider first forces F1, F2, ... acting through a common point (i.e.
concurrent forces) and in the same plane (coplanar).

Figure 1.3

For equilibrium we have

n
F = ∑
i =1
Fi

where F is the resultant (the vector sum) of all the forces.

It follows that:

Fx = ∑ Fix
Fy = ∑ Fiy

where suffixes x and y denote components in these directions.

Example 1.1 - Resolution of forces at a point:

Fig 1.4 shows three forces acting at a point.

DEN-102: Handout No. 1 Page 4

 y

F1
45o

x
F3
F2

30o

If the system is in equilibrium, the three conditions for equilibrium

apply. Since all forces act through the same point, no moments
operate. Hence only resolution of forces in two orthogonal
directions can be used to determine the values of each force.

Resolving in the y direction

Fy = F1 y + F2 y + F3 y ⇒ Fy = F1 cos(45) + F2 cos(30) + F3 cos(90)

Resolving in the x direction

Fx = F1x + F2 x + F3 x ⇒ Fx = F1 sin(45) + F2 sin(30) + F3 sin(90)

Three unknown forces exist in these equations, F1, F2 and F3, but
there are only two equations. Hence the value of one of the forces
must be defined before the equations can be used to find all force
values absolutely. Set F1 = 10kN. Then

∴F2 = 8.16kN

∴F3 = 11.15kN

1.1.2 Coplanar, non-concurrent forces

The non-concurrent forces may, or may not, be parallel to each
other.

DEN-102: Handout No. 1 Page 5

Moments
lo

F
M

Figure 1.5

A moment is the product of a force acting on lever arm. Consider a

force, F, acting orthogonally at the end of a bracket (or rod) of
length lo. Under the action of the force, the rod will tend to rotate
unless it is constrained. If the rod is held, the action is a moment,
M, equal to

M = l o × F ⇒ M = l o ⋅ F ⋅ sin(90)

Clearly there must be a reaction, preventing the rotation occurring,

the restraint can be replaced by reaction in the opposite sense to
the action (Newton's 3rd law).

Parallel Coplanar Forces

X1

X2
F1 F2 A

F4
F3
X3
X4

Figure 1.6

The resultant force as well as the moment of all the forces about
any arbitrary point (e.g. A) must be zero for equilibrium. Thus

and

DEN-102: Handout No. 1 Page 6

where MA denotes the moment of the forces about an axis through
A, perpendicular to the plane.

Example 1.2
Consider the beam shown in Fig 1.5 subjected to a force, F, at
distance, a, from one support of the beam and b from the other.
Find out how this beam is in equilibrium.

F
a b

Figure 1.7

First replace the supports by equivalent restraints, in this case

vertical reactions, R1 and R2.
F
a b

A
R1 R2
Figure 1.8

(a) Resolving vertically, we obtain

F = R1 + R2

(b) Resolving horizontally does not add further information - there

are no horizontal forces.
(c) Take moments about any point in the system. To eliminate one
of the unknown reactions, we take the point through which one
of these act.

M A = − F ⋅ a + R2 ⋅ ( a + b ) = 0

DEN-102: Handout No. 1 Page 7

Note that the two forces act in opposite directions. R2 acts
anticlockwise, taken as positive, while F acts clockwise, taken as
negative, about A.

Non-parallel Coplanar Forces

F1

a3
F2

a1
a2 F3
a4
F4

Figure 1.9

In this case, the resultants in two directions as well as moments

about an arbitrary point must be zero, i.e.,

DEN-102: Handout No. 1 Page 8

Types of Structural and Solid Body Components

'Tie' prevents parts A and B

A B
F F from moving apart when a
pull is applied.

'Strut' prevents parts A and B

F
B A
F
moving towards each other
when compressive force is
applied.

'Column' supports mass of a

structure

'Cable' - flexible member under

tension

'Beam' – typically a horizontal

member carrying vertical
loading and supported from
below

Cantilever – a beam with one

end built in and the other free
W

Shaft – Transmits a torque T

DEN-102: Handout No. 1 Page 9

1.1.3 Typical X-Sections of Components

Angle
Channel

I-Section
T-section

Circular (solid)
Circular (hollow)

DEN-102: Handout No. 1 Page 10

1.1.4 Types of support and connections

(a) Built in support Equivalent force system

M
Rx

Ry

weld

There will be reactions Rx and/or Ry and a "fixing moment", M, at

the built-in end to keep beam in equilibrium. There is no linear or
angular displacement at the support.

(b) Pinned joint Equivalent force system

Rx

Ry

DEN-102: Handout No. 1 Page 11

There will be no moments but reactions Rx and Ry can exist. No
linear displacement will occur but angular displacement possible
(rotation of link)

(c) Roller support Equivalent force system

Ry

There will be no moment and only a vertical reaction, Ry. Both

horizontal and angular displacement can occur.

(d) Sliding support Equivalent force system

Rx

There is a reaction, Rx and there can be a moment. Vertical sliding

motion is possible.

DEN-102: Handout No. 1 Page 12

2. Stress and Strain

The concept of stress

Stress is defined as the transmission of force across a boundary.
Consider the solid body shown in Figure 2.1 separated by a plane
into two parts A and B. Each part exerts a
force on the other, and over a small area δA
B at a point P. Let this force on part A be δF
in an arbitrary direction.
A P
δF The intensity of this force is δF/δA and the
stress at P is defined as

δA δF
A P σ P = lim (2.1)
δA→0 δA

Figure 2.1

The units of stress are those of (Force/Area).

Hence, typical values are N/mm2 or MN/m2.

The force action has magnitude and direction and in addition, the
plane through P has an orientation. These three quantities must
be specified for the stress σ to be completely known.

Average stresses
2.1.1 Tensile stress.
When a tensile force is
applied to a member, it
gives rise to tensile stresses. F F
Figure 2.2 shows a rod of
uniform cross-section, area Area Ao
Ao, subjected to an axial
tensile force F. Figure 2.2

DEN-102: Handout No. 1 Page 13

The average stress throughout the rod is tensile and is given by

F
σ= (2.2)
Ao

Tensile stresses are assigned as positive (+ve). Tensile stresses

tend to stretch or extend the length of the member.
Consider the tapering Area A 1
rod of length l, of area A rea A o
A0 at one end and A1
at the other, shown in
F F
Figure 2.3. x
The force is uniform
l
throughout the rod.
The area at any distance
x from the smaller end is equal to: Figure 2.3

x
A = Ao + ( A1 − Ao ) (2.3)
l

The average stress at any value of x is tensile and is equal to

F F
σ av = = (2.4)
A A + x (A − A )
o 1 o
l

Hence the stress varies along the length of the rod, from a high
value of F/A0 at the smaller end to the lowest F/A1 at the wide end.

We may also look at

the stresses
p
generated in a bar F P Q F
in terms of the
forces transmitted q
along the bar in
conjunction with equilibrium. Figure 2.4

DEN-102: Handout No. 1 Page 14

Consider a
straight tie bar p p
of uniform X- F P Q F
section loaded
by an axial q q
tensile force, F (Figure 2.4). Figure 2.5

Imagine that the bar is 'cut' into two parts P and Q at some section
pq at right angles to the direction of F - that you are standing in
the middle holding the 2 parts together (Figure 2.5).

We can then consider a free body consisting of part P only, drawing

the FBD (Figure 2.6). Equilibrium
requires that the force F on its LHS be p
balanced by a force F to the right. This
force can only be the force applied by F P
part Q on part P at the cross-section
pq. We assume that the force F is
distributed uniformly over the cross-
q
section An. Figure 2.6

Since the stress acts normal to the area, it is called the normal
stress. In the example considered, the force F is tensile, i.e.
tending to stretch the bar. The normal stress, σ, is thus along the
outward normal to the surface, i.e. it is a tensile stress.

2.1.2 Compressive stress

When a compressive force is
applied to a member, it F F
gives rise to compressive
stresses. The average Area Ao
stress throughout the rod
shown in Figure 2.7 is compressive,
equal to Figure 2.7

F
σ =− (2.5)
Ao

Compressive stresses are assigned as negative. Compressive

stresses tend to compress or shorten the length of the member.

DEN-102: Handout No. 1 Page 15

Note that the average stress depends only on the total area of the
section at any position. Hence the shape of the section is
irrelevant. The shape of the section can be circular, square,
rectangular, hollow (tubular) or irregular.
F
2.1.3 Shear Stress
Materials can also be subjected to
shear force. A shear force tends to
slide (or shear) one part relative to
another. Consider the uniform cross- Area A o
section square bar, area Ao in a shear
block shown in Figure 2.8. The left F
hand side of the shear block is moving Figure 2.8
upwards under a force F, while the right
hand side is moving down under an equal
opposing force F.

The material in the block is F

sheared along the dotted line and
is subject to a shear force across
the section (Figure 2.9). The
average shear stress tending to cut Area Ao
or shear the bar is
F
F
τ= (2.6) Figure 2.9
Ao

Now suppose that the bar m

considered in Figure 2.5 F P Q F
is 'cut' along 'mn' where Θ
mn is not normal to F as n
shown in Figure 2.10.
Figure 2.10
Again consider part P as a free body
(Figure 2.11). Let A be the cross Fn
sectional area of section mn, larger m
F P
than the 'normal' cross sectional area, Θ Fs
An. In this case, n
part Q exerts a net force F onto P axially. Figure 2.11

DEN-102: Handout No. 1 Page 16

This may be resolved into

(a) a force, Fn = F sinΘ normal to area A

(b) a force, Fs = F cosΘ along the surface

As above the normal force Fn divided by A is the normal stress

denoted here by σ.

Fn
σ= (2.7)
A

The force Fs along the surface is called a shear force and Fs/A is
called the shear stress, denoted by τ:

Fs
τ= (2.8)
A

The shear stress τ will cause shear (angular) deformation.

In many situations, combinations of tensile, compressive and shear

stress can act simultaneously, e.g. stresses in a simple rivetted
joint.
2.1.4 Shear stress in a rivet in single shear
Consider the forces acting
on the rivet shown in
Rivet (dia. d)
Figure 2.12. The top plate, F t
which is being pulled to F
the left by force F, presses
against the top half of the
rivet and exerts a force F to the left. Figure 2.12

Similarly, the lower plate

exerts a force F on the lower
F
half of the rivet (Figure 2.13). σb
These two forces tend to shear m n Fs
σb
the rivet across its cross- F
section mn. This is resisted
by shear force FS (= F ) acting
over the cross-section of the rivet at mn.
Figure 2.13

DEN-102: Handout No. 1 Page 17

Thus, there are shear stresses τ acting across the section mn which
has an area of πd2/4. If we assume the stresses to be uniform over
this area, the shear stress is given by

F
τ=
(
πd 2 / 4 ) (2.9)

In addition to the shear stress, a contact stress called bearing

stress, is also induced. Assuming that the bearing stress is
uniformly distributed, it is defined as the force divided by the
projected area of the appropriate part of the rivet. Considering the
top half, the projected area is a rectangle of dimensions t x d. Thus

F
σb = (2.10)
txd

The bearing stresses are direct stresses which tend to crush the
rivet, and also cause permanent deformation or enlargement of the
rivet hole.

DEN-102: Handout No. 1 Page 18

2.1.5 Rivet in double shear
Consider a rivetted joint shown in Figure 2.14. In this case, too,
there will be
shear stresses F/2
and bearing F
stresses. If the
thickness of F/2
the three
plates is the F/2 σb
same, the F/2
bearing stress F 2σb
on the central F/2
third projected F/2 σb
area is twice
that on the top (a) (b)
and bottom (c)
third projected areas. Figure 2.14

As shown in Figure 2.14, the force tending to shear the rivet is F/2,
i.e. half that for a rivet in single shear. Therefore, the shear stress
in the rivet is given by
( F / 2)
τ= (2.11)
(πd 2 / 4)

Stress concentrations
The previous sections assumed that the applied force is uniformly
distributed across the loaded section, hence the use of the phrase
average stress. In situations where the cross-section changes
rapidly, local stresses can by higher than the average stress.
F F
Consider a wide strip,
subjected to an axial w 3 σav
tensile force, F, σav
containing an b x
elliptical hole of major a
diameter a and
minor diameter b as F F
shown in Figure 2.15.
Figure 2.15

DEN-102: Handout No. 1 Page 19

Let the width of the strip be w and its thickness be t as shown in Fig
3.15.
The average tensile stress in the strip, well away from the hole is

F
σ av = (2.12)
wx t

The stress at the tip of the ellipse is increased by a factor

(1 + 2(a / b))
For a circular hole (a = b), the stress at the horizontal diameter of a
circular hole is increased by [1+2(a/a)]= 3 over the average stress
in the strip.

The stress at point x is 3σav and decreases rapidly as the distance

from the hole increases, reducing to σav well away from the hole,
shown in Figure 2.15.

Since stress can lead to the failure of materials, failure is likely to

initiate from the positions of stress concentration. This has serious
implications in many engineering situations e.g. rivetted joints as
considered previously.

Non-uniform force / stress distribution

We have assumed that the normal and shear forces are uniformly
distributed over the surface, A. In general this need not be the case
and the force per unit area, over equal elements of area, will vary
with the position of the element.

Consider an element of area δA in the whole area A as shown in

Figure 2.16. Let δFn and δFs be the
normal and shear forces on this area. δF
n
The local normal and shear stresses will
be
δF dF
σ = lim n = n
δA→0 δA dA δA δF
(2.13) s
δFs dFS
τ = lim =
δA→0 δA dA
Figure 2.16

DEN-102: Handout No. 1 Page 20

It is clear that the total forces Fn and Fs are obtained from

dFn = σ ⋅ dA
∴ Fn = ∫ σ ⋅ dA (2.14)
A

Similarly, FS = ∫ τ ⋅ dA where the integrals are taken over the whole

A
area A.
Complementary Shear stress
A block of material is subject to shear stress,
τ, separated by distance y as shown in Figure τ
2.17.

The resultant shear forces constitute a couple

and would lead to rotation of the block. For y
equilibrium, additional shear stress must act
around the block in a complementary manner
so as to prevent the block from rotating, i.e
they should provide an equal and opposite τ
couple. This implies that each shear stress
acting on a particular plane must have its Figure 2.17
complement acting on a plane
orthogonal to it.

Consider a small element

of sides ∆x, ∆y and ∆z in τxy
rectangular cartesian co-
ordinate system (x,y,z) as D C
shown in Figure 2.18. Let
the shear stress on its top y τyx τyx ∆y
surface be τxy where x
indicates the direction in A B
which τ acts and y o x τxy
indicates that the normal
to the surface is in the y- ∆x
direction. The force on the
element in positive x-direction will be:
Figure 2.18
τ xy ⋅ ∆x ⋅ ∆z

DEN-102: Handout No. 1 Page 21

For the element to be in equilibrium, there must be a force of the
same magnitude but in the -ve x-direction. Thus shear stress on
lower face must be τxy in the negative x-direction.
The forces (τxy ∆x ∆y) in opposite directions on AB and CD are in
linear equilibrium but they constitute a clockwise couple:
r
M = τ xy ∆x∆y∆z (2.15)

For equilibrium there must be an anticlockwise couple of the same

magnitude. This can only be provided by the shear stresses acting
on the left and right faces AD and BC. Let τyx be the shear stress on
these faces [subscript y denotes the direction of the stress and x
the direction of the normal to the surface]. The forces on AD and
BC are τ xy ∆x∆y∆z and the anticlockwise couple provided by them is
s
M = τ yx ∆y∆z∆x (2.16)

For equilibrium, clockwise M = anticlockwise M. Therefore, from

Eq. 3.15 and 3.16,

τ xy = τ yx (2.17)

τxy and τyx are called complementary shear stresses. Henceforth

we shall use only one symbol τxy for shear stresses on all four faces.

In Figure 2.18, we assumed that τxy on the top face CD acts in the
+ve x-direction. This shear
stress is taken as positive.
y Outward Normal
If we had assumed that the
shear stress on the top face Positive Faces
is in the -ve x-direction then o x
all the signs of other stresses
would be reversed. To make
the sign convention clearer,
we first define the +ve face of
an element as one for which Negative Faces
the outward normal is in the
+ve direction of the axis as Figure 2.19
shown in Figure 2.19.

DEN-102: Handout No. 1 Page 22

Thus a shear stress is +ve if it acts in the +ve direction of the axes
on the +ve face and in the -ve direction of the axes on the -ve face.
Thus, stresses shown above are +ve. We have omitted
consideration of the normal and shear stresses in the z-direction.
However, everything will also work in this direction.

The concept of Strain

Under the action of the applied stress, the material deforms. Such
deformation is measured in terms of strain. Two types of strain
exist, matching the two types of stress. Normal stresses (tensile or
compressive) are associated with normal strains. Shear stresses
give rise to shear strains.

2.1.6 Normal strain

If a prism of
uniform material is
subject to a
uniform normal e
stress, it will L
elongate by an
amount e as shown in Figure 2.20.
Figure 2.20

The normal strain ε is defined as the extension (or compression)

per unit length, i.e.
e
ε= (2.18)
l
This is a dimensionless quantity. Tensile strains (i.e. those arising
from tensile stresses) are considered positive, since the sign of e is
positive. Compressive strains are assigned negative values.

2.1.7 Shear strain

A plane element (as the dotted square in Figure 2.21a) if subjected
to equal shears on all four faces will deform as shown in a
parallelogram. The angle subtended by any face w.r.t. either the
vertical or the horizontal plane changes by an amount φ . φ is
defined as the shear strain, and like normal strain, is a
dimensionless quantity. Usually in engineering applications, an
axis parallel to one of the

DEN-102: Handout No. 1 Page 23

deformed sides is taken φ φ
as shown in Figure
2.21b, and γ = 2φ is then
taken as “engineering”
shear strain. Note that
both γ and φ are angles (a) (b)
measured in radians.
Figure 2.21

2.1.8 Volumetric
strain

Volumetric strain is ∆ly

defined as the
change in the
volume divided by y
the original volume. ∆lz
z
∆V ∆ lx
εV = (2.19) x
Vo
Figure 2.22

Consider a rectangular parallelopiped whose sides before loading

are lx, ly, and lz in x, y and z directions respectively (Figure 2.22).

Therefore, the initial volume V = lx ly lz .

Suppose the element is loaded so that εx, εy, and εz are the strains
in the x, y and z directions respectively.

∆l x ∆l y ∆l z
εx = εy = εz =
lx ly lz

Then, the new lengths are l + ∆l or

DEN-102: Handout No. 1 Page 24

 l x (1 + ε x ) l y (1 + ε y ) l z (1 + ε z ) (2.20)

Therefore, the increase in volume is

[ ]
∆V = Vnew − Vold = [l x (1 + ε x )] ⋅ l y (1 + ε y ) ⋅ [l z (1 + ε z )] − l x l y l z (2.21)

We are dealing with small strains, i.e. ε of the order of 0.01 or less.
Therefore, ε2 and ε3 terms are negligble, and

∴ ∆V = l x l y l z (ε x + ε y + ε z ) = Vo (ε x + ε y + ε z ) (2.22)

The volumetric strain εv is then given by

∆V
∴ εV = =εx +εy +εz (2.23)
Vo

Stress-strain relationships
The relationship between the strain as shown in Figure 2.20 and
the stress applied to produce this strain can exhibit one of the four
characteristic behaviours illustrated in Figure 2.23a to 2.23d.

σ σ σ σ
x

x
ε ε ε ε
(a) (b) (c) (d)
Figure 2.23

In each of these, the slope of the curve relating stress to strain is

the vital factor. This slope is symbolised by E. E is defined as the
ratio of incremental change in stress (dσ) to incremental change in
strain (dε).

DEN-102: Handout No. 1 Page 25

• In Figure 2.23a, E = ∞ , i.e. there is no strain and the material is
said to be rigid.

• In Figure 2.23b, E = constant, and the stress-strain relationship

is linear.

• In Figure 2.23c, E = f(σ).

• In Figure 2.23d, E = 0: there is no definite strain for any value of

stress and the material is said to be plastic.

The behaviour on unloading can follow either the full or the dotted
curves shown in Figure 2.23b and 2.23c. Whenever the material
returns to its original dimension along the loading curve, it is said
to be elastic, but if it behaves as indicated by dotted lines, i.e. if it
departs from the original loading curve, it is said to be inelastic.

Materials which fracture when the strains are small are known as
brittle whilst materials which undergo considerable deformation
before failure are said to be ductile.
Real materials generally have complex stress-strain relationships,
but for purposes of analysis the behaviour is simplified into either:

• that exemplified by the single curve of Figure 2.24a, or

• behaviour as rigid material up to some yield stress σy, after which

it is plastic (Fig.2.24b), or

• a combination of linear elastic and plastic behaviour (Fig. 2.24c).

σ σ σ

O ε ε ε
(a) (b) (c)
Figure 2.24

DEN-102: Handout No. 1 Page 26

A uniform rod of length L
A
and cross-sectional area
A (Figure 2.25) which is
σ
subjected to a uniform
state of stress σ, i.e. a
force F=σA., and with a
F
} F

value of E which is L e
constant for the stress
range used, has Figure 2.25

e
Strain =
L
F
Stress =
A
Stress FL
Hence, E= = (2.24)
Strain Ae

This constant value of E is called Young’s modulus or Modulus of

Elasticity. At any point A in curve of Figure 2.24a, the stress-
strain relationship can be defined in two ways. One is the tangent
(dσ/dε), i.e. the slope of the curve at A, and this value is known as
the tangent modulus. The other is the slope σ/ε of the line OA, and
this is known as the secant modulus. For linear elastic material,
these two modulii are the same.
For shear strain as in Figure 2.21c, similar behaviour exists but the
relationship between the shear stress τ and the engineering shear
strain γ is known as the shear modulus or modulus of rigidity and
is denoted by G. It is given by
τ
G= (2.25)
γ

Poisson's Ratio
The element of Figure 2.26, which initially has the shape indicated
by the full lines, will deform under a tension as shown to the shape
indicated by the dotted lines.
The longitudinal strain εx is εy
accompanied by a lateral strain σx σx
εy and the ratio ex/ey is known
as the Poisson’s ratio and is
generally denoted by ν. εx
Figure 2.26

DEN-102: Handout No. 1 Page 27

If, in addition to the force producing stress σx, there is a force on
the perpendicular plane producing stress σy then the strains εx and
εy due to combined stresses are given by

εx =
1
(σ x − νσ y ) (2.26)
E

εy =
1
(σ y −νσ x ) (2.27)
E

Poisson’s ratio not only gives the relationship between the

longitudinal and lateral strains, but also enables us to formulate
the relationship between the modulus of elasticity and modulus of
rigidity. Consider an element ABCD of unit length sides as shown
in Figure 2.27 which is subjected to shear stresses τ on each face.

This produces a direct tension and compression of magnitude τ on

the two diagonals. The strain on the diagonal BD is therefore equal
to
A B B’
1 τ
(τ + ντ ) = (1 +ν ) (2.28)
E E

If the effect of the shear is to BB'

cause the deformation shown 2
by the dotted lines in Figure
2.27 with B moving to B’, then D C
BB’ is the shear strain and is τ τ
approximately equal to τ/G.

The increase in length of the

diagonal DB is BB’/√2, i.e. τ τ
τ/G√2, and since the original
length was √2, the strain is σn=τ σn=τ
τ/2G. But this strain
is also equal to (τ/E)(1+ν).
Figure 2.27

DEN-102: Handout No. 1 Page 28

Hence,
1 +ν 1
= (2.29)
E 2G
or
E
G= (2.30)
2(1 + ν )

Generalised Hooke’s Law

For a linear elastic material, the stress is directly proportional to
the strain with Young’s modulus E as the constant of
proportionality:

σ
σ = E ⋅ε or ε= (2.31)
E

The above relationship between stress and strain is deceptively

simple and is an incomplete statement of the state of stress in the
material. It considers only one stress and one strain (in the
direction of the applied stress). Let us index this direction as x.
Hence,
σx
εx = (2.32)
E
Due to the Poisson’s Ratio effect, there must also be strains in the
two orthogonal directions to direction x, which will be called
direction y and direction z, and are known as secondary strains
(Figure 2.28).

These strains are given by y

z
ν
ε y = −νε x = − σx (2.33)
E
x
and

ν
ε z = −νε x = − σx (2.34)
E
Figure 2.28

DEN-102: Handout No. 1 Page 29

Hence, the uniaxial stress, σx, applied in direction x gives rise to
three mutually perpendicular strains or a component of strain in
the three perpendicular directions.

σx − νσ x −νσ x
εx = εy = εz = (2.35)
E E E

Alternatively the single stress could be applied in direction y as σy

(where σy ≠ σx). The strains due to σy would be:

−νσ y
εx = in x-direction
E
σy
εy = in y-direction (2.36)
E
−νσ y
εz = in z-direction
E

Similarly, the single stress could have been applied in direction z as

σz (where σz ≠ σy ≠ σz). The strains due to σz would be:

−νσ z
εx = in x-direction
E
−νσ z
εy = in y-direction (2.37)
E
σz
εz = in z-direction
E

In the most general case, the three components of stress (σx, σy, σz),
could be simultaneously applied to a block of material. The
resulting strains in the three directions would then be the sum of
the separate components.
By simply superimposing (or adding) equations 2.35, 2.36 and
2.37, the general linear elastic stress-strain relationships are
derived:
σ x νσ y νσ z
εx = − −
E E E
−νσ z σ y νσ z
εy = + − (2.38)
E E E
νσ x νσ y σz
εz = − − +
E E E

DEN-102: Handout No. 1 Page 30

The equations are usually rearranged and written in the form:
[ ]
ε x = σ x −ν (σ y + σ z )
1
E
1
[
ε y = σ y −ν (σ x + σ z )
E
] (2.39)

[ ]
ε z = σ z −ν (σ x + σ y )
1
E
These are termed the Linear elastic constitutive equations or
Generalised Hooke’s Law. These equations form the basis for
analysis of many mechanics of materials problems, where the
deformation is (or can be assumed to be) linear elastic. They are
subject to the following constraints:

(a) Linear-elastic deformation

(b) Small scale deformation
(c) Time independent deformation

Thermal stresses and strains

Most engineering materials expand when their
temperature is raised. If this expansion is
l1
restrained, then stresses may be induced. A t te m p e ra tu re T 1
Consider a rod of length l1 at temperature T1
being held between rigid supports (Figure
2.29). If the temperature is raised to T2, then
the rod, if its expansion was not restrained, l2
would expand to a length l2. In restraining the A t te m p e ra tu re T 2
expansion, the supports exert a compressive
load on the rod. Figure 2.29

2.1.9 Thermal strain

We wrote the Generalised Hooke's
Law in the form
l
1
E
[ ]
ε x = σ x −ν (σ y + σ z ) (2.40)
l1
with similar equations for the linear
strains, εy and εz, in the y and z directions.
Figure 2.30

DEN-102: Handout No. 1 Page 31

Consider now a piece of material with length lo in the x-direction at
temperature To (Figure 2.30). If the temperature is raised to T and
the rod is free to expand, the increase in length will be

δl = l1 −l o = α (T − To ) (2.41)

where α is the coefficient of linear expansion of the rod material.

The quantity (δl/lo) is clearly a linear strain arising from
temperature changes. It is called the thermal strain and denoted
by εt. It is given by

δl
εt = = α (T − To ) (2.42)
lo

It is clear that if the material expands equally in all directions (α is

the same for x, y, and z directions as the material is isotropic) then
this strain will be equal in all directions.

Thus when temperature changes occur, the total strain in any

direction must also include the thermal strain

εx =
1
E
[ ]
σ x −ν (σ y + σ z ) + α (T − To ) = (ε x )σ + ε t (2.43)

where (εx)σ is the strain resulting from the stresses σx, σy, and σz.
Similar equations will hold for εy and εz . Note that εt is the same
for all directions.

2.1.10 Thermal stress

Let us return to the simple problem we
started with and find the stress arising from
the restrained expansion (Figure 2.31).
In the x-direction there is no net strain, l1
At temperature T1
i.e. εx = 0. Also σy = σz = 0 since there is no
restraint in these directions nor is there any
imposed load. Thus
1
0= [σ x − 0] + α (T2 − T1 ) ∴ σ x = −αE (T2 − T1 )
E l1
Since T2 > T2 it follows σx < 0, i.e. the At temperature T2
thermally induced stress is compressive
as we had intuitively expected. Figure 2.31

DEN-102: Handout No. 1 Page 32

2.1.11 Practical Example
A simple example is that of tie-bars in old houses. The walls of old
houses tend to buckle outwards with the lapse of time under the
weight of the roof or movement of the foundations. To stop the
house collapsing, tie-bars are inserted to pull the side walls
together. The ends of such bars are passed through ‘X’ or ‘S’ plates
to spread the load on the walls. Rather than tensioning the bars by
turning the nuts on the tie-bars, it was customary to light a fire
under the tie-bar, thus causing it to expand. Then the nuts could
be easily nipped up. As the tie-bar then cooled, it contracted,
pulling the walls together. Clearly the tie-bar had to be heated to
the right amount, otherwise problems could occur.

DEN-102: Handout No. 1 Page 33

3. Beams - flexural loaded members
Concentrated or
Distributed load point-load
non-uniform

Beam
Reaction
Reaction

Figure 3.1

Beam supports may be simple (e.g. roller supports) or built in.

The stresses (or internal forces) in the beam material will depend
on:
(a) magnitude and distribution of loading
(b) nature of the supports
(c) size and shape of the beam corss section

The strains (proportional to the changes in length and angle)

depend on
(a) stresses
(b) properties of the beam material (E, ν)

First we can consider the internal forces and moments reulting

from a specific loading case.

The concepts of 'shear force' and 'bending moment'

Consider e.g. the loading on a skate board. This is essentially a

simply supported beam with a load which is (nearly) concentrated
at a point.
W
Subsystem α
Subsystem β
Cut

α β
z
a b RB
RA
L

Figure 3.2

DEN-102: Handout No. 1 Page 34

3.1.1 Equilibrium

For equilibrium, resolve y-wise & take moments about A or B for

the whole system

Fy = O W -RA - RB = O
∴ RA + RB = W (3.1)

MB = O Wb - RAL = O
Wb
∴ RA = (3.2)
L
Wa
Therefore RB = (3.3)
L

Next, to find how the load is sustained internally by the beam

material consider subsystems A and B and draw free body
diagrams for them

W
M M F
α β
F
z
RA

Figure 3.3

Consider the equilibrium of subsystem A.

For vertical equilibrium there must be a force F = RA acting

downwards.
Further RA has a moment, RA z (clockwise) about an axis at the cut.
The subsystem will rotate unless there is a moment , M= RA z
(anticlockwise) about the cut.

The only way the balancing force and moment can be applied, is if
subsystem B applies them on bubsystem A at the cut. By Newton's
3rd law - A must apply an equial and opposite force and moment
on β.

DEN-102: Handout No. 1 Page 35

F is called the 'shear force'. It prevents A from shearing off relative
to B.

α F
β α

RA F
RA

Figure 3.4

M is called the 'bending moment'. It is associated with the tendency

of the beam to bend.

M
M
α
β
RA

Figure 3.5

Before considering how F and M vary along the length of a beam,

we need to adopt some coordinate axes and a sign convention for
+ve F and M.

Axes: we use a right handed co-ordinate system with axes ox, oy,
oz

z
O

x
y

Figure 3.6

i.e. if a screw with a right handed thread is turned in the sense oy

to oz (anticlockwise in figure), the screw would move in the +ve x-
direction.

DEN-102: Handout No. 1 Page 36

We shall take the beam to lie along the z-axis
W+

O z

x
y

Figure 3.7

For a horizontal beam we shall take +ve direction of load to be

vertically downwards, i.e. in the +ve y direction.

3.1.2 "+ve and -ve faces" of a part of a beam

-ve face +ve face

z z
outward
movement
from face

x z z+δz
y

Figure 3.8

Right handed face (i.e. one with larger value of z) is the +ve face or
the +ve face is one for which the outward normal is in the +ve
direction.

3.1.3 +ve Shear Force

The shear force is +ve if it acts in the +ve y-direction on the +ve face
and in the -ve y-direction on the -ve face.

-ve face +ve face

z

x
y

Figure 3.9

DEN-102: Handout No. 1 Page 37

3.1.4 +ve Bending Moment

The B.M. is +ve if it tends to "sag" the beam i.e. to make it concave
on upper face.

-
Sagging moment Hogging moment

Figure 3.10

Note: Some books use the opposite convention for +ve SF and +ve
BM. Whatever convention you want to use, stick to it.

Determination of Shear Force and Bending Moments

Obtain expressions for F and M at the section shown

section
W1 W2

R2
R1

Figure 3.11

Consider subsystem consisting of beam from one end (RH or LH

end) and draw a f.b.d. for it. Let F and M be the shear force & B.M.
at the section. Show them in the +ve sense in the f.b.d.
W1 W2

M
F

z2
R1
z1
z

Figure 3.12

DEN-102: Handout No. 1 Page 38

For vertical equilibrium

F + W1 + W2 - R1 = 0

F = R1 - W1 - W2 (1)

For moment equilibrium, take moments about axis shown

M + W1 z1 + W2 z2 - R1 z = 0

∴ M = R1 z - W1 z1 - W2 z2 (2)

From (1) F = sum of all upward forces on subsystem i.e. all

loads/reactons to the left of the section.

M = sum of all clockwise moments about axis due to forces to

left of the section.

Shear Force (SF) and Bending Moment (BM) diagrams - are

graphical displays of the variation of F & M along length of beam.

DEN-102: Handout No. 1 Page 39

Example 1

Simply supported beam with concentrated loads. e.g. Painter (mg =

700 N) on a wooden plank supported on tressles. Dimensions as
shown. Draw the shear force and bending moment diagrams

1 2 3 4

0x z
A C B
y
0.25 m 1m 2m 0.25 m
R1 R2

DEN-102: Handout No. 1 Page 40

Example 2

A cantilever beam with concentrated load

1 2

C 0x z

A B Mc
z y
a Rc
L

DEN-102: Handout No. 1 Page 41

The method of considering subsystems and drawings f.b.d. to get F
and M seems cumbersome but in initial stages it helps to get
insight into internal forces/moments. After practice, it is perfectly
accpetable to imagine a 'cut' anywhere along the beam and to find F
as the total upward force to the left of the cut. Similarly for M.

Some General Relations among Loading Intensity, Shear Force

and Bending Moment

x
y
z
δz

At any point z, load/length = w (in general, varying with z).

Consider a short length δz of beam which carries point loads and a
distributed load.

Over this length, load/length = w

∴ load = w δz

Let F, M = SF and BM on LH (-ve) face

and F + δF, M + δM = SF & B M on R H (+ve) face

load

M
M+δM
F

beam Wδz F+δF

δz

DEN-102: Handout No. 1 Page 42

For vertical equilibrium -F + F + δF + wδz ≈ 0
∴ δF ≈ - wδz
δF
=- w
δz
Αs δz → 0, δF/δz → dF/dz so that
dF
=-w (3.3.1)
dz
Note that w may vary along beam length i.e. w = w (z).

Between positions z1 and z2

F2 z2
F2 - F1 = ⌠ dF = - ⌠ wdz = (3.3.2)
⌡ ⌡
F1 z1

= minus area under graph of w v z from z1 to z2

Wδz

z1
Z2
z δz

If w = 0, then dF/dz = 0 i.e. F = constant

If there is a point load, e.g.

W F2
F1

then we have no variation of w with z, so for equilibrium

W + F2 - F1 = 0
∴ F2 - F1 = -W (3.3.3)

For moment equilibrium (moment about horizontal axis through O)

δz
(M + δM) - M - Fδ z + (w δz) 2 = 0

DEN-102: Handout No. 1 Page 43

 δz
δM = Fδz - (w 2 δz)

δM w
∴ = F- δz
δz 2

δM dM w
As δz → 0, → and δz = 0. Thus
δz dz 2

dM
∴ = F (3.3.4)
dz

Between z1 and z2
M2 z2
M2 - M1 = ⌠ dM = ⌠ Fdz = B
⌡ ⌡
M1 z1

= area under graph of F v z from z1 to z2

If a couple Mo is applied at a point along beam

M1 Mo
M2

then for equilibrium

M2 + Mo - M1 = O

∴ M2 - M1 = - Mo

DEN-102: Handout No. 1 Page 44

4. Deformations and stresses in pure bending

M M

F
F
F=M=0

Figure 4.1

Initially straight beam will bend due to application of M. The lower

part of beam is stretched (in tension) and uppermost part is
compressed.

There is also shear deformation due to shear force F. We can

neglect this as it is small for long slender beams.

Assumptions (refer to figure)

(a) Plane transverse sections remain plane after bending, i.e. PQ,
AC etc remain plane while PAB is bent to P'A'B'

(b) Beam bends in circular arc in vertical plane only; transverse

sections are perpendicular to circular arcs having common
centre of curvature. But radius of curvature & centre of
curvature will vary from point to point along beam.

(c) Beam is subject to pure couples (M) resulting in tension &

compression in the z-direction; there are no shear stresses.

(d) Radius of curvature, R >> transverse 'depth' of beam

(e) E is the same in tension and compression

(f) Beam x-section has vertical axis of symmetry 0y - simplifies

presentation.

DEN-102: Handout No. 1 Page 45

Neutral Surface

Consider initially parallel sections AC and BD, a distance δz apart.

After bending:

CD stretches to C'D' i.e. εz +ve

AB contracts to A'B' i.e.εz -ve

Between AB and CD there is some layer FG which neither stretches

nor contracts, i.e. for which εz = 0
∴ straight length FG = curved length F'G' = δz

Layer FG is called the NEUTRAL SURFACE and axis 0x on the x-

section is called the NEUTRAL AXIS. Take z-axis to be in the
neutral surface and through the line of symmetry of the cross
section.
-y
δz Neutral surface
x
A B Neutral axis
P A B

OZ x
O z F G
x H K
Q C D
+y
x Area εA C D

+y
Section XX

M M

R+y R δθ

P’
A’ B’

y F’ G’ Neutral surface
Q’ K’
H’

C’ D’

Figure 4.2

DEN-102: Handout No. 1 Page 46

 σz E
Bending equation - y =R

Consider layer HK at distance y from the neutral axis. This

stretches to H'K' in the bent position. The longitudinal (i.e. z-wise)
tensile strain in HK is

H'K' - HK
εz =
HK
(= increase in length
initial length
) (4.1)

Now HK = FG = F'G' = δz = Rδθ (4.2)

where R = radius of curvature of neutral surface

δθ = angle (in radians) subtended by F'G' at the centre of
curvature

Also H'K' = (R + y) δθ (4.3)

(R+y) δθ - Rδθ y
∴ εz = = (4.4)
Rδθ R
i.e. εz varies linearly from 0 on the neutral axis to a maximum at
the bottom (C'D') and top (A'B'). Since y is +ve at the bottom, εz is
+ve there. And εz is -ve at top.

From Hooke's Law

1 σz
εz = [σz - ν (σx + σy)] =
E E
since in this case σx and σy are zero. Substituting εz = σz/E in (4.4)
gives

σz y σz E
= i.e. =
E R y R
(46.5)

E
i.e. σz = y and varies linearly with y in same way as εz.
R

DEN-102: Handout No. 1 Page 47

 M E
Bending equation - Ix =R

Neutral axis OZ
A B

σz
OZ x
O

dF2=σ2dΑ
Area dΑ C D
y
Axis of symmetry

+y

Figure 4.3

The force, δFz on an element of area, δA

E
δFz = σz δA = y δA
R
and its resultant is obtained by integrating δFz over whole the area
A, i.e.

E E
Fz = ⌠ dFz = ⌠ ydA= ⌠ ydA (4.6)
⌡ ⌡R R⌡
A A A

Consider the equilibrium of portion P'Q'D'B'. Since loading on P'Q'

is just a pure couple M (clockwise) the resultant force on P'Q' = 0.
Thus, for equilibrium we must have Fz = 0. From 6.6 it follows that

E
Fz = ydA=O therefore ⌠ ydA=O (4.7)
R⌠⌡ ⌡
A A

But ⌠ y d A = "first moment" of area about the neutral axis and for
⌡
A
this to be 0, the neutral axis must pass through the "centroid" of
area A, i.e. point O in Figure is the centroid of area.

DEN-102: Handout No. 1 Page 48

The neutral axis passes through the "centroid" of area

Consider moment equilibrium of P'Q'D'B'. Since there is a clockwise

couple M at P'Q', there must be a counter-clockwise couple M at
B'D'. This arises from the total moment of forces δFz about neutral
axis xx.

moment of δFz about xx = δM = δFz y = y δFz

total moment M = ⌠ y dFz and since δFz = (E/R) y δA

⌡
A
E
∴ M = ⌠ y (E/R)y d A = y2 d A (4.8)
⌡ R⌠⌡
A A

But

∫y dA = I x
2
(4.9)
A

"second moment" of area about axis xx (analogous to "second

moment of mass" or "moment of inertia" used in dynamics)

Subscript x on I denotes that distance y in y2 dA is measured from

axis xx. Substitute 6.9 in 6.8

E M E
M= Ix = (4.10)
R Ix R

From 6.5 and 6.10

M σz E
Ix = y = R (4.11)

Eqn. 11.11 is called the "bending formula".

σz is called the axial stress due to bending moment M, i.e. the
"bending stress".
Since σz = (M/Ix)y then for a given M &y, σz depends on Ix.

Ix depends on geometry of beam x-section. The first step is to find

the neutral axis (which passes through the centroid)

DEN-102: Handout No. 1 Page 49

Co-ordinates of centroid of area

X X = X +x
O X-axis

Centre of
area
o x-axis
Y

Y =Y + y
dA

Y-axis y-axis

Figure 4.4

Consider an arbitrary x-section of area A, with neutral axes 0x, 0y

through 0.

Let OX, OY be axes parallel to 0x, 0y.

Coordinates of area δA are (x,y) or (X,Y) depending on axes chosen.
-- --
Coordinates of centroid 0 are (X , Y ).
By definition, the first moment of area A about axes 0x or 0y is
zero.
The first moment of δA about OX is YδA
∴ total moment about axis OX is
φx = ⌠ Y d A (4.12)
⌡
A
But since area A can be taken to be concentrated at centroid 0, we
have also
_
φx = AY (4.13)

From 4.12 and 4.13

_
AY = ⌠ Y d A
⌡
A
so that

DEN-102: Handout No. 1 Page 50

 _ 1
Y = ⌠ YdA (4.14)
A⌡
A
Similarly considering first moment about axis OY
-- 1
X = ⌠ XdA (4.15)
A ⌡
A

-- --
Equations 6.14, 6.15 give coordinates (X , Y ) of the centroid

Second moment of area

The second moment of area about axis OX is defined as

I X = ⌠ Y2 d A (4.16)
⌡
A

subscript X on I denotes that distance Y is measured from the X

axis. Similarly about 0x

Ix = ⌠ y2 d A (4.17)
⌡
A

Also IY = ⌠ X2 d A (4.18)
⌡
A

A
I y = ⌠ x2 d A (4.19)
⌡

DEN-102: Handout No. 1 Page 51

Theorem of parallel axes

Referring to figure in section 4.5

_
Y = Y + y.
_
Thus IX = ⌠ Y2 dA = ⌠ ( Y + y) 2 dA
⌡ ⌡
A A
_ 2 _
= ⌠ Y dA + ⌠ 2Y y dA + ⌠ y2 dA
⌡ ⌡ ⌡
A A A
_ 2 _
= Y ⌠ dA + 2Y ⌠ y dA + ⌠ y2 dA
⌡ ⌡ ⌡
A A A
⌠ d A = total area A , ⌠ y d A = 0 since this is the first moment
⌡ ⌡
A A
about centroid which must be 0 and ⌠ y2 d A = Ix by definition.
⌡
A
_ 2
_ 2
Thus Ix = A Y + Ix = Ix + AY (4.20)
_ 2 _ 2
Similarly Iy = A X + I y = Iy + A X (4.21)

This is the theorem of parallel axes.

Notation in books is often I = Ic + AH2 i.e. Ic replaces Ix (or Iy) as the
value of I about axis through centroid and H replaces distance

Polar second moment of area (see notes on Torsion)

OZ X X-axis
R axes OZ, oz are⊥ lar to figure
oz R = distance of δA from OZ
r
x x-axis r = distance of δA from oz
Y y
dA

Y-axis y-axis

Figure 6.5

By definition polar second moment about axis OZ is

DEN-102: Handout No. 1 Page 52

 JZ = ⌠ R2 dA = ⌠ (X2 + Y2) dA = ⌠ X2 d A + ⌠ Y2 dA
⌡ ⌡ ⌡ ⌡
A A
JZ = Ix + Iy (4.22)

This is theorm of perpendicular axes. Similarly

Jz = ⌠ r2 dA = ⌠(x2 + y2) dA = ⌠ x2 A + ⌠ y2 dA
⌡ ⌡ ⌡ ⌡
A A A A

Jz = Ix + Iy (4.23)

Second moments of area of some typical x-sections

4.1.1 Rectangle

Centroid o is at intersection of diagonals

B
X

Y = D/2
D o x

dA=Bdy

y Y

Figure 4.6

Consider
D/2 BD3
Ix = ⌠ y2 d A= ⌠ y2 Bdy = (4.24)
⌡ ⌡ 12
A -D/2

By inspection
DB3
Iy = (4.25)
12

Consider axis OX parallel to ox. By parallel axes theorem

DEN-102: Handout No. 1 Page 53

 _ 2
_
IX = Ix + A Y and Y = D/2 ; A = BD

BD3 D
( ) 2 = BD
3
= + BD (4.26)
12 2 3

_ 2
_
But IY = Iy + A X = Iy since X = o

4.1.2 Rectangular 'box' section

B
b
b

D d x d x

y y

Figure 4.7

Box section can be thought of as rectangle (1) minus rectangle (2).

Thus
Ix = Ix1 - Ix2
BD3 bd3 1
= - = (BD3 - bd3) (4.27)
12 12 12
For a 'square' box section (B = D ), b = d )
1
Ix = (D4 - d4) (4.28)
12

DEN-102: Handout No. 1 Page 54

4.1.3 Circle

Consider axis oz, ⊥ lar to plane of circle. The polar 2nd moment of
area is:

x
D
r

dr

Figure 6.8

Jz = ⌠ r2 d A = ⌠ D/2 r2 2πr dr
⌡ ⌡
A o
πD4
= (4.29)
32
(see notes on Torsion)

by ⊥ lar axes theorem Jz = Ix + Iy and by symmetry, Ix = Iy.

1 π
Thus 2Ix = 2Ix = Jz ∴ Ix = Iy = JZ = D4 (4.30)
2 64
4.1.4 Annular x-section

As for box section

Ix= Iy = Ix1 - Ix2 =
x
π
D1 D2 ( D24 - D14)
64

Figure 6.9

DEN-102: Handout No. 1 Page 55

4.1.5 Symmetrical I sections

By symmetry centroid is at intersection of diaganols AC and BD.

1
2 3
10 5
200 x

10

100
y

Figure 6.10

To calculate Ix, note:

∴ Ix = Ix1 - Ix2 - Ix3

but Ix3 = Ix2

∴ Ix = Ix1 - 2Ix2

all dimensions in mm

100x2003 (50-2.5) x 1803

∴ Ix = -2 mm4 = 20.50 x 106 mm4
12 12

To calculate Iy, note:

Iy = Iy1 + Iy2 + Iy3 and Iy2 = Iy3

180x53 10x1003
Iy = Iy1 + 2Iy2 = +2 mm4 = 1.67 x 106 mm4
12 12

DEN-102: Handout No. 1 Page 56

5. Deflection of Beams

M E σz
Bending formula: = = (5.1)
Ix R y

A beam subjected to loads will also be deflected

δz

Ο x

NA before
v loading
v+δv

y B NA after
A loading

Figure 5.1

Let v = deflection of the N.A. at distance z along the N.A. before

bending.

v is +ve downwards (in the +ve y direction ). Note: v varies with z ⇒

v = v(z)

Both v and its rate of change with z (= dv/dz) are very small,
i.e
v << z (5.2)
dv
<< 1 (5.3)
dz

Consider two neighbouring points A and B on the N.A. and distance

δz apart.

v = deflection at A
v + δv = deflection at B

φΑ = φ = angle between tangent at A and z-axis

φB = φ + δφ = angle between tangent at A and z-axis

DEN-102: Handout No. 1 Page 57

Note that with the beam bent as shown, φB is actually less than φA,
i.e. δφ < o

δs = arc length AB = R δθ (δθ in radians)

v+δv
v
δθ A
δs
R R
B
C δz

O
z
φΑ φΒ
v
v+δv
A
B
D
δφ

Figure 5.2

Right angled ∆ABC (see Figure .2)

AB2 = BC2 + CA2

(δs)2 = (δz)2 + (δv)2
δv
= (δz)2 [1 + (
δz
)2 ]
δv
≈ (δz)2 since << 1
δz
∴ δs = δz ≈ Rδθ
dθ 1
= (5.4)
dz R

From Figure 5.2

δθ = φA - φB = φ - (φ + δφ) = - δφ (5.5)

Therefore (5.4) becomes

dθ dφ 1
= - = (5.6)
dz dz R

DEN-102: Handout No. 1 Page 58

From Figure 5.2
δv dv
tan φ ≈ → as δz → o (5.7)
δz dz
Since φ is small, tan φ = φ (φ in radians)
dv
φ = tan φ = (5.8)
dz
dφ d2v
∴ = (5.9)
dz dz2

Replacing (5.9) in (5.6):

d2v 1
= - (5.10)
dz2 R
From (5.1)
M 1
= (5.11)
E Ix R

Therefore from (5.10) and (5.11)

d2 v M
dz2 = - (5.12)
EIx

Equation (5.12) is the differential equation which, when solved,

gives the deflection curve of the beam, i.e. v = v(z). Since it is 2nd
order, 2 boundary conditions are required to determine the
constants of integration. The equation is approximate since we have
considered small v, φ & dv/dz. It can be shown that, more exactly:

d2v
dz2 M
= - (5.13)
dv E Ix
[ 1+ ( )]
dz
2 3/2

DEN-102: Handout No. 1 Page 59

Example 1: Cantilever with end loads

W Mo

Ro

For equlibrium:
Ro = W and Mo = WL

For moment equilibrium

M + Wz = O
∴ M = Wz (A)

d2v M 1 W
= - = - (-Wz) = z (B)
dz2 EIx EIx EIx

Integrate twice with respect to z

dv W z2
= + A (C)
dz EIx 2

W z3
v = + Az + B (D)
EIx 6

Need 2 boundary conditions

at z = L, v=O (E)
z = L, dv/dz = O (F)

Since beam is "built in" at B

Using (F) in (C)

W L2 W L2
0 = +A ∴A= -
EIx 2 EIx 2

DEN-102: Handout No. 1 Page 60

Using (E) in (D)
W L3
0 = + AL + B
EIx 6

W L3 L2 WL3
B = [- + L]∴ B=
EIx 6 2 3EIx

Thus (D) becomes :

W z3 L2z 1 L3
v=
EIx 6
[-
2
+
3
]
Maximum v occurs at z = O, vmax = WL3/3EIx

DEN-102: Handout No. 1 Page 61

Example 3 Simply supported beam with u.d.l.

RA = RB = wL/2
wL wz2 w
At section, M = z- = [Lz - z2] (1)
2 2 2

d2v M w
∴ = - =- [Lz - z2] (2)
dz2 EIx 2EIx
dv w Lz2 z3
∴ =- [ - ]+A (3)
dz 2EIx 2 6
w Lz3 z4
∴v = - [ - ] + Az + B (4)
2EIx 6 24

At z = 0, v = 0; also z = L, v = 0, gives B = 0 and A = -

wL3/24EIx.
Thus (4) becomes:
w z4 L3z
v= -
12EIx
[ Lz3 -
2
-
2
] (5)

L 5 wL4
vmax occurs at z = and vmax = (6)
2 384 EIx

DEN-102: Handout No. 1 Page 62

Example 4 Simply suppoprted beam with concentrated load

C
1 2
W
A B

RA a RB
L

Load W acts at distance z from

Taking moments about B:
(L-a)
RA = W (1)
L

At section (1), i.e. o < z < a

(L-a)z
M=W (2)
L

d2v M W (L-a) z
∴ 2 = - = - (3)
dz EIx EIx L

dv W L-a z2
∴ = - + A1 (4)
dz EIx L 2
W L-a z3
∴ v=- + A1z + B1 (5)
EIx L 6

Eqns (2) to (5) form "equation set 1" which apply to points from A to
C (i.e. o < z < a) only. We are only free to impose boundary
conditions in this range.

At section (2) between C and B (a < z < L) :

(L-a)
M=W z - W (z-a) (6)
L

d2v M W (L-a)
∴
dz2 = -
EIx
=
EIx
[ -
L
z + (z-a)] (7)
dv W (L-a) z2 (z-a) 2
∴
dz
=
EIx
{ -
L 2
+
2
} + A2 (8)
W (L-a) z3 (z-a)3
∴v=
EIx
{ - L 6 + 6 } + A2z + B2 (9)

DEN-102: Handout No. 1 Page 63

"set 2" equations apply from C to B (i.e. a < z < L) only

We have 4 unknown constants A1, B1, A2, B2 so we need 4

boundary conditions.

Two are obvious:

(a) z = o, v = o using eqn (5) → not (9)
(b) z = L, v = o using eqn (9) → not (5)

The remaining two boundary conditions are obtained by noting that

at point C (i.e. z = a), both sets of eqns are valid. Thus slope (dv/dz)
and deflection (v) from both sets must be equal. Thus
dv dν
(c)   using eqn (4) =   using eqn (8)
dz z=a dz z=a

(d) (v)z=a using eqn (5) = (v)z=a using eqn (9)

W L-a a2 W L-a a2
Using (c) : - + A1 = - + A2
EIx L 2 EIx L 2

∴ A1 = A2 = A (10)

W L-a a3 W L-a a3
Using (d) - + Aa + B1 = - + Aa + B2
EIx L 6 EIx L 6

∴ B1 = B2 = B (11)

Then using (a) in (5) gives B=0 and using (b) in (9) [we cannot use
(b) in (5)]

W (L-a) L3 (L-a)3
0= {- + } +A L
EIx L 6 6

W I (L-a) L3 (L-a)3 W a(L-a) (2L-a)

∴A= { - } =
EIx L L 6 6 6EIx L
subst for A1 = A and A2 = A in (5) and (9) respectively

W L-a 3 W a(L-a) (2L-a)

(5) ⇒ v = - z + z
6EIx L 6EIx L

DEN-102: Handout No. 1 Page 64

 W L-a
= [a(2L-a) z-z3] for 0 ≤ z ≤ a (12)
6EIx L

similarly (9) becomes

W L-a L
v= [ a(2L-a) z - z3 + (z-a)3 ] for a ≤ z ≤ L
6EIx L L-a
(13)

In particular, at point C (z=a) deflection is obtained from (12) or (13)

as

W L-a W 2a2(L-a)2
vc = [ a2(2L-a - a3] = (14)
6EIx L 6EIx L

If loading is at mid-point (a = L/2) (BEAM EXPERIMENT)

L L
2 ( )2 (L - )3
W 2 2 WL3
vc = = (15)
6EIx L 43EIx

for rectangular x-section beam of breadth b and depth d, as in the

experiment,
Ix = bd3/12. Thus
WL3 12 L3 W
vc = x = (16)
48E bd3 4E bd3
∴ graph of vc v/s W/bd3 should give straight line whose slope =
L3
4E

from which E can be calculated as

E = L3/(4 * slope)

DEN-102: Handout No. 1 Page 65

6. Two-dimensional pin-jointed frameworks
A framework of members joined at their ends is called a truss.
Trusses are commonly used for bridges, roof support structures,
towers, antennae, etc. The structural members in trusses include
I-beams, channels, angles and tubular members.

In this course, we shall consider a particular type of truss, where

all of the members lie in one plane: such frameworks are called
two-dimensional or plane trusses. Plane trusses are often used in
pairs, e.g. one on each side of a bridge with cross-member
connections that provide stability to the assembled three-
dimensional structure also transfer the applied loads from
roadways to trusses. Similarly, a roof truss may be composed of
several plane trusses. Peep into the attic of your house to confirm
this !!

Figure 6.1
The members of a truss may be joined at their ends by welding,
rivets or bolts and gusset plates, pins, adhesives, etc.

Rivetted joint Frictionless point joint

Figure 6.2

DEN-102: Handout No. 1 Page 66

 In this course, we shall consider one type of connection only, by
means of frictionless pins, i.e. we shall deal with pin-jointed
trusses. There are two reasons for this. Firstly, we shall see that
this restriction simplifies greatly the analysis of a truss. Secondly,
it may turn out that for any truss with slender members, any type
of joint may be modelled (i.e. idealised) as a pin-joint for the
purpose of analysis: this gives reasonably good predictions of the
bar forces in the real structure, providing the centrelines of
adjoining members are concurrent at each joint.

Triangulated trusses
Three bars joined by pins at their ends forms a rigid framwork
(Figure 6.3a) whereas four bars form a non-rigid frame or
mechanism (Figure 6.3b). The addition of one extra bar to the truss
of Figure 6.3b transforms it into a triangulated truss, which
consists of a series of triangles. Triangulated trusses are rigid.

(a) (b) (c)

Figure 6.3

The trusses in Figures 6.3a and 6.3c are both statically determinate
because their bar forces can be determined solely from
considerations of equilbrium. It is a characteristic of statically
determinate trusses that the bars can be assembled without
forcing, even though they may not all be exactly of the correct
length. It follows that changes of length of the members of
statically determinate trusses due to such causes as change in
temperature or humidity (especially in case of wooden trusses) do
not produce additional stresses in the structure.

DEN-102: Handout No. 1 Page 67

The truss shown in Figure 6.4, obtained by adding a bar to the
truss of Figure 6.3c, is statically indeterminate, i.e. its bar forces
can not be determined by equilibrium conditions alone. Analysis of
statically indeterminate trusses is out of scope of this course.

Figure 6.4

The easiest way of designing a truss which is both rigid and

statically determinate is to form a series of triangles and then to
arrange for suitable supports. The design of more general, non-
triangulated trusses which are both rigid and statically determinate
requires a proper matrix analysis which again is not the part of this
course.

External loads and internal forces for a truss

A truss consists of slender elements and, for a general loading,
there will be the following three internal forces at any section of any
member:

• axial force T;
• bending moment M; and
• shear force S

Figure 6.5 shows a typical straight member AB of a truss and the

internal forces T,M, and S at a particular section. It is assumed
that member AB is rigidly connected to other members of the truss,
and hence all three internal forces can be applied at its ends.

DEN-102: Handout No. 1 Page 68

 P
SB
W
TB

TA
MA
SA MB
SB

TB
TA
MA SA

Figure 6.5

In this course, we shall consider only pin-jointed trusses subject to

point loads applied only at the joints. As these loads are usually
large in comparison to the self-weight of the members, we shall also
neglect the self-weight of the members.

Let us consider a typical member of such a truss, as shown in

Figure 6.6: this member is subjected to only one force applied at
each pin end. It follows from considerations of statical equilibrium
that the two forces at either end must be equal in magnitude,
opposite in direction and collinear. Thus, each member is in a
state of pure tension or pure compression, i.e. it is an axial force
member.

A B
T
T
S=0
A
T

M=0

Figure 6.6
In particular note that the bending moment M is zero at all sections
of this member. Bending moment could arise only if the pins were
not smooth, or if the member was loaded in between the joints, for
example, by a uniformly distributed weight. From a design
perspective, a pin-jointed truss carries the applied loads from the
DEN-102: Handout No. 1 Page 69
joints to the foundation by setting up tension and compression in
its members.

Calculation of bar forces in a pin-jointed truss

Two alternative methods can be used to calculate the axial forces in
the bars: the method of joints and the method of section.

6.1.1 Method of Joints

The method of joints provides a systematic way of calculating all
the bar forces in a structure by considering the equilibrium of
forces at each joint. As the forces are concurrent, we can resolve
them in two directions to obtain two independent equilibrium
equations. Thus, we can find two unknown bar forces (or
reactions) at each joint.

Example 6.1 Two bar truss: find the axial forces in the bar

C 45o A

Figure 6.7
Free-body diagram for joint A Triangle of forces

TAC = −W
TAB Y

45o
W
450
X
TAC
W TAB = 2W

DEN-102: Handout No. 1 Page 70

∑F y = 0 ⇒ TAB ⋅ sin(450 ) − W = 0 ∴ TAB = 2W

∑F x = 0 ⇒ TAC + TAB ⋅ cos(450 ) = 0 ∴ TAC = −W

We can check this result by drawing a triangle of forces.

Example 6.2 Four-bar truss: find the axial forces in the bar

D
B

45o

45o
A
C

Figure 6.8

FBD for joint A

TAB = 2W
TAC = −W

FBD for joint B Triangle of forces

TBD = 2W
Y

TBD
X 45o

45o
45o
TAB = 2W TBC = − 2W
TBC TAB = 2W

TBC 2W
∑F y =0⇒−
2
−
2
=0 ∴ TBC = − 2W (compression)

TBC 2W
∑F x = 0 ⇒ −TBD −
2
+
2
=0 ∴ TBD = 2W (tension)

We can check this result again by drawing a triangle of forces.

DEN-102: Handout No. 1 Page 71

6.1.2 Method of Section
If we are required to know the force in only one or two members of
a truss and the members are remote from supports, the method of
joints can involve a lot of work. In such cases, the method of
section can save much time. It involves following three steps:

1. Make an imaginary cut through the structure to create a free

body. The cut must pass through the member of interest. The
cut often divides the structure into two sections.

2. Draw a separate FBD for a section of the structure, imagining it

to be completely removed from the remainder of the structure
but subject to axial forces (of unknown magnitude) acting on
the cut member.

3. Resolve and/or take moments for the section as necessary. In

general, up to three independent equilibrium equations can be
obtained from which up to three unknown bar forces can be
determined.

In step 3, often the bar force of interest can be obtained directly by

means of a single equilibrium equation, either by taking moments
about a specially chosen point, or by resolving forces in a suitable
direction.

Example 6.3 14 bar tower with diagonals at 45°: find the axial
forces in bars DF and EG

W
L W
L
A
A
B C
2L B C

E D D
E

TOF
G F TEG
TEF F
J H

Figure 6.9

DEN-102: Handout No. 1 Page 72

The method of joints would involve analysing 5 joints (A-E) in
succession. Having cut the structure through the section DEFG,
we note that the upper section is subject to 3 unknown forces
which can be found from 3 equilibrium equations.

We can obtaine TDF directly from one of these equations, by taking

moments about E (where TEF and TEG intersect):

∑ M (E) = 0 ⇒ W ⋅ 2L − T DF ⋅L = 0 ∴ TDF = 2W

Similarly for bar EG

∑ M ( F ) = 0 ⇒ W ⋅ 3L − T
EG ⋅L = 0 ∴ TEG = −3W

How would you find TEF? (See Examples Paper).

Final remarks
If all joints of a truss are connected to more than two bars (where
each foundation reaction counts as one bar), it is necessary to start
by determining the support reaction by considering the overall
equilibrium of the structure (see Example 2.4).

The calculation of bar forces can be simplified by noting the

following special cases for unloaded joints:

T3=0
3 members of which
T2=T1 2 collinear
T1

T2=0
2 members

T1=0

T2
T3=T1 2 pairs of collinear
members
T1 T4=T1

Figure 6.10

DEN-102: Handout No. 1 Page 73

 Example 6.4 Roof truss, vertical load W at joint E: find the
axial force in DE.

D
D

W1
F TOE
F
E
E

A B
TCE

HA=0 C TAC C
RA RB
L L L L RB
L L

Figure 6.11

Whether we use the method of joints or the method of section, we

must start by finding reactions at A and B: we could not start at
joint E with the method of joint because there are three unknown
forces there and we only have two equilibrium conditions (in x and
y directions).

Considering the equilibrium of the whole structure:

∑M ( A) = 0 ⇒ RB ⋅ 4 L − W1 ⋅ L = 0 ∴ RB = 0.25W1

Considering the equilibrium of the right hand section:

∑M (C ) = 0 ⇒ RB ⋅ 2 L + TDE ⋅ 2 L = 0 ∴ TDE = − 2 RB ∴ TDE = −0.354W1

If instead we wanted TAC then:

∑M E = 0 ⇒ TAC = 0.75W1

or for TCE:

∑M A = 0 ⇒ TCE = −0.707W1

For W = 1, the full set of tensions turn out to be:

DEN-102: Handout No. 1 Page 74

 -0.354
-0.354
1

-0.354
-1.061

0.75 0.25

0.75 0.25

Figure 6.12

DEN-102: Handout No. 1 Page 75

7. Torsion

A common engineering mode of deformation is that of torsion-

where a solid/tubular member is subjected to a torque about its
long axis; this results in twitsting

7.1 Stress-strain considerations for a structural member in

torsion

Torsion refers to twisting of a member when loaded by couples that

produce rotation about its longitudinal axis.

T
F

d
F

Figure 7.1

In the case shown, the torque (or twisting moment or twisting

couple) is force x perpendicular distance between the forces F

torque T = F ⋅ d

For equilibrium of the member, there must be a torque T = Fd

acting in the direction opposite to the torque due to the applied
force F. In the figure above this acts the fixed (left hand) end.

DEN-102: Handout No. 1 Page 76

7.2 Torsion of thin walled cylindrical tube

AIM: to establish a relationship between τ and θ

Consider a thin-walled tube of mean radius r and wall thickness t

(<<r), subjected to a torque at each end, shown in Fig 4.2.

Assume that 'lower' end of tube is fixed.

Due to application of torque, a straight line AB which is parallel to

the tube axis twists through a small angle, φ, to AB'. Similarly CD
twists to CD'.

We assume AB' and CD' are straight lines.

An element PQRS deforms to P'Q'R'S'. The shear strain in the

element is equal to the decrease in angle QPS, which is initially 90o.
The decrease is seen to be angle φ. In fact the shear strain for any
other elements (e.g. ABDC) is also φ.
∴ shear strain = φ (7.1)

From BOB' length of arc BB' = rθ (7.2)

where θ = angle BOB' (in radians) is the "angle of twist" over the
whole length L of the tube.

L
dFs

dA
φ P Q B
A
dα
P’ Q’
T B’ θ T

S R D
C
S’ R’ r t
D’

B
A φ

θ
B’
O

Figure 7.2

DEN-102: Handout No. 1 Page 77

From BAB' length of arc BB'= Lφ (7.3)
where φ = BAB' (in radians) is the shear strain
Equating 7.2 and 7.3
Lφ = rθ
rθ
∴φ= (7.4)
L

The Shear modulus, G, for a material is given by Hooke's Law

stating that shear sress is proportional to shear strain in the linear
elastic region, just as normal stress and strain are related by
Young's modulus
τ
= G
φ
τ Gθ
∴ = (7.5)
r L
Gθ
or τ = r (7.6)
L
Since we are considering a thin tube, the variation of τ across the
tube thickness is negligible, i.e. τ is constant.

AIM: to establish a relationship between T and θ

Consider a small element of area subtending an angle dα at the

centre of the tube, see Figure 7.2

The shear force on element due to τ is

dFs = τ dA = τ (r dα) t (7.7)

Torque about axis due to dFs is

dT = dFs r = τ dΑ r (7.8)

T = ⌠ τ r dA
⌡
A
Gθ Gθ 2
as τ = r, T = ⌠ r dA
L ⌡ L
G,θ and L are constants
and ⌠ r2 dA = J, an expression called the polar second moment of
⌡
area
τ T
Therefore = (7.9)
r J

DEN-102: Handout No. 1 Page 78

 τ Gθ
and from 7.5 =
r L
τ Gθ T
∴ = = (7.10)
r L J

7.3 Torsion of hollow circular shaft (Thick walled tube)

Consider hollow shaft of inner diameter D1, outer diameter D2 and

length L subjected to torque T, shown in fig.4.3.

dr

T r1 T

C r
r

τ
r2

Figure 7.3

We assume
(a) shaft is straight and of uniform x-section
(b) torque is constant at all x-sections perpendicular to z-
axis
(c) plane sections remain plane after twisting
(d) radial lines remain radial
The hollow shaft can be thought of as being built from thin
tubes of thickness δr and mean radius r.

From thin tube theory we have

τ Gθ
=
r L

DEN-102: Handout No. 1 Page 79

 Gθ
τ = r (7.11)
L

Figure 7.4

Since G, θ, L are constant, τ = constant x r i.e. τ increases

linearly with r.
Gθ D1 Gθ D2
At inner radius τ = and at outer radius τ =
L 2 L 2

Torque on elemental ring of radius r and thickness δr is

dT = τ dΑ r = τ 2πr dr r (7.12)
Gθ Gθ
= r 2π r2 dr = 2πr3 dr
L L

D2/2 Gθ
total torque, T = ⌠ 2πr3 dr
⌡ L
D1/2

Gθ D2/2
∴ T= ⌠ r2 (2πr dr) (7.13)
L ⌡
D1/2

Now 2πrdr = dA = area of elemental ring

∴ ⌠ r2 2πr dr = ⌠ r2 dA = 2nd moment of area about
⌡ ⌡

axis = J
Gθ Gθ T
∴ T= J or =
L L J

DEN-102: Handout No. 1 Page 80

As Gθ/L = τ/r

τ Gθ T
= = (7.14)
r L J

To calculate the polar second moment of area, described in

equation 7.13

D2/2
J= ⌠ r2 (2πr)dr = 2π ⌠ r3dr
⌡ ⌡
D1/2

r4 π
= 2π [ ] = (D24 - D14) (7.15)
4 32

7.4 Torsion of solid circular shaft

This is just a special case of hollow shaft with D1 = 0 (no hole) and
D2 = D. Equation 7.14 applies with 7.15 changing to J = πD4/32

7.5 Concentric shafts

Take two shafts of different materials, 1 and 2

r3 1

r2
r1

Figure 7.7

DEN-102: Handout No. 1 Page 81

Using equilibrium T = T1 + T2

As we have considered before, from the geometry θ = θ 1 = θ2

the angle of twist is the same if there is no slip between the shafts
i.e. the ends are rigidly connected

T Gθ
Using =
J L

T1 L1 T2 L2
θ1 = θ 2 = =
G1 J1 G2 J2

T1 T2
As L1 = L2 =
G1 J1 G2 J2

When both shafts are made of the same material

T1 T2
=
J1 J2

DEN-102: Handout No. 1 Page 82

 Appendix

DEN-102: Handout No. 1 Page 83

8- Mohr’s Stress Circle
As derived in Handout No. 7, the direct and shear stresses on a
plane inclined at an angle θ w.r.t. the x-axis are given by

1 1
σ xθ = (σ x + σ y ) + (σ x − σ y ) cos(2θ ) + τ xy sin(2θ ) (8.1)
2 2
1 1
σ yθ = (σ x + σ y ) + (σ x − σ y ) cos(2θ ) − τ xy sin(2θ ) (8.2)
2 2
1
σ xθ = − (σ x + σ y ) + τ xy cos(2θ ) (8.3)
2

These are the parametric equations of a circle. If we choose a set of

rectangular axes and plot a point M of abscissa σxθ and
ordinate τθ for any value of the parameter θ, all the points will lie on
a circle. Let us try to plot this circle. For clarity, let us use the
following notation:
1 1
(σ x + σ y ) = σ (σ x − σ y ) = a τ xy = b
2 2
Then, equation 8.1 for σxθ and equation 8.2 for τθ become

σ xθ − σ = a cos(2θ ) + b sin(2θ ) (8.4)

τ θ = −a sin(2θ ) + b cos(2θ ) (8.5)

Squaring equations 8.4 and 8.5 and adding them, we obtain

(σ xy − σ ) +τ θ2= a 2 (cos 2 (2θ ) + sin 2 (2θ ) + b 2 (sin 2 (2θ ) + cos 2 (2θ ) +

2

+ 2ab sin( 2θ ) cos(2θ ) − 2ab sin( 2θ ) cos(2θ )

∴ (σ xθ − σ ) + τ θ = a + b
2 2 2 2
(8.6)

We used equation 8.1 for σxθ to obtain equation 8.6. We could use
equation 8.2 for σyθ and would still obtain equation 8.6. This is due
to the simple fact that when squaring the terms, their sign does not
matter. Therefore, for further clarity, let us replace σxθ by σθ in
equation 8.6. It is also worth noting that
2
1 
a 2 + b 2 =  (σ x + σ y ) + τ xy2 = R 2 (8.7)
2 

DEN-102: Handout No. 1 Page 84

Therefore, equation 8.6 becomes
(σ θ − σ )2 + τ θ2 = R 2 (8.8)

Equation 8.8 is an equation showing the variation between σθ and

τθ, i.e. it describes a curve which relates τθ (plotted on vertical axis,
for example) to σθ (plotted on horizontal axis). It is an equation of a
circle whose centre has coordinates (σ, 0) and whose radius is R as
shown in Figure 8.1. This circle is called Mohr’s stress circle.

τθ +
D

2θ σa
- A
+
C (σ ,0) B

E
-

Figure 8.1 Mohr’s Stress Circle

Any point on Mohr’s stress circle (e.g., point P in Figure 8.1)

represents values of direct and shear stresses on a plane inclined at
an angle θ w.r.t. the x-axis. Note that the angle of inclination the
radius of Mohr’s stress circle passing through point P w.r.t. the σθ
axis is 2θ. It is due to the fact that the parametric equations of
Mohr’s stress circle are in terms of 2θ and not θ. The four points A,
B, C, and D on Mohr’s stress circle (as shown in Figure 8.1) have
special significance. Points A and B lie on the σθ axis, i.e. there are
no shear stresses at points A and B. Therefore, they must
represent principal stresses. Point A represents minimum or
minor principal stress and point B represents maximum or major
DEN-102: Handout No. 1 Page 85
principal stress. Similarly, points D and E represent maximum
shear stress.

8.1. Construction of Mohr’s stress circle

8.1.1. Sign convention for direct and shear stresses
Before we study the method of construction of Mohr’s stress circle,
let us fix the sign convention for direct and shear stresses. In the
case of direct stresses, the tensile stress is taken positive and the
compressive stress is taken negative. In the case of shear stress,
if the action of shear stress would result in an anticlockwise
rotation of the element, it is taken as
positive. On the other hand, if the action
of shear stress results in a clockwise
rotation, it is taken as negative. This is
clearly illustrated in Figure 8.2. You can
use a reverse sign convention if you want
to. It will not make any difference on the
construction of Mohr’s stress circle. But
you must be consistent in the use of the
adopted sign convention when you
present the results obtained from Mohr’s
stress circle. Figure 8.2

8.2. Method of construction

We are given a plane stress system
shown in Figure 8.3 which consists τxy
of two-dimensional stress
components σx, σy, τxy and τyx ( = τxy)
on any pair of orthogonal reference σx
planes aligned in the x and y
directions. We are required to find
out the direct and shear stresses
on a plane inclined at an angle θ
w.r.t. the x-axis using Mohr’s stress
circle. The construction of Mohr’s
stress circle involves following
steps:
(Refer to Figure 8.4). Figure 8.3

1. Draw the direct stress (tensile +ve) axis and the shear stress
(anticlockwise +ve) axis and fix the scale of these axes. Make

DEN-102: Handout No. 1 Page 86

 sure that you take the same scale for both the axes otherwise
your Mohr’s stress circle will turn out to be an ellipse !

2. Plot point X which has coordinates (σx, τxy) and point Y which has
coordinates (σy, -τxy) as shown in Figure 8.4. At θ = 0 or 90°, σxθ =
σx, σyθ = σy and τθ = τxy. Thus, (σx, τxy) acting on the face normal to
x-axis and (σx, -τxy) acting on face normal to y-axis are two points
on Mohr’s stress circle. The shear stress for the face normal to y-
axis is negative because it applies a clockwise couple to the
element. Since the planes on which the stresses (σx, τxy) and (σx, -
τxy) act are orthogonal (θ = 90°), they will be plotted diametrically
opposite on Mohr’s stress circle. This is because Mohr’s stress
circle is parametric in terms of 2θ which will be equal to 180° for
the two orthogonal planes. In other words, the centre of Mohr’s
stress circle will lie on the line passing through points X and Y.

Figure 8.4

3. Join points X and Y by a line. This line will intersect the direct
stress axis at point C which is the centre of Mohr’s stress circle.
The radius of Mohr’s stress circle will then be either CX or CY.
The coordinates of the centre will be (σ, 0) where σ is the
arithmetic mean of the two direct stresses σx and σy. Draw
Mohr’s stress circle with centre at C and radius as CX (or CY).

4. The principal stresses are defined by points B and D where the

shear stresses are zero.

DEN-102: Handout No. 1 Page 87

5. The maximum shear stresses are defined by points G and H.
These two points lie on either end of the vertical diameter of
Mohr’s stress circle. Note that points B and D representing the
principal stresses are on either end of the horizontal diameter of
Mohr’s stress circle. Hence, the angle between the planes of
maximum shear and the principal planes should be 90°/2 = 45°
(as proved in Handout No. 7).

8.3. Pole on Mohr’s stress circle

By now, it may have become apparent to you that this business of

denoting the stresses on Mohr’s stress circle in terms of 2θ is not
very convenient. It would be better if we could somehow draw on
the Mohr’s stress circle the planes on which these stresses act. This
can be achieved if we define a point known as the pole of Mohr’s
circle. The essential characteristic of the pole is that a line passing
through the pole and a point representing direct and shear stresses
on Mohr’s stress circle will be parallel to the plane on which the
represented stresses act. There is only one pole for a Mohr’s stress
circle and it can be located as follows:

1- After constructing the Mohr’s stress circle using the method

given above, draw a line passing trhough point X which is parallel
to the vertical plane on which stresses (σ x ,τ xy ) act, as shown in
Figure 8.5.

DEN-102: Handout No. 1 Page 88

 Figure 8.5 Pole of Mohr’s stress circle and principal planes

2- Similarly, draw a line passing through point Y which is parallel

to the horizontal to the horizontal plane on which stresses (σ x ,−τ xy )
act. These two lines intersect each other on Mohr’s stress circle at a
point P. This point is the pole of Mohr’s stress circle since any line
passing through P and point representing a stress state (e.g. point
X or Y) is parallel to the plane on which the stresses act.

3- It becomes fairly simple now to obtain the directions of principal

plane or planes of maximum shear stress. All you have to do is to
draw a line passing through the pole and the point representing the
principal stress or maximum shear stress as shown in Figure 8.5.

4- Similarly, the direction of planes of maximum shear stress can

be obtained by drawing a line passing through the pole and one of
the points of maximum shear which are located on either end of the
vertical diameter of Mohr’s stress circle. (Figure 8.6).

DEN-102: Handout No. 1 Page 89

 Figure 8.6 Planes of maximum shear stress

5- Using the reverse procedure, the stresses on any plane inclined

at an angle θ w.r.t. the horizontal plane (or x-axis) can be obtained.
In this case, a line parallel to the plane in question is drawn
passing through the pole of Mohr’s stress circle. This line will cut
Mohr’s stress circle at a point S as shown in Figure 8.7. The direct
and shear stresses acting on this inclined plane are the co-
ordinates of point S. If a line perpendicular to the inclined plane is
drawn passing through the pole, it will cut Mohr’s stress circle at
point T as shown in Figure 8.7. It can be seen from Figure 8.7 that
point T is diametrically opposite to point S and that its coordinates
will give the direct and shear stresses on a plane which is
orthogonal to the inclined plane.

DEN-102: Handout No. 1 Page 90

 Figure 8.7 Stresses on any inclined plane

In each of the three cases, you can either measure the angles and
the stresses off the diagram or use the equations given in Handout
No.7 to compute them. However, the beauty of Mohr’s stress circle
is that all the angles and stresses can be obtained graphically and
therefore, there is not much point in using the equations once you
have succesfully drawn Mohr’s stress circle unless great accuracy
is required.

DEN-102: Handout No. 1 Page 91

9. Multi-dimensional Stress States

Multi-dimensional Stress States

AUTOMOTIVE Stress analysis of front suspension

Source: www.ansys.com

AEROSPACE Stress analysis of fuselage impact

Source: ONERA

DEN-102: Handout No. 1 Page 92

 MEDICAL
Stress analysis of hip endoprostheses
Source: www.endolab.de

Maximum Principle Stresses (Example)

DEN-102: Handout No. 1 Page 93