Daniele Dini
2014 Imperial College London
What is Stress Analysis?
Stress Analysis is the short version of Stress-Strain Analysis, the engineering dis-
cipline related to the theory and the methods developed by physicists, engi-
neers and mathematicians to determine the stresses and strains in materials
and structures subjected to forces or loads. This is one of the fundamental
subjects for civil, mechanical and aeronautical engineers and enables quanti-
tative analysis of the stresses and deformations caused by external loads ap-
plied to a body. In this context, stresses are local measures of forces and are
defined as forces per unit area, while strains are local measures of deforma-
tions and usually defined as linear deformations per unit length.
Given an engineering component or a structure, the properties of the materi-
als they are made of, the links between the parts of the structure, and the typi-
cal forces applied to them, the laws and methods taught in the Stress Analysis
course will enable you to obtain a quantitative description of the stress over all
those parts and joints, and the deformation caused by those stresses.
There is a strong link between stresses and strains as they cannot exist inde-
pendently from each other: stress causes strain and vice versa. The main differ-
ence between Stress Analysis and Mechanics (see ME1 Mechanics) is that in
the Mechanics syllabus deformations are neglected and all parts and struc-
tures are considered as rigid bodies. This is an idealisation which enables to
study the kinematics and dynamics of objects without having to consider de-
formations.
i
Contents
Chapter 1 Static equilibrium . . . . . . . . . . . . . . . . . 1
1.1 Dynamics and statics . . . . . . . . . . . . . . . . . . 1
1.1.1 Dynamics of a particle . . . . . . . . . . . . . . . 1
1.2 Static equilibrium . . . . . . . . . . . . . . . . . . . 2
1.2.1 Example: See-saw in equilibrium . . . . . . . . . . . . 2
1.3 Free body diagram . . . . . . . . . . . . . . . . . . 3
1.3.1 Analysis of free body diagrams . . . . . . . . . . . . . 3
1.3.2 Example: FBD analysis of bicycle . . . . . . . . . . . . 4
1.3.3 Example: FBD analysis of block on wedge . . . . . . . . . 5
1.4 Notes. . . . . . . . . . . . . . . . . . . . . . . 7
Chapter 3 Stress . . . . . . . . . . . . . . . . . . . . 29
3.1 Stress and strength . . . . . . . . . . . . . . . . . .29
3.2 Stresses in statically determinate systems . . . . . . . . . . .30
3.2.1 Thin, pressurised cylinders: hoop stress . . . . . . . . . .30
3.2.2 Thin, pressurised cylinders: axial stress . . . . . . . . . .32
3.2.3 Thin, pressurised spherical shell: stress . . . . . . . . . .33
3.2.4 Example: pressure vessel . . . . . . . . . . . . . .34
3.2.5 Example: pressure vessel bolts . . . . . . . . . . . . .34
3.2.6 Thin rotating ring . . . . . . . . . . . . . . . . .35
3.3 Notes. . . . . . . . . . . . . . . . . . . . . . .36
Contents ii
4.3 Poisson’s ratio . . . . . . . . . . . . . . . . . . . .40
4.3.1 Typical Poisson’s ratio values . . . . . . . . . . . . .41
4.4 Stress and strain in three dimensions . . . . . . . . . . . . .41
4.4.1 Hydrostatic stress and volumetric strain . . . . . . . . . .43
4.4.2 Relating elastic modulus, Poisson’s ratio and bulk modulus . . . .43
4.4.3 Example: volume change due to stress and temperature . . . .44
4.5 Strains in cylindrical and spherical shells. . . . . . . . . . . .45
4.6 Apparent modulus of elasticity. . . . . . . . . . . . . . .45
4.7 Volumetric strain within shells . . . . . . . . . . . . . . .46
4.7.1 Volumetric strain within a cylindrical shell . . . . . . . . .46
4.7.2 Volumetric strain within a spherical shell . . . . . . . . .47
4.7.3 Stress-strain relationships in polar coordinates . . . . . . . .47
4.8 Elastic strain energy . . . . . . . . . . . . . . . . . .48
4.9 Solving statically indeterminate stress-strain problems . . . . . . .49
4.9.1 Example: statically determinate problem . . . . . . . . .49
4.9.2 Example: statically indeterminate problem . . . . . . . . .50
4.9.3 Example: hydraulic cylinder with tie rods . . . . . . . . .51
4.9.4 Example: pressure vessel with hemispherical ends . . . . . .54
4.10 Notes . . . . . . . . . . . . . . . . . . . . . .54
Contents iii
6.3 Example: Problem 5.3 . . . . . . . . . . . . . . . . .83
6.4 Example: stress in an I-beam . . . . . . . . . . . . . . .85
6.5 Bars under combined axial and bending loads . . . . . . . . .86
6.5.1 Example: combined axial and bending loads . . . . . . . .87
6.6 Notes. . . . . . . . . . . . . . . . . . . . . . .88
Contents iv
Chapter 1 Static equilibrium
1
Stress analysis begins by dealing with structures and components which are in
static equilibrium — neither moving nor accelerating rapidly enough for inertial
Static equilibrium
forces to significantly load them.
On successfully completing this section, you will be able to:
• Define statics and dynamics
• Recall Newton’s laws of motion
• Define forces and moments
• Define static equilibrium
• Sketch free body diagrams for very simple systems
Static equilibrium
Second law
The acceleration of a particle as produced by a net force is directly pro-
portional to the magnitude of the net force in the same direction as the
net force, and inversely proportional to the mass of the particle:
F = mẍ.
Third law
For every action, there is an equal and opposite reaction.
∑i Fi = ∑ mẍ = 0
and
∑jMj = 0
i.e. both the net force and the net moment in every direction is zero. Such a
system is in static equilibrium.
1m 2m
Static equilibrium
1.3 Free body diagram
To analyse any complex engineering system, it is first broken down into the inter-
acting components it consists of, while noting the points at which each is connec-
ted to others. The whole system can then be predicted from a knowledge of how
each component behaves in isolation.
To analyse mechanisms and structures, they are disassembled into separate
free bodies. All of the points at which each free body contacted those sur-
rounding it are then replaced by equivalent forces and moments, and the free
body is analysed using static equilibrium equations.
Remember
Newton’s Third Law: When one body exerts a force on a second body, the second
body simultaneously exerts a force equal in magnitude and opposite in direction on
the first body.
Static equilibrium
Analyse the forces acting on the frame of a bicycle of mass MB loaded by a
rider of mass MR.
Before attempting to analyse stresses using a static equilibrium free body dia-
gram, we first need to be sure that the bike is in static equilibrium — i.e. it is
stationary or being ridden at constant speed, not accelerating or hitting a
bump.
1. Construct a free-body diagram (FBD) of the bicycle:
2. Mark the known forces acting on the bike frame: W 1 = M1g, W B = MBg
3. Add to the diagram variable reaction forces R1, R2 from the ground
Static equilibrium
5. In this case there are more unknowns than equations and we continue by
separating one section of the system as a new free body and returning to
Step 2.
Using Newton’s 3rd Law to separate subcomponents
It is not always necessary to begin with a free body diagram of the whole sys-
tem.
Because both masses move at constant or zero velocity, there are no accelera-
tions and the system is in static equilibrium.
1. Begin by drawing each component — block and wedge — as a separate
free body:
Static equilibrium
2. Mark the known forces m1g, m2g acting on each component.
3. Add to the diagram the variable, equal and opposite pairs of action reac-
tion forces (normal R1 and R2, tangential F1 and F2) between the block
and wedge surfaces.
Note
It doesn’t matter which direction is chosen for any action force arrow: if the action
and reaction arrows were both reversed, the resulting value of each force would have
the opposite sign.
Hence R1 = m1 + m2 g and F1 = 0.
Static equilibrium
Hence for vertical force equilibrium: R1 = m1 + m2 g
as before.
1.4 Notes
1.4 Notes 7
Static equilibrium
1
8
1.4 Notes
Static equilibrium
1
9
1.4 Notes
Chapter 2 Loaded frames
Pin jointed frames consist of straight bars loaded in tension or compression, as-
sembled with joints which are (or might become) free to rotate. Examples include 2
bridges, cranes and roof support structures for warehouses etc..
Loaded frames
B C D
A G F E
Loaded frames
A pin joint can support a horizontal or vertical force without movement but
cannot resist rotation.
V V V
H H
M or sliding joint can support only a vertical load without move-
A roller joint
ment.
V V V
H H
M
2.2 Example: plane frames
Find the reaction force at B in the plane frame shown:
B
10 kN 20 kNm 6m
3m
A C
10 m
Free body diagram:
B P in-joint
20
6m 10
A C
HA HC
VA VC
Resolve vertically: V A + V C = 0
2
Take moments clockwise about A: 10 × 3 + 20 − 10V C = 0
Loaded frames
and V C = 5 kN and V A = − 5 kN
HA
VA
and V C = 5 kN and V A = − 5 kN
MB=0 by definition!
20 HC
VC 5
VB 20 kNm 6m
C
5m HC
VC
Resolve horizontally: HB = HC
Resolve vertically: V B + V C = 0
Hence HC = HB = 5/6 kN
Loaded frames
10 kN 6m
A
HA
5m
VA
Resolve horizontally: 10 + HA = HB
Loaded frames
F1
RELATED LINKS
Canadian Institute of Steel Construction educational website on steel framed structures
VA VC
HA HC C
A
1.5m 1.2m
2m
B
10kN
What are the tensions in bars AB and BC and the horizontal and vertical reac-
tions at A and C?
1. Draw an FBD of the joint.
TAB
TBC
10kN
2. Resolve horizontally.
−T ABsin α + T BCsin β = 0
hence T BC = 34 T AB.
3. Resolve horizontally. 2
−T ABcos α + T BCcos β − 10 = 0
Loaded frames
4 3 3
T +4
5 AB
× 5 T AB = 10
9
4 + 4 T AB = 50.
Important
Notice that all tensions and reactions have been found by static equilibrium alone: this sys-
tem is statically determinate.
RELATED LINKS
Animations and additional material on pin-jointed frames
HA A HD D HC C
Loaded frames
B
A D
C
B
Loaded frames
This example demonstrates that the statical determinacy test for frames must be
used with care.
Determine whether the structure shown is statically determinate:
Loaded frames
2m
A 1m D 1m C
60kN
5. Resolve forces at A
TAB
A 45°
20 TAD
40
Vertically: T ABsin 45° + 20 = 0 hence T AB = − √2
= − 28.3 kN
TCB
TCD C 10 2
40
Loaded frames
Vertically: T CB = − 40 kN
Horizontally: hence T CD = − 20 kN
7. Resolve forces at D
TDB
20 α=63.44°
D 10
60
Vertically: T DBsin α − 60 = 0 kN
60
hence T DB = sin α
= 67.1 kN.
A G F E
HA
W
RA RE
Horizontal forces: HA = 0
Loaded frames
and horizontally at each:
In this example, start at A
TAB
60°
TAG
RA
Resolve vertically: T ABsin 60° + RA = 0
4 2
and T AG = 3 3
Wcos 60° =
3 3
W
3. If possible, choose the sequence of joints so that each one has only two
remaining unknowns.
In this example A, B, G, C, F, D, E.
a. Sketch an FBD at B:
TBC
60o
60o
TAB TBG
and T BC = 12 T AB − 12 T BG = − 4
3 3
W
b. Sketch an FBD at G:
60o! 60o
TAG TGF
2
W
Loaded frames
Resolve vertically: T BGsin 60° + T CGsin 60° = W
and T GF = 12 T BG − 12 T CG + T AG = 1
3
W
B -4k C -2k D
-4k
4k 2k -2k 2k -2k
A G! F E
2k 3k k
W
2W/3 W/3
A G F E
HA
W
RA RE
Horizontal forces: HA = 0
2
Loaded frames
2. Cut out from the frame a free body on which the bar forces which you
need to evaluate act as external forces:
In this example, cut bars CD, CF and (to release the free body) GF.
B C
TCD
TCF (√3/2)L
A G TGF F
RA=2W/3 W
Loaded frames
the material; inversely proportional to the compressive load on it
Buckling of a bar
Proportional to the second moment of area of its cross-section and the
Young’s modulus of its material; inversely proportional to the square of
its length.
Note
The second moment of area is a function of the beam cross-sectional shape, characterising
its resistance to bending.
VA VC
HA HC C
A
1.5m 1.2m
2m
B
W
T BC = − 0.6W
−T ABL2 2
AB = − 0.8W × 1.5 = 1.8W
−T BCL2 2
BC = − 0.6W × 2.0 = 2.4W 2
Using these results, as the magnitude of W increases:
Loaded frames
1. While W acts downwards, bar AB is under the greatest tensile load and
will be the first to fail by tensile yield.
2. While W acts upwards, bar AB is under the greatest compressive load and
will be the first to fail by compressive yield.
3. While W acts upwards, bar BC has the greatest value of −TL2 and will be
the first to fail by buckling.
A E
HA H G F
L
VA VE
Determine which bar carries the greatest tensile load and which is most likely
to buckle when
1. W acts downwards
2. W acts upwards.
1. Assume static equilibrium of the entire structure to determine reaction
forces.
Take moments clockwise about A: W L + 3L − 2V E L + 3L = 0,
hence V E = 12 W .
Horizontal equilibrium: HA = 0.
Loaded frames
TAB 60o
A
TAH H
W/2
B TBC
TBH
A
TAH H
W/2
Resolve vertically: T BH = 12 W
1
Resolve horizontally: T AH + T BC = 0 hence T BC = − 2
3W .
B TBC
THC
A
H THG
W/2
THG G THG
Resolve vertically: T CG = 0. 2
1. W acts downwards:
Loaded frames
Bar Tension/W Length/L Tension × (length)2 / WL2
AB –1 2 –4.0
AH √3/2 = 0.87 √3 2.61
BH 1/2 = 0.5 1 0.5
BC –√3/2 = –0.87 1 –0.87
HC –1/√2 = 0.71 √2 –1.41
HG (1 = √3)/2 = 1.37 1 1.37
2.9 Notes
2.9 Notes 26
Loaded frames
2
27
2.9 Notes
Loaded frames
2
28
2.9 Notes
Chapter 3 Stress
Stress is the mechanical force per unit area of the material through which the force
is transmitted.
On successfully completing this module, you will be able to:
3
• Define stress and strain
• Discuss material strength and links to failure
Stress
• Define statically determinate and statically indeterminate systems
• Analyse statically determinate systems, e.g. pressure vessels and blocks
• Build free body diagrams and sketch stresses acting over surfaces in sim-
ple systems
Fn
Mean normal stress σn = An
in units of Nm–2 (MN–2 etc.).
Another name for a stress or pressure of one Nm–2 is a pascal, symbol Pa (kPa,
MPa etc.).
Stress is defined as positive when the normal load is tensile:
Chapter 3 Stress 29
A shear stress acts on a surface area when a force is applied parallel to it:
As
Fs 3
Stress
Fs
The mean shear stress σs = As
has the same units as direct stress.
RELATED LINKS
Missouri University of Science and Technology MecMovies: see Ch 1, Stress
Stress
We will assume the wall thickness t to be small enough compared to the mean
radius Rm for the assumption that Rm ≈ Ri, where Ri is the inner radius, is val-
id.
1. The cylinder is in static equilibrium. Remove a semi-circular element of
the wall, Δz in length, and draw the free-body diagram showing the
forces acting on it.
The lower half of the circular section is removed, and replaced by stresses
σθ acting on the upper half which acted across the plane joining them.
Tip
This result can be derived more simply by noting that the total upward force FP due
to pressure is simply the product of the pressure and the area, projected in the verti-
cal direction, on which it acts.
4. Apply vertical equilibrium to show that the total vertical force is zero, and
solve for σθ.
3
Since FP = Fσ,2PRiΔz = 2σθtΔz
Stress
PRi
hence σθ = t
.
FP = πR2
iP
3. Determine the rightward force due to normal stress σz acting on the cut-
through wall.
Fσ = 2πRmtσz.
4. Apply horizontal equilibrium to show that the total axial force is zero, and
solve for σθ.
Stress
Note
A typical example of an open-ended cylinder is a pneumatic or hydraulic actuator,
from which the pressure force is applied to an external load.
t r y r
A
x
Pressure, P z
FP = πR2
iP
3. Determine the downward force due to normal stress σθ acting on the cut-
through wall circumference of width t.
Fσ = 2πRmtσθ.
4. Apply horizontal equilibrium to show that the total vertical force is zero,
and hence solve for σθ.
RELATED LINKS
Missouri University of Science and Technology MecMovies: see Ch 14, Pressure vessels
3
3.2.4 Example: pressure vessel
Stress
The 300 mm diameter, 6 mm thick pressure vessel shown is to be made up of
two hemispherical shells bolted together at flanges. The maximum allowable
tensile stress in the material to be used is 150 MPa, what is the maximum al-
lowable working pressure?
2 × 0.006
Hence Pmax = 0.15
× 150 × 106 Pa = 12 MPa.
Stress
3.2.6 Thin rotating ring
If a cylindrical ring rotates about its axis at high speed every element experiences
an outward centrifugal force, just as it would do from internal pressure. Like inter-
nal pressure, this develops a circumferential stress, which can be calculated by
analogy.
This analysis derives an equation for the circumferential stress in a ring of
mean diameter Rm, thickness t ≪ Rm and material density ρ, at a rotational
speed ω.
1. Draw the free body diagram for a circumferential element of included an-
gle δα, and determine the radial centrifugal force on it.
so that 3
Stress
ρR2 2
mω t
Pω = Ri
.
3.3 Notes
3.3 Notes 36
3
Stress
3.3 Notes 37
3
Stress
3.3 Notes 38
Chapter 4 Elastic strain
When a solid is put under stress, its shape changes slightly as internal bonds
stretch. The resulting small displacement of particles within a body — without
movement of the body as a whole — are defined as strain; if it disappears when
the stress is removed, the strain is elastic.
On successfully completing this section, you will be able to:
4
• Recall Hooke’s law for uniaxial tension and its generalised form for three-
dimensional systems
Elastic strain
• Determine stress and strains in thin-walled pressure vessels and solid
blocks
• Define volumetric strain
• Determine relationships between elastic constants
• Define and calculate strain energy for individual bodies loaded in tension
and shear
• Identify and analyse statically indeterminate systems
F
The direct stress at any cross-section is σ = A
.
The direct strain along the axis (in this case the longitudinal, or axial strain) is
ΔL
e= L .
RELATED LINKS
Missouri University of Science and Technology MecMovies: see Ch 1, Stress
Elastic strain
• If e returns to zero as soon as the load is reduced again to zero (e.g. as for
a rubber), the behaviour of the material is described as elastic.
• If the relationship between σ and e is linear up to some point, the behav-
iour of the material there is described as linearly elastic. Many materials
are linearly elastic at small strains.
For these linearly elastic materials Hooke’s Law holds:
σ
e = E,
RELATED LINKS
Animation: linear elastic stress and strain
Missouri University of Science and Technology MecMovies: see Ch 3, Mechanical Properties
Important
For most materials, every cross-sectional plane along the axis of a uniaxially extended bar
will contract by an equal strain in every direction. These in-plane strains will be a constant
proportion −ν of the axial strain.
For example:
Elastic strain
• If Poisson’s ratio is positive and less than 0.5, when the bar is extended in
one direction its volume will increase
• If Poisson’s ratio is negative, when the bar is extended in one direction its
volume will increase. Such auxetic materials are very rare.
Elastic strain
1. Tabulate the strains due to each stress in turn acting alone.
Direction x y z
Strain ex ey ez
2. Add the strain contributions from each stress to give the total strain in
each direction.
σx σy σz 1
ex = E −ν E −ν E = E σ x − ν σ y + σz
σy σz σx 1
ey = E −ν E −ν E = E σ y − ν σz + σ x
σz σx σy 1
ez = E −ν E −ν E = E σz − ν σ x + σ y
σ H = σ = σ y = σ z.
x
The strain will be equal in every direction too, and will cause a change in vol- 4
ume. The volumetric strain, ev is defined as the change in volume V divided by
the original volume:
Elastic strain
ΔV
ev = V .
Hence
ΔV = ΔxΔyΔz ex + ey + ez
ΔV
and hence eV = V
= e x + e y + ez .
3. Set all three stresses equal to the hydrostatic stress σH to derive the bulk 4
σH
modulus K = eV
Elastic strain
1 − 2ν
Hence eV = 3σH E
and
E
K = 3 1 − 2ν
or E = 3K 1 − 2ν .
4
If the cylinder expands under internal pressure, the axial strain is simply the
Elastic strain
increase in length divided by the original length:
ΔL
ez = L .
Remember
u/R is the circumferential strain, not the radial strain! There is no tensile stress to extend
any material in the radial direction.
For a spherical shell, the circumferential strain for any circumference is also
u/R.
The real radial strain in any shell is the increase in radial length, divided by the
original radial length, of an element of material in the cylindrical shell. The
average radial strain is the increase in wall thickness divided by the original
wall thickness.
σz
Elastic strain
From the generalised Hooke’s law equations with no temperature change:
1
ex = E σ x − ν σ y + σz + αΔT
δV = π ⋅ 2RLδR + πR2δL.
Elastic strain
A spherical shell has radius R.
4
δV = 3 π ⋅ 3R2δR.
Hence ev = 3eθ.
For any point in a spherical solid, the generalised Hooke’s law equations be-
come:
1
eθ = E σ θ − ν σ r + σ ϕ + αΔT
1
er = E σ r − ν σ ϕ + σ θ + αΔT
1
eϕ = E σϕ − ν σθ + σr + αΔT
Elastic strain
Load
Extension
The external work done during a small increment of extension δx is Fδx and
the total area under the curve for an extension of ΔL is 21 FΔL.
1
2
W ⅆL
The stored strain energy per unit volume is AL
.
Since strain e = δL/L, stress σ = F/ A and tensile modulus E = σ/e, the elastic
strain energy per unit volume stored by applying a direct stress is 21 σ2 /E.
Following the same argument, the elastic strain energy per unit volume stored
by applying a direct stress is 12 τγ = 12 τ2 /G.
Elastic strain
More unknowns
than equations
FBD, reactions,
tensions, etc Equations in
Statically terms of strains
indeterminate
Apply static Geometry of
equilibrium deformation
Statically
Stress-strain
determinate relationships
Equations in
terms of stresses
Solve all equations
1. Define the system, identifying all possible external loads and reactions
acting on it.
2. Draw a free body diagram and apply static equilibrium conditions to form
up to three equilibrium equations.
3. Apply the method of sections to successive subcomponents (e.g. bars) of
the system until there are as many equations as unknowns.
4. If there are enough equations, the system is statically determinate: solve
them all.
5. If there are more unknowns than equations, analyse the geometry of de-
formation to derive compatibility equations in terms of strain, deflection
or extension.
6. Use stress-strain (or load-deflection) relatonships to convert these com-
patibility equations into terms of stress (or load), until there are as many
equations as unknowns.
7. Solve all equations.
Elastic strain
2. Draw a free body diagram and apply static equilibrium conditions to form
up to three equilibrium equations.
σ2 F2 d2 F2L2
e2
= E2, so that A = E2 × L and d2 = A E .
2 2 2 2
F L1 L2
If A1 = A2 = A: d = d1 + d2 = A E1
+E .
2
Elastic strain
3. Apply the method of sections to successive subcomponents (e.g. bars) of
the system until there are as many equations as unknowns.
Here, for either end-cap F = F1 + F2
Hence F1 = E1 A1 × δL and F2 = E2 A2 × δL
δ δ δ
and F = F1 + F2 = E1 A1 × L
+ E2 A2 × L = L E1 A1 + E2 A2 .
E1 A 1
F1 = F and
E1 A1 + E2 A2
E2 A 2
F2 = F.
E1 A1 + E2 A2
Hydraulic cylinder 4
Elastic strain
Calculate the stresses in the rods and the cylinder at the cylinder design pres-
sure of 20 MPa.
The cylinder and tie rods are made from steel with a Poisson’s ratio of 0.3.
1. Define the system, identifying all possible external loads and reactions
acting on it.
There are no significant external loads on the system. Unless the piston
(not shown) is accelerating rapidly, the system is in static equilibrium.
Thickness of cylinder t = 4 mm
Diameter of each rod drod = 8 mm.
2. Draw a free body diagram and apply static equilibrium conditions to form
up to three equilibrium equations.
Free body diagram of either end cap:
end plate
tie rod
Hydraulic cylinder
Elastic strain
4. If there are enough equations, the system is statically determinate: solve
them all.
In this example there are two unknowns (σrod, σz, cyl) and one equation:
not enough.
5. If there are more unknowns than equations, analyse the geometry of de-
formation to derive compatibility equations in terms of strain, deflection
or extension.
For the cylinder of thickness t under internal pressure 20 MPa, geometry
of deformation gives
erod = ez, cyl.
1
ez, cyl = E σz, cyl − ν σθ, cyl + er, cyl
PR 20 × 40
eθ, cyl = t = 4
= 200 MPa
σr, cyl ≈ − P ≈ − 20 MPa, which is small compared to σθ, cyl and can be
neglected.
Hence
1
ez, cyl = E σz, cyl − 200ν MPa.
Elastic strain
The pressure vessel on an air-blast circuit breaker has a thin-walled cylindrical
section of mean radius R and thickness t. The vessel has hemispherical ends
and the radial expansion of the cylinder and hemispheres at their junction is
to be the same when it is subjected to pressure P:
RELATED LINKS
What's an air-blast circuit breaker?
Air-blast circuit breaker in action
Typical compressed air tank designs
4.10 Notes
55
4.10 Notes
Elastic strain
4
56
4.10 Notes
Chapter 5 Bending of beams
A beam is a structural element which is slender relative to its length, but supports
loads perpendicular to its longitudinal axis by resisting bending. In structural engi-
neering, at least, beams are usually mounted horizontally and are usually loaded
by bearing the weight of other elements.
A beam is generally supported at points other than those at which it is loaded.
Its function is:
1. To limit deflection when the load is applied or moved (e.g. as a vehicle 5
crosses a bridge)
2. To prevent any risk of the load becoming unsupported because excessive
Bending of beams
force breaks the beam (e.g. an aircraft fuselage becoming unsupported
when the wing root breaks).
The objective of our analysis is to predict the deflections and strength of a
beam given its length, cross-section and support conditions. The first step is to
summarise the applied loading as a distribution of bending moment and of
shear stress acting at every point along the beam length.
On successfully completing this module, you will be able to:
• Define a beam
• Recognise different types of beams
• Recognise and list possible supports and reaction forces
• Build free body diagrams of simple beams
• List the possible loads and recognise their engineering symbols
• Determine shear forces and bending moments in simple beams
• Sketch diagrams for shear forces and bending moments
• Derive mathematical expressions to link applied Loads, shear forces and
bending moments.
Bending of beams
A simply supported beam can bend freely at and between its supports, but it
can deflect only between supports.
Note
We did not consider whether or not each support could support loads along the axis, be-
cause these are beams (intended to resist bending) rather than bars (intended to resist ax-
ial force). A structural element can be both a beam and a bar — but we consider these two
functions separately.
Because it is not important to a beam whether a support can resist axial load
or not, a simple support is drawn simply as an abstract, triangular ‘point sup-
port’.
Bending of beams
Example 3: A horizontal beam built-in at both ends ends, and loaded by a con-
tinuous distribution of weight:
w / unit length
Each slice is subjected to a downward shear force of F on its left-hand face and,
in reaction, to an upward shear force of F on its right-hand face.
Now section the beam at a point B, distance x from the left-hand end:
A B
and draw regions to the left and right of B as free body diagrams:
A B 5
Bending of beams
For the free body AB:
Resolve vertically: F x = P
Resolve horizontally: H x = 0
Moments about B: M x = Px.
We see that each slice of a beam in static equilibrium must be loaded through
its surfaces by its neighbours in two ways:
1. A shear force created by the equal and opposite vertical loads from the
beam sections to its left and right
2. A bending moment created by the equal and opposite moments acting
on the beam sections to its left and right
Remember
A consistent sign convention is used to represent distributed load w, shear force F and
bending moment M on a beam element. All are positive as shown below:
w
F F
M M
A B
3 3 6
F F
M M
1. Treat the entire beam as a free body to calculate the reaction forces.
Resolve vertically: RA + R = 12 + 10
B
5
Take moments about A: 12 × 3 + 10 × 6 = RB × 12
Bending of beams
hence RB = 8 and RA = 12 + 10 − 8 = 14.
=14
hence F + 14 = 0, F = − 14.
b. Within 3 < x < 6:
12
=14
hence F + 14 − 12 = 0, F = − 2.
c. Within 6 < x < 12:
12 10
=14
F + 14 − 12 − 10 = 0, F = 8.
3. Draw shear force diagram:
A 5
=14
Bending of beams
hence M + 14x = 0, M = − 14x.
b. Within 3 < x < 6:
12
=14
12 10
=14
M + 14x − 12 x − 3 − 10 x − 6 = 0, M = − 96 + 8x.
5. Draw a bending moment diagram:
BM
A
B
–42
–48
1 5
Bending of beams
A B C
1 5
Resolve vertically: RA + RB = 6
Hence RA = 5 and RB = 1.
SF
–5
–5
Bending of beams
12 3
1 3 2
12 3
A B C D
1 3 2
Resolve vertically: RA + RB = 15
Hence RA = 11 and RB = 4.
–25
Bending of beams
M = 4x − 24
5. Draw a bending moment diagram.
BM
–4
–11
1 3 2
12 3
8 / unit length
A B C D
1 3 2
Bending of beams
SF
14
–25
BM
0
–25
–40
Bending of beams
as a force couple F0 separated by an arbitrarily small small distance e, where
F0 = eM0:
Resolving vertically: RA + RB = 0
M0 M0
Hence RB = L
and RA = − L
.
The value of e affects the shear force diagram only within a region of its own
size:
SF
A B
B
A
Bending of beams
A B
ⅆF
hence w = ⅆx
.
2. Take moments B:
ⅆM 1
M + ⅆx δx − M − Fδx − wδx × 2 δx = 0
ⅆM 1
so that ⅆx
− F − 2 wδx = 0
and as δx 0,
ⅆM
F = ⅆx .
w x = ax2 + bx + c
where w is the intensity of load at a distance x along the beam and a, b and c
are constants. The intensity of load is zero at each end and has its maximum
value of 4 kNm–1 at mid-span.
Apply the differential equations of equilibrium relating w, F and M to obtain
the maximum values and distributions of shear force and bending moment.
[8 kN, 15 kNm]
Hence w x = − 49 x2 + 83 x.
Bending of beams
ⅆF
2. Apply w = ⅆx
to determine the shear force distribution along the beam.
5.7 Notes
5.7 Notes 69
Bending of beams
5
70
5.7 Notes
Bending of beams
5
71
5.7 Notes
Chapter 6 Bending of beams: stress and
strain
Once the length, support conditions and applied loads for a beam have been ex-
pressed as shear force and bending moment distributions, stresses and strains can
be calculated at every point along it. To do so, the cross-sectional shape is ex-
pressed as a single, geometrical property: the second moment of area.
On successfully completing this module, you will be able to:
• Define bending stresses in beams
• List the assumptions made to derive the mathematical expressions for 6
stresses and strains in slender beams
A2 B2 A'2 B'2
This strain increases with radius, but the average strain must be zero if the
beam bears no axial load.
Important
There must be a neutral axis between the top and bottom of the beam at which the
strain is zero.
Hence for the element shown at a height y from the neutral axis,
A′B′ = R + y δθ
and
R + y δθ − Rδθ y
eAB = Rδθ
= R.
RELATED LINKS
Animation: assumptions for bending analysis
A B
Compression
This must be equal to the local bending moment M, and since from Step 1
α = E/R:
E bd3
M= .
R 12
We have shown that the distribution of stress σ and the local radius of bend-
ing curvature R at any point along a rectangular beam where the moment is M
are given by:
12M σ
= y = R.
E 6
3
bd
9 kN 40 mm
150 mm
2m
σ 12M
Since y
= ,
bd3
compression
Given conditions (1) and (2), our task is to find the location of the neutral axis.
1. Calculate the force transmitted through the cross-section for an arbitrary
choice of neutral axis.
The force transmitted through the elemental strip shown is σb y δy,
where σ = αy where σ = E/R.
Hence the total axial force is
+y1 +y1 +y1
∫ σb y ⅆy = ∫ αyb y ⅆy =α ∫ b y y ⅆy.
−y2 −y2 −y2
∫ b y y ⅆy = 0,
−y2
3. Calculate the bending moment for the same choice of neutral axis.
The moment due to the force calculated in Step 1 is σb y δy × y.
Hence the total bending moment is:
∫ σb y ydy = α ∫ b y y2dy = α I
−y2 −y2
where I is the second moment of area of the beam cross-section about the
chosen neutral axis.
σ M bd3
Hence y
= I where I = 12 as derived previously.
x bd3
x Ixx = 12
bd3
Iyy = 12
x x πa4
Ixx = Iyy = 4
b1 + b2 h3
Ixx = 36
x x b1 + b2 b2 2
1 + b1b2 + b2 h
Iyy = 36
Round tube
y
x x
πa4
For the outer area: Iext = 4
πb4
For the inner area: Iint = 4
π
For the composite: Itube = 4
a4 − b4
x x
y
6
bd3
For the outer area: Iext = 12
1
For the composite: Ibox = 12
bd3 − ec3 .
Area A
x x
a a
For the original axis xx passing through the centroid of this area the moment
of area was
y2
Ixx = ∫b y y2 ⅆy
y1
Ixx = ∫ y2 ⅆA.
A
2a ∫ y ⅆA = 0
A
by definition; hence
web
N A
flange
d t
td + td y = td × + td × d +
2 2
hence
3d t
y= + .
4 4
2. Determine second moment of area around the axis now located.
Using the parallel axis theorem
I=I x + A1 y − y1 2 +I x + A2 y − y 2 2
a1a′1 a2a′2
hence
6
td3 3d t d 2 dt 3 t 3d t 2
I= + dt + − + + dt d + − + .
12 4 4 2 12 2 4 4
1.
84
6.3 Example: Problem 5.3
6.4 Example: stress in an I-beam
A beam is to bear a uniformly distributed load of 120 kNm–1 over the 5 m span
between its simply supported ends. If the maximum allowable stress for the
material is 160 MPa determine whether an I-beam 528.8 mm in height, 6
200 mm wide and having a second moment of area of 404×10–6 m4 will be
If not, determine the length and thickness of flange reinforcement plates nee-
ded to make it so.
Flange plates are fixed to the outer surfaces of the upper and lower flanges of
an I-beam, and of equal width.
σ M
1. Since y
= I and all variables are specified except the bending moment
M, determine the maximum allowable bending moment:
ymax 160 × 106 × 404 × 10−6
Mallow = σmax = = 245 kNm.
I 528.8/2 × 10−3
2. Treat entire beam as FBD to determine reactions:
By symmetry, RA = RB.
0.2t 3 D+t 2
I = I0 + 2 + 0.2t
12 2
1
ymax = 2 D + t , and
6
Note
There are two practical cases in which combined bending and tension can
arise by a single force:
1. A bar, of finite thickness, end-loaded at a point which does not lie on the
centroid:
Axial stresses along the bar are the sum of those due to
• An axial force θ, plus
• Bending by a clockwise moment M = Fe applied at the end.
2. A bar which is end-loaded at an angle θ to its axis:
Axial stresses along the bar are the sum of those due to
• An axial force Fcos θ, plus
• Bending by a normal end load Fsin θ.
RELATED LINKS
Animation: bending moment, stress and strain
1m
10 KN
60 mm
σ M
=
y I
Bending moment M at distance x from support
M=W 1−x
hence Mmax = W × 1 Nm at support.
π 0.06 4
I= = 6.362 × 10−7 m4.
64
Hence
Mmax W
σmax = ± ymax = ± × 0.03 Pa
I 6.362 × 10−7
and for this to become equal to σa,
This result illustrates that for a typical bar geometry, stresses (and strains) due
to bending are much higher than those developed by stretching. If you want
to break a bar you don’t stretch it: you bend it against a support.
6.6 Notes
6.6 Notes 88
Bending of beams: stress and strain
6
89
6.6 Notes
Bending of beams: stress and strain
6
90
6.6 Notes
Chapter 7 Bending of beams: deflections
Beam analysis allows the radius of curvature to be calculated at every point along
a beam of known material, geometry and loading conditions. The final step is to
integrate this curvature distribution to yield deflections at every point; in statically
indeterminate cases, however, these deflections may themselves influence load-
ing.
On successfully completing this module, you will be able to:
• Define bending deflections and recall the moment-curvature relationship
• Apply the moment-curvature relationship to find slopes and deflections
at any point along the beam
• Solve standard beam problems and recall expressions for significant de- 7
flection values values for simple beams
1 ⅆ2v
=
R ⅆx2
and hence the moment-curvature relationship for a beam
ⅆ2v
EI = M.
ⅆx2
RELATED LINKS
Missouri University of Science and Technology MecMovies: see Ch 10, Beam Deflections
A X B
!
A X
ⅆ2v
EI = M = PL − Px.
ⅆx2
4. Integrate twice to obtain the displacement profile:
ⅆv x2
EI = PLx − P +A
ⅆx 2
ⅆv 1 L2 PL2
= PL2 − P = .
ⅆx EI 2 2EI
A X B
Moments about A: MA = wL × 12 L.
A /unit length
X
ⅆ2v 1
EI = M = w L − x 2.
ⅆx 2 2
ⅆv 1
b. At x = 0, slope ⅆx
= 0 and therefore A = 6 wL3.
ⅆv wL3
= .
ⅆx 6EI
A B
ⅆ2v w 2
EI =M= x − Lx .
ⅆx 2 2
w x4 x3
EIv x = −L + Ax + B
2 12 6
bP
Resolve vertically: V A + V B = P hence V A = L
.
A
X
ⅆ 2v 1 bP
EI = − x
ⅆx 2 L
b. a < x < L:
ⅆ 2v 2 bP
EI =P x−a − x
ⅆx 2 L
bP 3 P bP 3
− a + A1a + 0 = a−a 3− a + A2a + B2
6L 6 6L 7
hence A1a = A2a + B2
a x
∫0 ⅆx x≤a
∫ x−a n
ⅆx = .
∫ x − a n ⅆx x>a
const. x≤a
= x−a n+1
+ const. x>a
n+1
7
A B
A
X
b
M = − LP x + P x − a
ⅆ2v bP
EI =P x−a − x.
ⅆx 2 L
A B
Resolve vertically: A
RA + RB = w L − a X
1 a
hence R A = w L − a 2
− 2L .
X
1 a 1
M = − w L − a 2 − 2L x + 2 wP x − a 2.
A B
2. Represent the removal of the distributed load w per unit length from
x = b to x = L as the addition of a distributed load −w per unit length.
7
= +
A B A B
Hence
1 1
M = − RAx + 2 w x − a 2 − 2 w x − b 2
ⅆ2v w w
EI = M = − R Ax + x−a 2− x − b 2.
ⅆx 2 2 2
ⅆv x2 w w
EI = − RA + x−a 3− x−b 3+A
ⅆx 2 6 6
1. Determine (e.g. from Handbook of Data and Formulae) end deflection and
slope for a cantilever of length L1.
wL4
1
End deflection δ1 = 8EI
wL3
1
End slope θ = 6EI
.
2. Determine δ2 by geometry.
The beam in L1 < x < L2 is not loaded and therefore has no curvature at
any point.
Attention
2 equations and 3 unknowns! The reactions cannot be found by static equilibrium
alone.
Equilibrium gave two equations to determine the three unknowns and com-
patibility has given a third. Hence V A and MA can now be determined.
To impose the additional condition that the beam slope at both ends is zero,
the displacement profile v x must be determined from the bending moment
distribution
M = − RAx + M A + P x − a .
RELATED LINKS
Missouri University of Science and Technology MecMovies: see Ch 11, Indeterminate beams
7.5 Notes
104
7.5 Notes
Bending of beams: deflections
7
105
7.5 Notes
Chapter 8 Torsion of bars
Torsion is the twisting of an object by an applied moment. It is important in engi-
neering both as an unwanted result of the need to transmit a torque along a shaft,
and as the source of useful compliance in a coil spring.
On successfully completing this module, you will be able to:
• Define torsion, shear stress and shear strain
• Determine stresses and deformations in thin-walled circular and non-cir-
cular shafts under torsion
• Determine stresses and deformations in solid circular and hollow shafts
under torsion
• Define torsional stiffness
• Analyse circular shafts of varying cross-section subjected to torsional load 8
• Recall formulae for the study of stress and deformation in non-circular
Torsion of bars
shafts under torsion.
FROM
ENGINE
TO
DIFFERENTIAL
Clutch 1
Clutch 2
Note
8
In this case torsion of each shaft is unwanted, and must be minimised.
Torsion of bars
2. Torsion bar suspension for vehicles.
3. Torsion of the wire or bar which a torsion spring is made of allows the
spring to change in length when a force is applied to it.
Note
In both cases 2 and 3, torsion is required — and must be predicted by engineering analy-
sis.
RELATED LINKS
Video on torsion bar suspension
Online textbook on torsion bar suspension
Torsion of bars
Consider a cylinder or tube wall whose thickness t is small compared to its
mean radius Rm
P Q
R S
Assume that:
• Assume the angle of twist is small, so that tan ϕ = Rθ
.
L
Q'
P Q P'
S'
R S R'
Roθ
so that shear strain γ ≈ L
.
T
so that τ = . 8
2πR2
mt
Torsion of bars
τ T Gθ Gϕ
Hence R
= = L = R .
3
2πR t
Torsion of bars
2. Integrate from the axis r = 0 to the outside radius Ro to obtain the total
torque for a solid shaft consisting of coaxial, thin-walled tubes with thick-
ness ⅆr under equal twist:
Ro
T=G
θ
L ∫2πr3dr = G θL J
0
where
Ro
π 4 R0 πD4
J= ∫ 2πr 3dr =
2
[r ]0 =
32
.
0
3. For a hollow shaft, change the lower limit of the integral for J from zero to
the internal radius Ri:
R0
Ro
J= ∫2πr3dr = 2π [r4]R i
=
π
D4 − D4
32 o i .
0
πD4
where in this case J = . Compare this with the bending equation
32
σ M E
= = .
y I R
t/R0 Japprox / J
0.1 0.997
0.2 0.987
0.3 0.97
Torsion of bars
• Torsional stiffness at least 30 kNm rad–1.
If the outer diameter must be D0 = 50 mm, find the largest allowable inner
diameter, Di.
τ T Gθ
We know that = = , and maximum shear stress at outer diameter
r J L
τmax T max
= .
D0 /2 J
1. Calculate the J value required to satisfy torque and maximum shear stress
limits.
T max D0 /2 1000 × 0.025
J= = = 0.5 × 10−6 m4
τmax 50 × 106
Therefore, for maximum T ≥ 1 kNm: Di ≤ 32.5 mm; but for torsional stiffness
to exceed 30 kNm: Di ≤ 26.5
mm and we must choose the thicker tube on the second condition.
Torsion of bars
4. Torsional stiffness at least 30 kNm rad–1.
If the internal diameter Di = 50 mm, find the smallest allowable outer diame-
ter, Do.
τ T Gθ
We know that = = , and maximum shear stress at outer diameter
r J L
τmax T max
= .
D0 /2 J
1. Apply the torsional stiffness (K T) required (at least 30 kNm rad–1): re-ar-
τ T Gθ
range = = to give the torsional stiffness condition
r J L
TL T L L
J= = = KT .
Gθ θ G G
2. Hence for Al alloy of shear modulus G = 26.5 GPa,
0.5
J = 30000 × 9
= 5.66 × 10−7 m4.
26.5 × 10
π
3. Hence J = D4 − D4
i = 5.66 × 10
−7
, D40 − D4i = 5.76 × 10−6 and
32 0
with D40 = 6.25 × 10−6 − 5.76 × 10−6 = 1.2 × 10−5 m4 hence
D0 = 58.8 mm is required.
8
8.3 Shafts of varying diameter
Torsion of bars
Torsion of a shaft which consists of several sections with different constant diame-
ters can be analysed by treating each section separately, matching torque and ro-
tation at each junction.
Consider the case illustrated, which consists of just two sections.
A B C
T2 G2 J2
θ2
= L = K T2 = torsional stiffness of section 2.
2
Hence
TL1 T
θ1 = G J = K and
1 1 T1
TL2 T
θ2 = G J = K
1 2 T2
T1 G1 J1
2. The torsional stiffness of section 1 is defined as θ1
= L = K T1
1
T2 G2 J2
4. The torsional stiffness of section 2 is defined as θ2
= L = K T2
2
T T
Hence the total rotation of C relative to A is θ = θ1 + θ2 = K T1
+K .
T2
T
But θ
≡ K T, the torsional stiffness of the entire shaft.
1 1 1
Hence KT
=K +K .
T1 T2
Torsion of bars
a
2D D
1 2
T GJ G πD4
1. The original torsional stiffness was = =
θ L L 32
2GπD4
2. Therefore the required stiffness is .
32L
G π
3. The stiffness of section 1 after the collar is fitted will be K T1 = 2D 4
a 32
G π 4
4. Stiffness of section 2 after the collar is fitted will be K T2 = D
L − a 32
1 1 1
Hence, since = + ,
KT KT KT
1 2
32L 32a 32 L − a
4
= 4
+
2GπD Gπ 2D GπD4
L a
= + L−a
2 16
8L = a + 16 L − a
15a = 8L
8
and the required collar length isa = 15
L.
8
Table 8.4-1 Analytical results for torsion of square, triangular and elliptical cross-sections
Torsion of bars
Shape of cross-section τmax θ
Square 4.81
T
7.10
TL
a3 a4G
Eqilateral triangle 20
T
46
TL
a3 a4G
Ellipse 2 T a2 + b2 TL
π a3
πa3b3 G
Torsion of bars
is also T .
20T
2. Use τallow = from Table 8.4-1 Analytical results for torsion of square,
a3
triangular and elliptical cross-sections on page 115 to apply the stress
limit.
This condition requires T = 179.2 Nm.
TL
3. Use θallow = 46 from Table 8.4-1 Analytical results for torsion of
a4GAl
square, triangular and elliptical cross-sections on page 115 to apply the
twist limit.
This condition requires T = 24.12 Nm
4. Comparison these results to find the lowest torque limit:
Torque is thereore limited by the angle of twist, to T = 24.12 Nm.
5. For a circular cross-section, first calculate the radius of the equivalent
cross-section:
For Acircle = Atriangle, R = 14.850 mm.
6. Use results from 8.2.2 Solid and hollow circular shafts on page 109 to im-
plement the stress limit:
TR
From τallow = J
, this requires T = 288.06 Nm.
RELATED LINKS
Missouri University of Science and Technology MecMovies: see Ch 6, Torsion
Torsion of bars
T
Average shear stress: τavg = 2tAm
, where
τavg is the average shear stress acting over thickness of tube (changes with
location if t changes)
T is the resultant internal torque at cross-section
t is the thickness of the tube where τavg is to be determined
Am is the mean area enclosed within the boundary of the centreline of the
tube thickness
Torsion of bars
119
8.6 Notes
Torsion of bars
8
120
8.6 Notes