STRUCTURAL ANALYSIS

SAN2602
STUDY GUIDE

Open Rubric
Contents

LEARNING UNIT ONE - THE ANALYSIS OF STATIC DETERMINATE AND

INDETERMINATE STRUCTURE .............................................................................................

1.1 Introduction ................................................................................................................

1.2 Degree of indeterminacy ...........................................................................................

1.3 Tutorials ......................................................................................................................

LEARNING UNIT TWO -THE ANALYSES OF STATICALLY DETERMINATE

FRAMEWORKS ....................................................................................................................

2.1 Introduction ................................................................................................................

2.2 The analysis of frames ..............................................................................................

2.3 Tutorials ......................................................................................................................

LEARNING UNIT THREE - THE MOMENT AREA THEOREMS .............................................

3.1 Introduction ................................................................................................................

3.2 Application of moment area for the determination of slopes (rotations) .............

3.3 Application of moment area for determination of deflections...............................

3.4 The application of Moment area to Fixed-End Beams ...........................................

LEARNING UNIT FOUR - AXIALLY LOADED COMPRESSION MEMBERS.........................

4.1 Introduction ................................................................................................................

4.2 Bending and Euler buckling theory .........................................................................

4.3 Rankine-Gordon Equation ........................................................................................

4.4 Perry-Robertson Equation ........................................................................................

4.5 Tutorials ......................................................................................................................

LEARNING UNIT FIVE - COMBINED BENDING AND AXIAL STRESSES ............................

5.1 Introduction ................................................................................................................

5.2 Central loads ..............................................................................................................

 5.3 Loads eccentric to one axis ......................................................................................

5.4 Loads eccentric to both axes ...................................................................................

5.5 Middle third rule .........................................................................................................

5.6 Tutorials ......................................................................................................................

LEARNING UNIT SIX – SLOPE DEFLECTION .......................................................................

6.1 Introduction ................................................................................................................

6.3 Slope-Deflection method of analysis .......................................................................

6.4 Sign convention .........................................................................................................

6.5 Derivation of Slope-Deflection equation .................................................................

6.6 Standard fixed-end moments ...................................................................................

6.7 Application of Slope-Defection equations ..............................................................

6.8 Special cases .............................................................................................................

6.9 Tutorial........................................................................................................................

LEARNING UNIT SEVEN – MOMENT DISTRIBUTION METHOD OF ANALYSIS ................

7.1 Introduction ................................................................................................................

7.2 Hardy-Cross Moment Distribution Method .............................................................

7.3 Basic Design Procedure – No sway .........................................................................

7.4 Types of members and connections .......................................................................

7.5 Initial moments when joints are clamped, and loads applied ...............................

7.6 Distribution and carry-over .......................................................................................

7.7 Sinking of beam support ..................................................................................................

7.9 Simple framed structures .................................................................................................

7.10 Frames with sideways ....................................................................................................

7.11 Possible types of connections in sway consideration ................................................

7.12 Sway correction procedure ............................................................................................

7.13 Double bay portal frames ............................................. Error! Bookmark not defined.
7.14 Multi-storey frame ...........................................................................................................

7.15 Settlement of frame support ..........................................................................................

7.16 Tutorial .............................................................................................................................

LEARNING UNIT EIGHT – STRAIN ENERGY AND CASTIGLIANO THEOREM ...................

8.1 Introduction ................................................................................................................

8.2 Introduction to strain energy analysis .....................................................................

8.3 Strain energy due to direct axial stress ...................................................................

8.4 Strain energy due to bending stress .......................................................................

8.5 Strain energy due to shear stress ............................................................................

8.6 Strain energy due to torsional stress ......................................................................

8.7 In general ....................................................................................................................

8.8 First theorem of Castigliano .....................................................................................

8.9 Deflection of structures: Application of Castigliano 1 ...........................................

8.10 Second theorem of Castigliano ..............................................................................

8.11 Redundant structures: Application of Castigliano II ............................................

LEARNING UNIT NINE – VIRTUAL WORK ............................................................................

9.1 Introduction ................................................................................................................

9.2 Introduction to virtual work theory ..........................................................................

LEARNING UNIT TEN – PLASTIC THEORY OF STRUCTURES ...........................................

10.1 Introduction ..............................................................................................................

LEARNING UNIT 1:

THE ANALYSIS OF STATICALLY DETERMINATE

AND STATICALLY INDETERMINATE STRUCTURES
Learning outcomes

After studying this learning unit, you should be able to

 identify statically determinate and statically indeterminate structures

 describe the differences between statically determinate and statically

indeterminate structures

 determine or calculate the degree of indeterminacy

1.1 Introduction

The terms structure or structural system refer to the configuration or arrangement of

members for a given purpose. These structures are generally subjected to externally
applied forces (loading). Various methods of analysis are used to establish their

 stability

 strength

 suitability

The following can be analysed by using only the static equilibrium equations:

(a) the sum of forces acting vertically on the structures equal to zero

(b) the sum of the forces acting horizontally on the structure equal to zero

(c) the sum of the moments caused by the forces acting on the structure about a
point

These structures, which are analysed by using only the static equilibrium equations,
are termed statically determinate structures. Conversely, the structures that cannot
be analysed by using only the static equilibrium equations, are termed statically
indeterminate structures.
1.2 Degree of indeterminacy

A structure, which cannot be solved and whose member forces bending moments,
shear forces, support reactions and the like, and cannot be found by the use of the
three equations of static equilibrium, as is the case here, is known as a statically
indeterminate structure. The terms redundant structure or hyperstatic structure may
also be encountered. These terms are identical in meaning to statically indeterminate.

This learning unit will not show you how to solve any particular type of problem, but it
will help you decide on a method to solve a given problem. It will also assist you in
visualising how a structure “works”.

For different types of structures, different approaches are used to determine their
degree of indeterminacy. The following examples illustrate the possible or common
methods for determining the degree of indeterminacy.

1.2.1 Degree of indeterminacy for trusses (revision)

In the preceding module, we discussed that two equations are used to determine the
degree of indeterminacy for trusses.

1.2.1 a) Trusses not attached to supports

The following equation is used: 𝑏 2𝑗 3

where b = number of members

j = number of joints

If this equation is satisfied, then the structure is statically determinate.

Example 1.1
 Figure 1a Figure 1b Figure 1c

Figure 1a: b=6

j=4

6=2x4–3

6=5

6–5=1

Therefore, the structure is statically indeterminate to the first degree.

Figure 1b: b=5

j=4

5=2x4–3

5=5

5–5=0

Therefore, the structure is statically determinate.

Figure 1c: b=9

j=6

9=2x6–3

9=9
 9–9=0

Therefore, the structure is statically determinate.

1.2.1 b) Trusses attached to supports or abutments

If the trusses in example 1.1 above are attached to supports, then the equation is as
follows:

2𝑗 𝑏 𝑟

Example 1.2

Figure 2a Figure 2b Figure 2c

Figure 2a: b=6

j=4

r=3

6+3=2x4

9=8

9–8=1

Therefore, the structure is statically indeterminate to the first degree.

Figure 2b: b=5

 j=4

r=4

5+4=2x4

9=8

9–8=1

Therefore, the structure is statically indeterminate to the first degree.

Figure 2c: b=9

j=6

r=3

9+3=2x6

12 = 12

12 – 12 = 0

Therefore, the structure is statically determinate.

Further examples
 b=3
j=3
b=9
Thus: b = 2j – 3 j=6

Thus: b = 2j – 3

b = 13
j=8

Thus: b = 2j – 3

Frame

b = 10
j=5
Thus: b = 2j

The frame is statically determinate.

Reactions

Four unknowns
Three static equations
Thus, one redundant reaction
 b = 2j - 3 b + r = 2j
b = 8, j = 5 OR b = 8, r = 3, j = 5
8 = 10 - 3 = 7 8 + 3 = 2  5 (+1)
One redundant member  One redundant reaction

Figure 1.19
b = 2j - 3 b + r = 2j
b = 15, j = 8 OR b = 15, r = 3, j = 8
15 = 2  8 - 3 = 13 (+2) 15 + 3 = 2  8 (+2)
 Two redundant members  Two redundant members

1.2.2 Degree of indeterminacy for beams

(1) Simply supported beams

The three basic conditions for equilibrium, that is, H = 0, V = 0 and M = 0, are
sufficient to determine all reactions.

Consider the following cases:

 W1 W2

A B


VA VB

There are two unknown reactions: VA and VB.

H=0 – no equation
V = 0: VA + VB = W1 + W2
MA = 0: VB  = W1a + W2b

The two equilibrium equations can be solved for the two unknowns.

W/m

There are two unknown reactions: VA and VB

V = 0:

M about one of the supports = 0:

There are two unknowns and two available static equilibrium equations, therefore the
structure is statically determinate.
 W

A B
HA
a


VA VB

There are three unknowns: VA, VB and HA.

H = 0: HA = Wcos

V = 0: VA + VB = Wsin

MA = 0: a.Wsin = VB x L

Thus there are three equations with three unknowns.

(2) Cantilever beams

W
MA

HA



VA

There are three unknowns: HA, VA and MA.

H = 0: HA = Wcos
V = 0: VA = Wsin
MA = 0: MA = aWsin

The three equilibrium equations are sufficient to find the three unknowns.

(3) Propped cantilever beams

MA

w
A B


VA VB

There are three unknown reactions: VA, VB, MA.

V = 0: VA + VB = wL

MA = 0: –MA – VBL + =0

Only two equations of static equilibrium are available. A third equation must be sought
from the conditions of geometrical compatibility:

Zero deflection or a known deflection of B relative to A occurs.

OR
Zero rotation or a known rotation of A occurs.

Thus, the deflection (or rotation) of certain critical points becomes important in
affecting the magnitude of the external reactions. If B is allowed to settle relative to A,
the value of B would steadily reduce. This is not the case with statically determinate
structures.

The propped cantilever above is statically indeterminate to the first degree, or, it has
one redundant reaction. There are three unknown reactions VA, VB and MA. There are
two equations of static equilibrium, therefore, it is statically indeterminate to the first
degree.

A third equation can be set up by using moment-area or another method and the
geometrical compatibility condition that the deflection or rotation at the appropriate
point is known.

(4) Encastre beams

MA

A B


VA VB

There are four unknown reactions: VA, VB, MA and MB. Only two static equilibrium
equations are available (V = 0 and M = 0). Two further equations must be found by
using the conditions of geometrical compatibility.

The encastre beam above is statically indeterminate to the second degree or has two
redundant reactions.

(5) Continuous beams

 W

w
A B C

VB
VA VC

There are three unknowns: VA, VB and VC. There are two equilibrium equations;
therefore, the structure in the figure above has one redundant reaction. It is redundant
to the first degree. Any one of the three reactions may be taken as redundant.

(a) Two redundancies (b) Two redundancies

1.2.3 Degree of indeterminacy for frames

Consider the cantilever shown at the left. This has three support reactions at support
A: VA, HA and MA – all shown in the positive direction.

HA

MA
VA

The three equations of static equilibrium are, of course, available and all three
reactions can be found for any condition of loading.

Thus, the cantilever is statically determinate.

LEARNING UNIT 2:

ANALYSIS OF STATICALLY DETERMINATE

FRAMEWORKS

Learning outcomes

After studying this learning unit, you should be able to

• analyse statically determinate structural systems (frames)

• calculate support reactions

• determine internal actions (including shearing forces, bending moments and

axial forces) and present results graphically in the form of diagrams

2.1 Introduction

The analysis of statically determinate beams, which consist of horizontal members

only, was dealt with in the preceding module (Theory of Structures II). In this chapter,
we analyse structures made up of horizontal and vertical members that are connected
at the points where they meet by a hinge (which allows for the relative rotation of the
members) or by a rigid joint.

Before you begin to study these frameworks, it is necessary to define clearly what a
rigid joint is. As shown in the analysis of pin-jointed trusses, it is assumed (as the name
suggests) that wherever a member is connected to other members at a joint, or to a
support at a point of reaction, the connection is made by a pin which allows for the free
rotation of each member connected to that joint or support. It was also shown that
under such conditions, the force in each member will consist solely of an axial
compressive or tensile. However, if at the joint, each member is fixed firmly to another
member or members by welding or by enough bolts to ensure a stiff connection, then
each member is no longer able to rotate freely; if the joint as a whole undergoes
rotation, each member will retain the original angle it made with any member. Under
these conditions, the force in each member will no longer be just a pure axial tension
or compression, but in general each member will be subject to an axial force, a shear
force and a bending moment. Such a joint is known as a rigid joint, and a framework
containing one or more of these is known as a rigid-jointed framework.
2.2 The analysis of frames

Example 2.1

Consider the statically determinate frame shown below. The support conditions are
such that the frame is pinned at support A and supported by a roller at support B.

a) Determine the support reactions at supports A and B and sketch a free body
diagram of the frame.

b) Draw shear force, bending moment and axial force diagrams for the frame.

100 kN

60 kN
D
C

4 3

B
A HA

VA VB
2 2

Figure 2.2a

Solution:

(a) Support reactions:

FV  + = 0 VA + VB −100 = 0 or VA + VB = 100 (I )

+
FH ⎯⎯ →=0 H A + 60 = 0 or H A = −60 ( II )

MA + = 0 4  60 + 2 100 − 4VB = 0

4VB = 440 or VB = 110 ( III )
 100

60

60
A
110

10
FBD

(b)

10

11
0

60

SFD (kN)
 240

240
220

BMD (kNm)

110 (c)

10 (T)

AFD (kN)

Now consider the situation on this frame if the support at B is made pinned instead of
on rollers. There are now four support reactions, VA, HA, VB and HB. There are still only
three equations of static equilibrium available and of course four unknowns cannot be
found with just three equations.
 100 kN

60 kN

4 3
HB

HA

VB
VA 2 2

To establish this, apply the three equations to the frame as it now exists:

FV  + = 0 VA + VB −100 = 0 VA + VB = 100 (I )

+
FH ⎯⎯ →=0 H A + H B + 60 = 0 H A + H B = −60 (II )

mA = 0 −4VB + H B + 4  60 + 2 100 = 0 4VB − H B = 440 ( III )

With these three equations containing four unknowns, there are an infinite number of
solutions (i.e. any desired value may be assumed for one of the unknowns and this
will then fix the value of the other three).

You may think that taking moments about support B would provide a fourth equation;
however, if this is done the equation obtained will contain no new information that
could not have been obtained from the original three equations.

To establish this:

MB + = 0 4VA − H A + 60  3 − 100  2 = 0

or 4VA − H A = 20 ( IV )
However, equation IV is basically just a rearrangement of the previous three equations
as can be seen from the following:

4xI : 4VA + 4VB = 400

-III : – 4VB + HB = –440

4VA + HB = – 40

-II : –HA – HB = – 60

4VA – HA = 20

which is of course identical to Equation IV.

Thus, it can be seen that only three equations of static equilibrium can be obtained for
a two-dimensional or plane structure: by resolving forces vertically, by resolving forces
horizontally and by taking moments about any desired point. Having done this, taking
moments about a second support point will yield no additional information.

The frame just dealt with had four unknown reactions with only three available
equations. The number of unknowns exceeded the number of available equations by
one and the frame is therefore said to be one-degree statically indeterminate.

Now, presuming that a frictionless pin is used at the joint at point C, the top left-hand
corner of the frame, the frame will still remain stable and able to carry its loads.
 100
C
60

HB
HA

VB
VA

The bending moment at a frictionless pin must be zero. Therefore, the extra
information that is now available is that the sum of the moments of all the forces acting
either on the left-hand side or the right-hand side of C must be zero, and this will
provide the extra necessary equation.

60

HA
VA

LHS

–4 HA = 0

 HA = 0 IV

Subst. in (II) HB + 0 = –60

 HB = –60
Subst. in III 4VD + 60 = 440

380
VB = = 95
4

Subst. in I VA + 95 = 100

VA = 5

Now consider the frame ABCDE, fixed at A, pinned at E and with frictionless pins at B
and D. Its degree of indeterminacy may be found as follows:

One cut will divide the frame into two, but extra moments must be provided at B, D
and E to create two stable cantilevers.

Thus, 3 x 1 cut – 3 moments = 0, that is, the frame is statically determinate.

200 kN

100 kN
B
C 2
3
D 50 kN
A
2
E

Figure 2.2b

Moment 200
provided 1 cut 100
B

Moment 50
D
provided HA

MA
E Moment HE
VA
provided

VE
+ VA + VE = 200 (I)

+→ HA + HE = –150 (II)

A+ 100 x 3 + 200 x 4 + 50 x 1 + MA

–4 VE – HE x 1 = 0

or MA – 4 VE – HE = –1150 (III)

This frame may be analysed as follows:

+ VA + VE = 200 (I)

+→ HA + HE = –150 (II)

A+ 100 x 3 + 200 x 4 + 50 x 1 + MA

–4 VE – HE = 0

or MA – 4 VE – HE = –1150 (III)

B + LHS
MA – 3 HA = 0 (IV)

D + RHS

2 HE = 0  HE = 0 (V)

Subst. in (II) HA = –150

Subst. in (IV) MA + 450 = 0

 MA = –450

Subst. in (III) –450 – 4 VE = –1150 4 VE = 700

VE = 175

Subst. in (I) VA + 175 = 200

 VA = 25
 200

100
C
B

50
D
150
A

450

E 0
25

175

25

150 50

SFD (kN)

C
B 100

A
450 E
BMD (kNm)

Drawn on Tension side

Example 2.2

The statically-determinate frame ABC as shown in the figure below has a frictionless
pin at joint B.

Draw the bending moment, shear force and axial force diagrams for the frame.

50 KN/m C
B

2
150 KN

2
75 KN

A
2 2

Solution

a) Support reactions:

∑ FV ↑+ = 0

VA + VC − 50 × 2 = 0

VA + VC = 100 − − − − − − − − − − − − − − − − − − − − − − − − − (I)

∑ FH →+ = 0

HA + HC + 75 + 150 = 0

HA + HC = 225 − − − − − − − − − − − − − − − − − − − − − − − − − (II)
∑ MA ↷+ = 0

22
75 × 2 + 150 × 4 + 50 × − VC × 4 + HC × 6 = 0
2

6HC − 4VC = −850 − − − − − − − − − − − − − − − − − − − − − −(III)

∑ MB (LHS) ↷+ = 0

−75 × 4 − 150 × 2 − HA × 6 = 0

HA = −100 kN ←

Subst. HA in (II)

−100 + HC = −225

HC = −125 kN ←

Subst. HC in (III)

6 × (−125) − 4VC = −850

VC = 25 kN ↑

Subst. VC in (I)
VA + 25 = 100

VA = 75 kN

50 KN/m C
B
125

2
150 KN
25

2
75 KN

100

A
75
FBD

KN

125 (C)

75 KN
AFD
(C)
 1,5 m
75
50
56.25
25
125 250

25
SFD
200
(KN)
100 BMD

(KNm)

2.3 Tutorials

Further problems in statically determinate frames:

Draw bending moment, sheer force and axial force diagrams for each structure
under the given loads.

50 KN
50 KN 40 KN/m Hinge

20 KN E B C
B E
D D
C
20 KN 2m
3m G
5 KN/m

4m F
F

10kN 2m
A H

A
2m 4m 2m 2m 3m 1m 2m
 50 KN 15 KN/m 20 KN/m
B
D
C D E F C 2m
2m 3m
20 KN
20 KN E
B
3m
2m A
G

G
A F

2m 2m 1m
2m 5m 1m
LEARNING UNIT 3:

THE MOMENT-AREA THEOREMS

Learning outcomes

After studying this learning unit, you should be able to

• describe basic bending relationships

• calculate slopes (rotations) and deflections (displacements) on symmetrical

and symmetrically loaded, simply supported beams and cantilevered beams
subjected to common types of loading using the moment-area technique

• determine fixed-end (reactant) bending moments of fixed-end (encastre) beams

subjected to common types of loading using the moment-area method

3.1 Introduction

In this chapter, we revise bending relationships and introduce you to moment-area

theorems.

Consider any beam such as that shown carrying any pattern loading. The beam will
have the following parameters associated with it:

W1 W2

(i) Loading (w)

(ii) Shear (P)
(iii) Moment (M)
(iv) Slope (dy/dx)
(v) Deflection (y)
and it is necessary to establish relationships between these.

Consider a short section of the beam of length dx carrying a uniformly distributed load
(UDL) (w).

(M + dM)
M
w
F (F + dF) T2
T1
dx

For horizontal equilibrium T1 = T2 = T

Resolving vertically

 + F = wdx + F + dF = 0
T2
dF
 dF = wdx =w
dx
and F =  wdx

Taking moments about right-hand end +

𝑑𝑥
= 𝐹𝑑𝑥 − 𝑀 − 𝑤𝑑𝑥. + 𝑀 + 𝑑𝑀 = 0
2
𝑑𝑥
−𝑤𝑑𝑥. is a second-order differential and can be ignored.
2

𝑑𝑀
𝑑𝑀 = −𝐹𝑑𝑥 = −𝐹
𝑑𝑥

and M =  Fdx

 M E
= =
Recalling the basic bending equation y I R
 𝑀 1 𝑑2 𝑦
gives = = curvature = 2
𝐸𝐼 𝑅 𝑑𝑥

d2y M  dy  M
thus =  slope   =  dx
dx 2 EI  dx  EI

dy
and deflection (y)  dx
dx

The relationship between the five factors listed above is therefore as follows:
dF
loading (w) = dx

dM
 wdx = shear (F) = dx

d  dy 
EI  
 Fdx = moment (M) = dx  dx 

M  dy  d
 dx   ( y)
EI = slope  dx  = dx

dy
 dx dx = deflection (y)

It is evident that you can go from deflection to loading by a series of four differentiations
with a multiplication by El going from slope to moment, and from loading to deflection
by a series of integrations with division by El going from moment to slope.

3.2 Application of moment-area for the determination of slopes

(rotations)

 M E
= =
Consider the basic bending equation y I R , or in particular the latter part of this
M 1
=
equation which can be re-written as EI R where R is the radius of curvature to
which the beam is bent at any point along its span. It can be shown mathematically

1 d2y
= 2
that R dx for all practical purposes and the equation now becomes:

M d2y
=
EI dx 2

Applying the above equation between any two points on a beam subjected to a
bending moment and integrating the equation between these two points gives:

A B

MB
MA

BM diagram

M 2
B d y  dy 
B
 dy   dy 
=   − 
B
A EI
dx = 
A dx 2
dx =  
 dx  A  dx  B  dx  A

B M  dy   dy 
i.e. 
A EI
dx =   −  
 dx  B  dx  A

The right-hand side of the above equation gives the slope of the beam at point B minus
the slope of the beam at point A (i.e. the change in slope of the beam going from point
A to point B).

However, it is necessary to consider more carefully the left-hand side of the equation.

If the beam has a bending moment diagram of the form shown below and has a
constant value of flexural rigidity El, a diagram can be drawn representing
the values of M/EI. This will merely be a reproduction of the bending moment diagram
B M
scaled-down by the value of El and the expression
 A EI
.dx
gives the area of this M/EI
diagram between points A and B (i.e. the shaded area shown in the sketch on the left).
 A B

Therefore, the equation above can be put in words as follows:

The change in slope of a beam between any two points A and B is given by the
area of the M/EI diagram taken between these two points.

This is known as Mohr’s first theorem, or the first moment-area theorem, which is
generally referred to as “Mohr I”. It must be memorised and thoroughly understood.

To apply Mohr’s first theorem, it is necessary to calculate the area of an M/El diagram
between various points on a beam. Examples of this will now be considered.

For the purposes of this module, virtually every bending moment or M/EI diagram can
be taken as composed of combinations of triangles, trapeziums and parabolas. You
should, therefore, be thoroughly familiar with the properties of these figures.

Recall that the areas of these figures are given by the following formulae.

Triangle: 1/2 x base x perpendicular height

Trapezium: Mean height x base

Parabola: 2/3 or 1/3 the area of the circumscribing rectangle depending on which
segment of the parabola is being considered, that is,

h h h
h1

ℓ ℓ ℓ ℓ

(1) Simply supported symmetrical and symmetrically loaded beams

For example, find the slope at each end of a simply supported beam of span L, with
constant El and carrying a uniformly distributed load w.

Solution
 EI w
A B
ℓ
A B
Beam A + B = 2

Defected Shape

ℓ ℓ
A B
M

This beam is symmetrical; therefore, the slope at each end will be equal and opposite.
Thus, the change in slope going from one support to the other will be twice the slope
at each end:

M
The area of the EI diagram between A and B

2W 2 W 3
= . . =
3 8EI 12 EI

 change in slope between A and B

W 3
= 2 =
12 EI

 slope @ each end = 

W 3
=
24 EI
Example 3.1

Find the slope at each end of a simply-supported beam spanning 16 m and carrying a
UDL of 10 kN/m. El for the beam is 80 x 10 kNm2.

EI = 80 x 103
10 KN/m
320 KNm M 4 x 10-3
16
A B 16

Solution

𝑤𝑙2 10 × 162
𝑀𝑚𝑎𝑥 = = = 320 𝑘𝑁𝑚
8 8

M 320
max = = 4  10−3
EI 80  10 3

M
Area of EI diagram between A and B

= change in slope between A and B

2 128
= .4 10−3.16 = radians
3 3 103

128
=
 slope at A (ΦA) and B (ΦB) = half of change in slope between A and B 6 103

radians or 21.33 x 10-3 radians

Example 3.2

Find the slope at each end of the beam in example 3.1 above if the UDL is now
replaced by a point load of W kN at midspan.

M
Change in slope A → B = area of EI diagram
 W
EI

M
16
A B
A B
ℓ

1 w w 2
= . . = = 2
2 4 EI 8EI

w 2
 =
16 EI

Example 3.3

If the UDL in example 3.2 above is replaced by a central point load of 320 kN, calculate
the slope at each support.

320

16 16 x 10-3
1280

A B
16

w 320  16
M max = = = 1280 KNm
4 4

M max 1280
= = 16  10 −3
EI 80  103

M
Area of EI diagram between A and B

1
= .16  10−3  16 = 128  10−3 radians
2

 slope @ each end = 64 x 10-3 radians

 w2 320 162
= = = 64 10−3 radians as above
Alternatively, 16 EI 16  80 10 3

If it is required to find the slope at each end of the same beam when it is carrying both
loads (in example 3.1 and example 3.3) simultaneously, the simplest way to deal with
this is to consider each load separately and to add the slopes algebraically rather than
attempting to draw a bending moment diagram for the combined loads and then finding
the area of the combined M/El diagram.

320 KN
320 KN
10 KN/m 10 KN/m

A B

The unit for the forces is either kN or N. In general, when using moment-area
techniques with a beam carrying both UD and point loads, each load is dealt with
separately and the results are combined to give the final answer.

Final slope @ A and B

= 21.33 x 10-3 + 64 x 10-3

= 85.33 x 10-3 radians

(2) Beam symmetrically loaded over part of its span

Presuming that the beam in example 3.1 had a UDL of 10kN/m over the central 8 m
only of the beam and it is desired to find the slope at each end as before, Mohr I still
applies, but now the calculation of the M/El diagram is somewhat more involved and
it is best regarded as the sum of two triangles, a rectangle and a parabola as follows:
 10 KN/m 3 80

1 2 1a 160
4 8 4
4 8 4
40 KN 40 KN

Taking each area separately and dividing the sum by EI at the end gives:

A = 12 160  4 = 320
1.

1 A = 320

2. 160 x 8 = 1280

1280
2
3  80  8 =
3. 3

7040
𝑇𝑜𝑡𝑎𝑙 𝐴𝑟𝑒𝑎 =
3

𝑀 7040
𝑎𝑟𝑒𝑎 𝑜𝑓 𝑑𝑖𝑎𝑔𝑟𝑎𝑚 =
𝐸𝐼 3 × 80 × 103

= 29.33 10−3 radians = change in slope A → B

 slope @ A and B = 14.67 10−3 radians

Additional example:

The beam shown has an EI value of 67 x 103 kNm2. Calculate the slope at each end.

100 kN

75 kN/m

2 1 1 2
The beams dealt with to date have all been symmetrical and have been symmetrically
loaded, giving equal and opposite slopes at each support and a total change of slope
equal to twice the slope at each support.

It must be emphasised that this applies only to beams that are symmetrical and
symmetrically loaded. In all other cases, the application of Mohr’s first theorem will
only give the change in slope between the points considered and it may not be possible
to calculate the absolute value of the slope at either of these points but merely the
change in slope between them.

However, if a point on the beam is known where the slope is zero, the application of
this theorem will give the absolute value of the slope at any other point on the beam
(e.g. cantilever beams).

Example 3.4

Calculate the value of the slope at the free end of the cantilever shown:

W
EI constant
M
A
B Wℓ
ℓ

As support A is fixed slope @ A = 0

 change in slope from A → B gives the actual value of the slope at B.

W 2 W 2
M
1 W
. . =
That is, area of EI diagram = 2 EI 2 EI  slope @ B = 2 EI

Example 3.5
Find the slope at the free end of the cantilever under the two loading conditions
shown:

EI constant

𝑙 𝑤𝑙 2
𝑀𝑚𝑎𝑥 = 𝑤 × 𝑙 × =
2 2

𝑤𝑙 2
2𝐸𝐼

𝑀 1 𝑤𝑙2 𝑤𝑙3 𝑤𝑙3

𝐴𝑟𝑒𝑎 𝑜𝑓 𝑡ℎ𝑒 𝑑𝑖𝑎𝑔𝑟𝑎𝑚 = × 𝑙 × = ∴ 𝑠𝑙𝑜𝑝𝑒 𝑎𝑡 𝐵 =
𝐸𝐼 3 2𝐸𝐼 6𝐸𝐼 6𝐸𝐼

Solve for the loading condition below:

2ℓ ℓ
/3 /3 Ans:
=

Example 3.6

The cantilever shown has a flexural rigidity (EI) of 50 x 103 kNm2 and is loaded as
shown. Find the slope at the free end.

10 KN/m

2 6
Solution

1
2

3 180

300 BMD

Element Area

1 1 180 360
×6× =
3 𝐸𝐼 𝐸𝐼

2 180 360
2× =
𝐸𝐼 𝐸𝐼

3 1 120 120
×2× =
2 𝐸𝐼 𝐸𝐼

Total 840
𝐸𝐼

Slope at B
840
= = 16.8 × 10−3 𝑟𝑎𝑑𝑖𝑎𝑛𝑠
50 × 103

(3) Beams with varying EI

If the value of El changes along the span of a beam, this will not affect the bending
moment diagram but will obviously alter the M/EI diagram. This must be taken into
account and the appropriate value of El used when calculating the area of the M/El
diagram for different sections of the beam.

For example, presume that the cantilever in example 3.6 has an El of 60 × 103 for the
first 4 m of its span but that this changes to 30 × 103 for the final 4 m. It is still required
to calculate the slope at the free end.
This is dealt with by subdividing the area of the bending moment diagram into areas
that can be divided by the respective EI values as follows:

10 KN/m
A
EI = 60 x 103 EI = 30 x 103

2 2 4

2
3 1
80

4 180

300 BMD

Let Basic EI = 30 x 103 kNm2, then 60 x 103 = 2EI

Element Area

1 1 80 320
×4× =
3 𝐸𝐼 3𝐸𝐼

2 1 1
3 × 6 × 180 − 3 × 4 × 80 =
380
2𝐸𝐼 3𝐸𝐼

3 180 180
2× =
2𝐸𝐼 𝐸𝐼

4 1 120 120
×2× =
2 2𝐸𝐼 2𝐸𝐼

Total 1420
3𝐸𝐼

1420
∴ Slope @ the free end = = 15.778 × 10−3 𝑟𝑎𝑑𝑖𝑎𝑛𝑠
3 × 30 × 103
Solve for the loading condition and EI below:

Find the slope at each support of the simply-supported, variable El beam shown:

[Answer: 3.5 x 10-3 radians]

10 KN/m
A
EI = 40 x 103 EI= 80 x 103 mm2 EI=40x103

mm2 mm2
2 4 2

3.3 Application of moment-area for determination of deflections

We now return to the basic bending equation:

M d2y
=
EI dx 2

If both sides of this equation are multiplied by x it becomes:

M .x d2y
= x. 2
EI dx

If this equation is integrated between any two points (A and B) on a beam subjected
to a bending moment, it gives:

A B

x dx

B Mx B d2y
A EI
dx = A x.
dx 2
dx

If the right-hand side of this equation is integrated by parts, it becomes:

 B
B Mx  dy 
A EI
dx =  x
 dx
− y
A

It is now necessary to study the actual physical significance of both sides of this
equation.

B M M
 dx
It has already been seen that A EI gives the area of the EI diagram between A and B.

B Mx M
However, the expression
A EI
.dx
gives the first moment of area of the EI diagram
A Mx
between A and B taken about A. Equally, the expression
B EI
.dx
gives the first
M
moment of area of the EI diagram between B and A taken about B.

A B

To establish the significance of the right-hand side of the equation, consider a section
of a bent beam between point A and B.

yB

If the limits of integration are inserted into the above expression, then it becomes:

 dy   dy 
xB   − yB − x A   + y A
 dx  B  dx  A
If point A on the beam is taken as the origin (xA = 0 and yA = 0) the expression
becomes:

  dy  
 xB   − yB 
  dx  B 

From the figure, it can be seen that this is the vertical distance from point A to the
tangent to the beam drawn at point B.

B Mx   dy  
A EI
dx −  xB   − yB 
  dx  B  can be put into words as follows:
Thus, the equation

M
The moment of area of the EI diagram between any two points A and B on a beam
taken about point A gives the deflection of A relative to the tangent to the beam drawn
at point B.

This is known alternatively as Mohr’s second theorem or the second moment-area

theorem and is generally abbreviated to Mohr II.

It is a theorem of great importance and it forms the basis of much of the work that
follows in Theory of Structures and Structural Analysis. It is essential that you
memorise and thoroughly understand this theorem.

To be able to use this theorem, you should be able to calculate the first moment of
area of various geometric figures about any point, inside or outside the figure. It must
be remembered that virtually any bending moment diagram can be regarded as a
combination of rectangles, triangles, trapezia, parabolas or semi-parabolas.

These will now be dealt with in turn:

(1) Rectangle

ℓ/ ℓ/ ℓ
/

xG h xG h

A B C A B
ℓ d ℓ
 𝑙 ℎ𝑙 2
𝐴𝑟𝑒𝑎 (𝐴) = ℎ𝑙 𝑥̅𝐴 = 𝐴𝑥̅𝐴 =
2 2

About B, the result is the same:

h 2
Ax B =
2

about a third point C:

ℓ ℓ
𝐴 = ℎ. ℓ 𝑥𝐶 = (𝑑 + ) 𝐴𝑥𝐶 = ℎℓ(𝑑 + )
2 2

(2) Right-angled triangle

2L/3
ℓ/
2L/3

h x xG

A C B A
d C
ℓ d ℓ

ℎ𝑙 𝑙 ℎ𝑙2
𝐴= 𝑥̅𝐴 = 𝐴𝑥̅𝐴 =
2 3 6
2𝑙 ℎ𝑙2
𝑥̅𝐵 = 𝐴𝑥̅𝐵 =
3 3

About point C inside the triangle

2𝑙 ℎ𝑙 2𝑙
𝑥̅𝐶 = − 𝑑 𝐴𝑥̅𝐶 = ( − 𝑑)
3 2 3

About point C outside the triangle

2𝑙 ℎ𝑙 2𝑙
𝑥̅𝐶 = + 𝑑 𝐴𝑥̅𝐶 = ( + 𝑑)
3 2 3
(3) General triangle

a b

A B
ℓ

This is best dealt with by splitting into two triangles as follows:

A B
a b

ℎ𝑎 2𝑎 ℎ𝑎2
i) 𝐴= 𝑥̅𝐴 = 𝐴𝑥̅𝐴 =
2 3 3

hb b hab hb2
A= x A = a + , Ax A = +
ii) 2 3 2 6

adding the two:

ha 2 hab hb2
Ax A = + +
Total 3 2 6

(4) Trapezium
 a b

A B
ℓ

This is best treated as the sum of two triangles as follows:

a b

A B
ℓ

a a 2 2 a 2
A= xA = Ax A = xB = Ax B =
i) 2 3 6 3 3

b 2 b 2 b 2
A= xA = Ax A = xB = Ax B =
ii) 2 3 3 3 6

Total area
𝑎𝑙 𝑏𝐿 𝑙
𝐴= + = (𝑎 + 𝑏)
2 2 2

Moments of the areas about A and B

a a 2 2 a 2
A= xA = Ax A = xB = Ax B =
2 3 6 3 3

b 2 b 2 b 2
A= xA = Ax A = xB = Ax B =
2 3 3 3 6

𝑙2 𝑙2
𝑻𝒐𝒕𝒂𝒍 𝐴𝑥̅𝐴 = (𝑎 + 2𝑏) 𝑻𝒐𝒕𝒂𝒍 𝐴𝑥̅𝐵 = (2𝑎 + 𝑏)
6 6

(5) Parabola
 ℓ/ ℓ/

h
xG h
A B A B C
6 3
d ℓ
C

2
A= h
3

x A = xB = 2

h 2
Ax A = Ax B =
3

2h  
xC = d + 2 A xC = d + 
3  2

(6) Semi-parabola

h or h

ℓ ℓ

For the first type:

xG
B
ℓ

2 5 5h 2
A = .h. xA = Ax A =
3 8 12

3 h 2
xB = Ax B =
8 4

For the second type:

 h

1𝑙 3𝑙
4 4

1 1 ℎ𝑙 2
𝐴= ℎ𝑙 𝑥̅𝐴 = 𝑙 𝐴𝑥̅𝐴 =
3 4 12

3 h 2
xB = Ax B =
4 4

Knowing the properties of all the above geometrical sections means that the properties
M
of basically any EI diagram encountered can be determined.

Mohr’s second theorem states that the moment of the M/EI diagram taken between
any two points about one of these points gives the vertical deflection of that point
relative to the tangent drawn at the second point.

Example 3.7

Consider the cantilever beams subjected to loading as follows:

EI Constant
B
A

Wℓ BMD

ℓ
 Wℓ/EI
M/EI
ℓ

Thus, if it is known that at this second point, the beam is horizontal (i.e. the slope is
zero) and the deflection is also zero (e.g. as is the case at a built-in support), the
theorem can be used to find the absolute deflection of any other point on the beam.

1 𝑤𝑙 2 𝑤𝑙 2
𝐴 = 𝑤𝑙 = 𝑥̅𝐵 = 𝑙 𝐴𝑥̅𝐵 =
2 2𝐸𝐼 3 3𝐸𝐼

𝑤𝑙
∴ 𝐵𝑦 𝑀𝑜ℎ𝑟 𝐼 𝑠𝑙𝑜𝑝𝑒 𝑎𝑡 𝐵 (𝜃𝐵 ) =
2𝐸𝐼
and

𝑤𝑙2
𝐵𝑦 𝑀𝑜ℎ𝑟 𝐼𝐼 𝑑𝑒𝑓𝑙𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑡 𝐵 (∆𝐵 ) =
3𝐸𝐼

For the cantilever beam in example 3.7, the load (w) is = 10 kN, length (L) = 6 m and
EI = 80 x 103 kNm2, the maximum values of slope and deflection can be obtained as
follows:

10 kN

EI = 80x103 kNm2
B
A

6m

60
𝐸𝐼

BMD

1 60 180 2 720
𝐴= ×6× = 𝑥̅𝐵 = × 6 = 4 𝐴𝑥̅𝐵 =
2 𝐸𝐼 𝐸𝐼 3 𝐸𝐼

180
∴ 𝐵𝑦 𝑀𝑜ℎ𝑟 𝐼 𝑠𝑙𝑜𝑝𝑒 𝑎𝑡 𝐵 (𝜃𝐵 ) = = 2.25 × 10−3 radians
80 × 103

720
𝑎𝑛𝑑 𝐵𝑦 𝑀𝑜ℎ𝑟 𝐼𝐼 𝑑𝑒𝑓𝑙𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑡 𝐵 (∆𝐵 ) = = 9 × 10−3 𝑚 = 9 𝑚𝑚
80 × 103
 W EI Const

A B

1 𝑤𝑙2 𝑤𝑙3 3𝑙 𝑤𝑙4

𝐴= × ×𝑙 = 𝑥̅𝐵 = 𝐴𝑥̅𝐵 =
3 2𝐸𝐼 6𝐸𝐼 4 8𝐸𝐼

𝑤𝑙 3
∴ 𝐵𝑦 𝑀𝑜ℎ𝑟 𝐼 𝑠𝑙𝑜𝑝𝑒 𝑎𝑡 𝐵 (𝜃𝐵 ) =
6𝐸𝐼

and

𝑤𝑙4
𝐵𝑦 𝑀𝑜ℎ𝑟 𝐼𝐼 𝑑𝑒𝑓𝑙𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑡 𝐵 (∆𝐵 ) =
8𝐸𝐼

Similarly, for the cantilever beam subjected to a uniformly distributed load (UDL), if the
magnitude of the UDL is for example = 6 kN/m, the maximum values of slope and
deflection can be obtained as follows:

6 kN/m EI=80x103

6m

A B
108
𝐸𝐼

BMD

1 108 216 3 9 972

𝐴= ×6× = 𝑥̅𝐵 = × 6 = 𝐴𝑥̅𝐵 =
3 𝐸𝐼 𝐸𝐼 4 2 𝐸𝐼

216
∴ 𝐵𝑦 𝑀𝑜ℎ𝑟 𝐼 𝑠𝑙𝑜𝑝𝑒 𝑎𝑡 𝐵 (𝜃𝐵 ) = = 2.7 × 10−3 radians
80 × 103

972
𝑎𝑛𝑑 𝐵𝑦 𝑀𝑜ℎ𝑟 𝐼𝐼 𝑑𝑒𝑓𝑙𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑡 𝐵 (∆𝐵 ) = = 12.15 × 10−3 𝑚 = 12.15 𝑚𝑚
80 × 103
There may, of course, be points on a beam other than a built-in support where it is
known that the slope is zero and the tangent horizontal. For example, at the midspan
point of any symmetrical and symmetrically loaded simply supported beam, the
slope will be zero and the tangent will be horizontal. By Mohr II, the moment of the
M/EI diagram between the support and the midspan point taken about the support will
give the deflection of the support relative to the tangent drawn at midspan. Although
the support does not deflect, the midspan point does – and as the tangent here is
horizontal, Mohr II will give the deflection of this tangent, which is also the deflection
of the midspan point relative to the support (i.e. the absolute deflection of the midspan
point).

Defected Shape Defection

Consider the simply supported beams subjected to loading as follows:

Find the midspan deflection of a simply supported beam of span L and carrying a
central point load of W. El is constant.

W
EI constant

M
ℓ
A B A B
ℓ

Considering the area between support A and the midspan of the beam where the slope
is equal to zero.

A C
ℓ
/2

Mohr II says that the moment of the M/EI diagram between A and C taken about A
gives the deflection of A relative to the tangent drawn at C.
The tangent at C is horizontal; therefore, the distance at which A finishes up above
this tangent is in fact the downward movement of C (i.e. the absolute deflection of C).

Mohr II about A:

W 3
A xA Ax A  C =
48EI
W 2 W 3
16 EI 3 48EI

If the above beam carried a UDL in place of a point load, the M/EI diagram would be
a semi-parabola instead of a triangle and the calculation would be as follows:

A C
ℓ/2

A xA Ax A

2 W 2 W 3 5 5 W 3 5
. . = . = .
3 8EI 2 24 EI 8 2 16 24 EI 16

5w 4
= = C
384 EI

Again, for the simply supported beam subjected to a point load (w) at midspan, if the
magnitude of the point load (w), for example, is = 60 kN, length of the beam (L) = 8 m
and the flexural rigidity EI = 90x103 kNm2, the maximum deflection can be determined
as follows:

60 kN
120
120
𝐸𝐼
𝐸𝐼
8m
A B A B 4m

EI = 90x103 kNm2
 1 120 240 2 8 1920
𝐴= ×4× = 𝑥̅𝐴 = × 4 = 𝐴𝑥̅𝐴 =
2 𝐸𝐼 𝐸𝐼 3 3 3𝐸𝐼

1920
𝐵𝑦 𝑀𝑜ℎ𝑟 𝐼𝐼 𝑑𝑒𝑓𝑙𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑡 𝐶 (∆𝐶 ) = = 7.111 × 10−3 𝑚 = 7.111 𝑚𝑚
3 × 90 × 103

Similarly, if the beam is subjected to a UDL, then the maximum deflection can be
obtained as follows:

20 kN/m
4m 160 160
8m 𝐸𝐼 𝐸𝐼
A B A C
A 4m
EI = 90x103 kNm2 B
BMD

2 160 1280 5 5 3200

𝐴= ×4× = 𝑥̅𝐴 = × 4 = 𝐴𝑥̅𝐴 =
3 𝐸𝐼 3𝐸𝐼 8 2 3𝐸𝐼

3200
𝐵𝑦 𝑀𝑜ℎ𝑟 𝐼𝐼 𝑑𝑒𝑓𝑙𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑡 𝐶 (∆𝐶 ) = = 11.852 × 10−3 𝑚 = 11.852 𝑚𝑚
3 × 90 × 103

3.4 The application of moment-area to fixed-end beams

A fixed-end, built-in or encastre beam is a single-span beam where both supports are
completely restrained against rotation and vertical and horizontal displacements.

A B

If any loading is placed on such a beam then moments MA and MB will be developed
at the supports as follows:

W
MA MB

These moments are statically indeterminate, in other words, they cannot be calculated
by applying the three equations of static equilibrium. They will clearly depend on the
nature and magnitude of the loading in the beam and are known as fixed-end
moments (FEMs).
A major part of the utility of moment-area theorems lies in the calculation of these
fixed-end moments.

When dealing with a fixed-end beam, the final bending moment diagram is generally
regarded as composed of two individual diagrams. One diagram shows what the
bending moments caused by the loading would be if the beam were simply supported
and there were no fixed-end moments. This is known as the free bending moment
diagram. The second diagram shows the bending moments along the beam due to the
fixed-end moments only. This is known as the reactant bending moment diagram.

While the shape of the free bending moment diagrams will depend on the nature of
the loading, the reactant bending moment diagram will always be trapezoidal,
rectangular or triangular depending on the relative magnitudes and direction of the
fixed-end moments at either end.

Thus:

W
A B

FREE

BMD
A B

MA MB
 REACTANT

BMD
MB
MA

Final BMD

MB
MA

It should be noted that at each support of a fixed-end beam, the slope is zero (i.e. the
tangent is horizontal) and the deflection is zero. This means that the total change in
slope of the beam between the two supports is zero and that the deflection of each
support relative to the tangent drawn at the other is also zero.

Therefore, according to Mohr’s first theorem, the area of the final M/EI diagram
between the two supports will be zero; according to Mohr’s second theorem, the
moment of the M/El diagram between the two supports (taken about either support)
will also be zero.

These results can be used to find the fixed-end moments for any single-span beam.

Example 3.8

Find the fixed-end moments for the beam shown. El is constant.

W
A B

Free M/EI deg.

It has already been seen for a shape diagram that:

 𝑤𝑙2 𝑙 𝑤𝑙3
𝐴= 𝑥̅ = 𝐴𝑥̅𝐴 =
8𝐸𝐼 𝐴 2 16𝐸𝐼

MB
MA

It has already been seen for a trapezoidal shape that:

𝑙 𝑙2 𝑙2
𝐴= (𝑀𝐴 + 𝑀𝐵 ) 𝐴𝑥̅𝐴 = (𝑀𝐴 + 2𝑀𝐵 ) 𝑎𝑛𝑑 𝐴𝑥̅𝐵 = (2𝑀𝐴 + 𝑀𝐵 )
2𝐸𝐼 6𝐸𝐼 6𝐸𝐼

By Mohr I:

𝑤𝑙 2 𝑙
− (𝑀 + 𝑀𝐵 ) = 0
8𝐸𝐼 2𝐸𝐼 𝐴

𝑤𝑙
𝑀𝐴 + 𝑀𝐵 = − − − − − − − − − − (𝐼)
4

By Mohr II about A:

𝑤𝑙 3 𝑙2
− (𝑀 + 2𝑀𝐵 ) = 0
16𝐸𝐼 6𝐸𝐼 𝐴

3𝑤𝑙
𝑀𝐴 + 2𝑀𝐵 = − − − − − − − − − − (𝐼𝐼)
8

(𝐼𝐼) − (𝐼)

𝑊𝑙
𝑀𝐵 =
8

Substitute MB in (I):

𝑊𝑙 𝑤𝑙
𝑀𝐴 + =
8 4

𝑤𝑙
𝑀𝐴 =
4

In example 3.8, if the load (w) is equal to say 120 kN and the length of the beam (L) =
6 m. The moments generated at the supports (fixed-end moments) can be obtained
as follows:
 120 kN
A B
3m 3m

180
𝐸𝐼

Free M/EI deg.

MB
MA

1 180 540 6 1620

𝐴= ×6× = 𝑥̅𝐴 = = 3 𝐴𝑥̅𝐴 =
2 𝐸𝐼 𝐸𝐼 2 𝐸𝐼

By Mohr I:

540 6
− (𝑀 + 𝑀𝐵 ) = 0
𝐸𝐼 2𝐸𝐼 𝐴

540 2𝐸𝐼
𝑀𝐴 + 𝑀𝐵 = ×
𝐸𝐼 6

𝑀𝐴 + 𝑀𝐵 = 180 − − − − − − − − − − (𝐼)

By Mohr II about A:

1620 𝑙2
− (𝑀 + 2𝑀𝐵 ) = 0
𝐸𝐼 6𝐸𝐼 𝐴

1620 6𝐸𝐼
𝑀𝐴 + 2𝑀𝐵 = × 2
𝐸𝐼 6

𝑀𝐴 + 2𝑀𝐵 = 270 − − − − − − − − − − (𝐼𝐼)

(𝐼𝐼) − (𝐼)

𝑀𝐵 = 90 𝑘𝑁𝑚

Substitute MB in (I):
𝑀𝐴 + 90 = 180

𝑀𝐴 = 90 𝑘𝑁𝑚

120 kN
90 90
3m 3m

Beams subjected to a UDL:

EI Const.
W
A
B
ℓ

FREE

𝑤𝑙2
8
A B

MB
MA

2 𝑤𝑙2 𝑤𝑙3 𝑙 𝑤𝑙4

𝐴= × ×𝑙 = 𝑥̅𝐴 = 𝐴𝑥̅𝐴 =
3 8 12𝐸𝐼 2 24𝐸𝐼

By Mohr I:

𝑤𝑙 3 𝑙
− (𝑀 + 𝑀𝐵 ) = 0
12𝐸𝐼 2𝐸𝐼 𝐴

𝑤𝑙2
𝑀𝐴 + 𝑀𝐵 = − − − − − − − − − − (𝐼)
6

By Mohr II about A:

𝑤𝑙 4 𝑙2
− (𝑀 + 2𝑀𝐵 ) = 0
24𝐸𝐼 6𝐸𝐼 𝐴

𝑤𝑙2
𝑀𝐴 + 2𝑀𝐵 = − − − − − − − − − − (𝐼𝐼)
4
(𝐼𝐼) − (𝐼)

𝑤𝑙2
𝑀𝐵 =
12

Substitute MB in (I):

𝑤𝑙2 𝑤𝑙2
𝑀𝐴 + =
12 6

𝑤𝑙2
𝑀𝐴 =
12

For the simply supported beam above if the magnitude of the UDL is, for example, =
120 kN and the length is = 6 m, the fixed-end moments at the supports can be
calculated as follows:

EI Const.
120 kN/m
A B
6m

540
𝐸𝐼
A B

FREE BMD

MB
MA

2 540 2160 6 6480

𝐴= ×6× = 𝑥̅𝐴 = = 3 𝐴𝑥̅𝐴 =
3 𝐸𝐼 𝐸𝐼 2 𝐸𝐼

By Mohr I:

2160 6
− (𝑀 + 𝑀𝐵 ) = 0
𝐸𝐼 2𝐸𝐼 𝐴

2160 2𝐸𝐼
𝑀𝐴 + 𝑀𝐵 = ×
𝐸𝐼 6
𝑀𝐴 + 𝑀𝐵 = 720 − − − − − − − − − − (𝐼)

By Mohr II about A:

6480 62
− (𝑀 + 2𝑀𝐵 ) = 0
𝐸𝐼 6𝐸𝐼 𝐴

6480 6𝐸𝐼
𝑀𝐴 + 2𝑀𝐵 = × 2
𝐸𝐼 6

𝑀𝐴 + 2𝑀𝐵 = 1080 − − − − − − − − − − (𝐼𝐼)

(𝐼𝐼) − (𝐼)

𝑀𝐵 = 360 𝑘𝑁𝑚

Substitute MB in (I):

𝑀𝐴 + 360 = 720

𝑀𝐴 = 360 𝑘𝑁𝑚

EI Const.
120 kN/m
360 360
6m

Example 3.9

Find the fixed-end moments for the beam shown with a non-central point load. El is
constant.

W
a b

A B
ℓ

Free BM

ℓ = (a + b)
 MB
MA

A xA Ax A

Wab +a Wab ( + a )
2 EI 3 6 EI

By Mohr I:

𝑊𝑎𝑏 𝑙
− (𝑀 + 𝑀𝐵 ) = 0
2𝐸𝐼 2𝐸𝐼 𝐴

𝑊𝑎𝑏
𝑀𝐴 + 𝑀𝐵 = − − − − − − − − − − (𝐼)
𝑙

By Mohr II about A:

𝑊𝑎𝑏(𝑙 + 𝑎) 𝑙2
− (𝑀 + 2𝑀𝐵 ) = 0
6𝐸𝐼 6𝐸𝐼 𝐴

𝑊𝑎𝑏 𝑊𝑎2 𝑏
𝑀𝐴 + 2𝑀𝐵 = + 2 − − − − − − − − − − (𝐼𝐼)
𝑙 𝑙

(𝐼𝐼) − (𝐼)

𝑊𝑎2 𝑏
𝑀𝐵 = 2
𝑙

Substitute MB in (I):

𝑊𝑎2 𝑏 𝑊𝑎𝑏
𝑀𝐴 + =
𝑙2 𝑙

𝑤𝑎𝑏2
𝑀𝐴 = 2
𝑙

For the beam in example 3.9, if the load (w) and the length (L) were given (e.g. 100 kN
and 10 m, respectively) and the load is applied at 4 m from support A, the fixed-end
moments can be obtained as follows:
 100 kN
4m 6m

A B
10 m

240
𝐸𝐼
1 2

Free BM

MB
MA

1 240 480 2 8 1280

1. 𝐴= ×4× = 𝑥̅𝐴 = ×4= 𝐴𝑥̅𝐴 =
2 𝐸𝐼 𝐸𝐼 3 3 𝐸𝐼

1 240 720 1 4320

2. 𝐴= ×6× = 𝑥̅𝐴 = ( × 6 + 4) = 6 𝐴𝑥̅𝐴 =
2 𝐸𝐼 𝐸𝐼 3 𝐸𝐼

1200 5600
𝑇𝑜𝑡𝑎𝑙 𝐴= 𝐴𝑥̅𝐴 =
𝐸𝐼 𝐸𝐼

By Mohr I:

1200 10
− (𝑀 + 𝑀𝐵 ) = 0
𝐸𝐼 2𝐸𝐼 𝐴

1200 2𝐸𝐼
𝑀𝐴 + 𝑀𝐵 = ×
𝐸𝐼 10

𝑀𝐴 + 𝑀𝐵 = 240 − − − − − − − − − − (𝐼)

By Mohr II about A:

5600 102
− (𝑀 + 2𝑀𝐵 ) = 0
𝐸𝐼 6𝐸𝐼 𝐴
 5600 6𝐸𝐼
𝑀𝐴 + 2𝑀𝐵 = × 2
𝐸𝐼 10

𝑀𝐴 + 2𝑀𝐵 = 336 − − − − − − − − − − (𝐼𝐼)

(𝐼𝐼) − (𝐼)

𝑀𝐵 = 96 𝑘𝑁𝑚

Substitute MB in (I):

𝑀𝐴 + 96 = 240

𝑀𝐴 = 144 𝑘𝑁𝑚

100 kN
4m 6m

144 96
10 m

To find the fixed-end moments for a beam subjected to a couple somewhere along
its length, the procedure is as follows:

a b

C
A B

A 1

2
B

MB
MA
 𝐶𝑎2 2𝑎 𝐶𝑎3
𝐴1 = 𝑥̅𝐴 = 𝐴𝑥̅𝐴 =
2𝐸𝐼𝑙 3 3𝐸𝐼𝑙

𝐶𝑏 2 1𝑏 𝐶𝑎𝑏 2 𝐶𝑏 3
𝐴2 = 𝑥̅𝐴 = (𝑎 + ) 𝐴𝑥̅𝐴 = +
2𝐸𝐼𝑙 3 2𝐸𝐼𝑙 6𝐸𝐼𝑙

Total area:
𝐶𝑎2 𝐶𝑏 2 𝐶 𝑎2 − 𝑏 2
− = ( )
2𝐸𝐼𝑙 2𝐸𝐼𝑙 2𝐸𝐼 𝑙

Total moment of area about a:

𝐶𝑎3 𝐶𝑎𝑏 2 𝐶𝑏 3 𝐶 2𝑎3 − 3𝑎𝑏 2 − 𝑏 3

−( + )= ( )
3𝐸𝐼𝑙 2𝐸𝐼𝑙 6𝐸𝐼𝑙 6𝐸𝐼 𝑙

By Mohr I:

𝐶 𝑎2 − 𝑏 2 𝑙
( )− (𝑀 + 𝑀𝐵 ) = 0
2𝐸𝐼 𝑙 2𝐸𝐼 𝐴

𝐶 2
𝑀𝐴 + 𝑀𝐵 = (𝑎 − 𝑏 2 ) − − − − − − − − − − (𝐼)
𝑙2

By Mohr II about A:

𝐶 2𝑎3 − 3𝑎𝑏 2 − 𝑏 3 𝑙2
( )− (𝑀 + 2𝑀𝐵 ) = 0
6𝐸𝐼 𝑙 6𝐸𝐼 𝐴

𝐶
𝑀𝐴 + 2𝑀𝐵 = (2𝑎3 − 3𝑎𝑏 2 − 𝑏 3 ) − − − − − − − − − − (𝐼𝐼)
𝑙3
𝑙
If the right-hand side of equation (I) is multiplied by one (i.e. by ), the equation will not
𝑙

change. Knowing that l = a + b, equation (I) will become

𝐶 2 (𝑎 + 𝑏)
𝑀𝐴 + 𝑀𝐵 = 2
(𝑎 − 𝑏 2 ) ×
𝑙 𝑙

𝐶 3
𝑀𝐴 + 𝑀𝐵 = (𝑎 − 𝑎𝑏 2 + 𝑎2 𝑏 − 𝑏 2 ) − − − − − − − − − − (𝐼)
𝑙3
(𝐼𝐼) − (𝐼)

𝑊𝑎2 𝑏
𝑀𝐵 =
𝑙2

C
II − I : M B = 3
(a 3 − 2ab 2 − a 2b)

Ca
= 3
(a 2 − 2 2
− ab)

Ca
= 3
(a 2 − b 2 − ab − b 2 )

Ca
= 3  (a − b)(a − b) − b(a + b 

Ca
= 3  ( a − b) − b 

Ca
= 2
(a − 2 )

Cb
MA = − 2
(b − 2a )
Similarly,

For the beam in example above, if the actual magnitude of the applied moment (C)
and the length (L) were given as shown below, the fixed-end moments can be
calculated as follows:

3m 5m

150 kNm
A B

8m

56.25 18.75
A 1

2
B
93.75

18.75
 MB
MA

1 −56.25 84.375 2 168.75

1. 𝐴= ×3× =− 𝑥̅𝐴 = ×3=2 𝐴𝑥̅𝐴 = −
2 𝐸𝐼 𝐸𝐼 3 𝐸𝐼

1 93.75 234.375 1 14 1093.75

2. 𝐴 = ×5× = 𝑥̅𝐴 = ( × 5 + 3) = 𝐴𝑥̅𝐴 =
2 𝐸𝐼 𝐸𝐼 3 3 𝐸𝐼

150 925
𝑇𝑜𝑡𝑎𝑙 𝐴= 𝐴𝑥̅𝐴 =
𝐸𝐼 𝐸𝐼

By Mohr I:

150 8
− (𝑀 + 𝑀𝐵 ) = 0
𝐸𝐼 2𝐸𝐼 𝐴

150 2𝐸𝐼
𝑀𝐴 + 𝑀𝐵 = ×
𝐸𝐼 8

𝑀𝐴 + 𝑀𝐵 = 37.5 − − − − − − − − − − (𝐼)

By Mohr II about A:

925 82
− (𝑀 + 2𝑀𝐵 ) = 0
𝐸𝐼 6𝐸𝐼 𝐴

925 6𝐸𝐼
𝑀𝐴 + 2𝑀𝐵 = × 2
𝐸𝐼 8

𝑀𝐴 + 2𝑀𝐵 = 86.719 − − − − − − − − − − (𝐼𝐼)

(𝐼𝐼) − (𝐼)

𝑀𝐵 = 49.219 𝑘𝑁𝑚

Substitute MB in (I):

𝑀𝐴 + 49.219 = 37.5

𝑀𝐴 = −11.719 𝑘𝑁𝑚
𝐅𝐨𝐫 𝐞𝐱𝐚𝐦𝐩𝐥𝐞, 𝐭𝐡𝐞 𝐛𝐞𝐧𝐝𝐢𝐧𝐠 𝐦𝐨𝐦𝐞𝐧𝐭𝐬 𝐠𝐞𝐧𝐞𝐫𝐚𝐭𝐞𝐝 𝐚𝐭 𝐭𝐡𝐞 𝐬𝐮𝐩𝐩𝐨𝐫𝐭𝐬 𝐰𝐢𝐥𝐥 𝐡𝐚𝐯𝐞 𝐭𝐡𝐞 𝐬𝐚𝐦𝐞

𝐝𝐢𝐫𝐞𝐜𝐭𝐢𝐨𝐧 𝐨𝐩𝐩𝐨𝐬𝐢𝐧𝐠 𝐭𝐡𝐞 𝐝𝐢𝐫𝐞𝐜𝐭𝐢𝐨𝐧 𝐨𝐟 𝐭𝐡𝐞 𝐚𝐩𝐩𝐥𝐢𝐞𝐝 𝐦𝐨𝐦𝐞𝐧𝐭.

3m 5m

150 kNm
11.79 49.219

8m

(1) Sinking supports

If support B of the fixed-end beam shown settles downwards by an amount  relative

to support A but without rotation of either support, bending moments will be induced
at each support of the beam. You should be able to calculate these.

A B

There is no loading on the beam and therefore no free bending moment diagram.

The reactant bending moment diagram will be as shown on the left below.

The change in slope from A to B is zero and the deflection of support B relative to the
tangent at A will be:

MA MB

By Mohr I:

(M A + M B ) = 0  M A = −M B (I )
2 EI

By Mohr II about B:
 2
6 EI 
(2M A + M B ) = − 2 M A + M B = − 2
( II )
6 EI

(𝐼𝐼) − (𝐼)

6𝐸𝐼∆
𝑀𝐵 = +
𝑙2

From I:

6𝐸𝐼∆
𝑀𝐴 = −
𝑙2

A B

For the example above, if the magnitude of the deflection (∆) and the length of the
beam (L) were given, say as, ∆ = 6 mm and L = 7 m, and the flexural rigidity EI = 85 ×
103 kNm2, the fixed-end moments can be obtained as follows.

By Mohr I:

7
(𝑀 + 𝑀𝐵 ) = 0
2𝐸𝐼 𝐴

𝑀𝐴 + 𝑀𝐵 = 0 − − − − − − − − − − (𝐼)

By Mohr II about A:

72
(𝑀 + 2𝑀𝐵 ) = −∆ (𝐶𝑜𝑛𝑠𝑖𝑑𝑒𝑟𝑖𝑛𝑔 𝑎 𝑠𝑖𝑔𝑛 𝑐𝑜𝑣𝑒𝑛𝑡𝑖𝑜𝑛 𝑤ℎ𝑒𝑟𝑒 𝑑𝑒𝑓𝑙𝑒𝑐𝑡𝑖𝑜𝑛𝑠
6𝐸𝐼 𝐴

𝑑𝑜𝑤𝑛𝑤𝑎𝑟𝑑𝑠 𝑎𝑟𝑒 𝑛𝑒𝑔𝑎𝑡𝑖𝑣𝑒)

6 × 85 × 103
𝑀𝐴 + 2𝑀𝐵 = −6 × 10−3 ×
72

𝑀𝐴 + 2𝑀𝐵 = −62.449 − − − − − − − − − − (𝐼𝐼)

(𝐼𝐼) − (𝐼)

𝑀𝐵 = −62.449 𝑘𝑁𝑚
Substitute MB in (I):

𝑀𝐴 − 62.449 = 0

𝐹𝑜𝑟 𝑒𝑥𝑎𝑚𝑝𝑙𝑒 𝑡ℎ𝑒 𝑚𝑜𝑚𝑒𝑛𝑡𝑠 𝑤𝑖𝑙𝑙 ℎ𝑎𝑣𝑒 𝑡ℎ𝑒 𝑠𝑎𝑚𝑒 𝑑𝑖𝑟𝑒𝑐𝑡𝑖𝑜𝑛 𝑜𝑝𝑝𝑜𝑠𝑖𝑛𝑔 𝑡ℎ𝑒 𝑑𝑒𝑓𝑙𝑒𝑐𝑡𝑖𝑜𝑛 𝑚𝑜𝑡𝑖𝑜𝑛.

62.499

A B
62.499

If support B had been pinned instead of fixed, but still settled by ∆ relative to A, the
bending moment at B would be zero and the reactant bending moment diagram would
be as shown:

MA

ℓ
A B

Mohr II about B:

𝑀𝐴 𝑙 2𝑙 3𝐸𝐼∆
× = −∆ ∴ 𝑀𝐴 = −
2𝐸𝐼 3 𝑙2

that is, half the value in the previous case.

(2) Rotating supports

Constant EI
A B

l
The change in slope of the beam going from A to B will be from –θ to θ (i.e. +θ). It can
be seen from the figure above that the deflection of B relative to the tangent drawn at
A will be ℓθ.

MA
MB

Mohr I:

2 EI
M A + M B  =  MA + MB = −I
2 EI

Mohr II about B:

2
6 EI
 2M A + M B  =  2M A + M B = − II
6 EI

4 EI 2 EI
II − I : MA = and MB = −

A
A B

Again, if the rotation (θ), length (L) and flexural rigidity (EI) were given, for example,
θ = 4x10-3 radians, L = 6 m and EI = 60x103 kNm2, the fixed-end moments could be
calculated as follows:

By Mohr I:

6
(𝑀 + 𝑀𝐵 ) = 𝜃
2𝐸𝐼 𝐴

2𝐸𝐼
𝑀𝐴 + 𝑀𝐵 = 𝜃 ×
6

−3
2 × 60 × 103
𝑀𝐴 + 𝑀𝐵 = 4 × 10
6

𝑀𝐴 + 𝑀𝐵 = 80 − − − − − − − − − − (𝐼)
By Mohr II about B:

62
(2𝑀𝐴 + 𝑀𝐵 ) = 𝑙 × 𝜃
6𝐸𝐼

6𝐸𝐼
(2𝑀𝐴 + 𝑀𝐵 ) = 𝑙 × 𝜃 ×
62

6 × 60 × 103
2𝑀𝐴 + 𝑀𝐵 = 6 × 4 × 10−3 ×
62

2𝑀𝐴 + 𝑀𝐵 = 240 − − − − − − − − − − (𝐼𝐼)

(𝐼𝐼) − (𝐼)

𝑀𝐴 = 160 𝑘𝑁𝑚

Substitute MB in (I):

160 + 𝑀𝐵 = 80

𝑀𝐵 = −80 𝑘𝑁𝑚

𝐹𝑜𝑟 𝑒𝑥𝑎𝑚𝑝𝑙𝑒, 𝑡ℎ𝑒 𝑚𝑜𝑚𝑒𝑛𝑡𝑠 𝑤𝑖𝑙𝑙 ℎ𝑎𝑣𝑒 𝑡ℎ𝑒 𝑠𝑎𝑚𝑒 𝑑𝑖𝑟𝑒𝑐𝑡𝑖𝑜𝑛 𝑜𝑝𝑝𝑜𝑠𝑖𝑛𝑔 𝑡ℎ𝑒 𝑟𝑜𝑡𝑎𝑡𝑖𝑜𝑛 𝑚𝑜𝑡𝑖𝑜𝑛.

If support B had been pinned rather than fixed, the bending moment at B would of
course be zero and the reactant bending moment would be calculated as follows:

MA
0

1 𝑀𝐴 𝑀𝐴 𝑙 2𝑙 𝑀𝐴 𝑙2
𝐴= ×𝑙× = 𝑥̅𝐵 = 𝐴𝑥̅𝐵 =
2 𝐸𝐼 2𝐸𝐼 3 3𝐸𝐼

By Mohr II

𝑀𝐴 𝑙2
=𝑙×𝜃
3𝐸𝐼

3𝐸𝐼 3𝐸𝐼𝜃
𝑀𝐴 = 𝑙 × 𝜃 × =
𝑙2 𝑙
that is, three-quarters of the value in the previous case.

If the values of L, θ and EI used previously are applied, the fixed-end moment at A
would be:

3𝐸𝐼𝜃 3 × 60 × 103 × 4 × 10−3

𝑀𝐴 = = = 120 𝑘𝑁𝑚
𝑙 6

3.5 Tutorial

1. Using the moment-area method, determine the maximum slope (rotation) and
deflection (displacement) for the two cases a) and b) below.

a)

40kN
5kN/m
A B

4m 2m

b)

80kN 80kN

A B

2m 4m 2m

2. For the encastre beams shown below, calculate the fixed-end (reactant) bending
moments in each case using the moment-area method.
a)

120kN
40kN/m
B
A

2m 4m 2m

b) Under the loading support A rotated by 3.2 x 10-3 radians.

2500kN

B
A

3m 4m

c) Under the loading support B subsided by 8 mm.

300kN

B
A

5m 3m
LEARNING UNIT 4:
AXIALLY LOADED COMPRESSION MEMBERS

Learning outcomes
After studying this learning unit, you should be able to
• define the slenderness ratio
• identify about which axis a structural member subjected to compressive axial
loading will buckle
• describe the relationship between buckling stress and the slenderness ratio
• calculate the radius of gyration and the slenderness ratio
• determine buckling stresses according to the Euler method, Rankine-Gordon
method and Perry-Robertson equation
• calculate the buckling load

4.1 Introduction
A strut may be defined as a member carrying an axial compressive load whose length
is generally, but not always, large in comparison with its lateral dimensions.

When a member whose length is much greater than its lateral dimensions is subjected
to a compressive load, it tends to fail by buckling (i.e. by lateral or sideways
movement of the member at some point along its length).

Thus, a long slender concrete column would fail by buckling, whereas a short stocky
concrete member (e.g. a concrete test cube) would fail through crushing of the
concrete.

Short member fails by crushing

Long member fails by buckling

To understand this phenomenon, it is necessary to study the mechanism of buckling

and determine what properties of the section govern whether it would fail by buckling
or crushing and at what point the changeover from one type of failure to the other
would occur.

4.2 Bending and Euler buckling theory

Consider an originally perfectly straight member loaded with an axial compressive
force as shown below. The ends of the member are assumed to be fixed in position,
but free to rotate (i.e. the member has pinned ends).

The member is assumed to have the following properties:

Length L
Area A
Second moment of area I
Young’s modulus E

All of these are expressed in their appropriate dimensions.

The deflection at any point of the member (denoted y) may be upwards (in which case
it will be positive) or downwards (in which case it will be negative). With an upwards
or positive deflection, the bending moment will be hogging or negative; while with a
downwards or negative deflection, the bending moment will be sagging or positive.

P y P P
-y

It is also known from the theory of bending that:

d2y
M = EI
dx 2
d2y
thus EI = − Py
dx 2
d 2 y Py
or + =0
dx 2 EI
P
Let 2 =
EI

d2y
the n 2
+ 2 y = 0
dx

This is an ordinary, second-order, linear differential equation with constant coefficients.

To solve it, consider the following equation:

y = ACos  x + BSin  x

Differentiating with respect to x:

dy
= −  A Sin  x +  B Cos  x
dx

Differentiating again
d2y
2
= −  2 ACos  x −  2 B Sin  x
dx
−𝜇2 [𝐴𝐶𝑜𝑠𝜇𝑥 + 𝐵𝑆𝑖𝑛𝜇𝑥]

= − 2 y
d2y d2y
 2 = −  y or
2
2
+  2 y = 0, The original equation.
dx dx

Thus y = ACos  x + B Sin  x is a solution of the equation and it is now necessary to

find the constants of integration A and B. This is done by considering the boundary
conditions of the member (that at x = C, y = 0; and at x = 1, y also = 0).

x=0 y=o
 ACos 0 + B Sin 0 = 0
A 1 = 0 A= 0
and y = B Sin  x

x= y=0
 B Sin  = 0

B cannot be zero as that would indicate zero deflection anywhere along the length of
the beam, which is clearly not the case.

Thus: Sin ℓ = 0

The solutions of this equation are as follows:

𝜇ℓ = 0
𝜇ℓ = 𝜋
𝜇ℓ = 2𝜋
and in general
𝜇ℓ = 𝑛𝜋 where 𝜋 is an integer.
The lowest value of P at which the strut will buckle is obviously when n = 1, and in this
𝜋 𝜋2
case: 𝜇 = or𝜇2 =
ℓ ℓ2

P 2  2 EI
 = 2 or P= 2
EI

This load is known as the Euler buckling load, named in honour of the mathematician
who first developed this theory, and is denoted as Pe (the “e” standing for Euler).

Example 4.1
A strut, pin-jointed at both ends has a square cross-section, 60 mm x 60 mm and is
1.5 m long. Young’s modulus is 200 GPa. Calculate the lowest value of load at which
the strut will buckle and calculate the stress existing in the strut at that load.

Solution

𝜋2 𝐸𝐼
𝑃𝑒 = 𝑘𝑁 𝐸 = 200 × 106 𝑘𝑁⁄𝑚2
𝑙2

0.64
I= = 1.08 10−6 m4
12
= 1.5m
 2  200 106 1.08 10−6
 Pe =
1.52
𝑃𝑒 = 947 𝑘𝑁

P 947 103 N
Stress = = = 263 MPa
A 60  60mm2

This strut has a pinned joint at either end. If the end conditions had been different, it
is likely that the buckling load would also have been different. It is now necessary to
investigate this fact. For example, consider a cantilever subjected to an axial
compressive load. The properties of the cantilever are taken as the same as the
previous strut, with the exception of the end conditions. It is required to find the
buckling load for the cantilever.
 EI

Again y = ACos  x + B Sin  x

Taking the free end of the cantilever as the origin:

@x =0 y=o  ACos 0 + B Sin 0 = 0
or A = 0 y = B Sin  x
𝑑𝑦 𝑑𝑦
@𝑥 =𝑙 =0 = 𝜇𝐵𝐶𝑜𝑠𝜇𝑥
𝑑𝑥 𝑑𝑥
  B Cos  = 0

Again, B cannot be zero.

n
Cos  = 0 giving solutions  =
2
n
or  = P
EI
=
2
The value of n giving the lowest value of P is n = 1.
 2 EI
 Pe =
(2 ) 2
An identical expression to that obtained for a pinned strut except that (2ℓ) replaces ℓ.

At this stage, it is convenient to introduce the concept of effective length (denoted as

le), which can replace the term l in the equation and which will allow for the fact that
the end conditions for any strut may be other than pinned.

Effective length le = Kℓ where ℓ = actual length, and

K = relevant multiplying factor for the end
conditions.

 2 EI
Thus, for a member pin-jointed at each end, that is, le = 1 and Pe = 2 .

 2 EI  2 EI  2 EI
For a cantilever, le = 2l and Pe = = = [ K = 2]
(2 ) 2 ( e) 2 ( K )2
Struts with end conditions other than pinned or completely free will obviously have
different values of effective length than the two cases we already dealt with. These
may be found by substituting the appropriate values of end displacements and
rotations into the basic equation.

A diagram showing the values of effective lengths corresponding to certain frequently

encountered types of end conditions is reproduced below.

ℓ
ℓe = 0.7L ℓe = 0.5L
ℓe = L ℓe =2L
L
0.7 ℓ 0.5 ℓ
2ℓ

K= 1 K= 0,7 K= 0,5 K= 2

For practical purposes, such as design, it is normally more useful to know the stress
that exists in a section when buckling is just about to occur rather than the force at
which this will happen. The stress may be found by modifying the equation for the
Euler buckling load as follows:

 2 EI
Pe =
( e) 2
𝑃𝑒 𝜋2 𝐸𝐼 𝜋2 𝐸 𝐼
Stress 𝜎𝑒 = = = 𝑥
𝐴 𝐴(ℓ𝑒)2 (ℓ𝑒)2 𝐴

I
But = r 2 where r (radius of gyration) = I
A
A

 2 Er 2  2E
 e = or c =
( e) 2  e
2

 
 r 

e  K 
The expression  or  , which is effective length divided by the radius of
r  r 
gyration, is a very important concept in structural analysis and is called the
slenderness ratio.
It can be seen from the expression for the Euler buckling stress

 2E
c =
( e / r)
2

that the higher the value of the slenderness ratio, the lower the value of the stress at
which the strut will buckle. Conversely, the lower the slenderness ratio, the higher will
be the value of stress in the strut at buckling (i.e. the stronger the strut will be).

It must be realised at this stage that a strut in general has an infinite number of
slenderness ratios, as both the effective length and the radius of gyration may be
different for every axis selected passing through the cross-section of the strut.

Example 4.2

Consider a steel strut with a solid rectangular cross-section 75 mm × 50 mm and 3 m

long. Young’s modulus is 200 GPa. The end connections are as shown on the
following diagram and are such that the strut is fully fixed about both axes at its lower
end. At its upper end, the strut is fully fixed about its minor principal axis (i.e. it cannot
deflect or rotate about its minor principal axis [Y-Y]), but it is pinned about its major
principal axis (i.e. it cannot deflect but it can rotate about its major principal axis [X-
X]).

It is required to find the buckling stress for the strut and the corresponding axial load,
which will cause this to occur.
 Y

X X

3m

Y
50

Y
X
75
X
Y
Pinned Fixed

ℓe = 0,7 ℓ
ℓ ℓe = 0,5 ℓ

Fixed Fixed

Solution
About the X-X axis:

𝑙𝑒 = 0.7𝑙 = 0.7 × 3000 = 2100 𝑚𝑚

I x− x bd 312 d2 d 75
rx − x = = = = =
A bd 12 2 3 2 3

𝑙𝑒 2100
= 21.7 𝑚𝑚 = = 97
𝑟 21.7
 2E  2  200 103
e = 2
= MPa = 210 MPa
  (97) 2
 e
r 

𝑃𝑒 = 𝜎 × 𝐴 = 210 × 103 × (0.05 × 0.075) = 788 𝑘𝑁

About the Y-Y axis:

e = 0,5 = 0,5  3000 = 1500 mm

 50 1500
ry − y = = 14.4 mm e
= = 104
2 3 ry − y 14.4

𝜋 2 × 200 × 103
𝜎𝑒 = = 183 𝑀𝑃𝑎
1042
𝑃𝑒 = 𝜎 × 𝐴 = 183 × 103 × (0.05 × 0.075) = 686𝑘𝑁

The strut will therefore buckle about the Y-Y axis with a stress of 183 MPa when a
load of 686 kN is applied. It clearly will never be able to sustain the load of 788 kN,
which would cause buckling about the X-X axis because by that time the strut would
already have buckled about the Y-Y axis.
 
Thus, a strut will buckle about that axis about which its slenderness ratio  e  is a
 r 
maximum.

Exercise 4.3

(a) Look up the values of I and r for both axes of a steel (grade 300W) section –
350 × 165 × 46I – in a table of steel section properties (e.g. the SAISC
Handbook).
(b) If that I section were to be used as a strut that is 5 m long so that it is fully fixed
at both ends for bending about the weaker axis, and fully pinned at both ends
for stronger axis bending, about which axis would it fail?

2
If the expression for the critical Euler buckling stress  e = r E 2 is plotted against
 e
 
 r 

 
the slenderness ratio  e  a graph such as that shown below will be obtained.
 r 
 

e

(ℓe/r)

It can be seen from this graph that for large values of slenderness ratio, the
buckling stress will be low; however, for small values of slenderness ratio, it
tends to be high, may be very high and may in fact exceed the yield stress
of the material. In this case, it is clear that the strut will fail by crushing the material
before it buckles.

For example, consider a rectangular concrete test specimen 100 mm × 100 mm and
1.2 m long. Young’s modulus for the concrete is 13 GPa and its characteristic crushing
strength is 40 MPa. If the critical buckling stress for the specimen is calculated, the
result is as follows:

1.2 m

100
100

P
  2E
e = 2 e = (pinned ends) = 1200 mm
 e
 
 r 

bd 3
For a rectangle I = A = bd
12
bd 3 d2 d
r= = =
12bd 12 2 3

100
 here r = = 28.9 mm
2 3
1200
e
= = 41.6
r 28.9

 2 13 103
 e = = 74.25 MPa
(41.6) 2

This value, of course, exceeds the value of the crushing stress for the concrete (which
is 40 MPa) and it is clear that the specimen would fail by crushing the concrete before
it has the opportunity to buckle. This behaviour is characteristic of struts, which have
relatively small values of slenderness ratio.

Clearly it is necessary to modify the theory already established in order to determine

whether any particular strut will crush or buckle, and at what value of stress or load it
will do either of these.

A possible approach is to modify the graph of buckling stress against the slenderness
ratio by inserting a flat top in the graph at the point where the buckling stress becomes
equal to the crushing stress as follows:
 Yield Stress (y)

e

ℓe
/r

However, this has some of the following disadvantages:

• The graph cannot be expressed as a single mathematical function for

calculation purposes.

• Because the compressive stresses due to the direct load and those due to
bending caused by the eccentricity of this load at the centre of the strut interact,
and also because no strut is both perfectly straight and homogenous, it is found
that the failure stresses predicted by this theory do not correspond with the
values established through practical testing. It is difficult to modify the graph to
account for these discrepancies.

• It is only at fairly high values of ℓe/r that the results of practical tests become
reasonably close to what the theory predicts.

• For struts with a very low value of slenderness ratio, the failure stress is
effectively the yield stress (y), as the buckling stress is very large compared to
the yield stress.

• For struts with very high slenderness ratios, the failure stress is effectively the
buckling stress (e) as this is very small compared with the yield stress.

• At values of the intermediate slenderness ratio, the two phenomena (yielding

and buckling) interact and the theory breaks down.

These compression members may be classified into three categories:

(1) Short, square members, which tend to crush rather than buckle. These have
low values of ℓe/r. For these, the failure stress c is given by c = y (crushing
or yield stress).

(2) Long, slender members, which tend to buckle in a perfect elastic manner.
These have rather longer values of slenderness ratio. The failure stress c is
given by c = e (Euler stress).

(3) Members that exhibit intermediate behaviour between the first two categories.
There is some degree or yielding in the member cross-section before the pure
Euler load is reached, and this has the effect of reducing the failure load. These
struts will still buckle rather than crush, but the plastic yielding affects the
behaviour.

The struts that fail by plastic buckling have very difficult failure behaviour from an
analysis perspective. Unfortunately, most real compression members of practical
dimensions have slenderness ratios that fall into this intermediate category.

4.3 Rankine-Gordon equation

Rankine and Gordon suggested a simple, mathematical approximation for dealing with
intermediate behaviour struts.

If the actual failure stress for the strut is designated c, the following relationship is
assumed:

1 1 1
= +
𝜎𝑐 𝜎𝑦 𝜎𝑒
1
For low values of slenderness ratios, e becomes very large and becomes very
𝜎𝑒

small.
1 1
Thus: = or  c =  y
c y
1
For high values of slenderness ratio, e becomes very small and very large
𝜎𝑒

1 1 1
compared to . 𝑇ℎ𝑢𝑠 = 𝑜𝑟 𝜎𝑐 = 𝜎𝑒
𝜎𝑦 𝜎𝑐 𝜎𝑒
Since this equation holds for the extreme cases of slenderness ratios, it is reasonable
to assume that it also gives a good approximation for struts with intermediate values
of slenderness ratios.

1 1 1
Thus = +
c y e

e + y
=
 y e

 y e
 c =
e + y

y y
= =
  
2
1+ y   e

 e 1+ y  r 
 2E
y
, which is a dimensionless constant for any particular material, is denoted “a”.
 2E

y
 c = 2
 
1+ a  e 
 r 

Consider once more the concrete test specimen dealt with previously. It was
rectangular, 100 mm × 100 mm and 1.2 m long. Young’s modulus was 13 GPa and
crushing strength was 40 MPa. The buckling stress for this specimen was calculated
at 74.25 MPa and it was clear that the actual stress in the specimen could never reach
this value as it would crush at a value of 40 MPa first.

However, under actual testing conditions, the specimen would fail before even
reaching the crushing stress because of the interaction caused by the two different
effects of the loading: the direct compression plus the extra compressive stress
caused by the bending, which in turn is due to the fact that at midspan, the member
will deflect, producing an eccentricity of the compressive load – hence a bending
moment and a bending compressive stress.

Applying the Rankine equation to this member:

 y
c = 2
 
1+ a  e 
 r 
y 40
a= = = 312 10−6
 E 2
 13 103
2

e
= 41, 6
r
40 40
c = =
1 + 312 10  ( 41, 0 )
−6 2
1.54

= 26 MPa

Thus, the specimen would fail at a compressive stress of 26 MPa, which is

considerably less than both the calculated Euler buckling stress of 74.25 MPa and the
concrete crushing stress of 40 MPa.

The load which would cause this failure stress can be calculated as:

26 x 1002/103 = 260 kN

This example makes it clear that a relatively slender specimen, such as the one just
dealt with, would not be satisfactory for carrying out tests on concrete cube strengths
owing to its slenderness and tendency to buckle. It would indicate a concrete cube
strength far less than the correct value. For this reason, standard dimensions for
concrete compressive test specimens of 150 × 150 × 150 are specified. If such a
standard specimen is considered for the same concrete with characteristic strength of
40 MPa, Young’s modulus of 13 GPa, applying the Rankine equation, would give the
following result:

y 150
c = 2
r= = 43.3 mm e = 150 mm
  2 3
1+ a  e 
 r 
150 40
e
= = 3.46 a= = 312 10 −6 [As previously]
r 43.3  13 103
2
 40 40
c = −6
=
1 + 312 10  (3.46) 2
1.0037
= 39.9 𝑀𝑃𝑎

This gives a much more realistic estimate of the actual cube strength of the concrete
and is the reason why such dimensions are standard for concrete test specimens.

4.4 Perry-Robertson equation

It should be borne in mind that the Rankine-Gordon equation is based on an
approximation. Perry and Robertson suggested an alternative that is based on a more
complicated version of the Euler formula.

Struts with intermediate values of slenderness are particularly sensitive to out-of-

straightness and the presence of residual stresses. For intermediate struts with out-
of-straightness effects, the failure stress (Stress at first yield – Onset of plasticity) is
given by c, where:

1
 c =  y + (1 +  ) e  −
2  
1
4 y
2
+ (1 +  ) e  −  y e 
Where y = yield stress
e = Euler stress, and
= is a variable whose value is determined in such a way as to reflect
the degree of initial out-of-straightness.

𝑙𝑒
In the Perry-Robertson version of the above equation, 𝜂 = 0.003
𝑟

This value for  was determined empirically and so should always be subject to a
factor of safety.

Example 4.4
A rectangular bar of grade 300W steel has the dimensions 100 mm × 150 mm × 2 m
length and is to be used as a pin-ended compression member. Assuming a load factor
of 1.5, what is the permissible load that can be resisted by this member?

Young’s modulus for steel = 200 GPa

Solution

We must determine the failure levels for each of the three possible modes of
behaviour, that is, the strut behaves as a short member, as a slender member, or as
an intermediate member. We can then compare these loads and determine the failure
load.

(1) Crushing or yielding failure (short column behaviour)

c = y = 300 MPa
 Failure load P = 300 x Area
Area = 100 x 150 = 15 x 103 mm2
P = 300 x 15 = 4500 kN

(2) Elastic buckling failure (Euler theory) (slender column behaviour)

 2E
c = e = 2
 
 e
r 


b =100

X X d =150

ℓe = ℓ
A A
Y
Section A – A

d
For a rectangular cross-section r=
2 3
Bending about X-X axis,
150
𝑟= = 43.3
2√3
𝑙𝑒 2000
∴ = = 46.2
𝑟 43.3

Bending about Y-Y axis,

100
𝑟= = 28.9 𝑚𝑚 ∴
2√3
𝑙𝑒 2000
= = 69.2
𝑟 28.9

Buckling will take place about the Y-Y axis more readily.

𝜋 2 × 200 × 103
𝜎𝑒 = = 412 𝑀𝑃𝑎
69.22
𝑃𝑒 = 412 × 103 × (0.1 × 0.15) = 6180 𝑘𝑁

(3) Plastic buckling failure (Perry-Robertson formula) (intermediate behaviour)

1
 c =  y + (1 +  ) e  −
2  
1
4 y
2
+ (1 +  ) e  −  y e 
𝑙𝑒
𝜂 = 0.003 = 0.003 × 69.2 = 0.2076
𝑟

1 1
𝜎𝐶 = [300 + (1 + 0.2076) × 412] − √{ [300 + (1 + 0.2076) × 412]2 − 300 × 412}
2 4

398.8 − 188.2 = 210.6 𝑀𝑃𝑎

The failure load 𝑃𝑒 = 210.6 × 103 × (0.1 × 0.15) = 3159 𝑘𝑁

Thus, we found that failure takes place by plastic buckling with the section behaving
like an “intermediate slenderness” strut. The working load is found from
𝑃 3159
𝑊𝑜𝑟𝑘𝑖𝑛𝑔 𝑙𝑜𝑎𝑑 = = = 2106 𝑘𝑁
𝑙𝑜𝑎𝑑 𝑓𝑎𝑐𝑡𝑜𝑟 1.5

SABS equations

Consult the structural steel book (the “red book”) and South African National
Standards (SANS) to familiarise yourself with the equations provided for the
determination of buckling stress and buckling load.

Example 4.5
A column consists of four grade 300W steel 100 x 75 x 10 unequal angles arranged
as shown in the sketch. The column, which is 10 m long, is so connected that it is
pinned at each end about its major principal axis X-X and pinned about one end but
fixed at the other, about its minor principal axis Y-Y.

The required properties of a 100 x 75 x 10 angle are summarised in the attached table.
The yield stress for grade 300W steel is 300 MPa and Young’s modulus is 200 GPa.

Calculate

(a) The Euler buckling load for the column

(b) The crushing load for the column
 Y
Y

X X 300 X X

ax

ay

Y
Y

250

A = 1661 mm2
Ixx = 1.618 x 106 mm4
Iyy = 0.776 x 106 mm4
ax = 31.9 mm
ay = 19.5 mm

Solution

Section properties A = 4 x 1661 = 6644 mm2

𝐼𝑥𝑥 = 4[1.1618 × 106 + 1661(150 − 31.9)2 ]

= 99.1 × 106 𝑚𝑚4

99.1 × 106
𝑟𝑥𝑥 = √ = 122 𝑚𝑚
6644

𝐼𝑦𝑦 = 4[0.776 × 106 + 1161(125 − 19.5)2 ]

= 77.1 × 106 𝑚𝑚4

77.1 × 106
𝑟𝑦𝑦 = √ = 108 𝑚𝑚
6644

ex − x = 10 000 e y − y = 0.7 x 10 000 = 7 000

 𝑙𝑒 10000 𝑙𝑒 7000
( ) = = 82 ( ) = = 64.8
𝑟 𝑥𝑥 122 𝑟 𝑦𝑦 108

The higher value of 82 is critical. The column will buckle about the X-X axis.
𝜋2 𝐸 𝜋2 ×200×103
(a) 𝜎𝑒 = 2 = = 294 𝑀𝑃𝑎
𝑙
( 𝑒⁄𝑟 ) 822

𝑃𝑒 = 𝜎 × 𝐴 = 294 × 103 × 6644 = 1953 𝑘𝑁

(b) 𝜎𝑦 = 300 𝑀𝑃𝑎

∴ 𝑃𝑒 = 𝜎𝑦 × 𝐴 = 300 × 103 × 6644 = 1993 𝑘𝑁

4.5 Tutorials

Question 1

A strut fabricated from grade 300W steel (i.e. yield stress of 300 MPa) has the cross-
section shown in the figure below. The strut is so supported that its effective length
about all the axes is 13 m. The yield stress of the steel is 300 MPa and Young’s
modulus is 206 GPa.

Calculate

(a) The axial compressive force that would produce yield stress in the strut.
[39.000 kN]

(b) The Euler buckling load for the strut. [24.300 kN]
(c) The failure load for the strut according to Rankine’s theory. [15.000 kN]

50

150

150

50

100 100 100 100

In the experimental determination of the failure loads for 12.5 mm diameter, steel, pin-
ended columns; two of the values obtained were:

(i) length 500 mm, load 9800 N

(ii) length 200 mm, load 26400 N

(a) State, with reasons, whether these values conform with the Euler theory.

(b) Assuming that both values conform with the Rankine formula, determine the
yield stress (σy) and the constant (a) in this formula, E = 200 GPa.

Answer

a) (i) Yes. Approximately Pe = 9460 N

(ii) No. Pe = 59139 N. It crushes before it can buckle.
b) σy = 320 MPa a = 119 x 10-6

Question 2

An axially loaded column fabricated from grade 300W steel has a rectangular cross-
section 400 mm × 500 mm with two circular holes, each 200 mm in diameter, drilled
in the positions shown in the figure below.

Take Young’s modulus as 206 GPa and yield stress of 300 MPa.

Support conditions for the column are such that it has an effective length about its X-
X or major axis of 15 m, while that about its Y-Y or minor axis is 13.75 m.

Calculate

(1) The axial compressive force which would produce yield stress in the section
(2) The Euler buckling load for the column
(3) The failure load for the column according to the Rankine theory

125

200
Dia

135

X X

125

200
Dia

125

Y
200 200

Question 3
A column consists of two grade 300W steel channels with the dimensions shown and
arranged as in the figure below. The column is 5 m long and is connected so that it is
pinned at each end about its major principal axis X-X, but pinned at one end and fixed
at the other about its minor principal axis Y-Y. The yield stress for grade 300W steel
is 300 MPa and Young’s modulus is 206 GPa.

Calculate
(a) The axial compressive force that would produce yield stress in the column
(b) The Euler buckling load for the column
(c) The failure load on the column according to the Rankine theory

Y
10 10

15

X X 270

15

100 100

(All dimensions in mm)

LEARNING UNIT 5:
COMBINED BENDING AND AXIAL STRESSES

Learning outcomes

After studying this learning unit, you should be able to

• describe the stress generated due to axial loading applied at the centroid of the
section of a structural member or pad foundation
• describe the resultant stresses generated due to loading applied at an
eccentricity
• determine the magnitude of the resultant stress generated at a point due to
central loads applied at an eccentricity, and draw a resultant stress diagram
along the edges of a section or pad foundation

5.1 Introduction
Combined bending and direct stress were dealt with in the previously taught Theory
of Structures module. The effect on an elastic structure of any loading system will be
the algebraic sum of the effects of the individual components of that loading system.

5.2 Central loads

Consider a rectangular structural section with cross-sectional area A, as shown in the
figure below, subjected to a compressive force W acting at its centroid. This force will
produce a uniform compressive stress equal to W/A on the section.
 y

x x

[Plan Area = A] y

W
Elevation

Stress Compression

5.3 Loads eccentric to one axis

If it is presumed that the line of action of the load is moved by a distance e from the Y-
Y axis while remaining on the X-X axis (as shown in figure 1 below), this will cause an
increase in the compressive stress on the left-hand side of the section and a decrease
in that on the right-hand side.

x x

[Plan Area = A]
y

e
W
Elevation

Compression
Stress

Figure 1
If the distance e is increased further it is clear that the stress on the right-hand side of
the section will firstly become zero and then become tensile, as shown in figure 2.

x x

[Plan Area = A]
y

e
W
Elevation

Compression Tension
Stress

Figure 2

If the condition where the load W is applied at distance e from the Y-Y axis is
considered in more detail as in figure 3(a) below, the force system of figure 3(a) can
be replaced by the sum of the two force systems: a force (w) applied at the centroid
and a moment (we) as shown in figure 3(b), which produces stress distributions as
shown in figure 3(c). These stresses are added algebraically to obtain the resultant
stress distribution, as given in figure 3(d).
 FIG 3
y
e
W

(a)
y

e
W W

W (b)

W W

(c) W

We
W

+
(d)

(T)
W
/A (c) (c)

(e)

(T)

(c)

(f)

Figure 3

Thus, a force W acting at a distance e from a centroidal axis of a section can be

replaced by a force W acting on the centroidal axis plus a moment “We” acting in the
appropriate direction about this axis.
The stresses produced by the force and the moment are shown in the diagrams in
figure 3(e) and these can be combined to give the final stress diagram shown in figure
3(f).

To calculate the total stress at any point on the section, the procedure is to calculate
the direct axial stress W/A and the bending stress Wex/I, and to add these together
algebraically to obtain the resultant stress.

In these problems, compressive stresses are normally taken as positive and tensile as
negative – although this is an arbitrary convention and the opposite may be adopted
if desired.

Note: Bending is taking place about the Y-Y axis in this illustration.  I is Iyy. The
variable x refers to the x-coordinate of the point at which the stress is determined, and
the eccentricity e is given in terms of the x-coordinate of the point of application of the
load.

Example 5.1

Calculate the value of the resultant stress acting along lines A, B, C, D and E of the
rectangular section shown, which is subjected to a compressive force of 480 kN acting
at an eccentricity, e, of 75 mm to the left of the Y-Y axis of the rectangle.

A B C D E
50 50 50 50

75

120

480 KN

200

Solution:

A = 200 x 120 = 24 x 103 mm2

I = 120 x 2003/12 = 80 x 106 mm4

W/A = 480 x 103/[24 x 103]

= +20 MPa (compression is taken as positive)

M = 480 x 75 = 36 x 106 N-mm

At each section  = W/A ± Mx/I.

At A x = 100

W Mx
= +
A I

36 106 100
= 20 + = 20 + 45
80 106
= 65 MPa (compression)

At B x = 50
36 106  50
 = 20 + = 44.5 MPa (c)
80 106

At C x=0  = 20 + 0 = 20 MPa (c)

At D x = 50 but bending stress is now tensile.

36 106  50
 = 20 − = 20 − 22.6 = 2.5(T ) MPa
80 106

36×106 ×100
At E 𝑥 = 100 𝜎 = 20 − = 20 − 45 = −25
80×106

= 25 MPa (T)

If all the results calculated are plotted on a diagram, this will appear as shown on the
previous page. It should be noted that the distribution of stress across the section is a
straight line and therefore only two results are necessary to plot the diagram, find the
stress at any point and find the position of zero stress.
 44.4

-25

20

44,5

65

To calculate the position where resultant stress is zero (i.e. the neutral axis of the
section as distinct from the centroidal axis), x is inserted as a variable in the expression
for resultant stress. This is put equal to zero and the resulting equation solved to find
x. In this case, Mx/I will clearly be tensile and therefore negative.

W Mx
Thus: − =0
A I

36 × 106 𝑥 36𝑥
20 − = 0 = 20 𝑥 = 44.4 𝑚𝑚
80 × 106 80

Example 5.2

The T-section shown below is subjected to a compressive force of 100 kN acting on

its X-X axis and 25 mm from the left-hand face of the section. Draw a diagram showing
the distribution of resultant stress across the section.

25

x
100 20

100
KN

30 120

Section properties:
 Element b d A x Ax 𝒅𝒃𝟑 ̅ − 𝒙)𝟐
𝑨(𝒙
𝟏𝟐
1 30 100 3000 15 45000 225000 3333333.333
2 120 20 2400 90 216000 2880000 4166666.667
5400 261000 3105000 7500000

261000
𝑥̅ = = 48.333 𝑚𝑚 𝐼𝑦𝑦 = 31050000 + 7500000 = 10.605 × 106 𝑚𝑚4
5400

48.3 y
101.7

48.3 101.7
a
25

b
100
KN
a e = 48.3 – 25
e
= 23.3
y

W Wex
@ a−a = +
A I

100 103 100 103  23.3  48.3

= +
5.4 103 10.6 106
= 18.5 + 10.6 = 29.1 MPa (c)

𝑊 𝑊𝑒𝑥
@𝑏−𝑏𝜎 = −
𝐴 𝐼

100 103  23.3 101.7

= 18.5 −
10.6 106
= 18.5 − 22.4 = − 3.9 = 3.9 MPa (T )

W W cx
To calculate position of neutral axis − =0
A I
 I 10.6 106
x = = = 84.2 mm from centroid
Ac 5.4 103  23.3
or 84.2 + 48.3 = 132.5 from left-hand face

132.5

3.9 MPa (T)

29.1 (c) MPa

5.4 Loads eccentric to both axes

In the preceding examples, all the loads were eccentric to only one axis. Where the
load is eccentric to both the X-X and the Y-Y axes, the procedure is only slightly more
difficult.
 y
\ Comp
A B
A-B
e
W
W
ey Stress for
x bending about
x x-x only

Plan C D C-D

y Side Tension
W \ elevation

A-C B-D

Front
elevation

Tension

Comp

Stress for
bending about
y-y only

The figure above shows a load W that is eccentric to both axes on a rectangular
section ABCD. The front elevation is the familiar diagram showing the effect of bending
about the Y-Y axis. The side elevation shows bending about the X-X axis. To find the
stress at any point, you must combine the two effects. Thus, the stress at the point is
given by:

Wex.x Wey. y W
  + (Compression positive)
I yy I xx A

For example, in the figure above:

Wex .xA Wey. y A W

At A A = + + +
I yy I xx A

Wex .xB Wey. yB W

At B B = − + +
I yy I xx A
 Wex .xC Wey. yC W
At C  C = + − +
I yy I xx A

Wex .xD Wey. yD W

At D  C = − − +
I yy I xx A

Example 5.3

The rectangular section below is subjected to a load of 250 kN at the eccentricities

shown. Determine the stresses at each of the corners.

y
A \ B

250 KN

x x 2m

0,7 m
C D

y
\
3m

Solution

3000  20003
I xx = = 2 1012 mm4 A = 3000  2000
12
ey = 500 mm = 6 106 mm2

2000  30003
I yy = = 4.5 1012 mm4 W = 250 103 N
12
ex = 700 mm

Coordinates are : ax = −1500, a y = +1000

bx = +1500 by = +1000
cx = −1500 c y = −1000
d x = +1500 d y = −1000

W Wex .xA Wey . y A

 A = − +
A I yy I xx
 1 700 1500 500 1000 
= 250 103  − + 
 6 10 4.5 1012 2 1012 
6
 = 250  103 (166  10−9 − 233  10−9 + 250  10−9 ) = 45.8  10−3 MPa
= 45.8 kPa (Compression)
By similar calculations B = 162.3 kPa (compression)
C = 79.3 kPa (tension)
and D = 37.3 kPa (compression)

5.5 Middle-third rule

Certain materials used in engineering are incapable of taking tension, for example,
mass concrete or the mortar joints in brick or masonry structures. Thus, if a load is to
be placed on a structure containing such materials, it will be necessary to know within
what zone the load can be placed in order to avoid producing tension in the section.

Consider the rectangle shown below. If a load P is placed on the Y-Y axis at a distance
e from X-X, it could cause tension at A-A.

y
A \ A

d x x

B B

y
\b

The limiting case will be where the stress at A-A is zero, and in this case:

If the load is represented as P and Zxx =I/y, then

P Pe
− =0
A Zxx
P 6 Pe
 =
bd bd 2

6c
or =1
a

or c = d 6
This means that the maximum distance at which the load can be placed below the X-
X axis without producing tension at A-A is d/6.

Clearly, the same will apply with regard to producing tension along B-B, that is, the
maximum distance the load can be placed above X-X is also d/6.

Thus, it can be seen that the safe zone in which a compressive P can be placed without
causing tension anywhere in the section is a rectangle d/3 wide centred on the X-X
axis. This zone is known as the middle-third of the section.

d/3

d/6
d/3
d/6

d/3

However, it must be emphasised that this rule, known as the middle-third rule,
applies to rectangular sections only and the safe zone for any different section will
have a completely different shape.

5.6 Tutorials

Question 1

The cantilever shown below has a span of 9 m and a uniform rectangular cross-
section, 2 m deep by 1 m wide. It carries an axial tensile load of 24 kN and a transverse
load of 2 kN applied at its tip as shown. Calculate the stresses at points A, B and C of
the cross-section at the support of the cantilever.

[Answers: @ A 39 kPa (tension); @ B 15 kPa (compression);

@ C 1.5 kPa (compression)]
 1 000

A
2 KN

24 KN
C
9m 500

Question 2

The pre-stressed concrete T-section shown below is subjected to a compressive force

F as shown.

F F

240

80

NA
e
240

80

(a) Find the maximum permissible value of e if there is to be no tensile stress at

the top of the section.
(b) For this value of e, calculate the maximum permissible value of F if the
compressive stress in the section is not to exceed 20 MPa.

Answers

(a) 75.6 mm
(b) 288 kN
Question 3
A compressive force of 45 kN acts parallel to the longitudinal axis of the I-section
shown below and in the position indicated. If the area of the cross-section is 1500 mm2,
Ixx = 2.7 x 106 mm4 and Iyy = 0.9 x 106 mm4, find the positions of greatest tensile and
compressive stresses and the values of these.
[80 MPa (c) @ B; 20 MPa (t) @ D]

ℓy = 10

40

A B

20

45 KN ℓx = 30
100

D C
1
100
LEARNING UNIT 6:
SLOPE DEFLECTION

Learning outcomes

After studying this learning unit, you should be able to

• analyse both structurally determinate and indeterminate beams and portal

frames using the slope deflection method

• apply the slope deflection method in analysing structures with different end-
support conditions

• analyse continuous beams, which have a relative deflection or sinking of

supports

• draw dimensioned shear force diagrams and bending moment diagrams for
beams and frames

• determine support reactions for any given loading on beams and frames

6.1 Introduction

In this learning unit, we will revise the moment-area method and demonstrate the slope
deflection method using examples.

6.2 Derivation of slope deflection equations

We briefly restate Mohr’s first and second theorems.
 Any loading system
A l1 l2 B

A B
A1 A2 Bending moment diagram

x2 M
diagram
x1 EI
A B
A1 A2
EI 1 EI 2

B
Bent beam Tangent at B
A 
Tangent at A
A
Unbent beam
Figure 6.1

Mohr’s first theorem (Mohr I)

The change in slope between the tangents to any two points A and B in an elastically
M
bent member is equal to the area of the diagram between these points.
EI
B
M A1 A2
=  EI dx = EI
A 1
+
EI 2

Mohr’s second theorem (Mohr II)

The intercept of carried tangents at A and B at A is equal to the moment of the area
M
of the diagram between A and B about A.
EI

B
M A x1 A x 2
A =  xdx = 1 + 2
A
EI EI 1 EI 2

6.3 Slope deflection method of analysis

The basic assumptions are as follows:

(1) The effects of direct stress, shear and tension are negligible compared to those
due to bending and can be disregarded.

(2) The structure is elastic, and the simple theory of bending applies.

In this method, the structure investigated will be considered to consist of a number of

members spanning between joints. Each member will initially be assumed to have
the joints at its ends fully fixed against rotation and deflection. The fixed-end moments
due to the loading can be easily calculated from the previous theory.

In section 6.6, a number of standard fixed-end moments (FEMs) is given.

The bending moment diagram for each member would now consist of a trapezium,
labelled (i) to (iv) in the diagram, due to the end moments, superimposed upon the
free moment diagram. The position would be as shown in the figure below.
 w kN/m

A B
Joints A and B L
imagined fully fixed.
MA
(ii)
MB
(iii)
-
(i)
kNm - (iv)
+
A B

Figure 6.2
Bending moment diagram for member A-B

The problem now is to determine the modification which must be made to the end
moments MA and MB if joints A and B are released and allowed to rotate and deflect
up to its final shape.

6.4 Sign convention

 Undeflected member

A A
B

MAB
deflected  MBA
 
member

MA + MB
L B
SAB=
MA + MB
L L
SBA=
L
3
M AB
EI M
diagram
EI
M AB
2L
EI
3

Figure 6.3

Note. The shaded portion is self-cancelling when using Mohr I and Mohr II.

A typical member A-B, of length L is shown in the figure. Due to rotation at A and B
and deflection at B relative to A, end moments MAB, MBA and end shears SAB, SBA are
MA + MB
induced. The end shears SAB and SBA will be =
L

In the sketch above all quantities are drawn in the positive direction:

MAB, MBA – end moments acting in A-B – positive if clockwise

 A,  B – rotation of A and B, that is, the slopes of the tangents at A and B –
positive if clockwise
 – deflection of B relative to A – positive if B settles with respect to A (or
if the whole member A-B turns in a clockwise direction)
SAB, SBA – end shears acting on A-B – positive if the couple acts anticlockwise
(this couple balances the moments)
Also: L = length of A-B
 = angle between the tangents at A and B
  = B – A
 = intercept on the vertical axis through B, made by the tangents at A and
B
  =  – LA
EI = flexural rigidity of A-B (= constant)

6.5 Derivation of the slope deflection equations

To derive the slope deflection equations, we refer to figure 6.3 and the following
notation definitions thereof:

Mohr I

M
 = area of diagram between A and B
EI

1  M AB  1  M BA 
=   ( L) −   ( L)
2  EI  2  EI 
Equation (1)
1
= ( M AB − M BA )
2 EI

Mohr II

M
 = moment of area of diagram between A and B, about B
EI
M AB L  2 L  M BA L  L 
=  −  
2 EI  3  2 EI  3 
Equation (2)
L2
= (2 M AB − M BA )
6 EI
 L
From Equation (1):  = B – A = ( M AB − M BA )
2 EI

2EI
 MAB - MBA = ( B −  A ) Equation (3)
L

L2
From Equation (2):  =  – LA = (2M AB − M BA )
6EI

6EI   
 2MAB - MBA =  − A Equation (4)
L L 

On solving simultaneous Equations (3) and (4) for MAB and MBA, we get

2 EI  3 
MAB =  2 A +  B −  Equation (5)
L  L

and
2 EI  3 
MBA =   A + 2 B −  Equation (6)
L  L

Equations (5) and (6) represent the changes which would occur in the end moments
of a loaded member if the joints, initially imagined fully fixed, were released.

The actual end moments on the loaded member A-B will therefore be, by
superposition:

2 EI  3 
 2 A +  B −  + M AB
F
MAB = Equation (7a)
L  L 

2 EI  3 
  A + 2 B −  + M BA
F
MBA = Equation (7b)
L  L
 F F
where M BA , M BA are the fixed-end moments acting on A-B if it is imagined fully fixed
at A and B. These fixed-end moments follow the same sign convention as the end
moments MAB and MBA, namely positive clockwise.

Equation (7a) and (7b) are the slope deflection equations.

6.6 Standard fixed-end moments

For the purpose of applying the slope deflection equations, it is necessary to know the
values of fixed-end moments for various loadings.

(1) Uniformly distributed loading

W total

L
W = wL

(2) Central point load

W

(3) Non-central point load

W

a b
L
b=L-a

(4) Triangular loading

A B
x x
For the triangular loading:
Apply the formula for the non-central point load and integrate:

x( L − x )
2
 xw 
M F
= −  x 
A
 L L2
2
− wL2
L


w
M F
A = − 3  x 2 ( L − x ) dx =
L 0 30

Note. Many other standard cases of fixed-end moments can be found in The red
book.

6.7 Application of slope defection equations

Slope deflection equations can be used to analyse a wide variety of statically

indeterminate structures. The procedure is as follows.

(1) Divide the structure into a number of straight members of constant flexural
rigidity EI and write down the number of unknown rotations and deflections. At
most there can be one rotation and one deflection per joint.

(2) Evaluate the fixed-end moments MFAB and MFBA for each span. If it is not a
standard case it can easily be done using moment-area or mathematical
analysis.

(3) Set up the slope deflection equations for each end moment throughout the
structure, by substituting in Equation 7 (a) and (b).

(4) Set up equations relating the end moments, using conditions of static
equilibrium (V = 0, H = 0 and M = 0). Note that as many equilibrium
equations must exist as there are unknown rotations and deflections in
Equation (1).
(5) Use the equilibrium equations from Equation (4) to solve for these unknown
rotations and deflections. Note that this will result in as many simultaneous
equations as there are rotations and deflections.

(6) Substitute the values obtained from Equation (5) into the original slope
deflection equations to obtain values for all the end moments.

(7) Draw the bending moment diagram by super-imposing the free bending
moment diagram on the trapezoidal diagram due to the end moments.

(8) The end shears and reactions can be calculated by considering the static
equilibrium of each span.

Example 6.1

We are required to calculate all the bending moments and reactions for the beam
shown, then draw the bending moment and shear force diagrams. EI is constant.

100 kN

30 kN/m

A 2m B C

4m 4m

The solution to the above problem is carried out using steps (i) to (iv) below. The steps
are a summary of the procedure detailed in the previous page.

(i) Determine the known and unknown displacement and rotations at the supports.
Solution A = 0 A = 0
B = ? B = 0 Known and unknown ’s and ’s
C = 0 C = 0
(ii) Determine support moments using the slope deflection equations and
equilibrium equations.

For equilibrium of joint B:

MBA + MBC = 0

2 EI  3 
M BA =  2 B +  A −  + M BA
F

   
2 EI 100  4
= (2 B + 0 + 0) +
4 8
M BA = EI B + 50

2 EI  3 
M BC =  2 B +  C −  + M BC
F

   
30  ( 4)
2
2 EI
= (2 B + 0 + 0) −
4 12
M BC = EI B − 40
 ( EI B + 50) + ( EI B − 40) = 0
−5
B =
EI
2 EI  3 
M AB =  2 A +  B −  + M AB
F

   
2 EI  5  100  4
= 0 − + 0 −
4  EI  8
5
= − − 50 M AB = - 52,5 kNm
2
2 EI  3 
M BA =  2 B +  A −  + M BA
F

   
2 EI  2( −5)  100  4
=  + 0 + 0 + M BA = +45kNm
4  EI  8
2 EI  3 
M BC =  2 B +  C −  + M BC
F

   
 30( 4)
2
2 EI  −10
=  + 0 + 0 − M BC = −45kNm
4  EI  12

Note: MBA + MBC = = +45-45 = 0

 2 EI  3 
M CB =  2 C +  B −  + M CB
F

   

 30(4)
2
2 EI  5
= 0− + 0 + M CB = +37,5 kNm
4  EI  12

(iii) Calculate the support reactions by considering the static equilibrium of each
span.

100 kN 45
52.5 37.5
30 kN/m
2m 2m

4m 4m

48.125 61.875
51. 875 51.875
110

52,5 100 37,5

45 45 30 kN/m
2 2

4m 4m

VA VB1 VB2 VC

MB = 0 MB = 0

4VA – 200 + 45 – 52.5 = 0 4VC + 45 – 37.5 - 240 = 0

VA = 51.875 kN VC = 58.125
 VB = 48.125 kN  VB = 61.875
Total VB = 48.125 + 61.875
= 110 kN

(iv) Draw the bending moment and shear force diagrams.

 61.875
51.875

Shear force diagram (kN)

48.125
58.125

52.5 48.75
45
37.5

60
100
Bending moment diagram (kNm)

18.75

51.25

Example 6.2

We are required to calculate MAB for the beam shown below and sketch the bending
moment diagram. EI is constant.

100 kN
200 kNm
48 kN/m

A B C

10 m 2m
-400 kNm +400 kNm
48 kN/m 100 kN
-200 kNm

A C
B
10 m 2 m
B

The solution to the above problem is carried out using steps (i) to (iii) below:

(i) First, we determine the known and unknown displacement and rotations at
the supports.

Solution: A = 0 A = B = 0

B = ?

(ii) Then we determine support moments using the slope deflection equations.

Equilibrium of joint B

MBA + MBC = 0

2 EI
(0 + 2B – 0) + 400 – 200 = 0
10

which gives B = − 500

EI

Substitute back in slope deflection equation:

MAB = 2 EI (0 − 500 – 0) – 400

10 EI
MAB = –500 kNm (i.e. anticlockwise)

MBA = + 200 kNm (i.e. clockwise)

MBC = –200 kNm (i.e. anticlockwise)

 (iii) Draw the bending moment and shear force diagrams.

500

200

Bending moment - -
diagram [kNm]
+

Note: The reactions can now be found in the usual way and the shear force diagram
drawn. The position of the maximum positive bending moment can be found
from the shear force diagram and its value calculated.

Example 6.3

120 kN 70 kN 70 kN 50 kN
2.25 m 1.25 m 1.25 m 1.25 m 200 kN total

A EIAB= 8500 cm4 B EIBC= 6500 cm4 C EICD= 5500 cm4 D E

4.5 m 3.75 m 3.75 m 1.2 m

We are required to calculate the bending moments at the supports and draw the
bending moment diagram.

The solution to the above problem is carried out using steps (i) to (ii) below:

(i) Determine the known and unknown displacement and rotations at the supports.

Solution: A = ? C = ? All ’s = 0

B = ? D = ?

(ii) Determine support moments using the slope deflection equations.

Joint A: MA = 0
2EI AB F
(2A + B – 0) + M AB = 0 Equation (1)
 AB

− 120(4,4 )
F
M AB = = −67,5
8

120 kN
-67.5 kNm +67.5 kNm

A B
4.5 m

For equilibrium of joint B: MBA + MBC = 0

2EI BA F 2EI BC F Equation (2)

(2B + A) + M BA + (2B + C) + M BC = 0
 BA  BC

70 kN 70 kN
- kNm + kNm

B C
1.25 m 1.25 m 1.25 m

70(2 ,5)(1,25) 70(1,25)(2 ,5)

2 2

Where, M
F
=− −
(3,75) 2 (3,75) 2
BC

For equilibrium of joint C: MCB + MCD = 0

2EI CB F 2EI CD F
(2C + B) + M CB + (2C + D) + M CD = 0 Equation (3)
 CB  CD
 - kNm + kNm
200 kN total

D
3.75 m

−200(3,75)
where, M CD =
F

12

For equilibrium of joint D: MDC + MDE = 0

2EI DC F
(2D + C) + M DC - 60 = 0 Equation (4)
 DC

50 kN
-60 kNm

E
1.2 m
D

We now have four equations with four unknowns. After finding the values of these
unknowns, they must be re-substituted in the slope deflection equations to find the
support bending moments.

Example 6.4

In the portal frame shown, sway is prevented by a reaction at B. Draw the bending
moment and shear force diagrams and sketch the deflected shape. Determine the
force preventing sway. EI is constant.
 5m
B
-240 kNm
C C

5m 4m 4m

240 kN 240 kN

A
4m 4m

+240 kNm
D D

The solution to the above problem is carried out using steps (i) to (iii) below:

(i) First, determine the known and unknown displacement and rotations at the
supports.

A = D = 0

B = ?

C = ?

All ’s = 0

(ii) Then, determine moments using the slope deflection equations.

2 EI B
MAB = 2 EI 2EI (0 + B – 0) + 0 i.e. MAB =
5 5

4 EI B
MBA = 2 EI (2B + 0 – 0) + 0 i.e. MBA =
5 5

MBC = 2 EI (2B + C)

5

MCB = 2 EI (2C + B)

5

EI C
MCD = 2 EI (2C + 0 + 0) – 240 MCD = − 240
8 2
 EI C
MDC = 2 EI (0 + C + 0) + 240 MDC = + 240 + 240
8 4

MBA + MBC = 0:
4 EI B 2 EI
+ (2 + C) = 0 Equation (1)
5 5

MCB + MCD = 0:
2 EI 4 EI C
(2C + B) + – 240 = 0 Equation (2)
5 2

From Equation (1):

8EI B 2 EI C
+ =0 Equation (3)
5 5

From Equation (2):

13EI C 2 EI B
+ – 240 = 0 Equation (4)
10 5

Equation (2)  4:
8EI B 52 EI C
+ − 960 = 0 Equation (5)
5 10

48EI C
Equation (5) – Equation (3): + 960 = 0
10

200
giving C = –
EI

Substituting C in Equation (3) we get:

8EI B 2 EI  200 
+  
5 5  EI 

50
giving B = −
EI

Substituting back in the slope deflection expressions:

 2 EI  −50 
MAB =   MAB = –20 kNm
5  EI 

4 EI  −50 
MBA =   MBA = –40 kNm
5  EI 

2 EI  100 200 
MBC = − +  MBC = +40 kNm
5  EI EI 

2 EI  400 50 
MCB = −  −  MCB = +140 kNm
5  EI EI 

EI  200 
MCD =   -240 MCD = –140 kNm
2  EI 

EI  200 
MDC =   +240 MDC = +290 kNm
4  EI 

(iii) Draw the bending moment and shear force diagrams and determine support
reactions.

40 140
B C HC
HB 40

480
A + 240
HA

20

D HD
290
Free body diagram A-B Bending moment diagram [kNm] Free body diagram D-C

Free body diagram A-B Bending moment diagram [kNm] Free body diagram D-C
M = 0 MC = 0
5HB – 20 – 40 = 0 8HD – 240  4 – 290 + 140 = 0
HB = 12 kN giving HD = 138.75 kN
H = 0 H = 0
HA = 12 kN HC = 101,25 kN

H
B C

A 240 kN
12

20
VA D
138.75
VD 140

Frame as a whole:

H = 0: H + 138,75 + 101.25 = 240 + 12

 Force preventing sway = H = 133,25 kN

MA = 0: 5VD – 133,25  5 + 240  1 + 138,75  3 + 20 - 140 = 0

which gives VD = 36 kN

V = 0: VA = 36 kN

113,25 101,25
C
36
240

12

36
138.75

36

Shear force diagram (kN) Thrust diagram (kN)

 Deflected shape
° = points of contraflecture

6.8 Special cases

The slope deflection equations are repeated below:

2EI 3 F
MAB = (2A + B – )+ M AB Equation (7a)
L L

2EI 3 M BAF
MBA = (2B + A – )+ Equation (7b)
L L

F F
In each of the following three cases there is no lateral load, so that M AB = M BA = 0

(1) A pinned, B fixed. Moment applied at A.

MAB L
A B
A

B =  = 0

2EI
From Equation (7a) MAB = (2A + 0 – 0) + 0
L
 4EI A
MAB = Equation (8)
L

2EI
From Equation (7b) MBA = (0 + A - 0) + 0
L
2EI A
MBA = = 1/2 MAB Equation (9)
L

(2) A fixed, B fixed. B sinks  relative to A.

MAB MBA
A

L
B

A = 0 = B

From Equation (7a):

6EI
MAB = − Equation (10)
L2

From Equation (7b):

6EI
MBA = − Equation (11)
L2

(3 A pinned, B pinned. Moment applied at A.

B sinks  relative to A.

MAB
A
A 
B

MBA = 0 (pin end at right)

 2EI 3
From Equation (7b) MBA = (2B + A – )=0
L L
4 EI B 2 EI A 6EI
 =− + 2
L L L
From Equation (7a), using the above:
3EI 
MAB = (A – ) Equation (12)
L L

Similarly, for a pin end at the left:

3EI 
MBA = (B – ) Equation (13)
L L

6.9 Tutorial

(1) For the beam shown below, find the support moments and reactions, and draw
the shear force and bending moment diagrams. The relative flexural rigidities
(EI) for each span are indicated in the figure.

400 kN

30 kN/m

(8EI) (9EI) (4EI)

A 2m B C D

4m 6m 5m

(2) For the propped cantilever with the triangular loading shown below,
calculate the rotation B at B, the moment MA at the fixed support A, and the
vertical reactions, VA and VB, at supports A and B, respectively.
 MA kNm

w kN/m

A B

(3) Using slope deflection analysis, draw the shear force and bending moment
diagrams for the beam shown below. EI is constant.

30 kN/m

C
A B

7m 3m

(4) Draw the shear force and bending moment diagrams for the beam shown. EI is
constant.
12 kN/m

A B C D E
5m 8m 8m 3m
LEARNING UNIT 7:
MOMENT DISTRIBUTION METHOD OF ANALYSIS

Learning outcomes

After studying this learning unit, you should be able to

• analyse both structurally determinate and indeterminate beams and portal

frames using the moment distribution method

• apply the moment distribution method in analysing structures with different end-
support conditions, that is, pinned, fixed-support or cantilever end-support

• analyse frames which have a no-sway and sway action

• draw dimensioned shear force diagrams and bending moment diagrams for
beams and frames

• determine support reactions for any given loading on beams and frames

7.1 Introduction

In this learning unit, we will demonstrate the moment distribution method using
examples.

7.2 Hardy-Cross moment distribution method

Moment distribution (or the Hardy-Cross method or the method of successive

approximations) is a stiffness or equilibrium method of analysis fundamentally similar
to slope deflection. Like slope deflection, it deals with end conditions, but the
calculations are carried out in terms of end moments instead of end deformations, and
a repetitive procedure is used instead of simultaneous equations.

Each change in moment in the process corresponds to a readily visualised change in

deformation.
Again, the basic assumptions are that compared to the bending deflections of the
structure, those due to shear, direct stress and torsion are negligible, and that the
structure is elastic, that is, the simple theory of bending applies.

The notation and sign convention are the same as for the slope deflection method.

7.3 Basic design procedure – no sway

• In a beam: sway takes place as a vertical displacement of a support due to

settlement or jacking up.

• In a portal frame: sway takes place as a horizontal displacement of a joint due

to asymmetrical loading or due to an asymmetrical shape of the portal frame.

For a frame which does not sway, the basic design procedure is as follows:

(1) Clamp all joints against rotation.

(2) Apply the loads. Initial moments, termed fixed-end moments, will develop at the
ends of loaded members.

(3) Unclamp one of the joints. It will rotate until it is in equilibrium, and the end
moments of the members meeting at the joint will be modified. These changes
in end moments can be found easily, as will be shown later.

(4) Reclamp the joint and release another joint, and so on, until all the joints have
been released and reclamped.

(5) Repeat operations (3) and (4) until the changes in end moments are negligible.

The above steps can be illustrated as follows.

(a) One joint clamped

B
C Original loaded beam
A

B
C 1. Clamp joint B
A
 +12 (say) 2. Apply load
A B Fixed-end moments develop at
C
A and B. B-C not affected.
-12

3. Unclamp B
(-8) +12 -2
Unbalanced moment at B will
cause it to rotate anticlock-wise.
Changes in end moments will
-12 (-4) (-4) occur at A, B and C. The change
Figure 7.1 at B must be such that MBA + MBC
=0

(b) Two joints clamped

Original loaded beam

B C D
A

1. Joints B, C clamped, loads applied:

Fixed-end moments develop at A, B,
B C D
A C, D.

2. Joint B unclamped:
B will rotate and the end moments at
B C D A, B and C of A-B, B-C will be
A
changed.

B C D 3. Joint B reclamped, Joint C unclamped:

A

Figure 7.2
 C will rotate and the end moments at
B, C and D of B-C, C-D will be
changed
(b)

At stage 3 the deflected shape and end moments will approximate to that obtained if
the structure were loaded without restraint.

A closer approximation would be obtained if the cycle 2 and 3 were repeated.

7.4 Types of members and connections

Initially, we will only consider uniform members with either rigid or pinned connections
and fixed or free ends. We should regard the joints as something like gusset plates to
which the members are rigidly connected or pinned. The gusset plates, and not the
ends of members, are clamped and released.

Typical members may be

(1) (2)

Rigid Rigid Rigid Pinned

(3) (4)

Pinned Pinned Rigid Free

Figure 7.3
7.5 Initial moments when joints are clamped, and loads applied

MFBA
1. Rigid-rigid
A B
The end moments are the fixed-end
F
M AB F F
moments M AB , M BA

B
2. Rigid-pinned
A
2 EI
At pinned end: M B = 0 = (2 B ) + M FBA

At fixed end:
2 EI
MA = ( B ) + M AB
F

1 F
= M AB
F
− M BA
2

A BB
3. Pinned-pinned
Both end moments are zero.

A B

4. Rigid-free
MAB = Cantilever moment
Figure 7.4

7.6 Distribution and carry-over

The sketch below represents all types of members: [(1), (2a), (2b), (3) and (4)], and
connections [(A), (B), (C), (D), (E) and (F)].
 F
free
E

pinned
rigid (1)
pinned A B
pinned
rigid
rigid

D rigid

C
pinned

Figure 7.5

For equilibrium of joint (A):

MAB + MAC + MAD + MAE + MAF = 0

Suppose at some stage in the analysis all joints, including (A), are clamped, the above
condition is not satisfied, but that

MAB + MAC + MAD + MAE + MAF = M1

If joint (A) is now released the unbalanced moment M1 will cause it to rotate until
equilibrium is restored.

Let the changes in end moments at (A) be:

M 1AB , M 1AC ... M 1AF

1 1 1 1 1
Equation (1)
Then M AB + M AC + M AD + M AE + M AF = –M1
When joint A rotates through an angle A, the changes in the end moments will be:
At joint released (i.e. at A) At far end (carry-over moment)
(distributed moment.)
 2 EI   EI   EI 
(1) M AB =   2 A = 4   A M 1AB = 2   A
1
 L  AB  L  AB  L  AB

(2a)
 2 EI   2 EI 
M 1AC =   ( 2 A +  C ) M 1CA =   (  + 2 C ) = 0
 L  AC  L  AC A
1
 2 EI   1   3EI   C = − A
=   2 A −  A  =   A 2 2
 L  AC  2   L  AC

1
(2b) M AD = 0 M 1DA = 0
1
(3) M AE = 0 M 1EA = 0
1
(4) M AF = 0 M 1FA = 0

The changes in end moments at joints released are therefore proportional to:
 EI   EI 
4  , 3  ,0 ,0, 0.
 L  AB  L  AC

These are constants for each type of member and are termed the stiffness of the
member.

The stiffness of a member depends on the nature of the connections at both ends. If
these are not the same, the member has two stiffnesses, depending upon which joint
rotates.

The following table summarises the stiffness and relative stiffness for various types of
members.
 Table 7.6.1
Type of connection
Figure Relative stiffness k
7.5 At joint released At far joint Stiffness 4EI
(put =K)
L
(1) Rigid Rigid 4EI K
L
(2a) Rigid Pinned 3EI 3
K
L 4
(2b) Pinned Rigid 0 0
(3) Pinned Pinned 0 0
(4) Rigid Free 0 0

From Equation (1) and (2), and since change of end moment at the joint released is
proportional to stiffness:

M 1AB =
stiffness for AB
 stiffness
(
 −M1 )

M 1AC =
stiffness for AC
 stiffness
(
 −M1 )

M 1AD =
stiffness for AD
 stiffness
 −M1( )

M 1AE =
stiffness for AE
 stiffness
(
 −M1 )

M 1AF =
stiffness for AF
 stiffness
(
 −M1 )

Distribution factor

stiffness K
The ratios = = are called the distribution factors.
 stiffness K

Carry-over factors
The changes in end moments at the far joints are proportional to the changes at the
joints released. They are obtained by multiplying the latter by constants known as
carry-over factors.

1 1
M 1BA = M AB
2

M1CA = 0
M 1DA = 0
M 1EA = 0
M 1FA = 0

 There is only one carry-over factor = 1/2 for members of type 1. All the other
carry-over factors are zero.
Example 7.1

50 kN 40 kN
16 kN/m

(12EI) (3EI) (4EI)

A 4m B 3m C D

8m 6m 6m

Using the moment distribution method, calculate the support bending moments for the
beam shown above if (EI)AB : (EI)BC : (EI)CD = 12:3:4.

The solution to the above problem is carried out using steps (a) to (e) below:
Joint B Joint C

Member B-A B-C C-B C-D

  4 EI   4 EI   4 EI   3EI 
(a) Rotational        
 L  BA  L  BC  L  CB  L  CD
stiffness
4  12 43 43 3 4
= =2 = =2
= =6 = =2 6 6
8 6

6 2 2 2
= 0,75 = 0,25 = 0,5 = 0,5
(b) Distribution

6+2 6+2 2+2 2+2

factors

(c) Carry-over factors

Member A-B B-A B-C C-B C-D D-C

Carry-over 0 0.5 0.5 0.5 0 0

factors

(d) Initial moments (the joints are clamped and the loads applied)
MAB = WL 50  8 = –50 kNm
− =−
8 8
MBA = WL = +50 kNm
+
8
MBC = WL 40  6 = –30 kNm
− =−
8 8
MCB = WL = +30 kNm
+
8
MCD= = –72 kNm
WL 1  WL  − 96  6  1  96  6
− −  = − 
12 2  12  12  2  12

(e) Distribution table

MAB MBA MBC MCB MCD MDC
Distribution and carry- 1 1
/2 /2 0
over factors 0.75 0.25 0.5 0.5
Initial moments –50 +50 -30 +30 –72 0
B released Balance –15 –5
Carry-over –7.5 –2.5 0
 C released Balance +22.3 +22.3
Carry-over +11.2 0
B released Balance –8.4 –2.8
Carry-over –4.2 –1.4
C released Balance +0.7 +0.7
Carry-over +0.4 0
B released Balance –0.3 –0.1
Carry-over –0.2 –0.1
Balance 0 0
End moments -61.9 +26.3 –26.3 +49 –49 0

Alternative (standard) moment distribution procedure

(1) An alternative procedure (is shown below) where:
(a) All joints balanced
(b) All carry-overs recorded
in one operation, is faster and commonly used.

(2) Decide on the order of accuracy beforehand, and stick to it.

(3) For the alternative procedure draw a double line after each balance and end
on balance.
MAB MBA MBC MCB MCD MDC
1 1
Distribution and /2 /2 0
carry-over factors 0.75 0.25 0.5 0.5
Initial moments –50 +50 –30 +30 –72 0
Balance –15 –5 +21 +21
Carry-over –7.5 +10.5 –2.5 0
Balance –7.9 –2.6 +1.3 +1.3
Carry-over –4.0 +0.7 –1.3 0
Balance –0.5 –0.2 +0.7 +0.7
Carry-over –0.3 +0.4 –0.1 0
Balance –0.3 –0.1 +0.1 +0.1
 End moments – +26.3 –26.3 +49.2 – 0
61.8 48.9

The reactions can now be calculated in the usual way and the bending moment and
shear force diagrams drawn.

Example 7.2

Draw the shear force and bending moment diagrams for the beam shown below.
EIAB: EIBC: EI CD = 2:3:4

100 kN 80 kN 40 kN
30 kN/m

A B C D
2.5 m 2.5 m 6m 1.25 m 2.5 m 1.25 m

The solution to the above problem is carried out using steps (a) to (f) below:

Joint B Joint C

Member B-A B-C C-B C-D

 4 EI   4 EI   4 EI   3EI 
Rotational

       
stiffness

 L  BA  L  BC  L  CB  L CD
(a)

42 8 43 4 6 3  4 12
= =2 = =2 = =
= = 6 3 5 5
5 5
8 2 5 2 5 12
Distribu

= =
factors

5 =4 5 = 6
tion

2+8 2 + 12 5 11
(b)

9
2 + 85 9 5 2 + 5 11
12

(b) Initial moments

 -62.5 kNm 100 kN
+62.5 kNm
-90 kNm +90 kNm
30 kN/m

A B B C
2.5 m 2.5 m 6m

100  5 180  6
M BA = = +62.5 M CB = = +90
8 12
80 kN 40 kN
89.1 kNm

C D
1.25 m 2.5 m 1.25 m

(d) Moment distribution table

MAB MBA MBC MCB MCD MDC
4 5 5 6
Distribution factor /9 /9 /11 /11
1 1
Carry-over /2 /2 0
factors
Initial moments –62.5 +62.5 –90 +90 – 0
89.1
Balance +12.2 +15.3 –0.4 –0.5
Carry-over +6.1 –0.2 +7.7 0
Balance +0.1 +0.1 –3.5 –4.2
Carry-over +0.1 –1.8 +0.1 0
Balance +0.8 +1.0 0 –0.1
Carry-over +0.4 0 +0.5 0
Balance 0 0 –0.2 –0.3
Carry-over 0 –0.1 0 0
Balance 0 +0.1 0 0
 End moments –55.9 +75.6 –75.6 +94.2 – 0
94.2

(e) Support reactions

-55.9 kNm 100 kN 80 kN 40 kN

75.6 kNm 94.2 kNm
30 kN/m

A B C D

2.5 m 2.5 m 6m 1.25 m 2.5 m 1.25 m

-55.9 kNm 100 kN

75.6 kNm 75.6 kNm 94.2 kNm
30 kN/m

VA = 46.1 kN VB1 = 53.9 kN VB2 = 86.9 kN VC1 = 93.1 kN

2.5 m 2.5 m 6m

5VA + 75.6 – 100  2.5 – 55.9 = 0 6VB2 + 94.2 – 75.6 – 30  6  3 = 0

VA = 46.1 VB1 = 53.9 kN VB2 = 86.9 VC1 = 93.1 kN

80 kN 40 kN
94.2 kNm

VC2 = 88.8 kN VD = 31.2 kN

1.25 m 2.5 m 1.25 m

5VC2 – 94.2 – 40  1.25 – 80  3.75 = 0

VC2 = 88.8 VD = 31.2
(f) Shear force and bending moment diagrams

86.9
88,8

46,1
8,8
+ + +
-
x 6− x
31,2 =
53,9 86,9 931,
x
x = 2 ,9m
93.1
Shear force diagram (kN)

94,2
84,9
55,9 75,6

39

- -
+ + +
17

50,1 39
59,4 1
(30)(6) 2 = 135
8

Bending moment diagram (kNm)

7.7 Sinking of beam support

Beam A-B is fixed at both ends. Support B sinks an amount Δ. Find the moments
induced at A and B.

M
M

A

B

L
/6

- Bending moment diagram

M
+ M
L
/2

5L
/6

Figure 7.6

Mohr II

Intercept of carried tangents at A and B at B

M
= moment of diagram between A and B about B
EI

M  L   5L  M  L   L  4 ML2
− = −    +    = −
2 EI  2   6  2 EI  2   6  24 EI

6EI
M=
L2

If we were to revert back to the slope deflection and moment distribution sign
convention, the following sketch would be applicable.
  A
B

Figure 7.7

See also special case (2), section 2.7.

Example 7.3
A horizontal beam A-B-C-D, 400 mm deep, is continuous over three equal spans of
4 m each. End supports A and D are rigidly fixed, and all supports are on the same
level. I is constant = 200 x 106 mm4 and E = 200 GPa.

(i) If a load of 30 kN/m is uniformly distributed over spans A-B and B-C, determine
the maximum bending stress in the beam.

(ii) What will be the maximum bending stress at B, if B sinks 6 mm?

30 kN/m

A B C D

4m 4m 4m

The solution to the above problem is carried out using steps (a) to (i) below:

-40 kNm +40 kNm

A B
 (a) Distribution factors
EI
is constant,  distribution factors for joints B and C are 1/2 for both joints
L
(b) Carry-over factors
Member A-B B-A B-C C-B C-D D-C

Carry-over 0 0.5 0.5 0.5 0.5 0

factors

(c) Distribution table when supports remain level

MAB MBA MBC MCB MCD
MDC
1 1 1
Distribution and /2 /2 /2
carry-over factors 0. 5 0.5 0.5 0.5
Initial moments –40 +40 –40 +40 0 0
Balance –20 –20
Carry-over –10 –10
Balance +5 +5
Carry-over +2.5 +2.5
Balance –1.2 –1.2
Carry-over –0.6 –0.6
Balance +0.3 +0.3
Carry-over +0.1 +0.1
Balance 0 0
– +45.3 –45.3 +21.4 – –
37.4 21.2 10.6
 (e) Bending moment diagram

-37.4 kNm 45.3 kNm 21.3 kNm -10.6 kNm

A B C D

4m 4m 4m

-37.4 kNm 45.3 kNm

21.3 kNm

A B C
-10.6 kNm
4m 4m 4m

Mmax = 45.3 kNm

My 45,3  10 
6 400
Maximum stress  max = = 2 = 45,3N / mm 2
I 200  10 6

max = 45.3 MPa

(f) Support B sinks 6 mm

-90 -90 +90

+90

A 6mm
D
6mm
C
B

6EI
Moment induced in spans A-B and B-C =
L2
 6  200 106  200 106 10−12  6 10−3
= kNm
42
= 90kNm(AB and BA are negative and BC and CB positive)

(g) Moment distribution table

MAB MBA MBC MCB MCD
MDC
1 1 1
Distribution and /2 /2 /2
carry-over factors 0. 5 0.5 0.5 0.5
Initial moments –90 –90 +90 +90 0 0
Balance –45 –45
Carry-over –22.5 –
22.5
Balance +11.2 +11.2
Carry-over +5.6 +5.6
Balance –2.8 –2.8
Carry-over –1.4 –1.4
Balance +0.7 +0.7
Carry-over +0.3 +0.3
Balance –0.1 –0.1
Moments due to – –78.1 +78 +48 – –
sinking support B 84.1 47.9 23.9

(h) Final support moments

MAB = –37.4 – 84.1 = –121.5

MBA = +45.3 – 78.1 = –32.8
MBC = –45.3 + 78.1 = +32.8
MCB = +21.3 + 48 = +69.3
MCD = –21.3 – 48 = –69.3
MDC = –10.6 – 23.9 = –34.5
 (i) Bending moment diagram

69,3
34,5
30 kN/m
A B C D
121,5
32,8

121,5

44,3 69,3

18,3
60
60
32,8 34,5
41,7

Bending moment diagram (kNm)

The bending moment at B has reversed by support settlement (i.e. it is now a sagging
moment) and its magnitude is smaller.

B = 32.8 MPa

32.8
-

+
32.8

Note:

In example 7.3 two separate distributions were carried out. This can be combined, as
will be shown in example 7.4.
7.8 Cantilever at end
If a cantilever overhangs one support, the bending moment here is known. This
support can be treated as a free end. However, the moment at this support must
immediately be balanced to the known cantilever moment.

Example 7.4

A uniform beam A-B-C-D-E has all supports initially at the same level, but under
loading support C sinks 12 mm. I = 120 x 106 mm4 and E = 200 GPa. Find the bending
moments at A, B, C and D and draw the bending moment diagram.

120 kN 60 kN
30 kN/m
60 kN/m

A B C D E
6m 6m 2.25 m 2.25 m 1.5 m

The solution to the above problem is carried out using steps (a) to (h) below:

(a) Initial moments

-90 kNm 30 kN/m +90 kNm

A B
6 m

w 2 30(6 )
2
M BA = = = 90kNm
12 12

-180 kNm +180 kNm

60 kN/m

B C
6 m
 w 2 60(6 )
2
M CB = = = 180kNm
12 12

120 kN
-67.5 kNm +67.5 kNm

C D
4.5 m

WL 120  4,5
M DC = = = 67,5kNm
8 8

60 kN
-90 kNm

E
1.5 m
D

MD = 90 kNm

(b) Moment due to support C sinking by 12 mm

+85,3

12mm
-48 -48 +85,3

6𝐸𝐼∆ 6 × 200 × 106 × 120 × 106 × 10−12 × 12 × 10−3

𝑀𝐶𝐵 = − = − = −48 𝑘𝑁𝑚
𝑙2 62
 Joint B Joint C

Member B-A B-C C-B C-D

Rotational

4 EI 4 EI 4 EI
stiffness

3EI
(c)

6 6 6 4,5
(d) Distribution
factors

0.5 0.5 0.5 0.5

(e) Carry-over factors

Member A-B B-A B-C C-B C-D D-C

Carry-over
factors 0 0.5 0.5 0.5 0 0.5

(f) Moment distribution table

 MAB MBA MBC MCB MCD MDC MDE
Distribution 1 1 1
/2 /2 /2
and carry-
0.5 0.5 0.5 0.5
over factors
Loading M –90 +90 –180 +180 –67.5 +67.5 -90
Sinking M
–48 –48 +85.3 +85.3
Balance
+69 +69 –74.9 –74.9 –62.8 *
Carry-over +34.5 –37.5 +34.5 –31.4
Balance
+18.8 +18.8 –1.6 –1.6
Carry-over +9.4 –0.8 +9.4
Balance
+0.4 +0.4 –4.7 –4.7
Carry-over +0.2 –2.4 +0.2
Balance
+1.2 +1.2 –0.1 –0.1
Carry-over +0.6 0 +0.6
Balance
–0.3 –0.3
End -45.3 +179.4 -179.3 +95.1 -95.2 +90 -90
moments

* Joint D balanced
(g) Support reactions
-45.3 kNm 120 kN 60 kN
179.4
95.2 kNm 90 kNm
30 kN/m kNm 60 kN/m

A B C D E
6m 6m 2.25 m 2.25 m 1.5 m

VA = 67.7 kN VB = 112.3 + 194 = 306.3 kN VC = 166+60.5 = 226.5 kN VD = 59.5 + 60.5 = 119.5 kN

-45.3 +95.2
30 kN/m +179.4 kNm -179.4 kNm 60 kN/m
kNm kNm

VB1 VB2 VC1

VA
6 m 6 m

6VA + 179.4 = 30  6  3 + 45.3 6VB2 + 95 = 179.4 + 60  6  3

VA=67.7 kN VB1 = 112.3 kN VB2 = 194 kN VC1 = 166 kN
 120 kN
-95.2 60 kN
+90kNm
kNm -90 kNm

VC2 4.5 m VD1 VD2 1.5 m

5VC + 90 = 95.2 + 300

VC2 = 60.5 VD1 = 59.5

(h) Shear force and bending moment diagrams

194 kN

67.7 kN

60.5 kN 60 kN

E
A B C
59.5 kN
4m 4m 4m D

112.3 kN

166 kN

Shear force diagram

179.4 kNm 137.3

kNm
112.4 kNm 95.2 kNm 92.6 kNm 90 kNm
-45.3 kNm

135 kNm
135 kNm 270 kNm

A B C 42.4 kNm
22.6 kNm
4m 4m 132.7 kNm 4m D

Bending moment diagram

7.9 Simple framed structures

A simple framed structure is one in which sway does not occur.

Example 7.5

The figure shows a three-span continuous beam, fixed at A and monolithic with
columns at B and C. Find the final bending moments at A, B, C, E and F, the reaction
at D and sketch the bending moment diagram for the loaded frame.

60 kN 60 kN 80 kN

A = =
= B 30 kN/m C
D

3m 4.5 m 6m 1.5 m 1.5 m

4m

ICD =ICF = 1 IAB = IBE = 2 IBC = 3

The solution to the above problem is carried out using steps (a) to (e) below:

(a) Fixed-end moments

60 kN 60 kN
-60 kNm +60 kNm

A B
1.5 m 1.5 m 1.5 m

60(1,5)(3) 60(3)(1,5)
2 2
M AB = − −
(4,5)2 (4,5)2
= −40 − 20
= −60kNm
 -90 kNm +90 kNm
30 kN/m

B 6 m C

30(6 )
2
M BC = − = −90kNm
12

80 kN
-45 kNm

3 m
C D

80  3 1  80  3 
M CD = − −  
8 2 8 
= −30 − 15
= −45kNm

(b) Distribution factors

Joint B Relative stiffness k Total relative k

stiffness k
2I 8I Put the value, it is missing
B-A =
4,5 18
2I 12I 29 I 12
B-E = k = = 0,414  =1
3 18 18 29
3I 9I 9
B-C = = 0,310
6 18 29
 Joint C Relative stiffness k Total relative k
stiffness k
3I 6I 6
C-B = = 0,5
6 12 12
I 3I 12 I 3
C-F = k = = 0,25
4 12 12 12  =1
I  3  3I 3
= 0,25
C-D  =
3  4  12 12

(c) Moment distribution table

A-B B-A B-E B-C C-B C-F C-D D- E-B F-C

C
Distribution and ½ ½ 0
carry-over 0.276 0.414 0.310 0.5 0.25 0.25
factors
Initial moments –60 +60 0 –90 +90 0 –45 0 0 0
Balance +8.3 +12.4 +9.3 –22.5 –11.3 –11.3
Carry-over +4.2 –11.3 +4.7 +6.2 –5.7
Balance +3.1 +4.7 +3.5 –2.4 –1.2 –1.2
Carry-over +1.6 –1.2 +1.8 +2.4 –0.6
Balance +0.3 +0.3 +0.4 –0.9 –0.5 –0.5
Carry-over +0.1 –0.4 +0.2 +0.2 –0.2
Balance +0.1 +0.2 +0.1 –0.1 0 0
–54.1 +71.8 +17.8 –89.2 +70.8 –13.0 –58.0 +8.8 –6.5

(d) Reaction at D:
80 kN
-58 kNm
D

3 m
C VD

3VD + 58 = 80  1.5 VD = 20.67 kN

(e) Bending moment diagram

 71.8 kNm 70.8 kNm
54.1 kNm 58 kNm
A = =
= B C
D
17.8 kNm
13 kNm
54.8 kNm 31 kNm
24.1 kNm
30 kNm
3m 4.5 m 6m 1.5 m 1.5 m
4m
8.8 kNm
E

6.5 kNm
F

Example 7.6

In the portal frame shown, sway is prevented by a reaction at B. Draw the bending
moment and shear force diagrams and sketch the deflected shape. Determine the
force preventing sway. EI = constant.

[Note: This example has already been carried out using slope deflection analysis –
see example 6.4 in learning unit 6.]

5m
B
-240 kNm
C C

5m 4m 4m

240 kN 240 kN

A
4m 4m

+240 kNm
D D

The solution to this example is given in steps (a) to (e) below.

 Joint B Joint C

Member B-A B-C C-B C-D

(a) Rotational
4 EI 4 EI 4 EI 4 EI

stiffness
(b) Distribution 5 5 5 8

0.5 0.5 0,8 0,5

factors

0,8 + 0,5 0,8 + 0,5

= 0,615
= 0,385

(c) Carry-over factors

Member B-A B-C C-B C-D
Carry-over factors 0.5 0.5 0.5 0.5

(d) Distribution table

 A-B B-A B-C C-B C-D D-C
1 1 1
Factors /2 /2 /2
0.5 0.5 0.615 0.385
Initial moment 0 0 0 0 –240 +240
Balance +147.7 +92.3
Carry-over 0 +73.9 0 +46.2
Balance –37.0 –37.0 0 0
Carry-over –18.5 –18.5 0
Balance 0 0 +11.4 +7.1
Carry-over 0 +5.7 +3.6
Balance –2.9 –2.9 0 0
Carry-over –1.5 0 –1.5 0
Balance 0 0 +0.9 +0.6
Carry-over 0 +0.5 0 +0.3
Balance –0.3 –0.3 0 0
End moments –20 –40.2 +39.9 +140 –140 +290.1

5m
40 kNm B 140 kNm
C

5m 4m

240 kN
-20 kNm
A
4m

+290 kNm
D

(e) See the bending moment and shear force diagrams of example 6.4.

Example 7.7

A concrete box culvert structure is shown in the figure below. One metre length is
considered and the loading is as shown in the figure. Draw the bending moment
diagram.
 20 kN
2m 2m
A B
(1.5 EI)

(EI)
3m (EI)

10 kN 10 kN

1m
(1.5 EI) C
D
Upward soil pressure = 20/4= 5 kN/m

The solution to the above problem is carried out using steps (a) to (f) below:
Joint A
(Note that joints B, C and D are
similar)
2 Member A-B A-D
(a) Rotational

4(1,5EI ) 4 EI
stiffness

3
4

6 4
(b) Distribution

4 3
6 +4 6 +4
factors

4 3 4 3
9 8
= = 0,53 = = 0,47
17 17

(c) Carry-over factors

Member D-A A-D A-B B-A B-C C-B C-D D-C

Carry-over
factors 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
(d) Fixed-end moments

20 kN
-10 kNm +10 kNm

A B
4m

W 20  4
M AB = = = 10kNm
8 8

+6.67 kNm -6.67 kNm

5 kN/m
C
D 4 m

W 2 54(4 )
2
M DC = = = 6,67kNm
12 12

B -2.22 kNm

2m

10 kN

1m

+4.44 kNm
C

10(1) (2 )
2
10(1)(2)2
M BC = = − 2 . 22kNm ; 𝑀𝐶𝐵 = = +4.44𝑘𝑁𝑚
32 32
(e) Distribution of moments using normal tabular form
D-A A-D A-B B-A B-C C-B C-D D-C
1 1 1 1
Distr. and /2 /2 /2 /2
c.o. factors 0.47 0.47 0.53 0.53 0.47 0.47 0.53 0.53
FEMs –4.4 +2.2 –10.0 +10.0 –2.2 +4.4 –6.7 +6.7
Balance –1.1 +3.7 +4.1 –4.1 –3.7 +1.1 +1.2 –1.2
Carry-over +1.8 –0.5 –2.0 +2.0 +0.5 –1.8 –0.6 +0.6
Balance –1.3 +1.3 +1.3 –1.3 –1.3 +1.3 +1.3 –1.3
Carry-over +0.6 –0.6 –0.6 +0.6 +0.6 –0.6 –0.6 +0.6
Balance –0.6 +0.6 +0.6 –0.6 –0.6 +0.6 +0.6 –0.6

Final –5.0 +6.7 –6.6 +6.6 –6.7 +5.0 –4.8 +4.8

(f) Bending moment diagram

20 kN
2m 2m
6.7 kNm 6.7 kNm
A B
(1.5 EI)
(EI)

3m
(EI)

10 kN 10 kN
1m
(1.5 EI)
D C
5 kNm
5 kNm
End moments
 6.7 kNm 6.7 kNm
A B
20 kNm
(1.5 EI)

(EI)
3m
(EI)

6.7 kNm
6.7 kNm

1m
(1.5 EI)
D 10 kNm C
5 kNm 5 kNm

Bending moment diagram for the box girder.

Example 7.8

Find the unknown moments in the double storey frame shown and construct the
bending moment diagram. Find the reactions at the fixed supports A and F.

C 15 kN/m
D
(2EI)

3m (EI) (EI)

30 kN/m
B E
(4EI)

3m (EI) (EI)

A F
6m
The solution to the above problem is carried out using steps (a) to (f) below:

Joint C Joint B

Member C-D C-B B-C B-E B-A

4(2 EI ) 4 EI 4 EI 4(4 EI ) 4 EI
(a) Rotational

6 3 3 6 3
stiffness

8 EI 8 EI 8 EI 16EI 8 EI
= = = = =
6 6 6 6 6
(b) Distribution

8 8 8 16 8
8 + 16 + 8 8 + 16 + 8 8 + 16 + 8
factors

8+8 8+8
= 0,25 = 0,5 = 0,25
= 0,5 = 0,5

(c) Fixed-end moments

-90 kNm +90 kNm
30 kN/m

B E
6m

30  6 2
M EB = = +90
12

-45 kNm +45 kNm

15 kN/m

C 6m D

15  6 2
M DC = = +45
12
 (c) Moment distribution table
1
2

A-B B-A B-E B-C C-B C-D D-C D-E E-D E-B E-F F-E
Distribution
and carry-
0.25 0.5 0.25 0.5 0.5 0.5 0.5 0.25 0.5 0.25
over factors
FEMs 0 0 –90 0 0 –45 +45 0 0 +90 0 0
Balance
+22.5 +45 +22.5 +22.5 +22.5 –22.5 –22.5 –22.5 –45 –22.5
Carry-over +11.2 –22.5 +11.2 +11.2 –11.2 +11.2 –11.2 –11.2 +22.5 –11.2
Balance
+2.8 +5.7 +2.8 0 0 0 0 –2.8 –5.7 –2.8
Carry-over +1.4 –2.8 0 +1.4 –1.4 +2.8 –1.4
Balance
+0.7 +1.4 +0.7 –0.7 –0.7 +0.7 +0.7 –0.7 –1.4 –0.7
Carry-over +0.3 –0.7 –0.3 +0.3 +0.3 –0.3 –0.3 +0.3 +0.7 –0.3
Balance
+0.2 +0.5 +0.2 –0.3 –0.3 +0.3 +0.3 –0.2 –0.5 –0.2
Carry-over +0.1 –0.2 –0.1 +0.1 –0.1 +0.1 +0.2 –0.1
Balance
+0.1 +0.1 +0.1 –0.1 –0.1 –0.1
End moments +13 +26.3 –63.5 +37.1 +34.5 –34.4 +34.4 –34.5 –37.1 +63.5 –26.3 –13
(e) Bending moment diagram
1
34,5 /8 (15)(6)2 = 67,5
34,5
C D 34,5
+

33 37,1
37,1 63,5 37,1
63.5 63,5
B E 26,3
26,3
26,3 +

F H 71,5
H
1 2
/8 (3)(6) = 135
A 13
13

Bending moment diagram (kNm)

(f) Support reactions

B

+26.3 kNm

3m

HA +13 kNm
A

MB = 0: 3HA = 13 + 26.3

HA = 13.1 kN

HA = Reactions at A and F = 13.1 kN in direction as drawn

7.10 Frames with sidesway

The effect of sway can be taken into account by introducing a new type of operation:
the sway correction.

The joints are kept clamped against rotation, but the frame is free to sway until
equilibrium is reached.

Thus, the effect of sway is only found in the stanchions.

Δ Δ

B C

A D

Figure 7.8
7.11 Possible types of connections in sway consideration

Figure 7.9

Consider the frame shown which has every possible type of stanchion. The
condition for sway equilibrium is H = 0, that is,

SAB + SCD + SEF + SGH + H = 0 Condition (A)

Consider stanchion A-B

B
MBA
MB = 0
 SAB  AB + H (  AB – a) = MBA + MAB
a
M BA + M AB H(  AB − a )
that is, SAB = −
 AB  AB
SAB MAB

If we consider all the stanchions in a similar fashion, condition (A) becomes:

M BA + M AB H(  AB − a) M CD + M DC M + M FE M GH + M HG
− + + EF + +H=0
 AB  AB  CD  EF  GH

 AB 
That is, M BA + M AB + ( M DC ) + AB ( M EF ) − H( AB − a ) = 0
 CD  EF Equation (1)
(Note: MFE = MCD = MGH = MHG = 0)

Suppose that at some stage when the joints are clamped the end moments are
mAB, mBA …
And that Equation (1) above is not satisfied, but:
 AB 
m BA + mAB + ( mDC ) + AB ( mEF ) − H( AB − a) = m
 CD  EF

Now suppose the joints are kept clamped against rotation, but the frame is allowed to
sway until equilibrium is restored. Then, if the end moment changes due to sway are:

m1AB , m1BA …:

m1AB + m1BA +
 AB 1
 CD
( 
) ( )
m DC + AB m1EF = − m
 EF
If the frame swayed a distance  when sway correction is applied, the changes
in end moments would be:
 3  6EI AB
− 
2 EI AB
m1AB = 2 A +  B −   that is, mAB =
1

 AB  2AB
similarly 6EI AB
m1BA =
 2AB
− 

2 EI CD  3  m1CD = 0
m1CD = 2 C − =0
 CD   CD 
+3
i. e  C =
 CD

 3 
m1DC =
2 EI DC
2 C − =0 that is, m DC =
1 3EI DC
− 
 DC   CD   2DC
3EI EF
m1EF =
 2EF
− 

m1FE = 0
m1GH = 0
m1HG = 0
Sway stiffness

Hence sway stiffness of a member is the end moment developed when one end is
given a unit negative displacement relative to the other, the joints at both ends being
clamped against rotation, as illustrated in the sketch below:
 A A

B -1 B -1

Figure 7.10

From Equation (2), it follows that sway correction moments are proportional to sway
stiffness, that is:

sway stiffness of AB
m1AB =  ( M)
 stiffness
1
mCD = _______________etc.

sway stiffness
where, = SWAY FACTOR
 swaystiffness
7.12 Sway correction procedure

We will use the unit (arbitrary) displacement method of sway correction.

Step 1: Prevent the frame from swaying by providing a propping force and analyse
the frame to find the end moments (the no-sway analysis).

Step 2: Find the propping force which has prevented sway by applying the
condition for lateral equilibrium H = 0.

Step 3: Allow the frame to sway by applying stanchion end moments proportional to
sway stiffness. Analyse the frame to find all the end moments.

Step 4: From the condition H = 0, find the force causing the arbitrary sway.

Step 5: Combine the end moments from Steps 1 and 3 in such a proportion so as to
eliminate the propping force which was provided in Step 1.

P Q corrected to -P

= +

Initial frame No -sway analysis Apply arbitrary end moments

Calculate P Proportional to sway stiffness
Calculate Q and correct to -P

Figure 7.11
Example 7.9

We are required to sketch the shear force and bending moment diagrams for the frame shown
below.

60 kN
3m
B
C
EIBC = 2EI
1.5 m

30 kN EIAB = EICD = EI

3m

A D

6m

The solution to the above problem is carried out using steps (a) to (j) below:

Joint B Joint C

Member B-A B-C C-B C-D

3EI 4  2 EI 4  2 EI 4 EI
Rotational

6 6
stiffness

4,5 4,5
(a)

3 8 8 4
4,5 6 6 4,5
3 +8 8 +4
(b) Distribution

3 +8 4,5 6 8 +4
4,5 6 6 4,5 6 4,5
factors

= 0,4
= 0,67 = 0.6
= 0,33
(c) Sway stiffness
Member B-A A-B C-D D-C

3EI 0 6 EI 6EI
2 2
2
6 EI 6EI
3EI = =
=
(4,5) 2
(4,5)2 (4,5) 2
Ratio of sway 1:0:2:2
stiffness
(d) Initial member end moments

+25 kNm
B

1.5 m

30 kN

3m

+ 30(3) (1,5) 1  30(1,5) (3)

2 2
M BA = +   = 20 + 5
(4,5)2 2  (4,5)2 
= 25kNm

60 kN
-45 kNm +45 kNm

B C
6m

60(6 )
M CB = + = +45
8
(e) No-sway analysis
MAB MBA MBC MCB MCD MDC
1 1
Distribution and 0 /2 /2
carry-over factors 0.33 0.67 0.6 0.4
Initial moments 0 +25 –45 +45 0 0
Balance +7 +13 –27 –18
Carry-over 0 –14 +6 –9
Balance +5 +9 –4 –2
Carry-over 0 –2 +4 –1
Balance +1 +1 –2 –2
End moments 0 +38 –38 +22 –22 –10
(No-sway)

(f) Find propping force which has prevented sway

+38 kNm
B -22 kNm
C
1.5 m

4.5 m
3m

A SAB -10 kNm

D SDC

MB = 0 MC = 0
SAB = 30  1.5 – 38 4.5SDC = 22 + 10
SAB = 1.6 kN SDC = 7.1 kN

We consider the frame as a whole and solve for the propping force (P) kN as shown
in the following figure:
 60 kN

C
B P kN

1.5 m

30 kN
4.5 m

3m

SDC = 7.1 kN
A SAB = 1.6 kN D

H = 0: P + 7.1 + 1.6 = 30

giving P = 21.3 kN
(g) Arbitrary sway analysis

Apply arbitrary end moments to stanchions in proportion to sway stiffness.

MAB MBA MBC MCB MCD MDC

Distribution and 1 1
0 /2 /2
carry-over 0.33 0.67 0.6 0.4
factors
Initial moments 0 +20 0 0 +40 +40
Balance –7 –13 –24 –16
Carry-over 0 –12 –6 –8
Balance +4 +8 +4 +2
Carry-over 0 +2 +4 +1
Balance –1 –1 –2 –2
Carry-over 0 –1 0 –1
Balance 0 0 0 0
End moments 0 +16 –17 –24 +24 +32
(Arbitrary sway)

(h) Find force causing arbitrary sway

+16 kNm +24 kNm

B C

4.5 m 4.5 m

A SAB +32 kNm

D SDC

MB = 0 MC = 0
4.5 SAB = 16 4.5 SDC = 24 + 32
SAB = 3.6 kN SDC = 12.4 kN
Consider the frame as a whole and solve for the arbitrary sway force (Q) kN as shown
in the following figure:
B C
Q kN

4.5 m

SAB = 3.6 kN SDC = 12.4 kN

A D

H = 0: Q = 3.6 + 12.4
Q = 16 kN

(i) Sway correction

Propping force to prevent sway P = 21.3 kN (←)

Force causing arbitrary sway Q = 16 kN (←)
that is, force Q must be corrected to 21.3 kN (→)

MAB MBA MBC MCB MCD MDC

No-sway 0 +38 –38 +22 –22 –10
21.3 kN (←)
Arb sway 0 +16 –16 –24 +24 +32
16 kN (←)
Correction: 0 –21 +21 +32 –32 –43
− 21,3
16
correction factor = –1.3

Final moments 0 +17 –17 +54 –54 –53

Note:
Choose the initial moments for the arbitrary sway analysis of the same order of
magnitude as the final moments from the no-sway analysis.

60 kN

17 kNm 54 kNm
C
B

1.5 m

30 kN
4.5 m

3m

-53 kNm
SDC
A SAB D

+17 kNm
-54 kNm
B C

1.5 m

30 kN

4.5 m
3m

SAB
-53 kNm SDC
A
D

MB = 0 MC = 0
4.5 SAB = 30 x 1.5 – 17 4.5 SDC = 54 + 53
SAB = 6.2 kN SDC = 23.8 kN
 60 kN

-17 kNm +54 kNm

B C

MC = 0 6 SBC + 54 – 17 - 60  3 = 0 SBC = 23.8 kN

V = 0 SCB = 60 – 23.8 = 36.2 kN

(j) Shear force and bending moment diagram

23,8

36,2

Shear force diagram (kN)

54
35,5
17

54,5

Bending moment diagram (kNm)

Example 7.10

Draw the bending moment and shear force diagrams for the frame shown in the figure.
EI is constant.

30 kN/m
B
C E

6m

A D

4.5 m 2.3 m

The solution to the above problem is carried out using steps (a) to (j) below:

(a) Sway stiffness

Member A-B B-A C-D D-C

3EI 3EI 0
0 36 36
Ratio of sway 0:1:1:0
stiffness

Joint B Joint C

Member B-A B-C C-B C-D C-E

3EI 4EI 4EI 3EI 0
Rotational
stiffness

6 4 .5 4 .5 6
(b)

1 8 8 1 0
2 9 9 2
 Joint B Joint C

Member B-A B-C C-B C-D C-E

1 8 8 1 0
2 9 9 2
(c) Distribution factors

1 +8 1 +8 1 +8 1 +8
2 9 2 9 2 9 2 9

9 16 16 9
= 0,36 = 0,64 = 0,64 = 0,36
25 25 25 25

(d) Initial member end moments

-50.6 kNm +50.6 kNm

30 kN/m

B C
4.5 m

30  4.5 2
M CB = = +50.6
12

-79.4 kNm 30 kN/m

E
2.3 m
C

M CE = −30  2.3  1.15 = −79.4

(e) No-sway analysis

MAB MBA MBC MCB MCD MCE MDC
Distribution and 1
0 /2
carry-over factors 0 0.36 0.64 0.64 0.36 0
Initial moment 0 0 –50.6 +50.6 0 –79.4 0
Balance
 +18.2 +32.4 +18.4 +10.3 0
Carry-over 0 +9.2 +16.2 0
Balance –3.3 –5.9 –10.3 –5.9 0
Carry-over 0 –5.1 –2.9 0
Balance +1.8 +3.3 +1.9 +1.0
Carry-over 0 +1.0 +1.7 0
Balance –0.4 –0.6 –1.1 –0.6
Carry-over 0 –0.5 –0.3 0
Balance +0.2 +0.3 +0.2 +0.1
End moments 0 +16.5 –16.5 +74.6 +4.9 –79.4 0
(No-sway)

(f) Find propping force which has prevented sway

+16.5 kNm +4.9 kNm

B C

6m 6m

A SAB D SDC

MB = 0 MC = 0
6SAB = 16.5 6SDC = 4.9
SAB = 2.8 kN SDC = 0.8 kN

Consider the frame as a whole and solve for the propping force (P) kN as shown in
the following figure:
 30 kN/m

P kN B C

6m

SDC = 0.8 kN
A SAB = 2.8 kN D

H = 0
P = 2.8 + 0.8 = 3.6 kN

(e) Arbitrary sway analysis

Apply arbitrary end moments to stanchions in proportion to sway stiffness.

MAB MBA MBC MCB MCD MCE MDC

Distribution and 1
0 /2
carry-over factors 0 0.36 0.64 0.64 0.36 0
Initial moment 0 +10 +10 0
Balance –3.6 –6.4 –6.4 –3.6 0
Carry-over 0 –3.2 –3.2 0
Balance +1.1 +2.1 +2.1 +1.1 0
Carry-over 0 +1.0 +1.0
Balance –0.4 –0.6 –0.6 –0.4 0
Carry-over 0 –0.3 –0.3 0
Balance +0.1 +0.2 +0.2 +0.1
End moments 0 +7.2 –7.2 –7.2 +7.2 0
(h) Find force causing arbitrary sway

+7.2 kNm +7.2 kNm

B C

6m 6m

A SAB D SDC

MB = 0 6SAB = 7.2 MC = 0 SDC = 1.2 kN

SAB = 1.2 kN

Consider the frame as a whole and solve for the arbitrary sway force (Q) kN as shown
in the following figure:
Q kN
B C

6m

SDC = 1.2 kN
A SAB = 1.2 kN D

H = 0
Q = 1.2 + 1.2 = 2.4 kN

(i) Sway correction

 MBA MBC MCB MCD
No-sway P = 3.6 +16.5 –16.5 +74.6 +4.9 (1)
Arbitrary sway Q =2.4 +7.2 –7.2 –7.2 +7.2
−3,6 –10.7 +10.7 +10.7 –10.7 (2)
Correction
2,4
Final M (1) + (2) +5.8 –5.8 +85.3 –5.8

85.3 kNm
5.8 kNm 79.4 kNm

B C

5.8 kNm

6m

SDC
A SAB D

-5.8 kNm
+5.8 kNm
C
B

6m

SDC
SAB D
A

6SAB = 5.8 6SCD = 5.8

SAB = 1 kN SCD = 1 kN
 +5.8 kNm 85.3 kNm
79.4 kNm

SBC SCB SCE

4.5 m 2.3 m

MC = 0 4.5 SBC + 85.3 – 5.8 – 30  4.5  2.25 = 0 SCE = 30  2.3 = 69 kN

giving SBC = 49.9 kN

V = 0 SCB = 85.1 kN

(j) Bending moment and shear force diagrams

1
/8 (30)(4,5)2 = 75,9 69

49,9
85,3

5,8 79,4
5,8 5,8

85,1

1 1

Bending moment diagram (kNm) Shear force diagram (kN)

Example 7.11
B
20 kN C

EIAB = EIBC = 2EI

3m
EICD = EI

A D

5m

Draw the bending moment diagram for the frame shown above and calculate the
horizontal and vertical reactions at A and D.

(Note: In this problem there are no loads producing fixed-end moments, hence only
the arbitrary sway analysis and the sway correction need to be carried out.)

The solution to the above problem is carried out using steps (a) to (g) below:

Joint B Joint C

Member B-A B-C C-B C-D

4( 2 EI) 4(2 EI) 4(2 EI) 3EI
(a) Rotational

5 3
stiffness

3 5
8EI
8EI 8EI = = EI
= = 5
3 5

Joint B Joint C

Member B-A B-C C-B C-D

 8 8 8 5
(b) Distribution factors 3 5 5 5
8 +8 8 +8 8 +5 8 +5
3 5 3 5 3 5 5 5

= 0,625 = 0,375 =
8 5
13 =
13
= 0,6154
= 0,3846

(c) Sway stiffness

Member A-B B-A C-D D-C

6  ( 2 EI) 6  ( 2 EI) 3( EI) 0

(3) 2
(3) 2
(3) 2

12 EI
= 12 EI 3EI
9 = =
9 9
Ratio of sway 12:12:3:0
stiffness Which can be simplified further to:
4:4:1:0
(d) Arbitrary sway analysis

MAB MBA MBC MCB MCD MDC

1 1
Distribution and /2 0.625 0.375 /2 0.6154 0.3846 0
carry-over factors
Initial moments +40 +40 +10 0
Balance –25 –15 –6.2 –3.8
Carry-over –12.5 –3.1 –7.5 0
Balance +1.9 +1.2 +4.6 +2.9
Carry-over +1.0 +2.3 +0.6 0
Balance –1.4 –0.9 –0.4 –0.2
Carry-over –0.7 –0.2 –0.4 0
Balance +0.1 +0.1 +0.2 +0.2
Balance +27.8 +15.6 –15.6 –9.1 +9.1 0

(e) Find force causing arbitrary sway

-9.1 kNm
+15.6 kNm C
B

3m
3m

+27.8 kNm
D SDC
A SAB

MB = 0 MC = 0
3SAB + 27.8 + 15.6 = 0 3SDC + 9.1 = 0
SAB = –14.5 kN SCD = –3.0 kN

Consider the frame as a whole and solve for the arbitrary sway force (Q) kN as shown
in the following figure:
Q kN B C

3m

SAB = 14.5 kN SDC = 3 kN

A D

H = 0 Q = 17.5 kN

(f) Sway correction

MAB MBA MBC MCB MCD MDC

Arbitrary sway Q = 17.5 +27.8 +15.6 –15.6 –9.1 +9.1 0

−20 –31.8 –17.8 +17.8 +10.4 –10.4 0

Correction Qx
17,5

B 10.4 kNm
20 kN
C
17.8 kNm
5.8 kNm

3m

SAB = 16.5 kN SDC = 3.5 kN

A D

VA VB

MD = 0 5VA + 31.8 = 20  3 VA = 5.6 kN

 V = 0 VB = 5.6 kN

+10.4 kNm
-17.8 kNm C
B

3m
3m

-31.8 kNm
D SDC
A SAB

3SAB = 31.8 + 17.8 3SDC = 10.4

SAB = 16.5 kN SDC = 3.5 kN

(g) Bending moment diagram

+10.4 kNm

-17.8 kNm
B C -10.4 kNm

+17.8 kNm

-31.8 kNm D
A

Bending moment diagram

7.13 Double-bay portal frames

Double-bay portal frames are solved in the same manner as single-bay frames.

Example 7.12

Draw the bending moment diagram for the frame shown below. The relative values of
EI are given.
150 kN 200 kN
3m 3m 3m

B C
100 kN D

EIAB = 3EI EICF = 4EI

EIBC = EICD = 2EI EIDE = 2EI 6m

9m

A
F
9m 9m

The solution to the above problem is carried out using steps (a) to (j) below:

Joint B Joint C Joint D

Member B-A B-C C-B C-F C-D D-C D-E

4(3EI ) 4( 2 EI ) 4( 2 EI ) 4( 4 EI ) 4( 2 EI ) 4( 2 EI ) 4(2 EI )
(a) Rotational

9 9 9 9 9 6
stiffness

9
12EI 8 EI 8 EI 16 EI 8 EI 8 EI
= = = = = 8 EI =
9 9 9 9 9 = 6
9
 8 16 8
12 8 9 9 9 8 12
(b) Distribution 9 9 8 + 16 + 8 8 + 16 + 8
8 + 16 + 8 9 9
12 + 8 12 + 8 9 8 +8 12 + 8
8 9 9 9 9 9 = 0,25 9 6 8 9
factors

= 0,25 = 0,5
= 0,6 = 0,4 = 0,4 = 0,6

(c) Sway stiffness

Member B-A A-B C-F F-C D-E E-D

6( 3EI ) 6(3EI) 6( 4 EI ) 6(4 EI) 6( 2 EI ) 6( 2 EI )

2
9 2 2 9 2
6 2
9 9 62
18EI
= 24 EI 27 EI
18EI 24 EI = = 27 EI
81 = = 81 81 =
81 81 81
Ratio of sway 18:18 24:24 27:27
stiffness

(d) Fixed-end moments

150 kN 200 kN
-333 kNm +367 kNm

C D

3m 3m 3m

−150( 3)( 6) 200( 6)( 3)

2 2

M F
CD = −
92 92
= −333 kNm

𝐹
200(6)2 (3) 150(6)(3)2
𝑀𝐷𝐶 = −
92 92
= 367 𝑘𝑁𝑚
(e) No-sway analysis
1
/2
MAB MBA MBC MCB MCF MCD MDC MDE MED MFC
Distribution 1/ 1/ 1/ 1/
2 2 2 2

and carry- 0.6 0.4 0.25 0.5 0.25 0.4 0.6

over actors
FEM – +367
Balance +83 +167 333 –147 –
+83 220
Carry-over +42 –74 +42 – +84
Balance –25 –17 +19 +37 +19 –17 –25 110
Carry-over -12 +9 –9 –8 +9 –12 +18
Balance –5 –4 +4 +9 +4 –4 –5
Carry-over -3 +2 –2 –2 +2 –3 +4
Balance –1 –1 +1 +1 +1 –1 –1
End -15 –31 +31 +96 +214 – +251 – – +106
moments 310 250 125

(f) Find force preventing sway

-31 kNm +214 kNm
B C

-250 kNm
D
9m 9m

6m

SED
A SAB F SFC
E
-15 kNm +106 kNm -125 kNm

MB = 0 : MC = 0 : MD = 0

9 SAB = 15 +31 9 SFC = –106 – 214 6 SED = 125 + 250
SAB = 5.1 SFC = –35.6 SED = 62.5

Consider the frame as a whole and solve for the propping force (P) kN as shown in
the following figure:
 150 kN 200 kN
3m 3m 3m

B C P
100 kN
D

6m

9m

SED = 62.5 kN
E
SAB = 5.1 kN SFC = 35.6 kN
A F

9m 9m

H = 0 Propping force P = 100 + 35.6 – 5.1 – 62.5

= 68 kN

Sway frame to right with force of 68 kN

Note. Sway moments are proportional to sway factors.

(g) Arbitrary sway analysis 1/

2

MAB MBA MBC MCB MCF MCD MDC MDE MED MFC
Distribution and 1/ 1/ 1/ 1/
2 2 2 2

carry-over 0.6 0.4 0.25 0.5 0.25 0.4 0.6

factors
Sway moments –60 –60 –80 –90 –90 –80
Balance +36 +24 +20 +40 +20 +36 +54
Carry-over +18 +10 +12 +18 +10 +27 +20
Balance –6 –4 –8 –15 –8 –4 –6
Carry-over –3 –4 –2 –2 –4 –3 –8
Balance +2 +2 +1 +2 +1 +2 +2
Carry-over +1 0 +1 +1 0 +1 +1
Balance 0 0 0 –1 0 0
End moments –44 –28 +28 +24 –54 +30 +40 –40 –65 –67
(h) Find force causing arbitrary sway

-28 kNm -54 kNm

B C

-40 kNm
D
9m 9m

6m

SED
A SAB F SFC
E
-44 kNm -67 kNm -65 kNm

MB = 0 : MC = 0 : MD = 0

9 SAB = 44 + 28 9 SFC = 67 + 54 6 SED = 105
SAB = 8 kN SFC = 13.4 kN SED = 17.5 kN

Consider the frame as a whole and solve for the arbitrary sway force (Q) kN as shown
in the following figure:

B C Q
D

6m
9m

SED = 17.5 kN
E
SAB = 8 kN SFC = 13.4 kN
A F

9m 9m

H = 0 Q = 8 + 13.4 + 17.5 = 38.9 kN

(i) Sway correction
68
Correction factor = = 1.75
38,9
MAB MBA MBC MCB MCF MCD MDC MDE MED MFC
No-sway –15 –31 +31 +96 +214 –310 +251 –250 –125 +106
moment
Sway –44 –28 +28 +24 –54 +30 +40 –40 –65 –67
moment
1.75 x sway –77 –50 +50 +42 –94 +52 +70 –70 –114 117
moment
Final –92 –81 +81 +138 +120 –258 +321 –320 –239 –8
moments

150 kN 200 kN
3m 3m 3m
138 kNm
B C 321 kNm
100 kN
81 kNm 258 kNm D

120 kNm
6m
9m
-239 kNm

-92 kNm -8 kNm SED = 93.3 kN

E
SAB = 19.2 kN SFC = 12.4 kN
A F
9m 9m

9 SAB = 92 + 81 9 SFC = 120 – 8 6 SED = 239 + 321

SAB = 19.2 SFC = 12.4 SED = 93.3

Check: H = 100 + 12.4 – 93.3 – 19.2 = –0.1  (ok)

(j) Bending moment diagram

321
279 300
258
138

81

222
249

Bending moment diagram (kNm)

Free bending moment values:

150 kN 200 kN

C D

3m 3m 3m

6
/9  150 = 100 6
/9  200 = 133
67 50
3
/9  200 = 3
/9  150 =
167 183

M150 = 167  3 = 501 kNm

M200 = 183  3 = 549 kNm
7.14 Multi-storey frame

When a frame has more than one storey, the horizontal shear must balance for each
storey independent of the other storeys.

Figure 7.12(a) shows a double storey frame with any horizontal loading. The frame
will obviously sway to the right. However, it is not known by how much the sway will
be distributed to the top and bottom frames (the X and Y frames, respectively). Two
sidesway conditions must therefore be considered: Figure (b) and (c).

The design analysis procedure for a two-storey frame is as follows:

(1) Do a no-sway distribution of the moments due to the applied loading assuming
there is no joint translation and determine the unbalanced shear for each storey.
These are shown as S1 and S2 (positive to the right).

(2) Apply arbitrary sway to the top storey (figure (b)) and determine the unbalanced
shear for each storey: S3 for the top storey and S4 for the bottom storey. S3 will
now be a displacement force. S4 will be a prop force.

(3) Apply an arbitrary sway to the bottom storey (figure (c)) and determine the
unbalanced shear for each storey. S5 for the top storey which is a propping
force. S6 for the bottom storey, a displacement force.

(4) Sway correction must now be carried out to balance the shears S1 and S2.
Correction factors are applied to the conditions (b) and (c).
Let factor for (b) = x,
Let factor for (c) = y,
Then S3x + S5y = –S1

and S4x + S6y = –S2

The final moments are the summation of the moments from (1), (2) x and (3)y.
 1 1

S1 S3 S5
prop displace- prop
ment 2 2
Loading S2 S4 S6

prop prop displacement

(a) (b) (c)

Figure 7.12
7.15 Settlement of frame support

If settlement of a support of a portal frame occurs, then the fully clamped conditions
6EI
require a moment M = to be applied to each end of each beam.
L2
Problem 3.14.1
50 kN
3m 5m

C D

100 kN
4m
6m 2m

B E

5m

F
A Support F sinks by 5 mm

E = Constant = 210 GPa

KAB = KCD = KDE = KEF = 13.35  103 mm3

KBE = 26.7  103 mm3

I
where the relative stiffness factor, k =
L

You should attempt this problem (3.14.1) as part of the tutorial.

7.16 Tutorial

(1)
(a) A portal frame A-B-C-D is firmly fixed at A and D. The vertical legs A-B
and C-D are both 8 m high and each has a moment of inertia of 8 x
107 mm4. A-B is loaded with a horizontal central point load of 80 kN
working inwards while C-D is unloaded. The horizontal beam B-C is
12 m long, has a moment of inertia of 32  106 mm4, and carries a load
of 100 kN uniformly distributed over its whole length. A horizontal force
F is applied at C to prevent any sway. Draw a bending moment diagram
indicating the most important values, and also calculate the force F.

(b) Also draw the bending moment diagram if force F is removed.

(2) Do the necessary calculations and draw a dimensioned bending moment

diagram for the loaded portal frame shown in the figure. Also calculate the
horizontal and vertical reactions at the supports.

150 kN/m
100 kN B C

3m
IAB = IBC = 160 x 106 mm4 ICD = 480 x 106 mm4
A
6m

4m

(3) Draw a dimensioned bending moment diagram for the structure shown in the
figure. A is a rigid joint, and the moment of inertia of the beams A-B and A-C is
0.5I while that for the column A-D, is 2I. Also calculate the horizontal and
vertical reactions at D.
 80 kN

B A 4 kN/m C

3m 3m 6m

6m

(4)
(a) A rectangular portal A-B-C-D is fixed at A and D. A-B = 1.5 m, B-C =
3 m, and C-D = 2 m. EI is constant throughout. A horizontal inward force
= 25 kN acts at B. Find the bending moments at A, B, C and D. In giving
the bending moments, assume that compression on the outside of the
portal means a positive bending moment.

(b) Find the vertical and horizontal components of the reactions.

(5) Draw a dimensioned bending moment diagram for the portal frame A-B-C-D
shown in the figure. The moment of inertia of member B-C is 2I, while that for the
members A-B and C-D is I. Also calculate the horizontal and vertical reactions at
the supports.
 120 kN
1.5 m 1.5 m
40 kN/m B C

IBC = 2I
3m
IAB = ICD = I

6m
A

(6) Sketch the bending moment and shear force diagrams for the portal frame
shown. EI = constant.
B
C

2.5 m

56 kN

2.5 m

A D

2.94 m

(7) Use the moment distribution method to draw a dimensioned bending moment
diagram for the frame shown in the figure. E is constant throughout.
 4 kN
1.5 m
B
C
IBC = 2I
2m
4m
3 kN IAB = ICD = I
D

3m

3m

(8) The portal frame shown in the figure is prevented from swaying by the
application of a horizontal force at F. The moment of inertia of the beams of the
portal frame is twice that of the columns of the portal frame. The method of
moment distribution must be used to analyse this structure.

(a) Calculate the horizontal and vertical reactions as well as the fixing
moments at A, C and E. Also find the magnitude and the direction of the
horizontal force that prevents sway.

(b) Assume that the frame is allowed to sway and explain in words and
symbols how the final fixing moments may be found. (No calculations
are necessary).
 240 kN
2m
B C
F
IBC = 2I 3m
4m E
IAB = ICD = I

5m D

4m 4m

(9)

10 kN

A B
5m 2.5 m 2.5 m C

Beam A-B-C is simply supported on rigid supports at A and B and on an elastic

support at C, which sinks 1 mm per kN load. EI is constant = 50 MN m2. Find the
reactions at A, B and C and hence the distance that C sinks. Use moment distribution.

(10)
 6 kN 6 kN 10 kN

5 kN/m

(3I) (2I) (I)

A B C D

4.8m 2m 2m 2m 4m

(a) Beam A-B-C-D is fixed at A and simply supported at B and C. Draw the
bending moment and shear force diagrams.

(b) If C sinks by 5 mm, what are the new values of the bending moments?
E = 207 GPa
I = 810  106 mm4

(11) Use the moment distribution method to draw the dimensioned bending moment
and shear force diagrams for the frame shown in the figure. EI is constant
throughout.
90 kN/m
B C

2m

5m D

3m

(12)
(a) A rectangular portal A-B-C-D is fixed at A and D. A-B = 1.5 m, B-C =
3 m, and C-D = 2 m. EI is constant throughout. A horizontal inward force
= 25 kN acts at B. Find the bending moments at A, B, C and D. In giving
 the bending moments, assume that compression on the outside of the
portal means a positive bending moment.

(b) Find the vertical and horizontal components of the reactions.

(13) Use moment distribution to calculate the magnitude and direction of the
horizontal force required at B to prevent sway. EI of columns = 0.5 EI of beams.
27 kN/m
B D
F

3m

A C
E
4m 4m
LEARNING UNIT 8:
STRAIN ENERGY AND CASTIGLIANO’S THEOREM

Learning outcomes

After studying this learning unit, you should be able to

• distinguish between different equations for determining strain due to direct axial
stress, bending stress, shear stress and torsional stress, and a combination of
these stresses in pin-jointed frames, beams and portal frames

• find the forces in the members of structurally indeterminate pin-jointed frames,

and portal frames using Castigliano’s first and second theorem

• find the forces in pin-jointed frames and portal frames due to temperature
changes, shrinkage (expansion) of the members or settlement of the supports

• calculate the vertical and horizontal deflection of structures due to applied

loading

8.1 Introduction

In this learning unit, we demonstrate the strain energy method using examples and
give an overview of Castigliano’s first and second theorems.

8.2 Introduction to strain energy analysis

When a load is applied to a structure, each member of the structure is strained an

amount proportional to the stress it carries. The strain may be due to one or more
kinds of stress:

(1) Direct axial stress (tensile or compressive)

(2) Bending stress
(3) Shear stress
(4) Torsional stress
The result of these strains will be to distort the structure slightly, causing
displacement of the points of application of the external loads.

The displacement of these points means that external work has been done on the
structure (External work = Average force × Distance).

The amount of energy stored in a structure under a given loading system is defined
as the strain energy of the structure.

External work done = Internal work stored

= Increase in internal energy
= Strain energy

The expression for strain energy provides an extra equation required to solve
redundant structures (i.e. statically indeterminate structures).

8.3 Strain energy due to direct axial stress

Let us consider a bar of cross-sectional area A, length l, subject to an external force

F. The force is applied gradually with an initial value of zero and increased in
proportional value to the strain. The average value of the load = 1/2  the maximum
value of the load.

Extension

Work done


Load
 F

F
(a) (b)
Figure 8.1
 F
Change in length  = where E = Young’s modulus.
AE

Strain energy stored in bar = Work done

= Average force  distance moved

F F
Strain energy U = 
2 AE

F2 
that is, U =
2AE

If there are a number of members in a structure (struts and ties):

F2 
Total strain energy due to direct stress  = 
2AE

8.4 Strain energy due to bending stress


M
M

x


Figure 8.2

Consider an element of beam, length x, to which an external moment M is gradually

applied, causing a rotation as shown in the figure.

x = R Equation (1)

 f E M
From the theory of bending = =
y R I
1 M
 =
R EI

x
From Equation (1)  =
R
M
= x
EI
Strain energy stored in element = Work done

= Average moment  angle turned through

M M
=  x
2 EI

M2
Strain energy U = x
2 EI

Total strain energy in beam due to bending stress:


U=  dU
0


M2
U= 0 2 EI dx

If there are a number of members subject to bending in a structure:


M2
Total strain energy due to bending stress =   dx
0
2 EI

Example 8.1

Find the strain energy in a simply supported beam of span L carrying a central point
load W. EI = constant. Also find the deflection at the point load.
 W

B
A C
x
L
L
/2 /2

Consider portion A-B

Wx
M=
2

2
W 2 x2
M =
4

L
2
M 2 dx
Strain energy in A-B = 
0
2 EI
𝐿
2
1 𝑊 2 𝑥 2 𝑑𝑥
= ∫
2𝐸𝐼 4
0
𝐿
𝑊2 𝑥3 2
= [ ]
8𝐸𝐼 3 0

W 2 L3
=
192 EI

Consider portion B-C

Wx  L  WL Wx W
M= − W x −  = − = ( L − x)
2  2 2 2 2
W2
M2 = ( L − x) 2
4

L
M 2 dx
Strain energy in B-C = 
L 2 EI
2
 L
1 W2 W 2 L3
( )
2 EI L 4
− =
2
= L x dx
192 EI
2

(Alternatively: due to symmetry)

𝑊 2 𝐿3
Strain energy in B-C = strain energy in A-B =
192𝐸𝐼

2  W 2 L3
 Strain energy in whole beam =
192 EI

W 2 L3
=
96EI

Deflection at the point load

External work done = Average load x distance moved

W
= 
2

But strain energy = External work done

W W 2 L3
 =
2 96EI
WL3
giving  =
48EI

Example 8.2
Compare the strain energies due to bending of a uniform beam of span L, with a
central point load W, if

(a) both ends are fully restrained

(b) one end is fully restrained, and the other end is simply supported
(a) Both ends restrained

A B C

L L
/2 /2

Bending moment diagram

Find deflection at B:

Mohr II

B = intercept of carried tangents at A and B at B

M
= Moment of diagram between A and B about B
EI

1  1 WL L 1 L 1 WL L 5 L 
=      −    
EI  2 8 4 3 4 2 8 4 12 

WL3
B = − (i.e. deflection is downward)
192 EI

2 3 2 3
1 1 W L = W L
Strain energy U = W =
2 2 192 EI 384 EI
 (b) One end is fully restrained, and the other end is simply supported

A B C
L L
/2 /2
R

RL
+

= WL

Find R

Point C does not deflect, that is, intercept of carried tangents at A and C at C = 0.

Mohr II

0 = intercept of tangents at A and C at C

M
= Moment of diagram between A and C about C
EI

1 2 L 1 WL L  L L
=  RL  L  −    + 
2 3 2 2 2 3 2

5
which gives R = W
16
Find deflection at B:

B = intercept of carried tangents at A and B at B.

M
= Moment of diagram between A and B about B
EI

1 5 L L 1 5 L L 1 WL L L 
  WL   +  WL   −    
2 3 2 2 2 3 
=
EI  32 2 4 2 32

7 WL3
= − 
2 384 EI

3
7 WL
Strain energy U = W = 
1
2 4 384 EI

Comparing the strain energy in (a) and (b), we find that:

7
Strain energy in (b) =  strain energy in (a)
4
Example 8.3

Sketch the bending moment diagram for the structure A-B-C-D. Calculate the energy
stored in the structure and hence the vertical displacement of point D due to the
gradually applied load of 60 kN. Consider strain energy due to bending only. E = 210
GPa, IAB = 96  106 mm4, IBC = ICD = 48  106 mm4.

4m
150
A B

2m

D
VA = 60 kN C
60 kN

1,5 m
150

Bending moment
x1 90
x2 diagram (kNm)

90
x3

MC = MB = –60  1.5 = –90 kNm

MA = –60  2.5 = –150 kNm

M2
Total strain energy Utotal = 
2 EI
For A-B (origin at A): Mx 1 = –150 + 60x1 kNm

4
(150 − 60 x ) 2
U AB = 
0
2 E  96
dx  10 6 kNm

22500 + 3600 x 2 − 18000 x

4

=  10 6 dx
0
2 E  96
11400  10 6
10 6
 
4
= 22500 x + 1200 x 3 − 9000 x 2 = kNm
192 E 0 96 E

For B-C (origin at B): Mx 2 = 90 kNm

2
(90) 2 dx
 2 E  48  10
6
U BC = kNm
0

1 16200  10 6
=
96 E
8100 x 0 =
2

96 E

For C-D (origin at D): Mx 3 = 60x 3

1,5
3600 x 2 dx
U CD = 0 2 E  48
4050  10 6
1
 
1,5
= 1200 x 3 =
96 E 0 96 E

106
𝑈𝑡𝑜𝑡𝑎𝑙 = [11400 + 16200 + 4050]
96𝐸

31650 × 106
= = 1,570 𝑘𝑁𝑚
96 × 210 × 106

But: External work done = Total strain energy

 60
 = 1570
, kNm
2
1570
,  10 6
= mm
30
 = 52 ,3 mm (downwards)

8.5 Strain energy due to shear stress

It can similarly be shown that


S2
U= k
0
2GA
dx

where k = shape factor.

(Note: Shear stress is not uniform over a cross-section, and to allow for this a shape
factor k is introduced.)

S = gradually applied shear force

A = area over which shear force acts
G = modulus of rigidity
 = length

8.6 Strain energy due to torsional stress

It can similarly be shown that


T2
U= 0 2GJ dx
where J = polar moment of inertia for circular sections
T = torsional moment
G = modulus of rigidity
 = length

If the cross-section of the member under consideration is a shape other than circular
or tubular, the equation above is modified.
For example, for a rectangular member with dimensions h > b,


T 2L
U= 0 2Cb3hGdx
where

hb 3 16 b b 4 
C=  − 3.336 1 − 
16  3 h  12h 4 

8.7 In general

Total strain energy

   
F2 M2 S2 T2
U total =  dx +  dx +  k dx +  dx
0
2EA 0
2 EI 0
2GA 0
2GJ

In pin-jointed frames, generally, only the first term is of importance. In beams and
portal frames, generally, only the second term is of importance.

8.8 First theorem of Castigliano

Consider an elastic structural member carrying gradually applied loads W1 W2 W3 W4

... as shown, the deflection under each load being y1 y2 y3 y4 ...

(It is assumed that the reactions do not work. If there are displacements at the
reactions, they do work and must be considered as loads.)

W1 W2 W3 W4

y1 y2 y3 y4

y1 y2 y4

y3

Figure 8.3
External work done by the loads gradually applied is:

1
U= (W1y1 + W2y2 + W3y3 + W4y4 ...) Equation (1)
2

Now say load W3 is increased by W3, causing y1 y2 y3 y4 ... to be increased by

y1 y2 y3 y4 ... respectively.

The increase in external work done is:

1
U = W1y1 + W2y2 + W3y3 + W4y4 ... + W3 y3
2

(Note: W1 W2 W4 ... have already been fully applied

1
 Increase in work = W1y1 etc, not W1y1 ... )
2

W3 y3 is of the second order of smallness and therefore negligible.

 U = W1y1 + W2y2 + W3y3 + W4y4 Equation (2)

Now assume the member unloaded and W1 W2 (W3 + W3) W4 ... applied gradually.

The final deflections will be as before:

(y1 + y1), (y2 + y2) , (y3 + y3) , (y4 + y4) ...

Now total work done

1
= [W1 (y1 + y1) + W2 (y2 + y2] + (W3 + W3) (y3 + y3) + W4 (y4 + y4) ...]
2

1 1
= (W1y1 + W2y2 + W3y3 + W4y4 + ...) + (W1y1 + W2y2 + W3y3 + W4y4 ...)
2 2
1 1
+ W3y3 + W3 y3
2 2
 1 1
=U+ U + W3y3
2 2
But total work done = U + U
1 1
 U + U = U + U + W3y3
2 2

U
that is, y3 =
W3

U
In the limit: y3 =
W3

U U U
Similarly y1 = y2 = y4 = …
W1 W2 W4

U
In general terms: yw =
W

The first theorem of Castigliano can now be stated as follows:

The displacement of the point of application of any load along the line of action
of that load equals the partial differential coefficient of the total strain energy in
the structure with respect to that load.

Similarly, in general terms  M = U

M

The rotation of the point of application of any moment, in the direction of that
moment, equals the partial differential coefficient of the total strain energy in the
structure with respect to that moment.
8.9 Deflection of structures: application of Castigliano 1

The deflection of a joint in a statically determinate pin-jointed frame due to an external

loading system can now be found.

Consider a statically determinate pin-jointed frame subject to an external loading

system. Apply an imaginary (dummy) load P at the relative joint (say A) in the
direction of the required deflection at this point.

Let Fw = force in chosen member due to external loading system

Let Fp = force in chosen member due to load P only

that is, Fp = fpP where fp = force in member when P = 1

Total force in any member = Fw + Fp

= Fw + fpP

Strain energy for whole structure

F2 L
U =  (see section 4.2)
2AE


( FW + f P P) L
2

=
2AE

U
But by Castigliano I:  =
p

2( FW + f P P ) L  f p
that is, A = 
2 AE

FW f P L Pf p2 L
=  +
AE AE

FW f P L
But P = 0  A = 
AE
where  = deflection of point of application of dummy load P in the direction of P

Fw = force in members due to external loading

fw = force in members due to unit load acting at point of application of

dummy load P in the direction of the required deflection

A = area of member

L = length of member

E = Young’s modulus

Note: Instead of the areas of the members of a frame being given, it is sometimes
stated that all tension members are stressed to a given value (say 120 MPa) and all
compression members to a given stress (say 60 MPa). Then, if the load system is
F
known, the sign of the force in each member can be determined. The term in the
A
above expression must then be replaced by the given stress with the appropriate sign.

Similarly, for beam deflections:

L
M W m p dx
A = 
0
EI

where A = deflection at A

Mw = moment due to external loading system

mp = moment due to unit load at A in the direction of the required deflection

EI = rotational stiffness (or flexural rigidity)

Example 8.4

A parallel chord truss is loaded as shown in the sketch below. The tension members
all have an area of 25 cm2 while the compression members all have an area of
37.5 cm2.

Calculate (a) The vertical deflection of point P

(b) The horizontal deflection of point P
EI = 210 MPa

50 kN 100 kN 100 kN

1 2 3 4

10 12 14
16 17 1,8 m
9 11 13 15

5 P 6 7 8

2,5 m 2,5 m 2,5 m 2,5 m

(a) Vertical displacement of P

Step 1 Find the forces F in the members due to the loading.

50 100 100

-156,25 -243,16 -243,16 -190,97

-156,25 -190,97

112,5 kN 137,5 kN

Step 2 Apply unit load at P acting vertically down and calculate the force fp in
all members.
 -1,04 -0.69 -0.69 -0,35

+1,04 +0,35
1

0,75 0,25

Step 3 Tabulate as follows

Notes (1) Compression = – Tension = +

(2) E is constant and can be introduced after the summation

 Member Length L Area A Fw (kN) fp Fw fp L
(cm) (cm2) A

1 250 37.5 -156.25 –1.04 +1084

2 250 37.5 -243.16 –0.69 +1119
3 250 37.5 -243.16 –0.69 +1119
4 250 37.5 -190.97 –0.35 +446
5 250 25 0 0 0
6 250 25 +156.25 +1.04 +1625
7 250 25 +190.97 +0.35 +668
8 250 25 0 0 0
9 180 37.5 -112.5 –0.75 +405
10 308 25 +192.64 +1.28 +3038
11 180 37.5 -112.5 +0.25 –135
12 308 25 +107.03 –0.43 –567
13 180 37.5 -100 0 0
14 308 25 +64.22 +0.43 +340
15 180 37.5 -137.5 –0.25 +165
16 308 25 +235.32 +0.43 +1247
17 180 37.5 -137.5 –0.25 +165
FW f P L( cm)
 = +10719
(
A cm 2 )

Converting to kN and m units, and introducing E:

10719  10 −2
F f L
p =  W P = m = 0,0051 m
EA 210  10 6  10 −4

Vertical deflection of point P = 5.1 mm (down)

(b) Horizontal deflection of P

Assume the horizontal displacement is to the left.  Apply unit (dummy) load at P in
that direction.

Step 1 Find forces Fw in the members due to the loading as before.

Step 2 Apply unit load at P acting as shown and calculate forces fp in all members.

P
1
1 +1 +1 +1

Step 3 Tabulate as follows:

Member Length L Area A Fw (kN) fp FW f P L


(cm) (cm2) A

6 250 25 +156.25 +1 +1562.5

7 250 25 +190.97 +1 +1909.7
8 250 25 0 +1 0

FW f P L( cm)
 = +3472.2
(
A cm 2 )

Converting to kN and m units and introducing E:

FW f P L 3472,2  10 −2
p =  = m = 0,00165 m
EA 210  10 6  10 −4

Horizontal displacement of point P = 1.65 mm (to left)

Example 8.5
Calculate the vertical deflection under the load for the frame shown, which is pinned
to the supports at A and B. All members are 1300 mm2 in area, and E = 200 kN/mm2.
E

100 kN

(32 )

C
(1) D

(3)

HA HB (=150/3 )
A
(=150/3 ) B
(150) (250)
1m 1m 3m

Solution
The lengths of members and the reactions at A and B are calculated and are shown
in brackets in the above sketch.

The forces in the members (Fw) are calculated using the method of resolution of forces
at a joint.

Joint A

F
V = 0 Fsin = 150

H
 F1
Joint B
F2

 H = 0 F1sin

V = 0 250 + F1cos = F2

250

Joint D

F
V = 0 Fsin = 400

H H = 0 H = Fcos
Joint E
 V = 0 4002 cos - 100 = Fsin
400
 =

400 F

4002 100
(Note: Check by resolving forces at C) F=500 kN

The forces in the members due to the external loading system can now be shown on
the frame. (Forces Fw).

1
Place unit dummy load at E. The forces in the frame due to the dummy load = 
100
Fw and is shown in brackets.
E

100 kN

(1) Frame showing forces

(-4)
-400 Fw and (fp) in members
D
Tabulate as follows:
Member Length Area (cm2) Fw (kN) fp FWfp L
(m) A
A-C 2 13 + 300 +3 + 600 13 = 46.15
3 3

B-C 2 13 + 300 +3 + 600 13 = 46.15

3 3

B-D 3 13 -400 -4 + 1600 3 = 213.17

13

C-D 1 13 -400 -4 + 1600 13 = 123.08

C-E 5 13 +500 +5 12500 = 961.54

13
D-E 3 2 13 −400 2 −4 2 + 9600 2 = 1044.18
13

FW f P L( cm)
 = +2434.27
(
A cm 2 )
FW f p L 2434,27
y =  = m = +0,1217 m
AE 200  10 6  10 −4

Vertical deflection of E = 121.7 mm (down)

Example 8.6

The frame shown is supported at A and C, and a load of 10 kN acts down at B. An

upward force R is applied at E so that there is no vertical deflection at this point.
Calculate the value of R. The members of the frame all have the same cross-sectional
area and are made of the same material.

R
D
E F

A C
B
10 kN
3m 3m
Solution

R forms part of the external loading system. Calculate deflection at joint E and equate
to zero. The unknown lengths are shown in brackets in the above sketch.

Forces in members due to external Forces in members due to unit load at E as

loading system (Fw): shown:
R 1

D F D F

+R +1

A C
B A B C

10 kN

Joint D F2 Joint D
F2
 1
/2
F1 F1
R
(5- /2) 

V = 0 V = 0
R 1 5
F1 cos  = 5 – F1 cos  = F1 =
2 2 8
4 R
F1 = 5 –
5 2
H = 0
F1 = 25 4 − 58 R
 3
F2 = F1 sin  F2 =
8
H = 0
F2 = F1 Sin 

(
F2 = 25 − 5 R
4 8 5
3
)
F2 = 15 4 − 3 8 R

Tabulate as follows:

Members Length (m) Fw (kN) fp F w fp L

(5− R) ( )
1 4
A-D 4 – + – 5− R = R-10
2 2 2

(5− R) ( )
1 4
C-F 4 – + – 5− R = R-10
2 2 2
B-A 3 0 0 0
B-C 3 0 0 0

B-D 5 (
+ 25 4 − 5 8 R ) -
5
(25 4 − 5R 8 )(− 258 ) = +125R 64 − 625 32
8

B-F 5 (
+ 25 4 − 5 8 R ) -
5
(25 4 − 5R 8)(− 258 ) = + 125R 64 − 625 32
8

E-D 3 (
− 15 4 − 3 8 R ) +
3
( 4 8 8 )( )
- 15 − 3R 9 = + 27 R − 135
64 32
8

E-F 3 (
− 15 4 − 3 8 R ) +
3
(
4 8 8 )( )
– 15 − 3R 9 = + 27 R − 135
64 32
8
B-E 4 +R 1 R  4  1 = 4R

43R 135
 Fw fp L = −
4 2

1 1  43R − 135
E =  Fw fp L =  
AE AE  4 2 

But E = 0
 1  43R − 135 = 0
  
AE  4 2 
135  4
R= kN
43  2

12
R=6 kN = 6.28 kN
43
BENDING: BEAMS AND PORTAL FRAMES

We have seen that

L
MW mp
A = 
0
EI
dx (see section 4.8)

Example 8.7

Calculate the vertical and horizontal displacements of point A on the crane loaded as
shown below, and hence the resultant displacement of A. The crane is made of an I
section for which Ixx = 1522 cm4. E = 210 000 MPa.

A
4m

B 45
20 kN

3,4 m

C
Solution

Downward deflection of A

s A A
s

x x
B 20kN s y B 1

x
y
C x = scos45 = s/2
C
y = scos45 = s/2
8
208 Place unit load at A

s
Tabulate as follows (hogging bending moment = –)

Member Mw (kNm) m Mwm Mwm dx

20s −s 4
640
−20 x = − +10s2
A-B
 10s ds =
2
2 3
2 0

B-C − 8 3, 4
−20 8 +160
 160dy = 160  3,4
0

1
EI 
 MWm

1  640 
=  + 160  3,4
EI  3 

213 13 + 544
V = m = 0,237 m
21  10 7  1522  10 −8
V = 237 mm

Horizontal displacement of A
 s A 1
A
y

20
B B
y1

C C

(As before) Place unit load at A

Member Mw (kNm) m Mwm Mwm dx

A-B 20s −s 4
640
−20 x = − +10s2
 10s ds =
2
2 3
2 0

B-C −20 8 − ( 8 + y1 ) 160 + 20 8y1  (160 + 20

3 ,4

)
8 y1 dy
0

= 871

1
EI 
 MWm

1  640 
=  + 871
EI  3 

213 13 + 871
H = m = 0,339m
21  10 7  1522  10 −8
H = 339 mm

Resultant deflection
Resultant deflection H = 339

V = 237 R
R = (237) 2
+ ( 339)
2

R = 413.6 mm

Example 8.8

Calculate the horizontal displacement of point D in the figure shown below.

2m
B C

2m
4m D P

A EI = constant

Solution

P
1

P 2P
1 2

Place unit load at D

Tabulate as follows:
 Member s = 0 at Limit of s Mp m Mm

D-C D 0-2 –Ps –s Ps2

C-B C 0-2 –2P –2 4P
A-B A 0-4 2P – Ps 2–s 4P – 4Ps +
Ps2

U 1
EI 
p = = M p mds
P
over whole structure

2 2 4

EI p =  Ps ds +  4 Pds +  (4 P − 4 Ps + Ps )ds

2 2

0 0 0

 
4
2
 4s2 s3 
+ 4 P s0 + P 4 s +
3
= Ps 3 + 
2

0  2 3 0
= 8 P 3 + 8 P + 16 P − 32 P + 64 P 3
= 16 P
16 P
p =
EI

Example 8.9

Calculate VA, VB and MA for the beam shown below, using strain energy methods. EI
is constant.

10 kN
50 kN/m
MA C
A
10m 2m

VA
VB
Solution

Mmdx
VB does not deflect.   =0
EI

10 kN
MA =2620-10VB 10
x
MA
C C
A x x1 A x x1
10m 2m 10m 2m

VA=600-VB
VB 1 1

External loading Unit (dummy) load in lieu of (i.l.o). VB

Find MA and VA in terms of VB and loading system.

MA = 0 MA + 10 VB = 50 x 10 x 5 + 10 x 12
MA = 2620 – 10 VB

V = 0 VA = 510 – VB

Calculate Mw and m and tabulate. Hogging BM is –

Member x = 0 limits Mw (kN) m Mmdx

at
1000
A-B A 0-10 (510 − VB ) x − (2620 − 10VB ) −
50 x 2 -x +10 ( VB − 200)
3
2
(See below)
C-B C 0-2 -10x 1 0 0
1000
Mmdx = ( VB − 200)
3
Member A-B

10 10

 Mmdx =  (510 − V )
1
B x − 2620 + 10V B − 25x 2 ( − x + 10)dx
EI 0 0

1000
= (VB − 200,5) (obtained after lengthy arithmetic)
3
= 0 for zero deflection of V B
V B = 200,5 kN
V A = 309 ,5 kN
M A = 615 kNm

Example 8.10
Calculate H and MA for the structure
shown. EI is constant.
B
HB = H

3m
0,25m 5 kN
C
4m

MA

HA = H A Mmdx
Solution HB does not deflect,    =0
VA EI

B HB = H B 1
x 5 kN x

x1 x1
C
0,25
m
HA = H A
A
VA
External loading Unit (dummy) load in lieu of ( i.l.o). HB
 Calculate Mw and m and tabulate
Member x = 0 limit Mw m Mwmdx
at s
3
B-C B 0-3 HBx 1.x
3
 x3 
0 B =  = 9HB
2
H x dx H B
 3 0
7
C-A 5  5 
C 3-7 Hbx1 -
4
1.x1
  H
3
B x12 − x1  dx1
4 
7
 H B x13 − 5x12 
= 
 3 8 3
H  245 − 45
= B (343 − 27) −   − 25
3  8 
316 H B
= − 25
3
316H B
  M W mdx = 9HB + - 25
3
343H B
= − 25
3

1  343H B 
 − 25 = 0 for zero deflection of HB
EI  3 

75
that is, HB = HB = 0.219 kN
343

5
MA = 0 MA – 7HB + = 0 giving MA = 0.281 kNm
4
Example 8.11
Draw the bending moment
diagram for the portal frame
loaded as shown. EI = constant.

40 kN C

E
B
25 m

10 m
D

A
10 m 10 m

Solution

The horizontal reactions (HA = HD = H) do not deflect

Mm
 dx = 0
EI

s
y=0,65
40 kN C
x = 0,85
12,5

12,5
25
B 25 m 15

10 m 20
s
s D 1 1
A HD= H
HA = H
10 m 10 m

20 20 D

External loading Unit (dummy) load in lieu of (i.l.o) H

 VA = VD = 20 kN
HA = HD = H

Calculate MW and m and tabulate as follows

Mem- s=0 Limits Mw m Mwm

ber at
A-B A 0-10 –Hs –1,s +Hs2
B-E B 0-12.5 +20x – H(10+y) –(165–10H–0.6Hs)(10+0.6s)
= 20(0.8s)–H(10+0.6s) = –160s + 100H + 12Hs
= 16s – 10H – 0.6Hs –1(10+0.6s) – 9.6s2 + 0.36 Hs2
E-C C 12.5-25 +20x–H(10+y)–40 (x– –(165–10H–0.6Hs–325+400)
10) (10+0.6s)
= 20(0.8s)–H(10+0.6s) = –80s + 100H + 12Hs–4000
– 40(0.8s–10) + 9.6s2 + 0.366 Hs2
= 16s–10H–0.6Hs–32s –1(10+0.6s)
+400
D-C D 0-25 –Hs –1,s +Hs2

EIH = 0

10 12 ,5

 Hs ds +  (−160s + 100 H + 12 Hs − 9,6s + 0,36 Hs 2 ds )

2 2

0 0
25 25

+  (−80s + 100 H + 12 Hs − 4000 + 9,6s )

+ 0,36 Hs ds +  Hs 2 ds = 0
2 2

12 ,50 0

Lengthy arithmetic gives

13666.66 H = 43750

that is, H = 3.2 kN

 80

C
80
56
Free BM

32 40  20/4 = 200 kNm

200
32
B

144

A D

Bending moment diagram (kN)

8.10 Second theorem of Castigliano

Strain energy can also be used to find the forces in redundant members of imperfect
frames (i.e. frames with redundant members) provided the deflection, or lack of fit, of
the redundant member is known.

Consider a statically determinate frame A-C-B-D as shown in the figure (a) below.
L


C
(a)
A B
K

L


C (b)
R
B1
R
A B
K

increase in length of AB1

 Figure 8.4

Distance A-B = L, and member A-K is not connected where it crosses C-D. In figure
(a) it is also not yet connected at B.

Suppose that member A-K of length L –  and cross-sectional area A, was strained to
connect at B. Then both the frame and member A-K would be strained, and forces will
be induced in all members. The new frame is shown in figure (b). Let the force in A-K
be = R.

R
A
Increase in length KB1
For member AK : E=
orginal length L − 

R( L −  )
 KB1 =
AE

( L − ) + ( AE )
R L− 
The final length of member is A-B1 =

BB1 = L – A-B1

 RL R 
= L – ( L −  ) + −
 AE AE 

R
Since  is very small and E very large, the term may be neglected.
AE

 RL 
 BB1 = L −  L −  + 
 AE 

RL
that is, BB1 = −
AE

But, from Castigliano I, if a force R were applied at B in the direction of A:

 U 0
BB1 =
R

U0 RL
 =−
R AE

Where U0 = strain energy in original frame (figure (a))

By putting in the extra member A-K, the strain energy increases by an amount UAK
where

R 2 ( L − )
UAK =
2 AE


Again is negligible compared to L/E
E

R2L
 UAK =
2AE

Differentiating partially with respect to R:

UAK RL
=
R AE

The total strain energy in the redundant structure ACB1D = Utotal = U0 + UAK

Utotal U0 UAK

= +
R R R
Utotal RL RL
that is, =− +
R AE AE
Utotal
that is, =
R
The second theorem of Castigliano can now be stated as follows:

The partial differential coefficient of the total strain energy of a redundant

structure with respect to the force in the redundant member equals the initial
lack of fit of the member.

8.11 Redundant structures: application of Castigliano II

Consider any plane frame A-B-C-D with one redundant member, carrying a load W at
C as shown in (a) of the figure below.
B B
A A B A
1

1
C
C C
D D
D
(a)
W (b) W
(c)

Figure 8.5

Assume A-C is the redundant member and is too short by an amount . When A-C is
removed the frame becomes statically determinate as shown in figure (b). Let the force
in any one member (say A-B) of the frame in figure (b) = F0.

Now remove load W and apply unit load in place of the redundant member, as shown
in figure (c). Let the force in any one member (say A-B) of the frame in figure (c) = f.

Then, total force in any one member (say A-B) of the redundant frame = FD + fR (where
R is the actual force in the redundant member).

( FD + fR)2 L
 Strain energy in the chosen member (A-B) =
2AE

Total strain energy in redundant frame:

 U total =
( FD + fR ) L
2

2AE
U total 2( FD + fR ) Lf
=
R 2AE
F fL f 2 RL
=  D +
AE AE

U total
From Castigliano II : =
R
FD fL f 2L
  + R =
AE AE

FD fL
−
i. e. R = AE
2
f L

AE

For  = 0

FD fL
−
R= AE
f 2L

AE

The procedure for determining the forces in a singly redundant plane frame can now
be formulated as follows:

Step 1 Remove any one member and determine the forces in the statically
determinate frame F0.

Step 2 Remove the external loading and apply unit load in place of the member
removed and determine the forces f in the frame.

2
F fL f L
Step 3 Calculate D and for each member and summate in tabular form. Use
AE AE
the normal sign convention.
Step 4 Calculate the force in the selected member.
𝐹𝐷 𝑓𝐿
∑
𝑅= 𝐴𝐸
𝑓 2𝐿
∑
𝐴𝐸

Step 5 Find the forces in all the other members from F0 + f R for each member.

Similarly, it can be shown that in members subject to bending moments only

L
M W mdx
−
EI
R= L
0

m 2 dx
0 EI

where R could be a redundant reaction

 could be a sinking support
Example 8.12

The plane frame shown in the figure is pin-jointed throughout. All bars are of the same
material and have a constant length: cross-sectional area ratio. Determine the loads
in the members of the frame due to the given loading.
1,5 1,5 1,5

B C D
A

50 kN 100 kN
E F

G H

J K

Solution

Step 1 Remove redundant member BF and calculate the forces F0 in the frame as
shown in figure (a) below.

Step 2 Remove external loading and replace BF with unit load. Calculate the forces
f in the frame as shown in figure (b).

Step 3 Summation as shown in the table.

Step 4 Find R.

Step 5 Calculate the forces in the members as shown in the last column of the
table.
Note: Take tension negative
Take compression positive

50 B 50 C 100 D A B C D
A
1
502 502 1002
50 1
100 1
50 E F 100 E F
150
G H G H
150
J K J K

0 150

(a) Redundant member B-F (b) External loads removed. Unit load in
removed. Calculate F0. place of redundant member. Find f.
Tabulate as follows
Member F0 (kN) f F0f f2 fR F0 + fR
(kN)
A-B –50 0 0 –50
1 50 1
+ − +
B-C –50 2 2 2 –25 –75
C-D –100 0 0 –100
1 100 1
+ +
E-F +100 2 2 2 –25 +75.0
G-H 0 0 0 0
1 1
+
B-E 0 2 0 2 –25 –25.0
E-G 0 0 0 0
G-J 0 0 0 0
1 50 1
+ +
C-F +50 2 2 2 –25 +25.0
F-H +150 0 0 +150
H-K +150 0 0 +150
A-E +50 2 0 0 +70.71
B-F 0 –1 0 1 +35.36 +35.36
E-C −50 2 –1 50 2 1 +35.36 –35.36
G-F 0 0 0 0
J-H 0 0 0 0
F-D +100 2 0 0 +141.42
= +100 2 =4
= 141.42

FD fL
−
R= 2
AE  = 0( no lack of fit )
f L

AE
−141,42
∴𝑅= = −35,36 𝑘𝑁
4

Now the fR and F0 + fR columns in the table can be completed.

L
(Note: and E are constant for all members and cancel out.)
A

Example 8.13

The frame shown in the figure has members with cross-sectional areas as follows:

A-B = D-C = 3 cm2

 Remaining members = 1 cm2

Use the method of strain energy as applied to redundant frames to find the forces in
the members.

A D

B C

0,8m

10 kN 20 kN

Solution

Step 1 Remove redundant member B-D and calculate the forces F0 in the frame as
shown in figure (a) below.

Step 2 Remove external loading and replace B-D with unit load. Calculate forces f
in frame as shown in figure (b).

Step 3 Summation as shown in the table.

Step 4 Find R.

Step 5 Calculate forces in members.

 A A
A
D

1A
1A
10 20

1A

B C C
B

10 kN 20 kN

(a) (b)
Redundant member removed and Loads removed. Redundant member
F0 found. replaced by unit load and f found.

Tabulate (tension negative)

Membe Lengt Area F0 f F0 fL f 2L fR F0 + fR
r h (m) (cm2) (kN) A A (kN)

A-B 0.6 3 –10 3 –1.2 9 +0.82 –9.18

+
5 125
B-C 0.8 1 0 4 0 64 +1.09 +1.09
+
5 125
C-D 0.6 3 –20 3 2.4 9 +0.82 –19.18
+
5 125
D-B 1.0 1 0 –1 0 1 –1.36 –1.36
A-C 1.0 1 0 –1 0 1 –1.36 –1.36
 = -3.6  = 2.656
FD fL
−
R= AE = 0 + 3,6 = +1,36 kN
2
f L 2,656

AE

Now find F0 + fR and complete the last column of the table.

Example 8.14

By treating the reaction at C as a redundant reaction, calculate the force in each

L
member of the pin-jointed frame shown in the figure. Note: for all members is the
AE
same.

50 kN F
45

B
45

45 45
45 E
D

C-D = D-E = A-B = E-F

Solution

Step 1 Remove the redundant reaction at C and find the forces in all members due
to the external load of 50 kN.

50
L B L
A
VF =

L 90 L
C D E

VE=0
 2L
MF = 0: 50  = HE  L
2
100
HE = = 50 2
2

50 50
H = 0: HE = HF + HF = 50 2 – HF = 25 2
2 2

Joint F
F H = 0
25 2
FBF
= 25 2 FBF=50
2
FBF VF
50
V= 0 VF = VF = 25 2
2

Frame as a whole V = 0 : VE + VF = 25 2
VE = 25 2 – 25 2 = 0

Step 2 Remove external load and replace redundant reaction at C with unit load.
Calculate forces f.

VF = 2
2 2
A (L) B
(L)
2
2 2 2 2
1 3
C HE = 2
(L) D (L) E

1 VE = 1

MF = 0 1(2L) = HEL HE = 2

H = 0  HF = 2

Joint F: FFB = 2 2 VF = 2

V = 0 VE + 1 = VF = 2 VE = 1

Joint E
V = 0
FEB
FEB
=1
2
E
2 FEB = 2
FED
H = 0
FEB
1 + 2 = FED
2

FED = 3
Joint A
A V = 0
FAB
FAD 2
= FAD = 2
FAD 2 2
H = 0

2 2
FAB = + FAB = 2
2 2

Joint D
FDB V = 0

FDB 2
= FDB = 2
1 3 2 2
D
Step 3 Summation as shown in the table.
L
Note: = constant and need not be tabulated. Take tension = negative.
AE

Member F0 (kN) f F0 f f2 fR F0 + f R (kN)

A-B 0 +2 0 4 +33 +33
A-C 0 + 2 0 2 +16.5 2 +23.3
A-D +50 – 2 –50 2 2 –16.5 2 +26.7
C-D 0 –1 0 1 –16.5 –16.5
B-D –50 + 2 –50 2 2 +16.5 2 –26.7
B-E 0 + 2 0 2 +16.5 2 +23.3
B-F –50 +2 2 –100 2 8 +33 2 –3.3
D-E +50 2 –3 –150 2 9 –49.5 +21.2
F0 f = –350 2 f2 =30

Step 4 Find R
L
From Castigliano II: ( cancelled out)
AE

FD fL
−
R= AE  = 0( no lack of fit )
f 2L

AE

There is no settlement of support C, =0

that is, R =
(
− −350 2 ) = 16,5 kN
30

Step 5 Complete columns 6 and 7 of the table. The force in each member is given
in column 7.
Example 8.15

Three bars A-D, B-D and C-D meet in a pin joint at D. At A, B and C they are pinned
to a rigid support as shown in the figure. A-D, B-D and C-D have cross-sectional areas
of 2 cm2, 2 cm2 and 6 cm2, respectively. A-B-C-D forms a rectangle. If a force of 15 kN
acts at D as shown, use strain energy to find the forces in the bars. State whether they
are compressive or tensile.

A
B

C D

2m 13
3

15 kN 2

Solution

Step 1 Take member B-D as redundant, remove B-D and find forces in members
A-D and C-D.
Step 2 Remove the external load and let the unit load act in the redundant
member as shown.

A A
B

1 1
15

C D C
D
15 kN
Step 1 Step 2

Step 3 Tabulate as follows. Take tension as positive.

Membe Area Length F0 f F0 fL f 2L fR F0 +

r (cm2) (m) (kN) A fR
A
(kN)
− 3  3  3
D-A 2 3 +15 13 − 45   − 9 13   –5.3 +9.71
13  2   2

D-B 2 13 0 +1 0  13  +6.36 +6.36

1 
 2 

D-C 6 2 0 − 2 0 4  2 –3.54 –3.54

13 13  6 

 = − 45  3  = 2.944
 
13  2 

Step 4 Find R
From Castigliano II :
 FD fL
−
R= 2
AE  = 0( no lack of fit )
f L

AE
For no lack of fit  = 0

F0 fL
−
R = A (Note: E cancels out)
2
f L

A
 45 3 
− −  
 13 2 
R= = 6,36 kN
2,944
Step 5 The last two columns can now be completed. Note: Conversion factors for
area and length are the same top and bottom, hence they cancel out.
8.12 Strain energy in bending: Castigliano II

From previous sections, it has been shown that:

L
M W mdx
−
EI
R= L
0
(see section 8.10)
m 2 dx
0 EI

Example 8.16

A rectangular portal frame A-B-C-D is loaded as shown below. The distance between
A and D increases by 10 mm on application of the load. E = 200 GPa. Draw the
bending moment diagram for the loaded frame.

60 kN/m
E
B C
x2 x3

2m 2m
IAB = ICD = 22  106 mm4

IBC = 44  106 mm4

A x1 x4
H D H

VA VD

Solution

MA = 0 4VD = 60  2  1 VD = 30 kN
VA = 90 kN

Take H as the redundant

 M W mdx
 − 
Castigliano II: H= EI
2
m dx

EI
60 kN/m
B C B C
x2 E x3

x1
A D x4 D
1 1

30
90

Remove redundancy and Apply unit (dummy) load.

apply loading. Find Mw Find m
Tabulate as below (kN and m units)
Section x=0 limits Mw m 1 MW m m2 dx
at
EI  EI
dx  EI
5
10 6 10 6
22 E 0
A-B A 0-5 0 –1x1 0 x12 dx
22 E

10 6  125
=  
22 E  3 
2 2
60 x 2 10 6
 ( )
B-E B 0-2 90 x − –5 106 106
−5 90 x2 − 30 x22 dx2 ∫ 25𝑑𝑥2
2 44 E 44𝐸
44 E 0
0
2
10 6  3x 2 x 2  106
= (150)  2 − 2  = (25)
44 E  2 3 
0 22𝐸
−10
= ( 250)
22 E
2
10 6 106 10 6
2
C-E C 0-2 30x3 –5
44 E 44 E 
−5( 303 ) dx3
0 44 E 0
25dx 3

−106
= (150) =
10 6
(25)
22 E
22 E
D-C 10 6 10 6
5
D 0-5 0 –1x4 0
x
2
22 E 4 dx 4
22 E 0

10  125
6
=  
22 E  3 
10 6
( −400)  = 10  400 
6
=
22 E 22 E 3

6
− 106 − −2 + 10  400
−10  10 −3
( −400) 10
H= 22 E = 22  200  106
106  400  106  400
 
22 E  3  22  200  106  3
= 2 ,67 kN

The bending moment diagram can now be drawn.

MB = –2.67  5 = –13.35 kNm

MC = –13.35 kNm

ME = 30  2 – 13.35 = 46.65 kNm

Maximum free moment occurs at S. F is 0,

90 – 60x = 0
90
x= = 1.5 m from B
60

−60  1,52
Maximum free moment = 90  1.5 – –13.35
2
= 54.15 kNm

16,67

13,35 13,35

1,5m 54,15

Bending moment diagram (kN)

8.13 Temperature change, shrinkage and settlement

(1) Pin-jointed frames

FL
As the expression is the extension of a bar of length L, cross-sectional area A, of
EA
a material with Young’s modulus E, under action of force F, this could be replaced by
the extension due to any other effect, for example, temperature change.

e=Lt where L = length of member

 = coefficient of linear expansion
t = change in temperature

(2) Rigid frames

Remove the redundancies to allow the structure to move freely, thus determining the
lack of fit. Use Castigliano to obtain the values of the redundancies.
 V
V = th H

th
h h
M M
H H

h =

(a) Remove redundancies H and M (b) Remove redundancies H and V

U U
= t = th
H V

Figure 8.6: Temperature change

 
H
M
V

Figure 8.7: Settlement of support

8.14 Combined effect of direct stress, bending, shear and torsion

In general, where we have computed:

F 2 M2 S2 T2
U = dx +  dx +  k dx +  dx
2 EA 2 EI 2GA 2GJ
 U
and , we have ignored strain energy due to
P

(1) Bending, torsion and shear in pinned frames

(2) Direct stress, torsion and shear in beams and rigid frames

The figure below shows a cranked beam in which bending, shear and torsion occur
simultaneously.

C 1m D

2m
A 1
/2 kN/m downward
B

Elevation

1
/2 kN/m

A
B D
C
Figure 8.8

Beam with bending shear and torsion

One example will be done in order to compare the effect of various types of stress on
a structure.

Example 8.17
Find the vertical deflection of point A in the structure shown below.
10 kN A 1m Given:
B
I = 3000 cm4
A = 40 cm2
(W) =30kN 2m
k = 2.4 = shape factor (shear)
E = 210 kN/mm2
C
G = 80 kN/mm2
Solution

U FL FW L S S W M M
W = = ds +  k + ds
W EI  W GA  W EI  W

The thrusts, shears and bending moments in ABC will be:

W
10 A B 10 kNm W
10 W
1m

W = 30 2m W

10 kNm

10 kN
Thrust Shear Bending
W

(1) Thrust
F FL F
Member A-B F = –10 =0 =0
W EA W
𝜕𝐹 𝐹𝐿 𝜕𝐹 2𝑊
Member B-C F = –W = −1 =
𝜕𝑊 𝐸𝐴 𝜕𝑊 𝐸𝐴

2W 60
W = 0 + W =
EA EA
= 0,071 mm

(2) Shear

S SL
Member A-B S = –W = −1 k
W GA
S SL S
Member B-C S = 10 =0 k =0
W GA W
2,4W 72
W =
S = GA GA
= 0,225 mm
(3) Bending

M M M
1 1
1
= −S 0 EI W EI 0 Ws ds
= 2
Member AB M = - WS ds
W
Origin at A
1
30  S 3  10
W =   =
EI  3  0 EI

𝜕𝑀 𝐿 𝑀 𝜕𝑀 1 2
Member BC M = W – 10s = 1 ∫0 𝑑𝑠 = ∫ (𝑊 − 10𝑠)𝑑𝑠
𝜕𝑊 𝐸𝐼 𝜕𝑊 𝐸𝐼 0

Origin at B

1
 1
 2W − 20 =
40
1
W = Ws − 5s 2 =
EI 0 EI EI
10 + 40 50
Total  W = =
EI EI
= 7,937 mm

Total  = 8.233 mm
Thrust contributes 0.86%; Shear contributes 2.73%; Bending contributes 96.41%

8.15 Structures with direct stress and bending

Two examples of this type of structure will be given.

Example 8.18

In the structure shown in the figure, the pinned brace is made 2 mm too short and is
forced into place. Find the forces and moments in the members. The following data is
given.

Areas A-B = 50 cm2

C
A-C = 5cm2
IAB = 5000 cm4 R
3m
E= 200 kN/mm2
D B
A
2m 2m
10 kN
4m
Solution

A-C = 5 m

Let R, the tension in A-C be the redundancy.

U
Then =  = 0,002 m.
R

Considering the effects of direct stress and bending:

U FL F M M
=  +  ds = 0,002
R EA R EI R

(1) Direct stress

F F
Member A-C F = R = 1 F =R
R R

FL F R 5 10 4 R
= =
EA R E  5  10 −4 E

F F
Member A-B F = −0,8R = − 0,8 F = +0,64 R
R R

FL F 0,64 R  4 5,12 R  10 2 512 R

= = =
EA R E  50  10 −4 E E

(2) Bending

M M
Member A-D M = 0,6Rs = 0,6s M = 0,36Rs 2
R R
(origin at A)
2
M M 0,36R  s 3 
L 2
0,36Rs 2
0
EI R
ds = 0 EI ds =
EI  3  0
0,12 R
= 8
EI
0,96R
=
EI
Member D-B
 M
M = 0,6 Rs − 10( s − 2) = 0,6s
R
M
M = 0,36 Rs 2 − 6s 2 + 12 s
R

M M
4
(0,36Rs 2
) ds
− 6s 2 + 12s
0 EI R ds = 
2
EI
4 4 4
0,36R  s3  6  s3  12  s 2 
= − +
EI  3  2 EI  3  2 EI  2  2
0,12 R 2 6
=  56 −  56 +  12
EI EI EI
6,72 R 40
= −
EI EI

U 0,96R 6,72 R 40
E = 10 4 R + 512 R + + +
R I I I
7,68R 40
= 10512 R + −
I I
7,68R 40
= 10512 R + −8
−
5000  10 5000  10 −8

= 10512 R + 153600R − 800000

U
E = E = 10512 R + 153600R − 800000
R
0,002  200  10 6 = 10512 R + 153600R − 800000
giving R = 7.32 kN

4,39
5,86

5,86
4.39
2,42 kNm
4.39
5,86 5,86
A 2m 2m B
5,61
10 kN
Reactions and forces in members:

V = 0 VB + 4.39 – 10 = 0 VB = 5.61 kN
MB = 0 MB + 5.86  3 – 10  2 = 0
MB = 20 – 17.58 = 2.42 kNm

8.16 Tutorial

(1) Calculate the horizontal displacement of point C in the truss below. E =

210 GPa, A = 5000 mm2 for all members.
160 kN
4  3m

1 2 3 4 B
A 15
8 9 10 11 12 13 14
5 6 7
C
1,5m 33 m 1,5m

(2) The wall crane shown in the figure below has to carry a load of 50 kN
hanging from point D. All the members have the same cross-sectional area
viz 2500 mm2. Member A-B is 2 m in length. Using strain energy methods
determine the deflection of point C in magnitude and direction.

D
A
30

50 kN
A
60
90

C
60

B
(3) The pin-jointed frame shown in the figure carries a single vertical load P as
shown. All the members of the frame are made of the same material and
have the same length and cross-sectional area. Prove that joint C deflects
8.2 off the vertical.
P

600 600 600 600

C

(4) The figure shows a loaded plane frame with pin joints. Calculate the
magnitude and direction of the displacement of the node E under load.
Cross-sectional area of the tension members = 100 mm2 and cross-
sectional area of the remaining members = 200 mm2. E = 200 GPa.
 E

10 kN
D
60
60
90
C
60
60

A B

(5) Calculate the vertical displacement of joint D and the horizontal displacement
of joint C for the truss shown in the figure. The cross-sectional areas of the
members are 3000 mm2 for those in compression and 1200 mm2 for those in
tension. E = 200 kN/mm2.
B

D C
A
4m 4m

20 kN

(6) Use strain energy methods to determine the force in the redundant member F-
C of the frame shown below. All members are 20 cm2 cross-sectional area
except F-G, which is 40 cm2. E = constant = 207 kN/mm2.
F G

2m
A D
B
C

15 kN
32 m
(7) Find the force in the redundant member A-C of the frame shown below. The
areas of all members = 100 cm2. EI = 207 kN/mm2.
B C
90 kN

14,4 m
A D

14,4 m

(8) Use the first theorem of Castigliano to determine the deflection at B on the
simply supported beam A-D shown below.

20 kN
1 kN/m
A
2m 1,5m D
B C
6m

(9)

(a) Member A-B-C shown in the figure has a moment of inertia of 5000 cm4 and E
= 21 x 104 MPa. Use strain energy to find the downward deflection of C:
(i) As a result of bending in portion A-B
(ii) As a result of bending in portion B-C
What is the total downward deflection?

(b) Use strain energy to find the horizontal deflection of C as a result of bending.
(Note: Horizontal deflection as a result of bending in A-B will be negligible).
 B
A
<* 30

40 kN

4m 4m

(10)

(a) A cantilever 10 m long carries a point mass of 6000 kg at the end, and a
uniformly distributed load throughout of 20 kN/m. If the beam has a moment of
inertia of 150 000 cm4 and E = 21 x 107 kN/m2, use strain energy to find the
deflection at the end of the beam.

(b) Find the total strain energy of the beam.

(11) Find the upward deflection of B in the frame shown below. The frame has a stiff
joint at C and is encastre at D. The frame carries a load W at A. EI for C-D is
twice EI for A-C and B-C.
B C A
L
L
/2 /2

W
2L

(12) Use strain energy to find the forces in the members in the figure below.
(Hint: Take A-D as redundant). The areas of the members are:
 A A-D = 2 cm2
B-D = 2 cm2
1,5m
B C-D = 5 cm2

1,5m

C D

3m
5 kN

(13)

(a) Use strain energy to find the horizontal reactions at A and D of the portal frame
shown below.

(b) Draw the bending moment diagram for the portal frame. EI = constant.

30 kN/m
B E
C

6m

A D

4,5m 2,3m
(14) Draw the bending moment diagram for the frame shown. Joint B is a stiff joint
and EI = constant.
50 kN B

D 3m

C
A

2m 2m

4m 4m

(15) Using strain energy methods calculate the value of the horizontal and vertical
reactions at A for the frame shown in the figure and draw dimensioned shear
force and bending moment diagrams.

4m 12m

B 10 kN/m C
D

(2I) (2I)

8m (I)

(16) The frame shown in the figure is made of steel (E = 200 kN/mm2). The cross-
sectional area of the diagonals is 1000 mm2 and that of the other members
2000 mm2. During construction, member A-D was found to be 1 mm too short
and had to be forced into placed. Calculate the magnitude and nature of the
force developed in each member due to this misfit.
 3m

A B

4m

C D

(17) The members of the pin-jointed frame shown in the figure are all 3 m long. The
tension members have a cross-sectional area of 1500 mm2 and the
compression members have a cross-sectional area of 3000 mm2. Determine
the vertical deflection at C and also the horizontal movement on the rollers at
D. E = 200 GPa.
B

A A
C
60
30
D 100kN

(18) The truss shown in the figure is simply supported at A and B. The tie A-B is
stressed so that the deflection at C is zero. Calculate the magnitude of the
tensile force that has to be supplied in tie A-B.

20 kN

1m

D 1m
A B

3m 3m
(19) Use energy methods to calculate the vertical and horizontal deflection of point
C for member A-B-C shown in the figure. Member A-B-C has and I value of
6 500 cm4 and E = 210  103 MPa.

A
10 kN/m 40 kN

30
B C
1,500 1,500

(20) Calculate the reactions at A and B for the pin-jointed frame shown in the figure.
Then find the magnitude and the direction of the deflection of joint E. All the
struts in the frame have a cross-sectional area of 100 mm2 and all the ties a
cross-sectional area of 50 mm2. E = 210 GPa.

A C

2m

B D E
2m 2m

10 kN 10 kN

(21)

(a) Determine whether the pin-jointed frame shown in the figure is statically
determinate or not.

(b) Determine the forces in each of the members of the frame and state whether
they are tensile or compressive. All the members have the same cross-sectional
area and are made of the same material.
 20 kN

A E 600 B
30

8m

(22) A propped cantilever A-B carries a load of 3 kN/m over its full span of 6 m. The
upward reaction of the prop at B is 5 kN and end A is firmly fixed. Calculate the
deflection at B and draw fully dimensioned bending moment and shear force
diagrams for the structure. E = 200 GN/m2; I = 200  106 mm4.
LEARNING UNIT 9:
VIRTUAL WORK

Learning outcomes

After studying this learning unit, you should be able to

• find the forces in the members of structurally determinate pin-jointed frames

and beams using the virtual work method

9.1 Introduction

In this learning unit, we will demonstrate the virtual work method using examples.

9.2 Introduction to virtual work theory

In general, mathematical results concerning the virtual work done by a set of forces in
equilibrium undergoing virtual (or imaginary) displacements are of great use in
structural analysis. The work done by external forces moving through small
displacements compatible with the geometry of a structure is called the external
virtual work done.

Strain energy (learning unit 8) can be derived from the method of virtual work. Strain
energy is merely a restatement of the method of virtual work. What you will see
in this learning unit, will seem similar to the methods learned in learning unit 8 (strain
energy). Therefore, when you consult books on these two methods (i.e. virtual work
and strain energy), you may be confused because they are so similar.

9.3 The principle of virtual displacement

If a set of external forces acting on a structure is in equilibrium, then any virtual

(imaginary) rigid-body displacements given to the system cause virtual (imaginary)
work to be done by each force, and the total external work done is zero.
This principle may be used to obtain the equilibrium equations and hence the reactions
for statically determinate frames. Consider the statically determinate structure A-B-C
shown below.
4W 2W
hinge
B C
A (a)

L
/2 L
/2 L Statically determinate
/2 L L
structure
RA RB RC

A (b)

Virtual displacement

A given to RA

(c)

Virtual displacement

B given to RB

C (d)

Virtual displacement

C given to RC

Figure 9.1

Figures (b) (c) and (d) represent three sets of virtual displacements, each set
compatible with support conditions.

(1) Virtual displacement A given to RA

A 
External virtual work = RAA - 4W + 2W A = 0
2 4
 which gives RA = 3W/2

Note: Negative load is done if a load “moves back” (in this case the load of 4W down
moves upwards).

(2) Virtual displacement B given to RB

B 3 B
External virtual work = RBB - 4W – 2W =0
2 4

which gives RB = 7W/2

(3) Virtual displacement C given to RC

C
External virtual work = RCC – 2W =0
2

which gives RC = W

Notes:

(1) Any system of virtual displacement could be chosen. Those shown above are
convenient because they give each reaction directly.

R
(2) Virtual work done = R, not .
2

9.4 Virtual work

In the method of real work, the external work done by the applied loads moving
through their displacements equals the energy stored in the members (strain energy).

We can also equate the external work done to the work done on the members (the
virtual work).
Consider a member of length  m, with an axial for Fm. The length of the member is
changed by a virtual displacement   m, considered as an extension. Then work done
Fm  m
on the member = Fm  m (not ).
2
If the member were in tension and the virtual displacement shortened the member,
work would be done by the member and would be negative in the work balance
equation.

Consider a pin-jointed structure under the action of an external set of loads P. The
structure is in equilibrium, and the force in a particular member is Fm.

If all members of the structure are inextensible except for member m, then apply a
virtual displacement   m to the member. As a result, all joints in the structure move to
new positions and work will be done by the applied loading system.

Let j be the corresponding displacement of the load Pj in the direction of Pj.

External work done = Work done on the member

that is, Pj j = Fm   m

If all the members were allowed to extend:

 Pj j =  Fm   m

That is, Virtual work done by external loads = Virtual work done on members.

To use this method, a mutually compatible system of deformations (j, lm) must be
found. Note that the actual loading system has been used in conjunction with a
compatible system of virtual displacements.

9.5 Virtual work applied to statically determinate structures

Virtual displacements and actual loads can be used

OR
 Actual displacements and virtual loads

The latter is more commonly used in structural methods and is referred to as the unit
load method.

Consider the Warren truss below. The external loading system consists of loads P1
and P2 and the reactions at A and B. The deflection of joint E in the direction of the
arrow is required.

C D
E

F G B
A

P1 P2

Figure 9.2

P =  F 

The extensions   of all the members due to the external loading system can be found
by finding the forces in the members due to P1 and P2, and a knowledge of the load-
extension relationships.

In order to obtain only the displacement we are seeking on the left-hand side of the
equation, apply a unit load in the direction of the required displacement at E. This unit
load will produce forces in the members of the frame denoted by the set F1.

 1 x E = F1AC  AC + F1CD   CD + ...

The value of   for each member is the extension due to loads P1 and P2.

If all the members are uniform and linearly elastic:

F
 =
AE

F1 F
and E =
AE

Similarly, for bending effects:

M 1 Mdx MM 1dx
 =  = 

EI EI

9.6 Examples: statically determinate frames

Example 9.1

Calculate the vertical deflection of joint D in the Warren truss shown in the figure. The
length of all members = 5 m. The cross-sectional area of the top and bottom chord
members = 2500 mm2 and of the other members = 2000 mm2. E = 210 GPa.

C D E

F G B
A

50 kN 50 kN

Solution

Let F be the set of member forces due to the applied external loading system. Let F1
be the set of member forces due to unit vertical load at D (i.e. the virtual forces). Let e
= set of extensions due to set of F forces.
  F 
e = 
 AE 

Set of forces F: (found by resolution of forces at a joint)

C 57.8 D 57.8

57.8 57.8
57.7 57.7
A

28.9 57.9 28.9

50 50 kN 50 kN 50

Set of forces F1:

1

C 0.577 D 57.8
0.577

0.577
A

0.289 0.866 0.289

0.5 0.5
Tabulate as follows:

Member Length Area F (kN) e (mm) F1 F1e

(m) (mm2) F (mm)
=
AE

AC 5 2000 -57.7 -0.69 -0.577 0.398

AF 5 2500 +28.9 +0.275 +0.289 0.079
CF 5 2000 +57.8 +0.69 +0.577 0.398
CD 5 2500 -57.8 -0.55 -0.577 0.317
FD 5 2000 0 0 -0.577 0
FG 5 2500 +57.8 +0.55 +0.866 0.476
DE 5 2500 -57.8 -0.55 -0.577 0.317
DG 5 2000 0 0 -0.577 0
EG 5 2000 +57.8 +0.69 +0.577 0.398
EB 5 2000 -57.8 -0.69 -0.577 0.398
GB 5 2500 +28.9 +0.276 +0.289 0.079

F1e = 2.86

Vertical deflection of D = 2.86 mm (down)

Example 9.2

Calculate the vertical and horizontal components of the deflection at joint F, due to
the 20 kN load at E. Find the magnitude and direction of the resultant deflection at F.
The area of the compression members is 2 x 582 mm2 and of the tension members
582 mm2. E = 200 GPa.

B C D

A E F G H

20 kN

4 @ 4 m = 16 m

Solution

Let F be the set of member forces due to the applied external loading system.
Let F1 be the set of member forces due to:
(a) unit vertical load at joint F
(b) unit horizontal load at joint F
Let e = set of extensions due to set of F forces
F
(e = )
AE

Set of F forces: (found by resolution of forces at a joint) (kN)

 B 10 C 10 D

20
A 5 H
450 15 15 5
E F G

3
/4  20 = 15 kN 20kN 5kN

Set of F1 virtual forces due to unit vertical load at F

B 1 C 1 D

450 E G H
A F
0.5 0.5 0.5 0.5

0.5 1 0.5
Tabulate as follows (tension = positive):

Member Length A (mm2) F (kN) F F1 F1e (mm)

e =
(m) AE

AB 5.657 1164 -21.2 -0.515 -0.71 +0.37

BE 4 582 +20 +0.687 0 0
AE 4 582 +15 +0.515 +0.5 +0.26
BC 4 1164 -10 -0.172 -1 +0.17
BF 5.657 1164 -7.1 -0.172 +0.71 -0.12
EF 4 582 +15 +0.515 +0.5 +0.26
CD 4 1164 -10 -0.172 -1 +0.17
CF 4 0 0 0 0
DF 5.657 582 +7.1 +0.345 +0.71 +0.24
FG 4 582 +5 +0.172 +0.5 +0.09
DG 4 0 0 0 0
GH 4 582 +5 +0.172 +0.5 +0.09
DH 5.657 1164 -7.1 -0.172 -0.71 +0.12

F1e = 1.65
Vertical downward component of deflection = 1.65 mm

Set of F1 virtual forces due to unit horizontal load at F

B C D

A H
1 E 1 F 1 G

The only members that need to be tabulated are A-E and E-F

Member Length(m) A (mm2) F (kN) F F1 F1e (mm)

e =
AE
 A-E 4 582 +15 +0.515 +1 +0.515
E-F 4 582 +15 +0.515 +1 +0.515

 F1e = +1.03

Horizontal component of deflection = 1.03 mm to right

Resultant deflection at F

1.03mm
F
 R= (1,03) 2 + (1,65) 2
R = 1,95 mm
1,65
1.65mm  = tan −1
R 1,03
 = 58 0

9.7 Examples: statically determinate beams and structures

Example 9.3

A simply supported beam A-B carries a point load P at C as shown. Find the following:
(a) The deflection at C
(b) The slope at C
P

A B
C

a b

l
Solution

(a) Deflection at C

(1) Draw a bending moment diagram for the loading system.

(2) Apply unit load at C and draw a bending moment diagram for the unit load.

MM 1dx
(3) Find. 1.C = 
EI

x1 P x2

B
A
C

a b

l
Pb Pa
VA = VB =
 

Pab/l +

1) Bending moment diagram for loading system

 x1 1 x2

B
A
C

a b

l
b a
VA = VB =
 

ab/l +

(2) Bending moment diagram for unit load at C

For length A-C

𝑃𝑏𝑥1 𝑏𝑥1
𝑀= 𝑀1 =
ℓ ℓ
𝑎
𝑀𝑀1 𝑑𝑥1 1 𝑎 𝑃𝑏𝑥1 𝑏𝑥1
∫ = ∫ ( ) (− ) 𝑑𝑥
0 𝐸𝐼 𝐸𝐼 0 ℓ ℓ
1 𝑎 𝑃𝑏2 𝑥12 𝑃𝑏2 𝑎3
= ∫ 𝑑𝑥 1 =
𝐸𝐼 0 ℓ2 3𝐸𝐼ℓ2

For length B-C

𝑃𝑎𝑥2 𝑎𝑥2
𝑀= 𝑀1 =
ℓ ℓ
𝑏
𝑀𝑀1 𝑑𝑥2 1 𝑏 𝑃𝑎𝑥2 𝑎𝑥2
∫ = ∫ ( )( ) 𝑑𝑥2
0 𝐸𝐼 𝐸𝐼 0 ℓ ℓ

1 𝑏 𝑃𝑎2 𝑥22 𝑃𝑎2 𝑏 3

= ∫ 𝑑𝑥2 =
𝐸𝐼 0 ℓ2 3𝐸𝐼ℓ2
𝑃𝑎2 𝑏2 𝑃𝑎2 𝑏2
= 𝛥𝐶 = (𝑎 + 𝑏) =
3𝐸𝐼ℓ 3𝐸𝐼ℓ
(b) Slope of beam at C

(1) Draw a bending moment diagram for the loading system as before.
(2) Apply unit moment at C and draw a bending moment diagram for the unit load.

MM 1dx
(3) Find. 1   C = 
EI
x1 x2

1
B
A
C

a b

l
1 1
VA = VB =
 

- a/l

b/l +

3) Bending moment diagram for unit moment at C

For length A-C

𝑃𝑏𝑥1 𝑥1
𝑀= 𝑀1 = −
𝑙 𝑙
𝑎 𝑎 𝑎
𝑀𝑀1 𝑑𝑥1 1 𝑃𝑏𝑥1 𝑥1 1 𝑃𝑏𝑥12
∫ = ∫( ) (− ) 𝑑𝑥 = − ∫ 2 𝑑𝑥
𝐸𝐼 𝐸𝐼 𝑙 𝑙 𝐸𝐼 𝑙
0 0 0

𝑃𝑏𝑎3
=
3𝐸𝐼𝑙2
For length B-C

𝑃𝑎𝑥2 𝑥2
𝑀= 𝑀1 =
𝑙 𝑙
𝑏 𝑏 𝑏
𝑀𝑀1 𝑑𝑥 1 𝑃𝑎𝑥2 𝑥2 1 𝑃𝑎𝑥22
∫ = ∫( ) ( ) 𝑑𝑥2 = ∫ 2 𝑑𝑥2
𝐸𝐼 𝐸𝐼 𝑙 𝑙 𝐸𝐼 𝑙
0 0 0

𝑃𝑎𝑏 3
=
3𝐸𝐼𝑙2

MM 1dx Pab 3 Pba 3

1  C =  = −
EI 3EI 2 3EI 2
=
Pab 2
3EI 2 (
b − a2 )
Pab
= (b + a )(b − a )
3EI 2
Pab
= (b − a )
3EI

Example 9.4

Find the deflection at point A of the cantilevers shown. EI is constant.

A C
B

a l-a

Solution

(1) Draw the bending moment for the loading system.

(2) Apply unit load at A and draw the bending moment diagram.
 MM 1dx
(3) Find 1. A =  EI
P
x1 x2 P(l-a)

A C
B

a l-a

VC = P

P(l-a)
-

Bending moment diagram for loading system

1
x1 x2 1.l

A C
B

VC = 1

l
-

Bending moment diagram for unit load at A

For length A-B

origin at A: M=0 M1 = -1.x1

a a
MM 1dx1 1
0
EI
=
EI 0
0dx = 0

For length B-C

origin at B: M = –Px2 M1 = –1(a+x2)

−a −a
MM 1dx1 1

0
EI
=
EI  ( Px )(a + x )dx
0
2 2

−a

 ( Px )
1
= 2
2 + Px 2 a dx
EI 0
−a
1  Px 23 Px 22 a 
=  + 
EI  3 2 0

=
P
6 EI

2(  − a ) + 3a(  − a )
3 2

P
= (  − a ) 2 (2 + a )
6 EI
P
1  A = (  − a ) 2 (2 + a )
6 EI

9.8 Tutorial

(1) Using the method of virtual work, determine the vertical displacement of joint E
for the truss shown in the figure below.

E = 200,000 MPa
Member areas (A) = 250 mm2 (for all members)
 25 kN 10 kN

F E D
20 kN

4m
B C
A
15 kN

2@3m=6m

Figure 1

(2) Using the principle of virtual work, calculate the vertical deflection of point D
due to the load shown in the beam in the figure below. EI =137500 kNm2.

5 kN/m

B D

A C

10 m 10 m 5m

(3) Using virtual work, calculate the vertical deflection of joint B of the truss shown
in the figure below due to the applied loading indicated in the figure. E =
200 GPa for all members; A = 100 mm2 for all members.
 H 6 G 8 F

7
1 5 9 11
3 13

A 2 B 4 C 10 D 12 E

50 kN 75 kN 50 kN

4 @ 3 m = 12 m

(4) Using the principle of virtual work:

• Determine the reactions. (Answer: VB = 6.25 kN; VA = 13.75 kN; MA =
22.5 kNm)
• Draw the bending moment diagram.
• Determine the vertical deflection (ΔC) under the point load. (Answer: ΔC =
39.48/EI)

20 kN

A
B

3m 3m
LEARNING UNIT 10:
PLASTIC THEORY OF STRUCTURES

Learning outcomes
After studying this learning unit, you should be able to

• use the plastic theory to analyse determinate and indeterminate structures,

including pin-jointed frames, beams and portal frames, and also to explain and
calculate
o elastic bending
o plastic bending
o plastic hinge
o shape factor
o behaviour of a portal frame under increasing load
o partial collapse in beams

10.1 Introduction

In this learning unit, we will demonstrate the plastic theory method using examples.

10.2 Elastic bending

In elastic bending the neutral axis passes through the centroid of the cross-section,
and the ultimate moment of resistance is reached when the extreme fibres reach the yield
stress, either in tension or compression. At this point the stress distribution is linear across
the cross-section and the strain diagram is also linear, as shown in the figure below.

G compression
a
n.A

tension
fy

Strain Stress distribution

 Figure 10.1

In elastic bending:
I
maximum moment of resistance Me = fy
y
where I = moment of inertia
fy = yield stress
y = extreme fibre distance
or Me = fy Ze

I
where Ze = elastic section modulus =
y

10.3 Plastic bending

If additional moment beyond Me is applied to the section, it is assumed that the extreme
fibres yield without taking additional stress, and the next layer of fibres becomes stressed
to the yield point.

As the bending moment is further increased, this process continues until the whole cross-
section is stressed to the yield stress. The neutral axis also shifts in the process.
When the whole cross-section is stressed to yield stress, the area above the neutral
axis must be equal to the area below, since, for equilibrium, total compression must =
total tension.

That is, the plastic neutral axis is an equal area axis

fy fY fY

C C C C

T T T T

strain fY fY fY fY
Assumed stress-strain (a) at Me (b) at >Me (c) at >>Me (d) at plastic yield
diagram
Figure 10.2

When the whole cross-section is stressed to yield point, as in figure 10.2(d), then the
moment of resistance is known as the plastic moment of resistance Mp, and the
corresponding section modulus is known as the plastic section modulus Zp.
That is, Mp = fyZp

10.4 Plastic hinge

Any increase in the applied bending moment beyond Mp will cause the cross-section to
yield and rotate. However, the moment of resistance of the cross-section remains Mp. A
plastic hinge carrying Mp has thus developed.

If a structure develops sufficient plastic hinges to become a mechanism, the structure

will collapse. Some elementary mechanisms are shown below.

(a) Simply supported beam (b) Propped cantilever (c) Encastre beam

Elementary mechanisms
Figure 10.3

10.5 Shear factor

For any cross-section, the ratio

Zp  Mp 
= 
Ze  Me 
is known as the shape factor s
plastic section modulus plastic moment of resistance
S= =
elastic section modulus elastic moment of resistance
Example 10.1: Find the shape factor of a rectangular section d  b

b b
fy fy

C C
C C
n a n
d a
T T T
T
fy fy
Elastic stress Plastic stress
distribution distribution
Elastic bending Plastic bending
(n.a through centroid) (n.a = equal area axis)
bd 3  bd 
I XX = C = fy  
12  2

bd 2  bd 
Ze = T = fy  
6  2

bd 2
Me = fy  d  d
6 M p = C  = T 
 2  2

bd 2
= fy
4

bd 2
Zp =
4

bd 2
Mp
Shape factors s = = 4 2 = 1,5
Me bd
6
Example 10.2: Find the shape factor for the I-section shown. If the yield stress for the
material is 300 MPa, calculate the fully plastic moment of the section.

fy = 300 MPa

200 200 fy
10
C1

C2
n a

10 10 T2

T1
10
fy

Elastic bending Plastic bending

I XX =
1
12
(
200  400 3 − 190  380 3 ) Zp = 2[(200 × 10 × 195) + (190 × 10 × 95)]
= 1141000𝑚𝑚3
= 197,86  10 6 mm 4 𝑍𝑝 1141000
S= = =1,153
197,86  10 6
𝑍𝑒 989300
Ze =
200 M𝑃 = 1141000 × 300𝑁𝑚𝑚
= 989300mm 3 = 342 kNm
M e = 989 300  300 Nmm
= 297 kNm

342
Shape factor s = = 1,15
297
For I-sections generally:

Shape factor s = 1.15

Example 10.3: Find the shape factor for the T-section shown. If the yield stress for the
material is 300 MPa, calculate the fully plastic moment of the section.

fy = 300 MPa

300 300
20 20 C1
C2
X X

15 T
15

Elastic bending Plastic bending

Find position of centroid Find position of equal area axis
 (300)(20)(440) + (430)(15)(215) 15a =
1
(300  20 + 430  15)
xy =
(300)(20) + (430)(15) 2
a = 415mm
= 323,4mm
15  (15) 15  ( 415)
2 2
300  ( 20)
3

I XX = + 300( 20)(116,6)
2 Z p = 300  20  25 + +
12 2 2
= 1 443 400mm 3
15( 430)
2

+ + 430(15)(108,4)
2

12 M p = 1 443 400  300  10 −6 = 433kNm

= 256 948 000 mm 4 Mp 433
s= = = 182
,
256 948 000 Me 238,4
Ze =
323,4 Zp 1 443 400
or s = = = 182
,
= 794 500 mm 3
Ze 794 500
M e = 794 500  300 Nmm
= 238,4 kNm

10.6 Behaviour of portal frame under increasing load

kV

kH
B C D Mp = 100 kNm

EI = constant
E
A k = load factor

5m 10m

Figure 10.4
Consider the above portal frame carrying loads V and H, to which a load factor k has
been applied. Assume V = H = 1.

Stage 1
An elastic analysis (by moment distribution or other method) gives the bending moment
diagram shown in Stage 1 of table 10.1. When k = 39, the largest bending moment, that
at E, becomes 100, and a plastic hinge forms. Apart from E, the whole structure is elastic
and remains so as k is increased above 39. When k increases, E rotates freely, but the
bending moment at E stays 100.

Stage 2
Shows the effective structure which resists the loads when k > 39. It is now the original
frame with a frictionless hinge at E. Elastic analysis of this structure gives the change in
bending moments. In order to get total moments, these changes must be added to the
bending moments when k = 39. The maximum moment is under the vertical load at point
C.
MC = 82,7 + 2,47 k1
where k1 = the change in k
When k1 = 7, MC = 100, and the new k = 46.

Stage 3
Shows that there are now two hinges in the effective structure. The third elastic analysis
is rapid. A further hinge forms at D when k = 47.

The process is continued in Stage 4.

When k = 50 a fourth hinge forms at A and the structure becomes a mechanism and
collapses.

The four stages are shown in table 10.1.

Note: The following two points about the bending moments in the right-hand column of
the table:
(1) Equilibrium condition

The distribution of bending moments is in equilibrium with the applied loads.

(2) Yield condition

Nowhere do the bending moments exceed the plastic moment of resistance Mp of the
members. Furthermore, at Stage 4, the structure becomes a mechanism and is at the
point of collapse.

(3) Mechanism conditions

There are sufficient plastic hinges for the structure to become a mechanism.

The load factors at the point of collapse is called the

COLLAPSE LOAD FACTOR kc
Table 10.1: Stages in the plastic collapse of a frame
0.78 2.08 Plastic Hinge
1 80.9
Stage 1 30.4
Whole structure elastic, until first plastic hinge
D - -
5m B + 1.14 2.12
forms at E.
C + 82.7
A + 2.57
E 44.4 + 100
5m 10m

x change in load factor Load factor = 39.0

Stage 2
0.14 2.34
31.4 79.3
Effective structure contains a frictionless hinge
at E. Second plastic hinge forms at C.
2.47 + 100
100
2.81 0
64.2
Frictionless Hinge
x change in load factor Load factor = 46.0
Stage 3 2.98 4.04 33.4 100
Effective structure has frictionless hinges at C
and E. Third plastic hinge forms at D.
100
39.4 0 66.8 100

x change in load factor

Load factor = 46.7
Stage 4 5 50
0
Frictionless hinges at C, D and E. Fourth plastic
hinge forms at A. Structure is a mechanism and 100
fails.
10 100
x change in load factor Load factor = 50
It is possible to visualise another plastic collapse mechanism for the frame, for
example:
B D
k
C 100
E
A
300
Resulting bending moment diagram
Figure 10.5

The mechanism shown in figure 10.5 would occur at k = 80 (> kC = 50) and gives rise
to a bending moment diagram which satisfies the equilibrium and mechanism
conditions, but which violates the yield condition, because MC = 300 > 100. The value
of k here is called the upper bound condition.

10.7 Theorem of plastic analysis

(1) Lower bound theorem

If, in a structure subject to loading defined by a positive load factor k, a bending

moment distribution satisfying equilibrium and yield conditions can be found, then k is
less than or equal to the collapse load factor kC.

(2) Upper bound theorem

If, in a structure subjected to loading defined by a positive load factor k, a bending

moment distribution satisfying the equilibrium and mechanism conditions can be
found, then k is greater than or equal to the collapse load factor kC.

(3) Uniqueness theorem

If a structure is subjected to loading defined by a positive load factor k, such that the
resulting bending moment distribution satisfies all three collapse conditions, then k =
kC.
That is, if all three collapse conditions are satisfied, that state of collapse represents
the “worst case” collapse condition.

10.8 Number of hinges required in a mechanism

In general, the number of hinges required in a collapse mechanism is one more than
the degree of redundancy
n=r+1

However, there are two exceptions to this rule.

(1) Partial collapse occurs when part of the structure collapses, leaving the
remainder intact, for example, in the portal frame the beam section could
collapse leaving both columns intact. This happens with just three hinges in
place of the four required for a normal mechanism
n<r+1 (3 < 3 + 1)

H V

Figure 10.6: Partial collapse

(2) Over-collapse occurs by coincidence owing to the geometry of the structure

or the loading when two plastic hinges occur simultaneously. In the portal
frame, over-collapse could occur by a combination when the fourth and fifth
hinges coincidentally occur at the same time.
n > r + 1 (5 > 3 + 1)
 H V

Figure 10.7: Over-collapse

10.9 Number of hinges required in a mechanism

Consider the two-span continuous beam shown below. Two collapse modes may
occur: partial collapse when n = 2 and normal collapse when n = 3.

r = 2

n = 2 Partial collapse

n = 3

Figure 10.8: Collapse mechanisms in continuous beams

10.10 Static method (free and reactant bending moment diagram

method) for finding collapse loads 𝐖𝒄
The collapse load is calculated if the position of the plastic hinge is known and if at this
point the bending moment is MP. The actual bending moment in an indeterminate
beam can be obtained by combing the free and reactant bending moment, that is,

Actual BM = Free BM + Reactant BM

The procedure of analysing a structure using the static method is as follows:

1) Select redundant(s).

2) Draw a moment diagram for the determinate structure.

3) Draw a moment diagram for the structure loaded by redundant(s).

4) Sketch the composite moment diagram in such a way that a mechanism is

formed (sketch mechanism).

5) Compute the value of ultimate load by solving the equilibrium equation.

(1) Simply supported beam

Example 10.4

Consider the following simply supported beam with fully plastic moment MP. It is
required to calculate the collapse load.

Solution
The beam is statically determinate n = 1, therefore, only one plastic hinge will form at
collapse. Assume that the plastic hinges form under the point load.

A WC
A B
B

l/2 l/2 l/2 l/2

(a) Beam (b) Collapse mechanism

 +
WCl/4

(c) Bending moment diagram

Figure 10.9 (a), (b) and (c)

Thus the collapse load is calculated as:

Wc l
Mp =
4
4M p
 Wc =
l

(3) Beams with fixed-end supports

Example 10.5
Calculate the value for the collapse load for the fixed-end beam shown below.

Solution
The beam is statically indeterminate to the second degree, therefore, three plastic
hinges will form at collapse. It is assumed that the plastic hinges will form at the
restrained ends and under the point load.
r=2 n=2+1 that is, n = 3

Wab2/l2 Wa2b/l2
A B -
C -
+
a b A B
2Wa2b2/l3

(a) Beam (b) Elastic BM diagram

 -
Wab/l +

A B
(c) Free BM diagram (d) Reactant BM diagram

MP MP
WC
A
B -
-
Wab/l
+
a b A B
MP

(e) Collapse mechanism (f) = (c) + (d)

Wc ab 2M p l
2M p = Wc =
l ab
Figure 10.10

General rules in determining the positions of plastic hinges

It can be seen that if the positions where the plastic hinges will form can be forecast,
the collapse and working loads can easily be calculated. In the previous two examples,
hinges occurred at points of maximum moments. This is not always the case and the
following general rules will prove very valuable in determining the position of plastic
hinges.

Rule 1 Plastic hinges occur at built-in supports.

Rule 2 Plastic hinges form under point loads.
Rule 3 Plastic hinges form at internal supports of continuous beams.
Rule 4 Plastic hinges form where the beam changes cross-section.
Rule 5 Where the beam changes section, the hinge will form in the weaker
section.
Rule 6 At a support, the plastic hinge forms at the plastic moment of the weaker
member.
Rule 7 For continuous beams consider partial collapse span by span.
(3) Propped cantilever

Example 10.6
Consider the following propped cantilever with fully plastic moment MP. It is required
to calculate the collapse load.

Solution
The beam is statically indeterminate to the first degree, therefore, two plastic hinges
will form at collapse. It is assumed that the plastic hinges form at the support and under
the point load.
r=1 n=1+1=2

WC
MP

A B - bMP/l
A + B
a b MP
l MP

(a) Beam (b) = Free BM + Reactant BM

Figure 10.11

Actual BM = Free BM + Reactant BM

Wc ab bM p
− =Mp
l l
M p (a + 2b )
Wc =
ab

(4) Continuous beams

In continuous beams the following two points are important:

(a) Each span may have a different plastic moment. The plastic hinge forms at the
plastic moment of the weaker member.
(b) It is unlikely that every span will fail together. Each span must be checked
individually, and the span, or spans, with the lowest collapse load determines
the collapse of the whole beam. Hence, for a continuous beam you should
consider the partial collapse span-by-span.

10.11 Load factors

Plastic analysis gives the values of the loads at collapse, based on the plastic
moments of resistance.

Working loads are obtained by dividing the collapse loads by a factor known as the
load factor.

Collapse load
Load factor k =
Working load

Example 10.7
Calculate the collapse load WC for the continuous beam shown in the figure below and
draw the bending moment diagram at collapse. The plastic moment for each span is
given.
WC 1.5WC WC

A B C D

MP=600 kNm MP=900 kNm MP=600 kNm

4m 4m 6m 6m 4m 4m

Solution 600 600

- -

+ hinge in +
+ hinge in
weaker
member weaker
member
2WC 2WC
4,5WC
(moment in (moment in
s.s. span AB) (moment in
s.s. span CD)
s.s. span BC)
 Bending moment diagram at collapse

(a) If A-B and C-D collapses first

600

+
2WC = 900 kNm
600
WC = 450 kN

(b) If B-C collapses first

600 600

4.5 WC = 900+600=1500 kNm WC

+ = 333 kN

This is clearly the critical case

900

4.5 WC = 900 + 600 = 1500 kNm WC = 333 kN

This is clearly the critical case.

(c) Final collapse condition

333 kN 500 kN 333 kN

A B C D

Mechanism

600 600
667
667
- -
+ +
+
367 367

900
 Bending moment diagram at collapse

10.12 Beam spans with distributed loads

The position of the plastic hinge is not immediately obvious (unless loading and
geometry are symmetrical).

The plastic hinge occurs where the resultant bending moment is a maximum

dM
That is, where =0
dx
dM
But =N = shear force = 0
dx

Consider the following fixed-end supported beam with uniform loading but with a
change in cross-sectional area along its length (l).

< l/2

wC kN/m

2MP MP
A
B
A and B fully fixed. l
x

Plastic bending 2MP MP

moments
- -
+
A B
2MP

Figure 10.12
Now construct a free body diagram, cutting the beam at the span hinge point, that is,
at the point of maximum bending moment (and where shear force = 0).

< l/2
2MP
2MP wC kN/m 2MP
wC kN/m
Zero shear MP
at cut
A
l-x x B

Figure 10.13: Free body diagram

For right-hand free body, taking moments about B

wC x 2
− + 2Mp + Mp = 0
2

 wCx2 = 3Mp Equation (1)

For left-hand free body, taking moments about A

wC (l − x )
2

- 2Mp - 2Mp = 0
2

wC (l − x )
2

 = 4Mp Equation (2)

2

4
 Equation (1) gives:
3

4  wC x 2 
  = 4Mp Equation (3)
3 2 

wC ( L − x)
2 2
4  wC x
Equation (2) – Equation (3) =
2 3 2
 x2 + 6xl – 3l2 = 0
Solving gives x = (-3 ± 23)l
that is, x = 0,464 l
27,87Mp l
which gives wC = 2
for x <
l 2

10.13 The virtual work method for finding collapse loads

Failure in a portal frame can occur in one of three modes:

(1) Sideways sway due to horizontal forces – If the horizontal force (H) is large and
vertical loading (V) is small, then sway mechanism would be expected.

(2) Individual failure of a beam due to vertical forces – If V is large and H is small,
then a beam mechanism is more likely.

(3) A combination of (1) and (2) – If H and V are comparable in magnitude, then a
combined mechanism is expected.

The method of virtual work is based on two premises:

Premise 1 When a framed structure collapses, all deformation of the structure

occurs by rotation at the plastic hinges.

Premise 2 The principle of virtual work can be applied to these rotations.

VC  

HC 

2Mp
 
Mp Mp 
 

(a) Point of collapse (b) Small horizontal displacement

just before rotation at top of left column. Position
just after rotation

Figure 10.14
External (virtual) work done by applied loads = H + V
=  W
(over all loads)

Internal (virtual) work done = Mp + 2Mp (+)+Mp

= Mp
(over all plastic hinges)

But external (virtual) work = internal (virtual) work

 W = Mp

Further assumptions are:

Premise 3 At collapse the bending moments remain constant as the structure

deforms.
Premise 4 All axial load effects are ignored.

In the following sections, various structures will be analysed using the virtual work
method.

10.14 Fixed-end support beam (using virtual work)

WC

Mp
L L
/2 /2

-Mp +Mp -Mp Stage 1: Collapse mechanism

  
Stage 2: Virtual plastic rotations at hinges
- -

Figure 10.15
Stage 3 External work done = WC
Stage 4 Internal work done = –Mp (-) + Mp (2) – Mp (-)
= 4 M p

(Note: All work absorbed is positive.)

Stage 5 Geometry of mechanism

L
/2
L
= 
 2


Stage 6 Done = Work absorbed

Work WC = 4 M p
8M p
That is, WC L/2  = 4 M p  WC =
L

10.15 Portal frame with pinned bases

Consider the following portal frame:

Stage 2 Virtual plastic rotations at hinges

 VC
l/2

B MP
HC C

MP
MP

A
D

Figure 10.16

There are three collapse mechanisms:

VC
 VC  VC
HC   HC  
HC
B C
2  
   2
A D
D 

(a) Beam collapse (b) Sway collapse (c) Combined collapse

Figure 10.17

(a) Beam collapse

External work = Internal work

VCL/2 = Mp ( + 2 + ) = 4Mp

 Vl = 8M p Equation (1)
(b) Sway collapse

External work = Internal work

Hch = 2Mp

 Hh = 2M p Equation (2)

(c) Combined mechanisms collapse

Note: There is no plastic rotation at B.

(r = 1, n = 1 + 1 = 2)

External work (beam mechanism) + External work (sway mechanism)

= Internal work (beam mechanism) + Internal work (sway mechanism) – Internal
work at hinge which falls away.

VC L
+ Hch = 4Mp + 2Mp – 2Mp
2

2 Hh + Vl = 8M p Equation (3)

Interaction diagram of the three equations

It is frequently necessary to predict the mode of failure and to calculate what relative
magnitudes of HC ad VC will give rise to the different types of failure mechanism. This
can be deduced by constructing an interaction diagram.

The “most likely” mechanism depends on the relative values of VC and HC.

Plot equations (1), (2) and (3) on a graph with axes Vc l and H c h . This graph is called

the interaction diagram.

 Interaction diagram
9
A
8

Collapse
6 [Combined]
VC l [MPa]

5
Area within B[Combined
4
which frame or Sway]
will not collapse
3

1
Collapse [Sway]
C
0 0
0 0,5 1 1,5 2 2,5 3 3,5 4
HCh [MPa]

Combined mechanism Sway mechanism

Beam Mechanism V=2H

Figure 10.18: Interaction diagram of a portal frame with pinned bases

From the interaction diagram (figure 10.18), the following can be noted:

• At point A on the diagram (HC = 0 and VC= 8 Mp), collapse will be by combined
or beam mechanism or both (over collapse).

• On the boundary between points A and B, collapse will be by a combined

mechanism.

• At point B collapse would be by a combined or sway mechanism or both

(again over-collapse).

• While between B and C collapse would be by a sway mechanism.

If the values of h, l, and Mp are known, then the corresponding values of any
combination of HC and VC which will collapse together with the type of collapse
mechanism can be determined from the diagram. If the values of VC1 and HCh are
restricted so as to remain within the hatched area on the diagram, then the frame is
safe against collapse.

10.16 Fixed-base portal frame

Example 10.8

(a) Construct the interaction diagram for the fixed-base portal frame shown below.

(b) If V = 1.25 H what type of failure mechanism would you expect and what values
of V and H would occur?

(c) Draw the bending moment diagram for the frame at collapse.
VC
4m

B MP = 390
HC C
MP = 210

3m
MP = 210

D
5m

10 m

Figure 10.19

The solution to the above problem is carried out using steps (i) to (iii) next.

(i) Using virtual work theory, compute the loads (Vc and/or Hc) that would cause
the collapse of the portal frame for each of the three collapse mechanisms.
There are three collapse mechanisms

VC
  VC  VC
HC   HC HC  
C
B  
  D  
 
A 

(a) Beam collapse (b) Sway collapse (c) Combined collapse

Figure 10.20

(a) Beam collapse

2
From geometry:  = 
3
At joint, plastic hinge forms at Mp of weakest member.
External work = Internal work
VC = 210 + 390 (+ ) + 210
= 600 ( + )
2
VC.4 = 600 (1 + )
3
VC = 250 kN

(b) Sway collapse

From geometry:  = 5 = 3
5
= 
3
External work = Internal work
HC = 210 x 2 + 210 x 2
5
HC x 5 = 420 (1 + )
3
HC = 224 kN
(c) Combined collapse

External work (beam) + External work (sway) = Internal work

2 5
VC x 4 + HC x 5 = 600 (1 + ) + 420 (1 + ) – 210  – 210 
3 3
4 VC + 5HC = 1700

(ii) Plot the interaction diagram.

Note: When plotting the interaction diagram in this case values of h an l are known
and fixed. Therefore, it is not necessary to plot HCh and VCl as h/l cannot vary and the
actual values for HC and VC may be plotted directly.

Interaction diagram
450

400

350

300

250 A
VC [kN]

HC=170 kN
VC=[] kN
200

150

100

50

0
0 50 100 150 200 250 300 350 400
Hc [kN]

Combined mechanism Sway mechanism

Beam Mechanism V=1.25H

Figure 10.21: Interaction diagram

The portal frame will fail by combined mechanism marked by point A on the graph,
with:
HC = 170 kN and VC = 212.5 kN

(iii) Draw the bending moment diagram

The bending moment diagram at collapse for this combination of loads is:
212.5 kN
60 kNm 210 kNm

B - -
170 kN 210 kNm
60 kNm C -
+
390 kNm

+ D
140 kN
-

A 100 kN
210 kNm
30 kN
10 m

112.5 kN
Figure 10.22: Bending moment diagram (kNm)

10.17 Uniformly distributed load on portal frame beams

If a plastic hinge forms under the centre of a uniformly distributed load, the external
work done by the uniformly distributed load is
= 1/2 wL 
= 1/4 wL2 
 w kN/m
 

 
L
/4 L
/4

L
L /2 L
/2
External work done
= 2[wL/2 ] = 2[wL/2  L/4] = wL2/4

Figure 10.23

Example 10.9

Calculate the hinge moment at collapse for a load factor of 1.75 in the portal frame
loaded as shown.
30 kN/m
B C
10m
3m
50 kN

3m
A D
Solution

Load factor = 1.75. Multiply working loads by 1.75 to obtain collapse loads.

30  1,75 = 52,5 kN/m

50  1,75 1 redundancy

= 87.5  2 hinges

Collapse mechanisms

52,5 kN/m 52,5 kN/m

 C
B

2
87,5  87,5 
  



(1) Beam (2) Sway (3) Combined

(1) Beam mechanism

External work done = Internal energy absorbed

1
(52.5  10) 5 = Mp  4
2
which gives Mp = 328 kNm

(2) Sway mechanism

External work done = Internal energy dissipated

87.5  3 = Mp  2
giving Mp = 131.25 kNm

(3) Combined mechanism

In the combined mechanism Mp does not develop at midspan of B-C.

10 - x x

 



 = (10 – x)  = x

 10 − x 
  = 
 x 
10 − x 10
=+=+ =
x x
Virtual work equation

External work done = Internal energy absorbed

10 10 − x  
(52 ,5  10)(10 − x) + 87 ,5  3 = M p  + 
1
  + 
2  x x  

10 + 10 − x + x 
2625 – 262.5x + 262.5 = M p  x 

20
2887.5 – 262.5x = M p 
x

which gives Mp = 144.375x – 13.125x2

dMp
Mp is a maximum at =0
dx

that is, 26.25x = 144.375 that is, x = 5.5 m

That is, plastic hinge develops at 5.5 m from C.

 Mp = 144.375 (5.5) – 13.125 (5.5)2

Mp = 397 kNm

 Hinge moment at collapse = 397 kNm

10.18 Examples

Example 10.10

(a) A fixed-end beam shown in the figure has to be designed plastically with a load
factor of 2. Draw a dimensioned bending moment diagram for the beam at
collapse.
 (c) Calculate the plastic moment of the beam cross-section shown.

Yield stress = 250 MPa.

100 mm

20 kN 40 kN
10 mm

400 mm
4m 4m 4m
10 mm

20 mm

150 mm

40 kN 80 kN

B C Load factor = 2
A D
4m 4m 4m  Double point loads

Two redundancies

RA = (53,3)  2 + 1 = 3 hinges for mechanism

RD = (66,7)

Mp Mp

=(133,40) (=133,4) Collapse BM Diagram (kNm)

Mp
(80,0)
(=133,4)

Solution

Load factor = 2

 Double point loads

Two redundancies

 2 + 1 = 3 hinges for mechanism

Collapse bending moment diagram (kNm)

Moments about C from right: Mp = RD  4 – Mp Equation (1)

that is, 2Mp = 4RD

Moments about C from left: Mp = RA  8 – Mp – 40  4

Equation (2)
that is, 2Mp = 8RA - 160

From Equation (1): 4RD = 8RA – 160

Equation (3)
that is, RD = 2RA – 40

But RA + RD = 120
Equation (4)
that is, RD = –RA + 120

[Equation (3) – Equation (4)]: 0 = 3RA – 160

160
RA = = 53.3 kN
3
RD = 120 – RA

RD = 66.7 kN

From Equation (1): Mp = 2RD

that is, Mp = 133.4 kNm

MB = 53.3  4 – 133.4 = 80.0 kNm

The values can now be filled in on the bending moment diagram.

(b) 100 mm

10 a
n
Stress
diagram

20

150 mm
Position of neutral axis (equal area axis)

(20  150) + (370  10) + (100  10)

(20  150) + (10y) = giving y = 85 mm
2
Zp = (10010290)+(28510142.5)+(851042.5)+(1502095)
= 1017  103 mm3

250  10 3
Mp = fyZp =  1017  103 = 254.3 kNm
10 9
Mp = 254.3 kNm

Example 10.11

(a) Define, with the aid of sketches, the concept of a plastic hinge.

(b) A beam with the cross-section as shown in the sketch spans 6 m. The ends of
the beam are fixed. It safely carries a uniformly distributed load of 10 kN/m over
the whole span as well as a central point load of 40 kN. Using plastic theory
and a yield stress of 250 MPa, determine the load factor for the beam.

40 kN
20 mm
10 kN/m
10 mm

3m 3m
20 mm

100 mm
Solution

fy fy fy
(a)

fy fy fy
(1)Elastic stress distribution (2) Semi plastic (3) Plastic stress distribution

At ultimate elastic moment, Moment increased, some All fibres in plastic zone, as
the top and bottom fibres fibres in plastic zone, some moment increases to Mp.
just reach yield stress fy. in elastic zone due to Now rotation starts taking
stress-strain relationship. place with increase in
moment.

k  40 kN k  10 kN/m
 

2
3m 3m

Let load factor be k: Loading becomes as shown

Two redundancies: Collapse mechanism as shown

Zp = 2 [ (100  20  90) + (80  10  40) ]

= 424 000 mm3
Mp = fy Zp = 250  424 000 Nmm
= 106 kNm

External work done = Internal energy absorbed

2 [10k  3  1,5] + 40k (3) = Mp  4
210k = 106  4
 4  106
k= = 2.02 Load factor = 2.02
210

Example 10.12

(a) The figure (a) shows the symmetrical cross-section of a rectangular hollow
beam. Determine the plastic moment of the section about the horizontal axis
and show that its shape factor is 1.4. The yield stress is 250 MPa.

(b) The normal working loads of a beam are shown below. The beam has the cross-
section shown in figure (a). Calculate the load factor.

100 kN 100 kN

3m 3m 2m 4m

50

100 mm
(a)

(a) Zp = 2 [ (100  37.5) (37.5 + 37,5 2 ) + (37.5  50  37,5 2 ) ]

= 2 (210 937.5 + 35 156.25)

= 492 187.5 mm3

250  10 3
Mp = fyZp = (492 187.5) = 123.05 kNm
10 9
I 1
Me = fy I= [100  (150)3 – 50(75)3] = 26.367  106 mm4
y 12

250  10 3  26,367  10 6 
=   = 87.89 kNm
10 9  75 

Mp 123,05
Shape factor = = =1.41
Me 87,09
(b)
100 kN
100 kN

B 2m 4m C
A
3m 3m

Failure mechanisms

W
B C
2m 4m
 /2
3/2

(1) B-C collapses first

External work done = internal energy absorbed

3
W(2) = Mp ( + )
2
5
2W = 123 ( )
2
W = 153.75

153,75
Load factor = = 1.54
100

3m 3m
A  
B
2

(2) A-B collapses first

 External work done = Internal energy absorbed
W(3) = Mp (4)
3W = 123 (4)
W = 164 kN

164
Load factor = = 1.64
100

For design use the smaller load factor

Load factor = 1.54

Example 10.13

Calculate the safe value for the load W when it is applied to the portal frame in the
figure, if all members of the frame are made of 533 x 210 x 93 I-section of which the
yield stress is 250 MPa. Use a load factor of 2.

4m
3m

6m
Solution

2W 4m

Load factor =2
3m
6m  Double load W

 

2W Collapse mechanism

  = 2  = 3 = 6

  = 2


Virtual work equation

External work done = Internal energy absorbed

2W (6) = Mp ( + 2)

 Mp = 4W

From section tables Zp = 2366  103 mm3 for 533  210  93 I

250  10 3
Mp = fyZp =  2366  103 = 591.5 kNm
10 9

 4W = 591.5

W = 147.9 kN
10.19 Tutorial

(1) Analyse the portal frame shown in the figure according to the plastic theory
and determine whether the cross-section shown would be adequate for the
members of the frame. Use a load factor of 2 and a yield stress of 250
MPa.

14 kN 200 mm

35 kN
20

4m
3m

2m 2m 20

(2)
(a) Show, from first principles, that the fully plastic moment Mp of a mild steel beam
of rectangular cross-section is 50% greater than the elastic moment Me, which
develops when the beam reaches its limit of elasticity.

(b) An I-section, 250 mm deep and 125 mm wide, is made of mild steel. The
thickness of the web and the flanges are 12 mm and 20 mm respectively. Find
the ratio of the fully plastic moment and the elastic limit moment for the section.

(3) The cantilever shown in the figure is to be designed plastically with a load factor
of 2. Length b of the beam has no flange plates, but the remainder of the beam
has flange plates as shown in the cross-section. Determine the values required
for a and b. Yield stress = 320 MPa.
 a
40 kN

20 mm
b
4m

100 mm

(4)
(a) Determine the plastic moment for the cross-section shown in the figure. Yield
stress = 250 MPa.

(b) What is the value of the shape factor for this section?
100 mm

20 mm

(5)

(a) A 533 x 210 x 109 I-section acts as a fixed-end beam over a span of 6m. It
carries a uniformly distributed load of 20 kN/m. Calculate the magnitude of the
central point load that can still be safely placed on the beam. Use the plastic
theory.
Load factor = 2; yield stress = 250 MPa.
(b) Calculate the shape factor for the cross-section shown in the figure.

100 mm

20 mm

(6) A beam with a cross-section as shown in the figure is simply supported over a
span of 4 m. Together with plastic theory principles use a load factor of 2 and
a yield stress of 250 MPa to answer the following questions:

(a) What is the safe uniformly distributed load that the beam can carry?

(b) Over what length of the beam have fibres become plastic when the beam is at
the point of collapse?

(d) What safe uniformly distributed load can the beam carry if the ends are fixed
instead of simply supported?
160 mm

10 mm
(7) Analysing various collapse mechanisms and using a load factor of 2,
determine the maximum plastic moment occurring in the portal frame shown in
the figure. Using a yield stress of 250 MPa, determine whether the I-section
shown will be able to carry this moment. What is the shape factor of this
section?
100 m

40 kN 20 kN/m
10 mm
B C

10 mm
A D
D 10 mm

4m