of

Structu

I-,. ,
 CHAPTER I

DEFLECTIONS OF STRUCTURES

1.1 Introduction
A rigid body moves slightly when it is subjected to load or strained
due to any other reason such as change in temperature, shrinkage or creep
of the material of construction.
The displacements of variou s points from their original positions are
called dejlectioTls. In a broader sense, deflections comprise linear displa-
cements as well as angular rotations.
Resistance of a structural element to deflection is termed stiffness.
Stiffness of a structure is of no less importance than its strength. This is
because an excessive deflection of a structure may impair its proper fun-
ction. For example, excessive deflection of floor beams may cause cracking
of the plaster and produce drainage problems and excessive deflection
of a frame in a building may damage the exterior curtain walls and
inte rior partitions.
Not only is a knowledge of the deflections important for reasons as
those mentioned above, but also because it forms the basis of the analysis
of statical1y indeterminate structures. 'When the number of equations
of eq uilibrium together with any condition equations available from
particular datails of construction (section 2.4" Part I) is not sufficient
to determine all the external reactions of a structure and the internal forces
and moments in its component members, resort is made to compatibility

I
conditions . These conditions which are based on displacements and
rotations provide the additional number of equations required for the
analysis.

I
There are several methods for computing deflections. Of these
methods" the following are the most useful a nd widely used and therefore
will be considered in detail in this chapter.
I. The double integration method.
2. The moment-area method.

r 3. The elastic-load method.

4 . The conjugate beam method.

~
2

5. The method of virtual work

6. Graphical methods for truss deflections (Williot and Williot-Mohr
diagrams).
As will be explained subsequently, some of these methods are only
suitable to calculate one particular deflection component at an individual
point, while others may be used to calculate deflections at several points
simultaneously. Each method nas its limitations, advantages and disad-
vantages. The student should therefore be thoroughly familiar with as
many methods as possible and should understand which method is best
suited to a particular problem.
For instance, the first four methods are used to determine deflections
of structures whose members are mainly subjected to bending strains such
as beams and frames. The method of virtual work is used to determine
deflections due to all types of strains; bending, axial and shear as well as
those due to no-load causes. The graphical methods in (6) are applicable
only to trusses.

Part I
Classical Methods of BeaDl Deflections

1.2 The double integration method

It is customary to consider downward deflection at a point on a beam
as positive. Thus, if the axis of the beam is taken as the x-axis and the
beam deflection as the y co-ordinate as shown in Fig. 1.1 . . then the defle-
cted shape of the beam OT, as it is commonly called, the elastic curve of
the beam can be expressed as follows

+x

Elastic curve

+y
Fig. I.l

From calculus, the expression for the curvature is

-
R
I
= ~/[l
dx2
+ (~.)2]~
\dx
J.l
 3
cly
where R is the radius of curvature at any point on the ;beam, y and
dx
are the deHcction and the slope of the elastic curve at that point respect-

ively. In most practical cases the slope cly -Y

- is small and hence (cl
clx dx
)2
is very small and can be neglected in comparison with unity. Thus
equation 1.1 reduces to :
cl 2y
1.2
R clx 2

From equation 8.12 (Part I), the expression relating the curvature
to the applied moment is :
I M
1.3
R EI

Combining equations 1.2 and 1.3,

cl 2 y _ M
1.4
clx 2 - EI

The solution of the differential equation 1.4 is the classical method

of studying beam deflections. If the bending moment M is expressed in
terms. of x, then the slope of the elastic curve at any point on the beam
is found by integrating equation 1.4 once and the deflection is obtained
by integrating it twice. Before doing so, however, it is necessary to define
the sign conventions in order to decide whether the positive or negative
signs that appear in equation 1.4 may be adopted. Following the usual
sign convention for the moment and considering downward deflection

positive, it is seen from Fig. 1.1, that with positive M, dy decreases as

dx
cl 2y
x increases and thus - - is negative. Similarly, it may be shown that
clx 2
cly cl 2 y. ..
with negative M, increases with x and hence dx 2 IS pOSItIve. In
dx
cl 2 y
other words, --2 is always opposite in sign to M. Thus, the negative
dx
sign is used and equation 1.4 becomes:
M
1.5
EI

r
4

By carrying out the integration of equation 1.5 in two steps, the

expressions for slopes and deflectiDns of a beam can be obtained. The
resulting constants of integration arc determined from either the displa-
cement boundary conditions or from the requirements of continuity of
the elastic curve. Displacement boundary conditions refer to the known
conditions of deflection and slope at the supports. For example, the de-
flection at a roller or a hinge support is zeru, and both deflection and slope
at a fixed support are zero. On the other hand, continuity of the elastic
curve means that there is no sudden change in eith~r the deflection or
slope at any point along the curve .
The product' EI of the modulus of elasticity E and the moment
of inertia I is called the bemfing rigidity.
1.3 Applications to the douhle integration lIlethod
Several of the classical problems of beam deflections will now be solved
to illustrate the application of the double integration method. As will
be seen later, some of the given problems can be solved in a much simpler
manner by other methods. Nevertheless, they arc given on purpose to
point out the relative ease or difficulty of the method and show
its limitations.

Exalllple 1.1 A simply supported beam of constant hending rigidity EI

carries a uniformly varying load as shown in Fig. 1.2a. Determine the
slope and deflection equations, and find the position and value of the
maximum deflection.

(a)

O.S19L I
(Ll

Fig . 1.2

Solution: The reactions arc first found and are as indicated in Fig. 1.2a.
The bending moment at any section along the beam at a distance x from
a is given by :
wLx wx 3
M - - - - --
6 6L
 5

Applying equation 1.5,

d'y I (WLx WX 3)
dx 2 = - E1 -6- - 6L

Integrating twice,
4
dy
--- -1 (WLx2
-- wx
- -
dx E1 12 24L
I (wLx3 wx' )
Y = - E1 \ 3s - 120L + Ax + B

The two boundary conditions needed to determine the constants A and

Bare:
y = 0 at x = 0 and y = 0 at x = L
1
o= - E1 ( 0 - 0 +0+ B) and

0 = - - 1 (WL4
-- - - wL'-+AL+B)
E1 36 120L
7wL3
B = o and A = -
360

wx 4
dy
Therefore - - = - - 1 (WLx2
-- - - - - 7-WL3)
-
, dx E1 12 24 L 360

The maximum deflection,) Yrna;V occurs at the point where t he slope

. dy.
oftee
h l ashe curve, - - J IS zero.
dx
4 7 WL3)
-1 (WLx2
dy wx
- =0 -- - - --- -
dx E1 12 24 L 360
x = 0.519L.

Substituting this value of x in .the expressIon for y,

0.0065 wL4
Ymp = E1
The elastic curve of the beam is shown in Fig. 1.2b.
i
,
,-

. batnple 1.2 A cantilever of constant bending rigidity EI carries a COD-

centrated end load as shown in Fig. 1.3a. Determine the slope and deflec-
tion equations, and hence find the slope and deflection at the free end.

Fig. 1.3

Solution: Taking the origin at a, the bending moment at a section at a

distance x from a is given by :
M = - P (L-x)
Applying equation 1.5,
dOy P
- = - (L-x)
dx 2 EI

Integrating twice,

dy
dx
=.!.... (Lx
EI
_ x' +
2
A)
(~.
3
y = :1 - x6 + Ax + B)
The two boundary conditions needed to determine the constants A and
B are :
dy
- - - o and y = 0 at x = 0
dx
P P
o -
EI
(0 - 0 + A) and 0= -
EI
(O-O+O+B)

A OandB = O

Therefore, dy =...!:...
dx EI
(Lx _ ~)
2

y = :1 (~2 x; ) _
 7

At the free end x = L then,

a" =( dY)
dx b
= ~ (L' _
EI
L')
2
= PL'
2EI

Yb =...! (L3 _ L3) = PL3

EI 2 6 3 EI
The elastic curve of the beam is shown in Fig. D.3b.

Elaunple 1.3 A simply supported beam of constant bending rigidity EI

carries an end moment Mo as shown in Fig. 1.4a. Find the expressions
for the slope and deflection. Locate the position of maximum deflection
and calculate its value. Also calculate the slopes at the supports.

~)~f--_M. _ __
L
- ----14\
(b)

Fig. 1.4

Solution: The reactions are first found and are as indicated in Fig. 1.4a.
The bending moment at any section at a distanc... x from a is given by :

M=-(Mo- M:)
Applying equation 1.5,

d2 _1_ (M _ M.,x)
y
dx'
=
EI 0 L
Integrating twice,

~ = ~I
(M.,x - ~~2 + A)
y
1(M.,x2 M.,x3
=lli -2--61:"""+ Ax +B
)
The constants A and B are found from the two boundary conditions :
y = 0 at x = 0 and y = 0 at x = L
8

2
I
O=-(O-O+O+B)andO= - I (MoL L2 +AL+B)
- - _ -M o-
EI EI 2 6
MoL
B o and A= - - -
3
dy
Therefore, - - _
dx
_1_ (M.x
EI
Mox2 _
2L
MoL)
3
_ _ I (MoX2 _ Mox3 _ MoLX)
Y - EI 2 6 L 3
The maximum deflection, Ymax ' occurs at the point where the slope

of the elastic curve, dy ,is zero. Thus,

dx
dy = 0 = _ I (M x _ Mox2 _ MoL)
dx EI 0 2L 3
x = 0.423 L
Substituting this value into the expression for y,
- 0.064 MoL2
Ymax = - EI
The negative sign indicates upward deflection.
It will be interesting to compare this value of maximum deflection
with the value of the deflection at the centre, Yo . Yo is found by ~ubstitu
ting L/2 for x in the expression for y.
MoL 2 0.0625 MoL2
Yo = - l6EI =- EI
It is noticed that in this particular case, the maximum deflection occurs
at a point at a distance 0.077 L from the centre, and the difference bet-
ween the maximum deflection and that at the centre is only about 2%.
This is characteristic of simply supported beams';.e. regardless of the load
system on the beam. . the maximum deflection is always approximately
equal to that at the centre, and never occurs further than 0.0774 L either
side of the centre.
The slopes at supports a and b are fo und by substituting x = 0 and
x L in the expression for the slope.

a

=
( dx
dY)
--
a
= - --
MoL
3 E~
and a" = (~)
dx b
-
MoL
6EI
The elastic curve of the beam is S!l(lWn in Fig. l.4b.
 9

Example 1.4 A simply supported beam of constant bending rigidity EI

carries a concentrated load P as shown in Fig: 1.5. Find the expressions
for the slope and deflection. Locate the position of maximum deflection
and calculate its value.

0
(I b

Pb 0 I b Po
L L L

Fig. 1.5

Solution: The reactions are first found and are as indicated. Unlike the
preceding examples, the bending moment has two different equations
to apply to parts ac and be of the beam. Thcrcfore~ moment equations are
found for each part and the integrations arc carried o ut separately. Not~
that in writing the moment equations the same co-ordinate axes arc used
although a solution is also possible by taking two origins, onc at each end.
of the beam, but this may lead to confusion in the signs adopted.
Part ac.. x <a
Pb
M= x (a)
L
d 2y Pbx
- - = (b)
dx 2 LEI
dy Pbx 2
-- = +A (e)
dx 2LEI
Pbx'
y - - - - + Ax+B
6LEI
(d)

Part cb, x > a

Pb
M = --x - P (x - a) (e)
L
2
-d y = - - I [PbX
- - P (x-a) ] (f)
dx 2 EI L
dy = _ _1_ [ Pbx
2_ P (x- a) 2] + C
(g)
dx EI 2L 2
I [Pbx' P(x - a)']
y = - - -- +Cx + D (h)
EI 6L 6
1-

10
Four constants;A, B, C, and D appear in the two sets of equations. The
boundaries provide two conditions only,
y = 0 at x 0 (i)
y=o at x = L (ii)
The other two conditions required are obtained from the continuity
of the elastic curve, i.e. at point c the slope and deflection are the same
whether they are obtained from the expressions of part ac or those of part
cb. Mathematically,
dy = dy
at x = a, at x = a (iii)
dx dx
at x = a, y y at x = a (iv)
From condition (i) and equation (d), B = O. From condition (iii) and
equations (c) and (g), and noting that x = a, A = C. From condition (iv)
and equations (d) and (h) and noting that x = a, B = 0 and A = C,
D = o. From condition (ii) and equation (h), and noting that D = 0
Pb
and (Ira) = b, C = -- (L2_b 2) .
6EIL
Substituting the value of the constants of integration thus determine.d
into equations (c), (d), (g) and (h), the required expressions for the slope
and deflection are obtained.

Part ac
dy Pb
- - (L2 _ b2 _ 3 x 2) (i)
dx 6 ElL
Pbx
y = - - (L2 _ b2 _ x 2) (i)
6 ElL
Part cb

dy = ~ [L2 _ b2 _ 3 x2 + 3L (x _ a)2] (k)

dx 6IDL b

y = -
Pb- [L
6 ElL
- - (x -
b
a)3 + (L2 _ b 2) x-x 3 ] (I)

For a > b, the maximum deflection will occur in the left part to
which equations (i) and Ul apply. By equating the right hand side of
equation (i) to zero, the position of the point of maximum deflection is
located. .-
x = ~ (L2 _ b 2)/3
Substituting this value of x into equation U) the maximum deflection
is obtained.
 11

Pb 31
= (L2 _ b2) 2
9 J 3EIL
For a centrally loaded beam} the maximum deflection occurs at the
centre and its value is obtained by substituting x=b=L/2 into equation(j).
PL3
Ymu = 48 EI

ExaDlple 1.5 A simply supported beam of constant bending rigidity EI

is acted upon by a couple Mo applied at an intermediate point as shown
in Fig. 1.6 a. Find the expressions for the slope and deflection. Find also
the deflection at point c and the slopes at the supports.

Mo
(al a b

Mo a b Mo
L L L

Y,~I
(bl
I~ ~
Fig. 1.6

Solution: The reactions are first calculated and are as indicated. As in

the previous example, the bending moment has two different equations
to apply to parts ac and cb of the beam. The moment , slope and deflec-
tion equations are found in the usual way and are expressed separately
for the two parts as foHows :

Part ac, x < a

Mo
M = x (a)
L
d 2y = _ Mox
(b)
dx 2 ElL
dy Mx 2
dx
- - 2 ;IL +
A (c)

M x3
y __0_ +Ax + B (d)
6 ElL
 12

Part cb, x > a

M M.,x _ M (e)
L 0

2
=
d y
dx2 -r:-- Mo)
-
1 (MoX
EI (f)

~ = __1_( Mx Mx)
2
o _ +e (g)
dx EI 2L 0

y = _ _1_ (MoX3_ M.,x2) + ex + D (h)

EI 6L 2
The constants of integration are determined from the two boundary
conditions :
y=o at x 0 (i)
y = 0 at x L (ii)
and from the two continuity conditions:
dy dy
at x - a, = at x = a (iii)
dX dx
at x - a, y = y at x - a (iv)
From condition (i) and equation (d), B = O. From condition (iii) and
Moa
equations (c) and (g), A = e + Er' From condition (iv) and

equations (d) and (h) and noting the relation between A and e and that
M a2
B = 0, D =. 2~1 . Also from condition (ii) and equation (h),

Moa
and A = EI
Substituting the values of these constants into equations (c), (d), (g) and
(h) the required expressions for the slope and deflection are obtained.

Part ac,
dy Mo [
dx 6 ElL 6aL-3 (a' + x2 ) - 2 L' ] (i)

Mox [
y= 6 ElL 6aL - 3 a' - x2 - 2 L' ] (j)

.~
Part cb
dy Mo [
dx
=
6 ElL
6 xL - 3 (a 2 + x 2) - 2 L' ] (k)
 13

y= 6~IoL [3a 2 (L_x) _x(x 2 _3Lx+ 2L 2) ] (I)

The deflection at c is found by substituting x = a in either equations

OJ or (I).

Notice that if a > b this deflection is positive. . i.c. downward. Also

if point c is in the middle of the span, a = b = LJ2, it will have zero
deflection.
The slope at support a is found by substituting x = 0 in equation (i)
Mo
a. = -- (6 aL - 3 a2 - 2 L2 )
6 ElL '
Similarly the slope at support b is found by substituting x L
in equation (k) .
Mo
ab 6Eli (L2 - 3 a 2)

The elastic curve of the beam is shown in Fig. 1.6 b.

Example t.6 A c3:"ntilever carries a concentrated end load P as shown

in Fig. 1.7a. The beam's cross-section is such that in part ab the moment
of inertia is twice that in part be. Determine the slope and deflection
equations, and find the slope and deflection at the free end.

Ib_
_....J.. _ O.5L _ --I c
L
Ole,

Fig. 1.7

Solution: This problem is similar to the previous two examples. Only

in those, the moment equations were different along two parts of a pris-
matic (of constant I) beamJ while here there is only one moment equ~tion
which is valid for the entire length of the beam but the bending rigidity
El is different for parts ab and be of the beam. Each of these parts should
be considered separately.
14

Part ab, x < L/2

M = - P (L - x) (a)
d'y P (L - x)
dx'
- 2 EI
(b)

dy P (L - x)'
+A (c)
dx 4 EI

P (L - x)3
y= + Ax + B (d)
12 EI
Part be, x > L/2
M - - P (L - x) {e)
d 2y P (L - x)
(f)
dx 2 EI
dy P (L - x)'
+C (g)
dx 2 EI
P (L - x)3
y= 6EI +Cx+D (h)

Constants A and B are to be found from the boundary conditions at the

PL' PL3
fixed end a. A = 4 EI and B = - 12 EI' Constants C and D
are to be found from the continuity conditions at point b wbere there is
5 PL2
a sudden change in the beam's cross-section. C = - - and
16 EI
PL'
D - BEl"
Substituting the values of these constants into equations (c) (d), (g)
and (h) the required expressions for the slope and deflection along the
beam length are obtained.

Part ab,
dy P (L - x)2 PL2
+ (i)
dx 4 EI 4 EI
P (L - x)3 2
PL x
y + - - - PL' (j)
12 EI 4 EI 12 EI
Part be, """
dy P(L - x)2 5 PL2
+ (k)
dx 2 EI 16 EI
 15

P (L _ x) 3 5 PL 2x
(1)
y=
6 EI 16 EI +
The slope and deflection at the free end are found by substituting
x = L in equations (k) and (1) respectively. Thus,

Q o = (ddxY)c = ~
16 EI
5 PL' PL' 3 PL3
Yo = 16 EI 8 EI !6Er'
The elastic curve of the beam is shown in Fig. 1.7 h.
It will be noticed from the previous examples that the double inte-
gration method provides expressions for the slope and deflection from
which the slope and deflection at any point along the beam length may
be determined. The position and value of maximum deflection are also
readily obtainable. The method is useful in simple cases of prismatic
members where the bending moment along the entire length of the beam
may be expressed by a single equation. When this is not the case~
the mathematical solution becomes too involved and it is advisable to
use another more suitable method.

1.4 The m.om.ent-area m.ethod

According to equation 1.5, the differential equation of the elastic
curve is :
d 2y M
dx 2 EI

x,
x,
0 11 21 x

~
"', Y,

"', t

Am B.M.D.

r Y
Fig. 1.8
16

This equation is the basis of the moment-area method. It follows that

the restrictions imposed on the application of the double integration
method, still apply to this method. These restrictions are
(1 ) The material of construction is linearly elastic.
(2) Bending occurs about a principal axis of inertia.
(3) The deflections are small.
(4) The elastic curve is a continuous curve.

Fig. 1.8 shows a length 12 of an originally straight beam deflected

to the position 1'~2' under the given bending moments. Referring to
this figure and integrating both sides of equation 1.5" fe-produced above,
between x = Xl and x = X2 then,
2 2

S dx
-s 1
M
EI
dx

(~)'-(~:)2 ~S
M
EI dx 1.6

Since small deflections are being dealt with... a study of equation 1.6

shows that ~ = tan a = a in radians. Also, if each ordinate, M"

dx
of the R.M.D . is divided by its corresponding value of EI, the resulting
M
diagram will be called the EI' diagram and the integral on the right
M
hand side of equation 1.6 represents the area of the EI-diagram

between the chosen limits. For homogeneous beams of uniform cross-

section, i.e.EI=constant throughout the length of the beam, the right
hand side of equation 1.6 represents the area of the B.M.D. between the
shosen limits divided by E1. Accordingly, equation 1.6 may be re-written
as :

1.7

where al and a2 are ,the slopes of the tangents to the elastic curve at
points 1 and 2 and Am is the area of the B.M.D. between these two points.
 17

Equation 1.7 is known as the first moment-area theorem which may

be stated as :

"The change in slope, expressed in radians, between any two points on the elastic
_ .. 1 M
&UTVe IS C'i1,UU, to the net area of the - - -diagram between these two points".
EI
Multiplying both sides of equation 1.5 by x and integrating,

2 2

d2y Mx
S dx
2xdx
S EI
dx

[x(::)-y]: =-s Mx
EI
dx

[x, (:: ),- y, ] - [ X, (:: ),- y, ]=s~x 1.8

A study of this equation in conjunction with Fig. 1.8 shows that the
left hand side represents the vertical intercept through the origin between
the tangents to. the elastic curve at points I and 2. This distance is called
the tangential deviation. Also, integrating the product of the elemental
M

I
area of the EI -diagram, and the distance x gives the moment of

M
the area of the EI -diagram between the limi ts chosen about the

tangential deviation through the origin. Accordingly" equation L8 may

be re-written as :

Am
Tangential deviation EI x",
1.9

where Am is the area of the B.M.D. and Je,n, as shown in Fig. 1.8" is the
M
distance between the centroid of the EI -diagram and the origin.
 18

If the origin is transferred to point 1 then the tangential deviation Y12J

which is the deflection of point I with respect to the tangent to the elastic
M d-
curve at poi n t 2, is equal to the moment of the area ofthe EI - lagram

A
between point' I and 2 about roint I; YI2 = E~ XI (Fig. 1.9a).

Similarly, if the origin is transferred to point 2, the tangential deviation

Y2tJ which is the deftection cfpoiot 2 with respect to the tangent to the
elastic curve at point 1 is equal to the moment of the area of the

: -diagram between points I and 2 about point 2; Y21 = i; X2

(Fig. 1.9 b).

,.
E. curve E_curve
2
\ Y21

B_MD

x2 ..J
(a.) (b)

Fig_ 1.9

Equation 1.9 is the second moment-area theorem which may be stated

as:

"The deflection of any point relative to the tangent at another point on the elastic
curve is equal to the algebraic sum of the moments qf the area qf the
M .
- - -dtagram hetween these two points ahout the point where the deflection is
EI
desired" .
M
In the solution of ordinary problems, the EI -diagram consists of

triangles, rectangles and parabolas whose areas and centroids are easily
obtainable and the application of the two moment-area theorems to the
r
19

computation of slopes and deflections offer no difficulty. The following

remarks are useful in this respect.

(1) The slopes at various points on a beam can be determined after

the slope at some point has been found. This provides no difficulty in cases
of cantilevers or simple beams under symmetrical loading as in the first
case the slope is known to be zero at the fixed support" and in the second
case the slope is known to be zero at mid-span. In the case ofunsymmetri-
cally loaded beams, however.. the point of known slope is usually taken
at one of the supports where the slope is found by dividing the tangential
deviation of the other support by the distance between the two supports.

(2) The deflections at various poi nts generally call for the calculation
of the deflection at some point with respect to the tangent at another,
together with simple relations based on similar triangles.

(3) In applying the moment-area theorems, the directions of slope and

tang.ential deviation are usually obvious. However, sometimes this may
not be so. Following the same sign conventions considered till now for
the moment and deflection, and moving along the member from lefr to
right the change in slope will be such that the tangent on the right will
rotate anticlockwise with respect to the tangent on the left if the net area
of the B.M.D. between the two points is positive (Fig. 1.10 a). Conversely,
if the net area is negative, the tangant on the right will rotate clock-
wise with respect to the otheras shown in Fig. 1.10 h. Also, the tangential

+ B.M.D.

E curve
E.curve

(a) :7 (b)

Fig LID

deviation at a point will be above the tangent if the moment of the

M
area of t,h e EI -diagram about this p0int is positive (Fig. l.lla).

Conversely, it will be below the tangent if the moment of the area of the
'.

20

M t of the
EI -diagra m is negativ e (Fig. I.Il b). The sign of tbe momen
M .
area of the - -diagra m corresp onds to the sign of the B.M.D
EI

(al (bl

Fig. 1.11

1.5 Applic ations to the lDom.e nt-area lDetho d

be solved
Severa l of the classical problem s of beam deflect ions will now
t-area
to illustra te the applica tion and range of usefulness of the momen
method .
can-
&a..np le 1.7 Determ ine the slope and deflect ion at the free end of a
tilever of consta nt EI and loaded as shown in Fig. 1.12a.

L
,
p u..........

(bl
e-.-'=3- n
3
(c) "'. -J....\

Fig. 1.12
-----
The
Soluti on: The B.M.D . is first drawn and is as shown in Fig. 1.12b.
equal to the
change in slope of the elastic curve betwee n points a and b is
M lar casc.:o
area of the - -diagra m betwee n a and h. In this particu
EI
 21

this represents the slope at the free end as it is known that the tangent at
the fixed end has zero slope. Thus"

PL2 PL2
o_ "'b = _ _I (PL X L) =
EI 2 2EI 2EI

The positive sign indicates that the tangent to the elastic curve rotates
clockwise.
The deflection at point b with respect to the tangent at point a .. which
in this particular case is the actual deflection, can he found by taking

moments of the M -diagram between point a and b about h. Thus,

EI

Yb = _ _1_ ( PL 2 X 2 L) = _ PL3
EI 2 3 3 EI
The negative sign indicates that point b deflects below the tangent
at a, i.e. downwards .
The elastic curve is shown in Fig. 1.12 c.

ExaDlple 1.8 Determine the slope and the deflection at the free end of a .
cantilever of constant EI and loaded by a uniformly distributed load as
shown in Fig. l.l3 a.

w
(a) lR:dJ:!IIIID:ITL:ITIIIID:rrq: b
2

(bl
wI:
2 r---------
--'=.. 11-
4 4
,"", ---.i

I
(el

Fig. l.l3
-=-- y,

Solution: The B.M.D. is first drawn and is as shown in Fig. 1.13 b. The
r
calculations are carried out as in the lpreviolls example as follows:
wL'
[ 6 EI

[
22
The positive sign indicates that the tangent to the elastic curve rotates
clockwise.
I (I wL2 3 L) wL<
Yb = - EI 3 X L X 2 X 4 = - 8 EI

The negative sign indicates that point b moves downwards.

The elastic curve is shown in Fig. 1.13 c.

Exatnple 1.9 A simply supported beam of constant EI is loaded by a

uniformly dist ributed load as shown in Fig. 1.14 a. Calculate the maxi-
mum deflection.

w
(a) ! ! I I

(b)

lL~
16
0<,
(e)

Y,

Fig. 1.14

Solution: This problem needs more thought. In t he previous two

examples co ncerning cantilevers the tangent to the elastic curve at the
fixed support has beer known to be of zero slope and thus a direct solution
for slopes a nd deflections was possible. In this case, however, the direction
of a tangent to the elastic curve at some point should be determined first
in order to be able to calculate the actual slopes and deflections at various
points along the beam. Referring to Fig. 1.14 c, which shows the elastic
curve of the beam, the tangential deviation Yba is equal to the moment

of the area of the - Md'

- lagram b etween pomts
. a and b about point
EI
b. Thus, with reference to Fig. 1.14 b which shows the B.M.D.,

Yba =
I (2 L wL2 L) wL
EI 3 X 8 X 2' = 24 EI
4
,
I
I!
I
I 23

Yba WL3
aa = L - 24Ei' (i)

From symmetry of loading, the maximum deflection occurs at mid

span. Referring again to Fig. 1.14 c, the maximum deflection is given by

Ymax = aa X Lj2 - Yca (ii)

where Yca is the tangential deviation of point c with respect to the tangent
at point a, which according to the second moment-area theorem and with
reference to Fig. 1.14 b is given by :

Yea =
(2
I
El 3" X
L
2" X
wL2
8"" X
3
8 X
L)
2" =
wL
4

128 EI
(iii)

Substituting from equations (i) a nd (iii) into (ii),

wL' L wL4 5 wL4

Yro., 24 EI X 2" - 12d EI = 384 EI

Example 1.10 For the beam shown in Fig. 1.15 a, calculate the deflection
under the load and the maximum deflection if EI = 2500 m 2t.

41

(a) or x ---'d
c b
t--- - 6m - -- -I---3m
9m
(b)

(e) 0<,
~-----------k~~--~

Y,.

Fig . 1.15

Solution: Referring to Figs. 1.15 band c which show the B.M.D. and a
sketch o f the elastic curve,
24

144
Yha EI
(24 X 5 + 12 x 2) -EI
144
Q
a 9 EI -
-EI16
I 48
Yc ~ 6 Q
a - Yca ~ EI (6 x 16 - 24 x 2)
EI
48
2500 x 100 ~ 1.92 em.

Let the maximum deflection occur at point d at a distance x from a

then, ad = 0 and from the first moment-area theorem,

E~ (~ x2)
16 2x 2 /3 x ~ J 24 ~ 4.89 m.

The maximum deflection occurs at a distance 4.89 m. from a and its

value is given by :
Yb Q a X 4.89 - Yda

_ I ( 16 X 4.89 - 2- X 4.89 2 X
4.89)
-
52
~ -
EI 3 3 EI
52
- X 100 2.08 em.
2500

1.6 The elastic-load nlethod (analytical solution)

The moment-area method can be extended to develop another direct

method whereby the true slopes and deHections, rather than the change
in slopes and deflection deviations, arc directly computed. This method
is called the elastic-load method and its analytical treatment is developed
In the following.

Fig. 1.16 a shows an originally straight member ab which is subjected

M
to a general C3.SC of loading. The resulting EI -diagram and the clastic

curve arc shown in Fig. 1.16 band c respectively. It should be noted

that member ab could be a simple beam, a single span of a cOFltin-
uous beam , or an end-supported span of an overhanging beam. Referring
to Fig. I.l6 c, from the second moment-area theorem the tangential de-
viation Yba is given by :
 25

(a)

11
EI
(b)
x
dx

(c)

Fig. l.l6
b

J
MdX
= - - (L - xl
EI
a

a. = Y~a = LI JMdX
Er (L - xl
a

Mdx
EI

Now if it is imagined that member ab is subjected to a distributed

M
vertical load represented by the E(diagram, the limaginary reactions
, at supports a and b will be exactly equal to the respective values of a.
and a" given by he above two equations. This then proves the following
rule:

t "The slope of the elastic curve at a support a

erically equal to the value
M.
of the
of an end-supported heam is num-
reaction at a of a simply supported beam loaikd b)'

1M E(d,agram" .
I
,-

26

M
The - -diagram when used as load is referred to as elastic load, and
EI
hence the name . The corresponding load functions are elastic reaction, elastic
shearing force and elastic hending moment for the ordinary reaction, shearing
force and bending moment respectively.

Consider a point c on the elastic curve and let it be required to deter~

mine the slope a c . a c may be expre.ssed as :

It has been shown that aa is equal to the elastic reaction at support

a. The angle (aa - a c) is the change in slope between points a and c
which, according to the first moment-area theorem, is equal to the area

of the M -diagram between a and c. If this elastic load is subtracted

EI
from the elastic reaction at a the result will be the elastic shearing force
at point c. This then proves the following rule:
"TM slope of tM elastic curve at any point along an end-supported heam is num-
erically equal to tM value of tM elastic sMaring force at this point of a corresponding
simply supported heam".

Consider again point c and let it next be required to determine the

deflection Yc. From geometry of the 61astic curve shown in Fig. 1.16 c..
Yc = a. X a - Yca
where Yca is the tangential deviation of point c with respect to the tangent
at point a which, according to the second moment-area theorem, is equal
to the moment of the elastic load between points a and c about point c.
Also aa is the elastic reaction at support a which when multiplied by
the distance a gives its moment about point c. The rigbt hand side of the
above equation thus represents the elastic bending moment about point
c, and thus tbe following rule is proved.
"TM deflection at any point along an end-supported heam is nUmilrical1y equal
to tM value if tM elastic hending 11Wmilnt at this point of a corresponding simply
supported heam".
It remains to consider the sign conventions to define the absolute values
of the slopes and the deflections. If a positive M -diagram indicat-
El
M
es a downward elastic load and a negative E{diagram an upward
 27

load, the sign conventions used for ordinary load still hold. In other
words, if the elastic S.F.D. and elastic B.M.D. are plotted in the usual
way, a positive elastic shearing force means a positive slope of the tangent
to the elastic curve, i.e., a clockwise rotation of the tangent with respect
to its original position and conversely a negative elastic shearing force
means a negative slope, i.e. an anticlockwise rotation of the tangent with
respect to its or ginal position. Likewise, a positive elastic bending moment
indicates a positive deflection (downward) and a negative elast:c bending
moment indicates a negative deflection (upward).
1.7 Applications to the elastic-load method (analytical solution)
A number of numerical examples are given below to illustrate the
application and range of usefulness of the elastic-load method. It should
be emphasized that the method is applicable to any end-supported beam;
simple span, the end-supported span of an over han sing beam or a single
span of a continuous beam.
Example I." Using the elastic-load method find the deflection at point
C of the simple beam shown in Fig. 1.17a. Also, locate the po;ition and

calculate the value of the maximum deflection a~ong the beam if E =.

2100 t/cm 2 and I = 12500 cm 4 .

4
t
CI
Bm
(b)
r. I
I r.
f--
v.;

(e) a~~--------~r- __.-__~~ b

17 19
w,'
Fig. 1.17

Solution: The B.M.D. is shown in Fig. 1.17b. For ease of calculation,

the area of the B.M.D. is divided into regular parts whose areas and cen-
troids may be easily determined .
 28

4 x 8
WJ =-- 16 m 2 t at 2.67 m. from a
2
2 X 2
W2 = --- = 2 m 2t at 4.6.7 m. from a
2
W3 = 2 X 6 = 12 m 2 t at 5 m. from a
6 X 2
w. = - - - = 6m 2 t at 6.67 m. from a
2

1: M. = 0 = 16 X 2.67 + 2 X 4.67 + 12 X 5 + 6 X 6.67 - 8 rb

r. 19 m 2 t.

1: Y 0 = 16 + 2 + 12 + 6 - 19 - r.
ra - 17 m 2 t.

I ( 4) 46.67
Y, = EI 17 X 4 - 16 X 3" = Er
Now in order to avoid confusion, it is advisable to express the bending
rigidity EI in m-t units. Thus,

EI = 2100 X 12500 X 10-4 = 2625 m 2t.

46.67
y, X 100 = 1.78 em.
2625

By inspection, the maximum deflection occurs at a point d to the right

of c. Let this point be at a distance x from h. It is located from the con-
dition that the elastic shear there must be zero. Referring to Fig. 1.17 c,

w',=6, w'2=6(x-2), w'3=t(x-2)2

19 - 6 - 6 (x - 2) - 0.5 (x - 2)2 = 0

x2 + 8x - 46 = 0 , x = 3.874 m.
Taking moments about d,
3
I( 6 X 1.874 2 0.5 X 1.87 )
Ymax - - 19x3.874- 6x2.54I- - ---
EI 2 3

46.8
= -- X 100 - 1.79 em.
2625

.--
 29

From this example it is apparant that the maximum deflection occurs

very near to the centre of the span and that its value is nearly the same as
the deflection at the centre point. This perhaps is why in practice the
central deflection in simply supported beams.. rather than the maximum
deflection, is usually calculated regardless of the case of loading .

.Example 1.12 Calculate the deflection at point c in the overhanging

beam shown in Fig. 1.18a if EI ~ 2500 m 2 t.

(a) All be: 111(7\ ,e: X~ If'

I II I

15 3m_+-1.S 2

+
675 3m! 3m!
w,
(bl

I
W2 W3 4.Sm!
4ml
"-
"-
'-

Fig. 1.18

Solution: The B.M.Do .. drawn in parts" is shown in Fig. 1.18 h. The cor-
responding elastic reactions at support a are as indicated.

ra ~ 6.75 + 9 - 4 ~ 11.75 m 2 t

Referring to Fig. 1.18 b,

3 X 1.5
w, 2.25 m 2 t at 0.5 m. from c
2
2 1.5 2
~-x 15X-
-8 ~ 0.28m 2 t at 0.75 m. from c
3 .
 30

3.375 x 1.5
w, - 2.53 m't at 0.5 m. from c
2

x 1.5
- 0 .75 m't at 0.5 m. from c
2

I
y, = EI (i 1.75 X 1.5 - 2.25 X 0.5 - 0.28 X 0.75
- 2.53 X 0.5 + 0.75 X 0.5)
15.375
-- X 100 = 0.615 cm.
2500

It should be noticed that the B.M.D. is drawn in parts for the sole
reason of facili tating the calculation of the elastic loads and reactions.
Also, particular attention should be paid to the fact that the elastic reac-
tions considered should be only those due to the elastic load between the
two supports.

Example 1.13 Calculate the deflection at point c of the statically indeter-

minate beam shown in Fig. 1.19 a if EI = 12000 m't.

lBI 141 \1.5t

(a) ~r:=-Q_-I---~--%*""---,
liliI 3m 3 M2

mt
(b) 16
+

12.3 W. w.
3
(c) ---- - ---
I
I +
r,=O I r,
16

Fig. 1.19
 31

Solution: A solution by one of the methods available for the analysis of

statically indeterminate structures will produce the B.M.D. shown in
Fig. 1.19b. The B.M.D. on span ab drawn in parts is shown in Fig. 1.19c,
Referring to this figure,

I
20 X 3 , 4 X 3
= 30 m t, W2 = 6 m"
2 2
16 X 3
16 X 3 = 48 m't, \"14 = 24 m't
2
I W,
12.3 X 9
--'-'----'- = _ 55.35 m't
2
t W6
3 X
= --- = -
9
13.5 m"

I
2
Noting that the elastic reaction at the fixed support ra '= 0,
fb =
30 + 6 + 48 + 24 - 55.35 - 13.5 = 39.15 m't
I I
y, = El (39.15X 3-24xl + 3 x3x 1.5 +
3.1 X 3
2
111.6
x l) = El

~ 111.6
= _ _ X 100 = 0.93 em.
12000
E:s:am.ple 1.14 Determine the deflection at point e in span be of the con~
tinuou, beam shown in Fig. 1.20 a if EI = 2500 m'1.

d
6m_~~
32

Solution: A solution by the equation of three moments will produce the

R.M.D. shown in Fig. 1.20 b. The R.M.D. for span bc, drawn in parts
is shown in Fig. 1.20 c. Referring to this figure,
10 X 4
wI = - - ' - ' - = 20 m 2t at 2.67 m. from b
2

10 X 2
10 m 2t at 4.67 m. from b
2

2.44 X 6 - 14.64 m 2t at 3 m. from b.

3.56 - 2.44
X 6 = - 3.36 m 2t. at 4 m. from b.
2

(20 x2.67 + 10 x4.67 - 14.64x 3 - 3.36 X 4) = 7.12

6

I
y, EI (7.12 X 2-10 X 0.67 + 3.18 X 2x I +
3.56- 3.18 4 14.48 14.48
2
X2x-)
3
= EI = 2500 X 100 = 0.58cm~

1.8 The elastio-load D1ethod (graphical solution)

As for ordinary loads, only more so) it is more advantageous sometimes
to use a graphical procedure in applying the elastic-load method. This
M
is particularly true when the --diagram is very irregular in shape..
EI
or when the elastic curve is required as in the case of many structural
problems.

In such cases, the elastic shearing force and elastic bending moment
diagrams, which in effect arc the slope diagram and the elastic curve of
the beam, are obtained graphically in a manner similar to that used in
determining the ordinary S.F. and B.M.Ds. (Section 3. 9, Part 1. Briefly,
M
however, the - -diagram is divided into a number of strips, the areas
EI
of these strips arc calculated and plotted into a force polygon. A pole
at a known distance H is chosen and the corresponding link polygon
is drawn. This link polygon will touch the e~astic curve at points corre-
sponding to the ends of strips. The closing link which determines the
 33
datum line from which the deflections are to be measured is found from
the displacement conditions at the supports. For an end-supported beam,
the closing link is found by joining the projection of the two support points
on their corresponding links. The closed link pol ygon represents then
the elastic curve and the deflection at any point is the vertical intercept
of the link polygon measu red to the linear scale times the polar distance
H to the clastic-load scale. A ray in the polar diagram parallel to the clos-
i ng link defines the clastic rcactiOll'i. The elastic shearing force diagram
or the slope diagram is found by direct projection of the elastic loads on
their corresponding lines of action. The procedures arc illustrated in
Fig. 1.21.

M._D,agram
EI

h
r.
Slop .. diagram
w, 0

k----i---
Etas tic curve
r, W

~
L H J
Fig. 1.21

The scale of the slope diagram is the same as that of the elastic load.
Since th e latter is dimensionless, the slope is also dimensionless. This is
in agreement with the units of the radian angle in which Jthe slope
is expressed.
The scale of the elastic curve is made up as follows:
Linear scale X polar distance H X elastic load scale.
Since as mentioned above the elastic-load is dimensionless the unit of
deflection is that of length as it should be.
 34

The above procedure although basic in understanding the graphical

method and pointing out the analogy between the slope and deflection
diagrams on one hand and the S.F. and B.M.Ds. on the other is imprac-
ticable. This is because the values of the elastic loads thus calculated are
very small as the value of EI is usually great many times the area of
individual strips in the B.M.D.
Another simplified method, which the student 15 recommended to
adopt.. is discussed below with regard to Fig. 1.22.

8M 0

y' Slope diagram

Ela!;lic curve

Fig. 1.22

M
If instead of USIng the --diagram the area of the B.M.D. is
EI
considered as elastic loading, the diagrams derived from it will represent
respectively the slope and deflection each multiplied by EI. For example,
if at point c the vertical intercept in the deflection diagram is y then the
actual deflection is given by :

y (linear scale) X H (elasticload scale)

Yo
EI
 35

If further the polar distance H is taken equal to E1, then y measured

to I n ear scale will give the actual deflection at point c.But this again
leads to a difficulty in drawing the polar diagram as the elastic loads will
be exceedingly small compared to the polar distance EI if both are drawn
to the same scale. In order to overcome this difficulty, a separate scale
for EI, which is much bigger than that used for the elastic loads" is
adopted . Thus, the actual deflection at point c will be given by :

y (linear scale) X H (elastic-load scale)

Yc =
El (EI scale)
and since H = EI,

y ( linear scale x elastic-load scale)

EI scale

Similarly, if at c the vertical intercept in the slope diagram is y', then

the actual slope at c is given by :
y' (elastic-load scale)
ac ~

H (EI scale)

y' (elastic-load scale)

ac
EI

It is emphasized that in carrying out this method, the various scales

are of great importa nce and special care must be taken with regard to
them. Also .. like a ll graphical methods .. the degree of accuracy of this
method depends on the accuracy in tracing and scaling.

The graphical m ethod is a pplicable to cases other than that of the

simply supported beam considered above. The datum line from which
deflections are measured i s found from the d isplacement conditions a t
the suppo r ts. F or exa mple, in a n ove rhanging beam, the deflections a t
the supports are zero and the case is essentially similar to tha t of the
simply supported beam w here t he closi ng link or t he datum line is foun d
b y projecting the support points on the corresp onding links.

When parts of a beam h ave bendi ng m oments of d ifferent signs.. such

as in overhanging beams, t he direction of the vectors representing the
elastic loads in the polar diagram should correspond to the sign of the area
of the B.M.D.; positive areas represent downward loads and negative
areas represent upward loads.

r
36

1.9 Applications to the elastic-load method (grapltical solution)

The examples given below illustrate the method of solution and show
how the various scales are handled.

Exam.ple 1.15 Determine graphically the elastic curve of the beam in

Example 1.11 (Fig. 1.17). Indicate the position and value of the
maximum deflection and also the slope of the elastic curve at support a if
EI = 2625 n 2 t.

(al
t=4m-L 2
tt -l-r 4 2
b

L.5. ,lem~200em

j j
~)1ft1r
8m. E.L.S, 1em~10 ntt
E1.5. ,I em~1000 m2 t

r. 4 12 14 6rrfl r.
~I:::,
K G
r.
(e) !\
\
~
r
r.
-~-
-;:-{l'
0

(d)
~x~3.9_

~ -- -
'~ ~
EI

(el
J
. -t-
:-...~
v/
y ~ 1.6
Fig. 1.23

Solution: The B.M.D. is divided into four strips. The area of each strip
is as indicated in Fig. 1.23 b. The polar diagram is shown in Fig. 1.23 c.
The scales used are :

Linear scale, 1 em. = 200 em.

Elastic-load scale, 1 em. = 10m 2t,
 37

EI scale, I em. ~ 1000 m 2 !.

The scales for slope and deflection are thus,

10
for slope, I em. - - - ~ 3.8 X 10- 3 rad .
2625

200 X 10
for deflection, 1 em. = - - - - - 2 em.
1000

The slope and deflection diagrams are plotted in the manner described
previously and are as shown in Figs. 1.23 e and d respectively. The el-
astic curve of the beam (the deflection diagram on a horizontal base lir:e)
may be obtained by drawing another link polygon corresponding to 0'
as pole; 0' being on the same level as the point that defines the elastic
reactions on the force polygon.

As the point of maximum deflection occurs at zero slope, its location

is determined by drawing a tangent to the deflection diagram parallel
to the base line.

From Fig. 1.23 d, this point lies at a distance = l.95 em. from support
a, and the vertical intercept measures 0.9 cm. Thus,
1.95 X 200
= 3.9 m. from a
100

Yma, ~ 0.9 X 2 1.8 em.

From Fig. 1.23 e .

a. ~ 1.7 (3.8 X 10- 3 ) ~ 6.5 X 10- 3 rad.

It is seen that the position and value of the maximum deflection are
very nearly the same as those obtained previously by the analytical method
(Example 1.I1).

ExalDple 1.16 Determine graphically the elastic curve of the overhanging

beam shown in Fig. 1.24a. Indicate the values of the deflections at points
c and d if EI ~ 2500 n 2 t.

r
38

L.5. ,1em=150 em

E.L.5.,lem=5m't
5.5 E1.5. ,1 em=1000 m' t
2.9 6.9 5.5 i.9 2.2 4m' t

2.9
y=0.3
(d) J...----'

0
6.91=====::7

(e)

Fig. 1.24

Solution: The B.M.D. is first drawn and then divided into strips. The
individua l areas of these strips a re shown in Fig. 1.24 b. The polar
diagram and the corresponding elas tic curve are shown in Figs. 1.24c and
d respectively. The scales used are:
Linear Scale.. em. = 150 em .
Elas tic-load Scale, em. = 5 m 2t.
EI Scale, em. = 1000 m 2 t .
Thus, the deflection scale is made up as follows :
150 X 5
I em. = 0.75 em.
1000
From the elastic curve in Fig. 1.24 b, the ordinates at points c and dare
0.9 and 0.3 em. respectively. Thus,
Yo = 0.9 X 0.75 0.675 em . (downward)
Yd = 0.3 X 0.75 0.225 em. (upward)
 39

1.10 The conjugate beaD> Itlethod

In the development of the elastic-load method .. some restrictions have
been made on its application. These include:
(I) The point at which the slope or deflection is desired lies on an
end-supported beam. This applies to simple and continuous beams.
(2) The beam does not contain an intermediate hinge within the
span where the point considered lies.

The elastic-load method cannot therefore be applied to cantilevers,

overhangs of overhanging beams or beams having intermediate hinges.
It can be extended, however.. to develop another method whereby it will
again be only required to calculate elastic shearing forces and bending
moments to determine the slopes and the deflections regardless of the
abovementioned restrictions.
This method is called the conjugate beam method. A conjugate beam
means a substitute or corresponding beam. It is an imaginary beam
having a span length equal to that of the real beam and loaded, as in the
M
elastic-load method, by the - -diagram. Further its support conditions
EI
are generally changed from those of the real beam to meet the following
requirement:
The force boundary conditions in the conjugate beam should be
comparable to the displacement boundary conditions in the real beam.
Consider as a first example a cantilever. At the fixed support the real
beam has neither slope nor deflection, and therefore the conjugate beam
must have neither shearing force nor bending moment at this end. The
only way for a point not to have shearing force or bending moment is to
be a free end. Therefore a free end in the conjugate beam corresponds
to a fixed support in the real beam. On the other hand, at the free end of
the cantilever the real beam has finite values of slope and deflection. The-
refore, the conjugate beam must have finite values of shearing force and
bending moment. The only way to have both finite shearing force and
bending moment at a point is to be fixed. Therefore, a fixed support in
the conjugate beam corresponds to a free end in the real beam.

As second example~ consider a simply supported beam. At each of

the end supports the real beam has a finite value of slope and a zero de-
flection. The conjuga te beam must therefore have a finite value o,f the
40

shearing force and zero bending moment at the corresponding locations.

In order to achieve this condition offinite shearing force and zero bending
moment, the conjugate beam must also have simple supports. Therefore,
a simple support in the conjugate beam corresponds to a simple support
in the real beam .

As a third example, consider an intermediate support. At such a

location the real beam has a finite slope and zero deflection. The conjug-
ate beam must therefore have a finite shearing force and zero bending
moment at the corresponding location; a condition which is satisfied by
an intermediate hinge. Hence; an intermediate hinge in the conjugate
beam corresponds to an intermediate support in the real beam.

By similar reasoning, other support conditions of the conjugate beam

may be found. The various basic cases are illustrated in Fig. l.25.

REAL BEAM CONJUGATE BEAM

" FIXed support

~ D<=O.y=O
Fr:e end _
S.F. - 0, 8M. _0
t
..
FreE" end Fixed support

(1..10, y/O t ~ S.F./O, B.M.;iO

S implle suppor~ Simplle support

I
10<10, y=O S.F./O , B.M.:O

t (nller, ohi nge t

S.F/O B.M.:O

{ (n teLahingle t
rj.fO , y/O

Fig. 1.25
 41

Some examples of common types of statically determinate and indeter~

minatc beams tQg.ether with their corresponding conjugate beams are
shown in Fig . 1.26.

STAT. DEl. BEAMS

~'-----4

a
l""O---- --b

~ir----:R:"""*---'"c a

a ___*b______-7c~~d a b d
Ok: K

b o",-c_ _ _7:d
:&: r 1i.

d o~e----J.-J.
:&: ..a
STAT INDET BEAMS
~~O_ __ _ _~~b O~_ __ _ -7~

O'--___ _ _--"b

a b c d

~O b
A
c
:;;a:
d

Fig. 1.26

It should be noticed that all statically determinate beams have stati-

cally determinate conjugate beams. Statically indeterminate beams on
 42

the other hand have unstable conjugate beams. Nevertheless, the method
is. still applicable to t hem, and the unstable conjugate beams are always in
M
equilibrium under the applied - - -diagram of the real beam.
EI
The sign conventions used in the conjugate beam method are the same
as those given in the elastic-load method, i.c. elastic loads on the conjugate
beam are downward if the B.M.D. of the corresponding real beam is posi-
tive. . positive shearing force in the conjugate beam results in a clockwise
rotation of the tangent to the elastic curve, and positive bending moment
in the conjugate beam indicates downward deflection.

1.11 Applications to the conjugate bealll lIlethod

A number of examples is presented in the following to show the applica-
tion of the conjugate heam method : Some of these problems deal with
statically indeterminate beams. In such cases the B.M.D . is given and the
student should not worry himself, at this stage, about how these moments
have been obtained . This will be co nsidered later on in this book.

Example 1. 17 Calculate the deflections at points c and d and the slopes

at points a a nd b of th e beam show n in Fig. 1.27 if EI = 2500 m 2t.

21 21
HIm
( a) a
3 'I 3m
I
(b) 19.5 nl 1

(c) 4.5ml

.-
3
Fig. 1.27
 43

Solution: The elastic loading, drawn in parts, and the corresponding

conjugate beam are shown in Fig. 1.27 c. For simplicity, the conjugate
beam is divided into its component parts as shown in Fig. 1.27 p. This
figure also shows the elastic reactions.

I 2x3
y, = EI (9.5x3+ -2- X 1-
2
x3x4.5x
3 a3 X3 - -3x3
2 - xl)

16.875 16.875 X 100

= -- = 0.67 em. (downward)
EI 2500
I 4 X 2 2 5.67
y. = EI (-5.5 X 2 + -2- X -x 2) = - -
3 EI
5.67 X 100
= - 0.227 em., i.e. 0.227 em. (upward)
2500
I 9.5
" a -- -EI (9 .5) -- 2500
- -- + 0.0038 rad.

The positive sign indicates that the tangent to the elastic curve at a
rotates clockwise.
1.5 1.5
ad =- -= -- = - 0.0006 rad.
EI 2500
The negative sign indicates that the tangent to the elastic curve at d
rotates anticlockwise.

Es:alDple 1.18 Calculate the maximum vertical deflection in the beam

shown in Fig. 1.28a if EI = 5000 m 21.

(a) ~---{_o=--_ _ _ _+_~b

(b)

I
I.... 8mt 14.5

Fig. 1.28
44

Solution : The elastic loading and the reactions of the corresponding con-
jugate beam are shown in Fig. 1.28 h.
Let the point of zero elastic shear be at a distance x to the right of c.
4 X 2 2 X I 2xl
Q, = 4.5 +~ +- 2- - 2 = 0, x = 3.08 m.

I(
Y..", = -
EI
4.5 X 3.08 + 4x2
-2- X 4.41 + 2xI
-2 - X 4.75

_ 3.08 X 6.16 X 3.08) = 26.6

2 3 EI
26.5 X 100
= 0.53cm. (downward)
5000

Exam.ple 1.19 Calculate the slopes at the intermediate hinge c and the
maximum deflection along the beam shown in Fig. 1.29 a if EI =
6000 mIt.

(b)

6.75
Fig. 1.29

Solution: The elastic loading and the reactions of the corresponding

conjugate beam are shown in Fig. 1.29 b. Referring to this figure,

2.25 X 1.5
Qd = + 2 + 1.6875 m 2t.

1.6875
a,1 = - - = 0.28 X 10 3 rad. (clockwise)
6000
Q" = 7.3 + 1.6875 = 8.9875 mIt.
8.9875
a" = - - = 1.5 X 10 3 rad. (clockwise)
6000
 45

The maximum deflection occurs at the point of zero elastic shear in

the conjugate beam. Let this point be at a distance x to the right of c.

2.25 X 1.5 6 .75 x

Q. = + 7.3 - - x
4.5
X "2 = 0, x = 3.46 m.
2

2
I ( 6.75 X 3.46 3.46)
Ym .. = -
EI
1.6875 X 4.46 + 7.3 X 3.46 -
4.5 X 2
x- 3
22.5 22.5
= -- = - - X 100 = 0.375 em . (downward)
EI 6000

Example 1.20 Determine the maximum deflection in each span of the con-
tinuous beam shown in Fig. 1.30a if EI = 2500 m 2t.

b 17.5t c d

(a)~====6-m-_-_--=:;A~
~=--4--+1-2--:;:;A-l<;:-_
:;:c. -_-6--7~
3.56
2.44
(b) l-====::::::::::=-==-+~---}~-=--:::::::::=1
+
6.82.

(c)

2.44m't 3.56

10mt

Fig. 1.30

Solution: A solution by the method or the equation of three moments

will produce the R.M.D. shown in Fig. 1.30b. The clastic loading and
the reactions ofthe corresponding conjugate beam are shown in Fig. 1.30c.
Since the maximum deflection occurs at the point of zero elastic shear in
the conjugate beam, such a point is first located.
46

Deflection in span abo

Let the point of zero elastic shear be at a distance x from a~

2.44 x x
Qx= - 2.44 + - 6- X "2 = 0

x = 3.46 m. to the right of a.

I ( 2.44 X 3.46 2 3.46) 5.62

Ym~= EI - 2.44 X 3.46+ 6 X 2 X 3""
= - EI
5 .62 X 100
= - = - 0.225 em .., i.e. 0.225 em. (upward)
2500

Deflection in span be.

2.44 X 6
Q b= - 2 .44 + - 4.88
2

Let the point of zero elastic shear be at a distance x from b,

1.12 x 2 lOx x
Qx= 4.88 + 2.44 x + -6-x- 2- x- =0
4 2

" = 3.35 ffi. to the right of b.

3.35 2 1.12 X 3.35 2

Ym~= _1_ (4.88 X 3.35 + 2.44 X - 2- + 6 X 2 X
:\.35
3
EI
10 X 3.35 2 3.35 ) 15.4
4x2 X --
3 =Ei
15.4 X 100
- 0.616 em. (downward)
2500

Deflection in span cd.

Let the point of zero elastic shear be at a distance x from d.
3.56 x 2
Qx= 3.56- 6 X 2 = 0

x = 3.46 m. to the left of d.

I ( 3.56 X 3.46 2 3:46 ) 8.2

Ym",,= - - 3.56 X 3.46 + X -- = - -
EI 6 X 2 3 EI

8.2 X 100
----- = - 0.328 em., i.e. 0.328 em. (upward)
2500
I,
I
r

47

futaDlple 1.21 Determine the deflection and the slope to the left of

I
hinge d, the maximum deflection in span be and the deflection at the
free end e of the beam shown in Fig. 1.3 1a if EI ~ 2000 m't.

41 bt
I (a )
d b
111m

-L3.J.-3 6m

Gmt
3
(b)
+ +
4.5
6mt
6

(c) a

(d)

Fig. 1.31

Solution: The given beam is statically indeterminate, and a solution by

one of the methods used for the analysis of statically indeterminate beams
will produce the B.M.D. shown in Fig. 1.31b. The corresponding
conjugate beam and the applied elastic loads are shown in Fig. 1.31c.
Referring to this figure,

adl ~ _ 1_ (6 X 6 _ 6 X 6) ~ 0
EI 2 2

which means that the tangent to the elastic curve just to the left of
d remains horizontal.
48

Yd = _1_ (~ X ~ X 6 - ~ x ~) = ~
EI \ 2 2 2 EI
18
= 2000 x 100 = 0.9 em. (downward)

The maximum deflection in span be will occur at a point f where the J

elastic shear in the conjugate beam is zero. Let point f be at a distance
x to the right of b.

Referring to Fig. 1.31d which shows the elastic loads on span be,

3 X 6 ) -3
x6x4.Sx3 - 3x6x3 - -- X 2 =-
2 EI

WI = 3 x / EI, w2 = x 2 /4 EI,

Qr= - 3 + 3 x+ :2 _ ~; _C;2 _ x:) = 0

2x 3- IS x 2 + 36 x - 36 = 0 , x = 4.3 m.
3
1 [ 4.3 3 4.3 (3 X 4.3 2 4.3 )
Yr = Ymax = - - 3 X 4.3 - - X- - -
EI 12 2 2 4
4.3 4.3 4.3 2 4.3] 4.2
X -3- + 3 X 4.3 X """2 + -;- X = - EI 3
4.2 X 100
- = - 0.21 em., i.e. 0.21 em. (upward)
2000

Referring again to Fig. 1.31 d, and taking moments about point b for
the elastic loads on span be,

1 (2 3X6)
QC=6EI 3 x 6 x4.SX3-3X 6 x 3 - -- X 4
2
=-Ei6
Deflection at the free end e

1 ( 6 X 2 2 ) 20
Y. = EI 6 x~2 +n- -
2
X 3 X 2 = EI

20
= - - X 100 = 1 em. (downward)
2000
 49

1.12 Standard cases of bean> deflections

The student is advised to memorize the deflection and slope formulae

for the standard cases of loading for simply supported beams and cantil-
evers presented in Table 1.1. When the load consists of a combination of
the given standard cases, the deflection or slope are readily obtainable
by applying the principle of superposition" but when the loading is com-
plicated or the moment of inertia varies along the beam length, resort is
made to one of the given methods for deflection computation.

Beam& Loading Deflections Slopes

w /
b wLJ
;"~/~I~;21~
=5wt
c 384EI ~a=-oIb= 24EI

a
tV! Ll2 rLl2
+
w
A
-----l
b S= p@
c 48EI
PI?
()(a=-0I6'16EI

~
II _wL4 wI!
~
I I I I I I I I I
b-8EI ()(b= 6EI
L

~a t I) _ pt.l
b-3EI
PL2
()(b= 2EI
I L ~
Table Ll

EsaJn.ple I..n Determine the deflection at the free end of a cantilever

3m. long and acted on by a uniformly distributed load of It/m and a con-
centrated end load 1.5 t if EI = 1500 m 2 t.

Solution: If the uniformly distributed load acts alone y = wL'/8 EI,

and if the concentrated load acts alone, y = PL 3/3 EI.

Total deflection is therefore given by:

3
3X 3 )
100 = 1.57 cm.
 50

1.13 Beam.s of variable moment of inertia

The slopes and deflections determined so far have been for prismatic
beams" i.e . beams in which the moment of inertia is constant for all the
sections along the beam length. The principles derived in sections 1.4,
1.6, 1.8 and 1.10 are also applicable to non-prismatic beams only with
slight modification. This is considered below.
The basic relations derived for the slope and the deflection are:

a = S Mdx
EI'
y =SXMdX
EI

For prismatic beams these relations have become,

a = _I
EI
5 Mdx

y = -1-5EI
xMdX 1.10

which means that the slope is represented by the area of the B.M.D. and
the deflection by the moment of this area" each divided by EL

If" however, I is not constant the expressions in equation 1.10 for the
slope a nd the deflection become:

a = -1-5
E
Mdx =
I
-1-5 ~Mdx
EI, I
1.11

y = _I jXMdx = _1_
E I EI,
S~ I
xMdx

where Ir is a reference moment of inertia which is arbitrarily chosen but

usually taken equal to either the smallest or the la rgest value of the
moment of inertia along the beam.

A comparison between equations l.l 0 a nd 1.11 shows that the methods

already developed in sections 1.4, 1.6, 1.8 and l.I0 may be a pplied to
beams of variable moment of inertia provided that the bending rigidity
considered is Ell' and that instead of using the B.M.D. another diagram
formed by dividing the ordinates of the B.M.D. by the corresponding
moment of inertia and multiplying all by I, is used. The diagram so for-
med is called the modified B.M.D.

The application of the various methods to nonprismatic beams is

illustrated in the following examples.

1- - - - -
 51

Exa.m.ple 1.23 A simply suppor.ted beam carries a central concentrated

load P as shown in Fig. 1.32 a. The beam's cross-section is detailed such
that the moment of inertia of the central half is double that of the
remainder of the beam. Calculate the maximum deflection.

Solution: The B.M.D. under the given case of loading is shown in Fig.
1.32 b. Taking the smaller value of the moment of inertia as 1p the modi
fied B.M.D. will be as shown in Fig. 1.32 c. Using the moment-
area method, the maximum deflection is equal to the tangential deviation
of point a from a tangent at point c. For ease of calculation the modified
B.M.D. is divided into the regular parts indicated. Thus..

3 PL3
256 E1,

Example 1.24 A cantilever carries an end concentrated load P as shown

in Fig. 1.33 a. The hearn's cross-section is detailed such that the moment
of inertia of the half nearer to the fixed end is double that nearer to the
free end. Calculate the deflection under the load.
52

ZI
1/2 I LIZ

~
(b)

p~ p~
8 .M.D.
p~
-
(c) f- .
16"t 8
,
f8
I

Modified B.M.D

(d)
E . curvE'

Fig. 1.33
-------------
Solution: The R.M.D. of the beam under the given case of loading is .
shown in Fig. 1.33 b . Choosing I as I" the modified B.M.D. will be as
shown in Fig. 1.33 c. Using the moment-area method,

~ _1_ ( PL' X~ + PL' x 3L + PL' X 5L) ~ 3 PL'

EI, 8 3 8 4 16 6 16 EI,
3 PL'
i.e . (downward)
16EI,

Example 1.25 Re-solve the problem in example 1.23(Fig. 1.32) using the
elastic-load metbod.
Solution: Referring to Fig. 1.32 c which shows the modified R.M.D.,
PL' PL' PL' 5 PL'
r. rb ~ 64 + ~ + 128 ~ \28

Yc
I (5 PL' L
X PL" X~- L
PL' X ...!: _ PL' X I 2)
E1, 128 2 64 3 64 8 128
3 PL3
256 EI,
Example 1.26 For the cantilever beam shown in Fig. 1.34 a, calculate
the deflections at points c and d if the moment of inertia varies as
indicated. I I ~ 2I2 ~ 12000 cm 4, E ~ 2000 tfcm' .
 53

3t 6t 61
(a) a~__~__~d~____~__~________ t-__,{D
f
I
I
15 f, 15+2 I,
3m 15

t
( bJ

I
...
225
+
9 9

! (cJ

l,
t-
5 +

9 9
9 rrtt
(d) a b

L75
1725
6.75
6.75 27 6.75
Fig. 1.34

So'ution : The B.M.D. is shown in Fig. f .34 b. T aking I, = I" the modi-
fied B.M.D. will be as shown in Fig. 1.34 c. The conjugate bea m and the
e astic loads are shown in Fig. 1.34 d . The elastic reactions indicated are
calcula ted below .

L M. 0 = 6.75 X I + 27 X 3 + 6.75 X 5 - 9 X 2- rb X 6
b

I
6
(6.75 + 31 + 33 .75 - 13) = 17.25 tn 2 [ (upward)

L M, = 0 = 6.75 X 1.5 + 6.75 X 6 + 27 X 3 + 6.75 X 10

- 3 X 13/3 - 9 X 7 - 17.25 X II - rd X 3
63.25
22.75 m 2 t. (upward)
3
54
I:Y = 0=6.75+6.75+27+6.75-3-9 -22.75-17.25-r
ra = - 4.75 m 2 t., i.e. 4.75 m 2 t. (downward)
Ell = 12000 X 2000 X 10-' = 2'1GO m 2 t.
I
y, = Er;- (17.25 X 4.5 - 6.75 X 3.5 - 27 X 1.5
2.25 X 4.5 21
+ 2
X 1.5)=-
Ell
21
X 100 = 0.875 em. (downward)
2400
I 24.25
Yd=--(-4.75 X 3-6.75 X 1.5)=---
ill I Ell
24.25
= _ - - X 100 = - 1.01 em., i.e. 1.01 em. (upward)
2400
Example 1.27 For the overhang"ng beam shown in Fig. 1.35 a, calculate
the deflection and the slope at point c. The moment of inertia varies as
indicated. 1= 5000 em 4 and E = 2000 t fcm 2
4!

(a)4r_I_I_I_"~~11111 T ";"~
a'IIIJJCL[IIIJ~2=t/cm~11]J:C[IIJ~b~[[IIIJ3'C
I; IIJ
l
12m I
0-..1
( b) 8
8

16m I 12
B
(e)

- ; ; - . - - - 16

-'~ - - --,
(d) a ~=--------+ ---:--::-:~),k=~~

6(b
2133 32m'! Fig. 1.35
 55

Solution: The B.M.D. is shown in Fig. 1.35 b. Taki ng 21 = I" the modi-
fie~ B.M.D. will be as shown in Fig. 1.35 c. The conjugate beam and
the elastic loa ds are as shown in Fig. 1. 35d. The elastic reactions indicated
are calculated below.

b
LMb
a
o= 6r. +4 X 2/3 + 16 X 1-21.33 X 4-8 X 4/3-1.33 X I

f
I
ra 13.11 m 2 t. (upward)

f LY o= r, + 12 + 4 + 16 + 13.11 - 21.33 - 8- 1.33 - 1.33

r, - 13.11, i.e. 13.11 m 2 t (downwa rd)

b
L, Mb = 0 = m, + 12 X 2/3 - 13 II X 2 - 1.33 X I

m, = 19.55 m 3t. (a nticlockwise)

EI, = 2(2000 X 5000 X 10- 4 )= 2000 m 2 t.

13.11
= - - = 0.0065rad. (clockwise)
EI, 2000

me = 19.55
y, X 100 0.98 cm. (downward)
EI, 2000

Example 1.28 Fig.1.36 a shows a reinforced concrete cantilever. It has

a rectangular cross~section of constant breadth 0[30 em. and a depth vary-
ing from 40 cm. at the free end to 80 cm . at the fixed end. Estimate the
slope and the deflection at end b. E for concrete = 200 t/cm 2

Solution: The beam is divided into four strips. The values of the moments
of inertia at the ends of these strips have been calculated and are as indi-
cated. The moment of inertia between the ends of strips is assumed to
vary linearly. The B.M.D. is shown in Fig. 1.36 b. Taking I, =128 X 10 4
cm., the modified B.M.D. will be as shown in Fig. 1.36 c. The elastic
loads indicated are calc"Ilated below.

18 X 0.75
Wl= 6.75m 2 t_ at 2.75 m.from b
2
20.2 X 0.75
W2 = 7.6 m 2 t. at 2.5 m. from b.
2
w, = W2 = 7.6 m 2t. at 2 m. from b.
56

3m 61

(a)

'0
$1
.0
'$1
l{) "
'0

..."
K
l{)
N
CD "
K M
~
II
co II>
N co .!l "
-;; "
18 mt

(t)
~
w,

18
/
/ I
(e) / I /
/ I / /
/ I / /
a~ ____ ~ ____ ~ ____ ~ __ ~b

Fig . 1.36

21.3 x 0.75
W4 = 8 m 2t. at 1.75 m. from h.
2
Ws = W4 - 8 m 2 t. at 1.25 m. from h.
18.45 x 0.75
\V6 = 6.9 m 2 t. at 1 m. from h.
2
W7 -
w. - 6.9 m 2t. at 0.5 m. from h.

Er, = 128 X 10 4 X 200 X 10- 4 = 25630 ill 2t.

Using the moment-area method,

1
ab = - (6 .75 + 2 X 7.6 + 2 X 8 + 2 X 6.9)
Er,

51.75
=25600
--- 2.02 X 10- 3 rad. (clockwise)
 57
I
Yb = EI
, (6.75 x 2.75 + 7.6 X 2.5 + 7.6 X 2 +8 X 1.75

, +8 X 1.25 + 6.9 X I + 6.9 X 0.5)

I 87.1
X 100 = 0.34 em. (downward)
25600

Example 1.29 Find graphically the clastic curve of the beam shown in
f Fig. 1.37a due to the given case of loading. Indicate the values of the
maximum positive and maximum negative deflections. The moments of
inertia, in m4., at the various locations are as indicated. E = 210 tfcm2.

(')O~,I~'
Ul<D Q <DlS
00M MO'
Od....-;;ll)-.oo

rr _\I a ~
0

7.5 , 1 1 5 m
a
7.5

I
,~
f

(b)

55 42.5 108 35 6
29 35 90 80 108

90

29
I
...- - E I , - -
(c )
Fig. 1.37
58

as shown
Solutio n; Taking I, = 0.55 m', the modifie d B.M.D . will be
ted and located at the
in Fig. 1.37 b. The elastic loads have been calcula
centroi ds of individ ual strips as shown.

EI, = 210 X 0.55 X 10' = 11.55 X 10' m 21.

1.37 c.
The polar diagram with a polar distanc e = EI, is shown in Fig.
d. The base line
The corresp onding link polygo n is shown in Fig. 1.37
the pro-
from which the deflect ions are to be measur ed, is the line joining
jection of the two suppor t points on the link polygo n.

The scales used are :

Linear scale, 1 em. = 500 em.
Elastic -load scale, I em. = 100 m 21.
2
El r scale, I em. = 2 X 10 5 m t.
500 X 100
Deflect ion scale, 1 em. -'-:-'---'--'-:-,:7- = 0.25 em
2 X lOS
and the
From the link polygo n in Fig. 1.37 c, the ordina te at point c
vely:-
maxim um ordina te betwee n a and b measur e 0.8 and 0.32 em. respecti
Thus"
Yo 0.8 X 0.25 = 0.2 cm. (downw ard)
Yo 0.32 X 0.25 = 0.08 em. (upwar d)
located
The point where the maxim um negativ e deflect ion occurs is
line.
by drawin g a tangen t to the elastic curve paralle l to the base

1.14 GeoD1 etry of deflec ted frattle s

can be
The method s present ed for the calcula tion of beam deflecti ons
the deflecte d frame is
applied to frames . A sketch of the genera l shape of
of a frame,
necessa ry in this regard . On sketchi ng the deflect ed shape
the followi ng princip les are useful.
ement,
(1) At a fixed suppor t, the frame has neither rotatio n nor displac
while at a
at a hinged suppor t it has a finite slope but no displac ement,
roller suppor t it has both finite rotatio n and displac ement.

defor-
(2) The angle betwee n membe rs meetin g at a rigid joint before
mation remain s the same after deform ation.
by
(3) The deflect ed shape of a membe r in a frame is defined
For exampl e, a membe r free of
the elastic loading along its length.
 59

moments will remain straight, a nd a member subject to constant moment

will bend in a second order curve. Also the elastic curve of t he member
will h ave a point of inflection at the point of zero moment.

(4) Since the elastic curve ofa member is flat" the length ofa deflected
member is assumed to remain unchanged after bending . Thus if onc end
of a member moves some displacement along the member the other
end will move the same displacement.
Since no new theory is involved, the method for computing frame
deflections is best introduced by a number of examples.

Example 1.3D Find the horizontal displacement at support b of the frame

shown in Fig . 1.38a if it has a uniform bending rigidity, EI = 2500 m 2t.

4t Gmt
/ c
+
a
>-- L.L3
1
4m
0<,
0<,

bJ \
(a) (b)
-! xJ--
Fig. 1.38

Solution: The B.M.D. and a sketch of the elastic curve are shown in
Fig. 1.38 b. Using the elastic-load method,

nc =
I (6X3)
EI 2 =
9
2500 (anticlockwise)

Since joint c is rigid, cb will also rotate anticlockwise an angle aco

Further, since member cb is free of moment, it will remain straight
and the horizontal displacement at b is equal to the rotation a c times
length be.
9
x = - - X 4 X 100 1.44 em. (to the right)
b 2500

Example 1.31 For the frame in Fig. 1.38 a, find the horizontal displace-
ment at support b due to a unit horizontal load acting there as indicated
in Fig. 1.39 a.
60

4mt
a c
6
--...11 4m

(a)
-lx,;lxL
(b) ---1 x, L

Fig. 1.39

Solution: The B.M.D. and a sketch of the elastic curve of the frame are
shown in Fig. 1.39 b.

Using the elastic-load method,

1(4 X 6 2 ) 8
n, = - EI - 2- X s = -2500 ,i.e. 0.0032 rad. (clockwise)

The hor:zontal displacement at support b, x b ' consists of two parts; the

first part xb , is due to the rotation of member be and the second part x b"
is due to bending of member be.

X'b = 4 n, = 4 X 0.0032 X 100 = 1.28 cm.

Xb" is best found by the moment-area method as follows:

x "= _1_ (4 X 4 X ~ X 4) = 21.33

b EI 2 3 EI

21.33 X 100
- 2500
0.85 cm.

Xb = 1.28 + 0.85 2.13 cm. (to the left)

I ExanJple Ien Calculate 4a ~ 4C ~ ad .J Xc , xd and xb of the frarn.e
shown in Fig. 1.4Oa if it has a uniform bending rigidity EI ~ 25000 m 21.
61

(a)

18

21
31.5

a 0<.
(b) (e)

I Fig. 1.40

Solution : The B.M.D. and the clastic loads are shown in Fig. 1.40 b. A
study of the elastic. curve in Fig. 1.40 c shows that neither c nor d can
move vertically and that they can only move horizontally. Also the hori-
zontal displacement is the same for both joints. Applying the elastic-load
method to member cd, and by taking moments about c,
1 36
ad ~ -
6&
(54 X 1.5 + 4 .5 X 2 + 31.5 X 4) =-
&

= ~
25000
= 1.44 X 10-' rad. (antidockwise)

l
By taking momepts about d,
I 54
a

= - - (31.5 X
6 EI
2 + 4.5 X 4 + 54 X 4.5)
EI

~
54
= - - - = 2.16 X 10- 3 rad. (clockwise)
25000

I
62

Applyi ng
Since joint c is rigid, membe r ac will rotate the same angle nc.
the momen t-area method to membe r ac,

+ 18 ; 4.5 )
a - a = _1_ (18 X 1.5
a C EI

I 121.5
a, = EI (27 + 40.5 + 54)
EI
121.5
4.86 X 10- 3 rad. (clockwise)
25000

Applyi ng the momen t-area method to membe r ac,

a.
where xca is the deflect ion at c with respect to the tangen t at
I
x, = Xd = - [6 X 121.5 - (18 X 1.5 X 1.5/2 + 18 X 2.25 X 3)]
EI
587.25 587.25
- -- = - -- X 100 = 2.35 em . (to the right)
EI 25000

angle ad "
Since joint d is rigid, membe r db will rotate anticlockwise an
t. Thus,
Furthe r, since membe r db is free of momen ts, it will remain straigh
x b =xd+ 4ad
= 2.35 + 4 (1.44 X 10- 3) 100 = 2.93 em. (to the right )

at point d
ExaJD ple 1.33 Find the horizontal and vertica l displac ements
of the frame shown in Fig. 1.41 a. EI = 20000 m 2t.

x.
211m
16m!

r
b c
! 3m
6 d~

a
~ 4m-J
(a l
16
( bl (el

Fig. 1.41
 63

Solution: The B.M.D. is first drawn and will be as shown in Fig. 1.41 h.
From this diagram and noting the displacement conditions at the fixed
support, the elastic curve shown in Fig. 1.41c may be sketched.

Referring to Fig. 1.4lc,

Yd = Yc = Yl + Y2 and xd

~ Using the moment-area method,

I

I
" . - "b = - - - (16 X 6)
t EI

"b = 96/EI (clockwise)

I 288 X 100
Xl =lli(l6 X 6 X 3) = = 1.44 cm. (to the right)
20000
400 X 96
400" = 1.92 em.
b 20000

_I_(~X 4 X 16 X
3
-x 4)= 64 X 100 0.32 em.
EI 3 4 20000

Yd = 1.92 + 0.32 = 2.24 em. (downward)

"b-"c= _ _
EI
I_(~X4X
3
16)=-~
3 EI
117.33
"c =
EI

Since member cd is free of moments, it will remain straight. Thus,

300 X 117.33
x2 = 300 "c = = 1.76cm. (to the left)
20000

I xd = 1.76 - 1.44 = 0.32 em. (to the left)

1.15 Deflection due to shearing force

In the methods discussed in the previous sections, deflection due to
bending moments only has been considered, and the contribution of the
shearing forces to the final deflection has been neglected. This is justify-
able for most beams of practical dimensions as will be demonstrated
subsequently.

The formula for computing the shear deflection will be first derived . .
and then a number of standard cases worked out. In each case the
64

additional deflection arising from the shearing force will be calculated

separately and compared to the deflection due to the bending moment
assuming a finite value for the ratio of the shear modulus to the
modulus of elasticity; GJE = 0.4, which is approximately true for most
materials.

The variation of the shear stress on the section will be first neglected.
The shear stress q will thus be assumed as given by :

q = QlA ... 1.12

Consider now an element of a beam dx long and let the shearing

force over this length be Q as shown in Fig. 1.42 .

--.1
L dx --l
Fig. 1.42

Due to the positive shearing force indicated, the right face of the element
will slide downwards witb respect to the left face an amount dy,. The
shear strain y is given by :
dy,
y= 1.13
dx

Also, the shear strain y is related to the shear stress q through Hook's
law,
q
y= 1.14
G

Combining Equations 1.12, 1.13 and 1.14,

dy, Q
-
1.15
dx GA
 65

This is the differential equation of the elastic cu rve which when integrated
will give the d eflection due to shear Ys of any point along the beam. The
resulting constant o f integration is found, as usual, fr.om the displacement
boundary conditions .

. - SQdX 1.16
}, - GA

It can be shown that if the variation of th e shear stress on the cross-

section is taken into consideration~ equation 1.16 becomes:
QdX
y,~ k
S GA 1.17

where k is a constant called the shear shape fador. It depends on the

shape of the cross-section and has th e value of 6/5, 10/9 and 3/2 for
rectangular, circular and ho'low circular sect ions respective.y.

Eq uafon 1.17 may be re-stated as :

y, -
_ S"QdX
GA 1.18
,
where At = A/k is a reduced cross-sectional area, equal to 5Aj6 for
rectangular sections, 0.9 A for circular sections, 2A/3 for hollow circular
sections and the area of the web for I-sections.
In a member of uniform section, a study of equation 1.18 shows that
its right hand side represents the area of the S.F.D. divided by the shear
rigidity GA,.
As an example on the derivation of the shear shape factor. the rect-
angu lar section is considered. The proced ure is to find the internal strain
energy and then equate it to the external work done~' t Q d ys'

~
--tilY'd
J l
"I Q
J
d

[.-
dx --J
---.J
Lb-l
Fig. 1.43
'.
 66

Referring to Fig. 1.43, the shear stress at a fiber at a distance y from

the centroidal axis is given by :
QS
q =-
Ib
which on substituting for the value of the statical moment of area
S becomes:

q Q [""2
= 2I (d)2 _y2] 1.19

Since the strain energy due to shear per unit volume is q 2J2G" the strain
energy per elemental volume = b dy dx is :
q 2b
dU = - dxdy
2G

Substituting for the value of q from equation 1.19"

bQ2 (d)2 2
dU = 8 GP [""2 - y2] dx dy

Elemental strain energy for length dx along the beam,

dU -

On reduction this gives,

dU=~(Q2dX) 1.20
5 2 GA

Equating the shear strain energy in 1.20, to the corresponding external

work done t Q dy"

dy, ~ (QdX)
5 GA

Ys = ~
5
S Qdx
GA
1.21

A comparison between equations 1.17 and 1.21 yields the value of the
shear shape factor, and it is seen that it is equal to 6/5 for rectangular
cross-sections.
 67

ExaDlple 1.34 Calculate the shear deflection at the free end of a

cantilever under the two cases of loading shown in Figs. 1.44 a and h.

Pt wllm
~ ~
II ! [ I

,,
I

P
L

P wL
L

+ +
Ca} (b)

Fig. 1.44

Solution : The S.F.Ds. for the two cases of loading are as shown in Fig.
1.44. In case (a),

k JQdX = kPL
Y,
GA GA
PL3
Ym =
3 EI

For a rectangular section, k = 6/5, A = bd, and I = bd 3/12

Ym PL3 / kPL
= 10 G (~r
Y, 3 EI GA 3 E d

(~r
G 4
For = 0.4, Ym
E Y, 3
kwL2
In case (b), Y, -
2GA
wL4
Ym -
8 EI

~ (~r
Ym 5
-
Y, 2

For
G
E
0.4 , Yro = (~r
Y, d

Thus, it is ~een that Ys is of the order of 1% of Ym for a span

depth ratio Lid = 10.
68

Esample 1.35 Calculate the maximum shear deflection in a simple beam

under the two cases- of loading shown in Figs. 1.45 a and h.

wtlm
I ! I !

f' L 't
P12 t---,-----4 wLl2
+ +
L-_-_ ----'P/2 wU2
(a) (b)

Fig . 1.45

Solution: The S.F.Ds. for the two cases of loading are as shown in Fig.
1.45. In both cases the maximum deflection occurs at mid-span .

kPL PL3
In case (a), y, and Ym --
4GA 48 EI

In case (b), y, kC X
wL X ~) = kwL2
GA 2 2 2 8 GA
5wL4
and Ym 384 EI

For rectangular sections and G/E = 0.4, ym/y, :31 ( dL )2 and

- 5 ( -L-
12 d
)2 for cases (a) and (b) respectively.
From the standard cases given above, it is obvious that the deflection
due to shearing force is very small in comparison to that due to bending
moment except in the case of very deep beams with short span; small L id
ratio.

I. 16 lliustrative exatnples
In the following, a number' of illustrative examples is solved utiljsing
the methods presented in the previous sections for deflection computation .
The student is advised to solve these examples independently and then
check his results against those given . There are three possible sources
of confusion . These are:

-
 69

(1) The determination of the areas and centroids of the B.M.Ds' J

particularw for complicated cases of loading. This is usually overcome
by using B.M.Ds. drawn in parts; the area and centroid of each part of
which may be easily determined.
(2) The directions of slopes and deflections. A sketch of the elastic
curve of the deflected structure always helps in this respect. Sometimes,
however, it is not obvious which way the beam will deflect or rotate. In
such cases, the sign conventions for the method used should be strictly
followed.
(3) The units used. This may be the main source arerrar as the various
terms involved are usually expressed in different units. For instance, the
load is usually expressed in t., the moment of inertia I in cm 4 ,thc modulus
of elasticity E in t/cm2~ and the distances along the beam in m.

To illustrate this point, consider a cantilever acted upon by a concen-

trated load at the free end. The expressions for the slope and deflection
at the free end have been previously found and are as follows
PL2 PL3
a = and y =
2EI 3EI
Let the units of the various terms are those commonly used~ i.e. P (t.)~
L (m) ,E (t/em 2) and I (cm 4 )then the substituted values must be multiplied
by 10 4 in order to obtain the slope in radians and by 10 6 to obtain the
deflection in centimeters.

Further the student should notice that the solution of a single problem
is frequently sim plified by the use of morc than one mcthod for deflection
computations.

ExaInple 1.36 Calculate the magnitude and direction of the deflection

at the free end of the cantilever shown in Fig. 1.46 if it is made of a single
steel angle L 80 X 80 X 12.

0.
51
l
I~----'-
_ 1m --1

(a) (b) (c)

Fig. 1.46
70

Solution: For the equal angle used, the principal axes u and v
make angles of 45 with the legs of the section. Also., from the tables for
standard steel sections Iu = 161 cm 4 and Iv = 43 cm 4 .
Resolving the applied load into. two. components along the principal axes,
each component will be of magnitude P/../2'. Referring to. Fig. l.46c, .

PL3 0 .5 (100) 3
Yu = = 1.38 cm.
3 J 2 EI. 3 J 2 X 2000 X 43
PL3 0.5 (100)3
y" = 0.37 cm.
3 J 2 EIu 3 J 2 X 2000 X 161
y = J 1.38 2 + 0.37 2 = 1.43 cm.
0.37
tan a = yv I yu = 1.38
- = 0.2671 a = 15

The deflection at the free end makes an angle f1 with the vertical axis
where fJ = 45 - 15 = 30

Exa.nple 1.37 Calculate the horizontal and vertical displacements at point

d of the structure shown in Fig. 1.47a if EI = 3000 m 21.

d 4t

6m
b
1'i 2
JI
-1
(a)

---=::::::::
---------------------~
(b) B M.D.
~,

(c) Elastic Curve

Fig. 1.47
 71

Solution; The B.M.D. is first constructed and will be as shown in Fig.

1.47b. From this diagram and a knowledge of the supporting condition s
the elastic curve shown in Fig. L47c may be sketched. Since ther(~ is
no change in the length of member cd, the vertical disp acement at point
d is the same as that at point c. Referring to Fig. 1.47c,

Yd = Yo = y, + y,
Using the elastic-load method,

ab
_1_
= EI (~X
3 2
6X6) 12
EI
(anticloekwise)

200 X 12
y, = 200 nb = = 0.8 em . (upward)
3000

Using the moment-area method,

I 12 X 100
Yz = EI (6 X 2 X I) = 3000 0.4 em. (upward)

Yd = 0.8 + 0.4 = 1.2 cm. (upward)

Referring again to Fig. 1.47 c,

Using the moment-area method,

I
EI (6 X 2)

24 24 . .
ao =- I.e. - (anttcloekwlse)
EI EI
150 X 24
150 no = 1.2 em. (to the left)
3000

_1_
EI
(6 X 1.5. X I
2 )
=
4.5x100
3000
= 0.15 em. (to the left)

1.2 + 0.15 1.35 cm. (to the left).

Exalllple 1.38 Fig. 1.48 a shows a crossseetion of a long water tank.

Considering a strip I m. long, calculate the change in the width of the
tank at the water level. (Neglect own weight of tank) Yw = It I rn 3 ,
In 1 = 3000 m 2t and El z = 4500 m 2t .
72
c d
r
3 11
Wate r
-

lt1m3 3
L a
I
b 3 3
6m ~ (b) load D.
(a)
-=l t I=x2
x

4.5k\-----r----H.4.5 ()(o

~;:=:~
0(0

13.5mt
(d) Elastic Curve
(c) 8.M.D.
Fig. 1.48

Solution: The water in the tank will produce a uniformly distributed load
on the base of the tank of intensity = 3t/m. In the same time the hydro-
static pressure will produce a uniformly varing load on the walls of the
tank. The load intensity varies from zero at the water level to a maximum
value of 3t/m at the base of the tank. The load distribution on the sides
of the tank is thus as shown in Fig. 1.4B b. The corresponding B.M.D.
is shown in Fig. 1.48 c. From this diagram and a knowledgf of the SUP4
porting conditions" the deformed shape of tne tank may be sketched and
will be as shown in Fig. 1.48 d. Referring to this figure, the change in the
tank width = 2x.
Using the elastic~load method,

aa = _1_
El2
(..!.. X 3x13.5 -
3
4.5 X 3)
13.5
El2
(clockwise)

300 X 13.5
= 0 .9 em. (to the right)
4500
Using th e moment-area method and noting the properties of area of the
third degree parabola; A = 3/4 ab, y= 4/5 b from top,

I (3 4 ) 24.3
x2 = Ell 4 X 4.5 Y :: >: ""5 X 3 = Ell
24.3 X 100
= 0.81 em. (to the left)
3000
 73

x = 0.9 - 0.81 = 0.09 em. (to the right)

Thus, the.e will be a decrcase in the tank width = 0.18 em.

Example 1.39 Calculate the horizontal and vertical displacements

at point e of the structure shown in Fig. 1.49 a ifEI = 9000 m 2 t.

12

a
1.6 tim J 12
e
I-- 3 ---L-- 3 m _
1.8
25.2ml
(a) (b) B.M.D.

lYtlXfh
a b .)'l Y.

ra
..... -,.."
- ,/
,/
,/
/.
:ru
rb
3.6 3.6
19.8 37.8
(e) EIGsfu toads (d)Elasie Curve

Fig. 1.49

Solution: The B.M.D. is first drawn and will be as shown in Fig. 1.49b.
The elastic loads on span ab are shown in Fig. 1.49 e. A sketch of the de-
formed shape of the structure is shown in Fig. 1.49 d.

Referrmg to Fig. 1.49 c,

I 29.4
r
= 6- EI
(3.6 X 1.5 + 37.8 X 2 + 19.8 X 4 + 3.6 X 4.5) = -EI
I 35.4
rb - - - (3.6
6ID
X 1.5 + 19.8 X 2 + 37.8 X 4 + 3.6 X 4.5) = - -
EI
74

Applying the elastic-load method,

I 6
<10 = EI (29.4 - 3.6 - 19.8) = EI (clockwise)

Since there is no change in the length of member de, the horizontal

displacement at point e is the same as that at point d. Referring to
Fig. 1.49 d.
x. = Xd = XI + X2
300 X 6
"I = 300 "0 = 9000 - 0.2 em.

Applying the moment-area method to member cd,

I 54 54xIOO
x, = EI (12x3xI.5)= EI = 9000 = 0.6 cm.

X. = 0.2 + 0.6 = 0.8 cm. (to the right)

Referring again to Fig. 1.49 d,

Y. = YI + Y2 + Y3
I 63
YI = Yo = - (29.4 X 3 - 19.8 X I - 3.6 X 1.5)
EI EI
63 X 100
= 0.7 em.
9000

Applying the moment-area method to member cd,

I 42
= - - (12 X 3) , ad = - (clockwise)
EI EI
150 X 42
Y2 = 150, ad = 9000 = 0.7 em.

Applying the moment-area method to member de,

Y, = _1_ (12 X 1.5 X I) = 9 X 100 0.1 em.

EI 2 9000
Yo = 0.7 + 0.7 + 0.1 = 1.5 cm. (downward)
 Part 2
The Method of Virtual Work

I. 17 Introduction
The method of virtual work, which is sometimes referred to as the unit
load method, is very useful in determining deflections of structures. It
is most general in that it is applicable to all types of structures; beams,
frames, arches, trusses and composite structures and that it is used to deter-
mine load as well as no-load deflections such as those resulting from
temperature change, creep or flow of the material of construction and
fabrication errors.

1.18 The principle of virtual work

This principle is based on the principle of virtual displacements which
was first developed by Bernoulli in the beginning of the eighteenth century. \
It may be stated as :
"If a stmcture is in equilihrium under a system oj loads, henciforth referred to
as Q-load system, and remains so when it is subjected to a small virtual deformation,
the external virtual work done by the Q-load J)lstem as it moves through the virtual
displacement is equal to the internal virtual work done by the internal straining actions
in the structure produced by the Qload J)lstem as they move through the virtual
deformation" .
The term virtual deformation means that the action producing the
deformation is independent of the Q -Ioad system or that it is caused by
some additional action. Such action may be another load system, hence-
forth referred to as P-Ioad system. . temperature change... fabrication erroTS,
or in general any calculable effect. Also, the term virtual work refers to
the wGrk done by the Q.load system during the virtual deformation.
Again it should be emphasized that the external virtual work is the
work produced by the Q-Ioad system only as it moves through the virtual
displacement, and that the internal virtual work is the work done by the
internal straining actions produced by the Q-Ioad system as they move
through the virtual deformations.
The principle of virtual work may be thus expressed mathematically as :
... -1.22
76

where Wo and Wi represents the external and internal virtual work

respectively.
This expression is in accordance with the principle of conservation of
energy; the external work done is equal to the internal strain energy
stored in the structure.
Equation 1.22 is the basis of the method of virtual work for deflection
computation. With some ingenuity, it may be used to calculate any de~
flection component at any point of a given structure . This is done by
choosing a Q;load system consisting of a single load of unit value and
placed on the structure at the point where the deflection is required and
in its direction before the structure is subjected to the given P-Ioad system.
The P-load system will then be co nsidered as the source of the virtual dis-
placement, which in effect is the actual displacement required.
Because of this procedure, i. e., choosing the Q-Ioad system as a unit
load, the method of virtual work is sometimes called the unit load method.
The procedure of analysis may be best considered in regard to specific
types of structures. Before doing so, however, it will be advantageous to
develop relationships between the internal straining actions and the corres..
ponding deformations.
1.19 Straining actions-deformation relationships
The deformations of a structural member depend on the internal
straining actions in the member, the mechanical properties of its material
and the geometrical properties of its cross-section.
In the following, deformations resulting from the various straining
actions will be considered separately.
(al Deformation due to axial force, F
If a member is subjected to an axial force Fit will undergo a change
m length dS. This change in length may be elongation or shortening
depending on whether F is tensile or compressive.

F r-- - - - - - - - - - - , - - , F
~ 1...--
I. L W
Fig. 1.50

Consider a member of cross-sectional area A, length L and let it

be subjected to an axial force F as shown in Fig . 1.50. The normal stress
f is related to the normal strain through Hook's law,
 77

f
E
1.23

and for a member subject to an axial force, the stress is uniform and given
by:
F
f 1.24
A
From equations. l.23 and 1.24, the total change in length dS is given by :
FL
dS = 1.25
EA
(b) Deformation due to bending moment, M
If an element of a member dx long is subjected to a constant bending
moment M" the end cross-sections will undergo a rotation da as shown
in Fig. 1.51.
Referring to this figure, the rotation da which is equal to the change
in slope of the elastic curve in length dx is readi ly .obtainable from the
first moment-area theorem. Thus,
Mdx
da = 1.26
EI

dx

Fig. 1.51

where I is the moment of inertia of the cross-section about the axis

of bending.
 78

(e) Deformation due to shearing force, Q

If an element of a member dx long is subjected to a constant shearing
force Q the end cross-sections will slide with respect to each other a dis-
tance dys as shown in Fig. 1.52. The shear deformation 'Y is given by :

Fig. 1.52

dy, q
y = - = - 1.27
dx G
Assuming the shear stress to be uniform"
Q
q 1.28
A
Combining equations. 1.27 and 1.28,
Qdx
dy, = GA 1.29

If the variation of the shear stress across the section is taken into considera-
tion, the relative slide is given by :
Qdx
dy, = GA 1.30
,
where At is a reduced cross-sectional area and depends on the shape of
the section (section 1.15).

(d) Deformation due to torque, T

If an element of a cir~ular shaft dx long is subjected to a constant
torque T as shown in Fig. 1.53, the two faces of the element will rotate
with respect to each other an angle de. This angle of twist has
been derived in equation 9.20 of Part I,
Tdx
dB= - 1.31
GJ
 79

t( f)0 L-dx----1
Fig. 1.53

1.20 Truss deflections

Consider a truss loaded by a P-load system as shown in Fig. I.54a and
suppose the vertical deflection at jo:nt c be required.

P,

p.
(a) P-load system

(b) Q - load system

Fig. 1.54

The procedure of analysis will be as follows :

(1) The P-Ioad system is removed and a Q -Ioad system consisting of
a unit vertical load is placed at point c- Fig. L54b.
The unit load causes a set of internal forces in the truss members.
These forces are denoted by F I; the subscript 1 indicating that these forces
are produced by the unit load.
(2) The P-Ioad system is added to the truss already loaded by the unit
load . These loads produce another set of forces in the truss members.
These forces are denoted by F. ; the subscript 0 while meaningless at-this
 80

stage will be of significance later on in the analysis of statically indeter-

minate trusses. The Fo-forces cause internal _ deformations which, in
the particular case of trusses, are in the form of changes in length of the
various members. The change in length of individual truss members is
readily obtainable from equation 1.25,
FoL
d8 = - 1.32
EA
Also, the Fo-forces cause external displacements of the various joints of
the truss including joint c where the unit load is placed . This displacement
is denoted by Sc'
(3) Assuming the unit load to be the original load and that the P-Ioad
system to be the source producing the virtual displacement then,
to tal external virtual work = 1 X Sc 1.33
internal virtual work in each member = FI do 1.34
total internal virtual work 1.35
where the summation in equation 1.35 is carried out for all the truss
members.
Equating the external virtual work in equation 1.33 to the internal
virtual work in equation 1.35..
I X 8e = L F\ dS 1.36
Substituting for the values of dS from equation 1.32, equation 1.36
may be expressed as :
FoL
LF - 1.37
lEA
The change in length dS may be either elongation or shortening
depending on the sign of Fo- IfFo in a member is tensile it elongates and
dS is considered positive and conversely, if the force is compressive the
member shortens and dS is considered negative.
Equation 1.37 may be used to calculate any deflection component at
any joint due to a given P-Ioad system. The Fj -forces considered, how-
ever, should be those resulting from an appropriate Q-Ioad system.
Figs. 1.55 a -c show Q-Ioad systems suitable for the determination of
various deflection components that may be required. Fig. 1.55a shows
the Q-Ioad system for the calculation of the vertical deflection at joint d.
Fig. 1.55 b_ shows the Q-Ioad system for the calculation of the horizontal
displacement at the roller support b. Fig. 1.55c shows the Q-Ioad system
for the calculation of the displacement at joint c'in a chosen direction c'e.
 81

(a)

( b)

(c)

."

(d)
11
/
/
ti lIh

Fig. 1.55

Sometimes, especiaBy in cases concerned with the analysis of trusses

which arc internally statically indeterminate, need calls for the calculation
of the relative displacement between two joints. In such a case, the
Q -load system consists of two opposite unit loads applied at the two joints
and along the line joining them as shown in Fig. 1.55 d. This is obvious
as in this case the external virtual work = 1 X 3d (----") + 1 X 3e' (K)'
But since Sd and Set occur along the same line and in opposite dire-
ctions, the external virtual work = 1 (3d - Se') = Sde' which is the
relative displacement ofjoints d and e' along line de'. Also, it is sometimes
desired to calculate the rotation ofa truss member. In this case the Q-Ioad
 82

system consists of a unit couple or two equal and opposite loads normal to
the truss member, each of value of unity divided by the member length
as shown in Fig. I.SSe. This follows from the previous case. If two opposite
horizontal unit loads are applied atjoinls band b', the resulting displace-
ment will he the relative horizontal displacement between the two joints.
This displacement when divided by the length of the member h will give
its rotation. The same result could be obtained directly by selecting the
Q-Ioad system as consisting of two equal opposite loads each of value
of ljh.

As an illustration of the application of the method of virtual wotk to

the determination of truss deflections due to a set of applied loads,
a number of examples will be considered. On applying equation 1.37, a
tabular form for the solution is the best way of recording all the quantities.
A typical form is shown in Table 1.2. L and A represent the length and
cross-sectional area of the associated member listed in the first column.
Fo and Flare the forces in the same member due to the applied loads and
the unit load respectively. If E is constant for all the members, as is us-
FjFoL
ually the case, the sum of all the terms in the last column, ! - - - ,
A
is first found and then divided by the value of E to determine the
required dhplacement. Further, if the ratio L/A is assumed to be
constant for all the members, which is a common procedure in the
preliminary analysis of statically indeterminate trusses, the quantity
I: F, Fo is first found and then multiplied by a factor LjEA to determine
the rc:}uired disp'acement.

L A F, F,FoL
Member Fo
A
-
-- - - - - --
- - -
- - - - - -
- - - - - -
- I: F,FoL
A
Table 1.2

Particular attention should be paid to the signs. Fo and Fl are

considered positive when tep-sile and negative when compressive. If the
final result for the required displacement comes out positive, it occurs in
the direction of the unit load otherwise it occu rs in a direction opposite
 83

to that of the unit load. Further, if as usual the member forces are
expressed in tons, the modulus of elasticity E in t fem 2 and the
cross-sectional area A in em 2.. the unit of the displacement is similar to
that used in expressing the member length L.

I. 21 Applications to truss deflections

Exantple 1.40 For the truss shown in Fig. 1.56a) determine the vertical
deflection at joint c and the displacement at joint c relative to joint e' along
ee'. The cross-sectional areas.. in em 2.) are as indicated and E= 2000 tjern 2.

Solution: The F.-forces are shown in Fig. 1.56 b. The Q-load system
for calculating the vertical deflection at joint c consists of a unit vertical
load applied there. . and the corresponding member forces are readily
obtainable from F I = F./24. The Q-Ioad system for determining the

c' 20cm2 d' 20

\0<) 25 <$ 'l,.<' ~o

25
~__2~5~~~2~5__~__~2~5__~~25~~b
c d p
4 _~____ 3x4=12 m
241
- 161 -16
'10 '10
,~<) 0
,<)
0
.24 .8 .8
241
(b)F;,-Forcps
11
'0
---
-0.471 -0.47

-----
'$6' 0 _-~'O
o -0.35 ,<). 0 0
0 0 0
-0.94
11
(c) F,'-Forcps

Fig \.56

relative displacement of joints c and e' along line ee', together with the
corresponding member forces (Ft l-forces) are shown in Fig. 1.56 c. The
calculations are presented in Table 1.3.
 84

Member L A Fo F, F', F,FoL F;FoL

m cm" t. t. t. ~ -p;:-
ac 4 25 +24 +1 0 3.840 0
cd 4 25 +24 + 1 -0.94 3.840 -3.60
de 4 25 +8 +0.33 0 0.427 0
eb 4 25 +8 + 0.33 0 0.427 0
ac' 5 40 -30 - 1.25 0 4 .700 0
c/d' 4 20 - 16 - 0.67 -----0 A 7 2.140 +1.50
d'e' 4 20 -16 -----0.67 -----0 A 7 2.140 +1.50
e'b 5 40 -10 - 0.42 0 0.525 0
cc' 3 25 +24 + 1 - 0.35 2.880 -1.01
dd' 3 25 0 0 0 0 0
ee' 3 25 0 0 0 0 0
c'd 5 25 -10 -----0.42 +0 .58 0 .840 - 1.16
e'd 5 25 +10 +OA2 -----0.58 0.840 -1.16
22.6 -3.93
Table 1.3

Column F0 represents the forces in the truss members due to the

applied load, column F 1 gives the forces resulting from a unit load acting .
downward at joint c and column F'l gives the forces resulting from two
opposite unit loads as shown in Fig. 1.56c. The other columns in the
table are se1fexplainatory. The value orthe vertical displacement at joint .
c, 0CJ is obtained by dividing the algebraic sum of all the values in column
F,FoL/A by E. Since the length is expressed in meters, this value should
be multiplied by 100 to get the deflection in centimeter. Thus;
22.6 X 100
Sc = = 1.13 cm. (downward)
2000
Similarly the displacement of joint e' relative to c along line e'c, Se'c is
given by :
3.93 X 100
- 0.196 cm .
2000

The negative sign indicates that the assumed direction for the displacement
of e' relative to c is incorrect and that the two j oints move away from
each other.

Exan1ple 1.4 1 For the t r uss shown in Fig. 1.57a, calculate the vertical
deflection at joint c and the horizontal displacement at the roller support
b. All tension members are 10 em 2 and all compression members are
16 em" in area. E = 2000 t/cm".
 85

"0 0 0
0
,\(J -'l . 6,
11
lOt
lOt
(clF~Forces
(al (bl F:-Forcel;
0 I

Fig. 1.57

Solution: Since Land E are constant for all the members, and the cross-
sectional area is decided by the type of stress in the member, the calcula-
tion may be presented as in Table 104.

Type of Area Fo Fl F'1 FIFo F'I F O

Member
stress t. t. 1\ A
-em'
t.
- ----
ab - 16 -10 -I -1.16 +0.625 +0.725
be - 16 -10 -I 0 +0.625 0
ad + 10 +10 +1 0 +1.000 0
de + 10 +10 +1 0 +1.000 0
bd - 16 -10 -I 0 +0.625 0
+3.875 +0'725
Table 1.4

F 0 represents the forces in the truss members due to the applied

load (Fig. 1.57b), and F 1 represents the forces in the members due to a
unit downward vertical load ate. In this particular case Fl=Fo/lO. f't
represents the forces in the truss members due to a unit from right to left
horizontal load at b as shown in Fig. 1.57 c.

The value of the vertical displacement at joint c, 3c ' is obtained by

FF L
multiplying the sum of all the values in column _1_ 0 by a factor - .
A E
Thus,
3.875 X 250
0.484 em. (downward)
2000
Similarly the horizontal displacement at roller b,Ob' is obtained by
F'IFO L
multiplying the sum of the values in column ---;::- by E Thus.

0.725 X 250
0.091 em (to the left)
2000
86

ExaDlple 1.42 For the truss shown in Fig. 1.58 a,calculate the horizontal
and vertical displacements at joint c. The applied load P is such that all
terrsion rtre:rnbers are stressed to I t/tm-2 and all compression members
are stressed (0 0.6 tJem 2. E = 2000 tJem 2.

1t

e
o -2.33

(b) F,-Forces
Fig. 1.58

Solution: In this case there is no need for calculating the member forces
due to the applied load as the change in length of the truss members, dS..
can be calculated directly from d8 = fLJE where fe = 0.6 tJem 2 and
ft = 1 tfem 2. The calculations are presented in Table 1.5. Zero members
under the given case of loading have been omitted.

Member
Length Type of fo F, F',
foF,L foF',L
m. stress fa tJem 2 t. t.
ad 4 + 1 +2.33 +1.33 9.32 5.32
df 4.236 + 1 +1.41 +1.88 5.91> 8.00
fg 2 + 1 +1.00 +1.33 2.00 2.66
ge 2 + 1 +1.00 +1.33 2.00 2.66
be 4 - - 0.6 - 1.00 --2.33 2.40 5.6:1
eh 2.5 - - 0.6 0 -1.67 0 2.50
he 2.5 - -0.6 0 -1.67 0 2.50
ef 3 - -0.6 -1.00 - 1.33 1.80 2.40
ed 3 - - 0.6 0 -1.33 0 2.40
bd 5 - -0.6 -1.67 0 5.00 0
Table 1.5 28.48 34.04

The Frforces are due to a unit from left to right load at c as shown
in Fig. 1.58 b, while the F'I-forces are due to a unit downward load at
c as shown in Fig. 1.58 c.
The value of the horizontal displacement at joint cc' is obtained
from dividing the sum of all the values in column foF lL by E.
28.48 X 100
3. = = 1.42 em. (to the right)
2000
 87

Similarly, the value of the vertical displacement at joint c, YC)

is obtained from dividing the sum of all the values in column foF' 1L by E.
34.04 X 100
Y. = = 1.702 em. (downward)
2000
Example 1.43 For the truss shown in Fig. 1.59a, calculate the
horizontal displacement at support b and the vertical deflection at joint
e. Assume L/A=6 cm- 1 for all the truss members and E = 2000 t/cm 2

21 21 21 21 21 21
c " ci
~~~~--~~~~~
f' ti 4t
f
4
L
2 c d E'

.. Via 4 x4=16 m
(a)
2 2 2 2

o
It

o \) 1t

(d) Ii - ForcE's
Fig. 1.59

"
88

Solution: The Fo-forces indicated in Fig. 1.5gb are the member forces
due to the applied loads. The F I-forces are the member forces due to a
unit horizontal left-ta-right load acting at b as indicated in Fig. 1.59 c,
while the F't-forces are those due to a unit downward load acting at e
as shown in Fig. 1.59 d. Since both E and L/A are constant for all the
members, summations :E FIFo and k.F'lFo are first found and then the
results multiplied by L/EA to obtain the required displacements. The
calculations are presented in Table 1.6.

Member Fo(t) F, (t) F',(t) F,Fo F'tFo

ca -2 J 5 /3 0 0 0 0
ad +4 J5/3 + J5/2 0 3.33 0
de +6 +ll/B +1/2 B.25 3
ef +6 +1O/B +1 7.50 6
fb +4 +9/B +1/2 4.50 2
cia' 0 0 0 0 0
a'd' -2 -3/B -1/2 0.75 I
die' -2 -1/4 -I 0.50 2
e'f' 0 - l iB - 1/2 0 0
f'b' +4 0 0 0 0
cc' - 2 0 0 0 0
aa' - 20/3 -3/8 - 1/2 2.50 3.33
dd ' -2 + l /B - 1/2 -0.25 I
ee' 0 + I /B + 1/2 0 0
If' +2 + I /B + 1/2 0.25 I
bb' -2 0 0 0 0
ca' +4J2/3 0 0 0 0
a'd + !OJ 2/ 3 +3J2/B + J2/2 2.50 3.33
d'a 0 - J2/B + .j2j2 0 0
e'f - J2i2 - J2/B -J2i 2 0.50 2
f'b -4 J 2 - J2/B - y f -2 / 2 1.00 4
31.33 2B.66
Table 1.6
--
31.33 X 6
b 2000
0.094 em. (to the right)

2B.66 X 6
S = 0.086 em. (downward)
2000
 89

Example I ..... For the truss shown in Fig . 1.60, calculate the relative
horizontal displacement between joints c and e. Assume L/A= 8 cm- 1
for all the members and E = 2000 tlcm 2.

1t
I
l.Sm
e'
l- d
11
e
t>

2 -l- 2 4m 4
Fig. 1.60

r Solution: The calculations are presented in Table 1.7. The F.-forces

are the member forces due to the applied loaels wbile the F,-fOrces are
\
those due to two opposite unit tensile loads applied at c and e. The'
calculations are presented in Table 1.7.

Member F. F, F,F.
ac + 5 0 0
cd +7.08 -1.25 -8.85
de +2.85 - 1.14 -3.05
eb 0 0 0
a'e' -3.33 0 0
c'd' -5.42 +1.25 ~.75
d'e' -5.70 +1.14 ~.10
e'b' -2.85 0 0
aa' -2 0 0
cc' -1.25 +0 .75 -0.94
dd ' +5.25 -1.125 -5.90
ee' - I +0.375 -D.37
bb' - I 0 0
a'c +1.67 0 0
c'd -D.33 - I +0.33
de' +2.67 -I -2 .67
eb' +2.67 0 0
-34.30
Table 1.7
90

34.3 X 8
Oce = - = - 0.14 em.
2000
The negative sign indicates that points c and e move away from each
other.

I.ll No-load deflections of trusses

A study of equation 1.36 shows that deflections may be calculated
regardless of the source of dS. In section 1.20, the source considered has
been a set of applied external loads but the change in length of members
dS may be due to a change in temperature or fabrication errors which
are usually referred to as misfits. These two factors will be considered
separately below.

(a) Effect of temperature change

Structural members are strained by change in temperature in the
absence of external loads. For a slender member as in the case of
truss members, only axial strains need be considered. The change in
length do of a member of original length L and subject to a temperature
change t is given by :
do at t L 1.38

where at is the coefficient of expansion of the material of construction,

e.g. at = 1 X IO-sr C for both steel and concrete. For a rise in-tempera.
ture, d3 is positive and for a temperature drop d8 is negative . ' According
to the E.S.S., a change in temperature of 20C should be considered in
the analysis of structures subject to temperature variations.
Substit uting for the above value of d3 into equation 1.36, the d eflection
at any point c due to temperature change is given by
I X So = :EF I (at t L) 1.39

where t he summation is carried out o nly for the members subjected to

a temperat u r e cha nge . .-
(b) E ffect of misfits
T russ members are sometim es fa bricated longer or shor ter than their
geometrical lengths. This is do ne either due to fabrication error s or pur
posely to produce a pre-specified deflection or camber. In practice, the
camber is produced by fab ricating a n umber of m embers additionally long
or short. i .e. all the selected members are fabricated longer, or shorter,
than their geometrical l,ngth by the same amou n t (say a).
 91

Thus, equation 1.36 may be expressed as :

I X 8. = I:F ,a 1.40
In contrast to this, members may be fabricated proportionally long or
short, i.e. the selected members are fabricated longer or shorter than their
geometrical lengths by amounts proportional to these lengths. If the
I ratio of the change in length to the original length is {3, equation 1.36 may
f be restated as :
1.41

I where the summations in either equations 1.40 or 1.41 are carried out
only for members suffering from a change in length. dS is considered
positive if the member is fabricated longer than its geometrical length
and vice versa.
Deflections due to other no-load effects such as creep or flow of the
material of construction may be calculated in a similar manner.

~ 1.23 Applications to no-load deflections of trusses

Example 1.45 For the tcuss considered previously in example 1.40
(Fig. 1.56), calculate tlie vertical deflection at joint d due to a drop in

I temperature of 180 in the bottom chord members and an increase in

temperature of 18C in the top chord members and end posts.
a, = I X 1O-'rC.
Solution: The calculations are presented in Table l.8. The Ft-forces
represent the member forces due to a unit downward vertical load at d.
Since both the truss and the change in temperature are symmetrical"
calculations ,are carried out for one half of the truss and the final deflection
at d is obtained by multiplying the result by 2. Further, members which
are not subjected to a change in temperature are omitted .

t (0C)
Member F, (t) L (em.) d8 = a,tL F,d8

ac +2/3 -18 400 -0.072 -0.048

cd +2/3 - 18 400 -0.072 -D.048
ac' -5/6 + 18 500 +0.090 -0.075
c'd' --4/3 + 18 400 +0.072 -0.096
-0.267
Table 1.8
8. = - 2 X 0.267 = - 0.534 cm.

The negaJive sign indicates that the deflection occurs in a direction

opposite to that assumed for the unit load, i.c. upwards.
92

Example 1.46 Ifit is required to provide a 5 em . camber at the centre

of the truss in Fig. 1.61 by fabricating the end posts and the top chord
members additionally long, how much increase in the length of the above-
mentioned members will be required? What would be the decrease in
the length of each of the bottom chord members to produce the same
camber?

I
4m

LQ~~~~~~~~~~__~,
e f 9
6.3 =18 m
Fig. 1.61

Solution : Each of the end potts and chord memb= is assumed to be

increased in length by 1 em. The calculations ale presented in Table 1.9.
The F 1~ forces are the member forces due to a unit upward force at e..

Member F, a F,a
ac' 5{8 +1 5{8
c'd' 3{4 +1 3{4
die' 9{8 +1 9{8
5{2
Table 1.9

5
2
+ 5 em.

The positive sign indicates tha t the camber is in the direction assumed
fo r the unit load, i.e. upwards. The calculated value of the camber
happens to be the required value. If this were not the case, the actual
increase in the length of the chosen members would be found by proportion.

Assume next that each of the bottom chord members be shortened

by 1 em. The calculations for the corresponding camber are presented
in Table 1.10.
 93

Member F, a F,",-
ac ----{).375 I 0.375
cd ----{).375 - I 0.375
de -0.750 -I 0.750
1.500
Table I.IO

8. ~ 2 X 1.5 ~ 3 cm.

This deflection corresponds to a shortening in each of the bottom

chord members of I em. For the camber to be equal to the
d esired value of 5 cm., each of the bottom chord members should
5
be shortened by I X - ~ 1.67 cm.
3

1.24 Beun and frante deflections

Consider a beam loaded by a Pload system as shown in Fig. I .62a,
'. and let the vertical deflection at point c be required.

w
I I I
,I
( 0 ) P- Iood sys tem

~~--------~c--------~~
(b)Q - Iood system
r ~

F g . 1.62

The procedure of analysis is as follows :

(I) The P-Ioad system is removed and a Q load system, which fM
the purpose at hand consists of a unit vertical load at c, is placed at the
point where the deflection is required - Fig. 1.62 b .

The unit load causes 3: set of internal straining actions along the beam
length. These straining actions, which generally consist of N.F., S.F. and
B.M., are denoted by N bQI and M 1 ; the subscript 1 refers to the unit
load.
94

(2) The P-Ioad system is added to the beam already loaded by the
unit load. The P-Ioad system produces another set of straining actions
No,Qo and Mo. The corresponding deformations in an element of the
beam dx long arc found from equations 1.25, 1.30 and 1.26.
Nodx Q"d", = Mod"
d8 = EA dy, = GA da EI 1.42
,
Also} the P-load system causes external displacements of the various points
along the beam length including point c where the unit load is placed.
This displacement is denoted by 0c'
(3) Assuming that the unit load to be the original load and that the
P-Ioad system to be the source producing the virtual displacement then,
total external virtual work = 1 X 0c
and since the work done by a force is equal to the value of the force times
the displacement through which the force moves and the work done by
a moment is equal to the value of the moment fmes the angle, in radians,
through which the moment rotates, the internal virtual work in an element
of length dx is given by :

and thus the total internal virtual work is given by

Wi = f ,d8
N + SQ,dY, +SM,da
where the integration is carried out over the entire length of the beam.
Equating the external virtual work to the internal virtual work,

I X 80 = J f J
N,d8 + Q,dy, + M,da

Substituting for the values of d8,-ayjand da from 1.42, equation 1.43

1.43

I X 80 =
J J
may be expressed as :

N~ EA
S
Nodx
+ Qodx
Q, GA, + M,
Modx
Er 1.44

In most members of practical dimensions subject to bending, however,

the first two terms on the right hand side of equation 1.44 which represent
the contribution of the normal and shearing forces to the total deflection
is small compared to the third which represents the deflection due to
bending moment, and is therefore neglected. Thus, for all practical
purposes, equations 1.43 and 1.44 may be restated as :
r
l 95
t
1 x Bc= f M,da 1.45

1 X Bc= S M, M~~X .. . 1.46

The principle of virtual work can be easily extended to the calculation

of rotations. If a unit moment is applied at point c where the rotation c
is required then,

11
d

a a b

(a ) (b)

c
II It

l-

I a b
f a
,

I (c)
"
(d)
~

...t 1 mt

t ---- --
11
c
11
c(:

I
I
a
( .. )
b
~

a
(f)

l
F
Fig. 1.63

~
96

Ix a
c
~
j M
I
Modx
- -
EI 1.47

where Ml is the bending momcnt in the structure due to a unit moment

applied at point c where the rotation is required.. and Mo is the bending
moments due to the applied loads.

In general~ equations 1.46 and 1.47 may be used to determine any

displacement; linear or angular~ at a'9Y point on a beam or a frame due
to a given P-Ioad system. The Mt-moments, however, should be those
resulting from the appropriate Q.-Joad system.

Figs. 1.63a-f show Q-Ioad syst.c;:ms suitable for the determination of

various deflections that may be n~quired. Fig. 1.63a is the Q-Ioad system
for the determination of the vertical deflection at point c. Fig. 1.63b. is
the Q -Ioad system for the determination of the horizontal displacement
at point d. Fig. 1.63c is the Q-Ioad system for the determination of the
rotation at point e. Figs. 1.63 d,e and f are the Q-Ioad system for the
determination of the relative vertical) horizontal and rotation at point
c respectively.
It remains to mention that, except in the simple cases of members
having constant cross-section and where the moment may be expressed
by a single equation which is valid for the entire length of the member,
a purely mathematical evaluation of integrals of the form given in
equation 1.44 is often lengthy and tedious.
A simple and a rather quick method for the evaluation of these inte-
grals is presented in the following section.

1.25 Graphical evaluation of integrals

This method is applicable to integrals of the ferm f MIMo dx provided
that at least either the MI or the Mo-diagrams are bounded by a straight
line while the other may be bGunded by any curve, broken line or a single
straight IiItf.

This condition is always met in deflection problems as the MI-diagram

is generally caused by a unit load or couple, and hence consists of segments
bounded by straight lines.
 97

y-

M, 0

01-----_ x

Fig_ 1.64

Fig. 1.64, shows a segment of a general Modiagram and the corres-

ponding Mcdiagram. An ordinate to the latter may always be expressed
as a linear function of x; MI = kx
Substituting this value for MI in the integral under consideration,

f M,Modx = k f xModx

A liltle thought ~ill show thatf xModx is equal to the first moment

orthe area under the Mo diagram about the y-axis; Ao-x. Thus,

S M,Modx = kA;
but x ~,

hence S M,Modx = AoM, .-. 1.48

Equation 1.48 indicates that S M,Modx is equal to the a rea Ao of the

Modiagram, which is bounded by any curve, multiplied by the ordinate

M J in the other straight line-bounded diagram under the centroid of the

I
!
Mo-d iagram.
The final sign of the integral is positive if both the Mo and the M I -
diagrams have the same sign, otherwise it is negative.
It may be noticed that the form of the integral considered differs from
r' that usually encount~red in the deflection problems;
l

I
 98

j MIM..dx'SN1NOdx
EI EA'
S QIQodXandfTITodX by the rigidities El,
GA, GJ
FA, GAr and GJ. This creates no problem in the case of prismatic
members as the deflection is simply obtained by dividing the result
of the graphical multiplication by the appropriate rigidity. If the
section of a member varies along its length but remains constant along
parts of this length, such as the case in stepped columns for instance" equa-
tion 1.48 still holds provided that it is applied to individual lengths of
constant cross-section. If, however, the section varies from point to point
along the member length, the suggested method does not apply. In order
to overcome this difficulty,the length of the non-prismatic member is divi-
ded into increments, each assumed of constant cross-section. This approxi-
mation is usually permissible and the degree of accuracy of the result
increases as the number of increments considered increases.
As an example on graphical multiplication, consider a member of
constant cross-section, length L and let the Mo and MJ-diagrams be as
sbown in Fig. 1.65.

parabola 1h
-- -- - t---_'""--r
~:--- I- b

M,.O.
c
-- -- -- --~
d
I L .1
Fig. 1.65
For ease of calculations, both the Mo and Mrdiagrams ar divided
into regular parts whose areas and centroids may be easily determined
as shown.

J J
M1M0dx

A special form of
2
= -Lh
3
(C+d)
- aL (2C d) + bL
2 + -3 -+-
23 -2(2d
-3+3c)
-

M,Modx which is often encountered in the analysis

of statically indeterminate structures is r Mt dx, where M,-diagram is

 99

again assumed to be bounded by a straight line. In such a case, equation

1.48 is still applicable, but a simplified formula may be developed .

~--- I>
Q
- --- --
r=- _
Q 1--- -
L
-- b

Fig . 1.66

Referring to Fig. 1.66 and applying equa tion 1.48,

S
fM 2 dx =
I

On reduction,
11233233
- ( -2a + b- ) + -
M M dx = aL - + -a )
bL (2b

The sign of the product ab depends on the relative signs of a and h. Thus,
referring to Fig. 1.67 b which shows a possible M,-diagram, and applying
equation 1.48,
;
J Mfd x = ~ (a
2
+ b
2
- ab)

Fig. J .67

1.26 ApplicatioDs to beam. and CraIne deflections

As an application to the principles presented in section 1.24 on beam
and frame deflections and also on the graphical evaluation of integrals,

, a number of solved examples is given.

100

Exanlple 1.47 Using the method of virtual work, ca lculate the vertical
deflections at points c and d a nd the rotation at point b of the overhanging
beam shown in Fig. 1.68a if EI = 4000 m 2 t.

Solution: The Mo.D . due to the applied loads is shown in Fig. 1.68 b.
The Q -Ioad systems for the determination of the vertical deflectiops at
c and d and the rotation at b, and the corresponding B.M . Ds .~ are shown
succession in Figs. L68c, d and e.

3, =
S M1ModX

5.625
-- =
EI
5.625
-
= -

-
1 [
EI

X
-

100
1.5 X 6
2
1.5
X -
2
+ 2 (1.5

0.14 em. (dow.nward)

X 3
2
2
X -
3
X 3
)]

EI 4000

3d =
j M'IModX
EI
1.125
= - 1[1.5
EI
-
3
1.125 X 100
X 1.5
26
+ -X
3
2
1.5 -
3x6 1.5]
-- x -
2 2

EI = - 4000 0.028 em. (upward)

 101

Q
b
= J M,"Modx = _1 [3 X6 x 0.5 _ 1.5 x 6 X ~]
E1 EI 2 2 3
1.5
EI

= -1.5- = 0.375 X 10- 3 rad. (anticlockwise)

4000

Exantple 1.48 For the cantilever beam shown in Fig. 1.69 a, calculate
the vertical deflections at points f and g, the ro tation at point e and the
relative rotation at hinge g if EI = 50000 m 21.

(a)

,,'
(I) M, D. >---i"-"=::::::------+--4+t------=~--'
+
1.5
Fig. 1.69

Solution: The Mo.D. due to the applied loads is shown in Fig. 1.69 b.
The Q~load systems for the required deflections ,'and rotations and the
corresponding B.M.Ds. are shown in succession in Figs. 1.69cf.
102

0, =
j M tM
odx
EI
=1-
EI
[ 2 (2-x6x36x-x
3
5
8
337.5 337.5
- -- = -- X 100 = 0.675 em. (downward)
EI 50000

M'tMOdX =
a

=
S EI
-
1 [7.5 X 3
EI 2
x 2-
3

-
2
-
X

X
3

12
3 X 12 (2
+- - -X

X 36
2

X
3

1.5] = -
7.5 + -15)
229.5
--
3

3 EI
229.5 X 100
=-'--'....:.-- = 0.459 em. (upward)
50000

a = SMt"Modx = _1_ [ _ 15 X 3 X 1 _ 15 X 6X ~+ 15 X 6X 1
EI EI 2 2 3 2 2

+ 7.5 2X 3 2
X - X
3
0.5 + 0.5 X
2
12 (2
- X 7.5
3
+1 )
- X 15
3
2 1] 68.25 63.25
- - X 12 X 36 X - X 0.5 = - --
3 2 EI 50000
= -1.365 X 10-3 rad., i.e. 1.365 X 1O- 3 rad. (clockwise)

a = SMi"Modx =
1 [15 X 6
EI - 2 - X 0.5 -
7.5 X 3 (2
- 2- ;; X 1.5 +1 X ;;
1)
EI
1.5 X 12 (2
2 3
- X 7.5 +1
- X 15) -
3
15 X 6
2
X -
1
3

+ 32- X 12 X 36 X -
1.5] 118.5 118.5
2
. 3
= - -= - - = 2.37x 10- rad.
EI 50000

EzaJnple 1.49 Considering the deformations due to B.M.,N.F. and

S.F., calculate the vertical deflection at point c of the reinforced concrete
frame shown in Fig. 1.70 if it has a uniform rectangular section
60 X 30 em. E = 200 tfem 2 and G = 80 tfcm 2
 103

211m
c b
4 41

t-
4m

ItQ 4 -I-..-- 6 m --!- 6 ------I.- 4 -----I

(a)

16ml
16
I 16 +
36 3ml

(b) Mo D (e) M, D.

121 0.51

[ -u 16
(c) No.D.
-t 0.5
(t) N,.D.

121 05 t
+ + +
j
12 0.5
-
4
(d) a".D. (g) Q,.D.
Fig. 1.70

r Solution: The B.M.,N.F. and S.F.Ds. due to the applied loads are as
shown in succession in Figs. I. 70 b - d. The corresponding diagrams

[ due to a unit vertical load at C are given in Figs. 1.70 e Mg.

l
 104

A = 30 x 60 = 1800 em 2 EA = 200 x 1800 = 360000 t.

5 ?
A, = - X 1800 = 1500 em - GA, = 80 X 1500 = 120000 t.
6

B = [ M,Modl [N,Nodl SQ,Qodl

b J EI +J EA + GA,

M, Modi = - 2 [2- x6x3 6 ( -x3

5 ) - -
3 X-6 X 252
EI
S EI EI 3 8 2 16] =

= 252 X 100 = 2.333 em

10800

I 56
= - ( 12 X 4 X 0.5 + 16 X 4 X 0.5) = -
EA EA
56 X 100
--::==-
360000
= 0.015 em.

6) ~
S Q,Qodl
GA,
= _1- [2 (12 X
GA,
36 X \00
2
X 0.5] =
GA,

0.03 em
120000

Be = 2.333 + 0.015 + 0.030 = 2.378 em

against 2.333 cm. due to bending strains only.
Example 1.50 For the frame shown in Fig. 1.71a, calculate the hori-
zontal displacement and rotation a t roller support b if EI = 20000 m 2t.
Solution: The Mo.D . due to the applied load is shown in Fig. 1.71b.
The B.M.Ds . due to a unit horizontal load and a unit moment at bare
shown in Figs. 1.71e and d respectively.

M,MOdl I [16 X 5( 2 + 2- x 5 x4 (6.2 + 2.4)

S _'.<
0 )
Bb = - - - 2.4 + - X
EI EI 2 3 3 2

+-
16 X-5 ( 4 + 32- X 2.2) + 2- X 5 X 4 ( 6.2 +- 4)] = -
541.3
--
2 3 , 2 EI
541.3
= - - - X 100 = - 2.7 em ., i.e . 2.7 em. (to theright)
20000
 105

2 tIm
I I I I I I ! I I I I I I I I I

!
3m 3
I
~
4
-+2.4
b-.1
L a t
L4 4mJ f
(a) (b) ~.D.

6.2mt
O.Smt
/

"-

0
t
1mt

(c)MtD. l (d)M;.D.

Fig. 1.71

=_1_ [M'tMo dx = __1_[5 x 16(~ ~) ~ 5 x 4 x 3-

a
b
EI J
EI EI 2 2+6 +3 x 4

+5 x 16 x ~+~x 5 x 4 X I..] = _ 53.33 =_ 53.33

.2 3 3 4 EI 20000

=- 2.67 X 10- 3 rad., i.e. 2.67 X 10- 3 rad. (anti clockwise)

 106

&auaple 1.51 For three-hinged frame shown in Fig. l.72a, calculate

the vertical deflection and the relative rotation at the intermediate hinge
c. The moment of inertia varies as indicated. I = 125000 em 4 and
E = 2000 t/cm 2

211m
21 c +
1 6 LSI .25 36
Lb
I--.. 6 --L 6 m--4-. 3 j
(a)
(b) Mo D.
11 1m!
1m!
+
3
,
(e) M,D.

Fig. 1.72

Solution: The Mo.D. due to the applied loads is shown in Fig. 1.72b. The
M1 .D. due to a unit vertical load at c is shown in Fig. 1.72 CJ and the
M'I.D. d ue to two opposite unit couples is shown in Fig. 1.72d. Consider-
ing I = I" EI, = 125000 X 2000 X 10-4 = 25000 m't.

8e =
S +
MIMO d l l [ 2 7 X 6
EI = EI,

27 + 45)
2 X 2 -

3 12 (27
2
3" X 6 X 2.25 X 1.5

+ 45)
+2 2
(
1 2 12 X 3
X
2 2
X -
3

5 2
X 6 X 36 X - X 3 - - X 12 X 36 X 3
~ I -+ - -36 X 6
X 2 X -
I ]
8 3 2 2 1.5
= 454.5 = 454.5 100 = 1.82 em . (downward)
EI, 25000 X

co = JM' IModl = _ 1
_ [27 X 6 X ~_ ~ X 6 X 2.25 X ~+~ (27 + 45)
e EI EI, 2 3 3 2 I 2
~ I
X 12 X I -
2
- X 12 X 36 X
3
I
2
- + 36 2X 6 2
X - X
3
-1.5I ]=--
169.5
EI,
169.5 3
= - - = 6.78 X 10- rad.
25000
 107

1.27 Deflections of composite structures

The use of the method of virtual work to determine truss deflections
on one hand and the deflections of beams and frames on the other hand
have been presented separately in sections 1.20 and 1.24. The distinction
has been made on the basis that truss members are subject to axial strains
while beams and frames afe mainly subjected to bending strains; the
contribution of the shear and axial..strains to the overall deflections being
usually neglected.

There are some structures, however, in which some parts are subjected
mainly to bending strains while others are subjected to axial strains only.
Such structures are called composite structures and in computing their
deflections the effect of both the axial and bending strains should be
considered.
The expression for deflection may be stated as :

1.49

Under load, equation 1.49 may be restated as :

J
M,Modl
I X 8c = + 1.50
EI
where 3c is the deflection component desired at any point c, Ml and F t
represent the bending moments and axial forces in the various members
of the structure as a result of an appropriate Q-Ioad system, and Mo and
F 0 are, as before, the bending moments and axial forces due to the applied
loads.

1.28 Applications to deflections of composite structures

Esample 1.52 For the.trussed beam shown in Fig. 1.73 a, calculate the
vertical deflection at point c taking the effect of thrust in beam acb into
consideration. E = 2000 tfcm'.
For beam ab, I = 37000 em' and A = 132 cm'.
For ties, A = 36 em' and struts A = 28 em'.

Solution: The forces in the link members Fo, the bending moment diagram
Mo.D., and the normal force diagram, No.D., due to the applied loads

r
are as shown in Fig. 1.72 b. The corresponding values F" M,.D. and
N,.D. due to a unit vertical load at c are shown in Fig. 1.73 c.
108

I 211m

2~~i""'@""~
rt 4 --L- 4 ---I-4m----l-- 4 -.J
(al
211m

~r
2ml
- ~
.~
2
M,.D.
321 32 21
I I I j
NoD. NrD.

(bl (el

128 128 x 100

= -= l. 73 cm. (downward)
EI 7400
The contribution of the axial strains in the link members to the deflection
at c is calculated in Table l.1l.

Member
L A F, Fo F,FoL
cm. cm 2 t. t. A
ad 447 36 + ..j5 + 16..j5 993
de 800 36 +2 + 32 1422
eb 447 36 + ..j5 + 16..jS 923
dd' 200 28 -I - 16 114
ee' 200 28 -I - 16 114
Table l.ll 3566
 109
F,FoL 3566
L -- = -- = 1.783 em.
EA, 2000

N ,Nodx = 2 X 32 X 16 X 100 = 0.388 em .

8c
S
=
EA 2
1.730
264000
+ 1.783 + 0.388 = 3.901 em. (downward)
Es:a.m.ple 1.53 A structu~e consists of a con tinuous member abc hinged
4
'J3
9t
4m

(a)

4
r b

L 1--3m ---1
a

51.t

Yo.....
---
7.2t 24.6t

:v
( b) -14.4 28.8mt

N
en
7.21
24.6
19.2~ ~D. 124.6 NO'D.
267t
-1.L-

~
.....
,q>
", .

(c) -2 4mt I
....
to
N
1t
1/
2.67~ F,.D. 2.67 MfD.

Fig. 1.74
 lIO

at a and braced by a system oflink members as shown in Fig. 1.74 a . De-

termine the horizontal deflection at c due to a pull of 9t. acting as
indicated. For abc, A = 250 em 2 and I = 140000 cm 4 , and for all link
members A ~ 25 cm 2 and E ~ 2000 tfcm 2 .

Solution: The forces in the link members F o, the bending moment

diagram, Mo.D. and the normal force diagram No.D. due to the applied
loads are as shown in Fig. 1.74 h. The corresponding values due to a unit
horizontal load at c are shown in Fig. 1.74 c.

Sc
~ J M,Modl
EI +
:E F,FoL
EA, +
f N,Nodl
EA2

~ ~ (28. 8x4 X ..! X 4) _ 307.2

EI 2 3 EI
307.2 X 100
1.10 cm.
28000

The contribution of the axial strains in the link members to the deflection
at c is calculated in Table 1.12.

L F, Fo
Member F,FoL
cm. t. t.

ed 400 +2.67 +19 .2 20480

de 500 +3.33 +24 40000
db 300 -2.00 - 14.4 8640
69120

Table 1.12

F,FoL 69120
..,
:E --~ ~ 1.38 cm.
EA, 2000 X 25

j NtNodl
EA2
2.67 X 24.6 X 800
2000 X 250
~ O.II cm.

S. ~ 1.10 + 1.38 + 0.11 ~ 2.59 em. (to the right)

It is noticed that the contribution of the thrust in member abc to the

overall deflection at c is less than 5%.
 III

Example 1.54 A frame is provided by a tie across tbe eaves as shown

in Fig. 1.75 a. Determine the horizontal displacement at roller b neglect-
ing the effect of thrust in members resisting mainly bending moments.
E = 2000 t/cm 2 Jor frame EI = 20000 m 2 t., and fo r tie A = 20 em 2

211m

f
3

1
6m

Q b
I....- 6m 6 ---I

(a )

F, =3 1
F. =121
r

t (c) M, .~Ds.

Fig. 1.75

Solution: The bending moment diagram Mo.D. and the force in the
tie Fo due to the applied loads are shown in Fig. 1.75b. The correspond-
ing values due to a unit horizontal load at b are shown in Fig. 1.75c.

ab =
J M,Modl
EI

I
F,FoL
+E~

20000 (2 X
2
3""
X 6.72 X 9 X 3) 100
3 X 12 X 12 X 100
2000 X 20

- 1.21 - 1.08 - - 2.29 cm. i.e. 2.29 em. (to the right)

Notice that the contribution of the elongation in the tie to the total
deflection at b amounts to 47%.
112

2 21 (c) M,,~.Ds.

Fig. 1.76

Solution: The bending mement diagram, Mo.D ... and the forces in the
the link members Fo due to the applied loads are shown in" Fig. 1.76 h.
The corresponding values due to two opposite unit loads at joints e and
f are shown in Fig. 1.76 c.
MIMod! FIFoL
EA
8" =
S-----pJ - + L

o + ~ (1 X ~- 3 X ~ + 2 ..)"2 X ..)"2 _ 2)2 X ..) 2)

EA 4 4 2 2
1.5 L 1.5 X 10 2
- EA" = - 2000 = - 0.75 X 10- em.
The negative sing indicates that joints e and fmove towards each other.
 113

1.29 No-load deflections of beaJJ1s and fraJJ1es

Deflections of beams a.nd frames ma.y take place in the absence of
load. Two of the possible causes of these deflections will be considered
separately below.

a) Effect oftemperature change

As has been demonstrated in section 1.22, structural members are
strained by temperature changes in the absence of external loads. When
a member is subjected to a uniform change in temperature, an assumption
which is usually made in the case of slender members, it only elongates
-or shortens according to whether the temperature rises or drops. On the
other hand, if a member is subjected to a differential temperature change"
i.e. the temperature of its top fibers differs from that of the bottom fiber~,
cross-sections undergo relative rotations da in addition to a possible
change in length dB.

O<,l,dx
n
0<' d.
I I
~~2 I ~'TI
I, I,

do, .1 h \ Y
- -- do , h
- 1---

'2
lh{Zl 12 j~
~ f-- d. I I ~ f--dx I
OC, '2dx
I
0<, '.dx
(a) (b)

Fig. 1.77

Consider an element of a member as shown in Fig. 1. 77a, and assume

its cross-section be symmetrical, i.e. the centroidal axis lies at the middle
of its height. Let the temperature of the top fiber be raised by t, and
that at the bottom fiber by t 2 Further, let the temperature along the
depth of the member vary linearly. The expansion of the top fiber in an
element dx long will be att l d x and that of the bottom fiber att1dx.
On the assumptions made, the ::entroidal axis will expand by an
average amount of d3 given by :

dB = at'; t2)dx 1.51

114

The relative rotation of the two cross-section.. da, is easily found from
the geometry of the deformed element and is given by :

da= a,C2;:-t, )dX 1.52

A study of equation 1.43 shows that deflection may be computed

regardless of the source of do and da. Consequently, this equation may
serve for the determination of displacements due to temperature variations.
Substitu ting for the values of d3 and da from equations 1.51 and
1.52 into equation 1.43; dys being zero as there i s no relative sliding of
the cross-sections due to a uniform change in temperature along a member
length" the expression for deflection caused by thermal strains is obtained.

3, = SM, a, C2 ~ t, ) dx + j a, C' ; t2 )
N, dx 1.53

Usually the terms at ( t2 ~ t, ) and at (t, ; t2) are constant for

individual members and equation 1.53 may be restated as :

3, = L a, ( t2 ~ t, ) J M, dx + La, (t, ; t2 ) S N, dx

A little thought will show that S M,dx and J N,dx represent the areas

under the Ml and N ,-diagrams respectively. Thus,

... 1.54

where the summations in equation 1.54 include all the members which
are exposed to temperature changes.
When the cross-sections of the member under consideration is unsym-
tr + t2
metrical with respect to the centroidal axis the term -'--"--=" in
2

equations 1.53 and 1.54 must be replaced by t, + (t2 - t,) ~ ,where

y is the distance of the centroidal axis from the top fiber as indicated in
Fig. l.77b.
The signs of the integrals in equation 1.53 or the products in equation
1.54 are positive if the thermal strains are similar to those produced by
the corresponding straining actions due to the unit load, otherwise they
are negative.
 115

b) Effect of support movements

I The effect of support movements on the deflectiorrs at vario u s poi n ts of

a statically determinate str ucture can be readily obtained from the
geometry of the deflected shape of the structure. On sketching this shape,
con t inuity of the structure must be maintained and the displacement
condi tions at the supports must be fulfilled.
A lternatively equation 1.22 <fllay be used. In this case, however, it
should be remember ed that any statically determinate structure is not
st rai ned due to support movements. The intern al strain energy stored in
t he struct ure Wi is thus equal to zero . Consequently the exter nal work
d o ne We by all the load s, inclulding the reactions, on the structure must
zer o . E qu at ion 1.22 thus reduces to :
W, = 0 1.55
In a pplying eq uat ion l. 55 care m ust b e t a ken to use t he p r oper signs
for t he wor k do n e by the reactions. A r eact ion does posi tive wor k if its
point of applica ti on m oves i n t h e same se nse . Co nversly, a reaction d oes
negative work i f i ts p oi nt of appl ication m oves in an op posite sen se.

1.30 Applications to no-load deflections of beams and fraDles

ExalDple 1.56 Calculate t he. horizontal d isp lacement at th e r oller support

r
4m

6
La b
8 8m ---<

(a)
10m\

2/(5\

(e) M,.D.
(b) N,.D.
Fig. 1.78
116

b in the frame shown in Fig. 1.78 a due to the following temperature

variations : ,
a) a uniform drop in temperature of 20 c.
b) an increase in the outdoor temperature of 20 0 C accompanied by
a drop in the indoor temperature of 10 C.

The frame has a uniform rectangular section 30 X 60 em.

ex, = I X 10-'tC.
Solution: The norma] force diagram N ID. and the bending moment
diagram MJ oD. due to a unit horizontal load at b are as shown in Figs.
1.78 band c respectively.

In case (a)" since there is no differential change in temperature, the

first term in equation 1.54 will vanish. Thus"

at (t. ; t2 ) l: An. = I X 10-' X 20 (2 X

2
.J5 X 8.96) X 100

0.32 em . (to the left)

In case (b), Il" is given by :

8b at ( t2 -;: t. ) l: 1\",. + at ( t. ; t, ) An.

I X 10-' ( 10 0~6 20) [6 ; 6+( 6 ~ 10 ) 8.96] X 2 X 100

8.96 + 0.08
+ I X 10-' ( 20 -; 10)

9.04 em. (to the left)

c; 52 X 8.96) 100

ExaInple 1.57 Calculate the horizontal displacement at the roller support

b in the frame shown in Fig. 1.79 a due to the following temperature
variations:

a) a uniform rise in temperature of 20 C.

b) a drop in the indoor temperature of 10 C accompanied by an

equal rise in the outdoor temperature.

The frame has a uniform section of depth 50 em., at = I X 10-'.' " ::-.
 117

r
4m

L
I

a
1 Bm

I
L 8m
(a)
J
8ml
0.5 t 0.5 4

+1

11 ~o. s 11 ~0 . 5

( b )N,.D.
to.s (c) M,.D.

Fig. 1.79

Solution: The normal force diagram N I.D. and the bending moment
diagram MJ.D . due to a unit horizontal load at a are shown in Figs. 1.7gb
and c respectively.
In case a, the first term in equation 1.54 will vanish. Thus,
So = I X 10- ' X 20 (0.5 X 4 - 0.5 X 8 _ . I X 8) 100
= - 0.2 em., i.e . 0.2 em (to the left)

In case b, the second term in equation 1.54 will vanish. Thus,

So =
20
I X 10-' X - --
(4 X4+ 6 X
8X8)
8 + - 2 - 100
0.5 2
= 3.5 em. (to th~ right)
U8

ExaIIlple 1.58 The three-hinged frame shown in Fig. 1.80a has a

uniform symmetrical section 80 em. deep. Calculate the relative rotation
at the intermediate hinge if the indoor temperature drops 16C relative
to the outdoor temperature. at = 1 X 10- 5 re.

4m

I
6

b
I--- 6 i ' (a) 12 m

6 ~,mt
1-~":)'
\

( /

1/4{13 t 1/4/10

11121 1112 1112 t 1112

(b)N,.O. (c)M,.O.

Fig. 1.80 J

Solution: Referring to Fig. 1.80b and c which show the N,.D. and the
M1.D. due to two opposite unit couples applied at c,

a, = at C' ; t2) LAn I+ at C' -;:- t, ) LAml

= 1 X 10-' X 8 ( J-'-
I
X 2
J-13 + JI - X
4 13 4 10

-X -8 +2x 5-X 2 J 13+-5

+lxlO - ' X16- (2 X 4.j1O)
0.80 3 2 6 6
0.0045 rad.
 119
ExaD1ple 1.59 For the he am shown in Fig. 1.81a, calculate the vertical
displacement at point c and the relative rotation at hinge d due to the
following support movements:
a) a vertical settlement of 1.2 cm. at support h.
I b) a clockwise rotation of 0.005 rad. at support a.

l " a) ~a
~
l--- 2
d o } - - - - - -- --b*----"-"c
---1_ _ _ 4m _..c;;06:a0"".0 - 2 J
1t!
(b) ~~I_~o~-------------.------J-
!0.51 t1.51
(c) Gto.251 ):*t-------------,----
ml
10.251
Fig. 1.81

Solution: The Q-Ioad system for determining the vertical displacement

at c and the corresponding reactions are shown in Fig. 1.81 h.
Effect of t:he settlement at b on the deflection at c,
We 1 X 8< - 1.5 X 1.2 = 0
8< 1.8 cm. (downward)
Effect of the rotation at a on the deflection at c,
We 1 X 8< . + (1 X 0.005) 100 = 0
8< - - 0.5 em. , i.e. 0.5 em. (upward)
The Q-Ioad system for determining the relative rotation at d and the
corresponding reactions are shown in Fig. 1.81 c.
Effect of the settlement at b on the rotation at d,
We (1 X ad) 100 + 0.25 X 1.2 = 0
ad - - 0.003 rad.
Effect of the rotation at a on the rotation at d,
We 1 X ad - 1.5 X 0.005 = 0
ad 0.0075 rad.
120

Exalllple 1.60 For the frame shown in Fig. 1.79 a, calculate the horizontal
displacement at the roller support a due to a downwm-d displacement at
support b equal to 0.8 em. accompanied by a horizontal outward displa-
cement there equal to 0.6 em.

Solution: The Q-Ioad system for determining th e horizontal displacement

at a and the corresponding reaction s arc shown in Fig . 1.79 h.
Wo 1 X B. - 0_5 X 0_8 - 1 X 0_6 = 0

B. I em_ (to the right)

Part 3

Graphical Methods for Truss Deflections

1.31 Introduction
If the deflect ion at a single joint o f a tfUSS in a speci fied direction is
required, then the method of virtual work already described is the most
suitable. However, if displacements at all or a number of j oin ts arc
required, as is the case in many st ructural problems, a grap hi cal method
known as the displacement diagram method, is morc a dva ntageous to
the method of virtual work. As will be shown subsequently, the graphi cal
method is easily carried out once the changes i n length of individual truss
members arc found. The change in length of a member is given by
Under load, d B = FL/EA 1.56
Due to temperature change, dS = at t L 1. 57
Due to fabrication error a, dS = a 1.58
Due to fabrication error fJL, do = fJL 1.59

It should be remembered that J like all graphical methods, this method

requires care a nd reasonable degree of accu racy in tracing an d scaling.

1.32 Williot diagralll

Co ns ider the simple truss shown in Fig. 1.82a. The applied load P will
produce tension in both members ab and ae thus e10nnating them beyond
their original length. Member bc will be in compression and will hence
shorten . Assume that the changes in lengt h of these members have been
calculated using any of equations 1.56 - 1.59_ The deformed shape of the
 121

~~ ____________~ti

(b)

c,
(c )Williot diagram
Fig. 1.82

truss, indicated to an exaggerated scale by the dashed li nes in Fig. I.82a~

is determined in the manner described below .

In this p.articular example and by virtue of the support conditions"

joint a is fixed in position and member ab does not r otate, i.e. it remains
fixed in direction after deformation. Now it is imagined that the members
are disconnected at joint c and the change in length in member ac is al-
lowed to take place. Since a is fixed in position and ab is fixed in direction
and that member ab is in tension, the only possible movement ofb is hori-
zontal to the right to h'; bb' being the pre-calculated value of the change
\0 length in member ab~ do ab , Allowi ng next the change in length in
rnembers ac and bc to take place, ends c in these members will move to
points CI and C z as indicated. It is noticed that, being in tension, end c in
ac moves away from a by the pre-calculated va lue of the change in length
in member ac, d8 ac ' Similarly, being in compression, end c in bc
moves towards b by the pre-calculated value of the change in length in
member be, d8 bc ' Actually, ends c in members ac and bc; points C I and
e2, must coinside. This is achieved by rotating members ac and bc about
points a and b respectively. Thus an arc using a as centre and ac as radiu s,
and another arc using b as centre and bc as radius are drawn . The inter-
section of these two arcs determines the deflected position of joint c; point
c ' . The deflection at any joint can now be determined by measuring the
distance between the displaced and the origina l positions of this joint.
It should be remembered that the changes in length indicated in
122

Fig . 1.82 a are greatly exaggerated. If the changes in length, which are
extremely small relative to the original lengths of the members were
plotted to the same scale as the truss, the deformed truss would almost
coinside with its original undcrformcd shape and it would be practically
impossible to measure.. to any acceptable degree of accuracy, the displace-
ments of the joints.
Now, this same cause which makes the graphical method impracticable,
ofrers the key to a method that permi ts the drawing of the changes in
length of the members to any convenient scale separately from the truss.

Since the main difficulty arises from the fact that the changes in length
arc vcry small in comparison to the original lengths of the members, then
for all practical purposes the tangents to the arcs may be substi tuted for
the arcs a nd thus ends c in members ac and bc may be assumed to move
perpendicular to members ac and bc respectively. The deflected position
of joint c is thus determined and the deformed shape of the truss is as in-
dicated in dashed lines in Fig. 1.82 b. It may scem that thc deflected posi-
tion of joint c in Fig. 1.82 a diffcrs from that in Fig. 1.82 b but
this is mainly due to the exaggerated scale used in representing the changes
in length of the members.
As arcs about centres of rotation are substituted QY normals to the
original directions of the members it is no longer necessary to draw the
members of the truss and it is possible to obtain relative joint deflections
by a simplified diagram involving only the changes in length and the
normals to the original directions of the members. Making use of this
fact the change in length may be represented to any convenient scale.
Such a displacement diagram is called the Williot diagram after the French
engineer who suggested it and is shown in Fig. 1.82 c. The procedure of
plotting the Williot diagram for the given truss is as follows:
A member which remains fixed in direction after deformai.ion and a
point, at either end of the chosen member, which is assumed to be fixed
in position are chosen. In this example member ab is fixed in direction
and end a is fixed in position. Denoting the points in the Williot diagram
by dashed letters of the corresponding truss joints, point a' is established
first at a convenient location on the tracing paper. Since ab remains fixed
in direction, joint b moves horizontally to the right relative to a (ab being
in tensi~n) and dBab is drawn horizontally to the right of a to a chosen
convenient scale. This dB is shown in Fig. 1.82c as vector ab'. Next.
 123

point c' which corresponds to th e rema ining joint c of tri a ngle abc in th e
truss is determined. Since member'ac is in tension .. end C of t his m ember
will move away r ela tive to a along a line p ar a llel to ac a s indica-
ted by vector a'et Similarly~ since member be is in compression, end c
of this m ember will move towards the pre located point h' along a line
parallel to be as indicated by vector b'C2o Point c' is finally found by draw-
ing perpendicula rs to members a c and be a s shown in Fig. 1.82c. Vectors
a'b' and a'e' thus represent the deflections of joints band c relative to
joint a . And since in this particular example joint a is actually fixed in
position. . and member ab remains horizontal after d eforma tion, these
vectors represent the absolute deflections at these joints. Of course, the
lengths of these vectors should be measured to the same scale as tha t used
in representing the dS-values.
The same procedure ca n be applied to simple trusses with any number
of joints provided that the actual position of one end ofa member in the
deformed truss relative to its other end is known. Consider, for example,
the truss shown in Fig. 1.83 a.

r-- x.
Id . - P,
C c.

CK---~~~~~~I

I, P
o

(a)

(b) WiUiot diagram

Fig. 1.83

In this case m ember ab remains fixed in d irection and joint a i n posi-

tion a fter d eformation . The signs i ndicated on the members represent the
sign of the forces developed in the members due to the applied load as
well as the sign of deformation; positive for tension and elongation and
124

negative for compression and shortening. A sketch of the Williot diagram

is shown in Fig. 1.83 h.

Point a' is first established, and since under the particular case ofload-
ing considered member ah has a zero force it neither shortens nor elonga-
tes. The displacement of b relative to a is therefore zero and thus point
b' in the Williot diagram coinsides with point a' . In order to determine
c'. . consider the changes in length of the members connecting joint c to
the two pre-located joints in trainglc abc; points at and b'. Vector a'el
; . . plotted parallel to ac to represent the elongation of member ac . Point C2
coinsides with point h' as member be does not change in length. This
member, however, does rotate and its rotation is effected in the Williot
diagram by drawing a normal to bc through e'2' is the point ofinter-
section of this normal and the perpendicular to ac through c 1 With c'
now located and with b' as previously located, the next triangle to be dealt
with is bcd. The shortening of member cd as represented by vector c~
is plotted parallel to cd. Also the shortening of member bd as represented
by vector b'd 2 is plotted parallel to bd. The deflected position of joint
d will be point d' where did' and d 2 d' are perpendicular to cd and bd
respectively. The remaining unlocated point in triangle cde, point e', is
located next. From d' vector d~1 representing the shortening of member
ed is plotted parallel to cd . Also, from c' vector c'e2 representing the
elongation of member ec is plotted paranel to ec. e' will bc the point of
intersection of elc' and e2c' where these are perpendicular to cd and ec
respectively. Finally f' is located. Vector e'fl representing the elongation
of member efis set out from e' and parallel to ef. Also vector d'f2 repre-
senting the shortening of member dfis set out from point d' and parallel
to df. f' will then be the point of intersection of flf' and f2f where these
are perpendicular to members ef and dfrespectiveIy.
The deflection at any joint may be determined by measuring, to the
chosen deformation scale, from a' to the point on the Williot diagram
corresponding to this joint. For example, the total deflection at joint e
and its horizontal and vertical components are as indicated in Fig. 1.83b.
As a further illustration, the truss shown in Fig. 1.84a is considered.
The truss is assumed symmetrical and symmetrically loaded. As before,
the signs indicated on the members represent the sign of the forces in the
members and their deformations .
Owing to the complete symmetry of the truss and loading, member
 125

h does not rotate during deformation. Either end of this memb r may
be assumed to be fixed in position, and here joint c has been chos fi.

P, r. P, P,
fk-~~g~~~~~-~~
,

a
b

(a)

f, 1\--
a, -'
,'/
-'.

(b) Williot diagram (e)

Fig. 1.84

The procedure for constructing the Williot diagram remaIns (.sscn-

tially the same but it may serve a purpose ifit is outlined again.
A suitable scale for the deformatirm is chosen and point c' is establi-
shed at a convenient location on t e tracing paper. If c is fixed in position
and ch in directiofi then h can only move vertically upwards. Vector
3

2h' which represents this displacement to the chosen scale is plotted. Point
h' is established by plotting vectors h/b 1 and c'b 2 representing the shor-
tening and elongation of members bh and cb respectively, and finding
the point of intersection of the normals to these vectors through hi and
l
b 2 The same procedure is repeated until point a is located. The Williot
diagram for the left half of the truss is shown in Fig. 1.84 b. The deformed
shape of the whole truss is shown in Fig. 1.84 c. Since the truss is
symmetrical and symmetrically loaded, the deformed shape of the right
half would be in a corresponding position to the right of the centre-line.
In the actual truss, joint a, not joint c, is fixed in position, This creates
no problem since the displacements given by the Williot diagram are
relative. Thus, the absolute true deflections are determined by measuring
 126

from point a' in the Williot diagram to the other points representing the
rest at the joints. The vertical deflection at joint c, for instance.. is the
measured distance Yr;; shown in Fig. 1.84 c. Also, the horizontal displace-
ment at joint e is the measured distance xe where point e' in the Williot
diagram is symmetrically opposite to point a'.

In the Williot diagrams given, the change in length of the members

are represented by heavy lines. This is m eant as a help in following the
construction and the student is advised to do the sa me.

1.33 Properly oriented Williot diagratns

[fthe truss given in the previous example were not symmetrical or not
symmetrically loaded then member eh would not have !remain("d fixed
in direction after deformation and thus t he actual displacement of one end
of a member relative to its other end would have not been known. T\ever-
theless, a Williot diagram could have been drawn assuming a member
to remain fixed in direction and one of the joints at its ends to remain
fixed in position. Although the relative displ acements ofthc various joints
could still be obtained from such a diagram, yet the true displacements ot
the joints will remain unknown.

Now, if by some means the d isplacement of one end of a member

relative to its other e nd is determined a properly oriented Williot diagram,
from which the true displacements at all the joints may bc directly found,
could be drawn.

For example, consider the truss in the previous example, which is repro-
duced in Fig . 1.85 a . In this case,joint a is known to bc fixed in posilion
and computing the absolute displacement at joint b thus provides the
only additional information necessary to determine the actual relative
positions of ends a and b of member abo The absolute displacement at
joint b is the resultant of the displacements along member ab and per-
pendicular to it. The first displaccmrnt component is simply the change
in length of member ab (already calculated) while the latter may be
calculated by the method of virtual work. In so doing, only a set of forces
FI due to a unit vertical load at b need to be ealcuated; the Fo-system
of forces having been already computed during the usual course of the
analY!:iis.
"- 127

P, P,

[ (a)
h'
(b) Winie! diagram
c
Fig. 1.85

Referring to Fig. 1.85 h .. point a' in the Williot diagram is first establi-
shed then. point h' is located to its right by an amount equal to the elonga-
tion of member ab, d8 ab and downward by the calculated vertical deOec-
tion at joint b'Yb ' Points a' and b' thus located, point ' is determined next
and the Williot diagram is completed in the usual manner. It is noticed
that the joint displacements thus determined are identical to those
obtained in the previous example.
It should be emphasized that no restriction is imposed on this method
in that it can be appl ied to tfusses where no symmetry in shape or load
exists.

1.34 IlIustrative exalllpies

As an application to the principles presented in the preceding two
sections, a number of illustrative examples is given. The student is recom-
mended to solve these examples independently and then check his results
against those given. Two points, which have proved to be a source of con-
fusion to many students, are considered first.
a) Choice of scale
Any scale suitable to the space available may be chosen. The important
point is that if k dS instead of dS values are used in plotting the Williot
diagram, the scaled displacements should be divided by the factor k.
(b) Location of the assumed fixed point on the tracing paper
In order to avoid the risk of a part of the diagram falling outside the
128

bounds of the tracing paper) it is advisable to first make a sketch of the

Williot diagram in order to form a general idea about its shape and in
which direction it will tend to spread.

ExaJnple 1.61 Draw the Williot diagram for the truss shown in Fig. 1.86 a
and state the actual deflection at joint d. The relative areas orthe various
members are as i ndieated. E = 2000 tlem 2, A = 25 em 2.
1

2A 151 151

Gm A
A A A 3

e
A J
L-_-l<c'
_l-l_ 4 m J
(al
(blWilliol diagram
Fig. 1.86

Solution: The forces in the truss members arc found and the changes in
length dS orthe various members are worked out in JTable 1.13.

L A F
Member E dS
em. cm. 2 t.

ab 500 50 +37.5 +375

be 400 25 0 0
ed 300 25 - 15 - 180
de 400 25 -20 -320
cf 400 25 -20 -320
af 600 25 + 7.5 + 180
bf 500 25 -12 .5 - 250
bd 500 50 + 25 + 250
be 300 25 0 0
Table l./3

Point a is fixed in position and member af is fixed in direction. The

4:0rrcsponrling Williot diagram drawn with member deformations =
4dS-values is shown in Fig. I.86h. From this diagram the deflection
components of joint d are found by measurement to be :
 129

Yd 3.8 x 1/4 0.95 cm. (downward)

Xd - 1.3 x 1/4 0.32 cm. (to the left)

Example 1.62 Draw the Williot diagram for the truss shown in Fig.
1.87a if under the given loading struts are stressed to 0.8 t/cm 2 and ties
to 1.2 tlcm 2. State the vertical deflection at joint c and the horizontal
displacement at joint e. E = 2000 t/cm.2

t.
I,

"'1 9
p
h
p p
j

c
p
(al

'"
Fig. 1.87 f
"

Solution: By inspection the type of force in each member is found. The

dBvalues are calculated in Table 1.14. For symmetry, only one half of
the truss is considered.

Member ab bc fg gh af bg ch fb bh
- - - -- - - - ----
L cm. 400 400 400 400 300 300 300 500 500
F/A t/cm 2 0 +1.2 -0.8 -0.8 -0.8 -0.8 +1.2 +1.2 -0.8
,- dB cm. 0 +0.24-0.16 -0.16 -0.12\-0.12 +0.18 +0.3 -0.2

Table 1.14

Member ch remains fixed in direction and point c is assumed fixed in

position. The corresponding Williot diagram drawn with member
deformations = 2d8~values is shown in Fig. 1.87 h. From this diagram,
Yc = 4.4/2 = 2.2 cm. (downward)

\
 130

Point e'
in the Williot diagram is located symmetrical1y opposite to a',
and hence by measurement"
Xc ~ 1/2 ~ 0.5 cm (to the right)

Exatnple 1.63 Draw the Williot diagram for the symmetrical truss shown
in Fig. 1.88 a due to a unit horizontal load at point 12 as indicated.
Dimension and state the values of the vertical deflection atjoinl 6 and the
horizontal displacement at the roller support 12. Assume L/A = 5 cm- 1
for all mem bers and E ~ 2000 tlem 2

(a)
-...
if
y.

100 db-Values ,.

-x12/2_+-_____ '1212
(b) Willio! diagram

Fig. 1.88

Solution: The forces in the truss members due to the applied load are
calculated and listed in Table 1.15. Since L/A and E are constant
L
for all the members then dS is proportional to F and the -vlaue
EA
 131

may be included in the scale used in drawing the displacement diagram.

For symmetry, only half of the truss members are considered.

ImemberI12113/23/24134/45 ~13615615~~/
1 Ft . +1.58-1.58-1.50+1.80-1.80 +2 ro- -3 01+210
Table l.l5
Member 6-7 remains fixed in direction and point 6 is assumed fixed
in position. The corresponding Williot diagram, drawn with member
EA
deformations = - d8-values, will be as shown in Fig. 1.8Sb. The
4L
displacement scale is thus made up as follows:
4L 4x5
I em. (measured) = - = - - = 0.01 em.
EA 2000
Point 12' in the Williot diagram will be symmetrically opposite to
point 1'. The vertical deflection at the various joints are to be measured
from the line joining points I' and 12', while the horizontal displacements
are to be measured from point I' which represents the hinged support in
the truss. Thus,
Y6 2.9 X 0.01 = 0.029 em. (upward)
Xu = 2 (7.5 X 0.01) = 0.15 em. (to the left)

Esample 1.64 Draw a properly oriented Williot diagram for the truss
shown in Fig. 1.89 a. Indicate the vertical deflections at joints 6 and 8.
Assume L/A = 4 cm- 1 and E = 2000 t/cm 2 .

~ "'6-""_ 8
8
8

,I
Y7 r-
y.
lot 3
Io
8
4
'fa 4.51.1

~o
L I 4x4=16 m
(a)

(b)WilIiot diagram
Fig. 1.89
132

Solution: In this case joint a is truely fixed in position, and since member
a7 has a zero force, computing the vertical deflection at joint 7'Y7' provides
the remaining information necessary to locate the true position of joint
7 relative to joint a. This is done by the method of virtual work. The com-
putations are presented in Table 1.16. The PO l-forces are those due to
a unit downward vertical load at joint 7.

Member Fo 10 3 d8 FI FIFo

al -15 -30 -0. 75 +1 1.25

12 -15.8 -31.6 -1.07 + 16.95
23 -13.3 -266 -0.48 + 6.38
34 -13.3 -26 .6 - 0.48 + 6.38
45 -12.9 -25.8 -0.36 , 4.66
T
56 +11.4 +22.8 0 0
a7 0 0 0 0
78 +14.7 +29.4 +1 + 14.70
89 +12 +24 +0.33 + 4.00
9b - 10.7 -21.4 0 0
b6 -10.7 -21.4 0 0
17 +15.8 +31.6 +1.07 +16.95
28 - 2.8 - 5.6 -0.70 + 1.96
48 + 0 .6 + 1.2 +0. 14 + 0.08
59 +24.4 +48.8 +0.36 + 8.80
27 - 5.5 -11 +0.63 - 3.47
38 + 1.3 + 2.6 + 0.33 + 0.43
49 - 8.5 - 17 -0.13 + 1.10
5b - 25 -50 -0.25 + 6.25
96.42
Table 1.16

L 4
Y, = EA L FIFo = 2000 X 96.42 = 0.193 cm. (downward)

Having thus determined the vertical deflection at joint 7, a properly

oriented Williot diagram may now be plotted and will be as shown in Fig.
1.89b. This diagram is plotted with member deformations = 20d8-values.
Y6 3.6 X J/20 0.18 cm. (downward)
ys = 4.6 X 1/20 0.23 cm. (downward)

This method has the advantage of providing a check on the solution.

In this case for example point b ' must be on a horizontal line through
point a' if the solution is correct.
Also" as will be noticed later" this method leads to a smaller and more
compact displacement diagrams. This permits the use of a larger scale
 133

for drawing the displacement diagram in the available space. This in

turn permits a higher degree of accuracy.

Example 1.65 If each of the top chord members of the roof truss shown
in Fig. 1.90 a is fabricated Icm. longer and the bottom chord members
1.5 em. shorter than their corresponding geometrical lengths: plot the
shape of the bottom chord in the absence of load.

C'~l a~d~h
1 1 I 1-- - - 4 x 3 =12 m - - - - - ' 1
~ (al
Y,=Y

5"-"""~ JJ
83 9.3cm 8.3
(c)
a'
(b) Willie! diagram

Fig. 1.90

Solution: Due to complete symmetry) member gd will remain fixed in

direction. The Williot diagram for one half of the truss is shown in Fig.
1.90 h. This diagram is drawn with member deformations = 0.5 d8-
values. The vertical deflections at the bottom chord joints are found by
measurement from the Williot diagram.
Yc = y, = 4.15 X 1/0.5 = 8.3 cm. (upward)
Yd = 4.65 X 1/0.5 = 9.3 cm. (upward)

The deflected shape of the bottom chord is thus as shown in Fig. 1.90c.

1.35 Williot-Mohr diagram

If the displacement of one end of a member in a truss relative to its
other end is known (due to either symmetry of loading and truss or parti-
cular disposition of members and support conditions), then it will be
possible to draw a Williot diagram from which the true deflections at the
various joints may be determined.
134

This situation, although commonly encountered in practice, does not

constitute the general case where no member exists for which the true
displacement of one of its ends relative to the other is known. In
such a case, the difficulty may be overcome by determining the relative
displacement of the ends of a member analytically and then by drawing
a properly oriented Williot diagram as explained in section 1.33.
Another purely graphical ni'ethod is also possible. This method is, in
effect, a correction to improperly oriented Williot diagrams drawn on
the assumption that some arbitrarily selected member remail!.s fixed in
direction. The principles involved may be best explained in regard to a
specific example.
Consider a truss such as t hat shown in Fig. 1.91a. The signs indicated
on the members represent the type of deformation. The loads are assumed
to be unsymmetrical" and due to this lack of symmetry" no member is
known to remain fixed in direction after deformation. Nevertheless, a
Williot diagram can still be drawn on the assumption that some point
remains fixed in position and a member in direction. Here, point a is
fixed in position and member ab is assumed to remain fixed in direction.
The corresponding Williot diagram will be as shown in Fig. 1.91b, and
the deformed truss will assume the form shown dashed in Fig. 1.91c. The
scale of the deformations used in drawing t he deformed shape in Fig. 1.9lc
is 1/10 that used in drawing the Williot diagram in Fig. 1.91b. In this
figure it is seen that the roller support e is raised above its original position
by an amount f1. The displacement at e, however, cannot be but along
line abo It is necessary" therefore, to rotate the deformed truss as a whole
about the fixed point a until point e' is brought back to its original level
on line ae. The angle of rotation is thus given by :
!;.
d8= -
L
To explain the nature of the rotational displacements of the various
joints, joint g is considered with reference to Fig. 1.91d. Let dx and dy
be the horizontal and vertical displacement components of joint g due
to rotation only then from geometry,
dx R d8 sin8 = y d8
dy R d8 cos8 = x d8
~ence, 8 = R d8 1.60
 135

e" e'
/I
/1
/ 1
h"f-- -~d"
I'
1 " 1
1 ' I
g"t- - - ~d" c,.
1
1/1
I /
/1
I
.'
t---
,
--l b"
I
" I
, I
i \L
. ..b'
a" a,a
(el Mohr diagram (f) Williot-Mohr diagram

Fig. 1.91
136

dy x
and 1.61
dx y
Equations 1.60 and 1.61 and other similar sets of geometrical
relationships that could be found for other joints indicate that the
rotational displacements of the various joints are proportional to their
corresponding radial distances from the centre of rotation a and in a
direction perpendicular to them.
This has suggested a purely graphical method for the determination
of the displacements due to rotation.
In Fig. 1.91e, the truss is drawn with a base line a" e" eq ual to D. (in
Fig. 1.91 b) and geometrically similar to the actual truss and also with
all its members making an angle 90 with corresponding members in the
actual truss. It can be proved that the displacements of the various joints
due to rotation are equal to the distances between similarly identified
points to the centre of rotation a labelled a" in this diagram .
It will be required to prove that g"a" is perpendicular to line ag and
that g"a" is equal to ag times the angle of rotation dEL
Since triangle a"g"c" (in Fig. 1.91e) is similar to triangle agc
(in Fig. 1.91a), angle g"a"c" = angle gae, and since a"e" is perpendic-
ular to ac, g"a" is perpendicular to ago
Also, from similarity of the two figures in Figs. 1.91a and 1.91e,
a" gil ag
--- = -
a" e" ae

or a" g" = ( a"

-;;-
") ag

a" e" 6-
but - - = -=d8
ae L
then a" g" = ag de
and thus the required is proved .
The diagram shown In Fig. 1.91e is called Mo hr diagram after the
German engineer Otto Mohr who first suggested it.
If the M ohr diagram in Fig. 1.91e is superimposed on the Williot
diagram shown in Fig. 1.91 b, the diagram shown in Fig. 1.91f is obtained.
This diagram is known as the Williot-Mohr diagram. The total displace-
ment at any joint is the vector sum oftne displacements due to rotation and
deformation. The former is measured from the double-dashed point in
1-

137

the Mohr diagram to the fixed point a" (also labelled aiL while the
latter is measured from the fixed point to the single-dashed point in the
Williot diagram. In other words, the magnitude and direction of the
total displacement at a joint may be found directly by measurement from
the double-dashed points to the similarly identified single-dashed points_
This is indicated in Fig_ 1.91f for joint d.

If a joint other than the hinge a had been chosen as the fixed point,
a Williat diagram could still be drawn. The deformed truss is then trans-
lated as a whole so that point a' in the deformed truss coinsides with its
original position a. This corresponds to measuring the deformation dis-
placement of any joint from point a' to the single-dashed point in the Wil-
liat diagram similarly labelled as the joint in question. The rest of the
procedure is the same as explained before.

As an illustration, the same truss in Fig. 1.91a is considered. The

Williot-Mohr diagram assuming joint c as fixed in position and member
ch in direction is presented in Fig. 1.92b. The student is advised to compare
the Williot-Mohr diagram in Fig. 1.91f with that in Fig. 1.92 b. The
scale used in both diagrams is the same. It will be noticed that the displa-
cements measured from the double-dashed point to the similar1y labelled
single-dashed point in both diagrams arc identical. It will also be noticed
that the Mohr diagram in Fig. 1.92 b is smaller than that in Fig. 1.91 f.

f _ g h

P, P,
(a)

(b) WilUot- Mohr diagram

Fig. 1.92

This is because member ch assumed fixed in Fig. 1.92 rotates less than
member ab assumed fixed in Fig . 1.91. In the extreme case, when the
 138

member assumed fixed remains actually fixed in direction after deforma

tion, Mohr diagram vanishes. Further, it is noticed that in Fig. 1.91 f
the Mohr diagram lies on the left side of the base line aile" while it lies
on the right side of the base line a"e" in Fig. 1.92 h. This may lead to
some confusion and students often inquire as to which side of the base line
the Mohr diagram should be drawn. The answer is that it should always he
drawn such that all its members are perpendicular to the corresponding
members in the original truss, or more simply, it is drawn on the side
which allows it to take a position parallel and similar to the original truss
when rotated through an angle of 90.

1.36 Applications to Williot-Mohr diagram

In the following, a number of examples is given to illustrate the con
struction of Williot-Mohr diagrams. The student is advised to solve these
problems independently and check his results against those given. Before
doing so" however" he should remember that accuracy in tracing and
scaling is essential. Also, good drafting techniques are helpful. In this
respect" it is recommended to :
(I) use heavy lines to represent the scaled values of dS, light solid
lines the perpendiculars and light dashed lines the Mohr diagram,
(2) use single-dashed and double-dashed letters in the Williot and
Mohr diagrams respectively to refer to the joints in the tqlSS similarly
labelled.
This technique is followed throughout this "book.
On drawing a Williot-Mohr diagram, the student should also notice
the following remarks.
(1) On assuming a member fixed in direction, it is preferable to
choose a member which is likely to have the least rotation during deforma-
tion. This is usally done hy inspection and has the advantage of reducing
the size of the Mohr diagram and thus the final Williot-Mohr diagram
occupies a relatively smaller space.
(2) The base line on which the Mohr diagram is drawn is defined hy
two points. The first point (designated hy a single dashed and douhle
dashed letter) represents the joint at the hinged support. The second
point is related to the roller support, and its location is determined
by the intersection of two lines; the first being drawn through the
point representing the hinged support and perpendicular to the line join-
ing the two supports in the original truss, and the second line through the
 139

point representing the roUer support and parallel to its path of movement.
(3) After determining the base line, the-Mohr diagram is obtained
by drawing a structure similar to the given truss. The ratio of similarity
is determined by the length of the base line and the corresponding length
in the actual truss. Further, the position of the structure in the Mohr
diagram should be such that it may move to a position parallel and similar
to the original truss when rotated through an angle of 90.

ExaD1ple 1.66 Sketch a free-hand Williot-Mohr diagram for the truss

shown in Fig. 1.93 a. Assume appropriate d8-values consistent with
the signs of deformations indicated on the members. Indicate the
displacement at joint b on the final displacement diagram.

x-
II
/ I
d
/ I
d~ __

\
b
~1
Cal 2
a fixed point
ad fixed direction

CblWiHiot-Mohr diagram

Fig. 1.93

Solution: The Williot-Mohr diagram is shown in Fig. 1.93 b. Point a IS

fixed in position and member ad is assumed fixed in direction.

ElI:aD1ple 1.67 Sketch a free-hand Williot-Mohr diagram for the truss

shown in Fig. 1.94 a. Assume appropriate d3-values consistent with the
signs of deformations indicated on the members. Indicate the horizontal
and vertical displacements at joint e on the final displacement diagram.
140

f' f"

+ e
(a)
a fixed point
ac fix'e d direction

b'
(b) Williot-Mohr diagraM Fig. 1.94

Solution: The Williot-Mohr diagram is shown in Fig. 1.94 h . Point

a is fix.ed i.n p::>sition and member ac is assumed fixed in direction.
Example 1.68 Sketch a free-hand Williot-Mohr diagram for the truss
shown in Fig. 1.96a. Assume appropriate dS-values consistent w:ith
the signs of deformations indicated on the members. Indicate the
displacement at the roller support on the final displacement diagram.

//'\,
b.

/ I
f'
/ 11\

~:01:-_-:-_"ie,-;~
d
CJl
1.5m

0.5

3 --o-l2
/ I
/ (a)
c'(---~ ,
I / \ a fixed point
I /
/ ab fixed direction
151( I
" I,
I
(b) WiUiot - Mohr diagram
d'

Fig. 1.95
 141

Solution: The Williot-Mohr diagram is shown in Fig. 1.95 b. Point a

is fixed in position and member ab is assumed fixed in direction.

Example 1.69 Sketch a free-hand Williot-Mohr diagram for the truss

shown in Fig. 1.96 a. Assume appropriate dS-values consistent with the
signs of deformations indicated on the members. Indicate the hori-
zontal displacement at the roller support f on the final displacement

I diagram.

4m

4
b b c

~ 3x3

0'0-
=9 m
(a)
3J
j'

I Of
f'

: fiXl'd point
C9 fiXl'd dirl'ction

(b)Wllliot- Mohr diagram

e'

Fig. 1.96

Solution: The Williot-Mohr diagram is shown in Fig. 1.96 b. Point c

is anumed fixed in potion and member og is assumed fixed in direction.
 142

The student should pay particular attention to how the base line is
determined in each of the exaroples presented above. He is also advised
to re-sketch the given displacement diagrams independently making the
same assumptions given for the chosen fixed points and fixed directions.
He will also do good if he sketches the displacement diagrams making his
own assumptions for the fixed points and fixed directions.

Exan.ple 1.70 For the truss shown in Fig. 1.97a draw the Williot-Mohr
diagram, and then sketch the deflected shape of the bottom chord indi-
cating the values of the vertical deflections at the panel points on
the sketch. E = 2000 t/cm' and L/A = 4 cm-' for all the members.

B
Ii
8 8
8
r- 41 '" 4
4.5.LJ
~~a~~7~--~8~--~~~--~
-- 4.1\04=16 m

r---=:::::::J::;:_L.-I
. (al

~
0.19 0.23 0.1gem

(e)

20 do-Values

(b)WiHiot-Mohr diagram

Fig. 1.97

Solution ; The forces in the members and the corresponding da-values

have already been calculated in example 1.64. Using these values and
 143

assuming point a fixed in position and member al in direction, the Williot-

Mohr diagram will be as shown in Fig. 1.97 b. The displacement diagram
is drawn, as in example 1.64, with member deformations = 20 d3-valucs.
From this diagram, the vertical deflections at the joints in the bottom
chord are found by measurement and are as indicated in Fig. 1.97 c.

Exalllple 1.71 For the truss shownin Fig. 1.98 a, draw the Williot-Mohr
diagram indicating the value of the horizontal displacement at the roller
support b and check the value thus obtained analytically. E = 2000 tfem'
and LfA = 2.5 cm- 1 for all the members.

21 2 2 2 2 2
1 2 41
1
4
L
~ I
a
4x4:16m
(a)

5'
10'

100 do-Values
(b) Williol- Mohr diagram 3'

Fig. 1.98
144

Solution: The forces in the members due to the applied loads and the
corresponding dS-values are presented in Table 1.17. These are all what
is necessary to construct the Williot~Mohr diagram shown in Fig. 1.98 h.
This diagram is drawn with member deformations = 100 dS-values and
assuming joint 8 fixed in position and member 83 fixed in direction.
-
Member
F 10 3 dS F, 10 3F,dS
t. em. t.
1-2 0 0 0 0
2-3 -2 - 2.5 - 0.375 + 0.94
3-4 -2 - 2.5 -0.25 + 0.625
4-5 0 0 -0.125 0
5-6 +4 +5 0 0
7-a -1.5 - 1.87 0 0
a -8 +3 +3.75 +1.12 + 4.2
8-9 +6 +7.5 +1.375 +10.3
9-10 +6 +7.5 +1.25 + 9.4
IO-b +4 +5 + 1.125 + 5.62
1-7 -2 -2.5 0 0
2-a -6.7 -8.35 - 0.375 + 3.13
3-8 -2 -2.5 +0.125 - 0.31
4-9 0 0 +0.125 0
5-10 +2 +2 .5 +0.125 + 0.31
6-b -2 -2.5 0 0
2-7 +1.9 +2.37 0 0
2-8 +4.7 +5.85 +0.53 + 0.31
3-9 0 0 -0.177 0
4-10 -2.8 -3.5 -0.177 + 0.62
5-b -5.7 -7. 12 -0.177 + 1.26
39.195
Table 1.17

By measurement from the displacement diagram in Fig. 1.98 b,

3.9 .
xb = - = 0.039 em. (to the nght)
100

The F,-forces presented in column (4) of Table 1.17 are the member
forces due to a unit horizontalleft-to-right load. The last column repre-
sents 10 3 F ,dS-values, which when summed and :divided by 10' will give
the horizontal displacement required. Thus,
39.1 95
Xb = -- = 0.039 em. (to the right)
10 3

which is identical to the value obtained graphically.

 145
Exunple 1.72 For the truss shown in Fig. 1.99a, construct the Williot-
Mohr diagram assuming joint h fixed in position and member hd fixed
in direction. The dS-values, in centimeters, are as indicated on the
members. Indicate the horizontal and vertical displacements at roller f.

-0.2 d
f '0
,c;:,."'<..~
'0 .<$
3m ~ or 0 .c;:,'), N
Cl

t-
3 \!?
b
.015

,~~
.02 h

9 .0.05

~ a
4.4 :16m
(al

r X,

4 db-Values

(blWiliiot-Mohr diagram

Kid'
Fig. 1.99
Solution: The required Williot-Mohr diagram drawn to scale 1 em. =
0.25 em. is shown in Fig. 1.99 b. The absolute displacement at roller f is
represented by vector f"f'. Its horizontal and vertical components are:
Xc = 3.1 X 0.25 = 0.78 em. (to the right)
Yc = 1.55 X 0.25 = 0.39 em. (upward)
146

Part 4
Influence Lines of Displacements

1.37 Maxwelrs theorem. of reciprocal deflections

This theorem which is due to the English scientist Clerk Maxwell is
extremely important for not only docs it serve to reduce the amount of
work involved in the analysis of statically indeterminate structures but
also it is the base of constructing the influence lines of deflections and
hence those of various load functions for simple statically indeterminate
structures.
Although a general development of the theorem may be made, it is
thought advisable at this stage and for the particular purpose of construct-
ing the influence lines of displacements which win be dealt with in the
following section to restrict the development to a special case of loading,
Consider a simple beam as shown in Fig. 1.100. Let Smn be the de-
flection at point m due to a uni t load acting at n, and Snm be the deflec-
tion at point n due to a unit load acting at m as indicated.

Fig. 1.100

These deflections may best be found by the method of virtual work

and are as follows :

where Mo and Mm are the bending moments due to a unit load acting at
points nand m respectively. Thus,
8mn = 8um 1.62
 147

Equation 1.62 is the mathematical expression of the well known Max-

well's theorem of reciprocal deflections. It means that the deflection at
any point m due to a unit load at some other point n is equal to the deflec-
tion at point n due to a unit load at point m, where the deflection and
load at a given point have the same direction and sense.

The example used for developing the theorem has been a simple beam
but it is applicable to frames, trusses or any linearly elastic structure. Also,
the theorem need not be restricted to linear displacements and may be
extended to include rotational displacements. This will be considered
later on in chapter 2.

1.38 Influence lines of deflections

In chapter 6 of Part 1, the construction of influence lines has been
confined to load functions in statically determinate structures such as
reactions, N.F..J S.F. and B .M. at a section ofa beam or a frame or a force
in a member of a truss. Influence lines can also be drawn for deflection
at a point on a structure. These influence lines form the basis for drawing
the influence lines for various load functions in simple statically indeter-
minate structures. In this section, however, statically determinate struC-
tures will be dealt with, leaving the consideration of statical1y indetermin-
ate structures to a later stage.

For purpose of review, if the curve shown in Fig. 1.101 is the influence
line for the deflection at point c, then ordinate Yl is the deflection at point
c when a unit load is placed at 1 and Y2 is the deflection at c due to unit
load at2, etc. It follows that if a unit load is placed at a sufficient number of
positions along the beam and in each case the deflection at c is calculated
then these values form the ordinates to the influence line for the
deflection at point c.

a r I r ~
b

f~
c 2

Y,!
I.L Y,
3L='".
Fig. l.lOI
148

This procedure, although quite permissible, is obviously v~ry tedious.

Another simple and quick method of drawing the influence lines for de-
flecti~ns is offered by Maxwell's theorem of reciprocal 'deflections.

Consider for example the cantilever shown in Fig. 1.102a.. and suppose
that the influence line for the deflection at the free end b is required.

(a) ;0~ __________ ~c____~l~

I

(b)

I.L. y.

Fig. 1.102

Fig. 1.102 b shows the elastic curve of the cantilever due to the case of
loading indicated. Now if a unit load is placed on the cantilever at an
arbitrarily chosen point such as c, it is known from the reciprocal theorem
that the vertical deflection at b when the unit load is at c, Ybc.J is the same
as the vertical deflection at c when the unit load is at b, Yeb' By definition,
however, Ybc represents an ordinate to the influence line for Yb" Since
point c on the cantilever has been arbitrarily chosen, the influence
line for the deflection at point b is the same as the elastic curve of the can~
tilever due to placing a unit load at b.
The elastic curve is usually determined graphically but may also be
obtained by any of the analytical methods for beam deflections presented
in part I of this chapter.
As a second example, consider the truss shown in Fig. 1.103 a where
the load iJ; assumed to move on the top chord, and let the influence line
for the deflection at joint c be required. The deflected shape of the loaded
chord of the truss due to a unit load placed at c is shown in Fig. 1.103 h.
As in the previous example, when the unit load is at a joint such as d the
deflection at c,YcdJ will be the same as the deflection at d when the unit
load is at c,ydc. Thus, the deflected shape of the loaded chord of the truss,
shown in Fig. 1.103 b, represents the influence line for the deflection at
joint c. Such deflected shape is best obtained graphically by the use of
displacement diagrams.
~
I 149
I

(a)

(b)

Fig. 1.103

From the previous two examples, the following rule may be deduced:

"The irifluence line of d4fection at any point on a structure is found by placing

a unit load at this point and determining the corresponding d4fected shape of the loatkd
part of the structure".

1.39 Wustranve exalIlples

The application of Maxwell's theorem of reciprocal deflections and the
rule stated at the end of the preceding section for determining the influence
lines for deflections will be demonstrated by a number of examples.
Various types of structures will be considered.

Example 1.73 If load is transmitted to the cantilever shown in Fig.

1.104 a at panel points spaced at 1.5 m., construct the influence line for
the deflection at the free end b . EI = 10000 m 2t.

(a)

~
ma 6m

(t l.L.o o \
0.062 I
0.225
0.456
0.12 em
Fig. 1.104

Solution: "The required influence line is the elastic curve due to a unit
load applied at b. The equation of this elastic curve has already been
found in example 1.2 and is re-produced below.
150

Y = (3 Lx2 _ x 3)
6 EI
x 100
at x - 1.5 m., y = (3 x 6 X 1.5 2 _ 1.5 3 ) = 0.C62em.
6 x 10000
1 X 100
at x 3 m., y (3 X 6 X 32 - 3 3 ) = 0.225 em.
6 X 10000
1 X 100
at x = 4.5 m., y (3 x 6 x 4.5 2 - 4.5 3) = 0:456 em.
6 X 10000
1 X 100
at x = 6 ro., Y= (3 X 6 X 6 2 - 6 3) = 0.72 em.
6 X 10000
These ordinates are indicated on the influence line for the deflection
at b shown in Fig. 1.104 b.

ExalDple 1.74 If load is transmitted to the simple beam shown in

Fig. 1.105 a at the quarter points, construct the influence line for the
deflection at the mid point c. EI 6)00 m 2t.

lal
~

(b)l.L.oc
~ 0.183
;;;5
0.267
0.183 em

Fig. 1.105

Solution: The required influence line is the elastic curve of the beam due
to a unit load applied at point c. This elastic curve may be found graphi-
cally using the elastic-load method as explained in section 1.8. Alterna-
tively, the equation of the elastic curve may he derived using the double
integration method. Since the beam is symmetrically loaded, the equation
of only one half of the beam needs be derived. This has been done for the
left half of the beam and the equation is as follows:
1
y = -- (3L 2x - 4x 3 )
12EI
1 X 100
at x 2 m., y (3x 8 2 X 2-4 X 2 3) = 0.183cm.
12 X 16:J00
1 X 100 2
at x 4 m., Y 12 X 16000 (3 X 8 X 4-4 X 4 3) = 0.267cm.
 151

ce
The ordinates at all the panel points are indicat ed on the influen
line for the deflect ion at point c shown in Fig. 1.105 b.
at point c
EsaIDp le 1.75 Construct the influen ce line for the deflection
of inertia
on the non-prismatic beam shown in Fig. 1.106 a. The momen t
2
varies as indicat ed and E = 210 tfcm .

(a) c

q~ ~
~i
a~ ~ III
a
I~
Odd ~dd
a 0
_ II

E
1 75
.
III
15m

r-- ....
75,

en a a a to
N
v
C! 0 0 0 a ...,
N

(b)I.l. 9, 0 a .-----r I :=r----.. 0 a

[5Y' =c:tJ
Fig. 1.106
to a unit
Solutio n: The require d influen ce line is the elastic curve due
graphic ally in
load applied at point c. This has already been found
uced
examp le 1.29 (Fig. 1.37) using the elastic-load method and is re-prod
influen ce
on a horizon tal base line in Fig. 1.IC6 h. The ordina tes to the
line, in centim eters, are indicat ed at points spaced at 3.75 m.
ement
ExaDlple 1.76 Constr uct the influen ce line for the horizon tal displac
16000 m 2 t.
at the roller suppor t b of the frame shown in Fig. 1.107 a. EI=
4 4mt

8m
(a) (b) B.M.D.

lY ~0.2
0.2cm
(c) I.L. Oh

Fig. 1.107
152

Solution: The required influence line is the elastic curve of the loaded part
of the frame due to a unit horizontal load applied at b. The corresponding
B.M.D. is shown in Fig. 1.107 b. Since part cd is subjected to a uniform
elastic load of intensity M/EI = 4/16000, its elastic curve will assume a

parabola with maximum ordinate at the middle = ( __4_ X ~ ) 100

16000 8
= 0.2 em . Further, since parts ec and df are free of moment, they will
remain straight. For continuity of the elastic curve, these straight parts
should touch the parabola at c and d. From geometry, the ordinates of
2
the influence line at e and f are given by - (2 X 0.2) = 0.2 cm.
4
The influence line for the horizontal displacement at support b is thus
as shown in Fig. 1.107 c. The negative sign indicates that as the
load moves between c and d, b will move in a direction opposite to that
assumed for the unit load, i.e. outwards.

ExaD1ple I.n If load is transmitted to the arch shown in !Fig. 1.I08a

at points a-b, construct the influence line for the horizontal displacement
at support b. EI = 40000 m 2t.

Solution: The ordinates of the required influence line are identical to the
vertical deflections at points c, d, c, f and g due to a uni$ horizontal
load applied at b. These deflections can best be found gral'hically by the
method of clastic-load described in section 1.8. The B.M.D. due to a
unit horizontal load at b and the corresponding elastic loads are shown
in Fig. 1.108 b . It should be noted that the elastic loads act on the axis of
the arch at the centroids of the segments of the B.M.D. The polar diagram
with a polar distance equal to EI = 40000 m 2 t is shown in Fig. 1. 108 c.
The corresponding link polygon is shown in Fig. 1.108 d.

The scales used are :

Linear scale, cm. 400 cm.
Elastic-load scale, cm. 20 m 2 t.
EI scale, cm. 20000 m 2 t.
Thus, the deflection scale is made up as follows:
400 X 20
I em = = 0.4 cm.
20000 I
By measurement from the link polygon.J n Fjg. 1.108 d,
 153

(aJ Afm~ 6.4= 24 m

6mt
5.5 5.5

(b) 6.M.D.

9.4
9.4
20.2

23
O'?-_ _- l

23
(d l link Polygon
20.2

9.4
2
40000m t I

(el (el l.l. Db

Fig. 1.108

Y. = Y. - 1.3 X 0.4 = 0.52 em.

Y. = Yr - 2.2 X 0.4 = 0.88 em.
Y. = 2.4 X 0.4 = 0.96 em.

These ordinates are indicated on the influence line for the horizontal
displacement at support b shown in Fig. 1.108e. The negative sign indica-
tes that for any vertical downward load on the arch, support b will move
outwards.
154

Exam.ple 1.78 Draw the displacement diagram of the truss shown in

Fig. !.l09a due to a unit horizontal load applied at support b. Using this
diagram find then the displacement at b due to :
(a ) horizontal loads H, H = 2t .
(b) vertical loads V, V 8 t.
1
Assume LfA = 10 cm- and E = 2000 t fcm 2 .

H P

Ha n m k

al~----~ ____~~__~~__~~____-J
C d e

v v v

(al f-..- - - - 5x4:20m

(bl Williol- Mohr diagram

50 d6-Values

Fig. !.l09

Solution : The member forces due to a unit load acting at b as indicated

are listed in Table 1.18. Since LjEA is constant for all the truss members,
the member deformations, dB, are proportional to the member forces.
The constant of proportionality may be included in the scale used for
drawing the displacement diagram.
 155

Member Force Member Force Member Force Member Force

ac -1.4 op 0 ih +1.2 dl +0.5

cd -1.4 pn 0 hb +1.2 Ie 0
de -2.2 nm +0.8 on 0 If -0.5
ef -2.2 ml +0.8 an +0.5 fk 0
fg -1.8 Ik +1.6 ne 0 fj -2.0
gb -1.8 ~j +1.6 nd -D.5 Ii 0
ao 0 Jl +1.2 md 0 ig 0
gh 0

Table 1.18

Assuming point d to remain fixed in position and dl fixed in direction,

the resulting Williot-Mohr diagram drawn with 50 da-values will be as
shown in Fig. 1.109 b. By measurement from this diagram, the horizontal
displacements at points p and 0 due to the unit load at b are respectively:

a pb 3.1/50 = 0.C62 em. (to the left)

aOb 1.5/50 = 0.030 cm. (to the left)

Applying Maxwell's theorem of reciprocal deflections, the displacement

at b due to loads of 21. applied at each of points p and 0 is given by:
a b (due to H) = 2 (0.062 +0.030) =0.184 cm. (to the right)

Similarly, the vertical displacements at points c,d,e and f due to the

unit load at b are respectively:
aOb - 2.2/50 = 0.044 em. (upward)
adb 3.6/50 0.072 em. (upward)
a,b = 4.35/50 = 0.087 em. (upward)
aft> = 3.65/50 = 0.073 em. (upward)

Applying Maxwell's theorem of reciprocal deflections, the displacement

at b due to downward loads of 8 t applied at each of points c, d, e, and f
is given by :
a b (due to V)=8 (0.044+0.072+0.087+0.073)=2.208 em. (outward)

ExaInple 1.79 If load moves from g to k on the truss shown in

Fig. 1.110 a, construct the influence line for the vertical deflection at
joint f. L/A = 6 em- 1 and E = 2100 t/em 2.
156

r-gk------7~----~~----_*--~k

(a)
f

12 m - --:o:::t="'-- 9

0.011 em
(e)J.L.6, k===I~~~--.1=-___..,-_~
+
K:-
1/
~f" 0.074 em
fK 1
I " I ..
1 ~~
/ , ,
i"k I
1, 1/ i'
1 '1/.,
I / 1C
nK t I
1' I
>tb"
I
~'l~r--- --h~--.Ja' -
175 d6-Values

j'

f(b)WiUiol-Mohr diagram

"'-_ _ _ __ _....Jk'

Fig. l.1I0
 157

Solution: The member forces due to a unit downward load at f arc listed
in Table 1.19. Since L/EA is constant foraH the truss members, the
member deformations, d3, are proportional to the member forces. The
constant of proportionality, L/EA, may be included in tl)~ scale used for
drawing the displacement diagram.

Member gh hi IJ jk ab bc ce ef ga kf
Force +0.75 +2.25 +2.0 0 0 -1.5 -1.25 - I +'0.75 0

Member gb bh hc ci Ie OJ jf cd de
Force -1 .06 + 1.06- 1.06---{).IS+0. 1B-1.41 + 1.41-1.24-1.24

Table 1.19

Assuming point h to remain fixed in position and member he fixed in

direction, the resulting Williot-Mohr diagram drawn with 175 d8 - values
is shown in Fig. l.1l0 b.

The vertical deflections at the panel points g-k of the loaded chord.
due to a unit load at f form the ordinates of the required influence line.
By measurement,
3rk - 12.9/175 = 0.074 cm. (downward)

3r; - 7.7/175 - 0.044 cm. (downward)

3fh = 2/175 = 0.01l cm. (upward)

These main ordinates are indicated on the influence line for the verti-
cal deflection at f shown in Fig. 1.1l0 c. The negative sign indicates that
as the load moves between g and i the displacement at f is upward.

ExaD1ple 1.80 If load is transmitted to the lower chord of the truss shown
in Fig. 1.111a, construct the influence line for the horizontal displacem-
ent at the roller support b . L/A = 4 cm- 1 and E = 2000 t/cm 2

Solution: The member forces due to a unit horizontal load at b are listed
in Table 1.20. Since L/EA is constant for all the truss members, the d3 -
values are proportional to the member forces. The constant L/EA may
be included in the scale used for drawing the displacement diagram.
L
1
f
,
ISS

J 5 7 9 11 13
.
(a)
1~1
6m

18m 10
y
0013 0.022 0.023 0.014 em
(d I.L.~
______ ..,...-- I - I J=-"

~ 0.036 0.05B

14'
.... 1
\r-.,.p;;~:::; ..
(....
I
I
I .... )/12
I .... "" I

,"---1
K I
I ')I!

" I _--
I .... :>t6'
k I
I 2

iC .... '"
I
....

I .... /
.... I

.....
6

4"
/
175 do-Values K. . ....
I
( b)Williot- Mohr diagram I .... ""

14

Fig. 1.111
 159

Mcmb er 1-3 3-5 5-7 7-9 9-11 !l -13 9-2 2-4 4-6 6-8

Force +0.44 +0.88 + 1.32 + 1.76 0 o ~ 1.22~ 1.67 ~2.1 ~2 .54

Mcmber 8-10 10-12 12-14 a-I 1-2 2-3 3-4 4-5 5-6 6-7

Force - 1.32 o 0 +0.4 -0.4 + 0.4 ~0.4 +0.4 ~0 . 4 +0.4

Memb cr 7-8 8-9 9-10 1O-!l !l-12 12-13 13-14 b-8 bolO

Force ~0.4 ~1.58 +1.58 0 o o o ~1.74 +1.38

Table 1.20

Assu ming point 7 to remain fixed in position and 7-6 fixed in direction,
the res ulting Williot-Mohr diagram, drawn with member deformations
= 175 do - values, is shown in Fig. 1.111 h.
The vertical deflections at panel points a-14 will then form the ordin-
ates of the required influence line. By measurement..
S2 2.35Jl75 = ~ 0.013 cm.
S, 3.8/175 = - 0.022 cm.
S. = 4/175 = ~ 0.023 cm.
Ss = 2.45Jl75 ~ 0.014 cm.

S,o = 2.35/ 175 + 0.013 cm.

S12 = 6.3/175 + 0.036 cm.
S" = .10.1/175 + 0.058 cm.
These ordinates are indicated on the influence line for the horizontal
displacement at the roller support b shown in Fig. 1.111c. The negative
sign indicates left-te-right movement.
I
r

160

EXAMPLES TO BE WORKED OUT

Classical Methods of Bean> Deflections

(I) Using the double integration method, derive the equation of the
elastic curve of the cantilever shown in Fig. 1.112. Determine the slope
and deflection at the free end if. EI = 4000 m 2t.

Fig. l.l12 Fig. lo1I3

(2) Using the double integration method, derive the equation of the
elastic curve of the beam shown in Fig. 1.1I3. Determine the maximum
deflection and compare its value to the deflection at mid span. EI =
5000 m 2 t.

(3), (4) Using the double integration method, derive the equation
of the elastic curve of each of the beams in Figs. lo1I4 and lo1l5.
Determine the slopes at the supports and the maximum deflection
if I = 7500 em' and E = 2000 tfem 2 .

3mt 6mt 3mt Gmt

C-6m-i)G~-6m-i0
Fig. lo1l4 Fig. lo1I5

(5) Using the double integration method, derive the equation of the
elastic curve of the cantilever shown in Fig. 1.116. Determine the slope
and the deflection at the free end.

Sine curve- W

~ 6r1111~b
~ c
L
AI
>---

Fig. 1.116 Fig. lo1l7

r-------------------------------------------------~/~
 161

(6) Using the double integration method .. derive the expressions for
the slope and the deflection along the beam shown in Fig. 1.117, then
find the slopes at the supports and the deflection at the mid point c.

(7) , (8) Using the double integration method, derive the expressions
for the slope and the deflection along each of the beams in Figs. 1.118
and I. II 9, then find the slopes at the supports and the deflection at point c.

Mo
K;)b
"-1.
3
~ 11.3 -# 2L
""3
Fig. 1.118 Fig. 1.119

(9), (10) Using the double integration method, re-solve problems (I)
and (5) if the moment of inertia of half the beam nearer to the fixed
end is doubled.

(1 1)- (16) Using the moment-area method, re-solve problems (I) - (4),
(7) and (8).

(17) Using the moment-area method" find the rotation and deflection
at the free end of the cantilever shown in Fig. 1.120 if EI = 5000 m 2t.
What will be the percentage reduction in these values if:

f a) the moment of inertia of part ab is doubled?

b) the moment of inertia of part be is doubled?

!'ig. 1.120 Fig. 1.121

(18) Using the moment-area method find the deflection at mid span
of the beam shown in Fig. 1.121 ifE = 200 t/cm 2 and 1= 6x lOS cm 4
What will this deflection be if the moment of inertia of the middle half
of the beam is doubled?
(19) Using the moment-area method, calculate the deflection at
the mid point of span ab and the deflection at point C of the beam shown
in Fig. 1.122 if EI = 7500 m"t.
162

1.SP
Ll3

Fig. 1.122 Fig. 1.123

(20) Using the moment-area method, calculate the vertical deflection

at point c and the horizontal displacement at point d of the beam
shown in Fig. 1.123 if EI is uniform.

(21), (22) Using the moment-area method, calculate the absolute

displacement at point c of each of the beams shown in Figs. 1.124 and
1.125 if EI = 8000 m 2 t.

Fig. 1.124 Fig. l.l25

(23)-(26) Using the elastic-load method, re-solve problems (3),(4),(7)

and (18)
(27), (28) Using the elastic -load method, find graphically the slope
diagram and the elastic curve of each of the beams in Figs. 1.126 and
1.127 ifE = 2000t/cm 2 and 1= 12500 em'. Check analytically the slope
at support b and the deflection at points c and d.

4tl 12 t lt1m
21

~
Q Q
n c d c
I--- 4 m --1-. 2 ..j... 3 -< l-3m.....L 3
Fig. 1.126 Fig. l.l27
(29) Using the elastic - load method, calculate the deflection at the
mid point of each span of the continuous beam shown in Fig. 1.128
if Y b = Y c = 0.575 P and EI is uniform.

Q b ted '"'~
~ L--<

Fig. 1.128 Fig. 1.129

 163

(30) Using the clastic -load method, calculate the deflection at the
mid point of each span of the continuous beam shown in Fig. 1.129 if
Mb = Me = -wL2/1O and EI is uniform.

(31) Calculate the deflection at the mid point of the fixed beam shown
in Fig. 1.130 if M. = Mb = - wL'/12. Comp,"e this deflection to its
corresponding value in a simple beam.

2t

~
~a '"
I! I I I!! I j ~
!b~
1.5tJm -.12.5
a b
~L I 8m 2J
Fig. 1.130 Fig. 1.131

(32) Find the position and calculate the value of the maximum deflec-
tion in span ab of the beam shown in Fig. 1.131 if M. = - 13 m.t. and
EI = 10000 m 2 t.

(33)-(36) Obtain graphically the elastic curve of each of the beams

shown in Figs. 1.132 - 1.135 if the moment of inertia varies as indicated.
Assume EI = 10000 m 2t

A 1 121~21J
I--- 6m -f-- 3 -+- 3

Fig. 1.132 Fig. 1.133

t 16t 31
e
1
21 21
I
L 4
L::,.
I 4 In--l--- 4
bO
---+-- t. ...J
Fig. 1.134

1 21 101 21 1 :1
~3 -f-- 3 -l--3m __ 3
I~I
-+- 3 --l- 3 -l
I

<
Fig. 1.135
164

(37)-(40) Each of Figs. 1.136 - 1.139 shows the cross - section of a

long water tank. Considering a strip having a dimension of 1m. along
the tank and neglecting the own weight of the tank, calcualtc the change
lfl width of the tank at the water level.
In all the tanks, the moment of i ncrtia of the horizontal beam (11) is

,
assumed to be double that of the vertical sides (1 2 ) . I, ~ 2 12 ~ 300000
em 4 and E 200 tJcm 2 .

2.5 Water t
3 Water
1 t/m 3
L L 1 t/m 3

f-.-4m ~1 ! - - - 6 m -----J 1 l-
Fig. 1.136 Fig. 1.137

3
Water
1 tlm3 Water
r 3
L 1 tlm 3
~
1.5
4m 6m
1.5
I
Fig. 1.138 Fig. 1.139

(41)-(47) Using the conjugate beam method, re-solve problems (I),

(3), (I7), (19) , (30) and (31) .

(48) Using the conjugate beam method, calculate the deflection at the
mid point of span ab and hinge d of the beam shown in,Fig. 1.140. Find
also the slope at points a, b, c and just to the left and just to the right of
hinge d if EI = 2500 m 2t. Indicate the calculated values on a sketch of
the elastic curve.

211m
a
I-- 6m

Fig. 1.140

(49) Using the conjugate beam method, calculate the deflections and
slopes at a sufficient number of points along the beam shown in Fig. 1.141
to allow a proper sketch of its elastic curve. EI = 2500 m 2t.
 165

..6:1-- 3
Fig. 1.1 41
ons
(50),(5 1) Using the conj ugate beam method , calcula te the deflecti
beams in
and slopes at a sufficient numbe r of points on each of the
curve.
Figs 1.142 and 1.143 to allow a proper sketch of the elastic
EI = 6000 m 21.
curve.
Indicat e th e calcula ted values on a ncat sketch of the elastic

Fig. 1.142

211m
~~~LIIII~~I~ICf5fI!~=!cC~I~iIJCI[III~~,~bLI2III!:':r::!::~[I~I!I!~~=C~c
Fig. 1.143
in
(52) , (53) Calcula te the magnit ude of the load P acting as shown
on.
Figs. 1.144 and 1.145 such t hat point c will have zero deflecti

" 20~~
~ 60cm 1
~ 3m --'=
Fig. 1.144 Fig. 1.145

on at
(54) Using the conjug ate beam method , calcula te the deflecti
d of the
the mid point of span .hc and the deflect ion and slope at point
1.146 if Y = 31. and
statical ly indeter minate beam shown i n Fig. b

EI = 8000 m 2t .
166

(55) Using the conjugate beam method, calculate the deflection at the
mid points of spans ab and be and the deflection and slope at d of the
beam shown in Fig. 1.147 if Mb ~ - 6.33 m.l., Y a ~ 3.12 t and
EI ~10000 m 2 1.

Fig. l.l47

(56) -(61) Calculate the displacement components necessary to make

a proper sketch of the deformed shape of each of the frames in
Figs. 1.148-1.153 ifEI ~ 10000 m 2 t. Indicate the calculated values on a
neat sketch of the deformed shape.

211m 211m
b c 31
I c d
2
4

6m~ l 6m
b

Fig. l.l48 Fig. 1.149

111m 111m
b c
dl b cr 3
4

a d 1m
4m 2--1 L3m-i- 3
---' 0.51
Fig. l.l50 Fig. l.l51
 167

211m 1.2 tIm

0.75
C d e b cf
d 1.5 t

3
a
9m
aJ-1

1---4m

Fig . 1.152 Fig. J.l53

(62) Calculate the shear deflection at the free end of the cantilever
shown in Fig. 1.112 ifit is made of S.I.B. No. 34. G = 800 t/cm 2 .

(63) Calcula te the percentage change in t he deflection at point c on

the beam shown in Fig. l.U 9 if the effect of the shearing force is taken
into account. The bea m has a rectangular section 25 em X 40 em.
G = 0.4 E .

The Method of Virtual Work

(64) Using the method of virtual work, calculate the horizontal and
vertical displacements at joint C of the truss shown in Fig. 1. 154. Assume
L/A = 5 cm- 1 fo r all the m embers and E = 2000 t/cm 2

10-- - - 5x4 =20 m

Fig. J.l54 Fig. J.l55

(65) Using the method of virtual work, calculate the displacement

at the roller support b and the rotation of member cd of the truss shown
in Fig. 1.155. The cross-sectional area of all the horizontal and vertical

I members = 30 em 2.. the cross-sectional area of the remaining members =

20 em 2 and E = 2000 tlcm 2.
 168
,
(66) Using the method of virtual work, calculate the vertical displace-
ment at joint c and the rotation of member cd of the truss shown in Fig.
1.156. Assume LfA = 12 cm- 1 for all the members and E = 2000 tfcm 2 .

(67) Using the method of virtual work, caiculate the vertical displace-
Inents at joints c and d of the truss shown in Fig. 1.157. For all tension
members A = 30 em 2 and for all compression members A = 50 em 2 and
E = 2000 tfem 2.

(68),(69) Using the method of virtual work, calculate the relative dis~
placement between joints c and d and the displacement at the roller sup-
port b of each of the trussesshown in Figs. 1.158 and 1.159. Assume LfA
JO cm-1 for all the members and E = 2000 tfem 2

41 4 4t 4
5 3 4 4

:r Q
c
e

4x4=16m
d
b
1 51 1'5 21
. f
\5
1Ci Q
c

6x2=12m
2

Fig. 1.158 Fig. 1.l59

(70) Using the method of virtual work, calculate the vertical displace-
ment at joint c and the relative di~p~acement between joints d and e of
"the truss shown in Fig. 1.160. Assume L /A = 10 cm- 1 for all the members
and E = 2000 tfcm 2.
 169

10 10 10 10 101
r-- 4x2=8m ----"j ' 41
51
d 1.5
3m -t1.5

t
5
d '
1
6m

J.
4.3=12m

Fig. 1.160 Fig. 1.161

(71) Using the method of virtual work, calculate the horizontal displa-
cement at joint e and the relative displacement between joints c and d
of the truss shown in Fig. 1.161. Assume LjA = 15 cm- 1 for all the
members and E = 2000 tjem 2

(72) Using the method of virtual work, calculate the vertical displace-
ment at joint c and the displacement at the roller support b of the truss
shown in Fig. 1.162 . Assume LjA = 6 cm- 1 for all the members and
E = 2000 tjcm 2

121
~r-~~--~----~--.---~--~
3F "'-- ...J(

3m

T 2x4=8_-+-_ _ _ 4x4 =16m

Fig. 1.162

(73) Using the method of virtual work, calculate the vertical displace-
ments at joints c and d of the truss shown in Fig. 1.163. Assume LjA =
8 em- 1 for all the members and E = 2000 tjcm 2
I
,

,-
I

170

3 6 61

d c f
6m
a t5 ----.l
I-- 2x6 =12 ---.;;:;::;:~_ _ 4x6 =24m
Fig. 1.163

(74) Using the method of virtual work, calculate the vertical displa-
cement at joint c and the displacement at the roller support b of the truss
shown in Fig. 1.164. Assume L/A ~ 8 cm~l for all the members and E
2000 t/em 2 .

3D
2~-,~~--,,--~~~~ ,3m
i
16m 10m
Fig. 1.164

(75) Link ce' of the truss shown in Fig. 1.165 is provided with a turn-
buckle to adjust its length . Calculate the adjustment that should be made
in order to balance out the vertical displacement at joint b due to
the applied load. Assume L/A ~ \0 em~ 1 for all the members and E ~
2000 t/em 2

6m

~k-a'~~c~
3
~a~__~~~~~~~~~~~b
~_ _ 4x3 =12m _ _--1-_ 2x3=6

61
Fig. 1.165
 171

(76) If it is required to produce a 10 em. camber at mid span ofthe

truss shown in Fig. 1.155 by fabricating each of the top chord members
additionally long by a length a and each of the bottom chord members
additionally short by the same amount, what will be the change in length
a ?, and what will be the associated displacement at the roller support b?

(77) Determine the vertical displacement at joint e of the truss shown

in Fig. 1.156 in the absence ofload due to a drop in the temperature of
member bc of 30C. at = I X 10-' tC.

(78) If it is required to balance out the deflection at point c of the

truss shown in Fig. 1.157 due to the applied load by fabricating each of the
top chord members and the two end diagonals proportionalIy short, cal-
culate the necessary change in length of these members.

(79),(80) Calculate the displacement at the roller support b of the

trusses shown in Figs. 1.158 and 1.159 in the absence of load due to a rise
in the temperature of the top chord members = 30C accompanied by a
drop in the temperature of the bottom chord members ==-20C. at =
I X IO-'I"C.

(81) Calculate the horizontal and vertical displacements at joint e

of the truss shown in Fig. 1.161 in the absence ofload due to a from right
to left movement at support a = 2 em. accompanied by a downward
movement = 3 cm. at the same support.

(82) Calculate the horizontal displacement at joint f of the truss shown

in Fig. 1.160 in the absence of load if:
a) members ad, dc, ce, and eb are fabricated 2% shorter than their
geometrical length.
b) support b moves to the right' 0.8 em. and downwards 0.6 em.

(83) Calculate the vertical displacement at joint d of the trussshown

III Fig. 1.163 due to :

a) a downward movement at support a = 2 em. accompanied by

a from left to right movement = 2 cm. at the same support,
b) an inward 'movement at each of supports a and b = 1.5 em,

(84)- (94) Using the method of virtual work, re-solve problems I,

17-22 and 48 - 51.

(9.<;)-(100) Using the method of virtual work, re-solve problems 56-61.

172

(101) Using the method of virtual work, calculate the ,three displace-
ment components at point d of the frame shown in Fig. 1.166 if EI =
10000 m 2 t.

211m
1.51/m
b c

~1
I
4m
.
l{)
c
1
6m
-, a 1 ci

~3r
d a
~ 3......L4m----J 0,51 6 ----I-6m
Fig. 1.166 Fig, 1.167

(102) Using the method of virtual work, calculate the vertical dis-
placement and the relative rotation at the intermediate hinge c of the
three-hinged frame shown in Fig. 1.167 if EI = 25000 m 2 t.

(103) Using the method of virtual work, calculate the horizontal dis-
placement at the roller support b of the frame shown in Fig. 1.168 taking
into account the effect of axial and shear deformations.
What is the percentage error if this effect is neglected? A = 200 cm 2J
A, = 80 em 2, I = 100000 em', E = 2000 tfcm 2 and G = 800 tfem 2.

5 2 tim
e
~
21 , 21
d I ~
2
6m b I
La L-.. 4 --L4 m--L 4
J 5 --+---5m
Fig. 1.168 Fig. 1.169

(104) Using the method of virtual work, calculate the horizontal and
vertical displacements at point e of the frame shown in Fig. 1.169 if I
1750000 em' and E = 200 tfcm 2.

(105)-(108) Using the method of virtual work, re-solve problems 37-40.

 173

(109),(110) Using the method of virtual work, calculate the rotation

at support a and the horizontal displacement at support b of the frames
shown in Figs 1.170 and 1.171 if EI = 8000 m 2t.

,.---l! t
211m
a\i;=Iiifu:::r::o+-~ i ~ 3L? N

3 2.4 ~

~ 6m _--,,~2

Fig. 1.170
b
f Lab
L.....-4m--l--3 .2--1
Fig. 1.171

(111) Using the method of virtual work, calculate the relative displa-
cement between the ends of member ac of the frame shown in Fig. 1.172
if EI = 67500 m 21.
Check your results by calculating the displacements at points a and
c separately.

2t 2t1m 2t1m
I I ! II II !
31
c 41 b c r
6m t- d
e1
3

L a
6m 8m

L8m
L b

I 2-1
L a
~ 6 6m
J
5

Fig. 1.172 Fig. 1.173

(112) Using the method of virtual work, calculatc the vertical displa-
cement at point d and the relative displacement between the ends
of member cd or the frame shown in Fig. 1.173 if lab = Ier = 10000 m 2 t.,
Ibe = led = rd. = 30000 m 2t and E = 2000 tjcm 2.

(113) Using the method of virtual work, calculate the displacement

at the roller support b of the structure shown in Fig. 1.174 taking the
effect of the axial deformation into consideration.
A = 400 cm 2, 1 = 2 x 10. cm' and E = 200 t jem 2
174

10 lOt
10

5.5 6m
3.5
I I
6x4 =24m
Fig. 1.174

(114) Using the method of virtual work, calculate the absolute displa-
cement at joint d of the composite structure shown in Fig. 1.175. For
column abe I = 15000 cm 4, for truss members LjA = 10 cm- I and
E = 2000 tjern 2.

0.51 2.251
c
r 1
b 4.5
2.25
t
2

r
2
d 3

4m
41;
i
4m

a ~3X1.5=4.5J
Fig. 1.175
c a
,.,.,}
Fig. 1.176

(115) Using the method of virtual work, calculate the vertical dis-
placement at joint d of the structure shown in Fig. 1.176. For column
ab I = 1500 cm 4 , for the link be LjA =10 cm-I, for the truss members
LjA = 15 cm- I and E = 2000,tjcrn 2 .

(116) Using the method of virtual work, calculate the relative displa-
cement between joints g and h, and joints e and f of the trussed frame
shown in Fig. 1.177. For columns aec and bfd I = 36000 crn 4, for truss
members LjA = IOcrn- I and E = 2000 tj=2.

 175

6
3

b b
3 ----I--- 3 m ---l I-- 6m_..i-_ 6 -----<

Fig . 1.177 Fig. 1.I7S

(117) Using the method of virtual work, calculate the vertical displace-
ment and the relative rotation at the intermediate hinge c of the frame
with a tie shown in Fig. 1. I 78. For frame members EI = 6000 m 2 t , for
tie de L/A = 20 cm-1 and E = 2000 tlcm 2.
(liS) Using the method of virtual work, calculate the vertical displa-
cement and the relative rotation at the intermediate hinge C of the trussed
beam shown in Fig. 1.179. For beam ab EI = 20000 m 2 t, for all struts
L/A = 10 cm- I , for all ties L/A = 20 cm- I and E = 2000 t/cm 2 .

211m
r~!!!!!!!g!!!!!!!~

~ 4 dl 4 4m Ie 4 --l
Fig. 1.179

(119) Using the method of virtual work, calculate the displacement

at the roller support b of the structure shown in Fig. I.ISO. For arch EI
= 40000 m 2t and for tie EA = 200000 t.

10 lOt 10
10 10
c

~ _ _ _ 6x4=24m _____~:r

Fig. I.IS0
176

(120) Using the method of virtual work, calculate the rotation at sup-
port b and the horizontal displacement at point c of the structure shown
in Fig. 1.181. For column bc I = 20000 cm" for bracing members LfA
18 cm~l and E = 2000 tfcm 2

6 61 6

r c
2m,.
2m
t .
a tU ~ a
L1.5~
...... 3--<-4m~4 --,--3
Fig. 1.181 Fig. 1.182
-_/"'

(121) Using the method of virtual work, calculate the vertical displa- . " ' -__
~

cement at point c and the displacement at the roller support b

of the structure shown in Fig. 1.182. For frame EI = 20000 m 2tand for
tie de EA = 40000 t.

(122) Using the method of virtual work, calculate the vertical displace-
ment at joint c and the relative displacement between joints d and e of
the trussed frame shown in Fig. 1.183. For columns EI = 12000 m2t,
for truss members LfA = 25 cm~1 and E = 2000 tfcm 2.

2 2 1 c 1
3m 1m

T
'"
a 6
a b j
1---4x4=16m _ _-I

Fig. 1.183

(123) Using the method of virtual work, calculate the absolute dis-
dlacement at joint d of the structure shown in Fig. 1.184. For column
 177

ac 1 = 32250 cm', for tie bc LfA = 25 cm- I, for truss members L/A =
15 CIl)-l and E = 2000 tfcm 2

(124)-(126) Calculate the rotation at c and the vertical displacement

and relative rotation at the intermediate hinge d of each of the beams in
~.
Figs 1.140 - 1.142 in the absence of the applied loads if support b settles

II by 1.5 cm.

(127) Calculate the vertical displacement and relative rotation at

I each of the intermediate hinges d and e of the beam in Fig. 1.143 in the
absence of the applied load due to :
a) a clockwise rotation at the fixed support a = 0.005 rad.
b) a vertical settlement at support b = 3 em.

(128)-(131) Calculate the displacement at the roller support of each of

I
the frames shown in Figs 1.168, 1.171-1.173 due to the following tempera-
I
t ture variations in the absence of the applied loads:
a) a uniform drop in temperature of 20C.
b) an increase in the outdoor temperature of 200 accompanied by
a similar drop in the indoor temperature.

r c) an increase in the outdoor temperature of 30C accompanied by

a drop of 10 0 e in the indoor temperature.
The frame has a uniform section 50 em. deep and at = 1 X IO~5 re.
(132) Calculate the relative rotation at the intermediate hinge c of
the frame shown in Fig 1.167 in the absence of load due to an increase
in the outdoor temperature of 20C accompanied by a similar drop in the
indoor temperature. The frame has a uniform section of depth 60 cm.
and a, = 1 X 1O->rC.
(133) Calculate the three displacement components at d of the frame
shoWn in Fig. 1.166 due to a clockwise rotational slip of 0.002 rad.
at support a in the absence of the applied loads.

(134) Calculate the relative rotation at the intermediate hinge c of

the frame shown in Fig. 1.167 due to equal outward displacements of 1
cm. at each of supports a and b in the absence of the applied loads.
178

Graphical Methods for Truss Deflection

(135)-(137) Make a neat sketch of the displacement diagram of each

of the trusses shown in Figs. 1.154-1.156. Assume reasonable member
deformations consistent with the signs of the member forces under the
given case of loading.

(138)-(140) Construct the displacement diagram of each of the trusses

shown in Figs. 1.185-1.187 if under the given loading all tenison
members are stressed to 1.2 t/em 2 and all compression members to 0.8
tf cm 2 and E = 2000 t fcm 2 .

9
10 10
, 31 1
2m
4 "
2m
~, 'i
t
4m
2
~,
2

di
2
La
------< 1-1. 5-!-1.s::i \--2x2=4 ----I--

Fig. 1.185 Fig. 1.186 Fig. 1.187

(141),(142) Construct the displacement diagram of each of the trusses

shown in Figs. 1.88 and 1.89. Assume LfA = 10 cm-1 and E = 2000
tfcm 2 . Indicate the vertical displacement at joint c.

91
d
r
3m
~,

--I- 4

Fig. 1.188
- J

Fig. 1.189
 179

(l43),(I44) Construct the displacement diagram of each of the

t:cusses shown in Figs 1.190 and 1.191. The cro!ffi-sectional area-of all ten-
sion members = 16 em 2, the cross-sectional area of all compression
members = 20 em 2 and E = 2000 tlcm 2. Plot the deflected sbape of
the bottom chord indicating the vertical displacements at all the panel
points. Check the displacement at joint d analytically.

ftit11 'I~~---k-.:h
8 81 8 41
----13
41
m

Fig. 1.190 Fig. 1.191

(145),(146) Construct the displacement diagram of each of the trusses
shown in Figs. 1.192 and 1.193. Indicate the maximum vertical and hori-
zontal displacements. Check the values thus obtained analytically. Assume
L/A = 10 cm-1 and E = 2000 tlcm 2.
k
81 k
1
6
~--~~----~----~
;1
1~~--~dL-~~~~
3x4=12 m 1----- 3x8=24 m

Fig. 1.192 Fig. 1.193

(147)-(150) Make a neat sketch of the displacement diagram of each
of the trusses shown in Figs. 1.157, 1.158, 1.194 and 1.195. Assume rea-
sonable member deformations consistent with the signs of the member
forces under the given case of loading. Indicate in each case the member
assumed to remain fixed in direction.

2 3 5 1 41
h g f
I
3m

I c d
~
~ 4x3=12m

Fig. 1.194 Fig. 1.195.

 180

(151),(152) Construct the displacement diagram of each of the trusses

shown in Figs. 1.196 and 1.197 if under the given case ofloading all tension
members are stressed to I tfem 2 and all compression members to 0.6 t fem 2
and E = 2000 tfem 2. Indicate in each case the member assumed
to remain fixed in .direction.

1.5 1.51

7m
k1
!,,-+-~:.....-c~...J?..---Jn "*3
3
...i

:::...- - 3x4 =12 m L2X4 _.l--_3x4=12m ~

Fig. l.I96 Fig. l.I97

(153),(154) Construct the displacement diagram of each of the trusses

shown in Figs. 1.198 and 1.199. Assume L/A = lOcm- 1 and E=2000 t/cm 2
petermi!J;c the absolute displacement of joint h and indicate in each case
the member assumed to remain fixed in direction.

31

= __4x3=12m
Fig. l.I98 Fig. l.I99

(155),(156) Construct the displacement diagram of each of the trusses

shown in Figs. 1.162 and 1.164. Compare the vertical displacement at
joint c and the displacement at the roller support b obtained graphically
to their corresponding values obtained previously by the method of virtual
work

(157),(158) Construct the displacement diagram of each of the trusses

in problems 142 and 143 (Figs. 1.189 and l.I90) if the roller support settles
by 0.1 cm./t reaction.

(159),(160) Construct the displacement diagram of each of the trusses

in Figs. 1.195 and 1.196 if the rollor suppott settles by 0.2 cm./t. reaction.
 lSI

(161)-(164) Construct a properly oriented Williot diagram of each

of,the trusses in Figs_ 1.196-1.199. Check the values thus obtained against
those obtained from the corresponding Williot-Mohr diagram.

(165) Construct the displacement diagram of the truss shown in Fig.

1.200 due to a unit load applied at the roller support b. Using this diagram
determine the displacement at the roller due to :
a) horizontal loads H, H = 2 t.
b) vertical load V, V = 6 t.
Assume L/A 10 cm- l and E = 2000 tlem 2 .

~ .1
1.5
I
H
v
8x 3= 24 m _ _ _" -_ _=<

r Fig. 1.200

(l66) Construct the displacement diagram of the truss shown in

Fig. 1.201 due to a unit load applied at the roller support. Using this
diagram, calculate the displacement at the roller due to 25t. acting as
indicated. Check the value thus obtained analytically.
Assume L /A = 10 cm l and E = 2000 t/cm 2

Il
25t 6m

L 12m
Fig. 1.201
9
 182

Influence Lines of Displacements

Note: An influence line is not complete unless all the critical ordinates and signs are
clearly indicated.

(167) Draw the influence line for the deflection at the free end of a
cantilever 4 m. long if the load is transmitted to it at quarter points. EI
8000 m 2t.

(168) Draw the influence line for the deflection at the end of the over-
hang in an overhanging beam of span length = 8 m and overhang length=
= 2m. The load is transmitted to the beam at panel points spaced at 2m.
and EI ~ 10000 m 2t.

(169) ,(170) Disregarding the loads indicated, draw the influence line
for the deflection at the overhanging end of each of the beams in Figs-
1.134 and 1.135. EI ~ 10000 m 2t.

(171),{l72) Each of the two beams in Figs. 1.202 and 1.203 is braced
by a system oflink members as shown. Assuming that member cd is cut
along its length .. draw the influence line for the relative displacement be-
tween the ends of the cut member. EI (for beam ab) ~ 40000 m 2t.
a b

[E ~ Vd
LLll-_ ____~~ . --11~1--~. _ _ _ __
---'c 6x 3 = 18 m _

Fig. 1.202
c d

5x5:25m
Fig. 1.203

(l73), (174) Draw the influence line for the displacement/at the roller
support b of each of the frames in Figs. 1.204 and 1.205 if the load is trans-
mitted to the frames at panel points spaced at 3m and EI ~ 25000 m 2t.
 183

r
Gm
t
La 1
Gm

La b b ~
~ 12 m I 3-+- 3-1 L9m I 3-f
Fig. 1.204 Fig. 1.205
U

(175),{I76) Draw the influence line for the relative rotation at the
intermediate hinge c of each of the frames in Figs. 1.206 and 1.207 if the
load is transmitted to the frames at panel points spaced at 3m and EI =
25000 m 2 t.

I r c r
La
c
1
~
6m 6m

La b b ~
I 12m 3-,- 3-J 1--9m --1--3.-;
l
Fig. 1.206 Fig. 1.207

(177) Draw the influence line for the vertical displacement at joint
c of the truss shown in Fig. 1.208. Calculate then the extreme values of
the displacement at c if the truss is subjected to a panel D.L. = 6t
and L.L. = 10 t. Assume LjA = 10 cm- 1 and E = 2000 tjcm 2

}m (1)
~/SZSZT'\Z~
f- 3x8=24m ~ 2x6=12 ~
Fig. 1.208

(178) Draw the influence line for the vertical displacement at JOInt
c of the truss shown in Fig. 1.209. Assume LjA = 6 cm- 1 and E = 2000
tjcm 2
'
 184

1 LI--,-':---....L-..,,--
4
k- 2x4=8 - - t - - - - 3x4=12 m _ _--<

Fig. 1.209

(I79) Load is transmitted to the truss shown in Fig. 1.210 a t panel

points a - h. Assuming that member cd is cut alo'(g its length draw the
influence line for the relative displacement between the cut ends of this
member. Assume L/A ~ 10 cm- 1 and E ~ 2000 t /cm2 .

.L
1m
d

r
4

~~__+-__*c~~__~
3
19k-____~~______~____~~______~b
4x6=24m
Fig. 1.210
 CHAPTER 2

THE METHOD OF CONSISTENT DEFORMATIONS

2.1 Introduction
A statically indeterminate structure is one which cannot be analysed
by the application of the equations of statics alone. There are two main
types of statical indeterminancy;cxternal related to the external reactions,
and internal concerned with the internal straining actions in the various
members of the structure. At this stage, the student should be quite famil~
iar with the criteria ofstabilitY3staticai determinancy and indeterminancy.
For a useful discussion on the subject, the student is referred to sections
2.6, 4.2 and 5.3 of Part l.
l The method of analysis presented in this chapter will be referred to
as the method of consistent deformations, although it is also known
r in other names such as the unit load method" flexibility method"

l force method" or the superposition equations method. All these names

are well justified and, as will be seen subsequently related to the method
of analysis.
The method of consistent deformations is applicable to the analysis
of all types of statically indeterminate structures. It can be used to analyse
beams, frames" arches, trusses or composite structures in which some
members are mainly subjected to bending and others to axial forces.
Moreover" it is used for analysing the structures under load as well as
no-load effects such as temperature change" shrinkage of the meterial of
construction, support movements or fabrication errors. The method is
basen on the principle of superposition which, for the purpose at hand"
may be stated as :
The deflection or slope at any point on a structure as a result of a number iffor-
ces, in:luding the reactions, acting simultaneously is equal to the algebraic sum if
the deflections, or slopes, at this particular point as a result oj these loads acting
individually .
From the discussion in section 3.7 of Part I" it will be seen that it is
valid provided that the body is elastic" the material of construction has
a linear stress-strain curve, and the geometry of the structure does no_
change during the application of load; elastic strains being neglected.
185
186

This principle and a knowledge of the initial supporting or continuity

conditions of the given structure, provide a mea ns for establishing a
number o f equations, which with the equations of statics arc sufficient to
determine the external reactions and the interna l straining actions in
all the m embe rs of the structure.
The basic principles of the method are quite easy to follow and can
be best d emonstrated by a series of examples. In order not to distract the
student's attention from the main principles, it is thought advisable at
this stage to develop the method with regard to a specific type of structure.
Statically indeterminate beams have been chosen for this purpose . L ater,
it will be an easy matter to exte nd the developed principles to other types
of structures .
2.2 Developm.ent of the m.ethod of consistent deformations
Consider the restrained beam shown in Fig. 2.la. Since ther e is one
reaction component in excess to the available equa tion s of equilibrium,
the beam is statically indeterminate to the fir st d egr ee. It follows that
one of the reaction components may be consider ed a s redundant.

(a)
4~: II ICons!.
II Ii II
El
~b
14 L

Fig. 2. 1
Let the vertical reaction component at end b be chosen as the redund-
a n t. When t his reaction i s temporarily removed, the remaining structure
is the cantilever shown in Fig. 2. lh. This stable and statically determin-
a te structure which h as resulted after the removal of the r edundant is
called the main system. Under the given load, end b in the main system
will deflect vertically an amount abO as indicated in Fig. 2.lb . The cor-
rect value of the redunda n t reaction at poi n t b X I is the value necessary
to b ri ng this point back to its original position of zero deflection. Under
 187

the action of XI, point b in the main system will deflect vertically some

[ amount which is impossible to calculate until the value of XI has been

determined. If, however, the vertical deflection at point'b in the main
f system due to a unit load acting there in the direction assumed for XI
I is 8bl as indicated in Fig.2.1c, then the deflection due to XI will be
X,3 bl . Now, in order to satisfy the displacement condition in the given
beam that the deflection at point b is zero. . the algebraic sum of the de-

I flections at point b in the main system due to the applied load and the
as-yet unknown redundant reaction must be zero. Mathematically, this
condition may be expressed as :
t SbO + X, Sbl = 0
I
I
2.1
I Sbl

I
The beam in Fig. 2.1a may have been made equally stable and
statically determinate by introducing a hinge at end a. The main system
will then be the simple beam shown in Fig. 2.1d, and the redundant X',
is chosen as the fixing moment at end a of the beam. Under the applied
load, the tangent to the elastic curve at point a will rotate an angle aaO'
The correct value of X't is the moment necessary to bring the tangent
to the elastiC curve at this point back to its original position of zero slope.
If the rotation at point a in the main system due to a unit moment acting
there is aa 1 as indicated in Fig. 2.le, then the condition that the slope
to the elastic curve at end a of the given beam is zero may be expressed
mathematically as :
aao + X'I aat = 0

X 't = - aaO 2.2

a,l
The numerical values of the displacements 8bO and 8bl and the
rotations al!0 and u a1 may be calculated by any suitable method. On
substituting these values in the superposition equations 2.1 and 2.2, care
must be taken to use the proper signs. For example, in the first main
system, 8bO and 8bl are considered positive when they occur in the same
direction assumed positive for the redundant XI' If the result in equation
2.1 comes out positive, XI actsin the assumed direction. If however, the
result comes out negative, Xl acts in a direction opposite to that assumed.
Using one of the methods for deflection computations described
in chapter 1 then,
wL4, L3
for the first main system, 3bO = --~--
8 EI
 188
Substituting these values in equation 2.1,
3wL
8
WL3 L
for th e second main system) uao and aat = 3 EI
24 EI
Substituting these values in equation 2.2 ..
wL2
X'! = - --
8
The directions of XI and X'I must have been assumed incorrectly
since their values came out negative.
Either equations 2.1 or 2.2 with the three equations of equilibrium
may be used to calculate the reactions and then the S.F. and B.M.Ds. may
be drawn as for statically determinate beams. Alternatively, the reactions,
shearing forces J bending moments or any other load function may be
obtained by applying the principle of superposition. Thus)
F ~ Fo XI FI + .. . 2 .3
where F is any load function in the given statically indeterminate bearn,1
Fo the corresponding value of the load function in the chosen main system
due to the applied loads, XI the calculated value of the redundan t and
F I the value of the load function under consideration in the main system
d ue to the case of loading X l = 1.
As an application to equation 2.3, consider the beam shown in Fig.
2 .2. The vertical reaction at the right support considered as red undant
was found to be XI ~ - 3wL/8 .
Main System Given Beam
w~
f"::' T w x,=,! eM, w
~!!! I ! I ! ~ ~5=1I 0:
1I D::I
I D:ICRbt
,
twL h R, h. R y,
wL
, (a l
, 5wL
- + 8 +
GoD. O,.D. Final S.F.D. 3
wL
w~ ( bl 8
2"
L

M,.D. Final B.M.D.

(c)
Fig. 2.2

r
 189

Referring to Fig. 2.2a, the reactions of the given beam are given by :
~ Ya = wL -83 5
wL (l) = S wL (upward)

r Ma
- WL2
2
3 wL2
-SwL(-L) = - - 8 -
WL2
i.c. - - (antidockwise)
I 8
~
I =
3
Xl = - -wL, i.c.
3wL
(upward)
8 8
(
With reference to Fig. 2.2 b, thc final S.F.D. for the given beam
is obtained by superimposing the Qo-diagram and the QJ-diagram times
(- 3wLj8) .
Similarly, with reference to Fig. 2.2 c, the final B.M.D. is obtained
by superimposing the Mo-diagram and the MI-diagram times ( - 3wLj8).
Qo and Mo-diagrams are the S.F. and B.M.Ds. of the chosen main
system due to the applied loads; Xl being zero. Also, QJ and Mt -diagrams
are the S.F. and B.M.Ds. of the main system due to the case of loading
XI = 1.
(As an exercise., the student is recommended to calculate the reactions
and draw the S.F. and B.M .Ds. for the same beam using the simple beam
main system and its corresponding value of the redundant.)
As a second example, consider a beam continuous over three supports
as shown in Fig. 2.3a.. and suppose that it is required to calculate its
reactions under the given: case of loading.
The beam is statically indeterminate to the first degree. The simple
beam shown in Fig. 2.3b may be chosen as a main system. Under the
given uniformly distri buted load point b will deflect vertically an amount
3 bO = 5w (2L) 4j384El = 5 wL 4/24 EI. Duder the unit vertical load
indicated in Fig. 2.3 c, point b will deflect an amount 3bl = 1 (2 L)3j48 El
=L 3/6 EI. The correct value of the redundant X" which in effect is the
vertical reaction at support b ..may be obtained from the condition that the
deflection at b in the given beam must be zero. Thus,
3bO + XI 3bl = 0

XI
3 bO
= -a:;-= -24EI
5WL4j 6El
L3
= --4-
5wL

The negative sign indicates that XI (= Vb) is upward.

From statics, Y a = Y c = - I ( 2wL - - - = - 3 wL (upward)

5WL)
248
190
w
(al
a b c
L L

X,=O
(b)
&.0

(el
~
r=1
&"
~
(d)

~ =-=
f =:j ~D

( e)
\- l-==J
Fig. 2.3

The overhanging beam shown in Fig. 2.3d could have been chosen as a
main system and the vertical reaction at c as the redundant X' I' Under the
given uniformly distributed load, point c will deflect an amount i\o which
is best found by the conjugate beam method; 3,0= wL 4/4 EI. Under the
unit vertical load indicated in Fig. 2.3e, point c will deflect an amount
Sci which is again best found by the conjugate beam method;
ScI = 2L3j3EI. The reaction at c X'I is the force necessary to bring point
c back to its original position of zero deflection. The following super-
position equation expresses this condition.
3'0 + X', 3" = 0
X' _ _ 3,0 __
, - 001 - 4 EI
WL4/23 La _
EI -
_ 3 wL
8' ,.e.
3wL
- - (upward)
8
3wL 5wL
From statics, Y a = -8- (upward) and Y b = -4- (upward)

These values arc identical to those obtained previously.

From the above two examples, it may be seen that there are more
than one way to reduce a given statically indeterminate structure to a
 191

stable and statically determinate one, i .e., there are more than one pos-
sible main system. One main system, however, may prove to be better
than the others as it leads to a quicker solution . In the first example con-
, sidered, there is a little preference between the two main systems chosen.
In the second example,the overhanging beam main system is not as simple
as the simple beam main system as the calculations of the deflection coef-
ficients DcQ and 0CI are more involved than those of DbO and Obi'

This demonstrates the importance of choosing a suitable main system.

2.3 Deflection notations

The met hod of analysis presented in the previous section is equally
applicable to structu res of higher degree 0[5tatical indeterminancy. How-
ever, before dealing with structures of higher degree of indeterminancy
and to avoid confusion, it will be advantageous to adopt notations for
deflection good enough to express any kind of oisplacement component
whether linear or angular, at any point on the structure due to any case
of loading. In this book, linear deflection components at any point will
be denoted by () and rotational deflection by a. A subscript added to
() or a refers to the particular point on the structure where the deflection
oc(.urs. Also, deflection components due to particular cases of loading
are usually required. In such cases, double subscripts are used. The first
subscript represents, as mentioned above, the point on the structure where
the deflection component occurs and the second the case of loading con-
sidered. Thus, if m is any point on a statically indeterminate structure,
deflection at m in the given structure
deflection at m in the ma~n system, all redundants X o
()ml deflection at m in the main system due to redundant Xl
8m2 deflection at m in the main system due to redundant X 2 =

8mm = deflection at m in the main system due to redundant ~ = 1

Similarly,
am = rotation at m In the given structure.
amo = rotation at m in the main system, all redundants X = 0
a ml = rotation at m in the main system, due to redundant Xl 1
am 2 = rotation at m in thc main system due to redundant X 2
..... ............ ......... . ... .... ........ ... .... ... . .. .........

a mm = rotation at m in the main system due to redundant Xm = I

192

It remains to mention that 8 and a. may be used to denote relative

linear and rotational displacements and that the letter y is ..ometimes
used to denote a linear displacement when it is required to distinguish
between two linear displacement compGnents at a single point on
a structure.

2.4 Generalized Maxwell's theorem. of reciprocal deflections

In section 1.37, a special case of Maxwell's theorem of reciprocal

deflections as related to vertical deflections has been developed. The
theorem, however, needs not be restricted to vertical deflections and may
be extended to include displacements in any direction as well as rotations.

One use of the theorem of reciprocal deflections has been already

demonstrated regarding the construction of influence lines for deflections.
The theorem has another use, of no less importance, in the solution of
statically indeterminate problems and its application generally leads to
a considerable reduction in the effort involved.

Before attempting to djscuss the analysis of structures of higher degree.

of indeterminancy, it will be advantageous therefore to establish the the-
orem of reciprocal deflections in its generalized form.

a) Linear displacements

Consider the linearly elastic structure shown in Fig. 2.4. Let 8mn
be the displacement at point m in direction m'm" due to a load P
acting at point n in direction n'n"-Fig. 2.4 a. Also let 80m be the
displacement at point n in direction n'n" due to an equal load P acting
at point m in direction m'm"-Fig. 2.4 h. These displacements may be
found independently by the application of the method of virtual work.
Thus, considering bending strains only,

=S
1)
rno
=
S
MnM'm dl
EI
and 1)
nm
M,.,M' n dl
EI

where ~ and Mm are the bending moments due to load P acting at

points nand m respectively'. and M 'm and Min are the bending moments
due to a unit load applied at the respective points.
Applying the principle of superposition,

M,. = PM', and M,.,= P M'm

 193

ni
p

(al
\m ./' ---- -; n

'/ \mo ri'

m"
~

,
\m
p J
n
m /" --- 00m
(bl

Y'/\m'
1
n" :>

Fig. 2.4

PM' M' dl PM'm M'dl

or
Hence, 8mn =
S n m and
EI
=
S
n
EI
2.4
i.e. the displacement at point m in a direction m'm" due to a load P acting
at some other point n in direction u/n", is equal to the displacement at
point n in the direction n'n" due to an equal load P acting at point m in
the direction m'm".
b) Angular displacements
Referring to the linearly elastic body shown in Fig. 2.5, let a mn be
the rotation at point m due to a moment M applied at n - Fig. 2.5a, and
let nnm be the rotation at point n due to an equal moment M acting at
point m-Fig. 2.5 b . Applying the method of virtual work,

a =SMnM'mdl and a =SMmM'ndl

- EI - m
where Mn and Mm are the bending moments due to a moment M acting
at nand m respectively, and M'm and M'n are the bending moments due
to a unit couple applied at the respective points. From the principle of
superposition, 1
MM'm-
194

(a)

(b)

Fig. 2.5

. MM' M' dI
Hence,

or
aUlD =
J n
EI
m and
=J 2.5
i.e. the rotation at point m (measured in the sense of a moment acting at
m) due to a m o ment acting at some other point D, is equal to the rotation
at n (measured in the sense of the moment at n) due to an equal moment
acting at m.
c) Linear and angular d isplacements
Referring to Fig. 2.6, let 8mn be the linear dispiacement at point m
in a direction m'm" due to a moment M acting at point n in the sense
indicated in Fig. 2.6a, and let anm be the angular displacement at
point n in the sense chosen for M due to a load P acting at m in the
direction m/m"-Fig. 2.6b. Applying the method ofvirlual work,

3
mn
=J MnM'm
EI
dI
and a
om
= S Mm M 'n dl
EI

where Mn represents the bending moments due to a moment M acting

at n, ~ the bending moments due to a load P acting at m" M'm the
bending moments due to a unit load at m and 1\1'n the bending moments
due to a unit couple at n .
 195

(a)

(b)

Fig. 2.6

From the principle of superposition,

M", = P M'm and Mn = M M'n

f
M' M' dl M'm M in dl
Hence, Smn = M n m and
EI EI

Now if M is numerically equal to P,

2.6

i.e. the li near displacement at a point m (measured in the direction and

sense of a load at m) due to a moment M acti ng at some other point n J
is equal to the angular displacement at n (measured in the sense of the
moment at n) due to a numerical1y equal load acting at m.

It should be emphasized that the application of Maxwell's theorem

of reciprocal deflections is not restricted to a particular type of structure.
It is equally applicable to beams. . frames, arches, trusses, composite stru-
ctures or generally to any linearly elastic body.

As this theorem is repeatedly used, it should be fully understood. The

examples shown in Fig. 2.7 may be helpful in this respect.
196

~
a 6}bC

b -
0",
Equal

1m!

1> 0
b
1 mt

Ct Equal

Fig. 2.7

2.5 Structures of higher degree of indeternlinancy

To illustrate the usc of the deflection Hotations proposed in section 2.3
and the application of the method of consistent deformations to the analysis
of structures of higher degree of indeterminancy, consider the beam shown
in Fig. 2.8a.

The beam is three times statically indeterminate. Let the moment at

end 1 X h the vertical reaction at support 2 X 2 and the vertical reaction
at support 3 X), be chosen as reclundants. The resulting main system
will then be the simple beam shown in Fig. 2.8 h. Under the applied loads,
end 1 in the main system will rotate an amount alO' point 2 will deflect
vertically downwards an amount 020 and point 3 an amount 830 - These
displacement components are shown in Fig. 2.8 b. The correct values
of the redundants XI' X 2 and XJ are those necessary to restore the beam
to its original condition of zero slope at end 1 and zero deflection at points
 197

2 and 3. Applying the three redundants individually to the main system,

the resulting displacement components at the locations of the redundant.
will be as indicated in Figs. 2.8 c - e.

~1'1
..-\\_{-.2
(al ~!--~
--Ll.-J1L--1~~'='
Wl.
I :o:::o::I?S\4
iDl

(b)~
~ , lx'=O 1 I ",.0
\~~_l~&,~o~____~&~JO~-'-'~
'0 1

("~ I", ... ' !

~) k t" .,1 ~1
(.) 1)"" _Ov j"..~
'Fig, 2.8

The conditions of zero slope at end 1, and zero deflection at points

2 and 3 may be expressed mathematically as :
a, - a,O + X, all +X + Xl al3
2 4 12 0
82 - 820 + X, 821 + 2 + X3 02l =
X 022 0
83 830 + X, 831 + Xi 832 + X3 833 0
The displacement components in these equations may be calculated
by any convenient method and the simultaneous solution of the equations
yields the unknown values of the redundants X" X 2 and X 3, These, in
effect, are the reaction components M b Y 2 and Y3' The remainder of the
reaction components, Xv Y I and Y. may be obtained from the three
equations of equilibrium.
The three superposition equations above, comprise twelve displace-
ment components. However, using Maxwell's theorem of reciprocal deflec-
tions, it is obvious that a12 = 821 , au = 831 and 823 = 832 , and
thus the number of displacement components to be calculated is
reduced to nine only.
198
This example demonstrates the amount of labour that could be saved
by the application of Maxwell's theorem of reciprocal deflecti"ns.
As a second example on structures of higher degree ofil1determinancy..
consider the fixed-beam shown in Fig. 2.9 a. In general, this beam is
statically indeterminate to the third degree but under vertical loads it is
indeterminate to the second degree only. This is because in this case there
are four unknown reaction components; a moment and a vertical reaction
at each of the two ends" and two useful equations of equilibrium; L Y = 0
and l' M = O. There is a number of possible main systems} two of which
are shown in Figs. 2.9 band e.

Fig. 2.9

Consider first the cantilever main system in Fig. 2.9 b. The redundants
chosen are the moment and the vertical reaction components at end 2.
These and the corresponding deflection components are indicated in Figs.
2.9 c and d . Using the moment-area method or the conjugate beam me-
thod, the displacement components are easily obtained and are as follows :
Pa 2 (3b + 2a) Pa 2
azo =
6 EI 2 EI
L2 L
a21 -
2 EI EI
L3 L2
--
3 EI
a22 -
2 EI
 199

The deflection conditions that must be satisfied in the given structure are:
a, - a,o + x I a21 + x, an 0
0.2 0.20 + XJ 0.21 + X 2 0.22 0
Substituting the displacement components in these two equations and
solving them simultaneously, the following values of the redundants are
obtained.
Pa 2 b Pa 2 (L + 2b)
XI - - and X, =
L2 L3
The negative sign indicates that X 2 acts in an opposite direction to that
assumed, i .e. upwards. The remainder of the reaction components may
be obtained from the equations of equilibrium.
Consider next the simple beam main system shown in Fig. 2.9 e. The
redundants chosen are the moments at both ends. These and the corres-
ponding deflection components. . which are best found by the elastic-load
method, afC in~icated in Figs. 2.9 f and g.
Pab Pab
0.10 = (a 2 + 3ab + 2b 2) alo = - - (b 2+3ab+2a 2)
6 EIL2 6 ElL2
L L
all all
3 EI 6 EI
L L
ail = a2l =
6 EI 3EI
(Notice the amount of labour that may be saved by applying Maxwell's
theorem on one hand and by making use of the symmetry of the structure
and loading on the other). The deflection conditions that must be
satisfied in the glven beam are:
al alO + X' I all + X', a" 0
a2 a,o + X'I a21 + X', an 0

Substituting the displacement components in these two equations and

solving simultaneously, the following values ofthe redunrlants are obtained.
Pab' Pba 2
X'I = -- --and X' ,= - -L2
-

l
L2
The negative signs ind~cate that both X ' 1 and X'2 act in directions
opposite to those assumed.
These values are consistent with those obtained usmg the cantilever
mai n system.
200

2.6 Support Move_ _

The method of consistent deformations is equally applicable to the
analysis of statically indeterminate structures due to support movements.
Consider the statically indeterminate beam shown in Fig. 2.8 and let
8" 6.2 and .63 be the actual displacements that occur in the original
structure. These may be due to partial rotation at the fixed end and set-
tlement of the two intermediate supports. Such a displacement must be
equa1 to the sum of the displacement components in the main system
due to the applied loads and all the redundants. Thus for locations 1,2
and 3 in the given beam,
at 410 + Xl all + X,
02 0'0 + X. 0" + X,
0, 0'0 + x. 0" 0" + X, 0" = 6,
where 8 1 , 1:::..2 and .6.3 are given support movements.
A simultaneous solution of the above equations gi"es the values of the
redundants X" X, and X, due to the combined effect of the applied
loads and the support movements. It is costumary, however, to analyse
structures for the effect of support movements separately for reasons such
as the desire to draw the envelop S.F. and B.M. curves due to the
combined effect of the applied loads and the estimated support movements.
In such cases the deflection components due to the applied loads; al0,
820 and 830" are all zero and the above equations reduce to :

X. a" + X, a" + X a" 8.

X. 0" + X, 0" + X, 0" 6,
X. 03. + X 0" + X, 0"
2 6,
Here, the deflection components all' a12' a13, 8 2 h and 833 may
be calculated and 8 1, b.2 and b.3 are known values. Further, in applying
superposition equations as these, particular attention must be paid to
the signs used. At any particular point all the displacements occuring in
the same direction have a similar sign. For example, if at point 3 down-
ward 8 is considered positiV'e then each of the displacement components
830, 831 , 3 2 and 033 is positive if it occurs in t he downward direction
otherwise it will be considered negative.

2.7 General procedure of analysis - discussion

While the method of consistent deformations has been developed by
discussions on statically indeterminate beams, it is completely general in
that it is equally applicable to the other types of statically indeterminate
 201

structures regardless of their degree ofindeterminancy. The only difference

may be in the method used for computing the displacements.
The procedure of analysis is summerized be1ow.
(I) The degree ofindeterminancy of the structure under consideration
is determined. For this.. the student is referred to sections 2.6,4.2 and 5.3
of Part 1. Briefly, however, as many supports are removed and as many
members are removed or cut to reduce the given statically indeterminate
structure to a stable and statically determinate onc. The degree of inde-
terminancy then is equal to the number of the removed restraints.
(2) Assuming the structure to be indeterminate to the n degree,
n redundant reaction components and or internal forces (or moments) are
removed to reduce it to a stable and statically determinate main system.
The deflections due to the applied loads at the points where the redundants
act; 3 10 , 320, . . . . . 3no are calculated. On making the choice
for the redundants, care should be taken not to fall in the serious mistake
of adopting an unstable main system. Further, several possible main
systems al~ays exist. One main system, however, may prove easier than
the rest. After some practice, the student is expected to make the proper
choice. The following remarks may be helpful in this respect.
(a) Whenever possible advantage should be taken of any symmetry
in the structure.
(b) Whenever possible a main system should be chosen such that the
effect of various cases of loading, applied loads or individual rcdundants
IS localized.

(3) Unit values of the n redundants X" X" ........ . .. Xn are

applied in succession to the main system and the flexibili~y coefficients;
3 11 , 3 12 , ... .. 3nn , are calculated.
In calculating these flexibility coefficients.. one method may prove
easier than the other. This is onc of the main reasons why the student
should be aquainted with all the methods available for deflection
computation .
(4) !fthe structure is to be analysed for the action of the applied loads
only.. the following set of superposition equations may be written.
a, = a,o+ X , all + X, a" + + X DaID = 0
a, = a,o + X , 8" + X, 8" + + x" 8'D = 0
... 2.7

aD= aDO + X, aD' + X, aD' + ...... + x" aDD = 0

 202

(5) If the structure is to be analysed for the effect of temperature

change only, the superposition equations become:
3. = 3.. + X. 3" + X 2 3'2 + . .. ... + Xn 3 = 0
32 = 42. + X. 32 + X 2 322 + ...... + X. 32 = 0
........................................... ... ...... .. . 2.8
.............. . .................... .. ... . -......... .
o. = 3 + X. 0., + X 2 0.2 + .. .... + X. 0 = 0
where 8It -' 8211 8nt are the displacements due to temperature
change at the points where the redundants act.
(6) If the structure is to be analysed for the effect of support movements
only, the superposition equations become:
3, = X, 0" + X 2 0'2 + + X. 0,. = 0,.
02 = X, 02' + X 2 022 + ..... + Xn 02. = 02.
. .. 2.9

8n = X t 8DI + X 2 802 + ..... + Xn 8nn = 8ns

where Sis' 82s ' . 8ns are known displacements.
(7) The sets of equations in either step (4), (5) or (6) are solved
simultaneously to obtain the unknown values of the redundants Xv X 2 ,
.... Xno If the value ofa redundant comes out positive, it acts in the
direction assumed and ifit comes out negative it acts in a direction oppos~
ite to that assumed.
(8) The precalculated values of the redundants in their proper direc-
tions are re.placed on the strucutre. Considering these redundants as
applied loads the remainder of the reaction components are calculated
by applying the equations of statics and then the analysis is completed in
the usual manner as for statically determinate structures. AlternativeIy,the
following superposition equation may be applied:
F = Fo + XiF, + X 2F2 + ........ + X.F. .., 2.10
where F is any load function, Fo the value of this load function in the main
system due to the applied loads, F, the value of the load function in the
main system due to Xl = I, etc.

, 2.8 Applications to statically indeterminate belUllS

The method of consistent deformations is not usually used for the

analysis of common types of continuous beams where the moment of
 203

inertia within individual spans is constant, and ,the beams are free of
intermediate hinges. This is because in such cases a solution by the method
of equation of three moments or the moment distribution method is
generally easier.

In the following, a number of examples is given. Although some of

the problems may be solved more easily by other methods, they are mainly
given to emphasize the principles of the method. The other problems are
examples of statically indeterminate beams where the method of consistent
deformations provides the easiest solution. Also, it should be noticed that
when the beams are analysed for the effect of applied loads only it is un-
necessary to substitute for the value of the bending rigidity EI as it always
cancels out. On the other hand, the absolute value of EI must appear in
the superposition equations for no-load effects.

ExalDple 2.1 Find the reactions and draw the S.F. and B.M.Ds for the
continuous beam shown in F-ig. 2.10a if it has constant EI.

Solution : The beam is statically indeterminate to the first degree. The

simple beam ab may be chosen as a main system and the redundant Xl.
is then the vertical reaction component at c. The main system and its
elastic curve are shown in Fig. 2.l0b, and the corresponding Mo.D. due
to the applied loads is shown in Fig. 2.lOc. The redundant is found from
the condition: Beo + Xl Bel = O.

Using the elastic-load method and with reference to Fig. 2.10 c,

20.5 X 3
WI 30.75 at 8.8 m. from b
2
20.5 X 4.8
w2 = 49.2 at 6.2 m. from b
2
24.5 X 4.8
W3 58.8 at 4.6 m. from b
2
24.5 X 3
- -- - = 36.75 at 2 m. from b.
2
LMb= 0 = ra X 10.8 - (30.75 X 8.8 + 49.2 X 6.2 + 58.8 X 4.6
+ 36.75 X 2)
918.5
r. = - - = 85
10.8
EI8,o= 85 X 6 - 30.75 X 4 - (20.5 X 3) 1.5 - (0.5 x2.5 X 3) I = 291
204

(a) o 61t

~ 3 J
6 9
( b)

(d)

8/3 ml

(e) M,.D.

6.9

6
(f) Reoclion'j's'--_ _ _-'-_ _ _--,,-_.....1_ _ _--I

1.76 t 1.84
7.16 7.16 t

(g)S.F.D.
1.76 1 76 +

1.84 1.84
4.24 4.24
7.44

(h) 8.M.D.
+ +
S.28mt 5.52

Fig. 2.10
 205

Fig.2.10d shows the elastic curve of the main system due to XI = It.
and the corresponding Mt.D. is shown in Fig. 2.10e.
Using again the elastic-load method and with reference to Fig. 2.10e,
Ws - 1/2 X 8/3 X 6 = 8 at 6.8 m. from b.
W6 1/2 X 8/3 X 4.8 = 6.4 at 3.2 m. from b.
LMb = 0 = r. X 10.8 - 8 X 6.8 - 6.4 X 3.2
fa = 6.9
EI8 cl = 6.9 X 6 - 8 X 2 = 25.4
8 291 .
XI - - -co- = - -- = - 11.4 t, I.e. 11.4 t. (upward)
8cl 25.4

Having thus found the value of the vertical reaction at c, the rest of the
reactions can be found from the equations of statics and are as indicated
in Fig. 2.10 f. The resulting S.F. and B.M.Ds. are shown in Figs. 2.10 g

l and h respectively.

Another stable and statically determinate structure may be obtained

by introducing a hinge at the central support. The main system thus con-
sists of two simple beams ac and be, and the redundant X' IJ which is the
connecting moment at the central support c, may be found from the

6t 9t
b
(a)
3 3m 3

(b)M .0
0
+ +

9mt 10.125
5/8
(e) M,.O

7.44 mt 10.125

(d) B.M.D. k=:::::::=-I---:7"'::C:.-_~~+-:---"'::::==::::,.I

+

Fig. 2.11
206

condition that the relative rotation at c must be zero. Mathematically this

condition is expressed as: aco +
X'l act = O. The relative rotations
cO and act are best found by the method of virtual work. As a
preliminary step towards this en~, the MooD. due to the applied loads
and the M 1 .D. due to two opposite unit couples applied at c are drawn.
These are as shown in Figs. 2.1 1 band c respectively. Referring to these
figures,

aco ~J MtMo dx
EI
_I_(~
EI 2
X
-
1
2
+
10.125 X 3
2
x-
2
3
x
5
8
1O.125x 1.8 ~) ~ 26.7
+ X
2 4 EI

~S M~dx
~:I
4.8
act 12+ - - xP ) ~ 3.6
EI (: X 3 EI

26.7
X't ~ - - ~ 7.44 m.t.
3.6
Having thus found the value of the redundant moment, the fi nal B.M.D.
is easily obtained by superimposing the Mo.D.to the Mt.D. times 7.44.The
result is shown in Fig. 2 .11 d, and it is seen that it is identical to that pre-
viously found in Fig. 2.10 h.

Exan>ple 2.2 Determine the reactions and draw the S.F. and B.M.Ds.
for the beam considered in the example 2.1 due to a settlement of2.56 cm.
at support c. EI ~ 2000 m 2t.

Solution : The beam is statically indeterminate to the first degree. The

simple beam ab may be chosen as a main system and the redundant reac-
tion at c Xl is found from the condition:
SeO + Xl Sci = Scs
In this equation SeO' which is the deflection at point c ln the
main system due to the applied loads, is zero. Sci which is the vertical
deflection at c due to a unit load applied there as indicated in Fig. 2.12 b
is given by :

s- ~ J-' M;dx ~ _I [~X (~)2 ~ (~)2] 25.6

c1 EI EI3 3+33 EI
25 .6
~--- X 100 1.28 cm. (downward)
2000
 207

1.11
(dlB.M . Qf.,-==-----------~
+
5.34

(el
2.56cm

(f)t.(D. '"""o-------------~
+
4.6mt
Fig. 2.12

2.56
Hence, Xl = -- = 2 t. (downward)
1.28

The reactions at supports a and b are found from statics. . and the cor-
responding S.F. and B.M.Ds. will be as shown in Figs. 2.12 c and d.

Now, suppose the beam had a bending rigidity EI = 4000 m 2 tinstead

of 2000 m 2t. In this case, 801 = 25.6 X 100/4000 = 0.64 em. and XI
=2.56/0.64=4t (downward). Consequently,the shearing forces and bending
moments along the beam are doubled. This demonstrates that the stiffer
the beam, the more affected it will be by support movements.

The overhanging beam shown in Fig. 2. 12 e could have been chosen

 208

as a main. system .. and the vertical reaction at b as the redundant X' I'
From the geometry of the deft.ected shape of the chosen main system,point b
will have a vertical deflection of2.56x 10.8/6=4.608 cm. (downward) . X',
is the force necessary to bring point b back to its original position of zero
deflection.. If the vertical deflection at point b due to a unit load acting
there as indicated in Fig. 2.12 f is 8bl then the following superposition
equation may be written.

4.608 - X', Sb' = 0

Using the method of virtual work, 8b1 is determined with reference
to Fig. 2.12 f as follows:

=s M~dx EI
= _I
EI
(~3 X 4.8 2
4.8
+ -3 X 4.8
2) =82.944
--
EI

82.944
X 100 4.147 cm.
2000

4.608
Hence, X', = - - = 1.11 t. (upward)
4.147
This is consistent with the value obtained from the first main system.

A comparison between the two solutions given above shows that the
first solution is easier since in the second the displacement at support b
due to settlement of support c had to be calculated first. This may be done
from the geometry of the deflected beam or by the application of equation
1.55. This, is why whenever support movements are involved it is advan
tageous to select the redundants at the supports where movements are
expected. It is repreated that while it is preferable and generally easier
to do so, it is not necessary as demonstrated by this example.

Esanlple 2.3 Calculate the reactions and draw the B.M.D. for the conti-
nuous beam shown in Fig. 2.13a. The moment of inertia varies as indicated.

Solution: The beam is statically indeterminate to the first degree. Choos

ing the redundant X, as the reaction at the middle support I,the Mo.D.
due to the applied loads on the chosen main system will be as shown in
Fig. 2.13 b. The redundant X, is found from the condition that the final
deflection at support I is zero. This condition is expressed as:
S,o - X, S11 = 0

.'- - - -
I
I 209
I

~ ~L-~ ____L-__ ~ ____4-__ ~ ____ ~ __ ~~J

3 -L 3 3m ___ 3 i 3 -1- 3
1---- I _ 21 -L I ----I

/
(b)MoD. I, +
I
36 36

w2
60 , 60mt

I W, W3 W.

I W, w. ~ 4.5 m.t.
r
;
(c)M,.D.

X,=1t

20.85

3 3
+ +
(d) B.M.D.
7.05 5.11 mt 5.11 7.05

Fig. 2.13

Referring to Fig.2.12 b and taking I as I"

36 X 3 54 36 X 3 54 60 X 3 90
--w- - -w -
2EI, - EI,' 2 - 2EI, - EI,' , - 2EI, EI,
60 X 3 90 3 X 6 18 3 X 3 4.5
W4 = 2EI = El ' Ws=
= EI,
-, W6=
2EI,
= -
EI,
r r EI,
Using the moment-area method
I
8)0 = -
EI,
(54 X 2 54 X 4 +
90 X 5 + + 90 X 7.5 -18 X 3 -4.5 X 7.5)
1361.25
EI,
The M,.D. due to a unit load at I 'is shown in Fig. 2.13 c. Referring
to this figure,
210
3 x 6 9 3 x 3 2.25 4.5 X 3 3.375
w ~--~-- w ~ --- w - ---
7 2E1 " E1' 8 2 (2E1 ,) - E1, ' - 2(2E1,) E1,
Using the moment-area method,
1 78.75
3" ~-- ~ (9 X
E1,
4 + 2.25x 7 + 3.375 X 8)
E1,
1361.25 E1,
Hence, X, ~ X - - ~ + 17.3 t.
E1, 78.75
The positive sign indicates that the reaction at support 1 is upward
as assumed.
From symmetry, Ro ~ R2 ~ (27 - 17.3)/2 ~ 4.85 t. (upward)
The bending moment at any section is found frqrn the relationship
M~Mo + X, M, and the final B.M.D. is obtained from superimposing
the Mo.D. to the M,.D. times 17.3. The result is shown in Fig. 2.13 d.

ExaDlple 2.4 Draw the B.M. and S.F.Ds. for the statically indeterminate
beam shown in Fig. 2.14 a. The moment of inertia varies as indicated.

51
(a)
3

(b) Mo.D. /' 18mt

(c) MrD.

9mt X=1t
1
w. W, w7
18 18mt

(d)B.M.D.
+
9
9
6
(e)S.F.D.
+ +

9 9
Fig. 2.14
 211

Solution: The beam is statically indeterminate to the first degree. Choos~

ing tbe redu.ud.ant Xl as the rea.ction at support 1, the main system will
be a cantilever. The corresponding l\1Q.D. is shown in Fig. 2.14 h. The
M(.D. due to a unit upward load at point I is shown in Fig. 2.14 c. The
redundant is found from the condition: 810 - Xl li11 = 0
Considering I = It' and referring to Fig. 2.14 b,
i
18 X 3 13.5 36 X 3 27 36 X 3 54

I
w( = 2(2E1,) E1, ' W2= 2(2EI,) = E1: W, = 2E1, EI,
81 X 3 121.5 81 X 3 121.5 153 X 3 229.5
w.= - --
~
w-
El ' 5 - 2EI t
= - -, W6 =
2EI, r EI, 2EI, E1,
i
Applying the moment - area method,
I
3(0 = -
E1,
(13.5 X I + 27 X 2 + 54 X 4 + 121.5 X 5 + 121.5 X 7

3577.5
+ 229.5 X 8) = - -
EI,

Referring to Fig. 2.14 c, and using the moment-area method,

3 X 3 2.25 6 X 3 9 9 X 6 27
-- w- =--, w9 =
w7 = 2 (2E1,) EI , ' 8 - 2EI, EI, 2E1, EI,

I 238.5
3((= -
EI,
(2.25 X 2 +9 X 5 + 27 X 7)
EI,

3577.5 EI,
X( =EI-
,
X
238.5
+ 15 t.

The positive sign indicates that the reaction at support I, is upward

as assumed.

L Y = 0 = Yo - 9 - 9 - 6 + 15, Yo = 9 t. (upward)
L Mo = 0 = Mo-9 x3 - 9x6 + 15x9, Mo = 18 m.t. (anti clockwise)

Now that the reactions are known the B.M.D. may be drawn from
the first principles or from the relationship, M = Mo + XtMto The
final B.M.D. is shown in Fig. 2.14d, and the S.F.D. is shown in
Fig. 2.14 e.

E,.ample 2.5 Draw the B.M. and S.F.Ds for the continuous beam shown
in Fig. LI5a if E1 is constant.
 212

HIm
(a)
b c
6m

3t
111m
(b)

6 6

(c ) MoD. 1<;:-- - - - - ,.-<'-- -+ - - - - --::>/-- -'\

+ +
4.5mt
6
o.s
(d)M

lmt

(e)

6mt 6

(t)B.M.D

3 3
2.5
(9)5.F.0
+ + +
1 1
3.5
Fig. 2.15

Solution: The beam is statically indeterminate to the first degree. Choos-

ing the redundant XI as the fixing moment at support a, the
resulting main system is the cantilever beam shown in Fig. 2.15 h. The
Mo.D. is shown in Fig. 2.15 c, and the Ml.D.is~own in Fig. 2.15 d . The
 ,
213

redundant is found from the condition that the slope a t the fixed support
in the original beam is zero; aa o + X I Q.al = 0
Applying the conjugate beam method and referring to Fig. 2.15 e,
the rotatians Q.ao and aat may be easily calculated and the result will
be as follows :
16.5 2.75
I
aoo = +F;I and aat =
~" E1 +
Hence, XI 16.5/2.75 = - 6m.t.

The final B.M.D. and S.F.D. aTC obtained in the usual manner and
are as shown in Figs. 2.15 f and g.

Another simpler main system could be obtained by disconnecting the

beam at the intermediate hinge d. The r edundant X ' I which is the S.F.
at d is obta;ned from the continuity condition that there is no relative
vertical displacement at d; OdO + X'I Odl = O. The displacements
adO and 3d1 are best found by the method of virtual work, and afC cal-

culated below in connection to Fig. 2.16.

(0)

12mt
G
tx;=o
+
l 4.5
. tX; =1t 3

I t
t
I
Gmt
Fig. 2.16
!I
=SM1ModX = (~x l-~3
,
~
8
dO

99
E1
_1
E12
12--x 3 x5)
x6x4.5x 1 . 5 -
2

E1

ll"-----------------------
 214

=SM2 J X =_1 (~ x
d
S
d
J
EI EI 3 6 2 + - 3 x3 2 + -6 2
X3) = -99
3 3 EI
Hence X'l = 99/99 = It.
which is consistent with the value obtained from the main system chosen
previously.

Lam.ple 2.6 Determine the reactions of the continuous beam shown in

Fig . 2.17 a due to :
(a) the given loads, (b) a clockwise rotation of 1.8 X 10- 3 rad. at a
and (c) an upward movement of 3 em. at support b. EI = 5000 m 2 t.

3
(a)

~=~1~:V~~___~~2~7~~__~i!_____(_b)_M
_O_D_l_8_m
_t---1!
1 (C)M,.D.

(~ml
91 151 (d) 61

rtrm~1- - + - - - - - - 1
\ l0:33 I t0.33 I (e) 01
~5.55ml
\ t062 I la.621 (I) of
Fig . 2.17

Solution: The beam is statically indeterminate to the first degree. A pos-

sible main system is obtained by introducing a hinge at a. The redund-
ant Xl which is the fixing moment at a is obtained from the condition
,

215

that the slope at the fixed support is zero; 0..0 +

X J a.1 = O. The
rotations 0..0 and Qal are found by the method of virtual work. Referring
to Figs. 2.17 band c which show the Mo and M,.Ds.,

M
a o =JM I odx=_1 [(9+3)27 x _I _(18 X 9)_1 ]=54
EI EI 2 2 2 3 EI

a I

=S M
2,
dx = _I
EI EI
(...!!..
3
X 12)=..2
EI

54
Hence, X, = - - = - 18 m.t., i.e. 18 m.t. (anticlockwise)
3

Having thus determined the redundant, the remaining reactions can

be easily found from the equations of statics and are as indicated in Fig.
2.17d. ./

The value of X'l due to a clockwise rotation at a is found from the

condition: X'I aal = 1.8 X 10-3

3 3
a.1 = EI = 5000 = 6 X 10- 4 rad. (clockwise)

1.8 X 10- 3
Hence, X 1-- -:---'-:-=:-;-
4 3 m. t. (clockwise)
6x10

The reactions due to the rotation at a will be as shown in Fig. 2. 17 e.

The reactions due to settlement at support b may be obtained by selec..

ting the redundant X" t as the vertical reaction at b and then find its value
from Lhe relation 8b1 X" 1 = 8bS ' This, however, requires fresh calcula..
tions for 8b1 Another easier method based on the original main system
is used. The necessary condition is: ass + Xl" aal = O.
If b rises by 3 em., this corresponds to an anticlockwise rotation at a
in the main system au = 3/900 = 3.33 X 10- 3 rad .(clockwise).

3.33 X 10- 3
Hence, X" I - - 5.55 m.t. (clockwise)
6x 10 4

The remaining reactions are found as usual by the application of the

equations of statics and are as indicated in Fig. 2.17 f.
216

2.9 Applications to statically indeterminate trusses

The applica.tion of the method of consis.tent deformation::. to tbe aD~lys.is.
of trusses may be best illustrated by a number of numerical examples. The
solution follows exactly the procedure summerized in section 2.7, and
no further comment is necessary except to point out that the method of
virtual work is used for displacement computations, and in so doing it is
advisable to arrange the calcuh~tions in a table form'.
Example 2.7 Calculate the forces in the members of the truss shown in
L
Fig. 2.18 a if all the members are assumed to have the same - -values.
EA
e

4m

3m
(al
X,=11

o .413

413P
(e) F,.D.
(b) F;,D

Fig. 2.18
 217

Solution: There is one member more tha n needed for a stable and stati-
cally determinate truss. Hence, the truss is statically indeterminate to te

r the first degree. (The degree of indeterminancy n may be found either

by inspection or from the relationship, r = m + r - 2 j). A possible
main system may be obtained by cutting the top member at any section
along its length. The redundant Xl is then the force in this member and
t
I
is obtained from the condition: Dio + Xl 3 11 = O. The Fo- forces due
to applied load and the F I-forces due to two opposite unit loads acting along
member cd are shown in Figs. 2.18 band c respectively. The calculations
are presented in Table 2.1

Member Fo F, F, Fo F,2 F
t ac 0
I + 4/3 0 16/9 +0.545 P

l cd
db
0
--4 P/3
+1
+4/3
0
-16 P/9
I
16/9
+0.410
-0.788
P
P

t ab
ad
0
+5 P/3
+1
-5/3
0
- 25 P/9
I
25/9
+0.410
+0.985
P
P

[,
f
bc 0 -5/3 0
--41 P/9
25/9
100/9
-0.685 P

I Table 2.1
,~ Since, as assumed, LjEA is constant for all the members,
and 8" = L/EA LF, 2
810 41 P 9
Hence, X, - - - = -- X - = 0.41 P
8" 9 100
Once the force in the redundant member cd is known) the forces in
the remaining members may be found from the superposition relation-
ship; F=Fo+F, X" and will be as given in the last column of Table 2.1.

Example 2.B : Calculate the forces in the members of the steel truss shown
in Fig. 2.19a. The cross-sectional areas (in em 2) are as indicated .

Solution: The truss is statically indeterminate to the first degree. A

possible main system may be obtained by cutting member cd at any
section along its length. The redundant X h which is the force in the
assumed cut member, is obtained from the condition: 8 10 + Xl 8 11 = O.
The Fo-forces due to the applied loads and the F,-forces due to
a unit tensile load acting along member cd are shown in Figs. 2.19 b
and c respectively. The calculations are presented in Table 2.2.
 81

(b ) Fa.D. ( C)F,.D.

Fig. 2.1 9

Mem- L A Fo F, F,FoL F,2L

ber em. em 2 t. t. F
A A
ab 400 40 - 4 -0.330 + 13.33 + l.ll - 3.9
ae 632 40 - 12.6 -0.526 + 10.5 + 4 .36 - 12.44
be 632 40 +12.6 - 0.526 - 10.5 + 4.36 + 12.76
de 400 20 0 + 1 0 + 20 - 0.305
ad 282 20 0 + 0.7 0 -+- 6.9 - 0.214
bd 282 20 0 + 0 .7 0 + 6.9 - 0.214
+ 13.33 + 43.63

Table 2.2

-----
I
!

t
I
,
13.33 43.63
219

r,
I Hence, XI = -
E

-
abO
abl
and 8" =

= -
13.33
- -
43.63
-
E

0.305 t.
i
The negative sign of Xl indicates that member cd is in compression. The
final forces in the members, given in the last column of the table,
are obtained from the superposition relationship, F= Fo + Fl XI' For
example, the final force in member ab is given by :
Fab = - 4 - 0.33 (- 0.305) = - 3.9 t.

Example 2.9 Find the forces in the members of the two-hinged truss
shown in Fig. 2.20 a. The cross-sectional areas (in em. 2 ) are as indicated.
15 t
50 e 50 a

25 em'
3 75 75

~50 f
50

b
1-.-4m 4 ~

(al
t -10
1St
-10

t -15 o
! .12.5 .12.5 -1.25 -115
~
t -7.5 -7.5 .0.75 .0.75
I
0 0 -1.25 -1.25
,t

rf (bl ~.D. (C)F,.D.

r
Fig. 2.20

f
220

A pos-
Solutio n: The truss is statically indeter minate to the first degree.
by ToileT.
a
sible main system may be obtaine d by replaci ng the hinge at b
at h, and
The redund ant XI is then the horizon tal reactio n compo nent
is found from the conditi on: SbO +
X,S., = O.
i-
The Fo-force and F,-forc es are shown in Figs. 2.20 band c respect
vely. The calcula tions are presen ted in Table 2.3.
Mem- L A Fo Fj F,FoL F,2L
F
ber cm. cm 2 t. t. -;;;:- ---;\
ac 600 50 - 7.5 +0. 75 --67.5 + 6.75 2.3
ce 400 50 -10 + 1 -SO + 8 -2.9
ed 400 50 -10 + 1 -SO + 8 -2.9
db 600 50 - 7.5 + 0.75 --67.5 + 6.75 -2.3
cf 500 75 +12.5 - 1.25 -104 +10,4 +3.65
fb 500 75 0 -1.25 0 +10.4 -S.85
df 500 75 +12.5 - 1.25 - 104 +10.4 +3.65
fa 500 75 0 - 1.25 0 +10.4 -S.85
ef 300 25 -15 0 0 0 -15
-503 +71.1
Table 2.3

Hence ,X t - ~ = 503 = + 7.1 t.

Sb' 71.1
The positive sign indicat es that the horizon tal reactio n compo nent at b
acts, as assumed, from right to left. The forces in the membe rs are (ound
from the relation ship: F = Fo + FIX 1 and are as indicat ed in the last
column of Table 2.3.
shown
ExaD1ple 2.10 Find the forces in membe rs ie, kd and ef of the truss
consta nt for all the truss membe rs.
in Fig. 2.21 a. L{EA is assume d

(a)

241
Fig. 2.21
A pos-
Solutio n: The truss is statical ly indeter minate to the first degree.
suppor t. The
sible main system may be obtained by removing the central
is found
redund ant X I is then the vertica l reactio n at support d J and
from the conditi on: SdO +
X, Sd' = o. The Fo-forces due to the applied
shown in
load and the F I-forces due to a unit downw ard load at d are
 221

Figs. 2.21 band c respectively. The calculations for the displacements

SdO and odl are presented in Table 2.4.

I X,=1t

~
Fig. 2.21

Member Fo FI FIFo F 12

ah -20 - 5/8 + 12.5 0.3904

hi -24 - 3/4 + 18 0.5625
ij --48 -3/2 + 72 2.2500
jk - 28.8 -1.8 + 52 32400
ab + 12 + 3/8 + 45 0.1406
bc +36 + 9/8 + 40.5 1.2656
cd +42 + 15/8 + 79 3.5156
bh +20 +5/8 + 12.5 0.3904
bi -20 - 5/8 + 12.5 0.3904
ic +20 +5/8 + 12.5 0.3g04
cj +10 - 5/8 - 6.25 0.3904
jd - 30 - 5/8 + 18.75 0.3904
kd +32 +2 +64 4.0000
gn -10 - 5/8 + 6.25 0.3904
nm -12 - 3/4 + 9 0.5625
ml - 24 - 3/2 +36 2.2500
Ik - 28.8 - 1.8 + 52 3.2400
gf + 6 +3/8 + 2.25 0.1406
fe + 18 + 9/8 + 20.25 1.2656
ed +30 +)5 /8 + 56.25 3.5156
fn +10 + 5/8 + 6.25 0.3904
fm -10 -5/8 + 6.25 0.3904
me +10 +5/8 + 6.25 0.3904
el -10 - 5/8 + 6.25 0.3904
Id -10 - 5/8 + 6.25 0.3904
605.75 30.63
Table 2.4
222

605.75
XI - 19.77 t. i.e. 19.77t. (upward)
30.63
The forces in the members of the truss are found from the relationship,
F = Fo + XIF I . Thus,
Fie = + 20 - 19.77 X 5/8 + 7.7 t.
Fkd = + 32 - 19.77 X 2 - 7.5 t.
Fer = + 18 - 19.77 X 9/8 - 4.2 t.
Alternative solution: Another possible main system may he obtained by
cutting member kd at any section along its length. The main system thus
consists of two simple trusses as shown in Fig. 2.22. The redundat XI'
is the force in the cut member. The Fo-forces due to the applied load and
the F I-forces due to a unit tensile force in member kd are shown in Figs.
2.22 a and b respectively. The calculations are presented in Table 2.5.
All the members having zero force have been omitted.
k

(b)F,.D
.3/16 _9/16
114 114
Fig. 2.22

Member Fo FI FIFo FI"

ah -10 - 5/16 + 3.125 0.0976

hi -12 -3/8 + 4.50 0.1406
IJ -24 -3/4 + 18.00 0.5625
jk 0 -0.9 0 0.8100
ab +6 +3/16 + 1.125 0.0351
be + 18 +9/16 + 10.125 0.3164
cd + 12 + 15/16 +11.25 0.8789
bh +10 +5fl6 + 3.125 0.0976
bi -10 -5/16 + 3.125 0.0976
Ie +10 +5/16 + 3.125 0.0976
eJ + 20 -5/16 - 6.25 0.0976
jd -20 -5fl6 + 6.25 0.0976
57.50 3.3291
Table 2.5
r.
r

223

1: F,Fo 57.5 and 1: Fl' = 2 X 3.3291

= + 1 7.6582
X', = - 57.5/7.6582 = - 7.5 t.
F;c = + 10 - 7.5 X 5/16 = + 7.7t.
Ftd = X', = - 7.5 t.
Fer = 0 - 7.5 X 9/16 = - 4.2 t.
ond~
It is seen that the forces thus obtaine d are identic al to their corresp
ing values obtain ed in the first solutio n.
ed in
The studen t is advised to attemp t the two solutio ns suggest
This examp le
order to apprec iate the time saved in the second solutio n.
choice for the
is giv~n to empha size the import ance of making a proper
main system.

bg of the
Exatnp le 2." C,lcula te the forces in membe rs fg and
truss shown in Fig. 2.23a. L{EA is assume d constan t.

21 f 9 h
t
(a) 3m
La
>-- 4 b

6I
4 c

61
4t4
21 -17
,,;,,0
"v.. .6

!
/

.13.5 .13.51 .12.5 .12.5

61 61 61

8.625 0 9.375
-0.81
XI =l
-0.6 -0.6 0 0 0
(c) ~.D .1
0 -0.8 0 0

-0.81

(d) F,.D
0 0 -0.6 0

0 0 0.8 0

Fig. 2.23
224

Solution : The truss is statically indeterminate to the second degree. A

main system is obtained by cutting each of members gb and gd at any section
along their lengths is shown in Fig. 2.23 b. The redundants X, and X"
which are the forces in the cut members, are obtained from the simultan-
eous solution of the two superposition equations :

8,. + X, 8 + X 2 8'2 ~ 0
11
82 + X, 82 , + X 2 822 ~ 0
These two equations comprise SIX displacements. From Maxwell's
theorem of reciprocal displacements, 3 12 = 821 and from symmetry
8 11 = 822 , Thus, the number of displacements to be calculated is
reduced to four only; 810' 3 1 b 8 12 and 820 , The calculations for these dis-
placements are presented in Table 2.6 and with reference to Figs. 2.33 b-d.

Mem-
ber
F. F, F2 F,F. ,
F 2 F,F> F 2F.

ab +13.5 0 0 0 0 0 0
be +13.5 --{l.8 0 -10.8 +0.64 0 0
cd +12.5 0 --{l.8 0 0 0 - 10
de +12 .5 0 0 0 0 0 0
af -14.375 0 0 0 0 0 0
fg - 17 --{l.8 0 +13.6 +0.64 0 0
gh -17 0 --{l.8 0 0 0 +13.6
he -15.625 0 0
bf
gc
+6
0
0
--{l.6
-D.6
0
--{l.6
- 3.6
+0.36
+0.36

0
+0.36
0
0
hd +6 0 --{l.6 0 0 0 - 3.6
fc + 4.375 0 + 4.375 0
he
gb
+ 5.625
+1
0 +1
0
0
0
+1
0 + 5.652
0
gd 0
+1
0 +1 0
+1
0 0 . 0
+3.575 +4.00 +0.36 +5.625
Table 2.6

8,. EA/LL F,F. ~

3.575 8" EA/L ~ L F,2 ~ 4.00
8'2 EA/L ~ L F ,F2 ~ 0.36 82 EA/L ~ L F 2 F. ~ 5.625

Substituting these values in the superposition equations and solving

simultaneously, Xl = F bg = - 0.77 t.
X 2 ~ F d ~ - 1.34t.
The force in member fg may be found from the relationship:
F ~ F. + X, F, + X 2 F2
Hence, F" ~ - 17 - 0.77 (-0.8) - 1.34 (0) ~ - 16.38 t.

\
 225

Esam.ple 2.12 Calcul ate the reactio ns of the continu ous truss shown
in
Fig. 2.24a if L/EA is assume d constan t.

~m ~1'CI~---r-7r<:I
:--r-~'-/
SVN[\;:~'-V-;>r~
9~ ~ -'-
Ii >---4. 4=16" , 1,6m
(a)
I 16m--l

)
~:
~~ZG
A
I (b)Main system No.1
Zero membe
, rs

15 o
(c) Main system No.2 (fo.D.)

(d) F,. O.

(e) '2-0.
Fig. 2.24

Solutio n: Two possibl e main systems are shown in Figs. 2.24

band c.
The first is obtain ed by removi ng the suppor ts at band c, and
the main
system is then the simple truss shown in Fig. 2.24 b with the
redund ants
as the vertica l reactio ns at band c. The second main system
is obtaine d
by cutting the two membe rs indicat ed, and the main system thus
consist s
of three simple trusses with the redund ants X'I and X' 2 as
the forces in
the cut membe rs. The solutio n given here is based on the second
main
system , and the studen t is recomm ended to re-solve the problem
using
226

the other suggested main system and compare the relative ease of the two
solutions.
The values of the redundants X', and X' 2 are obtained from the
simultaneous solution of the two superposition equations :

010 + X'I 0" + X'20" = 0

820 + X',o" + X'2022 = 0

Since L/EA values are the same for all the members, these equations
may be expressed as :

1: FIFo + X', L F2, + X'2 L F,F 2 = 0

1: F2FO + X', L F 2F, + X'2 L F22 = 0
Thus, with reference to Fig. 2.24,
L F,Fo = -2 (20 X 1/4) -2 (20x 3/4) - 2 (40 x 1/2) - 2 (25 X 5/16)
+
2 (25 X 5/16) = - 80

= 8 (5/16)2+4 (3/4)2+4 (1/4)2 + 4(1/2) 2+ 2 x 12= + 6.28

= 2 (1/2 X 1/2) + 4 (3/4 X 1/4) - 4 (5/16 x 5/16)= + 0.86

L F,F 2 = + 0.86
L F2, = + 6.28 from symmetry.

Substituting these values into the superposition equations and solving

them simultaneouslY:1
X', = + 12.7t. and X'2 = - 1.74 t.

The final reaction at any support may be expressed as :

R = Ro + X,'R, + X 2'R 2. Thus,

R. = 15 + 12.7 (- 3/16) - 1.74 (0) = 12.62 t. (upward)
Rb = 15 + 12.7 (3/8) - 1.74 (- 3/16) = 20.08 t. (upward)
Rc = 0 + 12.7 (3/16) - 1.74 (3/8) = 3.02 t. (downward)
R d = 0 + 12 .7 (0) - 1.74 (- 3/16) = 0.32 t. (uFward)
 227

Exatnple 2.13 Find the reactions and the forces in members ab, bf, cf and
cd of the truss shown in Fig. 2 .25a. LjEA is assumed constant.

2t 21
d
I
..., <D

Sm m
-
'"
-=:2'->lc'l-_I' 2 0

(a ) (b) Fa.D.

o 0 o 0

o
o 0
(c) F,.D (d) F,.D.
Fig. 2.25

r Solution ; The truss is statically indeterminate to the second degree. A

possible main system is obtained by replacmg the hinge at g by a roller
228

and cutting member eb at any section along its length. The redundants
Xl and X 2 are thus the force in member be and the horizontal reaction
at g. These can be obtained from the two superposition equations:
a lO + X, all + X2 a 12 = 0
agO + X, agI + X2 a g = 0

Since L/EA values are the same for all the members, these equations
may be expressed as :
k F,Fo + XI k F2, + X, k F,F, = 0
k F,Fo + XI k F2FI + X. k F2. = 0
With reference to Figs. 2.25 b - d,
k F ,Fo = (-0.25) (-3 .75) +
(-0.95)(-0.93) +(-0.95)(13.8) +(1) (-4.75)
=
0.94 + 8.8 - 13.1 - 4.75 = - 8.1 I
k F', = 0.25 2 + 0.95' X 2 + I' X 2 + 0.45'
= 0.06 + 1.8 + 2 + 0.2 = + 4.06

k FIF. = (.......{).25) (I) = - 0.25

k F.Fo = (I) (6.5) + 1 (-3.75) + I (-13) + (-1.75) (2.25) +
(- 1.75) (9) + (.......{).9) (9.5) + (.......{).9) (-3.75)
= 6.5 - 3.75-13-3.94- 15.75 - 8 .55+3.37 = - 35.12
k F'2 = 3 X 1" +
2 (1.75') 2 (0.9)'+
= 3 + 6.12 + 1.62 = 10.74
The superposition equations are thus as fallows:
- 8.11 + 4.06 XI - 0.25 X, = 0
- 35.12 - 0.25 XI + 10.74 X 2 = 0

Solving simultaneously, X, = + 2.2 t. and X. = + 3.32 t.

The final reactions and member forces are found from the expression,
F = Fo + F,X, + F 2X 2 and are follows:
X. = 6 + 0 (2.2) - I (3.32) = 2.68 t. (to the left)
Ya 6.25 t. (downward)
Y. 6.25 t. (upward)
X. - X. = 3.32 t. (to the left)
Fab - + 6.5 + 0 (2.2) + 1 (3.32) = + 9.82 t.

F br =- 3.75 - 0.25 (2.2) + 1 (3.32) = - 0.98 t.

Fer = - 4.75 + 1 (2.2) + 0 (3.32) = - 2.55 t.

Fed = + 9.3 + 0 (2.2) + 0 (3.32) = 9.3 t.

 229

2.10 Effect of fabrication errors on statically indetenninate trusses

If a member or a number of members in a statically determinate truss
are fabricated longer or shorter than their geometrical lengths, then no
initial forces arc developed in the truss members when they are assembled
and joined together. The truss will only assume a slightly different shape
to that originally intended. On the other hand, if these fabrication errors
take place in a statically indeterminate truss, initial forces are generally
deVeloped in its members in the absence of load.
c

X,=1t

a'k-----~

(a) ( b) (c)
Fig. 2.26

Consider, for example, the truss shown in Fig. 2.26 a and suppose that
member ad is fabricated too long by an amount a. A possible main system
may be obtained by assuming that member ad is cut. In the absence of
load there will be a lap = a at the cut section as shown in Fig. 2.26 b. The
redundant XI indicated in Fig. 2.26c) which is the force in the cut member)
is found from the condition: Olf + Xl all = 0
where 8 tl has the same meaning as before) and 8lf is the relative displa-
cement at the ends of the cut member due to fabrication error. For the
particular case considered) and for the main system chosen) 8 1r = + a.
In general 8lf is calculated from equation lAO; 8lf = :E aFt. Fora truss
statically indeterminate to the n dgree) the superposition equations become:
Xl all + X, aI, + Xn Oln + Olf = 0
Xl 0'1 + X, 022 + Xn o'n + 02f = 0
2.11

Xl anI + X, an' + .......... Xn ann + 0nf = 0

where (ilf' 3 2r.> 3nf are the displacements due to fabrica-
tion errors in the chosen main system at the points where the redundants
Xv X 2 , Xn act and in their direction and sense.
 230

L II Effect of teDlperature changes on statically indetenninate trnss.,,,

If a truss is subjected to temperature changes, its members elongate
or shorten according to whether the temperature rises or drops. These
changes in length when free to occur. . as is always the case in statically
determinate trusses, cause a slight change in the geometrical shape of the
truss and no forces are induced in its members. On the other hand if the
truss is statically indeterminate, forces are generally developed in its
members in the absence of load.

Fig. 2.27

Consider, for example, the two-hinged truss shown in Fig. 2.27a and
suppose that the top chord members are subjected to a temperature
change. A possible main system may be obtained by replacing the hinge
support at b by a roller. Due to temperature change point b will move
a displacement 3It and the redundant XI which is the horizontal reaction
component at b is found from the condition: 31~ + Xl 811 = 0
where 3 11 has the same meaning as before and SIt is the horizontal
displacement at b due to temperature change.
Displacements due to temperature changes are easily found from
equation 1.39; 81t = L Fl at tL
For a truss which is statically indeterminate to the n degree.. the
snpersitation equations become:

x, 3" + X 2 3'2 + Xn 3'0 + 3" o

X, 32 , + X 2 322 + Xo 320 + 3" o
2.12
. ... . .. .. .. .......... ..... .. ......... .. .. .... ...
X , 3n , + X 2 3n2 + ....... .. . X. 3nn + 3nt = 0
 231

where Sll' 82P 8nt are the displacements due to temperature

changes in the chosen main system at the points where the redundants
XI J X 2, .... . ... . . Xn act and in their direction and sense.

2.12 Effect of support 1D0veIDents on statically indetenninate trusses

The effect of settlement of supports on statically indeterminate trusses
can be evaluated in a manner similar to that presented in sections 2.10
and 2.11.
For a truss which is statically indeterminate to the n degree, the
superposition equations become:
X, + X 2 a12 + .. .. . . . .
all Xn a'n = ai'
X, a21 + X 2 a22 + ... . . . Xn a2n = a,.
2.13

where 81s ,8 2s ' . . . . . . . . . 8ns are the forced displacements due to

support movements in the chosen main system at the points where the
rerlundants Xb X 2 , . X D act' and in their sense and direction~

In order to explain the meaning of the forced displacements 8iS con..

sider the truss shown in Fig. 2.28a and assume that supports b,c and d
settle by 4, 1 and 3 em. respectively.

(a)

2
x, (b) X2
Fig. 2.28

A possible main system is obtained by removing supports band c and

the redundants XI and Xl will be the reactions at these supports. These
 232

are indicated in their assumed positive direction in Fig, 2.28 h. The cor-
responding forced displacements Sis will thus be:

Ots = - 3 em. and 821 = + I em. not - 4 em; and - 1 em. as one
might be lead to think.

2.13 Further examples on tru.sses involving no':"load effects

Exatnple 2.14 Find the forces in members ac, cd and be of the ,truss shown
in Fig. 2.29a if members cd and be were fabricated short and member ad
long by the amounts indicated. E = 2000 tlcm 2 and A = 20 cm 2 for all
the members.

Ci<---=-~O",.5~c~m~----jd
f '0
~ .
I
4

.4/3 .4/3

.,
h...--3m
(al (b)~.D.

Fig. 2.29

Solution: The truss is statically indeterminate to the first degree. A possible

main system may be obtained by cutting member cd at any Jsection along
its length and the redundant XI, which is the force in this member, is
obtained from the condition: Sa +
Xl 8 11 = O. The F1-forces are
shown in Fig. 2.29 b. Referring to the figure,

S(r = ~ aFt = (- 0.5 (I) + 0.6 (- 5/3) - 0.3 ( - 5/3) ) = - 1 cm.

F2(L 2 ( 25 16 4800
~ EA =EA 1 X300+ g X 500+- X 400 ) = -
9 EA
4800
0.12 em.
2000x20
 233

-I
XI = - - - = 8.33 t.
0.12.

F,d = XI = 8.33 t. +
F" 8.33 (4/3) = 11.11 t.
F., = 8.33 (-5/3) = - 13.89 t.

ExalDple 2.15 Calculate the force in member ce ofthe truss shown in Fig.
2.S0a due to a rise in temperature in the top chord members of 30C. For
all the members, L/A = 8 cm- 1 and E = 2000t/cm 2. a, = I X 1Q-'/"C.

d'
I
1.5
I
1.5
c'
d
e'

~ar'----7c~--------------~~~------~~
1.5
UJ
~I----__ 4m 4
(al

",~r----- <. 27
'l... 'l..'"
(~. -1.125' "
-I 1 "1-:::;
c - - t e'
.0.75 'k"
",.
d -I.//.
:l
.0.375
c",-_... X, =1 t _-~e

(bl F,.D

Fig. 2.30

Solution: The truss is statically indeterminate to the first degree.

A possible main system is obtained by cutting member ce at any section
along its length. The redundant X I1 which is the force in the cut member,
is obtained from the condition: Sit + XI 8 11 = O. The forces due to
XI = 1 have been calculated and are indicated in Fig. 2.30 b (zero
members having been omitted) . Referring to this figure,
8., = L FI at t L = I X 1Q-'x30 (1.25x250+1.I4x427) = 0.24cm.
234

F2 L 8
8" = :-1- = - - (0.752+ 1.252 X2 + 1.142 X 2+0.3752+12X3+1.125!)
EA 2000
= 0.0428 cm.

0.24
XI = Fe = - - - - = -
c 0.0428
5.6 t.

2.14 Applications to statically indeterminat e f r ames

The method of consistent deformations is used for the analysis of many
types of statically indeterminate frames. However.. it is seldom used for
analysing frames of relatively high degree of statical indeterminancy since
in such cases the slope-deflection and moment distribution methods (to
be considered later) usually provide easier solutions.
The application of the method to the analysis of statically indeterminate
frames will be demonstrated by several numerical examples.
Before presenting these examples, however, it will be advantageous
to emphasize the ideas and the procedure of analysis by considering the
simple two-hinged frame of constant EI shown in Fig. 2.3Ia.

w h h

r
h
+

>-- L ---'
Cal (blMO.D

Fig. 2.31

The frame is statically indeterminate to the first degree . It may be

reduced to a stable and statically determinate main system by replacing
the hinge at I by a roller. The horizontal reaction at I thus becomes the
redundant XI.
The Mo.D. of the chosen main system under the applied load is shown
in Fig. 2.31 b. As usual, it is plotted on the tension side.
 235

Under the applied load point I will move ho rizontally a distance 8 10 .

The correct value ofX t is the value necessary to bring point I back to its
original position. Under the action of Xv point I in the main system will
displace horizontally in the direction assumed for Xl some amount Xl 811~
where 8, t has the same meaning as before. The M}.D . due to the case
"f loading XI = I is shown in Fig. 2.3Ic. Again it should be noticed that
it is drawn on the tension side. The deflection condition in the given frame
that the horizontal displacement atthe hinge support is zero may be
stated as :

010 + XI 0" = 0,

The displacements 010 and 0" are best calculated by the method

S
of virtual work expressed as 8 10 =
f MIModl and
EI
S11 =

has the advantage that the complex nature of the deflected shape
M2d]
_ 1_
EI
. This

of the frame does not need to be considered at all. Also, confusion in

using the proper sign of displacements is avoided if the signs resulting
f
from the graphical evaluation of of integrals MIModl and M 2 dl f 1
are adopted; both Mo and Ml being drawn on the tension side.

Referring to Figs. 2.31 band c,

wL 3 h
8
10 = SMIModl = _ _ I (2L X wL2 X h) =
EI EI 3 8 12 EI

X -
h L
+-, X
2h3 3Lh2+
3h 2 = -'--:':::-:--
3 3 ) 3EI

WL3
, XI
4(2h2 + 3Lh)

Once the redundant is determined, the analysis may be completed

by applying the equations of statics or, easier still, by applying the
principle of superposition. For example, the final B.M.D. may b e obtained
by superimposing the Mo.D. to the MI.D. times XI '
 236

E>aunple 2.16 Draw the B.M.D. for the two-hinged frame shown in Fig.
2.32 a if EI is assumed constant.

211m
I! I I I I I I I

I
4m
tB
La b
6 ...h--12m
(a)

15.45ml

4 .712
IB t 18
(c) M,.o (d) 8 .M.0.

Fig. 2.32

Solution: The frame is statically indeterminate to the first degree. It

may be reduced to a stable and statically determinate main system by
replacing the hinge at b by a roller. The chosen redundant XI is then
the horizontal reaction component at b and is found from the condition:
8bO + X, 8b , = 0

Using the method of virtual work and with reference to Figs. 2.32 b
and c which show the Mo.D. and the M,.D. respectively,

8bo =
j M,Modl
EI
= - - 1 [(72
EI
X 7.2)
2
[10.67 + -2x 7.2 X 9 x 10
3

+( 72 X 12.64)
2
2
3
] 11083.2
1O.67+ - xI2.64x36xlO =-" - -
EI

r- -----------------------------------------------------
 2~7

=f M~dl = _ I (2 X - 8 X8 2 + -7.2 (8 2 +12 2 +8xI2)

Sbl
J EI EI 3 3
12.64 ) 2351.8
+ - - (12 2 + 8 2 + 12 X 8) = --
3 EI
11083.2
""""'=..,,-
2351.8
= 4.712 t. (to the left)

The final B.M.D. is best found by superimposing the Mo.D. to the

MID. times 4.712. This diagram is shown in Fig. 2.32 d.

Alternative solution: A possible main system may be obtained by

inserting a hinge at c. The main system is thus a three-hinged fI:ame
and the redundant Xl' is the connecting moment at c.

Cal ~.D. Cb) '-irD.

Fig. 2.33

The redundant is found from the condition: a c O + Xl' Ucl = O.

Using the method of virtual work and with reference to Figs. 2.33 a and
b which show the Mo.D. and MI.D,.,

aco =JM'Mo dl = _1_ [ 2 ( 48x & X.!.. X ~ ) + 48 X 7.2 X ~

EI EI 2 3 3 2 9
2 5 48 X 12.64 7 2
- - x 7.2x9x - + X --- X21.64 X36 X-5 ]
3 6 2 9 3 6
252.2
EI

ad
= S M2,dl
EI
I [8 4 7.2 ( 4 +1+ - 2 )
=EI 3 X 9 X 2 + - 3-9 3
12.64 16.32
+ (
1+4 + 2 ) ]=
3 3 3 EI
252.2
X'I = - 15.45 m. t.
16.32
238

The negative sign indicates that the moment at c acts i n a direction

opposite to that assumed for XI t" Both the value and direction of the
moment at c thus obtained ar e consistent with those resulting from the
first solution.
In all the following examples, only one solution is presen ted. The
student is recommended to consider alternative choices of the redunrlants
in these examples and try to ass~ss the relative simplicity of the numerical
calculations involved in each case.

Exa.nple 2.17 Calculate the r eactions and draw the B.M.D. for the frame
shown in Fig. 2.34 a. EI is assumed constan t.
28
211m
21
c +
0
16ml
6m

L a
6

L
L 8m
(a)
b

d
6 8

1.06
4.235

(c) M,.O. (d) F inal B.M.O

Fig. 2.34

Solution: The frame is statically indetermina te to the first degree. Choos-

ing the redundant Xl as the horizontal r eaction component at a, the
Mo.D. and the M,.D. will be as shown in Figs. 2.34 band c respectively.
 239

Referring to these figures and using the method of virtual work,

M.Mo dl I [ 2 28 X 8
8,0 =
S +
EI
12 x6
= - --x8xI6x7 + - - x 7.33
EI
2
3

x-x6+20x2x7+
4~2
2
2
6+-x
2 3 2 3
676.95
EI

=JM.2 dl = _I
EI EI 3
(~x
62 +
3
~
(6'+8 2 -+6x8)
3
+~x
8')

637.33
EI
0 676.96
x. =- - -
637.33
= - 1.06 t., i.e. 1.06 t. (to the left)
0
The remaining reactions are found from the equations of equilibrium
and the final B.M.D. is drawn in the usual manner. The result is shown
in Fig. 2.34 d.

Exunple 2.18 Draw the B.M.D. for the frame shown in Fig. 2.35a. EI
is assumed constant.

11
,--', 2 _

~"paDo~2jtj/miIII~__2j-i~~=~O~r-__6~ ~14~~
3

L6m_~b2f
(a)

,r-x,=lmt
l,r + 3mt

9t
(c)M,.D. (d)FinaI8.M.D.
Fig. 2.35
240

Solution : The frame is statically indeterminate to the first degree. Choos-

ing the redundant X"as the fixing-end moment at support a r the Mo.D.
and the M,.D. will be as shown in Figs. 2.35 band c respectively.
Referring to these figures and using the method of virtual work,

~aO =s M,Mo dl
EI
1(2 x6x9x - -
EI 3
1
2
6 X 6 x~) = 12
2 3 EI

=SM,2 EI
dl = _I
EI
(~
3
X

x, naO 12 . ..
= - - =- - = - 6 m.t., '.e. 6 m.t. (antIclockWlse)
aal 2
The final B.M.D. is found by superimposing the Mo.D. to the M,.D.
times (--6) . The resulting diagram is shown in Fig. 2.35 d.

Exa.nple 2.19 Draw the B.M.D. for the frame shown in Fig. 2.36 a. EI
is assumed constant.

I
6m
h
6_
c:
L a b l C
\...... 8 -4-8m---L 8
/\, (a)
,<:;)~
~%

(el M,.D. (d) Final B.M.U

t
Fig. 2.36

Solution: The frame 15 statically indeterminate to the first degree. Choos-

ing the force in link member cc' as the redundant, the Mo.D. and the
M . D. will be as shown in Figs. 2.36 band c respectively. Referring to
these figures and neglecting the axial deformation in member cc",
r
241

s = { MIMo dl .!. (_ 13.5x6 X:! x 2 - ~ x 6x4.5 xl

'" J EI El 2 3 3
13.5 X \0 2 2 4.5 X \0 2
X - X - x-
2 3 2 3

+ 4.52X 6 2
x-
3
X
2 )
= -234
-
EI

I (6 6
=f = -
EI
-
3
x2 2 +\0
'1
\0X
-x22 +-
3
2

+ -63 X 22 + -38 X 8
2)320
=-
EI

XI =-~=~ = 0.73t.
ad 320
The final B.M.D. is best found from the first principles and will be as
shown in Fig. 2.36 d.

Esa.nple 2.20 Draw the B.M.D. for the frame shown in Fig. 2.37a. EI
is assumed constant.

Him 111m
I I I I 1 I r I I I I I I I I I I " r I II I II I~=O
I I I
r
c
2
r
6",
lab a
I-~J.- 6 --1--6 In -l- 3--1 "" "
(a) (b)
=0

+
4O.5mt 10mt 16.5mt
(c) M,,-D. (d)M,-D. (e) Final B.M.D.
Fig. 2.37
242

Solution: In general, this frame is statically indeterminate to the second

degree as it has six reaction components and there are only four available
equations; three equations of equilibrium and one condition equation
provided by the intermediate hinge at c. A possible main system may be
obtained by disconnecting the frame at the itttermediate hinge and the
redundants will then be the horizontal and vertical reaction components
at c . Due to the complete symmetry in frame and loading, however, there
will be no relative vertical displacement at c and hence the vertical reac-
tion component there is zero. Thus, one redundant only remains to be
determined . Furthermore, as the frame is symmetrical and symmetrically
loaded, it is sufficient to consider half the frame only, and the chosen main
system will be as shown in Fig. 2.37 b. The Mo.D. and the M, .D. are
as shown in Figs.2.37 c and d respectively.

Referring to these figures and using the method of virtual work,

M,Mo dl
'0 =f EI
-1
EI
(1
- X 6.3 X 18 X 3- X 2
3
3 40.5 + 18
4
2
+ 18 X 2 X + X 3 X 4 --
2 3
2207
X 3 X 1.12 X 4 + 40.5 X 6 X 7 = - __
) EI

(~ 2
=J _ _ _
EI 3
X 22 +
3
(22+ 4 2 +2 x4)

+.2.. (3 X 4 2) + ~ (4 2 + \0 2 + 4 X 10))
3 3
387
EI

Substituting these displacements in the deflection equation 8co+Xt8ct =0,

5.7t.

The final B.M.D. is best found from the superposition equation,

M = Mo + X,M, and the result is shown in Fig. 2.37 e.
l
r 243

Exam.ple 2.21 Find the reactions and draw the B.M.D. for the frame
shown in Fig. 2.38a. EI is assumed constant.

I
2

111m
c
i 4m

a b c' J
'----- 6m 6 ~
(a)

16mt
+
+
4.5

, (b) Mo.D

r--- 6.32Srn
x2=ot
7.02
X=11

+ +
6mt
-;.
":J
+

1.76 0.61

(d) M .O. (e) Final 8.M.D.

2
Fig. 2.38

Solution: The frame is statically indeterminate to the second degree.

A possible main system may be obtained by replacing the hinge at b by
a roller and removing the force in the link ee'. The redundants Xl and
X 2 are obtained from the simultaneous solution of the two superposition
equations:
8. 0 + XI 8. 1 + X 2 8. 2 = 0
8eo + X I lid + X 2 8e2 = 0
244

Figures 2.38 b-d show the Mo.D., the M,.D. and the M 2.D. Referring
to these figures and using the method of virtual work,

_1_( 18 X 6 2 2
x- X 6 + - x6 X 4.5x3
EI 2 3 3
18 X 6 594
+ 2 EI

3b,=jM:dl
EI
= _ 1_ (2 X
EI
~X6'+...!.
3 3
X 3 X 6 2)
360
EI

3 = j M,M2 dl = __1_ (6 X 6 X 6) = _ 108

b2
EI EI 2 EI

3
00-
-J M2
M
EI
O dl _1_ ( 18 X 6
EI 2
X 2) = IE08
1

108
- -EI- J from Maxwell's theorem

3 =J M~ dJ = _I (~X 6' + 6.325 X 6') = 148

d ru ru 3 3 EI

Substituting these values in the two superposition equations, and solving

simultaneously,
X, = 1.83 t. and X 2 = 0.61 t.
The remaining reactions are best found from the application of the
superposition equation :

R = Ro + X,R, + X 2R 2. Thus,
X. 6 - 1.83 X 1 4.17 1. (to the left)
Y. 3 - 1 X 0.61 2.39 t. (downward)
Xb 0+1 X 1.83 - 1.83 t. (to the left)
Yb 3 - 2 X 0.61 - 1.78 t. (upward)
Yo X 2 = 0.61 1. (upward)

The final B.M.D. may now be determined from the first principles and
will be as shown in Fig. 2.38 e.
,0
245

E"aDlple 20n Draw the B.M.D. for the frame shown in Fig. 2.39a. EI
is assumed constant.

2t1m
I I! ! ' , ' I I ! II I I I

6m
a
I, '" 16m
___ b~"4 j--l
I (a)

! 48

~ 64mt
48

(C)M,.D.

41.4 mt
6
I 25.4

b25.6 30.6
(d) M,.D (e) Final B.M.D.

Fig. 2.39

Solution: The frame is statically indeterminate to the second degree. A

possible main system may be obtained by disconnecting the frame at the
intermediate hinge c. The main system thus consists of two cantilevers
and the redunclants XI and X 2 are the horizontal and vertical reaction
components at c. These may be obtained from the simultaneous solution
of the two superposition equations:
.8,0 +X I 8" + X 2 8" = 0
Y,o +X, yot + X, y" = 0
where 8 and y refer to the relative horizontal and vertical displacementi
of the two cantilevers at c.
 246

Figs. 2.39 b-d show the Mo.D., the MI.D. and the M 2.D. respectively.
Referring to these figures and using the method of virtual workJ

- _1_
EI
[2 (_1_3 X 8~54 X 64 X
+ 6 X 64 X 6 + 6 X 48 X 6]
4851.92
EI

=j Midi
EI
=-2(6-
EI 3
(3 2 + 92 + 8.54
3 X 9) + - - X
3
519.24
EI

= S MIM2 dl
EI
_ 0

M,Mo dl I ( I
YeO
S--"---"-- =

3
EI
-
EI
-64 X 6 X 8 - -
3
X 8.54 X 64

X '4 X 8 + 3I X 8.54 X 64 X 4"3 X 8 + 48 X 6 X8

)

768
EI

Yel
J M,MI dl
EI
o

Ye'
J Mi dl

1132.24
EI EI
- x3x8 2 + -
-2 (6
3
8-.X
3
58
42)

EI

Substituting these values in the two superposition equations and solving

them simultaneously, Xl = 9.35 t. and X 2 = 0.68 t.

The final B.M.D. may now be drawn and will be as shown in

Fig. 2.39 e.
 247

Exunple 2.23 Find the reactions of the frame shown in Fig. 2.40a if the
moment of inertia varies as indicated .

r
6m _ _. .:.6:...4ml

L
2

F
4
21

c
LL8m .1 ~~2J 34 ml

8ml 4-(2
(c) M,.D. (d, M,.D.
Fig. 2.40

Solution: The frame is statically indeterminate to the second degree.

A possible main system may be obtained by removing the two roller
supports at a and h. The redundants Xl and X 2 are found from the
simultaneous solution of the two superposition equations :

SaO + Xl Sal + X 2 Sa' = 0

SbO + Xl Sbl + X, Sb' = 0
, Figs. 2.40 h - d show the Mo.D., the MI .D. and the M,.D. respectively.
Referring to these figures and considering I = In
l.
!
248

= S MIMO dl
EI
_1_ ( __1_ x 10 X 64 X _3_ X 8 x _l-
EI, 3 4 5
- 64 x 2 x 8 X _I - 34 x 4 x 8 X _I ) = _ 1312
2 2 EI,
= S_M~;d;-I_
EI
_EI,1-(_1-2 -X 10 x 8 x _2_ x 8 x
3
.!...
5
+~ x 3 X 8 2 X _I ) = 234.66
3 2 E~

= _ _ 1 [(2J"2+4J'Z)4X8X _I ]
EI, 2 2
67.9
EI, EI,
M,Mo dl I [ 2 X 30 ( ' - 2 J- I
=s EI =

+ (2 J2
EI,

+ 4
- 2 V 2 +3
J"2) 4 X 34 X 2-] =
2)

264.1
3

2 2 EI,

_ 67.9 ~ from Maxwell's theorem

EI,
M~dl I (4-3 X 8 X -3I+ -43 (8+32 + 16) -2I)
s"
S
40.9
EI EI,

- --
EI,
Substituting these values in the superposItIon equations and solving
them simultaneously, XI = 7.2 t. (upward) and X, = 5.5. t. (u pward)
The remaining reaction components are found from statics and are as
follows:
Xc 3.9 t. (to the right)
Yc - 19.9 t. (upward)
Me = 7.6 m.t. (clockwise)
Particular attention should be paid to the manner in which the vary-
ing moment of inertia is handled. One value of I, not necessarily but
preferably one of the given values, is chosen as a reference value It and
all the other values of I are expressed in terms of it. This is done by divid-
ing each term in the expressions for the displacements by its corresponding
value of I and multiplying it by Ir- If this is done, the actual values of
the various moments of inertia need not be known.
 249

EsaDlple 1.14 Draw the B.M.D. for the frame shown in Fig. 2.41a. EI

r is assumed constant.

6 61

t
l. 3
d

t
I
t
6m

1--4 -1-4
b

~4 -1-4--1
(a)

f (c)M,.O. (d) M2 .D. (e) 8.M.D.

I
Fig. 2.41

Solution: In general. . the given frame is statically indeterminate to the

third degree. Due to symmetry, however, joint d has zero rotation and
zero horizontal displacement. Further, neglecting the axial deformation
of member bd, which is a common procedure,joint d has also zero vertical
displacement. Hence joint d may be considered fixed and then only one
half of the frame needs to be analysed. The reduced frame is hinged at
a and fixed at d, and thus is statically indeterminate to the second degree.
Choosing the horizontal and vertical reaction components fit a as the
redund.nts X, and X 2 , the Mo.D., the M,.D. and the M 2 .D. will be as
shown in Figs. 2.4Ib,c and d respectively. X, and X 2 are found from the
simultaneous solution of the two superposition equations:
250

8.0 + XI 8. 1 + X 2 8. 2 = 0
Y'o + XI Y.I + X 2 Y.2 = 0
Referring to Figs. 2.41 b - d, and using the method of virtual work,

8. 0 = S = -I
EI
(24 X 5 (6
2
+ I ) ) =420
-
EI

Sa' = S_M~rd_l_ = _I (_6 X 6 2 + _5 (6 2 + 9 2 + 6 X 9))

. EI . EI 3 3
642
EI

M,M, dl
EI EI
(5 X6 X 2 + 3-2X-5 2
(-
3
x4)
\
3 X 5 I 4)) = _ 300
+6x 5 x6+ - - (4 + -
2 3 EI
M,Mo dl
Y.o
=J EI
300
- I (5 X 24 (4 + - 2
EI 2 3
X 4)1 ) = -400
I EI

Y., = Sa2
EI
from Maxwell's theorem

Y., Mi dl
=5 213.3
EI
- -I- (5
EI
-- x4 2 + -5 (4 2 + 8 2 + 4
3 3
X 8) )

EI

Substituting these values in the superposition equations and solving

them simultaneously, X, = 0.648 t. (to the right) and X, = 2.79 t.(upward)

The final B.M.D. of one half of the frame is shown in Fig. 2.4Ie.

EXaIIlple 2.25 Draw the B.M.D. for the frame shown in Fig. 2.42a if
the moment of inertia varies as indicated.

Solution: The frame is externally statically determinate but generally

internally statically indeterminate to the third degree. Due to symmetry,
however, the shearing force at a section of the frame lying on the axis of
symmetry is zero. A possible main system is obtained by cutting the lower
member at its centre. The shearing force being zero at this section, the
remaining redundants to be determined are the axial force and bending
 251
I
31 3t
r 211m

r
r 41

~ 6m

21

8m ---=1=2-1
(a)

6 6mt
6 6mt

+
16

L--:)~----.J L-_~
X,=1t
..._ - - - '

(b)MDX,=O (c )M,.D
o
6
1 1
1.71

I I 8.29mt
X,=1mt

~~+
0.43
1 + 1
(d)M,.D. (e)Final B M 0

Fig. 2.42

moment at the cut section. These are denoted by XI and X 2 respectively

and are found from the two superposition equations:

~,o + X, ~II + X 2 ~'2 = 0

a,O + X, a" + X 2 al2 = 0

The Mo.D., the M" D. and the M 2 J;!. are shown in Figs. 2.42 b-d.
Referring to these figures and considering I = Ip then by using the method
of virtual work,
252

810
= J M,M.dl = -
EI
I I
\ (6 x 8 x 6 - -
EI,
2
3
I )
x 8 x 16 x 6)-
4
56
EI,

Mi dl
8 11
=J 216
EI
I ( (-x3x6
=-
EI,
8
3
I + 2 ( 6- x 6 2 )1 )
2) -
4 3

-
EI,

8'2
= S M,M 2 dl
EI
= _1_ ( (6 x 8 X I) _I + 2 (I X 6 X 3) I)
EI, 4
48
- --
EI,

a,. = S M 2M.dl -
EI EI,
(6 X8X 1)1-( ~ x 8 x 16x 1) +)
-
9.33
EI,

M 2M, dl 48
all
=s EI
8'2 -
EI,

a'2 = S Ml dl
EI
- I ( (8 X I x I ) - I + 2 (6 X I X I) I
EI, 2
I ) 18
+ (8 X I X 1)- = -
4 EI,

Substituting these values in the two superposition equations and solving

them simultaneously, X, = 0.356 t. and X, = - 0.43 m.t.

Having thus determined the straining actions at the cut section, the
B.M.D . for the entire frame may be obtained from the first principles
and will be as shown in Fig. 2.42 e.
 253

Esaoaple 1.26 Draw the B.M.D. for the frame shown in Fig. 2.43a. EI
is assumed constant.

III I I I I I I I I

.--
~d
6m
L a
b
L 7 I 7 -+-7m - .4-
'- 7---.J
(a)
12.3mt X =0

.- ~. 1~?
r=-==--~

--'-_....19
Lc) M.O
I

29.74mt

7
+
mt
X.=1t

7
(d)M O. , (e)Rnal B.M.D.

Fig. 2.43

Solution : In general, the frame is statically indeterminate to the fourth

dcgrcc. Due to symmetry, however, only one half of the frame needs to
be analysed. The redundant. chosen may be the horizontal reaction com-
ponent at the intermediate hinge X, (the vertical component being zero
due to complete symmetry in both frame and loading), and the force in
the link member X 2 These are found from the two superposition equations:
254

OcO + XI 0cl + X 2 OC2 = 0

OdO + XI Odl + X 2 0d2 = 0

The Mo.D., the MI .D. and the M 2.D. are shown in Figs. 2.43 bod.
Referring to these figures and using the method of virtual work. .

OcO
=JMIModl = - - I-
EI EI
C
- X 7.6 X 49 X-3
3 4
X 3)

279
- - E1

M; dl I (6- (9 2 +3 2 +3x9) + - 7.6 x3' )

Cl = J =--
EI EI 3 3
257
EI

Oc2
= SMI M 2 dl
EI = ~I [C; 252
9) 6 X 7] = - -E1
-

= M2 Mo dl = _ - I ((-x7X49x-x7)
I 3) 600
EI EI 3 4 EI
J
252
= 0 2= - -- from Maxwell's theorem.
c E1
M~dl + -63 X 3 72
) 408
0d2=
S ~-
EI
=--
EI

Substituting these values in the two superpOsItion equations and

solving them simultaneously, XI = 6.42 t. (to the left)
and X 2 = 5.7 t. (upward)

Example 2.27 Draw the B.M.D. for the frame shown in Fig. 2.44 a. The
moment of inertia varies as indicated.

Solution: The frame is statically indeterminate to the third degree. A

possible main system may be obtained by cutting member cd at its mid
point e. The chosen main system thus consists of two cantilevers and the
redundants XlJ X 2 and X3 are the normal force, shearing force and bend
ing moment at the cut section. These redundants may be obtained from
 255

the continuity of the elast ic curve; the relative horizontal and vertical
displacements and rotation being zero at the cut secti on. T hese cond itions
are expressed as :

8", + X , 8" + X , 8" + X, 8" 0

Y,o + X , y" + X, y" + X, y" = 0

4t X3=
d
r
C
21 X,=O
6m ! +

12mt
12 m
(a) (b) MoD.

6 t 6m t
+
X,=1t 6
~x,=1t
+
6 6
6 6mt
(e) M,.D. (d) M,.D.
r; X3=lmt
! lm t
+ 1

+
1.14

(e ) M,.D. (tl Fina l B.M.D

Fig. 2.44
The superposition equations comprise twelve displacement components
but using Maxwell's theorem of reciprocal deflections, it is obvious that
8e2 = Yeb 3e3 = acl and Yel = ac2' The MoD. , the M1.D., the M 2 D.,
and the M3.D. are shown in Figs. 2.44 b-e. Referring to these figures, and
using the method of virtual work ,
256

8eo = J M, M., dl
EI
= _ _1_ (6 x 6 x 12)=
EI 2
216
EI

8 e, = J _M
-;E",,~;d-l_ = :1 (: x 6 2
) = :

8e2 = J M, M 2 dl = 0
EI

8e, = J M,M, dl
EI
_~( 6 x 6 x
EI 2
I) =
36
EI

- J M2Mo dl = - - I (12 432

YcO - EI x 6 X 6)
EI EI

'Ycl = 8t2 = 0

l'
e2
= JMi
EI
dl
= _ 2_
EI
(~
3
x 6 2 X _1- +~ X
2 3
3 X 6 2)= 504
EI

_JM2 M,dl = 0
Yc> - EI

"cO ~ J M, Mo dl
EI
= -
I
EI
(12 X 6 X I)
72
EI

36
e,-
8 - - -EI

a e2 = 'Yc3 = 0

2 (l X 6 X I) + (I X 12 X I) 2 I) 18
= EI

Substituting these displacements in the superposition equations,

- 216 + 144X, - 36 X, = 0
- 432 + 504 X 2 = 0
72 - 36 X, + 18 X, = 0
Sol';ng simnItaneously>

X: 1 = + I t. , Xl = + -67 t. and X 1 = - 2 m.t.

 257

The final B.M.D. may now be drawn from the first principles and
will be as shown in Fig. 2.44 f.

It should be pointed out that this frame and similar problems can be
solved by other simpler methods which will be dealt with later on in this
book.

2.15 Efl'ea of teJl1perature changes on statieaUy indetenninate fr.unes

If a frame is subjected to temperature changes~ its members undergo
displacements. These displacements when free to occur, produce no stres-
ses in the frame. On the other hand, when these displacements are restra-
ined in any way, as is generally the case in statically indeterminate frames,
forces and moments are developed in the various members.

Consider for example the two-hinged frame shown in Fig. 2.45. If it

is subjected to a change in temperature, the span length will change
if support I were on a roller, an amount 8" which may be ealculated hy
the application of equation 1.54. In order that the frame. may maintain

t its original span length, a horizontal force X, should he applied at the

assumed free support.

X,
L

Fig. 2.45

The value of X, is determined from the condition that it should be

able to displace point 1 in the frame horizontally an amount 8 tt Thus,
o

8 + X, 811 = 0
"
where 8. t has the same meaning as before, and Sit is the horizontal dis-
placement at point 1 due to temperature change.
258

r
(a) (b)
Fig. 2.46

Consider next a frame which is statically indeterminate to the second

degree such as that shown in Fig2.46a. If a main system obtained by rerno-
ving the hinge support b is considered, then due to a uniform increase in
temperature, point b will assume the new position b' as indicated in Fig.
2.46 h. The redundants Xl and X 2 , which are the horizontal and vertical
reaction components at b are obtained from the two conditions t hat both
the horizontal and vertical displacements at point b are zero. Thus,

abt + XI abl + X 2 ab2 0

Ybt + XI Ybl + X 2 Yb2 0

0bb Ob2, jlbi and 'YbZ have the same meaning as before. Obt and jlbt
are the horizontal and vertical displacem( nts at point b due to
temperature change and may be obtained by the application of equation
1.54 or more readily by considering the geometry of the expanded frame.
Thus referring to Fig. 2.46 b,
Sbt = at t L a n d Yb, ~ at t (hi - h 2 )
It should be noticed that the directions assumed for the rerlunrlants
follow- those of the temperature displacements. Consequently, a positive
value for a redundant indicates tllat it acts in the assumed direction and
vice versa.

2.16 Effect of support"movem.ents op statically indeterminate frames

The effect of support movements on statically indeterminate frames

may be evaluated in a manner similar to that used in regard to external
loads and temperature changes. For a frame which is statically indetermin-
ate to the n degree, the superposition equations are those expressed
by equations 2.10 and reproduced below.
 259

Xl 0" + X 2 012 +
Xl 021 + X 2 022 +

where olsJ 82u . 0ns are the forced displacements due to support
movements at the locations where the redundants XI" X 2 , Xn act
and in their sense and direction.

It should be remembered thatin choosing a main system, it is preferable,

although not necessary, to choose the redundants at the points where
support movements occur.

2 z
(a) (e)
Fig. 2.47

C onsider for example the frame shown in Fig. 2.47a, and let it
be r equired to find the reactions du e to a vertical downward settlement =
a and a clockwise r otation = b at th e right support 1.

The best main system for the purpose a t hand i s obtained by fr eeing
support 1 as shown in Fig. 2.47h. Referring to this figure, t h e
superposi t ion equations are:

X l 0" + X 2 0 12 + X, 0" = 0
Xl Yll + X 2 Yl2 + X,Y13 a

X I al l + X2 a l2 + X 3 a l3 - b

By the simultaneous solu tion of these equations the redundants X h

X 2 a nd X 3, which are the reactions at t h e right support, are obtained.
The remaining r eactions may next be fou nd by applying the equations
of statics.
260

Another possible main system is obtained by replacing the fixed

support at I by a roller and the fixed support at 2 by a hinge. The redun-
dants indicated in Fig.2.47c are then obtained from the following
superpostion equations :
X'. 8" + X' 2 8 12 + X'3813 0
X'I al J + X'2 12 + X/3 aU
4 -b
X'I a21 + X'2 + X'3 a 23 +
a22 42s - 0
whener a2s = the rotation at 2 in the main system due to the given
support movements. az s may be calculated by the application of
equation 1.55. This, however, constitutes additional effort and this is why
it is preferable to choose the redundants at the points where support
movements occur.

2.17 Further esaD1ples on Cnunes involving no-load effects

Esunple 2.28 For the frame in problem 2. 17 (Fig. 2.34), find the reac-
tions due to :
a) a uniform rise in temperature in part ce of 20C.
b) a horizontal outward displacement of I em. at support a .
c) a vertical settlement of 2 em. at support h .
". = I X IO-'/"C. and E1 = 20000 m 2.t.

r
6
l8m
Q31
8 0.05 0.31
0.08
0.0125 t
0.0125 O.08t
(a) (b)
Fig. 2.48

,
Solution: The frame is statically indeterminate to the first degree.
A pomble main system may be obtained hy replacing t he hinge a t a by
a roller.
(a) Reactions due to rise in temperature.
,
I
i.
261

The redundant XL> which is the horizontal reaction component at a

is found from the condition: Sat +
XI 8a1 = O. Since part cde only
is subject to a uniform rise in temperature, Sat is readily obtainable from
the geometry of the expanded frame,
3at = at t L = 1 X 10-5 X 20 X 800 = 0.16 (to the left)
The horizontal displacement at a due to a unit left-to-right load
applied there has already been found in example 2. 17.
637.33 637.33
3., - - X 100 = 3.2 cm. (to the right)
EI 20000

X, ~ - 0.16
0.05 t. (to the right)
3a , 3.2

The remaining reaction components are found from statics and are
as indicated in Fig. 2.48a.
b ) Reactions due to horizontal outward movement at a,
The redundant X'I' which is again chosen as the horizontal reaction
component at a, due to a horizontal outward movement at a is found from
the condition: X'I Sal = Sas Thus,
- I
X'I = - = - 0.31 t., i.e. 0.31 t. (to the left)
3.2
The remaining reactions are easily foum! from statics and are as indica-
ted in Fig. 2 A8b.
c) Reactions due to vertical settlement at h.
The most suitable main system for the evaluation of the effect of
settlement at support b is obtained by replacing the hinge at b by a
vertical roller as shown in Fjg. 2.49a. The redundant X" I' which is the
vertical reaction at b, is found from the condition: X"I Sbl = SbS.

24mt

a 4 0.16
4t 0.16t
X:=l~
004
tl 004!
(0)1'01,0 (e )
Fig. 2.49
262

R"ferring to Fig. 2.49 b, which shows the M,.D.,

8., =J M~d1 = ~ (~x

EI EI 3
242+ ~(242+322+24X32)+~x
3 3
322)

10200
= 20000 X 100= 51 cm. (downward)

2
Hence, X", = - = 0.04 t. (downward)
51

The remaining reactions are found from statics and are as indicated
in Fig. 2.490.

It should be remembered that the main system used for requirements

(a) and (b) can still be used in evaluating the effect of the settlement at
support h. In this case, howcV'cr. . the horizontal displacement at point a
corresponding to the given vertical settlement at point b should be first
calculated. This is done by the application of equation 1.55. Referring
back to Fig. 2.34 c which shows the reactions due to a unit left-to-right
load applied at point a . .

W. 0 = I X 8 - 0.25 X 2
8a = 0.5 em. (to the right)

and the redundant at a is found from the condition:8as +X'" 18.1=0 Thus,
0.5
X"', = - - = - 0.16 t i.e. 0.16 t. (to the left)
2.96
which is consistent with the value obtained before using the other main
system.

Exanaple 2.29 For the frame in problem 2.18 (Fig. 2.35), draw the B.M.D.
in the absence of load if the fixed support rotates clockwise an angle of
0.003 rad. EI = 5000 m 2 t.

Solution: The main system adopted in example 2.18, i.e. that obtained
by replacing the fixed support at a by a hinge, is the system best suited for
the purpose at hand.

Assuming the moment at a to act clockwise as shown in Fig. 2.35c,

the redundant XI due to partial rotation at a is found from the condition:
 263

a" X, = 0.003
a.1 has already been calculated in example 2.18 and is given by :

a = - 2 = -2- = 4 X 10- 4 rad .

a' EI 5000

0.003
H ence, X, = 4= 7.5 m.t. (clockwise)
4 X 10
It is interesting to note that the positive moment at a produced by a
clockwise rotation of the fixed support = 0.003 rad. offsets the negative
moment developed there due to the applied load.

Exan>ple 2.30 Calculate the reactions of the frame in problem 2.22 (Fig.
2.39) in the absence of load if the indoor temperature drops 30C while
the outdoor temperature rises] DOC . T he frame has a uniform symmetrical
section 50 em. deep, at _ I X IO- ' I"C. a nd EI = 25000 m'.t.

81m t

b
a
(a) NrD. ( b) M.!:).

F ig. 2. 50

Solution : In general the given frame is statically indeterminate to the

second degree. A possible main system is obtained by disconnecting the
frame at the intermediate hinge. The two redundants Xl and X 2 will
then be the horizontal and vertical reaction components at c. Due to sym-
metry in the statically indeterminate part of the frame, however, there
will be no relative vertical displacement between the displaced ends at
c due to temperature change. Hence" X 2 is zero and XI is determined
from the condition: ~ct + XI ~cl = 0

ad has already been calculated in example 2.22 and is given by

519 519
= -- = - - X 100 = 2.076 cm.
EI 25000
264

Figs 2.50 a and b show the N , .D. and the M,.D. d ue to X, = I.

Referring to these figures and applying equation 1.54,

act = I X 10-5 X ~o x 2 ( .j73 x .j873) 100 - I x 10-5

40 8.54 x 3 3 +9 )
x- x
50
2(
2
+- 2
- x 6 100

- 0.160 - 0.078 = 0.082 em.

0 .082
Henee,X,= - - - = - 0.04 t.
2.076

From statics"
Ya Yb = 0
X. 0.04 t. (to the left)
Xb 0.04 t. (to the right)
M. Mb = - C.36 m.t.

2.18 Applications to statically indeterJD.inate com.posite structures

The principles used in the analysis of statically indeterminate composite
structures are the same as those already considered in regard to trusses,
beams and frames. The procedures of determining the degree of in deter-
minancy, choosing the redundants, reducing the structure to a stable and
statically determinate main system and setting out a number of deflection
equations remain as the basic approach. In computing the displacement
coefficients in the deflection equations, however" it is necessary to consider
the effect of axial deformation in members which are subject primarily
to axial forces on one hand and the bending deformation in members
which are mainly subject to bending moments on the other hand.

Since no new theory is involved. . the method is best explained through

discussions on specific examples.
 265

"an.ple 2.31 Find the forces in the link members. and draw the B.M.D.
for beam ab of the trussed beam shown in Fig. 2.51a. For beam ab, A =
120 em 2, I = 30000 em 4 and E = 2000 tfem 2. For link members,
A, = 20 em 2 and E = 2000 t fem 2 .

1.5 tim

(a)~4m
r- ~ 14-J-1
1.5t/m
4 4 --!-- 4
1I2t 1/2

~:;:S~t:~~
':0 I~I
48mt
(b) Mn. t;;.Os (c) M,., F,.Os.

Fig. 2.51

Solution: The beam is internally statically indeterminate to the first

degree. A possible main system may be obtained by cutting member cd
at any section along its length. The redundant XI' which is the force in
the cut member, may be found from the condition: 8 10 + Xl 8 11 = 0

where 3 refers to the relative displacement at the ends of the cut member.

EI = 2000 X 30000 X 10- 4 = 6000 m 2 t .

EA, = 2000 X 20 40000 t.
EA = 2000 X 120 240000 t.

Figs. 2.51 band c show the forces and moments in the chosen main
.ystem due to the applied loads and due to the case of loading X, = I.
Referring to these figures and using the method of virtual work,
266

=J M,M.dl
EI
=_2 (~X EI 3
5
8 x 48 X-
8
960
= - - - X 100 = -16cm.

J
6000

J
2
M2 dl F2L N dl
8" -- ~
'+1:'+'
EA, EA

=~(~X EI 3
3 2 )+ _ 1
EA,
[4(~)2X5+IX8]+
8
_1 (16X~X~)
EA 2 2
48 15.81 4)
= ( 6000 + 40000 + 240000 100
= 0.8 + 0.0395 + 0.0017 = 0.8412 em.
-16
Hence, X, =- --- = 19.0 t. (tension)
0.8412
The forces in the link members are readily obtainable from the relation-
ship: F = F,X,. Similarly, the final B.M.D. is obtained from the super-
position relationship: M = Mo + MIX, . The result is shown in Figs.
2.52 a and b respectively.
9mt

Fig. 2.52

It would be interesting to compare the negative moment at e in the

given trussed beam with that in a similar two-span continuous beam where
the moment at the central support is given by :

M =- -- =- wL2
8
1.5 X 8"
8
=- 12 m.t.

Alternative solution : The beam may be reduced to a stable and statically

determinate main system by inserting a hinge in the beam at e. The redun-
dant X'I is then the bending moment at e and is found from the
condition: aeo + X'I act = 0
where a. refers to the relative rotation at e.
 267

X;=O X;=1 m!
8 lifo 1i6J

~'Il"
.~

1m!

+ +
12 12m!
(a) Mo.,No.'~.Os. (b) Ml.,Nr'~ .Os.

Fig. 2.53

Figs. 2.53 a and b show the forces and moments in the chosen main
system due to the applied loads and due to the case of loading X', = I.
Referring to these figures and using the method of virtual work"
SN,Nodl
a", =
J M,Modl

~
EI

(~ X 8
+ 1:
F,FoL
EA, +
~) = 6000
EA

f
1:
M,Modl
EI

F,FoL =
=_
EI 3

L [~X l6X8+4(~X
X 12 X
2
64 X 100 = _

10) 5] = 84.3340000
X 100 =
1.067

0.211
EA, EA, 3 24

J N,Nodl
EA
I ( I )
= EA 8 X X 16 =
21.33 X 100
6240000 = 0.009

J
a.o = - 1.067 + 0.211 + 0.009 = - 0.847

M~dl
a., =

~(~ X I
J EI +
FiL
1: EA ,+
N:dl
EA

J Midl
-~-=
EI EI 3
FiL
1: - - = _ I [ 4 (5)
- 25 +
2) = 16xlOO =0.089
3 X 6000
( ~) 2
8 ]
= 1.756 X 100 = 0.0044
EA, EA, 24 3 40000
[(~)"
J a., N:dl =_1
EA EA 6
X 16] = 4 X 100
9 X 240000
= 0.089 + 0.0044 + 0.000 = 0.0934
=0.000

- 0.847
X', = -
0.0934
= 9 m.t. +
The positive sign indicates that the moment at e acts in the assumed
 268

direction. This is consistent with the value obtained from the first, and
obviously simpler, main system.

Example 2.32 A cantilever ab is braced by a system of link members aa

shown in Fig. 2.54 a. Draw the S.F. and B.M.Ds. for beam ab if
E = 2000 tlcm 2, I for beam = 20000 cm 4 and L/A for link members =
18 em-I. Neglect axial deformation in the beam.

36
61 61
~o~ ____~c~______~b
f~
1.5
L<>::------e:
t.:.- 2 m _I-'d_ o
(a)

2 2mt
6 61

)(,=1
1187t
8.85 14.7 8.85
.4/3 (d)
(c) M, .. F,.Os
18.3ml
121 12
+
+
2.8 2.8 5.7
(e) S.F.O (llS.M.O

Fig. 2.54

Solution: The structure is statically indeterminate to the first degree. A

possible main system may be obtained by cutting link member cd at
any section along its length. The redundant XI' which is the force in this
member, is then obtained from the condition: 3 10 + XI 8 11 = o.
 269

Figs. 2.54 band e show the forces and moments due to the applied loads
and due to the case ofloading X, = I. Referring to these figures: and
using the method of virtual work,

~'o

=
=
SM,MOdl

112
EI

EI =
=

112 X 100
4000
I [2X2
EI 2
= 2.S em.'
X
2
3 X 12 +
(12 + 36)
2

8" _ M:dI + 1: F:L

-EI EA

= ~In X3X2'+~X2']+~~J(ir+Gr+12]
10.67 X 100
+ -IS 50
X - = 0.266 + 0.050 = 0.316 em.
4000 2000 9

2.S
Hence, X, = - -- = - S.S5t. i.e. S.S5t. (tension)
0.316
The free body diagram together with the S.F. and B.M.Ds. are shown
10 Figs. 2.54 d,e and f.
Alternative solution:

Q
+
6/5mt
.31
e 0 -415
(a) M<f.~.DS. (b) M,., 'i.Ds.

Fig. 2,55

Considering the force in link member bd as redundant X', and

referring to Fig. 2.55,

8,. =
J M,M. dl

67.2
EI

= _ - - X 100 = -
=-
1 [12 X 2
EI

1.68 em.
2 X
2 6
3" X 5+ 24 X 2 X
6]
5

4000
 270

=j ill"" + MidI L
-
EA
1;
F2
I

= ~I
1
n 0)"+ i or]+ ~J or+ (;)"+
x x 3
18
12]

= --
4000
X 3.84 X 100 + 2000
-- X 2 = 0.096 + 0.018 = 0.114 em.

, -1.68
Hence, X 1= - -- =
0.114
+ 14.7 t.

which is consistent with the value obtained in the first solution.

Exa.nple 2.33 Draw the B.M.D. for the frame shown in Fig. 2.56 a. For
frame EI = 20000 m 2t. and for tie EA = 40000 t.

j I

b
~6m I 6 -..l
(a)

7.6B

F, =31 F=B.161

(d) 8.M.D.
Fig. 2.56

Solution : The frame is statically indeterminate to the first degree. A

possible main system may be obtained by replacing the hinge at b by a
. roller, and the redundant Xl is then the horizontal reaction componr.nt
at b. This is found from the condition: SbO + XI SbJ = O.
 271

The Mo.D. and the force in link member de due to the applied
loads are shown in Fig. 2.56 h. The corresponding values due to a
unit horizontal load at b are shown in Fig. 2.S6c. Referring to these
figures and using the method of virtual work,

~bO
=j - -100-(
20000
2
2X - x 6.72x9x3)
3
3 x 12 X 1200
40000
1.21 - 1.08 = - 2.29 em.

= J' __M-;,~ .,..d_l_

EI
F~L
+ - E-A- =
100x2
20000
(~ X 62
3
+ 6.72 X6 2)
3
3 2 X 1200
+ 40000
- 1.526 + 0.27 1.796 cm.

- 2.29
Hence, X 1 = 1.28 t. (to the left)
1.796

The final B.M.D . and the force in tie de are found by superposition and
are as indicated in Fig. 2.56 d.

Alternative solution: Another possible main system is obtained by cutting

link de and considering its force X'I as redundant. The value of
this redundant is obtained from the condition: 3 10 XII 3 11 = 0,+

--

(b)M,..F,.Ds.

Fig. 2.57

where 8 refers to the relative displacement between the cut ends of the
link.
272

The Mo.D. due to the applied loads and the M,.D. due to the case of
loading X,' = I are shown in Figs. 2.57 a and b re:optttively. Referring
to these figures and using the method of virtual work,>

M,Modl lOOx2 ( 24 x6 2 2-24 x 6.7

x- x
EI 20000 2 3 2
3
x-
2
x 2+ -
2
3
x 6.72 X 9 X I)
= -1.632 em.

M~dl 1 L _ :I..:OO
8"
=J EI +
X
EA
.:......:X
20000
I X 1200
- 6 X 22
..:.....:2:...,(
3'
6.72
+- - x2 2)
3

+ 40000
- 0.17 + 0.03
= 0.2 em.
-1.632
Hence, X', = - -....:..:.:..:.::. 8.16 t.
0.2

which is identical to the corresponding value obtained in the first

solution.

It is interesting to note that if the deformation in the tie is neglected

the value of the redundant will be, X', = 1.632/0.17 = 9.6 t. witb an
error of about 18%.

Example 2.34 Find the reactions and the force in the tie cd ofthe frame
shown in Fig. :1.58 a. For tie EA = 10000 t. and for frame EI = 20000
m 2t.

Solution : The frame is statically indeterminate to the second degree. A

possible main system may be obtained by replacing the hinge support at b
by a roller and cutting the tie at a section x along its length. The redun-
dants Xl and X 2 are therefore the horizontal reaction component at b
and the force in the tie respectively. These may be found from the
simultaneous solution of the two superposition equations :

8. 0 + X, 0., + X 2 8. 2 0
8xo + X, 8x ' + X 2 tx2 - 0
 273

,
3
1c d
6m
r La b
I I- 4 -I-- 6 m ---I-- 4 -'
(al

.-------,3 m t

(clM,.D.
Fig. 2.58

The Mo.D., the M,.D. and the M 2.D. a re shown in Figs. 2.58 b, c and
d respectively.
Referring to these figures and using the method of virtual work. .

8. 0 = J M,Mo dl
El
4480
= _ _1_(
El
5 X 40 (6 + 2) 2 + 40 X 8 X 9)
2

El
M'dl 1 ( -6- (6 2 ) 2 + - 5
8.,
=s I
El
= -
El 3 3
(6 2 + 9 2 + 6 x 9) 2

I + 9 X 8 X 9) = 1362
El
=S
I
I
8. 2

-
M,M2dl

336
El
El
_1 ( 5 X 3
El 2 (6 + 2) 2 + 3 X 8 X 9)

M2MO dl
8,0
=f 1360
El
= - - 1(5
El
X 40
2
X 2 X 2 + 40 X 8 x 3)

-
El
 274

336
- - - from Maxwell's theorem
EI

=J + IxL= _
EA EI
(2-3 (3)2X2 +8X3X3)

I X 16 102 16
+ EA = Ef+EA
134
EI

Substituting these values in the superposition equations,

- 4480 + 1362 XI + 336 X, 0
- 1360 + 336 XI + 134 X, 0

Solving these two equations simultaneously,

XI = 2.04 t. and X2 = 5.03 t.
The remaining reactions are found from statics and are as follows:
Xa = 2.04 t. (to the right) and Ya = Y b = IO t. (upward)

2.19 Deflections of statically indetern1inate structures

In the deve10pment of the method of virtual work for deflection com~

putations, there has been no restrictions on whether the structure was
statically determinate or indeterminate. It follows that the method of
virtual work is equally applicable to the determination of the deflection
of statically indeterminate structures.

Consider for example the statically indeterminate beam shown in

Fig. 2 .59 a, and let it be required to calculate the vertical deflection at
point c. Under the given load, the Mo.D. will be as shown in Fig. 2.59 b.
The first step for the calculation of the deflection at c is to place a unit
load there and find the corresponding MI.D . This has been done and the
result is presented in Fig. 2.59 c. Following the usual procedure of graph-
ical multiplication~
 275

! w tim

r (a)
1I2 c I Ll2~"
wL2J12

I +

I (c)M,.O .
wL2/B

1t

+
LIB

I (d)M;O
Ll2

~
Ll4
1t
_ _ _ _...:: ::"L ----- -j

1t

+
Ll4
Fig. 2.59

j -
_2_[
EI
wL2
12
X~(~-~)
2 S S
2 L wL2
+"3 X"'2 X- s- a X4"- a ]
(5 L L)

wL 4
384EI
is rather
The proced ure describ ed above, althoug h quite permiss ible"
minate
compli cated as it involve s the analysi s of the given statical ly indeter
a unit load
beam twice; once under the given load and anothe r under
r method "
placed at the point where the deflecti on is require d. Anothe
displac ements
and a much simple r one, is sugges ted by the fact that the
corresp on-
of the given statical ly indeter minate beam are identic al to their
by the applied
ding values in any chosen main system when acted upon
momen ts..
loads and the redund ant reactio ns, and Of interna l forces or
they may be
simulta neously . Theref ore, once the redund ants are found,
276

considered as external loads on the chosen main system and the required
deflection is found in the usual manner by placing a unit load on
the chosen main system. To generalize, any main system other than that
used in the analysis of the statically indeterminate structure may serve
the purpose. This is because while the final B.M.D. under the given case
of IOl.ding is unique, the main system used in the analysis is arbitrarily
chosen.
As an application to the above statement, consider the main system
shown in Fig. 2.59 d.

__1_ ( wL2 L L 2 L
x-x ---X-
EI 12 2 4 3 2
wL2 3 L) wL4
x 8
X -
8
X
2 = 384EI

f
Consider next the main system shown in Fig. 2.59 e.

8e = M;Modl _2_(..!. x ~ X wL2 x .2.. X~

EI EI 3 2 8 8 4
__l_ x ~ x~x wL')= wL4
2 2 4 12 384 EI

This demonstrates that on calculating the displacements of statically

indeterminate structures, the M1.D. of any statically determinate main
system may replace the corresponding B.M.D. in the original statically
indeterminate structure.

Esunple 2.35 Calculate the horizontal displacement at the eaves of the

frame shown in Fig. 2.41a if it has a uniform bending rigidity
EI = 2000 m 2 t.

1.1!-
6
J
'-- 8 m--l-- 4 -l- 4 -<
(b)M, D.

Fig. 2.60

(
I Solution: The final B.M.D. of the frame under the given case ofloading
has already been found in example. 2.24 (Fig. 2.41). This is considered
as the Mo.D. and is reproduced in Fig. 2.60 a. The best main system for
277

r the calculation of the horizontal displacement at the eaves is obtained by

removing the support at c and replacing the hinge at b by a roller. The
,
Mt.D: associated with this main system due to a unit horizontal load at
d is shown in Fig. 2.60 b.

Using the method of virtual work,

8
d
= S M,Mo dl = ' _1_ ( _ 3.89x6 x...!.x 6 _ 3.89X5 5
EI EI 2 3 2
x
+5.33X5 x4 _5.57X5 xl +5.33X5 X2)
2 2 2
29.28 29.28
=---= - - - X 100=-1.46cm. i.e. 1.46 cm.(to the left)
EI 2000

Example 2.36 C~lculate the vertical displacement at joint c of the truss

shown in Fig. 2. 21. L/A = 8 cm- 1 and E = 2000 t/cm 2 .

S:.lution : Furces Fo in the truss m:::mbers under the given case of loading
are readily obtainable from example 2.10 (Fig. 2.21), and are as given
in the second column of Table 2.7.

For c3.lculating the deflection at point cJa unit downward vertical

load is placed at c. The corresponding F I-forces in the given statically
indeterminate truss are easily deduced from the Foforces; F, = Fo/24.
8
- - - (83.09) =0.332 cm. (downward)
2000

A main system obtained by either removing the support at g or cutting

member kd ,Jay be conveniently used for the calculation of 8c' The forces
F'l in the members df the chosen main system due to a unit vertical load
at c are as listed in the fourth column of Table 2.7. The product F,'F.
is listed in column 5 of the table.

8
(83.24)= 0.332 cm. (downward)
2000
which is identical to the value obtained above by applying a unit load to
the original statically indeterminate truss.
Member F. F,F. F,' F,' F. Member F. F,F. ,
F' F,'F.
ah - 7.8 2.54 -5/12 + 3.24 gn +2 .3 +0. 22 0 0
hi -9.2 3.:4 -1/2 + 4.60 nm +2.8 +0.33 0 0
ij -18.4 14.10 -I + 18.40 ml +5.6 +1.30 0 0
jk + 6.7 1.87 0 0 Ik +6.7 +1.87 0 0
ab + 4.6 0.88 +1/4 + 1.1 5 gf -1.4 +0 .08 0 0
be +13 .8 7.93 +3/4 + 10.35 fe --4.2 +0.73 0 0
cd + 5.1 1.08 +1/2 + 2.55 ed -6.9 +1.98 0 0
bh + 7.7 2.47 +5/12 + 3.20 fn -2.3 +0.22 0 0
bi - 7.7 2.47 -5/12 + 3.20 fm +2.3 +0.22 0 0
ie + 7.7 2.47 +5/12 + 3.20 me -2.3 +0.22 0 0
ej +22.3 20.80 +5/6 + 18.60 el +2.3 +0.22 0 0
jd -17.7 13.05 -5/6 +14.75 Id +2.3 +0 .22 0 0
kd - 7.5 2.28 0
83,09 83 .24
Table 2.7
 279

2.20 Influence lines for silDple statically indeternUuate structures

Similar to any load function in statically determinate structures, the
influence line for a redundant in statically indeterminate structures may
be obtained by successive placing of a unit load at each possible position
on the structure and computing the value of the redundant under con-
sideration corresponding to each position of the unit load. These values
f- form the ordinates to the required influence line. Although this is quite
permissible, it is very tedious as it involves the analysis of the given static-
I, ally indeterminate structure several times; once for each position of the
unit load.

Another and a much simpler method of constructing the influence

lines for redundants follows from Maxwell's theorem of reciprocal deflec-
tions. This is explained in the following.
Consider the restrained beam shown in Fig. 2.61a and let it be required

(al ~a
I L

(bll.l.6b I

(cll.L. Xl

, Fig. 2.61

I to construct the influence line for the reaction at b. If the support at b

is removed and a unit downward load is applied there) then as has been
t demonstrated in section 2.17} the resulting elastic curve which is shown
in Fig. 2.61b is the influence line for the deflection at b.
!
i Now} if a unit load acts vertically downwards at an arbitrarily chosen
I
point then, by definition} the ordinate to the influence line at this point
lt will represent the deflection produced at b. For example} a unit load at
c will produce a deflection 8bc at b. The corresponding redundant reac-
! tion is obtained from the well known deflection equation: Xl = _ 8bc
S.,
~
l
280

where 8b1 is the vertical deflection at b due to a unit vertical load applied
there, which is also the ordinate 8bb ,tO the influence line for Sb .

Since point c W3.S arbitrarily chosen, it follows that the influence line
for the reaction at b is obtained by dividing the ordinates to the influence
line for the deflection at b by 8bb This influence line is shown in
Fig. 2.6Ic.

As a second example, consider the two-hinged truss shown in

Fig. 2.62a, aqd let the influence line for the redundant horizontal reaction

(a)

(c) I. L. X k-----.c:-~/--::/'1
b'<\fb

Fig. 2.62

at b,X, be required. If the hinge support at b is replaced by a roller, and

a unit horizontal load is applied there as indicated then the deflected
shape of the loaded chord ac, shown in Fig. 2.62 b, is the influence line
for the horizontal displacement at b, Sb. Now if a unit vertical load acts
downwards at any joint between a and c such as d, the corresponding
value of the redundant X is obtained from the deflection equation:

X = - Sbd The negative sign of X means that it acts in a direction

Sbl
opposite to that assumed, i.e. from right to left. Since joint d was
arbitrarily chosen,it may be concluded that if the ordinates to the influence
~ine for the displacement at b are divided by Sbl' the resulting diagram

(
L
I
f
i 281

I
is the influen ce line for the redund ant X. This influen ce line is shown
in Fig. 2.62 c.

It ml.y be stated, therefo re, that the influen ce li ne for any redund
ant
X is similar to the deflect ion curve drawn for the structu re when
X equals
unity. The values of the ordina tes are those of the deflect ion
curvc, each
r divided by the displac ement at the point where X acts and in
its directio n
I
I
for the same case of loading ; X = 1.

The commo n proced ure is to obtain the deflect ion curve of the
structu re
graphic ally while the displac ement at the point of applica
tion of the
redund ant is determ ined analyti cally. The method of elastic-
load is us-
ually used in the case of structu res which afC mainly subject ed
to bendin g
momen ts, and the displac ement diagram s are used in the case
of trusses.

To this point the constru ction of influen ce lines has been limited
to
redund ants. The constru ction of influen ce lines for other load
functio ns,
howeve r, creats no real problem as once the influen ce line for
the redund -
ant is constru cted, the influen ce line for any other desired load
functio n
is easily found by applyin g the superpo sition relation ship

where F IS any load functio n, I.L.:F is the influen ce line

o for this
load functio n in the chosen main system, and F 1 is the value
of F corres-
pondin g to the case of loading XI = I.

2.21 Wustr ative exaDl.p les on the influen ce lines

The princip les establis hed in section 2.20 will be applied to a

numbe r
of specific exampl es. In each exampl e a numbe r of typical influen
ce lines
is worked out. The studen t is advised to check the given influen
ce lines
and a ttempt the constru ction of additio na l influen ce lines for
other load
functio ns o f his choice.

Examp le 2.37 Constr uct the influen ce lines for the reactio ns at
suppor ts
a and c of the continu ous truss shown in Fig. 2.63a. Also constru
ct the
influen ce line for the force in membe r 23 and find the extrem
e values of
this force if the truss is subject ed to a uniform ly distribu ted
dead load,
g = .4 tim and live load w = 8 tim. L/A = 2.5 cm- 1 and E =
2000t/cm. 2
282

L
1....
4m
~al~--~5~~~6~~~--~--~~~
~ 3x 6:18 _..:!:=-_
(a)

1 2
/ /~
\ / 4
\Zj
a/ \ ~
a'___ _
10.5 5 6 'X: (
( b)

' 65

0.0376 em
(d ) 1.L. 6 c

(f) I.L. Yo

-1.5( I.L.X)
2m ! (e)Wil Iiot Diagra m
(g )l. L. r:;3 250 do-va lues
0.8 !

0.2 0.15

Fig. 2.63
degree. A
Solutio n: The given truss is staticaHy indeter minate to the first
t at c.
possible main system may be obtained by removing the roller suppor
t
,
I 283
t
The member forces and deformations due to X It. are presented in
Table 2.8.

Member a-I 1-2 2-3 3-4 a -5 5-6 6-c

F (t) -5/8 -3/4 -3/2 - 1.8 + 318 + 9/ 8 +15/8

d8 (10- cm) -0.78 -0.94 - 1.88 -2.25 + 0.47 + 1.41 +2.35
3
,
Member 1-5 2-6 3-6 3-c 4-c
t
,
5-2
F (t) + 5/ 8 -5/ 8 + 518 - 5/8 - 5/8 +2
d8 (10- em) + 0.78 + 0.78 + 0.78 -0.78 -0.78 + 2.50
3

I Table 2.8

I The corresponding displacement diagram drawn with 250 dS-values

and assuming member c4 to remain fixed in direction is shown in Fig.

~, 2.63 c . Indicated on this figure are the displacements at the panel points,
which represent the ordinates of the influence line for Be shown in Fig.
2.63 d. The influence line for X is obtained by dividing each of these
ordinates by Set- . The result is shown in Fig. 2 .63 c. The influence line

t for the reaction at a, shown Fig. 2 .63 f, is obtained from the relationship:
1.L. Y = 1.L.Yo + Y t (1.L. X); Y t = 0.5. Similarly, the influence line
for the force in member 23 is obtained from the superposition relationship:
1.L. F=1.L. Fo+F t (I.L.x); F t = - 1.5 The result is shown in Fig. 2.63 g
and the same influence line drawn on a horizontal base~line is shown in
Fig. 2.63 h. Referring to this figure,

f F 23 (due to D.L.) = 4 (-3.6) = - 14.4 1.

. . (0.2 x6 0.2+0.15
M".postllveF 23 (duetoL.L.)=8 - - + - - 0 - - x6+
0.15 X6)
2

I = 16.8
2
t.

0.35 X 6 0.35+0.8
2

0.8 X 0.6)
! M., negative F 23 (due to L.L.)=-8 ( 2 + 2 x6+ - -- -
2
=-45.6t.

I I
The extreme values are: + 2.4t. and 50 t.

ExaDlple 2.38 If load is transmitted to the lower chord of the truss shown
Fig. 2.64 a, construct the influence lines for the horizontal reaction at support
b and the force in member 5 - 7. L/A = 4 cm- 1 and E = 2000 tlcm 2.

t! Solution: The truss is statically indeterminate to the first degree. A pos-

sible main system may be obtained by replacing the hinge at b by a roller,
and the redundant X is then the horizontal reaction component at b. The

t inft.uence line for X is obtained from the relationship:

284

LL.X = - (LL. 8b )/8 bl 8bl is the horizontal displacement at b due to

a unit load applied there. The LL. for 8b has been found in example 1.80
(Fig. !.III c) and is re-produced in Fig. 2.64 b.

(a)

0..0.58 em.

(e )l.l. X
0..194 0..328 0.343 0..20.9
+ I
0..666 t.

I.l. Fa "-(l.l.X)

(d) l. L. ~1 1"""'==------'---""'...:-----1

0..876 20/9

(e 1I.l. ~ i-'=_---l'---.---l_---.l.:::::"..OCT_-,-_-11

1.0.67 t
Fig. 2.64

Referring back to the displacement diagram in Fig. 1.\11 b,8 bl =b"b'.

Tbus, by measurement 8bl = 11.7/175 = 0.067 em.
 285

The LL. for X is t hus as shown in Fig. 2.64 c . T he LL. for FL7

is obtained from the relationship: LL. F = LL. Fo + F, (LL.X) ,

F, = +4 /3 t. The LL. for F'-7 is thus as shown in Fig. 2.64 d. The
same I.L. drawn on a horizontal base-line is shown i n Fig . 2.64 e.

Exall1ple 2.39 Ifload is transm itted to the lower chord of the three-chord
truss shown in Fig. 2.65a, construct the influence lines for the forces in
members 1-4 and 5-12 . The cross-sectional area of all the verticals and
diagonals 1-4, 4 -5, 5-8 and 8-9 = 100 em 2, and the cross-sectional area
of the remaining members = 200 cm 2 . E = 2000 tlcm 2.

I
J
12
13
3m
"
t 5 7 9

t 6 8

~rl
1 - - - -- - 5x 6=30 m - - _ - ' = ,
tal 2T
12 024

r
~.
11

3 5

a~-*2--~~4~~--~6----~8~---'ln~4b
13

7 9
a

(b)

t
(d) l.L X(I.L r;;-11)

-3.35(J.L.X)

(e) 1.L. F,.4

0.781
0.2
0.11 0.04 12

0.41 (c)Willl01 Diagram

(fll.L. F,.L 25 d6-Values

Fig. 2.65

Solution: The truss is internally statically indeterminate to the first

degree. A possible main system may be obtained by cutting member 5-12
 286

at any section along its length, and the redundant X is then the force in
this member. The member forces and deformations due to X = It are
presented in Table 2.9

Member L (em. ) A (cm . 2 ) F (t) d8 (10- 3 em)

I - II 670 200 -3. 35 - 5.6
II - 12 610 200 -3.04 - 4.6
3 - II 300 100 + 1.00 + 1.5
5 - 12 400 100 + 1.00 + 2.0
a - I 430 200 0 0
I -3 600 200 +6 .00 + 9.0
3-5 600 200 +6.00 , 9.0
-L
a -2 300 200 0 0
2-4 6)0 200 0 0
4-6 6~0 200 -4 .00 - 6.0
I-4 670 100 -3 .35 -11.2
4 -5 670 100 + 1.I3 -I- 3.8
I -2 300 100 0 0
3-4 300 100 + 1.00 + 1.5
5-6 300 100 0 0
Table 2.9

The c:>rresponding displacement diagram drawn with 25d~-values

and assuming member 5 - 6 to remain fixed in direction is shown
in Fig. 2 .65 c. Indicated on this diagram are the displacements at the
panel points, which represent the ordinates of the influence line for the
relative displacement at the cut section. The influence line for X
is obtained by dividing these ordinates by S11 which is the relative dis-
placement at the cut section due to X = 1. This is approximately equal
to the relative displacement between joints 5 and 12 which, by measure
m~nt from the displacement diagram, is equal to 0.42cm. The influence
line for X, which is the required influence line for the force in member
5-12, is thus as shown in Fig. 2.65 d. The influence line for the force in
member 14 shown in Fig. 2.65e is obtained from the relationship:
LL. F = LL. Fo + F, (LL. X); F, = - 3.35. The same influence line
drawn on a horizontal base-line is shown in Fig. 2.65 f.

Example 2.40 If load is transmitted to the frame shown in Fig. 2.66 a

at panel points spaced at 3 m, construct the influence lines for the hori-
zontal reaction at b and the bending moment at section n at the middle
of cd . EI = 20000 m"t.
)

.- ---------
 287
c n d
r
3
La 6m

I 9m
(a)
4
6mt
3

10.5
1
13.5
I
16.5
I
(b)M, .D.
16.5
i
I 0.20
t
+
13."1------~0
(d)l.L. 6 b
0.34cm

10.

(e) l.L. X

0.252 t
-2.25
........ .........

.
\

0.38
Fig. 2.66
288

Solution : The given frame is statically indeterminate to the first degree.

A possible main system may be obtained by replacing the hinge at b by
a rollerJ and the redundant X is then the horizontal reaction component
at h. The influence line for X is obtained from the relationship:
LL. X= - (LL. 3 b) I 3bl ; 3bl is the horizontal displacement at b
due to a unit load applied there.
3bl is nIcuIated with reference to Fig.2.66 b which shows the B.M.D.
due to a unit load at h.

M 2 1 (3_ (3)2 +-(3

dl = - 9 2 + 6)+ 6- (6)2)
3bl =

270
f_1_
EI
270
EI 3 3
62 + 3 X
3

=- = - - X 100 = 1.35 cm.

EI 20000
The LL. for 3b is obtained graphically in the manner described in
section 1.8 and the result is shown in Fig. 2.66 d. The scales used are:
L.S., I cm = 200 cm, E.L.S., 1 cm. = 10 m 2t and ELS., I cm. =
8000 m 2t. The displacement scale is thus made up as follows:
200 X 10
Icm = = 0.25 cm.
8000
Thus.. by measurement, the ocdinates of the jnfluence line for Bb are
as indicated in Fig. 2.66 d. The ordinates of the influence line for X are
obtained by dividing the corresponding ordinates of the influence line for
3b by 3bh where 3bl = 1.35. The r esult'is shown in Fig. 2.66 e. The
influence line for M n, shown in Fig. 2.66 f, is obtained from the relation-
ship: LL. M = I.L. Mo + MI (LL.x); MI = - 4.5. The same influence
line drawn on a horizontal base-line is shown in Fig. 2.66 g.

Exan>ple 2.41 If load is transmitted to the arch with a tie shown in Fig.
2.67a at panel points a-b, construct the influence line for the bending
moment at section n at the middle of panel cd. For arch EI = 40000 m 2t
and for tie A = 100 cm. 2 and E = 2100 tlcm 2.

Solution: The given structure is statically indeterminate to the first

degree. A possible main system may be obtained by cutting the tie at any
section along its l ongth , and the redundant Xes then the force in the tie.
The influence line for X is obtained by dividing the ordinates of the in-
fluence line for 3b by 3 11 , where 3 JI is the relative displacement at the
cut section due to X = I. The LL. for 3b has been found in example
1.77 (Fig. 1.108 e) and is re-produced in Fig. 2.67 b.
 289

~ jm~.
6X4=24m~
(a)

0.88 0.96 em.

0.52

(e) I.L. X
4.5 I. L. Mo
/ -4.5(I.L.X )
/ + ~,..",.---,-===

(d) I. L. Mnf----'_ _ _ _ _ _ _~

Fig. 2.67

Referring to Fig. 1.108 b which shows the M,.D.,

S _JM,2 dl I xL
11- EI+EA

2 X 100 (5.3 4.47 4.04

= -- - (3.5)2 + - (3.5 2 + 5.5 2 + 3.5X5.5) + -
40000 3 3 3
2 2 I x24xlO0
)
(5.5 + 6 + 5.5 xg) + =1.236 + 0.011=1.247 cm.
2100 X 100
The LL. for X is thus as shown in Fig. 2.67 c. The positive sign indi-
cates that X acts as assumed in Fig. 1.108b. ( '
The LL. for M., shown in Fig. 2.67 d, is obtained from the relation-
ship: I.L. M = I.L. Mo +
M, (LL.X); M, = - 4.5. The same influence
line drawn on a horizontal base-line is shown in Fig. 2.67c.
290

R"n"ple 2.42 If load is transmitted to the trussed beam shown in Fig.

2.68a at panel points spaced at 3 m, construct the influence line for the
moment at point e. For beam I = 200000 em" and A = 300 cm 2 , for
linIc members A = 30 em 2 and E = 2000 t/cm 2 .

(a)

(b)M,,'i. ""lE"--L--t----'-------.------=~
~"."< -0.5 -0.5 ...,:,-~

(c)J.L.6

(d)J.L. X

(e)l.L. M~

( O I.L. M~
0.06
1.62 m t
Fig. 2.68

Solution : The given trussed beam is statically indeterminate to the first

degree. A possible main system may be obtained by cutting member dT
at any section along its length. Fig. 2.68b shows the bending moment
diagram and the forces in the link members due to X = L The corres-
ponding deflected shape of the beam which is also the influence line for
~
 291

the relative displacement at the cut section is shown in Fig. 2.68 c. This
is obtained analytically using the method of elastic-load.

EI = 2000 X 200000 X 10- 4 = 4{)000 m 2 t

Referring to Fig. 2.68 b,

I (3X6 ) 18
r. = EI -2- +3 X 3 = EI

= 0 =_1 (18 X 6 _ 1.5x3 XI) =51.75 = 51.75 X 100 O.l29 cm.

O. EI 2 EI 4{)000

I ( 3 X 6 ) 90 90 X 100
0. = or = - 18 X 6 - - - X 2 = - = 0.225 cm.
EI 2 EI 4{)000

I (3X6 ) 103.5 103.5 X 100

8. = - l8x9- - - x5--3x3x 1.5 = 4{)0()() = 0.259cm.
U 2 U

The influence line for X is obtained by dividing the ordinates of the

influence line for i5 by 3 11 where,

f
M,2dl 1 xL F,2 L
o11-
- EI
+ - - + :1;--
EAb EA

Referring again to Fig. 2.68 b,

= ~ [2 (~ X 3 2) + 6 X 3 X 3] + I X 1800
4{)0()() 3 2000 X 300

+ 100 [(12 X6 + 2 (0.5 2X3)+2 (1.122 X6.7)]

30 x2000
- 0.225 + 0.003 + 0.041 = 0.269 cm.

The influence line for X is thus as shown in Fig. 2.68d. The

influence line for ~ shown in Fig. 2.68 e, is obtained from the relation-
ship: I.L. M = I.L. Mo +M, (I.L. X); M, = - [3. The same influence .
line: drawn on a horizontal base-line is shown in Fig. 2.68 f.
("
292

Exunple 2.43 Construct the i nfluence line for the moment a t section
D at the middle of the Bow String Girder shown in Fig. 2.69 a. For girder
I = 200000 em', for vertica l link members A = 30 em" and for the
remaining link members A = 60 em". E = 2100 t/cm".

l~~dl
e' ,

3m b

rA~ 1 5x5 =25m _ _---=::r

4. X=1 t4 . 5
3 m t ,.

(b) 1-1,.0 .

7.5 18.75 22.5 18.75 7.5 7.5

18.75

(d) I.L X
18.75
I.l.M ~J~ _4.5 CI.L X)
7.5 V I
~
(e)

0)55 0.155

Fig. 2.69

Solution: The given structure is statically indeterminate to the first

degree. A possible main system may be obtained by cutting link d'e' at
any section along its length, and the redundant X is then the force in this
member . The influence line for X, shown in Fig. 2.69 d, is obtained by
dividing the infil,lence line for the relative displacement between the ends
of the cut member by 8". The influence line for the relative displace-
ment at the cut section is obtained graphically with reference to Figs.
2.69 b - d. The scales used are: (
 293

L.S.. I CID = 500 CID.

E.L.S. I em = 20 m 2 t.
EI.S.. I em = 20000 m 2to
The displacement scale is thus made up as follo'ws :
500 X 20
I em= 20000 = 05 . em .
M,2 d l F,2 L
8" =
S EI

Referring to Fig. 2.69 b,

+ l; - -
EA

j M,2 dl
--= -
EI
I(5
EI
2 X_
3
5
(3)2 + 2 X -(3 2 +4.5 2 +3x4.5) +5x4.5 2
3
)
2 75 275
= - = - - X 100 = 0.655 CID.
EI 42000

Member F, (t) L (em) A (em 2 ) d8(lo-2cm.)

die' 1.04 520 60 4.500

e'f' 1.04 520 60 4.500
dd ' 0.30 450 30 0 .064-
,
ee, 0.30 450 30 0.064-
ea Ll7 584 60 6.350
f'b LI7 584 60 6.350
ce' 0.60 300 30 1.720
ff' 0.60 300 30 1.720
d'e' 1.00 500 60 3.968
29.236
Table 2.10

The contribution d8 of the axial strains in the link members to 8 11

is calculated in Table 2.10.
Il" = 0.655 + 0.029 = 0.664 em.
The X-scale is thus made up as follows :
I em. =0.5/0.664 = 0.73 to

By measurement,the ordinates of the I.L. X, indicated in Fig. 2.69d, are:

y, = 1.3 X 0.73 = 0.95 t.
Y2 = 0.8 X 0.73 = 0.59 t.
The LL. M. shown in Fig. 2.69 e is obtained from tbe rdationship:
I.L. M = I.L. Mo + M, (LL. Xl; M, = - 4.5. The same I.L.
drawn on a horizontal base-line is shown in Fig. 2.69 f.
294

EXAM PLES TO BE WORK ED our

(1)-(6) For each of the statical ly indeter minate beams in Figs.
2.70-
2.75, sketch two possibl e main systems indicat ing the reclund ants
in each
case. Using one of the sugges ted main systems , draw the S.F. and
B.M.D s.

2t1m
d
c
-+- 2 --+-- I. m --I
Fig. 2.70

~o r 8
"
Ttlm
I 1I I I
1;- 3'1d
6m
~ 3 --I--- 3 --I- 3
Fig. 2.71
3 -'
2t1m 111m

Fig. 2.72

Fig. 2.73

~"m rOt
~~lllml
~
1i I'~ 21
3 -+-3m _ ._ 3 -+- 3 --<
n 1 ~
Fig. 2.74

~
I ,r 21r' 1m
~' l~~ r I
"'
3x3= 9 m 3x3= 9 m
~ (
Fig. 2.75
 295

of the
(7) - (12) State the degree of statical indder minan cy of each
method
beamsi B Figs. 2.76-2.81, then draw the S.F. and B.M. Ds. using the
of consistent deformations.

141 1S t
1St 2 tim "
~!!I!!!~
:a
a
l-- 2 _ 2 -+-- L m --l- 2 -<---
c
L ---l

Fig. 2.76

lSt/m "

2 -I-- 3m - I
Fig. 2.77

pt Sfl 1St Stl 1St ~

~ 2[ - [3 -l-213 ~
~ 3[ -<- 3[ -<- 213 ~m-l -I
I--- 3 -I--
Fig. 2.78

SII lSI
~ 21 1 21 ~
I- 3 -1-3 m..I- 3-l
Fig. 2.79

~
I
O! I I '2/
3t1m
I! I I

3 3 I Sm- --,
.- -+--
Fig. 2.8\
 296

(i3) Draw the B.M.D. for the beam in Fig.2.70 in the absence ofload
if EI = 5000 m 2t and,

a) support b settles by l.8 cm.,

b) support c has an anticlockwise rotational slip of 0.003 tad.

(14) Draw the B.M.D. for the beam in Fig.2.71 in the absence ofload
if EI = 6000 m 2t and,
a) support c settles by 3 cm.,
b) support a has a clockwise rotational slip of 0.0025 rad.

(15),(16) Calculate the value of the redundant chosen for the analys-
is of each of the beams in Figs.2.73 and 2.74 if in the absence of
load support b settles by 3 cm., I = 40000 cm 4 and E = 2000 t/em 2.

(17) Calculate the reactions of the beam in Fig. 2.76 in the absence
of load if support b settles by 4 cm. and EI = 5000 m 2 t.

(18)-(22) For each of the beams in Figs 2 .77-2.81 write down the
deftection equations required to determine a chosen set of redundants due
to some arbitrary support movements in the absence of the applied loads.

(23) Calculate the reaction at c and the force in the member. of the
truss shown in Fig.2.B2 if all the members are made of the same material
and have the same cross-sectional area.

~ L . L

2 6'
Fig. 2.82 Fig. 2.83

(24) Find the force in members be and bd of the truss shown in Fig.
2.83 if all the members are made of the same material and the cross-
sectional area of the diagonals equals double that of the remaining
members.
 297

(25),(26) Find the forces in the members of each of the trus..es shown
in Figs. 2.84 and 2.85 if L/EA is assumed COOOanL_

.- 51
2
51

~~--~--~~--~--~ I
2

6 4

Fig. 2.84 Fig. 2.85

(27),(28) Calculate the forces in the members of each of the trusses

shown in Figs. 2.86 and 2.87 if L/EA is assumed constant.

2 21

1
...
h_-=.9 3 L _f_--I
81 1---- 4x4:16m --~

Fig. 2.26 Fig. 2.87

(29) Calculate the forces in the members of the truss shown in Fig.

,
2.88 if L/EA is assumed constant.

6m
i- r-:;;>IIE--r~o---r---*-.--~--'
L~~~-L~~~~d-~
-+-__ 3X4: 24 _+__ 3x4:24 -=
30t
Fig. 2.88
 298

(30) Calculate the forces in the members of the truss shown in Fig.
2.89 if all the members are made of the same material, and the eross-
sectional area of the diagonals inclined at 45 = 30 em 2 and that of the
remaining members = 50 em 2.

(---~--~f---~--~~~~---1~

I- 6 12m

Fig. 2.89
6Y
(31) Calculate the forces in the members of the continuous truas
shown in Fig. 2.90 if L/EA is assumed constant.

Fig. 2.90

(32)-(35) Calculate the forces in the members of each of the trusses

shown in Figs 2.91 - 2.94 if L/EA is assumed constant.

"I
3
-t
3
41
-t
51 J
--I 2 ~3+- 5m -+-3-1
Fig. 2.91 Fig. 2.92
 299

4 4 41
211 4 4' 1
1.5 2 2 2
-+-
1.5 e 2
+
d 1.5 c d 2
t5
Q b
-,::;:::-_ 6 x 2=12 m,-_oof
1 Q

4x4:16m
1
Fig. 2.93 Fig. 2.94
r
(36),(37) Calculate the forces in the members of each of the trusses
shown in Figs. 2.95 and 2.96 if L/EA is assumed constant.

101

Fig. 2.95 Fig. 2.96

(38),(39) Calculate the forces in the members of each of the trusses

shown in Figs. 2.82 and 2.83 in the absence of load if member ad is fabri-
cated I em. shorter than its geometrical length. L/A = 10 cm- I
and E = 2000 tlem 2.

(40) Calculate the forces in the members of the truss shown in Fig.
2.84 in the absence of load if each of the bottom chord members is fabri-
cated I em. too short. L/A = 12 em- I and E = 2000 t/em 2

(41) Calculate the forces in the members of the truss shown in Fig.
2.88 in the absence of load if:
a) member cd is fabricated 4 em. too short,
b) each of the bottom chord members is fabricated I em. too short.
L/A = 6 cm- I and E = 2000 t/em 2

(42) Calculate the forces in t he members of the truss shown in Fig.

2.91 in the absence of load if each of the diagonals is fabricated 2 em.
too long. L/A = 8 em- I and E = 2000 t/em 2
 3CO

(43),(44) C~lculate the forces in the members of each of the trusses

shown in Figs 2.93 and 2.94 if member ed is fabricated 2 CID. too short.
L/A = 10 em- 1 and E = 2000 tfem".

(45)-(47) C~lculate the forces in the members of each of the trusseo

shown in Figs 2.87-2.89 in the absence of load if the temperature of the
bottom chord drops by 30"C. L/A = 10- 1, E = 2000 tfem" and ... =
I X Hr'rC.

(48)-(50) C~lculate the forces in the members of each of the trusseo

shown in Figs. 2.9:>, 2.95 and 2.96 in the absence of load if the
tem;>erature of the top chord rises by 30"C and that of the bottom chord
droP. by 30"C. LfA = 8 em- 1, E = 2000 t/cm" and ... = I X lo-'rC.

(51) C~lculate the reactions of the continuous truss shown in Fig.2.90

in the absence of load due to :
a) equal settlement at supports band c of 2 em.,
b) settlement of 3 cm. at support c,
c) settlement of 2, 5, I cm. at supports b, c and d respectively.
L/A = 8 em- 1 and E = 2000 tfcm".

(52) C~lculate the forces in the members of the truss shown in Fig. 2.96
in the absence of load due to :
a) settlement of 3 cm. at support b,
b) settlement of 2 cm. at support a,
c) equal settlement at supports a and b of 2 em.
LfA = 8 em- 1 and E = 2000 tfem".

(53)- (62) Sketch two possible main systems for each of the frames
shown in Figs. 2.97-2.106, then using one of the suggested main systems
draw the RM.Ds. if the moment of inertia varies as indicated.

211m 211m
I I I ! 111111111111'-.1

--';;:Cl -+__ 8 m

Fig. 2.97
r' ....... 6

Fig. 2.98
~i
12m_4
I
r
30\

I 2Q251

r
Sm 21
4\
\
t
1
4

~2'::j:::a _ 8m ~ Fig. 2.100

Fig. 2.99
211m 211m
41

r
31
4\

8
t 6 21
5
L J
1-_ _ 12m

Fig. 2.l01 Fig. 2.102

211m 21 2l
41 "1
1 "bo:ri1'i:5il/rmrno::t~ 2
~
E
6 a t
4
~
d 1-%_'_ 8 m --l 2--2- b

1 - - _ 8m _'::::;::::2J
Fig, 2.103 Fig. 2.104
211m
1.5t/m

f
S
c

Fig. 2.105 Fig, 2,106

302

(63)-(70) Sketch two possible main systems for each of the frames
shown in Figs. 2.107-2.114, then using oneof the suggested main
systems
draw the B.M.D s. if the momen t of inertia varies as indicat ed.

211m
I I II I I I , I I I Ii I I
d
c 20m 4
c T

-1-12m~~t
a
12m ----'
Fig. 2.107 Fig. 2.108

J. , 211m
I I I I I I I I I I I Ii I I I

211m
~2
-r
-t4
4
5
t- 1
Il
21 6

L 12m BY L
10 ----I
J
Fig. 2.109 Fig. 2.110

Fig. 2.l ll

211m
11111 11111 11111 11111 11111
1 illl'""1

-
Bm -..-. .-- B
Fig. 2.112
 303

21
21
,
"
t51/m
" 2
Q 21
1 I.

"
1--1. - I - I.rn--J

Fig. 2.113
8m
Fig.2.114
b
2r
(71)-(75) Draw the B.M.D. for each of the frames shown in Figs.
2.115-2.119 if the moment of inertia varies as indicated.

, 21 lOt
--,
~
81 2
21
E
:::
iI. 1
-11
I.
N

"- 8m ---l~t " 6m

Fig. 2.115 Fig. 2.116

3t
I. 11m 211m 211m
~lllllllllillii"

,
1.5 ___

1..5
21

21
~ 2 ~6m-1-- 6. - I { I-- l--- 8m

Fig. 2.117 Fig. 2.118

 304

211
41
3! 1.5!

1
6

L 12m 6 -----i
Fig. 2.[19

(76) a,lculate the reactions of the frame in problem 53 (Fig. 2.97)

in the absence of load due to :
a) a uniform increase in temperature of 300,
b) an increase in the outdoor temperature of 200 accompanied by
a similar drop in the indoor temperature. EI = 2000 m 2t and
a, = 1 X IO-'/"C.

(77) Calculate the reactions of the frame in problem 57 (Fig. 2.101)

in the absence of load due to :
a) a uniform inc:ease in temperature of 300,
b) an increase in the outdoor temperature of 300 e accompanied by a
drop of 100e in the indo~r temperature. The frame has a urjform section
30 X 80 em, E = 200 tlcm 2 and at = 1 X IO-'I"C.

(78)-(81) C,lculate the reactions of each of the frames in problems

61-64 (Figs. 2.105-2.108) in the absence ofload due to a uniform increase
in temperature of 20C. EI = 10000 m 2t and at = I X IO-s/"C.

(82),(83) Calculate the reactions of each of the frames shown in Figs.

2.120 and 2.121 due to a uniform increase in temperature of 200 if:
a) EI = 10000 m 2t, b) EI = 15000 m 21.

L
1.5

r
1..5
I
3
l 6

~6m_-I-_
Fig. 2.120
6
b
L 9m
Fig. 2.121
 305

(84) Calculate the reactions of the frame in problem 55 (Fig. 2.99)

I
in the absence of load due to :
a) a downward movement at support a of 2 cm."
b) an outward movement at support b = I em."
I = 80000 cm' and E = 210 t/cm 2

(85),(86) Calculate the reactions of each of the frames in problems

58 and 6~ (Figs. 2.102 and 2.104) in the absence of load due to :
a) a clockwise rotation at support a of 0.002 fad.,
b) a downward movement at support b of 2.4 cm. EI = 5000 m~t.

(87),(88) Calculate the reactions of each of the frames in problems

62 and 63 (Figs. 2.106 and 2.107) in the absence ofJoad due to :
a) a clockwise rotation at support a 0[0.005 fad.,
b) a clockwise rotation at support a of 0.003 rad. accompanied by
an anti clockwise rotation at support b of 0.003 rad. EI = 10000 m 2t.

(89) C,nsidering the internal forces at hinges c and d as redundants,

write down the deflection equations necessary to determine the effect of
a clockwise rotational slip of 0.0025 rad. at support a of the frame
in problem 74 (Fig. 2.108). Assume EI = 12000 m 2t for all the members.

(93),(91) Calculate the reactions of each of the frames shown in Figs.

2.120 and 2.121 due to the following support movements:
a) a clockwise rotation at support a of 0.005 Tad.,
b) a clockwise rotation at support a of 0.0025 rad., accompanied by
a downward movement of 2.5 em,
c) clockwise rotation at support a of 0.003 fad. accompanied by an
anticlockwise rotation at support b of the same magnitude.
EI = 10000 m 2t.

(92) Find the force in link cc' and calculate the reactions of beam
ab shown in Fig. 2.122 if EI for the beam = 4000 m 2t, area of the link
member = 10 em 2 and E = 2000 tlcm 2.

c ,.
~
r
lom 81 lom
211m
La ~ 1 b a !J
.f:i c M
~ 6m-\.. 3 -I- 3-t I--- 6
c
I 8m

Fig. 2.122 Fig. 2.123

 306

(93) Draw the B.M. and S.F.Ds. for beam ab shown in Fig. 2.123 if
the area oflink cc' = 20 em', I of the beam = 40000 em 4 and E = 2000
tfern' .
If the beam was supported by a rigid support at c, what would be the
percentage change in the maximum negative hendin~ moment?

(94) Find the force in member bd and draw the B.M.D. for beams ab
and cd in Fig. 2.124. EI for each beam = 2500 m't, area of
member bd = 8 em' and E = 2000 tlem'.

"0 b
t 3m
111m ~
dl,<-----l'r-....L-4i
c 3 -l-Jrn-..fI e'
Gm -1-2 -I-2...l
Fig. 2.124 Fig. 2.125

(95) Find the force in link ec' and draw the S.F. and B.M.Ds. for beams
ac and de in Fig. 2.125 ifEI for each beam = 8000 m't, area oflink =
20 cm' and E = 2000 tlem'.

(96) Find the force in the wire rope ce' and draw the S.F. and B.M.Ds.
for beam ab shown in Fig. 2.126 if EI for the beam = 2000 m 't, area of
wire rope = 5 em' and E = 1400 tlcm'.

e
Gm~l
121
Fig. 2.126 Fig. 2.127

(97) A simple beam ab is stiffened by a system of link members as

shown in Fig. 2.127. Calculate the forces in the link members and draw
the N.F., S.F. and B.M.Ds. for the beam. For the link members LIA =
20 em-t. E = 2000 tlem' and for the beam EI = 8000 m't.

r
I (98) Draw the S.F. and B.M.Ds. for beam ab in the structure shown
307

in Fig. 2.128. For the beam EI = 10000 m 2 t, for the link members L/A
= 20 cm- 1 and E = 2000 t/cm 2 .

21
25 25cm. 3m
10

20
10
20
~
3~3m 4m --!-. 2 2j-1
8t 181
Fig. 2.128 Fig. 2.129
(99) Calculate the forces in the truss members and draw the S.F. and
B.M.Ds. for beam ab in the structure shown in Fig. 2.129. Areas of indio
vidual link members are as indicated) I for the beam = 2000 em 4 and
E = 2000 t/cm 2

(100),(101) Draw the N.F., S.F. and B.M.Ds. for each of the beams
in Figs. 2.130 and 2.131 if EI for each beam = 10000 m 2 t, L/A for li~k
members = 15 cm- 1 and E = 2000 t/em 2

211m 211m
1~'I:Lim N~III1.~:VI'~
L2...i--4m--l.. 2 - l - r L2-1-4m -'- 2-l-2J
Fig. 2.130 Fig. 2.131
(102),(103) Draw the N.F., S.F. and B.M.Ds. for each of the frames
in Figs. 2.132 and 2.133 iffor the frame EI = 20000 m 2 tand for the tie
A = 20 cm 2 and E = 2000 t/em 2

1 I ,t ~ ,t/,n; I I I 1 1 1 [I1:1II::JIIIT::JJI=Li/I~mj'IITDIr:1---r
IIJI

l Sm
~-L!."--------<l>t-i
6m

~ 2 I- 6 ~6"'--l 2\- -120--- 8 -I-- 8m ~i

Fig. 2.132 Fig. 2.133
 308

(104),(105) Calculate the vertical deflection and rotation at the

middle of span ab of each of the beams in Figs . 2.70 and 2.78.
EI = 5000 m 2 t.

(106),(107) Calculate the vertical deflection at points m and n in

each of the beams in Figs. 2.73 and 2.74. I = 40000 cm 4 and E = 2000
t/cm 2

(108)-(110) Calculate the vertical deflection at joint d of each of the

trusses in Figs. 2.82, 2.83 and 2.85 ifL/A = 20 cm- 1 and E = 2000 t/cm 2

(11l),(1l2) Calculate the vertical deflection at joint m of each of the

trusses shown in Figs. 2.88 and 2.89 if L /A = 10 cm- 1 and E =
2000 t /cm 2

(l13)-(1l6) Calculate the vertical deflection at the intermediate hinge

c of each of the frames in Figs. 2.105-2.108. EI = 10000 m 2 t.

(1l7),(1l8) Calculate the horizontal displacement at the roof level

of each of the frames in Figs. 2.115 and 2.116. I = 50000 em' and E =
2000 t /em 2

(1l9) If load moves from a to C on the continuous truss shown in Fig.

2.134, construct the influence lines for Y a and Y b and the forces in the
two marked members. LjEA is assumed constant.

Fig. 2.134

(120) Ifload moves on the top chord of the truss shown in Fig. 2.135,
construct the influence lines for Y a, Y b and the forces in the three marked
members. LjEA is assumed constant.

3x4=12

Fig. 2.135
 r
I 309
I
I
I (121) Ifload mmres from a to c on each of the trusses shown in Figs.
2.136 and 2.137, construct the influence lines for the reactions and the
forces in the three marked members. LfEA is assumed constant.
t
i
, I

t---_2x6=12m~_ -I-_ _ 2X6: 1 2 Y

Fig. 2.136

~zS:ZVs:zsA
L3 I 2x6:12 I 2x6:12m 1 3J
Fig. 2.137
I

I (122) If load moves from a to c on the truss shown in Fig. 2.138,

construct the influence lines for the reactions and the forces in the three
marked members. LfEA is assumed constant.

I
~j
Fig. 2.138

(123) If load moves on the top chord of the two-hinged turss shown
in Fig. 2.139, construct the influence lines for the horizontal reaction and
the forces in the three marked members. L/EA is assumed constant.
310

~_ __ 6x3:16m

Fig. 2.139

(124) Ifload moves on the top chord ofthe truss shown in Fig. 2.140,
construct the influence lines for the reactions and the forces in the marked
members. L/EA is assumed constant.

~~~,,-.-~r-~~~~~-'-'

~_-..0.1 3.
--l-3
_ _-l--_3x3=9 ~
Fig. 2.140

(125) Ifload moves on the top chord of the truss shown in Fig. 2.141,
construct the influence lines for the reactions and the forces in the marked
members. L/EA is assumed constant.

~~-.--~-,--~--,-~

1-
-,:;;:...-_ _ _ 6x 4 =24 m
Fig. 2.141

(126) If load moves on the top chord of the truss shown in Fig. 2.142,
construct the influence lines for the forces in the four marked members.
L/EA is assumed constant.
!I

I
311

r fJ\CRVVl}
I Sxl.=2 0m 1
Fig. 2.142

in
(127) If load moves from a to b of the three-c hord truss shown
marked
Fig. 2.143.1 constru ct the influen ce lines for the forces in the three
membe rs. L/EA is assume d constan t.

I 7.5

-.l

~ _____ _____ _____ 6.5=4 0m ----- ----- ----- -1 .

Fig. 2.143

shown
(128) If load moves on the top chord of the three-c hord truss
the three
in Fig. 2.144) construct the influen ce lines for the forces in
marked membe rs. L{EA is assume d constan t.

6x3:2 4m
Fig. 2.144
--- +-1 d
Fig.
(129) If load is transm itted to the continu ous beam shown in
influen ce lines for the
2.145 at panel points spaced at 3 m J construct the
t.
reactions and the S.F. and B.M. at section d. EI = constan
312

a b a c

L Sm -...:;~- 4.5 _001-;_45 3-

Fig. 2.145

(130) Ifload is transmitted to the beam shown in Fig. 2.146 at panel

points spaced at 3 m, construct the influence lines for the reactions Va,
Ma'Yb and Y c and the S.F. at sections d and e. EI - constant.

Fig. 2.146

(131) If load moves from a to d on the frame shown in Fig. 2.147,

construct the influence lines for the reactions and the N.F., S.F. and B.M.
at sections p,b(left), b(right) and b(below). The moment of inertia varies
as indicated.

Ji ' I r 9mi
e __~bT-____-4~P____~Cr-~f
a~____-O
J, ,I '>-----7i~"l
d

L 9 -+4~ 16m
Fig. 2.147
4.5-1- ~ 9

(132) If the frame shown in Fig. 2.148 has constant bending rigidity
EI, construct the inftuence lines for the S.F. and B.M. at sections p,
b(Ieft), b(right) and b(below).

6m
L; J!
I-- 9 -~L:5I--Sm -Il51---- 9 ---l
Fig. 2.148
 313

(133) Ifload is. transmitted to the frame shown in Fig. 2.149 at panel
points spaced at 2 m, construct the influence lines for the reactions, S.F.
and B.M. at section p. The moment of inertia varies as indicated.

a c f d

3 I
I 5m L c'

1-4 -j-L-t-- - 10 m _ --I- 2 -I-- 4 .J

I Fig. 2. 149

(134) Ifload is transmitted to the arch with a tie shown in Fig. 2.150
at panel points a, I, 2J1 3, 4 and b, construct the influence lines for the force

t in the tie and the S.F. and B.M. at section p at the middle ofpanal 1-2.
For arch EI = 40000 mZtand fortieA = 100 ern" andE = 2000 tfem" .

2 3

5 x 4=20m
Fig. 2.150

(135) Ifload is transmitted to the arch with a tie shown in Fig . 2.1 51
at panel points a, 1,2,3,4,5 a nd b, construct the influence lines for the
force in the tie and the S.F. and B.M. at sections p and q a t the middle of
panels a- I and 1-2 respectively. For arch EI = 50000 m't a nd for tie
A = 100 em" and E = 2000 tfem 2
3
--=-r I
2 4
q
t 5
IS.S6m
~
6x4=24m

,t F ig. 2. 151
 314

in
(136)-(139) Load is transm itted to each of the trussed beams shown
infiuen ce
Figs. 2.152-2.155 at panel points spaced at 2m;' constru ct the
10000 m"t
fines for the N.F., S.F. and &M. <it ,ection p. For beam EI ""
and for link u',e~bers L/A = 15. cro- and E = 2000 t/cm
1 2

Fig. 2.152 Fig. 2.153

Fig. 2.154 Fig. 2.155

shown
(140),(141) Load is transm itted to each of the trussed beams
ct the in-
in Figs. 2."156 and 2.157 at panel points spaced at 2m, constru
beam,
fluence lines for the N.F., S.F. and B.M. at sections p and q. For
1 E 2000
EI= 90000 m't and for link membe rs L /A = 10 cm- and
=

t/em 2.

fTA~b
~..j-2-+-l.m--l- Fig. 2.156
" --!-2 +2 J
, ,
~
~m
~ ---f.
O.5B::.,-=:::.::r::==--.JFL--*_I!-L-=:::~I~
1-1_ _ _ 3 X 1.=12 m ---=1 1=-- 2xl.=6 ----l
Fig. 2.157

r
 CHAPTER 3
THE METHOD OF
THE EQ.UATION OF THREE MOMENTS

3.1 Introduction
A continuous beam is one of the most widely used type of statically
indeterminate structures. It is a beam spanning a relatively long opening
with a number of intermediate supports provided in order to reduce the
large moments that would otherwise develop.

There are several methods availabe for the analysis of statically in-
determinate beams. Perhaps. . the most common of these is the method
of the equation of three moments which, as will be shown subsequently,
stems from the method of consistent deformations presented in chapter
2. An equation of three moments establishes the relationship betYfeen
the bending moments at three successive supports of the beam.

For a beam continuous over n simple supports, (n-2) equations of three

moments may be written which, when solved simultaneously, give the
bending moments at all the (n-2) intermediate supports. The bending
moment at an end support is either zero or of a known value and sense
according to whether the end span is simply supported or has an
overhang.
The method of the equation of three moments is applicable to the
analysis of statically indeterminate beams provided that:

(I) the elastic curve of the beam is continuous; i.ew the beam has no
intermediate hinges;,
(2) the moment of inertia is constant within a single span.

If one or both of these conditions are violated, resort is generally made

to other methods.

The great advantage of the method of the equation of three moments

over the method of consistent deformations lies in the fact that there is no
need in this case for drawing the B.M.Ds. due to unit values of the redu-
ndants, determining the displacement coefficients in the superposition
equations and finally solving these equations.
315
316

3.2 Derivation of the equation of three _ _

The equation of three moments is based on the fact that the slope of
the beam at either side of any intermediate support is the same. This
results from the continuity of the elastic curve.

o ~l

Fig. 3.1

Fig. 3.1 represents a continuous beam carrying any system of loads.

An end span may he simply supported, fixed or with an overhang. The
supports will be numbered from left to rigbt hy 0, .,2, 3, .. n, and the
included spans will he designated by Lo .. L I b 1.,], .... ~'I).' The
index of each span refers to its left and right supports respeetiveIy.
In the following derivation, it will he assumed that :
(I) the hending rigidity EI may vary from span to span but remains .
constant within the same span,
(2) the supports do not settle,
(3) the beam has no intermediate binges.

Ca) 1 '2 2 ~J 3

f L'2 f L23 =-=t

1.,2
+L, i 2
1
.l.L
3

r~
L
1 ~

(b) M,
--
3 _-----

Fig. 3.2

Fig. 3.2a shows two successive spans LI2 and L 23 of a continuous

beam. The hending moment diagram, drawn in parts, is shown in
I Fig. 3.2b. The first part represents the moments produced by the applied
loads on a main system chosen as two simply supported beams. The
317

f second part represents the redundant moments. It should be noticed that

a general bending moment diagram which may correspond to any case
of loading is given with no reference to a particular case of loading. Wn
I and W23 represent the areas of the simple beam bending moment
diagrams for spans L12 and L 23 and at and a2 are, as indicated, the dis-

I tances of the centroids of these areas to the outer supports. Also" the
redundant bending moments are assumed positive.

The elastic curve of the part of the beam under consideration

is shown in Fig. 3.2c. From the continuity of the elastic curve, the slope
just to the left and right of support 2 must be the same. This may be
expressed mathematically as: al = a2" Thus,

Yl2 = Y32 3.1

t'2 L23
From the moment-area second theorem,

-
1- ( w Xa + -
M,L"
--X -L'2 +M2L'2
-- - 2 L12)
x- - 3.2
Y12 -
EI'2 12 I 2 3 2 3
1 ( M2L23 2 L23 M3 L23 L23)
Y32 = - EI W23 X a2 + - -2- X -3- + - -2- X ""3 3.3
23

The negative sign of Y32 follows from the sign convention used in the
moment-area method.

Substituting from equations 3.2 and 3.3. into equation 3.1 and simplifying,

M, LI2+2M2(LI2 + L23)+ M3 L23

1'2 1'2 123 123
= -6 (w1'2I2L'2a , + w" a 2)
123 L 23
3.4

A careful study of the right hand side of equation 3.4 shows that it
represents six times the elastic reaction at support 2 if the spans on either
side, both considered as simply supported, were loaded with the M/I-
diagram due to the applied loads on these spans. The sign convention
for the elastic reaction is the same as that used in the elastic-load
method; i.e. the reactiol\:will be considered positive when it acts upwards.
Equation 3.4 may therefore be written as :

23 23
M, L'2
- + 2 M2 (L12
- + -L ) + M3 -L = - 6 r2 3.5
1'2 1'2 123 123
~
I

I
~
318

If the moments of inertia of all the spans are constant; i.e.

112 = 123 = 134 = .... = I, equation 3.5 reduces to
M, LI2 + 2M, (LI2 + L,,) + M, L" ~ - 6 r, 3.6

where f2 in this case is the elastic reaction at support 2 with the simple
bending moment diagram on the spans on both "sides ()f this support as
loading.

Equations 3.4, 3.5 and 3.6 represent three different forms of the equa-
tion of three moments for support 2. Each represents the relationship
between the redundant moment M2 at support 2 and those at the supports
on its left and right . In general, an equation of three moments establishes
the relationship between the bending moments acting over three succes-
sive sUPEorts; and-hence the name.
A sufficient number of equations is obtained by writing equations of
three moments for all the intermediate supports. These equations, when
solved simultaneously, give the unknown values of the moments over the
supports. It should be noticed that at simply supported ends the bending
moments are known to be zero, and if an end span has an overhang, the
moment at the end support is determinate; i.e. it can be found from statics.
Further, a fixed end could be replaced by two supports having a span
between them of zero length. This is justified by the moment-area method
as such replacement is equivalent to the condition of zero slope at the
fixed end. Referring to Fig. 3.3a which shows a part of a continuous
beam with a fixed end, the relationship between M, and M, may be
obtained from the condition that the deflection at support 2 with respect
to a horizontal line through the fixed end is zero. Thus, from the moment-
area second theorem,
I (
Y21 ~ - - w 12
EI
a, + M, 2LI2 2
X - LI2
3
+ M,2LI2 X

W12 al
2 M,L 12 ~ - 6 -- 3.7
LI2
Replacing next the fixed end by two supports as shown in Fig. 3.3b,
and writing down the equation of three moments for support 1,

3.8
,-
f
319

01 2

Lar
M,i- - - - J

'42 1
(a) (b)

Fig. 3.3

Equations 3.7 and 3.8 are identical, and replacing a fixed end by two
supports as suggested is therefore justified .

It remains to mention that in deriving the equation of three moments,

all the moments including those at the supports were assumed positive.
Therefore, the simultaneous solution of the equations automatically gives
the right sign of the moments. If the sign ofa moment comes out negative,
as is usually the case, the moment produces tension in the upper fibers
and compression in the lower fibers, and if the sign of the moment comes
out positive, the moment produces tension in the lower fibers and
compression in the upper fibers.

Once the moments at the supports are determined, it will be an easy

matter to calculate the reactions and hence construct the S.F.D. This is
explained in the following.

Consider any two successive spans of a continuous beam as shown in

Fig. 3.4 a, and let it be required to determine the reaction at the inter-
mediate support 2. Let Y, and Y, be the simple beam reactions at support
2 due to the loads on the spans to its left and right. In addition to these
,I reactions there will be reactions Y'I and Y'r due to the moments at the
supports. Assuming the direction of these moments as indicated in Fig.
3.4 b, then,

Y' = M, -M,
and ,
L"
The total reaction at an intermediate support is thus given by :
Y = (Y, + Y',) + (Y, + Y',)
"!
320

w l~ IPz
I
, I ! i

P
(

(a)
f L12 L23

(b)
G
t
M
l
M2 M3~
Simple beam reactions

l End momt>nt re-actions ~

~ Support shears
Final reactions
Y2t

M2
M
(c) 1
+ +

Fig 3.4

T he first part, (Y I+ V'I), represents the shearing force at a section

just to th e left of the support, while the second part, (Y, +V',), is the
shearing force just to the right of the support.

The B.M.D. may be constructed from the first principles OT, easier
still, by superimposing the free moments on the redundant moments.

Referring again to the two spans shown in Fig. 3.4, and assuming the
directions of the moments at the supports to be as indicated, the free and
the redundant moments drawn separately will be as presented in Fig. 3.40.
These two diagrams, when superimposed, give the final B.M.D. shown
in Fig. 3Ad.
 321

33 IDastrative e:KaD1ples
The following examples will illustrate the method of analysis of
statically indeterminate beams using the equation of three moments and
the principles discussed in the preceding section. For further emphasis,
the procedures are summerised.

(I) A neat sketch of the beam indicating all the applied loads is made. If
an end is fixed, the fixed support is replaced by two simple supports
with zero span between them.

(2) All the supports of the beam including the two supports that replace
a fixed support, if any, are numbered from left to right.

(3) Simple beam B.M.Ds., regarding each span as a separate simply

supported beam, are drawn and the corresponding elastic reactions
at all the intermediate supports are calculated.

(4) The equations of three moments for all the intermediate supports
are formed.

(5) All the equations are solved simultaneously for the unknown moments
at the supports.

For symmetrical beams symmetrically loaded, work in steps 2-5

may be reduced by noting that moments at symmetrically placed
supports are equal.

(6) The final B.M.D. is obtained from the first principles or by superim-
posing the simple beam moments on those due to the calculated
moments at the supports.

(7) The reactions are calculated and the S.F.D. is drawn.

baJnple 3.1 For the continuous beam shown in Fig. 3.5 a, draw the
B.M. and S.F.Ds. The bending rigidity EI is constant throughout the
beam.
Solution: The simple B.M.D. is shown in Fig. 3.5 b. Since EI is constant,
form 3.6 of the equation of three moments may be used and the elastic
reaction is calculated from the simple B.M.D. alone. Thus, referring
to Fig. 3.5 b,

r = 2 (~X !- XWL2) = wL3

I 3 2 8 12
322

wi
(a) ~!!!! ! I I I I I I I
!I!:ttl
I I I I I I I I I I I I I I

L L

( b)

+ +

WL2/8 w~/8

(d)
w W""~
wLl2 wLl2

~wLl8 WLl8t t
3wLl8 5wLl8
"
lOWLl8t

3wLl8 5WL/8~
(e)S.F.D~ I~
~I ~
-----=::.j 5w Ll8 3wl/B
Fig. 3.5

Equation of t h ree m oments fo r support 1,

M"LD' + 2 M, (LD' + L' 2) + M2 L12

M" = M2 = 0 and Lu, = L' 2 = L

The B.M.D. is constructed by d rawing an ordinate M, at support 1

on the upper side and j oining it to the zero ordinates at supports 0 and
 323

3 by straight lines. The simple positive B.M.D. is then hung on these as

base lines. The final B.M.D. is thus given by the hatched area in Fig.
3.5c.

The reactions are calculated with reference to Fig.3.5d.

wL wL 3
Y O =Y2 = - - - = - wL (upward)
2 8 8

(upward)

The S.F.D. shown in Fig. 3.5e may be easily drawn by starting from
support 0 at an ordinate = + 3 wL/8 and decreasing linearly by an
amount of wL to -5wL/8 at support I, and then increasing abruptly
by 10 wL/8 to + 5wL/8 and again decreasing linerly between supports
I and 2 by an amount of wL to - 3wL/8 at support 2.

Exan>ple 3.2 For the continuous beam shown in Fig. 3.6 a, draw the
B.M. and S.F.Ds. The bending rigidity EI is constant throughout the
beam.

Solution: The simple B.M.D. is shown in Fig. 3.6 b. Referring to this

figure,

r =
,
2(~X
2
PL X~)
4 2
= PL2
8

Equation of three moments for support 1,

- 6 r,
Mo = M2 - 0 and L o, = L'2 = L
4 M, L - 6 X PL2/8 , M, = - 3PL/16

The final bending moment is shown in Fig. 3.6 c.

The reactions are calculated with reference to Fig. 3.6 d, and are as
follows:

P I 3PL 5P
Yo = Y 2 = - - -
X16=i6 (upward)
2 L
324

Y = 2(
1
!2 + L~ X 3WL2) = II P
16 8 'upward)

The S.F.D. is drawn in the usual manner and is as shown in Fig. 3.6e.

P P
(a ) 2

L12 _ 4--- lI2 -"'F-- lI2 _ ..l._ lI2 ~~

l L

(b)

+ +

Pl/4 PL/4
3PlI16

(c)B.M.D.f.:;:-_ _ _ _--;;,L-_ _~~----~

5Pl/32 5P1I32
P 3Pl/16 P
(d)
~P-/2------~-----~~
13pn6
5pn6
3PI16t
l1PI16
t

t
lIP/8
lIPI16 llPI16
p!!:.pn~6!!'-----'~i!!;16
(e)SF.QL-__+~_+_------~---_4----~
SPI16 SPI16
l1P/16 llPI16

Fig. 3.6
 325

Es:alDple 3.3 For the continuous beam shown in Fig. 3.7 a, draw the
B.M. and S.F.Ds. The bending rigidity El is constant throughout the
beam. Find the position and value of the maximum positive moment in
each span.
.
w I
X' '1i
(a) I I I I ! I

L L

(b)

wL2/8 will8
3wL"/28 wL"/14
(c)BMD.

wet1l,

~~t~~'~~-
w
(d)
wL
2 2 2
~ 3wL3wLt tWL WL~
28 28 28 26
j
llWL 17wL 15wL 13wL
28 28 28 28

~3~
lw1l28 15w1l28 13w1l28
(e)S.F.D. +

13wll28
17w1l28
Fig. 3.7

Solution : The simple B.M.D. is shown in Fig. 3.7 b. Referring to this

figure,

r = r = r = 2 (~ X !c- X WL2) = wL3

, 2 3 3 2 8 12

Mo = M. = 0 and from symmetry M, = M 3

Therefore, there are two unknown support moments only which may be
obtained by writing the equations of three moments for supports 1
and 2.
326

Equation of three momess for support 1,

Mo Lo, + 2 M, (Lo' + L'2) + M2 L'2 - - 6 r,
4M,L + M2L = - 6 X wL3/12
4 M, + M2 = - wL2/2 ( i)

Equation of three moments fOT support 2:0

M,L'2 + 2 M2 (L'2 + L 2,) + M,L = - 6 r2
"
M, L + 4 M2 L + M, L - 6 X wL'/12
M, + 2 M2 = - wL2/4 (ii)

A simultaneous solution of equations (i) and (ii) leads to :

M, = - 3 wL2/28 and M2 = -- wL2/14

The RM.D. is shown in Fig. 3.7c.

The reactions are calculated with reference to Fig. 3.7 d, and are as
follows :
wL I 3 wL2 II wL
Yo = Y. = - - - X - - = - - (upward)
2 L 28 28
WL
Y, = Y, = ( -
2
+ L-1 3 WL2) (WL 1 WL2)
X - - + - +-X-
28 2 L 28
8 wL
= - - (upward)
7
WL 1 WL2) 13 wL
Y2 = 2 ( - - - X - = - - (upward)
2 L 28 14

The S.F.D. is thus as shown in Fig. 3.7e.

Let the maximum positive moment in span 01 occur at a distance x
from 0 then,
II wL II L
Q=~-wx=O, x-2il 0.393 L

II wL II L W ( II L) 2
M .. = - - X --- - -- = 0.077 wL2
x 28 28 2 28
Let the maximum positive moment in span 12 occur at a distance x
from 1 then,
15wL 15 L
Q = - - - wx = 0 , x = - = 0.536 L
28 28
15 wL 15 L w (15 L)2 3wL2 2
M..ax =~ X 282 28 - 2s = 0.037 wL
 327

Although it is customary to draw the B.M.D. as shown in Fig.3.7c,

yet such a diagram will be oflittle use to designers. A more useful presen-
tation of the B.M.D. will be as shown in Fig. 3.S.

0.107w~
~ oonw'
f"'37~~1'''
I 0.077wL2
1-0.393l.J
1--0.536l--l
Fig. 3.S

Esample 3.4 For the continuous beam of constant EJ shown in Fig. 3.9 a,
draw the B.M. and S.F.Ds .

2tJm
(a)

Sm
Sm

(b)

+
9 8 8ml
12
8

(dB.M.D.

S
8m 9 41
211m
1
SU~6
(d)
S 4

!133 13lt t033 033}

2.67
tL
5
it3
4.S7 ~ --,r--'
13.67 t 7.67 I

~
6.33 6.33 5
(e)S.F.D. + + 1

2.67 2.67 3 3
7.33
Fig. 3.9
328

Solution: The simple B.M.D. is shown in Fig. 3.9 h. Referring to this

figure,

r, 2
= - X 3 X 9
3
+ 6-I [2-X2-
12 ( 4 + 2)
- + 4-X2-
3
12 X 2-
3

r2 = -I [2-X-
12 X 2- X 2 4 X 12 ( 2 4) ] + -- +-

e;
62323

+ ~ [ 8 2)] = 32 m 2t.

Mo = M3 = 0, and there are only two unknowns support moments

which may be determined from the equations of three moments for sup~
ports I and 2.
Equation of three moments for support I,

Mo Lo, + 2 M, (Lo, + L(2) + M'2L" - 6 r,

2 M, X 12 + M2 X 6 = - 6 X 38
4M, + M2 = - 38 (i)

Equation of three moments for support 2,

M, L" + 2 M2 (L" + L 23) + M, L 23 - 6r2
6 M, + 2 M2 X 12 = - 6 X 32
M, + 4 M2 = - 32 (ii)
A simultaneous solution of equations (i) and (ii) leads to :
M, = - 8 mt. and M2 = - 6 mt.
The B.M.D. is shown in Fig. 3.9 c.
The reactions are calculated with reference to Fig. 3.9 d" and are as
follows:
6 X 2
-8 = 4.67 t. (upward)
2 6

( 6 ; 2 + -f) +( 9 : 4 + 8 ;;- 6 )=13.6 (upward)

Y2 ( 9 : 2_ 8 -;;- 6 ) + (4 + +) = 7.67 t. (upward)

6
Y3 = 4 - -
= 3 t. (upward)
6
The S.F.D. is thus as shown in Fig. 3.9 e.
 329

Esample 3.S For the continuous beam of constant EI shown in Fig.

3.10 a, draw the B.M. and S.F.Ds.

211m
(al
6m

(b)

+
a 9 mt
6.33

(cJB.M.D.f-- L
~2~.B:3~~~~='~~____~__~~t=~~
3mt

+ 3.34 3.3m---l 4+3mt 1.5 t

(d l I----'---~ _,
" 4 6
q
633
II
2 tim
II, II " 'p3mt
7.5
10.88 o.aat t055 055+
t312 4.BB 6.55 6.95
.~

11 431
.55t
""",---",3.12 1.5 1.
(elS.F.O. L-..:!+=-+----+---=::::"...,--t---=---l
4.8B 4.881
5.45
Fig. 3.10

Solution: The fixed support is replaced by an additional span 01 of zero

length. The elastic reactions are calculated from the simple beam B.M.D.
shown in Fig. 3.10 b. Note that the diagram for the end span which has
an overhang is drawn in the same way as for a simply supported beam.
All what the overhang has to do with the analysis is to provide a moment
of a known value and sense at the end support. Thus)
 330

M, = 1.5 x 2 = - 3 m.t.
1
r, = - x 4 X 8 = 8 m2t
2 2
I 4x8 I 2
r, = - x - -
2 2 2
x -
3
+-
x6x9 = 26 m 2 t.

Equation of three moments for support I,

M. Lo, + 2 M, (r.;" + L 12) + M, LI2 - 6 r,
2 M, x 4 + 4 M, = - 6 x 8
2M, + M, = - 12 (i)

Equation of three moments for support 2,

M, LI2 + 2 M, (L'2 + L 23) + M3 L21 - 6r,
4 M, + 20 M, - 3 x 6 = - 6 X 26
2 M, + 10 M2 = - 69 (ii)

A simultaneous solution of equations (i) and (ii) leads to :

M, = - 2.83 m.t. and M, = - 6.33 m.t.

The B.M.D. is thus as shown in Fig. 3.10 c.

The reactions are calculated with reference to Fig. 3.10 d, and are
as follows :

8 6.33 - 2.83 =4 -0.88 = 3.12 t. (upward)

2 4

6.33 ~ 2.83) (2 X 6 6.33 - 3 )

4 + 2 + 6

(4 + 0.88) + (6 + 0.55) = 11.43 t. (upward)

2 X 6 6.33 -' 3
Y3 = + 1.5 - = 6.95 t. (upward)
2 6

The S.F.D. is thus as shown in Fig. 3.10 e.

.'

.- - - -- - -
 331

ExaDlple 3.6 For the statically indeterminate beam of constant EI shown

in Fig. 3.1 La, dra.w the .B.M.D.

21

2.5m
II:=::II:o::J'::r~:i/mi:o::IIIfCI:I:q~
15

n
(a)

! 8m 2 J
m
I
(b'fK=:/1 !

12 mt
13m!

(e)

Fig. 3.Il

Solution: The key figure to the solution is shown in Fig. 3.11 b. The
fixed support is replaced by two simple supports with zero span between
them, the simple B.M.D. is drawn, and the bending moment M at support
1.5 X 22
2 is given by : M = 2 X 2.5 - = + 2 m.t.
2
(Note that the determinate moment at support 2 is positive)

rl = '21(2'3 X 8 X 12 ) = 32 m 2 t.

Equation of three moments for support 1,

MoLol + 2 MI (Lol + L 12 ) + M2 LI2 = - 6 rl
2 MI X 8 + 2 X 8 = - 6 X 32, MI = - 13 m .t
The final B.M.D. is thus as shown in Fig. 3.1 I c.
332

Elaunple 3.7 For the statically indeterminate beam of constant EI shown

in Fig. 3.12 a, draw the B.M. and S.F.Ds.

...
2.4 3.6t tS.33 +
4
6.67 t
6.73ml

2.64
1.01
(d)B.M.O.
... ... ...
3.53ml 3.2 4.09 4.64

j)~ t5 f;)G
1.6 7.S
111m
(e)
3.8 3.6 4 3.st 1.25 1.25

~112 1.12t to.95 09?l to.17 017~

4.92 4.95 2.55 1.08 1.0~
P66 --------.r- v
9.871 t f1.l7 1
4.9 .951
-'
2.68
+
(f)S.F.O. +
0.32
1.08
1.92
4.921
Fig. 3.12
rr
333

Solution: For ease of calculating thf elastic reactions and convenience

in checking their values, the simple B.M.D. is drawn in parts as shown
in Fig. 3.12 c rather than that drawn in Fig. 3.12 b although the latter
is quite permissible.
The elastic reactions are calculated with reference to Fig. 3.12 c and
will be as follows :
r, = 9 + 10 + 3.6 + 5.33 27.93 mOt.
r2 = 8 + 6.67 - 3.9 = 10.77 mOt.
Equation of three moments for support I,
MoLo, + 2 M, (Lo, + L 12) + M2 LI2 - 6r,
24 M, + 6 M2 = - 6 X 27.93
4 M, + M2 = - 27.93 (i)
Equation C!f three moments for support 2,

M,L,; + 2 M2 (LI2 + L,,) + M, L 2, = - 6 r2

6 M, + 24 M2 = - 6 X 10.77
M, + 4 M, = - 10.77 (ii)

The simultaneous solution of equations (i) and (ii) leads to :

M, = - 6.73 m.t. and M2 = - 1.01 m.t.
The final B.M.D. is thus as shown in Fig. 3.12 d.
The reactions are calculated with reference to Fig. 3.12 e, and are
as follows :
I X6 + 1.6 6.73
-~::-'-- - - - = 3.8 - 1.12 = 2.68 t. (upward)
2 6
Y, = 3.8 + 1.12 + _1_ (4.5 X 4 +3 X 2) +
6.73 - 1.01
6 6
= 3.8 + 1.12 + + 0.95
4 = 9.87 t. (upward)

Y2 = - I (4.5
2 + 3 X 4) - .
095 - -
X 7.5 1.01
-.L- -
6 6 ' 6
= 3.5 - 0.95 - 1.25 + 0.17 = 1.47 t . (upward)

Y, = -~5- _ -I~l
-= 1.25-0.17 = LOSt.
(
upward)
6 6

The S.F.D. is thus as shown in Fig. 3.12 f.

334

EsaDlple 3.8 For the statically indeterminate beam of constant EI shown

in Fig. 3.13a, draw the B.M. and S.F.Ds.

41 4t
2 tim
(a)
, 1 2
2.5 - -1-- 2.5 m_-1-- 2.5 Ir'

(0)
10mt

+
16,,\1 14m! 16

(e )B.M.D. L-_~""'--r__r ---.----?"-_---l

11.51
8
6.5
(d l5.F.D. L-_ +___--'
.5::=.-..,=c-_ _ _ _--i

11.5
Fig. 3. 13

Solution : Each of the fixed supports is replaced by two simple supports

with zero span between them. There are two unknown moments at sup-
ports 1 and 2 . From symmetry, however, Ml = Ml and one is left with
one unknown only which may be determined from the equation of three
moments for support 1. The simple B.M.D., drawn in parts, is shown
in Fig. 3.13 b. Referring to this figure.

r. = _ 1_ ( 2 .5 +7.5 X 10 + ~ X 7.5 X 14) = 6Om 2t.

223
I Equation of three moments for support I,
335

MoLo, +2 M, (Lo, L 12 ) + + M,L 12 = - 6 r,

22.5 M, = - 6 X 60 , M, = - 16 m.t.
The B.M.D. and the S.F.D. are constructed in the usual manner and
will be as indicated in Figs. 3.13 c and d respectively.

Exatnple 3.9 For the beam shown in Fig. 3.9 a (example 3.4), draw the
B.M. and S.F.Ds. ifthe moment of inertia varies as indicated in Fig.3.14a.

211m
(al
1:'
I ! ! ! !

(bl
+
9
12ml
B.05
6.1Bm!

(el B.M.Q k------/------\----,,,L..--"'----7/

+

Ct
~(;4
B.05ml 41

~6
211m
(d)

6 4

134 to. 31 0. 31 l t1.03 103l

734t f631
------v--
269t t 503
------v---
Z. 97 t
13.651f 7.72! t
6.31 6.31 5.03
+ +
(e )S.F.O. l.:+t...:::~:----I--+-_=_-+---'==j=~_i
2.69 2.69
F
, 7.34
Fig. 3.14
336

Solution : For ease of reference, the simple B.M.D. is re~produced in

Fig. 3.14 h. Since the bending rigidity varies from span to span, form
3.5 of the equation of three moments is used. For simplicity, the relative
ralues of the moments of inertia, rather than their absolute values, are
used. This is quite permissible as multiplying both sides of equation 3.4
by any value I, will not change the equation. Assuming 10 = I, and
referring to Fig. 3.14 b,

2
x 3 X 9 _ 1_ _I [2 X 12 (4 ~)
3 + 1.2 X 6 2 + 3
4 xI2 2 ]
+ 2 X 3"" X 4 = 34.67

I
- 1-.2- X-6- [2;12 X ! x 2+ 4;12 (2++)]

+ +[8 ( 6 ; 2 )] 29.33

Equation of three moments for support 1,

Mo -Lo.- + 2 M. ( -Lo.-+-LI2-) + M 2-LI2-= - 6 r.

10 10 112 112

2M. (6 + _ 6 _ ) 1.2
+ M2 X _ 6_ = _ 6 X 34.67
1.2

22 M. + 5 M2 = - 208 ( i)

Equation of three moments for support 2,

M. ~
112
+ 2 M2 (~+~)+
112 123
M3 L"
123
=-6r2

_ 6 _ M. + 2 M2 ( _ 6 _ + 6) = - 6 X 29.33
1.2 I~

5 M. + 22 M2 = - 176 (ii)

Solving equations (i) and (ii) simultaneously,

M. = - 8.05 m.t. and M2 = - 6.18 m.t.

The final B.M.D. is thus as shown in Fig. 3.14 c.

 337

The reactions are calculated with reference to Fig. 3.14 d and are as
follows :

6 X 2 8.05
-=-.:.:...~ _ - - = 4.66 t. (upward)
2 6

Y, (
6 X
2
2+ S.05 )
6 +
(9 6 X 4
+
S.05 -
6
6.IS )

e:
13.65 t . (upward)

Yz 2 S.05 -;;- 6.1S) + (4 + 6.~S) = 7.72 t. (upward)

6.IS
Y3 = 4 - - - = 2.97 t. (upward)
6

The S.F.D . is thus as shown in Fig. 3.14 e.

Esample 3.10 For the beam shown in Fig. 3.10 a (example 3.5) and
re-produced in Fig. 3.15 a, draw the B.M.D. if I'2 = 16 X 10 4 em'.
and 123 = 20 X 10 4 em4.

(a)
11.1- 2 -1 -2- f- t

--'-;;~~:~'::~:II i:I~I:rDII:II I;I~q....._21_5Jt!

:r:r
l 1:1

(b)

+
+
8mt
9

6.15 mt
3
(c)B.MI.q.::
-:..;,.".;:---:--:;,L-----;-~--__r_-___,L=-.:::::"_I

Fig. 3.15

t
f
 338

Solution: Let 112 ~ I" 112 : 123 = 1 : 1.25. Referring to Fig. 3.15 b,

1 4 X 8
- x -- ~8
2 2

1 X 4-X-8 + -1- (1- X-X

2 22.4
2 2 1.25 2 3

Equation of three- moments for support 1,

2 M\ X 4 + M2 X 4 ~ - 6 X 8

2 M\ + M2 ~ - 12 ( i)

Equation of three moments for support 2,

- 6r2

4 M\ + 2 M2 (~+
1
~
1.25
)- 3 X
6
1.25
- 6 X 22.4

- 30 (ii)

Solving equations (i) and (ii) simultaneously,

M\ ~ - 2.92 m.t. and M2 ~ - 6.15 m.t.

The final B.M.D. is constructed in the usual manner and will be

as shown in Fig. 3.15c.
 339

EsaDlple 3.11 For the continuous beam of constant EI shown in Fig.

316 a, draw the B.M. and S.F.Ds.

( a)

(b)

- - -----
18
"-
-
~

~
"-
I (el "-

I
"- ./
"- ./
+ '-../
18
2 27ml
18 18m!

(d)BMD
+ +
9 9ml
18
9 9
61 6
+ +
(e )S.F.D.

6 6
9 9!

Fig. 3.16

Solution: Since there is an intermediate hinge in span be, the equation

of three moments cannot be applied directly, However, a solution based
on the equation of three moments may be carried out if it is noticed that
the beam can be divided into two stable separate parts as shown
in Fig. 3.16 b. Part dc is statically determinate and the equation of three
moments may be applied to the statically indeterminate beam ab after
 340

the effect of span de on part ad of the beam has been found. From statics,
the action of part de on the rest of the beam = 6 t. (downward)
The bending moment at support b is therefore given by

Mb = - 6 X 3 = - 18 m.t.

The calculations for part ad are as follows :

From the simple beam B.M.D. in Fig. 3.16 c,

Equation of three moments for support a,

MoLoa + 2 M. (Loa + Lab) + MbL' b = - 6ra
2 X 9 M. - 18 X 9 = - 6 X 81
M. = - 18. m.t.

The final B.M.D. and S.F.D. are therefore as shown in Figs. 3.16 d
and c.

(This problem has already been solved using the method of consistent
deformations, example 2.6. The student is advised to compare the relative
ease of the two solutions. Would a solution based on the equation of three
moments be possible if the intermediate hinge was introduced in span ab
instead of span be ?)

3.4 Settlement of supports

The equation of three moments presented in section 3.2 was derived
on the assumption that there was no settlement at any of the supports.
However, such settlement or sinking of the supports is quite possible in
practice and, as pointed out in chapter 2, generally causes a change in the
moment distribution in the beam. For this reason, it is important to find
a means by which the moments resulting from the settlement of one or
more supports may be calculated. One of the main advantages of the
method of equation of three moments is its applicability to the problem
of settlement of supports.

Fig.3 .17a represents any two successive spans in a continuous beam.

Let supports 1, 2 and 3 have settlements 3 1 ,3 2 and 33 respectively; 8
being considered positive when downwards.
 341

(a) 1 2 3
L2J
f L'2
f f
r --- M2

--- --
M,

-- --
~- M)
(b)
;
,f
t
f
t

l
f
r Fig. 3.17

l The elastic curve of the beam is assumed as shown in Fig. 3.17 c. The
unknown support moments are again found from the compatibility

t condition: al a2

Thus, referring to Fig. 3.17 C,

Y'2 + 8, - 82 Yn - 83 + 82 3.9

f
I Dividing the B.M.D. into triangles as shown in Fig. 3.17 b, and applying
! the moment-area second theorem,

I (M' L'2 x -L'2- + M2 L'2 x 2- L'2) ... 3.10

Y'2 = - -
El'2 2 3 2 3

Yn
= __1_ ( M2 L23 X 2L23 + M3 L 23 X L23) ... 3.11
El23 2 3 2 3
The negative sign for Y32 follows from the sign conventions used in the
moment-area method.
Substituting fmm equations 3.10 and 3.11 into equation 3.9 and
simplifying,

6E(8 2 - 8,+ 82 - 83 )
LI2 L 23
... 3.12
 342

When the moment of inertia I is constant for all the spans equation 3.12
becomes:

M,L'2+ 2 M2 (L 12 +L 2,)+M, L 2,=6EI

8-8,. 8-8,)
2 2
( -L'2- + - L23
- ... 3.13

Equations 3.12 and 3.13 represent two different forms of the equation of
three moments for support 2 for the moments induced in a continuous
beam due to settlements of supports.
It should be pointed out that if all the supports settle by the
same amount 8 1 = 82 = 83 .. = 8, the constant term in the
equations of three moments vanish, and consequently all the moments
will have zero value. In other words, if the beam sinks as a whole, no
additional moments will be induced in the beam. Only when there is a
differential settlement among the supports that additional moments
develop.

Ezam.ple 3.12 For the continuous beam shown in Fig. 3. 18 a, draw the
R.M. and S.F.Ds. if support I sinks by 1.5 cm. and the beam has a con-
stant EI 4000 m 2t.

~v_ _ _ _ _ _ _ _~_ _ _ _ _ _ _ _ _ _ _ _~2~______~3 (a)

~ ._ 6 6m --=F-- 6
2.92

k--------:;>.L---------=::""i (b) BM.D

4.6mt

~6 ~92
tC-0-18--0-76! rO-9-"---0.-9!
.
~-0.4-9--0-.4"'9l (e)

----v----
1.721" l. 43 t
kl.76 076 t 0.49 0.49
+ + (d)S.F'D
-
0.941 0.94
Fig. 3.18

.-------------------------
 343

Solution: There afe two unknown support moments. These may be deter-
mined from the two equations of three moments for supports 1 and 2 . .
Since the bending rigidity is constant throughout the beam, form 3.13
of the equation of three moments is used.

Equation of three moments for support 1,

8, - 80 8, - 82)
Mo 1.0, + 2 M, (Lo, + L 12 )+M2 LI2=6 EI ( + .:.!...-~
L o, LI2
1.5 - 0 1.5 - 0)
2 M, (6 + 8) + M2 X 8 = 6 X 4000 ( 600 + 800

28 M, + 8 M2 = 105 (i)
Equation of three moments for support 2,
82 82
M, LI2 + 2 M2 (LI2 + L 23 )+ M, L23=6EI ( - - + - -
-0, -0,)
LI2 L23
0 - 1.5 0-0)
M, X 8 + 2 M2 (8 + 6) =6x4000 ( + 600
800

8 M, + 28 M, = - 45 (ii)

A simultaneous solution of equations (i) and (ii) leads to :

M, = + 4.6 m.t. and M2 = - 2.92 m.t.
The B.M.D. is thus as shown in Fig. 3.18 b.

The reactions are found in the usual manner. Referring to Fig. 3.18 c,
4.6
= - = 0.78 t. (upward)
6

4.6 4.6 + 2.92 - 0.78 - 0.94 = 1.72 t. (downward)

6 8
r + 2.92 + -2.92 =
4.5
!, ---'---
8 6
0.94 + 0.49 1.43 t. (upward)

2.92
Y,= 6
0.49 t. (downward)

The S.F.D. is thus as shown in Fig . 3.18 d.

I
~.
A solution to the same problem, assuming EI = 8000 m 2 t., will lead to :

!
344

M, = + 9.2 m.t. and M2 = - 5.84 m.t.

This shows that the greater the value of EI or, to put it In general
terms, the greater the EI/L ratio for the spans, the greater are the moments
produced in the beam due to differential settlement.

ExalDple 3.13 For the continuous beam shown in Fig. 3.19 a J draw the
B.M. and S.F.Ds. if support 1 settles by 2 em., support 2 by 1 em., and
the end supports remain undisplaccd. The beam is made of steel (E =
2000 t/em 2 ), and its cross-section throughout is a S.LE. No. 38 (I =
24000 cm 4 ).

a 2 3 ( a)

6 8m 6

0..2
k----------:::::o-""""'===~___l (b)B.M.O.
+
2.2 mt

~2tr -:jY (el

0..367 0..367 io.3 a3t ~ 0..0.33
'- - - - - y - - - - '
0..033
'------v----'
0. .667 0..333

0..367 0..367
+ 0..0.33 0..0.33
(d) S.F.O
j
0..3 0..31

Fig. 3.19

Solution: In order that the resulting moments may be expressed in m-t,

units, the bending rigidity of the beam must be in m-t units.

EI = 2000 X 24000 X 10- 4 = 4800 m 2t

f 345

Equation of three moments for support 1.1

MoLo, + 2 M, (Lo, + L 12)+M2L 12 = 6 EI 3, - 30 + 3, L- 32)

(
Lo, '2
2 M, (6 + 8) + M2 X 8 = 6 X 4800 (2 - 0 + 2 - 1)
600 800
~ 7 M, + 2 M2 = 15 (i )

, Equation of three moments for support 2,

I M, L'2 + 2 M2 (L 12 + L 2,) + M,L21 = 6 EI(32 - 3, + 32 - 3,)

L'2 L 2,
I - 2 + I6iiO
6 X 4800 800 - 0)
t M, X 8 + 2 M2 (8 + 6) = (
2 M, + 7 M2 = 3 (ii)
A simultaneous solution of equations (i) and (ii) leads to :

M, = + 2.2 m.t. and M2 = - 0.2 m.t.

The B.M.D. is as shown in Fig. 3.19 h.

The reactions are calculated in the usual manner. Referring to Fig.

3.19 c,

2.2
= -6 = 0.367 t. (upward)

2.2 (2.2 + 0.2) )

Y, = - 6- 8 = -0.367-0.3 = 0.667 t. (downward

_ 2.2 + 0.2 + 0.2 = 0.3 + 0.033 = 0.333 t. (upward)

8 6

0.2
Y, = - - = 0.033 t. (downward)
6

,,
The S.F.D. is as shown in Fig. 3.19 d.

i
!
I-

t
346

Esam.ple 3.14 For the continuous beam shown in Fig. 3.20 a, draw the
B.M. and S.F.Ds. due to the given loads. If support 2 settles by 1.2 cm.,
calculate the percentage increase in the B.M . at support I. EI = 5000 m 2t.

(a}

(b)

12 12 12 12ml
6.1mt 6
(c)B.M.D.
+ + +
6mt
l.St/m (d)
2)
6 6 4

~1.28 U8t t 034

7.28 3.34
0.34ll~ ~t
4.n 2.66 2 6
----v- ---v--

10.62 t t 4.661 t
4.72 t
134 2 2
+ + 0.34 + (e)S.F.Q

2.66 2.66 4 41
7.28
Fig. 3.20

Solution: The simple B.M.D. is shown in Fig. 3.20 b. The elastic reactions
due to the applied loads are calculated with reference to this figure and
are as follows ;

2 I 12 + 4
r, ="3X4X I2 + -i X 2 X 12 =32 +48=80m 2 t.

r, = 48 + (I8 X 2 + 36 X 5) X 1/9= 48 + 24 = 72 m 2 t.

(
,
347

r Effect of applied loads only.

Equation of three moments for support 1,

MoLo, + 2 M, (Lo, + L 12 ) + M2L12 = - 6"

2 M, (8 + 12) + M2 X 12 = - 6 X 80
10 M, + 3 M2 = - 120 (i )

Equation of three moments for support 2,

M,L 12 + 2 M2 (L'2 + L 2,) + M,L 2, = - 6 '2
12 M, + 2 M2 (12 + 9) - 6 X 9 = - 6 X 72
2 M, + 7 M2 = - 63 (ii)

Solving equations (i) and (ii) simultaneously,

M, = - 10.2 m.t. and M2 = - 6.1 m.t.

Effect of settlement in the absence of load.

Equation of three moments for support 1,

2 M, X 20 + 12 M2 = 6 X 5000 ( a ;;0~2)
40 M, + 12 M2 = - 30 (iii)
Equation of three moments for support 2,

1.2 1.2 )
12 M, + 2 X 21 M2 = 6 X 5000 ( 1200 + 900

I 12 M, + 42 M2 = 70
A simultaneous ~olution of equations (iii) and (iv) leads to :
(iv)

l
M, = - 1.36 m.t. and M2 = 2.03 m .t.
1.36
Percentage increase in the moment at support = 10.2
-- X 100 = 13.4%
 348

3.5 Advantages and disadvantages of continuous beaDls

A comparison between the simple beam B.M.Ds. on one hand" and
the final B.M.Ds. of the continuous beams in example 3.l - 3.ll on the
other hand, shows that generally the greatest numerical value of the bend-
ing moment in a continuous beam is less than that for the same beam if
it were composed of a series of simple spans. Since the design is normally
governed by the greatest value of the bending moment along the beam.,
a continuous...,beam system leads to more economical design. Further, the
aver~ge numerical value of the bending moment in a continuous beam

is less than that in the corresponding series of simple beams) and also the
bending moment changes sign along the beam. Since the deflection is
mainly the integrated effect of the bending strains, a continu6us beam
system leads to a stiffer structure.
On the other hand, as illustrated by examples 3.l2, 3.l3 and 3.l4
continuous beams may be subjected to large additional moments if any
support settlement takes place whereas statically determinate beams are
not affected by such support movements. The moment due to settlement
may be large enough to change the sign of the bending moments due to
the given loading in some parts of the beam; a serious matter no doubt
for reinforced concrete structures where the tension side decides where
the main reinforcement is to be placed.
In a loaded continuous beam, two points of zero moment, called points
of inflection or contraflexture, usually occur in each span. If at these

(al

w
(b)

t--- L t, L J L

I" " r' l' r ,",i

I

w
(el
t~~8'8' L'
I-- L I L L ---l
Fig. 3.2l

r ------------------------------------------
 349

points proper hinges are inserted, a statically determinate beam, com-

monly called cantilever or Gerber beam, results. The B.M.D. of this heam
win not be different from that of the corresponding continuous beam.
This is illustrated in Fig. 3.21. Fig. 3.21a shows the B.M.D. of the con-
tinuous beam shown in Fig. 3.21b, and also for the cantilever beam shown
in Fig. 3.2Ic, which is identical to the previous one except in having two
hinges in the central span at the locations of the points of inflection.

3.6 Application of the !Dethod of the eqaation of three !DODlents

to statically indetenninate fralBes

The equation of three moments can be applied to the analysis of rigid-

jointed frames provided that :

(I) joint translation is not allowed,

(2) no more than two members are rigidly connected to a single joint.

Fig. 3.22 shows examples of frames satisfying these two conditions

Note that translation of the joints is generally prevented by virtue of the
disposition of the supports or due to complete symmetry in the frame and
loading.

b c b c
"

(a) (b)

b d

a c' d' f c
(e) (d)

Fig. 3.22

l
350

In Fig. 3.22 a, joint b can translate neither vertically nor horizontally

due to the support conditions of the frame, and twa members only meet
at joint h . In Fig. 3.22 b, neither joint b nor joint c can translate horizon..
tally due to the complete symmetry in both frame and loading. Further,
two members only meet at each of joints band c. In Fig. 3.22 c, ron of
joints b, c, d, and e can translate horizontally due to the complete sym-
metry in both frame and loading. On the other hand, three members
meet at joints c and d. Nevertheless the equation of three moments can
still be applied as members ee' and dd' are pin-connected to. the contin-
uous member bed e. Also, it may be argued that the equation of three
moments is inapplicable to the frame shown in Fig. 3.22d on the basis that
three members are rigidly connected to joint b. A little thought, however,
will show that overhang bd is statically determinate and all it has to do
with the analysis is to. provide a moment of known value and 3ense at
joint h.

The procedure of analysis is essentially the same as that outlined in

section 3.2 in regard to beams. Equations of three moments corresponding
to the number of intermediate joints, a fixed support being consideredas
one, are established. These equations when solved simultaneously, give
the unknown joint moments. Once the joint moments are determined, it
will be an easy matter to calculate the reactions and draw the N.F. and
S.F.Ds. by considering the equilibrium of individual members as free
bodies. The final B.M.D. may be obtained in the same manner but, easier
still, it may be obtained by superimposing the indeterminate moments
to the simple moments.

The application of the equation of three moments to the analysis

of statically indeterminate frames will be illustrated by a number of
examples.
 351

ExalDple 3. IS Usi ng the equation of t h ree mom en ts, const r uct t he

N .F.,S. F . and B.M.Ds . for t he fr a m e shown i n Fig . 3 .23 a. T he rel at ivr
moment s of i nert ia a re as ind icated.

4\
211m

i e

%1
b 21

6m
t
l
I
a b
12mt
I- 6 6m--l
( a) ( b)

I 4t
211m

+ 14 t

r
I
30ml
3t
~ (e) B M.D.
141 (d) Free body diagram
r
14
14 141 + 1 3
3 3t 3
2t

L-1I4 31
( ~) N.F.D. (IlS.F.D.

Fig . 3.23

Solution: There are two unknown moments at joints band c, but

from symmetry Mb = Me and one equation of three moments will be
sufficient to determine the unknown joint moments . The simple B.M.D.
is shown in Fig. 3.23 h. For ease of calculating the elastic reaction, it is
presented in parts.
352

Considering I Ir,

r. = +(~ X 6 X 36 + + X 6 X 12) = 90

Equation of three moments for joint b,

M Lab + 2 M (Lab + L b,) + M L b, =

1 b 1 2 '2
2 Mb (6 + 6) + Mb X 6 = - 6 X 90
Mb = M, = - 18. m.t.

The final B.M.D. may be constructed now and will be as shown in

Fig. 3.23 c. The free body diagram for members ab and bc are shown
in Fig. 3.23 d . From this diagram, the N.F. and S.F.Ds. shown in Figs.
3.23 e and f Can be easily established.

E",aD1ple 3.16 Using the equation of three moments, draw the N.F., S.F.
and B.M.Ds. for the frame shown in Fig. 3.24a ifit has constant EI.

Solution : There are two unknown joint moments Mb and Me- These
may be determined from the equations of three moments for joints band
c. The simple B.M.D. is shown in Fig. 3.24 b. Referring to this figure,
3 6
X
2
+ -32 X 4 X 24 = 9 + 64 = 73

2
- X 4 X 24 = 64
3

Equation of three moments for joint b,

MaLa. + 2 Mb (La. + Lbo) + M,L., - - 6 rb
2 Mb (6 + 8) + 8 M, - 6 X 73
14 Mb + 4 M, = - 219 (i)
Equation of three moments for joint c,
MbLb' + 2 M, (LbO + L,o) + MoL,o
8 M. + 16 M, = - 6 X 64
Mb + 2 M, = - 48 lii)
Solving equations (i) and (ii) simultaneously, Mb = - 1O.25m.t. and
~ = - 18.88 m.t. The final B.M.D. is found by superimposing the
simple bending moments in Fig. 3.24h to the indeterminate moments
 353

311m

c " b~O
9 + t 64
4t + 6ml 24mt

Q
~===- __ 8m
( a) ( b)
18.88
18.88mt
311m
3:12

13.061

0.875

(e) B.M.D.

10.921

10.92t
3.72 3.721
1 . 2
3.72

t 0.28 .72 13.08

10.92 (e) N.F.D. 028 (f) SFD

F ig. 3.24

t bus found. The result is shown in Fig. 3.24 c. The free-body diagram
_ of the frame is shown in Fig. 3.24 d. From this diagram, it will be an easy
matter to construct the N .F. and S.F.Ds. shown in Figs. 3.24 e and f respe-
ctively .
354

Example 3.17 For the frame of constant EI shown in Fig. 3.25 a, calcul-
ate the reactions and draw the B.M.D.

loSt 1 tim 4.51

2~
c d

I
4m
b
LL-l-Q 0_'_ _ 12 m
----1-1~J
Ca)

I
,
+
8.4mt

,1

I
10.5t
(bJ (e) B.M 0

Fig. 3.25

Solution: There are four unknown joint moments :tvl a , Me,M d and Mb
but from symmetry', Ma Mb and Me = Md and one is left with two
=
unknown moments only which may be determined from the equations
of three moments for joints a and c.
Referring to Fig. 3.25 b which shows the simple R.M.D., the elastic reac-
tions are calculated as follows:

rcc = i(1 X 12 X 18)

 355

rc = rei + rer = 3 + 72 = 75 m 2t.

Equation of three moments for joint a,
MoLo. + 2 Ma (Loa + Lac) + MeL.e
2M.x6+6Me~-6x6

2 M a+ M e = - 6 (i)
Equation of three moments for joint c,
MaLae + 2 Me (Lac + Led) + MdLed = - 6 rc
6 Ma + 2 Me (6 + 12) + 12 Me ~ - 6 X 75
M. + 8 Mc = - 75 (ii)
Solving equations (i) and (ii) simultaneously,
M. ~ Mb ~ + 1.8 m.t. and Mc ~ Md = - 9.6 m.t.
The final B.M.D. is obtained in the usual manner and is as shown in
Fig. 3.25 c. Indicated on the same figure are the reactions at support a.

Exam.ple 3.18 Using the equation of three moments, construct the B.M.D.
for the frame shown in Fig. 3.26 a if the moment of inertia varies
as indicated.

Solution: Overhangs eb and cf are statically determinate and all they

have to do with the analysis of the statically indeterminate bent abed is
to provide moments = 2 X 3 = 6 m.t. acting as indicated in Fig. 3.26 h.
These moments may be considered to act either at the ends of member
bc or members ab and cd acco rding to whether the main system con-
sidered is obtained by introducing hinges at the top of members ab and
cd or at the ends of member bc. The corresponding Mo.Ds. are shown
in Figs. 3.26 c and d.
Taking I = It' and considering the main system in Fig. 3.26 c ..
Ta = 0

~( ~ ~
I
r
rb X 6 X 36 -

Equation of three moments for joint a ..

6 X 6) 54

I
,
,
 356

21 2t 2t
2tfm
fe b 21 e

6m 1
a d

1=3 ~

12m
{al
"3J (bl
6
e b 6 e
+ +
36mt 36mt

(el (d)

18

12

"-;)"'"18-6:~'
12
hl +
18 mt

(e l (I) (g)B.M.D.

Fig. 3.26

2Ma +Mb "",O (i)

Equation of three moments for joint h,

Ma Lab + 2 Mb (Lab + LbC) + Mc Lbc = _ 6 rb

lab lab Ibc Ibc

6 Ma + 2 Mb (6 + 6) + Mb X 6 = - 6 X 54
Ma + 5Mb = - 54 (ii)

Solving equations (i) and (ii) simultaneously,

Ma = 6 m.t. and Mb = - 12 m.t.
The final moments at joint b are obtained by co-mbining the indeter-
minate moments thus found and the moments in the chosen main system.
The resulting free-body diagram of joint b will be as shown in Fig. 3.26e.
 357

Considering next the main system in Fig. 3.26 d,

6 X 6 I
ra =-- x - = 6
2 3

= ( 6X6
- 2- 2)
X31+ (2
3 x 6 X 3 6)1
2=84

Equation of three moments for joint a,

Mo Loa + 2 Ma (Lo, + Lab) + Mb Lab = - 6 ra

lOa lOa lab lab

2Ma X6+Mb X 6 = - 6 x 6

2 Ma + Mb = - 6 (iii)

Equation of three moments for joint b,

6 Ma + 2 Mb X 12 + Mb X 6 = - 6 X 84

M. + 5Mb = -84 (iv)

Solving equations (iii) and (iv) simultaneously,

M. = + 6 m.t . and Mb = - 18 m .t.

The final moments at joint b will thus be as shown in Fig. 3.26f,

which are identical to those obtained previously. The final B.M.D. is
obtained in the usual manner and is as shown in Fig. 3.26 g.
358

EXAMPLES TO BE WORKED OUT

(1)-(5) Draw the B.M. and S.F.Ds. for each of the continuous beams
shown in Figs. 3.27-3.31. EI ~ constant.

I--- L. m --t--- 2 --+- 2

Fig. 3.27

211m
Xliii'!'
1..--_ 6m
Fig. 3.28

211m
~IIIII!I

I--- 6m
Fig. 3.29

Fig. 3.30
~ " ,n'~" ,
I - - 6m
Fig. 3.31
it
---'!

(6)-(16) Dnw the B.M. and S.F.Ds. for each of the continuous beams
shown in Figs. 3.32-3.42. EI constant.

1.6 rim

Fig . 3.32

Fig . 3.33
 359

Fig. 3.34

~Q ILt Ii
JLt c 12t
_
I- 3 -L3m~ 3 -l- 3 -'zJ
Fig. 3.35

2-1
Fig. 3.36

1.5 11m
c
-1-_ _ 8 m ---l ~

Fig. 3.37

'II 181 611 161

Qi:~t('1'1 Illb~ c~
l- 2 -+-- L m ---I- 2 -t-- Z -I- 2 -4. 2 -J

Fig. 3.38
1.5 t

%
~
2tl 12t
it O.Bt/m
I I I I I I I I I I I I I ~.-l
Ifm2m

1-2-1-2+2 I Bm 11J

Fig. 3.39

. E:
~
Lomt

~I"I!"I"~
:tt/m

~3 __ 3 I 8m I

Fig.3 .40
 360

Fig. 3.41

Fig. 3.42

(17) The three-span continuous beam shown in Fig. 3.43 has a total
length of 3L. Determine the span lengths L 1, L2 and L} in terms of L
wLI2
so that M. = Mb = M
c
= - -- .
12
EI = constant.

(18) Determine the bending moments at the supports of the contin-

uous beam shown in Fig. 3.44 if: (a) end a is free, (b) end a is simply
supported, and (c) end a is fixed. EI = constant.

Q
lSl/m

L ..
iC
I
!!"Ii

Bm ~I

Fig. 3.44

(19) - (24) Re-solve problems 6,7,8,10,12 and 14 if the moment of inertia

varies from span to span as follows :
Problem 6 (Fig. 3.32), lab: Ibc : Icd = 2 : 1.2 : I
Pwblems 7 and 8 (Figs. 3.33 and 3.34) , lab: Ibc : ICd = 1.5 : 2 : I
Problems 10 and 12 (Figs. 3.36 and 3.38), lab: Ibc = 1.5 : 1
Problem 14 (Fig. 3.40), lab: Ibc : Icd =_ 1 : 2 : I
rr

I 361

I (25) - (31) Determine the bending moments at the supports of the beams
in problems 1,2,6,7,8,9 and 11 in the absence of the applied loads due
to the following support movements:

Problem 1(Fig. 3.27) , Bb 2.4 cm. El = 5000 ill 2t.

Problem 2(Fig. 3.28), Bb 2 cm. EI = 3000 m 2t.
Problem 6(Fig. 3.32), Bb 2 em and Be 0.5 em. EI = 4000m 2t.
Problem 7(Fig. 3.33), Be = 4cm. EI 6000 m 2 t.
Problem 8(Fig. 3.34), Be 3 cm. EI - 5000 m 2 t.
Problem 9(Fig. 3.35), Bb 2.5 cm. El 4000

m 2 t.
Problem II (Fig. 3.37), Bb 3.6 cm. El 4000 m 2 t.

(32) The continuous beam in Fig. 3.41 has uniform I = 9&00 cm 4

and E = 2000 tfcm 2. What will be the percentage change in the max;mum
shearing force and support moment if support c settles by 0.75 em ?

(33)-(43) U sing the equation of three moments, draw the N.F.,S.F.

and B.M.Ds. for the frames shown in Figs. 3.45-3.55. The moment of
i.nertia varies as indicated.

41
L r===::i:1'i:"'mC=II;:J
r
6m
2
CIL
61 21
___11m
61

4 1 1
L L
" - - 4x3 ,, 12m ------l t-.-_ _ 12m

Fig. 3.45 Fig. 3.46

1.5t1m 1.51/m
41 l-
21 IS

I 1
4S

Fig. 3.47
8m

Fig. 3.48
,
r
2r
362

1.5 II m

[i ::::'r:~ :l: ::::I

! 1.51
4 1
L 1.51
I 1.51/m I 311m
~4m - l - 4 -I I-- 3x4=12m - _
Fig. 3.49 Fig. 3.50

lOt

21 21
2t1m 1 6m 1 211m
-.I.
L..-- 9 _-'--_ 3x4:12m I 9--,
Fig. 3.51

2t I.t 4t 41 21

6m
r
8m
L 211m L.!- 1.5 11m

~8m - - I L3-l-- 4x3=12m --+- ~

Fig. 3.52 Fig. 3.53

61

I
3 ,
t 41
1.st/m

3m 3

~2 3d:!t m------J ~ 8m ;f .,-

Fig. 3.54 Fig. 3.55

 CHAPTER 4

THE METHODS OF ELASTIC CENTRE AND COLUMN ANALOGY

Part I
The Method of Elastic Centre

4.1 Introduction
The method of elastic centre which was suggested by the German
engineer Carl Culman stems from the method of consistent deformations
and provides a simplified solution to a special type of structures. Its
application is limited to three times statically indeterminate single span
frames or arches and closed frames or rings. Although it is applicable to all
frames of the abovementioned type, yet it is best suited to the analysis of
frames which have an axis of symmetry such as those shown in Fig. 4.1.

r
t
rI
,-

I,, " (a) " " (b) " " (e) "
I
!

'.i. .>..
(d) (e)
~
(f)

Fig. 4.1

The philosophy 'lfthe method lies in the fact that in general the analysis

I... of frames which are statically indeterminate to the third degree requires
the simultaneous solution of three equations if the method of consistent
deformations is used. Each of these equations generally contains all the
three unknown redundants. The method of elastic centre provides a means

r
I
363
 364

by which the three unknown redundants may be found by individual

solution, rather than by simultaneous solution, of three equations. As
will be seen subsequently, this has the advantage of reducing the number
of displacement components to be calculated in addition to saving in the
time required for the solution of the deflection equations involved.

4.2 Development of the method of elastic centre

The method may be developed with reference to the fixed arch shown
in Fig. 4.2 a.

(ai (b)

Fig. 4.2

A possible main system is shown in Fig. 4.2 b and the reclunrlants ,X,
Y and M are found from the usual deflection equations,
8. = 8,0 + X 8~l + Y 8'2 + M 8., = 0 4.1
Y. = Y,o + X Y.l + Y Y.2 + My., = 0 4.2
a. = a.o +X aaI + Y n a2 + M aa] = 0 4.3
Using the method of virtual work for the determination of the displa-
cement coefficients and noting that the moments produced by unit values
of X, Y and M are in succession,
M l = y,M2 = x and M, = I, equations 4.1, 4.2 and 4.3 may be
re-stated as :

S MlMOdl+XSM~dl
EI EI
-+ ySXYdl + MS y
EI EI
...'!.! = 0 ... 4.4

yS M~dl + MS Xdl
S M2Modi + X {
EI J xydl +
EI EI EI
= 0 ... 4.5

S M'Modl+xS~+
EI EI
YS~+MSMldl=O
EI EI

It may he noticed that dl/EI appears in every term of the deflection

... 4.6

r---------------------
 365

equations 4.4-4.6. Now, if an element dllong is imagined to have a width

equal to I/EI, then dl (I/EI) will represent an imaginary area called
the elastic area. The centroid of this area is called the elastic centre and is
denoted by c. In view of this,

J ~
XYdl = S EI = J ydA = 8. and fXdl JXdA = 8y
xydA =1"" JYdl Ei= r
4.7
Introducing the terms in equation 4.7 into equations 4.4-4.6,

S
M,Modl + X
EI lEI""
J
M!dl + I Y + 8 M = 0 ... 4.8

M 2 M o dl+ 1 x+ySMidl +8 M=O ... 4.9

EI xy EI y
S
M
MSM~EIdl
S M3
EI
O dl + 8 X + 8 Y +
. ,
= 0 . 4.10

The simplification obtained is based on the idea of connecting the assumed

free end of the structure to the elastic centre c by a rigid (undeformable).
arm. Fig. 4.2 c shows such an arm, and also three new redundants X.,
X 2 and X3 acting at c and replacing those at a. Obviously X., X 2 and
Xl are statically equivalent to X, Y and M, and their magnitude should
be such that to keep c in an identical state to that of a. Thus, referring to
the new rectangular axes wi th c as origin,
... 4.11
1'e aeo + XI 1'el + X2 1'e2 + X3 1'e3 = 0 ... 4.12
de.O + XI act + X2 de2 + X3 a e3 = 0 ... 4.13

Equations 4.11-4.13 may be expressed in a manner similar to that used

in equations 4.8-4.10. Thus,

S
M,M O dl + X
EI IJfM!EIdl + I "" X 2
+ 8. X3 = 0 ... 4.14

M 2Mo di + I X + X JM!dl + 8 X = 0 ... 4.15

EI "" I 2 EI y 3
S

.Jf M 3EIM .. dl +8. X ,+ 8 y

X
2 +
X SM~dl
3EI =0 ... 4.16
 366

Since by difinition of the centroid Sx = Sy = 0, equations 4.14, 4.15

and 4.16 reduce to :

MIModl + X SM;dl + I X = 0 ... 4.17

EI lEI xy2
S
M di dl
M2 o + I X + X JMi = 0 ... 4.18
EI xy I 2 EI
S

S M3Modl+ X SMjdl = 0
EI 3 EI
... 4.19

Equation 4.19 yields directly the value of X3 and the values of X 2 and
Xl are obtained from the simultaneous solution of equations 4.17 and
4.18.
If further either axis x or axis y is an axis of symmetry, I:w:.y = 0 and
equations 4.17 and 4.18 reduce to:

MIMOdl + X f M;dl = 0 X =_SMIModI/J M;dl

and
S EI I EI J ' I EI EI

(M2Modl +X (Midi = 0 X =_fM2Mo dl jjMi dl

J EI 2 EI J ' 2 J EI EI

4.3 Applications to the method of elastic centre

Example 4.1 Using the method ofelastic centre, calculate the reactions and
draw the B.M.D for the frame shown in Fig.4.3a (Note that this frame
has been previously analysed by the method of consistent deformations
in example 2.27 (Fig. 2.24).

Solution: The elastic centre will obviously be on the vertical line mid-
way between the two columns. Considering I = If.' the distance y bet-
ween the elastic centre and the top member will be :

y = 2 X 6 X 3 = 2 m.
2 X 6 + 0.5 X 12

A possible main system is obtained by removing the restraints at a. A

rigid arm is attached to the frame at a and extended to c where the redun-
dants XI' X 2 and Xl are assumed to act in the directions shown in Fig.
4.3 b.
I

I
367

I 12mt
I 21
6m I 1
l Q b
"- 12
I...--- 12 m ----l
(a) (b) (c)MoD

n:T'~ '"
~ \:1tLl ~,]j
4 4 6 X2"lt 6
(d)M,.D. (e)M 2D.

Fig. 4.3

Referring to Figs. 4.3 c-f, which show the Mo,D., the MI.D., the M 2 .D.
and the M,.D.,

~ = SMIModl = _I (4 X 6 X 12 _ 2 X 6 X 12) = ?!..

cO EI EI, 2 2 EI,

~c1 SM~t = = E
I 2
- -I J 2 : 6 (4 + 22_ 2 .X 4)+iXI2X 2X2]

72
EI,

~eO 72 .
XI = - - = - - = -It., I.e. It. (to the left)
~el 72

432
= SM2M odl = _1_ (12 X 6 X 6) = -
YeO EI EI , EI,

Ye2 = rM~dl = _ 1_ [2 X 6 X 6 X 6 + ~ X~(62+ 6 2 - 6X6)]

JEI EI, 23
504
EI,

Yeo 432 6 6
X2 = - - = - - = - - t., i.e - t . (downward)
Ye2 504 7 .7
 368

"co = SM 3M"dl =
EI
_1_ (12 x 6 x 1)
EJ.
72
EIr

"c3 = JM~dl EI
= _1_ (2 x 6 x 1 x 1 +! x 12 x 1 Xl=EJ.
EIr 2 )
18

X 3 = - -"co = - -72 = - 4 m .t., I.e.

. 4 m.t. (antIcI--'-'~)
ut,;AWlse
u e] 18
12m!
4!

3.14

1.14
3417
(a) (b) (dB.M.D.

Fig. 4.4

X" X 2 and X. are shown in their proper sense in Fig. 4.4a. Remember-
ing that a force is equivalent to a force and a couple, the redundant reac>-
tion components at a are easily found as follows :
X. = 1 t. (to the left), Y. = 6/7 t. (downward)
and M. = 4 + 1 X 4 - 6/7 X 6 = 2.86 m.t (anticIockwise)
The remaining reaction components may be found from statics and are
as indicated in Fig. 4.4 b. The final B.M.D. is shown in Fig. 4.4c.
Note that this diagram is identical to that previously obtained by the
application of the method of consistent deformations.

r,eu
41
L ~ 2m! 2

J~~l
2m .--- -v-- --,,.-'
T--'-t-- ."C

12 4 4
(a) (b)t.\,.D. (e)M\D

6mt

6 tl
Jlrll 6
6 (d)MxD.

Fig. 4.5
 369

Alternative solution : The frame can he equally solved hy cutting the

top member at a section taken at its mid point, and by assuming that two
rigid arms are attached to the two parts of the frame and extended to the
elastic centre c as shown in Fig. 4.5a. The Mo.D. due to the applied loads
and the B.M.Ds. due to unit values of Xl, X 2 and X3 are shown in sue..
cession in Figs. 4.5 hoe.

Referring to these figures,

3 = JMIModl = _I [(2 X 6 _ '1: X 6) 12] = _ E..

co EI EI, 2 2 EI,

2X 6 2
= JMfdl = _ 1- [ -3- I ,
(4 +22_2 X 4) + 2 x2xI2x2
EI EI, J
72
= -
EI,

3eo 72
XI = - - = - = + It.
3d 72

l Yeo = fM2Modl = _ _ 1- (12 X 6 X 6) = _ 432

ill EI, EI,

Yc2 = f~dl
EI
= _ 1- [2 X 6X6X6+!.. X
EI, 2 3
E.
(6 2 + 6 2 _6 X 6)]

504
EI,

X2 = - -Yeo = 432
- = + 6- t.
Ye2 504 7

72
a = JM1Modl = _ 1- (12 X 6 X I)
cO EI EI , EI,

f M~dl ~ 18
ac3 = = _1- (2 X I X 6 X I + X 12 X I X I = -
EI EIr 2 ) EI,

X, "C"" _ aeo = _..I!.. = - 4 m.t.

e, 18

which are consistent with the values obtained previously.

 370

EsaJnpJe 4.2 Using the method of elastic centre, construct the B.M.D. of
the frame shown in Fig. 4.6a if EI = constant.

31
211m

r
5m
e

1
La 7m

L:~6m~~ii
(a)

31 x=3m
211m
2

32 5
(b) (e)MoD. (d)M,.D.

1ml

3
(e) Mz.D. (t)MJ.D. (g) B.M.D.

Fig. 4.6

Solution: The elastic centre is determined first. This is done with respect
to the beam and the right column as follows:

x = 5.38 X 7 + 6 X 3 _ 55.7 '" 3 m.

5.38 + +
6 7 - 18.38 .
 371
/
y= 5.38 x 2.5 + 7 x 3.5 = 37.97 '" 2 m.
5.38 + 6 + 7 18.38
The frame is cut at a section just to th e right of corner b and two rigid
arms connect the two cut ends of the frame to the elastic centre as shown
in Fig. 4.6 b. The Mo.D. due to the applied loads and the B.M.Ds. due
to unit values of the redundants X h X 2 and X3 are shown in succession

in Figs. 4.6 c-f. The deflection equations will be :

Seo + X I Sci + X 2 Se2 = 0
Yeo + XI Yel + Xl Yc2 = 0
ncO + X3 acl = 0
Using the method of virtual work and referring to Fig. 4.6,

213.6
EI

Sci = SMi dl = _I [5.38 (2 2 + 32 _ 2 X 3) + 6 X 2 X 2

EI EI 3

+ 37- (2 2 + 52 _ 2 X 5) ] = - -
EI
81.1

Sel _ SMIM2 dl =~ [2 X 5.38(3 + ~) 3 X 5.38(3 + ~ X2)

EI EI 2 3 2 3
+3 X6 X 2 _ 3 X6 X 2 + 5 X 7 X 3 _ 2 X 7 X 3]
2 2 2 2
16.4
EI

'Yeo = SM 2 M odl = 2- [ 6 X 5.38 (3 +~ X 2) +~ X 6 X 36

EI EI 2 3 3

(4~ X 3 - 4~ X 3)- 32 X 7 X 3].

=-ru
710

16.4
'Yet = 8c2 = - - from Maxwell's theorem
EI
372

M2dl 1 [5.38 6
Yc2
J _2_",,_
EI EI
- - (3"+5"+3x5)+- (32+32-3X3)
3 3
169
+ 7 X 3 X 3] = -EI-
a
co
= SM M 3 Odl. 1 [5.38 X 6 X
EI EI 2
I
1+.3 X6X36XI+32X7Xl]

312.1
---
EI

a I
c
= SMi dl
EI
= -..r.. (5.38 X 1 X 1 + 6 X 1 X I + 7 X I X 1)
EI
18.38
EI
Substituting these values in the deflection equations,
- 213.6 + 81.1 XI + 16.4 X 2 = 0 (i )
- 710 + 16.4X I + 169X2 = 0 ( ii)
312.1 + 18.38 X3 = 0 (iii)

F rom . ( ... ) X 312.1

equatlOn III ~ 3 = -- - 17 m.t.
18.38
Solving equations (i) and (ii) simultaneously,
XI ~ 1.8 t. and X2 ~ 4 t.
The reactions may now be calculated and the final B.M.D. drawn and
will be as shown in Fig. 4.6g.
I 373

I
l-
Part 2
The Method of Colwnn Analogy
F
! 4.4 Introduction
The cJlumn analogy is a special method of analysis applicable to frames
of the one-Io:>p type and statically indeterminate up to and including the
third degree. Fig. 4.7 shows examples of such frames. The me~hod is

II based on the analogy between stresses in an eccentrically loaded short

column and the redundant moments in the frame, and hence the name.

Fig. 4.7

4.5 DevelopDlent of the Dlethod of colUDlD analogy

The method will be developed with reference to Fig. 4.8.

y
b,p:JITIrr:qc

-A-I<l ---I-- x
d
"
a

(al (bl (e)

Fig. 4.8
374

Fig. 4.8a shows a frame of the type mentio ned in section 4.3, Fig.
4.8b
a possible main system and Fig. 4.8c the same frame but with
a rigidar tn
connec ting the assume d released end d to the elastic centre c
. The red un-
dants are shown a s Xl , X z and Xl"
Denoti ng the momen ts created by the redund ants by Mi (the
sullix i
referring to the indeter minate momen ts), and the momen ts
produc ed by
the applied loads on the chosen main system by Mo. the final
bendin g
momen t at any section of the frame will be given by the algebra
ic sum
of Mo and M i - Thus,

. .. 4.20
where Mi is equal to the algebraic sum of the momen ts due
to X.., X 2
and Xl- Thus, referri ng to Fig. 4.8c, and noting that the momen
ts pro-
duced by unit values ofX 1, X and X3 are in succession Ml
2 = y, M 2 = x
and M, = I,
Mi = XI Y + X 2 X + X, . .. 4.21
A genera l solutio n may be obtaine d by expressing the conditi
ons of
deflection at c. Thus,

S
Mf dl
EI + X2
f M ,M 2 dl
EI

M,M,d l
+ X, EI
= 0 ... 4.22
f

Yc -
-S M 2 M odi
EI + XI
fM 2 M , dl
EI + X 2
S Midi
IiI

M,Mo dl
+ X,
fS M2M ,dl
EI
0
.. . 4.23

a
c --
S EI +X I
fM,M , dl
EI +X 2
M,M 2 dl
EI

dl
+X
J -SMi
p ; [ -- 0 ... 4.24

Remem bering that MI = y, M, = x and M, = I , equatio ns

4.22-4.24
may be re-stated as :

MOYdl +Xfy 2dl+X SXY dl + X SYdl_

'EI - O
EI I EI 2 EI ... 4.25
S
 375

j
2
MoXdl jXYdl dl X JXdl
JX + --0 ... 4.26
- - +X - EI +X2 -EI 'EI -
EI I

S S SdlEI
~
Mo dl Ydl xdl =0 4.27
S EI + XI EI + X 2 EI + X,

reference to
Following the notatio ns introdu ced in section 4.2 and with
the x and y axes,
I =s= oJ~ =sx =0
~= S~
A 'E
S
... 4.28a
EI I " EI

I S
2
X dl = I
EI ' '
SEI
2
y dl = I
x'
SXYd l = I
EI x, '" 4.28b

Also, a little though t will show that:

~ S
Modl = N SMOX dl= M
EI ' EI ' '
SMOY dl= M
EI x
.,. 4.29

I
t of
where N is the total area of the Mo/EI- diagram , Mx the first momen
area about
this area about the x-axis and ~ the first momen t of the
the y-axis.

I Substit uting the values in equatio ns 4.28 and 4.29 in equatio ns

4.25,
I 4.26 and 4.27,
4.30
Mx + Xlix + X2lx, = 0
4.31
M, + XII" + X2l, = 0
4.32
N + X,A = 0
solution
Equati on 4.32 gives the value of Xl directl y and a simulta neous
of equatio ns 4.30 and 4.31 yields the values of XI and X 2. Thus,
Mxl, - M,l x,
X, IXIy - Ixy 2

and
Substit uting for the values of Xl' X 2 and X, into equatio n 4.21
dropping the negative signs,

... 4.33
376

Thus it will be observed that the expression for the indeterminate moments
is analogous to the well known form used for the determination of stresses
In an eccentrically loaded column (equation 8.44, Part 1).
The indeterminate moment M j at any point on the frame is represented
by the stress in an eccentrically loaded section who!e centIC line dimensions
are iden tical to those of the given frame and whose width at any point is
proportional to llEI when loaded by the Mo-diagram. The Mo.D. may
thus be considered as the intensity ofload on the section of the analogous
column.

Having established the analogy between the stresses on the section of

the analogous column and the indeterminate moments in the given frame,
it is obvious that if the frame has an axis of symmetry I",y for the section
of the analogous column will be zero and equation 4.33 reduces to :

N M
+ _Y M,
M j
= x +-y 4.34
A Iy I,

Once the indeterminate moments are found .. the final moments are
obtained by applying the superposition relaticmhip in equalicn 4.20 ..
M = Mo + Mi.
4.6 Sign conventions
On applying the method of column analogy .. specific sign convention
must be adopted. It is felt .. however, that if full advantage is to be taken
of the m:::thod, complex sign convention should be avoided. This could
be simply achieved by considering Mo-moments as downward load on the
section of the analogous c::>lumn and following the sign convention used
in reg ard to the general equation of stress (section 8.11, Part 1) with which
the student is already familiar. Compressive stress will then correspond
to negative Mi (moment causing tension in the outer fibers), and conversly
tensile stress to positive moment (moment cau<jing tension in the inner
fibers of the frame).

4.7 Illustrative exatnples

The method will be illustrated by a number of examples. Before doing
this.. however.. it may be helpful to outline the procedure.
(1) A main system is chosen. The analysis is simplified by choosing a
main system that will provide a relatively less complicated Mo-diagram.
Further, different main systems may be chosen for the various loads on
 377

the frame without affecting the final result. This follows from the principle
of superposition and the fact that main systems are arbitrarily chosen.
(2) The relevant properties of the section of the analogous column are
calculated. On so doing, it should be noted that:
a) the width of the section at any point is proportional to IIEI, and set-
ting EI = 1 always results in simpler calculations,
b) a hinge support on the original frame is represented on the column
section by an infinite area concentrated at the position of the hinge;
i.e. A = co .. and the centroid coinsides with the location of the hinge,

c) if the frame has two hinge supports then. the line joining the correspond-
ing points on the section of the analogous coluri'm is a principal axis.
In such a case, the location of the centroid on this line is immaterial as
the moment of inertia about the other principal axis is infinite.

(3) T~e straining actions N, Mx and My corresponding to the Mo-

diagram a; load on the section of the analogous column are calculated
with reference to two rectangular axis through the centroid. Mo is consid-
ered as downward load on the section when positive and vice versa.

(4) Ignoring any sign conventions apart from those used in conjunc-
tion with the general equation of stress, the indeterminate moments are
found by applying either equation 4.33 or 4.34, _re-produced below.

N My M,
=-+ -x + -y
A Iy I,

II- The later equation is used only when the centroidal axes are principal
I axes.

In these equations compressive stresses indicate negative bending moments

F.
and tensile stresses positive bending moments.

(5) The final moments at the various points of the frame are obtained
from equation 4.20; M = Mo + M;.

I
378

Esample 4.3 Using the method of column analogy, construct the B.M.D.
for the frame shown in Fig. 4.9a. The relative moments of inertia are as
indicated.

tim 9
r e

~:
b 21 e e
6 1
1 a d
9
"
f---.. 12 m "3...J
(al (b) Mo.D.

l
9=2 N,=144 M =324 12.35
+
Nt'54
I EI=l
7.15ml
(el (d) 8.M.D.
Fig. 4.9

Solution: For ease of calculations, two different main systems are chosen
correspondi ng to two load systems; beam be and cantilever ecd loaded
seperately. The first main system is obtained by replacing the fixed supports
at a and d by simple supports and the second is obtained by removing the
fixed support at a. The resulting Mo-diagram is shown in Fig. 4.9b. The
section of the analogous column, assuming EI = 1, is shown in Fig. 4.9c.

Properties of the section of the analogous column,

A = 2 ( I X 6) + 12 X 1/2 = 18

Since the section is symmetrical about the y.axis the centroid lies on it at
a distance y from the upper edge.

2 XI X6 X 3
y= = 2
18
2
I = 2 -
,
6 X-
( 12
6 +6 X 12) +6 X 22 = 72
 379
6 X 122
Iy = 12 + 2 (6 X 6 2) = 504

Straining actions..

Nt = G X 12 X 36)i= 144

N2 = (9 X 6) I = 54
N = 144 + 54 = 198 ~
M, - 144 X 2- 54 X I = 234
.,.t
M" - 54 X 6 = 324 -+ +

The straining actions are indicated on the section of the analogous column
Fig. 4.9c.
III

Indeterminate moments,

Since the centroidal axes are principal axes, equation 4.34 is used" and,
the signs of the various terms are found by inspection.

M. = -
198
18 + -n
234x4
+
324x6
504 = - II + 13 +3.S5 = + 5.S5 m.t.

, 198
- -
IS
234 x 2 324x6
-- + -- = -
72 504
II - 6.5 +3.S5 = -13.65 m.t.

! 19S 234 x 2
-IS - -72- - - -=
324 x 6
- II - 6.5 -3.85 = -21.35 m .t.
504
t
198 234 x 4 324 x 6 .
M. = - - + ----- = - II + 13 -3.S5 = -1.85 m .t.
18 72 504

t
~-
Final moments,
M. = 0 + 5.S5 = 5.S5 m .t.
Mb = 0 - 13.65 = - 13.65 m .t.
M e (left) = 0- 21.35 =- 21.35 m .t.
Me (below) = + 9 - 21. 35 = - 12.35 m. t.
M. = 9- I.S5 = 7.15 m.t .

The final B.M.D. is thus as shown in Fig. 4.9 d.

380

Example 4.4 Using the method of column analogy, calculate the bending
moments at joints a; b, c, d, e, and (. and hence construct the B.M .D . for
the frame shown in Fj,!f'. 4.10a . The relative moments of inertia are as
indicated . ..

2.4 tim
I I I I ! II I I I II I ! II I
,-- d
41
e
3m c 41
t-
5
b
41 21 43.2

ta f
I---" 6 ---+- 6 ---' "
( a)

M,=279.2 114
N :43.2
--
Y
114
:43.2 N2=32.4
x

EI=1 112
1-. =5.53 532mt 11.91
(c) (d) B.M.D.

Fig. 4.\0

Solution: A main system is obtained by replacing the fixed supports at

a and f by simple supports. The resulting Mo-diagram is shown in Fig.
4. 10h . T he section of t he analogous column, assuming EI = 1, is shown
in Fig. 4.\Oc.
Properties of the section of the analogous column,
A = 1 X 5 + 0.25 X 6 + 0.25 X 3 + 0.25 X 6 + 0.5 X 8
= 5 + 1.5 + 0.75 + 1.5 + 4 = 12.75
The centroid is located with respect to two rectangular axes through a.
_
x =
1.5 X 3 + 0.75 X6 + 1.5 X 9 +4 X 12
= 5.53
12.75
_
Y =
5 X 2.5 + 1.5 X 5 +
0.75 X 6.5 + 1.5 X B + 4 X4
= 4.15
12.75
! I, = C~/2 + 5 X 1.65 2 ) + 1.5 X 0.S5 2
2
+C7~;32+0.75X2.352)
3S1

+ 1.5 X 3.S5 2 + (47: + 4' x 0.15 2) = 73.21

2 2
Iy = 5 X 5.53 2 + 1.5 X 6 +J.5 X 2.53 2) +0.75X0.47 2+ (1.5X6

I
( 12 12

+ 1.5 X 3.47 2 ) + 4 X 6.47 2 i" 357.00

f I,y = 5 (- 5.53) (- 1.65) + 1.5 ( - 2.53) (0.S5) + 0.75 (0.47) (2.35)

I + 1.5 (3.47) (3.S5) + 4 (6.47) (- 0.15) = 59.10

Straining actions,
I Nt = (2/3 X 6 X 43.2) 1/4 = 43.2
N2 = (43.2 X 3) 1/4 = 32.4
N3 = Nt = 43.2
N = 43.2 + 32.4 + 43.2 = lIS.S ~

From symmetry, the resultant of N 1, N2 and N J coinsides with the

location of N 2 - Hence. .
I M,. = IIS.S X 2.35 = 279.2 f
!
M" = lIS.S X 0.47 = 55.S_

~
, Indeterminate moments,
Since x and yare not principal axes, equation 4.33 has to be used.
In this case, the sign conventions should be strictly followed.
N = - IIS.S, M,. = - 279.2 and M" = - 55.S
M _ ..:'!. + M"I, - M,I,y x + M.I y - MyI,y Y
;- A I xy
I - I xy 2 I xy
I - I xy 2
" lIS.S (-55.S)(73.21)- (-279.2)(59.1) (-279.2)(357) -(-55.S)(59.1)
t
l
= - 12.75+ 73.21 X 357 - 59.12 x + 73.21 X 357 _ 59.12 Y
i
= -9.32 + 0.55 x-4.26 Y
i
I' M. = - 9.32 + 0.55 (- 5.53) - 4.26 (- 4.15) = 5.32 m.t.
Mb = - 9.32 + 0.55 (- 5.53) - 4.26 (0.S5) = - 15.9S m.t.
Me 9 - =.32 + 0.55 (0.47) - 4.26 (0.S5) = - 12.6S m.t.
382

Md =- 9.32 + 0.55 (0.47) - 4.26 (3 .85) = - 25.46 m.t.

M. = - 9.32 + 0.55 (6.47) - 4.26 (3.85) = - 22.17 m.t.
M. = - 9.32 + 0.55 (6.47) - 4.26 (- 4.15) = 11.91 m.t.

Final moments,
The final moments at joints a,b, e and f are the same as the indeterminate
moments since the Mo-moments at these joints are zero.
M, = 43.2 - 12.68 = 30.58 m.t.
Md = 43.2 - 25.46 = 17.74 m.t.

The final B.M .D. is thus as shown in Fig. 4.lOd.

Exantple 4.5 Resolve problem 4.4 if supports a and f are hinged.

1/t.

lIt.
blr-;~?==!.. c
N,
EI =I 112

A=oo
(a) (b) 8.M.D.
Fig. 4.11

Solution: The main system chosen in the previous example is used here.
The section of the analogo~s column is shown in Fig. 4.11a. It is noted
that infinite areas are concentrated at the locations of the hinge supports.
Thus, the centroid lies on the line connecting a and f which is taken as the
x-axis, and it is obvious that Iy = co and IXY = O.

Ix = C7;2+ 5 X 2.5 2) + 1.5 X 52 + (.75 :

1
2
3 + 0.75 X 6.5 2)
2
2
+ 1.5 X 8 + (47: + 8 X 42) = 292.8

N My Mx
= - + -x + -y
A Iy Ix
Mx
Since A = Iy = co = -y
I,
 383

Making use of the values of N calculated in the previous example,

.,.
Mx= 118.8 X 6.5 = 770 t
770 x 5
M,=O, Mb= - -- = - 13.15m.t., M, = +43 .2- 13.15=30.05 m .t.
292.8

. 770 X 8 770 X 8
Md = 43.2 - =22.2m.t., M,= - - 2Im.t., Mr= O.
292.8 292 .8

The final B.M.D . is thus as shown in Fig. 4.llb.

Exatnple 4.6 Using the method of column analogy, construct the B.M.D.
for the frame shown in Fig. 4. 12a . The relative moments of inertia are as
indicated.

4t1m
41

1 6m
d 1.
12m!
>--_ _ 12m
( a)

2. 31
(c) ( d ) B.M.D
Fig. 4.12

Solution: Two different main systems, corresponding to the concentrated

and uniformly distributed load systems.. are considered . The first is
obtained by removing the hinge support at d and the second by replacing
the fixed support at a by a roller. The resulting Mo-diagram is shown in
Fig. 4.12b.
Properties of the section of the analogous column,
384
there
Since suppor t d is hinged , an infinite area will be concen trated
co The ~cction of the analo-
and hence the centroid coinsides with d. A =
momen ts of
gous column , assumi ng EI = 1, is shown in Fig. 4.12c. The
throug h d
inertia and produc t of inertia with referen ce to x and y axes
arc to be found next.

Ix = C_5_X_0_;_~_X_7_.5_2 + 3.75 X 2.25 2) + 12 X 0.25 X 6 2

+ (I X6 X6
12
2
+6 X 32) = 216.7
.

X 122 )
Iy = 3.75 X 122 + (3 12 +3 X 6 2 = 684

Ia. = 3.75 (- 12) (2.25) + 3 (- 6) (6) + 6 to) (3) = - 202.5

Straini ng actions ,

N, = ( 15 X 7.5) I
2 2= 28.125 t
N2 = ( 2/3 X 12 X 72) 1/4 = 144 ~
...t
Mx = 28.125 + 144 X 6 = 835.87 5
My = 144 X 6- 28.125 X 12 = 526.5 ~+-

Indeter minate momen ts,

_ N MyI" --- M"I XY MXIy - MyIXY

M,_ + 2 x + 2 Y
A I"Iy --- IXY I"Iy --- IXY

(526.5) (216. 7) ~ (- 835.87 5) ( - 202.5) x

o+
216.7 X 684 - (- 202.5) 2
(- 835.87 5) (684) -(526 .5) (- 202.5) y - -0.51 5x-4. 338y
+ 216.7 X 684 - (- 202.5) 2
M. = - 0.515( - 12)-4 .338(- 1.5) - 12.687 m.t
Mb = - 0.515{ - 12)-4. 338(6 ) - -19 .848 m.t
Me = - 0.515(0 ) -4.33 8 (6) - - 26.028 m.t

Final momen ts,

M.= - 15 + 12.687 - - 2.313m .t
Mb= 0- 19.848 - -19.84 8m.t
Me = 0 - 26.028 - - 26.028 m.t
The final B.M.D . is thus as shown in Fig. 4.12d.
 385

I!.aJnple 4.7 Using the method of column analogy, construct the B.M.D.
for the frame shown in Fig. 4~13a. The moment of inertia varies as
indicated.

21 e

3m
21

31
14m
1 a
x
L 3 _____ 3m
(a)
(b) ""00
M - 6.57
.!:
I
58 Y
Tl2

=14.22

Ib9
:1

(e) (d)BM.O.

Fig. 4.13

Solution: A convenient main system is obtained by releasing the right

support. The resulting Mo.D. is thus as shown in Fig. 4.13b. The section
of the analogous column assuming EI = I is shown in Fig. 4.13c. Refer-
ring to this figure,
A = 1 X 3 + 3 X 1/2 + 5 X 1/3 = 3 + 1.5 + 1.67 = 6.17

1.5 X 1.5 + 1.67 X 4.5

X= = 1.58 m.
6.17

I:- y =
3 X 1.5 + 1.67 X 2
1.27 m.
6.17

Before proceeding any further" it will be helpful to discuss a particular

problem which is encountered quite often. This is the evaluation of the
moments and pr.oduct of inertia of a long and narrow rectangular section
with respect to axes which are at an angle with the rectangle. '
r y

/-l~-'---. v

Fig. 4.14

Referring to Fig. 4.14,

L/2

I. 2J = b dl (I sin 8)' = 2b sin2 8 L 3 /24

o

Av'
1=- ." i-.3:;
12
L/2

:r,. = 2J bdl (I cos 8)' = 2b cos' 8 L 3/24

o
Ah'
:r,.=- 12
.. 4.36

L/2

I., 2J = b dl (I sin 8) (I cos 8) = 2 b sin 8 cos 8 L 3 /24

o
Avh
I.,. =- 4.37
12

Applying equations 4.35, 4.36 and 4.37, the moments and product of
inertia of the section of the analogous col umn are as follows :
.
 387

3 X 32 1.67 X 4 2
= + 3 X 0.23 2 + 1.5 X 1.272 + --,--
12 12
+ 1.67 X 0.73 2 = 7.94

1.5 X 32 2 1.67 X 3 2
1,. = 3 X 1.58 2 + + 1.5 X 0.08 + - --
12 12
+ 1.67 X 2.92' = 24.12
1.67 X 3 X 4
I.,. = 3 (- 0.23) (- 1.58)+ 1.5 (- 0.08) (1.27)
12
+ 1.67 (2.92) (- 0.73) = - 4.29

Straining actions,
3 X 6
N - - x l = 9t..+9
2
+
- 9 X 0.73 = 6.57 t. - 6.57

= 9 X 1.58 = 14.22_... -14.22

Indeterminate moments,

9 (-14.22) (7.9'1) - (- 6.57) (- 4.29)

=-+
9.17 7.94 X 24.12 - (- 4.29)"
x

(- 6.57) (24.12) -(-14.22) (- 4.29)

+ 7.94 X 24.12-(-4.29)2 Y
+ 1.4586 - 0.8152 x - 1.2681 Y

Mia = 1.4586 - 0.8152 (- 1.58) - 1.2681 ( - 1.73) = 4.94 m.t.

Mlb = 1.4586 - 0.8152 ( - 1.58) - 1.2681 (1.27) = 1.14 m.t.
Mi. = 1.4586 - 0.8152 (1.42) - 1.2681 (1.27) =- 1.31 m.t.
M .. = 1.458'; - 0.8152 (4.42) - 1.2681 ( - 2.73) = 1.31 m.t.

F.na1 moments,
The final moments at all the joints are the same as the indeterminate
moments except at a where M. = - 6 + 4.94 =- 1.06 m.t. The final
B.M.D is thus as shown in Fig. 4.13 d.

I
388

&alllple 4.8 The frame shown in Fig. 4.1Sa has stepped columns. Draw
the B.M.D. for the frame under the given crane loads if the moment of
inertia varies as indicated.

d
31
9 mt

1--_12m _ _ 9 3
(al (bl"'\,.D.

,,\,407
3EI,11 y .~~____~==~!J8mt

2.06
.09

0.54 3.4
(el (dl B.M.D.
Fig. 4.15

SoIutioll : A convenient main system is obtained by cutting member cd

at any section along its length. The resulting Mo.D. is thus as shown in
Fig.4.15b. The section of the analogous column assuming 3EI = I is
shown in Fig. 4.15c. Referring to this figure,
A = 12 X I + 2 (3 X 3 +2 X 6) = 54
2 (3 X 3 X 1.5 +6 X 2 X 6)
Y = 54. = 3.17

3 X 33 2 X 63
I, = (12 X I) 3.172 + 2 ( '--'1"-2- + 9 X 1.67 2 + 12

+ 12 X 2.83 2) = 448

I X 12 3
=----+
12
2 (9 X 62 + 12 X 6 2 ) = 1656
 389

Straining actions,
NI = (3 X 6 X 2) = + 36
Nz = (9 X 6 X 2) = + 108
N = 36 + 108 = + 144
M. = (36 + 108) 2.83 = 407.f
My = (108 - 36) 6 = 432-++

Indeterminate moments,
N M. My 144 407 432
M; = A 1. Y I; x = + 54 448 Y 1656 x

M; = 2.67 0.9085 Y 0.2609 x

M. = 2.67 + 0.9085 (5.83) + 0.2609 (6) = 9.54 m.t.

M, = 2,67 - 0.9085 (0.17) + 0.2609 (6) = 4.09 m.t .

. = 2.67 - 0.9085 (3.17) + 0.2609 (6) = 1.36 m.t.

Md = 2.67 - 0.9085 (3.17) - 0.2609 (6) = - 1.78 m.t.

Mr = 2.67 - 0.9085 (0.17) - 0.2609 (6) = 0.94 m.t.

Mb = 2.67 + 0.9085 (5.83) - 0.2609 (6) = 6.40 m.t.

Final moments,
M, = 9.54 - 9.00 = 0.54 m.t.
M. (belo:w) = 4.09 - 9.00 =- 4.91 m.t.

M. (above) = 4.09 m.t.

Me = 1.36 m.t.

Md =- 1.78 m.t.

Mr (above) = 0.94 m.t.

Mr (below) = 0.94 - 3.00 = - 2.06 m.t.

Mb = 6.40 - 3 = 3.40 m.t.

The final B.M.D. is thus as shown in Fig. 4.15 d.
,

I.
I

390

Ezaxnple 4.9 Using the method of column analogy, construct the B.M.D.
for the frame shown in Fig. 4.16a. The moment of inertia varies
as indicated.

31 b
1.21

(c) (d)B.M.D.

Fig. 4.16

Solution : A convenient main system is obtained by releasing support d.

The resulting Mo.D. is thus as shown in Fig. 4.16b. The section of the
analogous column assuming EI = I is shown in Fig. 4.16c. Referring
to this figure,

5
A = 2 (5 X I) + 6 X - = 15
6
2 (5x I x2) 4
y
15 3

2 2
5 X 6 [5 X 3 2]
= 12 + 2 ---'-:-1-2:"" +5 (4.5) = 225
,-

Straining actions,

N =+ (
12 X
2
5) 1=+30
 391

M,: = 30 x 5 = 150 -++

Indeterminate moments"

M, = ~ M. y M,x
1\ Ix If
30 40 X 3 150
= 15 80 Y 225 x
= 2 1.5y 0.66 x

M. = 2 + 1.5( !) + 0.66 (6) + 10 m.t.

Mb 2 - 1.5( : ) + 0.66 (3) = + 2 m.t.

Me = 2 - 1.5( ; ) - 0.66 (3) - - 2 m.t.

M. = 2 + l.5( :) - 0.66 (6) - +2 m.t.

Final moments,

The final moments of all the joints are the same as the indeterminate
moments except at joint a where M. = - 12 + \0 = - 2 m.t. The
final B.M.D. is thus as shown in Fig. 4.l6d.
392

Ezample 4.10 Using the method of column analogy, construct the B.M.D.
for the frame of constant EI shown in Fig. 4.17a.

211m
1D ::ITII:JIITIII~i:r:TIID
LCi1: 64mt
3m 64'---~~~--~16~
fa e 8
6
b &4 48
.....Lam --1-4-1
(a) 'b) M. D

fM x=4852.8
41.39 ml
Y x My =768
2f.2~5~m~~~~~~---
..-...., . .
r
3.75
+-.
I N2:384
I-
N,=384
6t6m-l N31

EI=l
25.54 30.69
(e) (d) B.M.D.

Fig. 4.17

Solution: A main system is obtained by disconnecting the hinge at c.

The resulting Mo.D. is shown in Fig. 4.17b. The section of the analogous
column and the corresponding straining actions is shown in Fig. 4.17c.
It should be noted that since an infinite area is concentrated at the loca-
tion of the hinge" the centroid coinsides with c. Also, since"the indetermin-
ate bent is symmetrical, x and yare the principal axes; IXY = O.

Properties of the section of the analogous column,

A
I X 63 32
2 ---
( 12
+6 X 62 + 8.55212X + 8.55 X
)
1.5 2 = 519.2

2
Iy = 2 (6 X 82 + 8.55 12X 8
+ 8.55 X 4
2) 1132.4

Straining actions"
NJ = (64 X 6) = 384
 393

N2 = N3 = (0.33 X 64 X 8.55) 1 = 182.4

N. = (48 X 6) 1 = 288
M. = (384 + 288) 6 + 2 X 182.4 X 2.25 = 4852.8t
M, = (384 - 288) 8 = 768-++
Indeterminate moments,
M. M, 4852.8 768
M, = -Ix y + -Iy x =- -y +-
519.2
- x = 9.346 Y + 0.678 x
1132.4
M. = 9.346 X 9 + 0.678 X 8 = 89.54m.t.
Md = 9.346 X 3 + 0.678 X 8 = 33.46 m.t.
M, = 9.346 X 3- 0.678 X 8 = 22.61 m.t.
M,. = 9.346 X 9- 0.678 X 8 = 78.69 m.t.
Final moments,
M. = - 64 +
89.54 = 25.54 m.t.
Md = - 64 +
33.46 = - 30.54 m.t.
M, (above) = - 64 + 22.61 = - 41.39 m.t.
+
M. (below) = - 48 22.61 = - 25.39 m.t.
Mb = - 48 +
78.69 = 30.69 ITLt.

The final B.M.D. is thus as shown in Fig. 4.17d.

4.8 Application of the D1ethod of colUIDD analogy to statically

iadeterDlinate beams
The method of column analogy is applicable to the analysis of single
span beams of the type shown in Fig. 4.18 whether they are prismatic or
nonprismatic.

Fig. 4.18

The procedure of analysis is essentially the same as that outlined in

section 4.4 in regard to frames. In this case, however, the section of the
analogous column is a narrow strip of zero moment of inertia about the
x~a](is,
and hence equation 4.33 reduces to :
N M,
M, = A+lx
y
... 4.38
and the final moment is again determined from the superposition
equation, M = ~ M, +
394

E>cunple 4.11 Determine the end moments for the beam shown in Fig.
4.19a if the moment of inertia varies as indicated.

(a) ~~a 1--::2!-r------:

-;--1-';-- II~b~
LJ4 -l-LJ4 -+--LJ4 --t- LJ4

(b)"\,.D.

PU4

(e)

Fig. 4.19

Solution: A possible main system is obtained by replacing the fixed ends

by simple supports. The corresponding Mo.D. is shown in Fig. 4.19b.
Assuming EI = I, the section of the analogous column will be as shown
in Fig.4.16c.

Referring to this figure, A = 2 (0.25 L X I + 0.25 L X 0.5) = 0.75 L

I PL L ) (PL PL L) I] 5 PL2
N =2 [( "2Xa x 4" 1+ 4" x SX"4 ="64.1. "2
Since here My = 0, the moments at the fixed ends are given by :
5PL2/ 5 PL
M. =Mb=-M 0.75L=-~

which are the final moments since the Mo-moments at points a and bare
both zero.

It should be remembered that the main system considered was arbitra-

rily chosen. Any other main system would have yielded the same result .~

for the final moments, but the main system considered is the easiest as it
takes advantage ofthe symmetry ofloading on the section of the analogous
column.
 395

EzaJnple 4.12 Find the moment at end a and the reaction at support b
of the beam shown in Fig. 4.20a if the moment ofinertia varies as indicated.
wllm
(a)
a 21 1
LJ2 LJ2
(b)~.D.

112 EI=1
Ie)

Fig. 4.20

Solution : A possible main system is obtained by replacing the fixed sup-

port at a by a hinge. The corresponding Mo.D. is shown in Fig. 4.20b.
Since the beam is simply supported at b, the section ofthe analogous column
will have an infinite area concentrated at b and its centroid will coinside .
with b as shown in Fig. 4.20c.
Referring to this figure"
A = '"

= 0.5 (0.5 L) 3 + (0.5 X 0.5 L) (0.75 L) 2 + 1 (0.5 L),/12

12
3L3
+ (I X 0.5 L) (0.25 L)2 = 16
=_1
2
(.!
3
X ~ X WL2) = wL3
2 8 48
J.
2 L WL2) wL3
N2 = I ( 3 X '2 X ""8 = 24 J.
= WL3(-2 X.!:.. X ~) + WL3(~ X.!:..) = 7wL4 ++-
48 8 2 2 24 8 2 256
7wL4 L 7wL2
= - 256 X 3 L3/16=- 48
which is the final moment since the Mo-moment at a is zero.
7wL2 wL2 17wL
~Ma=0=LXYb+4i\-2 'Yb = 4 i \ '
396

4,9 Determination or stiffness and carry-over factors

Perhaps the most important application of the method of column
analogy is the determination of the stiffness and carry-over factors of non-
prismatic members. A knowledge of these values is essential to the moment
distribution method of analysis which will be presented in chapter 6. At
this stage it will be advantageous to define these new terms.

Stiffness : The bending stiffness at end a of member ab, which is usually

denoted by Kabj is the moment required to produce a unit rotation at end
a while end b is supported in some particular manner; usually fixed.

Carry-over factor: The carry-over factor from end a to end b of

member ab, which is usually denoted by Cab' is the ratio of the moment
induced at end b to the moment applied at end a.

The stiffness and carry-over factors at an end of a member depend on

the shape of the member and the restraining conditions at its other end.

Consider a prismatic member ab such as that shown in Fig. 4.21, and

let it be required to determine the stiffness Kab and the carry-over factor
Cab' If a moment Mab is applied at end a causing this end to rotate' an
angle aa = I then by definition Mab = K ab Also, as a result of M ab,
some moment Mba will be induced at end b preventing this end from
rotation, and again by defini tion Mba = Cab K ab

(al

(bl

Fig. 4.21

In order to determine Kab and Cab the usual procedure would be to

load the section of the analogous column by the moments on the beam.
These moments should satisfy the following displacement boundary
conditions :
 397

(I) The net area of the (M/EI)-diagram between a and b should

be equal to aa since ab = O.

(2) The moment of the (M/EI) - diagram between a and b about a

should he equal to zero since there is no relative displacement between the
two ends.

A concentrated load = I rad. applied at the rotated end a satisfies

both these conditions,and can therefore replace the actual elastic loading
on the beam. Consequently, whenever the stiffness and the carry-over
factors are required, an equivalent concentrated load of a unit value, is
applied to the section of the analogous column at the point corresponding
to the rotated end of the beam. This rule applies regardless of the shape
of the beam.

Referring to Fig. 4.21 b which shows the section of the analogous

column corresponding to the prismatic beam considered"
N Mx X L/2)
K
ab -
-f
a -
--+-
A
- -I+ (IL3/12EI
I - L/EI
(L/2)

The stiffness of a prismatic member is thus given by :

Kab = 4 EI/L ... 4.39
I (I X L/2)
Kab Cab = fb = - - 3 (L/2) = - 2 EI/L
L/EI L /12 EI

Substituting for KabJ

I
Cab -- - -2

The negative sign results from the sign convention used in the method
of column analogy.

The carry-over factor from one end to the other end of a prismatic
member is given by :
I
C
a
b = -2 . ... 4.4{)

The student is advised to memorize the stiffness and ~rry-over factors

of prismatic members derived above as they will be needed later in regard
to the method of moment
, distribution.
 398

Esanlp le 4.13 Calcula te the stiffness factors at ends a and b and

the
carryover factors from a to b and from b to a of the nonpri smatic
beam
shown in Fig.4.2 2a.

Solutio n: The section of the analog ous column is shown in

Fig. 4.22h.
Referr ing to this figure ..
A 0.8 L (1/2EI) + 0.2 L (I/EI) = 0.6 L/EI
=
(0.4 L/EI) 0.4 L + (0.2 L/EI) 0.9 L
x = = 0.567 L
0.6 L/EI
1 1 0.4 L 1 1
Iy = 2EI (0.8 L)3 X 12 + EI (0.167 L)2 + EI (0.2 L3) X 12

+ 0.2 L (0.333 L) 2 = 0.055 L3/EI

EI
To determ ine Kab and Cab' place a unit load at a,
N Mx EI (1 X 0.567 L) (0.567 L) EI
K.b = f. = A +1 = 0.6 L + 0.055 L3
= 7.5 EI/L
EI (I X 0.567 L) (0.433 L) EI _ IfL
K b C b= fb = - - - - - 2.8 E
a a 0.6 L 0.055 L3
Cab = - 2.8/7.5 = - 0.37
To determ ine Kba and Cba' place a unit load at b,
N Mx EI (1 X 0.433 L) (0.433 L) EI
K .. = fb =A: +1 = 0.6 L + 0.055 L3
= 5.1 ElfL
= f ~ _ (I. X 0.433 L) (0.433 L) EI = _ 2.8 EI/L
=
Kba Cab a 0.6 L 0.055 L 3
C .. = -2.8/5 .1 = -0.55
 EXAMPLES TO BE WORKED OUT

(1)-(8) Using the method of elastic centre, construct the B.M.D

for each of the frames shown in Figs. 3.52,3.53 and 4.23-4.28. The
of inertia moment varies as indicated

8t

r
4m
--
E
Ul
4m

~3m 3 ----I
7-
Sm
Fig. 4.23 Fig. 4.24

2t1m
21
f
4m 21
31
21 4m
21

t Fig. 4.25
LI
1
__ 2..L-Sm
111m
31

Fig. 4.26
~2...J
400

(9)-(14) Using the method of column analogy, r.e-solve prt>blems

1-8 (Figs. 3.52. 3.53 and 4.23-4.28).
(15) - (18) Using the method of column ilJlalogy qmsttuct the
, B.M.D. for each of the frames in Figs. 3.47, 3.48, 3.54 and 3.55.
(19)-(26) Using the method of column analogy, ennstruct the B.M.
and S.F.Ds. for each of the frames shown in Figs. 4.29- 4.36. The moment
of inertia varies as indicated.

II ",:" ~
L
t
4 4 2t1m

L-""____ ~~
L4 -L--- 12 m ~j L- 3 - L - 6 m ---l.- 3 -I
FigA.29 Fig. 4.30

iii 311m

I 16
6m---l-~;f
4m

~2i L3 Fig. 4.32

Fig. 4.31

,11
3m
21 -,
2t 211m 2

6 J
21
I
i4

8m
J
Fig. 4.33 Fig. 4.34
 401

211m 8t 8t
I I .. I " I I i I I I 2~/.m

21 21 1
6m

..J 2 1-" --I-4m 4-1-4 t

Fig. 4.35 Fig. 4.36

(27)-(30) Using the method of column analogy, determine the

moments and stiffnesses at ends a and b, and the carry-over factors from a
to b and from b to a for each of the beams shown in Figs. 4.37 - 4.40.

~Q 21 O~ b~
6mt
~Q 151
t I b~
J 4m I ~ I 6m 3---1
Fig. 4.37 Fig. 4.38

F=::=l==~~1.6m l~! 14 t b~l

~~8
1- 3 -J,.....: 6m-l--3 .. /1 r I--
IS
4x3 =12m --J"
Fig. 4.39 Fig. 4.40
I
! 403

CHAPTER 5

r THE SLOPE - DEFLECTION METHOD

5.1 Force and displaceD1ent D1ethods of analysis

In all the methods of elastic structural analysis, two basic conditions
must be fulfilled. These are

f (1) Static equilibrium.

(2) Compatibility of deformations.
Two main methods of analysis are possible. In the first, static equil-
ibrium is always satisfied and equations are established to express the
second condition; compatibility of deformations in the structure. Since
these equations use forces as unknowns. . the method is referred to as the
force method. The method of consistent deformations presented in chapter
2 and the method of the equation of three moment.s presented in chapter
3 are examples of the force method. In the second method,compatibility of
deformations in the structure is always maintained and equations are
established to express the first condition;static equilibrium of the structure.
Since these equations use displacements as unknowns} the method is
referred to as the displacement method. The slope-deflection method to
be considered in this chapter and the moment distribution method to be
considered in the following chapter are examples of the displacement
method.

5.2 Degree of freedoD1

The number of possible joint rotations and independent joint transla-
tions in a structure is called the degrees of freedom of the structure.
There are three types of degrees of freedom

(1) The degree of freedom in rotation

The degree of freedom in rotation of a beam or a frame is equal to
the number of possible joint rotations. It follows that if:,
sr = degree of freedom in rotation}
j number of joints including supports}
f number of fixed supports,

s, = j - f 5.1

!
 404

(2) The degree of freedom in translation

The degree offreedom in translation of a frame is equal to the number

of independent joint translations which can be given to the frame.

Since in general each joint has two joint translations" the total number
of possible joint translations = 2 j. Since, on the other hand, each fixed
or hinged support prevents two of these translations, and each roller sup-
port or connecting member prevents one of these translations, the total
number of the available translational restraints is (2f + + +
2h r m),
where

f number of fixed supports"

h number of hinged supports,

r number of roller supports,

m number of members.

The degree of freedom in translation, sp is thus given by :

St = 2j - (2f + 2h + r + m) 5.2

(3) Combined degree offreedom

The combined degree of freedom of a frame" s, is the sum of its degrees

of freedom in rotation and translation. Thus,
5.3

As opposed to the degree of statical indeterminancy, n, which refers

to the number of redundant forces, s is sometimes called the degree of
kinematic indetenninancy. If s = 0, the structure is said to be kinematically
determinate. The fixed beam is an example of a kinematically determinate
structure.

5.3 Outlines of the slope-deflection D1eth~d

The slopedeflection method is applicable to the analysis of statically

indeterminate beams and frames. It is particularly useful for the analysis
of highly statically indeterminate structures which generally have a low
degree of kinematic indeterminancy.
~
I
I 405

II ~~a~ ________~b~________-TC________~d

e f

Fig. 5.1

This is demonstrated py the frame shown in Fig. 5.1. The frame is

nine times statically indeterminate. This calls for the simultaneous solution
of nine equations if the force method of analysis were to be used. On the

It other hand, the frame has only two unknown rotations.. ab and ac, and
is thus kinematically indeterminate to the second degree. This calls for
the simultaneous solution of two equations only if the slope-deflection
! method is used.

I
The general procedure for the analysis is outlined below:

(1) The degree of freedom in rotation and translation are found

separately. This is done either by inspection or by the application of
equations 5.1 and 5.2.

(2) The moments acting on the ends of individual members of the

structu(e are expressed in terms of the rotations of the joints at the ends
of the member, the rotation of the member itself (if any) and the load
acting on the member. This is done by the slope-deflection equations to
he derived in section 5.5.

(3) The conditions of static equilibrium provide as many equations

as there are joint and member rotations. When solved simultaneously,
these equations give the unknown rotations .

(4) Back-substitution of the thus-determined rotations into the slope-

deflection equations yields the moments acting on the ends of individual
members.

5.4 Sign conventions

Before proceeding to derive the basic slope-deflection equations, it is
necessary to establish specified sign conventions for the moments and
rotations.
406

Moments acting on the ends of members, joint and member rotations

are considered positive when occuriDg in a clockwise direction. For
example, a fixed beam acted upon by a downward load will have negative
and positive moments at its left and right ends respectively. Also, if the
right end of a member sinks with respect to the left end, the member
rotation is positive.
This sign convention, though suitable for considering joint equilibrium"
contradicts the graphical presentation of the bending moments along the
length of a member where moments producing tension in the lower fibers
of the member are considered positive and vice versa. Nevertheless, once
the end moments are obtained it is quite easy to change over to the usual
graphical convention for plotting B.M.Ds.. This is done by drawing free-
body diagrams for individual members with the end moments indicated
according to the sign convention used in the slope-deflection method.

5.5 Derivation of the slope-deflection equations

Fig. 5.2a shows a single member ab of a continuous beam or a frame .
As a result of the loads applied to the structure, the member considered
will assume the given general deformed shape. Ends a and b have been
caused to rotate through angles aa and ab and the axis of the member
has rotated an angle.p = t. /L as a result of end b settling an amount t.

Fig. 5.2

relative to end a. It is now required to find expressions for the final mom-
ents M .b and Mb acting on ends a and b in terms of the rotations aa
and Ub at joints a and b, the rotation of the member, ifJab and the loads
acting on the member.
 407

First assume, as in Fig. 5.2h.. that ends a and b are fixed, i.e. with ~ero
rotation at the ends. The loads on the member produce the fixed-end
momeu'ts MPab and MFba indicated in Fig. 5.2h in their assumed positive
direction. These fixed-end moments must be corrected to allow for the
end rotations a, and ab and the member rotation ",. The effects of
these rotation. will be found separately and the final result will then be
obtained by superposition.

Il,l Fig. 5.2c the tangent to the elastic curve at a is rotated by a moment
M'a while end b is kept fixed. Using the second moment-area theorem
it may be easily shown that M' b = t M' a' and from the first moment-
area theorem, a. = M' aL/4EI. H'ence,

M'. = 4 EI a./L and M'b = 2 EI a./L.

Similarly, with reference to Fig. 5.2d it may be shown that ;

In Fig. 5.2e, end b is shown displaced relative to end a by an amount!::i

while the tangents to the elastic curve at a and b are kept fixed against
rotation. The end moments Mill a and M'" b may again be expressed in
terms of the member rotation ,p using the moment-area theorems. Thus,

Mm. = Mm b = - 6 EI !::ilL' = - 6 EI of/L.

Ifnow, for each end ofthe member all the end moments corresponding
to the various deformations are added to the fixed-end moments, the fol-
lowing expressions are obtained :

... 5.4

Equations 5.4 are the basic slope-deflection equations. If there is no

relative displacement between the ends of the member, of = 0 and these
equations reduce to :

.. . 5.5
Mba = MF b. + 2 EI/L (2 ab + a. )
:-
408

It should be noted that the coefficient (2EIJL) appearing in the slope-

deflection equations-is generally different for each member in the structure.

If the (2EIJL)-values for all the members are made n times smaller (or
larger), the effect will only be to make all the rotations n times larger (or
smaller) while the products (2 EI/L) (2 a. + a. - 3.p) and (2 EI/L)
+
(2 a. a. - 3.p) in equations 5.4 or (2 EI/L) (2 a. + a.) and (2 EI/L)
(2 ab + aa) in equations 5.5 remain unchanged, and consequently the
values of the end moments remain 'also unchanged. Thus., if the
actual values of the rotations are not of interest, as is usually the case.f
the relative values of (2 EI/L) rather than their absolute values may be
used in the slope-deflection equations. If the relative values of (2 EI/L)
are denoted by K, the slope-deflection equations in 5.4 and 5.5 may be
restated -a s :

5.6
M = MF + K (2 a. + a. -3.p)
and

... 5.7

5.6 Fixed-end DlODlents

MFab and MFba appearing in the slope-deflection equations are the

moments that develop at the ends ofa member assuming it to be restrained
against translation and rotation as in the case of a fixed-end beam.

Expressions for these moments due to common cases of loading are

given in Fig. 5.3. Expression for other cases of loading may be found
from the first principles or from the given expressions using the principle
of superposition. It should be remembered that the given expressions ar~
for prismatic members.
 409

M~b Loading Case Mba

F

wI! w
-12 a~ I I
I ~b wI!
12
! L I
/.

I
r -7
Pab
z
~ I--- a
r
I
I b ~ I
Pba 2
t3

PL
8 ~I---- L12
r L12 ----' ~
PL
8

Pa(L-a)
L ~f-a-.J
PI
r L a ---< ~
Pa(L-a)
L

Mb(2a-b)
t3 ~I-- a :J b ~
I
Ma(2b-a)
t3

f
<>'tEl (12-11)
d t=- 12

11

Fig . 5.3
r-* - <>'lEJ(tZ-Il)
d

5.7 Applications to statically indeterminat e beams

In this section. . the appl ication of the slope-deflection equations to the

analysis of statically indeterminate beams will be considered. Beams
without joint translation or known joint translation are dealt with.

It should be noticed that only the relative values ofK need be known
for the analysis of beams having no joint translation . On the other hand,
the absolute values orK must appear in the slope-deflection equations for
beams with known joint translation.
410

EsaDlple 5.1 Draw the B.M. and S.F.Ds. for the beam shown in
Fig. 5.4a if the moment of inertia varies as indicated.

311m
~a , , , , 'c~
(a)
3 ...$",.,'___'
3 _b 2i6'~
1 I
(O)~, r' Jt ,'liT,,, i
5.5 7m! ~
(c)BMO~ /;\. L I
I ~ 8.51
~
~

(d)S.FOI
3.75
+ I
315
~ _
425 425 ~
9.5
Fig. 5.4

Solution: Since there is no joint translation, form 5.7 of the slope-deflection

equations may be used. The relative stiffnesses and fixed-end moments
are found first.

K,b : KbC = 1/6 : 21/6 = I : 2

MF,b = - 8 X 6/8 = - 6 m.t., MFba = 8 X 6/8 = + 6 m.t.
MP bc = - 3 X 6 2/12 = - 9 m.t., MFeb = 3 X 6 2/12 = + 9 m.t.

The slope-deflection equations are :

M ab =-6+I(ab)
Mba = + 6 + I (2 a b)
Mbc +
= - 9 2 (2 ab)
Meb = + +
9 2 (ab)
For the single unknown ab appearing in the slope-deflection equations"
ODe equation of statics is required. This is provided by the condition of
equilibrium at joint b; L Mb = O. Thus,
f

411

Mba + MbC = O.
Substituting from the slope-deflection equations into the joint equili-
brium equation,
6 + 2 ab - 9 + 4 ab = 0, ab = 1/2
By back-substitution in the slope-deflection equations,
Mab = - 6 1 + (1/2) = - 5.5 m.t.
Mba = +
6 + 1 (2x l/2) = 7 m.t.+
Mbc = - 9 + 2 (2 X 1/2) = - 7 m.t.
MCb = 9 + +
2 (1/2) = + 10 m.t.
With the end moments thus determined, the free-body diagrams for
the two members may be drawn as shown in Fig. S.4h. The corresponding
B.M. and S.F.Ds. are obtained in the usual manner and are as shown in
Figs. 5.4c and d respectively.
It should be noted that in this problem there are three unknown
moments and one unknown angle. It is obvious therefore that a solution
by the displacement method will be simpler than .a solution based on the
force method.

Elaunple 1.5 Draw the B.M. and S.F.Ds. for the beam shown in
Fig.5.5a if EI = constant.

(al ~a r ;"
3 --I--- 3
211m
I I I I

6m
r I
II~
p2.J
2tl
.C;I

f--

I
!
( b) 43~34732 -.,;1. 7_.2_1
. _*:JI:J:Jj?:iII'i~m':IlIIIII'
3.828 6.386
*1
5614 2

Il 4.943
(C)B.M'O [?'->, ""'Z7
I
3'i"-37!..!2,--;----,3"1.372
6.314mt
/?\-
6.386
h--.
~
~
4

1c::=+:C=l-I___+I""-2+-=:::::::::"....~=-_-f2=+=:i2t
! (d)S.F.O.

3.8'"'2"'"8----:3~.828 ~14
~ Fig. 5.5
 412

Solution : Since there is no joint translation, form 5.7 of the slope-deflec-

tion equations may be used.
The relative stiffnesses and fixed-end moments are as follows :
K'b : Kbc = 1/6 : 1/6 = I : I
MFab = - PL/8 = - 7.2 X 6/8 = - 5.4 m.t., MF b = + 5.4 m.t.
MPbC = - wL2/12 = - 2 X 6 2112 --: - 6 m.t., MFcb = + 6 m.t.
Here aa is known to be zero and there are two---unknown a. - values; Q.b

and ac . The slope-deflection equations are:

Mab = - 5.4 + I (2 aa + ab)'

M.b = - 5.4 + ab
Mba = + 5.4 + I (2 ab + aa),
Mba = + 5.4 + 2 ab
Mbc = - 6 + I (2 ab + Mbc = - 6 + 2 ab + a c
a c)'
Mcb = + 6 + I (2 a c. + MCb == + 6 + ab + 2 a c
ab)'
(Note that a slope-deflection equation needs not be written for can-
tilever cd as it is statically determinate. Considering a section just to the
right of c, Mcd = - 2 X 2 = - 4 m.t. The negative sign for the moment
is required since the moment acting on the end of member cd. is
anticlockwise .
With two unknown rotations, two equations of statics are required.
These are provided by the conditions of equilibrium at joints band c.
At joint b, Mba + Mbc = 0
At joint c, Mcb - 4 = 0
It is of particular importance to note that all the unknown moments
should be assumed positive while any known moment should be substituted
for in the equilibrium equation with its known value and sense of direction.
Substituting from the slope-deflection equations into the joint equili-
brium equations,

4 ab +
a c - 0.6 = 0 (i )
ab+2ac+2 =0 (ii)

Solving equations (i) and (ii) simultaneously,

ab = 0.457 and a c = - 1.23
Substituting these values in the slope-deflection equations..
Mab = - 5.4 + (0.457) = - 4.943 m.t.
M.,. = + 5.4 + 2 (0.457) = 6.314 m.t.
Mb< = - 6 + 2 (0.457) + (- 1.23) = - 6.314 m.t.
Mcb = + 6 + (0.457) + 2 (- 1.23) = 4.000 m.t.

I
 413

Having thus determined the end moments, the free-body diagrams

for all the members may be drawn as shown in Fig. 5.5b. The correspond
ing B.M. and S.F.Ds. are obtained in the usual manner and are as shown
in Figs. 5.5c and d respectively.

Example 5.3 AU'lly,e the continuous beam shown in Fig. 5.6a if the
moment of inertia varies along the beam as indicated.

19t e 41 t d
(a)
'@ I 1.21 :a
I I I ~
2 - l - - 4 ___ 2 -1--2 -I-- 2-1

(b)
2t1m
i [ I I r805mt

'~---L..--~~I~
t

l'
4.66 7.34 6.31t 269 5.03 2.97

80~_ - ____68
.1 mt
12
(e ) 8.M.D. ~=---:-h;--7--~--E-""7,.L.--=~~=---=-=t:~":>!
+

I
4.66 t 6.31 6.311
5.~
(d)S.F.D. ~ ~
~ 2.69 2.69 2.~ 2.97
7.34
Fig. 5.6

Solution: The relative stiffnesses and fixed-end momen ts are as follows:

K'b : K"" : Kcd = 1/6 : 1.21/6 : 1/6 = 5 : 6 : 5

MFab = - wL2/12 = - 2 X 6 2112 = - 6 m.t.,

M"bC = - Pab 2/L2 =- 9 X 2 X 4 2 /6 2 = - 8 m.t.

MFcb = Pba 2/L2 = 9 X 4 X 22/6 2 = + 4 m.t.

MPcd = - Pa (L - a)/L = - 4 X 2 X 416 = - 5.33 m.t.,

M"dc = + 5.33 m.t.

414

The slope-deflection equations are :

Mab = - 6 + 5 (2 aa + ab) Mab = - 6 + IO aa + 5 ab
Mba = + 6 + 5 (2 ab + aa) Mba = + 6 + 5 aa + IO ab
M bo = - 8 + 6 (2 ab + a o) , M bo = - 8 + 12 a b + 6 a~
Mob = + 4 + 6 (2 a o + ab) ,Mob = + 4 + 6 ab + 12 a o
Mod = - 5.33 + 5 (2 ao + ad) , Mod = - 5.33 + IO a o + 5 ad
Mdo = + 5.33 + 5 (2 ad + a o) , Mdo = + 5.33 + 5 a o + IO ad

It should be noted that in keeping with the general development of the

slope-deflection equations, the fixed-end moments MFab and MFde must
be included although it is known that the final moments at .a and bare
zero. Later, Mab = 0 and Mdc = 0 provide two of the required equations
of statics.
Inspection of the above slope-deflection equations shows that there
are four unknown rotations; aa ,ab , a c and ado Therefore, four equa-
tions of statics are required. These are provided by the conditions that
the sum of the moments at each of joints a,b,c and d is zero. Thus,
M'b = 0, Mba + Mho = 0, Mob + Mod = 0 and Mdo = O.

Substituting from the slope-deflection equations into the joint equil-

ibrium equations,

10 a. + 5 ab =6 (i )
5a.+ 22a.+ 6 "'0 =6 ( ii)
6 a. + 22 "0 +5 .... = 1.33 (iii)
5 "" + 10 "II = - 5.33 (iv)

Solving equations (i) - flV) simultaneuusIy,

a. = 0.663, "'" = - 0.126, Bo = 0.2.45 and ad =- o.W
Substituting these values into the slope-deflection equatiens~

Mab = - 6 + 10 (0.663) + 5 (- 0.126) = - 6 + 6.63 - 0.63 = 0

Mba = + 6 + 5 (0.663) + 10 (-0.126) = + 6 + 3.31-1.26 =8.05 m.t.
M bo = - 8 + 12 (-0.126) + 6 (0.245) = - 8- 1.512 + 1.47
= -8.05 m.t.
Mob = + 4 + 6 (-0.126)+12 (0.245) =+4--D.756+2.94D = 6.ISm.t.
Mod = - 5.33 + 10 (0.245) + 5 (- 0.66) = - 5.33 + 2.45 - 3.30
= -6.18 m.t.

Mdo = +5.33 + 5 (0.245) + 10 (-0.66) = + 5.33 + 1.23 - 6.6 = 0

 415

Having thus determined the end moments, the free-body diagrams for
all the members are obtained and the B.M. and S.F.Ds. are constructed
in the usual manner. The result is shown in Figs. 5.6 b-d.
It should be noted that in this problem there are two unknown
moments only although by the slope-deflection method four unknown
angles are involved. It follows that a solution by the force method will be
simpler than that worked out above.

Esa.nple 5.4 Analyse the beam shown in Fig. 5.7a in the absence of loads
for the following simultaneous support movements:
A downward movement of 0.3 em. in addition to a clockwise rotation
of 0.001 rad. at support a, a downward movement of 1.2 em. at support
b, and a downward movement of 0.6 em. at support c. EI = 5000 m 2t.

b c
(a)
1
I--- Sm S 32
d

(b) q?2ml

1.191
1}
1.19 0.66
t
0.66
3.22ml
(c) B.M.D.
+
3.93
1.19 1 1.19
(d} 5.F.o. +

0.66 0.66
Fig. 5.7

Solution: Since there are finite joint and member rotations, form 5.4 of
the slope-deflection equations should be used with the absolute (2EI/L)-
values. Further, as there are no loads all the fixed-end moments are zero.
In this case it is known that:
1.2-0.3 1.2-D.6
a. = + 0.001, "'ab= - 600
- = 0.0015 and "'be = -- 6
00
= -- 0.001

.-----------------------------------------------------------------
416

The slope-deflection equations are thus as folleWs :

Ma. = 2 (5000/6) (2 X 0.001 + a" - 3 X 0.0015)
M = 2 (5000/6) (2 a. + 0.001 - 3 X 0.0015)
M., = 2 (5000/6) (2 ". + a, + 3 X 0.001)
Me. = 2 (5000/6) (2 a.,. + a" + 3 X 0.001}
There are two unknown rotatioIlS'; Q.iJ..and -a.c ."l'hese'" may be found
from the conditions of equilibrium at joints band c. Thus~
and
Substituting from the slope-deflection equations into the above two
equilibrium equations"
4a.+ a,-0.5 X 10- 3 =0 (i)
a. + 2 a, + 3 X 10- 3 = 0 (ii)
Solving equations (i) and (ii) simultaneously,
a. = 0.57 X 10- 3 rad. and a, = - 1.79 X 10- 3 rad.

Substituting these values into the slopeMdeflectioD equations,

+
Mab = 2 (5000/6) (2 X 10- 3 0.57 X 10- 3 _ 4.5 X HI-3) = - 3.22 m.t.
M b = 2 (5000/6) (2 X 0.57 X 10- 3 + 1 X 10- 3 - 4.5 X JO-3)
= - 3.93 m.t.
M b, = 2 (5000/6) (2 X 0.57 X 10- 3 - 1.79 X to..- 3 + 3 X JO-3)
= 3.93 m.t.

M = 2 (5000/6) (- 2 X 1.79 X 10- 3 + 0.57 X 10- 3 + 3 X t!r"3}= n

In this type of problem, particular care should be taken with regard
to the proper signs of the known a and of angles and also the .....ls used.
H~vingdetermined the end moments, the free-body diagrams for
individual members are obtained and the B.M. and S.F.I:>$.. are construc-
ted as shown in Figs. 5.7 bod.

Example 5.5 For the restrained beam shown in Fig. 5.8, determine the
fixing moment at end a ifsupport b settles 8 per unit load and EI = con-
stant.

r I-- U2 ---4-1_
-
I
P

lI2
Fig. 5.8
 417

Solution ;~b= 8xY. = 8 (Pf2 +M..JL). Henc.e,~ = 8 (Pf2 + Ma.{L){L.

Since there is finite rotation of member ab, form 5.4 of the sIope-deHeetion
equations should be used. Thus,

1.1 = - PL/8 + (2 EI{L) [a. - 3 8 {P/2 + Ma.{L){L]

= - PL/8 + 2 EI aofL - 6 EI 8 {P/2 + M .fL){L2

M'a = + PL/8 + (2 EI{L) [2 ao ~3 8 {P/2 + M .fL){L]

= + PL/8 + 4 EI a.fL - 6 EI 8 {P/2 + M {L){L2
There is ODe unknown rotation, aa,- This is obtained from the con-
dition ; 1.1.. = 0 = PL/8 + 4 EI a.{L - 6 EI 8 {P/2 + Mab{L){L2 or,
2EI a,JL = 3EI 8 {P/2 + Ma.{L){L2- PL/16
Snb'tituting this value inti> the first slope-deHection equation,

1.1.. = - (PL/16 + EI 8 pf2L2) / {1/3 + EI 8 {L3)

If support b does not settie, 8 = 0 and 1.1 = - 3PL/ 16 which is the
known value for a restrained beam subject to a ceotral coz:centrated load
P. It is also noticed that sinking of support b reduces the reaction at b
and increases the end moment at a.

5.8 AppW:ations to static:ally indeterminate fraJnes with DO joint

traDslation

Frames that do not involve transIation of the joints may be solved using
form 5.5. or 5.7 of the sIope-deHeetion equations. In this case, as for static-

I
ally indeterminate beams, there will be as many joint equilibrium equations
as there are unknown joint rotations. Mter solvirg thefe <quatiCJ:s:" the
joint rotations obtained are 'back-substituted in the sIope-ddlection equa-
tions to determine the moments acting on the ends of the various members
and the analy.is is then completed in the usnal manner.

Translation of the joints is generally prevented as a result of the dis-

f
I
position of the supports and the supporting conditions of the frame or
complete symmetry in both frame and loading. Fig. 5.9 shows examples of
frames w ith no joint translation. Joint translation of the frames in .
Figs. 5.9a and b is prevented due to the supporting conditions. On the
~ other hand, joint translation of the frames in Figs. 5.9c and d, while gen-
I erally possible, do not occur due to the complete symmetry in frame and
I loading. It should be noticed that strictly speaking there are slight joint

~
418

(a)

(Il)

(c) (d)

Fig. 5.9
translation due to the axial defonnation of the members but this is norm-
ally neglected in the analysis of scch frames.
Ezunple 5.6 Draw the S.F. and B.M.Ds. for the frame shown in Fig. 5.IOa
if the moment of inertia varies as indicated.

51
1.5t/m
c
21
2.41
51T\. r
d
,. ~
I..- 4 m--l-- 4 ---J
(a)
8.51

2.4

2.4 z!. 4 4
(c)S.FD. (dlB.M.D
Fig. 5.10
 419

Solution: Since the frame is symmetrical and symmetrically loaded, there

will be no joint translation. The relative stiffnesses and fixed-end moment,s
are as follows :
Kab: K b, : K,. = 1/5 : 21/S : 1/5 = 4 : 5 : 4
MP b, = - (wL2/12 + PL/S) = - (1.5 X S2/12 + 5 X SIS) = -13m.t.
MF'b = + 13 m.t.

Here it is known that ua a.b = 0 .. and hence the slope-deflection equa-

tions are as follows:
Mab = 0 + 4 ( a b) ,Mab= 4u b
Mba = 0 + 4 (2 a b) ,Mba = 8 ab
M., = - 13 +
5 (2 a b +
a,) , M b , = - 13 +
10 ab +
5 ae
M,b = + 13 + 5 (2 a, + a b) , Meb = + 13 + 5 ab + 10 a,
M,. = 0 +
4 (2 a,) ,Mcd=8uc
M.,=O+4(a,) ,Mdc =4ac

There are two unknown rotations; ab and Q.c ' These may be found
from the conditions of equilibrium at joints band c. Thus,
and

Substituting from the sIope-deft(ction equations into the two Jomt

equilibrium equations and solving them simultaneously, a b = +
1 and
ab = - 1. The numerically equal values of ab and a~ are consistent
with the symmetry of the frame. (This symmetry could have lead to appre-
ciable 'implifications had it been utilized at the outset ofthe solution. This
will be discussed in detail in section 5.11).

Substituting the values of ab and a c into the slope-deflc::ction equations,

Mab = 4 (I) = 4 m.t.
M b = S (I) = S m.t.
M b, = - 13 + 10 (I) +
5 (- I) = - S m.t.
M,b = + 13 + 5 (I) + 10 (- I) = S m.t.
M,. = S (-I) = -Sm.t.
M., = 4(-1) =-4m.t.

The free-body diagram for each member is shown in Fig. 5.lOb while
the S.F. and B.M.Ds. are shown in Figs. 5.IOc and d respectively.
 420

&cunple 5.7 Analyse the frame shown in Fig. 5.lIa if EI = constant.

3t/m 18.875mt
3.72

13.08

O.2Bt
i--- 8m - - - I
(al 110.92t (bl
10.92 18.875
+

Q2 3.72 11081

0.28
(clS.F.D. (dlB.M.D.
Fig. 5.11

Solution: All the joints in the frame are prevented from translation by
virtue of the supporting conditions. The relative stiffuesses and fixed-end
moments are as follows:

K' b : K"" = 1/6 : 1/8 = 4 : 3

MFab = - 4 X 6/8 = - 3 m.t. "MFba = + 3 m.t.
MF"" = - 3 X 8 2/12 = - 16 m .t. , MFeb = + 16 m.t.
Here a.c = 0, and the slope-deflection equations are as follows:
Mab = - 3 + 4 (2 a a+ a b ) , Mab = - 3 + 8 aa + 4 ~
Mba = + 3 + 4 (2 ab + aJ , Mba = +- 3 + 4 "a + 8 ab
M"" = - 16 + 3 (2 a b ) , M"" = - 16 6 ~ +
Meb = + 16 + 3 ( "b) , Meb = + 16 + 3 "b
The two unknown angles aa and a b are found from the two joint
equilibrium conditions; Mab = 0 and Mba + Mbc = 0
Substituting from the slope-deflection equations into the joint equili-
brium equations and solvingsimultaneousIy, ". = - 5/48 and ab = 23/24.
Substituting these values back into the slope-deflection equations"
 421

Mab = - 3 + 8 (- 5/48) + 4 (23/24) = 0

Mba = + 3 + 4 (- 5/48) + 8 (23/24) = 10.25 m.t.
Mbe = - 16 + 6 (23/24) = - 10.25 m.t.

Meb = + 16 + 3 (23/24) = 18.875 m.t.

The free-body, S.F. and B.M.Ds. are shown in succession in Figs.5.llb,c
and d.

EsaJnple 5.8 Construct the B.M.D. for the frame shown in Fig.5.12a if
the relative moments of inertia are as indi:ated.

l.5l/m 61

a 61 b 41

1.51

d
12m 8
(a)
20.52ml
----- ____1,.96 ./4 __
27
+ 2.52

1.26
( b) B.M .D.

Fig. 5.12

Solution: The relative stiffnesses and fixed -end moments are as follows:
= =
r K ab : Kbe : Kbd : K ..
MFab =-
61/12 : 41/8 : 1.51/6 : 1/6
1.5 X 122/12 = - 18 m.t. ,MFba =
3 : 3 : 1.5 : I
+ 18 m.t.
MPbe = - 1.5 X 8 2/12 = - 8 m.t. , MFeb = + 8 m.t.

Here aa = ad = 0, and the slope~deflection equations are expressed in

terms of ab' a c and ac.
422

Mab = - 18 + 3 ( ab) , M' b = - 18 + 3 ab

Mba = + 18 + 3 (2 a b) , Mba = + 18 + 6 ab
Mbc = - 8 + 3 (2 ab + a c) , Mbc = - 8 + 6 ab + 3 a c
M cb =+8+3(2ac +ab) , Mcb = + 8 + 3 ab + 6 a c
M b = 0 + 1.5 (2 ab) J,M bd = 3 Ctb

Mdb = 0 + 1.5 ( a b) , Mdb = 1.5 ab

M~ = 0 + 1 (2 ac + a e) , Mce = 2 o.c + ete
Mec = 0 + 1 (2 a e + a c) ,Mec =Ctc +2ae

The three unknown angles CtbJ Ct c and Q.

e are found from the equili-
brium conditions of joints b"c and c. Thus"
At joint b, Mba + MbC + Mbd = 0
At joint c, M,b + Moo - 12 = 0
At joint c, Mcc = 0

Substituting from the slope-deflection equations into the joint

equilibrium equations and solving simultaneously"
ab = - 0.84, a c = 0.868 and a e = - 0.434

Substitutiug these values back into the slope-deflection equations,

Mob = - 18 + 3 (- 0.84) = - 20.52 m.t.
Mba = + 18 + 6 ( - 0.84) = + 12.96 m.t.
Mbc = - 8 + 6 (- 0.84) + 3(0.868)= -10.44 m.t.
M'b = + 8 + 3 (- 0.84) + 6 (0.63) = + 10.67 m.t.
Mbd = 3 (- 0.84) = - 2.52 m.t.
M = 1.5 (- 0.84) = - 1.26 m.t.
M~ = 2 (0.863) - 0.434 = + 1.34 m.t.
M ec = 0.868 + 2 (- 0.434) = 0

The final B.M.D. is thus as shown in Fig. 5.12h.

 423

5.9 Applications to frattles with a smgle degree of freedOID in

translation

In general... joints in frames translate as well as rotate. Joint translation

m':iy occur as a result of applied lateral loads or lack of symmetry in either
frame or loading.

(a) (b) (c)

0.

(d) ( e)
(f)

( g) ( h)

Fig. 5.13

Fig. 5.13 shows several examples of frames with different degrees of

freedom in translation. This can be found either by inspection or by the
application of equation 5.2 and the result is presented in Table 5.1.
424

Frame j f h r m St = 2j - (2f + 2h + r + m)
-- - - - - --
a 4 2 - - 3 2 x 4 - (2 x 2 + 3) = I
b 4 2 - - 3 2 x 4 - (2 x 2 + 3) = I
c 6 2 I - 5 2 x 6 - (2 x ~ + 2 x I + 5) - I
d 5 2 - - 4 2 x 5 - (2 x 2 + 4) = 2
e 6 2 - - 5 2 x 6 - (2 x 2 + 5) = 3
f 6 2 - - 6 2 x 6 - (2 x 2 + 6) = 2
g 7 3 - - 6 2 x 7 - (2 x 3 + 6) = 2
h to - I I 13 2 x to - (2 x I + I + 13) = 4
Table 5.1

It should be noticed that although frames d and e have two and three
degrees of freedom in translation, yet under symmetrical Joadir.g al1 joint
translations can be expressed in terms of a single joint translation.
In the analysis of frames with joint translation, ur.kr.cwn rr:eu:ber
rotations ,p in addition to unknown joint rotations a. ari~e. The nurr.her
of the unknown ljI-values corresponds to the degree of fre(clem in transla-
tion of the frame at hand. The calculation of the ",-values requires special
consideration.
Take, for example, the frame shown in Fig. 5.14a which has a single
degree offreerlom in translation . Due to the lack of symmetry, the tops
of c:>lumns ab and de will translate some amount 6..
Thus, .p.. = .po. = .p = 6th

l
(a) ( b)
Fig. 5.14

Of course at the beginning of the solution, the magnitude of the trans-

lation f:j" and hence the member rotation I/J, is not known.
 425

Referring to Fig. 5.14b, the moments acting on the ends of the columns
are shown in their assumed positive direction. From statics,

Xa = (M ab + Mbal/h and Xd = (M dc + Mcd)/h

and since there are no horizontal loads acting on the frame, the sum of
these reaction components must be zero. Thus._

This equation is commonly known as the shear equation. There will

always be as many shear equations as there are unknown member rotat..
ions. The student should note in particular that in the shear equations,
as in the joint equilibrium equations, all unknown moments should be
assumed positive.

The simultaneous solution of the shear equation and the other two
equations provided by the conditions of equilibrium at joints b and c;
Mba + MbC = 0 and Mcb + Mcd = 0, yields the unknown rotations
aa) a b and I{l.

Once these rotations are determined" the moments acting on the ends
of the various members may be found by back-substitution in the slope-
deflection equations and the analysis completed in the usual manner.

Exam.ple 5.9 Calculate the reactions of the frame shown in Fig. 5.15a if
the relative moments of inertia are as indicated.

10
, c e

6m I
2-+00f4-- . -.......96t
a d
:z r.

~~L~2J
l (a)
2.951 7.05t
Q96t

Fig. 5.15
r
4,26

Solution: There are three unknown rotations; two joint rotations Clb and
"e and one member rotation .p (= .pab = .ped). The relative stiffnesses
and fixed-end moments are as follows :
K'b : Kbe : Ked = 1/6 : 41/8 : 116 = I : 3 : I
M"ho = - 10 X 3.2 X 4.8 2/8 2 = - 11.52 m.t.
M"eb = +
10 X 4.8 X 3.2 2 /8 2 = 7.68 m.t.+
The slope-deflection equations are as follows:
M'b = 0 + I ("b-3.p) , Mab = (lb- 3 I/J
Mba = 0 + I (2 "b - 3 .p) , Mba = 2 ab - 3.p
Mbe = - + 3 (2 "b + "J , Mbe = - 11.52 + 6 "b + 3 a e
11.52
Meb = + 7.68 + 3 (2 "c + ab) , Mcb = + 7.68 + 3 "b + 6 a c
Mcd = 0 + I (2 "c - 3 .p) , Mcd = 2 "c - 3.p
Mdc = 0 + I ( ac - 3 .p) , M d, = "c - 3.p
At joint b, Mbo + Mbc = 0
At joint c, MCb + Mcd - 20 = 0
Shear equation, (Mob + Mba)J6 + (Mcd + M d ,)/6 +2 = 0

Substituting from the slope-deflection equatior:s into the three equations

of statics,
8 ab + 3 "e - 3 .p = 11.52 (i)
3 "b + 8 "c- 3.p = -7.68 (ii)
-3 "b - 3 "c + 12.p = 12 (iii)

Solving equations (i), (ii) and (iii) simultaneomly, ab = 1.49, a e = 1.65

and .p = 1.78.
Substituting these values into the slope-deflection equations,
Mab = 1.49 - 3 (1.78) =- 3.85 m.t.
Mba = 2 (1.49) - 3 (1.78) = - 2.37 m.t.
M be = - + 6 (1.49) + 3 (1.65) = + 2.37 m.t.
11.52
MCb = + 7.68 + 3 (1.49) + 6 (1.65) = + 22.05 m.t.
Mcd = 2 (1.65) - 3 (1.78) = - 2.05 m.t.
Mdc = 1.65 - 3 (1.78) = - 3.69 m.t.
The free-body diagram of each member of the frame is shown in
Fig. 5.15b. Indicated on the same diagram are the reactions at supports
a and d.
 427

Exa.nple 5.10 Construct-the R.M.D. rorthe frame shown in Fig. 5.16a

if EI = constant.

2t/m
rp:DIl=:ilIfu::a::r:I:j,
3m
61 4m

~ _ _ 12 m _ _ _-I
(a)
--Jf-fI==~.-l+

\
8.48
4.13
(c) B.MD.
Fig. 5.16

Solution: The relative stiffnesscs and fixed-end moments are as follows:

Kab : Kbc : KCd = 1/6 : 1/12 : 1/4 = 2 : 1 : 3
MFab = - 6 X 6/8 = - 4 .5 m.t. , MFba = + 4.5 m.t.
MFbc = - 2 X 122/12 = - 24m.t. , MFcb = + 24 m.t.
Neither the frame nor its loading is symmetrical. This lack of sym_
metry will cause member be to translate horizontally some amount b..
Thus"
.pab = !:l/6 = .p , .pbc = 0 and .pcd = !:l/4 = 1.5.p

There arc three unknown rotations; two joint rotations a b and a c and
onc member rotation ,p.

The slope-deflection equations are as foHows :

Mab = - 4.5 + 2 ( ab- 3 .p) , Mab = - 4.5 + 2 ab- 6 .p
Mba = + 4.5 + 2 (2 ab - 3 .p) , Mba = + 4.5 + 4 a b - 6 .p
Mbc = - 24 + 1 (2 a b + a c) , Mbc = - 24 + 2 ab + a c
428

Med = 0 + 3 (2 ne - 3 x 1.5.p) , Med = 6 ne - 13.5 .p

Mde = 0 + 3 ( ne - 3 X 1.5 .p) ,M de = 3 ne - 13.5 .p
At joint b, Mba + Mbe = 0
At joint c, MCb + MCd = 0

To construct the shear equation refer to Fig. S.16b, which shows the
free-body diagram of the frame. Consideration of the horizontal equili-
brium of the frame as a whole leads to the following shear equation:

X a +X d +6=0

(Mab; Mba 3) + (Med : Mde) + 6= 0

Substituting from the slope-deflection equations into the three equations
of statics..

6n b + ne - 61.p= 19.5 (i )

nb + 8 ne - 13.5 .p = - 24 (ii)

- 6 nb - 13.5 ne + 52.5 .p = 18 (iii)

Solving equations (i), (ii) and (iii) simultaneously, ab = 3.665, a e= - 3.838

and .p = - 0.225

Substituting these values into the slope-deflection equations,

M;b = - 4.5 + 2 (3.665) - 6 (- 0.225) = + 4.13 m.t.
Mba = + 4.5 + 4 (3.665) - 6 (- 0.225) = + 20.51 m.t.
Mbe = - 24 + 2 (3.665) + (- 3.838) = -20.51 m.t.
Meb = + 24 + (3.665) + 2 (- 3.838) = + 19.99 m.t.
Med = 6 (- 3.838) - 13.5 (- 0.225) = - 19.99 m.t.

Mde = 3 (- 3.838) - 13.5 (- 0.225) = - 8.48m.t.

The B.M.D. is thus as shown in Fig. 5.160.

I
I
429

Exam.ple 5.11 Construct the B.M.D. for the frame shown in Fig. 5.l7a
jf the moment of inertia varies as indicated.

b c
r l21

a
v. ,
I-- 3 -+-- 6m -I.-3...l (b)
(a)

(d) (e)B.M.D.

Fig. 5.17

Solution: In the previous examples on frames involving joint trar.~latioD,

all the members were normal to each other. This resulted in obvicus rela-
tionships between various member rotations. In the given frame, and
si.milar frames having inclined members" such relatiOI:!hiFS rue :r:ot cbvicus
and must be established. To do so, the members are imagined to be pin-
connected. Since the length of each member does not change and the
deflections are small, the displacement of one end of a member relative to
its other end is normal to the axis of this member. This leads to the joint
displacements indicated in Fig. 5.17b. It should be remembered that the
dashed lines in this figure represent the chords of the members after defor-
mation not the elastic curve of the frame. Triangle c e'e", enlarged in
Fig. 5.17c, shows the relationships between ".b> "be and "cd. Taking
~ as a reference value"

".b = "cd = " leos a = 5" 14 and "be = 2" tana = 3" 12
.p.b = .pcb = 5" 14x 5 =" 14 = .p
.pbe = - 3 " 12 X 6 = - " 14 = - .p
Note that "be is negative as member be rotates anticloc.kwise.
 430

Since no loads ac t on individual members, the fixed-end moments are

all zero. The relative stiffnesses are as folIows :
K,b : Kbe : Ked = 1/5 : 1.21/6 : 1/5 = 1 : 1
The slope-deflection equations are as follows :

M,b = 1 (a b - 3 of)
M b = 1 (2 ab - 3 of)
Mbe = 1 (2 a b + a e + 3 of)
Meb = 1 (2 a e + a b + 3 of )
Med = 1 (2 a e - 3 of)
Mde = 1 (a e - 3 of)

At joint b, M b + Mbe = 0
At joint c, Mea + MCd = 0

The shear equation requires special consideration as the horizontal

reaction components at a and d involve not only the end moments on
rrembcrs ab and cd and any lateral load acting on these members, but
also the vertical reaction components. Referring to Fig. 5.17d which shows
the free-body diagram a nd by considering the equilibrium of m~mber
ab,

Similarly, by considering the .e quilibrium of member cd,

Xd = (Med + M de )/4 - (M be + M eb)/8
Substituting these values into the shear equations, Xa + Xd + 3 = 0,
(M.b + M b.)/4 + (Med + M de )/4 - (Mbe + Meb)f4 + 3 = 0

Substituting from the slope-deflection equations into the three equations of

statics and simplifying)
4ao+ae =0 (i )

ao+ 4ae=0 (ii )

3of=2 (iii)

Solving equations, (i), (ii) and (iii) simultaneously, ab = 0, a e = 0 and

of = 2/3. Substituting these values into the slope-deflection equations,
Mab = M b = Med = Mde = - 2 m.t. and Mbc = Meb = 2 m.t.
The final B.M.D. is thus as shown in Fig. 5.17e.
 431

Esatnple 5.12 Construct the B.M.D. for the frame shown in Fig. 5.1Sa.
if the moment of inertia varies as indi cated.

("'l8.M.D.
(dl Yd
Fig. 5.1S

Solution: The displacement diagram is shown in Fig. 5.18b. The relative

displacements between the ends of each member are shown enlarged in
Fig. 5.1Se. Referring to this figure,
6 5 3
l:::,. b = = l:::. tan a = - 6
/:::.,.6. b = - - = - l:::,. and .6. bc
a c cos a 4 4
-56 36
-- = 1.33y. : .... y. : y.
4 X 3 4 X 5

Note that tPbc is negative as member be rotates anticlockwise.

12131
K,b: Kbc: K,d =- : - :- = 5: 10 : 9
335
The slope deflection equations are as follows:
M'b = 5 ( ab - 4 y.)
M b = 5 (2 ab - 4 y.)
Mbo = 10 (2 a b ao + + 3 y.)
M'b = 10 ( ab + 2 a, + 3 y.)
432

M.od = 9 (2 a e - 3 .p )
MOe = 9 (ae - 3 .p)

At joint b, Mba +
M"., = 0
At joint c, M.ob + Med = 0
Referring to Fig. 5.IBd which shows the free-body diagram, and by
considering the horizontal equilibrium of the whole frarnc3

M.b + Mbc M.od + Mde

3 + 4

Substituting from the slope-deflection equations into the three equations

of statics,
30~+lOae+ 10 .p = 0 ( i)
1O~+38ae+ 3 .p=0 (ii)
1O~ + 3 a e + 167.33 .p = 8 (iii)

Solving equations (i), (iil and (iii) simultaneously,

~ = - 0.016416, ae = 0.000468 and .p = 0.04878
Substituting these values into the slope dcllection equations,
\1,b = 5 (- 0.016416) - 20 (0.04878) = - 1.06 m.t.
Mba = 10 (- 0.016416) - 20 (0.04878) = - 1.14 m.t.
Mbc = 20 (-0.016416) + 10 (0.000468) +30 (0.04878)= +1.14 m.t.
M.ob = 10 (-0.016416) + 20 (0.000468) + 30 (0.04878)= +1.31 m.t.
M.,. = 18 (0.000468) - 27 (0.04878) = - 1.31 m.t.
MOe = 9 (0.000468) - 27 (0.04878) = + 1.31 m.t.

The final B.M.D. is thus as shown in Fig. 5.18 e.

5_10 Applications to fralnes with lDuitiple degrees or reedoJn in

translation
The principles discussed in the preceding section in regard to frames
with a single degree of freedom in translation are applicable to frames
with any degree offreedom. In this case, however, there will be as many
shear equations as there are independent joint translations. The difficulty
associated with such problems arise mainly from setting the necessary
shear equations and the simultaneous solution of the relatively large
number of equations involved. Since solving of equations is of no concern
in this book, only the slope-deJIection equations and the necessary condi-
tions of equilibrium will be given in the following examples.
 433

EsaInpJe 5.13 Write down, without solution, the slope-deflection equa-

tions and the conditions of equilibrium nec~'ary for a complete analysis
of the two-storey frame shown in Fig. 5.19. The relative stiflilesses of the
various members are as indicated.

111m
21
c K=2 d

3m 1 1
2tJm
41
b 4 e
4 3 3

f
.~
6m -----l

Fig. 5.19

Solution: From equations 5.1 and 5.2, the degree of freedom in rotation
sr = 4 and the degree of freedom in translation St = 2. Denoting the
rotation of members ab and fe by PI and the rotation of members be and
ed by "'2 the slope-deflection equations will be as follows :
M~b = 3(ab - 3 "'.)
Mba = 3(2 ~ - 3 "'.)
Mbc = I(2 ~ + a c - 3 "'2)
M.., = I(2 a c +~ ~ 3 "'2)
M", = - 6 + 4 (2 ~ + "')
M.b = + 6 + 4 (2 ac + ~)
Med = - 3 + 2 (2 a c ad) +
M de =+3+2(ad+ a e)
+
Moe = 1 (2 ad a. - 3 "'2)
M.d = I (2 a. +ad - 3 "'2)
M.r = 3 (2 a c - 3 "'.)
Mr. = 3 ( a c - 3 "'.)
The equations of equilibrium should be equal t6 the combined degree
offreedom = 4 +2 = 6. Four equations are provided by the joint equi-
librium conditions at b,c,d and e, and two equations are provided by the
horizontal equilibrium of each of the two storeys. Thus, the required
conditions are :
 434

M b, + Mbo + Mbe = 0 (i )

MOb + Mod = 0 ( ii )

Mdo+M",,=O (iii)

Med + Meb + Mer = 0 ( iv)

+M.d + 2 = 0 (v)
3

- -",,-b....:+---C.M.=ba + Mer + M re + 2 + 4 = 0
M (vi)
4 4

Example 5.14 Write down the shear equations required for the determina-
tion of the independent member rotations in the frame shown in Fig. S.20a.

e f p e f P
I
~ b'f---''--ic c c
hz

d gad 9
_+1_"_ Lz ---j" Xo Xd Xg
(a) (c) ( b)
Fig . 5.20

Solution: From equation 5.2, the degree offreedom in translation Sl = 2.

Joints a,d and g are fixed in position while joints band c may move hori-
zontally an amount L~l.l and joints e and f an amount .620

Form 5.6 of the slope-deflecuon equations is to be used for all the verti-
cal members and form 5.7 for the horizontal members. Further, it is
necessary to have two shear equations in order to determine the two inde-
pendent displacementS.6.1 and 6.2'
Considering members ec and fg as free-bodies,

__ M",o",e_+
_ -'-.M,-,e"o Mr. + Mgr
X - -+ and X. = ) -+
o hI (hI + h2
 435

Referring to the free-body diagram of part cefg shown in Fig. 5.20b,

Xc + X.-P=O (i )
Considering members ab and cd as free-bodies,

Referring to the free-body diagram of part abcd shown in Fig. 5.20c,

r The two shear equations (i) and (ii) in addition to the joint equilibrium
(ii )

equations are sufficient to determine the four unknown joint rotations;

Ub , c
U > ue and Uf and the two independent displacements ~I and 62'

5.1 I Applications to statically indeterlDinate CraDles with support

IDoveUlents
The slope-deflection method is applicatle to the analysis of statically
indeterminate frames with known support movements. In this regard,
form 5.4. of the slope-deflection equations is used. On so doing, the
absolute values of member stiffnesses should be used and the known joint
and member rotations should be expressed in radians. Further, it sh(;lUld
be noticed that in the absence of the applied loads all the fixed-end mom-
ents are zero.
Consider, for example, the frame shown in Fig. 5.21, and let it be
required to determine all the end moments resulting from a clockwise
rotational slip at support a = 8 and a vertical settlement at support d = L1.

c e
21

d f
~ !~--
.... 2L ~I~ L 3~
Fig. 5.21

Before wntmg up the slope-deflection equations, it is noticed that

Ua = + 9 and that .pbC = t:. /2L and .p~ =
t:. /L. Thus, one is
-
left with three unknown joint rotations; ab' U c and uc' and one unknown
member rotation .p ; .pab = .pcd = .p,f =.p. The values of these un-
knowns may be found from the three conditions of equilibrium at joints
b, c and e and the shear equation.
 436

Example 5.15 Analyse the frame shown in Fig. 5.16a in the absence of
load if support a has a clockwise rotational slip = 0.002 rad. and support
d settles by 4 cm. EI = 12000 m 2 t.

Solution: The absolute (2 EI/L)-values of the various members are first

evaluated and are as follows:
2 EI 2 X 12000
member ab'L = 6 = 4000 m.t.

2 EI 2 X 12000
member bc, - - = = 2000 m.t.
L 12
2 EI 2 X 12000
member cd'L = 4 = 6000 m.t.

Member rotations:
4
and
"'be = 1200
Here it is known that ua = 0.002 and + ad = O. Thus, the s1ope-deflection
equat ions are as follows :

Mab = 4000 [2 (0.002) + Ub - 3 (6~O)]

Mba = 4000 [ 2 ab + 0.002 - 3 (6~0)]
Mbe = 2000 [ 2 Ub + Ue - 3 C:OO)]
Meb = 2000 [ 2 Ue + Ub - 3 C:oo)]
Med = 6000 [ 2 ae - 3 (~O)]
Mde = 6000 [ Ue - 3 (~O)]
Joint equilibrium equations:
At joint b, Mba + Mbe = 0

At joint c, Mi,:b + Mcd = 0

 437

Shear equation:
_M_a",bc-+-,--_M--,b",a Mcd + Mdc
6 + 4 = 0

Suhstituting from the slope-deflection equations into the three equations

of statics.. .
600 nb + 100 nc - f::, = 0.6 (i)
100 nb + 800 nc - 2.25 f::, = 1.0 (ii)
- nb - 2.25 nc + 0.01458 f::, = 0.002 (iii)
Solving equations (i), (ii) and (iii) simultaneously,
nb = 0.0017, nc = 0.0031 and f::, = 0.7316

0.7316)
Mab = 4000 [ 2 (0.002) + 0.0017 - 3 ( 600 ] = + 8.4 m.t.

Mba = 4000 [ 2 (0.0017) + 0.002 - 3 C::;6)] = + 7.0 m.t.

Mbc 2000 [2 (0 .0017) + 0.0031 - 0.01] =- 7.0 m.t.

MCb 2000 [2 (0.0031) + 0.0017 - 0.01] = - 4.2 m.t.
0.7316)
MCd = 6000 [ 2 (0.0031) - 3 ( 400 ] = + 4.2 m.t.

Mdc = 6000 [ 0.0031 - 3 (0.~~~6)] = - 14.4 m.t.

S.12 Special sUnplifying technique

The application of the slope-deflection equations to the analysis of
beams and frames has been general with no reference being made to any
special procedures that may be taken to reduce the labour involved. Such
procedures, however. . are possible and exist when:
(I) hinged supports are present,
(2) there is symmetry in the structure .

~
i
The corresponding simplifications will be considered separately below:
(I) Modification for hinged ends

a
..b....
1 I D~
I L I

Fig. 5.22
438

Consider a member ab having a hinged support at a as shown

In Fig. 5 .22 . The basic slope-deflection equations for this m'l1lber are:
F 2 EI
Mab = M ab + L (2 aa + ab - 3 .p)
2 EI
Mba = MFba + L (2 a b + aa - 3 of)

Since Mab = 0, the first equation may be fe-arranged to give

aa =-MFab (4LEI)- ~b+~ of

Substituting this value of aa into the second equation,

Mba = ( M ba
F M\) 3
---+ EI
+L(ab - of) ... 5.8

Equation 5.8 may be fe-expressed as :

3 EI
Mba = M~a + L ( a b - .p ) ... 5.9
where ... 5.10
-F
A study of equation 5.10 shows that Mba represents the fixed-end moment
at end b when end a is hinged. Also, M~b adds to the numerical value
of M~a. For example if beam ab shown in Fig. 5.22 carries a uniformly
wL2 ( WL2) wL2
distributed load, M~a= I2 - 2 X 12 = -8-

Equation 5.9 expresses Mba in terms of ab and if; without involving

aa' This results in one less unknown to be determined.To illustrate the
simplification resulting from the appl ication of equation 5.9, two problems
previously solved using the basic slope-deflection equations will be con-
sidered.

Exal11pie 5.16 Re-analyse the beam in example 5.3(Fig. 5.6) using the
modified slope-deflection equations.
\
Solution: The slope-deflection equations are:
2 X 62
Mba = - - + 3 X 5 ( ab) = + 9 + 15 a b
8
M b, = - 8 + 2 X 6 (2 a b + aJ = - 8 + 24 a b + 12 a,
M,b = 4 + 2 X 6 (2 a, + ab) = 4 + 12 ab + 24 a c
M,d = - 8 +3 X 5 (a c) = - 8 + 15 a,
 r
I

,I
439

I Two unknown rotations appear in the slope-deflection equations. These

rotations may be determined from the equilibrium conditions at joints b

!, and c; Mba + Mbc = 0 and MCb + Mcd = O. Thus,

39 ab + 12 a c = - I
12 ab + 39 a c = 4
(i)
(ii)

Solving equations (i) and (ii) s~multaneously, ab= - 0.(63 and a c=0.122.
Substituting these values into the slope-deflection equations,
i
Mba = + 9 + 15 (- 0.(6)) = + 8.05 m.t.

I M bc = - 8 + 24 (- 0.(63) + 12 (0.122) = - 8.05 m.t .

MCb = + 4 + 12 (- 0.C63) + 24 (0.122) = 6.18 m.t.

I Mcd = - 8 + 15 (0.122) = - 6.18 m.t.

,I These m,ments are identical to those obtained by the lengthy solution

presented in example 5.3.

I Example 5.17 Re-analyse the frame in example 5.7 (Fig. 5.11) using the
modified slope-deflec tion equations.

Solution: The slope-deflection equations are:

Mba = + 4.5 + 3 X 4 ( ab)
MbC = - 16 + 2 X 3 (2 ab)
MCb = + 16 + 2 X 3 ( ab)
There is but one unknown rotation which may be determined from the
equilibrium condition at joint b; Mba + MbC = O. Thus,
ll.5
- ll.5 + 24 ab = 0, from which ab = - = 0.48
24

Substituting this value into the slope-deflection equations,

Mba = + 4.5+ 12 (0.48) = + 10.25 m.t.
M bc = - 16 + 12 (0.48) = - 10.25 m.t.
Meb = + 16 + 6 (0.48) = + 18.87 m.t.
f These values are identical to those obtained by the lengthy solution presen-
ted in example 5.7.

I (2) Allowance for symmetry

When a structure is symmetrical" rotations of the various joints on
one side of the axis of symmetry are related to those of the corresponding

I joints on the other side of the axis of symmetry. The type of loading then

I
440

deoides the nature of this relationship. Loads may be unsymm

etrical,
symmetrical, or antisym metrica l. Simplification results from
the latter
two cases of loading . This is illustra ted with reference to Fig.
5.23.

(a)

a
~j
()(
~ ----k b~
~I
OI b
I~
oCc~
c
l '-i
a old
(b)

r c

(c)

(d)

P b c PI2

=
d
7-
'"
(e)

Fig. 5.23

Symm etry: A study of the elastic curve of the structu res in Figs
5.23a
and b, which have a centre span, shows that Ba = - ad and
~ = - Be
and thus there arc two unknow n rotatio ns only instead
of four. In
Fig. 5.23 c, which shows a structure with no centre span, aa
= - ae
 441

and abl = - abr. Since, however, the elastic curve IS continuous, the
only way for Ubi to be numerically equal and of opposite sign to 0br is
that both of them to be equal to zero. In other words, the given beam
may be considered as one~span beam simply supported at one end and
fixed at the other end.
Antisymmetry : It is seen in Fig. S.23d that f ua = ad and U b = U c
Similarly, in Fig. S.23e, Ub = .. nc. Thus, in each of these two structures
the number of unknown joint rotations is reduced by one half
Unsymmetry : In general, loads are unsymmetrical, but any unsym-
metrical load system may be divided into a symmetrical and an antisym-
mctricalload systems, to which the above simplifications apply. Consider,
for example.. the frame shown in Fig. S.24a. The given loading is equi-
valent to the sum of the symmetrical loading shown in Fig. 5.24h and the
antisymmetricalloading shown in Fig. 5.24c. The original frame involves
the solution of six simultaneous equations in six unknown rotations; a~

w, w, fl/2
R R/2 Fl/2 Fl/2
e d

w
R P'/2 R/2 P,/2
tl e
+
Q f
" (a) (tl)
" " (e)
Fig. 5.24

a c' !fab (= !frc) and !fbc ( = '" cd)' The frame under symmetrical
a c ' ad'
, loading involves the solution of two simultaneous equations in two
unknown joint rotations; ab (= ~ a e) and a c (= ~ ad)' The
frame under the anti symmetrical loading involves the solution of four
simultaneous equations in four unknown rotations; ' ab (= a e)., a c (= ad)'
r
~
"'ab (= "'re) and "'b, (= "'cd)' Thus, the original problem which
involves the solution of six equations is replaced by two component pro-
blems; the first involving the solution of two equations and the second

l
,
four equations, and it is this procedure which causes a reduction in the
computation time.
442

Example 5.18 Re-analyse the frame in example 5.6 (Fig. 5.IOa) making
use of the symmetry in the frame and loading.
Solution: From symmetry there is no joint translation and a.b =- o.c'
The slope-deflection equations are :
M ab =O + 4("b) ,M.b = 4 db

M ba =O+4("b) ,Mba = 8a b
Mbc = -13 + 5 (2 "b- "b) , Mbc = - 13 + 5 "b
These moments are expressed in terms of a single rotation ab' This
is found from the equilibrium condition at joint h, ~A:ba + Mbc=O. Thus,
8 "b - 13 + 5 "b = O,from which "b = I.
By back-substitution, into the slope-deflection equations, Mab = + 4,
Mba = + 8 and Mbc = - 8 m.t.
These moments are identical to those previously obtained in exampleS.6

Example 5.19 Re-analyse the frame in example 5.11 (Fig. 5.17a) making
use of the antisymmetry in loading.
Solution: From the anti symmetry of loading, a b = o.c and one is left.
with one unknown joint rotation ab and one unknown member rotation "'.
The slope-deflection equations are :
Mab = (ab - 3 op)
Mba = (2 "b - 3 op)
Mbc = (2 "b + "b + 3 op )
MCb = (2 "b + "b + 3 op)
Mcd = (2 "b - 3 op )
M.c = ("b- 3 op)
The two unknown rotations involved are determined from the equi-
librium condition at joint b and the shear equation.

Mba + Moc = 0

Sub3tituting from the slope-deflection equations into the two equations

of statics and solving simultaneously, "b = 0 and op = 2/3 .
Sub';tituting these values into the slope-deflection equations,
Mab = Mba = Mcd = Mdc = - 2 m.t. and Mbc = Mcb = 2 m.t
These moments are identical to those obtained by the lengthy solution
presented in example 5.11.
 EXAMPLES TO BE WORKED OUT

(I) - (II) Using the method of slope-deflection draw the B.M.D. for
each of the beams shown in Figs. 3.32-3.42. EI = constant-

(12)- (17) Re-solve problems I, 2, 3, 5,7,9 if the moment of inertia

varies from span to span as follows:
Problem I (Fig. 3.32), lab: Ibc : I'd = 2 : 1.2 : I

Problems 2 and 3 (Figs. 3.33 and 3.34), lab: I b, : red = 1.5 : 2 : I

Problems 5 and 7 (Figs . 3.36 and 3.38), lab: I b, = 1.5 : I

Problem 9 (Fig. 3.40), lab: I b, : I,d = I : 2 : I

(18) -(20) Determine the moments at the supports of the beams .in
problems 1,4 and 6 in the absence ofthc applied loads due to the follow-
ing support movements:

Problem I (Fig. 3.32), 3b = 2 cm. and 3, == 0.5 cm. EI == 4000 m 2t.

Problem 4 (Fig. 3.35), i - 3b = 2.5 cm.

ii - aa = 0.005 cm. EI = 4000 m 2t.

Problem 6 (Fig. 3.37), i - a , = 0.0045 rad. anticlockwise.

ii - 8b = 3.6 cm. EI = 5000 m 2t-

(21)-(24) Using the slope-deflection method, draw the B.M.D . for

each of the frames shown in Figs. 3.45-3.57 if the moment ofinertia varies
as indicated.

(25) -(42) Using the slope-deflection method, draw the B.M.D. for
each of the frames shown in Figs. 4.23, 4.24, 4.25, 4.27, 4.28, 4.30, 4.32
and 4.34 if the moment of inertia varies as indicated.

! 443
!
l
 (43)-(44) Using the slope-deflection method, draw the B.M.D. for
each of the frames in Figs. 5.25 and 5.26. Check your results by any con-
vement method.

,
2m 211m
21
!
211m
I ! Ii

r
-:J

4
~ 1
21

! 4

8m
wk I 12m

Fig. 5.25 Fig. 5.26

(45)-(52) State the degrees of freedom in rotation and translation of

each of the frames shown in Figs. 5.27-5.34, then write down the equations
of statics (in terms of member end moments) which are necessary for the
determination of the independent joint and member rotations.

H, d e ,- c f

H
f: b
I c H h'b d
2

f e
,.
~L, --/-L2 ---l " L L --t

Fig. 5.27 Fig. 5.28

H b C e

e
L, -l-L2 --I

Fig. 5.29 Fig. 5.30

 445

H c d d f

t w
f
1

r p
b 1 h3
h2 h2
J
Ll L2
Fig. 5.31
-l 4Ll~L2 Fig. 5.32
h
fi. L3 -l

e f
H

~
~ d e
d
1 H
H':l! b c f 9 c
H 4h

h2 H b

'" "
I-L-I-L-I-L-\-L-J
h
"
Q

'"
I---- L ----I

Fig. 5.33 Fig. 5.34

 CHAPTER 6
f THE MOMENT DISTRIBUTION METHOD

6. I Intrcdllction
The method of moment distribution was introduced by Professor
Hardy Cross in 1932. It is applicable to the analysis of statically indeter
minate beams and frames, and is particulaz1y useful in the case of struc-
tures with high degree of statical indeterminancy.
The moment distribution method starts from the same assumption
made in the slope-deflection method of a structure whose joints are held
against rotation (and translation). Allowance for joint rotation, however,
is made by a process of successive approximations which can be carried
out to any desired degree of accuracy. This has the big advantage
of avoiding the solution of a number of simultaneous equations involved
in the slopedeflection and other methods of analysis, which usually forms
a major part of the computational work. The sign convention adopted

~
is the same as that previously used in conjunction with the slope-deflection
method, i.c. clockNise end moments, joint and member rotations are con~
sidered positive and vice versa.

t Before proceeding with the details of the process of moment distribu-

tion, it will be advantageous to define some terms and establish some rela-
tionships for prismatic members which will be repeatedly used during the
t course of analysis.
t 6.2 Absolute stiffness, K
I The absolute stiffness at end a of a member ab is the moment required

t ~. M~ Ma.

~~==L:::::::==----=---'~ Cfy;-L-5t
r (a) (b)
,I
,
j;
~-o-r---L--x-rMo<-:'~O> Cf-.:-< ~.
, (c) (d)

I:- Fig. 6.1

447
I

l -
448

to produce a uni t rotation at end a while end b is supported in some parti- .

cular manner, usually fixed.
The absolute sti.Jlless of a prismatic member is derived with reference
to Fig. 6.la. T he figure shows a prismatic member ab subject to a moment
Mab at end a while end b is held liKed. The relationship between the
applied moment and the rotation at end a is readily obtainabie from the
slope-deflection equations:

F
=Mab 2 EI
+ - - (2aa +ab) ... 6.1
L

2 EI ( )
= Mba +
F
----r:-- 2 ab ;t- aa .. 6.2

Since no load is applied between a and b, ~b (= M~a) = O.

Also Since end b is fixed, ab = O. Thus,
Mab = 4 EI aa/L ... 6.3
With aa = 1, the value of the absolute stiffness becomes:
K'b = 4 EI/L .. 6.4
i.e. the moment required to rotate end a of a prismatic member ab through
a unit rotation while end b is being liKed is 4 EI/L.

6.3 Modified stiffness, K'

Very often, structures p:Jssess special features. lWhen recognized and
the necessary modifications made, these special features lead to a simpli-
fied solution.

I. Modification for hinged ends

Fig. 6.1h shows a prismatic member ab subject to a moment Mab
at end a while end b is simply supported. The relationship between the
applied moment and the rotation at end a is best found from the slope-
deflection equations.
Since end b is simply supported, Mab = 0, and from :equation 6.2,
ab = - aa/2. Substituting this value into equation 6.1,
M1b = 3 EI aa/L ... 6.5

With 4a = I, the modified stiffness becomes :

K 'b = 3 EI/L .. . 6.6

i.e. the moment required to rotate end a of a prismatic member ab through

 449

a unit rotation while end b is simply supported is 3 ElfL.

Feom equations 6.4 and 6.6,7

K"b =+K.b ... 6.7

Le. the modified stiff'less at end a of a prismatic member ah when end

b is simply supported is three quarters the corresponding absolute stiffness.

(2) Modification for symmetry

Fig. 6.le shows a prismatic member ab subject to a moment Mab at
end a and a moment at end b, Mba = - M ab " The relationship between
the applied moment and the rotation at end a is best found from the slope-
deflection equations.
From symmetry of loading t, = - aao Substituting this value into
equation 6.1,

M' b = 2 EI aa/L ... 6.8

With a. = I, the value ofthe modified stifFaess becomes:
K'ab = 2 ElfL ... 6.9
i.e. the moment required to rotate end a of a prismatic member ab
through a unit rotation while end b is subjected to an equal and opposite
moment is 2 EI/L.
From oquations 6.4 and 6.9,

K ' ab= 2K.b

I
... 6.10

i.e. the modified stiff.less at an end of a prismatic member ab subjected

to symmetrical end moments is reduced to one half the corresponding
absolute stiffness.

(3) Modification for antisymmetry

Fig. 6.ld shows a prismatic member ab subject to a moment Mab at
end a and a moment at end b, Mba = M ab . The relationship between
moment Mab and the rotation at end a is best found from the slope-
deRection equations.
From antisymmetry of loading ~ = aa. Substituting this value into
equation 6.1,
M'b = 6 EI aafL ... 6.11
450

With 1, the value of the modified stiffness becomes,

aa =
K"b = 6 EljL ... 6.12
i.e. the moment required to rotate end a of a prismatic member ab through
a unit rotation while end b is subjected to a moment of equal magnitude
and acting in the same sense is 6 EI/L.
From equations 6.4 and 6.12"

K'," = ~ K' b ... 6.13

2
i.c. the modified stifihess at an end of a prismatic member ab subjected
to antisymmetrical end moments is one and halftimes the corresponding
absolute value.

6.4 Relative stiffness

Simplification always results from multiplying all the stiffllcsses of the
members by some constant quantity. The resulting values are then rela-
tive stiffuesscs. From equations 6.4 and 6.6, the relative stiffness of a
I
prismatic member fixed at the far end is proportional to - while that
L
of a prismatic member simply supported at the far end IS proportional to

~ U)
6.5 Carry-over factor
The ratio between the moment induced at end b to the moment pro-
ducing it at end a of a member ab is called the carry-over factor.
The carry-over factor for a prismatic member is best found from the
slope-deflection equations.
Referring to Fig. 6.la,
M' b = 4 EljL and M b 2 EljL
Cab = Mb,jM'b = Ij2 . .. 6.14
i .c. the moment carried-over from end a to end b of a prismatic member
ab; end b being fixed, is equal to one half the applied moment at a and
has the same sign.
Referring to Fig. 6.1h, it is obvious that no moment will be induced
a t end b due t o any applied moment at end. a. Thus. .
. .. 6.15
i.e. no m oment is carried-over from end a to end b of a member ab if
end b is simply supported.
 451

6.6 Distribution fac~rs, D

Consider three members 01" 02 and 03 rigidly connected at joint 0 and
let this joint be subjected to an external moment M as shown in
Fig. 6.2.

Fig. 6.2

M will cause the joint to rotate through an angle a. From the condition
that joint 0 is rigid"
... 6.16
Also for equilibrium"

M o' + Mo, + M o' = M ... 6.17

By definition,
M o ' = C K,a, Mo' = CK, a and M o' = CK, a ... 6.18
where Kh K2 and K3 are the relative stiff{lCSSeS o f members 01, 02 and
03, and C is a consta nt.
Substituting from eq uations 6. 18 into equation 6 .17..
C a (K, + K, + K,) = M
or C a = M/ (K, + K, + K, ) = M/ E K ... 6.19
Substituting for C a into equations 6 .1 8 ..
K, K2 K,
MOl = - M, Mo, = -M and Mo,= - M ... 6.20
EK EK rK
where the ratios KilL K , K 2 /r. K, .... are defined as the distribution
r factors.
Equations 6 .20 show that whenever an external moment is applied
to ajoint, it will be distributed among the members framing into thisjoint
in proportion to their distribution factors .
I
,

452

6.7 Outlines of the D10lDent distribution Dleth(ld

The general procedure of analysis is outlined below:
(1) Assuming the joints to be held against rotation (and translation),
the fi'C:cd-end mo:nents of individual members are calculated.

(2) In general, the fixed-cnd moments will be out of balance at the

various joints. Therefore, some. artificial locking moments will have to be
applied at the joints to keep them in equilibrium.

It should be noted that the moment required to lock a joint is equal

to the algebraic sum of the moments at the ends of the members connected
to this joint. This sum of moments is termed the unbalanced moment and
thus, the locking moment is given by :
locking moment at joint j = unbalanced moment at joint j

(3) B~ing artificial, the locking moments at the various joints do not
exist and must therefore be eleminated. This is done by applying a releasing
moment at the various joints each equal in magnitude and opposite in
direction to the corresponding locking moment. Thus,
releasing moment at joint j = ~ (unbalanced moment at joint .n
(4) Superimposing the fixed-end moments in step I to the moments
in step 3 due to the releasing moments will result in the final moments.
The fixed-end moments in step I are readily obtainable. Expressions
for these m::nlents due to common cases of loading are given in Fig. 5.3.
These vllues arc for prism3tic members. Fixed-end moments for non-
prisma.tic members may be found by any convenient"method. (See section
4.7).
The problem thus reduces to finding the moments throughout the
struc.ture due to a set of external (releasing) moments applied at the various
join~s" Moment diftribution as the name may suggest.. is a method for
accomplishing this goal.
The principles of the moment distribution process are best explained
in conjunction with a number of numerical examples. At first only beams
and frames composed of prismatic members and with no joint translation
will be discussed. Later.. when the student has grasped the basic ideas..
it will be an easy m3.tter to extend the method to structures with joint
translation and to those composed to non-prismatic members.
 453

DO joint
6.8 Applic ations to statica lly indete rmina te beaDls with
transl ation
The
Examp le 6.1 Draw the RM.D . for the beam shown in Fig. 6.3.
previously
moment of inertia varies as indicat ed. (This problem has been
solved by the slope-d eflectio n method - Examp le 5.1)

181 311m

~af.-- 3 1 ~I
! I !!

21 IC~
-I--- 3 , 6m ,I
K CD (1)
D.Fs. 'OJ 1113 2/31 ro
F.E.Mts. -6 .6 -9 .9
D.... ts. ..Q... .1 ,1 ...Q
C.O.Mfs. .0.5 .1

Fmol Mts -55 .7 _7 .10

10
55 7m! ~
B.M.D.~ ~
""'V7 'S= 7
Fig. 6.3

Fig. 6.3.
Solutio n: The solutio n is presented in steps with reference to
(1) The relative stiffness for members ab and be are:

K,b : K b, = .L : ~ =1 :2
6 6
The relative K-values are shown encircled in Fig. 6.3.
ing to
(2) The distrib ution factors (D.Fs.) are compu ted accord
the fixed end can be
D = KIf. K. At joints a and c, D,b=D ,b = 0 since
replaced by a membe r of infinite s~jffness.
I I 2 2
At joint b, Dba = - - = - and DbC = - - - - 3
1+2 3 1+2
r The distrib ution factors arc indicat ed in the boxes attache d
to the line

l
I
!
diagra m of the beam.

(3) The fixed-end momen ts (F.E.M ts.) are calcula ted and are as

PL 8 6
follows:

, X
M 'b -- - -8 = - -8- "" - 6 m.t. and M'b, = + t
6 m ..
,I
I
-----------------------------
454

WL2 3x6 2
M~ = _ - - =_ - - = - 9 m.t. and M~
12 12
= +
9m.t. ,-

These moments are indicated directly below the line diagram of the beam.

(4) The fixed-end moments in step 3 would not have developed unless
an artificial locking moment was applied at joint b .
Locking moment = + 6- 9 = - 3 m.t.

(5) The relea'ing moment = - (the locking moment) = + 3 m.t.

This m:>m3'nt is distributed to members ab and be according to the distri-
bution factors at joint h.
The distributed m:>rnents (D. Mts.) are indicated on the second line.

(6) A line is drawn bolow the distributed moments indicating that

the preceding moments are in equiJibrium.

(7) Hdf the distributed moments at one end of a member is carried

ov.::r to its other end. These cl.rried over moments (C.O. Mts.) are written
on a new line.
Steps 3,5 and 6 form a conplete cycle of mo:>ment distribution. Genera~ly
the curied over moments start a new cycle and as many cycles as neces-
sary for a rCl.son'1.blc degree of accuracy are carried out. It should always
be remembered that the last step in a cycle is to balance joints which can
rotate and carryover moments to joints which cannot.

(8) The final moments acting on the ends of the V'arious members
are obt3.ined by summing up the moments in the respective columns.
fhese moments are indicated on the last line.
(9) Having thus determined the moments acting on the ends of the
various members, it will be an easy matter to draw the required B.M.D.

Esan1ple 6.2 Analyse the continuous beam shown in Fig. 6.4 by moment
distribution. EI = constant.

Solution: The analysis of the beam is presented in Fig. 6.4.' In the pre-
vious example, one joint only was free to rotate. Thus, one cycle oflocking
and releasing the joint was ad~quate'Tor a complete and exact solution.
In this example joints band c are free to rotate. Mter the first cycle, the
carried over moments at joints band c are out of balance. These moments
can be looked upon as a new set of fheed-end moments and ,distributed
in the usual manner during a second cycle, and so on. In this example
 455
four cycles are carried out. It is seen that the carried over moments in the
four th cycle have become so small that they do not deserve further
dis tribu tion.

9mt 9mt
111m
~a1-4 C;~"
2 --l I
I

12m
I ! I
"~2:24 d~ -1
K
CD CD CD
D.Fs. 13 113 13 1/3
F.E.MIs- 0 -12 ,12 0
n.Mt!.. ~ LL- ~ ~
C.O.Mts. J. 0 - 2 ,2 0 -4
D.Mts. _0_ .1.33 .. 0.67 - 0.67 1.33 0 2
C.O.MIs. .0.67 ( J - 0.33 .0.33 0 - 0..67
n. Mts. _0._ .0.22 .0.11 - D.lt -0.22 0. 3
C.O.Mls. t 0.11 0 - 0..0.6 .0.06 0. - 0..11
D.Mts. _0_ .0.04 ~ - 0..0.2 ~o.D4 0. 4
C.O.Mts. ,0..0.2 -0..0.2
~
=
Final Mts. .1.80 -9.59 -9.59 .9.59 -9.59 -t.8D

Fig. 6.4

The symmetry in the beam and the loading suggests a simpler solution.
utilizing modified stiffness for member be as explained in section 6.3.
This simplified solu tion is presented in the following example.

Example 6,3 Analyse the symmetrical beam shown in Fig. 6.5 using modi-
fied stiffness for the central span. EI = constant.
I
9mt t

EA
9mt
~a
1tIm
! I ! I
I ~3 d~
1-4 ---12 I 12' m 121-4--J
K CD CD CD
K' CD -tX1:@ CD
D.Fs.
F.E.Mts.
t:tQ:[';-
-3
- ---1)0p.S!fD"f.2f-----_ _
0. -11
+___-J
D.Mts. ~ _,9_
.6 _. 1_4
C.O.Mts. .4.8
Final Mts. ":1.i
Fig. 6.5

: Solution : By using modified stiffness for the central span, the carry over
! moments across this span are actually eleminated. Also, from symmetry
of beam and loading the moments on each side of the centre line are

f
r
,

,.
,

456

symmetrical. Thus, it is possible to work with only one half of the beam
as shown in Fig. 6.5.
The student is advised to compare the relative ease of the solution pre.
sented above to that presented in the previous example for the same pro-
blem. He is recommended to adopt the later solution based on the use
of modified stiffness.

Example 6.4 Analyse the continuous beam shown in Fig. 6.6. The moment
of inertia varies as indicated.

311m
~__~____~~~IoID~C
21
-=t::'-- - 6 m
K 1
r1/3 2131 CD
O.Fs. lol .6 .S
IT
F.E.Mls -6 -9
D.Mts 0 .1 .2 -S C yel.1
e.O,MIS. .0.5 - 0- - 4.5 ~
D,Mls _0_ .1.5 ~ -=L- 2
e.O.MIS .075 0 -0.5 .1.5
D.Mls 0 .0.16 .0.34 _1.5 3
C.O.MIS. .0. 08 0 _0.75 .0.171-
O.Mls.
C.D.Mls.
_0_
.0.12
.0.25 .0.50
0 -0.08
~_
.0.25
4
1
.~
O.Mls. 0 .0.02 .0.06 5
C.O.Mls, .. 0.01 0 -0.12 0.03 1-
O.Mls 0 .0.04 .o.oa ~ 6
C.O.MIs. .0,02
~
= ~

,.6.97 -8.S7
FlOal Mts 4.52 0

Fig. 6.6

Solution: Following normal moment distribution procedure, the analysis

will be as shown in Fig. 6.6. It should be particularly noted that the simply
supported end c, as any other joint, is first locked such that the fixed-end
moments are deVeloped in member be. The fixed-end moment at c is
next eleminated during the course of moment distribution. It should also
be noted that the distribution factor at the simply supported end c = l.
This will be obvious if member bc is looked at as being connected to an-
other member of zero stiffness.
Although the above procedure is quite basic, it is seen that convergence
to a solution is rather slow. (Six cycles were carried out before the carried
over moments have become too small to influence the final result). This
is because the hinged end c of member bc keeps sending back relatively
 457

large carried over momen ts to joint h. Anothe r solutio n based

on the
knowle dge that the final momen t at the hinged end must
be zero
and utilizin g modified stiffness for member be is possible. This
solutio n
is presen ted in the followi ng examp le.

Examp le 5.6 Analys e the beam shown in Fig. 6.7 using modifie
d stiffness
for span be. The momen t of inertia varies as indicat ed.

18t 311m
~o
I--- 3 3 ' 1 1 1 1 1
21
6m ';
K G) CD
K' G) -fl(2=@
D.Fs. iol 0.4 0.6
F.E.Mt5. -6 .6 13 5
D.Mt!.. ~ ~ 4.5
C.O.M's.
~
final t-IIts. _4.5
~ ~

.9 - 9

Fig. 6.7

Solutio n: The analysi s of the beam is presen ted in Fig. 6.7.

It should be
noted that a modifie d stiffness of 3/4 X 2 = 1.5 is used for membe
r bc to
allow for the hinged end c. It should also be noted that a modifie
d fixed-
end momen t at joint b ( - 13.5 m.t.) is used to accoun t for the
hinged end
c. This modifie d fixed-e nd momen t may be calcula ted indepe
ndently
by momen t distribution as follows :
Assum ing both ends of membe r be to be fixed,
wL2 3x6 2
Mbo = - - - = - - - = - 9 m.t. and
Mob = + 9 m.t.
12 12
Now Mob cannot exist and should therefo re be elemin ated by
applyin g
an equal and opposit e momen t = - 9 m.t. The carried over momen
t from
c to b = ~ (- 9) = - 4.5 m .t. Thus, the final momen t at c = 0, as it
2
should be, and the modified fixed-e nd momen t at b = -
9 - 4.5 =
- 13.5 m.t. as indicat ed.
The studen t is advised to compar e the relative ease of the
solutio n
present ed above to that present ed in the previou s exampl e for
the same
problem . He is recomm ended to adopt the later solutio n based
on the use
of modifie d stiflhes s.
 458

Es:unple 6.6 Analyse the continuous beam shown in Fig. 6.8. EI = con-
stant. (This problem has been previously solved by the method of the
equation of three moments in ExamFle 3.5-Fig. 3.10).

~a r
1-- 2 --I- 2
~ I
2t/m
! ! ! !

6m ~
l~J
2...J
K 0) CD
K' CD fX2=@
D.Fs. 101 fill 1m! 11 tal
Rele-oslng joint c -6 .6 -3
- 15 -3 0
F.E.Mts. -4 .4 -7.5
D.Mts.
C.O.Mt~
~ .. 2.33 .1.17
~-
Finol Mts. - 2.83 .6.33 -6.33 -;) ::)

Fig. 6 .8

Solution: The analysis of the beam is presented in Fig. 6.B. The only
new point in this example is the manner of ha ndling the overhanging end.
The overhanging end is similar to the simply supported end in Example
6.5, but in determining the modified fixed-end moment at end b of member
be, it must be remembered that now the moment at c has a fixed value
1.5 X 2 = 3 m.t. However, instead of independently calculating the
modified fixed-end moment at end b (allowing for the simple support at
c and the applied loads on be including a clockwise moment of 3 m.t. at
c) , it is easier to first release joint c and allow it to rotate under the unbal-
anced moment there and then proceed with the moment distribution.
Further, si.nce a modified stiffness has been used for member bc, no
moments are carried over from b to c.

6.9 Applications to statically indeterminate frantes with DO joint

translation
The anl.ly-3ls of statically indeterminate frames with no joint translation
follows: exactly the same procedure presented in the preceding section
dealing with beams. In the case of frames, however, systematic arrange-
m ent o f th: calculations becomes m ore important. Many arrangements
h ave b~en suggested. . but the author prefers the arrangement indicated
in Fig. 6.9.

r- --------------------------------------------_________
 459

In this arrang ement distrib ution factors and momen ts are

written
paralle l to the membe r to which they belong . The values related
to one
end of a membe r are written next to this end on the clockw ise
side of the

00

c~

Fig. 6.9
membe r. Furthe r, the calcula ted values are written outwar
d from the
line diagram of the frame. This arrang ement is adopte d in the
followi ng
exampl es and the studen ts are advised to do the same in order
to reduce
the possibi lity of error.

Examp le 6.7 Analys e the frame shown in Fig. 6.lOa. The

momen t of
inertia varies as indicat ed. (This problem has been previou sly
solved by
the slope-d eflectio n method in Examp le 5.6 - Fig. 5.10)

5t .
1.5t1m
b

r
5
21
c

d
4 -1--4 m ----I "

l
(a) (b)
Fig. 6.10

, Solutio n: The analysi s is present ed in Fig. 6.lOb. A modifie

d stiffness is
used for membe r be to allow for symme try. The relati'le stiffnes
ses are thus
as follows :
460

K'b : Kbc = ; : }( 2:) = 8 :5

= K/'f. K
The distribu tion factors are calcula ted from the relation sbip, D
and are' as follows :
8 8 5 5
Db = - - = - a n d D bc = - - =13 -
8 5 13+ 8 5 +
The fixed-end momen t at end b of membe r bc is :
2
M' = _ (WL 2 + PL) = _ (1.5 X 8 +5 X 8) = _ 13 m.t.
b
c 12 8 12 8

Note that since a modified stiffhess for member be was used)

it has been
possible to work with only one half of the frame.

constan t.
Examp le 6.8 Analys e the frame shown in Fig. 6.lla. EI =
n method
(This proble m has been previou sly solved by the sloped eflectio
in Examp le 5.7 - Fig.5.11).
.18.875
:-r.m
-0- '
.16
3 tim

r
3
b c~

41
CD
3

La
8m
(bl
(al
Fig. 6.11

d stiffness
Solutio n: The analysis is presen ted in Fig. 6.lIb. A modifie
The relative stif-
is used for membe r ab to allow for the hinged end at a.
fnesses are thus as follows :

K.b: Kbc =+(+ ):-f = I: I

DbC = 0.5.
The distribution factors at joint b are thus: Dba = 0.5 and
allow for
A modified fixed-e nd momen t is used at end b of membe r ab to
 461

th .
e hinged end at a, M~ = + - 4x6
- + - 1(4-X6)
- = + 4.5m.t.
.' 8 2 8
wL2 3 x8 2
M"bc = - 12 = - -12 - = - 8 m.t. and ~b = + 8 m.t.
Example 6.9 Analyse the frame shown in Fig. 6.12a. The moment of
inertia yuies as indicated. (This problem has been previously solved by
the slope-deflection method in Example 5.8-Fig. 5.12)

12m
(a)

o
~
~~:::~
o
NO.4
ci ~o d_
- 18 8
I I I _4
..9-
- 2 .. 1.6
_0_ r:;, -0.64
.0.32 0.0.6
-0.32
CD
....lL.-
- 0.16
:iii.48 -10.56 I / /"'1-4
o ooho~o~ ~
I (b)
Fig. 6.12

I Solution: The analysis is presented in Fig. 6.12h. A modified stiffness is

used for member ce to allow for the hinged ::nd at c. The relative stiffnes-

I ses are thus as follows :

= ~:~ : ~(.!..) = 4:4:2:1

L K.b:K",,:Kbd:K,
e 12 8
1.51:
6 4 6
These values are shown encircled. The distribution factors are calculated

II from the relationship: D = K/I:. K and are as follows :

462

4 4
At joint b, Dba = 0.4, Db< =
4+4+2 4 + 4 + 2 = 0.4
2
and Dbd = = 0.2
4+4+2

At joint c, Dcb = _4_ = 0.8 and Dec = _1_ = 0.2

4+1 4+1
Note that Dee = 0 since the overhanging member has zero stiffness.

The fixed-end moments are:

1.5 X 122
18 m.t. and M~a = + 18 m.t.
12
2
1.5 X 8 = _ B m.t.
and M~b = + 8 m.t.
12

It should be noted that over only three cycles of moment distribution,

the largest moment agrees with its corresponding value obtained by the
slope-deflection method to within 0.2%.

6.1 0 Applications to fraD1es with a single degree of freedom in

translation
In the preceding section, the application of the moment distribution
method has been restricted to frames with no joint translation. As explained
in sections 3.6 and 5.9, joint translation is generally prevented due to
either complete symmetry in frame and loading or the disposition of the
supports.
Before proceeding with the analysis of frames with joint translation"
it will be necessary to develop expressions for the moments developed at
the ends of a prismatic member due to a relative joint translation while
both ends are being held against rotation.

Fig. 6.13
i
I-
463

,
r
Fig. 6.13a shows a fixed-end prismatic beam abo It is assumed that
end b is forced through some displacement!:::. relative to end a, while both
ends are being held against rotation. The relationship between the fixed-
end moments developed at ends a and b on one hand and the relative
!
displacement between the two ends on the other hand may be easily found

I
from the moment-area theorems or more readily from the slope-deflection
equations:

M,b = M~b + 2 :1 (2 a, + ab _ 3 b.) ... 6.21

L

Mba = M~, + 2:1 (2 ab + a, _ 3:) ... 6.22

Since no load is applied between a and b, M: b = M~a = O. Also,

since both ends are held against rotation, U a = Ub = O. Thus,

Mb = _ 6EIb. .. . 6.23
a L2

If one end is hinged as shown in Fig. 6.13 h, the modified fixed-cnd

moment at the other end may be found by moment distribution as follows :"
Assuming both ends to be fixed~ then according to equation 6.23~
- 6 Elb.
L2
Now if end b is hinged~ Mba cannot exist and should be eleminated
6 EI b.
by applying a moment = + at b. The carried over moment
L2

from b to a = + -I (6 EI b.) = + 3 EI f\, . Thus.. the modified

2 L2 L2
fixed-end moment at end a allowing for the hinged support at b is given
by :
- 3 El b.
... 6.24
L2
In order to illustrate the method of handling joint translation~ the
I simple frame shown in Fig. 6.14 is considered.
The applied loads shown in Fig. 6.14a are equivalent to the two load

.L
I
- systems shown in Fig. 6.14h and c .
The first load system consists of the applied loads in addition to some
artificial restraining force Ha preventing~ the lateral deflection or sway of
 464

the frame. With joint translation thus prevented, the resulting non-sway
moments may be determined by moment distribution as for any structure in-
volving no joint translation. Next the artificial restraining force H . may
be calculated by considering the equilibrium of the frame.

b C

d
z

(al (bl (c)

Fig.6.14

The second load system consists of a single force equal in magnitude

and oDoosite in direction to Hao Once the as~ociat(d moments, whic.h will
be referred to as the actual sway mument.s , are found the final moments
will be given by :
Final moments = Non-sway moments + Actual sway moments
At this stage" the student is supposed to be capable of determining the
non-sway moments and the problem thus reduces to finding the actual
sway moments.
A direct solution for the actual sway moments is not possible. However
they may be found indirectly by first determining sway 11lIJ17lenls emsi')tent
with some force He. having the same line of action as Hal and then finding
the actual sway moments by direct proportion. The procedure is outlined
belo
(I) Joints band c are forced through some displacement I:; while
being held against rotation. The corresponding fixed-end moments deve-
loped at the ends of the columns can be determined from equation 6.23.

These moments are proportional to (~~) . Since I:; is arbitrarily

chosen, any set of numerical values for the fixed-end moments may be

chosen so long as they are proportional to their respective values of (I 1:;)

LZ
(2) The frame is held against any further translation while joints b and
 465

c are allowed to rotate. The resulting sway moments are determined by

moment distribution as for zny structUJe involvir g r.o joint trznslaticn.

(3) The ""Iue of the force He c;>.Using the chosen displacement !!, and
consistent with the sway moments calculated in step 2 is determined by

r considering the equilibrium of the frame.

(4) Since the artificial restraining force cannot exist~ the actual sway
moments should correspond to a force which when added algcbraicly to
the artificial restraining force the latter is cancelled out. This condition
is expressed mathematically as ;
H. xJf" = 0 ... 6.25

where x is a factor to be applied to the sway moments determined in step 2

to obtain the actual sway moments"
Actual sway moments = x (sway moments)
It remains to mention that the positive sign in equation 6.25 is used
if both Ha and He have the same sense and vice versa.
The procedure outlined above will be illustrated by a number of
examples.

ExaJnple 6.10 Analyse the frame shown in Fig . 6.15a. EI = constant.

(This problem has been previously solved by the slope-deflection method
in Example 5.10 - Fig. 5.16)

Solution: The analysis is presented in steps.

(1) The relative sti1f'nesses and distribution factors arc evaluated.

K'b : K"" : Ked = ] - : ~ : 2. = 2 : I :3

6 12 4
These stiflnesses are shown encircled in Fig. 6.15 b.
2 2 I
At joint b, Dba = -- = - and Db< = - - - = -
2 +I 3 2 + I 3
I 3 3
At joint c, Deb = --- - = --- = -
I + 3 4 I + 3 4

These distribution factors are shown in the boxes attached to the line
diagram of the fr ame in Fig. 6.15 b .
(2) Assuming that the frame is pr evented from sway by a horizontal
force H., the fixed-end moments due to the applied loads are calculated.
 rp::r:::o:=r::.iI::i:I:r::r===ll2 tim

3 4m
61

12m
(a) .20.82
:orr
.. 0.5
:o:sr
.. 3.25
:--s-
.24
114 Ha0.991!

:::I~
"';..
I ~ ~ 113
mOONO:~N_24
..
I .... .. 6.5
N~<O60 I::JNO~OM
~
I I I
I ->
I 0
~
0:0:4
1
oool",ol~ol~l~
~
o~ol~ ~
I I I I

\V
I ..
.-0
..
IMI'
d ;;;"
....

(b) Non-5way MIs.

.6.6
:aT
_0.4
-0.4
.1.7
.5.S
o

~~
ul
I"
I'" I'"
0.....: 0
I
tD~
.. I
~ 113
N 0
+3.3
:rr-
or:~ol~ol~ol~l~
-0.9
-0.2
.0.07 . I I

:sm l~
0 I:l> 1:gIUl
0 0 MODCOf'-o
I" I .. I

(c) Sway MIs.

_ 0.89 0.991 I
.0.68 -0.68 .. 0.89

.1.96

.t.Ol
(d)Aclual Sway MI5.
Fig. 6.15
I
I
467
I -
6 x 6
-B- = - 4.5 m.t. and MFb, -- + 4 5 m .t.

2 X 122
12 = - 24m.t. and MF'b = + 24 m.t.
r
(3) The fixed-end moments in step 2 are distributed in the usual manner
to obtain the non-sway moments shown in Fig. 6.15b.

I (4) The artificial restraining force Ha is calculated by considering

the eq uili bri urn of the frame as follows !
3.13+ 19 .77
----'----'----'- -
6
3 = O. BI 7t . _

Xd = 10.41 + 20.82 /
7.B08 t. +-
4
L X= Ha + 6 + 0.BI7 .L 7.808, H, = 0.991 t. _
(5) The frame is forced through some displacement b. and the corres-
ponding fixed-end moments are calculated .
F 6 E1 b.
M=b = Mba = - 62 ~ say - 10

6 E1 b.
M~d = M~c = _~-' say-22.5

(6) The fixed-end moments in step 5 are distributed in the usual

manner to obtain the sway moments shown in Fig. 6.I5e.
(7) The consistent straining force H, is calculated by considering the
equilibrium of the frame as follows:
Xa 7.54 + 5.07 = 2.lt . +-
6
x. = _14_..:..c
55'-+ .:......::6~.6 = 5.29 t.+-
4
LX = 0=H,-2.1-5.29, He = 7.39 t._
(8) From equation 6.25, the multiplying factor x is obtained,
0.991 + 7.39 x = 0, x = -0.135
(9) Factor x is applied to the sway moments in Fig. 6.15 c to obtain
the actual sway moments shown in Fig. 6.15 d.
(10) The final moments are obtained by adding the non-sway moments
in Fig. 6.l5b to the actual sway moments in Fig. 6.15d. These will be as
follows:
 468

Mab = + 3.13 + 1.01 = + 4.14 m.t.

M b = + 19.77 + 0.68 = + 20.45 m.t.
M b, = - 19.77 - 0.68 = - 20.45 m.t.
M'b = + 20.82 - 0.89 = + 19.93 m.t.
M,d = - 20.82 + 0.89 = - 19.93 m.t.
M d, = - 10.41 + 1.96 = - 8.45 m.t.

These moments are very nearly the same as those previously obta ir.cd
by the slope-deflection method.

Exatnple 6.11 Analyse the frame shown in Fig. 6.16a. The moment of
inertia varies as indicated. (This problem has been previously solved by
the slope-deflection method in Example 5.9-Fig. 5.15)

Solution: The analysis is presented in steps.

(l) The relative st:& lcsses and distribution factors arc evaluated.

K'b : K b, : K,d = -.!. : ~ : ~ = I :3:I

6 8 6
These stifFnesses are shown encircled in Fig. 6.16b.
I I 3 3
At joint b" Dba = -- - - and Dbc = - - = _.-
1+3 4 1+3 4
3 3
At joint c, DCb = - - = - and
I +3 4

Note that the overhang has zero stiffness. The distribution factors are
shown in the boxes attached to the line diagram of the frame in Fig. 6.16b.

(2) Assuming that the frame is prevented from sway by a horizontal

force RaJ the fixed-end moments due t? the applied loads are calculated.

10 X 3.2 X 4.8 2
M'be ----;:--- = - 11.52 m.t.
82
, l O x 4.8 X 3.22
M'b = + + 7.68m.t.

Moe = - 10 X 2 = - 20 m.t.

(3) These moments are distributed to obtain the non-sway moments

shown in Fig. 6.16 b.
 469

101
c
e

6m I

a d
"3.2---,--
I.- " ' 4.8m----l- 2 "J
(a)

I
I

I
l
- I. 62
1. 22
0.65
~ 0.49

~ ~<O
.
- , 96I~
o 00 -; '10 C?0
'" IN 1"'cf1 c:
0
<0
C>

(tl) Non-sway Mt5.

I
t

J' ~

2r~
r ,
I
I ~9-;~'?1
,
F
o

(c) Sway MIs.

Fig. 6.16
(4) The artificial restraining force H. is calculated by considering the
equilibrium of the frame as follows :
- 0.98 + 1.96 -- 049 t.~
Xa -
6
470

__ 1.13 + 2.28 _
_ 0. 568 t .-+
6
1: X = 0 = H. 2 0.49 + + +
0.568, H. = 3.058 t. +--
(5) The frame is forced through some displacement", and the cor-
responding fixed-end moments are calculated.

F F F F - 6 EI '" - II.
Mab = Mba = MCd = Md c = 62 ~ say

(the numerical value of 11 has been chosen only for convenience to simplify
the moment distribution in the following step.)
(6) These fixed-end moments are distributed to obtain the sway
moments shown in Fig. 6.16c. N?te that a modified stiffness for member
be has been used to allow for the anti symmetry of loading. The corres-
ponding distribution factors have been calculated and are as indicated in
the boxes attached to the line diagram of the frame in Fig. 6.16c.
(7) The consistent straining force He is calculated by considering the
equilibrium of the frame,

He = 2 9 +
6 10) = 6.333 t. -+
(
(8) The multiplying factor is determined from equation 6 .25,
3.058 - 6 .333 x = 0, x = + 0.483
(9) The actual sway moments are obtained by multiplying the sway
moments in Fig. 6.16c by the factor x and will be as follows:
Mab = Mde = - 10 (0.483) = - 4.83 m.t.
M b = Med = - 9 (0.483) = - 4.35 m.t.
Mbe = Meb = + 9 (0.483) = + 4.35 m.t.
(10) The final moments are obtained by adding the actual sway
moments in step 9 to the non-sway moments in Fig. 6.16b. The result
will be as follows :
Mao = + 0.98 - 4.83 = - 3.85 m.t,
Mba = + 1.96 - 4.35 = - 2.39 m.t.
Mbe =- + 4.35 = + 2.39 m.t.
1.96
Meb = + 17.72 + 4.35 = + 22.07 m.t.
Med = + 2.28 - 4.35 = - 2.07 m.t.
Mde = + 1.13 - 4.83 =- 3.70 m.t.
I
I It is seen that these momen ts are very nearly the same as those
usly obtaine d by the slope-d eflectio n method .
previo-
471

momen t of
ExaD1 ple 6.12 Analys e the frame shown in Fig. 6. I7a. The
inertia varies as indicat ed.

181
b 21 cr I j
6 d-.i

a
" 12 m
(a)

:;1~oINOI"'~ ci 0;
I + I ~

.4
" --2
'\.3.; 0.5
~
'6~~I:::ol_II.Igl~
0 I

(b) Sway MIs.

Fig. 6.17

Solutio n: The analysi s is presented in steps.

,~
ed.
(1) The relative stiHOesses and distrib ution factors are evaluat

I
K'b : K b, : K,. = +: +(+)
2121 : = 2 :2:3
allow for the
Note that a modifie d stiffness for membe r cd has been used to
hinged end at d.

Ir 2
At joint b, Dba = - - = 0.5 and Db'
2+2
2
_ _ =0.5
2+2

-- -- -- -- -_________
472

2 3
At joint c .. Deb = - - = 0.4 and D,. = - - =0.6
2+3 2+3
(2) Since no load is applied to the members inbenveen the joints) the
non-sway moments are zero.
(3) The frame is forced through some displacement 6 and the corres-
ponding fixed-end moments are calculated.

say- !O

say - 20

Note that in calculating MCd the dfuct of the hinged end d is taken into
account.

(4) These moments are distributed to give the sway moments shown
in Fig. 6.17b.

(5) The consistent straining force He is calculated by considering the

equilibrium of the frame as follows:
8.375 + 6.75
X. - = 2.521 t. . -
6
8.9
x. = - = 2.966 t . . -
3
LX = 0 = H,-2.521-2.966, H, = 5.487 t. -+
(6) The actual sway moments, which in this case are also the final
moments" are obtained by direct proportion. The constant of proportion-
ality = 18/5.487 = 3.28. The final moments are thus as follows:
Mab = - 8.375 X 3.28 = - 27.47 m.t.
Mb =- 6.75 X 3.28 =- 22.14 m.t.
M b , =,+ 6.75 X 3.28 = + 22.14 m.t.
M'b = . + 8 .9. X 3.28 = + 29.19 m.t.
M,. = - 8.9 X 3.28 = - 29.19 m.t.

ExaIDple 6.13 Analyse the frame shown in Fig. 6 .18a. The moment of
inertia varies as indicated. (This problem has been previously solved by
the slope-deflection method in Example 5.II-Fig. 5.17)
 473

31 b c
r 1. 21
CD CD
d
~
Cl"
i' ~C)
I- 3 ..-J.-- 6m -I-3~
I ~"
(a ) (b ) Sway M15,

I " 4

12 12

12

'. (c) 4
Fig , 6.18

Solution: The analysis is presented in steps.

(1) The relative stiff.l esses and distribution factors are evaluated.

K'b : Koc : KCd =

I
5 : 1.5 ( I). 5I
1.2
- 6- . = I : 1.5 : 1

Note that a modified stiffness has been used for member be to allow for
the antisymmetry of loading. As explained previously, this permits con-
sidering onc half orthe frame as there will be no carryover moments
between the two halves.

At joint b, Do, = I = 0.4 and Doc = 1.5 = 0.6

I + 1.5 1 + 1.5
(2) Since no load is applied to the memhers inbetween the joints the
non-sway moments arc zero.
(3) The frame is forced through some displacement ;;, and the corres-
ponding fix:ed-end moments are calculated.

M~b = M~. = _ 6 E (I) 5(~ .25 ;;,) , say - 12

F F 6 E (1. 21) (1.5 ;;,)

Moc = MCb = + 62 ' say + 12
Note that as joints band c arc forced to the right through a
t
474

a relative
displac ement 1::::. ) membe rs a b a nd cd fotate clockwise with
with
displacement = 1.25 6" a nd tha t membe r be rota tes anticlo ckwise
th e positiv e sign
a relative d isplace men t = 1.5 6. . This a ccount s for
ements
for the corresp onding fixed-e nd momen ts . The relative displac
ement diagram in
mentio ned a bove have been obtaine d from the displac
Fig .5 .17e (Examp le 5.11 ) .
shown
(4) These momen t s a re distribu ted to obtain th e sway momen ts
ge has been
in Fig. 6.18b. Note that in dis tributin g the momen ts a dvanta
taken of the a nti symme try of loading .
ce to
(5) The consist ent strainin g force He is calcula ted with referen
6.l8c,
the free body diagram s of individ ual membe rs shown in Fig.
He = 9 +9= 18 t. ~

ents in
(6) The actual sway momen ts, which are a lso the final m om
a factor
thi s case, are obtaine d by multip lying th e sway momen ts by
3/ 18. The result is a s follows :

M ab = Mba = M cd = M dC = - 2 m.t.
M be = M eb = + 2 m.t.
he slope-
These momen ts are id entical to t hose previou sly obtaine d b y t
deflecti on method .

'" 6.11 Applic ation to f r ames with multip le degree s of freedo

m in
transl ation
ed
~ The basic princ\o le<; present ed in section 6 .1 0 can be easi ly extend
in t ranslat ion.
to cove r frames w ith more than one degree of freedom
na l solutio n
For a frame having St degree of freedom in transla tion, th e fi
m ay be obtaine d b y superim posing t he results of (St + I ) sol u tions. One
joints is
solutio n corresp onds to the case w here transla tion of a ll the
and one additio nal
preven ted by a system of artifici al restrain ing fo rces,
indepe n-
solu tion corresp ond s to each degree of freedom wher e only one
rema ining indepe ndent
dent joint transla tion is allowed to occur while the
be applied
joint transla tions are preven ted. The multipl ying factors to
n of simulta neous
to the various sway cases are obtaine d from the solutio St

of freedom
equatio ns . EJ.ch equatio n is associa ted with a single degree
the artifica l restrain-
and expresses the conditi on that the algebra ic sum of
g from
ing force in the non-sw ay case and the respect ive forces resultin
ind ividual sway solutio ns must be zero .

.
-
 475

H.I
f p f

Hb
t. b C b c
+
a d 9 a d 9
" %
" % ~

(a) (b) Non-sway case

Ho2 H.3
e

I
f f

Hz Hb3
b C b c

+
II a
,.
d
%

(c) Sway casel

9
%
a
~
d
,
(d) Sway case 2
9
0

F ig. 6. 19

To illustrate this procedure, consider the frame shown in Fig.6.19a.

This frame has two degrees of freedom in translation corresponding to
the independ ent translation of joints band c.

The final solution comprise the three separate cases shown in Figs.
6.19b-d. In the first case, joint translations are completely prevented by

~ forces Hbl and H el . These forces can be found by statics once the associ-
ated non-sway moments are determined. In the second case, joint b is
!, forced through some arbitrarily chosen displacement while joint e is held
I against translation. The induced sway moments are found as explained
in section 6.10 and the consistent joint forces Hb2 and He2 are found by
r
I
statics. Similar solution is carried out for the third case and forces Hb3

,
l and He] are determined.
476
in Fig.
Assum ing the various forces to act in the directio ns indicat ed
each of joints
6.19, the conditi ons for zero artifici al restrain ing forces at
band e may be express ed mathem atically as

At joint b, HbI + X Hb2 - Y Hb3 = 0

At joint c., Hel - x He2 + Y He3 = 0
cases 1
where x and y arc the multip lying factors to be applied to sway
the above
and 2 respect ively and are found by the simulta neous solutio n of
two equatio ns . The final momen ts afC thus given by :

Final momen ts = Non-sw ay momen ts + x (sway momen ts

1)
+ Y (sway momen ts 2)

6.12 Applic ation to struct ures cOnlpo sed of non-pr isDlati c

memb ers.
ing
The basic princip les of momen t distribu tion present ed in the preced
ble to
section s are quite genera l in the sense that they arc equally applica
sma-
beams and frames whethe r they are compos ed of prisma tic or non-pri
ions ~or
tic membe rs. The only differen ce lies in the fact that the express
so for have
the fixed-e nd momen ts, stiffness and carryo ver factors used
compos ed
been derived specifically for prisma tic membe rs. For structu res
be evaluat ed first.
of non-pr ismatic membe rs, these quantit ies have to
r 4 on
A conven ient method for so doing has been introdu ced in chapte
column analog y .
t
477

t EXAMPLES TO BE WORKED OUT

I- (1)-(15) Using the moment distribution rr.etbod, zna.!yse the statiozlly
t
I indeterminate beams in (Figs. 2.81,3.27 - 3.39, and 6.20).

3t/m

21
8m - - + -
Fig. 6.20

(16)-26) Using the moment distribution method, ana lyse the statically
r indeterminate frames in Figs. 3.45-3.55.
(27)- (30) Using the moment distribution TTethod, ap.aly,e the static .!l),
indeterminate frames in Figs. 4.25, 4 . ~8 ar..d 4.30. ./
(31)-(34) Using the moment distribution methcd, analy,e the statically
indeterminate frames in Figs. 4.24, 4 .27, 4.32 and 4.34 ..
(35), (36) Using the moment distribution methcd, analyse the statically
indeterminate frames in Figs. 5.25 and 5.26.
(37)-(42) Using the moment distributicn. method, analyse the statically
indeterminate frames in Figs. 6.21-6.26.

2t1m

~~
21

4m

J--_ _ 12m _ _+-z 4 ~ ~':3 -I-- 3 -l-2-1-4m-1

Fig. 6.21 Fig. 6.22

4t1m 8t

r
6m

'" 3x3=9 _-+_ __

I--- 12 m --_+_- 3x3=9 ---l
Fig. 6.23
 ,1/
,,
I

478

1t 11
3t1m

r
6m
21 21 21

~2 I----- 8
~

6m 8
.1
----' 2 L.-
Fig. 6.24

I
t
titl m
,---- 21
5m 3t/m 3t/m
~ 21 21
5 1
C/,:
"'
~1O 10m
~"
10 --+-_'_"_10 ---I '"

Fig. 6.25

1tim
31
I4
2.5t/m
l.1
i4

_+--___
" 12 m _ _ _+--,,_. _ 6 ---l

Fig . 6 .26

(43)-(47) Using the moment distribution method, analyse the static-

ally indeterminate frames in Figs. 6.2i~6.31.
 479

101 4t1m
11
21 21 31 --,
6 21 4.5 21 1.51 3

6 6m-.J 12m
Fig. 6.27
Fig. 6.28

411m 201
31
1
1
I
5m

I- '" 15m I " 3-1-4 -I.... 3 __

J
Fig. 6 .29

311m

r
6m 1
21
1 4
3t/m

8m -1:1
Fig. 6.30
'''~3
4
1

-L4m--J- 4-1
Fig. 6.31

(48) Using the moment distribution method, calculate the ratio of

the sway moment to the corresponding final moment at the various
joints of the frame shown in Fig. 6 .32.
 15t1m
21 21
4
3 tim 3t1m

I
41 41
I
-t
4
3t1m 311m
41 41 44m
I
7h~
~em - I e -~

Fig. 6.32

in
(49),(5 0) If load is transm itted to membe r ac of each of the frames
apart, constru ct tbe
Figs. 6.33 and 6.34 at panel points spaced at 2m.
b.
influence line for the momen t at a section just to the left of joint

.
d r
151 4
"0 b c 0 b c
21 21 41
zJ
1.51 31 6

y
6m

Lem
d
<
e L8m~~j
Fig. 6.33 Fig. 6.34