11

Beam Deections:
4th Order Method
and Additional Topics

111
Lecture 11: BEAM DEFLECTIONS: 4TH ORDER METHOD AND ADDITIONAL TOPICS

TABLE OF CONTENTS

Page
11.1. Fourth Order Method Description 113
11.1.1. Example 1: Cantilever under Triangular Distributed Load . . . 113
11.2. Superposition 114
11.2.1. Example 2: Cantilever Under Two Load Cases . . . . . . 114
11.2.2. Example 3: A Statically Indeterminate Beam . . . . . . . 115
11.3. Continuity Conditions 116
11.3.1. Example 4: Simply Supported Beam Under Midspan Point Load 116

112
 11.1 FOURTH ORDER METHOD DESCRIPTION

11.1. Fourth Order Method Description

The fourth-order method to find beam deflections gets its name from the order of the ODE to be
integrated: E Izz v I V (x) = p(x) is a fourth order ODE. The procedure can be broken down into the
following steps.
1. Express the applied load p(x) as function of x, using positive-upward convention. This step
may involve changing load signs as necessary, as in the example below.
23. Integrate p(x) twice to get Vy (x) and Mz (x)
4. Pause. Determine integration constants from static BCs, and replace in Mz (x). (If the constants
are too complicated when expressed in terms of the data, they might be kept in symbolic form
until later.)
58. From here on, same as the second order method.
An example of this technique follows.

11.1.1. Example 1: Cantilever under Triangular Distributed Load

This example has been worked out in the previous lecture using the second order method. It is
defined in Figure 11.1, which reproduces the figure of the previous Chapter for convenience.

Constant EIzz w(x) = wB x /L

y
wB
A x B
L

Figure 11.1. Beam problem for Example 1. The applied load w(x) = w B x/L is considered positive if it goes
downward, that is, if w B > 0. This is converted to a negative load p B (x) = w B x/L to insert in the ODEs.

From inspection the applied load is

wB x
.
p(x) = (11.1)
L
Notice the minus sign to pass from the users convention: w(x) > 0 if directed downward, to the
generic load convention: p(x) > 0 if directed upward. Integrating p(x) twice yields

wB x 2
Vy (x) = + C1 ,
p(x) d x =
2L
 (11.2)
wB x 3
Mz (x) = Vy (x) d x = C1 x + C2 .
6L
Apply now the static BCs at the free end A: Vy A = Vy (0) = C1 = 0 and Mz A = Mz (0) = C2 = 0.
Hence
wB x 3
Mz (x) = 12 w(x) x ( 13 x) = (11.3)
6L
113
Lecture 11: BEAM DEFLECTIONS: 4TH ORDER METHOD AND ADDITIONAL TOPICS

Constant EIzz w(x) = wB x /L

y
wB

A x B
L
P (a) Original problem
w(x) = wB x /L
y y
wB
= A x B + A x B
L L
P
(b) Decomposition into two load cases and superposition

Figure 11.2. Beam problem for Example 2.

From here on the steps are the same as in the second order method worked out in Lecture 10. The
deflection curve is
wB
v(x) = (x 5 5L 4 x + 4L 5 ) (11.4)
120E Izz L
The maximum deflection occurs at the cantilever tip A, and is given by

wB L 4
v A = v(0) = (11.5)
30E Izz

The negative sign indicates that the beam deflects downward if w B > 0.

11.2. Superposition

All equations of the beam theory we are using are linear. This makes possible to treat complicated
load cases by superposition of the solutions of simpler ones. Simple beam configurations and load
cases may be compiled in textbooks and handbooks; for example Appendix D of Beer-Johnston-
DeWolf. The following example illustrates the procedure.

11.2.1. Example 2: Cantilever Under Two Load Cases

Consider the problem shown in Figure 11.2(a). The cantilever beam is subject to a tip point force
as well as a triangular distributed load. This combination can be decomposed into the two load
cases shown in Figure 11.2(b). Both of these have been separately solved previously as Examples
1 and 2 of Lecture 10 (the latter also as the example in the previous section). The deflection curves
for these cases will be distinguished as v P (x) and v w (x), respectively.
We had obtained
P wB
v P (x) = (L x)2 (2L + x), v w (x) = (x 5 5L 4 x + 4L 5 ) (11.6)
6E Izz 120E Izz L

114
 11.2 SUPERPOSITION

Constant EIzz w(x) = wB x /L

y
wB
MB
A x B A B
L RA RB
(a) Original (statically indeterminate) problem (b) Support reactions

w(x) = wB x /L
y y
wB
= A x B + A x B
RA L L

(c) Decomposition into two load cases and superposition

Figure 11.3. Beam problem for Example 3.

The deflection under the combined loading is obtained by adding the foregoing solutions:
P wB
v(x) = v P (x) + v w (x) = (L x)2 (2L + x) (x 5 5L 4 x + 4L 5 ).
6E Izz 120E Izz L
(11.7)
The tip deflection is

P L3 wB L 4 L3
v A = v(0) = = (10P + w B L) (11.8)
3E Izz 30E Izz 30E Izz
Superposition can be also used for any other quantity of interest, for example transverse shear
forces, bending moments and deflection curve slopes. An application to statically indeterminate
beam analysis is given next.
11.2.2. Example 3: A Statically Indeterminate Beam
The problem is defined in Figure 11.3(a). The beam is simply supported at A and clamped at B. If
the supports are removed 3 reactions are activated: R A , R B and 
M B , as pictured 
in Figure 11.3(b).
But thereare only two nontrivial static equilibrium equations: Fy = 0 and Many point = 0
because Fx = 0 is trivially satisfied. Consequently the beam is statically indeterminate because
the reactions cannot be determined by statics alone. One additional kinematic equation is required
to complete the analysis.
We select reaction R A as redundant force to be carried along as a fictitious applied load. Removing
the support at A and including R A makes the beam statically determinate. See Figure 11.3(b). This
beam may be viewed as being loaded by a combination of two load cases: (1) the actual triangular
load w(x), and (2) a point load R A at A. But this is exactly the problem solved in Example 2, if we
replace P by R A . The deflection curve of this beam is
RA wB
v(x) = (L x)2 (2L + x) (x 5 5L 4 x + 4L 5 ). (11.9)
6E Izz 120E Izz L

115
Lecture 11: BEAM DEFLECTIONS: 4TH ORDER METHOD AND ADDITIONAL TOPICS

Now the tip deflection must be zero because there is a simple support at A. Setting v A = v(0) = 0
provides the value of R A :

RA L3 wB L 4 wB L
v A = v(0) = =0 RA = (11.10)
3E Izz 30E Izz 10

This reaction value can be substituted to complete the solution. For example, the bending moment
is
wB x 3 wB L x wB x 3 wB x
Mz (x) = R A x = = (3L 2 5x 2 ) (11.11)
6L 10 6L 30L

The moment is zero at A (x = 0), becomes positive for 0 < x < L 3/5 .7746 L, crosses zero
at x = 0.7746 L and reaches Mz B = w B L 2 /15 at the fixed end. The deflection is

wB L wB
v(x) = (L x)2 (2L + x) (x 5 5L 4 x + 4L 5 ), (11.12)
60E Izz 120E Izz L

which may be simplified to

wB
v(x) = x (L 2 x 2 )2 (11.13)
120 E Izz L

11.3. Continuity Conditions

If the applied load is discontinuous, i.e., not a smooth function of x, it is necessary to divide the
beam into segments separated by the discontinuity points. The ODEs are integrated over each
segment. These solutions are patched by continuity conditions expressing that the slope v (x)
and the deflection v(x) are continuous between segments. This matching results in extra relations
between integration constants, which permits elimination of all integration constants except those
that can be determined by the standard BCs. The procedure is illustrated with the next example.

11.3.1. Example 4: Simply Supported Beam Under Midspan Point Load

The problem is defined in Figure 11.4(a). The calculation of the deflection curve will be done by the
second order method. Divide the beam into two segments: AC, which extends over 0 x 12 L,
and CB, which extends over 12 L x L. For brevity, these are identified as segments 1 and 2,
respectively, in the equations below.
The expression of the bending moment over each segment is easily obtained from statics. From
symmetry, the support reactions are obviously R A = R B = 12 P as shown in Figure 11.4(b). By
inspection one obtains that the bending moment Mz (x), diagrammed in Figure 11.4(c), is

Mz1 (x) = P x over segment 1 (AC),
Mz (x) = 2 (11.14)
M (x) = P(L x) over segment 2 (CB).
z2 2

116
 11.3 CONTINUITY CONDITIONS

y Constant EIzz
P P
A x C B A segment 1 C segment 2 B

L/2 L/2 RA = P RB = P
L 2 2
(b) Support reactions and division
(a) Problem definition
into two segments

P
A segment 1 C segment 2 B
RA = P RB = P
2 2

Mz1 (x)= Px Mz2 (x)= P(L-x)

2 2
+ +

(c) Bending moment diagram

Figure 11.4. Beam problem for Example 4.

Integrate Mz /E Izz over each segment:

2
E Izz v (x) = P x + C1 over segment 1 (AC),
E Izz v (x) =
1 4 (11.15)
E I v (x) = P x (2L x) + C over segment 2 (CB).
zz 2 4 1

It is convenient to stop here and get rid of C1 to avoid proliferation of integration constants. To
do that, note that the midspan slope vC must be the same from both expressions: vC = v1 ( 12 L) =
v2 ( 12 L). Else the beam would have a kink at C. This is called a continuity condition. Equating
P L 2 /16 + C1 = (3/16)P L 2 + C1 yields C1 = C1 P L 2 /8, which is replaced in the second
expression above:
2
E Izz v (x) = P x + C1 over segment 1 (AC),
1 4
E Izz v (x) = (11.16)

E Izz v2 (x) = P x (2L
4
x) P L 2 + C over segment 2 (CB).
8 1

Now C1 is gone. Integrate again both segments:

3
E Izz v1 (x) = P x + C1 x + C2 over segment 1 (AC),
E Izz v(x) = 12

E Izz v2 (x) = x (3L x) P L 2 x + C x + C
2
P over segment 2 (CB).
12 8 1 2
(11.17)
To get rid of C2 we say that the midspan deflection vC must be the same from both expressions:
vC = v1 ( 12 L) = v2 ( 12 L). This continuity condition gives C2 = C2 + P L 3 /48, which replaced

117
Lecture 11: BEAM DEFLECTIONS: 4TH ORDER METHOD AND ADDITIONAL TOPICS

yields
3
E Izz v1 (x) = P x + C1 x + C2 over segment 1 (AC),
E Izz v(x) = 12

E Izz v2 (x) = x (3Lx)
2
P P L2 x + P L3 + C x + C over segment 2 (CB).
12 8 48 1 2
(11.18)
We have now only two integration constants. To determine C1 and C2 use the kinematic BCs at A
and B. v A = v1 (0) = C2 = 0 and v B = v2 (L) = 0 C1 = P L 2 /16. Substitution gives, after
some simplifications,

E Izz v1 (x) = P x (4x 2 3L 2 ) over segment 1 (AC),
E Izz v(x) = 48
E Izz v2 (x) = P (4x 3 12L x 2 + 9L 2 x L 3 ) over segment 2 (CB).
48
(11.19)
The midspan deflection, obtainable from either segment, is

P L3
vC = v1 ( 12 L) = v2 ( 12 L) = (11.20)
48E Izz

As can be seen the procedure is elaborate and error prone, even for this very simple problem. It can
be streamlined by using Discontinuity Functions (DFs), which are covered in Lecture 12.

118