A First Course in the Finite Element Method, 5th Edition Daryl L.

Logan

© 2012 Cengage Learning Engineering. All Rights Reserved. 1

1
Many structures, such as buildings and
bridges, are composed of frames and/or grids.

Beam element (arbitrarily oriented in a plane)+

axial nodal displacement d.o.f.
analyze planes frame.

2
CAE A&D LAB
 3
CAE A&D LAB
 5.1 Two-Dimensional Arbitrarily Oriented Beam Element

related local displacement to global displacement

u '   C S  u 
 '   v 
v    S C  

4
CAE A&D LAB
  u1 
 v1'    S C 0 0 0 0   v1 
 ' 
  0 0 1 0 0 0  1 
 beam element  1'     
v2   0 0 0  S C 0  u2 
2'   0 0 0 
0 0 1   v2 
 
2 
S C 0 0 0 0
 0 0 1 0 0 0
T    0 0 0  S C 0
 
 0 0 0 0 0 1 

beam element transformation matrix
The above equation indicates rotation is invariant with respect to either coordinate system.

5
CAE A&D LAB
 k   T T k '  T  Eq.(3.4.22)

u1 v1 1 u2 v2 2
 12 S 2 12 SC 6 LS 12 S 2 12 SC 6 LS 
 
 12 SC 12C 2 6 LC 12 SC 12C 2 6 LC 
EI  6 LS 6 LC 4 L2 6 LS 6 LC 2 L2 
k  3  
L  12 S 2 12 SC 6 LS 12 S 2 12 SC 6 LS 
 12 SC 12C 2 6 LC 12 SC 12C 2 6 LC 
 
 6 LS 6 LC 2 L2 6 LS 6 LC 4 L2 

 the global element stiffness matrix, beam element, shear + bending resistance 

6
CAE A&D LAB
 include axial effects
three d.o.f. per node  u , v , 
'
i
'
i i
'

7
CAE A&D LAB
 recall, axial effects (bar)
 f1'x  AE  1 1  u1' 
 '     ' Eq.(3.1.13)
 f2 x  L  1 1  u2 
beam
 f1'y   12 6 L 12 6 L   v1' 
 '  2  '
 m1  EI  6 L 4 L 6 L 2 L  1 
2

 '  3  '
 f 2 y  L  12 6 L 12 6 L   v2 
 m2'   2 
 ' 
 6 L 2 L2
6 L 4 L 
 2

8
CAE A&D LAB
 Combing the axial effects with the shear and principal bending moment effects.
 f1'x   C1 0 0 C1 0 0  u1' 
 '    
 f1 y   0 12C2 6C2 L 0 12C2 6C2 L   v1' 
 m1'   0 6C2 L 4C2 L2 0 6C2 L 2C2 L2  1' 
 '   
 f 2 x   C1 0 0 C1 0 0  u2' 
 f 2' y   0 12C2 6C2 L 0 12C2 6C2 L   v2' 
 '   
 m2   0 6C2 L 2C2 L2 0 6C2 L 4C2 L2  2' 
AE EI
C1  and C2  3
L L
where
 C1 0 0 C1 0 0 
 0 12 C 6C L 0  12 C 6C L 
 2 2 2 2 
 0 6C2 L 4C2 L 2
0 6C2 L 2C2 L  2
 k '    
 C 1 0 0 C1 0 0 
 0 12C2 6C2 L 0 12C2 6C2 L 
 
 0 6C2 L 2C2 L2 0 6C2 L 4C2 L2 
(axial effect ( x ' ), shear force effects ( y ' ), bending mement effect ( z ' ))

9
CAE A&D LAB
 Local  global

 u1'   C S 0 0 0 0   u1 
 ' 
 v1    S C 0 0 0 0   v1 
1'   0 0 1 0 0 0  1 
 '   
u2   0 0 0 C S 0  u2 
 v2'   0 0 0 S C 0   v2 
 '   
2   0 0 0 0 0 1  2 
C S 0 0 0 0
S C 0 0 0 0 

0 0 1 0 0 0
T    0 0 0 C S 0
 (axial , shear , bending effects )

0 0 0 S C 0
 
0 0 0 0 0 1

10
CAE A&D LAB
 k   T 
 k '  T 
T

 general transformed global stiffness matrix for a beam element that includes axial force, shear force, and
bending moment effects.
 12 I 2  12 I  6I  12 I 2   12 I  6I 
 AC 2
 S  A  2 
CS  S   AC 2
 S    A  2 
CS  S
 L     L 
2 2
 L L L L 
  12 I  12 I 2 6I  12 I   2 12 I 2  6 I 
  A  2 
CS AS 2
 C C   A  2 
CS   AS  2 C  C 
      
2
L L L L L L 
 6I 6I 6I 6I 
  S C 4I S  C 2I 
E
k     L L L L 
L   12 I 2   12 I  6I 12 I 2  12 I  6I 

  AC 2
 S    A  2 
CS S AC 2
 S  A  2 
CS S 
   
L 2
L  L L2
 L  L

  12 I   2 12 I 2  6I  12 I  12 I 2 6I 
  A  2 
CS   AS  C   C  A  2 
CS AS 2
 C  C
      
2 2
L L L L L L 
 6I 6I 6I 6I 
  S C 2 I S  C 4 I 
 L L L L 
 Can analysis a rigid plane frame.
A rigid plane frame: a series of beam elements rigidly connected to each other.
That is, the original angles made between elements at their joints remain unchanged after the deformation
due to applied loads or applied displacements.

CAE A&D LAB

 5.2 Rigid Plane Frame Examples

12
CAE A&D LAB
 5.3 Inclined or Skewed Supports-Frame Element

The frame element with inclined support at node 3

13
CAE A&D LAB
  u3'   cos  sin  0  u3 
 '    
 
T applied to node 3   v3     sin  cos  0   v3 
 '   0 1   
 3  0 3 
equation of the plane frame
Ti  f   Ti  K Ti  d 
T

 F1x   u1  0 
F  v 0
 1y   1 
 M1   1  0 
   
F
 2x   u 2 
   
   
T
 2y 
F  Ti K Ti  2 v
M    
 '2   2' 
 F3 x   u3 
F '   v'  0 
 3y   3 

M 3   
3  3 
'

14
CAE A&D LAB
  I   0  0 
 
Where Ti    0    
I 0
 0 0 t3 

 cos  sin  0
and t3     sin  cos  0 
 0 0 1 

15
CAE A&D LAB
 5.4 Grid Equations

A grid is a structure on which loads are applied perpendicular to the

plane of the structure, as opposed to a plane frame, where loads are
applied in the plane of the structure.

d.o.f vi'  y
ix'  x
iz'  z
any effect of axial displacement
is ignored;that is ui'  0.

16
CAE A&D LAB
 ' ' '
forces f , m , m
iy ix iz

Grid element do not resist axial loading;

that is, f  0
'
ix

17
CAE A&D LAB
 Step 1
Fig.5-15 shows the sign conventions for nodal torque
and angle of twist and for element torque.

Step 2
We assume linear angle-of-twist variation along xˆ
 '  a1  a2 x '
x0 ' ,
1x
xL '
2x

  2' x  1' x
 '
 
'
 x  1' x
 L 
 
1 x 
 '

   N1 N 2   ' 
'

 2 x 
 
'
x x'
N1  1  N2 
L L

18
CAE A&D LAB
 Step 3.
Assuming that all radial lines, such as
OA remain straight during twisting or torsional deformation.
the arc length AB is given by
AB= max dx '  Rd '
 the maximum shear strain
Rd '
 max 
dx '
at any radial position r.
d ' r '  '  2' x  1' x  ' 
  r '  2 x  1' x       x   '
1x 
dx L   L  
  G
  shear stress
G  shear modulus of the material
  shear strain

19
CAE A&D LAB
 Step 4.
From elementary mechanics
J
mx' 
R
 J  polar moment of inertia for the circular cross section 
mx' 
GJ '
L
 2 x  1' x 
by the nodal torque sign convention
m1' x  mx'
GJ '
m1' x 
L
 1x  2' x 

Similarly, m2' x  mx'

m 
'
2x
L
 2 x 1x 
GJ '
   '

20
CAE A&D LAB
 the torsion bar stiffness matrix equation
 m1' x  GJ  1 1 1' x 
 '   1 1   ' 
 2x 
m L    2x 
the stiffness matrix for the torsion bar is
GJ  1 1
 k  
'
 1 1 
L  

for various cross sections, there is different J

for instance, for cross sections made up of
thin rectangular shapes such as channels,
angles or I shapes, we approximate J by
1
J   bi ti3
3
where bi is the length of any element of the cross section and
t i is the thickness of any element of the cross section.
21
CAE A&D LAB
Table 5-1 Torsional constants J
and shear center SC for
various cross sections

22
CAE A&D LAB
 torsional+shear+bending
 the local stiffness matrix equation for a grid element
 12 EI 6 EI 12 EI 6 EI 
 0 0
L 3
L2
L3
L2 
 
GJ  GJ
 f1'y   0 0 0   v' 
 1
 '   L L  ' 
m 6 EI 
2 EI 1 x 
 1x   4 EI
 '
  0
 1z   L   1z 

'
m L L2
 '   ' 
 12 EI 
6 EI  v2 
 f2 y  
2 
0
 m2' x   L3
L  2' x 
 '   GJ   ' 
 m2 z 
  0 2 z 
 
 L 
 
Symmetry 4 EI 
 L 
 the local stiffness matrix for a grid element
23
CAE A&D LAB
 torsional+shear+bending
 the local stiffness matrix for a grid element
 12 EI 6 EI 12 EI 6 EI 
 0 0
L3
L 2
L3
L2 
 
 GJ GJ
0 0 0 
 L L 
 4 EI 6 EI 2 EI 
 2
0 
  kG'   
L L L 
 12 EI 6 EI 
 3
0 2 
 L L 
 GJ 
 0 
L
 
Symmetry 4 EI 
 L 
 the local stiffness matrix for a grid element
24
CAE A&D LAB
the transformation matrix relating local to global d.o.f for a grid
1 0 0 0 0 0 
0 C S 0 0 0 
 
0  S C 0 0 0 
TG    
 0 0 0 1 0 0 
0 0 0 0 C S 
 
 0 0 0 0  S C 
x j  xi z j  zi
C  cos   S  sin  
L L
the global stiffness matrix
kG   TG T kG'  TG 

25
CAE A&D LAB
5.5 Beam Element Arbitrarily Oriented in Space
The stiffness matrix for the beam element arbitrarily oriented
in space, or three dimensions.

26
CAE A&D LAB
 Bending in x - z plane ' '

the stiffness matrix due to bending in the x ' - z ' plane

due to m 'y
 12 L 6 L2 12 L 6 L2 
EI  3 2 3 
y  4L 6L 2L 
 k y   4
'

L  12 L 6 L2 
 3 
Symmetry 4L 
I y  the moment of inertia of the cross section about the
principal axis y ' , the weak axis; that is, I y  I z .

27
CAE A&D LAB
 Bending in the x' - y' plane
the stiffness matrix due to bending in the x ' - y ' plane
due to mz'
 12 L 6 L2 12 L 6 L2 
 
EI 4 L3
6 L2 2 L3 
 k z'   4z 
L  12 L 6 L2 
 
Symmetry 4 L3 

 ' AE  1 1 
+axial stiffness matrix   k    1 1  
 L  

 ' GJ  1 1 
+torsional matrix   k    1 1  
 L  

28
CAE A&D LAB
 u1' v1' w1' 1' x 1' y 1' z u2' v2' w2' 2' x 2' y 2' z
 AE AE 
 L 0 0 0 0 0  0 0 0 0 0 
L
 
 0 12 EI z 6 EI z 12 EI z 6 EI z 
0 0 0 0  0 0 0
 L3 L2 L3 L2 
 12 EI y 6 EI y 12 EI y 6 EI y 
 0 0 0  0 0 0  0  0 
 
3 2 3
L L L L2
 GJ GJ 
 0 0 0 0 0 0 0 0  0 0 
 L L 
 6 EI y 4 EI y 6 EI y 2 EI y 
 0 0  0 0 0 0 0 0 
 L2 L L2 L 
 0 6 EI z
0 0 0
4 EI z
0 
6 EI z
0 0 0
2 EI z 
 L2 L L2 L 
  k '    
  AE 0 0 0 0 0
AE
0 0 0 0 0 
 L L 
 12 EI z 6 EI z 12 EI z 6 EI 
 0  0 0 0  0 0 0 0  2z
 L3 L2 L3 L 
 12 EI y 6 EI y 12 EI y 6 EI y 
 0 0 0 0 0 0 0 0 
 L3 L2 L3 L2 
 GJ GJ 
 0 0 0  0 0 0 0 0 0 0 
L L
 
 0 6 EI y 2 EI y 6 EI y 4 EI y
0  0 0 0 0 0 0 
 L2 L L2 L 
 
 0 6 EI z 2 EI z 6 EI z 4 EI z 
0 0 0 0  0 0 0
 L2 L L2 L 
29
CAE A&D LAB
 the transformation from local to global axis
 k   T T k '  T 
  33 
 
  33 
  
T 
 33 

  33 
C ' C yx' C zx' 
 xx 
   C xy' C yy ' C zy ' 
 
 C xz' C yz' C zz' 


30
CAE A&D LAB
 the direction cosines of the x ' axis
x '  cos  xx' i  cos  yx' j  cos  zx' k
x2  x1
cos  xx'  l
L
y2  y1
cos  yx'  m
L
z  z1
cos  zx'  2 n
L

i j k
1
z  x'  y '  0 0 1
D
l m n
m l
y'   i j D  (l 2  m 2 )1/ 2
D D
31
CAE A&D LAB
 i j k
1
z x y 
' ' '
l m n
D
m l 0
ln mn
z'   i j  Dk
D D
 
 l m n
 
m l
  33    0
 D D 
 ln mn 
  D
 D D 
This vector   from the local coordinate system into the global one

32
CAE A&D LAB
 In summary, we have

m l
cos  xy'   , cos  yy' 
D D
ln
cos  zy '  0, cos  xz'  
D
mn
cos  yz'  , cos  zz'  D
D

33
CAE A&D LAB
 x' axis coincides with the global z,

'  uncertain
y

34
CAE A&D LAB
 '
the positive x axis in the same direction
 0 0 1

as the global z      0 1 0 
 
 1 0 0
'
the positive x axis opposite the global z
0 0 1

    0 1 0 
 
1 0 0 
35
CAE A&D LAB
 5.6 Concept of Substructure Analysis

Structures are too large to be analyzed as a single system (the

stiffness matrix and equations for solution exceed the memory
capacity tothe computer)
Separate the whole structure into smaller units  substructures

36
CAE A&D LAB
 Substructure B include the beams at the top (a-a).
The beams at the bottom (b-b) are included in substructure A.

37
CAE A&D LAB
 The force/displacement equations for substructure B
are partitioned with the interface displacements and
forces separated from the interior as follows:

  B
  B

 F   ii   ie    i 
B K K    d B

  B
i
 B   5.6.1
 F    K ei   K ee     d e  
B B
e

i the interface nodal forces and displacements.

e  the interior nodal forces and displacements to be eliminated by
static condensation.

 i   ii  i   ie  e 
F B
  K B
 d B
  K B
 d B
5.6.2 
F    Kei di    Kee d e 
B
e  B
 B
 B
 B
5.6.3

38
CAE A&D LAB
d     e   ei  i 
1
B
e   K   FBB
ee   K B
 d B
 5.6.4 

F    K   K   e
B B B 1 B
i ie ee F


  K    K   K   K eiB  d iB 
B
ii
B
ie
B
ee
1
  5.6.5

  F
B
i   K   K 
B
ie
B 1
ee F 
B
e

  B 1
 5.6.6 
B
and K ii   Kii    Kie   K ee   K eiB 
B B
 

39
CAE A&D LAB
 F B
i    
 F
B
i  K
B
ii  i
d B
 5.6.7 
Similarly, for substructure A and C,

 F iA   K iiA  K  di 

  A A
  
 5.6.8
ie
 A   A 
 e   ei 
 Kee  de 
A A
F K

F i
A
 F   
A
i  K
A
ii  i
d A
 5.6.9 

F C
i  F    
C
i  K
C
ii  i
d C
 5.6.10 
40
CAE A&D LAB
 Compatibility,

 i top   i bottom  and

d A
 d B
 i top   i bottom 
d B
 d C
 5.6.11

The interface displacements at the common locations where

cuts were made must be the same.

Direct superposition eqs. (5.6.7), (5.6..9), (5.6.10), the final

equation are expressed in terms of the interface displacements at
the eight interface nodes only as

Fi   F i   K ii di   5.6.12 

Eq.(5.6.12) gives the displacements at the interface nodes.

41
CAE A&D LAB
Ch-5

Example
Example 5.1
As the first example of rigid plane frame analysis, solve the
solve the simple “bent” shown in Figure 5-4.
The frame is fixed at nodes 1 and 4 and subjected to a
positive horizontal force of 10,000lb applied at node2 and
to a positive moment of 5000lb-in. applied at node 3. The
global-coordinate axes and the element lengths are shown
in Figure 5-4. Let E = 30 106psi and A = 10in2 for all
element, and let I = 200in4 for element 1 and 3,and
I = 100in4 for element 2.
Using Eq. (5.1.11), we obtain the global stiffness
matrices for each element.

Figure 5-4 Plane frame for analysis, also showing local

axis for each element
Element 1
For element 1. the angle between the global x and the local
x̂ axes is 90  (counterclockwise) because x̂ is assumed
to be directed from node 1 to node 2. Therefore,
x 2  x1  60  (60)
C  cos 90  (1)
 0
L 120
y 2  y1 120  0
S  sin 90  (1)
 1
L 120
12 I 12(200)
Also,   0. 167 in 2
(5.2.1)
L2
(10  12 ) 2

6I 6(200)
  10 . 0 in 3

L (10  12) 2
E 30  10 6
  250,000lb / in 3
L 10  12
Then, Using Eqs. (5.2.1) to help in evaluating Eq. (5.1.11) for
element 1, we obtain the element local stiffness matrix as

d1x d1 y 1 d 2x d2y 2
 0.167 0  10  0.167 0  10
 0 10 0 0  10 0 
 
  10 0 800 10 0 400  lb
 250,000
(1)
k 
  0 .167 0 10 0. 167 0 10  in.
 0  10 0 0 10 0 
 
  10 0 400 10 0 800 

where all diagonal terms are positive.

Element 2
For element 2, the angle between x and x̂ is directed from node
2 to node 3. Therefore,
C 1 S 0
12 I 12(100)
Also,   0. 0835 in 2
L2 120 2
6 I 6(100)
  5.0 in3
L 120
E
 250,000 lb/in3
L
Using the quantities obtained in Eqs. (5.2.3) in evaluating Eq.
(5.1.11) for element 2, we obtain
d 2x d 2 y 2 d3x d 3 y 3
 0.167 0 10 0.167 0 10 
 0 10 0 0 10 0 
 
 10 0 800 10 0 400  lb
k (2)  250, 000  
  0.167 0  10 0.167 0  10  in.
 0 10 0 0 10 0 
 
 10 0 400 10 0 800 
Element 3
For element 3, the angle between x and x̂ is 270 (or  90 )
because x̂ is directed from node 3 to node 4. Therefore,
C0 S  1
Therefore, evaluating Eq. (5.1.11) for element 3, we obtain
d 3 x d 3 y 3 d 4 x d 4 y 4
 0.167 0 10  0.167 0 10 
 0 10 0 0  10 0 
 
 10 0 800  10 0 400  lb
 250,000 
(3)
k 
  0. 167 0  10 0.167 0  10 in.
 0  10 0 0 10 0 
 
 10 0 400  10 0 800 
Superposition of Eqs. (5.2.2), (5.2.4), (5.2.5) and application of the boundary
condition d1 x  d1 y  1  0 and d 4 x  d 4 y   4  0
At nodes 1 and 4 yield the reduced set of equations for a longhand solution as
10,000 10.167 0 10  10 0 0  d 2 x 
 0   0 10 . 0835 5 0  0 . 0835 5  d 
    2 y 
 0   10 5 1200 0 5 200    2 
   250,000  
 0    10 0 0 10 . 167 0 10 d 3x 
 0   0  0.0835  5 0 10.0835  5  d 3 y 
    
 5000   0 5 200 10 5 1200  3 
Solving Eq. (5.2.6) for the displacements and rotations, we have
d 2 x   0.211in 
d   0.00148in 
 2y   
  2   0.00153rad 
 = 
d
  
3 x 0.209in 
d 3 y   0.00148in 
   
 3   0.00149rad 
The element forces can now be obtained using fˆ  kˆT d
for each element, as was previously done in solving truss
and beam problems.
0 1 0 0 0 0  d1 x  0 
 1 0 0 0 0 0  d 0 
  1y 
0 0 1 0 0 0  1  0 
Td    
 0 0 0 0 1 0 d
  2x  0 . 211 
0 0 0  1 0 0 d 2 y  0.00148 
  
 0 0 0 0 0 1  2  0.00153

On multiplying the matrices in Eq. (5.2.8), we obtain
 0 
 0 
 
 0 
Td   
 0 .00148 
  0.211 
 
 0.00153
Then
 10 0 0  10 0 0  0 
 0 0. 167 10 0  0. 167 10  0 
  
 0 10 800 0  10 400   0 
fˆ  kˆT d  250,000  
  10 0 0 10 0 0  0.00148 
 0  0.167  10 0 0.167  10   0.211 
  
 0 10 400 0  10 800   0.00153

Simplifying Eq. (5.2.10), we obtain the local forces acting on element 1 as

 fˆ1x    3700lb 
ˆ   
 1y  
f 4900 lb 
 mˆ  376,000lb  in.
ˆ  
1

 2x  
f 3700 lb 
 fˆ    4990lb 
 2y   

 mˆ 2   223,000lb  in 
Element 2
fˆ  5010lb
2x
fˆ2 y  3700lb mˆ 2  223,000lb  in.
fˆ3 x  5010lb fˆ3 y  3700lb mˆ 3  221,000lb  in.

Element 3
fˆ  3700lb
3x
fˆ3 y  5010lb mˆ 3  226,000lb  in.
fˆ4 x  3700lb fˆ4 y  5010lb mˆ 4  375,000lb  in.
Figure 5-5 Free-body diagrams of (a) element 1,
(b) element 2, (c) element 3
Considering the free body of element 1, the equilibrium
equations are
 Fxˆ : -4990 + 4990 = 0
 F : -3700 + 3700 = 0
yˆ

 M : 376,000  223,000  4990(120in.)  0

2

Considering moment equilibrium at node 2, we see from

Eqs. (5.2.12a) and (5.2.12b) that on element 1,
mˆ 2  223,000lb  in. , and the opposite value,
-223,000lb-in., occurs on element 2. Similarly, moment
equilibrium is satisfied at node 3,as m̂3 from element
ments 2 and 3 add to the 5000 lb-in. applied moment. That
is, from Eqs. (5.2.12a) and (5.2.12b) we have

-221,000 + 226,000 = 5000 lb-in

Example 5.2
To illustrate the procedure for solving frames subjected to
distributed loads, solve the rigid plane frame shown in Figure 5-6.
The frame is fixed at nodes 1 and 3 and subjected to a uniformly
distributed load of 1000lb/ft applied downward over element 2.
The global-coordinate axes have been established at node 1. The
element lengths are shown in the figure. Let
E  30  10 6 psi, A  100 in2, and I  1000 in4 for both element of
the frame.
We begin by replacing the distributed load acting on element 2
by nodal forces and moments acting at nodes 2 and 3. Using Eqs.
(4.4.5)-(4.4.7)(or Appendix D), the equivalent nodal forces and
moments are calculated as
wL (1000)40
f2y     20,000lb  20kip
2 2
wL2 (1000)40 2
m2     133,333lb  ft  1600k  in
12 12
 © 2002 Brooks/Cole Publishing / Thomson Learning™

Figure 5-6 (a)Plane frame for analysis and (b)equivalent

nodal forces on frame
 wL (1000)40
f3y     20,000lb  20kip
2 2
wL2 (1000)40 2
m3    133,333lb  ft  1600k  in.
12 12
We then use Eq. (5.1.11), to determine each element stiffness
matrix:

Element 1
 (1)  45 C  0.707 S  0.707 L(1)  42.4 ft  509.0in.
E 30  10 3
  58.93
L 509
50.02 49.98 8.33 
k (1)  58.9349.98 50.02  8.33
lb
in.
 8.33  8.33 4000 
Simplifying Eq. (5.2.14), we obtain
d2x d2y 2
2948 2945 491 
 2945  491 
lb
k (1) 2948
in.
 491  491 235,700
Element 2
 ( 2)  0 C  1 S 0 L( 2)  40 ft  480in
E 30  10 3
  62.50
L 480
100 0 0 
k ( 2)  62.50  0 0.052 12.5 
lb
in.
 0 12.5 4000
Simplifying Eq. (5.2.16), we obtain
d2x d2y 2
6250 0 0 
  0 781.25 
( 2)
k 3.25
 0 781.25 250,000
 F2 x  0  9198 2945 491  d 2 x 
    d 
 2y
F   20 
  2945 2951 290  2 y 
   491 290 485,700   
 2
M   1600    2 

Solving Eq. (5.2.18) for the displacements and the rotation

at node 2, we obtain

d 2 x   0.0033in 
   
 2y  
d   0.0097 in. 
    0.0033rad 
 2  
Recall that the local forces are given by fˆ  kˆT d . For
element 1, we then have
 0.707 0.707 0 0 0 0  0 
 0.707 0.707 0 0 0 0  0 
  
 0 0 1 0 0 0   0 
Td   .  
 0 0 0 0 . 707 0. 707 0  0. 0033 
 0 0 0  0.707 0.707 0   0.0097 
  
 0 0 0 0 0 1 
   0 .0033 
Simplifying Eq. (5.2.20) yields
 0 
 0 
 
 0 
Td   
  0. 00452 
  0.0092 
 
  0.0033 
Using Eq. (5.2.21) and Eq. (5.1.8) for k̂ , we obtain
 fˆ1x   5893 0 0  5893 0 0  0 
ˆ  
 f1 y   2.730 694.8 0  2.730 694.8   0 

 mˆ   117,900 0  694.8 117,900   0 
ˆ 
1
 
f
 2x   5893 0 0   0.00452 
 fˆ2 y   2.730  694.8    0.0092 
    
 mˆ   Symmetry 235,800   0.0033 

Simplifying Eq. (5.2.22) yields the local forces in element 1 as

fˆ1x  26.64kip fˆ1 y  2.268kip mˆ 1x  389.1k  in
fˆ  26.64kip
2x fˆ  2.268kip mˆ  778.2k  in.
2y 2x
For element 2, the local forces are given by Eq. (4.4.11)
because a distributed load is acting on the element. From
Eqs. (5.1.10) and (5.2.19), we thwn have
1 0 0 0 0 0  0.0033 
0 1 0 0 0 0  0.0097
  
0 0 1 0 0 0   0.0033
Td    
 0 0 0 1 0 0  0 
0 0 0 0 1 0  0 
  
0 0 0 0 0 1   0  
Simplifying Eq. (5.2.24), we obtain
 0.0033 
 0.0097 
 
 0.0033
 
 0 
 0 
 
 0 
Using Eq. (5.2.25) and Eq. (5.1.8) for k̂ , we have
 6250 0 0  6250 0 0   0.0033 
 3. 25 781 . 1 0  3. 25 781 .1   0.0097 
  
 250,000 0  781.1 125,000   0.0033
kˆ dˆ  kˆT d    
 6250 0 0  0 
 3.25  781.1   0 
  
 Symmetry 250,000  0 

Simplifying Eq. (5.2.26) yields

 20.63 
  2.58 
 
 832.57 
kˆ dˆ   
  20. 63 
 2.58 
 
 412.50
 fˆ2 x   20.63   0 
ˆ      20 
 2y  
f  2.58   
 mˆ   832.57   1600
ˆ   
2

 3x  
f  20 .63   0 
 fˆ   2.58    20 
 3y     
 mˆ 3   412.50  1600 
Simplifying Eq. (5.2.28), we obtain
fˆ  20.63kip
2x fˆ  17.42kip
2y mˆ 2  767.4k  in.
fˆ3 x  20.63kip fˆ3 y  22.58kip mˆ 3  2013k  in
Example 5.3
Solve the frame shown in Figure 5-8(a). The frame consists
of the three elements shown and is subjected to a 15-kip
horizontal load applied at midlength of element 1. Node 1,
2, and 3 are fixed, and the dimensions are shown in the
figure. Let E  30  10 6 psi, I  800 in4,and A  8 in2 for
all elements.
1. We first ex[ress the applied load in the element 1 local
coordinate system (here x̂ is directed from node 1 to
node 4.) This is shown in Figure 5-8(b).
2. Next, we determine the equivalent joint forces at each
end of element 1, using the table in Appendix D. (These
force are of opposite sign from what are traditionally
known as fixed-end forces in classical structural
analysis theory [1].) These equivalent forces (and
moments) are shown in Figure 5-8(c).
3. We then transform the equivalent joint forces from the
present local-coordinate-system forces into the
global-coordinate-system forces, using the equation
f  T fˆ , where T is defined by Eq. (5.1.10) These
T

global joint forces are shown in Figure 5-8(d).

4. Then we analyze the structure in figure 5-8(d), using the
equivalent joint forces (plus actual joint forces, if any) in
the usual manner.
5. We obtain the final internal forces developed at the ends
of each element that has an applied load (here element 1
only) by subtracting step 2 joint forces from step 4 joint
forces; that is, Eq (4.4.11) is applied locally to all
elements that originally had loads acting on them.

The solution of the structure as shown in figure 5-8(d) now

follows. Using Eq.(5.1.11), we obtain the global stiffness
matrix for each element.
 © 2002 Brooks/Cole Publishing / Thomson Learning™

Figure 5-8 Rigid frame with a load applied on an element

Element 1
For element 1, the angle between the global x axes is
63.43 because x̂ is assumed to be directed from node 1
to node 4. Therefore,

x 4  x1 20  0
C  cos 63.43  (1)
  0.447
L 44.7

y 4  y1 40  0
S  sin 63.43  (1)
  0.895
L 44.7

12 I 12(800) 6I 6(800)
  0.0334   8.95
L2
(44.7  12) 2
L 44.7  12

E 30  10 3
  55.9
L 44.7  12
Using the preceding results in Eq. (5.1.11) for k , we
obtain
d4x d4 y 4
90.9 178 448 
  178 359  224 
(1)
k
 448  224 179,000

Where only the parts of the stiffness matrix associated with

degrees of freedom at node 4 are included node 1 is fixed
and, hence, not needed in the solution for the nodal
displacements.
Element 3
For element 3, the angle between x and x̂ is zero
because x̂ is directed from node 4 to node 3. Therefore,
12 I 12(800)
C 1 S 0   0.0267
L2
(50  12) 2

6 I 6(800) E 30  10 3
  8.00   50
L 50  12 L 50  12
Substituting these results into k , we obtain
d4x d4y 4
400 0 0 
k   0 1.334 400 
(3)

 0 400 160,000
Since node 3 is fixed.
Element 2
For element 2, the angle between x and x̂ is 116.57
because from node 2 to node 4. Therefore,
20  40 40  0
C  0.447 S  0.895
44.7 44.7
12 I 6I E
2
 0.0334  8.95  55.9
L L L

Since element 2 has the same properties as element 1.

Substituting these results into k , we obtain
d4x d4 y 4
 90.9  178 448 
k ( 2)    178 359 224 
 448 224 179,000
since node 2 is fixed. On superimposing the stiffness
matrices given by Eqs. (5.2.30), (5.2.31), and (5.2.32), and
using the nodal forces given in Figure 5-8(d) at node 4 only,
we have
  750kip  582 0 896  d 4 x 
    d 
 0 
  0 719 400  4 y 
 900k  in  896 400 518,000   
    4 

Simultaneously solving the three equations in Eq. (5.2.33)

we obtain
d 4 x  0.0103in
d 4 y  0.000956in
4  0.00172rad
Next, we determine the element forces by again using
fˆ  kˆT d . In general, we have
 C S 0 0 0 0  d ix 
 S C 0 0 0 0  d 
   iy 
 0 0 1 0 0 0   i 
Td    
 0 0 0 C S 0  d jx 
 0 0 0  S C 0 d jy 
  
 0 0 0 0 0 1   j 
Thus, the preceding matrix multiplication yields
 Cd ix  Sd iy 
  Sd  Cd 
 ix iy 

 i 
Td   
 Cd jx  Sd jy 
 Sd jx  Cd jy 
 
 j 
Element 1
 0   0 
 0   0 
   
 0   0 
Td    
 ( 0.447 )( 0. 0103)  ( 0. 895)( 0.000956)    0.00374 
(0.895)(0.0103)  (0.447)(0.000956)  0.00963 
   
  0.00172   0.00172

Using Eq. (5.1.8) for k̂ and Eq. (5.2.36), we obtain

 447 0 0  447 0 0   0 
 0 1.868 500.5 0  1.868 500.5   0 
 
 0 500.5 179,000 0  500.5 89,490   0 
kˆT d   
  
  447 0 0 447 0 0   0.00374
 0  1.868  500.5 0 1.868  500.5   0.00963 
   
 0 500.5 89,490 0  500.5 179,000  0.00172
 © 2002 Brooks/Cole Publishing / Thomson Learning™

Figure 5-9 Free-body diagrams of all elements of the

frame in Figure 5-8(a)
element 1 as
 1.67   3.36  5.03kip 
 0.88  6.17    7.59kip 
     
ˆf (1)    158    900    1058k  in 
 1.67   3.36  1.68kip 
 0.88   6.71    5.83kip 
     
  311    900   589k  in 

Element 3
fˆ4 x  4.12kip fˆ4 y  0.687 kip mˆ 4  275k  in
fˆ3 x  4.12kip fˆ3 y  0.687 kip mˆ 3  137k  in

Element 2
fˆ2 x  2.44kip fˆ2 y  0.877kip mˆ 2  158k  in
fˆ4 x  2.44kip fˆ4 y  0.877 kip mˆ 4  312k  in
 M 4  589  275  312  2k  in. (close to zero)
 Fx  1.68(0.447)  5.83(0.895)  2.44(0.447)
 0.877(0.895)  4.12  0.0027kip (close to zero)
 Fy  1.68(0.895)  5.83(0.447)  2.44(0.895)
 0.877(0.447)  0.687  0.004kip

Thus, the solution has been verified to be correct within the

accuracy associated with a longhand solution.
Example 5.4
The bar element 2 is used to stiffen the cantilever beam
element 1,as show in Figure 5-10. Determine the
displacement at node 1 and the element forces. For the bar,
let A  10  10 3 m2. For the beam, let
A  2  10 3 m2, I  5  10 5 m4,and L  3 m. For both the
bar and the beam element, let E  210 GPa. Let the angle
between the beam and the bar be 45 . A downward force
of 500kN is applied at node 1.
For brevity’s sake, since nodes 2 and 3 are fixed, we keep
only the parts of k for each element that are needed to
obtain the global K matrix necessary for solution of the
nodal degrees of freedom. Using Eq. (3.4.23), we obtain k
for the bar as
(1  10 3 )(210  10 6 ) 0.5 0.5
k 
( 2)

(3 / cos 45) 0.5 0.5

 
or, simplifying this equation, we obtain
 d1 x d1 y
0.354 0.354 kN
 70  10 
( 2) 3
k 
0.354 0.354 m

© 2002 Brooks/Cole Publishing / Thomson Learning™

Figure 5-10 Cantilever beam with a bar element support

Using Eq. (5.1.11), we obtain
d1 x d1 y 1
2 0 0 
3  kN
k  70  10 0 0.067 0.10
(1)

m
0 0.10 0.20
Where E / L)  10 3 has been factored out in evaluating Eq.
(5.2.42).
We assemble Eqs. (5.2.41) and (5.2.42) in the usual manner
to obtain the global stiffness matrix as
2.354 0.354 0 
K  70  10 3 0.354 0.421 0.10
kN
m
 0 0.10 0.20 
The global equations are then written for node 1 as
 F1 x   0  2.354 0.354 0   d1 x 
    3  d 
 1y  
F   500   70  10  0. 354 0.421 0. 10  1y 
M   0   0 0.10 0.20  1 
 1  
Solving Eq. (5.2.44), we obtain
d1x  0.00338m d1 y  0.0225m 1  0.0113rad

In general, the local element force are obtained

using fˆ  kˆT d . For the bar element, we then have
 d 1x 
 fˆ1x  AE  1  1 C S 0 0   d1 y 
ˆ   
 f 3 x  L   1 1   0 0 C S  d 3 x 
d 3 y 
The matrix triple product of Eq. (5.2.46) yields (as one
equation)
AE
fˆ1x  (Cd 1x  Sd1 y )
L
Substituting the numerical values into Eq. (5.2.47), we obtain
ˆf  (1  10 3
m 2
)( 210  10 6
kN / m 2
) 2 
1x  (0.00338  0.0225)
4.24m  2 
Simplifying Eq. (5.2.48), we obtain the axial force in the bar
(element 2) as fˆ1x  670kN
where the negative sign means fˆ is in the direction opposite 1x

x̂ for element 2. Similarly, we obtain fˆ3 x  670 kN

© 2002 Brooks/Cole Publishing / Thomson Learning™

Figure 5-11 Free-body diagrams of the bar (element 2) and

beam (element 1) elements of figure 5-10
 fˆ  f and dˆ  d .
 fˆ1 x  C1 0 0   d 1x 
ˆ    d 
 1y   
 1y 
f 0 12 C 2 6 C 2 L
 mˆ   0 6C L 4C L2    
 1  2 2  1 
 fˆ1 x  2 0 0   0.00338 
ˆ  3   0.0225
 1y 
f  70  10  0 0 . 067 0 . 10  
 mˆ     0.0113 
 1  0 0 .10 0 . 20  
The matrix product then yields
fˆ1 x  473kN fˆ1 y  26.5kN mˆ 1  0.0kN  m
Similarly, using Eq. (5.1.6), we have at node 2,
 fˆ2 x   2 0 0   0.00338 
ˆ  3   0.0225
 2y 
f  70  10  0  0 . 067  0 . 10  
 mˆ   0 0.10 0.10   0.0113 
  2

The matrix product then yields

fˆ2 x  473kN fˆ2 y  26.5kN mˆ 2  78.3kN  m
Example 5.5
Analyze the grid shown in figure 5-18. The grid consists of
three element, is fixed at nodes 2, 3, and 4, and is subjected
to a downward vertical force (perpendicular to the x  z
plane passing through the grid element) of 100 kip. The
global-coordinate axes have been established at node 3, and
the element lengths are shown in the figure. Let
E  30  10 3 ksi, G  12  10 3 ksi, I  400 in4, and J  110
in4 for all element of the grid.© 2002 Brooks/Cole Publishing / Thomson Learning™

Figure 5-18 Grid for analysis showing local x̂ axis for

each element
d 2 y   2 x  2 z  0 d 3 y  3 x   3 z  0 d 4 y  4 x   4 z  0

Element 1
x 2  x1  20  0
C  cos   (1)
  0.894
L 22.36
z z 10  0
S  sin   2 (1) 1   0.447
L 22.36
12 EI 12(30  10 3 )(400)
  7.45
L3 (22.36  12) 3

6 EI 6(30  10 3 )(400)
  1000
L3 (22.36  12) 3

GJ (12  10 3 )(110)
  4920
L (22.36  12) 3

4 EI 4(30  10 3 )(400)
  179,000
L (22.36  12) 3
Considering the boundary condition Eqs. (5.4.20), using the results of Eqs.
(5.4.21) in Eq. (5.4.17) for k̂ G and Eq. (5.4.18) for T G , and then applying
Eq. (5.4.19), we obtain the upper left-hand 3 3 partitioned part of the globa;
stiffness matrix for element 1 as

1 0 0   7.45 0 1000  1 0 0 
k  0 0.894 0.447   0 0  0 0.894 0.447 
(1)
4920
0 0.447 0.894  1000 0 179, 000  0 0.447 0.894 
Performing the matrix multiplications, we obtain the global element grid
stiffness matrix
d1 y 1 2
 7.45  447  894 
  447 39,700 69,600 
kip
k (1)
in
  894 69,600 144,000
Element 2
For element 2, we assume the local x̂ axis to be directed
from node 1 to node 3 for the formulation of the element
stiffness matrix. We need the following expressions to
evaluate the element stiffness matrix:
x 3  x1  20  0
C ( 2)
  0.894
L 22.36
z z  10  0
S  3 ( 2) 1   0.447
L 22.36
Other expressions used in Eq.(5.4.17) are identical to those
Eqs. (5.4.21) for element 1 because E,G,I,J, and L are
identical. Evaluating Eq. (5.4.19) for the global stiffness
matrix for element for element 2, we obtain
d1 y 1x 1z
 7.45 447  894 
k ( 2) 
  447 39,700  69,600  kip
in.
 894  69,600 144,000 
Element 3
For element 3, we assume the local x̂ axis to be directed
from node 1 to node 4. We need the following expressions
to evaluate the element stiffness matrix:
x x 20  20
C  4 ( 3) 1  0
L 10
z 4  z1 0  10
S ( 3)
 1
L 10
12 EI 12(30  10 3 )(400)
  83.3
L 3
(10  12) 3

6 EI 6(30  10 3 )(400)
  5000
L 2
(10  12) 2

GJ (12  10 3 )(110)
  11,000
L (10  12)
4 EI 4(30  10 3 )(400)
  400,000
L (10  12)
 d1 y 1x 1z
 83.3 5000 0 
k (3) 
 5000 400,000 0   kip
in.
 0 0 11,000

d1 y 1x 1z
 98.2 5000  1790  d 1 y 
 
K G   5000 479,000 0  1x 
  1790 0 299,000 1z 
The force F1 y is negative because the load is applied in
the negative y direction. Solving for the displacement
and the rotations in Eq. (5.4.28), we obtain
d1 y  2.83in.
1x  0.0295rad
1z  0.0169rad
Element 1
Using Eqs. (5.4.17) and (5.4.18) for k G and T G and Eq.
(5.4.29), we obtain
1 0 0 0 0 0    2.83 
0  0.894 0.447 0 0 0   0.0295 
  
0  0.447  0.894 0 0 0   0.0169
TGd    
 0 0 0 1 0 0  0 
0 0 0 0  0.894 0.447   0 
  
0 0 0 
0  0.447  0.894  0 
Multiplying the matrices, we obtain
  2.83 
 0.0339
 
 0.00192 
TGd   
 0 
 0 
 
 0 
Then fˆ  kˆ G T G d becomes
 fˆ1 y   7.45 0 1000  7.45 0 1000    2.83 
  
mˆ
 1x   0 4920 0 0  4920 0   0.0339
 mˆ   1000 0 179,000  1000 0 89,500   0.00192 
ˆ  
1y
  
 2y  
f  7 .45 0  1000 7 .45 0  1000  0 
mˆ   0  4920 0 0 4920 0  0 
 2x
   
mˆ
 2 y    1000 0 89 ,500  1000 0 179 ,000 
 0 
Multiplying the matrices in Eq. (5.4.31), we obtain the local element forces
as
 fˆ1 y    19.2kip 
   
 mˆ 1 x    167 k  in 
 mˆ   2480k  in 
 ˆ 
1y

 2y  
f 19 . 2 kip 
mˆ 2 x   167 k  in 
   
mˆ 2 y    2660 k  in 
© 2002 Brooks/Cole Publishing / Thomson Learning™

Figure 5-19 Free-body diagrams of the elements of

Figure 5-18 showing local-coordinate systems for each
Element 2
Similarly, using fˆ  kˆ G T G d for element 2, with the direction
in Eqs. (5.4.23), we obtain
 fˆ1 y   7.45 0 1000  7.45 0 1000 
   
mˆ
 1x   0 4920 0 0  4920 0 
mˆ   1000 0 179,000  1000 0 89,500 
 ˆ 
1y

 f3 y    7 .45 0  1000 7.45 0  1000 
mˆ 3 x   0  4920 0 0 4920 0 
   
mˆ 3 y   1000 0 89,500  1000 0 179,000
Multiplying the matrices in Eq. (5.4.33), we obtain the
local element forces as

fˆ1 y  7.23kip
mˆ 1x   92.5k  in
mˆ 1z  2240k  in
f 3 y   7.23kip
mˆ 3 x  92.5k  in
mˆ 3 z   295k  in
Element 3
Finally, using the direction cosines in Eqs. (5.4.25), we obtain the
local element forces as
 fˆ1 y   83.3 0 5000  83.3 0 5000 
   
ˆ
 1x  
m 0 11 , 000 0 0  11, 000 0 
mˆ   5000 0 400,000  5000 0 200,000
 ˆ 
1y

 3y  
f  83 . 3 0  5000 83 . 33 0  5000 
mˆ 3 x   0  11,000 0 0 11,000 0 
   
m3 y  
ˆ 5000 0 200 ,000  5000 0 400 ,000 
1 0 0 0 0 0    2.83 
0 0  1 0 0 0   0.0295 
  
0 1 0 0 0 0   0.0169
  
 0 0 0 1 0 0  0 
0 0 0 0 0  1  0 
  
0 0 0 0 1 0   0  
Multiplying the matrices in Eq. (5.4.35), we obtain the
local element forces as

fˆ1 y   88.1kip
mˆ 1x  186k  in
mˆ 1z   2340k  in
f 3 y  88.1kip
mˆ 3 x   186k  in
mˆ 3 z   8240k  in
 © 2002 Brooks/Cole Publishing / Thomson Learning™

Figure 5-20 Free-body diagram of node 1 of Figure 5-18

Element 1

 f1 y  1 0 0    19.2 
      167 
 1x  
m  0  0. 894  0.447  
m  0 0.447  0.894  2480
 1z    
Simplifying, we obtain the global-coordinate force and
moments as
f 1 y  19.2kip m1x  1260k  in m1z  2150k  in
Where f 1 y  fˆ1 y because y  yˆ .
Element 2

 f1 y  1 0 0   7.23 
     92.5
 1x  
m  0  0.894 0. 447  
m  0  0.447  0.894  2240 
 1z    
Simplifying, we obtain the global-coordinate force and
moments as
f1 y  7.23kip m1x  1080k  in m1z  1960k  in
Element 3
 f1 y  1 0 0   88.1 
     186 
 1x  
m  0 0 1  
m  0  1 0  2340
 1z    
Simplifying, we obtain the global-coordinate force and
moments as
f1 y  88.1kip m1x  2340k  in m1z  186k  in
Then forces and moments from each element that are equal
in magnitude but opposite in sign will be applied to node 1.
Hence, the free-body diagram of node 1 is shown in Figure
5-20. Force and moment equilibrium are verified as
follows:
 F1 y  100  7.23  19.2  88.1  0.07 kip (close to zero)
 M 1x  1260  1080  2340  0.0k  in
 M 1z  2150  1960  186  4.00k  in (close to zero)
Thus , we have verified the solution to be correct within the
accuracy associated with a longhand solution.
Example 5.6
Analyze the grid shown in Figure 5-21.The grid consists of
two elements, is fixed at nodes 1 and 3, and is subjected to
a downward vertical load of 22 kN. The global-coordinate
axes and element lengths are shown in the figure. Let
E=210GPa, G=84GPa, I=16.6×10-5m4, and J=4.6×10-5m4.
As in Example .5, we use the boundary conditions and
express only the part of the stiffness matrix associated with
the degrees of freedom at node 2. The boundary conditions
at nodes 1 and 3 are
d1 y  1x  1z  0 d 3 y   3 x  3 z  0

Figure 5-21 Grid example

Element 1
For element 1, we have the local x̂ axis coincident with
the global x axis. Therefore, we obtain
x 2  x1 3 z 2  z1 3  3
C (1)
 1 S (1)
 0
L 3 L 3

Other expressions needs to evaluate the stiffness matrix are

12 EI 12(210  10 6 kN m 2 )(16.6  10 5 m 4 )
  1. 55  10 4

L3 (3m) 3
6 EI 6(210  10 6 )(16.6  10 5 )
  2 .32  10 4

L2 (3) 2
GJ (84  10 6 )(4.6  10 5 )
  1.28  10 3
L 3
4 EI 4(210  10 6 )(16.6  10 5 )
  4.65  10 4
L 3
Considering the boundary condition Eqs. (5.4.40), using the
results of Eqs. (5.4.41) in Eq. (5.4.17) for k̂ G and Eq.
(5.4.18) for T G , and then applying Eq. (5.4.19), we obtain
the reduced part of the global stiffness matrix associated
only with the degrees of freedom at node 2 as
1 0 0  1.55 0 - 2.32 1 0 0
k (1)  0 1 0   0 0.128 0  (10 4 ) 0 1 0
  
0 0 1  2.32 0 4.65  0 0 1

Since the local axes associated with element 1 are parallel

to the global axes, we observe that T G is merely the
matrix; therefore, k  kˆ . Performing the matrix
G G
multiplications, we obtain
 1.55 0 - 2.32 
k (1)   0 0.128 0  (10 4 ) kN
  m
 2.32 0 4.65 
Element 2
For element 2, we assume the local x̂ axis to be directed from
node 2 to node 3 for the formulation of k . Therefore,
x x 00 z z 03
C  3 ( 2) 2  0 S  3 ( 2) 2   1
L 3 L 3

 1 0 0  1.55 0 2.32   1 0 0
  0 0 1  0 0  (10 4 )  0 0 1
( 2)
k 0.128
 0 - 1 0  2.32 0 4.65   0 - 1 0

 1.55 2.32 0 
  2.32  4 kN
( 2)
k 4.65 0  (10 ) m
 0 0 0.128
Superimposing the global stiffness matrices from Eqs.
(5.4.42) and (5.4.44), we obtain the total gloal stiffness
matrix (with boundary conditions applied) as
 3.10 2.32 - 2.32 

K G   2.32 4.78 
0  (10 ) 4 kN

m
 2.32 0 4.78 
The grid matrix equation becomes
F2 y  22  3.10 2.32 - 2.32  d 2 y 
     (10 4 )
 2x
M  0 
  2 .32 4.78 0  2x 
M  0   2.32 0   
 2 z   4.78   2z 
Solving for the displacement and the rotations in Eq.
(5.4.46), we obtain
d 2 y  0.259  10 2 m
 2 x  0.126  10 2 rad
 2 z  0.126  10  2 rad
Element 1
Using Eq. (5.4.17) for k̂ G , Eq. (5.4.18) for T G , and Eqs.
(5.4.47), we obtain
1 0 0 0 0 0  0 
0 1 0 0 0 0  0 
  
0 0 1 0 0 0  0 
TGd    2 
 0 0 0 1 0 0   0 .259  10 
0 0 0 0 1 0  0.126  10  2 
  
 0 0 0 0 0 1   0.126  10 
2

Multiplying the matrices, we have

 0 
 0 
 
 0 
TGd  2 
  0 .259  10 
 0.126  10   2
 
 0.126  10 
2
Using Eqs. (5.4.17), (5.4.41), and (5.4.48), we obtain the local element forces as
 fˆ1 y  1.55 0 2.32 - 1.55 0 2.32   0 
    
mˆ 1x   0.128 0 0 - 0.128 0 
0 
mˆ   2.32   0 
 1z  4.65 - 2.32 0
  (10 ) 
4
  2 
ˆ
 2y 
f  1.55 0 - 2.32   0. 259  10 
   0.128 0   0.126  10  - 2
 mˆ 2 x    
mˆ 2 z  
Symmetry 4.65 
  0.126  10 
 -2

Multiplying the matrices in Eq. (5.4.49), we obtain

fˆ1 y  11.0 kN mˆ 1x  1.50 kN  m mˆ 1z  31.0 kN  m
fˆ2 y  11.0 kN mˆ 2 x  1.50 kN  m mˆ 1z  1.50 kN  m
Element 2
We can obtain the local element forces for element 2 in a
similar. Because the procedure is the same as that used to
obtain the element 1 local forces, we will not show the
details but will only list the final results:
fˆ2 y  11.0 kN mˆ 2 x  1.50 kN  m mˆ 2 z  1.50 kN  m
fˆ2 y  11.0 kN mˆ 2 x  1.50 kN  m mˆ 3 z  31.0 kN  m

Free-body diagrams showing the local element forces are

shown in Figure 5-22.
Example 5.9
Solve for the displacement and rotation at node 3 for the
beam in Figure 5-33 by using substructuring. Let
L  29  10 3 ksi and I  1000 in4.
To illustrate the substructuring concept, we divide the beam
into two substructures, labeled 1 and 2 in Fidure 5-34. The
10-kip force has been assigned to node 3 of substructure 2,
although it could have been assigned to either
substructure or a fraction of it assigned to each
substructure. © 2002 Brooks/Cole Publishing / Thomson Learning™

Figure 5-33 Beam analyzed by substructuring

© 2002 Brooks/Cole Publishing / Thomson Learning™

Figure 5-34 Beam of Figure 5-33 separated into substructures

The stiffness matrix for each beam element is given by Eq.
(4.1.14) as
16.78 12d3 y  7203  12d 2 y  7202  0
16.78  720d3 y  576003  720d 2 y  288002   0
16.78  12d3 y  7203  24d 2 y  2   20
16.78  720d3 y  288003  0d 2 y  1152002  0
  12 720   12 720   24
1
0   12 720   d3 y 
16.78       0 115200   720 28800    
  720 57600   720 28800       3 
1
0   12 720   24 0   20 
     0 115200   0 
  
0 720 28800    
 25.17 3020  d3 y  10 
 3020 483264       600 
  3   
 12 720 12 720   d3 y   10 
 720 57600 720     0 
  3    
28800
16.78  
 12 720 12  12 720  720  d 4 y   0 
 720 28800 720  720 57600  57600     1200 
  4   
   12 720   12 720   24 0   12 720   d3 y 
1

16.78       0 115200   720 28800    

  720 57600   720 28800       3 
1
10   12 720   24 0   0 
    0 115200  1200 
 0   720 28800    
 25.17 3020  d3 y  17.5
 3020 483264       300 
  3   
50.34 0  d3 y  27.5
 0    
 966528 
 3   300 
d3 y  0.5463 in.
3  0.0003104 rad