Chapter 4

ANALYSIS OF TRUSSES AND FRAMES

The derivation of element equations for one-dimensional structural elements is considered in
previous chapter. These elements can be used for the analysis of bar type systems like planar
trusses, space trusses, beams, continuous beams, planar frames, grid systems and space
frames. Pin-jointed bar elements are used in the analysis of trusses. In planar truss analysis,
each of the two nodes can have components of displacement parallel to x and y axes. In three
dimensional truss analysis, each of the two nodes can have components of displacement
parallel to x, y and z directions. Rigidly jointed bar (beam) elements are used in the analysis
of frames. Thus a frame or a beam element is a bar which can resist not only axial forces also
transverse loads and bending moments. In the analysis of planar frames, each of the two
nodes of an element will have two translational displacement components (parallel to global x
and y axes) and a rotational displacement (in the plane of x y). For a space frame element,
each of the two ends is assumed to have three translational displacement components (parallel
to global x, y and z axes) and three rotational displacement components (one in each of the
three planes xy, yz, zx).
If local axes for a finite element are not parallel to global axes for the whole structure,
rotation-of-axes transformations must be applied to element stiffnesses, nodal displacements,
and nodal loads. Thus, when the elements are assembled, the resulting nodal equilibrium
equations will pertain to the global directions at each node. (It is also possible to apply
translation-of-axes transformations in order to refer all nodal equilibrium equations to the
same origin, but this type of operation is unnecessary.)

4.1 TRUSSES
A truss structure consist only of two-force members. That is, every truss element is in direct
tension or compression. In a truss, it is required that all loads and reactions are applied only at
the joints, and that all members are connected together at their ends by frictionless pin joints.
Every engineering student has, in a course on statics, analyzed trusses using the method of
joints and the method of sections. These methods, while illustrating the fundamentals of
statics, become tedious when applied to large-scale statically indeterminate truss structures.
Further, joint displacements are not readily obtainable. The finite element method on the other
hand is applicable to statically determinate or indeterminate structures alike. The finite
element method also provides joint deflections. Effects of temperature changes and support
settlements can also be routinely handled.

56

4.1.2 Plane Trusses

The local coordinate system is the x axis running along the element, since a truss element
simply a two-force member. Consequently, the nodal displacement vector in local coordinates is

{q} = q1 , q2 T

(4.1)

The nodal displacement vector in global coordinates is now

{q} = q1 , q 2 , q 3 , q 4 T

(4.2)
x

q4

q2
q3

q2

q1
q1
x
Figure 4.2 Plane truss element.

Transformation between local and global coordinates can be written as

{q } = []{q}
where the transformation matrix [] is given by

[ ] =

l m 0 0

0 0 l m

(4.3)

The element stiffness matrix in global coordinates is given by Eq. which yields

l2
lm

2
[k ] = E e A e lm2 m
le
l
lm

2
lm m

l 2 lm

lm m 2
l2
lm

lm
m 2

(4.4)

57

Any element load vector, {f}, in the local coordinate system can be transformed into the
global coordinates as

{f } = [ ]T {f }
For example, the element temperature load in the global coordinates can be obtained as
l 0
l
m 0
m
1

{f T } = [ ] {f T } =
E e A e e Te = E e A e e Te

0 l
1
l

m
0 m

(4.5)

The element stresses in global coordinates is given by

1
Le

{} = [E][B][]{q} = E e

q1
q1

q
1 l m 0 0 q 2 E e

l m l m 2 (4.6)
=

L e 0 0 l m q 3 L e
q 3
q 4
q 4

Example 4.1 Consider the four-bar truss shown in Fig. E4.1a. It is given that E = 70 Gpa and
Ae = 200 mm2 for all elements. The system is subjected two loads of 25 kN and 20 kN.
(a) Determine the element stiffness matrix for each element
(b) Assemble the structural (global) stiffness matrix [K] for the entire truss.
(c) Using the elimination approach, solve for the nodal displacement.
(d) Recover the stresses in each element
(e) Calculate the reaction forces.

25 kN

300 mm

y
20 kN
x
400 mm
Figure E4.1

Solution:

Lets model the truss with the bar elements.

58
y
25 kN
Q8

Q6

Q7

Q5

4
3

Q2

Q4
Q1

Q3

20 kN

a) It is recommended that a tabular form be used for representing nodal coordinate data and
element information as shown below. The nodal coordinate data is
Node
1
2
3
4

x (mm)
0
400
400
0

y (mm)
0
0
300
300

The element connectivity table is

Element
1
2
3
4

1
1
3
1
4

2
2
2
3
3

Note that the user has a choice in defining element connectivity. For example, the
connectivity of element 2 can be defined as 2-3 instead of 3-2 as above. However,
calculations of the direction cosines will be consistent with adopted connectivity scheme. For
example, the direction cosines of element 3 are obtained as

l = (x 3 x 1 ) / l 3

= (400 0) / 500
= 0.8

m = (y 3 y1 ) / l 3

= (300 0) / 500
= 0.6

Similarly, using formulas in Eqs 4.6 and 4.7, together with the nodal coordinate data and
element connectivity information given above, we obtain the direction cosines table:
Element
1
2
3
4

Le
400
300
500
400

l
1.0
0.0
0.8
1.0

m
0.0
-1.0
0.6
0.0

59

Now, using Eq. 4.13, the element stiffness matrices for the elements can be written as
1

Global serbestlik derecesi

0 1 0 1
0 0 0 2
0 1 0 3

0 0 0 4

1
0
4
k1 = 7 x10 x 200

1
400

0
1

0 0
0 1
4
k 2 = 7 x10 x 200

0 0
300

0 1

0.64 0.48 0.64 0, 48 1

0.48 0.36 0.48 0.36
4
2
k 3 = 7 x10 x 200

0.64 0.48 0.64
0.48 5
500

0.36 6
0.48 0.36 0.48

0 0
0 1
0 0

0 1

5
6
3
4

0 1 0 7
0 0 0 8
0 1 0 5

0 0 0 6

1
0
4
k 4 = 7 x10 x 200

1
400

The global dofs are indicated in the stiffness matrices of elements. These numbering scheme
assist in assembling the various element stiffness matrices.
(b) The structural stiffness matrix [K] is now assembled from the element stiffness matrices.
By adding the element stiffness contributions, noting the element connectivity, we get
1

0
0
7.68 5.76
22.68 5.76 15.0
5.76
4.32
0
0
0
5.76 4.32

15.0
0
15.0
0
0
0
0

3
0
0
0
20.0
0
0
20.0
[K ] = 7 x10
0
0
22.68 5.76 15.0
7.68 5.76
3

0
0
20.0 5.76 24.32
5.76 4.32
0
0
0
0
0
15.0
15.0

0
0
0
0
0
0
0

0
0
0

0
0

0
0

1
2
3
4
5
6
7
8

(c) The structural stiffness matrix [K] given above needs to be modified to account for the
boundary conditions (Q1 = Q2 = Q4 = Q7 = Q8 = 0). The elimination approach will
be used here. The rows and columns corresponding to dofs 1,2,4,7, and 8, which
correspond to fixed supports, are deleted from the [K] matrix above. The reduced
finite element equations are given as

0
0 Q 3
20
15.0
7 x10 3

3
0
22.68 5.76 Q 5 = 10 0

3
25
0
5.76 24.32 Q 6

Solution of the above equations yields the displacements

Q3 = 0.571 mm,

Q5 = 0.119 mm,

Q6 = -0.469 mm,

The nodal displacement vector for the entire structure can therefore be written as

60

{Q} = [0

0 0.571 0 0,119 0.469 0 0] mm

T

(d) The stress in each element can now be determined from Eq. , as shown below. The
connectivity of element 1 is 1 2. Consequently, the nodal displacement vector for
T
element 1 given by {q} = 0 0 0.571 0 , and Eq. yields
0
0
7 x10

2
[ 1 0 1 0]
1 =
= 99.925 N/mm (MPa)
400
0.571
0
4

The stress in member 2 is given by

0.119
0.469
7 x10 4

2
2 =
[0 1 0 1]
= 109.43 N/mm (MPa)
0
.
571
300

0
Following similar steps, we get
3 = -26,068 MPa
4 = 20,825 MPa
(e) The final step is to determine the support reactions. We need to determine the reaction
forces along dofs 1, 2, 4, 7, and 8, which correspond to fixed supports. These are
obtained by substituting for {Q} into the original finite element equation
{R} = [K R ]{Q} {FR } . In this substitution only those rows of [K] corresponding to the
support dofs are needed, and {F} ={0} for these dofs. Thus we have
0
0
15814
R
0
- 7.68 - 5.76
0 0
1
22.68 5.76 - 15.0

0
.
571

0
0
- 5.76 - 4,32
0 0
3128
2 7 x10 3 5.76 4.32
0
= 21887 N
R 4 =
0
0
0
20.0
0
20.0 0 0

0.119
R
0
0
0
0
15
.
0
0
15
.
0
0

7
0.469 4165

R 8
0
0
0
0
0
0
0 0
0
0

61

q5

q2
q4

q2

q6
q1

q1
q3
x
z

Figure 4.1 Three dimensional truss element.

4.1.1 Three-Dimensional Trusses

The local coordinate system is the x axis running along the element, since a truss element
simply a two-force member. Consequently, the nodal displacement vector in local coordinates
is

{q} = q1 , q2 T

(4.7)

The nodal displacement vector in global coordinates is now

{q} = q1 , q 2 , q 3 , q 4 , q 5 , q 6 T

(4.8)

Transformation between local and global coordinates can be written as

{q } = []{q}

(4.9)

where the transformation matrix [] is given by

[] = ox
l

m ox

n ox

l ox

m ox

or we simply say
lox = l

mox = m

nox = n

0
n ox

(4.10)

62

[] =

l m n 0 0 0

0 0 0 l m n

(4.11)

The element stiffness matrix in global coordinates can be obtained using the formula

[k ] = [ ]T [k ][ ]

(4.12)

If we substitute Eqs. (3.13) and (4.5) into Eq. (4.6), we obtain

l2
lm
ln
l 2 lm ln

lm
m2
mn lm m 2 mn

mn
n 2 ln mn n 2
Ee Ae ln
[k ] =

le l 2 lm ln l 2
lm
ln

lm m 2 mn lm
m2
mn

ln mn n 2
ln
mn
n 2

(4.13)

Any element load vector, {f}, in the local coordinate system can be transformed into the
global coordinates as

{f } = [ ]T {f }

(4.14)

The element stresses in global coordinates is given by

{} = [E][B]{q} = [E][B][]{q}

(4.15)

4.2 FRAMES
Frames, like trusses, are skeletal structures composed of slender members. However,
unlike trusses, the members of a frame transmit shear and bending, as well as axial loads.
Therefore, the individual frame elements behave like a beam with superimposed axial load.
The joints of a frame are usually considered to be rigid, althoug pin joints may be found in the
structure. Moments can be transfered from one member to another member to another accross
a rigid joint, but not accross a pin.

4.2.1 SPACE FRAME ELEMENT

The 12 x 12 stiffness matrix given in Eq. (3.60) is with respect to the local xyz
coordinate system. Since the nodal displacements in the local and global systems are related
by the relation

63
q 1 l ox
q l
2 oy
q 3 l oz

q 4 0
q 5 0

q 6 0
=
q 7 0
q 8 0

q 9 0
q 0
10
0
q 11
q 0
12

m ox
m oy

n ox
n oy

0
0

0
0

0
0

0
0

0
0

0
0

0
0

0
0

m oz
0
0
0

n oz
0
0
0

0
l ox
l oy
l oz

0
m ox
m oy
m oz

0
n ox
n oy
n oz

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0
0

0
0
0

0
0
0

0
0
0

0
0
0

0
0
0

l ox
l oy
l oz

m ox
m oy
m oz

n ox
n oy
n oz

0
0
0

0
0
0

0
0
0

0
0
0

0
0
0

0
0
0

0
0
0

0
0
0

0
0
0

0
0
0

l ox
l oy
l oz

m ox
m oy
m oz

0 q1
0 q 2
0 q 3

0 q 4
0 q 5

0 q 6
(4.16)
0 q 7

0 q8

0 q 9

n ox q 10

n oy q 11
n oz q 12

The transformation matrix, [], can be identified to be

[ ]12 x12

[ ] [0] [0] [0]

[ ] [0] [0]
[0] [ ] [0]

[0] [0] [ ]

[0]
=
[0]

[0]

(4.17)

q11

q11

q8

q8
y

q5

q5
q2
y

q3

q4

q2 q1

q9
q9
q12

q1

q4

q3

q6

q6

x
z
Figure 4.3 Three dimensional frame element.

q10
q7
q7
q12

q10

64
where

[ ]

3x3

l ox
= l o y
l o z

n ox
n o y
n o z

m ox
m oy
m oz

(4.18)

and

[0]3 x3

0 0 0
= 0 0 0
0 0 0

(4.19)

Here lox, mox, nox denote the direction cosines of x-axis (line ij in Fig. 5.14); loy, moy, noy
represent the direction cosines of y-axis; and loz, moz, noz indicate the direction cosines of zaxis with respect to global x, y, z axes. It can be seen that finding the direction cosines of xaxis is a straight forward computation since
l ox =

x j xi
L

, m ox =

y j yi
L

n ox =

z j zi

(4.20)

where xk, yk, zk indicate the coordinates of node k (k= i, j) in the global system. However, the
computation of the direction cosines of y- and z-axes requires some special effort. Finally
the stiffness matrix of the element with reference to the global coordinate system can be
obtained as

[k ] = [ ]T [k ][ ]
(4.21)
x

q5

q4

q5

q4

q6

y
x

q2
y

q1

q2
q3

q1
q3
x

q6

65
Figure 4.4 Planar frame element.

4.2.2 Plane Frame Element

In the case of two-dimensional (planar) frame analysis, we need to use an element
having six degrees of freedom as shown in Fig. 5.17. This element is assumed to lie in xy
plane and has two axial and four bending degrees of freedom. By using a linear interpolation
model for axial displacement and a cubic model for the transverse displacement, and
superimposing the resulting two stiffness matrices, the following stiffness matrix can be
T
obtained (the vector {q} is taken as {q } = q1 , q 2 , q 3 , q 4 , q 5 , q 6 )

AL2

I zz
0

[k ] = EI3zz 0 2
L AL

I zz
0

12
6L
0

4L2
0

12 6L
6L

2L2

AL2
I zz
0
0

Symmetric

12

6L
4L2

(4.22)

It is to be noted that the bending and axial deformation effects are uncoupled while deriving
Eq. (4.21). Equation (4.21 can also be obtained as a special case of Eq. (3.60) by deleting
rows and columns 2, 4, 5, 8, 10, 11.)
In this case the stiffness matrix of the element in the global xy coordinate system can be
found as

[k ] = [ ]T [k ][ ]
where
l ox
l
oy
0
[] =
0
0

m ox
m oy
0
0
0
0

0 0
0 0
1 0
0 l ox
0 l oy
0 0

0
0
0
m ox
m oy
0

0
0
0

0
0

(4.23)

with (lox , mox) denoting the direction cosines of x-axis and (loy , moy) indicating the
direction cosines of y-axis with respect to the global xy system.
Example: The frame shown in Figure is made from St 37 steel and subjected to a distributed
load of 20 kN/m and a concentrated load of 40 kN.
(a) Model the frame with two beam elements.
(b) Calculate the element stiffness matrices and load vectors and then assemble them to
obtain the stiffness matrix and load vector of the frame.

66
(c) Solve the LAES for the nodal displacements and rotations. Plot the undeformed and
deformed structure.
(d) Obtain the element stresses.
(e) Calculate the reaction forces.
Cross-section: DIN 1025 St37 IPB 100
y
20 kN/m
40 kN
0.8 m

2m

2m

30

Solution:
y
Q5

Q2
Q1

Q4

Q3

Q6

Q8

x
3
Q9

a) The nodal coordinate table is

Node
1
2
3

x (m)
0
2
3.732

y (m)
0
0
-0.5

The element connectivity table is

Element
1
2

1
1
2

2
2
3

Q7

67

The direction cosines table is

Element
1
2

Le
2
2

l
1.0
0.866

m
0.0
-0.5

Properties of IPB 100 cross-section are

Area
Iyy
Izz

= 0.24800x10-2 m2
= 0.43227 x10-5 m4
= 0.16681 x10-5 m4

The element stiffness matrices can be calculated as

0.248 x102 x 22
0.43227 x105

9
5
0
200 x10 x 0.43227 x10

k 1 =
3
2
0.248 x102 x 22
0.43227 x105

12
6 x2

4 x 22

12 6 x 2
6 x2

0.248 x102 x 22
0.43227 x105
0

2 x 22

Symmetric

12

2
6 x 2
4 x2

2294.86

0
12
Symmetric

0
12
16

= 108067.5

0 2294.86
2294.86 0

12 12
0
0
12

0
12
8
0
12
16

The all properties of the element 2 are same with those of the element 1 in local axes, the
stiffness matrices equal

k 2 = k 1
Since the local axes of element one are parallel to global axes, there is no need the
transformations for the stiffness matrix and load vector of element 1.
k 1 = k 1

The stiffness matrix and load vector of element 2 should be transformed into global axes. The
transformation matrix can be calculated as
lox = cos 30 = 0.866 mox = cos 120 = -0.5, loy = cos -60 = 0.5 moy = cos 30 = 0.866,

68

0
0
0
0.866 0.5 0
0.5 0.866 0
0
0
0

0
1
0
0
0
0
[ ] =

0
0 0.866 0.5 0
0
0
0
0 0.5 0.866 0

0
0
0
0
1
0
Now, the element stiffness matrix of element 1 in global coordinates can be determined as
853.653

492.835 284.609

0.03112
0.05391
0.06570

T
[k ] = [ ] [k ][ ] = 109

853.652 492.835 0.03112 853.653

492.835 284.609 0.05391 492.835 284.609

0.05391 0.05879 0.03112 0.05391 0.06570

0.03112

Assembly procedure gives the stiffness matrix of the structure,

1138.25

0
0.06225

0
0.06225 0.06225

0
0.06570 1191.90
1138.25

0.06225 0.06225 492.853 284.671

[ K ] = 109 0

0
0.06225 0.05879 0.03112 0.00834 0.13140

0
0
0
853.652 492.835 0.03112 853.653

0
0
0
492.835 284.609 0.05391 492.835 284.609

0
0
0
0.03112
0.05391
0.05879
0.03112
0.05391
0.06570

Now, let us calculate element load vectors.

{f } = b2x L

1
b

by L

by L2

12

bx L
2

by L
2

b L2
y
12

For element 1 bx = 0, and by = -20 kN/m, and load vector obtained as follows

{f } = 0
1
b

20000 6666.7 0 20000 6666.7

A concentrated load is applied on element and equivalent element loads can be calculated as

69

0
Fx N1a

F N
40000 N
1b
y 1b

F
N

40000
N

{fb2 } = Fy N 2b = 0 2b
x 2a

Fy N 3b
40000 N 3b

Fy N 4 b x =0.8 40000 N 4 b x =0.8

The values of shape functions at the specified point on the element are obtained as
1
1
2 x 3 3x 2 L + L3
= 3 2(0.8) 3 3(0.8) 2 2 + 23 = 0.648
3
x = 0.8
2
L
1
1
( N 2 ) x =0.8 = 3 x 3 L 2 x 2 L2 + xL3 x =0.8 = 3 (0.8)3 2 2(0.8)2 22 + (0.8)23 = 0.288
2
L
1
1
( N 3 ) x =0.8 = 3 2 x 3 + 3x 2 L x =0.8 = 3 2(0.8)3 + 3(0.8)2 2 = 0.736
2
L
1 3
1
( N 4 ) x =0.8 = 3 x L x 2 L2 x =0.8 = 3 (0.8)3 2 (0.8)2 22 = 0.256
2
L
nd
and the 2 element load vector in local coordinates yields as follows

( N1b ) x =0.8 =

0
25920

11520

{fb2 } = 0

29440

10240
A transformation for the first element is not necessary since the local and global axes are in
same directions. The load vector of element 2 in the global axes can be obtained as
0
0
0 0 12960
0.866 0.5 0
0.5 0.866 0
0
0
0 25920 22447

0
0
1
0
0
0
11520
11520

T
{fb2 } = [] {fb2 } = 0
=

0
0 0.866 0.5 0 0 14720

0
0
0 0.5 0.866 0 29440 25495

0
0
0
0
1 10240 10240
0
The load vector of entire structure is

{F} = 0

20000 6668 12960 2447 18188 14720 25495 10240

Now, we can solve the LAES applying the boundary conditions as Q1 = Q2 = Q3 = Q7 = Q8 =

Q9 = 0.

70
1191.90 492.853 0.03112 Q4 12960

10 492.853 284.671 0.00834 Q5 = 2447

0.03112 0.00834 0.13140 Q6 18188

9

Q4 = 0.00004396 m, Q5 = 0.00008877 m, Q6 = 0.13841 rad

4.2.3 Beam Element

A beam element is one-dimensional element which can resist bending moments in its plane. If
the beam lies in the xy plane, the corresponding degrees of freedom are shown in Fig. 4.4. By
assuming a cubic interpolation polynomial for the transverse deflection of the beam (v), the
stiffness matrix can be derived as
6L 12 6L
12
6L 4L2 6L 2L2

[k ] = EI3z
L 12 6L 12 6L

2
6L 4L2
6L 2L

(4.24)

q3

q1

q4
q2

x
L

Figure 4.5 Beam element.

If the beam element has an arbitrary orientation in xy plane as shown in Fig. 4.5, the stiffness
matrix of the beam element in the global xy system is given by

[k ]6 x6 = [ ]T 6 x 4 [k ]4 x 4 [ ]4 x6

(4.25)

where the transformation matrix [], is given by

loy

0
[ ] =
0

moy

0
0

1 0
0 loy

0
0
moy
0

0
0

(4.26)

71

where (loz, moz) are the direction cosines of z-axis with respect to xy system.