3 - Bar Element

CIVL 7/8117 Chapter 3 - Truss Equations - Part 1 1/53
Chapter 3a – Development of Truss Equations
Learning Objectives
• To derive the stiffness matrix for a bar element.
• To illustrate how to solve a bar assemblage by the direct
stiffness method.
• To introduce guidelines for selecting displacement
functions.
• To describe the concept of transformation of vectors in
two different coordinate systems in the plane.
• To derive the stiffness matrix for a bar arbitrarily oriented
in the plane.
• To demonstrate how to compute stress for a bar in the
plane.
• To show how to solve a plane truss problem.
• To develop the transformation matrix in three-
dimensional space and show how to use it to derive the
stiffness matrix for a bar arbitrarily oriented in space.
• To demonstrate the solution of space trusses.
Development of Truss Equations

Having set forth the foundation on which the direct stiffness
method is based, we will now derive the stiffness matrix for
a linear-elastic bar (or truss) element using the general
steps outlined in Chapter 2.
We will include the introduction of both a local coordinate

system, chosen with the element in mind, and a global or
reference coordinate system, chosen to be convenient (for
numerical purposes) with respect to the overall structure.
We will also discuss the transformation of a vector from the

local coordinate system to the global coordinate system,
using the concept of transformation matrices to express
the stiffness matrix of an arbitrarily oriented bar element in
terms of the global system.

Next we will describe how to handle inclined, or skewed,
supports.
We will then extend the stiffness method to include space

trusses.
We will develop the transformation matrix in three-dimensional

space and analyze a space truss.
We will then use the principle of minimum potential energy

and apply it to the bar element equations.
Finally, we will apply Galerkin’s residual method to derive the

bar element equations.




Stiffness Matrix for a Bar Element

Consider the derivation of the stiffness matrix for the linear-
elastic, constant cross-sectional area (prismatic) bar
element show below.
This application is directly applicable to the solution of pin-

connected truss problems.

element show below.
where T is the tensile force directed along the axis at

nodes 1 and 2, x is the local coordinate system directed
along the length of the bar.

element show below.
The bar element has a constant cross-section A, an initial

length L, and modulus of elasticity E.
The nodal degrees of freedom are the local axial

displacements u1 and u2 at the ends of the bar.

du
The strain-displacement relationship is:     E
dx
From equilibrium of forces, assuming no distributed loads
acting on the bar, we get:
A x  T  constant
Combining the above equations gives:
du
AE  T  constant
dx
Taking the derivative of the above equation with respect to the
local coordinate x gives:
d  du 
 AE 0
dx  dx 

The following assumptions are considered in deriving the bar
element stiffness matrix:
1. The bar cannot sustain shear force: f1y  f2 y  0
2. Any effect of transverse displacement is ignored.
3. Hooke’s law applies; stress is related to strain:  x  E x


Step 1 - Select Element Type
We will consider the linear bar element shown below.
The bar element has a constant cross-section A, an initial

length L, and modulus of elasticity E.

Step 2 - Select a Displacement Function
A linear displacement function u is assumed: u  a1  a2 x
The number of coefficients in the displacement function, ai, is

equal to the total number of degrees of freedom associated
with the element.
Applying the boundary conditions and solving for the unknown
coefficients gives:
 u  u1   x x   u1 
u  2  x  u1 u   1    
 L   L  L  u2 

Step 2 - Select a Displacement Function
u 
Or in another form: u  N1 N2   1 
u2 
where N1 and N2 are the interpolation functions gives as:

x x
N1  1  N2 
L L
The linear displacement
function u plotted over the
length of the bar element is
shown here.

Step 3 - Define the Strain/Displacement
and Stress/Strain Relationships
du u2  u1
The stress-displacement relationship is: x  
dx L
Step 4 - Derive the Element Stiffness Matrix and Equations
We can now derive the element stiffness matrix as follows:

T  A x
Substituting the stress-displacement relationship into the
above equation gives:
 u  u1 
T  AE  2 
 L 

Step 4 - Derive the Element Stiffness Matrix and Equations
The nodal force sign convention, defined in element figure, is:

f1x  T f2 x  T
therefore, u u   u  u1 
f1x  AE  1 2  f2 x  AE  2 
 L   L 
Writing the above equations in matrix form gives:
 f1x  AE  1 1  u1 
   
f2 x  L  1 1 u2 
Notice that AE/L for a bar element is analogous to the spring

constant k for a spring element.

Step 5 - Assemble the Element Equations
and Introduce Boundary Conditions
The global stiffness matrix and the global force vector are
assembled using the nodal force equilibrium equations, and
force/deformation and compatibility equations.
n n
K  K    k ( e ) F  F    f ( e )
e 1 e 1
Where k and f are the element stiffness and force matrices

expressed in global coordinates.

Step 6 - Solve for the Nodal Displacements
Solve the displacements by imposing the boundary conditions
and solving the following set of equations:
F  Ku
Step 7 - Solve for the Element Forces

Once the displacements are found, the stress and strain in
each element may be calculated from:
du u2  u1
x    x  E x
dx L

Example 1 - Bar Problem
Consider the following three-bar system shown below. Assume
for elements 1 and 2: A = 1 in2 and E = 30 (106) psi and for
element 3: A = 2 in2 and E = 15 (106) psi.
Determine: (a) the global stiffness matrix, (b) the displacement

of nodes 2 and 3, and (c) the reactions at nodes 1 and 4.

For elements 1 and 2:
1 2 node numbers for element 1
1  30  10  6
1 1 lb 6  1 1 lb
k (1)  k (2)   1 1 in  10  1 1 in
30    
For element 3:
 2  15  106   1 1 lb 6  1 1 lb

k (3)
  1 1 in  10  1 1 in
30    
As before, the numbers above the matrices indicate the
displacements associated with the matrix.

Assembling the global stiffness matrix by the direct stiffness
methods gives: E1 E2 E3
 1 1 0 0 
 1 2 1 0 
K  106  
 0 1 2 1
 
 0 0 1 1
Relating global nodal forces related to global nodal
displacements gives:
 F1x   1 1 0 0   u1 
F   2 1 0  u2 
 2x  6  1  
   10
F3 x   0 1 2 1 u3 
F4 x   
 0 0 1 1 u4 

The boundary conditions are: u1  u4  0
 F1x   1 1 0 0   0 
F   2 1 0  u2 
 2x  6  1  
   10
F3 x   0 1 2 1 u3 
F4 x   
 0 0 1 1  0 
Applying the boundary conditions and the known forces

(F2x = 3,000 lb) gives:
3,000  6  2 1  2 
u
   10    
 0   1 2  u3 

Solving for u2 and u3 gives: u2  0.002 in

u3  0.001 in
The global nodal forces are calculated as:

 F1x   1 1 0 0   0  2,000 
F   2 1 0  0.002  3,000 
 2x  6 1
   10     lb
F3 x   0 1 2 1  0.001  0 
F4 x   
 0 0 1 1  0   1,000 

Selecting Approximation Functions for Displacements
Consider the following guidelines, as they relate to the one-
dimensional bar element, when selecting a displacement
function.
1. Common approximation functions are usually

polynomials.
2. The approximation function should be continuous within
the bar element.

function.
3. The approximating function should provide interelement

continuity for all degrees of freedom at each node for
discrete line elements, and along common boundary lines
and surfaces for two- and three-dimensional elements.

function.
For the bar element, we must ensure that nodes common
to two or more elements remain common to these
elements upon deformation and thus prevent overlaps or
voids between elements.
The linear function is then called a conforming (or

compatible) function for the bar element because it
ensures both the satisfaction of continuity between
adjacent elements and of continuity within the element.

function.
4. The approximation function should allow for rigid-body
displacement and for a state of constant strain within the
element.
Completeness of a function is necessary for

convergence to the exact answer, for instance, for
displacements and stresses.

The interpolation function must allow for a rigid-body
displacement, that means the function must be capable of
yielding a constant value.
Consider the follow situation: u  a1 a1  u1  u2
Therefore: u  N1u1  N2u2   N1  N2  a1
Since u = a1 then: u  a1   N1  N2  a1
This means that: N1  N2  1
The displacement interpolation function must add to unity at

every point within the element so the it will yield a constant
value when a rigid-body displacement occurs.

Transformation of Vectors in Two Dimensions
In many problems it is convenient to introduce both local and
global (or reference) coordinates.
Local coordinates are always chosen to conveniently

represent the individual element.
Global coordinates are chosen to be convenient for the whole

structure.

Given the nodal displacement of an element, represented by
the vector d in the figure below, we want to relate the
components of this vector in one coordinate system to
components in another.

Let’s consider that d does not coincide with either the local or
global axes. In this case, we want to relate global
displacement components to local ones. In so doing, we will
develop a transformation matrix that will subsequently be
used to develop the global stiffness matrix for a bar element.

We define the angle  to be positive when measured
counterclockwise from x to x’. We can express vector
displacement d in both global and local coordinates by:
d  u1i  v1 j  u1i  v1 j

Consider the following diagram:
Using vector addition: a  b  i
Using the law of cosines, we get: | a |  | i | cos  | a |  cos 
Similarly: | b |  | i | sin | b |  sin


The vector a is in the i direction and b is in the j direction,

therefore:
a  | a | i   cos  i b  | b |   j    sin   j 

The vector i can be rewritten as: i  cos  i  sin j

The vector j can be rewritten as: j  sin i  cos  j
Therefore, the displacement vector is:

u1i  v1 j  u1  cos  i  sin j   v1  sin i  cos  j   u1i  v1 j

Combining like coefficients of the local unit vectors gives:

u1 cos  v1 sin  u1 C  cos  u1   C S  u1 
   
u1 sin  v1 cos  v1 S  sin v1   S C  v1 

The previous equation relates the global displacement d to
the d local displacements
 C S
The matrix is called the transformation matrix:  
 S C 
The figure below shows u expressed in terms of the global
coordinates x and y.
u   Cu  Sv

Example 2 - Bar Element Problem
The global nodal displacement at node 2 is u2 = 0.1 in and
v2 = 0.2 in for the bar element shown below. Determine the
local displacement.
Using the following expression we just derived, we get:
u   Cu  Sv
u2  cos 60o (0.1)  sin60o (0.2)  0.223 in

Global Stiffness Matrix
We will now use the transformation relationship developed
above to obtain the global stiffness matrix for a bar element.

We known that for a bar element in local coordinates we have:
 f1x  AE  1 1  u1 

    f   kd
f2x  L  1 1 u2 
We want to relate the global element forces f to the global

displacements d for a bar element with an arbitrary
orientation.
 f1x   u1 
f   
 1y   v1 
 k  f = kd
f2 x  u2 
f2 y  v 2 

Using the relationship between local and global components,
we can develop the global stiffness matrix.
We already know the transformation relationships:
u1  u1 cos   v1 sin u2  u2 cos   v 2 sin
Combining both expressions for the two local degrees-of-

freedom, in matrix form, we get:
 u1  d = T *d
 u1  C S 0 0   v1 
   
u2   0 0 C S  u2  C S 0 0 
T*   
v 2   0 0 C S

A similar expression for the force transformation can be
developed.
 f1x 
 
 f1x  C S 0 0   f1y 
     f  T*f
f2x   0 0 C S  f2 x 
f2 y 
Substituting the global force expression into element force

equation gives:
f  = kd  T * f  kd
Substituting the transformation between local and global
displacements gives:
d = T * d  T * f  kT *d

The matrix T* is not a square matrix so we cannot invert it.
Let’s expand the relationship between local and global
displacement.
 u1   C S 00   u1 
 v    S 00  v1 
 1  C
  d = Td
 
u2   0 0 C S  u2 
v 2   0 
0 S C  v 2 
where T is:
C S 0 0
 S C 0 0
T 
0 0 C S
 
0 0 S C 

We can write a similar expression for the relationship between
local and global forces.
 f1x   C S 0   f1x 
0
f     
 1y  S C 0 0   f1y 
     f  = Tf
f
 2x   0 0 C S  f2 x 

f2y   0 
0 S C  f2 y 
Therefore our original local coordinate force-displacement

expression
 f1x  AE  1 1  u1 

    f  = kd
f2x  L  1 1 u2 

May be expanded:  f1x  1 0 1 0   u1 
f  
 1y  AE
0 0 0 0  v1 
     
f2 x  L  1 0 1 0  u2 
f2y   
0 0 0 0  v 2 
The global force-displacement equations are:

f   kd  Tf  kTd
Multiply both side by T -1 we get: f  T -1 k T d

where T-1 is the inverse of T. It can be shown that: T 1  T T

The global force-displacement equations become: f = T TkTd
Where the global stiffness matrix k is: k  T TkT
Expanding the above transformation gives:

 C 2 CS C 2 CS 
 2
CS S 
2
AE  CS S
k
L  C 2 CS C 2 CS 
 
 CS S
2
CS S2 
We can assemble the total stiffness matrix by using the above

element stiffness matrix and the direct stiffness method.
n n
K  K    k ( e ) F  F    f ( e ) F  Kd
e 1 e 1

Local forces can be computed as:
 f1x  1 0 1 0   u1  1 0 1 0   C S 00   u1 
f  
 1y  AE
0 0 0 0  v1  AE  0 0 0 0   S C 0 0   v1 
          
f
 2 x L  1 0 1 0  u2  L  1 0 1 0  0 0 C S  u2 
    
f2y  0 0 0 0  v 2  0 0 0 0  0 0 S C  v 2 
 f1x   Cu1  Sv1  Cu2  Sv 2 

f    
 1y  AE  0
   
f
 2 x L  Cu1  Sv 1  Cu 2  Sv 2 
 
f2y   0 

For the bar element shown below, evaluate the global stiffness
matrix. Assume the cross-sectional area is 2 in2, the length
is 60 in, and the E is 30 x 106 psi.
 C 2 CS C 2 CS 
 2
CS S 
2
AE  CS S
k
L  C 2 CS C 2 CS 
 
 CS S
2
CS S2 
3 1
Therefore: C  cos30o  S  sin30o 
2 2

The global elemental stiffness matrix is:
 3 3 
 3  3
4 4 4 4
 
k
 
(2 in 2 ) 30  106 psi 

3
4
1
4
 3
4
1 
4 
60 in  3 3 
 3 3
 4 4 4 4 
 3 3 
  1 1 
4 4 4 4 
Simplifying the global elemental stiffness matrix is:

 0.750 0.433 0.750 0.433 
 0.433 0.250 0.433 0.250  lb
k  106   in
 0.750 0.433 0.750 0.433 
 
 0.433 0.250 0.433 0.250 

Computation of Stress for a Bar in the x-y Plane
For a bar element the local forces are related to the local
displacements by:
 f1x  AE  1 1  u1 
   
f2x  L  1 1 u2 
The force-displacement equation for f2x is:

AE u 
f2x   1 1  1 
L u2 
The stress in terms of global displacement is:
 u1 
 
C S 0 0  v1  E
     Cu1  Sv1  Cu2  Sv 2 
E
   1 1 
L  0 0 C S  u2  L
v 2 

For the bar element shown below, determine the axial stress.
Assume the cross-sectional area is 4 x 10-4 m2, the length is
2 m, and the E is 210 GPa.
The global displacements are known as

u1 = 0.25 mm, v1 = 0, u2 = 0.5 mm,
and v2 = 0.75 mm.
E

L
 Cu1  Sv1  Cu2  Sv 2 
210  106  1 3 1 3 
   (0.25)  (0)  (0.5)  (0.75) KN m
2  2 4 2 4 
  81.32  103 kN m  81.32 MPa
2

Solution of a Plane Truss
We will now illustrate the use of equations developed above
along with the direct stiffness method to solve the following
plane truss example problems.
A plane truss is a structure composed of bar elements all lying

in a common plane that are connected together by
frictionless pins.
The plane truss also must have loads acting only in the
common plane.

Example 5 - Plane Truss Problem
The plane truss shown below is composed of three bars
subjected to a downward force of 10 kips at node 1. Assume
the cross-sectional area A = 2 in2 and E is 30 x 106 psi for all
elements.
Determine the x and y displacement at node 1 and stresses in
each element.

Element Node 1 Node 2  C S

1 1 2 90o 0 1
2 1 3 45o 0.707 0.707
3 1 4 0o 1 0

 C 2 CS C 2 CS 
Example 5 - Plane Truss Problem 
AE  CS S
2 2
CS S 
k
L  C 2 CS C 2 CS 
The global elemental stiffness matrix are: 
 CS S
2
CS

S2 
u1 v1 u2 v2
0 0 0 0
 1 lb
(2 in )(30  10 psi ) 0
2 6
1 0
 in
element 1: C 0 S 1  k (1) 
120 in 0 0 0 0
 
0 1 0 1
u1 v1 u3 v3
1 1 1 1
 1 1 lb
(2 in 2 )(30  106 psi )  1 1
 in
element 2: C 2
S 2  k (2) 
2 2
240 2 in  1 1 1 1
 
 1 1 1 1
u1 v1 u4 v4
1 0 1 0
 0  lb
element 3: C 1 S  0 (2 in 2 )(30  106 psi )  0 0 0
 in
 k (3) 
120 in  1 0 1 0
 
0 0 0 0

The total global stiffness matrix is:
 1.354 0.354 0 0 0.354 0.354 1 0  element 1
 0.354 1.354 0 1 0.354 0.354 0 0 
 0 0 0 0 0 0 0 0 element 2
K  5  10  0.354
5 0 1 0 1 0 0 0 0  lb
in
element 3
 0.354 0 0 0.354 0.354 0 0

 0.354
1
0.354
0
0
0
0
0
0.354
0
0.354
0
0
1
0
0

 0 
 0 0 0 0 0 0 0
The total global force-displacement equations are:

 0   1.354 0.354 0 0 0.354 0.354 1 0  u1 
10, 000  0.354 1.354 0 1 0.354 0.354 0 0  v 1 
 F2 x   0 0 0 0 0 0 0 0  0 
 F2 y    0 
 F3 x   5  10  0.354
5 0 1 0 1 0 0 0 0
 F3 y   0.354 0 0 0.354 0.354 0 0

 0 
 F   0.354
1
0.354
0
0
0
0
0
0.354
0
0.354
0
0
1
0
0
  00 
 F   0 
  0 
4x
 4y   0 0 0 0 0 0 0

Applying the boundary conditions for the truss, the above
equations reduce to:
 0   1.354 0.354 0 0 0.354 0.354 1 0  u1 

10, 000  0.354 1.354 0 1 0.354 0.354 0 0  v 1 
 F2 x   0 0 0 0 0 0 0 0  0 
 F2 y  0  0 
 F3 x   5  10  0.354
5 0 1 0 1 0 0 0
 
 F3 y   0.354 0 0 0.354 0.354 0 0
 0 
 F   0.354
1
0.354
0
0
0
0
0
0.354
0
0.354
0
0
1
0
 0
0 0 
 F4 x   0  0
0 
 4y   0 0 0 0 0 0  

Applying the boundary conditions for the truss, the above
 0  5 1.354 0.354   1 
u
   5  10    
10,000  0.354 1.354  v1 
Solving the equations gives: u1  0.414  10 2 in

v1  1.59  10 2 in
E
The stress in an element is:   Cu 1  Sv 1  Cu 2  Sv 2 
L 
where i is the local node number


1 1 2 90o 0 1
2 1 3 45o 0.707 0.707
3 1 4 0o 1 0
E
 Cu 1  Sv 1  Cu 2  Sv 2 
L 
30  106
element 1  (1) 
120
 v1   3,965 psi
30  106
element 2  (2)  
120
(0.707)u1  (0.707)v1   1,471 psi


1 1 2 90o 0 1
2 1 3 45o 0.707 0.707
3 1 4 0o 1 0
E
 Cu 1  Sv 1  Cu 2  Sv 2 
L 
30  106
element 3  (3)

120
 u1   1,035 psi

Let’s check equilibrium at node 1:
F x  f1(2)
x cos(45) f1x
(3)
F y x sin(45) f1x 10,000 lb

 f1(2) (1)

Let’s check equilibrium at node 1:
F x  (1,471 psi )(2 in 2 )(0.707)  (1,035 psi )(2 in 2 )  0
F y  (3,965 psi )(2 in 2 )  (1,471 psi )(2 in 2 )(0.707)  10,000  0

Develop the element stiffness matrices and system equations
for the plane truss below.
Assume the stiffness of each element is constant. Use the

numbering scheme indicated. Solve the equations for the
displacements and compute the member forces. All elements
have a constant value of AE/L

Develop the element stiffness matrices and system equations
for the plane truss below.
Member Node 1 Node 2 Elemental 

Stiffness
1 1 2 k 0
2 2 3 k 3/4
3 1 3 k /2

Compute the elemental stiffness matrix for each element. The
general form of the matrix is:
 C 2 CS C 2 CS 
 2
CS S 
2
AE  CS S
k
L  C 2 CS C 2 CS 
 
 CS S
2
CS S2 

Stiffness
1 1 2 k 0
2 2 3 k 3/4
3 1 3 k /2

For element 1: u
1
v
1
u
2
v
2
 1 0 1 0  u1
 0 0 0 0 v1
k (1) k 
 1 0 1 0 u2
 
 0 0 0 0 v2

Stiffness
1 1 2 k 0
2 2 3 k 3/4
3 1 3 k /2

For element 2: u2 v2 u3 v3
 1 1 1 1 u2
 
k 1 1 1 1 v2
k (2)  
2  1 1 1 1 u3
 
 1 1 1 1 v3

Stiffness
1 1 2 k 0
2 2 3 k 3/4
3 1 3 k /2

For element 3: u1 v1 u3 v3
0 0 0 0 u1
0 1 0 1 v1
k (3) k 
0 0 0 0 u3
 
0 1 0 1 v3

Stiffness
1 1 2 k 0
2 2 3 k 3/4
3 1 3 k /2

Assemble the global stiffness matrix by superimposing the
elemental global matrices.
u1 v1 u2 v2 u3 v3
 2 0 2 0 0 0 u1 element 1
 0 2 0 0 0 2  v1 element 2
  element 3
k  2 0 3 1 1 1 u2
K  
2 0 0 1 1 1  1 v2
 0 0 1 1 1  1 u3
 
 0 2 1 1 1 3 v3

The unconstrained (no boundary conditions satisfied)
equations are:  2 0 2 0 0 0   u1   F1x 
 0    
 2 0 0 0 2  v1   F1y 
k  2 0 3 1 1 1  u2   P1 
     
2 0 0 1 1 1 1  v 2  P2 
 0 0 1 1 1 1  u3   F3 x 
    
 0 2 1 1 1 3   v 3   F3 y 
The displacement at nodes 1 and 3 are zero in both directions.

Applying these conditions to the system equations gives:
 2 0 2 0 0 0   0   F1x 
 0  
 2 0 0 0 2  0   F1y 
k  2 0 3 1 1 1  u2   P1 
     
2 0 0 1 1 1 1  v 2  P2 
 0 0 1 1 1 1  0   F3 x 
    
 0 2 1 1 1 3   0   F3 y 

Applying the boundary conditions to the system equations
gives:
k  3 1 u2   P1 
  
2  1 1  v 2   P2 
Solving this set of equations is fairly easy. The solution is:

P1  P2 P1  3P2
u2  v2 
k k

Using the force-displacement relationship the force in each
member may be computed.
Member (element) 1:  f1x   Cu1  Sv1  Cu2  Sv 2 
f   
 1y  0
 k 
C 1 S0 f
 2 x  Cu1  Sv 1  Cu 2  Sv 2
f2 y   
 0 
 P  P2 
f1x  k  Cu2   k   1    P1  P2  fy 1  0
 k 
 P  P2 
f2 x  k Cu2   k  1   P1  P2 fy 2  0
 k 

Member (element) 2: f2 x   Cu2  Sv 2  Cu3  Sv 3 
f   
 2y  0
1 1    k  
C S f3 x   Cu2  Sv 2  Cu3  Sv 3 
2 2 f3 y   
 0 
f2 x  k Cu2  Sv 2 
 P  P2   1   P1  3P2   1  
 k  1         2P2
 k   2  k   2 
f3 x  k  Cu2  Sv 2 
 P  P2   1   P1  3P2   1  
 k  1        2P2
 k   2   k  2 

Member (element) 3:  f1x   Cu1  Sv1  Cu3  Sv 3 
f   
 1y  0
 k 
C 0 S 1 f
  3 x  Cu1  Sv 1  Cu3  Sv 3
f3 y   
 0 
f1x  0 f1y  0
f3 x  0 f3 y  0
The solution to this simple problem can be readily checked by

using simple static equilibrium equations.

Consider the two bar truss shown below.
Determine the displacement in the y direction of node 1 and

the axial force in each element.
Assume E = 210 GPa and A = 6 x 10-4 m2

The global elemental stiffness matrix for element 1 is:
3
cos  (1)   0.6  0.36 0.48 0.36 0.48 
5 
210  106 (6  10 4 )  0.48 0.64 0.48 0.64 
k (1) 
4 5  0.36 0.48 0.36 0.48 
sin (1)   0.8  
5  0.48 0.64 0.48 0.64 
Simplifying the above expression gives:

u1 v1 u2 v2
 0.36 0.48 0.36 0.48 

 0.48 0.64 0.48 0.64 
k (1)  25,200 
 0.36 0.48 0.36 0.48 
 
 0.48 0.64 0.48 0.64 

The global elemental stiffness matrix for element 2 is:
cos  (2)  0 0 0 0 0 

4
0 1
(210  10 )(6  10 ) 0
6
1
k (2) 
4 0 0 0 0
sin (2)
1  
0 1 0 1
Simplifying the above expression gives:

u1 v1 u3 v3
0 0 0 0 
0 1.25 0 1.25 
k (2)  25,200  
0 0 0 0 
 
0 1.25 0 1.25 

The total global equations are:
 F1x   0.36 0.48 0.36 0.48 0 0   u1 
F   0.48   element 1
 1y   1.89 0.48 0.64 0 1.25   v1 
F2 x   0.36 0  u2  element 2
0.48 0.36 0.48 0
   25,200   
F
 2y   0.48 0.64 0.48 0.64 0 0  v 2 
F3 x   0 0 0 0 0 0  u3 
    
 3 y 
F  0 1.25 0 0 0 1.25  v 3 
The displacement boundary conditions are:

u1   u2  v 2  u3  v 3  0

The total global equations are:
 F1x   0.36 0.48 0.36 0.48 0 0   u1 
P  
 F1y
 0.48 1.89 0.48 0.64 0 1.25  v1 
F2 x   0.36 0.48 0.36 0.48 0 0  0 
   25,200   
 0.48 0.64 0.48 0.64 0 0  0 
F
 2y 
F3 x   0 0 0 0 0 0  0 
    
 3 y 
F  0 1.25 0 0 0 1.25   0 
By applying the boundary conditions the force-displacement

P  25,200(0.48  1.89v1 )

Solving the equation gives: v1  (2.1 105 )P  0.25
By substituting P = 1,000 kN and  = -0.05 m in the above

equation gives:
v1  0.0337m
The local element forces for element 1 are:

 u1  0.05 
 f1x   1 1 0.6 0.8 0 0  v1  0.0337 
   25,200    
f2 x   1 1  0 0 0.6 0.8   u2 
 v2 
The element forces are: f1x  76.6 kN f2 x  76.7 kN

Tension

The local element forces for element 2 are:
 u1  0.05 
 
 f1x   1 1 0 1 0 0  v1  0.0337 
   31,500     
f3 x   1 1 0 0 0 1  u3 
 v3 
The element forces are: f1x  1,061 kN f3 x  1,061 kN
Compression

Transformation Matrix and Stiffness Matrix for
a Bar in Three-Dimensional Space
Let’s derive the transformation matrix for the stiffness matrix for
a bar element in three-dimensional space as shown below:

The coordinates at node 1 are x1, y1, and z1, and the coor-
dinates of node 2 are x2, y2, and z2. Also, let x, y, and z be
the angles measured from the global x, y, and z axes,
respectively, to the local axis.

The three-dimensional vector representing the bar element is
gives as:
d  ui  vj  wk  u  i  v  j  w  k

Taking the dot product of the above equation with i gives:
u(i  i)  v ( j  i)  w (k  i)  u 
By the definition of the dot product we get:

x2  x1 y 2  y1 z2  z1
i  i   Cx j  i   Cy k  i   Cz
L L L
where L  ( x2  x1 )2  ( y 2  y1 )2  ( z2  z1 )2
Cx  cos x Cy  cos y Cz  cos z
where Cx, Cy, and Cz are projections of i on to i, j, and k,

respectively.

Therefore: u   Cx u  Cy v  Czw
The transformation between local and global displacements is:
 u1 
v  d  T *d
 1

 u1  Cx Cy Cz 0 0 0   w1 
   
u 
 2  0 0 0 C x
C y C z   u2  C C C 0 0 0 
v2  T*   x y z 
   0 0 0 C x C y Cz 
w 2 

The transformation from the local to the global stiffness matrix
is: C 0  x
C 0 
 y 
C 0  AE  1 1 Cx Cy Cz 0 0 0 
k  T TkT k z   1 1  0 0 0 C C C 
 0 Cx  L   x y z
 0 Cy 
 
 0 Cz 
 Cx2 C xC y C xC z Cx2 CxCy CxCz 

 Cy2 Cy2 
 C xC y Cy Cz CxCy Cy Cz 
AE  CxCz Cz2 Cy Cz Cz 
2
C y Cz CxCz
k  
L  Cx2 CxCy CxCz Cx2 C xC y C xC z 
 C C Cy2 Cy Cz C xC y Cy2 C y Cz 
 x y 
 CxCz Cy Cz Cz2 C xC z C y Cz Cz2 

The global stiffness matrix can be written in a more convenient

form as:
 Cx 2 C xC y C x Cz 
AE      
k   CxCy Cy 2
C y Cz 
L    
CxCz C y Cz Cz 2 


Example 8 – Space Truss Problem
Consider the space truss shown below. The modulus of
elasticity, E = 1.2 x 106 psi for all elements. Node 1 is
constrained from movement in the y direction.
To simplify the stiffness matrices

for the three elements, we will
express each element in the
following form:
AE    
k  
L   

Consider element 1: L(1)  ( x2  x1 )2  ( y 2  y1 )2  ( z2  z1 )2
L(1)  ( 72)2  (36)2  80.5 in
72
Cx   0.894
80.5
36
Cy   0.447
80.5
Cz  0
 0.80 0.40 0
   0.40 0.20 0

 0 0 0 

Consider element 1:
u v w u v w
1 1 1 2 2 2
(0.302 in )(1.2  10 psi )  

2 6
  lb
k    
in
80.5 in 

Consider element 2: L(2)  ( x3  x1 )2  ( y 3  y1 )2  ( z3  z1 )2
L(2)  ( 72)2  (36)2  (72)2  108 in

72
Cx   0.667
108
36
Cy   0.333
108
72
Cz   0.667
108
 0.444 0.222 0.444 

   0.222 0.111 0.444 

 0.444 0.444 0.444 

Consider element 2:
u v w u v w
1 1 1 3 3 3
(0.729 in )(1.2  10 psi )  

2 6
  lb
k    
in
108 in 

Consider element 3: L(3)  ( x 4  x1 )2  ( y 4  y1 )2  ( z4  z1 )2
L(3)  ( 72)2  ( 48)2  86.5 in
72
Cx   0.832
86.5
Cy  0
48
Cz   0.555
86.5
0.692 0 0.462 
   0 0 0 

0.462 0 0.308 

Consider element 3:
u v w u v w
1 1 1 4 4 4
(0.187 in )(1.2  10 psi )  

2 6
  lb
k    
in
86.5 in 
The boundary conditions are:
u2  v 2  w 2  0
u3  v 3  w 3  0
u4  v 4  w 4  0
v1  0

Canceling the rows and the columns associated with the
boundary conditions reduces the global stiffness matrix to:
u1 w1
 8,997 2,403 
K 
 2,403 4,398 
The global force-displacement equations are:

 8,997 2,403   u1   0 
 2,403   
 4,398  w1  1,000 
Solving the equation gives:
u1  0.0711 in w1  0.2662 in

It can be shown, that the local forces in an element are:
 ui 
v 
 i
 fix  AE  Cx Cy Cz Cx Cy Cz  w i 
    
f jx  L  Cx Cy Cz Cx Cy Cz   u j 
v j 
 
w j 
The stress in an element is:
 ui 
v 
 i
E w i 
  Cx  Cy  Cz Cx Cy Cz   
L uj 
v j 
 
w j 

The stress in element 1 is:  0.0711
 0 
 
1.2  106  0.2662
 (1)  0.894  0.447 0  0.894 0.447 0. 
0

80.5  
 0 
 
 (1)  948 psi  0 

 0 
 
1.2  106  0.2662
 (2)  0.667  0.333  0.667  0.667 0.333 0.667   
108  0 
 0 
 
 0 
 (2)  1,445 psi

 0 
 
1.2  106 0.2662 
 (3)  0.832 0 0.555  0.832 0  0.555   
86.5  0 
 0 
 
 0 
 (3)  2,843 psi

Problems:
2. Verify the global stiffness matrix for a three-dimensional
bar. Hint: First, expand T* to a 6 x 6 square matrix, then
expand k to 6 x 6 square matrix by adding the
appropriate rows and columns of zeros, and finally,
perform the matrix triple product k = TTk’T.
3a. Do problems 3.4, 3.10, 3.12, 3.15a,b, 3.18, 3.23, 3.37,

3.43, and 3.48 on pages 146 - 165 in your textbook “A
First Course in the Finite Element Method” by D. Logan.
End of Chapter 3a

3 - Bar Element

Uploaded by

Copyright:

Available Formats

You might also like

3 - Bar Element

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

3 - Bar Element

Uploaded by

Copyright:

Available Formats

CIVL 7/8117 Chapter 3 - Truss Equations - Part 1 1/53

Chapter 3a – Development of Truss Equations

Development of Truss Equations

We will include the introduction of both a local coordinate

We will also discuss the transformation of a vector from the

Development of Truss Equations

We will then extend the stiffness method to include space

We will develop the transformation matrix in three-dimensional

We will then use the principle of minimum potential energy

Finally, we will apply Galerkin’s residual method to derive the

Development of Truss Equations

Development of Truss Equations

Development of Truss Equations

Development of Truss Equations

Development of Truss Equations

Development of Truss Equations

Development of Truss Equations

Development of Truss Equations

Stiffness Matrix for a Bar Element

This application is directly applicable to the solution of pin-

Stiffness Matrix for a Bar Element

where T is the tensile force directed along the axis at

Stiffness Matrix for a Bar Element

The bar element has a constant cross-section A, an initial

The nodal degrees of freedom are the local axial

Stiffness Matrix for a Bar Element

Stiffness Matrix for a Bar Element

1. The bar cannot sustain shear force: f1y  f2 y  0

2. Any effect of transverse displacement is ignored.

3. Hooke’s law applies; stress is related to strain:  x  E x

Stiffness Matrix for a Bar Element

We will consider the linear bar element shown below.

The bar element has a constant cross-section A, an initial

Stiffness Matrix for a Bar Element

A linear displacement function u is assumed: u  a1  a2 x

The number of coefficients in the displacement function, ai, is

Stiffness Matrix for a Bar Element

where N1 and N2 are the interpolation functions gives as:

Stiffness Matrix for a Bar Element

Step 4 - Derive the Element Stiffness Matrix and Equations

We can now derive the element stiffness matrix as follows:

Stiffness Matrix for a Bar Element

The nodal force sign convention, defined in element figure, is:

Writing the above equations in matrix form gives:

Notice that AE/L for a bar element is analogous to the spring

Stiffness Matrix for a Bar Element

Where k and f are the element stiffness and force matrices

Stiffness Matrix for a Bar Element

Step 7 - Solve for the Element Forces

Stiffness Matrix for a Bar Element

Determine: (a) the global stiffness matrix, (b) the displacement

Stiffness Matrix for a Bar Element

 2  15  106   1 1 lb 6  1 1 lb

Stiffness Matrix for a Bar Element

Stiffness Matrix for a Bar Element

Applying the boundary conditions and the known forces

Stiffness Matrix for a Bar Element