Scribd
Upload
36 views4 pages

1dstaticbar Example

This document describes using a finite element method to solve for displacements and reactions in a bar with a non-uniform cross-sectional area subjected to tensile loading. The bar is divided into two elements and stiffness matrices are assembled to set up the equations of equilibrium. Gaussian elimination is used to solve for the nodal displacements and reaction force.

Uploaded by

Miguel Marshall
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
36 views4 pages

1dstaticbar Example

This document describes using a finite element method to solve for displacements and reactions in a bar with a non-uniform cross-sectional area subjected to tensile loading. The bar is divided into two elements and stiffness matrices are assembled to set up the equations of equilibrium. Gaussian elimination is used to solve for the nodal displacements and reaction force.

Uploaded by

Miguel Marshall
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

!

"#$%&'# )$*+,#-

.*/#' 0 123# 456 7#%/&*8 9:;:5

"# A bar of length L is restrained at one end and subjected to a tensile force at the
other end. The cross-sectional area varies linearly from A at the loaded end to 2A at
the restrained end. By dividing the structure into two finite elements, each with a
constant cross-section and linear displacement variation between nodes, find an
approximate solution to this problem. Assume in your calculations that L = 20mm, A
= 4mm
2
, E = 1000 N/mm
2
and P = 100 N.

Solution:



Figure shows the problem definition and all the given parameters. The figure also
shows the finite element discretization. The length of both elements is 10 mm.

Average area of element 1 = (2A+1.5A)/2 = 7A/4 = 7 x 4/4 = 7 mm
2

Average area of element 2 = (1.5A+1A)/2 = 5A/4 = 5 x 4/4 = 5 mm
2



Stiffness matrix:

The general relationship for an element stiffness matrix is given as



In the given problem c = 0 and thus the element stiffness for this problem is reduced
to
k
(e)
=
AE
l
!
"
#
$
%
&
(e)
1 '1
'1 1
(
)
*
+
,
-

The element stiffness matrix for the first element is

k
(1)
=
7 !1000
10
"
#
$
%
&
'
(1)
1 (1
(1 1
)
*
+
,
-
.
= 700
1 (1
(1 1
)
*
+
,
-
.

Similarly the stiffness matrix for the second element is calculated as
k
(2)
=
5 !1000
10
"
#
$
%
&
'
(2)
1 (1
(1 1
)
*
+
,
-
.
= 500
1 (1
(1 1
)
*
+
,
-
.


Assembly of element stiffness matrices gives the following global stiffness matrix

K=
700 ! 700 0
!700 700 +500 !500
0 !500 500
"
#
$
$
$
%
&
'
'
'
=
700 ! 700 0
!700 1200 !500
0 !500 500
"
#
$
$
$
%
&
'
'
'


Load vector:

General form of the load or forcing vector is



For an inner element (not for a boundary element) this is



Note that both element of the given problem are boundary elements. Note that
[FN
a
(e)
]
x=l
(2)
is not zero. Since a point load is given at node 3 as shown in the figure (P
= 100 N), F at this node is equal to100 N. Since T = 0 for the given problem, only
load is the load created by the point load P = F at node 3. Thus the element load
vectors are given as

f
(1)
=
0
0
!
"
#
$
%
&
and
f
(2)
=
0
100
!
"
#
$
%
&


Assembly of load vectors gives
F =
0
0 +0
100
!
"
#
$
#
%
&
#
'
#

Now the assembled system of simultaneous equations may be written as

700 ! 700 0
!700 1200 !500
0 !500 500
"
#
$
$
$
%
&
'
'
'
u
1
u
2
u
3
(
)
*
+
*
,
-
*
.
*
=
0
0
100
(
)
*
+
*
,
-
*
.
*


Apply boundary conditions and introduce reaction: Note that at the restrained end
of the elastic bar (at node 1) displacement is zero, i.e., u
1
= 0. This is one of the two
boundary conditions. The other boundary condition is the force boundary condition
with P = F = 100 N at node 3 already included into the load vector. Applying the
given boundary condition and introduction of reaction (at node 1) gives

700 ! 700 0
!700 1200 !500
0 !500 500
"
#
$
$
$
%
&
'
'
'
u
1
= 0
u
2
u
3
(
)
*
+
*
,
-
*
.
*
=
R
1
0
100
(
)
*
+
*
,
-
*
.
*
(1)

Since u
1
is known, the first equation can be eliminated and the matrix may be
simplified as

1200 !500
!500 500
"
#
$
%
&
'
u
2
u
3
(
)
*
+
,
-
=
0 + 700u
1
100
(
)
*
+
,
-
=
0
100
(
)
*
+
,
-


Gaussian elimination (Refer to Chapter 2 in notes):

To make the presentation simpler (I suggest you practice without this step) we rewrite
the system of equations separately as

1200 u
2
- 500 u
3
= 0
-500 u
2
+ 500 u
3
= 100

We need to now eliminate the first term of the second equation, i.e., -500 u
2
. This can
be done my multiplying the first equation by (500/1200) and adding the result to the
second equation. Then, replace the second equation with the new values, i.e., the
second equation is replaced by

1200 (500/1200)u
2
500u
2
-500 (500/1200)u
3
+ 500 u
3
= 0 x (500/1200) + 100

This transforms the second equation into

0 u
2
+ 500 u
3
(2500/12) u
3
= 100
or
0 u
2
+ (3500/12) u
3
= 100

Note that the first equation is unchanged. Now, the system can be rewritten in matrix
form as
1200 !500
0
3500
12
"
#
$
$
%
&
'
'
u
2
u
3
(
)
*
+
,
-
=
0
100
(
)
*
+
,
-


The coefficient matrix is now an upper triangular matrix. Now u
3
may be computed as

u
3
=
100
3500
12
=
1200
3500
=
12
35
mm
Now back substitution of u
3
into the first equation gives the value for u
2
. Substituting
we get
1200u
2
!500u
3
=1200u
2
!500"
12
35
= 0

This gives

1200u
2
=
6000
35

or
u
2
=
6000
35!1200
=
1
7
mm
The reaction at the first node (node number 1) may now be computed from Equation
(1) as

700u
1
! 700u
2
= R
1


This gives R
1
= -100 N.

Final answers:

u
2
=
1
7
mm, u
3
=
12
35
mm, and R
1
= !100N

You might also like