Professional Documents
Culture Documents
Yijun.Liu@uc.edu
http://urbana.mie.uc.edu/yliu
Copyright Notice
1997-2003 by Yijun Liu, University of Cincinnati.
All rights reserved. Permissions are granted for personal
and educational uses only. Any other uses of these lecture
notes (such as for classroom lectures outside the University
of Cincinnati, trainings elsewhere, and those of a
commercial nature) are not permitted, unless such uses
have been granted in writing by the author.
ii
Table of Contents
Copyright Notice ....................................................................................................... ii
Table of Contents ..................................................................................................... iii
Preface .......................................................................................................................v
Chapter 1. Introduction ............................................................................................1
I. Basic Concepts ...................................................................................................1
II. Review of Matrix Algebra .................................................................................7
III. Spring Element ..............................................................................................14
Chapter 2. Bar and Beam Elements.......................................................................25
I. Linear Static Analysis .......................................................................................25
II. Bar Element .....................................................................................................26
III. Beam Element .................................................................................................53
Chapter 3. Two-Dimensional Problems ................................................................75
I. Review of the Basic Theory ...............................................................................75
II. Finite Elements for 2-D Problems ..................................................................82
Chapter 4. Finite Element Modeling and Solution Techniques........................... 105
I. Symmetry ........................................................................................................ 105
II. Substructures (Superelements) ..................................................................... 107
III. Equation Solving......................................................................................... 109
IV. Nature of Finite Element Solutions ............................................................. 112
V. Numerical Error............................................................................................ 114
1997-2003 Yijun Liu, University of Cincinnati
iii
iv
Preface
These online lecture notes (in the form of an e-book) are intended to serve as
an introduction to the finite element method (FEM) for undergraduate students or
other readers who have no previous experience with this computational method.
The notes cover the basic concepts in the FEM using the simplest mechanics
problems as examples, and lead to the discussions and applications of the 1-D bar
and beam, 2-D plane and 3-D solid elements in the analyses of structural stresses,
vibrations and dynamics. The proper usage of the FEM, as a popular numerical
tool in engineering, is emphasized throughout the notes.
This online document is based on the lecture notes developed by the author
since 1997 for the undergraduate course on the FEM in the mechanical engineering
department at the University of Cincinnati. Since this is an e-book, the author
suggests that the readers keep it that way and view it either online or offline on
his/her computer. The contents and styles of these notes will definitely change
from time to time, and therefore hard copies may become obsolete immediately
after they are printed. Readers are welcome to contact the author for any
suggestions on improving this e-book and to report any mistakes in the
presentations of the subjects or typographical errors. The ultimate goal of this ebook on the FEM is to make it readily available for students, researchers and
engineers, worldwide, to help them learn subjects in the FEM and eventually solve
their own design and analysis problems using the FEM.
The author thanks his former undergraduate and graduate students for their
suggestions on the earlier versions of these lecture notes and for their contributions
to many of the examples used in the current version of the notes.
Yijun Liu
Cincinnati, Ohio, USA
December 2002
Chapter 1. Introduction
Chapter 1. Introduction
I. Basic Concepts
The finite element method (FEM), or finite element analysis
(FEA), is based on the idea of building a complicated object with
simple blocks, or, dividing a complicated object into small and
manageable pieces. Application of this simple idea can be found
everywhere in everyday life, as well as in engineering.
Examples:
Lego (kids play)
Buildings
Approximation of the area of a circle:
Element Si
i
R
2
Area of one triangle: S i = 2 R sin i
2
2
Area of the circle: S N = Si = 2 R N sin N R as N
i =1
where N = total number of triangles (elements).
Observation: Complicated or smooth objects can be
represented by geometrically simple pieces (elements).
Chapter 1. Introduction
Examples:
...
1997-2002 Yijun Liu, University of Cincinnati
Chapter 1. Introduction
Chapter 1. Introduction
Example:
Chapter 1. Introduction
Computer Implementations
Preprocessing (build FE model, loads and constraints)
FEA solver (assemble and solve the system of equations)
Postprocessing (sort and display the results)
Chapter 1. Introduction
Chapter 1. Introduction
(1)
(2)
where
a11
a
21
A = aij =
...
a
n1
x1
x
2
x = { xi } =
:
xn
a12
a22
...
[ ]
an 2
... a1n
... a2 n
... ...
... ann
b1
b
2
b = {bi } =
:
bn
(3)
Chapter 1. Introduction
w1
w = w2
w
3
v3 ]
v2
with
D = AB
with
d ij = a ij bij
Scalar Multiplication
A = [a ij ]
Matrix Multiplication
For two matrices A (of size lm) and B (of size mn), the
product of AB is defined by
C = AB
Chapter 1. Introduction
Transpose of a Matrix
If A = [aij], then the transpose of A is
[ ]
A T = a ji
Notice that ( AB ) T = B T A T .
Symmetric Matrix
A square (nn) matrix A is called symmetric, if
A = AT
Unit (Identity) Matrix
1 0
0 1
I=
... ...
0 0
or
a ij = a ji
... 0
... 0
... ...
... 1
Note that AI = A, Ix = x.
Determinant of a Matrix
The determinant of square matrix A is a scalar number
denoted by det A or |A|. For 22 and 33 matrices, their
determinants are given by
a b
det
= ad bc
c d
and
a11 a12
det a21 a22
a31 a32
Chapter 1. Introduction
a13
a23 = a11a22 a33 + a12 a23a31 + a21a32 a13
a33
a13a22 a31 a12 a21a33 a23a32 a11
Singular Matrix
A square matrix A is singular if det A = 0, which indicates
problems in the systems (nonunique solutions, degeneracy, etc.)
Matrix Inversion
For a square and nonsingular matrix A (det A 0 ), its
inverse A-1 is constructed in such a way that
AA 1 = A 1 A = I
The cofactor matrix C of matrix A is defined by
Cij = ( 1)i + j Mij
where Mij is the determinant of the smaller matrix obtained by
eliminating the ith row and jth column of A.
Thus, the inverse of A can be determined by
A 1 =
1
CT
det A
10
Chapter 1. Introduction
Examples:
a b
(1)
c d
1
d b
( ad bc) c a
Checking,
1
1
d b a b 1 0
a b a b
=
c d c d ( ad bc) c a c d = 0 1
1 1 0
(2) 1 2 1
0 1 2
3 2 1
3 2 1
1
2 2 1 = 2 2 1
=
(4 2 1)
1 1 1
1 1 1
Checking,
1 1 0 3 2 1 1 0 0
1 2 1 2 2 1 = 0 1 0
0 1 2 1 1 1 0 0 1
If det A = 0 (i.e., A is singular), then A-1 does not exist!
The solution of the linear system of equations (Eq.(1)) can be
expressed as (assuming the coefficient matrix A is nonsingular)
x = A 1b
Thus, the main task in solving a linear system of equations is to
found the inverse of the coefficient matrix.
1997-2002 Yijun Liu, University of Cincinnati
11
Chapter 1. Introduction
A( t ) = a ij ( t )
12
Chapter 1. Introduction
13
Chapter 1. Introduction
important is simple.
i
fi
ui
j
uj
fj
Two nodes:
i, j
Nodal displacements:
Nodal forces:
with = u j ui
Linear
Nonlinear
F
k
14
Chapter 1. Introduction
k ui f i
=
k u j f j
or,
ku = f
where
k = (element) stiffness matrix
u = (element nodal) displacement vector
f = (element nodal) force vector
Note that k is symmetric. Is k singular or nonsingular? That is,
can we solve the equation? If not, why?
15
Chapter 1. Introduction
Spring System
k1
k2
u1, F1
u2, F2
u3, F3
For element 1,
k1
k
1
k1 u1 f 11
=
k1 u2 f 21
element 2,
k2
k
2
k 2 u2 f 12
=
k 2 u3 f 22
16
Chapter 1. Introduction
That is,
F1 = k1u1 k1u2
F2 = k1u1 + ( k1 + k 2 )u2 k 2 u3
F3 = k 2 u2 + k 2 u3
In matrix form,
k1
k1
k k + k
1 1 2
k2
0
0 u1
k2 u2 =
k2 u3
F1
F2
F
3
or
KU = F
K is the stiffness matrix (structure matrix) for the spring system.
k1
k1
0
0
0
0 k
2
0 k 2
0 u1 f 11
0 u2 = f 21
0 u3 0
0 u1 0
k 2 u2 = f 12
k 2 u3 f 22
17
Chapter 1. Introduction
k1
k
1
0
0 u1 f 11
k 2 u2 = f 21 + f 12
k 2 u3 f 22
u1 = 0
and
F2 = F3 = P
we have
k1
k
1
0
k1
k1 + k 2
k2
0 0 F1
k 2 u2 = P
k 2 u3 P
which reduces to
k1 + k 2
k
k 2 u2 P
=
k 2 u3 P
and
F1 = k1u2
Unknowns are
u2
U=
u3
18
Chapter 1. Introduction
u
P
k
P
k
2
/
+
/
3
1
2
and the reaction force
F1 = 2 P
19
Chapter 1. Introduction
Example 1.1
k1
1
Given:
k2
2
k3
Find:
Solution:
(a) The element stiffness matrices are
100 100
k1 =
100 100
(N/mm)
(1)
200 200
k2 =
(N/mm)
200 200
(2)
100 100
k3 =
100 100
(3)
(N/mm)
20
Chapter 1. Introduction
u2
u3
u4
100
0
0
100
100 100 + 200
200
0
K=
200
200 + 100 100
0
0
0
100
100
or
0
0
100 100
100 300 200
0
K=
200 300 100
0
0
0
100 100
(4)
21
300 200 u2 0
200 300 u = P
Chapter 1. Introduction
(5)
(6)
(c) From the 1st and 4th equations in (4), we get the reaction forces
F1 = 100u 2 = 200 (N)
F4 = 100u 3 = 300 (N )
j j
Here i = 2, j = 3 for element 2. Thus we can calculate the spring
force as
u
F = f j = f i = [ 200 200] 2
u3
2
= [ 200 200]
3
= 200 (N)
Check the results!
1997-2002 Yijun Liu, University of Cincinnati
22
Chapter 1. Introduction
Example 1.2
k4
k1
2
F1
k2
F2
k3
Node i (1)
Node j (2)
1
2
3
4
4
2
3
2
2
3
5
1
23
Chapter 1. Introduction
u2
u2
k1
k1
k2
k2 =
k 2
u3
k2
k 2
u2
u5
k3
k 3
k4
k4 =
k 4
u1
k4
k 4
u2
u3
u4
k4
k1 + k 2 + k 4
k2
k1
0
k2
k2 + k3
0
k1
0
k3
0
k1
0
u5
0
0
k3
0
k 3
24
25
ui
fi
A,E
fj
L
L
length
cross-sectional area
elastic modulus
u = u( x )
displacement
= ( x)
strain
= ( x)
stress
Strain-displacement relation:
=
du
dx
(1)
Stress-strain relation:
= E
(2)
26
x
x
ui + u j
L
L
(3)
we have
=
u j ui
L
= E =
( = elongation)
E
L
(4)
(5)
We also have
=
F
A
(F = force in bar)
(6)
EA
= k
L
(7)
EA
is the stiffness of the bar.
L
27
k
k=
k
EA EA
k L
L
=
k EA EA
L
L
or
k=
EA 1 1
L 1 1
(8)
(9)
28
N j ( ) =
(10)
where
=
x
,
L
0 1
(11)
u = Ni
(12)
du d
= Nu = Bu
dx dx
(13)
d
N i ( )
dx
N j ( ) =
B = [ 1 / L 1 / L]
d
N i ( )
d
N j ( )
d
dx
(14)
29
(15)
1
1
T dV =
2
2
(u
B T EBu)dV
1 T
T
= u ( B EB )dV u
2
(16)
1
1
1
f i ui + f j u j = u T f
2
2
2
(17)
(18)
which gives
1 T
1
T
u ( B EB )dV u = u T f
2
2
T
(B EB )dV u = f
30
or
ku = f
(19)
where
k=
(B
EB )dV
(20)
k=
1 / L
EA 1 1
E [ 1 / L 1 / L] Adx =
1
/
L
L 1 1
(21)
31
Example 2.1
1 2A,E
1
2 A,E
2
u2
2 EA 1 1
k1 =
L 1 1
Element 2,
u2
u3
EA 1 1
k2 =
L 1 1
Imagine a frictionless pin at node 2, which connects the two
elements. We can assemble the global FE equation as follows,
32
2 2 0 u1 F1
EA
2 3 1 u2 = F2
L
0 1 1 u3 F3
Load and boundary conditions (BC) are,
u1 = u3 = 0,
F2 = P
FE equation becomes,
2 2 0 0 F1
EA
2 3 1 u2 = P
L
0 1 1 0 F3
Deleting the 1st row and column, and the 3rd row and column, we
obtain,
EA
3]{u2 } = { P}
[
L
Thus,
u2 =
PL
3EA
and
u1
0
PL
u2 =
1
u 3EA 0
3
Stress in element 1 is
33
u
1 = E 1 = EB 1u1 = E[ 1 / L 1 / L] 1
u2
u u
E PL
P
= E 2 1 =
0 =
L
L 3EA 3 A
Similarly, stress in element 2 is
u2
2 = E 2 = EB 2 u 2 = E [ 1 / L 1 / L]
u3
u u
E
PL
P
= E 3 2 = 0
=
3EA
3A
L
L
which indicates that bar 2 is in compression.
Check the results!
Notes:
In this case, the calculated stresses in elements 1 and 2
are exact within the linear theory for 1-D bar structures.
It will not help if we further divide element 1 or 2 into
smaller finite elements.
For tapered bars, averaged values of the cross-sectional
areas should be used for the elements.
We need to find the displacements first in order to find
the stresses, since we are using the displacement based
FEM.
34
Example 2.2
1
1
A,E
2
2
3
E = 2.0 104 N / mm 2 ,
L = 150 mm, =1.2 mm
Solution:
We first check to see if or not the contact of the bar with
the wall on the right will occur. To do this, we imagine the wall
on the right is removed and calculate the displacement at the
right end,
PL (6.0 104 )(150)
0 =
=
= 18
. mm > = 12
. mm
EA (2.0 104 )(250)
Thus, contact occurs.
The global FE equation is found to be,
35
1 1 0 u1 F1
EA
1 2 1 u2 = F2
L
0 1 1 u3 F3
The load and boundary conditions are,
F2 = P = 6.0 104 N
u1 = 0,
u3 = = 1.2 mm
FE equation becomes,
1 1 0 0 F1
EA
1 2 1 u2 = P
L
0 1 1 F3
The 2nd equation gives,
u2
EA
2
1
[
] = { P}
L
that is,
EA
EA
[ 2]{u2 } = P +
L
L
2 EA
and
36
u1 0
. ( mm)
u2 = 15
u 12
3 .
To calculate the support reaction forces, we apply the 1st
and 3rd equations in the global FE equation.
The 1st equation gives,
u1
EA
EA
F1 =
[ 1 1 0] u 2 = ( u 2 ) = 5.0 10 4 N
L
u L
3
and the 3rd equation gives,
u1
EA
EA
F3 =
0 1 1]u2 =
( u2 + u3 )
[
L
L
u
3
= 10
. 10 4 N
Check the results.!
37
Distributed Load
q
i
qL/2
qL/2
Wq =
qL
1
1
uqdx =
u( )q ( Ld ) =
u( ) d
2
2
2
qL
=
2
qL
=
2
[ N ( )
i
ui
N j ( ) d
u j
ui
d
[
] u
j
0
1 qL
=
2 2
qL ui
2 u j
qL / 2
uj
qL / 2
1
u
2 i
38
that is,
qL / 2
with f q =
qL
/
2
1
Wq = u T f q
2
(22)
(23)
which yields
ku = f + f q
(24)
f
qL
/
2
+
j
(25)
In an assembly of bars,
q
1
qL/2
1
qL
2
qL/2
3
39
ui
i
ui
vi
X
Local
Global
x, y
X, Y
ui' , v i'
ui , v i
1 dof at a node
2 dofs at a node
40
In matrix form,
m ui
ui' l
' =
v
m
l
v
i
i
(26)
or,
~
u i' = Tu i
where the transformation matrix
m
~ l
T=
m l
~
~
is orthogonal, that is, T 1 = T T .
(27)
(28)
~
T 0
with T =
~
0
T
(29)
or,
u = Tu
'
(30)
41
EA 0
L 1
0
0 1
0 0
0 1
0 0
0 ui' f i '
0 vi' 0
' =
0 u j f j'
0 v 'j 0
or,
k 'u' = f '
Using transformations given in (29) and (30), we obtain
k ' Tu = Tf
Multiplying both sides by TT and noticing that TTT = I, we
obtain
T T k ' Tu = f
(31)
(32)
42
Explicit form,
ui
vi
uj
vj
l2
lm l 2 lm
m2 lm m2
EA lm
k=
L l 2 lm l 2
lm
2
2
lm
m
lm
m
(33)
X j Xi
L
m = sin =
Yj Yi
L
(34)
ui
1 l m 0 0 vi
L 0 0 l m u j
v j
That is,
ui
v
E
i
= [ l m l m]
L
u j
v j
(35)
43
Example 2.3
A simple plane truss is made
of two identical bars (with E, A, and
L), and loaded as shown in the
figure. Find
1) displacement of node 2;
3
45o
2
P2
2
P1
Solution:
1
L 1 1
'
1
2
2
44
u1
v1
u2
v2
1 1 1
1
1
1
1
EA
k 1 = T1T k 1' T1 =
1
2 L 1 1 1
1 1 1
1
Element 2:
= 135o , l =
2
2
, m=
2
2
We have,
u2
v2
u3
v3
1 1 1 1
1 1
1
1
EA
k 2 = T2T k '2 T2 =
1 1
2 L 1 1
1 1 1 1
v1
u2
v2
u3
v3
1 1 1 0
0 u1 F1 X
1
1
1 1 1 0
0 v1 F1Y
0 1 1 u2 F2 X
EA 1 1 2
=
2
1 1 v2 F2Y
2 L 1 1 0
0
0 1 1
1 1 u3 F3 X
v
F
0
0
1
1
1
3 3Y
45
F2 X = P1 , F2Y = P2
Condensed FE equation,
EA 2 0 u2 P1
=
0
2
2L
v2 P2
Solving this, we obtain the displacement of node 2,
u2 L P1
=
v2 EA P2
Using formula (35), we calculate the stresses in the two bars,
0
E 2
L 0
2
1 =
[ 1 1 1 1] P = ( P1 + P2 )
L 2
EA 1 2 A
P2
P1
E 2
L P2
2
2 =
[1 1 1 1] 0 = ( P1 P2 )
L 2
EA 2 A
0
Check the results:
Look for the equilibrium conditions, symmetry,
antisymmetry, etc.
1997-2002 Yijun Liu, University of Cincinnati
46
Example 2.4
(Multipoint Constraint)
y
3
2
L
1 Y
3
45o
L = 1m,
A = 6.0 10 4 m2
E = 210 GPa ,
A = 6 2 10 4 m2 for element 3.
Determine the displacements and reaction forces.
Solution:
We have an inclined roller at node 3, which needs special
attention in the FE solution. We first assemble the global FE
equation for the truss.
Element 1:
= 90 o , l = 0,
m=1
47
u1
v1 u2 v2
0 0
0 0
0 1
( N / m)
0 0
0 1
Element 2:
= 0o ,
l = 1, m = 0
u2
v2 u3 v3
0 1
0 0
0 1
0 0
0
0
( N / m)
0
0
v1
u3
Element 3:
= 45o , l =
1
1
, m=
2
2
u1
v3
k3 =
0.5
0.5 0.5 0.5
2
( N / m)
48
15
. 0 1 0.5 0.5 v1 F1Y
1 0
0 u2 F2 X
1
5
1260 10
v = F
1
0
0
2 2Y
15
0.5 u3 F3 X
.
0
5
Sym.
.
v
F
3 3Y
2 u3
2
=
( u3 + v3 ) = 0,
2 v3 2
that is,
u3 v3 = 0
This is a multipoint constraint (MPC).
Similarly, we have a relation for the force at node 3,
2
F3 x ' =
2
2 F3 X
2
( F3 X + F3Y ) = 0,
=
2 F3Y 2
that is,
49
F3 X + F3Y = 0
Applying the load and BCs in the structure FE equation by
deleting 1st, 2nd and 4th rows and columns, we have
1 1 0 u2 P
1260 105 1 15
. 0.5 u3 = F3 X
1260 105 1 15
. 0.5 u3 = F3 X
0 0.5 0.5 u3 F3 X
which is
P
1 1
u
1260 105 1 2 2 = F3 X
u3
F
1
0
3X
The 3rd equation yields,
F3 X = 1260 105 u3
Substituting this into the 2nd equation and rearranging, we have
1 1 u2 P
1260 105
=
1 3 u3 0
Solving this, we obtain the displacements,
1997-2002 Yijun Liu, University of Cincinnati
50
1
u2
3 P 0.01191
=
=
( m)
5
u
P
0
003968
.
3 2520 10
2
1Y
5
0
0 u3 = 0.0 (kN)
F2Y = 1260 10 0
v 500
1 15
F
0
5
.
.
3X
3
500
F3Y
0
0.5
0.5
Aj u j = 0
j
51
3-D Case
y
Y
i
X
Z
Local
Global
x, y, z
X, Y, Z
ui , vi , wi
1 dof at node
3 dofs at node
52
vj, Fj
j
E,I
L
j, Mj
dv
dx
F = F ( x)
M = M ( x)
(36)
My
I
(37)
53
Direct Method
Using the results from elementary beam theory to compute
each column of the stiffness matrix.
vj
6 L 12 6 L vi Fi
12
2
2
EI 6 L 4 L 6 L 2 L i M i
=
L3 12 6 L 12 6 L v j Fj
6 L 2 L2 6 L 4 L2 M
j j
(38)
54
Formal Approach
Apply the formula,
L
k=
B T EIBdx
(39)
(40)
N 3 ( x ) = 3x / L 2 x / L
2
N 4 ( x ) = x 2 / L + x 3 / L2
Then, we can represent the deflection as,
v ( x ) = Nu
= [ N1 ( x)
N 2 ( x)
N 3 ( x)
vi
i
N 4 ( x )]
v j
j
(41)
55
(42)
+
L
L L L2 L3
L L2
L
(43)
1
1
T dV =
2
2
1
=
2
0 A
My 1 My
dAdx
I E I
L
1 d 2v
d 2v
T 1
M
Mdx =
EI 2 dx
dx
EI
2 dx 2
1
(Bu) T EI (Bu)dx
=
2
1 T
T
= u B EIBdx u
2
k=
B T EIBdx
56
12 EI
0
L3
6 EI
0
2
L
k=
EA
0
L
12 EI
3
0
L
6 EI
0
L2
0
6 EI
L2
4 EI
L
0
6 EI
L2
2 EI
L
EA
L
0
0
EA
L
0
0
0
12 EI
L3
6 EI
2
L
0
12 EI
L3
6 EI
2
L
0
6 EI
L2
2 EI
L
0
6 EI
2
L
4 EI
L
57
Example 2.5
Y
1
1
E,I
2
3
Given:
Find:
v2
6 L 12 6 L
12
2
2
EI 6 L 4 L 6 L 2 L
k1 = 3
L 12 6 L 12 6 L
6 L 2 L2 6 L 4 L2
v2
v3
6 L 12 6 L
12
2
2
EI 6 L 4 L 6 L 2 L
k2 = 3
L 12 6 L 12 6 L
6 L 2 L2 6 L 4 L2
58
v2
v3
6 L 12 6 L
0
0 v1 F1Y
12
6 L 4 L2 6 L 2 L2
0
0 1 M 1
0
12 6 L v2 F2Y
EI 12 6 L 24
=
0
8 L2 6 L 2 L2 2 M 2
L3 6 L 2 L2
0
0
12 6 L 12 6 L v3 F3Y
0
6L
2 L2 6 L 4 L2 3 M 3
0
Loads and constraints (BCs) are,
F2Y = P ,
M2 = M ,
v1 = v3 = 1 = 3 = 0
Reduced FE equation,
EI
L3
24 0 v2 P
0 8 L2 = M
2 24 EI 3 M
59
12 6 L
F1Y
2
M
1 EI 6 L 2 L v2
=
= 3
F
12
6
L
L
2
3Y
6L
M 3
2 L2
2 P + 3 M / L
1 PL + M
4 2 P 3 M / L
PL + M
Stresses in the beam at the two ends can be calculated using the
formula,
= x =
My
I
60
qL/2
qL/2
qL2/12
j
qL2/12
L
qL/2
qL
qL2/12
L
61
Example 2.6
y
E,I
L
Given:
Find:
f
m
E,I
L
where
f = pL / 2,
m = pL2 / 12
62
6 L 12 6 L v1 F1Y
12
2
2
EI 6 L 4 L 6 L 2 L 1 M 1
=
L3 12 6 L 12 6 L v2 F2 Y
6 L 2 L2 6 L 4 L2 M
2 2
M2 = m
12 6 L v2 f
6 L 4 L2 = m
=
3
6
EI
/
+
3
6
6
Lf
m
pL
EI
2
(A)
These nodal values are the same as the exact solution. Note
that the deflection v(x) (for 0 < x< 0) in the beam by the FEM is,
however, different from that by the exact solution. The exact
solution by the simple beam theory is a 4th order polynomial of
x, while the FE solution of v is only a 3rd order polynomial of x.
If the equivalent moment m is ignored, we have,
2
4
v 2
L 2 L f pL / 6 EI
=
3
2 6 EI 3 Lf pL / 4 EI
(B)
The errors in (B) will decrease if more elements are used. The
1997-2002 Yijun Liu, University of Cincinnati
63
2
1
where the result in (A) is used. This force vector gives the total
effective nodal forces which include the equivalent nodal forces
for the distributed lateral load p given by,
pL / 2
2
pL / 12
The correct reaction forces can be obtained as follows,
F1Y pL / 2 pL / 2 pL
=
= 2
2
2
M
pL
pL
5
/
12
/
12
1
pL / 2
Check the results!
64
Example 2.7
Y
P
E,I
1
2
1
L
Given:
2
3
L
Find:
Solution:
The beam has a roller (or hinge) support at node 2 and a
spring support at node 3. We use two beam elements and one
spring element to solve this problem.
The spring stiffness matrix is given by,
v3
k
ks =
k
v4
k
k
65
v1
v2
v3
v4
0
0
0 v1 F1Y
12 6 L 12 6 L
4 L2 6 L 2 L2
0
0
0 1 M 1
24
0
6L
0 v2 F2Y
12
EI
2
2
8L
0 2 = M 2
6L 2 L
L3
2
4
0
L
3 M 3
Symmetry
k ' v4 F4Y
in which
L3
k
k'=
EI
is used to simply the notation.
We now apply the boundary conditions,
v1 = 1 = v2 = v4 = 0,
M 2 = M 3 = 0,
F3Y = P
Deleting the first three and seventh equations (rows and
columns), we have the following reduced equation,
6 L 2 L2 2 0
8 L2
EI
L
k
L
6
12
6
'
v 3 = P
3
L
2 L2
6 L 4 L2 3 0
Solving this equation, we obtain the deflection and rotations at
node 2 and node 3,
66
2
3
2
PL
v3 =
7 L
EI (12 + 7 k ')
3
9
The influence of the spring k is easily seen from this result.
Plugging in the given numbers, we can calculate
2 0.002492 rad
v3 = 0.01744 m
0.007475 rad
3
1
=
.
kN
116
2
F
2
Y
F4Y 3.488 kN
Checking the results: Draw free body diagram of the beam
69.78 kN
1
69.78 kNm
50 kN
2
116.2 kN
3
3.488 kN
67
3000 lb
E, I, A
500 lb/ft
2
3
8 ft
X
12 ft
Given:
Find:
Solution:
For this example, we first convert the distributed load to its
equivalent nodal loads.
68
3000 lb
3000 lb
72000 lb-in.
3000 lb
72000 lb-in.
vi
EA
0
L
12 EI
0
3
L
6 EI
0
L2
k=
EA
0
L
12 EI
3
0
L
6 EI
0
L2
i
0
6 EI
L2
4 EI
L
0
6 EI
L2
2 EI
L
uj
EA
L
0
0
EA
L
0
0
vj
0
12 EI
L3
6 EI
2
L
0
12 EI
L3
6 EI
2
L
j
0
6 EI
L2
2 EI
L
0
6 EI
2
L
4 EI
L
69
Node i (1)
Node j (2)
1
2
3
1
3
4
2
1
2
v1
u2
v2
.
0
0
0
0
1417
141.7
0
0.784
56.4
0
0.784 56.4
56.4
5417
0
56.4 2708
0
4
k 1 = k 1 ' = 10
.
.
1417
0
0
1417
0
0
0
0
0.784 56.4
0.784 56.4
56.4
2708
0
56.4 5417
0
For elements 2 and 3, the stiffness matrix in local system is,
ui '
vi '
i '
uj '
vj'
j'
0
0
212.5
0
0
212.5
0
2.65
127
0
2.65 127
127
8125
0
127 4063
0
k 2 ' = k 3 ' = 104
212
5
0
0
212
5
0
0
.
.
0
2.65 127
0
2.65 127
0
127
4063
0
127
8125
where i=3, j=1 for element 2 and i=4, j=2 for element 3.
1997-2002 Yijun Liu, University of Cincinnati
70
0 0 1 0 0 0
T=
0
0
0
l
m
0
0 0 0 m l 0
0
0
0
0
0
1
We have
l = 0, m = 1
for both elements 2 and 3. Thus,
0
1
0
T=
0
0
1
0
0
0
0
0
0 0
0 0
1 0
0 0
0 1
0 0
0
0
0
1
0
0
0
0
0
0
0
71
u3
v3
u1
v1
0
127 2.65
0
127
2.65
0
212.5
0
0
212.5
0
0
8125 127
0
4063
127
k 2 = 10 4
.
.
2
65
0
127
2
65
0
127
0
212.5
0
0
212.5
0
0
4063 127
0
8125
127
and
u4
v4
u2
v2
127 2.65
127
0
0
2.65
0
212.5
0
0
212.5
0
0
8125 127
0
4063
127
k 3 = 104
.
.
2
65
0
127
2
65
0
127
0
212.5
0
0
212.5
0
0
4063 127
0
8125
127
Assembling the global FE equation and noticing the following
boundary conditions,
u3 = v3 = 3 = u4 = v4 = 4 = 0
F1 X = 3000 lb, F2 X = 0, F1Y = F2Y = 3000 lb,
M1 = 72000 lb in. , M 2 = 72000 lb in.
we obtain the condensed FE equation,
72
.
0
127 1417
0
0 u1
144.3
0
213.3
56.4
0
0.784 56.4 v1
56.4
13542
0
56.4 2708 1
127
4
10
u
.
.
1417
0
0
144
3
0
127
0
0
213.3 56.4 v2
0.784 56.4
56.4
2708
127
56.4 13542 2
0
3000
3000
72000
=
3000
72000
Solving this, we get
0.092 in.
u1
v 0.00104 in.
1
1 0.00139 rad
=
u
0
.
0901
in.
2
v2 0.0018 in.
. 10 5 rad
2 388
To calculate the reaction forces and moments at the two ends,
we employ the element FE equations for element 2 and element
3. We obtain,
73
F3 X 672.7 lb
F3Y = 2210 lb
M 60364 lb in.
3
and
F4 X 2338 lb
F4Y = 3825 lb
M 112641 lb in.
4
3000 lb
72000 lb-in.
3000 lb
72000 lb-in.
112641 lb-in.
60364 lb-in.
2338 lb
672.7 lb
2210 lb
3825 lb
74
for stresses,
x , y , z , xy , yz , zx
for strains.
and
y
yz
xy
zx
z
z
75
( z 0)
(1)
Plane strain:
z = yz = zx = 0
( z 0)
(2)
76
0 x x 0
/ E
1/ E
0 y + y 0
0
1 / G xy xy 0
(3)
or,
= E 1 + 0
where 0 is the initial strain, E the Youngs modulus, the
Poissons ratio and G the shear modulus. Note that,
G=
E
2(1 + )
(4)
=
y
2
1
xy
0
1
x x 0
y y 0
0 0 (1 ) / 2 xy xy 0
(5)
or,
= E + 0
where 0 = E 0 is the initial stress.
1997-2002 Yijun Liu, University of Cincinnati
77
The above relations are valid for plane stress case. For
plane strain case, we need to replace the material constants in
the above equations in the following fashion,
E
1 2
1
GG
E
(6)
1
E
1
y =
(1 + )(1 2 ) 0
0
xy
0
x x 0
0
y y 0
(1 2 ) / 2 xy xy 0
y 0 = T
0
xy 0
(7)
78
u v
v
u
, y = , xy =
+
y x
y
x
In matrix form,
x / x
0
u , or = Du
=
0
/
y
y
v
/ y / x
xy
(8)
(9)
where fx and fy are body forces (such as gravity forces) per unit
volume. In FEM, these equilibrium conditions are satisfied in
an approximate sense.
79
Boundary Conditions
ty
tx
y
St
x
Su
on Su
on S t
(10)
80
Example 3.1
A plate is supported and loaded with distributed force p as
shown in the figure. The material constants are E and .
y
p
x
The exact solution for this simple problem can be found
easily as follows,
Displacement:
p
y
E
p
x,
E
v =
p
,
E
y =
x = p,
y = 0,
u=
Strain:
x =
p
,
E
xy = 0
Stress:
xy = 0
81
u N 1
=
v 0
0
N1
N2
0
0
N2
u1
v
1
L
u2
L
v
2
M
or
u = Nd (11)
or
= Bd
(12)
82
1
1
T dV =
2
2
1
2
x + y y + xy xy ) dV
T
E ) dV =
(
1 T
E dV
2
1
= d T B T EB dV d
2
1
= d T kd
2
From this, we obtain the general formula for the element
stiffness matrix,
k=
B T EB dV
(13)
Note that unlike the 1-D cases, E here is a matrix which is given
by the stress-strain relation (e.g., Eq.(5) for plane stress).
The stiffness matrix k defined by (13) is symmetric since E
is symmetric. Also note that given the material property, the
behavior of k depends on the B matrix only, which in turn on the
shape functions. Thus, the quality of finite elements in
representing the behavior of a structure is entirely determined by
the choice of shape functions.
Most commonly employed 2-D elements are linear or
quadratic triangles and quadrilaterals.
83
u3
v
v1
1
(x1, y1)
v2
u
(x, y)
2
(x2, y2)
u1
u2
(14)
y = b6 ,
xy = b3 + b5
(15)
84
u N1
=
v 0
0
N1
N2
0
0
N2
N3
0
u1
v
1
0 u2
N 3 v 2
u3
v3
(16)
(17)
and
1997-2002 Yijun Liu, University of Cincinnati
85
1 x1
1
A = det 1 x 2
2
1 x3
y1
y2
y3
(18)
x
y 23
1
Bd
0
=
=
y
2
A
x32
xy
0
x32
y 23
y31
0
x13
0
x13
y 31
y12
0
x 21
u1
v
0 1
u2
x 21 (19)
v2
y12
u3
v3
B T EB dV = tA( B T EB )
(20)
86
=0
=b
=1
=1
(a, b)
2
(21)
Notice that,
N1 + N 2 + N 3 = 1
(22)
at node i;
at the other nodes
(23)
and varies linearly within the element. The plot for shape
function N1 is shown in the following figure. N2 and N3 have
similar features.
87
=0
N1
=1
1
2
1
(24)
or,
x = x13 + x 23 + x3
y = y13 + y 23 + y3
(25)
y u
u
x
x
J
=
u
y u
y
y
(26)
88
y13
,
y 23
J 1 =
1 y 23
2 A x 23
y13
x13
(27)
2 A x 23
y
u
y13
x13 u
1 y 23
=
2 A x 23
(28)
y13 u1 u3
x13 u2 u3
Similarly,
v
x 1 y 23
v =
x
2
A
23
y
y13 v1 v3
x13 v 2 v3
(29)
0
x32
y 23
y31
0
x13
0
x13
y31
y12
0
x 21
0
x 21
y12
(30)
89
90
3
y
v5
5
v6
u6
u5
v2
v1
1
u1
u4
u2
v4
x
(31)
91
(33)
N 4 = 4
N 5 = 4
N 6 = 4
in which = 1 . Each of these six shape functions
represents a quadratic form on the element as shown in the
figure.
=0
=1/2
6
=1
N1
4
u = N i ui ,
i =1
v = N i vi
(34)
i =1
92
v3
v4
=1
u3
v1
1
= 1
u4
v2
u2
u1
= 1
=1
u = N i ui ,
i =1
v = N i vi
(36)
i =1
93
=1
8
y
= 1
= 1
=1
(37)
94
1
N 5 = (1 )(1 2 )
2
1
N 6 = (1 + )(1 2 )
2
1
N 7 = (1 + )(1 2 )
2
1
N 8 = (1 )(1 2 )
2
Again, we have
u = N i ui ,
i =1
v = N i vi
(38)
i =1
95
Example 3.2
A square plate with a hole at the center and under pressure
in one direction.
y
p
A
FE Analysis:
From the knowledge of stress concentrations, we should
expect the maximum stresses occur at points A and B on the
edge of the hole. Value of this stress should be around 3p (=
300 psi) which is the exact solution for an infinitely large plate
with a hole.
96
No. Elem.
DOF
Max. (psi)
T6
966
4056
310.1
Q4
493
1082
286.0
Q8
493
3150
327.1
...
...
...
...
Q8
2727
16,826
322.3
Discussions:
Check the deformed shape of the plate
Check convergence (use a finer mesh, if possible)
Less elements (~ 100) should be enough to achieve the
same accuracy with a better or smarter mesh
Well redo this example in next chapter employing the
symmetry conditions.
1997-2002 Yijun Liu, University of Cincinnati
97
98
Transformation of Loads
Concentrated load (point forces), surface traction (pressure
loads) and body force (weight) are the main types of loads
applied to a structure. Both traction and body forces need to be
converted to nodal forces in the FEA, since they cannot be
applied to the FE model directly. The conversions of these loads
are based on the same idea (the equivalent-work concept) which
we have used for the cases of bar and beam elements.
qB
fB
qA
fA
s
B
L
B
A
Traction on a Q4 element
Suppose, for example, we have a linearly varying traction q
on a Q4 element edge, as shown in the figure. The traction is
normal to the boundary. Using the local (tangential) coordinate
s, we can write the work done by the traction q as,
L
Wq = t un ( s )q( s )ds
0
99
un ( s ) = (1 s / L )unA + ( s / L )unB
The traction q(s), which is also linear, is given in a similar way,
q( s ) = (1 s / L )q A + ( s / L )qB
Thus, we have,
Wq = t [ unA
0
L
1 s / L
q A
unB ]
1
s
/
L
s
/
L
[
ds
]
s / L
q B
= [ unA
( s / L )(1 s / L ) q A
(1 s / L ) 2
unB ]t
ds q
2
(
/
)(
1
/
)
(
/
)
s
L
s
L
s
L
0
B
= [ unA
unB ]
tL 2 1 q A
6 1 2 q B
B
B
Note, for constant q, we have,
f A qtL 1
=
f
2
B
1
For quadratic elements (either triangular or quadrilateral),
the traction is converted to forces at three nodes along the edge,
instead of two nodes.
Traction tangent to the boundary, as well as body forces,
are converted to nodal forces in a similar way.
100
Stress Calculation
The stress in an element is determined by the following
relation,
x
x
y = E y = EBd
xy
xy
(39)
101
1
( 1 2 ) 2 + ( 2 3 ) 2 + ( 3 1 ) 2
2
(40)
1 =
P
2 =
x + y
2
x + y
2
x y
2
+
+ xy
2
2
x y
2
+ xy
2
(41)
(42)
Averaged Stresses:
Stresses are usually averaged at nodes in FEA software
packages to provide more accurate stress values. This option
should be turned off at nodes between two materials or other
geometry discontinuity locations where stress discontinuity does
exist.
102
Discussions
T6 and Q8:
103
104
Types of Symmetry:
Reflective (mirror, bilateral) symmetry
Rotational (cyclic) symmetry
Axisymmetry
Translational symmetry
...
Examples:
105
Examples:
Cautions:
In vibration and buckling analyses, symmetry concepts, in
general, should not be used in FE solutions (works fine in
modeling), since symmetric structures often have antisymmetric
vibration or buckling modes.
106
Physical Meaning:
A finite element model of a portion of structure.
Mathematical Meaning:
Boundary matrices which are load and stiffness matrices
reduced (condensed) from the interior points to the exterior or
boundary points.
107
Disadvantages:
Increased overhead for file management
Matrix condensation for dynamic problems introduce
new approximations
...
108
Iterative Methods:
Solution time is unknown beforehand
Reduced storage requirement
Suitable for large problems, or bulky structures (large
bandwidth, converge faster)
Need solving again for different load cases
109
2
0 3 3 x 3 3
or Ax = b .
Forward Elimination:
Form
(1) 8 2 0
( 2) 2 4 3
(3) 0 3 3
2
1 ;
(3) 0 3
3
(2) +
2
2 ;
14
(3) (3):
3
(1) 8 2
0
( 2) 0 14 12
(3) 0 0
2
2
2 ;
12
Back Substitution:
x 3 = 12 / 2 = 6
x 2 = ( 2 + 12 x 3 ) / 14 = 5
x1 = ( 2 + 2 x 2 ) / 8 = 1.5
or
1.5
x = 5 .
6
110
or
aij x j = bi ,
i = 1, 2, ..., N .
j =1
( k + 1)
i 1
N
( k + 1)
(k )
b
a
x
a
x
,
i
ij
j
ij
j
j =1
j = i +1
for i = 1, 2, ..., N .
1
=
a ii
In vector form,
1
x ( k +1) = A D b A L x ( k +1) A L x ( k ) ,
where
A D = aii is the diagonal matrix of A,
A L is the lower triangular matrix of A,
T
such that A = A D + A L + A L .
Iterations continue until solution x converges, i.e.
x ( k + 1) x ( k )
x
(k )
111
u = N u
=1
Stiffening Effect:
FE Model is stiffer than the real structure.
In general, displacement results are smaller in
magnitudes than the exact values.
112
Exact Solution
FEM Solutions
No. of DOFs
113
V. Numerical Error
Error Mistakes in FEM (modeling or solution).
Type of Errors:
Modeling Error (beam, plate theories)
Discretization Error (finite, piecewise )
Numerical Error ( in solving FE equations)
u2
P
1
k1
k2
FE Equations:
k1
k
1
and
k 1 u1 P
=
k 1 + k 2 u 2 0
Det K = k 1 k 2 .
114
u2
P
k1
u2 = u1
u2 =
k1
u1
k1 + k 2
u1
P/k1
u2 = u1
u2
P
k1
u2 =
k1
u1
k1 + k 2
P/k1
u1
115
Type of Refinements:
h-refinement: reduce the size of the element (h refers to the
typical size of the elements);
p-refinement:
r-refinement:
Examples:
116
U = U i ,
i =1
U = U i* ,
*
i =1
M
U E = U E i ,
i =1
Ui =
1 T 1
2 E dV ;
V
i
U* =
i
1 *T 1 *
2 E dV ;
V
i
UEi =
1 T 1
2 E E E dV ;
V
i
117
UE
.
=
U + U E
(0 1)
Examples:
118
Applications:
Shear walls
Floor panels
Shelves
119
My
q(x,y)
Mxy
Qy
t
Mx
Qx
Mid surface
Mxy
Stresses:
z
yz
y
xy
xz
x
xy
120
t/2
( N m / m)
(1)
t/2
( N m / m)
(2)
( N m / m)
(3)
( N / m)
(4)
( N / m)
(5)
M y = t / 2 y zdz ,
Twisting moment (per unit length):
t/2
M xy = t / 2 xy zdz ,
Shear Forces (per unit length):
t/2
Q x = t / 2 xz dz ,
t/2
Q y = t / 2 yz dz ,
Maximum bending stresses:
( x ) max =
6M x
,
t2
( y ) max =
6M y
t2
(6)
121
w
x
w = w( x, y ),
w
u = z
,
x
w
v = z
.
y
( deflection )
(7)
122
Strains:
2w
x = z 2 ,
x
2w
y = z 2 ,
y
xy
(8)
2w
= 2 z
.
xy
1
xy
0
x
1
,
1
0
y
0 0 (1 ) / 2 xy
or,
2w
2
0 x2
x
1
E
w .
1
0
=
z
y
2
y 2
1
0 0 (1 ) 2 w
xy
xy
(9)
123
Governing Equation:
D 4 w = q ( x , y ) ,
(10)
where
4
4
4
( 4 + 2 2 2 + 4 ),
x
x y y
4
Et 3
D=
(the bending rigidity of the plate),
2
12(1 )
q = lateral distributed load (force/area).
Compare the 1-D equation for straight beam:
d 4w
EI
= q( x ) .
4
dx
Note: Equation (10) represents the equilibrium condition in
the z-direction. To see this, refer to the previous figure showing
all the forces on a plate element. Summing the forces in the zdirection, we have,
Q x y + Q y x + qxy = 0,
which yields,
Q x Q y
+
+ q( x , y ) = 0 .
x
y
Substituting the following relations into the above equation, we
obtain Eq. (10).
1997-2002 Yijun Liu, University of Cincinnati
124
M x M xy
,
+
x
y
Qy =
2w
2w
M x = D 2 + 2 ,
y
x
M xy
x
M y
y
2w
2w
M y = D 2 + 2 .
x
y
Boundary Conditions:
Clamped:
w = 0,
w
= 0;
n
(11)
Simply supported:
w = 0,
M n = 0;
(12)
Free:
Q n = 0,
M n = 0;
(13)
boundary
1997-2002 Yijun Liu, University of Cincinnati
125
Examples:
A square plate with four edges clamped or hinged, and
under a uniform load q or a concentrated force P at the center C.
z
C
L
y
L
Simply supported
0.00126 qL4/D
0.00406 qL4/D
Under concentrated
force P
0.00560 PL2/D
0.0116 PL2/D
in which: D= Et3/(12(1-v2)).
These values can be used to verify the FEA solutions.
126
yz 0 .
This means that a line which is normal to the mid surface before
the deformation will not be so after the deformation.
w
y
x
w
w
x
127
New relations:
u = z y ,
x = z
v = z x ;
(14)
y = z x ,
y
xy = z ( y x ),
y
x
w
+ y ,
xz =
x
w
x.
yz =
y
(15)
w
+ y = 0,
x
yz =
w
x = 0,
y
then we can recover the relations applied in the thin plate theory.
Main variables: w( x, y ), x ( x, y ) and y ( x, y ) .
The governing equations and boundary conditions can be
established for thick plate based on the above assumptions.
128
3
x
1
w w
w1 , ,
x 1 y 1
2
t
w w
w2 , ,
x 2 y 2
w,
w w
,
.
y y
w( x, y ) = N i wi + N xi ( ) i + N yi ( ) i ,
x
y
i =1
4
129
8-Node Quadrilateral
z
y
3
x
1
w, x and y.
On each element:
n
w( x, y ) = N i wi ,
i =1
n
x ( x, y ) = N i xi ,
i =1
n
y ( x, y ) = N i yi .
i =1
130
3
6
4
1
w w
, , x , y ;
x y
w w
At each node: w, x =
.
, y =
131
Test Problem:
z
P
C
y
L
L
x
wc ( PL2/D)
22
0.00593
44
0.00598
88
0.00574
1616
0.00565
Exact Solution
0.00560
132
Example:
Sea shell, egg shell (the wonder of the nature);
Containers, pipes, tanks;
Car bodies;
Roofs, buildings (the Superdome), etc.
Forces in shells:
Membrane forces + Bending Moments
(cf. plates: bending only)
133
internal forces:
p
membrane stresses
dominate
Shell Theory:
Thin shell theory
Thick shell theory
Shell theories are the most complicated ones to formulate
and analyze in mechanics (Russians contributions).
Engineering Craftsmanship
Demand strong analytical skill
134
Shell Elements:
v
u
Q4 or Q8 shell element.
135
v
i
136
Test Cases:
q
L/2
F
A
R
80o
L/2
F
Pinched Cylinder
Roof
F2
b
A
A
L
F1
F
Pinched Hemisphere
137
x
z
y,v
yx
yz
xy
zy
x
z
zx
xz
x, u
z, w
138
= { }= z
xy
yz
zx
or
[ ]
ij
(1)
[ ]
(2)
Strains:
x
y
z
= { }=
xy
yz
zx
or
ij
Stress-strain relation:
x
y
z
xy
yz
zx
0
0
0
v
v
1 v
v
1 v
0
0
0 x
v
1
0
0
0
v
v
v
1 2v
0
E
0
0
0
0 z
=
2
(1 + v )(1 2v )
xy
1 2v
0
0
0
0 yz
0
2
1
2
v
0
zx
0
0
0
0
or
= E
(3)
139
Displacement:
u ( x, y , z ) u1
u = v ( x, y , z ) = u 2
w( x , y , z ) u
( 4)
Strain-Displacement Relation:
w
v
u
, y =
, z =
,
x =
z
y
x
u w
w v
v u
+
xy = +
+
, yz =
, xz =
z x
y z
x y
(5)
or
ij =
1 ui u j
+ ,
2 x j xi
( i, j = 1, 2, 3)
1
ui , j + u j ,i )
(
2
( tensor notation)
or simply,
ij =
140
Equilibrium Equations:
x xy xz
+
+
+ fx = 0 ,
x
y
z
yx y yz
+ fy = 0 ,
+
+
z
y
x
zx zy z
+
+
+ fz = 0 ,
x
y
z
or
ij , j + f i = 0
( 6)
on ( specified traction )
( 7)
( traction ti = ij n j )
p
n
( = u + )
u
Stress Analysis:
Solving equations in (3), (5) and (6) under the BCs in (7)
provides the stress, strain and displacement fields (15 equations
for 15 unknowns for 3-D problems). Analytical solutions are
difficult to find!
1997-2002 Yijun Liu, University of Cincinnati
141
u=
N i ui
i =1
v=
N i vi
(8)
i =1
w=
N i wi
i =1
Nodal values
In matrix form:
N1
u
= 0
v
w
0
( 31)
or
0
N1
0
0
0
N1
N2
0
0
0
N2
0
0
0
N2
u1
v1
L
w1
L
(9)
u2
L
( 33N ) v2
w2
M ( 3N 1)
u=Nd
Using relations (5) and (8), we can derive the strain vector
=B d
(61) (63N)(3N1)
142
Stiffness Matrix:
k = B T E B dv
(10)
(3N)
(3N6)(66)(63N)
143
linear (4 nodes)
Hexahedron (brick):
linear (8 nodes)
Penta:
linear (6 nodes)
144
Element Formulation:
Linear Hexahedron Element
3
4
y
1
5
6
x
mapping (xyz)
(-1 ,, 1)
(-1,1,-1) 4
3 (1,1,-1)
(-1,1,1) 8
7 (1,1,1)
o
(-1,-1,-1) 1
(-1,-1,1) 5
2 (1,-1,-1)
6 (1,-1,1)
u = N i ui ,
i =1
i =1
1i =1
v = N i vi , w = N i wi
(11)
145
Shape functions:
1
N 1 ( , , ) = (1 ) (1 ) (1 ) ,
8
1
N 2 ( , , ) = (1 + ) (1 ) (1 ) ,
8
1
(12)
N 3 ( , , ) = (1 + ) (1 + ) (1 ) ,
8
M
M
1
N 8 ( , , ) = (1 ) (1 + ) (1 + ) .
8
Note that we have the following relations for the shape
functions:
N i ( j , j , j ) = ij , i, j = 1,2,L, 8.
8
N i ( , , ) = 1.
i =1
i =1
i =1
i =1
x = N i xi , y = N i yi , z = N i zi .
(13)
The same shape functions are used as for the displacement field.
Isoparametric element.
146
Jacobian Matrix:
u
u
=
u
x
x
z
z
u
x
u
y
u
z
(14)
Jacobian matrix
u
u
x
u
1 u
=J ,
y
u
u
z
u
=
N i
u
,
etc
.
i
i =1
and
v
v
x
v
1 v
=J ,
y
v
v
z
(15)
also for w.
147
u
x
v
x
y
y w
z z
= = x u = L use (15) = B d
xy +
yz x y
w v
zx +
y z
u
w
+
z x
where d is the nodal displacement vector,
i.e.,
= Bd
(16)
(61) (624)(241)
148
Strain energy,
1
1
U = T dV = ( E ) T dV
2V
2V
1
= T E dV
2V
=
1 T T
d B E B dV d
2
(17)
(18)
(2424) (246)(66)(624)
In coordinates:
dV = (det J ) d d d
(19)
1 1 1
k = B T E B (det J ) d d d
( 20)
1 1 1
( Numerical integration)
149
pA/12
Area =A
Stresses:
= E = EBd
Principal stresses:
1 , 2 , 3 .
von Mises stress:
1
e = VM =
( 1 2 ) 2 + ( 2 3 ) 2 + ( 3 1 ) 2 .
2
Stresses are evaluated at selected points (including nodes)
on each element. Averaging (around a node, for example) may
be employed to smooth the field.
Examples:
150
Baseball bat
shaft
r, u
z, w
r, u
r
rz
r
151
Displacement field:
u = u ( r , z ) , w = w( r , z )
Strains:
u
u
w
,
,
= ,
z =
r
r
z
w u
, ( r = z = 0)
rz =
+
r z
r =
r
d
( 21)
u
(r+u)d
rd
Stresses:
r
E
=
z (1 + v ) (1 2v )
rz
v
v
0
1 v
v 1 v
v
0
v 1 v
0
v
1 2v
0
0
0
r
( 22)
z
rz
152
Axisymmetric Elements
r, u
r, u
3
1
3-node element (ring)
k = B T E B rdr d dz
or
2 1 1
k=
B T E B r (det J ) d d d
0 1 1
1 1
= 2
B T E B r (det J ) d d
( 24)
1 1
153
Applications
Rotating Flywheel:
z
Body forces:
fr = r 2
fz = g
( gravitational force)
154
r0
q = ( p ) 2 r0
Press Fit:
r0
ri
ri +
ring ( Sleeve)
shaft
at r = ri :
uo ui =
MPC
i o
155
156
F(t)
I. Basic Equations
A. Single DOF System
k
m
f=f(t)
ku
c u&
m - mass
k - stiffness
c - damping
f ( t ) - force
f(t)
x, u
mu&&+cu& +k u = f(t) ,
(1)
2
2
where u is the displacement, u& = du / dt and u&& = d u / dt .
157
Free Vibration:
mu&&+k u = 0 .
(2)
i.e.,
m+k U = 0.
m+k = 0,
which yields
=
k
.
m
(3)
(4)
158
u = U s in w t
U
t
U
T = 1 /f
U n d a m p e d F r e e V ib r a t io n
(6)
c
(damping ratio).
cc
(7)
159
(8)
in which:
M mass matrix,
C damping matrix,
K stiffness matrix,
f forcing vector.
Physical meaning of Eq. (8):
Inertia forces + Damping forces + Elastic forces
= Applied forces
Mass Matrices
Lumped mass matrix (1-D bar element):
m1 =
AL 1 ,A,L 2
2 u1
u2
m2 =
AL
2
AL
m= 2
AL
0
1442424
3
diagonal matrix
160
m=
N T NdV
(9)
1
1 T
u& m u&
(cf. mv 2 )
2
2
1
1
T
= u& 2 dV = (u& ) u& dV
2 V
2 V
1
T
= (N u& ) (N u& )dV
2 V
1
= u& T N T N dV u&
V
2
1
42 43
1
[1 ]ALd
m =
V
1 / 3 1 / 6 u&&1
= AL
1 / 6 1 / 3 u&&2
(10)
161
Example
Simple Beam Element:
v2
v1
, A, L
m = NT NdV
V
13L v&&1
54
156 22L
4 L2
13L 3L2 &&1
AL 22L
=
13L
156 22L v&&2
420 54
2
2 &&
L
L
L
L
13
3
22
4
(11)
Choice I
s
m
kg
m/s2
N
kg/m3
Choice II
s
mm
Mg
mm/s2
N
Mg/mm3
162
[K
M u = 0
(13)
163
(14)
[K
2
i
M ui = 0 .
u iT K u
= 0,
u iT M u j = 0 ,
for i j,
(15)
164
u iT M u i = 1,
u iT K u i = i2 .
(16)
Note:
Magnitudes of displacements (modes) or stresses in normal
mode analysis have no physical meaning.
For normal mode analysis, no support of the structure is
necessary.
i = 0 there are rigid body motions of the whole or
a part of the structure.
apply this to check the FEA model (check for
mechanism or free elements in the models).
Lower modes are more accurate than higher modes in the
FE calculations (less spatial variations in the lower modes
fewer elements/wave length are needed).
165
Example:
y
v2
2
, A, EI
1
2
L
v2 0
K M = ,
2 0
K=
EI 12 6L
,
2
3
L 6L 4L
12 156
EVP:
M=
AL 156 22L
.
2
420 22L 4L
6L + 22L
= 0,
#3
#2
#1
EI 2 v 2 1
, = 1.38 ,
1 = 3.533
4
AL
2 1
L
1
1
EI 2 v 2
, = 7.62 .
2 = 34 .81
4
AL
2 2
L
Exact solutions:
1
EI 2
1 = 3.516
,
4
AL
EI 2
2 = 22.03
.
4
AL
166
III. Damping
Two commonly used models for viscous damping.
A. Proportional Damping (Rayleigh Damping)
C = M + K
(17)
,
,
1 =
+
2 =
+
2
21
2
2 2
Damping ratio
B. Modal Damping
Incorporate the viscous damping in modal equations.
167
[K
i M ui = 0 ,
i = 1,2,..., n
(18)
u iT K u
T
ui M u
j
j
= 0,
= 0,
for i j,
and
u iT M u i = 1 ,
T
2
u
K
u
=
i
i
i ,
for i = 1, 2, , n.
(nn )
[u 1
L u
(19)
0 22
M
T
(Spectralmatrix),
K = =
M
O 0
2
0
0
L
T M = I.
(20)
u = z1 u 1 + z 2 u 2 + L + z n u n = z ,
1997-2003 Yijun Liu, University of Cincinnati
(21)
168
where
z1 ( t )
z (t )
z= 2
M
z n ( t )
(22)
(proportional damping),
f (t) .
2 1
0
=
M
0
2 2
L
2
O
L
M
. (23)
2 n n
0
169
170
&& + Cu& + Ku = F
Mu
t
1sin
23
(25)
Harmonicloading
(26)
zi ( t ) =
pi i2
(1 ) + ( 2 ii )
2 2
i
sin( t i ),
(27)
where
/i
2 i i
arctan
, phase angle
=
i
1 i2
i = i ,
ci
ci
, damping ratio
=
=
i c
2
m
c
i
[K
+ i C 2 M u = F .
171
u(t)
t0 t1 t2
t n t n+1
172
&&
u
1
( u n + 1 u n 1 ),
2 t
1
( u n +1 2 u n + u
=
( t)2
=
n 1
M
( u n +1 2u n + u n 1 ) + C
( u n +1 u n 1 ) + Ku n = fn ,
2
2 t
( t )
which yields,
Au n +1 = F(t )
where
1
1
A
M
C,
=
+
2
t
2
(
)
t
F ( t ) = f n K 2 2 M u n 1 2 M 1 C u n 1 .
2t
( t )
( t )
173
= F (t)
n +1
where
1
C +
M ,
2
t
( t)
&& n ).
F ( t ) = f ( f n + 1 , , , t , C , M , u n , u& n , u
A = K +
e . g .,
1
.
2
1
,
=
2
1
4
174
B. Modal Method
First, do the transformation of the dynamic equations using
the modal matrix before the time marching:
u =
i =1
u i zi (t ) = z ,
i = 1,2,, m.
&z&i + 2 i i z& i + i z i = p i ( t ),
Modal Method
Small model
Large model
Single loading
Multiple loading
Shock loading
Periodic loading
175
176
I. Temperature Field
Fourier Heat Conduction Equation:
1-D Case:
T
,
(1)
f x = k
x
where,
fx = heat flux per unit area,
k = thermal conductivity,
x
T = T(x) = temperature.
3-D Case:
fx
T x
(2)
f y = T y ,
f
T z
z
0 0 k
1997-2003 Yijun Liu, University of Cincinnati
177
x
y
z
t
in which,
qv = rate of internal heat generation per unit volume,
c = specific heat,
= mass density.
For steady state ( T t = 0 ) and isotropic materials, we can
obtain:
k 2 T = q v .
(5)
x
T = T,
ST
on S T ;
T
= Q,
(6)
on S q .
n
Note that at any point on the boundary S = S T U S q , only one
type of BC can be specified.
1997-2003 Yijun Liu, University of Cincinnati
178
(7)
where,
KT = conductivity matrix,
T = vector of nodal temperature,
q = vector of thermal loads.
The element conductivity matrix is given by:
k T = B T BdV .
(8)
179
1-D Case:
At temperature T1
At temperature T2
o
(9)
in which,
= the coefficient of thermal expansion,
T = T2 T1 is the change of temperature.
Total strain,
= e + o
(10)
= E 1 + T ,
= E ( o ) .
(11)
(12)
180
o = y = T .
xy o 0
(13)
Plane Strain,
x
(1 + )T
o = y = (1 + )T .
xy o
(14)
181
3-D Case:
x
T
T
y
T
o = z =
.
0
xy
0
yz
zx o 0
(15)
(16)
Stress-Strain Relation:
= E e = E( o ) .
(17)
182
Further Reading
Further Reading
1.
2.
3.
4.
5.
6.
7.
183