Finite Element Method
CONTENTS
INTRODUCTION
LINEAR TRIANGULAR ELEMENTS
– Field variable interpolation
– Shape functions construction
– Using area coordinates
– Strain matrix
– Element matrices
LINEAR RECTANGULAR ELEMENTS
– Shape functions construction
– Strain matrix
– Element matrices
– Gauss integration
CONTENTS
LINEAR QUADRILATERAL ELEMENTS
– Coordinate mapping
– Strain matrix
– Element matrices
– Remarks
HIGHER ORDER ELEMENTS
INTRODUCTION
2D solid elements are applicable for the analysis
of plane strain and plane stress problems.
A 2D solid element can have a
triangular
,
rectangular
or
quadrilateral
shape with
straight
or
curved
edges.
A 2D solid element can deform only in the plane
of the 2D solid.
At any point, there are
two components
in the
x
and
y
directions for the displacement as well as
INTRODUCTION
For plane strain problems, the thickness of the
element is
unit
, but for plane stress problems, the
actual thickness
must be used.
In this course, it is assumed that the element has a
uniform
thickness
h
.
Formulating 2D elements with a given variation of
2D solids
–
plane stress and plane strain
LINEAR TRIANGULAR
ELEMENTS
Less accurate than quadrilateral elements
Used by most mesh generators for complex
geometry
A linear triangular element:
y, v
2 (x2, y2) (u2, v2) 3 (x3, y3)
(u3, v3)
A
fsx
Field variable interpolation
( , )
( , )
h
e
x y
x y
U
N
d
3 node at nts displaceme 2 node at nts displaceme 1 node at nts displaceme 3 3 2 2 1 1 v u v u v u e d 3 1 2 3 1 2 Node 2
Node 1 Node 3
0 0 0 0 0 0 N N N N N N N where x, u y, v
1 (x1, y1) (u1, v1)
2 (x2, y2) (u2, v2) 3 (x3, y3)
(u3, v3)
A
fsx
fsy
Shape functions construction
1 1 1 1
N
a
b x c y
2 2 2 2
N a b x c y
3 3 3 3
N
a
b x c y
i i i i
N a b x c y
Assume,
i= 1, 2, 3
1
TT
i
i i
i a
N x y b
c
p
p
Shape functions construction
Delta function property:
1 for
( , )
0 for
i j j
i j N x y
i j
1 1 1
1 2 2
1 3 3
( , ) 1
( , ) 0
( , ) 0
N x y N x y N x y
Therefore, 1 1 1 1 1 1 1 1
1 2 2 1 1 2 1 2 1 3 3 1 1 3 1 3
( , ) 1
( , ) 0
( , ) 0
N x y a b x c y
N x y a b x c y
N x y a b x c y
Solving, 2 3 3 2 2 3 3 2
1 , 1 , 1
2 e 2 e 2 e
x y x y y y x x
a b c
A A A
Shape functions construction
1 1
2 2 2 3 3 2 2 3 1 3 2 1
3 3
1
1 1 1
1 [( ) ( ) ( ) ]
2 2 2
1
e
x y
A x y x y x y y y x x x y
x y
P
Area of triangle Moment matrix
Substitute a1, b1 and c1 back into N1 = a1 + b1x + c1y:
1 2 3 2 3 2 2
1
[( )( ) ( )( )]
2 e
N y y x x x x y y
A
Shape functions construction
Similarly,
2 1 1
2 2 2
2 3 3
( , ) 0 ( , ) 1 ( , ) 0
N x y N x y N x y
2 3 1 1 3 3 1 1 3
3 1 3 1 3 3
1
[( ) ( ) ( ) ]
2 1
[( )( ) ( )( )]
2
e
e
N x y x y y y x x x y
A
y y x x x x y y
A
3 1 1
3 2 2
3 3 3
( , ) 0 ( , ) 0 ( , ) 1
N x y
N x y
N x y
3 1 2 1 1 1 2 2 1
1 2 1 2 1 1
1
[( ) ( ) ( ) ]
2 1
[( )( ) ( )( )]
2
e
e
N x y x y y y x x x y
A
y y x x x x y y
A
Shape functions construction
i i i i
N
a
b x c y
1
( )
2 1
( )
2 1
( )
2
i j k k j
e
i j k
e
i k j
e
a x y x y
A
b y y
A
c x x
A
where
i
j k
i= 1, 2, 3
J, k determined from cyclic
permutation
i = 1, 2
Using area coordinates
Alternative method of constructing shape
functions
i,1
j, 2
k, 3
x y
P
A1
1 2 2 2 3 3 2 2 3 3 2
3 3
1
1 1
1 [( ) ( ) ( ) ]
2 2
1
x y
A x y x y x y y y x x x y
x y
1 1
e
A L
A
2-3-P:
Similarly, 3-1-P A2
1-2-P A3
2 2
e
A L
A
3 3
e
A L
A
Using area coordinates
1 2 3
1
L
L
L
Partitions of unity:
3 2 2 3
2 2
1 2 3
1
e e e e
A
A
A
A
A
A
L
L
L
A
A
A
A
Delta function property: e.g. L1 = 0 at if P at nodes 2 or 3
Therefore,
1 1
,
2 2,
3 3N
L
N
L
N
L
( , )
( , )
h
x y
x y
16
Strain matrix
xx yy xy u x v y u v y x LU
where 0 0 x y y x L e eBd
LNd
LU
0 0 x y y x B LN N
1 2 3
1 2 3
1 1 2 2 3 3
0 0 0
0 0 0
a a a
b b b
b a b a b a
B
Element matrices
0
d ( d ) d d
e e e
h
T T T
e
V A A
V z A h A
k B cB B cB B cB
Constant matrix
T
e
hA
ek
B cB
0
d d d d
e e e
h
T T T
e
V A A
V x A h A
Element matrices
1 1 1 2 1 3
1 1 1 2 1 3
2 1 2 2 2 3
2 1 2 2 2 3
3 1 3 2 3 3
3 1 3 2 3 3
0 0 0
0 0 0
0 0 0
d
0 0 0
0 0 0
0 0 0
e
e
A
N N N N N N
N N N N N N
N N N N N N
h A
N N N N N N
N N N N N N
N N N N N N
mFor elements with uniform density and thickness,
A p n m p n m A L L L n p
A m 2 )! 2 ( ! ! ! d 3 2
1
Element matrices
2 0 2 . 1 0 2 0 1 0 2 1 0 1 0 2 0 1 0 1 0 2 12 sy hA e m x, u y, v1 (x1, y1)
(u1, v1)
2 (x2, y2)
(u2, v2)
3 (x3, y3)
(u3, v3)
A fsx fsy l f f l sy sx
e [ ] 2 3 d
T
N f y x e f f f l 0 0 2 1 3 2 fLINEAR RECTANGULAR
ELEMENTS
Non-constant strain matrix
More accurate representation of stress and strain
Shape functions construction
x, u y, v
1 (x1, y1) (u1, v1)
2 (x2, y2) (u2, v2)
3 (x3, y3) (u3, v3)
2a
fsy
fsx
4 (x4, y4) (u4, v4)
2b
Consider a rectangular element
1 1 2 2 3 3 4 4
displacements at node 1
displacements at node 2
displacements at node 3
displacements at node 4
e
u v u u u u u u
Shape functions construction
x, u y, v
1 (x1, y1)
(u1, v1)
2 (x2, y2)
(u2, v2)
3 (x3, y3)
(u3, v3)
2a
fsy
fsx
4 (x4, y4)
(u4, v4)
2b
1 (1, 1) (u1, v1)
2 (1, 1) (u2, v2)
3 (1, +1) (u3, v3)
2a
4 (1, +1) (u4, v4)
2b
, y b
a
x
( , ) ( , )
h
e
x y x y
U N d 1 2 3 4
3
1 2 4
Node 2 Node 3
Node 1 Node 4
0
0 0 0
0
0 0 0
N
N N N
N
N N N
N
where
Shape functions construction
) 1 )( 1 ( ) 1 )( 1 ( ) 1 )( 1 ( ) 1 )( 1 ( 4 1 4 4 1 3 4 1 2 4 1 1 N N N N 1 13 at node 1 4 1
1
1 3 at node 2 4
1 1
1 3 at node 3 4
1 1
1 3 at node 4 4
1
(1 )(1 ) 0
(1 )(1 ) 0
(1 )(1 ) 1
(1 )(1 ) 0
N N N N Delta function property 4
1 2 3 4
1
1[(1 )(1 ) (1 )(1 ) (1 )(1 ) (1 )(1 )]
i i
N N N N N
Partition of)
1
)(
1
(
41
j j
j
N
1 (1, 1) (u1, v1)
2 (1, 1) (u2, v2)
3 (1, +1) (u3, v3)
2a
4 (1, +1) (u4, v4)
2b
Strain matrix
a b a b a b a b b b b b a a a a 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 LN BElement matrices
, y b
a
x
dxdy = ab dd
Therefore,
d d
d 1 T
1 1 1 T
cB B
cB B
k h A abh
A
e
d d d d 1 d d
1 1 1
0 N N N N N N
N N
m T T
A T
A h T
V
e
V
x A
h A
abh
Element matrices
l f f l sy sxe [ ] 2 3 d
T
N f x, u y, v1 (x1, y1)
(u1, v1)
2 (x2, y2)
(u2, v2) 3 (x3, y3)
(u3, v3)
2a
fsy
fsx
4 (x4, y4) (u4, v4)
2b
For uniformly distributed load,
Gauss integration
For evaluation of integrals in ke and me (in practice)
In 1 direction: ( )d ( )
1 1
1 j j
m
j
f w f
I
m gauss points gives exact solution of
polynomial integrand of n = 2m - 1
1 1 1 1
1 1
( , )d d ( , )
y
x n
n
i j i j
i j
I f w w f
Gauss integration
m j wj Accuracy n
1 0 2 1
2 -1/3, 1/3 1, 1 3
3 -0.6, 0, 0.6 5/9, 8/9, 5/9 5 4 -0.861136, -0.339981,
0.339981, 0.861136
0.347855, 0.652145, 0.652145, 0.347855
7 5 -0.906180, -0.538469, 0,
0.538469, 0.906180
0.236927, 0.478629, 0.568889, 0.478629, 0.236927
9
6 -0.932470, -0.661209, -0.238619, 0.238619, 0.661209, 0.932470
0.171324, 0.360762, 0.467914, 0.467914, 0.360762, 0.171324
Evaluation of
m
e E.g. ) 1 )( 1 ( 4 ) 1 )( 1 ( ) 1 )( 1 ( 16 3 1 3 1 1 1 1 1 1 1 1 1 j i j i j i j i j i ij hab d d hab d d N N hab m
9 4 ) 1 1 1 )( 1 1 1 ( 4 3 1 3 1 33 hab habm
LINEAR QUADRILATERAL
ELEMENTS
Rectangular elements have limited application
Quadrilateral elements with unparallel edges are
more useful
Irregular shape requires coordinate mapping
Coordinate mapping
2 (x2, y2)
y
x 1 (1, 1) 2 (1, 1)
3 (1, +1) 4 (1, +1)
3 (x3, y3) 4 (x4, y4)
1 (x1, y1)
Physical coordinates Natural coordinates
( , ) ( , )
h
e
U N d (Interpolation of displacements)
( , )
( , )
eCoordinate mapping
( , )
( , )
eX N x
where x
y X , 1 1 2 2 3 3 4 4
coordinate at node 1
coordinate at node 2
coordinate at node 3
coordinate at node 4
e x y x y x y x y x ) 1 )( 1 ( ) 1 )( 1 ( ) 1 )( 1 ( ) 1 )( 1 ( 4 1 4 4 1 3 4 1 2 4 1 1 N N N N i i i x N
x ( , )
4 1
y Ny ( , )
4
Coordinate mapping
Substitute 1 into i i
i
x N
x ( , )
4 1
2 (x2, y2)
y
x 1 (1, 1) 2 (1, 1)
3 (1, +1) 4 (1, +1)
3 (x3, y3)
4 (x4, y4)
1 (x1, y1)
3 2 1 2 2 1 3 2 1 2 2 1 ) 1 ( ) 1 ( ) 1 ( ) 1 ( y y y x x x or ) ( ) ( ) ( ) ( 2 3 2 1 3 2 2 1 2 3 2 1 3 2 2 1 y y y y y x x x x x
Eliminating , { ( )} ( )
) ( ) ( 3 2 2 1 3 2 2 1 2 3 2 3 y y x x x y y x x
y
Strain matrix
y y N x x N N y y N x x N N i i i i ii i i
i i N N x N N y J or x y x y J
where (Jacobian matrix)
1 1
3
1 2 4
2 2
x y N
N N N
x y J
Strain matrix
1
i i
i i
N N
x
N N
y
J
Therefore,
N LN
B
x y
y x
0
0
Replace differentials of Ni w.r.t. x and y
with differentials of Ni w.r.t. and
Element matrices
Murnaghan (1951) : dA=det |J | dd
1 1
T
1 1
det
d d
e
h
k
B cB
J
Remarks
Shape functions used for interpolating the coordinates are
the same as the shape functions used for interpolation of the displacement field. Therefore, the element is called an
isoparametric element.
Note that the shape functions for coordinate interpolation
and displacement interpolation do not have to be the same.
Using the different shape functions for coordinate
interpolation and displacement interpolation, respectively,
will lead to the development of so-called subparametric or
HIGHER ORDER ELEMENTS
Higher order triangular elements
i (I,J,K) (0,0,p)
L1
L3 L2
(0,p1,1) (0,1,p1)
(1,0,p1) (2,0,p2)
nd = (p+1)(p+2)/2
I
J
K
p
Node i,
Argyris, 1968 :
1 2 3
( ) ( ) ( )
I J K
i I J K
N
l L l
L l
L
0 1 ( 1)
( )( ) ( )
( )
( )( ) ( )
L L L L L L
l L
HIGHER ORDER ELEMENTS
Higher order triangular elements (Cont’d)
x, u y, v
1
2 3
4 5 6
1 2 2 (2 1 1) 1
N N N L L
4 5 6 4 1 2
N N N L L
x, u y, v
1
2 3
4 5
6 7 8
9 10
1 2 3 1 1 1
1
(3 1)(3 2) 2
N N N L L L
4 9 1 2 1
9
(3 1) 2
N N L L L
10 27 1 2 3
N L L L
Cubic element
HIGHER ORDER ELEMENTS
Higher order rectangular elements
0
(0,m) (n,m)
i(n,m)
Lagrange type:
1 1
( ) ( )
D D n m
i I J I J
N
N N
l
l
0 1 1 1
0 1 1 1
( )( ) ( )( ) ( )
( )
( )( ) ( )( ) ( )
n k k n
k
k k k k k k k n
l
42
HIGHER ORDER ELEMENTS
Higher order rectangular elements (Cont’d)
1 2 3 4 5 6 7 8 9 1 1
1 1 1
1 1
2 2 1
1 1
3 2 2
1 1
4 1 2
1
( ) ( ) (1 ) (1 )
4 1
( ) ( ) (1 ) (1 )
4 1
( ) ( ) (1 )(1 )
4 1
( ) ( ) (1 )(1 )
4
D D
D D
D D
D D
N N N
N N N
N N N
N N N
1 1 5 3 1
1 1 6 2 3 1 1 7 3 2 1 1 8 1 1
1 1 2 2
9 3 3
1
( ) ( ) (1 )(1 )(1 ) 2
1
( ) ( ) (1 )(1 )(1 ) 2
1
( ) ( ) (1 )(1 )(1 ) 2
1
( ) ( ) (1 )(1 ) 2
( ) ( ) (1 )(1 )
D D
D D
D D
D D
D D
N N N
N N N
N N N
N N N N N N
HIGHER ORDER ELEMENTS
Higher order rectangular elements (Cont’d)
Serendipity type:
1 2
3
4
5
6 7
8 0
=1
=1
1 4
2 1
2
2 1
2
(1 )(1 )( 1) 1, 2, 3, 4 (1 )(1 ) 5, 7
(1 )(1 ) 6, 8
j j j j j
j j
j j
N j
N j
N j
HIGHER ORDER ELEMENTS
Higher order rectangular elements (Cont’d)
1 2
3 4
5 6
7 8 9
10
11
12
2 2
1 32
2 9
32
1 3 2
9 32
(1 )(1 )(9 9 10)
for corner nodes 1, 2, 3, 4
(1 )(1 )(1 9 )
for side nodes 7, 8, 11, 12 where 1 and
(1 )(1 )(1
j j j
j j j
j j
j j
N
j
N
j
N
1 3
9 )
for side nodes 5, 6, 9, 10 where and 1
j
j j
j
ELEMENT WITH CURVED
EDGES
4 3
8
7
6 1
4 2 5
3 6
1
2 3
4 5 6
3 4
6 7
COMMENTS (GAUSS
INTEGRATION)
When the Gauss integration scheme is used, one has to
decide how many Gauss points should be used.
Theoretically, for a one-dimensional integral, using m
points can give the exact solution for the integral of a polynomial integrand of up to an order of (2m1).
As a general rule of thumb, more points should be used for
COMMENTS (GAUSS
INTEGRATION)
Using a smaller number of Gauss points tends to
counteract the over-stiff behaviour associated with the
displacement-based method.
Displacement in an element is assumed using shape
COMMENTS ON GAUSS
INTEGRATION
Two Gauss points for linear elements, and two or three
points for quadratic elements in each direction should be sufficient for most cases.
Most of the explicit FEM codes based on explicit
CASE STUDY
CASE STUDY
Elastic Properties of Polysilicon Young’s Modulus, E 169GPa
Poisson’s ratio, 0.262 Density, 2300kgm-3
10N/m
CASE STUDY
CASE STUDY
CASE STUDY
CASE STUDY
Analysis no. 4: Von Mises stress distribution using 24 eight-nodal,
CASE STUDY
CASE STUDY
Analysis no.
Number / type of elements
Total number of nodes in
model
Maximum Von Mises Stress (GPa) 1 24 bilinear,
quadrilateral 41 0.0139
2 96 bilinear,
quadrilateral 129 0.0180
3 144 bilinear,
quadrilateral 185 0.0197
4 24 quadratic,
quadrilateral 105 0.0191 5 192 linear,