Finite Element Method

Chapter 10

Isoparametric Formulation
Basic Principle of Isoparametric Elements

Definition:
The term isoparametric (same parameters) is derived from the use of
the same shape (interpolation) functions N to define the elements
geometric shape as are used to define the displacements within the
element.

Alternatively:
The basic principle of isoparametric elements is that the interpolation
functions for the displacements are also used to represent the geometry
of the element.

4 4
u N i ui , v N i v i
i 1 i 1
4 4
x N i x i , y N i y i
i 1 i 1
Basic Principle of Isoparametric Elements

In this formulation, displacements are expressed in terms of the natural

(local) coordinates and then differentiated with respect to global
coordinates. Accordingly, a transformation matrix [J], called Jacobian, is
produced.

If the geometric interpolation functions are of lower order than the

displacement shape functions, the element is called subparametric. If the
reverse is true, the element is referred to as superparametric.

The isoparametric formulation is generally applicable to 1-, 2- and 3-

dimensional stress analysis. The isoparametric family includes elements
for plane, solid, plate, and shell problems. Also, it is applicable for
nonstructural problems.
Basic Principle of Isoparametric Elements

In this formulation, displacements are expressed in terms of the natural

(local) coordinates and then differentiated with respect to global
coordinates. Accordingly, a transformation matrix [J], called Jacobian, is
produced.

If the geometric interpolation functions are of lower order than the

displacement shape functions, the element is called subparametric. If the
reverse is true, the element is referred to as superparametric.

The isoparametric formulation is generally applicable to 1-, 2- and 3-

dimensional stress analysis. The isoparametric family includes elements
for plane, solid, plate, and shell problems. Also, it is applicable for
nonstructural problems.
Basic Principle of Isoparametric Elements

The isoparametric formulation makes it possible to generate elements that

are nonrectangular and have curved sides. So it can facilitate an
accurate representation of irregular elements.

Numerous commercial computer programs have adopted this formulation

for their various libraries of elements.
Basic Principle of Isoparametric Elements

As shown in the figure, the local (natural) coordinate system (,) for the two
elements have their origins at the centroids of the elements, with (,) varying
form 1 to 1. The natural coordinate system needs not to be orthogonal and
neither has to be parallel to the x-y axes.
The coordinate transformation will map the point (,) in the master element to
x(,) and y(,) in the slave element.
Examples
t t
1 1
1
s
1
y
s
x
t
t

s s
y
1
x
Step 2: Select Displacement Functions
In other words, we look for shape functions that map the regular shape
element in isoparametric coordinates to the quadrilateral in the x-y
coordinates whose size and shape are determined by the eight nodal
coordinates x1, y1, x2, y2, .., x4, y4.

x ( , ) a1 a 2 a 3 a 4
y ( , ) a 5 a 6 a 7 a 8
Step 2: Select Displacement Functions

x ( , ) a1 a 2 a3 a 4

y ( , ) 5
a a 6 a 7 a 8
a1
x ( , ) 1 0 0 0 0 a 2
0
y ( , ) 0 0 0 1
a8

x1 1 1 1 1 a1 a1 1 1 1 1 x1
x 1 1 1 1 a2 a 1
2 2 1 1 1 1 x2

x3 1 1 1 1 a3
a3 4 1 1 1 1 x3
x4 1
1 1 1 a4 a4 1
1 1 1 x4
Step 2: Select Displacement Functions

a1 1 1 1 1 x1
a 1 1 x 2
1 1 1
x ( , ) 1 2 1
a3 4 1 1 1 1 x3
a 4
1 1 1 1 x 4
1
[(1 )(1 ) x1 (1 )(1 ) x 2 (1 )(1 ) x 3 (1 )(1 ) x 4 ]
4
x1
y
1
x2 4
N i xi
x( , ) N1 0 N2 0 N3 0 N4 0 y 2 i 1

y ( , ) 0 N1 0 N2 0 N3 0 N 4 x3 4

y3 i 1
N i yi


x4
y
4
Step 2: Select Displacement Functions
Shape Function for 4-Nodes quadrilateral Elements

1 1
N 1 (1 )(1 ) , N 2 (1 )(1 )
4 4
1 1
N 3 (1 )(1 ) , N 4 (1 )(1 )
4 4

These shape functions are seen to map the (,) coordinates of any point in
the rectangular element in the above master element to x and y coordinates
in the quadrilateral (slave) element.
For example, consider the coordinates of node 1, where =-1,=-1 using the
above equation, we get x=x1 , y=y1
Step 2: Select Displacement Functions
Shape Function for 4-Nodes quadrilateral Elements

1 at node i
Ni
0 at all other nodes

n
Ni 1 , (i 1, 2, ,, n)
i 1

where n = the number of shape functions associated with number of nodes

Step 2: Select Displacement Functions
u1
v
1
u 2 4
0 v2
N i ui
u ( , ) N1 0 N2 0 N3 0 N4 i 1

v( , ) 0 N1 0 N2 0 N3 0 N 4 u3 4

v3 i 1
N i vi


u 4
v
4
u
[ N ] [d ]
v

where u and v are displacements parallel to the global x and y coordinates

Step 3: Define the Strain/Displacement and Stress/Strain Relationships

Using Chain Rule

f f x f y

x y
f f x f y

x y

f x y f

x


f x y f
y

Step 3: Define the Strain/Displacement and Stress/Strain Relationships

Using Chain Rule

N x y N


x

Can be We want to compute
computed
N x y N these for the B matrix

This is known as the

Jacobian matrix (J) for the
mapping (,) (x,y)

N N f f
x
x
1
J J
N N f f
y y
Basic Principle of Isoparametric Elements

Define the Strain/Displacement and Stress/Strain Relationships

x y 4 Ni 4
Ni

xi yi
[J ] i 1 i 1
x y 4 Ni 4
N i
x yi
i 1 i i 1

y y

1
[J ]
1

J x x



where
x y x y
J

Basic Principle of Isoparametric Elements

Define the Strain/Displacement and Stress/Strain Relationships

Since: N N
x
1
J
N N
y

N y y N
x 1


N J x x N

y

N 1 y N y N

x J
N 1 x N x N

y J
Basic Principle of Isoparametric Elements

Define the Strain/Displacement and Stress/Strain Relationships

u

x x
v
{ } y
y
xy u v
y x

y ( ) y ( )
0

x
1 x ( ) x ( ) u
y
J
0
v
xy x ( ) x ( ) y ( ) y ( )

{ } [ D ] [ N ] [d ]
Basic Principle of Isoparametric Elements

Define the Strain/Displacement and Stress/Strain Relationships

{ } [ D ] [ N ] [d ]

y ( ) y ( )
0

1 x ( ) x ( )
[ D ] 0
J
x ( ) x ( ) y ( ) y ( )

[ B] [ D] [N ]
(3 8) (3 2) (2 8)
Basic Principle of Isoparametric Elements
Derive the Element Stiffness Matrix and Equations

[k ] [ B]T [ D] [ B] t dx dy
A

f (x , y ) dx dy f ( , )
A A
J d d

1 1
[k ] [B ]T [D ][B ] t J d d
1 1
 Derive the Element Stiffness Matrix and Equations
The shape function are:
1 1
N1 (1 )(1 ) , N2 (1 )(1 )
4 4
1 1
N 3 (1 )(1 ) , N 4 (1 )(1 )
4 4
Their derivatives:
N1 1 N2 1 N3 1 N4 1
(1 ) , (1 ) , (1 ) , (1 )
4 4 4 4
and
N1 1 N2 1 N3 1 N4 1
(1 ) , (1 ) , (1 ) , (1 )
4 4 4 4

x y 4 Ni 4
Ni

xi yi
[J ] i 1 i 1
x y 4 Ni 4
N i
x yi
i 1 i i 1
 Basic Principle of Isoparametric Elements

Derive the Element Stiffness Matrix and Equations

x 4 Ni y 4 Ni
J11 xi , J12 yi
i 1 i 1
x 4 Ni y 4 Ni
J 21 xi , J 22 yi
i 1 i 1

x1 y1
J11 J12 N1, N 2 , N 3, N 4 , x2 y2
[J ]
N N 2 , N 3 , N 4 , x3 y3
21
J J 22 1,

x4 y4

J J11 J 22 J12 J 21

Explicit formulation for |J| for 4 node Element

0 1 1 y 1
1 y
2
J11 J12 1 0 1
J x x 2 x 3 x 4
J 21 J 22 8 1 1 0 1 y 3

1 1 0 y 4
 Basic Principle of Isoparametric Elements

Derive the Element Stiffness Matrix and Equations

[ B] [ D] [ N ]
( ) ( )
22
J J 0

12

1 ( ) ( ) N 1 0 N 0 N 0 N 0
[B ] J 21 0
2 3 4

J N 4
0 J 11
N1 0 N 2 0 N 3 0
( ) ( ) ( ) ( )
J 12

J J J 22

11 21

1
B1 B 2 B 3 B 4
J

J 22 N i , J 12 N i , 0


[ Bi ] 0 J 11 N i , J 21 N i ,

J N J N J 22 N i , J 12 N i ,
11 i , 21 i ,
 Basic Principle of Isoparametric Elements

The element body force matrix

{ f b } [ N ]T { X } dV [ N ]T { X } t dx dy
V A

1 1
{f b } [N ]T {X }t J d d
1 1

(8 1) (8 2) (2 1)

Xb
{X }
b
Y

and Xb and Yb are the weight densities (body weight/unit volume) in the x
and y directions, respectively.
 Basic Principle of Isoparametric Elements

The element surface force matrix

{ f s } [ N ]T {T } dS [ N ]T {T } t dx
S L
1
{f s } [N ]T {T } t J d
1

(4 1) (4 2) (2 1)
We assumed surface loading at edge with overall length L
(Since N1 = 0 and N2 = 0 along edge =1, and hence, no
nodal forces exist at nodes 1 and 2,
Note that for one-dimensional transformation | J | = L / 2. Also,
p, p are the pressure distributions in and , respectively.

f s3
f T
s3 1 N3 0 N4 0 p L
1 t d
f s4 0 N3 0 N 4 p 2
f s4

 Problem: Consider the following isoparamteric map

3
4
(6,6)
4 3 (3,6)
1 1
1

1 y
1 2 1 2
(3,1) (6,1)
ISOPARAMETRIC COORDINATES x
GLOBAL COORDINATES
Displacement interpolation
u N 1u1 N 2 u 2 N 3 u 3 N 4 u 4
v N 1v1 N 2 v 2 N 3 v3 N 4 v 4

Shape functions in isoparametric coord system

1 1
N1 (1 )(1 ) , N2 (1 )(1 )
4 4
1 1
N 3 (1 )(1 ) , N 4 (1 )(1 )
4 4
 The isoparamtric map
x N 1 ( , )x 1 N 2 ( , )x 2 N 3 ( , )x 3 N 4 ( , )x 4
y N 1 ( , ) y 1 N 2 ( , ) y 2 N 3 ( , ) y 3 N 4 ( , ) y 4

1 1 1 1
x (1 )(1 ) 3 (1 )(1 ) 6 (1 )(1 ) 6 (1 )(1 ) 3
4 4 4 4
3(3 )

2
1 1 1 1
y (1 )(1 ) 1 (1 )(1 ) 1 (1 )(1 ) 6 (1 )(1 ) 6
4 4 4 4
7 5

2
 The Jacobian matrix
3(3 )
x
since 2
7 5
y
2
x y 3
2 0
J x y



0 5
2

NOTE: The diagonal terms are due to stretching of the sides along the x-and y-
directions. The off-diagonal terms are zero because the element does not shear.

1 3 / 2 0 2 / 3 0 15
J
1

15 / 4 0 5 / 2
0 2 / 5
and J
4
 Hence, if I were to compute the columns of the B1 matrix along the
positive x-direction

J 22 N 1, s J 12 N 1, t 0

[B 1 ] 0 J 11 N 1, t J 21 N 1, s
J 11 N 1, t J 21 N 1, s J 22 N 1, s J 12 N 1, t

Hence

5 1 5 1
2 4 0
0 0
8

3 1 3 1
[B 1 ] 0 0 0
2 4 8
3 1
0
5 1 3 1 5 1
0
2 4 2 4
8 8
 Numerical Integration
Gauss Quadrature

Gauss quadrature implements a strategy of

positioning any two points on a curve to define a
straight line that would balance the positive and
negative errors.
Hence, the area evaluated under this straight line
provides an improved estimate of the integral.
Two points Gauss-Legendre Formula
Assume that the two Integration points are xo and x1 such that:

1
I f x dx c 0 f (x 0 ) c1f (x 1 )
1

c0 and c1 are constants, the function arguments x0 and x1 are

unknowns.(4 unknowns)
Two points Gauss-Legendre Formula

Thus, four unknowns 1

c0 f ( x0 ) c1 f ( x1 ) 1 dx 2
to be evaluated require 1
four conditions. 1
c0 f ( x0 ) c1 f ( x1 ) x dx 0
If this integration is 1
1
2
exact for a constant, c0 f ( x0 ) c1 f ( x1 ) x dx
2

3
1st order, 2nd order, and 1
1
3 order functions:
rd
c0 f ( x0 ) c1 f ( x1 ) x 3 dx 0
1
TwopointsGaussLegendreFormula

Solvingthese4equations,wecandeterminec0,c1,
x0 andx1.
The weighting factors are:c 0 c1 1
The Integration points are:
1
x0 0.5773503
3
1
x1 0.5773503
3
1 1
I f f
3 3
 TwopointsGaussLegendreFormula
Sinceweusedlimitsforthepreviousintegrationfrom1to1
andtheactuallimitsareusuallyfromatob,thenweneed
firsttotransformboththefunctionandtheintegrationfrom
thexsystemtothexdsystem

x b 1 f(x) f()
f(x1)

x a 1 f(xo)
ba ba
x
2 2
ba x
dx d xo x1
2 a b
-1 1
HigherPointsGaussLegendreFormula

n
I f ( ) d Wi f (i )
1

1
i 1

Weight Integrationpoint

n
I f ( ) d Wi f (i )
1

1
i 1

I W1 f (1) W 2 f (2 ) W n f (n )
MultiplePointsGaussLegendre
 Multiple Integration

Double integral:

d b
I f ( x, y ) dx dy
y c x a

In General transformation to natural

coordinate:

f ( x, y) dx dy f ( , ) J d d
A A
 1 1
I f ( ,) d d
1 1

n n
f ( , ) d d W ij f ( i , j )
1 1
I
1 1
i 1 j 1

Where Wij =Wi Wj

For n=2 Wij =Wi Wj=1
2 2
f ( , ) d d W ij f ( i , j )
1 1
I
1 1
i 1 j 1

1 1 1 1 1 1 1 1
f ( , ) f ( , ) f ( , )f ( , )
3 3 3 3 3 3 3 3

1 1
1 1 1 1 ,
, 3 3
3 3

1

1
1 1 1 1
, ,
3 3 3 3
Shape function for Isoparametric 6-Nodes triangular Elements

N 1 2 2 1 1
N 2 2 1
N 3 2 1
N 4 4 (1 )
N 5 4
N 6 4 (1 )
 Gauss Points for Isoparametric triangular Elements

Example 1. A n=1 point rule is exact for a polynomial

f (s, t ) ~ 1
s t
t

1/3 1 1 1
1 I f ,
2 3 3
1/3
s

1
 Gauss Points for Isoparametric triangular Elements

Example 2. A M=3 point rule is exact for a complete polynomial of degree 2

f (s, t ) ~ 1
s t
t s 2 st t2
1/2 1 1 1 1 1 1 1
1 I f , f ,0 f 0,
1 2 6 2 2 6 2 6 2
1/2
s
3
1
Example1

For the four-noded linear plane element shown with a uniform surface
traction along side 23, evaluate the force matrix by using the energy
equivalent nodal forces. Let the thickness of the element be h=0.1 in.
Example1

Shape functions N2 and N3 must be used,

as we are evaluating the surface traction
along side 23 (at s 1). Therefore, Eq.
(10.3.33) becomes
Example1

evaluated along s = 1
Example1
Flowchart to evaluate k by four-point Gaussian quadrature
 Higher-Order Shape Functions
In general, higher-order element shape functions can
be developed by adding additional nodes to the sides
of the linear element.
These elements result in higher-order strain variations
within each element, and convergence to the exact
solution thus occurs at a faster rate using fewer
elements.
Another advantage of the use of higher-order
elements is that curved boundaries of irregularly
shaped bodies can be approximated more closely than
by the use of simple straight-sided linear elements.
Three-noded linear strain bar isoparametric element
Three-noded linear strain bar isoparametric element

u
x
x
u u x x L
J
s x s s 2
u 1 u 2 u

x J s L s
Three-noded linear strain bar isoparametric element

u1
u 3
2s s 2s s
N i ,s u i 2s u 2
s i 1 2 2
u 3
u1
u 2 u 2s s 2s s 4s
u 2
x L s L L L
u 3
2s s 2s s 4s
B
L L L
The Stiffness Matrix of three-noded linear strain bar
1
[k ] [B ]T [D ][B ] A J ds
1

L
J
2
The Stiffness Matrix of three-noded linear strain bar
The Stiffness Matrix of three-noded linear strain bar
Shape function of a quadratic isoparametric element

or, in compact index notation, we express

where i is the number of the shape function

Shape function of a cubic isoparametric element
HW:
10.6, 10.8, 10.15 and 10.21