Scribd
Upload
62 views11 pages

Gauss Points: Transfer Integral From Physical To Computational Domain

This document discusses Gauss points and numerical integration in finite element analysis. It explains that Gauss points are used to evaluate integrals over an element by approximating them as a weighted sum of the integrated function evaluated at specific points. The number and location of Gauss points required for accurate integration depends on the order of the polynomial being integrated. The document also covers mapping elements to a standard coordinate system, evaluating the Jacobian of the mapping, and classifying different types of boundary conditions.

Uploaded by

rd_mech
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
62 views11 pages

Gauss Points: Transfer Integral From Physical To Computational Domain

This document discusses Gauss points and numerical integration in finite element analysis. It explains that Gauss points are used to evaluate integrals over an element by approximating them as a weighted sum of the integrated function evaluated at specific points. The number and location of Gauss points required for accurate integration depends on the order of the polynomial being integrated. The document also covers mapping elements to a standard coordinate system, evaluating the Jacobian of the mapping, and classifying different types of boundary conditions.

Uploaded by

rd_mech
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/ 11

Gauss Points

Transfer integral from physical to


computational domain
x2 +1
[k ] = [B(x )]T EA[B(x )]dx [B( )]T EA[B( )]d
x1 1

Numerical integration with limits [-1, +1]


available
+1 +1

f ( y )dy = (a + by )dy = 2a( Analytical )

[email protected]
1 1
+1

(a + by )dy = w f ( y ) = 2a,
1 1 y1 => [ 1,+1]
1
Gauss Points
Continue
+1

(a + by )dy = w f ( y ) = 2a + (0)b = w (a + by ) = w a + (w y )b
1
1 1 1 1 1 1 1

w1 = 2, y1 = 0
First integration point is at y1 = 0 and weight, w1 = 2

Higher order polynomial

[email protected]
+1

(a + by + cy )
+ ey 3 dy = w1 f ( y1 ) + w2 f ( y2 )
2

1
Gauss Points
Gauss point and weighting factor
1 0. 2.
2 1 3 1.
3 35 59
0. 89
4 0.861136 0.347854

[email protected]
0.339981 0.6521451
n Gauss points are required for accurate integration of a
polynomial of order of 2n-1.
Limits of Integration
Arbitrarily oriented quadratic elements in
physical domain - not compatible
Map into computational domain => master
element => compatible
Change limits of integration to [-1, +1]

[email protected]
Jacobian
Jacobian matrix maps from physical domain to
computation domain
( x, y , z )
J= => dV = Jdv
( , , )
Jacobian => ratio between volume of element
in physical domain and computational domain

[email protected]
If, J = 0 => Mapping to zero volume not
acceptable
Jacobian
Evaluation of Jacobian at Gauss points
numerical integration carried out at these
points
Jacobian value at various Gauss points basis
for element quality
Less variation in Jacobian value => good
element less distortions

[email protected]
Highly distorted element
Jacobian
Jacobian of a quadrilateral element
Geometric mapping to master y 3(2, 2)
element through shape functions
4(0, 1)

x = N 1 x1 + N 2 x 2 + N 3 x 3 + N 4 x 4
y = N 1 y1 + N 2 y 2 + N 3 y 3 + N 4 y 4 1(0, 0) 2(1, 0)
x

shape functions in and

N1 =
(1 )(1 ) , N2 =
(1 + )(1 )

[email protected]
4 4
N3 =
(1 + )(1 + ) , N4 =
(1 )(1 + )
4 4
Jacobian
Substitute nodal co-ordinates in geometric
interpolation function
x = (3 + 3 + + ), y = (3 + + 3 + )
1 1
4 4
x y
1 3 + 1+
[J ] = x y =
1 1 1
= J = + +
4 1+ 3 + 2 8 8

[email protected]

J > 0 for 1 1, 1 1 satisfied
Jacobian
Distorted element Geometric mapping is given by
y
x=
(1 + )(7 + 5 )
3(6, 6)
4(0, 6) 4
17 + 5 + 7 5
2(1, 5)
y=
4
J = 0 => = 0.4
1 1 & 1 1
1(0, 0) x

[email protected]
Jacobian is zero along line - = 0.4
Plotting this line in natural csys

Mapping not proper
Boundary Conditions
Classification of boundary conditions
Essential/Direct/Drichlet boundary condition =>
given directly on field variable
Ex. Displacement, rotation, temperature, pressure
Non-essential/Natural/Neumann boundary
condition => derivatives of field variables
Ex. Strain, temperature gradient etc
Cauchy boundary condition (combination of both

[email protected]
essential and non-essential BCs)
Heat conduction convection and buckling
Boundary Conditions
Essential and non-essential BCs together not
given at a single node
Neglect non-essential
No BC specified => Zero value to non-essential BC
Relation between order of governing DE and
order of BCs
2 m
+ ... = 0

[email protected]
x 2m

Order 0 to m-1 => essential boundary conditions


Order m to 2m-1 => non-essential boundary conditions

You might also like