Method of Finite Elements by Dr. Mojsilovic

Swiss Federal Institute of Technology
Page 1
M th d f Fi it El t I Method of Finite Elements I
Held by Dr. N. Mojsilovi
Assistant: Jianjun Qin, HIL E 13.1
Lectures homepage: http://www.ibk.ethz.ch/fa/education/
Course book: Finite Element Procedures by K.J. Bathe
Method of Finite Elements I
Oral examination, 30 minutes
Page 2
C t t Contents
An overview of the MFE I course
MFE development p
Introduction to the use of Finite Elements
Modelling the physical problem Modelling the physical problem
Finite elements as a tool for computer-aided
design and assessment design and assessment
Basic mathematical tools a review
Page 3
C O i Course Overview
26.02.2010 Introduction to the use of finite
l B i h i l l elements; Basic mathematical tools
*05.03.2010 Basic concepts of engineering analysis
12.03.2010 Displacement-based method of finite
elements
*19.03.2010 Formulation of finite elements
26.03.2010 Implementation 26.03.2010 Implementation
*16.04.2010 Isoparametric finite element matrices
23 04 2010 T u ele e t t ia ula ele e t 23.04.2010 Truss element, triangular element,
quadrilateral elements
Page 4
C O i Course Overview
30.04.2010 Element matrices in global coordinates
*07.05.2010 Beamelements and axisymmetric shell 07.05.2010 Beam elements and axisymmetric shell
elements
14.05.2010 Plate elements
*21.05.2010 Shell elements
28.05.2010 Solution of equilibriumequations in 28.05.2010 Solution of equilibrium equations in
static analysis
04.06.2010 Convergence, compatibility, o e ge e, o pa i i i y,
completeness, accuracy of the method of finite
elements
Page 7
MFE d l t MFE development
MFE is the confluence of three ingredients: matrix
structural analysis, variational approach and y pp
computer
1950s, M.J. Turner at Boeing (aerospace industry in
general): Direct Stiffness Method
Academia: J.H. Argyris, R.W. Clough (name finite
l ) HC M i d OC Zi ki i element), H.C. Martin and O.C. Zienkiewicz
popularisation
1960 M l h d D V b k V i ti l A h 1960s, Melosh and De Veubeke: Variational Approach
Commercial finite element computer codes
Page 8
I t d ti t th U f Fi it Introduction to the Use of Finite
Elements E e e s
Within the framework of continuummechanics Within the framework of continuum mechanics
dependencies between geometrical and physical
quantities are formulated on a differentially small q y
element and than extended to the whole continuum
As a result we obtain differential, partial differential , p
or integral equations for which, generally, an
analytical solution is not available they have to be
l d i i l d solved using some numerical procedure
MFE is based on the physical discretization of the
obser ed domain thus reducing the number of the observed domain, thus reducing the number of the
degrees of freedom; moreover the governing
equations are, in general, algebraic
equations are, in general, algebraic
Page 9
St i th MFE Steps in the MFE
Continuum is discretized in a mesh of finite elements
Elements are connected at nodes located on element
b d i boundaries
State of deformation, stresses, etc. in each element is
d ib d b i t l ti ( h ) f ti d described by interpolation (shape) functions and
corresponding values in the nodes; these node values
are basic unknowns of the method of finite elements are basic unknowns of the method of finite elements
The way in which these three steps are approached The way in which these three steps are approached
has a great influence on the results of the calculations
Page 10
B i T f MFE Basic Types of MFE
Direct MFE: analogue to displacement method
Variational MFE: based on the principle of
stationarity of a functional, which is usually total
t ti l l t t ti l potential energy or complementary potential energy
Residual MFE: based on the differential equations
th t d t d ib th bl that are used to describe the problem
Energy Balance MFE: based on the balance of
different energy types mostly used to solve different energy types, mostly used to solve
thermodynamic problems
Page 11
M d lli f th Ph i l P bl Modelling of the Physical Problem
Page 12
M d lli f th Ph i l P bl Modelling of the Physical Problem
MFE is only a way of solving the mathematical model
The solution of the physical problemdepends on the The solution of the physical problem depends on the
quality of the mathematical model the choice of the
mathematical model is crucial mathematical model is crucial
Thus, mathematical model must be reliable and effective
h h h i l d l i li bl if h i d The chosen mathematical model is reliable if the required
response can be predicted within a given level of accuracy
d th f h i measured on the response of a very comprehensive
mathematical model
The most effective mathematical model for the analysis is
the one that gives the required response with sufficient
d l
accuracy and at least costs
Page 13
Example Example
Complex physical problem modelled by a simple
th ti l d l
mathematical model
Page 14
Example Example
Detailed reference model 2D plane stress model
f MFE l i
for MFE analysis
Page 15
Example Example
Ch i f th ti l d l t d t d i d Choice of mathematical model must correspond to desired
response
The most effective mathematical model delivers reliable
answers with the least amount of effort
Any solution (including MFE) of a mathematical model is
limited to information contained in or fed into the model:
bad input bad output (garbage in garbage out)
Assessment of accuracy is based on comparisons with the y p
results from very comprehensive models but in practice
it has to be based on experience (experiments)
Page 16
MFE T l f CAD/CAM MFE as a Tool for CAD/CAM
Page 17
MFE T l f CAD/CAM MFE as a Tool for CAD/CAM
Practical application requires that solutions actica app icatio equi es t at so utio s
obtained by MFE are reliable and efficient
However it is also necessary that the use of MFE is However, it is also necessary that the use of MFE is
robust this implies that minor changes in any
input to an MFE analysis should not change the input to an MFE analysis should not change the
response quantity significantly
Engineer (user) should be able to judge the quality Engineer (user) should be able to judge the quality
of the obtained results (i.e. for plausibility)
Page 18
M t i Matrices
A matrix is an array of ordered numbers. A general matrix
consists of mn numbers arranged in m rows and n columns, g ,
thus the matrix is of order m x n (m by n). When we have
only one row (m = 1) or one column (n = 1), A is also called a y ( ) ( )
vector
Page 19
M t i Matrices
Example: Ax=b, where A is matrix of coefficients, x is an
array of unknowns and b an array of known quantities. y y q
Page 20
M t i Matrices
The transpose of the mxn matrix A, written as B = A
T
, is obtained by
interchanging the rows and columns in A, i.e. b
ij
=a
ji
If A A
T
i f ll h h b f d l i A If A = A
T
, it follows that the number of rows and columns in A are
equal and that a
ij
= a
ji
. Because m = n, we say that A is a square matrix
of order n, and because a
ij
=a
ji
, we say that A is a symmetric matrix. of order n, and because a
ij
a
ji
, we say that A is a symmetric matrix.
Note, symmetry implies that A is square, but not vice versa; i.e., a
square matrix need not be symmetric
S i i h l l h di l Square matrix with only zero elements, except on the diagonal
where they are unity, is called an identity or unit matrix I
Page 21
B d d M t i Banded Matrices
For symmetric banded matrix A we have a
ij
= 0 for
j > i+m
A
2m
A
+1 being the bandwidth j > i+m
A
, 2m
A
+1 being the bandwidth
If the half-bandwidth, m
A
, of a matrix is zero, we have
nonzero elements only on the diagonal of the matrix and nonzero elements only on the diagonal of the matrix and
denote it as a diagonal matrix (for example, unit matrix).
14
0 4 1 2
A A
a j m m = = > + =
14 A A
j
2 1 2 2 1 5
A
m + = + =
Page 22
O ti ith M t i Operations with Matrices
Matrix equality: A(mxp) = B(nxq) if and only if m = n
d d b and p = q and a
ij
= b
ij
Additi A( ) d B( ) b dd d if d l if Addition: A(mxp) and B(nxq) can be added if and only if
m = n and p = q. Thus
C = A + B i e c = a +b C = A + B i.e. c
ij
= a
ij
+b
ij
Scalar multiplication: Amatrix A is multiplied by a Scalar multiplication: A matrix A is multiplied by a
scalar c by multiplying all elements of A with c. Thus
B = cA i e b = ca B = cA, i.e. b
ij
= ca
ij
Page 23
Multiplication: Two matrices A(mxp) and B(nxq) can be
multiplied only if p = n. Thus C(mxq) and C = BA, i.e.
n
b a c
p y p ( q) ,
=
=
r
rj ir ij
b a c
1
Inversion: The inverse of matrix A is denoted A
-1
. If the
inverse matrix exist than we have equality AA
-1
= A
-1
A = I
A matrix with an inverse is said to be nonsingular; a
h l matrix without an inverse is a singular matrix
Page 24
Multiplication C=AB
Page 25
Inversion: AA
-1
= A
-1
A = I
Page 26
Commutative lawdoes not hold, i e AB BA Commutative law does not hold, i.e. AB BA
Distributive lawdoes hold, i.e. E = (A+B)C = AC+BC Distributive law does hold, i.e. E (A B)C AC BC
Associative law does hold, i.e. G = (AB)C =A(BC) = ABC , ( ) ( )
AB = CB does not imply that A = C p y
Special rule for the transpose of matrix product: p p p
(AB)
T
= B
T
A
T
Page 27
Matrices can be subdivided to facilitate matrix
manipulations
Partitioning lines must run completely across the original g p y g
matrix
Page 28
Th T f M t i The Trace of a Matrix
The trace of a matrix A is defined only if A is a square
matrix (nxn)
( )

=
n
ii
a tr A
matrix (nxn)
The trace of a matrix is a scalar value:
( )

= i
ii
a tr
1
A
The trace of a matrix is a scalar value:
Some rules:
tr(A+B) = tr(A) + tr(B) ( ) ( ) ( )
tr(cA) = c tr(A)
tr(AB) = tr(BA)
Page 29
Th T f M t i The Trace of a Matrix
The trace of a matrix A, tr(A) = 4+6+8+12=30
Page 30
Th D t i t f M t i The Determinant of a Matrix
The determinant of a matrix A is defined only if A is a
square matrix (nxn) square matrix (nxn)
The determinant of a matrix is a scalar value and is
obtained by means of the recurrence formula:
( )
n
j 1
d t 1 ) d t( A A

+
obtained by means of the recurrence formula:
( )
j j
i
j
a
1 1
1
1
det 1 ) det( A A

=
+
=
where A
1j
is the (n-1)x(n-1) matrix obtained by eliminating
the 1
st
row and the j
th
column from the matrix A e o a e j o u o e a i A
if A=[a
11
] then detA=a
11
[
11
]
11
Page 31
The determinant of a matrix is a scalar value and is obtained
by means of the recurrence formula:
( )
n
j
a
1
det 1 ) det( A A

+
by means of the recurrence formula:
( )
j j
i
j
a
1 1
1
1
det 1 ) det( A A

=
+
=
Page 32
The determinant of a matrix using the recurrence formula
along the first row(2 1 0):
( )
n
j 1
+
along the first row (2 1 0):
( )
j j
i
j
a
1 1
1
1
det 1 ) det( A A

=
+
=
Page 33
along the second row(1 3 1):
( )
2
d t( ) 1 d t
n
j +
A A
along the second row (1 3 1):
( )
2 2
1
det( ) 1 det
j
j j
i
a
=
=
A A
Page 34
along the third row(0 1 2):
( )
3
d t( ) 1 d t
n
j +
A A
along the third row (0 1 2):
( )
3 3
1
det( ) 1 det
j
j j
i
a
=
=
A A
Page 35
It is convenient to decompose a matrix A by so called LDL
decomposition: A=LDL
T
decomposition: A LDL
L is a lower triangular matrix with all diagonal elements
equal to 1 and D is a diagonal matrix with components d
ii
equal to 1 and D is a diagonal matrix with components d
ii
1 0 0
(
(
21
31 32
1 0
1
L l
l l
(
=
(
(

Thus the determinant of matrix A (nxn) can be obtained as:
n
=
=
i
ii
d
1
det A
i
Page 36
LDL decomposition: A=LDL
T
Page 37
S U f l O ti ith Some Useful Operations with
Determinants Determinants
d t(AB) d t(A) d t(B) det(AB) = det(A) det(B)
det(A
-1
) = 1/det(A)
det(I) = 1
Page 38
Te o Tensors
A set of quantities that obey certain transformation laws
relating the bases in one generalized coordinate systemto relating the bases in one generalized coordinate system to
those of another
Atensor consists of an array of a certain order (for A tensor consists of an array of a certain order (for
example: tensor of order 0 is a scalar, tensor of order 1 is a
vector) vector)
Bathe: An entity is called a second-order tensor if it has
nine components t
ij
, i=1,2,3 and j=1,2,3 in the unprimed nine components t
ij
, i 1,2,3 and j 1,2,3 in the unprimed
frame and nine components t
ij
in the primed frame and if
these components are related by the characteristic law e e o po e a e e a e y e a a e i i a
t
ij
=p
ik
p
jl
t
kl,
P being a rotation matrix
Page 39
St e Te o Stress Tensors
|
|
|
|
|
|
|
|
=
yz yy
xz xy xx

|
.
\
zz
y yy
Agraphical representation of a tensor is possible using A graphical representation of a tensor is possible using
Mohrs circles (for example: plane stress state shown on
figure above)
figure above)
Page 40
Variational Calculus Variational Calculus
Variational operator
Variations (of deformation) are small enough not to disturb ( ) g
the equilibrium and are consistent with the geometric constraint
of the system y
Some rules:
( ) F F F F
( ) u
d
d
d
du
=
|
.
|
\
|
( )
( ) ( ) ( ) F F F F F F
F F F F

2 1 2 1 2 1
2 1 2 1
+ =
=
( )
dx dx
. \
( ) ( ) ( )
( ) ( ) F F F F F
2
2 1 2 1 1
2 1 2 1 2 1
=
|
|
|
|
} }
=
a a
udx udx
F nF F
F
F
n n

1
2
2
2
=
|
.
\
} }
0 0
F nF F =

Method of Finite Elements by Dr. Mojsilovic

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Method of Finite Elements by Dr. Mojsilovic

Uploaded by

Copyright:

Available Formats

Swiss Federal Institute of Technology

You might also like