Calfem
Calfem
Calfem
Version 3.4
Version 3.4
ISBN: 91-8855823-1
E-mail address:
[email protected]
Homepage:
http://www.byggmek.lth.se/Calfem
Contacts:
The Division of Structural Mechanics
Lund University
PO Box 118
SE–221 00 Lund
SWEDEN
Phone: +46 46 222 0000
Fax: +46 46 222 4420
Preface
CALFEM R
is an interactive computer program for teaching the finite element method
(FEM). The name CALFEM is an abbreviation of ”Computer Aided Learning of the Finite
Element Method”. The program can be used for different types of structural mechanics
problems and field problems.
CALFEM, the program and its built-in philosophy have been developed at the Division of
Structural Mechanics, Lund University, starting in the late 70’s. Many coworkers, former
and present, have been engaged in the development at different stages.
This release represents the latest development of CALFEM. The functions for finite element
applications are all MATLAB functions (.m-files) as described in the MATLAB manual.
We believe that this environment increases the versatility and handling of the program
and, above all, the ease of teaching the finite element method. CALFEM also works with
Octave, presently with exception for some graphical functions.
The authors
Contents
1 Introduction 1–1
The finite element analysis can be carried out either interactively or in a batch oriented
fashion. In the interactive mode the functions are evaluated one by one in the MATLAB
command window. In the batch oriented mode a sequence of functions are written in a file
named .m-file, and evaluated by writing the file name in the command window. The batch
oriented mode is a more flexible way of performing finite element analysis because the
.m-file can be written in an ordinary editor. This way of using CALFEM is recommended
because it gives a structured organization of the functions. Changes and reruns are also
easily executed in the batch oriented mode.
A command line consists typically of functions for vector and matrix operations, calls to
functions in the CALFEM finite element library or commands for workspace operations.
An example of a command line for a matrix operation is
C = A + B
where two matrices A and B’ are added together and the result is stored in matrix C .
The matrix B’ is the transpose of B. An example of a call to the element library is
Ke = bar1e(k)
where the two-by-two element stiffness matrix Ke is computed for a spring element with
spring stiffness k, and is stored in the variable Ke. The input argument is given within
parentheses ( ) after the name of the function. Some functions have multiple input argu-
ments and/or multiple output arguments. For example
[lambda, X] = eigen(K, M)
computes the eigenvalues and eigenvectors to a pair of matrices K and M. The output
variables - the eigenvalues stored in the vector lambda and the corresponding eigenvectors
stored in the matrix X - are surrounded by brackets [ ] and separated by commas. The
input arguments are given inside the parentheses and also separated by commas.
The statement
help function
provides information about purpose and syntax for the specified function.
1–1 INTRODUCTION
The available functions are organized in groups as follows. Each group is described in a
separate chapter.
Groups of functions
General purpose
commands for managing variables, workspace, output etc
Matrix functions for matrix handling
Material functions for computing material matrices
Element functions for computing element matrices and element forces
System functions for setting up and solving systems of equations
Statement
functions for algorithm definitions
Graphics functions for plotting
INTRODUCTION 1–2
2 General purpose functions
The general purpose functions are used for managing variables and workspace, control of
output etc. The functions listed here are a subset of the general purpose functions described
in the MATLAB manual. The functions can be divided into the following groups
Purpose:
Remove variables from workspace.
Syntax:
clear
clear name1 name2 name3 ...
Description:
clear removes all variables from workspace.
clear name1 name2 name3 ... removes specified variables from workspace.
Note:
This is a MATLAB built-in function. For more information about the clear function,
type help clear.
Purpose:
Save session in a disk file.
Syntax:
diary filename
diary off
diary on
Description:
diary filename writes a copy of all subsequent keyboard input and most of the resulting
output (but not graphs) on the named file. If the file filename already exists, the
output is appended to the end of that file.
diary off stops storage of the output.
diary on turns it back on again, using the current filename or default filename diary
if none has yet been specified.
The diary function may be used to store the current session for later runs. To make
this possible, finish each command line with semicolon ’;’ to avoid the storage of
intermediate results on the named diary file.
Note:
This is a MATLAB built-in function. For more information about the diary function,
type help diary.
Purpose:
Display a variable in matrix bank on display screen.
Syntax:
disp(A)
Description:
disp(A) displays the matrix A on the display screen.
Note:
This is a MATLAB built-in function. For more information about the disp function,
type help disp.
Purpose:
Control output on the display screen.
Syntax:
echo on
echo off
echo
Description:
echo on turns on echoing of commands inside Script-files.
echo off turns off echoing.
echo by itself, toggles the echo state.
Note:
This is a MATLAB built-in function. For more information about the echo function,
type help echo.
Purpose:
Control the output display format.
Syntax:
See the listing below.
Description:
format controls the output format. By default, MATLAB displays numbers in a short
format with five decimal digits.
Command Result Example
format short 5 digit scaled fixed point 3.1416
format long 15 digit scaled fixed point 3.14159265358979
format short e 5 digit floating point 3.1416e+000
format long e 16 digit floating point 3.141592653589793e+000
Note:
This is a MATLAB built-in function. For more information about the format func-
tion, type help format.
Purpose:
Display a description of purpose and syntax for a specific function.
Syntax:
help function name
Description:
help provides an online documentation for the specified function.
Example:
Typing
yields
Ke=bar1e(ep)
----------------------------------------------------------
PURPOSE
Compute element stiffness matrix
for spring (analog) element.
Note:
This is a MATLAB built-in function. For more information about the help function,
type help help.
Purpose:
Retrieve variable from disk and load in workspace.
Syntax:
load filename
load filename.ext
Description:
load filename retrieves the variables from the binary file filename.mat.
load filename.ext reads the ASCII file filename.ext with numeric data arranged in m
rows and n columns. The result is an m-by-n matrix residing in workspace with the
name filename, i.e. with the extension stripped.
Note:
This is a MATLAB built-in function. For more information about the load function,
type help load.
Purpose:
Terminate CALFEM session.
Syntax:
quit
Description:
quit filename terminates the CALFEM without saving the workspace.
Note:
This is a MATLAB built-in function. For more information about the quit function,
type help quit.
Purpose:
Save workspace variables on disk.
Syntax:
save filename
save filename variables
save filename variables -ascii
Description:
save filename writes all variables residing in workspace in a binary file named file-
name.mat
save filename variables writes named variables, separated by blanks, in a binary file
named filename.mat
save filename variables -ascii writes named variables in an ASCII file named filename.
Note:
This is a MATLAB built-in function. For more information about the save function,
type help save.
GENERAL PURPOSE 2 – 10
type
Purpose:
List file.
Syntax:
type filename
Description:
type filename lists the specified file. Use path names in the usual way for your
operating system. If a filename extension is not given, .m is added by default. This
makes it convenient to list the contents of .m-files on the screen.
Note:
This is a MATLAB built-in function. For more information about the type function,
type help type.
2 – 11 GENERAL PURPOSE
what
Purpose:
Directory listing of .m-files, .mat-files and .mex-files.
Syntax:
what
what dirname
Description:
what lists the .m-files, .mat-files and .mex-files in the current directory.
what dirname lists the files in directory dirname in the MATLAB search path. The
syntax of the path depends on your operating system.
Note:
This is a MATLAB built-in function. For more information about the what function,
type help what.
GENERAL PURPOSE 2 – 12
who, whos
Purpose:
List directory of variables in matrix bank.
Syntax:
who
whos
Description:
who lists the variables currently in memory.
whos lists the current variables and their size.
Examples:
who
Your variables are:
A B C
K M X
k lambda
whos
name size elements bytes density complex
A 3-by-3 9 72 Full No
B 3-by-3 9 72 Full No
C 3-by-3 9 72 Full No
K 20-by-20 400 3200 Full No
M 20-by-20 400 3200 Full No
X 20-by-20 400 3200 Full No
k 1-by-1 1 8 Full No
lambda 20-by-1 20 160 Full No
Grand total is 1248 elements using 9984 bytes
Note:
These are MATLAB built-in functions. For more information about the functions,
type help who or help whos.
2 – 13 GENERAL PURPOSE
...
Purpose:
Continuation.
Syntax:
...
Description:
An expression can be continued on the next line by using ... .
Note:
This is a MATLAB built-in function.
GENERAL PURPOSE 2 – 14
%
Purpose:
Write a comment line.
Syntax:
% arbitrary text
Description:
An arbitrary text can be written after the symbol %.
Note:
This is a MATLAB built-in character.
2 – 15 GENERAL PURPOSE
3 Matrix functions
The group of matrix functions comprises functions for vector and matrix operations and
also functions for sparse matrix handling. MATLAB has two storage modes, full and sparse.
Only nonzero entries and their indices are stored for sparse matrices. Sparse matrices are
not created automatically. But once initiated, sparsity propagates. Operations on sparse
matrices produce sparse matrices and operations on a mixture of sparse and full matrices
also normally produce sparse matrices.
3–1 MATRIX
[]()=’. ,;
Purpose:
Special characters.
Syntax:
[]()=’. ,;
Description:
[] Brackets are used to form vectors and matrices.
() Parentheses are used to indicate precedence in arithmetic expressions and to
specify an element of a matrix.
= Used in assignment statements.
’ Matrix transpose. X’ is the transpose of X. If X is complex, the apostrophe
sign performs complex conjugate as well. Do X.’ if only the transpose of the
complex matrix is desired
. Decimal point. 314/100, 3.14 and 0.314e1 are all the same.
, Comma. Used to separate matrix subscripts and function arguments.
; Semicolon. Used inside brackets to end rows. Used after an expression to
suppress printing or to separate statements.
Examples:
By the statement
a=2
the scalar a is assigned a value of 2. An element in a matrix may be assigned a value
according to
A(2, 5) = 3
The statement
D = [ 1 2 ; 3 4]
results in matrix
1 2
D=
3 4
stored in the matrix bank. To copy the contents of the matrix D to a matrix E, use
E=D
The character ’ is used in the following statement to store the transpose of the matrix
A in a new matrix F
F = A
Note:
These are MATLAB built-in characters.
MATRIX 3–2
:
Purpose:
Create vectors and do matrix subscripting.
Description:
The colon operator uses the following rules to create regularly spaced vectors:
The colon notation may also be used to pick out selected rows, columns, and elements
of vectors and matrices:
Examples:
The colon ’:’ used with integers
d=1:4
d=[1 2 3 4]
resulting in
1 2 3 4
D = 2 4 6 8
3 6 9 12
D( : , 3 : 4 )
resulting in
3 4
D( : , 3 : 4 ) =
6 8
9 12
3–3 MATRIX
:
In order to copy parts of the D matrix into another matrix the colon notation is used
as
E( 3 : 4 , 2 : 3 ) = D( 1 : 2 , 3 : 4 )
Assuming the matrix E was a zero matrix before the statement is executed, the result
will be
0 0 0 0
0 0 0 0
E =
0 3 4 0
0 6 8 0
Note:
This is a MATLAB built-in character.
MATRIX 3–4
+−∗/
Purpose:
Matrix arithmetic.
Syntax:
A+B
A−B
A∗B
A/s
Description:
Matrix operations are defined by the rules of linear algebra.
Examples:
An example of a sequence of matrix-to-matrix operations is
D=A+B−C
b=c−A∗x
B = A/s
Note:
These are MATLAB built-in operators.
3–5 MATRIX
abs
Purpose:
Absolute value.
Syntax:
B=abs(A)
Description:
B=abs(A) computes the absolute values of the elements of matrix A and stores them
in matrix B.
Examples:
Assume the matrix
−7 4
C=
−3 −8
Note:
This is a MATLAB built-in function. For more information about the abs function,
type help abs.
MATRIX 3–6
det
Purpose:
Matrix determinant.
Syntax:
a=det(A)
Description:
a=det(A) computes the determinant of the matrix A and stores it in the scalar a.
Note:
This is a MATLAB built-in function. For more information about the det function,
type help det.
3–7 MATRIX
diag
Purpose:
Diagonal matrices and diagonals of a matrix.
Syntax:
M=diag(v)
v=diag(M)
Description:
For a vector v with n components, the statement M=diag(v) results in an n × n
matrix M with the elements of v as the main diagonal.
For a n × n matrix M, the statement v=diag(M) results in a column vector v with n
components formed by the main diagonal in M.
Note:
This is a MATLAB built-in function. For more information about the diag function,
type help diag.
MATRIX 3–8
full
Purpose:
Convert sparse matrices to full storage class.
Syntax:
A=full(S)
Description:
A=full(S) converts the storage of a matrix from sparse to full. If A is already full,
full(A) returns A.
Note:
This is a MATLAB built-in function. For more information about the full function,
type help full.
3–9 MATRIX
inv
Purpose:
Matrix inverse.
Syntax:
B=inv(A)
Description:
B=inv(A) computes the inverse of the square matrix A and stores the result in the
matrix B.
Note:
This is a MATLAB built-in function. For more information about the inv function,
type help inv.
MATRIX 3 – 10
length
Purpose:
Vector length.
Syntax:
n=length(x)
Description:
n=length(x) returns the dimension of the vector x.
Note:
This is a MATLAB built-in function. For more information about the length function,
type help length.
3 – 11 MATRIX
max
Purpose:
Maximum element(s) of a matrix.
Syntax:
b=max(A)
Description:
For a vector a, the statement b=max(a) assigns the scalar b the maximum element
of the vector a.
For a matrix A, the statement b=max(A) returns a row vector b containing the
maximum elements found in each column vector in A.
The maximum element found in a matrix may thus be determined by
c=max(max(A)).
Examples:
Assume the matrix B is defined as
−7 4
B=
−3 −8
d = −3 4
The maximum element in the matrix B may be found by e=max(d) which results in
the scalar e = 4.
Note:
This is a MATLAB built-in function. For more information about the max function,
type help max.
MATRIX 3 – 12
min
Purpose:
Minimum element(s) of a matrix.
Syntax:
b=min(A)
Description:
For a vector a, the statement b=min(a) assigns the scalar b the minimum element of
the vector a.
For a matrix A, the statement b=min(A) returns a row vector b containing the min-
imum elements found in each column vector in A.
The minimum element found in a matrix may thus be determined by c=min(min(A)).
Examples:
Assume the matrix B is defined as
−7 4
B=
−3 −8
d = −7 −8
The minimum element in the matrix B is then found by e=min(d), which results in
the scalar e = −8.
Note:
This is a MATLAB built-in function. For more information about the min function,
type help min.
3 – 13 MATRIX
ones
Purpose:
Generate a matrix of all ones.
Syntax:
A=ones(m,n)
Description:
A=ones(m,n) results in an m-times-n matrix A with all ones.
Note:
This is a MATLAB built-in function. For more information about the ones function,
type help ones.
MATRIX 3 – 14
red
Purpose:
Reduce the size of a square matrix by omitting rows and columns.
Syntax:
B=red(A,b)
Description:
B=red(A,b) reduces the square matrix A to a smaller matrix B by omitting rows and
columns of A. The indices for rows and columns to be omitted are specified by the
column vector b.
Examples:
Assume that the matrix A is defined as
1 2 3 4
5 6 7 8
A=
9 10 11 12
13 14 15 16
and b as
2
b=
4
3 – 15 MATRIX
size
Purpose:
Matrix dimensions.
Syntax:
d=size(A)
[m,n]=size(A)
Description:
d=size(A) returns a vector with two integer components, d=[m,n], from the matrix
A with dimensions m times n.
[m,n]=size(A) returns the dimensions m and n of the m × n matrix A.
Note:
This is a MATLAB built-in function. For more information about the size function,
type help size.
MATRIX 3 – 16
sparse
Purpose:
Create sparse matrices.
Syntax:
S=sparse(A)
S=sparse(m,n)
Description:
S=sparse(A) converts a full matrix to sparse form by extracting all nonzero matrix
elements. If S is already sparse, sparse(S) returns S.
S=sparse(m,n) generates an m-times-n sparse zero matrix.
Note:
This is a MATLAB built-in function. For more information about the sparse function,
type help sparse.
3 – 17 MATRIX
spy
Purpose:
Visualize matrix sparsity structure.
Syntax:
spy(S)
Description:
spy(S) plots the sparsity structure of any matrix S. S is usually a sparse matrix, but
the function also accepts full matrices and the nonzero matrix elements are plotted.
Note:
This is a MATLAB built-in function. For more information about the spy function,
type help spy.
MATRIX 3 – 18
sqrt
Purpose:
Square root.
Syntax:
B=sqrt(A)
Description:
B=sqrt(A) computes the square root of the elements in matrix A and stores the result
in matrix B.
Note:
This is a MATLAB built-in function. For more information about the sqrt function,
type help sqrt.
3 – 19 MATRIX
sum
Purpose:
Sum of the elements of a matrix.
Syntax:
b=sum(A)
Description:
For a vector a, the statement b=sum(a) results in a scalar a containing the sum of
all elements of a.
For a matrix A, the statement b=sum(A) returns a row vector b containing the sum
of the elements found in each column vector of A.
The sum of all elements of a matrix is determined by c=sum(sum(A)).
Note:
This is a MATLAB built-in function. For more information about the sum function,
type help sum.
MATRIX 3 – 20
zeros
Purpose:
Generate a zero matrix.
Syntax:
A=zeros(m,n)
Description:
A=zeros(m,n) results in an m-times-n matrix A of zeros.
Note:
This is a MATLAB built-in function. For more information about the zeros function,
type help zeros.
3 – 21 MATRIX
4 Material functions
The group of material functions comprises functions for constitutive models. The available
models can treat linear elastic and isotropic hardening von Mises material. These material
models are defined by the functions:
4–1 MATERIAL
hooke
Purpose:
Compute material matrix for a linear elastic and isotropic material.
Syntax:
D = hooke(ptype,E,v)
Description:
hooke computes the material matrix D for a linear elastic and isotropic material.
The variable ptype is used to define the type of analysis.
1 plane stress.
2 plane strain.
ptype =
3
axisymmetry.
4 three dimensional analysis.
The material parameters E and v define the modulus of elasticity E and the Poisson’s
ratio ν, respectively.
For plane stress, ptype=1, D is formed as
1 ν0
E
ν 10
D=
1 − ν2 1−ν
0 0
2
MATERIAL 4–2
mises
Purpose:
Compute stresses and plastic strains for an elasto-plastic isotropic hardening von
Mises material.
Syntax:
[es,deps,st]=mises(ptype,mp,est,st)
Description:
mises computes updated stresses es, plastic strain increments deps, and state variables
st for an elasto-plastic isotropic hardening von Mises material.
The input variable ptype is used to define the type of analysis, cf. hooke. The vector
mp contains the material constants
mp = [ E ν h ]
where E is the modulus of elasticity, ν is the Poisson’s ratio, and h is the plastic
modulus. The input matrix est contains trial stresses obtained by using the elas-
tic material matrix D in plants or some similar s-function, and the input vector st
contains the state parameters
p
st = [ yi σy ef f ]
at the beginning of the step. The scalar yi states whether the material behaviour
is elasto-plastic (yi=1), or elastic (yi=0). The current yield stress is denoted by σy
and the effective plastic strain by pef f .
The output variables es and st contain updated values of es and st obtained by
integration of the constitutive equations over the actual displacement step. The
increments of the plastic strains are stored in the vector deps.
If es and st contain more than one row, then every row will be treated by the com-
mand.
Note:
It is not necessary to check whether the material behaviour is elastic or elasto-plastic,
this test is done by the function. The computation is based on an Euler-Backward
method, i.e. the radial return method.
Only the cases ptype=2, 3 and 4, are implemented.
4–3 MATERIAL
dmises
Purpose:
Form the elasto-plastic continuum matrix for an isotropic hardening von Mises ma-
terial.
Syntax:
D=dmises(ptype,mp,es,st)
Description:
dmises forms the elasto-plastic continuum matrix for an isotropic hardening von Mises
material.
The input variable ptype is used to define the type of analysis, cf. hooke. The vector
mp contains the material constants
mp = [ E ν h ]
where E is the modulus of elasticity, ν is the Poisson’s ratio, and h is the plastic
modulus. The matrix es contains current stresses obtained from plants or some
similar s-function, and the vector st contains the current state parameters
p
st = [ yi σy ef f ]
where yi=1 if the material behaviour is elasto-plastic, and yi=0 if the material
behaviour is elastic. The current yield stress is denoted by σy , and the current
effective plastic strain by pef f .
Note:
Only the case ptype=2 is implemented.
MATERIAL 4–4
5 Element functions
5.1 Introduction
The group of element functions contains functions for computation of element matrices
and element forces for different element types. The element functions have been divided
into the following groups
Spring element
Bar elements
Heat flow elements
Solid elements
Beam elements
Plate element
For each element type there is a function for computation of the element stiffness matrix
Ke . For most of the elements, an element load vector f e can also be computed. These
functions are identified by their last letter -e.
Using the function assem, the element stiffness matrices and element load vectors are
assembled into a global stiffness matrix K and a load vector f . Unknown nodal values of
temperatures or displacements a are computed by solving the system of equations Ka = f
using the function solveq. A vector of nodal values of temperatures or displacements for a
specific element is formed by the function extract.
When the element nodal values have been computed, the element flux or element stresses
can be calculated using functions specific to the element type concerned. These functions
are identified by their last letter -s.
For some elements, a function for computing the internal force vector is also available.
These functions are identified by their last letter -f.
5.1 – 1 ELEMENT
5.2 Spring element
The spring element, shown below, can be used for the analysis of one-dimensional spring
systems and for a variety of analogous physical problems.
k
u1 u2
● ●
Spring k u P N
EA
Bar u P N
L
λA
Thermal conduction T H̄ H
L
1
Electrical circuit U I¯ I
R
kA
Groundwater flow φ H̄ H
L
πD4
Pipe network p H̄ H
128µL
5.2 – 1 ELEMENT
Interpretations of the spring element
Problem type Quantities Designations
k k spring stiffness
u1, P1 u2, P2 displacement
u
Spring ● ●
P element force
N N N spring force
● ●
L length
u1, P1 u2, P2 E modulus of elasticity
A area of cross section
Bar E, A
L u displacement
N N P element force
N normal force
L length
H1 T1 H T2 H2 λ thermal conductivity
Thermal
T temperature
conduction l
H̄ element heat flow
L H internal heat flow
R
R resistance
Electrical U1 U2 U potential
circuit I1
● ●
I2 I¯ element current
I I internal current
L length
Ground- H1 f1 H f2 H2 k permeability
water φ piezometric head
k
flow H̄ element water flow
L H internal water flow
L length
Pipe D pipe diameter
p1 D, m p2
network µ viscosity
H
(laminar H1 H2 p pressure
flow) L H̄ element fluid flow
H internal fluid flow
ELEMENT 5.2 – 2
The following functions are available for the spring element:
Spring functions
spring1e Compute element matrix
spring1s Compute spring force
5.2 – 3 ELEMENT
spring1e Spring element
Purpose:
Compute element stiffness matrix for a spring element.
k
u1 u2
● ●
Syntax:
Ke=spring1e(ep)
Description:
spring1e provides the element stiffness matrix Ke for a spring element.
The input variable
ep = [ k ]
Theory:
The element stiffness matrix Ke , stored in Ke, is computed according to
e k −k
K =
−k k
ELEMENT 5.2 – 4
Spring element spring1s
Purpose:
Compute spring force in a spring element.
N N
● ●
Syntax:
es=spring1s(ep,ed)
Description:
spring1s computes the spring force es in a spring element.
The input variable ep is defined in spring1e and the element nodal displacements ed
are obtained by the function extract.
The output variable
es = [ N ]
Theory:
The spring force N , or analog quantity, is computed according to
N = k [ u2 − u1 ]
5.2 – 5 ELEMENT
5.3 Bar elements
Bar elements are available for one, two, and three dimensional analysis. For the one
dimensional element, see the spring element.
Bar elements
u4 u5
u4
u3
u6
u2
u2
u3 u1
u1
bar2e bar3e
bar2g
5.3 – 1 ELEMENT
bar2e Two dimensional bar element
Purpose:
Compute element stiffness matrix for a two dimensional bar element.
u4
u3
(x2,y2)
y u2 x
E, A
u1
(x1,y1)
Syntax:
Ke=bar2e(ex,ey,ep)
Description:
bar2e provides the global element stiffness matrix Ke for a two dimensional bar ele-
ment.
The input variables
ex = [ x1 x2 ]
ep = [ E A ]
ey = [ y1 y2 ]
supply the element nodal coordinates x1 , y1 , x2 , and y2 , the modulus of elasticity E,
and the cross section area A.
Theory:
The element stiffness matrix Ke , stored in Ke, is computed according to
Ke = GT K̄e G
where
e EA 1 −1 nxx̄ nyx̄ 0 0
K̄ = G=
L −1 1 0 0 nxx̄ nyx̄
ELEMENT 5.3 – 2
Two dimensional bar element bar2g
Purpose:
Compute element stiffness matrix for a two dimensional geometric nonlinear bar.
u4
u3
(x2,y2)
y u2 x
A N
E, A,
u1
(x1,y1)
Syntax:
Ke=bar2g(ex,ey,ep,N)
Description:
bar2g provides the element stiffness matrix Ke for a two dimensional geometric non-
linear bar element.
The input variables ex, ey and ep are described in bar2e. The input variable
N=[N ]
Ke = GT K̄e G
where
1 0 −1 0 0 0 0 0
EA
0 0 0 0
N
0 1 0 −1
K̄e = +
L −1 0 1 0 L 0 0 0 0
0 0 0 0 0 −1 0 1
nxx̄ nyx̄ 0 0
nxȳ nyȳ 0 0
G=
0 0 nxx̄ nyx̄
0 0 nxȳ nyȳ
5.3 – 3 ELEMENT
bar2g Two dimensional bar element
ELEMENT 5.3 – 4
Two dimensional bar element bar2s
Purpose:
Compute normal force in a two dimensional bar element.
y N
N
Syntax:
es=bar2s(ex,ey,ep,ed)
Description:
bar2s computes the normal force in the two dimensional bar elements bar2e and bar2g.
The input variables ex, ey, and ep are defined in bar2e and the element nodal dis-
placements, stored in ed, are obtained by the function extract.
The output variable
es = [ N ]
Theory:
The normal force N is computed from
EA
N= [ −1 1 ] G ae
L
where E, A, L, and the transformation matrix G are defined in bar2e. The nodal
displacements in global coordinates
ae = [ u1 u2 u3 u4 ]T
are also shown in bar2e. Note that the transpose of ae is stored in ed.
5.3 – 5 ELEMENT
bar3e Three dimensional bar element
Purpose:
Compute element stiffness matrix for a three dimensional bar element.
u5
u4
u6 (x2,y2,z2)
u2 x E, A
y
u3 u1
(x1,y1,z1)
x
z
Syntax:
Ke=bar3e(ex,ey,ez,ep)
Description:
bar3e provides the element stiffness matrix Ke for a three dimensional bar element.
The input variables
ex = [ x1 x2 ]
ey = [ y1 y2 ] ep = [ E A ]
ez = [ z1 z2 ]
supply the element nodal coordinates x1 , y1 , z1 , x2 etc, the modulus of elasticity E,
and the cross section area A.
Theory:
The global element stiffness matrix Ke is computed according to
Ke = GT K̄e G
where
e EA 1 −1 nxx̄ nyx̄ nzx̄ 0 0 0
K̄ = G=
L −1 1 0 0 0 nxx̄ nyx̄ nzx̄
ELEMENT 5.3 – 6
Three dimensional bar element bar3s
Purpose:
Compute normal force in a three dimensional bar element.
N
y
N
x
z
Syntax:
es=bar3s(ex,ey,ez,ep,ed)
Description:
bar3s computes the normal force in a three dimensional bar element.
The input variables ex, ey, ez, and ep are defined in bar3e, and the element nodal
displacements, stored in ed, are obtained by the function extract.
The output variable
es = [ N ]
Theory:
The normal force N is computed from
EA
N= [ −1 1 ] G ae
L
where E, A, L, and the transformation matrix G are defined in bar3e. The nodal
displacements in global coordinates
ae = [ u1 u2 u3 u4 u5 u6 ]T
are also shown in bar3e. Note that the transpose of ae is stored in ed.
5.3 – 7 ELEMENT
5.4 Heat flow elements
Heat flow elements are available for one, two, and three dimensional analysis. For one
dimensional heat flow the spring element spring1 is used.
A variety of important physical phenomena are described by the same differential equa-
tion as the heat flow problem. The heat flow element is thus applicable in modelling differ-
ent physical applications. Table 3 below shows the relation between the primary variable
a, the constitutive matrix D, and the load vector fl for a chosen set of two dimensional
physical problems.
1 1
St. Venant torsion φ , 2Θ φ = stress function
G zy G zx
Gzy , Gzx = shear
moduli
Θ = angle of torsion
per unit length
5.4 – 1 ELEMENT
Heat flow elements
T3 T4 T3
● ● ●
T2
● T1
●
T1
● T2
●
flw2te flw2qe
flw2i4e
T7 T3 T4
● T8 ● T3
● ●
T4 ● T7
● ●
● T6
T8 ●
●
●
T1
● T2
● ● T5 ●
T1 T5 ● T6
T2
flw2i8e flw3i8e
ELEMENT 5.4 – 2
Two dimensional heat flow elements flw2te
Purpose:
Compute element stiffness matrix for a triangular heat flow element.
T3
(x3,y3) ●
y T2
●
T1 (x2,y2)
●
(x1,y1)
x
Syntax:
Ke=flw2te(ex,ey,ep,D)
[Ke,fe]=flw2te(ex,ey,ep,D,eq)
Description:
flw2te provides the element stiffness (conductivity) matrix Ke and the element load
vector fe for a triangular heat flow element.
The element nodal coordinates x1 , y1 , x2 etc, are supplied to the function by ex
and ey, the element thickness t is supplied by ep and the thermal conductivities (or
corresponding quantities) kxx , kxy etc are supplied by D.
ex = [ x1 x2 x3 ] kxx kxy
ep = [ t ] D=
ey = [ y1 y2 y3 ] kyx kyy
If the scalar variable eq is given in the function, the element load vector fe is com-
puted, using
eq = [ Q ]
Theory:
The element stiffness matrix Ke and the element load vector fel , stored in Ke and fe,
respectively, are computed according to
Ke = (C−1 )T B̄ D B̄ t dA C−1
T
A
fel = (C−1 )T
T
N̄ Q t dA
A
5.4 – 3 ELEMENT
flw2te Two dimensional heat flow elements
The evaluation of the integrals for the triangular element is based on the linear
temperature approximation T (x, y) and is expressed in terms of the nodal variables
T1 , T2 and T3 as
T (x, y) = Ne ae = N̄ C−1 ae
where
1 x1 y 1 T1
N̄ = [ 1 x y ] C = 1 x2 y 2 ae = T2
1 x3 y 3 T3
Ke = (C−1 )T B̄ D B̄ C−1 t A
T
QAt
fel = [ 1 1 1 ]T
3
ELEMENT 5.4 – 4
Two dimensional heat flow elements flw2ts
Purpose:
Compute heat flux and temperature gradients in a triangular heat flow element.
Syntax:
[es,et]=flw2ts(ex,ey,D,ed)
Description:
flw2ts computes the heat flux vector es and the temperature gradient et (or corre-
sponding quantities) in a triangular heat flow element.
The input variables ex, ey and the matrix D are defined in flw2te. The vector ed
contains the nodal temperatures ae of the element and is obtained by the function
extract as
ed = (ae )T = [ T1 T2 T3 ]
es = qT = [ qx qy ]
T ∂T ∂T
et = (∇T ) =
∂x ∂y
contain the components of the heat flux and the temperature gradient computed in
the directions of the coordinate axis.
Theory:
The temperature gradient and the heat flux are computed according to
∇T = B̄ C−1 ae
q = −D∇T
where the matrices D, B̄, and C are described in flw2te. Note that both the tem-
perature gradient and the heat flux are constant in the element.
5.4 – 5 ELEMENT
flw2qe Two dimensional heat flow elements
Purpose:
Compute element stiffness matrix for a quadrilateral heat flow element.
T4 T3
(x4,y4) ● ●
(x3,y3)
●
y
T5
(x1,y1)
●
T1 (x2,y2)
●
T2
x
Syntax:
Ke=flw2qe(ex,ey,ep,D)
[Ke,fe]=flw2qe(ex,ey,ep,D,eq)
Description:
flw2qe provides the element stiffness (conductivity) matrix Ke and the element load
vector fe for a quadrilateral heat flow element.
The element nodal coordinates x1 , y1 , x2 etc, are supplied to the function by ex
and ey, the element thickness t is supplied by ep and the thermal conductivities (or
corresponding quantities) kxx , kxy etc are supplied by D.
ex = [ x1 x2 x3 x4 ] kxx kxy
ep = [ t ] D=
ey = [ y1 y2 y3 y4 ] kyx kyy
If the scalar variable eq is given in the function, the element load vector fe is com-
puted, using
eq = [ Q ]
Theory:
In computing the element matrices, a fifth degree of freedom is introduced. The
location of this extra degree of freedom is defined by the mean value of the coordinates
in the corner points. Four sets of element matrices are calculated using flw2te. These
matrices are then assembled and the fifth degree of freedom is eliminated by static
condensation.
ELEMENT 5.4 – 6
Two dimensional heat flow elements flw2qs
Purpose:
Compute heat flux and temperature gradients in a quadrilateral heat flow element.
Syntax:
[es,et]=flw2qs(ex,ey,ep,D,ed)
[es,et]=flw2qs(ex,ey,ep,D,ed,eq)
Description:
flw2qs computes the heat flux vector es and the temperature gradient et (or corre-
sponding quantities) in a quadrilateral heat flow element.
The input variables ex, ey, eq and the matrix D are defined in flw2qe. The vector ed
contains the nodal temperatures ae of the element and is obtained by the function
extract as
ed = (ae )T = [ T1 T2 T3 T4 ]
es = qT = [ qx qy ]
T ∂T ∂T
et = (∇T ) =
∂x ∂y
contain the components of the heat flux and the temperature gradient computed in
the directions of the coordinate axis.
Theory:
By assembling four triangular elements as described in flw2te a system of equations
containing 5 degrees of freedom is obtained. From this system of equations the
unknown temperature at the center of the element is computed. Then according to
the description in flw2ts the temperature gradient and the heat flux in each of the
four triangular elements are produced. Finally the temperature gradient and the
heat flux of the quadrilateral element are computed as area weighted mean values
from the values of the four triangular elements. If heat is supplied to the element,
the element load vector eq is needed for the calculations.
5.4 – 7 ELEMENT
flw2i4e Two dimensional heat flow elements
Purpose:
Compute element stiffness matrix for a 4 node isoparametric heat flow element.
T4 T3
● ●
(x4,y4) (x3,y3)
y
T1
●
(x1,y1) T2
●
(x2,y2)
x
Syntax:
Ke=flw2i4e(ex,ey,ep,D)
[Ke,fe]=flw2i4e(ex,ey,ep,D,eq)
Description:
flw2i4e provides the element stiffness (conductivity) matrix Ke and the element load
vector fe for a 4 node isoparametric heat flow element.
The element nodal coordinates x1 , y1 , x2 etc, are supplied to the function by ex and
ey. The element thickness t and the number of Gauss points n
(n × n) integration points, n = 1, 2, 3
are supplied to the function by ep and the thermal conductivities (or corresponding
quantities) kxx , kxy etc are supplied by D.
ex = [ x1 x2 x3 x4 ] kxx kxy
ep = [ t n ] D=
ey = [ y1 y2 y3 y4 ] kyx kyy
If the scalar variable eq is given in the function, the element load vector fe is com-
puted, using
eq = [ Q ]
ELEMENT 5.4 – 8
Two dimensional heat flow elements flw2i4e
Theory:
The element stiffness matrix Ke and the element load vector fel , stored in Ke and fe,
respectively, are computed according to
Ke = BeT D Be t dA
A
fel = NeT Q t dA
A
T (ξ, η) = Ne ae
where
1 1
N1e = (1 − ξ)(1 − η) N2e = (1 + ξ)(1 − η)
4 4
1 1
e
N3 = (1 + ξ)(1 + η) N4 = (1 − ξ)(1 + η)
e
4 4
5.4 – 9 ELEMENT
flw2i4s Two dimensional heat flow elements
Purpose:
Compute heat flux and temperature gradients in a 4 node isoparametric heat flow
element.
Syntax:
[es,et,eci]=flw2i4s(ex,ey,ep,D,ed)
Description:
flw2i4s computes the heat flux vector es and the temperature gradient et (or corre-
sponding quantities) in a 4 node isoparametric heat flow element.
The input variables ex, ey, ep and the matrix D are defined in flw2i4e. The vector ed
contains the nodal temperatures ae of the element and is obtained by extract as
ed = (ae )T = [ T1 T2 T3 T4 ]
∂T 1 ∂T 1
∂x
∂y
x1 y1
∂T 2 ∂T 2
x2 y2
¯ )T =
et = (∇T
∂x ∂y
eci =
.. ..
.. ..
. .
. .
xn2 yn2
2 n2
∂T n ∂T
∂x ∂y
contain the heat flux, the temperature gradient, and the coordinates of the integra-
tion points. The index n denotes the number of integration points used within the
element, cf. flw2i4e.
Theory:
The temperature gradient and the heat flux are computed according to
∇T = Be ae
q = −D∇T
where the matrices D, Be , and ae are described in flw2i4e, and where the integration
points are chosen as evaluation points.
ELEMENT 5.4 – 10
Two dimensional heat flow elements flw2i8e
Purpose:
Compute element stiffness matrix for an 8 node isoparametric heat flow element.
T7 T3
●
●
T4
●
● T6
y T8 ●
● ●
T1 T5 ●
T2
x
Syntax:
Ke=flw2i8e(ex,ey,ep,D)
[Ke,fe]=flw2i8e(ex,ey,ep,D,eq)
Description:
flw2i8e provides the element stiffness (conductivity) matrix Ke and the element load
vector fe for an 8 node isoparametric heat flow element.
The element nodal coordinates x1 , y1 , x2 etc, are supplied to the function by ex and
ey. The element thickness t and the number of Gauss points n
(n × n) integration points, n = 1, 2, 3
are supplied to the function by ep and the thermal conductivities (or corresponding
quantities) kxx , kxy etc are supplied by D.
ex = [ x1 x2 x3 . . . x8 ] kxx kxy
ep = [ t n ] D=
ey = [ y1 y2 y3 . . . y8 ] kyx kyy
If the scalar variable eq is given in the function, the vector fe is computed, using
eq = [ Q ]
5.4 – 11 ELEMENT
flw2i8e Two dimensional heat flow elements
Theory:
The element stiffness matrix Ke and the element load vector fel , stored in Ke and fe,
respectively, are computed according to
Ke = BeT D Be t dA
A
fel = NeT Q t dA
A
T (ξ, η) = Ne ae
where
1 1
N1e = − (1 − ξ)(1 − η)(1 + ξ + η) N5e = (1 − ξ 2 )(1 − η)
4 2
1 1
N2 = − (1 + ξ)(1 − η)(1 − ξ + η)
e
N6 = (1 + ξ)(1 − η 2 )
e
4 2
1 1
N3 = − (1 + ξ)(1 + η)(1 − ξ − η)
e
N7 = (1 − ξ 2 )(1 + η)
e
4 2
1 1
N4 = − (1 − ξ)(1 + η)(1 + ξ − η)
e
N8 = (1 − ξ)(1 − η 2 )
e
4 2
ELEMENT 5.4 – 12
Two dimensional heat flow elements flw2i8s
Purpose:
Compute heat flux and temperature gradients in an 8 node isoparametric heat flow
element.
Syntax:
[es,et,eci]=flw2i8s(ex,ey,ep,D,ed)
Description:
flw2i8s computes the heat flux vector es and the temperature gradient et (or corre-
sponding quantities) in an 8 node isoparametric heat flow element.
The input variables ex, ey, ep and the matrix D are defined in flw2i8e. The vector ed
contains the nodal temperatures ae of the element and is obtained by the function
extract as
ed = (ae )T = [ T1 T2 T3 . . . T8 ]
∂T 1 ∂T 1
∂x
∂y
x1 y1
∂T 2 ∂T 2
x2 y2
¯ )T =
et = (∇T ∂x ∂y
eci =
.. ..
..
.. . .
. .
xn2 yn2
2 n
2
∂T n ∂T
∂x ∂y
contain the heat flux, the temperature gradient, and the coordinates of the integra-
tion points. The index n denotes the number of integration points used within the
element, cf. flw2i8e.
Theory:
The temperature gradient and the heat flux are computed according to
∇T = Be ae
q = −D∇T
where the matrices D, Be , and ae are described in flw2i8e, and where the integration
points are chosen as evaluation points.
5.4 – 13 ELEMENT
flw3i8e Three dimensional heat flow elements
Purpose:
Compute element stiffness matrix for an 8 node isoparametric element.
T4
T8 ● T3
● ●
T7
●
y
●
●
T1
● T2
T5 ●
T6
x
z
Syntax:
Ke=flw3i8e(ex,ey,ez,ep,D)
[Ke,fe]=flw3i8e(ex,ey,ez,ep,D,eq)
Description:
flw3i8e provides the element stiffness (conductivity) matrix Ke and the element load
vector fe for an 8 node isoparametric heat flow element.
The element nodal coordinates x1 , y1 , z1 x2 etc, are supplied to the function by ex,
ey and ez. The number of Gauss points n
(n × n × n) integration points, n = 1, 2, 3
are supplied to the function by ep and the thermal conductivities (or corresponding
quantities) kxx , kxy etc are supplied by D.
ex = [ x1 x2 x3 . . . x8 ] kxx kxy kxz
ey = [ y1 y2 y3 . . . y8 ] ep = [ n ] D=
kyx kyy kyz
ez = [ z1 z2 z3 . . . z8 ] kzx kzy kzz
If the scalar variable eq is given in the function, the element load vector fe is com-
puted, using
eq = [ Q ]
ELEMENT 5.4 – 14
Three dimensional heat flow elements flw3i8e
Theory:
The element stiffness matrix Ke and the element load vector fel , stored in Ke and fe,
respectively, are computed according to
Ke = BeT D Be dV
V
fel = NeT Q dV
V
T (ξ, η, ζ) = Ne ae
where
5.4 – 15 ELEMENT
flw3i8s Three dimensional heat flow elements
Purpose:
Compute heat flux and temperature gradients in an 8 node isoparametric heat flow
element.
Syntax:
[es,et,eci]=flw3i8s(ex,ey,ez,ep,D,ed)
Description:
flw3i8s computes the heat flux vector es and the temperature gradient et (or corre-
sponding quantities) in an 8 node isoparametric heat flow element.
The input variables ex, ey, ez, ep and the matrix D are defined in flw3i8e. The vector
ed contains the nodal temperatures ae of the element and is obtained by the function
extract as
ed = (ae )T = [ T1 T2 T3 . . . T8 ]
∂T 1 ∂T 1 ∂T 1
∂x ∂y ∂z
x1 y1 z1
∂T 2 ∂T 2 ∂T 2
x2 y2 z2
¯ )T =
et = (∇T ∂x ∂y ∂z eci = .. .. ..
.. .. .. . . .
. . .
xn3 yn3 zn3
∂T n
3 3 3
∂T n ∂T n
∂x ∂y ∂z
contain the heat flux, the temperature gradient, and the coordinates of the integra-
tion points. The index n denotes the number of integration points used within the
element, cf. flw3i8e.
Theory:
The temperature gradient and the heat flux are computed according to
∇T = Be ae
q = −D∇T
where the matrices D, Be , and ae are described in flw3i8e, and where the integration
points are chosen as evaluation points.
ELEMENT 5.4 – 16
5.5 Solid elements
Solid elements are available for two dimensional analysis in plane stress (panels) and plane
strain, and for general three dimensional analysis. In the two dimensional case there are
a triangular three node element, a quadrilateral four node element, two rectangular four
node elements, and quadrilateral isoparametric four and eight node elements. For three
dimensional analysis there is an eight node isoparametric element.
The elements are able to deal with both isotropic and anisotropic materials. The triangular
element and the three isoparametric elements can also be used together with a nonlinear
material model. The material properties are specified by supplying the constitutive matrix
D as an input variable to the element functions. This matrix can be formed by the functions
described in Section 4.
5.5 – 1 ELEMENT
Solid elements
u6 u8 u6
u5 u7 u5
● ● ●
u4
u3 u2
u2 ●
u1
u1 ● u4
●
u3
plante ●
planqe
u8 u6 u6
u7 u5 u8 u5
●
● ● u7
●
u2
u1
● u4
u2 u4
u3
u1 u3 ●
● ●
planre
plantce
plani4e
7 3 ●
● ●
● ●
4
● u2 ●
6
8 ● u1
● u3 ●
u2 ●
●
1 5
● ●
● 2
u1
●
plani8e
soli8e
ELEMENT 5.5 – 2
2D solid functions
plante Compute element matrices for a triangular element
plants Compute stresses and strains
plantf Compute internal element forces
planqe Compute element matrices for a quadrilateral element
planqs Compute stresses and strains
planre Compute element matrices for a rectangular Melosh element
planrs Compute stresses and strains
plantce Compute element matrices for a rectangular Turner-Clough element
plantcs Compute stresses and strains
plani4e Compute element matrices, 4 node isoparametric element
plani4s Compute stresses and strains
plani4f Compute internal element forces
plani8e Compute element matrices, 8 node isoparametric element
plani8s Compute stresses and strains
plani8f Compute internal element forces
3D solid functions
soli8e Compute element matrices, 8 node isoparametric element
soli8s Compute stresses and strains
soli8f Compute internal element forces
5.5 – 3 ELEMENT
plante Two dimensional solid elements
Purpose:
Compute element matrices for a triangular element in plane strain or plane stress.
u6
u5
(x3,y3) ●
u4
y u3
u2 ●
u1 (x2,y2)
●
(x1,y1)
Syntax:
Ke=plante(ex,ey,ep,D)
[Ke,fe]=plante(ex,ey,ep,D,eq)
Description:
plante provides an element stiffness matrix Ke and an element load vector fe for a
triangular element in plane strain or plane stress.
The element nodal coordinates x1 , y1 , x2 etc. are supplied to the function by ex and
ey. The type of analysis ptype and the element thickness t are supplied by ep,
and the material properties are supplied by the constitutive matrix D. Any arbitrary
D-matrix with dimensions from (3 × 3) to (6 × 6) may be given. For an isotropic
elastic material the constitutive matrix can be formed by the function hooke, see
Section 4.
ex = [ x1 x2 x3 ]
ep = [ ptype t ]
ey = [ y1 y2 y3 ]
D11 D12 D13 D14 [D15 ] [D16 ]
D21 D22 D23 D24 [D25 ] [D26 ]
D11 D12 D13
D31 D32 D33 D34
[D35 ] [D36 ]
D= D21 D22 D23 or D =
D41 D42 D43 D44 [D45 ] [D46 ]
D31 D32 D33
[D51 ] [D52 ] [D53 ] [D54 ] [D55 ] [D56 ]
[D61 ] [D62 ] [D63 ] [D64 ] [D65 ] [D66 ]
ELEMENT 5.5 – 4
Two dimensional solid elements plante
If uniformly distributed loads are applied to the element, the element load vector fe
is computed. The input variable
bx
eq =
by
Theory:
The element stiffness matrix Ke and the element load vector fel , stored in Ke and fe,
respectively, are computed according to
Ke = (C−1 )T B̄ D B̄ t dA C−1
T
A
fel = (C−1 )T
T
N̄ b t dA
A
with the constitutive matrix D defined by D, and the body force vector b defined by
eq.
The evaluation of the integrals for the triangular element is based on a linear dis-
placement approximation u(x, y) and is expressed in terms of the nodal variables u1 ,
u2 , . . . , u6 as
u(x, y) = Ne ae = N̄ C−1 ae
where
ux 1 x y 0 0 0
u= N̄ =
uy 0 0 0 1 x y
1 x1 y 1 0 0 0 u1
0 0 0 1 x1 y 1
u2
1 x2 y 2 0 0 0 u3
C= ae =
0 0 0 1 x2 y 2 u4
1 x3 y 3 0 0 0 u5
0 0 0 1 x3 y 3 u6
5.5 – 5 ELEMENT
plante Two dimensional solid elements
If a larger D-matrix than (3 × 3) is used for plane stress (ptype = 1), the D-matrix
is reduced to a (3 × 3) matrix by static condensation using σzz = σxz = σyz = 0.
These stress components are connected with the rows 3, 5 and 6 in the D-matrix
respectively.
If a larger D-matrix than (3 × 3) is used for plane strain (ptype = 2), the D-matrix
is reduced to a (3 × 3) matrix using εzz = γxz = γyz = 0. This implies that a
(3 × 3) D-matrix is created by the rows and the columns 1, 2 and 4 from the original
D-matrix.
Evaluation of the integrals for the triangular element yields
Ke = (C−1 )T B̄ D B̄ C−1 t A
T
At
fle = [ bx by bx by bx by ]T
3
ELEMENT 5.5 – 6
Two dimensional solid elements plants
Purpose:
Compute stresses and strains in a triangular element in plane strain or plane stress.
u6
u5 σyy
●
σxy
u4
y u3
u2 ●
σxx σxx
u1
● σxy
σyy
x
Syntax:
[es,et]=plants(ex,ey,ep,D,ed)
Description:
plants computes the stresses es and the strains et in a triangular element in plane
strain or plane stress.
The input variables ex, ey, ep and D are defined in plante. The vector ed contains
the nodal displacements ae of the element and is obtained by the function extract as
ed = (ae )T = [ u1 u2 . . . u6 ]
contain the stress and strain components. The size of es and et follows the size of D.
Note that for plane stress εzz = 0, and for plane strain σzz = 0.
Theory:
The strains and stresses are computed according to
ε = B̄ C−1 ae
σ= Dε
where the matrices D, B̄, C and ae are described in plante. Note that both the
strains and the stresses are constant in the element.
5.5 – 7 ELEMENT
plantf Two dimensional solid elements
Purpose:
Compute internal element force vector in a triangular element in plane strain or
plane stress.
Syntax:
ef=plantf(ex,ey,ep,es)
Description:
plantf computes the internal element forces ef in a triangular element in plane strain
or plane stress.
The input variables ex, ey and ep are defined in plante, and the input variable es is
defined in plants.
The output variable
Theory:
The internal force vector is computed according to
fie = (C−1 )T
T
B̄ σ t dA
A
where the matrices B̄ and C are defined in plante and σ is defined in plants.
fie = (C−1 )T B̄ σ t A
T
ELEMENT 5.5 – 8
Two dimensional solid elements planqe
Purpose:
Compute element matrices for a quadrilateral element in plane strain or plane stress.
u6
u8 3 u5
u7 ●
4
●
y u2
u1
● u4
(x1,y1) u3
●
2
x
Syntax:
Ke=planqe(ex,ey,ep,D)
[Ke,fe]=planqe(ex,ey,ep,D,eq)
Description:
planqe provides an element stiffness matrix Ke and an element load vector fe for a
quadrilateral element in plane strain or plane stress.
The element nodal coordinates x1 , y1 , x2 etc. are supplied to the function by ex and
ey. The type of analysis ptype and the element thickness t are supplied by ep,
and the material properties are supplied by the constitutive matrix D. Any arbitrary
D-matrix with dimensions from (3 × 3) to (6 × 6) may be given. For an isotropic
elastic material the constitutive matrix can be formed by the function hooke, see
Section 4.
ex = [ x1 x2 x3 x4 ]
ep = [ ptype t ]
ey = [ y1 y2 y3 y4 ]
D11 D12 D13 D14 [D15 ] [D16 ]
D21 D22 D23 D24 [D25 ] [D26 ]
D11 D12 D13
D31 D32 D33 D34
[D35 ] [D36 ]
D = D21 D22 D23 or D =
D41 D42 D43 D44 [D45 ] [D46 ]
D31 D32 D33
[D51 ] [D52 ] [D53 ] [D54 ] [D55 ] [D56 ]
[D61 ] [D62 ] [D63 ] [D64 ] [D65 ] [D66 ]
5.5 – 9 ELEMENT
planqe Two dimensional solid elements
If uniformly distributed loads are applied on the element, the element load vector fe
is computed. The input variable
bx
eq =
by
Theory:
In computing the element matrices, two more degrees of freedom are introduced.
The location of these two degrees of freedom is defined by the mean value of the
coordinates at the corner points. Four sets of element matrices are calculated using
plante. These matrices are then assembled and the two extra degrees of freedom are
eliminated by static condensation.
ELEMENT 5.5 – 10
Two dimensional solid elements planqs
Purpose:
Compute stresses and strains in a quadrilateral element in plane strain or plane
stress.
u6
u8 u5
● σyy
u7
●
σxy
y u2
u1 σxx σxx
● u4
u3 σxy
●
x σyy
Syntax:
[es,et]=planqs(ex,ey,ep,D,ed)
[es,et]=planqs(ex,ey,ep,D,ed,eq)
Description:
planqs computes the stresses es and the strains et in a quadrilateral element in plane
strain or plane stress.
The input variables ex, ey, ep, D and eq are defined in planqe. The vector ed contains
the nodal displacements ae of the element and is obtained by the function extract as
ed = (ae )T = [ u1 u2 . . . u8 ]
If body forces are applied to the element the variable eq must be included.
The output variables
contain the stress and strain components. The size of es and et follows the size of D.
Note that for plane stress εzz = 0, and for plane strain σzz = 0.
Theory:
By assembling triangular elements as described in planqe a system of equations con-
taining 10 degrees of freedom is obtained. From this system of equations the two
unknown displacements at the center of the element are computed. Then according
to the description in plants the strain and stress components in each of the four trian-
gular elements are produced. Finally the quadrilateral element strains and stresses
are computed as area weighted mean values from the values of the four triangular
elements. If uniformly distributed loads are applied on the element, the element load
vector eq is needed for the calculations.
5.5 – 11 ELEMENT
planre Two dimensional solid elements
Purpose:
Compute element matrices for a rectangular (Melosh) element in plane strain or
plane stress.
u8 u6
u7 u5
● ●
(x4,y4) (x3,y3)
y u2 u4
u1 u3
● ●
(x1,y1) (x2,y2)
Syntax:
Ke=planre(ex,ey,ep,D)
[Ke,fe]=planre(ex,ey,ep,D,eq)
Description:
planre provides an element stiffness matrix Ke and an element load vector fe for a
rectangular (Melosh) element in plane strain or plane stress. This element can only
be used if the element edges are parallel to the coordinate axis.
The element nodal coordinates (x1 , y1 ) and (x3 , y3 ) are supplied to the function by
ex and ey. The type of analysis ptype and the element thickness t are supplied by ep,
and the material properties are supplied by the constitutive matrix D. Any arbitrary
D-matrix with dimensions from (3 × 3) to (6 × 6) may be given. For an isotropic
elastic material the constitutive matrix can be formed by the function hooke, see
Section 4.
ex = [ x1 x3 ]
ep = [ ptype t ]
ey = [ y1 y3 ]
D11 D12 D13 D14 [D15 ] [D16 ]
D21 D22 D23 D24 [D25 ] [D26 ]
D11 D12 D13
D31 D32 D33 D34
[D35 ] [D36 ]
D = D21 D22 D23 or D =
D41 D42 D43 D44 [D45 ] [D46 ]
D31 D32 D33
[D51 ] [D52 ] [D53 ] [D54 ] [D55 ] [D56 ]
[D61 ] [D62 ] [D63 ] [D64 ] [D65 ] [D66 ]
ELEMENT 5.5 – 12
Two dimensional solid elements planre
If uniformly distributed loads are applied on the element, the element load vector fe
is computed. The input variable
bx
eq =
by
with the constitutive matrix D defined by D, and the body force vector b defined by
eq.
The evaluation of the integrals for the rectangular element is based on a bilinear
displacement approximation u(x, y) and is expressed in terms of the nodal variables
u1 , u2 , . . ., u8 as
u(x, y) = Ne ae
where
u1
ux N1e 0 N2e 0 N3e 0 N4e 0 u2
u= Ne = ae =
..
uy 0 N1e 0 N2e 0 N3e 0 N4e .
u8
With a local coordinate system located at the center of the element, the element
shape functions N1e − N4e are obtained as
1
N1e = (x − x2 )(y − y4 )
4ab
1
N2e = − (x − x1 )(y − y3 )
4ab
1
N3e = (x − x4 )(y − y2 )
4ab
1
N4e = − (x − x3 )(y − y1 )
4ab
where
1 1
a = (x3 − x1 ) and b = (y3 − y1 )
2 2
5.5 – 13 ELEMENT
planre Two dimensional solid elements
If a larger D-matrix than (3 × 3) is used for plane stress (ptype = 1), the D-matrix
is reduced to a (3 × 3) matrix by static condensation using σzz = σxz = σyz = 0.
These stress components are connected with the rows 3, 5 and 6 in the D-matrix
respectively.
If a larger D-matrix than (3 × 3) is used for plane strain (ptype = 2), the D-matrix
is reduced to a (3 × 3) matrix using εzz = γxz = γyz = 0. This implies that a
(3 × 3) D-matrix is created by the rows and the columns 1, 2 and 4 from the original
D-matrix.
Evaluation of the integrals for the rectangular element can be done either analytically
or numerically by use of a 2 × 2 point Gauss integration. The element load vector fle
yields
bx
by
bx
by
fle = abt
bx
by
bx
by
ELEMENT 5.5 – 14
Two dimensional solid elements planrs
Purpose:
Compute stresses and strains in a rectangular (Melosh) element in plane strain or
plane stress.
u8 u6
u7 u5
● ● σyy
σxy
y u2 u4 σxx σxx
u1 u3 σxy
● ●
σyy
x
Syntax:
[es,et]=planrs(ex,ey,ep,D,ed)
Description:
planrs computes the stresses es and the strains et in a rectangular (Melosh) element
in plane strain or plane stress. The stress and strain components are computed at
the center of the element.
The input variables ex, ey, ep and D are defined in planre. The vector ed contains
the nodal displacements ae of the element and is obtained by the function extract as
ed = (ae )T = [ u1 u2 . . . u8 ]
contain the stress and strain components. The size of es and et follows the size of D.
Note that for plane stress εzz = 0, and for plane strain σzz = 0.
Theory:
The strains and stresses are computed according to
ε = Be ae
σ=Dε
where the matrices D, Be , and ae are described in planre, and where the evaluation
point (x, y) is chosen to be at the center of the element.
5.5 – 15 ELEMENT
plantce Two dimensional solid elements
Purpose:
Compute element matrices for a rectangular (Turner-Clough) element in plane strain
or plane stress.
u8 u6
u7 u5
● ●
(x4,y4) (x3,y3)
y u2 u4
u1 u3
● ●
(x1,y1) (x2,y2)
Syntax:
Ke=plantce(ex,ey,ep)
[Ke,fe]=plantce(ex,ey,ep,eq)
Description:
plantce provides an element stiffness matrix Ke and an element load vector fe for a
rectangular (Turner-Clough) element in plane strain or plane stress. This element
can only be used if the material is isotropic and if the element edges are parallel to
the coordinate axis.
The element nodal coordinates (x1 , y1 ) and (x3 , y3 ) are supplied to the function by ex
and ey. The state of stress ptype, the element thickness t and the material properties
E and ν are supplied by ep. For plane stress ptype = 1 and for plane strain ptype = 2.
ex = [ x1 x3 ]
ep = [ ptype t E ν ]
ey = [ y1 y3 ]
If uniformly distributed loads are applied to the element, the element load vector fe
is computed. The input variable
bx
eq =
by
ELEMENT 5.5 – 16
Two dimensional solid elements plantce
Theory:
The element stiffness matrix Ke and the element load vector fel , stored in Ke and fe,
respectively, are computed according to
Ke = BeT D Be t dA
A
fel = NeT b t dA
A
where the constitutive matrix D is described in hooke, see Section 4, and the body
force vector b is defined by eq.
The evaluation of the integrals for the Turner-Clough element is based on a dis-
placement field u(x, y) built up of a bilinear displacement approximation superposed
by bubble functions in order to create a linear stress field over the element. The
displacement field is expressed in terms of the nodal variables u1 , u2 , . . ., u8 as
u(x, y) = Ne ae
where
u1
ux N1e N5e N2e N5e N3e N5e N4e N5e u2
u= Ne = ae =
..
uy N6e N1e N6e N2e N6e N3e N6e N4e .
u8
With a local coordinate system located at the center of the element, the element
shape functions N1e − N6e are obtained as
1
N1e = (a − x)(b − y)
4ab
1
N2e = (a + x)(b − y)
4ab
1
N3e = (a + x)(b + y)
4ab
1
N4e = (a − x)(b + y)
4ab
1
N5e = (b2 − y 2 ) + ν(a2 − x2 )
8ab
1
N6e = (a2 − x2 ) + ν(b2 − y 2 )
8ab
where
1 1
a = (x3 − x1 ) and b = (y3 − y1 )
2 2
5.5 – 17 ELEMENT
plantce Two dimensional solid elements
Evaluation of the integrals for the Turner-Clough element can be done either ana-
lytically or numerically by use of a 2 × 2 point Gauss integration. The element load
vector fle yields
bx
by
bx
by
fle = abt
bx
by
bx
by
ELEMENT 5.5 – 18
Two dimensional solid elements plantcs
Purpose:
Compute stresses and strains in a Turner-Clough element in plane strain or plane
stress.
u8 u6
u7 u5
● ● σyy
σxy
y u2 u4 σxx σxx
u1 u3 σxy
● ●
σyy
x
Syntax:
[es,et]=plantcs(ex,ey,ep,ed)
Description:
plantcs computes the stresses es and the strains et in a rectangular Turner-Clough ele-
ment in plane strain or plane stress. The stress and strain components are computed
at the center of the element.
The input variables ex, ey, and ep are defined in plantce. The vector ed contains the
nodal displacements ae of the element and is obtained by the function extract as
ed = (ae )T = [ u1 u2 . . . u8 ]
contain the stress and strain components. The size of es and et follows the size of D.
Note that for plane stress εzz = 0, and for plane strain σzz = 0.
Theory:
The strains and stresses are computed according to
ε = Be ae
σ=Dε
where the matrices D, Be , and ae are described in plantce, and where the evaluation
point (x, y) is chosen to be at the center of the element.
5.5 – 19 ELEMENT
plani4e Two dimensional solid elements
Purpose:
Compute element matrices for a 4 node isoparametric element in plane strain or
plane stress.
u6
u8 3 u5
u7 ●
4
●
y u2
u1
● u4
(x1,y1) u3
●
2
x
Syntax:
Ke=plani4e(ex,ey,ep,D)
[Ke,fe]=plani4e(ex,ey,ep,D,eq)
Description:
plani4e provides an element stiffness matrix Ke and an element load vector fe for a 4
node isoparametric element in plane strain or plane stress.
The element nodal coordinates x1 , y1 , x2 etc. are supplied to the function by ex and
ey. The type of analysis ptype, the element thickness t, and the number of Gauss
points n are supplied by ep.
The material properties are supplied by the constitutive matrix D. Any arbitrary D-
matrix with dimensions from (3 × 3) to (6 × 6) maybe given. For an isotropic elastic
material the constitutive matrix can be formed by the function hooke, see Section 4.
ex = [ x1 x2 x3 x4 ]
ep = [ ptype t n ]
ey = [ y1 y2 y3 y4 ]
D11 D12 D13 D14 [D15 ] [D16 ]
D21 D22 D23 D24 [D25 ] [D26 ]
D11 D12 D13
D31 D32 D33 D34
[D35 ] [D36 ]
D = D21 D22 D23 or D =
D41 D42 D43 D44 [D45 ] [D46 ]
D31 D32 D33
[D51 ] [D52 ] [D53 ] [D54 ] [D55 ] [D56 ]
[D61 ] [D62 ] [D63 ] [D64 ] [D65 ] [D66 ]
ELEMENT 5.5 – 20
Two dimensional solid elements plani4e
If different Di -matrices are used in the Gauss points these Di -matrices are stored
in a global vector D. For numbering of the Gauss points, see eci in plani4s.
D1
D2
D=
..
.
Dn2
If uniformly distributed loads are applied to the element, the element load vector fe
is computed. The input variable
bx
eq =
by
Theory:
The element stiffness matrix Ke and the element load vector fel , stored in Ke and fe,
respectively, are computed according to
Ke = BeT D Be t dA
A
fel = NeT b t dA
A
with the constitutive matrix D defined by D, and the body force vector b defined by
eq.
The evaluation of the integrals for the isoparametric 4 node element is based on a
displacement approximation u(ξ, η), expressed in a local coordinates system in terms
of the nodal variables u1 , u2 , . . ., u8 as
u(ξ, η) = Ne ae
where
u1
ux N1e 0 N2e 0 N3e 0 N4e 0 u2
u= Ne = ae =
..
uy 0 N1e 0 N2e 0 N3e 0 N4e .
u8
1 1
N1e = (1 − ξ)(1 − η) N2e = (1 + ξ)(1 − η)
4 4
1 1
e
N3 = (1 + ξ)(1 + η) N4 = (1 − ξ)(1 + η)
e
4 4
5.5 – 21 ELEMENT
plani4e Two dimensional solid elements
and where
∂ ∂ ∂x ∂x
∂x ∂ξ ∂ξ ∂η
= (JT )−1 J=
∂ ∂ ∂y ∂y
∂y ∂η ∂ξ ∂η
If a larger D-matrix than (3 × 3) is used for plane stress (ptype = 1), the D-matrix
is reduced to a (3 × 3) matrix by static condensation using σzz = σxz = σyz = 0.
These stress components are connected with the rows 3, 5 and 6 in the D-matrix
respectively.
If a larger D-matrix than (3 × 3) is used for plane strain (ptype = 2), the D-matrix
is reduced to a (3 × 3) matrix using εzz = γxz = γyz = 0. This implies that a
(3 × 3) D-matrix is created by the rows and the columns 1, 2 and 4 from the original
D-matrix.
ELEMENT 5.5 – 22
Two dimensional solid elements plani4s
Purpose:
Compute stresses and strains in a 4 node isoparametric element in plane strain or
plane stress.
u6
u8 u5
● σyy
u7
●
σxy
y u2
u1 σxx σxx
● u4
u3 σxy
●
x σyy
Syntax:
[es,et,eci]=plani4s(ex,ey,ep,D,ed)
Description:
plani4s computes stresses es and the strains et in a 4 node isoparametric element in
plane strain or plane stress.
The input variables ex, ey, ep and the matrix D are defined in plani4e. The vector ed
contains the nodal displacements ae of the element and is obtained by the function
extract as
ed = (ae )T = [ u1 u2 . . . u8 ]
ε1 ε1yy [ε1zz ] 1
γxy 1
[γxz ] 1
γyz
xx x1 y1
ε2 2
[ε2zz ] 2 2
γyz
2 x2 y2
xx εyy γxy [γxz ]
et = εT =
.. .. .. .. .. ..
eci =
.. ..
. . . . . . . .
2
εnxx εnyy
2 n2
εzz n2
γxy n2
γxz n2
γyz xn2 yn2
contain the stress and strain components, and the coordinates of the integration
points. The index n denotes the number of integration points used within the ele-
ment, cf. plani4e. The number of columns in es and et follows the size of D. Note
that for plane stress εzz = 0, and for plane strain σzz = 0.
5.5 – 23 ELEMENT
plani4s Two dimensional solid elements
Theory:
The strains and stresses are computed according to
ε = Be ae
σ=Dε
where the matrices D, Be , and ae are described in plani4e, and where the integration
points are chosen as evaluation points.
ELEMENT 5.5 – 24
Two dimensional solid elements plani4f
Purpose:
Compute internal element force vector in a 4 node isoparametric element in plane
strain or plane stress.
Syntax:
ef=plani4f(ex,ey,ep,es)
Description:
plani4f computes the internal element forces ef in a 4 node isoparametric element in
plane strain or plane stress.
The input variables ex, ey and ep are defined in plani4e, and the input variable es is
defined in plani4s.
The output variable
Theory:
The internal force vector is computed according to
fie = BeT σ t dA
A
where the matrices Be and σ are defined in plani4e and plani4s, respectively.
Evaluation of the integral is done by Gauss integration.
5.5 – 25 ELEMENT
plani8e Two dimensional solid elements
Purpose:
Compute element matrices for an 8 node isoparametric element in plane strain or
plane stress.
7 3
●
●
4
●
6
8 ●
y ●
u2
1 5
● ●
u1 2
(x1,y1) ●
x
Syntax:
Ke=plani8e(ex,ey,ep,D)
[Ke,fe]=plani8e(ex,ey,ep,D,eq)
Description:
plani8e provides an element stiffness matrix Ke and an element load vector fe for an
8 node isoparametric element in plane strain or plane stress.
The element nodal coordinates x1 , y1 , x2 etc. are supplied to the function by ex and
ey. The type of analysis ptype, the element thickness t, and the number of Gauss
points n are supplied by ep.
The material properties are supplied by the constitutive matrix D. Any arbitrary
D-matrix with dimensions from (3 × 3) to (6 × 6) may be given. For an isotropic
elastic material the constitutive matrix can be formed by the function hooke, see
Section 4.
ex = [ x1 x2 . . . x8 ]
ep = [ ptype t n ]
ey = [ y1 y2 . . . y8 ]
D11 D12 D13 D14 [D15 ] [D16 ]
D21 D22 D23 D24 [D25 ] [D26 ]
D11 D12 D13
D31 D32 D33 D34
[D35 ] [D36 ]
D = D21 D22 D23 or D =
D41 D42 D43 D44 [D45 ] [D46 ]
D31 D32 D33
[D51 ] [D52 ] [D53 ] [D54 ] [D55 ] [D56 ]
[D61 ] [D62 ] [D63 ] [D64 ] [D65 ] [D66 ]
ELEMENT 5.5 – 26
Two dimensional solid elements plani8e
If different Di -matrices are used in the Gauss points these Di -matrices are stored
in a global vector D. For numbering of the Gauss points, see eci in plani8s.
D1
D2
D=
..
.
Dn2
If uniformly distributed loads are applied to the element, the element load vector fe
is computed. The input variable
bx
eq =
by
fel = NeT b t dA
A
with the constitutive matrix D defined by D, and the body force vector b defined by
eq.
The evaluation of the integrals for the isoparametric 8 node element is based on a
displacement approximation u(ξ, η), expressed in a local coordinates system in terms
of the nodal variables u1 , u2 , . . ., u16 as
u(ξ, η) = Ne ae
where
u1
ux N1e 0 N2e 0 . . . N8e 0 u2
u= Ne = ae =
..
uy 0 N1e 0 N2e . . . 0 N8e .
u16
5.5 – 27 ELEMENT
plani8e Two dimensional solid elements
and where
∂ ∂ ∂x ∂x
∂x ∂ξ ∂ξ ∂η
= (JT )−1 J=
∂ ∂ ∂y ∂y
∂y ∂η ∂ξ ∂η
If a larger D-matrix than (3 × 3) is used for plane stress (ptype = 1), the D-matrix
is reduced to a (3 × 3) matrix by static condensation using σzz = σxz = σyz = 0.
These stress components are connected with the rows 3, 5 and 6 in the D-matrix
respectively.
If a larger D-matrix than (3 × 3) is used for plane strain (ptype = 2), the D-matrix
is reduced to a (3 × 3) matrix using εzz = γxz = γyz = 0. This implies that a
(3 × 3) D-matrix is created by the rows and the columns 1, 2 and 4 from the original
D-matrix.
ELEMENT 5.5 – 28
Two dimensional solid elements plani8s
Purpose:
Compute stresses and strains in an 8 node isoparametric element in plane strain or
plane stress.
3 σyy
7 ●
●
4 σxy
●
6
8 ●
y ●
u2 σxx σxx
1 5
● ●
σxy
u1 2
●
σyy
x
Syntax:
[es,et,eci]=plani8s(ex,ey,ep,D,ed)
Description:
plani8s computes stresses es and the strains et in an 8 node isoparametric element in
plane strain or plane stress.
The input variables ex, ey, ep and the matrix D are defined in plani8e. The vector ed
contains the nodal displacements ae of the element and is obtained by the function
extract as
ed = (ae )T = [ u1 u2 . . . u16 ]
ε1xx ε1yy [ε1zz ] 1
γxy 1
[γxz ] 1
γyz
x1 y1
ε2xx ε2yy [ε2zz ] 2
γxy 2
[γxz ] γyz
2 x2 y2
et = εT =
.. eci = .. ..
.. .. .. .. .. . .
. . . . . .
εnxx
2
εnyy
2
εnzz
2 n
γxy
2 n
γxz
2 n
γyz
2 xn2 yn2
contain the stress and strain components, and the coordinates of the integration
points. The index n denotes the number of integration points used within the ele-
ment, cf. plani8e. The number of columns in es and et follows the size of D. Note
that for plane stress εzz = 0, and for plane strain σzz = 0.
5.5 – 29 ELEMENT
plani8s Two dimensional solid elements
Theory:
The strains and stresses are computed according to
ε = Be ae
σ=Dε
where the matrices D, Be , and ae are described in plani8e, and where the integration
points are chosen as evaluation points.
ELEMENT 5.5 – 30
Two dimensional solid elements plani8f
Purpose:
Compute internal element force vector in an 8 node isoparametric element in plane
strain or plane stress.
Syntax:
ef=plani8f(ex,ey,ep,es)
Description:
plani8f computes the internal element forces ef in an 8 node isoparametric element
in plane strain or plane stress.
The input variables ex, ey and ep are defined in plani8e, and the input variable es is
defined in plani8s.
The output variable
Theory:
The internal force vector is computed according to
fie = BeT σ t dA
A
where the matrices Be and σ are defined in plani8e and plani8s, respectively.
Evaluation of the integral is done by Gauss integration.
5.5 – 31 ELEMENT
soli8e Three dimensional solid elements
Purpose:
Compute element matrices for an 8 node isoparametric solid element.
4
8 ● 3
● ●
7
u2 ●
y
u1
u3 ●
●
(x1,y1,z1)
● 2
5 ●
6
x
z
Syntax:
Ke=soli8e(ex,ey,ez,ep,D)
[Ke,fe]=soli8e(ex,ey,ez,ep,D,eq)
Description:
soli8e provides an element stiffness matrix Ke and an element load vector fe for an 8
node isoparametric solid element.
The element nodal coordinates x1 , y1 , z1 , x2 etc. are supplied to the function by ex,
ey and ez, and the number of Gauss points n are supplied by ep.
(n × n) integration points, n = 1, 2, 3
The material properties are supplied by the constitutive matrix D. Any arbitrary
D-matrix with dimensions (6 × 6) may be given. For an isotropic elastic material the
constitutive matrix can be formed by the function hooke, see Section 4.
D11 D12 · · · D16
ex = [ x1 x2 . . . x8 ]
D21 D22 · · · D26
ey = [ y1 y2 . . . y8 ] ep = [ n ] D= .
. .. .. ..
ez = [ z1 z2 . . . z8 ] . . . .
D61 D62 · · · D66
If different Di -matrices are used in the Gauss points these Di -matrices are stored
in a global vector D. For numbering of the Gauss points, see eci in soli8s.
D1
D2
D=
..
.
Dn3
ELEMENT 5.5 – 32
Three dimensional solid elements soli8e
If uniformly distributed loads are applied to the element, the element load vector fe
is computed. The input variable
bx
eq =
by
bz
Theory:
The element stiffness matrix Ke and the element load vector fel , stored in Ke and fe,
respectively, are computed according to
Ke = BeT D Be dV
V
fel = NeT b dV
V
with the constitutive matrix D defined by D, and the body force vector b defined by
eq.
The evaluation of the integrals for the isoparametric 8 node solid element is based
on a displacement approximation u(ξ, η, ζ), expressed in a local coordinates system
in terms of the nodal variables u1 , u2 , . . ., u24 as
u(ξ, η, ζ) = Ne ae
where
u1
ux Ne 0 0 N2e 0 0 . . . N8e 0 0
1 u2
u=
uy N = 0 N1e 0
e
0 N2e 0 . . . 0 N8e 0
ae =
..
uz 0 0 N1 0 0 N2 . . . 0 0 N8
e e e .
u24
1 1
N1e = (1 − ξ)(1 − η)(1 − ζ) N5e = (1 − ξ)(1 − η)(1 + ζ)
8 8
1 1
N2e = (1 + ξ)(1 − η)(1 − ζ) N6 = (1 + ξ)(1 − η)(1 + ζ)
e
8 8
1 1
N3 = (1 + ξ)(1 + η)(1 − ζ)
e e
N7 = (1 + ξ)(1 + η)(1 + ζ)
8 8
1 1
N4 = (1 − ξ)(1 + η)(1 − ζ)
e
N8 = (1 − ξ)(1 + η)(1 + ζ)
e
8 8
Be = ∇N
˜ e
5.5 – 33 ELEMENT
soli8e Three dimensional solid elements
where
∂
0 0
∂x
∂
0 0
∂y ∂ ∂
∂ ∂ξ
∂x
0 0
∂z ∂
T −1 ∂
∇
˜ = = (J )
∂ ∂
∂y ∂η
0
∂y ∂x ∂ ∂
∂ ∂ ∂z ∂ζ
0
∂z ∂x
∂ ∂
0
∂z ∂y
∂x ∂x ∂x
∂ξ ∂η ∂ζ
∂y ∂y ∂y
J=
∂ξ ∂η ∂ζ
∂z ∂z ∂z
∂ξ ∂η ∂ζ
ELEMENT 5.5 – 34
Three dimensional solid elements soli8s
Purpose:
Compute stresses and strains in an 8 node isoparametric solid element.
σyy
4
3 σyx
8 ●
●
σyz
●
7 σxy
y u1 ● σzy σxx
u2 σzx
u3 ●
● σzz σxz
(x1,y1,z1)
● 2
5 ●
6
x
z
Syntax:
[es,et,eci]=soli8s(ex,ey,ez,ep,D,ed)
Description:
soli8s computes stresses es and the strains et in an 8 node isoparametric solid element.
The input variables ex, ey, ez, ep and the matrix D are defined in soli8e. The vector
ed contains the nodal displacements ae of the element and is obtained by the function
extract as
ed = (ae )T = [ u1 u2 . . . u24 ]
ε1xx ε1yy ε1zz 1
γxy 1
γxz 1
γyz x1 y1 z1
ε2xx ε2yy ε2zz 2
γxy 2
γxz 2
γyz x2 y2 z2
et = ε = T
.. .. .. .. .. .. eci = .. .. ..
. . .
. . . . . .
3 3 3 3 3 3
εnxx εnyy εnzz n
γxy n
γxz n
γyz xn3 yn3 zn3
contain the stress and strain components, and the coordinates of the integration
points. The index n denotes the number of integration points used within the ele-
ment, cf. soli8e.
5.5 – 35 ELEMENT
soli8s Three dimensional solid elements
Theory:
The strains and stresses are computed according to
ε = Be ae
σ=Dε
where the matrices D, Be , and ae are described in soli8e, and where the integration
points are chosen as evaluation points.
ELEMENT 5.5 – 36
Three dimensional solid elements soli8f
Purpose:
Compute internal element force vector in an 8 node isoparametric solid element.
Syntax:
ef=soli8f(ex,ey,ez,ep,es)
Description:
soli8f computes the internal element forces ef in an 8 node isoparametric solid element.
The input variables ex, ey, ez and ep are defined in soli8e, and the input variable es
is defined in soli8s.
The output variable
Theory:
The internal force vector is computed according to
fie = BeT σ dV
V
where the matrices B and σ are defined in soli8e and soli8s, respectively.
5.5 – 37 ELEMENT
5.6 Beam elements
Beam elements are available for two, and three dimensional linear static analysis. Two
dimensional beam elements for nonlinear geometric and dynamic analysis are also available.
Beam elements
u11
u5
u4 u8
u6
u7 u10
u9
u5 u12
u2
u2
u3 u1 u1 u4
u3
u6
beam2e
beam2t
beam2w beam3e
beam2g
beam2d
2D beam elements
beam2e Compute element matrices
beam2s Compute section forces
beam2t Compute element matrices for Timoshenko beam element
beam2ts Compute section forces for Timoshenko beam element
beam2w Compute element matrices for beam element on elastic foundation
beam2ws Compute section forces for beam element on elastic foundation
beam2g Compute element matrices with respect to geometric nonlinearity
beam2gs Compute section forces for geometric nonlinear beam element
beam2d Compute element matrices, dynamic analysis
beam2ds Compute section forces, dynamic analysis
3D beam elements
beam3e Compute element matrices
beam3s Compute section forces
5.6 – 1 ELEMENT
beam2e Two dimensional beam element
Purpose:
Compute element stiffness matrix for a two dimensional beam element.
u5
u6 u4
(x2,y2)
y u2 x E, A, I
u3 u1
(x1,y1)
Syntax:
Ke=beam2e(ex,ey,ep)
[Ke,fe]=beam2e(ex,ey,ep,eq)
Description:
beam2e provides the global element stiffness matrix Ke for a two dimensional beam
element.
The input variables
ex = [ x1 x2 ]
ep = [ E A I ]
ey = [ y1 y2 ]
eq = qx̄ qȳ
then contains the distributed loads per unit length, qx̄ and qȳ .
ELEMENT 5.6 – 2
Two dimensional beam element beam2e
qy
qx
y
Theory:
The element stiffness matrix Ke , stored in Ke, is computed according to
Ke = GT K̄e G
where
EA
L
0 0 − EA
L
0 0
12EI 6EI
0 L3 L2
0 − 12EI
L3
6EI
L2
6EI 4EI 2EI
0 L2 L
0 − 6EI
L2 L
e
K̄ =
− EA 0 0 EA
0 0
L L
0 − 12EI − 6EI 0 12EI
− 6EI
L3 L2 L3 L2
6EI 2EI
0 L2 L
0 − 6EI
L2
4EI
L
nxx̄ nyx̄ 0 0 0 0
nxȳ nyȳ 0 0 0 0
0 0 1 0 0 0
G=
0 0 0 nxx̄ nyx̄ 0
0 0 0 nxȳ nyȳ 0
0 0 0 0 0 1
5.6 – 3 ELEMENT
beam2e Two dimensional beam element
fle = GT f̄le
where
qx̄ L
2
qȳ L
2
qȳ L2
12
f̄le =
qx̄ L
2
qȳ L
2
qȳ L2
−
12
ELEMENT 5.6 – 4
Two dimensional beam element beam2s
Purpose:
Compute section forces in a two dimensional beam element.
Vn
Mn Nn
y V
M1 M
M N
N1 V1 V
N
x
Syntax:
es=beam2s(ex,ey,ep,ed)
es=beam2s(ex,ey,ep,ed,eq)
[es,edi,eci]=beam2s(ex,ey,ep,ed,eq,n)
Description:
beam2s computes the section forces and displacements in local directions along the
beam element beam2e.
The input variables ex, ey, ep and eq are defined in beam2e, and the element displace-
ments, stored in ed, are obtained by the function extract. If distributed loads are
applied to the element, the variable eq must be included. The number of evaluation
points for section forces and displacements are determined by n. If n is omitted, only
the ends of the beam are evaluated.
The output variables
es = [ N V M ] edi = [ ū v̄ ] eci = [ x̄ ]
consist of column matrices that contain the section forces, the displacements, and
the evaluation points on the local x-axis. The explicit matrix expressions are
0
N1 V1 M1 ū1 v̄1
x̄2
N2 V2 M2
ū2 v̄2
..
es = .. .. .. edi = .. .. eci =
.
. . . . .
x̄n−1
Nn Vn Mn ūn v̄n
L
5.6 – 5 ELEMENT
beam2s Two dimensional beam element
Theory:
The evaluation of the section forces is based on the solutions of the basic equations
d2 ū d4 v̄
EA + qx̄ = 0 EI − qȳ = 0
dx̄2 dx̄4
From these equations, the displacements along the beam element are obtained as the
sum of the homogeneous and the particular solutions
ū(x̄)
u= = uh + up
v̄(x̄)
where
qx̄ Lx̄ x̄
(1 − )
ūp (x̄) 2EA L
uh = N̄ C−1 G ae up = =
v̄p (x̄) qȳ L2 x̄2 x̄
(1 − )2
24EI L
and
1 0 0 0 0 0
u1
0 0 1 0 0 0
1 x̄ 0 0 0 0 0 0 0 1 0 0 u2
N̄ = C= ae = ..
0 0 1 x̄ x̄2 x̄3 1 L 0 0 0 0
.
0 0 1 L L 2 L3 u6
0 0 0 1 2L 3L2
dū d3 v̄ d2 v̄
N = EA V = −EI M = EI
dx̄ dx̄3 dx̄2
Examples:
Section forces or element displacements can easily be plotted. The bending moment
M along the beam is plotted by
>> plot(eci,es(:,3))
ELEMENT 5.6 – 6
Two dimensional Timoshenko beam element beam2t
Purpose:
Compute element stiffness matrix for a two dimensional Timoshenko beam element.
u5
u6 u4
(x2,y2)
y u2 x E,
E,A,
G,I A, I, ks
u3 u1
(x1,y1)
Syntax:
Ke=beam2t(ex,ey,ep)
[Ke,fe]=beam2t(ex,ey,ep,eq)
Description:
beam2t provides the global element stiffness matrix Ke for a two dimensional Timo-
shenko beam element.
The input variables
ex = [ x1 x2 ]
ep = [ E G A I ks ]
ey = [ y1 y2 ]
eq = qx̄ qȳ
then contains the distributed loads per unit length, qx̄ and qȳ .
5.6 – 7 ELEMENT
beam2t Two dimensional Timoshenko beam element
qy
qx
y
Theory:
The element stiffness matrix Ke , stored in Ke, is computed according to
Ke = GT K̄e G
with
12EI
µ=
L2 GAks
ELEMENT 5.6 – 8
Two dimensional Timoshenko beam element beam2ts
Purpose:
Compute section forces in a two dimensional Timoshenko beam element.
Vn
Mn Nn
y V
M1 M
M N
N1 V1 V
N
x
Syntax:
es=beam2ts(ex,ey,ep,ed)
es=beam2ts(ex,ey,ep,ed,eq)
[es,edi,eci]=beam2ts(ex,ey,ep,ed,eq,n)
Description:
beam2ts computes the section forces and displacements in local directions along the
beam element beam2t.
The input variables ex, ey, ep and eq are defined in beam2t. The element displace-
ments, stored in ed, are obtained by the function extract. If distributed loads are
applied to the element, the variable eq must be included. The number of evaluation
points for section forces and displacements are determined by n. If n is omitted, only
the ends of the beam are evaluated.
The output variables
consist of column matrices that contain the section forces, the displacements and
dv̄
rotation of the cross section (note that the rotation θ is not equal to dx̄ ), and the
evaluation points on the local x-axis. The explicit matrix expressions are
0
N1 V1 M1 ū1 v̄1 θ1
x̄2
N2 V2 M2
ū2 v̄2 θ2
..
es = .. .. .. edi = .. .. .. eci =
.
. . . . . .
x̄n−1
Nn Vn Mn ūn v̄n θn
L
5.6 – 9 ELEMENT
beam2ts Two dimensional Timoshenko beam element
Theory:
The evaluation of the section forces is based on the solutions of the basic equations
d2 ū d3 θ d4 v̄
EA + qx̄ = 0 EI − qȳ = 0 EI − qȳ = 0
dx̄2 dx̄3 dx̄4
(The equations are valid if qȳ is not more than a linear function of x̄). From these
equations, the displacements along the beam element are obtained as the sum of the
homogeneous and the particular solutions
ū(x̄)
u = v̄(x̄) = uh + up
θ(x̄)
where
qx̄ Lx̄ x̄
(1 − )
2EA L
ūp (x̄)
2 2
qȳ L x̄ x̄ 2 qȳ Lx̄ x̄
uh = N̄ C−1 G ae up = p
v̄ (x̄) = (1 − ) + (1 − )
24EI L 2GA ks L
θp (x̄)
qȳ L2 x̄ 2x̄ x̄
(1 − )(1 − )
12EI L L
and
1 x̄ 0 0 0 0
EI
N̄ =
0 0 1 x̄ x̄2 x̄3
α=
2 GA ks
0 0 0 1 2x 3(x + 2α)
1 0 0 0 0 0
u1
0 0 1 0 0 0
0 0 0 1 0 6α u2
C= ae = ..
1 L 0 0 0 0
.
0 0 1 L L2 L3 u6
0 0 0 1 2L 3(L2 + 2α)
dū dv̄ dθ
N = EA V = GA ks ( − θ) M = EI
dx̄ dx̄ dx̄
ELEMENT 5.6 – 10
Two dimensional beam element beam2w
Purpose:
Compute element stiffness matrix for a two dimensional beam element on elastic
foundation.
kt
ka
Syntax:
Ke=beam2w(ex,ey,ep)
[Ke,fe]=beam2w(ex,ey,ep,eq)
Description:
beam2w provides the global element stiffness matrix Ke for a two dimensional beam
element on elastic foundation.
The input variables ex and ey are described in beam2e, and
ep = [ E A I ka kt ]
contains the modulus of elasticity E, the cross section area A, the moment of inertia I,
the spring stiffness in the axial direction ka , and the spring stiffness in the transverse
direction kt .
The element load vector fe can also be computed if uniformly distributed loads are
applied to the element. The optional input variable eq, described in beam2e, contains
the distributed loads.
5.6 – 11 ELEMENT
beam2w Two dimensional beam element
Theory:
The element stiffness matrix Ke , stored in Ke, is computed according to
Ke = GT (K̄e + K̄es )G
ELEMENT 5.6 – 12
Two dimensional beam element beam2ws
Purpose:
Compute section forces in a two dimensional beam element on elastic foundation.
V2
M2 N2
y
M1
N1 V1
x
Syntax:
es=beam2ws(ex,ey,ep,ed)
es=beam2ws(ex,ey,ep,ed,eq)
Description:
beam2ws computes the section forces at the ends of the beam element on elastic
foundation beam2w.
The input variables ex, ey, ep and eq are defined in beam2w, and the element dis-
placements, stored in ed, are obtained by the function extract. If distributed loads
are applied to the element the variable eq must be included.
The output variable
N1 V1 M1
es =
N2 V2 M2
5.6 – 13 ELEMENT
beam2ws Two dimensional beam element
Theory:
The section forces at the ends of the beam are obtained from the element force vector
P̄ = [ −N1 − V1 − M1 N2 V2 M2 ]T
computed according to
The matrices K̄e and G, are described in beam2e, and the matrix K̄es is described in
beam2w. The nodal displacements
ae = [ u1 u2 u3 u4 u5 u6 ]T
ELEMENT 5.6 – 14
Two dimensional beam element beam2g
Purpose:
Compute element stiffness matrix for a two dimensional nonlinear beam element.
u5
u6 u4
(x2,y2)
y u2 x E, A, I, N
u3 u1
(x1,y1)
Syntax:
Ke=beam2g(ex,ey,ep,N)
[Ke,fe]=beam2g(ex,ey,ep,N,eq)
Description:
beam2g provides the global element stiffness matrix Ke for a two dimensional beam
element with respect to geometrical nonlinearity.
The input variables ex, ey, and ep are described in beam2e, and
N=[N ]
contains the value of the predefined normal force N , which is positive in tension.
The element load vector fe can also be computed if a uniformly distributed transverse
load is applied to the element. The optional input variable
eq = [ qȳ ]
then contains the distributed transverse load per unit length, qȳ . Note that eq is a
scalar and not a vector as in beam2e
5.6 – 15 ELEMENT
beam2g Two dimensional beam element
Theory:
The element stiffness matrix Ke , stored in the variable Ke, is computed according to
Ke = GT K̄e G
kL kL 1 k 2 L2 1 3
φ1 = cot φ2 = φ3 = φ1 + φ2
2 2 12 (1 − φ1 ) 4 4
1 3
φ4 = − φ1 + φ2 φ5 = φ1 φ2
2 2
with
π√
k= ρ
L
kL kL 1 k 2 L2 1 3
φ1 = coth φ2 = − φ3 = φ1 + φ2
2 2 12 (1 − φ1 ) 4 4
1 3
φ4 = − φ1 + φ2 φ5 = φ1 φ2
2 2
with
π√
k= −ρ
L
N L2
ρ=−
π 2 EI
ELEMENT 5.6 – 16
Two dimensional beam element beam2g
The equivalent nodal loads fle stored in the variable fe are computed according to
fle = GT f̄le
where
T
1 L 1 L
f̄le = qL 0 ψ 0 − ψ
2 12 2 12
5.6 – 17 ELEMENT
beam2gs Two dimensional beam element
Purpose:
Compute section forces in a two dimensional nonlinear beam element.
V2
M2 N2
y
M1
N1 V1
Syntax:
es=beam2gs(ex,ey,ep,ed,N)
es=beam2gs(ex,ey,ep,ed,N,eq)
Description:
beam2gs computes the section forces at the ends of the nonlinear beam element
beam2g.
The input variables ex, ey, and ep are defined in beam2e, and the variables N and eq
in beam2g. The element displacements, stored in ed, are obtained by the function
extract. If a distributed transversal load is applied to the element, the variable eq
must be included.
The output variable
N1 V1 M1
es =
N2 V2 M2
ELEMENT 5.6 – 18
Two dimensional beam element beam2gs
Theory:
The section forces at the ends of the beam are obtained from the element force vector
P̄ = [ −N1 − V1 − M1 N2 V2 M2 ]T
computed according to
P̄ = K̄e G ae − f̄le
The matrix G is described in beam2e. The matrix K̄e and the nodal displacements
ae = [ u1 u2 u3 u4 u5 u6 ]T
5.6 – 19 ELEMENT
beam2d Two dimensional beam element
Purpose:
Compute element stiffness, mass and damping matrices for a two dimensional beam
element.
u5
u6 u4
(x2,y2)
y u2 x E, A, I, m
u3 u1
(x1,y1)
x
Syntax:
[Ke,Me]=beam2d(ex,ey,ep)
[Ke,Me,Ce]=beam2d(ex,ey,ep)
Description:
beam2d provides the global element stiffness matrix Ke, the global element mass
matrix Me, and the global element damping matrix Ce, for a two dimensional beam
element.
The input variables ex and ey are described in beam2e, and
ep = [ E A I m [ a0 a1 ] ]
contains the modulus of elasticity E, the cross section area A, the moment of inertia
I, the mass per unit length m, and the Raleigh damping coefficients a0 and a1 . If a0
and a1 are omitted, the element damping matrix Ce is not computed.
ELEMENT 5.6 – 20
Two dimensional beam element beam2d
Theory:
The element stiffness matrix Ke , the element mass matrix Me and the element
damping matrix Ce , stored in the variables Ke, Me and Ce, respectively, are computed
according to
5.6 – 21 ELEMENT
beam2ds Two dimensional beam element
Purpose:
Compute section forces for a two dimensional beam element in dynamic analysis.
V2
M2 N2
y
M1
N1 V1
Syntax:
es=beam2ds(ex,ey,ep,ed,ev,ea)
Description:
beam2ds computes the section forces at the ends of the dynamic beam element
beam2d.
The input variables ex, ey, and ep are defined in beam2d. The element displace-
ments, the element velocities, and the element accelerations, stored in ed, ev, and ea
respectively, are obtained by the function extract.
The output variable
N1 V1 M1
es =
N2 V2 M2
ELEMENT 5.6 – 22
Two dimensional beam element beam2ds
Theory:
The section forces at the ends of the beam are obtained from the element force vector
P̄ = [ −N1 − V1 − M1 N2 V2 M2 ]T
computed according to
The matrices K̄e and G are described in beam2e, and the matrices M̄e and C̄e are
described in beam2d. The nodal displacements
ae = [ u1 u2 u3 u4 u5 u6 ]T
Note that the transposes of ae , ȧe , and äe are stored in ed, ev, and ea respectively.
5.6 – 23 ELEMENT
beam3e Three dimensional beam element
Purpose:
Compute element stiffness matrix for a three dimensional beam element.
u11
u8
(x2,y2,z2)
u7 u10
u9
u5 u12
u2
y y
x
(x1,y1,z1)
u1 u4
u3 z
u6
x
z
Syntax:
Ke=beam3e(ex,ey,ez,eo,ep)
[Ke,fe]=beam3e(ex,ey,ez,eo,ep,eq)
Description:
beam3e provides the global element stiffness matrix Ke for a three dimensional beam
element.
The input variables
ex = [ x1 x2 ]
ey = [ y1 y2 ] eo = [ xz̄ yz̄ zz̄ ]
ez = [ z1 z2 ]
supply the element nodal coordinates x1 , y1 , etc. as well as the direction of the local
beam coordinate system (x̄, ȳ, z̄). By giving a global vector (xz̄ , yz̄ , zz̄ ) parallel with
the positive local z̄ axis of the beam, the local beam coordinate system is defined.
The variable
ep = [ E G A Iȳ Iz̄ Kv ]
supplies the modulus of elasticity E, the shear modulus G, the cross section area
A, the moment of inertia with respect to the ȳ axis Iy , the moment of inertia with
respect to the z̄ axis Iz , and St Venant torsional stiffness Kv .
ELEMENT 5.6 – 24
Three dimensional beam element beam3e
The element load vector fe can also be computed if uniformly distributed loads are
applied to the element. The optional input variable
then contains the distributed loads. The positive directions of qx̄ , qȳ , and qz̄ follow
the local beam coordinate system. The distributed torque qω̄ is positive if directed
in the local x̄-direction, i.e. from local ȳ to local z̄. All the loads are per unit length.
Theory:
The element stiffness matrix Ke is computed according to
Ke = GT K̄e G
where
k1 0 0 0 0 0 −k1 0 0 0 0 0
0 12EI z̄
0 0 0 6EIz̄
0 − 12EIz̄
0 0 0 6EIz̄
3 2 3 L2
L L L
0 0 12EI L3
ȳ
0 − 6EIȳ
L2
0 0 0 − 12EIȳ
L3
0 − 6EIȳ
L2
0
0 0 0 k2 0 0 0 0 0 −k2 0 0
6EIȳ 4EIȳ 6EIȳ 2EIȳ
0 0 − L2 0 L 0 0 0 0 L 0
L2
0 6EI z̄
0 0 0 4EIz̄
0 − 6EIz̄
0 0 0 2EIz̄
2 2
K̄e = L L L L
−k1 0 0 0 0 0 k1 0 0 0 0 0
0 − 12EI z̄
0 0 0 − 6EI z̄
0 12EI z̄
0 0 0 − 6EI z̄
L3 L2 L3 L2
12EIȳ 6EIȳ 12EIȳ 6EIȳ
0 0 − L3 0 L2 0 0 0 L3
0 L2 0
0 0 0 −k2 0 0 0 0 0 k2 0 0
6EIȳ 2EIȳ 6EIȳ 4EIȳ
0 0 − L2 0 L 0 0 0 L2
0 L 0
0 6EI
L 2
z̄
0 0 0 2EIz̄
L
0 − 6EIz̄
L2 0 0 0 4EIz̄
L
EA GKv
in which k1 = L
and k2 = L
, and where
nxx̄ nyx̄ nzx̄ 0 0 0 0 0 0 0 0 0
nxȳ nyȳ nzȳ 0 0 0 0 0 0 0 0 0
nxz̄ nyz̄ nzz̄ 0 0 0 0 0 0 0 0 0
0 0 0 nxx̄ nyx̄ nzx̄ 0 0 0 0 0 0
0 0 0 nxȳ nyȳ nzȳ 0 0 0 0 0 0
0 0 0 nxz̄ nyz̄ nzz̄ 0 0 0 0 0 0
G=
0 0 0 0 0 0 nxx̄ nyx̄ nzx̄ 0 0 0
0 0 0 0 0 0 nxȳ nyȳ nzȳ 0 0 0
0 0 0 0 0 0 nxz̄ nyz̄ nzz̄ 0 0 0
0 0 0 0 0 0 0 0 0 nxx̄ nyx̄ nzx̄
0 0 0 0 0 0 0 0 0 nxȳ nyȳ nzȳ
0 0 0 0 0 0 0 0 0 nxz̄ nyz̄ nzz̄
5.6 – 25 ELEMENT
beam3e Three dimensional beam element
In the transformation matrix G, nxx̄ specifies the cosine of the angle between the x
axis and x̄ axis, and so on.
The element load vector fle , stored in fe, is computed according to
fle = GT f̄le
where
qx̄ L
2
qȳ L
2
qz̄ L
2
qω̄ L
2
qz̄ L2
−
12
qȳ L2
12
f̄le =
qx̄ L
2
qȳ L
2
qz̄ L
2
qω̄ L
2
qz̄ L2
12
qȳ L2
−
12
ELEMENT 5.6 – 26
Three dimensional beam element beam3s
Purpose:
Compute section forces in a three dimensional beam element.
My,n
Vy,n Tn
Nn
Vz,n
Mz,1 Mz,n
My
Vz,1 Mz
y y Vy
x Vz T
N1 N
T1 z Vy,1
N
T Vz Vy
My,1
Mz My
x
z
Syntax:
es=beam3s(ex,ey,ez,eo,ep,ed)
es=beam3s(ex,ey,ez,eo,ep,ed,eq)
[es,edi,eci]=beam3s(ex,ey,ez,eo,ep,ed,eq,n)
Description:
beam3s computes the section forces and displacements in local directions along the
beam element beam3e.
The input variables ex, ey, ez, eo, and ep are defined in beam3e, and the element
displacements, stored in ed, are obtained by the function extract. If distributed
loads are applied to the element, the variable eq must be included. The number of
evaluation points for section forces and displacements are determined by n. If n is
omitted, only the ends of the beam are evaluated.
The output variables
consist of column matrices that contain the section forces, the displacements, and
the evaluation points on the local x̄-axis. The explicit matrix expressions are
N1 Vȳ1 Vz̄1 T Mȳ1 Mz̄1
N2 Vȳ2 Vz̄2 T Mȳ2 Mz̄2
es =
.. .. .. .. .. ..
. . . . . .
Nn Vȳn Vz̄n T Mȳn Mz̄n
5.6 – 27 ELEMENT
beam3s Three dimensional beam element
0
ū1 v̄1 w̄1 ϕ̄1
x̄2
ū2 v̄2 w̄2 ϕ̄2
..
edi = .. .. .. .. eci =
.
. . . .
x̄n−1
ūn v̄n w̄n ϕ̄n
L
Theory:
The evaluation of the section forces is based on the solutions of the basic equations
d2 ū d4 v̄
EA + qx̄ = 0 EIz − qȳ = 0
dx̄2 dx̄4
d4 w̄ d2 ϕ̄
EIy 4 − qz̄ = 0 GKv 2 + qω̄ = 0
dx̄ dx̄
From these equations, the displacements along the beam element are obtained as the
sum of the homogeneous and the particular solutions
ū(x̄)
v̄(x̄)
u=
= uh + up
w̄(x̄)
ϕ̄(x̄)
where
qx̄ Lx̄ x̄
(1 − )
2EA L
ūp (x̄) qȳ L2 x̄2 x̄
(1 − )2
v̄p (x̄) 24EIz L
uh = N̄ C−1 G ae up =
=
w̄p (x̄) qz̄ L2 x̄2 x̄
ϕ̄p (x̄) (1 − )2
24EIy L
qω̄ Lx̄ x̄
(1 − )
2GKv L
and
1 x̄ 0 0 0 0 0 0 0 0 0 0
0 0 1 x̄ x̄2 x̄3 0 0 0 0 0 0
N̄ =
0 0 0 0 0 0 1 x̄ x̄2 x̄3 0 0
0 0 0 0 0 0 0 0 0 0 1 x̄
ELEMENT 5.6 – 28
Three dimensional beam element beam3s
1 0 0 0 0 0 0 0 0 0 0 0
0 0 1 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 1 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 1 0
0 0 0 0 0 0 0 1 0 0 0 0 u1
0 0 0 1 0 0 0 0 0 0 0 0 u2
C=
ae = ..
1 L 0 0 0 0 0 0 0 0 0 0 .
0 0 1 L L 2 L3 0 0 0 0 0 0
u12
0 0 0 0 0 0 1 L L 2 L3 0 0
0 0 0 0 0 0 0 0 0 0 1 L
0 0 0 0 0 0 0 1 2L 3L2 0 0
0 0 0 1 2L 3L2 0 0 0 0 0 0
dū d3 v̄ d3 w̄
N = EA Vȳ = −EIz Vz̄ = −EIy
dx̄ dx̄3 dx̄3
dϕ̄ d2 w̄ d2 v̄
T = GKv Mȳ = −EIy 2 Mz̄ = EIz 2
dx̄ dx̄ dx̄
Examples:
Section forces or element displacements can easily be plotted. The bending moment
Mȳ along the beam is plotted by
>> plot(eci,es(:,5))
5.6 – 29 ELEMENT
5.7 Plate element
Only one plate element is currently available, a rectangular 12 dof element. The element
presumes a linear elastic material which can be isotropic or anisotropic.
Plate elements
u12 u9
u11 u8
u10 u7
u3 u6
u2 u5
u1 u4
platre
Plate functions
platre Compute element matrices
platrs Compute section forces
5.7 – 1 ELEMENT
platre Plate element
Purpose:
Compute element stiffness matrix for a rectangular plate element.
u12 u9
(x4 ,y4) u11 u8
y u3 u6
u2 u5
(x1 ,y1)
u1 u4 (x2 ,y2)
x t
z
Syntax:
Ke=platre(ex,ey,ep,D)
[Ke,fe]=platre(ex,ey,ep,D,eq)
Description:
platre provides an element stiffness matrix Ke, and an element load vector fe, for a
rectangular plate element. This element can only be used if the element edges are
parallel to the coordinate axis.
The element nodal coordinates x1 , y1 , x2 etc. are supplied to the function by ex and
ey, the element thickness t by ep, and the material properties by the constitutive
matrix D. Any arbitrary D-matrix with dimensions (3 × 3) and valid for plane stress
may be given. For an isotropic elastic material the constitutive matrix can be formed
by the function hooke, see Section 4.
D11 D12 D13
ex = [ x1 x2 x3 x4 ]
ep = [ t ] D= D21 D22 D23
ey = [ y1 y2 y3 y4 ]
D31 D32 D33
If a uniformly distributed load is applied to the element, the element load vector fe
is computed. The input variable
eq = [ qz ]
ELEMENT 5.7 – 2
Plate element platre
Theory:
The element stiffness matrix Ke and the element load vector fel , stored in Ke and fe
respectively, are computed according to
Ke = (C−1 )T B̄ D̃ B̄ dA C−1
T
A
fel = (C−1 )T
T
N̄ qz dA
A
t3
D̃ = D
12
w(x, y) = Ne ae = N̄ C−1 ae
where
N̄ = 1 x y x2 xy y2 x3 x2 y xy 2 y3 x3 y xy 3
1 −a −b a2 ab b2 −a3 −a2 b −ab2 −b3 a3 b ab3
0 0 1 0 −a −2b 0 a2 2ab 3b2 −a −3ab2
3
0 −1 0 2a b 0 −3a −2ab −b2
2
0 3a2 b b3
1 a −b a −ab b2 a3 −a2 b ab2
2
−b3 −a b −ab3
3
0 0 1 0 a −2b 0 a2 −2ab 3b2 a3 3ab2
0 −1 0 −2a b 0 −3a 2ab −b2
2
0 2
3a b b3
C= 2
1 a b a ab b2 a3 a2 b ab2 b3 ab3
ab3
0 0 1 0 a 2b 0 a2 2ab 3b2 a 3
3ab2
0 −1 0 −2a −b 0 −3a −2ab −b2
2
0 −3a b −b3
2
1 −a b a −ab b2 −a3 a2 b −ab2
2
b3 −a3 b −ab3
0 0 1 0 −a 2b 0 a2 −2ab 3b2 −a3 −3ab2
0 −1 0 2a −b 0 −3a 2ab −b2
2
0 −3a2 b −b3
ae = [ u1 u2 ... u12 ]T
and where
1 1
a = (x3 − x1 ) and b = (y3 − y1 )
2 2
5.7 – 3 ELEMENT
platre Plate element
where
∂2
∂x2
∗ ∂2
∇=
∂y 2
∂2
2
∂x∂y
Evaluation of the integrals for the rectangular plate element is done analytically.
Computation of the integrals for the element load vector fle yields
T
1 Ly Lx 1 Ly Lx 1 Ly Lx 1 Ly Lx
fle = qz Lx Ly − − − −
4 24 24 4 24 24 4 24 24 4 24 24
where
Lx = x3 − x1 and Ly = y3 − y1
ELEMENT 5.7 – 4
Plate element platrs
Purpose:
Compute section forces in a rectangular plate element.
u12 u9 Vyz
Mxy
(x4 ,y4) u11 u8
Myy
u10 u7 (x3 ,y3) Vxz
Mxy Mxx Mxy
y u3 u6 Vxz Mxx
u2 Myy
u5
(x1 ,y1)
Vyz Mxy
u1 u4 (x2 ,y2)
x t
z
Syntax:
[es,et]=platrs(ex,ey,ep,D,ed)
Description:
platrs computes the section forces es and the curvature matrix et in a rectangular
plate element. The section forces and the curvatures are computed at the center of
the element.
The input variables ex, ey, ep and D are defined in platre. The vector ed contains the
nodal displacements ae of the element and is obtained by the function extract as
5.7 – 5 ELEMENT
platrs Plate element
Theory:
The curvatures and the section forces are computed according to
κxx
κ= −1 e
κyy = B̄ C a
κxy
Mxx
M=
Myy = D̃ κ
Mxy
Vxz
V= =∇
˜ M
Vyz
where the matrices D̃, B̄, C and ae are described in platre, and where
∂ ∂
∂x 0
∂y
∇
˜ =
∂ ∂
0
∂y ∂x
ELEMENT 5.7 – 6
6 System functions
6.1 Introduction
The group of system functions comprises functions for the setting up, solving, and elimi-
nation of systems of equations. The functions are separated in two groups:
Ka = f
where K is the global stiffness matrix and f is the global load vector. Often used static
system functions are assem and solveq. The function assem assembles the global stiffness
matrix and solveq computes the global displacement vector a considering the boundary
conditions. It should be noted that K, f , and a also represent analogous quantities in
systems others than structural mechanical systems. For example, in a heat flow problem
K represents the conductivity matrix, f the heat flow, and a the temperature.
Dynamic system functions are related to different aspects of linear dynamic systems of
coupled ordinary differential equations according to
C ḋ + K d = f
6.1 – 1 SYSTEM
6.2 Static system functions
The group of static system functions comprises functions for setting up and solving the
global system of equations. It also contains a function for eigenvalue analysis, a function
for static condensation, a function for extraction of element displacements from the global
displacement vector and a function for extraction of element coordinates.
6.2 – 1 SYSTEM
assem Static system functions
Purpose:
Assemble element matrices.
i j
.. ..
k11 k12 .. ..
k21 .. ..
i j
e e e
kii kije i kii + kii kij + kij i
..
..
..
..
..
..
e e e e
kji kjj j kji + kji kjj + kjj j
..
..
..
.. ..
..
..
e ..
K .. ..
i = dofi .. .. knn
j = dofj
K
Syntax:
K=assem(edof,K,Ke)
[K,f]=assem(edof,K,Ke,f,fe)
Description:
assem adds the element stiffness matrix Ke , stored in Ke, to the structure stiffness
matrix K, stored in K, according to the topology matrix edof.
The element topology matrix edof is defined as
where the first column contains the element number, and the columns 2 to (ned + 1)
contain the corresponding global degrees of freedom (ned = number of element de-
grees of freedom).
In the case where the matrix Ke is identical for several elements, assembling of these
can be carried out simultaneously. Each row in Edof then represents one element,
i.e. nel is the total number of considered elements.
el1 dof1 dof2 . . . dofned
el2 dof1 dof2 . . . dofned
Edof =
.. .. .. .. one
row f or each element
. . . .
elnel dof1 dof2 . . . dofned
If fe and f are given in the function, the element load vector f e is also added to the
global load vector f .
SYSTEM 6.2 – 2
Static system functions coordxtr
Purpose:
Extract element coordinates from a global coordinate matrix.
ln
kn
● ● ● ● ●
(xn ,yn )
nel
● ● ● ● ● Ex = x1 x2 x7 x6
..
.
● ● ● ● ●
Ey = y1 y2 y7 y6
l6
5 ..
.
k6
(x6 , y6 )
● ● ● ● ●
l1 1 l2 2 3 4 l5
k1 k2 k5
● ● ● ● ●
(x1 , y1 ) (x2 , y2 ) (x5 , y5 )
nen =4
Syntax:
[Ex,Ey,Ez]=coordxtr(Edof,Coord,Dof,nen)
Description:
coordxtr extracts element nodal coordinates from the global coordinate matrix Coord
for elements with equal numbers of element nodes and dof’s.
Input variables are the element topology matrix Edof, defined in assem, the global
coordinate matrix Coord, the global topology matrix Dof, and the number of element
nodes nen in each element.
x1 y 1 [z1 ] k1 l1 ... m1
x2 y 2 [z2 ] k2 l2 ... m2
x3 y 3 [z3 ] k3 l3 ... m3
Coord = Dof = nen = [ nen ]
.. .. .. .. .. ..
. . . . . ... .
xn y n [zn ] kn ln ... mn
The nodal coordinates defined in row i of Coord correspond to the degrees of freedom
of row i in Dof. The components ki , li and mi define the degrees of freedom of node
i, and n is the number of global nodes for the considered part of the FE-model.
6.2 – 3 SYSTEM
coordxtr Static system functions
The output variables Ex, Ey, and Ez are matrices defined according to
x1 1 x2 1 x3 1 ... xnen 1
x1 2 x2 2 x3 2 ... xnen 2
Ex =
.. .. .. .. ..
. . . . .
x1 nel x2 nel x3 nel ... xnen nel
where row i gives the x-coordinates of the element defined in row i of Edof, and
where nel is the number of considered elements.
The element coordinate data extracted by the function coordxtr can be used for
plotting purposes and to create input data for the element stiffness functions.
SYSTEM 6.2 – 4
Static system functions eigen
Purpose:
Solve the generalized eigenvalue problem.
Syntax:
L=eigen(K,M)
L=eigen(K,M,b)
[L,X]=eigen(K,M)
[L,X]=eigen(K,M,b)
Description:
eigen solves the eigenvalue problem
| K − λM |= 0
where K and M are square matrices. The eigenvalues λ are stored in the vector L
and the corresponding eigenvectors in the matrix X.
If certain rows and columns in matrices K and M are to be eliminated in computing
the eigenvalues, b must be given in the function. The rows (and columns) that are
to be eliminated are described in the vector b defined as
dof1
dof2
b= ..
.
dofnb
The computed eigenvalues are given in order ranging from the smallest to the largest.
The eigenvectors are normalized in order that
XT MX = I
6.2 – 5 SYSTEM
extract Static system functions
Purpose:
Extract element nodal quantities from a global solution vector.
..
.
ai
ai u1
aj =
.. aj u2 edof = [eln i j m n ]
am u3 ed = [ u1 u2 u3 u4 ]
am
an u4
an
..
.
Syntax:
ed=extract(edof,a)
Description:
extract extracts element displacements or corresponding quantities ae from the global
solution vector a, stored in a.
Input variables are the element topology matrix edof, defined in assem, and the global
solution vector a.
The output variable
ed = (ae )T
where row i gives the element displacements for the element defined in row i of Edof,
and nel is the total number of considered elements.
SYSTEM 6.2 – 6
Static system functions extract
Example:
For the two dimensional beam element, the extract function will extract six nodal
displacements for each element given in Edof, and create a matrix Ed of size (nel × 6).
u1 u2 u3 u4 u5 u6
u1 u2 u3 u4 u5 u6
Ed = .. .. .. .. .. ..
. . . . . .
u1 u2 u3 u4 u5 u6
6.2 – 7 SYSTEM
insert Static system functions
Purpose:
Assemble internal element forces in a global force vector.
f1
f2
..
..
fi e fi + f i e
e
fj e fj + f j
..
fe ..
i = dofi fn
j = dofj
f
Syntax:
f=insert(edof,f,ef)
Description:
insert adds the internal element load vector fie , stored in ef, to the global internal
force vector f, stored in f, according to the topology matrix edof. The function is for
use in nonlinear analysis.
The element topology matrix edof is defined in assem. The vector f is the global
internal force vector, and the vector ef is the internal element force vector computed
from the element stresses, see for example plani4f.
SYSTEM 6.2 – 8
Static system functions solveq
Purpose:
Solve equation system.
Syntax:
a=solveq(K,f)
a=solveq(K,f,bc)
[a,r]=solveq(K,f,bc)
Description:
solveq solves the equation system
Ka=f
where the first column contains the row numbers and the second column the corre-
sponding prescribed values.
If r is given in the function, support forces are computed according to
r=Ka−f
6.2 – 9 SYSTEM
statcon Static system functions
Purpose:
Reduce system of equations by static condensation.
Syntax:
[K1,f1]=statcon(K,f,b)
Description:
statcon reduces a system of equations
Ka =f
by static condensation.
The degrees of freedom to be eliminated are supplied to the function by the vector
dof1
dof2
b= ..
.
dofnb
K1 a1 = f1
SYSTEM 6.2 – 10
6.3 Dynamic system functions
The group of system functions comprises functions for solving linear dynamic systems by
time stepping or modal analysis, functions for frequency domain analysis, etc.
Note: Eigenvalue analysis is performed by using the function eigen; see static system
functions.
6.3 – 1 SYSTEM
dyna2 Dynamic system functions
Purpose:
Compute the dynamic solution to a set of uncoupled second-order differential equa-
tions.
Syntax:
X=dyna2(w2,xi,f,g,dt)
Description:
dyna2 computes the solution to the set
Note:
The accuracy of the solution is not a function of the output time increment dt, since
the command produces the exact solution for straight line segments in the loading
time function.
See also:
gfunc
SYSTEM 6.3 – 2
Dynamic system functions dyna2f
Purpose:
Compute the dynamic solution to a set of uncoupled second-order differential equa-
tions.
Syntax:
Y=dyna2f(w2,xi,f,p,dt)
Description:
dyna2f computes the solution to the set
Example:
The dynamic solution to a set of uncoupled second-order differential equations can
be computed by the following sequence of commands:
>> g=gfunc(G,dt);
>> p=fft(g);
>> Y=dyna2f(w2,xi,f,p,dt);
>> X=(real(ifft(Y.’)))’;
where it is assumed that the input variables G, dt, w2, xi and f are properly defined.
Note that the ifft command operates on column vectors if Y is a matrix; therefore
use the transpose of Y. The output from the ifft command is complex. Therefore
use Y.’ to transpose rows and columns in Y in order to avoid the complex conjugate
transpose of Y; see Section 3. The time response is represented by the real part of
the output from the ifft command. If the transpose is used and the result is stored
in a matrix X, each row will represent the time response for each equation as the
output from the command dyna2.
See also:
gfunc, fft, ifft
6.3 – 3 SYSTEM
fft Dynamic system functions
Purpose:
Transform functions in time domain to frequency domain.
Syntax:
p=fft(g)
p=fft(g,N)
Description:
fft transforms a time dependent function to the frequency domain.
The function to be transformed is stored in the vector g. Each row in g contains
the value of the function at equal time intervals. The function represents a span
−∞ ≤ t ≤ +∞ ; however, only the values within a typical period are specified by g.
The fft command can be used with one or two input arguments. If N is not specified,
the numbers of frequencies used in the transformation is equal to the the numbers
of points in the time domain, i.e. the length of the variable g, and the output will
be a vector of the same size containing complex values representing the frequency
content of the input signal.
The scalar variable N can be used to specify the numbers of frequencies used in the
Fourier transform. The size of the output vector in this case will be equal to N.
It should be remembered that the highest harmonic component in the time signal
that can be identified by the Fourier transform corresponds to half the sampling
frequency. The sampling frequency is equal to 1/dt, where dt is the time increment
of the time signal.
The complex Fourier coefficients p(k) are stored in the vector p, and are computed
according to
N
(j−1)(k−1)
p(k) = x(j)ωN ,
j=1
where
ωN = e−2πi/N .
Note:
This is a MATLAB built-in function.
SYSTEM 6.3 – 4
Dynamic system functions freqresp
Purpose:
Compute frequency response of a known discrete time response.
Syntax:
[Freq,Resp] = freqresp(D,dt)
Description:
freqresp computes the frequency response of a discrete dynamic system.
D is the time history function and dt is the sampling time increment, i.e. the time
increment used in the time integration procedure.
Resp contains the computed response as a function of frequency. Freq contains the
corresponding frequencies.
Example:
The result can be visualized by
>> plot(Freq,Resp)
>> xlabel(’frequency (Hz)’)
or
>> semilogy(Freq,Resp)
>> xlabel(’frequency (Hz)’)
The dimension of Resp is the same as that of the original time history function.
Note:
The time history function of a discrete system computed by direct integration behaves
often in an unstructured manner. The reason for this is that the time history is a
mixture of several participating eigenmodes at different eigenfrequencies. By using a
Fourier transform, however, the response as a function of frequency can be computed
efficiently. In particular it is possible to identify the participating frequencies.
6.3 – 5 SYSTEM
gfunc Dynamic system functions
Purpose:
Form vector with function values at equally spaced points by linear interpolation.
g(t)
(t4,g(t4))
●
(t2,g(t2))
●
t
●
(t1,g(t1))
● ●
(t3,g(t3)) (tN,g(tN))
●
(t5,g(t5))
Syntax:
[t,g]=gfunc(G,dt)
Description:
gfunc uses linear interpolation to compute values at equally spaced points for a
discrete function g given by
t1 g(t1 )
t2 g(t2 )
G=
.. ,
.
tN g(tN )
SYSTEM 6.3 – 6
Dynamic system functions ifft
Purpose:
Transform function in frequency domain to time domain.
Syntax:
x=ifft(y)
x=ifft(y,N)
Description:
ifft transforms a function in the frequency domain to a function in the time domain.
The function to be transformed is given in the vector y. Each row in y contains
Fourier terms in the interval −∞ ≤ ω ≤ +∞.
The fft command can be used with one or two input arguments. The scalar variable
N can be used to specify the numbers of frequencies used in the Fourier transform.
The size of the output vector in this case will be equal to N. See also the description
of the command fft.
The inverse Fourier coefficients x(j), stored in the variable x, are computed according
to
N
−(j−1)(k−1)
x(j) = (1/N ) y(k)ωN ,
k=1
where
ωN = e−2πi/N .
Note:
This is a MATLAB built-in function.
See also:
fft
6.3 – 7 SYSTEM
ritz Dynamic system functions
Purpose:
Compute approximative eigenvalues and eigenvectors by the Lanczos method.
Syntax:
L=ritz(K,M,f,m)
L=ritz(K,M,f,m,b)
[L,X]=ritz(K,M,f,m)
[L,X]=ritz(K,M,f,m,b)
Description:
ritz computes, by the use of the Lanczos algorithm, m approximative eigenvalues and
m corresponding eigenvectors for a given pair of n-by-n matrices K and M and a
given non-zero starting vector f.
If certain rows and columns in matrices K and M are to be eliminated in computing
the eigenvalues, b must be given in the command. The rows (and columns) to be
eliminated are described in the vector b defined as
dof1
dof2
b=
.. .
.
dofnb
Note:
If the number of vectors, m, is chosen less than the total number of degrees-of-
freedom, n, only about the first m/2 Ritz vectors are good approximations of the
true eigenvectors. Recall that the Ritz vectors satisfy the M-orthonormality condition
XT M X = I,
SYSTEM 6.3 – 8
Dynamic system functions spectra
Purpose:
Compute seismic response spectra for elastic design.
Syntax:
s=spectra(a,xi,dt,f)
Description:
spectra computes the seismic response spectrum for a known acceleration history
function.
The computation is based on the vector a, that contains an acceleration time history
function defined at equal time steps. The time step is specified by the variable dt.
The value of the damping ratio is given by the variable xi.
Output from the computation, stored in the vector s, is achieved at frequencies
specified by the column vector f.
Example:
The following procedure can be used to produce a seismic response spectrum for a
damping ratio ξ = 0.05, defined at 34 logarithmicly spaced frequency points. The
acceleration time history a has been sampled at a frequency of 50 Hz, corresponding
to a time increment dt = 0.02 between collected points:
>> freq=logspace(0,log10(2^(33/6)),34);
>> xi=0.05;
>> dt=0.02;
>> s=spectra(a,xi,dt,freq’);
>> loglog(freq,s,’*’)
6.3 – 9 SYSTEM
step1 Dynamic system functions
Purpose:
Compute the dynamic solution to a set of first order differential equations.
Syntax:
Tsnap=step1(K,C,d0,ip,f,pbound)
[Tsnap,D,V]=step1(K,C,d0,ip,f,pbound)
Description:
step1 computes at equal time steps the solution to a set of first order differential
equations of the form
The command solves transient field problems. In the case of heat conduction, K and
C represent the n × n conductivity and capacity matrices, respectively.
The initial conditions are given by the vector d0 containing initial values of d. The
time integration procedure is governed by the parameters given in the vector ip
defined as
list of list of
nsnap nhist
moments dofs
where dt specifies the time increment in the time stepping scheme, T total time and
α a time integration constant; see [1]. The parameter nsnap denotes the number of
snapshots stored in Tsnap. The selected elapsed times are specified in (timei ... ),
whereas nhist is the number of time histories stored in D and V. The selected degrees-
of-freedom are specified in (dofi ... ). The following table lists frequently used values
of α:
The matrix f contains the time-dependent load vectors. If no external loads are
active, the matrix corresponding to f should be replaced by []. The matrix f contains
the time-dependent prescribed values of the field variable. If no field variables are
prescribed the matrix corresponding to pbound should be replaced by []. Matrix f is
organized in the following manner:
SYSTEM 6.3 – 10
Dynamic system functions step1
time history of the load at dof1
time history of the load at dof2
f=
.. .
.
time history of the load at dofn
The dimension of f is
The time history functions can be generated using the command gfunc. If all the
values of the time histories of f or pbound are kept constant, these values need to be
stated only once. In this case the number of columns in f is one and in pbound two.
It is highly recommended to define f as sparse (a MATLAB built-in function). In
most cases only a few degrees-of-freedom are affected by the exterior load, and hence
the matrix contains only few non-zero entries.
The computed snapshots are stored in Tsnap, one column for each requested snapshot
according to ip, i.e. the dimension of Tsnap is (number of dofs) × nsnap. The
computed time histories of d and ḋ are stored in D and V, respectively, one line
for each requested degree-of-freedom according to ip. The dimension of D is nhist ×
(number of timesteps + 1).
6.3 – 11 SYSTEM
step2 Dynamic system functions
Purpose:
Compute the dynamic solution to a set of second order differential equations.
Syntax:
Dsnap=step2(K,C,M,d0,v0,ip,f,pdisp)
[Dsnap,D,V,A]=step2(K,C,M,d0,v0,ip,f,pdisp)
Description:
step2 computes at equal time steps the solution to a second order differential equa-
tions of the form
where dt specifies the time increment in the time stepping scheme, T the total time
and α and δ time integration constants for the Newmark family of methods; see
[1]. The parameter nsnap denotes the number of snapshots stored in Dsnap. The
selected elapsed times are specified in (timei ... ), whereas nhist is the number of
time histories stored in D, V and A. The selected degrees-of-freedom are specified in
(dofi ... ). The following table lists frequently used values of α and δ:
1 1
α= 4
δ= 2
Average acceleration (trapezoidal) rule,
1 1
α= 6
δ= 2
Linear acceleration,
1
α=0 δ= 2
Central difference.
The matrix f contains the time-dependent load vectors. If no external loads are active,
the matrix corresponding to f should be replaced by []. The matrix pdisp contains
the time-dependent prescribed displacement. If no displacements are prescribed the
matrix corresponding to pdisp should be replaced by [].
SYSTEM 6.3 – 12
Dynamic system functions step2
time history of the load at dof1
time history of the load at dof2
f= .. .
.
time history of the load at dofn
The dimension of f is
The time history functions can be generated using the command gfunc. If all the
values of the time histories of f or pdisp are kept constant, these values need to be
stated only once. In this case the number of columns in f is one and in pdisp two.
It is highly recommended to define f as sparse (a MATLAB built-in function). In
most cases only a few degrees-of-freedom are affected by the exterior load, and hence
the matrix contains only few non-zero entries.
The computed displacement snapshots are stored in Dsnap, one column for each
requested snapshot according to ip, i.e. the dimension of Dsnap is (number of dofs) ×
nsnap. The computed time histories of d, ḋ and d̈ are stored in D, V and A,
respectively, one line for each requested degree-of-freedom according to ip. The
dimension of D is nhist × (number of timesteps + 1).
6.3 – 13 SYSTEM
sweep Dynamic system functions
Purpose:
Compute complex frequency response functions.
Syntax:
Y=sweep(K,C,M,p,w)
Description:
sweep computes the complex frequency response function for a system of the form
[K + iωC − ω 2 M]y(ω) = p.
Here K, C and M represent the m-by-m stiffness, damping and mass matrices, re-
spectively. The vector p defines the amplitude of the force. The frequency response
function is computed for the values of ω given by the vector w.
The complex frequency response function is stored in the matrix Y with dimension
m-by-n, where n is equal to the number of circular frequencies defined in w.
Example:
The steady-state response can be computed by
>> X=real(Y*exp(i*w*t));
>> Z=abs(Y);
SYSTEM 6.3 – 14
7 Statements and macros
Statements describe algorithmic actions that can be executed. There are two different
types of control statements, conditional and repetitive. The first group defines conditional
jumps whereas the latter one defines repetition until a conditional statement is fulfilled.
Macros are used to define new functions to the MATLAB or CALFEM structure, or to
store a sequence of statements in an .m-file.
Control statements
if Conditional jump
for Initiate a loop
while Define a conditional loop
Macros
f unction Define a new function
script Store a sequence of statements
7–1 STATEMENTS
if
Purpose:
Conditional jump.
Syntax:
if logical expression
..
.
elseif logical expression
..
.
else
..
.
end
Description:
if initiates a conditional jump. If logical expression produces the value True the
statements following if are executed. If logical expression produces the value False
the next conditional statement elseif is checked.
elseif works like if. One or more of the conditional statement elseif can be added after
the initial conditional statement if.
If else is present, the statements following else are executed if the logical expressions
in all if and elseif statements produce the value False. The if loop is closed by end to
define the loop sequence.
The following relation operators can be used
== equal
>= greater than or equal to
> greater than
<= less than or equal to
< less than
∼= not equal
Note:
This is MATLAB built-in language.
STATEMENTS 7–2
for
Purpose:
Initiate a loop.
Syntax:
for i = start : inc : stop
..
.
end
Description:
for initiates a loop which terminates when i>stop. The for loop is closed by end to
define the loop sequence.
Examples:
for i = 1 : 10 i takes values from 1 to 10.
for i = 1 : 2 : 10 i equals 1, 3, 5, 7, 9.
for i = 20 : -1 : 1 i equals 20, 19 ... 2, 1.
Note:
This is MATLAB built-in language.
7–3 STATEMENTS
while
Purpose:
Define a conditional loop.
Syntax:
while logical expression
..
.
end
Description:
while initiates a conditional loop which terminates when logical expression equals
False. The while loop is closed by end to define the loop sequence.
The different relation operators that can be used can be found under the if command.
Examples:
A loop continuing until a equals b
while a∼=b
..
.
end
Note:
This is MATLAB built-in language.
STATEMENTS 7–4
function
Purpose:
Define a new function.
Syntax:
function[ out1 , out2 , ... ]=name( in1 , in2 , ... )
Description:
name is replaced by the name of the function. The input variables in1, in2, ... can be
scalars, vectors or matrices, and the same holds for the output variables out1, out2,
... .
Example:
To define the CALFEM function spring1e a file named spring1e.m is created. The file
contains the following statements:
function [Ke]=spring1e(k)
% Define the stiffness matrix
% for a one dimensional spring
% with spring stiffness k
Ke=[ k, -k; -k, k ]
Note:
This is MATLAB built-in language.
7–5 STATEMENTS
script
Purpose:
Execute a stored sequence of statements.
Syntax:
name
Description:
name is replaced by the name of the script.
Example:
The statements below are stored in a file named spring.m and executed by typing
spring in the MATLAB command window.
Note:
This is MATLAB built-in language.
STATEMENTS 7–6
8 Graphics functions
The group of graphics functions comprises functions for element based graphics. Mesh
plots, displacements, section forces, flows, iso lines and principal stresses can be displayed.
The functions are divided into two dimensional, and general graphics functions.
8–1 GRAPHICS
axis
Purpose:
Plot axis scaling and appearance.
Syntax:
axis([xmin xmax ymin ymax])
axis([xmin xmax ymin ymax zmin zmax])
axis auto
axis square
axis equal
axis off
axis on
Description:
axis([xmin xmax ymin ymax]) sets scaling for the x- and y-axes on the current 2D plot.
axis([xmin xmax ymin ymax zmin zmax]) sets the scaling for the x-, y- and z-axes on
the current 3D plot.
axis auto returns the axis scaling to its default automatic mode where, for each plot,
xmin = min(x), xmax = max(x), etc.
axis square makes the current axis box square in shape.
axis equal changes the current axis box size so that equal tick mark increments on
the x- and y-axes are equal in size. This makes plot(sin(x),cos(x)) look like a circle,
instead of an oval.
axis normal restores the current axis box to full size and removes any restrictions on
the scaling of the units. This undoes the effects of axis square and axis equal.
axis off turns off all axis labeling and tick marks.
axis on turns axis labeling and tick marks back on.
Note:
This is a MATLAB built-in function. For more information about the axis function,
type help axis.
GRAPHICS 8–2
clf
Purpose:
Clear current figure (graph window).
Syntax:
clf
Description:
clf deletes all objects (axes) from the current figure.
Note:
This is a MATLAB built-in function. For more information about the clf function,
type help clf.
8–3 GRAPHICS
eldia2
Purpose:
Draw the section force diagrams of a two dimensional beam element.
y y
x x
y y
x x
Syntax:
eldia2(ex,ey,es,plotpar,sfac)
eldia2(ex,ey,es,plotpar,sfac,eci)
[sfac]=eldia2(ex,ey,es)
[sfac]=eldia2(ex,ey,es,plotpar)
Description:
eldia2 plots a section force diagram of a two dimensional beam element in its global
position.
The input variables ex and ey are defined in beam2e and the input variable
S1
S2
es = ..
.
Sn
consists of a column matrix that contains section forces. The values in es are com-
puted in beam2s. It should be noted, however, that whereas all three section forces
are computed in beam2s only one of them shall be given as input to eldia2 by es.
GRAPHICS 8–4
eldia2
specifies the local x̄-coordinates of the quantities in es. If eci is not given, uniform
distance is assumed.
Limitations:
Supported elements are two dimensional beam elements.
8–5 GRAPHICS
eldisp2
Purpose:
Draw the deformed mesh for a two dimensional structure.
Syntax:
[sfac]=eldisp2(Ex,Ey,Ed)
[sfac]=eldisp2(Ex,Ey,Ed,plotpar)
eldisp2(Ex,Ey,Ed,plotpar,sfac)
Description:
eldisp2 displays the deformed mesh for a two dimensional structure.
Input variables are the coordinate matrices Ex and Ey formed by the function co-
ordxtr, and the element displacements Ed formed by the function extract.
The variable plotpar sets plot parameters for linetype, linecolor and node marker.
Limitations:
Supported elements are bar, beam, triangular three node, and quadrilateral four
node elements.
GRAPHICS 8–6
eldraw2
Purpose:
Draw the undeformed mesh for a two dimensional structure.
Syntax:
eldraw2(Ex,Ey)
eldraw2(Ex,Ey,plotpar)
eldraw2(Ex,Ey,plotpar,elnum)
Description:
eldraw2 displays the undeformed mesh for a two dimensional structure.
Input variables are the coordinate matrices Ex and Ey formed by the function co-
ordxtr.
The variable plotpar sets plot parameters for linetype, linecolor and node marker.
elnum=Edof(:,1)
Limitations:
Supported elements are bar, beam, triangular three node, and quadrilateral four
node elements.
8–7 GRAPHICS
elflux2
Purpose:
Draw element flow arrows for two dimensional elements.
Syntax:
[sfac]=elflux2(Ex,Ey,Es)
[sfac]=elflux2(Ex,Ey,Es,plotpar)
elflux2(Ex,Ey,Es,plotpar,sfac)
Description:
elflux2 displays element heat flux vectors (or corresponding quantities) for a number
of elements of the same type. The flux vectors are displayed as arrows at the element
centroids. Note that only the flux vectors are displayed. To display the element mesh,
use eldraw2.
Input variables are the coordinate matrices Ex and Ey, and the element flux matrix
Es defined in flw2ts or flw2qs.
The variable plotpar sets plot parameters for the flux arrows.
Limitations:
Supported elements are triangular 3 node and quadrilateral 4 node elements.
GRAPHICS 8–8
eliso2
Purpose:
Display element iso lines for two dimensional elements.
Syntax:
eliso2(Ex,Ey,Ed,isov)
eliso2(Ex,Ey,Ed,isov,plotpar)
Description:
eliso2 displays element iso lines for a number of elements of the same type. Note that
only the iso lines are displayed. To display the element mesh, use eldraw2.
Input variables are the coordinate matrices Ex and Ey formed by the function co-
ordxtr, and the element nodal quantities (e.g displacement or energy potential) matrix
Ed defined in extract.
If isov is a scalar it determines the number of iso lines to be displayed. If isov is a
vector it determines the values of the iso lines to be displayed (number of iso lines
equal to length of vector isov).
isov = [ isolines]
isov = [ isovalue(1) ... isovalue(n) ]
The variable plotpar sets plot parameters for the iso lines.
Limitations:
Supported elements are triangular 3 node and quadrilateral 4 node elements.
8–9 GRAPHICS
elprinc2
Purpose:
Draw element principal stresses as arrows for two dimensional elements.
Syntax:
[sfac]=elprinc2(Ex,Ey,Es)
[sfac]=elprinc2(Ex,Ey,Es,plotpar)
elprinc2(Ex,Ey,Es,plotpar,sfac)
Description:
elprinc2 displays element principal stresses for a number of elements of the same type.
The principal stresses are displayed as arrows at the element centroids. Note that
only the principal stresses are displayed. To display the element mesh, use eldraw2.
Input variables are the coordinate matrices Ex and Ey, and the element stresses
matrix Es defined in plants or planqs
The variable plotpar sets plot parameters for the principal stress arrows.
Limitations:
Supported elements are triangular 3 node and quadrilateral 4 node elements.
GRAPHICS 8 – 10
figure
Purpose:
Create figures (graph windows).
Syntax:
figure(h)
Description:
figure(h) makes the h’th figure the current figure for subsequent plotting functions.
If figure h does not exist, a new one is created using the first available figure handle.
Note:
This is a MATLAB built-in function. For more information about the figure function,
type help figure.
8 – 11 GRAPHICS
fill
Purpose:
Filled 2D polygons.
Syntax:
fill(x,y,c)
fill(X,Y,C)
Description:
fill(x,y,c) fills the 2D polygon defined by vectors x and y with the color specified by
c. The vertices of the polygon are specified by pairs of components of x and y. If
necessary, the polygon is closed by connecting the last vertex to the first.
If c is a vector of the same length as x and y, its elements are used to specify colors
at the vertices. The color within the polygon is obtained by bilinear interpolation in
the vertex colors.
If X, Y and C are matrices of the same size, fill(X,Y,C) draws one polygon per column
with interpolated colors.
Example:
The solution of a heat conduction problem results in a vector d with nodal tem-
peratures. The temperature distribution in a group of triangular 3 node (nen=3)
or quadrilateral 4 node (nen=4) elements, with topology defined by edof, can be
displayed by
[ex,ey]=coordxtr(edof,Coord,Dof,nen)
ed=extract(edof,d)
colormap(hot)
fill(ex’,ey’,ed’)
Note:
This is a MATLAB built-in function. For more information about the fill function,
type help fill.
GRAPHICS 8 – 12
grid
Purpose:
Grid lines for 2D and 3D plots.
Syntax:
grid on
grid off
grid
Description:
grid on adds grid lines on the current axes.
grid off takes them off.
grid by itself, toggles the grid state.
Note:
This is a MATLAB built-in function. For more information about the grid function,
type help grid.
8 – 13 GRAPHICS
hold
Purpose:
Hold the current graph.
Syntax:
hold on
hold off
hold
Description:
hold on holds the current graph.
hold off returns to the default mode where plot functions erase previous plots.
hold by itself, toggles the hold state.
Note:
This is a MATLAB built-in function. For more information about the hold function,
type help hold.
GRAPHICS 8 – 14
plot
Purpose:
Linear two dimensional plot.
Syntax:
plot(x,y)
plot(x,y,’linetype’)
Description:
plot(x,y) plots vector x versus vector y. Straight lines are drawn between each pair
of values.
Various line types, plot symbols and colors may be obtained with plot(x,y,s) where s
is a 1, 2, or 3 character string made from the following characters:
– solid line . point y yellow
: dotted line o circle m magenta
-. dashdot line x x-mark c cyan
-- dashed line + plus r red
* star g green
b blue
w white
k black
Default is solid blue lines.
Example:
The statement
plot(x,y,’-’,x,y,’ro’)
plots the data twice, giving a solid blue line with red circles at the data points.
Note:
This is a MATLAB built-in function. For more information about the plot function,
type help plot.
8 – 15 GRAPHICS
pltscalb2
Purpose:
Draw a scale bar.
Syntax:
pltscalb2(sfac,magnitude)
pltscalb2(sfac,magnitude,plotpar)
Description:
pltscalb2 draws a scale bar to visualize the magnitude of displayed computational
results. The input variable sfact is a scale factor determined by the function scalfact2
and the variable
magnitude = [ S x y ]
specifies the value corresponding the length of the scale bar S, and (x, y) are the
coordinates of the starting point. If no coordinates are given the starting point will
be (0,-0.5).
The variable plotpar sets the the scale bar colour.
plotpar=[color ]
color = 1 black
2 blue
3 magenta
4 red
GRAPHICS 8 – 16
print
Purpose:
Create hardcopy output of current figure window.
Syntax:
print [filename]
Description:
print with no arguments sends the contents of the current figure window to the
default printer. print filename creates a PostScript file of the current figure window
and writes it to the specified file.
Note:
This is a MATLAB built-in function. For more information about the print function,
type help print.
8 – 17 GRAPHICS
scalfact2
Purpose:
Determine scale factor for drawing computational results.
Syntax:
[sfac]=scalfact2(ex,ey,ed)
[sfac]=scalfact2(ex,ey,ed,rat)
Description:
scalfact2 determines a scale factor sfac for drawing computational results, such as
displacements, section forces or flux.
Input variables are the coordinate matrices ex and ey and the matrix ed containing
the quantity to be displayed. The scalar rat defines the ratio between the geometric
representation of the largest quantity to be displayed and the element size. If rat is
not specified, 0.2 is used.
GRAPHICS 8 – 18
text
Purpose:
Add text to current plot.
Syntax:
text(x,y,’string’)
Description:
text adds the text in the quotes to location (x,y) on the current axes, where (x,y)
is in units from the current plot. If x and y are vectors, text writes the text at all
locations given. If ’string’ is an array with the same number of rows as the length of
x and y, text marks each point with the corresponding row of the ’string’ array.
Note:
This is a MATLAB built-in function. For more information about the text function,
type help text.
8 – 19 GRAPHICS
title
Purpose:
Titles for 2D and 3D plots.
Syntax:
title(’text’)
Description:
title adds the text string ’text’ at the top of the current plot.
Note:
This is a MATLAB built-in function. For more information about the title function,
type help title.
GRAPHICS 8 – 20
xlabel, ylabel, zlabel
Purpose:
x-, y-, and z-axis labels for 2D and 3D plots.
Syntax:
xlabel(’text’)
ylabel(’text’)
zlabel(’text’)
Description:
xlabel adds text beside the x-axis on the current plot.
ylabel adds text beside the y-axis on the current plot.
zlabel adds text beside the z-axis on the current plot.
Note:
This is a MATLAB built-in function. For more information about the functions,
type help xlabel, help ylabel, or help zlabel.
8 – 21 GRAPHICS
9 User’s Manual, examples
9.1 Introduction
This set of examples is defined with the ambition to serve as a User’s Manual. The
examples, except the introductory ones, are written as .m-files (script files) and supplied
together with the CALFEM functions.
MATLAB introduction
Static analysis
Dynamic analysis
Nonlinear analysis
The MATLAB introduction examples explain some basic concepts and introduce a set
of standard MATLAB functions usable in the finite element context. The static linear
examples illustrate finite element analysis of different structures loaded by stationary loads.
The dynamic linear examples illustrate some basic features in dynamics, such as modal
analysis and time stepping procedures. The examples of nonlinear analysis cover subjects
such as second order theory and buckling.
9.1 – 1 EXAMPLES
9.2 MATLAB introduction
The examples in this section illustrate basic MATLAB concepts such as handling of
workspace, variables and functions. The examples are:
MATLAB introduction
exi1 Handling matrices
exi2 Matrix and array operations
exi3 Create and handle .m-files
exi4 Display formats
exi5 Create a session .log-file
exi6 Graphic display of vectors
9.2 – 1 EXAMPLES
exi1 MATLAB introduction
Purpose:
Show how to create and handle matrices in MATLAB.
Description:
The following commands create a scalar x, two vectors u and v and two matrices A
and B.
Lines starting with the MATLAB prompt >> are command lines while the other lines
show the results from these commands.
>> x=7
x =
7
>> u=[1 2 3 4]
u =
1 2 3 4
>> v=[0:0.4:2]
v =
>> A=[1 2; 3 4]
A =
1 2
3 4
>> B=[5 6; 7 8]
B =
5 6
7 8
>> who
A B u v x
EXAMPLES 9.2 – 2
MATLAB introduction exi1
>> whos
>> u
u =
1 2 3 4
>> size(u)
ans =
1 4
>> clear x
>> A(2,2)=9
A =
1 2
3 9
9.2 – 3 EXAMPLES
exi1 MATLAB introduction
>> s=A(:,1)
s =
1
3
>> t=A(2,:)
t =
3 9
>> K=zeros(4,4)
K =
0 0 0 0
0 0 0 0
0 0 0 0
0 0 0 0
>> H=[A;B]
H =
1 2
3 9
5 6
7 8
>> J=[A B]
J =
1 2 5 6
3 9 7 8
EXAMPLES 9.2 – 4
MATLAB introduction exi2
Purpose:
Show some examples of basic matrix and element-by-element operations.
Description:
Consider the following matrices
a= 5 12 3 b= 1 0 4
1 6 3 2 5 4
A= B=
2 8 4 7 2 0
>> A’
ans =
1 2
6 8
3 4
>> a’
ans =
5
12
3
>> A+B
ans =
3 11 7
9 10 4
>> A-B
ans =
-1 1 -1
-5 6 4
Note that if the result of an operation is not assigned to a specific variable, the
answer is temporarily stored in the variable ans.
9.2 – 5 EXAMPLES
exi2 MATLAB introduction
Multiplication of matrices
>> a*b’
ans =
17
>> a’*b
ans =
5 0 20
12 0 48
3 0 12
>> A’*B
ans =
16 9 4
68 46 24
34 23 12
>> C=B*A’
C =
44 60
19 30
>> D=A*B
??? Error using ==> *
Inner matrix dimensions must agree.
>> inv(C)
ans =
0.1667 -0.3333
-0.1056 0.2444
EXAMPLES 9.2 – 6
MATLAB introduction exi2
>> det(C)
ans =
180
>> a.*b
ans =
5 0 12
>> A.*B
ans =
2 30 12
14 16 0
ans =
0 1 0 -1 0
9.2 – 7 EXAMPLES
exi3 MATLAB introduction
Purpose:
Show how to handle script files and function files.
Description:
When starting a MATLAB session the default working directory is according to
initial settings, for example C:\USER. A new working directory can be chosen by
typing for example
>> cd A:
>> test
A =
0 4
2 8
EXAMPLES 9.2 – 8
MATLAB introduction exi3
B =
3 9
5 7
C =
20 28
46 74
The second type of .m-files is the function file. The first line of these files contains
the word function. The example below is a function that computes the second and
third power of a scalar.
function [b,c]=func1(a)
% ----- function file ’func1.m’ -----
b=a*a;
c=a*a*a;
% ------------- end -----------------
The semi-colon prohibits the echo display of the variables on the screen.
The file can be created using an ordinary editor, and it must be saved as func1.m i.e.
the file name without extension must be the same as the function name.
The second and third power of 2 are calculated by typing
>> [b,c]=func1(2)
producing
b =
4
c =
8
9.2 – 9 EXAMPLES
exi4 MATLAB introduction
Purpose:
Show different display formats.
Description:
Consider the following matrix operation
C =
0.0079 0.0110
0.0181 0.0292
The result from the computation of C above is shown in the default format, displaying
four significant decimal digits.
Other formats can be defined by the command format
C =
0.00788332676389 0.01103665746945
0.01813165155696 0.02916830902641
C =
7.8833e-003 1.1037e-002
1.8132e-002 2.9168e-002
C =
7.883326763894364e-003 1.103665746945211e-002
1.813165155695703e-002 2.916830902640915e-002
EXAMPLES 9.2 – 10
MATLAB introduction exi5
Purpose:
How to make a command window session .log-file.
Description:
The diary and echo commands are useful for presentation purposes since the complete
set of statements can be saved together with some selected results.
The command
>> diary filename
saves all subsequent terminal input and resulting output in the file named filename
on the default device. The file is closed using
Normally, the statements in an .m-file do not display during execution. The com-
mands echo on and echo off allow the statements to be viewed as they execute.
Execution of test.m yields
>>test
C=A*B/2537
C =
0.0079 0.0110
0.0181 0.0292
9.2 – 11 EXAMPLES
exi6 MATLAB introduction
Purpose:
How to display vectors and handle the graphics window.
Description:
The contents of a vector versus the vector index or a vector versus another vector
can be displayed in the graphics window. Consider the vectors
x= 1 2 5 y= 5 22 16
The function
>> plot(y)
>> plot(x,y)
The commands
title(’text’ )
xlabel(’xlabel’ )
ylabel(’ylabel’ )
write text as a title of the current plot, and xlabel and ylabel as labels of the coordinate
axis.
EXAMPLES 9.2 – 12
9.3 Static analysis
This section illustrates some linear static finite element calculations. The examples deal
with structural problems as well as field problems such as heat conduction.
Static analysis
exs1 Linear spring system
exs2 One-dimensional heat flow
exs3 Plane truss
exs4 Plane truss analysed using loops
exs5 Simply supported beam
exs6 Plane frame
exs7 Plane frame stabilized with bars
exs8 Two dimensional diffusion
The introductory example exs1 illustrates the basic steps in the finite element method for
a simple structure of one-dimensional linear springs. The linear spring or analogy element
is also used in example exs2 to solve a problem of heat conduction in a wall. A plane
truss consisting of three bars is analysed in exs3. In example exs4 a plane truss consisting
of ten bars is analysed using loops. First the analysis is performed by defining coordinate
data for each element directly, and then it is shown how this data can be obtained from
global coordinate data. A simply supported beam is analysed in example exs5. Element
forces and the support forces are calculated. A two dimensional plane frame is analysed
in example exs6. A structure built up from both beams and bars is analysed in example
exs7. Graphics facilities are also explained in examples exs6, exs7, and exs8.
Note: The examples listed above are supplied as .m-files under the directory examples.
The example files are named according to the table.
9.3 – 1 EXAMPLES
exs1 Static analysis
Purpose:
Show the basic steps in a finite element calculation.
Description:
The general procedure in linear finite element calculations is carried out for a simple
structure. The steps are
Consider the system of three linear elastic springs, and the corresponding finite
element model. The system of springs is fixed in its ends and loaded by a single load
F.
k
2k
F
2k
2
1
2 3
● ● ●
1
3
The computation is initialized by defining the topology matrix Edof, containing ele-
ment numbers and global element degrees of freedom,
>> Edof=[1 1 2;
2 2 3;
3 2 3];
>> K=zeros(3,3)
K =
0 0 0
0 0 0
0 0 0
EXAMPLES 9.3 – 2
Static analysis exs1
and the load vector f (3×1) with the load F = 100 in position 2.
f =
0
100
0
Element stiffness matrices are generated by the function spring1e. The element prop-
erty ep for the springs contains the spring stiffnesses k and 2k respectively, where
k = 1500.
Ke1 =
1500 -1500
-1500 1500
>> Ke2=spring1e(ep2)
Ke2 =
3000 -3000
-3000 3000
The element stiffness matrices are assembled into the global stiffness matrix K ac-
cording to the topology.
>> K=assem(Edof(1,:),K,Ke2)
K =
3000 -3000 0
-3000 3000 0
0 0 0
>> K=assem(Edof(2,:),K,Ke1)
K =
3000 -3000 0
-3000 4500 -1500
0 -1500 1500
9.3 – 3 EXAMPLES
exs1 Static analysis
>> K=assem(Edof(3,:),K,Ke2)
K =
3000 -3000 0
-3000 7500 -4500
0 -4500 4500
The global system of equations is solved considering the boundary conditions given
in bc.
a =
0
0.0133
0
r =
-40.0000
0
-60.0000
Element forces are evaluated from the element displacements. These are obtained
from the global displacements a using the function extract.
>> ed1=extract(Edof(1,:),a)
ed1 =
0 0.0133
>> ed2=extract(Edof(2,:),a)
ed2 =
0.0133 0
>> ed3=extract(Edof(3,:),a)
ed3 =
0.0133 0
EXAMPLES 9.3 – 4
Static analysis exs1
>> es1=spring1s(ep2,ed1)
es1 =
40
>> es2=spring1s(ep1,ed2)
es2 =
-20
>> es3=spring1s(ep2,ed3)
es3 =
-40
9.3 – 5 EXAMPLES
exs2 Static analysis
Purpose:
Analysis of one-dimensional heat flow.
Description:
Consider a wall built up of concrete and thermal insulation. The outdoor temperature
is −17◦ C and the temperature inside is 20◦ C. At the inside of the theral insulation
there is a heat source yielding 10 W/m2 .
Ti = 20°C
heat source 10W/m2
T0 = -17°C
T 1 T2 T3 T4 T 5 T6
1 2 3 4 5
The wall is subdivided into five elements and the one-dimensional spring (analogy)
element spring1e is used. Equivalent spring stiffnesses are ki = λA/L for thermal
conductivity and ki = A/R for thermal surface resistance. Corresponding spring
stiffnesses per m2 of the wall are:
A global system matrix K and a heat flow vector f are defined. The heat source inside
the wall is considered by setting f4 = 10. The element matrices Ke are computed
using spring1e, and the function assem assembles the global stiffness matrix.
The system of equations is solved using solveq with considerations to the boundary
conditions in bc. The prescribed temperatures are T1 = −17◦ C and T6 = 20◦ C.
>> Edof=[1 1 2
2 2 3;
3 3 4;
4 4 5;
5 5 6];
EXAMPLES 9.3 – 6
Static analysis exs2
>> K=zeros(6);
>> f=zeros(6,1); f(4)=10
f =
0
0
0
10
0
0
>> [a,r]=solveq(K,f,bc)
a =
-17.0000
-16.4384
-15.8607
19.2378
19.4754
20.0000
r =
-14.0394
0.0000
-0.0000
0
0.0000
4.0394
9.3 – 7 EXAMPLES
exs2 Static analysis
The temperature values Ti in the node points are given in the vector a and the
boundary flows in the vector r.
After solving the system of equations, the heat flow through the wall is computed
using extract and spring1s.
>> ed1=extract(Edof(1,:),a);
>> ed2=extract(Edof(2,:),a);
>> ed3=extract(Edof(3,:),a);
>> ed4=extract(Edof(4,:),a);
>> ed5=extract(Edof(5,:),a);
>> q1=spring1s(ep1,ed1)
q1 =
14.0394
>> q2=spring1s(ep2,ed2)
q2 =
14.0394
>> q3=spring1s(ep3,ed3)
q3 =
14.0394
>> q4=spring1s(ep4,ed4)
q4 =
4.0394
>> q5=spring1s(ep5,ed5)
q5 =
4.0394
The heat flow through the wall is q = 14.0 W/m2 in the part of the wall to the left
of the heat source, and q = 4.0 W/m2 in the part to the right of the heat source.
EXAMPLES 9.3 – 8
Static analysis exs3
Purpose:
Analysis of a plane truss.
Description:
Consider a plane truss consisting of tree bars with the properties E = 200 GPa,
A1 = 6.0 · 10−4 m2 , A2 = 3.0 · 10−4 m2 and A3 = 10.0 · 10−4 m2 , and loaded by a
single force P = 80 kN. The corresponding finite element model consists of three
elements and eight degrees of freedom.
4 8
3 7
3
1.2 m 2
2 6
1 5
1
P
1.6 m
>> Edof=[1 1 2 5 6;
2 5 6 7 8;
3 3 4 5 6];
>> K=zeros(8);
f=zeros(8,1); f(6)=-80e3;
The element property vectors ep1, ep2 and ep3 are defined by
>> E=2.0e11;
>> A1=6.0e-4; A2=3.0e-4; A3=10.0e-4;
>> ep1=[E A1]; ep2=[E A2]; ep3=[E A3];
and the element coordinate vectors ex1, ex2, ex3, ey1, ey2 and ey3 by
9.3 – 9 EXAMPLES
exs3 Static analysis
The element stiffness matrices Ke1, Ke2 and Ke3 are computed using bar2e.
>> Ke1=bar2e(ex1,ey1,ep1)
Ke1 =
1.0e+007 *
7.5000 0 -7.5000 0
0 0 0 0
-7.5000 0 7.5000 0
0 0 0 0
>> Ke2=bar2e(ex2,ey2,ep2)
Ke2 =
1.0e+007 *
0 0 0 0
0 5.0000 0 -5.0000
0 0 0 0
0 -5.0000 0 5.0000
>> Ke3=bar2e(ex3,ey3,ep3)
Ke3 =
1.0e+007 *
Based on the topology information, the global stiffness matrix can be generated by
assembling the element stiffness matrices
>> K=assem(Edof(1,:),K,Ke1);
>> K=assem(Edof(2,:),K,Ke2);
>> K=assem(Edof(3,:),K,Ke3)
EXAMPLES 9.3 – 10
Static analysis exs3
K =
1.0e+008 *
Columns 1 through 7
0.7500 0 0 0 -0.7500 0 0
0 0 0 0 0 0 0
0 0 0.6400 -0.4800 -0.6400 0.4800 0
0 0 -0.4800 0.3600 0.4800 -0.3600 0
-0.7500 0 -0.6400 0.4800 1.3900 -0.4800 0
0 0 0.4800 -0.3600 -0.4800 0.8600 0
0 0 0 0 0 0 0
0 0 0 0 0 -0.5000 0
Column 8
0
0
0
0
0
-0.5000
0
0.5000
a =
1.0e-002 *
0
0
0
0
-0.0398
-0.1152
0
0
9.3 – 11 EXAMPLES
exs3 Static analysis
r =
1.0e+004 *
2.9845
0
-2.9845
2.2383
0.0000
0.0000
0
5.7617
The vertical displacement at the point of loading is 1.15 mm. The section forces es1,
es2 and es3 are calculated using bar2s from element displacements ed1, ed2 and ed3
obtained using extract.
>> ed1=extract(Edof(1,:),a);
>> N1=bar2s(ex1,ey1,ep1,ed1)
N1 =
-2.9845e+004
>> ed2=extract(Edof(2,:),a);
>> N2=bar2s(ex2,ey2,ep2,ed2)
N2 =
5.7617e+004
>> ed3=extract(Edof(3,:),a);
>> N3=bar2s(ex3,ey3,ep3,ed3)
N3 =
3.7306e+004
i.e., the normal forces are N1 = −29.84 kN, N2 = 57.62 kN and N3 = 37.31 kN.
EXAMPLES 9.3 – 12
Static analysis exs4
Purpose:
Analysis of a plane truss.
Description:
Consider a plane truss, loaded by a single force P = 0.5 MN.
2m
A = 25.0 · 10−4 m2
E = 2.10 · 105 MPa
P = 0.5 MN
2m 2m 30o
The corresponding finite element model consists of ten elements and twelve degrees
of freedom.
y
2 6 10
1 1 5 3 9
● ● ●
7
5 6
10
4 8 12
9.
3 2 7 11 x
4
● ● ●
>> Edof=[1 1 2 5 6;
2 3 4 7 8;
3 5 6 9 10;
4 7 8 11 12;
5 7 8 5 6;
6 11 12 9 10;
7 3 4 5 6;
8 7 8 9 10;
9 1 2 7 8;
10 5 6 11 12];
9.3 – 13 EXAMPLES
exs4 Static analysis
A global stiffness matrix K and a load vector f are defined. The load P is divided
into x and y components and inserted in the load vector f.
>> K=zeros(12);
>> f=zeros(12,1); f(11)=0.5e6*sin(pi/6); f(12)=-0.5e6*cos(pi/6);
The element matrices Ke are computed by the function bar2e. These matrices are
then assembled in the global stiffness matrix using the function assem.
>> Ex=[0 2;
0 2;
2 4;
2 4;
2 2;
4 4;
0 2;
2 4;
0 2;
2 4];
>> Ey=[2 2;
0 0;
2 2;
0 0;
0 2;
0 2;
0 2;
0 2;
2 0;
2 0];
All the element matrices are computed and assembled in the loop
The displacements in a and the support forces in r are computed by solving the
system of equations considering the boundary conditions in bc.
EXAMPLES 9.3 – 14
Static analysis exs4
a =
0
0
0
0
0.0024
-0.0045
-0.0016
-0.0042
0.0030
-0.0107
-0.0017
-0.0113
r =
1.0e+005 *
-8.6603
2.4009
6.1603
1.9293
0.0000
-0.0000
-0.0000
-0.0000
0.0000
0.0000
0.0000
0.0000
The displacement at the point of loading is −1.7 · 10−3 m in the x-direction and
−11.3 · 10−3 m in the y-direction. At the upper support the horizontal force is
−0.866 MN and the vertical 0.240 MN. At the lower support the forces are 0.616
MN and 0.193 MN, respectively.
Normal forces are evaluated from element displacements. These are obtained from
the global displacements a using the function extract. The normal forces are evaluated
using the function bar2s.
ed=extract(Edof,a);
9.3 – 15 EXAMPLES
exs4 Static analysis
>> N
N =
1.0e+005 *
6.2594
-4.2310
1.7064
-0.1237
-0.6945
1.7064
-2.7284
-2.4132
3.3953
3.7105
>> Coord=[0 2;
0 0;
2 2;
2 0;
4 2;
4 0];
>> Dof=[ 1 2;
3 4;
5 6;
7 8;
9 10;
11 12];
>> [ex,ey]=coordxtr(Edof,Coord,Dof,2);
EXAMPLES 9.3 – 16
Static analysis exs5
Purpose:
Analysis of a simply supported beam.
Description:
Consider the simply supported beam loaded by a single load f = 10000 N, applied
at a point 1 meter from the left support. The corresponding finite element mesh is
also shown. The following data apply to the beam
f = 10 kN
9m
2 5 8 11
1 4 7 10
3 6 9 12
1 2 3
>> Edof=[1 1 2 3 4 5 6
2 4 5 6 7 8 9
3 7 8 9 10 11 12];
The system matrices, i.e. the stiffness matrix K and the load vector f, are defined by
The element property vector ep, the element coordinate vectors ex and ey, and the
element stiffness matrix Ke, are generated. Note that the same coordinate vectors
are applicable for all elements because they are identical.
9.3 – 17 EXAMPLES
exs5 Static analysis
>> Ke=beam2e(ex,ey,ep)
Ke =
1.0e+008 *
3.1710 0 0 -3.1710 0 0
0 0.0234 0.0351 0 -0.0234 0.0351
0 0.0351 0.0703 0 -0.0351 0.0351
-3.1710 0 0 3.1710 0 0
0 -0.0234 -0.0351 0 0.0234 -0.0351
0 0.0351 0.0351 0 -0.0351 0.0703
Based on the topology information, the global stiffness matrix can be generated by
assembling the element stiffness matrices
>> K=assem(Edof,K,Ke);
Finally, the solution can be calculated by defining the boundary conditions in bc and
solving the system of equations. Displacements a and support forces r are computed
by the function solveq.
>> Ed=extract(Edof,a);
>> es1=beam2s(ex,ey,ep,Ed(1,:));
>> es2=beam2s(ex,ey,ep,Ed(2,:));
>> es3=beam2s(ex,ey,ep,Ed(3,:));
EXAMPLES 9.3 – 18
Static analysis exs5
Results
a = r =
0 1.0e+003 *
0
-0.0095 0
0 6.6667
-0.0228 -0.0000
-0.0038 0
0 0.0000
-0.0199 -0.0000
0.0047 0
0 0.0000
0 0.0000
0.0076 0
3.3333
-0.0000
es1 =
1.0e+004 *
0 -0.6667 0.0000
0 -0.6667 2.0000
es2 =
1.0e+004 *
0 0.3333 2.0000
0 0.3333 1.0000
es3 =
1.0e+004 *
0 0.3333 1.0000
0 0.3333 -0.0000
9.3 – 19 EXAMPLES
exs6 Static analysis
Purpose:
Analysis of a plane frame.
Description:
A frame consists of one horizontal and two vertical beams according to the figure.
q0
P
The corresponding finite element model consists of three beam elements and twelve
degrees of freedom.
5 8
4 3 7
6 9
● ●
1 2
2 11
1 10
3● 12 ●
A topology matrix Edof, a global stiffness matrix K and load vector f are defined. The
element matrices Ke and fe are computed by the function beam2e. These matrices
are then assembled in the global matrices using the function assem.
>> Edof=[1 4 5 6 1 2 3;
2 7 8 9 10 11 12;
3 4 5 6 7 8 9];
>> E=200e9;
>> A1=2e-3; A2=6e-3;
>> I1=1.6e-5; I2=5.4e-5;
>> ep1=[E A1 I1]; ep3=[E A2 I2];
EXAMPLES 9.3 – 20
Static analysis exs6
>> Ke1=beam2e(ex1,ey1,ep1);
>> Ke2=beam2e(ex2,ey2,ep1);
>> [Ke3,fe3]=beam2e(ex3,ey3,ep3,eq3);
>> K=assem(Edof(1,:),K,Ke1);
>> K=assem(Edof(2,:),K,Ke2);
>> [K,f]=assem(Edof(3,:),K,Ke3,f,fe3);
The system of equations are solved considering the boundary conditions in bc.
a = r =
0 1.0e+004 *
0
0 0.1927
0.0075 2.8741
-0.0003 0.0445
-0.0054 0
0.0075 0.0000
-0.0003 -0.0000
0.0047 -0.0000
0 0
0 0.0000
-0.0052 -0.3927
3.1259
0
The element displacements are obtained from the function extract, and the function
beam2s computes the section forces.
>> Ed=extract(Edof,a);
>> es1=beam2s(ex1,ey1,ep1,Ed(1,:),eq1,21)
es1 =
1.0e+004 *
9.3 – 21 EXAMPLES
exs6 Static analysis
>> es2=beam2s(ex2,ey2,ep1,Ed(2,:),eq2,21)
es2 =
1.0e+004 *
>> es3=beam2s(ex3,ey3,ep3,Ed(3,:),eq3,21)
es3 =
1.0e+004 *
A displacement diagram is displayed using the function eldisp2 and section force
diagrams using the function eldia2.
>> figure(1)
>> plotpar=[2 1 0];
>> eldraw2(ex1,ey1,plotpar);
>> eldraw2(ex2,ey2,plotpar);
>> eldraw2(ex3,ey3,plotpar);
>> sfac=scalfact2(ex3,ey3,Ed(3,:),0.1);
>> plotpar=[1 2 1];
>> eldisp2(ex1,ey1,Ed(1,:),plotpar,sfac);
>> eldisp2(ex2,ey2,Ed(2,:),plotpar,sfac);
>> eldisp2(ex3,ey3,Ed(3,:),plotpar,sfac);
>> axis([-1.5 7.5 -0.5 5.5]);
>> pltscalb2(sfac,[1e-2 0.5 0]);
>> axis([-1.5 7.5 -0.5 5.5]);
>> title(’displacements’)
EXAMPLES 9.3 – 22
Static analysis exs6
>> figure(2)
>> plotpar=[2 1];
>> sfac=scalfact2(ex1,ey1,es1(:,1),0.2);
>> eldia2(ex1,ey1,es1(:,1),plotpar,sfac);
>> eldia2(ex2,ey2,es2(:,1),plotpar,sfac);
>> eldia2(ex3,ey3,es3(:,1),plotpar,sfac);
>> axis([-1.5 7.5 -0.5 5.5]);
>> pltscalb2(sfac,[3e4 1.5 0]);
>> title(’normal force’)
>> figure(3)
>> sfac=scalfact2(ex3,ey3,es3(:,2),0.2);
>> eldia2(ex1,ey1,es1(:,2),plotpar,sfac);
>> eldia2(ex2,ey2,es2(:,2),plotpar,sfac);
>> eldia2(ex3,ey3,es3(:,2),plotpar,sfac);
>> axis([-1.5 7.5 -0.5 5.5]);
>> pltscalb2(sfac,[3e4 0.5 0]);
>> title(’shear force’)
>> figure(4)
>> sfac=scalfact2(ex3,ey3,es3(:,3),0.2);
>> eldia2(ex1,ey1,es1(:,3),plotpar,sfac);
>> eldia2(ex2,ey2,es2(:,3),plotpar,sfac);
>> eldia2(ex3,ey3,es3(:,3),plotpar,sfac);
>> axis([-1.5 7.5 -0.5 5.5]);
>> pltscalb2(sfac,[3e4 0.5 0]);
>> title(’moment’)
displacements
0 0.01
−1 0 1 2 3 4 5 6 7
9.3 – 23 EXAMPLES
exs6 Static analysis
normal force
0 30000
−1 0 1 2 3 4 5 6 7
shear force
0 30000
−1 0 1 2 3 4 5 6 7
moment
0 30000
−1 0 1 2 3 4 5 6 7
EXAMPLES 9.3 – 24
Static analysis exs7
Purpose:
Set up a frame, consisting of both beams and bars, and illustrate the calculations by
use of graphics functions.
Description:
A frame consists of horizontal and vertical beams, and is stabilized with diagonal
bars.
y
14 17
2 P 13 6 16
15 18
2 4
10
8
8 11
1 7 5 10
9 12
1 3
9.
7
2 5
0 1 4
3 6
0 1 x
The frame with its coordinates and loading is shown in the left figure, and the finite
element model in the right. In the following, the statements for analysing the frame
are given as an .m-file.
The matrices of the global system i.e. the stiffness matrix K, the load vector f, the
coordinate matrix Coord, and the corresponding degrees of freedom matrix Dof are
defined by
Coord=[0 0;
1 0;
0 1;
1 1;
0 2;
1 2];
9.3 – 25 EXAMPLES
exs7 Static analysis
Dof=[1 2 3;
4 5 6;
7 8 9;
10 11 12;
13 14 15;
16 17 18];
The material properties, the topology, and the element coordinates for the beams
and bars respectively, are defined by
ep1=[1 1 1];
Edof1=[1 1 2 3 7 8 9;
2 7 8 9 13 14 15;
3 4 5 6 10 11 12;
4 10 11 12 16 17 18;
5 7 8 9 10 11 12;
6 13 14 15 16 17 18];
[Ex1,Ey1]=coordxtr(Edof1,Coord,Dof,2);
ep2=[1 1];
Edof2=[7 1 2 10 11;
8 7 8 16 17;
9 7 8 4 5;
10 13 14 10 11];
[Ex2,Ey2]=coordxtr(Edof2,Coord,Dof,2);
eldraw2(Ex1,Ey1,[1 3 1]);
eldraw2(Ex2,Ey2,[1 2 1]);
The element stiffness matrices are generated and assembled in two loops, one for the
beams and one for the bars. The element stiffness functions beam2e and bar2e use
the element coordinate matrices ex and ey. These matrices are extracted from the
global coordinates Coord by the function coordxtr above.
EXAMPLES 9.3 – 26
Static analysis exs7
for i=1:6
Ke=beam2e(Ex1(i,:),Ey1(i,:),ep1);
K=assem(Edof1(i,:),K,Ke);
end
for i=1:4
Ke=bar2e(Ex2(i,:),Ey2(i,:),ep2);
K=assem(Edof2(i,:),K,Ke);
end
The global system of equations is solved considering the boundary conditions in bc,
and the deformed frame is displayed by the function eldisp2, where the displacements
are scaled by the variable sfac.
Ed1=extract(Edof1,a);
Ed2=extract(Edof2,a);
[sfac]=scalfact2(Ex1,Ey1,Ed1,0.1);
eldisp2(Ex1,Ey1,Ed1,[2 1 1],sfac);
eldisp2(Ex2,Ey2,Ed2,[2 1 1],sfac);
1.8
1.6
1.4
1.2
0.8
0.6
0.4
0.2
0
−0.5 0 0.5 1 1.5
9.3 – 27 EXAMPLES
exs8 Static analysis
Purpose:
Analysis of two dimensional diffusion.
Description:
y
c = 10 -3 kg/m3 13 14 15
0.1 m
7 8
10 11 12
5 6
7 8 9
c=0 c=0
3 4
4 5 6
1 2
1 2 3
c=0 0.1 m x
Description:
Consider a filter paper of square shape. Three sides are in contact with pure water
and the fourth side is in contact with a solution of concentration c = 1.0·10−3 kg/m3 .
The length of each side is 0.100 m. Using symmetry, only half of the paper has to
be analyzed. The paper and the corresponding finite element mesh are shown. The
following boundary conditions are applied
c(0, y) = c(x, 0) = c(0.1, y) = 0
c(x, 0.1) = 10−3
The element topology is defined by the topology matrix
>> Edof=[1 1 2 5 4
2 2 3 6 5
3 4 5 8 7
4 5 6 9 8
5 7 8 11 10
6 8 9 12 11
7 10 11 14 13
8 11 12 15 14];
EXAMPLES 9.3 – 28
Static analysis exs8
The system matrices, i.e. the stiffness matrix K and the load vector f, are defined by
Because of the same geometry, orientation, and constitutive matrix for all elements,
only one element stiffness matrix Ke has to be computed. This is done by the function
flw2qe.
>> Ke=flw2qe(ex,ey,ep,D)
>> Ke =
Based on the topology information, the global stiffness matrix is generated by as-
sembling this element stiffness matrix Ke in the global stiffness matrix K
>> K=assem(Edof,K,Ke);
Finally, the solution is calculated by defining the boundary conditions bc and solving
the system of equations. The boundary condition at dof 13 is set to 0.5·10−3 as an
average of the concentrations at the neighbouring boundaries. Concentrations a and
unknown boundary flows r are computed by the function solveq.
>> bc=[1 0;2 0;3 0;4 0;7 0;10 0;13 0.5e-3;14 1e-3;15 1e-3];
>> [a,r]=solveq(K,f,bc);
>> Ed=extract(Edof,a);
9.3 – 29 EXAMPLES
exs8 Static analysis
Results
a= r=
1.0e-003 * 1.0e-003 *
0 -0.0165
0 -0.0565
0 -0.0399
0 -0.0777
0.0662 0.0000
0.0935 0
0 -0.2143
0.1786 0.0000
0.2500 0.0000
0 -0.6366
0.4338 0.0000
0.5494 -0.0000
0.5000 0.0165
1.0000 0.7707
1.0000 0.2542
Es =
-0.0013 -0.0013
-0.0005 -0.0032
-0.0049 -0.0022
-0.0020 -0.0054
-0.0122 -0.0051
-0.0037 -0.0111
-0.0187 -0.0213
-0.0023 -0.0203
EXAMPLES 9.3 – 30
Static analysis exs8
The following .m-file shows an alternative set of commands to perform the diffusion
analysis of exs8. By use of global coordinates, an FE-mesh is generated. Also plots
with flux-vectors and contour lines are created.
K=zeros(15); f=zeros(15,1);
Coord=[0 0 ; 0.025 0 ; 0.05 0
0 0.025; 0.025 0.025; 0.05 0.025
0 0.05 ; 0.025 0.05 ; 0.05 0.05
0 0.075; 0.025 0.075; 0.05 0.075
0 0.1 ; 0.025 0.1 ; 0.05 0.1 ];
Dof=[1; 2; 3
4; 5; 6
7; 8; 9
10;11;12
13;14;15];
eldraw2(Ex,Ey,[1 3 0],Edof(:,1));
pause; clf;
for i=1:8
Ke=flw2qe(Ex(i,:),Ey(i,:),ep,D);
K=assem(Edof(i,:),K,Ke);
end;
bc=[1 0;2 0;3 0;4 0;7 0;10 0;13 0.5e-3;14 1e-3;15 1e-3];
9.3 – 31 EXAMPLES
exs8 Static analysis
[a,r]=solveq(K,f,bc)
Ed=extract(Edof,a);
for i=1:8
Es(i,:)=flw2qs(Ex(i,:),Ey(i,:),ep,D,Ed(i,:))
end
0.1
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01 0.02
0
−0.02 0 0.02 0.04 0.06 0.08
Flux vectors
EXAMPLES 9.3 – 32
Static analysis exs8
0.1
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
−0.02 0 0.02 0.04 0.06 0.08
Contour lines
With the MATLAB functions colormap and fill a color plot of the concentrations can
be obtained.
colormap(’jet’)
fill(Ex’,Ey’,Ed’)
axis equal
9.3 – 33 EXAMPLES
9.4 Dynamic analysis
This section concerns linear dynamic finite element calculations. The examples illustrate
some basic features in dynamics such as modal analysis and time stepping procedures.
Dynamic analysis
exd1 Modal analysis of frame
exd2 Transient analysis
exd3 Reduced system transient analysis
exd4 Time varying boundary condition
Note: The examples listed above are supplied as .m-files under the directory examples.
The example files are named according to the table.
9.4 – 1 EXAMPLES
exd1 Dynamic analysis
Purpose:
Set up the finite element model and perform eigenvalue analysis for a simple frame
structure.
Description:
Consider the two dimensional frame shown below. A vertical beam is fixed at its
lower end, and connected to a horizontal beam at its upper end. The horizontal
beam is simply supported at the right end. The length of the vertical beam is 3 m
and of the horizontal beam 2 m. The following data apply to the beams
8 11 14
2m 7 10 13
9 12 15
3 4
5
4
3m
2
1
3
a) b)
The structure is divided into 4 elements. The numbering of elements and degrees-of-
freedom are apparent from the figure. The following .m-file defines the finite element
model.
EXAMPLES 9.4 – 2
Dynamic analysis exd1
2.5
1.5
y
0.5
0
-0.5 0 0.5 1 1.5 2 2.5
x
9.4 – 3 EXAMPLES
exd1 Dynamic analysis
b=[1 2 3 14]’;
[La,Egv]=eigen(K,M,b);
Freq=sqrt(La)/(2*pi);
Note that the boundary condition matrix, b, only lists the degrees-of-freedom that
are zero. The results of these commands are the eigenvalues, stored in La, and the
eigenvectors, stored in Egv. The corresponding frequencies in Hz are calculated and
stored in the column matrix Freq.
2.5
6.983
1.5
y
0.5
0
-0.5 0 0.5 1 1.5 2 2.5
x
EXAMPLES 9.4 – 4
Dynamic analysis exd1
An attractive way of displaying the eigenmodes is shown in the figure below. The
result is accomplished by translating the different eigenmodes in the x-direction, see
the Ext-matrix defined below, and in the y-direction, see the Eyt-matrix.
9.4 – 5 EXAMPLES
exd2 Dynamic analysis
Purpose:
The frame structure defined in exd1 is exposed in this example to a transient load.
The structural response is determined by a time stepping procedure.
Description:
The structure is exposed to a transient load, impacting on the center of the vertical
beam in horizontal direction, i.e. at the 4th degree-of-freedom. The time history of
the load is shown below. The result shall be displayed as time history plots of the
4th degree-of-freedom and the 11th degree-of-freedom. At time t = 0 the frame is at
rest. The timestep is chosen as ∆t = 0.001 seconds and the integration is performed
for T = 1.0 second. At every 0.1 second the deformed shape of the whole structure
shall be displayed.
force (N)
0.15 sec
1000
The load is generated using the gfunc-function. The time integration is performed
by the step2-function. Because there is no damping present, the C-matrix is entered
as [ ].
dt=0.005; T=1;
% --- the load -----------------------------------------------
G=[0 0; 0.15 1; 0.25 0; T 0]; [t,g]=gfunc(G,dt);
f=zeros(15, length(g)); f(4,:)=1000*g;
% --- boundary condition, initial condition ------------------
bc=[1 0; 2 0; 3 0; 14 0];
d0=zeros(15,1); v0=zeros(15,1);
% --- output parameters --------------------------------------
ntimes=[0.1:0.1:1]; nhist=[4 11];
% --- time integration parameters ----------------------------
ip=[dt T 0.25 0.5 10 2 ntimes nhist];
% --- time integration ---------------------------------------
k=sparse(K); m=sparse(M);
[Dsnap,D,V,A]=step2(k,[],m,d0,v0,ip,f,bc);
The requested time history plots are generated by the following commands
EXAMPLES 9.4 – 6
Dynamic analysis exd2
figure(1), plot(t,D(1,:),’-’,t,D(2,:),’--’)
grid, xlabel(’time (sec)’), ylabel(’displacement (m)’)
title(’Displacement(time) for the 4th and 11th’...
’ degree-of-freedom’)
text(0.3,0.009,’solid line = impact point, x-direction’)
text(0.3,0.007,’dashed line = center, horizontal beam,’...
’ y-direction’)
0.005
-0.005
-0.01
0 0.2 0.4 0.6 0.8 1
time (sec)
The deformed shapes at time increment 0.1 sec are stored in Dsnap. They are visu-
alized by the following commands:
Eyt=Ey-4;
for i=6:10;
Ext=Ex+(i-6)*3; eldraw2(Ext,Eyt,[2 3 0]);
Edb=extract(Edof,Dsnap(:,i));
eldisp2(Ext,Eyt,Edb,[1 2 2],sfac);
Time=num2str(ntimes(i)); text(3*(i-6)+.5,-2.5,Time);
end
9.4 – 7 EXAMPLES
exd2 Dynamic analysis
EXAMPLES 9.4 – 8
Dynamic analysis exd3
Purpose:
This example concerns reduced system analysis for the frame structure defined in
exd1. Transient analysis on modal coordinates is performed for the reduced system.
Description:
In the previous example the transient analysis was based on the original finite element
model. Transient analysis can also be employed on some type of reduced system,
commonly a subset of the eigenvectors. The commands below pick out the first two
eigenvectors for a subsequent time integration, see constant nev. The result in the
figure below shall be compared to the result in exd2.
0.005
−0.005
TWO EIGENVECTORS ARE USED
−0.01
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
time (sec)
9.4 – 9 EXAMPLES
exd3 Dynamic analysis
DsnapR=Egv(:,1:nev)*Dsnapr; DR=Egv(nhist,1:nev)*Dr;
% --- plot time history for two DOF:s ------------------------
figure(1), plot(t,DR(1,:),’-’,t,DR(2,:),’--’)
axis([0 1.0000 -0.0100 0.0200])
grid, xlabel(’time (sec)’), ylabel(’displacement (m)’)
title(’Displacement(time) at the 4th and 11th’...
’ degree-of-freedom’)
text(0.3,0.017,’solid line = impact point, x-direction’)
text(0.3,0.012,’dashed line = center, horizontal beam,’...
’ y-direction’)
text(0.3,-0.007,’2 EIGENVECTORS ARE USED’)
EXAMPLES 9.4 – 10
Dynamic analysis exd4
Purpose:
This example deals with a time varying boundary condition and time integration for
the frame structure defined in exd1.
Description:
Suppose that the support of the vertical beam is moving in the horizontal direction.
The commands below prepare the model for time integration. Note that the structure
of the boundary condition matrix bc differs from the structure of the load matrix f
defined in exd2.
displacement (m)
0.1 sec
0.02
0.00
-0.01
0.2 0.4 0.6 0.8 1.0 time (sec)
dt=0.002; T=1;
% --- boundary condition, initial condition ------------------
G=[0 0; 0.1 0.02; 0.2 -0.01; 0.3 0.0; T 0]; [t,g]=gfunc(G,dt);
bc=zeros(4, 1 + length(g));
bc(1,:)=[1 g]; bc(2,1)=2; bc(3,1)=3; bc(4,1)=14;
d0=zeros(15,1); v0=zeros(15,1);
% --- output parameters --------------------------------------
ntimes=[0.1:0.1:1]; nhist=[1 4 11];
% --- time integration parameters ----------------------------
ip=[dt T 0.25 0.5 10 3 ntimes nhist];
% --- time integration ---------------------------------------
k=sparse(K); m=sparse(M);
[Dsnap,D,V,A]=step2(k,[],m,d0,v0,ip,[],bc);
% --- plot time history for two DOF:s ------------------------
figure(1), plot(t,D(1,:),’-’,t,D(2,:),’--’,t,D(3,:),’-.’)
grid, xlabel(’time (sec)’), ylabel(’displacement (m)’)
title(’Displacement(time) at the 1st, 4th and 11th’...
’ degree-of-freedom’)
text(0.2,0.022,’solid line = bottom, vertical beam,’...
’ x-direction’)
text(0.2,0.017,’dashed line = center, vertical beam,’...
’ x-direction’)
9.4 – 11 EXAMPLES
exd4 Dynamic analysis
0.005
-0.005
-0.01
-0.015
-0.02
0 0.2 0.4 0.6 0.8 1
EXAMPLES 9.4 – 12
Dynamic analysis exd4
9.4 – 13 EXAMPLES
9.5 Nonlinear analysis
This section illustrates some nonlinear finite element calculations.
Nonlinear analysis
exN1 Second order theory analysis of a frame
exN2 Buckling analysis of a frame
Note: The examples listed above are supplied as .m-files under the directory examples.
The example files are named according to the table.
9.5 – 1 EXAMPLES
exn1 Nonlinear analysis
Purpose:
Analysis of a plane frame using second order theory.
Description:
The frame of exs6 is analysed again, but it is now subjected to a load five times
larger than in exs6. Second order theory is used.
q0
P
The element model consisting of three beam elements and twelve degrees of freedom
is repeated here.
5 8
4 3 7
6 9
● ●
1 2
2 11
1 10
3● 12 ●
Edof=[1 4 5 6 1 2 3 ;
2 7 8 9 10 11 12;
3 4 5 6 7 8 9];
EXAMPLES 9.5 – 2
Nonlinear analysis exn1
E=200e9;
A1=2e-3; A2=6e-3;
I1=1.6e-5; I2=5.4e-5;
ep1=[E A1 I1]; ep3=[E A2 I2];
eq3=[-50e3];
The beam element function of the second order theory beam2g requires a normal
force as input variable. In the first iteration this normal force is chosen to zero.
This means that the first iteration is equivalent to a linear first order analysis using
beam2e. Since the normal forces are not known initially, an iterative procedure has
to be applied, where the normal forces N are updated according to the results of
the former iteration. The iterations continue until the difference in normal force of
the two last iteration steps is less than an accepted error eps, (N−N0)/N0 < eps.
The small value given to the initial normal force N (1) is to avoid division by zero
in the second convergence check. If N does not converge in 20 steps the analysis is
interrupted.
K=zeros(12,12);
f=zeros(12,1);
f(4)=10e3;
Ke1=beam2g(Ex(1,:),Ey(1,:),ep1,N(1));
Ke2=beam2g(Ex(2,:),Ey(2,:),ep1,N(2));
[Ke3,fe3]=beam2g(Ex(3,:),Ey(3,:),ep3,N(3),eq3);
K=assem(Edof(1,:),K,Ke1);
K=assem(Edof(2,:),K,Ke2);
[K,f]=assem(Edof(3,:),K,Ke3,f,fe3);
9.5 – 3 EXAMPLES
exn1 Nonlinear analysis
Ed=extract(Edof,a);
es1=beam2gs(Ex(1,:),Ey(1,:),ep1,Ed(1,:),N(1))
es2=beam2gs(Ex(2,:),Ey(2,:),ep1,Ed(2,:),N(2))
es3=beam2gs(Ex(3,:),Ey(3,:),ep3,Ed(3,:),N(3),eq3)
N0=N;
N=[es1(1,1) es2(1,1) es3(1,1)];
if (n>20)
disp(’The solution doesn’’t converge’)
return
end
end
Displacements and element forces from the linear elastic analysis and from the second
order theory analysis respectively:
a = a =
0 0
0 0
0 0
0.0377 0.0452
-0.0014 -0.0014
-0.0269 -0.0281
0.0376 0.0451
-0.0016 -0.0016
0.0233 0.0239
0 0
0 0
-0.0258 -0.0296
EXAMPLES 9.5 – 4
Nonlinear analysis exn1
es1 = es1 =
1.0e+005 * 1.0e+005 *
es2 = es2 =
1.0e+005 * 1.0e+005 *
es3 = es3 =
1.0e+005 * 1.0e+005 *
Using the second order theory, the horizontal displacement of the upper left corner
of the frame increases from 37.7 to 45.2 mm. The moment in the lower left corner
increases from 2.2 kNm to 8.0 kNm.
9.5 – 5 EXAMPLES
exn2 Nonlinear analysis
Purpose:
Buckling analysis of a plane frame.
Description:
The frame of exn1 is in this example analysed with respect to security against buckling
for a case when all loads are increased proportionally. The initial load distribution
is increased by a loading factor alpha until buckling occurs, i.e. the determinant of
the stiffness matrix K passes zero, or the solution does not converge. For each value
of alpha a second order theory calculation of type exn1 is performed. The horizontal
displacement a4 and the moment MA are plotted against alpha. The shape of the
buckling mode is also plotted using the last computed displacement vector before
buckling occurs.
P P
H
A 2, I 2, E E = 200 GPa
4.0 m A1 = 2.0 · 10−3 m2
A1, I1, E A 1, I1, E
I1 = 1.6 · 10−5 m4
A2 = 6.0 · 10−3 m2
A B I2 = 5.4 · 10−5 m4
6.0 m
H = 1.0 kN
P = 150.0 kN
The element model consists of three beam elements and twelve degrees of freedom.
5 8
4 3 7
6 9
● ●
1 2
2 11
1 10
3● 12 ●
Edof=[1 4 5 6 1 2 3 ;
2 7 8 9 10 11 12;
3 4 5 6 7 8 9];
EXAMPLES 9.5 – 6
Nonlinear analysis exn2
E=200e9;
A1=2e-3; A2=6e-3;
I1=1.6e-5; I2=5.4e-5;
ep1=[E A1 I1]; ep3=[E A2 I2];
f0=zeros(12,1);
f0(4)=1e3; f0(5)=-150e3; f0(8)=-150e3;
j=0;
for alpha=1:0.2:10
j=j+1;
N=[0.01 0 0];
N0=[1 1 1];
eps=0.0001;
n=0;
while(abs((N(1)-N0(1))/N0(1))>eps)
n=n+1;
K=zeros(12,12);
f=f0*alpha;
Ke1=beam2g(Ex(1,:),Ey(1,:),ep1,N(1));
Ke2=beam2g(Ex(2,:),Ey(2,:),ep1,N(2));
Ke3=beam2g(Ex(3,:),Ey(3,:),ep3,N(3));
K=assem(Edof(1,:),K,Ke1);
K=assem(Edof(2,:),K,Ke2);
K=assem(Edof(3,:),K,Ke3);
Ed=extract(Edof,a);
9.5 – 7 EXAMPLES
exn2 Nonlinear analysis
es1=beam2gs(Ex(1,:),Ey(1,:),ep1,Ed(1,:),N(1));
es2=beam2gs(Ex(2,:),Ey(2,:),ep1,Ed(2,:),N(2));
es3=beam2gs(Ex(3,:),Ey(3,:),ep3,Ed(3,:),N(3));
N0=N;
N=[es1(1,1),es2(1,1),es3(1,1)];
if(n>20)
disp([’Alpha= ’,num2str(alpha), ...
’: The solution doesn’’t converge.’])
break
end
end
Kred=red(K,bc(:,1));
if (det(Kred)<=0)
disp([’Alpha= ’,num2str(alpha), ...
’: Determinant <= 0, buckling load passed.’])
break
end
if(n>20)
break
end
disp([’Alpha= ’,num2str(alpha),’ is OK! ’, int2str(n), ...
’ iterations are performed.’])
disp([’ ’])
deform(j)=a(4);
M(j)=r(3);
loadfact(j)=alpha;
bmode=a;
end
EXAMPLES 9.5 – 8
Nonlinear analysis exn2
The requested plots of the horizontal displacement, the moment MA , and the shape
of the buckling mode are generated by the following commands
figure(1), clf
plot(deform(:),loadfact(:),’+’,deform(:),loadfact(:),’--’)
axis([0 0.4 0 7]), grid
xlabel(’Horizontal displacement (m)’), ylabel(’alpha’)
title(’Displacement(alpha) for the upper left corner’)
figure(2), clf
plot(M(:),loadfact(:),’+’,M(:),loadfact(:),’--’)
axis([0 0.4e6 0 7]), grid
xlabel(’Moment in A (Nm)’), ylabel(’alpha’)
title(’Supporting moment M-A(alpha)’)
9.5 – 9 EXAMPLES
exn2 Nonlinear analysis
4
alpha
0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
Horizontal displacement (m)
4
alpha
0
0 0.5 1 1.5 2 2.5 3 3.5 4
Moment in A (Nm) 5
x 10
EXAMPLES 9.5 – 10
Index