Professional Documents
Culture Documents
MIT Solidanalisis PDF
MIT Solidanalisis PDF
Finite Element
Procedures for Solids
and Structures
Linear Analysis
Klaus-JOrgen Bathe
Professor of Mechanical Engineering, MIT
Lectures
LECTURE 1
46 MINUTES
I-I
SolIe basic ccnacepls of eugiDeeriDg ualysis
Variational formulation
1-2
Some basic concepts 01 engineering aulysis
INTRODUCTION TO LINEAR
ANALYSIS OF SOLIDS AND STRUCTURES
1·3
Some basic concepts of engineering analysis
REMARKS
Physical problem
Revise (refine)
the model?
I
1
: I,-__S_ol_v_e_th_e_m_o_d_el_ _
I ~
~- - - iL-_I_n_te_r.;..p_re_t_t_h_e_re_s_u_lt_S_....J
1-4
SolIe basic concepts of engiDeering analysis
10 ft
15 ft
I,. 12 at 15°
K~~~~~-~,-Fault
\\(no restraint assumed)
Altered' grit
.,
E= toEc
Analysis of dam.
1·5
Some basic concepts of engineering analysis
B .
t
W
o
E~~;;C=-------_ ....
F
Finite element mesh for tire
inflation analysis.
1·6
SolDe basic concepts of engineering analysis
l,W
PINCHED CYLINDRICAL
p SHELL
~ C -200 -
BENDING MOMENT DISTRIBUTION ALONG DC OF DISPLACEMENT DISTRIBUTION ALONG DC OF
PINCHED CYLINDRICAL SHELL PINCHED CYLINDRICAL SHELL
1-7
SoBle basic concepts 01 engineering analysis
- idealization of system
- formulation of equili
brium equations
- solution of equations
- interpretation of results
1·8
Some basic concepts of engineering analysis
SYSTEMS
1·9
Some basic concepts of engineering analysis
Example:
Physical layout
ELEMENTS
U1 U3
~ : ~\l) .. ..
I 1
,
k, u1 -- F(' )
F(4)
1
'4 [-11-1]["1] . [F1 4' ]
F(4)
3
u2 1 U F(4)
3 3
-,F(2)
F(2 )
--- 2
u,
'2 [
-1
1-1]["I]fF}]
1 u F( 2 )
2
u2
2
- F(S)
2
F(S)
3
-r ]f
F(3) -- 2
'3 [ ]
1 u
1 P
F(3)l]
-1 2 2
1·10
SolIe basic cOIcepls of engineering analysis
Element interconnection
requirements :
F(4) + F(S) = R
3 3 3
KU = E.
Equilibrium equations
KU= R (a)
T
U = [u
··
···
- 1
+k4
T = [R
-R 1 k + k + k ~ -k - k
... 1 2 3.·'" ....2.......3 .......: ..-k....4 ..... ...
K = -k 2 - k3
··
~
..
k2 + k3 + k S~ -k S
· .
... ...............•................•.....•..... ...
·· ..
1·11
Some basic concepts of engineering analysis
~= t
i =1
~(i)
where
: :]
o 0
etc ...
1·12
Some basic concepls 01 engineering analysis
u,
·
................. : :.............. .
· .
... ··· .
.....
. ~
···· ..
·
u,
K1 .
... : :.............. .
· .
K=
... ·· ... .
·· :
U1
............................... : .
~ ~~.~
K ••• ~~
••••••••• ••••••••••• :............... ~1---.JI.l\fl--r/A~,1\1\1\~~r/A
1·13
SOlDe basic concepts of engineering analysis
u,
... ·: .: .
·· ..
K=
·· ..
..
• 'O • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •
··
u,
+ K4 ;
K 1 +K 2 + K 3 ;-K 2 -K 3
'O'O'O · .
'O'O'O'O'O'O'O'O'O'O'O'O'O'O:'O'O'O'O'O'O'O'O'O'O'O'O'O'O:'O'O'O'O'O'O'O'O'O'O'O'O'O'O •
·· ..
K=
· ..
··· ..
'O'O • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •
.
+ K4 ; u,
K 1 +K 2 + K 3 ~-K2 -K 3 -K
4
'O'O • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •
K= -K 2 -K 3 ~K2+ K3 + K5 -K
5
'O'O • • • • • • • • • • • • • • • : •••••••••••••••••••••••••••••
K
I
1·14
Some basic concepts of engineering analysis
an -
~-
0 (b)
1
U=~UT!!!
T
W = U R
KU = R
1·15
SOlDe basic concepts of engineering analysis
PROPAGATION PROBLEMS
m, a a
M= a m2 a
a a m3
EIGENVALUE PROBLEMS
Av = ABv
v is a vector of order n
A is a scalar
1·16
Some basic concepts of engineering analysis
~lU+KU=O
Let
Then we obtain
2
_w ~~ sin W(t-T)
+ -K <p
-
sin W(t-T)= -0
w, ,~,
w3 ' ~3
1·17
ANALYSIS OF
CONTINUOUS SYSTEMS;
DIFFEBENTIAL AND
VABIATIONAL
FOBMULATIONS
LECTURE 2
59 MINUTES
2-1
Analysis 01 continnous systems; differential and variational lonnDlations
2·2
Analysis of continuous systems; differential and variational formulations
BASIC CONCEPTS
OF FINITE • Some additional
ELEMENT ANALYSIS basic concepts are
CONTINUOUS SYSTEMS
used in analysis of
continuous systems
• We discussed some
basic concepts of
analysis of discrete
systems
CONTINUOUS SYSTEMS
differential variational
formulation formulation
t
Weighted residual Ritz Method
methods
Galerkin _.._-----41~_
least squares
....
finite element method
2·3
Analysis of continuous systeDlS; differential and ,arialionalIOl'llulali.
Young's modulus, E
~Lt: ) mass density,
cross-sectional area, A
R.. ~-------
r
~ .-;+~~ dx
I~
Area A, mass density p
2
aA Ix + A ~a
oX X
I
dx - aA Ix = p A au
~
The constitutive relation is
a
au
= E ax
Combining the two equations above
we obtain
2·4
baIysis 01 COitiDlOU systems; differatial aDd variationaliOl'lDDlatiODS
u(x,O} = °
~
at (x ' O) = °
In general, we have
Definition:
We call this problem a Cm-1
variational problem.
2·5
Analysis 01 continuous systems; differential and variatioD,a1 fOl'llolatiODS
We have in general
II=U-W
fL L
II = J } EA
o
au 2
(--)
ax dx -
i
o
u f
B
dx - u
LR
and
u
o=0
and we have 0 II =0
- 6u R = 0
L
or
2·6
lIiIysis of ..IiDIGUS systems; differential and variatiooallormulatioDS
dx + [EA ~I
ax x=L
- EA ~\
dX
x=o
and
EAax-
au I x=L=R
Also f
B
= -A p -
a2u and
, at 2
hence we have
2·7
Analysis of cODtiDaoas syst_ diIIereatial and variatioul fOlllalatiODS
Total Potential IT
I
Use oIT = 0 and essential "b.c.
~
Principle of Virtual _ solve
Displacements problem
I
Integration by parts
~
Differential Equation
of Equilibrium _solve
and natural b.c. problem
2·8
balysis of aDa. syst-: diBerential and variatiouallnaiatiOlS
(3.6)
B.[</>]
1
= q.,
1
= 1 ,2, •••
i
at boundary (3.7)
The basic step in the weighted
residual (and the Ritz analysis)
is to assume a solution of the
form
(3.10)
n
R = r - L2mCL a· f.] (3.11 )
1 =1 1 1
2·9
Analysis 01 C.tinnoDS systems; differential and variational 10000nlations
Galerkin method
fD f. 1
R dD=O ;=1,2, ••• ,n (3.12)
a ;=1,2, ••• ,n
aa.1
2·10
Analysis of continuous systems; differential and variational formulations
Properties
• By invoking 0 II = 0 we minimize
the violation of the internal equilibrium
requirements and the violation of
the natural b.c.
KU =R
Example
2
Area = 1 em
( R = 100 N
........_-
.F-
x,u --- - -
--.;;B;",.,. -- -- - ~---r;;;-==-e-
C
I-...--~~---·-I-
100 em
..- -------·-I
80 em
2·11
Analysis of COitiDlOIS systems; differeatial ad ,ariali" fOllDaialiODS
Here we have
180
IT
1
=
o
12 EA(~)2
ax dx - 100 u Ix = 180
u = a1x + a
2 i
Case 2
u
~
= I1JO 0< x < 100
180
oIT =
1 o
(EA ~~) o(~~) dx - 100 OU Ix=180
=0
£ o
(~~u)( EA ~~) I
dx = 100 OU x=180
J V
ET
-
T
-
dV = IT. F.
1 1
2·12
Analysis of continuous systems; differential and variational formulations
Exact Solution
~
ax (EA ~)
ax = 0
EA ~
ax = 100
x=180
The solution is
2·13
Analysis of continuous systems; differential and variational formulations
Case 1
180
dx+ I
2 f
100
(1+ x-l00)2
40
E [0.4467 116
116 34076
and
128.6
a1 = ---=E=--- a 2 -- - 0.341
E
u 12C.6 0.341 2
= E x - E x
a = 128.6 - 0.682 x
2·14
Analysis of continuous systems; differential and variational formulations
Case 2
Here we have
E
n=2 J 100
(100 u B) dx+ I2
1 2
f
180
(1+ x -l00)2
40
a 100
[15.4 -13]
[~:] [~oo]
E
240 =
-13 13
U
B
= 10000
E
U
c
11846.2
E
and
2-15
Aulysis of COilinDmas systems; diUerenliai and varialiOlla1I01'1BDlaIiGlS
EXACT
15000
-- --- ~
E
10000 ~~.-.-. -::.:--~
E
" Sol ution 2
5000
-E-
100 180
CALCULATED DISPLACEMENTS
(J
100 -I=:::==-==_==_:=os:=_=_=,==_=_==
"" EXACT
"I~
50
~
I
L._._.
-< ,J
~._._
SOLUTION 1
SOLUTION 2
-+ ---,~--------r-------~X
100 180
CALCULATED STRESSES
2·18
balysis of coatiDloas systms; diBerenlial ud variational fonnllatioas
2·17
FORMULATION OF THE
DISPLACEMENT-BASED
FINITE ELEMENT
METHOD
LECTURE 3
58 MINUTES
3·1
Formulation of the displacement-based finite element method
3·2
Formulation of the displacement-based finite element method
FORMULATION OF
THE DISPLACEMENT -The formulation is
BASED FINITE really a modern appli
ELEMENT METHOD cation of the Ritz/
Gelerkin procedures
- A very general discussed in lecture 2
formu lation
-Consider static and
-Provides the basis of dynamic conditions, but
almost all finite ele linear analysis
ment analyses per
formed in practice
3·3
FOl'Dlulation of the displaceDlent·based finite e1mnent lDethod
fB
X f~ i
FX
fB fS i
FZ
Z Z
uT = [u V w] (4.2)
3·4
Formulation of the displacement-based finite element method
where
3·5
Formulation of the displaceaenl-based filile e1eDlenl .ethod
,,
""
x,u
Finite element
"T
!! = [U, V, W, U2V2W2 ••• UNVNW N]
3·&
'OI'IIalation of the displaceDlenl-based filile eleDlenl method
(4.12)
'iTl
If
~
~
1 B(m) Tc(m)B(m)dv(m)j U=
v(m) - l- - £1 ---- ~(m) = f.(m) ~(m)
j [I T L l(m)
T (m)
!!(m) 1.B dV(m)
j (-£ )(m)
-
= B(m) l..u·)
--
I 1
_"
m V
, .
I
_~m
( )T
m
El;rm) - B(m)TTI(m) dv(m)j
-
-US
-(m) T
-___.r__.........1 ... ~
------.
"<I:::
(4.13)
3·7
Formulation of the displacement-based finite element .ethod
We obtain
KU =R (4.14)
where
R=.Ba + Rs - R1 + ~ ( 4. 15)
K= ~f
J
B(m)Tc(m)B(m)dV(m)
- m V(m)- - - - (4.16)
R =
~
"'1.
~ lm) -
H(m)TfB(m)dV(m)
-
(4.17)
R
-S
"'1
= HS (m)Tfs(m)dS(m) (4.18)
~ ~m) - -
R
~
=-F (4.20)
(m)T -B(m)
~B = ~ fV(m) .!:!. [1.
B(m) -B(m) •• (m)
1. = 1. - p!!
_ p(m).!:!.(m)~]dV(m) (4.21 )
MD+KU= R (4.22)
3·8
Formulation of the displacement-based finite element method
~a ~b ~a ~b
+
~a t!t>b -~
= (4.38)
.. ..
~a~+~a~=~-~b~-~b~
(4.39)
~=~a~+~b~+~a~+~b~
(4.40)
i A
!
• V Transformed
I
!
;-
Global degrees
of freedom
- \eedom
degrees of
f
C,-:~:e)
V
~ I
~
(restrained\
rl'f u
a]
L
COS a -sin
T=
[ sin a cos a
'/.. U= T IT
3·9
Formulation of the displacement-based finite element method
.th
1
.th
J
column
For the transformation on the
total degrees of freedom we use
(4.41 )
1. ••
i th row 1
j
cos a.
!
-s ina.
so that
T=
.. 1
Mu+Ku=R (4.42) }h sin a. cos a.
1
where
L
Fig. 4.11. Skew boundary condition
imposed using spring element.
,
k U. = k b (4.44)
where
k » k ..
"
___ 3·10
FormDlation of the displacement·based finite element method
Example analysis
z area =1
100
x y
area =9
100
80
element ®
Finite elements
J~
I" 100
-I
~I
80
3·11
Formulation of the displacement·based finite element method
Element
interpolation functions
1.0
I ... --I
L
y
H(l} = [(l-L)
- 100 100
a]
v(m} = H(m}U
!:!.(2} = [ a (1- L)
80 :0]
!!(l)=[ 1 1 a]
100 100 av = B(m}U
ay - -
1 1
!!(2) = [ a 80 80]
3·12
FOI'IDDlation of the displacement·based finite element method
stiffness matrix
- 1
100
5.= (1 HEllO
a
l~O [-l~O l~O o}Y
a
U
1
- 80
1
80
Hence
E [ 2.4 -2.4
=240 -2.4 15.4
a -13
3·13
GENERALIZED
COORDINATE FINITE
ELEMENT MODELS
LECTURE 4
57 MINUTES
4·1
Generalized coordinate finite element models
Example results
4-2
Generalized coordinate finite eleDlent models
DERIVATION OF SPECIFIC
FINITE ELEMENTS
• Generalized coordinate
~(m) = i
finite element models
V(m)
B(m)T C(m) B(m) dV (m)
In essence, we need
aW) = J
V(m)
H(m)T LB(m) dV (m)
• Convergence of
R(m)
!!S
= f HS(m)T f S(m) dS (m)
(m) - -
analysis results
S
etc.
4·3
Ge.raJized coordiDale finite elementlDOIIeIs
Hale
\
I
6
I
\
-\
\
- 1ZI
\
u(x,y), v(x,y)
are non-zero
w= 0 , E zz = 0
4·4
Generalized coordinate finite element models
Structure and loading
are axisymmetric.
j(
I
I
I
I,
I
\--
I
(before deformation)
(after deformation)
SHELL
4·5
Generalized coordinate finite element models
Displacement
Problem Components
Bar u
Beam w
Plane stress u, v
Plane strain u, v
Axisymmetric u,v
Three-dimensional u,v, w
Plate Bending w
Bar
Problem
(E"...,)
-
Strain Vector ~T
Beam [IC...,]
Plane stress (E"..., El'l' )' "7)
Plane strain (E..., EJ"7 )'..7)
Axisymmetric [E..., E"77 )'''7 Eu )
Three-dimensional [E..., E"77 Eu )'''7 )'76 )'...,)
Plate Bending (IC..., 1(77 1("7)
.
Nolallon: E..
au au
= ax' £7 =
au + au
a/ )'''7 = ay ax'
aw 1 aw 1
••• , IC..., = -dx ' IC = - OyZ,IC.., = 2
aoy
w 1
Z 77
0x
Table 4.2 (b) Corresponding Kine
matic and Static Variables in Various
Problems.
4·&
Generalized coordinate finite element models
Bar [T;u,]
Beam [Mn ]
Plane stress [Tn TJIJI T"'JI]
Plane strain [Tn TJIJI T"'JI]
Axisymmetric [Tn TJIJI T"'JI Tn]
Three-dimensional [Tn TYJI Tn T"'JI TJI' Tu ]
Plate Bending [Mn MJIJI M"'JI]
Bar E
Beam El
~.]
1 v
E v 1
Plane Stress 1-1':&
[o 0 1
4·7
Generalized coordinate finite element models
(4.47)
2
w(x,y) = Y, + Y2 x + Y3Y+ Y4xy + Y5x + •..
(4.48)
For three-dimensional solid elements
4·8
Generalized coordinate finite element models
Hence, in general
u = ~ ex (4.50)
(4.51/52)
(4.53/54)
(4.55)
Example
r Nodal point 6
0
lp
9
Element 0
5 8
CD @
Y.V V7
Y.V
1 4 7
X.V X.V V7
4·9
Generalized coordinate finite element models
2
LJ2.= US --II--.......- - - - - - - - -....~
U{X,y)] (2)
[ v{x,y)
= H(2) u
--
where
uT = [U 1
-
4·10
Generalized coordinate linite element models
or
U(X,y)] =_~l!.
[ v(x,y)
where
and
Defining
we have
Q = Aa.
Hence
H=iPA- 1
4·11
Generalized coordinate finite element models
Hence
H
-
= fll4 (1+x ) ( Hy) :
a
I ••• I
:
:
: (1 +x )( 1+y) :
a
I
and
UJ VJ U z t': u. v.
U2 U3 U4 Us U6 U7 Us U9 U1a
H'ZJ
-
= [00
I 0 : H IJ H 17 : 0 0: HI.
HI.
o : H ZJ H 21 : H:: H: 6 : 0 0: H..
H 16 : Hu :
Ha :
VI -element degrees of freedom
4·12
Generalized coordinate finite element models
where
E - au . E _ av. _ au + av
xx - ax' yy - ay , Yxy - ay ax
Hence
where
I
[~
1 0 y'O
I
0 0 0 1
I 0
I = I
0 1 X 10
I
4·13
Generalized coordinate finite element models
1
MECHANICAL IDEALIZATION
KINEMATICS, e.g. truss
plane stress
three-dimensional
Kirchhoff plate
etc.
MATERIAL, e.g. isotropic linear YIELDS:
elastic
Mooney-Rivlin rubber GOVERNING DIFFERENTIAL
etc. EQUATIONS OF MOTION
LOADING, e.g. concentrated e.g.
centrifugal
etc.
BOUNDARY CONDITIONS, e.g. prescribed
..!.. .!!!)
ax (EA ax = - p(x)
displacements
etc.
1
FINITE ELEMENT SOLUTION
YIELDS:
CHOICE OF ELEMENTS AND APPROXIMATE RESPONSE
SOLUTION PROCEDURES SOLUTION OF MECHANICAL
IDEALIZATION
4·14
Generalized coordinate finite element models
SECTION
ERROR ERROR OCCURRENCE IN discussing
error
4·15
Generalized coordinate finite element models
CONVERGENCE ~ compatible
Assume a compatible
LW layout
element layout is used,
then we have monotonic
convergence to the
solution of the problem CD incompatible
layout
~
governing differential
equation, provided the
elements contain:
t:=
1) all required rigid
body modes
2) all required constant
strain states
no. of elements
If an incompatible element
layout is used, then in addition
every patch of elements must
be able to represent the constant
strain states. Then we have
convergence but non-monotonic
convergence.
4·16
Geuralized coordinate finite e1eJDeDt models
I , 7 "
I
I
I
I
/ " 'r-
i I (
>
I I 1-- - -- " ,;
" /
......~_Q
I I
I I
Rigid body
translation
and rotation;
element must
be stress
free.
4·17
Generalized coordinate filite elellent .adels
10
Young's r------,
I modulus = 1.0 I
I
I -l Poisson's
1.
t I
I
I
I
I
ratio" 0.30
01 I
I
I
I -------
I
• I
_1
Rigid body mode Al = 0 Rigid body mode A2 = 0
\
-- ...--,.".-~\
\
\
\ (' --- --
_-I
I
I
\ \ I
\ I
\ \
\ \ \
-- -
\ .J \ I
\,. I
\ ..... ..... f
..... I
'J
Rigid body mode A3 =0 Flexural mode A4 =0.57692
\
\
\
\
~-
...... "
\
\ \
\
I
'-
.----"'"="""- \
--~
\
\
\
\...~
-, \
.... .... \~
,-----, I
r-------- 1
I I I I
I I I
I
I I I I
I I I :
I I I I
I I I I
I I
I I IL .JI
4·18
Generalized coordiDate finite element lDodels
(0 ·11
® G)
~
17
'c.
IT
~
,I>
® ®-. ® .f:
/ @
@)
20
IS
Point 2
Oyy(N/m )
A 1066
B 716
C 359
D 1303
E 1303
4·19
IMPLEMENTATION or
METHODS IN
COMPUTER PROGRAMS;
EXAMPLES SAP, ADINA
LECTURE 5
56 MINUTES
5·1
"pi_entation of metllods in computer prograDlS; examples SIP, ADlRA
Out-of-core solution
5·2
l_pIg_talioa of _ethods in CODIpDter program; mDlples SAP, ADINA
IMPLEMENTATION OF
T
THE FINITE ELEMENT K(m) = 1. B(m) C(m)B(m) dV(m)
- V(m)- --
METHOD
R(m) = 1. H(m)T fB(m) dV (m)
-B v(m) - -
We derived the equi
librium equations H(m) B(m) N = no. of d.o.f.
- -
of total structure
kxN .hN
tl ~
kxn R,xn
5·3
IDlpl_81taliol of Dlethods in toDlpuler progrw; mDlples SAP, ADIlI
l Sz :: 6
t W :: 3
Z
x
r----- y V::2 Sy:: 5
/U:: 1
Fig. A.1. Possible degrees of
/Sx:: 4 freedom at a nodal point.
ID(I,J) =
_. - nodal point
... -
I-
Degree of
freedom
....i
5·4
. . . .taIiOi of IIeIWs in coapler P.... UUlples SIP, ABilA
E· lOS N/em
2 .,-0.15
2
t
4
E. 2 x l(Jl1 N lem2
II'" 0.20
5 _ _3
10
8 -9
t
number
<D ®
E = l(Jl1 N/cm 2
., - 0.15
2 E = 2 x l(Jl1 N Icm2
t" =0.20
t
8
4 _1 7 -7
~
Node Temperature at
\
Degree of
bottom face = 70'C freedom
number
1 1 1 0 0 0 0 0 0
1 1 1 0 0 0 0 0 0
1 1 1 1 1 1 1 1 1
10 =
1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 ] 1
5·5
IJDpleDIeDtatiOD of methods in CODIpater.programs; examples SAP, ADIIA
and then
0 0 0 1 3 5 7 9 11
0 0 0 2 4 6 8 10 12
0 0 0 0 0 0 0 0 0
10=
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
Also
5·6
Implementation of methods in computer programs; examples SAP, ADINA
For compacted I
matrix 1 2 3 4 5 6 7 8
..,
For !1 oJ 4 0 0 0 0 1 2
LMT = [3 4 0 0 0 0 1 2]
5·7
Implementation of methods in computer programs; examples SAP, ADINA
L MT = [5 6 0 0 0 0 3 4]
for element 3,
T
L M = [9 10 3 4 1 2 7 8]
,. mK =3
" ·1
SkYline
k l l k 12 0 k 14 .~ 0 0 0
kn k 23
k 33
0
k 34
o"
o
J"0
k 36
0
------m
'0 0 6
=3
0
'-
k 44 k 45 k 46 0" 0
K=
k ss k S6 Fig. A.3. Storage scheme used for a
k 66 typical stiffness matrix.
Symmetric
5·8
_pl• •taiioo of lDeIJaods in COIDpater prograJDS; eDIIIples SAP, ADINA
. - - , COLUMN HEIGHTS
I I I
X 0 1
0 0 10 0 0
x = NONZERO ELEMENT o 0 0 0:0 0:0
0= ZERO ELEMENT xix x 010 0 x
XIX 0 010 0 0
XIX 0 0 X 0 0
X 0 X 10 0 0
I xxlxXIO
xix XiX
SYMMETRIC IX l
X X
XIX
IX
I
X0
X0
XIX
XiX
XiX
BLOCK 1
I
X
BLOCK 2 ~---~
~_BLOCK 4
5·9
IIIlpl• •tation of methods in computer programs; examples SAP, ABilA
7 9 10
,~ 1 2 3 4 5 6 8 11 12 13
14 15 16 17 18 19 20
~ 21 22 23 24 25 26 27 28 29 30 31 32 33
(a) Bad nodal point numbering,
mk + 1 = 46.
1 4 6 9 11 14 16 19 21 24 26 29 31
2 7 12 17 22 27 32
~,
, 3 5 8 10 13 15 18 20 23 5 283033
2
5·10
"pI• •tation of Ilethods in cOIlpuler program; exallples SAP, ADINA
START
END
5-11
Implementation of methods in computer programs; examples SAP, ADINA
I
-'-------4
ONE - DIMENSIONAL
ELEMENT
! RING ELEMENT
; - - . - - - - - - - - - - - - -... y
z 2
5-12
Implementation of metbods in computer programs; examples SAP, ADINA
y 2
---5
y
Fig. 15. Three-dimensional beam
element
p A.45.
5·13
Implementation of methods in computer programs; examples SAP, ADINA
-- --. __e_
•
---L~---
--
•
• 3-16 NODES
y
TRANSITION
x ELEMENT
5·14
FOBMULATION AND
CALCULATION OF
ISOPABAMETBIC
MODELS
LECTURE 6
57 MINUTES
6·1
FOl'Dlolation and calculation of isoparmetric models
6·2
FOI'DlDlatiOl ud calculation of isopariUHbic models
FORMULATION AND
CALCULATION OF ISO
PARAMETRIC FINITE
ELEMENTS
interpolation matrices
and element matrices
-We considered earlier
(lecture 4) generalized
coordinate finite
element models
lsoparametric Elements
Basic Concept: (Continuum Elements)
Interpolate Geometry
N N N
x=L h. x. ; y=
I I
L h. y. ;
I I
z=L h. z.
I I
i=l i =1 i=l
Interpolate Displacements
N N N
u= 1: h. u.
I I
v= L h.v.
I I
w= L h.w.
I I
i =1 i == 1 i =1
N = number of nodes
&·3
Formulation and calculation of isoparametric models
6·4
FOI'Ilalation and calcalatioD 01 isoparametric models
~~==-=l======~I=-==r======r ~1 ==
6·5
F. .utioa and calculation of isoparalDebic lDodels
S
1 1
Ill( ~
ll(
~ J
-
1
-
2/D element, 2x2 units
Similarly 3/D element 2x2x2 units
(r-s-taxes)
1 - D Element
2 Nodes:
-11.0
~~ -+- .. _ h1 = %(1 + r)
2 -r 1
-r
Formulation ud calculation 01 isoparUletric lIodeis
1.0
-- - -
e_----......:::::...::::=--
-...
-..:...::-:::;. h2 = Y.z(1- r) - Y.z(1- r 2 )
2 3 1
2 - 0 Element
4 Nodes:
/-r-r----+~~-r
h1 =~(1 + r)(1 + 5)
3 4
Similarly
h =%(1- r)( 1 + 5)
2
h3 =%(1- r)(1- 5)
h4 = %(1 + r)(1-s)
6-7
Formulation and calculation of isoparametric models
first obtain h
S:
....--+-+--+~_ ..... 1
-+--+-I--I--I---I--------I--.._r
h
1
= %(1 + r)(1 + s)
-%h S
3 4
6·8
Formulation and calculation of isoparametric models
y \
\ s=o
6 ---;q----- ---..
8 r
\
\
3
\
\ r =0 r = +1
r =-1
x
(a) Four to 9 variable-number-nodes
two-dimensional element
6·9
Fonnulation and calculation of isoparametric models
x==r
y == s
z == t
6·10
Formulation and calculation of isoparametric models
ah 0 ah 0
1 4 u
ar ar 1
E
l
r
ah ah v1
1 4
0 0
E
SS as as u
2
[Y
rs ah ah 3h ah
1 1 4 4
as at' as ar v4
\.. I
v-
B
We note again r==x
s=y
GENERAL ELEMENTS
s r = +1
s = +1
Y,v
r - - - t - - - 4 _r •
6·11
Formulation and calculation of isoparametric models
[ : ] = [::
as as :]l~]
as ay Aside:
cannot use
a a ar
or
ax ar ax + ...
---
-aar = -J a
ax (in general)
a _ 1 a
a-x- J- ar (5.25)
dv = det J dr ds dt
6·12
FOI'Dlalation ud calculation of isoparmebic .odels
Element 1
z.._----+----......- " · r
...
3, '4-
+-
"""1--1-----------t..~1
6 em.
X
Element 2
I
2.
.....------- 1+
....
.;
3 o
cDG-W\ J =
1 1
213 2
&-13
Formulation and calculation of isoparametric models
Element 3
\c.1V\
2.
-+---~~- -,c
1c.W'I : '3,.~
't' .....,..
.I.
...L... -"'------14-
31
I•
2 c l"l1
-, (1 +5)]
(3+r)
r=-I
Natural space
3
•
I I
,-. -I'
L/4
6·14
Formulation and calculation of isoparametric models
Here we have
3
x=L: h . x. 9
1 1
L
x =-4(1+r ) 2
i =1
hence
-J = [!:..2 + !'-2 LJ
and
or
Since
r = 2.Jf- 1
We note
1 singularity at X = 0 !
/x
6·15
Formulation and calculation of isoparaDlebic Dlodels
Numerical Integration
Gauss Integration
Newton-Cotes Formulas
-K = !:J
' " a··k
I,J,k
F·· k
IJ -IJ
x
6·16
FORMULATION OF
STRUCTURAL
ELEMENTS
LECTURE 7
52 MINUTES
7·1
Formulation of structural elements
7·2
FOI'IIDlati.... slnclDrai e1U11DIs
Continuum
" particles remain on
Approach
a straight line during
deformation"
Use the general
principle of virtlial e.g.
displacements, but beam
7·3
Formulation of structural elements
Neutral
axis
Beam
.. section Boundary conditions between
x beam elements
Deformation of cross-section dw _ dw
wi = wi ; dx -0 - dx +0
x-0 x+0 x x
./
Neutral
axis
WI - Wi
x- O x+ O
Beam
section
7·4
Formulation of structural elements
We use
dw
S=--y
dx
(5.48)
(5.49)
L
_ (L
J
o
pw dx -Lo m S dx
(5.50)
L
+ GAkJ (~~ - S) o(~~ - S) dx
o
-io
L
p oW dx -io L
m oS dx = 0
(5.51)
7·5
Formulation of structural elements
7·6
Formulation of structural elements
q q
W = ~ h.w. B = ~ h.e. (5.52)
L..J 1 1 L..J 1 1
i =, i =,
w=1
H- /U'
-'
B = .:...:.s-
H U
(5.53)
dW = BU' ~ =B U
dX 1-/ - ' dX ~-
Where
T
Q. = [w, W
q
8, 8qJ
~ = [h, hq 0 OJ
~ = [0 0 h, hqJ (5.54)
and
7·7
Formulation of structural elements
So that
1 T
K = E1
f
-1
~ ~ det J dr
+ GAk
t -1
T
(~-tla) (~-~)det J dr
(5.56)
and
R=
f
-1
~ p det J dr
+/ -1
~ m det J dr (5.57)
GAk (5.60)
ex. = IT
Hence
- use parabolic (or higher-order)
elements
. discrete Kirchhoff theory
- reduced numerical integration
7-8
Formulation of structural elements
Here we use
q
+~ ' b h Q,V k
2 L.- k k sx
k=l
q q
Q,y(r,s,t) = L hk Q,Yk +i L akh k Q,V~y
k=l k=l
q (5.61)
So that
1 0
u (r,s,t) = x- x
v (r,s,t) = ly _ 0y (5.62)
w (r,s,t) =
1z- 0z
and
q t q k
u(r,s, t) = L: hku k +"2 L: akh k Vtx
k=l k=l
q
+ t .E bkh k V~x
k=l
q t q
v(r,s,t)=L: hkv k +2 L
k=l k=l
q
+tL:
k=l
q
w(r,s,t)=L:
k=l
(5.63)
7·10
Formulation of structural elements
vk = ~ e
ak
...:..s
x
'is (5.65)
where
exk
e =
~
eyk (5.66)
ezk
£nn
q
where
T=
~
u [Uk vk wk exk eyk ezk ] (5.68)
and then also have
T E a a £
nn nn
Tn~
= a Gk a Yn~
TnI';; 0 a Gk Ynl;;
(5.77)
7-11
Formulation of structural elements
.... -- ----
Hence
dl\
E
XX dX
dS
E = z _-.1. (5.79)
yy dy
dS X _ dS y
Yxy dy dX
dW
Yyz dy - Sy
= (5.80)
dW+ S
-
Yzx dX x
7·12
Formulation of structural elements
and
LXX 1 v a
L = z_E_ v 1 a
yy 2
l-v
L a a l-v
xy 2
(5.81)
L
aw
yz ay - By
E
= 2(1+v) (5.82)
L aw + B
ZX ax x
II=-
1 dz dA
2
L
xy
-fw P dA
A
(5.83)
7·13
Formulation of structural elements
IT = t iTA
.<q, .<dA + t // f,;
A
y dA
-I: A
P dA (5.84)
where
as
_ .-J.-
K = ; y = (5.86)
ay
aw + s
as x _ ~ ax x
ay ax
1 v 0
C =.
Eh 3 v 1 0
~ 12(l-v 2 ) 1-v
0 0
2
1
Ehk (5.87)
f.s = 2{1+v) [ 0
7·14
Formulation of structural elements
-fw p dA =0
A (5.88)
q
w=~h.w.
LJ 1 1
i=l
q
=~
S
y LJ h.1 exi (5.89)
i=l
and
q q
x ~h.x.
= LJ Y=~h.y.
1 1 LJ 1 1
i=l ;=1
7·15
Formulation of structural elements
s
Mid-surface
r
\....-~-----t~
(5.90)
7·16
Formulation of structural elements
Therefore,
where (5.91)
(5.92)
To express Y~ in terms of
rotations at the nodal- point k
we define
k
°V
-1 = (e-y x Ov-nk ) / Ie- yx-°Vkl
n (5.93a)
then
k k
V
..:...n = - °Vk
~ O',k + °V
-1 Sk (5.94)
7·17
Finally, we need to recognize the
use of the following stress-strain
law
l = ~h ~ (5.100)
1 v a a a a
1 a a a a
Jl a a a
1_~2
T
~h=~h ( 1-v
-2- a a ) !2sh
1-v
-2- a
symmetric 1-v
2
(5.101)
16· node parent element with cubic interpolation 2
I- 5 -I
• •
2
• •
64£>-[>
000
o \'.' .\
Variable - number - nodes shell element
7·18
Formnlalion of structural elements
a) Shell intersections
7·19
NUMERICAL
INTEGRATIONS,
MODELING
CONSIDERATIONS
LECTURE 8
47 MINUTES
8·1
Numerical integrations, modeling considerations
Practical considerations
Calculation of stresses
8·2
Numerical integrations. modeling considerations
NUMERICAL INTEGRATION.
SOME MODELING CONSIDERATIONS
• Gauss integration
• Practical considerations
• Choice of elements
We had
K = f BT C B dV (4.29)
- V - --
M = J p HT H dV (4.30)
- V --
R=
~
f HT fB dV
V- -
(4.31 )
T
HS fS dS
-s = Sf
R
- -
(4.32)
8·3
Numerical integrations, modeling considerations
+1 +1
M= ff p tl T tt det J dr ds
-1 -1
etc...
8·4
Numerical integrations, modeling considerations
-K=L~a.·.F
. 4J lJ -lJ..
1 J
where
i, j denote the integration points
a. .. = weight coefficients
IJ
F··
-IJ B·· T -C -IJ
= -IJ B·· detJ··
~J
- r
- r = ±O.577
5 = ±O.577
, r = ±O.775 5 =± 0.775
\ \ r= 0 5=0
8·5
Numerical integrations. modeling coDSideratiODS
I
I
x
a a+b b
-2-
.-8
Numerical integrations, modeling considerations
I actual function F
2 nd order interpolating
~~~~polynomial in x .
a a+b b
2
etc....
(5.123)
n = number of intervals
interpolating polynomial is of
order n.
8·7
Numerical integrations, modeling considerations
Upper Bound on
Error R. as
Number of a Function of
Intervals n q q Cn2 cn3 q C·5 Cn6 the Derivative of F
1 1
"2 T 10-I(b-a}lF"(r)
1 4 1 10-3(b-a)5PV(r)
2 6" 6" 6"
1 3 3 1
3 "8 "8 "8 "8 1O-3(b-a)5F'V(r)
7 32 12 32 7 10-6(b-a)7FVI(r)
4 90 90 90 90 90
19 75 50 50 75 19 10-6(b-a)7Fv'(r)
5 288 288 US 288 ill 288
41 216 27 272 27 216 41 lO-'(b-a)'FVIU(r)
6 840 840 840 840 840 840 840
f a
b
F(r)dr" U F(r 1 ) + u2 F(r2 ) + ••.
1
8·8
Numerical integrations, modeling considerations
n rj /X,
Now let,
ri be a sampling point and
eli be the corresponding weight
8·9
Numerical integrations, modeling considerations
f
-1
+1
f
-1
+1
F(r,s) dr ds = I:1
"1 f
-1
+1
F(ri's) ds
(5.131)
or
f f
-1
+1
-1
+1
F(r,s)drds= I:
i ,j
,,;,,/(ri'sj)
(5.132 )
ff1
-1
+1
-1
+1
-1
+1
F(r,s,t}drdsdt
~a.·a.·a.kF(r.,s.,tk)
= LJ 1 J 1 J
i,j,k
(5.133 )
and a··k = a. Q. Q .
IJ I J k
8·10
Numerical integrations, modeling considerations
8·11
Numerical integrations, modeling considerations
Demonstrative example
Stress calculations
(5.136)
8-12
Numerical integrations. modeling considerations
thickness = 1 cm
A
-p = 3xl0 7 N/cm
2
1~[
E
<3> e. CD \) = 0.3
I c· 1> p = 300 N
:... 3c.m.
..,- 3 Coft'1.
-of
'100 N!Crrt'l.
A /
8 ... = a.
8·13
Numerical integratiODS. modeling coDSideratioDS
'?
, A
,
,~
,
@ B
C.
<D
4l v = 0.3
I
, s_ a~ "? P = lOON
" "
8 & <D
Co Co
174+ /lA/e-t'-
A A
® B 8 <D
Co
c
~I ".00 "'Ie-."
We need
• parabolic/undistorted elements
usually most effective
8-14
Numerical integrations, modeling considerations
TWO-DIMENSIONAL 8-node or
PLANE STRESS
PLANE STRAIN
AXISYMMETRIC
9-node
D
D
THREE-DIMENSIONAL 20-node
4-node
..... -/
PLATE 9-node
SHELL 9-node or
16-node
L7
~~
8·15
Numerical integrations, modeling considerations
4/'1ode
elEJmerrt
S node
e 1er1l(1'It.
.
i
g I'\oole
~kl'7ll'"t
~
I
1
J
I
119
8 B U.s
4- I\oc:(e
ele,"~"t.
VA
4 node
A A
e Iem tnt"" 'Ve- llA ~
4- node
el .... ~I\"t
c C U,
Constraint uA = (uC + uB)/2
equations:
vA = (vC + vB)/2
/.
!
8·16
SOLUTION OF
FINITE ELEMENT
EQUILIBRIUM
EQUATIONS
IN STATIC ANALYSIS
LECTURE 9
60 MINUTES
9·1
Solution of IiDile e1eDleul equilihrilll equations iu slatic aaalysis
Static condensation
Substructuring
Multi-level substructuring
Frontal solution
9·2
SoJutiOD of filile e1emenl equilihrillD equations in slatic analysis
5 -4 , 0 U, 0
-4 6 -4 , U2 ,
, -4 6 -4 U3
=
0
(8.2)
0 , -4 5 U4 0
9·3
Solation of finite element eqailihriUl equations in static analysis
5 -4 1 o
r------------
olI l!
5 -5
16 1 (8.3)
I
I
o I_~ 29 -4
: 5 5
I
o: -4 5
5 -4 o o
14 16
o 5-5
r--------
~ _20
= (8.4)
o 0:
I 7 7
I
o 0: _ 20 65
I 7 14
I
9·4
Solation of finite element eqailillriUl equations in static analysis
STEP 3:
5 -4 1 0 U1 0
14 16
0
S -s 1 U2 1
15 20
.- 8 (8.5)
0 0 -T U3 "7
7
r---
0 0 0 I 5 U4 7
I
I "6 "6
I
12
=5
1 - (- 156) U3 - (1) U4 _ 13
U =--------:;-;;----- (8.6)
2 14 -S
5
19 36 7
o-
(-4) 35 - (1)15 - ( 0 )"5 _ 8
U1 =----~---- 5 - "5
9·5
Solution of finite element eqDilihriDlD equations in slatic analysis
STATIC CONDENSATION
Hence
and
-1)
!!a = Ba - ~c .!Sec-1 ~
---------
(
~a - ~ilC .!Sec.!Sea
aa K
Example
5
tee
I
I -4
r:~a
1 0 U1 0
I
---+------------
-4 6 -4 1 U2 1
=
1 -4 6 -4 U3 0 so that
0 1 -4 5 U4 0
'---y----'
~c 14 16 1
5 -5
~a
K =
16 29 -4
Hence (8.30) gives -a. a -5 5
~ - ,- -
6 -4 -4 [1/5] [-4 1 0] 1 -4 5
~
-
Kaa = -4 6 -4 1
and we have obtained the 3x3
1 -4 5 0
'-- - 1....- unreduced matrix in (8.3)
9·8
SoIltiOl of finite elemelt eqlilihrilDl equations in static aualysis
5 -4 0 VI
-4
1
6 -4
-4 6 -4
U2
U3
:1
0 1 -4 5 U4 :1
14
"5
-!§
5
U2
-!§
5
29
-5 -4 U3 0
-4 5 V4 0
9-7
Solutiou of finite element eqDilihriom equations in static analysis
SUBSTRUCTUR ING
......--.- L
-~-o--o
-?-?
l
e--c>---n -6
50x50 32x32
Example
17 -20
~~
6L
-20 48
[
3 -28
9·8
SoIali. oIliDile e1emeal eqailihrilDl eqaaliODS ia stalic aaalysis
EA, [ '7
3
6"L
-20
Static condensation of U2 gives
EA,
6L
Ir 7
3
3] -[-20] [lJ[-20
25 -28 48
or
ll. EA, [ 1
9 L -1
and
9·9
Solution of fiDile elemeul equilibrilll equati. in slatic aDalysis
Multi-level Substructuring
-
A 2A 4A, SA, I 16A,
, , '~
-\&-o=2:E~f' · 'n-~
\
U Ur;, U6 U7 Us Ug
u. I U2 U3 Rs
Bar with linearly varying area
-I
U,
- u2
I 1-
.-u
u3
---I
U,
• 3
---I
U,
I
- I
u3
I 1-
Us
_I
U,
I
• I 1-
Us
_I
U,
I I I
- I
Us.Rr;,
I I I 1-
ug
-. U,
I I I I I I I 1-
ug
(c) Third-level substructure and
actual structure.
'-10
Solution of fiDile e1eDleul equilihrio equti. ill static analysis
Frontal Solution
-----
---
m m+3
Element 1 Element 4
9·11
Solution of finite element equilibrium equations in static analysis
L D LT FACTORIZATION
- Basic Step
L- 1
--1 -K = -1
K
Example:
5 -4 a 5 -4 a
4 ~4 16
5 -4 6 -4 a 5 5
=
1 a -4 6 -4 a _16 29
-4
-5 5 5
a a a a -4 5 a -4 5
We note
4 4
5 -5
L-1- =
1 ~1
-1 a 1 a
5 S-
o a a a a a
9·12
Solution of finite element equilibrium equations in static analysis
x x x x x
x x x x
S :=
x ....... x upper
triangular
x x matrix
x
x
Hence
or
where
9·13
Solution of finite eleJDent equilihriDII equations in static analysis
where
t = L D~
SOLUTION OF EQUATIONS
Using
T
K= L 0 L (8.16)
we have
L V= R (8.17)
o LT U = V (8.18)
where
V := L-I (8.19)
- -n-l
and
(8.20)
9·14
Solution of finite element equilibrimn equations in static analysis
5 -4 1
6 -4 1
6 -4
5
4
~ 5 4
-5
5
5
14 14
- -4 5 -4
5
6 -4 6 -4
5
5
5 4
~ 1 5 4 1
-5 5 -5 5
14 8 14 8
1 5 -7
5 7
15 15
-4 T -4
T
5 5
9·15
Solation of finite element eqailihriam eqaati. in static analysis
X=NONZERO ELEMENT
0= ZERO ELEMENT
_~ COLUMN HEIGHTS
o 0 000
o 0
'-----,
000
X 000 X
o 0 000
o 0 x 0 0
o X 000
X X X X 0
X X X X
SYMMETRIC X XX
X
X
SKYLINE
o 0 000
o
L...-_
0 000
X 0 0 0 X
X 0 0 0 X
X 0 X 0 X
X X X 0 X
X X X X X
X X X X
X X X
X X
X
9-16
Solution of finite element equilibrium equations in static analysis
x = NONZERO ELEMENT
0= ZERO ELEMENT
COLUMN HEIGHTS
I I I
-x 0 0 0 10 0:0
o 0 0 0:0 010
xix x 010 0 x
XlX 0 010 0 0
xIx 0 0 x 0 0
x 0 X\O 0 0
x xix XIO
xix xix
SYMMETRIC Ix XlX
xIx
Ix
9·17
SOLUTION OF
FINITE ELEMENT
EQUILIBRIUM
EQUATIONS
IN DYNAMIC ANALYSIS
LECTURE 10
56 MINUTES
10·1
Solotion of finite e1mnent eqoiIihrio equations in dynaDlic analysis
Implementation of methods
Integration errors
10·2
Solution of finite element equilihriDl equations in dyDalDic ualysis
DIRECT INTEGRATION
SOLUTION OF EQUILIBRIUM
EQUATIONS IN DYNAMIC
ANALYSIS
MU+CU+KU=R
-- -- -- -
• explicit, implicit • selection of solution
integration time step (b. t)
MU + C U + K U = R (9.1)
or
10·3
Solution of finite elelleul equilihrilll equatiolS in dynaJDic analysis
Load description
time
time
--
Fig. 1. Evaluation of externally
applied nodal point load vector
tR at time t.
10·4
Solation of finite eleDlent eqailibrimn equations in dynanaic analysis
(9.6)
where we note
! t!! =(~!(mT!!
Computational considerations
(9.7)
10·5
Solution of finite element equilibrium equations in dynamic analysis
Tn
l'It er = TI ; Tn = smallest natural
period in the system
• number of operations
ex no. of elements and no. of
time steps
10·6
Solution of finite elelDent eqoiIibriDII eqoatiou in dynandc analysis
(9.28)
{9.29J
an implicit integration scheme solution
is obtained using
a. = la , 0 =~
which is the
constant-average-acceleration
method
(Newmark's method)
• number of operations
== ~n m2 + 2 n mt
10·7
Solution of finite element equilibriDII equations in dynmic analysis
Accuracy considerations
~1U+KU=R
and
where
2
K ¢. :: w·1 ~1 <p.
--1 --1
Using
¢"1 K ¢ = 0. 2
where
.. 2 T
x.1 + w.1 x.1 = ~.
~1-
R i=l, ... ,n
10·8
Solution 01 finite eleDlent equilibriDll equations in dynaDlic analysis
r~O+KU=R
.. 2
x·1 + w.1 x·1 i=l, ... ,n
with
n
U= ~<I>.x.
- ~-l 1
i =1
.. 2
x+w x=r
.. 2
x+ w x = a
o·x= a 0·· 2
x= -w
10·9
Solotion of finite element eqoiIihriDl equations in dynandc analysis
19.0 19.0
Houbolt 15.0
15.0 method
-
w
le
....
E!:.
C
5..
.g 7.0
le
0 7.0
'"CC/I ~
0 >
"iii '"
u
"8
.;:
5.0 '"
"0 5.0
'"Co '"
"0
~
.~
'"
C/I
:!
Q.
c 3.0 E 3.0
'"
l:! '"8.
tf :!
c
'.01~4t€
'"
l:!
'"
Q" 1.~
1.0 1.0
~ PE
0.06 0.10 0.14 0.18 0.06 0.10 0.14 0.18
Fig. 9.8 (a) Percentage period elonga- Fig. 9.8 (b) Percentage period elonga-
tions and amplitude decays. tions and amplitude decays.
4t-----r----:--r--r----r-----,...-----,-----,
equation
..x + 2~wx. + w2x = S 1. n p t
31----+--f-+-+----t-----t-----.,t----'-1
2
...
o
'0 static
~
"t:J
response
CtI
o 1
CJ
'ECtI
c::
>
o
1 2 3
10·10
SoIIIi. of filile 81• •1 eqailihrillD eqaaliOlS in dJllillic analysis
g
7T D:.r • 1.05
- nYNAMIC RESPONSE
_ .. - STATIC RESPONSE
gi
+ ~ =0.05
r.it
z~~.:.::::7'--
!2C 1\
~
'"- .j.,
'
fs!
,;1
• 1
!
T
81
74- ._--+-- -t-" - -- .... __..--t- ---+-._--+_.. - ........-... _-.-1
'c.oe o.."~ fl. 'JO n. 7~ I. 00
I I ,.,~.
DLF .. 0.50
- DYNAMIC RESPONSE
--- STATIc.: RESPONSE
.f... = 3.0
w
.... ,
--~
---- -- - ./
....
__
-----=~---':....;-,,---=
==_7'~--- /"7--_____ . . /'
_.~~.:.--==---/-
+
g,
::i- +-~--+---+--- " .. -------t-----+---+I- - t - _ _---+--+1~--+--+I::-:---+----,+1:::---+----,+1::-:---+------::+-'::-:---+-----:<'
c.':;: C.25 ~.I)C :."L :.00 : . .?~ I.SO 1.75 2.00 2.25 2.50 2.75 3.00
TIllE
10·11
Solution of finite element equilibrium equations in dynamic analysis
(9.100)
(9.101 )
10·12
SoIaliOi .. filile 81• •1eqailihriDl eqaali_ in dJUlDic ualysis
a, = '/2I1t
... 0 O· 0··
2. Inltlahze !!., !!., !!. ;
M = a~ + a,f.
10·13
Solution of finite element equilibrium equations in dynamic analysis
10·14
SoI.ti. of filile elOl.1 equilihriDl equations in dynamic analysis
Newmark Method:
10·15
MODE SUPERPOSITION
ANALYSIS; TIME
BISTORY
LECTURE 11
48 MINUTES
11·1
Mode slperpClilion analysis; lillie bistory
Practical considerations
11·2
Mode superposition analysis; time history
!L = 1:. !(t)
nxl nxn nxl
P = transformation matrix
Using
on
MU + c 0 + KU= R (9.1)
we obtain
C fT ~ f ;
R = PT R (9.32)
11·3
Mode sDperJMlilion ualysis; tiDle history
M0+ K U =0 (9.34)
Using
(9.36)
i =j
M'" "
J 1== 0'
-<P 1" _.:t:..
T i ., j
(9.37)
2
< W (9.38)
- n
11·4
Mode superposition analysis; time history
Defining
(9.39)
we can write
(9.40)
and have
¢T M¢ = I (9.41)
Now using
(9.44)
11·5
Mode SUperpClitiOD aualysis; tilDe bislory
Analysis with Damping Neglected
(9.45)
(9.46)
with
T a
X'I1 t=O = lj). -M
-1 -U
(9.47)
• .T O'
X'I1 cp.M
=-'-- -1 - -U
t=O
+ a..1
sin1 + 8. cos w·t
w.t 1 1
where a.i and 8i are determined
from the initial conditions in (9.47).
And then
(9.49)
11-&
Mode sDperp.ition analysis; time history
4f----..-----:--..--r----,..----~---_r_---...,
equation
••
x + 2E;,wx. + W2X = S 1. n P t
31-_ _-+_ _+--+-+-_ _+-_ _-+ -+-_ _--.,
0 2
....
0
.....
u static
.....
CtI
response
-0
CtI
0
u
E
CtI
r:::::
>- ~= \-0
0
2 3
Hence we use
uP
--
= ~ ¢. x·1 (t)
~--l
i =1
where
uP - U
(9.50)
11·7
Mode superposition analysis; time history
Static correction
r.1 = ¢.
T R
-1-
Then
and
K flU fiR
Recall, we have
(9.43)
If the damping is proportional
T
¢. C (po = 2w. E;,. cS. . (9.51)
-1 ---J 1 1 1J
and we have
i=l, ... ,n
(9.52)
11·8
Mode superposition analysis; time history
(9.56)
C
-
= a
--
~1 +-
B-K (9.55)
example:
Assume ~, = 0.02
w, = 2
calculate a and B
We use
2w.1 ~.
1
or
'/
a + Bw:- 2w.1 ~.
- - 1 1
11·9
Mode superposition analysis; time history
a + 4ii = 0.08
a + 913 = 0.60
C = -0.336 M + 0.104 K
2
a + 13 w. = 2w. [, .
1 1 1
2
a + SW.
1
E,. =
1 2w.
1
= -
a + -13 w
2w. 2 i
1
11·10
Mode sDperp.ition analysis; time history
Response solution
with
r·1
TO
xi I t = 0 "--. !i!i .!:L
• ITO'
xi t = 0 = !i f1 .!:L
and then
P
uP ~¢. x. (t)
LJ-1 1
i =1
Practical considerations
- it may be important to
calculate E p (t) or the
static correction.
11·11
SOLUTION METHODS
FOR CALCULATIONS
OF FREQUENCIES
AND MODE SBAPES
LECTURE 12
58 MINUTES
12·1
Solution methods for calculations of frequencies and mode shapes
Large eigenproblems
12·2
Solatiu methods for calculations of frequencies and mode shapes
Standard EVP:
r! = \ !
nxn
Generalized EVP:
2
!sP.-=\!i! - (\=w )
Quadratic EVP:
12·3
Solation lIethods lor calcalatiou oIlreqa.cies and lIode shapes
p(A)
p(A) = det(K - A ~)
In buckling analysis
.!$.!=A~!
where
p(A) = det (~ - A ~)
p(A)
12·4
Solution methods for calculations of frequencies and mode shapes
.. --
.. ~ -
(~ - ~ ~)! = n .K 2£.
.Ki=;\!ii
M=I::I::T i=hTjJ
hence
~ :t = ;\ i ; K = 1::- 1 K [- T
or
M= W02 W
T etc ...
12·5
SolotiOl .elhods lor calcolations oIlreqoeacies ud .ode sllapes
1.3 . .. 1n
Equation: ~P_=A~~
e.g. Inverse It.
! ~+l = M ~
~+l
• Forward Iteration
• Rayleigh Quotient Iteration
can be employed to cal
culate one eigenvalue
and vector, deflate then
to calculate additional
eigenpair
12·6
Solution methods for calculations of frequencies and mode shapes
! ~ = A ~ ~ ~ (K - A M) ¢ 0
Hence
,
,,
Newton Iteration
2 n
p(A) aO + alA + a2A + ... + anA
= det L D LT = II d ..
Explicit polynomial iteration: -- - . II
I
12-7
Solution methods for calculations of frequencies and mode shapes
]J.1- 1
p (A) / (A-A,)
I
I
III
Convergence guaranteed to A1 ' then
A2 , etc. but can be slow when we
calculate multiple roots.
12·8
SaI.liOi .6Jds for calculali. of freql8cies iIIld ole shapes
3rd associated
constraint problem
2nd associated
constraint problem
1st associated
constraint problem
. ·
1
:::}
2 3 4
. . ..
! <p = A!11 9·
t
~
·
.. ·
·
; ;
... .
.
.. -..... .
12·9
Solution lDethods lor calculations 01 frequencies ud lDode shapes
3) STURM SEQUENCE METHODS
T
~ - ].lS. ~ = h Qh
Calculate
,
Count number of negative elements
in Q and use a strategy to isolate
eigenvalue(s) .
interval
,
].ls
1 ].lS2
Tf ..
---
• Need to take care in L D L aetonzatlon
4) TRANSFORMATION METHODS
<PTK<P=A
~!=A~!--T--
j<P
- -M<P =
I
-
<P = [~,... ~ J; _H
n = [Al ... 'n]
A
12·10
5oI1tiOi .elhods for calculations 01 frequencies ad .ode shapes
T T T
~ ... ~ ~l f f1 ~ ... ~-~
T T T
~ ... ~ f 1 !i f 1 ~ ... ~-l
e.g. generalized Jacobi
method
• Here we calculate all eigenpairs
simultaneously
• Expensive and ineffective
(impossible) or large problems.
• Determinant search
to get near a root
12·11
Solution methods for calCl1atiou of frequencies and mode sJlapes
p(A)
/
~-
. ....
A
L D LT = n d ..
= det - -- ; 11
n is normally = 1.0
n = 2. , 4. , 8. ,... when convergence
is slow
Same procedure can be employed to
obtain shift near A; , provided
P( A) is deflated of A1' . . . ,A; _1
12·12
Solution lOethods for calculations of freqoBcies ud lOode shapes
lJi+ 1
k =1,2, ...
•~+l = - ~+l
T - ~
(~+l !i ~+l)
- T
~+l !i ~k
p (~+l) =- T ~
~+l !i ~+l
12·13
Solution methods lor calculations oIlrequencies ud mode shapes
Advantage:
Calculates only eigenpairs actually
required; no prior transformation
of eigenproblem
Disadvantage:
Many triangular factorizations
-T
~+1 = 4+1 K
-~+1
-T
~+1 = 4+1 ~1
4+1
12·14
Solution methods for calculations of frequencies and mode shapes
CONDITION:
starting subspace spanned
by X, must not be orth
ogonal to least dominant
subspace required.
eigenvalue
p eigenvalues flS
T
!5. - flS t1 = ~ Q~
when
< tal
12·15
Solution methods for calculations of frequencies and mode shapes
Starting Vectors
Two choices
1) ~l x. = e.,
~ ~
x = random vector
4
2) Lanczos method
Here we need to use q much
larger than p.
Checks on eigenpairs
12·16
MIT OpenCourseWare
http://ocw.mit.edu
The following may not correspond to a particular course on MIT OpenCourseWare, but has been
provided by the author as an individual learning resource.
For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.