Vous êtes sur la page 1sur 35

Introduction to Finite Element Methods

MCE 565 Wave Motion & Vibration in Continuous Media Spring 2005 Professor M. H. Sadd

Need for Computational Methods


Solutions Using Either Strength of Materials or Theory of
Elasticity Are Normally Accomplished for Regions and Loadings With Relatively Simple Geometry Many Applicaitons Involve Cases with Complex Shape, Boundary Conditions and Material Behavior Therefore a Gap Exists Between What Is Needed in Applications and What Can Be Solved by Analytical Closedform Methods This Has Lead to the Development of Several Numerical/Computational Schemes Including: Finite Difference, Finite Element and Boundary Element Methods

Introduction to Finite Element Analysis


The finite element method is a computational scheme to solve field problems in engineering and science. The technique has very wide application, and has been used on problems involving stress analysis, fluid mechanics, heat transfer, diffusion, vibrations, electrical and magnetic fields, etc. The fundamental concept involves dividing the body under study into a finite number of pieces (subdomains) called elements (see Figure). Particular assumptions are then made on the variation of the unknown dependent variable(s) across each element using so-called interpolation or approximation functions. This approximated variation is quantified in terms of solution values at special element locations called nodes. Through this discretization process, the method sets up an algebraic system of equations for unknown nodal values which approximate the continuous solution. Because element size, shape and approximating scheme can be varied to suit the problem, the method can accurately simulate solutions to problems of complex geometry and loading and thus this technique has become a very useful and practical tool.

Advantages of Finite Element Analysis


-

Models Bodies of Complex Shape

- Can Handle General Loading/Boundary Conditions - Models Bodies Composed of Composite and Multiphase Materials - Model is Easily Refined for Improved Accuracy by Varying Element Size and Type (Approximation Scheme) - Time Dependent and Dynamic Effects Can Be Included - Can Handle a Variety Nonlinear Effects Including Material Behavior, Large Deformations, Boundary Conditions, Etc.

Basic Concept of the Finite Element Method


Any continuous solution field such as stress, displacement, temperature, pressure, etc. can be approximated by a discrete model composed of a set of piecewise continuous functions defined over a finite number of subdomains. One-Dimensional Temperature Distribution
T T

Exact Analytical Solution

Approximate Piecewise Linear Solution

Two-Dimensional Discretization

u(x,y)
-1

-2

-3 4 3 .5 3 2 .5 2 1 .5 y 1 -1 -0 .5 0 0 .5 1 1 .5 2 2 .5 3

Approximate Piecewise Linear Representation

Discretization Concepts
T E xact T emperature D is tribution, T(x)

Finite Elem ent Discretization


T1

Linear Interpolation Model (Four E lem ents)


T2 T2 T3 T3 T4 T4 T5

Q uadratic Interpolation M odel (Tw o E lem ents) T


1

T2

T3

T3

T4

T5

T
T1 T2 T3 T4 T5

T
T1 T2 T3 T4 T5

x Piecewise Linear Appr oxim ation


Tem perature Continuous but w ith D iscontinuous Temperature G radients

x Piecewise Q uadratic A pproximation


Tem perature and Temperature Gradients Continuous

Common Types of Elements


One-Dimensional Elements Line Rods, Beams, Trusses, Frames Two-Dimensional Elements Triangular, Quadrilateral Plates, Shells, 2-D Continua

Three-Dimensional Elements Tetrahedral, Rectangular Prism (Brick) 3-D Continua

Discretization Examples

One-Dimensional Frame Elements

Two-Dimensional Triangular Elements

Three-Dimensional Brick Elements

Basic Steps in the Finite Element Method Time Independent Problems


Domain Discretization Select Element Type (Shape and Approximation) Derive Element Equations (Variational and Energy Methods) Assemble Element Equations to Form Global System

[K]{U} = {F}
[K] = Stiffness or Property Matrix {U} = Nodal Displacement Vector {F} = Nodal Force Vector - Incorporate Boundary and Initial Conditions - Solve Assembled System of Equations for Unknown Nodal Displacements and Secondary Unknowns of Stress and Strain Values

Common Sources of Error in FEA


Domain Approximation Element Interpolation/Approximation Numerical Integration Errors (Including Spatial and Time Integration) Computer Errors (Round-Off, Etc., )

Measures of Accuracy in FEA


Accuracy Error = |(Exact Solution)-(FEM Solution)| Convergence Limit of Error as: Number of Elements (h-convergence) or Approximation Order (p-convergence) Increases Ideally, Error p 0 as Number of Elements or Approximation Order p g

Two-Dimensional Discretization Refinement

(Node) y

(Discretization with 228 Elements)

y y (Triangular Element)

(Discretization with 912 Elements)

One Dimensional Examples Static Case


Bar Element Uniaxial Deformation of Bars Using Strength of Materials Theory u1 1 Differential quation : d  (au )  cu  q ! 0 dx Boundary Condtions pecification : du u,a dx 2 u2
U1

Beam Element Deflection of Elastic Beams Using Euler-Bernouli Theory w1 1 w2


U2

Differenti al quation : d 2w d2  2 (b 2 ) ! f ( x ) dx dx Boundary Condtions pecificat ion : dw d 2w d d 2w ,b 2 , (b ) w, dx dx dx dx 2

Two Dimensional Examples


Triangular Element Scalar-Valued, Two-Dimensional Field Problems
J

Triangular Element Vector/Tensor-Valued, TwoDimensional Field Problems v3 3 u3 v2 u2

3 2 1
J
Example i erential Equation :
x 2J x 2J  ! f ( x, y ) xx 2 xy 2

J

v1 1
Q2 u  Q2 v 

u1
isplacements

Elasticity Field Equationsin Terms o

x xu xv E   Fx ! 0 2(1  R ) xx xx xy x xu xv E   Fy ! 0 2(1  R ) xy xx xy oundary onditons

oundary ondtions Speci ication :


J,

dJ xJ xJ nx  n y ! dn xx xy

xu xv xu xv  C12 n x  C66  n y Tx ! C11 xy xx xx xy xu xv xu xv  C22 n y Ty ! C66  nx  C12 xy xx xx xy

Development of Finite Element Equation


The Finite Element Equation Must Incorporate the Appropriate Physics
of the Problem For Problems in Structural Solid Mechanics, the Appropriate Physics Comes from Either Strength of Materials or Theory of Elasticity FEM Equations are Commonly Developed Using Direct, VariationalVirtual Work or Weighted Residual Methods

Direct Method
Based on physical reasoning and limited to simple cases, this method is worth studying because it enhances physical understanding of the process

Variational-Virtual Work Method


Based on the concept of virtual displacements, leads to relations between internal and external virtual work and to minimization of system potential energy for equilibrium

Weighted Residual Method


Starting with the governing differential equation, special mathematical operations develop the weak form that can be incorporated into a FEM equation. This method is particularly suited for problems that have no variational statement. For stress analysis problems, a Ritz-Galerkin WRM will yield a result identical to that found by variational methods.

Simple Element Equation Example Direct Stiffness Derivation


u1 F1 1 k 2 u2 F2

Equilibrium at Node 1 F1 ! ku1  ku 2 Equilibrium at Node 2 F2 ! ku1  ku 2 or in Matrix Form k  k


Stiffness Matrix

u F  k 1 1 u ! F k 2 2 [ K ]{u} ! {F }
Nodal Force Vector

Common Approximation Schemes One-Dimensional Examples


Polynomial Approximation Most often polynomials are used to construct approximation functions for each element. Depending on the order of approximation, different numbers of element parameters are needed to construct the appropriate function.

Linear

Quadratic Special Approximation

Cubic

For some cases (e.g. infinite elements, crack or other singular elements) the approximation function is chosen to have special properties as determined from theoretical considerations

One-Dimensional Bar Element


Approximation : u ! ] k ( x )uk ! [ N ]{d }
k

du d d[N ] {d } ! [ B]{d } ! ] k ( x )u k ! dx dx k dx Stress - Strain Law : W ! Ee ! E [ B]{d } Strain : e !

WHedV ! P u
; L 0

i i

 Pj u j  fHudV
;

{ d }T A[ B]T E [ B]dx{d } ! { d }T
0

Pi L  { d }T A[ N ]T fdx 0 Pj L L T A[ B] E[ B]dx{d } ! {P}  A[ N ]T fdx


0

[ K ] ! A[ B]T E[ B]dx ! Sti ness Matrix


0

[ K ]{d } ! {F }

P i L {F } !  [ N ]T fdx ! Loading Vector 0 P j u i {d } ! ! Nodal isplacement Vector u j

One-Dimensional Bar Element


Axial Deformation of an Elastic Bar x f(x) = Distributed Loading Typical Bar Element A = Cross-sectional Area E = Elastic Modulus

du Pi !  AE i dx

ui (i)

uj
(j)

Pj !  AE

du j dx

(Two Degrees of Freedom)

Virtual Strain Energy = Virtual Work Done by Surface and Body Forces

Wij Heij dV ! Ti n Hui dS  Fi Hui dV


St V

For One-Dimensional Case

WHedV ! P u
;

i i

 Pj u j  fHudV
;

Linear Approximation Scheme


ui (1) pproximate lastic L isplacement (2) uj x (local coordinate system) u(x) u1 ! a1 u2 ! a1  a 2 L (1)
]1(x)

u ! a1  a 2 x

u  u1 x x u ! u1  2 x ! 1  u1  u 2 L L L ! ] 1 ( x )u1  ] 2 ( x )u2 u1 u A ! 1  x x 1 ! [ N ]{d } u!? 1 2 u2 L L u 2 [ N ] ! pproximation unction atrix {d } ! odal Displacement Vector

(2)
]2(x)

1 x

(1)

(2)

]k(x) Lagrange Interpolation Functions

Element Equation
Linear Approximation Scheme, Constant Properties
1  1 1 L L AE [ K ] ! A[ B]T E[ B]dx ! AE [ B]T [ B] dx ! AE L  L! 0 0 1 L L L L x  L P1 L P1 T L dx ! P1  Af o L 1 {F } !  A[ N ] fdx !  Af o 0 x P2 P2 0 P2 2 1 L u1 {d } ! ! odal Displacement Vector u2
1  1  1 1

u P 1 AE  1 1 1 1 Af o L [ ]{d } ! {F } 1  1 u ! P  2 1 L 2 2

Quadratic Approximation Scheme


u1 (1) (2) L
Approximate Elastic Displacement u1 ! a1 u ! a1  a 2 x  a 3 x
2

u2 (3)

u3

u(x)

L L2 u2 ! a1  a 2  a 3 2 4 2 u3 ! a1  a 2 L  a3 L

(1)

(2)
]2(x)

(3)

u ! ]1 ( x )u1  ] 2 ( x )u2  ] 3 ( x )u3 u ! ?]1 ] 2 u 1 ] 3 Au2 ! [ N ]{d } 3 u

]1(x)

]3(x)

1 x

Element Equation u F 7  8 1 1 1 AE  8 16  8 u2 ! F2 3L u F 1  8 7 3 3 (1) (2) (3)

Lagrange Interpolation Functions


Using Natural or Normalized Coordinates
1 , i! j ] i (\ j ) ! 0 ,i{ j

\
(1)

1 e \ e 1

(2)

1 (1  \) 2 1 ] 2 ! (1  \) 2
]1 !

\
(1) (2) (3)

1 ]1 !  \(1  \) 2 ] 2 ! (1  \)(1  \) ]3 !
]1 ! 

1 \(1  \) 2

\
(1) (2) (3) (4)

1 1 9 (1  \)(  \)(  \) 3 3 16 27 1 (1  \)(1  \)(  \) ]2 ! 16 3 1 27 (1  \)(1  \)(  \) ]3 ! 3 16 9 1 1 ] 4 !  (  \)(  \)(1  \) 16 3 3

Simple Example
P A1,E1,L1 1 (1)
Global Equation Element 1 U P (1) 1  1 0 1 1 A1 E1  1 1 0 U 2 ! P2(1) L1 0 0 3 0 0 U

A2,E2,L2 2 (2) (3)


Global Equation Element 2 0 1 0 U 0 0 A2 E 2 0 1  1 U 2 ! P1( 2 ) L2 0  1 1 3 2( 2 ) U P Take Zero Distributed oading f !0

ssembled lobal ystem quation

A1 E1 L 1  A1 E1 L1 0

A1 E1 L1 A1 E1 A2 E 2  L1 L2 AE  2 2 L2 

P1(1) P1 U1 AE  2 2 U 2 ! P2(1)  P1( 2 ) ! P2 L2 ( 2) U P 3 P2 3 A2 E 2 L2 0

Simple Example Continued


P A1,E1,L1 1 (1) (2)
A1 E1 L 1  A1 E1 L1 0 A1 E1 L1 A1 E1 A2 E 2  L1 L2 AE  2 2 L2 

A2,E2,L2 2 (3)
Reduced lobal ystem quation P1(1) 0 AE  2 2 U 2 ! 0 L2 U A2 E 2 3 P L2 0 AE L 0 2  1 U 2 !  1 1 U P 3

oundary onditions U1 ! 0 P2( 2 ) ! P P2(1)  P1( 2 ) ! 0

or Uniform Properties A, E , L

olving U 2 !

2 PL PL , U3 ! , P1(1) !  P AE AE

A1 E1 A2 E 2 L  L 2 1  A2 E 2 L2

A2 E 2 U 0 L2 2 ! A2 E 2 3 U P L2

One-Dimensional Beam Element


Deflection of an Elastic Beam f(x) = Distributed Loading x I = Section Moment of Inertia E = Elastic Modulus Typical Beam Element w2 w1 U1 U2 ;
(1)
d 2w d d 2w , Q2 ! EI EI Q1 ! 2 2 dx dx 1 dx 1 d 2w d d 2w , Q4 !  EI Q3 !  EI dx 2 2 dx 2 2 dx dw dw u1 ! w1 , u2 ! U1 !  , u3 ! w2 , u4 ! U 2 !  dx 1 dx 2

(2)

V1

M1

M2 V2

(Four Degrees of Freedom)

Virtual Strain Energy = Virtual Work Done by Surface and Body Forces

WHedV ! Q u
; 0

1 1

 Q2 u2  Q3u3  Q4 w4  fHwdV
; L 0

EI [ B] [ B]dx{d } ! Q1u1  Q2 u 2  Q3u3  Q4 w4  f [ N ]T dV

Beam Approximation Functions


To approximate deflection and slope at each node requires approximation of the form w( x ) ! c1  c2 x  c3 x 2  c4 x 3 Evaluating deflection and slope at each node allows the determination of ci thus leading to w( x ) ! J1 ( x )u1  J2 ( x )u2  J3 ( x )u3  J4 ( x )u 4 , where Ji are the Hermite ubic Approximation Functions

Beam Element Equation


EI [ B]T [ B]dx{d } ! Q1u1  Q2 u2  Q3u3  Q4 w4  f [ N ]T dV
0 0 L L

u1 u 2 {d } ! u 3 4 u

[ B] !

d[N ] dJ dJ dJ3 dJ4 ![ 1 2 ] dx dx dx dx dx


J1 6 L 2 J  fL L T f [ N ] dx ! f dx ! 0 J3 12 6 4 L J

 3L  6  3L 6 2 L 3L L2 2 EI  3L 2 L [ K ] ! EI [ B]T [ B]dx ! 3 0 3L 6 3L L 6 3L 2 L2  3L L2

6 Q1  3L  6  3L u1 6 2 u Q 3L L2 2 2 fL L  2 EI  3L 2 L !  3L 6 3L 3 3 12 6 u Q L3  6 L u Q 3L 2 L2 4 4  3L L2

FEA Beam Problem


f a
1 (1)
6 / a3 2  3 / a  6 / a 3 2 EI 2  3 / a 0 0  3 / a2  6 / a3

Uni orm EI b
2

(2)
Element 1
 3 / a2

(3)
U Q 0 0 1 6 1(1) a (1) U  Q 0 0 2 2 U Q 0 0 3 fa 6 3(1) !   (1) U Q 0 0 4 12 a 4 0 0 U 0 0 5 0 0 U 0 0 6

2/ a 3/ a 0 0
2

3 / a2 6/ a 0 0
3

1/ a 3 / a2 2/a 0 0

1/ a

3 / a2

leme t 2
0 0 0 2 EI 0 0 0

0 0 0

0 0 6 / b3

0 0  3 / b2 2/b 3 / b2 1/ b

0 0  6 / b3 3 / b2 6 / b3 3 / b2

0  3 / b2 0  6 / b3 0  3 / b2

0 U1 0 U 0 2 2 U Q  3 / b 3 1( 2 ) ! (2) U Q 1/ b 4 2 U Q 3 / b 2 5 3( 2 ) (2) U Q 2 / b 6 4 0

FEA Beam Problem


1 (1) (2)
Global Assembled System
6 / a 3 2 EI  3 / a2  6 / a3  3 / a2

2 (3)

0 0
 6 / a3

2/a

3 / a2 6 / a3  6 / b3

1/ a 3 / a 2  3 / b2 2/ a  2/b

3 / a2 6 / a3

U 1 6 Q1(1) a ( U  Q21) 0 2 U Q  3 / a2 3 fa 6 3(1)  Q1( 2 ) !   (1) ( 2) U Q 1 / a 4 12 a 4  Q2 2 0 Q3( 2 ) U5 3/ a ( 0 Q4 2 ) U 2 / a 6 0 Matching Conditions

o U1 ! w
(1) 1

ary Co
(1) 1

itio s
!0,Q
( 2) 3

! 0 , U2 ! U

!Q

(2) 4

!0

( ( Q3(1)  Q1( 2 ) ! 0 , Q41)  Q22 ) ! 0

e
6 / a 3  6 / b 3 2 EI

ce System 3 / a2 6 / a3

2/a  2/b

Solve System or Primary Unknowns U1 ,U2 ,U3 ,U4 Nodal Forces Q1 and Q2 an Then e etermined

3 / a 2  3 / b2

 6 / a3

6 0  3 / a 3 U1 U a 0 1 / a 2 fa !   U 0 0 3 / a 2 3 12 U 0 0 2 / a 4

Special Features of Beam FEA

Analytical Solution Gives ubic e lection urve

Analytical Solution Gives Quartic e lection urve

FEA Using Hermit ubic Interpolation Will Yield esults That Match Exactly With ubic Analytical Solutions

Truss Element
Generalization of Bar Element With Arbitrary Orientation
y

k=AE/L x

s ! sin U , c ! cos U

Frame Element
Generalization of Bar and Beam Element with Arbitrary Orientation w1 w2 U2 U1 u1 ; u2 P1 P2 (1) (2) L M2 M1 V1 V2
AE L 0 0 AE  L 0 0

0 12 EI L3 6 EI L2 0


0 6 EI L2 4 EI L 0


AE L 0 0 AE L 0 0


0 12 EI L3 6 EI  2 L 0 12 EI L3 6 EI  2 L

12 EI L3 6 EI L2

6 EI L2 2 EI L

6 EI P1 u1 2 L w Q 2 EI 1 1 Q U L 1 2 ! u2 P 0 2 w Q 6 EI 2 3  2 2 4 Q U L 4 EI L

Element Equation an Then e otated to Accommodate Arbitrary Orientation

Some Standard FEA References


Bathe, K.J., Finite Element Procedures in Engineering Analysis, Prentice-Hall, 1982, 1995. Beer, G. and Watson, J.O., Introduction to Finite and Boundary Element Methods for Engineers, John Wiley, 1993 Bickford, W.B., A First Course in the Finite Element Method, Irwin, 1990. Burnett, D.S., Finite Element Analysis, Addison-Wesley, 1987. Chandrupatla, T.R. and Belegundu, A.D., Introduction to Finite Elements in Engineering, Prentice-Hall, 2002. Cook, R.D., Malkus, D.S. and Plesha, M.E., Concepts and Applications of Finite Element Analysis, 3rd Ed., John Wiley, 1989. Desai, C.S., Elementary Finite Element Method, Prentice-Hall, 1979. Fung, Y.C. and Tong, P., Classical and Computational Solid Mechanics, World Scientific, 2001. Grandin, H., Fundamentals of the Finite Element Method, Macmillan, 1986. Huebner, K.H., Thorton, E.A. and Byrom, T.G., The Finite Element Method for Engineers, 3rd Ed., John Wiley, 1994. Knight, C.E., The Finite Element Method in Mechanical Design, PWS-KENT, 1993. Logan, D.L., A First Course in the Finite Element Method, 2nd Ed., PWS Engineering, 1992. Moaveni, S., Finite Element Analysis Theory and Application with ANSYS, 2nd Ed., Pearson Education, 2003. Pepper, D.W. and Heinrich, J.C., The Finite Element Method: Basic Concepts and Applications, Hemisphere, 1992. Pao, Y.C., A First Course in Finite Element Analysis, Allyn and Bacon, 1986. Rao, S.S., Finite Element Method in Engineering, 3rd Ed., Butterworth-Heinemann, 1998. Reddy, J.N., An Introduction to the Finite Element Method, McGraw-Hill, 1993. Ross, C.T.F., Finite Element Methods in Engineering Science, Prentice-Hall, 1993. Stasa, F.L., Applied Finite Element Analysis for Engineers, Holt, Rinehart and Winston, 1985. Zienkiewicz, O.C. and Taylor, R.L., The Finite Element Method, Fourth Edition, McGraw-Hill, 1977, 1989.