SUBJECT :-FEM

TOPIC:- Galerkin’s and Raleigh-Ritz Method

By:-Nikhil gedam (roll no 10)

:-Abhijit(roll no.05)
 The Finite Element Method
• Evolved first from the matrix methods of structural analysis in the early
1960’s
• Uses the algorithms of linear algebra
• Later found to have a more fundamental foundation
• The essential features are in the formulation
• There are two alternative formulations that are broadly equivalent in most
circumstances
– Variational formulations, e.g. the Rayleigh-Ritz method
– Weak or weighted residual formulations, e.g.the Galerkin method
• Both approaches lead to integral equations instead of differential equations
(the strong form)
 Integral Formulations
Consider the strong form of a linear partial differential equation, e.g. 3-D
Poisson’s equation with zero boundary conditions:

 2u  2u  2u
    f ( x, y , z )
x 2
y 2
z 2 On region R

u 0 on boundary S

Strong Form Lu = f

Variational Principle, e.g. minimum potential energy

u  min  (Lv  2f )  v d V

v R
Weighted Residual (weak) form, e.g. virtual work

 (Lu  f )  w d V  0
R
 Weak Form for 2-D Poisson’s Equation
 2u  2u
   f ( x, y )
x 2
y 2 On region S

u0 on boundary C
Weak form
  2u  2u 
   2  2   w d xdy   f  w d xdy
S  x y  S

Integrate by parts
0 0
 u w u w  u u
S  x x y y 
 d xd y  C y w d x  C x w d y  S f  w d xdy

Where, u and w vanish at the boundary

 Galerkin’s Method for 2-D Poisson’s Equation
• Choose a finite set of approximating (trial) functions, i(x,y), i = 1, 2, …, N
• Allow approximations to u in the form
U(x,y) = U11 + U22 + U33 + … + UNN
(that can also satisfy the essential boundary conditions)
• Solve N discrete equations for U1, U2, U3, …, UN

   1  N    i   1   N   i 
  1 x
U  ...  U N 
x   x 
  U1
y
 ...  U N 
y  y 
d xdy
S
  f i dxdy
s
   K ij U j   Fi
j
 Galerkin’s Method for 2-D Poisson’s Equation
[K]U = F

[K] is the stiffness matrix and F is the load (RHS) vector

  i  j  i  j 
K ij     d xd y  K ji
S
x x y y 
Fi   f i dxdy
S
[K] is symmetric and positive definite
 Comments on Galerkin’s Method
• Galerkin is more general than Rayleigh-Ritz. If we add u/x, symmetry
& the variational principle are lost, but Galerkin still works
• If w is chosen as Dirac delta functions at N points, weighted residuals
reduces to the collocation method
• If w is chosen as the residual functions Lu-f, weighted residuals reduces
to the least squares method
• By choosing w to be the approximating functions, Galerkin’s method
requires the error (residual) in the solution to be orthogonal to the
approximating space.
• The integration by parts (Green-Gauss theorem) automatically
introduces the Neumann (natural) boundary conditions
• The Dirichlet (essential) boundary conditions must be satisifed explicitly
when solving [K]U=F
• Since discretized integrals are sums, contributions from many elements
are assembled into the global stiffness matrix by addition.
• The Ritz-Galerkin FEM finds the approximate solution that minimizes the
error in the energy
 V-5-1 Approximate Methods

(I) Method of Weighted Residuals (MWR)

L[u ] in 0 D
+homo. b.c’s in B
Assume approx. solution
n
u  un   Cii
i 1
where each trial function i satisfies the b.c’s
The residual
Rn  L[un ]
In this
method (MWR), Ci are chosen such that Rn is forced to be zero in an average
sense.

i.e. < wj, Rn > = 0, j = 1,2,…,n

where wj are the weighting functions..
8
(II) Galerkin Method
wj are chosen to be the trial functions  j hence the trial functions is chosen
as members of a complete set of functions.

Galerkin method force the residual to be zero w.r.t. an orthogonal complete set.

Ex: Torsion of a square shaft

 2  2
 0 on x  a , y  a
(i) one – term approximation

 1  c1 ( x 2  a 2 )( y 2  a 2 )
Ri   2 1  2  2c1[( x  a)2  ( y  a)2 ]  2
1  ( x 2  a 2 )( y 2  a 2 )
9
 a a
From  
a a
R11dxdy  0
5 1
 c1 
8 a2
therefore
5
1  2
( x 2
 a 2
)( y 2
 a 2
)
8a
the torsional rigidity

D1  2G   dxdy  0.1388G (2a) 4

R

the exact value of D is

Da  0.1406G (2a) 4
the error is only -1.2%

10
(ii) two – term approximation

 2  ( x 2  a 2 )( y 2  a 2 )[c1  c2 ( x 2  y 2 )]
By symmetry → even functions
 R2   2  2
1  ( x 2  a 2 )( y 2  a 2 )
2  ( x 2  a 2 )( y 2  a 2 )( x 2  y 2 )
From 
R
R21dxdy  0

and 
R
R22 dxdy  0

we obtain

1295 1 525 1
c1  , c2 
2216 a 2 4432 a 2
therefore

D2  2G   2 dxdy  0.1404G (2a ) 4 the error is only -0.14% 11

R
(II) Raleigh-Ritz Method

This is used when the exact solution is impossible or difficult to obtain.

n
First, we assume the approximate solution as : u  CU
i 1
i i

Where, Ui are some approximate function which satisfy the b.c’s.

Then, we can calculate extreme I .
I I
I  I (c1 , , cn ) choose c1 ~ cn i.e.   0
c1 cn

Ex: y  xy   x , y (0)  y (1)  0

Sol :
From
1
1  2 1 2 
 y  xy  x  ydx  0  I 
1
0  2  y   xy  xy  dx
0 2 

12
Assuming that
y  x  1  x   c1  c 2 x  c3 x 2  

(1)One-term approx
y  c1 x  1  x   c1  x  x 2  y  c1  1  2 x 
1 2 x 2 2 2 
Then,  1  0  1   1   1  
1
I c  c 1  4 x  4 x 2
 c x  2 x 3
 x 4
 c x x  x dx
2 2 
c12  4  c12  1 2 1   1 1   19 c 2  c1
 1  2         c1    1
2  3 2 4 5 6 3 4 120 12
I 19 1
0  c1   0  c1  0.263  y (1)  0.263x(1  x)
c1 60 12
(2)Two-term approx

y  x(1  x)(c1  c2 x)  c1 ( x  x 2 )  c2 ( x 2  x3 )

13
Then y  c1 (1  2 x)  c2 (2 x  3x 2 )

 1 2
I (c1 , c2 )   [ c1  1  4 x  4 x 2   2c1c2  2 x  7 x 2  6 x 3 
1

0 2
1 2 3
2 2 3 4
  2
 
c3 (4x 12x  9x )  c1 x  2x4  x5  2c1c2 x4  2x5  x6 
 
 c2 2 ( x 5  2 x 6  x 7 )  c1  x 2  x3   c2  x3  x 4  ]dx
c12  4 1 2 1  7 3 1 1 1
 1  2       c1c2 1      
2  3 4 5 6  3 2 5 3 7
c12  4 9 1 2 9 c c
  3     1  2
2 3 5 6 7 8  12 20
19 2 11 107 2 c1 c2
 c1  c1c2  c2  
120 70 1680 12 20

14
 I 19 11 1
0  c1  c2 
c1 60 70 12
I 11 109 1
0  c1  c2 
c2 70 840 20
0.317 c1 + 0.127 c2 = 0.05 c = 0.177 , c
1 2 = 0.173

 y (2)  (0.177 x  0.173x 2 )(1  x)

( It is noted that the deviation between the successive approxs. y(1) and y(2) is
found to be smaller in magnitude than 0.009 over (0,1) )

15