M.

Vable

MEEM 5170: Rayleigh-Ritz

Fall 2015

Rayleigh-Ritz Method

Linear functional: l 1 u 2 v 1 l u 2 l v

Bilinear functional B 1 u 1 2 u 2 v 1 B u 1 v 2 B u 2 v

Symmetric bilinear functional: B u v = B v u ; u and v can be vectors.

Rayleigh-Ritz method is a formal process of minimizing the functional (potential energy) given below using a series of kinematically admissible displacement functions to produce a set of algebraic equations in the unknown constants of the series approximation.

B u 1 v 1 2 v 2 1 B u v 1 2 B u v 2

1
--- B u u l u
2

Rayleigh-Ritz method is applicable to conservative systems that may be linear or non-linear systems.

Approximation
n

u x f0 +
Cf
j 1 j j

where, Cj are constants to be determined. Cj are the generalized coordinates as the variation of them represents the variation of the displacement curve.
1. f 0 is chosen such that the non-homogeneous part of all essential boundary conditions are satisfied.
2. f j are chosen such that the homogeneous part of all essential boundary conditions are satisfied. In other words, f j are a set of kinematically admissible functions.
3. f j must be sufficiently differentiable. In other words, the highest derivative in the functional must be defined.
4. f j must be independent. Otherwise, there exist values of Cj such that

j 1 Cj fj

= 0 and our matrix in the set of algebraic equations

will be singular.
5. f j must be complete. Otherwise, we may get large errors in our solution from our missing terms.
6. If all the above are satisfied then we have correctly chosen are functions, but the choice may not be a good choice because the series
converges very slowly.

M. Vable

MEEM 5170: Rayleigh-Ritz

Fall 2015

Algebraic equation
1
= --- B u 1 u 2 l u 1
2

u1 x f0 +
Cf
j 1 j j

u2 x f0 +
C f
k 1 k k

and
n

l u1 l f0 +
C f l f0 +
C lf
j 1 j j
j 1 j j
n

B u 1 u 2 B f 0 +
C f u B f 0 u 2 +
C B fj u 2 B f 0 f 0 +
C f +
C B f j f 0 +
C f or
j 1 j
j1 j
j 1 j j 2
k 1 k k
k 1 k k
n

B u 1 u 2 B f 0 f 0 +
C B f 0 f k +
C B f j f 0 +
C
C B f j f k
k 1 k
j1 j
j 1 j k 1 k
n

B u 1 u 2 B f 0 f 0 +
C B f 0 f j + B f j f 0 +
C C B f j f k
j 1 j
j 1 k 1 j k
1
1 n
1
n
n
C C Bf f
= --- B f 0 f 0 l f 0 + ---
C l fj --- B f 0 f j + B f j f 0
2
2 j 1 k 1 j k j k j 1 j
2
1
1 n
n
n
Cj C k B f j f k
= 0 + ---
Cj l fj --- B f 0 f j + B f j f 0

1
k

1
j

1
2
2

where

1
= --- B f 0 f 0 l f 0
2

We take the first variation of potential energy and set it equal to zero to minimize the potential energy.

1 n
1
n
n
= ---
Cj C k B f j f k + Cj C k B f j f k
Cj l fj --- B f 0 f j + B f j f 0 or

1
k

1
j

1
2
2

1 n
1
n
n
n
n
= ---
Cj C k B f j f k +
Cj C k B f k f j
Cj l fj --- B f 0 f j + B f j f 0
k 1 j 1
j 1
2 j 1 k 1
2

Symmetric Bilinear functionals: B f j f k = B f k f j ; B f 0 f j = B f j f 0

1 n
n
n
= ---
Cj C k B f j f k + B f j f k
Cj l fj B f 0 f j =
j 1
2 j 1 k 1
=

j 1 k 1 Cj Ck B fj fk j 1 Cj l fj B f0 fj
n

= 0

K jk = B f j f k

where

j 1 k 1 Cj B fj fk Ck l fj B f0 fj

k 1 B fj fk Ck l fj B f0 fj
Matrix Form: K C = R

j 1 Cj k 1 B fj fk Ck l fj B f0 fj
j = 1 to n
R j = l fj B f 0 f j

= 0 or

M. Vable

MEEM 5170: Rayleigh-Ritz

Fall 2015

K is called the stiffness matrix and because the bilinear functional is symmetric, the stiffness matrix is symmetric.
If B u u 0 for any u, then the stiffness matrix K is positive definitea property that is used in solution procedures of algebraic

equations and in eigenvalue problems of dynamic systems. The convergence of the series is in the following sense
uN uM

if

NM

As degree of freedom increases, the potential energy will decrease.

Potential Energy at equilibrium:

*

Let Cj represent the solution of the algebraic equations (the values at equilibrium), that is

k 1 B fj fk Ck

= l fj B f 0 f j

*
* n
*
*
*
*
1 n
1 n
n
n
Cj
C j l fj B f 0 f j
= 0 + ---
B f j f k C k
C j l fj B f 0 f j = 0 + ---
C j l fj B f 0 f j
j

1
j

1
j

1
j

1
k

1
2
2

or
*
*
*
1 n
1 n
C l fj B f 0 f j = 0 ---
C R
= 0 ---
2 j 1 j
2 j 1 j j

or

*
*
*
1 n
C R = W 2
0 = ---
2 j 1 j j

At equilibrium, the potential energy of the system is negative of half the work potential.

M. Vable

MEEM 5170: Rayleigh-Ritz

Fall 2015

C.1 A beam and its loading are as shown below. Use the Rayleigh-Ritz method with one and two parameters to determine the deflection at x = 0.25L, x = 0.5L, x = 0.75L, and x = L, and the potential energy function. Compare your results with the analytical solution.
Assume that EI is constant for the beam.
y

w
x

A
L