Rayleigh-Ritz Method in Elasticity Analysis
Translated by
ScribdTranslationsRayleigh-Ritz Method in Elasticity Analysis
Translated by
ScribdTranslationsSimón Bolívar University Introduction to the MEF
3. Rayleigh-Ritz Method
The solution to the elasticity problem consists of finding the function des-
valid placement for the entire domain Ω that meets the conditions of
contour.
The Rayleigh-Ritz method proposes an approximate solution to solve
the problem of elasticity in its energetic formulation. The approximate solution
is based on the method of separation of variables (i.e. the spatial component
is independent of the temporal component) and on the idea that a function
continuation can be expressed as a linear combination of a large number of
functions, i.e.:
u≈ =Φ q (3.1)
The functions these function calls of Ritzy must meet the si-
following conditions:
They must verify the essential boundary or displacement conditions for
the entire boundary Γ of the solid.
They must be continuous and differentiable for the entire domain Ω and of a certain order.
equal to or greater than that of the differential operators present in the equations.
If the number as the Ritz functions tend to infinity, the approximate solution tends to
to the exact solution.
The matrix Φ in the most general case has dimension 3× and the vector q has
dimension.
3.1. Internal energy of deformation (stiffness matrix)
Introducing the Rayleigh-Ritz approach in the internal energy of deformation
we have:
1 1
= u L D Lu = q Φ L D LΦ q (3.2)
2 2
Defining matrix B as the result of the application of the differential operator
to the matrix of Rits functions, i.e.:
B = LΦ (3.3)
Euro Casanova, 2005 14
Simón Bolívar University Introduction to the MEF
The strain vector takes the form:
=Bq (3.4)
The stress vector takes the form:
=DB q (3.5)
The internal energy of deformation take the form:
1
= q B D B q (3.6)
2
From this last expression, it is evident that the coefficient vector can go out.
from the integral, since it does not depend on the spatial coordinates, then:
1
K q =q (3.7)
2
Where the global stiffness matrix is denoted and is calculated as:
K=B D B (3.8)
This matrix is of dimension ×y turns out to be symmetric, i.e. K=K.
3.2. Kinetic energy (mass matrix)
From the Rayleigh-Ritz approximation, it is possible to calculate the function
speedu˙as:
≈u̇ Φ q =Φq˙ (3.9)
Introducing this approach into kinetic energy, we have:
1 1
= u̇u̇ = Φ
q̇q̇ (3.10)
2 2
In this expression, it is evident that the coefficient vector q˙ can be taken out of the
integral, then:
1
=q˙ M q˙ (3.11)
2
The global mass matrix is called and is calculated as:
M= Φ Φ (3.12)
This matrix is of dimension and it turns out to be symmetric, i.e. M = M.
Euro Casanova, 2005 15
Universidad Sim´on Bol´ıvar Introduction to the MEF
3.3. Work of conservative forces (force vector)
Introducing the Rayleigh-Ritz approximation in the expression for work
of external conservative forces and noting that q = Φ q, we have:
u g +u t+ u p
=q Φ g +q Φ t + q Φp
=q f
Dondefse refers to the global force vector and is calculated as:
f=Φ g +Φ t + Φ p (3.13)
This vector is of dimension.
3.4. Equations of motion
Taking into account the results presented above, the potential energy
Total and kinetic energy take the following discrete forms.
1
P=qKq-qf 3.14
2
1
=q˙ M q˙ (3.15)
2
Note that the dimension of the matrices and vectors that make up these expressions
is equal to the number of Ritz functions used in the approximation,
i.e. Consequently, these expressions can be introduced into the problem
dynamic formulated in section 2.1.2, having then La equations
barn of the form:
Π
− + =0 =1 (3.16)
˙
By introducing the total potential energy Π into the -th Lagrange equation
and taking into account that the stiffness matrix is symmetric (i.e. K=K ⇒
Euro Casanova, 2005 16
Simón Bolívar University Introduction to the MEF
= ), we have:
Π 1
= q K q-q f
2
1
= q K q− q f
2
1
= −
2
1
= + −
2
1
= ( + )−
2
= −
As for writing for the degrees of freedom, we have:
Π
= −
Π Π
= − ⇒ =Kq−f
..
.
Π
= −
In the case of kinetic energy, similarly and taking into account that the
Euro Casanova, 2005 17
Simón Bolívar University Introduction to the MEF
The mass matrix depends on the spatial coordinates, we have:
=Mq˙
˙
=Mq¨
˙
Finally, the discrete problem is written as:
Mq¨+Kq=f 3.17
It is important to emphasize that the essential boundary conditions are already included.
within this formulation since the Ritz functions used verify them.
Consequently, it is only necessary to include the initial conditions, i.e.:
u=u in Ω=
u̇ = u̇ in Ω=
In general, what is sought when solving this problem is the value of q, since
with this vector it is possible to calculate the displacement function through:
u = Φ q (3.18)
Once the function has been obtained we can proceed to estimate other functions
of interest such as deformations and stresses:
=Bq =DB q=D (3.19)
3.5. Limitations of the Rayleigh-Ritz method
The main advantage of the Rayleigh-Ritz method is its conceptual simplicity and its
excellent precision when choosing Ritz functions, even with a reduced number,
that are quite similar to the function that is desired to be approximated. However, this
the method presents several disadvantages that make its systematic application difficult
(programmable) for the resolution of general elasticity problems. Between
We have these disadvantages:
Euro Casanova, 2005 18
Simón Bolívar University Introduction to the MEF
The Ritz functions must be chosen by the user and in general, they are
different for each problem. In fact, the precision of the method depends
strongly from the chosen functions.
The chosen Ritz functions must be valid for the entire domain of
problem, which represents a difficulty for real problems where the
domains have complex three-dimensional geometries.
Ritz functions must satisfy the essential boundary conditions.
of the domain, which can represent a huge difficulty for a problem
real.