Finite Element Method

By Dr. Sohail Iqbal

 Modeling & Simulation
• Modeling and simulation is the foundation of predictive and functional design
• Idea of modeling and simulation is inherent in the basic concept of science
• To quantify the physical observations, physical laws are expressed in the form of
mathematical equations
• Since physical laws generally change in space and time , the equations
describing the physical systems are generally differential equations
• Analytical solution of such equations are possible on fast digital
computers(simulations)
• Numerical solutions are approximate solutions which can approach the exact
solution with better numerical algorithms and sufficient computing power
• The finite element method (FEM) provides a powerful numerical simulation
technique to simulate complicated scientific and engineering models
Finite Difference Method
Based on Taylor series expansion of the function leading
to numerical approximation of derivatives.
The differential equations are then changed to finite
difference equations, which are then solved at discrete
mesh points subject to specified initial and boundary
conditions.
Finite Volume Method
Based on integrating the differential equations over a
control volume and then applying the finite difference
techniques
The integral approach ensures the conservation of
physical quantities
 Numerical Simulation Methods
DETERMINISTIC METHODS
• Based on equation
• Generally discretize a continuum represented by
differential equations into finite domains
• Represented by discrete equations
STOCHASTIC METHODS
• Based on probabilities
 Formulations
Variational Formulation
• Uses energy minimization based on Rayleigh-
Ritz method of variational formulation
Weighted Residual Formulation
• Uses error minimization based on methods
such as Galerkin method of the weighted
residual formulation
 Variational Method

• Based on minimization and maximization of functionals subject to

specified constraints for the numerical approximation of an ordinary
or partial differential equation.
• Equation is put into equivalent weighted integral form and then an
approximate solution over the entire domain is assumed to be a
linear combination of the type
Ū(x)
U(x) represents an approximate solution in terms of
approximately chosen functions øi and undetermined coefficients
ci.
• Creation of approximate solution or trial function u(x) involves
creating a function øi that
1. satisfies boundary conditions
2. contains arbitrary adjustable constants Ci
• These coefficient are determined such that the integral equation
 Variational Method
Total P.E can be written as a functional in the
form of a definite integral that has an integrand
involving a single independent variable, x, a
function u(x), and possibly a derivative of u(x)
w.r.t x
 Variational Method
• Different approximation methods i.e Rayleigh- Ritz,
Galerkin, and least square methods differ from each
other in the choice of integral form, weight function and
approximation function.
• The essence of the approach is to find the functional in
the form of total potential energy of the system and to
invoke the stationarity of potential energy subject to
specified constraints to ensure equilibrium.
• The variational method suffer from the disadvantage
that the approximation functions for problems with
arbitrary domains are difficult to construct.
 The Rayleigh Ritz Method
• If variational form of the problem exist i.e we have a functional for the
problem, an approximate solution for the dependent variable can be
made by creating a solution function that minimizes the functional. The
selection of the solution function involves assuming a function that
1. satisfies boundary conditions
2. contain arbitrary constants that can be adjusted.
• The Rayleigh Ritz Method assumes a trial solution function of the form
Ū(x)
Substituting into function

• The functional now involves n arbitrary constants, Ci which are

determined by finding the extreme of the functional w.r.t each
constant. The resulting Ū(x) provides the approximate solution
 Principle of Minimum Potential Energy
For stable equilibrium a body must have minimum total potential
energy
Total P.E = U+V
Strain P.E U=
Force P.E V= -Fu

The true configuration of the deformed elastic continuum yields

a minimum value of total P.E. This is basics of the development
of approximation techniques for the equilibrium elasticity
problem.
The Principle of Minimum Potential Energy and the strong
formulation are exactly equivalent statements of the same
problem.

The exact solution (uexact) that satisfies the strong form, renders
the potential energy of the system a minimum.

So, why use the Principle of Minimum Potential Energy?

The short answer is that it is much less demanding than the
strong formulation. The long answer is, it
1. requires only the first derivative to be finite
2. incorporates the force boundary condition automatically. The
admissible displacement (which is the function that you need to
choose) needs to satisfy only the displacement boundary
condition
 Weak Formulations
• Weak formulations are tools for the analysis of mathematical equations
that permit the transfer of concepts of linear algebra to solve problems in
other fields such as partial differential equations.
• In a weak formulation, an equation is no longer required to hold
absolutely (and this is not even well defined) and has instead
weak solutions only with respect to certain "test vectors" or "
test functions".
• Note : For details please visit
http://
math.stackexchange.com/questions/408615/conceptual-difference-between-
strong-and-weak-formulations
 Rayleigh-Ritz Principle
The minimization of the potential energy is difficult to perform exactly.
The Rayleigh-Ritz principle is an approximate way of doing this.
Step 1. Assume a solution

Step 2. Plug the approximate solution into the potential energy

Step 3. Obtain the coefficients ao, a1, etc by setting boundary conditions
 FE Modification of the Rayleigh-Ritz Method

• In the Rayleigh-Ritz method

o A single trial function is applied throughout the entire region
o Trial functions of increasing complexity are required to model all but
the simplest problems
• The FE approach
o uses comparatively simple trial functions that are applied piece-wise
to parts of the region
o These subsections of the region are then the finite elements

14
 Galerkin Method
• Galerkin method incorporates differential
equations in their weak form.
 Finite Element Method (FEM)

Definition:
• Finite element method (FEM) is a numerical technique for finding approximate solutions
using the given boundry conditions.

Description:

• FEM gives approximate solution of the given problem.

• It is done by dividing the given model into smaller bodies and then perform analysis.
This is called Discretization.

• Connecting points of these bodies are called Nodes.

• The accuracy of solution depends upon the size of discreted body.

Fine discretization would give accurate results.

• Primary unknowns are displacements from which we find secondary unknowns which
16
are stresses.
 FEM METHODS
S.No Direct Approach Variational Approach Weighted Residual Approach
. (Rayleigh-Ritz) (Galerkin)
This approach can This approach can also Solution to non-linear and
be used only for be used for non-linear non-structural problems was
1 relatively simple problems achieved through the method
problems of weighted residuals (MWR)

In direct approach Variational approach It does not rely on a

we can directly write relies on the calculus variational statement. This
down the elemental of variation and approach is required in order
2 equations through involves extremizing a to generate the necessary and
physical reasoning functional sufficient number of
simultaneous equations to
solve for approximate solution
 FEM METHODS
S.No Direct Approach Variational Approach Weighted Residual Approach
. (Rayleigh-Ritz) (Galerkin)
Direct method is used It can be used for 3D It can be used for 3D
for one dimensional problems of complex problems which can not even
cases (straight shapes be solved by Variational
3 elements, like straight approach
bar and straight
beam)

One dimensional For complex For advanced types of

elements are simple geometries and problems, where the
enough that the curved surfaces this application of variational
characteristic matrix characteristic matrix principle is limited, the MWR
(structural stiffness cannot be obtained was found to provide the
4 matrix) can usually be using direct method. ideal theoretical basis for a
formulated by the Therefore, Variational much wider basis of problems
“direct method” approach is used as opposed to the Rayleigh-
Ritz method (RR)
 FEM METHODS
S.No Direct Approach Variational Approach Weighted Residual
. (Rayleigh-Ritz) Approach (Galerkin)

5 Direct method is used It is a systematic Functional needed for a

to determine the procedure for producing variational approach
element behavior more general FE cannot be written for
through matrix algebra approximations. The every Degree equations.
procedure requires only Galerkin method (GM) is
that a functional* be a way of formulating an
available approximate solution
when one knows the DE
but not the functional
 Idealization
Mathematical Models
• “A model is a symbolic device built to simulate and
predict aspects of behavior of a system.”
• Abstraction of physical reality
Implicit vs. Explicit Modelling
• Implicit modelling consists of using existent pieces of
abstraction and fitting them into the particular situation
(e.g. Using general purpose FEM programs)
• Explicit modelling consists of building the model from
scratch
 Dicretization
1. Finite Difference Discretization
• The solution is discretized
• Stability Problems
• Loss of physical meaning
2. Finite Element Discretization
• The problem is discretized
• Physical meaning is conserved on elements
• Interpretation and Control is easier
 Solution
1. Linear System Solution Algorithms
• Gaussian Elimination
• Fast Fourier Transform
• Relaxation Techniques

2. Error Estimation and Convergence Analysis

 Interpretations
Physical Interpretation:
The continous physical model is divided into finite
pieces called elements and laws of nature are applied
on the generic element. The results are then
recombined to represent the continuum.
Mathematical Interpretation:
The differetional equation reppresenting the system is
converted into a variational form, which is approximated
by the linear combination of a finite set of trial
functions.
 Notation
Elements are defined by the following
properties:
1. Dimensionality
2. Nodal Points
3. Geometry
4. Degrees of Freedom
5. Nodal Forces
Element Types
Historical Background
 Chronicle of Finite Element Method
Year Scholar Theory

1941 Hrennikoff Presented a solution of elasticity problem using one-dimensional elements.

1943 McHenry Same as above.

1943 Courant Introduced shape functions over triangular subregions to model the whole region.

1947 Levy Developed the force (flexibility) method for structure problem.

1953 Levy Developed the displacement (stiffness) method for structure problem.

1954 Argyris & Kelsey Developed matrix structural analysis methods using energy principles.

1956 Turner, Clough, Derived stiffness matrices for truss, beam and 2D plane stress elements. Direct
Martin, Topp stiffness method.

1960 Clough Introduced the phrase finite element .

1960 Turner et. al Large deflection and thermal analysis.

1961 Melosh Developed plate bending element stiffness matrix.

1961 Martin Developed the tetrahedral stiffness matrix for 3D problems.

1962 Gallagher et al Material nonlinearity.

 Chronicle of Finite Element Method
Year Scholar Theory

1963 Grafton, Strome Developed curved-shell bending element stiffness matrix.

1963 Melosh Applied variational formulation to solve nonstructural problems.

1965 Clough et. al 3D elements of axisymmetric solids.

1967 Zienkiewicz et. Published the first book on finite element.

1968 Zienkiewicz et. Visco-elasticity problems.

1969 Szabo & Lee Adapted weighted residual methods in structural analysis.

1972 Oden Book on nonlinear continua.

1976 Belytschko Large-displacement nonlinear dynamic behavior.

~1997 New element development, convergence studies, the developments of

supercomputers, the availability of powerful microcomputers, the development of user-
friendly general-purpose finite element software packages.
 Finite Element Analysis Steps
Classifying the problem:
The First step is to identify the problem that what physical conditions are given
and what is required to be found.
Creating a Model:
The analysis is applied to an appropriate model representing the original physical
model. Analyzing a model is much easier than dealing with the original problem.
Discretization of the Model
The model is divided into a finite continuous mesh to be analysed using FEA.
Defining relations
Appropriate relations/ Stiffness matrices are formed to find out primary unknowns
( Displacements ).
Solving for unknowns
Primary unknowns are then used to find secondary unknowns which are stresses
etc.
Interpreting the results
Results are then interpreted on the original physical problem.
30
 Finite Element Analysis

Steps used in FEA (cont) :

31
 Finite Element Analysis
Advantages:

• FEA is applicable to any field problem.

• There is no restriction of shape or size of the model.

• Boundry conditions or loading conditions can be taken anywhere on

the model according to the required unknowns.

• Can also deal with composite materials.

• Mesh size can easily be changed according to the desired accuracy.

32
 Finite Element Analysis Applications

It is used to analyze structural and non structural problems

– Stress analysis
– Buckling
– Vibration analysis
– Fluid flow
– Heat transfer
– Distribution of electric and magnetic potentials etc

33
 Finite Element Analysis Software’s

• Includes:
– NASTRAN/PATRAN, CREO, ADAMS, INTELLI-SUITE,
ANSYS, COSMOS, NISA, ALGOR, CATIA etc
• Involves:
– Pre-processing
– Numerical analysis
– Post-processing

34
 1D elements & computations procedure
• 1D elements include
– Straight bar loaded axially
– Straight beam loaded laterally
– Bar conducts heat or electricity etc
• Degree of freedom (DOF) is number of parameters that may vary
independently
• Total DOF in structure = number of nodes * DOF at each node
• Order of structural stiffness = Total DOF in structure * number of
nodes in structure
• i.e spring element with 2 nodes has order of structural stiffness of
2*2

35
 References
• Concepts & applications of finite element analysis 4th edition,
by Robert D. cook
• Fundamentals of Finite element analysis 1st edition, by David
V. Hutton
• Class notes
• Wikipedia
• Web Links

36