The Finite Element Analysis

PREPARED BY
MANJUNATHA T V
ASSISTANT PROFESSOR
CAMBRIDGE INSTITUTE OF TECHNOLOGY
 Course Outline

 Module1: Introduction To Finite Element Method

 Module 2: One Dimensional Analysis Of Bars
 Module 3: beams and shafts
 Module 4: heat transfer problems
 Module 5: axisymmetric sold elements and dynamic
considerations.
 MODULE:1- STEPS IN FEA

 Step 1-: Discretization of Continuum

 Step 2-: Selection of Displacement Function
 Step 3-: Formulation of Elemental Stiffness Matrix
 Step 4- : Formulation of Global Stiffness Matrix
 Step 5-: Formulation of Global Force Vector
 Step 6-: Formation of overall equilibrium equation
 Step 7-: Applying or Incorporation of Boundary Condition
 Step 8- : Solution For The Continuum Nodal Displacements
 Step 9-:Computation of Element Strain And Stress
 Finite element method

 The Finite Element Method (FEM) is a numerical technique for

finding approximate solutions to boundary value problems for
partial differential equation.
 In simple terms, FEM is a method for dividing up a very
complicated problem into small elements that can be solved in
relation to each other.
Finite Element Method
 Discretization of Continuum
considerations
 First step of FEA, Dividing the continuum in to small regions of finite
dimensions called as finite elements.
 Assemblage of theses finite elements connected at the finite number
of joints called as nodes.
Considerations
 Type of element
 Size of element
 Number of element
 Location of nodes
 Node numbering scheme
 Selection of Displacement Function

 Polynomial Function
Ex:

 Trigonometric Function
 Advantages of FEA
 It can readily handle complex geometry.
 It can handle complex analysis types like vibration ,heat transfer ,fluids etc.
 FEA can easily applied to irregular shaped object composed of several materials
having mixed boundary conditions
 Can handle complex or any type of loading: i. Node-based loading (point
loads). ii. Element-based loading (pressure, thermal, inertial forces). iii. Time or
frequency dependent loading. Applicable to linear and nonlinear problems
 It can handle complex restraints: Indeterminate structures can be analyzed.
 It can handle bodies comprised of nonhomogeneous materials: Can handle
bodies comprised of non-isotropic materials: Orthotropic & Anisotropic.
 Special material effects are handled such as temperature dependent properties
plasticity , creep ,etc. Applicable to steady-state, time dependent and eigenvalue
 Disadvantages FEA

 A specific numerical result is obtained for a specific problem.

 Computational time involved in the solution of the problem is high.
 The FEM is applied to an approximation of the mathematical model of a
system (the source of so-called inherited errors).
 Experience and judgment are needed in order to construct a good finite
element model.
 A powerful computer and reliable FEM software are essential.
 Input and output data may be large and tedious to prepare and interpret.
 Numerical errors such as the limitation of the number of significant digits,
rounding –off occur very often.
 Fluid elements with boundaries at infinity can be computed and treated by
using boundary element method. Mistake by users can be fatal.
 FEA Applications

1. Structural Analysis: linear and non-linear models

1. Vibrational Analysis 2. Fatigue Analysis 3. buckling analysis 4. Crash analysis
of cars trains air crafts
2. structural analysis : 1.Heat Transfer Analysis : consist of a steady state or
transient transfer. 2.fluid flow analysis: fluid flow through pipes
3. Eigen value problems, Transient problems.
4. 5. Bio-medical problems, aerospace engineering
6. Nuclear engineering
7.Analysis of robots, computer chip, analysis earthquakes, analysis of casting
forging welding and machining process.
8. Geo-mechanic problems: analysis of soils, underground openings, rock joint and
soil structure interaction problems
 FEM Help the Design Organization

 Simulation using the FEM also offers important business advantages to the
design organization:
 Reduced testing and redesign costs thereby shortening the product
development time.
 Identify issues in designs before tooling is committed.
 Refine components before dependencies to other components prohibit
changes.
 Optimize performance before prototyping.
 Discover design problems before litigation.
FEA fundamentals
 Software packages for FEA

 • ANSYS
 • NASTRAN
 • PATRAN
 • NISA / DISPLAY III
 • LS DYNA
 • HYPERMESH
 • CATIA
 • Pro-E(CREO)
 • SOLID WORKS
 • COSMOS
Numerical Methods
Types of Elements
Types of Elements
Types of Elements
 Size of Element

 FINE E;EMNTS
 ASPECT RATIO
 MORE COMPUTATIONAL TIME
 Number of elements
 SIZE OF ELEMNTS
 NUMBER OF DEGREE OF FREEDOM
 MORE ACCURATE
 MORE COMPUTAIONAL TIME
 HIGH STORAGE CAPACITY
OR
MEMORY
Location Of Nodes
 Node Numbering Scheme
 HALF BAND WIDTH B=(D+1)f
 HALF BAND WIDTH B=(D+1)f
 .
Plane strain
plane stress
Examples of plane stress
plane stress
Plane Strain
Simplex, complex and multiplex
elements
Simplex, complex and multiplex elements
 Truss
 Truss members are for the analysis of skeletal structural type systems .
Assemblage of slender bars fastened at their ends by pins, balls or by socket
joints as hinges.
 A truss element is a straight bar of an arbitrary cross-section, which can
deform only in its axis direction when it is subjected to axial forces.
 Truss elements are also termed as bar elements. planar trusses and space
trusses.
 In planar trusses, there are two components in the x and y directions for the
displacement as well as forces at a node. In this members lie in the plane and
loaded in same plane.
 For space trusses, there will be three components in the x, y and z directions
for both displacement and forces at a node. Member not lying in same plane.
TRUSS APPLICATIONS
Truss applications
Analysis of truss
Analysis of truss
Analysis of truss
Analysis of truss
 MNTamin, CSMLab

Truss
 Truss members are for the analysis of skeletal structural type
systems . Assemblage of slender bars fastened at their ends by
pins, balls or by socket joints as hinges.
 A truss element is a straight bar of an arbitrary cross-section,
which can deform only in its axis direction when it is subjected
to axial forces.
 Truss elements are also termed as bar elements. planar trusses
and space trusses.
 In planar trusses, there are two components in the x and y
directions for the displacement as well as forces at a node. In
this members lie in the plane and loaded in same plane.
 For space trusses, there will be three components in the x, y and
z directions for both displacement and forces at a node. Member
not lying in same plane.

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

TRUSS APPLICATIONS

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

Plane Trusses
A typical two-dimensional plane truss is shown. It comprises of two-force
members, connected by frictionless joints. All loads and reaction forces are
applied at the joints only.
Note: There are two displacement components at a given node j, denoted
by Q2j-1 and Q2j.

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

Local & Global Coordinate Systems

Local and global coordinate systems are

shown at left.
In local coordinate (x’), every node has
one degree of freedom, while in global
coordinate (x, y), every node has two
degrees of freedom.
The nodal displacements, in the local
coordinate system is,

q   q
' '
1
' T
q 
2

In the global coordinate system,

q  q1 q4 
T
q2 q3
Local and global coordinate systems.

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

4-5 Relation Between Coordinate Systems
Consider a deformed truss member as shown. We can now establish a
relationship between {q’} and {q} as follows:

q1'  q1 cos   q2 sin 

q2'  q3 cos   q4 sin 

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

Direction Cosines
To eliminate the  terms from previous equations, we define direction cosines,
such that
l  cos  and m  sin 

The relation between {q’} and {q} can now be written as

q1'  q1 (l )  q2 (m)
q2'  q3 (l )  q4 (m)

which can be written in matrix form as

 q1 
 q1'   l m 0 0  q2 
 '  q 
q 2   0 0 l m  3 
q4 

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

Or, in a condensed matrix form

q  Lq
'

where [L] is a rectangular matrix called the transformation matrix,

given by,

l m 0 0 
L   
 0 0 l m 

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

Formula for Evaluating l and m
Using a trigonometry relation, we observe

x2  x1
l  cos  
le
y2  y1
m  sin  
le
le  ( x2  x1 ) 2 +
 ( y2  y1 ) 2

Note: Coordinates (xi, yi) are based on local coordinate system.

PLANE TRUSS ELEMENTS
 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

Element Stiffness Matrix

A truss element is a one-dimensional (bar) element, when it is viewed in local
coordinate system.
Thus, element stiffness matrix for a truss element in local coordinate,
AE  1 1
k truss  k bar
'

le  1 1 
The internal strain energy in the truss element, in local coordinate system
is,
1 'T '

U e  q k truss q '
'

2

Substitutingq  L q
'
We need the expression
we get, for [k] when viewed in
global coordinate…
1
U  L q k truss L q
' T '
e
2
1 T
 
U e'  q L  k truss L  q
2
T '

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

In the global coordinates system,
1 T
Ue  q k truss q
2
Since internal strain energy is independent of coordinate system, Ue = U’e.

k truss   L  k truss  L 
T '
Therefore,
 l 0
m 0 
k truss    AE  1 1  l m 0 0 
 0 l  le  1 1  0 0 l m 
 
 0 m 
Simplifying,
 l2 lm l 2lm 
 
AE  lm m2 lm m 2 
k truss  k  
le  l 2 lm l2 lm 
 
 lm m
2
lm m 2 

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

Stress Calculations
Normal stress in a plane truss element, in local coordinate system is,

  E B q ' 

q '  L q
In the global coordinate system, since

  E  B    L   q

Expanding the [B] and [L] matrices,

 q1 
 
1  l m 0 0  q2 
  E 1 1    q 
l  0 0 l m  3

 q4 

 q1 
q 
E  
Or    l  m l m 2 
l q3 
q4 
PLANE TRUSS ELEMENTS
 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

Example -1
For a plane truss shown, determine: a) displacements at nodes 2 and 3; b) stresses
in each element; c) reaction forces at the supports. Use: Ee = 29.5 x 106 psi and Ae
= 1 in2 for all elements.

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

Solution:-
STEP-1. Establish nodal coordinate data and element connectivity information

Nodal coordinate
Node x y
1 0 0
2 40 0
3 40 30
4 0 30

Element connectivity
Element Node 1 Node 2
1 1 2
2 3 2
3 1 3
4 4 3

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

STEP-2. Compute the direction cosines
Element le l m

1 40 1 0

2 30 0 -1

3 50 0.8 0.6

4 40 1 0

STEP-3. Write the stiffness matrix for each element

For Element ;-1
1 0 1 0
 0
29.5 106  1  0 0 0
k (1) 
40  1 0 1 0
 
0 0 0 0

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

For elements 2, 3, and 4
0 0 0 0
 0  1
29.5  106 1 0 1
k ( 2) 
30 0 0 0 0
 
0  1 0 1

 0.64 0.48  0.64  0.48

 0.36  0.48  0.36
29.5  106  1  0.48
k (3) 
50  0.64  0.48 0.64 0.48 
 
  0.48  0.36 0.48 0 .36 

1 0 1 0
 0
29.5 106 1  0 0 0
k ( 4) 
40  1 0 1 0
 
0 0 0 0

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

Step-4. Assemble the global stiffness matrix [K] for the entire truss structure

 22.68 5.76 15.0 0 7.68 5.76 0 0

 5.76 4.32 0 0 5.76 4.32 0 0 

 15.0 0 15.0 0 0 0 0 0
 
29.5 106  0 0 0 20.0 0 20.0 0 0
K  
600  7.68 5.76 0 0 22.68 5.76 15.0 0
 
 5.76 4.32 0 20.0 5.76 24.32 0 0
 0 0 0 0 15.0 0 15.0 0
 
 0 0 0 0 0 0 0 0 

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

5. Assemble the global system of linear equations.

 22.68 5.76 15.0 0 7.68 5.76 0 0   Q1   0 

 5.76 4.32
 0 0 5.76 4.32 0 0  Q 2   0 
 15.0 0 15.0 0 0 0 0 0  Q3   20000 
6     
29.5 10  0 0 0 20.0 0 20.0 0 0  Q 4   0 
  
600  7.68 5.76 0 0 
22.68 5.76 15.0 0 Q5   0 
 
 5.76 4.32 0 20.0 5.76 24.32 0 0  Q6   25000 
 0     
0 0 0 15.0 0 Q
15.0 0  7   0 
 
 0 0 0 0 0 0 0 0  Q8   0 

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

Step-6. Impose boundary conditions & write the reduced system of linear equations

We have; Q1 = Q2 = Q4 = Q7 = Q8 = 0. fixed

Using elimination method, the above equations reduced to

15 0 0  Q3   20000 

29.5 10  6
 Q    0 
0 22.68 5.76  5   
600 
 0 5.76 24.32 Q6   25000

Step-7. Solve the reduced equations using the Gaussian elimination method,
we get

Q3   27.12  10 
3

   3 
Q5    5.65  10  in
Q   22.25  10 3 
 6  

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

Step-8. Compute the stress in each element
Element 1
 0 
 
29.5  106  0 
1   1 0 1 0 3 
 20000 psi
40 27.12  10 
 0 
Element 2
 5.65 103 
 3 
29.5 106  22.25 10 
2  0 1 0  1 3 
 21880 psi
30  27.12 10 
 0 
In the same manner,

 3  5208 psi
 4  4167 psi

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

Step-9. Determine support reactions

Using the 1st, 2nd, 4th, 7th, and the 8th equations, we get

PLANE TRUSS ELEMENTS

 MNTamin, CSMLab

SME 3033 – FINITE ELEMENT METHOD

 R1   15833
 R   3126 
 2   
 R4    21879  lb
 R    4167 
 7  
 R8   0 

PLANE TRUSS ELEMENTS