A Course Material on

Finite Element Analysis

By

Mr. M.SURENDRAN ME., (Ph.D)

R.GOPINATH ME
ASSISTANT PROFESSOR

DEPARTMENT OF MECHANICAL ENGINEERING

AL AMEEN ENGINEERING COLLEGE

ERODE 638 104

 QUALITY CERTIFICATE

This is to certify that the e-course material

Subject Code : ME6603

Subject : Finite Element Analysis

Class : III Year Mechanical

being prepared by me and it meets the knowledge requirement of the university curriculum.

Signature of the Author

Name: M.SURENDRAN M.E., (Ph.D)

Designation: ASSISTANT PROFESSOR

Signature of HOD

Name: PROF. G.THANGAVEL M.E., M.B.A.,

SEAL
 CONTENTS
S.NO TOPIC PAGE NO

UNIT-1 FINITE ELEMENT FORMULATION OF BOUNDARY VALUE PROBLEMS

1.1 INTRODUCTION 1

1.1.1 A Brief History of the FEM 1

1.1.2General Methods of the Finite Element Analysis 1

1.1.3General Steps of the Finite Element Analysis 1

1.1.4 Objectives of This FEM 2

1.1.5 Applications of FEM in Engineering 2

1.2 WEIGHTED RESIDUAL METHOD 2

1.3 THE GENERAL WEIGHTED RESIDUAL STATEMENT 5

1.4 WEAK FORMULATION OF THE WEIGHTED RESIDUAL 5
1.5 PIECE WISE CONTINUOUS TRIAL FUNCTION 6

1.6 EXAMPLES OF A BAR FINITE ELEMENT 8

1.6.1 Rigid Body 13

1.7 PRINCIPLE OF STATIONERY TOTAL POTENTIAL PSTP) 19

1.7.1 Potential energy in elastic bodies 19
1.7.2 Principle of Minimum Potential Energy 19

1.8 RAYLEIGH RITZ METHOD (VARIATIONAL APPROACH) 24

1.9 ADVANTAGES OF FINITE ELEMENT METHOD 24

1.10 DISADVANTAGES OF FINITE ELEMENT METHOD 24

UNIT 2 ONE DIMENSIONAL FINITE ELEMENT ANALYSIS

2.1 ONE DIMENSIONAL ELEMENTS 25

2.2 LINEAR STATIC ANALYSIS( BAR ELEMENT) 28

2.3 BEAM ELEMENT 28

2.4 1-D 2-NODED CUBIC BEAM ELEMENT MATRICES 33

2.5 DEVELOPMENT OF ELEMENT EQUATION 34

2.6 BEAM ELEMENT 42
2.6.1 ELEMENT MATRICES AND VECTORS 45
UNIT 3 TWO DIMENSIONAL FINITE ELEMENT ANALYSIS

3.1 INTRODUCTION 54

3.2 THREE NODED LINEAR TRIANGULAR ELEMENT 54

3.3 FOUR NODED LINEAR RECTANGULAR ELEMENT 55

TWO-VARIABLE 3-NODED LINEAR TRIANGULAR
3.4 56
ELEMENT

3.5 STRAIN STRESS RELATION 60

3.5.1 Plane stress conditions 61
3.5.2 Plane strain conditions 61
GENERALIZED COORDINATES APPROACH TO NODEL
3.6 65
APPROXIMATIONS

3.7 ISOPARAMETRIC ELEMENTS 66

STRUCTURAL MECHANICS APPLICATIONS IN 2
3.8 71
DIMENSIONS
UNIT 4 DYNAMIC ANALYSIS USING ELEMENT METHOD

4.1 INTRODUCTION 88
4.1.1 Fundamentals of Vibration 88
4.1.2 Causes of Vibrations 88
4.1.3 Types of Vibrations 88

4.2 EQUATION OF MOTION 89

4.3 CONSISTENT MASS MATRICES 94

4.3.1 Single DOF System 94
4.3.2.Multiple DOF System 98

4.4 VECTOR ITERATION METHODS 99

4.5 MODELLING OF DAMPING 102

4.5.1 Proportional Damping (Rayleigh Damping) 102
4.5.2 Frequency Response Analysis 105
4.6 TRANSIENT RESPONSE ANALYSIS 106
4.6.1Cautions in Dynamic Analysis 107
UNIT -5 APPLICATIONS IN HEAT TRANSFER &FLUID MECHANICS

5.1 ONE DIMENSIONAL HEAT TRANSFER ELEMENT 111

5.1.1Strong Form for Heat Conduction in One Dimension
111
with Arbitrary Boundary Conditions
5.1.2Weak Form for Heat Conduction in One Dimension
112
with Arbitrary Boundary Conditions

5.2 APPLICATION TO HEAT TRANSFER 2-DIMENTIONAL 112

5.2.1Strong Form for Two-Point Boundary Value Problems 112
5.2.2Two-Point Boundary Value Problem With
112
Generalized Boundary Conditions
5.2.3 Weak Form for Two-Point Boundary Value Problems 114

5.3 SCALE VARIABLE PROBLEM IN 2 DIMENSIONS 114

5.4 2 DIMENTIONAL FLUID MECHANICS 117

QUESTION BANK 120

TEXT BOOKS:
1. P.Seshu, Text Book of Finite Element Analysis, Prentice-Hall of India Pvt. Ltd. New
Delhi, 2007.
2. J.N.Reddy, An Introduction to the Finite Element Method, McGraw-Hill International
Editions(Engineering Mechanics Series), 1993.
3. Cook,Robert.D., Plesha,Michael.E & Witt,Robert.J. Concepts and Applications of
Finite Element Analysis,Wiley Student Edition, 2004.
4. Chandrupatla & Belagundu, Introduction to Finite Elements in Engineering, 3rd
Edition, Prentice-Hall of India, Eastern Economy Editions.
ME6603 FINITE ELEMENT ANALYSIS L T P C
3 0 03
OBJECTIVES:
1. To introduce the concepts of Mathematical Modeling of Engineering Problems.
2. To appreciate the use of FEM to a range of Engineering Problems

UNIT I INTRODUCTION 9
Historical Background Mathematical Modeling of field problems in Engineering Governing Equations
Discrete and continuous models Boundary, Initial and Eigen Value problems Weighted Residual
Methods Variational Formulation of Boundary Value Problems RitzTechnique Basic concepts of the
Finite Element Method.

UNIT II ONE DIMENSIONAL PROBLEM 9

One Dimensional Second Order Equations Discretization Element types- Linear and Higher order
Elements Derivation of Shape functions and Stiffness matrices and force vectors- Assembly of Matrices
Solution of problems from solid mechanics and heat transfer. Longitudinal vibration frequencies and mode
shapes. Fourth Order Beam Equation Transverse deflections and Natural frequencies of beams.

UNIT III TWO DIMENSIONAL SCALAR VARIABLE PROBLEMS 9

Second Order 2D Equations involving Scalar Variable Functions Variational formulation Finite Element
formulation Triangular elements Shape functions and element matrices and vectors. Application to Field
Problems Thermal problems Torsion of Non circular shafts Quadrilateral elements Higher Order
Elements.

UNIT IV TWO DIMENSIONAL VECTOR VARIABLE PROBLEMS 9

Equations of elasticity Plane stress, plane strain and axisymmetric problems Body forces and
temperature effects Stress calculations Plate and shell elements.

UNIT V ISOPARAMETRIC FORMULATION 9

Natural co-ordinate systems Isoparametric elements Shape functions for iso parametric elements One
and two dimensions Serendipity elements Numerical integration and application to plane stress
problems Matrix solution techniques Solutions Techniques to Dynamic problems Introduction to
Analysis Software.

OUTCOMES:
Upon completion of this course, the students can able to understand different mathematical Techniques
used in FEM analysis and use of them in Structural and thermal problem

TEXT BOOK:
1. Reddy. J.N., An Introduction to the Finite Element Method, 3rd Edition, Tata McGraw-Hill, 2005
2. Seshu, P, Text Book of Finite Element Analysis, Prentice-Hall of India Pvt. Ltd., New Delhi, 2007.
REFERENCE BOOKS:
1. Rao, S.S., The Finite Element Method in Engineering, 3rd Edition, Butterworth Heinemann, 2004
2. Logan, D.L., A first course in Finite Element Method, Thomson Asia Pvt. Ltd., 2002
3. Robert D. Cook, David S. Malkus, Michael E. Plesha, Robert J. Witt, Concepts and Applications of
Finite Element Analysis, 4th Edition, Wiley Student Edition, 2002.
4. Chandrupatla & Belagundu, Introduction to Finite Elements in Engineering, 3rd Edition, Prentice
Hall College Div, 1990
5. Bhatti Asghar M, Fundamental Finite Element Analysis and Applications, John Wiley & Sons, 2005
(Indian Reprint 2013)

1
ME6603 Finite Element Analysis
UNIT I
FINITE ELEMENT FORMULATION OF BOUNDARY VALUE PROBLEMS
1.1 INTRODUCTION
The finite element method constitutes a general tool for the numerical solution of partial
differential equations in engineering and applied science
The finite element method (FEM), or finite element analysis (FEA), is based on the idea of
building a complicated object with simple blocks, or, dividing a complicated object into small and
manageable pieces. Application of this simple idea can be found everywhere in everyday life as well as
in engineering.
Examples:
Lego (kidsplay) Buildings
Approximation of the area of a circle:

Element Si

i
R

Why Finite Element Method?

Design analysis: hand calculations, experiments, and computer simulations
FEM/FEA is the most widely applied computer simulation method in engineering
Closelyintegrated with CAD/CAM applications
1.1.1 A Brief History of the FEM
1943 --- Courant (variational method)
1956 --- Turner, clough, martin and top(stiffness)
1960 --- Clough (finite element plan problems)
1970 --- Applications on mainframe computer
1980 --- Microcomputers, pre and post processors
1990 --- Analysis of large structural systems
1.1.2 General Methods of the Finite Element Analysis
1. Force Method Internal forces are considered as the unknowns of the problem.
2. Displacement or stiffness method Displacements of the nodes are considered as the
unknowns of the problem.
1.1.3 General Steps of the Finite Element Analysis

Discretization of structure
Numbering of Nodes and Elements
Selection of Displacement function or interpolation function
Define the material behavior by using Strain Displacement and Stress Strain
relationships
 Derivation of element stiffness matrix and equations
Assemble the element equations to obtain the global or total equations
Applying boundary conditions
Solution for the unknown displacements computation of the element strains and stresses
from the nodal displacements
Interpret the results (post processing).

1.1.4 Objectives of This FEM

Understand the fundamental ideas of the FEM
Know the behavior and usage of each type of elements covered in this course
Be able to prepare a suitable FE model for given problems
Can interpret and evaluate the quality of the results (know the physics of the problems)
Be aware of the limitations of the FEM (dont misuse the
FEM - a numerical tool)

1.1.5 Applications of FEM in Engineering

Mechanical/Aerospace/Civil/Automobile Engineering Structure analysis
(static/dynamic, linear/nonlinear) Thermal/fluid flows
Electromagnetics
Geomechanics
Biomechanics

1.2 WEIGHTED RESIDUAL METHOD

It is a powerful approximate procedure applicable to several problems. For non structural
problems, the method of weighted residuals becomes very useful. It has many types. The popular
four methods are,
1. Point collocation method,
Residuals are set to zero at n different locations X i, and the weighting function wi
is denoted as (x - xi).

(x xi) R (x; a1, a2, a3 an) dx = 0

2. Subdomain collocation method

3. Least square method,

[R (x; a1, a2, a3 an)]2 dx = minimum.

4. Galerkins method. wi = Ni (x)

Ni (x) [R (x; a1, a2, a3 an)]2 dx = 0, i = 1, 2, 3, n.
Problem I
Find the solution for the following differential equation.
 EI qo =0
The boundary conditions are u(0)=0, (0)=0,
(L)=0, (L)=0,

Given: The governing differential equation

EI qo =0
Solution: assume a trial function
Let u(x) = a0+a1x+a2x2+a3x3+a4x4..
Apply 1st boundary condition
x=0, u(x)=0
0=a0+0
a0=0

Apply 2nd boundary condition

x=0, =0
a1=0

Apply 3rd boundary condition

x=L, =0
a2=-[3a3L+6a4L2]

Apply 4th boundary condition

x=L, =0
a3=-4a4L

Substitute a0, a1, a2 and a3values in trial function

u(x)= 0+0-[3a3L+6a4L2] -4a4L
u(x)= a4[6 L2x2-4 Lx3+ x4]
= a4[6 L2 (2x)-12 Lx2+ 4x3]
=24 a4
R= EI qo =0
a4=
Substitute a4values in u(x)
u(x) = [x4-4Lx3+6L2x2]

Result:
Final solution u(x) = [x4-4Lx3+6L2x2]

Problem 2
The differential equation of a physical phenomenon is given by
 + = 4, 0 1
The boundary conditions are: y(0)=0
y(1)=1
Obtain one term approximate solution by using galerkin method
Solution:
Here the boundary conditions are not homogeneous so we assume a trial function as,
y=a1x(x-1)+x
first we have to verify whether the trial function satisfies the boundary condition or not
y=a1x(x-1)+x
when x=0, y=0
x=1, y=1
Resuldual R:
Y=a1x(x-1)+x=a1(x2-x)+x
=a1(2x-1)+1
=2 1

Substitute value in given differential equation.

2a1+y=4x
Substitute y vlue
R=2a1+a1x(x-1)+x-4x
In galerkins method

Substitute wi and R value in equation
a1=0.83
So one of the approximate solution is, y= 0.83x(x-1)+x
= 0.83x2-0.83x+x
y=0.83 x2+0.17x
Problem 3
Find the deflection at the center of a simply supported beam of span length l subjected to
uniform distributed load throughout its length as shown using (a) point collection method (b) Sub-
domain method (c)least squared and (d) galerkinn method.
Solution:

EI - = 0, 0

The boundary condition are y=0, x= 0and y=

EI =0 at x=0 and x=

Where, EI =

Let us select the trail function for deflection as,

y= a sin /
1.3 THE GENERAL WEIGHTED RESIDUAL STATEMENT
 After understanding the basic techniques and successfully solved a few problem general
weighted residual statement can be written as

R dx=0 for i= 1,2,..n

Where wi=Ni
The better result will be obtained by considering more terms in polynomial and trigonometric series.
1.4 WEAK FORMULATION OF THE WEIGHTED RESIDUAL STATEMENT.

The analysis in Section as applied to the model problem provides an attractive perspective to the
solution of certain partial differential equations: the solution is identified with a point, which
minimizes an appropriately constructed functional over an admis- sible function space. Weak
(variational) forms can be made fully equivalent to respective strong forms, as evidenced in the
discussion of the weighted residual methods, under certain smoothness assumptions. However, the
equivalence between weak (variational) forms and variational principles is not guaranteed: indeed, there
exists no general method of construct-

ing functionals I [u], whose extremization recovers a desired weak (variational) form. In this

sense, only certain partial differential equations are amenable to analysis and solution by
variational methods.

Vainbergs theorem provides the necessary and sufficient condition for the equivalence of a
weak (variational) form to a functional extremization problem. If such equivalence holds, the functional
is referred to as a potential.

Theorem (Vainberg)
Consider a weak (variational) form

G(u, u) := B(u, u) + (f, u) + (q , u)q = 0 ,

where u U , u U0 , and f and q are independent of u. Assume that G pos- sesses a

Gateaux derivative in a neighborhood N of u, and the Gateaux differen- tial Du1 B(u, u2) is
continuous in u at every point of N .

Then, the necessary and sufficient condition for the above weak form to be derivable from a
potential in N is that
Du1 G(u, u2) = Du2 G(u, u1) ,
Namely that Du1 G(u, u2) be symmetric for all u1, u2 = U0 and all u = N .
Preliminary to proving the above theorem, introduce the following two lemmas:

Lemma 1 Show that Dv I[u] = lim

 In the above derivation, note that operations and |=0 are not interchangeable (as they
both refer to the same variable ), while lim0 and |=0 are interchangeable, conditional upon
sufficient smoothness of I [u].

Lemma 2 (Lagranges formula)

Let I [u] be a functional with Gateaux derivatives everywhere, and u, u + u be any points
of U. Then,

I [u + u] I [u] = Du I [u + u] 0 < < 1.

To prove Lemma 2, fix u and u + u in U, and define function f on R as

f() := I[u + u] .

It follows that

F = df f ( + ) f ()
d = lim

0
= lim I [u + u + u] I [u + u] = Du I [u + u] ,
0
Where Lemma 1 was invoked. Then, u s i n g the standard mean-value theorem of
calculus,

1.5 PIECE WISE CONTINUOUS TRIAL FUNCTION

In weighted residual method the polynomial and trigonometric series are used as trial function.
This trial function is a single composite function and it is valid over the entire solution domain this
assumed trial function solution should match closely to the exact solution of the differential equation
and the boundary conditions, it is nothing but a process of curve fitting. This curve fitting is carried
out by piecewise method i.e., the more numbers of piece leads better curve fit. Piecewise method can
be explained by the following simple problem.
We know that the straight line can be drawn through any two points.
Let, (x)=sin is the approximated function for straight line segments.
One straight line segment
Two straight line segment

One Spring Element

x
i j
fi ui uj fj
 Two nodes: i, j
Nodal displacements: ui, uj (in, m, mm)
Nodal forces: fi, fj (lb, Newton) Spring constant (stiffness): k (lb/in,
N/m, N/mm)

Spring force-displacement relationship:

Linear

F Nonlinear

k F/ (> 0) is the force needed to produce a unit stretch.

We only consider linear problems in this introductory course. Consider the equilibrium of
forces for the spring.
At node 1 we have

fi F k(u j ui ) kui kuj

and at node j,
fj F k(u j ui ) kui kuj

In matrix form,
k k ui fi
k k uj fj
or, where

(element) stiffness matrix

u = (element nodal) displacement vector

f = (element nodal) force vector

Note:
That k is symmetric. Is k singular or non singular? That is, can we solve the
equation? If not, why?
Problem 4
To find the deformation of the shape
X

K1 K2
u1F1 u2F2 u3F3
1 2 3

For element 1,

k1 k 1 u2 f1 2
k1 k1 u3 f2 2
element 2,
2
k2 k2 u2 f
2
k2 k2 u3 f
where fI at node 2 F2
M is the (internal) force acting on local node i of element Consider the quilibrium of
forces at node
Checking the Results

Deformed shape of the structure

Balance of the external forces
Order of magnitudes of the numbers

Notes about the Spring Elements

Suitable for stiffness analysis

Not suitable for stress analysis of the spring itself
Can have spring elements with stiffness in the lateral direction,
Spring elements for torsion, etc.
1.6 EXAMPLES OF A BAR FINITE ELEMENT
The finite element method can be used to solve a variety of problem types in
engineering, mathematics and science. The three main areas are mechanics of materials, heat
transfer and fluid mechanics. The one-dimensional spring element belongs to the area of
mechanics of materials, since it deals with the displacements, deformations and stresses
involved in a solid body subjected to external loading.
Element dimensionality:

An element can be one-dimensional, two-dimensional or three-dimensional. A spring element

is classified as one-dimensional.

Geometric shape of the element

The geometric shape of element can be represented as a line, area, or volume. The one-
dimensional spring element is defined geometrically as:

Spring law
The spring is assumed to be linear. Force (f) is directly proportional to deformation () via the
spring constant k, i.e.

Types of degrees of freedom per node

Degrees of freedom are displacements and/or rotations that are associated with a node. A one-
dimensional spring element has two translational degrees of freedom, which include, an axial
(horizontal) displacement (u) at each node.

Element formulation
There are various ways to mathematically formulate an element. The simplest and limited
approach is the direct method. More mathematically complex and general approaches are energy
(variation) and weighted residual methods.
 The direct method uses the fundamentals of equilibrium, compatibility and spring law from a
sophomore level mechanics of material course. We will use the direct method to formulate the one-
dimensional spring element because it is simple and based on a physical approach.
The direct method is an excellent setting for becoming familiar with such basis concepts of
linear algebra, stiffness, degrees of freedom, etc., before using the mathematical formulation
approaches as energy or weighted residuals.
Assumptions
Spring deformation

The spring law is a linear force-deformation as follows:

f=k

f - Spring Force (units: force)

k - Spring Constant (units: force/length)

- Spring Deformation (units: length)

Spring Behaviour:
A spring behaves the same in tension and compression.
Spring Stiffness:
Spring stiffness k is always positive, i.e., k>0, for a physical linear system.
Nodal Force Direction:
Loading is uniaxial, i.e., the resultant force is along the element. Spring has no resistance to
lateral force.
Weightless Member:
Element has no mass (weightless).
Node Location:
The geometric location of nodes I and J cannot coincide, i.e., xi xj. The length of the element
is only used to visually see the spring.

A column of KE is a vector of nodal loads that must be applied to an element to sustain a

deformed state in which responding nodal DOF has unit value and all other nodal DOF are zero. In
other words, a column of KE represents an equilibrium problem.
Example, uI = 1, uJ = 0.

Spring element has one rigid body mode.

Inter-Element Axial Displacement

The axial displacement (u) is continuous through the assembled mesh and is described by a
linear polynomial within each element. Each element in the mesh may be described by a different
linear polynomial, depending on the spring rate (k), external loading, and constraints on the element.

Inter-Element Deformation

The deformation () is piecewise constant through the assembled mesh and is described by a
constant within each element. Each element in the mesh may be described by a different constant,
depending on the spring constant (k), external loading, and constraints on the element.

Inter-Element Internal Axial Force

The internal axial force (f) is piecewise continuous through the assembled mesh and is
described by a constant within each element. Each element in the mesh may be described by a different
constant, depending on the spring constant, external loading, and constraints on the element.
1.6.1 Rigid Body

A body is considered rigid if it does not deform when a force is applied. Consider rigid and
non-rigid bars subjected to a gradually applied axial force of increasing magnitude as shown.

The reader should note the following characteristics of rigid and non-rigid (flexible) bodies:

Force Magnitude - Even if forces are large, a rigid body does not deform. A non-rigid body will
deform even if a force is small. In reality, all bodies deform.
Failure - A rigid body does not fail under any load; while a non-rigid body will result either in
ductile or brittle failure when the applied load causes the normal stress to exceed the breaking
(fracture) stress b of the material. Brittle failure occurs when the applied load on the non-rigid
bar shown above causes the breaking strength of the bar to be exceeded.
Material - The material is not considered in a rigid body. Since a rigid body does not deform (
= 0) this is equivalent to an infinite modulus of elasticity. In contrast the modulus of elasticity
for a non-rigid material is finite, e.g., for steel, Esteel = 30 x 106 psi. (200 GPa). For rigid and
non-rigid bars the material laws are:

Rigid Body Motion

Rigid body motion occurs when forces and/or moments are applied to an unrestrained mesh
(body), resulting in motion that occurs without any deformations in the entire mesh (body). Since no
strains (deformations) occur during rigid body motion, there can be no stresses developed in the mesh.

A rigid body in general can be subjected to three types of motion, which are translation,
rotation about a fixed axis, and general motion which consists of a combination of both translation and
rotation. These three motion types are as follows:

Translation - If any line segment on the body remains parallel to its original direction during
the motion, it is said to be in translation. When the path of motion is along a straight line, the motion is
called rectilinear translation, while a curved path is considered as a curvilinear translation. The
curvilinear motion shown below is a combination of two translational motions, one horizontal motion
and one vertical motion.
 Rotation About a Fixed Axis - If all the particles of a rigid body move along circular paths,
except the ones which lie on the axis of rotation, it is said to be in rotation about a fixed axis.

General Motion - Any motion of a rigid body that consists of the combination of both
translations

There are six rigid body modes in general three-dimensional situation; three translational
along the x, y, and z axes and three rotational about x, y, and z axes. Illustrations of these rigid body
modes are presented as follows:
 Translational Rotational
Rigid Body Modes Rigid Body Modes

x-direction about x-axis

y-direction about y-axis

z-direction about z-axis

1-D 3-NODED QUADRATIC BAR ELEMENT

Problem 6

A single 1-D 3-noded quadratic bar element has 3 nodes with local coordinates as shown in
Figure

1 u1 2 u2 3 u3 Note that node 2 is at the midpoint of

the element.
l
x0 x xl
2

The chosen approximation function for the field variable u is u a bx cx 2

Let the field variable u have values u1 , u2 and u3 at nodes 1, 2 and 3, respectively.

To find the unknowns a, b and c, we apply the boundary conditions

at x 0, u u1 u1 a a u1
u3 4u2 3u1
l l l2 b
at x , u u2 u2 a b +c l
2 2 4 solving
2
at x l , u u3 2
u1 a bl cl c 2 u3 2u2 u1
l

Substituting the values of a, b and c in equation (1) and collecting the coefficients of u1 , u2 and u3

u N1u1 N 2u2 N 3u3

 x x2 x 2x N1 3 x
N1 1 3 2 2 1 1 4 2
l l l l x l l
x x2 x x N 2 4 x
Where N2 4 4 2 4 1 8 2
l l l l x l l
x x2 x 2x N 3 1 x
N3 2 2 1 4 2
l l l l x l l

Derivation of stiffness matrix for 1-D 3-noded quadratic bar element:

3 x
l 4 l 2

4 x
B 8 2
T

l l
1 x
l 4 l2

D E for a bar element (1-D case - only axial stress x and strain x exist x E x )
l l

volume
dV Adx A dx
0 0
since the cross-sectional area A is constant for the total length of the bar.

3 x

l 4
l2
l
4 x 3
k A 8 2 E 4 2 4
x x 1 x
8 2 4 2 dx
0
l l l l l l l l
1 x
l 4 l2

3 x 3 x 3 x 4 x 3 x 1 x
l 4 l 2 l 4 l 2 4 2 8 2 4 2 4 2
l l l l l l l l

4 x
k AE 8 2
x 4 x 4 x 4 x 1
l
x 3
4 2 8 2 8 2 8 2 4 2 dx
0
l l l l l l l l l l l l
1 x 3 x 1 x 4 x 1 x 1 x
4 2 4 2 4 2 8 2 4 2 4 2
l l l l l l l l l l l l

k B D B dV
T

Volume

N1 N 2 N 3 3 x 4 x 1 x
B 4 2 8 2 4 2
x x x l l l l l l
To determine K 11 :

3 4 x 3 4 x
l l
9 12 x 12 x 16 x 2
K11 AE 2 2 dx AE 2 3 3 4 dx
0
l l l l 0
l l l l

Integrating and applying limit we get,

l
9 24 x 16 x 2
l
9 x 24 x 2 16 x 3 9l 24l 2 16l 3
K11 AE 2 3 4 dx AE 2 4 AE 2 3 4
0
l l l l 2l 3 3l 0 l 2l 3l

9 12 16 27 36 16 AE
K11 AE AE
l l 3l 3l 3l 7

7 AE
K11
3l

To determine K 12 and K 21 :
l
3 4 x 4 8 x
l
12 24 x 16 x 32 x 2
K12 AE 2 2 dx AE 2 3 3 4 dx
0
l l l l 0
l l l l

l
12 x 40 x 2 32 x 3
K12 AE 2 4
l 2l 3 3l 0

l
12 x 40 x 2 32 x 3 12l 40l 2 32l 3 12 20 32 36 60 32
K12 AE 2 4
AE 2 3 4 AE AE
l 3l
3
l 2l 3l 0 l 2l 3l l 3l
AE
K12 8 K 21
3l

To determine K 13 and K 31 :
l
3 4 x 1 4 x
l
3 12 x 4 x 16 x 2
K13 AE 2 2 dx AE 2 3 3 4 dx
0
l l l l 0
l l l l

l
3 16 x 16 x 2
l
3 x 16 x 2 16 x 3 3l 16l 2 16l 3 3 8 16
K13 AE 2 3 4 dx AE 2 3 4 AE 2 3 4 AE
0
l l l l 2l 3l 0 l 2l 3l l l 3l
9 24 16 AE
K13 AE
3l 3l 1 K 31
To determine K 22

l
4 8 x 4 8 x
l
16 32 x 32 x 64 x 2
K 22 AE 2 2 dx AE 2 3 3 4 dx
0
l l l l 0
l l l l

l
16 64 x 64 x 2
l
16 x 64 x 2 64 x 3
K 22 AE 2 3 4 dx AE 2 3
4
0 l l l l 2l 3l 0

16l 64l 2 64l 3 16 32 64 48 96 64 AE

K 22 AE 2 3 4 AE AE
l 3l 3l 16
l 2l 3l l 3l

16 AE
K 22
3l

To determine K 23 and K 32
l
4 24 x 32 x 2
l
4 x 24 x 2 32 x 3
K 23 AE 2 3 4 dx AE 2 4
0
l l l l 2l 3 3l 0

4l 24l 2 32l 3 4 12 32 12 36 32 AE
K 23 AE 2 3 4 AE AE
l l 3l 3l 8
l 2l 3l 3l

8 AE
K 23 K 32
3l

To determine K 33

l
1 4 x 1 4 x
l
1 4 x 4 x 16 x 2
K33 AE 2 2 dx AE 2 3 3 4 dx
0
l l l l 0
l l l l
l
4 8 x 1 4 x
l
4 16 x 8 x 32 x 2
K 23 AE 2 2 dx AE 2 3 3 4 dx
0
l l l l 0
l l l l
l
1 8 x 16 x 2
l
x 8 x 2 16 x 3
K 33 AE 2 3 4 dx AE 2 3 4
0
l l l l 2l 3l 0

l 8l 2 16l 3 1 4 16 3 12 16 AE
K33 AE 2 3 4 AE AE
l l 3l 3l 7
l 2l 3l 3l

7 AE
K33
3l
 7 8 1
AE
Assembling, we get k 8 16 8
3L
1 8 7

1.7 PRINCIPLE OF STATIONERY TOTAL POTENTIAL (PSTP)

1.7.1 Potential energy in elastic bodies
Potential energy is the capacity to do the work by the force acting on deformable bodies; the
forces acting on a body may be classified as external forces and internal forces. External forces are the
applied loads while internal force is the stresses developed in the body. Hence the total potential
energy is the sum of internal and external potential energy.
Consider a spring mass system let its stiffness be k and length L, due to a force P let it extend
by u
The load P moves down by distance u. hence it loses its capacity to do work by P u. the
external potential energy in this case is given by.
H = -P u

Average force =

The energy stored in the spring due to strain = Average force x Deflection

= xu

= K u2

Total potential energy in the spring = K u2 - P u

1.7.2 Principle of Minimum Potential Energy

From the expression for total potential energy,
= U+H

= +
In principle of virtual work =

=0
Hence we can conclude that a deformable body is in equilibrium when the potential energy is
having stationary value.
 Hence the principle of minimum potential energy states among all the displacement equations
that internal compatibility and the boundary condition those that also satisfy the equation of
equilibrium make the potential energy a minimum is a stable system

Problem 7

k1 k2 P k
3
x
1 2 3 4

Given: For the spring system shown above,

k1 100 N / mm,
k 2 200 N / mm,
k 3 100 N / mm
P 500 N, A
u1 0 p
u4 0 p
l
y
Find: (a) The global stiffness matrix
(b) Displacements of nodes 2 and 3 i
(c) The reaction forces at nodes 1 and 4 n
(d) the force in the spring 2 g
Solution:
t
(a) The element stiffness matrices are
h
e
100 100
k 1 (N/mm) (1)
100 100

200 200
k 2 (N/mm) (2)
200 200

100 100
k 3 (N/mm) (3)
100 100
 u1 u2 u3 u4
100 100 0 0

100 100 200 200 0

K
0 200 200 100 100

200 300 100

0 100 100

which is symmetric and banded.

Equilibrium (FE) equation for the whole system is

100 100 0 0 u1 F1

100 300 200 0 u2 0

(4)
0 200 300 100 u3 P

0 0 100 100 u4 F4

(b) Applying the BC

300 200 u2 0
(5)
200 300 u3 P

Solving Eq.(5), we obtain

2
u2 P / 250 (6)
(mm)
3
u3 3P / 500

st th
(c) From the 1 and 4 equations in (4), we get the reaction forces

F1 100u 2 200 (N)

F4 100u 3 300 (N)

(d) The FE equation for spring (element) 2 is

200 200 ui fi

200 200 uj fj

Here i = 2, j = 3 for element 2. Thus we can calculate the spring force as

Problem 8
4 k4
F1

k1 1
2
4 1
k2 k3 F2

For the spring system with arbitrarily numbered nodes and elements, as
shown above, find the global stiffness matrix.

Solution:
First we construct the following
Element Connectivity Table

Element Node i (1) Node j (2)

1 4 2
2 2 3
3 3 5
4 2 1

Which specifies the global node numbers corresponding to the local node numbers for
each element? Then we can write the element stiffness matrices as follows
 u4 u2 u2 u3

k1 k1 k2 k2
k k 2
k1 k1 k2 k2

u3 u5 u2 u1

k3 k3 k4 k4
k 3 k 4
k3 k3 k4 k4

Finally, applying the superposition method, we obtain the global stiffness matrix as
follows

We may note that N1 and N2 obey the definition of shape function that is the shape
function will have a value equal to unity at the node to which it belong and zero value at other
nodes.

u1 u2 u3 u4 u5
k4 k4 0 0 0
k4 k1 k2 k4 k2 k1 0
K 0 k2 k2 k3 0 k3
0 k1 0 k1 0
0 0 k3 0 k3
1.8 RAYLEIGH RITZ METHOD (VARIATIONAL APPROACH)
It is useful for solving complex structural problems. This method is possible
only if a suitable functional is available. Otherwise, Galerkins method of weighted
residual is used.

Problems (I set)
1. A simply supported beam subjected to uniformly distributed load over entire
span. Determine the bending moment and deflection at midspan by using Rayleigh
Ritz method and compare with exact solutions.

2. A bar of uniform cross section is clamed at one end and left free at another end
and it is subjected to a uniform axial load P. Calculate the displacement and stress in
a bar by using two terms polynomial and three terms polynomial. Compare with
exact solutions.
1.9 ADVANTAGES OF FINITE ELEMENT METHOD
1. FEM can handle irregular geometry in a convenient manner.
2. Handles general load conditions without difficulty
3. Non homogeneous materials can be handled easily.
4. Higher order elements may be implemented.

1.10 DISADVANTAGES OF FINITE ELEMENT METHOD

1. It requires a digital computer and fairly extensive
2. It requires longer execution time compared with FEM.
3. Output result will vary considerably.
ME6603 Finite Element Analysis

UNIT II
ONE DIMENSIONAL FINITE ELEMENT ANALYSIS
2.1 ONE DIMENSIONAL ELEMENTS

Bar and beam elements are considered as One Dimensional elements. These
elements are often used to model trusses and frame structures.

Bar, Beam and Truss

Bar is a member which resists only axial loads. A beam can resist axial,
lateral and twisting loads. A truss is an assemblage of bars with pin joints and a frame
is an assemblage of beam elements.

Stress, Strain and Displacement

Stress is denoted in the form of vector by the variable x as x, Strain is denoted
in the form of vector by the variable x as e x, Displacement is denoted in the form of
vector by the variable x as ux.

Types of Loading

(1) Body force (f)

It is a distributed force acting on every elemental volume of the body. Unit is

Force / Unit volume. Ex: Self weight due to gravity.

(2) Traction (T)

It is a distributed force acting on the surface of the body. Unit is Force / Unit
area. But for one dimensional problem, unit is Force / Unit length. Ex: Frictional
resistance, viscous drag and Surface shear.

(3) Point load (P)

It is a force acting at a particular point which causes displacement.

Finite Element Modeling

It has two processes.

(1) Discretization of structure

(2) Numbering of nodes.



CO ORDINATES
(A) Global co ordinates,
(B) Local co ordinates and
(C) Natural co ordinates.
Natural Co Ordinate ()
Integration of polynomial terms in natural co ordinates for two dimensional
elements can be performed by using the formula,

Shape function
N1N2N3 are usually denoted as shape function. In one dimensional
problem, the displacement
u = Ni ui =N1 u1
For two noded bar element, the displacement at any point within the
element is given by,
u = Ni ui =N1 u1 + N2 u2
For three noded triangular element, the displacement at any point
within the element is given by,

u = Ni ui =N1 u1 + N2 u2 + N3 u3
v = Ni vi =N1 v1 + N2 v2 + N3 v3
Shape function need to satisfy the following
(a) First derivatives should be finite within an element; (b) Displacement should
be continuous across the element boundary

Polynomial Shape function
Polynomials are used as shape function due to the following reasons, (1)
Differentiation and integration of polynomials are quite easy.
(2) It is easy to formulate and computerize the finite element equations.
(3) The accuracy of the results can be improved by increasing the order of
 Properties of Stiffness Matrix

1. It is a symmetric matrix,
2. The sum of elements in any column must be equal to zero,
3. It is an unstable element. So the determinant is equal to zero.
Problem (I set)
1. A two noded truss element is shown in figure. The nodal displacements are
u1 = 5 mm and u2 = 8 mm. Calculate the displacement at x = , 1/3 and .

Problem (II set)

1. Consider a three bar truss as shown in figure. It is given that E = 2 x 10 5
N/mm2. Calculate
(a) Nodal displacement,
(b) Stress in each member and
(c) Reactions at the support. Take Area of element 1 = 2000 mm2, Area of
element 2 = 2500 mm2, Area of element 3 = 2500 mm2.

Types of beam
1. Cantilever beam,
2. Simply Supported beam,
3. Over hanging beam,
4. Fixed beam and
5. Continuous beam.

Types of Transverse Load

1. Point or Concentrated Load,
2. Uniformly Distributed Load and
3. Uniformly
 Problem (III set)
1. A fixed beam of length 2L m carries a uniformly distributed load of w (N/m)
which runs over a length of L m from the fixed end. Calculate the rotation at Point
B.

2.2 LINEAR STATIC ANALYSIS( BAR ELEMENT)

Most structural analysis problems can be treated as linear static problems, based on the
following assumptions

1. Small deformations (loading pattern is not changed due to the deformed

shape)
2. Elastic materials (no plasticity or failures)
3. Static loads (the load is applied to the structure in a slow or steady fashion)
Linear analysis can provide most of the information about the behavior of a
structure, and can be a good approximation for many analyses. It is also the bases of
nonlinear analysis in most of the cases.

2.3 BEAM ELEMENT

A beam element is defined as a long, slender member (one dimension is much larger
than the other two) that is subjected to vertical loads and moments, which produce vertical
displacements and rotations. The degrees of freedom for a beam element are a vertical
displacement and a rotation at each node, as opposed to only an horizontal displacement at
each node for a truss element.

Degrees of Freedom

Degrees of freedom are defined as the number of independent coordinates necessary

to specify the configuration of a system. The degrees of freedom for a general situation
consists of three translations in the x, y, and z directions and three rotations about the x, y, and
z axes. A one-dimensional beam element has four degrees of freedom, which include, a
vertical displacement and a rotation at each node.
Assumptions

Nodal Forces and Moments

Forces and moments can only be applied at the nodes of the beam element, not
between the nodes. The nodal forces and moments, , are related to the nodal displacements
and rotations, through the element stiffness matrix, .

Constant Load
The loads that are applied to the beam element are assumed to be static and not to
vary over the time period being considered, this assumption is only valid if the rate of change
of the force is much less than the applied force (F >> dF/dt). If the loads vary significantly,
(if the variation in load is not much less than the applied force) then the problem must be
considered as dynamic.

Weightless Member
The weight (W) of the beam is neglected, if it is much less than the total resultant
forces (F) acting on the beam. If the weight of the beam is not neglected, then its effects must
be represented as vertical forces acting at the nodes, by dividing up the weight and lumping it
at the nodes, proportionally according to it's placement along the beam.

Prismatic Member
The beam element is assumed to have a constant cross-section, which means that the
cross-sectional area and the moment of inertia will both be constant (i.e., the beam element is
a prismatic member). If a beam is stepped, then it must be divided up into sections of
constant cross-section, in order to obtain an exact solution. If a beam is tapered, then the
beam can be approximated by using many small beam elements, each having the same cross-
section as the middle of the tapered length it is approximating. The more sections that are
used to approximate a tapered beam, the more accurate the solution will be.

The moment of inertia is a geometric property of a beam element, which describes the
beams resistance to bending and is assumed to be constant through the length of the element.
The moment of inertia can be different along different axes if the beam element is not
symmetric, we use the moment of inertia (I) of the axis about which the bending of the beam
occurs
Where (Iz) refers to the moment of inertia, resisting bending about the "z" axis and (Iy) about
the "y" axis.

The Beam Element is a Slender Member

A beam is assumed to be a slender member, when it's length (L) is more than 5 times
as long as either of it's cross-sectional dimensions (d) resulting in (d/L<.2). A beam must be
slender, in order for the beam equations to apply, that were used to derive our FEM
equations.

The Beam Bends without Twisting.

It is assumed that the cross-section of the beam is symmetric about the plane of
bending (x-y plane in this case) and will undergo symmetric bending (where no twisting of
the beam occurs during the bending process). If the beam is not symmetric about this plane,
then the beam will twist during bending and the situation will no longer be one-dimensional
and must be approached as an unsymmetric bending problem (where the beam twists while
bending) in order to obtain a correct solution.

Cross Section Remains Plane

When a beam element bends, it is assumed that it will deflect uniformly, thus the
cross section will move uniformly and remain plane to the beam centerline. In other words,
plane sections remain plane and normal to the x axis before and after bending.

Axially Rigid
The one-dimensional beam element is assumed to be axially rigid, meaning that there
will be no axial displacement (u) along the beams centriodal axis. This implies that forces
will only be applied perpendicular to the beams centriodal axis. The one-dimensional beam
element can be used only when the degrees of freedom are limited to vertical displacements
(perpendicular to the beams centriodal axis) and rotations in one plane. If axial displacements
are present then a one-dimensional bar element must be superimposed with the one-
dimensional beam element in order to obtain a valid solution.

Homogenous Material
A beam element has the same material composition throughout and therefore the same
mechanical properties at every position in the material. Therefore, the modulus of elasticity E
is constant throughout the beam element. A member in which the material properties varies
from one point to the next in the member is called inhomogenous (non-homogenous). If a
beam is composed of different types of materials, then it must be divide up into elements that
are each of a single homogeneous material, otherwise the solution will not be exact.
Isotropic Material
A beam element has the same mechanical and physical properties in all directions,
i.e., they are independent of direction. For instance, cutting out three tensile test specimens,
one in the x-direction, one in the y-direction and the other oriented 45 degrees in the x-y
plane, a tension test on each specimen, will result in the same value for the modulus of
elasticity (E), yield strength and ultimate strength . Most metals are considered
isotropic. In contrast fibrous materials, such as wood, typically have properties that are
directionaly dependant and are generally considered anisotropic (not isotropic).

The Proportional Limit is not Exceeded

It is assumed that the beam element is initially straight and unstressed. It is also
assumed that the material does not yield, therefore the beam will be straight after the load is
released. These assumptions mean that the beam must be made of an elastic material, one
which will return to it's original size and shape when all loads are removed, if not stressed
past the materials elastic or proportional limit. It is also assumed that the beam is not stressed
past the proportional limit, at which point the beam will take a permanent set and will not
fully return to it's original size and shape, when all loads are removed. Below the proportional
limit an elastic material is in the linear elastic range, where the strain ( ) varies linearly with
the applied load and the stress ( ) varies linearly according to: , where E is the
modulus of elasticity.

Rigid Body Modes for the One-Dimensional Beam Element

Rigid body motion occurs when forces and/or moments are applied to an unrestrained
mesh (body), resulting in motion that occurs without any deformations in the entire mesh
(body). Since no strains (deformations) occur during rigid body motion, there can be no
stresses developed in the mesh. In order to obtain a unique FEM solution, rigid body motion
must be constrained. If rigid body motion is not constrained, then a singular system of
equations will result, since the determinate of the mesh stiffness matrix is equal to zero (i.e.,
).

There are two rigid body modes for the one-dimensional beam element, a translation
(displacement) only and a rotation only. These two rigid body modes can occur at the same
time resulting in a displacement and a rotation simultaneously. In order to eliminate rigid
body motion in a 1-D beam element (body), one must prescribe at least two nodal degrees of
freedom (DOF), either two displacements or a displacement and a rotation. A DOF can be
equal to zero or a non-zero known value, as long as the element is restrained from rigid body
motion (deformation can take place when forces and moments are applied) .

For simplicity we will introduce the rigid body modes using a mesh composed of a
single element. If only translational rigid body motion occurs, then the displacement at local
node I will be equal to the displacement at local node J. Since the displacements are equal
there is no strain developed in the element and the applied nodal forces cause the element to
move in a rigid (non-deflected) vertical motion (which can be either up as shown below or it
can be in the downward direction depending on the direction of the applied forces).
 This rigid body mode can be suppressed by prescribing a vertical nodal displacement.

If rotational rigid body motion occurs, then the rotation at local node I will be equal to
the rotation at local node J (i.e., in magnitude and direction). In this situation the nodal forces
and/or moments applied to the element, cause the element to rotate as a rigid body (either
clockwise as shown below or counterclockwise depending on the direction of the applied
forces and/or moments).

This rigid body mode can be suppressed by prescribing a nodal translation or rotation.

If translational and rotational rigid body motion occurs simultaneously then:

Simple Examples of Beam Problems

with and without Rigid Body Motion
Determinant
Stable/Unsta Rigid Body
of Mesh
Case ble Mode(s) Equations
Stiffness
Structure Present
Matrix

Dependent
Unstable and
Equations

Dependent
Unstable
Equations
 Dependent
Unstable
Equations

Independen
Stable None
t Equations

Independen
Stable None
t Equations

2.4 1-D 2-NODED CUBIC BEAM ELEMENT MATRICES

A single 1-d 2-noded cubic beam element has two nodes, with two degrees of
freedom at each node (one vertical displacement and one rotation or slope). There is a total of
4 dof and the displacement polynomial function assumed should have 4 terms, so we choose
a cubic polynomial for the vertical deflection. Slope is a derivative of the vertical
deflections.

The vertical displacement v a bx cx 2 dx 3 ..(1)

dv
The slope b 2cx 3dx 2 ..(2)
dx

Apply the boundary conditions

at x 0, v v1 v1 a a v1
at x 0, 1 1 b b 1
3 1
2 2
c v v1 21 2
at x l , v v2 v2 a bl cl 2 +dl 3 l l
solving
at x l , 2 2 b 2cl 3dl 2 2 1
d 3 v1 v2 2 1 2
l l

Substituting the values of a, b, c and d in equation (1), and collecting the coefficients of
v1 , 1 , v2 , 2 we obtain

v N1v1 N 21 N 3v3 N 4 2

where
 x2 x3 x 2 x3
N1 1 3 2 , N2 x 2 ,
l2 l3 l l2
x2 x3 x 2 x3
N3 3 2 2 3 , N4 2
l l l l

2.5 DEVELOPMENT OF ELEMENT EQUATION

dx R

x
R y R
y d2y
y 2
R R dx

v N1v1 N21 N3v2 N42

d 2v d2
x y y N1v1 N 21 N3v2 N 4 2
dx 2 dx 2

v1

d N1
2 2
d N2 2
d N3 d N 4 1
2
x y
dx
2
dx 2 dx 2 dx 2 v2

2

B a

x B a

We Know that,
 K B D B dv
T

d 2 N1
2
dx
d 2 N2
2 d 2N d 2 N2 d 2 N3 d 2 N4
K y dx2 E y 2 1 dv
volume d N3 dx dx 2 dx 2 dx 2
dx 2
2
d N4

dx 2

d 2 N 2 d 2 N1 d 2 N 2 d 2 N1 d 2 N 3 d 2 N1 d 2 N 4
1
2 2 2 2

dx dx dx
2 2 2
dx dx dx dx
2 2
d N 2 d N1
2
d 2 N2 d 2 N 2 d 2 N3 d 2 N2 d 2 N4
dx 2 dx 2 2 2 2 2 2
2 dx dx dx dx dx
K E 0 y 2
l
dAdx
d 2 N3 d 2 N 4
2
2
d 2 N3 d 2 N 2 d 2 N3
d N2 3 d N2 1
v
2 2 2
dx dx 2
dx dx dx dx 2
dx
2
d 2 N 4 d 2 N1 d 2 N 4 d 2 N 2 d 2 N 4 d 2 N3 d 2 N4
dx 2 dx 2 2 2 2 2 2
dx dx dx dx dx
l
Where, 0
y 2 dA I

d 2 N1
2
dx
d 2 N2
v dx 2 d N d 2 N3 d 2 N4
2
d 2 N2

K EI 0 d 2 N dx 2 1
dx
3
dx 2 dx 2 dx 2
dx 2
2
d N4

dx 2

Where,

3x 2 2 x3 dN1 6 x 6 x 2 d 2 N1 6 12 x
N1 1 2 3 2 3 2 3
l l dx l l dx 2 l l

2 x 2 x3 dN 2 4 x 3x 2 d 2 N 2 4 6 x
N2 x 2 1 2 2
l l dx l l dx 2 l l
 3x 2 3x3 dN 3 6 x 6 x 2 d 2 N 3 6 12 x
N3 2 3
l2 l3 dx l2 l3 dx 2 l l

x3 x 2 dN 4 3 x 2 x d 2 N4 6x 2
N4 2
l2 l dx l2 l dx 2 l l

6 12 x
l2 l3

4 6 x
2
l 6 12 x 4 6 x 6 12 x 6 x 2
K EI 0
l l
2 l 3 l l 2 l 2 l 3 l 2 l dx
6 12 x l
2 3
l l
6x 2
2
l l

l 6 12 x 6 12 x
K11 EI 2 3 2 3 dx
0
l l l l

l 6l 12 x 6l 12 x
K11 EI dx
0
l3 l3

l 36l 72 xl 72 xl 144 x
2 2
K11 EI dx
0
l6

l 36l 144 xl 144 x 2

2
K11 EI 6 6 6 dx
0
l l l

l
36 xl 2 144 x 2l 144 x3
K11 EI 6
l 2l 6 3l 6 o

36 72 48
K11 EI 3 3 3
l l l

12EI
K11
l3

l 6 12 x 4 6 x
K12 EI 2 3 2 dx
0
l l l l
 l 6l 12 x 4l 6 x
K12 EI dx
l3 l
0 2

l 24l 48 xl 36 xl 72 x
2 2
K12 EI dx
0
l5

l 24l 84 xl 72 x 2
2
K12 EI 5 5 5 dx
0
l l l

l
l 24 xl 84 x 2l 72 x3
2
K12 EI 5 5
0
l 2l 5 3l 0

24 42 24
K12 EI 2 2 2
l l l

6EI
K12
l2

l 6 12 x 6 12 x
K13 EI 2 3 2 3 dx
0
l l l l

l 6l 12 x 6l 12 x
K13 EI dx
l3 l
0 3

l 36l 72 xl 72 xl 144 x
2 2
K13 EI dx
0
l6
l
l 36 xl 144 x 2l 144 x3
2
K13 EI 6
0
l 2l 6 3l 6 0

36 72 48
K13 EI
l3

12EI
K 13
l3

l 6 12 x 6 x 2
K14 EI 2 3 2 dx
0
l l l l

l 6l 12 x 6 x 2l
K14 EI dx
l3 l
0 2
 l 12l 36 xl 24 xl 72 x
2 2
K14 EI dx
0
l5

l 12l 60 xl 72 x 2
2
K14 EI 5 5 5 dx
0
l l l

l
12 xl 2 60 x 2l 72 x3
l
K14 EI 5 5
0
l 2l 5 3l 0

30 12 24
K14 EI
l2

6EI
K14
l2

l 4 6 x 6 12 x
K 21 EI 2 2 3 dx
0
l l l l

6EI
K 21
l2

l 4 6 x 4 6 x
K 22 EI 2 2 dx
0
l l l l

l 4l 6 x 4l 6 x
K 22 EI dx
l l
0 2 2

l 16l 2 24 xl 24 xl 36 x 2
K 22 EI dx
0
l4

l 16l 48 xl 36 x 2
2
K 22 EI 4 4 4 dx
0
l l l

l
l16 xl 2 48 x 2l 36 x3
K 22 EI 4 4
0
l 2l 4 3l 0

16 24 12
K 22 EI
l

4EI
K 22
l
 l 4 6 x 6 12 x
K 23 EI 2 2 3 dx
0
l l l l

l 4l 6 x 6l 12 x
K 23 EI dx
l l
0 2 3

l 24l 36 xl 48 xl 72 x
2 2
K 23 EI dx
0
l5

l 24l 84 xl 72 x
2 2
K 23 EI dx
0
l5
l
l 24 xl 2 84 x 2l 72 x3
K 23 EI 5 5
0
l 2l 5 3l 0

24 42 24
K 23 EI
l2

6EI
K 23
l2

l 4 6x 6x 2
K 24 EI 2 2 dx
0
l l l l

l 4l 6 x 6 x 2l
K 24 EI dx
l l
0 2 2

l 8l 24 xl 12 xl 36 x
2 2
K 24 EI dx
0
l4

l 8l 36 xl 36 x
2 2
K 24 EI dx
0
l4
l
l 8 xl 2 36 x 2l 36 x3
K 24 EI 4 4
0
l 2l 4 3l 0

18 12 8
K 24 EI
l

2EI
K 24
l
 l 6 12 x 6 12 x
K 31 EI 2 3 2 3 dx

0 l l l l

12EI
K 31
l3

l 6 12 x 4 6 x
K 32 EI 2 3 2 dx

0 l l l l

6EI
K 32
l2

l 6 12 x 6 12 x
K 33 EI 2 3 2 3 dx

0 l l l l

l 6l 12 x 6l 12 x
K 33 EI dx
l l
0 3 3

l 36l 72 xl 72 xl 144 x
2 2
K 33 EI dx
0
l6

l 36l 144 xl 144 x

2 2
K 33 EI dx
0
l6
l
36 xl 2 144 x 2l 144 x3
l
K33 EI 6
0
l 2l 6 3l 6 0

36 72 48
K 33 EI
l3

12EI
K 33
l3

l 6 12 x 6 x 2
K 34 EI 2 3 2 dx

0 l l l l

l 6l 12 x 6 x 2l
K 34 EI dx
l l
0 3 2

l 12l 24 xl 36 xl 72 x
2 2
K34 EI dx
0
l5
 l 12l 60 xl 72 x
2 2
K34 EI dx
0
l5
l
12 xl 2 60 x 2l 72 x3
l
K34 EI 5 5
0
l 2l 5 3l 0

12 30 24
K 34 EI
l2

6EI
K 34
l2

l 6x 2 6 12 x
K 41 EI 2 2 3 dx
0
l l l l

6EI
K 41
l2

l 6x 2 4 6 x
K 42 EI 2 2 dx
0
l l l l

2EI
K 42
l

l 6x 2 6 12 x
K 43 EI 2 2 3 dx
0
l l l l

6EI
K 43
l2

l 6x 2 6x 2
K 44 EI 2 2 dx
0
l l l l

l 6 x 2l 6 x 2l
K 44 EI dx
l l
0 2 2

l 4l 12 xl 12 xl 36 x
2 2
K 44 EI dx
0
l4

l 4l 24 xl 36 x
2 2
K 44 EI dx
0
l4
 l
l 4 xl 2 24 x 2l 36 x3
K 44 EI 4 4
0
l 2l 4 3l 0

12 4 12
K 44 EI
l

4EI
K 44
l

Therefore K is

12 6 12 6
l3 l2 l3 l2

6 4 6 2
2 l2 l
K EI l l

12 6 12 6
l3 l2 l3 l2
6 2 6 4
2
l l l2 l

2.6 BEAM ELEMENT

A beam is a long, slender structural member generally subjected to transverse loading

that produces significant bending effects as opposed to twisting or axial effects. An elemental
length of a long beam subjected to arbitrary loading is considered for analysis. For this
elemental beam length L, we assign two points of interest, i.e., the ends of the beam, which
become the nodes of the beam element. The bending deformation is measured as a transverse
(vertical) displacement and a rotation (slope). Hence, for each node, we have a vertical
displacement and a rotation (slope) two degrees of freedom at each node. For a single 2-
noded beam element, we have a total of 4 degrees of freedom. The associated forces are
shear force and bending moment at each node.

1 M1
2 M2
v1 v2 F1 F2
Nodal displacements Nodal forces

1st
vertical
degree vi or v1 shear force at node Fi or F1
displacement at 1 1
of i
node i corres-
freedom
2nd pond-
slope or rotation at bending moment at
degree of 2 i or 1 ing to M i or M 1 2
node i node i
freedom
3rd vertical 3 v j or v2 shear force at node F j or F2 3
degree of displacement at i
freedom node j
4th
slope or rotation at bending moment at
degree of 4 j or 2 M j or M 2 4
node j node j
freedom

The stiffness term kij indicates the force (or moment) required at i to produce a unit
deflection (or rotation) at j, while all other degrees of freedom are kept zero.
Sign conventions followed
Upward forces are positive and upward displacements are positive.
Counter-clockwise moments are positive and counter-clockwise rotations are positive.
Formulae required cantilever beam subjected to concentrated load and moment.

P PL2 ML
M
2 EI EI
PL3 ML2

3EI 2 EI

2.6.1 ELEMENT MATRICES AND VECTORS

Derivation of first column of stiffness matrix: v1 1, 1 v2 2 0 , i.e., allow the first

degree of freedom to occur and arrest all other DoF. (The deformed configuration is shown in
Figure 2).

Initially you have a horizontal beam element. Since v2 2 0 , we can fix node j. To produce
an upward deflection at node i (i.e., allowing first degree of freedom to occur), apply an
upward force k11 (first suffix indicates the force or moment DoF and the second suffix
k11 L3
indicates the displacement or rotational DoF). v1 upwards. Refer table for
3EI
displacement DoF number and force DoF number. Now the beam configuration is given by
Figure 1. We can observe from the figure that the slope at node i is not zero. To make the
slope at i equal to zero, we need to apply a counter-clockwise moment k21 . Refer Figure 2.
k21 L2
But this moment k21 will produce a downward deflection at node i. Refer Figure
2 EI
3. In order to have a resultant unit upward displacement at node i, upward displacement
produced by force k11 must be greater than the downward displacement produced by the
k11 L3 k21 L2
moment k21 . i.e., 1 ..(1). At the same time, the negative slope produced at
3EI 2 EI
node i by the force k11 must be cancelled by the positive slope produced by the moment k21 .
k11 L2 k21 L
i.e., .(2). Solving these two equations, k11 and k21 are found. The fixed end
2 EI EI
 reaction force and the reaction moment are assumed to be acting upwards and
counterclockwise, respectively. Now use force equilibrium equation to find fixed end
reaction force k31 . Fy 0 k11 k31 0 and moment equilibrium equation about
node i to find fixed end reaction moment k41 .... M i 0 k21 k31 L k41 0 .

k41

k11 Figure 2. Figure 3.

k11 Figure 1. k11 k31
Figure 4.

k21
k21
1unit

k11 12 EI
L3

k21 6 EI
L2
12 EI
k31 3
L
6 EI
k
41 L2
Derivation of second column of stiffness matrix: v1 0, 1 1, v2 2 0 , i.e., allow the
second degree of freedom to occur and arrest all other DoF. (The deformed configuration is
shown in Figure 2).

Initially you have a horizontal beam element. Since v2 2 0 , we can fix node j. To produce
a counterclockwise (positive) rotation or slope at node i (i.e., allowing second degree of
k L
freedom to occur), apply a counterclockwise moment k22 . 1 22 . Refer Figure 1. This
EI
2
k L
moment k22 will produce a downward deflection 22 . This downward deflection should be
2 EI
canceled by applying an upward force k12 at node i. The upward deflection produced by k12 is
k12 L3 k L2 k L3
. Refer Figure 2. Equating these two deflections 22 12 (1) But this upward
3EI 2 EI 3EI
k12 L2
force k12 will also produce a negative slope at node i which is . Refer Figure 3. Hence
2 EI
the rotation produced by k22 should be greater than that produced by k12 so that the resultant
k22 L k12 L2
rotation is 1 radians. 1 .(2). Solving these two equations, k12 and k22 are
EI 2 EI
found. The fixed end reaction force and the reaction moment are assumed to be acting
upwards and counterclockwise, respectively. Now use force equilibrium equation to find
fixed end reaction force k32 Fy 0 k12 k32 0 and moment equilibrium
equation about node i to find fixed end reaction moment k42 ....
M i 0 k 22 k32 L k 42 0 .

k22 k12 6 EI
k22 L2
k42
k22 4 EI
k12 k32 L
Figure 1.
Figure 4.
6 EI
k32 2
L
2 EI
k22 1 rad k
42 L

k12
k12 Figure 2. Figure 3.

Derivation of third column of stiffness matrix: v1 0, 1 0, v2 1, 2 0 , i.e., allow the

third degree of freedom to occur and arrest all other DoF. (The deformed configuration is
shown in Figure 2).
 Initially you have a horizontal beam element. Since v1 1 0 , we can fix node i. To produce
an upward deflection at node j (i.e., allowing third degree of freedom to occur), apply an
upward force k33 .

k33 L3
v2 upwards. Now the beam configuration is given by Figure 1. We can observe from
3EI
the figure that the slope at node j is not zero. To make the slope at j equal to zero, we need to
apply a clockwise moment k43 . Refer Figure 2. But this moment k43 will produce a downward
k43 L2
deflection at node j. Refer Figure 3. In order to have a resultant unit upward
2 EI
displacement at node j, upward displacement produced by force k33 must be greater than the
k33 L3 k43 L2
downward displacement produced by the moment k43 . i.e., 1 ..(1). At the
3EI 2 EI
same time, the positive slope produced at node j by the force k33 must be cancelled by the
k33 L2 k43 L
negative slope produced by the moment k43 . i.e., .(2). Solving these two
2 EI EI
equations, k33 and k43 are found. The fixed end reaction force and the reaction moment are
assumed to be acting upwards and counterclockwise, respectively. Now use force equilibrium
equation to find fixed end reaction force k13 Fy 0 k13 k33 0 and moment
equilibrium equation about node i to find fixed end reaction moment k23 ....
M i 0 k 23 k33 L k 43 0 .
k43
k23
k13 12 EI
L3
k13
Figure 4.
k33
k33 k23 6 EI
Figure 1. L2
12 EI
k43 k33
L
3

k43 6 EI
1unit k 2
43 L

Figure 2.
k33
Figure 3.

Derivation of fourth column of stiffness matrix: v1 1 0, v2 0, 2 1 , i.e., allow the

fourth degree of freedom to occur and arrest all other DoF. (The deformed configuration is
shown in Figure 2).

Initially you have a horizontal beam element. Since v1 1 0 , we can fix node i. To produce a
counterclockwise (positive) rotation or slope at node j (i.e., allowing fourth degree of freedom
k L
to occur), apply a counterclockwise moment k44 . 2 44 . Refer Figure 1. This moment k44 will
EI
 k44 L2
produce a upward deflection . This upward deflection should be canceled by applying a
2 EI
k L3
downward force k34 at node j. The downward deflection produced by k34 is 34 . Refer Figure
3EI
2 3
k L k L
2. Equating these two deflections 44 34 (1) But this downward force k34 will also
2 EI 3EI
k L2
produce a negative slope at node j which is 34 . Hence the rotation produced by k44 should be
2 EI
k L k L2
greater than that produced by k34 so that the resultant rotation is 1 radians. 44 34 1 .(2)
EI 2 EI
Refer Figure 3. Solving these two equations, k34 and k44 are found. The fixed end reaction force
and the reaction moment are assumed to be acting upwards and counterclockwise, respectively.
Now use force equilibrium equation to find fixed end reaction force k14
F y 0 k14 k34 0 and moment equilibrium equation about node i to find fixed end
reaction moment k24 .... M i 0 k24 k34 L k44 0 .

k24 k44 k14 6 EI

L2

k44 k24 2 EI
Figure 1. k14 k34 L
Figure 4.
6 EI
k34 2
k34 k34 L
4 EI
k
44 L
1 k44

Figure 3.
Figure 2.

Problem
Find the slopes at the supports and support reaction forces and support reaction moments for the
beam shown in Figure. Take E=210 GPa, I = 210-4 m4. Daryl Logan P4-24 page 208.

kN
5
m

5m 4m
2
Finite element representation of the problem
v2
Conversion of UDL into nodal forces and nodal moments

qL qL
Force kN kN
kN 2 2
q

Lm qL2 qL2
Moment kN m kN m
12 12

for element 1,

12.5 kN 12.5 kN
5 m kN
5 10.416667 kN m 10.416667 kN m

m

for element 2,
10 kN 10 kN
kN
5

4m
6.66667 kN m 6.66667 kN m
4 kN 4
EI 210 GPa 2 10 m 210 10 2 2 10 m 42000 kN-m2
4 6 4

Stiffness matrix for element 1

4, 032 10, 080 4032 10, 080

10, 080 33, 600 10, 080 16,800
K
(1)

4032 10, 080 4, 032 10, 080

10, 080 16,800 10, 080 33, 600

Stiffness matrix for element 2

7,875 15, 750 7,875 15, 750

15, 750 42, 000 15, 750 21, 000
K
(2)

7,875 15, 750 7,875 15, 750

15, 750 21, 000 15, 750 42, 000

Assembly of finite element equations

 F1 12.5 F1 12.5 Support reaction moments at all simply
M 10.416667 0 10.416667

1 supported ends are zero. M 1 M 2 M 3 0
F2 12.5 10 F2 22.5 All support reaction forces are unknowns.

M 2 10.416667 6.66667 0 3.75
F2 10 F2 10

M 3 6.66667 0 6.66667
support applied forces
reactions

v1 v2 v3 0 1 ? 2 ? 3 ?

F1 12.5 4, 032 10, 080 4, 032 10, 080 0 0 0

0 10.416667 10, 080
33, 600 10, 080 16,800 0 0 1
F2 22.5 4, 032 10, 080 4, 032 7,875 10, 080 15, 750 7,875 15, 750 0

0 3.75 10, 080 16,800 10, 080 15, 750 33, 600 42, 000 15, 750 21, 000 2
F3 10 0 0 7,875 15, 750 7,875 15, 750 0

0 6.66667 0 0 15, 750 21, 000 15, 750 42, 000 3

Eliminating the first, third and fifth rows and columns of the stiffness matrix, the reduced matrix
becomes

10.416667 33, 600 16,800 0 1

3.75 16,800 75, 600 21, 000 2
6.66667 0 21, 000 42, 000 3

Solving these equations

1 3.59623 10 4 rad 2 9.9206349 10 5 rad 3 1.0912698 10 4 rad

Substituting these values in the assembled matrix to find the support reactions, we find

F1 12.5 10, 080 3.59623 10 10, 080 9.92 10

4 5

F 22.5 10, 080 3.59623 10 5, 670 9.92 10 15, 750 1.0913 10

2
4 5 4

F3 10 15, 750 9.92 10 15, 750 1.0913 10

5 4

F1 12.5 2.625 kN F1 9.875 kN

F2 22.5 5.9062 kN which means F2 28.406 kN
F3 12.5 3.2812 kN F3 6.71869 kN
 kN
It is verified that the total applied load 5 5m 45kN is equal to the sum of the support
m
reaction forces (9.875+28.406+6.71869 = 45 kN).

Total force and moment diagram

10.416667 kN-m 3.75 kN-m 6.66667 kN-m

9.875 kN 28.406 kN 6.71869 kN

Individual force and moment diagrams

10.416667 kN-m 2.70816 kN-m 6.45813 kN-m 6.66667 kN-m

2.624832 kN 2.624832 kN 3.2811975 kN 3.2811975 kN

Individual force and moment calculations

Element 1

F1(1) 10, 080 3.59623 104 10, 080 9.92 105 2.624832 kN

M 1(1) 33, 600 3.59623 10 16,800 9.92 10 10.416 kN-m

4 5

F2(1) 10, 080 3.59623 10 10, 080 9.92 10 2.624832 kN

4 5

M 2(1) 16,800 3.59623 10 33, 600 9.92 10 2.70816 kN-m

4 5

Element 2

F1(1) 15, 750 9.92 10 5 15, 750 1.0913 10 4 3.2811975 kN

M 1(1) 42, 000 9.92 10 5 21, 000 1.0913 10 4 6.45813 kN-m
F2(1) 15, 750 9.92 10 5 15, 750 1.0913 10 4 3.2811975 kN
M 2(1) 21, 000 9.92 10 5 42, 000 1.0913 104 6.66667 kN-m
PROBLEM

Given that E=210 GPa and I=410-4 m4, cross section of the beam is constant.
Determine the deflection and slope at point C. calculate the reaction forces and moments.
DARYL LOGAN P 171-172

1kN

20 kN-m
A C
B
3m 3m

Solution:-

Degree of freedom in numbers:-

2 4 6

1 3 5

Degree of freedom of forces and moments:-

M1 M2 M3

F1 F2 F3
Degree of freedom of displacement and rotation:-

1 2 3

v1 v2 v3
Stiffness matrix for element 1 and 2:-
 12 EI 6 EI 12 EI 6 EI
l3
l2 l3 l2

6 EI 4 EI 6 EI
2
2 EI
l2 l
K K
l l
1 2

12 EI 6 EI
2
12 EI 6 EI
2
l3 l l3 l
6 EI 2 EI 6 EI 4 EI
2 2
l l l l

12 18 12 18
36 18 18
6 18
K K
1 2

12 18 12 18

18 18 18 36
Assembling:-

F1 12 18 12 18 0 0 v1
M 18
1
36 18 18 0 0 1
F2
6 12 18 24 0 0 0 v2
3.1
M2 18 18 0 72 18 18 2
F3 0 0 12 18 12 18 v3

3
M 0 0 18 18 18 36 3

Boundary condition:-

F2=-10 kN; M2=20 kN-m v1 =v3= 1= 3=0

Therefore first, second, fifth, sixth columns are ineffective

and hence the reduced matrix is given by

F2 24 0 v2
3 .1 1 0
6
0
m2 7 2 2

Deflection and slope at point c:-

V2= -1.3410-4 m = -0.134 mm

2= 8.9610-5 rad
 Reaction forces and moments:-

F2 12 18
m 18 18
2 6
3.1 10
F
3 12 18
m3
18 18
F1=10000N

M1=12500N-m

F3=0

M3= -2500N-m
12,500N-m 17,500N-m

10,000N 10,000N

individual element forces and moments are

2,500N-m 2500N-m

0 0

10kN

12.5kN-m 20kN-m
2.5kN-m

10kN
ME6603 Finite Element Analysis

UNIT III
TWO DIMENSIONAL FINITE ELEMENT ANALYSIS
3.1 INTRODUCTION
Two dimensional elements are defined by three or more nodes in a two dimensional
plane (i.e., x, y plane). The basic element useful for two dimensional analysis is the
triangular element.

Plane Stress and Plane Strain

The 2d element is extremely important for the Plane Stress analysis and Plane
Strain analysis.
Plane Stress Analysis:
It is defined to be a state of stress in which the normal stress () and shear
stress () directed perpendicular to the plane are assumed to be zero.
Plane Strain Analysis:
It is defined to be a state of strain in which the normal to the xy plane and the
shear strain are assumed to be zero.

3.2 THREE NODED LINEAR TRIANGULAR ELEMENT

The physical domain considered is geometrically a 2-Dimensional domain, i.e., an area with
uniform thickness and the single variable can be one of pressure, temperature, etc. (a scalar
quantity, not a vector quantity). An example is the temperature distribution in a plate. At each
point there can be only one temperature. We consider such an area meshed with triangular
elements. Each triangular element has three nodes, (i.e., one node at each corner). Let us consider
one such element with coordinates x1 , y1 , x2 , y2 and x3 , y3 . The single variable (for example,
temperature) at these nodes 1, 2 and 3 are u1 , u2 and u3 , respectively. If so, then the unknown single
variable u (temperature) at any non-nodal point x, y in the 2-D domain can be expressed in terms
of the known nodal variables (temperatures) u1 , u2 and u3 .
 Let us assume that the single variable can be expressed as

u c1 c2 x c3 y

In order to find the three unknowns c1 , c2 and c3 , we apply the boundary conditions

at x1 , y1 , u u1 u c1 c2 x1 c3 y1 y
at x2 , y2 , u u2 u c1 c2 x2 c3 y2
at x3 , y3 , u u3 u c1 c2 x3 c3 y3 3 (x3,y3)
Writing the above three equations in matrix form u3
u1 1 x1 y1 c1 u2

u2 1 x2 y2 c2 u1 2 (x2,y2)
u 1 x y3 c3 1 (x1,y1)
3 3

We need to find c1 , c2 and c3 x

1
c1 1 x1 y1 u1

c2 1 x2 y2 u2
c 1 x y3
3 3 u3
1
1 x1 y1 1 2 3
1 x
y2
1
1 2 3 where 2 A 1 2 3 and
2
2A
1 x3 y3 1 2 3
i x j yk xk y j i y j yk i x j xk
1 x2 y3 x3 y2 1 y2 y3 1 x2 x3
2 x3 y1 x1 y3 2 y3 y1 2 x3 x1
3 x1 y2 x2 y1 3 y1 y2 3 x1 x2

A is the area of the triangle.

c1 1 2 3 u1
1
c2 1 2 3 u2
c 2 A 3 u3
3 1 2

Substituting the values of c1, c2 and c3 in u c1 c2 x c3 y , we get

u N1u1 N 2u2 N3u3

1
where Ni i i x i y , i 1, 2,3
2A
3.3 FOUR NODED LINEAR RECTANGULAR ELEMENT

y
4 (0,b) 3 (a,b)

x
1 (0,0) 2 (a,0)
x

Let us assume that the single variable can be expressed as

u x , y c1 c2 x c3 y c4 x y (1)

This polynomial contains four linearly independent terms and is linear in x and y, with a
bilinear term in x and y. The polynomial requires an element with four nodes. There are two
possible geometric shapes: a triangle with the fourth node at the centroid of the triangle or a
rectangle with nodes at the vertices.

A triangle with a fourth node at the center does not provide a single-valued variation of u at
inter-element boundaries, resulting in incompatible variation of u at inter-element boundaries and is
therefore not admissible.

The linear rectangular element is a compatible element because on any side, the single
variable u varies only linearly and there are two nodes to uniquely define it.

Here we consider an approximation of the form given in eqauation (1) and use a rectangular
element with sides a and b. For the sake of convenience we choose a local coordinate system
x , y to derive the interpolation functions.
In order to find the three unknowns c1 , c2 and c3 , we apply the boundary conditions
at 0, 0 , u u1 u c1
at a, 0 , u u2 u c1 c2 a
at a, b , u u3 u c1 c2 a c3ab
at 0, b , u u3 u c1 c4b
Solving for c1, c2, c3 and c4
y
u u u u u1 u2 u3 u4
c1 u1 , c1 2 1 , c3 4 1 , c4 ,
a b ab
3.4 TWO-VARIABLE 3-NODED LINEAR TRIANGULAR ELEMENT
y
v3

3 x1 , y1 u3
v1

u1 v2
1 x1 , y1
u2
2 x2 , y2
x

Figure shows a 2-D two-variable linear triangular element with three nodes and the two dof at each
node. The nodes are placed at the corners of the triangle. The two variables (dof) are displacement
in x-direction (u) and displacement in y-direction (v). Since each node has two dof, a single element
has 6 dof. The nodal displacement vector is given by

u1
v
1
u
U 2
v2
u3

v3

We select a linear displacement function for each dof as

u x, y c1 c2 x c3 y
v x, y c4 c5 x c6 y

where u x, y and v x, y describe displacements at any interior point x, y of the element.

The above two algebraic equations can also be written as

c1
c
2
u 1 x y 0 0 0 c3

v 0 0 0 1 x y c4
c5

c6
Using steps we had developed for the 2-D single-variable linear triangular element, we can write

c1 1 2 3 u1
1
c2 1 2 3 u2
c 2 A 3 u3
3 1 2

c4 1 2 3 v1
1 v
c5 1 2 3 2
c 2 A 3 v3
6 1 2

and using the interpolation functions we had developed for the 2-D single-variable linear triangular
element, we can write

u x, y N1u1 N 2u2 N 3u3

v x, y N1v1 N 2 v2 N 3v3

where

1
Ni i i x i y , i 1, 2,3
2A

u1
v
1
u x, y N1 0 N2 0 N3 0 u2

Writing the above equations in matrix form v x, y 0 N1 0 N2 0 N 3 v2 The
u3

v3
U N a
strains associated with the two-dimensional element are given by
 u

N i
x x i
v x
y and note that
N i
y i
xy u v y

y x
u1
v
u1 1
u N1 N 2 N 3 N1 N 2 N 3 u
N1u1 N 2u2 N 3u3 u2 0 0 0 2
x x x x x x x x v2
u3 u3

v3
u1
v
v1 1
v N1 N 2 N 3 N1 N 2 N 3 u2
N1v1 N 2 v2 N 3v3 v2 0 0 0
y y y y y y y y v2
3
v
u3

v3
u1
v
1
u v N1 N1 N 2 N 2 N 3 N 3 u2
N1u1 N 2u2 N 3u3 N1v1 N 2 v2 N 3v3
y x y x y x y x y x v2
u3

v3
N1 N 2 N 3 u1 u1
0 0 0 v v
x x x x 1 1 0 2 0 3 0 1
N1 N 2 N 3 u2 u
y 0 0 0 0 1 0 2 0 3 2
y y y v2 v
xy N 1 1 2 2 3 3 2
N1 N 2 N 2 N 3 N 3 u3 u3

1

y x y x y x v3 v3

31 B 36 a61
 D
x x

y D y

xy xy
D B a

The stiffness matrix is given by K 66 B D B dV t B D B dxdy .

T T
63 33 36
Volume Area

where t is the thickness of the plate. The integrand B D B is not a function of x and y and
T

hence can be taken outside the integral to yield

K tA B D B
T

D matrix is the material constitutive matrix, either for the plane-stress case or for the plane-strain
case depending on the problem in hand.

and substituting them back in u, we get

u N1u1 N 2u2 N 3u3 N 4u4

x y
where N1 1 1
a b
x y
N 2 1
a b
x y
N3
ab
xy
N 4 1
ab

3.5 STRAIN STRESS RELATION

x y
x z
E E E
y
y x z
E E E
y z
z x
E E E
3.5.1 Plane stress conditions
x , y and xy are present.
z xz yz 0.
since z 0, from equations 1 and 2
x y
x
E E
y
y x
E E
solving the above two equations
for x and y , we get

x
E
x y and

1 2
y
E
x y

1 2
E
xy G xy xy
21

1 E 2

1 21
2 xy

1 1 E
xy
1 2
21
E 1
xy xy

1 2 2
writing x , y and xy in a matrix form

x
E 1 0 x
y 1 0 y
1
2
1
xy 0 0 xy
2

3.5.2 Plane strain conditions

x , y and xy are present.
z xz yz 0.
z is not zero.
since z 0, we get from equation 3
z x x
substituti ng z in equations 1 and 2

x
x
y 2 x
y
E E E

y x y 2 x
y
E E E
rearranging the terms we get

x y
x
E
E

1 2 1

y
y x 1
E E
1 2
mutiplying by X by and Y by 1 -
y 2

x x 1 2
E E

1

1 y x 1 1 y 1 1 2
E E
y

x 1 2
E E

1 1 2
adding the above two equations to eliminate x
y
x 1 y
E
2
1 1 1 2
y
x 1 y
E
2
1 1 1 1
y
x 1 y 1 2 1 1
E
x 1 y y
1

E
2
1 2 2

x 1 y
y E
1 1 2
similarly
1 x y
x E
1 1 2
andas before
E
xy xy
21
writing x , y and xy in matrixform
ME2353 Finite Element Analysis

x 1 0 x

y
E 1 0 x
1 1 2 1 2 xy

xy 0 0
2

It is difficult to represent the curved boundaries by straight edges element a large number of
element may be used to obtain reasonable resembalance between original body and the
assemblage

Two-Dimensional Problems
Review of the Basic Theory
In general, the stresses and strains in a structure consist of six components:

sx , sy , s z , txy , t yz , tzx for stresses,

and e x , e y , ez , g,xy, g,yz , g,zx, for strains.

sy

t yz t xy

y sx
tzx
sz
x
z
Under contain conditions, the state of stresses and strains can be simplified. A general
3-D structure analysis can, therefore, be reduced to a 2-D analysis.
t zx 0 (e z 0) (1)
Plane (2-D) Problems
Plane stress:
sz tyz
 y y

Plane strain:

A long structure with a uniform cross section and transverse loading along its
length (z-direction).

y y

p
x z

Stress-Strain-Temperature (Constitutive) Relations

For elastic and isotropic materials, we have,

ex 1/ E n /E 0 sx ex 0
ey n /E 0 sy ey 0
1/ E
gxy 0 0 1/ G txy gxy 0

where e0 is the initial strain, E the Youngs modulus, n the Poissons ratio and G the
shear modulus. Note that,G =
( )
 which means that there are only two independent materials constants for homogeneous and
isotropic materials.
We can also express stresses in terms of strains by solving the above equation,

The above relations are valid for plane stress case. For plane strain case, we
need to replace the material constants in the above equations in the following fashion,

n
n
1- n
For example, the stress is related to strain by

Initial strains due to temperature change (thermal loading) is given by,

where a is the coefficient of thermal expansion, T the change of temperature. Note that
if the structure is free to deform under thermal loading, there will be no (elastic) stresses in the
structure.

3.6 GENERALIZED COORDINATES APPROACH TO NODEL APPROXIMATIONS

ty
p
y tx

St
Su
x
The boundary S of the body can be divided into two parts, Su and St. The boundary
conditions (BCs) are described as, in which tx and ty are traction forces (stresses on the boundary)
and the barred quantities are those with known values.

In FEM, all types of loads (distributed surface loads, body forces, concentrated forces
and moments, etc.) are converted to point forces acting at the nodes.
Exact Elasticity Solution

The exact solution (displacements, strains and stresses) of a given problem must satisfy the
equilibrium equations, the given boundary conditions and compatibility conditions (structures
should deform in a continuous manner, no cracks and overlaps in the obtained displacement field)
3.7 ISOPARAMETRIC ELEMENTS

In one dimensional problem, each node is allowed to move only in x direction.

But in two dimensional problem, each node is permitted to move in the two directions i.e., x
and y.

The element connectivity table for the above domain is explained as table.

Element (e) Nodes

(1) 123
(2) 234
(3) 435
(4) 536
(5) 637
(6) 738
(7) 839
(8) 931

Constant Strain Triangular (CST) Element
A three noded triangular element is known as constant strain triangular (CST)
element. It has six unknown displacement degrees of freedom (u 1v1, u2v2, u3v3).

Shape function for the CST element Shape function N1 = (p1 + q1x + r1y) / 2A
 Shape function N2 = (p2 + q2x + r2y) / 2A
Shape function N3 = (p3 + q3x + r3y) / 2A

Displacement function for the CST element

u1
v1

0 u2
u(x, y) N1 0 N2 0 N3
Displacement function u = X

v(x, y) 0 N1 0 N2 0 N 3 v2
u3

v3

Strain Displacement matrix [B] for CST element

q1 0 q2 0 q3 0
1 0 r1 0 r2 0 r3

Strain Displacement matrix [B] =
2A
r1 q1 r2 q2 r3 q3
Where, q1 = y2 y3 r1 = x3 x2
q2 = y3 y1 r2 = x1 x3
q3 = y1 y2 r3 = x2 x1
Stress Strain relationship matrix (or) Constitutive matrix [D] for two
dimensional element


1 v v v 0 0 0

v 1 v v 0 0 0

v v 1 v 0 0
0
E
 Stress Strain relationship matrix for two dimensional plane stress problems
The normal stress z and shear stresses xz, yz are zero.

1 v 0
E
v 1 0
[D] = 1 v 2
1 v
0 0
2

Stress Strain relationship matrix for two dimensional plane strain

problems
Normal strain e z and shear strains exz, eyz are zero.

Stiffness matrix equation for two dimensional element (CST element)

Stiffness matrix [k] = [B]T [D] [B] A t

q1 0 q2 0 q3 0
1 0 r1 0 r2 0 r3

2A
[B] =
r1 q1 r2 q2 r3 q3
For plane stress problems,

1 v 0
E
v 1 0
[D] = 1 v 2
1 v
0 0
2

For plane strain problems,

Temperature Effects
Distribution of the change in temperature (T) is known as strain. Due to the
change in temperature can be considered as an initial strain e0.
= D (Bu - e0)

Galerkin Approach

Stiffness matrix [K] e = [B]T [D][B] A t.

Force Vector {F}e = [K]e {u}
 Linear Strain Triangular (LST) element
A six noded triangular element is known as Linear Strain Triangular (LST)
element. It has twelve unknown displacement degrees of freedom. The displacement
functions of the element are quadratic instead of linear as in the CST.

Problem (I set)
1. Determine the shape functions N1, N2 and N3 at the interior point P for the
triangular element for the given figure.

The two dimensional propped beam shown in figure. It is divided into two CST
elements. Determine the nodal displacement and element stresses using plane stress
conditions. Body force is neglected in comparison with the external forces.
Take, Thickness (t) = 10mm,
Youngs modulus (E) = 2x105 N/mm2,
Poissons ratio (v) = 0.25.
3. A thin plate is subjected to surface traction as in figure. Calculate the global stiffness matrix.

Scalar variable problems

In structural problems, displacement at each nodal point is obtained. By using
these displacement solutions, stresses and strains are calculated for each element. In
structural problems, the unknowns (displacements) are represented by the
components of vector field. For example, in a two dimensional plate, the unknown
quantity is the vector field u(x, y), where u is a (2x1) displacement vector.
3.8 STRUCTURAL MECHANICS APPLICATIONS IN 2 DIMENSIONS

Elasticity equations are used for solving structural mechanics problems. These
equations must be satisfied if an exact solution to a structural mechanics problem is to be
obtained. Thest are four basic sets of elasticity equations they are

Strain displacement relationship equations

Stress strain relationship equations
Equilibrium equations
Compatibility equations

TRUSS ELEMENT

A truss element is defined as a deformable, two-force member that is subjected to

loads in the axial direction. The loads can be tensile or compressive. The only degree of
freedom for a one-dimensional truss (bar) element is axial (horizontal) displacement at each
node.

Assumptions for the One-Dimensional Truss Element

Prismatic Member

The truss element is assumed to have a constant cross-section, i.e., it is a prismatic

member. If a truss structure is stepped, then it must be divided up into sections of constant
cross-section in order to obtain an exact solution as shown below.

If a truss structure is tapered, then it can be approximated by using many small truss
elements, each having the same cross-section as the middle of the tapered length it is
approximating. The more sections that are used to approximate a tapered truss, the more
accurate the solution will be.
Weightless Member

The weight (W) of the truss is neglected since it is assumed to be much less than the
total resultant forces (F) acting on the truss. If the weight of the truss is not neglected, then
its effects must be represented as vertical forces acting at the nodes. But since truss element
is defined as two-force member it cannot have any vertical (shear) force, thus the member
weight has to be neglected. If shear forces exist, then a beam element must be used to model
the structure.

Nodal Forces

For one-dimensional truss element, forces (loads) can only be applied at the nodes of
the element, but not between the nodes. This is consistent with the FEM equations which
relate nodal forces to nodal displacements through the stiffness matrix.

Axially Loaded

For one-dimensional truss element, forces (loads) can only be applied at the centroid
of the element cross-sectional area.

Buckling Effect not Considered

A bar element can be subjected to either tensile or compressive forces. Tensile forces
can be applied to a bar of any cross-sectional area or member length, and failure is
associated with sudden fracture or general yielding. When compressive forces are applied to
a member, it can either fail due to crushing or buckling. Buckling is present when the
member bends and laterally deflects as shown on the right figure below.

Buckling is not accounted for in the formulation of the truss element. Members that
do not buckle are classified as short columns and members that buckles are classified as
long columns. The structural response of a short column can be predicted with a truss
element.

To determine if buckling will occur the reader should refer to a mechanics of

material textbook. We will now introduce a simple geometric guideline to determine if
buckling might occur. If the ratio between the member length and the least dimension of the
cross-section is equal or less than 10, the member is considered a short column and buckling
will not occur, i.e.,
Two examples include

In the second case if a bar element is subjected to a compressive force, the element will not
predict the buckling response. One should note that the above geometric rule is a simple
guideline, however, in reality buckling depends not only on the member length and cross-
sectional area, but material properties and support conditions.

Isotropic Material

A truss element has the same mechanical and physical properties in all directions, i.e., they
are independent of direction. For instance, cutting out three tensile test specimens, one in the
x-direction, one in the y-direction and the other oriented 45 degrees in the x-y plane, a
tension test on each specimen, will result in the same mechanical values for the modulus of
elasticity (E), yield strength y and ultimate strength u. Most metals are considered
isotropic. In contrast fibrous materials, such as wood, typically have properties that are
directionally dependant and are generally considered anisotropic (not isotropic).

Constant (Static) Load

The loads that are applied to the truss element are assumed to be static and not to vary over
the time period being considered. This assumption is only valid if the rate of change of the
force is much less than the applied force (F >> dF/dt), i.e., the loads are applied slowly. If
the loads vary significantly, (if the variation in load is not much less than the applied force)
then the problem must be considered as dynamic.

Poisson's Effect not Considered

Poisson's ratio is a material parameter. Poisson's effect is when a uniform cross-section bar
is subject to a tensile load, and the axial stretching is accompanied by a contraction in the
lateral dimension. For one-dimensional truss element., this effect is neglected for simplicity,
i.e., v = 0.

Cross Section Remains Plane

For one-dimensional element, although the force(s) are acting on only the centroid of the
truss (bar) element, it is assumed that it has a uniform effect to the plane. Thus the cross
section will move uniformly and remain plane and normal to the axial axis before and after
loading.
Homogenous Material

A truss element has the same material composition throughout and therefore the same
mechanical properties at every position in the material. Therefore, the modulus of elasticity
E is constant throughout the truss element. A member in which the material properties varies
from one point to the next in the member is called inhomogenous (non-homogenous). If a
truss is composed of different types of materials, then it must be divide up into elements that
are each of a single homogeneous material, otherwise the solution will not be exact.

The left figure shows a composite bar composed of brass and aluminum. This structure can
be divided into two elements as shown on the right, one element for the brass with E 1 = 15 x
106 psi and one for the aluminum with E2 = 10 x 106 psi.

TRUSS ELEMENT (OR SPAR ELEMENT OR LINK ELEMENT)

Differentiate between a truss and a frame.

Truss Frame

Concentrated loads, uniformly distributed

Only concentrated loads act.
loads, moments, all can act.
Loads can be applied at the joints and/or
Loads act only at the joints.
in-between the joints
Truss members undergo only axial Frame members can undergo axial and
deformation (along the length of the bending deformations (translations as well as
member). rotations).

A grid is a structure on which loads applied perpendicular to the plane of the

structure, as opposed to a plane frame, where loads are applied in the plane of the structure.

6.7.1 Derivation of stiffness matrix and finite element equation for a truss element.
There are two joints for an arbitrarily inclined single truss element (at an angle , positive
counter-clockwise from +ve x-axis). For each joint i, there are two degrees of freedom, i.e.,
a joint can have horizontal displacement ui and vertical displacement vi . Hence, for a
single truss element, there are 4 degrees of freedom. The nodal displacement degrees of
freedom and the nodal force degrees of freedom are shown in the following figure.
 u j
vj F jy vj

y uj F jx

y x y x
Fiy
vi ui

Fix vi
ui x

Note that the deformations occurring in the truss members are so small that they are only
axial. The axial displacement of the truss can be resolved along horizontal x-axis and
vertical y-axis. But in our derivation, let us resolve the horizontal and vertical displacements
(in xy-axes) of a joint along and perpendicular to the truss member (in xy -axes). Refer to
the Figure in the next page. Note ui sin component acting towards negative y -direction
and all other components acting towards in +ve x - and y -directions.

ui cos
ui ui cos vi sin
ui
vi ui sin vi cos
ui sin

u j u j cos v j sin
vi cos

vi vj u j sin v j cos



vi sin

The above equations can be written in the matrix form as follows

u cos sin 0 0 ui
v
i sin cos 0 0 vi

u j 0 0 cos sin u j
vj 0 0 sin

cos v j
u T u where T is the transformation matrix

It is important to note that the displacements vi and vj are both zero since there can be no
displacements perpendicular to the length of the member. Also T T
1 T
 Similarly, we resolve forces along the length of the member (positive x direction)
and perpendicular to the length of the member (positive y direction)

Fix cos sin 0 0 Fix

F
iy sin cos 0 0 Fiy

Fjx 0 0 cos sin Fjx
Fjy 0 0 sin

cos Fjy
F T F where T is the transformation matrix

The arbitrarily inclined truss member can be thought of as a simple bar element
oriented at the same angle . Hence, we can write the finite element equation for this
inclined bar element (in xy coordinate system) as

Fix 1 0 1 0 ui
F
iy AE 0 0 0 0 vi

F jx L 1 0 1 0 u j
F jy
0 0 0 0 vj

F k u
Substituting F and u from the previous equations, we can write
T F k T u

Pre-multiplying the above equation by T ,

1

T T F T k T u
1 1

But T T 1 and the above equation can be written as

1

F k u where k T k T
1

Carrying out the matrix multiplication for k , we obtain

Fix c2 cs c 2 cs ui
F
iy AE cs s 2 cs s 2 vi

Fjx L c cs c cs u j
2 2

Fjy
cs s
2
cs s 2 v j
where c cos 2 and s sin 2 .

Computation of strain and stress in the truss element

The change in length of the truss member is equal to the change in axial displacement of the
truss member in the xy co-ordinate system
u j ui
u j cos v j sin ui cos vi sin
ui
v
i
cos sin cos sin
u j
v j

Strain in the truss element is given by e , i.e.,
L
ui

cos sin cos sin vi

e

L u j
v j
Stress in the truss element is given by e E e , i.e.,
ui

cos sin cos sin vi
e E
L u j
v j
Problem

The two-element truss is subjected to external loading as shown in figure. Using the
same node and element numbering as shown in figure, determine the displacement
components at node 3, the reaction components at nodes 1 and 2, and the element
displacement, stresses and forces. The elements have modulus of elasticity E1 = E2 = 1010
lb
2
and cross-sectional areas A1 = A2 = 1.5 in 2
in

(0, 40) 300 lb 500 lb

3
1 (40, 40)

2
(0,0)
 1Finite element model 1 4
3
Degree of freedom 6 3

3
5

2
1
2

For element 1

For element 2
3
vj

j uj

vi
vj
ui uj

vi 1 3
i j

b. Nodal forces
i
element 1 element 2

Fiy

Fjy
Fjx

1 Fix
3
i j
FINITEFELEMENT EQUATION
ix

Fix c2 cs c2 cs ui

Fiy AE cs s2 cs s2 vi
F uj
L c2 cs c2 cs
jx

2

F v
jy cs s2 cs s j
For element 1

AE 1 .5 1 0 1 0 6 lb
4 5, 2 .6 5 1 6 5 1 0 5 2
L 5 6 .5 6 8 5 in
1 2 5 6
0.5 0.5 0.5 0.5 1 1.325826 1.325826 1.325826 1.325826

AE 0.5 0.5 0.5 0.5 5 2 1.325826 1.325826 1.325826 1.325826
K 10
(1)

L 0.5 0.5 0.5 0.5 5 1.325826 1.325826 1.325826 1.325826

0.5 0.5 0.5 0.5 6 1.325826 1.325826 1.325826 1.325826

A E 1 .5 1 0 1 0 6 For element 2
0, 3 .7 5 1 0 5
L 40

3 4 5 6
1 0 1 0 3 3 .7 5 0 3 .7 5 0
0 4 0
AE 0 0 0 0 0 0
K 105
(2)

L 1 0 1 0 5 3 .7 5 0 3 .7 5 0

0 0 0 0 6 0 0 0 0

Assembly of finite element e

1 2 3 4 5 6
F1x 1 1.325 1.325 0 0 1.325 1.325 u1

F1y 2 1.325 1.325 0 0 1.325 1.325 v1
F 3 0 0 3.75 0 3.75 0 u2
2 x 105 quation
F2 y 4 0 0 0 0 0 0 v2
5 1.325 1.325 3.75 0 5.0751 1.325 u3
F3x
F 6 1.325 1.325 0 0 1.325 1.325 v3
3y
Applying boundary conditions knowns unknowns

F3 x 5 0 0 lb F1 x
F3 x 500 lb
F3 y 3 0 0 lb F1 y
F3 y 300 lb
u1 0 F2 x
u1 0 v1 0 F2 y
u2 0 u2 0 u3
v1 0 v2 0 v3
v2 0

F1x 1.325826 1.325826 0 0 1.325826 1.325826 0

F1 y 1.325826 1.325826 0 0 1.325826 1.325826 0
F 0 0 3.75 0 3.75 00
2 x 105
F2 y 0 0 0 0 0 00
1.325826 1.325826 3.75 0 5.0751826 1.325826 u3
500
300
1.325826 1.325826 0 0 1.325826 1.325826 v3

u3 and v3
Eliminating the 1st, 2nd, 3rd and 4th column to solve

500 105 5.075u3 1.325v3

300 105 1.325u3 1.325v3

Solve the equation and

F1x 300 lb
4
u3 5.33 10 in F1 y 300 lb
F2 x 200 lb
3
v3 1.731 10 in F2 y 0

Finding element stresses

For element 1
 E 10 10 6 lb
1 [u5 cos 0 u3 sin 0 ] (0.5333 1) (0 0) 133.325 2
L 40 in

E
2 (u 5 cos 45 u 6 sin 45 ) ( u1 cos 45 u 2 sin 45 )
L

10 10 6
(0.5333 10 3 cos 45 ) (1.731 10 3 sin 45 ) 0
56.57

10 10 6 lb
0.0003771 0.001224 283.03 2
56.57 in

PROBLEM

To illustrate how we can combine spring and bar element in one structure, we can
solve the two-bar truss supported by a spring as shown below. Both bars have E = 210 GPa
and A = 5.0 x10-4 m2. Bar one has a length of 5 m and bar two a length of 10 m. the spring
stiffness is k = 2000 kN/m.

25kN
2

5m 1
o
45
3 2
10m 1

3 k=2000 kN/m

Solution :
Given : E = 210 GPa
N
A = 5.0 x10-4 m2 , L1 = 5 m, L2 = 10 m, K = 2 x 106
m2
N
NOTE: A spring is considered as a bar element whose stiffness is 2 x 106 2
m
STEP 1 : Finite Element Representation Of Forces And Displacements
4
Displacements v2

u2 3 Total # Of Degrees Of Freedom: 8

2

v63 1
v12
u3 5
3 2 u1 1
1

3 8
v4

Forces u4 7
4
F2Y

F2X
2

F3Y 1
F1Y
F3X
3 2 F1X
1

3
F4Y

F4X
4

Step 2: Finite Element Equations

Element 1:
=1350 2
l2=cos2 =0.5
m2=sin2 =0.5 1 135O

lm=cos sin =-0.5

1 x
 1 2 3 4
1 0.5 0.5 0.5 0.5
5 10 4
m 2
210 10 kN/m
6 2
2 0.5 0.5 0.5 0.5
K
(1)

5m 3 0.5 0.5 0.5 0.5

4 0.5 0.5 0.5 0.5
1 2 3 4
1 1 1 1 1
1 1 1
5 2 1
K 105 10
(1)

3 1 1 1 1

4 1 1 1 1

Element 2:
180O
=1800
l2=cos2 =1 2
m2=sin2 =0 3 x
1
lm=cos sin =0

1 2 5 6
1 1 0 1 0
5 10 4
m 2
210 10 kN/m 6 2
2 0 0 0 0
K
(2)

10 m 3 1 0 1 0

4 0 0 0 0
1 2 5 6
1 1 0 1 0
0 0 0
5 2 0
K 105 10
(2)

5 1 0 1 0

6 0 0 0 0

Element 3:
=2700
l2=cos2 =0
m2=sin2 =1
lm=cos sin =0
 1 2 7 8
1 0 0 0 0
0 1
6 2 0 1
K 10
(3)

7 0 0 0 0

8 0 1 0 1

STEP 3: Combination Of Finite Element Equations

1 2 3 4 5 6 7 8
F1 X 1 210 105 105 105 105 0 0 0 u1
F 2 105 125 105 105 0 0 0 20 v
1 Y 1
F2 X 3 105 105 105 105 0 0 0 0 u2

F2Y 5 4 105 108 105 105 0 0 0 0 v2
10
F3 X 5 105 0 0 0 105 0 0 0 u3

F3Y 6 0 0 0 0 0 0 0 0 v3

F4 X 7 0 0 0 0 0 0 0 0 u4

F 8 0 20 20 v
4Y 0 0 0 0 0 4

STEP 4: Applying Boundary Conditions:

Since nodes 1, 2, and 3 are fixed, we have

u2 = v2 = 0; u3 = v3 = 0; u4 = v4 = 0;

F1x = 0 and F1y = -25 kN

0 210 105 105 105 105 0 0 0 u1

25 105 125 105 105 0 0 0 20 v
1
F2 X 105 105 105 105 0 0 0 0 0

F2Y 105 108 105 105 0 0 0 0 0
10
5
105
F3 X 0 0 0 105 0 0 0 0

F3Y 0 0 0 0 0 0 0 0 0

4X
F 0 0 0 0 0 0 0 0 0

F 0 20 20 0
4Y 0 0 0 0 0
Check whether there are as many unknowns as knowns.

STEP 5: SOLVING THE EQUATIONS:

Reduced matrix:

0 210 105 u1
10 5

25 105 125 v1
On solving,
u1=-1.724 x 10-3 m v1=-3.4482 x 10-3 m

Find the reactions at supports by substituting the known nodal values

F2x = -18.104 kN F2y = 18.1041 kN

F3x = 18.102 kN F3y = 0

F4x = 0 F4y = 6.89 kN

STEP 6: Post Processing

Stress in element 1:
1.724

E 3 3.4482
(1) l m l m10
L 0
0
(1) 51.2 MPa (Tensile)

Stress in element 2:
1.724

E 3 3.4482
(1) l m l m10
L 0
0

(2) 36.2 MPa (Compressive)

PROBLEM
A circular concrete beam structure is loaded as shown. Find the deflection of points
at 8,16, and the end of the beam. E = 4 x 106 psi
 y

12 in 3 in 50000 lb
x

24 in

Solution

The beam structure looks very different from a spring. However, its behavior is
very similar. Deflection occurs along the x-axis only. The only significant difference
between the beam and a spring is that the beam has a variable cross-sectional area. An
exact solution can be found if the beam is divided into an infinite number of elements,
then, each element can be considered as a constant cross-section spring element, obeying
the relation F = ku, where k is the stiffness constant of a beam element and is given by
k = AE/L.

In order to keep size of the matrices small (for hand- calculations), let us divide the
beam into only three elements. For engineering accuracy, the answer obtained will be in
an acceptable range. If needed, accuracy can be improved by increasing the number of
elements.
As mentioned earlier in this chapter, spring, truss, and beam elements are line-
elements and the shape of the cross section of an element is irrelevant. Only the cross-
sectional area is needed (also, moment of inertia for a beam element undergoing a
bending load need to be defined). The beam elements and their computer models are
shown

Here, the question of which cross-sectional area to be used for each beam section
arises. A good approximation would be to take the diameter of the mid-section and use
that to approximate the area of the element.

k1 k2 k3

k1 k2 k3
1 2 3
1 2 2 3 3 4
Cross-sectional area
The average diameters are: d1 = 10.5 in., d2 = 7.5 in., d3 = 4.5. (diameters are taken at the
mid sections and the values are found from the height and length ratio of the triangles
shown in figure 2.10), which is given as

12/L = 3/(L-24), L = 32

Average areas are:

A1 = 86.59 in2 A2 = 56.25 in2 A3 = 15.9 in2

24 in

12 in d1 d2 d3

3 in

Original Averaged 8 8 8 L- 24

Stiffness

k1 = A1 E/L1 = (86.59)(4 106/8) = 4.3295 107 lb./in., similarly,

k2 = A2 E/L2 = 2.8125 107 lb./in.
k3 = A3 E/L3 = 7.95 106 lb./in.

Element Stiffness Equations

[K(1)] = 43.295 107 1 -1

-1 1

Similarly,

[K(2)] = 28.125 106 1 -1

-1 1

[K(3)] = 7.9500 106 1 -1

-1 1
Global stiffness matrix is

43.295 -43.295 0 0
[Kg] = -43.295 43.295+28.125 -28.125 0 106
0 -28.125 28.125+7.95 -7.95
0 0 -7.95 7.95

Now the global structural equations can be written as,

43.295 -43.295 0 0 u1 F1
6
10 -43.295 71.42 -28.125 0 u2 = F2
0 -28.125 36.075 -7.95 u3 F3
0 0 -7.95 7.95 u4 F4

Applying the boundary conditions: u1 = 0, and F1 = F2 = F3 = 0, F4 = 5000 lb., results in

the reduced matrix,

71.42 -28.125 0 u2 0
6
10 -28.125 36.075 -7.95 u3 = 0
0 -7.95 7.95 u4 5000
Solving we get,

u2 0.0012
u3 = 0.0029 in.
u4 0.0092

The deflections u2, u3, and u4 are only the approximate values, which can be
improved by dividing the beam into more elements. As the number of elements increases,
the accuracy will improve.
ME6603 Finite Element Analysis

UNIT IV
DYNAMIC ANALYSIS USING ELEMENT METHOD

4.1 INTRODUCTION

It provides the basic equations necessary for structural dynamical analysis and developed
both the lumped and the consistent mass matrix involved in the analysis of bar beam and spring
elements.

4.1.1 Fundamentals of Vibration

Any motion which repeats itself after an interval of time is called vibration or oscillation or
periodic motion

All bodies possessing mass and elasticity are capable of producing vibration.

4.1.2 Causes of Vibrations

o Unbalanced forces in the machine. These force are produced from within the machine
itself
o Elastic nature of the system.
o Self excitations produced by the dry friction between the two mating surfaces.
o External excitations applied on the system.
o Wind may causes vibrations
o Earthquakes may causes vibrations

4.1.3 Types of Vibrations

1.According to the actuating force

Free or natural vibrations

Forced vibrations

Damped vibrations

Undamped vibrations

2.According to motion of system with respect to axis

Longitudinal vibrations

Transverse vibrations

Torsional vibrations
4.2 EQUATION OF MOTION
There is two types of equation of motion
Longitudinal vibration of beam or axial vibration of a rod
Transverse vibration of a beam

z y

Mid surface 4
3

x
1 2
w w t w w
w1 , , w2 , ,
x 1 y 1 x 2 y 2

DOF at each node: w w

w, v , .
y y

On each element, the deflection w(x,y) is represented by

w(x, y) N xi ( w N yi ( w
Ni wi )i )i ,
i 1 x y

where Ni, Nxi and Nyi are shape functions. This is an incompatible element! The
stiffness matrix is still of the form

k = BTEBdV ,
where B is the strain-displacement matrix, and E the stress- strain matrix.
Minding Plate Elements:
4-Node Quadrilateral 8-Node Quadrilateral
Three independent fields.
Deflection w(x,y) is linear for Q4, and quadratic for Q8.

Discrete Kirchhoff Element:

Triangular plate element (not available in ANSYS). Start with a 6-node riangular element,

z y 3

4 6

1 2
t 5 x

DOF at corner nodes: w, w w

, , x, ; y
x y

DOF at mid side nodes: Total DOF x, y.

= 21.

Then, impose conditions

xz 0, etc., at selected
yz
nodes to reduce the DOF (using relations in (15)). Obtain:

z y 3

1 2
x

At each node: w, x w w
, y .
x y

Total DOF = 9 (DKT Element).

Incompatible w(x,y); convergence is faster (w is cubic along each edge) and it is efficient.
Test Problem:

P
y
C
L
L
x L/t = 10, = 0.3
ANSYS 4-node quadrilateral plate element.

ANSYS Result for wc

Mesh wc ( PL2/D)
2 2 0.00593
4 4 0.00598
8 8 0.00574
16 16 0.00565
: :
Exact Solution 0.00560

Question:Converges from above? Contradiction to what we learnt about the nature of

the FEA solution?
Reason: This is an incompatible element ( See comments on p. 177).

Shells and Shell Elements

Shells Thin structures witch span over curved surfaces.

Example:

Sea shell, egg shell (the wonder of the nature); Containers, pipes, tanks;
Car bodies;
Roofs, buildings (the Superdome), etc.
Forces in shells:

Membrane forces + Bending Moments

(cf. plates: bending only)

Example: A Cylindrical Container.

internal forces:

p
p
membrane stresses
dominate
 Shell Theory:
Thin shell theory
Shell theories are the most complicated ones to formulate and analyze in mechanics
(Russians contributions).
Engineering Craftsmanship Demand strong analytical skill
Shell Elements:

plane stress element plate bending element

flat shell element

cf.: bar + simple beam element => general beam element.

DOF at each node:

w
v

u x
y
Q4 or Q8 shell element.
Curved shell elements:

z
i w
v

i u x
y
 Based on shell theories;
Most general shell elements (flat shell and plate elements are subsets);
Complicated in formulation.
Test Cases:
L/2
q L/2
F
A
R A
80o R

F
Roof Pinched Cylinder

F2
F R F
b
A

A F L
F1
F

Pinched Hemisphere

4.3CONSISTENT MASS MATRICES

Natural frequencies and modes

F(t)
Frequency response (F(t)=Fo sinwt) Transient

response (F(t) arbitrary)

4.3.1 Single DOF System
k

m - mass
f=f(t)
m k - stiffness
c c - damping
Free Vibration:
f(t) = 0 and no damping (c = 0)
Eq. (1) becomes
mu ku
(meaning: inertia force + stiffness force = 0)
Assume:

u(t) U sin (wt) ,

where is the frequency of oscillation, U the amplitude.
Eq. (2) yields

2
U m sin( t) kU sin( t) 0
2
i.e., w m k U 0.

For nontrivial solutions for U, we must have

2
w m k 0,

which yields

k
w .
m

This is the circular natural frequency of the single DOF system (rad/s). The cyclic frequency
(1/s = Hz) is
w
f ,
2p
 u = U s in w t
U

t
U
T=1/f

Undamped Free Vibration

With non-zero damping c, where

0 c cc 2mw 2 km (cc = critical damping)

we have the damped natural frequency:

wd 2
w 1 x ,

where x c
(damping ratio).
cc

For structural damping: 0 x 0.15 (usually 1~5%)

wd w.

Thus, we can ignore damping in normal mode analysis.

u

Damped Free Vibration

4.3.2.Multiple DOF System

Equation of Motion

Equation of motion for the whole structure is

Mu
Cu Ku f (t) , (8)

in which: u nodal displacement vector,

M mass matrix,
C damping matrix,
K stiffness matrix,
f forcing vector.

Physical meaning of Eq. (8):

Inertia forces + Damping forces + Elastic forces

= Applied forces

Mass Matrices

Lumped mass matrix (1-D bar element):

rAL 1 r,A,L 2 rAL

1
2 2
u1 u2
Element mass matrix is found to be

rAL
0
m 2
0 rAL
2

diagonal atrix
 Simple Beam Element:
v1
v2
q 1
r, A, L q2

m rNT NdV
V

156 22L 54 13L V1

rAL 22L 4L2 13L 3L2 Q1
420 54 13L 156 22L v2
13L 3L2 22L 4L2 Q2

Units in dynamic analysis (make sure they are consistent):

Choice I Choice II
t (time) s s
L (length) m mm
m (mass) kg Mg
a (accel.) m/s2 mm/s2
f (force) N N
r (density) kg/m3 Mg/mm3

4.4 VECTOR ITERATION METHODS

Study of the dynamic characteristics of a structure:

natural frequencies normal modes shapes)

Let f(t) = 0 and C = 0 (ignore damping) in the dynamic equation (8) and obtain
 Ku
Mu 0

Assume that displacements vary harmonically with time, that is,

u (t ) u sin( w t ),
u(t ) w u cos( w t ),
(t )
u 2 w t ),
w u sin(

where u is the vector of nodal displacement amplitudes.

Eq. (12) yields,

2
K w M u 0

This is a generalized eigenvalue problem (EVP).

Solutions?

This is an n-th order polynomial of from which we can find n solutions (roots) or
eigenvalues

i (i = 1, 2, , n) are the natural frequencies (or characteristic frequencies) of the

structure (the smallest one) is called the fundamental frequency. For each gives one
solution (or eigen) vector

2
K w i M ui 0.

u i (i=1,2,,n) are the normal modes (or natural modes, mode shapes, etc.).
 Properties of Normal Modes

u iT Ku j 0,

u iT M u j 0, for i j,

if wi w j . That is, modes are orthogonal (or independent) to each other with respect to K and
M matrices.

Note:

Magnitudes of displacements (modes) or stresses in normal mode analysis have no physical

meaning.

For normal mode analysis, no support of the structure is necessary.

i = 0 there are rigid body motions of the whole or a part of the structure. apply this to check
the FEA model (check for mechanism or free elements in the models).

Lower modes are more accurate than higher modes in the FE calculations (less spatial variations
in the lower modes fewer elements/wave length are needed).

Example:

v2

r, A, EI q2

1 2
L

2
K wM

EI 12
K 3
L
 EVP: 12 156l 6L
22Ll
2 0,
6L 22Ll 4L 2
4L l

2 4
in which l w rAL / 420 EI .
Solving the EVP, we obtain,
1
2 v2 1
w1 3.533 EI
, ,
4 1.38
rAL q2 1 L
#3 #2
1
2 v2 1
w2 34.81 EI
#1 4 , .
7.62
rAL q2 2 L
Exact solutions:
1 1
2 2
w1 3.516 EI w2 22.03 EI
4 , 4 .
rAL rAL

4.5 MODELLING OF DAMPING

Two commonly used models for viscous damping.

4.5.1 Proportional Damping (Rayleigh Damping)

C M K
(17)
where the constants & are found from

with
1 , 2 , 1 & 2 (damping ratio) being selected.
 Damping ratio

Modal Damping

Incorporate the viscous damping in modal equations.

Modal Equations
Use the normal modes (modal matrix) to transform the coupled system of
dynamic equations to uncoupled system of equations.
We have

2 i 1,2,..., n
K M
i u 0i , (18)

where the normal mode u i satisfies:

T
u i K u j 0,
T for i j,
u i M u j 0,

and

u iT M u i 1,
for i = 1, 2, , n.
u iT K u i i
2
,
Form the modal matrix:

(n n ) u 1 u 2 L u n

are called principal coordinates.

Substitute (21) into the dynamic equation:

M &z& C z& K z f ( t ).

T
Pre-multiply by , and apply (20):

&z& C z& z p ( t ),

where C I (proportional damping),

p T
f (t ) .

Using Modal Damping

Can verify that

Transformation for the displacement vector,

z 1 (t) z 2 (t) M

z n (t)

Equation (22) becomes,

2 ( t ),
i z i p
&z&i 2 i i z& i i = 1,2,,n. (24)
i

Equations in (22) or (24) are called modal equations. These are uncoupled, second-
order differential equations, which are much easier to solve than the original dynamic
equation (coupled system).

To recover u from z, apply transformation (21) again, once z is obtained from (24).
 Notes:

Only the first few modes may be needed in constructing the modal matrix (i.e.,
could be an n m rectangular matrix with m<n). Thus, significant reduction in the
size of the system can be achieved.

Modal equations are best suited for problems in which higher modes are not
important (i.e., structural vibrations, but not shock loading).

4.5.2 Frequency Response Analysis

(Harmonic Response Analysis)

Ku E u
Harmonicloading
(25)

Modal method: Apply the modal equations,

2 sin t,
&zi 2 i i Zk i i zi pi i=1,2,,m. (26)
These are 1-D equations. Solutions are

2
p i i
zi (t) sin( t
2 2 2
(1 ) (2i )
zi

i i

where

ci ci
/i i , damping ratio
cc 2m i

Recover u from (21).

Direct Method: Solve Eq. (25) directly, that is, calculate

the inverse. With u i t
ue (complex notation), Eq. (25)
becomes

This equation is expensive to solve and matrix is ill- conditioned if is close

4.6TRANSIENT RESPONSE ANALYSIS
(Dynamic Response/Time-History Analysis)
Structure response to arbitrary, time-dependent loading.

f(t)

t
u(t)

Compute responses by integrating through time:

u1

u n u n+1
u2

t0 t1 t2 t n t n+1
B. Modal Method

First, do the transformation of the dynamic equations using the modal matrix before the
time marching:

Then, solve the uncoupled equations using an integration method. Can use, e.g.,
10%, of the total modes (m= n/10).
Uncoupled system, Fewer equations,
No inverse of matrices,
More efficient for large problems.
4.6.1Cautions in Dynamic Analysis
Symmetry: It should not be used in the dynamic analysis (normal modes, etc.)
because symmetric structures can have antisymmetric modes.

Mechanism, rigid body motion means = 0. Can use this to check FEA models
to see if they are properly connected and/or supported.

Input for FEA: loading F(t) or F( ) can be very complicated in real applications
and often needs to be filtered first before used as input for FEA.

Examples
Impact, drop test, etc.

PROBLEM

In the spring structure shown k1 = 10 lb./in., k2 = 15 lb./in., k3 = 20 lb./in., P= 5 lb.

Determine the deflection at nodes 2 and 3.

k1 k2 k3
o o o o
1 2 3 4

Figure 2.4

Solution:

Again apply the three steps outlined previously.

Step 1: Find the Element Stiffness Equations

Element 1:
1 2
[K(1)] = 10 -10 1
-10 10 2

Element 2: 2 3
[K(2)] = 15 -15 2
-15 15 3

Element 3: 3 4
(3)
[K ] = 20 -20 3
-20 20 4
Step 2: Find the Global stiffness matrix

1 2 3 4
1 10 -10 0 0 10 -10 0 0
2 -10 10 + 15 -15 0 = -10 25 -15 0
3 0 -15 15 + 20 -20 0 -15 35 -20
4 0 0 -20 20 0 0 -20 20

Now the global structural equation can be written as,

F1 10 -10 0 0 u1
F2 = -10 25 -15 0 u2
F3 0 -15 35 -20 u3
F4 0 0 -20 20 u4

Step 3: Solve for Deflections

The known boundary conditions are: u1 = u4 = 0, F3 = P = 3lb. Thus, rows and columns 1 and 4 will drop
out, resulting in the following matrix equation,

0 25 15 2
= =
3 15 35 3

Solving, we get u2 = 0.0692 & u3 = 0.1154

PROBLEM
In the spring structure shown, k1 = 10 N/mm, k2 = 15 N/mm, k3 = 20 N/mm, k4 = 25 N/mm, k5 = 30
N/mm, k6 = 35 N/mm. F2 = 100 N. Find the deflections in all springs.

k1

k3
k2 F2 k6
k4

k5
Solution:

Here again, we follow the three-step approach described earlier, without specifically
mentioning at each step.

Element 1: 1 4
[K(1)] = 10 -10 1
-10 10 4

Element 2: 1 2
[K(2)] = 15 -15 1
-15 15 2

Element 3: 2 3
[K(3)] = 20 -20 2
-20 20 3

Element 4: 2 3
(4)
[K ] = 25 -25 2
-25 25 3

Element 5: 2 4
(5)
[K ] = 30 -30 2
-30 30 4

Element 6: 3 4
[K(6)] = 35 -35 3
-35 35 4

The global stiffness matrix is,

1 2 3 4

10+15 -15 0 -10 1

[Kg] = -15 15+20+25+30 -20-25 -30 2
0 -20-25 20+25+35 -35 3
-10 -30 -35 10+30+35 4

And simplifying, we get

25 -15 0 -10
[Kg] = -15 90 -45 -30
0 -45 80 -35
-10 -30 -35 75
And the structural equation is,

F1 25 -15 0 -10 u1
F2 = -15 90 -45 -30 u2
F3 0 -45 80 -35 u3
F4 -10 -30 -35 75 u4

Now, apply the boundary conditions, u1 = u4 = 0, F2 = 100 N. This is carried out by

deleting the rows 1 and 4, columns 1 and 4, and replacing F2 by 100N. The final matrix
equation is,

100 90 -45 u2
Which 0 = -45 80 u3 gives

Deflections:

Spring 1: u4 u1 = 0

Spring 2: u2 u1 = 1.54590

Spring 3: u3 u2 = -0.6763

Spring 4: u3 u2 = -0.6763

Spring 5: u4 u2 = -1.5459

Spring 6: u4 u3 = -0.8696
 UNIT V
APPLICATIONS IN HEAT TRANSFER &FLUID MECHANICS
5.1 ONE DIMENSIONAL HEAT TRANSFER ELEMENT
In structural problem displacement at each nodel point is obtained. By
using these displacement solutions, stresses and strains are calculated for each
element. In structural problems, the unknowns are represented by the
components of vector field. For example, in a two dimensional plate, the
unknown quantity is the vector field u(x,y),where u is a (2x1)displacement
vector.
Heat transfer can be defined as the transmission of energy from one
region another region due to temperature difference. A knowledge of the
temperature distribution within a body is important in many engineering
problems. There are three modes of heat transfer.
They are: (i) Conduction
(ii) Convection
(iii) Radiation
5.1.1Strong Form for Heat Conduction in One Dimension
with Arbitrary Boundary Conditions

Following the same procedure as in Section, the portion of the

boundary where the temperature is prescribed, i.e. the essential boundary is
denoted by T and the boundary where the flux is prescribed is recommended
for Science and Engineering Track. Denoted by q ; these are the
boundaries with natural boundary conditions. These boundaries are
complementary, so

q = ; q \ T 0:

With the unit normal used in , we can express the natural boundary
condition as qn q. For example, positive flux q causes heat inflow (negative
q ) on the left boundary point where qn q q and heat outflow (positive
q ) on the right boundary point where qn q q.

Strong form for 1D heat conduction problems

5.1.2Weak Form for Heat Conduction in One Dimension with
Arbitrary Boundary Conditions

We again multiply the first two equations in the strong form by the weight
function and integrate over the domains over which they hold, the domain for the
differential equation and the domain q for the flux boundary condition, which
yields ws dx with w
Recalling that w 0 on T and combining with gives

Weak form for 1D heat conduction problems

Find T x 2 U such that
Notice the similarity between
5.2 APPLICATION TO HEAT TRANSFER TWO-DIMENTIONAL

5.2.1Strong Form for Two-Point Boundary Value Problems

The equations developed in this chapter for heat conduction, diffusion and
elasticity problems are all of the following form:

Such one-dimensional problems are called two-point boundary value

problems. gives the particular meanings of the above variables and parameters
for several applications. The natural boundary conditions can also be
generalized as (based on Becker et al. (1981))

0 on :

Equation is a natural boundary condition because the derivative of the

solution appears in it. reduces to the standard natural boundary conditions
considered in the previous sections when bx 0. Notice that the essential
boundary condition can be recovered as a limiting case of when bx is a
penalty parameter, i.e. a large number In this case, and Equation is
called a generalized boundary condition.
An example of the above generalized boundary condition is an elastic bar
with a spring attached as shown in In this case, bl k and reduces to

E(n-l) ( k-u) u 0 at x l;
 where k is the spring constant. If the spring stiffness is set to a very large
value, the above boundary condition enforces u; if we let k 0, the above
boundary condition corresponds to a prescribed traction boundary. In practice,
such generalized boundary conditions are often used to model the influence of
the surroundings. For example, if the bar is a simplified model of a building and its
foundation, the spring can represent the stiffness of the soil.

5.2.2 Two-Point Boundary Value Problem With Generalized Boundary

Conditions

u(l)
uk
- ku(l) t

An example of the generalized boundary for elasticity problem.

Another example of the application of this boundary condition is

convective heat transfer, where energy is transferred between the surface of the
wall and the surrounding medium. Suppose convective heat transfer occurs at x
l. Let T l be the wall temperature at x l and T be the temperature in the
medium. Then the flux at the boundary x l is given by ql hT l T , so
bl h and the boundary condition is

where h is convection coefficient, which has dimensions of W m 2 o

C 1 . Note that when the convection coefficient is very large, the temperature
T is immediately felt at x l and thus the essential boundary condition is
again enforced as a limiting case of the natural boundary condition.

There are two approaches to deal with the boundary condition . We will
call them the penalty and partition methods. In the penalty method, the
essential boundary condition is enforced as a limiting case of the natural
boundary condition by equating bx to a penalty parameter. The resulting
strong form for the penalty method is given in.

General strong form for 1D problems-penalty method

f 0 on ;
STRONG AND WEAK FORMS FOR ONE-DIMENSIONAL PROBLEMS

In the partition approach, the total boundary is partitioned into the natural
boundary, and the complementary essential boundary, The natural boundary condition has
the generalized form defined by The resulting strong form for the partition method is
summarized in.

5.2.3 Weak Form for Two-Point Boundary Value Problems

In this section, we will derive the general weak form for two-point boundary value
problems. Both the penalty and partition methods described in will be considered. To
obtain the general weak form for the penalty method, we multiply the two equations in the
strong by the weight function and integrate over the domains over which they hold: the
domain for the differential equation and the domain for the generalized boundary
condition.

5.3 SCALE VARIABLE PROBLEM IN 2 DIMENSIONS

x1
y
2
x1

x N1 0 N2 0 N3 0 N4 0 y2
u

y 0 N1 0 N2 0 N3 0 N 4 x3
y3

x4

y 4

N1=1/4(1-) (1-); N2=1/4(1+) (1-); N3=1/4(1+) (1+); N4=1/4(1-) (1+).

Equation of Stiffness Matrix for 4 noded isoparametric quadrilateral element

J 11 J 12
J ;
J 21 J 22
STRONG AND WEAK FORMS FOR ONE-DIMENSIONAL PROBLEMS

N 2 N 3 N 4
N1 0
0 0 0

N 2 N 3 N 4
J 22 J 12 0 N1 0
B 1
0 0 0 0

J 0 0 J 21 J 11
N1 N 2 N 3 N 4
J 21 J 11 J 12 0 0 0 0

J 22

0 N1 N 2 N 3 N 4
0 0 0

1 v v 0
E
[D] v 1 0
(1 v )
2
1 v , for plane stress conditions;
0 0
2


E 1 v v 0

[D] v 1 v 0 , for plane strain conditions.
(1 v)(1 2v) 1 2v
0 0
2

Equation of element force vector


F e [N ] ;
Fx T

Fy
N Shape function, Fx load or force along x direction,
Fy load or force along y direction.

Numerical Integration (Gaussian Quadrature)

The Gauss quadrature is one of the numerical integration methods to calculate the
definite integrals. In FEA, this Gauss quadrature method is mostly preferred. In
this method the numerical integration is achieved by the following expression,
1 n

f (x)dx w f (x )
1 i 1
i i

Table gives gauss points for integration from -1 to 1.

 Number of Location Corresponding Weights
Points xi wi
n

1 x1 = 0.000 2.000

2 1
x1, x2 = 0.577350269189
3 1.000

5
x1, x3
3
0.774596669241 0.555555
3 5 9

x2=0.000 8
0.888888
9

4 x1, x4 = 0.8611363116 0.3478548451

x2, x3 = 0.3399810436 0.6521451549

Problem (I set)

x
1

1. Evaluate I cos
1
2
dx , by applying 3 point Gaussian quadrature and

compare with exact solution.

x 1
1
1
2
I 3e x dx , using one point and two point
x 2
2. Evaluate

Gaussian quadrature. Compare with exact solution.

3. For the isoparametric quadrilateral element shown in figure, determine the
local co ordinates of the point P which has Cartesian co-ordinates (7, 4).
 4. A four noded rectangular element is in figure. Determine (i) Jacobian
matrix, (ii) Strain Displacement matrix and (iii) Element Stresses. Take
E=2x105N/mm2,= 0.25, u=[0,0,0.003,0.004,0.006, 0.004,0,0]T, = 0, =0.
Assume plane stress condition.

5.4 2 DIMENTIONAL FLUID MECHANICS

The problem of linear elastostatics described in detail in can be extended to include the
effects of inertia. The resulting equations of motion take the form

+ f = u in I ,
n = t on q I ,
u = u on u I ,
u(x1 ,x2 ,x3 , 0) = u0 (x1 ,x2 ,x3 ) in ,
v(x1 , x2 ,x3 ,0) = v0 (x1 ,x2 ,x3 ) in ,
 where u = u(x1 , x2 , x3 , t) is the unknown displacement field, is the mass density, and I
= (0, T ) with T being a given time. Also, u0 and v0 are the prescribed initial displacement
and velocity fields. Clearly, two sets of boundary conditions are set on u and q ,
respectively, and are assumed to hold throughout the time interval I . Likewise, two sets of
initial conditions are set for the whole domain at time t = 0. The strong form of the
resulting initial/boundary- value problem is stated as follows: given functions f , t, u , u0 and
v0 , as well as a constitutive equation for , find u in I , such that the equations are
satisfied.
A Galerkin-based weak form of the linear elastostatics problem has been derived in Sec-
tion In the elastodynamics case, the only substantial difference involves the inclusion
R
of the term w u d, as long as one adopts the semi-discrete approach. As a result, the
weak form at a fixed time can be expressed as
Z Z Z Z
w u d + s w : d = w f d + w t d .

Following the development of Section 7.3, the discrete counterpart of can be written as
Z Z Z w t d .
h
wh uh d + (wh ) D(uh ) d wh f d +
=

Following a standard procedure, the contribution of the forcing vector Fint,e due to
interele- ment tractions is neglected upon assembly of the global equations. As a result, the
equations is give rise to their assembled counterparts in the form

Mu + Ku = F ,

where u is the global unknown displacement vector1 . The preceding equations are, of course,
subject to initial conditions that can be written in vectorial form as u(0) = u0 and v(0) = v0
.
The most commonly employed method for the numerical solution of the system of cou-
pled linear second-order ordinary differential equations is the Newmark method. This
method is based on a time series expansion of u and u := v. Concentrating on the time
interval (tn ,tn+1 ], the Newmark method is defined by the equations

1 2
un+1 = un + vn tn + [(1 2)an + 2an+1 ]tn ,
2
vn + [(1 )an + an+1 ]tn ,
vn+1 =
 It is clear that the Newmark equations define a whole family of time integrators.
It is important to distinguish this family into two categories, namely implicit and explicit
integrators, corresponding to > 0 and = 0, respectively.

The overhead hat symbol is used to distinguish between the vector field u and the
solution vector uemanating from the finite element approximation of the vector field u.

The general implicit Newmark integration method may be implemented as follows: first,
solve (9.18)1 for an+1 , namely write

an+1 = (un+1 un vn tn ) an
t2n

Then, substitute (9.19) into the semi-discrete form (9.17) evaluated at tn+1 to find that

M +K un+1 = Fn+1
n
.

After solving for un+1 , one may compute the acceleration an+1 from and the velocity
vn+1 from.
Finally, the general explicit Newmark integration method may be implemented as follows:
starting from the semi-discrete equations evaluated at tn+1 , one may substitute u n+1from
to find that

Man+1 = K(un + vn tn +an tn ) + Fn+1 .

If M is rendered diagonal (see discussion in Chapter 8), then an+1 can be determined
without solving any coupled linear algebraic equations. Then, the velocities bv n+1 are
immediately computed from (9.18)2 . Also, the displacements u n+1 are computed from
indepen-dently of the accelerations an+1 .
 QUESTION BANK
PART A QUESTIONS WITH ANSWERS

UNIT 1
1. What is meant by finite element?
A small units having definite shape of geometry and nodes is called finite element.
2. What is meant by node or joint?
Each kind of finite element has a specific structural shape and is inter- connected with
the adjacent element by nodal point or nodes. At the nodes, degrees of freedom are
located. The forces will act only at nodes at any others place in the element.
3. What is the basic of finite element method?
Discretization is the basis of finite element method. The art of subdividing a structure
in to convenient number of smaller components is known as discretization.
4. What are the types of boundary conditions?
Primary boundary conditions
Secondary boundary conditions
5. State the methods of engineering analysis?
Experimental methods
Analytical methods
Numerical methods or approximate methods
6. What are the types of element?
7. 1D element
2D element
3D element
8. State the three phases of finite element method.
Preprocessing
Analysis
Post Processing
9. What is structural problem?
Displacement at each nodal point is obtained. By these displacements solution
stress and strain in each element can be calculated.
10. What is non structural problem?
Temperature or fluid pressure at each nodal point is obtained. By using these
values properties such as heat flow fluid flow for each element can be calculated.

10. What are the methods are generally associated with the finite element analysis?
Force method
Displacement or stiffness method.
11. Explain stiffness method.
Displacement or stiffness method, displacement of the nodes is considered as the
unknown of the problem. Among them two approaches, displacement method is
desirable.
12. What is meant by post processing?
Analysis and evaluation of the solution result is referred to as post processing.
Postprocessor computer program help the user to interpret the result by displaying
them in graphical form.
13. Name the variation methods.
Ritz method.
Ray-Leigh Ritz method.
14. What is meant by degrees of freedom?
When the force or reaction act at nodal point node is subjected to deformation. The
deformation includes displacement rotation, and or strains. These are collectively
known as degrees of freedom
15. What is meant by discretization and assemblage?
The art of subdividing a structure in to convenient number of smaller components
is known as discretization. These smaller components are then put together. The
process of uniting the various elements together is called assemblage.
16. What is Rayleigh-Ritz method?
It is integral approach method which is useful for solving complex structural
problem, encountered in finite element analysis. This method is possible only if a
suitable function is available.
17. What is Aspect ratio?
It is defined as the ratio of the largest dimension of the element to the smallest
dimension. In many cases, as the aspect ratio increases the in accuracy of the
solution increases. The conclusion of many researches is that the aspect ratio
18. What is truss element?
The truss elements are the part of a truss structure linked together by point joint
which transmits only axial force to the element.
19. What are the h and p versions of finite element method?
It is used to improve the accuracy of the finite element method. In h version, the
order of polynomial approximation for all elements is kept constant and the numbers of
elements are increased. In p version, the numbers of elements are maintained constant and
the order of polynomial approximation of element is increased.
20. Name the weighted residual method
Point collocation method
Sub domain collocation method
Least squares method
Galerkins method.

UNIT 2

21. List the two advantages of post processing.

Required result can be obtained in graphical form. Contour diagrams can be used to
understand the solution easily and quickly.
22. During discretization, mention the places where it is necessary to place a
node?
Concentrated load acting point
Cross-section changing point
Different material interjections
Sudden change in point load
23. What is the difference between static and dynamic analysis?
Static analysis: The solution of the problem does not vary with time is known as
static analysis
Example: stress analysis on a beam
Dynamic analysis: The solution of the problem varies with time is known as
dynamic analysis
 24. Name any four FEA softwares.
ANSYS
NASTRAN
COSMOS
25. Differentiate between global and local axes.
Local axes are established in an element. Since it is in the element level, they change
with the change in orientation of the element. The direction differs from element to
element.

Global axes are defined for the entire system. They are same in direction for all the
elements even though the elements are differently oriented.
26. Distinguish between potential energy function and potential energy functional
If a system has finite number of degree of freedom (q 1,q2,and q3), then the potential
energy expressed as,
= f (q1,q2,and q3)
It is known as function. If a system has infinite degrees of freedom then the potential
energy is expressed as
dy d y
f x , y ,
2
, ....dx
2

dx dx
27. What are the types of loading acting on the structure?
Body force (f)
Traction force (T)
Point load (P)
28. Define the body force
A body force is distributed force acting on every elemental volume of the
body Unit: Force per unit volume.
Example: Self weight due to gravity
29. Define traction force
Traction force is defined as distributed force acting on the surface of the
body. Unit: Force per unit area.
Example: Frictional resistance, viscous drag, surface shear
30. What is point load?
Point load is force acting at a particular point which causes displacement.
 31. What are the basic steps involved in the finite element modeling.
Discretization of structure.
Numbering of nodes.
32. Write down the general finite element equation.
F K u

33. What is discretization?

The art of subdividing a structure in to a convenient number of smaller components is
known as discretization.
34. What are the classifications of coordinates?
Global coordinates
Local coordinates
Natural coordinates
35. What is Global coordinates?
The points in the entire structure are defined using coordinates system is known as
global coordinate system.
36. What is natural coordinates?
A natural coordinate system is used to define any point inside the element by a set
of dimensionless number whose magnitude never exceeds unity. This system is
very useful in assembling of stiffness matrices.
37. Define shape function.
Approximate relation (x,y) = N1 (x,y) 1 + N2 (x,y) 2 + N3 (x,y) 3
Where 1, 2, and 3 are the values of the field variable at the nodes N1, N2, and
N3 are the interpolation functions.
N1, N2, and N3 are also called shape functions because they are used to express the
geometry or shape of the element.
38. What are the characteristic of shape function?
It has unit value at one nodal point and zero value at other nodal points. The sum of
shape function is equal to one.
39. Why polynomials are generally used as shape function?
Differentiation and integration of polynomial are quit easy.
The accuracy of the result can be improved by increasing the order of the
polynomial. It is easy to formulate and computerize the finite element equations
 40. How do you calculate the size of the global stiffness matrix?
Global stiffness matrix size = Number of nodes X Degrees of freedom per node

UNIT 3
41. Write down the expression of stiffness matrix for one dimensional bar element.
AE 1 1
K
l 1 1
42. State the properties of stiffness matrix
It is a symmetric matrix
The sum of elements in any column must be equal to zero
It is an unstable element. So the determinant is equal to zero.
43. Write down the expression of stiffness matrix for a truss element.
u
1
u 1 l m 0 0 u2

u 0 0l mu

2 3

u
4
44. Write down the expression of shape function N and displacement u for one
dimensional bar element.
U= N1u1+N2u2
N1= 1-X /
l N2 = X / l
45. Define total potential energy.
Total potential energy, = Strain energy (U) + potential energy of the external forces
(W)
46. State the principle of minimum potential energy.
Among all the displacement equations that satisfied internal compatibility and the
boundary condition those that also satisfy the equation of equilibrium make the
potential energy a minimum is a stable system.
47. Write down the finite element equation for one dimensional two noded bar element.
AE 1 1
K

l 1 1

48. What is truss?

A truss is defined as a structure made up of several bars, riveted or welded together.
49. States the assumption are made while finding the forces in a truss.
All the members are pin jointed.
The truss is loaded only at the joint
The self weight of the members is neglected unless stated.
50. State the principles of virtual energy?
A body is in equilibrium if the internal virtual work equals the external virtual work
for the every kinematically admissible displacement field
51. What is essential boundary condition?
Primary boundary condition or EBC Boundary condition which in terms of field
variable is known as Primary boundary condition.

52. Natural boundary conditions?

Secondary boundary natural boundary conditions which are in the differential form
of field variable is known as secondary boundary condition
53. How do you define two dimensional elements?
Two dimensional elements are define by three or more nodes in a two dimensional
plane. The basic element useful for two dimensional analysis is the triangular
element.
54. What is CST element?
Three noded triangular elements are known as CST. It has six unknown
displacement degrees of freedom (u1, v1, u2, v2, u3, v3). The element is called
CST because it has a constant strain throughout it.
55. What is LST element?
Six nodded triangular elements are known as LST. It has twelve unknown
displacement degrees of freedom. The displacement function for the elements are
quadratic instead of linear as in the CST.
56. What is QST element?
Ten nodded triangular elements are known as Quadratic strain triangle. It is also
called as cubic displacement triangle.
58. What meant by plane stress analysis?
Plane stress is defined to be a state of stress in which the normal stress and shear
stress directed perpendicular to the plane are assumed to be zero.
 60. Define plane strain analysis.
Plane strain is defined to be state of strain normal to the xy plane and the shear
strains are assumed to be zero.

UNIT 4

61. Write down the stiffness matrix equation for two dimensional CST elements.

Stiffness matrix K B D B At
T

BT -Strain displacement D-Stress strain matrix B-Strain displacement matrix

62. Write down the stress strain relationship matrix for plane stress conditions.

1 0
E
1 12 0 1 0

12
0 0
2

63. What is axisymmetric element?

Many three dimensional problem in engineering exhibit symmetry about an axis of
rotation such type of problem are solved by special two dimensional element called
the axisymmetric element
64. What are the conditions for a problem to be axisymmetric?
The problem domain must be symmetric about the axis of revolution All
boundary condition must be symmetric about the axis of revolution All
loading condition must be symmetric about the axis of revolution
65. Give the stiffness matrix equation for an axisymmetric triangular
element. Stiffness matrix K B D B 2rA
T

66. What is the purpose of Isoparametric element?

It is difficult to represent the curved boundaries by straight edges finite elements. A
large number of finite elements may be used to obtain reasonable resemblance
between original body and the assemblage.
67. Write down the shape functions for 4 noded rectangular elements using natural
coordinate system.
 N1 1 11 N2 1 11
4 4

N3 1 1 1 N4 1 1 1
4 4
68. Write down Jacobian matrix for 4 noded quadrilateral elements.
J J
J 11 12

J J
21 22
69. Write down stiffnes matrix equation for 4 noded isoparametric quadrilateral elements.

Stiffness matrix K t B T D B J
1 1

1 1

70. Define super parametric element.

If the number of nodes used for defining the geometry is more than of nodes used
for defining the displacement is known as super parametric element
71. Define sub parametric element.
If the number of nodes used for defining the geometry is less than number of nodes
used for defining the displacement is known as sub parametric element.
72. What is meant by Isoparametric element?
If the number of nodes used for defining the geometry is same as number of nodes
used for defining the displacement is known as Isoparametric element.
73. Is beam element an Isoparametric element?
Beam element is not an Isoparametric element since the geometry and displacement
are defined by different order interpretation functions.
74. What is the difference between natural coordinate and simple natural
coordinate?
L1 = 1-x/l
L2 = x/l
75. What is Area coordinates?
L1 = A1/A L2 = A2/A L3 = A3/A
76. What is simple natural coordinate?
A simple natural coordinate is one whose value between -1 and 1.
77. Give example for essential boundary conditions.
The geometry boundary condition are displacement, slope.
78. Give example for non essential boundary conditions.
The natural boundary conditions are bending moment, shear force
79. What is meant by degrees of freedom?
When the force or reaction act at nodal point node is subjected to deformation. The
deformation includes displacement rotation, and or strains. These are collectively
known as degrees of freedom.
80. What is QST element?
Ten noded triangular elements are known as Quadratic strain triangle. It is also
called as cubic displacement triangle.

UNIT 5

81. What meant by plane stress analysis?

Plane stress is defined to be a state of stress in which the normal stress and shear
stress directed perpendicular to the plane are assumed to be zero.
82. Define plane strain analysis.
Plane strain is defined to be state of strain normal to the x,y plane and the shear
strains are assumed to be zero.

83. What is truss element?

The truss elements are the part of a truss structure linked together by point joint
which transmits only axial force to the element.

84. List the two advantages of post processing.

Required result can be obtained in graphical form.
Contour diagrams can be used to understand the solution easily and quickly.
85. What are the h and p versions of finite element method?
It is used to improve the accuracy of the finite element method. In h version, the order
of polynomial approximation for all elements is kept constant and the numbers of
elements are increased. In p version, the numbers of elements are maintained constant
and the order of polynomial approximation of element is increased.
86. During discretization, mention the places where it is necessary to place a node?
Concentrated load acting point Cross-
section changing point Different
material inter junction point Sudden
change in point load
87. What is the difference between static and dynamic analysis?
Static analysis: The solution of the problem does not vary with time is known as
static analysis
Example: stress analysis on a beam
Dynamic analysis: The solution of the problem varies with time is known as
dynamic analysis
Example: vibration analysis problem.
88. What is meant by discretization and assemblage?
The art of subdividing a structure in to convenient number of smaller components is
known as discretization. These smaller components are then put together. The
process of uniting the various elements together is called assemblage.
89. What is Rayleigh-Ritz method?
It is integral approach method which is useful for solving complex structural
problem, encountered in finite element analysis. This method is possible only if a
suitable function is available.
90. What is Aspect ratio?
It is defined as the ratio of the largest dimension of the element to the smallest
dimension. In many cases, as the aspect ratio increases the in accuracy of the solution
increases. The conclusion of many researches is that the aspect ratio should be close
to unity as possible.
91. What is essential boundary condition?
Primary boundary condition or EBC, Boundary condition which in terms of field
variable is known as Primary boundary condition
92. Natural boundary conditions.
Secondary boundary natural boundary conditions which are in the differential
form of field variable is known as secondary boundary condition.
93. How do you define two dimensional elements?
Two dimensional elements are define by three or more nodes in a two dimensional
 plane. The basic element useful for two dimensional analysis is the triangular
element.
94. State the principles of virtual energy?
A body is in equilibrium if the internal virtual work equals the external virtual work
for the every kinematically admissible displacement field..

96. What is non-homogeneous form?

When the specified values of dependent variables are non-zero, the boundary
conditi said to be non-homogeneous.
97. What is homogeneous form?
When the specified values of dependent variables is zero, the boundary condition are
said to be homogeneous.
98. Define initial value problem.
An initial value problem is one in which the dependent variable and possibly is
derivatives are specified initially.
99. Define boundary value problem.
A differential equation is said to describe a boundary value problem if the dependent
variable and its derivatives are required to take specified values on the boundary.