1

Lecture 1: Introduction

1.1.1 Introduction
The Finite Element Method (FEM) is a numerical technique to find approximate solutions of partial
differential equations. It was originated from the need of solving complex elasticity and structural
analysis problems in Civil, Mechanical and Aerospace engineering. In a structural simulation, FEM
helps in producing stiffness and strength visualizations. It also helps to minimize material weight
and its cost of the structures. FEM allows for detailed visualization and indicates the distribution of
stresses and strains inside the body of a structure. Many of FE software are powerful yet complex
tool meant for professional engineers with the training and education necessary to properly interpret
the results.
Several modern FEM packages include specific components such as fluid, thermal,
electromagnetic and structural working environments. FEM allows entire designs to be constructed,
refined and optimized before the design is manufactured. This powerful design tool has significantly
improved both the standard of engineering designs and the methodology of the design process in
many industrial applications. The use of FEM has significantly decreased the time to take products
from concept to the production line. One must take the advantage of the advent of faster generation
of personal computers for the analysis and design of engineering product with precision level of
accuracy.

1.1.2 Background of Finite Element Analysis

The finite element analysis can be traced back to the work by Alexander Hrennikoff (1941) and
Richard Courant (1942). Hrenikoff introduced the framework method, in which a plane elastic
medium was represented as collections of bars and beams. These pioneers share one essential
characteristic: mesh discretization of a continuous domain into a set of discrete sub-domains, usually
called elements.
• In 1950s, solution of large number of simultaneous equations became possible because of the
digital computer.
• In 1960, Ray W. Clough first published a paper using term “Finite Element Method”.
• In 1965, First conference on “finite elements” was held.
• In 1967, the first book on the “Finite Element Method” was published by Zienkiewicz and
Chung.
• In the late 1960s and early 1970s, the FEM was applied to a wide variety of engineering
problems.
 2

• In the 1970s, most commercial FEM software packages (ABAQUS, NASTRAN, ANSYS,
etc.) originated. Interactive FE programs on supercomputer lead to rapid growth of CAD
systems.
• In the 1980s, algorithm on electromagnetic applications, fluid flow and thermal analysis were
developed with the use of FE program.
• Engineers can evaluate ways to control the vibrations and extend the use of flexible,
deployable structures in space using FE and other methods in the 1990s. Trends to solve fully
coupled solution of fluid flows with structural interactions, bio-mechanics related problems
with a higher level of accuracy were observed in this decade.

With the development of finite element method, together with tremendous increases in computing
power and convenience, today it is possible to understand structural behavior with levels of
accuracy. This was in fact the beyond of imagination before the computer age.

1.1.3 Numerical Methods

The formulation for structural analysis is generally based on the three fundamental relations:
equilibrium, constitutive and compatibility. There are two major approaches to the analysis:
Analytical and Numerical. Analytical approach which leads to closed-form solutions is effective in
case of simple geometry, boundary conditions, loadings and material properties. However, in reality,
such simple cases may not arise. As a result, various numerical methods are evolved for solving such
problems which are complex in nature. For numerical approach, the solutions will be approximate
when any of these relations are only approximately satisfied. The numerical method depends heavily
on the processing power of computers and is more applicable to structures of arbitrary size and
complexity. It is common practice to use approximate solutions of differential equations as the basis
for structural analysis. This is usually done using numerical approximation techniques. Few
numerical methods which are commonly used to solve solid and fluid mechanics problems are given
below.

• Finite Difference Method

• Finite Volume Method
• Finite Element Method
• Boundary Element Method
• Meshless Method

The application of finite difference method for engineering problems involves replacing the
governing differential equations and the boundary condition by suitable algebraic equations. For
 3

example in the analysis of beam bending problem the differential equation is reduced to be solution
of algebraic equations written at every nodal point within the beam member. For example, the beam
equation can be expressed as:
d 4w q
4
= (1.1.1)
dx EI
To explain the concept of finite difference method let us consider a displacement function variable
namely w = f ( x)

Fig. 1.1.1 Displacement Function

Now, w  f (x  x)  f (x)

dwΔw f(x + Δx) - f(x) 1
So, = Lt = Lt =  wi+1 - wi  (1.1.2)
dxΔx Δx0 ΔxΔx
0 h
Thus,
d 2w d  1  1 1
=   wi+1 - wi  = 2  wi+2 - wi+1 - wi+1 + wi  = 2  wi+2 - 2 wi+1 + wi  (1.1.3)
dx 2
dx  h  h h
d 3w 1
=  wi+3 - wi+2 - 2wi+2 + 2wi+1 + wi+1 - wi 
dx 3 h 3 (1.1.4)
1
= 3  wi+3 - 3wi+2 + 3wi+1 - wi 
h
 4

d 4w 1
=  wi+4 - wi+3 - 3wi+3 + 3wi+2 + 3wi+2 - 3wi+1 - wi+1 + wi 
dx 4 h 4
1
= 4  wi+4 - 4wi+3 +6wi+2 - 4wi+1 + wi  (1.1.5)
h
1
= 4  wi+2 - 4wi+1 +6wi - 4wi-1 + wi-2 
h
Thus, eq. (1.1.1) can be expressed with the help of eq. (1.1.5) and can be written in finite difference
form as:
q 4
( wi − 2 − 4 wi −1 + 6 wi − 4 wi +1 + wi + 2 ) = h (1.1.6)
EI

Fig. 1.1.2 Finite difference equation at node i

Thus, the displacement at node i of the beam member corresponds to uniformly distributed load can
be obtained from eq. (1.1.6) with the help of boundary conditions. It may be interesting to note that,
the concept of node is used in the finite difference method. Basically, this method has an array of
grid points and is a point wise approximation, whereas, finite element method has an array of small
interconnecting sub-regions and is a piece wise approximation.
Each method has noteworthy advantages as well as limitations. However it is possible to
solve various problems by finite element method, even with highly complex geometry and loading
conditions, with the restriction that there is always some numerical errors. Therefore, effective and
reliable use of this method requires a solid understanding of its limitations.

1.1.4 Concepts of Elements and Nodes

Any continuum/domain can be divided into a number of pieces with very small dimensions. These
small pieces of finite dimension are called ‘Finite Elements’ (Fig. 1.1.3). A field quantity in each
element is allowed to have a simple spatial variation which can be described by polynomial terms.
Thus the original domain is considered as an assemblage of number of such small elements. These
elements are connected through number of joints which are called ‘Nodes’. While discretizing the
structural system, it is assumed that the elements are attached to the adjacent elements only at the
nodal points. Each element contains the material and geometrical properties. The material properties
inside an element are assumed to be constant. The elements may be 1D elements, 2D elements or 3D
elements. The physical object can be modeled by choosing appropriate element such as frame
 5

element, plate element, shell element, solid element, etc. All elements are then assembled to obtain
the solution of the entire domain/structure under certain loading conditions. Nodes are assigned at a
certain density throughout the continuum depending on the anticipated stress levels of a particular
domain. Regions which will receive large amounts of stress variation usually have a higher node
density than those which experience little or no stress.

Fig. 1.1.3 Finite element discretization of a domain

1.1.5 Degrees of Freedom

A structure can have infinite number of displacements. Approximation with a reasonable level of
accuracy can be achieved by assuming a limited number of displacements. This finite number of
displacements is the number of degrees of freedom of the structure. For example, the truss member
will undergo only axial deformation. Therefore, the degrees of freedom of a truss member with
respect to its own coordinate system will be one at each node. If a two dimension structure is
modeled by truss elements, then the deformation with respect to structural coordinate system will be
two and therefore degrees of freedom will also become two. The degrees of freedom for various
types of element are shown in Fig. 1.1.4 for easy understanding. Here ( u , v, w ) and (θ x , θ y , θ z )
represent displacement and rotation respectively.
 6

Fig. 1.1.4 Degrees of Freedom for Various Elements

 7

Lecture 2: Basic Concepts of Finite Element Analysis

1.2.1 Idealization of a Continuum

A continuum may be discretized in different ways depending upon the geometrical configuration of
the domain. Fig. 1.2.1 shows the various ways of idealizing a continuum based on the geometry.

Fig. 1.2.1 Various ways of Idealization of a Continuum

1.2.2 Discretization of Technique

The need of finite element analysis arises when the structural system in terms of its either geometry,
material properties, boundary conditions or loadings is complex in nature. For such case, the whole
structure needs to be subdivided into smaller elements. The whole structure is then analyzed by the
assemblage of all elements representing the complete structure including its all properties.
 8

The subdivision process is an important task in finite element analysis and requires some
skill and knowledge. In this procedure, first, the number, shape, size and configuration of elements
have to be decided in such a manner that the real structure is simulated as closely as possible. The
discretization is to be in such that the results converge to the true solution. However, too fine mesh
will lead to extra computational effort. Fig. 1.2.2 shows a finite element mesh of a continuum using
triangular and quadrilateral elements. The assemblage of triangular elements in this case shows
better representation of the continuum. The discretization process also shows that the more accurate
representation is possible if the body is further subdivided into some finer mesh.

Fig. 1.2.2 Discretization of a continuum

1.2.3 Concepts of Finite Element Analysis

FEA consists of a computer model of a continuum that is stressed and analyzed for specific results.
A continuum has infinite particles with continuous variation of material properties. Therefore, it
needs to simplify to a finite size and is made up of an assemblage of substructures, components and
members. Discretization process is necessary to convert whole structure to an assemblage of
members/elements for determining its responses. Fig. 1.2.3 shows the process of idealization of
actual structure to a finite element form to obtain the response results. The assumptions are required
to be made by the experienced engineer with finite element background for getting appropriate
response results. On the basis of assumptions, the appropriate constitutive model can be constructed.
For the linear-elastic-static analysis of structures, the final form of equation will be made in the form
of F=Kd where F, K and d are the nodal loads, global stiffness and nodal displacements respectively.
 9

Classical
Actual Structure
∂σ x ∂σ y ∂σ z
+ + + FΩ =
0
∂x ∂y ∂z
(Partial Differential Equations)

Assumptions
Equilibrium Equations
Stress-Strain Law
Compatibility Conditions

   d    d F d    d F d

T T T

  
(Principles of Virtual Work)

FEM
Structural Model
F  =  K d 
(Algebraic Equations)

Fig. 1.2.3From classical to FE solution

Varieties of engineering problem like solid and fluid mechanics, heat transfer can easily be solved by
the concept of finite element technique. The basic form of the equation will become as follows
where action, property and response parameter will vary for case to case as outlined in Table 1.2.1.
 10

{F } [=
K ]{d } OR {d } [ K ] {F }
−1
=

Action Property Response

Table 1.2.1 Response parameter for different cases

Property Action Response
Solid Stiffness Load Displacement
Fluid Viscosity Body force Pressure/Velocity
Thermal Conductivity Heat Temperature

1.2.4 Advantages of FEA

1. The physical properties, which are intractable and complex for any closed bound solution,
can be analyzed by this method.
2. It can take care of any geometry (may be regular or irregular).
3. It can take care of any boundary conditions.
4. Material anisotropy and non-homogeneity can be catered without much difficulty.
5. It can take care of any type of loading conditions.
6. This method is superior to other approximate methods like Galerkine and Rayleigh-Ritz
methods.
7. In this method approximations are confined to small sub domains.
8. In this method, the admissible functions are valid over the simple domain and have nothing
to do with boundary, however simple or complex it may be.
9. Enable to computer programming.

1.2.5 Disadvantages of FEA

1. Computational time involved in the solution of the problem is high.
2. For fluid dynamics problems some other methods of analysis may prove efficient than the
FEM.

1.2.6 Limitations of FEA

1. Proper engineering judgment is to be exercised to interpret results.
2. It requires large computer memory and computational time to obtain intend results.
3. There are certain categories of problems where other methods are more effective, e.g., fluid
problems having boundaries at infinity are better treated by the boundary element method.
 11

4. For some problems, there may be a considerable amount of input data. Errors may creep up
in their preparation and the results thus obtained may also appear to be acceptable which
indicates deceptive state of affairs. It is always desirable to make a visual check of the input
data.
5. In the FEM, many problems lead to round-off errors. Computer works with a limited number
of digits and solving the problem with restricted number of digits may not yield the desired
degree of accuracy or it may give total erroneous results in some cases. For many problems
the increase in the number of digits for the purpose of calculation improves the accuracy.

1.2.7 Errors and Accuracy in FEA

Every physical problem is formulated by simplifying certain assumptions. Solution to the problem,
classical or numerical, is to be viewed within the constraints imposed by these simplifications. The
material may be assumed to be homogeneous and isotropic; its behavior may be considered as
linearly elastic; the prediction of the exact load in any type of structure is next to impossible. As
such the true behavior of the structure is to be viewed with in these constraints and obvious errors
creep in engineering calculations.
1. The results will be erroneous if any mistake occurs in the input data. As such, preparation of
the input data should be made with great care.
2. When a continuum is discretised, an infinite degrees of freedom system is converted into a
model having finite number of degrees of freedom. In a continuum, functions which are
continuous are now replaced by ones which are piece-wise continuous within individual
elements. Thus the actual continuum is represented by a set of approximations.
3. The accuracy depends to a great extent on the mesh grading of the continuum. In regions of
high strain gradient, higher mesh grading is needed whereas in the regions of lower strain,
the mesh chosen may be coarser. As the element size decreases, the discretisation error
reduces.
4. Improper selection of shape of the element will lead to a considerable error in the solution.
Triangle elements in the shape of an equilateral or rectangular element in the shape of a
square will always perform better than those having unequal lengths of the sides. For very
long shapes, the attainment of convergence is extremely slow.
5. In the finite element analysis, the boundary conditions are imposed at the nodes of the
element whereas in an actual continuum, they are defined at the boundaries. Between the
nodes, the actual boundary conditions will depend on the shape functions of the element
forming the boundary.
 12

6. Simplification of the boundary is another source of error. The domain may be reduced to the
shape of polygon. If the mesh is refined, then the error involved in the discretised boundary
may be reduced.
7. During arithmetic operations, the numbers would be constantly round-off to some fixed
working length. These round–off errors may go on accumulating and then resulting accuracy
of the solution may be greatly impaired.
 13

Lecture 3: Introduction to Elasticity

1.3.1 Stresses and Equilibrium

Let consider an infinitesimal element of sides dx, dy and dz as shown in Fig. 1.3.1. The stresses are
acting on the elemental volume dV because of external and or body forces. These stresses can be
represented by six independent components as given below.
{σ } = σ x ,
T
σ y , σ z , τ xy , τ yz , τ zx  (1.3.1)
Here, σ x , σ y and σ z are normal stresses and τ xy , τ yz and τ zx are shear stresses. Applying the
conditions of static equilibrium for forces along the direction of X axis (i.e., ∑F
x = 0 ), following
expression will be obtained.

∂σ x ∂τ yx ∂τ
dxdydz + dxdydz + zx dxdydz + FΩx dxdydz =
0 (1.3.2)
∂x ∂y ∂z

Fig. 1.3.1Stresses on an infinitesimal element

 14

Where, FΩx is the component of body force along x direction. Now, dividing dxdydz on the above
expression, following equilibrium condition is obtained.
∂σ x ∂τ yx ∂τ zx
+ + =
− fx (1.3.3)
∂x ∂y ∂z

Similarly, applying equilibrium condition along Y and Z directions, one can find the following
relations.
∂τ xy ∂σ y ∂τ zy
+ + = − fy (1.3.4)
∂x ∂y ∂z
∂τ xz ∂τ yz ∂σ z
+ + =− fz (1.3.5)
∂x ∂y ∂z

Here, FΩy and FΩz are the component of body forces along Y and Z directions respectively. Satisfying
moment equations (i.e.,= ∑ M x 0;= ∑ M y 0 and= ∑ M z 0; ), one can obtain the following
relations.

=τ xy τ=
yx ; τ yz τ zy and
= τ xz τ zx (1.3.6)

Using eq. (1.3.6), the equilibrium equations (1.3.3 to 1.3.5), can be rewritten in the following form.
∂σ x ∂τ xy ∂τ zx
+ + = − FΩx
∂x ∂y ∂z
∂τ xy ∂σ y ∂τ yz
+ + = − FΩy (1.3.7)
∂x ∂y ∂z
∂τ zx ∂τ yz ∂σ z
+ + = − FΩz
∂x ∂y ∂z

Eq. (1.3.7) is known as equation of equilibrium.

Let assume an element of area ∆Γ on the surface of the solid in equilibrium (Fig.1.3.2) and FΓx, FΓy
and FΓz are the components of external forces per unit area and are acting on the surface.
Consideration of equilibrium along the three axes directions gives the following relations.
σ x l + τ xy m + τ zx n =
FΓx
τ xy l + σ y m + τ yz n =
FΓy (1.3.8)
τ zx l + τ yz m + σ z n =
FΓz

Here, l, m and n are the direction cosines of the normal to the boundary surface. Eq. (1.3.8) is known
as static boundary condition.
 15

Fig. 1.3.2 Forces acting on an elemnt on the boundary

1.3.2 Strain-Displacement Relations

The displacement at any point of a deformable body may be expressed by the components of u, v
and w parallel to the Cartesian coordinate’s axes. The components of the displacements can be
described as functions of x, y and z. Displacements basically the change of position during
deformation. If point P (x,y,z) is displaced to P’ (x’,y’,z’), then the displacement along X, Y and Z
direction (Fig. 1.3.3) will become
x’ = x + u or u = x’ – x
y’ = y + v or v = y’ – y
z’ = z + w or w = z’ – z
Therefore, the normal strain can be written as:
∆u ∂u ∆L
=εx =
Lt (As ε = for uniform strain in axial member)
∆x →0 ∆x ∂x L
∂v ∂w
=
Similarly, εy = and ε z
∂y ∂z
 16

Fig. 1.3.3 Deformation of an elastic body

Let consider points P, Q and R are before deformation and points P’,Q’ and R’ are after deformation
as shown in Fig. 1.3.4 below. Now for small deformation, rotation of PQ will become
∆v ∂v
=θ1 lim
=
∆x →0 ∆x ∂x
∆u ∂u
be: θ 2 lim
Similarly, rotation of PR due to deformation will = =
∆y →0 ∆y ∂y
Thus, the total change of angle between PQ and PR after deformation is as follows which is defined
as shear strain in X-Y plane.
∂v ∂u
γ xy = γ yx =θ1 + θ 2 = +
∂x ∂y

Fig. 1.3.4 Derivation of shear strain

 17

Similarly, shear strains in Y-Z and X-Z plane will become

∂w ∂v ∂w ∂u
γ yz =
γ zy =+ ; γ xz =
γ zx =+
∂y ∂z ∂x ∂z
The strain can be expressed as partial derivatives of the displacements u, v and w. The above
expressions for strain-displacement relationship are true only for small amplitude of deformation.
However, the strain-displacement relations are expressed by the following equations for large
magnitude of deformation.
∂u 1  ∂u   ∂v   ∂w  
2 2 2

εx = +   +   +    (1.3.9)
∂x 2  ∂x   ∂x   ∂x  

∂v 1  ∂u   ∂v   ∂w 
2 2 2

ε y = +   +   +    (1.3.10)
∂y 2  ∂y   ∂y   ∂y  

∂w 1  ∂u   ∂v   ∂w  
2 2 2

εz = +   +   +    (1.3.11)
∂z 2  ∂z   ∂z   ∂z  
∂v ∂u  ∂u ∂u ∂v ∂v ∂w ∂w 
γ xy = + + + +
∂x ∂y  ∂x ∂y ∂x ∂y ∂x ∂y 
(1.3.12)

∂w ∂v  ∂u ∂u ∂v ∂v ∂w ∂w 
γ yz = + + + +
∂y ∂z  ∂y ∂z ∂y ∂z ∂y ∂z 
(1.3.13)

∂u ∂w  ∂u ∂u ∂v ∂v ∂w ∂w 
γ zx = + + + + (1.3.14)
∂z ∂x  ∂z ∂x ∂z ∂x ∂z ∂x 

The eqs.(1.3.9 to 1.3.14) are known as Green-Lagrange strain displacement equation. The
components of the strain εx, εy, εz, γxy, γyz and γzx define the state of strains in the deformed body, and
can be written in a matrix form as

   x  y  z  xy  yz  zx 

T
(1.3.15)

The relations given in eqs.(1.3.9 to 1.3.14) are non-linear partial differential equations in the
unknown component of the displacements. In case of small deformations, the products and squares
of the first derivatives are assumed to be negligible compared with the derivatives themselves in
many problems of stress analysis. Thus the strain-displacement relations in eqs. (1.3.9 to 1.3.14 )
reduce to linear relations as follows.
 18

∂u
εx =
∂x
∂v
εy =
∂y
∂w
εz =
∂z
(1.3.16)
∂v ∂u
γ= +
∂x ∂y
xy

∂w ∂v
γ= +
∂y ∂z
yz

∂u ∂w
γ= +
∂z ∂x
zx

Eq. (1.3.16) is known as Von-Karman strain displacement equation. The above equation can be
expressed in a matrix form as given below.
∂ 
 ∂x 0 0
 ∂ 
ε x   0 0
ε   ∂y 
 y  ∂ u 
 ε z   0 ∂z   v 
0
 =∂ ∂   (1.3.17)
γ xy   0   w
γ yz   ∂y ∂x  
   ∂ ∂
γ zx   0 ∂z ∂y 


∂ 0
∂
 ∂z ∂x 

The above assumption will be incorrect in case of large deformation problems. In these cases,
geometric nonlinearity has to be considered.

1.3.3 Linear Constitutive Relations

Hooke’s law states that the six component of stress may be described as linear function of six
components of strain. The relation for a linear elastic, anisotropic and homogeneous material are
expressed as follows.
 19

σ x  C11 C12 . . . C16  ε xε 

σ  C
 y   21 C 22 . . . C 26   ε y 
σ z   . .   ε z 
 =   (1.3.18)
τ xy   . .  γ xy 
τ yz   . .  γ yz 
    
τ zx  C 61 C 62 . . . C 66  γ zx 

or {σ } = [C ]{ε } (1.3.19)

Where [C] is constitutive matrix. If the material has three orthogonal planes of symmetry, it is said
to be orthotropic. In this case only nine constants are required for describing constitutive relations as
given below.

 σ x  C11 C12 C13 0 0 0  ε x 

σ   C 22 C 23 0 0 0  ε 
 y   y 
σ zτ   C 33 0 0 0   ε z 
 =   (1.3.20)
τ xy   C 44 0 0  γ xy 
τ yz   Symmetry C 55 0  γ yz 
    
 τ zx   C 66  γ zx 

The inverse relation for strains and stresses may be expressed as

   C1     D (1.3.21)

An isotropic is one for which every plane is a plane of symmetry of material behavior and only two
constants (Young Modulus, E and Poisson ratio µ) are required to describe the constitutive relation.
The following equation includes the effect due to temperature changes as may be necessary in
certain cases of stress analysis.

 x   1   0  x  1

  
0 0

 
  
 
 
 

    1  0 0 0     

 y 
   

. 
 

1

 
  1 1 0 0 0  .    
1
   T  
z      
or      (1.3.22)

 
 xy  E  2 1    0 0  .   
0


 
   
 
 



 yz 
  2 1    0  

. 
 

0

       
 
 
 zx  
 2 1   

 zx

  

0

 20

T and α in eq. (1.3.22) denote the difference of temperature and coefficient of thermal expansion
respectively.
The inverse relation of stresses in terms of strain components can be expressed as
(1 − µ ) µ µ 0 0 0 
σ x   (1 − µ ) µ 0 0 0  ε x  1
σ     1
  y  (1 − µ ) 0 0 0  ε y   
 1 − 2µ   ε  EαT 1
σ z   0  z  −
  = E 0   (1.3.23)
τ
   2  γ xy  1 − 2µ 0
1 − 2µ
xy
τ yz   0  γ yz  0
   2    
τ zx   1 − 2µ  γ zx  0
 
 2 
E
where E =
(1 + µ )(1 − 2µ )

1.3.4 Two-Dimensional Stress Distribution

The problems of solid mechanics may be formulated as three-dimensional problems and finite
element technique may be used to solve them. In many practical situations, the geometry and loading
will be such that the problems may be formulated to two-dimensional or one-dimensional problems
without much loss of accuracy. The relation between strain and displacement for two dimensional
problems can be simplified from eq. (1.3.16) and can be written as follows.
∂u
εx =
∂x
∂v
εy = (1.3.24)
∂y
∂v ∂u
γ= +
∂x ∂y
xy

The above expression can be written in a combined form:

∂ 2ε x ∂ ε y ∂ γ xy
2 2

+ 2 = (1.3.25)
∂y 2 ∂x ∂x∂y
Eq. (1.3.25) is the compatibility equation since it states the geometric requirements. This condition
will ensure adjacent elements to remain free from discontinuities such as gaps and overlaps.

1.3.4.1 Plane stress problem

The plane stress problem is characterized by very small dimensions in one of the normal directions.
Some typical examples are shown in Fig. 1.3.5. In these cases, it is assumed that no stress
 21

component varies across the thickness and the stress components σz, τxz and τyz are zero. The state of
stress is specified by σx, σy and τxy only and is called plane stress.

Fig. 1.3.5 Plane stress example: Thin plate with in-plane loading

The stress components may be expressed in terms of strain, which is as follows.

 
 x      1
 1  0  x
    E     ET  
 1 0   y  1 (1.3.26)
 y  1 2     1  
 xy  
0 0
1   xy 
 0
 2 

The strain components can also be expressed in terms of the stress, which is given below.

 x   1  0   x  1
 

 
   
 
  
  1
  y    1   
0    y  T 1 (1.3.27)

  
 E 0    

 xy 
   0 2 1    


 xy 
 





0


It can also be shown that
−µ
εz = (ε x + ε y ) + 1 + µ αT and γ yz = γ zx = 0 (1.3.28)
1− µ 1− µ

1.3.4.2 Plane strain problem

Problems involving a long body whose geometry and loading do not vary significantly in the
longitudinal direction are referred to as plane strain problems. Some typical examples are given in
Fig. 1.3.6. In these cases, a constant longitudinal displacement corresponding to a rigid body
 22

translation and displacements linear in z corresponding to rigid body rotation do not result in strain.
As a result, the following relations arise.
ε z = γ yz = γ zx = 0 (1.3.29)
The constitutive relation for elastic isotropic material for this case may be given by,

 
σ x  1 − µ µ 0  ε x  1
  E   E α T  
σ y  =  µ 1− µ 0  ε y  − 1 (1.3.30)
τ  (1 + µ )(1 − 2 µ )  
1 − 2µ  γ  1 − 2 µ 0
 xy   0 0  xy   
 2 

Also σ z = µ (σ x + σ y ) − EαT and τ yz = τ zx = 0 (1.3.31)

The strain components can be expressed in terms of the stress as follows.

 x   0    


  1    1  


 x  
1
 
 

   
 
  1  0  y   1   T 1 (1.3.32)
 y
  
 E  0  
 
 



 xy  
 0 2 
 
 
  xy 
 

0



(a) Retaining wall (b) Dam

Fig. 1.3.6 Plane strain examples

 23

1.3.4.3Axisymmetric Problem
Many problems in stress analysis which are of practical interest involve solids of revolution subject
to axially symmetric loading. A circular cylinder loaded by a uniform internal or external pressure,
circular footing resting on soil mass, pressure vessels, rotating wheels, flywheels etc. The strain-
displacement relations in these type of problems are given by
∂u
εx =
∂x
u
εθ =
x
(1.3.33)
∂v
εy =
∂y
∂u ∂v
γ= +
∂y ∂x
xy

The two components of displacements in any plane section of the body along its axis of symmetry
define completely the state of strain and therefore the state of stress. The constitutive relations are
given below for such types of problems.

σ x  (1 − µ ) µ µ 0  ε 
σ   µ − µ µ  x 
 y E  1 0  ε y 
 =  µ µ 1− µ 0  ε  (1.3.34)
σ θ  (1 + µ )(1 − 2µ )  1 − 2µ   
θ
τ xy   0 0 0  γ 
 2   xy 
 24

Lecture 4: Steps in Finite Element Analysis

1.4.1 Loading Conditions

There are multiple loading conditions which may be applied to a system. The load may be internal
and/or external in nature. Internal stresses/forces and strains/deformations are developed due to the
action of loads. Most loads are basically “Volume Loads” generated due to mass contained in a
volume. Loads may arise from fluid-structure interaction effects such as hydrodynamic pressure of
reservoir on dam, waves on offshore structures, wind load on buildings, pressure distribution on
aircraft etc. Again, loads may be static, dynamic or quasi-static in nature. All types of static loads
can be represented as:
• Point loads
• Line loads
• Area loads
• Volume loads
The loads which are not acting on the nodal points need to be transferred to the nodes properly using
finite element techniques.

1.4.2 Support Conditions

In finite element analysis, support conditions need to be taken care in the stiffness matrix of the
structure. For fixed support, the displacement and rotation in all the directions will be restrained and
accordingly, the global stiffness matrix has to modify. If the support prevents translation only in one
direction, it can be modeled as ‘roller’ or ‘link supports’. Such link supports are commonly used in
finite element software to represent the actual structural state. Sometimes, the support itself
undergoes translation under loadings. Such supports are called as ‘elastic support’ and are modeled
with ‘spring’. Such situation arises if the structures are resting on soil. The supports may be
represented in finite element modeling as:
• Point support
• Line support
• Area support
• Volume support

1.4.3 Type of Engineering Analysis

Finite element analysis consists of linear and non-linear models. On the basis of the structural system
and its loadings, the appropriate type of analysis is chosen. The type of analysis to be carried out
depends on the following criteria:
 25

• Type of excitation (loads)

• Type of structure (material and geometry)
• Type of response

Considering above aspects, types of engineering analysis are decided. FEA is capable of using
multiple materials within the structure such as:
• Isotropic (i.e., identical throughout)
0
• Orthotropic (i.e., identical at 90 )
• General anisotropic (i.e., different throughout)
The Equilibrium Equations for different cases are as follows:
1. Linear-Static:
Ku = F (1.4.1)
2. Linear-Dynamic
Mu(t ) + Cu (t ) + Ku (t ) =
F (t ) (1.4.2)
3. Nonlinear - Static
Ku + FNL =
F (1.4.3)
1. Nonlinear-Dynamic
Mu(t ) + Cu (t ) + Ku (t ) + F (t ) NL =
F (t ) (1.4.4)
Here, M, C, K, F and U are mass, damping, stiffness, force and displacement of the structure
respectively. Table 1.4.1 shows various types of analysis which can be performed according to
engineering judgment.

Table 1.4.1 Type of analysis

Excitation Structure Response Basic analysis type

Static Elastic Linear Linear-Elastic-Static Analysis

Static Elastic Nonlinear Nonlinear-Elastic-Static Analysis
Static Inelastic Linear Linear-Inelastic-Static Analysis
Static Inelastic Nonlinear Nonlinear-Inelastic-Static Analysis
Dynamic Elastic Linear Linear-Elastic-Dynamic Analysis
Dynamic Elastic Nonlinear Nonlinear-Elastic-Dynamic Analysis
Dynamic Inelastic Linear Linear-Inelastic-Dynamic Analysis
Dynamic Inelastic Nonlinear Nonlinear-Inelastic-Dynamic Analysis
 26

1.4.4 Basic Steps in Finite Element Analysis

The following steps are performed for finite element analysis.
1. Discretisation of the continuum: The continuum is divided into a number of elements by
imaginary lines or surfaces. The interconnected elements may have different sizes and
shapes.
2. Identification of variables: The elements are assumed to be connected at their intersecting
points referred to as nodal points. At each node, unknown displacements are to be prescribed.
3. Choice of approximating functions: Displacement function is the starting point of the
mathematical analysis. This represents the variation of the displacement within the element.
The displacement function may be approximated in the form a linear function or a higher-
order function. A convenient way to express it is by polynomial expressions. The shape or
geometry of the element may also be approximated.
4. Formation of the element stiffness matrix: After continuum is discretised with desired
element shapes, the individual element stiffness matrix is formulated. Basically it is a
minimization procedure whatever may be the approach adopted. For certain elements, the
form involves a great deal of sophistication. The geometry of the element is defined in
reference to the global frame. Coordinate transformation must be done for elements where it
is necessary.
5. Formation of overall stiffness matrix: After the element stiffness matrices in global
coordinates are formed, they are assembled to form the overall stiffness matrix. The
assembly is done through the nodes which are common to adjacent elements. The overall
stiffness matrix is symmetric and banded.
6. Formation of the element loading matrix: The loading forms an essential parameter in any
structural engineering problem. The loading inside an element is transferred at the nodal
points and consistent element matrix is formed.
7. Formation of the overall loading matrix: Like the overall stiffness matrix, the element
loading matrices are assembled to form the overall loading matrix. This matrix has one
column per loading case and it is either a column vector or a rectangular matrix depending on
the number of loading cases.
8. Incorporation of boundary conditions: The boundary restraint conditions are to be
imposed in the stiffness matrix. There are various techniques available to satisfy the
boundary conditions. One is the size of the stiffness matrix may be reduced or condensed in
its final form. To ease computer programming aspect and to elegantly incorporate the
boundary conditions, the size of overall matrix is kept the same.
9. Solution of simultaneous equations: The unknown nodal displacements are calculated by
the multiplication of force vector with the inverse of stiffness matrix.
 27

10. Calculation of stresses or stress-resultants: Nodal displacements are utilized for the
calculation of stresses or stress-resultants. This may be done for all elements of the
continuum or it may be limited to some predetermined elements. Results may also be
obtained by graphical means. It may desirable to plot the contours of the deformed shape of
the continuum.

The basic steps for finite element analysis are shown in the form of flow chart below:

Fig. 1.4.1 Flowchart for steps in FEA

 28

1.4.5 Element Library in FEA Software

A real structure can be modeled with various ways with appropriate assumptions. The structure may
be divided into following categories:
• Cable or tension structures
• Skeletal or framed structures
• Surface or spatial structures
• Solid structures
• Mixed structures

The configuration of structural elements depends upon the geometry of the structural system and the
number of independent space coordinates (i.e., x, y and z) required to describe the problem. Thus, the
element can be categorized as one, two or three dimensional element. One dimensional element can
be represented by a straight line whose ends will be nodal points. The skeletal structures are
generally modeled by this type of elements. The pin jointed bar or truss element is the simplest
structural element. This element undergoes only axial deformation. The beam element is another
type of element which undergoes in-plane transverse displacements and rotations. The frame
element is the combination of truss and beam element. Thus, the frame element has axial and in-
plane transverse displacements and rotations. This element is generally used to model 1D, 2D and
3D skeletal structural systems. Two-dimensional elements are generally used to model 2D and 3D
continuum. These elements are of constant thickness and material properties. The shapes of these
elements are triangular or rectangular and it consists of 3 to 9 or even more nodes. These elements
are used to solve many problems in solid mechanics such as plane stress, plane strain, plate bending.
Three-dimensional element is the most cumbersome which is generally used to model the 3-D
continuum. The elements have 6 to 27 numbers of nodes or more. Because of large degrees of
freedom, the analysis is time consuming using 3-D elements and difficult to interpret its results.
However, for accurate analysis of the irregular continuum, 3-D elements are useful. To analyze any
real structure, appropriate elements are to be assigned for the finite element analysis. In standard
FEA software, following types of element library are used to discretize the domain.
• Truss element
• Beam element
• Frame element
• Membrane/ Plate/Shell element
• Solid element
• Composite element
• Shear panel
• Spring element
 29

• Rigid/Link element
• Viscous damping element
The different types of elements available in standard finite element software are shown in Fig. 1.4.2.

1D Elements (Truss, beam, grid and frame)

2D Elements (Plane stress, Plane strain, Axisymmetric, Plate and Shell)

3D Elements

Fig. 1.4.2 Various types of elements for computer modeling

 1

Lecture 1: Virtual Work and Variational Principle

2.1.1 Introduction
Finite element formulation can be constructed from governing differential equations over a domain.
This can be formulated by various ways like Virtual work method, Variational method, Weighted
Residual Method etc.

2.1.2 Principle of Virtual Work

The principle of virtual work is a very useful approach for solving varieties of structural mechanics
problem. When the force and displacement are unrelated to the cause and effect relation, the work is
called virtual work. Therefore, the virtual work may be caused by true force moving through
imaginary displacements or vice versa. Thus, the principle of virtual work can be divided into two
categories: (a) principle of virtual forces and (b) principle of virtual displacements. The principle of
virtual forces establishes the compatibility conditions. The principle of virtual displacements
establishes the conditions of equilibrium and is used in the displacement model of the finite element
technique.
The external virtual work is the work done by real load moving through imaginary
displacements in a structure. These loads include both the load distributed over the entire surface and
volume. Thus, the virtual work done by the external force is:
Fx  Fx 
   
WE   u v w Fy d   u v w Fy d
    (2.1.1)

 z 
F 
Fz 
Where, δu, δv and δw are the components of the virtual displacements in x, y and z direction
respectively. FΓx, FΓy and FΓz are the surface forces and FΩx, FΩy and FΩz are the body forces in x, y
and z direction respectively. In the above equation, the integration is carried out over the entire
surface in the first term and over the entire volume in the second term. The above expression can be
rewritten as:
WE    d F d    d F d
T T

  (2.1.2)

Here, d  u v w .
T
For the three dimensional stress-strain condition, there are six
components of stresses ( x ,  y , z ,  xy ,  yz ,  zx ) and six components of strains in virtual

displacement fields (  x ,  y ,  z ,  xy ,  yz ,  zx ). Therefore, the virtual internal work can be

expressed as follows:
 2

 x 
 
  y 
  
U    x  y  z  xy  yz  zx  z d
 xy  (2.1.3)

 
 yz 
 
 zx 
Or
U     d
T

 (2.1.4)
According to principle of virtual work, the work done by external forces due to the virtual
displacement of a structure in equilibrium is equal to the work done by the internal forces for the
virtual internal displacement. Therefore, WE  U Thus eqs. (2.1.2) and (2.1.4) can be made equal
and can be related as follows:

  d F d    d F d     d

T T T

   (2.1.5)

2.1.3 Variational Principle

Variational formulation is the generalized method of formulating the element stiffness matrix and
load vector using the variational principle of solid mechanics. The strain energy in a structural body
is given by the relation
1
= ∫∫∫ {ε } {σ } d Ω
T
U
2 Ω (2.1.6)
  x , y , z ,  xy ,  yz ,  zx  .
T
For a 3D structural problem, stress has six components:

   x ,  y ,  z ,  xy ,  yz ,  zx  . Now the strain-

T
Similarly, there are six components of strains:

displacement relationship can be expressed as {ε } = [ B ]{d } , where {d} is the displacement vector in
x, y and z directions and [B] is called as the strain displacement relationship matrix. Again, the
stress can be represented in terms of its constitutive relationship matrix: {σ } = [ D ]{ε } . Here [ D ]
is called as the constituent relationship matrix. Using the above relationship in the strain energy
equation one can arrive
1
[ B ]{d } [ D ]{ B}{d } d Ω
T
U ∫∫∫
2 Ω (2.1.7)
Applying the variational principle one can express
 3

∂U
{F }
= = ∫∫∫ [ B ] [ D ][ B ] d Ω {d }
T

∂ {d } Ω (2.1.8)
Now, from the relationship of { F } = [ K ]{d } , one can arrive at the element stiffness matrix as:

=[K ] ∫∫∫ [ B ] [ D ][ B ] d Ω
T

Ω (2.1.9)

Thus, by the use of variational principle, the stiffness matrix of a structural element can be obtained
as expressed in the above equation.

2.1.4 Weighted Residual Method

Virtual work and Variational method are applicable and adequate for most of the problems.
However, in some cases functional analogous to potential energy cannot be written because of not
having clear physical meaning. For some applications, such as in fluid mechanics problem,
functional needed for a variational approach cannot be expressed. For some types of fluid flow
problems, only differential equations and boundary conditions are available. For Such problems
weighted residual method can be used for obtaining the solutions. Approximate solutions of
differential equation satisfy only part of conditions of the problem. For example a differential
equation may be satisfied only at few points, rather than at each. The strategy used in weighted
residual method is to first take an approximate solution and then its validity is assessed. The
different methods in weighted Residual Method are
• Collocation method
• Least square method
• Method of moment
• Galerkin method
The mathematical statement of a physical problem can be defined as:
In domain Ω,
Du  f  0 (2.1.10)
Where,
D is the differential operator
u = u(x) = dependent variables such as displacement, pressure, velocity,
potential function
x = independent variables such as coordinates of a point
f = a function of x which may be constant or zero

If u is an approximate solution then residual in domain Ω,

R  Du  f (2.1.11)
According to the weighted residual method, the weak form of above equation will become
 4

w i R d  0 for i 1,2,3,...,n


or (2.1.12)

 w Du  f d  0
i

Where weighting function wi = wi(x) is chosen from the approximate basis function used for
constructing approximated solution u .
 5

Lecture 2: Galerkin Method

2.2.1 Introduction
Galerkin method is the most widely used among the various weighted residual methods. Galerkin
method incorporates differential equations in their weak form, i.e., before starting integration by
parts it is in strong form and after by parts it will be in weak form, so that they are satisfied over a
domain in an integral. Thus, in case of Galerkin method, the equations are satisfied over a domain in
an integral or average sense, rather than at every point. The solution of the equations must satisfy the
boundary conditions. There are two types of boundary conditions:
• Essential or kinematic boundary condition
• Non essential or natural boundary condition
4y
For example, in case of a beam problem ( EI  q  0 ) differential equation is of forth order.
x 4
As a result, displacement and slope will be essential boundary condition where as moment and shear
will be non-essential boundary condition.

2.2.2 Galerkin Method for 2D Elasticity Problem

For a two dimensional elasticity problem, equation of equilibrium can be expressed as
x  xy
  Fx  0 (2.2.1)
x y

 xy  y
  Fy  0 (2.2.2)
x y
Where, Fx and Fy are the body forces in X and Y direction respectively. Let assume,

x and y are surface forces in X and Y direction and  as angle made by normal to surface
with X– axis (Fig. 2.2.1). Therefore, force equilibrium of element can be written as:
Fx PQ t  x OP t   xy OQ t
OP OQ
Fx  x   xy  x cos    xy sin   x cos    xyCos 90  
PQ PQ
Thus, Fx  x    xy m (2.2.3)
Where, ℓ and m are direction cosines of normal to the surface. Similarly,
Fy   xy    y m (2.2.4)
 6

Fig. 2.2.1 Elemental stress in 2D

Adopting Galerkin’s approach

     
  x   xy  F u    xy   y  F v dxdy  0
   x x    x y x   (2.2.5)
 y  

Where u and v are weighting functions i.e elemental displacements in X and Y directions
respectively. Now one can expand above equation by using Green’s Theorem.
Green Theorem states that if  x, yand  x, y are continuous functions then their first and
second partial derivatives are also continuous. Therefore,
        2  2     
   dxdy     2  2 dxdy    
 x x y y   x
  y 
 
  x

 y
m ds

(2.2.6)

 
Assuming,   x ;  u;  0 one can rewrite with the use of above relation as
x y
x  u 
 x
u dx dy   x
x
dx dy   x  u ds (2.2.7)

 
Similarly, assuming    y ;  0 and  v
x y
 y  v
 y
v dx dy    y
y
dx dy    y m v ds (2.2.8)
 7

 
Again, assuming    x y ;  v; 0
x y

 x y  v
 y
v dx dy    x y
x
dx dy    x y  v ds (2.2.9)

 
And assuming,    x y ;  0;  u
x y
 x y  u 
 y
u dx dy    x y
y
dx dy    x y m u ds

Putting values of eqs. (2.2.7), (2.2.8) and (2.2.9), in eq. (2.2.5), one can get the following relation:

     
 x u    y v   xy v   xy u dx dy
 x y x y 

 x u   y mv   xy v   xy mu  ds   Fx u dx dy   Fyv dx dy  0

(2.2.10)
Rearranging the terms of above expression, the following relations are obtained.
     
 x u    y v   xy v   xy u dx dy   Fx u  Fyv dx dy
 x y x y 

 x    xy muds    xy    y mvds  0 (2.2.11)

Here, Fx and Fy are the body forces and u & v are virtual displacements in X and Y directions
respectively.

Considering first term of eq. (2.2.11), virtual displacement u is given to the element of unit
thickness. Dotted position in Fig. 2.2.2 shows the virtual displacement. Thus, work done by x :
   
x dy u  u dx   x dyu  x u dxdy (2.2.12)
 x  x

Similarly, considering second term of eq. (2.2.11), virtual work done by body forces is

 F x u  Fyvdx dy
 8

Putting eqs. (2.2.3) & (2.2.4) in third term of eq. (2.2.11) we get the virtual work done by surface
forces as:

F x uds   Fyvds

Fig. 2.2.2 Element subjected to stresses

Due to virtual displacement u , change in strain  x is given by:

  
u  u dx   u
 x  
 x    u  (2.2.13)
dx x
The virtual work done by x is x . x .dxdy . Similarly all the individual term in the first term of
eq. (2.2.11) can be derived from eq. (2.2.13) which will be as follows:


 x x
u dxdy   x x dxdy

 y y vdxdy   y y dxdy (2.2.14)



   


 xy  x
  v  
y
u     xy xy dxdy




Now, the work done by internal forces will be

 9

U   x  x  y y  xy xy  dxdy (2.2.15)

If external work done is represented by WE and U is the internal work done then,
U  w E  0 or U  w E (2.2.16)
Thus in elasticity problems, Galerkin’s method turns out to be the principle of virtual work, which
can be stated that “A Deformable body is said to be in equilibrium, if the total work done by external
forces is equal to the total work done by internal forces.” The work done above is virtual as either
forces or deformations are also virtual. Thus, Galerkin’s approach can be followed in all problems
involving solution of a set of equations subjected to specified boundary values.

2.2.3 Galerkin Method for 2D Fluid Flow Problem

Let consider the two dimensional incompressible fluid equation which can be expressed by pressure
variable only as follows.
2 p  0 (2.2.17)
Where p is the pressure inside the fluid domain. The above equation can be expressed in 2D form as:
 2p  2p
 0
x 2 y 2
or (2.2.18)
p,ii  0
Applying weighted residual method, the weak form of the above equation will become

w i p,ii d  0 (2.2.19)


Integrating by parts of the above expression, the following relation can be obtained.

w i p,i d   w i,i p,i d   0

 

or w i,i p,i d    w i p,i d (2.2.20)

 

If the nodal pressure and interpolation functions are denoted by p and N respectively, then the
pressure at any point inside the fluid domain can be expressed as
p   N p
Similarly, the weighted function can also be written with the help of interpolation function as
w   N w 
 10

 
Thus, pi,i   L p   L  N p   Bp , where,  L     = differential operator.
 x y 
Similarly, w i,i   L W   L  N w    Bw 
Thus,  w i,i p,i d     w   B  B p d
T T
(2.2.21)


p
 w i p,i d =  w   N  d
T T
(2.2.22)
 
n
Here, Γdenotes the surface of the fluid domain and n represents the direction normal to the surface.
Thus, from eq. (2.2.20), one can write the expression as:
p
Thus,  w   B  Bp d   w   N  d
T T T T

 
n
Or, G p  S (2.2.23)

Where,
 T   T  
G     B  Bd     N   N    N   N d
T

 
 x x y y 
(2.2.27)
p
and S   N  d
T


n
Here, n is the direction normal to the surface. Thus, solving the above equation with the prescribed
boundary conditions, one can find out the pressure distribution inside the fluid domain by the use of
finite element technique.
 11

Lecture 3: Finite Element Method: Displacement Approach

2.3.1 Choice of Displacement Function

Displacement function is the beginning point for the structural analysis by finite element method.
This function represents the variation of the displacement within the element. On the basis of the
problem to be solved, the displacement function needs to be approximated in the form of either
linear or higher-order function. A convenient way to express it is by the use of polynomial
expressions.

2.3.1.1 Convergence criteria

The convergence of the finite element solution can be achieved if the following three conditions are
fulfilled by the assumed displacement function.

a. The displacement function must be continuous within the elements. This can be ensured by
choosing a suitable polynomial. For example, for an n degrees of polynomial, displacement
function in I dimensional problem can be chosen as:
u=α 0 + α1 x + α 2 x 2 + α 3 x3 + α 4 x 4 + ..... + α n x n (2.3.1)
b. The displacement function must be capable of rigid body displacements of the element. The
constant terms used in the polynomial (α0 to αn) ensure this condition.
c. The displacement function must include the constant strains states of the element. As
element becomes infinitely small, strain should be constant in the element. Hence, the
displacement function should include terms for representing constant strain states.

2.3.1.2 Compatibility
Displacement should be compatible between adjacent elements. There should not be any
discontinuity or overlapping while deformed. The adjacent elements must deform without causing
openings, overlaps or discontinuous between the elements.
Elements which satisfy all the three convergence requirements and compatibility condition
are called Compatible or Conforming elements.

2.3.1.3 Geometric invariance

Displacement shape should not change with a change in local coordinate system. This can be
achieved if polynomial is balanced in case all terms cannot be completed. This ‘balanced’
representation can be achieved with the help of Pascal triangle in case of two-dimensional
polynomial. For example, for a polynomial having four terms, the invariance can be obtained if the
following expression is selected from the Pascal triangle.
u =α 0 + α1 x + α 2 y + α 3 xy (2.3.2)
 12

The geometric invariance can be ensured by the selection of the corresponding order of terms on
either side of the axis of symmetry.
1

x y

x2 xy y2

x3 x2y xy2 y3

x4 x3y x2y2 xy3 y4

Fig. 2.3.1 Pascal’s Triangle

2.3.2 Shape Function

In finite element analysis, the variations of displacement within an element are expressed by its
nodal displacement ( u = ∑ N i ui ) with the help of interpolation function since the true variation of
displacement inside the element is not known. Here, u is the displacement at any point inside the
element and ui are the nodal displacements. This interpolating function is generally a polynomial
with n degree which automatically provides a single-valued and continuous field. In finite element
literature, this interpolation function (Ni) is referred to “Shape function” as well. For linear
interpolation, n will be 1 and for quadratic interpolation n will become 2 and so on. There are two
types of interpolation functions namely (i) Lagrange interpolation and (ii) Hermitian interpolation.
Lagrange interpolation function is widely used in practice. Here the assumed function takes on the
same values as the given function at specified points. In case of Hermitian interpolation function, the
slopes of the function also take the same values as the given function at specified points. The
derivation of shape function for varieties of elements will be discussed in subsequent lectures.

2.3.3 Degree of Continuity

Let consider φ as an interpolation function in a piecewise fashion over finite element mesh. While
such interpolation function φ can be ensured to vary smoothly within the element, the transition
between adjacent elements may not be smooth. The term Cm is considered to define the continuity of
a piecewise displacement. A function Cm is continuous if its derivative up to and including degree m
are inter-element continuous. For example, for one dimensional problem, φ= φ(x) is C0 continuous if
φ is continuous, but φ,x is not. Similarly, φ= φ(x) is C1 continuous if φ and φ,x are continuous, but φ,xx
is not. In general, C0 element is used to model plane and solid body and C1 element is used to model
beam, plate and shell like structure, where inter-element continuity of slope is necessary to ensure.
Let assume a linear function for bar like element: φ= 1 α 0 + α1 x This function is C0 continuous as
 13

φ1,x is discontinuous. If the interpolation function is considered as φ2 =α 0 + α1 x + α 2 x 2 then

φ2,=
x α1 + 2α 2 x is also continuous but φ2, xx = 2α 2 is discontinuous. As a result, this function φ2 will
become C1 continuous.

2.3.4 Isoparametric Elements

If the shape functions (Ni) used to represent the variation of geometry of the element are the same as
the shape functions (N´i) used to represent the variation of the displacement then the elements are
called isoparametric elements. For example, the coordinates (x,y) inside the element are defined by
the shape functions (Ni) and displacement (u,v) inside the element are defined by the shape functions
(N´i) as below.

x  Ni x i u  N i u i (2.3.3)

y  N i yi v  N vi i

If Ni = N´i, then the element is called isroparametric. Fig. 2.3.2(a) shows the two dimensional 8 node
isoparametric element.

If the geometry of element is defined by shape functions of order higher than that for representing
the variation of displacements, then the elements are called superparametric (Fig. 2.3.2(b)).

If the geometry of element is defined by shape functions of order lower than that for representing the
variation of displacements then the elements are called subparametric (Fig. 2.3.2(c)).

Fig. 2.3.2 Shape functions for geometry and displacements

 14

2.3.5 Various Elements

Selection of the order of the polynomial depends on the type of elements. For example, in case of
one dimensional element having single degrees of freedom with two nodes, the displacement
function can be chosen as = u α 0 + α1 x . However, if the same has two degrees of freedom at each
node, then the chosen displacement function should be u =α 0 + α1 x + α 2 x 2 + α 3 x3 . Various types of
elements used in finite element analysis are given below:

1. One dimensional elements.

(a) Two node element
(b) Three node element

Fig. 2.3.3 One dimensional elements

2. Two dimensional elements

(a) Triangular element
(b) Rectangular element
(c) Quadrilateral element
(d) Quadrilateral formed by two triangles
(e) Quadrilateral formed by four triangles

Few of the elements with number of nodes are shown in Fig. 2.3.4.
 15

Fig. 2.3.4 Two dimensional elements

3. Three dimensional elements.

(a) Tetrahedron
(b) Rectangular brick element
Few of the three dimensional solid elements are shown in Fig. 2.3.5.
 16

Fig. 2.3.5 Three dimensional elements

 17

Lecture 4: Stiffness Matrix and Boundary Conditions

2.4.1 Element Stiffness Matrix

The stiffness matrix of a structural system can be derived by various methods like variational
principle, Galerkin method etc. The derivation of an element stiffness matrix has already been
discussed in earlier lecture. The stiffness matrix is an inherent property of the structure. Element
stiffness is obtained with respect to its axes and then transformed this stiffness to structure axes. The
properties of stiffness matrix are as follows:
• Stiffness matrix is symmetric and square.
• In stiffness matrix, all diagonal elements are positive.
• Stiffness matrix is be positive definite

2.4.2 Global Stiffness Matrix

A structural system is an assemblage of number of elements. These elements are interconnected
together to form the whole structure. Therefore, the element stiffness of all the elements are first
need to be calculated and then assembled together in systematic manner. It may be noted that the
stiffness at a joint is obtained by adding the stiffness of all elements meeting at that joint.
To start with, the degrees of freedom of the structure are numbered first. This numbering will
start from 1 to n where n is the total degrees of freedom. These numberings are referred to as degrees
of freedom corresponding to global degrees of freedom. The element stiffness matrix of each
element should be placed in their proper position in the overall stiffness matrix. The following steps
may be performed to calculate the global stiffness matrix of the whole structure.
a. Initialize global stiffness matrix [K ] as zero. The size of global stiffness matrix will be equal
to the total degrees of freedom of the structure.
b. Compute individual element properties and calculate local stiffness matrix [ k ] of that
element.
c. Add local stiffness matrix [ k ] to global stiffness matrix [K ] using proper locations
d. Repeat the Step b. and c. till all local stiffness matrices are placed globally.

The steps to be followed in the computer program are shown in the form of flow chart in Fig. 2.4.1
for assembling the local stiffness matrix to global stiffness matrix.
 18

Fig. 2.4.1 Assemble of stiffness matrix from local to global

 19

2.4.3 Boundary Conditions

Under this section, procedure to include the effect of boundary condition in the stiffness matrix for
the finite element analysis will be discussed. The solution cannot be obtained unless support
conditions are included in the stiffness matrix. This is because, if all the nodes of the structure are
included in displacement vector, the stiffness matrix becomes singular and cannot be solved if the
structure is not supported amply, and it cannot resist the applied loads. A solution cannot be
achieved until the boundary conditions i.e., the known displacements are introduced.
In finite element analysis, the partitioning of the global matrix is carried out in a systematic
way for the hand calculation as well as for the development of computer codes. In partitioning,
normally the equilibrium equations can be partitioned by rearranging corresponding rows and
columns, so that prescribed displacements are grouped together. For example, let consider the
equation of equilibrium is expressed in compact form as:
{F } = [ K ]{d } (2.4.1)
Where,
[K] is the global stiffness matrix,
{d} is the displacement vector consisting of global degrees of freedom, and
{F} is the load vector corresponding to degrees of freedom.
By the method of partitioning the above equation can be partitioned in the following manner.
{ Fα }  [ Kαα ]  Kα β   {dα }
 =   (2.4.2)
{ F }   K   K 
 β    βα   ββ    β  {d }
Where, subscripts α refers to the displacements free to move and β refers to the prescribed support
displacements. As the prescribed displacements {dβ} are known, eq. (2.4.2) may be written in
expanded form as:
{Fα }= [ Kαα ]{dα }+  Kαβ  {d β } (2.4.3)
Thus it is possible to obtain the free displacement of the structure {dα} as
{dα } = [ Kαα ] {{Fα } - }
 Kαβ  {d β }
-1
(2.4.4)
If the displacements at supports {dβ} are zero, then the above equation can be simplified to the
following expression.
{dα } = [ Kαα ] {Fα }
-1
(2.4.5)
Thus, by rearranging assembled matrix, the portion corresponding to the unknown displacements in
eq. (2.4.4) can be taken out for the solution purpose. This is possible as the known displacements
{dβ} are restrained, i.e., displacements are zero. If the support has some known displacements, then
eq. (2.4.4) can be used to find the solution. If the few supports of the structures yield, then the above
method may be modified by partitioning the stiffness matrix into three parts as shown below:
 20

F   K    K    K    d 

          
F     K   K  

 K    d  

          
 
F    K   K    K    d  
              (2.4.6)

Here, α refers to unknown displacement; β refers to known displacement (≠0) and γ refers to zero
displacement. Thus, the above equation can be separated and solved independently to find required
unknown results as shown below.
F    K  d    K   d     K   d  
or,  K  d   F    K   d   as d    0
Thus, d    K  
1
F   K  d 
  
(2.4.7)
For computer programming, several techniques are available for handling boundary conditions. One
of the approaches is to make the diagonal element of stiffness matrix corresponding to zero
displacement as unity and corresponding all off-diagonal elements as zero. For example, let consider
a 3×3 stiffness matrix with following force-displacement relationship.
 F1   k11 k12 k13   d1 
    
 F2  =  k21 k22 k23  d 2 
 F  k k33   d3  (2.4.8)
 3   31 k32
Now, if the third node has zero displacement (i.e., d3= 0) then the matrix will be modified as follows
to incorporate the boundary condition.
 F1   k11 k12 0   d1 
    
 F2  =  k21 k22 0  d 2 
0  0 1   d3  (2.4.9)
   0

Thus, while inverting whole matrix, d3 will become zero automatically.

To incorporate known support displacement in computer programming following procedure may be

adopted. Considering the displacement d2 has known value of δ, 1st row of eq. (2.4.8) can be written
as:
F1 = k11 × d1 + k12 × d 2 + k13 × d3
(2.4.10)
Or
F1 − k12 × δ = k11 × d1 + k13 × d3 (2.4.11)

Now the 2nd row of eq. (2.4.8) has to become:

 21

{δ } = {d 2 } (2.4.12)
Similarly 3rd row will be:
F3 − k32 × δ = k31 × d1 + k33 × d3 (2.4.13)
Thus above three equations can be written in a combined form as
 F1 − k12δ   k11 0 k13   d1 
    
 δ  =  0 1 0  d 2 
 F − δ  k  
 32   31 0 k33   d3  (2.4.14)
Another approach may also be followed to take care the known restrained displacements by
assigning a higher value δ (say δ =1020) in the diagonal element corresponding to that displacement.
 F1   k11 k12 k13   d1 
  k  
δ ×10 × k=
20
22   21 k22 ×1020 k23  d 2 
   k31 k33   d3 
 F3  k32
(2.4.15)
 1020  k 22  k 21d1  k 22 1020  d 2  k 23  d 3
As d3 is corresponding to zero displacement, the above equation can be simplified to the following.
 1020  k 22  k 21d1  k 22 1020  d 2
or 1020  k 22  k 22 1020  d 2
 d 2   known displacement is ensured

If the overall stiffness matrix is to be formed in half band form then the numbering of nodes should
be such that the bandwidth is minimum. For this the labels are put in a systematic manner
irrespective of whether the joint displacements are unknowns or restraints. However, if the unknown
displacements are labeled first then the matrix operations can be restricted up to unknown
displacement labels and beyond that the overall stiffness matrix may be ignored.
 1

Lecture 1: Natural Coordinates

Natural coordinate system is basically a local coordinate system which allows the specification of a
point within the element by a set of dimensionless numbers whose magnitude never exceeds unity.
This coordinate system is found to be very effective in formulating the element properties in finite
element formulation. This system is defined in such that the magnitude at nodal points will have
unity or zero or a convenient set of fractions. It also facilitates the integration to calculate element
stiffness.

3.1.1 One Dimensional Line Elements

The line elements are used to represent spring, truss, beam like members for the finite element
analysis purpose. Such elements are quite useful in analyzing truss, cable and frame structures. Such
structures tend to be well defined in terms of the number and type of elements used. For example, to
represent a truss member, a two node linear element is sufficient to get accurate results. However,
three node line elements will be more suitable in case of analysis of cable structure to capture the
nonlinear effects. The natural coordinate system for one dimensional line element with two nodes is
shown in Fig. 3.1.1. Here, the natural coordinates of any point P can be defined as follows.
x x
N1  1 and N 2  (3.1.1)
l l
Where, x is represented in Cartesian coordinate system. Similarly, x/l can be represented as ξ in
natural coordinate system. Thus the above expression can be rewritten in the form of natural
coordinate system as given below.
N1  1 x and N 2  x (3.1.2)
Now, the relationship between natural and Cartesian coordinates can be expressed from eq. (3.1.1) as
1 1 1  N1 
      (3.1.3)
 x 0 l   N 2 
Here, N1 and N2 is termed as shape function as well. The variation of the magnitude of two linear
shape functions (N1 and N2) over the length of bar element are shown in Fig. 3.1.2. This example
displays the simplest form of interpolation function. The linear interpolation used for field variable φ
can be written as
f x   f1 N1  f2 N 2 (3.1.4)
 2

Fig. 3.1.1 Two node line element

Fig. 3.1.2 Linear interpolation function for two node line element

Similarly, for three node line element, the shape function can be derived with the help of natural
coordinate system which may be expressed as follows:
 3


 3x 2 x 2  

1  2  
 l  
 N1 
   l  1  3  2 2 

    2 
 
 
 (3.1.5)

   2 
    4x 4x
       
  
2
N N 4 4

 
 
 l l 2

 
 


 
 N  
  x 2x 
 
  2  2


3

2
  

  


 l l2 

The detailed derivation of the interpolation function will be discussed in subsequent lecture. The
variation of the shape functions over the length of the three node element are shown in Fig. 3.1.3

Fig. 3.1.3 Variation of interpolation function for three node line element

Now, if φ is considered to be a function of L1 and L2, the differentiation of φ with respect to x for
two node line element can be expressed by the chain rule formula as
d  L1  L 2
 .  . (3.1.6)
dx L1 x L 2 x
Thus, eq.(3.1.4) can be written as
L1 1 L2 1
  and  (3.1.7)
x l x l
Therefore,
 4

d 1    
   
dx l  L2 L1 
(3.1.7)

The integration over the length l in natural coordinate system can be expressed by
p !q !
L L2 q dl 
p
l (3.1.9)
l
1
 p  q 1!
Here, p! is the factorial product p(p-1)(p-2)….(1) and 0! is defined as equal to unity.

3.1.2 Two Dimensional Triangular Elements

The natural coordinate system for a triangular element is generally called as triangular coordinate
system. The coordinate of any point P inside the triangle is x,y in Cartesian coordinate system. Here,
three coordinates, L1, L2 and L3 can be used to define the location of the point in terms of natural
coordinate system. The point P can be defined by the following set of area coordinates:
𝐴1 𝐴2 𝐴3
𝐿1 = ; 𝐿2 = ; 𝐿3 = (3.1.10)
𝐴 𝐴 𝐴
Where,
𝐴1 = Area of the triangle P23
𝐴2 = Area of the triangle P13
𝐴3 = Area of the triangle P12
A=Area of the triangle 123
Thus,
𝐴 = 𝐴1 + 𝐴2 + 𝐴3
and
𝐿1 + 𝐿2 + 𝐿3 = 1 (3.1.11)
Therefore, the natural coordinate of three nodes will be: node 1 (1,0,0); node 2 (0,1,0); and node 3
(0,0,1).
 5

Fig. 3.1.4 Triangular coordinate system

The area of the triangles can be written using Cartesian coordinates considering x, y as coordinates
of an arbitrary point P inside or on the boundaries of the element:

1 𝑥1 𝑦1
1
A = 2
�1 𝑥2 𝑦2 �
1 𝑥3 𝑦3
1 𝑥 𝑦
1
A1 = 2
�1 𝑥2 𝑦2 �
1 𝑥3 𝑦3
1 𝑥 𝑦
1
A2 = � 1 𝑥3 𝑦3 �
2
1 𝑥1 𝑦1
1 𝑥 𝑦
1
A3 = 2
�1 𝑥1 𝑦1 �
1 𝑥2 𝑦2

The relation between two coordinate systems to define point P can be established by their nodal
coordinates as
1 1 1 1 𝐿1
𝑥
� �=� 1𝑥 𝑥2 𝑥3 � �𝐿2 � (3.1.12)
𝑦 𝑦1 𝑦2 𝑦3 𝐿3
Where,
𝑥 = 𝐿1 𝑥1 + 𝐿2 𝑥2 + 𝐿3 𝑥3
 6

𝑦 = 𝐿1 𝑦1 + 𝐿2 𝑦2 + 𝐿3 𝑦3
The inverse between natural and Cartesian coordinates from eq.(3.1.12) may be expressed as
𝐿1 𝑥2 𝑦3 −𝑥3 𝑦2 𝑦2 − 𝑦3 𝑥3 − 𝑥2 1
1
�𝐿2 � = 2𝐴 �𝑥3 𝑦1 −𝑥1 𝑦3 𝑦3 − 𝑦1 𝑥1 − 𝑥3 � �𝑥 � (3.1.13)
𝐿3 𝑥1 𝑦2 −𝑥2 𝑦1 𝑦1 − 𝑦2 𝑥2 − 𝑥1 𝑦
The derivatives with respect to global coordinates are necessary to determine the properties of an
element. The relationship between two coordinate systems may be computed by using the chain rule
of partial differentiation as
  L1  L 2  L 3
 .  .  .
x L1 x L 2 x L3 x
b1  b  b 
 .  2.  3.
2A L1 2A L 2 2A L3 (3.1.14)
3
bi 
 .
i1 2A Li

Where, b1 = y2 – y3; b2 = y3 – y1 and b3 = y1 – y2. Similarly, following relation can be obtained.

 3
c 
 i .
y i1 2A Li (3.1.15)

Where, c1 = x3 – x2; c2 = x1 – x3 and c3 = x2 – x1. The above expressions are looked cumbersome.
However, the main advantage is the ease with which polynomial terms can be integrated using
following area integral expression.

p!q!r!
L L 2 q L3r dA 
p
2A (3.1.16)
A
1
p  q  r  2!

Where 0! is defined as unity.

3.1.3 Shape Function using Area Coordinates

The interpolation functions for the triangular element are algebraically complex if expressed in
Cartesian coordinates. Moreover, the integration required to obtain the element stiffness matrix is
quite cumbersome. This will be discussed in details in next lecture. The interpolation function and
subsequently the required integration can be obtained in a simplified manner by the concept of area
coordinates. Considering a linear displacement variation of a triangular element as shown in Fig.
3.1.5, the displacement at any point can be written in terms of its area coordinates.
 7

u  1L1   2 L 2   3L3
u    
T
Or, (3.1.17)

where,    L1 L3  and    1  2  3 

T T
L2
𝐴1 𝐴2 𝐴3
And 𝐿1 = ; 𝐿2 = ; 𝐿3 = (3.1.18)
𝐴 𝐴 𝐴
Here, A is the total area of the triangle. It is important to note that the area coordinates are dependent
as 𝐿1 + 𝐿2 + 𝐿3 = 1 . It may be seen from figure that at node 1, L1 = 1 while L2 = L3 = 0. Similarly
for other two nodes: at node 2, L2 = 1 while L1 = L3 = 0, and L3 = 1 while L2 = L1 = 0. Now,
substituting the area coordinates for node 1, 2 and 3, the displacement components at nodes can be
written as
 u1  1 0 0
   
u i   u 2   0 1 0   (3. 1.19)
 
u 3  0 0 1
Thus, from the above expression, one can obtain the unknown coefficient  :
1 0 0  u1 
   
   0 1 0 u 2 
 (3. 1.20)
 
 0 0 1 u 3 
  

Fig. 3.1.5 Area coordinates for triangular element

Now, eq.(3.1.17) can be written as:

 8

1 0 0  u1   1 0 0
    
 
T 
u   0 1 0 u 2    0 1 0 u i 
 T
 (3. 1.21)
 
0 0 1 u 3   0 0 1
     
The above expression can be written in terms of interpolation function as u  N u i 
T

Where,
 1 0 0
 
N   L1 L2 L3  0 1 0   L1 L 2 L3 
T
(3. 1.22)
 
 0 0 1
 
Similarly, the displacement variation v in Y direction can be expressed as follows.

v   N  v i 
T
(3.1.23)
Thus, for two displacement components u and v of any point inside the element can be written as:
 0  ui 
u  N 
T T

d  =   =  T T  
(3.1.24)
v  0  N   vi 
Thus, the shape function of the element will become
 L1 L2 L3 0 0 0
 N     (3.1.25)
0 0 0 L1 L2 L3 
It is important to note that the shape function Ni become unity at node i and zero at other nodes of
the element. The displacement at any point of the element can be expressed in terms of its nodal
displacement and the interpolation function as given below.
u = N1u1 + N 2u2 + N 3u3
(3.1.26)
v = N1v1 + N 2v2 + N 3v3
 9

Lecture 2: Triangular Elements

The triangular element can be used to represent the arbitrary geometry much easily. On the other
hand, rectangular elements, in general, are of limited use as they are not well suited for representing
curved boundaries. However, an assemblage of rectangular and triangular element with triangular
elements near the boundary can be very effective (Fig. 3.2.1). Triangular elements may also be used
in 3-dimensional axi-symmetric problems, plates and shell structures. The shape function for
triangular elements (linear, quadratic and cubic) with various nodes (Fig. 3.2.2) can be formulated.
An internal node will exist for cubic element as seen in Fig. 3.2.2(c).

Fig. 3.2.1 Finite element mesh consisting of triangular and rectangular element

Fig. 3.2.2 Triangular elements

In displacement formulation, it is very important to approximate the variation of displacement in the

element by suitable function. The interpolation function can be derived either using the Cartesian
coordinate system or by the area coordinates.
 10

3.2.1 Shape function using Cartesian coordinates

Polynomials are easiest way of mathematical operation for expressing variation of displacement. For
example, the displacement variation within the element can be represented by the following function
in case of two dimensional plane stress/strain problems.
u = α⍺0 + α⍺1 x + α⍺2 y
(3.2.1)
v = α⍺3 + α⍺4 x + ⍺α5 y
(3.2.2)
where α0, α1, α2 ….. are unknown coefficients. Thus the displacement vectors at any point P, in the
element (Fig.3.2.3) can be expressed with the following relation.
𝛼0
⎧𝛼1 ⎫
𝑢 1 𝑥 𝑦 0 0 0 ⎪𝛼2 ⎪
{𝑑} = � � = � � (3.2.3)
𝑣 0 0 0 1 𝑥 𝑦 ⎨𝛼3 ⎬
⎪𝛼4 ⎪
⎩𝛼5 ⎭
Or, {d}=[ 𝜙]{ α } (3.2.4)

Fig. 3.2.3 Triangular element in Cartesian Coordinates

Similarly, for “m” node element having three degrees of freedom at each node, the displacement
function can be expressed as
𝑢 = 𝛼0 + 𝛼1 𝑥 + 𝛼2 𝑦 + 𝛼3 𝑥 2 + 𝛼4 𝑥𝑦 + 𝛼5 𝑦 2 + . . … . . . +𝛼𝑚−1 𝑦 𝑛
𝑣 = 𝛼𝑚 + 𝛼𝑚+1 𝑥 + 𝛼𝑚+2 𝑦 + 𝛼𝑚+3 𝑥 2 + 𝛼𝑚+4 𝑥𝑦+. . . . . +𝛼2𝑚−1 𝑦 𝑛 (3.2.5)
2 𝑛
𝑤 = 𝛼2𝑚 + 𝛼2𝑚+1 𝑥 + 𝛼2𝑚+2 𝑦 + 𝛼2𝑚+3 𝑥 + 𝛼2𝑚+4 𝑥𝑦+. . . . . +𝛼3𝑚−1 𝑦
 11

Hence, in such case,

𝑢 {ϕ}𝑇 0 0
{𝑑} = � 𝑣 � = � 0 𝑇
{ϕ} 0 � {𝛼} (3.2.6)
𝑤 0 0 {ϕ}𝑇
Where, {𝛼}𝑇 = [𝛼0 𝛼1 . . . 𝛼3𝑚−1 ] and, [𝜙]𝑇 = [1 𝑥 𝑦 𝑥 2 𝑥𝑦 . . . . 𝑦 𝑛 ]
Now, for a linear triangular element with 2 degrees of freedom, eq. (3.2.3) can be written in terms of
the nodal displacements as follows.
𝑢1 1 𝑥1 𝑦1 0 0 0 𝛼0
⎧𝑢2 ⎫ ⎡1 𝑥2 𝑦2 0 0 0 ⎤ ⎧𝛼1 ⎫
⎪𝑢 ⎪ ⎢1 𝑥 𝑦 0 0 0 ⎥ ⎪𝛼 ⎪
{𝑑} = 𝑣3 = ⎢ 3 3 ⎥ 2 (3.2.7)
⎨ 1 ⎬ ⎢0 0 0 1 𝑥1 𝑦1 ⎥ ⎨𝛼3 ⎬
⎪𝑣2 ⎪ ⎢0 0 0 1 𝑥2 𝑦2 ⎥ ⎪𝛼4 ⎪
⎩𝑣3 ⎭ ⎣0 0 0 1 𝑥3 𝑦3 ⎦ ⎩𝛼5 ⎭

Where, {d} is the nodal displacements. To simplify the above expression for finding out the shape
function, the displacements in X direction can be separated out which will be as follows:

𝑢1 1 𝑥1 𝑦1 ⍺∝0
{𝑢𝑖 }=� 2 � = �1 𝑥2 𝑦2 � �⍺∝1 �
𝑢 (3.2.8)
𝑢3 1 𝑥3 𝑦3 ⍺∝2
To obtain the polynomial coefficients, {α} the matrix of the above equation are to be inverted. Thus,
1
 0  1 x1 y1 
 u1   x 2 y3  x 3 y 2 x 3 y1  x1 y3 x1 y 2  x 2 y1  u1 
       
 y2 
1   1 x 2 u   1  y 2  y3
 y3  y1 y1  y 2  u 2 
    2  2A   
y3    u 
 2  1 x 3
  u 3   x3  x 2 x1  x 3 x 2  x1   3 
 a1 a2 a 3  u1 
1    
 b1 b2 b 3  u 2  (3.2.9)
2A   
 c1 c2 c3 u 3 
Where, A is the area of the triangle and can be obtained as follows.
1 𝑥1 𝑦1
1
A = 2 �1 𝑥2 𝑦2 � (3.2.10)
1 𝑥3 𝑦3
Now, eq. (3.2.1) can be written from the above polynomial coefficients.
 12

1 
u  x 2 y3  x 3 y 2    y 2  y3  x   x 3  x 2  y u1
2A 
1 
  x 3 y1  x1y3    y3  y1  x   x1  x 3  y u 2 (3.2.11)
2A 
1 
  x1y 2  x 2 y1    y1  y 2  x   x 2  x1  y u 2
2A 
Thus, the interpolation function can be obtained from the above as:

 

1 
           


  x 2 y 3 x 3 y 2 y 2 y 3 x x 3 x 2 y 

 N1 
   2A 

     
N   N 2   
1  x 3 y1  x1y3    y3  y1  x   x1  x 3  y  
      (3.2.12)

 
 

2A 

 N
   1
3  

              

x
 1 2 y x y
2 1 y1 y 2 x x 2 x 1 y 
 2A
 


Such three node triangular element is commonly known as constant strain triangle (CST) as its strain
is assumed to be constant inside the element. This property may be derived from eq. (3.2.1) and
eq.(3.2.2). For example, in case of 2-D plane stress/strain problem, one can express the strain inside
the triangle with the help of eq.(3.2.1) and eq.(3.2.2):
u  ( 0  1x   2 y)
x    1
x x
v  ( 3   4 x   5 y)
y    5 (3.2.13)
y y
v u
 xy    2  2
x y
CST is the simplest element to develop mathematically. As there is no variation of strain inside the
element, the mesh size of the triangular element should be small enough to get correct results. This
element produces constant temperature gradients ensuring constant heat flow within the element for
heat transfer problems.

3.2.2 Higher Order Triangular Elements

Higher order elements are useful if the boundary of the geometry is curve in nature. For curved case,
higher order triangular element can be suited effectively while generating the finite element mesh.
Moreover, in case of flexural action in the member, higher order elements can produce more
accurate results compare to those using linear elements. Various types of higher order triangular
 13

elements are used in practice. However, most commonly used triangular element is the six node
element for which development of shape functions are explained below.

3.2.2.1 Shape function for six node element

Fig. 3.2.4 shows a triangular element with six nodes. The additional three nodes (4, 5, and 6) are
situated at the midpoints of the sides of the element. A complete polynomial representation of the
field variable can be expressed with the help of Pascal triangle:
φ ( x, y ) =α 0 + α1 x + α 2 y + α 3 x 2 + α 4 xy + α 5 y 2
(3.2.14)

Fig. 3.2.4 (a) Six node triangular element (b) Lines of constant values of the area coordinates

Using the above field variable function, one can reach the following expression using interpolation
function and the nodal values.
6
 x, y   N i  x, yi (3.2.15)
i1

Here, the every shape function must be such that its value will be unity if evaluated at its related
node and zero if evaluated at any of the other five nodes. Moreover, as the field variable
representation is quadratic, each interpolation function will also become quadratic. Fig. 3.2.4(a)
shows the six node element with node numbering convention along with the area coordinates at three
corners. The six node element with lines of constant values of the area coordinates passing through
the nodes is shown in Fig. 3.2.4(b). Now the interpolation functions can be constructed with the help
of area coordinates from the above diagram. For example, the interpolation function N1 should be
unity at node 1 and zero at all other five nodes. According to the above diagram, the value of L1 is 1
 14

at node 1 and ½ at node 4 and 6. Again, L1 will be 0 at nodes 2, 3 and 5. To satisfy all these
conditions, one can propose following expression:
 1
N1  x, y  N1 L1 ,L 2 ,L3   L1 L1  
 2 (3.2.16)

Evaluating the above expression, the value of N1 is becoming ½ at node 1 though it must become
unity. Therefore, the above expression is slightly modified satisfying all the conditions and will be as
follows:
 1
N1  2L1 L1    L1 2L1 1
 2 (3.2.17)

Eq. (3.2.17) assures the required conditions at all the six nodes and is a quadratic function, as L1 is a
linear function of x and y. The remaining five interpolation functions can also be obtained in similar
fashion applying the required nodal conditions. Thus, the shape function for the six node triangle
element can be written as given below.
N1  L1 2L1 1
N 2  L 2 2L 2 1
N 3  L3 2L3 1 (3.2.18)
N 4  4L1L 2
N 5  4L 2 L3
N 6  4L3L1
Such six node triangular element is commonly known as linear strain triangle (LST) as its strain is
assumed to vary linearly inside the element. In case of 2-D plane stress/strain problem, the element
displacement field for such quadratic triangle may be expressed as
u ( x, y ) =α 0 + α1 x + α 2 y + α 3 x 2 + α 4 xy + α 5 y 2
v ( x, y ) =α 6 + α 7 x + α 8 y + α 9 x 2 + α10 xy + α11 y 2 (3.2.19)

So the element strain can be derived from the above displacement field as follows.
u
x   1  2 3 x   4 y
x
v
y    8  10 x  211y (3.2.20)
y
v u
 xy     2   4 x  2 5 y   7  2 9 x  10 y
x y
The above expression shows that the strain components are linearly varying inside the element.
Therefore, this six node element is called linear strain triangle. The main advantage of this element is
that it can capture the variation of strains and therefore stresses of the element.
 15

3.2.3 Construction of Shape Function by Degrading Technique

Sometimes, the geometry of the structure or its loading and boundary conditions are such that the
stresses developed in few locations are quite high. On the other hand, variations of stresses are less
in some areas and as a result, refinement of finite element mesh is not necessary. It would be
economical in terms of computation if higher order elements are chosen where stress concentration
is high and lower order elements at area away from the critical area. Fig. 3.2.5 shows graphical
representations where various order of triangular elements are used for generating a finite element
mesh.

Fig. 3.2.5 Triangular elements with different number of nodes

Fig. 3.2.5 contains four types of element. Type 1 has only three nodes, type 2 element has five
nodes, type 3 has four nodes and type 4 has six nodes. The shape function for 3-node and 6-node
triangular elements has already been derived. The shape functions of 6-node element can suitably be
degraded to derive shape functions of other two types of triangular elements.

3.2.3.1 Five node triangular element

Let consider a six node triangular element as shown in Fig. 3.2.6(a) whose shape functions and nodal
displacements are (N1, N2, N3, N4, N5, N6) and (u1, u2, u3, u4, u5, u6) respectively. Similarly, for a five
node triangular element as shown in Fig. 3.2.6(b), the shape functions and nodal displacements are
considered as (N’1, N’2, N’3, N’4, N’5) and (u’1, u’2, u’3, u’4, u’5) respectively. Thus, the displacement
at any point in a six node triangular element will become
 16

u  N1u1  N 2 u 2  N 3u 3  N 4 u 4  N 5u 5  N 6 u 6 (3.2.21)
Where, N1, N2, …, N6 are the shape functions and is given in eq.(3.2.18). If there is no node between
2 and 3, the displacement along line 2-3 is considered to vary linearly. Thus the displacement at an
assumed node 5´ may be written as
u 2  u3
u '5 
2 (3.2.22)
Substituting, the value of u’5 for u5 in eq.(3.2.21) the following expression will be obtained.
u 2  u3
u  N1u1  N 2 u 2  N 3u 3  N 4 u 4  N 5  N6u 5 (3.2.23)
2

Fig. 3.2.6 Degrading for five node element

Thus, the displacement function can be expressed by five nodal displacements as:
 N   N 
u  N1u1   N 2  5  u 2   N 3  5  u 3  N 4 u 4  N 6 u 5
 2  2 (3.2.24)

However, the displacement function for the five node triangular element can be expressed as
u  N1u1  N 2 u 2  N 3 u 3  N 4 u 4  N 5 u 5 (3.2.25)
Comparing eq.(3.2.24) and eq.(3.2.25) and observing node 6 of six node triangle corresponds to
node 5 of five node triangle, we can write
N5 N
N '1  N1, N 2  N 2  , N 3  N 3  5 , N 4  N 4 an dN 5  N 5
2 2 (3.2.26)
Hence, the shape function of a five node triangular element will be
 17

N1  L1 2L1 1

4L 2 L3
N 2  L 2 2L 2 1   L 2 1 2L1 
2
N 4L L
N 3  N 3  5  L3 2L3 1  2 3  L3 1 2L1 
2 2 (3.2.27)
N 4  4L1L 2
N 5  4L3L1
Thus, for a five node triangular element, the above shape function can be used for finite element
analysis.
 18

Lecture 3: Rectangular Elements

Rectangular elements are suitable for modelling regular geometries. Sometimes, it is used along with
triangular elements to represent an arbitrary geometry. The simplest element in the rectangular
family is the four node rectangle with sides parallel to x and y axis. Fig. 3.3.1 shows rectangular
elements with varying nodes representing linear, quadratic and cubic variation of function.

Fig. 3.3.1 Rectangular elements

3.3.1 Shape Function for Four Node Element

Shape functions of a rectangular element can be derived using both Cartesian and natural coordinate
systems. A four term polynomial expression for the field variable will be required for a rectangular
element with four nodes having four degrees of freedom. Since there is no complete four term
polynomial in two dimensions, the incomplete, symmetric expression from the Pascal’s triangle may
be chosen to ensure geometric isotropy.

3.3.1.1 Shape function using Cartesian coordinates

For the derivation of interpolation function, the sides of the rectangular element (Fig. 3.3.2) are
assumed to be parallel to the global Cartesian axes. From the Pascal’s triangle, a linear variation may
be assumed to define filed variable to ensure inter-element continuity.
f  x, y   aaaa
0  1 x  2 y  3 xy (3.3.1)
 19

Fig. 3.3.2 Rectangular element in Cartesian coordinate

Applying nodal conditions, the above expression may be written in matrix form as
f1  1 x1 y1 x1 y1  a0 
     
f2  1 x2 y2 x2 y2  a1 
      (3.3.2)
f3  1 x3 y3 x3 y3  a2 
    
f4  1 x4 y4 x4 y4  a3 
The unknown polynomial coefficients may be obtained from the above equation with the use of
nodal field variables.
1
a0 
  1 x1 y1 x1 y1  
 f1 

    
a1  1 x2 y2 x2 y2 

f2  

      
   (3.3.3)

a  1
  x3 y3 x3 y3  
f3 

a  
f 
2
 1 x4 y4  
 3
   x4 y4  4
 
Thus, the field variable at any point inside the element can be described in terms of nodal values as
1
a0  1 x1 y1 x1 y1  f1 
     
a 1 x2 y2 
f  x, y   1 x y xy  1   1 x y xy  
x2 y2
 f2 
a2 
1 x3 y3 x3 y3  f3 
  1  
a3   x4 y4 x4 y4  f4 
(3.3.4)
f1 
 
f 
  N1 N2 N3 N 4  2 
f3 
 
f4 
From the above expression, the shape function Ni can be derived and will be as follows.
 20

 x − x2   y − y4 
N1 =   
 x1 − x2   y1 − y4 
 x − x1   y − y3 
N2 =   
 x2 − x1   y2 − y3 
(3.3.5)
 x − x4   y − y2 
N3 =   
 x3 − x4   y3 − y2 
 x − x3   y − y1 
N4 =   
 x4 − x3   y4 − y1 
Now, substituting the nodal coordinates in terms of (x1, y1) as (–a, –b) at node 1; (x2, y2) as (a, –b) at
node 2; (x3, y3) as (a, b) at node 3 and (x4, y4) as (–a, b) at node 4 the above expression can be re-
written as:
1
N1 = ( x − a )( y − b )
4ab
1
N2 = ( x + a )( y − b )
4ab
(3.3.6)
1
N3 = ( x + a )( y + b )
4ab
1
N4 = ( x − a )( y + b )
4ab
Thus, the shape function N can be found from the above expression in Cartesian coordinate system.

3.3.1.2 Shape function using natural coordinates

The derivation of interpolation function in terms of Cartesian coordinate system is algebraically
complex as seen from earlier section. However, the complexity can be reduced by the use of natural
coordinate system, where the natural coordinates will vary from -1 to +1 in place of –a to +a or –b to
+b. The transformation of Cartesian coordinates to Natural coordinates are shown in Fig. 3.3.3.
 21

Fig. 3.3.3 Four node rectangular element

From the figure, the relation between two coordinate systems can be expressed as
x−x y− y
=ξ = and η (3.3.7)
a b
Here, 2a and 2b are the width and height of the rectangle. The coordinate of the center of the
rectangle can be written as follows:
x1 + x2 y1 + y4
=x = and y (3.3.8)
2 2
Thus, from eq. (3.3.7) and eq.(3.3.8), the nodal values in natural coordinate systems can be derived
which is shown in Fig. 3.3.4(b). With the above relations variations of x & h will be from -1 to +1.
Now the interpolation function can be derived in a similar fashion as done in section 3.3.1.1. The
filed variable can be written in natural coordinate system ensuring inter-element continuity as:
f x, h  aa
0  1x  a2 h  a3xh (3.3.9)
The coordinates of four nodes of the element in two different systems are shown in Table 3.3.1 for
ready reference for the derivation purpose. Applying the nodal values in the above expression one
can get
f1  1 1 1 1  a0 
    
 
f2   1 1 1 1 a1  (3.3.10)
f3  1 1 1 1  a2 
     
f4  1 1 1 1 a3 
 22

Table 3.3.1 Cartesian and natural coordinates for four node element

Node Cartesian Coordinate Natural Coordinate

x y ξ η
1 x1 y1 -1 -1
2 x2 y2 1 -1
3 x3 y3 1 1
4 x4 y4 -1 1

Thus, the unknown polynomial coefficients can be found as

  1 1 1 1  1 
a0  f1 
 1 1 1 1  f1 
 
      
    

  1 1 1 1 
a1    1 1 1
f2  1 1 

f2 
        (3.3.11)
a2 
  1 1
 1 1    4 1 1
f3 
 1 1  
f3 

 
   
 
  1 1   
a
 3 1 1 1 1  f
 4  1 1 f4 



The field variable can be written as follows using eq.(3.3.9) and eq.(3.3.11).
a0  1 1 1 1  f1 
    
a  1 1 1 1 1 f2 
f x, h  1 x h xh 1   1 x h xh   
a2  4 1 1 1 1  f3 
   1 1  
a3   1 1 f4 
(3.3.12)
f1 
 
f 
  N1 N 2 N 3 N 4  2 
f3 
 
f4 
Where, Ni are the interpolation function of the element in natural coordinate system and can be
found as:


 1 x1 h 

 

 4 

  
 



N 
  1  x 1  h  
 
1
 
N2   
   4 
Ni       (3.3.13)

 N3 
 
 1  x1  h

 
 
 


N4  
 4 



 1 x1  h

 


 4 

 23

3.3.2 Shape Function for Eight Node Element

The shape function of eight node rectangular element can be derived in similar fashion as done in
case of four node element. The only difference will be on choosing of polynomial as this element is
of quadratic in nature. The derivation will be algebraically complex in case of using Cartesian
coordinate system. However, use of the natural coordinate system will make the process simpler as
the natural coordinates vary from -1 to +1 in the element. The variation of filed variable ϕ can be
expressed in natural coordinate system by the following polynomial.
f x, h  aa
0  1x  a2 h  a3x  a4 xh  a5 h  a6 x h  a7 xh
2 2 2 2
(3.3.14)
It may be noted that the cubic terms ξ3 and η3 are omitted and geometric invariance is ensured by
choosing the above expression. Fig. 3.3.4 shows the natural nodal coordinates of the eight node
rectangle element in natural coordinate system.

The nodal field variables can be obtained from the above expression after putting the coordinates at
nodes.
f1  1 1 1 1 1 1 1 1 a0 
    
f2  1 1 1 1 1 1 1 1  a1 
 
f  1 1
 3   1 1 1 1 1 1  a2 
f  1 1 1  
1 1 1 1 1 a3 
fi    4        Aai  (3.3.15)
f5  1 0 1 0 0 1 0 0  a4 
    
f6  1 1 0 1 0 0 0 0  a5 
     
f7  1 0 1 0 0 1 0 0  a6 
  
f  1 1 0
 8   1 0 0 0 0  a7 

Fig. 3.3.4 Natural coordinates of eight node rectangular element

Replacing the unknown coefficient αi in eq.(3.3.14) from eq.(3.3.15), the following relations will be
obtained.
 24

f x, h  1 x h x 2 xh h2 x 2 h xh2   A fi 

1

1 1 1 1 2 2 2 2  f1  
  

0 0 0 0 0 2 0 2  f2 
  
0
 0 0 0 2 0 2 0  

f3 
  
2 1  1 1 1 1 2 0 2 0   f4  

 1 x h x xh h x h xh  
2 2 2
 

0 
4  1 1 1 1 0 0 0 f5  
  
1 1 1 1 
0 2 0 2 f6  
  

1 1 1 1 2 0 2 0   f7

  
 
1 1 1 1 0 2 0 f8 
2  
 

f1 
 
f2 
 
f3 
 
f
  N1 N2 N3 N4 N5 N6 N7 N8  4 
f5 
 
f6 
 
f7 
f 
 8  (3.3.16)
Thus, the interpolation function will become

1 x1 hx  h 1 1  x1 hx  h 1

N1  ; N2 ;
4 4
1  x1  hx  h 1 1 x1  hx  h 1
N3  ; N4  ;
4 4
1  x1 x1 h 1  x1  h1 h
N5  ; N6  ;
2 2
1  x1 x1  h 1 x1  h1 h
N7  ; N8 
2 2 (3.3.17)

The shape functions of rectangular elements with higher nodes can be derived in similar manner
using appropriate polynomial satisfying all necessary criteria. However, difficulty arises due to the
inversion of large size of the matrix because of higher degree of polynomial chosen. In next lecture,
the shape functions of rectangular element with higher nodes will be derived in a much simpler way.
 25

Lecture 4: Lagrange and Serendipity Elements

In last lecture note, the interpolation functions are derived on the basis of assumed polynomial from
Pascal’s triangle for the filed variable. As seen, the inverse of the large matrix is quite cumbersome
if the element is of higher order.

3.4.1 Lagrange Interpolation Function

An alternate and simpler way to derive shape functions is to use Lagrange interpolation polynomials.
This method is suitable to derive shape function for elements having higher order of nodes. The
Lagrange interpolation function at node i is defined by
n
   j     1   
2 ....  i1    i1 ....   n 
f i ( )   
j1  i   j  i  
1  i   i1  i  
2 .... i   i1 .... i   n 
ji
(3.4.1)
The function fi (ξ) produces the Lagrange interpolation function for i node, and ξj denotes ξth

coordinate of jth node in the element. In the above equation if we put ξ = ξj, and j ≠ i, the value of the
function fi(ξ) will be equal to zero. Similarly, putting ξ = ξi, the numerator will be equal to
denominator and hence fi(ξ) will have a value of unity. Since, Lagrange interpolation function for ith
node includes product of all terms except jth term; for an element with n nodes, fi(ξ) will have n-1
degrees of freedom. Thus, for one-dimensional elements with n-nodes we can define shape function
as N i ()  f i () .

3.4.1.1 Shape function for two node bar element

Consider the two node bar element discussed as in section 3.1.1. Let us consider the natural
coordinate of the center of the element as 0, and the natural coordinate of the nodes 1 and 2 are -1
and +1 respectively. Therefore, the natural coordinate ξ at any point x can be represented by,

2  x  x1 
 1
l (3.4.2)

Fig. 3.4.1 Natural coordinates of bar element

 26

The shape function for two node bar element as shown in Fig. 3.4.1 can be derived from eq.(3.4.1) as
follows:

  2  1 1
N1  f1     1  
1   2  1  (1) 2
  1   1 1 (3.4.3)
N 2  f1     1  
 2  1  1 (1) 2

Graphically, these shape functions are represented in Fig.3.4.2.

Fig. 3.4.2 Shape functions for two node bar element

3.4.1.2 Shape function for three node bar element

For a three node bar element as shown in Fig. 3.4.3, the shape function will be quadratic in nature.
These can be derived in the similar fashion using eq.(3.4.1) which will be as follows:

2   
   3 ( )  1 1
N1   f1     1
 
2  1   3 
1   1(2) 2

1   
   3   1 1
N 2    f 2      1   2  (3.4.4)
 2  
1  2   3  1 1

  
1   2   1() 1
N 3   f3      1
 
3  
1  3   2  2(1) 2
 27

Fig. 3.4.3 Quadratic shape functions for three node bar element

3.4.1.3 Shape function for two dimensional elements

We can derive the Lagrange interpolation function for two or three dimensional elements from one
dimensional element as discussed above. Those elements whose shape functions are derived from
the products of one dimensional Lagrange interpolation functions are called Lagrange elements. The
Lagrange interpolation function for a rectangular element can be obtained from the product of
appropriate interpolation functions in the ξ direction [fi(ξ )] and η direction [fi(η)]. Thus,
N i 
,   f i 
fi   Where, i = 1,2,3, …., n-node (3.4.5)
The procedure is described in details in following examples.

Four node rectangular element

The shape functions for the four node rectangular element as shown in the Fig. 3.4.4 can be derived
by applying eq.(3.4.3) eq.(3.4.5) which will be as follows.
2    2 
  
N1 
,   f1 
f1   
` 2   1  2 
1   (3.4.6)
 1  1
1
   1  1  
1  (1) 1  (1) 4

Similarly, other interpolation functions can be derived which are given below.
1
N 2  f1    1  1  
,   f 2 
4
1
` N 3 
,   f 2 
f 2    1  1   (3.4.7)
4
1
N 4 
,   f1 
f 2    1  1  
4
These shape functions are exactly same as eq.(3.3.13) which was derived earlier by choosing
polynomials.
 28

Fig. 3.4.4 Four node rectangular element

Nine node rectangular element

In a similar way, to the derivation of four node rectangular element, we can derive the shape
functions for a nine node rectangular element. In this case, the shape functions can be derived using
eq.(3.4.4) and eq.(3.4.5).
1 1 1
N1 
,   f1 
f1      1  
 1   1 1
2 2 4 (3.4.8)
In a similar way, all the other shape functions of the element can be derived. The shape functions of
nine node rectangular element will be:
1 1
N1    1 1, N 2     1 1
4 4
1 1
N 3    1  1, N 4    1  1
4 4
N 5   1     11  2 
1 1
2
 1, N 6   (3.4.9)
2 2
N 7   1    11 2 
1 1
2
  1, N8  
2 2
N 9  1   2 1   2 
 29

Fig. 3.4.5 Nine node rectangular element

Thus, it is observed that the two dimensional Lagrange element contains internal nodes (Fig. 3.4.6)
which are not connected to other nodes.

Fig. 3.4.6 Two dimensional Lagrange elements and Pascal triangle

3.4.2 Serendipity Elements

Higher order Lagrange elements contains internal nodes, which do not contribute to the inter-
element connectivity. However, these can be eliminated by condensation procedure which needs
extra computation. The elimination of these internal nodes results in reduction in size of the element
matrices. Alternatively, one can develop shape functions of two dimensional elements which contain
nodes only on the boundaries. These elements are called serendipity elements (Fig. 3.4.7) and their
interpolation functions can be derived by inspection or the procedure described in previous lecture
 30

(Module 3, lecture 3). The interpolation function can be derived by inspection in terms of natural
coordinate system as follows:
(a) Linear element
1
N i 
,   1  i 1  i  (3.4.10)
4
(b) Quadratic element
(i) For nodes at   1,   1
1
N i 
,   1  i 1  
i  i  i  1 (3.4.11a)
4
(ii) For nodes at   1,   0

,   1  i 1  2 
1
N i  (3.4.11b)
2
(iii) For nodes at   0,   1

1   2 1  i 
1
N i 
,  (3.4.11c)
2
(c) Cubic element
(i) For nodes at   1,   1

i  9 
N i 
, 
1
1  i 1    2  2  10
32  (3.4.12a)
1
(ii) For nodes at   1,   
3

1  i 1  2 1  9i 

9
N i 
,  (3.4.12b)
32
And so on for other nodes at the boundaries.
 31

Fig. 3.4.7 Two dimensional serendipity elements and Pascal triangle

Thus, the nodal conditions must be satisfied by each interpolation function to obtain the functions
serendipitously. For example, let us consider an eight node element as shown in Fig. 3.4.8 to derive
its shape function. The interpolation function N1 must become zero at all nodes except node 1, where
its value must be unity. Similarly, at nodes 2, 3, and 6, ξ = 1, so including the term ξ − 1 satisfies the
zero condition at those nodes. Similarly, at nodes 3, 4 and 7, η = 1 so the term η − 1 ensures the zero
condition at these nodes.

Fig. 3.4.8 Two dimensional eight node rectangular element

Again, at node 5, (ξ, η) = (0, −1), and at node 8, (ξ, η) = (−1, 0). Hence, at nodes 5 and 8, the term (ξ
+ η + 1) is zero. Using this reasoning, the equation of lines are expressed in Fig. 3.4.9. Thus, the
 32

interpolation function associated with node 1 is to be of the form N1  y1 h 1x 1x  h  1

where, ψ1 is unknown constant. As the value of N1 is 1 at node 1, the magnitude unknown constant
ψ1 will become -1/4. Therefore, the shape function for node 1 will become
1
N1   1 h1 x x  h  1 .
4
Similarly, ψ2 will become -1/4 considering the value of N2 at node 2 as unity and the shape function
1
for node 2 will be N 2  y2 h 1x  1x  h 1   1  x1 hx  h 1 . In a similar
4
fashion one can find out other interpolation functions from Fig. 3.4.9 by putting the respective values
at various nodes. Thus, the shape function for 8-node rectangular element is given below.

Fig. 3.4.9 Equations of lines for two dimensional eight node element

N 5  1   2 1  ,
1 1
N1   1  1  1    ,
4 2
N 6  1  1  2 ,
1 1
N 2   1  1  1    ,
4 2
(3.4.13)
N 7  1   2 1  ,
1 1
N 3   1  1  1    ,
4 2
N 4   1  1  1     and N8  1  1  2 
1 1
4 2
 33

It may be observed that the Lagrange elements have a better degree of completeness in polynomial
function compare to serendipity elements. Therefore, Lagrange elements produce comparatively
faster and better accuracy.
 34

Lecture 5: Solid Elements

There are two basic families of three-dimensional elements similar to two-dimensional case.
Extension of triangular elements will produce tetrahedrons in three dimensions. Similarly,
rectangular parallelepipeds are generated on the extension of rectangular elements. Fig. 3.5.1 shows
few commonly used solid elements for finite element analysis.

Fig. 3.5.1 Three-dimensional solid elements

Derivation of shape functions for such three dimensional elements in Cartesian coordinates are
algebraically quite cumbersome. This is observed while developing shape functions in two
dimensions. Therefore, the shape functions for the two basic elements of the tetrahedral and
parallelepipeds families will be derived using natural coordinates.

The polynomial expression of the field variable in three dimensions must be complete or incomplete
but symmetric to satisfy the geometric isotropy requirements. Completeness and symmetry can be
 35

ensured using the Pascal pyramid which is shown in Fig. 3.5.2. It is important to note that each
independent variable must be of equal strength in the polynomial.

Fig. 3.5.2 Pascal pyramid in three dimensions

The following 3-D quadratic polynomial with complete terms can be applied to an element having
10 nodes.
f x, h, z   aa
0  1x  a2 h  a3z  a4 x  a5 h  a6z  a7 xh  a8 hz  a9zx
2 2 2
(3.5.1)
However, the geometric isotropy is not an absolute requirement for field variable representation to
derive the shape functions.

3.5.1 Tetrahedral Elements

The simplest element of the tetrahedral family is a four node tetrahedron as shown in Fig. 3.5.3. The
node numbering has been followed in sequential manner, i.e, in this case anti-clockwise direction.
Similar to the area coordinates, the concept of volume coordinates has been introduced here. The
coordinates of the nodes are defined both in Cartesian and volume coordinates. Point P(x, y, and z)
as shown in Fig. 3.5.2 is an arbitrary point in the tetrahedron.
 36

Fig. 3.5.3 Four node tetrahedron element

The linear shape function for this element can be expressed as,
{N} = [ L1 L4 ]
T
L2 L3
(3.5.2)
Here, L1 , L2 , L3 , L4 are the set of natural coordinates inside the tetrahedron and are defined as follows
Vi
Li =
V (3.5.3)
Where Vi is the volume of the sub element which is bound by point P and face i and V is the total
volume of the element. For example L1 may be interpreted as the ratio of the volume of the sub
element P234 to the total volume of the element 1234. The volume of the element V is given by the
determinant of the nodal coordinates as follows:
1 1 1 1
1 x1 x2 x3 x4
V=
6 y1 y2 y3 y4 (3.5.4)
z1 z2 z3 z4
The relationship between the Cartesian and natural coordinates of point P may be expressed as

1   1 1 1 1   L1 
x  x x4   L2 
   1 x2 x3
 =  
 y   y1 y 4   L3 
y2 y3 (3.5.5)
 z   z1 z2 z3

z 4   L4 
It may be noted that the identity included in the first row ensure the matrix invertible.
 37

L1 + L2 + L3 + L4 = 1
(3.5.6)
The inverse relation is given by
 L1  V1 a1 b1 c1  1 
L  
 2  1 V2 a2 b2 c2   x 
 =  
 L3  6V V3 a3 b3 c3   y  (3.5.7)
 
 L4  V4 a4 b4 c4   z 

Here, Vi is the volume subtended from face i and terms ai , bi , and ci represent the projected area of
face i on the x, y, z coordinate planes respectively and are given as follows:
ai = ( z j y k − z k y j ) + ( z k y l − z l y k ) + ( z l y j − z j y l )
bi = ( z j xk − z k x j ) + ( z k xl − z l xk ) + ( z l x j − z j xl )
(3.5.8)
ci = ( y j x k − y k x j ) + ( y k xl − y l x k ) + ( y l x j − y j xl )
i, j , k , l will be in cyclic order (i.e., 1 2  3  4  1). The volume coordinates fulfil all nodal
conditions for interpolation functions. Therefore, the field variable can be expressed in terms of
nodal values as
f  x, y, z   L1f1  L2f2  L3f3  L4f4
(3.5.9)
Though the shape functions (i.e., the volume coordinates) in terms of global coordinates is
algebraically complex but they are straightforward. The partial derivatives of the natural coordinates
with respect to the Cartesian coordinates are given by
∂Li ai ∂Li b ∂Li c
= , = i , = i
∂x 6V ∂y 6V ∂z 6V (3.5.10)
Similar to area integral, the general integral taken over the volume of the element is given by,
p!q!r!s!
∫ L L L L dV = ( p + q + r + s + 3)!.6V
p q r s
1 2 3 4
(3.5.11)
v

The four node tetrahedral element is a linear function of the Cartesian coordinates. Hence, all the
first partial derivatives of the field variable will be constant. The tetrahedral element is a constant
strain element as the element exhibits constant gradients of the field variable in the coordinate
directions.
Higher order elements of the tetrahedral family are shown in Fig. 3.5.1. The shape functions for such
higher order three dimensional elements can readily be derived in volume coordinates, as for higher-
order two-dimensional triangular elements. The second element of this family has 10 nodes and a
cubic form for the field variable and interpolation functions.
+++++++++++
3.5.2 Brick Elements
 38

Various orders of elements of the parallelepiped family are shown in Fig. 3.5.1. Fig. 3.5.4 shows the
eight-node brick element with reference to a global Cartesian coordinate system and then with
reference to natural coordinate system. The natural coordinates for the brick element can be relate
Cartesian coordinate system by
x−x y− y z−z
=ξ = , η = and ζ (3.5.12)
a b c
Here, 2a, 2b and 2c are the length, height and width of the element. The coordinate of the center of
the brick element can be written as follows:
x1 + x2 y1 + y4 z1 + z5
=x = , y = and z (3.5.13)
2 2 2
Thus, from eq.(3.5.12) and eq.(3.5.13), the nodal values in natural coordinate systems can be derived
which is shown in Fig. 3.5.4(b). With the above relations variations of x, h & z will be from -1 to +1.
Now the interpolation function can be derived in several procedures as done in case of two
dimensional rectangular elements. For example, the interpolation function can be derived by
inspection in terms of natural coordinate system as follows:
1
N i 
, ,   1  i 1  i 1  i  (3.5.14)
8

Fig. 3.5.4 Eight node brick element

By using field variable the following terms of the polynomial may be used for deriving the shape
function for eight-node brick element.
f x, h, z   aa
0  1x  a2 h  a3z  a4 xh  a5 hz  a6zx  a7zhx
(3.5.14)
The above equation is incomplete but symmetric. However, such representations are quite often used
and solution convergence is achieved in the finite element analysis. Again, the shape functions for
 39

three dimensional 8-node or 27-node brick elements can be derived using Lagrange interpolation
function. For this we need to introduce interpolation function in the ζ-direction. Thus, for example,
the Lagrange interpolation function for a three dimensional 8 node brick element can be obtained
from the product of appropriate interpolation functions in the ξ, η and ζ directions. Therefore, the
shape function will become
N i 
, ,   f i 
fi  fi   Where, i = 1,2,3, …., n-node (3.5.15)
Thus using the Lagrange interpolation function the shape function at node 1 can be expressed as
2    
   2    2 
N1 
, ,   f1 
f1  f1   
2   1  
1   2   1  2  (3.5.16)
 1  1  1 1
    1  1  1  
1  (1) 1  (1) 1  (1) 4
Using any of the above concepts, the interpolation function for 8-node brick element can be found as
follows:

1 1
N1  1  1  1  , N1  1  1  1  ,
4 4
1 1
N 3  1  1  1  , N 4  1  1  1  ,
4 4 (3.5.17)
1 1
N 5  1  1  1  , N 6  1  1  1  ,
4 4
1 1
N 7  1  1  1  , N8  1  1  1  
4 4

The shape functions of rectangular parallelepiped elements with higher nodes can be derived in
similar manner satisfying all necessary criteria.
 40

Lecture 6: Isoparametric Formulation

3.6.1 Necessity of Isoparametric Formulation

The two or three dimensional elements discussed till now are of regular geometry (e.g. triangular
and rectangular element) having straight edge. Hence, for the analysis of any irregular geometry, it is
difficult to use such elements directly. For example, the continuum having curve boundary as shown
in the Fig. 3.6.1(a) has been discretized into a mesh of finite elements in three ways as shown.

(a) The Continuum to be discritized (b) Discritization using Triangular Elements (c)
Discritization using rectangular elements (d) Discritization using a combination of
rectangular and quadrilateral elements
 41

Fig 3.6.1 Discretization of a continuum using different elements

Figure 3.6.1(b) presents a possible mesh using triangular elements. Though, triangular elements can
suitable approximate the circular boundary of the continuum, but the elements close to the center
becomes slender and hence affect the accuracy of finite element solutions. One possible solution to
the problem is to reduce the height of each row of elements as we approach to the center. But,
unnecessary refining of the continuum generates relatively large number of elements and thus
increases computation time. Alternatively, when meshing is done using rectangular elements as
shown in Fig 3.6.1(c), the area of continuum excluded from the finite element model is significantly
adequate to provide incorrect results. In order to improve the accuracy of the result one can generate
mesh using very small elements. But, this will significantly increase the computation time. Another
possible way is to use a combination of both rectangular and triangular elements as discussed in
section 3.2. But such types of combination may not provide the best solution in terms of accuracy,
since different order polynomials are used to represent the field variables for different types of
elements. Also the triangular elements may be slender and thus can affect the accuracy. In Fig.
3.6.1(d), the same continuum is discritized with rectangular elements near center and with four-node
quadrilateral elements near boundary. This four-node quadrilateral element can be derived from
rectangular elements using the concept of mapping. Using the concept of mapping regular triangular,
rectangular or solid elements in natural coordinate system (known as parent element) can be
transformed into global Cartesian coordinate system having arbitrary shapes (with curved edge or
surfaces). Fig. 3.6.2 shows the parent elements in natural coordinate system and the mapped
elements in global Cartesian system.
42
 43

(a) Natural Coordinate System (b) Global Coordinate System

Fig. 3.6.2 Mapping of isoparametric elements in global coordinate system

3.6.2 Coordinate Transformation

The geometry of an element may be expressed in terms of the interpolation functions as follows.
n
x= N1 x1 + N 2 x2 + ... + N n x=
n ∑N x
i =1
i i

n
(3.6.1)
y= N1 y1 + N 2 y2 + ... + N n y=
n ∑N y
i =1
i i

n
z= N1 z1 + N 2 z2 + ... + N n z=
n ∑N z
i =1
i i

Where,
n=No.of Nodes
N i =Interpolation Functions
x i ,yi ,zi =Coordinates of Nodal Points of the Element
One can also express the field variable variation in the element as
 44

n
φ (ξ ,η , ζ ) = ∑ N i (ξ ,η , ζ ) φi (3.6.2)
i =1

As the same shape functions are used for both the field variable and description of element
geometry, the method is known as isoparametric mapping. The element defined by such a method is
known as an isoparametric element. This method can be used to transform the natural coordinates of
a point to the Cartesian coordinate system and vice versa.

Example 3.6.1
Determine the Cartesian coordinate of the point P (ξ= 0.8, η= 0.9) as shown in Fig. 3.6.3.

Fig. 3.6.3 Transformation of Coordinates

Solution:
As described above, the relation between two coordinate systems can be represented through their
interpolation functions. Therefore, the values of the interpolation function at point P will be
(1 − ξ )(1 − n) (1 − 0.8)(1 − 0.9)
=N1 = = 0.005
4 4
(1 + ξ )(1 − n) (1 + 0.8)(1 − 0.9)
=N2 = = 0.045
4 4
(1 + ξ )(1 + n) (1 + 0.8)(1 + 0.9)
=N3 = = 0.855
4 4
(1 − ξ )(1 + n) (1 − 0.8)(1 + 0.9)
=N4 = = 0.095
4 4

Thus the coordinate of point P in Cartesian coordinate system can be calculated as

 45

4
x = ∑ N i xi = 0.005 ×1 + 0.045 × 3 + 0.855 × 3.5 + 0.095 ×1.5 = 3.275
i =1
4
y = ∑ N i yi = 0.005 ×1 + 0.045 ×1.5 + 0.855 × 4.0 + 0.095 × 2.5 = 3.73
i =1

Thus the coordinate of point P (ξ= 0.8, η= 0.9) in Cartesian coordinate system will be 3.275, 3.73.

Solid isoparametric elements can easily be formulated by the extension of the procedure followed for
2-D elements. Regardless of the number of nodes or possible curvature of edges, the solid element is
just like a plane element which is mapped into the space of natural co-ordinates, i.e,
ξ = ±1,η = ±1, ζ = ±1 .

3.6.3 Concept of Jacobian Matrix

A variety of derivatives of the interpolation functions with respect to the global coordinates are
necessary to formulate the element stiffness matrices. As the both element geometry and variation of
the shape functions are represented in terms of the natural coordinates of the parent element, some
additional mathematical obstacle arises. For example, in case of evaluation of the strain vector, the
operator matrix is with respect to x and y, but the interpolation function is with x and h . Therefore,
the operator matrix is to be transformed for taking derivative with x and h . The relationship between
two coordinate systems may be computed by using the chain rule of partial differentiation as
  x   y   x  y
  and   (3.6.3)
x  x x  y x h  x h  y  h
The above equations can be expressed in matrix form as well.

   x y    
 

      

 x 
  x x   

x  
 x 

          J  
   (3.6.4)

  
 x y     
       
 

   h  
 y  
 y 
 h 
   h  
  

 x y 
  n
 x
The matrix [J] is denoted as Jacobian matrix which is: 
x 
 . As we know, x = ∑ N i xi
 x y  i =1
 
 h h 
n
∂ ∑ N i xi
∂x n
∂N i
where, n is the number of nodes in an element. Hence, J=
11 =
∂ξ
i =1

∂ξ
= ∑ ∂ξ
i =1
xi

Similarly one can calculate the other terms J12, J21 and J22 of the Jacobian matrix. Hence,
 46

 n N i n
N i 
 x  x yi 
 i1 x i 
 J    n i 1
 (3.6.5)
  N i xi
n
N i 
 i1 h 
i 1 h
yi 
 

From eq. (3.6.4), one can write


  
 
  
 


 
x   x 

 
     
 J
1 
(3.6.6)

   
  
 
 y  
 


 
  
 x 

 J11* J12* 
Considering  * are the elements of inverted [J] matrix, we may arise into the following
J * 
J 22
 21 
relations.
  
 J11*   J12* 
x x h
(3.6.7)
  
 J 21
*
  J 22
*

y x h

Similarly, for three dimensional case, the following relation exists between the derivative operators
in the global and the natural coordinate system.

 ∂   ∂x ∂y ∂z   ∂  ∂
 ∂ξ   ∂ξ ∂ξ ∂ξ     ∂x 
     ∂x   
 ∂   ∂x ∂y ∂z   ∂  ∂
=  =   [J ]  (3.6.8)
 ∂η   ∂η ∂η ∂η   ∂y   ∂y 
 ∂   ∂x ∂y ∂z   ∂  ∂
      
 ∂ζ   ∂ζ ∂ζ ∂ζ   ∂z   ∂z 

Where,
 47

 ∂x ∂y ∂z 
 ∂ξ ∂ξ ∂ξ 

 ∂x ∂y ∂z 
[J ] =  (3.6.9)
 ∂η ∂η ∂η 
 ∂x ∂y ∂z 
 ∂ζ ∂ζ ∂ζ 

[J] is known as the Jacobian Matrix for three dimensional case. Putting eq. (3.6.1) in eq. (3.6.9) and
after simplifying one can get
 ∂N i ∂N i ∂N i 
 ∂ξ xi ∂ξ
yi
∂ξ 
zi

n
 ∂N ∂N i ∂N i 
[ J ] = ∑  i xi (3.6.10)
∂η 
yi zi
i =1
 ∂η ∂η
 ∂N i ∂N i ∂N i 
 ∂ζ xi ∂ζ
yi
∂ζ 
zi


From eq. (3.6.8), one can find the following expression.

∂  ∂ 
 ∂x   
   ∂ξ 
∂ −1  ∂ 
  = [J ]   (3.6.11)
 ∂y   ∂η 
∂  ∂ 
   
 ∂z   ∂ζ 

 J11* J12* J13* 

 * * 
Considering [ J ] =  J 21
−1 *
J 22 J 23  we can arrived at the following relations.
 J 31
* *
J 32 * 
J 33 

   
 J11*   J12*   J13* 
x x h z
   
 J 21
*
  J 22
*
  J 23
*
 (3.6.12)
y x h z
   
 J 31
*
  J 32
*
  J 33
*

z x h z
 48

Lecture 7: Stiffness Matrix of Isoparametric Elements

3.7.1 Evaluation of Stiffness Matrix of 2-D Isoparametric Elements

For two dimensional plane stress/strain formulation, the strain vector can be represented as

 u   
 
 u u 


   J11*   J12*  


 x   
 
  x  h 

 
 ex 
  

   v     
   y   
ee 


*
J 21
v
  J 22 *

v 
 (3.7.1)

 
 
 
y   x h 

 g 
 xy 
     

  v u 
 
  v  v  u  u 
    J *
  J *
  J *
  J *
 

 
 x y 

  x h x h 
11 12 21 22
  
 


The above expression can be rewritten in matrix form


 u 

 


  x 


 

 J11* J12*   u 
 0 0    
*   h 
e    0   (3.7.2)
*
0 J 21 J 22  
 
J*
 21
*
J 22 J11* J12   v 
* 

  x 

 


  v 


 
 
  h 

n
For an n node element the displacement u can be represented as, u   N i ui and similarly for v &
i 1

w. Thus,


 u   N N n 

 
  1  0  0 
 x   x x  u1 
 

    

 u 
   N1 N n   

 h 
   h
 
h
0  0 
 un 



       
 (3.7.3)
   N1 N n  
 v1 
  
v 
   0  0  

 x   x x   

 
 
 

  
N n  



v 

  0  0
N1
  n
 v 

  h 
 h 
  
 h

As a result, eq. (3.7.2) can be written using eq. (3.7.3) which will be as follows.
 49

 N N n  u1 
 1  0  0   
    : 
   
 J11  N N n  : 
0   1
(3.7.4)
0   
* *
 J12 0  0 
 u n 
   0 0 J*21 J*22    
 
 J* *  N1 N n  v1 
J*22   0  
*
 21 J11 J12  0 
     : 
  
 N1 N n  : 
 0    
  v n 
0
 
 
Or,
   Bd (3.7.5)

Where {d} is the nodal displacement vector and [B] is known as strain displacement relationship
matrix and can be obtained as
 N1 N n 
  0 
0 
   
 
 J11
* *   N N n 

J12 0 0  1 
0  0  (3.7.6)
*    
 B   0 0 J 21 J 22   
*

 J* J* J* J*   0 N1 N n 
 21 22 11 12    0  
    
 
 N1 N n 
 0  0  
   
It is necessary to transform integrals from Cartesian to the natural coordinates as well for calculation
of the elemental stiffness matrix in isoparametric formulation. The differential area relationship can
be established from advanced calculus and the elemental area in Cartesian coordinate can be
represented in terms of area in natural coordinates as:
dA  dx dy  J d
d (3.7.7)

Here J is the determinant of the Jacobian matrix. The stiffness matrix for a two dimensional
element may be expressed as

 k     B  D Bd  t   B  D  Bdxdy (3.7.8)

T T

 A

Here, [B] is the strain-displacement relationship matrix and t is the thickness of the element. The
above expression in Cartesian coordinate system can be changed to the natural coordinate system as
follows to obtain the elemental stiffness matrix
 50

1 1

k   t   B DB J d

T
d (3.7.9)
1 1

Though the isoparametric formulation is mathematically straightforward, the algebraic difficulty is

significant.

Example 3.7.1:
Calculate the Jacobian matrix and the strain displacement matrix for four node two dimensional
quadrilateral elements corresponding to the gauss point (0.57735, 0.57735) as shown in Fig. 3.6.4.

Fig. 3.7.1 Two dimensional quadrilateral element

Solution:
The Jacobian matrix for a four node element is given by,
 n N i n 
N i
 x  x yi 
 i1 x i 
 J    n i 1

  N i xi
n
N i 
 i1 h 
i 1 h
yi 
 

For the four node element one can find the following relations.
1 1  N1 1   N1 1 
N1  ,  , 
4  4  4
1  1  N 2 1   N 2 1 
N2  ,  , 
4  4  4
1  1   N 3 1   N 3 1  
N3  ,  , 
4  4  4
 51

1 1   N 4 1   N 4 1  
N4  ,  , 
4  4  4
Now, for a four node quadrilateral element, the Jacobian matrix will become
 N1 N 2 y1 
N 3 N 4   x 1
  
   y2     x 2
 J     
N N 2 y3 
N 3
 N 4   x 3
 1 
   y 4 
    x 4
 1  1  1   1     x1 y1 

  
  x2 y2 
 4 4 4 4
 
 1  1   1   1     x 3 y3 

  
 4 4 4 4   x 4 y 4 

Putting the values of ξ & η as 0.57735 and 0.57735 respectively, one will obtain the following.
N1 N1
 0.10566  0.10566
 
N 2 N 2
 0.10566  0.39434
 
N 3 N 3
 0.39434  0.39434
 
N 4 N 4
 0.39434  0.10566
 
4
N i
Hence, J11   xi  0.105661  0.105663  0.394343.5  0.394341.5  1.0
i 1 x
Similarly, J12 =0.64632, J21 =0.25462 and J22 =1.14962.
Hence,
1.00000 0.64632
J  
 0.25462 1.14962 
Thus, the inverse of the Jacobian matrix will become:
 * J12*   1.1671 0.6561
 J *    J11  
   J * *   
 21 J 22  0.2585 1.0152 
Hence strain displacement matrix is given by,
 52

 N1 N n 
  0  0 
   
 
 J11
* *   N N n 

J12 0 0  1  0  0 
 
 B   0 0 J*21 J*22    

 J* *  N1 N n 
J*22   0 
*
 21 J11 J12  0 
    
 
 N1 N n 
 0  0  
   
 1.1671 0.6561 0 0 

  0 0 0.2585 1.0152  
0.2585 1.0152 1.1671 0.6561
 
0.10566 0.10566 0.39434 0.39434 0 0 0 0 
 
0.10566 0.39434 0.39434 0.10566 0 0 0 0 
 
   
 0 0 0 0 0.10566 0.10566 0.39434 0.39434 

 0 0 0 0 0.10566 0.39434 0.39434 0.10566 

0.0540 0.3820 0.2015 0.5294 0 0 0 0 

 
 0 0 0 0 0.0800 0.4276 0.2984 0.2092 
 
0.0800 0.4276 0.2984 0.2092 0.0540 0.3820 0.2015 0.5294
 

3.7.2 Evaluation of Stiffness Matrix of 3-D Isoparametric Elements

Stiffness matrix of 3-D solid isoparametric elements can easily be formulated by the extension of the
procedure followed for plane elements. For example, the eight node solid element is analogous to the
four node plane element. The strain vector for solid element can be written in the following form.
 53

 ∂u 
 ∂x 
 
 ∂u 
 ∂y 
 
 ∂u 
 ∂z 
 ε x  1 0 0 0 0 0 0 0 0   ∂v 
 ε  0  
 y  0 0 0 1 0 0 0 0   ∂x 
 ε z  0 0 0 0 0 0 0 0 1   ∂v 
 =  
γ xy  0 1 0 1 0 0 0 0 0   ∂y 
γ yz  0 0 0 0 0 1 0 1 0   ∂v 
    
γ zx  0 0 1 0 0 0 1 0 0   ∂z 
 ∂w 
 ∂x 
 
 ∂w 
 ∂y 
 
 ∂w  (3.7.10)
 ∂z 

The above equation can be expressed as

 ∂u 
 ∂ξ 
 
 ∂u 
 ∂η 
 ∂u   
 ∂x   ∂u 
   ∂ζ 
 ∂v   
 ∂y   J11* J12* J13* 0 0 0 0 0 0   ∂v 
   
 ∂w  0 0 0 *
J 21 *
J 22 *
J 23 0 0 0   ∂ξ 
 
 ∂z  0 * 
 ∂v 
{ε } =
* *
0 0 0 0 0 J 31 J 32 J 33
=   *  
 ∂u + ∂v  0   ∂η 
* *
 J 21 J 22 J 23 J11* J12* J13* 0 0
 ∂y ∂x  0 0 0 *
J 31 *
J 32 *
J 33 *
J 21 *
J 22 * 
J 23 ∂v 
   *   
 ∂v + ∂w   J 31 J11* J12* J13*   ∂ζ 
* *
J 32 J 33 0 0 0
 ∂z ∂y   ∂w 
   
 ∂x + ∂w   ∂ξ 
 ∂z ∂x   ∂w 
 
 ∂η 
 ∂v  (3.7.11)
 
 ∂ζ 
 54

8
For an 8 node brick element u can be represented as, u   N i ui and similarly for v & w.
i 1

∂u 8
∂N ∂u ∂N ∂u ∂N
8 8
= ∑ i ui , = ∑ i ui & = ∑ i ui
∂ξ i =1 ∂ξ ∂η i =1 ∂η ∂ζ i =1 ∂ζ
∂v 8
∂N ∂v 8
∂N ∂v 8
∂N
= ∑ i vi , = ∑ i vi & = ∑ i vi (3.7.12)
∂ξ i =1 ∂ξ ∂η i =1 ∂η ∂ζ i =1 ∂ζ
∂w 8 ∂N i ∂w 8 ∂N i ∂w 8
∂N
=∑ wi , =∑ wi & = ∑ i wi
∂ξ i =1 ∂ξ ∂η i =1 ∂η ∂ζ i =1 ∂ζ

Hence eq. (3.7.11) can be rewritten as

 ∂N i 
 ∂ξ 0 0 
 
 ∂N i 
 J11* J12* J13* 0 0 0 0 0 0   0 ∂η
0 
   
 ∂N i 
* * *
0 0 0 J 21 J 22 J 23 0 0 0 
0 0  ui 
8 
0 0 0 0 0 0 *
J 31 *
J 32 * 
J 33 ∂ζ    (3.7.13)
{ε }  * ×∑   vi 
 J 21
* *
J11* J12* J13* 0  i =1  ∂N i ∂N i 
0   wi 
J 22 J 23 0 0
0 * * * * * *   ∂η ∂ξ
0 0 J 31 J 32 J 33 J J J 23  
 * 21 22

 J 31 J *
J *
0 0 0 J *
J *
J13*   ∂N i ∂N i 
 0 ∂η 
32 33 11 12
∂ζ
 
 ∂N i ∂N i 
 ∂ζ 0
∂ξ 

Thu, the strain-displacement relationship matrix [B] for 8 node brick element is
 55

 ∂N i 
 ∂ξ 0 0 
 
 ∂N i 
J *
J*
J*
0 0 0 0 0 0   0 ∂η
0 

11 12 13
  
 ∂N i 
* * *
0 0 0 J21 J22 J23 0 0 0 
0 0
8 
0 0 0 0 0 0 *
J 31 *
J 32 * 
J 33 ∂ζ 
[ B]  * ×∑  (3.7.14)
 J 21
*
J 22 *
J 23 J11* J12* J13* 0 0 0  i =1  ∂N i ∂N i 
 ∂η 0 
0 0 0 *
J 31 *
J 32 *
J 33 *
J 21 *
J 22 * 
J 23 ∂ξ
 *   
 J 31
*
J 32 *
J 33 0 0 0 J11* J12* J13 
*
 ∂N i ∂N i 
 0 ∂ζ ∂η 
 
 ∂N i ∂N i 
 ∂ζ 0
∂ξ 


The stiffness matrix may be found by using the following expression in natural coordinate system.
1 1 1

 k     B  D Bd    B  D Bdxdydz     [B]T [D][B]d

d d J (3.7.15)
T T

 V 1 1 1
 56

Lecture 8: Numerical Integration: One Dimensional

The integrations, we generally encounter in finite element methods, are quite complicated and it is
not possible to find a closed form solutions to those problems. Exact and explicit evaluation of the
integral associated to the element matrices and the loading vector is not always possible because of
the algebraic complexity of the coefficient of the different equation (i.e., the stiffness influence
coefficients, elasticity matrix, loading functions etc.). In the finite element analysis, we face the
problem of evaluating the following types of integrations in one, two and three dimensional cases
respectively. These are necessary to compute element stiffness and element load vector.

 
 d ;   , d d ;   , , d d d ; (3.8.1)

Approximate solutions to such problems are possible using certain numerical techniques. Several
numerical techniques are available, in mathematics for solving definite integration problems,
including, mid-point rule, trapezoidal-rule, Simpson’s 1/3rd rule, Simpson’s 3/8th rule and Gauss
Quadrature formula. Among these, Gauss Quadrature technique is most useful one for solving
problems in finite element method and therefore will be discussed in details here.

3.8.1 Gauss Quadrature for One-Dimensional Integrals

The concept of Gauss Quadrature is first illustrated in one dimension in the context of an integral in
1 x2
the form of I     d from
  f (x)dx . To transform from an arbitrary interval of x1≤ x ≤ x2
1 x1

to an interval of -1 ≤ ξ ≤ 1, we need to change the integration function from f(x) to ϕ(ξ) accordingly.
Thus, for a linear variation in one dimension, one can write the following relations.
1 1 
x x1  x 2  N1x1  N 2 x 2
2 2
1 1 1 1
so for   1, x  x1  x 2  x1
2 2
  1, x  x 2
x2 1
I f (x)dx    d

x1 1

Numerical integration based on Gauss Quadrature assumes that the function ϕ(ξ) will be evaluated
over an interval -1 ≤ ξ ≤ 1. Considering an one-dimensional integral, Gauss Quadrature represents
the integral ϕ(ξ) in the form of
1 n
I
1
  d   w i
  i   w1
 1   w 2  
 2   ..  w 
 (3.8.2)
i1
 57

Where, the ξ1, ξ2, ξ3, ..., ξn represents n numbers of points known as Gauss Points and the
corresponding coefficients w1, w2, w3, …, wn are known as weights. The location and weight
coefficients of Gauss points are calculated by Legendre polynomials. Hence this method is also
sometimes referred as Gauss-Legendre Quadrature method. The summation of these values at n
sampling points gives the exact solution of a polynomial integrand of an order up to 2n-1. For
example, considering sampling at two Gauss points we can get exact solution for a polynomial of an
order (2×2-1) or 3. The use of more number of Gauss points has no effect on accuracy of results but
takes more computation time.

3.8.2 One- Point Formula

Considering n = 1, eq.(3.8.2) can be written as
1

1

( )d  w1
( 1) (3.8.3)

Since there are two parameters w1 and 1 , we need a first order polynomial for ϕ(ξ) to evaluate the

   a 0  a1 ,
eq.(3.8.3) exactly. For example, considering, 
1
Error   a 0  a1
d  w1
 1  0
1

 2a 0  w1 a 0  a11   0

 a 0 2  w1   w1a11  0 (3.8.4)

Thus, the error will be zero if w1  2 and 1  0 . Putting these in eq.(3.8.3), for any general ϕ, we
have
1
I   
  d  2 0 (3.8.5)
1

This is exactly similar to the well known midpoint rule.

3.8.3 Two-Point Formula

If we consider n = 2, then the eq.(3.8.2) can be written as
1

1
  d  w1
  1   w 2
 2 (3.8.6)

This means we have four parameters to evaluate. Hence we need a 3rd order polynomial for ϕ(ξ) to
exactly evaluate eq.(3.8.6).
   a 0  a1  a 22  a 33
Considering, 
 1 
Error    a 0  a1  a 2  2  a 3
3
 d    w1  2 
 1   w 2
  1 
 58

a a 2  w1 a 0  a11  a 2 12  a 313   w 2 a 0  a1 2  a 2  2 2  a 3 23   0

2
2 0
3
2 
 2  w1  w 2  a 0   w11  w 2  2  a1    w112  w 2  2 2  a 2   w113  w 2  23  a 3  0
 3 

Fig 3.8.1 One-point Gauss Quadrature

Requiring zero error yields

w1  w 2  2
w11  w 2  2  0
2 (3.8.7)
w112  w 2  22 
3
w11  w 2  2  0
3 3

These nonlinear equations have the unique solution as

w1  w 2  1 1   2  1 3  0.5773502691 (3.8.8)
From this solution, we can conclude that n-point Gaussian Quadrature will provide an exact solution
if ϕ(ξ) is a polynomial of order (2n-1) or less. Table 3.8.1 gives the values of w1 and 1 for Gauss
Quadrature formulas of orders n = 1 through n = 6. From the table it can be observed that the gauss
 59

points are symmetrically placed with respect to origin and those symmetrical points have the same
weights. For accuracy in the calculation maximum number digits for gauss point and gauss weights
should be taken. The Location and weights given in the Table 3.8.1 must be used when the limits of
integration ranges from -1 to 1. Integration limits other than [-1, 1], should be appropriately changed
to [-1, 1] before applying these values.

Table 3.8.1 Gauss points and corresponding weights

Number of Gauss Point Location, 1 Weight, w1
points, n
1 0.0 2.0
2 ±0.5773502692 (= 1 3) 1.0
3 0.0 0.8888888889 (=8/9)
±0.7745966692 (=  6 ) 0.5555555556 (=5/9)
4 ±0.3399810436 0.6521451549
±0.861363116 0.3478548451
5 0.0 0.5688888889
±0.5384693101 0.4786286705
±0.9061798459 0.2369268851
6 ±0.2386191861 0.4679139346
±0.6612093865 0.3607615730
±0.9324695142 0.1713244924

Example 1:
1 2x 
Evaluate I   e x  2 dx using one, two and three point gauss Quadrature.

0  x  2 

Solution:
Before applying the Gauss Quadrature formula, the existing limits of integration should be changed
from [0, 1] to [-1, +1]. Assuming,   a  bx , the upper and lower limit can be changed. i.e., at x =
0, ξ = -1 and at x = 1, ξ = +1. Thus, putting these conditions and solving for a & b, we get a = -1 and
b = 2. The relation between two coordinate systems will become   2x 1 and d  2dx .
Therefore the initial equation can be written as
 60

    1 
   
 2  
1  
 1 2
  2 
I   e  dx
1      
2
  1  2 

  2  


1 1  2 4   1 
1

2 1
Or, I   e  d
 1  8 
2

Using one point gauss Quadrature:

w1  2, 1  0 and
I  2 0
1  4 
Or I  2  e0.5    2.22015
 2  7 
Using two point gauss Quadrature:
w1  w 2  1
1  0.5773502692
 2  0.5773502692
Putting these values and calculating, I  2.39831

Using three point gauss Quadrature:

w1 = 0.555555556
ξ1 = −0.774596669
w2 = 0.888888889
ξ 2 = 0.000000000
w3 = 0.555555556
ξ3 = 0.774596669
and I  2.41024
This may be compared with the exact solution as Iexact  2.41193
 61

Lecture 9: Numerical Integration: Two and Three Dimensional

Numerical integrations using Gauss Quadrature method can be extended to two and three
dimensional cases in a similar fashion. Such integrations are necessary to perform for the analysis of
plane stress/strain problem, plate and shell structures and for the three dimensional stress analysis.

3.9.1 Gauss Quadrature for Two-Dimensional Integrals

For two dimensional integration problems the above mentioned method can be extended by first
evaluating the inner integral, keeping η constant, and then evaluating the outer integral. Thus,

 n  n  n 
,  d d    w i 
i ,  d   w j  w i i , j
1 1 1
I    
     
1 1 1    
 i1  i1  i1
Or,
n n
 i , j
I   w i w j (3.9.1)
i1 j1

In a matrix form we can rewrite the above expression as

φ (ξ1 ,η1 ) φ (ξ1 ,η2 ) φ (ξ1 ,ηn )   w1 
 
 φ (ξ 2 ,η1 ) φ (ξ 2 ,η2 ) φ (ξ 2 ,ηn )   w2 
I ≈ [ w1  wn ]
  
w2 (3.9.2)
 
  
φ ( ξ n , η1 ) φ ( ξ n , η 2 ) φ ( ξ n , η )
n   wn 

Example 1:
yd  4 x  b3

Evaluate the integral: I    1 x  2  y dxdy

2 2

yc4 x a  2

Solution:
Before applying the Gauss Quadrature formula, the above integral should be converted in terms of

and and the existing limits of y should be changed from [-4,4] to [-1, 1] and that of x is from
[2,3] to [-1,1].
 62

b  a  b  a    5 d
x   ; dx 
2 2 2 2
d  c  d  c 
y   4; dy  4d
2 2
1 1 1 1
 3   
2

I  2      
d d     , d d
2
 2   2 4
1 1 1 1

 3   
2

 ,   2 
  2  4  2 3   1 2
2 2 2
where
 2 

Fig. 3.9.1 Gauss points for two-dimensional integral

 63

1 1 1 1
1   ; 1   ; 2  ; 2 
3 3 3 3
 1   2 
2 2

 1 , 1   2 3   1    54.49857


 3  3
 
2
 3  1 
3   
2
 2  4   118.83018
 2 , 1   
 
 2   3 
 
 
2
 3  1 
3   
2
 2  4   0.61254
 2 , 2   
 
 2   3 
 
 
2
 3  1 
3   4 
2

 1 , 2   
  2    0.28093
 2   3
 

 
 1 , 1   1 , 2  w1 
I  w 1 w 2    
 2 , 1   2 , 2  w 2 


 54.49857 0.28093 1

 1 1  
118.83018 0.61254 1

= 174.22222 agrees with the exact value 174.22222

3.9.2 Gauss Quadrature for Three-Dimensional Integrals

In a similar way one can extend the gauss Quadrature for three dimensional problems also and the
integral can be expressed by.
n n n
 i , j, k 
 , ,  d d d   w i w j w k 
1 1 1
I    (3.9.3)
1 1 1
i1 j1 k 1

The above equation will produce exact value for a polynomial integrand if the sampling points are
selected as described earlier sections.
 64

3.9.3 Numerical Integration of Element Stiffness Matrix

As discussed earlier notes, the element stiffness matrix for three dimensional analyses in natural
coordinate system can be written as
1 1 1

 k     B  D Bd    B  D Bdxdydz     [B] [D][B]d

(3.9.4)
T T T
d d J
 V 1 1 1

Here, [B] and [D] are the strain displacement relationship matrix and constitutive matrix respectively
and integration is performed over the domain. As the element stiffness matrix will be calculated in
natural coordinate system, the strain displacement matrix [B] and Jacobian matrix [J] are functions
of 
, and . In case of two dimensional isoparametric element, the stiffness matrix will be
simplified to
1 1

k   t   [B] [D][B]d
T
d J (3.9.5)
1 1

This is actually an 8×8 matrix containing the integrals of each element. We do not need to integrate
elements below the main diagonal of the stiffness matrix as it is symmetric. Considering,
 ,   t[B]T [D][B] J , the element stiffness matrix will become after numerical integration as

n n
 i , j
 k    w i w j (3.9.6)
i1 j1

Using a 2×2 rule, we get

 k   w12
 1 , 1   w1w 2
 1 , 2   w 2 w1
 2 , 1   w 22
 2, 2 (3.9.7)

Where w1  w 2  1.0, 1  1  0.57735...., and  2  2  0.57735.... Here, wn is the weight

factor at integration point n. A suitable computer program can be written to calculate the element
stiffness matrix through the numerical integration. The process of obtaining stiffness matrix using
Gauss Quadrature integration will be demonstrated through a numerical example in module 5.

3.10.4 Gauss Quadrature for Triangular Elements

The procedure described for the rectangular element will not be applicable directly. The Gauss
Quadrature is extended to include triangular elements in terms of triangular area coordinates.

I    L1 , L 2 , L3  d A  w i Li1 , Li2 , Li2 

n
(3.9.8)
A i1

Where, L terms are the triangular area coordinates and the wi terms are the weights associated with
those coordinates. The locations of integration points are shown in Fig. 3.9.2.
 65

Fig. 3.9.2 Gauss points for triangles

The sampling points and their associated weights are described below:
For sampling point =1 (Linear triangle)
1
w1  1 L11  L12  L13  (3.9.9)
3
For sampling points =3 (Quadratic triangle)
1 1
w1  L11  L12  , L13  0
3 2
1 1
w2  L21  0, L22 L23  (3.9.10)
3 2
1 1 1
w3  L31  , L32  0, L33 
3 2 2
For sampling point = 7 (Cubic triangle)
27 1
w1  L11  L12  L13 
60 3
8 1
w2  L21  L22  , L23  0
60 2
8 1
w3  L31  0, L32  L23 
60 2
8 1
w4  L41  L43  , L42  0 (3.9.11)
60 2
3
w5  L51  1, L52  L53  0
60
3
w6  L61  L63  0, L62  1
60
3
w7  L71  L72  0, L73  1
60
 66

3.10.5 Gauss Quadrature for Tetrahedron

The Gauss Quadrature for triangles can be effectively extended to include tetrahedron elements in
terms of tetrahedron volume coordinates.

I    L1 , L 2 , L3 , L 4  d A  w i Li1 , Li2 , Li3 , Li4 

n

i1
A (3.9.12)
Where, L terms are the volume coordinates and the wi terms are the weights associated with those
coordinates. The locations of Gauss points are shown in Fig. 3.9.3.

Fig. 3.9.3 Gauss points for tetrahedrons

The sampling points and their associated weights are described below:
For sampling point = 1 (Linear tetrahedron)
1
w1  1 L11  L12  L13  L14  (3.9.13)
4
For sampling points = 4 (Quadratic tetrahedron)
1
w1  L11  0.5854102, L12  L13  L14  0.1381966
4
1
w2  L22  0.5854102, L21  L23  L24  0.1381966
4
(3.9.14)
1
w3  L33  0.5854102, L31  L32  L34  0.1381966
4
1
w4  L44  0.5854102, L41  L42  L43  0.1381966
4
For sampling points = 5 (Cubic tetrahedron)
 67

4 1
w1   L11  L12  L13  L14 
5 4
9 1 1
w2  L21  , L22  L23  L24 
20 3 6
9 1 1
w3  L32  , L31  L33  L34  (3.9.15)
20 3 6
9 1 1
w4  L43  , L41  L42  L44 
20 3 6
9 1 1
w5  L54  , L51  L52  L43 
20 3 6
 68

Worked out Examples

Example 3.1 Calculation of displacement using area coordinates
The coordinates of a three node triangular element is given below. Calculate the displacement at
point P if the displacements of nodes 1, 2 and 3 are 11 mm, 14mm and 17mm respectively using the
concepts of area coordinates.

1 𝑥1 𝑦1 1 2 3
1 1 1 8
A = 2 �1 𝑥2 𝑦2 � = �1 5 4� =
2 2
[(30-12) - (12-9) + (8-15)] = 2 = 4
1 𝑥3 𝑦3 1 3 6

1 𝑥 𝑦 1 3 4
1 1 1 4
𝐴1 = 2
�1 𝑥2 𝑦2 �= �1 5
2
4� = 2 [(30-12) - (18-12) + (12-20)] = 2 = 2
1 𝑥3 𝑦3 1 3 6

Fig. Ex.3.1 Nodal coordinates of a triangular element

1 𝑥 𝑦 1 3 4
1 1 1 2
𝐴2 = 2
�1 𝑥3 𝑦3 �= �1 3
2
6� = 2 [(9-12) - (9-8) + (18-12)] = 2 = 1
1 𝑥1 𝑦1 1 2 3
 69

1 𝑥 𝑦 1 3 4
1 1 1 2
𝐴3 = 2
�1 𝑥1 𝑦1 �= �1 2
2
3� = 2 [(8-15) - (12-20) + (9-8)] = 2 = 1
1 𝑥2 𝑦2 1 5 4

𝐴1 2
𝑁1 = = 4 = 0.5
𝐴

𝐴2 1
𝑁2 = = 4 = 0.25
𝐴

𝐴3 1
𝑁3 = = 4 = 0.25
𝐴

u = 𝑁1 𝑢1 +𝑁2 𝑢2 +𝑁3 𝑢3
= 0.5 x 11 + 0.25 x 14 + 0.25 x 17 = 13.25 mm
 70

Example 3.2 Derivation of shape function of four node triangular element

Derive the shape function of a four node triangular element.

Fig. Ex.3.2 Degrading for four node element

The procedure for four node triangular element is the same as five node triangular element to derive
its interpolation functions. Here, node 5 and 6 are omitted and therefore displacements in these
nodes can be expressed in terms of the displacements at their corner nodes. Hence,
u 2  u3 u1  u 3
u '5  an d u '6 
2 2 (3.11.1)
Substituting the values of u’5 and u’6 in eq.(3.3.8), the following relations can be obtained.
u 2  u 3   u 3  u1 
u  N1u1  N 2 u 2  N 3 u 3  N 4 u 4  N 5  N6
2 2
(3.11.2)
 N   N   
N  N6 
  N1  6  u1   N 2  5  u 2   N 3  5  u  N4u 4
 2  
 2   2  3
Now, the displacement at any point inside the four node element can be expressed by its nodal
displacement with help of shape function.
u  N1u1  N 2 u 2  N 3 u 3  N 4 u 4 (3.11.3)
Comparing eq. (3.11.2) and eq. (3.11.3), one can find the following relations.
N6 4L L
N1  N1   L1 2L1 1  3 1  L1 1 2L 2 
2 2
N 4L L
N 2  N 2  5  L 2 2L 2 1  2 3  L 2 1 2L1 
2 2 (3.11.4)
N  N6 4L L  4L3 L1
N 3  N 3  5  L3 2L3 1  2 3  L3
2 2
N 4  N 4  4L1L 2
Thus, the shape functions for the four node triangular element are
 71

N1  L1 1 2L 2 
N 2  L 2 1 2L1 
(3.11.5)
N 3  L3
N 4  4L1L 2

Example 3.3 Numerical integration for two dimensional problems

3

Evaluate the integral: I   x  11x  32 dx using one, two and three point gauss Quadrature.
2

2

Also, find the exact solution for comparison of accuracy.

Solution:
The existing limits of integration should be changed from [-2, +3] to [-1, +1]. Assuming,   a +bx ,
the upper and lower limit can be changed. i.e., at x1  2, 1  1 and at x2  3,  2  1 . Thus,
putting these limits and solving for a & b, we get a = -0.2 and b = 0.4. The relation between two
coordinate systems will become:
5  1
  0.2  0.4x or x  and dx  2.5d
2

Thus, the initial equation can be written as

3  5  121
 5  1 
I    x  11x  32 dx = 2.5   
 11  32d 
 2   2 
2

1 
 
2 

(i) Exact Solution:

3

I    x 2  11x  32 dx
2

 x3 11x 2 
3

   32 x 
3 
 2  2
 99   8 
 9   96    22  64
 2   3 
 37.5  83.33333  120.83333
Thus, Iexact = -120.83333
 72

(ii) One Point Formula:

1

I   
  d  w1
 1
1

For one point formula in Gauss Quadrature integration, w1  2, 1  0 . Thus,

 5 0  12 
I1  2  2.5    11 5 0  1  32
   2  
 2  
 1 11 
 5    32  131.25
 4 2 
Thus, % of error = (120.83333-131.25)×100/120.83333 = 8.62%

(iii) Two Point Formula:

Here, for two point formula in Gauss Quadrature integration,
1
w1  w 2  1.0 and 1   2   . Thus,
3
 1   w2
I 2  w1  2
 5 
2
 5    5 
2
 5  
             
  3 1   3 1    3 1   3 1 
1.0  2.5    11   32  1.0  2.5    11   32
 2   2    2   2  
         
         

 0.88996 10.37713  32 2.5  3.77671  21.3771  32 2.5

 48.3333 2.5
= 120.83325
Thus, % of error = (120.83333-120.83325)×100/120.83333 = 6.62×10-05

(iv) Three Point Formula:

Here, for three point formula in Gauss Quadrature integration,
w1  0.8889, 1  0.0
w2  0.5556,  2  0.7746
w3  0.5556, 3  0.7746
Thus,
 1   w2
I 3  w1  3
 2   w3
 73

 5 0  12 5 0  1 

I 3  0.8889  2.5    11  32
 2 
 2 
 5 0.7746  12 
 0.5556  2.5    11 5 0.7746  1  32
  
 2 2 
 5 0.7746  12 
 0.5556  2.5    11 5 0.7746  1  32
  
 2 2 
I 3  0.8889  2.50.25  5.5  32
 0.5556  2.55.9365  26.8015  32
 0.5556  2.5 2.0635 15.8015  32
 2.523.3336  0.4100  25.4120
 2.5 48.3356  120.839
Thus, % of error = (120.83333-120.839)×100/120.83333 = 4.69×10-03. However, difference of
results will approach to zero, if few more digits after decimal points are taken in calculation.

Example 3.4 Numerical integration for three dimensional problems

1 1 1
Evaluate the integral: I     1 2 1
 3  2 d
2 2 2
d d
1 1 1

Solution:
Using two point gauss Quadrature formula for the evaluation of three dimensional integration, we
have the following sampling points and weights.
w1  w 2  1
1  0.5773502692
 2  0.5773502692
1  0.5773502692
 2  0.5773502692
1  0.5773502692
 2  0.5773502692

 , ,   1 2 1

Putting the above values, in   3  2 one can find the following values
2 2 2

in 8 (i.e., 2 × 2 × 2) sampling points.

 74

 1 , 1 , 1   160.8886

 1 , 1 , 2   0.8293

 1 , 2 , 1   11.5513

 1 , 2 , 2   0.0595

 2 , 1 , 1   0.8293

 2 , 1 , 2   0.0043

 2 , 2 , 1   0.0595

 2 , 2 , 2   0.0003

2 2 2
 i , j, k 
Now, I   w i w j w k 
i1 j1 k 1

 1 , 1 , 1   w1w1w 2
Thus, I  w1w1w1  1 , 1 , 2     w 2 w 2 w 2
 2 , 2 , 2  = 174.222, where as
Iexact = 174.222.
 1

Lecture 1: Stiffness of Truss Members

4.1.1 Introduction
Analysis of frame structures can be carried out by the approach of stiffness method.
However, such types of structures can also be analyzed by finite element method. A unified
formulation will be demonstrated based on finite element concept in this module for the
analysis of frame like structures. A truss structure is composed of slender members pin
jointed together at their end points. Truss element can resist only axial forces (tension or
compression) and can deform only in its axial direction. Therefore, in case of a planar truss,
each node has components of displacements parallel to X and Y axis. Planar trusses lie in a
single plane and are used to support roofs and bridges. Such members will not be able to
carry transverse load or bending moment. The major benefits of use of truss structures are:
lightweight, reconstructable, reconfigurable and mobile. Configuration of few standard truss
structures are shown in Fig. 4.1.1.

Fig. 4.1.1 Configuration of various truss structures

4.1.2 Element Stiffness of a Truss Member

Since, the truss is an axial force resisting member, the displacement along its axis only will
be developed due to axial load. Therefore, using Pascal’s triangle, the displacement function
of truss member for development of shape function can be expressed as:

α 
u ( x ) =α 0 + α1 x =[1 x ]  0 
α1  (4.1.1)
 2

Fig. 4.1.2 Axial force on the member along X axis

Applying boundary conditions as shown in Fig. 4.1.2:

At x= 0, u(0)= u 1 and at x=L, u(L) = u 2
u 2 − u1
Thus, α 0 = u1 and α 1 = Therefore,
L .
 x
u ( x ) = 1 − u1 + u 2 = [N ]{u}
x
(4.1.2)
 L L
Here, N is the shape function of the element and is expressed as:

[N ] = 1 − x x
(4.1.3)
 L L 
So we get the element stiffness matrix as

=[k ] ∫∫∫ [ B ] [ D ][ B ] d Ω
T
(4.1.4)
Ω

d [N ]  1 1 
Where, [B ] = = − 
dx  L L
So, the stiffness matrix will become:
 1
−   1 1 AE  1 − 1
= ∫ [B ] E [B ]Adx = AE ∫  L −
L L
=
T

L − 1 1 
dx
0 0

1  L
 L 
 L 

Thus, the stiffness matrix of the truss member along its member axis will be:

AE  1 − 1
[k ] =
L − 1 1 
(4.1.5)

4.1.3 Element Stiffness of Truss Member with Varying Cross Section

 3

Now, let us find the stiffness matrix of a pin-jointed member of length L with respect to local
axis, having cross sectional areas A1 and A2 at the two ends of the member as shown in the
figure below.

Fig. 4.1.3 Member with varying cross section

From the above figure, the cross sectional area at a distance of x from left end can be
expressed as:
A − A1
Ax = A1 + 2 x (4.1.6)
L
As it is a pin-jointed member, the displacement at any point may be expressed in terms of
nodal displacement as= u N1u1 + N 2u2 .
Similarly the cross sectional area at any point may be represented in terms of the cross
sectional area of the two ends. Thus Ax = N 1 A1 + N 2 A2

x x
Where the shape functions are: N1 =1 − ; N2 =
L L
Now, the strain may be written as:

∂u ∂N1 ∂N 2 1 1 u1 
εx = u 2 = − u1 + u 2 = 1 [− 1  
= u1 + 1]  = [B ]{u} (4.1.7)
∂x ∂x ∂x L L L 
u 2 

As the stress is proportional to strain according to Hook’s law, the stress-strain relationship
will be as follows:
u1 
 
σ x = Eε x = [− 1 1]  = E [B ]{u}
E
(4.1.8)
L 
u 2 


Now the strain energy may be expressed as

L L
U = ∫ ε x σ x dv = ∫ ε x Eε x Ax dx = ∫ {u} [B ] E [B ]{u}Ax dx
1 T 1 T 1 T T
(4.1.9)
2V 20 20

Applying Castigliano’s theorem, the force will become:

 4

T u1 
 
L
∂U
L

∫ [− 1 1] [− 1 1]Ax dx   = [k ]{d }
E
{F } = = ∫ [B ] E [B ]{u}Ax dx = 2
T

∂{u} 0 L 0 u 2 

(4.1.10)

Thus, the stiffness matrix will be:

1 − 1 L
E  1 − 1  A − A1 2 
L

[k ] = E    A1 + A2 − A1 
∫
x dx = 2    A1 x + 2 x 
L2 − 1 1 0  L  L − 1 1   2L 0


E  1 − 1  A2 − A1   1 − 1
= + =
E
( + )
L − 1 1   − 1 1 
A A A
2  2 L
1 1 2
  (4.1.11)

4.1.4 Generalized Stiffness Matrix of a Plane Truss Member

Let us consider a member making an angle ‘θ’ with X axis as shown in the figure below. By
resolving the forces along local X and Y direction, the following relations are obtained.

=Fx1 Fx1 cos θ + Fy1 sin θ

=Fx 2 Fx 2 cos θ + Fy 2 sin θ
(4.1.12)
− Fx1 sin θ + Fy1 cos θ
Fy1 =
− Fx 2 sin θ + Fy 2 cos θ
Fy 2 =

Where, Fx1 and Fx 2 are the axial forces along the member axis X . Similarly, Fy1 and Fy 2 are
the forces perpendicular to the member axis X .

Fig. 4.1.4 Inclined truss member

 5

The relationship expressed in eq. (4.1.12) can be rewritten in matrix form as follows:

 Fx1   cos θ sinθ 0 0   Fx1 

    
 Fy1   − sinθ cos θ 0 0   Fy1 
 =   (4.1.13)
 Fx 2   0 0 cos θ sinθ   Fx 2 
   − sinθ

cos θ   Fy 2 
 Fy 2   0 0

Now, the above equation can be expressed in short as:

{F } = [T ]{F } (4.1.14)

Here, [T] is called transformation matrix. This relates between the global (𝑋, 𝑌 axis) and
member axis (𝑋�, 𝑌� axis). Similarly, the relations of nodal displacements between two
coordinate systems may be written as:

{d } = [T ]{d } (4.1.15)

Again, the equation stated in (4.1.5) can be generalized and expressed with respect to the
member axis including force and displacement vector as:

 Fx1  1 0 −1 0   u1 
  
 Fy1  AE  0 0 0 0   v1 
 =   (4.1.16)
 Fx 2  L  −1 0 1 0  u2 
 Fy 2   
  0 0 0 0   v2 

Where, the nodal forces in Y direction are zero. The above equation may also be expressed in
short as:

{F } = [k ]{d } (4.1.17)

Where, the matrices in the above equation are written with respect to the member axis. Now,
eq. (4.1.17) can be rewritten with the use of eq. (4.1.14) and (4.1.15) as given below.

[T ]{F } = [k ][T ]{d } (4.1.18)

Or,
{F } = [T ]−1 [k ][T ]{d } (4.1.19)
 6

Here, the transformation matrix [T] is orthogonal, i.e., [T]-1 is equal to [T]T. Therefore, from
the above relationship, the generalized stiffness matrix can be expressed as:
[k ] = [T ]T [k ][T ] (4.1.20)

Thus,

cos θ − sin θ 0 0  1 0 −1 0   cos θ sin θ 0 0 

 sin θ co θ s 0  
0  AE  0 0 0 0   − sin θ co θ s 0 0 
[ k ] =  0 0 cos θ − sin θ  L  −1 0 1 0  0 0 cos θ sin θ 
    
 0 0 sin θ co θ s 0 0 0 0  0 0 − sin θ co θ s

(4.1.21)
Or,

 cos 2 θ sin θ cosθ − cos 2 θ − sin θ cosθ 

 
 sin θ cosθ sin 2 θ − sin θ cosθ − sin 2 θ 
[k ] = AE   (4.1.22)
L  − cos 2 θ − sin θ cosθ cos θ
2
sin θ cosθ 
 
− sin θ cosθ
 − sin 2 θ sin θ cosθ sin θ 
2

The above stiffness matrix can be used for the analysis of two-dimensional truss problems.
 7

Lecture 2: Analysis of Truss

4.2.1 Element Stiffness of a 3 Node Truss Member

Fig. 4.2.1 3-node truss member

Here, the displacement function using Pascal’s triangle can be expressed as:
α 0 
 
u ( x ) =α 0 + α1 x + α 2 x = 1 x x  α1 
2 2
(4.2.1)
α 
 2

Applying boundary conditions:

At x= 0, u(0)= u 1 , x=L/2, u(L/2) = u and at x=L, u(L) = u
2 3

And solving for α 0 , α1 and α 2

2u1 − 4u 2 + 2u 3
α 0 = u1 , α = −3u1 + 4u2 − u3 and α 2 =
1
L L3
Therefore,
 3x 2 x 2   4x 4x2   x 2x2 
u ( x ) = 1 − + 2  u 1 +  − 2  u2 +  − + 2  u 3 = [ N ]{u} (4.2.2)
 L L   L L   L L 

Here, N is the shape function of the element and is expressed as:

 3 x 2 x 2   4 x 4 x 2   x 2x2  
[ ]  1 − + 2   − 2 
N = − + 2   (4.2.3)
 L L   L L   L L 

Now, the element stiffness matrix can be written as

=[k ] ∫∫∫ [ B ] [ D ][ B ] d Ω
T
(4.2.4)
Ω

d [ N ]  3 4x 4 8x 1 4x 
Where, [ B ] = =− + − − +
d x  L L2 L L2 L L2 

So, the stiffness matrix will be:

[k ] ∫∫∫ [ B ] [ D ][ B ] d Ω = ∫ [ B ] E [ B ] Adx
L T
=
T
0
Ω
 8

3 4𝑥
⎧− + 2 ⎫
𝐿 𝐿 ⎪
𝐿⎪
4 8𝑥 3 4𝑥 4 8𝑥 1 4𝑥
= 𝐴𝐸 � − 2 × �− + 2 − − + 2 � 𝑑𝑥
0 ⎨ 𝐿 𝐿 ⎬ 𝐿 𝐿 𝐿 𝐿2 𝐿 𝐿
⎪ 1 4𝑥 ⎪
⎩− 𝐿 + 𝐿2 ⎭

16𝑥 2 24𝑥 40𝑥 32𝑥 2 16𝑥 16𝑥 2

⎡ 9+ − −12 + − 2 3− + 2 ⎤
⎢ 𝐿2 𝐿 𝐿 𝐿 𝐿 𝐿 ⎥
𝐴𝐸 𝐿 ⎢ 40𝑥 32𝑥 2 64𝑥 64𝑥 2 24𝑥 32𝑥 2 ⎥
= 2 � ⎢−12 + − 2 16 − + 2 −4 + − 2 ⎥ 𝑑𝑥
𝐿 0 𝐿 𝐿 𝐿 𝐿 𝐿 𝐿 ⎥
⎢ 2
⎢ 16𝑥 16𝑥 24𝑥 32𝑥 2 8𝑥 16𝑥 2 ⎥
⎣ 3 − + −4 + − 2 1− − 2 ⎦
𝐿 𝐿2 𝐿 𝐿 𝐿 𝐿

(4.2.5)

After integrating the above equation, the stiffness matrix of the 3-node truss member will
become:
7 −8 1
AE
[𝑘] = �−8 16 −8� (4.2.6)
3L
1 −8 7

4.2.2 Worked Out Example

Analyze the truss shown below by finite element method. Assume the cross sectional area of
the inclined member as 1.5 times the area (A) of the horizontal and vertical members. Assume
modulus of elasticity is constant for all the members and is E.

Fig. 4.2.2 Plane truss

 9

Solution

The analysis of truss starts with the numbering of members and joints as shown below:

Fig. 4.2.3 Numbering of members and nodes

The member information for the truss is shown in Table 4.2.1. The member and node
numbers, modulus of elasticity, cross sectional areas are the necessary input data. From the
coordinate of the nodes of the respective members, the length of each member is computed.
Here, the angle θ has been calculated considering anticlockwise direction. The signs of the
direction cosines depend on the choice of numbering the nodal connectivity.

Table 4.2.1 Member Information for Truss

Member Starting Ending Value Area Modulus of

No. Node Node of θ Elasticity
1 1 2 90° A E
2 2 3 315° 1.5A E
3 3 1 180° A E

Now, let assume the coordinate of node 1 as (0, 0). The coordinate and restraint joint
information are given in Table 4.2.2. The integer 1 in the restraint list indicates the restraint
exists and 0 indicates the restraint at that particular direction does not exist. Thus, in node no.
2, the integer 0 in x and y indicates that the joint is free in x and y directions.

Table 4.2.2 Nodal Information for Plane Truss

Node No. Coordinates Restraint List

x y x y
1 0 0 1 1
2 0 L 0 0
3 L 0 1 1
 10

The stiffness matrices of each individual member can be found out from the stiffness matrix
equation as shown below.

 cos 2 θ cos θ sin θ − cos 2 θ − cos θ sin θ 

 
 cos θ sin θ sin 2 θ − cos θ sin θ − sin 2 θ 
[k ] = AE  
L  − cos 2 θ − cos θ sin θ cos 2 θ cos θ sin θ 
 
− cos θ sin θ
 − sin 2 θ cos θ sin θ sin θ 
2

Thus the local stiffness matrices of each member are calculated based on their individual
member properties and orientations and written below.
1 2 3 4 3 4 5 6
0 0 0 0 1 1 −1 −1 1 3
   
   
0 1 0 − 1 −1 1 1 − 1 4
AE 
2 3 AE 
[k ]1 =   [k ]2 =  
L 0 0 0 0 3
4 2 L − 1 1 1 − 1 5
   
   
0 −1 0 1  4  1 −1 −1 1  6
and
5 6 1 2
1 0 −1 0 5
 
 
0 0 0 − 1 6
AE 
[k ]3 =  
L − 1 0 1 0 1
 
 
 0 0 0 0  2
Global stiffness matrix can be formed by assembling the local stiffness matrices into globally.
Thus the global stiffness matrix are calculated from the above relations and obtained as
follows:
1 2 3 4 5 6
1

6
 11

1 0 0 0 −1 0
0 1
 0 −1 0 0 
3 3 3 3
0 0 − − 
 4 2 4 2 4 2 4 2 
 3 
[K ] = AE  0 − 1 − 3
1+
3 3
− 
L  4 2 4 2 4 2 4 2
− 1 0 − 3 3 3 3 
+1 −
 4 2 4 2 4 2 4 2
 3 3 3 3 
0 0 − − 
 4 2 4 2 4 2 4 2 

 Fx1  0
F  0
 y1   
 Fx 2  2 P 
{F } =
The equivalent load vector for the given truss can be written as:=   
 Fy 2  P
 Fx 3  0
   
 Fy 3  0

Let us assume that u and v are the horizontal and vertical displacements respectively at joints.
Thus the displacement vector will be expressed as follows:

u1   0 
v   0 
 1  
u  u 
{d } =  2  =  2 
v 2  v 2 
u 3   0 
   
 v3   0 

Therefore, the relationship between the force and the displacement will be:
 12

1 0 0 0 −1 0 
0 1 0 −1 0 0 

 Fx1   3 3 3 3  0
F  0 0 − −  0
 y1   4 2 4 2 4 2 4 2   
 2 P  AE  3 3 3 3  u2 
  =  0 −1 − 1+ −   
P L
 4 2 4 2 4 2 4 2  v2 
 Fx 3   3 3 3 3  0
   −1 0 − +1 −   
 Fy 3   4 2 4 2 4 2 4 2 0
 3 3 3 3 
0 0 − −
 4 2 4 2 4 2 4 2 

From the above relation, the unknown displacements u2 and v2 can be found out through
computer programming. However, as numbers of unknown displacements in this case are
only two, the solution can be obtained by manual calculations. The above equation may be
rearranged with respect to unknown and known displacements in the following form:

 Fα   kαα kαβ   dα 
 =  
 Fβ   kβα kββ  d β 

Thus the developed matrices for the truss problem can be rearranged as:

2P  3 −3 −3 
3
0 0
4 2
 4 2 4 2 4 2  u2
P  − 3 3 3 −3 
1+ 0 −1 v2
4 2 4 2 4 2 4 2 
 
Fx1 AE   0
= −1 0 
L 
0 0 1 0
Fy1  0 −1 0 1 0 0  0
 −3 3 3 3 
 −1 0 +1 − 
Fx3 4 2 4 2 4 2 4 2 0
 3 −3 3 3 
0 0 −
4 2 4 2 
Fy3 0
 4 2 4 2
.
The above relation may be condensed into following

 3 −3 
2 P  AE  4 2 4 2  u 
 =   2
 P  L  −3 1+
3   v2 
4 2 4 2 

 13

The unknown displacements can be derived from the relationships expressed in the above
equation.
−1
 3 −3   3 3 
  1+
u2  AE 4 2 4 2  2 P  4 2 L  4 2 4 2  2 P 
  =     
 v2  L  −3 3   P  3 AE  3 3  P 
4 2 1+
 4 2   4 2
 4 2 

Thus the unknown displacement at node 2 of the truss structure will become:

 8 2
u2  PL 3 + 
 =  3 
 v2  AE  3 
 

Support Reactions:

The support reactions {Ps} can be determined from the following relation:
{Ps } =
− { Pcs } +  K βα  {dα }

Where, {Pcs} correspond to equivalent loadings at supports. Thus, the support reaction of the
present truss structure will be:

 0 0 
 0  0 
0  −1 
0    8 2   − 3P 
{Ps } =   AE  −3
− +
3  PL 3 +  = 
  AE  3  − 2 P 
0  L  4 2 4 2
 3   
0   3 −3   2P 
 
4 2 4 2

Member End Actions:

Now, the member end actions can be obtained from the corresponding member stiffness and
the nodal displacements. The member end forces are derived as shown below.

Member –1

 0 
 Fmx1  0 0 0 0     0 
  0 1 0 −1  0   
 Fmy1  AE     PL −3P 
  = 8 2   
 Fmx 2  L 0 0 0 0  3 +  AE  0 
  3 
 Fmy 2  0 −1 0 1 
 3 
 3P 
 14

Member – 2
 8 2
 Fmx 2   1 −1 −1 1  3 +   2P 
F   −1 1 1 −1  3  −2 P 
 my 2  3 AE    3  PL  
  =  
 Fmx 3  4 2 L  −1 1 1 −1   AE −2 P 
 Fmy 3    0   2 P 
 1 −1 −1 1 
 0 

Member –3
 Fmx 3  1 0 − 1 0 0 0
F  
 my 3  AE  0 0 0 0 0 PL 0
  
 =   = 
 Fmx1  L − 1 0 1 0 0 AE 0
  0
 Fmy1  0 0 0 0 0

Thus the member forces in all members of the truss will be:
 3P   3P 
 2
  
{ m}  ( ) ( )  =
2
F = − 2 P + 2 P −2 2 P 
   0 
  
0


The reaction forces at the supports of the truss structure will be:
 0 
 −3P 
{FR } =  
−2 P 
 2 P 

Thus the member force diagram will be as shown in Fig. 4.2.4.

Fig. 4.2.4 Member Force Diagram

 15

Lecture 3: Stiffness of Beam Members

4.3.1 Introduction
A beam is a structural member which is capable of withstanding load primarily by resisting
bending. The primary tool for analysis of beam is the Euler–Bernoulli beam equation. Other
methods for determining the deflection of beams include "slope deflection method" and
"method of virtual work". For calculation of internal forces of beam include "moment
distribution method", force or flexibility method and stiffness method. However, all these
methods have limitations if either of geometry, loading, material properties or boundary
conditions becomes arbitrary in nature. Finite element techniques can well handle such cases
and relieve the analyzer of making simplifications to arrive approximate solutions.

4.3.2 Derivation of Shape Function

The degrees of freedom at each node for a beam member will be (i) vertical deflection and
dv
(ii) rotation. For a beam member, the slope of the elastic curve θ is given by: θ = , where
dx
the variable v is the displacement function of the beam. As the beam has two degrees of
freedom at each node, the variation of v will be cubic and can be expressed using Pascal’s
triangle as:
α 0 
α 
v(x ) = α 0 + α 1 x + α 2 x 2 + α 3 x 3 = [1 x x 2 ]
3  1
x   (4.3.1)
α 2 
α 3 
and
α 0 
α 
θ=
dv
[  
= 0 1 2 x 3x 2  1  ] (4.3.2)
dx α 2 
α 3 

Fig. 4.3.1 Beam element

 16

Now, applying boundary conditions, the following expressions from the above relations can
be obtained:

At x=0:
α 0  α 0 
α  α 
 
V1 = [1 0 0 0]  1  ; θ1 = [ 0 1 0 0]  1  ;
α 2  α 2 

α 3 
 α 3 

At x=L:
α 0  α 0 
α  α 
3  1 2  1
V2 = 1 L L2 L    θ 2 = 0 1 2 L 3L   
;
α 2  α 2 
α 3  α 3 

Thus combining the above expressions one can write:

V1  1 0 0 0  α 0 
θ 
 1
0 1 0 0  α1 
=  =  
1 L L2 L3  α 2 
[ A]{α } (4.3.3)
V2 
θ 2   2
0 1 2 L 3L  α 3 

So,
−1
α 0  1 0 
0 0 V1   1 0 0 0 
V1 
α   0  θ   0 1 0 0   
 1  1 0
 0  2  3 1 θ 1 
 =   = − −
2 3
−  
(4.3.4)
α 2  1 L 
L2 L3 V2   L2 L L2 L  V2 
α 3  0 2
1 2 L 3L  θ 2   2 1 2
− 3
1  θ 
 2
 3
 L L2 L L2 

Therefore,
 1 0 0 0
V1  V1 
 0
 3 1 0 0    θ 
1 θ1 
[
v(x ) = 1 x x2 ]
x 3 − 2 −
2 3
−    = [N 1 N2 N3
 
N 4 ] 1  (4.3.5)
 l l l2 l  V2  V2 
 2 1 2 1  θ  θ 2 
− 3  2
 l 3 l2 l l 2 
Where,
3 2 2 3 2 2 x3 3x 2 2 x 3 x2 x3
N1 = 1 − x + x ; N 2 = x − x + ; N 3 = − and N 4 = − +
L2 L3 L L2 L2 L3 L L2
(4.3.6)
 17

N is called shape function which interpolates the beam displacement in terms of its nodal
displacements.

4.3.3 Derivation of Element Stiffness Matrix

Now, the strain displacement relationship matrix [B] can be expressed from the following
expressions with the help of eq. (4.3.1):
α 0 
 
d 2v α 1 
= [0 6 x ]  = [B ]{α } = [B ][A] {d }
−1
χ= 0 2 (4.3.7)
dx 2 α 2 
 
α 4 

1 0 0 0  V1 
0 0  θ 
Where, [ B ] [ 0=
1 0  2
= 0 2 6 x ] ; [ A]  = ; {d }  
1 L L2 L3  V2 
  θ 2 
0 1 2 L 3L2 

From the moment curvature relationship, we can write:

d 2v
= EI [B ][ A] {d }
−1
M = EIχ = EI 2
(4.3.8)
dx
Strain energy,

∫ {d } [A ] [B] [B][A ]{d }dx

L L
U = ∫ [χ ] [M ]dx =
1 T EI T −1 T T −1
(4.3.9)
0
2 2 0

Thus,
∂U
[ ] [B] [B][A ]{d }dx
L
{F } = = EI ∫ A −1
T T −1
(4.3.10)
∂{d } 0

So, the stiffness matrix will be:

[k ] = EI ∫ [A ] [B] [B][A ]dx = EI [A ] ∫ [B] [B]dx[A]

L L
−1 T T −1 −1 T T −1
(4.3.11)
0 0

0 0 0 0 0  0 0 0 0 
L  L
 
0 
∫[ ]
0 0 0 
L
0 0 0
B T [B ]dx = ∫  [0 0 2 6 x ]dx = ∫ 
0 0
Now, dx =  
0 0
2 0 0 4 12 x  0 0 4L 6 L2 
 
0
6 x  0 12 x 36 x 2   
0 0 0 6 L2 12 L3 
(4.3.12)
 18

So,
0 0 0 0 
 
0 0  −1
[ ]
0 0
[k ] = EI A −1 T  [A]
0 0 4L 6 L2 
 
0 3
0 6 L2 12 L 

 3 2 
1 0 − 0  1 0 0 0 
L2 L3  0 0 0

 1  
0 1 −
2
 0 0 0 0  0 1 0 0 
[k ] = EI  L L2    3 2 3 1
2 − − − 
0 − 3  0
3 2 0 4L 6 L   L2
0 L L2 L
 L2 L   2
 1 2 1 
 1 1  0 0 6 L2 12 L3   3 − 3
0 0 −   L L2 L L2 
 L L2 

0  12 6 12 6 
0 0 0 6  1 0 0  L3 L2 
   6
L2 L3

0 6
0  0 1 0 4 6
0 0 −2  2 − 2 − 2
= EI   3 2 3 1  = EI  L L L L 
 
0 0 0 − 6 − L2 −
L L2 L
 − 12 6
− 2
12
− 2 
6
  2  L3 L3
 1 2 1  6
L L
0 0 2 6l   3 − 3 2 6 4 
 L L2 L L2   − 2 
 L2 L L L 

Thus, the element stiffness of a beam member is:

 12 6 L −12 6 L 
 6 L 4 L2 −6 L 2 L2 
EI
[ k ] = 3  −12 −6 L 12 −6 L  (4.3.13)
L
 2 
 6 L 2 L −6 L 4 L 
2

4.3.4 Generalized Stiffness Matrix of a Beam Member

Consider a beam member making an angle ‘θ’ with X axis as shown in Fig 4.3.2 below. By
resolving the forces along local X and Y direction, the following relations are obtained.
 19

=Fx1 Fx1 cos θ + Fy1 sin θ

=Fx 2 Fx 2 cos θ + Fy 2 sin θ
− Fx1 sin θ + Fy1 cos θ
Fy1 =
(4.3.14)
− Fx 2 sin θ + Fy 2 cos θ
Fy 2 =
M1 = M1
M2 = M2

Where, Fx1 and Fx 2 are the axial forces along the member axis X . Similarly, Fy1 and Fy 2 are

the forces perpendicular to the member axis X . M 1 and M 2 are the moment about its axis at
node 1 and 2 respectively.

Fig. 4.3.2 Inclined beam member

The relationship expressed in eq. (4.3.14) can be rewritten in matrix form as follows:

 Fx1   cos θ sinθ 0 0 0 0   Fx1 

  
 Fy1   − sinθ cos θ 0 0 0 0   Fy1 
 M 1   0 0 1 0 0 0   M 1  (4.3.15)
 =  
 Fx 2   0 0 0 cos θ sinθ 0   Fx 2 
 Fy 2   0 0 0 − sinθ cos θ 0   Fy 2 
    
 M 2   0 0 0 0 0 1   M 2 

Now, the above equation can be expressed in short as:

{F } = [T ]{F } (4.3.16)
 20

Similarly, the displacement vector in local coordinate system (𝑋�, 𝑌�) may be transformed to
global (𝑋, 𝑌) coordinate system by the following relation.

{d } = [T ]{d } (4.3.17)

The force-displacement relation in local coordinate system may be expressed as:

0 0 0 0 0 0 
 6 EI   
 Fx1  0 12 EI 6 EI 12 EI
0 − 3 u1
   L3 L2 L L2   
 Fy1   6 EI 4 EI 6 EI 2 EI   v1 
  0 0 − 2  
 M1   L2 L L L  θ1 
 =   (4.3.18)
 Fx2  0 0 0 0 0 0  u 
  2
  12 EI 6 EI 12 EI 6 EI   v 
 Fy2  0 − 3 − 0 − 2  2
M   L L2 L3 L  
θ
 2  6 EI 2 EI 6 EI 4 EI   2 
0 0 − 2 
 L2 L L L 

The matrices in the above equation are written with respect to the member axis. Now, the eq.
(4.3.18) can be rewritten as follows with the use of eqs. (4.3.16) and (4.3.17).

[T ]{F } = [k ][T ]{d } (4.3.19)

Or,
{F } = [T ]  k  [T ]{d }
−1
(4.3.20)

Here, the transformation matrix [T] is orthogonal. Thus, from the above relationship, the
generalized stiffness matrix can be expressed as:

[k ] = [[T ]T [k ]][T ] (4.3.21)

Considering λ = cos θ and µ = sin θ the above expression can be written as follows:
 21

0 0 0 0 0 0 
 12 6 12 6 
λ −µ 0 0  0 − 3 λ µ 0
L2  
0 0 0 0 0 0
µ   L3 L2 L
 λ 0 0 0 0  6 4 6 2  −µ λ 0 0 0 0 
 0 0 − 2 
0 0 1 0 0 0  L2 L  0 0 1 0 0 0
[ k ] = EI  
L L
 
0 0 0 λ − µ 0 0 0 0 0 0 0  0 0 0 λ µ 0
 
0 0 0 µ λ 0  12 6 12 6  0 0 0 −µ λ 0
  0 − 3 − 2 0 − 2  
0 0 0 0 0 1  L L L3 L  0 0 0 0 0 1
 6 2 6 4 
0 0 − 2 
 L2 L L L 
(4.3.22)

Thus, the generalized stiffness matrix of a beam member is derived as:

 12 µ 2 12 µλ 6µ 12 µ 2 12 µλ 6µ 
− − − −
 L3
 L3 L2 L3 L3 L2 
 12 µλ 12λ 2 6λ 12 µλ 12λ 2 6λ 
 − L3 L3 L2 L3
− 3
L L2 
 
 − 6µ 6λ 4 6µ 6λ
− 2
2 
 2
L2 L2 L 
[ k ] = EI  L 2 L L
 (4.3.23)
 − 12 µ 12 µλ 6µ 12 µ 2 12 µλ
− 3
6µ 
 L3 L3 L2 L3 L L2 
 
 12 µλ 12λ 2
− 3
6λ
− 2
12 µλ
− 3
12λ 2 6λ
− 2
 L3 L L L L3 L 
 6µ 6λ 2 6µ 6λ 4 
 − 2 − 2 
 L L2 L L2 L L 
 22

Lecture 4: Analysis of Continuous Beam

4.4.1 Equivalent Loading on Beam Member

In finite element analysis, the external loads are necessary to be acting at the joints, which
does not happen always; as some forces may act on the member. The forces acting on the
member should be replaced by equivalent forces acting at the joints. These joint forces
obtained from the forces on the members are called equivalent joint loads. These joint loads
are combined with the actual joint loads to provide the combined joint loads, which are then
utilized in the analysis.

4.4.1.1 Varying Load

Let a beam is loaded with a linearly varying load as shown in the figure below. The
equivalent forces at nodes can be expressed using finite element technique. If w(x) is the
function of load, then the nodal load can be expressed as follows.
{Q} = ∫ [ N ] w ( x ) dx
T

(4.4.1)
The loading function for the present case can be written as:
w2 − w1
w ( x=
) w1 + x
L (4.4.2)

Fig. 4.4.1 Varying load on beam

From eqs. (4.4.1) and (4.4.2), the equvalent nodal load will become
 L  2 x 3 3x 2    7 w 3w
∫  3 − 2 + 1 w( x )dx   1 + 2  L 

 F1   0L  3
L L    20 20  
    2
   1 2  2 
 M 1   ∫  L2 − L + x  w( x )dx    20 + 30  L 
x 2 x w w
{Q} =   =  0L   
 =  3w 7 w  (4.4.3)
    1 + 2  L 

3 2
 F2    − 2 x 3 x
( )
   ∫0  L3
+  w x dx
L2    20 20  
M 2   L 3   w w  2 
x x 
2

∫  2 −  w( x )dx   − 30 − 20  L 
1 2

 0  L L     
 23

Now, if w1=w2=w, then the equivalent nodal force will be:

 wL 
 2 
 
 wL 
2

 
{Q} =  12  (4.4.4)
 wL 
 2 
 2
− wL 
 12 

4.4.1.2 Concentrated Load

Consider a force F is applied at a point is regarded as a limiting case of intense pressure over
infinitesimal length, so that p(x)dx approaches F. Therefore,
{Q} [ N ] p ( x ) dx
∫=
T

N 
T
= *
 F (4.4.5)

Fig. 4.4.2 Concentrated load on beam

Here, [N*] is obtained by evaluating [N] at point where the concentrated load F
is applied. Thus,

 2 x3 3x 2   2a 3 3a 2 
 3 − 2 + 1  3 − 2 + 1
 L3 L   L3 L 
 x 2 x2   a 2a 2 
 L2 − L + x   L2 − L + a 
[ N *] = 3 2 
at dista nce a  3 2  (4.4.6)
 − 2 x + 3x   − 2a + 3a 
 L3 L2   L3 L2 
 3 2   3 2 
 x −x   a −a 
 L2 L   L2 L 
 24

 2a 3 3a 2 
 3 − 2 + 1
 F1   3
L L 
    a 2 a 2
 
 M 1    L2 − L + a  
Therefore, {Q} =   =   
 (4.4.7)
 F2    −  2 a 3
3 a 2
 
+ 2 

    L3 L 
M 2   
  a − a  
3 2

  L2 L  

Now, if load F is acting at midspan (i.e., a=L/2), then equivalent nodal load will be

 F 
 2 
 
 FL 
 
{Q} =  8  (4.4.8)
 F 
 2 
 FL 
− 
 8 

With the above approach, the equivalent nodal load can be found for various loading function
acting on beam members.

4.4.2 Worked Out Example

Analyze the beam shown below by the stiffness method. Assume the moment of inertia of
member 2 as twice that of member 1. Find the bending moment and reactions at supports of
the beam assuming the length of span, L as 4 m, concentrated load (P) as 15 kN and udl, w as
4 kN/m.

Fig. 4.4.3 Example of a continuous beam

 25

Solution
Step 1: Numbering of Nodes and Members

The analysis of beam starts with the numbering of members and joints as shown below:

Fig. 4.4.4 Numbering of nodes and members

The member AB and BC are designated as (1) and (2). The points A,B,C are designated
by nodes 1, 2 and 4. The member information for beam is shown in tabulated form as shown
in Table 4.4.1. The coordinate of node 1 is assumed as (0, 0). The coordinate and restraint
joint information are shown in Table 4.4.2. The integer 1 in the restraint list indicates the
restraint exists and 0 indicates the restraint at that particular direction does not exist. Thus, in
node no. 2, the integer 0 in rotation indicates that the joint is free rotation.

Table 4.4.1Member Information for Beam

Member Starting node Ending node Rigidity modulus

number

1 1 2 EI
2 2 3 2EI

Table 4.4.2 Nodal Information for Beam

Node No. Coordinates Restraint List

x y Vertical Rotation
1 0 0 1 1
2 L 0 1 0
3 2L 0 1 0
 26

Step 2: Formation of member stiffness matrix:

The local stiffness matrices of each member are given below based on their individual
member properties and orientations. Thus the local stiffness matrix of member (1) is:

1 2 3 4

 12 EI 6 EI 12 EI 6 EI 
 L3 − 1
L2 L3 L2 
 6 EI 4 EI 6 EI 2 EI 
 − 2  2
[k ]1 =  12
L2 L L L 
− EI 6 EI
− 2
12 EI
− 2 
6 EI 3
 L3 L L3 L 
 6 EI 2 EI 6 EI 4 EI 
 − 2  4
 L2 L L L 

Similarly, the local stiffness matrix of member (2) is:

3 4 5 6
 24 EI 12 EI 24 EI 12 EI 
 L3 − 3
L2 L3 L2 
 12 EI 8 EI 12 EI 4 EI 
 − 2  4
[k ]2 = L L 
2
L L
− 24 EI 12 EI
− 2
24 EI
− 2 
12 EI 5
 L3 L L3 L 
 12 EI 4 EI 12 EI 8 EI 
 − 2  6
 L2 L L L 

Step 3: Formation of global stiffness matrix:

The global stiffness matrix is obtained by assembling the local stiffness matrix of members
(1) and (2) as follows:

1 2 3 4 5 6

 12 EI 6 EI 12 EI 6 EI 
 L3 − 0 0  1
L2 L3 L2
 6 EI 4 EI 6 EI 2 EI 
 2
− 2 0 0  2
 L L L L 
− 12 EI 6 EI 36 EI 6 EI 24 EI 12 EI 
− 2 − 3
[K ] 
= L
3
L L L2 L3 L2 
6 EI 2 EI 6 EI 12 EI 12 EI 4 EI 
 2
− 2  4
 L L L2 L L L 
 0 − 2 
24 EI 12 EI 24 EI 12 EI
0 − 3 − 2 5
 L L L3 L 
 12 EI 4 EI 12 EI 8EI 
 0 0 − 2  6
 L2 L L L 
 27

Step 4: Boundary condition:

The boundary conditions according to the support of the beam can be expressed in terms of
the displacement vector. The displacement vector will be as follows

 0
 0
 
 0
{d } =  
θ 2 
 0
 
θ 3 
 

Step 5: Load vector:

The concentrated load on member (1) and the distributed load on member (2) are replaced by
equivalent joint load. The equivalent joint load vector can be written as

Fig. 4.4.5 Equivalent Load

 P 
 − 
2
 PL 
 − 
 8 
  P wL  
 −  2 + 2  
{F } =  PL wL2 
 − 
 8 12 
 wL 
 − 
 2 
 wL2 
 12 

Step 6 : Determination of unknown displacements:.

The unknown displacement can be obtained from the relationship as given below:
 28

{ F } = [ K ] {d }
{d } = [ K ] { F }
−1

 12 EI 
−1  P 
6 EI 12 EI
− 3
6 EI
0 0   − 
 L3 L2 L L2 
2

   −
PL

 0   6 EI 4 EI
− 2
6 EI 2 EI
0 0 
 0   L2 L L L   8 
   12 EI 12 EI    P wL  
 −  + 2  
6 EI 36 EI 6 EI 24 EI
 0  − L3 − 2
L L L2
−
L3 L2  ×   2
  =  6 EI 4 EI   PL wL2 
θ 2  
2 EI 6 EI 12 EI 12 EI
− 2   − 
0  L  8 12 
2
L L2 L L L 
   0 24 EI 12 EI 24 EI 12 EI  
θ 3   0 − 3 − 2 − 2   −
wL

L L L3 L 
 12 EI 4 EI 12 EI 8EI   2
2

 0 0 − 2   wL 
 L2 L L L   12 

The above relation may be condensed into following

−1
12 EI 4 EI   PL wL2   PL wL2 
 −  −
θ 2   L L  8 12  L  2 −1  8 
12 
 =    × =
  −1 3   
θ3   4 EI 8 EI     wL
2 2
 wL  20 EI 
 L L   12   12 

 PL wL2 
θ 2  L  4 − 4 
  =  2 
θ 3  20 EI − PL + wL 
 8 3 

PL2 wL3
θ2 = −
80 EI 80 EI
PL wL3
θ3 = − +
160 EI 60 EI

Step 7: Determination of member end actions:

The member end actions can be obtained from the corresponding member stiffness and the
nodal displacements. The member end actions for each member are derived as shown below.

Member-(1)
 29

 12 EI 6 EI   3P 3wL 
 40 − 40 
6 EI 12 EI
 L3 − 
L2 L3   
2

PL wL2 
L 0
 F1   6 EI 2 EI  
M  
4 EI 6 EI
− 2  0   −
 1 L  L2 L L L ×  =  40 40 
=   3P 3wL 
  6 EI   0
 F2  20 EI − 3
12 EI 6 EI 12 EI
 − 2 − 2  PL wL  −
 2 + 
M 2  L L L3 L  −   40 40 
 6 EI 2 EI 6 EI 4 EI   4 4   PL wL2 
 − 2   20 − 20 
 L2 L L L 

Member-(2)
 24 12 24 12   wL 3P 
− +
 L2 L L2 L   0   20 40 
 F2   12 12   PL wL2   3PL wL2 
M   − 4   −  − 
4  =  40
8
 2  EI  L
  = L  ×  4  30 
12   wL 3P 
 F3  L − 2
24 12 24 0
− −  2  − −
 M 3   L L L2 L  − PL + wL   20 40 
 12   8 3   wL2 
8  
12
 4 − 
 L L   12 

Actual member end actions:

Member (1)
 3P 3wL   P   23P 3wL 
 40 − 40  
 F1   2 
2   40 − 40 
   2 
   PL − wL   PL   6 PL − wL 
 M 1   40 40   8  
  =  3P 3wL  +  P  =  40 40 
 2  −
F +     17 P 3wL 
+ 
M   40 40   2   40 40 
 2   PL wL2   PL 
 3PL wL2 
 20 − 20  − 8  − 40 − 20 
 
Member (2)
 wL 3P   wL 
 20 + 40   2  11wL + 3P 
 F2   2  2   20 40 
   3PL − wL   wL   3PL wL2 
M 2   40 12  =  + 
  =  wL 3P  +  wL
30
  40 20 
 3  −
F −  9 wL 3P 
 M 3   20 40   2   − 
  wL2
wL 2
 20 40 
  −  
  12  
0
 12
 23P 3wL 
−

 R A   40 40 

The support reactions at the supports A, B and C are {FR } =  RB  =  25wL + P 
 R   40 2
 C   9wL 3P 
+
 20 4 
Putting the numerical values of L, P and w (P=15, L=4, w=4) the member actions and support
reactions will be as follows:
 30

Member end actions:

 F2  9.925  F1  7.425
       
M 2   7.7   M 1   7.4 
 =  ,  = 
 F3  6.075  F2  7.575
       
 M 3   0  M 2   − 7.7 

Support reactions:
 R A  7.425
{FR } =  RB  =  17.5 
 R  6.075
 C  
 31

Lecture 5: Plane Frame Analysis

4.5.1 Introduction
The plane frame is a combination of plane truss and two dimensional beam. All the members
lie in the same plane and are interconnected by rigid joints in case of plane frame. The
internal stress resultants at a cross-section of a plane frame member consist of axial force,
bending moment and shear force.

4.5.2 Member Stiffness Matrix

In case of plane frame, the degrees of freedom at each node will be (i) axial deformation, (ii)
vertical deformation and (iii) rotation. Thus the frame members have three degrees of
freedom at each node as shown in Fig. 4.5.1 below.

Fig. 4.5.1 Plane frame element

Therefore, the stiffness matrix of the frame in its local coordinate system will be the
combination of 2-d truss and 2-d beam matrices:
u1 v1 θ1 u2 v2 θ2

𝐴𝐸 𝐴𝐸
⎡ 𝐿 0 0 − 0 0 ⎤
𝐿
⎢ 0 12𝐸𝐼 6𝐸𝐼
0 −
12𝐸𝐼 6𝐸𝐼 ⎥
⎢ 𝐿3 𝐿2 𝐿3 𝐿2 ⎥
6𝐸𝐼 4𝐸𝐼 6𝐸𝐼 2𝐸𝐼
⎢ 0 0 − ⎥
���� 𝐿2 𝐿2
[𝑘]=⎢ 𝐴𝐸 𝐿
𝐴𝐸
𝐿 ⎥
(4.5.1)
⎢− 0 0 0 0 ⎥
⎢ 𝐿 12𝐸𝐼 6𝐸𝐼
𝐿
12𝐸𝐼 6𝐸𝐼 ⎥
⎢ 0 −
𝐿3
−
𝐿2
0
𝐿3
− 2⎥
𝐿
⎢ 6𝐸𝐼 2𝐸𝐼 6𝐸𝐼 4𝐸𝐼 ⎥
⎣ 0 𝐿2 𝐿
0 −
𝐿2 𝐿 ⎦

4.5.3 Generalized Stiffness Matrix

In plane frame the members are oriented in different directions and hence it is necessary to
transform stiffness matrix of individual members from local to global co-ordinate system
 32

before formulating the global stiffness matrix by assembly. The generalized stiffness matrix
of a frame member can be obtained by transferring the matrix of local coordinate system into
its global coordinate system. The transformation matrix can be expressed as:

cos 𝜃 sin 𝜃 0 0 0 0
⎡− sin 𝜃 cos 𝜃 0 0 0 0⎤
⎢ ⎥
0 0 1 0 0 0⎥
[T]=⎢ (4.5.2)
⎢ 0 0 0 cos 𝜃 sin 𝜃 0⎥
⎢ 0 0 0 − sin 𝜃 cos 𝜃 0⎥
⎣ 0 0 0 0 0 1⎦

Now, the generalized stiffness matrix of the member can be obtained from the relation of
� ][𝑇] . Thus considering 𝜆 = cos 𝜃 and 𝜇 = sin 𝜃 the stiffness matrix in global
[𝐾] = [𝑇]𝑇 [𝐾
coordinate system can be written as follows:
AE AE
⎡ 0 0 − 0 0 ⎤
⎢ L L ⎥
⎢ 0 12EI 6EI 12EI 6EI ⎥
0 − 3
λ −µ 0 0 0 0 ⎢ L3 L2 L L2 ⎥
⎡µ λ 0 0 0 0⎤ ⎢ 6EI 4EI 6EI 2EI ⎥
⎢ ⎥ ⎢ 0 0 − 2
[K] = EI ⎢0 0 1 0 0 0⎥ × ⎢ L 2 L L L ⎥
⎢0 0 0 λ −µ 0⎥ − AE AE ⎥
⎢0 0 0 µ λ 0⎥ ⎢ 0 0 0 0 ⎥
⎢ L L ⎥
⎣0 0 0 0 0 1⎦ 12EI 6EI 12EI 6EI
⎢ 0 − 3 − 2 0 − 2⎥
⎢ L L L3 L ⎥
⎢ 6EI 2EI 6EI 4EI ⎥
⎣ 0 L 2 L
0 − 2
L L ⎦
λ µ 0 0 0 0
⎡−µ λ 0 0 0 0⎤
⎢ ⎥
⎢ 0 0 1 0 0 0⎥
×
⎢ 0 0 0 λ µ 0⎥
⎢ 0 0 0 −µ λ 0⎥
⎣ 0 0 0 0 0 1⎦

EA 12EI EA 12EI 6EI EA 2 12EI 2 EA 12EI 6EI

⎡ � λ2 + 3 μ2 � � λμ − 3 λμ� − μ �− λ − 3 μ � �− λμ + 3 λμ� − μ⎤
⎢ L L L L L2 L L L L L2 ⎥
EA
⎢ � λμ − 12EI EA 12EI 6EI EA 12EI EA 2 12EI 2 6EI ⎥
λμ� � μ2 + 3 λ2 � λ �− λμ + 3 λμ� �− μ − 3 λ � λ
⎢ L L3 L L L2 L L L L L2 ⎥
⎢ 6EI 6EI 4EI 6EI 6EI 2EI ⎥
⎢ − 2 μ λ μ − 2 λ
=⎢ L L2 L L2 L L ⎥
EA 2 12EI 2 EA 12EI 6EI EA 2 12EI 2 EA 12EI 6EI ⎥
⎢ �− λ − 3 μ � �− λμ + 3 λμ� μ � λ + 3 μ � � λμ − 3 λμ� μ ⎥
⎢ L L L L L2 L L L L L2 ⎥
⎢�− EA 12EI EA 12EI 6EI EA 12EI EA 2 12EI 2 6EI
λμ + 3 λμ� �− μ − 3 λ2 �
2
− 2 λ � λμ − 3 λμ� � μ + 3 λ � − 2 λ⎥
⎢ L L L L L L L L L L ⎥
⎢ 6EI 6EI 2EI 6EI 6EI 4EI ⎥
⎣ − 2 μ λ μ − 2 λ
L L2 L L2 L L ⎦
 33

(4.5.3)

4.5.4 Worked Out Example

Analyse the plane frame shown below by the stiffness method. Assume the modulus of
elasticity of the horizontal member is 1.5 times that of the vertical member and length of
the vertical member is 1.5 times that of horizontal member. Find the bending moment and
reactions at support assuming the length, cross section area and modulus of elasticity of
vertical member as 3.0 m, 0.4 x 0.4 m2 and 2 x 1011 N/mm2, respectively.

Fig. 4.5.2 Plane frame

Solution

Step 1: Numbering of Nodes and Members

The numbering of members and joints of the plane frame are as shown below:

Fig. 4.5.3 Numbering of Nodes and Members

The members AB and BC are designated as (1) and (2). The points A, B and C are designated
by nodes 1, 2 and 3. The member information for the frame is shown in tabulated form as
shown in Table 1(a). The coordinate of node 1 is assumed as (0,0). The coordinate and
restraint joint information are shown in Table 1(b). The integer 1 in the restraint list indicates
 34

the restraint exists and 0 indicates the restraint at that particular direction does not exist.
Thus, in node no. 2, the integer 0 all the restraint type indicates that the joint is in free all the
three directions.

Table 4.5.1 Member Information for Beam

Member number Starting node Ending node Rigidity modulus
1 1 2 EI
2 2 3 1.5EI

Table 4.5.2 Nodal Information for Beam

Node no. Coordinates Restraint list
X Y Axial Vertical Rotation
1 0 0 1 1 1
2 0 1.5L 0 0 0
3 L 1.5L 1 1 1

Step 2: Formation of member stiffness matrix:

The individual member stiffness matrices can be found out directly from eqn. shown above.
Thus the stiffness matrices of each member in global coordinate system are given below
based on their individual member properties and orientations. Thus the stiffness matrix of
member (1) is:

1 2 3 4 5 6
12EI 6EI 12EI 6EI
⎡ (1.5L)3 0 − − 0 − 1
(1.5L)2 (1.5L)3 (1.5L)2 ⎤
⎢ AE AE ⎥
⎢ 0 0 0 − 0 2
(1.5L) (1.5L) ⎥
⎢ 6EI 4EI 6EI 2EI ⎥
−
⎢ (1.5L)2 0 0 ⎥ 3
(1.5L) (1.5L)2 1.5L
[𝑘]1 =⎢ 12EI 6EI 12EI 6EI ⎥
⎢− (1.5L)3 0
(1.5L)2 (1.5L)3
0
(1.5L)2 ⎥ 4
⎢ AE AE ⎥
⎢ 0 −
(1.5L)
0 0
(1.5L)
0 ⎥ 5
⎢ 6EI 2EI 6EI 4EI ⎥
⎣− (1.5L)2 0
(1.5L) (1.5L)2
0
(1.5L) ⎦ 6

Similarly, the stiffness matrix of member (2) is :

4 5 6 7 8 9
 35

A(1.5 E) A(1.5 E)
⎡ L
0 0 − L
0 0 ⎤ 4
⎢ 0
12(1.5 E)I 6(1.5 E)I
0 −
12(1.5 E)I 6(1.5 E)I ⎥
5
⎢ L3 L2 L3 L2 ⎥
⎢ 6(1.5 E)I 4(1.5 E)I 6(1.5 E)I 2(1.5 E)I ⎥
0 2
0 − L2
6
[𝑘]2 = ⎢ A(1.5 E) L L
A(1.5 E)
L ⎥
⎢− 0 0 0 0 ⎥ 7
L L
⎢ 12(1.5 E)I 6(1.5 E)I 12(1.5 E)I 6(1.5 E)I
⎥
⎢ 0 − L3 − 0 − L2 ⎥ 8
L2 L3
⎢ 6(1.5 E)I 2(1.5 E)I 6(1.5 E)I 4(1.5 E)I ⎥ 9
⎣ 0 L2 L
0 − L2 L ⎦

Step 3 : Formulation of global stiffness matrix:

The global stiffness matrix is obtained by assembling by assembling the local stiffness matrix
of member (1) and (2) as follows:
1 2 3 4 5 6 7 8 9
32EI 8EI 32EI 8EI 1
⎡ 9L3 0 − 2 − 0 − 0 0 0 ⎤
3L 9L3 3L2
⎢ 0 2AE
0 0 −
2AE
0 0 0 0 ⎥ 2
⎢ 3L 3L ⎥
⎢ 8EI2 0
8EI 8EI
0
4EI
0 0 0 ⎥ 3
3L 3L 3L2 3L
⎢ 32EI 8EI 32EI 1.5EA 8EI 1.5EA
⎥
⎢− 9L3 0 ( 3 + ) 0 − 0 0 ⎥ 4
3L2 9L L 3L2 L
⎢ 2AE 2AE 18EI 9EI 18EI 9EI ⎥ 5
[K] = ⎢ 0 − 0 0 ( + ) 0 −
3L 3L L3 L2 L3 L2 ⎥
⎢ 8EI 4EI 8EI 9EI 8EI 6EI 9EI 3EI ⎥ 6
− 0 ( 2+ ) 0 − 2
⎢ 3L2 3L 3L2 L2 3L L L L ⎥
⎢ 0 1.5AE 1.5AE
0 0 − O O 0 0 ⎥ 7
⎢ L L ⎥
18EI 9EI 18EI 9EI 8
⎢ 0 0 0 0 − − 0 − 2⎥
L3 L2 L3 L
⎢ 9EI 3EI 9EI 6EI ⎥ 9
⎣ 0 0 0 0
L2 L
0 − 2
L L ⎦
Step 4: Boundary conditions:
The boundary conditions according to the support of the frame can be expressed in terms of
the displacement vector. The displacement vector will be as follows:
0
⎡ 0 ⎤
⎢ 0 ⎥
⎢ ⎥
⎢𝛿𝑥𝐵 ⎥
{𝑑} = ⎢𝛿𝑦𝐵 ⎥
⎢ 𝜃𝐵 ⎥
⎢ 0 ⎥
⎢ 0 ⎥
⎣ 0 ⎦
Here, 𝛿𝑥𝐵 , 𝛿𝑦𝐵 and 𝜃𝐵 indicate the displacement in X-direction, displacement in Y-direction
and rotation at point B.

Step 5: Load vector:

 36

The distributed load on member (2) can be replaced by its equivalent joint load as shown in
the figure below.

Fig. 4.5.4 Equivalent Joint Loads

Thus, the equivalent joint load vector can be written as

0
⎡ 0 ⎤
⎢ 0 ⎥
⎢ ⎥
⎢ 0 ⎥
⎢ − 𝑤𝐿 ⎥
⎢ 2 ⎥
{𝐹} = ⎢ 𝑤𝐿2 ⎥
⎢− 12 ⎥
⎢ 0 ⎥
⎢ 𝑤𝐿 ⎥
⎢− ⎥
⎢ 2 ⎥
⎢ 𝑤𝐿2 ⎥
⎣ 12 ⎦
Step 6: Determination of unknown displacements:
The unknown displacements can be obtained from the relationship of {F} = [K]{d} or
{d} =[k]-1 {F}. Now eliminating the rows and columns in the stiffness matrix and force
matrix, corresponding to zero elements in displacement matrix, the reduced matrix will be as
follows.
−1
32𝐸𝐼 1.5𝐸𝐴 8𝐸𝐼
⎡� 3 + � 0 ⎤ 0
𝛿𝑥𝐵 ⎢ 9𝐿 𝐿 3𝐿2 ⎥ ⎡ 𝑤𝐿 ⎤
�𝛿𝑦𝐵 � = ⎢ 0 �
2𝐴𝐸 18𝐸𝐼
+ 3 �
9𝐸𝐼 ⎥ ⎢− 2 ⎥
3𝐿 𝐿 𝐿2 ⎢ ⎥
𝜃𝐵 ⎢ ⎥ ⎢ 𝑤𝐿2 ⎥
⎢ 8𝐸𝐼 9𝐸𝐼 8𝐸𝐼 6𝐸𝐼 ⎥ −
� + �⎦ ⎣ 12 ⎦
⎣ 3𝐿 2 𝐿2 3𝐿 𝐿
Thus, the unknown displacements will be:
𝛿𝑥𝐵 1 0.04327𝑤
�𝛿𝑦𝐵 � = 10 �−1.7127𝑤 �
10
𝜃𝐵 −5.4978𝑤
Step 7: Determination of member end actions:
The member end actions can be obtained from the corresponding member stiffness and the
nodal displacements. The member end actions for each member are derived as shown below.
 37

Member – (1)
In case of member (1), the member forces will be:{𝐹𝑚 }1 = [𝐾](1) {𝑑}(1)
𝐹𝑥
⎡ 1⎤ 56.17 0 −126.4 −56.17 0 −126.4
𝐹𝑦 ⎡ 0 7110 0 0 −7110 0 ⎤
⎢ 1⎥ ⎢ ⎥
⎢ 𝑀1 ⎥ −126.4 0 379.2 126.4 0 189.6 ⎥
⎢𝐹𝑥 2 ⎥ = 10 ⎢⎢−56.17
6
0 379.2 56.17 0 126.4 ⎥
⎢ ⎥ ⎢
𝐹
⎢ 𝑦 2⎥ 0 −7110 0 0 7110 0 ⎥
⎣−126.4 0 189.6 126.4 0 379.2 ⎦
⎣ 𝑀2 ⎦
0
⎡ 0 ⎤
⎢ ⎥
0
×⎢ −12
⎥
⎢ 4.327 𝑋 10 𝑤 ⎥
⎢−1.7127 𝑋 10−10 𝑤 ⎥
⎣−5.4978 𝑋 10−10 𝑤 ⎦

0.0697𝑤
⎡ 1.2177𝑤 ⎤
⎢ ⎥
−0.10479𝑤 ⎥
=⎢
⎢−0.06925𝑤 ⎥
⎢−1.21661𝑤 ⎥
⎣−0.20793𝑤 ⎦

It is to be noted that {Fm} are the end actions due to joint loads. Hence it must be added to the
corresponding end actions in the restrained structure in order to obtain the end actions due to
the loads. Therefore, {Fm}actual are the true member end actions due to actual loading system
can be expressed as
{𝐹𝑚 }𝑎𝑐𝑡𝑢𝑎𝑙 = {Fm} + {Ffm}

Where, {Ffm} are the end actions in the restrained structure. Since there is no load acting on
member (1), the actual end actions will be:

0.0697𝑤 0 0.0697𝑤
⎡ 1.2177𝑤 ⎤ ⎡0⎤ ⎡ 1.2177𝑤 ⎤
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
{Fm}actual = ⎢−0.10479𝑤 ⎥ + ⎢0⎥ = ⎢−0.10479𝑤 ⎥
⎢−0.06925𝑤 ⎥ ⎢0⎥ ⎢−0.06925𝑤 ⎥
⎢−1.21661𝑤 ⎥ ⎢0⎥ ⎢−1.21661𝑤 ⎥
⎣−0.20793𝑤 ⎦ ⎣0⎦ ⎣−0.20793𝑤 ⎦

Member (2)
In similar way, the member forces in member (2) will be {Fm}(2) = [K](2){d}(2)
 38

𝐹𝑥
⎡ 2⎤ 16 0 0 −16 0 0 4.327 𝑋 10−12 𝑤
⎡ 0 ⎤ ⎡ ⎤
⎢ 𝐹𝑦2 ⎥ 0.284 0.426 0 −0.284 0.426 −1.7127 𝑋 10−10 𝑤
⎢ 𝑀2 ⎥ ⎢ ⎥ ⎢ ⎥
9⎢ 0 0.426 0.853 0 −0.426 0.426 ⎥ ⎢−5.4978 𝑋 10−10 𝑤 ⎥
⎢𝐹𝑥 3 ⎥ = 10 ×
⎢ −16 0 0 16 0 0 ⎥ ⎢ 0 ⎥
⎢ ⎥ ⎢ ⎥
⎢𝐹𝑦 3 ⎥ 0 −0.284 −0.426 0 0.284 −0.426 ⎢ 0 ⎥
⎣ 0 0.426 0.426 0 −0.426 0.853 ⎦ ⎣ 0 ⎦
⎣ 𝑀3 ⎦
0.069232𝑤
⎡−0.28325𝑤 ⎤
⎢ ⎥
−0.54215𝑤 ⎥
=⎢
⎢−0.06923𝑤 ⎥
⎢ 0.283245𝑤 ⎥
⎣ −0.3076𝑤 ⎦

The actual member forces in the member (2) will be:

0.069232𝑤 0 0.0692𝑤
⎡−0.28325𝑤 ⎤ ⎡ 1.5𝑤 ⎤ ⎡ 1.2167𝑤 ⎤
⎢ ⎥ ⎢ ⎥ ⎢ ⎥
−0.54215𝑤 ⎥ ⎢ 0.75𝑤 ⎥ ⎢ 0.2078𝑤 ⎥
{Fm}actual = ⎢ + =
⎢−0.06923𝑤 ⎥ ⎢ 0 ⎥ ⎢−0.0692𝑤 ⎥
⎢ 0.283245𝑤 ⎥ ⎢ 1.5𝑤 ⎥ ⎢ 1.7832𝑤 ⎥
⎣ −0.3076𝑤 ⎦ ⎣−0.75𝑤 ⎦ ⎣−1.0576𝑤 ⎦

Lecture 6 Analysis of Grid and Space Frame

4.6.1 Introduction
The property of a grid member is basically a combination of 2-d beam with torsional effect.
The plane frame is assumed to be loaded in its own plane where as loading in the grid is
normal to its plane. As a result torsional effects are included in the grid analysis. Thus the
grid member can withstand bending moment, shear force as well as torsional moment.

4.6.2 Element Stiffness Matrix for Grid Members

The degrees of freedom at each node of the grid member will be (i) vertical deformation and
(ii) rotation in two different directions.
 39

Fig. 4.6.1 Degrees of freedom of grid element

Therefore, the stiffness matrix of the grid in its local coordinate system will be:

θx1 θz1 v1 θx2 θy2 v2

 GI x GI x 
 L 0 0 − 0 0 
Mx1
L
 4 EI y 6 EI y 2 EI y 6 EI y 
 0 − 0  Mz1
 L L 2
L L2 
 6 EI y 12 EI y 6 EI y 12 EI y 
 0 − 0 − −  Fy1
[k ] =  GI L2 L3
GI x
L2 L3  (4.6.1)
− x 0 0 0 0  Mx2
 L L 
 2 EI y 6 EI y 4 EI y 6 EI y 
 0 − 2
0  Mz2
 L L L L2 
 0 6 EI y 12 EI y 6 EI y 12 EI y 
− 0 Fy2
 L2 L3 L2 L3 

Here, the G is the modulus of torsional rigidity.

4.6.3 Generalized Stiffness Matrix
The generalized stiffness matrix of a grid member can be obtained by transferring the matrix
of local coordinate system into its global coordinate system. The transformation matrix can
be expressed as:

cos 𝜃 sin 𝜃 0 0 0 0
⎡− sin 𝜃 cos 𝜃 0 0 0 0⎤
⎢ ⎥
0 0 1 0 0 0⎥
[T]=⎢
⎢ 0 0 0 cos 𝜃 sin 𝜃 0⎥
⎢ 0 0 0 − sin 𝜃 cos 𝜃 0⎥
⎣ 0 0 0 0 0 1⎦
 40

Now, the generalized stiffness matrix of the member can be obtained from the relation
of [𝑘] = [𝑇]𝑇 �𝑘��[𝑇] . Thus considering 𝜆 = cos 𝜃 and 𝜇 = sin 𝜃 the stiffness matrix in
global coordinate system can be written as follows:

xGI GIx
⎡ L 0 0 − 0 0 ⎤
L
𝜆 −𝜇 0 0 0 0 ⎢ 0 4EIy
−
6EIy
0
2EIy 6EIy ⎥
⎡𝜇 ⎢ L2 ⎥
0⎤ ⎢
L L2 L
𝜆 0 0 0 6EIy 12EIy 6EIy 12EIy
⎢
0 0 1 0 0
⎥
0⎥ ⎢ 0 − L2 0 − − L3 ⎥
L3 L2 ⎥
[𝑘] = ⎢ × GIx GIx
⎢0 0 0 𝜆 −𝜇 0 ⎥ ⎢− 0 0 0 0 ⎥
⎢0 0 0 𝜇 𝜆 0⎥ ⎢ L 2EIy 6EIy
L
4EIy 6EIy ⎥
⎣0 0 0 0 0 1⎦ ⎢ 0 − 0 − L2 ⎥
L L2 L
⎢ 6EIy 12EIy 6EIy 12EIy ⎥
⎣ 0 L2
− L3
0 L2 L3 ⎦
𝜆 𝜇 0 0 0 0
⎡−𝜇 𝜆 0 0 0 0⎤
⎢ ⎥
0 0 1 0 0 0⎥
×⎢
⎢ 0 0 0 𝜆 𝜇 0⎥
⎢0 0 0 −𝜇 𝜆 0⎥
⎣0 0 0 0 0 1⎦

 GI x 2 4 EI y 2  GI x 4 EI y  6 EI y GI x 2 4 EI y 2  GI 2 EI y  6 EI y 
  −  x + λµ
 λ + µ  L − L λµ L2
µ − λ + µ  −
L2
µ
 L L   L L  L L  
  GI x 4 EI y  GI x 2 4 EI y 2 6 EI y  GI 2 EI y  GI x 2 2 EI y 2 6 EI y 
  − λµ µ + λ − λ −  x + λµ − µ + λ λ 
  L L  L L L2  L L 
 L L L2 
 6 EI y 6 EI y 12 EI y 6 EI y 6 EI y 12 EI y 
 2
µ − 2 λ µ − 2 λ − 
=  L L L3 L2 L L3 
 GI x 2 4 EI y 2  GI x 2 EI y  6 EI y GI x 2 4 EI y 2  GI x 4 EI y  6 EI y 
− L λ + L µ −  + λµ µ λ + µ 
 L − L λµ
 − 2 µ
 L L  L2 L L   L
 
  GI x 2 EI y  GI 2 EI y 2 6 EI y  GI x 4 EI y  GI x 2 4 EI 6 EI y 
  + λµ − x µ2 + λ − λ 
 L − L λµ
 µ + λ2 λ 
y
 L2 L2
  L L  L L   L L 
 6 EI y 6 EI y 12 EI y 6 EI y 6 EI y 12 EI y 
 − 2 µ λ − − 2 µ λ 
 L L2 L3 L L2 L3 

(4.6.2)

4.6.4 Worked Out Example

Analyze the grid shown below by the stiffness method. Draw the shear force and bending
moment diagram assuming the cross sectional area and modulus of elasticity of each member
as 0.3×0.3 m2 and 2×1011 N/m2 respectively. Assume EI = 3GJ. The length of member AB
and BC is 4 m and 5 m respectively.
 41

Fig. 4.6.2 Grid structure

Solution

Step 1: Numbering of Nodes and Members

The numbering of members and joints of the plane frame are as shown in the figure below:

Fig. 4.6.3 Numbering of nodes and members

The member AB and BC are designated as (1) and (2). The points A, B and C are designated
by nodes 1, 2 and 3. The member information for the grid is shown in tabulated form as
shown in Table 4.6.1. The coordinate of node 1 is assumed as (0, 0). The coordinate and
restraint joint information are shown in Table 4.6.2. The integer 1 in the restraint list indicates
the restraint exists and 0 indicates the restraint at that particular direction does not exist.
Thus, in node no. 2, the integer 0 all the restraint type indicates that the joint is free in all the
three directions.

Table 4.6.1 Member Information

Member number Starting node Ending node
1 1 2
2 2 3
 42

Table 4.6.2 Member Coordinates

ode coordinates Restraint list
No x z Vertical Rotation Rotation
1 0 0 1 1 1
2 4 0 0 0 0
3 4 5 1 1 1

Step 2: Formation of member stiffness matrix:

The individual member stiffness matrices can be found out directly. Thus the stiffness
matrices of each member in global coordinate system are given below based on their
individual member properties and orientations. As the member AB is horizontal, i.e., θ = 0,
the values of Cos θ = 1 and Sin θ = 0. Thus the stiffness matrix of member (1) is:

1 2 3 4 5 6
 GJ GJ  1
 L 0 0 − 0 0 
L
 12 EI 6 EI 12 EI 6 EI 
 0 0 −  2
 L3 L2 L3 L2 
 0 6 EI 4 EI 6 EI 2 EI 
0 − 2 3
[k ]AB 
= L2 L L L 
GJ GJ 
− 0 0 0 0  4
 L L 
 0 12 EI 6 EI 12 EI 6 EI 
− − 0 − 2 5
 L3 L2 L3 L 
 6 EI 2 EI 6 EI 4 EI 
 0 0 − 2  6
 L2 L L L 

Assuming EI=3GJ=3K, the above equation can be written as

1 2 3 4 5 6
 K K 
 L 0 0 − 0 0 
1
L
 36 K 18 K 36 K 18 K 
 0 0 −  2
 L3 L2 L3 L2 
 0 18 K 12 K 18 K 6K 
0 − 2 3
[k ]AB 
= L2 L L L 
K K 
− 0 0 0 0  4
 L L 
 0 − 2 
36 K 18 K 36 K 18 K
− − 0
 L3 L2 L3 L  5
 18 K 6K 18 K 12 K 
 0 0 − 2 
 L2 L L L  6
 43

As the member BC member is also horizontal, the value of Cos θ = 1 and Sin θ = 0 and thus,
the stiffness matrix will be:

6 5 4 7 8 9
 K K 
 L 0 0 − 0 0 
6
L
 36 K 18 K 36 K 18 K 
 0 0 −  5
 L3 L2 L3 L2 
 0 18 K 12 K 18 K 6K 
0 − 2 4

[k ]BC =  K L 2
L L L 
K 
− 0 0 0 0  7
 L L 
 0 36 K 18 K 36 K 18 K 
− 3 − 2 0 − 2
 L L L3 L  8
 18 K 6K 18 K 12 K 
 0
Step 3: Formation of global stiffness 0
matrix: − 2 
 L 2
L L 9 L 
The global stiffness matrix can be obtained by assembling the local stiffness matrix of
members (AB) and (BC). Now looking at the grid structure, the displacements at the fixed
supports, are known and all are equal to zero. Only the displacement at co-ordinates 4, 5, 6
are unknown. So the global system stiffness matrix, corresponding to the displacement at co-
ordinate 4, 5, 6 will be:

𝐾 12𝐾 18𝐾
⎡𝐿𝐴𝐵 𝐿𝐵𝐶 + 0 ⎤
𝐿𝐵𝐶 2
⎢ 18𝐾 36𝐾 36𝐾 18𝐾 ⎥
[𝐾] = ⎢ 𝐿 2 + − 2 ⎥
𝐵𝐶 𝐿𝐴𝐵 3 𝐿𝐵𝐶 3 𝐿𝐴𝐵
⎢ 18𝐾 𝐾 12𝐾 ⎥
⎣ 0 − +
𝐿𝐴𝐵 2 𝐿𝐵𝐶 𝐿𝐴𝐵 ⎦

2.65 0.72 0
= K �0.72 0.8505 −1.125�
0 −1.125 3.2

Step 4: Boundary condition:

The boundary conditions according to the support of the grid structure can be expressed in
terms of the displacement vector. The displacement vector will be as follows
 44

0
0
 
0
 
d 4 
{d } =  d5 
d 
 6
0
 
0
0
 
Here, d4, d5, and d6 indicate the displacement vectors at point B.

Step 5: Load vector:

The distributed load on member (1) can be replaced by its equivalent joint load as shown in
the figure below.

Fig. 4.6.4 Equivalent load

Thus the equivalent load vector will be:

 0 
 wL 
− 2 
 wL2 
− 
 12 
 0 
{P} =  − wL 
 22 
 wL 
 12 
 0 
 
 0 
 0 
 
 45

Step 6: Determination of unknown displacements:.

The unknown displacements can be obtained from the relationship of { F } = [ K ]{d } or

{d } = [ K ] { F } .
−1
Now, eliminating the rows in the force matrix, corresponding to zero
element in displacement matrix, the reduced matrix will be as follows.
0
𝑑4 −1 ⎡ 4𝑤 ⎤
2.65 0.72 0
⎢− ⎥
�𝑑5 � = k −1 �0.72 0.8505 −1.125� ⎢ 2 ⎥
𝑑6 0 −1.125 3.2 ⎢ 16𝑤 ⎥
⎣ 12 ⎦
0
1 0.662 −1.047 −0.368 −2𝑤
= �−1.047 3.856 1.3556 � � 4 �
𝑘
−0.368 1.355 0.789 𝑤
3

Thus, the unknown displacements will be:

𝑑4
1 1.603𝑤
�𝑑5 � = � −5.905𝑤 �
𝐾
𝑑6 — 1.658𝑤

Step 7: Determination of member end actions:

The member end actions can be obtained from the corresponding member stiffness and the
nodal displacements. The member end actions for each member are derived as shown below.

Member - AB
In case of member (AB), the member forces will be: {Fm}(AB) = [K](AB) {d}(AB)
1 −1
⎡ 0 0 0 0 ⎤
⎢ 4 4 ⎥
⎢ 0 36 18 36 18 ⎥
0 − 3
𝑇1 ⎢ 43 42 4 42 ⎥ 0
⎡𝐹 ⎤ 18 12 18 6 ⎥ ⎡ ⎤
1 ⎢ 0
⎢ ⎥ 0 0 − ⎢ ⎥
⎢
⎢𝑀1 ⎥ = 𝐾 ⎢ 42 4 42 4 ⎥× 1⎢ 0
⎥ 𝐾 ⎢ 1.603𝑤 ⎥⎥
⎢ 𝑇2 ⎥ 1 1
⎢ 𝐹2 ⎥ ⎢− 0 0 0 0 ⎥ ⎢−5.905𝑤 ⎥
⎢ 4 4 ⎥ ⎣−1.658𝑤 ⎦
⎣𝑀2 ⎦ 36 18 36 18
⎢ 0 − − 0 − ⎥
⎢ 43 42 43 42 ⎥
⎢ 18 6 18 12 ⎥
⎣ 0 4 2 4
0 − 2
4 4 ⎦

Thus,
𝑇1 −0.4𝑤
⎡𝐹 ⎤ ⎡
1 1.456𝑤 ⎤
⎢ ⎥ ⎢ ⎥
⎢𝑀1 ⎥ = ⎢ 4.156𝑤 ⎥
⎢ 𝑇2 ⎥ ⎢ 0.4𝑤 ⎥
⎢ 𝐹2 ⎥ ⎢−1.456𝑤 ⎥
⎣𝑀2 ⎦ ⎣ 1.70𝑤 ⎦
 46

It is to be noted that {Fm} are the end actions due to joint loads. Hence it must be added to the
corresponding end actions in the restrained structure in order to obtain the end actions due to
the loads. Therefore, {Fm} Actual are the true member end actions due to actual loading system
and can be expressed as
{Fm} Actual = {Fm} + {Ffm}

Where, {Ffm}are the end actions in the restrained structure. Since there is no load acting on
member (1), the actual end action will be:
0
⎡ 4𝑤 ⎤
⎢ ⎥
−0.4𝑤 2
⎡ 1.456𝑤 ⎤ ⎢ ⎥ ⎡−0.4𝑤 ⎤
42 𝑤
⎢ ⎥ ⎢ ⎥ ⎢ 3.46𝑤 ⎥
4.156𝑤 ⎥ ⎢ 12 ⎥ ⎢ 5.49𝑤 ⎥
{𝐹𝑚 }𝐴𝑐𝑡𝑢𝑎𝑙 = ⎢ + =
⎢ 0.4𝑤 ⎥ ⎢ 0 ⎥ ⎢ 0.40𝑤 ⎥
⎢−1.456𝑤 ⎥ ⎢ 4𝑤 ⎥ ⎢ 0.54𝑤 ⎥
⎣ 1.70𝑤 ⎦ ⎢ 2 ⎥ ⎣ 0.34𝑤 ⎦
⎢ 42 𝑤 ⎥
⎣− 12 ⎦

Member - BC

In similar way, the member forces in member (BC) will be: {Fm}(BC) = [K](BC) {d}(BC)
1 −1
⎡ 0 0 0 0 ⎤
⎢ 5 4 ⎥
⎢ 0 36 18 36 18 ⎥
0 − 3
𝑇2 ⎢ 53 52 5 52 ⎥ −1.658𝑤
⎡𝐹 ⎤ 18 12 18 6 ⎥ ⎡−5.905𝑤 ⎤
2 ⎢
⎢ ⎥ ⎢ 0 0 − 2 ⎢ ⎥
⎢𝑀2 ⎥ = 𝐾 52 5 5 5 ⎥ × 1 ⎢ 1.603 ⎥
⎢ 𝑇3 ⎥ ⎢ 1 1 ⎥ 𝐾⎢ 0 ⎥
𝐹 ⎢ − 0 0 0 0 ⎥ ⎢ 0 ⎥
⎢ 3⎥ 5 5
⎢ ⎥ ⎣ ⎦
⎣𝑀3 ⎦ 0
⎢ 0 − 36 − 18 0 36
−
18⎥
⎢ 53 52 53 52 ⎥
⎢ 18 6 18 12 ⎥
⎣ 0 5 2 5
0 − 2
5 5 ⎦

Thus,
 47

𝑇2 −0.33𝑤
⎡𝐹 ⎤ ⎡ ⎤
⎢ ⎥ ⎢−0.55𝑤 ⎥
2

⎢𝑀2 ⎥ = ⎢−0.40𝑤 ⎥
⎢ 𝑇3 ⎥ ⎢ 0.33𝑤 ⎥
⎢ 𝐹3 ⎥ ⎢ 0.55𝑤 ⎥
⎣𝑀3 ⎦ ⎣−2.33𝑤 ⎦

Since there is no load acting on member (BC), the actual end action will be:
−0.33𝑤 0 −0.33𝑤
⎡−0.55𝑤 ⎤ ⎡0⎤ ⎡−0.55𝑤 ⎤
⎢−0.40𝑤 ⎥ ⎢ ⎥ ⎢−0.40𝑤 ⎥
{𝐹𝑚 }𝐴𝑐𝑡𝑢𝑎𝑙 = ⎢ ⎥ + ⎢0⎥ = ⎢ ⎥
⎢ 0.33𝑤 ⎥ ⎢0⎥ ⎢ 0.33𝑤 ⎥
⎢ 0.55𝑤 ⎥ ⎢0⎥ ⎢ 0.55𝑤 ⎥
⎣−2.33𝑤 ⎦ ⎣0⎦ ⎣−2.33𝑤 ⎦
Thus, the reaction forces at the support and load at the joints will be:

𝑇1 −0.4𝑤
⎡𝐹 ⎤ ⎡ 3.46𝑤 ⎤
⎢𝑀1 ⎥ ⎢ 5.49𝑤 ⎥
⎢ 1⎥ ⎢ ⎥
⎢ 𝑇2 ⎥ ⎢ 0 ⎥
⎢ 𝐹2 ⎥ = ⎢ 0 ⎥
⎢𝑀2 ⎥ ⎢ 0 ⎥
⎢ 𝑇3 ⎥ ⎢ 0.33𝑤 ⎥
⎢ 𝐹3 ⎥ ⎢ 0.55𝑤 ⎥
⎣𝑀3 ⎦ ⎣−2.33𝑤 ⎦

4.6.5 Analysis of Space Frame

Space frames are an increasingly common architectural technique especially for large roof
spans in commercial and industrial buildings. The rigid jointed frames such as building
frames are usually three dimensional space structures. Thus in case of certain structures like
multi-storeyed buildings, it is necessary consider 3-dimensional effects for analysis. The
space frame constitutes the final step of increasing complexity. It consists of plane frame and
grid actions. The displacement and rotation vector associated with each joint have three
components in case of space frame structures. There are six equilibrium equations associated
with each joint. The degrees of freedom at each node of the space frame member will be (i)
displacement in three perpendicular directions and (ii) rotations in three different directions.
Therefore, the degrees of freedom in each node of the member will be six as shown in the
figure below. The stiffness matrix in local coordinate system considering all possible degrees
of freedom will be as given in Table 4.6.1.
 48

Fig. 4.6.5 Degrees of freedom for space frame member

Table 4.6.1 Stiffness matrix of space frame member

1 2 3 4 5 6 7 8 9 10 11 12

 EAX EAX 
1  L 0 0 0 0 0 − 0 0 0 0 0 
L
 
 0 12 EI Z
0 0 0
6 EI Z
0 −
12 EI Z
0 0 0
6 EI Z 
2  L3 L2 L3 L2 
 12 EIY 6 EIY 12 EIY 6 EIY 
 0 0 0 − 0 0 0 − 0 − 0 
3  L3 L2 L3 L2 
 GI X GI X 
 0 0 0 0 0 0 0 0 − 0 0 
4  L L 
 6 EIY 4 EIY 6 EIY 2 EIY 
5  0 0 − 0 0 0 0 0 0 
 L2 L L2 L 
 6 EI Z 4 EI Z 6 EI Z 2 EI Z 
6  0 0 0 0 0 − 0 0 0
L2 L L2 L 
 
 − EAX 0 0 0 0 0
EAX
0 0 0 0 0 
7  L L 
 12 EI Z 6 EI Z 12 EI Z 6 EI Z 
8  0 − 0 0 0 − 0 0 0 0 − 2
 L3 L2 L3 L 
 12 EIY 6 EIY 12 EIY 6 EIY 
9  0 0 − 0 0 0 0 0 0 
 L3 L2 L3 L2 
 GI X GI X 
10  0 0 0 − 0 0 0 0 0 0 0 
 L L 
 6 EIY 2 EIY 6 EIY 4 EIY 
11  0 0 − 0 0 0 0 0 0 
L2 L L2 L
 
 0 6 EI Z 2 EI Z 6 EI Z 4 EI Z 
12 0 0 0 0 − 0 0 0
 L2 L L L 

The generalized stiffness matrix of a rigid jointed space frame member can be obtained by
transferring the matrix of local coordinate system into its global coordinate system. The
transformation matrix will become a square matrix of size 12×12 in this case as the degrees
of freedom for each node/joint is six.
Lecture 1: Constant Strain Triangle

The triangular elements with different numbers of nodes are used for solving two dimensional
solid members. The linear triangular element was the first type of element developed for the
finite element analysis of 2D solids. However, it is observed that the linear triangular element is
less accurate compared to linear quadrilateral elements. But the triangular element is still a very
useful element for its adaptivity to complex geometry. These are used if the geometry of the 2D
model is complex in nature. Constant strain triangle (CST) is the simplest element to develop
mathematically. In CST, strain inside the element has no variation (Ref. module 3, lecture 2) and
hence element size should be small enough to obtain accurate results. As indicated earlier, the
displacement is expressed in two orthogonal directions in case of 2D solid elements. Thus the
displacement field can be written as
u
d     (5.1.1)
v
Here, u and v are the displacements parallel to x and y directions respectively.

5.1.1 Element Stiffness Matrix for CST

A typical triangular element assumed to represent a subdomain of a plane body under plane
stress/strain condition is represented in Fig. 5.1.1. The displacement (u, v) of any point P is
represented in terms of nodal displacements
u = N1u1 + N 2u2 + N 3u3
(5.1.2)
v = N1v1 + N 2 v2 + N 3v3
Where, N1, N2, N3 are the shape functions as described in module 3, lecture 2.

Fig. 5.1.1 Linear triangular element for plane stress/strain

The strain-displacement relationship for two dimensional plane stress/strain problem can be
simplified in the following form from three dimensional cases (eq.1.3.9 to1.3.14).

u 1  u   v  
2 2

x      
x 2  x   x  

v 1  u   v  
2 2

y      
y 2  y   y  
(5.1.3)
 
v u  u u v v 
 xy     
x y  x y x y 

In case of small amplitude of displacement, one can ignore the nonlinear term of the above
equation and will reach the following expression.
∂u
εx =
∂x
∂v
εy = (5.1.4)
∂y
∂v ∂u
γ= +
∂x ∂y
xy

Hence the element strain components can be represented as,

 ∂u ∂N ∂N ∂N
ε x = =1 u1 + 2 u2 + 3 u3
 ∂x ∂x ∂x ∂x
 ∂v ∂N1 ∂N 2 ∂N
ε=
ε y == v1 + v2 + 3 v3
 ∂y ∂y ∂y ∂y
 ∂u ∂v ∂N1 ∂N ∂N ∂N ∂N ∂N
γ xy = + = u1 + 2 u2 + 3 u3 + 1 v1 + 2 v2 + 3 v3
 ∂y ∂x ∂y ∂y ∂y ∂x ∂x ∂x
Or,

 ∂N ∂N 2 ∂N 3   u1 
 1 0 0 0  u 
 ε x   ∂x ∂x ∂x  2
   ∂N1 ∂N 2 ∂N 3  u3 
=ε =εy   0 0 0   (5.1.5)
γ   ∂y ∂y ∂y   v1 
 xy   ∂N ∂N 2 ∂N 3 ∂N1 ∂N 2 ∂N 3   v2 
 
1

 ∂y ∂y ∂y ∂x ∂x ∂x   v3 

Or, ε = [ B ] {d } (5.1.6)
In the above equation [B] is called as strain displacement relationship matrix. The shape
functions for the 3 node triangular element in Cartesian coordinate is represented as,

 

1 
           


  x 2 y 3 x 3 y 2 y 2 y 3 x x 3 x 2 y 


 N  2A 
 1    1  

 
N 2     x 3 y1  x1y3    y3  y1  x   x1  x 3  y  
     2A  

   

 3
 N   1 
   

  x1y 2  x 2 y1    y1  y 2  x   x 2  x1  y  
 2A
 

Or,
 1
 


    x   y  

 
1 1 1

 N1   2A 
     

  1
 N 2     2  2 x   2 y   (5.1.7)

 
 
 2A 


 3
 N  
  1 


  3  3 x  3 y

 2A
 

Where,
 1   x 2 y 3  x 3 y 2 ,  2   x 3 y1  x1y3 ,  3   x1y 2  x 2 y1 ,
1   y 2  y3 , 2   y3  y1 , 3   y1  y 2 , (5.1.8)
1   x 3  x 2 ,  2   x 2  x1 ,  3   x 2  x1 ,
Hence the required partial derivatives of shape functions are,
∂N1 β1 ∂N 2 β 2 ∂N 3 β3
= , = , = ,
∂x 2 A ∂x 2 A ∂x 2 A
∂N1 γ 1 ∂N 2 γ 2 ∂N 3 γ 3
= , = , = ,
∂y 2 A ∂x 2A ∂x 2 A
Hence the value of [B] becomes:
 ∂N ∂N 2 ∂N 3 
 1 0 0 0 
 ∂x ∂x ∂x 
 ∂N1 ∂N 2 ∂N 3 
[ B] =  0 0 0 
 ∂y ∂y ∂y 
 ∂N1 ∂N 2 ∂N 3 ∂N1 ∂N 2 ∂N 3 
 
 ∂y ∂y ∂y ∂x ∂x ∂x 
 β1 β2 β3 0 0 0
1 
Or, [ B] =  0 0 0 γ1 γ 2 γ 3  (5.1.9)
2A
 γ 1 γ2 γ3 β1 β 2 β3 
 According to Variational principle described in module 2, lecture 1, the stiffness matrix is
represented as,
=[k ] ∫∫∫ [ B ] [ D ][ B ] d Ω
T
(5.1.10)
Ω

Since, [B] and [D] are constant matrices; the above expression can be expressed as

=[k ] [=
B ] [ D ][ B ] ∫∫∫ d V [ B ] [ D ][ B ]V
T T
(5.1.11)
V

For a constant thickness (t), the volume of the element will become A.t . Hence the above
equation becomes,
[ k ] = [ B ] [ D ][ B ] At
T
(5.1.12)
For plane stress condition, [D] matrix will become:
 
1 µ 0 
E  
[ D] = 2 
µ 1 0  (5.1.13)
1− µ 
1− µ 
0 0 
 2 
Therefore, for a plane stress problem, the element stiffness matrix becomes,
 β1 0 γ1 
β 0 γ 2   
 2 1 µ 0   β1 β 2 β3 0 0 0
 β3 γ3   
0   0 0 γ 3 
Et 0
[k ] =   µ 1 0 γ1 γ 2 (5.1.14)
4 A (1 − µ 2 )  0 γ1 β1  
1 − µ   γ 1 γ 2 γ3 β1 β 2 β3 
0 γ2 β2   0 0 
  2 
 0 γ3 β3 
Or,
 2 (1 + µ ) β γ 
 β1 + Cγ 1 β1β 2 + Cγ 1γ 2 β1β3 + Cγ 1γ 3 µβ1γ 2 + C β 2γ 1 µβ1γ 3 + C β3γ 1 
2
1 1
 2 

β + Cγ
2 2
β 2 β3 + Cγ 2γ 3 µβ 2γ 1 + C β1γ 2
(1 + µ ) β γ 
µβ 2γ 3 + C β 3γ 2  (5.1.15)
 2 2 2 2
 2 
Et
[k ] =  (1 + µ ) β γ 
4 A (1 − µ ) 
2
β32 + Cγ 32 µβ3γ 1 + C β1γ 3 µβ 3γ 2 + C β 2γ 3 3 3 
 2 
 γ + Cβ
2
1 1
2
γ 1γ 2 + C β1β 2 γ 1γ 3 + C β1β3 
 Sym. γ 22 + C β 22 γ 2γ 3 + C β 2 β 3 
 
 γ 32 + C β32 

Where, C =
(1 − µ )
2
Similarly for plane strain condition, [D] matrix is equal to,
  
(1 − µ ) µ 0 
E  
= [ D]  µ (1 − µ ) 0  (5.1.16)
(1 + µ )(1 − 2µ )  1 − 2µ 
 0 0 
 2 
Hence the element stiffness matrix will become:
 M β12 + γ 12 M β1β 2 + γ 1γ 2 M β1β3 + γ 1γ 3 ( µ + 1) β1γ 1 µβ1γ 2 + β 2γ 1 µβ1γ 3 + β3γ 1 
 
 M β 22 + γ 22 M β 2 β 3 + γ 2γ 3 µβ 2γ 1 + β1γ 2 ( µ + 1) β 2γ 2 µβ 2γ 3 + β3γ 2 
 M β32 + γ 32 µβ3γ 1 + β1γ 3 µβ3γ 2 + C β 2γ 3 ( µ + 1) β3γ 3 
(5.1.17)
Et
[k ] =  
2 A (1 + µ )  M γ 12 + β12 M γ 1γ 2 + β1β 2 M γ 1γ 3 + β1β3 
 Sym. M γ 22 + β 22 M γ 2γ 3 + β 2 β 3 
 
 M γ 32 + β32 

Where M= (1 − µ )

5.1.2 Nodal Load Vector for CST

From the principle of virtual work,

   d    u  F d    u  F d

T T T
(5.1.18)
  

Where, FΓ, and FΩ are the surface and body forces respectively. Using the relationship between
stress-stain and strain displacement, one can derive the following expressions:
D ][ B ]{d } , δ {ε } [ B ]δ=
{σ } [ = {d } and δ {u} [ N ]δ {d }
(5.1.19)
Hence eq. (5.1.18) can be rewritten as,
 NS  F d   dT  N T F d (5.1.20)
  d B DBdd    d 
T T T T

 
  

Or,
  B  D  Bdd    N  F d    N  F d
S T (5.1.21)
T T

  
s
Here, [N ] is the shape function along the boundary where forces are prescribed. Eq.(5.1.21) is
equivalent to  k d  F , and thus, the nodal load vector becomes

F    N  F d    N  F d

T T
S
 

  (5.1.22)
For a constant thickness of the triangular element eq.(5.1.22) can be rewritten as

F  t   N  F ds  t   N  F dA

T T
S
  (5.1.23)
S A

For the a three node triangular two dimensional element, one can represent F and F as,
 Fx  Fx 
F     and F    
Fy  Fy 
Fx   0 
For example, in case of gravity load on CST element, F   
    
Fy  g
For this case, the shape functions in terms of area coordinates are:
 L1 L 2 L3 0 0 0
 N     (5.1.24)
0 0 0 L1 L 2 L3 
As a result, the force vector on the element considering only gravity load, will become,
L 0  0   0 
1
    0 
L 0  0   
2
  
L 0  0   0 
  dA  t  
0 
F  t  3
 dA  gt   dA (5.1.25)
0 L   L g   L 
A
 1
 g A
 
1 A
 
1

0 L  L g L 

     
2 2 2

0 L 
 3 L g 3 L 3

The integration in terms of area coordinate is given by,

p! q! r !
∫A L1 L2 L3dA = ( p + q + r + 2 )! 2 A
p q r
(5.1.26)

Thus, the nodal load vector will finally become


 0  
  0  

 
  
  
 



0 
  0  0
 
  
 
 
 



0 
  0   
    0

 1!0!0!  
   gAt

 
 

(5.1.27)
  
 
  
0
 
2A gAt
F  gt  (1  0  0  2)!       

 
 

3 
 3 1


 0!1!0! 
 
 gAt  
 

 2A    1

 (0  1  0  2)!   
 3  
 


 
 
 
 

1


 0!0!1!    gAt 

 2A 
 
 

 (0  0  1  2)! 

  
  3 
Lecture 2: Linear Strain Triangle

5.2.1 Element Stiffness Matrix for LST

In case of CST, it is observed that the strain within the element remains constant. Though, these
elements are able to provide enough information about displacement pattern of the element, but
it is unable to provide adequate information about stress inside an element. This limitation will
be significant enough in regions of high strain gradients. The use of a higher order triangular
element called Linear Strain Triangle (LST) significantly improves the results at these areas as
the strin inside the element is varying. The LST element has six nodes (Fig. 5.2.1) and hence,
twelve degrees of freedom. Thus the displacement function can be chosen as follows.
u =α 0 + α1 x + α 2 y + α 3 x 2 + α 4 xy + α 5 y 2
(5.2.1)
v =a6 + a7 x + a8 y + a9 x 2 + a10 xy + a11 y 2

Fig. 5.2.1 Linear strain triangle element

Therefore, the element strain matrix is obtained as

u
x   1  2 3 x   4 y
x
v (5.2.2)
y    8  10 x  211y
y
v u
 xy    ( 2   7 )  ( 4  2 9 )x  (2 5  10 )y
x y
In the area coordinate system as discussed in module 3, lecture 3 we can write the shape function
for the six node triangular element as
N1  L1 2L1 1 N 2  L 2 2L 2 1 N 3  L3 2L3 1
(5.2.3)
N 4  4L1L 2 N 5  4L 2 L3 N 6  4L3L1
The displacement (u,v) of any point within the element can be represented in terms of their nodal
displacements with the use of interpolation function.
6
u = ∑ N i ui
i =1
6
(5.2.4)
v = ∑ N i vi
i =1

Using eq.(5.2.4) we can rewrite eq.(5.2.2) as,

 u1 
u 
 2
u3 
 
 ∂N ∂N 2 ∂N 3 ∂N 4 ∂N 5 ∂N 6  u4 
 1 0 0 0 0 0 0  u 
 ∂x ∂x ∂x ∂x ∂x ∂x  5
 ∂N1 ∂N 2 ∂N 3 ∂N 4 ∂N 5 ∂N 6  u6 
ε = 0 0 0 0 0 0  
 ∂y ∂y ∂y ∂y ∂y ∂y   v1 
 ∂N1 ∂N 2 ∂N 3 ∂N 4 ∂N 5 ∂N 6 ∂N1 ∂N 2 ∂N 3 ∂N 4 ∂N 5 ∂N 6   v2 
  
 ∂y ∂y ∂y ∂y ∂y ∂y ∂x ∂x ∂x ∂x ∂x ∂x   v3 
v 
 4
 v5 
v 
 6
Or,
ε = [ B ] {d } (5.2.5)
Where,
 ∂N ∂N 2 ∂N 3 ∂N 4 ∂N 5 ∂N 6 
 1 0 0 0 0 0 0 
 ∂x ∂x ∂x ∂x ∂x ∂x 
 ∂N1 ∂N 2 ∂N 3 ∂N 4 ∂N 5 ∂N 6  (5.2.6)
[ B] =  0 0 0 0 0 0 
 ∂y ∂y ∂y ∂y ∂y ∂y 
 ∂N1 ∂N 2 ∂N 3 ∂N 4 ∂N 5 ∂N 6 ∂N1 ∂N 2 ∂N 3 ∂N 4 ∂N 5 ∂N 6 
 
 ∂y ∂y ∂y ∂y ∂y ∂y ∂x ∂x ∂x ∂x ∂x ∂x 

Using Chain rule,

∂N1 ∂N1 ∂L1 ∂N1 ∂L2 ∂N1 ∂L3
= . + . + .
∂x ∂L1 ∂x ∂L2 ∂x ∂L3 ∂x
As discussed in module 3, lecture 1, we can write the above expression as,
 ∂N1 b1 ∂N1 b2 ∂N1 b3 ∂N1
= . + . + .
∂x 2 A ∂L1 2 A ∂L2 2 A ∂L3
∂N1 b1
= . ( 4 L1 − 1)
∂x 2 A
Similarly we can evaluate expressions for other terms and can be written as,
∂N1 b1 ∂N 2 b2 ∂N 3 b3
= . ( 4 L1 − 1) = . ( 4 L2 − 1) = . ( 4 L3 − 1)
∂x 2 A ∂x 2 A ∂x 2 A
∂N 4 ∂N 5 ∂N 6
=4 ( L2b1 + L1b2 ) =4 ( L3b2 + L2b3 ) = 4 ( L1b3 + L3b1 )
∂x ∂x ∂x
And,
∂N1 a1 ∂N 2 a2 ∂N 3 a3
= . ( 4 L1 − 1) = . ( 4 L2 − 1) = . ( 4 L3 − 1)
∂y 2 A ∂y 2 A ∂y 2 A
∂N 4 ∂N 5 ∂N 6
=4 ( L2 a1 + L1a2 ) =4 ( L3 a2 + L2 a3 ) =4 ( L1a3 + L3 a1 )
∂y ∂y ∂y
Where,
a1 =
x2 − x3 a2 =
x3 − x1 a3 =
x1 − x2
b1 =
y2 − y3 b2 =
y3 − y1 b3 =
y1 − y2

The stiffness matrix of the element is represented by,

=[k ] ∫∫∫ [ B ] [ D ][ B ] d Ω
T
(5.2.7)
Ω

The, [D] matrix is the constitutive matrix which will be taken according to plane stress or plane
strain condition. The nodal strain and stress vectors are given by,
{ε n } = {ε x1 ε x 2 ε x3 ε y1 ε y 2 ε y 3 γ xy1 γ xy 2 γ xy 3}
T

(5.2.9a)
{σ n } = {σ x1 σ x 2 σ x3 σ y1 σ y 2 σ y 3 τ xy1 τ xy 2 τ xy 3}
T

(5.2.9b)

 [ Bn1 ] [ 0] 
{ε n } =  [ 0] [ Bn 2 ] {d } (5.2.10)
[ Bn 2 ] [ Bn1 ] 
Referring to section 3.3.1, using proper values of area coordinates in [B] matrix, one can find
 3b1 −b2 −b3 4b2 0 4b3 
1 
[ Bn1 ] = −b1 3b2 −b3 4b1 4b3 0  (5.2.11a)
2A
 −b1 −b 3b3 0 4b2 4b1 
And,
 3a1 −a2 −a3 4a2 0 4a3 
[ Bn 2 ] = −a1 3a2 −a3 4a1 4a3 0 
1 
(5.2.11b)
2A
 −a1 −a 3a3 0 4a2 4a1 
Thus, the element stiffness can be evaluated by putting the values from eq. (5.2.11) in eq. (5.2.7).

5.2.2 Nodal Load Vector for LST

Similar to 3-node triangular element, the load will be lumped at each node which can be
computed using the earlier expression,

F    N  F d    N  F d

T T
S
  (5.2.12)
 

And for element with constant thickness,

F  t   N  F ds  t   N  F dA
T T
S
 
(5.2.13)
S A

5.2.3 Numerical Example using CST

Determine the displacements at the nodes for the following 2D solid continuum considering a
constant thickness of 25 mm, Poisson’s ratio, µ as 0.25 and modulus of elasticity E as 2 x 105
N/mm2. The continuum is discritized with two CST plane stress elements.

Fig. 5.2.2 Geometry and discretization of the continuum

The element 1 is connected with node 1, 3 and 4 and let assume its Cartesian coordinates are (x1,
y1), (x3, y3) and (x4, y4) respectively. If we consider nodes 1, 3 and 4 are similar to node 1, 2 and
3 in eq.(5.1.9) then the [B] can be written as
  β1 β 2 β3 0 0 0 
[ B ] =  0 0 0 γ 1 γ 2 γ 3 
1 
2A
 γ 1 γ 2 γ 3 β1 β 2 β3 
By introducing values of β & γ discussed in previous lecture note, we can get value of [B] as
 0 1 −1 0 0 0 
1 
=[ B] 0 0 0 −3 0 3 
1500 
 −3 0 3 0 1 −1
For plain stress problem, putting the values of E and µ one can find the following values.
 
1 µ 0  16 4 0 
E   4 ×104  
= [ D] = µ 1
1 − µ 2 
0   4 16 0 
3
1− µ   0 0 6 
0 0 
 2 
Therefore the stiffness matrix for the element 1 will be
[ k ]1 = tA [ B ] [ D ][ B ]
T

Putting values of t, A, [B] & [D]we will get,

750 0 -750 0 -250 -250 

0 222.2222 -222.2222 -166.6667 0 166.6667 

-750 -222.2222 972.2222 166.6667 250 -416.6667 
[ k ]1 =4 ×104 × 0 -166.6667 166.6667 2000 0 -2000

 
-250 0 250 0 83.3333 -83.3333 
 
 250 166.6667 -416.6667 -2000 -83.3333 2083.3333
Similarly element 2 is connected with nodes 1, 2 and 3 and global coordinates of these nodes are
(x1, y1), (x2, y2) and (x3, y3) respectively. For this element, by proceeding in a similar manner to
element 1 we can calculate [B] matrix as,
 −1 1 0 0 0 0 
1 
= [ B]  0 0 0 0 −3 3
1500
 0 −3 3 −1 1 0 
Hence, the elemental stiffness matrix becomes,
  222.2222 -222.2222 0 0 166.6667 -166.6667 
-222.2222 972.2222 -750 250 -416.6667 166.6667 

0 -750 750 -250 250 0 
[ k ]2 =4 ×10 × 
4

0 250 -250 83.3333 -83.3333 0 
166.6667 -416.6667 250 -83.3333 2083.3333 -2000 
 
-166.6667 166.6667 0 0 -2000 2000 

By assembling the stiffness matrices into global stiffness matrix [K],

972.2222 -222.2222 0 -750 0 166.6667 -416.6667 -250 
-222.2222 972.2222 -750 0 250 -416.6667 166.6667 0 
 
0 -750 972.2222 -222.2222 -416.6667 250 0 166.6667 
 
4  -750 0 -222.2222 972.2222 166.6667 0 250 -416.6667 
[ K ] =×
4 10 ×
0 250 -416.6667 166.6667 2083.3333 -83.3333 0 -2000 
 
166.6667 -416.6667 250 0 -83.3333 2083.3333 -2000 0 
-416.6667 166.6667 0 250 0 -2000 2083.3333 -83.3333 
 
 250 0 166.6667 -416.6667 -2000 0 -83.3333 2083.3333

Now, applying equation [ F ] = [ K ]{d } , the following expression can be written.

 Fu1  972.2222 -222.2222 0 -750 0 166.6667 -416.6667 -250   u1 
F  -222.2222 972.2222 -750 0 250 -416.6667 166.6667 0  u 
 u2    2
 Fu 3  0 -750 972.2222 -222.2222 -416.6667 250 0 166.6667  u3 
    
 u4 
F 4  -750 0 -222.2222 972.2222 166.6667 0 250 -416.6667  u4 
  =× ×
  v1 
4 10
0
 Fv1  
250 -416.6667 166.6667 2083.3333 -83.3333 0 -2000
 
 Fv 2  166.6667 -416.6667 250 0 -83.3333 2083.3333 -2000 0   v2 
  -416.6667  
 Fv 3  166.6667 0 250 0 -2000 2083.3333 -83.3333   v3 
 
F   250 2083.3333  v4 
 v4  0 166.6667 -416.6667 -2000 0 -83.3333

Putting boundary conditions u=

1 v=
1 u=
2 u=
4 v=
4 0 and adopting elimination technique for
applying boundary condition we get expression,
 0  972.2222 250 0  v2 
    u 
25000  =4 × 10 ×  250
4
2083.3333 2000  3
 0     
   0 2000 2083.3333   v3 
Solving the above expression, the unknown nodal displacements may be obtained as follows.
v2 25.96 ×10−5 mm, u3 =
= −10.02 ×10−5 mm and
= v3 96.92 ×10−5 mm.
Lecture 3: Rectangular Elements

The rectangular elements are widely used for solving two dimensional continuums. The main
advantage of this type of element is the easy formulation and easy development of computer
code. The element stiffness of such elements is derived here using the concept of isoparametric
formulation.

5.3.1 Computation of Element Stiffness

In case of a four node rectangular element, the geometry and displacement filed can be expressed
in terms of their nodal values with the help of interpolation function. As the formulation will be
isoparametric, the interpolation function will become same for expressing both the variables.
Thus, coordinates and displacements at any point inside the element (Fig. 5.3.1) can be expressed
as
x  N1 x1  N 2 x2  N 3 x3  N 4 x4
(5.3.1)
y  N1 y1  N 2 y2  N 3 y3  N 4 y4
And
u  N1u1  N 2u2  N 3u3  N 4u4
(5.3.2)
v  N1v1  N 2 v2  N 3v3  N 4 v4

The above equations can be written in matrix form as


 x1 


 


 y1


x  


 2 
 


   N1
x 0 N2 0 N3 0 N4 0  y2 

    
   (5.3.3)

 
 x 0
 N1 0 N2 0 N3 0  
N 4   x3 

 

 y3 
 

 



x4 


 y4 


 

And
 
 u1 


 


 v1


 


 u 2
  

u
   N1 0 N2 0 N3 0 N4 0  v2 
      (5.3.4)

v 
  0
 N1 0 N2 0 N3 0 N 4  u3 

 
 v3 


 


 u 
4

 
v4 

 


The shape function four node rectangular element is derived and shown in module 3, lecture 4.
However the shape functions are reproduced here for easy reference for the derivation of the
stiffness matrix.

 1 x1 h 

 

 4 

 N1  
 


 
  1  x 1  h  
   
N2   

 
Ni      4
 (5.3.5)

 N3 
 
 1  x1  h

 
 
 


N4  
 4 



 1 x1  h

 


 4 


Fig. 5.3.1 Four node rectangular element

The strain-displacement relationship for two dimensional plane stress/strain problem can be
simplified in the following form from three dimensional cases (eq.1.3.9 to1.3.14).

u 1  u   v  
2 2

x      
x 2  x   x  

v 1  u   v  
2 2

y      
y 2  y   y  
(5.3.6)
 
v u  u u v v 
 xy     
x y  x y x y 
In case of small amplitude of displacement, one can ignore the nonlinear term of the above
equation and will reach the following expression.
∂u
εx =
∂x
∂v
εy = (5.3.7)
∂y
∂v ∂u
γ= +
∂x ∂y
xy

Using the shape function the above expression can be written as

 u1 
 
 ∂N1 ∂N 2 ∂N 3 ∂N 4   v1 
 0 0 0 0  u 
εx   ∂x ∂x ∂x ∂x  2
   ∂N1 ∂N 2 ∂N 3 ∂N 4   v2 
εy  = 0 0 0 0   [ B ]{d } (5.3.8)
   ∂y ∂y ∂y ∂y  u3 
γ xy   ∂N1 ∂N1 ∂N 2 ∂N 2 ∂N 3 ∂N 3 ∂N 4 ∂N 4   v3 
  
 ∂y ∂x ∂y ∂x ∂y ∂x ∂y ∂x  u4 
v 
 4
Here, [B] is known as strain displacement relationship matrix. The derivatives of the shape
functions are calculated using the chain rule.
∂N i ∂N i ∂x ∂N i ∂y
= . + .
∂ξ ∂x ∂ξ ∂y ∂ξ
(5.3.9)
∂N i ∂N i ∂x ∂N i ∂y
= . + .
∂η ∂x ∂η ∂y ∂η
Here, i is referred to number of nodes in an element and will be 4 in this case. Converting above
expression in matrix form
  N i   x y   N i   N i 
       
 x   x x   x   x 
       J   (5.3.10)
 N i   x y   N i   N i 
      
 h   h h   y   y 
The matrix [J] is referred to Jacobian matrix which is discussed in Lecture 7, module 3. Using
eq. (5.3.1) one can write
x N1 N N N
 x1  2 x2  3 x3  4 x4 (5.3.11)
x x x x x
Putting the values of the nodal coordinates and shape functions of the four node element in the
above equation the following relations will be obtained.
x 11 h 11 h 11  h 11  h a
 0  a  a  0 
x 4 4 4 4 2
(5.3.12a)
Similarly,
y 11 h 11 h 11  h 11  h
 0  0  b  b  0
x 4 4 4 4
(5.3.12b)
x 11 x  11  x  11  x  11 x 
 0  a  a  0  0 (5.3.12c)
h 4 4 4 4
y 11 x  11  x  11  x  11 x  b
 0  0  b  b 
h 4 4 4 4 2
(5.3.12d)
Substituting above values in Jacobian matrix the following relations will be obtained.
a  2 
 0  0
   
 J    2  and  J    a 
1
(5.3.13)
 b  2
0  0 
 2   b 
Thus, eq.(5.3.10) can be written as
 N i   N   N   N 
 i   2 0   i   2 . i 
 x 
    J 1  x    a   x   a x 
    (5.3.14)
 N i   N i   2   N i   2 N i 
     0    . 
 y   h   b   h   b h 
After derivation of the shape functions expressed in eq.(5.3.5), the following values will be
obtained.
 N1 2 N1 1 h N 2 1 h N 3 1  h N 4 1  h
 .  ;  ;  ; 
x a x 2a x 2a x 2a x 2a (5.3.15)
N1 2 N1 1 x N 2 1  x N 3 1  x  N 4 1 x 
 .  ;  ;  ; 
y b h 2b y 2b y 2b y 2b
So, the strain displacement relationship matrix, [B] will become as follows.
 (1 − η ) (1 − η ) (1 + η ) (1 + η ) 
− 0 0 0 − 0 
 2a 2a 2a 2a 
 (1 − ξ ) (1 + ξ ) (1 + ξ ) (1 − ξ )  (5.3.16)
[ B] =
 0 − 0 − 0 0 
 2b 2b 2b 2b 
 (1 − ξ ) (1 − η ) (1 + ξ ) (1 − η ) (1 + ξ ) (1 + η ) (1 − ξ ) (1 + η ) 
− − − − 
 2b 2a 2b 2a 2b 2a 2b 2a 

The element stiffness matrix will become

[k ] t =
∫∫ [ B ] [ D ][ B ] dx dy t ∫∫ [ B ] [ D ][ B ] J dξ dη
T T
= (5.3.17)
It is seen that the above is expressed in terms of ξ and η and hence can be numerically integrated
by the Gauss Quadrature rule. The stiffness matrix for each element can be found which needs
to be globally assembled for getting the global stiffness matrix to obtain the solution. The
stiffness matrix of higher order rectangular element can be derived in a similar fashion. For
example, in case of eight node rectangle element, the size of [B] matrix will become 16 × 3
which was 8 × 3 for four node element. Thus the size of element stiffness for eight node element
will become 16 × 16.

5.3.2 Computation of Nodal Loads

If a distributed load acts on a side of a four node rectangular element, the nodal load vector can
be calculated the similar procedure as discussed in case of triangular element. If an element as
shown below is subjected to a linearly varying intensities of load at its one side, then the
magnitude of this at any point on the side can be expressed by its interpolation function as
follows.
1 − η 1 + η  qx 2 
qx =    (5.3.18)
 2 2   qx 3 
Here, qx2 and qx3 are the force intensities per unit length at nodes 2 and 3 respectively. The load
at nodes can be calculated from the following expression.
{Fx }
= ∫ {N } q d Γ
S
2 x 1 (5.3.19)
As ξ=1 along the side 2-3, the interpolation function will become
  (1 − ξ )(1 − η ) 
   0 
 4   
 (1 + ξ )(1 − η )   (1 − η ) 

 
  2 
=N 2S ={ } 4
   (5.3.20)
 (1 + ξ )(1 + η )   (1 + η ) 
 4   2 
   
 (1 − ξ )(1 + η )   0 

 4 


If the element thickness is t, then dΓ1 =t.dl. Thus the eq.(5.3.19) can be replaced as
 0 

 (1 − η ) 
+1
  1 −η 1 + η  qx 2 
{Fx } = t ∫  2    dl (5.3.21)
−1  (1 + η )  
2 2   qx 3 
 2 
 
 0 

Fig. 5.3.2 Varying load on a four node element

After integrating the above expression, the nodal load vector along x direction will become as
follows.
 0 
 2q + q 
{Fx } =  x 2 x3 
t
(5.3.22)
3  q x 2 + 2q x 3 

 0 

Lecture 4: Numerical Evaluation of Element Stiffness

Derivation of element stiffness for a four node rectangle element has been demonstrated in last
lecture. The stiffness matrix of each element can be calculated easily by developing a suitable
computer algorithm. To help students for developing their own computer code, a numerical
example has been solved and demonstrated here.

5.4.1 Numerical Example

Calculate the stiffness matrix for the given four node rectangular element by the Gauss
Quadrature integration rule using one point and two point formula assuming plane stress
formulation. Consider, the thickness of element = 20 cm, E=2 × 103 kN/cm2 and µ =0.

Fig. 5.4.1 Element Dimension

5.4.2 Evaluation of Stiffness using One Point Gauss Quadrature

For the calculation of stiffness matrix, first, 1×1 Gauss Quadrature integration procedure has
been carried out. Thus, the natural coordinate of the sampling point will become 0,0 and weight
will become 2.0 which is shown in the figure below.

Fig. 5.4.2 Natural coordinates for one point Gauss Quadrature

For a four node quadrilateral element, the shape functions and their derivatives are as follows.

(1 − ξ )(1 − η ) (1 + ξ )(1 − η ) (1 + ξ )(1 + η ) (1 − ξ )(1 + η )

N1 = , N2 = and N 3 = and N 4 =
4 4 4 4
∂N1 − (1 − ξ ) ∂N 2 (1 − ξ ) ∂N 3 (1 + ξ ) ∂N 4 − (1 + ξ )
= ; = ; = ; =
∂ξ 4 ∂ξ 4 ∂ξ 4 ∂ξ 4
∂N1 − (1 − η ) ∂N 2 − (1 + η ) ∂N 3 (1 + η ) ∂N 4 (1 − η )
= ; = ; = ; =
∂η 4 ∂η 4 ∂η 4 ∂η 4

The Jacobian matrix can be found from the following relations.

 ∂N1 ∂N 2 ∂N 3 ∂N 4   x1 y1  x y1 
 −(1 − ξ ) +(1 − ξ ) +(1 + ξ ) −(1 + ξ )   1 
 ∂ξ  
∂ξ ∂ξ ∂ξ   x 2 y2   4 4 4 4   x2 y2 

|J| =   
 ∂N1 ∂N 2 ∂N 3 ∂N 4   x3 y3   −(1 − η ) −(1 + η ) +(1 + η ) +(1 − η )   x y3 

 ∂η    3
 ∂η ∂η ∂η   x4 y4   4 4 4 4   x y4 
 4

Considering the sampling point, (ξ=0 and η=0 ), the value of the Jacobian, [J] is

−1 +1 1 − 1  0 0 
 

[J ] =  4 4 4 4  70 0 
−1 −1 1 1  70 50
 
4 4 4 4   0 50

1 
 0
35 0  35  J11* J12* 
Thus, [J ] =   =
and [ J ]−1
 =   * * 
; and |𝐽| = 875
 0 25 0 1   J 21 J 22 
 25 
Now, the strain vector for the element will become
 ∂u 
 ∂ξ 
 
 J 11* J 12* 0 0   ∂u 
 *   ∂η 
[ε ] =  0 0 J 21 *
J 22  ×  ∂v 
 J 21

* *
J 22 J 11* J 12*   
 ∂ξ 
 ∂v 
 ∂η 
 
  u1 
 ∂N 1 ∂N 2 ∂N 3 ∂N 4  
 ∂ξ 0 0 0 0  v1
∂ξ ∂ξ ∂ξ  
  u 
0   ∂N 1 ∂N 2 ∂N 3 ∂N 4
 J 11* 0  
2
J 12* 0 0 0 0
 *   ∂η ∂η ∂η ∂η  2 
v
[ε ] =  0 0 *
J 21 J 22 × ∂N 1 ∂N 2 ∂N 3 ∂N 4  u 3 
 J 21
* *
J 22 J 11* J 12*   0 0 0 0  
  ∂ξ ∂ξ ∂ξ ∂ξ   v3 
 ∂N 1 ∂N 2 ∂N 3 ∂N 4   
 0 0 0 0 u4
 ∂η ∂η ∂η ∂η   
 v 4 

 J 11* J 12* 0 0 
 * 
[ε ] =  0  × [B']{d } = [ B]{d }
*
0 J 21 J 22
 J 21
* * * 
 J 22 J 11* J 12 

 1 1 1 1 
− 4 0 4 0 4
0 −
4
0
1 
 1 1 1 1   35 0 0 0
− 0 − 0  1
0 0
[B ] =
/  4 4 4 4  and [ B] =  0 0 0 [B']
 0 −1 0 1 0 − 
1 1  25 
 0
0 0 
4 4 4 4 1 1
  
0 −
1
0 −
1
0
1
0
1
 25 35 
 4 4 4 4

7.143 0 7.143 0 7.143 0 7.143 0 

 
[B]   0 10 0 10 0 10 0 10 
 
 10 7.143 10 7.143 10 7.143 10  7.143
 

For plane stress condition

 
1 µ 0  1 0 0 
0  = 2 × 10 0 1 0 
E 
[C ] = µ 1 3

1− µ 2  1− µ 
0 0  0 0 1 / 2
 2 
 1 0 0 
[C ][ B] = 2 × 10 0 1 0  ×
3

0 0 1 / 2

7.143 0 7.143 0 7.143 0 7.143 0 

 
 0 10 0 10 0 10 0 10 
 
 10 7.143 10 7.143 10 7.143 10  7.143
 
Assume the values of gauss weight, w = 2, the stiffness matrix [k] at this sampling point is
[k ] = tw [ B]T [C ] [ B] | J | , Where t is thickness of the element. Thus,
 −7.143 0 −10 
 0 −10 −7.143 

 7.143 0 −10 
  1 0 0 
−10 7.143  
10 × 0 1 0 
 0
[ k ] = 20 × 2 × 875 × 2 ×10 × 
3

7.143 0 10
  0 0 1/ 2 
 0 10 7.143  
 −7.143 0 4 
 
 0 10 −7.143 
 −7.143 0 7.143 0 7.143 0 −7.143 0 

× 0 −10 0 −10 0 10 0 10 
 −10 −7.143 −10 7.143 10 7.143 10 −7.143

 7.071 2.5 −0.714 −2.5 −7.071 −2.5 0.714 2.5 

 2.5 8.785 2.5 5.214 −2.5 −8.785 2.5 −5.214 

 −0.714 2.5 7.071 −2.5 0.714 −2.5 −7.071 2.5 
 
 −2.5 5.214 −2.5 8.785 2.5 −5.214 2.5 −8.785 
= 10 ×
3
 −7.071 −2.5 0.714 2.5 7.071 2.5 −0.714 −2.5 
 
 −2.5 −8.785 −2.5 −5.214 2.5 8.785 2.5 5.214 
 0.714 −2.5 −7.071 2.5 −0.714 2.5 7.071 −2.5 
 
 2.5 −5.214 2.5 −8.785 −2.5 5.214 −2.5 8.785 

5.4.3 Evaluation of Stiffness using Two Point Gauss Quadrature

In this case, 2×2 Gauss Quadrature integration procedure has been carried out to the calculate the
stiffness matrix of the same element for a comparison. The natural coordinate of the sampling
point is shown in the figure below.

Fig. 5.4.3 Natural Coordinates for Two Points Gauss Quadrature

The natural co-ordinates of the sampling points for 2×2 Gauss Quadrature integration are
1 +0.57735 +0.57735
2 -0.57735 +0.57735
3 -0.57735 -0.57735
4 +0.57735 -0.57735

For a four node quadrilateral element, the shape functions and their derivatives are as follows.

(1 − ξ )(1 − η ) (1 + ξ )(1 − η ) (1 + ξ )(1 + η ) (1 − ξ )(1 + η )

N1 = , N2 = , N3 = and N 4 =
4 4 4 4
∂N1 − (1 − ξ ) ∂N 2 (1 − ξ ) ∂N 3 (1 + ξ ) ∂N 4 − (1 + ξ )
= ; = ; = ; =
∂ξ 4 ∂ξ 4 ∂ξ 4 ∂ξ 4
∂N1 − (1 − η ) ∂N 2 − (1 + η ) ∂N 3 (1 + η ) ∂N 4 (1 − η )
= ; = ; = ; =
∂η 4 ∂η 4 ∂η 4 ∂η 4
The Jacobian matrix will be
 ∂N1 ∂N 2 ∂N 3 ∂N 4   x1 y1  x y1 
 −(1 − ξ ) +(1 − ξ ) +(1 + ξ ) −(1 + ξ )   1 
 ∂ξ  
∂ξ ∂ξ ∂ξ   x 2 y2   4 4 4 4   x2 y2 
| J | =   
 ∂N1 ∂N 2 ∂N 3 ∂N 4   x3 y3   −(1 − η ) −(1 + η ) +(1 + η ) +(1 − η )   x y3 

 ∂η    3
 ∂η ∂η ∂η   x4 y4   4 4 4 4   x y4 
 4
(a) At sampling point 1, (ξ=0.57735, η=0.57735)
The value of the Jacobian, [J] at sampling point 1 will become

 − (1 − 0.57735) + (1 − 0.57735) + (1 + 0.57735) − (1 + 0.57735)   0 0 

 4 4 4 4  70 0 
[J ] = 
− (1 − 0.57735) − (1 + 0.57735) + (1 + 0.57735) + (1 − 0.57735)  70 50
 
 4 4 4 4   0 50

1 
 0
35 0  35  J11* J12* 
Thus, [J ] =  0 25 =
and [ J ]−1
=   * * 
; Thus |𝐽| = 875
  0 1   J 21 J 22 
 25 
 ∂u 
 ∂ξ 
 
 J 11* J 12* 0 0   ∂u 
 *   ∂η 
[ε ] =  0 0 *
J 21 J 22  ×  ∂v 
 J 21

* *
J 22 J 11* J 12*   
 ∂ξ 
 ∂v 
 ∂η 
 
 u1 
 ∂N 1 ∂N 2 ∂N 3 ∂N 4  
 ∂ξ 0 0 0 0  v1
∂ξ ∂ξ ∂ξ  
  u 
0   ∂N 1 ∂N 2 ∂N 3 ∂N 4
 J 11* 0  
2
J 12* 0 0 0 0
 *   ∂η ∂η ∂η ∂η  v2 
[ε ] =  0 0 J *
J 22 ×
21
∂N 1 ∂N 2 ∂N 3 ∂N 4  u 3 
 J 21
*
J *
J *
J 12*   0 0 0 0  
 22 11
 ∂ξ ∂ξ ∂ξ ∂ξ   v3 
 ∂N 1 ∂N 2 ∂N 3 ∂N 4   
 0 0 0 0 u4
 ∂η ∂η ∂η ∂η   
 v 4 
 J 11* J 12* 0 0 
 * 
[ε ] =  0  × [B']{d } = [ B]{d }
*
0 J 21 J 22
 J 21

* *
J 22 J *
11 J 12* 
  1 − 0.57735 1 − 0.57735 1 + 0.57735 1 + 0.57735 
− 4
0
4
0
4
0 −
4
0 
 
 − 1 − 0.57735 0 −
1 + 0.57735
0
1 + 0.57735
0
1 − 0.57735
0 
 4 4 4 4 
′
[B ] =  
 1 − 0.57735 1 − 0.57735 1 + 0.57735 1 + 0.57735 
0 − 0 0 0 −
 4 4 4 4 
 1 − 0.57735 1 + 0.57735 1 + 0.57735 1 − 0.57735 
 0 − 0 − 0 0 
 4 4 4 4 

 −0.1057 0 0.1057 0 0.3943 0 −0.3943 0 

 −0.1057 0 −0.3943 0 0.3943 0 0.1057 0 
[ B '] = 
 0 −0.1057 0 0.1057 0 0.3943 0 −0.3943
 
 0 −0.1057 0 −0.3943 0 0.3943 0 0.1057 

Thus,
1 
 35 0 0 0  −0.1057 −0.3943
 0 0.1057 0 0.3943 0 0 
  
1   −0.1057 0 −0.3943 0 0.3943 0 0.1057 0 
[ B]  0 0 0 ×
 25   0 −0.1057 0 0.1057 0 0.3943 0 −0.3943
   
0 1 1
0  0 −0.1057 0 −0.3943 0 0.3943 0 0.1057 
 25 35 

 − 0.003 0 0.003 0 0.0113 0 − 0.0113 0 


[B] =  0 − 0.0042 0 − 0.0158 0 0.0158 0 0.0042 
− 0.0042 − 0.003 − 0.0158 0.003 0.0158 0.0113 0.0042 − 0.0113

For plane stress condition

 
1 µ 0  1 0 0 
0  = 2 × 10 0 1 0 
E 
[C ] = µ 1 3

1− µ 2  1− µ 
0 0  0 0 1 / 2
 2 
 1 0 0 
[C][B] = 2 × 10 0 1 0  ×
3

0 0 1 / 2

 − 0.003 0 0.003 0 0.0113 0 − 0.0113 0 


 0 − 0.0042 0 − 0.0158 0 0.0158 0 0.0042 
− 0.0042 − 0.003 − 0.0158 0.003 0.0158 0.0113 0.0042 − 0.0113

The values of gauss weights are wi=wj=1.0. Therefore, the stiffness matrix [k] at this sampling
point is [k ] = twi w j [ B ]ij [C ]ij [ B ]ij | J | , where t is thickness of the element. Thus at sampling point
T

1,
 −0.003 0 −0.0042 
 0 −0.0042 −0.003 

 0.003 0 −0.0158
  1 0 0 
−0.0158 0.003  
× 0 1 0  ×
0
[ k1 ]= 20 ×1×1× 875 × 2 ×103 × 
0.0113 0 0.0158  
   0 0 1/ 2 
 0 0.0158 0.0113  
 −0.0113 0 0.0042 
 
 0 0.0042 −0.0113
 −0.003 0 0.003 0 0.0113 0 −0.0113 0 
 0 −0.0042 0 −0.0158 0 0.0158 0 0.0042 

 −0.0042 −0.003 −0.0158 0.003 0.0158 0.0113 0.0042 −0.0113
 0.0632 0.0223 0.0848 −0.0223 −0.2358 −0.0834 0.0878 0.0834 
 0.0785 0.0834 0.2174 −0.0834 −0.2929 −0.0223 −0.0030 

 0.4672 −0.0834 −0.3162 −0.3109 −0.2358 0.3109 
 
 0.8866 0.0834 −0.8111 0.0223 −0.2929 
[ k=
1 ] 10 ×
4
 0.8795 0.3109 −0.3275 −0.3109 
 
 sym 1.0927 0.0834 0.0113 
 0.4755 −0.0834 
 
 0.2847 

(b) At sampling point 2, (ξ=-0.57735, η=0.57735)

The value of the Jacobian, [J] at sampling point 2 can be calculated in a similar way and finally
the strain-displacement relationship matrix and then the stiffness matrix [k2] can be evaluated
and is shown below.

 −0.0113 −0.0042 
0
 0 −0.0113
−0.0042

 0.0113 −0.0158
0
 
0 −0.0158
0.0113 
[ k2 ]= 20 ×1×1× 875 × 2 ×103  ×
0.003 0.0158 
0
 
 0 0.003 
0.0158
 −0.003 0.0042 
0
 
 0 −0.0030 
0.0042
 −0.0113 0 0.0113 0 0.0030 0 −0.0030 0 
 0 −0.0042 0 −0.0158 0 0.0158 0 0.0042 

 −0.0021 −0.0056 −0.0079 0.0056 0.0079 0.0015 0.0021 −0.0015

0.4756 0.0833 −0.3276 −0.0833 −0.2357 −0.0223 0.0878 0.0223 

 0.2847 0.3110 0.0112 −0.3110 −0.2929 −0.0833 −0.0030 

 0.8797 −0.3110 −0.3164 −0.0833 −0.2357 0.0833 
 
 1.0930 0.3110 −0.8113 0.0833 −0.2929 
[ k=
2 ] 10 ×
4
 0.4673 0.0833 0.0848 −0.0833
 
 sym 0.8868 0.0223 0.2174 
 0.0632 −0.0223
 
 0.0785 
(c) At sampling point 3, (ξ=-0.57735, η=-0.57735)
The value of the strain-displacement relationship matrix and then the stiffness matrix [k3] can be
evaluated and is shown below.

 −0.0113 0 −0.0158
 0 −0.0158 −0.0113

 0.0113 0 −0.0042 
 
0 −0.0042 0.0113 
[ k3 ]= 20 ×1×1× 875 × 2 ×103  ×
0.0030 0 0.0042 
 
 0 0.0042 0.0030 
 −0.0030 0 0.0158 
 
 0 0.0158 −0.0030 
 −0.0030 0 0.0113 0 0.0030 0 −0.0030 0 
 0 −0.0158 0 −0.0042 0 0.0042 0 0.0158 

 −0.0079 −0.0056 −0.0021 0.0056 0.0021 0.0015 0.0079 −0.0015

0.8797 0.3110 −0.3276 −0.3110 −0.2357 −0.0833 −0.3164 0.0833 

 1.0930 0.0833 0.0112 −0.0833 −0.2929 −0.3110 −0.8113

 0.4756 −0.0833 0.0878 −0.0223 −0.2357 0.0223 
 
 0.2847 0.0833 −0.0030 0.3110 −0.2929 
[ k=
3 ] 10 ×
4
 0.0632 0.0223 0.0848 −0.0223
 
 sym 0.0785 0.0833 0.2174 
 0.4673 −0.0833
 
 0.8868 

(d) At sampling point 4, (ξ=0.57735, η=-0.57735)

The value of the strain-displacement relationship matrix and then the stiffness matrix [k3] can be
evaluated and is shown below.
  −0.0030 0 −0.0158
 0 −0.0158 −0.0030 

 0.0030 0 −0.0042 
 
0 −0.0042 0.0030 
[ k4 ]= 20 ×1×1× 875 × 2 ×103  ×
0.0113 0 0.0042 
 
 0 0.0042 0.0113 
 −0.0113 0 0.0158 
 
 0 0.0158 −0.0113
 −0.0030 0 0.0030 0 0.0113 0 −0.0113 0 
 0 −0.0158 0 −0.0042 0 0.0042 0 0.0158 

 −0.0079 −0.0015 −0.0021 0.0015 0.0021 0.0056 0.0079 −0.0056 

0.4673 0.0833 0.0848 −0.0833 −0.2357 −0.3110 −0.3164 0.3110 

 0.8868 0.0223 0.2174 −0.0223 −0.2929 −0.0833 −0.8113

 0.0632 −0.0223 0.0878 −0.0833 −0.2357 0.0833 
 
 0.0785 0.0223 −0.0030 0.0833 −0.2929 
[ k=
4 ] 10 ×
4
 0.4756 0.0833 −0.3276 −0.0833
 
 sym 0.2847 0.3110 0.0112 
 0.8797 −0.3110 
 
 1.0930 

The stiffness matrix of the element can be computed as the sum of the values at the four
sampling points: [k ] = [k1 ] + [k2 ] + [k3 ] + [k4 ] . Thus, the final value of the stiffness matrix will
become

1.8857 0.5000 −0.4857 −0.5000 −0.9429 −0.5000 −0.4571 0.3110 

 2.3429 0.5000 0.4571 −0.5000 −1.1714 −0.5000 −1.6286 

 1.8857 −0.5000 −0.4571 −0.5000 −0.9429 0.5000 
 
2.3429 −0.5000 −1.6286 0.5000 −1.1714 
k ] 104 × 
[=  1.8857 0.5000 −0.4857 −0.5000 
 
 sym 2.3429 0.5000 0.4571 
 1.8857 −0.5000 
 
 2.3429 
Lecture 5: Computation of Stresses, Geometric Nonlinearity and Static Condensation

5.5.1 Computation of Stresses

After solving the static equation of {F} = [K]{d}, the nodal displacement {d} can be obtained in
global coordinate system. The element nodal displacement d  can then be calculated from the
nodal connectivity of the element. Using strain-displacement relation and then stress-strain
relation, the stress at the element level are derived.
σ
ε
D
= DB  d       (5.5.1)
Here,  is the stress at the Gauss point of the element as the sampling points for the integration
has been considered as Gauss points. Here, [D] is the constitutive matrix, [B] is the strain
displacement matrix of the element. As a result these stresses at Gauss points need to extrapolate
to the corresponding nodes of the element. It is well established that 2 × 2 Gauss integration
points are the optimal sampling points for two dimensional isoparametric elements. The ‘local
stress smoothing’ is a technique that can be used to extrapolate stresses computed at Gauss
points to nodal points. The stresses are computed at four Gauss points (I, II, III and IV) of an 8
node element as shown in Fig. 5.5.1. For example, at point III, r = s =1 an d ξ = η = 1 3.
Therefore the factor of proportionality is 3 ; i.e.,
r = ξ 3 and s = η 3 (5.5.2)
Stresses at any point P in the element are found by the usual shape function as
′ σ i for i = 1,2,3,4
σ P = ∑ N di (3.5.3)
In the above equation, σ P is σ x , σ y and τ xy at point P. N di
′ are the bilinear shape functions
written in terms of r and s rather than ξ and η as

′ = (1 ± r )(1 ± s )
1
N di (5.5.4)
4
′ are evaluated at r and s coordinates of point P. Let the point P coincides with the corner 1.
N di
To calculate stress σ x1 at corner 1 from σ x values at the four Gauss points, substitution of r and
s into the shape functions will give
σ xI =1.8666σ xI − 0.500σ xII + 0.134σ xIII − 0.500σ xIV (5.5.5)
Fig. 5.5.1 Natural coordinate systems used in extrapolation of stresses from Gauss points

The resultant extrapolation matrix thus obtained may be written as

σ 1  (
 1+ 3 2 ) − 0.5 (1 −3 2 ) − 0.5 
σ 
 2

 − 0.5 (
1+ 3 2 ) − 0.5 (
1− 3 2 

)
σ 3  (
 1− 3 2

) − 0.5 (
1+ 3 2 ) − 0.5 

σI 
 
σ 4  =  − 0.5 (
1− 3 2 ) − 0.5 (
1+ 3 2  ) σ 
 II 
σ 5 
 
( )
 1+ 3 4

( )
1+ 3 4 (
1− 3 4 ) ( )
1 − 3 4


σ III 
 (5.5.6)

σ 6  ( )
 1− 3 4 ( )
1+ 3 4 (
1+ 3 4 ) ( )
1 − 3 4 σ IV 
σ 
 7
(
 )
 1− 3 4 ( )
1− 3 4 (
1+ 3 4 ) ( )
1 + 3 4

σ 8  ( )
 1 + 3 4 ( )
1− 3 4 (
1− 3 4 ) ( )
1 + 3 4

Here, σ 1 , σ 2 …….. σ 8 are the smoothened nodal values and σ I … σ IV are the stresses at the
Gauss points. Smoothened nodal stress values for four node rectangular element can be also be
evaluated in a similar fashion. The relation between the stresses at Gauss points and nodal point
for four nodel element will be
  1 
1  3 
1
1 
3

 2 2 2 2 

1   3  I 
   1 3 1
2    2 1  1  
2  II 
    2 2
 III 
(5.5.7)
3   3 1 3 1
  1   1    

4   2 2 2 2  IV 
 3 
 1 1 
3

1
1 
 2 2 2 2 

The stress at particular node joining with more than one element will have different magnitude as
calculated from adjacent elements (Fig. 5.5.2(a)). The stress resultants are then modified by
finding the average of resultants of all elements meeting at a common node. A typical stress
distribution for adjacent elements is shown in Fig. 5.5.2(b) after stress smoothening.
 Fig. 5.5.2 Stress smoothening at common node

5.5.2 Geometric Nonlinearity

As discussed earlier, nonlinear analysis is mainly of two types: (i) Geometric nonlinearity and
(ii) Material nonlinearity. For geometric nonlinearity consideration, the relation between strain
and displacement is of utmost importance in the finite element formulation for stress analysis
problems. In case of plane stress/strain problem, the nonlinear term of the strain expression are
dropped for the sake of simplicity in the analysis. However, for large displacement problems, the
nonlinear strain term plays a vital role to obtain accurate response. The generalized strain-
displacement relations for the two-dimensional plane stress/strain problems are rewritten here to
derive the nonlinear solution.

∂u 1  ∂u   ∂v  
2 2

εx =+   +  
∂x 2  ∂x   ∂x 
∂v 1  ∂u   ∂v  
2 2

εy = +   +   
∂y 2  ∂y   ∂y  
(5.5.8)
∂v ∂u  ∂u ∂u ∂v ∂v 
γ xy = + +  +
∂x ∂y  ∂x ∂y ∂x ∂y 

The displacements at any point inside the node are expressed in terms of their nodla
displacements. Thus,
n n
=u N u [ N ]{u } =
∑= and v ∑
= N v [ N ]{v }
i =1
i i i i
i =1
i i i i (5.5.9)
Therefore,
 ∂u ∂ [ N i ]
= = {ui } [ B1 ]{ui }
∂x ∂x
∂v
= [ B1 ]{vi }
∂x
∂u
= [ B2 ]{ui }
∂y
∂v
= [ B2 ]{vi }
∂y
(5.5.10)
Here, [B1] and [B2] are the derivative of the shape function [Ni] with respect to x and y
respectively. The vectors {ui} and {vi} represent the nodal displacements vectors in x and y
directions respectively. The vector of strains at any point inside an element, {ε } may be
expressed in terms of nodal displacement as
{ε } = [ B ] {d }
(5.5.11)
where [B] is the strain displacement matrix. {d} is the nodal displacement vector and may be
expressed as
{u }
{d } =  i 
{vi } 
(5.5.12)
The matrix [B] may be expressed with two components as
[=
B ] [ Bl ] + [ Bnl ]
(5.5.13)
where, [Bl] and [Bnl] are the linear and nonlinear part of the strain-displacement matrix
respectively and are expressed as follows:
 [B1 ] [0] 
[Bl ] =  [0] [B2 ]
[B2 ] [B1 ]
(5.5.14)
and
 1 T 
 2 {u } [B1 ] [B1 ] 2 {v } [B1 ] [B1 ] 
T 1 T T

 
[Bnl ] =  1 {u }T [B2 ]T [B2 ] 1 {v }T [B2 ]T [B2 ]
2 T 2 
 {u } [B2 ]T [B1 ] {v }T [B2 ]T [B1 ] 
 
(5.5.15)
5.5.2.1 Steps to include effect of geometrical nonlinearity
The nonlinear geometric effect of the structure at a particular instant of time can be obtained by
performing the following steps.
1. Calculation of displacement {d}1 considering linear part of strain matrix [Bl].
2. Evaluation of nonlinear part of the strain matrix [Bnl] (eq5.5.15) adopting {d}1 from
previous step.
3. Evaluation of total strain matrix [B] = [Bl] + [Bnl].
4. Calculation of displacement {d}2 considering both linear and nonlinear part of strain
matrix [B].
5. Repetition of steps 2 to 4 with {d}2, from which modified displacement, {d}3 are
obtained.
6. Step 5 is carried out until the displacements for two consecutive iteration converge
i.e.,
{d } j − {d } j −1
<ε
{d } j
Where ε is any pre-assigned small value and j is the number of iterations.

5.5.3 Static Condensation

The higher order Lagrangian elements (i.e., nine node, sixteen node rectangular element) contain
number of internal nodes. This is necessary sometimes for the completeness of the desired
polynomial used in displacement function for derivation of interpolation function. These internal
nodes are not connected to the adjoining elements in the assemblage (Fig. 5.5.3).
 Fig. 5.5.3 Internal nodes of Nine node elements

Thus, the displacements of these nodes are not required to formulate overall equilibrium
equations of the structure. This limits the usefulness of these elements. A technique known as
“static condensation” can be used to suppress the degrees of freedom associated with the internal
nodes in the final computation. The technique of static condensation is explained below. The
equilibrium equation for a system are expressed in the finite element form as
F  =  K d  (5.5.16)
Where, {F}, [K] and {d} are the load vector, stiffness matrix and displacement vector for the
entire structure. The above equation can be rearranged by separating the relevant terms
corresponding to internal and external nodes of the elements.
Fi   K ii   K ie di 
 =    (5.5.17)
Fe   K ei   K ee  d e 
    
Here, {di} and {de} are the displacement vectors corresponding to internal and external nodes
respectively. Similarly, {Fi} and {Fe} are force vectors corresponding to internal and external
nodes. Now, the above expression can be written in the following form separately.

Fi    Kii di    Kie de  (5.5.18)

Fe    K ei di    K ee de  (5.5.19)

The stiffness matrix and nodal load vector corresponds to the internal nodes can be separated out.
For this, eq.(5.5.18) can be rewritten as
1 1
di    Kii  Fi    Kii   Kie de  (5.5.20)
Substituting the value of {di} obtained from the above equation in eq.(5.5.19), the following
expression will be obtained.
 Fe    K ei  Kii  Fi    Kii   Kie de    K ee de 
1 1
(5.5.21)

Here, the equations are reduced to a form involving only the external nodes of the elements. The
above reduced substructure equations are assembled to achieve the overall equations involving
only the boundary unknowns. Thus the above equation can be rewritten as

F   K
e ei  Kii  Fi    K ee    K ei  Kii   Kie de  de 
1 1
(5.5.22)

Or
Fc  =  K c de  (5.5.23)

Fc  = Fe  -  K ei  Kii  Fi  and  K c  =  K ee  -  K ei  Kii   Kie de  . Here, [Kc]
-1 -1
Where,
is called condensed or reduced stiffness matrix and {Fc} is the condensed or effective nodal load
vector corresponding to external nodes of the elements. In this process, the size of the matrix for
inversion will be comparatively small. The unknown displacements of the exterior nodes, {de}
can be obtained by inverting the matrix [Kc] in eq.(5.5.23). Once, the values of {de} are obtained,
the displacements of internal nodes {di} can be found from eq.(5.5.20).
Lecture 6: Axisymmetric Element

5.6.1 Introduction
Many three-dimensional problems show symmetry about an axis of rotation. If the problem
geometry is symmetric about an axis and the loading and boundary conditions are symmetric
about the same axis, the problem is said to be axisymmetric. Such three-dimensional problems
can be solved using two-dimensional finite elements. The axisymmetric problem are most
conveniently defined by polar coordinate system with coordinates (r, θ, z) as shown in Fig. 5.6.1.
Thus, for axisymmetric analysis, following conditions are to be satisfied.
1. The domain should have an axis of symmetry and is considered as z axis.
2. The loadings on the domain has to be symmetric about the axis of revolution, thus they
are independent of circumferential coordinate θ.
3. The boundary condition and material properties are symmetric about the same axis and
will be independent of circumferential coordinate.

Fig. 5.6.1 Cylindrical coordinates

Axisymmetric solids are of total symmetry about the axis of revolution (i.e., z-axis), the field
variables, such as the stress and deformation is independent of rotational angle θ. Therefore, the
field variables can be defined as a function of (r, z) and hence the problem becomes a two
dimensional problem similar to those of plane stress/strain problems. Axisymmetric problems
includes, circular cylinder loaded with uniform external or internal pressure, circular water tank,
pressure vessels, chimney, boiler, circular footing resting on θsoil
• P( r,z, ) mass, etc.

5.6.2 Relation between Strain and Displacement

An axisymmetric problem is readily described in cylindrical polar coordinate system: r, z and θ.
Here, θ measures the angle between the plane containing the point and the axis of the coordinate
system. At θ = 0, the radial and axial coordinates coincide with the global Cartesian X and Y
coordinates. Fig. 5.6.2 shows a cylindrical coordinate system and the definition of the position
ˆ zˆ and ˆ be unit vectors in the radial, axial, and circumferential directions at a point
vectors. Let r,
in the cylindrical coordinate system.

Fig. 5.6.2 Cylindrical Coordinate System

If the loading consists of radial and axial components that are independent of θ and the material
is either isotropic or orthotropic and the material properties are independent of θ, the
displacement at any point will only have radial (𝑢𝑟 ) and axial (𝑢𝑧 ) components. The only stress
components that will be nonzero are 𝜎𝑟𝑟 , 𝜎𝑧𝑧 , 𝜎𝜃𝜃 𝑎𝑛𝑑 𝜏𝑟𝑧 .

(a) Element in r-z plane (b) Element in r-θ plane

 Fig. 5.6.3 Deformation of the axisymmetric element
A differential element of the body in the r-z plane is shown in Fig. 5.6.3(a). The element
undergoes deformation in the radial direction. Therefore, it initiates increase in circumference
and associated circumferential strain. Let denote the radial displacement as u, the circumferential
displacement as v, and the axial displacement as w. Dashed line represents the deformed
positions of the body in Fig. 5.6.3(b). The radial strain can be calculated from the above diagram
as
1  u  u
=
εr u +- u =
× dr  (5.6.1)
dr  r  r
Since the rz plane is effectively the same as a rectangular coordinate system, the axial strain will
become
1  w  w
=
εz ×dzw+- w =  (5.6.2)
dz  z  z

Considering the original arc length versus the deformed arc length, the differential element
undergoes an expansion in the circumferential direction. Before deformation, let the arc length is
assumed as ds = rdθ. After deformation, the arc length will become ds = (r+u) dθ. Thus, the
tangential strain will be
r +u  d  - rd u
ε =  (5.6.3)
rd  r
Similarly, the shear strain will be

u w
 rz  
z r (5.6.4)
 r  0 and  z  0

Thus, there are four strain components present in this case and is given by
 ∂u  ∂ 
 ∂r   ∂r 0
εr     
   ∂w  0 ∂
εz   ∂z   ∂z   u 
{ε } =
=   =     (5.6.5)
 εθ   u  1 0  
 w
γ rz   r  r 
 ∂u ∂w  ∂ ∂
 +   
 ∂z ∂r   ∂z ∂r 
5.6.3 Relation between Stress and Strain
The stress strain relation for axisymmetric case can be derived from the three dimensional
constitutive relations. We know the stress-strain relation for a three-dimensional solid is

1−𝜇 𝜇 𝜇 0 0 0
⎡ ⎤
𝜎𝑥 ⎢ 𝜇 1−𝜇 𝜇 0 0 0 ⎥ ϵ𝑥
⎧𝜎𝑦 ⎫ ⎢ ⎥ ⎧ϵ ⎫
𝑦
⎪𝜎 ⎪ ⎢ 𝜇 𝜇 1−𝜇 0 0 0 ⎥ ⎪ϵ ⎪
𝐸
𝑧
= (1+𝜇)(1−2𝜇) ⎢ ⎥ 𝑧
𝜏 0 ⎥ ⎨𝜈𝑥𝑦 ⎬
1−2𝜇
⎨ ⎬𝑥𝑦 ⎢ 0 0 0 2
0
⎪𝜏𝑦𝑧 ⎪ ⎢ 1−2𝜇 ⎥ ⎪𝜈𝑦𝑧 ⎪
⎩𝜏𝑧𝑥 ⎭ ⎢ 0 0 0 0 2
0 ⎥ ⎩𝜈𝑧𝑥 ⎭
⎢ 1−2𝜇 ⎥
⎣ 0 0 0 0 0 2 ⎦
(5.6.6)

The stresses acting on a differential volume of an axisymmetric solid under axisymmetric

loading is shown in Fig. 5.6.4.

Fig. 5.6.4 Stresses acting on a differential volume

Now, comparing the stress-strain components present in the axisymmetric case, the stress-strain
relation can be expressed from the above expression as follows
 𝜎𝑟 1−𝜇 𝜇 𝜇 0 𝜀
⎧ ⎫ ⎡ ⎤⎧ 𝑟 ⎫
⎪𝜎𝑧 ⎪ ⎢ 𝜇 1−𝜇 𝜇 0 ⎥ ⎪𝜀𝑧 ⎪
𝐸
= ⎢ ⎥ (5.6.7)
⎨𝜎𝜃 ⎬ (1+𝜇)(1−2𝜇) ⎢ 𝜇 𝜇 1−𝜇 0 ⎥ ⎨𝜀𝜃 ⎬
⎪ ⎪ ⎢ 1−2𝜇 ⎥
⎪ ⎪
⎩𝜏𝑟𝑧 ⎭ ⎣ 0 0 0 2 ⎦
⎩𝜈𝑟𝑧 ⎭

Thus, the constitutive matrix [D] for the axisymmetric elastic solid will be
1−𝜇 𝜇 𝜇 0
⎡ ⎤
⎢ 𝜇 1−𝜇 𝜇 0 ⎥
𝐸
[D] = (1+𝜇)(1−2𝜇) ⎢ ⎥ (5.6.8)
⎢ 𝜇 𝜇 1−𝜇 0 ⎥
⎢ 1−2𝜇 ⎥
⎣ 0 0 0 2 ⎦

5.6.4 Axisymmetric Shell Element

A cylindrical liquid storage container like structures (Fig. 5.6.5) may be idealized using
axisymmetric shell element for the finite element analysis. It may be noted that the liquid in the
container may be idealized with two dimensional axisymmetric elements. Let us consider the
radius, height and, thickness of the circular tank are R, H and h respectively.

Fig. 5.6.5 Thin wall cylindrical container

The strain energy of the axisymmetric shell element (Fig. 5.6.6) including the effect of both
stretching and bending are expressed as
H
U = ε ∫+( N
N y εθy θ+ M χy y2πRdy)
1
(5.6.9)
20
Here, Ny and Nθ are the membrane force resultants and My is the bending moment resultant. The
shell is assumed to be linearly elastic, homogeneous and isotropic. Thus the force and moment
resultants can be expressed in terms of the mid-surface change in curvature χy as follows.

Fig 5.6.6 Axisymmetric plate element

Here, the strain-displacement relation is given by

 {σ} = [ D ] {ε} (5.6.10)

In which,
 
N y  ε y  1 µ 0 
     Eh 
 {σ } =  Nθ  ,  {ε } = ε θ  and  [ D ] =µ 1 0  (5.6.11)
M  χ  1− µ 2 2 
 y  y 0 0 h 
 12 
The generalized strain vector can be expressed in terms of the displacement vectors as follows.
{ε } = [ B ] {d } (5.6.12)
Where,
  ∂
 0 ∂y 
 
u   1 
 {d } =   and  [ B ] =  0 (5.6.13)
v  
R

 ∂ 
2

 − ∂y 2 0 
 
Here, u and v are the displacement components in two perpendicular directions. With the use of
stress and strain vectors, the potential energy expression are written in terms of displacement
vectors as

∫ ({d } [ B ] [ D ][ B ]dy
{d } )
H
1
U = πR× 2
T T
(5.6.14)
2 0

Thus, the element stiffness are derived as

H
 [ k ] = 2πR ∫ [ B ] [ D ][ B ] dy
T
(5.6.15)
0

Similarly, neglecting the rotary inertia, the kinetic energy can be expressed as

{d})
∫ ({d} [ N ] m [ N ]dy
H
1 T
T = πR× 2
T
(5.6.16)
2 0

{}
Where, m denotes the mass of the shell element per unit area and   d represents the velocity
vector. Thus, the element mass matrix is given by
Le

 [ M ] = 2πRm ∫ [ N ] [ N ] dy
T
(5.6.17)
0

Lecture 7: Finite Element Formulation of Axisymmetric Element

Finite element formulation for the axisymmetric problem will be similar to that of the two
dimensional solid elements. As the field variables, such as the stress and strain is independent of
rotational angle θ, circumferential displacement will not appear. Thus, the displacement field
variables are expressed as
n
u r,z  =  N i r,z ui
i=1
n
(5.7.1)
wr,z  =  N i r,z  wi
i=1
Here, ui and wi represent radial and axial displacements respectively at nodes. Ni (r, z) are the
shape functions. As the geometry and field variables are independent of rotational angle θ, the
interpolation function Ni (r, z) can be expressed similar to 2-dimensional problems by replacing
the x and y terms with r and z terms respectively.
5.7.1 Stiffness Matrix of a Triangular Element
Fig. 5.7.1 shows the cylindrical coordinates of a three node triangular element. Hence the
analysis of the axisymmetric element can be approached in a similar way as the CST element.
Thus the field variables of such an element can be expressed as
u =α 0 + α1r + α 2 z
(5.7.2)
w =α 3 + α 4 r + α 5 z
Or,
{d } = [φ ]{α } (5.7.3)
Where,
1 r z 0 0 0 
u  ,
 [φ ] =  {α } = {α 0 α1 α 2 α 3 α 4 α 5 }
T
{d } =   and
 w 0 0 0 1 r z 
Using end conditions,
 u1  1 ri zi 0 0 0  α 0 
 u  1 r zj 0 0 0  α1 
 2  j

 u3  1 rk zj 0 0 0  α 2 
 =   (5.7.4)
 w1  0 0 0 1 ri zi  α 3 
 w2  0 0 0 1 rj z j  α 4 
    
 w3  0 0 0 1 rk zk  α 5 
Or,
{d } = [ A]{α }
(5.7.5)
[ A] {d }
⇒ {α } =
−1

{}
Here d are the nodal displacement vectors.
 Fig. 5.7.1 Axisymmetric three node triangle in cylindrical coordinates

Putting above values in eq.(5.7.3), the following relations will be obtained.

][ A] {d } [ N ]{d }
{d } [φ=
−1
= (5.7.6)
Or,
 r1 
r 
 2
 u   Ni Nj Nk 0 0 0   r3 
{=
d} =    
N k   z1 
(5.7.7)
 w  0 0 0 Ni Nj
 z2 
 
 z3 
Using a similar approach as in case of CST elements, the three shape functions [ N1 , N 2 , N 3 ] can
be assumed as,

1
N1 (=
r, z ) ( r2 z3 − r3 z2 ) + ( z2 − z3 ) r + ( r3 − r2 ) z 
2A 
1
N 2 (=
r, z ) ( r3 z1 − r1 z3 ) + ( z3 − z1 ) r + ( r1 − r3 ) z 
2A 
1
N 3 (=
r, z ) ( r1 z2 − r2 z1 ) + ( z1 − z2 ) r + ( r2 − r1 ) z 
2A 
 Or,
1
N i ( r , z=
) (α i + r βi + zγ i )
2A
N j ( r , z=
)
1
2A
(α j + r β j + zγ j ) (5.7.8)

1
N k ( r , z=
) (α k + r β k + zγ k )
2A

Where,
αi =
rj zk − rk z j αj =
rk zi − ri zk αk =
ri z j − rj zi
βi =
z j − zk βj =
zk − zi βk =
zi − z j
γi =
rk − rj γi =
ri − rk γi =
rj − ri (5.7.9)

2 A=
1
2
( ri z j + rj zk + rk zi − ri zk − rj zi − rk z j )
Putting the value of {u,w} in eq. (5.7.7) from eq. (5.6.5),
 ∂N i ∂N j ∂N k 
 0 0 0 r 
 ∂r ∂r ∂r 1
 
 Ni Nj Nk   r2 
 r 0 0 0   r3 
{ε } =
r r
  [ B ]{d } (5.7.10)
 ∂N i ∂N j ∂N k   z1 
 0 0 0
∂z ∂z ∂z   z2 
  
 ∂N i ∂N j ∂N k ∂N i ∂N j ∂N k   z3 
 ∂z ∂z ∂z ∂r ∂r ∂r 
Thus, the strain displacement matrix can be expressed as,
 βi βj βk 0 0 0
 Nj 
N Nk
1  i 0 0 0
[ B] =  r r r  (5.7.11)
2A 
0 0 0 γi γj γk 
 
 γ i γj γk βi βj β k 

ri + rj + rk
Where, r = . Thus the stiffness matrix will become
3
=[k ] ∫∫∫ [ B ] [ D ][ B ] d Ω
T

2π
= [k ] ∫ ∫ [ B ] [ D ][ B ] r dθ dA 2π ∫ ∫ [ B ] [ D ][ B ] r dr dz
∫=
T T
Or, (5.7.12)
0
Since, the term [B] is dependent of ‘r’ terms; the term [ B ] [ D ][ B ] cannot be taken out of
T

integration. Yet, a reasonably accurate solution can be obtained by evaluating the [B] (denoted as
[B]) matrix at the centroid.
Hence, [ k ] = 2π r [ B ] [ D ][ B ] ∫ ∫ dr dz
T

Or,
[ k ]  [ B ] [ D ][ B ] 2π rA
T
(5.7.13)

5.7.2 Stiffness Matrix of a Quadrilateral Element

 50

The strain-displacement relation for axisymmetric problem derived earlier (eq.(5.6.5)) can be
rewritten as
 ∂u 
 ∂u   ∂r 
 ∂r   
  1 0 0 0 0   ∂u 
εr 
ε   ∂w  0 0 0 1 0   ∂z 
 ∂z  
{ε } =
= z 
  =  
 
1   ∂w  (5.7.14)
 εθ   u  0 0 0 0 
r   ∂r 
γ rz   r    
 ∂u ∂w  0 1 1 0 0   ∂w 
 +   ∂z 
 ∂z ∂r  u 
 

Applying chain rule of differentiation equation we get,

 ∂u   ∂u 
 ∂r   ∂ξ 
   J*  
*
0   ∂u 
 ∂u   11
J12 0 0

 ∂z   J*21 J*22 0 0 0   ∂η 
    
 ∂w  = 0 0 *
J11 *
J12 0   ∂w  (5.7.15)
 ∂rξ  0 J*21 J*22

0  ∂ 
  
0
 
 ∂w   0 0 0 0 1   ∂w 
 ∂z   ∂η 
 u   
   u 
Hence, the strain components are calculated as
 ∂u 
 ∂ξ 
 
0 0 0 0   J11 0   ∂u 
* *
1 J12 0 0
ε r    * 
ε   0 0 0 1 0   J 21 J*22 0 0 0   ∂η 
 z  
 = 1  0 0   ∂w 
* *
0 J11 J12
ε θ   0 0 0 0 
r   0 0 J*21 J*22

0   ∂ξ 
γ rz    
0 1 1 0 0   0 0 0 0 1   ∂w 
 ∂η 
 
 u 
Or,
 51


 u 

 


  

 
 
 J11

* *
J12 0 0 0  
   u 
  r  
 

  0 
  0    
*
  0 J J*22 
 z   
 
21
   
1  
 w (5.7.16)
  
    
  0 0 0 0
r   
   
 
 rz 
   J* J*22 * * 
0  w 
 
 21 J11 J12
 


  

 


 
 u 

 u u w w 
With the use of interpolation function and nodal displacements,  , , ,  can be expressed
 x h x h 
for a four node quadrilateral element as

 u 
   N1 N 2 N 3 N 4  u1 


   0   

x 
0 0 0

 
  x x x x  u2 

    

 u   N1 N 2 N 3 N 4   u3 
 

 
  0 0 0 0   

  h   h h h h  u4 

      

 (5.7.17)
 w  N1 N 2 N 3 N 4  
 w1 

   0 0 0 0   

 x   x x x x   w2 


 
 
 

  w   N1 N 2 N 3 
N 4   

 
  0 0 0 0 

w3

  h   h h h 
h   
    
 
 w4 


Putting eq. (5.7.17) in eq. (5.7.16) we get,

 N1 N 2 N 3 N 4 
 0 0 0 0   u1 

      

  u2 
 J11
* *   N N 2 N 3 N 4  

0   u3 

 r    J 0 0 0  1
 12 0 0 0 
   0 0 J*21 J*22 0       

  z     u4 
   1   0 N1 N 2 N 3 N 4   
 

  
    0 0 0 0 0 0 0 
 w1 

   r      
   


 rz    J* J* J* J* 0   w 2 

 21 22 11 12  0 N1 N 2 N 3 N 4  
 

 0 0 0  
 w 
3
       
 
w 4 

 N 4  
 
 N1 N2 N3 N4 N1 N2 N3
(5.7.18)
Thus, the strain displacement relationship matrix [B] becomes
 52

 N1 N 2 N 3 N 4 
 0 0 0 0 
     
 
 * * 
0 0  N1 N 2 N 3 N 4 
 J11 J12 0 0 0 0 0 
0
 0 J 21 J*22 0  
*    

 B   
1   0 N1 N 2 N 3 N 4  (5.7.19)
0 0 0 0 0 0 0
      
 * r 
 J 21 J*22 J11

* *
J12 0  N1 N 2 N 3 N 4 
 0 0 0 0
    

 N
 1 N2 N3 N4 N1 N2 N3 N 4 
For a four node quadrilateral element,
(1ξ− )1( η− )          and     
N1 = ⇒
∂N1
=
−
( 1η− ) ∂N1 1( ξ− )
=−
4ξ 4 ∂ η 4 ∂
(1ξ+ 1) ( η− )          and   
N2 = ⇒
∂N 2 ( 1η− )
=
∂N1 1( ξ+ )
=− (5.7.20)
4ξ 4 ∂ η 4 ∂

=N3
(1ξ+ 1) ( η+ )
⇒=
∂N 2 ( 1η+ )
         and   
∂N1 1( ξ+ )
=
4ξ 4 ∂ η 4 ∂
(1ξ− )1( η+ )          and   
N4 = ⇒
∂N 2
=
−
( 1η+ ) ∂N1 1( ξ− )
=
4ξ 4 ∂ η 4 ∂
Thus, the [B] matrix will become
 J11
* *
J12 0 0 0

0 0 J 21 J*22 0
*

 B   1  
0 0 0 0
r 
 *
 J 21 J 22 J11 J12 0
* * *
 
 1 1 1   1   
   0 0 0 0 
 4 4 4 4 
 
 1 1   1   1 
   0 0 0 0 
 4 4 4 4 
 
 1 1 1   1   
 0 0 0 0  
 4 4 4 4 
 
 1 1   1   1 
 0 0 0 0  
 4 4 4 4 
 
 
 11 1  1 1  1   11   11 1  1 1  1   11  
 4 4 4 4 4 4 4 4 

(5.7.21)
The stiffness matrix for the axisymmetric element finally can be found from the following
expression after numerical integration.
+1+1
=[k ] ∫ [ B=
] [ D ][ B ] d Ω ∫ ∫ [ B] [ D][πr.B].2.dξdηJ
T T
(5.7.22)
Ω −1−1
 53

Lecture 8: Finite Element Formulation for 3 Dimensional Elements

5.8.1 Introduction
Solid elements can easily be formulated by the extension of the procedure followed for two
dimensional solid elements. A domain in 3D can be discritized using tetrahedral or hexahedral
elements. For example, the eight node solid brick element is analogous to the four node rectangular
element. Regardless of the possible curvature of edges or number of nodes, the solid element can be
mapped into the space of natural co-ordinates, i.e the ξ = ±1,η = ±1, ζ = ±1 just like a plane element.

Fig. 5.8.1Eight node brick element

For three dimensional cases, each node has three degrees of freedom having u, v, and w as
displacement field in three perpendicular directions (X, Y and Z). In this case, one additional
dimension increases the computational expense manifolds.

5.8.2 Strain Displacement Relation

The strain vector for three dimensional cases can be written in the following form
 54

 ∂u 
 ∂x 
 
 ∂u 
 ∂y 
 
 ∂u 
 ∂z 
 ε x  1 0 0 0 0 0 0 0 0   ∂v 
 ε  0  
 y  0 0 0 1 0 0 0 0   ∂x 
 ε z  0 0 0 0 0 0 0 0 1   ∂v 
 =  
γ xy  0 1 0 1 0 0 0 0 0   ∂y 
γ yz  0 0 0 0 0 1 0 1 0   ∂v 
    
γ zx  0 0 1 0 0 0 1 0 0   ∂z 
 ∂w 
 ∂x 
 
 ∂w 
 ∂y 
 
 ∂w  (5.8.1)
 ∂z 

The following relation exists between the derivative operators in the global co-ordinates and the
natural co-ordinate system by the use of chain rule of partial differentiation.
 ∂   ∂x ∂y ∂z   ∂ 
 ∂ξ   ∂ξ
   ∂ξ ∂ξ   ∂x 
 ∂   ∂x ∂y ∂z   ∂ 
 =   (5.8.2)
 ∂η   ∂η ∂η ∂η   ∂y 
 ∂   ∂x ∂y ∂z   ∂ 

 ∂ζ   ∂ζ ∂ζ ∂ζ   ∂z 
Where the Jacobian Matrix will be
 ∂x ∂y ∂z 
 ∂ξ
 ∂ξ ∂ξ 
[J ] =  ∂x ∂y ∂z 
(5.8.3)
∂η ∂η ∂η 
 ∂x ∂y ∂z 
 
 ∂ζ ∂ζ ∂ζ 
For an isoparamatric element the coordinates at a point inside the element can be expressed by its
nodal coordinate.
n n n
=x i i
=i 1 =i 1
∑=
N x ; y ∑ N=
y and z ∑ N z i i
=i 1
i i (5.8.4)
 55

Substituting the above equations into the Jacobian matrix for an eight node brick element, we get

 ∂N i ∂N i ∂N i 
 ∂ξ xi ∂ξ 
yi zi
 ∂ξ
 ∂N i x ∂N i ∂N i 
8
[J ] = ∑ yi zi (5.8.5)
i =1
 ∂η i ∂η ∂η 
 ∂N ∂N i ∂N i 
 i xi yi zi 
 ∂ζ ∂ζ ∂ζ 
u 
 
The strain displacement relation is given by {ε } = [B ]{d }, where, {d } =  v  .
 w
 
The displacements in the x, y and z direction are u, v, and w respectively. Let consider the inverse of
Jacobian matrix as
 J 11* J 13* 
J 12*
 * * 
[J ]−1=  J 21 *
J 22
J 23  (5.8.6)
 J 31
* ** 
 J 32
J 33 
Thus, the relation between two coordinate systems can be rewritten as
−1
∂  ∂x ∂y ∂z   ∂   ∂ 
 ∂x   ∂ξ ∂ξ ∂ξ   ∂ξ   ∂ξ 
       J11
* *
J12 *
J13  
∂  ∂x ∂y ∂z   ∂   * *  ∂ 
=  =   ∂η   J 21
*
J 22 J 23   (5.8.7)
 ∂y   ∂η ∂η ∂η    *   ∂η 
 J 31
* *
J 32 J 33 
∂  ∂x ∂y ∂z   ∂    ∂ 
   ∂ζ ∂ζ ∂ζ   ∂ζ   
 ∂z       ∂ζ 

Thus, one can write the following relations

∂u ∂u ∂u ∂u 8
 ∂N i ∂N i ∂N i  8
= J11* + J12* + J13* = ∑  J11* + J12* + J13*  ui = ∑ ai ui
∂x
= ∂ξ ∂η ∂ζ i 1 =
 ∂ξ ∂η ∂ζ  i 1

∂u * ∂u * ∂u * ∂u
8
 * ∂N i * ∂N i * ∂N i 
8
= J 21 + J 22 + J 23 = ∑  J 21 + J 22 + J 23  i ∑ bi ui
u =
∂y
= ∂ξ ∂η ∂ζ i 1 =
 ∂ξ ∂η ∂ζ  i 1

∂u * ∂u * ∂u * ∂u
8
 * ∂N i * ∂N i * ∂N i 
8
= J 31 + J 32 + J 33 = ∑  J 31 + J 32 + J 33  i ∑ ci ui
u =
∂z
= ∂ξ ∂η ∂ζ i 1 =
 ∂ξ ∂η ∂ζ  i 1

Similarly,
∂v 8 ∂v 8 ∂v 8 ∂w 8 ∂w 8 ∂w 8
= ∑ = ai vi ; ∑ i i ∂z ∑
= b v ; = c v ; ∑ i i ∂y ∑
=
a w ; b w= and ∑ ci wi
∂x i 1 = ∂y i 1 = ∂x i 1 = ∂z i 1
i i i i
= i 1 = i 1 =

Using above relations, the strain vector can be written as

 56

 ∂u 
 ∂x 
 
 ∂v 
εx   ∂y   ai 0 0
     0 
ε y   ∂w  0 bi
u 
 ε z   ∂z  8 0 ci   i 
{ε } =
0
=   ∂u =  ∑   v  (5.8.8)
γ xy   + ∂v  i =1  bi ai 0 i 
bi   i 
w
γ yz   ∂y ∂x  0 ci
     
γ zx   ∂v + ∂w   ci 0 ai 
 ∂z ∂y 
 
 ∂u + ∂w 
 ∂z ∂x 

Now, the strain displacement relationship matrix [B] can be identified from the above equation by
comparing it to {ε } = [B ]{d }.

5.8.3 Element Stiffness Matrix

The element stiffness matrix can be generated similar to two dimensional case using the following
relations.
+1 +1 +1

[ k ] = ∫ ∫ ∫ [ B ] T [ D ][ B ]dξ dη dζ J (5.8.9)
−1 −1 −1

The size of the constitutive matrix [D] for solid element will be 6 × 6 and is already discussed in
module 1, lectures 3. For eight node brick element, the size of stiffness matrix will become 24 × 24
as number of nodes in one element is 8 and the degrees of freedom at each node is 3. It is well
established that 2 × 2 × 2 Gauss integration points are the optimal sampling points for eight node
isoparametric brick elements.

5.8.4 Element Load Vector

The forces on an element can be generated due to its self weight or externally applied force which
may be concentrated or distributed in nature. The distributed load may be uniform or non-uniform.
All these types of loads are to redistributed to the nodes using finite element formulation.

5.8.4.1 Gravity load

The load vector due to body forces in general is given by
={Q} ∫[N ]
T
{ X }d Ω (5.8.10)
Ω
 57

where {X } is the body forces per unit volume. The nodal load vector at any node i may be expressed
as
={Qi } ∫[N ] i
T
{ X }d Ω (5.8.11)
Ω

In case of gravity load, the force will act in the global negative Z direction. Therefore,
Ni 0 0   0 
   
[ N i ] =  0 N i 0  and { X } =  0  (5.8.12)
 0 0 N i   
− ρ g 
Here, the mass density of the material is ρ and the acceleration due to gravity is g. Thus, eq.(5.8.11)
will become
 0 
= {Qi } ∫  0  d Ω (5.8.13)
Ω 
− Ni ρ g 
For isoparametric element the, the above expression will become
 0 

+1 +1 +1 
{Qi } = ∫ ∫ ∫  0  J dξdηdζ (5.8.14)
−1 −1 −1
− N ρ 
 i 
Using Gauss Quadrature integration rule, the above expression may be evaluated as
 0 
n n n
 
{ Qi } = ∑∑∑ wi w j wk J ( ξi ,η j ,ζ k )  0  (5.8.15)
=i 1 =j 1 =
k 1 − N ρ g 
 i ( ξi ,η j ,ζ k )
Where, n is the number of nodes in an element. For eight node linear brick element the value of n
will be 8 and the integration order suggested is 2×2×2. Similarly, for twenty node quadratic brick
element, the value of n will be 20 and the integration order suggested is 3×3×3.

5.8.4.2 Surface pressure

Let assume a uniform surface pressure of intensity q is acting normal to the element face. The load
vector due to surface pressure is given by
{Q} = ∫ ∫ [ N s ]T { p}dA (5.8.16)

The nodal load at any node i may be expressed as

{Qi } = ∫ ∫ [ N i s ]T { p}dA (5.8.17)

In case of surface load, the value of [ N is ] in the above equation will become
 58

 N is 0 0 
 
[ N is ] =  0 N i
s
0  (5.8.18)
0
 0 N is 

Here N is is the interpolation function for the node i. For example, the value of N is can be obtained
by substituting ξ =1 in N is for face 1. Thus, the surface pressure is expressed as,
 ql3 
 
{ p} = qm3  (5.8.19)
 qn 
 3
Where, l3 , m3 , n3 are the direction cosines. Thus, eq.(5.8.17) can be expressed using eq.(5.8.19) in the
following form.
 N is ql3 
{Qi } = ∫−1 ∫−1  N is qm3 dA
+1 +1
(5.8.20)
 N s qn 
 i 3
The value of dA is can be evaluated considering the cross product of vectors along the natural
coordinates parallel to the loaded faces of the element. Thus,
dA = e2 × e3 dηdζ (5.8.21)

5.8.5 Stress Computation

Using the relation of {F} = [K]{d}, the unknown nodal displacement vector {d} are calculated in
global coordinate system. Once the nodal displacements are obtained, the strain components as each
node can be computed using strain-displacement relations for each element. Similarly element stress
can be calculated using stress-strain relation. These stresses at Gauss points are extrapolated to the
corresponding nodes of the element to find the nodal stresses. In general, for three dimensional state
of stress there are at least three planes, called principal planes. The corresponding stress vector is
perpendicular to the plane and where there are no normal shear stresses. These three stresses which
are normal to these principal planes are called principal stresses. The principal stresses σ1, σ2, and σ3
are computed from the roots of the cubic equation represented by the determinant of the flowing.

σ x −σ τ xy τ xz
τ xy σ y −σ τ yz =
0 (5.8.22)
τ xz τ yz σ z −σ
The characteristic equation has three real roots σi, due to the symmetry of the stress tensor. The
principal stresses are arranged so that σ1 > σ2 > σ3. The maximum shear stress can be computed
from the following relations.
 59

 σ1 − σ 2 σ2 −σ3 σ 3 − σ1 
τ max = largest of  , and  (5.8.23)
 2 2 2 
These three shear stress components will occur on planes oriented at 450 from the principal planes.
The distortion energy theory suggests that the total strain energy can be divided into two
components. They are (i) volumetric strain energy and (ii) distortion or shear strain energy. It is
anticipated that yield develops if the distortion component exceeds that at the yield point for a simple
tensile test. From the concept of distortion energy theory, the equivalent stress which is historically
known as Von Mises stress are defined as
 (σ 1 − σ 2 ) 2 + (σ 2 − σ 3 ) 2 + (σ 3 − σ 1 ) 2 
σe =   (5.8.24)
 2 
 
The Von Mises stresses offer a measure of the shear or distortional stress in the material. In general,
this type of stress tends to cause yielding in metals.
 60

Worked out Examples

Example 5.1 Calculation of nodal loads on a triangular element

A CST element as shown in Fig. 5.I gets axial loading of (Fx1) 10 kN/m in X direction and (Fy1) 20
kN/m in Y direction. Compute the nodal loads in the element.

Fig. 5.I Distributed loading on a triangular element

From the above figure, the length of sides 1, 2 and 3 are calculated and will be 10, 8 and 6 cm
respectively. First, let consider side 1:
l1

{F }1 = ∫ [ N ] {FΓ } ds
T

Or,
0 0 0 0 0
L 0  L 0  L 
 2  2  2
l1
L 0   Fx1  l1
 L3 0  10  l1
 L3 
{F }=1 ∫  3    ds= ∫0  0    ds= 10 × ∫  ds
0 
0 0   Fy1   0  0  0
0
0 L2  0 L2  0
     
 0 L3   0 L3  0
p p !q !
Putting, ∫ L
1 L2 q ds =
( p + q + 1)!
l , we will get,
 61

0  0   0 
1  1  0.5
     
l1 1  0.1 1  0.5
{F }1 =10 × ×   =10 × ×   = 
2 0  2 0   0 
0  0   0 
     
0  0   0 
Similarly for side 2,
 L1 0  L1 0 0   0 
0 0  0 0  0   0 
     
l2
L 0   Fx 2  l2
 L3 0  0  l2  0   0 
{F }2 = ∫ 3    ds = ∫0  0    ds = 20 × ×   =   kN
0 
0 L1   Fy 2   L1  20  2 1  0.8
0 0 0 0 0   0 
       
 0 L3   0 L3  1  0.8
Since no force is acting on side 3,
0 
0 
 
0 
{F }3 =  
0 
0 
 
0 
Hence, the nodal load vector in all the nodes in x and y directions will become,
0 
0.5
 
0.5
{F } =   kN.
0.8
0 
 
0.8
 1

Lecture 1: Introduction to Plate Bending Problems

6.1.1 Introduction
A plate is a planer structure with a very small thickness in comparison to the planer dimensions. The
forces applied on a plate are perpendicular to the plane of the plate. Therefore, plate resists the applied
load by means of bending in two directions and twisting moment. A plate theory takes advantage of this
disparity in length scale to reduce the full three-dimensional solid mechanics problem to a two-
dimensional problem. The aim of plate theory is to calculate the deformation and stresses in a plate
subjected to loads. A flat plate, like a straight beam carries lateral load by bending. The analyses of
plates are categorized into two types based on thickness to breadth ratio: thick plate and thin plate
analysis. If the thickness to width ratio of the plate is less than 0.1 and the maximum deflection is less
than one tenth of thickness, then the plate is classified as thin plate. The well known as Kirchhoff plate
theory is used for the analysis of such thin plates. On the other hand, Mindlin plate theory is used for
thick plate where the effect of shear deformation is included.

6.1.2 Notations and Sign Conventions

Let consider plates to be placed in XY plane. Representation of plate surface slopes W,x, W,y by right
hand rule produces arrows that point in negative Y and positive X directions respectively. Both surface
slopes and rotations are required for plate elements. Signs and subscripts of rotations and slopes are
reconciled by replacing θx by Ψy and θy by -Ψx

Fig. 6.1.1 Notations and sign conventions

6.1.3 Thin Plate Theory

 2

Classical thin plate theory is based upon assumptions initiated for beams by Bernoulli but first applied to
plates and shells by Love and Kirchhoff. This theory is known as Kirchhoff’s plate theory. Basically,
three assumptions are used to reduce the equations of three dimensional theory of elasticity to two
dimensions.
1. The line normal to the neutral axis before bending remains straight after bending.
2. The normal stress in thickness direction is neglected. i.e., σ z = 0 . This assumption converts
the 3D problem into a 2D problem.
3. The transverse shearing strains are assumed to be zero. i.e., shear strains γxz and γyz will be
zero. Thus, thickness of the plate does not change during bending.

The above assumptions are graphically shown in Fig. 6.1.2.

Fig. 6.1.2 Kirchhoff plate after bending

6.1.3.1 Basic relationships

𝑡
Let, a plate of thickness t has mid-surface at a distance 2 from each lateral surface. For the analysis
purpose, X-Y plane is located in the plate mid-surface, therefore z=0 identifies the mid-surface. Let u, v,
w be the displacements at any point (x, y, z).
 3

Fig. 6.1.3 Thin plate element

Then the variation of u and v across the thickness can be expressed in terms of displacement w as
𝜕𝑤 𝜕𝑤
u = -z 𝜕𝑥
v = -z 𝜕𝑦
(6.1.1)
Where, w is the deflection of the middle plane of the plate in the z direction. Further the relationship
between, the strain and deflection is given by,
𝜕𝑢 𝜕2 𝑤
εx = 𝜕𝑥 = -z 𝜕𝑥 2
= z χx
𝜕𝑣 𝜕2 𝑤
εy = 𝜕𝑦
= -z 𝜕𝑦 2
= z χy (6.1.2)
𝜕𝑢 𝜕𝑣 𝜕2 𝑤
γxy = 𝜕𝑦
+ 𝜕𝑥 = -2z 𝜕𝑥𝜕𝑦 = z χxy
where,
ε corresponds to direct strain
γ corresponds to shear strain
χ corresponds to curvature along respective directions.
Or in matrix form, the above expression can written as
𝜕2
𝜀𝑥 ⎡ 𝜕𝑥 2 ⎤
⎢ 𝜕2 ⎥
� 𝜀𝑦 � = −𝑧 ⎢ 𝜕𝑦 2 ⎥ 𝑤 (6.1.3)
γxy
⎢ 𝜕2 ⎥
⎣𝜕𝑥𝜕𝑦⎦
Or,
ε = −z∆𝑤 (6.1.4)
Where, ε is the vector of in-plane strains, and ∆ is the differential operator matrix.

6.1.3.2 Constitutive equations

 4

From Hooke’s law,

σ = [ D]ε (6.1.5)
Where,
 
1 υ 0 
E  
[ D] = 2 
υ 1 0  (6.1.6)
(1 − υ )  1−υ 
0 0 
 2 
Here, [D] is equal to the value defined for 2D solids in plane stress condition (i.e., σ z = 0 ).

6.1.3.3 Calculation of moments and shear forces

Let consider a plate element of 𝑑𝑥 × 𝑑𝑦 and with thickness t. The plate is subjected to external
uniformly distributed load p. For a thin plate, body force of the plate can be converted to an equivalent
load and therefore, consideration of separate body force is not necessary. By putting eq. (6.1.4) in eq.
(6.1.5),
− z [ D ] ∆w
σ= (6.1.7)
It is observed from the above relation that the normal stresses are varying linearly along thickness of the
plate (Fig. 6.1.4(a)). Hence the moments (Fig. 6.1.4(b)) on the cross section can be calculated by
integration.
M X  t /2
   t/2  t3
M = M y  =∫ σ zdt =−  ∫ z 2 dt  [ D ] ∆w =− [ D ] ∆w (6.1.8)
 M  −t / 2  −t / 2  12
 xy 
 5

(a) Stresses in plate

(b) Forces and moments in plate

Fig. 6.1.4 Forces on thin plate

 6

On expansion of eq. (6.1.8) one can find the following expressions.

Et 3  ∂2w ∂2w 
MX =
−  2 + υ =DP ( χ x + υχ y )
12 (1 − υ 2 )  ∂x ∂y 2 

Et 3  ∂2w ∂2w 
My =
−  2 +υ 2  =DP ( χ y + υχ x ) (6.1.9)
12 (1 − υ 2 )  ∂y ∂x 

Et 3 ∂2w D (1 − υ )
M xy = M yx = = − P χ xy
12 (1 + υ ) ∂x∂y 2
Where, DP is known as flexural rigidity of the plate and is given by,
Et 3
DP = (6.1.10)
12 (1 − υ 2 )
Let consider the bending moments vary along the length and breadth of the plate as a function of x and
∂M x
y. Thus, if Mx acts on one side of the element, M='
Mx + dx acts on the opposite side. Considering
∂x
x

equilibrium of the plate element, the equations for forces can be obtained as
∂Qx ∂Qy
+ +p= 0 (6.1.11)
∂x ∂y
∂M x ∂M xy
+ =
Qx (6.1.12)
∂x ∂y
∂M xy ∂M y
+ =
Qy (6.1.13)
∂x ∂y
Using eq.6.1.9 in eqs.6.1.12 & 6.1.13, the following relations will be obtained.
∂  ∂2w ∂2w 
Qx =
− DP  2 + 2  (6.1.14)
∂x  ∂x ∂y 
∂  ∂2w ∂2w 
Qy = − DP  2 + 2  (6.1.15)
∂y  ∂x ∂y 
Using eqs. (6.1.14) and (6.1.15) in eq. (6.1.11) following relations will be obtained.
∂4w ∂4w ∂4w p
+ 2 + =
− (6.1.16)
∂x 4
∂x ∂y
2 2
∂y 4
DP

6.1.4 Thick Plate Theory

Although Kirchhoff hypothesis provides comparatively simple analytical solutions for most of the cases,
it also suffers from some limitations. For example, Kirchhoff plate element cannot rotate independently
of the position of the mid-surface. As a result, problems occur at boundaries, where the undefined
transverse shear stresses are necessary especially for thick plates. Also, the Kirchhoff theory is only
 7

applicable for analysis of plates with smaller deformations, as higher order terms of strain-displacement
relationship cannot be neglected for large deformations. Moreover, as plate deflects its transverse
stiffness changes. Hence only for small deformations the transverse stiffness can be assumed to be
constant.
Contrary, Reissner–Mindlin plate theory (Fig. 6.1.5) is applied for analysis of thick plates, where
the shear deformations are considered, rotation and lateral deflections are decoupled. It does not require
the cross-sections to be perpendicular to the axial forces after deformation. It basically depends on
following assumptions,
1. The deflections of the plate are small.
2. Normal to the plate mid-surface before deformation remains straight but is not necessarily
normal to it after deformation.
3. Stresses normal to the mid-surface are negligible.

Fig. 6.1.5 Bending of thick plate

Thus, according to Mindlin plate theory, the deformation parallel to the undeformed mid surface, u and
v, at a distance z from the centroidal axis are expressed by,
 8

𝑢 = 𝑧𝜃𝑦 (6.1.17)
v = −zθ𝑥 (6.1.18)
Where θx and θy are the rotations of the line normal to the neutral axis of the plate with respect to the x
and y axes respectively before deformation. The curvatures are expressed by
∂θ
χx = y (6.1.19)
∂x
∂θ
χy = − x (6.1.20)
∂y
Similarly the twist for the plate is given by,
 ∂θ ∂θ 
=χ xy  y − x  (6.1.21)
 ∂y ∂x 
Using eqs.6.1.9-6.1.10, the bending stresses for the plate is given by
 
 Mx  1 υ 0   χx 
  Et 3   
My  = 2 
υ 1 0   χy  (6.1.22)
 M  12 (1 − υ )  1 − υ   χ xy 
 xy  0 0 
 2 
Or
{M } = [ D ]{ χ } (6.1.23)

Further, the transverse shear strains are determined as

∂w
γ xz= θ y + (6.1.24)
∂x
∂w
γ yz =−θ x + (6.1.25)
∂y
The shear strain energy can be expressed as
1 2
Us = 2 𝛼𝐺𝐴 ∬𝐴 �(𝛾𝑥 )2 + �𝛾𝑦 � � 𝑑𝑥 𝑑𝑦
1 𝜕𝑤 2 𝜕𝑤 2
= 2 𝛼𝐺𝐴 ∬𝐴 ��𝜃𝑦 + 𝜕𝑥
� + �−𝜃𝑥 + 𝜕𝑦 � � 𝑑𝑥 𝑑𝑦 (6.1.26)
𝐸
Where, G = 2(1+𝜇). The shear stresses are
𝜏𝑥𝑧 𝐸 1 0 𝛾𝑥
�𝜏 � = 2(1+𝜇) � �� � (6.1.27)
𝑦𝑧 0 1 𝛾𝑦
Hence the resultant shear stress is given by,
𝑄𝑥 𝐸𝑡𝛼 1 0 𝛾𝑥
�𝑄 � = 2(1+𝜇) � �� � (6.1.28)
𝑦 0 1 𝛾𝑦
Or,
{𝑄} = [𝐷𝑠 ]{𝛾} (6.1.29)
 9

Here "𝛼" is the numerical correction factor used to characterize the restraint of cross section against
warping. If there is no warping i.e., the section is having complete restraint against warping then α = 1
and if it is having no restraint against warping then α = 2/3. The value of α is usually taken to be π2/12 or
5/6. Now, the stress resultant can be combined as follows.
{M } [ D ] 0  { χ }
 =
[ Ds ] {γ } 
(6.1.30)
 {Q}   0
Or,
𝑀𝑥 1 𝜇 0 0 0 χ
⎧ 𝑀 ⎫ ⎡ 𝐸𝑡 3 ⎤ ⎧ χ𝑥 ⎫
⎪ 𝑦 ⎪ ⎢12(1−𝜇2 ) �𝜇 1 0 � 0 0 ⎥⎪ 𝑦 ⎪
𝑀𝑥𝑦 = ⎢ 0 0
1−𝜇
0 0 χ (6.1.31)
⎥ 𝑥𝑦
⎨𝑄 ⎬ 2 ⎨ 𝛾𝑥 ⎬
⎪ 𝑥⎪ ⎢ 0 0 0 𝐸𝑡
�
𝛼 0 ⎥⎪ ⎪
� 𝛾
⎩ 𝑄𝑦 ⎭ ⎣ 0 0 0 2(1+𝜇) 0 𝛼 ⎦ ⎩ 𝑦 ⎭
Or,
𝜕𝜃𝑦
⎧ 𝜕𝑥 ⎫
𝑀𝑥 1 𝜇 0 0 0 ⎪ 𝜕𝜃𝑥 ⎪
⎧ 𝑀 ⎫ ⎡ 𝐸𝑡 3 𝜇 1 0 ⎤ ⎪ − 𝜕𝑦 ⎪
⎪ 𝑦 ⎪ ⎢12(1−𝜇2 ) � � 0 0 ⎪ ⎪
1−𝜇 ⎥ 𝜕𝜃𝑦 𝜕𝜃𝑥
𝑀𝑥𝑦 = ⎢ 0 0 2 0 0 ⎥ 𝜕𝑦 − 𝜕𝑥 (6.1.32)
⎨𝑄 ⎬ ⎨ ⎬
⎪ 𝑥⎪ ⎢ 0 0 0 𝐸𝑡
�
𝛼 0 ⎥ ⎪ 𝜃 + 𝜕𝑤 ⎪
�
⎩ 𝑄𝑦 ⎭ ⎣ 0 0 0 2(1+𝜇) 0 𝛼 ⎦ ⎪
𝑦 𝜕𝑥 ⎪
⎪ 𝜕𝑤⎪
⎩−𝜃𝑥 + 𝜕𝑦 ⎭
The above relation may be compared with usual stress-strain relation. Thus, the stress resultants and
their corresponding curvature and shear deformations may be considered analogous to stresses and
strains.

6.1.5 Boundary Conditions

For different boundaries of the plate (Fig. 6.1.6), suitable conditions are to be incorporated in plate
equation for solving the governing differential equations. For example, following conditions need to
satisfy along y direction of the plate for various boundaries.
 10

Fig. 6.1.6 Plate with four boundaries

1. Simply support edge (Along y direction)

w ( x, =
y ) 0, M
= x [ x const & 0 ≤ y ≤ b]
0 =
2. Clamped Edge (Along y direction)
∂w
w (=x, y ) 0, (=
x, y ) 0 = [ x const & 0 ≤ y ≤ b]
∂x
3. Free Edge (Along y direction)
∂M xy
M= 0, Qx + = 0, [=x const & 0 ≤ y ≤ b ]
∂x
x

Similar to the above, the boundary conditions along x direction can also be obtained. Once the
displacements w(x,y) of the plate at various positions are found, the strains, stresses and moments
developed in the plate can be determined by using corresponding equations.
 11

Lecture 2: Finite Element Analysis of Thin Plate

6.2.1 Triangular Plate Bending Element

A simplest possible triangular bending element has three corner nodes and three degrees of freedom per
nodes �𝑤, 𝜃𝑥 , 𝜃𝑦 � as shown in Fig. 6.2.1.

Fig. 6.2.1 Triangular plate bending element

As nine displacement degrees of freedom present in the element, we need a polynomial with nine
independent terms for defining, w(x,y) . The displacement function is obtained from Pascal’s triangle by
choosing terms from lower order polynomials and gradually moving towards next higher order and so
on.
1

x y

x2 xy y2

x3 x2y xy2 y3

x4 x3y x2y2 xy3 y4

Thus, considering Pascal triangle, and in order to maintain geometric isotropy, we may consider the
displacement model in terms of the complete cubic polynomial as,
w ( x, y ) =α 0 + α1 x + α 2 y + α 3 x 2 + α 4 xy + α 5 y 2 + α 6 x 3 + α 7 ( x 2 y + xy 2 ) + α 8 y 3 (6.2.1)
 12

Corresponding values for �𝜃𝑥 , 𝜃𝑦 �are,

∂w
θx = =α 2 + α 4 x + 2α 5 y + α 7 ( x 2 + 2 xy ) + 3α 8 y 2 (6.2.2)
∂y
∂w
θy =− = −α1 − 2α 3 x − α 4 y − 3α 6 x 2 − α 7 ( 2 xy + y 2 ) (6.2.3)
∂x
In the matrix form,
α 0 
α 
 1
α 2 
w
1 x y x 2

xy y 2 x3 ( x y + xy )
2 2
y 3  α 
 3 (6.2.4)
   
θ x  0 0 1

0 x 2y 0 ( x + 2 xy )
2
3 y 2  α 4 
 
θ 
 y 0 −1 0 −2 x − y 0 −3 x 2
 − ( 2 xy + y )
2
0  α 5 
α 6 
α 
 7
α 8 
Putting the nodal displacements and rotations for the triangular plate element as shown in Fig. 6.2.1 in
the above equation, one can express following relations.
 w1  1 0 0 0 0 0 0 0 0 
  0 0 1   α 0 
θ x1  
0 0 0 0 0 0  α1 
θ y1  0 −1 0 0 0 0 0 0 0  
  1 0 y2   α 2 
0 0 y22
0 0 y23   
 w2   α3
  0 0 1  0 0 2 y2 0 0 3 y22   
θ x 2  =   α 4  (6.2.5)
θ  0 −1 0 0 − y2 0 0 − y22 0  
α5
 y 2  1 x y x32 x3 y3 y32 x33 ( x32 y3 + x3 y32 ) y32  α 
 w3   3 3
6
  0 0 1 2
3 y3  α 7 
θ x 3   0 x3 −2 y3 0 ( 2 x3 y3 + x32 )
θ y 3    
  0 −1 0 −2 x3 − y3 0 3 x32 ( − y32 + 2 x3 y3 ) 0  α 8 
Or,
{α } = [ Φ ] {di }
−1
(6.2.6)
Further, the curvature of the plate element can be written as
𝜕2 𝑤
𝜒𝑥 ⎧− 𝜕𝑥 2 ⎫
⎪ 𝜕2 𝑤⎪
� 𝜒𝑦 � = − 𝜕𝑦 2 (6.2.7)
𝜒𝑥𝑦 ⎨ 2 ⎬
⎪ 2𝜕 𝑤 ⎪
⎩ 𝜕𝑥𝜕𝑦 ⎭
Again, from eq. (6.2.1) the following equations can be obtained.
 13

∂2w
=2α 3 + 6α 6 x + 2α 7 y
∂x 2
∂2w
=2α 5 + 2α 7 x + 6α 8 y
∂y 2 (6.2.8)
∂ w
2
α 4 2α 7 ( x + y )
=+
∂x∂y

The above equation is expressed in matrix form as

α 0 
α 
 1
α 2 
 
 χx  0 0 0 −2 0 0 −6 x −2 y 0  α 3 
    
 χy 
= 0 0 0 0 0 −2 0 −2 x −6 y  α 4  (6.2.9)
χ  0 0 0 0 2 0 0 4( x + y) 0  α 5 
 xy   
α 6 
α 
 7
α 8 
Or,
{χ } = [ B ]{α }
Thus,
χ } [ B ][ Φ ] {d }
{=
−1

(6.2.10)
Further for isotropic material,
 
 Mx  1 υ 0   χx 
  Et 3   
My  = 2 
υ 1 0   χy  (6.2.11)
 M  12 (1 − υ )  1 − υ   χ xy 
 xy  0 0 
 2 
Or
{M } [=
D ]{ χ } [ D ][ B ][ Φ ] {d }
−1
= (6.2.12)
Now the strain energy stored due to bending is
1 𝑎 𝑏 1 𝑎 𝑏
𝑈 = 2 ∫0 ∫0 [𝜒]𝑇 {𝑀}𝑑𝑥𝑑𝑦 = 2 ∫0 ∫0 {𝑑}𝑇 [𝜙 −1 ]𝑇 [𝐵]𝑇 [𝐷][𝐵][𝜙 −1 ]{𝑑}𝑑𝑥𝑑𝑦 (6.2.13)
Hence the force vector is written as
𝜕𝑈 𝑎 𝑏
{𝐹} =
𝜕{𝑑}
= [𝜙 −1 ]𝑇 ∫0 ∫0 [𝐵]𝑇 [𝐷][𝐵]𝑑𝑥𝑑𝑦 [𝜙 −1 ]{𝑑} = [𝑘]{𝑑} (6.2.14)
Thus, [k] is the stiffness matrix of the plate element and is given by
 14

𝑎 𝑏
[𝑘] = [𝜙 −1 ]𝑇 ∫0 ∫0 [𝐵]𝑇 [𝐷][𝐵]𝑑𝑥𝑑𝑦 [𝜙 −1 ] (6.2.15)
For a triangular plate element with orientation as shown in Fig. 6.2.2, the stiffness matrix defined in
local coordinate system [k] can be transformed into global coordinate system.
[𝐾] = [𝑇]𝑇 [𝑘 ] [𝑇] (6.2.16)
Where, [K] is the elemental stiffness matrix in global coordinate system and [T] is the transformation
matrix given by
1 0 0 0 0 0 0 0 0 
0 l m 0 0 0 0 0 0 
 x x 
0 l y my 0 0 0 0 0 0 
 
0 0 0 1 0 0 0 0 0 
[T ] = 0 0 0 0 lx mx 0 0 0  (6.2.17)
0 0 0 0 l y my 0 0 0 
0 0 0 0 0 0 1 0 0 
 
 0 0 0 0 0 0 0 l x mx 
0 0 0 0 0 0 0 l m 
 y y

Here, (lx , mx) and (ly, my) are the direction cosines for the lines OX and OY respectively as shown in Fig.
6.2.2.

Fig. 6.2.2 Local and global coordinate system

6.2.2 Rectangular Plate Bending Element

A rectangular plate bending element is shown in Fig. 6.2.3. It has four corner nodes with three degrees
of freedom �𝑤, 𝜃𝑥 , 𝜃𝑦 � at each node. Hence, a polynomial with 12 independent terms for defining w(x,y)
is necessary.
 15

Fig 6.2.3 Rectangular plate bending element

Considering Pascal triangle, and in order to maintain geometric isotropy the following displacement
function is chosen for finite element formulation.
w(x, y)   0  1x   2 y   3 x 2   4 xy   5 y 2   6 x 3   7 x 2 y
(6.2.18)
  8 xy 2   9 y3  10 x 3 y  11xy3
Hence Corresponding values for �𝜃𝑥 , 𝜃𝑦 �are,
w
x    2   4 x  2 5 y   7 x 2  2 8 xy  3 9 y 2  10 x 3  311xy 2 (6.2.19)
y
w
y    1  2 3 x  4 y  3 6 x 2  2 7 xy  8 y 2  310 x 2 y 11y3 (6.2.20)
x
The above can be expressed in matrix form
3 
  0 
 w  1 x y x 2 2 3 2 2 3 3  
  
xy y x x y xy y x y xy    

   0 0 1 0 x 2y 0 x2 2xy 3y 2 x3 3xy 2   1 
 x      
 y  0 1 0 2x y 0 3x 2 2xy y 2 0 3x 2 y y3  
11
(6.2.21)
In a similar procedure to three node plate bending element the values of {α} can be found from the
following relatios.
 16

 w1  1 0 0 0 0 0 0 0 0 0 0 0   0 
     
  x1  0 0 1 0 0 0 0 0 0 0 0 0  1 
    
  y1  0 1 0 0 0 0 0 0 0 0 0 0    2 
 
  
w 2  1 0 b 0 0 b2 0 0 0 b3 0 0    3 
  0  
0 1 0 0 2b 0 0 0 3b 2 0 0    4 
 x 2    
 y2  0 1 0 0 b 0 0 0 b 2 0 0 b3    5 
    
 w 3  1 a b a2 ab b 2 a3 a 2 b ab 2 b3 a 3b ab3    6 
    
 x3  0 0 1 0 a 2b 0 a2 2ab 3b 2 a3 3ab3    7 
  0  
 y3   1 0 2a b 0 3a 2 2ab b 2 0 3a 2 b b3    8 
 
w 4  1 a 0 a2 0 0 a3 0 0 0 0 0    9 
    
 x 4  0 0 1 0 a 0 0 a2 0 0 a3 0  10 
    
 y4  0 1 0 2a 0 0 3a 2 0 0 0 0 0  11 
(6.2.22)
Thus,
{α } = [Φ ] {d }
−1
(6.2.23)
Further,
𝜕2 𝑤
𝜒𝑥 ⎧− 𝜕𝑥 2 ⎫
⎪ 𝜕2 𝑤⎪
� 𝜒𝑦 � = − 𝜕𝑦 2 (6.2.24)
𝜒𝑥𝑦 ⎨ 2 ⎬
⎪ 2𝜕 𝑤 ⎪
⎩ 𝜕𝑥𝜕𝑦 ⎭

Where,
2w
 2 3  6 6 x  2 7 y  610 xy
x 2
2w (6.2.25)
 2 5  2 8 x  6 9 y  611xy
y 2
2w
  4  2 7 x  2 8 y  310 x 2  311y 2
xy
Thus, putting values of eq. (6.2.25) in eq. (6.2.24), the following relation is obtained.
0 
 
  0 0 0 2 0 0 6x 2y  6xy   

 x 
  0 0 0  
 



 y    0 0 0 0 0 2 2x 6y 6xy   
1 (6.2.26)
 
0 0 0    

 
 0 0 0 0 2 0 4y 6y 2  
 


2

 xy    0 4x 0 6x
  
 11
Or,
   B  -1  d  (6.2.27)
 17

Further in a similar method to triangular plate bending element we can estimate the stiffness matrix for
rectangular plate bending element as,
 
 Mx  1 υ 0   χx 
  Et 3   
My  = 2 
υ 1 0   χy  (6.2.28)
 M  12 (1 − υ )  1 − υ   χ xy 
 xy  0 0 
 2 
Or
{M } [=
D ]{ χ } [ D ][ B ][ Φ ] {d }
−1
= (6.2.29)
The bending strain energy stored is
1 𝑎 𝑏 1 𝑎 𝑏
𝑈 = 2 ∫0 ∫0 [𝜒]𝑇 {𝑀}𝑑𝑥𝑑𝑦 = 2 ∫0 ∫0 {𝑑}𝑇 [𝜙 −1 ]𝑇 [𝐵]𝑇 [𝐷][𝐵][𝜙 −1 ]{𝑑}𝑑𝑥𝑑𝑦 (6.2.30)
Hence, the force vector will become
𝜕𝑈 𝑎 𝑏
{𝐹} =
𝜕{𝑑}
= [𝜙 −1 ]𝑇 ∫0 ∫0 [𝐵]𝑇 [𝐷][𝐵]𝑑𝑥𝑑𝑦 [𝜙 −1 ]{𝑑} = [𝑘]{𝑑} (6.2.31)
Where, [k] is the stiffness matrix given by
𝑎 𝑏
[𝑘] = [𝜙 −1 ]𝑇 ∫0 ∫0 [𝐵]𝑇 [𝐷][𝐵]𝑑𝑥𝑑𝑦 [𝜙 −1 ] (6.2.30)
The stiffness matrix can be evaluated from the above expression. However, the stiffness matrix also can
be formulated in terms of natural coordinate system using interpolation functions. In such case, the
numerical integration needs to be carried out using Gauss Quadrature rule. Thus, after finding nodal
displacement, the stresses will be obtained at the Gauss points which need to extrapolate to their
corresponding nodes of the elements. By the use of stress smoothening technique, the various nodal
stresses in the plate structure can be determined.
 18

Lecture 3: Finite Element Analysis of Thick Plate

6.3.1 Introduction
Finite element formulation of the thick plate will be similar to that of thin plate. The difference will be
the additional inclusion of energy due to shear deformation. Therefore, the moment curvature relation
derived in first lecture of this module for thick plate theory will be the basis of finite element
formulation. The relation is rewritten in the below for easy reference to follow the finite element
implementation.
𝑀𝑥 1 𝜇 0 0 0 χ𝑥
⎧ 𝑀 ⎫ ⎡ 𝐸𝑡 3 ⎤ ⎧ χ ⎫
⎪ 𝑦 ⎪ ⎢12(1−𝜇2 ) �𝜇 1 0 � 0 0 ⎥ ⎪ 𝑦⎪
𝑀𝑥𝑦 = ⎢ 0 0 2
1−𝜇
0 0 χ (6.3.1)
⎥ 𝑥𝑦
⎨𝑄 ⎬ ⎨ 𝛾𝑥 ⎬
⎪ 𝑥⎪ ⎢ 0 0 0 𝐸𝑡
�
𝛼 0 ⎥⎪ ⎪
� 𝛾
𝑄
⎩ 𝑦⎭ ⎣ 0 0 0 2(1+𝜇) 0 𝛼 ⎦ ⎩ 𝑦 ⎭
Or,
{M } [ D ] [ 0]  { χ }
 =   (6.3.2)
 {Q}   [ 0] [ Ds ] {γ } 
The above relation is comparable to stress-strain relations.
{σ }P = [C ]P {ε }P (6.3.3)
Where,
χ𝑥
⎧χ ⎫
⎪ 𝑦⎪
{𝜀}𝑃 = χ𝑥𝑦 = [𝐵]{𝑑𝑖 } (6.3.4)
⎨ 𝛾𝑥 ⎬
⎪ ⎪
⎩ 𝛾𝑦 ⎭
Where [B] is the strain displacement matrix and {di} is the nodal displacement vector. Thus, combining
eqs. (6.3.3) and (6.3.4), the following expression is obtained.
{σ }P = [C ]P [ B ]{di } (6.3.5)

6.3.2 Strain Displacement Relation

Let consider a four node isoparametric element for the thick plate bending analysis purpose. The
variation of displacement w and rotations, θx and θy within the element are expressed in the form of
nodal values.
4
w = ∑ N i wi
i =1
4
θ x = ∑ N iθ xi (6.3.6)
i =1
4
θ y = ∑ N iθ yi
i =1
 19

Where, the shape function for the four node element is expressed as,
1
N i =(1 + ξ ξi )(1 + ηηi ) (6.3.7)
4
Here, ξi and ηi are the local coordinates ξ and η of the ith node. Using eq. (6.3.6), eq. (6.3.4) can be
rewritten as
4
∂N i
χ x = ∑ θ yi
i =1 ∂x
4
∂N i
χ=
y ∑
i =1
−θ xi
∂y
∂N i 4 4
∂N i
=χ xy yi
=i 1 =
∑θ
− ∑ θ xi
∂y i 1 ∂x
(6.3.8)

∂N4 4
=γx i
i

∂x
=i 1 =i 1
∑w
+ ∑ θ yi N i
4
∂N 4
=γ y ∑ wi i + ∑ θ xi N i
=i 1 = ∂y i 1
The above can be expressed in matrix form as follows:

 ∂N i 
 0 0
∂x 
 
∂N i
 χx   0 − 0 
χ   ∂y 
 y   ∂N ∂N i 

{ε }P =
= χ xy   0 − i {d } (6.3.9)
 ∂x ∂y  i
γ   
 x   ∂N i
 γ y   0 Ni 
∂x 
 
 ∂N i − Ni 0 
 ∂y 
Here, for a four node quadrilateral element, the nodal displacement vector {di} will become
𝑇
{𝑑𝑖 } = �𝑤1 , 𝜃𝑥1 , 𝜃𝑦1 , 𝑤2 , 𝜃𝑥2 , 𝜃𝑦2 , 𝑤3 , 𝜃𝑥3 , 𝜃𝑦3 , 𝑤4 , 𝜃𝑥4 , 𝜃𝑦4 � (6.3.10)
Thus, the strain-displacement relationship matrix will be
 20

 ∂N1 ∂N 2 ∂N 3 ∂N 4 
 0 0
∂x
0 0
∂x
0 0
∂x
0 0
∂x 

 0 ∂N1 ∂N 2 ∂N 3 ∂N 4
 − 0 0 − 0 0 − 0 0 − 0 
∂y ∂y ∂y ∂y
 
 0 ∂N ∂N1 ∂N ∂N 2 ∂N ∂N 3 ∂N ∂N 4 
[ B] =  − 1
∂x ∂y
0 − 2
∂x ∂y
0 − 3
∂x ∂y
0 − 4
∂x ∂y 
 
 ∂N1 ∂N 2 ∂N 3 ∂N 4 
 ∂x 0 N1 0 N2 0 N3 0 N4 
∂x ∂x ∂x
 
 ∂N1 − N1 0
∂N 2
− N2 0
∂N 3
− N3 0
∂N 4
− N4 0 
 ∂y ∂y ∂y ∂y 
(6.3.11)
Or,
[ B ] = [ B1 ](5×3) [ B2 ](5×3) [ B3 ](5×3) [ B4 ](5×3)  (6.3.12)

Now, eq. (6.3.5) can be expressed using above relation as

4
={σ }P [C=
] P [ B ] {d i } [ C ] P ∑ [ B ] {d i } (6.3.13)
i =1

Or,
[C ]P [ B ] = [CP B1 ](5×3) [CP B2 ](5×3) [CP B3 ](5×3) [CP B4 ](5×3)  (6.3.14)

Considering the ith sub-matrix of the above equation,

 − µ t 2  ∂N i  t 2  ∂N i 
   
 0 1 − µ  ∂y  1 − µ  ∂x 
 −t 2  ∂N i  υ t  ∂N i
2

    
 0 1 − µ  ∂x  1 − µ  ∂y 
 
Et −t 2  ∂N i  t 2  ∂N i  
[CBi ] = 
    (6.3.15)
12 (1 + µ )  0 2  ∂x  2  ∂y  
 
  ∂N i  
6α  ∂x 
 6α N i 
 0 
  ∂N i  
6α   
  ∂y  −6α N i 0 

The bending and shear terms form above equation are separated and written as
 21

  t 2  N  t 2  N i  
  i  
  1    y  1    x   0 0 0  
 0   
   0 
0  
 t 2  N i  t  N i   0
2

      0 0  
Et   0 1   x  1    y   0
CBi      6  N 
121       2
   t 2
 N    i 
t N   x Ni  
     
i
i
 0 
 0 
2 x   
2  y   
   N i 
   
 0 0 0   y N i 0  
  
  0 0 0  
(6.3.16)
The above expression can be written in compact form as
CBi  = CBi b + CBi s (6.3.17)
Here, the contributions of bending and shear terms to stress displacement matrix is denoted as [CBi]b and
[CBi]s respectively. Generally, the contribution due to bending, [CBi]b in eq. (6.3.17) is evaluated
considering 2×2 Gauss points where as the shear contribution [CBi]s is evaluated considering 1×1 Gauss
point.

6.3.3 Element Stiffness Matrix

The expression for element stiffness matrix is
 k    A  B Cp  Bdx dy
T
(6.3.18)

In natural coordinate system, the stiffness matrix is expressed as

 k    A  B Cp  B J d d
T
(6.3.19)

Using value of [C]P form eq. 6.3.1 and [Bi] from 6.3.11, the product of  B Cp  Bis evaluated as
T
 22

 N i N i 
 0 0 0 
 x y 
 
 N i N i 
 k    B C  B   0
T
  0 N i 
   p   y x 
 
 N N i 
 i 0 Ni 0 
 x y 
  i1,2,3,4
 N i 
 0 0 
   x 
  
    0 0   N 
 1  0      0  i 0 
 Et 3
   0 0   y 
 2  1 0      
 l 1      0 0   N N i 
  1  
   0  i 
 0   x y 
 
0
  2  
  N
 0 0 0 Et 1 0   i 0 Ni 

      x 
 0 0 0 21   0 1   
 N i N 0 
 y i
  i1,2,3,4
Or in short, (6.2.20)
 k   k12   k13   k14  
  11       
 
k   k 22   k 23   k 24  
 k    BT C  B    21        (6.2.21)
  p    k 32   k 33   k 34  
  k 33       

k   k 42   k 43   k 44  
  41        
Where,
 
 
 
 
 
  N i N i N i N i  N i N i 
 6    6 Nj 6 Nj 
  x x y y  y y 
 
  t 2  N N j    t  N N j 
2
   i     i 

N i 1    y y    1    y x 
 k ij   Et       
  121      6N i
y  t 2  N N  
  t 2  N N  
j
   i j
  6N i N j    i  
  2  x x    2  x y  
 
  t 2  N N j    t 2  N N j  
   i    i 

N i  1    x y   1    x x  
       
 6N i
 x  t 2  N N    t 2  N N   
j 
   i     i
  
j


  2  y x    2  y y   
 23

(6.3.22)
By separating the bending and shear terms from above equation,
 
 
0 
 0 0 
 
   t  N N j  t  N N j 
  t  N i N j   t  N i N j  
2 2 2 2
 k ij   Et   i    i 
  1 1 2    0 1    y y  2  x x   
 1    y x  2  x y 

    
  t  N N j  t  N N j    t 2  N N j  t 2  N N j  

2 2
  i   i    i    i 
0  1    x y  2  y x   1    x x  2  y y  
 
   N N N i N i  N N i 
 i i
   i Nj Nj
   x x y y  y x 
 
  
6  0 
N i
Ni Ni N j
 y 
  
 N i
N 0 N N 
 x
j i j 
 
(6.3.23)
Thus, the matrix  k  can now be written as the sum of bending and shear contributions
 
k  k  k
   b  s (6.2.24)
Or,
 k   k12   k13   k14     k11   k12   k13   k14  
  11  b  b  b   b    s  s  s  s 
   
  k 21   k 22   k 23   k 24     k 21   k 22   k 23   k 24  
k    b  b  b  b   s

 s  s  s  (6.2.25)
  k 
  33  b  k 32   k 33   k 34     k 33   k 32   k 33   k 34  
  b  b   b   s  s  s  s 
k 
  41  b  k 42   k 43   k 44     k 41   k 42   k 43   k 44  
 b  b   b   s  s  s  s 
The stiffness matrix k  can be evaluated from the following expression by substituting
 k  for  BT C  B in eq. (6.3.19) and is given as
  p

1 1
 k   1  
 1  k  J d d (6.2.26)

Here, J is the determinate of the Jacobian matrix. The Gauss Quadrature integration rule is used to
compute the stiffness matrix [k].

6.3.4 Nodal Load Vector

Considering a uniformly distributed load q on the plate, the equivalent nodal load vector can be
calculated for finite element analysis from the flowing expression.
 24

 Fx  q 1 1 q

     
Qi  = M x    ξNdη 0 d  A  N i 0 J d (6.3.27)
  A i    
M x  0 -1 -1
0
Using Gauss Quadrature integration rule the above expression can be evaluated as,
 N i 
 
Q = q  wi w j J  0 
n n
(6.3.28)
 
 0 i=1,2,3,4
i=1 i=1

The nodal load vectors from each element are assembled to find the global load vector at all the nodes.
 25

Lecture 4: Finite Element Analysis of Skew Plate

6.4.1 Introduction
Skew plates often find its application in civil, aerospace, naval, mechanical engineering structures.
Particularly in civil engineering fields they are mostly used in construction of bridges for dealing
complex alignment requirements. Analytical solutions are available for few simple problems. However,
several alternatives are also available for analyzing such complex problems by finite element methods.
Commonly used three discretization methods for skew plates are shown in Fig 6.4.1.

(b) Discretization using combination of

(a) Discretization using rectangular plate
rectangular and triangular plate
elements
elements

(c) Discretization using skew plate element

Fig 6.4.1 Discretization of a skew plate

 26

If the skew plate is discretized using only rectangular plate elements, the area of continuum excluded
from the finite element model may be adequate to provide incorrect results. Another method is to use
combination of rectangular and triangular elements. However, such analysis will be complex and it may
not provide best solution in terms of accuracy as, different order of polynomials is used to represent the
field variables for different types of elements. Another alternative exist using skew element in place of
rectangular element.

6.4.2 Finite Element Analysis of Skew Plate

Let consider a skew plate of dimension “2a” and “2b” as shown in Fig. 6.4.2. Let the skew angle of the
element be “ϕ”. It is possible for the parallelogram shown in Fig. 6.4.3 to map the coordinate from
orthogonal global coordinate system to a skew local coordinate system. If the local coordinates are
represented in the form of ξ, η, then the relationship can is represented as,
x     cos , y   sin  (6.4.1)
Hence,
  ycos ec,   x  ycot  (6.4.2)

Fig. 6.4.2 Skew plate in global coordinate system

 27

Fig. 6.4.3 Point “P” in global and local coordinate system

It is important to note that the terms (ξ, η) in the above equations represent the absolute coordinate of the
point “P” in skew coordinate system, not the natural coordinates. Since the above element has four
corner nodes and each node have three degrees of freedom present, a polynomial with minimum 12
independent terms are necessary for defining the displacement function w(x,y). Considering Pascal
triangle, and in order to maintain geometric isotropy, the displacement function may be considered as
follows:
w   0  1   2   3 2   4   5 2   63   7 2
(6.4.3)
  82 + 93  103  113
Hence corresponding values for �𝜃𝑥 , 𝜃𝑦 �are,
w
     2   4  2 5   7 2  2 8  3 9 2  103  311 2  (6.4.4)

w
   1  2 3   4  3 6 2  2 7   8 2  310 2  113 (6.4.5)

Or, in matrix form,
  
 w  1    2  2
 3  2 2 3  3 3   0 
    
   0 0 1 0  2 0  2 3
2 2
3 3 2   1 
       
   0 1 0 2  0 3 2 2 2 0 3 2 3   
 11
(6.4.6)
Or,
 28

{d } = [ Φ ]{α } (6.4.7)
The value of [α] can be determined using value of �𝑤, 𝜃𝜉 , 𝜃𝜂 � at four nodes as
 w1  1 0 0 0 0 0 0 0 0 0 0    0 
0
     
  x1  0 0 1 0 0 0 0 0 0 0 0   1 
0
 
  y1  0 1 0 0 0 0 0 0 0 0 0    2 
0
  
  
w 2  1 0 b 0 0 b2 0 0 0 b3 0 0    3 
  0  
0 1 0 0 2b 0 0 0 3b 2 0 0    4 
 x 2    
 y2  0 1 0 0 b 0 0 0 b 2 0 0 b3    5 
    
 w 3  1 a b a2 ab b 2 a3 a 2 b ab 2 b3 a 3b ab3    6 
    
 x3  0 0 1 0 a 2b 0 a2 2ab 3b 2 a3 3ab3    7 
  0  
 y3   1 0 2a b 0 3a 2 2ab b 2 0 3a 2 b b3    8 
 
w 4  1 a 0 a2 0 0 a3 0 0 0 0 0    9 
    
 x 4  0 0 1 0 a 0 0 a2 0 0 a3 0  10 
    
 y4  0 1 0 2a 0 0 3a 2 0 0 0 0 0  11 
(6.4.8)
Or,
{α } = [ Φ ] {di }
−1
(6.4.9)
Further, considering eq. (6.4.2)
  

 0,  cos ec
x y 

 (6.4.10)
  
 1,   cot  

x y 


Again,
𝜕2 𝑤
𝜒𝑥 ⎡− 𝜕𝑥 2 ⎤
⎢ 𝜕2 𝑤⎥
� 𝜒𝑦 � = ⎢− 𝜕𝑦 2 ⎥ (6.4.11)
𝜒𝑥𝑦 ⎢ 2𝜕2 𝑤 ⎥
⎣ 𝜕𝑥𝜕𝑦 ⎦
The values in right hand side of the eq. (6.4.11) can be calculated by using chain rule as,
w w  w  w
 .    (6.4.12)
x  x  x 
Therefore,
2w 2w
 2 (6.4.13)
x 2 
 29

Similarly,
w w  w  w w
      cos   cosec
y  y  y  
Thus, further derivation provides
2w 2  w
2
2w 2w
 cosec  2  cot  2  2cos  cosec
2
(6.4.14)
y 2   
And,
2w   w  2w 2w
     cot  2  cosec 
2

xy x  y 
(6.4.15)
 
Hence eq. (6.4.11) is converted to
  2 w    2 w 

 2   2 
 x   1  
 
  2 w   0 0
   2 w 
    cot 2  cosec 2 -cotcosec  
   2 
(6.4.16)
 y 2  
 2  -2cot 0 cosec 2   2 
 2 w   2 w 
   
 xy    
Or,
 x,y    H   ,  (6.4.17)
Further, by partial differentiation of eq. (6.4.3),
2w 
  2 3  6 6  2 7  610 
 2 

2w 
 2  2 5  2 8  6 9  611  (6.4.18)
 

2w 
2  4 7   4 8  610 2  6112 
 
Or, in matrix form,
 30

  2 w 
 
  2 
  0 0 0 2 0 0 6 2 0 0 6 0   0 
  2 w    
  2 


0 0 0 0 0  2 0 0  2   6 0  6 
   
    
 2   0 0 0 0 0 0 4 4 6 2  11
62 
0 0  
 2 w 
 
  
(6.4.19)
Or,
 ,   B  B1  di 
Or,
 x,y    H  B1  di  (6.4.20)
Again,
M x,y   D x,y  (6.4.21)
Where [D] for plane stress condition is
 
 
 1  0 
Eh 3  
D    1 0  (6.4.22)
l21   2   
 1  
0 0 
 2 
Using eq. (6.4.20) in eq. (6.4.21),
M x,y   D H  B A1  d (6.4.23)
The expression for bending strain energy stored,
a b

  x,y  M x,y dxdy

1
U 
T
(6.4.24)
2 0 0

Hence force vector,

U
a b
1   BT  H  T  D   H   B 1  d dxdy
 d 0 
T
        
0
(6.4.25)
a b
 1   1 
     B  H   D   H   B   J dd d
T T T

0 0
Where,
 31

 x y 
 
   x, y      1 0 
 J     
   x
   sin  (6.4.26)
   ,     y   cos  sin 

   

From expression in eq. (6.4.25)

U
F    k d (6.4.27)
 d
Hence,
a b

 k   sin  1    B,   H   D   H   B,  dd 1 
T T T
      
0 0

(6.4.28)
Thus, the element stiffness matrix of a skew element for plate bending analysis can be evaluated from
the above expression using Gauss Quadrature numerical integration.
 32

Lecture 5: Introduction to Finite Strip Method

6.5.1 Introduction
The finite strip method (FSM) was first developed by Y. K. Cheung in 1968. This is an efficient tool for
analyzing structures with regular geometric platform and simple boundary conditions. If the structure is
regular, the whole structure can be idealized as an assembly of 2D strips or 3D prisms. Thus the
geometry of the structure needs to be constant along one or two coordinate directions so that the width
of the strips or the cross-section of the prisms does not change. Therefore, the finite strip method can
reduce three and two-dimensional problems to two and one-dimensional problems respectively. The
major advantages of this method are (i) reduction of computation time, (ii) small amount of input (iii)
easy to develop the computer code etc. However, this method will not be suitable for irregular geometry,
material properties and boundary conditions.

6.5.2 Finite Strip Method

To understand the finite strip method, let consider a rectangular plate with x and y axes in the plane of
the plate and axis z in the thickness direction as shown in Fig. 6.5.1. The corresponding displacement
components of the plate are denoted as u, v and w.

Fig. 6.5.1 Finite strip in a plate

The strips are assumed to be connected to each other along a discreet number of nodal lines that coincide
with the longitudinal boundaries of the strip. The general form of the displacement function in two
dimensions for a typical strip is given by
𝑤 = 𝑤(𝑥, 𝑦) = ∑𝑛𝑚=1 𝑓𝑚 (𝑦)𝑋𝑚 (𝑥) (6.5.1)
Here, the functions ƒm(y) are polynomials and the functions Xm(x) are trigonometric terms that satisfy the
end conditions in the x direction. The functions Xm(x) can be taken as basic functions (mode shapes) of
the beam vibration equation.
 33

𝑑4 𝑋 𝜇4
𝑑𝑦 4
− 𝐿4 𝑌 = 0 (6.5.2)
Here L is the length of beam strip and µ is a parameter related to material, frequency and geometric
properties. The general solution of the above equation will become
𝑛𝜋𝑥 𝑛𝜋𝑥 𝑛𝜋𝑥 𝑛𝜋𝑥
𝑋𝑚 (𝑥) = 𝐶1 sin � 𝐿
� + 𝐶2 cos � 𝐿
� + 𝐶3 sinh � 𝐿
� + 𝐶4 cosh � 𝐿
� (6.5.3)
Four conditions at the boundaries are necessary to determine the coefficients C1 to C4 in the above
expression.

6.5.2.1 Boundary conditions

According to different end conditions eq. (6.5.3) can be solved. Solution of the above equation is
evaluated for few boundary conditions in the below.
(a) Both end simply supported
For simply supported end, following conditions will arise:
(i) At one end (say at x =0) displacement and moment will be zero: x 0  = x" 0  = 0
(ii) At other end (at x =L) displacement and moment will be zero: x  L = x"  L = 0
Thus, considering above boundary conditions, eq. (6.5.3) yields to the following mode shape function:
m
 μ πx 
Where,  m  ,2.......upto n t h term
nπx
X m  x =  sin = sin  m  (6.5.4)
n=1 l  l 
Since the functions Xm are mode shapes, they are orthogonal and therefore, they satisfy the following
relations:
𝐿
∫0 𝑋𝑚 (𝑥)𝑋𝑛 (𝑥)𝑑𝑥 = 0 𝑓𝑜𝑟 𝑚 ≠ 𝑛 (6.5.5)
And
𝐿
∫0 𝑋"(𝑥)𝑚 𝑋"(𝑥)𝑛 𝑑𝑥 = 0 𝑓𝑜𝑟 𝑚 ≠ 𝑛 (6.5.6)
The orthogonal properties of Xm(x) result in structural matrices with narrow bandwidths and thus
minimizing computational time and storage. Using relation in eq. (6.5.4), eq. (6.5.1) can now be written
as
𝑛𝜋𝑥
𝑤 = 𝑤(𝑥, 𝑦) = ∑𝑚
𝑛=1 𝑓𝑚 (𝑦). sin � 𝐿
� (6.5.7)
(b) Both end fixed supported
In case of fixed supported end at both the side, the following boundary conditions will be adopted:
(i) At one end (say at x =0) displacement and slope will be zero: x 0  = x' 0  = 0
(ii) At other end (at x =L) displacement and slope will be zero: x  L = x'  L = 0
For the above boundary conditions, eq. (6.5.3) yields to the following:
 μ x  μ y      
X m  x = sin  αm cos
- sin h  m- cos - h   μm y   μm y  (6.5.8)
 L   L  m   L   L 
 
 34

 2m +1 μ - sinhμ sin

Where  μm = 4.73,7.8532,10.996........ π  and αm = m m
 2 μ - coshμ cos m m
(c) One end simply supported, other end fixed
(i) At simply supported end (say at x =0) displacement and moment will be zero: x 0  = x" 0  = 0
(ii) At fixed end (say at x =L) displacement and slope will be zero: x  L = x'  L = 0
Thus, the solution of eq. (6.5.3) will become
 μm x   
 h  -
X m  x =αsinsin  μm x  (6.5.9)
 l  m  l 
4m  1 sinμm
where, μm  3.9266,7.0685,10.2102.......  and αm =
4 sinhμm
(d) Both end free
If both the end of the strip element is free, the following boundary conditions will be assumed:
(i) At one end (say at x =0) moment and shear will become zero: x" 0  = x"' 0  = 0
(ii) At other end (at x =L) moment and shear will become zero: x"  L = x'"  L = 0
Thus, for the above end conditions, eq. (6.5.3) yields to the following:
2x
X 1 = 1,μ =1 0 & X =2 1 - ,μ =2 1
L
(6.5.10)
 μ x μ x  μm x μm x 
X m  x = sin  mα  + - m 
cossin h +m cosh 
 l  l  l l 
sinμm - sin hμm 2m - 3
Where, αm = and μm = 4.73,7.8532, 10.996....... π, for m = 3,4 ,- - - 
cosμm - coshμm 2

6.5.3 Finite Element Formulation

In this section, finite element solution for a finite strip will be evaluated considering simply supported
conditions at both the end. As a result, the functions ƒm(Y) in eq. (6.5.7) can be expressed for the
bending problem as
𝑓𝑚 (𝑌) = 𝑤(𝑦) = ∝0 +∝1 𝑦 +∝2 𝑦 2 +∝3 𝑦 3 (6.5.11)
Applying boundary conditions of the strip plate of width b, the following relations will be obtained.
𝑤0 1 0 0 0 𝛼0
𝜃 0 1 0 0 � �𝛼1 �
� 𝑤𝑥0 � = � 𝛼2 (6.5.12)
0 1 𝑏 𝑏2 𝑏3
𝜃𝑥0 0 1 2𝑏 3𝑏 2 𝛼3
Thus, the nodal displacement can be written in short as
{𝑑} = [𝐴]{𝛼} (6.5.13)
Thus, the unknown coefficient α are obtained from the following relations.
{𝛼} = [𝐴]−1 {𝑑} (6.5.14)
 35

The formulation of the finite strip method is similar to that of the finite element method. For example,
for a strip subjected to bending, the moment curvature relation will become
𝜕2 𝑤
− 𝜕𝑥 2 ⎫
𝑀𝑥 𝐷𝑥 𝐷1 0 ⎧ ⎪ 𝜕2 𝑤 ⎪
� 𝑀𝑦 � = �𝐷1 𝐷𝑦 0 � − 2
𝜕𝑦
(6.5.15)
𝑀𝑥𝑦 0 0 2𝐷𝑥𝑦 ⎨ 𝜕2 𝑤 ⎬
⎪2 ⎪
⎩ 𝜕𝑥𝜕𝑦⎭
Where Mx , My and Mxy are moments per unit length and [D] is the elasticity matrix. From eq. (6.5.7)
the following expressions are evaluated to incorporate in the above equation.
𝜕2 𝑤 𝑛𝜋 2 𝑛𝜋𝑥
− 𝜕𝑥 2 = ∑𝑚
𝑛=1 𝑤(𝑦) � 𝐿 � 𝑠𝑖𝑛 𝐿
𝜕2 𝑤 𝑛𝜋𝑥
− 𝜕𝑦 2 = − ∑𝑚
𝑛=1 −(2𝛼2 + 6𝛼3 𝑦)𝑠𝑖𝑛 𝐿
(6.5.16)
𝜕2 𝑤 𝑛𝜋 𝑛𝜋𝑥
2 𝜕𝑥𝜕𝑦 = ∑𝑚 2
𝑛=1 2(𝛼1 + 2𝛼2 𝑦 + 3𝛼3 𝑦 ) � 𝐿 � 𝑐𝑜𝑠 𝐿
The above expression is written in matrix form in the below
𝜕2 𝑤
⎧ − 𝜕𝑥 2 ⎫
⎪ 𝜕2 𝑤 ⎪ ∑ 𝑤(𝑦)𝑁 2 𝑠𝑖𝑛𝑁𝑥
− 2 = �− ∑ −(2𝛼2 + 6𝛼3 𝑦)𝑠𝑖𝑛𝑁𝑥 � (6.5.17)
⎨ 𝜕𝑦 2
⎬ 2
∑ 2(𝛼1 + 2𝛼2 𝑦 + 3𝛼3 𝑦 )𝑁𝑐𝑜𝑠𝑁𝑥
⎪2 𝜕 𝑤 ⎪
⎩ 𝜕𝑥𝜕𝑦⎭
𝑛𝜋
Here, 𝐿 is denoted as N. Rearranging the above expression, one can find the following.
𝜕2 𝑤
⎧ − 𝜕𝑥 2 ⎫
⎪ 𝜕2 𝑤 ⎪ sin 𝑁𝑥 0 0 𝑤(𝑦)𝑁 2
− 𝜕𝑦 2 = � 0 sin 𝑁𝑥 0 �� −(2𝛼2 + 6𝛼3 𝑦) �
⎨ 2 ⎬ 0 0 cos 𝑁𝑥 2(𝛼1 + 2𝛼2 𝑦 + 3𝛼3 𝑦 )𝑁
2
⎪2 𝜕 𝑤 ⎪
⎩ 𝜕𝑥𝜕𝑦⎭
𝛼0
sin 𝑁𝑥 0 0 𝑁2 𝑁2𝑦 𝑁2𝑦2 𝑁2𝑦3
𝛼
=� 0 sin 𝑁𝑥 0 ��0 0 −2 −6𝑦 � �𝛼1 � (6.5.18)
2
0 0 cos 𝑁𝑥 0 2𝑁 4𝑁𝑦 6𝑁𝑦 2 𝛼3
Thus, in short, the curvature and moment equation will become
{𝜒} = [𝐻(𝑥)][𝐵(𝑦)]{𝛼} = [𝐻(𝑥)][𝐵(𝑦)][𝐴]−1 {𝑑} (6.5.19)
{𝑀} = [𝐷]{𝜒} = [𝐷][𝐻(𝑥)][𝐵(𝑦)][𝐴] {𝑑} −1
(6.5.20)
Now, the strain energy for the bending element can be written similar to plate bending formulation.
𝐿 𝑏 𝐿 𝑏
1 1
𝑈 = � �{𝜒}𝑇 {𝑀} 𝑑𝑥 𝑑𝑦 = � �{𝛼}𝑇 [𝐴−1 ]𝑇 [𝐵(𝑦)]𝑇 [𝐻(𝑥)]𝑇 [𝐷][𝐻(𝑥)][𝐵(𝑦)][𝐴]−1 {𝑑} 𝑑𝑥 𝑑𝑦
2 2
0 0 0 0
(6.5.21)
Thus the force vector can be derived as
 36

𝜕𝑈 𝐿 𝑏
{𝐹} =
𝜕{𝑑}
= ∫0 ∫0 [𝐴−1 ]𝑇 [𝐵(𝑦)]𝑇 [𝐻(𝑥)]𝑇 [𝐷][𝐻(𝑥)][𝐵(𝑦)][𝐴]−1 𝑑𝑥 𝑑𝑦 {𝑑} = [𝑘]{𝑑}
(6.5.22)
Thus, the stiffness matrix of a strip element can be obtained from the following expression.
𝐿 𝑏
[𝑘] = ∫0 ∫0 [𝐴−1 ]𝑇 [𝐵(𝑦)]𝑇 [𝐻(𝑥)]𝑇 [𝐷][𝐻(𝑥)][𝐵(𝑦)][𝐴]−1 𝑑𝑥 𝑑𝑦
𝐿 𝑏
= [𝐴−1 ]𝑇 ∫0 ∫0 [𝐵(𝑦)]𝑇 [𝐻(𝑥)]𝑇 [𝐷][𝐻(𝑥)][𝐵(𝑦)] 𝑑𝑥 𝑑𝑦 [𝐴]−1 (6.5.23)
𝐿
The stiffness matrix [k] can be simplified by integrating the term∫0 [𝐻(𝑥)]𝑇 [𝐷][𝐻(𝑥)] 𝑑𝑥 as follows.
𝐿
∫0 [𝐻(𝑥)]𝑇 [𝐷][𝐻(𝑥)] 𝑑𝑥
sin 𝑁𝑥 0 0 𝐷𝑥 𝐷1 0 sin 𝑁𝑥 0 0
𝐿
= ∫0 �
0 sin 𝑁𝑥 0 � �𝐷1 𝐷𝑦 0 �� 0 sin 𝑁𝑥 0 � 𝑑𝑥
0 0 cos 𝑁𝑥 0 0 2𝐷𝑥𝑦 0 0 cos 𝑁𝑥
𝐷𝑥 sin 𝑁𝑥 𝐷1 sin 𝑁𝑥 0 sin 𝑁𝑥 0 0
𝐿
= ∫0 � 𝐷1 sin 𝑁𝑥 𝐷𝑦 sin 𝑁𝑥 0 �� 0 sin 𝑁𝑥 0 � 𝑑𝑥
0 0 2𝐷𝑥𝑦 cos 𝑁𝑥 0 0 cos 𝑁𝑥
𝐷𝑥 sin2 𝑁𝑥
𝐷1 sin2 𝑁𝑥 0
𝐿 2 2
= ∫0 � 𝐷1 sin 𝑁𝑥
𝐷𝑦 sin 𝑁𝑥 0 � 𝑑𝑥 (6.5.24)
2
0 0 2𝐷𝑥𝑦 cos 𝑁𝑥
Here, the terms Dx, Dy and Dxy are constant and not varied with x or y. Following integrations are carried
out to simplify the above expression further.
𝐿 𝐿 1−𝑐𝑜𝑠2𝑁𝑥 1 𝑠𝑖𝑛2𝑁𝑥 𝐿 1 𝑠𝑖𝑛2𝑁𝐿
Now ∫0 𝑠𝑖𝑛2 𝑁𝑥 𝑑𝑥 = ∫0 � 2
� 𝑑𝑥 = 2 �𝑥 − 2𝑁
� = 2 �𝐿 − 2𝑁
�
0
𝑛𝜋
𝑛𝜋 𝐿 1 𝑠𝑖𝑛2 𝐿 1
Putting, 𝑁 = 𝐿
finally ∫0 𝑠𝑖𝑛2 𝑁𝑥 𝑑𝑥 will become 2 �𝐿 − 𝑛𝜋
𝐿
� = 2𝐿
2
𝐿
𝐿 𝐿 1+𝑐𝑜𝑠2𝑁𝑥 1 𝑠𝑖𝑛2𝑁𝑥 𝐿 1
Similarly; ∫0 𝑐𝑜𝑠 2 𝑁𝑥 𝑑𝑥 = ∫0 � 2
� 𝑑𝑥 = 2 �𝑥 + 2𝑁
� =2𝐿
0
Thus,
𝐷𝑥 𝐷1 0
𝐿 𝐿 𝐿
∫0 [𝐻(𝑥)]𝑇 [𝐷][𝐻(𝑥)] 𝑑𝑥 = 2 �𝐷1 𝐷𝑦 0 �=
2
[𝐷] (6.5.25)
0 0 2𝐷𝑥𝑦
Using eq. (6.5.25), the expression for stiffness matrix [k] in eq. (6.5.23) is simplified as follows.
𝑏 𝐿
[𝑘] = [𝐴−1 ]𝑇 ∫0 [𝐵(𝑦)]𝑇 [𝐷][𝐵(𝑦)] 𝑑𝑦 [𝐴−1 ]
2
2
𝑁 0 0
⎡ 2 ⎤ 𝐷𝑥 𝐷1 0 𝑁2 𝑁2𝑦 𝑁2𝑦2 𝑁2𝑦3
𝐿 𝑏 𝑁 𝑦 0 2𝑁
= 2 [𝐴−1 ]𝑇 ∫0 ⎢ 2 2 ⎥ �𝐷1 𝐷𝑦 0 �� 0 0 −2 −6𝑦 � 𝑑𝑦[𝐴]−1
⎢𝑁 𝑦 −2 4𝑁𝑦 ⎥
0 0 2𝐷𝑥𝑦 0 2𝑁 4𝑁𝑦 6𝑁𝑦 2
⎣𝑁 2 𝑦 3 −6𝑦 6𝑁𝑦 2 ⎦
 37

𝑁 2 𝐷𝑥 𝑁 2 𝐷1 0
𝑏⎡ ⎤ 2 𝑁2𝑦
𝐿 𝑁 2 𝑦𝐷𝑥 𝑁 2 𝑦𝐷1 4𝑁𝐷𝑥𝑦 ⎥ 𝑁 𝑁2𝑦2 𝑁2𝑦3
⎢
= [𝐴−1 ]𝑇 � ⎢ 2 2 �0 0 −2 −6𝑦 � 𝑑𝑦[𝐴]−1
2 𝑁 𝑦 𝐷𝑥 − 2𝐷1 𝑁 2 𝑦 2 𝐷1 − 2𝐷𝑦 8𝐷𝑥𝑦 𝑁𝑦 ⎥
𝑜 ⎢ 2 3 ⎥ 0 2𝑁 4𝑁𝑦 6𝑁𝑦 2
⎣𝑁 𝑦 𝐷𝑥 − 6𝑦𝐷1 𝑁 2 𝑦 3 𝐷1 − 6𝑦𝐷𝑦 12𝑁𝑦 2 𝐷𝑥𝑦 ⎦

(𝑁 4 𝐷𝑥 ) (𝑁 4 𝑦𝐷𝑥 ) (𝑁 4 𝑦 2 𝐷𝑥 − 2𝑁 2 𝐷1 ) (𝑁 4 𝑦 3 𝐷𝑥 − 6𝑁 2 𝑦𝐷1 )
⎡ ⎤
⎢ (𝑁 4 𝑦𝐷𝑥 ) �𝑁 4 𝑦 2 𝐷𝑥 + 8𝑁 2 𝐷𝑥𝑦 � �𝑁 4 𝑦 3 𝐷𝑥 − 2𝑁 2 𝑦𝐷1 + 16𝑁 2 𝑦𝐷𝑥𝑦 � �𝑁 4 𝑦 4 𝐷𝑥 − 6𝑁 2 𝑦 2 𝐷1 + 24𝑁 2 𝑦 2 𝐷𝑥𝑦 �⎥
𝑏⎢ ⎥
𝐿 𝑁 4 𝑦 4 𝐷𝑥 − 4𝑁 2 𝑦 2 𝐷1 𝑁 4 𝑦 5 𝐷𝑥 − 2𝐷1 𝑁 2 𝑦 3 − 6𝑁 2 𝑦 3 𝐷1
⎢ � � � � ⎥
= [𝐴−1 ]𝑇 � ⎢ +4𝐷 𝑦 + 32𝑁 2 2
𝑦 𝐷 𝑥𝑦 −12𝑦𝐷 𝑦 + 48𝑁 2 3
𝑦 𝐷 𝑥𝑦 ⎥ 𝑑𝑦[𝐴]
−1
2 (𝑁 4 2
𝑦 𝐷 − 2𝑁 2
𝐷 ) �𝑁 4 𝑦 3 𝐷𝑥 − 2𝑁 2 𝑦𝐷1 + 8𝑁 2 𝑦𝐷𝑥𝑦 �
0 ⎢ 𝑥 1 4 5
𝑁 𝑦 𝐷𝑥 − 6𝑁 𝑦 𝐷1 2 3 4 6
𝑁 𝑦 𝐷𝑥 − 6𝑁 𝑦 𝐷1 2 4 ⎥
⎢(𝑁 4 𝑦 3 𝐷𝑥 − 6𝑁 2 𝑦𝐷1 ) �𝑁 4 𝑦 4 𝐷𝑥 − 6𝑁 2 𝑦 2 𝐷1 + 24𝑁 2 𝑦 2 𝐷𝑥𝑦 � 2 3 2 4 2 ⎥
⎢ � −2𝑁 𝑦 𝐷 1 + 12𝑦𝐷 𝑦 � � −6𝑁 𝑦 𝐷1 + 36𝑦 𝐷𝑦 � ⎥
2 3 2 4
⎣ +48𝑁 𝑦 𝐷𝑥𝑦 +72𝑁 𝑦 𝐷𝑥𝑦 ⎦
(6.5.26)
Thus, by putting the assumed shape function, the stiffness matrix of a strip element can be evaluated
numerically using Gaussian Quadrature or other numerical integration methods.
 38

Lecture 6: Finite Element Analysis of Shell

6.6.1 Introduction
A shell is a curved surface, which by virtue of their shape can withstand both membrane and bending
forces. A shell structure can take higher loads if, membrane stresses are predominant, which is primarily
caused due to in-plane forces (plane stress condition). However, localized bending stresses will appear
near load concentrations or geometric discontinuities. The shells are analogous to cable or arch structure
depending on whether the shell resists tensile or, compressive stresses respectively. Few advantages
using shell elements are given below.
1. Higher load carrying capacity
2. Lesser thickness and hence lesser dead load
3. Lesser support requirement
4. Larger useful space
5. Higher aesthetic value.

The example of shell structures includes large-span roof, cooling towers, piping system, pressure vessel,
aircraft fuselage, rockets, water tank, arch dams, and many more. Even in the field of biomechanics,
shell elements are used for analysis of skull, Crustaceans shape, red blood cells, etc.

6.6.2 Classification of Shells

Shell may be classified with several alternatives which are presented in Fig 6.6.1.
 39

Fig 6.6.1 Classification of shells

Depending upon deflection in transverse direction due to transverse shear force per unit length, the shell
can be classified into structurally thin or thick shell. Further, depending upon the thickness of the shell
in comparison to the radii of curvature of the mid surface, the shell is referred to as geometrically thin or
thick shell. Typically, if thickness to radii of curvature is less than 0.05, then the shell can be assumed as
a thin shell. For most of the engineering application the thickness of shell remains within 0.001 to 0.05
and treated as thin shell.

6.6.3 Assumptions for Thin Shell Theory

Thin shell theories are basically based on Love-Kirchoff assumptions as follows.
1. As the shell deforms, the normal to the un-deformed middle surface remain straight and normal
to the deformed middle surface undergo no extension. i.e., all strain components in the direction
of the normal to the middle surface is zero.
2. The transverse normal stress is neglected.
Thus, above assumptions reduce the three dimensional problems into two dimensional.

6.6.4 Overview of Shell Finite Elements

Many approaches exist for deriving shell finite elements, such as, flat shell element, curved shell
element, solid shell element and degenerated shell element. These are discussed briefly bellow.
 40

(a) Flat shell element

The geometry of these types of elements is assumed as flat. The curved geometry of shell is obtained by
assembling number of flat elements. These elements are based on combination of membrane element
and bending element that enforced Kirchoff’s hypothesis. It is important to note that the coupling of
membrane and bending effects due to curvature of the shell is absent in the interior of the individual
elements.

(b) Curved shell element

Curved shell elements are symmetrical about an axis of rotation. As in case of axisymmetric plate
elements, membrane forces for these elements are represented with respected to meridian direction
as(𝑢, 𝑁𝑧 , 𝑀𝜃 ) and in circumferential directions as(𝑤, 𝑁𝜃 , 𝑀𝑧 ). However, the difficulties associated with
these elements includes, difficulty in describing geometry and achieving inter-elemental compatibility.
Also, the satisfaction of rigid body modes of behaviour is acute in curved shell elements.

(c) Solid shell element

Though, use of 3D solid element is another option for analysis of shell structure, dealing with too many
degrees of freedom makes it uneconomic in terms of computation time. Further, due to small thickness
of shell element, the strain normal to the mid surface is associated with very large stiffness coefficients
and thus makes the equations ill conditioned.

(d) Degenerated shell elements

Here, elements are derived by degenerating a 3D solid element into a shell surface element, by deleting
the intermediate nodes in the thickness direction and then by projecting the nodes on each surface to the
mid surface as shown in Fig. 6.6.2.

(a) 3D solid element (b) Degenerated Shell element

 41

6.6.2 Degeneration of 3D element

This approach has the advantage of being independent of any particular shell theory. This approach can
be used to formulate a general shell element for geometric and material nonlinear analysis. Such element
has been employed very successfully when used with 9 or, in particular, 16 nodes. However, the 16-
node element is quite expensive in computation. In a degenerated shell model, the numbers of unknowns
present are five per node (three mid-surface displacements and two director rotations). Moderately thick
shells can be analysed using such elements. However, selective and reduced integration techniques are
necessary to use due to shear locking effects in case of thin shells. The assumptions for degenerated
shell are similar to the Reissner-Mindlin assumptions.

6.6.5 Finite Element Formulation of a Degenerated Shell

Let consider a degenerated shell element, obtained by degenerating 3D solid element. The degenerated
shell element as shown in Fig 6.6.2(b) has eight nodes, for which the analysis is carried out. Let (𝜉, 𝜂)
are the natural coordinates in the mid surface. And ς is the natural coordinate along thickness direction.
The shape functions of a two dimensional eight node isoparametric element are:
1 xhxhxxh
1   1 1  1 1 
N1  N5 
4 2
1  xhxhxhh
1   1 1  1  1 
N2  N6 
4 2 (6.6.1)
1  xhxhxxh
1    1 1  1 1  
N3  N7 
4 2
1 xhxhxhh
1    1 1 1  1 
N4  N8 
4 2
The position of any point inside the shell element can be written in terms of nodal coordinates as

 x  
 
 xi 
 
 xi 
 

  
     

  8
1  V   1  V   
 y   N i xh ,   yi    yi   (6.6.2)

 
 

 2 
 
 2 
 
 


 z  
 zi  
 zi  bottom 
i 1
 
 
  top
   

Since, ς is assumed to be normal to the mid surface, the above expression can be rewritten in terms of a
vector connecting the upper and lower points of shell as
 x   x   xi    xi   xi  
    i       
8
1      V     
 y   N i xh ,   yi    yi     yi    yi  
  i1  2 
     2      

 z    
 z 
 
 z 

  
 z 

 
 z 
 
  i top i bottom   i top i bottom 

Or,
 42

 x
  
 xi 
  

  
  V 

  8
 

   i
y  N  xh
,   i 
y  V3i 
(6.6.3)

 
 i 1

 
 2  

 
 z 
 
 

 zi 
 

Where,
 xi   x   xi    xi   xi 
  1  i     











 
 yi    yi    yi   and, V3i   yi    yi  (6.6.4)
  2       
 
 
 

 zi  

 zi 
top 
 zi 

bottom   

 zi 

top


 zi bottom



Fig. 6.6.3 Local and global coordinates

For small thickness, the vector V3i can be represented as a unit vector tiv3i:
 x  xi  
 y 
8
  V 
,  yi   ti v3i 

  
N i xh (6.6.5)
i 1

 
 2 

 z  
 zi  
th
Where, ti is the thickness of shell at i node. In a similar way, the displacement at any point of the shell
element can be expressed in terms of three displacements and two rotation components about two
orthogonal directions normal to nodal load vector V3i as,
 u   ui  
  8
   Vti ai 
  
v  N i xh
,  vi   v1i v2i   (6.6.6)
i 1

 
 2  b 
 i 
w 
wi  
Where, (𝛼𝑖 , 𝛽𝑖 ) are the rotations of two unit vectors v1i & v2i about two orthogonal directions normal to
nodal load vector V3i.The values of v1i and v2i can be calculated in following way:
The coordinate vector of the point to which a normal direction is to be constructed may be defined as
x  xiˆ  yjˆ  zkˆ (6.6.7)
 43

In which, iˆ, ˆj , kˆ are three (orthogonal) base vectors. Then, V1i is the cross product of iˆ & V3i as shown
below.
V1i  iˆV3i & V2i  V3i V1i (6.6.8)
and,
V1i & v2i  V2i
v1i  (6.6.9)
V1i V2i

6.6.5.1 Jacobian matrix

The Jacobian matrix for eight node shell element can be expressed as,
 8 N 8
* N i
8
N 
   xi  txi*  i   y  tyi    zi  tzi*  i 
 i1 xxx
i
  
 i 1 i 1

 8  8
 8
 
 J      xi  txi*  i   yi  tyi*  i   zi  tzi*  i 
N N N
(6.6.10)
 i1 hhh
i 1  i 1  
 8 8 8 
 
  Nx *
i  Ny *
i  Nz *
i 
 i 1 i 1 i 1 

6.6.5.2 Strain displacement matrix

The relationship between strain and displacement is described by
e    B d  (6.6.11)
Where, the displacement vector will become:
d   u1 v1 w1 v11v21  u8 v8 w8 v18v28 
T
(6.6.12)
And the strain components will be
 u 
 
 x 
 v 
 
 y 
 
 u v 

e      (6.6.13)
 y x 
 
 v  w 
 z y 
 
 w u 
  
 x z 
Using eq. (6.6.6) in eq. (6.6.13) and then differentiating w.r.t. (𝜉, 𝜂, 𝜍) the strain displacement matrix
will be obtained as
 44

 u v w  
 N i 
 
 N   
 N  
  
 
 
 V i  
V i 
 xxx   
 xx 
 
  
 
  x 

   b1   a1 
T T

 
 
 
   
 
  
 u v w  N N i    t v  N i   
    i ui vi wi   i 2i V
8 8 8
tv
   b 2    i 1i V  a
 hhhhh
        
2      
i 1 2  h  
  2
  i1  i 1 
 u v w 

 0   
 N    b 3  i 

 N



 a3  i
  
 
 

i 
 

i 

 V V V      
 

(6.6.14)

6.6.5.3 Stress strain relation

The stress strain relationship is given by
s    D e  (6.6.15)
Using eq. (6.6.11) in eq. (6.6.15) one can find the following relation.
s    D  B d  (6.6.16)
Where, the stress strain relationship matrix is represented by
1 m 0 0 0 
 
m 1 0 0 0 
 
 1  m 
 0 0 0 
0
E  2  (6.6.17)
 D   
1 m 2  1 
am 
0 0 0 0 
 2 
 
 1 
am
0 0 0 0
 2 
The value of shear correction factor a is considered generally as 5/6. The above constitutive matrix can
be split into two parts ([Db] and [Ds] )for adoption of different numerical integration schemes for
bending and shear contributions to the stiffness matrix.
 Db   0 

 D        (6.6.18)
 0  
 Ds 

Thus,
 
 
1 m 0 
E  
 Db   2 
m 1 0  (6.6.19)
1 m  
 1 m 
0 0 
 2 
and
E a  1 0
 Ds     (6.6.20)
2 1  m  0 1
 45

It may be important to note that the constitutive relation expressed in eq. (6.6.19) is same as for the case
of plane stress formulation. Also, eq. (6.6.20) with a multiplication of thickness h is similar to the terms
corresponds to shear force in case of plate bending problem.

6.6.5.4 Element stiffness matrix

Finally, the stiffness matrix for the shell element can be computed from the expression
 k     B   D  B  d 
T
(6.6.21)
However, it is convenient to divide the elemental stiffness matrix into two parts: (i) bending and
membrane effect and (ii) transverse shear effects. This will facilitate the use of appropriate order of
numerical integration of each part. Thus,
 k    k b   k s (6.6.22)
Where, contribution due to bending and membrane effects to stiffness is denoted as [k]b and transverse
shear contribution to stiffness is denoted as [k]s and expressed in the following form.
 k b    B b  D b  B b d  and  k s    B s  D s  B s d 
T T
(6.6.23)
Numerical procedure will be used to evaluate the stiffness matrix. A 2 ×2 Gauss Quadrature can be used
to evaluate the integral of [k]b and one point Gauss Quadrature may be used to integrate [k]s to avoid
shear locking effect.
 1

Lecture 1: Finite Elements for Elastic Stability

7.1.1 Introduction
There are two types of structural failure associated: namely (i) material failure and (ii) geometry or
configuration failure. In case of material failure, the stresses exceed the permissible values which may
lead to formation of cracks In configuration failure, the structure is unable to maintain its designed
configuration under the external disturbance even though the stresses are in permissible range. The
stability loss due to tensile forces falls in broad category of material instability. The loss of stability
under compressive load is termed as structural or geometric instability which is commonly known as
buckling. Thus, buckling is considered by a sudden failure of a structural element subjected to large
compressive stress, where the actual compressive stress at the point of failure is less than the ultimate
allowable compressive stresses. Buckling is a wide term that describes a range of mechanical
behaviours. Generally, it refers to an incident where a structural element in compression deviates from a
behaviour of elastic shortening within the original geometry and undergoes large deformations involving
a change in member shape for a small increase in force. Various types of buckling may occur such as
flexural buckling, torsional buckling, torsional flexural buckling etc. Here, the use of finite element
technique for elastic stability analysis of various structural members will be briefly discussed.

7.1.2 Buckling of Truss Members

For a truss member, the axial strain can be expressed in terms of its displacements
u 1  u   v  
2 2

x  +       (7.1.1)
x 2  x   x  
Here, u and v are the displacements in local X and Y directions respectively. Here, the strain-
displacement relation is nonlinear. The total potential energy in the member with uniform cross sectional
area and subjected to external forces can be written as
L
EA
=Π 
2 0
ε x 2 dx -  P1v1 + P2 v2  (7.1.2)

Where, P1 and P2 are the external forces in nodes 1 and 2, v1 and v2 are the vertical displacements in
nodes 1 and 2, E and A are the modulous of elasticity and cross sectional area respectively (Fig. 7.1.1).
The length of the member is considered to be L.
 2

Fig.7.1.1 Truss element in local coordinate system

The above expression can be rewritten in terms of displacement variables as

EA  u 1  u   v   
2
L 2 2

2 0  x 2  x   x   

=Π  +       dx -  P1v1 + P2 v2 

L   u 2  v 2  
EA  u   u 
3 2
    
2 2

  +        +        dx -  P1v1 + P2 v2 
v u 1
2 0  x   x 
=
x   x   4  x   x   
  (7.1.3)
Neglecting higher order terms and considering EA u = P the above expression can simplify to the
x
following form.
EA  u  P  v 
L 2 L 2

2 0  x  2 0  x 
=Π   dx +   dx -  P1v1 + P2 v2  (7.1.4)

Considering nodal displacements at nodes 1 and 2 as u1, v1 and u2, v2, the displacement at any point
inside the element can be represented in terms of its interpolation functions and nodal displacements.
The shape function for a two node truss element is

[N ] = 1 − x x
(7.1.5)
 L L 
Equation expressed in eq. (7.1.5) is expressed in terms of the nodal displacement vectors {ui} and {vi}
with the help of interpolation function.
L L
1 P
=
Π  ui   N ' EA N 'ui  dx +  vi   N '  N 'vi  dx -  P1v1 + P2 v2 
T T T T
(7.1.6)
2 0 2 0
Where,
u  v 
=  N ui  =  N' ui  and =  N vi  =  N' vi  (7.1.7)
x x x x
Making potential energy (eq. 7.1.6) stationary one can find the equilibrium equation as
 3

L L

F  =   N ' EA N ' dx ui + P   N '  N ' dx vi 

T T
(7.1.8)
0 0

Or,
F  =  k A ui   P  kG vi  (7.1.9)
Where, [kA] is the axial stiffness of the member and [kG] is the geometric stiffness of the member in its
local coordinate system and can be derived as follows.
 1 
- 
L  1 1 AE  1 -1
L

 k A  =   N ' EA N ' dx = AE 0  L  -  dx =  

T
(7.1.10)
 1   L L  L  -1 1 
0
 
 L 

 1 
- 
L  1 1 1  1 -1
L

 kG  =   N '  N ' dx = 0  L  -
  dx =  
T
(7.1.11)
  
 
  -1 1 
0   1 L L L
 L 
However, in a generalised form to accommodate both the direction, these stiffness matrices can be
written as
 1 0 1 0  0 0 0 0 
   
AE  0 0 0 0 1 0 1 0 1
k A = and  kG  = (7.1.12)
L 1 0 1 0 L 0 0 0 0 
 0 0 0 0  0 1 0 1 
   

The generalised stiffness matrix of a plane truss member in global coordinate system can be derived
using the transformation matrix as derived module 4, lecture 1. The transformation matrix can be
recalled and written here as follows.
 cos θ sinθ 0 0 
 − sinθ cos θ 0 0 
[T ] = 

0 0 cos θ sinθ 
 
 0 0 − sinθ cos θ 
(7.1.13)
Thus, the stiffness matrices with respect to global coordinate system will become
 K A  = T   k A T  and  KG  = T   kG T 
T T
(7.1.14)
Here, [KA] and [KG] are the axial stiffness and geometric stiffness of the member in global coordinate
system. The force-displacement relationship in global coordinate system can be written from eq. (7.1.9)
as
 4

F  =  K A d   P  KG d    K A   P  KG  d  (7.1.15)

Where, {d} is the displacement vector in global coordinates. If the external force is absent in eq.
(7.1.15), the value of P will be considered to be undetermined as.
[ K A ] {d } + P [ K G ] {d } =
0 (7.1.16)
The above equation can be solved as an eigenvalue problem to calculate buckling load P.

7.1.3 Buckling of Beam-Column Members

Let consider a pin ended column under the action of compressive force P. The elastic and geometric
stiffness matrices can be developed from 1st principles for a beam-column element which can be used in
the linear elastic stability analysis of frameworks. Considering small deflection approximation to the
curvature, the total potential energy is computed from the following.
2 2
𝐸𝐼 𝐿 𝑑2 𝑤 𝑃 𝐿 𝑑𝑤 2 𝐿 𝐸𝐼 𝑑2 𝑤 𝑃 𝑑𝑤 2
𝛱= 2
∫0 � 𝑑𝑥 2 � 𝑑𝑥 + ∫ � � 𝑑𝑥
2 0 𝑑𝑥
= ∫0 � 2 � 𝑑𝑥 2 � + 2
� 𝑑𝑥 � � 𝑑𝑥 (7.1.17)

Here, w is the transverse displacement and I is the moment of inertia of the member.

Fig. 7.1.2 Beam-column member

Considering nodal displacements at nodes 1 and 2 as w1, θ1 and w2, θ2, the displacement at any point
inside the element can be represented in terms of its interpolation functions and nodal displacements.
𝑤1
𝜃
𝑤 = [𝑁1 𝑁2 𝑁3 𝑁4 ] �𝑤1 � = [𝑁]{𝑑} (7.1.18)
2
𝜃2
Thus, the above energy equation can be rewritten as
1 L   N "  EI  N "  d  + d  P  N '  T  N '  d  dx
2 0 
 
T T
=Π d    
(7.1.19)
     
Where,
 5

d 2w d 2 d wd
2
= 2  N d  =  N" d  and =  N d  =  N' d  (7.1.20)
dx dx dx dx
Applying the variational principle one can express

=
∂Π
∂ {d }
T
(
{F } = ∫0  N "  EI [ N "] + P [ N '] [ N '] dx {d }
l T
) (7.1.21)

Thus, the stiffness matrix will be obtained as follows which have two terms.
[k ] = (
∫0  N  EI [ N "]
l
" T
) dx +P∫ ([ N '] [ N ']) dx =
l

0
T
[k ] + P [k ] F G (7.1.22)

The first term resembles ordinary stiffness matrix for the bending of a beam. So this matrix is called
flexural stiffness matrix. The second matrix is known as geometric stiffness matrix as it only depends on
the geometrical parameters. Thus, the flexural stiffness matrix [kF] and geometric stiffness matrix [kG]
can be derived from the following expressions.
[ kF ] ∫=
0  ( T

0 )
 N "  EI [ N "] dx and [ kG ] ∫ ([ N '] [ N ']) dx
l l T
(7.1.23)

The above matrices can be derived from the assumed interpolation function. For example, the
interpolation functions for a two node beam element are expressed by the following equation.
𝑥2 𝑥3 𝑥2 𝑥3 𝑥2 𝑥3 𝑥2 𝑥3
[𝑁] = ��1 − 3 2 + 2 3 � , �𝑥 − 2 + 𝐿2 � , �3 𝐿2 − 2 𝐿3 � , �− + 𝐿2 �� (7.1.24)
𝐿 𝐿 𝐿 𝐿
Now, the first and second order derivative of the above function will become
𝑥 𝑥2 𝑥 𝑥2 𝑥 𝑥2 𝑥 𝑥2
[𝑁 ′ ] = ��−6 2 + 6 3 � , �1 − 4 + 3 2 � , �6 2 − 6 3 � , �−2 + 3 2 �� (7.1.25)
𝐿 𝐿 𝐿 𝐿 𝐿 𝐿 𝐿 𝐿
6 𝑥 4 𝑥 6 𝑥 2 𝑥
�𝑁 " � = ��− 𝐿2 + 12 𝐿3 � , �− 𝐿 + 6 𝐿2 � , �𝐿2 − 12 𝐿3 � , �− 𝐿 + 6 𝐿2 �� (7.1.26)
Using the above expressions in eq. (7.1.23), flexural and geometric stiffness matrices can be derived and
obtained as follows.
12 6𝐿 −12 6𝐿
2
𝐸𝐼
[𝐾𝐹 ] = 3 � 6𝐿 4𝐿 −6𝐿 2𝐿2 � (7.1.27)
𝐿 −12 −6𝐿 12 −6𝐿
6𝐿 2𝐿2 −6𝐿 4𝐿2

36 3𝐿 −36 3𝐿
2
[𝐾𝐺 ] =
1
� 3𝐿 4𝐿 −3𝐿 −𝐿2 � (7.1.28)
30𝐿 36 −3𝐿 36 −3𝐿
3𝐿 −𝐿2 −3𝐿 4𝐿2

Due to external forces {F}, one can find the displacement vectors from the following equation.
{F }
= [=
k ] {d } [ k F ] {d } + P [ k G ] { d } (7.1.29)
The above is the beam-column equation in finite element form. In the absence of transverse load, the
member will become column and P will be considered to be undetermined.
[ k F ] {d } + P [ kG ] { d } =
0 (7.1.30)
Denoting P as -λ we can reach to the familiar eigenvalue problem as given below.
 6

[ k F ] {d } = λ [ kG ] {d } (7.1.31)
Solving above equation, values of λ and associated nodal displacement vectors can be obtained.
Mathematically this can be expressed as
( [ k ] − λ [ k ] ) {d } =
F G 0 (7.1.32)
Thus,
[ k F ] − λ [ kG ] =
0 (7.1.33)
For the matrix of size n, one can find n+1 degree polynomial in λ. The smallest root of the above
equation will become the first approximate buckling load. From this value λ, one can find a set of ratios
for the nodal displacement components. From this, the first buckling mode shape can be calculated.
Higher mode approximations can also be found in a similar process. The procedure to determine the
critical load by the above method is illustrated in the following example.

Example 7.1.1
Consider a column with one end clamped and other end free as shown in Fig. 7.1.2.

Fig. 7.1.3 Column with one end fixed and other end free

Now, the finite element equation of the column considering single element will be

12 6𝐿 −12 6𝐿 36 3𝐿 −36 3𝐿
𝐸𝐼 6𝐿 4𝐿2 −6𝐿 2𝐿2 � − λ 1 � 3𝐿 4𝐿2 −3𝐿 −𝐿2 � = 0
�
𝐿3 −12 −6𝐿 12 −6𝐿 30𝐿 36 −3𝐿 36 −3𝐿
6𝐿 2𝐿2 −6𝐿 4𝐿2
3𝐿 −𝐿2 −3𝐿 4𝐿2
 7

The boundary conditions for this member are given by

dw
At x = 0, d1  0 and 1  0
dx
Thus, according to the above boundary conditions, the first and second rows as well as columns of the
equation are deleted and rewritten in the following form.
𝐸𝐼 12 −6𝐿 𝜆 36 −3𝐿
� 3� 2�− � �� = 0
𝐿 −6𝐿 4𝐿 30𝐿 −3𝐿 4𝐿2
Or
4𝐸𝐼 2𝜆𝐿 12𝐸𝐼 6𝜆 6𝐸𝐼 𝜆 2
� − � � 3 − � − �− 2 + � = 0
𝐿 15 𝐿 5𝐿 𝐿 10
Solving the above expression, the critical value of λcr and thus the critical value of force Pcr will become
as
𝐸𝐼
𝜆𝑐𝑟 = 𝑃𝑐𝑟 = 2.486 2
𝐿
It is important to note that the exact value for such clamped-free column is
2 EI 2 EI EI
Pcr    2.467
2L
2
Le 2 L2
The finite element result has slight deviation from the exact result. This difference can be minimized by
the increase of number of elements in the column as we know, more we subdivide the continuum, better
we can obtain the result close to the exact one.

7.1.4 Buckling of Plate Bending Elements

The elastic stability analysis of rectangular plates is discussed in this section. The total potential energy
for plate are expressed as
2
𝐷 𝑎 𝑏 𝜕2 𝑤 𝜕2 𝑤 𝜕2 𝑤
𝜋 = 2 ∫−𝑎 ∫−𝑏 �(∇2 w)2 + 2(1 − 𝜇) ��𝜕𝑥𝜕𝑦� − � 𝜕𝑥2 � � 𝜕𝑦 2 ��� 𝑑𝑥𝑑𝑦
1 𝑎 𝑏 𝜕𝑤 2 𝜕𝑤 𝜕𝑤 𝜕𝑤 2
− 2 ∫−𝑎 ∫−𝑏 �𝐹𝑥 � 𝜕𝑥 � + 2𝐹𝑥𝑦 𝜕𝑦 𝜕𝑥
+ 𝐹𝑦 � 𝜕𝑦 � � 𝑑𝑥𝑑𝑦 (7.1.34)
Here, Fx, Fy, and Fxy, are the in-plane edge load and compressive load is considered as positive. For,
finite element formulation the deflection in above expression needs to convert in terms of nodal
displacements in the element. Following the derivations in beam-column member (eqs. 7.1.18 to 7.1.22),
the flexural and geometric stiffness for the plate element can be derived. Thus, the above expression can
be derived to the following form using interpolation functions.
1 𝐹𝑥 𝐹𝑦 𝐹𝑥𝑦
𝜋 = 2 {𝑑}𝑇 [𝑘𝐹 ]{𝑑} − {𝑑}𝑇 [𝑘𝐺𝑥 ]{𝑑} − {𝑑}𝑇 �𝑘𝐾𝐺𝑦 �{𝑑} − {𝑑}𝑇 �𝑘𝐺𝑥𝑦 �{𝑑} (7.1.35)
2 2 2
Here [𝑘𝐹 ] is the flexural stiffness matrix. The other stiffness matrices are analogous to the geometric
stiffness matrices of the plate and can be expressed as
𝑘𝐺𝑥 = ∬⌊𝑁′𝑥 ⌋ {𝑁′𝑥 }𝑑𝑥 𝑑𝑦
 8

𝑘𝐺𝑦 = ∬�𝑁′𝑦 � �𝑁′𝑦 �𝑑𝑥 𝑑𝑦 (7.1.36)

𝑘𝐺𝑥𝑦 = ∬⌊𝑁′𝑥 ⌋ �𝑁′𝑦 �𝑑𝑥 𝑑𝑦
Where [𝑁′𝑥 ] and �𝑁′𝑦 � indicate partial derivative of [𝑁] with respect to x and y respectively. Thus, the
equation of buckling becomes
[𝑘𝐹 ]{𝑑} − 𝐹𝑥 [𝑘𝐺𝑥 ]{𝑑} − 𝐹𝑦 �𝑘𝐺𝑦 �{𝑑} − 𝐹𝑥𝑦 �𝑘𝐺𝑥𝑦 �{𝑑} = 0 (7.1.37)
If the in-plane loads have a constant ratio to each other at all time during their buildup, the above
equation can be expressed as follows
[𝑘𝐹 ]{𝑑} = 𝑃∗ �𝛼[𝑘𝐺𝑥 ] + 𝛽�𝑘𝐺𝑦 � + 𝛾�𝑘𝐺𝑥𝑦 ��{𝑑} (7.1.38)
The term 𝑃∗ is called the load factor, and 𝛼, 𝛽, and 𝛾 are constants relating the in-plane loads in the plate
member. Solving the above expression, the buckling mode shapes are possible to determine.
 9

Lecture 2: Finite Elements in Fluid Mechanics

7.2.1 Governing Fluid Equations

The fluid mechanics topic covers a wide range of problems of interest in engineering applications.
Basically fluid is a material that conforms to the shape of its container. Thus, both the liquids and gases
are considered as fluid. However, the physical behaviour of liquids and gases is very different. The
differences in behaviour lead to a variety of subfields in fluid mechanics. In case of constant density of
liquid, the flow is generally referred to as incompressible flow. The density of gases not constant and
therefore, their flow is compressible flow. The Navier–Stokes equations are the fundamental basis of
almost all fluid dynamics related problems. Any single-phase fluid flow can be defined by this
expression. The general form of motion of a two dimensional viscous Newtonian fluid may be expressed
as
1
  v j vi, j vi, ji  f i
p i , vi
 (7.2.1)

Here,
 = kinematic viscosity
 = mass density of fluid
vi = Velocity components
f i = Body forces
p = fluid pressure
The suffix ,j and ,ji are the derivatives along j and j & i direction respectively. The dot represents the
derivative with respect to time. Neglecting non linear convective terms, viscosity and body forces, eq.
(7.2.1) can be simplified as:
p,i  v i  0 (7.2.2)
Now, the continuity equation of the fluid is expressed by

p  c 2 v k,k  0 (7.2.3)
Here, c is the acoustic wave speed in fluid. In the above expression, two sets of variables, the velocity
and the pressure are used to describe the behaviour of fluid. Now it is possible to combine equation
(7.2.2) and (7.2.3) to obtain a single variable formulation. For the small amplitude of fluid motion, one
can assume

vi  u i (7.2.4)
Where u i is the displacement component of fluid. To obtain single variable formulation, eq. (7.2.4) may
be substituted into eq. (7.2.3) and one can get

p  c 2 u k,k  0 (7.2.5)
Integrating eq. (5) w. r. t. time we have
 10

p   c 2 u k,k (7.2.6)
Now differentiating the above expression w. r. t. xi following expression will be arrived:

p,i  c 2 u k,ki (7.2.7)

Substituting the above in to eq. (7.2.2) one can have

v i c 2 u k,ki  0 (7.2.8)

Thus, eq. (7.2.8) is expressed in terms of displacement variables only and known as displacement based
equation.

Similarly, it is possible to obtain the fluid equation in terms of pressure variable only. Differentiating eq.
(7.2.3) w. r. t. Time, the following expression can be obtained.

p  c 2 v k,k  0 (7.2.9)

Again, differentiating eq. (7.2.2) w. r. t. xi, we have
p,ii v i,i  0 (7.2.10)
From eqs. (7.2.9) and (7.2.10), the pressure based single variable expression can be arrived as given
below.


p  c 2 p,ii  0 (7.2.11)
The above expression is basically the Helmholtz wave equation for a compressible fluid having acoustic
speed c.
1
2 p  2

p0 (7.2.12)
c
Thus, the general form of fluid equations of 2D linear steady state problems can be expressed by the
Helmholtz equation. For incompressible fluid c becomes infinitely large. Hence for incompressible
fluid, eq. (7.2.12) can be written as
2 p  0 (7.2.13)
For the ideal, irrotational fluid flow problems, the field variables are the streamline, ψ and potential φ
functions which are governed by Laplace’s equations
2 2
 0
x 2 y 2
 2  2
 0
x 2 y 2 (7.2.14)
 11

Derivation of the above expression and many other related equations can be found in details in fluid
mechanics related text books.

7.2.2 Finite Element Formulation

The equation of motion of fluid can be expressed in various ways and some of those are shown in
previous section. Finite element form of those expressions can be derived using various methods
considering pressure, displacement, velocity, velocity potential, stream functions and their
combinations. Here, displacement and pressure based formulations will be derived using finite element
method.

7.2.2.1 Displacement based finite element formulation

Consider the equation (7.2.8) which can be expressed only in terms of displacement variables.

u i c 2 u k,ki  0 (7.2.15)

Here, u is the displacement vector. Now, the weak form of the above equation will become

 w u c u  d  0
2
i i k,ki (7.2.16)

Performing integration by parts on the second terms, one can arrive at the following expression:

w i u i d  w i c 2 u k,k d   w i,i c 2 u k,k d  0



o ,r w i u i d   w i,i c 2 u k,kid   w i c 2 u k,k d

(7.2.17)

  

Now from earlier relation (eq. 7.2.6) we have, p   c 2 u k,k . Thus, the above equation may be
written as:

w i u i d   w i,i c 2 u k,ki d   w i pd (7.2.18)

  

In case of fluid filled rigid tank, the weighting function wi must satisfy the condition w i n i  0 on its
rigid boundary. Therefore, the above eq. will become

 w u  w i i i,i c 2 u k,k d   w i n i p

(7.2.19)
 
p

For finite element implementation of the above expression, let consider the interpolation function as N
and u as the nodal displacement vector. Thus,
 12

u  Nu and w  Nw (7.2.20)
Now the divergence of the displacement vector can be expressed as:
u i,i  Lu  LNu  Bu (7.2.21)

Where L      is the differential operator. Thus eq. (7.2.19) may be written as

 x y 

wT   T  
 N Nu  B c Bu d  w  N np d
T 2 T T
(7.2.22)
 
p

Or,

 M u   K u   F (7.2.23)

Where,
 M      N   N  d
T

 K    c 2  B   B  d 
T
(7.2.24)


F    N  n p d
T


p

Using eq. (7.2.23), the displacements in fluid domain can be determined under external forces applying
proper boundary conditions.

7.2.2.2 Pressure based finite element formulation

The Helmholtz equation (7.2.12) for a compressible fluid in two dimension can be used to determine the
pressure distribution in the fluid domain using finite element technique.
1
p,ii  2 p  0 (7.2.25)
c
The weak form of the above expression can be written as
 1 
 w i p,ii  c2 p d  0 (7.2.26)


Now, performing integration by parts on the first term, the following expression can be obtained.
1
w i p,i d   w i,i p,i d  
c2
   0
w i pd (7.2.27)
  

Thus,
1
c2  w p d   w
i p d   w i p,i d 
i,i ,i (7.2.28)
 
 13

Assuming interpolation function as N and p as the nodal pressure vector, the pressure (p) at any point
can be written as: p  Np and similarly, w  Nw . The divergence of the pressure can be expressed
   . Again,
as: p, i  Lp  LNp  Bp , where, L    
 x y 
w i,i  Lw  LNw  Bw
p   Nw   Np
   w T N T Np
 (7.2.29)
w i 
T

Thus, eq. (7.2.28) will become:
1    w BT Bpd  w N T p d
T T T

c2  w N T Npd   n
(7.2.30)
  
Or,
 E p  G p  B (7.2.31)
Where,
1 T
E 
c2   N   N d

T
G     B  B d (7.2.32)

T p
B    N  d

n
Applying boundary conditions, eq. (7.2.31) can be solved to calculate the dynamic pressure developed in
the fluid under applied accelerations on the domain.

7.2.3 Finite Element Formulation of Infinite Reservoir

Let consider an infinite reservoir adjacent to a dam like structure. In such case, if the dam is vibrated, the
hydrodynamic pressure will be developed in the reservoir which can be calculated using above method.
For finite element analysis, it is necessary to truncate such infinite domain at a certain distance away
from structure to have a manageable computational domain. The reservoir has four sides (Fig. 7.2.1) and
as a result four types of boundary conditions need to be specified.
B  B1  B2  B3  B4 (7.2.33)
 14

Fig. 7.2.1 Reservoir and its boundary conditions

(i) At the free surface (Γ1)

Neglecting the effects of surface waves of the water, the boundary condition of the free surface may be
expressed as
p  x, H, t   0 (7.2.34)
Here, H is the depth of the reservoir. However, sometimes, the effect of surface waves of the water
needs to be considered at the free surface. This can be approximated by assuming the actual surface to
be at an elevation relative to the mean surface and the following linearised free surface condition may be
adopted.
1 ∂p
p + = 0 (7.2.35)
g ∂y
Thus, the above expression may be written in finite element form as
p 1
B1      R1 p (7.2.36)
n g
In which,
 R1     N   N d
T
(7.2.37)
1

(ii) At the dam-reservoir interface (Γ2)

At the dam-reservoir interface, the pressure should satisfy
∂p
(0, y, t) = ρ ae iω t (7.2.38)
∂n
 15

where aeiωt is the horizontal component of the ground acceleration in which, ω is the circular frequency
of vibration and i = − 1 , n is the outwardly directed normal to the elemental surface along the interface.
In case of vertical dam-reservoir interface ∂p ∂n may be written as ∂p ∂x as both will represent normal to the
element surface. For an inclined dam-reservoir interface ∂p ∂x cannot represent the normal to the element
surface. Therefore, to generalize the expressions ∂p ∂n is used in eq. (7.2.38). If {a} is the vector of nodal
accelerations of generalized coordinates, {B2} may be expressed as
B2     R 2 a (7.2.39)
where,
 R 2  =    N r  T  N d  d
T
(7.2.40)
2

Here, [T] is the transformation matrix for generalized accelerations of a point on the dam reservoir
interface and [Nd] is the matrix of shape functions of the dam used to interpolate the generalized
acceleration at any point on their interface in terms of generalized nodal accelerations of an element.

(iii) At the reservoir bed interface (Γ3)

At the interface between the reservoir and the elastic foundation below the reservoir, the accelerations
should not be specified as rigid foundation because they depend on the interaction between the reservoir
and the foundation. However, for the sake of simplicity, the reservoir bed can be assumed as rigid and
following boundary condition may be adopted.
p
B3    x, 0, t   0 (7.2.41)


(iv) At the truncation boundary (Γ4)

The specification of the far boundary condition is one of the most important features in the finite
element analysis of a semi-infinite or infinite reservoir. This is due to the fact that the developed
hydrodynamic pressure, which affects the response of the structure, is dependent on the truncation
boundary condition. The infinite fluid domain may truncated at a finite distance away from the structure
for finite element analysis satisfying Sommerfeld radiation boundary condition. Application of
Sommerfeld radiation condition at the truncation boundary leads to
p
B4   (L, y, t) = 0 (7.2.42)
x
Here, L represents the distance between the structure and the truncation boundary. Thus, the
hydrodynamic pressure developed on the dam-reservoir interface can be calculated under external
excitation by the use of finite element technique.
 16

Lecture 3: Dynamic Analysis

The finite element method is a powerful device for analyzing the dynamic response of structures as in
case of static analyses. The responses such as displacements, velocities, strains, stresses become time
dependent in dynamic analysis. The dynamic equation of motion of a structure can be derived from
Hamilton’s principle using Lagrange equation.

7.3.1 Hamilton’s Principle

Hamilton’s principle is a simple but powerful tool to derive discritized dynamic system equations. It can
be stated as “Of all the possible time histories of displacement the most accurate solution makes the
Lagrangian functional a minimum provided the following conditions are satisfied.”
• the essential or the kinematic boundary conditions,
• the compatibility equations, and
• the conditions at initial (t1) and final time (t2).
First condition of the above ensures that the displacement constraints are fulfilled. Second condition
makes sure that the displacements are continuous in the problem domain and the third condition
necessitates the displacement history to satisfy the constraints at the initial and final times. Hamilton’s
principle allows one to assume any set of displacements, as long as it satisfies the above three
conditions. It is basically a variational formulation method. Just like in the other variational
formulations, a functional is used here. Mathematically, Hamilton’s principle states:
t2

δ ∫ Ldt = 0 (7.3.1)
t1

The Langrangian functional, L, is defined as follows using a set of permissible time histories of
displacement.
L  T (7.3.2)
Where T is the kinetic energy and  is the potential energy which is basically elastic strain energy. The
kinetic energy, T is defined in the integral form as
1
T   x T x d (7.3.3)
2 
Here Ω represents the whole volume of the domain, ρ is the mass density of structure and x is the set of
time histories of velocities. The strain energy in the entire domain of elastic structure can be expressed
as
1 1
   e T s d   ee T
D d (7.3.4)
2  2 
Here ε are the strains obtained from the set of time histories of displacements. Here, L can be expressed
in terms of the generalized variables x1 , x 2  x n an d x 1 , x 2  x n . Here, xi is the displacement and
the velocity is expressed by
 17

d
x i  xi  (7.3.5)
dt
Considering xi as generalised displacement, the Lagrange’s equation of motion are expressed as
d  L  L
  0 where i  1 to n
dt  x  x
(7.3.6)
i i

7.3.2 Finite Element Form in MDOF System

Consider a two-degree of freedom system as shown in the figure below.

Fig. 7.3.1 Two Degree Freedom System

The kinetic and potential energy can be expressed by following form:

1 1
T  m1x 12  m2 x 22
2 2 (7.3.7)
  k1x12  k 2  x 2  x1 
1 1 2

2 2
Now, substituting the values in Lagrangean L, we can obtain
1 1 1 1 1
L  T   m1x 12  m 2 x 22  k1x12  k 2 x 22  k 2 x1x 2  k 2 x12 (7.3.8)
2 2 2 2 2
Thus,
d  L  L
  m1x1  k1x1  k 2 x 2  k 2 x1
dt  x 1  x1
 m1x1  k1x1  k 2  x 2  x1   0
(7.3.9)
d  L  L
  m 2 x 2  k 2 x 2  k 2 x1
dt  x 2  x 2
 m 2 x 2  k 2  x 2  x1   0
Above equations can be written in the matrix form as shown below:
 18

 
 m1 0  
x1  k1  k 2  k 2   x1  0
 0 m       (7.3.10)
 x
2  
   k k  x 2  
 0

  2 2 2 

Thus, for a multi degree freedom system, the above expression can be written as
 M x   K x  0 (7.3.11)
For a steady state condition, one can consider x = a sin t . Thus, x = -2 a sin t = -2 x . As a result,
the above equation can be re-written as
 K x  2  M x   0
 K   2  M  x   0
 
 2    M 1  K 
  (7.3.12)
Thus, for a single degree of freedom system, the fundamental frequency can be found as
K
 (7.3.13)
M

7.3.3 Solid Body with Distributed Mass

The velocity vector x can be expressed in terms of the elemental nodal velocities with the help of
interpolation functions as x  Nx . Here, x is the nodal velocity, and N is the interpolation function. Thus
the expression of kinetic energy can be re-written as
1 1
T   x T x d  x T  N T N d x (7.3.14)
2  2 
In the above equations, the mass matrix can be obtained as
1
M   N T N d (7.3.15)
2 
Volume integral has to be taken over the volume of the element to find the mass matrix of that element.

7.3.3.1 Mass matrix for truss element

In case of truss member, the degrees of freedom at each node is one because of its only axial
deformation. Thus for a two node truss element, the interpolation function will become
 N   1 
Here,   x / L and thus, dx  Ld . Hence, the mass matrix can be expressed as
 19

 l
 M     N T Nd  A 0 N T Ndx  AL 0 N T Nd


1 l  1  1

2

0  1 2 d

l
 AL    1 d  AL (7.3.16)
   
   
0

 13  2 3 
1

  1 1
    
  AL  3 6  AL  2 1
 AL  3 2 3

  2 3 3  1 1 6 1 2
   
 3  0  6 3 
 2 3
Thus, the consistent mass matrix will be
AL  2 1
M    (7.3.17)
6 1 2
Whereas the lumped mass matrix can be expressed as follows.
AL 1 0
M    (7.3.18)
2  0 1
Now, to obtain a generalised mass matrix of a two dimensional truss member, consider Fig. 7.3.2.

Fig. 7.3.2 Two Degree Freedom System

The displacement vector and the shape function for such case are expressed as


 u1 


 
 v1 
 N N2 0 
d    and  N    1
0
0
(7.3.19)

 u2  N1 0 N 2 

 


 2
 v 


Thus, the mass matrix can be derived as given below.

 20

1 0 
 
l 0 1   1 0  0
 M     N T Nd  AL 0  
  0 1 0 
d (7.3.20)
   0   
 0  

After performing integration, the mass matrix will be obtained as
 2 0 1 0
 
pAL  0 2 0 1
M  (7.3.21)
6  1 0 2 0
 0 1 0 2
 

The lumped mass matrix of two dimensional truss member will become

1 0 0 0
 
pAL 0 1 0 0

M= (7.3.22)
2 0 0 1 0
0 1
 0 0

7.3.3.2 Mass matrix for beam element

The beam element has two degrees of freedom at each node, one transverse displacement (v) and the
other one is an angle of rotation (𝜃). So, a beam with two nodes has a total of four degrees of freedom.
From earlier notes in module 4, lecture 3, the interpolation function of a beam is obtained as
3 2 2 3 2 2 x3
N1  1 x  x , N 2  x  x  2,
L2 L3 L L (7.3.23)
2 3 2 3
3x 2x x x
N 3  2  3 and N4    2
L L L L
The consistent mass matrix can be following earlier process.
l
 M     N T Nd  A 0  N   N dx
T

Substitution for the interpolation functions and performing the required integrations, one can arrive the
following mass matrix for the beam element.
156 22𝐿 54 −13𝐿
𝜌𝐴𝐿 22𝐿 4𝐿2 13𝐿 −3𝐿2
[M] = 420 � � (7.3.24)
54 13𝐿 156 −22𝐿
−13𝐿 −3𝐿2 −22𝐿 4𝐿2

Similarly, the lumped mass matrix of the beam element is

 21

1 0 0 0
𝜌𝐴𝐿 0 1 0 0
[M] = � � (7.3.25)
2 0 0 1 0
0 0 0 1

7.3.3.3 Mass matrix for plane frame element

The plane frame element has three degrees of freedom at each node: (i) axial displacement (u), (ii)
transverse displacement (v) and (iii) angle of rotation (𝜃). So it has a total of six degrees of freedom for
a two node plane frame element and its shape functions are given by
x 3 2 2 3 2 2 x3
N1  1 , N 2  1 x  x , N 3  x  x  2,
L L2 L3 L L (7.3.26)
2 3 2 3
x 3x 2x x x
N4  , N 5  2  3 and N6    2
L L L L L
Similar to earlier case, one can derive the consistent mass matrix of a two node plane frame member
which will become as given below.

1/3 0 0 1/6 0 0
⎡ ⎤
⎢ 13/35 11𝐿/210 0 9/70 −13𝐿/420⎥
⎢ 𝐿2 /105 0 13𝐿/420 −𝐿2 /140 ⎥
[M] =𝜌𝐴𝐿 ⎢ ⎥ (7.3.27)
1/3 0 0
⎢ ⎥
⎢𝑆𝑦𝑚. 13/35 −11𝐿/210⎥
⎣ 𝐿2 /105 ⎦

The lumped mass matrix of such two dimensional frame element will be
1 0 0 0 0 0
⎡0 1 0 0 0 0⎤
⎢ ⎥
𝜌𝐴𝐿 0 0 1 0 0 0
[M] = 2 ⎢ ⎥ (7.3.28)
⎢0 0 0 1 0 0⎥
⎢0 0 0 0 1 0⎥
⎣0 0 0 0 0 1⎦

7.3.3.4 Mass matrix for space frame element

In case of space frame, each node has six degrees of freedom: (i) three displacements and (ii) three
rotations. Therefore, for a two node space frame element there will be twelve degrees of freedom and
the consistent mass matrix of the element in the local xyz system can be derived as
 22

1 
 
3 
 
 13 
0 
 35 
 
0 13 
 0 
 35 
 
0 0 0
J 
 
 3A 
 11 L2 
0 0  L 0 
 210 105 
 
 11 L2 
0 L 0 0 0 
 210 105 

 M   AL  

1 0 0 0 0 0
1 
6 3 
 
 9 13 13 
0 0 0 0 L 0 
 70 420 35 
 
 9 13 13 
0 0 0  L 0 0 0 
 70 420 35 
 
0 J J 
 0 0 0 0 0 0 0 
 6A 3A 
 L2 L2 
0 13

11 
 0 L 0 0 0 0 L 0 
 420 140 210 105 
 L2 2 
 0  13 L  0 
11 L 
 0 0 0 L 0 0 0
105 
 420 140 210

(7.3.29)

In a similar process, one can derive the mass matrix of other types of structures such as two dimensional
plane stress/strain element, three dimensional solid element, plate bending element, shell element etc.

7.3.4 Time History Analysis

The dynamic equation of motion of a multi-degree freedom system can be written as
 M x  Cx    K x  F t  (7.3.30)
Where, [C] is the damping matrix and {F(t)} is the time dependent force vector. The element matrices of
the above equation can be expressed as follows.
 23

1
M     N    N  d
T

2 
1
C    N    N  d
T
(7.3.31)
2 
1
K     B  D Bd
T

2 
In a linear dynamic system, these values remain constant throughout the time history analysis. Damping
is a mechanism which dissipates energy, causing the amplitude of free vibration to decay with time. The
various types of damping which influences structural dynamical behaviour are viscous damping,
hysteresis damping, radiation damping etc. Viscous damping exerts force proportional to velocity, as
exhibited by the term. A formulation for this kind of problem was developed by Rayleigh. In this case,
the energy dissipated is proportional to frequency and to the square of amplitude. Viscous damping is
provided by surrounding gas or liquid or by the viscous damper attached to the structure. Radiation
damping refers to energy dissipation to a practically unbounded medium, such as soil that supports
structure. Among the various types of physical damping, viscous damping is easy represent
computationally in dynamic equations. Fortunately, damping in structural problems is usually small
enough regardless of its actual source. Its effect on structural response is modelled well by regarding it
as viscous. The Rayleigh damping defines the global damping matrix [C ] as a linear combination of the
global mass and stiffness matrices.
[C ] = α ′[M ] + β ′[K ] (7.3.32)
Here, α ′ and β ′ are the stiffness and mass proportional damping constants respectively. The
relationship between the fraction of critical damping ratio, ξ , α and β at frequency ω is given by
1 β
ξ =  αω +  (7.3.33)
2 ω
Damping constants α ′ and β ′ are determined by choosing the fractions of critical damping ( ξ1′ and ξ 2′ )
at two different frequencies (ω 1 and ω2) and solving simultaneously as follows
2 1 2 2  2 11 
 
22 12 
(7.3.34)
2 1 2 2 1 
 
22 12 
There are several numerical integration schemes are available to obtain the time history response of the
structural system. Amongst, the implicit type of methods such as the Newmark’s method is one of the
most popular one.