Modelling and Finite Element Analysis 16ME6DCMFE

Unit 1

Portions for CIE1

Fundamental concepts: Principles of Elasticity:

Equilibrium equations,
strain displacement relationships in matrix form
Generalised Hooke’s Law, Constitutive relationships(3D)
Constitutive relationships for plane stress, plane strain , Axi-symmetric
Potential energy and equilibrium,PMPE and numericals on Springs in combination subjected to
forces
Rayleigh-Ritz method applied to simple problems on axially loaded members,,with point loads and
distributed loads.
Gauss elimination method and
Gaussian quadrature-1pt,2pt and 3 pt formula.
commercial packages-preprocessor, solver and post processor .

1.Explain the significance of Equilibrium equations. Derive the force equilibrium conditions
for 2D state of stress. Write equilibrium equations for 3D.
2.Sketch the variation of stresses in a 3D differential element subjected to body forces.
Establish the equations of force equilibrium in differential form
3.Write down (do not derive) the equations of equilibrium in differential form in terms of
stresses and body forces for a three dimensional element volume with the help of a neat
sketch
4.Explain principal stress, maximum stress and Von Mises stress and bring out the importance of
these in the stress analysis.
5.The stress components at a point in a body are given by x=3xy2z+2x ,y=5xyz+2y
,z=3xy2z+2x ,xy=0, yz=xz=3xy 2 z+2xy. Determine whether these components of stress
satisfy the equilibrium equations or not at the point (1,-1,2).If not then determine the suitable
body force vector required at this point so that these stress components are in equilibrium.
6.Write strain displacement relations for 2D and for 3D in Cartesian coordinates .
7.Strain in a rod of length “L” fixed at one end and subjected to axial loading is given by
x=1+2x2.Find the tip displacement .
8.The displacement field for a body is given by u=(x2+y2)i+(3+z)j+(x2+2y)k.Find the rectangular
strain components at (3,1,-2)

Department of Mechanical Engineering, BMSCE , 2020 1

 Modelling and Finite Element Analysis 16ME6DCMFE
9. Explain plane stress and plane strain problems. Give examples and write stiffness and
compliance matrices
10. Explain Plane stress ,Plane strain and axisymetric problems with examples.write
stiffness matrix for the same

11. A.Identify the idealization of 2D problems shown. State the geometric and
corresponding stress/strain conditions and write stiffness matrices.

• 1.Thin Plate with Hole 2.Thin cantilever plate

 xy y
 xy
x

11b Stress Analysis of the prototype seatbelt component shown in the figure below is to be carried out
for a tensile load. Identify the idealization of 2D problem shown. State the geometric and corresponding
stress/strain conditions and write stiffness matrices.

Given 2 D member has in plane dimensions(x,y) very large compared to Out of plane dimension(z).Since

Department of Mechanical Engineering, BMSCE , 2020 2

 Modelling and Finite Element Analysis 16ME6DCMFE
the plate is thin in the z-direction, there can be little variation in the stress components through the thickness.
Thus they will be approximately zero throughout the entire domain. Under these assumptions, the stress field
can be taken as

 x   x ( x, y),  y   y ( x, y), xy   xy ( x, y)
 z   xz   yz  0, Also,  xz   yz  0,  z  0

   D 

OR

11c) Stress Analysis of the thick Aluminum tang shown in the figure below is to be carried
out. Identify the idealization of 2D problem. State the geometric and corresponding
stress/strain conditions and write stiffness matrices.

Given 2 D member has in plane dimensions(x,y) relatively less compared to Out of plane
dimension(z). If the body forces and tractions on lateral boundaries are independent of the z-
coordinate and have no z-component, then the deformation field can be taken in the reduced form
u  u ( x, y ) , v  v ( x, y ) , w  0
 z   xz   yz  0
Under these assumptions, the stress field can be taken as
 x   x ( x, y),  y   y ( x, y), xy   xy ( x, y)
 z   xz   yz  0, xz   yz  0, Also,  z  0

Department of Mechanical Engineering, BMSCE , 2020 3

Modelling and Finite Element Analysis 16ME6DCMFE
 
1   0 
Plane Strain    D , D 3 x 3 
E
  1  0 
1  1  2   1  2 
 0 0 
 2 
12. The state of strain at a point is given by x=0.0015, y=-0.0025, z = yz= xz = 0, xy=
-0.004.Determine the stress tensor at the point . E =208 Gpa. and =0.28.
13. Determine the strain tensor for the stress tensor at a point given by
210 140 00
[] = 140 -240 00 where E=208 GPa and =0.3
00 00 00
14. Explain Axi-symmetric problems. Give examples and write stiffness and compliance
matrices.
15. Explain with examples of bar and beam ,essential and Non essential boundary
conditions
16. Explain boundary value problems and Initial value problems with example.
17. Write admissible functions for
i. Axial bar fixed at one end ii. Axial bar fixed at both ends
iii. Cantilever beam , iv Simply supported beam(Polynomial and trigonometric)
iv.Beam fixed at both ends
v.Propped cantilever beam
18. Using matrix notation, develop an expression for the total potential energy functional
for a 3D elastic solid subjected to body forces, surface forces and point loads.
19. State principle of Minimum Potential Energy.
20. Obtain displacements of the nodes in the spring system problems below using PMPE

21. Explain Rayleigh-Ritz method applied to continuum. What are its disadvantages
22. Obtain expression for displacement for uniform bar fixed at one end and loaded by a
point load at free end in tension using Rayleigh Ritz method .

Department of Mechanical Engineering, BMSCE , 2020 4

 Modelling and Finite Element Analysis 16ME6DCMFE
23. Obtain expression for displacement for uniform bar fixed at one end and loaded by a
point load at free end in compression using Rayleigh Ritz method .
24. Obtain expression for displacement for uniform bar under uniformly distributed
axial load of intensity q=C .where C is a constant using Rayleigh Ritz method .Use
second order polynomial
25. Using Rayleigh Ritz method obtain expressions for displacement and stress for
uniform bar shown in figure. Normalize the values if P=A=L=E=1

26. Use Rayleigh Ritz method to find displacement of any point of the rod shown.Also
determine the displacement of the midpoint.

27. Use Rayleigh Ritz method to find displacement of any point of the rod shown above
when the bottom is not fixed. Also determine the displacement of the free end .
28. Demonstrate the importance of convergence using first order and second order
polynomial in Rayleigh Ritz method to obtain expressions for displacement and stress
for uniform bar fixed at one end x=0 and subjected to linearly varying distributed axial
load of intensity q = Cx, where C is a constant.
29. Explain preprocessor,solver/processor and post processor applied to commercial
finite element package.

1. Give the principle involved in deriving Gauss quadrature formula for a simple function
F( r )=1 +2r+3r2+4r3, F( r) integrated between –1 to +1 06
2. Explain 1pt and 2 pt. Gaussian quadrature method
3. Evaluate the integral I =  (1   2 )(4   2 )dd using two point formula.
11
1 1
4. Evaluate I =
1  (3e x  x 2 
( x  2)
)dx using one point and two point Gauss Quadrature.

5. Using Gaussian Quadrature formula, evaluate the following

1 dx
  (
1
(1  2r  3r 2  4r 3 )dr   2   3 )dd
2
i. 11. iii.
1 1 x 11

Department of Mechanical Engineering, BMSCE , 2020 5

 Modelling and Finite Element Analysis 16ME6DCMFE
6. Using Gaussian Quadrature formula, evaluate the problems when integration limits are
not -1 to +1 like


3
(1  2r  3r 2  4r 3 )dr
0

7. Explain steps involved/algorithm in Gauss elimination method

8. Numericals on Gauss elimination method

21.Solve the following system of equations by Gaussian elimination method

x+y+z=9
x - 2y + 3z =8
2x + y- z = 3

1. Theory of Elasticity 1.1 INTRODUCTION:

Theory of elasticity deals with the stress and displacements in elastic solids generated by
external forces. Theory of elasticity is concerned with evaluation of 15 Unknowns: 6
Stresses,6 Strain and 3 displacements and for this one needs to understand the relevant 15
equations i.e Equilibrium equations (3),Strain-displacement equations (6) and
Constitutive equations (6)

Value of Elasticity Theory (SADD)

 Develops “Exact” Analytical Solutions For Problems of Limited Complexity
 Provides Framework for Understanding Limitations of Strength of Materials Models
 Establishes Framework for Developing Linear Finite Element Modeling
 Generates Solutions for Benchmark Comparisons with FEA Solutions
In elementary mechanics of materials (the first undergraduate course in solid mechanics)
as well as in an introductory Theory of elasticity course, five fundamental assumptions
are normally made about the characteristics of the materials for which the analysis is
valid.

Department of Mechanical Engineering, BMSCE , 2020 6

Modelling and Finite Element Analysis 16ME6DCMFE
These assumptions require the material to be,
1.2 Assumptions
 Continuum: The body is continuous, so displacements, Strains and stresses, can be
expressed by continuous functions in space.
 Homogeneous: The body is homogeneous, i.e., the elastic properties are the same
throughout the body. Elastic constants will be independent of the location in the body.
 Isotropic: The body is isotropic so that the elastic properties are the same in all
directions. Thus the elastic constants will be independent of the orientation of coordinate
axes.
 The two independent elastic E constants are E → Young’s modulus
G
 G → Shear modulus 2(1  )

Linear Elastic: The body is perfectly elastic

Obeys Hook's law of elasticity i.e linear relations between stress components and strain
components. = E. ,E → Young’s modulus  = G., G → Shear modulus
F(
) F ()

Elastic(Non linear) Linear Elastic

D(
D(
)

The displacements and strains are small:

• The displacements components of all points of the body during deformation are very
small compared with its original dimensions
• The equations of Elasticity are considerably simplified.
• Also be called small deformation theory, small displacement theory, or small
displacement-gradient theory.

Department of Mechanical Engineering, BMSCE , 2020 7

Modelling and Finite Element Analysis 16ME6DCMFE
1.3 Types of Forces

fx 
1.BodyForce
f 3x1  fy fz T Body Forces
 Acts on the volume(mass) of the body.
 Dimension is Force/Volume ,
 Examples :gravitational force ,Inertia
forces (in motion), Magnetic force.
Component of in X, Y, Z directions are

2. Surface force T3 x 1 Cantilever Beam Under Self-Weight Loading

 Surface force(often termed surface Traction) : Acts on the surface of the body.
 Dimension is force/Area, e.x., N/m2
 Example:Contact forces ,Aerodynamic
pressures,friction hydrostatic pressure. T
   T T T 
 Component of T in X, Y, Z directions are T  x z
 
3 x1
y

Forces: 3. Point load

P 
i 3x1

 Point load (often termed Concentrated Load ) : Idealised as acting at a point on the body.
 Dimension is force, e.x., N.
T
 Component of P in X, Y, Z directions are Pi 3 x1  Px Py Pz 
 

Department of Mechanical Engineering, BMSCE , 2020 8

Modelling and Finite Element Analysis 16ME6DCMFE
1.4 Stress at a point

F
Stress at a point: definition  R  lim
A0 A
Fn
Normal stress (normal component)  nn  lim
A0 A

Fs
Shear stress (parallel component)   lim  2 R   2 nn   2
A0 A

1.5 Positive Stress system

 Stress is a second order tensor.
 It needs two subscript
 The first subscript indicates the plane on which it
acts by its normal and
 the second subscript its direction

 xx Direction

Plane(normal)
 xy   xy  12 x-plane, y-direction
Stress acting on an element cube

 xx  xy  xz 
 
ij

  yx  yy  yz 

 zx  zy  zz 


Department of Mechanical Engineering, BMSCE , 2020 9

Modelling and Finite Element Analysis 16ME6DCMFE
In contracted notation, Stress Is written in the form of a vector as
 6 X 1   xx  yy  zz  xy  yx  zx T
Development of stress concept
Sl.No Diagrams Components Definition
1.Force Force components

Push or Pull

2.Traction Force per unit area

On a surface of
specified
orientation.
(Force intensity)
3.Surface Equilibrium SurfaceStress A pair of equal and
Stress components opposite traction
on a surface of
specified
orientation

4. Stress at Stress Surface stresses on

point components planes of all
orientations
through a point.

A Similarities between stresses and tractions

• 1 Same dimensions (force per unit area)
• 2 The normal stress acting on a plane matches the normal traction
• B Differences between stresses and tractions
• 1 Stresses are tensor quantities and tractions are vectors.
• 2 The stress state is defined at a point using a fixed reference frame, whereas a
traction is defined on a plane with a reference frame that floats with the plane.

Department of Mechanical Engineering, BMSCE , 2020 10

Modelling and Finite Element Analysis 16ME6DCMFE
1.6 Equilibrium equations
• The state of stress varies from point to point in a loaded member in general.
• This variation is governed by the condition that each and every differential element
considered should satisfy the conditions of equilibrium
Force equilibrium: Fx  0, Fy  0 Fz  0

• Moment equilibrium: M x  0, M y  0 M z  0

1.6.1 : 2 D equilibrium equations

• Consider a 2D element of size dx,dy from a loaded 2D structure.

• The element is shown under positive stress system
Let the body forces(expresses in terms of force /volume be
T
f 2 x1  f x f y 
 
Equilibrium Equations 2D
 y
y  dy
y  yx
 yx  dy  x  yx
y F x 0 
x

y
 fx  0
 xy  
fy  xy  dx  Fy  0  xxy  yy  f y  0
x
x  x M 0   xy   yx
fx x  dx z

 xy x

 yx

y

Department of Mechanical Engineering, BMSCE , 2020 11

Modelling and Finite Element Analysis 16ME6DCMFE
1.6.2 Equilibrium Equations 3D
 Consider A Three dimensional Body Occupying Volume V ,having surface S
 Constrained on some region (Displacement B.C) and with Traction B.C
and subjected to Body forces
Consider the equilibrium of an element volume to obtain the 3 force equilibrium
equations of elasticity

 x  yx  zx
Fx  0,    fx  0
x y z
 xy  y  zy
Fy  0,    fy  0
x y z
  yz  z
FZ  0, xz    fz  0
x y z

Moment Equilibriu m
M x  0 , yz   zy ,  M y  0, xz   zx ,  M z  0 , xy   yx
Stress tensor is symmetric

Department of Mechanical Engineering, BMSCE , 2020 12

Modelling and Finite Element Analysis 16ME6DCMFE

1.7 Displacement:

• Pattern of Deformation

1.Rigid Body Motion

Zero Strains
Two-Dimensional Example

Rigid body Rotation Rigid body Translation

Displacement vector
Displacement of a material point P
inside a body, before and after the
deformation

• Initial position of the material

points of the body is described by the

Department of Mechanical Engineering, BMSCE , 2020 13

Modelling and Finite Element Analysis 16ME6DCMFE
coordinates x, y, z of the generic point P

• Its displacement is defined by vector PP’ with components u,v,w in the reference
directions x, y, z, respectively.

• The position of the point after the deformation is therefore given by the coordinates x + u,
y + v, z + w.
• If the material is continuous before and after the deformation, the functions u(x, y, x),
v(x, y, z), w(x, y, z) are continuous functions of the position coordinates of the body
before the deformation, x, y, z.

• Let Intial position be

• P (x, y,z) and displaced position be , P' (x', y',z',).
• Displacement vector PP' is denoted by u.
• The displacement vector u has components Ux=u, Uy=v
and Uz=w along the x, y and z axes respectively, and can
be expressed as, u  iu  jv  kw

2. Deformation

1.
1

1
Biaxial stretch 1.

Infinitesimal Strain (Small Strains)

Department of Mechanical Engineering, BMSCE , 2020 14

Modelling and Finite Element Analysis 16ME6DCMFE
1.The Engineering Normal Strain
• Expressed as the change in length ΔL per unit of the original length Lo of the line element
or fibers.
• If Lo is the original length and L is the final length ,then
L  Lo L
 
Lo Lo
• Measures of strain are often expressed in parts per million or microstrains.

2. Shear Strain ()

• Shear strain measures changes in angles in terms of radians with respect to two specific
directions initially perpendicular to each other.

1.8 Strain Displacement Relation: 2D

• Consider deformation of a the two-dimensional rectangular element with original

dimensions dx by dy represented by ABCD.
• After deformation, the element takes a rhombus form as A’B’C’D’
• A1B11C11D11 shown in the dotted outline corresponds to rigid body displacement .

Department of Mechanical Engineering, BMSCE , 2020 15

Modelling and Finite Element Analysis 16ME6DCMFE
• Reference point A is taken at location (x,y), and the displacement components of this
point are thus u(x,y) and v(x,y).
u v
• The corresponding displacements of point B are u  dx and v  dx
x x
u v
• The corresponding displacements of point D are u  dy and v  dy
y y

Longitudinal Strain  x
A1B1  AB
x  , A1B1  AB (1   x )  dx(1   x )
AB
u
Horizontal projection of A1 B1  dx  dx
x
v
Vertical projection = B11 B1  dx
x
A B   dx(1   )
1 1 2
x
2

u   v 
2 2

  dx  dx    dx 
 x   x 
 u  u  v 
2 2

 (1   x2  2 x )  1     2   
 x  x  x 
 Assuming small deformations and strains and neglecting product of Smaller
quantities
u
x 
x

Similarly Longitudinal strain in Y direction

A1D1  AD v
y   y 
AD y
Shear Strain 
Shear strain measures changes in angles in terms of radians with respect to two specific
 xy  B11 A1 B1  D11 A1 D1
 1   2
 v u 
 dx  dx 
directions initially perpendicular to each other. x y 
 xy 
dx
 v u 
 xy   
 x y 

Department of Mechanical Engineering, BMSCE , 2020 16

Modelling and Finite Element Analysis 16ME6DCMFE
u v  v u 
x  , y   xy    
x y  x y 

Extending the logic to 3D

u v w
x  , y  , z 
x y z
 v u   w v 
 xy    ,  yz    
 x y   y z 
 w u 
 xz   
 x z 

1.9 Generalised Hooke’s Law

• The relationships between the stress and strain components are termed Constitutive
equations.
• The Constitutive equations is based on Experimental observations and established
principles.

Hooke’s Law
Robert Hooke (1635-1703) :Established tension is proportional to the stretch
• Hooke’s law established the notion of (linear) elasticity, but not yet in a way that was
expressible in terms of stress and strain.

 For Linear Elastic material ,Stress is proportional to Strain and vica-versa

 is   or   E , E is the constant of Proportionality
 is   or   G , G is the constant of Proportionality
 The relation between the normal stress σ and the longitudinal strain in the same
direction ε is called the longitudinal modulus of elasticity or Young’s modulus E of the
material.
 The relation between the Shear stress and the Shear strain is called the Shear modulus of
elasticity or Modulus of Rigidity G of the material.

Department of Mechanical Engineering, BMSCE , 2020 17

Modelling and Finite Element Analysis 16ME6DCMFE

Generalised Hooke’s Law

For Linear Elastic material “Each of the six Stress components may be expressed as a
linear function of the six strain components and Vica versa”.i.e  6 x1  C 6 x6  6 x1
 x  C11 C12 C13 C14 C15 C16    x    x   S11 S12 S13 S14 S15 S16   x 
  C C22 C23 C24 C25 C26    y    y  S 21 S 22 S 23 S 24 S 25 S 26   y 
 y   21
 z  C31 C32 C33 C34 C35 C36    z    z   S31 S32 S33 S34 S35 S36   z 
         
 xy  C41 C42 C43 C44 C45 C46   xy   xy  S 41 S 42 S 43 S 44 S 45 S 46   xy 
 yz  C51 C52 C53 C54 C55 C56   yz   yz   S51 S52 S53 S54 S55 S56   yz 
         
 xz  C61 C62 C63 C64 C65 C66   xz   xz   S61 S62 S 63 S64 S65 S66   xz 

Hooke's Law in Compliance Form

Similarly  6 x1  S 6 x 6  6 x1

 Note that the stiffness matrix is traditionally represented by the symbol C OR D, while S
is reserved for the compliance matrix!
 The generalized Hooke’s law is an assumption, which is reasonably accurate for many
material subjected to small strain, for a given temperature, time.
 The 36 coefficients C11 to C66 are called elastic coefficients

Hooke's Law in Stiffness Form

 For Isotropic material the number of independent elastic constants reduces
two 2.i.e Young's modulus E and the Poisson's ratio v.
   D 
1    0 0 0 
  1   0 0 0 

   1  0 0 0 
 1  2 
D  C   E
 0 0 0 0 0 
(1  )(1  2 )  2 
1  2
 0 0 0 0 0 
 2 
 1  2 
 0 0 0 0 0
2 

Department of Mechanical Engineering, BMSCE , 2020 18

Modelling and Finite Element Analysis 16ME6DCMFE
Hooke's Law in Compliance Form    S  
1   0 0 0 
  1  0 0 0 
 
1    1 0 0 0 
S     Note :   G , G 
E
E0 0 0 21   0 0  21  
0 0 0 0 21   0 
 
 0 0 0 0 0 21  

1.10 : 2 D Problems
 By virtue of Geometry, Loading and Material Property ,certain class of problems can be
reduced in dimension from 3D to 2 D and sometimes 1 D without much loss of accuracy.
 This saves considerable memory space and computational time.
Two vs Three Dimensional Problems(Sadd)

2D elastic problems
• PLANE STRESS
• PLANE STRAIN
• The basic theories of plane strain and plane stress represent the fundamental plane
problem in elasticity.

Plane Stress Problems

• Consider a 2 D member
whose in plane dimensions(x,y)
are very large compared to
Out of plane dimension(z)
• The domain is bounded two stress free planes z
= h,

Department of Mechanical Engineering, BMSCE , 2020 19

Modelling and Finite Element Analysis 16ME6DCMFE
• Since the plate is thin in the z-direction, there can be little variation in the stress
components through the thickness.
• Thus they will be approximately zero throughout the entire domain.  z   xz   yz  0
• Under these assumptions, the stress field can be taken as
 x   x ( x, y),  y   y ( x, y), xy   xy ( x, y)
 z   xz   yz  0, Also,  xz   yz  0,  z  0

Plane Stress :Examples

• 1.Thin Plate with Hole 2.Thin cantilever plate
•

 xy y
 xy
x

Rotating disc/Flywheel
Plane Stress    D 

 x   
1  0   x  
 
 y  
E
 1  
0   y  z    y
1 
x

  1   1    
2

 xy  0 0   xy
 2  

[D] matrix for the plane stress case is

 
E 1  0 
D3 x3   1 0 
1  2  1  
0 0 
 2 

Department of Mechanical Engineering, BMSCE , 2020 20

Modelling and Finite Element Analysis 16ME6DCMFE
Plane Strain Problems
• Consider a 2 D member whose in plane dimensions(x,y) are very small compared to out
of plane dimension(z)

• If the body forces and tractions on lateral boundaries are independent of the z-coordinate
and have no z-component, then the deformation field can be taken in the reduced form
u  u ( x, y ) , v  v ( x, y ) , w  0
 z   xz   yz  0
• Under these assumptions, the stress field can be taken as
 x   x ( x, y),  y   y ( x, y), xy   xy ( x, y)
 z   xz   yz  0, xz   yz  0, Also,  z  0
Plane Strain :Examples
1.Wall of a Dam
2.Strip footing
3. Rotating Shaft/
cylinder
4. Long cylindrical pressure vessel subjected
to internal/external pressure
Plane Strain    D 

 x 
  E

1   
0   x 
 
 z    x   y 
 y     1  0   y 
  1  1  2   0 1  2   
 xy     xy
2  
0

[D] Matrix for the plane strain case is
 
1   0 
D3 x3 
E
  1  0 
1  1  2   1  2 
 0 0 
 2 

Department of Mechanical Engineering, BMSCE , 2020 21

Modelling and Finite Element Analysis 16ME6DCMFE

Sl.No Plane Stress Plane Strain

Dimensions In plane dimensions are Out of plane dimensions are very

very large compared to out large compared to in plane
of plane dimensions dimensions
Stress Out of Plane Stresses are
zero  z   xz   yz  0
 z    x   y 
 xz   yz  0

Strain  xz   yz  0 Out of Plane Strains are Zero

  z   xz   yz  0
z    y
1 
x

Department of Mechanical Engineering, BMSCE , 2020 22

Modelling and Finite Element Analysis 16ME6DCMFE
Axisymmetric Analysis

Problems involving three- dimensional axisymmetric solids or solids of revolution,

subjected to axisymmetric loading, reduce to simple two dimensional problems. Because
of total symmetry about the Y axis, as seen in the fig. all deformations and stresses are
independent of the rotational angle θ. Thus the problem needs to be looked at as a two
dimensional problem in XY, define on the revolving area. Examples: Pressurized
cylinders, cooling towers and revolving bodies like Disk type Flywheel, Shafts.

Because of symmetry about the z axis, the stresses are independent of the  coordinate.
Therefore, all derivatives with respect to  vanish, and the displacement component v

Department of Mechanical Engineering, BMSCE , 2020 23

Modelling and Finite Element Analysis 16ME6DCMFE
(tangent to the  direction), the shear strains r and z and the shear stresses r and z
are all zero.

Boundary Conditions and Types

Given the geometry of the body, the manner in which it is locates, i,e., specified boundary
conditions, the distribution of body forces, and the elastic constants, it is required to determine
stresses, strains and displacements as a function of location.
Boundary conditions play a very essential role in properly formulating and solving elasticity
problems, it is important to acquire a clear understanding of their specification and use.Improper
specification results in either no solution or a solution to a different problem than what was
originally sought.
The common types of boundary conditions for elasticity applications normally include
specification of how the body is being supported or loaded.

 This concept is mathematically formulated by specifying either the displacements or tractions

at boundary.Depending upon the boundary conditions, problems can be classified as
1. Traction conditions 2. Displacements conditions 3. Mixed conditions

(n)
T
S S St
Su
R
R R

u
Traction Conditions Displacement Conditions Mixed Conditions

Department of Mechanical Engineering, BMSCE , 2020 24

Modelling and Finite Element Analysis 16ME6DCMFE

FIGURE: Examples boundary conditions.

Boundary conditions are normally specified using the coordinate system describing the problem, and thus
particular components of the displacements and tractions are set equal to prescribed values.

For displacement-type conditions, such a specification is straightforward, and a common example includes
fixed boundaries where the displacements are to be zero.
For traction boundary conditions, the specification can be a bit more complex.
Traction conditions :Traction: Distributed force per unit area

p x 
 
T T S  p y 
p 
nz  z
S p
n z
p
n y
n p
S y
x x
n x 
 
n  n y 
If the unit outward normal to ST
n 
 z
p x   x nx   xy n y   xz nz
p y   xy nx   y n y   yz nz
Then
p z   xz nx   zy n y   z nz
Department of Mechanical Engineering, BMSCE , 2020 25
Modelling and Finite Element Analysis 16ME6DCMFE

2.1 Boundary Value Problems

• The objective of most analyses is to determine functions, called dependent variable.
• They are governed by a set of Differential equation posed in a domain  and satisfy some
boundary conditions on the boundary 
• Governing differential equations plus boundary conditions are set to state a problem in
Strong Form.
• Whatever functions that define these field variables have to be differentiable up to the
order of the partial differential equations that exist in the strong form of the system
equations.
• Obtaining the exact solution for a strong form of the system equation is usually very
difficult for practical engineering problems.
• A weak form of the system equations is usually created using one of the following widely
used methods:
 Energy principles (see, e.g. Reddy, 1984)
 Weighted residual methods (see, e.g. Ziekiewicz and Taylor, 2000)
• A integral expression such as Functional that implicitly contains the differential equations
is called the weak form.
• The energy principle can be categorized as a special form of the variational principle
which is particularly suited for problems of the mechanics of solids and structures.
• The Strong form states conditions that must be met at every material Point, wheras the
weak form states the condition in an average sense
• Both are valid statements of the problem.
• In Functional approach, the Functional is extremised.
 This satisfies 1.Differential equation and
 2. Certain boundary conditions called Natural/Nonessential B.C
• A functional, such as Potential energy  contains integrals that span the entire domain.

A: Area of Cross section , U(x) : Axial displacement,E: Young’s

2.1a BAR :
Modulus, AE : Axial Rigidity

du du
 ( x )  E ( x )  E , x 
dx dx
du
Internal Axial force p  A  EA
Department of Mechanical Engineering, BMSCE , 2020 dx 26
q : Distribute d force
P : Prescribed external load
Modelling and Finite Element Analysis 16ME6DCMFE
Primary Variable: Quantities which has to be Continuous in the domain . Secondary variables
need not be Continuous

For Bar problem, The primary variable is displacement u which has to be continuous.
Otherwise, there will be opening or overlap. The secondary variable is strain i.e du/dx.It can be
discontinuous. In the stepped bar shown, stress is discontinuous at change of cross section and
hence strain also.

2.2 Boundary conditions

Boundary conditions can be classified into two types based on Variational approach

1. Essential boundary conditions:

2. Natural boundary conditions

Essential boundary conditions are associated with Primary Variable and Natural/Non-essential
boundary conditions are associated with Secondary Variable.

2.2a :Boundary conditions :Bar

• Thus for Bar, essential boundary are associated with u and natural boundary
conditions are associated with du/dx.

•
• In the fig. shown Essential B.C is u=0 @x=0(Prescribed displacement)

• Non essential B.C is du/dx=P/(AE) @x=L (prescribed force/derivative of

displacement

Department of Mechanical Engineering, BMSCE , 2020 27

 Modelling and Finite Element Analysis 16ME6DCMFE
2.2b: Boundary conditions :Beam

Beams are slender members used for supporting transverse loading Loads could be concentrated
loads, distributed loads and moments.I: Moment of Inertia about N.A. , v : Transverse
displacement,E: Young’s Modulus, EI : Flexural Rigidity

v : Transverse Displaceme nt
dv
: Slope or rotation about z axis
dx
d2v d2v M
Bending moment M  EI  
dx 2 dx 2 EI
dM d3v d3v V
Shear Force V   EI  
dx dx 3 dx 3 EI

Governing Differenti al equation

d4v
 q( x )  0 0  x  L,
dx 4
q is the rate of loading

• The primary variables( which has to be Continuous) are Transverse displacement v and
rotation dv/dx and The secondary Variables (Which can be discontinuous) are

d 2v d 2v M
Bending moment M  EI 2
 2

dx dx EI
3 3
dM d v d v V
Shear Force V   EI 3
 3

dx dx dx EI

Department of Mechanical Engineering, BMSCE , 2020 28

Modelling and Finite Element Analysis 16ME6DCMFE
 Thus for Beam,

• Essential boundary are associated with v and slope dv/dx

• Natural boundary conditions are associated with

d2v M d2v
2
 as Bending moment M  EI 2
dx EI dx
d3v V dM d3v
3
 as Shear Force V   EI 3
dx EI dx dx

For the Cantilever shown In the figure Essential boundary are At X=0, v=0 and Slope
dv/dx=0
d2v M
Natural boundary conditions are at x=L
2
 0
dx EI
d3v P

dx 3 EI
 In Solid/Structural mechanics,

• Essential B.C are also known as Geometric B.C(Essential to be satisfied)

• Non Essential B.C are known as Force/dynamic B.C.(Automatically

satisfied in Functional approach)

 Essential B.C are also known Dirichlet B.C

 Non Essential B.C are known as Neumann B.C

Department of Mechanical Engineering, BMSCE , 2020 29

Modelling and Finite Element Analysis 16ME6DCMFE

BAR Beam

d du  d4v
Differential equation EA ( x )  q( x )  0  q( x )  0
dx  dx  dx 4
Order of D.E 2m=2 2m=4,m=2

2m m=1

Primary variable Axial Displacement u Transverse Displacement v

&
Order m-1
Slope/rotation
dv/dx

Secondary Variable Strain du/dx

m to 2m-1 Or Stress E. du/dx

Essential /Geometric Associated with u

B.C m-1 =0

Order m-1or less

Natural/Force/ Dynamic . Associated with d2v M d3v V

2
 , 3
Order m to 2m-1 du/dx dx EI dx EI
order 2 and 3
m=1

Department of Mechanical Engineering, BMSCE , 2020 30

Modelling and Finite Element Analysis 16ME6DCMFE
Functional :Potential Energy
• Functional is an integral expression that implicitly contains the Governing differential
equation for a particular problem.
• For structures problem Potential Energy  is a functional.
• Potential Energy has two components
1.U,Strain energy due to internal forces/Stresses
2. ,Work potential due to applied loads

2.3a Potential energy of a single spring

SEP , U 
0  F  
F

k 2
as Stiffness k 
F
2 2 2 
Work Potential    F
Potential Energy   U  
k 2
  F
2
2.3b Potential Energy of an Elastic Body
Consider a linear elastic three dimensional body occupying volume v, with body forces and
having surface s subjected to traction

SEP U

Consider a small element of Volume dV. Let the state

of stress be
 6 X 1   xx  yy  zz  xy  yz  zx T
and the corresponding strain be
 6 X 1   xx  yy  zz  xy  yz  zx T
Strain energy in the element of volume dV (Area under stress
Strain diagram).dV

Department of Mechanical Engineering, BMSCE , 2020 31

Modelling and Finite Element Analysis 16ME6DCMFE

dU 
1
 xx xx   yy yy   zz zz   xy xy   yz yz   zx zx dV
2
1
  1Tx 6  6 x1 dV
2
1 
SEP U     1 x 6  6 x1 dV
T
Total (1A)
V2 

Work Potential  due to B.F

Let body force vector acting on elemental volume dv be
 f 3 X 1  f x f
y
f
z
T
and the correspond ing displaceme ntvector beu3 X 1  u v wT

Work potential d   f .u  f .v  f .w dw  uT  f 3 x1 dv

 x
BF y z  1x 3
Total Work potential due to BF
    uT  f 3 x1 dv (1B )
BF 1x 3
V
Work Potential  due to Traction
Let Traction force vector acting on elemental surface ds be T 3 X 1  T x  Ty Tz T
beu3 X 1  u v w
T
and the correspond ing displaceme ntvector

Work potential d T   T x .u  T y .v  Tz .w ds   u1 x 3 T 3 x1 ds
T

Total Work potential due to Traction

 T    u1 x 3 T 3 x1 ds
T
(1C )
Work Potential  due to Point Loads
S

Let Point load force vector acting at a point be

Pi 3 X 1  Px Py Pz T
beui 3 X 1  u w
T
and the correspond ing displaceme ntvector v

Work potential d PL   Pix .ui  Piy .v i  Piz .w i   ui 1 x 3 Pi 3 x1
T

Total Work potential due to Po int loads at i locations

 PL   ui 1 x 3 Pi 3 x1
T
(1D )
i
Total Potential energy

  U   BF   T   PL
1 
     1 x 6  6 x1 dV   u1 x 3  f 3 x1 dv   u1 x 3 T 3 x1 ds  ui 1 x 3 Pi 3 x1
T T T T

V2  i
V S

Department of Mechanical Engineering, BMSCE , 2020 32

Modelling and Finite Element Analysis 16ME6DCMFE

2.3c: POTENTIAL Energy Of A Bar

𝐱𝟐

𝐈 = ∫ 𝐅(𝐱, 𝐲, 𝐲 ′ , 𝐲" )𝐝𝐱

𝐱𝟏

𝑳 𝑳 𝑳
𝟏 𝒅𝒖 𝟐
𝚷 = ∫ 𝑬𝑨 ( ) 𝒅𝒙 − ∫ 𝒒 𝒖𝒅𝒙 − ∫ 𝑻 𝒖𝒅𝒙 − 𝚺𝒖𝒊 𝑷𝒊
𝟐 𝒅𝒙
𝟎 𝟎 𝟎

2.3d: PE FOR A BEAM

𝐱𝟐

𝐈 = ∫ 𝐅(𝐱, 𝐲, 𝐲 ′ , 𝐲" )𝐝𝐱

𝐱𝟏

𝑳 𝟐 𝑳
𝑬𝑰 𝝏𝟐 𝒗 𝝏𝒗
𝚷= ∫ ( 𝟐 ) 𝒅𝒙 − ∫ 𝒒𝒗𝒅𝒙 − 𝑷𝒗@𝒙=𝑳 − 𝑴𝒐 ( )
𝟐 𝝏𝒙 𝝏𝒙 @𝒙=𝑳
𝟎 𝟎

Department of Mechanical Engineering, BMSCE , 2020 33

Modelling and Finite Element Analysis 16ME6DCMFE
2.4 Admissible configuration/Displacement
• A configuration which satisfies the internal compatibity and Essential Boundary
conditions.
• Polynomial or Trigonometric functions can be used to represent Admissible functions
• Polynomial are
 Inherently continuous
 Easy to integrate or differentiate
 Accuracy can be increased by increasing its order

2.4a Admissible Configuration :Cantilever Beam

Non Admissible : Cantilever Beam

Department of Mechanical Engineering, BMSCE , 2020 34

Modelling and Finite Element Analysis 16ME6DCMFE
2.5 Principle of Minimum Potential Energy(PMPE)

 For Conservative system ,off all the kinematically admissible displacement fields,those
corresponding to equilibrium extremize the total potential energy.

 If the extremum is minimum,the equilibrium state is stable

Three states of Equilibrium

Stable Unstable Neutral

( min ) ( max ) Same 

2.6 PMPE :Discrete System

1. Single spring

F k 2 F
SEP , U   as Stiffness k 
2 2 
Work Potential    F
Potential Energy   U  
k 2
   F
2
d
From PMPE  0
d
K  F  0
K  F : Equilibriu m equation

Department of Mechanical Engineering, BMSCE , 2020 35

Modelling and Finite Element Analysis 16ME6DCMFE

2.PMPE: Spring in series

k 112 k 2 22 k 3 23
SEP , U   
2 2 2
1  Q1  0,  2  Q 2  Q1 ,  3  Q 3  Q 2
Work potential
    F2 Q 2  F3Q 3
Potential Energy   U  

k 1Q12 k 2 Q 2  Q1 2 k 3 Q 3  Q 2 2
  
2 2 2
 F2Q 2  F3Q 3


From PMPE  0, i  1,2,3
Q i
Equilibriu m equations


Q1 k1  k 2  k2 0   Q1   0 
    
   k2 k2  k3  k 3  Q 2     F2 
Q 2  
  0  k3 k 3  Q 3   F3 

Q 3

Department of Mechanical Engineering, BMSCE , 2020 36

Modelling and Finite Element Analysis 16ME6DCMFE
3. PMPE

1 1
𝑆𝐸𝑃 𝑈 = 𝑘1 𝛿12 + 𝑘2 𝛿22 + 𝑘3 𝛿32 + 𝑘4 𝛿42
2 2
𝑊ℎ𝑒𝑟𝑒 𝛿1 = (𝑄2 − 𝑄1 ) , 𝛿2 = (𝑄1 − 0),𝛿3 = (𝑄3 − 𝑄1 ) , 𝛿4 = (0 − 𝑄3 )
𝑊𝑃 Ω = −(−𝐹1 )𝑄1 − 𝐹2 𝑄2 − 𝐹3 𝑄3

• In the matrix form

Department of Mechanical Engineering, BMSCE , 2020 37

Modelling and Finite Element Analysis 16ME6DCMFE
This has subsitution mistake
Check

Subsituting the values , In the matrix form

Solving

2.9 Rayleigh-Ritz method

2.9.a :Rayleigh-Ritz method:

• Variational approach attributed to Lord Rayleigh (1842-1919) & Walter Ritz (1878-1909)
• Rayleigh-Ritz method is a classical Variational method for finding approximate solution
to a given problem.
• The need to solve differential equations can be avoided by applying Rayleigh-Ritz
method to a Functional.
• Rayleigh-Ritz method solution is rarely exact but becomes more accurate as more d.o.f is
used
• The method involves an approximate solution assumed for the displacement fields and
for expressing the total potential energy

𝑢 = 𝜙0 (𝑥, 𝑦, 𝑧) + ∑ 𝑎𝑖 𝜙𝑖 (𝑥, 𝑦, 𝑧)
𝑖=1

Department of Mechanical Engineering, BMSCE , 2020 38

Modelling and Finite Element Analysis 16ME6DCMFE

𝑣 = 𝜓0 (𝑥, 𝑦, 𝑧) + ∑ 𝑏𝑖 𝜓𝑖 (𝑥, 𝑦, 𝑧)
𝑖=1

𝑤 = Γ0 (𝑥, 𝑦, 𝑧) + ∑ 𝑐𝑖 Γ𝑖 (𝑥, 𝑦, 𝑧)
𝑖=1
• The displacements must by kinematically admissible, that is they must satisfy essential
boundary conditions and internal compatibility.
• Where, 𝜙𝑖 , 𝜓𝑖 , Γ𝑖 (for i=1,2…n) are known as independent functions or trial
functions defined over entire domain
• Usually taken as either as polynomials or trigonometric series.
• Coefficients 𝑎𝑖 , 𝑏𝑖 , 𝑐𝑖 are unknown Rayleigh-Ritz parameters to be determined
• The total potential energy is expressed in terms of Rayleigh-Ritz parameters.
𝜋𝑝 = 𝜋𝑝 (𝑎1, 𝑎2 , … … … … . . 𝑎𝑟 , 𝑏1 , 𝑏2 , … … … … … … … … … . 𝑏𝑟 , 𝑐1 , 𝑐2 , … . 𝑐𝑟 )
• Using RR principle, is extremised with respect to RR parameters.
• This results in linear algebraic equations, solving which RR parameters can be obtained.
Thus solution can be obtained

2.9b: STEPS in RR Method

1. Formulate the PE functional for the given problem.
2. Identify the Essential Boundary conditions and Obtain Admissible function which should
satisfy
the Essential/Geometric boundary conditions and internal compatibility.
. Usually taken as either as polynomials or trigonometric series consisting of Ritz parameters or
Generalized coordinates𝑎𝑖
3. Express the total potential energy in terms of Rayleigh-Ritz parameters.
4. Minimize the PE functional with respect to n Ritz parameters so as to obtain n equations with
n unknown
𝜕Π
= 0𝑖 = 1, 2 … … … … … . 𝑛
∂𝑎𝑖
5. Solve these equations to obtain n𝑎𝑖
6. Substitute these𝑎𝑖 in Admissible function to obtain the field variable i.e displacement.
• Obtain derived quantities like strain and stresses.

Department of Mechanical Engineering, BMSCE , 2020 39

Modelling and Finite Element Analysis 16ME6DCMFE

1
𝑆𝐸𝑃, 𝑈 = ∫ 𝜎𝑥 𝜀𝑥 𝑑𝑉
2
𝑣
𝑑𝑢
𝜎 = 𝜎𝑥 = 𝐸𝜀𝑥 , 𝜀 = 𝜀𝑥 =
𝑑𝑥

𝑙
1 𝑑𝑢 2
𝑆𝐸𝑃, 𝑈 = ∫ 𝐸𝐴 ( ) 𝑑𝑥
2 𝑑𝑥
0
WP due to distributed load 𝑞 = 𝑓𝑥 𝐴
𝐿

Ω𝐵𝐹 = − ∫ 𝑞 𝑢𝑑𝑥
0

WP due to Traction load 𝑇 = 𝑇𝑥 . Perimeter

𝐿

Ω𝑟 = − ∫ 𝑇 𝑢𝑑𝑥
0

WP due to point load 𝑃, Ω𝑃𝐿 = −𝑃𝑢𝑝

Department of Mechanical Engineering, BMSCE , 2020 40

Modelling and Finite Element Analysis 16ME6DCMFE

2.9c: Rayleigh-Ritz method: BAR Problem

1. Bar fixed at one end and Loaded by point load P atFree end.

•
A:Area of C/S, E: Young’s ModulusEBC is u=0 at x=o

Assume Trial function in polynomial form

(Minimum first order as there is one EBC)

Step1. Admissible function

• Trial function𝑢 = 𝑎0 + 𝑎1 𝑥(1)

𝐸𝐵𝐶: 𝐴𝑡 𝑥 = 0, 𝑢 = 0 ∴ 𝑎0 + 𝑎1 (0) ∴ 𝑎0 = 0

∴ 𝑢 = 𝑎1 𝑥 (2), Admissible function

Step2.Potential Energy

1 𝐿 𝑑𝑢 2
Π − ∫0 𝐸𝐴 ( ) 𝑑𝑥 − 𝑃𝑢@𝑥=1
2 𝑑𝑥

• Substituting (2)
1 𝐿
Π − 2 ∫0 𝐸𝐴 (𝑎1 )2 𝑑𝑥 − 𝑃𝑎1 𝐿

1
Π − 2 𝐸𝐴 𝑎1 2 𝐿 − 𝑃𝑎1 𝐿

𝒅𝝅
Step3.By RR Method𝒅𝒂 = 𝟎, 𝑬𝑨𝑳𝒂𝟏 − 𝑷𝑳 = 𝟎 (𝟑)
𝟏

𝑷
Step4 .Solving for Ritz Parameter 𝒂𝟏 = 𝑬𝑨 (𝟒)

Step5: Displacement field, (4) in (2)

𝑃
𝑢 = 𝐸𝐴 𝑥 (5)

𝒅𝒖 𝑷
Step6. Stress𝝈𝒙 = 𝑬𝜺𝒙 = 𝑬 𝒅𝒙 = 𝑨 (𝟔)

Department of Mechanical Engineering, BMSCE , 2020 41

Modelling and Finite Element Analysis 16ME6DCMFE
Problem 2..Bar fixed at one end and Loaded by linearly varying distributed load q=Cx.

A:Area of C/S, E: Young’s ModulusEBC is u=0 at x=o

Assume Trial function in polynomial form

(Minimum first order as there is one EBC)

Step1. Admissible function

• Trial function𝑢 = 𝑎0 + 𝑎1 𝑥(1)

𝐸𝐵𝐶: 𝐴𝑡 𝑥 = 0, 𝑢 = 0 ∴ 𝑎0 + 𝑎1 (0) ∴ 𝑎0 = 0

∴ 𝑢 = 𝑎1 𝑥 (2), Admissible function

Step2.Potential Energy

1 𝐿 𝑑𝑢 2 𝐿
Π = 2 ∫0 𝐸𝐴 (𝑑𝑥 ) 𝑑𝑥 − ∫0 𝑞 𝑢𝑑𝑥 Substituting (2)

1 𝐿 𝑑𝑢 2 𝐿
Π = 2 ∫0 𝐸𝐴 (𝑑𝑥 ) 𝑑𝑥 − ∫0 𝐶 𝑥𝑢𝑑𝑥 𝑞 = 𝐶𝑥

1 𝐿
Π = 2 𝐸𝐴 (𝑎1 )2 𝑑𝑥 − ∫0 𝐶 𝑎1 𝑥 2 𝑑𝑥

1 𝐶𝑎1 3
Π= 𝐸𝐴 𝑎1 2 𝐿 − 𝐿
2 3
Step3.By RR Method

𝑑𝜋 𝐶 3
= 0, 𝐸𝐴𝐿𝑎1 − 𝐿 = 0 (3)
𝑑𝑎1 3
𝐶𝐿2
Step4 .Solving for Ritz Parameter𝑎1 = 3𝐸𝐴 (4)

Step5: Displacement field, (4) in (2)

𝐶𝐿2
𝑢= 𝑥 (5)
3𝐸𝐴
𝑑𝑢 𝐶𝐿2
Step6. Stress𝜎𝑥 = 𝐸𝜀𝑥 = 𝐸 𝑑𝑥 = (6)
3𝐴

Department of Mechanical Engineering, BMSCE , 2020 42

Modelling and Finite Element Analysis 16ME6DCMFE

Problem 3. Same problem by

2nd order polynomial

Step1. Admissible function

• Trial function𝑢 = 𝑎0 + 𝑎1 𝑥 + 𝑎2 𝑥 (1)

𝐸𝐵𝐶: 𝐴𝑡 𝑥 = 0, 𝑢 = 0 ∴ 0 = 𝑎0 + 𝑎1 (0) +
𝑎2 02 ∴ 𝑎0 = 0

∴ 𝑢 = 𝑎1 𝑥 + 𝑎2 𝑥 2 (2), Admissible function

Step2.Potential Energy
𝐿 𝐿
1 𝑑𝑢 2
Π = ∫ 𝐸𝐴 ( ) 𝑑𝑥 − ∫ 𝑞 𝑢𝑑𝑥
2 𝑑𝑥
0 0

𝒅𝒖
From (2), 𝒖 = 𝒂𝟏 𝒙 + 𝒂𝟐 𝒙𝟐 (𝒅𝒙 ) = 𝒂𝟏 + 𝟐𝒂𝟐 𝒙 and 𝒒 = 𝑪𝒙

𝐿 𝐿
1
Π = ∫ 𝐸𝐴 (𝑎1 + 2𝑎2 𝑥)2 𝑑𝑥 − ∫ 𝐶 (𝑎1 𝑥 + 𝑎2 𝑥 2 )𝑥 𝑑𝑥
2
0 0

𝐿 𝐿
𝐸𝐴
Π= ∫(𝑎1 2 + 4𝑎1 𝑎2 𝑥 + 4𝑎2 2 𝑥 2 ) 𝑑𝑥 − ∫ 𝐶 (𝑎1 𝑥 2 + 𝑎2 𝑥 3 )𝑑𝑥
2
0 0

𝐸𝐴 𝐿 4 𝐿3 𝑎2 4
Π= 2
∫0 [𝑎12 𝐿 + 2𝑎1 𝑎2 𝐿2 + 3 𝑎22 𝐿3 ] − 𝑐 [𝑎1 3
+ 4
𝐿] (3)

Step3.By RR Method

𝜕𝜋 𝐸𝐴 𝐶
= 0, [2𝑎1 𝐿 + 2𝑎2 𝐿2 ] − 𝐿3 = 0 (3𝑎)
𝜕𝑎1 2 3

𝜕𝜋 𝐸𝐴 8 𝐶
= 0, [2𝑎1 𝐿2 + 𝑎2 𝐿2 ] − 𝐿4 = 0 (3𝑏)
𝜕𝑎2 2 3 3
𝐶𝐿2 4 𝐶𝐿2
3𝑎 → 𝑎1 + 𝑎2 𝐿 = 3𝐸𝐴 (3𝑐) 3𝑏 → 𝑎1 + 3 𝑎2 𝐿 = 4𝐸𝐴 (3𝑑)

Department of Mechanical Engineering, BMSCE , 2020 43

Modelling and Finite Element Analysis 16ME6DCMFE
Step4 .Solving for Ritz Parameters,3c & 3d
7𝐶𝐿2 𝐶𝐿
Solving, 𝑎1 = 12𝐸𝐴 , 𝑎2 = − 4𝐸𝐴 (4)

Step5:Displacement field, (4) in (2)

7𝐶𝐿2 𝐶𝐿
(4)𝑖𝑛 (2), 𝑢 = 𝑥 − 4𝐸𝐴 𝑥 2 (5) Displacement field: Quadratic
12𝐸𝐴

Step6 Stress

𝑑𝑢 7𝐶𝐿2 𝐶𝐿2
𝜎𝑥 = 𝐸𝜀𝑥 = 𝐸 𝑑𝑥 = − 𝑥 (6) Stress:Linear
12𝐴 2𝐴

Convergence

Particular First Order Second Order

Admissible 𝑢 = 𝑎1 𝑥 𝑢 = 𝑎1 𝑥 + 𝑎2 𝑥 2
Function(u)
Displacement(u) 𝐶𝐿2 7𝐶𝐿2 𝐶𝐿 2
𝑢= 𝑥 𝑥− 𝑥
3𝐸𝐴 12𝐸𝐴 4𝐸𝐴
Stress(x) 𝐶𝐿2 7𝐶𝐿2 𝐶𝐿
𝜎𝑥 = − 𝑥
3𝐴 12𝐴 2𝐴
(Constant) (Linear)

Problem: Bar fixed at one end and Loaded by quadratically varying distributed load q=Cx2.

Convergence:Observations

• Fig.shows Exact and Approximate Solutions

• As order of Polynomial increases, Accuracy increases.
• Approximate displacements are more accurate compared to Stresses.
• Error in satisfying D.E and NBC
• Using Cubic polynomial and satisfying Natural B.C
𝑑𝑢
𝜎𝑥 = 𝐸 { } =0
𝑑𝑥 @𝑥=0

𝐶𝐿2 𝐶
Yield Exact solutions𝑎1 = 2𝐸𝐴 , 𝑎2 = 0, 𝑎3 = − 6𝐸𝐴

Department of Mechanical Engineering, BMSCE , 2020 44

Modelling and Finite Element Analysis 16ME6DCMFE

Problem: .Bar fixed at both ends and Loaded by quadratically varying distributed load q=Cx2.

Department of Mechanical Engineering, BMSCE , 2020 45

Modelling and Finite Element Analysis 16ME6DCMFE
Problem 4. Using Rayleigh Ritz method obtain expressions for displacement and stress
for uniform bar shown in figure. Normalize the values if P=A=L=E=1

• Step1:Admissible Function
• Based on the Rayleigh-Ritz
• method, let us assume axial displacement at any point as𝑢 = 𝑎0 + 𝑎1 𝑥 + 𝑎2 𝑥 2 (1)
• Where a0, a1 and a2 are Rayleigh-Ritz constants, which are to be determined.
• The assumed displacement𝑢 = 𝑎0 + 𝑎1 𝑥 + 𝑎2 𝑥 2 should satisfy the following essential
conditions At x=0 and L, u=0
• u=0 at x=0,𝑎0 = 0
• 𝐸𝐵𝐶: 𝐴𝑡 𝑥 = 2𝐿, 𝑢 = 0 ∴ 0 = 𝑎1 (2𝐿) + 𝑎2 (2𝐿)2 ∴ 𝑎1 = −2𝑎2 𝐿
• Admissible function is
• 𝑢 = 𝑎2 (𝑥 2 − 2𝑥𝐿) (2), Admissible function
• Step 2 P.E
2𝐿 2
1 𝑑𝑢
Π = ∫ 𝐸𝐴 ( ) 𝑑𝑥 − 𝑃𝑢@𝑥=𝑙
2 𝑑𝑥
0

𝑑𝑢
From (2), 𝑢 = 𝑎2 (𝑥 2 − 2𝑥𝐿) (𝑑𝑥 ) = 2𝑎2 (𝑥 − 𝐿) and 𝑢@𝑥=𝐿 = −𝑎2 𝐿2

2
Π = 𝐸𝐴𝐿3 𝑎2 2 − 𝑃𝑎2 𝐿2 (3)
3
• Step3.By RR Method
𝑑𝜋 4
= 0, 𝐸𝐴𝐿3 𝑎2 + 𝑃𝐿2 = 0 (3)
𝑑𝑎2 3

• Step4. Solving (3)

3𝑃
𝑎2 = − 4𝐸𝐴𝐿 (4)

• Step5:Displacement field, (4) in (2)

3𝑃 3𝑃𝑥
𝑢 = − 4𝐸𝐴𝐿 (𝑥 2 − 2𝑥𝐿) Or 𝑢 = 4𝐸𝐴𝐿 (2𝐿 − 𝑥) (5)

Department of Mechanical Engineering, BMSCE , 2020 46

Modelling and Finite Element Analysis 16ME6DCMFE
3𝑃𝐿
• At mid Point, x=L,𝑢@𝑥=𝐿 = 4𝐸𝐴
𝑑𝑢 3𝑃
• Step6:Stress𝜎𝑥 = 𝐸𝜀𝑥 = 𝐸 𝑑𝑥 = 2𝐴𝐿 (𝐿 − 𝑥) (6)

• At mid Point, x=L,𝜎𝑥 = 0

• If A=P=E=L=1,at midpoint ,u= 0.75

Comparison

From RR Method
𝟑𝑷𝒙
𝒖= (𝟐𝑳 − 𝒙)Quadratic
𝟒𝑬𝑨𝑳

𝒅𝒖 𝟑𝑷
𝝈𝒙 = 𝑬 𝒅𝒙 = 𝟐𝑨𝑳 (𝑳 − 𝒙) Linear

2.10 : Rayleigh-Ritz method

• Functional form of D.E should be available

• Admissible function satisfying EBC and internal compatibility should be found
• Applicable with simple geometry.
• Automatically satisfies NBC.
• Extended to FEM as Piecewise RR Method

Department of Mechanical Engineering, BMSCE , 2020 47

 Modelling and Finite Element Analysis 16ME6DCMFE
3.1 Integration:

Reasons for numerical integration

There are several reasons for carrying out numerical integration.

1. The integrand f(x) may be known only at certain points, such as obtained by sampling.
Some embedded systems and other computer applications may need numerical integration
for this reason.
2. A formula for the integrand may be known, but it may be difficult or impossible to find
an antiderivative that is an elementary function. An example of such an integrand is f(x) =
exp(−x2), the antiderivative of which (the error function, times a constant) cannot be written
in elementary form.
3. It may be possible to find an antiderivative symbolically, but it may be easier to compute a
numerical approximation than to compute the antiderivative. That may be the case if the
antiderivative is given as an infinite series or product, or if its evaluation requires a special
function that is not availabl

3.2 Quadrature:

Quadrature can be defined as a numerical integration of a function. Quadrature rule is an

approximation of the definite integral of a function, usually stated as a weighted sum of function
values at specified points within the domain of integration.

In FEM one, two, three dimensional problems especially for computation of the element stiffness or
for the element nodal load vector are required to solve. Several methods are available to solve such
definite integrals.

Two most widely used methods are

Department of Mechanical Engineering, BMSCE , 2020 48

 Modelling and Finite Element Analysis 16ME6DCMFE
1. Gaussian Quadrature
2. Newton cotes Quadrature
1.3 Newton cotes Quadrature

Newton–Cotes quadrature is a group of formulae for numerical integration (also called quadrature)
based on evaluating the integrand at equally spaced points.

1 Trapezoid rule

2 Simpson's rule

3 Simpson's 3/8 rule

For the Newton–Cotes rules to be accurate, the step size h needs to be small, which means that the
interval of integration must be small itself, which is not true most of the time. For this reason, one
usually performs numerical integration by splitting into smaller subintervals, applying a Newton–
Cotes rule on each subinterval, and adding up the results.

3.3 Gaussian Quadrature

The gauss quadrature is the approximation for the definite integrals; it includes some specific
functions called weight function and some sampling points called gauss points through which the
approximation method has been carried out.

For example, consider an integral such as,

This definite integral can be approximated by a property weighted sum of particular values of
function f( ) at specific point , .i.e. , f( ) with in the limits o integration.

Thus the value of integral can be approximated as

Department of Mechanical Engineering, BMSCE , 2020 49

 Modelling and Finite Element Analysis 16ME6DCMFE

Where is called sampling point or gauss point

f( is the value of functions at gauss points I (i=1 to n) and

is the weight function

Now Equation 2 can be expressed as,

Note:

Newton Cotes forms pick equally spaced points in the interval of integration,Gaussian
quadrature picks the best points. For this reason Gaussian quadratureis more accurate and uses
less panels. ... Gaussian does not include the endpoints. Higher forms of Newton's also tend to
have large coefficients

1. Gauss quadrature method is especially adopted for the integrals whose variables are specified in
natural coordinates system and the integration limit should be from -1 to +1 .
2. The function to be integrated should be in polynomial form of the order (2n-1) or less for ‘n’
sampling points during approximation.

Mathematically,

For n=1, the function should be in (2n-1) degree (i.e. 2×1-1=1 degree) of polynomial
For example f(x)= + x
For n=2 , f(x) must be a polynomial of the order (2×2-1=3) or less
i.e. f(x)= + x
or f(x)= + x+
or f(x)= + x+ + which is the maximum degree for n=2.

In gauss quadrature method the Gaussian points will be located at equal distanes from the origin in t
he opposite directions i.e. , they are symmetrically located in the curves specified by the integral .
The distance or gauss points from the origin (i.e. , the location of gauss points) and the values of
weight function at the gauss points

Department of Mechanical Engineering, BMSCE , 2020 50

 Modelling and Finite Element Analysis 16ME6DCMFE

3.4 One point formula

Let us consider the integral with n=1 (i.e. , one point approximation as

Since n=1 , the function f(x) can have order as (2n-1) = (2*1-1)=1

Hence , f(x) may be written as

f(x)= + x

Now EQN 1 becomes

OR ERROR=

i.e. + + =0

+ + =0

(2) +

i.e. 2

i.e. . =0

to satisfy the equation (2) i.e. is the mid point of the integral and the weight

Department of Mechanical Engineering, BMSCE , 2020 51

 Modelling and Finite Element Analysis 16ME6DCMFE
3.5 Two point formula

Consider two point approximation. The integral function for n=2 can be written as

For n=2, the function for f(x) may be have the polynomial order = (2n-1)= (2*2-1)=3

i.e. f(x)= + x+ +

Now from eqn 1

Error=

i.e. + +

Which yields

Solving the equation 1, we have

,and

=0.57735

Hence, equation (1) can be written as

Department of Mechanical Engineering, BMSCE , 2020 52

 Modelling and Finite Element Analysis 16ME6DCMFE
3.6 Three point formula

Consider three point approximations. The integral function for n=3 can be written

For n=3 , the function f(x) may have the polynomial order= (2n-1)= (2*3)-1=5

I.e. f(x)= + x+ +

From eqn 1

Error=

i.e. +

After simplification we get

(2

+ (

Which yields

Department of Mechanical Engineering, BMSCE , 2020 53

 Modelling and Finite Element Analysis 16ME6DCMFE
Solving equation 1 we have

Hence the equation 1 can be written as,

2.4 Problems:

1. Using Gaussian quadrature formula evaluate

Solution:
To find sampling points (n)
Order of polynomial = 3
(2n-1) = 3
i.e. Sampling point, n=2
Thus, for two sampling points use Two-point formula

Where,
Thus,
At

Department of Mechanical Engineering, BMSCE , 2020 54

 Modelling and Finite Element Analysis 16ME6DCMFE
At

1 1
2. Evaluate I =  1
(3e x  x 2 
( x  2)
)dx using one point and two point Gauss Quadrature.

Solution:

One-point formula: n=1 (sampling points)

i.e.,

Where,
1
 x2 
Given that, 3e x
( x  2)

Two –point formula: n=2 (sampling points)

Where,
1
 x2 
3e x ( x  2)

At ;

At ;

Department of Mechanical Engineering, BMSCE , 2020 55

 Modelling and Finite Element Analysis 16ME6DCMFE
When the limits are not from -1 to 1

Department of Mechanical Engineering, BMSCE , 2020 56

Modelling and Finite Element Analysis 16ME6DCMFE

Department of Mechanical Engineering, BMSCE , 2020 57

 Modelling and Finite Element Analysis 16ME6DCMFE
3. Evaluate the value of the integral .
Soln.: Degree/order of the polynomial= 1
2n-1=1

I= where
But we know that mapping function for change of intervals,

p=

Where, a=2 and b=4 (given)

i.e. , p=

i.e. , p= x+3

Department of Mechanical Engineering, BMSCE , 2020 58

Modelling and Finite Element Analysis 16ME6DCMFE

Department of Mechanical Engineering, BMSCE , 2020 59

Modelling and Finite Element Analysis 16ME6DCMFE

Department of Mechanical Engineering, BMSCE , 2020 60

Modelling and Finite Element Analysis 16ME6DCMFE

Department of Mechanical Engineering, BMSCE , 2020 61

Modelling and Finite Element Analysis 16ME6DCMFE

Department of Mechanical Engineering, BMSCE , 2020 62

 Modelling and Finite Element Analysis 16ME6DCMFE

Advantages of Gaussian quadrature over Newton cotes

Gaussian quadrature is more accurate than the Newton-Cotes quadrature in the following sense:

1. Both Gaussian quadrature and Newton-Cotes quadrature use the similar idea to do the
approximation, i.e. they both use the Lagrange interpolation polynomial to approximate the
integrand function and integrate the Lagrange interpolation polynomial to approximate the given
definite integral.

2. When the same number of nodes is used, the algebraic degree of precision of the Gaussian
quadrature is higher than that of the Newton-Cotes quadrature.

Solution of Linear Algebraic quations

Department of Mechanical Engineering, BMSCE , 2020 63

 Modelling and Finite Element Analysis 16ME6DCMFE
A system of linear Algebraic equations is nothing but a system of ' n' algebraic equations
satisfied by a set of n unknown quantities. The aim is to find these n unknown quantities
satisfying the n equations. It is a very common practice to write the system of n equations in
matrix form as Ax = b where A is an n x n, non-singular matrix and x and b are n x 1 matrices
out of which b is known.
.
There are two types of methods exists to solve these linear systems.
1. Direct Methods

These are the methods have number of steps known apriori

Some of the important direct methods are

i. Gaussian Elimination and ii.LU Decomposition

2. Iterative Methods:

i.Jacobi-Method,ii.Gauss-Seidel Method

iii.Relaxation Method

A system of linear equations has no solution, or exactly one solution, or infinitely many
solutions.
A system of linear equations is said to be consistent if it has either one solution or infinitely
many solutions.
A system of linear equation is said to be inconsistent if it has no solution.

The essential information of a linear system can be recorded compactly in a rectangular array
called a matrix.
For the following system of equations,
x1  2 x2  x3  0
2 x2  8 x3  8
4 x1  5 x2  9 x3  9,
coefficient matrix augmented matrix
 1 2 1
 1 2 1 0
 0 2 8  0
  2 8 8
 
 4 5 9   4
 5 9 9 

GAUSS ELIMINATION METHOD FOR SOLVING LINEAR ALGEBRAIC EQUATION –
WITH NUMERICAL EXAMPLES :

Department of Mechanical Engineering, BMSCE , 2020 64

 Modelling and Finite Element Analysis 16ME6DCMFE
There are two basic steps in this elimination method. They are
Forward elimination
Back substitution.

SSOLVE BY GAUSS ELIMINATION METHOD.

10
X + Y
2 + Z1 = 13 (1)

X
2 + 10
Y + Z2 = 14 (2)

X1 + Y2 + 10
Z = 13 (3)
ELIMINATION STEP
R X Y Z B
1 10.00 2.00 1.00 13.00 Equation 1
2 2.00 10.00 2.00 14.00 Equation 2
Equation 3
3 1.00 2.00 10.00 13.00

4 1.00 0.20 0.10 1.30 Divide throughout equation [1] by the coefficient of X i.e.. 10

5 0.00 9.60 1.80 11.40 Multiply [4] by the coefficient of X of [2] and subtract it from [2]

6 0.00 1.80 9.90 11.70 Multiply [4] by the coefficient of X of [3] and subtract it from [3]
Divide equation [5] through out by the coefficient of Y of
7 1.00 0.19 1.19
equation [5] i.e.. 9.6
8 0.00 9.56 9.56 Multiply [7] by the coefficient of Y of [6] and subtract it from [6]
BACK SUBSTITUTION STEP From [8] we have z = 1
From [7] we have y = 1.000 From [4] we have x =1

NOTE :FORWARD ELIMINATION

4 1.00x + 0.20y + 0.10z = 1.30

7 0x + 1.00y + 0.19z = 1.19
8 0x + 0y + 9.56z = 9.56
BACK SUBSTITUTION STEP
From [8] we have z = 1.000
From [7] we have y = 1.000
From [4] we have x = 1.000

2.SOLVE BY GAUSS ELIMINATION METHOD.

Department of Mechanical Engineering, BMSCE , 2020 65

Modelling and Finite Element Analysis 16ME6DCMFE

2X + 1Y + -1 Z = 7 (1)

1X + -2 Y + 3 Z = 12 (2)

2X + 4Y + -3 Z = 10 (3)
ELIMINATION STEP
RX Y Z B
1 2.00 1.00 -1.00 7.00 Equation 1
2 1.00 -2.00 3.00 12.00 Equation 2
3 2.00 4.00 -3.00 10.00 Equation 3
Divide through out equation [1] by the
4 1.00 0.50 -0.50 3.50
coefficient of X i.e.. 10
Multiply [4] by the coefficient of X of [2] and
5 0.00 -2.50 3.50 8.50
subtract it from [2]
Multiply [4] by the coefficient of X of [3] and
6 0.00 3.00 -2.00 3.00
subtract it from [3]
Divide equation [5] through out by the
7 1.00 -1.40 -3.40
coefficient of Y of equation [5] i.e.. 9.6
Multiply [7] by the coefficient of Y of [6] and
8 0.00 2.20 13.20
subtract it from [6]

BACK SUBSTITUTION STEP From [8] we have z = 6.000

From [7] we have y = 5.000 From [4] we have x = 4.000

3.SOLVE BY GAUSS ELIMINATION METHOD.

10 X +4 Y+ 1Z = 24 (1)

Department of Mechanical Engineering, BMSCE , 2020 66

Modelling and Finite Element Analysis 16ME6DCMFE

4 X +6 Y+ 3Z = 28 (2)

2 X +4 Y+ 8Z = 30 (3)

ELIMINATION STEP

R X Y Z B

1 10.00 4.00 1.00 24.00 Equation 1

2 4.00 6.00 3.00 28.00 Equation 2

3 2.00 4.00 8.00 30.00 Equation 3

4 1.00 0.40 0.10 2.40 Divide through out equation [1] by the coefficient of X i.e.. 10

5 0.00 4.40 2.60 18.40 Multiply [4] by the coefficient of X of [2] and subtract it from [2]

6 0.00 3.20 7.80 25.20 Multiply [4] by the coefficient of X of [3] and subtract it from [3]

7 1.00 0.59 4.18 Divide equation [5] through out by the coefficient of Y of equation [5] i.e.. 9.6

8 0.00 5.91 11.82 Multiply [7] by the coefficient of Y of [6] and subtract it from [6]

BACK SUBSTITUTION STEP From [8] we have z =

From [7] we have y = 3.000 From [4] we have x =

Partial pivoting
4. Consider x + x + x = 6, 3x + 3x + 4x = 20, 2x + x + 3x = 13
1 2 3 1 2 3 1 2 3

Department of Mechanical Engineering, BMSCE , 2020 67

Modelling and Finite Element Analysis 16ME6DCMFE
Partial pivoting : Since the largest element in the first column is at 3x which is not in
1
the pivotal position, perform the row transformation R « R . Now the system is
1 2
3x + 3x + 4x = 20
1 2 3
x +x +x =6
1 2 3
2x + x + 3x = 13
1 2 3
R ® R - 1/3 R
2 2 1
R ® R - 2/3 R
3 3 1

Since this is a '0' at the pivotal position i.e., at second row second column
apply R «R (interchange rows two and three)
2 3

Now the augmented matrix is in diagonal form(the part of coefficient matrix A).
Back substitution : From the last row : -1/3x = -2/3 Þ x = 2
3 3
Now from the second row
-x = -1/3 - 1/3x Þ x =1
2 2 2
and from the first row 3x = 20 - 3x - 4x(2) Þ x =3
1 1 1
T
Hence the solution is X = (3 1 2)

Department of Mechanical Engineering, BMSCE , 2020 68

Modelling and Finite Element Analysis 16ME6DCMFE

6.1 Commercial FEA program consists of three basic modules:

1. Pre processor
2. Processor/Solver
3. Post processor.

PRE-PROCESSOR

1. Read control parameters

2. Read/Generate nodal coordinates and boundary conditions
3. Read/Generate elements connectivity and elements loads
4. Read material properties or constitutive matrices
5. Read nodal loads and loadings conditions

PROCESSOR/SOLVER
1. Computer parameters for memory/file managemant
2. Computer elements matrices and vectors
3. From global matrices and vectors
4. Enforce SPC, MPC boundary conditions.
5. Solve governing matrix equations and compute the unknown values of
primary variables.
6. Compute additional derived variables such as reaction forces, element
strains, stresses and heat flow etc.

POST PROCESSOR
1. Print/plot deformed mesh over undeformed mesh
2. Print/plot contours of displacements
3. Print/plot contours of stresses
4. Display locations of max./min. stress
5. Print/plot contours of failure index
6. Animate dynamic model behaviour.
7. Sort element stresses in order of magnitude.
8. Produce color-coded temperature plots.

Department of Mechanical Engineering, BMSCE , 2020 69

Modelling and Finite Element Analysis 16ME6DCMFE
In general, a finite element solution may be broken into the following three stages. This is a
general guideline that can be used for setting up any finite element analysis.
1. Preprocessing: defining the problem; the major steps in preprocessing are given below:
o Define keypoints/lines/areas/volumes (Or Building a solid model)
o Define element type and material/geometric properties
o Mesh lines/areas/volumes as required
The amount of detail required will depend on the dimensionality of the analysis (i.e.
1D, 2D, axi-symmetric, 3D).
2. Solution: assigning loads, constraints and solving; here we specify the loads (point or
pressure), constraints (translational and rotational) and finally solve the resulting set of
equations.
3. Postprocessing: further processing and viewing of the results; in this stage one may wish
to see:
o Lists of nodal displacements
o Element forces and moments
o Deflection plots
o Stress contour diagrams

The objective of this chapter is outline a general analysis procedure to be used to solve a
simulation. Regardless of the physics of the problem, the same general procedure can be
followed.

4.1 Every analysis involves four main steps: Preliminary Decisions

•Preliminary Decisions

–Which analysis type?

–What to model? Preprocessing
–Which element type?

•Preprocessing
–Define Material Solution
–Create or import the model geometry
–Mesh the geometry
•Solution
–Apply loads Postprocessing
–Solve
•Postprocessing
–Review results
–Check the validity of the solution

Department of Mechanical Engineering, BMSCE , 2020 70