MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

A - INTRODUCTION AND OVERVIEW

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 1

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

Course Content:

A – INTRODUCTION AND OVERVIEW

Numerical method and Computer-Aided Engineering; Physical
problems; Mathematical models; Finite element method;.

B – REVIEW OF 1-D FORMULATIONS

Elements and nodes, natural coordinates, interpolation function, bar
elements, constitutive equations, stiffness matrix, boundary
conditions, applied loads, theory of minimum potential energy; Plane
truss elements; Examples.

C – PLANE ELASTICITY PROBLEM FORMULATIONS

Constant-strain triangular (CST) elements; Plane stress, plane
strain; Axisymmetric elements; Stress calculations; Programming
structure; Numerical examples.

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 2

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

COMPUTER-AIDED ENGINEERING (CAE)

The use of computers to analyze and simulate the

function (structural, motion, etc.) of mechanical,
electronic or electromechanical systems.

• Computer-Aided Design (CAD)

- Drafting
- Solid modeling
- Animation & visualization
- Dimensioning & tolerancing
• Engineering Analyses
- Analytical & numerical methods

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 3

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

The Process of FE analysis

Ref: K.J. Bathe, Finite Element Procedures, Prentice-Hall, 1996

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 4

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

PHYSICAL PROBLEM AND MATHEMATICAL MODEL

Buckling of Euler Column

d2y P
2
+ y=0
dx EI

Eigenvalue problem
k1 + k 2 − k 2   x1  2 1
m 0   x1 
 −k   =ω   
 2 k 2   x2  0 m2   x2 

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 5

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

EXAMPLE OF A MODEL
Steady-state heat conduction through a thick wall

d  dT   k11 k12 L k1L  T1   R1 

k  +Q=0 k k 22 L k 2 L  T2   R2 
dx  dx   21  = 
 M  M   M 
Solution : T ( x)    
k L1 kL2 L k LL + h  TL   RL + hT∞ 

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 6

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

REVIEW OF MATRIX ALGEBRA

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 7

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS
Matrix Algebra
In this course, we need to solve system of linear equations in the
form
a11 x1 + a12 x2 + ... + a1n xn = b1
a21 x1 + a22 x2 + ... + a2 n xn = b2 (2-1)
............................................
an1 x1 + an 2 x2 + ... + ann xn = bn
where x1, x2, …, xn are the unknowns.

Eqn. (2-1) can be written in a matrix form as

[A]{x} = {b} (2-2)

where [A] is a (n x n) square matrix, {x} and {b} are (n x 1) vectors.

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 8

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

The square matrix [A] and the {x} and {b} vectors are is given by,

 a11 a12 ... a1n   x1   b1 

a ... a2 n  x  b 
a22  2  2
[A] =  21 , {x} =   , and {b} =  
 : : : :  : :
   xn  bn 
an1 an 2 ... ann 

… (2-3)

Note:
Element located at ith row and jth column of matrix [A] is denoted by aij. For
example, element at the 2nd row and 2nd column is a22.

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 9

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS
Matrix Multiplication
The product of matrix [A] of size (m x n) and matrix [B] of size (n x
p) will results in matrix [C], with size (m x p).

[A] [B] = [C ] (2-4)

( m x n) ( n x p ) (m x p)
Note: The (ij)th component of [C], i.e. cij, is obtained by taking the DOT
product,

cij = (ith row of [ A]) ⋅ ( jth column of [ B]) (2-5)

Example: 1 4 
2 1 3 5 − 2   7 15
 0 − 2 1   = - 10 7 
  0 3   

(2 x 3) (3 x 2) (2 x 2)
INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 10
MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS
Matrix Transposition
If matrix [A] = [aij], then transpose of [A], denoted by [A]T, is given by
[A]T = [aji]. Thus, the rows of [A] becomes the columns of [A]T.

 1 − 5
Example: 0 6 
[ A] =  
− 2 3 
 
 4 2 

 1 0 − 2 4
Then, [ A] = 
T

 − 5 6 3 2 

Note: In general, if [A] is of dimension (m x n), then [A]T has the dimension
of (n x m).
INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 11
MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS
Transpose of a Product
The transpose of a product of matrices is given by the product of the
transposes of each matrices, in reverse order, i.e.

([ A][ B ][C ])T = [C ]T [ B ]T [ A]T (2-6)

Determinant of a Matrix
Consider a 2 x 2 square matrix [x],

 x11 x12 
[x] = 
 x21 x22 
The determinant of this matrix is give by,

det [x ] = x11 x22 − x21 x12 (2-7)

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 12

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS
EXAMPLE
Given that:
 1 0  3 1 0
[ A] =   [C ] =  
 − 1 4   − 1 0 3 
 3 1 2 1
 
[ D] = 1 4 0 {E} = 2
2 0 3 3 
 

Find the product for the following cases:

a) [A][C]
b) [D]{E}
c) [C]T[A]

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 13

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS
Solution of System of Linear Equations
System of linear algebraic equations can be solved for the unknown using
the following methods:
a) Cramer’s Rule
b) Inversion of Coefficient Matrix
c) Gaussian Elimination
d) Gauss-Seidel Iteration

Example: Solve the following SLEs using Gauss elimination method.

2 x1 + 1x2 − 3 x3 = 11 (i)
4 x1 − 2 x2 + 3x3 = 8 (ii)
− 2 x1 + 2 x2 − 1x3 = −6 (iii)

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 14

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

Gauss Elimination Method

Reducing a set of n equations in n unknowns to an equivalent triangular
form (forward elimination). The solution is determined by back substitution
process.

Basic approach
-Any equation can be multiplied (or divided) by a nonzero scalar
-Any equation can be added to (or subtracted from) another equation
-The position of any two equations in the set can be interchanged

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 15

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

Example: Solve the following SLEs using Gaussian elimination.

2 x1 + 1x2 − 3 x3 = 11 (i)
4 x1 − 2 x2 + 3x3 = 8 (ii)
− 2 x1 + 2 x2 − 1x3 = −6 (iii)

Eliminate x1 from eq.(ii) and eq.(iii). Multiply eq.(ii) by 0.5 we get,

2 x1 + 1x2 − 3 x3 = 11 (i)
2 x1 − 1x2 + 1.5 x3 = 4 (ii) *
− 2 x1 + 2 x2 − 1x3 = −6 (iii)

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 16

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

Subtract eq.(ii)* from eq.(i), we obtain

2 x1 + 1x2 − 3x3 = 11 (i)

0 x1 + 2 x2 − 4.5 x3 = 7 (ii) * *
− 2 x1 + 2 x2 − 1x3 = −6 (iii)
Add eq.(iii) with eq.(i), yields

2 x1 + 1x2 − 3 x3 = 11 (i)
0 x1 + 2 x2 − 4.5 x3 = 7 (ii) * *
0 x1 + 3 x2 − 4 x3 = 5 (iii) *

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 17

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS
Eliminate x2 from eq.(iii)*. Multiply eq.(ii)** by 3 and eq.(iii)* by 2
we get
2 x1 + 1x2 − 3 x3 = 11 (i)
0 x1 + 6 x2 − 13.5 x3 = 21 (ii) * * *
0 x1 + 6 x2 − 8 x3 = 10 (iii) * *

Subtract eq.(iii)** from eq.(ii)***, we obtain

2 x1 + 1x2 − 3 x3 = 11 (i)
0 x1 + 2 x2 − 4.5 x3 = 7 (ii) * *
0 x1 + 0 x2 − 5.5 x3 = 11 (iii) * * *
From eq.(iii)*** we determine the value of x3, i.e.

11
x3 = = −2
− 5 .5

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 18

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

Back substitute value of x3 into eq.(ii)** and solve for x2, we get

7 + 4.5(−2)
x2 = = −1
2
Back substitute value of x2 and x3 into eq.(i) and solve for x1, we
get

x1 = 6

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 19

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS
Example
Solve the following systems of linear equations by using the
Gaussian elimination method.

a) − x1 + 3x2 − 2 x3 = 2
2 x1 − 4 x2 + 2 x3 = 1
0 x1 + 4 x2 + x3 = 3

b) 2 x1 + 1x2 − 3x3 = 11
4 x1 − 2 x2 + 3x3 = 8
− 2 x1 + 2 x2 − 2 x3 = −6

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 20

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

Steps in solving a continuum problem by FEM

Identify and understand the problem
(This essential step is not FEM)

Select the solution domain
Select the solution region for analysis.

Discretize the continuum
Divide the solution region into finite number of elements,
connected to each other at specified points / nodes.

Select interpolation functions
Choose the type of interpolation function to represent the
variation of the field variables over the element.

Derive element characteristic matrices and vectors
Employ direct, variational, weighted residual or energy
balance approach.
[k](e) {φ}(e) = {f}(e)

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 21

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

Steps > (Continued)

Assemble the element characteristic matrices and vectors

 Combine the element matrix equations and form the matrix
equations expressing the behavior of the entire solution region /
system.
 Modify the system equations to account for the boundary conditions
of the problem.

Solve the system equations

Solve the set of simultaneous equations to obtain the unknown
nodal values of the field variables.

Make additional computations, if desired

Use the resulting nodal values to calculate other important
parameters.

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 22

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

What is the problem? P

Solution region of
interest

Stress concentration along a shaft

Other examples:
-Scratches on a tensile surface
-Oil groove on a shaft
-Threaded connections Task:
-Rivet holes under tension To simulate stress and strain fields in the
vicinity of a sharp notch under tensile load

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 23

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS
FE procedures:

Select the solution domain

Discretize the continuum

Choose inerpolation functions

Derive element characteristic
matrices and vectors

Assemble element
characteristic matrices and
vectors

Solve the system equations

Make additional
computations, if desired
INTRODUCTION AND OVERVIEW
FE software user steps:

Draw the model geometry

Mesh the model geometry

Select element type

(The FEA software was written to do this)
Input material properties

(The computer will assemble it)
Input specified load and boundary conditions

(The compiler will solve it)
Request for output

Post-process the result file
M.N. Tamin, CSMLab, UTM 24
MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

EXAMPLE: Stresses in a C(T) specimen

Draw the model geometry

Mesh the model geometry

Select element type
[20-node Hexahedral elements]

Input material properties
[Elastic modulus, Poisson’s ratio]

Input specified load and BC
[Pin loading, displacement rate,
zero displacement at lower pin]

Request for output
[Displacement, strain and stress
components]

Post-process the result file

INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 25

MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

EXAMPLE: Stresses in a C(T) specimen

Check for obvious mistakes/errors

Extract useful information

Explain the physics/ mechanics

Validate the results
INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 26
MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

MODELING CAPABILITIES – Microelectronic device reliability

- Prediction of spatial distribution of
physical parameters

Heat sink

Silicon Die
Solder Substrate
mask Solder joints

Motherboard PCB εin

0.10
T
183 - Prediction of damage 0.08

125 evolution characteristics 0.06

TD1

TC1
0.04
25 MT1
0.02
-40 Time
Re-flow 0.00
0 1 2 3 4 5
N
INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 27
MMJ1153 – COMPUTATIONAL METHOD IN SOLID MECHANICS

MODELING CAPABILITIES – Deflection of composite laminates beam

wx 3
EI y = (x − L)
6

Al Comp-0º Comp-90º
E E11 E22
• 14 layers [0/45/90/-45/45/-45/02]s
70 GPa 44.74 GPa 12.46 GPa
• Layer thickness = 0.3571 mm
INTRODUCTION AND OVERVIEW M.N. Tamin, CSMLab, UTM 28