Week 8 MECH3361

1
Recap
Contact Problems
Cases Tangential def Normal Displacement
contact stress



* ) ( 4
3
2 1
2 1 3
E
P
R R
R R
a
+
=

1
2
2
2
1
2
1
1 1
*

|
|
.
|

\
|

+

=
E E
E
v v


2
2 1
2 1 3
*
) (
16
9
|
.
|

\
|
+
=
E
P
R R
R R
o

2
2
0
1
a
r
q q =

( ) P E
R R
R R
q
2
2
2 1
2 1
3
3
0
*
6
|
|
.
|

\
| +
=
t



* 4
3
1
3
E
P
R a =


2
1
3
*
1
16
9
|
.
|

\
|
=
E
P
R
o

2
2
0
1
a
r
q q =

( ) P E
R
q
2
2
1
3
3
0
*
1 6
|
|
.
|

\
|
=
t



* ) ( 4
3
1 2
2 1 3
E
P
R R
R R
a

=


2
2 1
1 2 3
*
) (
16
9
|
.
|

\
|
=
E
P
R R
R R
o

2
2
0
1
a
r
q q =

( ) P E
R R
R R
q
2
2
2 1
1 2
3
3
0
*
6
|
|
.
|

\
|
=
t



* ) (
4
2 1
2 1 2
E
p
R R
R R
b
+
=
t

|
.
|

\
|
+

+
|
.
|

\
|
+

=
407 . 0
2
ln
1
407 . 0
2
ln
1
2
2
2
2
2
1
1
2
1
3
b
R
E
b
R
E p
v
v
t
o

2
2
0
1
b
x
q q =

p E
R R
R R
q *
1
2 1
2 1 2
0
|
|
.
|

\
| +
=
t

Stress Methods (Flowchart): Select stress function | (combination of several) Calculate
stresses
2
2
y
xx
c
c
=
|
o ,
2
2
x
yy
c
c
=
|
o ,
y x
xy
c c
c
=
|
o
2
Use B.C. to determine coefficients
Fully determine stress functions Hookes law to determine strains Strain-Displacement
to derive displacement function Disp B.C. to fully determine displacement functions.
Solution to a circular hole in a plate

|
|
.
|

\
|
+ =
|
|
.
|

\
|
+
|
|
.
|

\
|
+ =
|
|
.
|

\
|
+ +
|
|
.
|

\
|
=
u o
u o
u o
u
uu
2 sin
2 3
1
2
2 cos
3
1
2
1
2
2 cos
4 3
1
2
1
2
2
2
4
4
4
4
2
2
2
2
4
4
2
2
r
a
r
a T
r
a T
r
a T
r
a
r
a T
r
a T
r
rr

When 1 2 cos = u ,
uu
o reaches the maximum ( ) T a r 3
max
= = =
uu uu
o o
Strain energy density (SED) | | 2 / ) ( 2
zx zx yz yz xy xy zz zz yy yy xx xx
U c o c o c o c o c o c o + + + + + =
Distortion energy density (DED) | |
2
1 3
2
3 2
2
2 1
) ( ) ( ) (
6
1
o o o o o o
v
+ +
+
=
E
U
d



T
a
b
u
T
Week 8 MECH3361
2
x
y
z
o
nn
Chapter 7 Plasticity and Failure

7.1 Introduction
- Plasticity should be avoided and most of mechanical system should work in elastic
regime (e.g. gear and spring).
- Plastic permanent deformation should be made in metal forming.

Questions:
- Are the basic equations still valid?
- Criteria to enable us to present
- Relationship of stress strain
- Can we still use the methodology established in the sites


7.2 Octahedral Shear Stress

Definition:
Considering the principal directions as the coordinate axes, a
special oriented plane which normal vector makes equal
angles with each of the principal axes, i.e. having direction
cosines equal to 3 / 1 = = = n m l , is called an octahedral
plane.

Octahedral normal stress
o
o
From Chapter 1 stress transformation, we have
zx zy xy zz yy xx nn
nl mn lm n m l o o o o o o o 2 2 2
2 2 2
+ + + + + =
Since the octahedral plane is rotated from the principal stress direction whose shear
components are zero, thus we can have
) (
3
1
3
1
3
1
3
1
3 2 1
2
3
2
2
2
1
2
3
2
2
2
1
o o o o o o o o o o + + = |
.
|

\
|
+ |
.
|

\
|
+ |
.
|

\
|
= + + = n m l
nn

Octahedral normal stress:
) (
3
1
3 2 1
o o o o o + + = =
nn o

o
o is the mean principal stress or hydrostatic stress. This value is the same in all eight
octahedral planes.

Octahedral shear stress
The octahedral shear stress
0
t = resultant shear stress on an octahedral plane

Total stress on the Octahedral plane:

( ) ( ) ( ) ( ) ( ) ( )
) (
3
1
3
1
3
1
3
1
2
3
2
2
2
1
2
3
2
2
2
1
2
3
2
2
2
1
2 2 2 2
o o o o o o
o o o o o o o
+ + = |
.
|

\
|
+ |
.
|

\
|
+ |
.
|

\
|
=
+ + = + + = n m l
nz ny nx


The octahedral shear stress
0
t can be calculated from the total stress and normal stress as
Week 8 MECH3361
3
| |
| |
) 2 ( ) 2 ( ) 2 (
3
1
2 2 2 ) ( 2
3
1
2 2 2
9
1
) (
9
3
) (
3
1
) (
3
1
2
3 3 2
2
2
2
3 3 1
2
1
2
2 2 1
2
1
3 1 3 2 2 1
2
3
2
2
2
1
3 1 3 2 2 1
2
3
2
2
2
1
2
3
2
2
2
1
2
3 2 1
2
3
2
2
2
1
2 2
o o o o o o o o o o o o
o o o o o o o o o
o o o o o o o o o o o o
o o o o o o o o t
+ + + + + =
+ + + + =
+ + + + + + + =
(

+ + + + = =
o o

Octahedral shear stress:
2
1 3
2
3 2
2
2 1
) ( ) ( ) (
3
1
o o o o o o t + + =
o


7.2 Distortion Energy

Work-energy principle:
External work = Internal Strain energy (if no heat energy loss)

Strain energy density (strain energy per unit volume) (SED)
| | ) ( 2
2
1
zx zx yz yz xy xy zz zz yy yy xx xx
U c o c o c o c o c o c o + + + + + =

Apply the Hookes law to replace strain by stress:
| | ) )( 1 ( 2 ) ( 2 ) (
2
1
2 2 2 2 2 2
zx yz xy xx zz zz yy yy xx zz yy xx
E
U o o o v o o o o o o v o o o + + + + + + + + =
Or apply the Hookes law to replace stress by strain:
( )
(

+ + + + + +

= ) ( 2 ) (
2 1
2 2 2 2 2 2
2
1 zx yz xy zz yy xx
I G U c c c v c c c
v
v
c


When the x, y, z are principal directions, U can be express in terms of principal stresses

| |
( ) | |
d v
U U
E
I K
E
U
+ =
+ +
+
+ =
+ + + + =
2
1 3
2
3 2
2
2 1
2
1
1 3 3 2 2 1
2
3
2
2
2
1
) ( ) ( ) (
6
1
2
1
) ( 2 ) (
2
1
o o o o o o
v
o o o o o o v o o o
c

where ( )
2
1
2
1
c
I K U
v
= ,
c
1
I is the first strain invariant and K is bulk modulus.
Since
c
1
I =volume change rate, therefore ( )
2
1
2
1
c
I K U
v
= represents SED due to volume
change caused by hydrostatic stress.
| |
2
1 3
2
3 2
2
2 1
) ( ) ( ) (
6
1
o o o o o o
v
+ +
+
=
E
U
d

corresponds the shape change due to shear stress (recall the octahedral shear stress),
namely distortion energy density (DED)



Week 8 MECH3361
4
7.3 Plasticity under Simple Tension

Typical tensile testing diagram:
The yielding occurs after yielding limit in the test as shown.
Unloading leads to plastic strain (permanent strain) and
elastic strain (recoverable strain).


Beyond this yielding point, material may behave in
different ways of the plastic deformation as shown below:

Mild steel Stainless steel Al alloy

To simplify such curves, people introduced some simplified models:


Remarks:
- When the unloading takes place after plasticity has occurred, the deformation will be
completely elastic. Thus we can use the method of elastic analysis to study the stress
and deformation during unloading.
- After unloading, then load it again. The strain-stress behaviour will be first elastic. If
the material has a hardening behaviour, it will not yield until the reloading stress
reaches new yield stress, namely Y. Since Y is higher than Y, the material in
reloading shows a hardening behaviour.
- Increase in yield stress and loading-reloading phenomena is called work-hardening

The bilinear model sometime is not accurate enough. For this reason some nonlinear models
may be needed.
Power law:
n
Ac o =
where material constants A= strength coefficient and n=strain-hardening exponent.
Rambery-Osgood model:
n
E
k
E
|
.
|

\
|
+ =
o o
o
K and n are material constants to be determined by experiments.
Curve fitting: Sometime fit the testing curves is needed as

+ + +
=
m
m
A A A
E
c c
c
o

1 0

Week 8 MECH3361
5
Example 7.1
A cylinder bar made of steel is subjected to a uniform tension P along its axis. The maximum
axial strain was measured to be 005 . 0
max
= c .
Youngs modulus E=210GPa and its yield stress is
Y=600MPa and the materials stress-strain curve
shows almost no hardening. (a) Calculate the plastic
strain in the bar at the above max strain. (b) If the bar
is unloaded at this max strain, what is the residual
strain remaining in the bar after complete unloading
(short answer question, 2009 Final Exam).

Soln:
(a) No hardening, the steel can be idealised as an elastic/perfectly plastic material as shown.
003 . 0
10 200
10 600
9
6
=

= = =
E
Y
Y e
c c
Hence at the maximum strain of 005 . 0
max
= c , the plastic strain in the tested bar is
002 . 0 003 . 0 005 . 0
max
= = =
e
p
c c c
(b) When unloading, all the elastic strain will disappear and the plastic strain remains. Thus
the residual strain after complete unloading is equal to 002 . 0 =
p
c .

Example 7.2 For the structure loaded at joint O shown below, determine (a) elastic limit P
e
;
(b) plastic limit load P
p
.

Soln
Step 1: determine internal forces. From the F.B.D. (right sub-figure)

P N N P N N N F
N N N N F
y
x
= + = + + =
= = + =

2 1 3 2 1
3 1 3 1
cos 2 , 0 cos cos : 0
, 0 sin sin : 0
u u u
u u

We cannot solve for three unknowns using the two equations. Other compatibility condition is
needed.
Step 2: Compatibility condition of strain (initial length of Bars 1 and 3 is u cos / ' L L = ):
Elongation of bar 1 and 3 is approximately u cos d ,
( )
u c
u
u c u
c
2
2
2
1
cos
cos /
cos
'
cos
= = =
L
L
L
d

Step 3: Stress inside the bars

= +
=
A P/ cos 2
2 1
3 1
o u o
o o

Step 4: Calculate the stresses
From Hookes law:
1 1
c o E = and
2 2
c o E = u
c
c
o
o
2
2
1
2
1
cos = = u o o
2
2 1
cos =
Plug into stress eqn#2: ( ) A P/ cos cos 2 cos 2
2
2
2 2 1
= + = + o u u o o u o
h h
L
d
u u
1 2 3
P
N
1
N
2
N
3
u u
L
/
c
o
s
u
A
B
C
O
E=Youngs modulus
Y=Yield stress
o
c
c
e
c
p
c
max
Week 8 MECH3361
6
We obtain:
) cos 2 1 (
3
2
u
o
+
=
A
P
Thus:
) cos 2 1 (
cos
3
2
1
u
u
o
+
=
A
P

Step 5: Elastic deformation and initial yielding:
Since u o o
2
2 1
cos = ,
2 1
o o s , which implies that yielding occurs in Bar 2 first.
Lets assume that the corresponding load in this bar reaches the yield limit Y =
2
o . This is
) cos 2 1 (
3
2
u
o
+
=
A
P
) cos 2 1 ( ) cos 2 1 (
3 3
2
u u o + = + = YA A P
e
) cos 2 1 (
3
u + =YA P
e

Step 6: Elastic-plastic deformation and plastic limit load:
Bar 2 reaches elastic limit and undergoes plastic deformation if further increasing the load.
Since the material is elastic/perfectly plastic, we always have Y =
2
o . Thus from
A P/ cos 2
2 1
= +o u o , one can have
) 1 cos 2 ( cos ) ( 2 ) ( + = + = u u YA YA Y A P
This reaches plastic limit where all the bars start undergoing different degrees of plastic
deformation: ) 1 cos 2 ( + = u YA P
p

Step 7: Comparison of Elastic-plastic loading: 1
cos 2 1
cos 2 1
) cos 2 1 (
) 1 cos 2 (
3 3
>
+
+
=
+
+
=
u
u
u
u
YA
YA
P
P
e
p

For example, = 45 u , 41 . 1 =
e
p
P
P
, which means that if we allow the frame to work in its
elastic-plastic deformation regime, the load-carrying capacity is much increased.

7.4 Plasticity under Complex Stress States
Yield Criterion and initial yield condition
It is assumed that when the principal stresses and some
parameters satisfy a critical condition, plastic yielding
occurs. Mathematically,
( ) 0 , , , , , ,
2 1 3 2 1
=
n
k k k F o o o
is called yield criterion or initial yield condition (Initial =
before the critical state is reached, deformation is purely
elastic). If plot it graphically, ( ) 0 , , , , , ,
2 1 3 2 1
=
n
k k k F o o o

is a surface in the space of coordinates
3 2 1
, , o o o (as shown),
namely initial yield surface.

Tresca criterion
Tresca observed that material flow seems to be along the
direction of the maximum shear stress.
Margin ( )
Tresca Tresca
k k F 2 2
3 1 max
= = o o t
Or:
Tresca
k 2
3 1
= o o
is called Tresca criterion.
To determine
Tresca
k , one can conduct an uniaxial tension test, in which metal will yield when
Y =
1
o . Thus: 2 / 2 /
1
Y k
Tresca
= =o
Finally, the Tresca criterion becomes:
Y =
3 1
o o
Week 8 MECH3361
7
von Mises criterion
Experiment has shown that metals do not yield under high hydrostatic stress. Huber (1904)
proposed that plastic yielding occurs when the distortion energy density equal or exceeds that
of the same material under uniaxial tension. In 1913, von Mises suggested that metal will
yield under a combination of the principal stresses

2 2
1 3
2
3 2
2
2 1
2 ) ( ) ( ) (
vm
k = + + o o o o o o
which is called von Mises criterion.
To determine
vm
k , one can conduct an uniaxial tension test in the 1
st
principal stress direction,
where
2
o and
3
o vanish, as.

2 2
1
2
1
2 2
1
2 2 ) 0 ( ) 0 0 ( ) 0 (
vm
k = = + + o o o
Thus Y k
vm
= =o . The von Mises criterion becomes

2 2
1 3
2
3 2
2
2 1
2 ) ( ) ( ) ( Y = + + o o o o o o
Or in Cartesian coordinate system

2 2 2 2 2 2 2
2 ) ( 6 ) ( ) ( ) ( Y
zx yz xy xx zz zz yy yy xx
= + + + + + o o o o o o o o o

Example 7.3
A thin square plate is subjected to a set of uniform stresses on its edges
as show, where =0.5. Y=600MPa. Find the maximum stress o beyond
which plastic deformation appears (Final exam 2012).
Soln
Step 1: determine the principal stresses: Since plane stress problem
0 =
zz
o ,
|
.
|

\
|
+ = +
|
.
|

\
|

+
= +
|
|
.
|

\
|

+
=
2 2
2
2
2
3 , 1
4 1 1
2
) (
2
0
2
0
2 2

o
o
o o
o
o o o o
o
xy
yy xx yy xx

Since
0 0 4 1 1
3
2
s s + o

Thus principal stresses |
.
|

\
|
+ = =
|
.
|

\
|
+ + =
2
3 2
2
1
4 1 1
2
, 0 , 4 1 1
2

o
o o
o
o

Step 2: According to the Tresca criterion Y =
3 1
o o :
Y = + =
|
.
|

\
|
+
|
.
|

\
|
+ + =
2 2 2
3 1
4 1 4 1 1
2
4 1 1
2
o
o

o
o o
Thus the maximum stress must be:
( )
MPa
Y
54 . 422
2
600
5 . 0 4 1
600
4 1
2 2
= =
+
=
+
=

o
Step 3: according to the von Mises criterion
2 2
1 3
2
3 2
2
2 1
2 ) ( ) ( ) ( Y = + + o o o o o o :
) 4 1 ( 4
4
4 1 1
4
4 1 1
4
4 1 1
2
4 1 1
2
4 1 1
2
0 0 4 1 1
2
) ( ) ( ) (
2
2 2
2
2 2
2
2
2
2 2
2
2
2
2
2
1 3
2
3 2
2
2 1

o
o o o o o o
+ + |
.
|

\
|
+ + |
.
|

\
|
+ + =
(

|
.
|

\
|
+ + |
.
|

\
|
+ +
(

|
.
|

\
|
+ +
(

|
.
|

\
|
+ + =
+ +

o o
o
o
Week 8 MECH3361
8
| | | |
2 2 2 2
2
2 2 2 2 2
2
2 3 1 2 ) 4 1 ( 6 2
4
) 4 1 ( 4 4 1 2 ) 4 1 ( 1 4 1 2 ) 4 1 ( 1
4
Y = + = + + =
)
`

+ +
(

+ + + +
(

+ + + + =
o
o

o

MPa
Y
56 . 453
75 . 1
600
5 . 0 3 1
600
3 1
2 2
= =
+
=
+
=

o

Tresca criterion is lower in this case.

7.5 Failure Theories
Ductile materials:
- The maximum Shear Stress Theory:

o
o o
Y
s
3 1

where o is the safety factor

- The maximum distortion energy theory:

o
o o o o o o
2
2
1 3
2
3 2
2
2 1
2
) ( ) ( ) (
Y
= + +

Brittle materials
There is no yield for brittle material in general.
- The maximum normal stress theory

o
o
o
f
s
1

f
o is the failure normal stress

- The maximum normal strain theory

o
c
c
f
s
1

f
c is the failure normal strain.
Since from Hookes law: | | ) (
1
3 2 1 1
o o v o c + =
E


o
o
o o v o
f
s + ) (
3 2 1


Week 8 MECH3361
9
Chapter 8 Finite Element Method

8.1 Introduction
Concept
The finite element method (FEM), or finite element analysis (FEA), is based on the idea of
building a complicated object with simple blocks, or, dividing a complicated object into small
and manageable pieces.
Application of this simple idea can be found everywhere in daily life and engineering.
- Lego (kids toy) for complex shapes/patterns
- Bricks for Buildings
- Approximation of the area of a circle:
Triangular area of each element: u sin
2
1
2
R S
i
= .
) (
2
sin
2
1
2
1
2
= =

=
N when R
N
R N S S
N
i
i N
t
t


Why Finite Element Method?
- Applications of analytical solutions are limited. Experimental solutions are expensive
and time-consuming.
- Design analysis: hand calculations, experiments, and computer simulations
- FEM/FEA is the most widely applied computer simulation method in engineering
- Closely integrated with CAD/CAM applications

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

A Brief History of the FEM
- 1943 Courant (Variational methods)
- 1956 Turner, Clough, Martin and Topp (Stiffness)
- 1960 Clough (Finite Element, plane problems)
- 1970s Applications on mainframe computers
- 1980s Microcomputers, pre- and postprocessors
- 1990s Analysis of large structural systems

Procedures
- Divide structure into pieces (elements with nodes)
- Describe the behaviours of the physical quantities on each element
- Connect (assemble) the elements at the nodes to form an approximate system of
equations for the whole structure
- Solve the system of equations involving unknown quantities at the nodes (e.g.,
displacements)
- Calculate desired quantities (e.g., strains and stresses) at selected elements

Week 8 MECH3361
10
Computer Implementations
- Preprocessing (build FE model, material properties, loads and boundary conditions)
- FEA solver (assemble and solve the system of equations)
- Postprocessing (sort and display the results)

Commercial FEM Packages
- ANSYS (General purpose, PC and workstations)
- ABAQUS (General purpose, Nonlinear and dynamic analyses)
- Strand7 (General purpose, fast solver and user-friendly)
- Cosmosworks

Type of elements


8.2 Spring Element

Step 1: Single Spring System
- Two nodes: i, j
- Nodal displacements: u
i
, u
j
(m) (x-displacement only)
- Nodal forces:
j i
f f , (N)
- Spring constant (stiffness) k (N/m)
Spring force-displacement relation: A = k f
spring

where elongation
i j
u u = A
Consider the equilibrium of forces for the spring.
At node i, we have : 0 = +
spring i
f f
j i i j spring i
ku ku u u k k f f = = A = = ) (
At node j, : 0 =
spring j
f f
j i i j spring j
ku ku u u k k f f + = = A = = ) (
In a matrix form:
)
`

=
)
`

j
i
j
i
f
f
u
u
k k
k k

Or: f ku=
where: k = elemental stiffness matrix
u = (element nodal) displacement vector
f = (element nodal) force vector
Note that k is symmetric. Is k singular or nonsingular?

Multiple spring system
For element 1:

=
)
`

) 1 (
2
) 1 (
1
2
1
1 1
1 1
f
f
u
u
k k
k k

node
element
node
(a) 1D (b) 2D (c) 3D
f
s
p
r
i
n
g
Week 8 MECH3361
11
For element 2:

=
)
`

) 2 (
2
) 2 (
1
3
2
2 2
2 2
f
f
u
u
k k
k k

where
) (
1
m
f and
) (
2
m
f is the (internal) force acting on local node i of element m (i = 1, 2).

Consider equilibrium of the spring system at nodes: (F
1
, F
2
, F
3
are total forces in the nodes)
- At node 1:
) 1 (
1 1
f F =
- At node 2:
) 2 (
1
) 1 (
2 2
f f F + =
- At node 3:
) 2 (
2 3
f F =
Thus:

+ =
+ + = + + =
=
3 2 2 2 3
3 2 2 2 1 1 1 3 2 2 2 2 1 1 1 2
2 1 1 1 1
) ( ) ( ) (
u k u k F
u k u k k u k u k u k u k u k F
u k u k F

Or in a total matrix for:

(
(
(

+

3
2
1
3
2
1
2 2
2 2 1 1
1 1
0
0
F
F
F
u
u
u
k k
k k k k
k k

i.e. F KU =
where K is named the global stiffness matrix (structure matrix) for the spring system.

An alternative way of assembling the whole stiffness matrix: Enlarging the stiffness
matrices for elements 1 and 2 according to the nodal numbers, we have:
Element #1:

=
)
`

) 1 (
2
) 1 (
1
2
1
1 1
1 1
f
f
u
u
k k
k k

3 #
2 #
1 #
0 0 0 0
0
0
) 1 (
2
) 1 (
1
3
2
1
1 1
1 1
Node
Node
Node
f
f
u
u
u
k k
k k

(
(
(

Element #2:

=
)
`

) 2 (
2
) 2 (
1
3
2
2 2
2 2
f
f
u
u
k k
k k

3 #
2 #
1 # 0
0
0
0 0 0
) 2 (
2
) 2 (
1
3
2
1
2 2
2 2
Node
Node
Node
f
f
u
u
u
k k
k k

(
(
(

Add these two expanded equations:

|
|
|
.
|

\
|
(
(
(

+
(
(
(

) 2 (
2
) 2 (
1
) 1 (
2
) 1 (
1
3
2
1
2 2
2 2 1 1
1 1
0
0 0
0
0 0 0
0 0 0
0
0
f
f f
f
u
u
u
k k
k k k k
k k

Thus

+ =

(
(
(

+

3
2
1
) 2 (
2
) 2 (
1
) 1 (
2
) 1 (
1
3
2
1
2 2
2 2 1 1
1 1
0
0
F
F
F
f
f f
f
u
u
u
k k
k k k k
k k

Boundary and load conditions (for a problem as shown below)

Fix the left end 0
1
= u and the spring system is loaded by external forces: P F F = =
3 2

Reaction
force F
u
1
=0 u
2
, P u
3
, P
Week 8 MECH3361
12

(
(
(

+

P
P
F
u
u
k k
k k k k
k k
1
3
2
2 2
2 2 1 1
1 1
0
0
0

Because of zero displacement in u
1
=0, the above matrix Eq can be separated (equivalent) into

)
`

=
)
`

+
P
P
u
u
k k
k k k
3
2
2 2
2 2 1

and
1 2 1
F u k =
The unknowns are
)
`

=
3
2
u
u
U and reaction force F
1

Solve the equation set, we obtain the displacements:
)
`

+
=
)
`

2 1
1
3
2
/ / 2
/ 2
k P k P
k P
u
u

Thus the reaction force: P u k F F 2
2 1 1
= = =

Remarks
- Both elemental stiffness and global stiffness matrices are symmetric and positive
definite.
- Check the equilibrium of the external forces
- Suitable for other stiffness analysis (e.g. torsional spring etc)

Example 8.1
For a three-spring system shown, mm N k / 100
1
= , mm N k / 200
2
= , mm N k / 100
3
= ,
N P 500 = and 0
4 1
= = u u . Determine: (a) global stiffness matrix, (b)
2
u and
3
u , (c) reaction
forces at nodes 1 and 4, (d) force in spring #2.

Solution
Step 1: Elemental stiffness matrices
(

=
100 100
100 100
1
k ,
(

=
200 200
200 200
2
k ,
(

=
100 100
100 100
3
k
Step 2: Expand the elemental stiffness matrices according to the node numbers

4 3 2 1
0 0 0 0
0 0 0 0
0 0 100 100
0 0 100 100

1
u u u u
(
(
(
(

= k ,
(
(
(
(

=
0 0 0 0
0 200 200 0
0 200 200 0
0 0 00 0

2
k ,
(
(
(
(

=
100 100 0 0
100 100 0 0
0 0 0 0
0 0 0 0

3
k
Step 3: Superposition (add) all elemental stiffness matrices for global stiffness matrix

(
(
(
(

+
+

= + + =
100 100 0 0
100 100 200 200 0
0 200 200 100 100
0 0 100 100

3 2 1
k k k K
Week 8 MECH3361
13
Thus the global stiffness matrix:
(
(
(
(

=
100 100 0 0
100 300 200 0
0 200 300 100
0 0 100 100
K
which is symmetric and banded.

Step 4: Global equilibrium equation system

+
+
=

(
(
(
(

4
1
) 3 (
2
) 3 (
1
) 2 (
2
) 2 (
1
) 1 (
2
) 1 (
1
4
3
2
1
0
100 100 0 0
100 300 200 0
0 200 300 100
0 0 100 100
F
P
F
f
f f
f f
f
u
u
u
u

Step 5: Apply boundary condition 0
4 1
= = u u

(
(
(
(

4
1
3
2
0
0
0
100 100 0 0
100 300 200 0
0 200 300 100
0 0 100 100
F
P
F
u
u

Delete (or separate) the 1
st
and 4
th
rows and columns, we have non-vanishing equations

)
`

=
)
`

P u
u 0
300 200
200 300
3
2
i.e.

= +
=
P u u
u u
3 2
3 2
300 200
0 200 300


Step 6: Solve for the linear equation set to obtain unknown displacements
Solve for the equations: ) (
3
2
500 / 3
250 /
3
2
mm
P
P
u
u
)
`

=
)
`

=
)
`

Step 7: Reaction forces
From the dropped equations:

=
=
4 3
1 2
100
100
F u
F u

= = =
= = =
N u F
N u F
300 3 100 100
200 2 100 100
3 4
2 1


Example 8.2
For the spring system with arbitrarily numbered
nodes and elements, as shown above, find the
global stiffness matrix.

Solution
Step 1: Construct the element connectivity (4
elements and 5 nodes)

which specifies the global node numbers corresponding to the local node numbers for each
element.
Week 8 MECH3361
14
Step 2: Elemental stiffness matrices
2 4
1 1
1 1
1
u u
k k
k k
(

= k

3 2
2 2
2 2
2
u u
k k
k k
(

= k

5 3
3 3
3 3
3
u u
k k
k k
(

= k

1 2
4 4
4 4
4
u u
k k
k k
(

= k

Step 3: Expand the elemental stiffness matrices
For the un-ordered stiffness matrices (the stiffness matrix should be positive definite)
2 4
1 1
1 1
1
u u
k k
k k
(

= k

= +
=
) 1 (
2 2 1 4 1
) 1 (
1 2 1 4 1
f u k u k
f u k u k

= +
=
) 1 (
1 4 1 2 1
) 1 (
2 4 1 2 1
f u k u k
f u k u k

4 2
1 1
1 1
1
u u
k k
k k
(

= k

1 2
4 4
4 4
4
u u
k k
k k
(

= k

2 1
4 4
4 4
4
u u
k k
k k
(

= k

5 4 3 2 1
1 1
1 1
1
0 0 0 0 0
0 0 0
0 0 0 0 0
0 0 0
0 0 0 0 0
u u u u u
k k
k k
(
(
(
(
(
(

= k

5 4 3 2 1
2 2
2 2
2
0 0 0 0 0
0 0 0 0 0
0 0 0
0 0 0
0 0 0 0 0
u u u u u
k k
k k
(
(
(
(
(
(

= k

5 4 3 2 1
3 3
3 3 3
0 0 0
0 0 0 0 0
0 0 0
0 0 0 0 0
0 0 0 0 0
u u u u u
k k
k k
(
(
(
(
(
(

= k

5 4 3 2 1
4 4
4 4
4
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0
0 0 0
u u u u u
k k
k k
(
(
(
(
(
(

= k

Step 4: Superposition (add) for global stiffness matrix

5 4 3 2 1
3 3
1 1
3 3 2 2
1 2 4 2 1 4
4 4
0 0 0
0 0 0
0 0
0
0 0 0
u u u u u
k k
k k
k k k k
k k k k k k
k k
(
(
(
(
(
(

+
+ +

= K


8.2 Bar Elements
Bar Elements in 1-D
Known: L-length of bar, A-cross-sectional area, E-
Youngs modulus, displacement ) (x u u = , strain
) (x c c = , stress ) (x o o =
- Strain-displacement:
dx
du
= c
- Hookes law: c o E =
Week 8 MECH3361
15
Finite Element Method 1
Step 1: Assuming that the displacement u is varying linearly along the axis of the bar, i.e.,

j i
u
L
x
u
L
x
x u +
|
.
|

\
|
= 1 ) (
in which, when
j j i i j i
u u u L x u L x u u u x u x = + = = = = + = = = 1 0 ) ( , ; 0 1 ) 0 ( , 0
Step 2: Calculate strain and stress

L
u u
L
i j

= =
A
c (A=elongation = relative displacement between node i and j)

L
E
E
A
= = c o
For a bar,
A
F
= o A A
A
o k
L
EA
A
L
E
A F =
|
.
|

\
|
=
|
.
|

\
|
= = , where stiffness
L
EA
k =
Step 3: Elemental stiffness matrix:
The bar is acting like a spring in this case and we conclude that element stiffness matrix is

(

=
(
(
(

=
(

=
1 1
1 1
L
EA
L
EA
L
EA
L
EA
L
EA
k k
k k
k
Step 4: Equilibrium equation:
)
`

=
)
`

j
i
j
i
f
f
u
u
L
EA
1 1
1 1

Remarks
- Degree of Freedom (dof) - Number of components of displacement vector at a node.
For 1-D bar element: one dof at each node (i.e. the u displacement).
- Physical Meaning of the Coefficients in k
The jth column of k (here j = 1 or 2) represents the forces applied to the bar to
maintain a deformed shape with unit displacement at node j and zero displacement at
the other node.

Finite Element Method 2 A Formal Approach
Lets derive the same stiffness matrix for the bar using a formal approach which can be
applied to many other more complicated situations.
Step 1: Displacement: Define two linear shape functions (interpolation function) as follows
=1 ) (
i
N , = ) (
j
N
where natural coordinate:
L
x
= ( 1 0 s s ). As such, we change the coordinate system from
[x
1
,x
2
] to [0,1]
Express displacement as
j j i i
u N u N u x u ) ( ) ( ) ( ) ( + = = (interpolation of displacement)
where when 0 = ,
i j i j j i i
u u u u N u N u = + = + = = 0 1 ) 0 ( ) 0 ( ) 0 (
when 1 = ,
j j i j j i i
u u u u N u N u = + = + = = 1 0 ) 1 ( ) 1 ( ) 1 (
in a matrix form: Nu =
)
`

=
j
i
j i
u
u
N N u ] [ , where
)
`

=
j
i
u
u
u is nodal displacement vector.
Step 2: Strain-displacement
( ) Bu u N Nu =
|
.
|

\
|
= = =
dx
d
dx
d
dx
du
c (note that vector u is not a function of x)
Week 8 MECH3361
16
where B is the elemental strain-displacement matrix.

| | | |
| | | |
(

= =
|
.
|

\
|
=
= =
|
.
|

\
|
=
L L L L
x
dx
d
d
d
dx
d
N N
d
d
N N
dx
d
dx
d
j i j i
1 1 1
1 1 1
) ( ) ( ) ( ) (

N B

Step 3: Stress (Hookes law)
Bu E E = = c o
Step 4: Apply Strain energy-work principle
( ) u B B u (Bu) ) B (u (Bu) Bu) (
(

= = = =
} } } }
V
T T
V
T T
V
T
V
T
dV E dV E dV E dV U
2
1
2
1
2
1
2
1
c o
The work done by the two nodal forces is: { } f u
T
2
1
2
1
2
1
2
1
=
)
`

= + =
j
i
j i j j i i
f
f
u u u f u f W
Since W U = we have ( ) f u u B B u
T
2
1
2
1
=
(

}
V
T T
dV E
Thus: ( ) f u B B =
(

}
V
T
dV E
This is f ku=
where ( )
}
=
V
T
dV EB B k is the stiffness matrix

( )
(

=
(
(
(

=
(
(
(

=
(
(
(

=
(

=
}
} }
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1 1 1
2 2
2 2
2 2
2 2
2 2
2 2
L
EA
L
L L
L L
EA dx
L L
L L
EA
dx
L L
L L
EA Adx
L L
E
L L
V
V V
T
k

which is the same as we derived using the direct method.
This is a general result which can be used for the construction of other types of elements. This
expression can also be derived using other more rigorous approaches, such as the Principle of
Minimum Potential Energy, or the Galerkins Method.
Note that the strain energy in the element can be written as
Ku u
T
V
T
dV U
2
1
2
1
= =
}
c o
Example 8.3 (2011 Quiz question) The three bar structure is fully clumped at ends A and D as
shown. Solve the following problems by using finite element method. (1) Write the elemental
stiffness matrices. (2) Give the global equilibrium equation. (3) Calculate the displacement at
nodes B and C. (4)Calculate the reaction forces at points A and D.

Solution
Step 1: Connectivity: Elements: AB=(1), BC=(2), CD=(3). Nodes: A=1, B=2, C=3, D=4.
1m 1m 1m
EA=2
EA=1
EA=2
x
P
2P
A B C D
Week 8 MECH3361
17
Step 2: Elemental stiffness matrices of bars
Bar elements:
(

=
(

=
1 1
1 1
L
EA
k k
k k
k
Thus:
(

=
2 2
2 2
1
k ,
(

=
1 1
1 1
2
k ,
(

=
2 2
2 2
3
k
Step 3: Expand the elemental stiffness matrices according to unknown vector u:

=
4
3
2
1
u
u
u
u
u

4 3 2 1
0 0 0 0
0 0 0 0
0 0 2 2
0 0 2 2

1
u u u u
(
(
(
(

= k ,
(
(
(
(

=
0 0 0 0
0 1 1 0
0 1 1 0
0 0 0 0

2
k ,
(
(
(
(

=
2 2 0 0
2 2 0 0
0 0 0 0
0 0 0 0

3
k
Step 4: Superposition (add) all elemental stiffness matrices for global stiffness matrix

i.e. the global stiffness matrix:
(
(
(
(

=
2 2 0 0
2 3 1 0
0 1 3 2
0 0 2 2
K
, which is symmetric and banded
Step 5: Equilibrium equation system: f Ku=

(
(
(
(

4
3
2
1
4
3
2
1
2 2 0 0
2 3 1 0
0 1 3 2
0 0 2 2
F
F
F
F
u
u
u
u

Step 6: Apply boundary condition 0
4 1
= = u u

(
(
(
(

4
1
3
2
2
0
0
2 2 0 0
2 3 1 0
0 1 3 2
0 0 2 2
F
P
P
F
u
u

Delete the 1
st
and 4
th
rows and columns,
)
`

=
)
`

P
P
u
u
2 3 1
1 3
3
2
i.e.

= +
=
P u u
P u u
2 3
3
3 2
3 2

Step 7: Solve for the linear equation set to obtain displacements
Solve for the equations:
)
`

=
)
`

8 / 7
8 / 5
3
2
P
P
u
u

Step 8: Reaction forces: From the dropped equations:

=
=
4 3
1 2
2
2
F u
F u

= = =
= = =
4
7
8
7
2 2
4
5
8
5
2 2
3 4
2 1
P P
u F
P P
u F

(
(
(
(

+
+

= + + =
2 2 0 0
2 2 1 1 0
0 1 1 2 2
0 0 2 2

3 2 1
k k k K