HANOI UNIVERSITY OF SCIENCE AND TECHNOLOGY

UNIVERSITY OF MECHANICS

Finite Element Methods

Prepared by
Prof. M.C. Nguyen, Dept of Mechatronics

1
 UNIT - 1
• Introduction to FEM:
– Stiffness equations for a axial bar element in local co-ordinates
using Potential Energy approach and Virtual energy principle
– Finite element analysis of uniform, stepped and tapered bars
subjected to mechanical and thermal loads
– Assembly of Global stiffness matrix and load vector
– Quadratic shape functions
– properties of stiffness matrix

2
 Axially Loaded Bar
Review:

Stress: Stress:

Strain: Strain:

Deformation: Deformation:

3
 Axially Loaded Bar
Review:

Stress:

Strain:

Deformation:

4
 Axially Loaded Bar – Governing Equations
and Boundary Conditions

• Differential Equation
d  du 
 EA(x)   f (x)  0 0  x L
dx  dx 
• Boundary Condition Types
• prescribed displacement (essential BC)

•prescribed force/derivative of displacement

(natural BC)

5
 Axially Loaded Bar –Boundary Conditions

• Examples
• fixed end

• simple support

• free end

6
 Potential Energy
• Elastic Potential Energy (PE)
- Spring case
Unstretched spring

PE  0
Stretched bar
1 2
PE  kx
2
x
- Axially loaded bar

undeformed: PE  0
L
deformed: PE   Adx
1
20
- Elastic body


1 T
PE  σ εdv
2V
7
 Potential Energy
• Work Potential (WE)
f
P f: distributed force over a line
A B P: point force
L
u: displacement
WP    u  fdx  P  u B
0

• Total Potential Energy

L L
  Adx   u  fdx  P  uB
1
20 0

• Principle of Minimum Potential Energy

For conservative systems, of all the kinematically admissible displacement fields,
those corresponding to equilibrium extremize the total potential energy. If the
extremum condition is a minimum, the equilibrium state is stable.
8
Potential Energy + Rayleigh-Ritz Approach
Example:
f
P
A B

Step 1: assume a displacement field u   aii x  i  1 to n

i
 is shape function / basis function
n is the order of approximation

Step 2: calculate total potential energy

9
 Potential Energy + Rayleigh-Ritz
Approach
Example:
f
P
A B

Step 3:select ai so that the total potential energy is minimum

10
 Galerkin’s Method
Example:
f
P
A B
d  du  Seek an approximation u~so
EA(x)   f ( x)  0
dx  dx  d  du~  
ux  0  0 V i  dx 
w  EA( x)
dx   f ( x)dV

0

EA( x)
du
P u~x 0  0
dx xL du~
EA( x) P
dx xL

In the Galerkin’s method, the weight function is chosen to be the same as the shape
function.

11
 Galerkin’s Method
Example:
f
P
A B

d  du~   du~dw i du~

L L L

  dx 
wi  EA( x)   f ( x)dV 0  EA( x) dx   wi fdx wi EA( x) 0
V
dx   0
dx dx 0
dx 0
1 2 3

3
12
Finite Element Method – Piecewise
Approximation
u

13
 FEM Formulation of Axially Loaded
Bar – Governing Equations
• Differential Equation
d  du 
EA(x)   f (x)  0 0  x L
dx  dx 

• Weighted-Integral Formulation
w d EA(x) du   f (x) dx
 0
L
 0
 dx  dx  

• Weak Form
L
L
 dw  du    du 
0     EA(x)   wf (x)dx  w EA(x) 
0
dx  dx    dx  0
14
 Approximation Methods – Finite
Element Method
Example:

Step 1: Discretization

Step 2: Weak form of one element P2

P1
x1 x2
x
x2
 dw  du    du  2
 dx  EA( x) dx   w( x) f ( x)dx  w( x)EA( x) dx   0
x1 x1

x2
 dw  du  
x  dx 
 EA( x)   w( x) f ( x) dx  w x 2 P2  w x1 P1  0
1
dx   15
 Approximation Methods – Finite
Element Method
Example (cont):

Step 3: Choosing shape functions

- linear shape functions u 1u1  2u2
x 
  
x1 l x2

x2 x x  x1 1  1 
1  ; 2  1  ; 2 
l l 2 2

  2 x  x 1 1; x 
  1l  x
1
l 2

16
 Approximation Methods – Finite
Element Method
Example (cont):

Step 4: Forming element equation

E,Aare constant 2 x
EA EA
u2   1 f dx  P1
x2
uu
x2
1 u1 
Let w  1, weak form becomes   l EA  2 l 1 dx   f1dx   P1 2  P1 1 0 l l x1
x1 x1

2 x
1 EA EA
u2 u 1 u2   2 f dx  P2
x2 x2
Let w  2, weak form becomes  u1 
 l  EA  l dx   f2 dx   P2 2  P2 1 0 l l
x 
x1
1 x 1

 x2 
 1 
 fdx
EA  1 1 u1   x1   P1   f1   P1 
        
l  1 1  u2   x2   P2   f 2  P2 
  fdx
2 
 x1 
17
 Approximation Methods – Finite
Element Method
Example (cont):

Step 5: Assembling to form system equation

Approach 1:
 1 1 0 0 u1 I   f1I  P1 I 
      
Element 1: E I A I  1 1 0 0 u2I   f 2I   P2I 
  
l I  0 0 0 0  0   0   0 
 0 0 0 0      
  0  0   0 
0 0 0 0  0   0   0 
Element 2: E II AII 0 1 1 0 u1II   f1II   P1II 
        
l II 0 1 1 0  u2II  f 2II   P2II 
0 0 0 
 0 0   0  0 
0 0 0 00  0  0 
Element 3:
E III AIII 0 0 0 00  0  0 
  
lIII 0 0 1 1u 1III   f1III   P1 III 
     
0 0 1   III   f III   P III  18
 1  u2   2   2 
 Approximation Methods – Finite
Element Method
Example (cont):

Step 5: Assembling to form system equation

Assembled System:

I I
 E I AI E A 
 lI lI
0 0 
 
  E I AI E I AI E II AII
II
E A
II
 u1   f1   P1  f1I   P1I 

 0  u2  2f  2P    f I  f II   P I II 
lI lI l II l II 1   2  P1 
          II
2
   II III 
 E II AII E II AII E III AIII E III AIII   u3   f 3   P3  f
 2  f III
1  P
 2  P 1 
 0  II  III  III
l l II l l  u4   f 4   P4   f 2 III   P2 III 
 
 0 E III AIII E III AIII 
0  III
 l l III 

19
 Approximation Methods – Finite
Element Method
Example (cont):

Step 5: Assembling to form system equation

Approach 2: Element connectivity table Element 1 Element 2 Element 3

kije  K IJ 1 1 2 3

2 2 3 4

local node global node index

(i,j) (I,J)

20
 Approximation Methods – Finite
Element Method
Example (cont):

Step 6: Imposing boundary conditions and forming condense system

Condensed system:  E I AI E II AII

II II

E A
 I  0 
 l l II l II
E A  u   f   0 
 E II AII E II AII E III AIII  III III  2    2    

  l II l II l III
u   f   0 
l III   3   3   
  u  f4  P 
 E III AIII E III AIII   4 
 0  III 
 l l III  21
 Approximation Methods – Finite
Element Method
Example (cont):

Step 7: solution

Step 8: post calculation

d2 d2
u  u11  u 22   du  u 1d1  u 2   E  Eu 1d1  Eu 2
dx dx dx dx dx

22
 Summary - Major Steps inFEM
• Discretization
• Derivation of element equation
•weak form
•construct form of approximation solution
over one element
•derive finite element model
• Assembling – putting elements together
• Imposing boundary conditions
• Solving equations
23
• Postcomputation
 Exercises – Linear Element
Example 1:

E= 100 GPa,A = 1 cm2

24
 Linear Formulation for Bar Element
u u1 f(x) u2

P1 P2
x
x=x1 L= x2-x1 x= x2

P1   f1  K11 K12 u1 

P    f      
 2   2   K12 K 22 u2 

 di d j 
x2 x2

where K ij   EA dx  K ji , f i   if dx


x  dx dx 
1 x1

 1 2 

x=x1 x=x2
x 25
Higher Order Formulation for Bar Element
u u1 u2 u3

x
1 2 3

u(x)  u11 (x) u22 (x) u33 (x)

u u1 u2 u3 u4

x
1 2 3 4

u(x)  u11 ( x )  u22 ( x )  u33 ( x )  u44 ( x )

u u1 u2 u3 u4 …………… un

x
1 2 3 4 …………… n

u(x)  u11 ( x )  u22 ( x )  u33 ( x )  u44 ( x )        unn ( x )

26
Natural Coordinates and InterpolationFunctions
=-1  =1
x x=x1 x= x2
x 0 x l
x  x  x1
x  x1  x2
Natural (or Normal) Coordinate:  
2
l /2
=-1  =1
1 1
1   , 2 
1 2 2 2
=-1  =1    1  1
1  , 2     1  1, 3 
2 2
1 2 3
9  1   
27
    1   1
  1 
       2
1 , 1 

=-1  =1 1 16  3  3 16  3 

2 4 3   27   1  1   1,  4  9  1  1 

  1 
1 3 16  3 16  3  3

27
 Quadratic Formulation for Bar Element
 P1   f1   K 11 K 12 K 13 u1 
P    f   K  u 
 2   2   12 K 22 K 23   2 
P   f  K  u3 
 3   3   13 K 23 K 33 
where Kij  x EA di  ddx
2
 j  EA1 i  d d j d 2  K
 d d  l
ji
x1  dx dx  1

x2 1
l
and f i   i f dx   i f  d , i, j  1, 2, 3
2
x1 1

 


=-1 =0 =1 28

 Quadratic Formulation for Bar Element
u1 f(x) u2 u3
P2
P1 P3
x1 x2 x3
=-1 =0 =1

   1  1
u(  )  u1 1 (  )  u 2 2 (  )  u3 3 (  )  u 1  u2  1  1 u 3
2 2

  1  1

1  , 2    1  1, 3 
2 2
x  x1  x2 l d  dx d  2
2
l /2 2 dx l

d1 2 d1 2 1 d2 2 d2 4 d3 2 d3 2 1

  ,   ,  
dx l d  l dx l d  l dx l d  l
29
 Exercises – Quadratic Element
Example 2:

E= 100 GPa,A1 = 1 cm2; A1 = 2 cm2

30
 Some Issues
Non-constant cross section:

Interior load point:

Mixed boundary condition:

k

31
 UNIT – 2
Finite Element Analysis of Trusses

32
 Plane Truss Problems
Example 1: Find forces inside each member. All members have
the same length.

33
 Arbitrarily Oriented 1-D Bar Element on 2-DPlane
Q2 , v2 P2 , u2
P2, u2

P1, u1 

P1 , u1 Q1 , v1

34
Relationship Between Local Coordinates and Global
Coordinates

 u1   cos sin  0 0 u1 

v  0   sin 
 1   cos 0 0  
  1 
v
  sin   u2 
 2   0
u 0 cos
v2 0  0  
0  sin  cos  v2 

35
 Relationship Between Local Coordinates and Global
Coordinates

 P1  cos sin  0 0  P1

 0  sin   
   cos 0 0  1 
Q
   
P
 2  0 0 cos sin   P2 
0   
 0 0  sin  cos Q2

36
 Stiffness Matrix of 1-D Bar Element on 2-D Plane
Q2 , v2 P2 , u2
P2, u2

 1 0 1 0
 0 0 

P1, u1  AE  0 0
Kij 
L 1 0 1 0
P1, u1 Q1 , v1 0 0 
 0 0 

 P1 cos   sin  0 0   cos sin  0 0 u1 

   0   sin  0  v1 
Q1    sin  cos  cos 
 P
0
cos 
 
 Kij 
 sin  
0
cos

 
sin  u2 
 2  0 0  0 0

Q2   0  
0 sin  cos   0 0  sin  cos v2

 P1  cos2  sin  cos  cos2   sin  cos u1 

Q   v 
 1   AE  sin  cos  sin 2   sin  cos  sin 2  1
 P  
 
L   cos 2   
 sin cos cos 2  sin  cos u2 
2
 
Q2 sin  cos  sin 2  sin  cos sin  v2
2
37
 Arbitrarily Oriented 1-D Bar Element in 3-D Space

z
2
P2 , u 2

-x, -x, -xare the Direction

x- y Cosines of the bar in the x-
-x
y-z coordinate system
1 x-
x
P1 , u1

 u1   x x  x 0 0 0  u1   P1   x x  x 0 0 0  P1
 v  0   y  y 0  v1 
Q  0  
 1   y y  y 0 0 0 Q1 
 1   y 0 0    
 w1  0  z z  z 0 0 0  w 1   R1  0   z z  z 0 0 0  R1 
 0    0  P 
 x 
u 0 0 x x  x  u2  P
 2   0 0 x x 2
 2  
v 2  0  0 0 0 y y  y  v 2  Q2  0  0 0 0 y y  y Q2 
         
w2  0 0 0 0 z z  z w2 R 
 2   0
0 0 0 z z  z R2
38
 Stiffness Matrix of 1-D Bar Element in 3-D Space
z
2
P2 , u 2

 P1  1 0 0  1 0 0  u1 
Q  0  0 0 0 0 0 0  v 1  0 
x-  1   
-x y  R1  0  AE  0 0 0 0 0 0 w 1  0
    
 2 
P L  1 0 0 1 0 0  u2 
1 x-  Q2  0  0 0 0 0 0 0  v2  0 
x     
P1 , u 1 R2 0  0 0 0 0 0 0w2  0

 P1   x2  x x  x x  2 x x  x x   u1 

 
x
Q 
 1   x  x
 2
x  x  x  x  x   x2  x x  v1 
 R1  AE  xxx xx x x x x x   2       2  w1 
    
 P2  L    x
2
x x  x x  x2 x x  x x   u2 
Q2       x2   x x  x  x  x2  x x   v2 
   x x
 
R2   x x   x  x   x2  x x  x x  x  w2 
2

39
Matrix Assembly of Multiple Bar Elements
 P1  1 0 1 0u1 
Q   0 0 0 0  v 
 1 AE    1
Element I  P 1  u 
 2 L  0 1 0  2
Q2   
 0 0 0 0 v2 
 P2   1  3 1 3  u 2 
  
Element II

Q2   AE  3 3 3  3 v 2
 P    
 3 4L 1 3 1  3 u3 
 
Q3  3 3  3 3 v3

 P1  1 3 1  3u1 
Q  

 1   AE  3 3  3  3  v1 
Element III  P    
 3  4L  1  3 1 3 u 3 

Q3  3  3 3 3 v3
40
Matrix Assembly of Multiple Bar Elements
 P1 4 0 4 0 0 u1 
Q    v 
 1  0  1 
 P2  AE  0 0 00 00 u2 
Element I     
Q
  2 4L  4 0 4 0 0 v2 
 P3  u3 
   0  
Q3  0 0 0 0 0 v3
 P1 0 0 0 0 0 0 u1 
Q  0 0 0 00 0  v 
0 
 1  0 0 0 0 0 0
1

 P2  AE 0 0 1  3 1 3 u2 
Element II   0 0 0 0 0 0  
Q
 2 4L 0 3 3 3  3 v2 
 P3 0 0 3u3 
   0 1 3 1  3  
Q3 0 0 3 3  3  v3 
 P1  1 3 0 0 1  3 u1 
Q    
 1  3 3 0 0  3  3  v1 
Element III  P2  AE  0 0 0 0 0 0  u2 
    
 2
Q 4L  0 0 0 0 0 0 v2 
 1 
 P3
   3 0 0 1 3 u 3 
   
Q3  3 3 0 0 3 3 v3 41
 Matrix Assembly of Multiple Bar Elements
 R1   4 1 0 3 4 0 1  3 u1 
S  v 
 1 0  3 0 3 0 0  3 3  1 
R2  AE   4 0 41 0 3 1 3  u2 
S     
 2  4L  0 0 0 3 03 3  3  v 2 
 R3   1  3 1 3 1 1 3  3 u3 
    
S 3    3 3 3 3 3 3 3  3 v3
Apply known boundary conditions

 R1  ?   5 3 4 0 1  3 u1  0 
 S  0   
 1   3 3 0 0  3  3   v1  ? 
R2  F  AE   4 0 5  3 1 3  u2  ? 
S     
 2  ?  4 L  0 0  3 3 3  3  v2  0 
 R3  ?   1  3 1 u  0
 3 2 0
   3 
S 3  ?    3 3 3 3 0 6 v3 0 
42
 Solution Procedures
 R2  F   4 0 5  3 1 3 u1  0 
 S  0  
 1   3 3 0 0  3  3   v 1  ? 
 R1  ?  AE  5 3 4 0 1  3  u2  ? 
S     
 2  ?  4 L  0 0  3 3 3  3  v 2  0 
 R3  ?  u  0
0
    3 
S 3  ?   13  33  31 3 2 
6  v3 0  
3 0
u2= 4FL/5AE, v1= 0

 R2  F   4 0 5  3 1 3   u1  0 
S 0    v  0 
 1   3 3 0 0  3  3  1 
4FL 
 R1  ?  AE  5
 3 4 0 1  3 u2 
S    5 AE 
 2  ?  4 L  0 0  3 3 3  3  v2  0 
 R3  ?   1 3  32 0  
 1 u  0
    3 
S 3  ?   3 3 3 3 0 6   v3  0 
43
 Recovery of Axial Forces
 P1 1 0 1 0  u1  0   4 
Q     v1  0   5
 0 
 1  AE  0 0 0 0 
Element I  P    4FL   F  4 
 2 L 1 0 1 0  u2    5
  5AE 
Q2 0 0 0 0 
  v2  0   0 

 1 
 P2   1  3 1 3 u2  4FL 
 5 
Q    5AE   3 
Element II  2  AE  3 3 3  3  v  0  5
 P   4L     F  1 
 3  u  0 
2
1  
 3 
3 1
 5 
3   v  0   
3
Q3  3 3  3 3
 3   5 
 P1  1 3 1  3u1  0  0
Q      
 1   AE  3 3  3  3 v1  0  0
Element III
 P    
 
3 u3  0 0
 3  4L  1  3 1

Q3  3  3 3 3 v3 0 0 44
 Stresses inside members

4F
Element I 
4F 4F 5A
P1  P2 
5 5

3
Q3  F
1 5
P3  F
5

Element II 3
Q2  F
5

1
P2  F
5

Element III

45
Finite Element Analysis of
Beams

46
 Bending Beam
y
Review
M M
x

Pure bending problems:

Normal strain: x   y

Ey
Normal stress:  x 


Normal stress with bending moment:

  x ydA  M
1 d2 y
Moment-curvature relationship: 1 M M  EI  EI
  dx2
 EI
My
Flexure formula: x   I   y 2 dA
I
47
 Bending Beam
y
Review q(x)

Relationship between shear force, bending moment and

transverse load:

dV dM
q V
dx dx
4
Deflection: EI d y  q
dx 4
Sign convention:
M + M - M

V + V - V

48
 Governing Equation and Boundary Condition
• Governing Equation
d 2  d 2 v(x) 
 EI
2  2
  q(x)  0, 0<x<L
dx  dx 
• Boundary Conditions -----
dv d 2v d  d 2v 
v?&  ? & EI 2  ? &  EI 2   ? ,
 at x  0
dx dx dx  dx 
dv d 2v d  d 2v 
v? &  ? & EI 2  ? &  EI 2   ? ,
 at x  L
dx dx dx  dx 
Essential BCs– if v or is specdifviedat the boundary.

{ Natural BCs– if or 2
EI d v
dx2
dx
is specifi ed2 atthe boundary.
d d v
 EI 2 
dx  dx  49
 Weak Formulation for Beam Element
• Governing Equation
d 2  d 2 v(x) 
2
 EI 2
  q(x)  0, x1  x  x 2
dx  dx 

• Weighted-Integral Formulation for one element

x2
 d 2  d 2 v(x)  
0   w(x)  2  EI 2
  q(x)dx
x1  dx  dx  

• Weak Form from Integration-by-Parts ----- (1st time)

x
x2
 dw d  d v 
2  d  d v2 2

0     EI 2
  wq  dx w  EI 2 
x 
1
dx dx  dx   dx  dx  x
1
50
 Weak Formulation
• Weak Form from Integration-by-Parts ----- (2nd time)
x x
x2
d w  d v 
2 2 d  d v 2
dw  d v 
2 2 2

0    2  EI 2   wq  dx w  EI   EI 2 
x 
dx  dx  
2
dx  dx  x dx  dx  x
1 1 1

y V(x1) q(x) V(x2)

M(x1) M(x2)
x
x = x1 L= x2-x1 x = x2

x
 d 2 w  d 2v 
x2
  dw 
2

0    2  EI 2   wq  dx  wV  M
x1 
dx  dx    dx  x1

51
 Weak Formulation
• Weak Form
x
x2
 d 2 w  d 2v    dw  2
0    2  EI 2   wq  dx  wV  M
x1 
dx  dx    dx  x1

y(v) Q1 q(x) Q3

Q2 Q4
x
x = x1 L= x2-x1 x = x2

Q1  V x1 , Q2  M x1 , Q3  V x2 , Q4  M x2 

x2
 d 2 w  d 2v   dw dw
    
x  dx2  dx 2  
EI wq dx w(x1
)Q1
w(x 2 )Q3 
dx 1
Q2 
dx 2
Q4
1
52
 Ritz Method for Approximation
y(v) Q1 q(x) Q3

Q2 Q4
x
x = x1 L= x2-x1 x = x2
n
Let v(x)   u j j (x) and n  4
j 1

u1  vx1 ; u 2  ; u3  vx 2 ; u 4 
where dv dv
;
dx x x1 dx x x2
x2  d 2w  4 d 2 j   dw dw
  dx 2  EI 
 uj
dx 2 
  wq  dx  w(x1 )Q1  w(x 2 )Q3 
 dx 1
Q2 
dx 2
Q4
x1   j1
Let w(x)= i (x), i = 1, 2, 3, 4

 d 2i  d 2 j  
id i d
x2 4

  dx2  
 EI u j
dx 2 
  i 
q

dx   i (x1 )Q1   i (x )Q
2 3
  Q2
dx 1 dx 2
Q4
x1   j1 53
 Ritz Method for Approximation

y(v) Q1 Q3

Q2 Q4
x
x = x1 L= x2-x1 x = x2


   di 
 i x1 Q1  Q2  i
 dx x 
  x2
Q3 
 di 

 dx
  4
Q4   Kiju j qi
  j 1
  1  x2  

x2
 di 2 d 2 j  x2

where K ij   EI  2 dx and qi  iqdx

x
 dx dx 2 
x1   1

54
 Ritz Method forApproximation

y(v) Q1 Q3

Q2 Q4
x
x = x1 L= x2-x1 x = x2


 
 1 x1

 d1 

 dx x 
  
1 x2
 d1  
 
 dx x  
 1  2 
 
 

 2 x1
 d2 

 dx x 
  
2
 d2  Q1   K11 K12 K13
  
 dx x  Q   K
K14 u1  q1 
   
 1 
x2
 2
K 22 K 23 K 24  u2  q2 
       
2 21


 
 3 x1
 d3 

 dx x 
  
3 x
 d3   Q3   K 31 K32 K33
 
 dx x Q4  K 41 K42 K43
K34 u3  q3 

K44 u4  q4
 1   2 
2



 
 4 x1
 d4

 dx



 
4 x
 d4 
 
 dx x  
where K K 
  x1   2 
2
ij ji
55
 Selection of Shape Function
The best situation is -----

 
 1 x
 1
 d1 




 
1 x
 d1  




 dx x1   dx x2  
2

 
 

 d2 
     d2  1
 
 dx x  0
0 0 0
 dx x  1 0 0
2
 2 x1  1 
x2
 2
   Interpolation
 d3   0 0 1 0

 
 3 x1
 d3 




 
3 x
  
 dx x  0
 0 0 1

Properties

 dx x1   2 
2



 
 4 x1
 d4





 
4 x
 d4 




  dx x1   dx x2 
2

Q1   K11 K12 K13 K14 u1  q1 

Q   K K 24   q 
 2    12  2    2 
K 22 K 23 u
Q3   K13 K 23 K 33 K 34  u3  q3 

Q4   K14 K24 K34 K44 u4 q4
56
Derivation of Shape Function for Beam Element –
Local Coordinates
How to select i???
v( )  u11  u22  u33  u44
dv( ) d d d d
and  u1 1  u2 2  u3 3  u 4 4
d d d d d
dv1 dv2
where u1  v1 u2  u3  v2 u4 
d d

Let i  ai  bi  ci 2  di 3

Find coefficients to satisfy the interpolation properties.

57
Derivation of Shape Function for BeamElement

How to select i???

e.g. Let 1  a1  b1  c1 2  d1 3

1  1  2 2   
1
4

1
Similarly 2  1  2 1   
4
3  1  2 2   
1
4
2  1  2   1
1
4
58
Derivation of Shape Function for BeamElement

In the global coordinates:

l dv1 l dv 2
v( x)  v11 ( x)  2 ( x)  v23 ( x)  4( x)
2 dx 2 dx

  1x  x   1x  x  
2 3

 1 3   2  
  x2  x1   x2  x1  
 
2 
1    x  x1
  
2
 x  x 1  
 2 l 1
 x2  x1  
    
2

3 
 3   3 1x  x xx
 2 1  
4   x2  x1   x2  x1  
 
  x  x1 
2 x  x
2 1 
l x  x 
1  
  
  x 2  x1  x2  x1 

59
Element Equations of 4th Order 1-D Model
y(v) u1 q(x) u3

u2 u4
x
x = x1 L= x2-x1 x = x2

1 4
1 1
2 3
x=x2
x=x1

Q1  q1   K11 K12 K13 K14 u1 

Q  q   K  
 2    2    12 K 22 K 23 K 24  u 2 
  
Q
 3   3 q  K K K K 34 u3 

13 23 33

Q4  q4   K14 K 24 K 34 K44 u4

x2
 di 2 d 2 j  x2

where K ij   EI  2 dx  K ji and qi  i qdx

x
 dx dx  2
x1   1 60
Element Equations of 4th Order 1-D Model
y(v) u1 q(x) u3

u2 u4
x
x = x1 L= x2-x1 x = x2

Q1   q1  6 3L 6 3L  u1  v1 
Q  q
2EI  3L 2L2  3L L  u2  1 
2
 2

    q   L3 
2

Q3   3   6  3L 6  3Lu3  v2 
  3L 2L2  u   
Q4  q4 
2
3L L
  4 2

x2

where qi   i qdx
x1

61
 Finite Element Analysis of 1-D Problems -
Applications
Example 1.
F
L L L

Governing equation:

d 2  d 2v 
 EI 2   q(x)  0
2 
0  x L
dx  dx 
Weak form for one element

x2
 d 2 w d 2v 
x  dx 2 dx 2
 EI  wq 
dx  wx1Q1
dw
Q2  wx 2 Q3 
dw
Q4  0
1
 dx x1 dx x2

where Q1  V (x1 ) Q2  M (x1 ) Q3  V (x 2 ) Q4  M (x 2 )

62
 Finite Element Analysis of 1-D Problems
Example 1.
l dv1 l dv 2
Approximation function: v( x)  v11 ( x)  2 ( x)  v 23 ( x)  4( x)
2 dx 2 dx
3 
 xx 
2
xx 
 1 3 1   2 1  
  x2  x1   x2  x1  
 
2 
1   2  x  x1
   x  x1 1  
 2 l  x2  x1  
    
2

3 
 x 
 3   3 1   2 1 
x  x  x 
4   x2  x1   x2  x1  

2  x  x  2 x  x 
 l x  x1  x  x   x  x  
1 1

  2 1  2 1

1 4
x=x1 2 3

x=x2
63
 Finite Element Analysis of 1-D Problems
Example 1.

Finite element model:

Q1  6 3L 6 3L v1 
Q   3L 2L2  3L L2  
2EI
   L3   1 
2

Q3   6  3L 6  3Lv2 
Q4  3L 
 L2  3L 2L2 2 

Discretization:

P1, v1 I P2, v2 II P3, v3 III P4, v4

M1 , 1 M2 , 2 M3 , 3 M4 , 4

64
Matrix Assembly of Multiple Beam Elements
Q1I   6 3L 6 3L 0 0 0 0  v1 
Q I  3L 2L2 3L L2 0 0 0 0  
 2   1
Q3I  6 3L 6 3L 0 0 0 0 v2 
 I   
Element I Q4   2EI 3L L
2
3L 2L2 0 0 0 0  2 
  

 
0 L3
0 0 0 0 0 0 0 0  v
 3
    
0  0 0 0 0 0 0 0 0
    3
0  0 0 0 0 0 0 0 0  v 4 
0    
   0 0 0 0 0 0 0 0  4 
0  0 0 0 0 0 0 0 0  v1 
0  0 0 0 0 0 0 0 0 1 
    
Q1 
II
0 0 6 3L 6 3L 0 0 v2 
 II  0 0 3L 2L2 3L L2 0 0    
 2  2EI 
Q
    2 
 II  
Element II Q
 3
L3
0 0 6 3L 6 3L 0 0   v3 
  
Q4II   0  0 3L L2
3L 2L 2
0 0 3
   
 
0  0 0 0 0 0 0 0 0   4
v
0   
0 0 0 0 0 0 0 0 4  65
 Matrix Assembly of Multiple Beam Elements
 0  0 0 0 0 0 0 0 0   v1 
 0  0 0 0 0 0 0 0 0  1 
    
 0  0 0 0 0 0 0 0 0  v2 
Element III
 0 


  2EI 0  0 0 0 0 0 0 0  2 
 III    
Q
 1  L3
 0 0 0 0 6 3L 6 3L v3 
 
Q2III  0 0 0 0 3L 2L2 3L L2  3 
 III    
Q
 3   0 0 0 0 6 3L 6 
3L v4 
 
Q4III  0 0 0 0 3L L
2
3L 2L  4 
2 

 P1  6 3L 6 3L 0 0 0 0 v1 
M 
 3L 2L 2
 3L L2 0 0 0 0 1 
  1

 P2   6  3L 66  3L  3L 6 3L 0 0 v2 
M    
 2   2EI  3L L2  3L  3L 2L2  2 L2  3L L2 0 0 2
P  
  L3 
0 0 6 3L  6  6  3L  3L 6 3L  v 3
 
3
M 3   0 0 3L L2  3L  3L 2L2  2L2  3L L2  3 
   0  
 P4 0 0 0 6  3L 6  3Lv4 
 
M4  0 0 0 0 3L L2  3L 2L2 4 
66
 Solution Procedures
Apply known boundary conditions
 P1 ?  6 3L 6 3L 0 0 0 0 v1  0 
 M ?  3L 2L2  3L L2 0 0 0 0   0 
 1    1 
 P2 ?   6  3L 12 0 6 3L 0 0 v2  0
M  
 2  0   2EI  3L
 L2 0 4L2  3L L2 0 0  2  ? 
P?  L3  0 0 6  3L 12 0 6 3L v 3  0
 3  
 M3  0 
 0
 0 3L L2 0 4L2  3L L2 3  ? 
   
 4
P  F   0 0 0 0  6  3L 6  3Lv4  ? 
M 4  0   
0 0 0 0 3L L2  3L 2L2 4  ? 

 M2  0   3L L2 0 4L2  3L L2 0 0 v1  0 
 M  0 0
 0 3L L 2
0 4L2  3L L2 1 0 
 3
  
P4  F  0 0 0 0 6  3L 6  3Lv2  0
M    
 4  0   2EI  0 0 0 0 3L L2  3L 2L2  2  ? 
P?  6  
6 0  v 3  0 
3
  L 3L 3L 0 0 0
 
1
 M1  ?   3L 2L
2
 3L L2 0 0 0 0 3  ? 
   6  3L 12  
 P  ?  0 6 3L 0 0  v 4  ? 
2
 
P3 ?   0 0  6  3L 12 0 6 3L 4  ? 67
 Solution Procedures

 M2  0   3L L2 0  3L 4L2 L2 0 0 v1  0 
 M  0 0
 0 3L 0 L2 4L2  3L L2 1 0 
 3
  
P4  F  0 0 0 6 0  3L 6  3Lv2  0
M    
 4  0   2EI  0 0 0 3L 0 L2  3L 2L2   v3  0 
P?  6  
6 0   2  ? 
3
  L 3L 0 3L 0 0
 
1
 M1  ?   3L 2L
2
 3L 0 L 2
0 0 0 3  ? 
   6  3L 12   
 P 2 ?   6 0 3L 0 0
 4
v  ? 
P3 ?   0 0  6  12  3L 0 6 3L 4  ?

 M2  0  4L2 L2 0 0 2  ?  P1 ?   3L 0 0 0 2 

 M  0  2EI  L2 4L2  3L L2   M  ? 2EI  L2  
 3 ? 0 0 0
 3    3  0 3L 0 0  
3
  v4  ? 
1 3

P4 F  L  0  3L 6  3L    P2 ?  L  v 4
M4  0  L2  3L 2L2

 0 4  ? P3 ?  3L 0  6 3L 4

68
Shear Resultant & Bending Moment Diagram

3
F F
7

P2

9
F
2 7
FL
7

1
FL
7

69
FL
 Plane Flame
Frame: combination of bar and beam

Q1 , v1
E, A, I, L Q3 , v2
P1, u1 P2, u2
Q2 , 1 Q4 , 2
 AE  AE 
 L 0 0 0 0 
L
 12EI 6EI 12EI 6EI   u1 
 P1   0 0  3 
   L3 L2 L L2   v 
Q1   6EI 4EI 6EI 2EI   1 
 2
Q2   0 L 1 
0
L2 L L
 P    AE AE  u 
 2   0 0 0 0  2 
Q3   L L   v2 
   0 
12EI

6EI
0
12EI
 2 2 
6EI
Q4   L3 L2 L3 L 
 6EI 2EI  6EI 4EI 
0 0
 L2 L L2 L  70
 Finite Element Model of an Arbitrarily Oriented
Frame
y

 x

 x

71
 Finite Element Model of an Arbitrarily Oriented
Frame
local

 12EI  6EI  12EI  6EI 

 L3 0 0
global L2 L3 L2 
 AE AE  u
 P1   0 0 0  0  1 
   L L   v1 
Q1   6EI 4EI 6EI 2EI   
Q2   L2 0
L L2
0
L 1 
 P    12EI 6EI 12EI 6EI  u2 
 2   3 0 0   
Q3   L L2 L3
v2 
2
L
   0 AE
0 0
AE
0  

Q4   L L  2 
  6EI 2EI 6EI 4EI 
 L2 0 0
L L2 L 
72
 Plane Frame Analysis - Example

Rigid Joint Hinge Joint

F F F F

Beam II Bar

Beam I Beam

73
 Plane Frame Analysis
Q3 , v2
Q4,  2
P2, u2

P1, u1
Q2 , 1
Q1 , v1
 12EI 6EI 12EI 6EI  I
 L3 0  2  3 0  2 
L L L
 AE AE  u
 P1  0   1
I
 0 0 0 
   L L   v1 
Q1    6EI 4EI 6EI 2EI   
L  1 
Q2  0 0
 L L L2
2
 P  12EI 6EI 12EI 6EI  u2 
 2   L3 0 0   
Q3  L2 L3 L  v 
2

   
2
AE AE
Q4 
 0  0 0 0  2 
 L L 
  6EI 0
2EI 6EI
0
4EI 
 L2 L L2 L  74
 Plane Frame Analysis

Q1 , v2 Q3 , v3

P1, u2 P2, u3

Q2 , 2 Q4 , 3

 AE  AE 
 L 0 0 0 0 
L
 6EI u2 
 P1 
II
12EI 6EI 12EI
 0 0  
Q   L3 L2 L3 L2  v 
 2  0 6EI 4EI 6EI 2EI   2 
 2
L 2 
Q2  0
 L L
2
L
 P  AE AE u 
 2  0 0 0 0  3 
Q3   L L   v3 
   0 
12EI

6EI
0
12EI
 2 4 
6EI
Q4   L3 L2 L3 L 
 6EI 2EI  6EI 4EI 
0 0
 L2 L L2 L  75
 Plane Frame Analysis
 12EI 6EI 12EI 
I
 AE  AE 
0   6EI 0   L 0 0 0 0
 L3 L2  L 
 L3 L2  u  12EI 6EI  12EI 6EI u 
AE P 
II
 P1    1
I
 0  0
L2  
0 AE 1 0 2
   L 0  0   v1  Q   L3 L2 L3
v
Q1    6EI 4EI L
    2  0 6EI 4EI  6EI 2EI   2 
0
Q2   1  L  2 
6EI 2EI Q2  0
 P  L2 L 0   L2 L L2
 12EI 6EI L2 L  u     A AE  
 2  L3 0 12EI 6EI  2  P2  E 0 0 0 0  u 3 
Q 3 0
 L2  v2  Q 3  L v 
   0 
AE L3 L2
      L
 12EI  6EI 12EI 6EI  3 
Q4  0 AE 0  2 4 
 L 0 0   2 Q4   0 L3 L2 L3 L 
  6EI 0
2EI L   6EI 2EI 6EI 4EI 
 L2 6EI 4EI  0  2
 L 
L 0 L2
L2 L 0 L L

76
Plane Frame Analysis

1
F
2

1
F
2

1
FL
8

3
FL
8

77
 UNIT- 3
Finite element analysis constant
strain triangle

78
Finite element formulation for 2D:

Step 1: Divide the body into finite elements connected to each

other through special points (“nodes”)
py
3 v3
px
4 3
u3
v4 u 1 
v 2 Element ‘e’ v2 v 
1  1
4
u ST u4 u 2 
v1 2 u2  
y v 2 
x y d 
Su  u3 
1 u1 v 3 
v  
x x u 4 
u v 
 4
79
 u (x, y)  N1(x, y)u1  N2(x, y) u2 N3(x, y) u3  N4(x, y) u4
v (x, y)  N1(x, y)v1  N2(x, y)v2 N3(x, y) v3  N4(x, y)v4

u1 
v 
 1
u 2 
 
u (x, y)  N 1 0 N2 0 N3 0 N4 0  v 2 
u  
v (x, y)  0 N1 0 N2 0 N3 0 N 4 u 3 
v3 
 
u 4 
v 
uNd  4

80
TASK 2: APPROXIMATETHESTRAIN and STRESSWITHIN EACH
ELEMENT

Approximation of the strain in element ‘e’

u (x, y) N1(x, y) N2 (x,y) N (x, y)  N4 (x, y) u

x   u1  u2  3 u3
x x x x x
4

v (x, y) N1(x, y) N2 (x, y) N3(x, y) N4 (x,y)

y   v  v  v  v4
y y y y y
1 2 3

u (x, y) v (x, y) N1(x, y) N1(x, y)

 xy    u  v1  ......
y x y x
1

81
   x
 
   y 
xy 

u1 
v 
N (x, y) N 2(x, y) N 3(x, y) N 4(x, y)  1 
1 0 0 0 0 u 2
 x x x x  
N1(x, y) N2 (x, y) N3(x, y) N4 (x,y)v2 
 0 0 0 0  
 y y y y  u 3 
N (x, y) N1(x, y) N2 (x, y) N2 (x, y) N3(x, y) N3(x, y) N4 (x, y) N 4(x , y)v 
1  3
 y x y x y x y x u 
4
 B v 
4 

εBd

82
Summary: For each element
Displacement approximation in terms of shape functions
uNd

Strain approximation in terms of strain-displacement matrix

εBd
Stress approximation
  DB d
Element stiffness matrix

k   e B D B dV
T
V

Element nodal load vector

f   e N T X dV   e
T
N T S dS
V ST
f f
b S 83
Constant Strain Triangle (CST) : Simplest 2D finite element

v1
v3
1 u1
(x1,y 1) (x3,y3)
v2 v u3
y u 3
(x,y)
u2
2 (x ,y )
2 2

• 3 nodes per element

• 2 dofs per node (each node can move in x- and y- directions)
• Hence 6 dofs per element

84
The displacement approximation in terms of shape functions is

u (x,y)  N1u1  N2u2  N3u3

v(x,y)  N1v1  N 2 v 2  N 3 v 3
u1 
v 
 1
u (x, y)  N 1 0 N2 0 N3 0 u 2 
u 

v (x, y)   0 N1 0 N2 0 N 3  v 2 
u 3 
 
u 21  N26 d61 v3

N1 0 N2 0 N3 0
N
0 N1 0 N2 0 N 3 

85
 Formula for the shape functions are
a1  b1x  c1 y
v1 N1 
v3 2A
1 u1 a2  b2 x  c2 y
(x1,y 1) (x3,y 3) N2 
v2 v u3 2A
u a3  b3 x  c3 y
y (x,y)
3
N3 
2A
where u2
2 (x ,y
2 )2

x
1 x 1 y1 
1 
A  area of triangle  det 1 x 2 y 
2
2
1 x 3 y 3 
a1  x 2 y3  x3 y 2 b1  y 2  y 3 c1  x3  x 2
a 2  x3 y1  x1 y 3 b2  y3  y1 c 2  x1  x3
a 3  x1 y 2  x 2 y1 b3  y1  y 2 c3  x 2  x1
86
 Properties of the shape functions:

1. The shape functions N1, N2 and N3 are linear functions of x and

y
N2
N1 N3
1
1 1
1 3 1
y 3 2 3

1
2 2

1 at node 'i'
Ni  
0 at other nodes
87
2. At every point in the domain

N
i1
i 1
3

N x i i x
i1
3
N y i i y
i1

88
3. Geometric interpretation of the shape functions
At any point P(x,y) that the shape functions are evaluated,

A1
N1 
A
A2
N2 
P (x,y)
A
1 A3
A2 N3 
A3 A
y A1 3

89
Approximation of the strains
 u 
 x 
 x   
v
   y 

 
  Bd
y
   
 xy   u v 

 y x 


N1(x,y) N2(x,y) N3(x,y) 

 0 0 0 
 x x x 
N1(x,y) N2(x,y) N3(x, y) 
B 0 0 0
 y y y 
N (x, y) N(x,y) N(x,y) 2N(x,y) N3(x, y) N3(x, y)
1 1 2 
 y x y x y x 
b1 0 b2 0 b3 0 
  0 c1 0 c2 0 c3
1
2A
c1 b1 c2 b2 c3 b3

90
Inside each element, all components of strain are constant: hence
the name Constant Strain Triangle

Element stresses (constant inside each element)

  DB d

91
IMPORTANT NOTE:
1. The displacement field is continuous across element boundaries
2. The strains and stresses are NOT continuous across element
boundaries

92
Element stiffness matrix
t
k   e B D B dV
T
V

Since B is constant
A

k  B D B e dV  BT D B At t=thickness of the element

T
V A=surface area of the element

93
Element nodal load vector

f   e N T X dV   e N T T S dS
V ST
f f
b S

94
Element nodal load vector due to bodyforces

f b   e N X dV  t  e N X dA
T T
V A

t N X dA 
 f b1x   Ae
1 a

fb1y
fb3y   t  e N1 X b dA 
1 fb1x  fb1y   A 
t N X dA
 b 2 x   Ae 2 a
 f 

fb2y
Xb fb3x fb   
 b2 y  t Ae N 2 X b
Xa 3 f dA
y (x,y)
 f b3x   
fb2x   t Ae N 3 X a dA
2  f b3 y   
x
t Ae N 3 X b dA 

95
EXAMPLE:

If Xa=1 and Xb=0

t N X dA 
 fb1x   Ae
1 a
  t  e N 1dA 
tA 
    3

 f 


t A e
N 1 X b dA
 
A
0 
0
tA 
b1y
 
 f b 2 x    e N 2 X a dA t N dA  
 
   Ae 2
t
f b   A 3
 f b2 y  t Ae N 2 X b dA  0  0
 f b3x      tA 
  t  e N 3 X a dA  Ae 3   
t N dA
f b3 y   A   0   3 
t Ae N 3 X b dA 0

96
Element nodal load vector due to traction

  e N T S dS
T
f S ST

EXAMPLE:

fS1y
fS3y  t
T
1 f e
N T S dS
S l13 along 13

fS1x fS3x
y 3

97
Element nodal load vector due to traction

EXAMPLE:

fS2y f S  t
T
e
N T S dS
l23 along 23
(2,2)

y
2 fS2x
fS  t
2x l N2 (1) dy 1
1 e along 23
TS  
23

0  1 2
fS3y  t   2 1  t
1  2
3 f x
(0,0) (2,0) S3x Similarly, compute

fS  0
2y

fS  t
3x

fS  0
3y

98
Recommendations for use of CST

1.Use in areas where strain gradients aresmall

2.Use in mesh transition areas (fine mesh to coarse mesh)

3.Avoid CST in critical areas of structures (e.g.,stress

concentrations, edges of holes, corners)

4.In general CSTs are not recommended for general analysis

purposes as a very large number of these elements are required
for reasonable accuracy.

99
 Example 1000 lb
y 300 psi
3 2

El 2 Thickness (t) = 0.5 in

2 in E=30×106 psi
El 1
=0.25
1
x
4 3 in

(a) Compute the unknown nodal displacements.

(b) Compute the stresses in the two elements.

100
Realize that this is a plane stress problem and therefore we need to use

1 
  0  3.2 0.8 0 
D  E  1 0   0.8 3.2 0  10 7 psi
1  2  1  
 0 0  0 0 1.2
2

Step 1: Node-element connectivity chart

ELEMENT Node 1 Node 2 Node 3 Area

(sqin)
1 1 2 4 3
2 3 4 2 3

Node x y
1 3 0 Nodal coordinates
2 3 2
3 0 2
4 0 0 101
 Step 2: Compute strain-displacement matrices for the elements

Recall b1 0 b2 0 b3 0 with

b1  y2  y3 b2  y3  y1 b3  y1  y2
B  0 c1 0 c2 0 c3
1
2A  c1  x3  x2 c2  x1  x3 c3  x2  x1
c1 b1 c2 b2 c3 b3

For Element #1: 2(2) y1  0; y2  2; y3  0

x1  3; x2  3; x3  0
Hence
b1  2 b2  0 b3  2
c1  3 c2  3 c3  0
4(3) 1(1) Therefore  2 0 0 0 2 0 
B(1)   0 3 0 3 0 0 
(local numbers within brackets) 1
6
3 2 3 0 0 2
For Element #2:

 2 0 0 0 2 0
B(2)   0 3 0 3 0 0
1
6
3  2 3 0 0 2 102
Step 3: Compute element stiffness matrices

k  AtB D B  (3)(0.5)B
(1) (1)T (1) (1)T
D B(1)
0.9833  0.5  0.45 0.2  0.5333 0.3 

 1.4 0.3 1.2 0.2  0.2

 0.45 0 0  0.3 7
  10
 1.2  0.2 0 
 0.5333 0 
 
 0.2 
u1 v1 u2 v2 u4 v4

103
k  AtB D B  (3)(0.5)B D B(2)
(2) (2)T (2) (2)T

0.9833  0.5  0.45 0.2  0.5333 0.3 

 
 1.4 0.3 1.2 0.2  0.2

 0.45 0 0  0.3 7
  10
 1.2  0.2 0 
 0.5333 0 
 
 0.2 
u3 v3 u4 v4 u2 v2

104
Step 4: Assemble the global stiffness matrix corresponding to the nonzero degrees of
freedom
Notice that
u3  v3  u 4  v4  v1  0
Hence we need to calculate only a small (3x3) stiffness matrix

 0.983  0.45 0.2 u

  0.45 0  10 7 1
K 0.983 u2
 0.2 0 1.4 v2
u u2 v2
1

105
Step 5: Compute consistent nodal loads

 f1x   0 
  
f  f 2 x    0 
f  f 
 2 y   2 y

f 2 y  1000  f S
2y

The consistent nodal load due to traction on the edge 3-2

f S2 y  
3 x
x0
N 3 32 (300)tdx N 2 32 
3
 (300)(0.5)  N 3 32 dx
3

 150 3
x x0
dx 3 2
x0 3
 x2 
3
 9
 50    50    225 lb
 2 0 2
106
Hence
f 2 y  1000 f S
2y

 1225 lb

Step 6: Solve the system equations to obtain the unknown nodal loads

Kdf

 0.983  0.45 0.2u1   0 

   
107   0.45 0.983 0  u 2    0 

 0.2 0   v2  1225 
1.4 

Solve to get

 u1   0.2337 10 4 in 
   4 
 2 
u  0.1069 10 in 
v   0.9084 10 4 in
 2  
107
Step 7: Compute the stresses in theelements

In Element #1

 (1)  D B(1)d(1)

With

d (1)  u1 v1 u2 v2 u4 v4 
T


 0.2337 104 0 0.1069 104  0.9084 104 
0 0

Calculate

 114.1 
 (1)  1391.1 psi
  76.1 

108
 In Element #2

 (2)  DB(2) d(2)

With

d (2)  u3 v3 u4 v4 u2 v2 
T


 0 0 0 0 0.1069 104  0.9084 104 
Calculate

 114.1 
 (2)   28.52  psi
363.35

Notice that the stresses are constant in each element

109
 UNIT - 4
Heat transfer analysis

110
 Thermal Convection

Newton’s Law of Cooling

q  h(Ts  T )

h: convective heat transfer coefficient (W m2  C o )

111
Thermal Conduction in 1-D
Boundary conditions:

Dirichlet BC:

Natural BC:

Mixed BC:

112
 Weak Formulation of 1-D HeatConduction
(Steady State Analysis)
• Governing Equation of 1-D Heat Conduction -----
d  dT ( x) 
   ( x) A( x)   AQ( x)  0 0<x<L
dx  dx 

• Weighted Integral Formulation -----

 d  dT (x)  
L
0   w( x)   ( x) A( x)   AQ( x) dx
0  dx  dx  
• Weak Form from Integration-by-Parts -----
L
 dw  dT   dT 
0     A   wAQ dx  w  A 
L

0
dx  dx    dx  0
113
 Formulation for 1-D Linear Element
T1 T2

x f1 f2
1 2
x1 x2

T T
f1 (x)   A , f 2 (x)  A
x 1 x 2

Let T (x)  T11(x)  T22 (x)

x2  x x  x1
1( x)  , 2( x) 
l l 2T2
1T1

x1 x2 114
 Formulation for 1-D Linear Element

Let w(x)= i (x), i = 1, 2

2  x2  d d    x2

0   T j    A  dxi dxj  dx   i AQ dx  i ( x 2 ) f 2  i ( x1 ) f1 

j1  x1    x1
2
  KijTj  Qi  i (x2 ) f 2  i (x1 ) f1 
j 1

 f1  Q1   K11 K12 T1 

f      
 2  Q2   K12 K22 T2 

 d i d  j 
x2 x2
dT dT
where K ij    A  dx dx  dx, Qi   i AQ dx, f1   A dx , f2  A
x1   x1 x1 dx x2
115
Element Equations of 1-D Linear Element

T1 T2

x f1 f2
1 2
x1 x2

 f1  Q1  A  1 1T1 
 f      
 2  Q2  L 1 1 T2 
x2
dT dT
where Qi   i AQ dx, f1   A , f2   A
x1
dx x x1
dx xx2
116
 1-D Heat Conduction - Example
A composite wall consists of three materials, as shown in the figure below.
The inside wall temperature is 200oC and the outside air temperature is 50oC
with a convection coefficient of h = 10 W(m2.K). Find the temperature along
the composite wall.
1  70W m  K , 2  40W m  K , 3  20W m  K 
t1  2cm, t2  2.5cm, t3  4cm

1 2 3
T0 200 C
o T  50o C

t1 t2 t3
x

117
Thermal Conduction and Convection-
Fin
Objective: to enhance heat transfer

Governing equation for 1-D heat transfer in thin fin

d dT 
 A c AQ
c 0
dx  dx 
w t
2h(T  T )  dx  w  2h(T  T )  dx  t 2h(T  T )  w  t 
Qloss  
Ac  dx Ac
x
dx
d dT 
 A c  Ph T  T   A Q
c  0
dx  dx 

where P  2 w  t 

118
 Fin - Weak Formulation
(Steady State Analysis)
• Governing Equation of 1-D Heat Conduction -----
d  dT ( x) 
   ( x) A( x)   Ph T  T   AQ  0 0<x<L
dx  dx 

• Weighted Integral Formulation -----

 d  dT (x)  
L
0   w( x)   ( x) A( x)   Ph(T  T )  AQ( x) dx
0  dx  dx  
• Weak Form from Integration-by-Parts -----
L
 dw 
L
dT    dT 
0    A   wPh(T  T  )  wAQ  dx  w   A 
0
dx  dx    dx 0
119
 Formulation for 1-D Linear Element

Let w(x)= i (x), i = 1, 2

2 x 2
di d j  
x 2

0  Tj     A  Phi j  dx   i  AQ  PhT dx

j1  x 
1
dx dx   x 1

 i ( x2 ) f 2  i ( x1 ) f1 
2
  KijTj  Qi  i (x2 ) f 2  i (x1 ) f1 
j 1

 f1  Q1   K11 K12 T1 

f      
 2  Q2   K12 K22 T2 
 d i d  j 
x2 x2

where K ij     A  Phi j  dx, Qi   i  AQ  PhT dx,

x1 
dx dx x1

dT dT
f1    A , f2   A 120
dx x x1 dx x x2
 Element Equations of 1-D Linear Element

T1 T2

x f1 f2
1 2
x=0 x=L

 f1   Q1    A  1 1 Phl 2 1   T1 

f           
 2  Q2   L  1 1  6  1 2   T2 

x2
dT dT
where Qi   i AQ  PhT dx, f1    A , f2   A
x1
dx x x1
dx x x2

121
Time-Dependent Problems

122
 Time-Dependent Problems
In general, ux, t 
Key question: How to choose approximate functions?

Two approaches:

ux, t   u j j x, t 

ux, t   u j tj x

123
Model Problem I – Transient Heat Conduction

u   u 
c   a   f x, t 
t x  x 
Weak form:

 w u u
x2

0  a  cw  wf dx  Q1w(x 1 )  Q2 w(x2 )
x1 
x x t 

Q1   a  ; Q2  a 
du du
 dx x1  dx x2

124
 Transient Heat Conduction
n
let: ux, t   u j tj x and w  i x
j 1

 w u u
x2

0  a  cw  wf dx  Q1w(x1 )  Q2w(x 2 )
x1 
x x t 

K uM u  ODE!

F x2
i
x2

K ij   a j
dx M ij   ci j dx
x1
x x x2 x1

Fi   i fdx  Qi
x1 125
 Time Approximation – First Order ODE

 bu  f t u0 u 0
du
a 0t T
dt
Forward difference approximation - explicit

 t
uk 1  uk f k  buk 
a
Backward difference approximation - implicit

t
uk 1  uk  f  buk 
a  bt
k
126
 Stability Requirment

2
t  t cri 
1 2max

where
K M uQ

Note: One must use the same discretization for solving

the eigenvalue problem.

127
Transient Heat Conduction - Example
u  2 u
 0 0  x 1
t x 2
u
u0, t  0 1,t  0 t0
t
ux,0 1.0

128
Transient Heat Conduction - Example

129
Transient Heat Conduction - Example

130
Transient Heat Conduction - Example

131
Transient Heat Conduction - Example

132
Transient Heat Conduction - Example

133
 UNIT – 5
Dynamic Analysis

134
 Axi-symmetric Analysis
Cylindrical coordinates: r,  , z 
x  r cos ; y  r sin  ; z  z

• quantities depend on r and z only

• 3-D problem 2-D problem

135
Axi-symmetric Analysis

136
Axi-symmetric Analysis – Single-Variable Problem

1  u(r, z)    u(r, z) 
  11
ra    22
a   a00 u  f (r, z)  0
r r  r  z  z 

Weak form:

w  u  w  u  
0     a11    22  00
a  a wu  wf (r, z) rdrdz
 
r  r  z 
e
z  
 wq ds n
e

u(r, z) u(r, z)
where qn  a11 nr  a22 nz
r z

137
Finite Element Model – Single-Variable Problem

u   u j j where  j (r, z)   j (x, y)

j

Ritz method: w  i
n
Weak form
 ij j i i
K e e
u  f e
 Q e

j 1

 i  j i  j 
where
K    11
e a  a22  a00
i j  rdrdz
r r z z
ij
 e 

fi e    frdrdz
e
i

Qie   i qnds
e
138
 Single-Variable Problem – Heat Transfer
Heat Transfer:

1  T (r, z)  T (r, z) 
 r r  rk r    k   f (r, z)  0
  z  z 

Weak form

 w  T  w  T  
 
0 
e  
k
 
k
 r  r  z  z  wf (r, z)  rdrdz
 
 wq ds n
e

T (r, z) T (r, z)
where qn  k nr  k nz
r z

139
 3-Node Axi-symmetric Element

T (r, z)  T11  T22  T33

3

r2 z3  r3z2 
1 
1 r z 
1  z2  z 3 
2 Ae  r r 
 3 2 

 r3 z1 r1z3 
2 2 
1 r z   z  z 
 3 1 
2 Ae  rr 
 1 3 

 r1 z2 r 2z 1 
3
1 rz 
 z1 z 2 
2 Ae  r r 
 2 1 

140
 4-Node Axi-symmetric Element


T (r, z)  T11  T22  T33  T44

4 3
b

1  1   1   2   1  

1 2

a  a  b a b
z
3    4  1  
r ab  a b

141
Single-Variable Problem – Example

Step 1: Discretization

Step 2: Element equation

 i  j   
K   
e
  i j rdrdz
r r z z 
ij
e

fi e   i frdrdz Qie   i qnds

e e

142
 Review of CSTElement
• Constant Strain Triangle (CST) - easiest and simplest
finite element
– Displacement field in terms of generalized coordinates

– Resulting strain field is

– Strains do not vary within the element. Hence, the name

constant strain triangle (CST)
• Other elements are not so lucky.
• Can also be called linear triangle because displacement field is
linear in x and y - sides remain straight.

143
 Constant Strain Triangle
• The strain field from the shape functions looks like:

– Where, xi and yi are nodal coordinates (i=1, 2, 3)

– xij = xi - xj and yij=yi - yj
– 2A is twice the area of the triangle, 2A =x21y31-x31y21
• Node numbering is arbitrary except that the sequence 123
must go clockwise around the element if A is to be positive.

144
 Constant Strain Triangle
• Stiffness matrix for element k =BTEBtA
• The CSTgives good results in regions of the FE
model where there is little strain gradient
– Otherwise it does not work well.

145
 Linear Strain Triangle
• Changes the shape functions and results in
quadratic displacement distributions and
linear strain distributions within the element.

146
 Linear Strain Triangle

• Will this element work better for the problem?

147
 Example Problem
• Consider the problem we were1k looking at:
1 in.

1k
5 in.
0.1 in.

I  0.113 / 12  0.008333 in4

  M  c  1 0.5  60 ksi
I 0.008333
    0.00207
E
2
ML 25
   0.0517 in.
2EI 2  29000  0.008333

148
 Bilinear Quadratic
• The Q4 element is a quadrilateral element
that has four nodes. In terms of generalized
coordinates, its displacement field is:

149
 Bilinear Quadratic
• Shape functions and strain-displacement
matrix

150
 Example Problem
• Consider the problem we were0.1klooking at:
1 in.
0.1k

5 in.
0.1 in.

I  0.1 13 / 12  0.008333in4

  M  c  1 0.5  60ksi
I 0.008333
    0.00207
E
PL3 0.2  125
   0.0345in.
3EI 3  29000  0.008333

151
 Quadratic Quadrilateral Element
• The 8 noded quadratic quadrilateral element
uses quadratic functions for the displacements

152
 Quadratic Quadrilateral Element
• Shape function examples:

• Strain distribution within the element

153
 Quadratic Quadrilateral Element
• Should we try to use this element to solve our
problem?

• Or try fixing the Q4 element for our purposes.

– Hmm… tough choice.

154
 Isoparametric Elements and Solution
• Biggest breakthrough in the implementation of the
finite element method is the development of an
isoparametric element with capabilities to model
structure (problem) geometries of any shape and size.
• The whole idea works on mapping.
– The element in the real structure is mapped to an
‘imaginary’ element in an ideal coordinate system
– The solution to the stress analysis problem is easyand
known for the ‘imaginary’ element
– These solutions are mapped back to the element in the
real structure.
– All the loads and boundary conditions are also mapped
from the real to the ‘imaginary’ element in this approach

155
 Isoparametric Element
3
4 
(x3, y3) 4
(x4, y4) 3 (1, 1)
(-1, 1)

1 2
Y,v 2 (-1, -1) (1, -1)
1
(x1, y1) (x2, y2)

X, u

156
 Isoparametric element
• The mapping functions areqx u
 ite simple:

1

x
 2
 x3 
X  N1 N 2 N3 N4 0 0 0 0  x4 
   
Y   0 0 0 0 N1 N2 N3 N 4  y1 
 y2 
 
 y3 
y4 

N1 1 (1  )(1 ) Basically, the x and y coordinates of any point in the
4 element are interpolations of the nodal (corner)
1
N 2  (1  )(1 ) coordinates.
4
From the Q4 element, the bilinear shape functions
N 3  1 (1  )(1 )
4 are borrowed to be used as the interpolation
1 functions. They readily satisfy the boundary values
N 4  (1  )(1 )
4 too. 157
 Isoparametric element
• Nodal shape functions for duisplacements 1

u
 2
u3 
u N1 N 2 N3 N4 0 0 0 0 u 4 
  
v   0 0 0 0 N1 N2 N3 N4  v1 
v2 
 
v3 
v4

N1 1 (1  )(1 )

4
1
N 2  (1  )(1 )
4
N 3  1 (1  )(1 )
4
1
N 4  (1  )(1 ) 158
4
• The displacement strain relationships:
 u   u     u  
x 
 X   X   X
 v   v     v  
y 
 Y   Y   Y
 u 
 u    
0
 
0    

 x   X    X X  
u 
     0    
y  
v   0  
Y Y     v 
   Y         
 xy   u  v     
 Y X    Y Y X X   v 

 

B u t , i t is too difficult to obtain

 a n d 
X X 159
 Isoparametric Element
Hence we will do it another way
u  u  X  u  Y
 X  Y 
u  u  X  u  Y
 X  Y 
X   Ni X Y Ni
 u  X Y   u   Y
  i   i
     X 
 u   
X Y   u  X N
 iX i
Y N
  i Yi
         
   Y 

It is easier to obtain
X and Y
 u   u 
 
 X  1   
X Y      J  
    u   u 
J  X Y   Jacobian Y  
 
 
defines coordinate transformation
169
 Gauss Quadrature
• The mapping approach requires us to be able to
evaluate the integrations within the domain (-1…1) of
the functions shown.
• Integration can be done analytically by using closed-
form formulas from a table of integrals (Nah..)
– Or numerical integration can be performed
• Gauss quadrature is the more common form of
numerical integration - better suited for numerical
analysis and finite element method.
• It evaluated the integral of a function as a sum of a
finite numb1er of terms n
I   d becomes I  Wii
1 i 1
161
 Gauss Quadrature
• Wi is the ‘weight’ and i is the value of f(=i)

162
 Numerical Integration
b
Calculate: I   f xdx
a

• Newton – Cotes integration

• Trapezoidal rule – 1st order Newton-Cotes integration

f (b)  f (a)
f (x)  f 1(x)  f (a)  (x  a)
ba
b b
f (a)  f (b)
I   f (x)dx   f1 (x)dx  (b  a)
a a
2

• Trapezoidal rule – multiple application

b xn x1 x2 xn

I   f (x)dx   fn (x)dx   f (x)dx   f (x)dx   f

(x)dxa x0 x0 x1 x n1

h n1

I   f (a)  2  f (xi )  f (b) 
2 i1  163
 Numerical Integration
b
Calculate: I   f xdx
a

• Newton – Cotes integration

• Simpson 1/3 rule – 2nd order Newton-Cotesintegration
(x  x1 )(x  x2 ) (x  x0 )(x  x2 ) (x  x0 )(x  x1)
f (x)  f 2(x)  f (x0 )  f (x1 )  f (x2 )
(x0  x1 )(x0  x2 ) (x1  x 0)(x 1 x )2 (x2  x0 )(x2  x1 )

b b
f (x0)  4 f (x1 )  f (x2 )
I   f (x)dx   f 2 (x)dx  (x2  x0 )
a a
6

164
 Numerical Integration
b
Calculate: I   f xdx
a

• Gaussian Quadrature
Trapezoidal Rule: Gaussian Quadrature:

f (a)  f (b)
I  (b  a)
2
I  c0 f (x0 )  c1 f (x1)
(b  a) (b  a)
 f (a)  f (b)
2 2

Choose c0 , c1, x0 , x1 according to certain criteria

165
 Numerical Integration
b
Calculate: I   f xdx
a
1
• Gaussian Quadrature I  f x dx  c f x  c f x 
0 0 1 1  cn1 f xn1 
1
• 2pt Gaussian Quadrature
 1   1 
1
I   f xdx  f    f
1  3  3

• 3pt Gaussian Quadrature

1
I   f xdx  0.55 f  0.77 0.89  f 0 0.55 f 0.77
1

Let: ~x  1 2(x  a)

ba

1 
b 1
1 1
 f (x)dx  (b  a)  f  (a  b)  (b  a)x~dx~
2 2 2 
a 1
166