0% found this document useful (0 votes)
31 views

Two Dimensional Elements-Thermal Problems

This document discusses two-dimensional thermal problems and the finite element method used to solve them. It begins by describing types of 2D problems including problems with vector and scalar variables. It then presents the governing equation for 2D heat transfer by conduction and convection. Next, it derives the weak form of the governing equation and the element stiffness matrix formulation. The document provides an example problem and walks through constructing the element stiffness matrices and solving the system of equations to obtain temperatures at nodes. It concludes by defining strain-displacement relationships for a constant strain triangle element.

Uploaded by

abhi
Copyright
© © All Rights Reserved
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
31 views

Two Dimensional Elements-Thermal Problems

This document discusses two-dimensional thermal problems and the finite element method used to solve them. It begins by describing types of 2D problems including problems with vector and scalar variables. It then presents the governing equation for 2D heat transfer by conduction and convection. Next, it derives the weak form of the governing equation and the element stiffness matrix formulation. The document provides an example problem and walks through constructing the element stiffness matrices and solving the system of equations to obtain temperatures at nodes. It concludes by defining strain-displacement relationships for a constant strain triangle element.

Uploaded by

abhi
Copyright
© © All Rights Reserved
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 70

TWO DIMENSIONAL ELEMENTS- THERMAL PROBLEMS

LECTURE 9

1
Types of 2D Problems
VECTOR VARIABLE PROBLEMS
e.g. Structural problems
SCALAR VARIABLE PROBLEMS
e.g. Torsion of non-circular shafts,
Heat transfer through fins

2
3
Governing Equation for 2D Heat transfer
by conduction and convection

 T  T 
2 2
k  2  2   h(T  T )  0
 x y 

Weak form of the equation

T w T w
 x x dxdy   y y dxdy   hTw( x, y)dxdy
  hT w( x, y )dxdy
4
 Ni N j Ni N j 
Kij condn.  k  dxdy   dxdy
 x x y y 

1
N i (x,y) = (αi + βi x + γi y)
2 Ae

 12   12 1 2   1 2 1  3   1 3 
k  
K  condn.
e
  1 2   1 2 2   2
2 2
 2  3   2 3 
4A
 1  3   1 3  2  3   2 3  3   3 
2 2

5
hpl 2 1
k conv   
6 1 2
 p 1

 q1  0
  hlT  
q 2   1
q  2  
 3 1
6
PROBLEM 1:

7
8
Element i j k
1 1 2 3
2 5 4 3
3 1 3 5
9
10
 12   12 1  2   1 2 1  3   1 3 
k  
[k ] 
e
  2 1   1 2 2   2
2 2
 2  3   2 3 
4A
 1  3   1 3  2  3   2 3  3   3 
2 2

Element 1 and 2 Element 3


β1 = -0.15 , 1 = 0, β1 = 0.15 , 1 = -0.4
β2 = 0.15 , 2 = -0.4 β2 = 0.3 , 2 = 0
β3 = 0 , 3 = 0.4 β3 = -0.15 , 3 = 0.4

11
 0.0225  0.0225 0 
2
[k ]cond  [k ]1cond 
1.5  0.0225 0.1825  0.16
4  
 0.4  0.15  0  0.16 0.16 
2

 0.028125  0.028125 0 
 10  0.028125 0.228125  0.2
 
 0  0.2 0.2 

 0.28125  0.28125 0 
 
  0.28125 2.28125  2
 0 2 2 
12
 0.1825  0.045  0.1825
[k ]3cond 
1.5   0.045 0 .09  0.045 
1  
4  2   0.4  0.15  0.1825  0.045 0.1825 
2

1.14  0.28125  0.86 



 0.5625 
 0.28125
 1.14 

13
hpl 2 1
k conv   
6 1 2
 p 1

Element i j k
1 1 2 3
2 5 4 3
3 1 3 5
14
0 0 0
hl  
2
[k ]  [k ]
1
 0 2 1
6 
conv conv

0 1 2

0 0 0 

 0 2.5 1.25 
0 1.25 2.5 

0  0 
hlT    
Q 1  93.75
2    
  
1 93.75 15
 [k ]condn  [k ]conv
e
[k ] Thermal

 0.28125  0.28125 0 
1 2 
[k ]th  [k ]th   0.28125 4.78 
 0.75
 0  0.75 4.5 

 1.14  0.28125  0.86 


3 
[k ]th   0.28125 0.5625 
 0.28125
  0.86  0.28125 1.14 
16
T1   0 
T   
 2  93.75

 
[k ]th T3   [Q] G [Q]G  93.75  93.75
 
T  
93.75

 4  0 
T5 
 1.42125  0.28125  0.28125 0  0.86 
 0.28125 4.78  0.75 0 0 
 
[k ]G   0.28125  0.75 9.5625  0.75  0.28125
 
 0 0  0.75 4.78  0.28125
  0.86 0  0.28125  0.28125 1.42125 
17
 1.42125  0.28125  0.28125 0  0.86  T1  0
 0.28125 4.78  0.75 0 0  T  1
   2   
 0.28125  0.75 9.5625  0.75  0.28125 T3   93.752
   1
 0 0  0.75 4.78  0.28125 T4   
  0.86 0  0.28125  0.28125 1.42125  T5  0

Substitute forT4 & T5as 180º and


evaluate T1, T2 ,T3

18
 1.42125  0.28125  0.28125 0  0.86  T1 
 0.28125 4 .78  0.75 0 0  T 
   2 
 0.28125  0.75 9.5625  0.75  0.28125 T3  
  
 0 0  0.75 4.78  0. 28125 T4 
  0.86 0  0.28125  0.28125 1.42125  T5 
 0  0.86 *180 
 1 
  T1 = 124.5ºC
93.752  (0.28125  0.75) *180
 

1
 T2 = 34.0ºC
 0 
T3 = 45.4ºC

19
20
21
u v u v
 xx  ,  yy  , and  xy  
x y y x
1
 xx  ( 1u1   2 u 2   3 u 3 )
2A
1
 yy  ( 1u1   2 u 2   3 u 3 )
2A
22
Variation of Shape functions for CST element

23
STIFFNESS MATRIX FOR BI LINEAR
RECTANGULAR ELEMENT

 x  y   x  y 
N1  1  1   N 3    
 2a  2b   2a  2b 

 x  y   x  y 
N 2   1   N 4  1   
 2a  2b   2a  2b 

24
2 a 2bdN1 dN1
k11    . dxdy
0 0 dx dx

25
2
1 
2 a 2by 
k11    2 1   dxdy
0 0 4a
 2b 
b

3a

2
2 a 2b 1 1  y 
k12      1   dxdy
0 0 2a 2a  2b 
b

3a
26
2 a 2b 1 1  y 
k13      1   dxdy
0 0 2a 2a  2b 
b

6a

2 a 2b 1  y  y 
k14    2  1   dxdy
0 0 4a
 2b  2b 
b

6a

27
1 
2 a 2b y 
k22    2 1   dxdy
0 0 4a
 2b 
b

3a

2 a 2b 1  y 
k23    2 1   dxdy
0 0 4a
 2b 
b

6a
28
2 a 2b 1  y 
k24     2 1   dxdy
0 0 4a  2b 
b

6a

2 a 2b 1  y 
2

k33    2  2  dxdy
0 0 4a
 4b 
b

3a
29
2 a 2b 1  y2 
k34     2  2  dxdy
0 0 4a  4b 
b

3a

2a dN 4 dN 4
2b
K 44    dxdy
0 0 dx dx
2
2 a 2b 1 y b
  2 2
dxdy 
0 0 4a 4b 3a
30
31
 2 2 1 1   2 1 1 2 
 2 2 1 1   1 2 2 1 
b  k a 
k  
6a  1 1 2 2  6b  1 2 2 1
   
 1 1 2 2   2 1 1 2 

 2( a 2
 b 2
) a 2
 2b 2
 ( a 2
 b 2
) (b 2
 2 a 2
)
 2 2(a 2  b 2 ) (b 2  2a 2 ) (a 2  b 2 ) 
k  a  2b
2


6ab  (a 2  b 2 ) (b 2  2a 2 ) 2(a 2  b 2 ) (2b 2  a 2 ) 
 2 
 (b  2a ) (a  b ) a  2b 2(a  b ) 
2 2 2 2 2 2 2

32
33
 x  y xy
N 1  1  1   N3 
 3  2  6
x y y x
N 2  1   N 4  1  
3 2 2 3

34
35
VECTOR VARIABLE PROBLEMS

36
Three dimensional stresses
37
Stresses on an elemental cuboid
38
xx + xy + xz + Bx = 0
x y z

zx + zy + zz + Bz = 0 Force


x y z Equilibrium
Equations
yx + yy + yz + By = 0
x y z

Mx = 0 ,My = 0 & Mz = 0 yields


xy = yx ; yz = zy ; zx = xz (2)

39
Strain – displacement relations:-
xx = u
x
yy = v
y
zz = w
z
xy = v + u
x y
yz = w + v
y z
zx = w + u
x z
40
Stress – Strain Relations:-
xx = xx -  ( yy + zz)
E E
yy = yy -  (xx + zz)
E E
zz = zz -  (xx + yy)
E E
xy = xy / G

yz = yz/ G

zx = zx/ G

41
Where E = Young’s Modulus
G = Shear Modulus = E
2 (1 + )
 = Poisson’s ratio

42
The equations (6) can be written in matrix form as

xx 1 - - 0 0 0 xx
yy - 1 - 0 0 0 yy
zz = 1 - - 1 0 0 0 zz
xy E 0 0 0 2(1+) 0 0 xy
yz 0 0 0 0 2(1+) 0 yz
zx 0 0 0 0 0 2(1+) xz

. {} = [C] {}


 {} = [C]-1 {}
= [D] {}
Here the matrix [D] is called the constitutive
matrix given by 43
1-   0 0 0
1-2 1-2 1-2
[D] = E  1-  0 0 0
1+ 1-2 1-2 1-2
  1- 0 0 0
1-2 1-2 1-2
0 0 0 ½ 0 0
0 0 0 0 ½ 0
0 0 0 0 0 ½

44
[D]

45
46
47
48
49
50
u1 
v 
 xx   1 0 2 0 3 0  1

  1   u 2 
 yy   Bd   0 1 0 2 0  3  
  2A  v2 
 xy    1 1  2 2 3  3 
u 3 
 
 v3 

51
{} = [D] {}= DBd
52
STRAIN DISPLACEMENT RELATIONS

{}= Λu = B d

Where B = = ΛN

STRESS STRAIN RELATIONS

{} = [D] {}= DBd

53
2-D APPROXIMATIONS OF 3 – D
PROBLEMS
There exists several problems in solid
mechanics that can be formulated as
three Dimensional problems and the
finite element technique can be used to
solve them.
 However it may turn out to be costly and
time consuming to perform Finite
Element Analysis of 3 D problems.

54
 In several practical situations the
geometry and loading may be such that
the problem can be reduced from 3 D to
2 D or from 2D to 1D.

 The two dimensional idealizations in


stress analysis include

i. PLANE STRESS problems


ii. PLANE STRAIN problems
iii. AXISYMMETRIC problems
55
PLANE STRESS: - A 3D problem can be
reduced to a plane stress condition if it is
characterized by very small dimensions in
one of the normal directions.

A thin plate with a cut out subjected to in-


plane loading.

Thin plate subjected to in-plane loading


56
In these cases the stress components z, xz,
& yz are zero and it is assumed that no stress
component varies across the thickness. The
state of stress is then specified by x, y and
xy only, (functions of x & y) and is called
plane stress. The stress strain relations are
given by
x 1  0 x
y = E  1 0 y
xy 1 – 2 0 0 1– xy
2

57
PLANE STRAIN:- There exist problems
involving very long bodies i.e. a body
whose geometry and loading do not vary
significantly in the longitudinal direction.
Such problems are referred to as plane
strain problems.
Some typical examples include a long
cylinder such as a tunnel, culvert or buried
pipe, a laterally loaded retaining wall, a
long earth dam, and a loaded semi-infinite
half space such as a strip footing on a soil
mass.
58
A long dam

59
In all these problems, the dependant variable
can be assumed to be functions of only x & y
co-ordinates provided that we consider a
cross-section some distance away from the
two ends.
If we further assume that ‘w’ the displacement
component in the ‘z’ direction is zero at every
cross-section, then the non-zero strain
components will be
x = u ; y = v ; xy = u + v
x y y x
and the strain components
60
z, xz, yz will vanish. The dependant stress
variables are x, y & xy and the constitutive
relation for an elastic isotropic material is given
by
x (1-)  0 x
y = E  (1-) 0 y
xy (1 + ) (1 – 2) 0 0 (1 – 2 ) xy
2
It is important to note here that only εz = 0 but z  0.
εz = z -  x   y  
E E E
z  - x  y

61
AXISYMMETRIC PROBLEMS:- Many
engineering problems involve solids of
revolution (axisymmetric solids) subject to
axially symmetric loading.
Examples are a circular cylinder loaded by
uniform internal or external pressure or other
axially symmetric loading as shown in

and a semi – infinite half space loaded by a


circular area. eg., a circular footing on a soil
mass. 62
Because of symmetry the stress components
are independent of the angular co-ordinate ‘’
and hence all the derivatives with respect to ‘ ‘
vanish and the components , ,z, x , y
are zero. The strain displacement relation are
given by
εr = u ; ε = u ; εz = w rz = u + w
x r z z r
The constitutive relations is

63
Now the strain energy stored in an element is given
by
U = ½ v { ε }T { }dv
= ½ v { ε }T [D] { ε }dv
= ½ v [B]T {d} [D] [B] {d}dv

The work done by nodal forces is given by


W = ½ v {F} {d}dv
Equating for a conservative system we get
v ([B]T [D] [B])dv {d} = {F}

i.e. [K] {d} = {F}

where [K] = v[B]T [D] [B] dv


64
Problem 2:- Assuming plane stress conditions evaluate
the stiffness matrix for the element shown in Fig.
Assume E= 2 x 105 N/cm2 and = 0.3. u1 = 0.000, v1
= 0.0025, u2 = 0.0012, v2 = 0.000, u3 = 0.0000 & v3 =
0.0025.

65
 1 = y2 – y 3 = 0 – 1 = -1
 2 = y3 – y 1 =1+1= 2
3 = y1 – y2 = -1– 0 = -1
1 = -(x2 – x3) = 0 – 2 = -2
2 = -(x3 – x1) =0–0= 0
3 = -(x1 – x2) =2–0= 2

66
. A=½xbxh=½x2x2=2

 1 0 2 0  3 0 u1
{} = 1 0 1 0 2 0 3 v1
2A 1 1 2 2 3 3 u2
v2
u3
v3
= {B] {d}

-1 0 2 0 -1 0
[B] = 1 0 -2 0 0 0 2
2(2) -2 -1 0 2 2 -1
67
1  0
[D] = E  1 0
1 – 2 0 0 1-
2

1 0.3 0
= 2 x 105 0.3 1 0
1 – (0.3)2 0 0 1 – 0.3
2

68
69
70

You might also like