EPT371 Finite Element Analysis

SEM 01 2016/2017
TUTORIAL

1. Find the nodal displacements and element internal forces of the two-spring system,
given that

k1 = 10 N/mm k2 = 20 N/mm
F2x = 50 N F3x = 80 N
k1 k2

F2x F3x
Element 1: Element 2:
ì f1x ü é 10 - 10ù ì d1x ü ì f 2 x ü é 20 - 20ù ìd 2 x ü
í ý=ê úí ý í ý=ê úí ý
î f 2 x þ ë- 10 10 û îd 2 x þ î f 3 x þ ë- 20 20 û îd 3 x þ

Using the direct equilibrium method, the global equations

ì F1x ü é 10 - 10 0 ù ì d1x ü
ï ï ê úï ï
í F2 x ý = ê- 10 30 - 20ú íd 2 x ý
ïF ï ê 0 - 20 20 úû ïîd 3 x ïþ
î 3x þ ë

Introducing the boundary condition, d1x = 0, the global equations reduce to

ì F2 x ü é 30 - 20ù ìd 2 x ü
í ý=ê úí ý
î F3 x þ ë- 20 20 û î d 3 x þ

Using the given values for F2x and F3x, the nodal displacements are
ì50ü é 30 - 20ù ìd 2 x ü
í ý=ê ú í ý or
î80þ ë- 20 20 û î d 3 x þ

Thus, d 2x = 13 mm and d 3x = 17 mm

The reaction force F1x = -130 N

Internal forces of element 1: Internal forces of element 2:

ì f 1x ü é 10 - 10ù ì 0 ü ì- 130ü ì f 2 x ü é 10 - 10ù ì13ü ì- 40ü
í ý=ê úí ý = í ýN í ý=ê úí ý = í ýN
î f 2 x þ ë- 10 10 û î13þ î 130 þ î f 3 x þ ë- 10 10 û î17þ î 40 þ

2. Determine the displacement of nodes and reaction forces of the spring system
shown.

1 30 N/mm

50 N/mm
45 N 30 N
2 3 4

25 N/mm
Element 1: Element 2: Element 3:
ì f 1x ü é 30 - 30ù ì d 1x ü ì f 2 x ü é 25 - 25ù ìd 2 x ü ì f 3 x ü é 50 - 50ù ìd 3 x ü
í ý=ê úí ý í ý = ê úí ý í ý=ê úí ý
î f 3 x þ ë- 30 30 û îd 3 x þ î f 3 x þ ë- 25 25 û î d 3 x þ î f 4 x þ ë- 50 50 û îd 4 x þ

Using the direct equilibrium method, the global equations

ì F1x ü é 30 0 - 30 0 ù ì d 1x ü é 30 0 - 30 0 ù ì d1x ü
ïF ï ê 0 - 25 ú ï ï ê
0 ú ïd 2 x ï ê 0 - 25 0 úú ïïd 2 x ïï
ï 2x ï ê 25 25
í ý= í ý= í ý
ï F3 x ï ê- 30 - 25 30 + 25 + 50 - 50ú ïd 3 x ï ê- 30 - 25 105 - 50ú ï d 3 x ï
ïî F4 x ïþ êë 0 0 - 50
ú ê
50 û ïîd 4 x ïþ ë 0 0
ú
- 50 50 û ïîd 4 x ïþ

Introducing the boundary condition, d1x = d 2x = 0, the global equations reduce

ì F3 x ü é 105 - 50ù ì d 3 x ü ì45ü
í ý=ê úí ý = í ý N
î F4 x þ ë- 50 50 û îd 4 x þ î30þ

Thus,
d 3 x = 1.36 mm d 4 x = 1.96 mm

The reaction forces,

ì 0 ü
ï ï
ì F1x ü é30 0 - 30 0ù ï 0 ï ì- 40.9ü
í ý ê = úí ý=í ýN
î F2 x þ ë 0 25 - 25 0û ï1.36ï î - 34.1þ
ïî1.96ïþ
3. Find the nodal displacements and reaction force of the bar-spring system.

1 2 3
1
3000 N

2m

k = 20 N/mm E = 300 x 109 N/m2

A = 600 mm2
Element 1 (Spring): Element 2 (Bar):
ì f 1x ü 3 é 20 - 20ù ì d 1x ü ì f 2 x ü AE é 1 - 1ù ìd 2 x ü
í ý = 10 ê úí ý í ý= í ý
î f 2x þ ë- 20 20 û îd 2 x þ î f 3x þ L êë- 1 1 úû î d 3 x þ

ì f 2 x ü 600 ´ 10 -6 ´ 300 ´ 10 9 é 1 - 1ù ìd 2 x ü
í ý= ê- 1 1 ú íd ý
î f 3x þ 2 ë ûî 3x þ

Using the method of equilibrium, the structural (global) equation is obtained:

ì F1x ü é 20,000 - 20,000 0 ù ì d1 x ü
ï ï ê úï ï
í F2 x ý = ê- 20,000 90,020,000 - 90,000,000ú íd 2 x ý
ïF ï ê - 90,000,000 90,000,000 úû ïî d 3 x ïþ
î 3x þ ë 0
Applying the boundary condition d3x = 0 and substituting the values for the global
forces, we get,
ì3000ü é 20,000 - 20,000 0 ù ì d1 x ü
ï ï ê úï ï
í 0 ý = ê- 20,000 90,020,000 - 90,000,000ú íd 2 x ý
ïF ï ê - 90,000,000 90,000,000 úû ïî 0 ïþ
î 3x þ ë 0

Partitioning (or deleting due to zero-displacement boundary condition), we obtain:

ì3000ü é 20,000 - 20,000 ù ì d1x ü
í ý=ê úí ý
î 0 þ ë - 20,000 90,020,000û îd 2 x þ

Solving the equations, the global displacements

d1x = 0.15 m d2x = 0.03 × 10 -3 m

The reaction force,

F3x = -2700 N

4. Determine the nodal displacements, the reactions and the forces in each element.

2
400 mm2
300 mm 4
3
1 2

F 3
1 2

200 mm 200 mm 250 mm

E = 120 × 103 N/mm2

F = 200 kN

Element 1: Element 2:
ì f1x ü A1 E1 é 1 - 1ù ì d1x ü ì f 2 x ü A2 E 2 é 1 - 1ù ìd 2 x ü
í ý= í ý í ý= í ý
î f 2x þ L1 êë- 1 1 úû îd 2 x þ î f 3x þ L2 êë- 1 1 úû îd 3 x þ
ì f1 x ü 3 é 180 - 180ù ì d1x ü ì f 2x ü 3 é 180 - 180ù ìd 2 x ü
í ý = 10 ê úí ý í ý = 10 ê úí ý
î f 2x þ ë- 180 180 û îd 2 x þ î f 3x þ ë- 180 180 û îd 3 x þ

Element 3:
é 1 - 1ù ìd 3 x ü
ì f 3 x ü A3 E 3
í ý= ê- 1 1 ú íd ý
î f 4x þ L3 ë ûî 4x þ
ì f 3x ü 3 é 192 - 192ù ìd 3 x ü
í ý = 10 ê úí ý
î f 4x þ ë- 192 192 û îd 4 x þ
Using the method of equilibrium, the structural (global) equation is obtained:
ì F1x ü é 180 - 180 0 0 ù ì d1x ü
ïF ï ê- 180 360 - 180 0 úú ïïd 2 x ïï
ï 2x ï 3ê
í ý = 10 í ý
ï F3 x ï ê 0 - 180 372 - 192ú ï d 3 x ï
ïî F4 x ïþ ê ú
ë 0 0 - 192 192 û ïîd 4 x ïþ

Applying the boundary conditions, substituting F2x, and partitioning we get,

ì200 ´ 10 3 ü 3é
360 - 180ù ìd 2 x ü
í ý = 10 ê úí ý
î 0 þ ë - 180 372 û îd 3 x þ

Thus,
d2x = 0.73 mm d3x = 0.35 mm

The reactions,
ì 0 ü
ì F1x ü 3é
180 - 180 0 0 ù ïï0.73ïï ì- 131.9 ü 3
í ý = 10 ê í ý=í ý10 N
î F4 x þ ë 0 0 - 192 192úû ï0.35ï î - 68.1 þ
ïî 0 ïþ
The forces in Element 1:
ì f1 x ü 3é
180 - 180ù ì 0 ü ì- 131.9ü 3
í ý = 10 ê úí ý=í ý10 N
î f 2x þ ë- 180 180 û î0.73þ î 131.9 þ

The forces in Element 2:

ì f 2x ü 3 é 180 - 180ù ì0.73ü ì 68.1 ü 3
í ý = 10 ê úí ý=í ý10 N
î f 3x þ ë- 180 180 û î0.35þ î- 68.1þ

The forces in Element 3:

ì f 3x ü 3é
192 - 192ù ì0.35ü ì 68.1 ü 3
í ý = 10 ê úí ý=í ý10 N
î f 4x þ ë- 192 192 û î 0 þ î- 68.1þ

5. For the plane truss shown, solve for the horizontal and vertical displacements of
node 1 and determine the stress in each element. All elements have E = 210 GPa
and A = 4 × 10 -4 m2.

1 3m
2
3m

45 o 3 4
10 kN
1 3m

20 kN
Element 1:
ì f1 x ü é0 0 0 0 ù ì d1 x ü
ïf ï ê0 280 0 - 280ú ï d ï
ï 1y ï ú ïí 1 y ïý
í ý = 10 ê
5
f
ï 2x ï ê 0 0 0 0 ú ïd 2 x ï
ïî f 2 y ïþ ê ú
ë0 - 280 0 280 û ïîd 2 y ïþ
Element 2:
ì f 1x ü é 140 140 - 140 - 140ù ì d 1x ü
ïf ï ê 140 ï ï
ï 1y ï 5ê
140 - 140 - 140úú ï d1 y ï
í ý = 10 í ý
ï f 3x ï ê- 140 - 140 140 140 ú ï d 3 x ï
ïî f 3 y ïþ ê ú
ë- 140 - 140 140 140 û ïîd 3 y ïþ
Element 3:
ì f1 x ü é 280 0 - 280 0ù ì d1x ü
ïf ï ê 0 ï ï
ï 1y ï 5ê
0 0 0úú ï d 1 y ï
í ý = 10 í ý
ï f 4x ï ê- 280 0 280 0ú ïd 4 x ï
ïî f 4 y ïþ ê ú
ë 0 0 0 0û ïîd 4 y ïþ

Assembling the element equations yields to the structural equations as

ì F1x ü é 420 140 0 0 - 140 - 140 - 280 0ù ì d 1x ü
ïF ï ê 140 ï ï
ï 1y ï ê 420 0 - 280 - 140 - 140 0 0úú ï d1 y ï
ï F2 x ï ê 0 0 0 0 0 0 0 0ú ï d 2 x ï
ï ï ê úï ï
ï F2 y ï 5ê 0 - 280 0 280 0 0 0 0ú ïd 2 y ï
í ý = 10 ê í ý
ï F3 x ï - 140 - 140 0 0 140 140 0 0ú ï d 3 x ï
ê ú
ï F3 y ï ê - 140 - 140 0 0 140 140 0 0ú ï d 3 y ï
ï ï ê- 280 ï ï
ï F4 x ï 0 0 0 0 0 280 0ú ïd 4 x ï
ê ú
ï F4 y ï 0ûú ïîd 4 y ïþ
î þ ëê 0 0 0 0 0 0 0

Introducing the boundary conditions and partitioning the structural equations, we get
ì F1x ü 5é
420 140 ù ì d 1x ü ì - 10,000 ü
í ý = 10 ê úí ý = í ý
î F1 y þ ë140 420û îd 1 y þ î- 20,000þ

Thus,
d1x = -8.93 × 10 -5 m d1y = -44.6 × 10 -5 m

Stress in Element 1:
ì d1x ü
ï ï
c s 0 0ùï d1 y ï
210 ´ 10
[- 1 1]éê
9
s (1) = ú í ý = 31.2 MPa (T)
3 ë0 0 c s û ï d 2 x ï
ïîd 2 y ïþ
Stress in Element 2:
ì d1 x ü
ï ï
210 ´ 10 9 é c s 0 0ù ï d1 y ï
s =
( 2)
[- 1 1]ê ú í ý = 26.5 MPa (T)
3 ë0 0 c s û ï d 3 x ï
ïîd 3 y ïþ
Stress in Element 3:
ì d1 x ü
ï ï
c s 0 0ù ï d 1 y ï
210 ´ 10
[- 1 1]éê
9
s ( 3) = ú í ý = 6.25 MPa (T)
3 ë0 0 c s û ïd 4 x ï
ïîd 4 y ïþ

6. For the two-bar truss shown, determine the displacements of node 1 and the stress
in element 1-3. Show that element 1-3 new orientation can reduce the stress in
element 1-2.
20 kN

500 mm
1
2

E = 200,000 MPa 300 mm

A = 300 mm2 3
400 mm
Element 1-2:
ì f1 x ü é 120 0 - 120 0ù ì d1x ü
ïf ï ê 0 ï ï
ï 1y ï 3ê
0 0 0úú ï d 1 y ï
í ý = 10 í ý
ï f 2x ï ê- 120 0 120 0ú ïd 2 x ï
ïî f 2 y ïþ ê ú
ë 0 0 0 0û ïîd 2 y ïþ
Element 1-3:
ì f1 x ü é 76.8 - 57.6 - 76.8 57.6 ù ì d 1x ü
ïf ï ê- 57.6 43.2 ï ï
ï 1y ï 3ê
57.6 - 43.2úú ï d 1 y ï
í ý = 10 í ý
ï f 3x ï ê- 76.8 57.6 76.8 - 57.6 ú ïd 3 x ï
ïî f 3 y ïþ ê ú
ë 57.6 - 43.2 - 57.6 43.2 û ïîd 3 y ïþ
Assembling the element equations yields to the structural equations as
ì F1x ü é 196.8 - 57.6 - 120 0 - 76.8 57.6 ù ì d1x ü
ïF ï ê- 57.6 43.2 ï ï
ï 1y ï ê 0 0 57.6 - 43.2úú ï d1 y ï
ïï F2 x ïï ê
3 - 120 0 120 0 0 0 ú ïïd 2 x ïï
í ý = 10 ê úí ý
ï F2 y ï ê 0 0 0 0 0 0 ú ïd 2 y ï
ï F3 x ï ê- 76.8 57.6 0 0 76.8 - 57.6 ú ï d 3 x ï
ï ï ê úï ï
îï F3 y þï ëê 57.6 - 76.8 0 0 - 57.6 43.2 ûú îïd 3 y þï
Introducing the boundary conditions and partitioning the structural equations, we get
ì F1x ü 3 é 196.8 - 57.6ù ì d1x ü ì 0 ü
í ý = 10 ê ú í ý=í ý
F
î 1y þ ë- 57.6 43.2 û î 1 y þ î- 20,000þ
d
Thus,
d1x = -0.222 mm d1y = -0.759 mm

Stress in Element 1-3:

ì d 1x ü
ï ï
200 ´ 10 9 é c s 0 0ù ï d 1 y ï
s (1-3) = [- 1 1]ê ú í ý = 111 MPa (C)
0.5 ë0 0 c s û ïd 3 x ï
ïîd 3 y ïþ

Stress in Element 1-2:

ì d1x ü
ï ï
200 ´ 10 9 é c s 0 0ù ï d1 y ï
s (1- 2) = [- 1 1]ê ú í ý = 88.8 MPa (T)
0 .5 ë0 0 c s û ï d 2 x ï
ïîd 2 y ïþ

6.1 The nodal coordinates of the triangular element are shown in Figure P6.1. At the
interior point P, the x-coordinate is 3.5 and N1 = 0.25. Determine N2, N3 and the y-
coordinate at point P.

3 (3,5)
y

P
2 (5,3)

1 (1.5,1)
x

Figure P6.1
Solution:

From Eq.(6.15a), we get,

x = N1x1 + N2x2 + N3x3 ; 3.5 = (0.25)(1.5) + N2(5) + N3(3) ; 3.125 = 5N2 + 3N3

y = N1y1 + N2y2 + N3y3 ; y = (0.25)(1) + N2(3) + N3(5) ; y-0.25 = 3N2 + 5N3

At all points, N1 + N2 + N3 = 1 ; and, 0.75 = N2 + N3

3.125 = 5N2 + 3(0.75 – N2) ; 0.875 = 2N2 ;

Therefore, N2 = 0.4375 and N3 = 0.3125 #

And, y = 0.25 + (3)(0.4375) + (5)(0.3125) = 3.125 #

6.5 For the triangular element shown in Figure P6.5, obtain the strain-displacement
relation matrix [B] and determine the strains ex, ey, and gxy.

q1 = 0.002 q2 = 0.003
y q6 q3 = 0.0015 q4 = 0.002
q5 = 0.001 q6 = 0.004

3 (4,4) q5 q4

q3
q2 2 (6,2)

q1
1 (1,1)
x

Figure P6.5
Solution:

éb i 0 bj 0 bm 0ù é- 2 0 3 0 -1 0 ù
1 ê ú 1 ê
[B] = ê 0 gi 0 gj 0 g m ú = ê 0 - 2 0 - 3 0 5 úú
2A 12
êg i bi gj bj gm b m úû êë- 2 - 2 - 3 3 5 - 1úû
ë

The strains can be found using Eq-6.41,

ì ui ü
ïv ï
ìe x ü éb i 0 bj 0 bm 0 ùï i ï
ï ï 1 ê ú ïï u ïï
{e } = í e y ý = ê 0 gi 0 gj 0 g m úí j ý
ïg ï 2 A ê g v
î xy þ ë i bi gj bj gm b m úû ï j ï
ïu m ï
ï ï
ïîv m ïþ

ì 0.002 ü
ï 0.003 ï
ìe x ü é- 2 0 3 0 - 1 0 ùï ï
ï ïï
{e } = ïí e y ïý = 1 êê 0 - 2 0 - 3 0 5 úú ïí
0 . 0015
ý
ïg ï 12 ê- 2 - 2 - 3 3 ï 0.002 ï
î xy þ ë ú
5 - 1û
ï 0.001 ï
ï ï
ïî 0.004 ïþ

ìe x ü ì- 0.0005ü
{e } = í e y ý = í 0.008 ïý
ï ï 1 ï
ïg ï 12 ï- 0.0075ï
î xy þ î þ
6.11 For the configuration shown in Figure P6.11, determine the deflection at the point
of load application using a one-element model. If a mesh of several triangular
elements is used, comment on the stress values in the elements close to the tip.

90 N

35 mm
2 (0,25) 45 N
1 (35,25)

25 mm t = 8 mm
E = 69,900 MPa
ν = 0.3

(0,0)
Figure P6.11

Solution:

The stiffness matrix for an element is

[k ] = tA[B]T [D][B]

é 0 0 25 0 - 25 0 ù
[B] = 1 ê 0 - 35 0 0 0 25 úú
0.875 ´ 10- 3 ê
êë- 35 0 0 25 25 - 25úû

é 0 0 28.57 0 - 28.57 0 ù
ê
= 10 ê 0
3
- 39.99 0 0 0 28.57 úú
êë- 39.99 0 0 28.57 28.57 - 28.57 úû
The material property matrix is given by

é ù
ê 1 n 0 ú é76.8 23 0 ù
E 9ê
[D] = ên 1 0 ú = 10 ê 23 76.8 0 úú
1- v ê
2
1 -n ú
ê0 0 ú êë 0 0 26.9úû
ë 2 û

Then,
é 0 0 - 1076ù
ê - 920 - 3072 0 úú
ê
[B ]T [D ] = 1012 êê 2194 657 0 ú
ú
ê 0 0 768 ú
ê- 2194 - 657 768 ú
ê ú
ëê 657 2194 - 768 ûú
The element stiffness matrix is

é 0 0 - 1076ù
ê - 920 - 3072 0 ú
ê úé 0 0 28.57 0 - 28.57 0 ù
ê úê
28.57 úú
2194 657 0
[k ] = tA ´10 ê
15
úê 0 - 39.99 0 0 0
ê 0 0 768 ú ê- 39.99 28.57 28.57 - 28.57úû
768 ú ë
0 0
ê- 2194 - 657
ê ú
êë 657 2194 - 768 úû

é 43040 0 0 30741 30741 - 30741ù

ê 0 122880 - 26284 0 26284 - 87767ú
ê ú
ê 0 - 26280 62682 0 - 62682 18770 ú
= 8 ´ 0.4375 ´ 10 9 ê ú
ê - 30720 0 0 21942 21942 - 21942ú
ê - 30720 26280 - 62682 21942 84624 - 40712ú
ê ú
ëê 30720 - 87760 18770 - 21942 - 40712 84624 ûú

Simplifying the equations we get,

é 150640 0 0 107593 107593 - 107593 ù

ê 0 430080 - 91994 0 91994 - 307184ú
ê ú
ê 0 - 91980 219387 0 - 219387 65695 ú
[k ] = 10 ê
9
ú
ê- 107520 0 0 76797 76797 - 76797 ú
ê- 107520 91980 - 219387 76797 296184 - 142492 ú
ê ú
ëê 107520 - 307160 65695 - 76797 - 142492 296184 ûú

For the one element model,

ì F1x ü ì R1x ü ì d 1x ü ì 0 ü
ïF ï ï R ï ïd ï ï 0 ï
ï 1y ï ï 1y ï ï 1y ï ï ï
ïï F2 x ïï ïï 45 ïï ïïd 2 x ïï ïï d 2 x ïï
í ý=í ý = [K ]í ý = [K ]í ý
ï F2 y ï ï- 90ï ïd 2 y ï ïd 2 y ï
ï F3 x ï ï R3 x ï ïd 3x ï ï 0 ï
ï ï ï ï ï ï ï ï
ïî F3 y ïþ ïî R3 y ïþ ïîd 3 y ïþ ïî 0 ïþ

Applying the boundary conditions,

ì 45 ü 9 é 219387 0 ù ìd 2 x ü
í ý = 10 ê í ý.
î- 90þ ë 0 76797úû îd 2 y þ

Therefore, d2x = 0.000205 mm and d2y = -0.001171 mm

10.1 Consider a brick wall of thickness L = 30 cm, K = 0.7 W/m•oC. The inner surface
is at 40oC, and the outer surface is exposed to cold air at -10oC. The heat-transfer
coefficient associated with the outside surface is h = 50 W/m2•oC. Determine the
steady-state temperature distribution within the wall and also the heat flux
through the wall. Use a two-element model, and assume it is a one-dimensional
flow.

Solution

Method 1 (Solution based on text book).

The 2 elements conductivity matrices are,

0.7 é 1 - 1ù
kT(1) = kT( 2)
0.15 êë- 1 1 úû

Thus, the global K,

é 1 -1 0 ù
K = 4.667êê- 1 2 - 1úú
êë 0 - 1 1 úû

Now, since convection occurs at node 3, the constant h = 50 is added to the (3,3)
location of K. Therefore,

é 1 -1 0 ù
ê
K = 4.667ê- 1 2 - 1 úú
êë 0 - 1 11.714úû

Since no heat generation Q occurs in this problem, the heat rate vector R consists only
of hT∞ in the last row. That is,

R = [0 0 50 ´ (-10) ] = [0 0 - 500]
T T

The specified temperature boundary condition T1 = 40 oC, will now be handled by the
penalty approach. We choose C based on,

C = max K ij ´ 104 = 4.667 ´ 11.714 ´ 104 = 546692.38

Now C gets added to (1,1) location of K, while CT1 is added to the first row of R. The
resulting equations are
 é117141 - 1 0 ù ìT1 ü ì21867695ü
ê ï ï ï ï
4.667 ê - 1 2 - 1 úú íT2 ý = í 0 ý
êë 0 - 1 11.714úû îT3 þ î - 500 ïþ
ï ï ï

The solution is,

T = [40 16.1 - 7.8]

T o
C

Therefore, the heat flux through the wall is,

dT (T - T )
q (1) = - k = -0.7 2 1 = -0.7(-159.333) = 111.53 W/m2. #
dx x2 - x1

Method 2 (Based on notes)

For element 1,
1 - 1ù hPL é2 1 ù
[ke ] = AK xx éê ú + ê ú
é0 0ù
+ hAê ú
L ë- 1 1 û 6 ë1 2û ë0 1 û

Assuming the element having a unit cross-sectional area → A = 1 m2. For element 1,
there is no convection occurs → h = 0.
1 - 1ù é 4.667 - 4.667ù
[k1 ] = (1)(0.7) éê =
0.15 ë- 1 1 úû êë- 4.667 4.667 úû

For element 2, there is an additional (convection) term owing to heat loss from the flat
surface at its right end. Hence,

0 0ù é 4.667 - 4.667ù é0 0 ù é 4.667 - 4.667ù

[k2 ] = [k1 ] + hAéê ú=ê ú+ê ú=ê ú
ë0 1 û ë- 4.667 4.667 û ë0 50û ë- 4.667 54.667 û
To obtain the element force matrices, we use Eq.7.41, Eq.7.42, Eq.7.43 and Eq.7.47.
However, in this question, Q = 0 (no heat source), q* = 0, and there is no convection
except from the right end.

Therefore, for element 1,

ì f1x ü ì0ü
=
For element 2, íî f 2 x ýþ íî0ýþ

ì f2x ü ì0ü ì0ü ì 0 ü

í ý = hT¥ Aí ý = (50)(-10)(1) í ý = í ý
î f3x þ î1þ î1 þ î- 500þ

The assemblage of the element matrices and the force matrix yields,

é 4.667 - 4.667 0 ù ìt1 ü ì F1 ü

ê- 4.667 9.333 - 4.667ú ït ï = ï 0 ï
ê úí 2 ý í ý
êë 0 - 4.667 54.667 úû ît3 þ î- 500ïþ
ï ï ï
Substituting the known temperature t1 = 40 oC, and transposing known terms to the right
side, we have

é1 0 0 ù ì t1 ü ì 40 ü
ê0 9.333 - 4.667ú ït ï = ï186.68ï
ê úí 2 ý í ý
ï ï ï ï
ëê0 - 4.667 54.667 ûú ît3 þ î - 500 þ

Solving the equations simultaneously, we get

t2 = 16.1 oC t3 = -7.8 oC

The heat flux for element 1 is,

é 1 1 ù ì 40 ü
q (1) = - K xx [ B]{t} = -0.7 ê- í ý = 111.53 W/m . #
2

ë 0.15 0.15 úû î16.1þ

10.2 Heat is entering into a large plate at the rate of q0 = -450 W/m2 as shown. The
plate is 100 mm thick. The outside surface of the plate is maintained at a
temperature of 15 oC. Using two finite elements, solve for the vector of nodal
temperatures T. Thermal conductivity k = 1.0 W/m•oC.

15 oC

q0

100 mm

Solution:

The 2 elements conductivity matrices are,

1.0 é 1 - 1ù
kT(1) = kT( 2)
0.05 êë- 1 1 úû

Thus, the global K,

é 1 -1 0 ù
K = 20 êê- 1 2 - 1úú
êë 0 - 1 1 úû

Since no heat generation Q occurs in this problem, the heat rate vector R is added with
the specified heat flux on the boundary, q 0 (as per Eq.10.29) which resulted to

R = [- (-450) 0 0) ]
T
The resulting equations for the system are

é 20 - 20 0 ù ìT1 ü ì450ü
ê- 20 40 - 20ú ïT ï = ï 0 ï
ê úí 2 ý í ý
ëê 0 - 20 20 ûú îT3 þ î 0 ïþ
ï ï ï

The boundary condition T3 = 15 oC is handled by the elimination approach.

The resulting equations are

é 20 - 20ù ìT1 ü ì450ü T To

ê- 20 40 ú íT ý = í300ý . Thus [T1 T2] = [60 37.5] C #
ë ûî 2 þ î þ

10.3 The outside of a heating tape is insulated, while the inside is attached to one face
of a 4 cm thick stainless steel plate (k = 16.6 W/m•oC). The outer face of the
plate is exposed to the surroundings, which are at a temperature of 20 oC. Heat is
supplied at a rate of 300 W/m2. Determine the temperature of the face to which
the heating tape is attached.

Solution:

The 2 elements conductivity matrices are,

16.6 é 1 - 1ù
kT(1) = kT( 2)
0.02 êë- 1 1 úû

Thus, the global K,

é 1 -1 0 ù
K = 830êê- 1 2 - 1úú
ëê 0 - 1 1 úû

Now, since convection occurs at node 3, the constant h = 10 is added to the (3,3)
location of K. Therefore,

é 830 - 830 0 ù
K = ê- 830 1660 - 830úú
ê
êë 0 - 830 840 úû

Since no heat generation Q occurs in this problem, the heat rate vector R consists only
of hT∞ in the last row and it is added with the specified heat flux on the boundary, q0
which resulted to

R = [- ( -300) 0 (10 ´ 20) ]

T

The resulting equations for the system are

é 830 - 830 0 ù ìT1 ü ì300ü
ê- 830 1660 - 830ú ïT ï = ï 0 ï
ê úí 2 ý í ý
êë 0 - 830 840 úû îT3 þ î200ïþ
ï ï ï

Solving the equations we get , [T1 T2 T3]T = [50.7 50.4 50]T oC #

10.4 Consider a pin fin having a diameter of 0.3 in and length of 6 in. At the root, the
temperature is 147 oF. The ambient temperature is 80 oF and h = 5
BTU/(hr•ft2•oF). Take k = 25.5 BTU/(hr•ft•oF). Assume that the tip of the fin is
insulated. Using a two-element model, determine the temperature distribution
and heat loss in the fin.

Solution:
Element matrices,

Ph le é2 1 ù pd (5)(0.25) é2 1 ù é2 1 ù
hT = ê ú = ê ú = 33.333ê ú
Ac 6 ë1 2û (pd / 4)6 ë1 2 û
2
ë1 2û

k é 1 - 1ù é 1 - 1ù
kT =
le ê- 1 1 ú = 102ê- 1 1 ú
ë û ë û

PhT¥ le ì1ü pd (5)(80)(3) ì1ü ì1ü

r∞ = í ý= í ý = 8000í ý
Ac 2 î1þ (pd / 4)(2) î1þ
2
î1þ

Upon assembly the above matrices, we obtain

é 2 1 0ù
HT = 33.333êê1 4 1 úú
êë0 1 2úû

é 1 -1 0 ù
KT = 102êê- 1 2 - 1úú
ëê 0 - 1 1 úû

ì1 ü
ï ï
R∞ = 8000í2ý
ï1 ï
î þ
Thus,

(HT + KT)T= R∞ →
é é66.667 33.333 0 ù é 102 - 102 0 ù ù ìT1 ü ì 8000 ü
êê ú ê úúï ï ï ï
ê ê33.333 133.333 33.333ú + ê- 102 204 - 102ú ú íT2 ý = í16000ý
êë êë 0 33.333 66.667úû êë 0 - 102 102 úû úû ïîT3 ïþ ïî 8000 ïþ

é 168.667 - 68.667 0 ù ìT1 ü ì 8000 ü

ï ï ï ï
→ - 68.667 337.333 - 68.667ú íT2 ý = í16000ý
ê
ê ú
êë 0 - 68.667 168.667 úû ïîT3 ïþ ïî 8000 ïþ

Given the boundary condition, T1 = 147 oF. Using the elimination approach, we get

é 337.333 - 68.667ù ìT2 ü ì16000ü ì- 10094.05ü ì26094.049ü

ê- 68.667 168.667 ú íT ý = í 8000 ý - í ý=í ý
ë ûî 3 þ î þ î 0 þ î 8000 þ

The solution is

[T2 T3] = [94.9 86.0]oF

The total heat loss in the fin can be computed as

H = å He
e

The loss He in each element is

H e = h(Tav - T¥ ) As

where As is the surface area and Tav is the average temperature within the element.

Thus, H = 5 ´ 0.0196(120.95 - 80) + 5 ´ 0.0196(90.45 - 80) = 5.04 BTU/hr #

OOOOOOO