You are on page 1of 24

Chapter 3

SOLUTIONS

3.1.Sol.
  πx   πy   3πx   3πy  
 sin   sinh   sin   sinh   
θ T − T1
=     a +  a   a  + ...
4 a
=
θm Tm − T1 π   πb   3πb  
 sinh   3 sinh   
  a   a  

x = 0.3 m, y = 0.2 m, a = 0.6 m and b = 0.4 m, T1 = 100ºC and Tm = 400ºC


πx π π
= ∴ sin   = 1
a 2 2
πy π π
= ∴ sinh   = 1.254
a 3 3
x 3π  3π 
3π = ∴ sin   = −1
a 2  2 
y
3π =π ∴ sinh (π) ≈ 11.65
a
πb 3π  2π 
= ∴ sinh   = 4.02
a 3  3 
 πb 
3   = 2π ∴ sinh (2π) = 270
 a 
Tp − T1 4 1 × 1.254 (−1) × 11.65 
∴ = ×
Tm − T1 π  4.02 270 

Tp − 100 4
= [0.312 − 0.043] = 4 (0.269)
400 − 100 π π
4
∴ Tp = 100 + 300 × × 0.269 = 203°C
π
If Tm = −60°C, then

95
Tp − 100 4
= (0.269)
− 60 − 100 π
160 × 4
∴ Tp = 100 − (0.269) = 100 − 55 = 45°C.
π
The heat flow along the boundary BC per unit depth is given by the following
equation
  πb     3πb   
 cosh   − 1  cosh  − 1 
4π Kθm   a  + 1  a   + ...
q=    
a 2  sinh  πb   3  sinh  3πb   
     
  a     a   
 πb   2π 
cosh   = cosh   = 4.14
 a   3 
 3πb 
cosh   = cosh (2π) = 270
 a 
Substituting these values in the above equation
4π × 35 × 300  4.14 − 1 1 270 × 1
q=  4.02 + 3 × 270 
0.36  
=366520 [0.78 + 0.332] = 407570 Watts = 407.6 kW
If Tm = −60°C then θm = Tm − T1 = −60 − 100 = −160°C
(−160)
For this case q = 407.6 × = −217.4 kW
300
This shows that there is absorption of heat at the boundary surface.

3.2.Sol.
We can use the following equation for finding out the temperature distribution
  πy   3πy  
T − T1 4  − π   πx  1 − a   3πx 
= (e ) ⋅ sin   + (e) ⋅ sin   + ...
Tm − T1 π   a 3  a  
 
a = 0.5 m, x = 0.2 m, y = 0.5 m
πy π × 0.5
= =π ∴ (e) − π = 0.0125
a 0.5

96
πx π × 0.2 2π  2π 
= = ∴ sin   = 0.9511
a 0.5 5  5 
3πy 1
= 3 × π = 3π ∴ ( e ) − 3π =
a 12000
 3πy 
− 
 3πx 
∴ (e)  a  sin →0
 a 

Substituting these values the above equation


Tp − 100 4
= [0.0425 × 0.9511] = 0.0514
500 − 100 π

Tp = 100 + 400 × 0.0514 = 100 + 20.56 = 120.56°C

3.3.Sol.
This problem can be solved by using the following expression:

T − T1 4 ∞  nπx  − ⋅y
= ∑ sin   ⋅ ( e) a
Tm − T1 π n =1,3  a 

The given data is


T1 = 0, Tm = 100°C


400 ∞  nπx  − ⋅y
∴ T= ∑
π n =1,3
sin 
 a 
 ( e) a

If the width a = π, then

400 ∞
T= ∑ sin (nx) ⋅ (e)− ny
π n =1,3

3.4.Sol.
The problem can be solved by using the following expression

97
 nπy 
∞ sinh  
T − T1 4 1  nπx   a 
= ∑ ⋅ sin  ⋅
Tm − T1 π n =1, 3 n  a  sinh  nπb 
 
 a 
The given data is
T1 = 0°C, Tm = 100°C, a = b = 100 cm

Substituting these values in the above equation


 nπy 
∞ sinh  
400 1  nπx   100 
T ( x , y) = ∑ sin  ⋅
π n =1,3 n  100  sinh (nπ)

Now substituting x = y = 50 cm
 nπ 
∞ sinh  
400 1  nπ   2 
T0 (at mid point) = ∑
π n =1, 3 n
sin   ⋅
 2  sinh (nπ)

 nπ 
∞ sinh  
400 1  2  200 ∞ 1 1
= ∑
π n =1,3 n
× (1) ⋅
 nπ   nπ 
= ∑
π n =1,3 n  nπ 
2 sinh   ⋅ cosh   cosh  
 2   2   2 
Taking n = 1 and n = 3
 
200  1 1 1 
Ta =  + ⋅ 
π   nπ  3  3π  
 cosh  2  cosh  
 2 

200  1 1 1  200
=  + ⋅  = = 25.5°C
π  2.507 3 54.97  π × 2.507

3.5.Sol.
The temperature at the edge are shown in Fig. Prob. 3.5.
The temperature solution T (x,y) is given by
T ( x, y) = T2 ( x , y) + T4 ( x , y)

98
Fig. Prob. 3.5

Where T2 (x,y) and T4 (x,y) are given by the following expressions



 nπx   nπy 
T2 ( x , y) = ∑ D n sin   ⋅ sinh  
n =1  a   a 
a
2  nπx 
Where Dn = ∫
 nπb  0
f 2 ( x ) ⋅ sin 
 a 
 ⋅ dx
a sinh  
 4 

 nπy   nπx 
And T4 ( x , y) = ∑ Bn sin b 
 ⋅ sinh 
 b 

n =1

b
2  nπy 
Where Bn =
 nπa 0∫ f 4 ( y) ⋅ sin 
 b 
 dy
b sinh  
 b 
In this problem, the given data is
a = b = L and f 2 ( x ) = f 4 ( y) = 100°C

 nπx   nπy 
∴ T2 ( x , y) = ∑ Dn sin L 
 ⋅ sinh 
 L 

n =1

L
2  nπx 
and Dn = ∫
L sinh (nπ) 0
100 ⋅ sin 
 L 
 ⋅ dx

99
L
200  L  nπx  200
=  − cos   = [1 − cos nπ]
L sinh (nπ)  nπ  L  0 nπ sinh (nπ)

400 ∞ 1
= ∑
n n =1,3 n sinh (nπ)
as (1 − cos nπ) = 0 for n = 2,4,6.

As (1 − cos nπ) = 2 for n = 1 , 3, 5


400 ∞ 1  nπx   nπy 
∴ T2 ( x , y) = ∑
π n =1,3 n sinh nπ
⋅ sin 
 L 
 ⋅ sinh 
 L 
 (1)


 nπy   nπx 
and T4 ( x , y) = ∑ Bn sin L 
 ⋅ sinh 
 L 

n =1,3

L
2  nπy  400 ∞ 1
and Bn = ∫
L sinh (nπ) 0
100 ⋅ sin 
 L 
dy = ∑
π n =1,3 n sinh (nπ)

400 ∞ 1  nπy   nπy 


∴ T4 ( x , y) = ∑
π n =1,3 n sinh (nπ)
⋅ sin 
 L 
 ⋅ sinh 
 L 
 (2)

∴ T ( x , y) = T2 ( x , y) + T4 ( x , y)

The temperature at any point can be calculated by combining the values of the
expressions 1 and 2.

3.6.Sol.
The temperature at the edge are shown in Fig. Prob.3.6.
The temperature solution T(x,y) is given by
T ( x , y) = T1 ( x , y) + T3 ( x , y)
Where T1 ( x , y) and T3 ( x , y) are given by the following expressions

  nπy   nπb   nπy   nπx 
T1 ( x , y) = ∑ Dn sinh  a 
 − tanh 
 a 
 cosh   sin 
 a   a 

n =1, 2 
a
2  nπx 
Where Dn = − ∫
 nπb  0
f1 ( x ) sin 
 a 
 dx
a tanh  
 a 

100
Fig. Prob. 3.6.


 nπy    nπx   nπa   nπx 
and T3 ( x, y) = ∑ Bn sin 
 b 
 sinh 
 b 
 − tanh 
 b 
 cosh 
 b 

n =1, 2
b
2  nπy 
Where Bn = ∫
 nπa  0
f3 ( y) sin 
 b 
 dy
b tanh  
 b 

In this problem, the given data is


a = b = L and f1 ( x ) = f 3 ( y) = 100°C

 nπx    nπy   nπy 
T1 ( x, y) = ∑ Dn sin   sinh 
L   L 
 − tanh (nπ) ⋅ cosh 
 L 

n =1

L
2  nπx  200 L
Where Dn = ∫
L tanh (nπ) 0
100 ⋅ sin 
 L 
 ⋅ dx = − × [1 − cos nπ]
L tanh (nπ) nπ

200 − 400 ∞ 1
=−
nπ tanh (nπ)
(1 − cos nπ) = ∑
π n =1,3 n tanh (nπ)

 nπx 
∞sin  
400  L   nπy   nπy  
∴ T1 ( x , y) = − ∑ 
π n =1,3 n tanh (nπ) 
sinh 
 L 
 − tanh (nπ) ⋅ cosh
 L 
 (1)


 nπy    nπx   nπx 
T3 ( x , y) = ∑ Bn sin   sinh 
L   L 
 − tanh (nπ) ⋅ cosh 
 L 

n =1

101
L
2  nπy  400 ∞ 1
Where Bn = − ∫
L tanh (nπ) 0
100 ⋅ sin 
 L 
 ⋅ dy = − ∑
π n =1,3 n tanh (nπ)

 nπy 
sin 
∞ 
400  L    nπx   nπx 
∴ T3 ( x , y) = − ∑ 
π n =1,3 n tanh (nπ) 
sinh 
 L 
 − tanh (nπ) ⋅ cosh 
 L 
 (2)

∴ T ( x , y) = T1 ( x , y) + T3 ( x , y)

The temperature at any point can be calculated by combining the expressions


(1) and (2).

3.7.Sol.
The temperature along the edges are shown in Fig. Prob. 3.7.

Fig. Prob. 3.7.

The temperature solution T (x,y) is given by


T (x,y) = T3 (x,y) = T4 (x,y).

Where T3 (x,y) and T4 (x,y) are given by the following expressions.



 nπy    nπx   nπx 
T3 ( x , y) = ∑ Bn sin   sinh   − tanh (nπ) ⋅ cosh  
n =1  L   L   L 
L
2  nπy 
Where Bn = − ∫
L tanh (nπ) 0
f3 ( y) ⋅ sin 
 L 
 ⋅ dy

102
As f 3 ( y) = 100°C given
L
2  nπy  400 ∞ 1
∴ Bn = − ∫
L tanh (nπ) 0
100 ⋅ sin 
 L 
 dy = − ∑
π n =1,3 n tanh (nπ)

 nπx 
sin  ∞ 
400  L    nπx   nπx 
∴ T3 ( x , y) = − ∑ 
π n =1,3 n tanh (nπ) 
sinh 
 L 
 − tanh (nπ) ⋅ cosh 
 L 
 (1)

And T4 (x,y) is given by



 nπy   nπx 
T4 ( x , y) = ∑ Bn sin  L 
 ⋅ sin 
 L 

n =1,3

L
2  nπy 
Where Bn = ∫
L sinh (nπ) 0
f 4 ( y) ⋅ sin 
 L 
 ⋅ dy

As f 4 ( y) = 100°C.
L
2  nπy  400 ∞ 1
∴ Bn = − ∫
L sinh (nπ) 0
100 ⋅ sin 
 L 
 ⋅ dy = − ∑
π n =1,3 n sinh (nπ)

(nπy)  nπx 
∞ sin ⋅ sinh  
400 L  L 
∴ T4 ( x , y) = ∑
π n =1,3 n sinh (nπ)
(2)

∴ T (x, y) = T3 ( x , y) + T4 ( x , y).
The temperature at any point is given by combining the expressions (1) and (2).

3.8.Sol.
The given condition of temperature are shown in Fig. Prob.3.8.
The controlling differential equation is given by:
∂ 2T ∂ 2T
+ =0
∂x 2 ∂y 2
Assuming the solution of the above equation in the form as
T = X .Y

103
d 2X d 2Y
∴ Y= +X =0
dx 2 dy 2
1 d 2X 1 d 2Y
∴ 2
= − 2
= ±k 2
X dx Y dy

Fig. Prob. 3.8.

The solution must contain sine and cosine terms in x, therefore - K2 should be taken
d 2X 2 d 2Y
∴ 2
+ XK = 0 and 2
− YK 2 = 0
dx dy
∴ X= (A cos Kx + B sin Kx) and Y = (C1 cosh Ky + D1 sinh Ky)
∴ T = (A cosKx + B sin Kx) (C1 cosh Ky + D1 sinh Ky)

Now using the first boundary condition


T(0, y) = (A + 0)[C1 cosh Ky + D1 sinh Ky] = 0
∴ A=0
∴ X = B sin Kx [C1 cosh Ky + D1 sinh Ky] = sin Kx [C cosh Ky + D sinh Ky]

Using the second boundary condition


T(L, y) = sin KL [C cosh Ky + D sinh Ky] = 0
∴ sin KL = 0 = sin n π

∴ K=
L
 nπx    nπy   nπx 
∴ T ( x, y) = sin   C cosh   + D sinh  
 L   L   L 

104
∂T  nπx    nπ   nπy   nπ   nπy 
= sin   C  −  sinh   + D   cosh  
∂y  L   L   L   L  L 

 nπ   nπx    nπy   nπy  


=   sin   − C sinh   + D cosh  
 L   L   L   L 

Now using the third boundary condition


∂T  nπ   nπx 
( x ,0) =   sin   [0 + D ] = 0
∂y  L  L 
∴ D=0
 nπx   nπy   nπx   nπy 
∴ T ( x , y) = sin   ⋅ C cosh   = C ⋅ sin   ⋅ cosh   (a)
 L   L   L   L 

Now using the fourth boundary condition


 nπx 
T( x , L) = C sin   ⋅ cosh (nπ) = f(x)
 L 

The value of 'C' can be given as


L
2  nπx 
C ⋅ cosh (nπ) = ∫ f ( x ) ⋅ sin   ⋅ dx
L0  L 

Substituting the given value of f(x), we get


L ∞ ∞
2T  nπx  2T 2L 4Ti 1
C ⋅ cosh(nπ) = i ∫ sin 
L 0  L 
 ⋅ dx = i ∑ = ∑
L n =1,3 nπ π n =1,3 n
4Ti 1
∴ C=
π n ⋅ cosh (nπ)
Substituting this value of C in equation (a), we get
 nπx   nπy 
∞ sin   ⋅ cosh  
4Ti  L   L 
T ( x , y) = ∑
π n =1,3 n ⋅ cosh (nπ)

 (2n − 1)πx   (2n − 1)πy 


∞ sin   ⋅ cosh  
4T
= i ∑  L   L 
π n =1, 2 (2n − 1) cosh[(2n − 1)π]

105
3.9.Sol.
The controlling differential equation for the above problem is
∂ 2T 1 ∂T 1 ∂ 2T
+ + =0
∂r 2 r ∂r r 2 ∂θ2

The general solution of this equation is given by


T = (Cr λ + Dr −λ )[A1 cos λθ + B1 sin λθ]

The given boundary conditions are


T(r, 0) = 0 and T(r, π ) = 0
T(0, θ) = 0 (as given) but it must be finite and T(a , θ) = f (θ)

Now using the boundary condition.


 D
T (0, θ) = 0 =  0 +  [A1 cos λθ + B1 sin λθ]
 0
∴ D=0
∴ T = Cr λ [A1 cos λθ + B1 sin λθ] = r λ [A cos λθ + B sin λθ]

Now using the boundary condition


T ( r , 0) = r λ [ A + 0] = 0
∴ A=0

∴ T = r λ ⋅ B sin λ θ.

Now using the third boundary condition


T (r, π) = 0 = r λ ⋅ B sin λπ
As B≠0 ∴ sin λπ = 0 = sin nπ
∴ λ=n

∴ T = Br n ⋅ sin (nθ) = ∑ Bn r n sin (nθ) (a)
n =1, 2

Now using the last condition



T (a , θ) = f (θ) = ∑ Bn ⋅ a n ⋅ (sin nθ)
n =1, 2

106
The constant Bn can be calculated by using the Fourier sine series expansion
π
2
Bn ⋅ a = ∫ f (θ) ⋅ sin nθ ⋅ dθ
n
π0
π
2 1
π a n ∫0
∴ Bn = ⋅ f (θ) ⋅ sin nθ ⋅ dθ

Substituting the value of Bn in the equation (a), we get the expression for
temperature distribution.
(A) If f (θ) = Ta is given, then
π ∞
2Ta 1 2Ta  1  4T 1
Bn = ⋅ n ∫ sin nθ dθ =
π a 0 n 
π ⋅ a n
(1 − cos(1 − cos nπ) = a
 π
∑ n
n =1,3 n ⋅ a

Substituting this value in the equation (a), we get


∞ n
4Ta 1 r 
∴ T=
π
∑ n  a  sin (nθ).
n =1,3

∞ 2 n −1
4Ta 1 r
=
π
∑ (2n − 1) ⋅  a  ⋅ sin [(2n - 1) θ]
n =1,2

π
(B) If f (θ) = 500, 0<θ< and
2
π
= 50(π − θ), <θ<π
2

Then, the constant Bn is given by


2 1  
π/2 π
Bn = ⋅ n  ∫ 50.θ sin nθ.dθ + ∫ 50( π − θ) ⋅ sin nθ.dθ
π a  0 π/2 

100 1  
π/2 π
= ⋅ n  ∫ θ sin nθ.dθ + ∫ (π − θ) ⋅ sin nθ.dθ
π a  0 π/2 
π
100 1   cos nθ  sin nθ π / 2   cos nθ  sin nθ 
= ⋅ θ − +  + ( π − θ)− − 
π an   n  n 2 0   n  n 2 
π/2

107
100 1  π  nπ  1  nπ    π  nπ  1  nπ 
= ⋅ n  − ⋅ cos   + 2 ⋅ sin   +  cos   + 2 sin  
π a  2n  2  n  2   2n  2  n  2  

100 1 2  nπ  200 1  nπ 
= ⋅ n ⋅ 2 sin   = 2 ⋅ 2 sin  
π a n  2  πa n  2 

This expression is valid only for n = 1, 3, 5.


200 ∞ 1  nπ 
∴ Bn = ∑ 2
π n =1,3 a n 2
⋅ sin  
 2 

Substituting this value of Bn in the equation (a), we get


n
200 ∞ 1  r   nπ 
T= ∑ 2
  ⋅ sin   sin nθ
π n =1,3 n  a   2 
2 n −1
200 ∞ 1 r  (2n − 1) π 
= ∑  
π n =1, 2 (2n − 1) 2  a 
⋅ sin 
 2  ⋅ sin [(2n - 1)θ]

2 n −1
200 ∞ (−1) n +1  r 
= ∑ ⋅ 
π n =1, 2 (2n − 1) 2  a 
⋅ sin [(2n − 1)θ].

3.10.Sol.
The heat flow is given by
q = K ⋅ Fs (Tp − Ts )
Where Tp and Ts are pipe surface and ground surface temperatures.
K = Conductivity of soil.
and Fs = Conduction shape factor of the pipe.
2πL
Fs (From table) =
  2H  
 log e  
L  L 
loge   1 +
 r  L 
 log e   
 r 
L = 30 m, r = 0.05 m, H = 2 m

108
2π × 30
∴ Fs =
  4 
 loge   
 30    30  
log e   1+
 0.05   log  30  
 e 
  0.05  
60π 60π
= ≈ = 42.8
 30   4  6.4 − 2.01
log e   + log e  
 0.05   30 
q = 0.3 × 12.8 × (150 − 20) = 1670 Watts.

2.11.Sol.
The heat flow is given by the equation
q = KFs (Ts − Tg )

where Ts and Tg are sphere surface and ground surface temperatures.


The value of Fs can be calculated by using the formula given in the table for the given
configuration.
4πr
Fs =
r
1−
2H
r = 2 m and H = 10 m.
4π × 2 8π
∴ Fs = = = 27.9
2 0 .9
1−
20
Q q = 0.3 × 27.9(200 − 20) = 1505 Watts.

2.12.Sol.
If we define θ = T − T1, then the equation for hear distribution in the plate
∂ 2θ ∂ 2θ
+ =0
∂x 2 ∂y 2

109
with the boundary conditions.
(i) θ=0 at x = 0,L

(ii) θ=0 at y=∞


(iii) θ = f ( x ) − T1 = f ( x ) at y = 0

Using the separation of variable method, we assume a product solution


θ = XY where X = X (x) and Y = Y (y). This reduces the Laplace equation for θ to
1 d 2X 1 d 2Y
= −
X dx 2 Y dy 2

d 2X
or 2
+ λ2 X = 0
dx
d 2Y
2
+ λ2 Y = 0
dy
With the general solution
X = C1 cos λx = C 2 sin λx

Y = C3eλy + C 4e − λy

∴ θ = (C1 cos λx + C 2 sin λx) (C3eλy + C4e −λy )


The application of boundary condition to the above equation gives
(i) when θ=0 at x = 0 ; C1 = 0
(ii) when θ=0 at y = ∞ ; C3 = 0

∴ θ = C 2C 4e − λy (sin λx ) = Ce −λy (sin λx )


(iii) when θ=0 at x = L, then sin λL = 0

or λ= ; n = 1, 2, 3, ….
L

 nπx 
Hence θ= ∑ Cn e−(nπ / L) y  sin L 

n =1

(iv) when θ = F ( x ) at y = 0, we have

110

nπx
F( x ) = ∑ Cn sin L
n =1

Which is a Fourier expression of F (x) in an infinite sine series


L
2 nπx
∴ C n = ∫ F ( x ) sin dx
L0 L

Hence the final solution is


L
2 ∞ − ( nπ / L ) y nπx nπx
θ = T − T1 = ∑
L n =1
e sin
L 0∫ F (x) sin
L
dx

Which is the required expression for the temperature distribution in the semi-
infinite plate.

3.13. Sol.
The heat flow is given by the equation
q = KFs (Tp1 − Tp 2 )

where Tp1 and Tp2 are pipe surface temperatures.


The value of Fs can be calculated by using the formula given in the table for the
given configuration.
2πL
Fs =
 2 2 2
−1 D − r1 − r2
cosh  
 2r1r2 
r1 = 0. m, r2 = 0.1 m, D = 1 m and L = 50 m
2π × 50
∴ Fs =
1 − 0.09 − 0.01
cosh −1  
 2 × 0.3 × 0.1 
100π 100π
∴ Fs = −1
= = 101
cosh (11.2) 3.1

∴ q = 0.3 × 101 (150 − 10) = 4242 Watts.

111
1
ww (mass of water) = πr22 ⋅ V ⋅ ρ = π (0.1) 2 × × 1000 = 0.52 kg/sec
60
Now q = m ⋅ c p ⋅ (∆T)

q 4242
∴ ∆T = = = 1.94°C.
m ⋅ c p 0.52 × 4200

3.14. Sol.
The heat flow is given by
q = KFs (Ts − Tg )
q 2000
∴ Fs = = = 6.25 m.
K (Ts − Tg ) 1.6 × 200

The shape factor for the given system is given by


2πa
Fs = where a >> b and H > 2b
 2πH 
log e  
 b 
2π × 3
∴ 6.25 =
 2π × H 
log e  
 0.6 
∴ loge (10.5 H) = 3
( e) 3 20
∴ H= = = 1.91 m.
10.5 10.5

3.15.Sol.
The temperature distribution and heat flow lines are plotted as shown in fig.
Prob. 3.15. The heat flow rate can be calculated from the plotted field in Fig. Prob 3.15. The
isotherms and heat flow lines intersect to form curvilinear squares. The shape factor can be
readily determined by noting that the number of temperature increments is 8; the number of
heat flow channels for the square plate is 16. The shape factor is thus
N 16
S= = =2
M 8

and the heat loss through the square plate becomes


q = k ⋅ S ⋅ ∆T = k ⋅ 2 ⋅ ∆T = 200k.

112
3.16.Sol.
The following technique is employed to find the steady state temperature
distribution:
(i) Temperature at the boundaries are fixed as shown in Fig. Prob.3.16.
(ii) Divide the section into a number of arrays to get nodes (marked 1, 2, 3, ........ 16,
as shown)
(iii) Assign initial temperature values that might appear reasonable and compatible
with the boundary conditions at the nodes. These intial temperatures are shown to
the right of the nodal points,
(iv) Calculate the residuals for each of the nodal points by using Eq.
T1 + T2 + T3 + T4 − 4T0 = R. .
(v) Compare all the equations for maximum residual and then irrespective of the
sign (+ve or -ve) make that particular residual zero by changing the
temperature at that node.
(vi) Now calculate the new value of residuals for the affected neighbouring
nodal points.
(vii) Continue repeating steps (v) and (vi) until all the residuals for different nodal
points are reduced to zero or almost zero.
Following this procedure, the temperature obtained at different nodes, in various
interations are as shown in Fig. Prob. 3.16. The final equations obtained are:

Node No. Sum of Neighbouring Node Temp 4 × Node Temp. Residue


1. (100 + 22.96 + 0 + 60) − 4 × 45.856 = − 0.464
2. (33.61 + 0 + 11.565 + 45.856) − 4 × 22.96 = − 0.809
3. (22.96 + 5.01+ 0 +17.885) − 4 × 11.565 = − 0.405
4. (0 + 0 + 7.69 + 11.565) − 4 × 5.01 = − 0.785
5. (100 + 33.6 + 59.82 + 45.856) − 4 × 60.0 = − 0.724
6. (60 +17.615 + 22.96 + 33.34) − 4 × 33.61 = − 0.525
7. (33.61+7.69+11.565 + 17.59) − 4 × 17.615 = − 0.005
8. (17.615 + 0 + 5.01+7.57) − 4 × 7.69 = − 0.565
9. (60 + 45.535 + 33.34+100) − 4 × 59.82 = − 0.405
10. (59.82 + 17.59 + 22.72 + 33.61) − 4 × 33.34 = − 0.535
11. (17.615 + 11.3 + 33.34 + 7.57) − 4 × 17.59 = − 0.185
12. (17.59 + 7.69+0+4.79) − 4 × 7.57 = − 0.21
13. (22.06 + 100 + 59.82 + 0) − 4 × 45.535 = − 0.26
14. (33.34 + 0 + 45.535 + 11.3) − 4 × 22.72 = − 0.705
15. (17.59 + 0 + 4.79 + 22.7.2) − 4 × 11.3 = − 0.10
16. (2.87 + 0 + 0 + 11.3) − 4 × 4.79 = − 0.01

113
The final temperature distribution is as shown in Fig. Prob. 3.16.
(viii) The amount of heat supplied is equal to that conducted from the left hand
boundary to each of the internal nodes adjacent to it.
∆y M
q total = k ⋅ (
∑ T1, j − T2, j
∆x j= 2
)

Since in this problem, we have a square ajay, ∆y / ∆x and the expression for q becomes
M
(
q total = k ⋅ ∑ T1, j − T2, j )
j= 2

= k (54.144 + 40 + 40.18 + 54.465) = 188.789k

(ix) The amount of heat removed is the sum of that conducted from all adjacent
internal nodes to those nodes at 0°C. The total rate of heat removal is
determined by the sum of heat transfers from the top, bottom and right
hand surfaces.
q total = k [(45.856 + 22.96 + 11.565 + 5.01) +
(45.535 + 22.72 + 11.3 + 4.79) +
(5.01 + 7.69 + 7.57 + 4.79)] = 194.796 k

This result should equal the value obtained for total heat input. The difference is
due to the coarse grid spacing and the approximations used.

3.17.Sol.
Because the duct cross-section is symmetrical, we need to analyse only a
quarter section. A teledeltos paper model is shown in Fig. Prob.3.17 (b).
The flux plot is shown in Fig. Prob. 3.17 (a). The heat flow rate can be
calculatedfrom the flux plot. The shape factor can be readily determined by noting that the
number of temperature increments M is 8, the number of heat flow channels for the quarter
section is 19 or N = 76. The shape factor is thus
N 76
S= = = 9.5
M 8

and the heat loss thruogh the duct per unit depth becomes
q = k ⋅ s ⋅ ∆T = (1.0 × 10 −2 )(9.5)(120) = 11.4 W/m

The method seems to give reasonable good results for the examples considered.
Although with the advent of digital computers, these analogues have been rendered non-

114
essential, they are still useful in providing a simple application of heat-transfer phenomena
for experimental and teaching purposes. The electrical analogue is of practical value
where the use of conducting paper analogue may be preferred in order to avoid the delay
involved in compiling and testing a computer programme.

Fig. Prob. 3.17.( b)

Fig. Prob. 3.17. (c)

115
3.18.Sol.
The cross-section of the duct is as shown in Fig. Prob. 3.18. Since the duct is
symmetrical, only a quarter section is considered for evaluating the steady state
temperature distribution in the insulation. The temperatures at the boundaries are fixed
as shown in Fig. Prob. 3.18.
The final equations obtained are:

Node No. Sun of Neighbouring Node Temp Node Temp. Residue


1. (100 + 110 + 150 + 67.8) − 4 × 107 = − 0.2
2. (63 + 30 + 107 + 70) − 4 × 67.8 = + 1.25
3. (67.8 + 100 + 30 + 54.5) − 4 × 63 = +0.3
4. (150 + 79.3 + 63 + 107) − 4 × 100 = − 0.7
5. (30 + 79.3 + 63 + 45.43) − 4 × 54.5 = − 0.23
6. (103.83 + 54.5 + 100 + 60.5) − 4 × 79.3 = +1.63
7. (150+ 112.84 + 74 + 79.3) − 4 × 103.83 = +0.5
8. (103.83+ 116.85+ 150 + 81.71) − 4 × 112.8 = +1.19
9. (112.8+ 120 + 86.4+ 1.50) − 4 × 116.85 = +1.8
10. (54.5 + 37.24 + 60.5 + 30) − 4 × 116.85 = +1.8
11. (45.43 + 74 + 79.3 + 44.2) − 4 × 60.5 = +0.5
12. (103.83 + 50.5 + 81.71 + 60.5) − 4 × 74 = +0.54
13. (112.8 + 54.87 + 74 + 86.4) − 4 × 81.71 = +1.19
14. (116.85 + 57.77 + 90 + 81.72) − 4 × 86.43 = +0.62
15. (45.43 + 44.24 + 30 + 30) − 4 × 37.24 = +0.71
16. (50.5 + 60.5 + 30 + 37.24) − 4 × 44.2 = +1.44
17. (74.30 + 44.2 + 54.87 + 30) − 4 × 50.5 = +0.5
18. (57.77 + 50.5 + 81.7 + 30) − 4 × 54.87 = +0.5
19. (86.43 + 33 + 60 + 54.87) − 4 × 57.77 = +0.69

The last values of temperatures calculated can be assumed to be the final


temperatures because all the residue are now less than ±2. The temperature distribution in
the insulation is as shown in Fig. Prob. 3.17.
The heat flow rate can be calculated as follows :
(a) The amount of heat supplied is equal to that conducted from the inner side of the duct to
each of the internal nodes adjacent to it.
qin = 1.0 × 10−2 [(40 + 43 + 50) + (46.47 + 37.2 + 33.15 + 30)]

= 2.372 W/m (for quarter section)

116
(b) The amount of heat removed is sum of that conducted from all internal nodes adjacent to
the outer wall of duct (at 30°C).
q out = 1.0 × 10−2 [(40 + 37.8 + 24.5 + 15.43 + 7.24) +
(7.24) + 14.2 + 20.5 + 24.87 + 27.79 + 30]
= 2.488 W/m (for quarter section)
The total heat input is not precisely equal to total heat output because the residues
were not reduced to zero, resulting in approximate temperatures at the nodes. Also the
grid used was coarse. A finer grid could lead to better results.
The approach described here is intuitive, and is of limited scope. With the advent of
high speed digital computers numerical techniques have been and are being devised to
handle problems of almost any degree of complexity. The results obtained by these
techniques are available only at a few discrete points, but with the help of interpolating
functions and graphical output devices. A complete distribution of temperature or heat flow
can be obtained.
The two numerical approximation methods commonly used are ( i ) finite-
difference method, (ii) finite element method. The finite difference method are relatively
straightforward and are more frequently used. The fundamental operation in this method is
the replacement of each derivative in the governing partial differential equation of the
problem by an algebraic approximation a each discrete point in the region. For example, if a
region of interest is divided into N discrete points (nodes by a suitable mesh arrangement, the
process of approximation by finite differences will give N simultaneous algebraic equations, the
solution of which provides a numerical value of the variable (e .g. temperature) for
each mesh point.
In the finite element approach. «l.e region of interest is divided into a number of elements
(usually triangular or quadrilateral) and the unknown function, viz., temperature, uniquely
specified throughout the field by a discrete number. of values associated with the node
points of the elements. This method, through not as simple and straightforward the finite
difference method, has several advantages in the solution of conduction heat transfer
problem of curved boundaries and complex geometry

3.19.Sol.
As the walls of the furnace are thick , this is a case of three- dimensional
conduction.
Inside surface area = 2 (4 × 3 + 3 × 2 + 2 × 4) = 52 m 2
Length of edges = ∑ y = 4 (4 + 3 + 2 + 3) = 36 m

Thickness ∆x = 0.2 m

117
Mean Area A m = A1 + Sedge∆x + Scorner ⋅ ∆x

= A1 + 0.54 y ⋅ ∑ ∆x + 8 (0.15) ⋅ (∆x ) 2

= 52 + (0.54) (36) (0.2) + 8 (0.15) ⋅ (0.2) 2

= 52 + 3.808 + .0.048 = 55.936 m 2


Hence
k ⋅ A m (T1 − T2 ) (1.6) (55.936) (170)
q= =
∆x (0.2)

= 76073 W = 76.073 kW.

118

You might also like