CHENG 314 Heat Transfer

Dr. Zakir Hossain

E.mail: [email protected]
Office: 15-312
Tel: 1787-6374

Department of Chemical Engineering

University of Bahrain

1
Two Dimensional Heat
Conduction
Chap 4

2
The Heat Equation and Methods of Solution
General Heat conduction:

  T    T    T  T
 k    k    k   q  c p
x  x  y  y  z  z  t

For 2D, SS, without Heat source

For 2D, SS with Heat source

2T 2T q  x, y 
 2  0
x2
y k
3
The Heat Equation and Methods of Solution

• Solution Methods:
– Exact/Analytical: Separation of Variables (Section 4.2)
 Limited to simple geometries and boundary conditions.

 
– Approximate/Graphical q  0 : Flux Plotting (Section 4 S.1)
–  Of limited value for quantitative considerations but a quick aid to
establishing physical insights.

– Approximate/Numerical: Finite-Difference, Finite Element or Boundary

Element Method.
 Most useful approach and adaptable to any level of complexity.
 Readily extended to 3D problems

4
 Review: Introduction to Finite Difference

• Numerical solutions can give answers at only discrete points in the domain,
called grid points.

5
Review: Introduction to Finite Difference

f f  fi
Forward: ( ) i 1 / 2  i 1
x x

f f i  f i 1
Backward: ( ) i 1 / 2 
x x

f f  f i 1
Central: ( )i  i 1
x 2x

Note: f i 1  f ( xi 1 )
f i  f ( xi )
f i 1  f ( xi 1 )
6
 Review: Introduction to Finite Difference

Second derivatives:

2 f f  2 f i  f i 1
* Central difference: ( 2 )i  i 1
x (x) 2

2 f f  2 f i 1  f i
* Forward difference: ( 2 )  i  2
x (x) 2

* Backward difference: 2 f f i  2 f i 1  f i 2
( 2)
x (x) 2

7
The Nodal Network and Finite-Difference Approximation
• The nodal network identifies discrete
points at which the temperature is
to be determined and uses an
m,n notation to designate their
location.

What is represented by the

temperature determined at a nodal
point, as for example, Tm,n?

Each node represents a certain region, and its temperature is a

measure of average temperature of the region. Here temperature
for node (m, n) may be viewed as the average temperature of
surrounding shaded area.

8
The Nodal Network and Finite-Difference Approximation

• A finite-difference approximation
is used to represent temperature
gradients in the domain.

How is the accuracy of the solution affected by construction of the nodal

network? What are the significance of selection of a fine or a coarse mesh?
9
Finite Difference Form of the Heat Equation
Temperature gradient can be written as

10
Finite Difference Form of the Heat Equation
Now

Tm1,n  Tm1,n  Tm,n 1  Tm,n1

Tm,n 
4

Interpretation

The temperature of each interior node is the arithmetic

average of the temperature of the four neighboring
nodes. 11
The Energy balance Method-alternative method
Consider a volume element of size Δx ×
Δy ×1 centered about a general interior
node (m,n) in a region with heat
generation q and thermal conductivity
k

Since the actual direction of Heat flow (into or out of the node) is often
unknown , it is convenient to formulate energy balance by assuming that
all the heat flow is into the node. Such a condition is of course impossible
but if the rate equations are expressed in a manner consistent with this
assumption , the correct form of the finite difference equation is obtained.

12
The Energy balance Method-alternative method

E in  q  0

Rate of heat Rate of heat Rate of change

conduction at the generation of the energy
left, top, right, and + inside the = content of the
bottom surfaces element element

. Eelement
qcond,left  qcond,top  qcond,right  qcond,bottom  q  0
element t

13
 Tm1,n  Tm,n Tm,n1  Tm,n Tm1,n  Tm,n Tm,n1  Tm,n .
ky  kx  ky  kx  qm,n xy  0
x y x y

Dividing each term by and simplifying give

.
Tm1,n  2Tm,n  Tm1,n Tm,n 1  2Tm,n  Tm,n 1 q m,n
   0
x 2
y 2
k 14
In finite difference analysis, usually a square mesh is used for
simplicity and thus Δx and Δy are taken to be the same.
Then Δx = Δy and the relation above simplifies to
.
q m,n (x) 2
Tm1,n  Tm1,n  Tm,n 1  Tm,n 1  4Tm,n   0
k

When there is no heat generation in the medium, the

finite differences equation for an interior node further
simplified to

T T  T T
T node
 left top right bottom

4
15
16
17
18
 Example 1
Consider steady heat transfer in an L-shaped solid body whose cross section is given
in Fig. 1. heat transfer in the direction normal to the plane of the paper is negligible
and thus heat transfer in the body is two-dimensional. The thermal conductivity of the
body is 15 W/m.K, and heat is generated in the body at a rate of q = 2 ×106 W/m3. the
left surface of the body is insulated and the bottom surface is maintained at a uniform
temperature of 90oC. The entire top surface is subjected to convection to ambient air
at T = 25oC with a convection coefficient h = 80 W/m2Kand the right surface is
subjected to heat flux at a uniform rate of qR = 5000 W/m2. the nodal network of the
problem consists of 15 equally spaced nodes with Δx = Δy = 1.2 cm, as shown in the
figure. Six of the nodes are at the bottom surface, and thus their temperatures are
known. Obtain the finite difference equations at the remaining nine nodes and
determine the nodal temperatures by solving them.

Fig. 1

19
Node 1
.
x y (T2  T1 ) x (T4  T1 ) x y
h, T∞ 0h (T  T1 )  k k  e1  0
2 2 x 2 y 2 2
1 2

-2.064 T1 + T2 +T4 = -11.2

4
h, T∞
1 2

Node 2
4 5

.
y ( T3  T2 ) ( T T ) y ( T1  T2 ) y
hx ( T  T2 )  k  kx 5 2  k  e2 x  0
2 x y 2 x 2

T1 – 4.128 T2 + T3 + 2T5 = -22.4

20
 .
Node 3 x y x ( T6  T3 ) y ( T2  T3 ) x y
h(  )( T  T3 )  k k  e3  0
2 2 2 y 2 x 2 2
h, T∞
2 3
T2 – 2.128 T2 + T6 = -12.8
h, T∞

This node is on insulated boundary and can be treated as

Node 4
interior node by replacing the insulation by a mirror. This
put a reflected image of node 5. The general interior node
1 relation for the steady two dimentional case I gives
mirror

5 5 .
4 y
T5  T1  T5  T10  4 T4` e4 x  0
k

10
T1 – 4T4 + 2T5 = -109.2 21
 Node 5
y
2 T4  T2  T6  T11  4 T5  e5 x  0
k
4 5 6
T2 + T4 - 4T5 + T6 = -109.2

11

Node 6
x y y ( T7  T6 ) ( T T )
h(  )( T  T6 )  k  kx 12 6
2 2 2 x y
3
( T5  T6 ) x ( T3  T6 ) 3 xy
 ky k  e6  0
6 7 x 2 y 4
5

12
T3 + 2 T5 – 6.128 T6 + T7 = -212 22
 Node 7
y ( T8  T7 ) ( T  T7 )
h x ( T  T7 )  k  kx 13
2 x y
6 7 8
y ( T6  T7 ) y
k  e7 x  0
2 x 2
13
T6 – 4.128T7 + T8 = -202.4

Node 8 This node is identical to node 7, and

T7 – 4.128T8 + T9 = -202.4
Node 9
h, T∞
8 9 x y x ( T15  T9 ) y ( T8  T9 ) x y
h( )( T  T9 )  qR k k  e9  0
qR 2 2 2 y 2 x 2 2

15
T8– 2.064 T9 = -105.2
23
 -2.064 T1 + T2 +T4 = -11.2
T1 – 4.128 T2 + T3 + 2T5 = -22.4
T2 – 2.128 T2 + T6 = -12.8
T1 – 4T4 + 2T5 = -109.2
T2 + T4 - 4T5 + T6 = -109.2
T3 + 2 T5 – 6.128 T6 + T7 = -212
T6 – 4.128T7 + T8 = -202.4
T7 – 4.128T8 + T9 = -202.4
T8– 2.064 T9 = -105.2

Solving these equations we get [ 1) matrix inversion or 2) gauss seidel

iteration]

T1 = 112.1oC T2 = 110.8oC T3 = 106.6oC

T4 = 109.4oC T5 = 108.1oC T6= 108.1oC
T7 = 97.3oC T8= 96.3oC T9 = 97.6oC 24
 Example 2

• For symmetryT1 =T2 and T3 =T4

• k = 10 W/m.K

a) Find all nodal temperature as shown in Figure

b) Estimate heat flows at the boundaries

25
4.3 The Conduction Shape Factor
In many instances, two- or three-dimensional conduction problems (where
only two T limits are involved) may be rapidly solved by utilizing existing
solutions to the heat diffusion equation. These solutions are reported in
terms of a shape factor S.

S: conduction shape factor

k: the thermal conductivity of the medium between the surfaces

•The conduction shape factor depends on the geometry of the system

only. Conduction shape factors are applicable only when heat transfer
between the two surfaces is by conduction.

Relationship between the conduction

shape factor and the thermal resistance
26
Shape factors may also be defined for one-dimensional
geometries:

Plane wall: S= A/L

Cylindrical wall: S=2πL/ln(r2/r1)

Spherical wall: S= 4πr1r2/(r2- r1).

27
28
29
 Example 3
An alberta oil pipe line is buried in the earth at a depth of 1 m. The
horizontal pipe is a thin-walled of outside diameter of 50 cm. The pipe
is 1 m long and the averaged temperature of the oil is 100C and the
ground soil temperature is at -20 C (ksoil=0.5 W/m.K).

a) Estimate the heat loss per unit length of pipe.

b) If the mass flow rate of the oil is 2 kg/s and the specific heat of the
oil is 2 kJ/kg.K, determine the temperature change in 1 km of pipe
length.
T2
c) Plot T distribution for a long pipe
(assume L=4.74 km)
z=1 m

T1 30
 Example 3

T2
a) From Table 4.1, case 1.
L>>D, z>3D/2

z=1 m Assumptions:
• steady state
• deep ground T same as surface T
• negligible resistance of the pipe wall and for
convection from oil
T1 • isothermal surface exposed to soil

2L 2 (1)
S   3.02
ln( 4 z / D) ln( 4 / 0.5)
q  kS(T1  T2 )  (0.5)(3.02)(100  20)
 181.2(W ) heat loss for every meter of pipe

31
 Example 3 (cont.)
b) Temperature change in 1 m of pipe length.

q 181.2
q  mC
 P T , T    0.045( C)
 P 2000 * 2
mC

So, ΔT=0.045x1000=45C for every 1 km of pipe length.

Therefore, the total temperature variation can be significant if the pipe is

very long.

• Heating might be needed to prevent the oil from freezing up.

• The heat transfer can not be considered constant for a long pipe

32
 Example 3 (cont.)
c)

Ground at -20C

Heat transfer to the ground (q)

m C pT m C p (T  dT )
Length dx

33
 Example 3 (cont.)
Heat Transfer at section with a temperature T(x)
2k(dx)
q= (T  20)  1.51(T  20)( dx )
ln(4z / D)
Energy balance: mC P T  q  mC
 P (T  dT )
dT dT
 P
mC  1.51(T  20)  0,  0.000378dx, integrate
dx T  20
T ( x )  20  Ce 0.000378 x , at inlet x = 0, T(0) = 100 C, C = 120
T(x) = -20 + 120e 0.000378 x
100
• Temperature drops exponentially
from the initial temp. of 100C
50
T( x ) • It reaches 0C at x=4740 m,
therefore, reheating is required for
0
every 4.74km.

50
0 1000 2000 3000 40005000 34
x
Acknowledgement

35
Algor: How many elements?

Elements: 9 Time: 6s
Nodes: 16 Memory: 0.239MB
Algor: How many elements?

Elements: 16 Time: 6s
Nodes: 25 Memory: 0.255MB
Algor: How many elements?

Elements: 49 Time: 7s
Nodes: 64 Memory: 0.326MB
Algor: How many elements?

Elements: 100 Time: 7s

Nodes: 121 Memory: 0.438MB
Algor: How many elements?

Elements: 324 Time: 7s

Nodes: 361 Memory: 0.910MB
Algor: How many elements?

Elements: 625 Time: 9s

Nodes: 676 Memory: 1.535MB
Algor: How many elements?

Elements: 3600 Time: 15s

Nodes: 3721 Memory: 7.684MB
 Algor: How many elements?

Automatic Mesh

Elements: 334 Time: 7s

Nodes: 371 Memory: 0.930MB
Algor Results Options
 Algor: How many elements?
Smaller Elements Fewer Elements

• Higher accuracy • Faster

• More time, memory • Less storage space
46
47
Once the value of the shape factor is known for a specific geometry, the
total steady heat transfer rate can be determined from the following
equation using the specified two constant temperatures of the two
surfaces and the thermal conductivity of the medium between them.

48