hol29362_ch02 10/13/2008 18:58

C H A P T E R

2 Steady-State Conduction—
One Dimension

2-1 INTRODUCTION
We now wish to examine the applications of Fourier’s law of heat conduction to calculation
of heat flow in some simple one-dimensional systems. Several different physical shapes
may fall in the category of one-dimensional systems: cylindrical and spherical systems are
one-dimensional when the temperature in the body is a function only of radial distance
and is independent of azimuth angle or axial distance. In some two-dimensional problems
the effect of a second-space coordinate may be so small as to justify its neglect, and the
multidimensional heat-flow problem may be approximated with a one-dimensional analysis.
In these cases the differential equations are simplified, and we are led to a much easier
solution as a result of this simplification.

2-2 THE PLANE WALL

First consider the plane wall where a direct application of Fourier’s law [Equation (1-1)]
may be made. Integration yields

kA
q=− (T2 − T1 ) [2-1]
x
when the thermal conductivity is considered constant. The wall thickness is x, and T1
and T2 are the wall-face temperatures. If the thermal conductivity varies with temperature
according to some linear relation k = k0 (1 + βT ), the resultant equation for the heat flow
is
 
k0 A β 2
q=− (T2 − T1 ) + (T2 − T1 ) 2
[2-2]
x 2
If more than one material is present, as in the multilayer wall shown in Figure 2-1, the
analysis would proceed as follows: The temperature gradients in the three materials are
shown, and the heat flow may be written
T2 − T1 T3 − T2 T4 − T3
q = −kA A = −kB A = −kC A
xA xB xC
Note that the heat flow must be the same through all sections.

27

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.27 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

28 2-3 Insulation and R Values

Figure 2-1 One-dimensional heat transfer through a composite wall and electrical analog.

A
Temperature q
profile
q q RA RB RC

T1 ΔxA T2 ΔxB T3 ΔxC T4

A B C
kAA kBA kCA

1 2 3 4

Solving these three equations simultaneously, the heat flow is written

T1 − T4
q= [2-3]
xA /kA A + xB /kB A + xC /kC A

At this point we retrace our development slightly to introduce a different conceptual view-
point for Fourier’s law. The heat-transfer rate may be considered as a flow, and the combina-
tion of thermal conductivity, thickness of material, and area as a resistance to this flow. The
temperature is the potential, or driving, function for the heat flow, and the Fourier equation
may be written

thermal potential difference

Heat flow = [2-4]
thermal resistance
a relation quite like Ohm’s law in electric-circuit theory. In Equation (2-1) the thermal
resistance is x/kA, and in Equation (2-3) it is the sum of the three terms in the denominator.
We should expect this situation in Equation (2-3) because the three walls side by side act as
three thermal resistances in series. The equivalent electric circuit is shown in Figure 2-1b.
The electrical analogy may be used to solve more complex problems involving both
series and parallel thermal resistances. A typical problem and its analogous electric circuit
are shown in Figure 2-2. The one-dimensional heat-flow equation for this type of problem
may be written
Toverall
q=  [2-5]
Rth

where the Rth are the thermal resistances of the various materials. The units for the thermal
resistance are ◦ C/W or ◦ F · h/Btu.
It is well to mention that in some systems, like that in Figure 2-2, two-dimensional
heat flow may result if the thermal conductivities of materials B, C, and D differ by an
appreciable amount. In these cases other techniques must be employed to effect a solution.

2-3 INSULATION AND R VALUES

In Chapter 1 we noted that the thermal conductivities for a number of insulating materials are
given in Appendix A. In classifying the performance of insulation, it is a common practice

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.28 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER 2 Steady-State Conduction—One Dimension 29

Figure 2-2 Series and parallel one-dimensional heat transfer through a

composite wall and electrical analog.

B
F

A C E

G
D

1 2 3 4 5

RB RF
q
RC

T1 RA RE
RD RG

T2 T3 T4 T5

in the building industry to use a term called the R value, which is defined as
T
R= [2-6]
q/A

The units for R are ◦ C · m2/W or ◦ F · ft 2 · h/Btu. Note that this differs from the thermal-
resistance concept discussed above in that a heat flow per unit area is used.
At this point it is worthwhile to classify insulation materials in terms of their application
and allowable temperature ranges. Table 2-1 furnishes such information and may be used
as a guide for the selection of insulating materials.

2-4 RADIAL SYSTEMS

Cylinders
Consider a long cylinder of inside radius ri , outside radius ro , and length L, such as the
one shown in Figure 2-3. We expose this cylinder to a temperature differential Ti − To and
ask what the heat flow will be. For a cylinder with length very large compared to diameter,
it may be assumed that the heat flows only in a radial direction, so that the only space
coordinate needed to specify the system is r. Again, Fourier’s law is used by inserting the
proper area relation. The area for heat flow in the cylindrical system is
Ar = 2πrL

so that Fourier’s law is written

dT
qr = −kAr [2-7]
dr
or
dT
qr = −2πkrL
dr

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.29 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

30 2-4 Radial Systems

Table 2-1 Insulation types and applications.

Thermal
Temperature conductivity, Density,
Type range, ◦ C mW/m · ◦ C kg/m3 Application
1 Linde evacuated superinsulation −240–1100 0.0015–0.72 Variable Many
2 Urethane foam −180–150 16–20 25–48 Hot and cold pipes
3 Urethane foam −170–110 16–20 32 Tanks
4 Cellular glass blocks −200–200 29–108 110–150 Tanks and pipes
5 Fiberglass blanket for wrapping −80–290 22–78 10–50 Pipe and pipe fittings
6 Fiberglass blankets −170–230 25–86 10–50 Tanks and equipment
7 Fiberglass preformed shapes −50–230 32–55 10–50 Piping
8 Elastomeric sheets −40–100 36–39 70–100 Tanks
9 Fiberglass mats 60–370 30–55 10–50 Pipe and pipe fittings
10 Elastomeric preformed shapes −40–100 36–39 70–100 Pipe and fittings
11 Fiberglass with vapor −5–70 29–45 10–32 Refrigeration lines
barrier blanket
12 Fiberglass without vapor to 250 29–45 24–48 Hot piping
barrier jacket
13 Fiberglass boards 20–450 33–52 25–100 Boilers, tanks,
heat exchangers
14 Cellular glass blocks and boards 20–500 29–108 110–150 Hot piping
15 Urethane foam blocks and 100–150 16–20 25–65 Piping
boards
16 Mineral fiber preformed shapes to 650 35–91 125–160 Hot piping
17 Mineral fiber blankets to 750 37–81 125 Hot piping
18 Mineral wool blocks 450–1000 52–130 175–290 Hot piping
19 Calcium silicate blocks, boards 230–1000 32–85 100–160 Hot piping, boilers,
chimney linings
20 Mineral fiber blocks to 1100 52–130 210 Boilers and tanks

Figure 2-3 One-dimensional heat flow

through a hollow cylinder
and electrical analog.

L
ro r dr
ri

q
Ti To

ln(ro/ri )
R th =
2 π kL

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.30 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER 2 Steady-State Conduction—One Dimension 31

Figure 2-4 One-dimensional heat flow through multiple cylindrical sections

and electrical analog.

r1
r2
T2 r3
T3
T1 q
r4
A T4 T1 RA T2 RB T3 RC T4
B
ln(r2冫r1) ln(r3冫r2) ln(r4冫r3)
C
2πkAL 2π kBL 2π kCL

with the boundary conditions

T = Ti at r = ri
T = To at r = ro

The solution to Equation (2-7) is

2πkL (Ti − To )
q= [2-8]
ln (ro /ri )
and the thermal resistance in this case is
ln (ro /ri )
Rth =
2πkL
The thermal-resistance concept may be used for multiple-layer cylindrical walls just as it
was used for plane walls. For the three-layer system shown in Figure 2-4 the solution is
2πL (T1 − T4 )
q= [2-9]
ln (r2 /r1 )/kA + ln (r3 /r2 )/kB + ln (r4 /r3 )/kC
The thermal circuit is also shown in Figure 2-4.
Spheres
Spherical systems may also be treated as one-dimensional when the temperature is a function
of radius only. The heat flow is then
4πk (Ti − To )
q= [2-10]
1/ri − 1/ro
The derivation of Equation (2-10) is left as an exercise.
Multilayer Conduction EXAMPLE 2-1
An exterior wall of a house may be approximated by a 4-in layer of common brick [k =
0.7 W/m · ◦ C] followed by a 1.5-in layer of gypsum plaster [k = 0.48 W/m · ◦ C]. What thick-
ness of loosely packed rock-wool insulation [k = 0.065 W/m · ◦ C] should be added to reduce the
heat loss (or gain) through the wall by 80 percent?

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.31 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

32 2-4 Radial Systems

Solution
The overall heat loss will be given by
T
q= 
Rth

Because the heat loss with the rock-wool insulation will be only 20 percent (80 percent reduction)
of that before insulation

q with insulation R without insulation
= 0.2 =  th
q without insulation Rth with insulation

We have for the brick and plaster, for unit area,

x (4)(0.0254)
Rb = = = 0.145 m2 · ◦ C/W
k 0.7
x (1.5)(0.0254)
Rp = = = 0.079 m2 · ◦ C/W
k 0.48
so that the thermal resistance without insulation is
R = 0.145 + 0.079 = 0.224 m2 · ◦ C/W

Then
0.224
R with insulation = = 1.122 m2 · ◦ C/W
0.2
and this represents the sum of our previous value and the resistance for the rock wool

1.122 = 0.224 + Rrw

x x
Rrw = 0.898 = =
k 0.065
so that
xrw = 0.0584 m = 2.30 in

EXAMPLE 2-2 Multilayer Cylindrical System

Figure Example 2-2
A thick-walled tube of stainless steel [18% Cr, 8% Ni, k = 19 W/m · ◦ C] with 2-cm inner diam-
Stainless steel
T1 = 600˚C
eter (ID) and 4-cm outer diameter (OD) is covered with a 3-cm layer of asbestos insulation
[k = 0.2 W/m · ◦ C]. If the inside wall temperature of the pipe is maintained at 600◦ C, calculate
the heat loss per meter of length. Also calculate the tube–insulation interface temperature.
r1 r2
Solution
r3 Figure Example 2-2 shows the thermal network for this problem. The heat flow is given by
Asbestos q 2π (T1 − T2 ) 2π (600 − 100)
= = = 680 W/m
T2 = 100˚C L ln (r2 /r1 )/ks + ln(r3 /r2 )/ka (ln 2)/19 + (ln 5 )/0.2
2
T1 T2 This heat flow may be used to calculate the interface temperature between the outside tube wall
and the insulation. We have
ln (r2冫r1) ln (r3冫r2)
2π ksL 2π kaL q Ta − T2
= = 680 W/m
L ln (r3 /r2 )/2πka

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.32 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER 2 Steady-State Conduction—One Dimension 33

where Ta is the interface temperature, which may be obtained as

Ta = 595.8◦ C

The largest thermal resistance clearly results from the insulation, and thus the major portion of the
temperature drop is through that material.

Convection Boundary Conditions

We have already seen in Chapter 1 that convection heat transfer can be calculated from

qconv = hA (Tw − T∞ )

An electric-resistance analogy can also be drawn for the convection process by rewriting
the equation as

Tw − T∞
qconv = [2-11]
1/hA
where the 1/hA term now becomes the convection resistance.

2-5 THE OVERALL HEAT-TRANSFER

COEFFICIENT
Consider the plane wall shown in Figure 2-5 exposed to a hot fluid A on one side and a
cooler fluid B on the other side. The heat transfer is expressed by
kA
q = h1 A (TA − T1 ) = (T1 − T2 ) = h2 A (T2 − TB )
x
The heat-transfer process may be represented by the resistance network in Figure 2-5, and
the overall heat transfer is calculated as the ratio of the overall temperature difference to
the sum of the thermal resistances:

TA − TB
q= [2-12]
1/h1 A + x/kA + 1/h2 A

Figure 2-5 Overall heat transfer through a plane wall.

Fluid A

TA

q
TA T1 T2 TB
T1
q
Δx
Fluid B

h1 T2 1 1
h1A kA h2A
h2

TB

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.33 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

34 2-5 The Overall Heat-Transfer Coefficient

Figure 2-6 Resistance analogy for hollow cylinder with convection boundaries.

Fluid B
q

Fluid A TA Ti To TB

1 1
hiAi hoAo
1 2

Observe that the value 1/ hA is used to represent the convection resistance. The overall
heat transfer by combined conduction and convection is frequently expressed in terms of
an overall heat-transfer coefficient U, defined by the relation
q = UA Toverall [2-13]

where A is some suitable area for the heat flow. In accordance with Equation (2-12), the
overall heat-transfer coefficient would be
1
U=
1/h1 + x/k + 1/h2

The overall heat-transfer coefficient is also related to the R value of Equation (2-6) through
1
U=
R value
For a hollow cylinder exposed to a convection environment on its inner and outer surfaces,
the electric-resistance analogy would appear as in Figure 2-6 where, again, TA and TB are
the two fluid temperatures. Note that the area for convection is not the same for both fluids
in this case, these areas depending on the inside tube diameter and wall thickness. The
overall heat transfer would be expressed by
TA − TB
q= [2-14]
1 ln (ro /ri ) 1
+ +
hi A i 2πkL ho A o
in accordance with the thermal network shown in Figure 2-6. The terms Ai and Ao represent
the inside and outside surface areas of the inner tube. The overall heat-transfer coefficient
may be based on either the inside or the outside area of the tube. Accordingly,
1
Ui = [2-15]
1 Ai ln (ro /ri ) Ai 1
+ +
hi 2πkL A o ho
1
Uo = [2-16]
Ao 1 Ao ln (ro /ri ) 1
+ +
A i hi 2πkL ho
The general notion, for either the plane wall or cylindrical coordinate system, is that
UA = 1/ Rth = 1/Rth,overall

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.34 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch02 10/29/2008 14:15

CHAPTER 2 Steady-State Conduction—One Dimension 35

Calculations of the convection heat-transfer coefficients for use in the overall heat-transfer
coefficient are made in accordance with the methods described in later chapters. Some typi-
cal values of the overall heat-transfer coefficient for heat exchangers are given in Table 10-1.
Some values of U for common types of building construction system are given in Table 2-2
and may be employed for calculations involving the heating and cooling of buildings.

Table 2-2 Overall heat transfer coefficients for common construction systems according to James
and Goss [12].

Description of construction system U, Btu/hr · ft2 · ◦ F U, W/m2 · ◦ C

1 2 × 3 in double-wood stud wall, 406 mm OC, polyisocyanurate 0.027 0.153

(0.08-mm vapor retarder, 19-mm insulation), fiberglass batts
in cavity, 12.7-mm plywood
2 2 × 4 in wood stud wall, 406 mm OC, polyisocyanurate 0.060 0.359
foil-faced, fiberglass batts in cavity, 15-mm plywood
3 2 × 4 in wood stud wall, 406 mm OC, 38-mm polyisocyanurate, 0.039 0.221
foil-faced, cellular polyurethane in cavity, 19-mm plywood
4 2 × 4 in wood stud wall, 406 mm OC, 15-mm exterior sheathing, 0.326 1.85
0.05-mm polyethylene vapor barrier, no fill in cavity
5 Nominal 4-in concrete-block wall with brick facade and 0.080 0.456
extruded polystyrene insulation
6 2 × 4 in wood stud wall, 406 mm OC, fiberglass batt insulation 0.084 0.477
in cavity, 16-mm plywood
7 2 × 4 in wood stud wall, 406 mm OC, fiberglass batt insulation 0.060 0.341
in cavity, 16-mm plywood, clay brick veneer
8 2 × 4 in wood stud wall, 406 mm OC, fiberglass batt in cavity, 0.074 0.417
13-mm plywood, aluminum or vinyl siding
9 2 × 4 in wood stud wall, 406 mm OC, polyurethane foam 0.040 0.228
in cavity, extruded polystyrene sheathing, aluminum siding
10 2 × 4 in steel stud wall, 406 mm OC, fiberglass batts 0.122 0.691
in cavity, 41-mm air space, 13-mm plaster board
11 Aluminum motor home roof with fiberglass insulation 0.072 0.41
in cavity (32 mm)
12 2 × 6 in wood stud ceiling, 406 mm OC, fiberglass 0.065 0.369
foil-faced insulation in cavity, reflective airspace (ε ≈ 0.05)
13 8-in (203-mm) normal-weight structural concrete (ρ = 2270 kg/m3 ) 0.144 0.817
wall, 18-mm board insulation, painted off-white
14 10-in (254-mm) concrete-block-brick cavity wall, 0.322 1.83
no insulation in cavities
15 8-in (203-mm) medium-weight concrete block wall, 0.229 1.3
perlite insulation in cores
16 8-in (203-mm) normal-weight structural concrete, 0.764 4.34
(ρ = 2270 kg/m3 ) including steel reinforcement bars
(Note: actual thickness of concrete is 211 mm.)
17 8-in (203-mm) lightweight structural concrete (ρ = 1570 kg/m3 ) 0.483 2.75
including steel reinforcement bars
(Note: Actual thickness of concrete is 210 mm.)
18 8-in (203-mm) low-density concrete wall (ρ = 670 kg/m3 ) 0.216 1.23
including steel reinforcement bars
(Note: Actual thickness of concrete is 216 mm.)
19 Corrugated sheet steel wall with 10.2-in (260-mm.) 0.030 0.17
fiberglass batt in cavity
20 Corrugated sheet steel wall with (159-mm) fiberglass batt in cavity 0.054 0.31
21 Metal building roof deck, 25 mm polyisocyanurate, foil-faced 0.094 0.535
(ε ≈ 0.03), 203-mm reflective air space
22 Metal building roof deck, 25-mm foil-faced polyisocyanurate, 0.065 0.366
38-mm fiberglass batts in cavity

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.35 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

36 2-5 The Overall Heat-Transfer Coefficient

EXAMPLE 2-3 Heat Transfer Through a Composite Wall

“Two-by-four” wood studs have actual dimensions of 4.13 × 9.21 cm and a thermal conductivity
of 0.1 W/m · ◦ C. A typical wall for a house is constructed as shown Figure Example 2-3. Calculate
the overall heat-transfer coefficient and R value of the wall.

Figure Example 2-3 (a) Construction of a dwelling wall; (b) thermal resistance
model.

Outside air convection, h = 15 W/m2 • ˚C

Common brick, k = 0.69 8 cm

1.9 cm, k = 0.96

Gypsum
sheath
1.9 cm, k = 0.48
40.6 cm
Insulation, k = 0.04
2 x 4 studs
Inside air convection, h = 7.5 W/m2 • ˚C
(a)
R sheath R insul R sheath
outside inside

Tair Tair
outside inside

R convection R brick R convection

outside inside

R sheath R stud R sheath

outside inside

(b)

Solution
The wall section may be considered as having two parallel heat-flow paths: (1) through the studs,
and (2) through the insulation. We will compute the thermal resistance for each, and then combine
the values to obtain the overall heat-transfer coefficient.
1. Heat transfer through studs (A = 0.0413 m2 for unit depth). This heat flow occurs through six
thermal resistances:
a. Convection resistance outside of brick
1 1
R= = = 1.614 ◦ C/W
hA (15)(0.0413)

b. Conduction resistance in brick

0.08
R = x/kA = = 2.807 ◦ C/W
(0.69)(0.0413)

c. Conduction resistance through outer sheet

x 0.019
R= = = 0.48 ◦ C/W
kA (0.96)(0.0413)

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.36 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER 2 Steady-State Conduction—One Dimension 37

d. Conduction resistance through wood stud

x 0.0921
R= = = 22.3 ◦ C/W
kA (0.1)(0.0413)

e. Conduction resistance through inner sheet

x 0.019
R= = = 0.96 ◦ C/W
kA (0.48)(0.0413)

f. Convection resistance on inside

1 1
R= = = 3.23 ◦ C/W
hA (7.5)(0.0413)

The total thermal resistance through the wood stud section is

Rtotal = 1.614 + 2.807 + 0.48 + 22.3 + 0.96 + 3.23 = 31.39 ◦ C/W [a]
2. Insulation section (A = 0.406 − 0.0413 m2 for unit depth). Through the insulation sec-
tion, five of the materials are the same, but the resistances involve different area terms,
i.e., 40.6 − 4.13 cm instead of 4.13 cm, so that each of the previous resistances must be mul-
tiplied by a factor of 4.13/(40.6 − 4.13) = 0.113. The resistance through the insulation is
x 0.0921
R= = = 6.31
kA (0.04)(0.406 − 0.0413)

and the total resistance through the insulation section is

Rtotal = (1.614 + 2.807 + 0.48 + 0.96 + 3.23)(0.113) + 6.31 = 7.337 ◦ C/W [b]
The overall resistance for the section is now obtained by combining the parallel resistances in
Equations (a) and (b) to give

1
Roverall = = 5.947 ◦ C/W [c]
(1/31.39) + (1/7.337)
This value is related to the overall heat-transfer coefficient by
T
q = UAT = [d]
Roverall

where A is the area of the total section = 0.406 m2 . Thus,

1 1
U= = = 0.414 W/m2 · ◦ C
RA (5.947)(0.406)

As we have seen, the R value is somewhat different from thermal resistance and is given by
1 1
R value = = = 2.414◦ C · m2/W
U 0.414
Comment
This example illustrates the relationships between the concepts of thermal resistance, the overall
heat-transfer coefficient, and the R value. Note that the R value involves a unit area concept, while
the thermal resistance does not.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.37 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

38 2-5 The Overall Heat-Transfer Coefficient

EXAMPLE 2-4 Cooling Cost Savings with Extra Insulation

A small metal building is to be constructed of corrugated steel sheet walls with a total wall surface
area of about 300 m2 . The air conditioner consumes about 1 kW of electricity for every 4 kW
of cooling supplied.1 Two wall constructions are to be compared on the basis of cooling costs.
Assume that electricity costs $0.15/kWh. Determine the electrical energy savings of using 260 mm
of fiberglass batt insulation instead of 159 mm of fiberglass insulation in the wall.Assume an overall
temperature difference across the wall of 20◦ C on a hot summer day in Texas.
Solution
Consulting Table 2-2 (Numbers 19 and 20) we find that overall heat transfer coefficients for the
two selected wall constructions are
U(260-mm fiberglass) = 0.17 W/m2 · ◦ C

U (159-mm fiberglass) = 0.31 W/m2 · ◦ C

The heat gain is calculated from q = UAT , so for the two constructions
q (260-mm fiberglass) = (0.17)(300)(20) = 1020 W

q (159-mm fiberglass) = (0.31)(300)(20) = 1860 W

Savings due to extra insulation = 840 W

The energy consumed to supply this extra cooling is therefore

Extra electric power required = (840)(1/4) = 210 W

and the cost is

Cost = (0.210kW)(0.15$/kWh) = 0.0315 $/hr

Assuming 10-h/day operation for 23 days/month this cost becomes

(0.0315)(10)(23) = $7.25/month

Both of these cases are rather well insulated. If one makes a comparison to a 2 × 4 wood stud wall
with no insulation (Number 4 in Table 2-2) fill in the cavity (U = 1.85 W/m2 · ◦ C), the heating
load would be
q = (1.85)(300)(20) = 11,100 W

and the savings compared with the 260-mm fiberglass insulation would be
11,100 − 1020 = 10,080 W

producing a corresponding electric power saving of $0.378/h or $86.94/month. Clearly the insu-
lated wall will pay for itself. It is a matter of conjecture whether the 260-mm of insulation will
pay for itself in comparison to the 159-mm insulation.

1
This is not getting something for nothing. Consult any standard thermodynamics text for the reason for this
behavior.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.38 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER 2 Steady-State Conduction—One Dimension 39

Overall Heat-Transfer Coefficient for a Tube EXAMPLE 2-5

Water flows at 50◦ C inside a 2.5-cm-inside-diameter tube such that hi = 3500 W/m2 · ◦ C. The
tube has a wall thickness of 0.8 mm with a thermal conductivity of 16 W/m · ◦ C. The outside of
the tube loses heat by free convection with ho = 7.6 W/m2 · ◦ C. Calculate the overall heat-transfer
coefficient and heat loss per unit length to surrounding air at 20◦ C.
Solution
There are three resistances in series for this problem, as illustrated in Equation (2-14). With
L = 1.0 m, di = 0.025 m, and do = 0.025 + (2)(0.0008) = 0.0266 m, the resistances may be cal-
culated as
1 1
Ri = = = 0.00364 ◦ C/W
hi Ai (3500)π(0.025)(1.0)
ln (do /di )
Rt =
2πkL
ln(0.0266/0.025)
= = 0.00062 ◦ C/W
2π(16)(1.0)

1 1
Ro = = = 1.575 ◦ C/W
ho Ao (7.6)π(0.0266)(1.0)
Clearly, the outside convection resistance is the largest, and overwhelmingly so. This means that it
is the controlling resistance for the total heat transfer because the other resistances (in series) are
negligible in comparison. We shall base the overall heat-transfer coefficient on the outside tube
area and write
T
q =  = UAo T [a]
R

1 1
Uo =  =
Ao R [π(0.0266)(1.0)](0.00364 + 0.00062 + 1.575)
= 7.577 W/m2 · ◦ C

or a value very close to the value of ho = 7.6 for the outside convection coefficient. The heat
transfer is obtained from Equation (a), with

q = UAo T = (7.577)π(0.0266)(1.0)(50 − 20) = 19 W (for 1.0 m length)

Comment
This example illustrates the important point that many practical heat-transfer problems involve
multiple modes of heat transfer acting in combination; in this case, as a series of thermal resis-
tances. It is not unusual for one mode of heat transfer to dominate the overall problem. In this
example, the total heat transfer could have been computed very nearly by just calculating the free
convection heat loss from the outside of the tube maintained at a temperature of 50◦ C. Because
the inside convection and tube wall resistances are so small, there are correspondingly small tem-
perature drops, and the outside temperature of the tube will be very nearly that of the liquid inside,
or 50◦ C.

2-6 CRITICAL THICKNESS OF INSULATION

Let us consider a layer of insulation which might be installed around a circular pipe, as
shown in Figure 2-7. The inner temperature of the insulation is fixed at Ti , and the outer

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.39 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

40 2-6 Critical Thickness of Insulation

Figure 2-7 Critical insulation thickness.

h, T
ri
+ ro
Ti Ti T

ln (ro冫ri) 1
2π kL 2π roLh

surface is exposed to a convection environment at T∞ . From the thermal network the heat
transfer is
2πL (Ti − T∞ )
q= [2-17]
ln (ro /ri ) 1
+
k ro h
Now let us manipulate this expression to determine the outer radius of insulation ro , which
will maximize the heat transfer. The maximization condition is
 
1 1
−2πL (Ti − T∞ ) − 2
dq kro hro
=0=  
dro ln (ro /ri ) 1 2
+
k ro h

which gives the result

k
ro = [2-18]
h
Equation (2-18) expresses the critical-radius-of-insulation concept. If the outer radius is less
than the value given by this equation, then the heat transfer will be increased by adding more
insulation. For outer radii greater than the critical value an increase in insulation thickness
will cause a decrease in heat transfer. The central concept is that for sufficiently small values
of h the convection heat loss may actually increase with the addition of insulation because
of increased surface area.

EXAMPLE 2-6 Critical Insulation Thickness

Calculate the critical radius of insulation for asbestos [k = 0.17 W/m · ◦ C] surrounding a pipe
and exposed to room air at 20◦ C with h = 3.0 W/m2 · ◦ C. Calculate the heat loss from a 200◦ C,
5.0-cm-diameter pipe when covered with the critical radius of insulation and without insulation.
Solution
From Equation (2-18) we calculate ro as
k 0.17
ro = = = 0.0567 m = 5.67 cm
h 3.0
The inside radius of the insulation is 5.0/2 = 2.5 cm, so the heat transfer is calculated from Equation
(2-17) as
q 2π (200 − 20)
= = 105.7 W/m
L ln (5.67/2.5) 1
+
0.17 (0.0567)(3.0)
Without insulation the convection from the outer surface of the pipe is
q
= h(2πr)(Ti − To ) = (3.0)(2π)(0.025)(200 − 20) = 84.8 W/m
L

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.40 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER 2 Steady-State Conduction—One Dimension 41

So, the addition of 3.17 cm (5.67 − 2.5) of insulation actually increases the heat transfer by
25 percent.
As an alternative, fiberglass having a thermal conductivity of 0.04 W/m · ◦ C might be
employed as the insulation material. Then, the critical radius would be
k 0.04
ro = = = 0.0133 m = 1.33 cm
h 3.0
Now, the value of the critical radius is less than the outside radius of the pipe (2.5 cm), so addition
of any fiberglass insulation would cause a decrease in the heat transfer. In a practical pipe insulation
problem, the total heat loss will also be influenced by radiation as well as convection from the
outer surface of the insulation.

2-7 HEAT-SOURCE SYSTEMS

A number of interesting applications of the principles of heat transfer are concerned with
systems in which heat may be generated internally. Nuclear reactors are one example;
electrical conductors and chemically reacting systems are others. At this point we shall
confine our discussion to one-dimensional systems, or, more specifically, systems where
the temperature is a function of only one space coordinate.

Plane Wall with Heat Sources

Consider the plane wall with uniformly distributed heat sources shown in Figure 2-8. The
thickness of the wall in the x direction is 2L, and it is assumed that the dimensions in
the other directions are sufficiently large that the heat flow may be considered as one-
dimensional. The heat generated per unit volume is q̇, and we assume that the thermal
conductivity does not vary with temperature. This situation might be produced in a practical
situation by passing a current through an electrically conducting material. From Chapter 1,

Figure 2-8 Sketch illustrating

one-dimensional
conduction problem with
heat generation.

q• = heat generated per

unit volume
x=0

TO
TW

L TW
L x

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.41 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

42 2-7 Heat-Source Systems

the differential equation that governs the heat flow is

d 2 T q̇
+ =0 [2-19]
dx2 k
For the boundary conditions we specify the temperatures on either side of the wall, i.e.,
T = Tw at x = ± L [2-20]

The general solution to Equation (2-19) is

q̇ 2
T =− x + C1 x + C2 [2-21]
2k
Because the temperature must be the same on each side of the wall, C1 must be zero. The
temperature at the midplane (x = 0) is denoted by T0 and from Equation (2-21)
T0 = C2

The temperature distribution is therefore

q̇ 2
T − T0 = − x [2-22a]
2k
or  x 2
T − T0
= [2-22b]
Tw − T0 L
a parabolic distribution. An expression for the midplane temperature T0 may be obtained
through an energy balance. At steady-state conditions the total heat generated must equal
the heat lost at the faces. Thus
  
dT
2 −kA = q̇A 2L
dx x=L

where A is the cross-sectional area of the plate. The temperature gradient at the wall is
obtained by differentiating Equation (2-22b):
  
dT 2x 2
= (Tw − T0 ) = (Tw − T0 )
dx x=L L 2
x=L L

Then
2
−k(Tw − T0 ) = q̇L
L
and

q̇L2
T0 = + Tw [2-23]
2k
This same result could be obtained by substituting T = Tw at x = L into Equation
(2-22a).
The equation for the temperature distribution could also be written in the alternative
form
T − Tw x2
=1− 2 [2-22c]
T0 − Tw L

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.42 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER 2 Steady-State Conduction—One Dimension 43

2-8 CYLINDER WITH HEAT SOURCES

Consider a cylinder of radius R with uniformly distributed heat sources and constant thermal
conductivity. If the cylinder is sufficiently long that the temperature may be considered a
function of radius only, the appropriate differential equation may be obtained by neglecting
the axial, azimuth, and time-dependent terms in Equation (1-3b),
d 2 T 1 dT q̇
+ + =0 [2-24]
dr 2 r dr k
The boundary conditions are
T = Tw at r = R
and heat generated equals heat lost at the surface:

dT
q̇πR L = −k2πRL
2
dr r=R

Since the temperature function must be continuous at the center of the cylinder, we could
specify that
dT
=0 at r = 0
dr
However, it will not be necessary to use this condition since it will be satisfied automatically
when the two boundary conditions are satisfied.
We rewrite Equation (2-24)
d 2 T dT −q̇r
r + =
dr 2 dr k
and note that  
d 2 T dT d dT
r + = r
dr 2 dr dr dr
Then integration yields
dT −q̇r 2
r = + C1
dr 2k
and
−q̇r 2
T= + C1 ln r + C2
4k
From the second boundary condition above,

dT −q̇R −q̇R C1
= = +
dr r=R 2k 2k R

Thus
C1 = 0
We could also note that C1 must be zero because at r = 0 the logarithm function becomes
infinite.
From the first boundary condition,
−q̇R2
T = Tw = + C2 at r = R
4k

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.43 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

44 2-8 Cylinder with Heat Sources

so that
q̇R2
C2 = Tw +
4k
The final solution for the temperature distribution is then
q̇
T − Tw = (R2 − r 2 ) [2-25a]
4k
or, in dimensionless form,
T − Tw  r 2
=1− [2-25b]
T0 − Tw R
where T0 is the temperature at r = 0 and is given by
q̇R2
T0 = + Tw [2-26]
4k
It is left as an exercise to show that the temperature gradient at r = 0 is zero.
For a hollow cylinder with uniformly distributed heat sources the appropriate boundary
conditions would be
T = Ti at r = ri (inside surface)
T = To at r = ro (outside surface)

The general solution is still

q̇r 2
T =− + C1 ln r + C2
4k
Application of the new boundary conditions yields
q̇ 2 r
T − To = (ro − r 2 ) + C1 ln [2-27]
4k ro
where the constant C1 is given by
Ti − To + q̇ (ri2 − ro2 )/4k
C1 = [2-28]
ln (ri /ro )

EXAMPLE 2-7 Heat Source with Convection

A current of 200 A is passed through a stainless-steel wire [k = 19 W/m · ◦ C] 3 mm in diameter.
The resistivity of the steel may be taken as 70 μ · cm, and the length of the wire is 1 m. The
wire is submerged in a liquid at 110 ◦ C and experiences a convection heat-transfer coefficient of
4 k W/m2 · ◦ C. Calculate the center temperature of the wire.
Solution
All the power generated in the wire must be dissipated by convection to the liquid:
P = I 2 R = q = hA (Tw − T∞ ) [a]

The resistance of the wire is calculated from

L (70 × 10−6 )(100)
R=ρ = = 0.099 
A π(0.15)2

where ρ is the resistivity of the wire. The surface area of the wire is πdL, so from Equation (a),
(200)2 (0.099) = 4000π(3 × 10−3 )(1)(Tw − 110) = 3960 W

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.44 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER 2 Steady-State Conduction—One Dimension 45

and
Tw = 215◦ C [419◦ F]

The heat generated per unit volume q̇ is calculated from

P = q̇V = q̇πr 2 L

so that
3960
q̇ = = 560.2 MW/m3 [5.41 × 107 Btu/h · ft 3 ]
π (1.5 × 10−3 )2 (1)
Finally, the center temperature of the wire is calculated from Equation (2-26):
q̇ro2 (5.602 × 108 )(1.5 × 10−3 )2
T0 = + Tw = + 215 = 231.6◦ C [449◦ F]
4k (4)(19)

2-9 CONDUCTION-CONVECTION SYSTEMS

The heat that is conducted through a body must frequently be removed (or delivered) by
some convection process. For example, the heat lost by conduction through a furnace wall
must be dissipated to the surroundings through convection. In heat-exchanger applications
a finned-tube arrangement might be used to remove heat from a hot liquid. The heat transfer
from the liquid to the finned tube is by convection. The heat is conducted through the
material and finally dissipated to the surroundings by convection. Obviously, an analysis
of combined conduction-convection systems is very important from a practical standpoint.
We shall defer part of our analysis of conduction-convection systems to Chapter 10
on heat exchangers. For the present we wish to examine some simple extended-surface
problems. Consider the one-dimensional fin exposed to a surrounding fluid at a temperature
T∞ as shown in Figure 2-9. The temperature of the base of the fin is T0 . We approach the
problem by making an energy balance on an element of the fin of thickness dx as shown in
the figure. Thus

Energy in left face = energy out right face + energy lost by convection

The defining equation for the convection heat-transfer coefficient is recalled as

q = hA (Tw − T∞ ) [2-29]

where the area in this equation is the surface area for convection. Let the cross-sectional
area of the fin be A and the perimeter be P. Then the energy quantities are

dT
Energy in left face = qx = −kA
dx

dT
Energy out right face = qx+dx = −kA
dx x+dx
 
dT d 2 T
= −kA + 2 dx
dx dx
Energy lost by convection = hP dx (T − T∞ )

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.45 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

46 2-9 Conduction-Convection Systems

Figure 2-9 Sketch illustrating one-dimensional

conduction and convection through a
rectangular fin.

d qconv = h Pdx (T − T∞)

qx qx+dx
Z
dx
L
x
Base

Here it is noted that the differential surface area for convection is the product of the perimeter
of the fin and the differential length dx. When we combine the quantities, the energy balance
yields

d 2 T hP
− (T − T∞ ) = 0 [2-30a]
dx2 kA
Let θ = T − T∞ . Then Equation (2-30a) becomes

d 2 θ hP
− θ=0 [2-30b]
dx2 kA
One boundary condition is
θ = θ0 = T0 − T∞ at x = 0

The other boundary condition depends on the physical situation. Several cases may be
considered:
CASE 1 The fin is very long, and the temperature at the end of the fin is essentially
that of the surrounding fluid.
CASE 2 The fin is of finite length and loses heat by convection from its end.
CASE 3 The end of the fin is insulated so that dT/dx = 0 at x = L.
If we let m2 = hP/kA, the general solution for Equation (2-30b) may be written
θ = C1 e−mx + C2 emx [2-31]

For case 1 the boundary conditions are

θ = θ0 at x = 0
θ=0 at x = ∞

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.46 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER 2 Steady-State Conduction—One Dimension 47

and the solution becomes

θ T − T∞
= = e−mx [2-32]
θ0 T0 − T∞
For case 3 the boundary conditions are
θ = θ0 at x = 0
dθ
= 0 at x = L
dx
Thus
θ0 = C1 + C2
0 = m(−C1 e−mL + C2 emL )

Solving for the constants C1 and C2 , we obtain

θ e−mx emx
= −2mL
+ [2-33a]
θ0 1 + e 1 + e2mL

cosh [m(L − x)]

= [2-33b]
cosh mL
The hyperbolic functions are defined as
ex − e−x ex + e−x
sinh x = cosh x =
2 2
x
sinh x e − e −x
tanh x = =
cosh x ex + e−x
The solution for case 2 is more involved algebraically, and the result is
T − T∞ cosh m (L − x) + (h/mk) sinh m (L − x)
= [2-34]
To − T∞ cosh mL + (h/mk) sinh mL

All of the heat lost by the fin must be conducted into the base at x = 0. Using the
equations for the temperature distribution, we can compute the heat loss from

dT
q = −kA
dx x=0

An alternative method of integrating the convection heat loss could be used:

L L
q= hP(T − T∞ ) dx = hP θ dx
0 0

In most cases, however, the first equation is easier to apply. For case 1,
√
q = −kA (−mθ0 e−m(0) ) = hPkA θ0 [2-35]

For case 3,
 
1 1
q = −kAθ0 m −2mL
− [2-36]
1+e 1 + e+2mL
√
= hPkA θ0 tanh mL

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.47 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

48 2-10 Fins

The heat flow for case 2 is

√ sinh mL + (h/mk) cosh mL
q = hPkA (T0 − T∞ ) [2-37]
cosh mL + (h/mk) sinh mL
In this development it has been assumed that the substantial temperature gradients occur
only in the x direction. This assumption will be satisfied if the fin is sufficiently thin. For
most fins of practical interest the error introduced by this assumption is less than 1 percent.
The overall accuracy of practical fin calculations will usually be limited by uncertainties in
values of the convection coefficient h. It is worthwhile to note that the convection coefficient
is seldom uniform over the entire surface, as has been assumed above. If severe nonuniform
behavior is encountered, numerical finite-difference techniques must be employed to solve
the problem. Such techniques are discussed in Chapter 3.

2-10 FINS
In the foregoing development we derived relations for the heat transfer from a rod or fin
of uniform cross-sectional area protruding from a flat wall. In practical applications, fins
may have varying cross-sectional areas and may be attached to circular surfaces. In either
case the area must be considered as a variable in the derivation, and solution of the basic
differential equation and the mathematical techniques become more tedious. We present
only the results for these more complex situations. The reader is referred to References 1
and 8 for details on the mathematical methods used to obtain the solutions.
To indicate the effectiveness of a fin in transferring a given quantity of heat, a new
parameter called fin efficiency is defined by
actual heat transferred
Fin efficiency = = ηf
heat that would be transferred
if entire fin area were
at base temperature
For case 3, the fin efficiency becomes
√
hP kA θ0 tanh mL tanh mL
ηf = = [2-38]
hPLθ0 mL
The fins discussed were assumed to be sufficiently deep that the heat flow could be
considered one-dimensional. The expression for mL may be written

hP h(2z + 2t)
mL = L= L
kA kzt
where z is the depth of the fin, and t is the thickness. Now, if the fin is sufficiently deep, the
term 2z will be large compared with 2t, and

2hz 2h
mL = L= L
ktz kt
Multiplying numerator and denominator by L1/2 gives

2h 3/2
mL = L
kLt
Lt is called the profile area of the fin, which we define as
Am = Lt

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.48 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER 2 Steady-State Conduction—One Dimension 49

so that 
2h 3/2
mL = L [2-39]
kAm
We may therefore use the expression in Equation (2-39) to compute the efficiency of a fin
with insulated tip as given by Equation (2-38).
Harper and Brown [2] have shown that the solution in case 2 may be expressed in
the same form as Equation (2-38) when the length of the fin is extended by one-half the
thickness of the fin. In effect, lengthening of the fin by t/2 is assumed to represent the
same convection heat transfer as half the fin tip area placed on top and bottom of the fin. A
corrected length Lc is then used in all the equations that apply for the case of the fin with
an insulated tip. Thus
t
Lc = L + [2-40]
2
The error that results from this approximation will be less than 8 percent when
 1/2
ht 1
≤ [2-41]
2k 2
If a straight cylindrical rod extends from a wall, the corrected fin length is calculated
from
πd 2/4
Lc = L + = L + d/4 [2-42]
πd
Again, the real fin is extended a sufficient length to produce a circumferential area equal to
that of the tip area.
Examples of other types of fins are shown in Figure 2-10. Figure 2-11 presents a
comparison of the efficiencies of a triangular fin and a straight rectangular fin corresponding
to case 2. Figure 2-12 shows the efficiencies of circumferential fins of rectangular cross-
sectional area. Notice that the corrected fin lengths Lc and profile area Am have been
used in Figures 2-11 and 2-12. We may note that as r2c /r1 → 1.0, the efficiency of the
circumferential fin becomes identical to that of the straight fin of rectangular profile.
It is interesting to note that the fin efficiency reaches its maximum value for the trivial
case of L = 0, or no fin at all. Therefore, we should not expect to be able to maximize fin
performance with respect to fin length. It is possible, however, to maximize the efficiency
with respect to the quantity of fin material (mass, volume, or cost), and such a maximization
process has rather obvious economic significance. We have not discussed the subject of
radiation heat transfer from fins. The radiant transfer is an important consideration in a

Figure 2-10 Different types of finned surfaces. (a) Straight fin of

rectangular profile on plane wall, (b) straight fin of
rectangular profile on circular tube, (c) cylindrical tube
with radial fin of rectangular profile, (d) cylindrical-spine
or circular-rod fin.

(a) (b) (c) (d )

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.49 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

50 2-10 Fins

Figure 2-11 Efficiencies of straight rectangular and triangular fins.

0.9
t L  2t rectangular
0.8 Lc 
L triangular
0.7 L

Fin efficiency, ηƒ
tLc rectangular
0.6 Am  t
2 Lc triangular

0.5

0.4

0.3 t

0.2 L
0.1

0
0 0.5 1 1.5 2 2.5
Lc3/2(h/kAm)1/ 2

Figure 2-12 Efficiencies of circumferential fins of rectangular

profile, according to Reference 3.

1
0.9 L = r2 − r1
t
0.8 Lc = L +
2
t r2c = r1 + L c
0.7
Fin efficiency, ηf

Am = tL c
0.6 r2c冫r1
L r1
0.5 r2
0.4 1
2
0.3
3
0.2
4 5
0.1
0
0 0.5 1 1.5 2 2.5 3
Lc3Ⲑ2(hⲐkAm)1Ⲑ 2

number of applications, and the interested reader should consult Siegel and Howell [9] for
information on this subject.
In some cases a valid method of evaluating fin performance is to compare the heat
transfer with the fin to that which would be obtained without the fin. The ratio of these
quantities is

q with fin ηf Af hθ0

=
q without fin hAb θ0

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.50 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER 2 Steady-State Conduction—One Dimension 51

where Af is the total surface area of the fin and Ab is the base area. For the insulated-tip
fin described by Equation (2-36),

Af = PL
Ab = A

and the heat ratio would become

q with fin tanh mL
=√
q without fin hA/kP

This term is sometimes called the fin effectiveness.

Thermal Resistance for Fin-Wall Combinations

Consider a fin attached to a wall as illustrated in either Figure 2-11 or Figure 2-12. We
may calculate a thermal resistance for the wall using either Rw = x/kA for a plane wall,
or Rw = ln (ro /ri )/2πkL for a cylindrical wall. In the absence of the fin the convection
resistance at the surface would be 1/hA. The combined conduction and convection resistance
Rf for the fin is related to the heat lost by the fin through
θo
qf = ηf Af hθo = [2-43]
Rf
Figure 2-13 Heat
or, the fin resistance may be expressed as loss from fin-wall
combination.
1
Rf = [2-44]
η f Af h Wall, Rw

The overall heat transfer through the fin-wall combination is then Fin, h, T

Ti − T∞
qf = [2-45]
Rwf + Rf
Ao
where Ti is the inside wall temperature and Rwf is the wall resistance at the fin position.
This heat transfer is only for the fin portion of the wall. Now consider the wall section
shown in Figure 2-13, having a wall area Ab for the fin and area Ao for the open section of
Ab Af
the wall exposed directly to the convection environment. The open wall heat transfer is
Ti − T∞
qo = [2-46]
Rwo + Ro

where now
1
Ro = [2-47]
hAo
and Rwo is the wall resistance for the open wall section. This value is Rwo = x/kw Ao for
a plane wall, where x is the wall thickness. A logarithmic form would be employed for a
cylindrical wall, as noted above. The total heat lost by the wall is therefore

qtotal = qf + qo [2-48]

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.51 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

52 2-10 Fins

which may be expressed in terms of the thermal resistances by

 
1 1
qtotal = (Ti − T∞ ) +
Rwf + Rf Rwo + Ro
Rwo + Ro + Rwf + Rf
= (Ti − T∞ ) [2-49]
(Rwf + Rf )(Rwo + Ro )

Conditions When Fins Do Not Help

At this point we should remark that the installation of fins on a heat-transfer surface will not
necessarily increase the heat-transfer rate. If the value of h, the convection coefficient, is
large, as it is with high-velocity fluids or boiling liquids, the fin may produce a reduction in
heat transfer because the conduction resistance then represents a larger impediment to the
heat flow than the convection resistance. To illustrate the point, consider a stainless-steel
pin fin that has k = 16 W/m · ◦ C, L = 10 cm, d = 1 cm and that is exposed to a boiling-water
convection situation with h = 5000 W/m2 · ◦ C. From Equation (2-36) we can compute
q with fin tanh mL
=√
q without fin hA/kp
 1/2
5000π(1 × 10−2 )(4) −2
tanh (10 × 10 )
16π(1 × 10−2 )2
=  1/2
5000π(1 × 10−2 )2
(4)(16)π(1 × 10−2 )
= 1.13

Thus, this rather large pin produces an increase of only 13 percent in the heat transfer.
Still another method of evaluating fin performance is discussed in Problem 2-68. Kern
and Kraus [8] give a very complete discussion of extended-surface heat transfer. Some
photographs of different fin shapes used in electronic cooling applications are shown in
Figure 2-14. These fins are obviously not one-dimensional, i.e., they cannot be characterized
with a single space coordinate.

Cautionary Remarks Concerning Convection Coefficients for Fins

We have already noted that the convection coefficient may vary with type of fluid, flow
velocity, geometry, etc. As we shall see in Chapters 5, 6, and 7, empirical correlations for
h frequently have uncertainties of the order of ±25 percent. Moreover, the correlations are
based on controlled laboratory experiments that are infrequently matched in practice. What
this means is that the assumption of constant h used in the derivation of fin performance
may be in considerable error and the value of h may vary over the fin surface. For the
heat-transfer practitioner, complex geometries like those shown in Figure 2-14 must be
treated with particular care. These configurations usually must be tested under near or actual
operating conditions in order to determine their performance with acceptable reliability.
These remarks are not meant to discourage the reader, but rather to urge prudence when
estimating the performance of complex finned surfaces for critical applications.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.52 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER 2 Steady-State Conduction—One Dimension 53

Figure 2-14 Some fin arrangements used in electronic cooling applications.

Source: Courtesy Wakefield Engineering Inc., Wakefield, Mass.

Influence of Thermal Conductivity on Fin

Temperature Profiles EXAMPLE 2-8
Compare the temperature distributions in a straight cylindrical rod having a diameter of 2 cm
and a length of 10 cm and exposed to a convection environment with h = 25 W/m2 · ◦ C, for
three fin materials: copper [k = 385 W/m · ◦ C], stainless steel [k = 17 W/m · ◦ C], and glass
[k = 0.8 W/m · ◦ C]. Also compare the relative heat flows and fin efficiencies.
Solution
We have
hP (25)π(0.02) 5000
= =
kA kπ(0.01)2 k

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.53 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

54 2-10 Fins

The terms of interest are therefore

Material hP m mL
kA
Copper 12.99 3.604 0.3604
Stainless steel 294.1 17.15 1.715
Glass 6250 79.06 7.906

These values may be inserted into Equation (2-33a) to calculate the temperatures at different
x locations along the rod, and the results are shown in Figure Example 2-8. We notice that the
glass behaves as a “very long” fin, and its behavior could be calculated from Equation (2-32). The
fin efficiencies are calculated from Equation (2-38) by using the corrected length approximation
of Equation (2-42). We have
d 2
Lc = L + = 10 + = 10.5 cm
4 4

Figure Example 2-8

1.0 Copper, k = 385 W冫m • ˚C

h = 25 W冫m2 • ˚C
d = 2 cm
0.8 L = 10 cm

0.6
Stainless steel, k = 17 W冫m • ˚C
θ
θo

0.4

0.2
Glass, k = 0.8 W冫 m • ˚C

2 4 6 8 10
x, cm

The parameters of interest for the heat-flow and efficiency comparisons are now tabulated as

Material hP kA mLc
Copper 0.190 0.3784
Stainless steel 0.0084 1.8008
Glass 3.9 × 10−4 8.302

To compare the heat flows we could either calculate the values from Equation (2-36) for a unit
value of θ0 or observe that the fin efficiency gives a relative heat-flow comparison because the
maximum heat transfer is the same for all three cases; i.e., we are dealing with the same fin size,
shape, and value of h. We thus calculate the values of ηf from Equation (2-38) and the above
values of mLc .

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.54 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER 2 Steady-State Conduction—One Dimension 55

q relative to
Material ηf copper, %
Copper 0.955 100
Stainless steel 0.526 53.1
Glass 0.124 12.6

The temperature profiles in the accompanying figure can be somewhat misleading. The glass has
the steepest temperature gradient at the base, but its much lower value of k produces a lower
heat-transfer rate.

Straight Aluminum Fin EXAMPLE 2-9

An aluminum fin [k = 200 W/m · ◦ C] 3.0 mm thick and 7.5 cm long protrudes from a wall,
as in Figure 2-9. The base is maintained at 300◦ C, and the ambient temperature is 50◦ C with
h = 10 W/m2 · ◦ C. Calculate the heat loss from the fin per unit depth of material.
Solution
We may use the approximate method of solution by extending the fin a fictitious length t/2 and
then computing the heat transfer from a fin with insulated tip as given by Equation (2-36). We
have
Lc = L + t/2 = 7.5 + 0.15 = 7.65 cm [3.01 in]

 

hP h(2z + 2t) 1/2 2h

m= = ≈
kA ktz kt

when the fin depth z  t. So,

 1/2
(2)(10)
m= = 5.774
(200)(3 × 10−3 )

From Equation (2-36), for an insulated-tip fin

√
q = (tanh mLc ) hPkA θ0

For a 1 m depth
A = (1)(3 × 10−3 ) = 3 × 10−3 m2 [4.65 in2 ]

and
q = (5.774)(200)(3 × 10−3 )(300 − 50) tanh [(5.774)(0.0765)]
= 359 W/m [373.5 Btu/h · ft]

Circumferential Aluminum Fin EXAMPLE 2-10

Aluminum fins 1.5 cm wide and 1.0 mm thick are placed on a 2.5-cm-diameter tube to dissipate the
heat. The tube surface temperature is 170◦ , and the ambient-fluid temperature is 25◦ C. Calculate
the heat loss per fin for h = 130 W/m2 · ◦ C. Assume k = 200 W/m · ◦ C for aluminum.
Solution
For this example we can compute the heat transfer by using the fin-efficiency curves in
Figure 2-12. The parameters needed are
Lc = L + t/2 = 1.5 + 0.05 = 1.55 cm
r1 = 2.5/2 = 1.25 cm
r2c = r1 + Lc = 1.25 + 1.55 = 2.80 cm

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.55 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

56 2-10 Fins

r2c /r1 = 2.80/1.25 = 2.24

Am = t(r2c − r1 ) = (0.001)(2.8 − 1.25)(10−2 ) = 1.55 × 10−5 m2
   1/2
3/2 h 1/2 3/2 130
Lc = (0.0155) = 0.396
kAm (200)(1.55 × 10−5 )

From Figure 2-12, ηf = 82 percent. The heat that would be transferred if the entire fin were at the
base temperature is (both sides of fin exchanging heat)
2 − r 2 )h(T − T )
qmax = 2π(r2c 1 0 ∞
= 2π(2.82 − 1.252 )(10−4 )(130)(170 − 25)
= 74.35 W [253.7 Btu/h]

The actual heat transfer is then the product of the heat flow and the fin efficiency:
qact = (0.82)(74.35) = 60.97 W [208 Btu/h]

EXAMPLE 2-11 Rod with Heat Sources

A rod containing uniform heat sources per unit volume q̇ is connected to two temperatures as shown
in Figure Example 2-11. The rod is also exposed to an environment with convection coefficient h
and temperature T∞ . Obtain an expression for the temperature distribution in the rod.

Figure Example 2-11

T1 T2

h, T∞
q•

qx A qx+dx

dx

Solution
We first must make an energy balance on the element of the rod shown, similar to that used to
derive Equation (2-30). We have
Energy in left face + heat generated in element
= energy out right face + energy lost by convection

or
 
dT dT d2T
−kA + q̇A dx = −kA + 2 dx + hP dx (T − T∞ )
dx dx dx

Simplifying, we have
d2T hP q̇
− (T − T∞ ) + = 0 [a]
dx2 kA k

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.56 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER 2 Steady-State Conduction—One Dimension 57

or, with θ = T − T∞ and m2 = hP/kA

d2θ q̇
− m2 θ + = 0 [b]
dx k
We can make a further variable substitution as
θ = θ − q̇/km2

so that our differential equation becomes

d 2 θ
− m2 θ = 0 [c]
dx2
which has the general solution
θ = C1 e−mx + C2 emx [d]
The two end temperatures are used to establish the boundary conditions:
θ = θ1 = T1 − T∞ − q̇/km2 = C1 + C2
θ = θ2 = T2 − T∞ − q̇/km2 = C1 e−mL + C2 emL

Solving for the constants C1 and C2 gives

(θ1 e2mL − θ2 emL )e−mx + (θ2 emL − θ1 )emx
θ = [e]
e2mL − 1

For an infinitely long heat-generating fin with the left end maintained at T1 , the temperature
distribution becomes
θ /θ1 = e−mx [f]
a relation similar to Equation (2-32) for a non-heat-generating fin.
Comment
Note that the above relationships assume one-dimensional behavior, i.e., temperature dependence
only on the x-coordinate and temperature uniformity across the area A. For sufficiently large heat
generation rates and/or cross-section areas, the assumption may no longer be valid. In these cases,
the problem must be treated as multidimensional using the techniques described in Chapter 3.

2-11 THERMAL CONTACT RESISTANCE

Imagine two solid bars brought into contact as indicated in Figure 2-15, with the sides of the
bars insulated so that heat flows only in the axial direction. The materials may have different
thermal conductivities, but if the sides are insulated, the heat flux must be the same through
both materials under steady-state conditions. Experience shows that the actual temperature
profile through the two materials varies approximately as shown in Figure 2-15b. The tem-
perature drop at plane 2, the contact plane between the two materials, is said to be the result of
a thermal contact resistance. Performing an energy balance on the two materials, we obtain

T1 − T2A T2A − T2B T2B − T3

q = kA A = = kB A
xA 1/ hc A xB

or
T1 − T3
q= [2-50]
xA /kA A + 1/hc A + xB /kB A

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.57 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

58 2-11 Thermal Contact Resistance

where the quantity 1/hc A is called the thermal contact resistance and hc is called the contact
coefficient. This factor can be extremely important in a number of applications because of
the many heat-transfer situations that involve mechanical joining of two materials.
The physical mechanism of contact resistance may be better understood by examining
a joint in more detail, as shown in Figure 2-16. The actual surface roughness is exaggerated
to implement the discussion. No real surface is perfectly smooth, and the actual surface
roughness is believed to play a central role in determining the contact resistance. There are
two principal contributions to the heat transfer at the joint:

1. The solid-to-solid conduction at the spots of contact

2. The conduction through entrapped gases in the void spaces created by the contact

The second factor is believed to represent the major resistance to heat flow, because the
thermal conductivity of the gas is quite small in comparison to that of the solids.

Figure 2-15 Illustrations of thermal-contact-resistance

effect: (a) physical situation;
(b) temperature profile.

q A B q

Δ xA Δ xB

(a)

T1

T2A

T2B

T3

1 2 3 x
(b)

Figure 2-16 Joint-roughness model for analysis of

thermal contact resistance.

T2B
Lg
T2A

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.58 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER 2 Steady-State Conduction—One Dimension 59

Table 2-3 Contact conductance of typical surfaces.

1/hc
Roughness Temperature, Pressure, h · ft2 · ◦ F/
Surface type μ in μm ◦C atm Btu m2 · ◦ C/W × 104

416 Stainless, ground, air 100 2.54 90–200 3–25 0.0015 2.64
304 Stainless, ground, air 45 1.14 20 40–70 0.003 5.28
416 Stainless, ground, with 100 2.54 30–200 7 0.002 3.52
0.001-in brass shim, air
Aluminum, ground, air 100 2.54 150 12–25 0.0005 0.88
10 0.25 150 12–25 0.0001 0.18
Aluminum, ground, with 100 2.54 150 12–200 0.0007 1.23
0.001-in brass shim, air
Copper, ground, air 50 1.27 20 12–200 0.00004 0.07
Copper, milled, air 150 3.81 20 10–50 0.0001 0.18
Copper, milled, vacuum 10 0.25 30 7–70 0.0005 0.88

Designating the contact area by Ac and the void area by Av , we may write for the heat
flow across the joint
T2A − T2B T2A − T2B T2A − T2B
q= + kf Av =
Lg /2kA Ac + Lg /2kB Ac Lg 1/hc A

where Lg is the thickness of the void space and kf is the thermal conductivity of the fluid
which fills the void space. The total cross-sectional area of the bars is A. Solving for hc , the
contact coefficient, we obtain
 
1 Ac 2kA kB Av
hc = + kf [2-51]
Lg A kA + kB A

In most instances, air is the fluid filling the void space and kf is small compared with kA and
kB . If the contact area is small, the major thermal resistance results from the void space. The
main problem with this simple theory is that it is extremely difficult to determine effective
values of Ac , Av , and Lg for surfaces in contact.
From the physical model, we may tentatively conclude:
1. The contact resistance should increase with a decrease in the ambient gas pressure when
the pressure is decreased below the value where the mean free path of the molecules
is large compared with a characteristic dimension of the void space, since the effective
thermal conductance of the entrapped gas will be decreased for this condition.
2. The contact resistance should be decreased for an increase in the joint pressure since
this results in a deformation of the high spots of the contact surfaces, thereby creating a
greater contact area between the solids.
A very complete survey of the contact-resistance problem is presented in References 4,
6, 7, 10, 11. Unfortunately, there is no satisfactory theory that will predict thermal contact
resistance for all types of engineering materials, nor have experimental studies yielded
completely reliable empirical correlations. This is understandable because of the many
complex surface conditions that may be encountered in practice.
Radiation heat transfer across the joint can also be important when high temperatures are
encountered. This energy transfer may be calculated by the methods discussed in Chapter 8.
For design purposes the contact conductance values given in Table 2-3 may be used
in the absence of more specific information. Thermal contact resistance can be reduced
markedly, perhaps as much as 75 percent, by the use of a “thermal grease” like Dow 340.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.59 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

60 List of Worked Examples

Influence of Contact Conductance

EXAMPLE 2-12 on Heat Transfer
Two 3.0-cm-diameter 304 stainless-steel bars, 10 cm long, have ground surfaces and are exposed
to air with a surface roughness of about 1 μm. If the surfaces are pressed together with a pressure
of 50 atm and the two-bar combination is exposed to an overall temperature difference of 100◦ C,
calculate the axial heat flow and temperature drop across the contact surface.
Solution
The overall heat flow is subject to three thermal resistances, one conduction resistance for each
bar, and the contact resistance. For the bars
x (0.1)(4)
Rth = = = 8.679◦ C/W
kA (16.3)π(3 × 10−2 )2

From Table 2-2 the contact resistance is

1 (5.28 × 10−4 )(4)
Rc = = = 0.747◦ C/W
hc A π(3 × 10−2 )2

The total thermal resistance is therefore


Rth = (2)(8.679) + 0.747 = 18.105

and the overall heat flow is

T 100
q=  = = 5.52 W [18.83 Btu/h]
Rth 18.105

The temperature drop across the contact is found by taking the ratio of the contact resistance to
the total thermal resistance:
Rc (0.747)(100)
Tc =  T = = 4.13◦ C [39.43◦ F]
Rth 18.105

In this problem the contact resistance represents about 4 percent of the total resistance.

REVIEW QUESTIONS
1. What is meant by the term one-dimensional when applied to conduction problems?
2. What is meant by thermal resistance?
3. Why is the one-dimensional heat-flow assumption important in the analysis of fins?
4. Define fin efficiency.
5. Why is the insulated-tip solution important for the fin problems?
6. What is meant by thermal contact resistance? Upon what parameters does this resistance
depend?

LIST OF WORKED EXAMPLES

2-1 Multilayer conduction
2-2 Multilayer cylindrical system
2-3 Heat transfer through a composite wall
2-4 Cooling cost savings with extra insulation
2-5 Overall heat-transfer coefficient for a tube

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.60 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER2 Steady-State Conduction—One Dimension 61

2-6 Critical insulation thickness

2-7 Heat source with convection
2-8 Influence of thermal conductivity on fin temperature profiles
2-9 Straight aluminum fin
2-10 Circumferential aluminum fin
2-11 Rod with heat sources
2-12 Influence of contact conductance on heat transfer

PROBLEMS
2-1 A wall 2 cm thick is to be constructed from material that has an average thermal
conductivity of 1.3 W/m · ◦ C. The wall is to be insulated with material having an
average thermal conductivity of 0.35 W/m · ◦ C, so that the heat loss per square meter
will not exceed 1830 W. Assuming that the inner and outer surface temperatures of
the insulated wall are 1300 and 30◦ C, calculate the thickness of insulation required.
2-2 A certain material 2.5 cm thick, with a cross-sectional area of 0.1 m2 , has one side
maintained at 35◦ C and the other at 95◦ C. The temperature at the center plane of
the material is 62◦ C, and the heat flow through the material is 1 kW. Obtain an
expression for the thermal conductivity of the material as a function of temperature.
2-3 A composite wall is formed of a 2.5-cm copper plate, a 3.2-mm layer of asbestos, and
a 5-cm layer of fiberglass. The wall is subjected to an overall temperature difference
of 560◦ C. Calculate the heat flow per unit area through the composite structure.
2-4 Find the heat transfer per unit area through the composite wall in Figure P2-4.
Assume one-dimensional heat flow.
Figure P2-4

kA = 150 W/m•˚C
kB = 30
kC = 50 Ac = 0.1 m2
kD = 70
B
AB = AD q
A C
T = 370˚C
D
T = 66˚C

2.5 cm 7.5 cm 5.0 cm

2-5 One side of a copper block 5 cm thick is maintained at 250◦ C. The other side is
covered with a layer of fiberglass 2.5 cm thick. The outside of the fiberglass is main-
tained at 35◦ C, and the total heat flow through the copper-fiberglass combination is
52 kW. What is the area of the slab?
2-6 An outside wall for a building consists of a 10-cm layer of common brick and a
2.5-cm layer of fiberglass [k = 0.05 W/m · ◦ C]. Calculate the heat flow through the
wall for a 25◦ C temperature differential.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.61 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

62 Problems

2-7 One side of a copper block 4 cm thick is maintained at 175◦ C. The other side is
covered with a layer of fiberglass 1.5 cm thick. The outside of the fiberglass is
maintained at 80◦ C, and the total heat flow through the composite slab is 300 W.
What is the area of the slab?
2-8 A plane wall is constructed of a material having a thermal conductivity that varies
as the square of the temperature according to the relation k = k0 (1 + βT 2 ). Derive
an expression for the heat transfer in such a wall.
2-9 A steel tube having k = 46 W/m · ◦ C has an inside diameter of 3.0 cm and a tube wall
thickness of 2 mm. A fluid flows on the inside of the tube producing a convection
coefficient of 1500 W/m2 · ◦ C on the inside surface, while a second fluid flows across
the outside of the tube producing a convection coefficient of 197 W/m2 · ◦ C on the
outside tube surface. The inside fluid temperature is 223◦ C while the outside fluid
temperature is 57◦ C. Calculate the heat lost by the tube per meter of length.
2-10 A certain material has a thickness of 30 cm and a thermal conductivity of
0.04 W/m · ◦ C. At a particular instant in time, the temperature distribution with x,
the distance from the left face, is T = 150x2 − 30x, where x is in meters. Calculate
the heat-flow rates at x = 0 and x = 30 cm. Is the solid heating up or cooling down?
2-11 A 0.025-mm-diameter stainless steel wire having k = 16 W/m · ◦ C is connected to
two electrodes. The length of the wire is 80 cm and it is exposed to a convection
environment at 20◦ C with h = 500 W/m2 · ◦ C. A voltage is impressed on the wire
that produces temperatures at each electrode of 200◦ C. Determine the total heat lost
by the wire.
2-12 A wall is constructed of 2.0 cm of copper, 3.0 mm of asbestos sheet
[k = 0.166 W/m · ◦ C], and 6.0 cm of fiberglass. Calculate the heat flow per unit
area for an overall temperature difference of 500◦ C.
2-13 A certain building wall consists of 6.0 in of concrete [k = 1.2 W/m · ◦ C], 2.0 in of
fiberglass insulation, and 38 in of gypsum board [k = 0.05 W/m · ◦ C]. The inside
and outside convection coefficients are 2.0 and 7.0 Btu/h · ft 2 · ◦ F, respectively. The
outside air temperature is 20◦ F, and the inside temperature is 72◦ F. Calculate the
overall heat-transfer coefficient for the wall, the R value, and the heat loss per unit
area.
2-14 A wall is constructed of a section of stainless steel [k = 16 W/m · ◦ C] 4.0 mm thick
with identical layers of plastic on both sides of the steel. The overall heat-transfer
coefficient, considering convection on both sides of the plastic, is 120 W/m2 · ◦ C.
If the overall temperature difference across the arrangement is 60◦ C, calculate the
temperature difference across the stainless steel.
2-15 An ice chest is constructed of Styrofoam [k = 0.033 W/m · ◦ C] with inside dimen-
sions of 25 by 40 by 100 cm. The wall thickness is 5.0 cm. The outside of the chest
is exposed to air at 25◦ C with h = 10 W/m2 · ◦ C. If the chest is completely filled
with ice, calculate the time for the ice to completely melt. State your assumptions.
The enthalpy of fusion for water is 330 kJ/kg.
2-16 A spherical tank, 1 m in diameter, is maintained at a temperature of 120◦ C and
exposed to a convection environment. With h = 25 W/m2 · ◦ C and T∞ = 15◦ C, what
thickness of urethane foam should be added to ensure that the outer temperature of
the insulation does not exceed 40◦ C? What percentage reduction in heat loss results
from installing this insulation?
2-17 A hollow sphere is constructed of aluminum with an inner diameter of 4 cm and an
outer diameter of 8 cm. The inside temperature is 100◦ C and the outer temperature
is 50◦ C. Calculate the heat transfer.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.62 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER2 Steady-State Conduction—One Dimension 63

2-18 Suppose the sphere in Problem 2-16 is covered with a 1-cm layer of an insulating
material having k = 50 m W/m · ◦ C and the outside of the insulation is exposed to
an environment with h = 20 W/m2 · ◦ C and T∞ = 10◦ C. The inside of the sphere
remains at 100◦ C. Calculate the heat transfer under these conditions.
2-19 In Appendix A, dimensions of standard steel pipe are given. Suppose a 3-in schedule
80 pipe is covered with 1 in of an insulation having k = 60 m W/m · ◦ C and the
outside of the insulation is exposed to an environment having h = 10 W/m2 · ◦ C and
T∞ = 20◦ C. The temperature of the inside of the pipe is 250◦ C. For unit length of
the pipe calculate (a) overall thermal resistance and (b) heat loss.
2-20 A steel pipe with 5-cm OD is covered with a 6.4-mm asbestos insulation
[k = 0.096 Btu/h · ft · ◦ F] followed by a 2.5-cm layer of fiberglass insulation
[k = 0.028 Btu/h · ft · ◦ F]. The pipe-wall temperature is 315◦ C, and the outside insu-
lation temperature is 38◦ C. Calculate the interface temperature between the asbestos
and fiberglass.
2-21 Derive an expression for the thermal resistance through a hollow spherical shell
of inside radius ri and outside radius ro having a thermal conductivity k. (See
Equation 2–10.)
2-22 A 1.0-mm-diameter wire is maintained at a temperature of 400◦ C and exposed to
a convection environment at 40◦ C with h = 120 W/m2 · ◦ C. Calculate the thermal
conductivity that will just cause an insulation thickness of 0.2 mm to produce a
“critical radius.” How much of this insulation must be added to reduce the heat
transfer by 75 percent from that which would be experienced by the bare wire?
2-23 A 2.0-in schedule 40 steel pipe (see Appendix A) has k = 27 Btu/h · ft · ◦ F. The fluid
inside the pipe has h = 30 Btu/h · ft 2 · ◦ F, and the outer surface of the pipe is cov-
ered with 0.5-in fiberglass insulation with k = 0.023 Btu/h · ft · ◦ F. The convection
coefficient on the outer insulation surface is 2.0 Btu/h · ft 2 · ◦ F. The inner fluid tem-
perature is 320◦ F and the ambient temperature is 70◦ F. Calculate the heat loss per
foot of length.
2-24 Derive a relation for the critical radius of insulation for a sphere.
2-25 A cylindrical tank 80 cm in diameter and 2.0 m high contains water at 80◦ C. The
tank is 90 percent full, and insulation is to be added so that the water temperature
will not drop more than 2◦ C per hour. Using the information given in this chapter,
specify an insulating material and calculate the thickness required for the specified
cooling rate.
2-26 A hot steam pipe having an inside surface temperature of 250◦ C has an inside
diameter of 8 cm and a wall thickness of 5.5 mm. It is covered with a 9-cm layer of
insulation having k = 0.5 W/m · ◦ C, followed by a 4-cm layer of insulation having
k = 0.25 W/m · ◦ C. The outside temperature of the insulation is 20◦ C. Calculate the
heat lost per meter of length. Assume k = 47 W/m · ◦ C for the pipe.
2-27 A house wall may be approximated as two 1.2-cm layers of fiber insulating board,
an 8.0-cm layer of loosely packed asbestos, and a 10-cm layer of common brick.
Assuming convection heat-transfer coefficients of 12 W/m2 · ◦ C on both sides of the
wall, calculate the overall heat-transfer coefficient for this arrangement.
2-28 Calculate the R value for the following insulations: (a) urethane foam, (b) fiberglass
mats, (c) mineral wool blocks, (d) calcium silicate blocks.
2-29 An insulation system is to be selected for a furnace wall at 1000◦ C using first a layer
of mineral wool blocks followed by fiberglass boards. The outside of the insulation
is exposed to an environment with h = 15 W/m2 · ◦ C and T∞ = 40◦ C. Using the
data of Table 2-1, calculate the thickness of each insulating material such that the

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.63 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

64 Problems

interface temperature is not greater than 400◦ C and the outside temperature is not
greater than 55◦ C. Use mean values for the thermal conductivities. What is the heat
loss in this wall in watts per square meter?
2-30 Derive an expression for the temperature distribution in a plane wall having uni-
formly distributed heat sources and one face maintained at a temperature T1 while
the other face is maintained at a temperature T2 . The thickness of the wall may be
taken as 2L.
2-31 A 5-cm-diameter steel pipe is covered with a 1-cm layer of insulating material hav-
ing k = 0.22 W/m · ◦ C followed by a 3-cm-thick layer of another insulating material
having k = 0.06 W/m · ◦ C. The entire assembly is exposed to a convection surround-
ing condition of h = 60 W/m2 · ◦ C and T∞ = 15◦ C. The outside surface temperature
of the steel pipe is 400◦ C. Calculate the heat lost by the pipe-insulation assembly
for a pipe length of 20 m. Express in Watts.
2-32 Derive an expression for the temperature distribution in a plane wall in which dis-
tributed heat sources vary according to the linear relation
q̇ = q̇w [1 + β(T − Tw )]
where q̇w is a constant and equal to the heat generated per unit volume at the wall
temperature Tw . Both sides of the plate are maintained at Tw , and the plate thickness
is 2L.
2-33 A circumferential fin of rectangular profile is constructed of stainless steel with
k = 43 W/m · ◦ C and a thickness of 1.0 mm. The fin is installed on a tube having
a diameter of 3.0 cm and the outer radius of the fin is 4.0 cm. The inner tube is
maintained at 250◦ C and the assembly is exposed to a convection environment
having T∞ = 35◦ C and h = 45 W/m2 · ◦ C. Calculate the heat lost by the fin.
2-34 A plane wall 6.0 cm thick generates heat internally at the rate of 0.3 MW/m3 . One
side of the wall is insulated, and the other side is exposed to an environment at 93◦ C.
The convection heat-transfer coefficient between the wall and the environment is
570 W/m2 · ◦ C. The thermal conductivity of the wall is 21 W/m · ◦ C. Calculate the
maximum temperature in the wall.
2-35 Consider a shielding wall for a nuclear reactor. The wall receives a gamma-ray flux
such that heat is generated within the wall according to the relation
q̇ = q̇0 e−ax
where q̇0 is the heat generation at the inner face of the wall exposed to the gamma-ray
flux and a is a constant. Using this relation for heat generation, derive an expression
for the temperature distribution in a wall of thickness L, where the inside and outside
temperatures are maintained at Ti and T0 , respectively. Also obtain an expression
for the maximum temperature in the wall.
2-36 Repeat Problem 2-35, assuming that the outer surface is adiabatic while the inner
surface temperature is maintained at Ti .
2-37 Rework Problem 2-32 assuming that the plate is subjected to a convection environ-
ment on both sides of temperature T∞ with a heat-transfer coefficient h. Tw is now
some reference temperature not necessarily the same as the surface temperature.
2-38 Heat is generated in a 2.5-cm-square copper rod at the rate of 35.3 MW/m3 . The rod
is exposed to a convection environment at 20◦ C, and the heat-transfer coefficient is
4000 W/m2 · ◦ C. Calculate the surface temperature of the rod.
2-39 A plane wall of thickness 2L has an internal heat generation that varies according
to q̇ = q̇0 cos ax, where q̇0 is the heat generated per unit volume at the center of the

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.64 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER2 Steady-State Conduction—One Dimension 65

wall (x = 0) and a is a constant. If both sides of the wall are maintained at a constant
temperature of Tw , derive an expression for the total heat loss from the wall per unit
surface area.
2-40 A certain semiconductor material has a conductivity of 0.0124 W/cm · ◦ C. A rect-
angular bar of the material has a cross-sectional area of 1 cm2 and a length of 3 cm.
One end is maintained at 300◦ C and the other end at 100◦ C, and the bar carries a cur-
rent of 50 A. Assuming the longitudinal surface is insulated, calculate the midpoint
temperature in the bar. Take the resistivity as 1.5 × 10−3  · cm.
2-41 The temperature distribution in a certain plane wall is

T − T1
= C1 + C2 x2 + C3 x3
T2 − T1

where T1 and T2 are the temperatures on each side of the wall. If the thermal con-
ductivity of the wall is constant and the wall thickness is L, derive an expression for
the heat generation per unit volume as a function of x, the distance from the plane
where T = T1 . Let the heat-generation rate be q̇0 at x = 0.
2-42 Electric heater wires are installed in a solid wall having a thickness of 8 cm and
k = 2.5 W/m · ◦ C. The right face is exposed to an environment with h = 50 W/m2 · ◦ C
and T∞ = 30◦ C, while the left face is exposed to h = 75 W/m2 · ◦ C and T∞ = 50◦ C.
What is the maximum allowable heat-generation rate such that the maximum tem-
perature in the solid does not exceed 300◦ C?
2-43 Two 5.0-cm-diameter aluminum bars, 2 cm long, have ground surfaces and are joined
in compression with a 0.025-mm brass shim at a pressure exceeding 20 atm. The
combination is subjected to an overall temperature difference of 200◦ C. Calculate
the temperature drop across the contact join.
2-44 A 3.0-cm-thick plate has heat generated uniformly at the rate of 5 × 105 W/m3 . One
side of the plate is maintained at 200◦ C and the other side at 45◦ C. Calculate the
temperature at the center of the plate for k = 16 W/m · ◦ C.
2-45 Heat is generated uniformly in a stainless steel plate having k = 20 W/m · ◦ C.
The thickness of the plate is 1.0 cm and the heat-generation rate is 500 MW/m3 .
If the two sides of the plate are maintained at 100 and 200◦ C, respectively, calculate
the temperature at the center of the plate.
2-46 A plate having a thickness of 4.0 mm has an internal heat generation of 200 MW/m3
and a thermal conductivity of 25 W/m · ◦ C. One side of the plate is insulated
and the other side is maintained at 100◦ C. Calculate the maximum temperature in
the plate.
2-47 A 3.2-mm-diameter stainless-steel wire 30 cm long has a voltage of 10 V impressed
on it. The outer surface temperature of the wire is maintained at 93◦ C. Calculate the
center temperature of the wire. Take the resistivity of the wire as 70 μ · cm and
the thermal conductivity as 22.5 W/m · ◦ C.
2-48 The heater wire of Example 2-7 is submerged in a fluid maintained at 93◦ C. The con-
vection heat-transfer coefficient is 5.7 kW/m2 · ◦ C. Calculate the center temperature
of the wire.
2-49 An electric current is used to heat a tube through which a suitable cooling fluid
flows. The outside of the tube is covered with insulation to minimize heat loss to
the surroundings, and thermocouples are attached to the outer surface of the tube to
measure the temperature. Assuming uniform heat generation in the tube, derive an
expression for the convection heat-transfer coefficient on the inside of the tube in

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.65 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

66 Problems

terms of the measured variables: voltage E, current I , outside tube wall temperature
T0 , inside and outside radii ri and ro , tube length L, and fluid temperature Tf .
2-50 Derive an expression for the temperature distribution in a sphere of radius r with
uniform heat generation q̇ and constant surface temperature Tw .
2-51 A stainless-steel sphere [k = 16 W/m · ◦ C] having a diameter of 4 cm is exposed to
a convection environment at 20◦ C, h = 15 W/m2 · ◦ C. Heat is generated uniformly
in the sphere at the rate of 1.0 MW/m3 . Calculate the steady-state temperature for
the center of the sphere.
2-52 An aluminum-alloy electrical cable has k = 190 W/m · ◦ C, a diameter of 30 mm,
and carries an electric current of 230 A. The resistivity of the cable is 2.9 μ · cm,
and the outside surface temperature of the cable is 180◦ C. Calculate the maximum
temperature in the cable if the surrounding air temperature is 15◦ C.
2-53 Derive an expression for the temperature distribution in a hollow cylinder with heat
sources that vary according to the linear relation
q̇ = a + br
with q̇i the generation rate per unit volume at r = ri . The inside and outside temper-
atures are T = Ti at r = ri and T = To at r = ro .
2-54 The outside of a copper wire having a diameter of 2 mm is exposed to a convection
environment with h = 5000 W/m2 · ◦ C and T∞ = 100◦ C. What current must be
passed through the wire to produce a center temperature of 150◦ C? Repeat for an
aluminum wire of the same diameter. The resistivity of copper is 1.67 μ · cm.
2-55 A hollow tube having an inside diameter of 2.5 cm and a wall thickness of 0.4 mm
is exposed to an environment at h = 100 W/m2 · ◦ C and T∞ = 40◦ C. What heat-
generation rate in the tube will produce a maximum tube temperature of 250◦ C for
k = 24 W/m · ◦ C?
2-56 Water flows on the inside of a steel pipe with an ID of 2.5 cm. The wall thickness is
2 mm, and the convection coefficient on the inside is 500 W/m2 · ◦ C. The convec-
tion coefficient on the outside is 12 W/m2 · ◦ C. Calculate the overall heat-transfer
coefficient. What is the main determining factor for U ?
2-57 The pipe in Problem 2-56 is covered with a layer of asbestos [k = 0.18 W/m · ◦ C]
while still surrounded by a convection environment with h = 12 W/m2 · ◦ C. Calcu-
late the critical insulation radius. Will the heat transfer be increased or decreased by
adding an insulation thickness of (a) 0.5 mm, (b) 10 mm?
2-58 Calculate the overall heat-transfer coefficient for Problem 2-4.
2-59 Calculate the overall heat-transfer coefficient for Problem 2-5.
2-60 Air flows at 120◦ C in a thin-wall stainless-steel tube with h = 65 W/m2 · ◦ C.
The inside diameter of the tube is 2.5 cm and the wall thickness is 0.4 mm.
k = 18 W/m · ◦ C for the steel. The tube is exposed to an environment with
h = 6.5 W/m2 · ◦ C and T∞ = 15◦ C. Calculate the overall heat-transfer coefficient
and the heat loss per meter of length. What thickness of an insulation having
k = 40 mW/m · ◦ C should be added to reduce the heat loss by 90 percent?
2-61 An insulating glass window is constructed of two 5-mm glass plates separated by an
air layer having a thickness of 4 mm. The air layer may be considered stagnant so
that pure conduction is involved. The convection coefficients for the inner and outer
surfaces are 12 and 50 W/m2 · ◦ C, respectively. Calculate the overall heat-transfer
coefficient for this arrangement, and the R value. Repeat the calculation for a single
glass plate 5 mm thick.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.66 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER2 Steady-State Conduction—One Dimension 67

2-62 A wall consists of a 1-mm layer of copper, a 4-mm layer of 1 percent carbon steel, a
1-cm layer of asbestos sheet, and 10 cm of fiberglass blanket. Calculate the overall
heat-transfer coefficient for this arrangement. If the two outside surfaces are at
10 and 150◦ C, calculate each of the interface temperatures.
2-63 A circumferential fin of rectangular profile has a thickness of 0.7 mm and is installed
on a tube having a diameter of 3 cm that is maintained at a temperature of 200◦ C.
The length of the fin is 2 cm and the fin material is copper. Calculate the heat lost
by the fin to a surrounding convection environment at 100◦ C with a convection
heat-transfer coefficient of 524 W/m2 · ◦ C.
2-64 A thin rod of length L has its two ends connected to two walls which are maintained
at temperatures T1 and T2 , respectively. The rod loses heat to the environment at T∞
by convection. Derive an expression (a) for the temperature distribution in the rod
and (b) for the total heat lost by the rod.
2-65 A rod of length L has one end maintained at temperature T0 and is exposed to an
environment of temperature T∞ . An electrical heating element is placed in the rod
so that heat is generated uniformly along the length at a rate q̇. Derive an expression
(a) for the temperature distribution in the rod and (b) for the total heat transferred
to the environment. Obtain an expression for the value of q̇ that will make the heat
transfer zero at the end that is maintained at T0 .
2-66 One end of a copper rod 30 cm long is firmly connected to a wall that is maintained
at 200◦ C. The other end is firmly connected to a wall that is maintained at 93◦ C.
Air is blown across the rod so that a heat-transfer coefficient of 17 W/m2 · ◦ C is
maintained. The diameter of the rod is 12.5 mm. The temperature of the air is 38◦ C.
What is the net heat lost to the air in watts?
2-67 Verify the temperature distribution for case 2 in Section 2-9, i.e., that

T − T∞ cosh m(L − x) + (h/mk) sinh m(L − x)

=
T0 − T∞ cosh mL + (h/mk) sinh mL

Subsequently show that the heat transfer is

√ sinh mL + (h/mk) cosh mL

q= hPkA (T0 − T∞ )
cosh mL + (h/mk) sinh mL

2-68 An aluminum rod 2.0 cm in diameter and 12 cm long protrudes from a wall that is
maintained at 250◦ C. The rod is exposed to an environment at 15◦ C. The convection
heat-transfer coefficient is 12 W/m2 · ◦ C. Calculate the heat lost by the rod.
2-69 Derive Equation (2-35) by integrating the convection heat loss from the rod of case 1
in Section 2-9.
2-70 Derive Equation (2-36) by integrating the convection heat loss from the rod of case 3
in Section 2-9.
2-71 A long, thin copper rod 5 mm in diameter is exposed to an environment at 20◦ C.
The base temperature of the rod is 120◦ C. The heat-transfer coefficient between the
rod and the environment is 20 W/m2 · ◦ C. Calculate the heat given up by the rod.
2-72 A very long copper rod [k = 372 W/m · ◦ C] 2.5 cm in diameter has one end main-
tained at 90◦ C. The rod is exposed to a fluid whose temperature is 40◦ C. The heat-
transfer coefficient is 3.5 W/m2 · ◦ C. How much heat is lost by the rod?
2-73 An aluminum fin 1.5 mm thick is placed on a circular tube with 2.7-cm OD. The fin
is 6 mm long. The tube wall is maintained at 150◦ C, the environment temperature

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.67 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

68 Problems

is 15◦ C, and the convection heat-transfer coefficient is 20 W/m2 · ◦ C. Calculate the

heat lost by the fin.
2-74 A straight fin of rectangular profile has a thermal conductivity of 14 W/m · ◦ C,
thickness of 2.0 mm, and length of 23 mm. The base of the fin is maintained at a
temperature of 220◦ C while the fin is exposed to a convection environment at 23◦ C
with h = 25 W/m2 · ◦ C. Calculate the heat lost per meter of fin depth.
2-75 A circumferential fin of rectangular profile is constructed of a material having
k = 55 W/m · ◦ C and is installed on a tube having a diameter of 3 cm. The length of fin
is 3 cm and the thickness is 2 mm. If the fin is exposed to a convection environment
at 20◦ C with a convection coefficient of 68 W/m2 · ◦ C and the tube wall temperature
is 100◦ C, calculate the heat lost by the fin.
2-76 The total efficiency for a finned surface may be defined as the ratio of the total
heat transfer of the combined area of the surface and fins to the heat that would be
transferred if this total area were maintained at the base temperature T0 . Show that
this efficiency can be calculated from
Af
ηt = 1 − A (1 − ηf )
where
ηt = total efficiency
Af = surface area of all fins
A = total heat-transfer area, including fins and exposed tube or other surface
ηf = fin efficiency
2-77 A triangular fin of stainless steel (18% Cr, 8% Ni) is attached to a plane wall main-
tained at 460◦ C. The fin thickness is 6.4 mm, and the length is 2.5 cm. The envi-
ronment is at 93◦ C, and the convection heat-transfer coefficient is 28 W/m2 · ◦ C.
Calculate the heat lost from the fin.
2-78 A 2.5-cm-diameter tube has circumferential fins of rectangular profile spaced at
9.5-mm increments along its length. The fins are constructed of aluminum and
are 0.8 mm thick and 12.5 mm long. The tube wall temperature is maintained
at 200◦ C, and the environment temperature is 93◦ C. The heat-transfer
coefficient is 110 W/m2 · ◦ C. Calculate the heat loss from the tube per meter
of length.
2-79 A circumferential fin of rectangular profile surrounds a 2-cm-diameter tube. The
length of the fin is 5 mm, and the thickness is 2.5 mm. The fin is constructed of mild
steel. If air blows over the fin so that a heat-transfer coefficient of 25 W/m2 · ◦ C is
experienced and the temperatures of the base and air are 260 and 93◦ C, respectively,
calculate the heat transfer from the fin.
2-80 A straight rectangular fin 2.0 cm thick and 14 cm long is constructed of steel and
placed on the outside of a wall maintained at 200◦ C. The environment temperature
is 15◦ C, and the heat-transfer coefficient for convection is 20 W/m2 · ◦ C. Calculate
the heat lost from the fin per unit depth.
2-81 An aluminum fin 1.6 mm thick surrounds a tube 2.5 cm in diameter. The length
of the fin is 12.5 mm. The tube-wall temperature is 200◦ C, and the environment
temperature is 20◦ C. The heat-transfer coefficient is 60 W/m2 · ◦ C. What is the heat
lost by the fin?
2-82 Obtain an expression for the optimum thickness of a straight rectangular fin for a
given profile area. Use the simplified insulated-tip solution.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.68 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER2 Steady-State Conduction—One Dimension 69

2-83 Derive a differential equation (do not solve) for the temperature distribution in Figure P2-83
a straight triangular fin. For convenience, take the coordinate axis as shown in
Figure P2-83 and assume one-dimensional heat flow. L
2-84 A circumferential fin of rectangular profile is installed on a 10-cm-diameter tube t
maintained at 120◦ C. The fin has a length of 15 cm and thickness of 2 mm. The fin
is exposed to a convection environment at 23◦ C with h = 60 W/m2 · ◦ C and the fin x

conductivity is 120 W/m · ◦ C. Calculate the heat lost by the fin expressed in watts. x=O
2-85 A long stainless-steel rod [k = 16 W/m · ◦ C] has a square cross section 12.5 by
12.5 mm and has one end maintained at 250◦ C. The heat-transfer coefficient is
40 W/m2 · ◦ C, and the environment temperature is 90◦ C. Calculate the heat lost by
the rod.
2-86 A straight fin of rectangular profile is constructed of duralumin (94% Al, 3% Cu)
with a thickness of 2.1 mm. The fin is 17 mm long, and it is subjected to a convec-
tion environment with h = 75 W/m2 · ◦ C. If the base temperature is 100◦ C and the
environment is at 30◦ C, calculate the heat transfer per unit length of fin.
2-87 A certain internal-combustion engine is air-cooled and has a cylinder constructed
of cast iron [k = 35 Btu/h · ft · ◦ F]. The fins on the cylinder have a length of 58 in
and thickness of 18 in. The convection coefficient is 12 Btu/h · ft 2 · ◦ F. The cylinder
diameter is 4 in. Calculate the heat loss per fin for a base temperature of 450◦ F and
environment temperature of 100◦ F.
2-88 A 1.5-mm-diameter stainless-steel rod [k = 19 W/m · ◦ C] protrudes from a wall
maintained at 45◦ C. The rod is 12 mm long, and the convection coefficient is
500 W/m2 · ◦ C. The environment temperature is 20◦ C. Calculate the temperature of
the tip of the rod. Repeat the calculation for h = 200 and 1500 W/m2 · ◦ C.
2-89 An aluminum block is cast with an array of pin fins protruding like that shown in
Figure 2-10d and subjected to room air at 20◦ C. The convection coefficient between
the pins and the surrounding air may be assumed to be h = 13.2 W/m2 · ◦ C. The pin
diameters are 2 mm and their length is 25 mm. The base of the aluminum block may
be assumed constant at 70◦ C. Calculate the total heat lost by an array of 15 by 15,
that is, 225 fins.
2-90 A finned tube is constructed as shown in Figure 2-10b. Eight fins are installed as
shown and the construction material is aluminum. The base temperature of the fins
may be assumed to be 100◦ C and they are subjected to a convection environment
at 30◦ C with h = 15 W/m2 · ◦ C. The longitudinal length of the fins is 15 cm and
the peripheral length is 2 cm. The fin thickness is 2 mm. Calculate the total heat
dissipated by the finned tube. Consider only the surface area of the fins.
2-91 Circumferential fins of rectangular profile are constructed of aluminum and attached
to a copper tube having a diameter of 25 mm and maintained at 100◦ C. The length
of the fins is 2 cm and thickness is 2 mm. The arrangement is exposed to a convec-
tion environment at 30◦ C with h = 15 W/m2 · ◦ C. Assume that a number of fins is
installed such that the total fin surface area equals that of the total surface fine area in
Problem 2-90. Calculate the total heat lost by the fins.
2-92 A 2-cm-diameter glass rod 6 cm long [k = 0.8 W/m · ◦ C] has a base temperature of
100◦ C and is exposed to an air convection environment at 20◦ C. The temperature
at the tip of the rod is measured as 35◦ C. What is the convection heat-transfer
coefficient? How much heat is lost by the rod?
2-93 A straight rectangular fin has a length of 2.5 cm and a thickness of 1.5 mm. The
thermal conductivity is 55 W/m · ◦ C, and it is exposed to a convection environment

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.69 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

70 Problems

at 20◦ C and h = 500 W/m2 · ◦ C. Calculate the maximum possible heat loss for a
base temperature of 200◦ C. What is the actual heat loss?
2-94 A straight rectangular fin has a length of 3.5 cm and a thickness of 1.4 mm. The
thermal conductivity is 55 W/m · ◦ C. The fin is exposed to a convection environment
at 20◦ C and h = 500 W/m2 · ◦ C. Calculate the maximum possible heat loss for a base
temperature of 150◦ C. What is the actual heat loss for this base temperature?
2-95 A circumferential fin of rectangular profile is constructed of 1 percent carbon steel
and attached to a circular tube maintained at 150◦ C. The diameter of the tube is
5 cm, and the length is also 5 cm with a thickness of 2 mm. The surrounding air
is maintained at 20◦ C and the convection heat-transfer coefficient may be taken as
100 W/m2 · ◦ C. Calculate the heat lost from the fin.
2-96 A circumferential fin of rectangular profile is constructed of aluminum and surrounds
a 3-cm-diameter tube. The fin is 2 cm long and 1 mm thick. The tube wall temperature
is 200◦ C, and the fin is exposed to a fluid at 20◦ C with a convection heat-transfer
coefficient of 80 W/m2 · ◦ C. Calculate the heat loss from the fin.
2-97 A 1.0-cm-diameter steel rod [k = 20 W/m · ◦ C] is 20 cm long. It has one end main-
tained at 50◦ C and the other at 100◦ C. It is exposed to a convection environment at
20◦ C with h = 50 W/m2 · ◦ C. Calculate the temperature at the center of the rod.
2-98 A circumferential fin of rectangular profile is constructed of copper and surrounds a
tube having a diameter of 1.25 cm. The fin length is 6 mm and its thickness is
0.3 mm. The fin is exposed to a convection environment at 20◦ C with h = 55
W/m2 · ◦ C and the fin base temperature is 100◦ C. Calculate the heat lost by the fin.
2-99 A straight rectangular fin of steel (1% C) is 2 cm thick and 17 cm long. It is placed on
the outside of a wall which is maintained at 230◦ C. The surrounding air temperature
is 25◦ C, and the convection heat-transfer coefficient is 23 W/m2 · ◦ C. Calculate the
heat lost from the fin per unit depth and the fin efficiency.
2-100 A straight fin having a triangular profile has a length of 5 cm and a thickness of
4 mm and is constructed of a material having k = 23 W/m · ◦ C. The fin is exposed
to surroundings with a convection coefficient of 20 W/m2 · ◦ C and a temperature of
40◦ C. The base of the fin is maintained at 200◦ C. Calculate the heat lost per unit
depth of fin.
2-101 A circumferential aluminum fin is installed on a 25.4-mm-diameter tube. The length
of the fin is 12.7 mm and the thickness is 1.0 mm. It is exposed to a convection
environment at 30◦ C with a convection coefficient of 56 W/m2 · ◦ C. The base tem-
perature is 125◦ C. Calculate the heat lost by the fin.
2-102 A circumferential fin of rectangular profile is constructed of stainless steel (18% Cr,
8% Ni). The thickness of the fin is 2.0 mm, the inside radius is 2.0 cm, and the length
is 8.0 cm. The base temperature is maintained at 135◦ C and the fin is exposed to a
convection environment at 15◦ C with h = 20 W/m2 · ◦ C. Calculate the heat lost by
the fin.
2-103 A rectangular fin has a length of 2.5 cm and thickness of 1.1 mm. The thermal
conductivity is 55 W/m · ◦ C. The fin is exposed to a convection environment at
20◦ C and h = 500 W/m2 · ◦ C. Calculate the heat loss for a base temperature of
125◦ C.
2-104 A 1.0-mm-thick aluminum fin surrounds a 2.5-cm-diameter tube. The length of
the fin is 1.25 cm. The fin is exposed to a convection environment at 30◦ C with
h = 75 W/m2 · ◦ C. The tube surface is maintained at 100◦ C. Calculate the heat lost
by the fin.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.70 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER2 Steady-State Conduction—One Dimension 71

2-105 A glass rod having a diameter of 1 cm and length of 5 cm is exposed to a convection

environment at a temperature of 20◦ C. One end of the rod is maintained at a tem-
perature of 180◦ C. Calculate the heat lost by the rod if the convection heat-transfer
coefficient is 20 W/m2 · ◦ C.
2-106 A stainless steel rod has a square cross section measuring 1 by 1 cm. The rod length
is 8 cm, and k = 18 W/m · ◦ C. The base temperature of the rod is 300◦ C. The rod
is exposed to a convection environment at 50◦ C with h = 45 W/m2 · ◦ C. Calculate
the heat lost by the rod and the fin efficiency.
2-107 Copper fins with a thickness of 1.0 mm are installed on a 2.5-cm-diameter tube.
The length of each fin is 12 mm. The tube temperature is 275◦ C and the fins are
exposed to air at 35◦ C with a convection heat-transfer coefficient of 120 W/m2 · ◦ C.
Calculate the heat lost by each fin.
2-108 A straight fin of rectangular profile is constructed of stainless steel (18% Cr, 8% Ni)
and has a length of 5 cm and a thickness of 2.5 cm. The base temperature is main-
tained at 100◦ C and the fin is exposed to a convection environment at 20◦ C with
h = 47 W/m2 · ◦ C. Calculate the heat lost by the fin per meter of depth, and the fin
efficiency.
2-109 Acircumferential fin of rectangular profile is constructed of duralumin and surrounds
a 3-cm-diameter tube. The fin is 3 cm long and 1 mm thick. The tube wall temperature
is 200◦ C, and the fin is exposed to a fluid at 20◦ C with a convection heat-transfer
coefficient of 80 W/m2 · ◦ C. Calculate the heat loss from the fin.
2-110 A circular fin of rectangular profile is attached to a 3.0-cm-diameter tube maintained
at 100◦ C. The outside diameter of the fin is 9.0 cm and the fin thickness is 1.0 mm.
The environment has a convection coefficient of 50 W/m2 · ◦ C and a temperature
of 30◦ C. Calculate the thermal conductivity of the material for a fin efficiency of
60 percent.
2-111 A circumferential fin of rectangular profile having a thickness of 1.0 mm and a length
of 2.0 cm is placed on a 2.0-cm-diameter tube. The tube temperature is 150◦ C, the
environment temperature is 20◦ C, and h = 150 W/m2 · ◦ C. The fin is aluminum.
Calculate the heat lost by the fin.
2-112 Two 1-in-diameter bars of stainless steel [k = 17 W/m · ◦ C] are brought into end-
to-end contact so that only 0.1 percent of the cross-sectional area is in contact at
the joint. The bars are 7.5 cm long and subjected to an axial temperature difference
of 300◦ C. The roughness depth in each bar (Lg /2) is estimated to be 1.3 μm. The
surrounding fluid is air, whose thermal conductivity may be taken as 0.035 W/m · ◦ C
for this problem. Estimate the value of the contact resistance and the axial heat flow.
What would the heat flow be for a continuous 15-cm stainless-steel bar?
2-113 When the joint pressure for two surfaces in contact is increased, the high spots of
the surfaces are deformed so that the contact area Ac is increased and the rough-
ness depth Lg is decreased. Discuss this effect in the light of the presentation of
Section 2-11. (Experimental work shows that joint conductance varies almost directly
with pressure.)
2-114 Two aluminum plates 5 mm thick with a ground roughness of 100 μin are bolted
together with a contact pressure of 20 atm. The overall temperature difference across
the plates is 80◦ C. Calculate the temperature drop across the contact joint.
2-115 Fins are frequently installed on tubes by a press-fit process. Consider a circumferen-
tial aluminum fin having a thickness of 1.0 mm to be installed on a 2.5-cm-diameter
aluminum tube. The fin length is 1.25 cm, and the contact conductance may be

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.71 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

72 Problems

taken from Table 2-2 for a 100-μin ground surface. The convection environment is
at 20◦ C, and h = 125 W/m2 · ◦ C. Calculate the heat transfer for each fin for a tube
wall temperature of 200◦ C. What percentage reduction in heat transfer is caused by
the contact conductance?
2-116 An aluminum fin is attached to a transistor that generates heat at the rate of 300 mW.
The fin has a total surface area of 9.0 cm2 and is exposed to surrounding air at 27◦ C.
The contact conductance between transistor and fin is 0.9 × 10−4 m2 · ◦ C/W, and
the contact area is 0.5 cm2 . Estimate the temperature of the transistor, assuming the
fin is uniform in temperature.
2-117 A plane wall 20 cm thick with uniform internal heat generation of 200 kW/m3 is
exposed to a convection environment on both sides at 50◦ C with h = 400 W/m2 · ◦ C.
Calculate the center temperature of the wall for k = 20 W/m · ◦ C.
2-118 Suppose the wall of Problem 2-117 is only 10 cm thick and has one face insulated.
Calculate the maximum temperature in the wall assuming all the other conditions
are the same. Comment on the results.
2-119 A circumferential fin of rectangular profile is constructed of aluminum and placed
on a 6-cm-diameter tube maintained at 120◦ C. The length of the fin is 3 cm and its
thickness is 2 mm. The fin is exposed to a convection environment at 20◦ C with
h = 220 W/m2 · ◦ C. Calculate the heat lost by the fin expressed in Watts.
2-120 A straight aluminum fin of triangular profile has a base maintained at 200◦ C and is
exposed to a convection environment at 25◦ C with h = 45 W/m2 · ◦ C. The fin has
a length of 8 mm and a thickness of 2.0 mm. Calculate the heat lost per unit depth
of fin.
Figure P2-122 2-121 One hundred circumferential aluminum fins of rectangular profile are mounted on
a 1.0-m tube having a diameter of 2.5 cm. The fins are 1 cm long and 2.0 mm thick.
q The base temperature is 180◦ C, and the convection environment is at 20◦ C with
h = 50 W/m2 · ◦ C. Calculate the total heat lost from the finned-tube arrangement
over the 1.0-m length.
2-122 The cylindrical segment shown in Figure P2-122 has a thermal conductivity of
100 W/m · ◦ C. The inner and outer radii are 1.5 and 1.7 cm, respectively, and the
surfaces are insulated. Calculate the circumferential heat transfer per unit depth for
an imposed temperature difference of 50◦ C. What is the thermal resistance?
q 2-123 The truncated hollow cone shown in Figure P2-123 is used in laser-cooling applica-
tions and is constructed of copper with a thickness of 0.5 mm. Calculate the thermal
Figure P2-123 resistance for one-dimensional heat flow. What would be the heat transfer for a
2.0 cm temperature difference of 300◦ C?
2-124 A tube assembly is constructed of copper with an inside diameter of 1.25 cm, wall
thickness of 0.8 mm, and circumferential fins around the periphery. The fins have
a thickness of 0.3 mm and length of 3 mm, and are spaced 6 mm apart. If the
4.0 cm convection heat transfer coefficient from the tube and fins to the surrounding air
is 50 W/m2 · ◦ C, calculate the thermal resistance for a 30-cm length of the tube-
fin combination. What is the fin efficiency for this arrangement? If the inside tube
temperature is 100◦ C and the surrounding air temperature is 20◦ C, what is the heat
loss per meter of tube length? What fraction of the loss is by the fins?
4.0 cm
2-125 Calculate the R value for the fin-tube combination in Problem 2-116.
2-126 Repeat Problem 2-124 for aluminum fins installed on a copper tube.
2-127 Repeat Problem 2-125 for aluminum fins installed on a copper tube.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.72 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER2 Steady-State Conduction—One Dimension 73

2-128 A stainless-steel rod having a length of 10 cm and diameter of 2 mm has a resistivity

of 70 μ · cm and thermal conductivity of 16 W/m · ◦ C. The rod is exposed to a
convection environment with h = 100 W/m2 · ◦ C and T = 20◦ C. Both ends of the
rod are maintained at T = 100◦ C. What voltage must be impressed on the rod to
dissipate twice as much heat to the surroundings as in a zero-voltage condition?
2-129 Suppose the rod in Problem 2-128 is very long. What would the zero-voltage heat
transfer be in this case?
2-130 Suppose the cylindrical segment of Problem 2-122 has a periphery exposed to a
convection environment with h = 75 W/m2 · ◦ C and T∞ = 30◦ C instead of to the
insulated surface. For this case, one end is at 50◦ C while the other end is at 100◦ C.
What is the heat lost by the segment to the surroundings in this circumstance? What
is the heat transfer at each end of the segment?

Design-Oriented Problems
2-131 Suppose you have a choice between a straight triangular or rectangular fin con-
structed of aluminum with a base thickness of 3.0 mm. The convection coefficient
is 50 W/m2 · ◦ C. Select the fin with the least weight for a given heat flow.
2-132 Consider aluminum circumferential fins with r1 = 1.0 cm, r2 = 2.0 cm, and thick-
nesses of 1.0, 2.0, and 3.0 mm. The convection coefficient is 160 W/m2 · ◦ C. Com-
pare the heat transfers for six 1.0-mm fins, three 2.0-mm fins, and two 3.0-mm fins.
What do you conclude? Repeat for h = 320 W/m2 · ◦ C.
2-133 “Pin fins” of aluminum are to be compared in terms of their relative performance
as a function of diameter. Three “pins” having diameters of 2, 5, and 10 mm with
a length of 5 cm are exposed to a convection environment with T∞ = 20◦ C, and
h = 40 W/m2 · ◦ C. The base temperature is 200◦ C. Calculate the heat transfer for
each pin. How does it vary with pin diameter?
2-134 Calculate the heat transfer per unit mass for the pin fins in Problem 2-133. How
does it vary with diameter?
2-135 A straight rectangular fin has a length of 1.5 cm and a thickness of 1.0 mm. The con-
vection coefficient is 20 W/m2 · ◦ C. Compare the heat-transfer rates for aluminum
and magnesium fins.
2-136 Suppose both fins in Problem 2-129 are to dissipate the same heat. Which would be
lower in weight? Assume that the thickness is the same for both fins but adjust the
lengths until the heat transfers are equal.
2-137 Insulating materials are frequently installed with a reflective coating to reduce the
radiation heat transfer between the surface and the surroundings. An insulating mate-
rial is installed on a furnace oven wall that is maintained at 200◦ C. The energy cost
of the fuel firing the oven is $8.25/GJ and the insulation installation must be justified
by the savings in energy costs over a three-year period. Select an appropriate insula-
tion from Table 2-1 and/or Table A-3 and determine a suitable quantity of insulation
that will pay for itself over a three-year period. For this computation assume that the
outer surface of the insulation radiates like a blackbody and that the heat loss can
be determined from Equation (1-12). For the calculation use Table 1-2 as a guide
for selecting the convection heat-transfer coefficient. Next, consider the same type
of insulating material but with a reflective coating having  = 0.1. The radiation
transfer may still be calculated with Equation (1-12). Determine the quantity of the

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.73 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

74 Problems

reflective insulating material required to be economical. How much higher cost per
unit thickness or volume could be justified for the reflective material over that of
the nonreflective? Comment on uncertainties which may exist in your analysis.
2-138 A thin-wall stainless-steel tube is to be used as an electric heating element that
will deliver a convection coefficient of 5000 W/m2 · ◦ C to water at 100◦ C. Devise
several configurations to accomplish a total heat transfer of 10 kW. Specify the
length, outside diameter, wall thickness, maximum tube temperature, and necessary
voltage that must be imposed on the tube. Take the resistivity of stainless steel as
70 μ · cm.
2-139 Thin cylindrical or spherical shells may be treated as a plane wall for sufficiently large
diameters in relation to the thickness of the shell. Devise a scheme for quantifying
the error that would result from such a treatment.
2-140 A 2.5-cm-diameter steel pipe is maintained at 100◦ C by condensing steam on the
inside. The pipe is to be used for dissipating heat to a surrounding room at 20◦ C by
placing circular steel fins around the outside surface of the pipe. The convection loss
from the pipe and fins occurs by free convection, with h = 8.0 W/m2 · ◦ C. Examine
several cases of fin thickness, fin spacing, and fin outside diameters to determine the
overall heat loss per meter of pipe length. Take k = 43 W/m · ◦ C for the steel fins
and assume h is uniform over all surfaces. Make appropriate conclusions about the
results of your study.
2-141 A pipe having a diameter of 5.3 cm is maintained at 200◦ C by steam flowing inside.
The pipe passes through a large factory area and loses heat by free convection
from the outside with h = 7.2 W/m2 · ◦ C. Using information from Table 2-1 and/or
Table A-3, select two alternative insulating materials that could be installed to lower
the outside surface temperature of the insulation to 30◦ C when the pipe is exposed
to room air at 20◦ C. If the energy loss from the steam costs $8.00/109 J, what are
the allowable costs of the insulation materials per unit volume to achieve a payback
period of three years where
(energy cost saved per year) × 3
= (cost of installed insulation/unit volume) × volume
2-142 It is frequently represented that the energy savings resulting from installation of extra
ceiling insulation in a home will pay for the insulation cost within a three-year period.
You are asked to evaluate this claim. For the evaluation it may be assumed that 1 kW
of electrical input to an air-conditioning unit will produce about 1.26 × 104 kJ/h of
cooling and that electricity is priced at $0.085/kWh. Assume that an existing home
has ceiling insulation with an R value of 7.0◦ F · ft 2 · h/Btu and is to be upgraded
to an R value of either 15 or 30. Choose two alternative insulation materials from
Table 2-1 and/or Table A-3 and calculate the allowable costs per unit volume of
insulating material to accomplish the three-year payback with the two specified R
values. For this calculation, (energy cost saved/year) × 3 = (insulation cost per unit
volume) × volume. Make your own assumptions regarding (1) temperature differ-
ence between the interior of the house and the attic area and (2) the hours of operation
for the air-conditioning system during an annual period. Comment on the results and
assumptions.
2-143 A finned wall like that shown in Figure 2-10a is constructed of aluminum alloy with
k = 160 W/m · ◦ C. The wall thickness is 2.0 mm and the fins are straight with rectan-
gular profile. The inside of the wall is maintained at a constant temperature of 70◦ C
and the fins are exposed to a convection environment at 25◦ C with h = 8 W/m2 · ◦ C
(free convection). The assembly will be cast from the aluminum material and must

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.74 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

CHAPTER2 Steady-State Conduction—One Dimension 75

dissipate 30 W of heat under the conditions noted. Assuming a square array, deter-
mine suitable combinations of numbers of fins, fin spacing, dimension of the square,
and fin thickness to accomplish this cooling objective. Assume a uniform value of
h for both the fin and wall surfaces.
2-144 Repeat Problem 2-143 for cooling with forced convection, which produces a con-
vection coefficient of h = 20 W/m2 · ◦ C.
2-145 Consider a pin fin as shown in Figure 2-10d. Assume that the fin is exposed to an
evacuated space such that convection is negligible and that the radiation loss per
unit surface area is given by
qrad /A = σ(T 4 − Ts4 )
where  is a surface emissivity constant, σ is the Stefan-Boltzmann constant, and
the temperatures are expressed in degrees Kelvin. Derive a differential equation for
the temperature in the pin fin as a function of x, the distance from the base. Let
T0 be the base temperature, and write the appropriate boundary conditions for the
differential equation.
2-146 Consider two special cases for the fin in Problem 2-145: (a) an insulated-tip fin
losing heat by radiation and (b) a very long fin losing heat by radiation. Write the
appropriate boundary conditions for these two cases.
2-147 Consider another special case for the fin of Problem 2-145; where the surrounding
radiation boundary temperature is negligible, that is,
Ts4
T 4
Write the resulting simplified differential equation under this condition.

REFERENCES
1. Schneider, P. J. Conduction Heat Transfer, Reading, Mass.: Addison-Wesley Publishing
Company, 1955.
2. Harper, W. B., and D. R. Brown. “Mathematical Equations for Heat Conduction in the Fins of
Air-cooled Engines,” NACA Rep. 158, 1922.
3. Gardner, K. A. “Efficiency of Extended Surfaces,” Trans. ASME, vol. 67, pp. 621–31, 1945.
4. Moore, C. J. “Heat Transfer across Surfaces in Contact: Studies of Transients in One-dimensional
Composite Systems,” Southern Methodist Univ., Thermal/Fluid Sci. Ctr. Res. Rep. 67-2, Dallas,
Tex., March 1967.
5. Ybarrondo, L. J., and J. E. Sunderland. “Heat Transfer from Extended Surfaces,” Bull. Mech.
Eng. Educ., vol. 5, pp. 229–34, 1966.
6. Moore, C. J., Jr., H. A. Blum, and H. Atkins. “Subject Classification Bibliography for Thermal
Contact Resistance Studies,” ASME Pap. 68-WA/HT-18, December 1968.
7. Clausing, A. M. “Transfer at the Interface of Dissimilar Metals: The Influence of Thermal Strain,”
Int. J. Heat Mass Transfer, vol. 9, p. 791, 1966.
8. Kern, D. Q., and A. D. Kraus. Extended Surface Heat Transfer. New York: McGraw-Hill, 1972.
9. Siegel, R., and J. R. Howell. Thermal Radiation Heat Transfer. 2d ed., New York: McGraw-Hill,
1980.
10. Fried, E. “Thermal Conduction Contribution to Heat Transfer at Contacts,” Thermal Conductivity,
(R. P. Tye, Ed.) vol. 2, New York: Academic Press, 1969.
11. Fletcher, L. S. “Recent Developments in Contact Conductance Heat Transfer,” J. Heat Transfer,
vol. 110, no. 4(B), p. 1059, Nov. 1988.
12. James, T. B., and W. P. Goss. Heat Transmission Coefficients for Walls, Roofs, Ceilings, and
Floors. Altanta: American Society of Heating, Refrigeration, and Air-Conditioning Engineers,
1993.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.75 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch02 10/13/2008 18:58

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.76 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch03 10/29/2008 15:34

C H A P T E R

3 Steady-State Conduction—
Multiple Dimensions

3-1 INTRODUCTION
In Chapter 2 steady-state heat transfer was calculated in systems in which the temperature
gradient and area could be expressed in terms of one space coordinate. We now wish to
analyze the more general case of two-dimensional heat flow. For steady state with no heat
generation, the Laplace equation applies.
∂2 T ∂2 T
+ 2 =0 [3-1]
∂x2 ∂y

assuming constant thermal conductivity. The solution to this equation may be obtained by
analytical, numerical, or graphical techniques.
The objective of any heat-transfer analysis is usually to predict heat flow or the tem-
perature that results from a certain heat flow. The solution to Equation (3-1) will give the
temperature in a two-dimensional body as a function of the two independent space coor-
dinates x and y. Then the heat flow in the x and y directions may be calculated from the
Fourier equations
∂T
qx = −kAx [3-2]
∂x
∂T
qy = −kAy [3-3]
∂y

These heat-flow quantities are directed either in the x direction or in the y direction. The
total heat flow at any point in the material is the resultant of the qx and qy at that point.
Thus the total heat-flow vector is directed so that it is perpendicular to the lines of constant
temperature in the material, as shown in Figure 3-1. So if the temperature distribution in
the material is known, we may easily establish the heat flow.

3-2 MATHEMATICAL ANALYSIS OF

TWO-DIMENSIONAL HEAT CONDUCTION
We first consider an analytical approach to a two-dimensional problem and then indicate the
numerical and graphical methods that may be used to advantage in many other problems.

77

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.77 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

78 3-2 Mathematical Analysis of Two-Dimensional Heat Conduction

It is worthwhile to mention here that analytical solutions are not always possible to obtain;
indeed, in many instances they are very cumbersome and difficult to use. In these cases
numerical techniques are frequently used to advantage. For a more extensive treatment of
the analytical methods used in conduction problems, the reader may consult References 1,
2, 10, and 11.
Consider the rectangular plate shown in Figure 3-2. Three sides of the plate are main-
tained at the constant temperature T1 , and the upper side has some temperature distribution
impressed upon it. This distribution could be simply a constant temperature or something
more complex, such as a sine-wave distribution. We shall consider both cases.
To solve Equation (3-1), the separation-of-variables method is used. The essential
point of this method is that the solution to the differential equation is assumed to take a
product form
T = XY where X = X(x)
[3-4]
Y = Y(y)

The boundary conditions are then applied to determine the form of the functions X and Y .
The basic assumption as given by Equation (3-4) can be justified only if it is possible to
find a solution of this form that satisfies the boundary conditions.

Figure 3-1 Sketch showing the heat flow in

two dimensions.
qy
+
qy = ⫺kA y ∂T qx
∂y q
=
Iso
the
rm

qx = ⫺kA x ∂T
∂x

Figure 3-2 Isotherms and heat flow lines in a

rectangular plate.

y
T=f(x)

T=T1 H

T=T1

T=T1 x
W

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.78 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 79

First consider the boundary conditions with a sine-wave temperature distribution

impressed on the upper edge of the plate. Thus
T = T1 at y = 0
T = T1 at x = 0
T = T1 at x = W [3-5]
 πx 
T = Tm sin + T1 at y = H
W
where Tm is the amplitude of the sine function. Substituting Equation (3-4) in (3-1) gives

1 d2X 1 d2Y
− = [3-6]
X dx2 Y dy2
Observe that each side of Equation (3-6) is independent of the other because x and y are
independent variables. This requires that each side be equal to some constant. We may thus
obtain two ordinary differential equations in terms of this constant,
d2X
+ λ2 X = 0 [3-7]
dx2
d2Y
− λ2 Y = 0 [3-8]
dy2
where λ2 is called the separation constant. Its value must be determined from the boundary
conditions. Note that the form of the solution to Equations (3-7) and (3-8) will depend on
the sign of λ2 ; a different form would also result if λ2 were zero. The only way that the
correct form can be determined is through an application of the boundary conditions of
the problem. So we shall first write down all possible solutions and then see which one fits
the problem under consideration.
For λ2 = 0:
X = C1 + C2 x
Y = C3 + C4 y [3-9]
T = (C1 + C2 x)(C3 + C4 y)
This function cannot fit the sine-function boundary condition, so the λ2 = 0 solution may
be excluded.
For λ2 < 0:
X = C5 e−λx + C6 eλx
Y = C7 cos λy + C8 sin λy [3-10]
T = (C5 e−λx + C6 eλx )(C7 cos λy + C8 sin λy)
Again, the sine-function boundary condition cannot be satisfied, so this solution is excluded
also.
For λ2 > 0:
X = C9 cos λx + C10 sin λx
Y = C11 e−λy + C12 eλy [3-11]
T = (C9 cos λx + C10 sin λx)(C11 e−λy + C12 eλy )
Now, it is possible to satisfy the sine-function boundary condition; so we shall attempt to
satisfy the other conditions. The algebra is somewhat easier to handle when the substitution
θ = T − T1

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.79 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch03 10/29/2008 15:34

80 3-2 Mathematical Analysis of Two-Dimensional Heat Conduction

is made. The differential equation and the solution then retain the same form in the new
variable θ, and we need only transform the boundary conditions. Thus
θ=0 at y = 0
θ=0 at x = 0 [3-12]
θ=0 at x = W
πx
θ = Tm sin at y = H
W
Applying these conditions, we have
0 = (C9 cos λx + C10 sin λx)(C11 + C12 ) [a]

0 = C9 (C11 e−λy + C12 eλy ) [b]

0 = (C9 cos λW + C10 sin λW)(C11 e−λy + C12 eλy ) [c]

πx
Tm sin = (C9 cos λx + C10 sin λx)(C11 e−λH + C12 eλH ) [d]
W
Accordingly,
C11 = −C12
C9 = 0

and from (c),

0 = C10 C12 sin λW(eλy − e−λy )

This requires that

sin λW = 0 [3-13]

Recall that λ was an undetermined separation constant. Several values will satisfy Equation
(3-13), and these may be written
nπ
λ= [3-14]
W
where n is an integer. The solution to the differential equation may thus be written as a sum
of the solutions for each value of n. This is an infinite sum, so that the final solution is the
infinite series
∞
nπx nπy
θ = T − T1 = Cn sin sinh [3-15]
W W
n=1

where the constants have been combined and the exponential terms converted to the hyper-
bolic function. The final boundary condition may now be applied:
∞
πx  nπx nπH
Tm sin = Cn sin sinh
W W W
n=1

which requires that Cn = 0 for n > 1. The final solution is therefore

sinh(πy/W)  πx 
T = Tm sin + T1 [3-16]
sinh(πH/W) W

The temperature field for this problem is shown in Figure 3-2. Note that the heat-flow lines
are perpendicular to the isotherms.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.80 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 81

We now consider the set of boundary conditions

T = T1 at y = 0
T = T1 at x = 0
T = T1 at x = W
T = T2 at y = H

Using the first three boundary conditions, we obtain the solution in the form of Equation
(3-15):
∞
nπx nπy
T − T1 = Cn sin sinh [3-17]
W W
n=1

Applying the fourth boundary condition gives

∞
 nπx nπH
T2 − T1 = Cn sin sinh [3-18]
W W
n=1

This is a Fourier sine series, and the values of the Cn may be determined by expanding
the constant temperature difference T2 − T1 in a Fourier series over the interval 0 < x < W.
This series is
∞
2  (−1)n+1 + 1 nπx
T2 − T1 = (T2 − T1 ) sin [3-19]
π n W
n=1

Upon comparison of Equation (3-18) with Equation (3-19), we find that

2 1 (−1)n+1 + 1
Cn = (T2 − T1 )
π sinh(nπH/W) n

and the final solution is expressed as

∞
T − T1 2  (−1)n+1 + 1 nπx sinh(nπy/W)
= sin [3-20]
T2 − T1 π n W sinh(nπH/W)
n=1

An extensive study of analytical techniques used in conduction heat transfer requires

a background in the theory of orthogonal functions. Fourier series are one example of
orthogonal functions, as are Bessel functions and other special functions applicable to
different geometries and boundary conditions. The interested reader may consult one or
more of the conduction heat-transfer texts listed in the references for further information
on the subject.

3-3 GRAPHICAL ANALYSIS

Consider the two-dimensional system shown in Figure 3-3. The inside surface is maintained
at some temperature T1 , and the outer surface is maintained at T2 . We wish to calculate the
heat transfer. Isotherms and heat-flow lanes have been sketched to aid in this calculation.
The isotherms and heat-flow lanes form groupings of curvilinear figures like that shown in
Figure 3-3b. The heat flow across this curvilinear section is given by Fourier’s law, assuming
unit depth of material:
T
q = −k x(1) [3-21]
y

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.81 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

82 3-3 Graphical Analysis

Figure 3-3 Sketch showing element used for curvilinear-square

analysis of two-dimensional heat flow.

(a)

q
Δy

q
Δx
ΔT

(b)

This heat flow will be the same through each section within this heat-flow lane, and the total
heat flow will be the sum of the heat flows through all the lanes. If the sketch is drawn so
that x ∼
= y, the heat flow is proportional to the T across the element and, since this heat
flow is constant, the T across each element must be the same within the same heat-flow
lane. Thus the T across an element is given by
Toverall
T =
N

where N is the number of temperature increments between the inner and outer surfaces.
Furthermore, the heat flow through each lane is the same since it is independent of the
dimensions x and y when they are constructed equal. Thus we write for the total heat
transfer
M M
q= k Toverall = k(T2 − T1 ) [3-22]
N N
where M is the number of heat-flow lanes. So, to calculate the heat transfer, we need only
construct these curvilinear-square plots and count the number of temperature increments
and heat-flow lanes. Care must be taken to construct the plot so that x ≈ y and the lines
are perpendicular. For the corner section shown in Figure 3-3a the number of temperature
increments between the inner and outer surfaces is about N = 4, while the number of heat-
flow lanes for the corner section may be estimated as M = 8.2. The total number of heat-flow
lanes is four times this value, or 4 × 8.2 = 32.8. The ratio M/N is thus 32.8/4 = 8.2 for the
whole wall section. This ratio will be called the conduction shape factor in subsequent
discussions.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.82 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 83

The accuracy of this method is dependent entirely on the skill of the person sketching
the curvilinear squares. Even a crude sketch, however, can frequently help to give fairly
good estimates of the temperatures that will occur in a body. An electrical analogy may be
employed to sketch the curvilinear squares, as discussed in Section 3-9.
The graphical method presented here is mainly of historical interest to show the relation
of heat-flow lanes and isotherms. It may not be expected to be used for the solution of many
practical problems.

3-4 THE CONDUCTION SHAPE FACTOR

In a two-dimensional system where only two temperature limits are involved, we may define
a conduction shape factor S such that
q = kS Toverall [3-23]

The values of S have been worked out for several geometries and are summarized in
Table 3-1. A very comprehensive summary of shape factors for a large variety of geometries
is given by Rohsenow [15] and Hahne and Grigull [17]. Note that the inverse hyperbolic
cosine can be calculated from

cosh−1 x = ln(x ± x2 − 1)

For a three-dimensional wall, as in a furnace, separate shape factors are used to calculate
the heat flow through the edge and corner sections, with the dimensions shown in Figure 3-4.
When all the interior dimensions are greater than one-fifth of the wall thickness,
A
Swall = Sedge = 0.54D Scorner = 0.15L
L
where
A = area of wall
L = wall thickness
D = length of edge
Note that the shape factor per unit depth is given by the ratio M/N when the curvilinear-
squares method is used for calculations.

Figure 3-4 Sketch illustrating dimensions for use in calculating

three-dimensional shape factors.

L
D L

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.83 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch03 10/29/2008 15:34

84 3-4 The Conduction Shape Factor

Table 3-1 Conduction shape factors, summarized from References 6 and 7.

Note: For buried objects the temperature difference is T = Tobject − Tfar field . The far-field
temperature is taken the same as the isothermal surface temperature for semi-infinite media.

Physical system Schematic Shape factor Restrictions

Isothermal cylinder of radius r 2πL Lr

Isothermal
buried in semi-infinite medium cosh−1 (D/r)
having isothermal surface
2πL Lr
D
ln(D/r) D > 3r
r

Isothermal sphere of radius r 4πr

buried in infinite medium
r

Isothermal sphere of radius r Isothermal 4πr

buried in semi-infinite medium 1 − r/2D
having isothermal surface
T = Tsurf − Tfar field D
r

Conduction between two r1 r2 2πL Lr

 2 
isothermal cylinders of length L D −r12 −r22 LD
cosh−1 2r1 r2
buried in infinite medium
D

Row of horizontal cylinders Isothermal 2πL D > 2r

S=   
of length L in semi-infinite ln 1 sinh(2πD/l)
πr
medium with isothermal
surface
D

Buried cube in infinite 8.24L

medium, L on a side

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.84 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch03 10/29/2008 15:34

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 85

Table 3-1 (Continued).

Physical system Schematic Shape factor Restrictions

Isothermal cylinder Isothermal

2πL L  2r
of radius r placed in ln(2L/r)
semi-infinite medium
as shown L

2r

Isothermal   −0.59  −0.078

1.685L log 1 + ba b
Isothermal rectangular See Reference 7
parallelepiped buried in c
semi-infinite medium having b
isothermal surface

c
L
a

Plane wall A One-dimensional heat flow

L
A

Hollow cylinder, length L ri

2πL Lr
+ ro ln(ro /ri )

Hollow sphere 4πro ri

ri
ro − ri
+ ro

Thin horizontal disk buried Isothermal 4r D=0

in semi-infinite medium 2r D 8r D  2r
with isothermal surface 4πr D/2r > 1
π/2 − tan−1 (r/2D) tan−1 (r/2D) in radians
Hemisphere buried in Isothermal 2πr
semi-infinite medium +
T = Tsphere − Tfar field r

Isothermal sphere buried in Insulated 4πr

semi-infinite medium with 1 + r/2D
D
insulated surface +
T∞ r

Two isothermal spheres r1 

4πr2
D > 5rmax
r2
buried in infinite medium r2 (r1 /D)4 2r
+ + r1 1− − D2
1−(r2 /D)2
D

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.85 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 11/6/2008 9:46

86 3-4 The Conduction Shape Factor

Table 3-1 (Continued).

Physical system Schematic Shape factor Restrictions

Thin rectangular plate of πW D=0

Isothermal
length L, buried in semi-infinite ln(4 W/L) W>L
medium having isothermal D
L 2π W D
W
surface
W ln(4 W/L) W>L
2π W W
L
ln(2π D/L) D > 2W

Parallel disks buried in
π

4πr
 D > 5r
t1 t2
infinite medium r 2 − tan −1 (r/D) tan−1 (r/D)
in radians

Eccentric cylinders of length L 2πL L
r2

r2  2 
r1 +r22 −D2
r1 cosh−1 2r1 r2
+ +
D

Cylinder centered in a square 2πL L
W

of length L ln(0.54 W/r)
L

r
W +

Horizontal cylinder of length L Isothermal 2πL

centered in infinite plate ln(4D/r)
D
r
+

Isothermal

Thin horizontal disk buried in Insulated 4πr D/2r > 1

semi-infinite medium with π/2 + tan−1 (r/2D) tan−1 (r/2D)
adiabatic surface in radians
D
T = Tdisk − Tfar field

2r

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.86 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 87

Buried Pipe EXAMPLE 3-1

A horizontal pipe 15 cm in diameter and 4 m long is buried in the earth at a depth of 20 cm.
The pipe-wall temperature is 75◦ C, and the earth surface temperature is 5◦ C. Assuming that the
thermal conductivity of the earth is 0.8 W/m · ◦ C, calculate the heat lost by the pipe.
Solution
We may calculate the shape factor for this situation using the equation given in Table 3-1. Since
D < 3r,
2πL 2π(4)
S= = = 15.35 m
cosh−1(D/r) cosh−1(20/7.5)
The heat flow is calculated from
q = kST = (0.8)(15.35)(75 − 5) = 859.6 W [2933 Btu/h]

Cubical Furnace EXAMPLE 3-2

A small cubical furnace 50 by 50 by 50 cm on the inside is constructed of fireclay brick
[k = 1.04 W/m · ◦ C] with a wall thickness of 10 cm. The inside of the furnace is maintained
at 500◦ C, and the outside is maintained at 50◦ C. Calculate the heat lost through the walls.
Solution
We compute the total shape factor by adding the shape factors for the walls, edges, and corners:
Walls: A (0.5)(0.5)
S= = = 2.5 m
L 0.1
Edges: S = 0.54D = (0.54)(0.5) = 0.27 m

Corners: S = 0.15L = (0.15)(0.1) = 0.015 m

There are six wall sections, twelve edges, and eight corners, so that the total shape factor is
S = (6)(2.5) + (12)(0.27) + (8)(0.015) = 18.36 m

and the heat flow is calculated as

q = kST = (1.04)(18.36)(500 − 50) = 8.592 kW [29.320 Btu/h]

Buried Disk EXAMPLE 3-3

A disk having a diameter of 30 cm and maintained at a temperature of 95◦ C is buried at a depth of
1.0 m in a semi-infinite medium having an isothermal surface temperature of 20◦ C and a thermal
conductivity of 2.1 W/m · ◦ C. Calculate the heat lost by the disk.
Solution
This is an application of the conduction shape factor relation q = kS T . Consulting Table 3-1 we
find a choice of three relations for S for the geometry of a disk buried in a semi-infinite medium
with an isothermal surface. Clearly, D  = 0 and D is not large compared to 2r, so the relation we
select for the shape factor is for the case D/2r > 1.0:

4πr
S=
[π/2 − tan−1 (r/2D)]

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.87 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

88 3-5 Numerical Method of Analysis

Note that this relation differs from the one for an insulated surface by the minus sign in the
denominator. Inserting r = 0.15 m and D = 1.0 m we obtain
4π(0.15) 4π(0.15)
S= −1
= = 1.26 m
[π/2 − tan (0.15/2)] [π/2 − 0.07486]

For buried objects the shape factor is based on T = Tobject − Tfar field . The far-field temperature
is taken as the isothermal surface temperature, and the heat lost by the disk is therefore
q = kST = (2.1)(1.26)(95 − 20) = 198.45 W

EXAMPLE 3-4 Buried Parallel Disks

Two parallel 50-cm-diameter disks are separated by a distance of 1.5 m in an infinite medium
having k = 2.3 W/m · ◦ C. One disk is maintained at 80◦ C and the other at 20◦ C. Calculate the
heat transfer between the disks.
Solution
This is a shape-factor problem and the heat transfer may be calculated from
q = kST

where S is obtained from Table 3-1 as

4πr
S= for D > 5r
[π/2 − tan−1 (r/D)]

With r = 0.25 m and D = 1.5 m we obtain

4π(0.25) 4π(0.25)
S= −1
= = 2.235
[π/2 − tan (0.25/1.5)] [π/2 − 0.1651]

and
q = kS T = (2.3)(2.235)(80 − 20) = 308.4 W

3-5 NUMERICAL METHOD OF ANALYSIS

Figure 3-5 Sketch An immense number of analytical solutions for conduction heat-transfer problems have
illustrating nomenclature used been accumulated in the literature over the past 150 years. Even so, in many practical
in two-dimensional numerical situations the geometry or boundary conditions are such that an analytical solution has not
analysis of heat conduction.
been obtained at all, or if the solution has been developed, it involves such a complex series
solution that numerical evaluation becomes exceedingly difficult. For such situations the
m, n + 1 most fruitful approach to the problem is one based on finite-difference techniques, the basic
principles of which we shall outline in this section.
Δy Consider a two-dimensional body that is to be divided into equal increments in both
m − 1, n m, n m + 1, n the x and y directions, as shown in Figure 3-5. The nodal points are designated as shown,
the m locations indicating the x increment and the n locations indicating the y increment.
Δy
We wish to establish the temperatures at any of these nodal points within the body, using
m, n − 1
Equation (3-1) as a governing condition. Finite differences are used to approximate differ-
Δx Δx ential increments in the temperature and space coordinates; and the smaller we choose these
finite increments, the more closely the true temperature distribution will be approximated.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.88 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 89

The temperature gradients may be written as follows:


∂T Tm+1,n − Tm,n
≈
∂x m+1/2,n x

∂T Tm,n − Tm−1,n
≈
∂x m−1/2,n x

∂T Tm,n+1 − Tm,n
≈
∂y m,n+1/2 y

∂T Tm,n − Tm,n−1
≈
∂y m,n−1/2 y
 
∂T ∂T
 −
∂2 T ∂x m+1/2,n ∂x m−1/2,n Tm+1,n + Tm−1,n − 2Tm,n
≈ =
∂x m,n
2 x (x)2
 
∂T ∂T
 −
∂ T
2 ∂y m,n+1/2 ∂y m,n−1/2 Tm,n+1 + Tm,n−1 − 2Tm,n
≈ =
∂y2 m,n y (y)2
Thus the finite-difference approximation for Equation (3-1) becomes
Tm+1,n + Tm−1,n − 2Tm,n Tm,n+1 + Tm,n−1 − 2Tm,n
+ =0
(x)2 (y)2

If x = y, then
Tm+1,n + Tm−1,n + Tm,n+1 + Tm,n−1 − 4Tm,n = 0 [3-24]

Since we are considering the case of constant thermal conductivity, the heat flows may all
be expressed in terms of temperature differentials. Equation (3-24) states very simply that
the net heat flow into any node is zero at steady-state conditions. In effect, the numerical
finite-difference approach replaces the continuous temperature distribution by fictitious
heat-conducting rods connected between small nodal points that do not generate heat.
We can also devise a finite-difference scheme to take heat generation into account. We
merely add the term q̇/k into the general equation and obtain
Tm+1,n + Tm−1,n − 2Tm,n Tm,n+1 + Tm,n−1 − 2Tm,n q̇
+ + =0
(x)2 (y)2 k

Then for a square grid in which x = y,

q̇(x)2
Tm+1,n + Tm−1,n + Tm,n+1 + Tm,n−1 + − 4Tm,n−1 = 0 [3-24a]
k
To utilize the numerical method, Equation (3-24) must be written for each node within the
material and the resultant system of equations solved for the temperatures at the various
nodes. A very simple example is shown in Figure 3-6, and the four equations for nodes 1,
2, 3, and 4 would be
100 + 500 + T2 + T3 − 4T1 = 0
T1 + 500 + 100 + T4 − 4T2 = 0

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.89 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

90 3-5 Numerical Method of Analysis

Figure 3-6 Four-node problem.

T = 500˚C

T = 100˚C
T = 100˚C
1 2

3 4

T = 100˚C

100 + T1 + T4 + 100 − 4T3 = 0

T3 + T2 + 100 + 100 − 4T4 = 0

These equations have the solution

T1 = T2 = 250◦ C T3 = T4 = 150◦ C

Of course, we could recognize from symmetry that T1 = T2 and T3 = T4 and would then
only need two nodal equations,
100 + 500 + T3 − 3T1 = 0
100 + T1 + 100 − 3T3 = 0

Once the temperatures are determined, the heat flow may be calculated from
 T
q= k x
y

where the T is taken at the boundaries. In the example the heat flow may be calculated
at either the 500◦ C face or the three 100◦ C faces. If a sufficiently fine grid is used, the two
values should be very nearly the same. As a matter of general practice, it is usually best to
take the arithmetic average of the two values for use in the calculations. In the example, the
two calculations yield:
500◦ C face:
x
q = −k [(250 − 500) + (250 − 500)] = 500k
y
100◦ C face:
y
q = −k [(250 − 100) + (150 − 100) + (150 − 100) + (150 − 100)
x
+ (150 − 100) + (250 − 100)] = −500k

and the two values agree in this case. The calculation of the heat flow in cases in which
curved boundaries or complicated shapes are involved is treated in References 2, 3, and 15.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.90 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch03 10/29/2008 15:34

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 91

When the solid is exposed to some convection boundary condition, the temperatures
at the surface must be computed differently from the method given above. Consider the
boundary shown in Figure 3-7. The energy balance on node (m, n) is
Tm,n − Tm−1,n x Tm,n − Tm,n+1 x Tm,n − Tm,n−1
−k y −k −k
x 2 y 2 y
= h y(Tm,n − T∞ )

If x = y, the boundary temperature is expressed in the equation

 
h x h x 1
Tm,n +2 − T∞ − (2Tm−1,n + Tm,n+1 + Tm,n−1 ) = 0 [3-25]
k k 2

An equation of this type must be written for each node along the surface shown in
Figure 3-7. So when a convection boundary condition is present, an equation like (3-25) is
used at the boundary and an equation like (3-24) is used for the interior points.
Equation (3-25) applies to a plane surface exposed to a convection boundary condition.
It will not apply for other situations, such as an insulated wall or a corner exposed to
a convection boundary condition. Consider the corner section shown in Figure 3-8. The
energy balance for the corner section is
y Tm,n − Tm−1,n x Tm,n − Tm,n−1 x y
−k −k =h (Tm,n − T∞ ) + h (Tm,n − T∞ )
2 x 2 y 2 2
If x = y,
 
h x h x
2Tm,n +1 −2 T∞ − (Tm−1,n + Tm,n−1 ) = 0 [3-26]
k k

Other boundary conditions may be treated in a similar fashion, and a convenient sum-
mary of nodal equations is given in Table 3-2 for different geometrical and boundary
situations. Situations f and g are of particular interest since they provide the calcula-
tion equations that may be employed with curved boundaries, while still using uniform
increments in x and y.

Figure 3-7 Nomenclature for nodal Figure 3-8 Nomenclature for nodal
equation with convective equation with convection at
boundary condition. a corner section.

m − 1, n m, n
Δy
m, n + 1 2
T∞ Δy T∞
m − 1, n m, n Δy
m, n − 1
Δy m − 1, n − 1
Δx
q 2
Δx
m, n − 1
Δx
2 Surface
Δx

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.91 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

92 3-5 Numerical Method of Analysis

Table 3-2 Summary of nodal formulas for finite-difference calculations. (Dashed lines indicate element volume.)†
Nodal equation for equal increments in x and y
Physical situation (second equation in situation is in form for Gauss-Seidel iteration)

(a) Interior node 0 = Tm+1,n + Tm,n+1 + Tm−1,n + Tm,n−1 − 4Tm,n

m, n + 1 Tm,n = (Tm+1,n + Tm,n+1 + Tm−1,n + Tm,n−1 )/4

Δy
m − 1, n m, n m + 1, n

Δy
m, n − 1

Δx Δx

 
(b) Convection boundary node 0 = hx 1
k T∞ + 2 (2Tm−1,n + Tm,n+1 + Tm,n−1 ) −
hx + 2 T
k m,n

m, n + 1 Tm−1,n +(Tm,n+1 +Tm,n−1 )/2+Bi T∞

Tm,n = 2+Bi
Δy
m − 1, n h, T∞ Bi = hx
k
m, n
Δy
m, n − 1

Δx

 
(c) Exterior corner with convection boundary 0 = 2 hx
k T∞ + (Tm−1,n + Tm,n−1 ) − 2
hx + 1 T
k m,n
h, T∞
m − 1, n (Tm−1,n + Tm,n−1 )/2 + Bi T∞
m, n Tm,n = 1+Bi
Δy
Bi = hx
k
m, n − 1

Δx

 
(d) Interior corner with convection boundary 0 = 2 hx hx T
k T∞ + 2Tm−1,n + Tm,n+1 + Tm+1,n + Tm,n−1 − 2 3 + k m,n

m, n + 1 Bi T∞ +Tm,n+1 +Tm−1,n +(Tm+1,n + Tm,n−1 )/2

Tm,n = 3+Bi

Bi = hx
k
m − 1, n m, n m + 1, n

Δy h, T∞
m, n − 1

Δx

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.92 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch03 10/29/2008 15:34

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 93

Table 3-2 (Continued).

Nodal equation for equal increments in x and y
Physical situation (second equation in situation is in form for Gauss-Seidel iteration)

(e) Insulated boundary 0 = Tm,n+1 + Tm,n−1 + 2Tm−1,n − 4Tm,n

Tm,n = (Tm,n+1 + Tm,n−1 + 2Tm−1,n )/4

m, n + 1
Insulated

m − 1, n m, n

Δy
m, n − 1

Δx
 
(f ) Interior node near curved boundary‡ 0 = b(b 2+ 1) T2 + a +
2 T 2 2 1 1
1 m+1,n + b + 1 Tm,n−1 + a(a + 1) T1 − 2 a + b Tm,n
m, n + 1
h ,T∞ 3
2 c Δ x Δy
1 b Δx
m − 1, n m, n m + 1, n
a Δx
Δy
m, n − 1

Δx Δx
 
(g) Boundary node with convection along 0= √ b T1 + √ b T3 + a+1 hx
b Tm,n + k ( c + 1 + a + b )T∞
2 2 2
a2 + b2 c2 + 1
curved boundary—node 2 for (f ) above§
   
− √ b +√b + a+1 2 2 2 hx
b + ( c + 1 + a + b ) k T2
a2 + b2 c2 +1

† Convection boundary may be converted to insulated surface by setting h = 0 (Bi = 0).

‡ This equation is obtained by multiplying the resistance by 4/(a + 1)(b + 1)
§ This relation is obtained by dividing the resistance formulation by 2.

Nine-Node Problem EXAMPLE 3-5

Consider the square of Figure Example 3-5. The left face is maintained at 100◦ C and the top face
at 500◦ C, while the other two faces are exposed to an environment at 100◦ C:
h = 10 W/m2 · ◦ C and k = 10 W/m · ◦ C

The block is 1 m square. Compute the temperature of the various nodes as indicated in
Figure Example 3-5 and the heat flows at the boundaries.
Solution
The nodal equation for nodes 1, 2, 4, and 5 is
Tm+1,n + Tm−1,n + Tm,n+1 + Tm,n−1 − 4Tm,n = 0

The equation for nodes 3, 6, 7, and 8 is given by Equation (3-25), and the equation for 9 is given
by Equation (3-26):
hx (10)(1) 1
= =
k (3)(10) 3

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.93 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

94 3-5 Numerical Method of Analysis

Figure Example 3-5 Nomenclature for Example 3-5.

T = 500˚C

1 2 3

T = 100˚C
1m
4 5 6
T∞ =
100˚C

7 8 9

1m

The equations for nodes 3 and 6 are thus written

2T2 + T6 + 567 − 4.67T3 = 0
2T5 + T3 + T9 + 67 − 4.67T6 = 0

The equations for nodes 7 and 8 are given by

2T4 + T8 + 167 − 4.67T7 = 0
2T5 + T7 + T9 + 67 − 4.67T8 = 0

and the equation for node 9 is

T6 + T8 + 67 − 2.67T9 = 0
We thus have nine equations and nine unknown nodal temperatures. We shall discuss solution
techniques shortly, but for now we just list the answers:

Node Temperature, ◦ C

1 280.67
2 330.30
3 309.38
4 192.38
5 231.15
6 217.19
7 157.70
8 184.71
9 175.62
The heat flows at the boundaries are computed in two ways: as conduction flows for the 100 and
500◦ C faces and as convection flows for the other two faces. For the 500◦ C face, the heat flow
into the face is
 T   
q= kx = (10) 500 − 280.67 + 500 − 330.30 + (500 − 309.38) 12
y
= 4843.4 W/m

The heat flow out of the 100◦ C face is

 T   
q= ky = (10) 280.67 − 100 + 192.38 − 100 + (157.70 − 100) 12
x
= 3019 W/m

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.94 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 95

The convection heat flow out the right face is given by the convection relation

q= hy(T − T∞ )
   
= (10) 13 309.38 − 100 + 217.19 − 100 + (175.62 − 100) 12

= 1214.6 W/m

Finally, the convection heat flow out the bottom face is


q= hx(T − T∞ )
     
= (10) 13 (100 − 100) 12 + 157.70 − 100 + 184.71 − 100 + (175.62 − 100) 12

= 600.7 W/m

The total heat flow out is

qout = 3019 + 1214.6 + 600.7 = 4834.3 W/m

This compares favorably with the 4843.4 W/m conducted into the top face. A solution of this
example using the Excel spreadsheet format is given in Appendix D.

Solution Techniques
From the foregoing discussion we have seen that the numerical method is simply a means
of approximating a continuous temperature distribution with the finite nodal elements. The
more nodes taken, the closer the approximation; but, of course, more equations mean more
cumbersome solutions. Fortunately, computers and even programmable calculators have
the capability to obtain these solutions very quickly.
In practical problems the selection of a large number of nodes may be unnecessary
because of uncertainties in boundary conditions. For example, it is not uncommon to have
uncertainties in h, the convection coefficient, of ±15 to 20 percent.
The nodal equations may be written as
a11 T1 + a12 T2 + · · · + a1n Tn = C1
a21 T1 + a22 T2 + · · · = C2
a31 T1 + · · · = C3 [3-27]
....................................
an1 T1 + an2 T2 + · · · + ann Tn = Cn

where T1 , T2 , . . . , Tn are the unknown nodal temperatures. By using the matrix notation
⎡C ⎤ ⎡T ⎤
⎡a ⎤ 1 1
11 a12 · · · a1n
⎢ C2 ⎥ ⎢ T2 ⎥
⎢ a21 a22 · · · ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⎥ ⎢ · ⎥ ⎢ · ⎥
[A] = ⎢ a31 ··· ⎥ [C] = ⎢ ⎥ [T ] = ⎢ ⎥
⎣ .................. ⎦ ⎢ · ⎥ ⎢ · ⎥
⎣ · ⎦ ⎣ · ⎦
an1 an2 · · · ann
Cn Tn

Equation (3-27) can be expressed as

[A][T ] = [C] [3-28]

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.95 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

96 3-5 Numerical Method of Analysis

and the problem is to find the inverse of [A] such that

[T ] = [A]−1 [C] [3-29]

Designating [A]−1 by
⎡b b12 · · · b1n ⎤
11
⎢b b22 · · · ⎥
[A]−1 = ⎣ 21
.................. ⎦
bn1 bn2 · · · bnn

the final solutions for the unknown temperatures are written in expanded form as

T1 = b11 C1 + b12 C2 + · · · + b1n Cn

T2 = b21 C1 + · · ·
[3-30]
...................................
Tn = bn1 C1 + bn2 C2 + · · · + bnn Cn

Clearly, the larger the number of nodes, the more complex and time-consuming the solution,
even with a high-speed computer. For most conduction problems the matrix contains a
large number of zero elements so that some simplification in the procedure is afforded. For
example, the matrix notation for the system of Example 3-5 would be

⎡ −4 1 0 1 0 0 0 0 0 ⎤ ⎡ T1 ⎤ ⎡ −600 ⎤
⎢ 1 −4 1 0 1 0 0 0 0 ⎥ ⎢ T2 ⎥ ⎢ −500 ⎥
⎢ ⎥⎢ ⎥ ⎢ ⎥
⎢ 0 2 −4.67 0 0 1 0 0 0 ⎥ ⎢ T3 ⎥ ⎢ −567 ⎥
⎢ ⎥⎢ ⎥ ⎢ ⎥
⎢ 1 0 0 −4 1 0 1 0 0 ⎥ ⎢ T4 ⎥ ⎢ −100 ⎥
⎢ ⎥⎢ ⎥ ⎢ ⎥
⎢ 0 1 0 1 −4 1 0 1 0 ⎥ ⎢ T5 ⎥ = ⎢ 0⎥
⎢ ⎥⎢ ⎥ ⎢ ⎥
⎢ 0 0 1 0 2 −4.67 0 0 1 ⎥ ⎢ T6 ⎥ ⎢ −67 ⎥
⎢ ⎥⎢ ⎥ ⎢ ⎥
⎢ 0 0 0 2 0 0 −4.67 1 0 ⎥ ⎢ T7 ⎥ ⎢ −167 ⎥
⎣ 0 0 0 0 2 0 1 −4.67 1 ⎦ ⎣ T8 ⎦ ⎣ −67 ⎦
0 0 0 0 0 1 0 1 −2.67 T9 −67

We see that because of the structure of the equations the coefficient matrix is very sparse. For
this reason iterative methods of solution may be very efficient. The Gauss-Seidel iteration
method is probably the most widely used for solution of these equations in heat transfer
problems, and we shall discuss that method in Section 3-7.

Software Packages for Solution of Equations

Several software packages are available for solution of simultaneous equations, including
MathCAD (22), TK Solver (23), Matlab (24), and Microsoft Excel (25, 26, 27). The spread-
sheet grid of Excel is particularly adaptable to formulation, solution, and graphical displays
associated with the nodal equations. Details of the use of Excel as a tool for such problems
are presented in Appendix D for both steady-state and transient conditions.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.96 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 97

Figure 3-9 General conduction node.

An example of an Excel worksheet is shown in Figure 3-9, which includes a schematic

of a protruding fin cooled by a convection environment, numerical solution displayed to
match the geometric configuration, and four graphical presentations of the results. While one
might regard this presentation as graphical overkill, it does illustrate a variety of options
available in the Excel format. Several examples are discussed in detail in Appendix D,
including the effects of heat sources and radiation boundary conditions.
Other methods of solution include a transient analysis carried through to steady state
(see Chapter 4), direct elimination (Gauss elimination [9]), or more sophisticated iterative
techniques [12]. An Excel spreadsheet solution treating Example 3-5 as a transient problem
carried through to steady state is given Appendix D.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.97 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

98 3-6 Numerical Formulation in Terms of Resistance Elements

3-6 NUMERICAL FORMULATION IN TERMS

OF RESISTANCE ELEMENTS
Up to this point we have shown how conduction problems can be solved by finite-difference
approximations to the differential equations. An equation is formulated for each node and
the set of equations solved for the temperatures throughout the body. In formulating the
equations we could just as well have used a resistance concept for writing the heat transfer
between nodes. Designating our node of interest with the subscript i and the adjoining nodes
with subscript j, we have the general-conduction-node situation shown in Figure 3-10. At
steady state the net heat input to node i must be zero or
 Tj − Ti
qi + =0 [3-31]
Rij
j

where qi is the heat delivered to node i by heat generation, radiation, etc. The Rij can take
the form of convection boundaries, internal conduction, etc., and Equation (3-31) can be
set equal to some residual for a relaxation solution or to zero for treatment with matrix and
iterative methods.
No new information is conveyed by using a resistance formulation, but some workers
may find it convenient to think in these terms. When a numerical solution is to be performed
that takes into account property variations, the resistance formulation is particularly useful.
In addition, there are many heat-transfer problems where it is convenient to think of con-
vection and radiation boundary conditions in terms of the thermal resistance they impose on
the system. In such cases the relative magnitudes of convection, radiation, and conduction
resistances may have an important influence on the behavior of the thermal model. We
shall examine different boundary resistances in the examples. It will be clear that one will
want to increase thermal resistances when desiring to impede the heat flow and decrease
the thermal resistance when an increase in heat transfer is sought. In some cases the term
thermal impedance is employed as a synonym for thermal resistance, following this line of
thinking.
For convenience of the reader Table 3-3 lists the resistance elements that correspond
to the nodes in Table 3-2. Note that all resistance elements are for unit depth of material
and x = y. The nomenclature for the table is that Rm+ refers to the resistance on the
positive x side of node (m, n), Rn− refers to the resistance on the negative y side of node
(m, n), and so on.

Figure 3-10 General conduction

node.

3 4

Ri3
Ri4
Ri2 Etc.
i
Rij
2

Ri1 j

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.98 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch03 10/29/2008 15:34

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 99

Table 3-3 Resistances for nodes of Table 3-2 x = y, z = 1.

Physical situation Rm+ Rm− Rn+ Rn− V

(a) Interior node 1 1 1 1 (x)2

k k k k
1 1 2 2 (x)2
(b) Convection boundary h x k k k 2
2 2 2 2 (x)2
(c) Exterior corner, convection h x k h x k 4
2 1 1 2 3(x)2
(d) Interior corner, convection† k k k k 4
1 2 2 (x)2
(e) Insulated boundary ∞ k k k 2
(f ) Interior node near 2 2a 2b 2 0.25(1 + a)(1 + b)(x)2
(b+1)k (b+1)k (a+1)k (a+1)k
curved boundary
to node to node 1 to node 2 to node
(m + 1, n) (m, n − 1)
√
R23 = 2 cbk + 1
2
(g) Boundary node with
curved boundary √ V = 0.125[(2 + a) + c](x)2
R21 = 2 abk+ b
2 2
node 2 for ( f ) above
R2−∞ = √ 2 √ 
hx c 2 + 1 + a2 + b2
Rn− = k(a2b
+ 1) to node (m, n)
† Also R = 1/ hx for convection to T .
∞ ∞

The resistance formulation is also useful for numerical solution of complicated three-
dimensional shapes. The volume elements for the three common coordinate systems are
shown in Figure 3-11, and internal nodal resistances for each system are given in Table 3-4.
The nomenclature for the (m, n, k) subscripts is given in Table 3-3, and the plus or minus
sign on the resistance subscripts designates the resistance in a positive or negative direction
from the central node (m, n, k). The elemental volume V is also indicated for each coor-
dinate system. We note, of course, that in a practical problem the coordinate increments are
frequently chosen so that x = y = z, etc., and the resistances are simplified.

3-7 GAUSS-SEIDEL ITERATION

When the number of nodes is very large, an iterative technique may frequently yield a more
efficient solution to the nodal equations than a direct matrix inversion. One such method is
called the Gauss-Seidel iteration and is applied in the following way. From Equation (3-31)
we may solve for the temperature Ti in terms of the resistances and temperatures of the
adjoining nodes Tj as

qi + (Tj /Rij )
j
Ti =  [3-32]
(1/Rij )
j

The Gauss-Seidel iteration makes use of the difference equations expressed in the form
of Equation (3-32) through the following procedure.
1. An initial set of values for the Ti is assumed. This initial assumption can be obtained
through any expedient method. For a large number of nodes to be solved on a computer
the Ti ’s are usually assigned a zero value to start the calculation.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.99 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

100 3-7 Gauss-Seidel Iteration

Figure 3-11 Volume of resistance elements: (a) cartesian, (b) cylindrical, and
(c) spherical coordinate systems.

y
Rn+

x
z
Rm+ z Rn+

Δy φ
Δφ Rm+
R k+ Δz
Δx φ

(a)

rm
z Rk+

Rk+
φ
Δφ Rn+ θ
θ
Δθ
y
x rm
y

x
R m+ (c)

(b)

Table 3-4 Internal nodal resistances for different coordinate systems.

Cartesian Cylindrical Spherical
Nomenclature for increments x, m r, m r, m
y, n φ, n φ, n
z, k z, k θ, k
Volume element V x y z rm r φ z rm
2 sin θ r φ θ

Rm+ x r r
y z k (rm +r/2) φ z k (rm +r/2)2 sin θ φ θ k
Rm− x r r
y z k (rm −r/2) φ z k (rm −r/2)2 sin θ φ θ k
y rm φ φ sin θ
Rn+ x z k r z k r θ k
y rm φ φ sin θ
Rn− x z k r z k r θ k
Rk+ z z θ
x y k rm φ r k sin(θ+θ/2) r φ k
Rk− z z θ
x y k rm φ r k sin(θ−θ/2) r φ k

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.100 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 101

2. Next, the new values of the nodal temperatures Ti are calculated according to Equation
(3-32), always using the most recent values of the Tj .
3. The process is repeated until successive calculations differ by a sufficiently small amount.
In terms of a computer program, this means that a test may be inserted to stop the
calculations when
|Tin+1 − Tin | ≤ δ for all Ti
where δ is some selected constant and n is the number of iterations. Alternatively, a
nondimensional test may be selected such that
 
 Tin+1 − Tin 
≥  
T 
in

Obviously, the smaller the value of δ, the greater the calculation time required to obtain
the desired result. The reader should note, however, that the accuracy of the solution to
the physical problem is not dependent on the value of δ alone. This constant governs the
accuracy of the solution to the set of difference equations. The solution to the physical
problem also depends on the selection of the increment x.
As we noted in the discussion of solution techniques, the matrices encountered in the
numerical formulations are very sparse; they contain a large number of zeros. In solving
a problem with a large number of nodes it may be quite time-consuming to enter all these
zeros, and the simple form of the Gauss-Seidel equation may be preferable.

Nodal Equations for x = y

For nodes with x = y and no heat generation, the form of Equation (3-32) has been
listed as the second equation in segments of Table 3-2. The nondimensional group
hx
= Bi
k
is called the Biot number. Note that equations for convection boundaries may be converted
to insulated boundaries by simply setting Bi = 0 in the respective formula.

Heat Sources and Boundary Radiation Exchange

To include heat generation or radiation heat transfer in the nodal equations for x = y,
one need only add a term
qi /k
to the numerator of each of the equations. For heat sources
qi = q̇V

where q̇ is the heat generated per unit volume and V is the volume of the respective node.
Note that the volume elements are indicated in Table 3-2 by dashed lines. For an interior
node V = xy, for a plane convection boundary V = (x/2)y, for an exterior corner
V = (x/2)(y/2), etc.
For radiation exchange at boundary note,

qi = qrad,i × A

where A is the surface area of the node exposed to radiation, and qrad,i is the net radiation
transferred to node i per unit area as determined by the methods of Chapter 8.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.101 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

102 3-8 Accuracy Considerations

For the common case of a surface exposed to a large enclosure at radiation temperature
of Tr , the net radiation to the surface per unit area is given by Equation (1-12),
 

qrad,i = σεi Tr4 − Ti4

where εi is the emissivity of note i and all temperatures must by expressed in degrees
absolute.

3-8 ACCURACY CONSIDERATIONS

We have already noted that the finite-difference approximation to a physical problem
improves as smaller and smaller and smaller increments of x and y are used. But we
have not said how to estimate the accuracy of this approximation. Two basic approaches
are available.
1. Compare the numerical solution with an analytical solution for the problem, if available,
or an analytical solution for a similar problem.
2. Choose progressively smaller values of x and observe the behavior of the solution. If
the problem has been correctly formulated and solved, the nodal temperatures should
converge as x becomes smaller. It should be noted that computational round-off errors
increase with an increase in the number of nodes because of the increased number of
machine calculations. This is why one needs to observe the convergence of the solution.
It can be shown that the error of the finite-difference approximation to ∂T/∂x is of the order
of (x/L)2 where L is some characteristic body dimension.
Analytical solutions are of limited utility in checking the accuracy of a numerical model
because most problems that will need to be solved by numerical methods either do not have
an analytical solution at all or, if one is available, it may be too cumbersome to compute.

Energy Balance as Check on Solution Accuracy

In discussing solution techniques for nodal equations, we stated that an accurate solution of
these equations does not ensure an accurate solution to the physical problem. In many cases
the final solution is in serious error simply because the problem was not formulated correctly
at the start. No computer or convergence criterion can correct this kind of error. One way
to check for formulation errors is to perform some sort of energy balance using the final
solution. The nature of the balance varies from problem to problem but for steady state it
always takes the form of energy in equals energy out. If the energy balance does not check
within reasonable limits, there is a likelihood that the problem has not been formulated
correctly. Perhaps a constant is wrong here or there, or an input data point is incorrect, a
faulty computer statement employed, or one or more nodal equations incorrectly written.
If the energy balance does check, one may then address the issue of using smaller values
of x to improve accuracy.
In the examples we present energy balances as a check on problem formulation.

Accuracy of Properties and Boundary Conditions

From time to time we have mentioned that thermal conductivities of materials vary with
temperature; however, over a temperature range of 100 to 200◦ C the variation is not great
(on the order of 5 to 10 percent) and we are justified in assuming constant values to simplify
problem solutions. Convection and radiation boundary conditions are particularly notorious
for their nonconstant behavior. Even worse is the fact that for many practical problems the

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.102 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 11/6/2008 9:46

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 103

basic uncertainty in our knowledge of convection heat-transfer coefficients may not be

better than ±25 percent. Uncertainties of surface-radiation properties of ±10 percent are
not unusual at all. For example, a highly polished aluminum plate, if allowed to oxidize
heavily, will absorb as much as 300 percent more radiation than when it was polished.
These remarks are not made to alarm the reader, but rather to show that selection of a
large number of nodes for a numerical formulation does not necessarily produce an accurate
solution to the physical problem; we must also examine uncertainties in the boundary
conditions. At this point the reader is ill-equipped to estimate these uncertainties. Later
chapters on convection and radiation will clarify the matter.

Gauss-Seidel Calculation EXAMPLE 3-6

Apply the Gauss-Seidel technique to obtain the nodal temperatures for the four nodes in Figure 3-6.
Solution
It is useful to think in terms of a resistance formulation for this problem because all the connecting
resistances between the nodes in Figure 3-6 are equal; that is,

y
x 1
R= = = [a]
k
y k
y k
Therefore, when we apply Equation (3-32) to each node, we obtain (qi = 0)


kj Tj
j
Ti =
[b]
kj
j

Because each node has four resistances connected to it and k is assumed constant,

kj = 4k
j

and
1
Ti = Tj [c]
4
j

We now set up an iteration table as shown and use initial temperature assumptions of 300 and
200◦ C. Equation (c) is then applied repeatedly until satisfactory convergence is achieved. In the
table, five iterations produce convergence with 0.13 degree. To illustrate the calculation, we can
note the two specific cases below:
(T2 )n=1 = 14 (500 + 100 + T4 + T1 ) = 14 (500 + 100 + 200 + 275) = 268.75
(T3 )n=4 = 14 (100 + T1 + T4 + 100) = 14 (100 + 250.52 + 150.52 + 100) = 150.26

Number of
iterations n T1 T2 T3 T4

0 300 300 200 200

1 275 268.75 168.75 159.38
2 259.38 254.69 154.69 152.35
3 251.76 251.03 151.03 150.52
4 250.52 250.26 150.26 150.13
5 250.13 250.07 150.07 150.03

Note that in computing (T3 )n=4 we have used the most recent information available to us for T1
and T4 .

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.103 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch03 10/30/2008 19:17

104 3-8 Accuracy Considerations

EXAMPLE 3-7 Numerical Formulation with Heat Generation

We illustrate the resistance formulation in cylindrical coordinates by considering a 4.0-mm-
diameter wire with uniform heat generation of 500 MW/m3 . The outside surface temperature of the
wire is 200◦ C, and the thermal conductivity is 19 W/m · ◦ C. We wish to calculate the temperature
distribution in the wire. For this purpose we select four nodes as shown in Figure Example 3-7a.
We shall make the calculations per unit length, so we let z = 1.0. Because the system is one-
dimensional, we take φ = 2π. For all the elements r is chosen as 0.5 mm. We then compute the
resistances and volume elements using the relations from Table 3-4, and the values are given below.
The computation of Rm+ for node 4 is different from the others because the heat-flow path is shorter.
For node 4, rm is 1.75 mm, so the positive resistance extending to the known surface temperature is
r/2 1
Rm+ = =
(rm + r/4) φ z k 15πk
The temperature equation for node 4 is written as
2749 + 6πkT3 + 15πk(200)
T4 =
21πk
where the 200 is the known outer surface temperature.

rm , Rm+ , Rm− , V = rm r φ z, qi = q̇ V ,

Node mm ◦ C/W ◦ C/W μm3 W

1 0.25 1 ∞ 0.785 392.5

2πk
2 0.75 1 1 2.356 1178
4πk 2πk
3 1.25 1 1 3.927 1964
6πk 4πk
4 1.75 1 1 5.498 2749
15πk 6πk

Figure Example 3-7a Figure Example 3-7b Comparison of

Example Schematic. analytical and
numerical solution.

25
1 2 3 4

20
T – T w , ˚C

15

10
Analytical
Numerical
5

0
0.5 1.0 1.5 2.0
r, mm

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.104 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 11/6/2008 9:46

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 105

A summary of the values of (1/Rij ) and Ti according to Equation (3-32) is now given to be
used in a Gauss-Seidel iteration scheme.


1 qi + (Tj /Rij )
Node , W/◦ C Ti = 
Rij (1/Rij )

1 2πk = 119.38 T1 = 3.288 + T2

2 6πk = 358.14 T2 = 3.289 + 13 T1 + 23 T3
3 10πk = 596.90 T3 = 3.290 + 0.4T2 + 0.6T4
4 21πk = 1253.50 T4 = 2.193 + 27 T3 + 142.857

Thirteen iterations are now tabulated:

Node temperature, ◦ C
Iteration n T1 T2 T3 T4

0 240 230 220 210

1 233.29 227.72 220.38 208.02
2 231.01 227.21 218.99 207.62
3 230.50 226.12 218.31 207.42
4 229.41 225.30 217.86 207.30
5 228.59 224.73 217.56 207.21
6 228.02 224.34 217.35 207.15
7 227.63 224.07 217.21 207.11
8 227.36 223.88 217.11 207.08
9 227.17 223.75 217.04 207.06
10 227.04 223.66 216.99 207.04
11 226.95 223.60 216.95 207.04
12 226.89 223.55 216.93 207.03
13 226.84 223.52 216.92 207.03
Analytical 225.904 222.615 216.036 206.168
Gauss-Seidel check 225.903 222.614 216.037 206.775
Exact solution
of nodal equations 226.75 223.462 216.884 207.017

We may compare the iterative solution with an exact calculation which makes use of Equation
(2-25a):
q̇
T − Tw = (R2 − r 2 )
4k
where Tw is the 200◦ C surface temperature, R = 2.0 mm, and r is the value of rm for each node.
The analytical values are shown following iteration 13, and then a Gauss-Seidel check is made on
the analytical values. There is excellent agreement on the first three nodes and somewhat less on
node 4. Finally, the exact solutions to the nodal equations are shown for comparison. These are
the values the iterative scheme would converge to if carried far enough. In this limit the analytical
and numerical calculations differ by a constant factor of about 0.85◦ C, and this difference results
mainly from the way in which the surface resistance and boundary condition are handled. A
smaller value of r near the surface would produce better agreement. A graphical comparison of
the analytical and numerical solutions is shown in Figure Example 3-7b.
The total heat loss from the wire may be calculated as the conduction through Rm+ at node 4.
Then
T − Tw
q= 4 = 15πk(207.03 − 200) = 6.294 kW/m [6548 Btu/h · ft]
Rm+

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.105 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/30/2008 16:47

106 3-8 Accuracy Considerations

This must equal the total heat generated in the wire, or

q = q̇V = (500 × 106 )π(2 × 10−3 )2 = 6.283 kW/m [6536 Btu/h · ft]
The difference between the two values results from the inaccuracy in determination of T4 . Using
the exact solution value of 207.017◦ C would give a heat loss of 6.2827 kW. For this problem the
exact value of heat flow is 6.283 kW because the heat-generation calculation is independent of the
finite-difference formulation.

Heat Generation with Nonuniform

EXAMPLE 3-8 Nodal Elements
A layer of glass [k = 0.8 W/m · ◦ C] 3 mm thick has thin 1-mm electric conducting strips
attached to the upper surface, as shown in Figure Example 3-8. The bottom surface of the
glass is insulated, and the top surface is exposed to a convection environment at 30◦ C with
h = 100 W/m2 · ◦ C. The strips generate heat at the rate of 40 or 20 W per meter of length. Deter-
mine the steady-state temperature distribution in a typical glass section, using the numerical method
for both heat-generation rates.

Figure Example 3-8 (a) Physical system, (b) nodal boundaries.

T∞ = 30˚C
3.0 cm 3.0 cm
Heater 3 mm
1 mm Glass

Insulation
(a)

T∞ = 30˚C
5 mm
Heater 1 mm
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28

(b)

Solution
The nodal network for a typical section of the glass is shown in the figure. In this example we have
not chosen x = y. Because of symmetry, T1 = T7 , T2 = T6 , etc., and we only need to solve for
the temperatures of 16 nodes. We employ the resistance formulation. As shown, we have chosen
x = 5 mm and y = 1 mm. The various resistances may now be calculated:
Nodes 1, 2, 3, 4:
1 1 k(y/2) (0.8)(0.001/2)
= = = = 0.08
Rm+ Rm− x 0.005
1
= hA = (100)(0.005) = 0.5
Rn+
1 kx (0.8)(0.005)
= = = 4.0
Rn− y 0.001

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.106 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 107

Nodes 8, 9, 10, 11, 15, 16, 17, 18:

1 1 ky (0.8)(0.001)
= = = = 0.16
Rm+ Rm− x 0.005
1 1 kx
= = = 4.0
Rn+ Rn− y
Nodes 22, 23, 24, 25:
1 1 k(y/2)
= = = 0.08
Rm+ Rm− x
1 kx
= = 4.0
Rn+ y
1
=0 (insulated surface)
Rn−
The nodal equations are obtained from Equation (3-31) in the general form
 
(Tj /Rij ) + qi − Ti (1/Rij ) = 0

Only node 4 has a heat-generation

 term, and qi = 0 for all other nodes. From the above resistances
we may calculate the (1/Rij ) as

Node (1/Rij )

1, 2, 3, 4 4.66
8, . . . , 18 8.32
22, 23, 24, 25 4.16

For node 4 the equation is

(2)(0.08)T3 + 4.0T5 + (0.5)(30) + q4 − 4.66T4 = 0

The factor of 2 on T3 occurs because T3 = T5 from symmetry. When all equations are evaluated
and the solution obtained, the following temperatures result:

q/L, W/m
Node
temperature, ◦ C 20 40

1 31.90309 33.80617
2 32.78716 35.57433
3 36.35496 42.70993
4 49.81266 69.62532
8 32.10561 34.21122
9 33.08189 36.16377
10 36.95154 43.90307
11 47.82755 65.65510
15 32.23003 34.46006
16 33.26087 36.52174
17 37.26785 44.53571
18 46.71252 63.42504
22 32.27198 34.54397
23 33.32081 36.64162
24 37.36667 44.73333
25 46.35306 62.70613

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.107 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

108 3-8 Accuracy Considerations

The results of the model and calculations may be checked by calculating the convection heat lost
by the top surface. Because all the energy generated in the small heater strip must eventually be
lost by convection (the bottom surface of the glass is insulated and thus loses no heat), we know
the numerical value that the convection should have. The convection loss at the top surface is
given by

qc = hi Ai (Ti − T∞ )


x x
= (2)(100) (T1 − T∞ ) + x(T2 + T3 − 2T∞ ) + (T4 − T∞ )
2 2

The factor of 2 accounts for both sides of the section. With T∞ = 30◦ C this calculation yields
qc = 19.999995 for q/L = 20 W/m
qc = 40.000005 for q/L = 40 W/m

Obviously, the agreement is excellent.

Composite Material with Nonuniform

EXAMPLE 3-9 Nodal Elements
A composite material is embedded in a high-thermal-conductivity material maintained at 400◦ C
as shown in Figure Example 3-9a. The upper surface is exposed to a convection environment at
30◦ C with h = 25 W/m2 · ◦ C. Determine the temperature distribution and heat loss from the upper
surface for steady state.
Solution
For this example we choose nonsquare nodes as shown in Figure Example 3-9b. Note also that
nodes 1, 4, 7, 10, 13, 14, and 15 consist of two materials. We again employ the resistance
formulation.
For node 1:
1 kA (2.0)(0.005)
= = = 0.6667
Rm+ x 0.015
1 kA (0.3)(0.005)
= = = 0.15
Rm− x 0.01
1
= hA = (25)(0.005 + 0.0075) = 0.3125
Rn+
   
1 kA kA (0.3)(0.005) + (2.0)(0.0075)
= + = = 1.65
Rn− y L y R 0.01

For nodes 4, 7, 10:

1 (2.0)(0.01)
= = 1.3333
Rm+ 0.015
1 (0.3)(0.01)
= = 0.3
Rm− 0.01
1 1
= = 1.65
Rn+ Rn−

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.108 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch03 10/29/2008 15:34

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 109

Figure Example 3-9 (a) Physical system, (b) nodal boundaries.

T∞ = 30˚C h = 25 W冫 m2 • C
1 cm 1.5 cm

k = 0.3 W冫m • ˚C
ρ = 2000 kg冫 m3
1 cm c = 0.8 kJ冫 kg • ˚C

k = 2.0 W冫 m • ˚C
ρ = 2800 kg冫 m3
c = 0.9 kJ冫 kg • ˚C

T = 400˚C
(a)

1 2 3 2 1

4 5 6 5 4

7 8 9 8 9

10 11 12 11 10

13 14 15 14 13

(b)

For node 13:

1 (2.0)(0.005) + (0.3)(0.005)
= = 0.76667
Rm+ 0.015
1 (0.3)(0.01)
= = 0.3
Rm− 0.01
1
= 1.65
Rn+
1 (0.3)(0.0075) + (0.3)(0.005)
= = 0.375
Rn− 0.01
For nodes, 5, 6, 8, 9, 11, 12:
1 1 (2.0)(0.01)
= = = 1.3333
Rm+ Rm− 0.015
1 1 (2.0)(0.015)
= = = 3.0
Rn+ Rn− 0.01
For nodes 2, 3:
1 1 (2.0)(0.005)
= = = 0.6667
Rm+ Rm− 0.015

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.109 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

110 3-8 Accuracy Considerations

1
= hA = (2.5)(0.015) = 0.375
Rn+
1
= 3.0
Rn−

For nodes 14, 15:

1 1 (2.0)(0.005) + (0.3)(0.005)
= = = 0.76667
Rm+ Rm− 0.015
1
= 3.0
Rn+
1 (0.3)(0.015)
= = 0.45
Rn− 0.01

We shall use Equation (3-32) for formulating the nodal equations. For node 1, (1/Rij ) = 2.7792,
and we obtain
1
T1 = [(400)(0.15) + (30)(0.3125) + T2 (0.6667) + 1.65T4 ]
2.7792

For node 3, (1/Rij ) = 4.7083, and the nodal equation is
1
T3 = [T2 (0.6667)(2) + 3.0T6 + (0.375)(30)]
4.7083
The factor of 2 on T2 occurs because of the mirror image of T2 to the right of T3 .
A similar procedure is followed for the other nodes to obtain 15 nodal equations with the 15
unknown temperatures. These equations may then be solved by whatever computation method is
most convenient. The resulting temperatures are:

T1 = 254.956 T2 = 247.637 T3 = 244.454

T4 = 287.334 T5 = 273.921 T6 = 269.844
T7 = 310.067 T8 = 296.057 T9 = 291.610
T10 = 327.770 T11 = 313.941 T12 = 309.423
T13 = 343.516 T14 = 327.688 T15 = 323.220
The heat flow out the top face is obtained by summing the convection loss from the nodes:

qconv = hAi (Ti − T∞ )
= (2)(25)[(0.0125)(254.96 − 30) + (0.015)(247.64 − 30)
+ (0.0075)(244.45 − 30)]
= 382.24 W per meter of depth

As a check on this value, we can calculate the heat conducted in from the 400◦ C surface to nodes
1, 4, 7, 10, 13, 14, and 15:
 T
qcond = kAi
x

0.3
qcond = 2 [(0.005)(400 − 254.96) + (0.01)(400 − 287.33) + (0.01)(400 − 310.07)
0.01
+ (0.01)(400 − 327.77) + (0.0225)(400 − 343.52) + (0.015)(400 − 327.69)
+ (0.0075)(400 − 323.22)]
= 384.29 W per meter of depth

The agreement is excellent.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.110 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 111

Radiation Boundary Condition EXAMPLE 3-10

A 1-by-2-cm ceramic strip [k = 3.0 W/m · ◦ C, ρ = 1600 kg/m3 , and c = 0.8 kJ/kg · ◦ C] is
embedded in a high-thermal-conductivity material, as shown in Figure Example 3-10, so that
the sides are maintained at a constant temperature of 900◦ C. The bottom surface of the ceramic is
insulated, and the top surface is exposed to a convection and radiation environment at T∞ = 50◦ C;
h = 50 W/m2 · ◦ C, and the radiation heat loss is calculated from
q = σA (T 4 − T∞
4 )

where

A = surface area
σ = 5.669 × 10−8 W/m2 · ◦ K 4
= 0.7
Solve for the steady-state temperature distribution of the nodes shown and the rate of heat loss.
The radiation temperatures are in degrees Kelvin.

Figure Example 3-10

2 cm
h, T∞ = 50˚C
1 2 3
4 5 6
1 cm T = 900˚C

7 8 9

T = 900˚C
Insulated

Solution
We shall employ the resistance formulation and note that the radiation can be written as
T − T∞
q = σ A(T 4 − T∞
4 )= [a]
Rrad

1
= σ A(T 2 + T∞
2 )(T + T )
∞ [b]
Rrad
From symmetry T1 = T3 , T4 = T6 , and T7 = T9 , so we have only six unknown nodes. The resis-
tances are now computed:
Nodes 1, 2:
1 1 kA (3.0)(0.0025) 1 (3.0)(0.005)
= = = = 1.5 = = 3.0
Rm+ Rm− x 0.005 Rn− 0.005

1
= hA = (50)(0.005) = 0.25 [c]
Rn+,conv
1
= σ A(T 2 + T∞
2 )(T + T )
∞
Rn+,rad

The radiation term introduces nonlinearities and will force us to employ an iterative solution.
Nodes 4, 5:
1 kA (3.0)(0.005)
All = = = 3.0
R x 0.005

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.111 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

112 3-8 Accuracy Considerations

Nodes 7, 8:
1 1 1
= = 1.5 = 3.0
Rm+ Rm− Rn+
Because the bottom surface is insulated, 1/Rn− = 0. We now use Equation (3-32)

(Tj /Rij )
Ti =  [3-33]
(1/Rij )

and tabulate:

Node (1/Rij )

1 6.25+1/Rrad
2 6.25+1/Rrad
4 12
5 12
7 6
8 6

Our nodal equations are thus expressed in degrees Kelvin because of the radiation terms and
become
1
T1 =  [1.5T2 + 3T4 + (1.5)(1173) + (323)(0.25)
(1/Rij )
+ σ (0.005)(T12 + 3232 )(T1 + 323)(323)]
1
T2 =  [1.5T1 (2) + 3T5 + (323)(0.25)
(1/Rij )
+ σ (0.005)(T22 + 3232 )(T2 + 323)(323)]
1 [(1173)(3.0) + 3T + 3T + 3T ]
T4 = 12 1 [2T (3.0) + 3T + 3T ]
T5 = 12
1 7 5 4 2 8
T7 = 16 [(1173)(1.5) + 3T4 + 1.5T8 ] T8 = 16 [2T7 (1.5) + 3T5 ]

The radiation terms create a very nonlinear set of equations. The computational algorithm we shall
use is outlined as follows:
1. Assume T1 = T2 = 1173 K.

2. Compute 1/Rrad and (1/Rij ) for nodes 1 and 2 on the basis of this assumption.
3. Solve the set of equations for T1 through T8 .
4. Using new values of T1 and T2 , recalculate 1/Rrad values.
5. Solve equations again, using new values.
6. Repeat the procedure until answers are sufficiently convergent.
The results of six iterations are shown in the table. As can be seen, the convergence is quite rapid.
The temperatures are in kelvins.

Iteration T1 T2 T4 T5 T7 T8

1 990.840 944.929 1076.181 1041.934 1098.951 1070.442

2 1026.263 991.446 1095.279 1068.233 1113.622 1090.927
3 1019.879 982.979 1091.827 1063.462 1110.967 1087.215
4 1021.056 984.548 1092.464 1064.344 1111.457 1087.901
5 1020.840 984.260 1092.347 1064.182 1111.367 1087.775
6 1020.879 984.313 1092.369 1064.212 1111.384 1087.798

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.112 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 11/6/2008 9:46

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 113

As a practical matter, the iterations would be carried out using a commercial software package
(such as those mentioned in References 22–27) and only the final set of values would be displayed
on the computer.
At this point we may note that in a practical problem the value of
will only be known within
a tolerance of several percent, and thus there is nothing to be gained by carrying the solution to
unreasonable limits of accuracy.
The heat loss is determined by calculating the radiation and convection from the top surface
(nodes 1, 2, 3):


qrad = σ
Ai (Ti4 − 3234 )
= (5.669 × 10−8 )(0.7)(0.005)[(2)(1020.884 − 3234 ) + 984.3134 − 3234 ]
= 610.8 W/m depth


qconv = hAi (Ti − 323)
= (50)(0.005)[(2)(1020.88 − 323) + 984.313 − 323] = 514.27 W
qtotal = 610.8 + 514.27 = 1125.07 W/m depth

This can be checked by calculating the conduction input from the 900◦ C surfaces:

T
qcond = kAi
x
(2)(3.0)
= [(0.0025)(1173 − 1020.879) + (0.005)(1173 − 1092.369)
0.005
+ (0.0025)(1173 − 1111.384)]
= 1124.99 W/m depth

The agreement is excellent.

Use of Variable Mesh Size EXAMPLE 3-11

One may use a variable mesh size in a problem with a finer mesh to help in regions of large
temperature gradients. This is illustrated in Figure Example 3-11, in which Figure 3-6 is redrawn
with a fine mesh in the corner. The boundary temperatures are the same as in Figure 3-6. We wish
to calculate the nodal temperatures and compare with the previous solution. Note the symmetry
of the problem: T1 = T2 , T3 = T4 , etc.
Figure Example 3-11

∆x 500˚C
3
∆y 5 6 7 7
3 8 9 10 10
11 12 1 2
100˚C 100˚C

∆y

13 14 3 4

∆x
100˚C

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.113 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

114 3-8 Accuracy Considerations

Solution
Nodes 5, 6, 8, and 9 are internal nodes with x = y and have nodal equations in the form of
Equation (3-24). Thus,
600 + T6 + T8 − 4T5 = 0
500 + T5 + T7 + T9 − 4T6 = 0
100 + T5 + T9 + T11− 4T8 = 0
T8 + T6 + T10 + T12 − 4T9 = 0

For node 7 we can use a resistance formulation and obtain

1/R7−6 = k
k(x/6 + x/2)
1/R7−500◦ = = 2k
y/3
1/R7−10 = 2k

and we find
1000 + T6 + 2T10 − 5T7 = 0

Similar resistances are obtained for node 10.

1/R10−9 = k
1/R10−7 = 2k = 1/R10−1

so that
2T7 + T9 + 2T1 − 5T10 = 0

For node 1,
k(y/6 + y/2)
1/R1−12 = = 2k
x/3
k(x/6 + x/2)
1/R1−3 = = 2k/3
y
1/R1−10 = 2k

and the nodal equation becomes

3T12 + 3T10 + T3 − 7T1 = 0

For node 11,

k(y/6 + y/2)
1/R11−100◦ = 1/R11−12 = = 2k
x/3
1/R11−8 = k
k(x/3)
1/R11−13 = = k/3
y

and the nodal equation becomes

600 + 6T12 + 3T8 + T13 − 16T11 = 0

Similarly, the equation for node 12 is

3T9 + 6T11 + 6T1 + T14 − 16T12 = 0

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.114 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 115

For node 13,

k y
1/R13−100◦ = = 3k = 1/R13−14
x/3
1/R13−11 = 1/R13−100 = k/3

and we obtain
1000 + 9T14 + T11 − 20T13 = 0

Similarly for node 14,

100 + 9T13 + 9T3 + T12 − 20T14 = 0

Finally, from resistances already found, the nodal equation for node 3 is
200 + 9T14 + 2T1 − 13T3 = 0

We choose to solve the set of equations by the Gauss-Seidel iteration technique and thus write
them in the form Ti = f(Tj ). The solution was set up on a computer with all initial values for the
Ti ’s taken as zero. The results of the computations are shown in the following table.

Number of iterations
Node 2 10 20 30 50

1 59.30662 232.6668 247.1479 247.7605 247.7875

2 59.30662 232.6668 247.1479 247.7605 247.7875
3 50.11073 139.5081 147.2352 147.5629 147.5773
4 50.11073 139.5081 147.2352 147.5629 147.5773
5 206.25 288.358 293.7838 294.0129 294.023
6 248.75 359.025 366.9878 367.3243 367.3391
7 291.45 390.989 398.7243 399.0513 399.0657
8 102.9297 200.5608 208.4068 208.7384 208.753
9 121.2334 264.2423 275.7592 276.2462 276.2677
10 164.5493 302.3108 313.5007 313.974 313.9948
11 70.95459 156.9976 164.3947 164.7076 164.7215
12 73.89051 203.6437 214.5039 214.9634 214.9836
13 70.18905 115.2635 119.2079 119.3752 119.3826
14 62.82942 129.8294 135.6246 135.8703 135.8811

Again, the results of the various sets of iterations are shown merely to illustrate the rapidity of
convergence. In actual practice only the final set of values would be displayed on the computer.
Note that these solutions for T1 = T2 = 247.79◦ C and T3 = T4 = 147.58◦ C are somewhat below the
values of 250◦ C and 150◦ C obtained when only four nodes were employed, but only modestly so.

Three-Dimensional Numerical Formulation EXAMPLE 3-12

To further illustrate the numerical formulation, consider the simple three-dimensional block shown
in Figure Example 3-12a. The block has dimensions of 3 × 4 × 4 cm with the front surface exposed
to a convection environment with T∞ = 10◦ C and h = 500 W/m2 · ◦ C. The four sides are main-
tained constant at 100◦ C and the back surface is insulated. We choose x = y = z = 1 cm and
set up the nodes as shown. The front surface has nodes 11, 12, 13, 14, 15, 16; the next z nodes are
21, 22, 23, 24, 25, 26 and so on. We shall use the resistance formulation in the form of Equation
(3-32) to set up the nodal equations.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.115 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

116 3-8 Accuracy Considerations

Figure Example 3-12a Schematic.

T = 100⬚C
walls

Insulated
51 52 53 back surface

54 55 56

11 12 13 5
es k = 2.0 W/m • ˚C
lan 4
z-p
14 15 16 3

2 ⌬x = ⌬y = ⌬z = 1 cm

1
Convection front y
surface, h = 500 W/m • ˚C z
T∞ = 10⬚C
T11 = T12 = T14 = T16 x
T12 = T15
(a)

Figure Example 3-12b Results.

100

90

80 T11, etc.

70
Temperature, ⬚C

60
T12, etc.

50

40

30

20

10
1 2 3 4 5
z-plane
(b)

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.116 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 117

Solution
All of the interior nodes for z-planes 2, 3, 4 have resistances of
1/R = k A/x = (2)(0.01)2 /0.01 = 0.02 = 1/R11−21 = 1/R21−22 , etc.

The surface conduction resistances for surface z-plane 1 are

1/R11−12 = k A/x = (2)(0.01/2)/0.01 = 0.01 = 1/R11−14 , etc.

The surface convection resistances are

1/R11−∞ = h A = (500)(0.01)2 = 0.05

For surface nodes like 11 the (1/Rij ) term in Equation (3-32) becomes

(1/R11−j ) = (4)(0.01) + 0.02 + 0.05 = 0.11

while, for interior nodes, we have


(1/R21−j ) = (6)(0.02) = 0.12

For the insulated back surface nodes


(1/R51−j ) = (4)(0.01) + (0.02) = 0.06

There are 30 nodes in total; 6 in each z-plane. We could write the equations for all of them but
prefer to take advantage of the symmetry of the problem as indicated in the figure. Thus,
T11 = T13 = T14 = T16 and T12 = T15 , etc.

We may then write the surface nodal equations as

T11 = [0.05 T∞ + 0.02 T21 + (0.01)(100 + 100 + T14 + T12 )]/0.11
T12 = [0.05 T∞ + 0.02 T22 + (0.01)(100 + T11 + T15 + T13 )]/0.11

Inserting T∞ = 10 and simplifying we have

T11 = (2.5 + 0.02 T21 + 0.01 T12 )/0.1
T12 = (1.5 + 0.02 T22 + 0.02 T11 )/0.1

Following the same procedure for the other z-planes we obtain

T21 = (200 + T11 + T31 + T22 )/5
T22 = (100 + T12 + T32 + T21 )/5
T31 = (200 + T21 + T41 + T32 )/5
T32 = (100 + T22 + T42 + T31 )/5
T41 = (200 + T31 + T51 + T42 )/5
T42 = (100 + T32 + T52 + 2 T41 )/5
T51 = (2 + 0.02 T41 + 0.01 T52 )/0.05
T52 = (1 + 0.02 T42 + 0.02 T51 )/0.05

Solving the 10 equations gives the following results for the temperatures in each z-plane.
z-plane Node 1 Node 2

1 45.9 40.29
2 84.36 80.57
3 95.34 93.83
4 98.49 97.93
5 99.16 98.94

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.117 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch03 10/29/2008 15:34

118 3-9 Electrical Analogy for Two-Dimensional Conduction

Figure Example 3-12b gives a graphical display of the results, and the behavior is as expected.
The temperature drops as the cooled front surface is approached. Node 2 is cooled somewhat more
than node 1 because it is in contact with only a single 100◦ surface.

Comments
While this is a rather simple three-dimensional example, it has illustrated the utility of the resistance
formulation in solving such problems. As with two-dimensional systems, variable mesh sizes, heat
generation, and variable boundary conditions can be accommodated with care and patience.

Remarks on Computer Solutions

It should be apparent by now that numerical methods and computers give the engineer
powerful tools for solving very complex heat-transfer problems. Many large commercial
software packages are available, and new ones appear with increasing regularity. One char-
acteristic common to almost all heat-transfer software is a requirement that the user under-
stand something about the subject of heat transfer. Without such an understanding it can
become very easy to make gross mistakes and never detect them at all. We have shown
how energy balances are one way to check the validity of a computer solution. Sometimes
common sense also works well. We know, for example, that a plate will cool faster when
air is blown across the plate than when exposed to still air. Later, in Chapters 5 through 7,
we will see how to quantify these effects and will be able to anticipate what influence they
may have on a numerical solution to a conduction problem. A similar statement can be
made pertaining to radiation boundary conditions, which will be examined in Chapter 8.
These developments will give the reader a “feel” for what the effects of various boundary
conditions should be and insight about whether the numerical solution obtained for a prob-
lem appears realistic. Up to now, boundary conditions have been stipulated quantities, but
experienced heat-transfer practitioners know that they are seldom easy to determine in the
real world.

3-9 ELECTRICAL ANALOGY FOR

TWO-DIMENSIONAL CONDUCTION
Steady-state electric conduction in a homogeneous material of constant resistivity is anal-
ogous to steady-state heat conduction in a body of similar geometric shape. For two-
dimensional electric conduction the Laplace equation applies:

∂2 E ∂2 E
+ 2 =0
∂x2 ∂y

where E is the electric potential. A very simple way of solving a two-dimensional heat-
conduction problem is to construct an electrical analog and experimentally determine the
geometric shape factors for use in Equation (3-23). One way to accomplish this is to use
a commercially available paper that is coated with a thin conductive film. This paper may
be cut to an exact geometric model of the two-dimensional heat-conduction system. At
the appropriate edges of the paper, good electrical conductors are attached to simulate
the temperature boundary conditions on the problem. An electric-potential difference is
then impressed on the model. It may be noted that the paper has a very high resistance in

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.118 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 119

comparison with the conductors attached to the edges, so that a constant-potential condition
can be maintained at the region of contact.
Once the electric potential is impressed on the paper, an ordinary voltmeter may be used
to plot lines of constant electric potential. With these constant-potential lines available, the
flux lines may be easily constructed since they are orthogonal to the potential lines. These
equipotential and flux lines have precisely the same arrangement as the isotherms and heat-
flux lines in the corresponding heat-conduction problem. The shape factor is calculated
immediately using the method which was applied to the curvilinear squares.
It may be noted that the conducting-sheet analogy is not applicable to problems where
heat generation is present; however, by addition of appropriate resistances, convection
boundary conditions may be handled with little trouble. Schneider [2] and Ozisik [10]
discuss the conducting-sheet method, as well as other analogies for treating conduction
heat-transfer problems, and Kayan [4, 5] gives a detailed discussion of the conducting-sheet
method. Because of the utility of numerical methods, analogue techniques for solution of
heat-transfer problems are largely of historical interest.

3-10 SUMMARY
There is a myriad of analytical solutions for steady-state conduction heat-transfer problems
available in the literature. In this day of computers most of these solutions are of small
utility, despite their exercise in mathematical facilities. This is not to say that we cannot
use the results of past experience to anticipate answers to new problems. But, most of the
time, the problem a person wants to solve can be attacked directly by numerical techniques,
except when there is an easier way to do the job. As a summary, the following suggestions
are offered:
1. When tackling a two- or three-dimensional heat-transfer problem, first try to reduce it
to a one-dimensional problem. An example is a cylinder with length much larger than
its diameter.
2. If possible, select a simple shape-factor model that may either exactly or approximately
represent the physical situation. See comments under items 4 and 5.
3. Seek some simple analytical solutions but, if solutions are too complicated, go directly
to the numerical techniques.
4. In practical problems, recognize that convection and radiation boundary conditions are
subject to large uncertainties. This means that, in most practical situations, undue concern
over accuracy of solution to numerical nodal equations is unjustified.
5. In general, approach the solution in the direction of simple to complex, and make use
of checkpoints along the way.

REVIEW QUESTIONS
1. What is the main assumption in the separation-of-variables method for solving Laplace’s
equation?
2. Define the conduction shape factor.
3. What is the basic procedure in setting up a numerical solution to a two-dimensional
conduction problem?

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.119 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

120 Problems

4. Once finite-difference equations are obtained for a conduction problem, what methods
are available to effect a solution? What are the advantages and disadvantages of each
method, and when would each technique be applied?
5. Investigate the computer software packages that are available at your computer center
for solution of conduction heat-transfer problems.

LIST OF WORKED EXAMPLES

3-1 Buried pipe
3-2 Cubical furnace
3-3 Buried disk
3-4 Buried parallel disks
3-5 Nine-node problem
3-6 Gauss-Seidel calculation
3-7 Numerical formulation with heat generation
3-8 Heat generation with nonuniform nodal elements
3-9 Composite material with nonuniform nodal elements
3-10 Radiation boundary condition
3-11 Use of variable mesh size
3-12 Three-dimensional numerical formulation

PROBLEMS
3-1 Beginning with the separation-of-variables solutions for λ2 = 0 and λ2 < 0 [Equations
(3-9) and (3-10)], show that it is not possible to satisfy the boundary conditions for
the constant temperature at y = H with either of these two forms of solution. That is,
show that, in order to satisfy the boundary conditions
T = T1 at y = 0
T = T1 at x = 0
T = T1 at x = W
T = T2 at y = H
either a trivial or physically unreasonable solution results when either Equation (3-9)
or (3-10) is used.
3-2 Write out the first four nonzero terms of the series solutions given in Equation
(3-20). What percentage error results from using only these four terms at y = H
and x = W/2?
3-3 A horizontal pipe having a surface temperature of 67◦ C and diameter of 25 cm is
buried at a depth of 1.2 m in the earth at a location where k = 1.8 W/m · ◦ C. The earth
surface temperature is 15◦ C. Calculate the heat lost by the pipe per unit length.
3-4 A 6.0-cm-diameter pipe whose surface temperature is maintained at 210◦ C passes
through the center of a concrete slab 45 cm thick. The outer surface temperatures of
the slab are maintained at 15◦ C. Using the flux plot, estimate the heat loss from the
pipe per unit length. Also work using Table 3-1.
3-5 A 2.5-cm-diameter pipe carrying condensing steam at 101 kPa passes through the
center of an infinite plate having a thickness of 5 cm. The plate is exposed to room

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.120 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 121

air at 27◦ C with a convection coefficient of 5.1 W/m2 · ◦ C on both sides. The plate is
composed of an insulation material having k = 0.1W/m · ◦ C. Calculate the heat lost
by the steam pipe per meter of length.
3-6 A heavy-wall tube of Monel, 2.5-cm ID and 5-cm OD, is covered with a 2.5-cm
layer of glass wool. The inside tube temperature is 300◦ C, and the temperature at
the outside of the insulation is 40◦ C. How much heat is lost per foot of length? Take
k = 11 Btu/h · ft · ◦ F for Monel.
3-7 A symmetrical furnace wall has the dimensions shown in Figure P3-7. Using the flux
plot, obtain the shape factor for this wall.

Figure P3-7

1m 3m

2m

4m

3-8 A furnace of 70- by 60- by 90-cm inside dimensions is constructed of a material

having a thermal conductivity of 0.5 Btu/h · ft · ◦ F. The wall thickness is 6 in. The
inner and outer surface temperatures are 500 and 100◦ F, respectively. Calculate the
heat loss through the furnace wall.
3-9 A cube 35 cm on each external side is constructed of fireclay brick. The wall thickness
is 5.0 cm. The inner surface temperature is 500◦ C, and the outer surface temperature
is 80◦ C. Compute the heat flow in watts.
3-10 Two long cylinders 8.0 and 3.0 cm in diameter are completely surrounded by a medium
with k = 1.4 W/m · ◦ C. The distance between centers is 10 cm, and the cylinders are
maintained at 200 and 35◦ C. Calculate the heat-transfer rate per unit length.
3-11 A 10-cm-diameter sphere maintained at 30◦ C is buried in the earth at a place where
k = 1.2 W/m · ◦ C. The depth to the centerline is 24 cm, and the earth surface temper-
ature is 0◦ C. Calculate the heat lost by the sphere.
3-12 A 20-cm-diameter sphere is totally enclosed by a large mass of glass wool. A heater
inside the sphere maintains its outer surface temperature at 170◦ C while the temper-
ature at the outer edge of the glass wool is 20◦ C. How much power must be supplied
to the heater to maintain equilibrium conditions?
3-13 A large spherical storage tank, 2 m in diameter, is buried in the earth at a location
where the thermal conductivity is 1.5 W/m · ◦ C. The tank is used for the storage of
an ice mixture at 0◦ C, and the ambient temperature of the earth is 20◦ C. Calculate
the heat loss from the tank.
3-14 A 2.5-cm-diameter pipe carrying condensing steam at 101 kPa passes through the
center of a square block of insulating material having k = 0.04 W/m · ◦ C. The block
is 5 cm on side and 2 m long. The outside of the block is exposed to room air at
27◦ C and a convection coefficient of h = 5.1 W/m2 · ◦ C. Calculate the heat lost by
the steam pipe.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.121 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

122 Problems

3-15 The solid shown in Figure P3-15 has the upper surface, including the half-cylinder
cutout, maintained at 100◦ C. At a large depth in the solid the temperature is 300 K;
k = 1 W/m · ◦ C. What is the heat transfer at the surface for the region where L = 30 cm
and D = 10 cm?

Figure P3-15

L
+

3-16 In certain locales, power transmission is made by means of underground cables. In

one example an 8.0-cm-diameter cable is buried at a depth of 1.3 m, and the resistance
of the cable is 1.1 × 10−4 /m. The surface temperature of the ground is 25◦ C, and
k = 1.2 W/m · ◦ C for earth. Calculate the maximum allowable current if the outside
temperature of the cable cannot exceed 110◦ C.
3-17 A copper sphere 4.0 cm in diameter is maintained at 70◦ C and submerged in a large
earth region where k = 1.3 W/m · ◦ C. The temperature at a large distance from the
sphere is 12◦ C. Calculate the heat lost by the sphere.
3-18 Two long, eccentric cylinders having diameters of 20 and 5 cm, respectively, are
maintained at 100 and 20◦ C and separated by a material with k = 2.5 W/m · ◦ C. The
distance between centers is 5.5 cm. Calculate the heat transfer per unit length between
the cylinders.
3-19 Two pipes are buried in the earth and maintained at temperatures of 200 and 100◦ C.
The diameters are 9 and 18 cm, and the distance between centers is 40 cm. Calculate
the heat-transfer rate per unit length if the thermal conductivity of earth at this location
is 1.1 W/m · ◦ C.
3-20 A hot sphere having a diameter of 1.5 m is maintained at 300◦ C and buried in a
material with k = 1.2 W/m · ◦ C and outside surface temperature of 30◦ C. The depth
of the centerline of the sphere is 3.75 m. Calculate the heat loss.
3-21 A scheme is devised to measure the thermal conductivity of soil by immersing a long
electrically heated rod in the ground in a vertical position. For design purposes, the rod
is taken as 2.5 cm in diameter with a length of 1 m. To avoid improper alteration of the
soil, the maximum surface temperature of the rod is 55◦ C while the soil temperature is
10◦ C. Assuming a soil conductivity of 1.7 W/m · ◦ C, what are the power requirements
of the electric heater in watts?
3-22 Two pipes are buried in an insulating material having k = 0.8 W/m · ◦ C. One pipe is
10 cm in diameter and carries a hot fluid at 300◦ C while the other pipe is 2.8 cm in
diameter and carries a cool fluid at 15◦ C. The pipes are parallel and separated by a
distance of 12 cm on centers. Calculate the heat-transfer rate between the pipes per
meter of length.
3-23 At a certain location the thermal conductivity of the earth is 1.5 W/m · ◦ C. At this
location an isothermal sphere having a temperature of 5◦ C and a diameter of 2.0 m
is buried at a centerline depth of 5.0 m. The earth temperature is 25◦ C. Calculate the
heat gained by the sphere.
3-24 Two parallel pipes are buried very deep in the earth at a location where they are
in contact with a rock formation having k = 3.5 W/m · ◦ C. One pipe has a diameter

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.122 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 123

of 20 cm and carries a hot fluid at 120◦ C while the other pipe has a diameter of
40 cm and carries a cooler fluid at 20◦ C. The distance between centers of the pipes is
1.0 m and both pipes are very long in respect to their diameters and spacing. Calculate
the conduction heat transfer between the two pipes per unit length of pipe. Express
as W/m length.
3-25 Steam pipes are sometimes carelessly buried in the earth without insulation. Consider
a 10-cm pipe carrying steam at 150◦ C buried at a depth of 23 cm to centerline. The
buried length is 100 m. Assuming that the earth thermal conductivity is 1.2 W/m · ◦ C
and the surface temperature is 15◦ C, estimate the heat lost from the pipe.
3-26 A hot steam pipe, 5 cm in diameter and carrying steam at 150◦ C, is placed in the center
of a 15-cm-thick slab of lightweight structural concrete. The outside of the concrete
slab is exposed to a convection environment that maintains the top and bottom of the
sheet at 20◦ C. Calculate the heat lost per unit length of pipe.
3-27 Seven 1.0-cm-diameter tubes carrying steam at 100◦ C are buried in a semi-infinite
medium having a thermal conductivity of 1.2 W/m · ◦ C and surface temperature of
25◦ C. The depth to the centerline of the tubes is 5 cm and the spacing between centers
is 3 cm. Calculate the heat lost per unit length for each tube.
3-28 Two parallel pipes 5 cm and 10 cm in diameter are totally surrounded by loosely
packed asbestos. The distance between centers for the pipes is 20 cm. One pipe
carries steam at 110◦ C while the other carries chilled water at 3◦ C. Calculate the heat
lost by the hot pipe per unit length.
3-29 A long cylinder has its surface maintained at 135◦ C and is buried in a material having
a thermal conductivity of 15.5 W/m · ◦ C. The diameter of the cylinder is 3 cm and
the depth to its centerline is 5 cm. The surface temperature of the material is 46◦ C.
Calculate the heat lost by the cylinder per meter of length.
3-30 A 2.5-m-diameter sphere contains a mixture of ice and water at 0◦ C and is buried in a
semi-infinite medium having a thermal conductivity of 0.2 W/m · ◦ C. The top surface
of the medium is isothermal at 30◦ C and the sphere centerline is at a depth of 8.5 m.
Calculate the heat lost by the sphere.
3-31 An electric heater in the form of a 50- by-100-cm plate is laid on top of a semi-infinite
insulating material having a thermal conductivity of 0.74 W/m · ◦ C. The heater plate is
maintained at a constant temperature of 120◦ C over all its surface, and the temperature
of the insulating material a large distance from the heater is 15◦ C. Calculate the heat
conducted into the insulating material.
3-32 A thin isothermal disk, having a diameter of 1.8 cm, is maintained at 40◦ C and buried
in a semi-infinite medium at a depth of 2 cm. The medium has a thermal conductivity
of 0.8 W/m · ◦ C and its surface is maintained at 15◦ C. Calculate the heat lost by
the disk.
3-33 Two parallel pipes, each having a diameter of 5 cm, carry steam at 120◦ C and chilled
water at 5◦ C, respectively, and are buried in an infinite medium of fiberglass blanket
(k = 0.04 W/m · ◦ C). Plot the heat transfer between the pipes per unit length as a
function of the centerline spacing between the pipes.
3-34 A small furnace has inside dimensions of 60 by 70 by 80 cm with a wall thickness of
5 cm. Calculate the overall shape factor for this geometry.
3-35 A 15-cm-diameter steam pipe at 150◦ C is buried in the earth near a 5-cm pipe car-
rying chilled water at 5◦ C. The distance between centers is 15 cm and the thermal
conductivity of the earth at this location may be taken as 0.7 W/m · ◦ C. Calculate the
heat lost by the steam pipe per unit length.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.123 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

124 Problems

3-36 Derive an equation equivalent to Equation (3-24) for an interior node in a three-
dimensional heat-flow problem.
3-37 Derive an equation equivalent to Equation (3-24) for an interior node in a one-
dimensional heat-flow problem.
3-38 Derive an equation equivalent to Equation (3-25) for a one-dimensional convection
boundary condition.
3-39 Considering the one-dimensional fin problems of Chapter 2, show that a nodal equa-
tion for nodes along the fin in the Figure P3-39 may be expressed as


hP(x)2 hP(x)2
Tm +2 − T∞ − (Tm−1 + Tm+1 ) = 0
kA kA
Figure P3-39

T∞
Δx

m−1 m m+1
Base

Δx

3-40 Show that the nodal equation corresponding to an insulated wall shown in
Figure P3-40 is
Tm,n+1 + Tm,n−1 + 2Tm−1,n − 4Tm,n = 0
Figure P3-40
Δx
2
m, n + 1 Insulated
surface

m − 1, n m, n Δ y

Δy
m, n − 1

Δx

Figure P3-41 3-41 For the insulated corner section shown in Figure P3-41, derive an expression for the
Insulated nodal equation of node (m, n) under steady-state conditions.
surfaces 3-42 Derive the equation in Table 3-2f.
3-43 Derive an expression for the equation of a boundary node subjected to a constant heat
m, n m+1, n flux from the environment. Use the nomenclature of Figure 3-7.
3-44 Set up the nodal equations for a modification of Example 3-7 in which the left half of
the wire is insulated and the right half is exposed to a connection environment with
m, n–1 m+1, n–1
h = 200 W/m2 · ◦ C and T = 20◦ C.
3-45 In a proposed solar-energy application, the solar flux is concentrated on a 5-cm-OD
stainless-steel tube [k = 16 W/m · ◦ C] 2 m long. The energy flux on the tube surface
is 20,000 W/m2 , and the tube wall thickness is 2 mm. Boiling water flows inside the
tube with a convection coefficient of 5000 W/m2 · ◦ C and a temperature of 250◦ C.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.124 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 125

Both ends of the tube are mounted in an appropriate supporting bracket, which main-
tains them at 100◦ C. For thermal-stress considerations the temperature gradient near
the supports is important. Assuming a one-dimensional system, set up a numerical
solution to obtain the temperature gradient near the supports.
3-46 An aluminum rod 2.5 cm in diameter and 15 cm long protrudes from a wall main-
tained at 300◦ C. The environment temperature is 38◦ C. The heat-transfer coefficient
is 17 W/m2 · ◦ C. Using a numerical technique in accordance with the result of Prob-
lem 3-39, obtain values for the temperature along the rod. Subsequently obtain the
heat flow from the wall at x = 0. Hint: The boundary condition at the end of the rod
may be expressed by



h x hP(x)2 h x hP(x)2
Tm + + 1 − T∞ + − Tm−1 = 0
k 2kA k 2kA

where m denotes the node at the tip of the fin. The heat flow at the base is

qx = 0 = −kA
x (Tm+1 − Tm )

where Tm is the base temperature and Tm+1 is the temperature of the first increment.
3-47 Repeat Problem 3-46, using a linear variation of heat-transfer coefficient between
base temperature and the tip of the fin. Assume h = 28 W/m2 · ◦ C at the base and
h = 11 W/m2 · ◦ C at the tip.
3-48 For the wall in Problem 3-6 a material with k = 1.4 W/m · ◦ C is used. The inner and
outer wall temperatures are 650 and 150◦ C, respectively. Using a numerical technique,
calculate the heat flow through the wall.
3-49 Repeat Problem 3-48, assuming that the outer wall is exposed to an environment at
38◦ C and that the convection heat-transfer coefficient is 17 W/m2 · ◦ C. Assume that
the inner surface temperature is maintained at 650◦ C.
3-50 Repeat Problem 3-4, using the numerical technique.
3-51 In the section illustrated in Figure P3-51 the surface 1-4-7 is insulated. The convection
heat transfer coefficient at surface 1-2-3 is 28 W/m2 · ◦ C. The thermal conductivity
of the solid material is 5.2 W/m · ◦ C. Using the numerical technique, compute the
temperatures at nodes 1, 2, 4, and 5.
Figure P3-51
Insulated

1 4 7
T ∞ = 0˚C
30 cm
2 5 8
h = 28 W/ m2 • ˚C

3 6 9
30 cm
T7 = T8 = T9 = 38˚C
T3 = T6 = 10˚C

3-52 A glass plate 3 by 12 by 12 in [k = 0.7 W/m · ◦ C] is oriented with the 12 by 12 face in

a vertical position. One face loses heat by convection to the surroundings at 70◦ F. The
other vertical face is placed in contact with a constant-temperature block at 400◦ F.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.125 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

126 Problems

The other four faces are insulated. The convection heat-transfer coefficient varies
approximately as
hx = 0.22(Ts − T∞ )1/4 x−1/4 Btu/h · ft 2 · ◦ F
where Ts and T∞ are in degrees Fahrenheit, Ts is the local surface temperature, and x
is the vertical distance from the bottom of the plate, measured in feet. Determine the
convection heat loss from the plate, using an appropriate numerical analysis.
Figure P3-53 3-53 In Figure P3-53, calculate the temperatures at points 1, 2, 3, and 4 using the numerical
700˚C method.
3-54 For the block shown in Figure P3-54, calculate the steady-state temperature distribu-
4 1 tion at appropriate nodal locations using the numerical method. k = 3.2 W/m · ◦ C.

100˚C 400˚C Figure P3-54

200˚C
2 3

Insulated
500˚C
8 cm

T∞ = 100˚C
h = 50 W/m2 • ˚C

150˚C
5 cm
3-55 The composite strip in Figure P3-55 is exposed to the convection environment at
300◦ C and h = 40 W/m2 · ◦ C. The material properties are kA = 20 W/m · ◦ C,
kB = 1.2 W/m · ◦ C, and kC = 0.5 W/m · ◦ C. The strip is mounted on a plate main-
tained at the constant temperature of 50◦ C. Calculate the heat transfer from the strip
to plate per unit length of strip. Assume two-dimensional heat flow.
Figure P3-55
T∞ = 300˚C

A 0.5 cm
B 1.5 cm

C T = 50˚C 2.0 cm

6.0 cm

3-56 The fin shown in Figure P3-56 has a base maintained at 300◦ C and is exposed to
the convection environment indicated. Calculate the steady-state temperatures of the
nodes shown and the heat loss if k = 1.0 W/m · ◦ C.
Figure P3-56
h = 40 W/m2 • ˚C T ∞ = 20˚C
1 2 3 4
1.0 cm
5 6 7 8
300˚C
1.0 cm

2 cm 2 cm
8 cm

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.126 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 127

3-57 Calculate the steady-state temperatures for nodes 1 to 16 in Figure P3-57. Assume
symmetry.
Figure P3-57
1 cm
1 2

3 4

1 cm
5 6

h = 30 W/m2 • ˚C
7 8
T∞ = 10˚C

9 10 11 12

1 cm
Insulated

Insulated
13 14 15 16

1 cm

0.5 1.5 cm 1 cm
200˚C cm
k = 10 W/m • ˚C

3-58 Calculate the steady-state temperatures for nodes 1 to 9 in Figure P3-58.

Figure P3-58

h = 25 W/ m2 •˚C
T∞ = 5˚C

1 2 3

4 5 6
T = 100˚C
Insulated

7 8 9

T = 100˚C

Δx = Δy = 25 cm
k = 2.3 W/ m •˚C

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.127 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

128 Problems

3-59 Calculate the steady-state temperatures for nodes 1 to 6 in Figure P3-59.

Figure P3-59
h = 12 W/ m2 •˚C
T∞ = 15˚C

1 2

3 4

T = 50˚C T = 50˚C

5 6

T = 50˚C
Δx = Δy = 25 cm
k = 1.5 W/ m •˚C

3-60 Calculate the temperatures for the nodes indicated in Figure P3-60. The entire outer
surface is exposed to the convection environment and the entire inner surface is at
a constant temperature of 300◦ C. Properties for materials A and B are given in the
figure.

Figure P3-60
T∞ = 10˚C
h = 125 W/m2 •˚C

1 2 3 4
A
5 6 7 8

B
9 10 11 12

T = 300˚C 13 14 15

kA = 10 W/m •˚C
kB = 40 W/m •˚C
Δ x = Δy = 1 cm

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.128 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 129

3-61 A rod having a diameter of 2 cm and a length of 10 cm has one end maintained at
200◦ C and is exposed to a convection environment at 25◦ C with h = 40 W/m2 · ◦ C.
The rod generates heat internally at the rate of 50 MW/m3 and the thermal con-
ductivity is 35 W/m · ◦ C. Calculate the temperatures of the nodes shown in the
Figure P3-61 assuming one-dimensional heat flow.
Figure P3-61

h = 40 W/m2 • C
T = 200˚C

1 2 3 4 5
2 cm

Δ x = 2 cm

3-62 Calculate the steady-state temperatures of the nodes in Figure P3-62. The entire outer
surface is exposed to the convection environment at 20◦ C and the entire inner surface
is constant at 500◦ C. Assume k = 0.2 W/m · ◦ C.
Figure P3-62

h = 10 W/m2 • ˚C
T = 20˚C

1 2 3 4 5

20 cm
6 7 8 9 10

11 12
500˚C
20 cm
13 14

40 cm
10 cm 10 cm 20 cm

k = 0.2 W/m • ˚C

3-63 Calculate the steady-state temperatures for the nodes indicated in Figure P3-63.
Figure P3-63

h = 75 W/m2 • ˚C
T∞ = 0˚C
1 cm

1 2 100˚C
3 4 0.25 cm
5 6
100˚C
Insulated
k = 4.0 W/m • ˚C

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.129 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

130 Problems

3-64 The two-dimensional solid shown in Figure P3-64 generates heat internally at the rate
of 90 MW/m3 . Using the numerical method calculate the steady-state nodal temper-
atures for k = 20 W/m · ◦ C.
Figure P3-64

h = 100 W/m2 • ˚C
T = 20˚C

1 2 3

T = 100˚C
4 5 6
Insulated

7 8 9

10 11 12

Insulated
Δ x = Δy = 1 cm
k = 20 W/m • ˚C
q• = 90 MW/m3

3-65 Two parallel disks having equal diameters of 30 cm are maintained at 120◦ C and
34◦ C. The disks are spaced a distance of 80 cm apart, on centers, and immersed in
a conducting medium having k = 3.4 W/m · ◦ C. Assuming that the disks exchange
heat only on the sides facing each other, calculate the heat lost by the hotter disk,
expressed in watts.
3-66 The half-cylinder has k = 20 W/m · ◦ C and is exposed to the convection environment
at 20◦ C. The lower surface is maintained at 300◦ C. Compute the temperatures for the
nodes shown in Figure P3-66 and the heat loss for steady state.
Figure P3-66

10 cm h = 50 W/m2 • ˚C
T∞ = 20˚C
1 2 3
5
4 7
6

3-67 A tube has diameters of 4 mm and 5 mm and a thermal conductivity 20 W/m · ◦ C.

Heat is generated uniformly in the tube at a rate of 500 MW/m3 and the outside
surface temperature is maintained at 100◦ C. The inside surface may be assumed to be
insulated. Divide the tube wall into four nodes and calculate the temperature at each
using the numerical method. Check with an analytical solution.
3-68 Repeat Problem 3-67 with the inside of the tube exposed to a convection condition
with h = 40 W/m2 · ◦ C. Check with an analytical calculation.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.130 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 131

3-69 Rework Problem 3-57 with the surface absorbing a constant heat flux of 300 W/m2
instead of the convection boundary condition. The bottom surface still remains
at 200◦ C.
3-70 Rework Problem 3-60 with the inner surface absorbing a constant heat flux of 300 W/m2
instead of being maintained at a constant temperature of 300◦ C.
3-71 Rework Problem 3-64 with the surface marked at a constant 100◦ C now absorbing a
constant heat flux of 500 W/m2 . Add nodes as necessary.
3-72 The tapered aluminum pin fin shown in Figure P3-72 is circular in cross section with
a base diameter of 1 cm and a tip diameter of 0.5 cm. The base is maintained at 200◦ C
and loses heat by convection to the surroundings at T∞ = 10◦ C, h = 200 W/m2 · ◦ C.
The tip is insulated. Assume one-dimensional heat flow and use the finite-difference
method to obtain the nodal equations for nodes 1 through 4 and the heat lost by the
fin. The length of the fin is 6 cm.

Figure P3-72

Insulated
1 2 3 4
T0 = 200˚C

3-73 Write the nodal equations 1 through 7 for the symmetrical solid shown in
Figure P3-73. x = y = 1 cm.

Figure P3-73

h, T∞

1 2

3 4 5 6 7 Insulated

T = 100˚C

3-74 Obtain the temperature for nodes 1 through 6 shown in Figure P3-74.
x = y = 1 cm.

Figure P3-74
T = 100˚C

T = 40˚C 1 3 5 T = 100˚C

2 4 6

T = 0˚C

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.131 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

132 Problems

3-75 Write the nodal equations for nodes 1 through 9 shown in Figure P3-75.
x = y = 1 cm.

Figure P3-75
T = 100˚C

Insulated 1 2 3 T = 50˚C

4 5 6

7 8 9

Constant
"
heat flux q冫A = q"

3-76 Write the nodal equation for nodes 1 through 12 shown in Figure P3-76. Express the
equations in a format for Gauss-Seidel iteration.

Figure P3-76

h, T∞

T = 50˚C k = 10 W/m • ˚C
1 2 3 4
5 6 7 8
T = 100˚C h = 30 W/m2 • ˚C

9 10 11 12
T = 150˚C T∞ = 15 ˚C

Insulated

3-77 Sometimes a square grid is desired even for a circular system. Consider the quadrant of
a circle shown in Figure P3-77 with r = 10 cm. x = y = 3 cm and k = 10 W/m · ◦ C.
Write the steady-state nodal equations for nodes 3 and 4. Make use of Tables 3-2
and 3-4.

Figure P3-77
h = 30 W冫m2 • ˚C, T∞ = 20˚C

7
r = 10 cm
3 Δ x = Δy = 3 cm
2 4 6
1
5

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.132 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 133

3-78 Taking Figure P3-78 as a special case of Table 3-2(f), write the nodal equations for
nodes (m, n) and 2 for the case of x = y.

Figure P3-78

h, T∞

1 Δy冫2
m, n m + 1, n

m, n − 1
Δx Δx
Δx = Δy

3-79 Repeat Problem 3-78 for a slanted surface that is insulated; i.e., h = 0.
3-80 If the slanted surface of Problem 3-78 is isothermal at T∞ , what is the nodal equation
for node (m, n)?
3-81 The slanted intersection shown in Figure P3-81 involves materials A and B. Write
steady-state nodal equations for nodes 3, 4, 5, and 6 using Table 3-2(f and g) as a
guide.

Figure P3-81

2
10
A
1 3 5 Δy冫2
11

4 Δx Δy
2

9 6 7
B Δy
8

Δx Δx
Δ x = Δy

3-82 A horizontal plate, 25 by 50 cm, is maintained at a constant temperature of 78◦ C and

buried in a semi-infinite medium at a depth of 5 m. The medium has an isothermal
surface maintained at 15◦ C and a thermal conductivity of 2.8 W/m · ◦ C. Calculate the
heat lost by the plate.
3-83 A cube 20 cm on a side is maintained at 80◦ C and buried in a large medium at 10◦ C
with a thermal conductivity of 2.3 W/m · ◦ C. Calculate the heat lost by the cube.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.133 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

134 Problems

How does this compare with the heat that would be lost by a 20-cm-diameter sphere?
Compare these heat transfers on a unit-volume basis.
3-84 A long horizontal cylinder having a diameter of 10 cm is maintained at a temperature
of 100◦ C and centered in a 30-cm-thick slab of material for which k = W/m · ◦ C. The
outside of the slab is at 20◦ C. Calculate the heat lost by the cylinder per unit length.
3-85 Work Problem 3-84 using the flux plot.
3-86 A horizontal plate 20 by 150 cm is buried in a large medium at a depth of 2.0 m and
maintained at 50◦ C. The surface of the medium is at 10◦ C and has k = 1.5 W/m · ◦ C.
Calculate the heat lost by the plate.
3-87 A thin disk 10 cm in diameter is maintained at 75◦ C and placed on the surface of
a large medium at 15◦ C with k = 3 W/m · ◦ C. Calculate the heat conducted into the
medium.
3-88 Repeat Problem 3-87 for a square 10 cm on a side. Compare the heat transfers on a
per unit area basis.
3-89 A hot steam pipe 10 cm in diameter is maintained at 200◦ C and centered in a square
mineral-fiber insulation 20 cm on a side. The outside surface temperature of the
insulation is 35◦ C. Calculate the heat lost by a 20-m length of pipe if the thermal
conductivity of the insulation can be taken as 50 mW/m · ◦ C.
3-90 A pipe having a diameter of 10 cm passes through the center of a concrete slab having
a thickness of 70 cm. The surface temperature of the pipe is maintained at 100◦ C by
condensing steam while the outer surfaces of the concrete are at 24◦ C. Calculate the
heat lost by the pipe per meter of length.
3-91 Consider a circumferential fin of rectangular profile as shown in Figure 2-12. Set up
nodal equations for a fin of thickness t, heat transfer coefficient h, thermal conductivity
k, and heat generation rate q as a function of radial coordinate r, taking increments
of r. Write the nodal equations for the node adjacent to the base temperature T0 , a
node in the middle of the fin, and the node at the end of the fin.
3-92 Set up a nodal equation for the geometry of Problem 2-123, using increments in the
height of the truncated cone as the one-dimensional variable. Then work the problem
with the numerical method and compare with the one-dimensional analytical solution.
3-93 Set up nodal equations for the geometry of Problem 2-122, using increments in an
angle θ as the one-dimensional variable. Then work the problem using the numerical
method and compare with the one-dimensional analytical solution.
3-94 A cube 30 cm on a side is buried in an infinite medium with a thermal conductivity
of 1.8 W/m · ◦ C. The surface temperature of the cube is 30◦ C while the temperature
of the medium is 10◦ C. Calculate the heat lost by the cube.
3-95 A thin horizontal disk having a diameter of 15 cm is maintained at a constant surface
temperature of 87◦ C and buried at a depth of 20 cm in a semi-infinite medium with
an adiabatic surface. The thermal conductivity of the medium is 2.7 W/m · ◦ C and the
temperature of the medium a large distance away from the disk (not the adiabatic the
surface temperature) is 13◦ C. Calculate the heat lost by the disk in watts.
3-96 A copper rod has an internal heater that maintains its surface temperature at 50◦ C
while it is buried vertically in a semi-infinite medium. The rod is 2 cm in diamter and
40 cm long and the isothermal surface of the medium is at 20◦ C. Calculate the heat
lost by the rod if the thermal conductivity of the medium is 3.4 W/m · ◦ C.
3-97 Rework Problem 2-122, using a numerical approach with five nodes operating in
increments of the radial angle θ, and compare with the analytical results of
Problem 2-122.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.134 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 135

Design-Oriented Problems
3-98 A liner of stainless steel (k = 20 W/m · ◦ C), having a thickness of 3 mm, is placed on
the inside surface of the solid in Problem 3-62. Assuming now that the inside surface
of the stainless steel is at 500◦ C, calculate new values for the nodal temperatures in
the low-conductivity material. Set up your nodes in the stainless steel as necessary.
3-99 A basement for a certain home is 4 × 5 m with a ceiling height of 3 m. The walls
are concrete having a thickness of 10 cm. In the winter the convection coefficient
on the inside is 10 W/m2 · ◦ C and the soil on the outside has k = 1.7 W/m · ◦ C.
Analyze this problem and determine an overall heat transfer coefficient U defined
by qloss = UAinside (Tinside − Tsoil ). Determine the heat loss when Tinside = 26◦ C and
Tsoil = 15◦ C.
3-100 A groundwater heat pump is a refrigeration device that rejects heat to the ground
through buried pipes instead of to the local atmosphere. The heat rejection rate for
such a machine at an Oklahoma location is to be 22 kW in a location where the
ground temperature at depth is 17◦ C. The thermal conductivity of the soil at this
location may be taken as 1.6 W/m · ◦ C. Water is to be circulated through a length of
horizontal buried pipe or tube with the water entering at 29◦ C and leaving at 23.5◦ C.
The convection coefficient on the inside of the pipe is sufficiently high that the inner
pipe wall temperature may be assumed to be the same as the water temperature.
Select an appropriate pipe/tube material, size, and length to accomplish the required
cooling. You may choose standard steel pipe sizes from Table A-11. Standard tubing
or plastic pipe sizes are obtained from other sources. Examine several choices before
making your final selection and give reasons for that selection.
3-101 Professional chefs claim that gas stove burners are superior to electric burners
because of the more uniform heating afforded by the gas flame and combustion prod-
ucts around the bottom of a cooking pan. Advocates of electric stoves note the lack of
combustion products to pollute the air in the cooking area, but acknowledge that gas
heat may be more uniform. Manufacturers of thick-bottomed cookware claim that
their products can achieve uniformity of cooking as good as gas heat because of the
“spreading” of heat through an 8-mm-thick aluminum layer on the bottom of the pan.
You are asked to verify this claim. For the evaluation assume a 200-mm-diameter
pan with an 8-mm-thick aluminum bottom and the interior exposed to boiling water,
which produces h = 1500 W/m2 · ◦ C at 1 atm (100◦ C). Observe the approximate
spacing for the circular element in an electric burner and devise an appropriate
numerical model to investigate the uniformity-of-heating claim. Consider such fac-
tors as contact resistance between the burner element and the bottom of the pan, and
radiation transfer that might be present. Consider different heating rates (different
burner element temperatures) and their effect. When the study is complete, make
recommendations as to what the cookware manufacturers might prudently claim for
their thick-bottomed product. Discuss uncertainties in your analysis.
3-102 The fin analyses of Section 2-10 assumed one-dimensional heat flows in the fins.
Devise a numerical model similar to that shown in Problem 3-57 to examine the
validity of this assumption. Restrict the analysis to aluminum with k = 200 W/m · ◦ C.
Examine several different combinations of fin thickness, fin length, and convection
coefficient to determine the relative effects on temperature variation across the fin
thickness. State conclusions as you think appropriate.
3-103 A small building 5 m wide by 7 m long by 3 m high (inside dimensions) is mounted
on a flat concrete slab having a thickness of 15 cm. The walls of the building are

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.135 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch03 10/29/2008 15:34

136 References

constructed of concrete also, with a thickness of 7 cm. The inside of the building is
used for cold storage at −20◦ C and the outside of the building is exposed to ambient
air at 30◦ C, with a convection coefficient of 15 W/m2 · ◦ C. The inside convection
coefficient for the building is estimated at 10 W/m2 · ◦ C and the floor slab is in contact
with earth having k = 1.8 W/m · ◦ C. The earth temperature may be assumed to be
15◦ C. Calculate the heat gained by the building in the absence of any insulating
material on the outside. Next, select two alternative insulation materials for the
outside of the building from Table 2-1 and/or Table A-3. The insulation objective
is to raise the outside surface temperature of the insulation to 26◦ C for the ambient
temperature of 30◦ C. The refrigeration system operates in such a manner that 1 kW
will produce 4000 kJ/hr of cooling, and electricity costs $0.085/kWh. Economics
dictates that the insulation should pay for itself in a three-year period. What is the
allowable cost per unit volume of insulation to accomplish this payback objective,
for the two insulating materials selected? Suppose an outside surface temperature of
24◦ C is chosen as the allowable value for the insulation. What would the allowable
costs be for a three-year payback in this case? Make your own assumptions as to the
annual hours of operation for the cooling system.

REFERENCES
1. Carslaw, H. S., and J. C. Jaeger. Conduction of Heat in Solids, 2d ed. Fair Lawn, NJ: Oxford
University Press, 1959.
2. Schneider, P. J. Conduction Heat Transfer. Reading, MA: Addison-Wesley, 1955.
3. Dusinberre, G. M. Heat Transfer Calculations by Finite Differences, Scranton, PA: International
Textbook, 1961.
4. Kayan, C. F. “Heat Transfer Temperature Patterns of a Multicomponent Structure by Comparative
Methods,” Trans ASME, vol. 71, p. 9, 1949.
5. Rudenberg, R. Die Ausbreitung der Luft-und Erdfelder und Hochspannungsleitungen, besonders
bei Erd-und Kurzschlussen, Elektrotech. Z., vol. 46, p. 1342, 1925.
6. Andrews, R. V. “Solving Conductive Heat Transfer Problems with Electrical-Analogue Shape
Factors,” Chem. Eng. Prog., vol. 51, no. 2, p. 67, 1955.
7. Sunderland, J. E., and K. R. Johnson. “Shape Factors for Heat Conduction through Bodies with
Isothermal or Convective Boundary Conditions,” Trans. ASHAE, vol. 70, pp. 237–41, 1964.
8. Richtmeyer, R. D. Difference Methods for Initial Value Problems. New York: Interscience
Publishers, 1957.
9. Crank, J., and P. Nicolson. “A Practical Method for Numerical Evaluation of Solutions of P. D.
E. of Heat Conduction Type,” Proc. Camb. Phil. Soc., vol. 43, p. 50, 1947.
10. Ozisik, M. N. Boundary Value Problems of Heat Conduction. Scranton, PA: International
Textbook, 1968.
11. Arpaci, V. S. Conduction Heat Transfer. Reading, MA: Addison-Wesley, 1966.
12. Ames, W. F. Nonlinear Partial Differential Equations in Engineering. New York: Academic
Press, 1965.
13. Myers, R. F. Conduction Heat Transfer. New York: McGraw-Hill, 1972.
14. Adams, J. A., and D. F. Rogers. Computer Aided Analysis in Heat Transfer. New York:
McGraw-Hill, 1973.
15. Rohsenow, W. M., and J. P. Hartnett, eds. Handbook of Heat Transfer. 2nd ed. New York:
McGraw-Hill, 1988.
16. Kern, D. Q., and A. D. Kraus. Extended Surface Heat Transfer. New York: McGraw-Hill, 1972.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.136 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

CHAPTER 3 Steady-State Conduction—Multiple Dimensions 137

17. Hahne, E., and U. Grigull. “Formfaktor und Formwiderstand der stationaren mehr-dimensionalen
Warmeleitung,” Int. J. Heat Mass Transfer, vol. 18, p. 751, 1975.
18. Chapra, S. C., and R. P. Canale. Numerical Methods for Engineers. 3rd ed. McGraw-Hill, 1996.
19. Constantinides, A. Applied Numerical Methods with Personal Computers. McGraw-Hill, 1987.
20. Patankar, S. V. Numerical Heat Transfer and Fluid Flow. Hemisphere Publishing, 1980.
21. Minkowycz, W. J., E. M. Sparrow, G. E. Schneider, and R. H. Pletcher. Handbook of Numerical
Heat Transfer. New York: Wiley, 1988.
22. ——. Mathcad 8, Cambridge, MA: Mathsoft, Inc., 1999.
23. ——. TK Solver, Rockford, Ill.: Universal Technical Systems, 1999.
24. Palm, W. MATLAB for Engineering Applications. New York: McGraw-Hill, 1999.
25. Gottfried, B. Spreadsheet Tools for Engineers—Excel 97 Version. New York: McGraw-Hill,
1998.
26. Holman, J. P. What Every Engineer Should Know About EXCEL, Chap. 5. Boca Raton, FL: CRC
Press, 2006.
27. Orvis, W. J., Excel for Scientists and Engineers, 2nd ed. San Francisco: SYBEX, 1996.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.137 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch03 10/14/2008 18:0

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.138 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

C H A P T E R

4 Unsteady-State Conduction

4-1 INTRODUCTION
If a solid body is suddenly subjected to a change in environment, some time must elapse
before an equilibrium temperature condition will prevail in the body. We refer to the equilib-
rium condition as the steady state and calculate the temperature distribution and heat transfer
by methods described in Chapters 2 and 3. In the transient heating or cooling process that
takes place in the interim period before equilibrium is established, the analysis must be
modified to take into account the change in internal energy of the body with time, and the
boundary conditions must be adjusted to match the physical situation that is apparent in the
unsteady-state heat-transfer problem. Unsteady-state heat-transfer analysis is obviously of
significant practical interest because of the large number of heating and cooling processes
that must be calculated in industrial applications.
To analyze a transient heat-transfer problem, we could proceed by solving the general
heat-conduction equation by the separation-of-variables method, similar to the analytical
treatment used for the two-dimensional steady-state problem discussed in Section 3-2. We
give one illustration of this method of solution for a case of simple geometry and then refer
the reader to the references for analysis of more complicated cases. Consider the infinite
plate of thickness 2L shown in Figure 4-1. Initially the plate is at a uniform temperature Ti ,
and at time zero the surfaces are suddenly lowered to T = T1 . The differential equation is

∂2 T 1 ∂T
= [4-1]
∂x2 α ∂τ

The equation may be arranged in a more convenient form by introduction of the variable
θ = T − T1 . Then
∂2 θ 1 ∂θ
= [4-2]
∂x2 α ∂τ
with the initial and boundary conditions

θ = θi = Ti − T1 at τ = 0, 0 ≤ x ≤ 2L [a]

θ=0 at x = 0, τ > 0 [b]

θ=0 at x = 2L, τ > 0 [c]

139

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.139 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

140 4-1 Introduction

Figure 4-1 Infinite plate

subjected to sudden
cooling of surfaces.

Ti

T1 x
2L

Assuming a product solution θ(x, τ) = X(x)H(τ) produces the two ordinary differential
equations
d2X
+ λ2 X = 0
dx2
dH
+ αλ2 H = 0
dτ
where λ2 is the separation constant. In order to satisfy the boundary conditions it is necessary
that λ2 > 0 so that the form of the solution becomes
θ = (C1 cos λx + C2 sin λx)e−λ
2 ατ

From boundary condition (b), C1 = 0 for τ > 0. Because C2 cannot also be zero, we find
from boundary condition (c) that sin 2Lλ = 0, or
nπ
λ= n = 1, 2, 3, . . .
2L
The final series form of the solution is therefore
∞
nπx
Cn e−[nπ/2L] ατ sin
2
θ=
2L
n=1

This equation may be recognized as a Fourier sine expansion with the constants Cn deter-
mined from the initial condition (a) and the following equation:

1 2L nπx 4
Cn = θi sin dx = θi n = 1, 3, 5, . . .
L 0 2L nπ
The final series solution is therefore
∞
θ T − T1 4  1 −[nπ/2L]2 ατ nπx
= = e sin n = 1, 3, 5 . . . [4-3]
θi Ti − T1 π n 2L
n=1

We note, of course, that at time zero (τ = 0) the series on the right side of Equation (4-3)
must converge to unity for all values of x.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.140 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 141

In Section 4-4, this solution will be presented in graphical form for calculation purposes.
For now, our purpose has been to show how the unsteady-heat-conduction equation can be
solved, for at least one case, with the separation-of-variables method. Further information
on analytical methods in unsteady-state problems is given in the references.

4-2 LUMPED-HEAT-CAPACITY SYSTEM

We continue our discussion of transient heat conduction by analyzing systems that may be
considered uniform in temperature. This type of analysis is called the lumped-heat-capacity
method. Such systems are obviously idealized because a temperature gradient must exist
in a material if heat is to be conducted into or out of the material. In general, the smaller
the physical size of the body, the more realistic the assumption of a uniform temperature
throughout; in the limit a differential volume could be employed as in the derivation of the
general heat-conduction equation.
If a hot steel ball were immersed in a cool pan of water, the lumped-heat-capacity
method of analysis might be used if we could justify an assumption of uniform ball tem-
perature during the cooling process. Clearly, the temperature distribution in the ball would
depend on the thermal conductivity of the ball material and the heat-transfer conditions
from the surface of the ball to the surrounding fluid (i.e., the surface-convection heat-
transfer coefficient). We should obtain a reasonably uniform temperature distribution in the
ball if the resistance to heat transfer by conduction were small compared with the convection
resistance at the surface, so that the major temperature gradient would occur through the
fluid layer at the surface. The lumped-heat-capacity analysis, then, is one that assumes that
the internal resistance of the body is negligible in comparison with the external resistance.
The convection heat loss from the body is evidenced as a decrease in the internal energy
of the body, as shown in Figure 4-2. Thus,
dT
q = hA(T − T∞ ) = −cρV [4-4]
dτ
where A is the surface area for convection and V is the volume. The initial condition is
written
T = T0 at τ = 0
so that the solution to Equation (4-4) is
T − T∞
= e−[hA/ρcV ]τ [4-5]
T0 − T∞

Figure 4-2 Nomenclature for single-lump heat-capacity

analysis.

dT
q = hA (T – T∞) = –cρV dττ
S
T0

1
Cth =ρ cV hA

T∞

(a) (b)

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.141 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

142 4-2 Lumped-Heat-Capacity System

where T∞ is the temperature of the convection environment. The thermal network for the
single-capacity system is shown in Figure 4-2b. In this network we notice that the thermal
capacity of the system is “charged” initially at the potential T0 by closing the switch S.
Then, when the switch is opened, the energy stored in the thermal capacitance is dissipated
through the resistance 1/hA. The analogy between this thermal system and an electric system
is apparent, and we could easily construct an electric system that would behave exactly like
the thermal system as long as we made the ratio
hA 1 1
= Rth = Cth = ρcV
ρcV Rth Cth hA
equal to 1/Re Ce , where Re and Ce are the electric resistance and capacitance, respectively.
In the thermal system we store energy, while in the electric system we store electric charge.
The flow of energy in the thermal system is called heat, and the flow of charge is called
electric current. The quantity cρV/hA is called the time constant of the system because it
has the dimensions of time. When
cρV
τ=
hA
it is noted that the temperature difference T − T∞ has a value of 36.8 percent of the initial
difference T0 − T∞ .
The reader should note that the lumped-capacity formulation assumes essentially uni-
form temperature throughout the solid at any instant of time so that the change in internal
energy can be represented by ρcVdT/dτ. It does not require that the convection boundary
condition have a constant value of h. In fact, variable values of h coupled with radiation
boundary conditions are quite common. The specification of “time constant” in terms of
the 36.8 percent value stated above implies a constant boundary condition.
For variable convection or radiation boundary conditions, numerical methods (see
Section 4-6) are used to advantage to predict lumped capacity behavior. A rather general
setup of a lumped-capacity solution using numerical methods and Microsoft Excel is given
in Section D-6 of the Appendix. In some cases, multiple lumped-capacity formulations can
be useful. An example involving the combined convection-radiation cooling of a box of
electronic components is also given in this same section of the Appendix.
Applicability of Lumped-Capacity Analysis
We have already noted that the lumped-capacity type of analysis assumes a uniform temper-
ature distribution throughout the solid body and that the assumption is equivalent to saying
that the surface-convection resistance is large compared with the internal-conduction resis-
tance. Such an analysis may be expected to yield reasonable estimates within about 5 percent
when the following condition is met:
h(V/A)
< 0.1 [4-6]
k
where k is the thermal conductivity of the solid. In sections that follow, we examine those
situations for which this condition does not apply. We shall see that the lumped-capacity
analysis has a direct relationship to the numerical methods discussed in Section 4-7. If one
considers the ratio V/A = s as a characteristic dimension of the solid, the dimensionless
group in Equation (4-6) is called the Biot number:
hs
= Biot number = Bi
k
The reader should recognize that there are many practical cases where the lumped-capacity
method may yield good results. In Table 4-1 we give some examples that illustrate the
relative validity of such cases.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.142 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch04 10/30/2008 17:34

CHAPTER 4 Unsteady-State Conduction 143

Table 4-1 Examples of lumped-capacity systems.

Approximate
value of h, h(V/A)
Physical situation k, W/m · ◦ C W/m2 · ◦ C k

1. 3.0-cm steel cube cooling in room air 40 7.0 8.75 × 10−4

2. 5.0-cm glass cylinder cooled by a 50-m/s airstream 0.8 180 2.81
3. Same as situation 2 but a copper cylinder 380 180 0.006
4. 3.0-cm hot copper cube submerged in water 380 10,000 0.132
such that boiling occurs

We may point out that uncertainties in the knowledge of the convection coefficient of
±25 percent are quite common, so that the condition Bi = h(V/A)/k < 0.1 should allow for
some leeway in application.
Do not dismiss lumped-capacity analysis because of its simplicity. Because of uncer-
tainties in the convection coefficient, it may not be necessary to use more elaborate analysis
techniques.
Steel Ball Cooling in Air EXAMPLE 4-1
A steel ball [c = 0.46 kJ/kg · ◦ C, k = 35 W/m · ◦ C] 5.0 cm in diameter and initially at a uniform
temperature of 450◦ C is suddenly placed in a controlled environment in which the temperature
is maintained at 100◦ C. The convection heat-transfer coefficient is 10 W/m2 · ◦ C. Calculate the
time required for the ball to attain a temperature of 150◦ C.
Solution
We anticipate that the lumped-capacity method will apply because of the low value of h and high
value of k. We can check by using Equation (4-6):
h(V/A) (10)[(4/3)π(0.025)3 ]
= = 0.0023 < 0.1
k 4π(0.025)2 (35)

so we may use Equation (4-5). We have

T = 150◦ C ρ = 7800 kg/m3 [486 lbm /ft 3 ]
T∞ = 100◦ C h = 10 W/m2 · ◦ C [1.76Btu/h · ft 2 · ◦ F]
T0 = 450◦ C c = 460 J/kg · ◦ C [0.11 Btu/lbm · ◦ F]

hA (10)4π(0.025)2
= = 3.344 × 10−4 s−1
ρcV (7800)(460)(4π/3)(0.025)3

T − T∞
= e−[hA/ρcV ]τ
T0 − T∞
150 − 100 −4τ
= e−3.344×10
450 − 100
τ = 5819 s = 1.62 h

4-3 TRANSIENT HEAT FLOW IN A

SEMI-INFINITE SOLID
Consider the semi-infinite solid shown in Figure 4-3 maintained at some initial temperature
Ti . The surface temperature is suddenly lowered and maintained at a temperature T0 , and we

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.143 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

144 4-3 Transient Heat Flow in a Semi-Infinite Solid

Figure 4-3 Nomenclature for transient

heat flow in a semi-infinite
solid.

T0
T1
∂T
qo = –kA
∂x x =0

seek an expression for the temperature distribution in the solid as a function of time. This
temperature distribution may subsequently be used to calculate heat flow at any x position
in the solid as a function of time. For constant properties, the differential equation for the
temperature distribution T(x, τ) is
∂2 T 1 ∂T
= [4-7]
∂x2 α ∂τ
The boundary and initial conditions are
T(x, 0) = Ti
T(0, τ) = T0 for τ > 0

This is a problem that may be solved by the Laplace-transform technique. The solution is
given in Reference 1 as
T(x, τ) − T0 x
= erf √ [4-8]
Ti − T0 2 ατ
where the Gauss error function is defined as
 √
x 2 x/2 ατ
e−η dη
2
erf √ = √ [4-9]
2 ατ π
It will be noted that in this definition η is a dummy variable and the integral is a function of
its upper limit. When the definition of the error function is inserted in Equation (4-8), the
expression for the temperature distribution becomes
 x/2√ατ
T(x, τ) − T0 2
e−η dη
2
=√ [4-10]
Ti − T0 π
The heat flow at any x position may be obtained from
∂T
qx = −kA
∂x
Performing the partial differentiation of Equation (4-10) gives
 
∂T 2 −x2/4ατ ∂ x
= (Ti − T0 ) √ e √ [4-11]
∂x π ∂x 2 ατ
Ti − T0 −x2/4ατ
= √ e
πατ

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.144 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 145

Figure 4-4 Response of semi-infinite solid to (a) sudden change in surface temperature and
(b) instantaneous surface pulse of Q0 /A J/m2 .

10

1
0.1
0.9
0.8
0.01
0.7
0.6 τ (sec)
0.001
0.5 0.01
0.4 0.05
0.0001 0.1
0.3 1.0
0.2 5
0.00001
0.1 20
0 100
0 0.5 1 1.5 2 2.5 0.000001
0.01 0.1 1 10
x
2 ατ x2
4ατ
(a) (b)

At the surface (x = 0) the heat flow is

kA(T0 − Ti )
q0 = √ [4-12]
πατ
The surface heat flux is determined by evaluating the temperature gradient at x = 0 from
Equation (4-11). A plot of the temperature distribution for the semi-infinite solid is given
in Figure 4-4. Values of the error function are tabulated in Reference 3, and an abbreviated
tabulation is given in Appendix A.

Constant Heat Flux on Semi-Infinite Solid

For the same uniform initial temperature distribution, we could suddenly expose the surface
to a constant surface heat flux q0 /A. The initial and boundary conditions on Equation (4-7)
would then become
T(x, 0) = Ti

q0 ∂T
= −k for τ > 0
A ∂x x=0

The solution for this case is

√  2  
2q0 ατ/π −x q0 x x
T − Ti = exp − 1 − erf √ [4-13a]
kA 4ατ kA 2 ατ

Energy Pulse at Surface

Equation (4-13a) presents the temperature response that results from a surface heat flux that
remains constant with time. A related boundary condition is that of a short, instantaneous
pulse of energy at the surface having a magnitude of Q0 /A. The resulting temperature
response is given by
T − Ti = [Q0 /Aρc(πατ)1/2 ] exp(−x2/4ατ) [4-13b]

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.145 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch04 10/30/2008 17:34

146 4-3 Transient Heat Flow in a Semi-Infinite Solid

In contrast to the constant-heat-flux case where the temperature increases indefinitely for
all x and times, the temperature response to the instantaneous surface pulse will die out with
time, or
T − Ti → 0 for all x as τ → ∞
This rapid exponential decay behavior is illustrated in Figure 4-4b.

Semi-Infinite Solid with Sudden Change

EXAMPLE 4-2 in Surface Conditions
A large block of steel [k = 45 W/m · ◦ C, α = 1.4 × 10−5 m2/s] is initially at a uniform temperature
of 35◦ C. The surface is exposed to a heat flux (a) by suddenly raising the surface temperature to
250◦ C and (b) through a constant surface heat flux of 3.2 × 105 W/m2 . Calculate the temperature
at a depth of 2.5 cm after a time of 0.5 min for both these cases.
Solution
We can make use of the solutions for the semi-infinite solid given as Equations (4-8) and (4-13a).
For case a,
x 0.025
√ = = 0.61
2 ατ (2)[(1.4 × 10−5 )(30)]1/2
The error function is determined from Appendix A as
x
erf √ = erf 0.61 = 0.61164
2 ατ

We have Ti = 35◦ C and T0 = 250◦ C, so the temperature at x = 2.5 cm is determined from

Equation (4-8) as
x
T(x, τ) = T0 + (Ti − T0 ) erf √
2 ατ
= 250 + (35 − 250)(0.61164) = 118.5◦ C

For the constant-heat-flux case b, we make use of Equation (4-13a). Since q0 /A is given as
3.2 × 105 W/m2 , we can insert the numerical values to give
(2)(3.2 × 105 )[(1.4 × 10−5 )(30)/π]1/2 −(0.61)2
T(x, τ) = 35 + e
45
(0.025)(3.2 × 105 )
− (1 − 0.61164)
45
= 79.3◦ C x = 2.5 cm, τ = 30 s

For the constant-heat-flux case the surface temperature after 30 s would be evaluated with x = 0
in Equation (4-13a). Thus,
(2)(3.2 × 105 )[(1.4 × 10−5 )(30)/π]1/2
T(x = 0) = 35 + = 199.4◦ C
45

EXAMPLE 4-3 Pulsed Energy at Surface of Semi-Infinite Solid

An instantaneous laser pulse of 10 MJ/m2 is imposed on a slab of stainless steel having properties
of ρ = 7800 kg/m3 , c = 460 J/kg · ◦ C, and α = 0.44 × 10−5 m2 /s. The slab is initially at a uniform
temperature of 40 ◦ C. Estimate the temperature at the surface and at a depth of 2.0 mm after a time
of 2 s.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.146 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 147

Solution
This problem is a direct application of Equation (4-13b). We have Q0 /A = 107 J/m2 and at x = 0
T0 − Ti = Q0 /Aρc(πατ)1.2
= 107/(7800)(460)[π(0.44 × 10−5 )(2)]0.5 = 530◦ C

and
T0 = 40 + 530 = 570◦ C
At x = 2.0 mm = 0.002 m,

T − Ti = (530)exp[−(0.002)2 /(4)(0.44 × 10−5 )(2)] = 473◦ C

and
T = 40 + 473 = 513◦ C

Heat Removal from Semi-Infinite Solid EXAMPLE 4-4

A large slab of aluminum at a uniform temperature of 200◦ C suddenly has its surface temperature
lowered to 70◦ C. What is the total heat removed from the slab per unit surface area when the
temperature at a depth 4.0 cm has dropped to 120◦ C?
Solution
We first find the time required to attain the 120◦ C temperature and then integrate Equation (4-12)
to find the total heat removed during this time interval. For aluminum,
α = 8.4 × 10−5 m2 /s k = 215 W/m · ◦ C [124 Btu/h · ft · ◦ F]

We also have
Ti = 200◦ C T0 = 70◦ C T(x, τ) = 120◦ C

Using Equation (4-8) gives

120 − 70 x
= erf √ = 0.3847
200 − 70 2 ατ

From Figure 4-4 or Appendix A,

x
√ = 0.3553
2 ατ
and
(0.04)2
τ= = 37.72 s
(4)(0.3553)2 (8.4 × 10−5 )
The total heat removed at the surface is obtained by integrating Equation (4-12):
 τ  τ 
Q0 q0 k(T0 − Ti ) τ
= dτ = √ dτ = 2k(T0 − Ti )
A 0 A 0 πατ πα
 1/2
37.72
= (2)(215)(70 − 200) = −21.13 × 106 J/m2 [−1861 Btu/ft2 ]
π(8.4 × 10−5 )

4-4 CONVECTION BOUNDARY CONDITIONS

In most practical situations the transient heat-conduction problem is connected with a con-
vection boundary condition at the surface of the solid. Naturally, the boundary conditions
for the differential equation must be modified to take into account this convection heat

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.147 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

148 4-4 Convection Boundary Conditions

transfer at the surface. For the semi-infinite-solid problem, the convection boundary con-
dition would be expressed by

Heat convected into surface = heat conducted into surface

or 
∂T
hA(T∞ − T)x=0 = −kA [4-14]
∂x x=0

The solution for this problem is rather involved and is worked out in detail by Schneider
[1]. The result is
     √ 
T − Ti hx h2 ατ h ατ
= 1 − erf X − exp + 2 × 1 − erf X + [4-15]
T∞ − Ti k k k

where
√
X = x/(2 ατ)
Ti = initial temperature of solid
T∞ = environment temperature
This solution is presented in graphical form in Figure 4-5.
Solutions have been worked out for other geometries. The most important cases are
those dealing with (1) plates whose thickness is small in relation to the other dimensions,
(2) cylinders where the diameter is small compared to the length, and (3) spheres. Results
of analyses for these geometries have been presented in graphical form by Heisler [2],
and nomenclature for the three cases is illustrated in Figure 4-6. In all cases the convection
environment temperature is designated as T∞ and the center temperature for x = 0 or r = 0 is
T0 .At time zero, each solid is assumed to have a uniform initial temperature Ti . Temperatures
in the solids are given in Figures 4-7 to 4-13 as functions of time and spatial position. In
these charts we note the definitions

θ = T(x, τ) − T∞ or T(r, τ) − T∞
θi = Ti − T∞
θ0 = T0 − T∞

If a centerline temperature is desired, only one chart is required to obtain a value for θ0 and
then T0 . To determine an off-center temperature, two charts are required to calculate the
product
θ θ0 θ
=
θi θi θ0

For example, Figures 4-7 and 4-10 would be employed to calculate an off-center temperature
for an infinite plate.
The heat losses for the infinite plate, infinite cylinder, and sphere are given in
Figures 4-14 to 4-16, where Q0 represents the initial internal energy content of the body in
reference to the environment temperature

Q0 = ρcV(Ti − T∞ ) = ρcVθi [4-16]

In these figures Q is the actual heat lost by the body in time τ.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.148 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch04 10/30/2008 17:34

Figure 4-5 Temperature distribution in the semi-infinite solid with convection boundary condition.

0.1

h(ατ )1/2/k
1− (T− T∞)/(Ti − T∞) = 1− q
qi

∞
3
0.01 2
1

0.5
0.3
0.2

0.001 0.1

0.05

0.0001
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6
x
(4ατ )1/2

Figure 4-6 Nomenclature for one-dimensional solids suddenly subjected to convection

environment at T∞ : (a) infinite plate of thickness 2L; (b) infinite cylinder of
radius r0 ; (c) sphere of radius r0 .

x
r r0
+ +
r0 r
L L
T0 = centerline temperature T0 = centerline axis temperature T0 = center temperature
(a) (b) (c)

149

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.149 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
 150
hol29362_Ch04

# 101675
10/30/2008
20:6

Title: Heat Transfer 10/e

Cust: McGraw-Hill
Server:
Figure 4-7 Midplane temperature for an infinite plate of thickness 2L: (a) full scale.

1.0
0.7

Au: Holman
0.5
0.4
0.3 90 100
80
0.2 14
70
12 60
10
0.1 50

Pg. No.150
9

1.
0
0.07 8 40 45

0.
8
7

0.
0.05 0. 6 30
35
0.04 7
5

6 0.5
0.03
25

0.4 0.3
20

0.1
0.02

0
1 6

0.2
3
18

= (T0 − T∞)/(Ti − T∞)

0.0
2.5

θi
θ0
0.01

1.2
L
k冫h

2.0 1.8
1.6 1.4
0.007
0.005

K/PMS 293
0.004

Short / Normal / Long

0.003
0.002

0.001
0 1 2 3 4 6 8 10 12 14 16 18 20 22 24 26 28 30 40 50 60 70 80 90 100 110 130 150 200 300 400 500 600 700
ατ 冫L2 = Fo
(a)

Publishing Services
DESIGN SERVICES OF
S4CARLISLE
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 151

Figure 4-7 (Continued). (b) expanded scale for 0 < Fo < 4, from Reference 2.

1.0 100
25
18
16
0.7 10
8
7
θ 0  θ i = (T0 − T∞)(Ti − T∞)

6
0.5 5
4
0.4

khL = 1Bi
3
0.3
2.5

0.2 2.0
1.8

1.6
1.4
0.1 0.1 0.2 0.3 0.4 0.5 0.6 0.8 1.0 1.2
0 1 2 3 4
ατ = Fo
L2
(b)

If one considers the solid as behaving as a lumped capacity during the cooling or heating
process, that is, small internal resistance compared to surface resistance, the exponential
cooling curve of Figure 4-5 may be replotted in expanded form, as shown in Figure 4-13
using the Biot-Fourier product as the abscissa. We note that the following parameters apply
for the bodies considered in the Heisler charts.

(A/V)inf plate = 1/L

(A/V)inf cylinder = 2/r0
(A/V)sphere = 3/r0

Obviously, there are many other practical heating and cooling problems of interest. The
solutions for a large number of cases are presented in graphical form by Schneider [7], and
readers interested in such calculations will find this reference to be of great utility.

The Biot and Fourier Numbers

A quick inspection of Figures 4-5 to 4-16 indicates that the dimensionless temperature
profiles and heat flows may all be expressed in terms of two dimensionless parameters
called the Biot and Fourier numbers:
hs
Biot number = Bi =
k
ατ kτ
Fourier number = Fo = 2 =
s ρcs2

In these parameters s designates a characteristic dimension of the body; for the plate it is
the half-thickness, whereas for the cylinder and sphere it is the radius. The Biot number
compares the relative magnitudes of surface-convection and internal-conduction resistances

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.151 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
 152
hol29362_Ch04

# 101675
10/30/2008

Title: Heat Transfer 10/e

17:34

Figure 4-8 Axis temperature for an infinite cylinder of radius r0 : (a) full scale.

Cust: McGraw-Hill
Server:
1.0
0.7
0.5
0.4

Au: Holman
5
0.3
3.5 4
0.2 25
2.5 3.0
20
18 100
0.1 1. 2 16
8 .0

Pg. No.152
14 90

1 .
0.07 12 80

4
1.6
70
0.05

10 9
60
0.04

1.2
7

1.0
6
0.03
4

0.8
50 5
40

0.6
0
0.02

.
35

= (T0 − T∞)/(Ti − T∞)

4
30

θi
θ0
0.5

0
0.3
0.01

0.2 0.1
0.007 r0
kⲐh

K/PMS 293
0.005

Short / Normal / Long

0.004
0.003
0.002

0.001
0 1 2 3 4 6 8 10 12 14 16 18 20 22 24 26 28 30 40 50 60 70 80 90 100 110 120 130 140 150 200 300 350

Publishing Services
ατ = Fo

DESIGN SERVICES OF
S4CARLISLE
r02
(a)
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 153

Figure 4-8 (Continued). (b) expanded scale for 0 < Fo < 4, from Reference 2.

1.0 100
50
25
0.7 20
16
14
θ0 θi = (T0 − T∞)(Ti − T∞)

12
0.5
9
0.4 8

khr0 = 1Bi
7
0.3 6
5

0.2 4
3.5
3.0

2.5
0.1 0 0.2 0.4 0.6 0.8 1.0 1.2 1.6 1.8 2.0
0 1 2 3 4
ατ = Fo
r02
(b)

to heat transfer. The Fourier modulus compares a characteristic body dimension with an
approximate temperature-wave penetration depth for a given time τ.
A very low value of the Biot modulus means that internal-conduction resistance is
negligible in comparison with surface-convection resistance. This in turn implies that the
temperature will be nearly uniform throughout the solid, and its behavior may be approxi-
mated by the lumped-capacity method of analysis. It is interesting to note that the exponent
of Equation (4-5) may be expressed in terms of the Biot and Fourier numbers if one takes
the ratio V/A as the characteristic dimension s. Then,
hA hτ hs kτ
τ= = = Bi Fo
ρcV ρcs k ρcs2

Applicability of the Heisler Charts

The calculations for the Heisler charts were performed by truncating the infinite series
solutions for the problems into a few terms. This restricts the applicability of the charts to
values of the Fourier number greater than 0.2.
ατ
Fo = > 0.2
s2
For smaller values of this parameter the reader should consult the solutions and charts given
in the references at the end of the chapter. Calculations using the truncated series solutions
directly are discussed in Appendix C.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.153 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
 154
hol29362_Ch04

# 101675
10/30/2008

Title: Heat Transfer 10/e

17:34

Cust: McGraw-Hill
Figure 4-9 Center temperature for a sphere of radius r0 : (a) full scale.

Server:
1.0
0.7
0.5
0.4

Au: Holman
0.3
90 100
0.2 70 80
3.5 14
12 50 6900
2.6
0.1 2.8

Pg. No.154
40
2. 2 10
0.07 2 .4 - 45

30
9
0.05
35
8

20
16
2.0
0.04
1.8 7 25
0.03
18
1. 6
6
0.02

= (T0 − T∞)/(Ti − T∞)

5

1.

θi
θ0
4
4

1.2
0.01
0.007

K/PMS 293
r0

1.0

Short / Normal / Long

0.005 k冫h

0.7
5
0.004

0.05
0.5
0.003

0.4
0.3
0.002

0.2
0
0.001
0 0.5 1.0 1.5 2 2.5 3 4 5 6 7 8 9 10 15 20 25 30 35 40 45 50 90 130 170 210 250

Publishing Services
DESIGN SERVICES OF
ατ Ⲑr02 = Fo

S4CARLISLE
(a)
hol29362_ch04 10/14/2008 19:33

Figure 4-9 (Continued). (b) expanded scale for 0 < Fo < 3, from Reference 2.

1.0
100
50
35
0.7 30
25
18
14
0.5
θ0 /θ i = (T0 − T∞)/(Ti − T∞)

12
10
0.4 9

k/hr0 = 1/Bi
8
0.3 7
6

0.2
5

0.1 0.5 0.75 1.0 1.2 1.6 2.0 2.4 2.8 3.0 3.5
0 0.1 0.05 0.2 0.35
0 0.5 1.0 1.5 2.0 2.5 3.0
ατ = Fo
r20
(b)

Figure 4-10 Temperature as a function of center temperature in an

infinite plate of thickness 2L, from Reference 2.

0
1.0
x / L = 0.2
0.9

0.8 0.4
θ θ 0 = (T − T∞)/(T0 − T∞)

0.7

0.6 0.6

0.5

0.4
0.8
0.3
0.9
0.2

0.1 1.0

0
0.01 0.02 0.05 0.1 0.2 0.5 1.0 2 3 5 10 20 50 100
k = 1
hL Bi

155

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.155 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

Figure 4-11 Temperature as a function of axis temperature in an

infinite cylinder of radius r0 , from Reference 2.

0
1.0 r/ r = 0.2
0

0.9
0.4
0.8

θ /θ 0 = (T − T∞)/(T0 − T∞)
0.7

0.6 0.6

0.5

0.4
0.8
0.3
0.9
0.2

0.1 1.0

0
0.01 0.02 0.05 0.1 0.2 0.5 1.0 2 3 5 10 20 50 100
k = 1
hr0 Bi

Figure 4-12 Temperature as a function of center temperature for a

sphere of radius r0 , from Reference 2.

0
1.0
r/r0 = 0.2

0.9

0.8 0.4

0.7
= (T − T∞)/(T0 − T∞)

0.6
0.6
0.5

0.4
θ0
θ

0.3 0.8

0.2 0.9

0.1
1.0
0
0.01 0.02 0.05 0.1 0.2 0.5 1.0 2 3 5 10 20 50 100
k = 1
hr0 Bi

156

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.156 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

Figure 4-13 Temperature variation with time for solids that may be
treated as lumped capacities: (a) 0 < BiFo < 10,
(b) 0.1 < BiFo < 1.0, (c) 0 < BiFo < 0.1.
Note: (A/V)inf plate = 1/L, (A/V)inf cyl = 2/r0 ,
(A/V)sphere = 3/r0 . See Equations (4-5) and (4-6).

0.1

0.01

θ
θi

0.001

0.0001

0.00001
0 1 2 3 4 5 6 7 8 9 10
h(A/V)τ
BiFo = ␳c
(a)
1

0.9

0.8

0.7
θ
θi
0.6

0.5

0.4

0.3
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
h(A/V)τ
BiFo = ␳c
(b)

157

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.157 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

Figure 4-13 (Continued).

0.99

0.98

0.97

0.96

θ
θi 0.95

0.94

0.93

0.92

0.91

0.9
0 0.02 0.04 0.06 0.08 0.1
h(A/V)τ
BiFo = ␳c
(c)

Figure 4-14 Dimensionless heat loss Q/Q0 of an infinite plane of thickness 2L with time,
from Reference 6.
1.0
0.9
0.8
0.7
001

0.2
0.02

0.1
0.05

0.5
5
0.01
2
0.00
0.00

1
2

0.6
10

20

50
= 0.

Q
Q0 0.5
hL /k

0.4
0.3
0.2
0.1
0
10 −5 10 −4 10 −3 10 −2 10 −1 1 10 10 2 10 3 10 4
h 2ατ
= Fo Bi2
k2

158

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.158 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 159

Figure 4-15 Dimensionlesss heat loss Q/Q0 of an infinite cylinder of radius r0 with time,
from Reference 6.
1.0
0.9
0.8
0.7

0.2
0.02

0.1
0.05

0.5
5
0.01
2
0.00

0.00
0 .0 0

1
2

5
0.6

10

20

50
Q
0 k=

Q0 0.5
hr /

0.4
0.3
0.2
0.1
0
10 −5 10 −4 10 −3 10 −2 10 −1 1 10 10 2 10 3 10 4
h2ατ
= Fo Bi2
k2

Figure 4-16 Dimensionless heat loss Q/Q0 of a sphere of radius r0 with time, from
Reference 6.
1.0
0.9
0.8
0.7
0.00 1

0.2
0.02

0.1
0.05

0.5
5
0.01
2
0.00

0.00

1
2

0.6
10

20

50

Q
0 k=

Q0 0.5
hr /

0.4
0.3
0.2
0.1
0
10 −5 10 −4 10 −3 10 −2 10 −1 1 10 10 2 10 3 10 4
h2ατ
= Fo Bi2
k2

Sudden Exposure of Semi-Infinite

Slab to Convection EXAMPLE 4-5
The slab of Example 4-4 is suddenly exposed to a convection-surface environment of 70◦ C with
a heat-transfer coefficient of 525 W/m2 · ◦ C. Calculate the time required for the temperature to
reach 120◦ C at the depth of 4.0 cm for this circumstance.
Solution
We may use either Equation (4-15) or Figure 4-5 for solution of this problem, but Figure 4-5 is
easier to apply because the time appears in two terms. Even when the figure
√ is used, an iterative
√
procedure is required because the time appears in both of the variables h ατ/k and x/(2 ατ).

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.159 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

160 4-4 Convection Boundary Conditions

We seek the value of τ such that

T − Ti 120 − 200
= = 0.615 [a]
T∞ − Ti 70 − 200

We therefore try values of τ and obtain readings of the temperature ratio from Figure 4-5 until
agreement with Equation (a) is reached. The iterations are listed below. Values of k and α are
obtained from Example 4-4.
√ x
h ατ T − Ti
√ from Figure 4-5
τ, s k 2 ατ T∞ − Ti
1000 0.708 0.069 0.41
3000 1.226 0.040 0.61
4000 1.416 0.035 0.68

Consequently, the time required is approximately 3000 s.

EXAMPLE 4-6 Aluminum Plate Suddenly Exposed to Convection

A large plate of aluminum 5.0 cm thick and initially at 200◦ C is suddenly exposed to the convection
environment of Example 4-5. Calculate the temperature at a depth of 1.25 cm from one of the faces
1 min after the plate has been exposed to the environment. How much energy has been removed
per unit area from the plate in this time?
Solution
The Heisler charts of Figures 4-7 and 4-10 may be used for solution of this problem. We first
calculate the center temperature of the plate, using Figure 4-7, and then use Figure 4-10 to calculate
the temperature at the specified x position. From the conditions of the problem we have
θi = Ti − T∞ = 200 − 70 = 130 α = 8.4 × 10−5 m2/s [3.26 ft 2/h]
2L = 5.0 cm L = 2.5 cm τ = 1 min = 60 s
k = 215 W/m · ◦ C [124 Btu/h · ft · ◦F]
h = 525 W/m2 · ◦ C [92.5 Btu/h · ft 2 · ◦F]
x = 2.5 − 1.25 = 1.25 cm

Then
ατ (8.4 × 10−5 )(60) k 215
= = 8.064 = = 16.38
L2 (0.025)2 hL (525)(0.025)
x 1.25
= = 0.5
L 2.5
From Figure 4-7
θ0
= 0.61
θi
θ0 = T0 − T∞ = (0.61)(130) = 79.3

From Figure 4-10 at x/L = 0.5,

θ
= 0.98
θ0
and
θ = T − T∞ = (0.98)(79.3) = 77.7
T = 77.7 + 70 = 147.7◦ C

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.160 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 161

We compute the energy lost by the slab by using Figure 4-14. For this calculation we require the
following properties of aluminum:
ρ = 2700 kg/m3 c = 0.9 kJ/kg · ◦ C

For Figure 4-14 we need

h2 ατ (525)2 (8.4 × 10−5 )(60) hL (525)(0.025)
= = 0.03 = = 0.061
k2 (215)2 k 215

From Figure 4-14

Q
= 0.41
Q0
For unit area
Q0 ρcVθi
= = ρc(2L)θi
A A
= (2700)(900)(0.05)(130)
= 15.8 × 106 J/m2

so that the heat removed per unit surface area is

Q
= (15.8 × 106 )(0.41) = 6.48 × 106 J/m2 [571 Btu/ft 2 ]
A

Long Cylinder Suddenly Exposed to Convection EXAMPLE 4-7

A long aluminum cylinder 5.0 cm in diameter and initially at 200◦ C is suddenly exposed to a
convection environment at 70◦ C and h = 525 W/m2 · ◦ C. Calculate the temperature at a radius of
1.25 cm and the heat lost per unit length 1 min after the cylinder is exposed to the environment.
Solution
This problem is like Example 4-6 except that Figures 4-8 and 4-11 are employed for the solution.
We have
θi = Ti − T∞ = 200 − 70 = 130 α = 8.4 × 10−5 m2/s
r0 = 2.5 cm τ = 1 min = 60 s
k = 215 W/m · ◦ C h = 525 W/m2 · ◦ C r = 1.25 cm
ρ = 2700 kg/m3 c = 0.9 kJ/kg · ◦ C

We compute
ατ (8.4 × 10−5 )(60) k 215
= = 8.064 = = 16.38
r2 (0.025)2 hr0 (525)(0.025)
0
r 1.25
= = 0.5
r0 2.5

From Figure 4-8

θ0
= 0.38
θi
and from Figures 4-11 at r/r0 = 0.5
θ
= 0.98
θ0

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.161 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

162 4-5 Multidimensional Systems

so that
θ θ θ
= 0 = (0.38)(0.98) = 0.372
θi θi θ0
and
θ = T − T∞ = (0.372)(130) = 48.4
T = 70 + 48.4 = 118.4◦ C

To compute the heat lost, we determine

h2 ατ (525)2 (8.4 × 10−5 )(60) hr0 (525)(0.025)
= = 0.03 = = 0.061
k2 (215)2 k 215

Then from Figure 4-15

Q
= 0.65
Q0
For unit length
Q0 ρcVθi
= = ρcπr02 θi = (2700)(900)π(0.025)2 (130) = 6.203 × 105 J/m
L L
and the actual heat lost per unit length is
Q
= (6.203 × 105 )(0.65) = 4.032 × 105 J/m [116.5 Btu/ft]
L

4-5 MULTIDIMENSIONAL SYSTEMS

The Heisler charts discussed in Section 4-4 may be used to obtain the temperature distri-
bution in the infinite plate of thickness 2L, in the long cylinder, or in the sphere. When a
wall whose height and depth dimensions are not large compared with the thickness or a
cylinder whose length is not large compared with its diameter is encountered, additional
space coordinates are necessary to specify the temperature, the charts no longer apply, and
we are forced to seek another method of solution. Fortunately, it is possible to combine the
solutions for the one-dimensional systems in a very straightforward way to obtain solutions
for the multidimensional problems.
It is clear that the infinite rectangular bar in Figure 4-17 can be formed from two
infinite plates of thickness 2L1 and 2L2 , respectively. The differential equation governing
this situation would be
∂2 T ∂2 T 1 ∂T
+ 2 = [4-17]
∂x 2 ∂z α ∂τ
and to use the separation-of-variables method to effect a solution, we should assume a
product solution of the form
T(x, z, τ) = X(x)Z(z)
(τ)

It can be shown that the dimensionless temperature distribution may be expressed as a

product of the solutions for two plate problems of thickness 2L1 and 2L2 , respectively:
     
T − T∞ T − T∞ T − T∞
= [4-18]
Ti − T∞ bar Ti − T∞ 2L1 plate Ti − T∞ 2L2 plate

where Ti is the initial temperature of the bar and T∞ is the environment temperature.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.162 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 163

Figure 4-17 Infinite rectangular bar.

z x

2L 2
2L 1

For two infinite plates the respective differential equations would be

∂2 T1 1 ∂T1 ∂2 T2 1 ∂T2
= = [4-19]
∂x2 α ∂τ ∂z2 α ∂τ
and the product solutions assumed would be
T1 = T1 (x, τ) T2 = T2 (z, τ) [4-20]

We shall now show that the product solution to Equation (4-17) can be formed from a simple
product of the functions (T1 , T2 ), that is,
T(x, z, τ) = T1 (x, τ)T2 (z, τ) [4-21]

The appropriate derivatives for substitution in Equation (4-17) are obtained from Equa-
tion (4-21) as
∂2 T ∂ 2 T1 ∂2 T ∂ 2 T2
= T2 = T1
∂x2 ∂x2 ∂z2 ∂z2
∂T ∂T2 ∂T1
= T1 + T2
∂τ ∂τ ∂τ
Using Equations (4-19), we have
∂T ∂ 2 T2 ∂ 2 T1
= αT1 2 + αT2 2
∂τ ∂z ∂x
Substituting these relations in Equation (4-17) gives
 
∂2 T1 ∂ 2 T2 1 ∂ 2 T2 ∂ 2 T1
T2 + T1 2 = αT1 2 + αT2 2
∂x2 ∂z α ∂z ∂x

or the assumed product solution of Equation (4-21) does indeed satisfy the original dif-
ferential equation (4-17). This means that the dimensionless temperature distribution for
the infinite rectangular bar may be expressed as a product of the solutions for two plate
problems of thickness 2L1 and 2L2 , respectively, as indicated by Equation (4-18).
In a manner similar to that described above, the solution for a three-dimensional block
may be expressed as a product of three infinite-plate solutions for plates having the thickness
of the three sides of the block. Similarly, a solution for a cylinder of finite length could be

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.163 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

164 4-5 Multidimensional Systems

expressed as a product of solutions of the infinite cylinder and an infinite plate having
a thickness equal to the length of the cylinder. Combinations could also be made with
the infinite-cylinder and infinite-plate solutions to obtain temperature distributions in semi-
infinite bars and cylinders. Some of the combinations are summarized in Figure 4-18, where
C(
) = solution for infinite cylinder
P(X) = solution for infinite plate
S(X) = solution for semi-infinite solid

Figure 4-18 Product solutions for

temperatures in
multidimensional systems:
(a) semi-infinite plate;
(b) infinite rectangular bar;
(c) semi-infinite rectangular
bar; (d ) rectangular
parallelepiped;
(e) semi-infinite cylinder;
( f ) short cylinder.

P (X ) S (X 1 ) P (X1 ) S (X2 )

2L 1 2L 2 2L 1
(a) (b)

S (X ) P (X1) P (X2) P (X1) P (X2 ) P (X3)

2L 3

2L2 2L1 2L2 2L1

(c) (d )

C (Θ) P (X )
C (Θ) S (X )

2L
x

2r0 2r0
(e) (f)

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.164 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch04 10/30/2008 20:6

CHAPTER 4 Unsteady-State Conduction 165

The general idea is then

       
θ θ θ θ
=
θi combined θi intersection θi intersection θi intersection
solid solid 1 solid 2 solid 3

Heat Transfer in Multidimensional Systems

Langston [16] has shown that it is possible to superimpose the heat-loss solutions for one-
dimensional bodies, as shown in Figures 4-14, 4-15, and 4-16, to obtain the heat for a
multidimensional body. The results of this analysis for intersection of two bodies is
         
Q Q Q Q
= + 1− [4-22]
Q0 total Q0 1 Q0 2 Q0 1

where the subscripts refer to the two intersecting bodies. For a multidimensional body
formed by intersection of three one-dimensional systems, the heat loss is given by
                  
Q Q Q Q Q Q Q
= + 1− + 1− 1−
Q0 total Q0 1 Q0 2 Q0 1 Q0 3 Q0 1 Q0 2
[4-23]

If the heat loss is desired after a given time, the calculation is straightforward. On the other
hand, if the time to achieve a certain heat loss is the desired quantity, a trial-and-error or
iterative procedure must be employed. The following examples illustrate the use of the
various charts for calculating temperatures and heat flows in multidimensional systems.

Semi-Infinite Cylinder Suddenly Exposed

to Convection EXAMPLE 4-8
A semi-infinite aluminum cylinder 5 cm in diameter is initially at a uniform temperature of 200◦ C.
It is suddenly subjected to a convection boundary condition at 70◦ C with h = 525 W/m2 · ◦ C.
Calculate the temperatures at the axis and surface of the cylinder 10 cm from the end 1 min after
exposure to the environment.
Solution
This problem requires a combination of solutions for the infinite cylinder and semi-infinite slab in
accordance with Figure 4-18e. For the slab we have
x = 10 cm α = 8.4 × 10−5 m2/s k = 215 W/m · ◦ C

so that the parameters for use with Figure 4-5 are

√
h ατ (525)[(8.4 × 10−5 )(60)]1/2
= = 0.173
k 215
x 0.1
√ = = 0.704
2 ατ (2)[(8.4 × 10−5 )(60)]1/2
From Figure 4-5  
θ
= 1 − 0.036 = 0.964 = S(X)
θi semi-infinite slab
For the infinite cylinder we seek both the axis- and surface-temperature ratios. The parameters
for use with Figure 4-8 are
k ατ θ0
r0 = 2.5 cm = 16.38 = 8.064 = 0.38
hr0 r2 θi
0

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.165 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

166 4-5 Multidimensional Systems

This is the axis-temperature ratio. To find the surface-temperature ratio, we enter Figure 4-l1, using
r θ
= 1.0 = 0.97
r0 θ0

Thus  
θ 0.38 at r = 0
C(
) = =
θi inf cyl (0.38)(0.97) = 0.369 at r = r0

Combining the solutions for the semi-infinite slab and infinite cylinder, we have
 
θ
= C(
)S(X)
θi semi−infinite cylinder
= (0.38)(0.964) = 0.366 at r = 0
= (0.369)(0.964) = 0.356 at r = r0

The corresponding temperatures are

T = 70 + (0.366)(200 − 70) = 117.6 at r = 0
T = 70 + (0.356)(200 − 70) = 116.3 at r = r0

Finite-Length Cylinder Suddenly Exposed

EXAMPLE 4-9 to Convection
A short aluminum cylinder 5.0 cm in diameter and 10.0 cm long is initially at a uniform temperature
of 200◦ C. It is suddenly subjected to a convection environment at 70◦ C, and h = 525 W/m2 · ◦ C.
Calculate the temperature at a radial position of 1.25 cm and a distance of 0.625 cm from one end
of the cylinder 1 min after exposure to the environment.
Solution
To solve this problem we combine the solutions from the Heisler charts for an infinite cylinder and
an infinite plate in accordance with the combination shown in Figure 4-18f. For the infinite-plate
problem
L = 5 cm
The x position is measured from the center of the plate so that
x 4.375
x = 5 − 0.625 = 4.375 cm = = 0.875
L 5
For aluminum
α = 8.4 × 10−5 m2/s k = 215 W/m · ◦ C
so
k 215 ατ (8.4 × 10−5 )(60)
= = 8.19 = = 2.016
hL (525)(0.05) L2 (0.05)2
From Figures 4-7 and 4-10, respectively,
θ0 θ
= 0.75 = 0.95
θi θ0
so that  
θ
= (0.75)(0.95) = 0.7125
θi plate
For the cylinder r0 = 2.5 cm
r 1.25 k 215
= = 0.5 = = 16.38
r0 2.5 hr0 (525)(0.025)

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.166 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 167

ατ (8.4 × 10−5 )(60)

= = 8.064
r02 (0.025)2

and from Figures 4-8 and 4-11, respectively,

θ0 θ
= 0.38 = 0.98
θi θ0

so that  
θ
= (0.38)(0.98) = 0.3724
θi cyl

Combining the solutions for the plate and cylinder gives

 
θ
= (0.7125)(0.3724) = 0.265
θi short cylinder

Thus
T = T∞ + (0.265)(Ti − T∞ ) = 70 + (0.265)(200 − 70) = 104.5◦ C

Heat Loss for Finite-Length Cylinder EXAMPLE 4-10

Calculate the heat loss for the short cylinder in Example 4-9.
Solution
We first calculate the dimensionless heat-loss ratio for the infinite plate and infinite cylinder that
make up the multidimensional body. For the plate we have L = 5 cm = 0.05 m. Using the properties
of aluminum from Example 4-9, we calculate
hL (525)(0.05)
= = 0.122
k 215

h2 ατ (525)2 (8.4 × 10−5 )(60)

= = 0.03
k2 (215)2

From Figure 4-14, for the plate, we read

 
Q
= 0.22
Q0 p

For the cylinder r0 = 2.5 cm = 0.025 m, so we calculate

hr0 (525)(0.025)
= = 0.061
k 215
and from Figure 4-15 we can read  
Q
= 0.55
Q0 c
The two heat ratios may be inserted in Equation (4-22) to give
 
Q
= 0.22 + (0.55)(1 − 0.22) = 0.649
Q0 tot

The specific heat of aluminum is 0.896 kJ/kg · ◦ C and the density is 2707 kg/m3 , so we calculate
Q0 as
Q0 = ρcVθi = (2707)(0.896)π(0.025)2 (0.1)(200 − 70)
= 61.9 kJ

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.167 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

168 4-6 Transient Numerical Method

The actual heat loss in the 1-min time is thus

Q = (61.9 kJ)(0.649) = 40.2 kJ

4-6 TRANSIENT NUMERICAL METHOD

The charts described in Sections 4-4 and 4-5 are very useful for calculating temperatures
in certain regular-shaped solids under transient heat-flow conditions. Unfortunately, many
geometric shapes of practical interest do not fall into these categories; in addition, one is
frequently faced with problems in which the boundary conditions vary with time. These
transient boundary conditions as well as the geometric shape of the body can be such that
a mathematical solution is not possible. In these cases, the problems are best handled by
a numerical technique with computers. It is the setup for such calculations that we now
describe. For ease in discussion we limit the analysis to two-dimensional systems. An
extension to three dimensions can then be made very easily.
Consider a two-dimensional body divided into increments as shown in Figure 4-19.
The subscript m denotes the x position, and the subscript n denotes the y position. Within
the solid body the differential equation that governs the heat flow is
 2 
∂ T ∂2 T ∂T
k + = ρc [4-24]
∂x2 ∂y2 ∂τ

assuming constant properties. We recall from Chapter 3 that the second partial derivatives
may be approximated by

∂2 T 1
≈ (Tm+1,n + Tm−1,n − 2Tm,n ) [4-25]
∂x 2 (x)2

∂2 T 1
≈ (Tm,n+1 + Tm,n−1 − 2Tm,n ) [4-26]
∂y 2 (y)2

Figure 4-19 Nomenclature for numerical

solution of two-dimensional
unsteady-state conduction
problem.

m, n + 1

Δy
m − 1, n m, n m + 1, n

Δy
m, n − 1

Δx Δx

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.168 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 169

The time derivative in Equation (4-24) is approximated by

p+1 p
∂T Tm,n − Tm,n
≈ [4-27]
∂τ τ
In this relation the superscripts designate the time increment. Combining the relations above
gives the difference equation equivalent to Equation (4-24)
p p p p p p p+1 p
Tm+1,n + Tm−1,n − 2Tm,n Tm,n+1 + Tm,n−1 − 2Tm,n 1 Tm,n − Tm,n
+ = [4-28]
(x)2 (y)2 α τ
Thus, if the temperatures of the various nodes are known at any particular time, the
temperatures after a time increment τ may be calculated by writing an equation like
p+1
Equation (4-28) for each node and obtaining the values of Tm,n . The procedure may be
repeated to obtain the distribution after any desired number of time increments. If the
increments of space coordinates are chosen such that
x = y
p+1
the resulting equation for Tm,n becomes
α τ
p   4α τ

p+1 p p p
Tm,n = T + Tm−1,n + Tm,n+1 + Tm,n−1 + 1 − Tp [4-29]
(x)2 m+1,n (x)2 m,n
If the time and distance increments are conveniently chosen so that
(x)2
=4 [4-30]
α τ
it is seen that the temperature of node (m, n) after a time increment is simply the arithmetic
average of the four surrounding nodal temperatures at the beginning of the time increment.
When a one-dimensional system is involved, the equation becomes
 
α τ  p p 2α τ
Tmp+1 = T + T + 1 − Tp [4-31]
(x)2 m+1 m−1
(x)2 m
and if the time and distance increments are chosen so that
(x)2
=2 [4-32]
α τ
the temperature of node m after the time increment is given as the arithmetic average of the
two adjacent nodal temperatures at the beginning of the time increment.
Some general remarks concerning the use of numerical methods for solution of transient
conduction problems are in order at this point. We have already noted that the selection of
the value of the parameter
(x)2
M=
α τ
governs the ease with which we may proceed to effect the numerical solution; the choice
of a value of 4 for a two-dimensional system or a value of 2 for a one-dimensional system
makes the calculation particularly easy.
Once the distance increments and the value of M are established, the time increment
is fixed, and we may not alter it without changing the value of either x or M, or both.
Clearly, the larger the values of x and τ, the more rapidly our solution will proceed.
On the other hand, the smaller the value of these increments in the independent variables,
the more accuracy will be obtained. At first glance one might assume that small distance
increments could be used for greater accuracy in combination with large time increments

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.169 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

170 4-6 Transient Numerical Method

to speed the solution. This is not the case, however, because the finite-difference equa-
tions limit the values of τ that may be used once x is chosen. Note that if M < 2 in
p
Equation (4-31), the coefficient of Tm becomes negative, and we generate a condition that
will violate the second law of thermodynamics. Suppose, for example, that the adjoining
p p
nodes are equal in temperature but less than Tm . After the time increment τ, Tm may not
be lower than these adjoining temperatures; otherwise heat would have to flow uphill on
the temperature scale, and this is impossible. A value of M < 2 would produce just such an
effect; so we must restrict the values of M to

(x)2 M ≥ 2 one-dimensional systems
=
α τ M ≥ 4 two-dimensional systems

This restriction automatically limits our choice of τ, once x is established.

It so happens that the above restrictions, which are imposed in a physical sense, may also
be derived on mathematical grounds. It may be shown that the finite-difference solutions will
not converge unless these conditions are fulfilled. The problems of stability and convergence
of numerical solutions are discussed in References 7, 13, and 15 in detail.
The difference equations given above are useful for determining the internal temper-
ature in a solid as a function of space and time. At the boundary of the solid, a convection
resistance to heat flow is usually involved, so that the above relations no longer apply. In
general, each convection boundary condition must be handled separately, depending on the
particular geometric shape under consideration. The case of the flat wall will be considered
as an example.
For the one-dimensional system shown in Figure 4-20 we may make an energy balance
at the convection boundary such that

∂T
−kA = hA(Tw − T∞ ) [4-33]
∂x wall
The finite-difference approximation would be given by
y
−k (Tm+1 − Tm ) = h y(Tm+1 − T∞ )
x
Figure 4-20 Nomenclature for
numerical solution of
unsteady-state conduction
problem with convection
boundary condition.

m–1 m m+1

Environment
T∞

Surface,Tw = Tm+1

Δx Δx

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.170 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch04 10/30/2008 17:34

CHAPTER 4 Unsteady-State Conduction 171

or
Tm + (h x/k)T∞
Tm+1 =
1 + h x/k
To apply this condition, we should calculate the surface temperature Tm+1 at each time
increment and then use this temperature in the nodal equations for the interior points of the
solid. This is only an approximation because we have neglected the heat capacity of the
element of the wall at the boundary. This approximation will work fairly well when a large
number of increments in x are used because the portion of the heat capacity that is neglected
is then small in comparison with the total. We may take the heat capacity into account in a
general way by considering the two-dimensional wall of Figure 3-7 exposed to a convection
boundary condition, which we duplicate here for convenience as Figure 4-21. We make a
transient energy balance on the node (m, n) by setting the sum of the energy conducted and
convected into the node equal to the increase in the internal energy of the node. Thus
p p p p p p
Tm−1,n − Tm,n x Tm,n+1 − Tm,n x Tm,n−1 − Tm,n
k y +k +k
x 2 y 2 y
p+1 p
p x Tm,n − Tm,n
+h y(T∞ − Tm,n ) = ρc y
2 τ
p+1
If x = y, the relation for Tm,n becomes

p+1 α τ h x p p p
Tm,n = 2 T∞ + 2Tm−1,n + Tm,n+1 + Tm,n−1
(x)2 k
  
(x)2 h x p
+ −2 − 4 Tm,n [4-34]
α τ k

The corresponding one-dimensional relation is

   
α τ h x p (x)2 h x
Tmp+1 = 2 T ∞ + 2Tm−1 + − 2 − 2 Tmp [4-35]
(x)2 k α τ k

Notice now that the selection of the parameter (x)2/α τ is not as simple as it is for the
interior nodal points because the heat-transfer coefficient influences the choice. It is still

Figure 4-21 Nomenclature for nodal

equation with convective
boundary condition.

m, n + 1
T∞
m − 1, n m, n
Δy
Δy
q

m, n − 1
Δx
2 Surface
Δx

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.171 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch04 10/30/2008 17:34

172 4-6 Transient Numerical Method

p p
possible to choose the value of this parameter so that the coefficient of Tm or Tm,n will be
zero. These values would then be
⎧

⎪ h x
⎪ + 1 for the one-dimensional case
(x)2 ⎨
2
=
k 
ατ ⎪ ⎪ h x
⎩2 + 2 for the two-dimensional case
k

To ensure convergence of the numerical solution, all selections of the parameter (x)2/α τ
must be restricted according to
⎧

⎪ h x
⎪ + 1 for the one-dimensional case
(x)2 ⎨
2
≥
k 
ατ ⎪ ⎪ h x
⎩2 + 2 for the two-dimensional case
k

Forward and Backward Differences

The equations above have been developed on the basis of a forward-difference technique
in that the temperature of a node at a future time increment is expressed in terms of the
surrounding nodal temperatures at the beginning of the time increment. The expressions
p+1
are called explicit formulations because it is possible to write the nodal temperatures Tm,n
p
explicitly in terms of the previous nodal temperatures Tm,n . In this formulation, the calcula-
tion proceeds directly from one time increment to the next until the temperature distribution
is calculated at the desired final state.
The difference equation may also be formulated by computing the space derivatives
in terms of the temperatures at the p + 1 time increment. Such an arrangement is called
a backward-difference formulation because the time derivative moves backward from the
times for heat conduction into the node. The equation equivalent to Equation (4-28) would
then be
p+1 p+1 p+1 p+1 p+1 p+1 p+1 p
Tm+1,n + Tm−1,n − 2Tm,n Tm,n+1 + Tm,n−1 − 2Tm,n 1 Tm,n − Tm,n
+ = [4-36]
(x)2 (y)2 α τ
The equivalence to Equation (4-29) is
−α τ  p+1  4α τ

p p+1 p+1 p+1
Tm,n = T + T + T + T + 1 + T p+1 [4-37]
(x)2 m+1,n m−1,n m,n+1 m,n−1
(x)2 m,n
We may now note that this backward-difference formulation does not permit the explicit
calculation of the T p+1 in terms of T p . Rather, a whole set of equations must be written
for the entire nodal system and solved simultaneously to determine the temperatures T p+1 .
Thus we say that the backward-difference method produces an implicit formulation for the
future temperatures in the transient analysis. The solution to the set of equations can be
performed with the methods discussed in Chapter 3.
The Biot and Fourier numbers may also be defined in the following way for problems
in the numerical format:
h x
Bi = [4-38]
k
α τ
Fo = [4-39]
(x)2
By using this notation, Tables 4-2 and 4-3 have been constructed to summarize some typical
nodal equations in both the explicit and implicit formulations. For the cases of x = y

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.172 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch04 10/30/2008 17:34

CHAPTER 4 Unsteady-State Conduction 173

displayed in Table 4-2, the most restrictive stability requirement (smallest τ) is exhibited
by an exterior corner node, assuming all the convection nodes have the same value of Bi.

Table 4-2 Explicit nodal equations. (Dashed lines indicate element volume.)†
Stability
Physical situation Nodal equation for x = y requirement

p+1 p p p p
(a) Interior node Tm,n = Fo Tm−1,n + Tm,n+1 + Tm+1,n + Tm,n−1 Fo ≤ 14
p
+[1− 4(Fo)]Tm,n
m, n + 1 p+1 p p p p
Tm,n = Fo Tm−1,n + Tm,n+1 + Tm+1,n + Tm,n−1

p p
Δy −4Tm,n + Tm,n
m − 1, n m, n m + 1, n

Δy
m, n − 1

Δx Δx

p+1 p p p p
(b) Convection boundary node Tm,n = Fo [2Tm−1,n + Tm,n+1 + Tm,n−1 + 2(Bi)T∞ ] Fo(2 + Bi) ≤ 12
p
+[1 − 4(Fo) − 2(Fo)(Bi)]Tm,n
m, n + 1 p+1 p p p p
Tm,n = Fo [2Bi (T∞ − Tm,n ) + 2Tm−1,n + Tm,n+1
Δy p p p
+Tm,n−1 − 4Tm,n ] + Tm,n
m − 1, n h, T∞
m, n

Δy
m, n − 1

Δx
p+1 p p p
(c) Exterior corner with convection Tm,n = 2(Fo) [Tm−1,n + Tm,n−1 + 2(Bi)T∞ ] Fo(1 + Bi) ≤ 14
boundary p
+[1 − 4(Fo) − 4(Fo)(Bi)]Tm,n
h, T∞ p+1 p p p
m − 1, n Tm,n = 2Fo [Tm−1,n + Tm,n−1 − 2Tm,n
m, n p p p
+2Bi(T∞ − Tm,n )] + Tm,n
Δy
m, n − 1

Δx

p+1 p p
(d) Interior corner with convection Tm,n = 23 (Fo) [2Tm,n+1 + 2Tm+1,n Fo(3 + Bi) ≤ 34
boundary p p p
+2Tm−1,n + Tm,n−1 + 2(Bi)T∞ ]
4 p
m, n + 1 +[1 − 4(Fo) − 3 (Fo)(Bi)]Tm,n
p+1 p p
Tm,n = (4/3)Fo [Tm,n+1 + Tm+1,n
p p p p p
+Tm−1,n − 3Tm,n + Bi (T∞ − Tm,n )] + Tm,n
m − 1, n m, n m + 1, n

Δy h, T∞
m, n − 1

Δx

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.173 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch04 11/6/2008 10:1

174 4-6 Transient Numerical Method

Table 4-2 (Continued).

Stability
Physical situation Nodal equation for
x =
y requirement
p+1 p p p
(e) Insulated boundary Tm,n = Fo [2Tm−1,n + Tm,n+1 + Tm,n−1 ] Fo ≤ 14
p
+[1 − 4(Fo)]Tm,n
m, n + 1

m − 1, n m, n Insulated

∆y
m, n − 1

∆x

† Convection surfaces may be made insulated by setting h = 0 (Bi = 0).

Table 4-3 Implicit nodal equations. (Dashed lines indicate volume element.)
Physical situation Nodal equation for
x =
y


p+1 p+1 p+1 p+1
(a) Interior node [1 + 4(Fo)]Tm,n − Fo Tm−1,n + Tm,n+1 + Tm+1,n

p+1 p
+ Tm,n−1 − Tm,n = 0
m + 1, n

∆y
m – 1, n m, n m + 1, n

∆y
m – 1, n

∆x ∆x


p+1 p+1 p+1
(b) Convection boundary node [1 + 2(Fo)(2 + Bi)]Tm,n − Fo Tm,n+1 + Tm,n−1

p+1 p+1 p
+2Tm−1,n + 2(Bi)T∞ − Tm,n = 0
m, n + 1
∆y
m – 1, n h, T∞
m, n
∆y
m, n − 1

∆x

p+1 p+1 p+1
(c) Exterior corner with convection boundary [1 + 4(Fo)(1 + Bi)]Tm,n − 2(Fo) Tm−1,n + Tm,n−1
h, T∞ 
m – 1, n p+1 p
+2(Bi)T∞ − Tm,n = 0
m ,n
∆y
m, n − 1

∆x

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.174 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch04 10/30/2008 17:34

CHAPTER 4 Unsteady-State Conduction 175

Table 4-3 (Continued).

Physical situation Nodal equation for x = y


Bi p+1 2(Fo)  p+1 p+1
(d) Interior corner with convection 1 + 4(Fo) 1 + Tm,n − × 2Tm−1,n + Tm,n−1
3 3 
boundary p+1 p+1 p+1 p
+ 2Tm,n+1 + 2Tm+1,n + 2(Bi)T∞ − Tm,n = 0

m, n + 1

m – 1, n m, n m + 1, n

Δy h, T∞
m, n − 1

Δx

p+1 p+1 p+1 p+1 p
(e) Insulated boundary [1 + 4(Fo)]Tm,n − Fo 2Tm−1,n + Tm,n+1 + Tm,n−1 − Tm,n = 0

m, n + 1
Insulated

m – 1, n m, n

Δy
m, n − 1

Δx

The advantage of an explicit forward-difference procedure is the direct calculation

of future nodal temperatures; however, the stability of this calculation is governed by the
selection of the values of x and τ. A selection of a small value of x automatically
forces the selection of some maximum value of τ. On the other hand, no such restriction
is imposed on the solution of the equations that are obtained from the implicit formulation.
This means that larger time increments can be selected to speed the calculation. The obvious
disadvantage of the implicit method is the larger number of calculations for each time step.
For problems involving a large number of nodes, however, the implicit method may result in
less total computer time expended for the final solution because very small time increments
may be imposed in the explicit method from stability requirements. Much larger increments
in τ can be employed with the implicit method to speed the solution.
Most problems will involve only a modest number of nodes and the explicit formulation
will be quite satisfactory for a solution, particularly when considered from the standpoint
of the more generalized formulation presented in the following section.
For a discussion of many applications of numerical analysis to transient heat conduction
problems, the reader is referred to References 4, 8, 13, 14, and 15.
It should be obvious to the reader by now that finite-difference techniques may be
applied to almost any situation with just a little patience and care. Very complicated problems
then become quite easy to solve with only modest computer facilities. The use of Microsoft
Excel for solution of transient heat-transfer problems is discussed in Appendix D.
Finite-element methods for use in conduction heat-transfer problems are discussed in
References 9 to 13. A number of software packages are available commercially.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.175 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

176 4-7 Thermal Resistance and Capacity Formulation

4-7 THERMAL RESISTANCE AND CAPACITY

FORMULATION
As in Chapter 3, we can view each volume element as a node that is connected by thermal
resistances to its adjoining neighbors. For steady-state conditions the net energy transfer
into the node is zero, while for the unsteady-state problems of interest in this chapter the
net energy transfer into the node must be evidenced as an increase in internal energy of the
element. Each volume element behaves like a small “lumped capacity,” and the interaction
of all the elements determines the behavior of the solid during a transient process. If the
internal energy of a node i can be expressed in terms of specific heat and temperature, then
its rate of change with time is approximated by
p+1 p
E T − Ti
= ρcV i
τ τ
where V is the volume element. If we define the thermal capacity as
Ci = ρi ci Vi [4-40]
then the general resistance-capacity formulation for the energy balance on a node is
 Tjp − Tip p+1
Ti − Ti
j
qi + = Ci [4-41]
Rij τ
j

where all the terms on the left are the same as in Equation (3-31). The resistance and volume
elements for a variety of geometries and boundary conditions were given in Tables 3-3 and
3-4. Physical systems where the internal energy E involves phase changes can also be
accommodated in the above formulation but are beyond the scope of our discussion.
The central point is that use of the concepts of thermal resistance and capacitance
enables us to write the forward-difference equation for all nodes and boundary conditions
in the single compact form of Equation (4-41). The setup for a numerical solution then
becomes a much more organized process that can be adapted quickly to the computational
methods at hand.
Equation (4-41) is developed by using the forward-difference concept to produce an
p+1
explicit relation for each Ti . As in our previous discussion, we could also write the energy
balance using backward differences, with the heat transfers into each ith node calculated in
terms of the temperatures at the p + 1 time increment. Thus,
 Tjp+1 − Tip+1 p+1
Ti − Ti
p
qi + = Ci [4-42]
Rij τ
j

Now, as before, the set of equations produces an implicit set that must be solved
p+1
simultaneously for the Ti , etc. The solution can be carried out by a number of methods
as discussed in Chapter 3. If the solution is to be performed with a Gauss-Seidel iteration
p+1
technique, then Equation (4-42) should be solved for Ti and expressed as
 p+1 p
qi + (T /Rij ) + (Ci /τ)Ti
p+1 i j
Ti =  [4-43]
(1/Rij ) + Ci /τ
j

It is interesting to note that in the steady-state limit of τ → ∞ this equation becomes

identical with Equation (3-32), the formulation we employed for the iterative solution in
Chapter 3.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.176 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 177

The stability requirement in the explicit formulation may be examined by solving

p+1
Equation (4-41) for Ti :
  Tjp  τ  
p+1 τ  1 p
Ti = qi + + 1− T [4-44]
Rij Ci Ci Rij i
j j

The value of qi can influence the stability, but we can choose a safe limit by observing the
behavior of the equation for qi = 0. Using the same type of thermodynamic argument as
p
with Equation (4-31), we find that the coefficient of Ti cannot be negative. Our stability
requirement is therefore
τ  1
1− ≥0 [4-45]
Ci Rij
j

Suppose we have a complicated numerical problem to solve with a variety of boundary

conditions, perhaps nonuniform values of the space increments, etc. Once we have all the
nodal resistances and capacities formulated, we then have the task of choosing the time
increment τ to use for the calculation. To ensure stability we must keep τ equal to or
less than a value obtained from the most restrictive nodal relation like Equation (4-45).
Solving for τ gives
 Ci 
τ ≤  for stability [4-46]
(1/Rij )
j
min

While Equation (4-44) is very useful in establishing the maximum allowable time
increment, it may involve problems of round-off errors in computer solutions when small
p+1
thermal resistances are employed. The difficulty may be alleviated by expressing Ti in
the following form for calculation purposes:
⎡ ⎤
p+1 τ ⎣  Tjp − Tip
Ti = qi + ⎦+Tp [4-47]
i
Ci Rij
j

In Table 4-2 the nodal equations for x = y are listed in the formats of both
equations (4-44) and (4-47). The equations listed in Table 4-2 in the form of Equation
(4-47) do not include the heat-source term. If needed, the term may be added using
qi = q̇i Vi

where q̇i is the heat generation per unit volume and Vi is the volume element shown by
dashed lines in the table. For radiation input to the node,
qi = qi, rad × Ai

where qi,rad is the net radiant energy input to the node per unit area and Ai is the area of
the node for radiant exchange, which may or may not be equal to the area for convection
heat transfer.
We should remark that the resistance-capacity formulation is easily adapted to take into
account thermal-property variations with temperature. One need only calculate the proper
values of ρ, c, and k for inclusion in the Ci and Rij . Depending on the nature of the problem
and accuracy required, it may be necessary to calculate new values of Ci and Rij for each
time increment. Example 4-17 illustrates the effects of variable conductivity.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.177 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

178 4-7 Thermal Resistance and Capacity Formulation

Steady State as a Limiting Case of Transient Solution

As we have seen, the steady-state numerical formulation results when the right side of
Equation (4-41) is set equal to zero. It also results when the calculation of the unsteady case
using either Equation (4-44) or (4-47) is carried out for a large number of time increments.
While the latter method of obtaining a steady-state solution may appear rather cumbersome,
it can proceed quite rapidly with a computer. We may recall that the Gauss-Seidel iteration
method was employed for the solution of many steady-state numerical problems, which
of course entailed many computer calculations. If variable thermal resistances resulting
from either variable thermal conductivities or variations in convection boundary condi-
tions are encountered, the steady-state limit of a transient solution may offer advantages
over the direct steady-state solution counterpart. We will recall that when variable thermal
resistances appear, the resulting steady-state nodal equations become nonlinear and their
solution may be tedious. The transient solution for such cases merely requires that each
resistance be recalculated at the end of each time increment τ, or the resistances may be
entered directly as variables in the nodal equations. The calculations are then carried out
p+1
for a sufficiently large number of time increments until the values of the Ti no longer
change by a significant amount. At this point, the steady-state solution is obtained as the
resulting values of the Ti .
The formulation and solution of transient numerical problems using Microsoft Excel
is described in Section D-5 of the Appendix, along with worked examples. An example
is also given of a transient solution carried forward a sufficient length of time to achieve
steady-state conditions.
EXAMPLE 4-11 Sudden Cooling of a Rod
A steel rod [k = 50 W/m · ◦ C] 3 mm in diameter and 10 cm long is initially at a uniform temperature
of 200◦ C. At time zero it is suddenly immersed in a fluid having h = 50 W/m2 · ◦ C and T∞ = 40◦ C
while one end is maintained at 200◦ C. Determine the temperature distribution in the rod after 100 s.
The properties of steel are ρ = 7800 kg/m3 and c = 0.47 kJ/kg · ◦ C.

Figure Example 4-11

Δx
2
T0 = 200˚C
1 2 3 4
T∞ = 40˚C
Δx Δx Δx Δx
Δ x = 2.5 cm

Solution
The selection of increments on the rod is as shown in the Figure Example 4-11. The cross-sectional
area of the rod is A = π(1.5)2 = 7.069 mm2 . The volume element for nodes 1, 2, and 3 is
V = Ax = (7.069)(25) = 176.725 mm3

Node 4 has a V of half this value, or 88.36 mm3 . We can now tabulate the various resistances
and capacities for use in an explicit formulation. For nodes 1, 2, and 3 we have
x 0.025
Rm+ = Rm− = = = 70.731◦ C/W
kA (50)(7.069 × 10−6 )

and
1 1
R∞ = = = 84.883◦ C/W
h(πd x) (50)π(3 × 10−3 )(0.025)

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.178 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 179

C = ρc V = (7800)(470)(1.7673 × 10−7 ) = 0.6479 J/◦ C

For node 4 we have

1 x
Rm+ = = 2829◦ C/W Rm− = = 70.731◦ C/W
hA kA
ρc V 2
C= = 0.3240 J/◦ C R∞ = = 169.77◦ C/W
2 hπdx
To determine the stability requirement we form the following table:

 Ci
Node (1/Rij ) Ci  ,s
(1/Rij )
1 0.04006 0.6479 16.173
2 0.04006 0.6479 16.173
3 0.04006 0.6479 16.173
4 0.02038 0.3240 15.897

Thus node 4 is the most restrictive, and we must select τ < 15.9 s. Since we wish to find the
temperature distribution at 100 s, let us use τ = 10 s and make the calculation for 10 time
increments using Equation (4-47) for the computation. We note, of course, that qi = 0 because
there is no heat generation. The calculations are shown in the following table.

Node temperature
Time
increment T1 T2 T3 T4
0 200 200 200 200
1 170.87 170.87 170.87 169.19
2 153.40 147.04 146.68 145.05
3 141.54 128.86 126.98 125.54
4 133.04 115.04 111.24 109.70
5 126.79 104.48 98.76 96.96
6 122.10 96.36 88.92 86.78
7 118.53 90.09 81.17 78.71
8 115.80 85.23 75.08 72.34
9 113.70 81.45 70.31 67.31
10 112.08 78.51 66.57 63.37

We can calculate the heat-transfer rate at the end of 100 s by summing the convection heat
losses on the surface of the rod. Thus
 Ti − T∞
q=
Ri∞
i

and
 
200 − 40 112.08 + 78.51 + 66.57 − (3)(40) 1 1
q= + + + (63.37 − 40)
(2)(84.883) 84.883 169.77 2829
= 2.704 W

Implicit Formulation EXAMPLE 4-12

We can illustrate the calculation scheme for the implicit formulation by reworking Example 4-11
using only two time increments, that is, τ = 50 s.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.179 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

180 4-7 Thermal Resistance and Capacity Formulation

For this problem we employ the formulation indicated by Equation (4-43), with τ = 50 s.
The following quantities are needed.
Ci  1 Ci
+
Node τ i Rij τ
1 0.01296 0.05302
2 0.01296 0.05302
3 0.01296 0.05302
4 0.00648 0.02686

We have already determined the Rij in Example 4-11 and thus can insert them into Equation (4-43)
p
to write the nodal equations for the end of the first time increment, taking all Ti = 200◦ C. We use
the prime to designate temperatures at the end of the time increment. For node 1,
200 T2 40
0.05302T1 = + + + (0.01296)(200)
70.731 70.731 84.833
For node 2,
T1 T3 40
0.05302T2 = + + + (0.01296)(200)
70.731 70.731 84.833
For nodes 3 and 4,
T2 40T4
0.05302T3 = + + + (0.01296)(200)
70.73170.731 84.833
T3 40 40
0.02686T4 = + + + (0.00648)(200)
70.731 2829 169.77
These equations can then be reduced to
0.05302T1 − 0.01414T2 = 5.8911
−0.01414T1 + 0.05302T2 − 0.01414T3 = 3.0635
−0.01414T2 + 0.05302T3 − 0.01414T4 = 3.0635
−0.01414T3 + 0.02686T4 = 1.5457

which have the solution

T1 = 145.81◦ C T2 = 130.12◦ C
T3 = 125.43◦ C T4 = 123.56◦ C

We can now apply the backward-difference formulation a second time using the double prime to
designate the temperatures at the end of the second time increment:
200 T2 40
0.05302T1 = + + + (0.01296)(145.81)
70.731 70.731 84.833
T1 T3 40
0.05302T2 = + + + (0.01296)(130.12)
70.731 70.731 84.833
T2 T4 40
0.05302T3 = + + + (0.01296)(125.43)
70.731 70.731 84.833
T3 40 40
0.02686T4 = + + + (0.00648)(123.56)
70.731 2829 169.77
and this equation set has the solution
T1 = 123.81◦ C T2 = 97.27◦ C
T3 = 88.32◦ C T4 = 85.59◦ C

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.180 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 181

We find this calculation in substantial disagreement with the results of Example 4-11. With a
larger number of time increments, better agreement would be achieved. In a problem involving a
large number of nodes, the implicit formulation might involve less computer time than the explicit
method, and the purpose of this example has been to show how the calculation is performed.

Cooling of a Ceramic EXAMPLE 4-13

A 1 by 2 cm ceramic strip [k = 3.0 W/m · ◦ C] is embedded in a high-thermal-conductivity material,
as shown in Figure Example 4-13, so that the sides are maintained at a constant temperature
of 300◦ C. The bottom surface of the ceramic is insulated, and the top surface is exposed to a
convection environment with h = 200 W/m2 · ◦ C and T∞ = 50◦ C. At time zero the ceramic is
uniform in temperature at 300◦ C. Calculate the temperatures at nodes 1 to 9 after a time of
12 s. For the ceramic ρ = 1600 kg/m3 and c = 0.8 kJ/kg · ◦ C. Also calculate the total heat loss in
this time.

Figure Example 4-13

2 cm
h, T∞ = 50˚C
1 2 3
4 5 6
1 cm T = 300˚C

7 8 9

300˚C
Insulated

Solution
We treat this as a two-dimensional problem with x = y = 0.5 cm. From symmetry T1 = T3 ,
T4 = T6 , and T7 = T9 , so we have six unknown nodal temperatures. We now tabulate the various
nodal resistances and capacities. For nodes 4 and 5
x 0.005
Rm+ = Rm− = Rn+ = Rn− = = = 0.3333
kA (3.0)(0.005)

For nodes 1 and 2

x (0.005)(2)
Rm+ = Rm− = = = 0.6667◦ C/W Rn− = 0.3333◦ C/W
kA (3.0)(0.005)
1 1
Rn+ = = = 1.0◦ C/W
h x (200)(0.005)

For nodes 7 and 8

Rm+ = Rm− = 0.6667◦ C/W Rn+ = 0.3333◦ C/W Rn− = ∞

For nodes 1, 2, 7, and 8 the capacities are

ρc(x)2 (1600)(800)(0.005)2
C= = = 16 J/◦ C
2 2
For nodes 4 and 5
C = ρc(x)2 = 32 J/◦ C

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.181 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

182 4-7 Thermal Resistance and Capacity Formulation

The stability requirement for an explicit solution is now determined by tabulating the following
quantities:

 1 Ci
 ,s
Node Rij Ci (1/Rij )
1 7 16 2.286
2 7 16 2.286
4 12 32 2.667
5 12 32 2.667
7 6 16 2.667
8 6 16 2.667

Thus the two convection nodes control the stability requirement, and we must choose τ ≤ 2.286 s.
Let us choose τ = 2.0 s and make the calculations for six time increments with Equation (4-47).
We note once again the symmetry considerations when calculating the temperatures of nodes 2,
5, and 8, that is, T1 = T3 , etc. The calculations are shown in the following table.

Node temperature
Time increment T1 T2 T4 T5 T7 T8
0 300 300 300 300 300 300
1 268.75 268.75 300 300 300 300
2 258.98 253.13 294.14 294.14 300 300
3 252.64 245.31 289.75 287.55 297.80 297.80
4 284.73 239.48 285.81 282.38 295.19 293.96
5 246.67 235.35 282.63 277.79 292.34 290.08
6 243.32 231.97 279.87 273.95 289.71 286.32

The total heat loss during the 12-s time interval is calculated by summing the heat loss of each
node relative to the initial temperature of 300◦ C. Thus

q= Ci (300 − Ti )

where q is the heat loss. For this summation, since the constant-temperature boundary nodes
experience no change in temperature, they can be left out. Recalling that T1 = T3 , T4 = T6 , and
T7 = T9 , we have

Ci (300 − Ti ) = nodes (1, 2, 3, 7, 8, 9) + nodes (4, 5, 6)
= 16[(6)(300) − (2)(243.2) − 231.97 − (2)(289.71)
− 286.32] + 32[(3)(300) − (2)(279.87) − 273.95]
= 5572.3 J/m length of strip

The average rate of heat loss for the 12-s time interval is
q 5572.3
= = 464.4 W [1585 Btu/h]
τ 12

EXAMPLE 4-14 Cooling of a Steel Rod, Nonuniform h

A nickel-steel rod having a diameter of 2.0 cm is 10 cm long and initially at a uniform temperature
of 200◦ C. It is suddenly exposed to atmospheric air at 30◦ C while one end of the rod is maintained
at 200◦ C. The convection heat-transfer coefficient can be computed from
h = 9.0 T 0.175 W/m2 · ◦ C

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.182 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 183

where T is the temperature difference between the rod and air surroundings. The properties of
nickel steel may be taken as k = 12 W/m · ◦ C, c = 0.48 kJ/kg · ◦ C, and ρ = 7800 kg/m3 . Using the
numerical method, (a) determine the temperature distribution in the rod after 250, 500, 750, 1000,
1250 s, and for steady state; (b) determine the steady-state temperature distribution for a constant
h = 22.11 W/m2 · ◦ C and compare with an analytical solution.

Figure Example 4-14

200 τ=0

175

τ = 250 s
150
T,˚C

125 τ = 500 s

τ = 750 s
100 τ = 1000 s
τ = 1250 s
Steady state
75 x
T1 T2 T3 T4 T5
Δ x = 2 cm

1 2 3 4 5

Solution
Five nodes are chosen as shown in Figure Example 4-14 with x = 2.0 cm. The capacitances are
then
(7800)(480)π(0.02)2 (0.02)
C1 = C2 = C3 = C4 = = 23.524 J/◦ C
4
C5 = 12 C1 = 11.762 J/◦ C

The resistances for nodes 1, 2, 3, and 4 are

1 1 kA (12)π(0.02)2
= = = = 0.188496
Rm+ Rm− x (4)(0.02)
1
= hP x = (9.0)π(0.02)(0.02)(T − 30)0.175 = (1.131 × 10−2 )(T − 30)0.175
R∞
For node 5
1
= 0.188496
Rm−
1 π(0.02)2
= hA = 9.0 (T − 30)0.175 = (2.827 × 10−3 )(T − 30)0.175
Rm+ 4

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.183 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

184 4-7 Thermal Resistance and Capacity Formulation

1 1
= = (5.655 × 10−3 )(T − 30)0.175
R5∞ 2R1∞

where T∞ = 30◦ C for all nodes. We can compute the following table for worst-case conditions
of T = 200◦ C throughout the rod. The stability requirement so established will then work for all
other temperatures.

Ci
   ,s
Node (1/Rij )min (1/Rij )

1 0.4048 58.11
2 0.4048 58.11
3 0.4048 58.11
4 0.4048 58.11
5 0.2093 56.197

Thus, time steps below 56 s will ensure stability. The computational procedure is compli-
cated by the fact that the convection-resistance elements must be recalculated for each time step.
Selecting τ = 50 s, we have:

Node τ/Ci
1 2.1255
2 2.1255
3 2.1255
4 2.1255
5 4.251

We then use the explicit formulation of Equation (4-47) with no heat generation. The computational
algorithm is thus:
1. Compute R∞ values for the initial condition.
2. Compute temperatures at next time increment using Equation (4-47).
3. Recalculate R∞ values based on new temperatures.
4. Repeat temperature calculations and continue until the temperature distributions are obtained
at the desired times.
Results of these calculations are shown in the accompanying figure.
To determine the steady-state distribution we could carry the unsteady method forward a
large number of time increments or use the steady-state method and an iterative approach. The
iterative approach is required because the equations are nonlinear as a result of the variations in
the convection coefficient.
We still use a resistance formulation, which is now given as Equation (3-31):
 Tj − Ti
=0
Rij

The computational procedure is:

1. Calculate R∞ values for all nodes assuming all Ti = 200◦ C.
2. Formulate nodal equations for the Ti ’s.
3. Solve the equations by an appropriate method.
4. Recalculate R∞ values based on Ti values obtained in step 3.
5. Repeat the procedure until there are only small changes in Ti ’s.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.184 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 185

The results of this iteration are shown in the following table:

Iteration T1 , ◦ C T2 , ◦ C T3 , ◦ C T4 , ◦ C T5 , ◦ C
1 148.462 114.381 92.726 80.310 75.302
2 151.381 119.557 99.409 87.853 83.188
3 151.105 119.038 98.702 87.024 82.306
4 151.132 119.090 98.774 87.109 82.396

This steady-state temperature distribution is also plotted with the transient profiles.
The value of h for Ti = 200◦ C is 22.11 W/m2 · ◦ C, so the results of the first iteration correspond
to a solution for a constant h of this value. The exact analytical solution is given in Equation (2-34)
as
θ T − T∞ cosh m(L − x) + [h/km] sinh m(L − x)
= =
θ0 T0 − T∞ cosh mL + [h/km] sinh mL
The required quantities are
   
hP 1/2 (22.11)π(0.02) 1/2
m= = = 19.1964
kA (12)π(0.01)2

mL = (19.1964)(0.1) = 1.91964
22.22
h/km = = 0.09598
(12)(19.1964)

The temperatures at the nodal points can then be calculated and compared with the numerical
results in the following table. As can be seen, the agreement is excellent.

Node x, m (θ/θ0 )num (θ/θ0 )anal Percent deviation

1 0.02 0.6968 0.6949 0.27
2 0.04 0.4964 0.4935 0.59
3 0.06 0.3690 0.3657 0.9
4 0.08 0.2959 0.2925 1.16
5 0.1 0.2665 0.2630 1.33

We may also check the heat loss with that predicted by the analytical relation in Equation (2-34).
When numerical values are inserted we obtain
qanal = 11.874 W

The heat loss for the numerical model is computed by summing the convection loss from the six
nodes (including base node at 200◦ C). Using the temperatures for the first iteration corresponding
to h = 22.11 W/m2 · ◦ C,


q = (22.11)π(0.02)(0.02) (200 − 30) 12 + (148.462 − 30)
+ (114.381 − 30) + (92.726 − 30) + (80.31 − 30)

1 
+ (75.302 − 30) + (22.11)π(0.01)2(75.302 − 30)
2
= 12.082 W

We may make a further check by calculating the energy conducted in the base. This must be the
energy conducted to node 1 plus the convection lost by the base node or
(200 − 148.462)
q = (12)π(0.01)2 + (22.11)π(0.02)(0.01)(200 − 30)
0.02
= 12.076 W

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.185 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

186 4-7 Thermal Resistance and Capacity Formulation

This agrees very well with the convection calculation and both are within 1.8 percent of the
analytical value.
The results of this example illustrate the power of the numerical method in solving problems
that could not be solved in any other way. Furthermore, only a modest number of nodes, and
thus modest computation facilities, may be required to obtain a sufficiently accurate solution. For
example, the accuracy with which h will be known is typically ±10 to 15 percent. This would
overshadow any inaccuracies introduced by using relatively large nodes, as was done here.

EXAMPLE 4-15 Radiation Heating and Cooling

The ceramic wall shown in Figure Example 4-15a is initially uniform in temperature at 20◦ C and
has a thickness of 3.0 cm. It is suddenly exposed to a radiation source on the right side at 1000◦ C.
The left side is exposed to room air at 20◦ C with a radiation surrounding temperature of 20◦ C.
Properties of the ceramic are k = 3.0 W/m · ◦ C, ρ = 1600 kg/m3 , and c = 0.8 kJ/kg · ◦ C. Radiation
heat transfer with the surroundings at Tr may be calculated from
qr = σ A(T 4 − Tr 4 ) W [a]
where σ = 5.669 × 10−8 , = 0.8, and T is in degrees Kelvin. The convection heat-transfer coef-
ficient from the left side of the plate is given by
h = 1.92T 1/4 W/m2 · ◦ C [b]
Convection on the right side is negligible. Determine the temperature distribution in the plate
after 15, 30, 45, 60, 90, 120, and 150 s. Also determine the steady-state temperature distribution.
Calculate the total heat gained by the plate for these times.
Solution
We divide the wall into five nodes as shown and must express temperatures in degrees Kelvin
because of the radiation boundary condition. For node 1 the transient energy equation is


5/4 p+1 p
p4 p k
p p
 x T1 − T1
σ 2934 − T1 − 1.92 T1 − 293 + T2 − T1 = ρc [c]
x 2 τ
Similarly, for node 5

 p+1 p
p4 k
p p
 x T5 − T5
σ 12734 − T5 + T4 − T5 = ρc [d]
x 2 τ
Equations (c) and (d ) may be subsequently written
 
p+1 τ p2 p p k p
T1 = σ (2932 + T1 )(293 + T1 )(293) − 1.92(T1 − 293)1/4 (293) + T2
C1 x
 
τ p2 p p k p
+ 1− σ (2932 + T1 )(293 + T1 ) − 1.92(T1 − 293)1/4 + T1 [e]
C1 x
 
p+1 τ p2 p k p
T5 = σ (12732 + T5 )(1273 + T5 )(1273) + T4
C5 x
 
τ p2 p k p
+ 1− σ (12732 + T5 )(1273 + T5 ) + T5 [f]
C5 x
where C1 = C5 = ρcx/2. For the other three nodes the expressions are much simpler:
 
p+1 τ k p p 2k τ p
T2 = (T1 + T3 ) + 1 − T2 [g]
C2 x C2 x

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.186 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

Figure Example 4-15 (a) Nodal system, (b) transient response, (c) heat added.

Steady
state

1100

1000
τ = 150 s
τ = 120 s
900
τ = 90 s

800
Temperature, ˚K τ = 60 s

τ = 45 s
700

τ = 30 s
600

τ = 15 s
Radiation source at 1000˚C 500
1 2 3 4 5

qrad 400

Δ x = 0.75 cm 293 300 τ =0

3 cm 200˚K
Room at 20˚C T1 T2 T3 T4 T5
(a) (b)

15,000

12,000

9000
Q(τ ), kJ

6000

3000

6 18 30 60 90 120 150
τ, s
(c)

187

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.187 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

188 4-7 Thermal Resistance and Capacity Formulation

 
p+1 τ k p p 2k τ p
T3 = (T2 + T4 ) + 1 − T3 [h]
C3 x C3 x
 
p+1 τ k p p 2k τ p
T4 = (T3 + T5 ) + 1 − T4 [i]
C4 x C4 x
where C2 = C3 = C4 = ρc x. So, to determine the transient response, we simply choose a suitable
value of τ and march through the calculations. The stability criterion is such that the coefficients
of the last term in each equation cannot be negative. For Equations (g), (h), and (i) the maximum
allowable time increment is
C x (1600)(800)(0.0075)2
τ max = 3 = = 12 s
2k (2)(3)
p
For Equation ( f ), the worst case is at the start when T5 = 20◦ C = 293 K. We have
(1600)(800)(0.0075)
C5 = = 4800
2
so that
4800
τ max =
(5.669 × 10−8 )(0.8)(12732 + 2932 )(1273 + 293) + 3.0/0.0075
= 9.43 s
p
For node 1 [Equation (e)] the most restrictive condition occurs when T1 = 293. We have
C1 = C5 = 4800

so that
4800
τ max =
(5.669 × 10−8 )(0.8)(2932 + 2932 )(293 + 293) + 3.0/0.0075
= 11.86 s

So, from these calculations we see that node 5 is most restrictive and we must choose τ < 9.43 s.
The calculations were performed with τ = 3.0 s, and the results are shown in
Figure Example 4-15b, c. Note that a straight line is obtained for the steady-state temperature
distribution in the solid, which is what would be expected for a constant thermal conductivity. To
compute the heat added at any instant of time we perform the sum

Q(τ) = Ci (Ti − 293) [ j]

and plot the results in Figure Example 4-15c.

EXAMPLE 4-16 Transient Conduction with Heat Generation

The plane wall shown has internal heat generation of 50 MW/m3 and thermal properties of k =
19 W/m · ◦ C, ρ = 7800 kg/m3 , and C = 460 J/kg · ◦ C. It is initially at a uniform temperature of
100◦ C and is suddenly subjected to the heat generation and the convective boundary conditions
indicated in Figure Example 4-16A. Calculate the temperature distribution after several time
increments.
Solution
We use this resistance and capacity formulation and write, for unit area,
1/R12 = kA/x = (19)(1)/0.001 = 19,000 W/◦ C

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.188 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch04 10/30/2008 17:34

CHAPTER 4 Unsteady-State Conduction 189

Figure Example 4-16a

hA = 400 W冫m2 • ˚C hB = 500 W冫m2 • ˚C

TA = 120˚C TB = 20˚C

1 2 3 4 5 6

Δx Δx

Δ x = 1.0 mm

All the conduction resistances have this value. Also,

1/R1A = hA = (400)(1) = 400 W/◦ C
1/R1B = hA = (500)(1) = 500 W/◦ C

The capacities are

C1 = C6 = ρ(x/2)c = (7800)(0.001/2)(460) = 1794 J/◦ C
C2 = C3 = C4 = C5 = ρ(x)c = 3588 J/◦ C

We next tabulate values.

Ci
 
Node (1/Rij ) Ci (1/Rij )
1 19,400 1794 0.092
2 38,000 3588 0.094
3 38,000 3588 0.094
4 38,000 3588 0.094
5 38,000 3588 0.094
6 19,500 1794 0.092

Any time increment τ less than 0.09 s will be satisfactory. The nodal equations are now written
in the form of Equation (4-47) and the calculation marched forward on a computer.
The heat-generation terms are
qi = q̇ Vi
so that
q1 = q6 = (50 × 106 )(1)(0.001/2) = 25,000 W
q2 = q3 = q4 = q5 = (50 × 106)(1)(0.001) = 50,000 W

The computer results for several time increments of 0.09 s are shown in the following table.
Because the solid stays nearly uniform in temperature at any instant of time it behaves almost like
a lumped capacity. The temperature of node 3 is plotted versus time in Figure Example 4-16B to
illustrate this behavior.
Number of time increments (τ = 0.09 s)
Node 5 20 100 200
1 106.8826 123.0504 190.0725 246.3855
2 106.478 122.8867 190.9618 248.1988
3 106.1888 122.1404 190.7033 248.3325
4 105.3772 120.9763 189.3072 246.7933
5 104.4622 119.2217 186.7698 243.5786
6 102.4416 117.0056 183.0735 238.6773

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.189 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_Ch04 10/30/2008 17:34

190 4-7 Thermal Resistance and Capacity Formulation

Number of time increments (τ = 0.09 s)

Node 500 800 1200 3000
1 320.5766 340.1745 346.0174 347.2085
2 323.6071 343.5267 349.4654 350.676
3 324.2577 344.3137 350.2931 351.512
4 322.5298 342.536 348.5006 349.7165
5 318.4229 338.1934 344.0877 345.2893
6 311.9341 331.2853 337.0545 338.2306

Figure Example 4-16b

350

300
Temperature, node 3, ˚C

250

200

150

100
0 20 40 60 80 100
Time, s

EXAMPLE 4-17 Numerical Solution for Variable Conductivity

A 4.0-cm-thick slab of stainless steel (18% Cr, 8% Ni) is initially at a uniform temperature of
0◦ C with the left face perfectly insulated as shown in Figure Example 4-17a. The right face is
suddenly raised to a constant 1000◦ C by an intense radiation source. Calculate the temperature
distribution after (a) 25 s, (b) 50 s, (c) 100 s, (d) an interval long enough for the slab to reach a
steady state, taking into account variation in thermal conductivity. Approximate the conductivity
data in Appendix A with a linear relation. Repeat the calculation for the left face maintained at 0◦ C.

Figure Example 4-17a

Insulated or constant 0˚C

1000˚C

1 2 3 4

Δx = 1 cm

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.190 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 191

Solution
From Table A-2 we have k = 16.3 W/m · ◦ C at 0◦ C and k = 31 W/m · ◦ C at 1000◦ C. A linear
relation for k is assumed so that
k = k0 (1 + βT)

where T is in degrees Celsius. Inserting the data gives

k = 16.3(1 + 9.02 × 10−4 T) W/m · ◦ C

We also have ρ = 7817 kg/m3 and c = 460 J/kg · ◦ C, and use the thermal resistance-capacitance
formula assuming that the resistances are evaluated at the arithmetic mean of their connecting
nodal temperatures; i.e., R3−4 is evaluated at (T3 + T4 )/2.
First, the thermal capacities are evaluated for unit area:
C1 = ρ(x/2)c = (7817)(0.01/2)(460) = 17,980 J/m2 · ◦ C
C2 = C3 = C4 = ρ(x)c = (7817)(0.01)(460) = 35,960 J/m2 · ◦ C

For the resistances we have the form, for unit area,

1/R = k/x = k0 (1 + βT)/x

Evaluating at the mean temperatures between nodes gives

1/R1−2 = (16.3)[1 + 4.51 × 10−4 (T1 + T2 )]/0.01 = 1/R2−1
1/R2−3 = (16.3)[1 + 4.51 × 10−4 (T2 + T3 )]/0.01 = 1/R3−2
1/R3−4 = (16.3)[1 + 4.51 × 10−4 (T3 + T4 )]/0.01 = 1/R4−3
1/R4−1000 = (16.3)[1 + 4.51 × 10−4 (T4 + T1000 )]/0.01 = 1/R1000−4

The stability requirement is most severe on node 1 because it has the lowest capacity. To be on the
safe side we can choose a large k of about 31 W/m · ◦ C and calculate
(17,980)(0.01)
τ max = = 5.8 s
31

The nodal equations are now written in the form of Equation (4-47); that is to say, the equation
for node 2 would be

p+1 τ p p p p
T2 = 1630 [1+ 4.51 × 10−4 (T1 + T2 )](T1 − T2 )
C2

−4 p p p p p
+1630 [1 + 4.51 × 10 (T3 + T2 )](T3 − T2 ) + T2

A computer solution has been performed with τ = 5 s and the results are shown in the tables. The
steady-state solution for the insulated left face is, of course, a constant 1000◦ C. The steady-state
distribution for the left face at 0◦ C corresponds to Equation (2-2) of Chapter 2. Note that, because
of the nonconstant thermal conductivity, the steady-state temperature profile is not a straight line.

Temperatures for left face at constant 0◦ C, τ = 5 s

Node 25 s 50 s 100 s Steady state
1 0 0 0 0
2 94.57888 236.9619 308.2699 317.3339
3 318.7637 486.5802 565.7786 575.9138
4 653.5105 748.1359 793.7976 799.7735

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.191 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

192 4-8 Summary

Temperatures for left face insulated, τ = 5 s

Node 25 s 50 s 100 s Steady state
1 30.55758 232.8187 587.021 1000
2 96.67601 310.1737 623.5018 1000
3 318.7637 505.7613 721.5908 1000
4 653.5105 752.3268 855.6965 1000

These temperatures are plotted in Figure Example 4-17b.

Figure Example 4-17b

Left face constant at 0˚C

1000 Left face insulated

Steady state

+
τ =100 s
800 τ =100 s

+
+
τ =50 s
+
Temperature,˚C

+
600
+

τ =25 s
400
+

+
+

200
+

00 1 2 3 4
x cm

The purpose of this example has been to show how the resistance-capacity formulation can
be used to take into account property variations in a rather straightforward way. These variations
may or may not be important when one considers uncertainties in boundary conditions.

4-8 SUMMARY
In progressing through this chapter the reader will have noted analysis techniques of varying
complexity, ranging from simple lumped-capacity systems to numerical computer solutions.
At this point some suggestions are offered for a general approach to follow in the solution
of transient heat-transfer problems.
1. First, determine if a lumped-capacity analysis can apply. If so, you may be led to a
much easier calculation.

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.192 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services
hol29362_ch04 10/14/2008 19:33

CHAPTER 4 Unsteady-State Conduction 193

2. Check to see if an analytical solution is available with such aids as the Heisler charts
and approximations.
3. If analytical solutions are very complicated, even when already available, move directly
to numerical techniques. This is particularly true where repetitive calculations must be
performed.
4. When approaching a numerical solution, recognize the large uncertainties present in
convection and radiation boundary conditions. Do not insist upon a large number of
nodes and computer time (and chances for error) that cannot possibly improve upon
the basic uncertainty in the boundary conditions.
5. Finally, recognize that it is a rare occurrence when one has a “pure” conduction problem;
there is almost always a coupling with convection and radiation. The reader should keep
this in mind as we progress through subsequent chapters that treat heat convection and
radiation in detail.

REVIEW QUESTIONS
1. What is meant by a lumped capacity? What are the physical assumptions necessary for
a lumped-capacity unsteady-state analysis to apply?
2. What is meant by a semi-infinite solid?
3. What initial conditions are imposed on the transient solutions presented in graphical
form in this chapter?
4. What boundary conditions are applied to problems in this chapter?
5. Define the error function.
6. Define the Biot and Fourier numbers.
7. Describe how one-dimensional transient solutions may be used for solution of two-
and three-dimensional problems.

LIST OF WORKED EXAMPLES

4-1 Steel ball cooling in air
4-2 Semi-infinite solid with sudden change in surface conditions
4-3 Pulsed energy at surface of semi-infinite solid
4-4 Heat removal from semi-infinite solid
4-5 Sudden exposure of semi-infinite slab to convection
4-6 Aluminum plate suddenly exposed to convection
4-7 Long cylinder suddenly exposed to convection
4-8 Semi-infinite cylinder suddenly exposed to convection
4-9 Finite-length cylinder suddenly exposed to convection
4-10 Heat loss for finite-length cylinder
4-11 Sudden cooling of a rod
4-12 Implicit formulation
4-13 Cooling of a ceramic
4-14 Cooling of a steel rod, nonuniform h
4-15 Radiation heating and cooling
4-16 Transient conduction with heat generation
4-17 Numerical solution for variable conductivity

# 101675 Cust: McGraw-Hill Au: Holman Pg. No.193 K/PMS 293 DESIGN SERVICES OF

Title: Heat Transfer 10/e Server: Short / Normal / Long

S4CARLISLE
Publishing Services