CHAPTER 19

THERMAL PROPERTIES

PROBLEM SOLUTIONS

Heat Capacity

19.1 Estimate the energy required to raise the temperature of 2 kg (4.42 lb m) of the following materials
from 20 to 100°C (68 to 212°F): aluminum, steel, soda–lime glass, and high-density polyethylene.

Solution

The energy, E, required to raise the temperature of a given mass of material, m, is the product of the specific
heat, the mass of material, and the temperature change, DT, as

E = c p m DT

The DT in this problem is equal to 100°C – 20°C = 80°C (= 80 K), while the mass is 2 kg, and the specific heats are
presented in Table 19.1. Thus,

E(aluminum) = (900 J/kg- K)(2 kg)(80 K) = 1.44 ´ 105 J

E(steel) = (486 J/kg- K)(2 kg)(80 K) = 7.78 ´ 104 J

E(glass) = (840 J/kg- K)(2 kg)(80 K) = 1.34 ´ 105 J

E(HDPE) = (1850 J/kg- K)(2 kg)(80 K) = 2.96 ´ 105 J

 19.2 To what temperature would 25 lb m of a 1025 steel specimen at 25°C (77°F) be raised if 125 Btu of
heat is supplied?

Solution

We are asked to determine the temperature to which 25 lb m of steel initially at 25°C would be raised if 125

Btu of heat is supplied. This is accomplished by utilization of a modified form of Equation 19.1 as

DQ
DT =
m cp

in which DQ is the amount of heat supplied, m is the mass of the specimen, and cp is the specific heat. From Table
19.1, cp = 486 J/kg-K for steel, which in Customary U.S. units is just

æ 2.39 ´ 10 -4 Btu/lb - °F ö
c p = (486 J/kg - K) ç m
÷ = 0.116 Btu/lb m - °F
è 1 J/kg - K ø

Thus

125 Btu
DT = = 43.1 °F
(25 lb m )(0.116Btu/lb m - °F )

and

Tf = T0 + DT = 77°F + 43.1°F = 120.1°F (49°C)

 19.3 (a) Determine the room temperature heat capacities at constant pressure for the following materials:
aluminum, silver, tungsten, and 70Cu-30Zn brass. (b) How do these values compare with one another? How do
you explain this?

Solution

(a) This problem asks that we determine the room-temperature heat capacities at constant pressure, Cp, for
aluminum, silver, tungsten, and 70Cu-30Zn brass. All we need do is multiply the cp values in Table 19.1 by the atomic

weights (values are found inside the front cover), taking into account the conversion from grams to kilograms (for the
atomic weights). Thus, for Al

C p = (900 J/kg- K)(1 kg/1000 g)(26.98 g/mol)= 24.3 J/mol- K

For Ag

C p = (235 J/kg- K)(1 kg/1000 g)(107.87 g/mol)= 25.4 J/mol- K

For W

C p = (138 J/kg- K)(1 kg/1000 g)(183.84 g/mol)= 25.4 J/mol- K

For brass it is first necessary to compute the alloy atomic weight (Aave) using Equation 4.11a as follows:

100
Aave =
CCu C Zn
+
ACu AZn

100
=
70 wt% 30 wt%
+
63.55 g / mol 65.41 g / mol

= 64.09 g/mol
Thus
C p = (375 J/kg- K)(1 kg/1000 g)(64.09 g/mol)= 24.0 J/mol- K

(b) These values of Cp are very close to one another because room temperature is considerably above the
Debye temperature for these metals; therefore, the values of Cp should be about equal to 3R [(3)(8.31 J/mol-K) = 24.9

J/mol-K], which is indeed the case for all four of these metals.
 19.4 For aluminum, the heat capacity at constant volume Cv at 30 K is 0.81 J/mol-K, and the Debye
temperature is 375 K. Estimate the specific heat (a) at 50 K and (b) a t 425 K.

Solution

(a) For aluminum, Cv at 50 K may be approximated by Equation 19.2, since this temperature is significantly
below the Debye temperature (375 K). The value of Cv at 30 K is given, and thus, we may compute the constant A as

Cv 0.81 J / mol - K
A = 3
= = 3.00 ´ 10 -5 J/mol - K 4
T (30 K) 3

Therefore, at 50 K

C v = AT3 = (3.00 ´ 10 -5 J/mol - K 4 ) (50 K) 3 = 3.75 J/mol - K

and

cv = (3.75 J/mol - K)(1 mol/26.98 g)(1000 g/kg) = 139 J/kg - K

(b) Since 425 K is above the Debye temperature, a good approximation for Cv is

C v = 3R

= (3)(8.31 J/mol- K) = 24.9 J/mol- K

And, converting this to specific heat

cv = (24.9 J/mol - K)(1 mol/26.98 g)(1000 g/kg) = 923 J/kg - K

 19.5 The constant A in Equation 19.2 is 12π 4R/5 q D
3 , where R is the gas constant and θ is the Debye
D

temperature (K). Estimate θD for copper, given that the specific heat is 0.78 J/kg-K at 10 K.

Solution

For copper, we want to compute the Debye temperature, qD, given the expression for A in Equation 19.2 and

the heat capacity at 10 K. First of all, let us determine the magnitude of A, as

Cv
A =
T3

(0.78 J/mol - K)(1 kg /1000 g)(63.55 g/mol)

=
(10 K) 3

= 4.96 ´ 10 -5 J/mol - K 4

As stipulated in the problem statement

12 p 4 R
A =
5q 3D

Or, solving for qD

æ 12 p 4 R ö1/3
q D = çç ÷÷
è 5A ø

é (12)(p) 4 (8.31 J/mol - K) ù1/ 3

= ê ú = 340 K
ë (5)( 4.96 ´ 10 -5 J/mol - K 4 ) û
 19.6 (a) Briefly explain why Cv rises with increasing temperature at temperatures near 0 K. (b) Briefly
explain why Cv becomes virtually independent of temperature at temperatures far removed from 0 K.

Solution

(a) The reason that Cv rises with increasing temperature at temperatures near 0 K is because, in this

temperature range, the allowed vibrational energy levels of the lattice waves are far apart relative to the available
thermal energy, and only a portion of the lattice waves may be excited. As temperature increases, more of the lattice
waves may be excited by the available thermal energy, and, hence, the ability of the solid to absorb energy (i.e., the
magnitude of the heat capacity) increases.
(b) At temperatures far removed from 0 K, Cv becomes independent of temperature because all of the lattice

waves have been excited and the energy required to produce an incremental temperature change is nearly constant.
 Thermal Expansion

19.7 An aluminum wire 10 m (32.8 ft) long is cooled from 38 to -1°C (100 to 30°F). How much change in
length will it experience?

Solution

In order to determine the change in length of the aluminum wire, we must employ a rearranged form of
Equation 19.3b and using the value of al taken from Table 19.1 [23.6 ´ 10-6 (°C)-1 ] as

Dl = l0a l DT = l0a l (Tf - T0 )

[ ]
= (10 m) 23.6 ´ 10 -6 (°C) -1 (-1°C - 38°C)

= - 9.2 ´ 10 -3 m = - 9.2 mm (-0.36 in.)

 19.8 A 0.1 m (3.9 in.) rod of a metal elongates 0.2 mm (0.0079 in.) on heating from 20 to 100°C (68 to
212°F). Determine the value of the linear coefficient of thermal expansion for this material.

Solution

The linear coefficient of thermal expansion for this material may be determined using a rearranged form of
Equation 19.3b as

Dl Dl 0.2 ´ 10 -3 m
al = = =
l0 DT l0 (Tf - T0 ) (0.1 m)(100 °C - 20°C)

= 25.0 ´ 10-6 (°C) -1

 19.9 Briefly explain thermal expansion using the potential energy-versus-interatomic spacing curve.

The phenomenon of thermal expansion using the potential energy-versus-interatomic spacing curve is
explained in Section 19.3.
 19.10 Compute the density for nickel at 500°C, given that its room-temperature density is 8.902 g/cm3.
Assume that the volume coefficient of thermal expansion, αv, is equal to 3αl.

Solution

In this problem we are asked to determine the density of nickel at 500°C. Let us use as the basis for this
determination 1 cm3 of material at 20°& ZKLFK KDVDP DVVRI J LWLVDVVXP HG W
KDWW
KLVP DVVZLOOUHP DLQ FRQVW
DQW
u pon heating to 500°C. Let us compute the volume expansion of this cubic centimeter of nickel as it is heated to
500°C. A volume expansion expression is given in Equation 19.4—viz.,

DV
= a v DT
V0

or

DV = V0a vDT

Also, a v = 3a l, as stipulated in the problem. The value of a l given in Table 19.1 for nickel is 13.3 ´ 10-6 (°C)-1.

Therefore, the volume, V, of this specimen of Ni at 500°C is just

( )
V = V0 + DV = V0 1 + a vDT = V0 1 + 3a l DT ( )

{ [ ]
= (1 cm 3 ) 1 + (3) 13.3 ´ 10 -6 (°C)-1 (500°C - 20°C) }

= 1.01915 cm 3

Thus, the dens ity is just the 8.902 g divided by this new volume—i.e.,

8.902 g
r = = 8.735 g/cm 3
1.01915 cm3
 19.11 When a metal is heated its density decreases. There are two sources that give rise to this
diminishment of ρ: (1) the thermal expansion of the solid, and (2) the formation of vacancies (Section 4.2).
Consider a specimen of copper at room temperature (20°C) that has a density of 8.940 g/cm3. (a) Determine its
density upon heating to 1000°C when only thermal expansion is considered. (b) Repeat the calculation when the
in troduction of vacancies is taken into account. Assume that the energy of vacancy formation is 0.90 eV/atom, and
that the volume coefficient of thermal expansion, αv is equal to 3αl.

Solution

(a) In this portion of the problem we are asked to determine the density of copper at 1000°C on the basis of
thermal expansion considerations. The basis for this determination will be 1 cm3 of material at 20°C; this volume of
copper has a mass of 8.940 g, which mass is assumed to remain constant upon heating to the 1000°C. Let us first
compute the volume expansion of this cubic centimeter of copper as it is heated to 1000°C. According to Equation
19.4 volume expansion is equal to

DV
= a vDT
V0

where av, the volume coefficient of thermal expansion, as stipulated in the problem statement, is equal to 3al. The
value of al given in Table 19.1 for copper is 17.0 ´ 10-6 (°C)-1 . Therefore, the volume of this specimen of Cu at

1000°C (V) is equal to

V = V0 + DV = V0 + V0a vDT = V0 (1 + a vDT)

é ù
= V0 (1 + 3a l DT) = V0 êë1 + 3a l (Tf - T0 )úû

{ [ ]
= (1 cm 3) 1 + (3) 17.0 ´ 10 -6 (°C)-1 (1000 °C - 20°C) }

= 1.04998 cm 3

Thus, the density is just the 8.940 g divided by this new volume—i.e.,

8.940 g
r = = 8.514 g/cm 3
1.04998 cm3

(b) Now we are asked to compute the density at 1000°C taking into consideration the creation of vacancies
which will further lower the density. To begin, this determination requires that we calculate the number of vacancies
using Equation 4.1. But it first becomes necessary to compute the number of Cu atoms per cubic centimeter (NCu) at

1000°C using Equation 4.2. Thus,

N A rCu
NCu =
ACu

(6.022 ´ 10 23 atoms / mol)(8.514 g/cm3)

=
63.55 g/mol

= 8.07 ´ 10 22 atoms/cm3

Now, from Equation 4.1, the total number of vacancies, Nv, is computed as

æ Q ö
N v = NCu exp ç- v ÷
è kT ø

é 0.90 eV / atom ù
= (8.07 ´ 10 22 atoms/cm 3 ) exp ê- -5 ú
ë (8.62 ´ 10 eV / K) (1000 + 273 K) û

= 2.212 ´ 10 19 vacancies/cm3

We now want to determine the number of vacancies per unit cell, which is possible if the unit cell volume is multiplied
by Nv. The unit cell volume (VC) may be calculated using Equation 3.5 taking n = 4 inasmuch as Cu has the FCC

crystal structure. Thus, from a rearranged form of Equation 3.5

nACu
VC =
rCu N A

(4 atoms/unit cell)(63.55 g/mol)

=
(8.514 g/cm3)(6.022 ´ 10 23 atoms / mol)

= 4.958 ´ 10 -23 cm3/unit cell

Now, the number of vacancies per unit cell, n v, is just

nv = N vVC
 = (2.212 ´ 1019 vacancies/cm3 )(4.958 ´ 10-23 cm3/unit cell)

= 0.001097 vacancies/unit cell

What this means is that instead of there being 4.0000 atoms per unit cell, there are only 4.0000 – 0.001097 = 3.998903
atoms per unit cell. And, finally, the density may be computed using Equation 3.5 taking n W
KXV

nACu
rCu =
VC N A

(3.998903 atoms/unit cell)(63.55 g/mol)

=
(4.958 ´ 10 -23 cm3/unit cell)(6.022 ´ 10 23 atoms / mol)

= 8.512 g/cm3

Thus, the influence of the vacancies is almost insignificant--their presence reduces the density by only 0.002 g/cm3
(from 8.514 g/cm3 to 8.512 g/cm3 ).
 19.12 The difference between the specific heats at constant pressure and volume is described by the
expression

a 2v v0T
c p - cv = (19.10)
b

where αv is the volume coefficient of thermal expansion, v0 is the specific volume (i.e., volume per unit mass, or the
reciprocal of density), β is the compressibility, and T is the absolute temperature. Compute the values of cv at room
temperature (293 K) for copper and nickel using the data in Table 19.1, assuming that αv = 3αl and given that the
values of β for Cu and Ni are 8.35 ´ 10 -12 and 5.51 ´ 10 -12 (Pa) -1, respectively.

Solution

This problem asks that we calculate the values of cv for copper and nickel at room temperature using
Equation 19.10, the data in Table 19.1, given that av = 3al, and also values of the compressibility. From Equation

19.10

a 2v v0T
cv = c p -
b

And, from Table 19.1 and the problem statement

cp(Cu) = 386 J/kg-K

cp(Ni) = 443 J/kg-K
av(Cu) = (3)[17.0 ´ 10-6 (°C)-1] = 5.10 ´ 10-5 (°C)-1
av(Ni) = (3)[(13.3 ´ 10-6 (°C)-1] = 3.99 ´ 10-5 (°C)-1

b(Cu) = 8.35 ´ 10-12 (Pa)-1

b(Ni) = 5.51 ´ 10-12 (Pa)-1

The specific volume is just the reciprocal of the density; thus, in units of m3/kg

1 æ 1 ö æ 1000 g ö æ 1 m ö3
v0 (Cu) = =ç ÷ç ÷ç ÷ = 1.119 ´ 10 -4 m 3/kg
r è 8.94 g/cm 3 ø è kg ø è 100 cm ø

æ 1 ö æ 1000 g ö æ 1 m ö3
v0 (Ni) = ç ÷
3 çè kg ÷ø ç 100 cm ÷
= 1.124 ´ 10 -4 m3 / kg
è 8. 90 g/cm ø è ø
Therefore, for copper

a 2v (Cu) v0 (Cu)T
cv (Cu) = c p (Cu) -
b(Cu)

= 386 J/kg - K -
[5.10 ]
´ 10 -5 (°C)-1 (1.119 ´ 10 -4 m3 / kg) (293 K)
8.35 ´ 10 -12 (N/m2 )-1

= 376 J/kg-K

And, also for nickel

cv (Ni) = 443 J/kg - K -

[3.99 ´ 10 -5 (°C)-1 ] (1.124 ´ 10 -4 m3 / kg) (293 K)
5.51 ´ 10 -12 (N/m2 )-1

= 433 J/kg-K
 19.13 To what temperature must a cylindr ical rod of tungsten 10.000 mm in diameter and a plate of 316
stainless steel having a circular hole 9.988 mm in diameter have to be heated for the rod to just fit into the hole?
Assume that the initial temperature is 25°C.

Solution

This problem asks for us to determine the temperature to which a cylindrical rod of tungsten 10.000 mm in
diameter must be heated in order for it of just fit into a 9.988 mm diameter circular hole in a plate of 316 stainless steel,
assuming that the initial temperature is 25°C. This requires the use of Equation 19.3a, which is applied to the
diameters of both the rod and hole. That is

d f - d0
= a l (Tf - T0 )
d0

Solving this expression for d f yields

é ù
d f = d0 ê1 + a l (T f -T0 )ú
ë û

Now all we need do is to establish expressions for d f(316 stainless) and d f (W), set them equal to one another, and
solve for Tf. According to Table 19.1, al(316 stainless) = 16.0 ´ 10-6 (°C)-1 and al (W) = 4.5 ´ 10-6 (°C)-1. Thus

d f (316 stainless) = d f (W)

[ {16.0 ´ 10 -6 (°C)-1} (Tf - 25°C)]

(9.988 mm) 1 +

= (10.000 mm) 1 +[ {4.5 ´ 10 -6 (°C)-1} (Tf - 25°C)]

Now solving for Tf gives Tf = 129.5°C
 Thermal Conductivity

19.14 (a) Calculate the heat flux through a sheet of steel 10 mm (0.39 in.) thick if the temperatures at the
W
ZR IDFHVDUH DQG & DQG ) DVVXP HVW
HDG\ -state heat flow. (b) What is the heat loss per hour if
2 2
the area of the sheet is 0.25 m (2.7 ft )? (c) What will be the heat loss per hour if soda–lime glass instead of steel is
used? (d) Calculate the heat loss per hour if steel is used and the thickness is increased to 20 mm (0.79 in.).

Solution

(a) The steady-VW

DWH KHDWIOX[ W
KURXJ K W
KH SODW
H P D\ EH FRP SXW
HG XVLQJ ( TXDW
LRQ W
KH W
KHUP DO
conductivity for steel, found in Table 19.1, is 51.9 W/m-K. Therefore,

DT
q =-k
Dx

é (100 + 273 K) - (300 + 273 K) ù

= - (51.9 W/m - K) ê ú
ë 10 ´ 10 -3 m û

= 1.04 ´ 10 6 W/m2

(b) Let dQ/dt represent the total heat loss such that

dQ
= qAt
dt

where A and t are the cross-sectional area and time, respectively. Thus,

dQ
= (1.04 ´ 10 6 J/s - m2 )(0.25 m 2 ) (60 s/min)(60 min/h)
dt

= 9.3 ´ 10 8 J/h (8.9 ´ 10 5 Btu/h)

(c) If soda-lime glass is used (k = 1.7 W/m-K, Table 19.1),

dQ DT
= - kAt
dt Dx

æ - 200 K ö
= - (1.7 J/s - m - K)(0.25 m 2 ) (3600 s/h) ç ÷
è 10 ´ 10 -3 m ø
 = 3.06 ´ 10 7 J/h (2.9 ´ 10 4 Btu/h)

(d) If the thickness of the steel is increased to 20 mm, then

dQ DT æ - 200 K ö
= - kAt = - (51.9 W/m - K)(0.25 m 2 ) (3600 s/h) ç ÷
dt Dx è 20 ´ 10 -3 m ø

= 4.7 ´ 10 8 J/h (4.5 ´ 10 5 Btu/h)

 19.15 (a) Would you expect Equatio n 19.7 to be valid for ceramic and polymeric materials? Why or why
not? (b) Estimate the value for the Wiedemann–Franz constant L [in Ω-W/(K) 2] at room temperature (293 K) for
the following nonmetals: silicon (intrinsic), glass-ceramic (Pyroceram), fused silica, polycarbonate, and
polytetrafluoroethylene. Consult Tables B.7 and B.9 in Appendix B.

Solution

(a) Equation 19.7 is not valid for ceramic and polymeric materials since, in the development of this
expression, it is assumed that free electrons are responsible for both electrical and thermal conduction. Such is the
case for most metals. For ceramics and polymers, free electrons are the primary contributors to the electrical
conductivity. However, free electrons do not contribute significantly to the thermal conductivity. For ceramics,
thermal conduction is primarily by means of phonons; for polymers, the energy transfer is made by chain vibrations,
translations, and rotations.
(b) Estimated room-temperature values of L, in W-W/(K)2, for the several materials are determined below.
Electrical conductivity values were determined by taking reciprocals of the electrical resistivities given in Table B.9,
$ SSHQGL[ % W
KHUP DOFRQGXFW
LYLW
LHVDUHW
DNHQ IURP 7 DEOH% LQ W
KHVDP HDSSHQGL[ Note: when a range of values is
given in these tables, an average value is used in the computation.)

For intrinsic silicon

k 141 W/m - K
L = = = 1203 W - W/K 2
sT é 1 ù
ê ú (293 K)
ë 2500 (W - m) û

For Pyroceram glass-ceramic

3.3 W / m - K
L = = 2.3 ´ 10 12 W - W/K 2
é 1 ù
ê ú (293 K)
ë 2 ´ 10 -14 (W - m) û

For fused silica

1.4 W / m - K
L = = 4.8 ´ 10 15 W - W/K 2
é 1 ù
ê 18 ú (293 K)
ë 10 (W - m) û

For polycarbonate
 0.20 W / m - K
L = = 1.4 ´ 10 11 W - W/K 2
é 1 ù
ê ú (293 K)
ë 2 ´ 10 14 (W - m) û

For polytetrafluoroethylene

0.25 W / m - K
L = = 8.5 ´ 10 13 W - W/K 2
é 1 ù
ê 17 ú (293 K)
ë 10 (W - m) û
 19.16 Briefly explain why the thermal conductivities are higher for crystalline than noncrystalline
ceramics.

Solution

Thermal conductivities are higher for crystalline than for noncrystalline ceramics because, for
noncrystalline, phonon scattering, and thus the resistance to heat transport, is much more effective due to the highly
disordered and irregular atomic structure.
 19.17 Briefly explain why metals are typically better thermal conductors than ceramic materials.

Solution

Metals are typically better thermal conductors than are ceramic materials because, for metals, most of the
heat is transported by free electrons (of which there are relatively large numbers). In ceramic materials, the primary
mode of thermal conduction is via phonons, and phonons are more easily scattered than are free electrons.
 19.18 (a) Briefly explain why porosity decreases the thermal conductivity of ceramic and polymeric
materials, rendering them more thermally insulative. (b) Briefly explain how the degree of crystallinity affects the
thermal conductivity of polymeric materials and why.

Solution

(a) Porosity decreases the thermal conductivity of ceramic and polymeric materials because the thermal
conductivity of a gas phase that occupies pore space is extremely small relative to that of the solid material.
Furthermore, contributions from gaseous convection are generally insignificant.
(b) Increasing the degree of crystallinity of a semicrystalline polymer enhances its thermal conductivity;
the vibrations, rotations, etc. of the molecular chains are more effective modes of thermal transport when a crystalline
structure prevails.
 19.19 For some ceramic materials, why does the thermal conductivity first decrease and then increase
with rising temperature?

Solution

For some ceramic materials, the thermal conductivity first decreases with rising temperature because the
scattering of lattice vibrations increases with temperature. At higher temperatures, the thermal conductivity will
increase for some ceramics that are porous because radiant heat transfer across pores may become important, which
process increases with rising temperature.
 19.20 For each of the following pairs of materials, decide which has the larger thermal conductivity.
Justify your choices.
(a) 3 XUHFRSSHU DOXP LQXP EURQ]H ZW &X-5 wt% Al).
(b) Fused silica; quartz.
(c) Linear polyethylene; branched polyethylene.
(d) Random poly(styrene-butadiene) copolymer; alternating poly(styrene-butadiene) copolymer.

Solution

This question asks for us to decide, for each of several pairs of materials, which has the larger thermal
conductivity and why.
(a) Pure copper will have a larger conductivity than aluminum bronze because the impurity atoms in the
latter will lead to a greater degree of free electron scattering.
(b) Quartz will have a larger conductivity than fused silica because fused silica is noncrystalline (whereas
quartz is crystalline) and lattice vibrations are more effectively scattered in noncrystalline materials.
(c) The linear polyethylene will have the larger conductivity than the branched polyethylene because the
former will have the higher degree of crystallinity. Linear polymers have higher degrees of crystallinity than
branched polymers. Since heat transfer is accomplished by molecular chain vibrations, and the coordination of these
vibrations increases with percent crystallinity, the higher the crystallinity, the greater the thermal conductivity.
(d) The alternating poly(styrene-butadiene) copolymer will have a higher crystallinity than the random
copolymer; alternating copolymers crystallize more easily than random ones. The influence of crystallinity on
conductivity is explained in part (c).
 19.21 We might think of a porous material as being a composite wherein one of the phases is a pore phase.
Estimate upper and lower limits for the room-temperature thermal conductivity of a magnesium oxide material
having a volume fractio n of 0.30 of pores that are filled with still air.

Solution

This problem asks that we treat a porous material as a composite wherein one of the phases is a pore phase,
and that we estimate upper and lower limits for the room-temperature thermal conductivity of a magnesium oxide
material having a 0.30 volume fraction of pores. The upper limit of k (k upper) may be determined using Equation 16.1

with thermal conductivity substituted for the elastic modulus, E. From Table 19.1, the value of k for MgO is 37.7
W/m-K, while for still air in the pore phase, k = 0.02 W/m-K (Section 19.4). Thus

kupper = Vp kair + VMgO kMgO

= (0.30)(0.02 W/m-K) + (0.70)(37.7 W/m-K) = 26.4 W/m-K

For the lower limit we employ a modification of Equation 16.2 as

kair kMgO
klower =
Vp kMgO + VMgO kair

(0.02 W/m - K)(37.7 W / m - K)

= = 0.067 W/m - K
(0.30)(37.7 W / m - K) + (0.70)(0.02 W / m - K)
 19.22 Nonsteady-state heat flow may be described by the following partial differential equation:

¶T ¶2T
= DT 2
¶t ¶x

where DT is the thermal diffusivity; this expression is the thermal equivalent of Fick’s second law of diffusion
(Equation 5.4b). The thermal diffusivity is defined according to

k
DT =
rc p

In this expression, k, ρ, and cp represent the thermal conductivity, the mass density, and the specific heat at
constant pressure, respectively.
(a) What are the SI units for DT?
(b) Determine values of DT for aluminum, steel, aluminum oxide, soda–lime glass, polystyrene, and nylon
6,6 using the data in Table 19.1. Density values are included in Table B.1, Appendix B.

Solution

(a) The units of DT are

k (J /s - m - K)
DT = = m2 /s
r(kg / m3) c p (J / kg - K)

(b) The values of DT for the several materials are given below. (Note: values for k and cp are taken from

7 DEOH GHQVLW
\ YDOXHVDUHIURP 7 DEOH% $ SSHQGL[ % DQG FRQYHUW
HG W
R XQLW
VRI NLORJ UDP VSHUP HW
HUFXEHG

For aluminum

k 247 W/m - K
DT = = = 1.0 ´ 10 -4 m 2 /s
r cp ( 2.71 ´ 10 3 kg/m3) (900 J/kg - K)

For steel

51.9 W/m - K
DT = = 1.4 ´ 10 -5 m2 /s
(7.9 ´ 10 3kg/m3) (486 J/kg - K)

For aluminum oxide

 39 W/m - K
DT = = 1.26 ´ 10 -5 m2 /s
(4 ´ 10 3 kg/m3) (775 J/kg - K)

For soda-lime glass

1.7 W/m - K
DT = = 8.1 ´ 10 -7 m2 /s
(2.5 ´ 10 3 kg/m3) (840 J/kg - K)

For polystyrene

0.13 W/m - K
DT = = 1.06 ´ 10 -7 m2 /s
(1.05 ´ 10 3 kg/m3) (1170 J/kg - K)

For nylon 6,6

0.24 W/m - K
DT = = 1.3 ´ 10 -7 m2 /s
(1.14 ´ 10 3 kg/m3) (1670 J/kg - K)
 Thermal Stresses

19.23 Beginning with Equation 19.3, show that Equation 19.8 is valid.

Solution

We want to show that Equation 19.8 is valid beginning with Equation 19.3. Upon examination of Equation
19.3b,

Dl
= a l DT
l0

it may be noted that the term on the left-hand side is the same expression as that for the definition of engineering
VW
UDLQ ( TXDW
LRQ W
KDWLV

Dl
e=
l0

Furthermore, elastic stress and strain are related through Hooke's law, Equation 6.5:

s = Ee

Making appropriate substitutions and algebraic manipulations gives

Dl s
= e = = a l DT
l0 E

Or, solving for s

s = Ea l DT

which is the form of Equation 19.8.

 19.24 (a) Briefly explain why thermal stresses may be introduced into a structure by rapid heating or
cooling. (b) For cooling, what is the nature of the surface stresses? (c) For heating, what is the nature of the
surface stresses?

Solution

(a) Thermal stresses may be introduced into a structure by rapid heating or cooling because temperature
gradients will be established across the cross section due to more rapid temperature changes at the surface than
within the interior; thus, the surface will expand or contract at a different rate than the interior and since this surface
expansion or contraction will be restrained by the interior, stresses will be introduced.
(b) For cooling, the surface stresses will be tensile in nature since the interior contracts to a lesser degree
than the cooler surface.
(c) For heating, the surface stresses will be compressive in nature since the interior expands to a lesser
degree than the hotter surface.
 19.25 (a) If a rod of 1025 steel 0.5 m (19.7 in.) long is heated from 20 to 80°C (68 to 176°F) while its ends
are maintained rigid, determine the type and magnitude of stress that develops. Assume that at 20°C the rod is
stress free. (b) What will be the stress magnitude if a rod 1 m (39.4 in.) long is used? (c) If the rod in part (a) is
cooled from 20 to -10°C (68 to 14°F), what type and magnitude of stress will result?

Solution

(a) We are asked to compute the magnitude of the stress within a 1025 steel rod that is heated while its ends
are maintained rigid. To do this we employ Equation 19.8, using a value of 207 GPa for the modulus of elasticity of
steel (Table 6.1), and a value of 12.0 ´ 10-6 (°C)-1 for al (Table 19.1). Therefore

s = Ea l (T0 - Tf )

[ ]
= (207 ´ 10 3 MPa ) 12.0 ´ 10 -6 (°C)-1 (20°C - 80°C)

= – 150 MPa (– 21,800 psi)

The stress will be compressive since its sign is negative.

(b) The stress will be the same [– 150 MPa (– 21,800 psi )], since stress is independent of bar length.
(c) Upon cooling the indicated amount, the stress becomes

s = Ea l (T0 - Tf )

[ ]
= (207 ´ 10 3 MPa ) 12.0 ´ 10 -6 (°C)-1 [(20°C - (-10°C)]

= +74.5 MPa (+10,900 psi)

This stress will be tensile since its sign is positive.

 19.26 A copper wire is stretched with a stress of 70 MPa (10,000 psi) at 20°C (68°F). If the length is held
constant, to what temperature must the wire be heated to reduce the stress to 35 MPa (5000 psi)?

Solution

We want to heat the copper wire in order to reduce the stress level from 0 3DW
R 0 3D LQ GRLQJ VR ZH
reduce the stress in the wire by 70 MPa – 35 MPa = 35 MPa, which stress will be a compressive one (i.e., s = – 35
MPa). Solving for Tf from Equation 19.8 [and using values for E and al of 110 GPa (Table 6.1) and 17.0 ´ 10-6 (°C)-1

(Table 19.1), respectively] yields

s
Tf = T0 -
Ea l

- 35 MPa
= 20°C -
(110 [
´ 10 3 MPa) 17.0 ´ 10 -6 (°C)-1 ]

= 20°C + 19°C = 39°C (101°F)

 19.27 If a cylindrical rod of nickel 100.00 mm long and 8.000 mm in diameter is heated from 20°C to
200°C while its ends are maintained rigid, determine its change in diameter. You may want to consult Table 6.1.

Solution

This problem asks for us to determine the change in diameter of a cylindrical nickel rod 100.00 mm long and
8.000 mm in diameter when it is heated from 20°C to 200°C while its ends are maintained rigid. There will be two
contributions to the diameter increase of the rod; the first is due to thermal expansion (which will be denoted as Dd 1),

while the second is from Poisson's lateral expansion as a result of elastic deformation from stresses that are
established from the inability of the rod to elongate as it is heated (denoted as Dd 2). The magnitude of Dd 1 may be

computed using a modified form of Equation 19.3 as

Dd1 = d0 a l (Tf - T0 )

From Table 19.1 the value of al for nickel is 13.3 ´ 10-6 (°C)-1. Thus,

[ ]
Dd1 = (8.000 mm) 13.3 ´ 10 -6 (°C)-1 (200 °C - 20°C)

= 0.0192 mm

Now, Dd 2 is related to the transverse strain (ex) according to a modified form of Equation 6.2 as

Dd2
= ex
d0

Also, transverse strain and longitudinal strain (ez) are related according to Equation 6.8:

e x = - ne z

where n is Poisson’s ratio. Substitution of this expression for ex into the first equation above leads to

Dd2
= - ne z
d0

Furthermore, the longitudinal strain is related to the modulus of elasticity through Equation 6.5—i.e.,
 s
ez =
E

And, therefore,

Dd2 s
= -n
d0 E

Now, from Equation 19.8 stress is equal to

s = Ea l (T0 - Tf )

which, when substituted into the preceding equation leads to

Dd2 nEa l (T0 - Tf )

= - = - na l (T0 - Tf )
d0 E

Solving for Dd 2 and realizing that, for nickel, n = 0.31 (Table 6.1) yields

Dd2 = - d0 na l (T0 - Tf )

= - (8.000 mm)(0.31) [13.3 ´ 10 -6 (°C)-1](20°C - 200 °C)

= 0.0059 mm

Finally, the total Dd is just Dd 1 + Dd 2 = 0.0192 mm + 0.0059 mm = 0.0251 mm.

 19.28 The two ends of a cylindrical rod of 1025 steel 75.00 mm long and 10.000 mm in diameter are
maintained rigid. If the rod is initially at 25°C, to what temperature must it be cooled to have a 0.008-mm
reduction in diameter?

Solution

This problem asks for us to determine to what temperature a cylindrical rod of 1025 steel 75.00 mm long and
10.000 mm in diameter must be cooled from 25°C in order to have a 0.008 mm reduction in diameter if the rod ends are
maintained rigid. There will be two contributions to the diameter decrease of the rod; the first is due to thermal
contraction (which will be denoted as Dd 1), while the second is from Poisson's lateral contraction as a result of elastic

deformation from stresses that are established from the inability of the rod to contract as it is cooled (denoted as
Dd 2). The magnitude of Dd 1 may be computed using a modified form of Equation 19.3b as

Dd1 = d0 a l (Tf - T0 )

Now, Dd 2 is related to the transverse strain (ex) according to a modified form of Equation 6.2 as

Dd2
= ex
d0

Also, transverse strain and longitudinal strain (ez) are related according to Equation 6.8:

e x = - ne z

where n is Poisson’s ratio. Substitution of this expression for ex into the first equation above leads to

Dd2
= - ne z
d0

Furthermore, the longitudinal strain is related to the modulus of elasticity through Equation 6.5—i.e.,

s
ez =
E

And, therefore,

Dd2 s
= -n
d0 E
Now, from Equation 19.8 stress is equal to

s = Ea l (T0 - Tf )

which, when substituted into the preceding equation leads to

Dd2 nEa l (T0 - Tf )

= - = - na l (T0 - Tf )
d0 E

And, solving for Dd 2 from this expression

Dd2 = - d0 na l (T0 - Tf )

The total Dd is just Dd = Dd 1 + Dd 2, and

Dd = d0a l (Tf - T0 ) + d0 na l (Tf - T0 ) = d0a l (Tf - T0 )(1 + n)

The values of n and al for 1025 steel are 0.30 and 12.0 ´ 10-6 (°C)-1, respectively (Tables 6.1 and 19.1). Incorporating,
into the above equation, these values, as well as those for Dd, d 0, and T0 cited in the problem statement gives

[ ]
- (0.008 mm) = (10.000 mm) 12.0 ´ 10 -6 (°C)-1 (Tf - 25°C ) (1 + 0.30)

And, finally, solving the above expression for Tf yields Tf = – 26.3°C.

 19.29 What measures may be taken to reduce the likelihood of thermal shock of a ceramic piece?

Solution

According to Equation 19.9, the thermal shock resistance of a ceramic piece may be enhanced by increasing
the fracture strength and thermal conductivity, and by decreasing the elastic modulus and linear coefficient of
thermal expansion. Of these parameters, sf and al are most amenable to alteration, usually be changing the

composition and/or the microstructure.

 DESIGN PROBLEMS

Thermal Expansion

19.D1 Railroad tracks made of 1025 steel are to be laid during the time of year when the temperature
averages 10°C (50°F). If a joint space of 4.6 mm (0.180 in.) is allowed between the standard 11.9-m (39-ft) long
rails, what is the hottest possible temperature that can be tolerated without the introduction of thermal stresses?

Solution

For these railroad tracks, each end is allowed to expand one-half of the joint space distance, or the track may
expand a total of this distance (4.6 mm). Equation 19.3a is used to solve for Tf, where the value al for the 1025 steel
[12.0 ´ 10-6 (°C)-1] is found in Table 19.1. Thus, solving for Tf from Equation 19.3a leads to

Dl
Tf = + T0
a l l0

4.6 ´ 10 -3 m
= + 10°C
[12.0 ´ 10-6 (°C)-1](11.9 m)

= 32.2°C + 10°C = 42.2°C (108°F)

 Thermal Stresses

19.D2 The ends of a cylindrical rod 6.4 mm (0.25 in.) in diameter and 250 mm (10 in.) long are mounted
between rigid supports. The rod is stress free at room temperature [20°C (68°F)] ; and upon cooling to -40°C (-
40°F), a maximum thermally induced tensile stress of 125 MPa (18,125 psi) is possible. Of which of the following
metals or alloys may the rod be fabricated: aluminum, copper, brass, 1025 steel, and tungsten? Why?

Solution

This is really a materials selection problem in which we must decide for which of the five metals listed, the
stress in the rod will not exceed 125 MPa (18,125 psi), when it is heated while its ends are mounted in rigid supports.
Upon examination of Equation 19.8, it may be noted that all we need do is to compute the EalDT product for each of

the candidate materials, and then note for which of them the stress is less than the stipulated maximum. [The value of
DT is T0 – Tf = 20°C – (– 40°C) = 60°C.] These parameters and their product for each of the alloys are tabulated below.
(Modulus of elasticity values were taken from Table 6.1, while the al values came from Table 19.1.)

Alloy al (°C)-1 E (MPa) alEDT (MPa)

Aluminum 23.6 ´ 10-6 69 ´ 103 98

Copper 17.0 ´ 10-6 110 ´ 103 112
Brass 20.0 ´ 10-6 97 ´ 103 116
1025 Steel 12.0 ´ 10-6 207 ´ 103 149
Tungsten 4.5 ´ 10-6 407 ´ 103 110

Thus, aluminum, copper, brass, and tungsten are suitable candidates.

 19.D3 (a) What are the units for the thermal shock resistance parameter (TSR)? (b) Rank the following
ceramic materials according to their thermal shock resistance: glass-ceramic (Pyroceram), partially stabilized
zirconia, and borosilicate (Pyrex) glass. Appropriate data may be found in Tables B.2, B.4, B.6, and B.7 of
Appendix B.

Solution

(a) This portion of the problem asks that we cite the units for the thermal shock resistance parameter (TSR).
From Equation 19.9

s f ( N / m2 ) k (W / m - K)
TSR = = W /m
E( N / m2 ) a l (°C)-1

(Note: in reducing units in the above expression, we have assumed that units of temperature in K and °C are
equivalent)

(b) Now we are asked to rank glass-ceramic (Pyroceram), partially-stabilized zirconia, and borosilicate (Pyrex)
glass as to their thermal shock resistance. Thus, all we need do is calculate, for each, the value of TSR using
Equation 19.9. Values of E, sf, al, and k are found, respectively, in Tab les B.2, B.4, B.6, and B.7, Appendix B. (Note:

whenever a range for a property value in these tables is cited, the average of the extremes is used.)
For the glass-ceramic

sf k
TSR =
E al

(247 MPa)(3.3 W/m - K)

= = 1045 W/m
(120 [
´ 10 3 MPa) 6.5 ´ 10 -6 (°C)-1 ]

For partially-stabilized zirconia

(1150 MPa)(2.7 W/m - K)

TSR = = 1578 W/m
(205 [
´ 10 3 MPa) 9.6 ´ 10 -6 (°C)-1 ]

And, for borosilicate glass

(69 MPa)(1.4 W/m - K)

TSR = = 418 W/m
(70 [
´ 10 3 MPa) 3.3 ´ 10 -6 (°C)-1 ]
 Thus, these materials may be ranked according to their thermal shock resistance from the greatest to the
least as follows: partially-stabilized zirconia, glass-ceramic, and borosilicate glass.
 19.D4 Equation 19.9, for the thermal shock resistance of a material, is valid for relatively low rates of
heat transfer. When the rate is high, then, upon cooling of a body, the maximum temperature change allowable
without thermal shock, ∆Tf, is approximately

sf
DT f =
Ea l

where σf is the fracture strength. Using the data in Tables B.2, B.4, and B.6 (Appendix B), determine ∆Tf for a glass-
ceramic (Pyroceram), partially stabilized zirconia, and fused silica.

Solution

We want to compute the maximum temperature change allowable without thermal shock for these three
ceramic materials, which temperature change is a function of the fracture strength, elastic modulus, and linear
coefficient of thermal expansion. These data and the DTf's are tabulated below. (Values for E, sf, and al are taken

from Tables B.2, B.4, B.6 in Appendix B.)

Material sf (MPa) E (MPa) al (°C)-1 DTf (°C)

Glass ceramic 247 120 ´ 103 6.5 ´ 10-6 317

Zirconia 1150 205 ´ 103 9.6 ´ 10-6 584
Fused silica 104 73 ´ 103 0.4 ´ 10-6 3562