LAPPEENRANTA UNIVERSITY OF TECHNOLOGY

Faculty of Technology
LUT Energy
Lecture Note 7

FUNDAMENTALS OF HEAT TRANSFER

Lecture Note
Ari Vepslinen
Janne Pitknen
Timo Hyppnen

2011

ISBN 978-952-265-127-3
ISBN 978-952-265-128-0 (PDF)
ISSN 1798-1336

Preface
Following over 170+ pages and additional appendixes are formed based on content of Course: Fundamentals of Heat
Transfer. Mainly this summarizes relevant parts on Book of Fundamentals of Heat and Mass Transfer (Incropera),
but also other references introducing same concepts are included. Students point of view has been considered with
following highlights:
Relevant topics are presented in a nutshell to provide fast digestion of principles of heat transfer.
Appendixes include terminology dictionary.
Totally 22 illustrating examples are connecting theory to practical applications and quantifying heat
transfer to understandable forms as: temperatures, heat transfer rates, heat fluxes, resistances and etc.
Most important Learning outcomes are presented for each topic separately.
The Book, Fundamentals of Heat and Mass Transfer (Incropera), is certainly recommended for those going beyond
basic knowledge of heat transfer.
Lecture Notes consists of 3 primary content-wise objectives:
Give understanding to physical mechanisms of heat transfer,
Present basic concepts and terminology relevant for conduction, convection and radiation
Introduce thermal performance analysis methods for steady state and transient conduction systems.
In first chapter basic concepts of heat transfer are introduced. Conservation of Energy or 1st Law of
Thermodynamics is presented as a general tool for heat transfer analysis. 3 different heat transfer modes:
conduction, convection and radiation; are shortly introduced. Definitions for common concepts and variables related
to thermal system analysis are summarized.
Stationary conduction chapter begins with introduction of Fouriers Law of heat conduction and its analogy to other
physical experimental based Laws of nature. Thermal property characteristics of materials relevant for conduction
heat transfer are summarized before introduction to General Heat Diffusion Equation. Differential energy balances
in different coordinate systems are formed for 1-dimensional steady state conduction analysis. Also concepts of
Thermal Resistance and Overall Heat Transfer Coefficient are presented and relevance as thermal circuit or
system analysis tool is shown.
Heat transfer mechanisms and methods for heat transfer performance analysis relevant for structural heat transfer
enhancement, more commonly known as Fins, are introduced. A short summary of different designs and
performance values is given as an ending of a Fin chapter.
Lumped Capacitance Method is presented as thermal performance analysis method of transient uniform
temperature profile systems. Spatial effects in transient conduction are shown by analytical approach: forming and
solving differential energy conservation equations in Cartesian, Cylindrical and Spherical coordinate systems.
Importance of Boundary Conditions for thermal system performance and analytical solutions are highlighted.
Fundamentals of convection heat transfer include introduction to physical phenomenon involving velocity, thermal
and concentration boundary layer developments are presented. Related flow dynamic and heat & mass transfer
equations are derived from fundamental conservation laws to provide scientific basis for engineering heat transfer
methods and tools for solving problems of convection. Definitions of dimensionless parameters and equations are
presented and relevance of them is shown in differential conservation balance equations and in empirical convection
correlations. All basic forms of convection are discussed. Forced and free convection, as well as external and
internal convection, are separately handled. Boiling and condensation as special modes of convection are shortly
introduced as a last, but not least meaningful, part of summarized fundamentals of convection heat transfer.
Types and thermal principles of most important industrial application of heat transfer, heat exchanger, are
highlighted. Two main thermal performance and design methods, Logarithmic mean temperature difference and
efficiency NTU method, are presented by thermal theory based to solution step wise approach with design charts
applicable for common types of heat exchangers.
Physical principles and quantities of most complex, and thus perhaps most interesting, form of heat transfer,
radiation, are introduced by means of Stefan-Boltzmanns and Wien displacement Laws and Black body radiation
functions. Spectral radiation properties of material surfaces are highlighted before giving finally two fundamental
radiation heat transfer solution methods: (1) based on Kirchoffs Law and (2) Radiation between two bodies.
Wishing Warm Autumn for All Interested,
Ari Vepslinen

Table of Contents
1.

INTRODUCTION ................................................................................................................... 1
1.1
1.2
1.3

2.

STATIONARY CONDUCTION .............................................................................................. 9

2.1
2.2
2.3
2.4
2.5

2.6
3.

Empirical Correlations
Analytical Solution for Flat Plate
Cylinder in Cross Flow
Sphere
Selection of Correlation
Learning Outcomes

FORCED CONVECTION - INTERNAL .............................................................................. 87

8.1
8.2
8.3
8.4

9.

Boundary Layers
Boundary Layer Equations
Similarity
Analogies
Learning Outcomes

FORCED CONVECTION - EXTERNAL ............................................................................. 73

7.1
7.2
7.3
7.4
7.5
7.6

8.

Classification of Convection Heat Transfer

Problem of Convection
Learning Outcomes

BOUNDARY LAYER THEORY........................................................................................... 60

6.1
6.2
6.3
6.4
6.5

7.

The Lumped Capacitance Method

Transient Conduction with Spatial Effects
Learning Outcomes

FUNDAMENTALS OF CONVECTION................................................................................ 55
5.1
5.2
5.3

6.

Fin Theory
Fins with Uniform Cross-Section
Performance
Learning Outcomes

UNSTEADY-STATE CONDUCTION .................................................................................. 39

4.1
4.2
4.3

5.

Fouriers Law
Thermal Properties
General Heat Diffusion Equation
1-Dimensional, Steady-state Conduction
2-Dimensional, Steady-state Conduction
Learning Outcomes

FINS ...................................................................................................................................... 31
3.1
3.2
3.3
3.4

4.

Modes of Heat Transfer

Conservation of Energy
Applications

Hydrodynamics
Thermal Perfromance
Convection Heat Transfer Correlations
Learning Outcomes

FREE CONVECTION ......................................................................................................... 101

9.1
9.2
9.3

Analytical Solutions for Flat Plate

Empirical Correlations
Learning Outcomes

10. BOILING & CONDENSATION .......................................................................................... 111

10.1
10.2
10.3
10.4
10.5
10.6

Principles of Boiling
Pool Boiling
Forced convection Boiling
Principles of Condensation
Drop-wise Condensation
Learning Outcomes

11. HEAT EXCHANGERS ....................................................................................................... 132

11.1 Heat Exchanger Types
11.2 Energy Balance

11.3 LMTD Method

11.4 Effectiveness-NTU Method
11.5 Learning Outcomes

12. RADIATION ...................................................................................................................... 150

12.1 Principles
12.2 Radiation Quantities
12.3 Radiation Properties
12.4 Radiation between Two Surfaces (Bodies)
12.5 Learning Outcomes

APPENDIXES
I Dimensionless Numbers
II Terminology
III Thermal Properties

NOMENCLATURE
SYMBOLS
Roman
A
B
c
C

C
C*
Cf
d
dh
D
e
E
f
F
g
G
h
hm
hfg
jH
k
K
L
m
NA
NTU
Nu
p
P
Pr
q
q
qm
qV
r
R
R

Re
s
S
St
t
T
U
v
V
w
W

[m2]
[m]
[ J / kgK ]
[mol/ m3]

Area
Width
Specific Heat
Concentration

[W / K ]
[-]
[-]
[m]
[m]
[m2/s], [m]
[m]
[W/m2]
[-]
[-]
[m2/s]
[ kg / m2s ], [ W / K ]
[ W / m2 K ]
[W/m2K]
[J/kg]
[-]
[ W / mK ]
[-]
[m]
[ kg ]
[mol/s]
[-]
[-]
[ Pa ]
[-]
[W]
[-]
[W]
[ W / m2 ]
[ kg / s ]
[ m3 / s ]
[m]
[K/W]
[ m2K / W ]
[ m2K / W ]
[ m2K / W ]
[-]
[m]
[m]
[-]
[m]
[s]
[K]
[ W / m2 K ]
[ m3 / kg ]
[m3]
[m/s]
[W]

Heat Capacity Rate

Ratio of Heat Capacity Rates in -Ntu Method.
Fanning Friction Factor (Friction Coefficient)
Diameter
Hydraulic Diameter
Diffusivity, Diameter
Surface Roughness
Emissive Power
Darcy (Moody) Friction Factor
Correction Factor in LMTDMethod
Gravitational Acceleration
Mass Velocity, Conductance
Convection Heat Transfer Coefficient
Convection Mass Transfer Coefficient
Latent Heat
Colburn J-Factor, St Pr2/3
Thermal Conductivity
Unit Resistance
Length
Mass
Convection Mass Transfer Rate
Number of Transfer Units, Dimensionless Conductance
Nusselt Number
Pressure
1. Temperature Effectiveness
2. Power
Prandtl Number
Heat Transfer Rate
Heat Flux
Mass Flow Rate
Volume Flow Rate
Radius
1. Thermal Resistance
Thermal Resistance per Area
1. RTc Contact Resistance per Area
2. RF Fouling Resistance per Area
Reynolds Number
1. Wall thickness, fin spacing
Spacing
Stanton Number
1. Thickness
2. Time
Temperature
Overall Heat Transfer Coefficient
Specific Volume
Volume
Flow Velocity
Work

Greek

f
o
p

SUPERSCRIPTS

*
SUBSCRIPTS
avg
c
D
e
f

ff
fr
h
i
L
lm
m

max
min
o
p
rad
s
t
T
tb
w

[m2 / s ]
[-]
[K-1]
[m2 / m3]
[m]
[m]
[-]
[-]
[kg / m3]
[W/m2K4]
[-]
[-]
[-]
[ Pa s ] or [kg/sm]
[m2 / s]
[N/m2]
[s]
[C]
[rad]

Per length
Per area
Per volume
Dimensionless

average
1. cold side of heat exchanger
2. core of the heat exchanger
diagonal
entry to heat exchanger core
1. fouling
2. fin
3. fluid
free-flow
frontal
hot side of heat exchanger
1. inlet
2. inside
longitudinal
logarithmic mean
1. mass
2. mean
3. modified
maximum
minimum
1. outlet
2. outside
isobaric
radiation
surface
total
transverse
tube-to-baffle
wall

1. Thermal Diffusivity, = K / Cp
2. Absorptivity: 0-1
Volumetric thermal expansion coefficient
Heat Transfer Area per Volume
Boundary layer thickness
Gap between plates
1. Heat Exchanger Effectiveness
2. Emissivity: 0-1
Density
Stefan-Boltzmann constant
Fin Efficiency
Overall Surface Efficiency
Pump/Fan Efficiency
Dynamic Viscosity
Kinematic Viscosity
1.Shear stress
2. Time constant
1. Temperature difference
2. Zenith angle

1 INTRODUCTION
1.1 MODES OF HEAT TRANSFER
As simply described, fundamental engineering heat transfer knowledge consist knowhow of evaluating rates three modes of heat transfer with specific conditions, properties
and geometries, and further applying that to design and performance analysis of heat
exchangers.
Table 1-1 Three modes of heat transfer and their basic rate equations

Conduction

Convection

Radiation

Heat transfer across medium.

Heat transfer between moving

Heat transfer in form of

fluid and surface.

electromagnetic waves emitted

by surfaces at a finite
temperature.

kA

dT
dx

hA TS

A TS4

T4

Figure 1-1 Heat transfer modes: Conduction, Convection and Radiation

1.1.1 Conduction
Conduction is transfer of energy from more energetic particles to less energetic ones
due to interaction between atomic and molecular particles.
In Solids conduction is due to combination of
(1) Vibrations of the molecules in a lattice and (2) Energy transport by free
electrons
In gases and liquids conduction is
(1) Collision and (2) Diffusion of molecules during their random motion.

Fouriers Law of Heat Conduction

q

Figure 1-2

kA

dT
dx

(1-1)
k
A

Thermal conductivity [kJ/mK]

Area perpendicular to direction of heat transfer [m2]

dT dx

Temperature gradient [K/m]

One-dimensional heat transfer by conduction

Thermal conductivity is rate of heat transfer through a unit thickness of material per
unit area per unit temperature difference. Thermal conductivity is property of material
and dependent on temperature (read chapter 2). Just to give the idea on range of
conductivities: changing from insulation to metal thermal conductivity varies with
0.001-100 W/mK.

Figure 1-3 Range of Conductivities

1.1.2 Convection
Convection is classified according the nature of the flow:
Free (Natural) Convection Flow is induced by buoyancy forces, which are
caused by temperature variations formed due to heat transfer in the fluid.
Forced Convection Flow is caused by external means: fan, pump, wind, etc.
Commonly, convection is sensible energy transfer of the fluid. Two special convection
heat transfer cases associated with phase change between liquid and vapour states of
fluid are (1) boiling and (2) condensation.

Figure 1-4

Convection heat transfer classification: forced, free, (boiling, condensation)

Convection heat transfer is combination of two mechanisms:

Random molecular motion diffusion (Conduction)
And bulk/macroscopic fluid motion.
Two types of boundary layers are associated with fluid flow near surface and
convective heat transfer. Hydrodynamic Boundary Layer is region of the fluid, where
velocity varies from zero at surface to velocity u of bulk fluid flow. Thermal
Boundary Layer is region of the fluid, where temperature varies from surface
temperature to bulk fluid temperature. At the interface (surface) temperatures of solid
and fluid are same and velocity of fluid is zero and heat is transferred only by random
molecular motion (conduction).

Figure 1-5

Three Boundary layers: temperature, concentration and velocity boundary layers

Newtons Law of Cooling

Convection heat transfer rate is expressed commonly as Newtons law of cooling:
q

hA TS

(1-2)

TS

Convection heat transfer coefficient [W/Km2]

Area perpendicular to heat transfer [m2]
Surface temperature [K]

Fluid temperature [K]

h
A

Convection heat transfer coefficient is experimentally determined parameter and it is

function of surface geometry, nature of fluid motion, properties of fluid and bulk fluid
velocity.

1.1.3 Radiation
Radiation is energy emitted by matter that is at nonzero temperature in form of
electromagnetic waves. Electromagnetic waves are caused by electronic configuration
changes of atoms and molecules. Contrarily to conduction and convection, energy
transfer due radiation is most efficient in vacuum. Even radiation is volumetric
phenomenon, it is typically considered as transfer phenomenon between solid surfaces.
Stefan-Boltzmann Law
Surface that emits maximum rate of radiation is called blackbody or ideal radiator.
Upper limit of emissive heat transfer is given by Stefan-Boltzmann law as
Eb
TS4
(1-3)
Eb

= Emissive power of blackbody [W/m2]

= Stefan-Boltzmann constant [W/m2K4]

TS

= Absolute temperature of the surface [K]

Heat flux emitted by real surface is less than that of blackbody at the same temperature
and is given as
E

TS4

(1-4)

= Emissive power of real surface [W/m2]

= Emissivity, 0

Emissivity depends strongly on material and finish of surface. Absorptivity

is
another surface radiation property. Absorptivity is the fraction of the radiation energy
incident on a surface that is absorbed and its value varies between 0 and 1. Blackbody
is perfect absorber (
1 ). Radiation incident on a surface from its surroundings is
called irradiation, G, and rate, which irradiation is absorbed to surface is
Gabs

(1-5)

Gabs
G

= Absorbed radiation [W/m ]

= Absorptivity, 0
1
2
= Irradiation [W/m ]

Figure 1-6 Radiation heat transfer

Kirchhoffs Law of Radiation

Emissivity and absorptivity of a surface are equal at the same temperature and
wavelength. Usually, dependence on temperature and wavelength are ignored by
approximating emissivity and absorptivity to be equals (a gray surface).
q' '

TS4

T4

(1-6)

TS

Radiation heat flux [W/m2]

gray surface emissivity
Temperature at surface [K]

Temperature at surroundings [K]

1.2 CONSERVATION OF ENERGY

1st Law of Thermodynamics
The Law of conservation of energy over a time interval: The increase in the amount of
energy stored in a control volume must equal the amount of energy that enters the
control volume, minus the amount of energy that leaves the control volume. For closed
system 1st law of thermodynamics over a time interval can be stated as
E sttot

Q W
Esttot
Q
W

(1-7)
= Change in total energy stored in system
= Net heat transferred to system
= Net work done by the system

Figure 1-7 Energy balance for a) a closed system over a time interval b) open system (control volume) at
instant time given as rates.

Total energy consists of mechanical energy, which is combination of kinetic and

potential energies, and internal energy. While studying heat transfer, thermal energy is
form of internal energy to be focused on. Statement of the first law of thermodynamics
that is suitable for heat transfer analysis for control volume (also open system) can be
given as Thermal and mechanical energy equation:
dE st
(1-8)
E st
Ein Eout E g
dt
Est
Ein / out
Eg

= Stored thermal and mechanical energy

= Energy entering/leaving system
= Thermal energy generation

Energy storage and generation are volumetric phenomena and are usually proportional
to magnitude of volume. For control volume, thermal energy generation can be
chemical, electrical, electromagnetic or nuclear energy conversion. The inflow and
outflow terms are surface phenomena and are generally proportional to surface area.
Energy can be transferred across surface of control volume in forms of heat, work and
mass containing thermal and mechanical energies.

For steady-state open system having no thermal energy generation, energy equation
over control volume can be written as
m(ut

pv

1 V 2 gz )in
2
ut
pv
1 V2
2
gz

m(ut

pv

1 V2
2

gz )out

q W

0 (1-9)

= Specific internal energy

= Specific flow work
= Specific kinetic energy
= Specific potential energy

Figure 1-8 Energy balance for a steady-flow, open system

For Systems having negligible kinetic and potential energy changes and negligible
work, this can be further reduced for ideal gases or incompressible liquids to simplified
steady-flow thermal energy equation:
q

1.2.1

mc p (Tout

Tin )

(1-10)

Surface energy Balance

The most applied form of conservation of energy equation in heat transfer problems is
surface energy balance
E In

EOut

(1-11)

, which in case of conduction and convection is

q conv

q cond .

(1-12)

Figure 1-9 Surface Energy Balance: Conduction Convection (Wendl)

1.3 APPLICATIONS
Fields of heat transfer applications are for example:
Heat Exchangers: at power plants, etc.
Cooling of Electronic equipments
Buildings: insulations and air-conditioning
Refrigeration
Human body, etc.
Engineering Heat Transfer problems can be divided to two groups:
1) Rating: Determination of heat transfer rate for system having specified
temperature difference.
2) Sizing: Determination of size of the system to transfer heat at specified rate for
a specified temperature difference.

2 STATIONARY CONDUCTION
2.1 FOURIERS LAW
1-Dimensional Steady-State form of Fouriers Law states for conduction heat transfer
rate
qx

kA

dT
dx

(2-1)

, and for conduction heat flux

q"x

dT
.
dx

(2-2)

Direction of conduction heat flow is always normal to surface of constant temperature,

isothermal surface. Generalization of conduction rate equation to 3-dimensional form
gives
q"

k T

T
T
j
i k
y
x

T
k
z

(2-3)

2.1.1 Analogy of Fouriers, Ohms and Ficks Laws

Origin of Fouriers Law is phenomenological meaning that it is derived from observer
phenomenon rather than first principles of physics. Fourier Law has various numbers of
important analogies: Ohms and Ficks Laws being examples of electrical and mass
transfer analogies.
Table 2-1 Analogy of Fouriers, Ohms and Ficks Laws
Fouriers
Law

Conduction heat flux

q"

k T

Ohms Law

Electrical current density

J"

Ficks Law

Mass diffusion flux

q"

D m

Thermal conductivity k
Temperature gradient T
Electrical conductivity
Voltage gradient V
Diffusion coefficient D
Mass concentration gradient

2.2 THERMAL PROPERTIES

Thermophysical properties of materials are classified to two categories: transport and
thermodynamic properties. Thermal conductivity is transport property, and density and
specific heat are common thermodynamic properties (Table 2-2).
Table 2-2 Transport and thermodynamical material properties (Note analogy in transport properties)
Transport
Thermal conductivity

Heat Transfer, Conduction: q

Diffusivity

Mass Transfer, Diffusion: q

Viscosity

Friction, shear stress:

"

"

k T
D m

Thermodynamic
Density
Specific heat

cp

Volumetric heat capacity

Describes the ability of system to store thermal energy

cp

Thermal diffusivity
k c p [m2/s] is ratio of heat conducted through the material to
heat stored per unit volume.

2.2.1 Thermal Conductivity

As stated earlier in Fouriers Law, thermal conductivity is defined as
"
qx
k
[W/mK]
T/ x

(2-4)

Conductivity depends on physical atomic and molecular structure of matter, which are
related to state of matter.
Solids: Thermal conductivity for solids can be expressed as
k

ke

kl

(2-5)

ke
kl

Conductivity associated with freely moving electrons

Conductivity related to vibration of lattice

Structure of refractory materials is porous and their effective thermal conductivity is

formed as a sum of different heat transfer modes: conduction, convection and radiation.
Liquids and gases: For fluids thermal energy is transported with molecular motion. As
molecular spacing is much larger in case of liquid and gases compared to solids,
thermal energy transport is less effective, therefore meaning smaller conductivities than
that of solids. Similarly gases have generally smaller conductivities than liquids.
Typical conductivities and temperature dependence of materials are shown in Figure
2-1 figure.
10

Figure 2-1

Range and temperature dependency of thermal conductivities of various materials

11

2.3 GENERAL HEAT DIFFUSION EQUATION

Suitable statement of the first law of thermodynamics for heat transfer analysis for
control volume, as described earlier, can be given as Thermal and mechanical energy
equation:
E st

dE st
dt

Ein

Eout

Est
Ein / out
Eg

Eg

(2-6)

= Stored thermal and mechanical energy

= Energy entering/leaving system
= Thermal energy generation

Figure 2-2 Differential Cartesian control volume for conduction analysis

The conduction heat rates at opposite surfaces can be given as Taylor series expansion:
qx

dx

qx

qy

dy

qy

qz

dz

qz

qx
dx
x

(2-7)

qy

dy
y
qz
dz
z

(2-8)
(2-9)

Energy source term can written as

Eg

q dx dy dz

(2-10)

q = Energy generation rate in unit volume W/ m 3

Change of thermal energy in control volume with time can be written as

E st

cp

T
dx dy dz
t

(2-11)

12

Writing energy equation by substituting conduction rates as energy flow terms,

following form can be obtained:

qx

qy

qz

qx
dx
x

qx

qy

qx
dx
x

qy
dy
y

qy

qz
dz
z

dy

qz

qdxdydz

qz
dz
z

cp

q dxdydz

T
dxdydz
t

cp

T
dxdydz
t

(2-12)

Conduction heat rates can be written according Fouriers Law:

qx

k dy dz

T
x

(2-13)

qy

k dx dz

T
y

(2-14)

qz

k dx dy

T
z

(2-15)

By substituting conduction heat rates to energy equation the general form of heat
diffusion equation (also heat equation) in Cartesian coordinates can be obtained
k

T
x

T
y

T
z

cp

T
t

(2-16)

Heat diffusion equation is simplified, if thermal conductivity is constant (isotropic

conditions), and can be written as
2

q
k

1 T
t

(2-17)

Initial Condition
As heat equation is first order in time, only one condition, termed initial condition has
to be specified. Typical condition is known temperature distribution T(x, y, z, t=0) at
time t.

13

Boundary Conditions
For each coordinate (2nd order) in heat equation two boundary conditions has to given
to describe system. Typical boundary conditions are shown in Figure 2-3.

Figure 2-3 Boundary conditions for heat diffusion equation at the surface (Incropera)

14

2.4 1-DIMENSIONAL, STEADY-STATE CONDUCTION

General heat diffusion equation can written under steady-state, one-dimensional
conditions with no heat generation as

d 2T
dx 2

d dT
dx dx

(2-18)

One-dimensionality approximation is appreciable, when temperature gradient in one

direction is significantly greater than in others, that is
T
T
,
x
y

T
x

T
.
z

2.4.1 Plain Wall

Let us consider a plane wall (Figure 2-4), where heat is transferred via convection from
hot fluid to one surface, via conduction across wall and via convection from other
surface to cold fluid.

Figure 2-4 a) Temperature distribution of plain wall b) equivalent thermal circuit

For determining temperature distribution and conduction heat transfer rate across wall
general heat diffusion equation is written for one-dimensional conduction and constant
thermal conductivity as

d 2T
dx 2

d dT
dx dx

(2-19)

, which can be further solved as follows

dT
dx

0dx

dT
=C1
dx

15

dT

C1dx

and gives general solution:

T x

C1 x C2

(2-20)

Boundary Conditions: in case both surface temperatures are known

1. x = 0: T = Ts,1

Ts,1

C1 0 C2 eli C2

2. x = L: T = Ts,2

Ts ,2

C1 L

Ts,1 eli C1

Ts,1
Ts, 2

Ts ,1
L

From boundary conditions general solution forms temperature profile solution for plain
wall:
T x

Ts,2

Ts ,1

x
L

Ts,1

(2-21)

Conduction Heat Transfer Rate is given by Fouriers Law and by substituting solved
boundary condition results following form is achieved:
qx

kA

dT d
=
dx dx
qx

dT
dx

Ts ,2

Ts,1

k
L

Ts,1

Ts,2

Ts,1
L

kA
T
T
L s,1 s,2

(2-22)

2.4.2 Thermal Resistance & Overall Heat Transfer Coefficient

Thermal resistance for conduction in a plane wall (Figure 2-4) is determined as
Rt ,cond

L
.
Ak

(2-23)

Conduction heat transfer rate across plane wall is

Ak
(T
T )
L s,1 s,2

T
Rt ,cond .

(2-24)

, which is analogical to definition of electrical current with relationship of voltage and

electrical resistance
U
R .

16

Thermal resistance for convection and radiation are defined similarly:

1
hA
1
hr A

Rt , conv .

Rt , rad

Ts

(2-25)

Ts T
qrad

(2-26)

Total resistance can be calculated for resistances in series with

Rt ,tot

Rt ,i

(2-27)

And for resistances in parallel with

I
Rt ,tot

I
Rt ,i

(2-28)

Conduction heat transfer rate in plane wall (Figure 2-4) can be expressed as
qx

,1
i

,2

UA(T

Rt ,tot

,1

,2 )

(2-29)

And Overall heat transfer coefficient is defined as

1
.
ARtot

(2-30)

The Equivalent Thermal Circuit

Based on equivalent thermal circuits (Figure 2-4) for plain wall heat transfer rate can
determined as

qx

Ts ,1
1 h1 A
,1

Ts ,1 Ts , 2
L kA

Ts 2 T , 2
1 h2 A

(2-31)

2.4.3 Contact Heat Resistance

Surface roughness in composite systems forms additional heat resistance, which is
called thermal contact resistance.
For mating surface system shown in Figure 2-5, thermal contact resistance is defined as

R' 't , c

TA TB
q' ' x

(2-32)

17

Figure 2-5 Temperature drop due thermal contact resistance

For smooth surfaces in contact with small characteristic gap width L, contact resistance
can be approximated with relationship to interfacial gas conductivity as

R"t , c

(2-33)

k gas

Generally contact resistance is experimentally defined for different interfacial fluids,

contacting materials and surface roughness.
Example I: Wall structure - Resistances in series
A typical wall structure of a Finnish detached house is described on a picture below. Calculate the heat
loss through the wall and the temperature distribution in the wall, when temperature outer side of the wall
is -27 C and inner side of the wall +20 C.
Structure
Heat resistance
Inner surface

Gypsum

Polyethen plastic
Insulator

Weather shield Board

Air gap
Brick

Outer surface

R1 = 1/(hconv + hw) = 0,15 m2K/W

hconv = 3 W/ m2K
hw = 3,7 W/ m2K
k = 0,23 W/ mK
s = 0,012 m
R2 = s/k = 0,052 m2K/W
k = 0,045 W/ mK
s = 0,10 m
R4 = s/k = 2,22 m2K/W
k = 0,09 W/ mK
s = 0,02 m
R5 = s/k = 0,222 m2K/W
R6 = 0,2 m2K/W
s = 0,015 m
k = 0,07 W/ mK
s = 0,08 m
R7 = s/k = 0,144 m2K/W
R8 = 1/(hconv + hw) = 0,05 m2K/W
hconv = 15 W/ m2K
hw = 5 W/ m2K

ASSUMPTIONS: From the surface (5) to the air and further from the air to the surface (6) heat is
transferred mainly by convection. If we assume that the heat transfer coefficient for air is h = 10 W/m2K,
the heat resistance R6 is then R6 = 1/10 + 1/10 = 0,2 m2K/W. Notice that the conduction schema (s/k)
would give a noticeably different value for the heat resistance in the air gap.
At first we calculate the overall heat resistance, and after that heat loss power and finally the temperature
variations.

18

= (0,15 + 0,052 + 0 + 2,22 + 0,222 + 0,20 + 0,114 + 0,05) m K/W

= 3,0 m K/W

Overall heat transfer coefficient:

Heat loss:

= 0,332 W/ m K

= 0,322 W m K [20

Temperature differences:
W
m K
=
= 15,6
0,15
= 2,3 K
m
W
= 15,6 0,052 = 0,8 K
= 15,6 0 = 0 K
= 15,6 2,22 = 34,6 K
= 15,6 0,222 = 3,5 K
= 15,6 0,2 = 3,1 K
= 15,6 0,114 = 1,7 K
=
= 15,6 0,05 1,0 K

( 27)] K = 15,6 W/m

Temperatures
= 17,7 C (= )
= 16,9 C
= 16,9 C
17,7 C
21,2 C
24,3 C
26,0 C
27 C (= )

In this example the convection heat transfer coefficients hconv = 3 W/m2K (inner surface) and hconv = 15
W/m2K (outer surface) are based on empirical and average values. At the outer side of the wall hconv
depends strongly on wind and varies after that.

Example II: Window - Circuit Analogy

A manufacturer of pre-fabricated components for high-rise buildings is determining the thermal rating of
a new panel design. The system consists of a header made up of an exterior cladding material having
thermal conductivity kc, and an inner insulator of conductivity ki, window glass with conductivity k g, and
a footer of the same construction as the header. It can be assumed that heat transfer occurs one
dimensionally through the panel. The panel height and width are w and L and the header, pane, and footer
each have a height of w = 3. The total thickness is t, half of which is cladding and half of which is
insulating material.
What would the thermal circuit for this system be assuming that there is a thermal contact resistance of
Rc between the cladding and insulator components?

19

ANALYSIS:
The thermal resistance at the outer side of the panels surface (Eq. 2-25).
=

The conductive resistances for glass, cladding and insulation from Eq. 2-22.
=
=
=

/3

/2
/3

/2
/3

The contact resistance between the cladding and insulation (Eq. 2-33).
=
Thermal circuit:

2.4.4

The Cylinder

Cylindrical systems usually have temperature gradient in radial direction only and can
therefore be treated as one-dimensional in spherical coordinate system. For onedimensional, steady-state conditions with no heat generation radial conduction heat
equation can be expressed as
1
T
kr
0
(2-34)
r r
r
By estimating constant thermal conductivity k and further differentiating as follows
dT
dr
dr

kd r

d r

0 dr

dT
r
= C1 :
dr
dr

dT

The general solution can be obtained:

20

dT
dr
C1

0dr
dr
r

Tr

C1 ln r C2

(2-35)

Figure 2-6 Hollow cylinder: Temperature distribution and equivalent thermal circuit

Boundary Conditions: for hollow cylinder shown in Figure 2-6

1. r = r1: T = Ts,1

=>

Ts,1 = C1 lnr1+C2

2. r = r2: T = Ts,2

=>

Ts,2 = C1 lnr2+C2

By solving constants as follows

C1

Ts,1-Ts,2= C1 lnr1 - C1 lnr2

Ts,1

Ts,1 - Ts,,2
r
In 1
r2

ln r1 C2

C2

Ts ,1 Ts , 2
ln r1 ln r2

Ts ,1

Ts ,1 Ts , 2
ln

Ts ,1 Ts , 2
r
ln 1
r2

r1
r2
ln r1

Solution for temperature distribution can be obtained:

T (r )

T (r )

C1 ln r C2

Ts,1 Ts,2
r
ln 1
r2

Ts ,1 Ts , 2
r
ln 1
r2
ln r

ln r1

ln r Ts ,1

Ts ,1

21

Ts ,1 Ts , 2
r
ln 1
r2

ln r1

(2-36)

r
r1

Ts ,1 Ts , 2 ln

T (r )

Ts ,1

r
ln 1
r2

(2-37)

Conduction Heat Transfer Rate is given by Fouriers Law and by substituting solved
boundary conditions following form for conduction heat rate is achieved:
dT
dr

qr

kA

dT
dr

d T (r )
dr

Ts,1 Ts,2
r ln

k 2 rL Ts ,1
r
r ln 1
r2

qr

r1
r2
Ts,2

(2-38)

Total thermal resistance for hollow cylinder can be defined from equivalent thermal
circuit and radial heat transfer can be expressed as

Rtot

qr

1
A1h1
T

,1

ln r2 / r1
2 Lk
T

Rtot

,2

1
A2h2

1
D1 Lh1

ln D2 / D1
2 Lk

1
D2 Lh2

(2-39)

(2-40)

22

2.4.5 Conduction with Internal Heat Source

Let us consider steady-state, one-dimensional conduction in plane wall (Figure 2-7)
with constant thermal conductivity having appropriate form of heat equation as
d 2T
dx 2

q
k

(2-41)

The general solution of heat equation is:

T x

q 2
x
2k

C1 x

C2

(2-42)

Figure 2-7 Conduction in plane wall with uniform heat generation - Boundary conditions: a)
asymmetrical b) Symmetrical c) adiabatic surface at mid plane.

Boundary Conditions
General solution can be further solved for three common boundary conditions (a-c) as
follows:

a) Asymmetrical case: known surface temperatures

T(-L) = Ts,1
T(L) = Ts,2
By solving integration constants temperature profile can be expressed as
T x

qL
1
2k

x2
L2

Ts ,2

Ts,1 x
2
L

23

Ts,1

Ts ,2
2

(2-43)

b) Symmetrical case
Resulting temperature profile can be expressed as
T x

qL
1
2k

x2
L2

Ts ,

(2-44)

With maximum temperature at midline of wall, x = 0

qL2
T ( 0) T0
Ts
2k

(2-45)

c) Adiabatic surface at midline

Symmetric temperature profile (b) results temperature gradient at midline to be (dT/dx)
= 0, which means zero heat flux across midline. Thus temperature profile equation is
same for (b) symmetric and (c) adiabatic surface at midline cases.
Surface temperature for cases (b) and (c) can be solved from energy equation. By
considering that all thermal energy generated in wall is transferred to boundary via
convection energy equation and surface temperature can be written as
Eg

qV

Ts

hA(Ts

T )

(2-46)

qV
.
hA

(2-47)

24

2.5 TWO DIMENSIONAL STEADY-STATE CONDUCTION

2.5.1

Heat Diffusion Equation

In two dimensional conduction temperature distribution is characterized by two spatial

coordinates T(x,y). Heat flux vector is characterized by two directional components,
qx and qy.

Figure 2-8Isotherms of two dimensional conduction

General form of heat diffusion equation in Cartesian coordinates was

T
x

T
y

T
z

cp

T
t

Assuming steady-state, two-dimensional conduction in a rectangular domain with

constant thermal conductivity and heat generation, the heat equation is

d 2T
dx 2

d 2T
dy 2

q
k

(2-48)

Differential heat equation can be treated with different solution methods

1) Exact/Analytical: Separation of Variables
Limited to simple geometries and boundary conditions
Conduction shape factor is based on analytical solutions
2) Approximate/Graphical: Flux Plotting
Havent been considered here
limited value for quantitative considerations but a quick aid to
establishing physical insights

25

3) Approximate/Numerical: Finite-Difference/Finite Volume, Finite Element or

Boundary Element Method.
Havent been considered here
Most useful approach and adaptable to any level of complexity

2.5.2

Conduction Shape Factor

Two- or three-dimensional heat transfer in a medium bounded by two isothermal

surfaces at T1 and T2 may be represented in terms of a conduction shape factor S. Heat
transfer rate between these isotherms can be described with conduction resistance or
with shape factor
=

(2-49)

thus, for conduction shape factor following relation can be written

=

(2-50)

Following tables give shape factors for two and three dimensional isotherm cases.
Table 2-3 Conduction Shape factors in three Coordinates

26

Table 2-4 Two and three dimensional conduction shape factors

27

28

Example III: Oil pipe under soil Shape factor

A long cylindrical pipe with diameter of
= 0,5 is placed 1,5 m under the ground surface. The pipe is
covered a cellular glass insulation with thickness of 0,1 m and is filled with oil with temperature of 120
C. The temperature of the ground surface is 0 C. Calculate the heat loss per unit length of the pipe.

ASSUMPTIONS: (1) Temperature of oil is uniform at any instant, (2) Radiation exchange with the
surroundings is negligible, (3) Constant properties
PROPERTIES: Table A.3, Soil (T = 300 K): k = 0,52 W/mK; Table A.3, Soil (T = 365 K): k = 0,069
W/mK;
ANALYSIS:
Conduction circuit:

The conduction resistances:

=
=

ln(
1

ln(0,7 m 0,5 m)
0,776
=
mK/W
0,069 W/mK

cosh (

cosh (

The heat transfer rate per unit length:

=

= 84 W/m

(120

0)C

(0,776 + 0,653) W/mK

29

1,5 m/0,7 m) 0,653

=
mK/W
0,52 W/mK

= 84 W/m

2.6 LEARNING OUTCOMES

Chapter 2.4 consisted from following main concepts of one-dimensional, steady-state
conduction (Table 2-5). Additionally very basic approach to 2-D conduction was
presented as charts of shape factor.
Table 2-5 Learning Outcomes: One-dimensional, steady-state conduction

Level of
Knowledge

Concept
1D Conduction

Apply,
Understand,
describe

Fouriers Law
Thermal conductivity temperature dependency
Heat equation solutions in different coordinates (Summarized in
Table 2-6)

Circuit Analogy & heat resistances

1D conduction with internal heat sources
2D Conduction
Apply,
Understand,
describe

Shape factor

Table 2-6 summarizes general heat equation solutions for one-dimensional, steady-state
conduction heat transfer in different coordinates.1

Table 2-6 1-dimensional, steady-state solutions for Heat Equation

Note that with internal heat source heat equation has different boundary condition dependent solutions.

30

FINS

Target of most frequent application of extended surfaces, fin, is to increase surface to

enhance convection heat transfer between solid and fluid. Heat transfer from fin
involves with combined conduction-convection effects. According to Newtons law of
cooling convection heat transfer rate is
q

hA TS

(3-1)

h
A

TS

= Convection heat transfer coefficient [W/Km2]

= Area perpendicular to direction of heat transfer [m2]
= Surface temperature [K]

= Fluid temperature [K]

Generally, methods for increasing convective heat transfer include

Increasing of temperature difference (Ts-T )
Enhancing convection heat transfer coefficient h by increasing fluid velocity
Increase of heat transfer surface area A

Fins

In practice, diversity of fin designs are only limited by imagination, cost of

manufacturing and conditions involved in heat transfer application. Figure 3-1 shows
designs of internally and externally finned tubes, and strip plate fins in plate-fin heat
exchanger.

Figure 3-1 (a) Types of internally and externally finned tubes, (b) rectangular offset strip plate fins
(Bejan)

31

3.1 FIN THEORY

To obtain general energy equation for extended surfaces energy balance over control
volume has to be written. For steady-steady state and zero heat generation conditions
with approximation of one-dimensional conduction in fin energy balance (Figure 3-2)
can be expressed as
qx

qx

qx
dx

dx

qx

dq conv
qx
dx
x

qx
dx
x

(3-2)

dq conv

(3-3)

Figure 3-2 Energy balance for control volume in Fin

For a general case cross-sectional area is dependent on distance x from base of the fin
as: Ac = Ac(x). Conduction heat transfer rate at distance x from base of the fin is
qx

kAc

T
x

(3-4)

Convection from fin surface occurs within fin surface temperature T(x) = Ts(x) and
convective heat transfer rate from fin surface can be written as
dqconv

hdAs (Ts T )
dAs

(3-5)
surface area of fin control volume

By substituting heat transfer rate equations to energy equation following general form
of 1-dimensional conduction-convection in fin is achieved

d
dT
Ac
dx
dx

h
T T
k

dAs
dx

32

(3-6)

d 2T
dx 2

1 dAc dT
Ac dx dx

h dAs
T T
kAc dx

(3-7)

3.2 FINS WITH UNIFORM CROSS-SECTION

In case of uniform fin cross-sectional area (

dAc
dx

0 ), general fin heat equation can be

written as

d 2T
dx 2

hP
T
KAc

(3-8)

= Perimeter of Fin [m]

Temperature difference of fin and boundary, excess temperature, can written with
(T

T )

(3-9)

And resulting a simplified form of heat equation as

d 2
dx 2

m2

(3-10)

hP
kAc

m2

General solution for linear, homogenous 2nd order differential heat equation is
C1e mx

C2 e

mx

(3-11)

Boundary Conditions
at the base of fin (x=0):
0

Tb T

C1e m 0 C 2 e

m0

C1 C 2
at fin tip (x=L): four typical boundary conditions are
b

A) Convection heat transfer:

h ( L)

B) Adiabatic no heat transfer:

d / dx x

C) Defined temperature:
D) Infinite Fin:

kd / dx x

( L)
(L

),

(L) = 0

Temperature profile and heat transfer rate solutions are presented in (Table 3-1).

33

Table 3-1 Temperature distributions and fin heat transfer rate for uniform cross-section fins

3.3 PERFORMANCE
Evaluation of enhance of heat transfer with fins due increased heat transfer area and
parallel additional conduction resistance can be done with fin effectiveness. Fin
effectiveness is defined as ratio of the fin heat transfer rate to the heat rate that would
exist without the fin and is expressed as
qf
f

hAc ,b

Ac,b

(3-12)

= Area of fin at base

= Temperature difference of fin base and boundary

With approximation of equivalent convection coefficient of finned and smooth nonfinned base surfaces, estimation fin effectiveness for infinite fin (tip condition D) can
be written as

hPkAc
f

hA c

Pk
hAc

(3-13)

From fin effectiveness approximation for infinite fin some heat transfer enhancement
conclusions can be drawn
Selection of high conductivity material increases fin heat transfer (aluminium,
copper, etc.)
Increase of ratio of perimeter and cross-sectional area by using thin closely
spaced fins enhances heat transfer
Fins are more effective in case of low convection coefficient meaning gas rather
than liquid and in case of free convection.
Generally, the use of fins is not justified unless effectiveness

34

2.

Fin resistance, describing also fin performance, is defined by treating base and fluid
temperature difference as driving potential for heat transfer as follows
b

Rt , f

Tb

qf
qf
Thermal resistance for non-finned surface can be written as
Rt .b

1
hAc ,b

qb

(3-14)

(3-15)

In consequence fin effectiveness can be expressed as ratio of thermal resistance as

Rt ,b
f

Rt , f

(3-16)

Fin efficiency is defined as ratio of real heat transfer from fin and heat transfer from fin
having surface entirely at base temperature.
qf
f

q max

qf
hA f

(3-17)
b

Figure 3-3 Fin efficiencies of common fin shapes

35

Overall Surface Efficiency characterizes thermal performance of combination of an

array of fins and the attachment base surface and is defined as

qtot
q max

q tot
hAtot

(3-18)
b

Total area of finned surface is calculated as

Atot

NA f

Ab

(3-19)

N
Af
Ab

Number of fins
Surface area of a single fin
Surface area of a base

Total heat transfer rate from finned surface, fins and non-finned base surface, by
assuming equivalent convection coefficients for finned and non-finned surfaces can be
expressed as
qtot

NAf
Atot

Af h

Ab h

(3-20)

(3-21)

Example IV: Rod penetrating to Furnace Fin theory

A rod of diameter D = 25 mm and thermal conductivity k = 60 W/mK protrudes normally from a furnace
wall that is at Tw = 200 C and is covered by insulation of thickness Lins = 200 mm. The rod is welded to
the furnace wall and is used as a hanger for supporting instrumentation cables. To avoid damaging the
cables, the temperature of the rod at its exposed surface, T0, must be maintained below a specified
operating limit of Tmax = 100 C. The ambient air temperature is T = 25 C, and the convection
coefficient is h = 15 W/m2K
a)

Derive an expression for the exposed surface temperature T0 as a function of the prescribed
thermal and geometrical parameters. The rod has an exposed length L0 and its tip is well
insulated.
b) Will a rod with L0 = 200 mm meet the specified operating limit? If not, what design parameters
would you change? Consider another material, increasing the thickness of the insulation, and
increasing the rod length. Also, consider how you might attach the base of the rod to the furnace
wall as a means to reduce T0.
ASSUMPTIONS: (1) Steady-state
conditions,
(2)
One-dimensional
conduction in rod, (3) Negligible
thermal contact resistance between the
rod and hot hot furnace wall, (4)
Insulated section of rod, Lins,
experiences no lateral heat losses, (5)
Convection coefficient uniform over
the exposed portion of the rod, L0, (6)
Adiabatic tip condition for the rod and
(7) Negligible radiation exchange
between rod and its surroundings.

36

ANALYSIS: (a) The rod can be modeled as a thermal network comprised of two resistances in series and
behaving as a fin with an adiabatic tip condition, Rfin,
=

, (table 3-1)

From the thermal network,

=

=
+

(b) Substituting numerical values

(0,025 m)
= 4,909 10
4

0,025 m = 7,854 10

= (15 W/m

= 15 W/m

7,854 10

7,854 10

= 0,0347 W

=
=

0,200 m
60 W/mK 4,909 10
0,0347 W

= 25 C +

m ) (60 W/mK 4,909 10

60 W/mK 4,909 10

= 6,790 K/W

sinh(6,324 m

6,298
(200
6,790 + 6,298

m ) = 6,324 m

0,2 m)

= 6,298 K/W

25) C

= 109 C >

Design changes to reach the correct value for T0:

1)
2)
3)
4)
5)
6)

Decreasing Ac by using a tube instead of a rod.

Decreasing the thermal conductivity by changing rods material (stainless steel, for example).
Introducing contact resistance between the wall and the rod by tack welding
Increasing the length of fin
Increasing the insulation thickness
Increasing the convection coefficient with air handling unit.

37

3.4 LEARNING OUTCOMES

Besides understanding principles of fin theory three important concepts where
described in Fins Chapter as stated in Table 3-2.
Table 3-2 Learning Outcomes: Fins

Level of
Knowledge

Concept

Understand,
Describe

1-dimensional conduction-convection characteristics of fin

and related formulation of fin equation

Apply,
Understand,
describe

Fin equation solutions for constant diameter fins:

temperatures and heat transfer rates
Single fin performance: effectiveness, resistance and
efficiency
Finned surface: efficiency and heat transfer rates

38

4 UNSTEADY-STATE CONDUCTION
Unsteady, or transient, heat transfer phenomenon arises commonly with changed
boundary conditions. This chapter introduces solution methods for time dependent
conduction problems.

4.1

THE LUMPED CAPACITANCE METHOD

For transient heat transfer processes, where temperature gradients in solid are
negligible, a simple approached called Lumped Capacitance Method can be utilized.
Let us consider a quenching of a hot metal forging that is initially at uniform
temperature Ti and is cooled by immersing it into a pool containing liquid of lower
temperature T . Time dependent cooling of metal forging occurs due convection heat
transfer and decreasing temperature difference of metal and liquid. In Lumped
capacitance method spatially uniform temperature profile change of forging is assumed,
which practically acceptable for solids having high thermal conductivity compared to
convection or overall heat transfer coefficient between solid and boundary.

Figure 4-1 Cooling of a hot metal forging

With assumption of non-existing temperature gradient, Fouriers law of heat conduction

is not available for solution. Energy equation over metal control volume can be
expressed as
E st

qconv

(4-1)

Vc p

dT
dt

hAs ( T

T ).

(4-2)

By notating temperature difference as

temperature T
temperature gradient

= (T-T ) with assumption of constant boundary

dT
dt

39

d
dt

(4-3)

Vc p d
.
(4-4)
hAs dt
Integral solution with initial conditions at start time t = 0 and initial temperature T(0)
=Ti for energy equation is
energy equation

Vc p
hAs

ln

T
Ti

(4-5)

T
T

hAs
t
Vcp

(4-6)

Thermal Time Constant: By defining thermal time constant as follows

1
( Vc p )
As h

Rt
Ct

Rt C t

(4-7)

Convection heat transfer resistance

Lumped thermal capacitance

Time dependent solution for temperature can be expressed as

T
Ti

T
T

t/ t

(4-8)

Time dependent temperature behaviour of metal is characterized by exponentially

decaying temperature difference between solid and liquid. Increase in either convection
heat transfer resistance Rt or lumped thermal capacitance Ct will decrease temperature
response of solids. (Figure 4-2)

Figure 4-2 Transient temperature response of lumped capacitance solids with different thermal time
constants.

40

Total heat transferred till time t can be solved by integrating convection heat transfer
rate over time with substitution of time dependent temperature solution as follows
t

qdt

hAs

dt

( Vc p )

1 exp(

) .

(4-9)

4.1.1 Validity Criteria

Suitable and convenient validity criteria for use of simplest transient heat transfer
problem solving method refer to assumption of uniform temperature profile of solid.
Considering a steady-state conduction in plane wall (Figure 4-3) with constant
temperature Ts,1 in other wall and other cooled by convection with fluid of temperature
T < Ts,1, surface energy balance can be written as
kA
(T
L s,1

Ts ,2 )

hA(Ts ,2

T )

(4-10)

By restating energy equation, Biot number can be written

Bi

hL
k

Ts,1 Ts,2
Ts,2

( L / kA)
(1 / hA)

Rcond
.
Rconv

(4-11)

Biot number is a dimensionless parameter describing the ratio of conduction in solid

and convection resistances, and the ratio of temperature drop in solids to temperature
difference with solid surface and liquid boundary. Figure 4-3 shows temperature
profiles of solid plane wall with different Biot numbers and it can noted that with
decreasing Biot number uniform temperature profile is approached.

Figure 4-3 Steady-state temperature distributions of plane wall with different Biot numbers

Generally, Biot number Bi << 1 is considered as criteria for uniform temperature

profile in solid during transient heat transfer. This condition can be interpreted as The
41

conduction resistance within solid is much less compared to convection resistance

across fluid boundary layer. Figure 4-4 presents strong effect of Biot number to solid
temperature profile behavior with in a plane wall cooled symmetrically with convection
on both surfaces.

Figure 4-4 Transient temperature distribution for different Bi Numbers in a plain wall

Characteristic length is defined as ratio of solid volume to surface area and for the use
of thermal calculations of complicated shapes

LC

V
AS

(4-12)

Fourier number, known also as a dimensionless time, characterizes together with Biot
number transient conduction phenomenon. Fourier number is defined as

Fo

t
2
C

(4-13)

With Fourier and Biot numbers time dependent temperature equation can formulated as
follows

Bi Fo

hLC t
k L2C

T T
Ti T

hAS t
Vc
hAs
t
Vc p

(4-14)

Bi Fo

42

(4-15)

Example V: A Thermocouple junction Lumped Capacitance Method

A thermocouple junction, which may be approximated as sphere, is to be used for temperature
measurement in a gas stream. The convection coefficient between the junction surface and the gas is h =
400 W/m2K and the junction thermophysical properties are k = 20 W/mK, c = 400 J/kgK and = 8500
kg/m3. Determine the junction diameter needed for the thermocouple to have a time constant of 1 s. If the
junction is at 25 C and is placed in a gas stream that is at 200 C, how long will it take for the junction
to reach 199 C.
ASSUMPTIONS:
1. Temperature of junction is
uniform at any instant
2.

Radiation exchange with the

surroundings is negligible

3.

Losses by conduction through

the leads are negligible

4.

Constant properties

ANALYSIS:
1. From equation 4-7 and the fact that
/6

for a sphere, it follows that

1

=
=

400 W/m
1s
= 7,06 10
kg
8500
400J/kgK
m

With Lc = r0/3 it then follows from equation 4-11 that

3)

400 W/m

3,53 10
20 W/mK

= 2,35 10

Lumped capacitance method may be used to an excellent approximation.

2. From equation 4-5 the time required for the junction to reach T = 199 C

=
=

8500

= 5,2 s

/6)
ln
)

ln

kg
7,06 10 m 400J/kgK (25 200)K
m
ln
(199 200)K
400 W/m K

43

4.2 1D TRANSIENT CONDUCTION WITH SPATIAL EFFECTS

Alternative approaches for lumped capacitance method have to be used with conduction
problems having not negligible temperature profiles of solids. Analytical solutions can
be derived for simple geometries from fundamental heat transfer equations and as a
basis general heat diffusion equation can be written in one-dimensional form with no
heat generation as
2

dT
dt

T
x2

(4-16)

4.2.1 Dimensionless Governing Equations

By forming dimensionless forms of numbers and variables a total number of dependent
variables in governing equations can be reduced. For the use in transient heat transfer
problems a dimensionless temperature difference is defined as

T T
Ti T

,0

T ( x, t )
Ti (x,0)

(4-17)

Transient temperature profile

Initial temperature profile

Similarly a dimensionless spatial coordinate for plane wall can be defined as

x
L

, 1
L

(4-18)

Half thickness of the wall

Temperature, spatial and time derivates can be written in dimensionless form as

2

Ti

L x

, T

Ti

T
(4-19)

Fo

Dimensionless heat equation can then be written as

Ti T 2
L2 x*2

Ti T

L2

(4-20)

Fo

(4-21)

Fo

44

4.2.2 Plain Wall

Let us consider a plain wall having uniform temperature profile, which is then suddenly
exposed to convection conditions.

Figure 4-5 Plain wall under convection conditions

Initial Conditions

x , Fo

Dimensionless temperature difference in the

beginning: wall temperature profile is initial
temperature profile.

Boundary Conditions

0
x

Heat transfer rate at midline of symmetrical

wall is zero

Bi
x

1, Fo

Conduction heat transfer rate is equal to

convection at wall.
Solution of dimensionless transient heat equation can now be expressed with 3
f x , Fo, Bi , when ordinary form would require
dimensionless variables as
representation as T

x, t , Ti , T , L, k , , h .

Exact solution of dimensionless heat equation is

Cn e

2
n Fo

* cos

n 1

Coefficient : C n

Eigenvalue :

4 sin n
2 n sin 2
tan

(4-22)
n

Bi

45

Approximate Solution
Infinite series solution can be approximated with first term, if Fourier number
C1e

2
1 Fo

* cos

4 sin 1
2 1 sin 2

Coefficient : C1

0.2:

Eigenvalue, 0

(4-23)
1

tan

Bi

Transient temperature profile can be solved from by formatting solution for temperature
at midline of wall (x*=0) as

T0 T
Ti T

cos

2
1 Fo

C1e

(4-24)

(cos0=1)

Heat transfer rate from other surface to boundary can be written based on energy
balance as
Q

E st

c p T x, t dV

Vc p Ti
v

(4-25)

Vc p Ti

C1
e
L

2
1 Fo

sin

Vc p Ti

sin

Maximum heat energy that could be transferred is

Q0

Vcp Ti T

(4-26)

And relationship of energy transfer can be comprised as

Q
Q0

sin

Q0 .

(4-27)

4.2.3 Cylinder
General heat diffusion equation for cylindrical coordinates one-dimensional conditions
with no heat generation are written as

46

T
1
r
r
r r

1 T
t

(4-28)

Figure 4-6 Infinite Cylinder or Sphere

One-dimensional conduction, responding infinite length, in case of cylinder is

reasonable approximation, if length to radius ratio L/r0 10.
Initial Conditions: Uniform initial temperature: t = 0, T(r,t=0)=Ti
Boundary Conditions: At surface conduction rate equals to convection heat transfer rate
and at midpoint conduction rate is zero
T
r r0 ,
k
h T r r0 , t T
r r r0
r

0,

T
r

0
r 0

Exact solution of dimensionless heat equation

T T
Ti T

Cn e

Jo

n 1

Cn

2
n Fo

J o2

(4-29)

J1

J 12

Quantities of Bessel functions of first kind: J0 and J1 are presented at

roots of

J1
Jo

Bi

are positive

(4-30)

And values are shown in Table 4-1 as function of Biot number. Dimensionless radius,
heat transfer coefficient and time are defined as

r
,
ro

Bi

hro
,
k

Fo

t
2
o

47

(4-31)

Approximate Solution
As in case of plain wall, infinite series solution can be approximated with first term, if
Fourier number 0.2. Relationship of transferred and maximum energy transfer can be
expressed as
2 0
Q
(4-32)
1
J1 1
Q0
1
, where temperature of axis (r =0, Jo(0)=1 ) is calculated from
2
1 Fo

C1e

(4-33)

4.2.4 Sphere
General heat diffusion equation for spherical coordinates under steady-state, onedimensional conditions with no heat generation are written as

1
T
r2
2
r
r r

1 T
t

(4-34)

Initial and boundary conditions are same than in case of cylinder:

Initial Conditions: Uniform initial temperature: t = 0, T(r,t=0)=Ti
Boundary Conditions: At surface conduction rate equals to convection heat transfer rate
and at midpoint conduction rate is zero
T
r r0 ,
k
h T r r0 , t T
r r r0
r

0,

T
r

0
r 0

Exact solution of dimensionless heat equation

2
T T
1
Cn e n Fo
sin
Ti T
n 1
nr

r
(4-35)

4 sin n
n cos
2 n sin 2 n

Cn

Quantities of Bessel functions of first kind: J0 and J1 (Table 4-1),

of
1

cot

Bi

are positive roots

(4-36)

And values are shown in Table 4-1 as function of Biot number. Dimensionless radius,
heat transfer coefficient and time are defined as

48

hro
,Fo
k

r
, Bi
ro

(4-37)

2
o

Approximate Solution
As in case of plain wall, infinite series solution can be approximated with first term, if
Fourier number 0.2. Relationship of transferred and maximum energy transfer can be
expressed as

Q
Q0

0
3
1

sin

cos

(4-38)

, where temperature of centre ( r =0, Jo(0)=1 ) is calculated similarly to cylinder as

0

C1e

2
1 Fo

(4-39)

49

Table 4-1 (a) One term approximation coefficients to series solution for transient 1D conduction and (b)
Bessel Functions of the First Kind

50

4.2.5 Semi-Infinite Solid

By definition, Semi-infinite solid means a solid that extends to infinity in all but one
direction. It is a practical approximation for heat transfer problems. General heat
diffusion equation for semi-infinite solid can be written under steady-state, onedimensional conditions with no heat generation as
2

1 T
t

(4-40)

Initial Conditions: Uniform initial temperature

T(x,t=0)=Ti

Boundary Conditions: semi-infinity approximation: temperature at infinite distance

from surface approaches initial uniform temperature
T(x
, t) =Ti

Figure 4-7 Transient temperature profiles in semi-infinite solids with 3 surface conditions: constant
temperature and heat flux, and surface convection.

Analytical solutions of heat equation can be derived for three surface condition cases
presented at Figure 4-7:

51

Case 1: Constant surface Temperature T(x=0, t) = Ts

T ( x, t ) Ts
Ti Ts
q s (t )

k (Ts

erf

(4-41)

Ti )

(4-42)

Case 2: Constant surface heat flux

T ( x , t ) Ti

2q 0

exp

Case 3: Surface Convection

T ( x , t ) Ti
T
Ti

erfc

x
2

T
x

q0
x 0

x2
4 t

q0 x
x
erfc
k
2 t

T
x

hT x 0

(4-43)

x 0

hx
exp
k

h2 t
k

erfc

x
2

h
t

t
k

(4-44)

Definition and values of Gaussian error function, erf, and complementary error function
is presented at Table 4-2.

52

Table 4-2 Gaussian Error functions

53

4.3 LEARNING OUTCOMES

Table 4-3 Learning Outcomes: 1-dimensional transient heat transfer

Level of
Knowledge

Concept
Lumped Capacitance

Describe
Understand

Convection Heat storage lumped capacitance energy

equation

Apply
describe
understand

Lumped capacitance Time & temperature solutions + validity

criteria
Concepts: time constant, Bi, Fo, characteristic length
Spatial Effects 2

Describe
Understand

Form of heat equation

Concepts: dimensionless temperature difference

Apply
describe
understand

One term approximate solutions: for different geometries and

boundary conditions

Not currently included to content of course: Fundamentals of Heat Transfer

54

5 FUNDAMENTALS OF CONVECTION
In principle, convection is energy transfer between a surface and a fluid flowing over
the surface. Convection heat transfer is combination of two mechanisms: random
molecular motion (diffusion or conduction) and bulk/macroscopic fluid motion
(advection). There are two major objectives in convection heat transfer analysis
presented here:
Understanding the physical phenomenology associated with boundary layers underlying the
convection heat transfer.
Gain the knowhow to solve convection heat transfer problems with relevant knowledge of
empirical friction and heat transfer coefficients.

5.1 CLASSIFICATION OF CONVECTION HEAT TRANSFER

Convection is classified according the nature of the flow:
Free (Natural) Convection Flow is induced by buoyancy forces, which are
caused by temperature variations formed due to heat transfer in the fluid.
Forced Convection Flow is caused by external means: fan, pump, wind, etc.
o Convection is classified also as 1) external or 2) internal, depending on
whether fluid is flowing over a surface or inside a channel.
Two special convection heat transfer cases associated with phase change
between liquid and vapour states of fluid are boiling and condensation. Both
processes may occur under either free or forced convection conditions.

Figure 5-1

Convection heat transfer classification: forced, free, boiling and condensation

55

5.2 PROBLEM OF CONVECTION

Problem of convection is seen as determination of local or/and average convection
coefficients. Local and total heat transfer rates are solved from rate equations, which
are dependent on knowledge of local convection coefficients h or average convection
coefficients h .
Newtons Law of Cooling
Convection heat transfer rate is expressed commonly as Newtons law of cooling:
q

hA TS

(5-1)

TS

= Convection heat transfer coefficient [W/Km2]

= Area perpendicular to direction of heat transfer [m2]
= Surface temperature [K]

= Fluid temperature [K]

h
A

5.2.1 Variables effecting to Convection

Convection heat transfer rate and convection heat transfer coefficient are affected by
fluid properties: density, thermal conductivity, dynamic viscosity, specific heat
geometry and roughness of surface
Mode of fluid flow: laminar or turbulent
Mode of convection: free, forced, boiling or condensation

5.2.2 Convection heat transfer coefficient

Convection heat transfer coefficient is experimentally determined parameter and tables
below show typical ranges for convection heat transfer coefficient.
Table 5-1 Ranges of convection coefficients

Type of Convection
Free convection of gases
Free convection of liquids
Forced convection of gases
Forced convection of liquids
Boiling & condensation

h [W/Km2]
2-25
10-1000
25-250
50-20 000
2500-100 000

56

Table 5-2 Convection heat transfer coefficients (Lienhard)

Heat Transfer
Based on Newtons Law of cooling convection heat transfer rate can be formulated
with local convection heat transfer coefficient h as follows
q ,,

h Ts

T
(5-2)

h Ts

T dAs

As

Ts

hdAs
As

Average convection coefficient is integral of local value over total surface area and with
this total heat transfer rate based on average convection heat transfer coefficient can be
then expressed as

1
hdAs
As A
s

(5-3)

q hAs Ts T

(5-4)

In special case of flat plate, local convection coefficient varies only with distance x and
thus average convection coefficient can be formulated as

1L
hdx
L0

(5-5)

57

Figure 5-2 Local and total convection heat transfer. (a) Wing surface (b) Flat plate

Mass Transfer
Convection mass transfer from surface is analogical to convection heat transfer. Total
convection mass transfer rate can be expressed with average convection heat
NA

h m As C A, s

C A,

CA

mol

(5-6)

Concentration of species A [ mol

m3

, where average convection mass transfer coefficient is defined as

hm

1
As

hm dAs

(5-7)

As

For flat plate, average convection mass transfer coefficient can be written analytically
to convection heat transfer coefficient as
L

hm

1
hm dx
L0

(5-8)

In special condition for ideal gases with saturation vapour pressure at surface, surface
concentration can be written as

C A, s

p A, sat Ts

(5-9)

RTs
pA,sat(Ts)

Saturation vapour pressure at surface temperature Ts

58

5.3 LEARNING OUTCOMES

Table 5-3 Learning Outcomes: Fundamentals of Convection

Level of
Knowledge
Understand
Describe

Concept
Convection modes
Newtons Law of Convection
Convection heat transfer coefficient: Range, Local or average? How
its formed?

Apply
describe
understand

Newtons Law of Convection

59

6 BOUNDARY LAYER THEORY

Formation and behaviour of boundary layers are fundamental phenomenon describing
convection heat and mass transfer characteristics in fluid motion over solid surface. The
fluid motion, flow, over a surface is divided to two different regions:
Boundary Layer, a thin fluid layer near surface, having large velocity, temperature or/and
concentration gradients to opposite direction of flow and high shear stresses.
Region outside of boundary layer having small shear stresses, thermal and mass transfer
gradients.

6.1 BOUNDARY LAYERS

6.1.1 Velocity Boundary Layer
Velocity Boundary Layer is region of the fluid, where velocity varies from zero at
surface to velocity 0.99 u of bulk fluid flow and boundary layer thickness can be
written as

y u 0,99u

(6-1)

Velocity boundary layer and surface friction exists always in a flow over any surface.
In a flow over a flat plate (Figure 1-5) boundary layer velocity profile changes with
flow moving further on a plate. Inside the velocity boundary layer shear stress forces
are large and form velocity gradients opposite to surface direction. A dimensionless
measure of ratio of shear stress and velocity used in fluid mechanics is a local friction
coefficient

Cf

s
2

(6-2)

u /2

Figure 6-1 Velocity boundary layer in a flow over a flat plate

Shear stress for Newtonian fluids can be determined with velocity gradient as
u
s
yy 0
Dynamic viscosity

60

(6-3)

6.1.2 Temperature Boundary Layer

Temperature Boundary Layer is region of the fluid, where temperature varies from
surface temperature to bulk fluid temperature. At the interface (surface) temperatures of
solid and fluid are same and velocity of fluid is zero and heat is transferred only by
random molecular motion (conduction). Temperature boundary layer is characterized
by temperature gradients and heat transfer.

Figure 6-2 Temperature boundary layer in a flow over a flat plate

Development of temperature boundary layer is analogical to formation of velocity

boundary layer. Thickness of temperature boundary layer is defined as

Ts T
Ts T

0,99

(6-4)

Within boundary layer in a case of flat plate temperature is function of both distances
from the tip and height from plate: T = T(y,x). Out of boundary layer temperature is
bulk fluid temperature T T .
Local surface heat flux at any distance from tip can be expressed according to Fouriers
Law with temperature gradient and conductivity as
q "s

kf

T
y y
0

(6-5)

Definition for convection heat flux is

q"

h Ts T

(6-6)
"

At surface convection heat transfer is only conduction, ( q s

convection heat transfer coefficient is

T
y y 0
Ts T

q " ), and thus a local

kf
h

(6-7)

61

Thickness of temperature boundary layer is growing similarly with developing velocity

boundary layer and consequently temperature gradients are decreasing.

6.1.3 Concentration Boundary Layer

Concentration Boundary Layer is similar to velocity and temperature boundary layers.
It is characterized by concentration gradients and convection mass transfer rates.
Thickness of concentration boundary layer is defined as

C A, s C A y

0,99

C A, s C A,

(6-8)

Figure 6-3 Concentration boundary layer in a flow over a flat plate

Molar flux by diffusion from surface any distance from tip of flat plate (Figure 6-3) can
be written with Ficks Law as
N "A, s

D A, B

CA
y

(6-9)
y 0

DA, B

Diffusion coefficient

On the other hand, convective mass transfer is defined as

N "A, s

hm C A, s

C A,

(6-10)

Convection mass transfer coefficient can thus be described analogically to heat

transfer coefficient as

CA
Y y 0
C A, s C A,

D A, B
hm

(6-11)

62

6.1.4 Laminar and Turbulent Flows

Both surface friction and convection transfer rates depends strongly on flow condition
and primarily weather the boundary layer is turbulent or laminar. Table 6-1 describes
basic differences of laminar and turbulent flow conditions.
Table 6-1 Laminar versus Turbulent flow

Laminar

Turbulent

Streamlined and highly ordered

Irregular and random 3D motion of

parcels of fluid
Highly fluctuating velocity
components
Chaotic fluctuation conditions result
as higher order transfer of
momentum, temperature and
concentration

At steady-state conditions no time

dependent fluctuation of velocity
Momentum, temperature and
concentration are diffusing
through the boundary layer

Figure 6-4 shows boundary layer development in a flow on flat plat. Thickness of
velocity boundary layer grows continuously with decreasing velocity gradients. Flow
conditions are streamlined in laminar region until transition zone is reached.
Conditions in transition zone fluctuate between laminar and turbulent flow. Turbulent
boundary layer can be divided to
1) viscous sub-layer: diffusion and linear velocity profile,
2) buffer layer: diffusion and turbulent mixing, and
3) turbulent zone: turbulent mixing dominating transport phenomenon.

Figure 6-4 Development of velocity boundary layer on flat plate

Dimensionless grouping of parameters describing ratio of inertia to viscous forces is

Reynolds Number and is defined for plate as
Re

u x

(6-12)

63

The condition for change from laminar to turbulent flow in case of flat plate is
determined by critical Reynolds number Re c 5 *105 , or 105 Re c 3 * 106
depending on surface properties.

Figure 6-5 Velocity boundary layer thickness

and a local heat transfer coefficient for flow
over a flat plate

Figure 6.5 shows velocity boundary layer development with transition from laminar to
turbulent regions in flow over flat plate. Local heat transfer coefficient is highly
dependent on flow condition with its velocity gradients and transport properties.
Thickening of boundary layer can be seen as decaying heat transfer at both laminar and
turbulent regions, while a dramatic increase of heat transfer coefficient occurs in
transition zone caused by enhanced turbulent transport conditions. Effect of turbulent
conditions is similar in case of convection mass transfer from surface.

Example VI: Air flow over plate Boundary conditions

Consider airflow over a flat plate of length L = 1 m under conditions for which transition occurs at xc =
0,5 m based on the critical Reynolds number, Rex,c = 5 105.
a) Evaluating the thermophysical properties of air at 350 K, determine the air velocity.,
b) In the laminar and turbulent regions, the local convection coefficients are, respectively,
( )
( )
and
Where, at T = 350 C,
= 8,845 W/m
K, and
= 49,75 W/m
K. Develop an
expression for the average convection coefficient,
), as a function of distance from the
leading edge, , for the laminar region,
.
c) Develop an expression for the average convection coefficient
), as a function of distance
from the leading edge, x, for the turbulent region,
.
d) On the same coordinates, plot the local and acerage convection coefficients,
respectively, as a function of for
.

64

, and

ASSUMPTIONS:
(1) Steady-state conditions,
(2) Constant properties

PROPERTIES:
Table A.4, air (T = 350 K):
k = 0,030 W/mK,
= 20,92 10-6 m2/s,
Pr = 0,700.

ANALYSIS:
a) At 350 K with
Re

=
=

= 0,5 m,

Re

10

20,92 10
0,5 m

m /s

= 20,9 m/s

b) From Eq. 5-3, the average convection coefficient in the laminar region

c)

=2

( )=

( )

( )

=2

(1)

The average convection coefficient in the turbulent region

( )=

+ 1,25

( )
(

)]

( )

.
=

(2)

d) The local and average coefficients, Eqs. (1) and (2) are plotted below as a function of x for the
range
.

65

6.2 BOUNDARY LAYER EQUATIONS

Fluid motion at boundary layer acts according to fundamental laws of nature:
conservation of mass, energy and chemical species as well as momentum, Newtons 2nd
law of motion, are valid for fluid flow characterization.

Figure 6-6 Development of temperature, concentration and velocity boundary layers for curved surface

Here conservation equations are presented for two-dimensional arbitrary surface with
following approximations:
Laminar flow & Steady-state conditions
Constant fluid properties
gradients to direction of Flow:
2

1) Velocity

u
x2

u
y2

T
x2

2) temperature:

p
x

3) pressure:

T
y2
p
x

Conservation of Mass
Overall Continuity Equation can be derived from writing mass balance for differential
two-dimensional control volume as
u
x

v
y

(6-13)

For incompressible flow, meaning constant density, continuity equation reduces to

66

u
x

v
y

(6-14)

Newtons Second Law of Motion

Momentum equation to x-coordinate direction can be reduced with assumptions made
to

u
x

u
y

1 p
x

fluid motion x -momentum

leaving control volume

(6-15)

y2

net pressure force

viscous shear stresses

Conservation of Energy
Energy equation for boundary layer control volume may be expressed with conditions
earlier presented as
2

T
v
y

T
u
x

net rate of thermal energy

leaving control system
with fluid motion

T
y2

v
cp

u
y

(6-16)

net rates of y-direction conduction

viscous dissipatio n

6.3 SIMILARITY
Similarity of conservation equations presented, applicable for low-speed, forced
convection flows, can be further developed by forming dimensionless parameters and
governing equations. Independent dimensionless spatial coordinates are defined as
x*

x
L

y*

y
L

; L Characteristic length of surface

Dimensionless dependent velocity components are defined as

u*

u
V

v*

v
V

; V Bulk flow velocity

Dimensionless temperature and concentration differences can be expressed as

T*

T TS
T TS

CA

CA
CA

CAS
C AS

6.3.1 Dimensionless Groups and Equations

Table 6-2 shows dimensionless forms of conservation equations with three important
dimensionless similarity parameters:
Reynolds number:

VL
VL
v
The ratio of inertia to viscous forces
Re L

67

Prandtl number:

Pr

The relative effectiveness of momentum and energy transport by

diffusion in the velocity and thermal boundary layers
Schmidt number:

v
D AB
The relative effectiveness of momentum and mass transport by
diffusion in the velocity and concentration boundary layers
Sc

Table 6-2 Dimensionless boundary layer conservation equations and y-direction boundary conditions
(Incropera)

Following dimensionless parameters describing convection fluid dynamic, and

convection heat and mass transfer functional dependences have been derived by
similarity approach and can be used in formation of experimental correlations:

2
f x * , Re L
V / 2 Re L
Dimensionless surface shear stress

Friction Coefficient: C f

S
2

hL
f Re L , Pr
kf
Ratio of convection to conduction heat transfer

Nusselt Number: Nu

If functional relation is known, convection heat transfer coefficient h can be

determined.

hm L
f Re L , Sc
D AB
Dimensionless concentration gradient

Sherwood Number: Sh

If functional relation is known, convection mass transfer coefficient hm can

be determined for described geometry.

68

6.3.2 Boundary layer Thickness

Similarity relationships can be used for determining approximate relationships between
boundary layer thicknesses as follows (in usual applications exponent n can be
approximated with value 1/3):
Ratio of velocity and temperature boundary layer thicknesses as
n

Pr n

(6-17)

Ratio of velocity and concentration boundary layer thicknesses as

n

= Sc n

DAB

(6-18)

Ratio of temperature and concentration boundary layer thicknesses as

n

Sc
= Le
(6-19)
Pr
Lewis number ratio of thermal and mass diffusivities
n

D AB
Le

69

6.4 ANALOGIES
As heat and mass transfer conservation equations are of the same form, analogous, and
related to velocity field with dimensionless parameters: Reynolds, Prandtl and Schmidt
numbers, the boundary layer temperature and concentration profiles are same form with
similar boundary conditions.
Table 6-3 Functional relations relevant to boundary layer analogies (Incropera)

6.4.1 Heat and mass transfer analogy

For particular geometry with similar boundary conditions following heat and mass
transfer analogy can be written
Nu
Pr n

Sh
Sc n

(6-20)

Exponent value most applications

This can be further developed to ratio of heat and mass transfer convection coefficients:

h
hm

k
D A, B Le n

c p Le1

(6-21)

Example VII: Evaporative cooling of beverage bottle Cooling Beer at Hot Beach
ASSUMPTIONS:
(1) Heat and mass transfer analogy is
applicable
(2) Vapor displays ideal gas behaviour,
(3) Radiation effects are negligible,
(4) Air properties may be evaluated at a
mean boundary layer temperature
assumed to be 300 K.

Beer bottle inside a wet towel at hot beach conditions

PROPERTIES: Table A-4: air (300 K): = 1,16 kg/m
= 22,5 10 m /s, = 26,3 10 W/mK

ANALYSIS:
1) Conservation of energy:

70

= 1,007 kJ/KgK

= 14,82 10

m /s,

"
qconv
.

"
qevap

hT

Ts

h fg n "A

2) Evaporative mass flux of water

hm (

nA

A , sat

(Ts )

3) Heat and mass transfer analogy

h
k
1
c p Le
hm D A, B Le n
Where

A,

4) Bottle temperature difference to surroundings from equations above

D Le
h fg AB
k

Ts

5) Ideal gas law:

=

3
A, sat

Ts

A,

By setting
= 0, an iterative equation is formed from eq. (4) and (5). Beginning the iteration with
value = 280 K:

= 313 K

Comments:
1)

280 K

kg
100 KJ/kg 200 kmol

kJ
8,315 kmol K 1,007 kJ/KgK 1,16 kg/m
5000 N/m 10 kJ/Nm
280 K

22,5 10 m /s
20 10 m /s

As such hot conditions, a 7C Beer would be worth of its weight in gold?

6.4.2 Reynolds Analogy

Table 6-3 shows similarity with velocity, thermal and concentration boundary layers
and with similar boundary conditions all the engineering parameters combining
equation can be formed as Reynolds Analogy. Validity of this analogy requires Pr and
Sc 1 and dp*/dx* 0. Modified Reynolds or Chilton-Colburn analogies containing
correction functions have been developed for wider operating range as
Reynolds Analogy

Chilton-Colburn Analogies

Cf

Re L
2

Cf
2
Cf
2

71

Nu

Sh

St Pr 2 / 3

jH

St m Pr 2 / 3

jm

(6-22)

0.6

0.6

Pr

Sc

60

(6-23)

3000 (6-24)

6.5 LEARNING OUTCOMES

Table 6-4 Learning Outcomes: Boundary Layer Theory

Level of
Knowledge
Understand
Describe

Concept

Development of Boundary Layers: Thermal, Velocity and

Concentration

Transition from laminar to turbulent

Laminar Boundary layer Conservation equations: terms and
their meaning
Concepts: Dimensionless Numbers: Re, Pr, Sc; Boundary
layer thickness
Apply
understand
describe

Heat and mass transfer analogy

72

7 FORCED CONVECTION EXTERNAL FLOW

In forced convection fluid is forced to flow over a surface by external means such as
pump, fan or wind. In external flow boundary layers develop freely without disturbance
of adjacent surfaces and at flow region out of boundary layer has negligible velocity,
temperature and concentration gradients.
An engineering problem of forced convection at external flow is generally to find
convection coefficients for described geometries. Here the principles and functional
forms of convection correlations for low-speed forced convection flow are presented.
As discussed in boundary layer analogies chapter, local and average convection heat
transfer coefficients have dimensionless functional form as
Nu x

f ( x * , Re x , Pr)

Nu x

f (Re x , Pr)

And respective functional form for mass transfer coefficients as

Shx

f x * , Re L , Sc

Sh x

f Re x , Sc

There are two options to obtain convection coefficient functional relationships:

Analytical approach appropriate for simple geometries only
Experimental or empirical approach also complex geometries

7.1 EMPIRICAL CORRELATIONS

Experimental approach for determining convection coefficient for geometry, such as
flat plate, could be done by electrically heating plate and simultaneously cooling it with
convective fluid flow. Heating power would result as known convective heat flux, if
insulation would be appropriate. With measurement of surface temperature, and fluid
temperature and heat flux, convection coefficient can be calculated based on Newtons
law of Cooling. From fluid properties and flow velocity descriptive dimensionless
numbers: Reynolds, Prandtl and Nusselt numbers can be defined. As the functional
dependence of Nusselt, Reynolds and Prandtl numbers is key question for determining
general heat transfer coefficient correlation for described geometry, a correlation form
to be fitted with test data may for example be chosen as
Nu x

C Re mx Pr n

(7-1)

By repeating measurements with variation of Nu, Pr and Re by changing for example

convection flow velocity and temperature, characteristic dimensions of geometry and
fluid, correlation parameters for convection coefficient can be solved (Figure 7-1).
Correlation for dimensionless heat transfer coefficient, Nusselt number, obtained is
valid for similar surface geometry and type of flow.

73

Figure 7-1 Dimensionless fitting of convection heat transfer coefficient test data

The effect of varying fluid properties with changing temperature at boundary layer can
be accounted by calculating film temperature
Ts

Tf

(7-2)

Or by adding property correction to convection coefficient correlation as

r

C Re mx Pr n Pr / PrS

Nu x

or
C Re mx Pr n

Nu x

(7-3)

7.2 ANALYTICAL SOLUTION FOR FLAT PLATE

Velocity component are presented with stream function (x, y) as
u

(7-4)

Similarity solution f and similarity variable

f( )

is then defined as
(7-5)

vx / u

y u / vx

(7-6)

7.2.1 Velocity Boundary Layer

By using similarity approach and differentiating velocity components momentum
equation can be written as

74

d3 f
d 3

d2 f
d 2

(7-7)

Solutions for differential equation are presented in Table 7-1. It can noted that
boundary layer thickness (u/u =0.99) corresponds similarity value =5 and thus
boundary layer thickness in laminar flow over flat plate is
5x
5x

u / vx

(7-8)

Re x

Similarly wall shear stress and then local friction coefficient can be written as

C f ,x

u
y
s ,x
2
x

0.332u

u /x

(7-9)

y 0

u /2

0.664 Re x1 / 2

(7-10)

Table 7-1 Analytical solutions of laminar flow over a flat plate

7.2.2 Thermal Boundary Layer

With dimensionless temperature T* = [(T-Ts)/ (T-T )] energy equation can be written in
similarity form as

d 2T *
d 2

Pr dT *
f
d
2

(7-11)

75

For Prandtl number Pr > 0.6, solution of differential energy equation gives correlation
for temperature gradient as

dT *
d

0.332 Pr1 / 3

Local Nusselt number has consequently a form

Nu x

hx x
k

0.332 Re1x/ 2 Pr 1 / 3

Pr 0.6

(7-12)

7.3 CYLINDER IN A CROSS FLOW

In a flow across cylinder (Figure 7-2) maximum pressure (p = pmax) is at forward
stagnation point, which also has zero fluid velocity u = 0, and locates just in front of
cylinder. Velocity u increases and pressure decreases with flow towards zenith. After
change of pressure gradient to dp/dx > 0, velocity decelerates and velocity gradient at
surface eventually becomes zero (Figure 7-3). As a result occur boundary layer
separation and wake formation in downstream region.

Figure 7-2 Boundary layer formation and turbulent separation on a cylinder in a cross flow.

Figure 7-3 Velocity profiles in turbulent wake formation in a cross flow over circular cylinder

For a cylinder Reynolds number is defined as

76

Re L

VD
v

VD

(7-13)

Figure 7-4 presents experimental results of convection coefficient variation with

angular coordinate and Reynolds number in air cross flow over circular cylinder.

Figure 7-4 Local convection heat transfer correlations for air cross flow over cylinder

7.3.1 Empirical Convection Heat Transfer Correlations

Overall average convection heat transfer coefficient correlations are interesting in
engineering applications. In following three correlations for cylinders in a cross flow
are presented.3
Hilperts Correlation

Nu D

hD
k

C Re mD Pr 1 / 3

Pr 0.7

(7-14)

Correlations coefficients are presented at Table 7-2 and properties are evaluated at film
temperature.
Table 7-2 Correlation coefficients for Hilperts Correlation

All heat transfer correlations are applicable to mass transfer problems via analogy treatment: average Nusselt number is replaced
by Sherwood number and Prandtl number by Schmidt number.

77

Convection correlation coefficients for other shapes than circular cylinders can be
obtained from Table 7-3.
Table 7-3 Correlation coefficients for Hilperts Correlation for noncircular cylinders

Pr
Nu D C Re Pr
Prs
Pr 10
n 0.37
Pr 10
n 0.36
m
D

Zukauskas

1/ 4

(7-15)

All the properties are calculated at bulk flow temperature t and correlation is valid for

0.7

Pr 500

1 Re D

10 6

Correlation coefficients are presented at Table 7-4.

Table 7-4 Correlation coefficients for Zukauskas correlation

Churchill & Bernstein

Nu D

0.3

0.62 Re1D/ 2 Pr 1/ 3
1 (0.4 / Pr) 2 / 3

1/ 4

Re D
282000

5/8

4/5

(7-16)

All material properties should be evaluated at film temperature. Entire range of

Reynolds numbers is covered and Pr > 0.2 is recommended.

78

7.4 SPHERE
Boundary layer phenomenon in flow over sphere is very much similar to those at cross
flow over a circular cylinder. Convection heat transfer coefficient correlations is
suggested by Whittaker as
1/ 4
1/ 2
D

Nu D

2 (0.4 Re

0.06 Re

2/3
D

) Pr

0.4

(7-17)
s

All the properties are calculated at bulk flow temperature t except surface viscosity
and correlation is valid for
0.71 Pr

380
7.6 *10 4

3.5 Re D
1.0 ( /

) 3.2

Another average convection heat transfer correlation for sphere is presented by Ranz &
Marshall
Nu D

2 0.6 Re1D/ 2 Pr1/ 3

, where first term corresponds pure conduction to stationary medium as

Nu D

7.5 FLOW ACROSS BANKS OF TUBES

Common heat exchanger design in industrial applications is a bank of tubes in a cross
flow. Arrangement of tubes are either aligned or staggered. Dimensions of tube bank
design are characterized by tube diameter D, longitudinal pitch ST and transverse pitch
SL (Figure 7-5).

Figure 7-5 Tube bank in a cross flow and arrangements of tubes: a) Aligned b) staggered tube
arrangements in a bank

Heat transfer coefficient depends on tube position in bank. Flow conditions and thus
heat transfer coefficient in a first row of tubes are approximately equal to single tube in

79

a cross flow. Heat transfer coefficient increases in next inner rows due to turbulent flow
conditions caused by the first tube rows. Usually after fourth or fifth row heat transfer
and flow conditions are stabilized and after this only minor change of heat transfer
coefficient occurs.
Beyond the first tube aligned tubes are in wake turbulence flow caused by upstream
tubes. In heat transfer point of view, aligned designs with SL /ST<0.7 are not
recommended. In staggered designs heat transfer is enhanced by tortuous flow,
particularly with small Reynolds numbers Re <100.

Figure 7-6 Flow Conditions for a) aligned and b) staggered

7.5.1 Grimisons Correlation

For air flow across tube bundles of 10 or more rows NL, Grimison has suggested
following correlation for dimensionless heat transfer coefficient
Nu D

C1 Re1D/ 2

(7-18)

NL 10
2000 Re D
Pr

40000

0.7

Where correlation constants are shown in Table 7-5 and Reynolds number for cross
flow is determined as
V max D
(7-19)
Re D

80

Table 7-5 Correlation constants for cross airflow over a tube bank

All properties are evaluated in film temperature. Maximum fluid velocity is calculated
at smallest cross-section. For aligned arrangement this is

Vmax

ST
ST

(7-20)

Heat transfer correlation is extended for other fluids as

Nu D

1.13C1 Re1D/ 2 Pr1 / 3

(7-21)

NL 10
2000 Re D
Pr

40000

0.7

7.5.2 Zukauskas Correlation

Another heat transfer correlation is recently suggested by Zukauskas for cross flow
over tube banks as

Nu D

C Re

m
D ,max

NL

Pr

0.36

Pr
Prs

1/ 4

(7-22)

20

1000 Re D 2 *10 6
0.7 Pr 500

Where correlation constants are shown in Table 7-5 and all properties except Prs are
calculated in arithmetic mean temperature of fluid inlet and outlet temperatures.

81

Table 7-6 Zukauskas correlation constants

7.5.3 Log-mean Temperature Difference

Heat transfer rate per unit length of tubes can be solved from
q N h D Tlm
N

(7-23)
Number of tubes in bank

Where appropriate form of temperature difference is log-mean temperature difference

defined as

Tlm

(Ts Ti ) (Ts To )
T To
ln s
Ts Ti
Ti
Ts

(7-24)

Inlet fluid temperature

Outlet fluid temperature

Fluid Outlet temperature may estimated from

Ts To
Ts Ti

exp
NT

DN h
VN T S T c p

(7-25)

number of tubes in transverse lane

82

7.6 SELECTION OF CORRELATION

Following steps should be considered, while selecting correlation for specific
application:
1)
2)
3)
4)
5)

Flow conditions: Laminar or turbulent

Validity range: Re, Pr
Use or availability: Local / average
Geometry: plate, cylinder, sphere, tube bank, etc.
Reference temperature of fluid properties

Table 7-7 Summary of convection correlations for external flow

83

Example VIII: Air flow over tube bank Cross flow HX

Air is used for cooling water in a cross flow tube bank. The tube bank is staggered and the air is forced to
flow across the tube bank with velocity of = 6 m/s. The air upstream temperature is 15 C. The tube
outside diameter is 16,4 mm and the longitudinal and transverse pitches are
= 34,3 mm and
=
31,3 mm. There are seven rows of tubes in the airflow direction and eight tubs per row. Under typical
operating conditions the cylinder surface temperature is at 70 C.
a) Determine the air-side convection coefficient and the rate of heat transfer for the tube bundle.
b) What is the air-side pressure drop?
PROPERTIES: Table A-4:
air ( = 15 C):
= 1,217 kg/m
= 1007 J/KgK
= 14,82 10 m /s,
= 0,0253 W/mK,
Pr = 0,710.

ASSUMPTIONS:
(1) Steady-state conditions,
(2) Negligible radiation effects,
(3) Negligible effects of change in air temperature across tube bank on air
properties.

air ( = 70 C):
Pr = 0,701. ( = 43 C):
= 17,4 10 m /s,
= 0,0274 W/mK,
Pr = 0,705

ANALYSIS:
a) From equation 7-22 (Zukauskas correlation), the air-side Nusselt number is
Pr

Re

Pr
Pr

+ ( /2) = 37,7 mm is greater than (

Since
=
occurs on the transverse plane . Hence
=
Re

=
=

)/2, the maximum velocity

31,3 mm
6 m/s = 12,6 m/s
(31,1 16,4) mm

m
12,6 s 0,0164 m
=
= 13,943
14,82 10 m /s

31,3 mm
= 0,91 < 2
34,3 mm

It follows that from Tables 7-6

= 0,35

= 0,34

= 0,6

= 0,95 0,34 13,943

84

= 0,95
0,71

0,71
0,701

= 87,9

From equation 7.67

Heat flux is calculated from eq. 7-23 and 7-24

b) The pressure drop from equation

85

7.7 LEARNING OUTCOMES

Table 7-8 Learning Outcomes: Forced Convection - External

Level of
Knowledge

Concept

Understand
Describe

Principle of analytical solution

Development of Boundary Layers: cylinder
Basis for Selection of correlation

Apply
understand
describe

Empirical Nu correlations: plate, cylinder, sphere & tube bank

Log-mean temperature difference bank of tubes

86

8 FORCED CONVECTION INTERNAL FLOW

Unlike in external flow, boundary layer development is constrained by surface in
internal flow. Within internal flow, hydrodynamic entrance or fully developed flow
regions determine flow conditions besides laminar or turbulent flow type.

8.1 HYDRODYNAMICS
Let us study laminar flow in a circular tube, where fluid enters with uniform velocity.
The viscous effect of tube walls will form boundary layer developing with increasing
distance x from entrance. Simultaneously uniform velocity, inviscid flow, region
shrinks until boundary layers merge at centreline of tube. Point of merging boundary
layers is end of hydrodynamic entrance region and beginning of fully developed flow
region.

Figure 8-1 Laminar hydrodynamic boundary layer development in tube flow

Reynolds number for circular tube flow is defined as

um D u m D
Re D
v
um
Mean velocity in tube cross-section

(8-1)

In a fully developed flow, critical Reynolds number corresponding start of turbulence

flow behaviour is
Re D ,C

2300

Although larger Reynolds number (~10000) is needed for achieving fully turbulent
conditions.
For laminar flow, one correlation form for hydrodynamic entry length is given as

x fd ,h
D

0.05 Re D

(8-2)

lam

For turbulent flow, hydrodynamic entry length is not dependent on Reynolds number
and a rough approximation is expressed as

87

10

x fd ,h
D

60

(8-3)

turb

Mean velocity for incompressible flow can solved form

r

um

2 o
u (r , x) rdr
ro2 0

(8-4)

And for laminar fully developed flow velocity profile is

u (r )
um

21

r
r0

(8-5)

Shape of laminar velocity profile in fully developed region is parabolic, while turbulent
profile is more flat due to turbulent radial mixing. At entrance region velocity gradients
are high near wall surface and decreases together with increasing boundary layer
thickness towards fully developed region.

8.2 THERMAL PERFORMANCE

Let us now look to thermal behaviour of uniform temperature fluid flow entering to
circular tube. When tube surface temperature is higher than fluid temperature,
convection heat transfer develops a thermal boundary layer. Thermally fully developed
region is eventually reached and the shape of temperature profile depends on whether
boundary condition is
a) uniform surface temperature
or
b) constant surface heat flux.

Figure 8-2 Thermal boundary layer development

88

For laminar flow, one correlation form for thermal entry length is given as

x fd ,t
D

0.05 Re D Pr

(8-6)

lam

For extremely high Prandtl numbers (oils) a practical assumption is velocity profile as
fully developed also throughout entrance region. For turbulent flows, high Reynolds
number, conditions are independent from Pr and first rule of thumb approximation
would be

x fd ,t
D

10

(8-7)

turb

Newtons law of cooling for internal flows has a form

q s''

h(TS

Tm )

(8-8)

Where mean temperature is reference temperature and can be solved from

r

Tm

2 o
uTrdr
u m ro2 0

(8-9)

In thermally fully developed flow with constant properties local convection heat
transfer coefficient is constant (Figure 8-3). Even mean temperature changes with
distance from entry of tube, dimensionless temperature profile is independent of
distance:

Ts ( x) T (r , x)
x Ts ( x) Tm ( x)

(8-10)

fd ,t

Figure 8-3 Axial variation of convection heat transfer coefficient

89

Example IX: Tube flow

What are mean temperature and Nusselt number for tube flow with known velocity and temperature
profiles?

ASSUMPTIONS: constant properties, incompressible

ANALYSIS: Mean temperature - Definition
r

2 0
uTrdr
2
u m r0 0

Tm

r2
2C1
T
s
u m r02
2

2C1

r
2C1
Ts 0
2
u m r0
2
Ts

r0
2

r
ro

Nu

hD
k

rdr

C2 2
r0
4

q s''

Convection coefficient

r0

Note! here C1

Heat Flux Fourier Law

Nusselt number

C2 r 4
4 r02
2

C2
2

C2 1

r2
C2
2
C2

Ts

u m r0

r0

T
r

um

k C2 2
r ro

qs''
Ts Tm

( 4 k / ro ) * 2 ro
k

90

r
ro2

2C2
r ro

2C2 (k / ro )
C2 / 2
8

4k
ro

k
ro2

8.2.1 Energy Balance

Mean temperature variation in tube can be solved by energy balance approach. If
steady-state convection flow occurs in tube with negligible viscous dissipation,
incompressibility or ideal gas assumption with low pressure variation is reasonable,
energy equation can be written to form

qconv

m c p (Tm ,o Tm ,i )

(8-11)

With constant heat flux boundary condition, mean temperature can be solved from
energy equation as
Tm ( x) Tm ,i

q s'' P

(8-12)

mcp

With constant surface temperature boundary condition, mean temperature has following
relationship to flow distance x

Ts Tm ( x)
Ts Tm ,i

exp

Px h

(8-13)

mcp
h

Average convection coefficient from entry to distance x

Surface temperature and flow mean temperatures in two common boundary conditions:
constant surface heat flux and constant surface temperature; are shown in Figure 8-4.

Figure 8-4 Axial temperature variations for heat transfer in a tube a) constant surface heat flux b)
constant surface temperature

91

8.3 CONVECTION HEAT TRANSFER CORRELATIONS

Convection heat transfer correlations are summarized in table below and following
chapter gives a brief introduction to principles beyond the equations.
Table 8-1 Convection heat transfer correlations for flow in circular tube

92

8.3.1 Laminar Flows

Energy balance for differential element in laminar, fully developed flow of
incompressible, constant property fluid can be written with axial advection and radial
conduction terms as

T
dx
x

(d m)c p

qr

qr
dr
r

qr

qr
dr
r

(8-14)

Figure 8-5 Differential energy balance for in laminar, fully developed flow of incompressible, constant
property fluid

By substituting definitions of mass flow rate dm

u 2 rdr and radial heat transfer rate
qr
k ( T / r )2 rdx , energy equation reduces to

T
x

r r

T
r

(8-15)

Constant heat flux - Fully developed conditions

For constant heat flux energy equation can be written as

1
T
r
r r
r

2u m dTm
dx

r
ro

(8-16)

Differential equation has a solution as constant Nusselt number

Nu d

hD
k

4.36

Constant surface temperature - Fully developed conditions

For constant heat flux energy equation can be written as

1
T
r
r r
r

2u m dTm
1
dx

r
ro

Ts T
Ts Tm

Solution for above differential equation is a constant Nusselt number

93

(8-17)

Nu d

3.66

Conclusion: Nusselt number is constant and independent of Re, Pr and distance x for
laminar, fully developed internal flow in circular tubes.

8.3.2 Entry Region

Figure 8-6 a presents Nusselt number solutions for thermal entry length with
assumption of fully developed velocity profile, which would be in cases with unheated
starting length or with large Prandtl number fluids, i.e. oils. Heat transfer coefficient for
fully developed velocity profile is independent on viscosity and thus to change of
Prandtl number.
For simultaneous laminar velocity boundary layer and thermal boundary layer
development viscous forces, and Prandtl number, have effect to entry length heat
transfer solution. Figure 8-6 a presents also combined entry length solution for Prandtl
number 0.7, which is usually suitable in case of gas flows.
Correlation for average Nusselt number at entrance region for thermal entry length or
low Prandtl number (Pr<5) problems, correlation has been presented as (Figure 8-6 b)

Nu D

3.66

0.0668( D / L) Re D Pr
2/3
1 0.04 ( D / L) Re D Pr

(8-18)

For combined entry length problems Sieder and Tate correlation has form of

Nu D

Re D Pr
1.86
L/ D

0.14

1/ 3

0.60 Pr 5
s

94

(8-19)

Figure 8-6 Nusselt numbers entry length solutions for laminar flow in circular tube a) local Nusselt
number solutions b) average Nusselt number correlations and test results

8.3.3 Pressure drop and friction factor in fully developed flow

Moody friction factor is defined as

(dp / dx ) D
u m2 / 2

(8-20)

Fanning friction coefficient is related to friction factor as

Cf

f
4

(8-21)

um2 / 2

Moody friction factor can be written for fully developed laminar flow as

64
Re D

(8-22)

Friction factors are presented at Table 8-1 for different flow regimes and in Figure 8-7.
Pressure drop for fully developed flow between positions x1 and x2 is
p

p2
p1

dp

u m2
( x1
f
2D

x2 )

95

(8-23)

Figure 8-7 Friction factor for fully developed flow in circular tube

8.3.4 Turbulent Flows

Classical expression for solving local Nusselt number for fully developed turbulent
flow in smooth circular tube is presented by Colburn.
Colburn Analogy

Colburn equation

Cf
2
Nu D

f
8

St Pr 2 / 3

Nu D
Pr 2 / 3
Re D Pr

0.023 Re D4 / 5 Pr1 / 3

(8-24)

(8-25)

Dittus-Boelter equations are suitable for moderate temperature differences of surface

and mean fluid temperature, where properties are evaluated in mean temperature:
Nu D

0.023 Re 4D/ 5 Pr 0.4 (Heating)

(8-26)

Nu D

0.023 Re 4D/ 5 Pr 0.3 (Cooling)

(8-27)

L / D 10
Re D

10000

0.6 Pr 160

96

Temperature dependent property variations are taken account in correlation proposed

by Sieder and Tate with viscosity ratio as
0.14

Nu D

0.023 Re

4/ 5
D

Pr

1/ 3

(8-28)
s

For long tubes average dimensionless heat transfer coefficient is approximately same as
in local fully developed region as entry length is relatively short (10<x/D<60):
Nu D

Nu D

(8-29)

And for short tubes average Nusselt number can be evaluated with following expression
Nu D
Nu D, fd

C
x/ D

(8-30)

Coefficients C and m depends on dimensions and form of inlet and entry region as well
as Prandtl and Reynolds numbers.

8.3.5 Noncircular Tubes

A simple way of approximating heat transfer for noncircular tubes is utilizing circular
tube results with hydraulic diameter (effective diameter). Hydraulic diameter is defined
as
Dh

4 AC
P

(8-31)

AC
P

Cross-sectional area
Wetted Perimeter

For laminar flows, specifically in sharp tube with sharp corners, circular tube
correlations are less accurate and Nusselt number correlations for fully developed
conditions in laminar flow are listed in Table 8-2.

97

Table 8-2 Nusselt numbers & friction factors for fully developed laminar tube flows in differing crosssection

Example X: Heating engine oil Tube flow with constant surface temperature
Engine oil flows at a rate of 1 kg/s through a 5 mm diameter straight tube. The oil has an inlet
temperature of 45 C and it is desired to heat the oil to a mean temperature of 80 C at the exit of the
tube. The surface of the tube is maintained at 150 C. Determine the required length of the tube. Hint;
Calculate the Reynolds numbers at the entrance and exit of the tube before proceeding with your
analysis.
ASSUMPTIONS:
(1) Steady-state,
(2) Incompressible flow,
(3)
Negligible
viscous
dissipation.
PROPERTIES: Engine oil (Ti = 45 C = 318 K):

98

; (To = 80 C = 318 K):

From properties tables (Incr.Table A-5), this occurs at a transition temperature of

= 325 K = 52 C.
Now we proceed to analyze separately the heat transfer in the laminar and turbulent regions.
PROPERTIES (Laminar region): Table A-5,
= (45 C + 52 C) 2 = 48,5 C = 321,5 K :
= 1999 J/kgK,
= 13,2 10 Ns/m , = 0,143 W/mK, Pr = 1851.
Re

1 kg/s
= 1930
0,005 m 13,2 10 Ns/m

The hydrodynamic and thermal entry lengths are given by (Eq. 8-2, 8-6)

x fd ,h

0.05 Re D D

x fd ,t

x fd , h Pri

0,05 1930 0,005 m

0,48 m

0,48 m 1851 890 m

Flow is hydrodynamically developed but thermally developing, (Eq. 8-18)

3.66

Nu D

0.0668( D / L) Re D Pr
2/3
1 0.04 ( D / L) Re D Pr

h1 D / k1
(1)

To solve laminar region length L1 we need also an equation for mean temperature variation, expressed
between the inlet and the transition point. (Eq. 8-13)

Ts Tm ( x)
Ts Tm,i

h1 L1

m cp

m cp,l

h1 L1

Dx h

exp

ln

Ts

Tm, t

Ts

Tm,i

1kg/s 1999 J/kgK 150 C 52 C

ln
150 C 45 C
0,005m

(2)

8780W/mK

The right values of

= 3,66 we find.

and

are solved by iterating Eqs (1) and (2). Beginning with the estimate

= 105 W/m K (Eq. 1),

= 83,6 m (Eq. 2)

= 22,3 (Eq. 1),

= 638 W/m K (Eq. 1),

= 13,8 m (Eq.

2)

= 18,7 (Eq. 1),

= 535 W/m K (Eq. 1),

= 16,4 m (Eq.

2)

= 16,9 (Eq. 1),

= 484 W/m K (Eq. 1),

= 18,1 m (Eq.

2)

= 3,66,
1 round
st

2nd round

3rd round

10th round

99

= 16,9 (Eq. 1),

= 484 W/m K (Eq. 1),

11th round

PROPERTIES (Turbulent region):

=
10 Ns/m ,
= 0,139 W/mK, Pr = 834.
Re

= 18,1 m (Eq.

= 66 C = 339 K,

= 2072 J/kgK,

2)

= 5,62

= 4530

We assume that the flow is fully-developed hydrodynamically and thermally and use Eq 8.62 from Table
8-1.
=

= (0,790 lnRe
=

1,64)

= (0,790 ln (4530)

1,64)

= 0,0398

= 184

= 5120 W/m K

The required length L2 can be found from (Eq. 8-13) expressed between the inlet and the transition point.

L2

m cp,2
D h2

ln

Ts Tm,o

1kg/s 2072 J/kgK

150 C 80 C
ln
0,005 m 5120 W/m 2 K 150 C 52 C

Ts Tm,t

The total required length is L

L1

L2

18,1 m 8,7 m

8,7 m

26,8 m

COMMENT: If we had assumed the flow to be turbulent throughout, we would have get higher average
Nusselt number and significantly lower total length (L = 11,9 m). The variation of properties with
temperature can be very important for some fluids such as oils.

8.4 LEARNING OUTCOMES

Table 8-3 Learning Outcomes: Forced Convection - Internal

Level of
Knowledge
Understand
Describe

Concept

Hydrodynamic and thermal entry region: convection coefficient

behaviour

Fully developed flow

Apply
understand
describe

Empirical

&

Nu

correlations:

turbulent/laminar,

developed/entry effects

Log-mean temperature difference fluid mean temperature

Concepts: hydraulic diameter

100

fully

9 FREE CONVECTION
Free or natural convection is formed due to buoyancy forces caused by density
gradients and gravitational body forces that are proportional to density. In most cases
density gradients are formed due to temperature gradients. Velocities in free convection
cases are smaller than in forced convection, and thus resulting as much smaller heat
transfer coefficients.
Importance of free convection is relevant in many multimode heat transfer effects,
commonly with radiation, and free convection is preferred choice for cooling
applications from operation cost minimization point of view.
Applications
Cooling of electrical equipments
Natural circulation boilers
Cooling of pipelines
Windows and radiators
Condensers of refrigeration devices
Flow dynamics of water systems and atmosphere
Free boundary flows are not bounded by surface and may occur in form of plumes or
buoyant jets.

Figure 9-1 Free boundary layer development a) plume above heated fire b) buoyant jet with from heat
discharge

Classical free convection flow bounded by surface is heated vertical wall. Boundary
layer develops to surface of wall, when lighter heated fluid on wall starts to flow
upwards. Velocity is zero both on wall and at quiescent fluid region, out of boundary
layer area.

101

Figure 9-2 Free convection boundary layer development on vertical heated plate

9.1 ANALYTICAL SOLUTIONS FOR VERTICAL PLATE

Conservation principles are similar to those in forced convection. Mass, momentum and
energy conservation equation set is valid for free convection boundary layer flows with
negligible viscous effect and incompressible flow assumptions, (except in x-momentum
equation density change resulting in buoyancy flow with viscous effect), are
u
x

Continuity

v
y

(9-1)

1 p
x

u
u
v
x
y

X-momentum

Energy

T
u
x

g v

y2

(9-2)

T
v
y

T
y2

(9-3)

By substituting pressure gradient dp / dx

g in quiescent boundary region outside
from boundary layer to x-momentum equation following momentum equation can be
written

u
u
v
x
y

y2

(9-4)

If density differences causing free convection flow are only due to temperature
differences, free convection flow can be related to fluid property, volumetric thermal
expansion coefficient, which is defined as

102

1
v

v
T

1
T

1
T

(9-5)

Specific volume [m3/kg]

Volumetric thermal expansion coefficient describes density change response to

temperature change at constant pressure.
Boussinesq approximation states for density difference that
(T T )

(9-6)

Giving a form for x-momentum conservation equation as

u

u
x

u
y

g T T

Buoyancy force

Momentum of Convection

y2

(9-7)

Viscous force

For ideal gas (pv = RT) volumetric thermal expansion coefficient can be written as

1
T

1
T

p
RT 2

(9-8)

For liquids and non-ideal gases, temperature (&pressure) dependent correlations and
property tables are available.
Similarity
By defining dimensionless parameters as
x
L
u
u0

x*

u*

y*

v*

y
L
v
u0

L = Characteristic length of surface

T*

T T
TS T

X-momentum conservation equation can be written in dimensionless form

2
g TS T
u*
1
T*
2
Re L y *2
u0

u*
u*
u*
v*
x*
y*

(9-9)

By choosing arbitrary reference velocity u0 as

u0

T ) L3

g (TS

(9-10)

v2

Reynolds number for free convection may be presented as

Re L

g TS

T ) L3 / v 2

1/ 2

103

(9-11)

Grashof Number
Definition of Grashof number, dimensionless parameter relevant for free convection, is
a square of Reynolds number described above. Grashof number describes ratio of
buoyancy forces to viscous forces and is defined as
Gr

T L3

g Ts

(9-12)

Grashof number in free convection flows corresponds to Reynolds number in forced

convection flows. As Nusselt number is function of Reynolds and Prandtl numbers in
forced convection cases, heat transfer coefficient h(Gr) is function of Grashof number
in free convection problems. Convection domination mode assessment can be done
with Grashof and Reynolds numbers as
Table 9-1 Free or forced convection dominating
Gr/Re2 1
Gr/Re2 << 1
Gr/Re2 >> 1

Mixed free and forced convection

Forced convection
Free convection

9.1.1 Laminar Free Convection Solutions

Solutions for conservation equations for vertical heated plate can be presented in
similarity parameter graphs (Figure 9-3). Boundary conditions for problem are
y

0: u

: u

v 0 T

0 T

TS

104

Figure 9-3 Free convection boundary layer solutions for laminar flow on vertical surface a) velocity b)
temperature profiles

Local convection heat transfer coefficient hx for laminar free convection on vertical
plate is

Nu x

g (Pr)

hx x
k

Grx
4

1/ 4

dT *
d

Grx
4

1/ 4

g (Pr)

(9-13)

0,75 Pr1 / 2
(0,609 1,221Pr1 / 2 1,238 Pr)1 / 4

Remembering that Prandtl number was defined as ratio of kinetic viscosity and thermal
diffusivity as follows
cp
cp
Pr
(9-14)
k
k
Average convection heat transfer coefficient h is obtained by integrating local value
over wall length as
Nu L

hL
k

4 GrL
3 4

1/ 4

g (Pr)

105

4
Nu L
3

(9-15)

9.1.2 Turbulence
As in forced convection, disturbances in flow may be amplified and turbulent flow may
be induced in free convection flows. Similarly heat transfer is more effective in case of
turbulent flow. Heat transfer coefficients for turbulent flows have to be defined from
experimental correlations.
Rayleigh Number is used for correlating transition from laminar to turbulent boundary
layer, which is dependent on ratio of buoyancy and viscous forces. Definition of
Rayleigh number is product of Grashof and Prandtl number and is defined for vertical
plates as
g (Ts T ) L3
(9-16)
Ra Gr Pr
In free convection transition from laminar to turbulent flow occurs at Rayleigh number
exceeding approximately 109.

9.2 EMPIRICAL CORRELATIONS

Generally free convection heat transfer correlations are in form
Nu L

hL
k

CRa Ln

(9-17)

n = 1 / 4 for laminar (Vertical plates: C=0.59)

n = 1 / 3 for turbulent (Vertical plates: C=0.1)

Fluid properties are defined in film temperature T f

TS

T / 2.

9.2.1 Plates
Empirical correlation for vertical plate over entire range or Rayleigh number is
proposed as

Nu L

0,387 Ra 1L/ 6

0,825
1

0, 492 / Pr

Laminar flow, 10

Turbulent flow, 10

9 / 16 8 / 27

(9-18)

Ra L 10 9
RaL 1013

Average Nusselt number for horizontal plates can approximated with following
correlations:
Heated surface up or cooled surface down
Nu L

0,54 Ra1L/ 4

10 4

Ra L

10 7

(9-19)

Nu L

0,15 Ra1L/ 3

10 7

Ra L

1011

(9-20)

106

Heated surface down or cooled surface up

Nu L

0,27 Ra1L/ 4

10 5

Ra L

1010

(9-21)

9.2.2 Long Horizontal Cylinder

Free convection boundary layer is formed across heated cylinder in quiescent fluid.
Local Nusselt number depends on formation of laminar boundary layer as shown in
Figure 9-4.

Figure 9-4 boundary layer development and Nusselt number distribution on heated horizontal cylinder

Correlation form for average Nusselt number is given by Churchill and Chu as
0,387 Ra1D/ 6

0,60

Nu D

0,559 / Pr

9 / 16 8 / 27

Ra D 1012

(9-22)

9.2.3 Sphere
Average Nusselt number correlation for sphere is proposed by Churchill as
Nu D

0,589 Ra1D/ 4
1 (0, 469 / Pr) 9 / 16

4/ 9

(9-23)

Pr 0,7
RaD 1011

As Rayleigh number approaches zero, heat transfer coefficient approaches 2, which

corresponds conduction heat transfer coefficient of sphere in stationary infinite
medium.

107

Example XI: Analytical & Empirical Solutions Free Convection over vertical plate
A square aluminum plate 5 mm thick and 200 mm on a side is heated while vertically suspended in
quiescent air at 40 C. Determine the average heat transfer coefficient for the plate when its temperature
is 15 C by two methods: using results from the similarity solution to the boundary layer equations, and
using results from an empirical correlation.

PROPERTIES: Table A-4,

) 2 = (40 + 15)C 2 = 300 K, 1 atm)
Air ( = (
= 15,89 10 m /s, = 0,0263 W/mK, = 22,5 10
Pr = 0,707.

ASSUMPTIONS:
(1) Uniform plate surface
temperature,
(2) Quiescent room air,
(3) Surface radiation exchange
with surroundings negligible,
(4) Perfect gas behavior for air,
=1 .

m /s,

ANALYSIS: Rayleigh number (Eq. 9-15)

Ra =
(

Ra =

)(

= 1,827 10 < 10 , (Laminar)

Nusselt number (Uniform plate surface temperature, quiescent air) Eq. 9-14:
=

g(Pr)

(1)

Interpolation of the Figure 9-3 (temperature gradient at the surface as a function of the Prandtl number)
Eq. 9-12
g(Pr) = (

(2)

Substituting numerical values with with Gr = Ra /Pr

g(Pr) = (
=

4 Gr
3 4

g(Pr) =

= 4,42 W/m K

Nu L

hL

is given by Eq. 9-17

0,492 / Pr

9 / 16 8 / 27

0,387Ra1L/ 6
1

0,387 Ra 1L/ 6

0,825

k
0,825
L

= 0,501

0,0263 W/mK 4 1,827 10 /0,707

0,20 m
3
4

The appropriate empirical correlation for estimating

hLL
k

0,492 / Pr

9 / 16 8 / 27

108

0,501

hL

0,387 1,827 107

0,0263 W/mK
0,825
0,2 m

0,492 / 0,707

1/ 6

9 / 16 8 / 27

4,87 W/m2 K

Comments:
1)

In this case, both analytics and empirics, give a close to each other solution for free convection heat
transfer coefficient

Example XII: Free convection between Fins Cooling of Parked Motor Cycle
Engine
Consider an array of vertical rectangular fins, which is to be used to cool an electronic device mounted in
quiescent air at
= 27 C (1 atm). Each fin has = 20 mm and has = 150 mm and operates at an
approximately uniform temperature of = 77 C
a) Viewing each fin surface as a vertical plate in an infinite quiescent medium, briefly describe
why there exists an optimum fin spacing . Using figure 9-3, estimate the optimum value of
for prescribed conditions.
b) For the optimum value of and a fin thickness of = 1,5 mm, estimate the rate of heat transfer
from the fins for an array of width
= 355 mm.
3 ASSUMPTIONS:
(1) Fins are isothermal
(2) Radiation effects are
negligible,

PROPERTIES: Table A-4,

Air ( = ( + ) 2 = (27 + 77)C 2 = 52 C = 325 K):
= 18,41 10 m /s, = 0,0282 W/mK, Pr = 0,703.

(3) Air is quiescent.

ANALYSIS:
a) If fins are too close, boundary layers on adjoining surfaces will coalesce and heat transfer will
decrease. If fins are too far apart, the surface area becomes too small and heat transfer decrease.
=2
. From Fig. 9-3, the edge of boundary layer corresponds to
=

Gr
4

Gr =

= 1,5 10

( )=(

(1 325 K) (77 27)

(18,41 10 m /s)

9,81 m/s

= 0,017 m = 17 mm

b) The number of fins N can be found as

=

355 mm
= 10
(34 + 1,5) mm

Heat transfer rate:

(

)=2

109

(0,150 m)

34 mm

The appropriate empirical correlation for estimating

hLH
k

Nu H

hL

0,492 / Pr

9 / 16 8 / 27

0,387 Ra 1L/ 6
1

0,387 Ra 1L/ 6

0,825

k
0,825
H

is given by Eq. 9-17

0,492 / Pr

9 / 16 8 / 27

1/ 6

hL

0,387 1,5 10 7 0,703

0,0282 W/mK
0,825
9 / 16 8 / 27
0,150 m
1 0,492 / 0,703
= 2 10 5,95

W
0,15 m 0,02 m (350
m K

5,95 W/m 2 K

300) K = 17,5 W

COMMENTS:
1) Further reduction in S would compensate for the effect of fluid entrapment due to boundary
layer merger.
2) From a more rigorous treatment
10
is obtained for the prescribed conditions
joining of boundary layers dont decrease heat transfer some much than increased heat transfer
surface will enhance it -> optimization problem

9.3 LEARNING OUTCOMES

Table 9-2 Learning Outcomes: Free Convection

Level of
Knowledge

Concept

Apply,
Understand,
describe

Similarity solutions for laminar free convection (Figure 9-3)

Concepts: Ra, Gr, Re and empirical Nu correlations, volumetric
expansion
Distinguish free from forced convection: Gr limit

Understand,
describe

Boundary layer development in different geometries

Terms of conservation equations
transition from laminar to turbulent

110

10 BOILING AND CONDENSATION

Boiling and condensation are convection processes occurring at solid-liquid or solidvapour interface and are associated with phase change of a fluid. High heat transfer
rates with small temperature differences are achieved due to latent heat of phase
change. Two other important fluid properties effecting to boiling or condensation are
liquid-vapour surface tension and density difference of two phases. Density difference
of two phases induces a buoyancy force that causes fluid convective motion promoting
heat transfer together with latent heat effect.
Applications
Power Plants
o Boiler: vaporization of pressurized liquid
o Condenser: expanded vapour condensed back to feed water
Refrigerators and heat pumps: evaporators and condensers
Advanced electronic cooling
Cooking: Boiling 3 min eggs & Nature: Rain, misty lake

10.1 PRINCIPLES OF BOILING

Evaporation at solid-liquid interface is termed as boiling. Boiling starts with formation
of vapour bubbles, if surface temperature of solid rises over fluid saturation temperature
corresponding liquid pressure. Newtons law of cooling for boiling heat transfer rate is
q s''

h(TS

Tsat )

h Te

Te
Tsat ( p )

(10-1)
Excess temperature
Fluid saturation temperature

Boiling convection heat transfer is affected by the dynamics of formation, growth and
movement of vapour bubbles and related liquid motion near surface.

10.1.1 Classification
Characteristic for pool boiling is relatively quiescent liquid, and bubble formation and
free convection induced liquid motion near surface. Forced convection boiling is on the
other hand characterized with externally induced fluid flow together with free
convection and bubble-dynamics induced mixing.
Sub-cooled boiling means vapour formation in liquid having temperature below its
saturation temperature. Boiling vapour formed at surface might later condense in liquid.
In saturated boiling temperature of liquid slightly exceeds saturation temperature.

111

10.2 POOL BOILING

Pool boiling heat flux plotted against excess temperature is called boiling curve (Figure
10-1). Boiling curve reveals characteristics of different boiling regimes of pool boiling.
Different boiling regimes are classified according excess temperature and each mode
has its own bubble-fluid behaviour and heat transfer characteristics.

Figure 10-1 Boiling curve for saturated water at atmospheric pressure

10.2.1 Free Convection Boiling

Free convection boiling exists below ~5C excess temperature ( Te <5C). In free
convection boiling fluid motion is determined by free convection and thus free
convection heat transfer correlations can be used in this boiling region.

10.2.2 Nucleate Boiling

Onset of nucleate boiling (ONB, Te =5C) is starting point of nucleate boiling (Figure
10-1, point A), which continues until point of critical heat flux (point C, Te ,C 30C).

112

Region A-B: Isolated bubbles are form and rise from surface. Bubble separation from
surface enhances fluid mixing near surface and increases both heat transfer coefficient
and rate. Heat transfer is mostly from surface to fluid in boiling convective motion.
Region B-C: Vapour escapes as jets and bubbles and merges to slugs of vapour as
bubble formation is further increased (Figure 10-2). Heat transfer coefficient increases
until maximum point a, when it starts to decrease with excess temperature caused by
increased vapour bubble formation on surface. Heat flux increases with excess
temperature until critical heat flux ( Te ,C 30C), end of nucleate boiling regime.
Increased vapour formation prevents fluid to wet surface and thus local maximum heat
flux is reached.

Figure 10-2 Nucleate boiling with vapour jets

10.2.3 Transition boiling

In transition region 30C < Te <120C) vapour bubble formation is rapid and
conditions at surface oscillates between nucleate and film boiling. Fraction of vapour
film coverage increases with excess temperature. Heat transfer coefficient and heat
transfer rate are decreasing caused by lower conductivity of vapour compared to liquid.

Figure 10-3 Transition boiling

113

10.2.4 Film Boiling

At Leidenfrost point (point D) minimum heat flux is reached and surface is covered
totally by vapour film. Small water droplets are evaporated in vapour film. Heat
transfer occurs by conduction and radiation through vapour blanket. Increase of heat
transfer in film boiling region with excess temperature is enhanced by increase of
surface radiation.

Figure 10-4 Film boiling

Increase of heat flux from point of maximum heat flux (point C) causes fast increase of
surface temperature due to boiling crisis (or burnout point). Boiling crisis is a change of
boiling regime from nucleate to film boiling with fast departure from boiling curve.

10.2.5 Pool Boiling Correlations

Nucleate Pool Boiling
Rohsenow correlation is widely used for determine nucleate boiling heat transfer rate
q

''
s

h fg

1/ 2
l

c p ,l Te
C S , f h fg Prln

114

(10-2)

Table 10-1 Coefficient values for Rohsenow correlation for various surface-liquid combinations

Rohsenow correlation applies only for clean surfaces and error marginal is 100%
when predicting heat fluxes.
Boiling phenomenon in nucleate region is similar to forced convection with fluid
motion induced by rise of bubbles. Forced convection heat transfer correlations are
generally in form of

Nu L

C fc Re L fc Pr

n fc

(10-3)

Heat fluxes
Prediction of critical, maximum, heat flux is important in sense of avoiding boiling
crisis. Expression for maximum heat flux is presented as

''
max

C h fg

1/ 4

(10-4)

2
v

Leading constant C depends weekly on surface geometry. For spheres, horizontal

cylinders and large finite heated surfaces value /24=0.131 can be used. Critical heat
flux has strong dependence to pressure due to surface tension and heat of vaporization.
Film Boiling
For film boiling on a cylinder or sphere heat transfer correlation obtained from laminar
film condensation analogy is proposed as
Nu D

h conv D
kv
C

v v k v (Ts

h 'fg

1/ 4

Tsat )

Constant: 0.62 for cylinders, 0.67 for spheres

115

(10-5)

Corrected latent heat accounting for the heat needed to maintain film temperature above
saturation temperature may be approximated as
h 'fg

h fg

0.8c p ,v (TS

Tsat )

(10-6)

At temperatures above 300C radiation heat transfer has significant influence on total
heat transfer. Effective radiation coefficient is expressed as

hrad

(Ts4 Tsat4 )
Ts Tsat

(10-7)
Emissivity of solid
Stefan-Boltzmann constant

Total heat transfer coefficient for film boiling accounting convection and radiation parts
is suggested by Bromley to be solved from
h

4/3

4/3

h conv

4/ 3

h rad h

1/ 3

(10-8)

Example XIII: Boiling water in Coffee Machine Heater element

A heater element of 5-mm diameter is maintained at a surface temperature of 350 C when immersed
horizontally in water under atmospheric pressure. The element sheath is stainless steel with a
mechanically polished finish having an emissivity of 0.25.
a) Calculate the electrical power dissipation and the rate of vapor production per unit heater length.
b) If the heater were operated at the same power dispation rate in the nucleate boiling regime, what
temperature would the surface achieve? Calculate the rate of wapor production per unit length
for this operating condition.
c) Sketch the boiling curve and represent the two operating conditions of parts a) and b). Compare
the results of your analysis. If the heater element is operated in the power controlled mode,
explain how you would achieve these two operating conditions beginning with a cold element.

ASSUMPTIONS:
(1) Steady-state conditions,
(2) Water exposed to standard atmospheric pressure and uniform temperature
PROPERTIES: Table A-6, Saturated water, liquid (
= 100 C): = 957,9 kg/ m ,
=
(
+ 0,80
4217 J/kgK, = 279 10 Ns/ m , Pr = 1,76,
= 2257 kJ/kg,
= 100 C):
= 0,5955 kg/
) = 2654 kJ/kg, = 58,9 10 N/m; Saturated water, vapour (
) 2 500 C):
m , Table A-4, Water vapor ( = (
= 0,4405 kg/ m ,
= 4217 J/kgK,
= 0,0339 W/mK
= 38,68 10 m /s.

ANALYSIS:
a) Since
> 120 C, the element is operating in the film-boiling (FB) regime. The electrical
power dissipation per unit length is
( )(
)
(1)

116

total heat transfer coefficient is

(2)
The convection coefficient, correlation eq. 10-5 ( = 0,62):
(

0,0339 W/mK
0,005 m

9,81 m/s (957,9 0,5955) kg/ m 2654 kJ/kg (0,005 m)

38,68 10 m /s 0,0339 W/mK (350 100)K

0,62
= 225

W
m K

The radiation coefficient, Eq. 10-7, where the total hemispherical emissivity = 0,25 and the

Stefan-Bolzman constant
(

=
=

= 2,57 10
)

0,25 2,57 10

W/m K

W/m K (623
(623 373) K

373 ) K

= 7,4 W/m K

Substituting numerical values into Eg. (1) and (2)

= 231 W/m K
(

= 231

0,005 m)(623

373) K = 907 W/m

907 W/m
= 57,8 kW/m
0,005 m

The evaporation rate per unit length is

m =

= 1,4 kg/ms

b) For the same heat flux,

regime.
=

= 57,8 kW/m , using the Rohsenow for the nucleate boiling (NB)
)

Pr

Where, from Table 10-1, for stainless steel mechanically polished finish with water,
0,0132 and = 1,0

117

Pr

= 7,6 K

= 107,6 C

The evaporation rate per unit length is

m =

= 1,4 kg/ms

The two operating conditions are shown on the boiling curve. The element can be operated at
NB with the same heat flux
= 107,6 C

= 57,8 kW/m , with a surface temperature of

COMMENTS:
1) If the element is cole, it would be brought to the NB condition following the arrow of the boiling
curve. To reach the FB regime, the heat flux is increased to the critical heat flux (CHF). The
burnout condition will then likely occur. If not, the FB point is reached by decreasing the heat
flux.

118

10.3 FORCED CONVECTION BOILING

In forced convection boiling fluid flow is due to combination of bulk fluid motion and
buoyancy effects of vapour bubbles.

10.3.1 Two-Phase Flow

Internal, forced convection boiling is referred as two-phase flow and is associated with
bubble growth at inner surface of heated tube through which liquid flows. The boiling
flow goes through different two-phase flow regions.
1) Let us consider a vertical tube that is subjected to a constant surface heat flux, and
where laminar sub-cooled liquid flow enters initially in single-phase forced convection
mode.

Figure 10-5 Flow regimes for forced convection boiling in a tube

2) Vaporization is initiated in sub-cooled flow boiling region, where wall temperature

exceeds saturation temperature. Radial temperature differences are significant as
bubbles are adjacent to heated wall and sub-cooled liquid. Thickness of bubble region
increases with downstream of flow and eventually core of liquid reaches saturation
temperature of fluid, which is point of beginning of saturated flow boiling.

119

3) Saturated flow boiling

In saturated flow boiling region mean velocity of fluid increases remarkably due to
increase of vapour mass fraction.
First stage is known as bubbly flow regime.
2nd stage, slug-flow regime has bubble coalescence forming vapour slugs.
3rd stage, annular-flow regime has liquid film on surface and vapour flowing at
larger velocity in the core of tube
Mist Regime: as liquid film dries eventually at transition zone, vapour carries
remaining liquid in droplets with high velocity at core of tube.
4) Vapour forced convection
When liquid droplets are completely evaporated flow regime changes to vapour forced
convection. Here superheated vapour is in single-phase flow mode having significantly
higher velocity than liquid entering to tube.
Figure 10-5 shows heat transfer coefficient behaviour at different flow regimes of twophase flow. At sub-cooled boiling region heat transfer coefficient increase is around
order of magnitude and smallest heat transfer coefficients exist at vapour forced
convection regime. Following correlation has been defined for saturated flow boiling
region in smooth circular tubes:
0.7

h
hsp

0.1

0.6683

0.16

(1 X ) 0.64 f ( Fr ) 1058

"
S

(1 X ) 0.8 Gs , f (10-9)

m" h fg

Stratification parameter f(Fr) is unity for vertical tubes and horizontal tubes of Fr<0.04.
For horizontal tubes having Fr>0.04 f(Fr) =2.63Fr0.3. Single-phase convection
coefficient hsp is associated with liquid forced convection and is defined at saturation
temperature.
Following is valid for liquid phase Froude number
2

Fr

"
s

m" h fg

1
gD

(10-10)

Table 10-2 Coefficient Gs,f values for different surface-liquid combinations

120

Mean vapour mass fraction can be solved from equation

X ( x)

q S" Dx

(10-11)

mh fg
x Distance from point, where vapour mass fraction exceeds zero

10.4 PRINCIPLES OF CONDENSATION

Reduction of vapour temperature below its saturation temperature results as
condensation. In industrial applications condensation occurs commonly in contact
between cool surface and vapour. Vapour releases it latent energy, which is transferred
to cool surface and condensate is simultaneously formed.

Figure 10-6 Modes of Condensation a) film condensation b) droplet condensation on a surface c)

homogenous condensation or fog formation d) Direct contact condensation

10.4.1 Surface Condensation

Film Condensation
Condensate liquid forms a film flowing across cooling surface. Condensation latent
heat conducts through liquid film to cooling surface from liquid film surface. Film
condensation is most usual type of condensation in industrial equipments.
Drop-wise Condensation
Condensate forms droplets on cool surfaces cracks, pits and cavities. Size of the
droplets covering typically 90% of surface varies from few micrometers to
agglomerations visible to the naked eye. Drop-wise condensation mode conditions may
be created with surface coatings that inhibit wetting.

121

10.4.2 Homogenous Condensation

Vapour condensates homogenously to liquid droplets with decreasing temperature.
Liquid droplets are in form of fog. Temperature change can be caused for example
velocity-pressure change of vapour.

10.4.3 Direct Condensation

Condensation of vapour occurs by cooling in a cold liquid, where vapour is sprayed.
Other option is to spray liquid coolant to vapour, which causes vapour condensation to
surface of liquid droplets.

10.4.4 Condensers
Condensate forms thermal resistance to heat transfer between vapour and cool surface.
Condensate layer thickness increases to direction of flow and heat resistance increases
with thickness. Thus short vertical surfaces or horizontal cylinders are preferred in
design of condensers. Typical condenser is formed from horizontal tube bundles
through which liquid coolant flows and around where vapour condensates.
For performance of condenser in terms of heat transfer rate and condensation, dropwise condensation mode is preferable instead of film condensation operation. An order
of magnitude higher heat transfer rates are achievable with drop-wise condensation
compared to film condensation. In practice, condenser design calculations are done for
film condensation mode, because difficulty of maintaining drop-wise condensation
conditions.

10.5 FILM CONDENSATION

Total condensation heat transfer rate to cooling surface can be written in form of
Newtons Law of Cooling:
q

h L A(Tsat Ts )

(10-12)

Condensation mass flow rate can be then expressed with heat transfer rate as
m

q
h 'fg

h L A(Tsat
h 'fg
h 'fg

Ts )

(10-13)

h fg (1 0.68 Ja)

Modified latent heat

Dimensionless parameter, Jakob Number, describes the ratio of maximum sensible

energy absorbed by the liquid to latent energy absorbed by liquid during condensation
Ja

c p (Tsat Ts )

(10-14)

h fg

122

10.5.1 Laminar film condensation on a vertical plate

Let us consider laminar condensation film formation on a vertical plate. Film flows
downward influenced by the gravity with increasing condensate film thickness and
mass flow rate caused by continuous vapour condensation. Liquid-droplet interface is
approximately at saturation temperature. Thermal boundary layer develops between
cooling surface and vapour-liquid interface. Velocity gradient in both vapour and liquid
is formed due shear stress forces. (Figure 10-7 a)

Figure 10-7 laminar film condensation boundary layers on a vertical plate

Analytical solution for laminar film condensation on a vertical plate is presented by

Nusselt and it contains a set of approximations and assumptions (Figure 10-7):
Constant properties in laminar flow
Pure vapour at uniform saturation temperature
Shear stress negligible in liquid-vapour interface: velocity gradient opposite to surface at film
vapour interface is zero
Momentum and energy transfer by advection of liquid in condensate film is assumed to be
negligible, because of the low flow velocity of film: Heat transfer is only due to conduction
from liquid-vapour surface to cooling surface and results a linear temperature profile in
condensate film.

123

Newtons second law of motion can be then written x- momentum equation as

u
u
v
x
y

1 p
x

By stating pressure gradient dp / dx

2

(10-15)

y2

g , momentum equation reduces to

(10-16)

Velocity profile can be obtained by integrating as

u( y)

g(

1 y
2

(10-17)

Condensate mass flow rate per unit width b is then

( x)

( x)

m
b

g l(

u ( y )dy

(10-18)

For thickness of condensate layer following expression can be obtained

1/ 4

4k l l (Tsat
g l( l

( x)

Ts ) x
'
v ) h fg

(10-19)

Local heat transfer coefficient can be written with liquid conductivity and condensate
layer thickness as

hx

g l(

kl

l
l

(Tsat

) k l3 h 'fg

1/ 4

(10-20)

Ts ) x

Average heat transfer coefficient can be solved by integrating over length of plate as

hL

0.943

g l(
l

(Tsat

) k l3 h 'fg

1/ 4

(10-21)

Ts ) L

Average Nusselt number can thus be written as

Nu L

hLL
kl

0.943

g l(
l

k l (Tsat

124

) h 'fg L3
Ts )

1/ 4

(10-22)

10.5.2 Turbulent Film Condensation

Reynolds number of sliding water film on a vertical plate is defined as

Re

4 l um

4
l

(10-23)

Flow conditions in film condensation on vertical plate are laminar and wave free if
Re
30 . Reynolds number for laminar wave free condensation film can be further
modified by substituting laminar condensate film flow and thickness equations
described above to Re definition as

Re

4g l (

v)

2
l

4/3

k L (T
T)
3.78 l ' sat2 s1 / 3
l h fg (v l / g )

(10-24)

By combining definition of Reynolds number and condensate mass flow rate equation
for condensation heat transfer coefficient can be achieved as
hL

Re

h 'fg

(10-25)

4 L Tsat Ts

Transition to laminar and wavy film flow occurs after Reynolds number exceeds 30.
Moreover turbulent conditions are reached with Reynolds number exceeding
approximately 1800.

Figure 10-8 Film condensation on vertical plate: a) condensate flow rate b) Flow regimes

From following correlations Reynolds number can be solved in wavy laminar (

30 Re 1800 ) flow or at turbulent flow ( 1800 Re ):
0.82

Re

3.7 k l L (Tsat Ts )
'
2
1/ 3
l h fg (v l / g )

4.8

125

(10-26)

4/3

Re

0.069k l L (Tsat Ts ) 0.5

Prl 151Prl0.5 253
'
2
1/ 3
h
v
g
(
/
)
l fg
l

(10-27)

Figure 10-9 Modified Nusselt number for condensation on a vertical plate

10.5.3 Radial Systems

Film Condensation on Outer Surfaces
Typical condenser design tube arrangement is vertical tier of horizontal tubes (Figure
10-10d). Heat transfer coefficient on outer surface reduces from the first tube row to
lower tube row levels caused by increased thickness of condensate film layer.

Figure 10-10 Film condensation flow on a) sphere b) single horizontal tube c) vertical tier of horizontal
tubes with continuous condensate sheet and d) with dripping condensate

126

Extension of Nusselt analysis to spheres and horizontal tubes outer surface laminar
film condensation leads to solution of average heat transfer coefficient as

hD

g l(
l

) k l3 h 'fg

(Tsat

1/ 4

(10-28)

Ts ) D

C = 0.826
C = 0.729

Sphere
Tube

Average heat transfer coefficient for vertical tier of N horizontal tubes can be solved
from
h D,N

g l(

0.729

N l (Tsat

) k l3 h 'fg

1/ 4

(10-29)

Ts ) D

Film Condensation in horizontal Tubes

Refrigeration and air-conditioner condensers typically have vapour condensing inside
horizontal or vertical tubes. Film condensation conditions at inner surface of tube
depend strongly on vapour velocity. With low velocities condensate flows from upper
portion of the tube to the bottom, where it flows to longitudinal direction pushed by
vapour flow shear forces.

Figure 10-11 Film condensation in a horizontal tube a) low vapour velocities b) large vapour velocities

One low velocity criteria presented in form of Reynolds number is

Re v ,i

v m ,v

35000

(10-30)

And related recommended average heat transfer coefficient is written as

hD

0.555

g l(
l

(Tsat

) k l3 h 'fg

Ts ) D

Where modified latent heat is determined as

3
h 'fg h fg
c p ,l (Tsat Ts )
8

127

1/ 4

(10-31)

(10-32)

10.6 DROP-WISE CONDENSATION

Drop-wise condensation is characterized by formation of droplets of varying size on
surface and subsequent growth followed by slide down on cooling surface when they
achieve critical size. Drop-wise condensation is one of the most efficient heat transfer
mechanisms and an order of a magnitude higher heat transfer coefficients are typical
compared to film condensation.
A simple heat transfer coefficient correlation for drop-wise condensation of steam on
copper surfaces is presented as
hdropwise

51100 2044Tsat

22 C Tsat

255310

Tsat

100 C

100 C

(10-33)

In heat exchanger applications having enhancement of heat transfer by promoting dropwise condensation, thermal resistance associated with other heat transfer mechanisms
may remarkably larger.
Example XIV: Drop-wise condensation Cool Koff Can in shower room
A beverage can with a diameter of 65 mm and a height of 120 mm had a uniform temperature of 5 C
when it is removed from a refrigerator. The can is set on a table in a room with an ambient air
temperature of 32 C and a relative humidity of 75 %. Estimate the condensation rate assuming Dropwise condensation.
3 ASSUMPTIONS:
(1) Condensation on top and bottom surface of
can neglected,
(2) Negligible non-condensibles in water
vapor-air,
(3) For film condensation, film thickness is
small compared to diameter of can.

PROPERTIES: Table A-6,

Water vapor ( = 32 C = 305 K):
= 0,04712 bar;
Water vapor (
= 0,04712 bar)
300 K = 27 C,
) 2 = 289 K)
Water liquid ( = (
= 4185 J/kgK.

= 2438 kJ/kg ;

From Eq. 10-13:

m=
Where

, with Ja =

)(

=
)

[1 + 0,68 Ja]

,n (Eq. 10-14):

= 2438 kJ/kg
= 2501 kJ/kg

1 + 0,68

4185 J/kgK (300 278)K

2438 kJ/kg

128

is the saturation temperature of the water vapor in air at 32 C having a relative humidity of
= 0,75.) For drop-wise condensation, the correlation of Eq. 10.33 with
= 300 K = 27 C yields.
h=h

= 51104 + 2044

h , [W/m2K]

22 C <

100 C

, [C]

= 51104 + 2044 27 = 106292 W/m K

Hence, the condensation rate is

W
( 0,065 m 0,125 m)(27
106292
m
K
m=
2501 kJ/kg

5)K

= 0,0229 kg/s

COMMENTS:
1) With film condensation mode cooling of can would take much more time, condensation rate
circa 20 times less

Example XV: Condenser of a steam power plant

The condenser of a steam power plant consists of AISI 302 stainless steel tubes ( = 15 W/mK), each of
outer and inner diameters
= 30 mm and
= 26 mm, respectively. Saturated steam at 0.135 bar
condenses on the outer surface of a tube, while water at a mean temperature of
= 290 K is in fully
developed flow through the tube.
a) For a water flow rate of = 0.25 kg/s, what is the outer surface temperature
of the tube
and the rates of heat transfer and steam condensation per unit tube length? (As a first estimate,
evaluate the properties of the liquid film at the saturation temperature.) If one wishes to increase
the transfer rates, what is the limiting factor that should be addressed?
b) Explore the effect of the water flow rate on
and the rate of heat transfer per unit length.

ASSUMPTIONS: (1) Negligible concentration of noncondensible gases in the steam. (2) Uniform tube
surface temperature, (3) Laminar film condensation, (4) Fully-developed internal flow, (5) Constant
properties.
PROPERTIES: Table A-6, Water ( = 290 K): = 0,00108 Ns/ m , = 0,598 W/mK, Pr = 7,56;
Saturated water, vapour ( = 0,135 bar,
= 325 K):
= 0,0904 kg/ m ,
= 2378 kJ/kg ;
Saturated water, liquid (
):
= 987kg/ m ,
= 4182 J/kgK, = 528 10 Ns/ m ,
= 0,645 W/mK

ANALYSIS: From the thermal circuit, the heat rate may be expressed as
=

(1)

129

Conductive heat resistance:

)

=(

Convective heat resistance:

Reynolds number: Re =

= 0,00152 mK/W

(2)

(3)
/

= 11336

(Turbulent)

Appropriate correlation for the Nusselt number is the Dittius-Boelter correlation, Eq. 8-23:
Nu = 0,023Re
=

Pr

0,023Re

Pr

Substituting numerical values to Eq. 3:

=

0,023Re
=

Pr

0,598 W/mK 0,023 11336

7,56

= 0,00588 mK/W

The resistance associated with the condensate film:

=(

(4)

Where the average convection coefficient for laminar film condensation on a horizontal (with
= 0,729) tube is:
h

h = 0,729

(5)

+ 0,68

(6)

The unknown surface temperature may be determined from an additional rate equation, such as
=

(7)

Iteration from equations (4), (5), (6) and (7), with

=
2449 kJ/kg

h =
25,9 kJ/kg

= 0,41 mK/W

130

= 300 K as a starting value, gives:

= 4481 W/m

= 323 K

10.7 LEARNING OUTCOMES

Table 10-3 Learning Outcomes: Boiling & Condensation

Level of
Knowledge
Apply,
Understand,
describe

Concept
Boiling
Distinguish pool boiling regimes
Concepts: Boiling curve, critical heat flux, Leidenfrost point,
ONB, excess temperature
Heat transfer correlation under different pool boiling regimes
Stages of forced convection boiling4

Understand,
describe

Condensation
Modes of condensation: film, droplet, homogeneous, direct
Film condensation flow regimes: laminar wave-free, Wavy
laminar, turbulent
Heat transfer correlation h calculation: (1) under different film

Apply,
Understand,
describe

condensation regimes (2)at tube internal and external surfaces (3)at drop-wise
condensation mode

Concepts: Jacob number, Newtons Law of cooling and

condensation rate

Not currently included to content of Course: Fundamentals of Heat Transfer

131

11 HEAT EXCHANGERS
11.1 HEAT EXCHANGER TYPES
Heat Exchangers involve heat exchange between two fluids (T1 T2) separated by a
solid wall and encompass a wide range of flow configurations. Classification is
commonly done according to flow arrangement and type of construction. This chapter
discusses only on regenerators, heat exchangers where heat transfer interlinked with
cyclic movement of heat transfer surface between cold and hot streams example
configurations and definitions of heat exchangers are given in Table 11-1.
Table 11-1 Recuperative air-preheater and regenerative shell-and-tube U-tube exchangers
Recuperators
Heat exchangers where two fluids are separated
by the heat transfer surface that normally has no
moving parts

Regenerators
Heat exchangers where the heat transfer is
performed through a material that is alternately
brought to contact with hot and cold streams,
storing energy from the hot stream and releasing it
into the cold one

Introduction to thermal design of heat exchangers is presented in this lecture note and
more comprehensive design guidelines can be found for example in Heat Exchanger
Thermal Design Guide (Saari J., 2010)

11.1.1

Concentric tube heat exchanger

Most simple and for many small applications cheapest configuration is concentric heat
exchangers, which can be constructed with two flow-arrangements: parallel and
counterflow.

Figure 11-1 Parallel and counterflow concentric tube heat exchangers

132

11.1.2

Cross-flow heat exchangers

Other flow-arrangement type is cross-flow heat exchangers (Figure 11-2), where hot
fluid flow is perpendicular to cold one.
For cross-flow over the tubes, fluid motion, and hence mixing, in the transverse
direction (y) is prevented by the finned tubes and in tube flow by tube walls. Crossflow over tubes without fins is called unmixed flow. Heat exchanger performance is
strongly influenced by mixing and heat exchange surface area. Plates can restrict
mixing, but on the other hand provide additional surface and act as fins for tubes.

Figure 11-2 Cross-flow HXs: (1) finned and both fluids unmixed (2) one fluid mixed and other unmixed

11.1.3

Shell-and-tube heat exchanger

On flow-arrangement point-of-view, Shell-and-tube heat exchangers are mix of

concentric and cross-flow heat exchangers (Figure 11-3): Baffles in shell side are used
to establish a cross-flow and to induce turbulent mixing of the shell-side fluid, both of
which enhance convection. Number of tube and shell passes are varied, e.g. Figure 1-1.

Figure 11-3 Shell-and-tube heat exchanger with one shell pass and one tube pass

Figure 11-4 Shell-and-tube heat exchangers: (1) one shell pass and two tube passes and (2) two shell
passes and four tube passes

133

11.1.4

Compact Heat Exchanger

Compact heat exchangers are characterized by large heat transfer surface areas per unit
volume (>400 m2/m3), small flow passages (Dh < 5mm), and laminar flow. Commonly
they are used for achieving large heat rates per unit volume, particularly when one or
both fluids are gas.
Several surface configurations exist: (1) Flat or circular tubes, (2) plate or circular fins,
(3) flow channels can be constructed from parallel plates, which may include fins or
corrugation. Figure 11-5 shows examples of compact heat exchanger core structures.

Figure 11-5 Compact heat exchanger structures (a) Fin-tube flat tubes, continuous plate fins (b)
Fin-tube: circular tubes, continuous plate fins (c) Fin-tube: circular tubes, circular fins (d)
Plate-fin: single pass) (e) Plate-fin: multipass

11.2 ENERGY BALANCE

Let us consider a parallel flow heat exchanger with cold and hot fluid streams (Figure
11-1). On the basis of 1st law of thermodynamics, the heat transfer rate q must equal the
rate of heat lost by the hot fluid stream and gained by the cold fluid stream:
(11-1)
(11-2)
(11-3)
C = Heat capacity rates [W/K]

134

Figure 11-6 Parallel flow of hot (h) and cold (c) fluid steams

11.2.1

Overall Heat Transfer Coefficient

Heat transfer rate of above described heat exchanger can be also written with equation
similar to Newtons Law of cooling

The central variables in any heat exchanger analysis are the heat transfer rate q [W],
heat transfer area A [m2], heat capacity rates C [W/K], and the overall heat transfer
coefficient U
. General definition for overall heat transfer coefficient U is

For tubular, unfinned heat exchangers overall heat transfer coefficient U can be stated
based on outer (o) and inner (i) convection and fouling resistances as

Typical values of U [W / m2K] for tubular, non-finned heat exchangers (Saari, 2010): *oils, ** water

Fluid 2
Fluid 1
gas, 1 bar
gas, 25 bar
liquid, high
liquid, low
phase change

gas, 1 bar
535
1060
1550
2070
2070

gas,

25 bar

10...60
100400
100400
150500
200500

135

liquid, high
viscosity*
1550
100400
100400
200500
200900

liquid, low
viscosity**
2070
150500
200500
4001700
5002000

phase change
2070
200500
200900
5002000
7002500

Heat transfer surface area A means the contact area between one of the fluids, and the
surface of the wall that separates the fluid5. Conductance of the heat exchanger is
defined as

Table 11-2 Example fouling factors

Example XVI: Overall Heat Transfer Coefficient Annular Finned HX

A heat recovery device involves transferring energy from the hot flue gases passing through an annular
region to pressurized water flowing through the inner tube of the annulus. The inner tube has inner and
outer diameters of 24 and 30 mm and is connected by eight struts to an insulated outer tube of 60-mm
diameter. Each strut is 3 mm thick and is integrally fabricated with the inner tube from carbon steel
(
).
The velocity of water flow through the inner tube is 0,161 kg/s at 300 K while flue gases at 800 K flow
through the annulus, maintaining a convection coefficient of 100 W/m2K on both the struts and the outer
surface of the inner tube. What is the rate of heat transfer per unit length of tube from gas to the water?

ASSUMPTIONS: (1) Steady-state conditions, (2) Constant properties, (3) One-dimensional conduction
in strut, (4) Adiabatic outer surface conduction, (5) Negligible gas-side radiation, (6) Fully-developed
internal flow, (7) Negligible fouling.
PROPERTIES: Table A-6, Water (

):

If the areas are different on each sides of the wall, the larger area is the one to be used in equation (1) as
the heat transfer area. The areas are typically significantly different from each other in the case in tubular
or extended-surface heat exchangers.

136

ln (

Reynolds number:

Re =

ln (30 24)
= 7,10 10
(50 W/mK) 1 m

K/W

= 9990 (Turbulent)

For turbulent flow Dittus-Boelter correlation (Eq. 8-22) gives

=

0,023 Re

= 1883 W m K
)

= (1883 W m K
(

W
0,613 mK

0,024 m

0,024 m 1 m)

The fin efficiency is defined as (Eq. 3-21)

=1

Pr

=8

= 7,043 10

5,83

K W

) = 2(0,015 m 1 m) = 0,03 m

= 2(

0,023 9990

= 8 0,03 m

= 0,31 m

0,03 m 1 m

0,003 m 1 m)

Fin efficiency for rectangular fins (Figure 3-2)

(

Where from table 3-1:

=[
=

100 W m K)

50

tanh (36,5 m
0,015 m)
= 0,911
36,5 m
0,015 m

=1
(

= (

0,03 m
(
0,31 m

W
0,003 m
mK

0,911) = 0,931

= (0,931 100 W m K 0,31 m )

) = (7,043 10

= 23,6 W K (800

= 36,5 m

K W 7,10 10

300)K = 11,8 kW

137

= 0,0347 K W

K/W 0,0347 K W)

= 23,6 W/K

11.2.2

Logarithmic temperature difference

Mean temperature difference (Figure 11-7) is accurately described by mean logarithmic

temperature difference defined as

For parallel flow heat exchanger temperature profiles and logarithmic temperature
difference are according

Parallel flow

Figure 11-7 Logarithmic nature of temperature difference parallel flow heat exchanger

For parallel flow heat exchanger temperature profiles and logarithmic temperature
difference are according

Counter flow

Figure 11-8 Logarithmic nature of temperature difference counter flow heat exchanger

Other than pure parallel or counter flow arrangements correction factor for temperature
difference have to be used as

138

Figure 11-9 and Figure 11-10 presents correction factors F for shell-and-tube and crossflow heat exchangers according inlet and outlet temperatures.

Figure 11-9 Correction factors (F) for Shell-and-tube heat exchangers

Figure 11-10 Correction factors (F) for Shell-and-tube heat exchangers

139

11.3 LMTD METHOD

11.3.1

Solving a sizing problem

1) Out of heat transfer rate q and fluid outlet temperatures T x,o, determine those
that are unknown on the basis of those that are known energy balance

2) Calculate logarithmic temperature difference

=

ln(

3) Determine F from graph or equation for the type of heat exchanger flow
arrangement used (Figure 11-9, Figure 11-10)

4) Calculate required heat transfer area A from

a. The heat transfer rate
q = U A Tm
b. and from the overall heat transfer coefficient U
1

11.3.2

Performance Calculation

If the LMTD method is used in performance calculations for which both outlet
temperatures must be determined from knowledge of the inlet temperatures, the
solution procedure is iterative.

Example XVII: LMTD Method Counter Flow HX with hot oil

A counter flow, concentric tube heat exchanger is designed to heat water from 20 C to 80 C using hot
oil, which is supplied to the annulus at 160 C and discharged at 140 C. The thin-walled inner tube has a
diameter of = 20 mm, and the overall heat transfer coefficient is 500 W/m 2K. The design conditions
calls for a total heat transfer rate of 3000 W.
a) What is the length of the heat exchanger?
b) After 3 years of operation, performance is degraded by fouling on the waterside of the
exchanger, and the water outlet temperature is only 65 C for the same fluid flow rates and inlet
temperatures. What are the corresponding values of the heat transfer rate, the outlet temperature
?
of the oil, the overall hat transfer coefficient, and the waterside fouling factor,

140

3 ASSUMPTIONS:
(1) Negligible heat loss to the surroundings
(2) Negligible tube wall conduction
resistance
(3) Constant properties.

ANALYSIS:
a) Log mean temperature difference
Where (for counter flow exchanger)
= (160

= (140

=
From

b) With
conditions.

Hence

80) C = 80 C

20)C = 120 C

120
80 C
= 98,7 C
ln (120 C 80 C)
=(

)
=

3000 W
= 0,968 m
0,020 m) 500 W m K 98,7C

, the following ratio may be formed in terms of the design and 3 year
=

= 1,333

1,333 = 3000 W 1,333 = 2250 W

Having determined the ratio of heat rates, it follows that

=

=
)

= 160

20 C
160

20 C
= 145 C
1,333

Log mean temperature difference is then

=

125
95 C
= 109,3 C
ln (125 C 95 C)

141

= 1,333

=
With

(
+

= 338 W m K

and
1

1
338

1
m K W = 9,59 10
500

m K W

COMMENTS:
1) Would you wait for another three years or plan maintenance clean-up?

11.4

FFECTIVENESS -NTU METHOD

The effectiveness-NTU method may be used without iteration for both design and
performance calculations.
Heat exchanger effectiveness is defined as the ratio of actual heat transfer rate in the
heat exchanger to the maximum heat transfer rate
=

(11-4)

Maximum heat transfer rate in heat exchanger is defined with temperature difference of
inlet flows and minimum heat capacity rate as
(11-5)

Hence, effectiveness can be also written with heat transfer rate definitions as follows
=

(11-6)

Total heat duty of heat exchanger is then simply

(11-7)

11.4.1

How to obtain effectiveness? -NTU Relationships

Defined and correlated functional relationships between effectiveness , heat capacity

ratio Cr and NTU are basis of -NTU method.
Effectiveness
=

Heat capacity ratio

142

Number of Transfer Units

Following tables shows effectiveness and NTU relations for different flow
arrangements as correlations and charts. If one of the fluids experiences a phase change
h, then the temperature of that fluid stream does not change, and C of that stream is
infinity. Thus it follows that, any finite C of the other stream therefore becomes the
Cmin, and the heat capacity rate ratio will be 0.
Table 11-3 Effectiveness relations

143

Table 11-4 Effectiveness of crossflow of heat exchangers: (a) both mixed and (b) other mixed and other
unmixed.

Table 11-5 NTU Relations

144

Table 11-6 Effectiveness of parallel, counterflow and two configurations of shell-and-tube heat
exchangers

145

11.4.2

Solving a sizing problem

Defining heat transfer surface area in design phase would follow four steps6:
1) Solve Heat Capacity Ratio C* and Efficiency
=

2) Using -NTU relationships of figures (or equations) find NTU

3) Solve required UA
=

4) Solve or estimate heat transfer area

1

11.4.3

Perf romance calculation

Steps for defining heat transfer rate and outlet temperatures in performance calculation
would be following:
1) Solve Overall heat transfer coefficient U and heat transfer area A
1

2) Using -NTU NTU

3) Solve efficiency of heat exchanger

(

4) Solve performance: heat transfer rate q and outlet temperatures

=

Depending on known variables, steps could be in other order also

146

Example XVIII: -NTU Method Double Pass shell-and-tube HX

A shell-and-tube heat exchanger consists of 135 thin-walled tubes in a double-pass arrangement, each of
12,5-mm diameter with a total surface area of 47,5 m2. Water (the tube-side fluid) enters the heat
exchanger at 15 C and 6,5 kg/s and is heated by exhaust gas entering at 200 C and 5 kg/s. The gas may
be assumed to have the properties of atmospheric air, and the overall heat transfer coefficient is
approximately 200 W/m2K.
a) What are the gas and water outlet temperatures?
b) Assuming fully developed flow, what is the tube-side convection coefficient?
c) With all other conditions remaining the same, plot the effectiveness and fluid outlet
temperatures as a function of the water flow rate over the range from 6 to 12 kg/s.
d) What gas inlet temperature is required for the exchanger to supply 10 kg/s of hot water at an
outlet temperature of 42 C, all other conditions remaining the same? What is the effectiveness
for this operating condition?
ASSUMPTIONS:
(1) Negligible heat lost to
surroundings,
(2) Fully developed conditions for
internal flow of water in tubes
(3) Exhaust gas properties are those
of air,
(4) The overall coefficient remains
unchanged for the operating
conditions examined.

PROPERTIES: Table A-6, Water (

300 K): = 997 kg/ m ,
mK, = 855 10 Ns/ m , Pr = 5,83; Table A-4, Air (1 atm,
= 1014 J/kgK

= 4179 J/kgK,
= 0,613 W/
= 400 K): = 0,8711 kg/ m ,

ANALYSIS:
a) Using the -NTU method, first find the capacity rates,
= 6,5
= 5,0
Recognizing that
=

J
kg
4179
= 27,164 W K
kgK
s

kg
J
1014
= 5,070 W K
s
kgK
=

and

= 0,19

200 W m K 47,5 m
= 5,070 W/K
5,070 W K

From table 11-16 for the shell and tube exchanger, with
0,78
=

200
200

147

15 C

= 5,070 W/K and

= 0,78

= 0,19

= 55,7 C

From energy balances of the two fluids,

+(

= 15 C + 0,19 (200

b) Reynolds number for estimating

Re

55,7)C = 42,4 C

for water

= 5736 (Turbulent)

From Dittus-Boelter correlation (Eq. 8-22)

=

c)

0,023 Re

Pr

W
0,613 mK

12,5 10

0,023 5736

5,83

= 2320 W m K

Using the IHT Heat Exchanger Tool, Shell and Tube, One-shell pass and N-tube passes, and the
prescribed properties, following the analysis of (a):

d) Using the above IHT model, the hot inlet temperature

, required to provide
with
= 42 C and the effectiveness for this condition are
= 74,4 C
= 0,55

COMMENTS:
1) for part (d) with m _h doubling, we should expect U to increase?

148

= 10 kg/s

11.5 LEARNING OUTCOMES

Table 11-7 Learning Outcomes: Heat Exchangers

Level of
Knowledge
Understand
Describe

Concept

Types and thermal principles: concentric, counter, parallel, cross

flow, shell-and-tube, plate & compact, etc.

Concepts: log-mean temperature difference, overall heat

transfer coefficient, energy balance
Apply
understand
describe

Two main thermal performance and design methods:

Logarithmic mean temperature difference and efficiency NTU method

149

12 RADIATION
Stefan-Boltzmanns law states heat flux emitted by black body

Stefan-Boltzmann-vakio 5,67 10 [
Temperature of black body [ ]

(12-1)

(12-2)

Radiation emitted by real surface can be written as

Emissivity

[ ]

Radiation net heat flux to real surface may be written according Kirchoffs law as
follows
( )

Absorptivity
( ) Blackbody radiation flux [
]
Irradiation = radiation flux from surroundings [

(12-3)

Three equations above are very fundamental tools for simple engineering radiation
calculations. They consists assumptions and thermal radiation concepts that are
beneficial to understand for being aware of applicability of basic radiation equations
and conditions requiring more detailed evaluation. Clarification to those radiation
principles is given in this chapter.

12.1 PRINCIPLES
All bodies constantly emit energy by process of electromagnetic radiation. The
intensity of energy flux depends upon (1) the temperature of the body and (2) the nature
of its surface. Radiation has dual nature: (1) particles, which are photons or quanta
(Planck) and (2) electromagnetic waves.
In all cases, radiation is characterized by a wavelength and frequency f, which are
related through the speed at which radiation propagates in the medium of interest
(according Maxwell):
(12-4)
c=2.998*108 m/s speed of light in vacuum

Thermal radiation is confined to the infrared, visible and ultraviolet regions of the
spectrum of electromagnetic radiation (Figure 12-1). It is originated from oscillations
and transitions of electrons of material at an absolute temperature T > 0K.

150

Figure 12-1 Spectrum of electromagnetic radiation

12.1.1

Blackbody

Black body is an idealization providing limits on radiation emission and absorption by

matter.
1) A blackbody is an Ideal emitter.
For a prescribed temperature and wavelength, no surface can emit more radiation than
a blackbody

2) A blackbody is a diffuse emitter.

Although the radiation emitted by a blackbody is a function of wavelength and
temperature, it is independent of direction

3) A blackbody absorbs all incident radiation: the ideal absorber

Regardless of wavelength and direction

The spectral (Planck) distribution of the blackbody emissive power (determined

theoretically and confirmed experimentally, Figure 12-2) is

(12-5)
C1 = 3.742*108 m4/m2 Radiation constant
C2 = 1.439*104 mK Radiation constant

Wiens displacement law states relationship between wavelengths of maximum

emissive power and temperature

(12-6)
C3 = 2898 mK Radiation constant

151

Figure 12-2 Spectral emissive power of blackbody

Some fundamental characteristics of spectral emissive power can be noted as

1) At any wavelength magnitude of emitted radiation increases with increasing T
2) The fractional amount of total blackbody emission appearing at lower wavelengths
increases with increasing T
3) T < 800K: emission is in the infrared region of the spectrum and is not visible to the
eye
4) Emission from the sun may be approximated as that from a blackbody at 5800K, and
significant fraction of radiation is in visible region of spectrum

12.1.2

Stefan-Boltzmann Law & Band Emission

Stefan-Boltzmann Law is the total emissive power of a blackbody obtained by

integrating the Planck distribution over all possible wavelengths
(12-7)
= 5.67 * 108 W/m2K4 Stefan-Boltzmann constant

152

Blackbody functions describe fraction of emission from blackbody at certain

wavelength interval. Fraction of blackbody emission in interval from 0 to (Figure
12-3) is only function of T as stated by

(12-8)

Thus for interval from

blackbody functions as

to

fraction of blackbody emission is given by difference of

(12-9)

Figure 12-3 Band emission fraction of emission

Figure 12-4 Fraction of Blackbody emission

in described wavelength interval (/band): 0-

function in spectral band (0- ) as function of T

Table 12-1 gives values for Blackbody functions with T.

153

Table 12-1 Blackbody radiation functions

154

12.1.3

Surface Response to Irradiation

Spectral irradiation, radiation incident to surface from surroundings with all

wavelengths, may act with three different ways (Figure 12-5) while facing
semitransparent surface
Reflection from the medium

Spectral reflectivity

Absorption within the medium

Spectral absorptivity

Transmission through the medium

Spectral transmissivity

Volumetric radiation balance for semitransparent medium is then spectrally

(12-10)

This can be further reduced to sum of spectral radiation properties

=1

(12-11)

For total, over wavelength distribution averaged properties, volumetric balance is

similarly
G=

(12-12)

=1

(12-13)

155

Figure 12-5 Semitransparent surface response to irradiation

For opaque material, material for which transmission is negligee (Gtr=0) and
irradiation response is defined in fraction of micrometer from incident surface,
irradiation is surface phenomenon and balances are written as
(12-14)
(12-15)

12.1.4

Kirchoffs Law

Kirchhoffs law equates the total, hemispherical emissivity of a surface to its total,
hemispherical absorptivity:
=

(12-16)

Derivation of Kirchoffs law requires strict conditions associated to be fulfilled:

Irradiation of the surface corresponds to emission from a blackbody at
the same temperature as the surface.
With this formulation, net radiation balance (Figure 12-6) for opaque material can be
stated as

(12-17)

156

Figure 12-6 Surface radiation balance for opaque material

However, for spectral, directional properties formulation can be applied without

restriction:
(12-18)

12.2 RADIATION QUANTITIES

12.2.1

Radiation Intensity & Emission

Let us first define Spectral Intensity

(12-19)
rate at which radiant energy is emitted at wavelength in the ( ) direction per Area of
emitting surface normal to this direction, per unit solid angle about this direction, and
per unit wavelength interval d about .

Figure 12-7 (a) Projected Area dA1 and (b) Solid angle

Solid angle is defined as

(12-20)

157

For complete hemisphere integration results as

=

sin

=2

[sr]

(12-21)

Figure 12-8 Solid angle subtended by dAn at point dA1

Good approximation for projected surface area is

for those surfaces
having radiation intensity independent from direction. These surfaces are called diffuse,
and such radiation isotropic.

Figure 12-9 (a) Diffuse radiation from Blackbody (left) and directional emission from real surface (right)
(b) Directional distribution

Total and spectral Emissive powers are

=

(12-22)
]

(12-23)

And if intensity is independent on direction (diffuse surface)

[

[W/m m]

158

(12-24)
(12-25)

12.2.2

From Irradiation and Emission to Radiosity

Irradiation is radiation incident to surface from all directions (Figure 12-10)

(12-26)

If radiation is approximated to originate from large surroundings (e.g. sky, furnace/

room walls) acting like blackbody it can be written
(12-27)

Figure 12-10 Radiosity from surface

The radiosity of an opaque surface accounts for all of the radiation leaving the surface
in all directions and may include contributions from both reflection and emission. For
radiosity it can be written as

(12-28)

If we consider these two terms, irradiation and radiosity, with definition of intensity, it
can be written first for spectral irradiation
(12-29)

and total irradiation

(12-30)

Similarly for radiosity, first spectral

(12-31)

, and then total radiosity as

(12-32)

159

Example IXX: Diffuse emitter Radiation quantities

Determine the fraction of the total, hemispherical emissive power that leaves a diffuse surface in the
directions /4
/2 and
.

ASSUMPTIONS: (1) Diffuse emitting surface.

ANALYSIS: the total, hemispherical emissive power is
=

)cos sin

For a diffuse surface Ie , , ) is independent of direction, and as given by Eq. 12-27,

=

The emissive power, which has directions prescribed by the limits on and , is
=

cos sin

It follows that

sin
2

[ ]

=
= 0,25
=

0,25

sin

/2

sin

/4

= 0,25

Example XX: Hot 5-mm-square Radiation quantities

Consider a 5-mm-square, diffuse surface Ao having a total emissive power of E0 = 4000 W/m2. The
radiation field due to emission into the hemispherical space above the surface is diffuse, thereby
providing a uniform intensity I( , ). Moreover, if the space is a nonparticipating medium (nonabsorbing, non-scattering, and non-emitting), the intensity is independent of radius for any ( , )
direction. Hence intensities at any points P1 and P2 would be equal.
a) What is the rate at which radiant energy is emitted by Ao, qemit?
b) What is the intensity Io,e of the radiation field emitted from the surface Ao?
c)

Beginning with Equation

)=

cos sin

and presuming knowledge of

the intensity Io,e, obtain an expression for qemit.

d) Consider the hemispherical surface located at r = R1 = 0,5 m. Using the conservation of energy
requirement, determine the rate at which radiant energy is incident on this surface due to
emission from Ao.

160

e)

Using Equation 12.6 (Incr.),, determine the rate at which radiant energy leaving Ao is
intercepted by the small area A2 located in the direction (45, ) on the hemispherical surface.
What is the irradiation on A2?
f) Repeat part (e) for the location (0, ). Are the irradiations at the two locations equal?
g) Using Equations 12.14, determine the irradiation G1 on the hemispherical surface at r = R1.
3 ASSUMPTIONS:
(1) Diffuse surface, Ao
(2) Medium above Ao is
almost non-participating,
(3)
.

ANALYSIS:
a) The radiant power leaving Ao by emission is
= 4000

W
(0,005 m 0,005 m) = 0,10 W
m

b) The emitted intensity is Io,e and is independent of direction since Ao is diffuse emitter.
=

= 1333 W/m

sr

The intensities at points P1 and P2 are also Io,e and the intensity in the directions shown in
the schematic above will remain constant no matter how far the point is from the surface
Ao since the space is nonparticipating.
c)

From knowledge of Io,e, the radiant power leaving Ao

=

cos sin
cos sin

= 0,10 W

d) Defining control surfaces above

through A1.

Ao and on A1, the radiant power leaving

Ao must pass

= 0,10 W

Average irradiation on the hemisphere, A1

e)

= 63,7 mW m

From the radiant power leaving Ao intercepted by A2, where A2 = 410-6 m2, located at (
= 45 , ) as per schematic, follows that

161

cos
=
Where (

cos

,where (

R = 4 10 m

), the direction normal to A2

= 1333

W
sr 25 10
m

= 45 C)

1
= 1,60 10
(0,5 m)

cos 45 1,60 10

sr

sr = 3,77 10

From the definition of irradiation

= 93,3 mW m

f) With A2, located at ( = 0, ) cos = 1, cos

g) Using Eg. 12.15 based upon Figure 12.10
=

=1

=
sin

= 63,7 mW/m
=

and

Example XXI: Sheet steel Radiation quantities

cos

Sheet steel emerging from the hot roll section of a steel mill has following properties: temperature of
1200 K, a thickness of = 3 mm, density of 7900 kg/m3 and specific heat of 640 J/kgK.
a) What is the total hemispherical emissivity?
b) Accounting for emission from both sides of the sheet and neglecting heat transfer from the
surroundings, determine the initial time rate of change of the sheet temperature
).
c) As the steel cools, it oxidizes and its total, hemispherical emissivity increases. The correlation
[1200 K
for the increasing emission is of the form =
(K)]. How long it will take for
the steel to cool from 1200 to 600 K?

ASSUMPTIONS: (1) Negligible conduction (in longitudinal direction), convection and radiation rom
surroundings, (2) Negligible transverse temperature gradients.
PROPERTIES: Steel (given): = 7900 kg/m , = 640 J/kg K, = 1200 i/T (K).
ANALYSIS:
a) The initial total hemispherical emissivity is
=

(1200 K)/

(1200 K)]

and integrating by parts using values from Table 12-1, find

= 1200

= 0,6 0,002 + 0,4(0,819

= 0,002;

= 7200

0,002) + 0,25(

162

0,819) = 0,373

= 0,819

b)

From an energy balance on a unit surface area of strip (top and bottom),
=d

dT
dt
c)

/dt

(1200

)
5,78 K/s

From the energy balance,

=

= d(

)/dt

0,373 5,67 10 W/m K (1200 K)

kg
0,003 m 640 J/kg K
7900
m

dt, =

kg
0,003 m 640 J/kg K
m
=
4800 0,373 5,67 10 W/m K
7900

1
600

1
1200

= 311 s

COMMENTS: Initially, from Eq. 1.9(Incr.),,

= 36.6 W/m K. Assuming a plate width
) ( /2)
.
=
of
= 1 m, the Rayleigh number may be evaluated from
= 1.8 10 . From Eq.
Assuming
= 300 K and evaluating properties at
= 750 K,
9.31(Incr.), Nu = 84, giving = 9.2 W/m K. Hence heat loss by radiation exceeds that associated
with free convection. To check the validity of neglecting transverse temperature gradients, compute
Bi =
/2)/ . With
= 36.6 W/m K and
= 28 W/m K, Bi = 0.002 << 1. Hence the
assumption is valid.

163

12.3 RADIATION PROPERTIES

12.3.1

Surface Emissivity

For most cases, reasonable approximation is that the hemispherical emissivity is equal
to the normal emissivity = n.
Ranges of normal emissivity for some metals and nonconductors are presented in
Figure 12-11. Polished metals have very low emissivity, while emissivity increases
strongly for unpolished and oxidized surfaces. Compared to metals nonconductors have
higher total normal emissivity.

Figure 12-11 Total, normal emissivity

Materials have different spectral emissivity characteristics, e.g.: for metals emissivity
decreases with increasing wavelength and nonmetals have they own spectral behavior
curves.

Figure 12-12 Spectral variation of emissivity

Radiation properties, and thus also emissivities, are dependent on temperature of

material surface. Figure 12-13 shows variety of thermal emissivity behavior.

164

Figure 12-13 Examples of temperature dependence of total, normal emissivity

12.3.2

Surface Reflectivity and Absorptivity

Most of opaque materials presented in Figure 12-14 have a strong dependence of

absortivity, and reflectivity ( =1- ), on wavelength.

Figure 12-14 Spectral normal reflectivity and absorptivity

Surfaces may be divided to two idealized categorizes according the way they reflect
radiation: (1) for diffuse surfaces reflected radiation is independent on reflection angle
and on angle of incident radiation, (2) specular surfaces reflect radiation with equal
angle to incident radiation (Figure 12-15).

165

Figure 12-15 Diffuse and specular reflectivity surfaces

12.3.3

Surface Transmissivity

Figure 12-16 shows spectral transmissivity of few example materials.

Figure 12-16 Spectral transmissivity of selected materials

12.4 RADIATION BETWEEN TWO SURFACES

12.4.1

The View Factor

View factor Fij is defined as fraction of the radiation leaving surface i that is
intercepted by surface j.

(12-33)
(12-34)

Summation rule for view factor is

(closed system)

(12-35)

(between surfaces)

(12-36)

166

, and reciprocity relation

(12-37)

12-17 (a) View factor associated with radiation exchange between two areas and (b) Radiosity J

For Blackbody Ai, (Ji= Ebi) radiation intercepted by surface Aj is

(12-38)

View factors are liste for example cases in Table 12-2.

12.4.2

Two Surface Enclosures

Between two surfaces a single expression for the rate of radiation transfer may be
inferred from a network representation of the exchange (Figure 12-18) as
=(

(12-39)

Figure 12-18 Two surface enclosures

This reduces for special condition for infinite parallel plates to form

167

(12-40)

Figure 12-19 Large (Infinite) Parallel Plates

168

Table 12-2 Example View Factors: 2- and 3-dimensional

169

Example XXII: Ice Hockey Hall Radiation exchange between two bodies
The assessment is to determine the ceiling irradiative properties for an ice rink in terms of ability to
maintain surface temperature above the dew point. The rink has a diameter of D = 50 m and a height and
a height of L = 10 m, and the temperatures of the ice and walls are -5 C and 15 C, respectively. The
rink air temperature is 15 C, relative humidity 70 %, and a convection coefficient of 5 W/m2K
characterizes conditions on the ceiling surface. The thickness and thermal conductivity of the ceiling
insulation are 0.3 m and 0.035 W/m2K, respectively, and the temperature of the outdoor air is -5 C.
Ceiling can be assumed as a diffuse-grey surface and the walls and ice may be approximated as
blackbodies.
a) Consider a flat ceiling having an emissivity of 0,05 (highly reflective panels) or 0,94 (painted
panels). Perform an energy balance on the ceiling to calculate the corresponding values of the
ceiling temperature. Will condensation occur for either or both of the emissivities?
b) For each of the emissivities, calculate and plot the ceiling temperature as a function of the
insulation thickness for
. Identify conditions for which condensation will occur
on the ceiling.

ASSUMPTIONS: (1) Rink comprised of the ice, walls and ceiling approximates a threesurface,diffusegray enclosure, (2) Surfaces have uniform radiosities, (3) Ice surface and walls are black, (4) Panels are
diffuse-gray, and (5) Thermal resistance for convection on the outdoorside of the ceiling is negligible
compared to the conduction resistance of the ceiling insulation.
PROPERTIES: Psychrometric chart (Atmospheric pressure; dry bulb temperature, Tdb = T ,i
= 15 C; relative humidity, RH = 70%): Dew point temperature, Tdp = 9.4 C.
ANALYSIS: Applying an energy balance to the inner surface of the ceiling and treating all
heat rates as energy outflows,

170

(4)
( )

Since the ceiling panels are diffuse-gray,

( )

(5)

. for parallel, coaxial disks

= 0,672

From the summation rule applied to the ice (i) and the reciprocity rule,

=1

(symmetry)

=1

=(

=(

)(

) = 0,410

where
/4 and
Using the foregoing energy balance, Eq. (1), and the rate equations, Eqs. (2-5), the ceiling temperature is
calculated using radiative properties for the two panel types,
Ceiling panel
(C)
Reflective

0,05

14,0

Paint

0,94

8,6

Condensation will occur on the painted panel since

(b) Applying the foregoing model for 0.1

<

1.0 m, the following result is obtained

For the reflective panel ( = 0.05), the ceiling surface temperature is considerably above the
For the reflective panel ( = 0.05), the ceiling surface temperature is considerably above dew point.
Therefore, condensation will not occur for the range of insulation thicknesses. For the painted panel ( =
0.94), the ceiling surface temperature is always below the dew point, and condensation occurs for the
range of insulation thicknesses.
COMMENTS: (Icehockey or figure skating?)
1) From the analysis, recognize that radiative exchange between the ice and the ceiling has the
dominant effect on the ceiling temperature. With the reflective panel, the rate is reduced nearly
20-fold relative to that for the painted panel. With the painted panel ceiling, condensation will
occur for most of the conditions likely to exist in the rink.

171

12.5 LEARNING OUTCOMES

Table 12-3 Learning Outcomes: Radiation

Level of
Knowledge

Concept

Apply
Understand
Describe

Concepts: Stefan-Boltzmanns and Wien displacement Laws

and Black body radiation functions,
Surface/body properties: black body, gray, diffusive, opaque
Two radiation heat transfer solution methods: (1) based on
Kirchoffs Law and (2) Radiation between two bodies

understand
describe

Spectral radiation properties of material surfaces

172

REFERENCES
Bejan A., and Kraus A. D., Heat Transfer Handbook, Hoboken, NJ: John Wiley &
Sons, Inc, 2003, ISBN: 0-471-39015-1.
Holman, J.P., Heat Transfer. 6th edition, McGraw-Hill Book Company Singapore.
Singapore, 1989.
Incropera Frank P, DeWitt David, Fundamentals of Heat and Mass Transfer. 5th edition.
John Wiley & Sons, Inc. United States, 2002.
Lampinen M. J., Assad El M., Kotiaho V., Lmmnsiirto-oppi, teknillinen
korkeakoulu, Otaniemi 2008.
Lienhard J. H. IV, Lienhard J. H. V: A Heat Transfer Textbook, 3rd Edition. 2008.
Luttunen E., Lmmnsiirron yhtlt ja taulukot, 2010.
Saari, J., Heat Exchanger Thermal Design Quide, LUT Energy Lecture Note 6, 2010.
Tynjl T., Sarkomaa P., Lmmnsiirto Luennot, 1999 - 2000.
Yunus A. Cenel, Heat Transfer A Practical Approach
Wendl, Michael C., Fundamentals of Heat Transfer Theory and Applications: 1999,
2003 and version 2.1 on August 2005.

173