CHAPTER 3 Heat Transfer Review

This chapter is intended to review those aspects of heat transfer

that are important in the design and analysis of solar collectors and
systems:

-review of radiation heat transfer

-review of convection correlations for internal flow and wind-
induced flow
-review of conduction heat transfer
-review of radiation characteristics of opaque materials

1 / 50
I/- The electromagnetic spectrum

Thermal radiation is electromagnetic energy that is propagated

through space at the speed of light.

The wavelengths of importance in solar energy and its applications

are in the ultraviolet to near-infrared range, that is, from 0.29 to
approximately 25 μm.

They all travel with the speed of light C and have a frequency ν such
that
C0
C = λν
=
n
where C0 is the speed of light in a vacuum and n is the index of
refraction of the medium.
2 / 50
 Figure 1 The spectrum of electromagnetic radiation

For some purposes in solar energy applications, it is necessary to

consider the energy of a particle or photon, which can be thought
of as an ‘‘energy unit’’ with zero mass and zero charge.

The energy of the photon is given by

E = hν
3 / 50
 The wavelength distribution of radiation emitted by a blackbody of
temperature T is given by Planck’s law

2π hC0 2 C1
=Eλb =
5  hC0   5  C2  
λ exp   -1 λ exp   -1
  λ kT     λT  

E λb is the emissive power of the blackbody

where h is Planck’s constant and k is Boltzmann’s constant
and the constants C1 and C2 are defined as

=C1 2π=
hC0 3.7405 × 10 W .µ m / m
2 8 4 2

4 / 50
 hC0
C2 = 14387.8µ m.K
=
k

Note that a blackbody is a perfect absorber and a perfect emitter.

By differentiating Planck’s distribution and equating to zero, the

wavelength corresponding to the maximum of the distribution can
be derived.

This leads to Wien’s displacement law, which can be written as

λmaxT = 2897.8 µ m.K

5 / 50
In engineering calculations the total energy is often of more
interest. By integrating Planck’s law over all wavelengths, the total
energy emitted per unit area by a blackbody of temperature T is
found to be
∞
= ∫ λb
= λ σ 4
Eb E d T
0

where σ is the Stefan-Boltzmann constant and

σ=5.6697×10 W/m K
-8 2 4

The fraction of radiation emitted by the blackbody of temperature T

between wavelengths 0 and λ can be written as

6 / 50
 λ
∫ Ebλ d λ
λ =
0
f 0− ∞
∫
0
Ebλ d λ

Using the expression of Ebλ this becomes

λ C1λ −5
∫
0   C2   
dλ
λ
 exp   -1 
∫=E λdλ   λT   λT C1 ( λT )
−5

d ( λT ) ≡ f 0−λT
∫
b
=f 0−λ 0
=
∞
σT   
∫ E λdλ  C2
4 0
b σ  exp   -1 
 λT
0
  

The value of this integral is then also the fraction of the blackbody
energy between 0 and λT; these values have been calculated and
tabulated by many authors:
7 / 50
8 / 50
Table 6 Fraction of
Blackbody Radiant
Energy between 0
and λT for Even
Increments of λT

9 / 50
Note that the emissive power E for a gray surface is defined as the
emissive power of a blackbody with the same temperature multiply
by a positive quantity ε less than 1

E = ε Eb
where 0 < ε < 1 is called the emissivity of the gray surface

II/- Radiation intensity and flux

The intensity I of radiation leaving a small surface ∆A in a solid

angle ∆ω normal to the surface is defined as

10 / 50
 ∆E
I = lim
∆A→0 ∆A ∆ω
∆ω → 0

Figure 2: Schematic of radiation flux.

The radiation flux leaving the surface per unit surface is then found
by integrating I over the hemisphere

11 / 50
 2π π 2
q=∫ ∫ I cos θ sin θ dθ dφ
0 0

For surfaces where intensity I is constant over all directions, the

above integral gives

q =πI
For a blackbody which is a diffuse emitter the emissive power Eb is
found from the blackbody intensity Ib by

Eb = π I b
12 / 50
The foregoing equations were written for total radiation but apply
equally well to monochromatic radiation:

Eλb = π I λb
The majority of heat transfer problems in solar energy applications
involve radiation between two surfaces.

For gray surfaces, the radiant heat leaving surface A1 towards

surface A2 is given as

13 / 50
 σ (T24 − T14 )
Q1 =
−Q2 =
1 − ε1 1 1 − ε2
+ +
ε1 A1 A1F12 ε 2 A2
F12 is called the configuration view factor between the 2 surfaces

For the special case where the two surfaces are infinite parallel
plates (as in flat-plate collectors) the areas A1 and A2 are equal and
the view factor F12 is unity. Under these conditions the above
equation becomes
Q σ (T24 − T14 )
=
A 1 1
+ −1
ε1 ε2

14 / 50
The second special case is for a small convex surface 1 surrounded
by a large enclosure, surface 2.
A1
Under these conditions, the area ratio A approaches zero, the
2
view factor F12 is unity, and above equation becomes; note that this
is the heat radiation received by surface A1 and coming from A2;

=Q1 ε1 A1σ (T − T 2
4
1
4
)
the opposite will be the radiation leaving surface A1

To predict the performance of solar collectors, it will be necessary

to evaluate the radiation exchange between a surface and the sky.

15 / 50
The sky can be considered as a blackbody at some equivalent sky
temperature Ts so that the actual net radiation between a horizontal
flat plate and the sky is given by

=Q ε Aσ (T 4 − Ts4 )
where TS is the sky temperature.

To retain the simplicity of linear equations, it is convenient to define

a radiation heat transfer coefficient.

The heat transfer by radiation between two arbitrary surfaces has

been given above.
If we define a heat transfer coefficient hr so that the radiation
exchange between the two surfaces is given by

16 / 50
 =Q A1hr ( T2 − T1 )

then it the heat transfer coefficient is given by

σ (T22 + T12 ) (T2 + T1 )

hr =
1 − ε1 1 (1 − ε 2 ) A1
+ +
ε1 F12 ε 2 A2

III/- Natural convection between flat parallel plates and between

concentric cylinders

17 / 50
The rate of heat transfer between two plates inclined at some angle
to the horizon is of obvious importance in the performance of flat-
plate collectors.

Free-convection heat transfer data are usually correlated in terms

of two or three dimensionless parameters: the Nusselt number Nu,
the Rayleigh number Ra, and the Prandtl number Pr.

Some authors correlate data in terms of the Grashof number, which

is the ratio of the Rayleigh number to the Prandtl number.

The Nusselt Nu, Rayleigh Ra, and Prandtl Pr numbers are given by

18 / 50
Where
h= heat transfer coefficient of the flowing fluid [W/m2 K]
L= plate spacing [m]
k = thermal conductivity of the fluid [W/m K]
g = gravitational constant [m/s2]
β’= volumetric coefficient of expansion [1/K] (for ideal gas, β’=
1/T )
∆T = temperature difference between plates [K]
ν = kinematic viscosity of fluid [m2/s]
19 / 50
 α = thermal diffusivity of fluid [m2/s]

For a natural convection of a fluid between two plates tilted from the
horizontal plane, the convection heat transfer between the two
plates is given empirically by the relation (Note that here the fluid is
moving in an enclosed space)

where the meaning of the + exponent is that only positive values of

the terms in the square brackets are to be used (i.e., use 0 if the term
is negative).

20 / 50
IV/- Heat Transfer Relations for Internal Flow

Heat transfer coefficients for common geometries are given in many

heat transfer books.

For fully developed turbulent liquid flow inside tubes (2300 < Re =
ρVDh/μ < 5 × 106 and 0.5 < Pr < 2000), authors suggest

where n=0.11 for heating and n=0.25 for cooling

the friction factor f of the tube is given as

21 / 50
For gases, the viscosity ratio should be replaced by the temperature
n
 Tw 
ratio  T 
For noncircular tubes the hydraulic diameter can be used for the
characteristic length in the preceding two equations. The hydraulic
diameter is defined Dh as

4 × ( flow area )
Dh =
wetted perimeter

22 / 50
 L
For short tubes with D > 1.0 and a sharp-edged entry, McAdams
recommends that the Nusselt number be calculated from

For short tubes the developing thermal and hydrodynamic

boundary layers will result in local Nusselt numbers for the case of
constant heat rate represented by an equation of the form

where the constants a, b, m, and n are given in the Table below:

23 / 50
 Table 14 Constants for Equation for Calculation of Local Nu for
Circular Tubes with Constant Heat Rate

In the study of solar air heaters and collector-storage walls it is

necessary to know the forced-convection heat transfer coefficient
between two flat plates.

For air the following correlation can be derived from the data of
Kays and Crawford (1980) for fully developed turbulent flow with

24 / 50
one side heated and the other side insulated (Note that here the
heat transfer concerns the fluid and the walls of the channel).

where the characteristic length is the hydraulic diameter.

For the case of parallel plates with constant temperature on one

side and insulated on the other side, Mercer et al. obtained the
following relation for 0.1 < Pr < 10:

25 / 50
V/- Wind convection coefficient

The heat loss from flat plates exposed to outside winds is important
in the study of solar collectors.

Sparrow et al. did wind tunnel studies on rectangular plates at

various orientations and found the following correlation over the
Reynolds number range of 2 × 104 to 9 × 104.

where the characteristic length is four times the plate area divided
by the plate perimeter.

26 / 50
The work of Polhausen suggests that this equation should be valid
for Reynolds numbers up to 106.

A simple relation by McAdams is sometimes used for fast

calculation
=
h 5.7 + 3.8V
where V is in m/s and h is in W/m2.K

It is probable that the effects of free convection and radiation are

included in this equation.
For this reason Watmuff et al. report that this equation should be

=
h 2.8 + 3.0V
27 / 50
VI/- HEAT CONDUCTION

Thermal conduction is the transfer of heat by the vibrations of

atoms, molecules and electrons in solids or in fluids without bulk
movement.

Although conduction also occurs in liquids and gases, heat transfer

in those cases is usually dominated by convection.

The heat flow P by conduction through a slab of material, area A,

thickness ∆x, surface temperature difference ∆T is given by Fourier
law:

28 / 50
 ∆T ∆T
P=
− kA =
− ( in W)
∆x ∆x kA

where k is the thermal conductivity of the solid (unit W.m-1. K-1 ), and
the negative sign indicates that heat flows in the direction of
decreasing temperature.

The thermal resistance of conduction is then given as:

∆x
R=
kA

VII/- Radiation Characteristics of Opaque Materials

29 / 50
For many solar energy calculations only two quantities are
required, the solar absorptance and the long-wave or infrared
emittance, usually referred to as just absorptance and emittance.

1°/- Absorptance and emittance

The monochromatic directional absorptance is a property of a

surface and is defined as the fraction of the incident radiation of
wavelength λ from the direction μ, φ (where μ is the cosine of the
polar angle and φ is the azimuthal angle) that is absorbed by the
surface.
It is given as

30 / 50
where subscripts a and i represent absorbed and incident.

When integrated over all wavelengths, the fraction of all the

radiation from the direction μ, φ that is absorbed by a surface is
called the directional absorptance of the surface and is defined by
the following equation:

As for the emission coefficient,

31 / 50
The monochromatic directional emittance of a surface is defined as
the ratio of the monochromatic intensity emitted by a surface in a
particular direction to the monochromatic intensity that would be
emitted by a blackbody at the same temperature, it is written as:

The directional total emittance is given by

32 / 50
It is the ratio of the emitted total intensity in the direction μ, φ to the
blackbody intensity.

The monochromatic hemispherical absorptance and emittance are

obtained by integrating over the enclosing hemisphere, as:

for the absorptance

and for the emittance

33 / 50
The total hemispherical absorptance and emittance are obtained by
integrating over all wavelengths and are defined by

for the absorptance

34 / 50
and for the emittance

2°/- Kirchoff’s law

We can always write

The monochromatic directional emittance is equal to the

monochromatic directional absorptance of the surface.
35 / 50
If the surface does not exhibit a dependence on the azimuthal angle,
then we have

and if the dependence on polar angle can also be neglected then

Kirchhoff’s law reduces to

and if the surface does not exhibit any wavelength dependency, that
is for a gray surface, the Kirchhoff’s law reduces to

α =ε

36 / 50
3°/- Reflectance of surfaces

Figure 3: Reflection from surfaces

The bi-angular reflectance or reflection function is used to relate the

intensity of reflected radiation in a particular direction by the
following equation

37 / 50
( µi , ϕi ) are the angular coordinates of the direction of the incident
radiation, and
( µ r ,ϕ r )
are the angular coordinates of the direction of the reflected
radiation

The energy that is reflected in all directions is given as

38 / 50
The monochromatic angular-hemispherical reflectance can then be
expressed as

The incident energy can be written in terms of the incident intensity

integrated over the hemisphere, as

Then the monochromatic hemispherical-angular reflectance is

given as

39 / 50
In terms of the reflectance function this equation can be written as

For the special case when the incident radiation is diffuse, the
monochromatic hemispherical-angular reflectance is identical to
the monochromatic angular-hemispherical reflectance.
That is

40 / 50
The total angular-hemispherical reflectance is given as:

The monochromatic hemispherical-hemispherical reflectance is

given as:

where,

41 / 50
the reflected monochromatic energy can be expressed in terms of
the reflection function and the incident intensity by

and the incident energy, expressed in terms of the incident

intensity, is

The total hemispherical reflectance is found by integration over all

wavelengths and finding the ratio

42 / 50
4°/- Relationships between absorptance, emittance, and reflectance

Consider a surface located in an isothermal enclosure maintained

at temperature T.
The monochromatic intensity leaving in a direction μ, φ an
infinitesimal area of the surface consists of emitted and reflected
radiation and must be equal to the incoming intensity Iλb :

43 / 50
(In fact this is an energy balance made on a location very close to
surface)

The emitted intensity is

and the reflected intensity is

Since the incident radiation is diffuse, the hemispherical-angular

reflectance is equal to the angular-hemispherical reflectance,
therefore we have

44 / 50
and using Kirchhoff’s law, we can write

When integrated over all directions, this relation clearly states that
for an opaque surface, energy from all directions, either
monochromatic or total, is either absorbed or reflected, and we
write

and for diffuse radiations

45 / 50
 ρ +α = ρ +ε = 1

5°/- Selective surfaces

Solar collectors must have high absorptance for radiation in the

solar energy spectrum, since 97% of the extraterrestrial solar
radiation is at wavelengths less than 3.0μm, and

At the same time, it is desirable to have the long-wave emittance of

the surface as low as possible to reduce radiative losses, since less
than 1% of the blackbody radiation from a 200 oC surface is at
wavelengths less than 3.0μm (the temperature of most flat-plate
collectors is less than 200 oC).

46 / 50
Surfaces with high absorptance for solar radiation and low
emittance for long-wave radiation are termed selective surfaces.

Table 9 Properties of Some Selective Surfaces

47 / 50
6°/- Angular dependence of solar absorptance

The directional absorptance for solar radiation of ordinary

blackened surfaces (such as are used for solar collectors) is a
function of the angle of incidence of the radiation on the surface.

An example of this dependence of absorptance on angle of

incidence is shown in the Figure below

48 / 50
 Figure 4 : Absorptance as a function of angle of incidence

A polynomial fit to the curve of this figure is

49 / 50
where α n is the absorptance for the radiation in the direction
normal to the surface.

END OF CHAPTER 3

50 / 50