Heat Transfer Lab Manual

ME 341A
(2023-24-II)

January 6, 2024
List of experiments

Experiment TA (2023-24-II)

1: Unsteady heat conduction Pawan Kumar

2a: Forced convection from a pin fin Yogesh Bhaskar
2b: Heat transfer through an extended surface Anuj Warghat
3: Critical heat flux in pool boiling Debartha Chaterjee
4: Measurement of emissivity (radiation) Arun Kumar
5a: Natural convection Nivesh Mishra and Atul Anand
5b: Thermocouple calibration Chandan Prasad
 Unsteady heat conduction
Experiment No. 1

Objectives
1. Estimate the heat transfer coefficient for transient heat transfer from a constant temperature
water bath to a metal (a) plate, (b) long cylinder, and (c) sphere of known thermal
conductivity.

2. To determine the thermal conductivity of a solid body using analytical transient-

temperature/heat flow charts from measurements taken on a similar shaped body, under
similar conditions, but having a different thermal conductivity.

The study is based on the use of analytical transient temperature charts (Heisler charts).

Theory
In the absence of any heat source, and given sufficient time, the body will reach a thermal
equilibrium with the ambient such that the temperature inside the body is constant with respect
to space and time. Prior to this state, unsteady heat transfer takes place between the body and the
ambient (reservoir of heat); and temperature within the body is a function of both time and space.
Within a body of arbitrary shape, the spatio-temporal temperature variation is given by the
three-dimensional heat diffusion equation. However, for shapes such as a plane wall, a long
cylinder, or a sphere, in the absence of an internal heat source, the heat equation may be
conveniently reduced to the one-dimensional heat equation, where the temperature variation is
only along the thickness of the wall or the radius of the cylinder/sphere. For the wall, the heat
equation assuming constant physical properties within the body reduces to:

𝜕 2𝑇 1 𝜕𝑇
=
𝜕𝑥 2 𝛼 𝜕𝑡

𝛼 = 𝑘/𝜌𝑐 𝑝 is the thermal diffusivity with 𝑘, 𝜌, and 𝑐 𝑝 being thermal conductivity, density,
and specific heat of the solid material, respectively. Considering the boundary conditions1 ,
𝜕𝑇 𝜕𝑇

= 0 and −𝑘 = ℎ 𝑇𝐿, 𝑡 − 𝑇∞ and the initial condition that the body has a uniform
𝜕𝑥 𝑡,𝑥=0 𝜕𝑥 𝑡,𝑥=𝐿
temperature 𝑇1 at 𝑡 = 0, the solution to the transient heat transfer problem can depend on several
parameters: 𝑇 = 𝑇 (𝑥, 𝑡, 𝑇1 , 𝑇∞ , 𝐿, 𝑘, 𝛼, ℎ). However, with appropriate non-dimensionalization,
the physical parameters form four non-dimensional groups:

𝑥 𝑇 − 𝑇∞ 𝛼𝑡 ℎ𝐿
𝑥∗ = , 𝜃= , 𝜏= = Fo, Bi =
𝐿 𝑇1 − 𝑇∞ 𝐿2 𝑘

and result in the following dimensionless functional dependence: 𝜃 = 𝜃 (𝑥 ∗ , Fo, Bi). Fo is the
dimensionless time, also known as Fourier number, which is representative of the heat conducted
with respect to the heat stored within the body. Bi is the ratio between resistances to conduction
within the body and to convective heat transfer outside, and is known as the Biot number. A small
Biot number (𝐵𝑖 < 0.1) signifies that temperature profile within the body may be considered
spatially uniform on account of less resistance to conduction of heat.
Exact solutions of the one-dimensional transient heat equations involve transcendental
equations or infinite series; however, for large non-dimensional times (Fo > 0.2), a one-term
approximation leads to an error of less than two percent.
The following are the one-term approximations2 for the spatio-temporal temperature profile
and the time-varying temperature at the mid-plane (plane wall), axis (cylinder), and the center
(sphere) for 𝜏 > 0.2:

2 2
Plane wall : 𝜃 𝑥 ∗ ,𝜏 = 𝐴1 𝑒 −𝜆1 𝜏 cos (𝜆 1 𝑥 ∗ ) 𝜃 𝑥=0,𝜏 = 𝐴1 𝑒 −𝜆1 𝜏
2 2
Cylinder : 𝜃 𝑟 ∗ ,𝜏 = 𝐴1 𝑒 −𝜆1 𝜏 𝐽0 (𝜆 1𝑟 ∗ ) 𝜃 𝑟=0,𝜏 = 𝐴1 𝑒 −𝜆1 𝜏 (1.1)
2 𝑟 ∗) 2
Sphere : 𝜃 𝑟 ∗ ,𝜏 = 𝐴1 𝑒 −𝜆1 𝜏 sin(𝜆 1
𝜆1 𝑟 ∗ 𝜃 𝑟=0,𝜏 = 𝐴1 𝑒 −𝜆1 𝜏
𝐴1 and 𝜆 1 are dependent on the shape (plane wall, cylinder, or sphere) and the Biot number,
Bi. The relations were also graphically provided by M. P. Heisler in 1947. The plots—also
known as Heisler charts—are provided for the three shapes in App. 1 of this chapter.
1𝑥 = 0 is the mid-plane of the plane wall, and correspondingly 𝐿 is half of the wall thickness.
2𝑟 ∗ = 𝑟/𝑟 0 , 𝑟 0 being the radius of the cylinder/sphere.

2
 Figure 1.1: Photographs of the major components of the experimental setup.

Figure 1.2: Details of the hot water bath with locations of the thermocouples 𝑇1 –𝑇3 .

Experimental setup
Figure 1.1 presents photographs of the main components of the unsteady heat transfer setup—the
hot water bath, which is equipped with a heater and a water circulation pump for water bath
temperature control; a heat transfer and temperature indicator unit HT10X, which acts as the
interface for water circulation pump control, as well as for the display of temperatures; and a
computer connected to the heat transfer unit for data logging. Additional details of the water
bath, in particular the thermocouple locations are shown via a schematic in Fig. 1.2.
The hot water bath is provided with a port on the top through which the various metal samples
can be immersed. Images of the metal shapes and the holder are shown in Fig. 1.3.

3
 Figure 1.3: The metal shapes investigated in the experiment.

Procedure
In the experiment, unsteady heat transfer from a constant temperature water bath to samples
(polished rectangular block, cylindrical rod, and sphere) made out of metal with known physical
properties (brass; 𝛼=3.7 × 10−5 m2 s−1 and 𝑘=121 W m−1 K−1 ) is analysed to obtain the heat
transfer coefficient, ℎ, for convective heat transfer between the bodies and the water bath.
In the second part, unsteady heat transfer is studied for geometrically-identical shapes made
out of steel. Since the shapes are identical to the corresponding brass samples in terms of
geometry and finishing, the heat transfer coefficient, ℎ calculated for brass is expected to describe
convective heat transfer for steel samples as well. Knowing the value of ℎ for the steel samples,
𝑘 can be obtained using the Heisler charts (or corresponding tables).
The following steps are to be followed for the experiments:

1. Switch on the front Main switch (If the panel meters do not illuminate, check the Residual-
Current Device (RCD) and any other circuit breakers at the rear of the service unit. All
the switches at the rear should be turned up)

4
 2. Check that the water bath is filled with enough water. Afterwards, switch on the electrical
supply to the water heater (switch on the RCD which is located on the connection box
adjacent to the water heater)

3. Ensure that the green light is illuminated on the water heater, indicating that electrical
power is being supplied to the unit

4. Adjust the thermostat setting on the water heater to 4 and check that the red light is
illuminated, indicating that power is being supplied to the heating element

5. Set the voltage to the circulating pump to 12 volts using the voltage control box on the
mimic diagram software display

6. Allow temperature of the water to stabilise (monitor the changing temperature, 𝑇1 )

7. The water temperature must be in the range 60 − 70 ◦ C for satisfactory operation. If

outside this range, adjust the thermostat and monitor, 𝑇1 , until the temperature is within
this range

8. Attach the brass cylinder to the shape holder (insert the insulated rod into holder and
secure using transverse pin), but do not hold the metal shape or subject it to a change in
temperature. Check that the thermocouple attached to the shape is connected to 𝑇3 on the
HT10XC and that the thermocouple wire is positioned in the slot at the top of the shape
holder

9. Check and confirm that the temperature of the shape has stabilised (same as air temperature,
𝑇2 )

10. Switch off electrical supply to the water bath (i.e., switch off the RCD on the connection
box) to minimise fluctuations in temperature in case the thermostat switches on/off.

11. Start continuous data logging by selecting the "GO" icon on the software toolbar

12. Allow temperature of the shape to stabilise at hot water temperature (monitor the changing
temperature, 𝑇3 , on the mimic diagram software screen)

13. When temperature, 𝑇3 , has stabilised , click the "STOP" icon to end data logging

14. Create new results sheet using the appropriate icon on the screen

5
 15. Switch on electrical supply to the water bath to allow the thermostat to maintain the water
temperature

16. Remove the brass cylinder from the shape holder and fit in the stainless steel cylinder

17. Repeat the above procedure to obtain the transient response for the stainless steel cylinder.
Remember to create a new results sheet afterwards for the next set of results

18. Remove the stainless steel cylinder from the shape holder

19. Response of the other shapes can also be determined using the same procedure as above

The following precautions must be observed:

1. Handle the metal shapes only via the insulated rods otherwise heat transferred to the shape
by holding it from the metal part will affect the initial temperature readings.

2. Ensure that the water bath is sufficiently filled with water before switching on the heater.

3. The water circulation loop should be switched off before taking measurements.

4. Ensure that the thermocouples are working and appropriately connected during the
experiments.

Results and discussion

Estimating the heat transfer coefficient for brass shapes
For each measurement, 𝑡 = 0 is established as the time when 𝑇2 approaches 𝑇1 . The time stamp for
each subsequent data point is adjusted according to this reference for 𝑡 = 0. Subsequently analysis
is performed with the data point where 𝑇3 is about 2–3 degrees below the final temperature of the
data set. The corresponding time stamp is referred to as 𝑡1 . Using 𝑇3 | 𝑡=0 as the initial temperature,
the water bath temperature, 𝑇1 , and 𝑇3 | 𝑡=𝑡1 , the non-dimensional temperature 𝜃 is obtained. The
Fourier number (the non-dimensional time) is obtained based on 𝑡1 ; the thermal diffusivity, 𝛼,
for brass; and the characteristic length of the shape. Based on these non-dimensional quantities
for the measurement point and the Heisler chart given in App. 1, the Biot number (ℎ𝐿/𝑘) is
obtained. Finally, the characteristic length and the thermal conductivity of brass will be used to
estimate ℎ.

6
Estimating thermal conductivity of steel
The heat transfer coefficient, ℎ estimated from the brass shapes can now be employed to get
the thermal conductivity, 𝑘, of steel. The procedure for obtaining the non-dimensional time
and temperature is the same as just discussed. 𝛼 for steel may be taken as 0.6 × 10−5 m2 s−1 .
ℎ𝐿
Once the Biot number is obtained from the Heisler charts, the relation Bi = together with
𝑘
the characteristics length scale and the heat transfer coefficient obtained for the different shapes,
ℎ, is used to estimate 𝑘.3

3 For steel, the typical thermal conductivity coefficient is 25 W m−1 K−1 .

7
Appendix 1: Heisler charts for a plate, cylinder, and sphere

8
9
10
Appendix 2: Observation table template

Temperature (◦ C)

Shape Metal 𝑻1 𝑻2 𝑻3 𝒕1 (min)

Plate brass
steel

Cylinder brass
steel

Sphere brass
steel

11
 Forced convection from a pin fin
Experiment No. 2a

Objectives
1. Experimentally investigate the temperature distribution along the length of a pin fin with
an insulated end under forced convection configuration.

2. Determine the heat transfer coefficient for the pin fin for the configuration.

3. Compare the temperature distribution obtained in the experiments with the theoretical
prediction for the same.

4. Evaluate the effectiveness and efficiency of the pin fin.

Theory
Convective heat transfer from a body to the surrounding fluid is given by the Newton’s law of
cooling:
𝑄¤ 𝑐𝑜𝑛𝑣 = ℎ𝐴𝑠 (𝑇𝑠 − 𝑇∞ ) (W)

where

 
−2 ◦ −1
ℎ = convective heat transfer coefficient Wm · C
 
𝐴𝑠 = heat transfer surface area m2

𝑇𝑠 = surface temperature ( ◦ C)

𝑇∞ = temperature of the surrounding fluid far away from surface ( ◦ C)

 The heat transfer coefficient for a body predominantly depends on the thermo-physical
properties of the surrounding fluid, shape of the body, and the relative velocity between the body
and the ambient fluid. In particular, an increase in the relative velocity increases ℎ (and thereby
convective heat transfer), such that it is practically advantageous in several situations to have
forced flow, using an external agent such as a fan, over the body. Such configurations fall under
the case of forced convection.
When a further increase in the heat transfer coefficient is not possible, heat transfer may be
enhanced by increasing the surface area, 𝐴𝑠 , by attaching extra material to the body in the form
of plates, rods, or annular rings to name a few. Such extended surfaces are referred to as fins. A
cylindrical rod used as an extended surface is known as a pin fin.
Conduction and convection are the primary heat transfer mechanisms for fins; and under
steady conditions, a temperature profile is established over the length of the fin depending on
the fin configuration. Analytical solutions exist for standard cases, including a pin fin with an
insulated end and with constant heat input from the other side (assuming a one-dimensional
case). As is typically the case, the fin is attached to a body with a higher temperature than the
surroundings; and the temperature along the fin decreases from the base towards the fin tip. At
any infinitesimal section along the pin fin, the heat loss due to conduction towards the tip and the
loss due to forced convection is balanced by heat input due to conduction from the side facing
the base of the fin and the following equation is obtained:
d𝑞 𝑐𝑜𝑛𝑑
+ ℎ𝑝 (𝑇 − 𝑇∞ ) = 0
d𝑥
Here, 𝑝 is the perimeter of the infinitesimal fin element. In combination with the Fourier’s law
of conduction,
d𝑇
𝑞 𝑐𝑜𝑛𝑑 = −𝑘 𝐴𝑐
d𝑥
where 𝑘 is the thermal conductivity of the material and 𝐴𝑐 is the cross-sectional area. Substituting,
we obtain the following equation for steady state heat transfer along a pin fin:
 
d d𝑇
𝑘 𝐴𝑐 − ℎ𝑝 (𝑇 − 𝑇∞ ) = 0
d𝑥 d𝑥
which can be simplified to
d2 𝜃
2
− 𝑚2 𝜃 = 0
d𝑥
𝜃 = 𝑇 − 𝑇∞
ℎ𝑝
𝑚2 =
𝑘 𝐴𝑐

2
Assuming (a) constant thermal conductivity of the material 𝑘, (b) constant heat transfer coefficient
ℎ along the pin fin, and (c) negligible heat transfer due to conduction at the insulated end (𝑋 = 𝐿);
the following analytical solution for the temperature profile along a pin fin can be derived:
𝜃 𝑇 − 𝑇∞ 𝑐𝑜𝑠ℎ 𝑚(𝐿 − 𝑥)
= =
𝜃 0 𝑇0 − 𝑇∞ 𝑐𝑜𝑠ℎ 𝑚𝐿
where
𝐿 = length of the fin
𝑇0 = temperature at the base
With this expression, the rate of heat transfer from the fin can be obtained using the Fourier’s
law of conduction at 𝑥 = 0 as equal to:
p
𝑄¤ 0 = 𝜃 0 ℎ𝑝𝑘 𝐴𝑐 𝑡𝑎𝑛ℎ 𝑚𝐿

One of the objectives of the experiment is to estimate the heat transfer coefficient, ℎ. This is
achieved by using the following correlation for heat transfer from a cylindrical surface under
various flow conditions:
𝑁𝑢 = 𝐶 𝑅𝑒 𝑛 𝑃𝑟 1/3

where


ℎ𝑑
𝑁𝑢 = the Nusselt number : the ratio between convective and conductive heat transfer
𝑘𝑎
across the boundary at the surface.
 
𝑉𝑎 𝑑
𝑅𝑒 = the Reynolds number : the ratio between inertial and viscous forces in a flow.
  𝜈
𝜇𝑐 𝑝
𝑃𝑟 = Prandtl number : the ratio of momentum (viscous) and thermal diffusions.
𝑘𝑎
The subscript, 𝑎 denotes a property of air flowing over the fin; 𝐶 and 𝑛 are experimentally-obtained
values that depend on the Reynolds number as given in App. 1.
The correlation requires estimation of various temperature-dependent properties of air that
appear in it. These properties are to be estimated at a ‘mean temperature’ 𝑇𝑚 of the air flow
participating in the convective heat transfer process; which may be taken as the average !of the air
Í5
1 𝑇𝑖
temperature far away from the fin, 𝑇6 and 𝑇𝑠 , the mean fin surface temperature .
5

Fin effectiveness

Effectiveness for a fin is defined as the ratio of heat transfers with and without fin. Without the fin,
heat transfer can be estimated as the convective heat transfer from a surface area approximately

3
equal to 𝐴𝑐 : 𝑄¤ 𝑛𝑜− 𝑓 𝑖𝑛 = ℎ𝐴𝑐 (𝑇1 − 𝑇∞ ). The effectiveness, 𝜀 then is given as:
r
𝑝𝑘
𝜀= 𝑡𝑎𝑛ℎ 𝑚𝐿
ℎ𝐴𝑐

Fin efficiency

Efficiency is defined as the ratio of the actual heat transfer from the fin to the maximum
possible heat transfer assuming a uniform temperature (𝑇1 ) throughout the fin surface (i.e.
𝑄¤ 𝑚𝑎𝑥 = ℎ𝐴 𝑓 𝑖𝑛 (𝑇1 − 𝑇∞ )). The efficiency for the pin fin therefore is equal to:

𝑡𝑎𝑛ℎ 𝑚𝐿
𝜂=
𝑚𝐿

Experimental setup
The pin fin to be investigated is shown schematically in Fig. 2.1. The pin fin is 160 mm long
cylinder with a diameter (cross-section) of 12 mm. The fin is heated on one side (𝑥 = 0) with a
electrical heater and is insulated at the other end (𝑥 = 𝐿). Thermocouples (𝑇1 –𝑇5 ) mounted on
the fin measure temperatures at five axial stations along the fin, as shown in the figure. The fin is
made out of brass (thermal conductivity, 𝑘 = 110 W m−1 K−1 ).
A schematic of the setup is shown in Fig. 2.1. The fin is mounted inside a rectangular duct
0.2 × 0.1 × 0.7 𝑚 open on one end and attached to a blower on the other. The pin fin is mounted
along the breadth of the duct, at an axial location 0.093 m from the open end and at half the
height of the duct. Another thermocouple (𝑇6 ) measuring the temperature of the air inside the
duct is mounted at an axial location 0.56 m from the open end of the duct. The heater and the
blower are connected to a control panel for controlling the heat input to the pin fin and the mean
flow velocity within the duct respectively. The heat input can be adjusted within a range 0 W to
60 W, while the flow velocity is kept constant throughout the experiment at 1.25 m s−1 .
The flow velocity is measured using an orifice flow measurement setup, where the velocity is
estimated from the pressure drop through the orifice (measured via a water column manometer).

Procedure
Experiments will be performed to obtain the temperature profiles and heat transfer coefficients,
ℎ (using the empirical correlation), at various heater power inputs. Experimental data for each

4
Figure 2.1: Schematic, dimensions of the pin fin, and the location of thermocouples 𝑇1 –𝑇5 .

Figure 2.2: Schematic of the forced convection over a pin fin setup used for the experiments.

5
case will comprise of (a) the temperature recordings and (b) the manometer readings (in order to
estimate the air flow velocity within the duct). The blower speed is not a variable, and will be
kept constant for all experiments. The scenario simulates the case where increasing the surface
area for heat transfer (using the fin) is the only possible way of enhancing heat transfer.
The following steps are recommended for the experiment:

1. Establish a steady mean flow in the duct by switching on the blower and waiting for
5–10 min for conditions to become steady.

2. Record temperatures 𝑇1 –𝑇6 using the selector switch on the thermotech temperature reading
device, as well as the manometer readings (difference between the liquid levels in the
liquid column). This recording will be for the case when no heat is supplied to the pin fin;
and is essentially a reference case.

3. Subsequently take recordings for increasing values of the heater power input. For each
measurement, it is important to wait 40–50 min for the system to achieve a steady state.

The following precautions must be observed:

1. The blower must be switched on before the heater is switched; otherwise there is risk of
damage due to overheating.

2. For the same reason, the blower must not be switched off while the heater is operating.

3. At the end of the experiment, the heater must be turned off first. Subsequently, wait
2–5 min before turning the blower off so that the heater has cooled down to a safe level.

Results and discussion

The heater input power, the pressure drop (manometer), and the temperatures are tabulated for all
the experiments performed. The mean air temperature, 𝑇𝑚 is inferred from 𝑇1 –𝑇6 as mentioned
previously. Properties of air required for Nusselt, Reynolds, and Prandtl numbers are estimated
based on this mean temperature, 𝑇𝑚 (tables given in the App. 2 towards the end of the chapter).
Reynolds number also requires the flow velocity, which is inferred from the pressure difference
measurements (see App. 3 for details). Once the Reynolds number has been estimated the
coefficient, 𝐶, and the index, 𝑛, in the empirical correlation between 𝑁𝑢, 𝑅𝑒, and 𝑃𝑟 are obtained

6
from the lookup table provided in App. 1. Subsequently, the heat transfer coefficient, ℎ, being
the only unknown in the correlation can be identified for each experiment and its variation with
the heater power input (alternatively, the fin base temperature 𝑇0 ) can be identified. The obtained
ℎ can then be plugged into the Newton’s law of cooling to obtain heat transfer from the fin.
Additionally, once the temperature profile along the length of the fin, and the heat transfer
coefficient have been experimentally obtained, the data are compared to theoretical predictions
of the temperature profile along the fin. Finally, the fin effectiveness and efficiency can be
estimated.
Deviations from theoretical predictions and potential causes are to be discussed in the reports.

7
Appendix 1: Reynolds number dependence of Nusselt number
correlation coefficient and index

Reynolds number C n

0.4–4 0.989 0.33

4–40 0.911 0.385
40–4000 0.683 0.466
4000–40000 0.293 0.618
40000–400000 0.27 0.805

Table 2.1: The dependence of the parameters of the empirical correlation between 𝑁𝑢, 𝑅𝑒, and
𝑃𝑟 on 𝑅𝑒, the Reynolds number.

Appendix 2: Temperature-dependent air properties

Temperature 𝝆 𝒄𝒑 𝝂 (×1 × 10−6 ) 𝒌 𝒂 (×1 × 10−3 ) Pr

(K) (kg m−3 ) (kJ kg−1 K−1 ) (m2 s−1 ) (W m−1 K−1 ) (-)

250 1.3947 1.006 11.44 22.3 0.720

300 1.1614 1.007 15.89 26.3 0.707
350 0.9950 1.009 20.92 30.0 0.700
400 0.8711 1.014 26.41 33.8 0.690

Table 2.2: The dependence of the air properties required for the Nusselt number correlation on
mean temperature.

8
Appendix 3: Orifice flow meter
Pressure drop across a standard orifice plate can be used to obtain the average speed through the
orifice based on the following relation:
   1/2   −1/2
𝜌𝑤
𝑣𝑜 = 𝑐 𝑑 2𝑔Δ𝐻 1 − 𝛽4
𝜌𝑎

where,

Δ𝐻 = difference in the heights of the liquid columns (m)

𝑣𝑜 = average speed of flow through the orifice plate
𝜌𝑤 = density of manometric fluid (1000 kg m−3 )
𝜌𝑎 = density of air
𝑐 𝑑 = coefficient of discharge for the orifice geometry (here 0.85)
𝛽 = ratio between the orifice and orifice meter pipe diameters (0.3875)

The mean velocity in the rectangular duct holding the pin fin can then be calculated based on
mass conservation. Assuming identical air density in the duct as well as in the orifice meter,
and a uniform velocity profile in the duct, the velocities in the duct and in the orifice meter
tube are related as: 𝑉𝑎 𝐴𝑑𝑢𝑐𝑡 = 𝑣𝑜 𝐴𝑜 ; where 𝐴𝑜 is the cross-sectional area of the orifice plate
𝜋𝑟 2 ; 𝑟 = 0.00775 m, and 𝐴𝑑𝑢𝑐𝑡 is the cross-sectional area of the duct 0.15 × 0.1 m2 .

Appendix 4: Observation table template

Temperature ◦ C

No. Time (min) Heater (W) Δ𝐻 (cm) 𝑇1 𝑇2 𝑇3 𝑇4 𝑇5 𝑇6

9
 Heat transfer through an extended surface
Experiment No. 2b

Objectives
1. Experimentally obtain the temperature profile along the length of an extended surface and
compare data with analytical predictions.

2. Estimate the heat transfer coefficient for a combined convective and radiative heat transfer
situation.

3. Determine the combined rate of heat transfer and compare this value with the calculated
heat input provided to the extended surface.

4. Estimate the thermal conductivity of the material used for the extended surface.

Theory
Part 1
The term extended surface is commonly used to describe the special case involving heat transfer
by conduction within a solid and heat transfer by convection (and/or radiation) from the surface
of the solid. The direction of heat transfer from the surface is perpendicular to the principal
direction of conduction within the solid. Due to the combined action of the modes of heat
transfer involved, a temperature gradient exists along the extended surface.
The temperature distribution along the surface can be used to estimate heat transfer from
the surface to its surroundings. Since radiation and natural convection from the surface occur
simultaneously, both of these effects must be included in the analysis. Steady-state energy
 Figure 2.1: Schematic: infinitesimal segment along an extended surface.

balance for an infinitesimal section along the axis of the extended surface is employed to obtain
an analytical expression for the temperature profile—assuming one-dimensional heat conduction,
as well as uniform material properties and a constant cross-sectional area.
For an infinitesimal section along the extended surface as shown in Fig. 2.1, the heat loss due
to conduction towards the tip (the colder end) and the loss due to convection and radiation is
balanced by heat input due to conduction from the side facing the hot end of the body. Heat loss
due to natural convection is given by the Newton’s law of cooling, while radiative heat transfer is
given by the Stefan-Boltzmann law:

𝑞 𝑐𝑜𝑛𝑣 = ℎ𝑐 𝑝Δ𝑥 (𝑇𝑥 − 𝑇𝑎 ) (W)

𝑞𝑟𝑎𝑑 = 𝜎𝜀𝐹 𝑇𝑥4 − 𝑇𝑎4


(W)

where

ℎ𝑐 = convective heat transfer coefficient (W m−2 ·◦ C−1 )

𝑝 = perimeter of the cross-section (𝜋𝐷; 𝐷 being the cross section diameter) (m)
𝑇 = temperature at a point along the axis of the surface (◦ C)
𝑇𝑥 = surface temperature at the location, 𝑥 (◦ C)
𝑇𝑎 = temperature of the surrounding fluid far away from surface (◦ C)
𝜎 = Stefan-Boltzmann constant (5.670 373 × 10−8 W m−2 K−4 )
𝜀 = emissivity of the material (0.85 for the present setup)
𝐹 = view factor (considered equal to 1 here)
Convective and radiative heat transfer may be combined  and expressed as 𝑞 𝑐𝑜𝑛𝑣+𝑟𝑎𝑑 =
4
𝑇 − 𝑇𝑎 4
ℎ𝑝Δ𝑥 (𝑇𝑥 − 𝑇𝑎 ). Here ℎ = ℎ𝑐 + ℎ𝑟 ; and ℎ𝑟 = 𝜎𝜀𝐹 𝑥 . Accordingly, the heat transfer
𝑇𝑥 − 𝑇𝑎

2
balance gives
d𝑞 𝑐𝑜𝑛𝑑
+ ℎ𝑝 (𝑇 − 𝑇∞ ) = 0
d𝑥
In combination with the Fourier’s law of conduction,

d𝑇
𝑞 𝑐𝑜𝑛𝑑 = −𝑘 𝐴𝑐
d𝑥

where 𝑘 is the thermal conductivity of the material (brass) and 𝐴𝑐 is the cross-sectional area.
Substituting, we obtain the following equation for steady state heat transfer along a pin fin:
 
d d𝑇
𝑘 𝐴𝑐 − ℎ𝑝 (𝑇𝑥 − 𝑇∞ ) = 0
d𝑥 d𝑥

which can be simplified to

d2 𝜃
2
− 𝑚2 𝜃 = 0
d𝑥
𝜃 (𝑥) = 𝑇𝑥 − 𝑇𝑎
ℎ𝑝
𝑚2 =
𝑘 𝐴𝑐
Assuming (a) constant thermal conductivity of the material 𝑘, (b) constant heat transfer coefficient
ℎ along the pin fin, and (c) negligible heat transfer at the tip (𝑥 = 𝐿); the following analytical
solution for the temperature profile along a pin fin can be derived:

𝜃 𝑇𝑥 − 𝑇𝑎 𝑐𝑜𝑠ℎ 𝑚(𝐿 − 𝑥)
= =
𝜃 0 𝑇1 − 𝑇𝑎 𝑐𝑜𝑠ℎ 𝑚𝐿

where
𝐿 = length of the fin
𝑇1 = temperature at the base (𝑥 = 0)
For 𝑥 < 𝐿, the function is monotonously decreasing: the temperature decreases from the heater
end to the tip of the extended surface. This is physically correct because as x increases, the net
heat loss due to convection and radiation also increases.
Part 2a—a simplified approach for the total heat transfer
In a simplified approach, the total rate of heat loss from the rod can be estimated as:

𝑞 𝑡𝑜𝑡𝑎𝑙 = ℎ𝐴𝑠 (𝑇𝑠 − 𝑇𝑎 ) (𝑊)

𝐴𝑠 is the total surface are of the extended surface: 𝐴𝑠 = 𝑝𝐿; and 𝑇𝑠 is the average temperature of
the rod surface, and may be taken as the average of all the eight surface temperature measurements:
Í8
𝑇𝑖
𝑇𝑠 = 1 .
8

3
 This method is accurate for a linear temperature profile along the extended surface, and is
an approximation for the actual heat transfer as we know that the temperature profile follows a
hyperbolic function.
Part 2b—an alternate approach for the total heat transfer
In the second approach, we estimate the total heat loss from the temperature gradient at 𝑥 = 0.
From the Fourier’s law of conduction we know that the heat input to the extended surface at
𝑥 = 0 is given by the following expression:

d𝑇
𝑞 𝑡𝑜𝑡𝑎𝑙 = −𝑘 𝐴𝑐
d𝑥 𝑥=0

d𝑇
is obtained from the hyperbolic cosine fit to the experimental temperature distribution.
d𝑥 𝑥=0
The value of 𝑘 for the material (brass) must be known for this approach. It is in turn estimated
from the relation, 𝑘 𝐴𝑐 𝑚 2 = ℎ𝑝. The heat transfer coefficient is obtained from the Rayleigh and
Nusselt number correlation containing the convective heat transfer coefficient, ℎ𝑐 , and with an
estimate for the radiative heat transfer coefficient, ℎ𝑟 , both of which are described next.
Rayleigh and Nusselt number correlation
We have the experimental correlation between the Nusselt and Rayleigh numbers for the
present case of combined buoyancy-driven natural convection as follows:

𝑁𝑢 = 𝐶 (𝑅𝑎 D ) 𝑛

where
 
ℎ𝑐 𝐷
𝑁𝑢 = Nusselt number : convective vs. conductive heat transfer ratio
𝑘𝑎
in the boundary layer (thermal conductivity 𝑘 𝑎 is for air)
𝛽𝑔(𝑇𝑠 − 𝑇𝑎 )𝐷 3
 
𝑅𝑎 D = Rayleigh number
𝜈𝛼
𝑔 = acceleration due to gravity
𝛽 = coefficient of volumetric expansion (1/T, for ideal gases, where T is the
absolute mean temperature)
𝜈 = the kinematic viscosity
𝛼 = the thermal diffusivity
The coefficient, 𝐶, and the index, 𝑛 in this empirical relation are experimentally-obtained
values. These are dependent on the Rayleigh number as tabulated in App. 1. The properties

4
 𝑇𝑠 + 𝑇𝑎
of air are obtained corresponding to the temperature estimate: , where 𝑇𝑠 is the mean
2
surface temperature and 𝑇𝑎 is the ambient temperature.
Radiative heat transfer coefficient
Each infinitesimal section of the extended surface will radiate differently owing to different
surface temperatures. The radiative heat transfer coefficient for a section at location 𝑥 with a
temperature 𝑇𝑥 , assuming a view factor of 1 and emissivity 0.85 is given as:
 4
𝑇𝑥 − 𝑇𝑎4

ℎ𝑟 𝑥 = 0.85𝜎
𝑇𝑥 − 𝑇𝑎
An appropriate estimate for average radiative heat transfer coefficient is to take the mean of the
radiative heat transfer coefficients calculated for each of the eight temperature measurements:
8
Õ
ℎ𝑟 = 1/8 ℎ𝑟 𝑥𝑖
1

Experimental setup
Figure 2.2 contains an annotated photograph of the extended surface. The extended surface (a
cylindrical rod, length 350 mm diameter 10 mm made of brass) is mounted on a stand made out
of non-conducting materials. On one side it is attached to an electrical heater and the other end
is insulated. Thermocouples 𝑇1 –𝑇8 measure the surface temperature at eight equally spaced
locations along the length of the extended surface.
The heater power (specifically voltage) and the thermocouple input is connected to the HT10X
heat transfer unit. It allows the variation of input heat to the extended surface and to read out the
temperatures at the eight locations.
The setup involves heat transfer via conduction within the extended surface and heat loss to
the environment via natural (buoyancy-driven) convection and radiation. The extended surface
is coloured black in order to increase heat loss via radiation.

Procedure
Experiments to be performed involve recording steady state temperatures for three different
voltages (approx.)—7 V, 9 V and 11 V. From the temperature records, estimates will be obtained
for the heat transfer coefficients and the thermal conductivity of the extended surface material;
and the experimentally-obtained temperature profile will be compared to analytical estimates.

5
Figure 2.2: Annotated photograph of the extended surface setup. Also marked are the thermo-
couple (𝑇1 –𝑇8 ) locations.

The following steps are recommended for the experiment:

1. Ensure that the setup has been setup appropriately and safety precautions have been
followed.

2. Set the heater voltage to 6 V and allow the system enough time to attain steady state.
Preheating the setup at 20 V till 𝑇1 reaches about 80 Celsius would help achieve steady
state faster.

3. Record temperatures 𝑇1 –𝑇8 , and the ambient temperature, which corresponds to thermo-
couple 𝑇9 using the temperature selector.

4. Switch on the HT10X front panel once the system has achieved steady state. Also record
the current input to the heater.

5. Subsequently take recordings for increasing values of the heater power input, 9 V and
11 V.

The following precautions must be observed:

1. To be able to use the 𝑁𝑢–𝑅𝑎 D correlation, it must be ensured that convective heat transfer
is solely due to natural convection.

2. It should be ensured that no radiation heat source other than the extended surface itself is
present in the vicinity of the setup.

6
Figure 2.3: Sample comparison between experimentally-obtained data and theoretical estimates
for the temperature profile along the extended surface.

3. The voltage setting must be increased/decreased gradually.

Results and discussion

In the first part of the analysis, the temperature profile recorded for each of the heater voltage
settings will allow the estimation of ‘𝑚’ through an iterative fitting procedure.4 Once an estimate
for an ‘average 𝑚’ is obtained, estimates for the temperature at each measurement location, 𝑇𝑥 is
obtained and compared with the experimentally-obtained data. Fig. 2.3 presents an example of
the comparison between experimentally-obtained and theoretical estimates of 𝑇𝑥 as a function of
𝑥. The plot is obtained for each measurement.
The cumulative influence of experimental errors is estimated from the difference in the
experimental data and theoretical estimate for the temperatures, 𝑇𝑥 .
In the second part of the analysis, the measured rate of heat loss 𝑞 𝑡𝑜𝑡𝑎𝑙 is estimated based on
the two approaches mentioned in the ‘Theory’ section of this document. The air properties are
𝑇𝑠 + 𝑇𝑎
obtained from appropriate tables, evaluated at the estimate for the mean air temperature, .
2
The estimates for the rates of heat loss are compared to the calculated rate of heat input from the
heater (𝑉 𝐼).

4 As an initial guess, the value 8.0 may be used.

7
 Finally, based on the estimates of the heat transfer coefficients, ℎ𝑐 , ℎ𝑟 ; and ‘𝑚’ the thermal
conductivity for the material of the extended surface is estimated.5
Deviations from theoretical predictions and potential causes are to be discussed in the reports.

Appendix 1: Rayleigh number dependence of Nusselt number

correlation coefficient and index

RaD C n

1 × 10−10 – 1 × 10−2 0.675 0.058

1 × 10−2 – 1 × 102 1.02 0.148
1 × 102 – 1 × 104 0.85 0.188
1 × 104 – 1 × 107 0.48 0.25
1 × 107 – 1 × 1012 0.125 0.333

Table 2.3: The dependence of the parameters of the empirical correlation between 𝑁𝑢 and 𝑅𝑎 𝐷
on 𝑅𝑎 𝐷 , the Rayleigh number.

Appendix 2: Temperature-dependent air properties

Temperature 𝝆 𝜶 (×1 × 10−6 ) 𝝂 (×1 × 10−6 ) 𝒌 𝒂 (×1 × 10−3 ) Pr

(K) (kg m−3 ) (m2 s−1 ) (m2 s−1 ) (W m−1 K−1 ) (-)

250 1.3947 15.9 11.44 22.3 0.720

300 1.1614 22.5 15.89 26.3 0.707
350 0.9950 29.0 20.92 30.0 0.700
400 0.8711 38.3 26.41 33.8 0.690

Table 2.4: The dependence of the air properties required for the Nusselt number correlation on
mean temperature.

5 The value for brass is 121 W m−2 K−1 .

8
Appendix 3: Observation table template

Temperature (◦ C)

time (min) 𝑇1 𝑇2 𝑇3 𝑇4 𝑇5 𝑇6 𝑇7 𝑇8 𝑇9

7V
𝐼 = A
𝑄¤ 𝑖𝑛 = W

9V
𝐼 = A
𝑄¤ 𝑖𝑛 = W

12 V
𝐼 = A
𝑄¤ 𝑖𝑛 = W

9
 Critical heat flux in pool boiling
Experiment No. 3

Objectives
Experimentally obtain the critical heat flux for pool boiling (water) at various bulk temperatures
and compare the results with the predictions from an analytical correlation (Zuber’s correlation
for critical heat flux) for the same.

Theory
Boiling is a liquid-to-vapor phase change process and occurs at a solid-liquid interface when the
liquid is brought in contact with a surface maintained at a temperature, 𝑇𝑠 , sufficiently above the
saturation temperature of the liquid, 𝑇𝑠𝑎𝑡 . Boiling is of particular interest because it is an effective
means of carrying heat from a heated element into a liquid. However, the heat transfer process
with respect to the relative temperature between the heater and the liquid is highly nonlinear
and is affected by the dynamics of the liquid-vapor interface formed at the heater surface. A
generic pool boiling curve, first introduced by Shiro Nukiyama in 19346 , based on very similar
experiments as will be carried out here is shown in Fig. 3.1. The temperature difference between
the liquid and the heater is given on the X-axis on a logarithmic scale and the Y-axis is the
heat flux. For comparison, the flow boiling curve where a relative flow velocity is maintained
between the liquid and the heater is also shown.

6 S. Nukiyama. The maximum and minimum values of the heat q transmitted from metal to boiling water under
atmospheric pressure. J. Jap. Soc. Mech. Eng., 37:367374, 1934. (transl.: Int. J. Heat Mass Transfer, vol. 9,
1966, pp. 14191433).
 Figure 3.1: Regimes of pool boiling and comparison to flow boiling.

As the heater temperature is increased, the heat flux increases sharply for a small increase in
Δ𝑇, reaches a maximum, falls again to reach a minimum heat flux, and subsequently increases
monotonously in the range, Δ𝑇 ≈ 100–1000 K. Note, however, that a Δ𝑇 of 1000 K is sufficiently
high for most metals to melt and further increase in Δ𝑇 is practically not important for most
practical liquid-heater combinations.
The boiling curve is divided into four major regimes depending on the heat flux and the
dynamics of the bubble/liquid-vapor interface. With reference to the boiling curve, these are
natural convection, nucleate boiling, transitional boiling, and film boiling.
At small Δ𝑇, heat is removed from the heater via natural convection: liquid in the vicinity of
the heater becomes superheated and continues to rise in temperature until bubbles begin to form.
Beyond the bubble inception point, heat transfer via nucleate boiling takes over. On the
boiling diagram this regime exists between points A and C. Initially isolated vapor bubbles form
at isolated nucleation sites on the heater and then condense before reaching the surface. As Δ𝑇
is increased, more and more nucleation sites become active, vapour bubbles grow in size as they

2
rise, and they rise all the way till the liquid surface. Beyond point B vapor formed at the heater
surface merges into large vapor pockets—slugs and columns.
The heat transfer peaks at point C, the critical heat flux (CHF) point, when vapour is unable
to escape without obstruction, and the heater surface is partly covered by a vapor layer that
inhibits heat transfer. With further increase in Δ𝑇, the portion of the heater that comes in contact
with liquid decreases and the effectiveness of vapor escape (therefore, heat flux) is increasingly
diminished. This is the transitional boiling regime.
Because of the reduced heat transfer, the temperature of the heater rapidly rises once the CHF
point is crossed and the system immediately jumps to the film boiling regime characterized by a
stable vapor blanket covering the heater. Heat transfer is greatly reduced because heat must be
conducted through the vapor film instead of through the liquid. Δ𝑇 is of the order of 1000 K and
most metals (incl. Nichrome used here as well as by S. Nukiyama in his first experiments) melt.
The minimum heat flux point, point D, is typically only obtained in the reverse direction (i.e.
starting the experiment at point E in the boiling diagram or beyond, temperature is gradually
reduced).
The peak heat flux at CHF point is of technical importance. Although the heat flux is quite
high at this point, it is dangerous to operate equipment close to this point because if heat input is
increased beyond this point, the system will suffer sudden and damaging increase of temperature.
Correlations have been proposed for predicting the peak and minimum heat flux by Zuber7 .
The correlation for peak heat flux (𝑞00𝑚𝑎𝑥 , point C) and minimum heat flux (𝑞00𝑚𝑖𝑛 , point D)8 in
saturated pool boiling are given as:

𝑞00𝑚𝑎𝑥 = 𝐶 𝜌𝑣1/2 𝜆 (𝜎𝑔 (𝜌𝑙 − 𝜌𝑣 )) 1/4

  1/4
𝜎𝑔 (𝜌𝑙 − 𝜌𝑣 )
𝑞00𝑚𝑖𝑛 = 𝐶 𝜌𝑣 𝜆
(𝜌𝑙 + 𝜌𝑣 ) 2

7 N. Zuber. Hydrodynamic aspects of boiling heat transfer. Report AECU- 4439, US Atomic Energy Commission,
June 1959. UCLA doctoral dissertation.
8 Asmentioned earlier, this point will not be observed in experiments described here.

3
where,

𝑞00 = heat flux per unit area

𝐶 = coefficient estimated 𝜋/24 by Zuber and adjusted to 0.131 for 𝑞¤ 𝑚𝑎𝑥 and 0.09 for 𝑞¤ 𝑚𝑖𝑛
based on experiments by Kutateladze and Berenson, respectively
𝜌𝑣 = density of vapor phase calculated at 𝑇𝑠𝑎𝑡 and 1 bar (0.6 kg m−3 )
𝜌𝑙 = density of liquid phase at 𝑇𝑠𝑎𝑡 and 1 bar (957.9 kg m−3 )
𝜆 = latent heat of vaporization at 1 bar (2.257 kJ kg−1 )
𝜎 = surface tension at 𝑇𝑠𝑎𝑡 (0.0589 N m−1 )
𝑔 = acceleration due to gravity

Experimental setup
Figure 3.2 shows the schematic of the pool boiling setup, consisting of the glass vessel (diameter
200 mm, height 100 mm) for holding distilled water. It is mounted on an elevated platform for
aiding in the imaging of boiling process (forward-scatter illumination configuration using an
LED panel for illumination) via a high speed camera. The temperature of the pool is monitored
by a thermocouple. As illustrated, the Nichrome test wire is immersed sufficiently below the
water surface to ensure proper visualization of the vapor bubbles. An image of the setup assembly
together with the heater and temperature control panel and the camera is shown in Fig. 3.3.
The setup consists of two heaters: a 1 kW Nichrome coil heater for controlling the bulk water
temperature, R1, and the other being the test Nichrome wire, R2. The test wire is 0.2 mm in
diameter and 100 mm in length. Heat input to the test heater, R2, is controlled by changing the
voltage across the wire; while heater R1 operates at a constant power setting. The corresponding
change in the current flowing through the wires is also monitored. The heat input to the wires
adjusted for the efficiency provided by the manufacturer (0.86) for a given voltage, 𝑉 and current
𝐼 is:
𝑞𝑖𝑛𝑝𝑢𝑡 = 0.86𝑉 𝐼

Procedure
This experimental set up is designed to study the pool boiling phenomenon up to critical heat
flux point. Pool boiling over the Nichrome wire can be visualised in the different regimes up

4
 Figure 3.2: Schematic of the pool boiling setup.

Figure 3.3: Photograph of the pool boiling setup.

to critical heat flux point beyond which the wire melts. The heat input to the wire is gradually
increased by increasing applied voltage across the test wire and the transition from convection to
nucleate boiling can be observed in the experiments.
The formation of bubbles and their growth in size and number can be visualised, followed
by the vigorous bubble formation and their immediate rise to surface and subsequently in the
breaking of the wire indicating the occurrence of critical heat flux point. The experiment is
repeated for various temperatures of the water in the container up to the saturation temperature.
The following steps are to be followed for the experiments:

1. The glass container is filled with about 3–4 litres of distilled water.

5
2. Immerse the pool temperature control heater R1 (1kW Nichrome coil) as well as the test
Nichrome wire heater. Ensure both the heaters are completely immersed and the Nichrome
heater is deep enough below the surface of water to be able to make observations on the
bubble formation and rising phenomena.

3. Set up the required electrical connections for the heaters

4. Bring the pool water to the desired temperature (40–100 ◦ C) using the temperature control
heater, R1.

5. Disconnect the temperature control heater, R1; and switch on the test heater R2.

6. Gradually increase the voltage of the test heater and record images of various regimes of
boiling as the heater voltage is increased.

7. The temperature difference between the wire and the pool temperature will increase
as the voltage is gradually increased, and as discussed, beyond a critical temperature
difference the Nichrome wire will melt. Note the corresponding voltage setting for each
pool temperature investigated. Slowly bring the voltage back to 0.

8. Repeat the experiment for different pool temperatures.

The following precautions must be observed:

1. Before starting experiments, ensure that the voltage controls for the heaters is set to zero.

2. The water contained is to be filled with enough distilled water to immerse the heaters
sufficiently.

3. Connect the test heater wire across the studs tightly.

4. Operate the variac very gently in steps and allow sufficient time in between

5. Once electrical connections are set up and the heater switched-on, do not touch water or
terminal points.

6. After attaining the critical heat flux condition, decrease the voltage slowly and bring it
back to zero

6
 Figure 3.4: Variation in the peak heat flux with increasing pool temperatures.

Results and discussion

The heater setting for which the wire breaks is a condition beyond point C in the boiling curve,
𝑉𝐼
Fig. 3.1. The heat input per unit area to the Nichrome calculated as 𝑞00𝑖𝑛𝑝𝑢𝑡 = 𝜂 , where 𝜂
𝜋𝑑𝐿
is the efficiency of the test heater setup provided by the manufacturer (0.86), and 𝑑, 𝐿 are the
dimensions of the wire. Once the heat input even slightly exceeds the critical heat flux, 𝑞00𝑚𝑎𝑥 , the
heat transfer to water decreases and therefore the temperature of the heater shoots up immediately
till the point Nichrome melts. Therefore, the heater setting (and the corresponding 𝑞00𝑖𝑛𝑝𝑢𝑡 ) at
the point the heater wire breaks corresponds to 𝑞00𝑚𝑎𝑥 . This value is compared to the predicted
critical heat flux using Zuber’s correlation.
Because the prediction is based on theoretical considerations and experiments involving flat
plate heaters, the experimentally-obtained value for the Nichrome wire and the predictions may
be expected to be different.
With respect to the pool temperature, critical heat flux decreases at the water pool temperature
approaches 𝑇𝑠𝑎𝑡 , as seen in the sample results in Fig. 3.4. This trend is well captured in the
literature and is represented by the relation:

 
𝐶𝐻𝐹𝑠𝑢𝑏 Δ𝑇𝑠𝑢𝑏
= 1+ (3.2)
𝐶𝐻𝐹𝑠𝑎𝑡 Δ𝑇𝑤

7
 In addition to these measurements, instantaneous images of the boiling phenomenon are
captured for various values of the test heater voltage.

Appendix 1: Observation table template

Condition for test Nichrome wire (R2) to melt

Pool water
temperature (◦ C) Ammeter reading (A) Voltmeter reading (V)

40
50
60
70
80
90
95
100 (𝑇𝑠𝑎𝑡 )

8
 Measurement of emissivity (radiation)
Experiment No. 4

Objectives
Experimentally obtain the emissivity of a test object at various surface temperatures using a
physically identical, lamp black-coated body.

Theory
Thermal radiation is emitted by matter due to energy transitions of atoms and molecules
constituting the matter. As temperature is the measure of the strength of molecular activity in the
matter, the emission of thermal radiation is (1) a property of all matter above absolute zero; and;
(2) is dependent on the temperature, 𝑇, of the matter under consideration. Thermal radiation
is an electromagnetic wave and does not require any material medium for propagation. The
wavelength range encompassed by thermal radiation is approximately 0.3 to 50 𝜇𝑚. In turn, this
wavelength range includes three sub-ranges; the ultraviolet, visible, and infrared ranges. All
bodies can emit radiation and have also the capacity to absorb all or part of the radiation coming
from the surroundings towards it. Most materials are opaque to thermal radiation and, therefore,
thermal radiation is a surface phenomenon for such materials.
In addition to the (surface) temperature, the amount of thermal radiation from a body depends
also on the material of the body and the physical properties of the surface (such as surface
roughness). It is convenient to represent the radiation from a body in terms of the maximum
radiation that can be emitted from the surface for a given temperature. The idealized body
which emits this amount of radiation is referred to as a blackbody. A blackbody absorbs all the
incident radiation regardless of the spectral distribution. The reflectivity and transmissivity of a
blackbody is equal to zero. The Stefan-Boltzmann law gives the total (over all wavelengths)
radiation energy emitted by a blackbody per unit time, per unit surface area as:

𝐸 𝑏 (𝑇) = 𝜎𝑇 4 (W m−2 )

Here, 𝜎 is the Stefan-Boltzmann constant and is equal to 5.670 373 × 10−8 W m−2 K−4 and the
temperature, 𝑇, is in K. Surfaces coated with lamp-black paint approach this idealized blackbody
behaviour.
The blackbody serves as a reference for thermal radiation, and it is convenient to express
radiation from a body with respect to radiation from a blackbody at the same temperature
as the test body. This ratio of radiation emitted by the surface at a given temperature to the
radiation emitted by a blackbody at the same temperature is the emissivity, 𝜀(𝑇, 𝜆, 𝜃, 𝜙), of
the corresponding surface. It is a function of the temperature, wavelength (𝜆), and directions
represented by the zenith (𝜃) and azimuth (𝜙) angles. The emissivity of a blackbody is, by
definition, equal to 1.
Emissivity averaged over all directions is called the spectral hemispherical, or monochromatic
emissivity, and is defined as:
𝐸𝜆 (𝜆, 𝑇)
𝜀𝜆 (𝜆, 𝑇) =
𝐸 𝑏𝜆 (𝜆, 𝑇)
𝐸𝜆 (𝜆, 𝑇) and 𝐸 𝑏𝜆 (𝜆, 𝑇) are the hemispherical monochromatic emissive powers of the surface
and the blackbody respectively. The total emissive powers over all wavelengths gives the total
emissivity, 𝜀(𝑇): ∫∞
𝐸 (𝑇) 𝐸𝜆 d𝜆
𝜀(𝑇) = = ∫ 0∞
𝐸 𝑏 (𝑇) 𝐸 𝑏𝜆 d𝜆
0
Knowing the total emissivity, 𝜀(𝑇) of a body at a given temperature allows the estimation of
thermal radiation power emitted by the body at that temperature and the use of this information
for heat transfer calculations.
The emissivity experiments described below involve the estimation of the emissivity of a
test plate by comparing the heat inputs required to maintain the test plate and a reference
plate (blackbody) at equal temperatures (at steady conditions). The emissivity variation with
temperature can be obtained by varying the target temperatures.

2
Figure 4.1: Schematic diagram of the setup employed for emissivity measurements. 1: dim-
merstat (reference plate), 2: main switch, 3: dimmerstat (test plate), 4: double pole
double throw switch, 5: test plate heater, 6: test plate, 7: asbestos sheet, 8: reference
plate heater, 9: reference plate, 10: enclosure, T1–T3: thermocouples, A: ammeter,
V: voltmeter, T: temperature indicator (range 0 to 300 ◦ C).

Experimental setup
The setup consists of two circular (diameter, D=160 mm) Aluminium plates identical geometri-
cally and in physical properties, and equipped with heating coils at the bottom. The heating coils
are essentially Nichrome strips wound on Mica sheets and sandwiched between Mica sheets. The
efficiency of heat transfer from the electrical heaters to the plates is specified by the manufacturer
as 86%. The plates are mounted on an asbestos cement sheet and are kept in an enclosure
(580 mm x 300 mm x 300 mm) so as to provide undisturbed natural convection surrounding. A
schematic of the setup is given in Fig. 4.1 and a photograph of the same is shown in Fig. 4.2.
The heat input to the plates, 𝑊1 (reference plate) and 𝑊2 (test plate) is varied by the respective
dimmerstats and the same is measured using an ammeter and a voltmeter/wattmeter (𝑊𝑖 = 𝑉𝑖 𝐼𝑖 )
with the help of double pole double-throw switches. The temperature of the plates, 𝑇𝑠 (reference
and test plates are set to the same temperature), and the temperature of the ambient within
the enclosure (𝑇𝑑 ) is measured by J-type thermocouples (Iron-Constantan) placed as shown in
Fig. 4.1.
The reference plate (Plate-1) is blackened by a thick layer of lamp black to achieve the
black body radiation for the plate whereas the plate-2 is the test plate whose emissivity to be
determined.

3
 Figure 4.2: Photograph of the emissivity measurement setup

Procedure
The heat input to the two plates is dissipated from the plates by conduction, convection, and
radiation. The experimental set up is designed in such a way that at steady state the heat
dissipation by conduction and convection is the same for both (reference and test) plates.
When their surface temperatures are brought to the same value, the difference in the heater
input readings is the difference in radiation from the plates due to different surface emissivities
(for the reference plate 𝜀 = 𝜀 𝑏 = 1).
The following steps are to be followed to obtain measurements from which the emissivity of
the test plate (plate-2) can be estimated:

1. Gradually increase the power input to the heater of the reference plate by varying the
resistance using the dimmerstat knob and set it to a value (e.g. 30 W, 50 W and 75 W).

2. Adjust the power input to the test plate to a value slightly less value than the reference
plate (about 27 W, 35 W and 55 W corresponding to the list in given in the previous step).

3. Constantly check the temperatures of the two plates in 10 min time intervals and adjust the
power input to the test plate by means of the dimmerstat such that the test plate attains the
same steady state temperature as the reference plate. This will require trial and error and
one has to wait sufficiently long (an hour or longer) to reach a steady state where both
plates have the same temperature.

4. Once steady state is achieved, record the heat inputs, 𝑊1 , 𝑊2 ; and the temperatures 𝑇𝑠 and
𝑇𝑑 .

4
Figure 4.3: Typical variation of the emissivity of an Aluminium plate with surface temperature.

5. The procedure is to be repeated for several plate surface temperatures (it will take less time
if the temperatures are increased for every iteration) in order to obtain data for estimating
the effect of surface temperature on emissitvity.

The following precautions must be observed:

1. Use stabilized A.C. single phase supply (preferably).

2. Always keep the dimmerstats at zero position before the start.

3. Increase the heater inputs gradually (no sudden changes).

4. Ensure that the reference plate is covered with a layer of lamp black uniformly.

Note: There is the possibility of getting absurd results if the supply voltage is fluctuating or if
the input is not adjusted till the satisfactory steady-state condition is reached.

Results and discussion

Tabulate the records of 𝑊1 , 𝑊2 , 𝑇𝑠 , and 𝑇𝑑 obtained for the experiments. The emissivity of the
test plate, 𝜀, can be estimated from this data by using the Stefan-Boltzmann law for the reference
and the test plates; and by considering that the difference in the heat inputs is due to the different

5
radiation powers emitted by the plates:
h  i
0.86 (𝑊1 − 𝑊2 ) = 𝜎 𝐴 (𝜀 𝑏 − 𝜀) 𝑇𝑠4 − 𝑇𝑑4

Because the reference plate is considered a blackbody, 𝜀 𝑏 = 1; 𝐴 is the surface area (exposed to
the ambient) of each plate.
The effect of surface temperature on the emissivity of the test plate can be obtained by
estimating the emissivity at different reference heater input powers.
As the temperature is increased, the surface becomes ‘dull’; and the emissivity is expected
to increase with temperature. Figure 4.3 presents the typical variation in the emissivity of an
aluminium plate with surface temperature.

6
 Natural convection
Experiment No. 5a

Objectives
Experimentally determine heat transfer coefficient for heat loss via natural convection from a
uniformly-heated (uniform heat flux) vertical cylinder.

Theory
There are certain situations in which a fluid motion is produced due to change in density
resulting from temperature gradients, which is a heat transfer mechanism called free (or, natural)
convection. Natural convection is the principal mode of heat transfer from pipes, refrigerating
coils, hot radiators, etc. Movement of fluid in free convection is due to the fact that fluid particles
in the immediate vicinity of a hot object become warmer than the surrounding fluid, resulting
in a local change of density. The warmer fluid would then be replaced by the colder fluid,
thus, creating convection currents (see Fig. 5.1). These currents originate when a body force
(gravitational, centrifugal, electrostatic, etc.) acts on a fluid having density gradients. The force
which induces these convection currents is called a buoyancy force, which is attributed to the
presence of a density gradient within the fluid and a body force. Therefore, Grashof number (Gr),
the ratio of buoyant force to viscous force, plays a very important role in natural convection.
In contrast to forced convection, natural convection phenomenon is due to temperature
difference between the surface and the fluid and is not created by any external agency. The test
section is a vertical, open-ended, cylindrical pipe dissipating heat from the internal surface. This
test section is electrically heated, imposing the circumstantially and axially constant wall heat flux.
Figure 5.1: Convective thermal and velocity boundary layers for natural convection. Adapted
from A Heat Transfer Book by Lienhard, J. H.

As a result of heat transfer to air from the internal surface of the pipe, temperature of air increases.
The resulting non-uniform density causes air in the pipe to rise. The present experimental setup
is designed to study natural convection phenomenon from a vertical cylinder in terms of the
variation of local heat transfer coefficient and conduct a comparison of experimental results with
values obtained using an appropriate correlation. When a hot body is kept in a still atmosphere,
heat is transferred to the surrounding fluid by natural convection. The fluid layer in contact with
the hot body gets heated and rises due to a decrease in its density and the cold surrounding fluid
rushes in to take its place. This process is continuous and heat transfer takes place due to this
relative motion of hot and cold particles. Here, the heat transfer coefficient is given by:

𝑞
ℎ= (5.1)
𝐴𝑠 (𝑇𝑠 − 𝑇𝑎 )
Where,
ℎ = Average surface heat transfer coefficient,
𝑞 = Heat transfer rate,
𝐴𝑠 = Area of heat transferring surface,
𝑇𝑠 = Average surface temperature ( ◦ C),
𝑇𝑎 = Ambient temperature in the duct ( ◦ C) = 𝑇8 ,

2
 𝑇1 +𝑇2 +𝑇3 +𝑇4 +𝑇5 +𝑇6 +𝑇7
𝑇𝑠 = 7

Surface heat transfer coefficient of a system transferring heat by natural convection depends
on the shape, dimension, and orientation of the body, the temperature difference between hot
body and surrounding fluid, and fluid properties like 𝑘, 𝜇,𝑝, etc.
Dependence of all the above-mentioned parameters is generally expressed in terms of
non-dimensional groups as follows:

𝑛
𝑔𝛽Δ𝑇 𝐿 3 n 𝜇𝑐 𝑃 o
 
ℎ𝐿
=𝐴 (5.2)
𝑘 𝑣2 𝑘
Here,
ℎ𝐿
Nusselt Number, Nu = 𝑘
𝑔𝛽Δ𝑇 𝐿 3
Grashof Number, Gr = 𝜈2
𝜇𝑐
Prandtl Number, Pr = 𝐾𝑝

where, 𝐴 and 𝑛 are constants depending on the shape and orientation of the heat transferring
surface, 𝐿 is a characteristic dimension of the surface, 𝑘 is the thermal conductivity of the fluid,
𝜈 is the kinematic viscosity of the fluid, 𝜇 is the dynamic viscosity of the fluid, 𝑐 𝑝 is the specific
heat of the fluid, 𝛽 is the coefficient of volumetric expansion of the fluid, and 𝑔 is the acceleration
due to gravity at the place of experiment.

Δ𝑇 = 𝑇𝑠 -𝑇𝑎
1 −1
For gases, 𝛽 = 𝑇 𝑓 +273.15 𝐾

where, 𝑇 𝑓 is mean film temperature given by:

𝑇𝑠 +𝑇𝑎
𝑇𝑓 = 2

For a vertical cylinder losing heat by natural convection, constants A and n of equation 5.2
have been determined and the following empirical correlations have been obtained:

ℎ𝑡ℎ 𝐿 35𝐿
𝑁𝑢 = = 0.59(𝐺𝑟.𝑃𝑟) 0.25 ; 104 < 𝐺𝑟 · 𝑃𝑟 < 109 , 𝐷 ≥ (5.3)
𝑘 𝐺𝑟 0.25

ℎ𝑡ℎ 𝐿 35𝐿
𝑁𝑢 = = 0.1(𝐺𝑟.𝑃𝑟) 1/3 ; 109 < 𝐺𝑟 · 𝑃𝑟 < 1012 , 𝐷 ≥ (5.4)
𝑘 𝐺𝑟 0.25
Here, 𝐿 is the length of cylinder and ℎ𝑡ℎ is theoretical heat transfer coefficient. All properties
of the fluid are evaluated at the mean film temperature, 𝑇 𝑓 .

3
Apparatus
The apparatus consists of a stainless-steel tube fitted in a rectangular duct in a vertical fashion.
Control panel for the natural convection apparatus is shown in Figure 5.2. Heat input to the heater
is measured by an ammeter and a voltmeter and is varied using a Dimmerstat. Temperatures
of the vertical tube are measured using seven thermocouples (𝑇1 to 𝑇7 ) and are marked on
the Temperature Indicator Switch of the instrument panel as shown in Figure 2. One more
thermocouple is used to measure ambient temperature. Schematics of this natural convection
apparatus are shown in Figures 5.3 and 5.4

Figure 5.2: Control panel for natural convection apparatus

The duct is open at the top and the bottom forms an enclosure, which serves the purpose of
undisturbed surroundings. One side of the duct is made up of Perspex for visualization. An
electric heating element is kept in the vertical tube to internally heat the tube surface. Heat
is lost from the tube to the surrounding air by natural convection. The vertical cylinder with
thermocouple positions is shown in Figure 5.4.
Possible flow patterns and the expected variation of local heat transfer coefficient are shown in
Figure 5.5. The tube has been polished to minimize radiation losses.

Specifications
• Diameter of the tube (d) is 38𝑚𝑚

4
Figure 5.3: Natural convection setup and instrument panel (for temperature measurement and
heat input control) employed in the experiments.

• Length of the tube (L) is 500𝑚𝑚

• Duct size (200𝑚𝑚 × 200𝑚𝑚 × 1000𝑚𝑚)

• Number of thermocouples: 8

• Thermocouple number 8 reads the Ambient Temperature and is kept in the duct.

• Temperature Indicator 0-300 (◦ C). Multi-channel type calibrated from iron-constantan

thermocouples with compensation of ambient from 0-50 (◦ C)

• Ammeter

• Voltmeter

• Dimmerstat

Experimental procedure
1. Switch on the supply and adjust the Dimmerstat to obtain the required heat input (say, 40
W, 60 W, and 70 W approximately)

5
Figure 5.4: Left: schematic of the natural convection setup. Right: details of the constant heat
flux vertical cylinder and thermocouple locations investigated in this study.

2. Monitor temperatures 𝑇1 through 𝑇8 every five minutes till steady-state is reached

3. Wait till the steady-state is reached. This can be confirmed from temperature readings (𝑇1
to 𝑇7 ),i.e., if they remain steady and do not register a change of more than 1 ◦ C per hour

4. Measure surface temperature at various points (𝑇1 to 𝑇7 )

5. Note the ambient temperature, 𝑇8

6. Repeat this experiment for different heat inputs (say, 40 W, 60 W, and 70 W, approximately)
by varying the Dimmerstat position

Precautions
1. Switch off the ceiling fan before giving supply to set-up. This is to ensure a natural
convection heat transfer environment

2. Adjust the temperature indicator to ambient level by using compensation screw before
starting the experiment (if needed)

3. Keep dimmerstat to zero volt position and increase it slowly

6
Figure 5.5: Variation of heat transfer coefficient (ℎ) along the height of the tube in free air
flow and dependence of this variation on the nature of flow with (1) showing curve
obtained from empirical correlations (2) showing curve obtained from experimental
data

4. Operate the change-over switch of temperature indicator gently from one position to other,
i.e., from position 1 to position 8

Results and discussion

1. Calculate the value of average surface heat transfer coefficient
Average heat transfer coefficient is given by:

𝑞
ℎ𝑎𝑣𝑔 = (5.5)
𝐴𝑠 (𝑇𝑠 − 𝑇𝑎 )

Where,
𝑞 = 𝑉 · 𝐼 (W)
𝐴𝑠 = surface area of vertical cylinder rod =𝜋𝑑𝐿 𝑚 2
𝑇𝑠 = Average surface temperature ( ◦ C),
𝑇𝑎 = Ambient temperature in the duct ( ◦ C) = 𝑇8 ,
𝑇1 +𝑇2 +𝑇3 +𝑇4 +𝑇5 +𝑇6 +𝑇7
𝑇𝑠 = 7

2. Calculate the value of local heat transfer coefficient and plot a graph of the local heat

7
 transfer coefficient along length of the rod.

3. Compare the experimental value of average heat transfer coefficient with the theoretical
value obtained by using correlations 5.3 and 5.4

8
 Thermocouple calibration
Experiment No. 5b

Objectives
To calibrate a J-Type thermocouple using a Fluke field metrology well.

Theory
A thermocouple is an electrical device consisting of two dissimilar electrical conductors forming
electrical junctions at differing temperatures. A thermocouple produces a temperature-dependent
voltage a result of the thermoelectric effect, and this voltage can be interpreted as a measure of
the temperature (difference).
The basic theory behind this is that a small voltage is generated when two different metals are
joined together. This is known as Seebeck effect or thermoelectric effect. The magnitude of the
voltage generated depends on the temperature of the junction. The junction can therefore be
used as a temperature transducer, converting temperature into voltage. This type of temperature
transducer is called a thermocouple.
The voltages V1 and V2 at the two junctions oppose each other and any current in the circuit
is due to the voltage difference V1 -V2 which will be due to temperature difference between the
two junctions (shown in Fig. 5.1).
Types of thermocouples
Commonly used thermocouple metal combinations include J-type (iron-constantan), K-type
(chromel-alumel), T-type (copper-constantan), E-type (chromel-constantan).
Principle of operation
In 1821, the German physicist Thomas Johann Seebeck discovered that when different metals
are joined at the ends and there is an imposed temperature difference between the junctions,
a magnetic field is observed. The magnetic field he observed was later shown to be due to
thermoelectric current. The Seebeck effect is responsible for the behavior of thermocouples,
which are used to approximately measure temperature differences. The magnitude of the voltage
depends on the types of wire being used. Generally, the voltage is in the microvolt range.

5.1 Apparatus
Field Metrology Well (FMW) is designed to be a stable and reliable heat source that can be
used in the laboratory. The special built-in features of FMW- 9143 make it extremely adaptable.
Exclusive voltage compensation allows technical experts to plug into mains power with voltage
ranging from 90 V to 250 V AC, without any degradation to the instrument. The range of
ambient temperature compensation is 13 ◦ C to 33 ◦ C. Gradient temperature compensation keeps
the axial gradient within specification over the entire temperature range of the instrument and
over the specified guaranteed operating temperature range. These combined features, along with
the rugged, compact, and light-weight design make this instrument ideal as a thermocouple
calibrator. Unique safety features also make this the safest heat source available. Distinctive
air-flow design keeps the probe and instrument cool, protecting users from heat effects. Block
temperature indicator shows the user when well temperature is above 50 ◦ C, letting users know
when it is safe to remove the insert or move the instrument. An LCD display continuously shows
many useful operating parameters including block temperature, current set point, block stability,
and heating and cooling status. Rugged in design and equipped with these special features, make
them ideal for the field or laboratory with the proper use. The instrument provides continued
accurate calibration of temperature. Before the use of the instrument, the user should be familiar
with its warnings, precautions, and operating procedures.

5.2 Procedure
1. Switch on the Field Metrology Well (FMW 9143) and Temperature scanner (1586 A)

2. Measure initial temperature from FMW-9143

3. Set temperature to 30 ◦ C in FMW-9143 and press enter. Thereafter, heating of the dry
well will begin and temperature will rise up from ambient to 30 ◦ C

2
 4. When the dry well reaches a temperature of 30 ◦ C, the buzzer of FMW- 9143 will beep. It
signals that the system is reaching steady-state

5. Let the instrument continue the heating process / leave as is till it reaches a steady-state.
The steady-state may be reached after 10-12 minutes of the buzzer beeping

6. After reaching the steady-state, press the SCAN/Monitor key of the temperature scanner
(1586). Let the process continue for 4-5 minutes. Thereafter, display of the temperature
scanner (1586) will show values of the measured temperature and the corresponding value
of emf

7. Note the readings for measured temperatures and emf from the temperature scanner (1586)

8. Dont stop the system and dont press any key of the experiment during the above step

9. Follow the above procedure for the second measurement of temperatures and emf

10. Repeat the experiment and take measurements at 40, 50, 60, 70, 80, 90, and 95 ◦ C

11. Measure emf values w.r.t. each reading of thermocouple

12. Draw the plot for (T𝑚𝑒𝑎𝑠 - T𝑎𝑚𝑏 ) vs emf

5.3 Technical Specifications

1. Temperature range: Ambient to 350 ◦ C

2. Mains voltage: 220 to 240 Volts

3. Frequency: 50/60 Hz

4. Power: 1800 Watt

5. Immersion (Well) depth: 150 mm (5.9 in)

6. Insert outer diameter (OD): 25.3 mm (1.00 in)

7. Size (𝐻 × 𝑊 × 𝐷): 290 mm x 185 mm x 295 mm

8. Weight: 7.3 Kg

3
 9. Ambient temperature: 27 ◦ C

10. Ambient relative humidity (ARH): 0 to 90%

5.4 Precautions
1. Dont turn off the instrument until the temperature of FMW−9143 goes down to 40 ◦ C

2. Dont touch any wire/thermocouple wires when the instrument is in running condition

3. Make sure that the thermocouple beads are properly made

4. Thermocouple beads should not be in contact with each other inside the FMW−9143

5. Ensure that there is no loose connection in the experimental set up

4
 Figure 5.1: Seeback/Thermoelectric effect.

Figure 5.2: J-type (iron-constantan) thermocouple.

Figure 5.3: Wire diagram of Thermocouple Calibration System

5
 Figure 5.4: Thermocouples Input Module

Figure 5.5: Thermocouple calibration system