MEM311

THERMAL & FLUID SCIENCE LABORATORY

MANUAL



Edited by Brandon Terranova, Eric Wargo, Ertan Agar, Chris
Dennison and E. Caglan Kumbur


Adopted from MEM 311 Thermal & Fluid Science Laboratory
by Baktier Farouk and David Stacck


LIST OF EXPERIMENTS:
EXPERIMENT 1: FLOW MEASURING DEVICES PAGE 1
EXPERIMENT 2: CONTROL VOLUME ENERGY AND ENTROPY
ANALYSIS IN A VORTEX TUBE
PAGE 10
EXPERIMENT 3: HEAT TRANSFER FROM A CIRCULAR CYLINDER PAGE 21
EXPERIMENT 4: PERFORMANCE ANALYSIS OF A STEAM TURBINE
POWER PLANT
PAGE 38
Experiment 5: Lift Characteristics of an Airfoil Section
PAGE 51

LIST OF APPENDICES:
APPENDIX A: MANOMETER PREPARATION AND OPERATION
Page 63
APPENDIX B: FLOW MEASURING USING A ROTAMETER
Page 66
APPENDIX C: ANALYSIS OF BIAS ERRORS AND EXPERIMENTAL
UNCERTAINTY
Page 68
APPENDIX D: FITTING CURVES TO EXPERIMENTAL DATA
Page 74



MEM311: Thermal and Fluid Science Laboratory Experiment 1: Flow Measuring Devices
1 | P a g e


EXPERIMENT 1:
FLOW MEASURING DEVICES
by Brandon Terranova 2010, adapted from Bakhtier Farouk and David Staack

A. OBJECTIVES

The objective of this experiment is to determine the discharge coefficient for an
orifice flow meter as a function of the Reynolds number.

B. THEORY
Review of Friction Factors in Pipe Flow
As a fluid flows through a pipe (or any other device, for that matter), area changes, friction
and heat transfer affect the properties in a flow system. By evaluating the forces acting on a
control volume in a pipe flow, the pressure drop for fully developed laminar pipe flow, Ap,
is related to the wall shear stress,

, by the equation

(1)
Where is the length of the pipe and

is the equivalent hydraulic diameter defined as

= 4(Cross-sectional area of flow) / (Perimeter wetted by fluid). Since the wall shear stress
is a complex function of the flow velocity, viscosity, density, wall surface roughness, etc.,
the pressure drop, Ap, is expressed as a product of a non-dimensional friction factor, , and
the dynamic pressure (

). Which includes the velocity and density of the fluid, V and

respectively. So the pressure drop for a horizontal pipe is given as:

(

(2)
Equating Eq. (1) and (2), we obtain

(3)
MEM311: Thermal and Fluid Science Laboratory Experiment 1: Flow Measuring Devices

2 | P a g e


Therefore, the friction factor, , is a measure of the shear stress at the wall.

The friction factor, or more generally, the effects of viscosity on fluid flow, can be
correlated using the flow Reynolds number (ratio of inertial forces to viscous forces) given
as

(4)
where and are the density and absolute viscosity of the fluid, respectively, and V is the
mean flow velocity.
Review of Bernoullis Equation
As explained above, friction forces induce an irreversible decrease in pressure. The
pressure can also change in a reversible way as described by Bernoullis equation. Because
the crux of Bernoullis principle is that along a streamline of flow, the increase in velocity
corresponds to drop in the static pressure of the fluid. While Bernoullis equation assumes
a lot of simplifications to your system (constant density (incompressible), steady flow, no
friction), it produces very accurate results compared to empirical evidence at low Mach
numbers.
Using Bernoullis equation, the conservation of mass and the fact that mass flow rate,

, is constant through the duct, we can write the pressure drop along the duct as a
function of only the upstream velocity and the change in area:

( (

)

(5)
Pipe Flowrate Meters
Equation 2 shows that the pressure drop through a pipe is a function of the velocity
of the flow through the system along with the friction factor. In fact, one method to
determine the flow rate of fluid through a piping system is to measure the pressure drop
through a device for which the friction factor and other losses are precisely known. These
are called obstruction flow meters.
MEM311: Thermal and Fluid Science Laboratory Experiment 1: Flow Measuring Devices
3 | P a g e

There are three basic designs used in obstruction flow meters as shown in Fig. 1. A
venturi flow meter offers the highest accuracy and the lowest overall pressure drop but is
more expensive to manufacture and accurately calibrate. Both the flow nozzle and the
orifice configurations have larger permanent pressure drops but are relatively simple to
manufacture.
When an orifice flow meter is placed in a pipe, the hole in the orifice essentially
forms a jet which expands to fill the whole pipe at some distance downstream of the plate.
Of course, frictional forces affect the pressure as the air is forced through the hole. In the
absence of viscous effects and under the assumption of a horizontal pipe, application of the
Bernoulli equation between points (1) and (2) in figure 1 gives the volumetric flow rate
through the orifice:

)
(

)


(6)

Figure 1. Schematic of three typical obstruction meters
MEM311: Thermal and Fluid Science Laboratory Experiment 1: Flow Measuring Devices

4 | P a g e

Where V2 is the velocity of the flow immediately after the obstruction meter (in the center
of the vena contracta) and o = d/D1 as labeled in figure 2.

Figure 2. Orifice meter detail

The Orifice Flow Meter
In this lab we will be examining an orifice flow meter. A typical orifice meter is constructed
by inserting between two flanges of a pipe a flat plate with a hole, as shown in figure 2. The
pressure at point (2) within the vena contracta is less than that at point (1).
1
Since the vena
contracta area A2, is less than the area of the hole, Ao, and the turbulent motion near the
orifice plate introduces losses that cannot be calculated theoretically. To take these effects
into account, the orifice discharge coefficient is used. The discharge coefficient is the ratio
of the mass flow rate at the discharge end of the orifice to that of an ideal orifice which
expands an identical working fluid from the same initial conditions to the same exit
pressure.
2
The following equation yields the volumetric flow rate for the orifice by
comparing the pressures on either side of the plate:

)
(

)


(7)
The Inlet Nozzle
The flow pattern for the inlet nozzle used in this experiment is closer to ideal than the
orifice meter flow. There is only a slight vena contracta and the secondary flow separation
is less sever, but there are still viscous effects. These are accounted for by the use of the
nozzle discharge coefficient, Cn, where
MEM311: Thermal and Fluid Science Laboratory Experiment 1: Flow Measuring Devices
5 | P a g e

)
(

)


(8)
with

, d being the inner diameter of the nozzle and | = d/D where D is the outer
diameter of the nozzle. Practically identical to equation 7, note that the pressure drop here
is measured across the nozzle, not the orifice.
Be careful not to confuse the areas, the diameters used, the coefficients and
location of pressures used to determine the respective coefficients.

C. EQUIPMENT

An image of the Armfield F6 Air Flow facility is shown in Fig. 3. The equipment consists of
a long smooth walled pipe (diameter D = 80mm) with an orifice plate of diameter d =
50mm. One end of the pipe is connected to a centrifugal fan via a conical inlet duct while
the other end (inlet nozzle) is open to the atmosphere. The inlet nozzle has an outer
diameter of 120mm and the inner diameter is equal to the pipe diameter. The inlet nozzle
discharge coefficient was determined previously to be

. Pressure taps are located

along the complete length of the pipe to allow measurement of the wall pressure as a
function of length. The centrifugal fan is mounted on a floor-standing metal frame and is
driven by a constant-speed meter. The fan discharge duct terminates is a flow control
damper and jet dispersion orifice gate, which is easily adjustable. This flow control damper
will be used to vary the airflow rate through the tube. Velocities between 0 and 35 m/s can
be obtained with this apparatus by adjusting the position of the flow damper (labeled in
figure 3).
Pressure tap Distance (cm)
1 0
2 7.5
3 31.5
4 79.5
5 137
6 148.5
7 159.5
8 183.5
9 208
10 232
MEM311: Thermal and Fluid Science Laboratory Experiment 1: Flow Measuring Devices

6 | P a g e

Table 1: Pressure tap distances

Figure 3. Armfield F6 flow facility with transducer array
A fourteen-tube manometer will be used to measure the pressures along the pipe. The
manometer is filled with red oil for easier reading. The specific gravity of this oil is 0.86. A
flow splitter (anti-vortex vanes) is fitted to the inlet of the pipe to prevent swirling of the
flow.
This experiment requires that the airflow rate through the pipe be determined
independently of the orifice plate. To accomplish this, a pre-calibrated inlet flow nozzle is
used. The pressure drop across the nozzle is calculated using equation 8.
In addition to the manometers, the pressure taps along the tube are also connected in
parallel to a transducer array. The pressure transducers convert the pressure force to an
electrical voltage which is then read by a signal conditioner and then displayed as inches of
water. There is one pressure transducer for every pressure tap along the tube. The first
display (labeled 1 in figure 3) is dedicated to the first pressure tap (inlet nozzle tapstarts
at gauge pressure, ie. Atmospheric pressure = 0). The second display (labeled 2 in figure 3)
reads the pressures of the remaining pressure taps as selected by the rotary switch (all
read atmospheric pressure to start, do not attempt to zero!. Use pressure taps 2-
10disregard the #1 setting on the switch, as it is dedicated to pressure tap 11, not used in
MEM311: Thermal and Fluid Science Laboratory Experiment 1: Flow Measuring Devices
7 | P a g e

this experiment). The last display (labeled 3 in figure 3) can be read by toggling the switch
next to it to display the pressure difference across the orifice plate (pt 6 pt 5). You must
have the switch toggled to the (#5, #6) position to be able to read the individual pressures
on display 2 using the rotary switch.
The pressure transducers have an uncertainty of 1%.

D. PROCEDURE

Throughout this lab, the manometers and transducers measurements will be used to infer
pressure differences along the pipe. A description of manometer operation is given in
appendix A.

1. Turn on the fan and set a low airflow by closing the flow control damper almost all
the way (do not ever fully close the damper!). Record the level of manometer
tubes and the transducer measurements for pressure taps 1 - 10 on your data
sheet. Be sure to also record the approximation error in your measurements.

2. Repeat step 4 for the remaining 9 damper settings.













MEM311: Thermal and Fluid Science Laboratory Experiment 1: Flow Measuring Devices

8 | P a g e

E. PRE-LAB

1. A 2-in. diameter orifice plate is inserted in a 3-in. diameter pipe. If the water
flowrate through the pipe is 0.90 cfs, determine the pressure difference indicated by
a manometer attached to the flow meter using the figure below with the calculated
Reynolds number.


2. Water flows through the orifice meter shown in the figure below at a rate of 0.10 cfs.
If d = 0.1 ft, determine the value of h.


MEM311: Thermal and Fluid Science Laboratory Experiment 1: Flow Measuring Devices
9 | P a g e


F. DATA ANALYSIS AND REPORTING REQUIREMENTS

1. Calculate the actual orifice volumetric flow rates (m
3
/s) for each of the 10
conditions from the manometer and transducer measurements using equation 6.

2. Calculate the orifice discharge coefficient, Co, defined by equation 7, and the
Reynolds number, defined by equation 4, for each condition as measured by the
manometer and transducers. Use the mean flow velocity,

, to calculate
the Reynolds number.

3. Write a formula for the orifice discharge coefficient Co

in terms of only constants and
the parameters which were directly measured. Make sure to include the
measurement errors from the experiment. For the transducers, multiply the
measurement error by the 1% transducer uncertainty. From this formula derive and
calculate the uncertainty (equation 13 in appendix C) in Co at each flow rate
condition and for both manometer and transducer measurements.

4. On the same graph, plot the orifice discharge coefficient for both manometer and
transducer readings, as a function of the Reynolds number. Be sure to include the
uncertainty error bars in your plots.

5. Plot the wall pressure measurements as a function of distance along the duct for
your 10 flow conditions. You should have 10 plots on a single graph for both
manometer and transducer measurements. In the plots, identify the curves and the
location of the orifice flow meter. What does this plot tell you about the effect of an
orifice flow meter on the air flow through a pipe? Why might this be an important
consideration when designing a piping system? What are some explanations for the
discrepancy between the manometer and transducer measurements?

G. REFERENCES

1. Munson et. al (2009). Fundamentals of Fluid Mechanics, 6
th
Ed., Wiley.

2. R. L. Daugherty and J. B. Franzini (1965), Fluid Mechanics, 6
th
Ed., McGraw-Hill.





MEM311: Thermal and Fluid Science Laboratory Experiment 2: Control Volume Energy and Entropy
Analysis in a Vortex Tube

10 | P a g e

EXPERIMENT 2:
CONTROL VOLUME ENERGY AND ENTROPY ANALYSIS IN A
VORTEX TUBE
by Ertan Agar 2010, adapted from Bakhtier Farouk and David Staack

A. INTRODUCTION AND OBJECTIVES
The vortex tube also known as the Ranque-Hilsch vortex tube is a unique device which
converts a flow of compressed gas into two streams one hotter and the other colder than
the gas supply temperature. It contains no moving parts and the mechanism of its
operation is still a subject of debate, yet the usually agreed upon explanation will be given
herein. This vortex effect was discovered by G. Ranque in 1928. The United States became
focused upon the vortex tube in 1947 when R. Hilsch published a technical paper reporting
research on the device (Ref. 1). Since that time, many technical applications of vortex tubes
for cooling, air conditioning, and drying have been developed (Ref. 2).
The vortex tube is a simple mechanical device that diverts a flow of compressed gas into
two separate streams, one hot and one cold relative to the gas supply temperature. They
are commonly used to prevent thermal damage by providing spot cooling to complex
mechanical or electrical systems. Other technical applications of this technology include air
conditioning, drying, and recovering waste pressure energy from both high and low
pressure sources. The general flow distribution inside a vortex tube is depicted below in
Figure 1.

Figure 4: General Flow Pattern inside a Vortex Tube
MEM311: Thermal and Fluid Science Laboratory Experiment 2: Control Volume Energy and Entropy
Analysis in a Vortex Tube
11 | P a g e

Figure 1-a shows high pressure compressed air entering the vortex tube. The compressed
air accelerates to a high rate of rotation due to its tangential injection point.[1] As a result, a
strong vortex flow is produced inside the tube. Figure 1-b shows hot air exiting the control
volume on the right side of the device. The remainder of the compressed gas is forced to
travel back across the high speed air stream and exit as extremely cold air as shown in
Figure 1-c. Through an energy and entropy analysis of this flow pattern, the first and
second laws of thermodynamics can be validated by experimentally determining the total
rate of entropy creation per mass of air flowing through the vortex tube.
The objective of this experiment is to apply a control volume energy and entropy analysis
to a practical engineering device, a vortex tube. The energy separation phenomenon
induced by the vortex fluid motion will be investigated and explained using basic
thermodynamic principles.

B. THEORY

Principles of Operation
On the basis of flow visualization studies, Hartnett and Eckert (Ref. 3) found that the axial
velocity component (velocity component along the length of the tube) was relatively small
over most of the radius of the tube. Therefore, the flow can be analyzed by evaluating one
plane through the vortex tube perpendicular to the tube axis as shown in Figure 3.
Hartnett and Eckert also observed that the flow consisted of a colder region in the center of
the tube that rotated as a solid body having a circumferential velocity, v
u
= e r, where e is
the constant angular velocity of the fluid. In fluids, this type of rotation is called a forced
vortex because it is vortical flow, which is induced by an external force, in this case, the
outer stream. In the outer stream, the circumferential velocity is proportional to 1/r and
therefore decreases as r increases, i.e., v
u
= K/r where K is a constant. This type of vortex is
called a free vortex, a common example of which is the vortical motion of water as it goes
down the drain in a bath tub. In a true free vortex, the circumferential velocity goes to zero
as r goes to infinity. Therefore, the outside stream only approximates a free vortex because
r must be less than or equal to rw, the radius of the tube. A flow with a forced vortex inside
and a free vortex outside is called a combined vortex and has a circumferential velocity
profile given by the following equations
v
u
= e r r s ro (1a)

v
u
= K/r r > ro (1b)
MEM311: Thermal and Fluid Science Laboratory Experiment 2: Control Volume Energy and Entropy
Analysis in a Vortex Tube

12 | P a g e

where ro is the radius of the central flow, as shown in Figure 2. Given this velocity profile,
we can determine why the temperature separation occurs and why the hotter outer stream
surrounds the cooler inner core.
We will begin this analysis by evaluating the pressure distribution from the centerline of
the tube to the outer wall. The streamlines for the flow in the vortex tube form closed
concentric circles. Evaluating F = ma normal to a streamline, as shown in Ref. 4, Section
3.3, we find the change of pressure in a direction normal to a streamline is give by
R
V
n
p
2

=
c
c
(2)
where R is the local radius of curvature of the streamline and V is the velocity along the streamline.
However, the streamlines in a combined vortex form closed concentric circles, as shown in
Figure 3. Therefore, R = r and the velocity, V, is simply vq. Also, in Eq. (2), the positive n
direction points toward the inside of the curved streamline, i.e., opposite to the positive
direction of the radial coordinate, r. Therefore,
r n c
c
=
c
c
(3)
Figure 2: Velocity profile in a vortex tube
MEM311: Thermal and Fluid Science Laboratory Experiment 2: Control Volume Energy and Entropy
Analysis in a Vortex Tube
13 | P a g e

Substituting these results into Eq. (2), we find the radial pressure distribution in a vortex to
be
r
v
n
p
2
u

=
c
c
(4)

Figure 3: Streamlines in a Vortex Tube
Substituting Eqs. (1a) and (1b), respectively, into Eq. (4), we obtain
r
r
p
2
e =
c
c
(5a)
and
3
2
r
K
r
p
=
c
c
(5b)
Equations (5a) and (5b) show that in both the free and forced vortex regions, the pressure
increases as r increases. Integrating these equations with respect to r, starting with a
known pressure p = p1, we find the pressure distribution in the vortex tube to be given by
1
2 2
p r 2 / 1 p + e =
r s ro (6a)
) r / 1 r / 1 ( K 2 / 1 p
2 2
o
2
+ = r > ro (6b)
where po is the pressure at r = ro. The value of po is found by evaluating Eq. (6a) at r = ro to
obtain
1
2 2
p r 2 / 1 p + e =
(7)
MEM311: Thermal and Fluid Science Laboratory Experiment 2: Control Volume Energy and Entropy
Analysis in a Vortex Tube

14 | P a g e

This pressure profile is plotted as a function of r in Figure 4. From this figure, we see that
the pressure is lowest at the center of the tube and increases to a maximum at the wall.
Herein lies the reason for the temperature difference between the inner and outer streams.
Recall the piston-cylinder devices that were studied extensively in basic thermodynamics.
The first law energy balance indicates that if a gas is adiabatically (no heat transfer)
compressed in a piston-cylinder, the internal energy and hence, the temperature, must
increase. As the gas enters the vortex tube, the viscosity of the fluid induces a vortical
motion which creates a forced vortex at the center of the tube. This flow produces the
pressure distribution given by Eqs. (6a) and (6b). The gas on the outside of the tube is
adiabatically compressed resulting in an increase in temperature. This gas is also at a
higher pressure so it can be drawn off at the control valve. The work to compress the outer
gas came from the gas near the centerline which is adiabatically expanded and cooled. The
cooler gas is confined to the inner core of the tube so it can be withdrawn from the
opposite end of the tube through an orifice plate. Note that even though the pressure is
lowest at the center of the tube, it is still greater than atmospheric pressure and will flow
out of the orifice. Therefore, the separation of the gas into two streams having different
temperature is caused by viscous forces in the gas which induces a pressure distribution in
the tube. The gas in the high pressure region is compressed and heated while that in the
low pressure region is expanded and cooled.

Figure 4: Radial pressure distribution
This description of the operation of a vortex tube resulted only after many experimental
observations and a detailed analysis. The first and second laws of thermodynamics,
however, present us with a simple way to evaluate any thermodynamic system to
determine whether it is thermodynamically valid. If we ever determine that a proposed
process violates either the first or second law, we know that the process is impossible. In
this experiment, you will perform a first and second law analysis of a vortex tube to
examine its performance.

MEM311: Thermal and Fluid Science Laboratory Experiment 2: Control Volume Energy and Entropy
Analysis in a Vortex Tube
15 | P a g e

First and Second Law of Thermodynamics for a Control Volume
All thermodynamic analyses begin by defining the system to be evaluated. A vortex tube is
an example of a steady-state, steady-flow device and is most easily represented by a control
volume. Reference 6, Section 5-4 (Reference 5, Section 4.5) gives the general first law
energy balance for such a control volume as

(8)
where is the heat, the work,

the mass flow rate in and out respectively,

the specific enthalpy of mass in and out,

the velocity of mass in and out,

the
elevation of mass in and out, and lastly, is gravitational acceleration. (The form of the first
law of thermodynamics given in Ref. 5 is slightly different than that given in Eq. 8.
Primarily, Ref. 5 defines the work done by a system as positive (ASME sign convention)
whereas this work is negative in Ref. 6 (scientific sign convention). The equations are
equivalent and you must be able to use both of them. The second law of thermodynamics is
given by the entropy generation principle for a steady-state, steady-flow control volume
(see Ref. 5, Section 6.2 or Ref. 6, Section 7-2) as

( )

( )

(9)
where

is the total rate of entropy change for the control volume, and

is the rate
of internal generation of entropy within the system (intrinsic entropy associated with
matter,

for irreversible systems,

for reversible systems, and system is

impossible if

). In this experiment, pressurized air is used to drive the vortex tube.

The ideal gas equations can therefore be used to evaluate Eqs. (8) and (9). See Ref. 5,
Section 2.8 and Section 5.9 (Ref. 6, Chapter 3 and Section 7-4) to review the application of
these equations to ideal gases. Pay particular attention to the evaluation of the change in
entropy for an ideal gas.







MEM311: Thermal and Fluid Science Laboratory Experiment 2: Control Volume Energy and Entropy
Analysis in a Vortex Tube

16 | P a g e

C. EQUIPMENT

Figure 5: Experimental Setup
Figure 5 shows the experimental setup and all the components and measuring devices.
Compressed air enters the system at the top left of the picture and goes through a de-
humidifier, then a pressure control and a pressure gage. Then it passes through a mass flow
meter. See Figure 6 for a close-up of the inlet setup.

Figure 6: Inlet setup
MEM311: Thermal and Fluid Science Laboratory Experiment 2: Control Volume Energy and Entropy
Analysis in a Vortex Tube
17 | P a g e

After the flow enters the system and has the properties desired, it then goes to the vortex
tube and separated into a hot and cold stream. Below Figure 7 is a close-up picture of the
vortex tube itself and the direction of the hot and cold streams.

Figure 7: Vortex Tube
After leaving the vortex tube both streams pass through a series of measurement devices
similar to the ones monitoring the inlet flow. For the cool stream there is a mass flow
meter, a temperature indicator connected to a thermocouple and a pressure gage, see
Figure 8. A throttle valve is used to control the cold air pressure.

Figure 8: Cold stream measuring device
The cold stream uses a pressure gauge and a temperature indicator in conjunction with a
digital mass flow meter. For the hot stream, pressure and temperature are similarly
measured, but a rotameter is used to measure mass flow rate. A rotameter has a ball in
between two tapered tracks and the air pressure flowing through the meter pushes the ball
up to a certain height and the height markings on the vertical meter correlate to a flow rate
(See Appendix B for rotameter function). A throttle valve and a muffler are also used on
both outlet air streams. See Figure 9 for the layout of the hot stream measuring devices.
MEM311: Thermal and Fluid Science Laboratory Experiment 2: Control Volume Energy and Entropy
Analysis in a Vortex Tube

18 | P a g e


Figure 9: Hot stream measuring devices
D. PROCEDURE
1. Familiarize yourself with the general performance of the vortex tube by
manipulating the pressure regulating valve and the discharge throttling valves
found on the hot and cold ends.
2. Close the hot flow valve and open cold valve as much as possible to experimentally
determine the maximum inlet pressure for the hot flow stream. This will allow for a
complete range of on-scale flow readings for all throttling valve positions.
3. Set the inlet pressure of the main flow to 60 psi. Open the hot stream flow,
establishing a constant mass flow rate. Take 5 readings at even intervals of cold
mass flow rate values.
4. Record all temperatures, pressures, and flow rates at steady state (typically takes 4-
5 minutes) for 5 evenly spaced cold mass flow rate values.
5. Repeat steps 2-4 for the hot mass flow rate intervals.
6. Repeat all experiments at 80 psi.

E. PRE-LAB
1. Complete the assigned thermodynamic problems. These are intended to provide a
review of the control volume analysis using the First and Second Law of
Thermodynamics.

A vortex tube has an air inlet flow at 20
o
C, 200 kPa and two exit flows of 100 kPa, one
at 0
o
C and the other 40
o
C. The tube has no external heat transfer and no work all the
MEM311: Thermal and Fluid Science Laboratory Experiment 2: Control Volume Energy and Entropy
Analysis in a Vortex Tube
19 | P a g e

flows are steady and have negligible kinetic energy. Find the fraction of the inlet flow
that comes out 0
o
C. Is this setup possible?

A Hilch (Vortex) tube has an air inlet mass flow of 50 SLPM at 20
o
C, 200 kPa and two
exit flows of 100 kPa one at 0
o
C and the other t 40
o
C. The tube has no external heat
transfer and no work and all the flows are at steady state and have negligible kinetic
energy. Find the fraction of the inlet flow that comes out at 0
o
C. Is this setup possible?

2. After reviewing the lab procedures, prepare a data sheet on which to record your
experimental measurements (pressure, temperature and flow rate for main stream, hot
stream and cold stream). Bring this sheet to the lab to record your data.

F. DATA ANALYSIS AND REPORTING REQUIREMENTS
1. Calculate the volumetric and mass flow rates from the rotameter and mass flow-meters
data. All volumetric flow rates (both cold plus main inlet flows) must be corrected to the
actual pressure and temperature of the flowing gas to obtain an accurate volumetric flow
rate. Hint: ideal gas state equation may be used.

2. Perform a mass balance for each set of operating conditions with using mass flow meters
data to determine the flow rate of rotameter. Then compare the calculated value with the
measured flow rate of rotameter obtained in the experiment. Use graph to show the relative
error.

3. Perform an energy balance (Eq. (8)) to solve for the rate of heat transfer for each set of
operating conditions from the control volume of the vortex tube. Discuss your result with
drawing heat transfer rate graph for different conditions.

4. Evaluate the total rate of entropy generation (Eq. (9)) flowing through the control volume
using your measurements and heat transfer rate calculated in 3 above. The property data
required can be found in Ref. 5 and should be included in the sample calculations. Discuss
your result with drawing rate of entropy generation graph for different conditions.

5. Do your results satisfy the Second Law of Thermodynamics (Increase-in-Entropy
Principle)? What does this imply about the process that occurs in a vortex tube?



MEM311: Thermal and Fluid Science Laboratory Experiment 2: Control Volume Energy and Entropy
Analysis in a Vortex Tube

20 | P a g e

G. REFERENCES
1. Hilsch, R., The Use of the Expansion of Gases in a Centrifugal Field as a Cooling Process,
Review of Scientific Instruments 18, 1947.
2. Hartnett, J.P. and Eckert, E.R.G., Experimental Study of the Velocity and Temperature
Distribution in a High-Velocity Vortex-Type Flow, Transactions of the ASME 79, 1957, pp.
751-758.
3. Eckert, E.R.G. and Drake, R.M., Jr., Analysis of Heat and Mass Transfer, McGraw-Hill, New
York, 1972, pp. 427-430.
4. Munson, B.R., Young, D.F., and Okiishi, T.H., Fundamentals of Fluid Mechanics, John Wiley
and Sons, 1990.
5. Black, W.Z. and Hartley, J.G., Thermodynamics (2
nd
edition), Harper Collins Publishers, New
York, 1991.
6. Wark, Kenneth, Jr., Thermodynamics (5
th
edition), McGraw-Hill, New York, 1988.
7. Holman, J.P., Experimental Methods for Engineers. 3
rd
ed., McGraw-Hill, 1978.

MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

21 | P a g e

EXPERIMENT 3: HEAT TRANSFER FROM A CIRCULAR
CYLINDER
by Eric Wargo 2010, adapted from Bakhtier Farouk and David Staack

A. OBJECTIVES

The objective of this experiment is to determine the natural convection, forced convection
and radiation heat transfer (qnatural convection + qforced convection + qradiation) from an electrically
heated horizontal cylinder. These values will be compared to existing heat transfer
correlations provided herein.

In this experiment, you will demonstrate how heat transfer from a heated surface to a
quiescent environment is a combination of several mechanism of heat loss. The relative
magnitudes of the natural convection, forced convection, and radiation heat transfer
coefficients depend on the surface temperature and flow velocity. Radiation becomes more
important as the surface temperature increases. Forced convection becomes more
important as the flow velocity increase. The problem will be analyzed using a control
volume under equilibrium conditions. For equilibrium, heat input to a surface must equal
the heat transferred from the surface to the surroundings.

B. THEORY

Natural and Forced Convection

Free convection heat transfer occurs whenever a body is placed in a fluid at a higher or
lower temperature. As a result of the temperature difference, heat is transferred between
the fluid and the body and causes a change in the density of the fluid layers in the vicinity of
the surface. This difference in density leads to an upward flow of the lighter fluid (Figure
5). If the motion of the fluid is caused solely by differences in density resulting from
temperature gradients, the associated heat transfer mechanism is called free or natural
convection. If the fluid motion is enhanced using a fan or otherwise forced by some device,
the heat transfer mechanism is called forced convection (Figure 6). Because the fluid
velocity is usually less in free convection than in forced convection, the rate of heat transfer
from a surface is also generally less. In this experiment, you will measure and compare the
magnitudes of the heat transfer rates for forced convection and free convection
configurations.

MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

22 | P a g e


Figure 5. Cross sectional view of a heated cylinder under natural convection.

Figure 6. Cross sectional view of a heated cylinder under forced convection.
The rate of heat transfer by convection (both forced and free) between a surface and a fluid
may be computed using the relation

) (1)

where

is the average convective heat transfer coefficient, A is the area available for heat
transfer, Ts is the surface temperature, and T is the ambient temperature. The relation
expressed by Equation 1 was originally proposed by the British scientist, Sir Isaac Newton
in 1701. Therefore, it is sometimes referred to as Newtons law of cooling. Even though
this equation has been used for many years to evaluate convective heat transfer, it is
actually more a definition of

than a law of convection. If the value of

is known for a
certain flow configuration, the evaluation of Equation 1 to determine the rate of heat
MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

23 | P a g e

transfer is straightforward. However, determining the appropriate convective heat
transfer coefficient is difficult because convection is a very complex phenomenon. The
value of

depends not only on the geometry of the surface of the object (both macroscopic
and microscopic surface characteristics) but on the velocity and physical properties of the
fluid, all of which affect the conditions on the boundary layer. Since these quantities are not
necessarily constant over a surface, the heat transfer coefficient may vary from point to
point. For this reason, we must distinguish between a local and average convective heat
transfer coefficient. The local coefficient, hc, is defined by

) (2)

while the average coefficient,

, can be defined in terms of the local value by

(3)

The primary problem in either forced or free convection is to determine the appropriate
local or average heat transfer coefficient. Many experiments have been performed to
measure these coefficients for a wide variety of geometries and flow configurations.
Numerical calculations have only recently become sufficiently exact to calculate the heat
transfer coefficients directly. However, heat transfer coefficients can be accurately
calculated for relatively simple flow configurations. Complex configurations must still be
determined experimentally. In this experiment, you will determine the heat transfer
coefficient of free convection and forced convection from a cylinder placed within a range
of flow conditions. Your results will be compared with existing heat transfer correlations.

Radiation Heat Transfer

Thermal radiation is heat transfer by the emission of electromagnetic waves which carry
energy away from the emitting object. For ordinary temperatures (i.e. less than red hot),
the radiation is in the infrared region of the electromagnetic spectrum. The relationship
governing radiation from hot objects is called the Stefan-Boltzmann law. The heat
transferred into or out of an object by thermal radiation is a function of several
components. These include its surface emissivity, surface area, temperature, and geometric
orientation with respect to other thermally participating objects. The heat loss rate caused
by radiation from a heated surface to the surroundings can be calculated by

) (4)

where is the Stefan-Boltzmann constant (= 5.67x10
-8
W/ m
2
K
4
), is the emissivity of the
surface, and Fsa is the view factor of the surface. The surface temperature is given by Ts,
and the temperature of the body receiving the radiation (the ambient environment) is
given by T. An object's surface emissivity is a function of its surface microstructure. The
view factor takes into account the geometric orientation of the surface to the external
environment. If all of the radiation emitted by the surface has a direct line of sight to the
MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

24 | P a g e

external environment, the view factor is equal to 1. Equation 4 can also be cast in the form
of Newtons law of cooling (Equation 1) as

) (5)

Comparing Equation 4 with Equation 5 we can define a radiative heat transfer coefficient as

)
(

)
(6)

Natural Convection, Forced Convection, and Radiation Heat Transfer

If a surface is at a temperature above that of its surroundings and is located in stationary or
moving air, heat will be transferred from the surface to the surroundings. This transfer of
energy will be a combination of natural convection, forced convection (if there is a driven
air flow) and radiation to the surroundings. As described above in natural convection, the
motion of the fluid is caused solely by differences in density resulting from the temperature
gradients. If the fluid motion is enhanced using a fan or otherwise forced by some device,
the heat transfer mechanism is called forced convection. Radiation heat transfer generally
becomes significant at surface temperatures well above room temperature. However, it
does play a role at lower temperatures and should be accounted for, especially when
comparing measurements with existing correlations for natural and forced convection.
Heat loss by conduction would normally be included in the analysis of a real application. In
this experiment, it is minimized by the design of the equipment and experimental
procedures.

The total heat lost by a surface can thus be presented as a linear superposition of the
aforementioned heat losses:

(7)

When there is no forced flow, the

term can be neglected. When there is an air flow

typically the heat lost due to the forced convection is significantly greater than that by
natural convection; thus, the

term can be neglected. Equation 7 can be written in

terms of the various heat transfer coefficients:

) .

/ (

) (8)


MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

25 | P a g e

Experimental Determination of the Heat Transfer Coefficients

Solving Equation 8 for the heat transfer coefficient, we obtain

)
(9)

The heat transfer coefficient can be determined by measuring all the quantities on the right
hand side of Equation 9 for a specific flow configuration and solving for

. The
temperatures Ts and T are easily measured using thermocouples. The area, A, is simply the
surface area available for convective heat transfer (sometimes called the wetted area). This
can be evaluated once the experimental configuration is defined. Therefore, the problem is
reduced to determining the rate of heat transfer,

. Recall that the first law of

thermodynamics for a control mass can be written in infinitesimal form as

(10)

Recall from thermodynamics that passing an electrical current, I, through an object is a
form of work given as

(11)

where V is the applied voltage, and R is the resistance. The relations in Equation 11 are
based on Ohms laws, V = IR and P = VI, where P is the power. Substituting Equation 11 into
Equation 10, dividing by a small increment of time t, and taking the limit as t goes to 0,
we obtain

(12)

Where is the rate of heat transfer and (V
2
/R) is the rate at which work is done. Since the
heated cylinder can be considered to be an incompressible substance, the change in its
internal energy is given as

(13)

where m is the mass of the cylinder, and c is the specific heat. Recall that for an
incompressible substance, cp = cv = c (see Section 2.8 of Reference [1], or Sections 4-7 of
Reference [3] for a review of these thermodynamic relationships). Since m and c are
constant, the rate of change of internal energy is

(14)

MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

26 | P a g e

If the electrical current to the cylinder is controlled so that the temperature remains
constant,

, and therefore,

, is identically equal to zero. Such a situation happens in

steady state operation. For this condition, Equation 12 reduces to the following form:

(15)

The negative sign in Equation 15 indicates that the heat transfer is out of the cylinder.
Substituting Equation 15 into Equation 9 yields

)
(16)

The surface temperature, TS, and the temperature of the cylinder, T, are interchangeable,
because it is assumed that there are no temperature gradients inside the cylinder. This is
true only if the cylinder is made of a material that conducts heat rapidly. The Biot number,
defined as

(17)

provides a measure of the accuracy of this assumption. In Equation 17, L is the
characteristic dimension (cylinder diameter, around which the fluid is flowing) and ks is the
thermal conductivity of the material. Physically, the Biot number is the ratio of the external
convective heat transfer rate to the internal conductive heat transfer rate. If Bi<<1, heat is
conducted within the material much faster than it is convected away from the cylinder. The
assumption of uniform temperature is then valid and Equation 16 can be used to evaluate
the overall rate of heat transfer. If Bi is greater than approximately 0.1, the non-
uniformities within the material must be accounted for when determining the heat transfer
coefficient.

Equation 16 shows that the heat transfer coefficient can be calculated by measuring the
electrical voltage and current supplied to the cylinder such that the cylinder temperature of
the cylinder is constant. Comparing Equation 16 and 8 we can simply write this as

(18)

which simply states that in steady state the heat flow out of the cylinder is equal to the
energy flow into the cylinder.

Heat Transfer Correlations

As shown by the equations above, the heat transfer coefficient is a dimensional number
that depends on the area (size) of the object being evaluated and the
temperature/properties of the fluid in which the object is immersed. When developing
MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

27 | P a g e

experimental correlations, it is convenient to non-dimensionalize

so that the results are

applicable to other configurations. The Nusselt number is a dimensionless heat transfer
parameter defined as

(19)

where L is the characteristic dimension of the geometry being evaluated (cylinder or
sphere diameter, around which the fluid is flowing), and kf is the thermal conductivity of
the fluid in which the object is immersed. The Nusselt number is defined as the ratio of
convective to conductive heat transfer across a boundary (the objects surface); thus,
radiation effects should not be included in its calculation. The overbar on the Nusselt
number indicates that it is formed from the average heat transfer coefficient and, therefore,
is a measure of the average heat transfer from the object.

Other dimensionless parameters are used that represent the air flow conditions. For forced
convection flows, the Reynolds number is used to correlate heat transfer data. The
Reynolds number is defined as

(20)

where is the density, V is the velocity, L is the characteristic length (cylinder diameter,
around which the fluid is flowing), is the absolute viscosity, and is the kinematic
viscosity and is equal to /.

In the case of natural convection, Nu depends on the Rayleigh number, Ra. The Rayleigh
number, in turn, is frequently written in terms of two other non-dimensional numbers,
namely the Grashof and the Prandtl numbers, Gr and Pr respectively. These are given by
the equations:


(

(21)

(22)


(

(23)

The new symbols in Equations 21 23 are explained in Table 2. Note: D is the
characteristic dimension (cylinder diameter, around which the fluid is flowing).

MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

28 | P a g e

Table 2. Symbol definitions for Equations 21 23.
Symbol Definition Value and/or Units
g gravitational acceleration 9.81 m/s
2
volume expansion coefficient, 1/Tfilm K
-1

kinematic viscosity, / m
2
/s
cp specific heat J/kgK
density kg/m
3
thermal diffusivity m
2
/s
k thermal conductivity W/mK

All of the thermophysical properties given in Equations 1923 are functions of the
temperature of the fluid. Since the temperature varies from the surface to the ambient, by
convention these properties are evaluated at the film temperature

(24)

Values of the thermophysical properties listed in Table 2 can be found in Appendix A of
Reference [2] or any other heat transfer textbook. At the very least, a reputable source
should be utilized and properly referenced in the lab report; online calculators are
convenient, but their range of accuracy may be questionable.

These equations form the basis for the experimental method to experimentally evaluate the
heat transfer coefficient. A more detailed discussion of forced and free convective heat
transfer can be found in Reference [2], Chapters 7 and 9, respectively. Specifics regarding
the heat transfer from a circular cylinder in cross flow may be found in Reference [2],
Section 7.4 (forced convection) and Section 9.6.3 (free convection).

Reference [2] gives the following empirical correlation for the Nusselt number, Nu, for
forced convection as a function of the Reynolds number:

()

(25)

where c and n are obtained from Table 3.

MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

29 | P a g e

Table 3. Reynolds number-Nusselt number empirical relation parameters.
Re c n
1 40 0.75 0.4
40 1000 0.51 0.5
10
3
2x10
5
0.26 0.6
2x10
5
10
6
0.076 0.7


Reference [3] gives the following empirical correlation for the Nusselt number, Nu, for
natural convection as a function of the Rayleigh number:

()

(26)

where c and n are obtained from Table 4.

Table 4. Raleigh number-Nusselt number empirical relation parameters.
Ra c n
10
-10
to 10
-2
0.675 0.058
10
-2
to 10
2
1.02 0.148
10
2
to 10
4
0.850 0.188
10
4
to 10
7
0.480 0.250
10
7
to 10
12
0.125 0.333

For a given Ra, the Nu value can be computed from Equation 15 and the corresponding
value from Equation 9.

There are other equations that are sometimes used to calculate the natural convection heat
transfer coefficient. One simplified, analytical relation that may be used to calculate the
average heat transfer coefficient from a horizontal cylinder is given by

0
(

(27)

where D is the cylinder diameter. Equation 27 yields

in W/m
2
K if the
temperatures are in Kelvin and D is in meters. In the experiment, you will compare your
MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

30 | P a g e

experimental results with the predictions of Equations 25, 26, and 27. Note: care must be
take to distinguish Ra and Re.

C. EQUIPMENT

1. Armfield HT14 Combined Convection and Radiation Accessory (see Figure 7 and Figure
8)
2. Armfield HT10X Heat Transfer Service Unit (see Figure 7 and Figure 9)
3. Computer on site or laptop for manual data entry (optional but highly recommended,
since you will be recording the surface temperature in 30 second time intervals for
durations sometimes totaling over 10 minutes)


Figure 7. Experimental set-up.
MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

31 | P a g e


Figure 8. Schematic View of the Armfield HT14
Combined Convection and Radiation Accessory.

Figure 9. Armfield HT10X Unit.
MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

32 | P a g e

The HT14 accessory consists of a centrifugal fan with vertical outlet duct at the top of
which is mounted a heated cylinder. The heated cylinder has an outside diameter of 10mm,
a heated length of 70mm and is internally heated throughout its length by an electric
heating element. The heating element is rated to produce 100 Watts nominally at 24V DC
into the cylinder. The power supplied to the heated cylinder can be varied and measured
on the HT10X unit. The mounting arrangement for the cylinder in the duct is designed to
minimize loss of heat by conduction to the wall of the duct. The surface of the cylinder is
coated with heat resistant paint which provides a consistent emissivity close to unity. A K-
type thermocouple is attached to the wall of the cylinder (T10), at mid position, allowing
the surface temperature to be measured under the various operating conditions.

A variable throttle plate at the inlet to the fan allows the velocity of the air through the
outlet duct to be varied, and a vane type anemometer within the fan outlet duct allows the
air velocity in the duct to be measured over the range 0-7 meters/sec. The inside diameter
of the outlet duct is 70mm (matching the length of the heated cylinder). A K-type
thermocouple is also located in the air duct (T9), allowing the ambient air temperature to
be measured upstream of the heated cylinder.

The HT10X unit is used to control the heating of the cylinder. The voltage is controlled by a
rotary knob, and the top digital display can be used to alternatively view the voltage,
current, and air velocity by using the selector knob. The temperature at the thermocouples
is shown on the bottom digital display, and the thermocouple selector knob determines
which temperature is displayed. Connections for additional accessories and computer data
acquisition and control are available on the service unit but are not used in this experiment.

D. PROCEDURE

***Before operating the equipment, please understand that the heated cylinder will
reach temperatures above 500C. Serious skin burns will result if the equipment is
mishandled. Please ask an instructor if you have any questions or concerns.

Equipment Set-up

Before proceeding with the exercise, ensure that the equipment has been prepared as
follows:

a. Locate HT14 Combined Convection and Radiation accessory alongside the HT10X Heat
Transfer Service Unit on a suitable bench.
b. Ensure that the horizontal cylinder is located at the top of the vertical metal duct with
the T10 thermocouple attached.
c. Connect the thermocouple attached to the horizontal cylinder to socket T10 on the front
of the service unit.
d. Connect the thermocouple located in the vertical duct to socket T9 on the service unit.
MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

33 | P a g e

e. Set the voltage control potentiometer to minimum (counterclockwise) and the selector
switch to MANUAL.
f. Connect the power lead from the heated cylinder on the HT14 to the socket marked O/P3
at the rear of the service unit.
g. Ensure that the service unit is connected to an electrical supply.

Experimental Procedures

You will perform 2 sets of experiments: one for natural convection, and another for forced
convection. For the natural convection conditions, you will measure the steady state
temperature achieved on the cylinder for 5 different heating voltages (4, 7, 10, 13, and 16
volts). For the forced convection cases, you will measure the steady state temperature
achieved on the cylinder for 5 different flow velocities (approximately equally spaced
between 0 m/s and the maximum attainable by the device), all at a heating voltage of 10
volts.

Perform the forced convection experiments first, as they achieve steady state more quickly.
Record the temperatures at time intervals of 30 seconds to be able to identify when steady
state has been achieved (steady state can be considered achieved when the surface
temperature (T10) remains constant for 2 minutes or four, 30 second readings).

1. Prepare a data sheet on which to record the raw data. It is highly recommended that you
record your data directly into a Microsoft Excel spreadsheet, since you will be recording the
surface temperature in 30 second time intervals for durations sometimes totaling over 10
minutes. The spreadsheet should have the following headings:

Table 5. Variables to record in datasheet.
Variable Symbol Units
Time t min:sec
Heater voltage V volts
Heater current I amps
Upstream air temperature, T9 T C or K
Cylinder surface temperature, T10 Ts C or K
Upstream air velocity Ua m/s

2. Turn on the front main switch (Figure 7) (if the panel meters do not illuminate, check
the circuit breakers at the rear of the service unit). All switches at the rear should be up.
MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

34 | P a g e

3. Turn on the fan switch and open the throttle plate until the knob stops turning, ensuring
the maximum air velocity is attained.
4. Set the heater voltage to 10 volts.
5. Record the experimental conditions as a function of time (30 second intervals) until a
steady state is achieved. Steady state can be considered achieved when the surface
temperature (T10) remains constant on the display for 2 minutes (four, 30 second
readings).
6. When the temperature is stable, record V, I, T9, T10, and Ua. Note: the digital display
outputs temperature readings in degrees Celsius.
7. Repeat steps 5 and 6 for an additional four flow velocities: 75%, 50%, 25%, 0% (letting
the heater remain at 10 volts). A flow velocity of 0% is best attained by shutting the fan
off. The throttle plate should be opened completely to allow for air to flow naturally
through the duct during free convection.
8. The last condition for forced convection above represents the first free convection
condition. Repeat steps 5 and 6 for heater voltages of 4, 7, 13, and 16 volts. In all you
should measure 9 conditions, and if you deem appropriate you may change the order.

Suggestion: This lab can be completed the fastest if (after obtaining data for forced
convection at 100% 25% flow velocity) the heated cylinder is allowed to cool back
down by turning the voltage off and throttling the flow velocity back to 100% for a few
minutes. Then, repeat steps 5 and 6 for heater voltages of 4, 7, 10, 13, and 16 volts with
the fan off and throttle plate open completely. This ensures that the cylinder is only ever
heating up to steady state for all conditions, rather than cooling down naturally (which
takes far more time).

The order will be as follows:

Voltage 10 10 10 10 cool
down
4 7 10 13 16
Flow rate 100% 75% 50% 25% 0% 0% 0% 0% 0%

9. Return the voltage to 0 volts, and turn off the unit.

Note: Do not set the heater voltage in excess of 16 volts when operating the cylinder in the
natural convection mode (no forced airflow). The life of the heating element will be
considerably reduced if operated at excessive temperature. If temperatures approach 550C
check with an instructor, and if the temperatures exceed 600C shut off the unit.






MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

35 | P a g e

Use the following data for your calculations:

Diameter of horizontal cylinder D = 0.01 m
Heated length of cylinder L = 0.07 m
Emissivity of surface of the cylinder = 0.75
Stefan-Boltzmann constant = 5.67x10
-8
W/m
2
K
4

View Factor Fs a = 1

E. PRE-LAB

Assume an experimental setup similar to the Armfield facility and an ambient temperature
of 25C.

1. What is the proper temperature unit (C or K) to be used in Equation 4? Why?
2. For a cylinder surface temperature of 500C, use Equation 4 to calculate the power lost
by radiative heat transfer.
3. For a cylinder surface temperature of 500C, calculate a Raleigh number using Equation
23 and appropriate properties based on the film temperature. Use this Raleigh number
to calculate a Nusselt number by Equation 26 and Table 4. Use Equation 19 and the
Nusselt number to calculate a corresponding free convection heat transfer coefficient for
the cylinder. Now use Equation 1 to calculate the heating power which is dissipated by
natural convection at this surface temperature.
4. For a cylinder surface temperature of 500C and a flow velocity of 5 m/s, calculate a
Reynolds number using Equation 20 and appropriate properties based on the film
temperature. Use this Reynolds number to calculate a Nusselt number by Equation 25
and Table 3. Use Equation 19 and the Nusselt number to calculate a corresponding
forced convection heat transfer coefficient for the cylinder. Now use Equation 1 to
calculate the heating power which is dissipated by forced convection at this surface
temperature.
5. Sum the powers lost by radiation, forced convection and free convection to get the total
heating power required to maintain this temperature. What percentage are the various
mechanisms of the total power? If there were no forced convection, how do radiation
and free convection compare? Briefly discuss these results.
6. After reviewing the lab procedures, prepare a data sheet to record your experimental
measurements. It is highly recommended that you create a data sheet in Microsoft Excel
and bring it to lab along with a laptop, so you can quickly key in temperature readings
during the lab. Bring this data sheet and/or a laptop computer to the lab to record your
data.


MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

36 | P a g e

F. DATA ANALYSIS AND REPORT REQUIREMENTS

1. For all 9 test conditions, plot the measured temperature as a function of time. How long
does steady state take to attain? What conditions determine steady state? What methods
could be employed to attain steady state faster?
2. For each of the five test conditions corresponding to natural convection:
a. Calculate and tabulate the following parameters using the raw data and provided theory.

Cylinder temperature

K
Film temperature Tfilm K
Rayleigh number Ra
Heat flow (power input)

W
Heat transferred by radiation

W
Heat transfer coefficient (radiation)

W/m
2
K
Heat transferred by natural convection

W
(Us Equation 18 and assume that forced convection is negligible)
Heat transfer coefficient (natural convection)

W/ m
2
K
Nusselt Number (natural convection) Nu

b. Compute the heat transfer coefficient and Nusselt number (natural convection)

from
the analytical relation, Equation 27. Compute a Nusselt number and heat transfer
coefficient from the calculated Raleigh number and empirical correlation, Equation 26.
Use Equation 19 to switch between h and Nu.
c. Compare the analytical and empirical calculations of

and

obtainable from
Question 2.b with the values for

and

calculated only from the experimental

measurements (Question 2.a) and discuss any differences in them. Plot the measured
and calculated values of

versus the surface temperature

on a single graph. [hint:

there should be 3 series on the plot]
d. On a single graph plot

, and

as a function of the average surface
temperature

. Comment on the relationship between

and

. Comment on the
relative importance of natural convection and radiation heat transfer. At what
temperature (if any) is

?
3. For each of the test conditions corresponding to forced convection:
a. Calculate and tabulate the following parameters using the raw data and equations given
above. (For cases where cylinder temperature distribution was not measured use the
temperature measurement at the steady state condition)

Cylinder temperature

K
Film temperature Tfilm K
Reynolds number Re
Heat flow (power input)

W
Heat transferred by radiation

W
MEM311: Thermal and Fluid Science Laboratory Experiment 3: Heat Transfer from a Circular Cylinder

37 | P a g e

Heat transfer coefficient (radiation)

W/ m
2
K
Heat transferred by forced convection

W
(Us Equation 18 and assume that natural convection is negligible)
Heat transfer coefficient (forced convection)

W/ m
2
K
Nusselt Number (forced convection) Nu

b. Compute a Nusselt number and heat transfer coefficient from the calculated Reynolds
number and empirical correlation, Equation 25. Use Equation 19 to switch between h
and Nu.
c. Compare the correlation calculated values of

and

obtainable from Question

3.b with the values for

and

calculated only from the experimental

measurements (Question 3.a) and discuss any differences in them. Plot the measured
and calculated values of

versus the surface temperature

on a single graph.
d. Plot

as a function of the flow velocity. Comment on the relationship.

e. Plot Nu vs. Re from Equation 25 for Re from 100 to 5000. On the same graph plot Nu vs.
Re attained from the experiment (Question 3.a). How close are the experimental data
points to the correlation line? Comment on the non-dimensional comparison.
4. For this lab, the surface temperature of the heated cylinder was only measured at a
single point. If several temperature readings were obtained from locations around the
cylinder diameter, what would you expect the circumferential distribution to look like?
Comment on the effects of using an average temperature versus a single point
temperature measurement on the calculations made above.

G. REFERENCES

[1] Holman, J. P. Experimental Methods for Engineers. 7th ed. Boston: McGraw-Hill, 2001.
Print.
[2] Incropera, F. P., and D. P. DeWitt. Fundamentals of Heat and Mass Transfer. 5th ed. New
York: Wiley, 2002. Print.
[3] Munson, B. R., D. F. Young, and T. H. Okiishi. Fundamentals of Fluid Mechanics. 4th ed.
New York: Wiley, 2002. Print.

MEM311: Thermal and Fluid Science Laboratory Experiment 4: Performance Analysis of a
Steam Turbine Power Plant

38 | P a g e

EXPERIMENT 4:
PERFORMANCE ANALYSIS OF A STEAM TURBINE POWER
PLANT
by Chris Dennison 2010, adapted from Bakhtier Farouk and David Staack

A. OBJECTIVE
The objective of this laboratory is to offer students hands-on experience with the operation
of a functional steam turbine power plant. A comparison of real world operating characteristics to
that of the ideal Rankine power cycle will be made.
The laboratory is conducted using a miniaturized steam turbine power plant. The apparatus
is scaled for educational use and utilizes components and systems similar to full-scale industrial
facilities. Students will be able to operate and analyze this system in detail, allowing them to
determine the efficiency of the facility and suggest possible modifications for further improvement.

Figure 10. Miniature steam turbine power plant
MEM311: Thermal & Fluid Science Laboratory Experiment 4: Performance Analysis of a
Steam Turbine Power Plant

39 | P a g e

B. THEORY
One of the most important ways to convert energy from fossil fuels, nuclear, and solar radiation is
through processes known as vapor power cycles. One example of the use of vapor power cycles is
electrical power plants. As engineers it is important to become familiar with these types of
systems. The first step in becoming familiar with these cycles is by studying the idealized cycles.
The ideal cycle for vapor power cycles can be modeled using the Rankine Cycle. This cycle is
composed of four components: a heater (boiler), a turbine, a condenser, and a pump. To complete
the system there must be some type of fluid flowing through the components, which is called the
working fluid. Most often the working fluid is water. As the working fluid passes through each of the
components it undergoes a process and ends up at a new state. Keeping in mind that the ideal
Rankine cycle is physically impossible, we define each process to involve no internal
irreversibilities. For the following it is necessary to number each of the states. State 1 is the state at
the boiler exit. State 2 is the turbine exit. State 3 is the condenser exit and state 4 is the pump exit.

Figure 11. Schematic of simple ideal Rankine cycle.



MEM311: Thermal and Fluid Science Laboratory Experiment 4: Performance Analysis of a
Steam Turbine Power Plant

40 | P a g e

Now the processes that the working fluid undergoes as it completes the cycle will be defined. First
is the heater, which in most cases is a boiler. As the fluid ends the cycle, at state 4, it is pumped into
the boiler. In the boiler, the working fluid is heated from sub-cooled liquid to saturated vapor. This
occurs at a constant pressure and is described in the following equation:

(1)
where

is the rate of heat addition relative to the mass flow rate of the working fluid passing
through the boiler. The value (

) is the difference in outlet and inlet enthalpies of the working

fluid.
Second is the turbine. Through the turbine the vapor leaving the boiler expands to the condenser
pressure. This is said to be isentropic expansion so that no heat transfer to the surroundings is
present. The equation that is used to describe this process is as follows:

(2)
where

is the rate of work being done relative to the mass flow rate through the turbine.
Again the difference in inlet and exit enthalpies of the working fluid is required.
Next the working fluid enters the condenser. At this stage heat is rejected from the vapor at a
constant pressure. Ideally, this continues until all of the vapor condenses to leave nothing but
saturated liquid. The equation for this is:

(3)
where

is the rate at which heat is transferred from the working fluid relative to the mass
flow rate. The value (

) is the difference between inlet and outlet enthalpies of the condenser.

Finally, the working fluid enters the pump. The fluid goes through an isentropic
compression process to reach the boiler pressure. The equation describing this is as follows:

(4)
where

is the rate of work being done relative to the mass flow rate through the pump.
Finally the difference in pump outlet enthalpy and inlet enthalpy is needed.
The efficiency of a given Rankine cycle may be computed as:

(5)

MEM311: Thermal & Fluid Science Laboratory Experiment 4: Performance Analysis of a
Steam Turbine Power Plant

41 | P a g e


Figure 3. P-v and T-s diagrams for the Rankine cycle.
In reality, no Rankine cycle is completely ideal. In particular, the compression of the working fluid
through the pump, and the expansion of the working fluid through the turbine are not actually
isentropic processes. Irreversibilities in these processes lead to increased power input to the pump,
and decreased power output from the turbine, both of which effectively lower the overall efficiency
of the system.
C. EXPERIMENTAL APPARATUS
The experimental hardware (RankineCycler) consists of multiple components that make up the
necessary components for electrical power generation (utilizing water as the working fluid).

Figure 4. The RankineCycler apparatus, including the data acquisition system.
MEM311: Thermal and Fluid Science Laboratory Experiment 4: Performance Analysis of a
Steam Turbine Power Plant

42 | P a g e

These components include:
BOILER
A stainless steel constructed, dual pass, flame-through tube type boiler, with super heat
dome, that includes front and rear doors. Both doors are insulated and open easily to reveal
the gas fired burner, flame tubes, hot surface igniter and general boiler construction. The
boiler walls are insulated to minimize heat loss. A side mounted sight glass indicates water
level.

Figure 5. Boiler with superheat dome
TURBINE
The axial flow steam turbine is mounted on a precision-machined stainless steel shaft,
which is supported by custom manufactured bronze bearings. Two oiler ports supply
lubrication to the bearings. The turbine includes a taper lock for precise mounting and is
driven by steam that is directed by an axial flow, bladed nozzle ring. The turbine output
shaft is coupled to an AC/DC generator.

Figure 6. Axial flow steam turbine wheel
MEM311: Thermal & Fluid Science Laboratory Experiment 4: Performance Analysis of a
Steam Turbine Power Plant

43 | P a g e

ELECTRIC GENERATOR
The electric generator, driven by the axial flow steam turbine, is of the brushless type. It is a
custom wound, 4-pole type and exhibits a safe/low voltage and amperage output. Both AC
and DC output poles are readily available for analysis (rpm output, waveform study,
relationship between amperage, voltage and power). A variable resistor load is operator
adjustable and allows for power output adjustments.
CONDENSER TOWER
The seamless, metal-spun condenser tower features 4 stainless steel baffles and facilitates
the collection of water vapor. The condensed steam (water) is collected in the bottom of the
tower and can be easily drained for measurement/flow rate calculations.
DATA ACQUISITION
The experimental apparatus is also equipped with an integral computer data acquisition
station, which utilizes National Instruments data acquisition software.
The fully integrated data acquisition system includes 9 sensors:
1. Boiler pressure
2. Boiler temperature
3. Turbine inlet pressure
4. Turbine inlet temperature
5. Turbine exit pressure
6. Turbine exit temperature
7. Fuel flow
8. Generator voltage output
9. Generator amperage output
The sensor outputs are conditioned and displayed in real time on screen. Data can be
stored and replayed. Run data can be copied off to a USB flash drive for individual student
analysis. Data can be viewed in Notepad, Excel and MSWord (all included).




MEM311: Thermal and Fluid Science Laboratory Experiment 4: Performance Analysis of a
Steam Turbine Power Plant

44 | P a g e

A schematic of the complete system is shown in Figure below.

Figure 7. RankineCycler schematic

*Note: When compared to figure 2 (simple ideal Rankine cycle), the steam ejected from the
RankineCycler turbine condenses into liquid via the condenser tower and then exits the
condenser into a collecting volume at the condensers base, rather than being pumped back
to the boiler. This represents the main difference between the simple Rankine cycle
described in figure 2, since the liquid exiting the condenser is dispensed, rather than
pumped back to the boiler.

General Safety
The RankineCycler operates at very high pressures and temperatures. It is essential for
the safety of everyone on the lab that certain safety precautions are adhered to at all times.
If these guidelines are not strictly followed, SERIOUS INJURY OR DEATH MAY RESULT.
-Do not touch any of the functional components during operation. These
components will be hot.
-Do not open the boiler during or immediately following operation. If the pressure
gauge indicates positive pressure, the boiler must remain closed.
-Do not exceed 120 psi boiler pressure.
Burner
Boiler
LP Natural
Gas Tank
Fuel flow
Boiler
pressure
Turbine Generator

Condensate
Collection Tank
Turbine exit temperature
and pressure
Variable
Load
Current and
voltage
Condenser
Turbine inlet temperature
and pressure
Boiler
temperature
MEM311: Thermal & Fluid Science Laboratory Experiment 4: Performance Analysis of a
Steam Turbine Power Plant

45 | P a g e

-Be very careful working near the condensing tower. The steam exiting the tower is
still extremely hot. Condensation drained from the bottom of the tower is also very
hot.
-If the scent of gas is detected at any time during operation, shut the equipment
down immediately.
-If any questions or concerns arise during equipment operation, notify the TA
immediately.

D. EXPERIMENTAL PROCEDURE
Do not begin operation without proper supervision. Prior to beginning any operation,
ensure that a trained lab technician, TA, or faculty member is present.
Prior to operation, familiarize yourself the following operator controls:
GAS VALVE
The gas valve is a simple two-position valve (On or Off). It is located on the far right
side of the slanted operator control panel. It will prevent gas flow to the burner
when in the off position- regardless of any other control positions/settings.

KEYED MASTER SWITCH
The systems electronic master switch is key operated and is located on the left side
of the operator control panel. This key switch supplies power to all electronic and
electrically operated components. A green indicator light, located directly above the
keyed master switch, will light when the master switch is selected to the on position
and power is available to the switch
BURNER SWITCH
The burner switch is labeled as such and is located next to the keyed master switch.
The burner switch powers the automatic gas valve and ignition controls. A green
indicator light, located directly above the burner switch, will light when the burner
switch is selected to the on position and power is available to the switch.
LOAD SWITCH
The load switch functions as a generator load disconnect switch.
LOAD RHEOSTAT CONTROL KNOB
The load rheostat control knob is connected in series with the load toggle switch
and generator DC output terminals. It provides a source of variable generator load.
AMP METER
The amp meter indicates generator load conditions.
MEM311: Thermal and Fluid Science Laboratory Experiment 4: Performance Analysis of a
Steam Turbine Power Plant

46 | P a g e

VOLTMETER
The voltmeter indicates the generator voltage output.
STEAM ADMISSION VALVE
The steam admission valve controls the steam flow rate to the steam turbine.

PRE-START
The TA will complete a pre-start procedure prior to the lab period. The pre-start
procedure includes safety and operability checks to ensure that the equipment is
functioning properly.

BOILER FILL
Fill the boiler according to the following procedure:
1. Verify that the boiler is empty
2. Fill the graduated cylinder with 6000mL of distilled water
3. Connect the fitting on the end of the plastic tube attached to the graduated cylinder to
the port on the lower, middle, back side of the boiler. The fitting should snap into place
4. Set the graduated cylinder on top of the condenser tower
5. Drain 500mL of distilled water into the boiler by operating the valve at the base of the
graduated cylinder
6. Record the water level indicated on the sight-glass attached to the boiler
7. Repeat steps 5 and 6 until all 6000mL of water have been drained into the boiler.
Record the total volume of water within the boiler, and the corresponding water level
each time.
The data recorded during this step will be used to develop a correlation relating boiler
water level to the remaining volume of water within the system. This correlation may be
found by entering the data into Excel, and obtain a curve fit of the data. The resulting
function can be used to calculate the total volume of water consumed during a steady-state
run.

If the duration of the run is known, the mass flow rate can be computed.

MEM311: Thermal & Fluid Science Laboratory Experiment 4: Performance Analysis of a
Steam Turbine Power Plant

47 | P a g e

where
[

]
duration of run ,-
START
1. Open LP bottle gas valve
2. Turn gas valve knob CCW to "on" position
3. Turn master switch on (observe green indicator light on)
4. Turn the load switch to the "on" position.
5. Set the load rheostat to ~1/2 maximum load.
6. Turn burner switch on (observe green indicator light on)
NOTE: Combustion blower starts automatically. Wait for 30 seconds. This will allow the lines
to purge. Then turn the burner switch to the "off" position and immediately back on (this step
can be eliminated from the start procedure if the system has previously been operated using
the currently attached LP source). This resets the starting cycle and assures that the lines are
purged. After approximately 20 seconds, the automatic gas valve will open and the burner
will light.
7. Boiler pressure indication should be observed within 3 minutes of ignition.
8. Allow the boiler pressure to increase to approximately 110 psi.
NOTE: SHUT OFF BURNER SWITCH IF THE BOILER PRESSURE EXCEEDES 120 PSIG.
9. Observe the voltmeter and gently open the steam admission valve. Regulate turbine
speed to indicate 7-10 volts. This will pre-heat the turbine components and the pipes.
Close valve after 30 seconds and wait for boiler pressure to return to 110 psi. Very
small leaks may be visible due to condensation and cold turbine bearing clearances.
This is normal and will stop after normal operating temperatures are attained.
10. Repeat step 9 two more times (three times total) to completely pre-heat the system.



MEM311: Thermal and Fluid Science Laboratory Experiment 4: Performance Analysis of a
Steam Turbine Power Plant

48 | P a g e

DATA COLLECTION
The T/A will create a folder for you on the data acquisition computer. The T/A will also be
responsible for designating file names for the data sets from each of the experimental runs.
1. Ensure that the load switch is on, and the load rheostat is set to ~ load (~12 oclock
position).
2. Allow the boiler pressure to return to 110 psi.
3. Gently open the steam admission valve, and make fine adjustments to achieve a steady-
state condition at ~90 psi. Each group must determine their own steady-state
tolerance. It is helpful to use the real-time boiler pressure display on the DAQ monitor.
4. Once a steady-state has been reached, simultaneously:
a. Record the starting water level.
b. Begin timing for ~3 minutes.
c. Begin data acquisition by clicking the play button in the DAQ interface.
During data acquisition, you may continue to make fine adjustments to the steam
admission valve to maintain your desired steady state.
5. After 3 minutes have passed, stop the data acquisition and record the final water level.
6. Open the Excel data file to ensure that data was recorded successfully.
7. Repeat steps 1 through 6 with the rheostat set to ~ load. Be sure to have the T/A
rename the data file prior to data acquisition. Otherwise, all of the data from the
previous run will be overwritten.
SHUT DOWN
1. Leave the steam admission valve open.
2. Move the burner switch to the "off" position.
3. Turn gas valve off.
4. Turn LP gas bottle valve off.
5. Slowly open the steam admission valve to release the remaining pressure. Do not allow
the generator voltage to exceed 12V.
6. Move the load rheostat to the no-load position
7. Move the load switch to the off position
8. Turn the master switch to the off position

MEM311: Thermal & Fluid Science Laboratory Experiment 4: Performance Analysis of a
Steam Turbine Power Plant

49 | P a g e

EMERGENCY SHUTDOWN
1. Unplug RankineCycler power cord
2. Move to a safe distance
3. If safety is not compromised: Turn burner switch off, turn load rheostat to
maximum, open steam admission valve to obtain maximum voltage.
E. PRE-LAB
Complete the following thermodynamic problem. This is intended to provide a
review of the Rankine cycle.
1. Consider a steam power plant that operates on a simple ideal Rankine cycle and has
a net power output of 30 MW. Steam enters the turbine at 7 MPa and 500C and is
cooled in the condenser at a pressure of 10 kPa by running cooling water from a
lake through the tubes of the condenser at a rate of 2000 kg/s. Show the cycle on a
p-v, and a T-s diagram with respect to saturation lines, and the determine (a) the
thermal efficiency of the cycle, (b) the mass flow rate of the system and (c) the
change in temperature of the cooling water.

2. Compute the Carnot efficiency of the cycle described in part 1. Discuss the reasons
for any difference between the Rankine and Carnot efficiencies. How would you
increase the efficiency of the Rankine cycle?

3. Read the entire lab procedure. Create a spreadsheet to record the data obtained in
the Boiler Fill steps. Print this spreadsheet and bring it with you to the lab (or
bring your laptop with the spreadsheet to the lab).
F. DATA ANALYSIS AND REPORTING REQUIREMENTS
1. From the time-averaged data collected by the computer, plot the state points 4, 1 and 2
(as shown in Figure 3). Why are we not asking you to plot state point 3?
2. Calculate the values of enthalpy and entropy for the state points 4, 1 and 2.
3. Compute the turbine work (Watts). The water level/volume correlation, obtained during
the Boiler Fill step, will yield the mass flow rate.
4. Compute the generator work from the measured voltage and current measurements.
MEM311: Thermal and Fluid Science Laboratory Experiment 4: Performance Analysis of a
Steam Turbine Power Plant

50 | P a g e

5. Compute the (rate of heat addition) to the boiler from the fuel flow rate and the
heating value of the fuel (propane).
6. Compute the cycle efficiency. Identify and discuss sources of inefficiency in the system.
Suggest modifications which would improve cycle efficiency.
Complete these steps at both operating conditions ( and load).

G. REFERENCES
1. RankineCycler Operations Manual, Turbine Technologies, Ltd., Chetek, WI.
2. Cengel, Y. A. and Boles , M. A., Thermodynamics, An Engineering Approach (4th
Edition), McGraw-Hill, 2002
3. "Rankine Cycle." Wikipedia, the Free Encyclopedia. Web.
<http://en.wikipedia.org/wiki/Rankine_cycle>.

MEM311: Thermal and Fluid Science Laboratory Experiment 5: Lift Characteristics of an Airfoil
Section

51 | P a g e

EXPERIMENT 5:
LIFT CHARACTERISTICS OF AN AIRFOIL SECTION
by Brandon Terranova 2010, adapted from Bakhtier Farouk and David Staack

A. OBJECTIVES

Determine lift force and coefficient for an airfoil section at different angles of attack for
several tunnel velocities.

B. THEORY
Airfoil geometry
The wing of an airplane is its primary lifting device. Lift is the force that counteracts the
aircraft weight and causes flight. The simplified cross-section of an infinite wing, an airfoil,
is often tested in wind tunnels to accurately and optimally design the wing of an aircraft.
The study of actual wing geometries (finite wings) is built on the understanding of the
idealized airfoil aerodynamics. A schematic of the airfoil geometry with associated
terminology is shown in Figure 12.



Figure 12: Airfoil geometry


This experiment uses the NACA-2415 airfoil, which is one of the NACA 4-digit series
standard airfoils. The National Advisory Committee on Aeronautics (NACA) studied the
characteristics of airfoils to develop a database for aeronautic engineering design.

Notice that this particular airfoil is not symmetric about the chord line. This type of airfoil
is generally referred to as a cambered airfoil. The 4-digit series shape is mathematically
identical to symmetrical airfoils with the exception that the mean camber line is bent. The
first digit indicates the maximum camber height (distance from the chord to the maximum
height of the mean camber line) m in hundredths of the chord (for NACA-2415 m = 0.02
or 2%). The second digit indicates the location of the maximum camber (as measured from
the leading edge) p in tenths of the chord (in this case p = 0.4 or 40%). The third and
Leading edge
Trailing edge
Mean camber line
Upper surface
Lower surface
Chord line
MEM311: Thermal and Fluid Science Laboratory Experiment 5: Lift Characteristics of an Airfoil
Section


52 | P a g e

fourth digits signify the maximum thickness t in percentage of the chord (in this case t =
15%). For a 4-digit cambered NACA airfoil, the following formulas are used to calculate the
mean camber line:

/

(1)

( )
( )

/

(2)
Where c is the chord length and x is the position along the chord from 0 to c. The
expression for the entire camber line would then be a piecewise function. The upper and
lower airfoil surfaces is given by

(3)

(4)
where

)
(5)
and

/ .

+
(6)
which is the equation corresponding to the shape of a symmetrical 4-digit NACA airfoil.
1

In addition to the standardization of certain airfoils, we are able to describe the arbitrary
shape of any cambered airfoil mathematically. Since the shape of the camber is an arbitrary
curve, we must use Fourier analysis to accurately define a function which reproduces the
curve. The idea behind Fourier analysis is that any function can be created as a sum of sine
and cosine curves.

From airfoil theory, we note that the camber line, regardless of shape, will be a streamline
of the flow around the corresponding airfoil. From this concept, an equation can be derived
that is a general form for any camber line:
MEM311: Thermal and Fluid Science Laboratory Experiment 5: Lift Characteristics of an Airfoil
Section

53 | P a g e


() (

)
(7)
Where is the wind velocity at infinity (refers to the unperturbed conditions very far
ahead of an aerodynamic body), and u is a coordinate transformed from
with representing the angle of attack, which is the angle formed between the incoming
wind and the chord line of the airfoil.

Sources of Lift, Drag and Moments
No matter how complex the airfoil geometry, or any shape for that matter, the aerodynamic
forces and moments on the body are due entirely to the pressure and shear stress
distributions over the body surface. As you know from fluid mechanics, the pressure
difference between the upper and lower wing surfaces arises from the velocity difference
between the flows. This can be visualized by noticing the spacing between the streamlines
in Figure 13.

Figure 13: Streamlines over an airfoil.
The more condensed streamlines are an indication of the greater fluid velocity above the
wing, and this corresponds to a decrease in pressure. The equation for dynamic pressure
illuminates this, showing how pressure can change due to the velocity of the fluid alone.

( )
The Fourier coefficients are found using
(8)

(9)

(10)
MEM311: Thermal and Fluid Science Laboratory Experiment 5: Lift Characteristics of an Airfoil
Section


54 | P a g e

(11)
But wait! This equation says that pressure should INCREASE when velocity is increased.
Whats going on here? Since Bernoullis equation involves more than just the dynamic
pressure, we need some more information to understand how lift occurs. Where is the
density of air, and the dynamic pressure is denoted with a q to distinguish it from the other
terms in Bernoullis equation for incompressible flow:

(12)


Where we use to represent the total pressure, which must be constant between one
region of a streamline of the fluid and another. This is the essence of Bernoullis equation:
the fact that for a steady flow, the total pressure must be the same between one region of a
fluid and another. Bernoullis equation as shown applies to the flow along a streamline
when 1) the fluid has constant density, 2) the flow is steady, and 3) there is no friction.
Figure 14a shows the pressure distribution with arises from zero angle of attack ( = 0) on a
cambered airfoil.

So back to the velocity vs. pressure issue. Since increasing the velocity obviously increases
the dynamic pressure, where does the drop in pressure come from? Since the pressure
must be constant between one region of the streamline and another, this increase in
velocity must be accompanied by a decrease in pressure somewhere! The decrease
happens in the static pressure. When the velocity of a fluid increases, the thermodynamic
properties of a fluid changes (density changes), thus the decrease in the static pressure is
what balances the equation. But you still might be wondering: Ok, so the pressures
balance outhow do we end up with a lower pressure on the wing if they are balanced??
The answer is that the static pressure is what is felt by the skin of the wing, thus this
decrease of static pressure is what is primarily responsible for lift!


Static pressure (due to the
thermodynamic
properties of the fluid)
(A) (B)
Gravity pressure (due to the weight of
the water on itselfneeded if at reference
height, h0)
MEM311: Thermal and Fluid Science Laboratory Experiment 5: Lift Characteristics of an Airfoil
Section

55 | P a g e

Figure 14: (a) Pressure distribution. (b) Shear stress distribution.

In addition to the pressure acting on the airfoil, friction occurs as the fluid moves over the
airfoil, resulting in a pulling on the airfoil which gives rise to the shear stress distribution
as seen in Figure 14b. This shear stress distribution generates a moment () around the
body. At the same time, the pressure distribution which is induced by the difference in
dynamic pressure, couples with the moment created by the shear stress, and the
cumulative effect of these forces is described by a resultant vector located at a point known
as the center of pressure as shown in Figure 15. This resultant vector can be broken into
its component vectors, which we refer to as drag () and lift ().

Figure 15: Lift and drag as resulting from pressure and shear stress.

Pressure and Lift Coefficients
The pressure coefficient is a very useful parameter for characterizing the flow of
incompressible fluids, while the lift coefficient is directly related to wing aerodynamics
(notice the inclusion of the wing area term in equation 14). The lift coefficient is related to
the lift force, which as you know arises from complicated physical interactions between the
moving fluid and the wing surface. In general, it is very difficult to analytically determine
the value of this coefficient (as well as coefficients for drag, axial, normal force and moment
coefficients) for arbitrary wing geometries, therefore it is usually determined empirically.
There exist other coefficients that serve other purposes, but for this lab we can confine our
interests to these four. The lift, drag and moment coefficients for an aerodynamic body can
be obtained by integrated the pressure and skin friction (related to the shear stress)
coefficients over the body surface from the leading to the trailing edge.
2
Notice that again,
all parameters are derived from only considering the pressure and shear stresses on the
wing.

(13) (14)
cp = pressure coefficient
cl = lift coefficient
p

= pressure at infinity
A = area of the airfoil, A = cS
MEM311: Thermal and Fluid Science Laboratory Experiment 5: Lift Characteristics of an Airfoil
Section


56 | P a g e

c = chord length of the airfoil S = span of the airfoil
For wind tunnel measurements, the conditions at infinity correspond to the unperturbed conditions
at the beginning of the test section.

Lift Calculation
As mentioned earlier, to calculate lift, we are primarily interested in the static pressure on
the various sections of the wing. In our laboratory experiment, the static pressure ps will be
measured at several locations along the lower and upper surfaces of the airfoil section.

The lift force on the airfoil can be approximately calculated from the measured pressure
distribution along the airfoil (lower and upper) surfaces. The pressure along the upper and
lower surfaces of the airfoil is measured at several tap locations as shown in table 1. Each
pressure tap is assumed to act over a small sectional area as shown in figure 5. The force
on each sectional area is given by the product of the static pressure on that section and the
sectional area. The force on the upper and lower surfaces can be obtained by summing the
components of the individual forces on each section. The net lift is obtained by adding the
net forces on the upper and lower surfaces. When summing the forces it should be noted
that the pressure acting on the upper surface pushes downward, and the pressure acting on
the lower surface pushes upward.

Tap ID X-location (in)
Tap location (X/c)
c = 6
Upper Surface
1 0.18 0.029
2 0.5 0.08
3 1.07 0.172
4 1.7 0.273
5 2.55 0.409
6 3.45 0.554
7 4.3 0.69
8 5.25 0.843
Lower Surface
9 (under 1) 0.16 0.026
10 (under 2) 0.5 0.08
11 (under 3) 0.95 0.152
12 (under 4) 1.4 0.225
13 (under 5) 2.2 0.353
14 (under 6) 3.1 0.498
15 (under 7) 3.94 0.632
16 (under 8) 4.96 0.796
Table 6: Pressure tap locations along the NACA 2415 airfoil section.
MEM311: Thermal and Fluid Science Laboratory Experiment 5: Lift Characteristics of an Airfoil
Section

57 | P a g e





Figure 16: Schematic for determining sectional areas associated


The area associated with each pressure tap will be taken as the average distance between
taps multiplied by the span of the airfoil as shown in figure 5. Starting at the leading edge
and the upper surface, pressure tap #1 will be assumed to act over the sectional length

(15)
Thus the area associated with tap #1 is

)
(16)
The sectional lift force associated with section #1 would then be

)
(17)
For tap #2, the sectional force would be

)
(18)
and similarly for the rest of the pressure taps. Be careful when computing the lift associated
with taps close to the leading and trailing edges, as they contain terms dissimilar from the
rest of the equations.
X1
X2
X3
MEM311: Thermal and Fluid Science Laboratory Experiment 5: Lift Characteristics of an Airfoil
Section


58 | P a g e

Wind Tunnel Design Theory
A wind tunnel is just a large Venturi tube. Wind tunnels are used in aerodynamic research to study
the effects of air moving past aerodynamic bodies. The objective of the design of a wind tunnel is to
create laminar (turbulence-free) air flow in the test section. The typical open-circuit wind tunnel
consists of three main parts which are labeled in figure 6. Air is drawn through the device by a fan
attached to the diffuser section. The compressor section is typically open to the atmosphere with the
exception of closely-spaced vertical and horizontal air vanes used to smooth out the turbulent
airflow before reaching the testing subject. The test section is where the laminar flow is
concentrated, and subsequent measurements are taken on the test subject.
The main operating principle of the wind tunnel geometry is that we can adjust the areas of the
compressor and diffuser sections to control the velocity in the test section. Flow rate can be
measured using the pressure difference between sections, as is done in a conventional venture tube
using Bernoullis equation and the continuity equation.


Figure 17: Open-circuit wind tunnel design

C. EQUIPMENT
Wind Tunnel
The model 1440 Flotek wind tunnel, as shown in figure 7, provides a full 12 x 12 (305
mm square) cross-section in the working area over 36 (914 mm) in length. Air is drawn
through the compressor intake cone into the test section by a variable speed fan. A plastic
honeycomb flow straightener attached to the mouth of the compressor assures a laminar air
flow through the test section. The entrance cone has a contraction ratio of 12:1 down to the
test section. The air velocity through the test section is variable up to 90 mph.

MEM311: Thermal and Fluid Science Laboratory Experiment 5: Lift Characteristics of an Airfoil
Section

59 | P a g e


Figure 18: The Flotek 1440 wind tunnel with detailed view of the NACA-2415 airfoil.
Airfoil
This lab uses a NACA-2415 airfoil. This airfoil has a chord length c = 6 in and a span of 11 11/16 in.
The airfoil has 16 pressure tappings, 8 on the upper surface, and 8 below.
Instrumentation
The system is connected to a computer, and includes an interface card and Labview
software. Data acquisition and computer control of the wind tunnel is accomplished by
using a 16 channel analog-to digital (A/D) and 2 channel digital-to-analog (D/A) converter
board. The pressure taps are connected a water-filled manometer array as well as to a set of
16 pressure transducers which converts the pressure into an analog voltage. The data
acquisition system works by converting this analog voltage signal to a digital signal which is
read by the computer. Similarly, converting a digital signal from the computer to an analog
voltage provides control for the motor which sets the angle of attack. The converter board is
located in a PCI slot inside the control computer, and a cable connects it to the wind tunnels
transducer box located directly behind the test section, underneath the exhaust section.

Software based on LabVIEWs visual programming language is used for system control and
data acquisition. A screenshot of the softwares instrument panel is shown in figure 8. Slider
bars (circled in red) are used to adjust the motor speed and angle of attack. The software
also has the capability of setting a particular speed in the test section by toggling the
velocity control (circled in blue) switch, entering the desired velocity into the setpoint
field, and adjusting the gain (this dictates how quickly the system arrives to the set velocity).

The airfoil pressure measurements are taken from the 8 pressure taps along the upper
surface of the airfoil and 7 along the lower surface. The remaining pressure tap (which only
has a manometer display) measures the static pressure in the wind tunnel test section.

MEM311: Thermal and Fluid Science Laboratory Experiment 5: Lift Characteristics of an Airfoil
Section


60 | P a g e

The static pressure in the wind tunnel test section is used to calculate the tunnel velocity
using Bernoullis equation. This calculation is hard-coded into the LabVIEW algorithm. The
total pressure which is required for that measurement is measured when the LabVIEW
program is first turned on (thus to operate properly the wind tunnel fan motor should be off
when starting the program). The relative pressures measured along the airfoil are shown in
two plots, one for the upper and one for the lower surface. Similar to how the static pressure
is calculated in the test section, the flow velocities are calculated from the pressure
information at each tap on the airfoil as well. The transducer offset values which in effect
determine the total pressures are measured when the program is started and are shown in
the diagnostics tab (circled in green). When the record button (circled in black) is pressed,
the pressure data, velocity data, motor rpm, and angle of attack are stored in a data file. This
file can be opened with Excel.

Subsequent use of the record button appends a line of
new data to the file. Test this feature until you are
comfortable taking data.


Figure 19: Wind tunnel instrumentation panel
MEM311: Thermal and Fluid Science Laboratory Experiment 5: Lift Characteristics of an Airfoil
Section

61 | P a g e

D. PROCEDURE
1. Begin by starting the software located on the desktop. The fan motor should be off
when the Labview program is started.

2. Once the program is running set the angle of attack to 0, turn the key and press the
green button on the Flotek fan power control (shown in Figure 18).

3. Once you are familiar with the system operation, press the record data button (circled
in black in Figure 19) and record data for a few moments - you should check the data
output with some test velocities.

4. For steady state tunnel velocities of 60 ft/s and 75 ft/s and 90 ft/s measure the
pressure distributions at -4, 0, 4, 8, 12, and 16 so that you make 18 measurements
in total.

Note: Since only 15 of the 16 pressure taps on the airfoil are digitally acquired, for each
condition you will need to manually write down the pressure on the 8
th
tap of the lower
surface from the manometer board. In order to correlate the manual data with the
computer acquired data you will also need to note the test section pressure
(manometer tube on far left) and atmospheric pressure from the manometer board
(manometer tube which is open to the environment). Be sure that your calculations
match reasonably well to the transducer measurements before you are confident of the
8
th
pressure tap reading.

E. PRELAB
1. Consider an airfoil section with only 3 pressure taps on the upper and lower surfaces.
The pressure readings (gauge) for the taps are given below:

Upper surface Lower surface
Pressure at tap #1 -17.5 lbf/ft
2
Pressure at tap #1 8.5 lbf/ft
2

Pressure at tap #2 -12.6 lbf/ft
2
Pressure at tap #2 0.6 lbf/ft
2

Pressure at tap #3 -7.1 lbf/ft
2
Pressure at tap #3 0.2 lbf/ft
2

Assume the airfoil chord is 3 inches and the span is 6 inches. The sectional length
assigned to each tap is 1 inch. Calculate the total lift on the airfoil where

(hint: when working in U.S. customary units remember the conversion 1 lbf =
32.2 lbm ft/s
2
)
2. Using equations 1-6, find an expression for the camber line, upper surface and lower
surface of the NACA-2415 airfoil. Plot all three equations on the same graph using some
computer algebra program. (Note that the mean camber line equation is a piecewise
MEM311: Thermal and Fluid Science Laboratory Experiment 5: Lift Characteristics of an Airfoil
Section


62 | P a g e

function. Also, if you get a couple discontinuities in the plots for the upper and lower
surface, dont worry about it. Your plot should look like the airfoil.)

3. a) Use the equation below to determine the center of pressure for the NACA-2415 airfoil
encountering a wind speed of 90 mph (possible in our wind tunnel!). b) Explain the
difference between center of pressure, center of gravity and aerodynamic center.

)]
Where again

is the coefficient of lift, and A1 and A2 are the Fourier coefficients. Note:
Use the

calculated in #1 for the lift in

.

F. DATA ANALYSIS AND REPORTING REQUIREMENTS

1. Calculate and plot on the same graph the pressure profile (measured static pressures) on
the upper and lower surfaces as a function of the chord position for the -4 and the 8
angle of attack for each of the three tunnel velocities (6 plots total). Discuss how the
pressure profiles change with angle of attack and velocity.

2. Calculate and plot on the same graph the pressure coefficient on the upper and lower
surfaces as a function of the chord position for -4, 0 and 8 angle of attack for 75 ft/s
velocity (3 plots total).

3. Plot the lift coefficient as a function of angle of attack for the three tunnel velocities on the
same graph (1 plot total).

G. REFERENCES

1. Moran, J. (2003). An Introduction to Theoretical and Computational Aerodynamics. Dover.

2. Anderson, John D. (2007). Fundamentals of Aerodynamics, Fourth Edition. McGraw-Hill.

3. Munson et. al (2009). Fundamentals of Fluid Mechanics, Sixth Edition. Wiley.

MEM311: Thermal and Fluid Science Laboratory Appendix A

63 | P a g e

APPENDIX A
MANOMETER PREPARATION AND OPERATION
by Brandon Terranova 2010, adapted from Bakhtier Farouk and David Staack
The fluid manometer provides a relatively simple method to measure fluid pressures under steady-state
conditions. A sketch of a simple U-tube manometer is shown in Fig. A.1. In this configuration, one side
of the manometer is open to the atmosphere and the other is open to the pressure to be measured.
The theory behind manometer operation (Ref. A.1, Section 2. and Ref. A.2, Section 6-3) shows that the
difference between the pressures on each side of the manometer are given by the hydrostatic equation,
i.e.,

) (A.1)

where

is the density of the fluid in the manometer and

is the density of the fluid transporting the

unknown pressure. Typically, the density of the fluid in the manometer is much larger than the density
of the fluid transmitting the pressure so Eq. (A.1) is often written as

(A.2)


Figure A.20
Note: The distance Ah is measured parallel to the gravitational force and that the differential pressure is
measured at the location of the manometer. If the location of the pressure source is at a different
elevation than the manometer, there could be appreciable error in the pressure determination
depending on the density of the transmitting fluid.

MEM311: Thermal and Fluid Science Laboratory Appendix A

64 | P a g e

The manometer can also be used to measure a pressure difference between two locations such as on
opposite sides of an orifice flow meter. In this configuration, shown in Fig. A. 1b, the manometer is
placed in a closed loop containing both the manometer fluid and the transmitting fluid. The pressure
difference between the two locations is calculated using Eq. (A.1).

Figure A.21

(A.3)

The experiments in this lab make use of manometers to measure pressure differences. A slight
difference between those used in the lab and shown in figure A.1 is that multiple manometer tubes
share the same reservoir. With reservoir manometers the cross sectional area of the reservoir is
significantly greater than that of the manometer tube. Thus a decrease in pressure causes an increase in
the height of the fluid in the tube but a negligible change in the height of the fluid in the reservoir. A
schematic of a reservoir type manometer Be sure to familiarize yourself with the configuration before
attempting to interpret your measurements.

Water manometers are simple to use once they have been properly prepared. Often, if the
apparatus has been standing for some time, air will become trapped in the pipes and in the water
columns of the manometer tubes. These air pockets will affect the pressure measurement because air
ha a different density than water. The manometer should be ready for operation when you report to
the lab. Improper use of the manometer during the experiment may cause air to become trapped in the
lines. Also, there are some experiments which, by nature, require moving the rubber tubes leading from
the manometer to various pressure taps in the system. This may result in air bubbles forming in the
manometer tubes. If this condition occurs during your experiment, you will have to remove the air
bubbles before beginning the experiment. This procedure varies depending on the exact configuration
but generally involves putting a pressure difference across the two ends of the nanometer. This will
induce a flow through the tube thereby forcing the bubble out the tube.

MEM311: Thermal and Fluid Science Laboratory Appendix A

65 | P a g e


Note: Extreme care must be exercised when using mercury manometers. Do not try to remove air from
a mercury manometer using the same procedures as for a water manometer. It is a much less frequent
problem because of the larger density of mercury. Because of the toxicity of mercury, spills can be
extremely hazardous. Also, never ignore or attempt to hide a mercury spill! Inform someone who can
properly clean up the spill and dispose of the mercury.

References

A.1. Munson, B.R., Young, D.F., and Okiishi, T.H., Fundamentals of Fluid Mechanics, John Wiley and
Sons, New York, 1990.
A.2. Holman, J.P., Experimental Methods for Engineers (7
th
edition), McGraw-Hill, New York, 2001.


MEM311: Thermal and Fluid Science Laboratory Appendix B

66 | P a g e

APPENDIX B
FLOW MEASURING USING A ROTAMETER
by Brandon Terranova 2010, adapted from Bakhtier Farouk and David Staack
Many devices and methods exist to measure the flow rate of a fluid. Several of the more popular of
these are described in Chapter 7 of Ref. B.1. An orifice flow meter and a venturi flow meter are
examples of obstruction-type flow meters because their operating principle is based on measuring
the response of the fluid to an obstruction in the flow. A rotameter is slightly different in that the
fluid flow rate is determined by measuring the response of an obstruction to the fluid motion. Flow
enters from the bottom of a rotameter, shown in Fig. B.1, and flows through a tapered vertical tube.
This flow causes a bob in the tube to move upward due to the fact that as flow increases, the area
around the float must also increase in accordance with the basic equation for volumetric flow rate
(Eq. B.1). The bob will rise to a position in the tube where the drag forces acting on the bob are just
balanced by the weight and buoyancy forces. The position of the bob is then taken as an indication
of the flow rate of the fluid. The drag force acting on the bob will be a function of not only the flow
rate of the fluid but how the tube is tapered and the shape of the bob. Therefore, each rotameter
must be calibrated to correctly account for these geometric effects.
(B.1)

The equations detailing the operating principle of a rotameter are presented in Ref. B.1,
Section 7-6.
*
(

(B.2)

where y = vertical displacement of the bob,

= density of the fluid,

= density of the bob, and C is

the calibration constant for the rotameter. While you may find different types of rotameters, most
are calibrated to read from the top of the bob. Since rotameters are typically used to measure the
flow rates of gases,

is much larger than

equation B.2 can be reduced to

(B.3)

where

= volumetric flow rate (m

3
/s). Equation (B.2) shows that we must know the density of a
fluid to determine either the volumetric or mass flow rate. The calibration constant, C, is a function
not only of the geometry of the tube and bob but also of the density of the fluid used during the
calibration. Luckily, we do not need separate rotameters calibrated for every fluid density that we
may encounter during our experiment. Instead, we can correct the mass flow rate given by Eq.
(B.3) for the density of our fluid if we know the conditions at which the rotameter was calibrated
and our test conditions. Rotameters and most gas flow devices are calibrated at standard
atmospheric condition, i.e., calibrated using air at 70F and 1 atm. Calibration tables or
MEM311: Thermal and Fluid Science Laboratory Appendix B

67 | P a g e


Figure B.22

graphs give

as a function of y where the subscript s indicates at standard temperature and

pressure. Therefore, the calibration constant quoted by the manufacturer can be interpreted as

(B.4)

Using the calibration constant given by the manufacturer, Eq. (B.3) becomes

(B.5)

Comparing equations B.3 and B.5:

(B.6)

For a given value of y, we can use equation B.6 to find the experimental calibration constant once a
flow measurement has been made.

References

B.1. Holman, J.P., Experimental Methods for Engineers (7
th
edition), McGraw-Hill, New York,
2001.

MEM311: Thermal and Fluid Science Laboratory Appendix C

68 | P a g e

APPENDIX C
ANALYSIS OF BIAS ERRORS AND EXPERIMENTAL UNCERTAINTY
by Brandon Terranova 2010, adapted from Bakhtier Farouk and David Staack
A. Terminology
As in all experimental investigations, any measured data are subject to experimental errors that cannot
be eliminated completely. These errors arise because of limitations in the accuracy of one or all of the
instruments used to make the measurements or to the variations between people using the instrument.
These errors must be quantified to provide a measure of the reliability or uncertainty associated with
the data.
Various types of errors can contribute to the uncertainty in fundamental measurements. These include
accidental errors (outright mistakes), bias errors (also referred to as fixed or systematic errors), and
random errors. Accidental errors cause repeated readings to differ without apparent reason and may be
attributed to instrument friction, time lag, and personal errors. These must be identified and corrected
in any experiment. If this type of error is not removed entirely, it must at least be made to effect the
experiment the same way every time. Then it is known as a bias error. Bias errors result when data are
shifted from actual values by some characteristic of the substance being measured or the measurement
technique. Bias errors account for reductions in the accuracy of the experimental data. Random errors
may be caused by random human error, random environmental changes, or any fluctuation of the
property being measured. The magnitude of this type of error is a measure of the precision of a
measurement. The physical difference between precision and accuracy as applied to experimental
measurements is illustrated below.
Because these errors are random, they are better expressed as an uncertainty and require that
probability concepts be applied to quantify experimental errors. A concise way to describe data that
consists of both bias errors and experimental uncertainty is to specify the mean value and an
uncertainty interval, i.e.
u m (xx percent uncertainty) (C.1)

Figure 23C: Precision and accuracy in measurements
MEM311: Thermal and Fluid Science Laboratory Appendix C

69 | P a g e

where m is the mean value, u is the uncertainty interval, and xx is the percent confidence we have in the
uncertainty level. The meaning of these terms is more transparent if Eq. (C.1) is expressed in words as
There is an xx percent probability that the true value of a property lies within u of the mean value,
m. The factor xx is often called the confidence level or interval.

As an example, consider measuring the length of a desk using a yardstick. If the measurement was made
several times, each measurement will probably be slightly different. If every student in the class made
the same measurement, these measurements would also vary. Its not that any single measurement was
wrong or right. The variations occur because the measurement technique and procedure (using a
yardstick) has inherent errors that limit the accuracy of the measurement. How can you express this
uncertainty about the true length of the table? The set of all the length measurements can be
represented as (x
1
, x
2
, x
3
, x
4
, , x
n
) where n is the number of measurements and is assumed, for now, to
be large (greater than ~30). The following statistical quantities can then be defined calculated using this
set.
Mean:

(C.2)

Deviation:

(C.3)

Average
Deviation:

(C.4)

Variance:

(C.5)

Standard
Deviation:

(C.6)
If we only have a limited set of data (less than 20 or so samples), we really dont have sufficient data to
accurately estimate the standard deviation using Eq. (C.6). The true value of the standard deviation can
then be estimated using the following equation

(C.7)
MEM311: Thermal and Fluid Science Laboratory Appendix C

70 | P a g e

Note that the factor (n-1) is used in equation C.7 instead of n in equation C.6.

We could also plot the number of times a specific value of x was obtained for the length of the table
from the entire set of data. As n goes to infinity (an infinite number of samples), a continuous curve
would be obtained. Assuming that the errors associated with these measurements are equally likely to
be positive and negative, the curve would form a normal or Gaussian distribution about the mean
value, . This distribution is shown below.
This figure shows that the standard deviation, o, is a measure of the width of the distribution.

In the above discussion, the arithmetic mean value has been assumed to be our best estimate of the
true value of a set of experimental measurements. The Gaussian distribution allowed us to estimate
how the data are distributed around the mean value. We still have a very important question to answer,
i.e., how well does this arithmetic mean approximate the true value, which is unknown? To obtain an
answer to this question experimentally, it would be necessary to repeat the entire set of measurements
and calculated a new mean value. This value would undoubtedly differ from the previous value because
of the same variations that produced the differences within any one data set. We could continue this
procedure to obtain a large number of mean values, estimate the standard deviation of the mean values
an, finally, the uncertainty in our estimate of the mean value. In other words, the mean value of all the
mean values of a large number of data sets is presumably the true value. Fortunately, this problem can
be treated with a statistical analysis that allows us to approximate the standard deviation of the mean
value using the standard deviation of a single set of data. This analysis will not be repeated here but
results in the following equation for the standard deviation of the mean,
m
o ,

(C.8)
Figure C24: Normal distribution about the mean value
MEM311: Thermal and Fluid Science Laboratory Appendix C

71 | P a g e


where o is the standard deviation given by equation C.6 or C.7 and n is the number of samples in the
data set. We could then say with 95 percent confidence, for example, that the true value of x was
within
m
o 2 of our calculated mean value. An example of the application of equation C.8 to a set of
data is given in Ref. C.1, Section 3-11. If n is very small (less than 15-20) the approximation for
m
o

given
by equation 8 becomes worse, and other methods must be used to estimate the standard deviation of
the mean value. This method, known as the t distribution test, will not be discussed here but can be
found practically any book on statistics (Ref. C.2, Section 7.4 and 9.3, for example).

B. Propagation of Errors in Experiments
The analysis of uncertainties and confidence intervals discussed above were all concerned with knowing
the true value of a single measured parameter such as the fluid height is a manometer, temperature,
mass flow rate, etc. However, in all our experiments, these parameters are used to calculate additional
quantities. Bias errors and uncertainties in the individual measurements propagate through the
equation, resulting in an overall uncertainty in the calculated result. This uncertainty can be estimated if
the uncertainties of all the individual parameters are known. For instance, what is the error in
where A and B are two measured quantities with errors AA and AB respectively?
A first thought might be that the error in Z would be just the sum of the errors in A and B. However, this
assumes that when combined, the errors in A and B have the same sign and maximum magnitude; that
is, they always combine in the worst possible way. This could only happen if the errors in the two
variables were perfectly correlated, (i.e. if the two variables were not really independent.)
If the variables are independent (they usually are), then sometimes the error in one variable will happen
to cancel out some of the error in the other and so, on the average, the error in Z will be less than the
sum of the errors in its parts. A reasonable way to try to take this into account is to treat the
perturbations in Z produced by perturbations in its parts as if they were "perpendicular" and added
according to the Pythagorean theorem,
()

()

(C.9)
This idea can be used to derive a general rule. Suppose there are two measurements, A and B, and the
final result is ( ) for some function f. If A is perturbed by AA then Z will be perturbed by
(

) (C.10)
Similarly, the perturbation in Z due to a perturbation in B is
(

) (C.11)
Combining these by the Pythagorean theorem yields
MEM311: Thermal and Fluid Science Laboratory Appendix C

72 | P a g e

((

) )

((

) )

(C.12)
We can then write down a general definition for the uncertainty in a variable Z by considering Z to be a
linear function of n independent variables, so that (

). If each x
i
has corresponding
uncertainties

, the relation for the uncertainty in Z is given as

(C.13)
When equation C.13 is applied, the confidence level for U
Z
is the same as for the estimates of

. The
value of Z is then reported as

(xx percent confidence level) or equivalently

(xx percent confidence level). An example of the application of equation C.13 is given below. Several
additional examples can be found in Ref. C.1, section 3-4.

Example
Consider a pressure measurement made using an open-end U-tube water manometer. The pressure is
then given by

(C.14)
where h is the height of the fluid column, p
o
is the atmospheric pressure measured with a barometer,
and is the density of the fluid. We find that the atmospheric pressure, p
o
= 29.6 inches Hg and h = 24
inches H
2
O. During our measurements, we determine that the scale on the barometer can be read to
0.1 inch Hg while that on the manometer can be read to 0.25 in H
2
O. The density of water and the
gravitational acceleration can be taken as constant, = 62.4 lbm/ft
3
and g = 32.2 ft/s
2
, respectively. As
always, g
c
= 32.2 lbm-ft/lbf-s
2
. The problem is to determine the pressure, p (in psia), and the uncertainty
in the measurement of p caused by the uncertainties in p
o
and h.

Solution
First, substituting the measured values of p
o
and h into equation C.14 and converting to psia, we find
that p = 15.41 psia. Since there was some uncertainty in reading the scale of both the barometer and
the manometer, the atmospheric pressure and the fluid height are properly expressed as p
o
= 29.6 0.1
inches Hg and h = 24 0.25 inches H
2
O. For this problem, equation C.13 can be expanded as follows:
MEM311: Thermal and Fluid Science Laboratory Appendix C

73 | P a g e

((

((

(C.15)

The derivatives in equation C.15 must now be evaluated using equation C.14.

Substituting these derivatives into equation C.15, we obtain

(C.16)
Substituting the known values into equation C.16, being aware that units of psia are desired and must
be consistent in the equation, we get the calculated pressure as p = 15.41 0.05 psia or p = 15.41 0.32
%.

References
C.1. Holman, J.P. (2001), Experimental Methods for Engineers, 7
th
edition, McGraw-Hill, New York.
C.2. Mendenhall, W., Reinmuth, J.E. (1989), and Beaver, R., Statistics for Management and
Economics, 6
th
edition, PWS-Kent Publishing Company, Boston.
C.3. Taylor, John R. (1982), An Introduction to Error Analysis: The Study of Uncertainties if
Physical Measurements. University Science Books.
C.4. P.V. Bork, H. Grote, D. Notz, M. Regler (1993), Data Analysis Techniques in High Energy
Physics Experiments. Cambridge University Press.
C.5. H. Coleman and W. Steele, Experimentation and Uncertainty Analysis for Engineers, (2
nd

edition), Wiley, New York, 1999

MEM311: Thermal and Fluid Science Laboratory Appendix D

74 | P a g e

APPENDIX D
FITTING CURVES TO EXPERIMENTAL DATA
by Brandon Terranova 2010, adapted from Bakhtier Farouk and David Staack
The procedures to perform curve fits of experimental data can be some of the most useful and insightful
means of data analysis. Unfortunately, the simplicity with which most current spreadsheet and plotting
programs perform curve fits has allowed students and engineers alike to fit data without analyzing
which type of equation should be best based on physical grounds. For example, suppose that you just
measured the pressure and specific volume of a gas at constant temperature. If you plot your
experimental data and attempt to fit it with a curve, it might be tempting to start trying all of the curve
fits available in your plotting program to determine which gives the lowest error. However, by doing
this, you are neglecting any physics that might determine the proper form of the curve. For a gas that is
represented by the ideal gas equation, an isotherm is expressed as

(D.1)
where the constant is the quantity (nRT). Therefore, it would only make physical sense to fit this data
with a hyperbolic curve. Any other fit that gives a low error (high concentration coefficient) over a
limited range of specific volume could hide important results. What if your plotting program doesnt
explicitly offer a hyperbolic curve fit? Practically any equation can be reduced to a linear function
provided that the variables are properly defined. This procedure is discussed in the following section.

A. General Curve Fitting

Suppose we have a set of measurement *

+ which we can plot on linear graph paper. However, the

relation

(D.2)
is suggested by either a trend in the data or applicable theory. To apply this type of curve fit, we must
determine a, b, and n. The simplest way to do this is to note that Eq. (D.2) can be written as
( )

Taking the natural logarithm of both sides of the equation, we obtain

( ) (D.3)

MEM311: Thermal and Fluid Science Laboratory Appendix D

75 | P a g e

Equation (D.3) has the form of the equation of a straight line, i.e., if we define
( ), m = n, , and . Plotting ( ) vs. on linear-linear graph
paper, we obtain








Where the constant a, b, and n are thus determined. Some common functional relations and
corresponding straight-line plot are given in Ref. D.1.

B. General Curve Fitting

All plotting programs can easily perform a linear curve fit but it is beneficial to understand how these
values are being obtained. Most of these programs use a method of least squares to fit polynomial
curves. The procedure for linear and quadratic curves is discussed below. Higher-order fits generally
become cumbersome to derive but follow the same procedure

Suppose we have a set of n measurements represented by
*

(D.4)
The sum of a set of squares of numbers *

+ about a given number

(D.5)
We want to minimize S with respect to

. At any maximum or minimum, the first derivative must be

zero, i.e.

(D.6)
Figure D.25
MEM311: Thermal and Fluid Science Laboratory Appendix D

76 | P a g e

which implies

(D.7)

This of course, is the definition for the mean value of *

+. From this, you can see that by calculating the

mean value of an array of numbers you are actually minimizing the function S given in Eq. (D.5).
Next, consider a two-dimensional set of n measurements represented as
*

+ (

) (

) (

) (

) (D.8)
where the error in y is independent of the magnitude of the error in x. Suppose we have a linear
relation between y and x.
Recall from the previous section that even if the relationship is exponential, it can be expressed in linear
form by redefining the variables. Define the general linear equation

(D.9)
where

is taken as the optimum y value to represent the data at a given x.

Then
(

)-

(D.10)
We wish to compute a and b such that S is a minimum. Therefore, S must be differentiated with respect
to both a and b and the resulting equations both set equal to zero:

)-

(D.11)

Figure D.26
MEM311: Thermal and Fluid Science Laboratory Appendix D

77 | P a g e

)-(

(D.12)
Solving for a and b leads to

(

)

(D.13)

(

)

(D.14)

The equation that results from this procedure,
(D.15)
is called a regression equation. As previously stated, this method may be used for any functional relation
that can be expressed in terms of a linear relationship. Recall that we previously found that y = ax
b

could be written as log y = log a + b log x.

The above procedure can also be performed to determine the coefficients of the quadratic equation
that best fits a set of data. In general, a quadratic equation is given as

(D.16)
the function S (Eq. (10)) is then given as
(

)-

(D.17)
Equation (21) is then differentiated with respect to a, b, and c and all three equations set equal
to zero, e.g.,

These three equations can then be solved for the three unknowns, i.e., a, b, and c to obtain the
quadratic equation that fits the given data. Higher-order least squares methods follow a similar
development procedure but will require solution of an (N+1) X (N+1) matrix, where N is the order of the
polynomial being used to fit the data.
References
D.1. Holman, J.P., Experimental Methods for Engineers (7
th
edition), McGraw-Hill, New York, 2001.
D.2. H. Coleman and W. Steele, Experimentation and Uncertainty Analysis for Engineers, (2
nd
edition),
Wiley, New York, 1999