MEAM 347

Mechanical Engineering Laboratory II

Laboratory Manual

Do Not Print this on CETS or SEAS Printers

Department of Mechanical Engineering

University of Pennsylvania
September, 1998
 TABLE OF CONTENTS

Introduction ................................................................................................................................... 3

Objectives............................................................................................................................ 3
Course Content ................................................................................................................... 3

Writing a Lab Report ...................................................................................................................... 8

Some General Comments..................................................................................................... 8

Types of Reports ................................................................................................................. 9
Contents of a Report ......................................................................................................... 11
Graphical Presentations ..................................................................................................... 18
Miscellaneous Helpful Hints............................................................................................... 19

Presenting Experimental Data........................................................................................................ 20

Introduction....................................................................................................................... 20
Causes and Types of Experimental Errors .......................................................................... 20
Error Analysis on a Commonsense Basis............................................................................ 22
Uncertainty Analysis .......................................................................................................... 23
Evaluation of Uncertainties for complicated Data Reduction .............................................. 28
Graphical Analysis and Curve Fitting ................................................................................. 30
General Considerations in Data Analysis ............................................................................ 31
Summary ........................................................................................................................... 31
Review Questions .............................................................................................................. 35
Problems ........................................................................................................................... 35

Experiment #1............................................................................................................................... 43
Lift and Drag on 1/48 Scale Model Aircraft

Experiment #2............................................................................................................................... 48
Fluid Friction Flow

Experiment #3............................................................................................................................... 51
Pressure Distribution Over an Airfoil

Experiment #4............................................................................................................................... 57
Cooling Tower Experiment

2
 1. INTRODUCTION

1.1 OBJECTIVES

The objectives of the MEAM 247 and 347 laboratory courses are:

(1) To teach you how to do the experimental work expected from modern engineering
professionals.
(2) To demonstrate to you physical concepts that you are learning in various engineering
courses. The experiments are not necessarily limited to material you have learned in class.
Some of the experiments are designed to make you think and introduce new concepts and
ideas that are beyond traditional classroom instruction.
(3) To stimulate your curiosity and imagination. Hopefully, the course will give you a sense of
measuring and discovering nature, and will introduce you to the excitement of research.
(4) To give you experience with modern instrumentation, and to acquaint you with computer-
aided data acquisition, control, real time data processing and graphical display of results.
(5) To train you in team work.
(6) To enhance your communication and writing skills. Great emphasis is put on the quality and
appearance of the laboratory reports.

1.2 COURSE CONTENT

The material covered and the experiments in the laboratory course are classified into four categories:
i. Experimental methods
ii. Computer aided data acquisition, processing and analysis
iii. Experiments in mechanics, thermal and fluid sciences and mechanical systems
iv. Computer integrated design and manufacturing

We briefly discuss each of these categories next.

Experimental methods

A good experimentalist must be equipped with analytical tools that are required to model physical
systems (including sensors), design experiments and analyze data. In addition, it is very important to
be ab yo youworrite a coherent and concise report. The basics of system dynamics, signal processing,
statistical analysis and technical writing will be taught in this section. Some of this material will be
covered in the classroom.

Computer aided data acquisition, processing and analysis

The primary advantage of computer-aided data acquisition and control of experiments is that the data
are acquired, reduced and analyzed, almost in real time, results are graphically displayed on the
computers' CRTs, and neat, formatted output which you can prepare to be ready for final submission,
can be almost immediately printed out. The experimental data can be compared either with theoretical

3
predictions (if available) or with known engineering correlations. Should discrepancies between
experimental observations and the expected results arise, you can either repeat the experiment or try
to establish the source of the discrepancy. (Contrast this with the more conventional mode of
operation in which you record the data manually, reduce and analyze it later at home. Perhaps you
discover, in the final stages of your analysis, that there is something wrong with the data, and the only
solution is to repeat the experiment another day. Of course, the graphs and tables must be
painstakingly generated manually!) In this section of the course, you will learn what is involved in
interfacing sensors and controllers to computers, how to set up computers to take data or to output
data to control the external world, and how to process, interpret and analyze experimental data that is
often corrupted by errors that are either predictable (explained by a suitable model) or random
(noise). You will use personal computers, data acquisition cards, and appropriate software to collect
data and then analyze it. You will learn to interface sensors and analyze their response characteristics.

Experiments in mechanics, thermal and fluid sciences and mechanical systems

In this section of the course, you will perform experiments that will enhance the materials you are
learning in various engineering courses.

In these experiments, the computer integrated environment with microcomputers and associated
instrumentation is typical of what engineers may expect to find in modern, automated industry. The
microcomputers will substantially enhance your learning environment since you can delegate tedious,
repetitive tasks to the machines and concentrate on observing the fundamental phenomena
demonstrated by the experiments. At the same time, the experiments have been designed to avoid the
use of "black boxes" (as much as possible) and "canned programs".

Computer integrated design and manufacturing

This section of the course is designed to train you in computer integrated engineering analysis, design
and manufacturing techniques. The emphasis is on the intellectual component in "integration" and its
importance in engineering. The experiments will give you a flavor of the automation that is typical in
the manufacturing industry. You will be able to design and draft a product, build a 3-D model of the
product, verify its manufacturability, and manufacture a prototype in the same environment. This
allows you to concurrently design and manufacture with databases integrated into one environment.
(Contrast this with the conventional approach in which a designer sketches out the design, has it
drawn by the draftsman, and then sends it to the shop floor to get it fabricated. Very often, the
designer does not take into consideration the actual manufacturing process and specifies parts that
cannot be machined. Or, the draftsman misinterprets the sketch. At every stage there is delay and the
process of transforming a conceptual idea into a prototype is stretched out to a point where it is a
financial disaster.) This idea of electronic prototyping is fast becoming a reality in today's industry.
You will use state of the art software and numerically controlled milling machines and lathes to build
small machine components out of machinable wax.

Course structure

4
Both MEAM 247 and 347 courses will be continuous through the academic year. Each course
consists of 8-10 laboratory sessions. You will work in groups, typically of 3, and prepare reports for
laboratory experiments.

Time and Place

The lab period for each group will be set during the organizational meeting. The main laboratory is
the GM Lab which is located in room 193 on the west side, ground floor of the Towne Building (on
the same floor as the SEAS machine shop and copy center). Experiments will be conducted in other
labs as well including the wind tunnel facility in 175 Towne, which is behind the machine shop, the
mechanical testing lab in the basement of LRSM, and the Chemical Engineering undergraduate lab in
116 Towne. The lab for each experiment will be noted on the master schedule.

You will be able to perform the experiments only during the scheduled times. However, students will
have additional access to the laboratory to use the computers and prepare for their next experiment.
This access is only during normal work hours - Monday through Friday, 9:00 a.m. to 4:30 p.m.

Books and Manuals

The primary text for this course is this manual to be purchased from the copy center (room 143,
Towne Building). The material specific to each experiment is also kept near the experimental set-up,
consult with the TA if you need it.

Groups

The laboratory exercises and report writing are to be done by an individual student or in teams of
three. You should select partners to form a laboratory team. Select your team members wisely, since
you will work together for the year. Team work is an important aspect of this course and it will
enhance your performance. All students who cannot identify partners should see the instructor or the
TA during the first week the lab meets.

Laboratory reports

You will submit a report for each experiment that you perform. All reports will be graded on the
technical content as well as the writing and presentation. The deadline for submitting rough and final
drafts are shown on the syllabus for your individual group. These deadlines are not changeable.

Technical report submission will be an important factor in your future career. It is an avenue for
communicating your ideas to your superiors and colleagues. The opinion that your superiors form of
you will depend heavily on the quality of such reports. You should make a considerable effort to
master the technique of report writing.

The laboratory exercise grade will be based on:

5
 (a) The quality of preparation for the laboratory session. This may include a short verbal or
written quiz before or during each experiment.
(b) Attendance and performance in the laboratory.
(c) Quality of the report: This will be based on the technical content as well as the quality of
presentation and technical writing. Late assignments will result in lower grades.

Preparation for the laboratory session

Before coming to the laboratory:

1. Read carefully and understand the description of the experiment in the lab manual. You may
go to the lab at an earlier date to look at the experimental facility and understand it better.
Consult the appropriate references to be completely familiar with the concepts and hardware.
At the beginning of the class, if the TA or the instructor finds that a student is not adequately
prepared, they will be graded zero for that experiment and not be allowed to take it.
2. If software is required for the experiment, it must be prepared and brought on floppy diskette
to the lab. You will not be allowed (or have time) to write the software during the laboratory
period.
3. Bring necessary material needed to perform the required preliminary analysis. It is a good
idea to do sample calculations and as much of the analysis as possible during the session. TA
help will be available. Errors in the procedure may thus be easily detected and rectified.
4. Check with one of the TAs one week before the experiment to make sure that you have the
handout for that experiment and all the apparatus.

After the laboratory session

1. Clean up your work area.

2. check with the TA or technician before you leave.
3. Make sure you understand what kind of report is to be prepared and when it is due.
4. Do sample calculations and some preliminary work to verify that the experiment was
successful.

Make-ups and late work

Students must participate in all laboratory exercises as scheduled. They must obtain permission from
the instructor for absence, which would be granted only under justifiable circumstances. In such an
event, a student must make arrangements for a make-up laboratory, which will be scheduled when
laboratory time is available.

Laboratory policies

1. Food and beverages are not allowed in the laboratory at any time.
2. Do not sit or place anything on instrument benches.
3. The teaching assistants and staff are concerned about your safety and for the equipment in the
laboratory. Follow their instructions. If you have any doubt about the way to operate some

6
 equipment, do not guess. Read the instructions carefully and if still in doubt, ask the staff for
assistance.
4. Organizing laboratory experiments requires the help of laboratory technicians and staff. Be
punctual. If you come late to the laboratory, you may be asked to leave. Remember that
some of the sessions include short tests before the experiment and no make-ups will be given.

Notices, Electronic Mail

The entire class will meet only once during the two semesters. Therefore, changes in schedule and
deadlines in class will be announced in the following ways:

1. Announcements will be posted on the notice board outside the GM Lab and on the eniac
newsgroup. It is your responsibility to check this board every week.
2. Messages will be posted on the electronic bulletin board on eniac upenn.meam.meam247 or
347. Read your mail at least twice a week and check the bulletin board.

The easiest way to get in touch with a TA, instructor or staff person is via electronic mail. In the
class notes, there are two pages which tell you how to use it. If you want to meet with an instructor,
send mail to the TA requesting an appointment with preferred times.

7
 2. WRITING A LAB REPORT

2.1 SOME GENERAL COMMENTS

The importance of good report writing and data presentation cannot be overemphasized. No matter
how good an experiment or how brilliant a discovery, it is worthless unless the information is properly
communicated to other people.

A man who uses a great many words to express his meaning is like a bad marksman who
instead of aiming a single stone at an object takes up a handful and throws at it in hopes he
may hit. -Samuel Johnson

Succinct and precise expression are essential for a good technical report. When graphs or tables will
present the idea clearly, use them, but do not also include a wordy explanation which tells the reader
what can be plainly seen by careful inspection of the graph.

Third-person past tense is generally accepted as the most formal grammatical style for technical
reports, and it is seldom incorrect to use such a style. In some instances, first person may be
employed in order to emphasize a point or to stress the fact that a statement is primarily the opinion
of the writer. The usual scientific writing style is also in the passive mode. Examples of the two
styles are:

Third person Equation (5) is recommended for the correlation in accordance with the
limitations of the data as discussed above.
First person We (I) recommended Eq. (5) for the final correlation in accordance with the
limitations of the data presented in our discussion above.

In the first-person statement, the writer is making the recommendation on a much more personal basis
than in the third-person statement. The selection of the proper statement is a matter of idiom, which
depends on many factors, including consideration of the person(s) who will read the report. For a
formal paper in a scientific journal, the third-person statement might be preferable, whereas the first-
person usage might be desirable in an engineering report to an individual.

A writer should make sure that the strong points of a report are clearly set forth and not buried in
excess words. Be specific if you have something to be specific about. Consider the following two
statements:

1. An analysis of the experimental data showed that the average deviation from the theoretical
values was less than 1 percent. Uncertainties in the primary data were shown to account for a
deviation of 0.5 percent. In view of this excellent agreement between Eq. (42) and the
experimental data, this relation is believed to be an adequate representation of the physical
phenomena and is recommended for calculation purposes.

2. The experimental data are in good agreement with the theoretical development. In view of
this favorable comparison the assumptions pertaining to the derivation are verified.

8
Note the difference between the two statements. The first statement is quite specific and leaves the
reader with a feeling of confidence in the experimental data and theoretical analysis. Upon examining
the second statement the reader will immediately ask: How good is "good"; how favorable is
"favorable"? The author has seen such statements applied to experimental data that differed from
theoretical values by such a large factor that the veracity of the writer might be questioned by a
careful reader.

A brief, general procedure for report writing might take the following form:

1. Make a written outline of the report with as much detail as possible.

2. Let the outline "cool" for a period of time while you direct your thoughts to other matters.
3. Go back to the outline and make whatever changes you feel are necessary.
4. Write the report in rough draft form as quickly as possible.
5. Let the report cool for a period of time.
6. Go back and make corrections on your draft. You will probably find that you did not say
things quite the way you would like to, did not include as much information as you wanted to,
or made several stupid mistakes in some of the data analysis.
7. If possible, have a colleague scrutinize the report carefully. This person should be one whose
competence you respect.
8. Consider your colleague's comments very carefully, and rewrite the report in its final form.

The above remarks on report writing can be considered as general ones which may be applicable to
the broad range of reports, papers, or monographs the engineer will be called upon to write. Of
course, a carefully constructed outline is always the best starting point and will help the writer to
make sure that all pertinent information is included.

2.2 TYPES OF REPORTS

Informational Reports

Informational reports usually follow a letter or interoffice memo format and are brief and to the point.
Two examples are given below.

MEMO FORMAT
TO: J. J. Brown
FROM: B. R. Smith
DATE:
SUBJECT: Test of TV Satellite Uplink Quality

In accordance with your request of _______, we conducted tests with the TV satellite uplink on
______. Signals were monitored and videotaped at locations in Boulder, Colorado; Andover,
Massachusetts; and Dayton, Ohio. All three sites reported satisfactory video quality at transmission

9
levels below 125 watts. This value is well below the maximum power level of 300 watts for the
uplink. Difficulties were experienced with the audio signal caused by 60-Hz interference from the on-
desk monitor. The interference was eliminated by better location of the desk microphone or using
and alternate lapel mike.

The three receiving sites are returning videotapes of all the tests to my office where they will be
available for your inspection. Based on our verbal communications with the sites, we believe that the
entire uplink operation is now delivering satisfactory video and audio signals.

LETTER FORMAT
Mr. J. R. Marshall
Acme Development Corp.
501 Main Street
Dallas, Texas 75201

Dear Mr. Marshall:

Following out meeting of ______ we conducted preliminary test of the HVAC system in your
two-story building at 10123 North Road. The purpose of these tests was to determine if the main
unit was performing in accordance with specifications. We made the following measurements:

1. Dry-bulb and wet-bulb temperatures for all inlet and exit airstreams of the main unit.
2. Air velocities for these airstreams.
3. Leakage from the outside air damper.
4. Power consumption of the main unit.

Calculations based on these measurements indicate that the main unit is capable of achieving
its rated cooling capacity; however, measurements on the outside air damper indicate a leakage of 15
percent when the damper is fully closed. This leakage places an additional load on the cooling unit so
that some deterioration in performance is experienced in very hot and humid weather.

We recommend that the outside air damper be replaced; therefore, we would expect entirely
satisfactory operation of the system.
Please let us know if we can be of further service.

Sincerely,.
R.W. Smith, P.E.
Consulting Engineer

Formal Reports

Formal reports are usually organized to include several of the elements described in the following
section. Lengths obviously vary with the complexity of the report. The format for the report may be

10
specified by the person or organization to receive the document without much discretion on the part
of the writer.

Papers and Journal Articles

Papers and journal articles written for the professional engineering or scientific community may
certainly be classified as "formal" reports. In many cases the journal or literature of the sponsoring
organization will include a section entitles "Information of Authors" which specifies the format to be
followed. The writer should pay close attention to such information because non-adherence to the
required format may result in rejection of the paper even if it is judged by the referees to have good
technical merit.

Tutorial Reports

A tutorial report may also be classified as "formal," but - in contrast to papers or journal articles - it is
not directed to a knowledgeable professional in the field. The purpose of the tutorial report is to
teach someone who is not well informed in the subject matter. For example, a manufacturer of
flowmeters might ask a number of its engineering staff to prepare a report which describes how the
flowmeters work, the difficulties to be encountered in installation, proper selection techniques, and
accuracies which may be expected. Such a report would probably be issued as a company brochure
directed at possible users of the product. The report must be factual and helpful to the reader, but it
does not convey "new" information as would a paper or a journal article.

Books and Monographs

We mention books and monographs mainly for the sake of completeness. Obviously, some large
reports are long enough to be called "books," and indeed, some are eventually published as books.
Book publishers frequently have their own standards for format and style.

2.3 CONTENTS OF A REPORT

We have mentioned that a good outline is always the starting point for a good report. To aid in this
outline construction we now give a brief discussion of different report elements. Not all these
elements will be included in every report, and the writer must choose those appropriate to the
audience.

Front Matter

Front matter includes the title page, with author affiliation(s), sponsor of report activity (if any), table
of contents, list of nomenclature, list of figures, preface (if any), and a letter of transmittal if required.

In many cases the letter of transmittal will be a simple interoffice memo for the company. For a major
report to a project sponsor the letter may be more elaborate.

11
A variety of schemes are used for organizing the table of contents. Many authors like to number
every paragraph, thereby creating many subheadings. A technical paper for a journal may or may not
use numbered headings and would not include a table of contents.

Nomenclature lists, identifying the symbols naming the used variables, are normally required for both
journal articles and comprehensive technical reports. They may not be necessary in brief reports for a
limited audience.

Common practice is to list terms or variables in alphabetical order, followed by Greek or foreign
symbols in order, and then followed by subscripts and superscripts. The units of all terms should be
given with the nomenclature list. If more than one set of units may be used, as with English and SI,
both should be stated. Standard abbreviations should be employed for stating the units. Some
examples of nomenclature listings are as follows:

a Local velocity of sound, m/s or ft/s

Ac Flow cross-sectional area, m2 or ft2
k Thermal conductivity, W / m ⋅°C or Btu/h ⋅ft⋅°F
Re=pud/µ Reynolds number, dimensionless
µ Dynamic viscosity, kg/m⋅s or lbm/h ⋅ft
p Density, kg/m3 or lbm/ft3

12
Subscripts
i Inside
o Outside
8 Free-stream conditions

When using built-up units, as with thermal conductivity k above, several choices of style are available,
such as:

Btu
1.
(h)(ft)(°F)
2. Btu/(h)(ft)(°F)
3. Btu/h ft °F
4. Btu/h-ft-°F
5. Btu/h/ft/°F

The first four styles all mean the same thing. Styles 1 and 2 are more work for the typist, style 34 is
easiest for the typist, and is common in typeset publications. Style 5 does appear from time to time,
but the multiple slash marks can be confused for multiple divisions unless the reader knows the
subject. In general, style 5 should be avoided. Inclusion of units in the nomenclature list is very
important. A reader unfamiliar with the terms finds the absence of units very annoying.

Lists of tables and figures are frequently included in large reports but are never used in papers for
journal publication. Inclusion of a preface is optional in reports: it is normally used only in
monographs (books).

Abstracts

Someone stated an old rule for army methods of communications: Tell them what you are going to
tell them; tell them; and then tell them what you told them. The abstract should attempt to
accomplish the first objective in a very short, succinct format, without mathematical formulations. It
should tell, specifically, but very briefly, what was done, and the conclusions which resulted from the
work. Keep in mind that there are many people who will read only the abstract of a report because of
heavy demands on their time. This is especially true of managers who must review a large number of
reports on a broad range of subjects. A well-written abstract becomes particularly useful for these
people, and may arouse their interest in examining the report in more detail.

Some writers may choose to use the terms "summary" or "executive summary" instead of abstract,
but the purpose of the section is the same.

Introduction

The introduction section is used to clearly state the motivation for performing the work, i.e., to define
the problem, and sometimes a review of previous work is given too.

Background and Previous Work

13
A survey of the literature may be appropriate and can be included under this title. It could also be
included in an introduction. In most cases the survey should not be simply chronological, but written
to present the current knowledge of the subject and the specific deficiencies in that knowledge.

This section can vary widely in length, depending on what one assumes for the knowledge level of the
intended reader. The length also depends on what the reader needs to know in order to understand
and appreciate the remainder of the report. It is rather easy to go off the deep end in this section
because with most technical subjects there are great volumes of background and previous work. The
writer should keep this section focused so that it both acknowledges previous work and clearly points
to the need for the current study.

Because of space limitations, most technical papers and journal articles try to minimize the length of
this section.

Theoretical Presentation(s)

In some reports a large section will be devoted to development of theoretical information applicable
to the subject. This section enables the reader to understand the implications of the experimental
work and aids in proper interpretation of the data.

Of course, some reports are strictly theoretical in nature, and thus this section forms the main body
for those reports. To encourage more efficient use of the reader's time, some writers may use this
section as a vehicle to summarize theoretical presentations and relegate long detailed derivations to
appendices. In this way, the reader may get to the heart of the presentation more quickly, while
retaining the option of examining details later.

In the theoretical presentations(s) one has the option of defining units for the various terms as they
are introduced, or specifying the units only in the nomenclature list. The latter practice is preferable
for papers because it conserves space. For unusual variables the units might be defined when they are
introduced.

Display of mathematical formulas should follow standard practice in the field.

Experimental Apparatus and Procedure

Sufficient information must be supplied on the apparatus and experimental procedures for the reader
to understand what was done. If the report is concerned with research and new knowledge, a rather
detailed discussion of the apparatus may be necessary. If test results are being reported in accordance
with standard procedures (ASME, ASTM, etc.), then the appropriate procedures can be cited without
furnishing details in the report. An error analysis and calibration procedures for the experiments must
be presented, as explained in Chapter 3.

14
Great variations in length of this section can be used. An extensive report might include all details of
the apparatus and instrumentation, while a technical paper or journal article would only give a brief
summary.

Results of Experiments

One normally will include a separate section in the report to give results of the experiments in a form
which is consistent with the needs of the intended audience. Clear tabular and graphical presentations
should be used as much as possible to conserve words in the report. Of course, some verbal
discussions of the graphs and tables must be given, but such discussion should focus the reader's
attention on salient features of the data, not just recite numbers or parameters which are obvious upon
inspection. In other words the written and graphical presentations should be complementary so that
the reader's time is conserved.

Section 2.5, on graphical presentations, gives suggestions for good practice.

Interpretation of Results

Once the experimental results have been presented in a clear form, the author has a responsibility to
interpret the results in the light of the theoretical presentation and the work of others in the same
general subject area. In this section the background, theoretical presentation, and experimental
results are all brought together to lead the reader to the conclusions of the study. It is very important
to ensure that the results can be justified based on lows of nature, even if only on a qualitative level.

Of course, there are many instances where one is concerned only with presentation of results and no
interpretation is required. For instance, a calibration test for a thermocouple calls only for a
calibration curve, table, or polynominal relation. The results speak for themselves.

Conclusions and Recommendations

By the time the reader has reached this section, most of the conclusions of the work should already
have been drawn. The object of the conclusion section is to collect all the important results and
interpretations in clear summary form. This is the section which tells the reader what was covered in
the body of the report. There will be many readers who will read only the abstract or conclusion
sections of a report, so it is especially important that they be carefully written.

Some writers like to include recommendations for action or further study in this section. A typical
action statement might be:

Calibration of the model 802 flowmeter has now been completed and production runs may
begin immediately.

While a further-study recommendation could be:

15
 The calibration procedure for the Model 802 flowmeter has now been established and reliable
data obtained for the temperature range of 50 to 100°F. It is recommended that further data
be obtained up to a temperature of 300°F. Once these data are obtained, production runs may
begin.

Acknowledgments

Many people may contribute to a project who are not listed as formal authors of the report. An
acknowledgment section may be used to recognize these contributions, as well as the sponsors of the
work (if applicable). For technical papers it is normally placed after the conclusions section, while for
other reports it may appear as a part of front matter.

Appendix Materials

Wide latitude is available to the writer in the types of material which may be placed in appendices. Of
course, the writer may choose to have no appendix at all and include all information in the body of the
report. Otherwise, one or more of the following might appear in the appendix:

1. Detailed mathematical derivations which are summarized in the body of the report, as
described under "Theoretical Presentation(s)" above.
2. Tabulations of raw experimental data which are summarized or correlated in the body of the
report.
3. Calibration information for instruments or sensors employed in the experiments.
4. Uncertainty analysis of the experiments, the results of which may be stated or summarized in
the body of the report.
5. Tabulations or graphs of material properties used in the report.
6. In some cases, calculation charts or materials obtained from other sources.
7. Detailed computer programs used in the work, which may be referenced in the report.

We can see that the general purpose of the appendix is to remove detailed clutter from the body of
the report so that the reader may appreciate the work and conclusions in a shorter period of time. For
those interested, the details are still available for study. Most technical papers and journal articles do
not have much appendix material (if any at all) because the very object of these papers is to present
just the essence of the work.

References and Bibliographies

Most technical reports require references to the work of others. References should be cited when a
work was used in writing the report. For example, the statement:

Cheesewright [1] discusses turbulent natural convection...

with a citation:

16
1. Cheesewright, R.: Turbulent Natural Convection from a Vertical Plane Surface, J. Heat Trans.,
vol. 90, p. 1, 1968.

is a valid reference citation. On the other hand, a standard textbook in heat transfer might have
general information on free convection and yet not be employed as a specific reference for the report.
It is common practice to list such sources under the title of Bibliography. Normally, only extensive
works like textbooks or monographs have separate bibliographies. Because of the extra space
consumed, bibliographies are almost never used in papers for journal publication.

Some latitude is available in the manner of literature citation. The above reference could be listed
without its title, i.e.,

1. Cheesewright, R., J. Heat Trans., vol. 90, p. 1, 1968.

Although such citations may be accepted by various journals, and do conserve space, they should be
avoided whenever possible because the absence of the article title is very annoying to a reader. The
object of a report is to communicate. A procedure which impedes communication is to be shunned.

An alternate citation technique is to list references in alphabetical order of the first author, then
second author, and then year. No specific reference number is given and citation in the body of the
report refers to the year of publication. When the same author has more than one citation in one year,
the year is followed by a, b, c, and so on. For one or two authors both names are cited; for more than
two authors only the first author followed by "et. al." is cited. The following examples illustrate this
technique.

1. Lee, Y., Korpela, S. A., and Horne, R. N., 1982, "Structure of Multi-Cellular Natural Convection
in a Tall Vertical Annulus," Proceedings, 7th International Heat Transfer Conference, U. Grigull
et al., eds., Hemisphere Publishing Corp., Washington, D.C., vol. 2, pp. 221-226.
2. Kwon, O.K., and Pletcher, R.H., 1981, "Prediction of the Incompressible Flow Over a Rearward-
Facing Step," Technical Report HTL-26, CFD-4, Iowa State Univ., Ames, Iowa.
3. Sparrow, E.M., 1980a, "Fluid-to-Fluid Conjugate Heat Transfer for a Vertical Pipe-Internal
Forced Convection and External Natural Convection," ASME Journal of Heat Transfer, vol. 102,
pp. 402-407.
4. Sparrow, E.M., 1980b, "Forced-Convection Heat Transfer in a Duct Having Spanwise-Periodic
Rectangular Protuberances," Numerical Heat Transfer, vol. 3, pp. 149-167.

The corresponding citations in the report would appear as:

1. Lee et al. (1982) discovered...

2. Kwon and Pletcher (1981) predicted...
3. Sparrow (1980a) studied...
4. Sparrow (1980b) observed...

17
The reader will notice that some of the above citations list the journal title in full (Journal of Heat
Transfer) while only an abbreviation (J. Heat Trans) is used in others. Either is correct, but some
journals and/or publishers follow their own standard format.

2.4 GRAPHICAL PRESENTATIONS

Most engineering reports include some form of graphical presentation: a simple plot of raw data, a
statistical distribution of data, correlation of data, and perhaps a comparison of experimental data
with analytical predictions. Some of the types of graphs used are:

1. General curves or "trends"

2. Detailed plots of experimental data
3. Accurate graphs which are used for calculation purposes
4. Schematic diagrams
5. Scale diagrams for experimental apparatus
6. Graphs in a format to be used for presentation

Style

Regardless of the purpose of the graph, there are certain elements of good style which should be
common to all.

1. Label the graph. Make sure all coordinates are labeled consistently, i.e., don't use all uppercase
on one coordinate and mixed upper- and lowercase on the other.

2. If numerical values of variables are graphed, make sure that each coordinate has scale markings
and that each label for a variable includes the units for that variable. Try to maintain the same unit
system for all variables, i.e., do not mix English and SI units.

3. Recognize that some graphs may be reduced in size later because of space limitations. This is
particularly true for those to be published in technical papers or symposia volumes. If any
reduction is anticipated, the size of the labels and unit specifications on the graph must be
increased so that they will be legible in their eventual format. For graphs prepared on computers
this will be an easy matter to arrange.

4. The purpose of graphs is to convey information. Sometimes this is best accomplished by plotting
several curves on one graph; sometimes not. For example, suppose that six pressure traverses are
taken along the length of a wind tunnel for six different flow rates. In this case it would probably
be best to include all six sets of data on one graph with each curve clearly labeled. A use of six
separate graphs would not communicate as well and would certainly consume more space. On
the other hand, suppose a heated plate is placed in the tunnel and heat transfer data collected for
the same six flow rates. It would not be prudent to present these data on the same graph as the
pressure data because the display would be too cluttered. A second graph would be in order.
There is no hard rule to apply in these circumstances. Common sense usually works fine.

18
5. There is a class of presentations called G and A graphs (for Generals and Admirals). They are
usually broad-brush and never loaded with a lot of data points or complicated mathematical
expressions. The purpose of these graphs is to present general trends without precise data points,
and they are normally employed for readers or viewers not very familiar with the subject.

6. Careful thought should be given to the selection of coordinates for a graph. Even with plots of
raw data one may wish to choose a logarithmic plot over a linear system if the data spans are a
very wide range and is expected to have an exponential trend.

Legends for Graphs

Although a pertinent part of the legend for a graph may be placed directly on the graph itself, the
curves are labeled typically with numerals (1, 2, 3, etc.) or letters (a, b, c, etc.), or by different line
texture or symbols, which are then defined or explained in the figure title placed under the figure.

Audiovisual presentations expose the audience more briefly to the graphs, and legends have to be
more clear than those used in a written report.

2.5 MISCELLANEOUS HELPFUL HINTS

As a further aid to the writer we give the following bits of advice which do not fall in the categories
of other sections on report writing.

1. Always give great attention to the audience for which the report is written.
2. Set up a list of symbols to be used in the report before starting to write. Then be sure to use these
symbols consistently as the report is written. Be particularly careful about switching between
uppercase and lowercase letters for the same variable.
3. In choosing symbols to be used in a report try to follow the practice used in standard textbooks or
technical journals in the field. Don't invent new symbols which will make the report hard for the
reader to understand.
4. Once writing is started, write the report as if you had to make it correct on the first draft. Take
your time. This will pay off later. Of course, unless you are very good, or very lucky, changes
will have to be made, but they usually can be minimized with this approach.
5. If you are not word-processing or typing the report yourself, consider the typist or word-
processor operator. Be careful with handwriting, and avoid confusion between English and Greek
symbols.

Acknowledgement: The material, in Sections 2 and 3, is reproduced (with some changes) from
"Experimental Methods for Engineers", J.P. Holman and W.J. Gajda Jr. Fifth Ed. (1989) McGraw-
Hill, Inc. with permission from the publisher.

19
 3. PRESENTING EXPERIMENTAL DATA

3.1 INTRODUCTION

This relatively brief summary should be augmented by readings from the reference list, especially
Holman [1].

Some form of analysis must be performed on all experimental data. The analysis may be a simple
verbal appraisal of the test results, or it may take the form of a complex theoretical analysis of the
errors involved in the experiment and matching of the data with fundamental physical principles.
Even new principles may be developed in order to explain some unusual phenomenon. Our
discussion in this chapter will consider the analysis of data to determine errors, precision, and general
validity of experimental measurements. The correspondence of the measurements with physical
principles is another matter, quite beyond the scope of our discussion. Some methods of graphical
data presentation will also be discussed. The interested reader should consult the monograph by
Wilson [2] for many interesting observations concerning correspondence of physical theory and
experiment.

The experimentalist should always know the validity of data. The automobile test engineer must
know the accuracy of the speedometer and gas gage in order to express the fuel-economy
performance with confidence. A nuclear engineer must know the accuracy and precision of many
instruments just to make some simple radioactivity measurements with confidence. In order to
specify the performance of an amplifier, an electrical engineer must know the accuracy with which the
appropriate measurements of voltage, distortion, etc., have been conducted. Many considerations
enter into a final determination of the validity of the results of experimental data, and we wish to
present some of these considerations in this chapter.

Errors will creep into all experiments regardless of the care which is exerted. Some of these errors
are of a random nature, and some will be due to gross blunders on the part of the experimenter. Bad
data due to obvious blunders may be discarded immediately. But what of the data points that just
"look" bad? We cannot throw out data because they do not conform with our hopes and expectations
unless we see something obviously wrong. If such "bad" points fall outside the range of normally
expected random deviations, they may be discarded on this basis of some consistent statistical data
analysis. The key here is "consistent". The elimination of data points must be consistent and should
not be dependent on human whims and bias based on what "ought to be". In many instances it is very
difficult for the individual to be consistent and unbiased. The pressure of a deadline, disgust with
previous experimental failures, and normal impatience all can influence rational thinking processes.
However, the competent experimentalist will strive to maintain consistency in the primary data
analysis. Our objective in this chapter is to show how one may go about maintaining this consistency.

3.2 CAUSES AND TYPES OF EXPERIMENTAL ERRORS

In this section we present a discussion of some of the types of errors that may be present in
experimental data and begin to indicate the way these data may be handled. First, let us distinguish
between single-sample and multisample data.

20
Single-sample data are those in which some uncertainties may not be discovered by repetition.
Multisample data are obtained in those instances where enough experiments are performed so that the
reliability of the results can be assured by statistics. Frequently, cost will prohibit the collection of
multisample data, and the experimentor must be content with single-sample data and prepared to
extract as much information as possible from experiments. The reader should consult Refs [1-3] for
further discussion on this subject, but we state a simple example at this time. If one measures
pressure with a pressure gage and a single instrument is the only one used for the entire set of
observations, then some of the error that is present in the measurement will be sampled only once no
matter how many times the reading is repeated. Consequently, such as experiment is a single-sample
experiment. On the other hand, if more than one pressure gage is used for the same total set of
observations, then we might say that a multisample experiment has been performed. The number of
observations will then determine the success of this multisample experiment in accordance with
accepted statistical principles.

The magnitude of an experimental error is ultimately unknown. If the experimenter know what the
error was, he or she would correct it and it would no longer be an error. In other words, the real
errors in experimental data are those factors that are always vague to some extent and carry some
amount of uncertainly. Our task is to determine just how uncertain a particular observation may be
and to devise a consistent way of specifying the uncertainty in analytical form. A reasonable
definition may be taken as the possible range that error may have. This uncertainty may vary a great
deal depending upon the circumstances of the experiment. Perhaps it is better to speak of
experimental uncertainty instead of experimental error because of magnitude of an error is always
uncertain. Both terms are used in practice, however, so the reader should be familiar with the
meaning attached to the terms and the ways that they relate to each other.

At this point, we may mention some of the types of errors that may cause uncertainty in an
experimental measurement. First, there can always be those gross blunders in apparatus or instrument
construction which may invalidate the data. Hopefully, the careful experimenter will be able to
eliminate most of these errors. Second, there may be certain fixed errors which will cause repeated
readings to be in error by roughly the same amount but for some unknown reason. These fixed errors
are sometimes called systematic errors. Third, there are the random errors, which may be caused by
personal fluctuations, random mechanical and electronic fluctuations in the apparatus or instruments,
various influences of friction, etc. These random errors usually follow a certain statistical distribution,
but not always. In many instances it is very difficult to distinguish between fixed errors and random
errors.

The experimentalist may sometimes use theoretical methods to estimate the magnitude of a fixed
error. For example, consider the measurement of the temperature of a hot gas stream flowing in a
duct with a mercury-in-glass thermometer. It is well known that heat may be conducted from the
stem of the thermometer into the surrounds. In other words, the fact that part of the thermometer is
exposed to the surroundings at a temperature different from the gas temperature to be measured may
influence the temperature of the stem of the thermometer. Therefore, the temperature we read on the
thermometer is not the true temperature of the gas, and it will not make any difference how many
readings are taken we shall always have an error resulting from the heat-transfer condition of the stem

21
of the thermometer. This is fixed error, and its magnitude may be estimated with theoretical
calculations based upon known heat transfer processes and thermal properties of the gas and the glass
thermometer.

3.3 ERROR ANALYSIS ON A COMMONSENSE BASIS

We have already noted that it is somewhat more explicit to speak of experimental uncertainty rather
than experimental error. Suppose that we have satisfied ourselves with the uncertainty in some basic
experimental measurements, taking into consideration such factors as instrument accuracy,
competence of the using the instruments, etc. Often, the primary measurements must be combined to
calculate a particular result that is desired. We shall be interested in knowing the uncertainty in the
final result due to the uncertainties in the primary measurements. This may be done by a
commonsense analysis of the data which may take many forms. To find the worst-case error, all the
errors in the primary measurements are combined in the most detrimental way . Consider the
calculation of electric power from

P = EI

where E (voltage) and I (current) are measured as

E = 100 V ± 2 V
I = 10 A ± 0.2 A

The nominal value of the power is 100 x 10 =1000 W. By taking the worst possible variations in
voltage and current, we could calculate.

Pmax = (100 + 2)(10 + 0.2) = 1040.4 W

Pmin = (100 - 2)(10 - 0.2) = 960.4 W

Thus, using this method of calculation, the uncertainty in the power is +4.04 percent, -3.96 percent.
It is quite unlikely that the power would be in error by these amounts because the voltmeter variations
would probably not correspond with the ammeter variations. When the voltmeter reads an extreme
"high," there is no reason why the ammeter must also read an extreme "high" at that particular instant;
indeed, this combination is most unlikely.

The simple calculation applied to the electric-power equation above is a useful way of inspecting
experimental data to determine what errors could result in a final calculation; however, the test is too
severe and should be used only for rough inspections of data. It is significant to note, however, that if
the results of the experiments appear to be in error by more than the amounts indicated by the above
calculation, then the experimenter had better examine the data more closely. In particular, the
experimenter should look for certain fixed errors in the instrumentation (such as the reading on the
bathroom scale when no weight is on it), which may be eliminated by applying either theoretical or
empirical corrections.

22
As another example we might conduct an experiment where heat is added to a container of water. If
our temperature instrumentation should indicate a drop in temperature of the water, our good sense
(i.e., knowledge of the laws of nature) would tell us that something is wrong and the data point(s)
should be thrown out. No sophisticated analysis procedures are necessary to discover this kind of
error.

3.4 UNCERTAINTY ANALYSIS

A method of estimating uncertainty in experimental results has been presented by Kline and
McClintock [3]. The method is based on a careful specification of the uncertainties in the various
primary experimental measurements. For example, a certain pressure reading might be expressed as
p = 100 kN/m2 ± 1 kN/m2

When the plus or minus notation is used to designate the uncertainty, the person making this
designation is stating the degree of accuracy with which he or she believes the measurement has been
made. We may note that this specification is in itself uncertain because the experimenter is naturally
uncertain about the accuracy of these measurements.

If a very careful calibration of an instrument has been performed recently, with standards of very high
precision, then the experimentalist will be justified in assigning a much lower uncertainty to
measurements than if they were performed with a gage or instrument of unknown calibration history.

To add a further specification of the uncertainty of a particular measurement, Kline and McClintock
propose that the experimenter specify certain odds for the uncertainty. The above equation for
pressure might thus be written

p = 100 kN/m2 ± 1 kN/m2 (20 to 1)

In other words, the experimenter is willing to bet with 20 to 1 odds that the pressure measurement is
within ± 1 kN/m2. It is important to note that the specification of such odds can only be made by
the experimenter based on the total laboratory experience.

Suppose a set of measurements is made and the uncertainty in each measurement may be expressed
with the same odds. These measurements are then used to calculate some desired result of the
experiments. We wish to estimate the uncertainty in the calculated result on the basis of the
uncertainties in the primary measurements. The result R is a given function of the independent
variables x1, x2, x3, ..., xn . Thus,

R = R(x1, x2, x3,....,xn ) (3.1)

Let wr be the uncertainty in this result and w1, w2, ..., wn be the uncertainties in the independent
variables. If the uncertainties in the independent variables are all given with same odds, then the
uncertainty in the result having these odds is given in Ref. [1] as

23
 1/2
 ∂R 
2
 ∂R 
2
 ∂R 
2

=  wn  
∂x1 1 
+ 
∂x2 2 
wR w w + L + 

    ∂x n   (3.2)

If this relation is applied to the electric power relation of the previous section, the expected
uncertainty (also called the root-mean-square error) is 2.83 percent instead of 4.04 percent.

Example 3.1 The resistance of a certain size of copper wire is given as

R = Ro [1 + α (T-20)]

where Ro = 6? ± 0.3 percent is the resistance at 20°C, α = 0.004°C-1 ± 1 percent is the temperature
coefficient of resistance, and the temperature of the wire is T = 30 ± 1°C. Calculate the resistance of
the wire and its uncertainty.

Solution. The nominal resistance is

R = (6)[1 + (0.004)(30 - 20)] = 6.24?

The uncertainty in this value is calculated by applying Eq. (3.2). The various terms are:

∂R
= 1 + α (T - 20 ) = 1 + (0.004)(30 - 20) = 1.04
∂Ro
∂R
= R o (T - 20 ) = (6)(30 - 20) = 60
∂α
∂R
= Ro α = (6)(0.004) = 0.024
∂T
w ro = (6)(0.003) = 0.018Ω
wα = (0.004)(0.01) = 4 x 10-5 °C-1
wT = 1°C

Thus, the uncertainty in the resistance is

wR = [(1.04)2(0.018)2 + (60)2(4x10-5)2 + (0.024)2(1)2]1/2
= 0.0305? or 0.49%

Particular notice should be given to the fact that the uncertainty propagation in the result wR
predicted by Eq. 3.2 depends on the squares of the uncertainties in the independent variables wn .
This means that if the uncertainty in one variable is significantly larger than the uncertainties in the
other variables, say, by a factor of 5 or 10, then it is the largest uncertainty that predominates and the
others may probably be neglected.

24
To illustrate, suppose there are three variables with a product of sensitivity and uncertainty
[(?R/?x)wx] of magnitude 1, and one variable with a magnitude of 5. The uncertainty in the result
would be
(52 + 12 + 12 + 12 ) = 28 = 5.29
1/2

The importance of this brief remark concerning the relative magnitude of uncertainties is evident
when one considers the design of an experiment procurement of instrumentation, etc. Very little is
gained by trying to reduce the "small" uncertainties. Because of the square propagation it is the large
ones that predominate, and any improvement in the overall experimental result must be achieved by
improving the instrumentation or technique connected with these relatively large uncertainties. In the
examples and problems that follow, both in this chapter and throughout the book, the reader should
always note the relative effect of uncertainties in primary measurements on the final result.

The reader is cautioned to examine possible experimental errors before the experiment is designed
and conducted. Equation (3.2) may be used very effectively for such analysis, as we shall see in the
sections and chapters that follow. A further word of caution may be added here. It is equally as
unfortunate to overestimate uncertainty as to underestimate it. An underestimate gives false security,
while an overestimate may make one discard important results, miss a real effect, or buy much too
expensive instruments. The purpose of this chapter is to indicate some of the methods for obtaining
reasonable estimates of experimental uncertainty.

In the previous discussion of experimental planning we noted that an uncertainty analysis may aid the
investigator in selecting alternative methods to measure a particular experimental variable. It may
also indicate how one may improve the overall accuracy of a measurement by attacking certain critical
variables in the measurement process. The next three examples illustrate these points.

Example 3.2 Selection of measurement method. A resistor has a nominal stated value of
10? ± 1 percent. A voltage is impressed on the resistor, and the power dissipation is to be calculated
in two different ways: (1) from P = E2/R and (2) from P = EI. In (1) only a voltage measurement will
be made, while both current and voltage will be measured in (2). Calculate the uncertainty in the
power determination in each case when the measured values of E and I are:

E = 100 V ± 1%
I = 10 A ± 1%

FIGURE EXAMPLE 3.2

Power measurement across a resistor.

Solution. The schematic is shown in the accompanying figure. For the first case we have

25
 ∂P ∂P 2
2E E
= = - 2
∂E R ∂R R

and we apply Eq. (3.2) to give

 2E 2 2  E2 2
2 1/2

wP =   w E + - 2  w R  (a)
 R 
  R  

Dividing by P = E2/R gives
 wE 2 w R  
2 1/2
wP
= 4  +    (b)
P  E R

Inserting the numerical values for uncertainty,

wP
[
= 4(0.01)2 + (0.01)2 ]
1/2
= 2.236%
P

For the second case we have

∂P ∂P
= I = E
∂E ∂I

and after similar algebraic manipulation, we obtain

wP  w E 2 w I  
2 1/2

=   +    (c)
P  E  I 

Inserting the numerical values of uncertainty,

wP
[
= (0.01) + (0.01) ]=
2 2 1/2
1.414%
P

Thus, the second method of power determination provides considerably less uncertainty than the first
method, even though the primary uncertainties in each quantity are the same. In this example the
utility of the uncertainty analysis is that it affords the individual a basis for selection of a measurement
method to produce a result with less uncertainty.

Example 3.3 Instrument selection. The power measurement in Example 3.2 is to be conducted by
measuring voltage and current across the resistor with the circuit shown in the accompanying figure.
The voltmeter has an internal resistance Rm, and the value of R is known only approximately.
Calculate the nominal value of the power dissipated in R and the uncertainty for the following
conditions:

R = 100? (not known exactly)

Rm = 1000? ± 5%

26
 I = 5 A ± 1%
E = 500 V ± 1%

FIGURE EXAMPLE 3.3

Effect of meter impedance on measurement.

Solution. A current balance on the circuit yields

I1 + I2 = I

E E
+ = I
R Rm

and
E
I1 = I - (a)
Rm

The power dissipated in the resistor is

E2
P = EI1 = EI - (b)
Rm

The nominal value of the power is thus calculated as

5002
P = (500)(5) - = 2250 W
1000

In terms of known quantities the power has the functional form P = f(E,I,Rm), and so we form the
derivatives
∂P 2E ∂P
= I- = E
∂E Rm ∂I

∂P E2
= 2
∂Rm Rm

The uncertainty for the power is now written as

 2E  2  E2  2 
2 1/2

w P =  I -  w E + E w 1 +  2  w Rm 
2 2 2
(c)
 Rm 
  Rm  


27
Inserting the appropriate numerical values gives

 
1/2
 10 4 
2
1000  2
2
w P = 5 -  5 + (25 x 10 )(25 x 10 ) + 25 x
4 -4
 (2500 )

 1000   10 6  

= [16 + 25 + 6.25] (5)
1/2

= 34.4 W
or
wP 34.4
= = 1.53%
P 2250

In order of influence on the final uncertainty in the power we have

1. Uncertainty of current determination

2. Uncertainty of voltage measurement
3. Uncertainty of knowledge of internal resistance of voltmeter

Comment. There are other conclusions we can draw from this example. The relative influence of
the experimental quantities on the overall power determination is noted above. But this listing may be
a bit misleading in that it implies that the uncertainty of the meter impedance does not have a large
effect on the final uncertainty in the power determination. This results from the fact that Rm>>R(Rm
= 10R). If the meter impedance were lower, say, 200? , we would find that it was a dominant factor
in the overall uncertainty. For a very high meter impedance there would be little influence, even with
a very inaccurate knowledge of the exact value of Rm. Thus, we are led to the simple conclusion that
we need not worry too much about the precise value of the internal impedance of the meter as long as
it is very large compared with the resistance we are measuring the voltage across. This fact should
influence instrument selection for a particular application.

3.5 EVALUATION OF UNCERTAINTIES FOR COMPLICATED DATA REDUCTION

We have seen in the preceding discussion and examples how uncertainty analysis can be a useful tool
to examine experimental data. In many cases data reduction is a rather complicated affair and is often
performed with a computer routine written specifically for the task. A small adaptation of the routine
can provide for direct calculation of uncertainties when analytical determination of the partial
derivatives in Eq. (3.2) is difficult. We still assume that this equation applies, although it could
involve several computational steps. We also assume that we are able to obtain estimates by some
means of the uncertainties in the primary measurements, i.e., w1, w2, etc.

Suppose a set of data is collected in the variables x1, x2,....,xn and a result calculated. At the
same time one may perturb the variables by ∆x1, ∆x2, and so on, and calculate new results. We
would have
R(x1) = R(x1, x2,....,xn )

R(x1 + ∆x1) = R(x1, + ∆x1, x2, ..., xn )

28
 R(x2) = R(x1, x2,...,xn )

R(x2 + ∆x2) = R(x1, x2 + ∆x2,...,xn )

For small enough values of ∆x the partial derivatives can be well approximated by the finite difference
expressions

∂R R(x1 + ∆x1 ) - R(x1 )

≈
∂x1 ∆x1

∂R R(x 2 + ∆x 2 ) - R(x2 )
≈
∂x2 ∆x 2

and these values could be inserted in Eq. (3.2) to calculate the uncertainty in the result.

At this point we must again alert the reader to the ways uncertainties or errors of instruments are
normally specified. Suppose a pressure gage is available and the manufacturer states that it is
accurate within ± 1.0 percent. This statement normally refers to percent of full scale. So a gage with
a range of 0 to 100 kPa would have an uncertainty of ± 10 percent when reading a pressure of only
10 kPa. Of course, this means that the uncertainty in the calculated result, either as an absolute value
or percentage, can vary widely depending on the range of operation of instruments used to make the
primary measurements. The above procedure can be used to advantage in complicated data-reduction
schemes.

A very full description of this technique and many other considerations of uncertainty analysis are
given by Moffat [4]. An example of an industry standard on uncertainty analysis is given in Ref. [5].

Example 3.5. Calculate the uncertainty of the wire resistance in Example 3.1 using the technique
described in this section.

Solution. In Example 3.1 we have already calculated the nominal resistance at 6.24? . We now
perturb the three variables Ro, α, and T by small amounts to evaluate the partial derivatives. We shall
take
∆Ro = 0.01 ∆α = 1 x 10-5 ∆T = 0.1
Then
R(Ro + ∆Ro) = (6.01)[1 + (0.004)(30-20)] = 6.2504

and the derivative is approximated as

∂R R(Ro + ∆Ro )- R 6.2504 - 6.24

≈ = = 1.04
∂Ro ∆Ro 0.01

29
or the same result as in Example 3.1. Similarly,

R(α + ∆α) = (6.01)[1 + (0.00401)(30-20)] = 6.2406

∂R R(α + ∆α )- R 6.2406 - 6.24

≈ = = 60
∂α ∆α 1 x 10-5

R(T + ∆T) = (6)[1 + (0.004)(30.1-20)] = 6.2424

∂R R(T + ∆T )- R 6.2424 - 6.24

≈ = = 0.24
∂T ∆T 0.1

All the derivatives are the same as in Example 3.1 so the uncertainty in R would be the same, or
0.0305? .

3.6 GRAPHICAL ANALYSIS AND CURVE FITTING

Successful analysis of experimental data requires good understanding of the physical processes behind
the data. Unless thought through carefully, curve-plotting and cross-plotting usually generate an
excess of displays, which are confusing not only to the management or supervisory personnel who
must pass on the experiments, but sometimes even to the experimenter.

Assuming that the engineer knows what is to be examined with graphical presentations, the plots may
be carefully prepared and checked against appropriate theories. Frequently, a correlation of the
experimental data is desired in terms of analytical expression between variables that were measured in
the experiment; the easiest to plot and understand is a linear relationship. It is most convenient, then,
to try to plot the data in such a linear form, which could sometimes be accomplished by a coordinate
transformation.

Table 3.1 summarizes several different types of functions and transformations that may be used to
produce straight lines on graph paper. The graphical measurements, which may be made to determine
the various constants, are also shown. It may be remarked that the method of least squares may be
applied to all these relations to obtain the best straight line to fit the experimental data. A number of
computer software packages are available to accomplish the functional plots illustrated in Table 3.1.
See, for example, Refs.. [6], [7], and [8].

Note that when using logarithmic or semilog graph paper is is unnecessary to make log calculations;
the scaling of the paper automatically accomplishes this.

3.7 GENERAL CONSIDERATIONS IN DATA ANALYSIS

Our discussions in this chapter have considered a variety of topics: statistical analysis, uncertainty
analysis, curve plotting, least squares, etc. With these tools the reader is equipped to handle a variety
of circumstances that may occur in experimental investigations. As a summary to this chapter let us

30
now give an approximate outline of the manner in which one would go about analyzing a set of
experimental data.

1. Examine the data for consistency. No matter how hard one tries, there will always be some
data points that appear to be grossly in error. If we add heat to a container of water, the
temperature must rise, and so if a particular data point indicates a drop in temperature for a
heat input, that point might be eliminated. In other words, the data should follow consistency
with laws of nature, and points that do not appear proper in that way should be eliminated. If
very many data points fall in the category of "inconsistent," perhaps the entire experimental
procedure should be investigated for gross mistakes or miscalculation.

2. Perform a statistical analysis of data where appropriate. A statistical analysis is only

appropriate when measurements are repeated several times. If this is the case, make estimates
of such parameters as standard deviation, etc.

3. Estimate the uncertainties in the results. We have discussed uncertainties at length.

Hopefully, these calculations will have been performed in advance and the investigator will
already know the influence of different variables by the time the final results are obtained.

4. Anticipate the results from theory. Before trying to obtain correlations of the experimental
data, the investigator should carefully review the theory appropriate to the subject and try to
glean some information that will indicate the trends the results may take. Important
dimensionless groups, pertinent functional relations, and other information may lead to a
fruitful interpretation of the data.

5. Validate the data. The experimental investigator should make sense of the data in terms of
physical theories or on the basis of previous experimental work in the field. Certainly, the
results of the experiments should be analyzed to show how they conform to or differ from
previous investigations or standards that may be employed for such measurements.

6. Correlate the data. Develop the mathematical relationship between the parameter of interest
and the independently measured variables which define it. For example, the equation Nu =
cRe0.8Pr0.4 is the mathematical relationship between the Nusselt number and the Reynolds and
Prandth numbers which are the independent variables sufficient to determine it.

3.8 SUMMARY

By now the reader will have sensed the central theme of this chapter as that of uncertainty analysis
and the use of this analysis to influence experiment design, instrument selection, and evaluation of the
results of experiments. At this point we must reiterate statements we have made before. We still
must recognize that uncertainty is not the same as error, even though some people interchange the
terms. The determination of "error" is eventually related to a comparison with a standard. Even then,
there is still "uncertainty" in the error because the "standard" has its own uncertainty.

31
Table 3.1
Methods of plotting various functions to obtain straight lines

32
Table 3.1
Methods of plotting various functions to obtain straight lines (Continued)

33
In the chapters which follow we shall examine a large number of instruments and measurement
devices and will see how the concepts of error, uncertainty, and calibration apply to each.

REVIEW QUESTIONS (Answers require also the reading of the references)

3.1 How does an error differ from an uncertainty?

3.2 What is a fixed error; random error?
3.3 Define standard deviation and variance.
3.4 In the normal error distribution, what does P(x) represent?
3.5 What is meant by measure of precision?
3.6 What is Chauvenet's criterion and how is it applied?
3.7 What are some purposes of uncertainty analyses?
3.8 Why is an uncertainty analysis important in the preliminary stages of experiment planning?
3.9 How can an uncertainty analysis help to reduce overall experimental uncertainty?
3.10 What is meant by standard deviation of the mean?
3.11 What is a least-squares analysis?
3.12 What is the correlation coefficient?
3.13 What is meant by a regression analysis?

PROBLEMS

3.1 The resistance of a resistor is measured 10 times, and the values determined are 100.0, 100.9,
99.3, 99.9, 100.1, 100.2, 99.9, 100.1, 100.0, and 100.5. Calculate the uncertainty in the
resistance.
3.2 A certain resistor draws 110.2 V and 5.3 A. The uncertainties in the measurements are ± 0.2
V and ± 0.06 A, respectively. Calculate the power dissipated in the resistor and the
uncertainty in the power.
3.3 A small plot of land has measured dimensions of 50.0 by 150.0 ft. The uncertainty in the 50-ft
dimension is ± 0.01 ft. Calculate the uncertainty with which the 150-ft dimension must be
measured to ensure that the total uncertainty in the area is not greater than 150 percent of that
value it would have if the 150-ft dimension were exact.
3.4 Two resistors R1 and R2 are connected in series and parallel. The values of the resistance are

R1 = 100.0 ± 0.1?
R2 = 50.0 ± 0.03?

Calculate the uncertainty in the combined resistance for both the series and the parallel
arrangements.
3.5 A resistance arrangement of 50? is desired. Two resistances of 100.0 ± 0.1? and two
resistances of 25.0 ± 0.02? are available. Which should be used, a series arrangement with the
25? resistors or a parallel arrangement with the 100-? resistors? Calculate the uncertainty
for each arrangement.
3.6 The following data are taken from a certain heat-transfer test. The expected correlation
equation is y=axb. Plot the data in an appropriate manner, and use the method of least
squares to obtain the best correlation.

34
x 2040 2580 2980 3220 3870 1690 2130 2420 2900 3310 1020

Calculate the mean deviation of these data from the best correlation.
3.7 A horseshoes player stands 30 ft from the target. The results of the tosses are

Deviation from Deviation from

Toss target, ft Toss target, ft
1 0 6 +2.4
2 +3 7 -2.6
3 -4.2 8 +3.5
4 0 9 +2.7
5 +1.5 10 0

On the basis of these data would you say that this is a good player or a poor player? What
advice would you give this player in regard to improving at the game?
3.8 Calculate the probability of drawing a full house (three of a kind and two of a kind) in the first
5 cards from a 52-card deck.
3.9 Calculate the probability of filling an inside straight with one draw from the remaining 48
cards of a 52-card deck.
3.10 A voltmeter is used to measure a known voltage of 100 V. Forty percent of the readings are
within 0.5V of the true value. Estimate the standard deviation for the meter. What is the
probability of an error of 0.75V?

3.11 In a certain mathematics course the instructor informs the class that grades will be distributed
according to the following scale provided that the average class score is 75:

Grade A B C D F
Score 90-100 80-90 70-80 60-70 below 60

Estimate the percentage distribution of grades for 5, 10, and 15 percent failing. Assume that
there are just as many A's as F's.
3.12 For the following data points y is expected to be a quadratic function of x. Obtain this
quadratic function by means of a graphical plot and also by the method of least squares.

x 1 2 3 4 5
y 1.9 9.3 21.5 42.0 115.7

3.13 It is suspected that the rejection rate for a plastic-cup-molding machine is dependent on the
temperature at which the cups are molded. A series of short tests is conducted to examine
this hypothesis with the following results:

Temperature Total production Number rejected

T1 150 12

35
T2 75 8
T3 120 10
T4 200 13

On the basis of these data do you agree with the hypothesis?

3.14 A capacitor discharges through a resistor according to the relation E/Eo=e-t/RC where
Eo=voltage at time zero; R=resistance; C=capacitance. The value of the capacitance is to be
measured by recording the time necessary for the voltage to drop to a value E1. Assuming
that the resistance is known accurately, derive an expression for the percent uncertainty in the
capacitance as a function of the uncertainty in the measurements of E1 and t.
3.15 In heat-exchanger applications, a log mean temperature is defined by

∆Tm =
(T - T )- (T
h1 c1 h2 )
- Tc 2
ln [(T - T )/(T
h1 c1 h2 ]
- T )
c2

where the four temperatures are measured at appropriate inlet and outlet conditions for the
heat-exchanger fluids. Assuming that all four temperatures are measured with the same
absolute uncertainty wT, derive an expression for the percentage uncertainty in ∆Tm in terms
of the four temperatures and the value of wT. Recall that the percentage uncertainty is

w ∆ Tm
x 100
∆Tm

3.16 A certain length measurement is made with the following results:

Reading 1 2 3 4 5 6 7 8 9 10
x, in 49.36 50.12 48.98 49.24 49.26 50.56 49.18 49.89 49.33 49.39

Calculate the standard deviation, the mean reading, and the uncertainty. Apply Chauvenet's
criterion as needed.
3.17 A citizen's traffic committee decides to conduct its own survey and analysis of the influence of
drinking on car accidents. By some judicious estimates the committee determines that in their
community 30 percent of the drivers on a Saturday evening between 10 p.m. and 2 a.m. have
consumed some alcohol. During this same period there were 50 accidents, varying from
minor scratched fenders to fatalities. In these 50 accidents 50 of the drivers had had
something to drink (there are 100 drivers for 50 accidents). From these data what conclusions
do you draw about the influence of drinking on car accidents? Can you devise a better way to
perform this analysis?
3.18 The grades for a certain class fall in the following ranges:

Number 10 30 50 40 10 8
Score 90-100 80-90 70-80 60-70 50-60 Below 50

36
 The arithmetic mean grade is 68. Devise your own grade distribution for this class. Be sure
to establish the criteria for the distribution.

3.19 A certain length measurement is performed 100 times. The arithmetic mean reading is 6.823
ft, and the standard deviation is 0.01 ft. How many readings fall within (a) ± 0.005 ft, (b)
± 0.02 ft, (c) ± 0.05 ft, and (d) 0.001 ft of the mean value?
3.20 A series of calibration tests is conducted on a pressure gage. At a known pressure of 1000
psia, it is found that 30 percent of the readings are within 1 psia of the true value. At a known
pressure of 500 psia, 40 percent of the readings are within 1 psia. At a pressure of 200 psia,
45 percent of the readings are within 1 psia. What conclusions do you draw from these
readings? Can you estimate a standard deviation for the pressure gage?
3.21 Two resistors are connected in series and have the following values:

R1 = 10,000 ? ± 5% R2 = 1 M? ± 10%

Calculate the percent uncertainty for the series total resistance.

3.22 Two groups of secretaries operate under the same manager. Both groups have the same
number of people, use the same equipment, and turn out about the same amount of work.
During one maintenance period, group A had 10 service calls on the equipment while group B
had only 6 calls. From these data would you conclude that group A was harder on the
equipment?
3.23 A laboratory experiment is conducted to measure the viscosity of a certain oil. A series of
tests gives the values as 0.040, 0.041, 0.041, 0.042, 0.039, 0.040, 0.043, 0.041, and 0.039
ft2/s. Calculate the mean reading, the variance, and the standard deviation. Eliminate any
data points as necessary.
3.24 The following data are expected to follow a linear relation of the form y = ax + b. Obtain the
best linear relation in accordance with a least-squares analysis. Calculate the standard
deviation of the data from the predicted straight-line relation.

x 0.9 2.3 3.3 4.5 5.7 6.7

y 1.1 1.6 2.6 3.2 4.0 5.0

3.25 The following data points are expected to follow a functional variation of y = axb. Obtain the
values of a and b from a graphical analysis.

x 1.21 1.35 2.40 2.75 4.50 5.1 7.1 8.1

y 1.20 1.82 5.0 8.80 19.5 32.5 55.0 80.0

3.26 The following data points are expected to follow a functional variation of y = aebx. Obtain
the values of a and b from a graphical analysis.

x 0 0.43 1.25 1.40 2.60 2.9 4.3

y 9.4 7.1 5.35 4.20 2.60 1.95 1.15

37
3.27 The following heat-transfer data points are expected to follow a functional form of N = aRb.
Obtain the values of a and b from a graphical analysis and also by the method of least squares.

R 12 20 30 40 100 300 400 1000 3000

N 2 2.5 3 3.3 5.3 10 11 17 30

What is the average deviation of the points from the correlating relationship?
3.28 In a student laboratory experiment a measurement is made of certain resistance by different
students. The values obtained were

Calculate the standard deviation, the mean reading, and the uncertainty.
3.29 In a certain decade resistance box resistors are arranged so that four resistances may be
connected in series to obtain a desired result. The first selector uses 10 resistances of 1000,
2000,...,9000, the second uses 10 of 100, 200,...,900, the third uses 10 of 10, 20,...,90, and
the fourth, 1,2,...,9? . Thus the overall range is 0 to 9999 ? . If all the resistors have an
uncertainty of 1.0 percent, calculate the percent uncertainties for total resistances of 9, 56,
148, 1252, and 9999 ? .
3.30 Calculate the chances and probabilities that data following a normal-distribution curve will fall
within 0.2, 1.2, and 2.2 standard deviations of the mean value.
3.31 Suggest improvements in the measurement uncertainties for Example 3.4 which will result in
reduction in the overall uncertainty of flow measurement to 1.0 percent.
3.32 What uncertainty in the resistance for the first part of Example 3.2 is necessary to produce the
same uncertainty in power determination as results from the current and voltage
measurements?
3.33 Use the technique of Sec. 3.5 with Example 3.4.
3.34 Use the technique of Sec. 3.5 with Examples 3.3 and 3.2.
3.35 Obtain the correlation coefficient for Prob. 3.24.
3.36 Obtain the correlation coefficient for Prob. 3.25.
3.37 Obtain the correlation coefficient for Probs. 3.26 and 3.27.
3.38 Obtain the correlation coefficient for Probs. 3.6 and 3.12.
3.39 For the heat exchanger of Prob. 3.15 the temperatures are measured as Th 1 = 100°C, Th 2 =
80°C, Tc1 = 75°C, and Tc2 = 55°C. All temperatures have an uncertainty of ± 1°C. Calculate
the uncertainty in ∆Tm using the technique of Sec. 3.5.
3.40 A radar speed-measurement device for state police is said to have an uncertainty of ± 4
percent when directed straight at an oncoming vehicle. When directed at some angle θ from
the straight-on position, the device measures a component of the vehicle speed. The police
officer can only obtain a value for the angle θ˜ through a visual observation having an
uncertainty of ± 10°. Calculate the uncertainty of the speed measurement for θ values of 0,
10, 20, 30, and 45°. Use the techniques of both Secs. 3.4 and 3.5.

38
3.41 An automobile is to be tested for its acceleration performance and fuel economy. Plan this
project taking into account the measurements which must be performed and expected
uncertainties in these measurements. Assume that three different drivers will be used for the
tests. Make plans for the number of runs which will be used to reduce the data. Also prepare
a detailed outline with regard to form and content of the report which will be used to present
the results.
3.42 A thermocouple is used to measure the temperature of a known standard maintained at 100°C.
After converting the electrical signal to temperature the readings are: 101.1, 99.8, 99.9, 100.2,
100.5, 99.6, 100.9, 99.7, 100.1, and 100.3. Using whatever criteria seem appropriate, make
some statements about the calibration of the thermocouple.
3.43 Seven students are asked to make a measurement of the thickness of a steel block with a
micrometer. The actual thickness of the block is known very accurately as 2.000 cm. The
seven measurements are: 2.002, 2.001, 1.999, 1.997, 1.998, 2.003, and 2.003 cm. Comment
on these measurements using whatever criteria you think appropriate.
3.44 A collection of 120 rock aggregate samples is taken and the volumes measured for each. The
mean volume is 6.8cm3 and the standard deviation is 0.7cm3. How many rocks would you
expect to have volumes ranging from 6.5 to 7.2cm3?
3.45 Plot the equation y=5e1.2x on semilog paper. Arbitrarily assign fictitious data points on both
sides of the line so that the line appears by eye as a reasonable representation. Then, using
these points, perform a least-squares analysis to obtain the best fit to the points. What do you
conclude from this comparison?
3.46 The following data are presumed to follow the relation y=axb. Plot the values of x and y on
log-log graph paper and draw a straight line through the points. Subsequently, obtain the
values of a and b. Then determine the values of a and b by the method of least squares.
Compute the standard deviation for both cases. If a packaged computer routine for the least-
squares analysis is available, use it.

x y
4 105
5.3 155
11 320
21 580
30 1050
50 1900

3.47 The variables x and y are related by the quadratic equation

y = 2 - 0.3x + 0.01x2

for 0<x<2. Compute the percentage uncertainty in y for uncertainties in x of ± 1, 2, and 3

percent. Use both an analytical technique and the numerical technique discussed in Sec. 3.5.
3.48 Reynolds numbers for pipe flow may be expressed as

•
4md
Re =
µ

39
 •
where m is the mass flow in kg/s, d is pipe diameter in m, and µ is viscosity in kg/ m ⋅s . In a
certain system the flow rate is 12 lbm/min, ± 0.5 percent, through a 0.5-in diameter ( ± 0.005-
in) pipe. The viscosity is 4.64 x 10-4 lbm/hr ⋅ft , ± 1 percent. Calculate the value of the
Reynolds number and its uncertainty. Use both the analytical and numerical techniques.
3.49 The specific heat of a gas at constant volume is measured by determining the temperature rise
resulting from a known electrical heat input to a fixed mass and volume. Then

P = EI = mcν∆T = mcν(T2-T1)
where the mass is calculated from the ideal gas law and the volume, that is,

p1V
m =
RT1
Suppose the gas is air with R = 287 J/ kg⋅K and cν=0.714 kJ/ kg⋅ °C, and the measurements
are to be performed on a 1-liter volume (known accurately) starting at p1 = 150kPa and T1 =
30°C. Determine suitable power and temperature requirements, assign some uncertainties to
the measured variables, and estimate the uncertainty in the value of specific heat determined.
3.50 A model race car is placed on a tethered circular track having a diameter of 10 m ± 1cm. The
speed of the car is determined by measuring the time required for traveling each lap. A hand-
held stopwatch is used for the measurement, and the estimated uncertainty in both starting and
stopping the watch is ± 0.2 sec. For a nominal speed of 100 mi/hr calculate the uncertainty in
the speed measurement when made over 1, 2, 3, and 4 laps.

REFERENCES

1. Holman, J.P. and Gajada, W.J. Jr., "Experimental Methods for Engineers", 5th ed., McGraw
Hill, N.Y. 1989.
2. Wilson, E.B.: "An Introduction to Scientific Research," McGraw-Hill Book Company, New
York, 1952.
3. Kline, S.J., and F.A. McClintock: Describing Uncertainties in Single-Sample Experiments,
Mech. Eng., p. 3, January 1953.
4. Moffat, R.J.: Contributions to the Theory of Single-Sample Uncertainty Analysis, J. Fluid
Engr., vol. 104, p. 250, June 1982.
5. "Measurement Uncertainty," ANSI/ASME Power Test Code 19.1-1985, American Society
Mechanical Engineers, New York, 1986.
6. Johnson, A.T.: Microcomputer Programs for Instrumentation and Data Analysis, Int. Appl.
Engineering Education, vol. 3, no. 2, p. 149, 1987.
7. Canale, R.P., and S.C. Chapra: "Electronic Tool Kit," McGraw-Hill Book Company, 1988.
8. Chapra, S.C., and R.P. Canale: "Numerical Methods for Engineers," 2d Ed., McGraw-Hill
Book Company, 1988.

40
 EXPERIMENT No. 1.
LIFT AND DRAG ON 1/48 SCALE MODEL AIRCRAFT
APPARATUS

WIND TUNNEL: The AEROLAB Educational Wind Tunnel is of the Eiffel or Open Circuit type
with a 12”x12” test section. The airspeed is infinitely variable from 0 to 145 mph with a 10 hp drive
operated by a variable frequency system. The tunnel achieves a high energy ratio through its large
contraction ratio of 9.5:1, its small-angle diffuser, efficient fan system and low turbulence level of
0.13%. The fan operates in an acoustically treated section to reduce the noise level. The tunnel is
equipped with an honeycomb and two anti-turbulence screens. An orifice ring encircles the upstream
end of the test section to provide the average static pressure of this section. This provides the static
pressure of the undisturbed flow which is used as the reference pressure for pressures measured on
models when the pressures are non-dimensionalized as pressure coefficients. This static pressure is
also used as the throat Venturi pressure in the measurement of airspeed.

START UP PROCEDURE: Before operating the tunnel make sure that the model to be tested is
securely fastened and all loose objects have been removed from the test section.
1. Keep the FAN switch OFF and turn the fan speed control fully counterclockwise.
2. Turn on panel power by deflecting METER switch ON. All meters should light.
3. Turn on the FAN switch. Slowly turn fan speed control clockwise until motor begins to turn.
4. Check to see that the direction of rotation is correct. Air should be blowing out the exhaust and
the fan blades should be turning clockwise looking through the exhaust port. If the motor is
running backwards, contact your TA or the lab technician.

ANGLE OF ATTACT: Angle of Attack is sensed by a potentiometer connected to the lower arm
shaft of the parallelogram linkage. The zero adjustment is a second potentiometer on the front panel.
The calibration can be checked by leveling and zeroing the sting and then adjusting the sting a known
number of degrees and reading the new indication. DO NOT EVER position sting at an angle of
attack greater than +30°. This places great pressure on the cable and will usually shear the
very fine gage wires inside the cable.

ZERO AIRSPEED: Setting the airspeed reading to zero before starting a run is complicated by the
internal square root circuit which cannot take the square root of negative numbers. This requires that
zero be adjusted from an initially positive reading until it just reaches zero.

MODEL INSTALLATION: A socket in the model is provided to mount the model on the sting.
Before mounting a model make sure that the set screw is retracted. Push the model gently to the stop
and tighten set screw with moderate torque.
As in most large subsonic tunnels, the model is sting mounted for balance readings. The model
positioning system is of the parallelogram type which rotates the model in yaw or pitch about a point
near the end of the sting, keeping the model substantially in the center of the test section. A turntable
in the floor of the test section permits rotation in yaw up to the limits imposed by the side walls
(±30°).
For airplane models the airplane center of gravity (C.G.) is used as a reference center. The
measured moment is referred to the balance moment center. The moment must be transferred to a

41
parallel axis through the C.G. Usually the C.G. of the model is in the vicinity of the sting balance
center so the moment transfer is small.

STING BALANCE: The sting balance is designed with a 3/8” diameter hardened stainless steel sting
equipped with separate strain gauges for Normal Force and Pitching Moment. Axial Force is
measured by parallel flexures which are designed for both strength and sensitivity. Normal Force is
differentially wired. Pitching Moment is measured by a completely separate Wheatstone bridge.
Because of this wiring, each component is direct reading without resorting to arithmetical
calculations. The maximum load ranges are:
Normal Force: 25 lbs.
Axial Force: 10 lbs.
Pitching Moment: 50 inch-lbs.
To protect the balance, a stop is provided to counteract abnormally large forces in the axial
direction. The outer end of the balance is cylindrical to accommodate models with a straight hole and
a transverse set screw. The balance is designed to rotate 90 in its socket so that Normal Force
becomes Side Force and Pitch Moment becomes Yawing Moment. This feature is important for
measurements on a model in yaw.
To convert the Normal Force and Axial Force to Lift and Drag components the following
formulas should be used:
Drag = N sin α + Acosα where α = angle of attack
Lift = Ncosα − Asin α N = Normal Force
A = Axial Force
BALANCE CALIBRATION: The strain gauge balance is both the most fragile and the most
expansive component of the system. The wires leading from the balance are only a few thousandths of
an inch in diameter and therefore cannot stand much tension or rough handling.
For checking the calibration of the balance, the procedure is as follows:
1. Install the 3/4” diameter aluminum alloy calibration barrel on the sting balance by sliding it back to
the stop and tightening the set screw.
2. Level the calibration barrel. The angle of attack readout should be adjusted to 0°.
3. Adjust the sting to an angle of attack at 30°.
4. Zero the balance panel meters: Normal Forces, Axial Forces and Pitching Moment. Make sure the
roll-pin at the end of the sting balance support bears against the near end of the slot marked NF.
5. Hang the weight holder onto the calibration barrel, and record the reading for the Normal and
Axial Forces.
6. Add a known weight, one pound at a time, onto the weight holder, and record the Normal and
Axial Forces. From the recorded data obtain a calibration factor relates the applied forces to the
readouts.
7. For Pitching Moment calibration, move the known weight forward two grooves (i.e. 2” forward
of the balance moment center). The Normal Force readout should remain approximately constant.
(Not used in this experiment.)

PRESSURE SYSTEM: Tunnel pressures are connected to the 24-input selector valve to the left of
the control panel. The control knob rotates an internal valve, which successively connects each of the
ports to a Sen-Sym transducer that offers the optimum combination of high accuracy, stability and
freedom from hysteresis. The transducer is of the differential type so pressures may be sensed relative

42
to any desired pressure. The most useful reference pressure is the static pressure of the approaching
or undisturbed flow which is averaged by the static pressure ring at the upstream end of the test
section. The reference port is therefore ordinarily connected to the static ring. The pressure range is
±20” of water.

PRESSURE CALIBRATION: This wind tunnel offers a uniform low-turbulence airstream at speeds
up to 145 mph. Accurate data can be obtained provided that the airspeed, or more appropriately the
dynamic pressure, is measured correctly along with forces, moments and pressures on models.
Starting out with the Airspeed calibration, the procedure is as follows:
1. Zero Airspeed, see the instruction.
2. Disconnect tube running from test section static ring.
3. Connect one of the lines from the tee (connected to the manometer) to the line just disconnect
from the static ring.
4. Apply suction to the tube connected to tee and hold the pressure when 5” to 6” of vacuum is
registered.
5. Read Airspeed on display.
6. Obtain the correction factor. The reading is considered “correct” when the speed (in miles per
hour) corresponds to the vacuum reading on the manometer. The Airspeed is based on standard
air density of 0.002378 slugs/ft 3 . The dynamic pressure q can therefore be calculated as 12 ρairV 2
where V is in ft/sec. The density of water in the manometer is taken at 62.3 lbs/ft 3 corresponding
to a temperature of 20°C centigrade.
The pressure in the static pressure ring relative to atmospheric tunnel inlet pressure is usually termed
∆h. For a typical tunnel, q/∆h = 1.02 over the range of 100 to 150 mph. This value is factored into
the airspeed versus Inches of Water Curve.

The pressure readout system can be calibrated by temporary disconnecting the tubing connected
to the center of the 24-inlet selection valve and connecting it to one leg of the U-tube manometer
supplied. Positive or negative pressure can be applied. The procedure is the same as the airspeed
calibration.

INTRODUCTION

FORCE BALANCE: In straight and level steady flight, the forces on an aircraft are in balance. To
maintain flight at a fixed altitude, the lift must balance the weight of aircraft. To maintain a fixed
flight speed, the thrust must balance the drag, as indicated in the figure 1.

Figure 1. Force balance on an aircraft in flight.

We will measure here Lift and Drag as a function of angle of attack α. Then we will calculate lift
and drag coefficients

43
 L D
CL = , CD = 1
ρU S 2 ρU S
1 2 2
2
and plot CL and CD as functions of α. Here ρ = air density = .002327slug ft 3 , U = Flight speed, S =
wing area. Corrections in a wind tunnel must be made as described below.

WIND TUNNEL CORRECTIONS: An airplane flies in practically a limitless volume of air (except
when near the ground) whereas an airplane model “flies” in a wind tunnel test section in a volume of
air which is more confined. This difference gives rise to a number of corrections which are applied to
wind tunnel data. Many are quite small and in this experiment we will consider only the correction
arising from the streamlines being forced straight by the flat walls of the wind tunnel whereas they are
curved at corresponding distances from the full scale airplane, the so-called wind tunnel wall
correction. The correction depends on the shape and sweep of the wing as well as its span relative to
the test section width, but as an approximation we suggest the following additive corrections to angle
of attack and to the drag coefficient for any given value of CL:

∆α = 1.5C L (in degrees)

∆C D = .02C L2

FORCE RESOLUTION: The force balance measures forces along the sting axis (axial force A) and
perpendicular to the sting axis (normal force N). These are resolved into Lift (perpendicular to the
flight path U) and Drag (parallel to the flight path U) through the angle of attack α, the angle between
the flight path and the sting axis, as indicated in figure 2,
Drag = N sin α + Acosα where α = angle of attack
Lift = Ncosα − Asin α N = Normal Force
A = Axial Force

Figure 2. Force resolution.

PROCEDURE

1. Get familiar with the software HP VEE, especially the function on the data acquisition.
2. Calibrate the angle of attack according to the manual. Zero the angle of attack. You may run the
calibration software and follow the instruction on screen. The software will calibrate Angle of
Attack, Normal and Axial Forces, and Pressure.
3. Calibrate the strain gauge balance according to the instruction listed in the APPARATUS or
follow the instruction of the calibration software.
4. Install the aircraft model according to the instruction.

44
5. Zero the airspeed. Set the normal and axial force components equal to zero. At the six values of
angle of attack, note axial and normal force components. These are due to the model weight and
must be subtracted from the aerodynamic forces measured when the wind is on.
6. Measure normal and axial forces for two windspeeds U=100mi/hr, U=130mi/hr for each of the six
values of angle of attack: α = − 4o,0o ,4o ,8o,12o ,16o . Please note that at each angle of attack, the
Normal and Axial forces should be zeroed.
7. Load the measured data into Excel (automatically done in HP-VEE), graph and check your data
in Excel.
8. Calculate Lift and Drag (including the wind tunnel corrections) for each of the 12 cases.
Calculate CL, CD using the density of air given, the airspeed, and the measured wing surface area
S=12.87in2 for the model.
9. Plot CL and CD vs. α for the two airspeeds. Characterize the two airspeeds by their Reynolds
numbers Re = ρUc µ where c= wing root chord = 8.15 cm and µ=shear viscosity of air
= 3.78 ×10-7 slug/ft ⋅sec . Note that CL, CD, Re are completely dimensionless so unit conversions
must be made as necessary. Forces A and N are measured in lbs. and airspeed U in mi/hr.

QUESTION
Since the plots are dimensionless, we may apply the results to low speed flight of the real aircraft.
With a gross weight of 22,500 lb., calculate the minimum take off and landing speeds of this aircraft
based on the wind tunnel measurements.

45
 EXPERIMENT NO. 2.
FLUID FRICTION FLOW

OBJECTIVE:
To investigate the frictional losses across pipes.

DISCUSSION:
In this experiment, you will be measuring frictional losses across pipes of different diameters and
smoothness as a function of flow rate, looking at the effect of valves and bends in the flow stream,
and examining the performance of several types of flow meters, including a Venturi meter, an orifice
meter, and a Pitot tube. The apparatus allows one to measure the pressure drop across the various
parts of the flow stream using either a mercury or water manometer and to determine the flow rate.
Tabulated data for the viscosity and density of water at given temperature allow friction factors to be
determine and compared to standard charts.
A brief description of how to do the measurements is attached. While the actual experiments are
straightforward, you should plan your time carefully so that you can study one aspect of friction
losses or flow metering fully, rather than taking a collection of unrelated results. Examples of
experiments which you could perform include the following:

1) Determine the relationship between head losses due to friction (h) and flow rate (Q) in smooth
tubes. What relationship do you expect to see in the laminar and turbulent flow regimes? Plot
log(h) versus log(Q) to see if this is observed. Calculate friction factors as a function of
Reynolds number from your data and see if it agrees with tabulated data. A roughened pipe is
also available for determining the effect of roughness on head losses.

2) Measure the head losses due to friction for each of the various fittings and vales as a function
of flow rate. The fittings include a sudden contraction and expansion, a 45° elbow, a 90° elbow
and bend, a T-junction, a Y-junction and a strainer. For valves, losses can be measured as a
function of how "open" the valves is for a gate valve, a ball valve, and a globe valve.

3) Determine the performance of the differential head devices for measurement of flow rate and
water velocity in a pipe. Examine the use of the orifice meter, the Venturi meter, and the Pitot
tube. (The theory of theses devices is described in most fluid mechanics texts.)

46
NOTES ON OPERATION OF THE APPARATUS:

The schematic of the experiment is shown in Fig.1.

Figure 1. General Schematic of the Fluid Friction Experiment

Flow rates through the apparatus may be adjusted by operation of outlet control valve (V6).
Simultaneous operation of inlet flow control valve (V2) will permit adjustment of the static pressure
in the apparatus together with the flow rate. Operation of by-pass flow control (V1) will permit
reduction in static pressure when operating at low flow rates. Fine outlet control valve (V5) will
permit accurate control at very low flow rates. Suitable selection and operation of these control
valves will enable tests to be performed at different, independent combinations of flow rate and
system static pressure.

Measurement of Flow Rates using the Volumetric Tank

The service module incorporates a molded volumetric measuring tank (22) which is stepped to
accommodate low or high flow rates. A stilling baffle is incorporated to reduce turbulence. A remote
sight gauge, consisting of a sight tube and scale, is connected to a tapping in the base of the tank and
gives an instantaneous indication of water lever. The scale is divided into zones corresponding to the
volume above and below the step in the tank. A dump valve in the base of the volumetric tank is

47
operated by a remote actuator. In operation, the volumetric tank is emptied by lifting the dump valve,
allowing the entrained water to return the sump (23). When test conditions have stabilized, the dump
valve is lowed, entraining the water in the tank.
Timing are taken as the water level rises in the tank. Low flow rates are monitored on the lower
portion of the scale corresponding to the small volume beneath the step. Larger flow rates are
monitored on the upper scale corresponding to the main tank. Before operation, the position of the
scale relative to the tank should be adjusted.
When extremely small volumetric flow rates are to be measured, the measuring cylinder should be
used rather than volumetric tank. When using the measuring cylinder, division of the flow to and from
the cylinder should be synchronized as closely as possible with the starting and stopping of a watch.
Do not attempt to use a definite time of a definite volume.

Operation of the Self-Bleeding Manometers

Both the mercury and pressured water manometers installed on the apparatus are fitted with quick
connection test probes and self-bleeding pipework.
Each test point on the apparatus is fitted with a self-sealing connection. In operation, the dust cover is
unscrewed from the test point selected and the required test probe inserted into the exposed
receptacle. Screwing the collar of the test probe onto the test point opens the integral valve
permitting measurement of the system pressure via the pipe connected to the probe. Each test probe
is connected to the limb of a manometer via a vented ball valve which is situated over the volumetric
tank. In operation, the connecting valves are set to the 90° position and the test probes screwed onto
the required test points. Pressure in the test pipe, drives fluid along the flexible connecting pipe
pushing air bubbles to the valve where the mixture of air and water is ejected into the volumetric tank
via the vent in the body. In this condition the valve connection to the manometer remains sealed
keeping the manometer fully primed.
When all air bubbles have been expelled at the vent, the valve is turned through 90° to the live
position connecting the test point directly to the manometer.
Prior to removal of the test probe, the valve is returned to 90° position to prevent loss of water from
the manometer. Using this procedure, the manometers once primed will remain free from air bubbles
ensured accuracy in readings.
The pressured water manometer incorporated a Schrader valve which is connected to the top
manifold. This permits the level at high static pressures.

48
 EXPERIMENT No. 3.
PRESSURE DISTRIBUTION OVER AN AIRFOIL

OBJECTIVE:
To measure the pressure distribution over a Clark Y-14 (CY-14) Airfoil at
various angles of attack.

INTRODUCTION:
An airfoil develops Lift at a positive angle of attack through generally lower pressures above the
wing and higher below with respect to the pressure of the approaching air. The lift on the airfoil
increases as the angle of attack increases until a flow separation occurs. Then the lift decreases
suddenly. This phenomenon is called stall.
The overall pressure distribution can be measured with small tubes embedded in the wing leading
to a suitable pressure transducer. The airfoil model is equipped with 18 pressure openings. The
openings are located 0, 5, 10, 20, 30, 40, 50, 60 and 70% chord on both upper and lower surfaces
and there is an additional opening at 80% chord on the upper surface, see the attached data sheet.

Calculation of the Lift From the Measured Pressure Distribution

Assume the pressure p is only a function of s (or x), please refer to the Figure 1 below. The force
on a strip of width b and length ds is

dF = p ⋅b ⋅ds . (1)
This force acts normal to the surface of the airfoil.

The force dF makes an angle θ with the normal to the center line of the cross-section (see Fig.2).

Let the vertical component of this force be dV. Then

dV = cos(α − θ )⋅dF . (2)

Note that θ is a function of s but it is small over most of the airfoil. Also, dx = cosθ ⋅ds.. Thus, we
approximate

V•
dF

y
s
x

b
α

Figure 1. Force dF on an airfoil with angle of attack α.

49
 dV
dF α
θ
s
x

Figure 2. Decomposition of force dF into a vertical component dV.

dx ≅ ds
. (3)
dV ≅cosα ⋅dF = p ⋅b ⋅cosα ⋅ds

Thus, the resultant vertical force is approximately

C
V ≅ b ⋅cosα ⋅∫0 p(x )⋅dx . (4)
Let pl be the pressure distribution on the lower surface and let pu be the pressure distribution on the
upper surface. The pressure coefficients are defined by
p − p∞ p − p∞
Cp,l = l and Cp,u = u (5)
q∞ q∞
where p∞ is the freestream pressure and q∞ = 1 2 ρo V ∞2 is the dynamic pressure. The coefficient of lift is
defined by
V
CL = . (6)
q∞ ⋅b⋅c

Using these definitions, the coefficient of lift CL can be expressed

1 x x   x
CL = cosα ⋅∫ Cp,l  − Cp,u  ⋅d  . (7)
0  c  c   c 

The resultant moment of the pressure distribution about the leading edge of the airfoil at small angles
of attack is given by
M = b ⋅∫0 (pl − pu )⋅xdx .
C
(8)

The coefficient of moment is defined by

x  x
M
[ ]
= ∫ Cp,l − Cp,u ⋅ d  .
1
CM = (9)
q∞ ⋅b ⋅c
2 0 c   c 

The location of the center of pressure is

CM
xp = . (9)
CL

PROCEDURE:

50
1. Calibrate the pressure reading of the wind tunnel system, following the instruction listed in the
Experiment No.1 or in the calibration software.

2. Install the pressure wing vertically in the wind tunnel and connect the pressure tubes, in order, to
the inlet nipples of the tunnel pressure system. Make sure that the static pressure of the test
section (the upstream orifice ring) is attached to the reference connection of the pressure
transducer. Any unconnected input nipples will register the difference between atmospheric and
the test section pressure which can be taken as q∞ , the dynamic pressure of the airstream.
Dividing the pressure measured at any point on the airfoil by q∞ provides the pressure coefficient
at that point, (p − pref ) q = C p .
∞

3. Operate the wind tunnel at two airspeeds of 100 and 130 mph and make pressure measurements
at angles of attacks of 0°, 3°, 6°, 9°, 12°, 15° and with small increment up to as far as the wind
tunnel allows. Check to see if you have observed the stall of the airfoil at large angle of attack.

4. Load the measured data into Excel (automatically done in HP-VEE), graph and check your data
in Excel.

5. Plot the pressure coefficients along the chord line on both the lower and upper surface of the
airfoil.

6. Integrate the pressure coefficients to determine Lift and Drag coefficients. Note that this method
does not measure viscous drag forces.

51
 Data Sheet for Experiment 3
Date

Test Section Airspeed

PRESSURE DATA
Station % of α1= α2= α3= α4= α5= α6= α7=
Number Chord
0 0
1 5
2 10
3 20
4 30
5 40
6 50
7 60
8 70
9 80
10 5
11 10
12 20
13 30
14 40
15 50
16 60
17 70

52
Ordinates of Airfoils

53
Section Characteristics for Miscellaneous Airfoils

54
 COOLING TOWER EXPERIMENT
Although the laboratory experiment will be carried out by a team of three students, each
student must familiarize him/herself with the laboratory instructions and submit an assignment prior to
the beginning of the lab session.
This laboratory experiment consists of three parts:
1. Preparation before the lab:
(i) Read carefully the laboratory notes.
(ii) Answer the preliminary questions in writing and submit the answers to the TA before the
start of the lab session. The answers should be worked out and submitted by each student.
Failure to submit these homeworks may result in banning the student from the lab session.
2. Laboratory Experiment:
(i) Make sure that you understand what you are doing. If in doubt, ask.
(ii) Make sure to collect all the needed data.
3. After the Laboratory:
Prepare a laboratory report. The report writing is a team effort. Only one report will be
submitted by each team. The report should be written clearly since, in addition to technical content, it
will be graded for the quality of the writing.

1. INTRODUCTION
Cooling towers are widely used to dissipate heat to the environment. Typically, a condenser
of a power plant or of an air-conditioning/chilling system is cooled by water. The warm
water is transmitted to the cooling tower,
where it is sprayed from the top of the tower.
The falling water passes through a series of
baffles intended to keep the water as a thin film
to increase its surface area and to promote
evaporation. Atmospheric air is supplied at the
tower's bottom. The air rises counter to the
direction of the falling water. The air flow can
be driven by either buoyancy (natural draft) or
by means of a fan (forced draft). As the two
streams come into direct contact, a small
fraction of the liquid water evaporates into the
moist air which exits the tower at a greater
humidity ratio than that of the incoming air.
The energy needed for the latent heat of
evaporation is mostly provided by the liquid
water, with the net result that the liquid exiting
the tower is at a lower temperature than the Fig. 1: A bank of cooling towers
water entering the tower. Since some of the
water evaporates in the process, a small
amount of makeup water needs to be added to
the liquid stream.

55
OBJECTIVES:
The objectives of this experiment are:
1. To familiarize you with a piece of frequently used thermodynamic equipment.
2. To use the first law of thermodynamics in the analysis of thermodynamic hardware.

Note:
This experiment involves a few new concepts that are not taught in MEAM 203. These new concepts
are essential for carrying out the experiment and they are described, as much as needed, in section 2
of this laboratory manual.

2. BACKGROUND MATERIAL

2.1 SPECIFIC HUMIDITY

The amount of water vapor carried by the air is specified conveniently by the specific humidity
(w). The specific humidity is the ratio between the mass of the water-vapor (mv) and that of the dry-
air (ma):
•
m m
w = v = •v . (1)
ma m a
The specific humidity is a dimensionless quantity. Although in most thermodynamic calculations one
defines the fractional mass as the ratio between the mass of a particular species and the total mass of
the mixture, in air - vapor systems, it is more convenient to use the dry-air mass rather then the total
mass (ma+mv) as the normalizing quantity since the mass of the dry-air typically does not change
during the process.

2.2 STATE EQUATIONS FOR THE AIR-VAPOR MIXTURE

The dry- air (air without any water vapor content) is a mixture of a number of components.
Since in our application the air's composition does not change, we can think of the air as if it were a
pure substance with representative molecules having an average molecular mass of Ma=28.97
kg/kmol. This molecular mass was calculated as the weighted average of the molecular masses of the
various components of air such as O2, N2 and Ar.
A simple way of thinking of an air - vapor mixture is to consider the air and the vapor as if
they were existing all by themselves in a volume, V, at a uniform temperature, T, each at its own
(partial) pressure. The total pressure of the mixture (or the atmospheric pressure in our case), p0, is
the sum of the partial pressures of the air, pa, and the vapor, pv.
p0 = pa + pv . (2)

56
This model is known as the Dalton Model and it is valid when the gasses are ideal and there are no
intermolecular forces. When the Dalton Model is valid, we say that the mixture is ideal. Not all
mixtures are ideal. The ideal mixture approximation provides a reasonably accurate description of the
behavior of air-vapor mixtures.
pa pv p

Dry Air at Vapor at volume Air -Vapor mixture

volume V & V & partial pressure, at volume, V, and
partial pressure, p p total pressure,
a v p=p +p
a v
Fig. 2: The Dalton Model: the dry-air and the vapor can be thought of as each of them existed all by itself at its own
partial pressure and occupied the entire volume of the mixture.

To the first approximation, we assume that both the dry-air and the water-vapor behave like
ideal gases. When we studied the properties of water, we emphasized over and over again that
superheated vapor (or steam) cannot be assumed to behave like an ideal gas. In this particular
example, however, such an assumption can be made because the partial pressure of the vapor, pv, is
very small.
We write the law of ideal gases separately for the dry-air and for the vapor:
n RT n RT
pa = a and pv = v , (3)
V V
where na and nv are, respectively, the number of moles of dry-air and vapor; R=8.314 kJ/kmol-K is
the universal constant of ideal gases; and T is the mixture's temperature expressed in degrees Kelvin
(K). Witness that the mixture as a whole also obeys the law of ideal gases. By adding the state
equations for the dry-air and the water-vapor, we get the state equation for the mixture:
= (na + nv ) .
n RT nv RT RT
p = pa + pv = a + (4)
V V V
The ideal gas law also provides us with the means of determining the masses of the dry-air and
the water-vapor in a given volume, V, and temperature, T. For example, the mass of the air,
m a = na Ma . (5)
Consequently,
paV pV
m a = Ma and m v = Mv v . (6)
RT RT
Furthermore, we can express the specific humidity, w, in terms of the partial pressures,
m M  p p pv
w = v =  v  v = 0.622 v = 0.622 . (7)
ma  Ma  pa pa p − pv

57
 The air cannot carry infinite amounts of water-vapor. Once the partial pressure of the water-
vapor, pv, exceeds the water saturation pressure, psat, water will start to condense. The partial
pressure of the vapor-water cannot be larger than the saturation pressure, pv<psat. The saturation
pressure is a function of the temperature. Since psat increases as the temperature increases, the
capacity of the air to carry water-vapor increases as the temperature increases.

2.3 RELATIVE HUMIDITY

The quantity that determines our level of comfort and is reported daily by the weathermen is
the relative humidity (φ) which represents the ratio between the amount of vapor the air actually
carries, mv, and the maximum amount of vapor the air could possibly carry, mv,max, at the same
temperature, T. According to equation (6),
p V pvV
m v, max = Mv sat and m v = Mv . (7)
RT RT
Thus, the relative humidity,
mv p
φ= = v . (8)
mv, max psat
When the air carries water-vapor to its full capacity, we say that the air is saturated, and φ=1.

2.4 THE INTERNAL ENERGY AND ENTHALPY OF AIR-VAPOR MIXTURES

The internal energy and enthalpy of air vapor mixtures are obtained simply by summing up the
corresponding properties of the individual constituents. Since both the dry-air and the water-vapor
are assumed to behave like ideal gases, the internal energy and the enthalpy are functions of the
temperature alone,
H mix = Ha + H v = ma ha + m v hv (kJ ) . (9)
It is convenient to define the specific enthalpy of the mixture,
hmix = ha + whv (kJ / kg dry − air) . (10)
Witness that the normalization is done with respect to the mass of the dry-air rather than the total
mass of the mixture.
In processes when the temperature changes are not large, the change in the dry-air's enthalpy
can be calculated using the air's specific heat at a constant pressure, i.e.,
ha,2 − ha,1 = Cpa (T 2 − T1 ), (11)
where Cpa=1.0035kJ/kg-K. The enthalpy of the water-vapor can be read from the steam tables using
the saturation data at the given temperature, i.e., hv(T)=hv,sat(T) and hl (T)=hl,sat(T).

2.5 THE DETERMINATION OF THE HUMIDITY AND DRY & WET BULB
TEMPERATURES

58
 The air's humidity is measured by an instrument known as a hygrometer. One common way of
measuring the specific humidity is through the use of an adiabatic saturation process. Consider, for
example, the apparatus depicted in the Fig. 3.

1 2

Saturated air
Air with an
unknown liquid water
specific humidity, w 3
Fig. 3: An adiabatic saturation process

Air containing an unknown amount of water-vapor is brought into contact with a body of
liquid water until the air becomes fully saturated. The apparatus is assumed to be well insulated so
that there is no thermal interaction with the environment and the process is adiabatic. The
temperatures at locations 1, 2, and 3 are monitored. The temperature at the inlet, T1, is known as the
dry-bulb temperature. The temperature of the fully saturated air at the exit, T2, is the wet-bulb
temperature.
Once initial transients die out (i.e., the temperatures T1 and T2 no longer change as functions
of time), the process is a steady-state-steady-flow process. Mass conservation statements can be
written separately for the air and the water. The mass flow rate of the air is the same at the inlet and
exit,
• • •
m a,1 = ma, 2 = m a . (12)
When writing mass conservation for the water, one needs to account for the fact that the
amount of water-vapor may vary during the process. For example, in the adiabatic saturation
process, liquid water evaporates. The mass conservation statement for the water is:
• • •
m v,2 = mv ,1 + ml ,3 , (13)
•
where m is the amount of make-up water one needs to add to maintain a fixed level of water in the
l ,3

liquid reservoir. It is convenient to normalize this last equation by the mass flow rate of dry air,
•
ml ,3
w 2 = w1 + • . (14)
ma
Since the process is adiabatic and there is no shaft work, the first law of thermodynamics
states that
m a (ha,1 + w1 hv,1 )+ ml ,3 hl,3 = ma (ha,2 + w 2hv ,2 ).
• • •
(15)
Using equations (14) and (15), we obtain a formula for w1,

59
 w1 =
(h a,2 − ha,1 )+ w 2 (hv,2 − hl ,3 )
. (16)
(hv ,1 − hl,3 )
Thus, by measuring the temperatures at locations 1, 2, and 3, one can compute the specific humidity
w1. Usually, the temperature of the make-up water (point 3) is not measured and it is assumed to
equal the temperature T2. This introduces only a negligibly small error.
In practice, the wet-bulb temperature is measured by a thermometer whose bulb is enclosed by
a wick moistened with water. The dry-bulb temperature is measured by simply placing a thermometer
is the air-vapor mixture.

Example: The dry and wet bulb temperatures of air at 101.325kPa are 25C and 15C, respectively.
Determine the specific and relative humidities of the air.
In order to carry out the calculations, I'll use my Maple-based thermodynamic tables. You can
also obtain the data from the tables at the end of your thermodynamics book. First, I read the
thermodynamic tables into my worksheet.
> read `Macintosh HD:Thermo Tool Box:ASME_SteamTables.m`;
Second, I insert into Maple the given information:
> p_ambient:=0.101325*MPa: T1:=(25+273.15)*K: T2:=(15+273.15)*K:
Cp_air:=1.0035*kJ/kg/K:
Third, I will evaluate all the needed thermodynamic properties. For the purpose of the Maple
calculations, I do not really need to view explicitly the values of the various properties; but I'll do so
for the benefit of those of you who work with hard copy thermodynamic tables.
> psat1:=SaturationPressureWater(T1);
psat1 := .003165990008 MPa
> psat2:=SaturationPressureWater(T2);
psat2 := .001703925815 MPa

> hv1:=EnthalpySaturatedVaporWater(T1);
kJ
hv1 := 2547.284813 ----
kg

> hv2:=EnthalpySaturatedVaporWater(T2);

kJ
hv2 := 2529.054737 ----

60
 kg

> hl3:=EnthalpySaturatedLiquidWater(T2);

kJ
hl3 := 62.94098261 ----
kg
Given the fact that the exiting air (point 2 in Fig. 3) is saturated (φ2=1), I can readily calculate
the specific humidity of the exiting air (equation 7).
w2:=.622*psat2/(p_ambient-psat2);
w2 := .01063873147
Finally, I have all the information needed for the evaluation of the specific humidity of the air,
w1 (equation 16):
> w1:=(Cp_air*(T2-T1)+w2*(hv2-hl3))/(hv1-hl3);
w1 := .006521368663
and the relative humidity (φ1).
> ph1:=pv1/psat1;
ph1 := .3320670449

Since in the course of the experiment you will need to evaluate the specific humidity a large
number of times, it is convenient to define a Maple function for the evaluation of specific humidity.

> ws:=proc(T_DB,T_WB,p_ambient)
> local Cp_a, psat, w_out, hv2, hv1, hl;
> Cp_a:= 1.0035*kJ/kg/K;
> psat:=SaturationPressureWater(T_WB);
> w_out:=0.622*psat/(p_ambient-psat);
> hv2:= EnthalpySturatedVaporWater(T_WB);
> hv1:= EnthalpySturatedVaporWater(T_DB);
> hl:= EnthalpySturatedLiquidWater(T_WB);
> (Cp_a*(T_WB-T_DB)+w_out*(hv2-hl))/(hv1-hl);
> end:

Example:

61
> ws(298*K, 288*K,0.1*MPa);
.006559
Similarly, we can write a Maple procedure that would allow us to evaluate the relative
humidity.

> relative_humidity:=proc(specific_humidity,temp, p0)

> local pv;
> pv:=p0*specific_humidity/(specific_humidity+0.622);
> pv/SaturationPressureWater(temp);
> end:

3. THE COOLING TOWER EXPERIMENT

3.1 THE EXPERIMENTAL SET-UP

The experimental apparatus is depicted schematically in Fig. 4. The cooling tower
accommodates counterflow of air and water. The air flows upwards while the water drips
downwards and spreads on the baffle plates. In the process, some of the water evaporates. Primarily
due to the latent heat of evaporation, the water cools down.
Prior to its entering the cooling tower, the water is heated (to simulate the "heating load") in
the "load tank". Subsequently, the water is pumped through a control valve that allows for the
adjustment of the flow rate and through a flow meter to the top of the column. After its temperature
is measured, the water is distributed over the top packing deck. The water spreads over the baffles
and flows downwards as a thin film whose surface is exposed to the air.
The cooled water leaves the cooling tower and collects into the "load tank." Before re-
entering into the tank, the temperature of the exiting water is measured. Due to evaporation, the level
of the water in the "load tank" tends to fall. This causes the float-operated needle valve to open and
transfer water from the "make-up tank" into the load tank. Under steady-state conditions, the rate at
which the water leaves the "make-up tank" is approximately equal to the rate of evaporation of the
water in the tower.
• •
The pump's power consumption is approximately W = − 100W , where W represents the
pump work and one can assume that all this power is consumed by the water. The water's flow rate is
measured by a rotometer.

62
63
Atmospheric air enters the fan at a rate that is controlled by the intake damper setting. The fan
discharges into a distribution chamber and the air passes dry and wet bulb thermometers before it
enters the column. The air flows upwards in the column. On leaving the top of the column, the air
passes through a droplet arrester which traps most of the entrained droplets and returns them to the
column. The air is then discharged to the atmosphere via an air flow-rate measuring orifice and wet
and dry bulb thermometers.
The air mass flow rate can be obtained from the formula:
• ∆p
m a = 0.0137 , (17)
vb
where ∆p is the pressure difference (expressed in mm of H2O) between the atmosphere and the air
stream; and vb (m3/kg) is the specific volume of the water-air mixture leaving the top of the column
(cooling tower).

3.2 ANALYSIS
We assume that the cooling tower is adiabatic (i.e., it does not interact thermally with its
environment). When analyzing the cooling tower, one can consider two different control volumes. In
the first case, the control volume includes the water re-circulation conduits, the load tank, the pump,
and the heater (Fig. 5a) . In the second case, the control volume excludes the water recirculation unit
(Fig. 5b). The boundaries of the control volumes are denoted by dashed lines in Figs. 5a and 5b.
Points (A) and (B) denote, respectively, the inlet and exit of the air-vapor mixture. Points (C) and
(D) denote, respectively, the inlet and exit of the water. Point (E) denotes the inlet of the make-up
water.

64
 • •
ma ma
B B
C C

Make-up Make-up
Water Tank Water Tank

A A
• •
ma ma

E E

D
D
water
pump pump
Fig 5.a: One possibility of a control volume that Fig 5.b: Another possibility of a control volume that
includes the load tank, the heater, and the pump. excludes the load tank, the heater, and the pump.

We analyze first the control volume shown in Fig. 5a. Under steady-state conditions, mass
conservation of air and water requires that:
m•  = m•  = m• (18)
 a A  a B a

and
m•  − m•  = m• , (19)
 v  B  v A  l E
where subscripts a, v, and l denote, respectively, air, vapor, and liquid.
The water mass-conservation can be rewritten in terms of the specific humidity as:
•
m l, E
w B − wA = • . (20)
ma
Next, we use the first law of thermodynamics to write the energy balance.
• • • •
Q
{ − W
{ = ma ((ha + wh v )B − (ha + wh v )A )− m l hl , E
14 4 4 4 4 24 4 4 4 4 3
Heat load pump work
Energy gained by the air − vapor mixture , (21)

65
 For the control volume shown in Fig. 5b, the water mass conservation statement is:
m•  − m•  = m• − m•  , (22)
 v  B  v  A  l  C  l D
and the energy conservation (the first law) is:
•  •   •  
0 = ma ((ha + wh v )B − (ha + wh v )A )−  ml hl  − ml hl  
14 4 4 4 4 24 4 4 4 4 3    

D 
Energy gained by the air − vapor mixture 14 4 44C 24 4 4 43 , (23)
Energy lost by the water

In the experiment, we will measure the temperatures at locations A, B, C, D, and E. Knowledge of

these temperatures will allow us to obtain the enthalpies needed to compute the energy balances.

4. PREPARATORY WORK
The requirements of this section must be completed before the laboratory session and handed to the
TA at the start of the lab. The preparatory work is designed to assure that you are well familiar with
the processes occurring in the cooling tower. This will minimize the possibility of you collecting
insufficient data during the experiment.

66
4.1 Read carefully the laboratory handout and make sure DATA
The numerical subscripts correspond
that you understand the material well. to the numerical designations in the
experimental set-up.
4.2 Visit the experimental set-up, and identify the various Air Outlet, Dry Bulb 21.2
temperature
components of the cooling tower.
TB,DB=T0 (C)
4.3 Suppose that the data given in the table on the right was Air Outlet, Wet 19.9
Bulb temperature
collected in an experiment. Determine:
TB,WB=T1 (C)
4.3.1 the specific and relative humidity at the air inlet Air Inlet, Dry Bulb 22.7
temperature
(point A);
TB,DB=T2 (C)
4.3.2 the specific and relative humidity at the air exit Air Inlet, Wet Bulb 13.5
temperature
(point B);
TB,WB=T3 (C)
4.3.3 the mass flow rate of the dry-air; Water Outlet 19.9
temperature
TD=T4 (C)
4.3.4 the flow rate of the make-up water; and Water Make-up 20.2
temperature
TE=T5 (C)
4.3.5 the heating load of the system. Water Inlet 27.9
temperature
TC=T7 (C)
Air Orifice 15
Differential Pressure
in mm H20

5. EXPERIMENTAL PROCEDURE
Necessary equipment: Graduated Cylinder, Stop Watch, Syringe, Distilled Water
In this section, some of the experimental procedures are described. The description is not
exhaustive and you will need to exercise common sense.
5.1 Make sure that the upper side of the water manometer is connected to the outlet located beneath
the orifice and the other side is exposed to the atmosphere.
5.2 Turn on the fan and set the damper so that the air pressure drop will be around 15 mm H20.
5.3 Make sure that the wicks of the wet-bulb thermometers are saturated with distilled water. These
wicks tend to dry-up during the experiment. Make sure to re-check them from time to time and refill
the wells with distilled water as needed.
5.4 Measure the dry and wet bulb temperatures at the air inlet and exit.

67
5.5 Turn on the pump and, using the Flow Control Valve, adjust the water flow rate to 25-30 g/s.
Warning: Do not turn on the heaters in the absence of water circulation!
5.6 The first set of experiments will be conducted in the absence of any heat load.
5.7 Once you have turned on the water pump, monitor the wet-bulb temperature at the cooling tower
exit as a function of time, i.e., measure this wet-bulb temperature every 30s and record the time and
temperature readings.
5.8 Once the temperature does not change as a function of time and steady state conditions are
established (this should take about 10-15 minutes) , fill the make-up water container up to the mark
and start the stop-watch.
5.9 Periodically, re-examine the water flow rate and, if needed, re-adjust it so as to keep the water
flow rate as constant as possible throughout each experiment. You may change the water flow rate
when you change the heat-load.
5.10 Read the seven temperatures (set the temperature switch to 0, 1, 2, 3, 4, 5, and 7) and
document the readings in the enclosed worksheets together with all the other relevant data. Repeat
these measurements three times. Note: switch setting 6 is inactive.
5.11 Wait a sufficient amount of time until at least 100 ml of water have drained from the make-up
tank. This may take as long as 10 minutes (this time interval decreases as the heating load increases).
5.12 Use the graduated cylinder to refill the make-up tank up to the mark. Take note of the water
level in the graduated cylinder before and after the re-fill operation so that you can obtain the amount
of make-up water added. As soon as you have added water up to the mark, stop the stopwatch.
Monitor the amount of water added and the amount of time that have elapsed.
5.13 Set the heating load to 0.5kW.
5.14 Repeat steps 5.7-5.12.
5.15 Set the heating load to 1.0kW.
5.16 Repeat steps 5.7-5.12.
5.17 Set the heating load to 1.5kW.
5.18 Repeat steps 5.7-5.12.
5.19 At the conclusion of the experiment, make sure to first turn off the heating load. Then turn off
the liquid pump, the air fan, and the power to the entire experiment.

6. REPORT
The requirements listed here represent the bare minimum that should be included in the report.
You are encouraged to further elaborate upon and find new ways to analyze the data. You are also
welcome to conduct additional experiments.

68
6.1 In the absence of water flow, calculate the specific and relative humidities of the air entering and
exiting the cooling tower. Would you expect the relative and specific humidities at the inlet and exit
to be the same? Explain.
6.2 Depict the exit-air wet-bulb temperature as a function of time for every change in the heating
load. Determine the time-constant of the system.
6.3 For all the heating loads, calculate the specific humidity and the relative humidity at the air inlet
and exit. Using the specific humidities obtained from the dry and wet bulb thermometers, calculate
the amount of make-up water and compare it to the amount of make-up water which you measured
directly. Determine the percentage difference between the calculated and measured make-up rates.
Comment about the possible reasons for any discrepancies. Plot the mass flow rate of the make-up
• •
Q− W
water normalized with the mass flow rate of dry air as a function of • .
ma
6.4 For all the heating loads, calculate the amount of energy gained by the air-vapor mixture and
compare it to the amount of energy lost by the water and to the amount of energy supplied by the
heater. Comment on any possible sources of measurement errors.

REFERENCES
1. Bench Top Cooling Tower H890, P.A. Hilton Ltd, Experimental Operating & Maintenance
Manual. A copy of the manual is available in the MEAM laboratory. You may borrow the manual
before the experiment.
2. Moran, M., J., and Shapiro, H., N., Fundamentals of Engineering Thermodynamics, Chapter 12.
3. Sonntag, R. E., and Van Wylen, G., J., 1991, Introduction to Thermodynamics, Classical and
Statistical, John Wiley, Chapter 11.
3. COOLING TOWERS ON THE NET, See, for example,
http://www.texasonline.net/us/history.htm.

69
 COOLING TOWER OBSERVATION SHEET #1
Date: __________ Atmospheric Pressure: _____________

Heating Load: ____________

Test # 1 2 3 4
Air Outlet, Dry Bulb
Temperature, TB,DB=T0 (C)
Air Outlet, Wet Bulb
Temperature, TB,WB=T1 (C)
Air Inlet, Dry Bulb
Temperature, TB,DB=T2 (C)
Air Inlet, Wet Bulb
Temperature, TB,WB=T3 (C)
Water Outlet Temperature
TD=T4 (C)
Water Make-up Temperature
TE=T5 (C)
Water Inlet Temperature
TC=T7 (C)
Air Orifice Differential
Pressure in mm H20
Water Flow Rate, ml,C (g/s)

Make-up Quantity (ml)

Time Interval to Measure

Make-up water (s)

Comments: ______________________________________________________________

________________________________________________________________________

________________________________________________________________________

70
 COOLING TOWER OBSERVATION SHEET #2
Date: __________ Atmospheric Pressure: _____________

Heating Load: ____________

Test # 1 2 3 4
Air Outlet, Dry Bulb
Temperature, TB,DB=T0 (C)
Air Outlet, Wet Bulb
Temperature, TB,WB=T1 (C)
Air Inlet, Dry Bulb
Temperature, TB,DB=T2 (C)
Air Inlet, Wet Bulb
Temperature, TB,WB=T3 (C)
Water Outlet Temperature
TD=T4 (C)
Water Make-up Temperature
TE=T5 (C)
Water Inlet Temperature
TC=T7 (C)
Air Orifice Differential
Pressure in mm H20
Water Flow Rate, ml,C (g/s)

Make-up Quantity (ml)

Time Interval to Measure

Make-up water (s)

Comments: ______________________________________________________________

________________________________________________________________________

________________________________________________________________________

71
 COOLING TOWER OBSERVATION SHEET #3
Date: __________ Atmospheric Pressure: _____________

Heating Load: ____________

Test # 1 2 3 4
Air Outlet, Dry Bulb
Temperature, TB,DB=T0 (C)
Air Outlet, Wet Bulb
Temperature, TB,WB=T1 (C)
Air Inlet, Dry Bulb
Temperature, TB,DB=T2 (C)
Air Inlet, Wet Bulb
Temperature, TB,WB=T3 (C)
Water Outlet Temperature
TD=T4 (C)
Water Make-up Temperature
TE=T5 (C)
Water Inlet Temperature
TC=T7 (C)
Air Orifice Differential
Pressure in mm H20
Water Flow Rate, ml,C (g/s)

Make-up Quantity (ml)

Time Interval to Measure

Make-up water (s)

Comments: ______________________________________________________________

________________________________________________________________________

________________________________________________________________________

72
 COOLING TOWER OBSERVATION SHEET #4
Date: __________ Atmospheric Pressure: _____________

Heating Load: ____________

Test # 1 2 3 4
Air Outlet, Dry Bulb
Temperature, TB,DB=T0 (C)
Air Outlet, Wet Bulb
Temperature, TB,WB=T1 (C)
Air Inlet, Dry Bulb
Temperature, TB,DB=T2 (C)
Air Inlet, Wet Bulb
Temperature, TB,WB=T3 (C)
Water Outlet Temperature
TD=T4 (C)
Water Make-up Temperature
TE=T5 (C)
Water Inlet Temperature
TC=T7 (C)
Air Orifice Differential
Pressure in mm H20
Water Flow Rate, ml,C (g/s)

Make-up Quantity (ml)

Time Interval to Measure

Make-up water (s)

Comments: ______________________________________________________________

________________________________________________________________________

________________________________________________________________________

73
 COOLING TOWER EXPERIMENT - EVALUATION SHEET
At the conclusion of the experiment, please fill this evaluation sheet and submit it to Professor Haim H. Bau's mail-box
in the MEAM department. Verbal comments will also be appreciated. The objective of this evaluation sheet is to
enable us to improve this laboratory experiment for the benefit of future generations of mechanical engineering
students. If needed, please add comments on additional sheets. Thank you.

About Yourself (Please circle the appropriate)

1. Sophomore / Junior / Senior

2. Did you take? MEAM203 yes / no MEAM338 yes / no

3. Name (optional): _____________________ E-mail address (optional): __________________

Laboratory Handout:
4. Was the laboratory handout clear? _______________________________________________

5. Did you detect any errors? where? _______________________________________________

6. Was the background material sufficient? ___________________________________________

Preparatory Assignment
7. Did the preparatory assignment help clarify relevant concepts? __________________________

8. Did the assignment help you utilize the laboratory time more effectively? __________________

9. Should any additional material be included in the preparatory assignment?

______________________________________________________________________________

Laboratory Session
10. Did the laboratory experiment run without glitches? _________________________________

11. Could these glitches be avoided by appropriate instructions that were missing from the handout?

_____________________________________________________________________

12. Should changes/additions be made in the laboratory instructions? ______________________

______________________________________________________________________________

Report
13. Did you use Maple? ________________________

General
14. Was this laboratory experiment useful? __________________________________________

74