Chemical engineering education ( Journal Site )

Material Information

Chemical engineering education
Alternate Title:
Abbreviated Title:
Chem. eng. educ.
Physical Description:
v. : ill. ; 22-28 cm.
American Society for Engineering Education -- Chemical Engineering Division
Chemical Engineering Division, American Society for Engineering Education
Place of Publication:
Storrs, Conn
Publication Date:
annual[ former 1960-1961]


Subjects / Keywords:
Chemical engineering -- Study and teaching -- Periodicals   ( lcsh )
serial   ( sobekcm )
periodical   ( marcgt )


Chemical abstracts
Additional Physical Form:
Also issued online.
Dates or Sequential Designation:
1960-June 1964 ; v. 1, no. 1 (Oct. 1965)-
Numbering Peculiarities:
Publication suspended briefly: issue designated v. 1, no. 4 (June 1966) published Nov. 1967.
General Note:
Title from cover.
General Note:
Place of publication varies: Rochester, N.Y., 1965-1967; Gainesville, Fla., 1968-

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 01151209
lccn - 70013732
issn - 0009-2479
lcc - TP165 .C18
ddc - 660/.2/071
System ID:

Full Text

^^f^Carol Hall

^^Clemson University^


This guide is offered to aid authors in preparing manuscripts for Chemical Engineering Education (CEE), a quarterly journal
published by the Chemical Engineering Division of the American Society for Engineering Education (ASEE).
CEE publishes papers in the broad field of chemical engineering education. Papers generally describe a course, a laboratory, a
ChE department, a ChE educator, a ChE curriculum, research program, machine computation, special instructional programs, or
give views and opinions on various topics of interest to the profession.

Specific suggestions on preparing papers *
TITLE Use specific and informative titles. They should be as brief as possible, consistent with the need for defining the
subject area covered by the paper.

AUTHORSHIP Be consistent in authorship designation. Use first name, second initial, and surname. Give complete mailing
address of place where work was conducted. If current address is different, include it in a footnote on title page.

ABSTRACT: KEY WORDS Include an abstract of less than seventy-five words and a list (5 or less) of keywords

TEXT We request that manuscripts not exceed twelve double-spaced typewritten pages in length. Longer manuscripts may
be returned to the authors) for revision/shortening before being reviewed. Assume your reader is not a novice in the field.
Include only as much history as is needed to provide background for the particular material covered in your paper. Sectionalize
the article and insert brief appropriate headings.

TABLES Avoid tables and graphs which involve duplication or superfluous data. If you can use a graph, do not include a
table. If the reader needs the table, omit the graph. Substitute a few typical results for lengthy tables when practical. Avoid
computer printouts.

NOMENCLATURE Follow nomenclature style of Chemical Abstracts; avoid trivial names. If trade names are used, define
at point of first use. Trade names should carry an initial capital only, with no accompanying footnote. Use consistent units of
measurement and give dimensions for all terms. Write all equations and formulas clearly, and number important equations

ACKNOWLEDGMENT Include in acknowledgment only such credits as are essential.

LITERATURE CITED References should be numbered and listed on a separate sheet in the order occurring in the text.

COPY REQUIREMENTS Send two legible copies of the typed (double-spaced) manuscript on standard letter-size paper.
Submit original drawings (or clear prints) of graphs and diagrams on separate sheets of paper, and include clear glossy prints of
any photographs that will be used. Choose graph papers with blue cross-sectional lines; other colors interfere with good
reproduction. Label ordinates and abscissas of graphs along the axes and outside the graph proper. Figure captions and legends
will be set in type and need not be lettered on the drawings. Number all illustrations consecutively. Supply all captions and
legends typed on a separate page. State in cover letter if drawings or photographs are to be returned. Authors should also include
brief biographical sketches and recent photographs with the manuscript.

Send your manuscript to
Chemical Engineering Education, c/o Chemical Engineering Department
University of Florida, Gainesville, FL 32611-6005

Chemical Engineering Education
Department of Chemical Engineering
University of Florida Gainesville, FL 32611
PHONE and FAX: 352-392-0861
Web Page:

T. J. Anderson

Phillip C. Wankat

Carole Yocum

James O. Wilkes
University of Michigan
William J. Koros
University of Texas, Austin

E. Dendy Sloan, Jr.
Colorado School of Mines

Gary Poehlein
Georgia Institute of Technology
Klaus Timmerhaus
University of Colorado

Dianne Dorland
University of Minnesota, Duluth
Thomas F. Edgar
University of Texas at Austin
Richard M. Felder
North Carolina State University
Bruce A. Finlayson
University of Washington
H. Scott Fogler
University of Michigan
David F. Ollis
North Carolina State University
Angelo J. Perna
New Jersey Institute of Technology
Ronald W. Rousseau
Georgia Institute of Technology
Stanley I. Sandler
University of Delaware
Richard C. Seagrave
Iowa State University
M. Sami Selim
Colorado School of Mines
James E. Stice
University of Texas at Austin
Donald R. Woods
McMaster University

Summer 1999

Chemical Engineering Education

Volume 33

Number 3

Summer 1999

178 Clemson University, Ron Grant, Charlie Gooding

184 Carol Hall of North Carolina State University, Richard M. Felder

190 Introducing Students to Basic ChE Concepts: Four Simple Experiments,
Duncan M. Fraser
198 Integrating Process Safety into ChE Education and Research,
M.S. Mannan, A. Akgerman, R.G. Anthony, R. Darby, P.T. Eubank, K.R.
204 Experiments with Integration of Early Engineering Education,
Vincent G. Gomes, Timothy A.G. Langrish

196 Speaking of Education II, Richard M. Felder

210 Acetone Production from Isopropyl Alcohol: An Example
Debottlenecking Problem and Outcomes Assessment Tool,
Joseph A. Shaeiwitz, Richard Turton
222 The Effective Use of Logbooks in Undergraduate Classes,
Jennifer I. Brand
244 How to Involve Faculty in Effective Teaching, Francesc Giralt, Joan
Herrero, Magda Medir, Francesc X. Grau, Joan R. Alabart
250 Computer-Mediated, Collaborative Learning in ChE at the University of
Ottawa, David G. Taylor
254 A Software Package for Capital Cost Estimation,
P.T. Vasudevan, Deepak Agrawal

216 Sequential Batch Processing Experiment for First-Year ChE Students,
Ronald J. Willey, J. Anthony Wilson, Warren E. Jones, John H. Hills
226 Two Simple Experiments for the Fluid-Mechanics and Heat-Transfer
Laboratory Class, Manual A. Alves, Alexandra M.F.R. Pinto, Jodo R.F.
Guedes de Carvalho
232 Experiments on Viscosity of Aqueous Glycerol Solutions Using a Tank-
Tube Viscometer, Kyung Kwon, Sammaiah Pallerla, Sanjeev Roy

238 Rate Measurement with a Laboratory-Scale Tubular Reactor, Wei-Yin

> 189 Letter to the Editor

CHEMICAL ENGINEERING EDUCATION (ISSN 0009-2479: USPS 101900) is published quarterly by the Chemical
Engineering Division, American Society for Engineering Education, and is edited at the University of Florida. Annual
subscription costs $40.00. Correspondence regarding editorial matter, circulation, and changes of address should be sent to
CEE, Chemical Engineering Department, University of Florida, Gainesville, FL 32611-6005. Copyright 1999 by the
Chemical Engineering Division, American Society for Engineering Education. The statements and opinions expressed in
this periodical are those of the writers and not necessarily those of the ChE Division, ASEE. Defective copies replaced if
notified within 120 days of publication. Write for information on subscription costs andfor back copy costs and availability.
POSTMASTER: Send address changes to CEE, Chemical Engineering Department., University of Florida, Gainesville, FL
32611-6005. Periodicals Postage Paid at Gainesville, Florida.

4. department

Clemson University

Clemson University Clemson, SC 29634-0909

Clemson University, the land-grant in-
stitution of South Carolina, is the re-
alization of a long-held dream of its
founder, Thomas Green Clemson. A Pennsyl-
vania native, Clemson developed a profitable
career as a young consulting and mining engi-

neer in Paris, Philadelphia, and Washington. In 1838 he married the daughter
of South Carolina statesman John C. Calhoun, and over the next fifty years he
developed an abiding love for upstate South Carolina and an intense interest in
the application of scientific principles to improve agriculture. Clemson man-
aged Calhoun's Fort Hill plantation, wrote and published extensively on
agricultural chemistry, and eventually served as U.S. Superintendent of Agri-
cultural Affairs. He bought Fort Hill in 1866 after Calhoun's death and spent
the last years of his life developing plans to create a "high seminary of
learning to benefit the agricultural and mechanical arts." Outliving his wife
and children,Clemson left the bulk of his estate to South Carolina upon his
death in 1888, with specific instructions in his will leading to the establishment

i ssa |w fi i'i-di"-
IiJsrLL~i~j A-

Chemical Engineering Education

of Clemson College on the Fort Hill site. The Calhoun mansion re-
mains in the center of campus today-a historic landmark to Clemson's
vision. The small town that borders the campus also bears his name.
Chemical engineering was first introduced as a course of study at
Clemson in 1917. At that time there was no chemical engineering
department or faculty, and the core of the curriculum was drawn
from courses in mathematics, physics, chemistry, and mechanical
engineering. In the spring of 1923, four students completed the
prescribed course of study and became Clemson's first chemical
engineering graduates. The Department of Chemical Engineering
was formally established in 1946, and the undergraduate program has
been accredited by ABET since 1959. The Master of Science program
was begun in 1960, and the Doctor of Philosophy program was added
in 1962. In 1965 the department awarded the first PhD in engineering
in the State of South Carolina.

Earle Hall was donated to the University by the Olin Charitable
Trust in 1958 specifically to house chemical engineering. The 50,000-
square-foot facility contains four classrooms, an auditorium, a li-
brary, a student lounge, a seminar room, a shop, a 9,000-square-foot
unit operations lab, several dedicated research labs, and faculty,
graduate student and administrative offices,
Over the past forty years, Earle Hall has undergone many renova-
tions, the most recent being the conversion of the old auditorium into
a modern, 68-seat seminar and teaching facility, complete with mul-
timedia projection equipment and a network connection at each seat.
As the department has matured, it has benefited enormously from the
generous support of alumni and corporate benefactors. Several re-
search labs have been refurbished in recent years to accommodate
growing programs and new faculty additions, and the Dow Chemical
Company Unit Operations Lab is currently undergoing additional
equipment upgrades and modifications.
The faculty of the department is also undergoing a transition, with
four new additions in the last few years, all with youthful exuberance
and excellent credentials. David Bruce and Mike Kilbey joined the
department in 1995, Scott Husson arrived in 1998, and Graham
Harrison will be on board in August 1999, after completing a post-
doc at the University of Melbourne. The faculty now totals twelve
tenured and tenure-track faculty members involved full time in the
teaching and research programs of the department (see Table 1).
Another new colleague, currently in industry, will also join the
faculty later this year. In addition, Y. T. Shah is Senior Vice Provost
and Chief Research Officer of the University, Steve Melsheimer is
Associate Dean for Undergraduate Studies in the College of Engineer-
ing and Science, and Bill Beckwith is Director of the General Engineer-
ing Program. Professor Emeritus Joe Mullins also remains active in the
support of numerous teaching and research activities of the department.

Over the last decade the senior chemical engineering class at
Clemson has averaged forty-five students per year, and our recent
graduates have taken jobs with over 100 different companies. About
Summer 1999


Charles H. Barron Jr. (DSc, University of Virginia, 1963)
Polymer reaction engineering, analysis of the effects ofphysical
interactions on molecular weight distribution; formation vstivem
for design database development

David A. Bruce (PhD, Georgia Institute of Technology, 1994)
Catalyst development for the petrochemical and pharmaceutical
industries and for pollution abatement, chiral eolites, solid super-
acidt, supported metal comple.res

Dan D. Edie (PhD. Unt of Vuguin. 19721 Dow Chemical Professor
Director o the Centerbr fr.Adcanced Engieerinp Fibers and Films
Polhmcr processing, formanon and characieritanon of high
performance fibers and composite maernals, mathematiL al
modeling, rheology

Charles H. Gooding (PhD, North Carolina State University, 1979)
Department Chair
Mass transfer. panrtularly application and modeling of membrane
separation technologies

James M. Halle sPhD. Universin of Fonrda. 1976i
Molecular dynamics and thermod namics, the use of computer
lsmulanon techniques to determine thermodynamic and transport
properties influids

Graham M. Harrison (PhD, Univ. of California, Santa Barbara,
Non-Newtonian fluid mechanics, optical and mechanical
techniquesfor e spenmenial characitria:ton ool homers.
molecular-based constilurln equatincr

Douglas E. Hrt(PhD. Prncelon LUniersit). 1989)
Polymerfilms: extrusion addtive diffusion, interfacial phenom-
ena, mass transfer modeling, polymer thermodymanics

Scott Husson (PhD, University of California, Berkeley, 1998)
Bioseparations, reversible complexation in adsorption and
extraction, environmentally benignprocessing

S. Michael Kilbey H (PhD, University of Minnesota, 1996)
Equilibrium and dynamic behavior of molecularly thin films,
interface modification using amphiphilic molecules and elen trcallt
conductive polymers, surface forces measurements

Amod A. Ogale (PhD, University of Delaware, 1986)
Polymerprocessing; compositeformation, charoactrrzanon
micromechanics and modeling; stereolithography and rapid

Richard W. Rice (PhD, Yale University, 1972)
Kinetc i and catahlst s, heterogenou.s calrasis in petrochemical
and read rete reactions. catalyst charactenrcwaon, enm'ronmentaltv
related catalysis

Mark C. Thies WPhD. Lnivserstr of Delauare. 19851
Thermodynamics and supercritr, al fluids, separation proc e issi
materials processing, phase behavior .ni conmples rmuur i.
environmental applications


Above: Experiments in Dr. Thies' lab are designed to
produce phase equilibrium data, usually at high tem-
peratures and pressures.
Left: Professors Husson and Thies discuss an on-line
analytical scheme.

10 to 15% of our BS graduates elect to continue their engi-
neering education by pursuing graduate study, and one or
two each year choose law school or medical school to further
their education. To ensure that students with such diverse
interests and career aspirations are well prepared, the Bach-
elor of Science program in chemical engineering at Clemson
emphasizes broad, fundamental principles in science and
engineering rather than narrow specializations. Over half of
Clemson's chemical engineering undergraduates also gain
valuable experience, career insight, and financial assistance
by participating in the Cooperative Education Program, which
requires at least three semester-long work periods in industry.
The Clemson chemical engineering faculty has placed a
high priority on undergraduate instruction since the forma-
tive years of the department under the leadership of Charlie
Littlejohn. Traditional methods are honed and applied con-
scientiously, and innovations are continuously being devel-
oped and tested to better reach today's students. For ex-
ample, to improve the communication skills of students,
Doug Hirt uses journal writing in most of his undergraduate
classes. He and Charles Barron also developed the concept
of evolving design projects several years ago with support
from the NSF SUCCEED coalition. In the introductory sopho-
more chemical engineering course, a new process flow sheet
is introduced each year, and student teams are formed to
study and solve material and energy balance problems. The
same flow sheet follows the students through the curricu-
lum, with new aspects being investigated in each course,
such as pump specification in fluid flow and heat exchanger
layout in heat and mass transfer. In the first semester of the
capstone design sequence the evolving design project culmi-
nates with an economic analysis.
Doug Hirt is a frequent speaker and author on the subject

of both evolving design projects and effective writing as-
signments in engineering education. He was honored re-
cently by the Chemical Engineering Division of the ASEE
with the 1998 Ray W. Fahien Award, which recognizes
outstanding teaching effectiveness and educational scholar-
ship. Doug also chairs the Teaching Effectiveness Commit-
tee in the College of Engineering and Science. Jim Haile
was also recognized in 1998 by ASEE's Chemical Engineer-
ing Division, winning the Corcoran Award for his series of
papers, "Toward Technical Understanding," which appeared
in the Summer 1997, Fall 1997, and Winter 1998 issues of
Chemical Engineering Education. In these papers Jim inves-
tigated the fundamental question of what is meant by under-
standing of technical material. He is now immersed in the
application phase of that quest as a teacher, investigating
new ways to inspire, probe, and open the minds of under-
graduates as they attempt to grasp the concepts of chemical
The commitment to teaching excellence is also in good
hands with the most recent additions to the chemical engi-
neering faculty. At the May, 1999, commencement exer-
cises, Mike Kilbey received Clemson's top teaching honor
for the year, the Alumni Master Teacher Award. It is rare for
a professor in one of the university's smaller departments to
receive this student-determined award; even more notable,
Mike is the youngest recipient ever. Though they haven't
yet received such singular awards, David Bruce and Scott
Husson have also demonstrated their ability to establish excel-
lent rapport with students at all levels and to guide them to
understanding the intricacies of chemical engineering.
Undergraduate students who complete Clemson's chemi-
cal engineering program develop a sense of community and
pride that is nurtured by small class sizes (usually 20 to 30
Chemical Engineering Education

In its 110-year history, Clemson University has matured from a small agricultural
college to a nationally recognized, comprehensive university.

students), accessibility of the faculty, a heavy emphasis
on team projects, and the professional and social activi-
ties of the AIChE student chapter. Enrichment opportu-
nities include the Co-op Program and the Senior Depart-
mental Honors Program through which academically
qualified students can participate in research activities
with the faculty and graduate students. The two-semes-
ter senior seminar series also serves to complement the
classroom and laboratory experience and prepare stu-
dents for entry into the profession. In the fall semester
the seniors learn about career opportunities, resume prepa-
ration, interviewing, and early career success skills from
a series of external speakers, including recent graduates
of the department. In the spring they refine their speaking
skills further and learn from each other about business
aspects of the chemical industry, professional ethics, safety,
the environment, and other topics researched and pre-
sented in 40-minute team seminars.

The Department offers advanced study leading to the
Master of Science and Doctor of Philosophy degrees.
The MS degree requires completion of a research thesis
and 24 semester credit hours of graduate-level courses.
Four core courses are required: Advanced Transport Phe-
nomena, Chemical Engineering Thermodynamics.
Chemical Engineering Kinetics, and Separation Pro-
cesses. Of the remaining twelve credit hours of technical
electives, at least six must be in chemical engineering.
For graduate students interested in polymers, transport
phenomena, or process simulation, an exciting new series
of interdisciplinary courses covering chemistry, flow be-
havior, transport phenomena and visual simulation of poly-
mer flow and orientation is offered through Clemson's
Center for Advanced Engineering Fibers and Films.
The MS Industrial Residency Program is a special
arrangement involving Clemson University, a graduate
student, and a sponsoring company. Typically an MSIRP
student initiates a research project at the sponsoring
company's site during the summer, completes the re-
quired 24 hours of course work in the following fall and
spring semesters, and then returns to the company for
another seven months of full-time research. This pro-
gram is restricted to U.S. citizens. Graduate residency
students are paid at the prevailing salary level for BS
chemical engineers for the ten months on site, but the
salary is distributed over the 19-month program. This
arrangement results in a monthly stipend about 40%
Summer 1999

The undergraduate unit ops lab
emphasizes planning and execu-
tion of experiments, analysis of
results, technical communication,
and teamwork.

higher than normal graduate assistantships. The sponsoring com-
pany also pays tuition and fees, and the research is conducted
under joint faculty and industrial supervision and produces a thesis
or an equivalent formal report. The MSIRP is an excellent pro-
gram for students who want to gain industrial experience while
pursuing a graduate degree, whether their ultimate intent is to
continue with the PhD or to enter industry full-time after earning
an MS degree.
Students in Clemson's PhD program must complete at least 36
hours of approved course work beyond the BS, including the MS
course requirements (or equivalent courses taken elsewhere) and
at least 12 hours in fields other than chemical engineering. PhD
students plan their course work with the approval of a research
adviser and advisory committee to ensure general competence in
chemical engineering, a comprehensive knowledge of the field of
specialization, and a mastery of research methods. Each PhD stu-
dent must pass a written comprehensive exam based on under-
graduate and graduate course work, and an oral comprehensive
exam, which consists of the presentation and defense of a formal
research proposal. After a student's research is completed, the final
requirement for the PhD is the oral defense of the dissertation.

The research interests of the faculty are summarized in Table 1.
Strong, multifaceted programs exist in materials science and engi-
neering, particularly polymer studies, thermodynamics and sepa-

such as the
designation of
the Center for
Fibers and
Films as an
Center hold
great promise
for the future.

ration processes, and kinetics and ca-
talysis. These programs encompass most
of the traditional branches of chemical
engineering as well as newer areas, such
as advanced rheology, supercritical flu-
ids, molecular simulation, and biotech-
nology. Research interests of the faculty
range from purely theoretical topics to
the analysis and improvement of full-
scale industrial processes. Two Chemi-
cal Engineering faculty were recognized
this year for their research accomplish-
ments: Mark Thies received the
McQueen Quattlebaum Faculty Achieve-
ment Award from the College of Engi-
neering and Science, and Dan Edie re-
ceived the Alumni Award for Outstand-
ing Achievement in Research from
Clemson University.
Research opportunities in the depart-
ment increased exponentially in 1998
with the National Science Foundation's
recognition of Clemson's Center for Ad-
vanced Engineering Fibers and Films
(CAEFF) as a national Engineering Re-
search Center. This signal event is ex-
pected to bring more than $100 million
in research support to Clemson over the
next ten years. "This award does more
than establish us as a national research
institution," said Thomas M. Keinath,
Dean of Clemson's College of Engineer-
ing and Science. "It challenges Clemson
to be a leader in the nation's revolution in
engineering research and education. What
we do in the coming years will have a
profound effect on the fiber and film in-
dustry as well as the nation's next genera-
tion of engineers and scientists."

The Department of Chemical Engineer-
ing resides in Clemson's College of En-
gineering and Science (COES), the larg-
est of the University's five colleges.
Other engineering disciplines in the Col-
lege include Civil, Electrical and Com-
puter, Mechanical, Industrial, Ceramic
and Materials, Biosystems (formerly Ag-
ricultural), and graduate-only programs
in Bioengineering and Environmental En-
gineering and Science. A major reorga-
nization of the University in 1995 also

The Center for Advanced Engineering Fibers and Films

The Center for Advanced Engineering Fibers and Films is a National Science Foundation
Engineering Research Center that comprises a partnership between Clemson University and the
Massachusetts Institute of Technology. The Center provides an integrated research and educa-
tion environment for the systems-oriented study of fibers and films. It is the only NSF ERC in
the nation to deal exclusively with fibers and films, an industry that accounts for 25% of the
manufacturing segment of the U.S. gross domestic product. The industry's manufacturing base
includes electronic components, fiber optic cables, synthetic fibers, multi-layer food-packaging
films, and reinforced composites used in construction and aircraft. Products to be affected-in
some cases, reinvented-as a result of Clemson research can be found in fields as diverse as
biomedicine, transportation, communication, and construction.
Through CAEFF, faculty who are recognized for their expertise in key areas of engineering
and science are partnering with fiber and film manufacturers to study polymeric fibers and films.
These interdisciplinary teams are providing the knowledge base necessary to advance technol-
ogy in engineering fibers and films and supporting an educational program to produce highly
qualified professionals to lead this vital materials industry into the 21st century. Much of the
work involves faculty members from the Chemical Engineering Department. In addition to
Dan Edie, who directs CAEFF, Mark Thies, Amod Ogale, Doug Hirt, Mike Kilbey, David
Bruce, and Graham Harrison are key participants. Dr. Hirt leads one of the Center's three
major research thrusts. Drs. Thies, Ogale, and Kilbey head three of the Center's eight
primary research topics.
Center facilities include its centralized research/teaching testbed, comprised of integrative
fiber and film processing laboratories, on-line measurement instrumentation, a molecular mod-
eling laboratory, and a virtual reality laboratory. Center researchers have access to an impressive
battery of sophisticated instruments including FTIR, IR, UV, Raman and mass spectrometers;
gas, liquid, gel permeation, and supercritical fluid chromatographs; thermal analysis instru-
ments; x-ray analysis instruments, Instron capillary as well as Rheometrics and Haake rotational
rheometers; and a central microscope facility. The Center also has several devices for the
preparation of fiber and film precursors, small- and pilot-scale fiber and film extrusion equip-
ment, compression and injection molding equipment for the fabrication of composites, and
instruments for the physical testing of fiber, film and composite samples.
In CAEFF, chemical engineering undergraduate and graduate students join interdisciplinary
research teams that are developing advanced process models capable of predicting final fiber
and film properties. This work focuses on integrating molecular information into continuum
models. To verify those models, CAEFF faculty, students, and industry partners are conducting
an extensive experimental program for precursors ranging from conventional polymers, such as
nylon and polyester, to liquid crystalline materials. The models are ultimately converted to 3-D
visual process simulations. The goal is to create a new class of virtual process models that would
allow fiber and film producers to develop new and improved products rapidly and efficiently. By
designing materials at the molecular level, CAEFF is pioneering engineering technology for the
21st century.
In addition to undergraduate and graduate research programs, CAEFF provides short courses
for industrial personnel, sponsors conferences and workshops, and pre-college outreach pro-
grams to attract younger students to engineering and science disciplines.
The National Science Foundation has committed $12 million in support for the Center in its
first five years, with the total NSF funding anticipated at more than $20 million. In addition, the
State of South Carolina and the University have committed $1 million per year, and industrial
partners have already pledged more than $1 million per year to support the Center's research and
education programs. Partnering industries include 3M, Allied Signal, Arteva Specialties, BP
Amoco, Celanese Acetate, Collins and Aikman, Cryovac Division of Sealed Air Corp., Dow
Chemical, DuPont, Kemet, Raytheon STX, MSNW, N.H. Andreas, and Shell Chemical.

For more information on CAEFF, visit the Center web site at

Chemical Engineering Education

brought the Departments of Chemistry, Computer Science,
Geological Sciences, Mathematical Sciences, and Physics
and Astronomy into the college as well as the School of
Textiles, Fibers, and Polymer Science. Clemson also offers a
comprehensive General Engineering program designed ex-
clusively for freshman engineering students and students
who transfer from one of the state's two-year institutions.
This course of study provides a solid grounding in the funda-
mentals of engineering while the student explores the many
options available in the engineering field. Upon completion
of the freshman curriculum, students select their specific
engineering major.
South Carolina has been very successful in cultivating
international investment in business and manufacturing, with
more that 500 companies representing 27 countries now
located in the state. Numerous international firms, including
Michelin North America and BMW Manufacturing Corpo-
ration have both national headquarters and manufacturing
plants in South Carolina. Clemson University has formed
academic and business partnerships with many of these firms
and the countries they represent, creating study-abroad pro-
grams that give engineering and science students a strong
competitive advantage.
The Engineering Program for International Careers (EPIC)
prepares engineering students to be more competitive in the
international arena. Key features of this program include:
Foreign language courses, including a summer immersion
program, to provide competency in French, German, Japanese
or Spanish.
An International Internship to provide experience living and
working in a foreign culture.
EPIC graduates receive a certificate to document completion
of the program.
The College of Engineering and Science is also committed
to student support. The Programs for Education Enrichment
and Retention (PEER) was begun to help underrepresented
students in the College of Engineering and Science. PEER
students are assigned in groups to a first-year PEER mentor,
who is a junior, senior, or graduate minority student of the
COES. The mentor meets with the PEER group regularly to
share information. The PEER office also sponsors study
halls, counseling, seminars, and social events. The Program
for Educational Enrichment and Retention has helped make
Clemson's graduation rate of African American engineering
students the 5th highest in the nation.
As female enrollment in the COES has grown over the last
two decades, the College responded with WISE (Women in
Science and Engineering). An outgrowth of the PEER pro-
gram, WISE encourages women to persist in preparing for
and obtaining careers in science and engineering and to help
them be successful in those careers. Academic assistance,
including mentoring, advising, tutoring, and study groups,
as well as a special resource library, is sponsored through the
Summer 1999

WISE office. Wise is definitely having an impact. Although
women make up only 22% of the COES undergraduate student
body (30% in chemical engineering), they won over 60% of
the student awards presented this year.
Computers are essential to today's electronic modes and
methods of communication as well as technical calculations.
For the 1998-99 and 1999-00 school years, the College of
Engineering and Science is hosting a Pilot Laptop Program for
undergraduates at Clemson. Students can purchase high per-
formance laptop computers at a discount. Special laptop courses
are held in classrooms equipped with ethernet connections at
every desk. Courses are being offered in English, math, chem-
istry, computer science, physics, history, and engineering.

Thomas Green Clemson's dream has become the nucleus
of agricultural, scientific, and technological advancement in
South Carolina. Full-time enrollment at Clemson University
is now approximately 16,300 including 3,700 graduate stu-
dents. Clemson offers 73 undergraduate and 70 graduate
areas of study in its five academic colleges. The University
is accredited by the Southern Association of Colleges and
Schools to award the bachelor's, master's, specialist and
doctoral degrees, and appropriate curricula are accredited by
various professional organizations and associations.
Nestled in the foothills of the Blue Ridge Mountains on
the shores of Lake Hartwell, Clemson offers the amenities of
a small, southern town while providing big-city opportuni-
ties. The local environs provide unlimited, year-round op-
portunities for outdoor recreation, including whitewater raft-
ing on the Chattooga River and watching the Tigers play
nearly every sport known to mankind. The University com-
munity provides and hosts numerous cultural events, and
both Atlanta, Georgia, and Charlotte, North Carolina, are
just two hours away via 1-85. Clemson is 50 minutes from
the regional GSP airport and only a half-hour from Greenville,
South Carolina, which claims the greatest number of engi-
neers per capital in the United States.
In its 110-year history, Clemson University has matured
from a small agricultural college to a nationally recognized,
comprehensive university. Achievements such as the Na-
tional Science Foundation's designation of the Center for
Advanced Engineering Fibers and Films as an NSF Engi-
neering Research Center hold great promise for the future.
With innovative faculty, curricula, facilities, and programs
that respond to the needs of students, citizens, and industry,
the Department of Chemical Engineering will continue to
contribute toward Clemson University's goal of preparing
students for 21st century careers.
Additional information about Clemson University, the De-
partment of Chemical Engineering, and the Center for Ad-
vanced Engineering Fibers and Films may be found at http:/
/ O

p' educator

Carol Hall

of North Carolina State University

North Carolina State University Raleigh, NC 27695

If you walk into 119A Riddick Labs at North Carolina
State University, the odds are you'll see two people in
there, facing each other in adjacent chairs in front of the
desk. One is Carol Hall, whose office it is. The other could
be anyone-an undergraduate, a graduate student (who might
or might not be her advisee), a postdoc (ditto), a faculty
colleague, or a former advisee. The visitor could be there to
talk about some subtle point of statistical thermodynamics or
molecular simulation or how to get a better grade on the next
test or what to do about a personal problem or just to shoot the
breeze for a while. The topic doesn't matter-the visitor will
have Carol's undivided attention for as long as he or she is
there, and so will the next one and the one after that. You can't
help but wonder how she ever manages to get anything done.
But she manages quite well. Between conversations in the
past decade or two she has managed to get enough done to
become a widely recognized leader in the field of molecular
thermodynamics, a first-rate teacher, and the mother of three
remarkably talented children. She also makes the best stuffed
cabbage you ever tasted in your life.

Carol Klein Hall grew up in Brooklyn, the daughter of
two unusual parents. Her father was an attorney, business-
man, and politician. He ran for Brooklyn borough president
but lost to the machine candidate; he served as a New York
City Transit Commissioner and fought (unsuccessfully) to
keep the subway fare at 50; he organized a network of 150
trucking companies to help supply Israel in the 1948 war;
and he coordinated the famous JFK birthday party at Madi-
son Square Garden. Her mother stayed at home while Carol
and her younger brother Mitchell were growing up,
worked in an administrative position in the U.S. District

Court in the 1970s, and then went back to get her college
degree at age 63.
When Carol was 14, her parents and the parents of a boy
who lived nearby decided that she and Henry would make a
nice couple. He was definitely interested but she had other
fish to fry and nothing came of it. He was crushed, but
recovered enough to go on to a rather successful show busi-
ness career, and Carol and Henry (Winkler, aka "The Fonz")
have remained in touch and still enjoy each other's company
whenever they have a chance to get together.
Carol gravitated into a science career at about the same
time the first satellite gravitated into orbit. The national
anxiety over Russian scientific supremacy triggered by Sput-
nik turned New York City science teachers into evangelists:
if their students could handle algebra, they were encouraged
to become scientists. Carol handled algebra very well, and
when her high school physics teacher, Dr. Herman Gerwitz,
assured her and her skeptical mother that "women could be
physicists too," that was that, and off she went to become a
physics major at Cornell.
There were forty physics majors in Carol's entering class
at Cornell, six of them women. There were twelve left in her
graduating class, of whom six were the same women. She
does not speak fondly of her undergraduate educational ex-
perience, having found most of her professors remote and
seemingly indifferent to the success or failure of their stu-
dents. When asked how she accounts for the remarkable
success of her female classmates, she replies that they recog-
nized that the support they would need to survive would
have to come from themselves, so they networked with one
another to provide it, and it worked.
Carol met Tom Hall in a sophomore physics lab at Cornell,

Copyright ChE Division of ASEE 1999

Chemical Engineering Education

... in the past decade or two [Carol] has managed to get enough done
to become a widely recognized leader in the field of molecular
thermodynamics, a first-rate teacher, and the mother of
three remarkably talented children.

She also makes the best stuffed cabbage
you ever tasted
Carol at the in your
tender age life.
of four
(below) and
a few years
all grown
up and
in 1967.

1967 also saw Carol's marriage to Tom,
shown at the right with her proud
parents Harris and Celia Klein.
S fCarol's own family, shown below,
Sn \ p i includes (bottom row) Katie (25),
Norah (16), Carol;
(top row) Adam (21) and Tom.
where he was an engineering physics major. They ignored each other
as sophomores, noticed each other as juniors, got married as seniors,
and went to graduate school together in physics at the State University
of New York at Stony Brook. Carol began her doctoral program by
studying critical phenomena, and as part of that project wrote her first
computer program, foreshadowing the major role that computer simu-
lation has played in her subsequent research. After a year she learned
about the statistical mechanics research being carried out by a new
professor named George Stell. She thought it sounded interesting,
found out more about it, became intrigued by the idea that simply
by knowing the forces between two molecules one could under-
stand the collective behavior of 1023 of them, and switched advi-
sors to become Stell's first Ph.D. student. The roots of her growth
as a researcher were planted.
Stell was Carol's first true mentor. He communicated his passion
for research and intellectual inquiry to her-sometimes in his office
and sometimes at his kitchen table. He showed her how research is
done and at the same time gave her the freedom and encouragement to
Summer 1999

develop her own style of approaching problems. That free-
dom was exactly what she needed to come into her own as a
scientist. Until then she had been the typical underconfident
college woman of her generation, believing that her some-
times less-than-outstanding performance in the classroom
was irrefutable evidence of her lack of ability. Once she
began working with Stell, she began to truly believe that Dr.
Gerwitz was correct about the possibility of women being
physicists. Her dissertation research was a study of how
repulsive and attractive intermolecular forces compete to
determine the shape of the phase diagram. She
spent many happy hours attempting to "think
like a molecule" and constructing questionable
analogies between the behavior of systems of MUCh
molecules and of mixed-gender groups. career

Carol and Tom received their doctorates on
the same day and at the graduation gave joint
speeches about the challenges that confront
two-career families. As if to illustrate their
speeches, they each got job offers but none
from the same place, so they went back to
Cornell and Carol took an unpaid postdoctoral
position in the Chemistry Department, work-
ing in the world-renowned chemical physics
group led by Ben Widom and Michael Fisher.
In her work with Widom, she explained why
the critical-point scaling relations break down
in the ideal Bose gas at greater than four di-
mensions. She and a postdoctoral colleague
also developed lattice gas models that explained
the peculiar hourglass-shaped phase diagrams
observed experimentally for certain gas mix-
tures at high pressures. Her daughter Katie (a
gifted actress, singer, and dancer now working
in New York) was born during this period.

Three years later when Carol was ready to find a paid
position, the notoriously bad job market of the early 1970s
had commenced and there were few academic openings in
physics to be found. She and Tom considered themselves
fortunate to land jobs together at Bell Laboratories, working
on the applied technology side of the house. She worked in the
economic modeling department, combining her knowledge of
probability theory with her native New Yorker's instinct for
competitive shopping to produce an innovative model for pric-
ing multiple products that competed with one another.
Although Carol enjoyed her work at Bell, she wanted to
get back into real science. The head of the Chemical Engi-
neering Department at Princeton University at the time was
Leon Lapidus, a visionary who believed that women would
be valuable additions to the ranks of chemical engineering
faculties. Since there were only two female chemical engi-
neering professors in the entire country at the time and hardly
any in the graduate student pipeline, Lapidus decided to recruit

from closely related disciplines. He invited Carol to consider a
career change, a chance at which she jumped. Sadly, Lapidus
died suddenly a month before she came on board.
The move from physics to chemical engineering proved to
be a major cultural shift, and forced her to re-evaluate her
position on the continuum between "rigorous" and "close
enough." She was bemused, for example, by the notion that
the number of theoretical stages in a separation process often
depended on how sharp your pencil was. She eventually
negotiated the transition, however-she learned the jargon,
got the grants, performed widely recognized
research into the phase behavior of metal hy-
C W drides, and won the Rheinstein Outstanding
Young Faculty Award. She also became a
S been popular teacher, teaching courses she had
d to never taken. She discovered that chemical
ing the engineering was a veritable gold mine of in-
!es to tellectual challenges, offering many exciting
ced by research problems that were ideally suited to
Approaches grounded in statistical thermody-
; in namics. Her son Adam (now a student of art
I fields and history at Tulane) and daughter Norah (a
helping talented and-as Katie was-stage-struck
ercome high school student) were born during her
Princeton years.

In 1985 Carol became the latest in a dis-
tinguished line of New Yorkers who made
the transition from thinking "North Carolina-
isn't that south of Jersey somewhere" to join-
ing the Chemical Engineering Department at
North Carolina State University. At N.C. State
her career has flourished. She has made
ground-breaking advances in the modeling of
polymer structures, trained 21 graduate stu-

dents and 5 postdoctoral fellows, published over 125 papers
in the most prestigious chemical engineering and physics
journals in the world, and won the NCSU Alumni Associa-
tion Outstanding Research Award and the Alcoa Foundation
Distinguished Engineering Research Award.

Carol's research on metal hydrides at Princeton was mo-
tivated in part by the energy crises of the early 1970s. Sky-
rocketing gasoline prices had led many people to look to
hydrogen as a clean alternative fuel, and metal hydrides
were strong candidates as hydrogen storage media. Carol
performed Monte Carlo simulations on lattice models of
metal hydrides and successfully explained the unusual phase
behavior that accounted for the metals' ability to absorb vast
quantities of hydrogen. At the same time, a team composed
of Carol, Princeton colleague Bill Russell, and Ph.D. student
Alice Gast (now a professor at Stanford) used perturbation
theory to explain the phase behavior that underlies polymer-
Chemical Engineering Education


success f
and then
others ov
to help t
form sup,


ly tries

Princeton faculty and
graduate students in 1978
included a very pregnant
(8 months) Carol (first
row, next to last).
Interested readers may
also spot other familiar
(youthful) faces in the
first row (left to right):
Bob Bratzler; Mort Kostin;
Dick Toner; Ernie
Ron Andres, with Bill
Russel peeking from
behind him;
Carol, with Bob Axtman
behind her and Dudley
Saville behind Bob;
and Joe Calo, with Dave
Ollis behind him to the
right and Bob
Prud'homme behind him
to the left.

induced colloidal precipitation. Their research demonstrated
that such systems can coexist in three phases analogous to
gas, liquid, and solid, a phenomenon that had been observed
experimentally in the preceding year. While doing all that
and having two babies and co-running a home, Carol still
found time to collaborate with Gene Helfand of Bell Labora-
tories to develop a model for conformational state relaxation
in polymers. The result was the so-called Hall-Helfand cor-
relation function, which has become a standard against which
to compare results of NMR relaxation and time-resolved
fluorescence spectroscopy experiments.
Towards the end of her time at Princeton, Carol became
interested in fluids containing what she calls "chainlike"
molecules, a subject that has become her main area of inter-
est at North Carolina State. For the past half-century, the
state of the art in polymer modeling has been the classical
work of Flory and Huggins, which is based on a model of
chain molecules confined to a lattice. While revolutionary
for its time, the Flory-Huggins model has a number of short-
comings, not least of which is that the so-called Flory pa-
rameter-which the model presumes constant-turns out to
depend on variables such as composition, polymer chain
length, and temperature.
In the mid-1980s Carol and her coworkers identified in-
trinsic differences between chains moving on a lattice and
chains moving in continuous space as the principal limita-
tion in the predictive capability of the Flory-Huggins theory.
By blending the probabilistic reasoning underlying the origi-
nal Flory approach with powerful theories of liquid-state
Summer 1999

physics that had previously only been applied to monatomic
and diatomic fluids, they developed the "Generalized Flory
Theory," providing an elegant and deceptively simple ap-
proach to constructing equations of state for mixtures of
fluids and polymer molecules made up of segments of dif-
ferent size, shape, energetic, and angular constraints. The
theory is computationally convenient and physically intui-
tive, and has been found to be remarkably accurate in exten-
sive tests against computer simulation data. Carol's papers
in this area have stimulated a flurry of follow-on research in
the chemical engineering thermodynamics community.
Carol's contributions also include trailblazing simula-
tions of the dynamics of entangled polymer melts. Her mo-
lecular simulations are considered computational tours de
force, spanning almost six orders of magnitude in time.
They are aimed at reconciling the two competing theories of
polymer entanglements: the older view that polymer chains
entwine to form a so-called "local knot" and the more recent
view that chains "reptate" through a tube formed of sur-
rounding molecules. The simulations show that both types
of entanglements influence polymer dynamics, but that they
occur on different time scales.
Carol's latest venture is into the area of protein aggrega-
tion, a phenomenon associated with (and possibly a cause
of) many degenerative diseases such as Alzheimer's,
Huntington's, and "mad cow" disease. She and her students
recently completed the first computer simulation of the si-
multaneous aggregation and folding of model proteins. One
of their major findings is that a crowded protein environ-

ment can distort the folding pathway, making it more likely
that pathological partially folded intermediates will be stabi-
lized and will then aggregate. Such "misassembly processes"
are hypothesized to be the catalyzing events for fibril forma-
tion, the physical manifestation of Alzheimer's disease.
Carol's work has gained considerable attention throughout
the scientific community. As of the writing of this article,
her papers have been cited 2500 times, and five of them have
been cited more than 140 times each. Her accomplishments
and widespread recognition led to her designation as Alcoa
Professor of Chemical Engineering at N.C. State.

Carol enjoys a reputation as an excellent undergraduate
teacher. Recalling the impersonal nature of her undergradu-
ate instruction and the difficulties it caused many of the
students, she makes a point of learning every student's name,
no matter how large the class. In her writing and lecturing,
she takes great pains to make the complex seem simple and
yet manages to leave her students with a deep understanding
of the physical ideas that underlie the formulas and algo-
rithms she presents. Long before "Writing Across the Cur-
riculum" became a buzzword in educational circles, she
recognized that the exercise of writing frequently leads to
improved clarity and depth of thinking, and she has become
locally famous for requiring the students in her thermody-
namics courses to write essays on concepts that have baffled
generations of engineering students. Remarkably, she reads
and provides feedback on every essay, even though class
sizes at N.C. State sometimes approach 100.
The insights Carol has acquired in several decades of
teaching thermodynamics will be reflected in an undergradu-
ate textbook scheduled to be published in 2000 by Oxford
University Press. The book's aims are to introduce the idea
that thermodynamic properties are a direct reflection of mo-
lecular size and shape and the nature of the interactions
among neighboring molecules; to illustrate thermodynamics
concepts with up-to-date examples that go beyond the usual
ones involving light hydrocarbons; and to help students for-
mulate a general strategy for setting up and solving most
thermodynamics problems.
A variety of professional service activities augment Carol's
career dossier. She has organized numerous sessions at pro-
fessional society meetings, initiated a reception for women
chemical engineers that has become a fixture of the annual
AIChE meeting, and is currently the consulting editor for
thermodynamics of the AIChE Journal. She serves on the
Executive Board of the National Programming Committee
and on the AIChE ABET Accreditation Team, and is on the
editorial boards of Molecular Physics and the Journal of
Chemical and Engineering Data.
Mentorship is at the heart of Carol's professional activi-
ties. Those constant conversations with graduate students in

Carol's office are not just incidental to her job-she consid-
ers them the most important part of it. In putting together an
award nomination package, the Chemical Engineering De-
partment recently asked some current students and former
advises to reflect on the impact she had on their profes-
sional and personal lives. Their comments paint a remark-
able portrait of a truly gifted role model to young people.
For example, "Dr. Hall is a special person to many graduate
students, not just her own." "I can only hope that I will touch
people's lives the way that Dr. Hall has." "I believe she is
genuinely interested in what I am going to become, not just
in my research results." "I [and indeed, all students who
interact with her] owe a great deal to her active interest in
our intellectual and personal growth, which is supplemented
by her terrific sense of humor, patience, much wisdom, and
undoubted scientific expertise."
Former advisee Professor Alice Gast wrote "Everyone
should have a list of the most influential people in their lives.
Carol is near the top of my list." Dr. Kevin Honnell of Los
Alamos National Labs wrote "As we began to work to-
gether, what impressed me much more than the research
itself was the quality of the training Carol provides to her
students and the exceptional care and commitment she in-
vests in their intellectual, social, and emotional well-being."
Dr. Mauricio Futran, recently promoted to Executive Direc-
tor of Chemical Process Development at Bristol-Myers
Squibb, commented, "Her support, encouragement, and ad-
vice do not stop at graduation. It is typical for her students to
stay closely in touch and to rely on her feedback through the
years, both on professional and personal matters." George
Stell has a right to feel proud of his first PhD student and of
himself. Good mentoring bears rich fruit.
Attracting and retaining women has widely been recog-
nized as an important challenge facing science and engineer-
ing education. Much of Carol's career has been devoted to
overcoming the obstacles to success faced by women in
technical fields and then helping others overcome them. She
learned the value of networking and mutual support as an
undergraduate at Cornell, as one of a handful of female
graduate students in physics at Stony Brook, and as the only
female engineering faculty member at Princeton, and she
constantly tries to help others form supportive networks.
When speaking to women students in her office and at
seminars, Carol offers insights based on her experience. She
tells her listeners about the research showing a drop in confi-
dence experienced by many women in science and engineer-
ing, both in school and in the workplace. She encourages
them to join forces and provide one another with the kind of
mutual support that helped her survive professionally, and
she gives them a modified version of the Herman Gerwitz
message, assuring them that women can indeed succeed
in engineering. No one is better qualified to deliver this
message. 0
Chemical Engineering Education

SW -letter to the editor

Dear Editor:
Ten years ago, Sanders and Sommerfeld[11 published an
interesting article presenting a laboratory experiment on com-
bined mass transfer and kinetics. Specifically, the increase
of the pH in an aqueous solution of acetic acid was followed
with a digital pH-meter during neutralization with commer-
cial antacid tablets. This experiment was selected mainly
because its cost is very low and because it allows an interest-
ing process to be studied.
Since approximately 1989, we have implemented this ex-
periment in our undergraduate laboratory on mass transfer
and kinetics, following all the instructions, including the
experimental system, procedure, theory, and the data analy-
sis, as presented in the above-mentioned paper.
Nevertheless, a closer study of the discussion presented by
these authors revealed important inconsistencies that cannot
be obviated, and the argument of the authors that the model
provides a good fitting of the pH (though the constants
reported are not correct) involving a very simple data reduc-
tion procedure seems to be not acceptable.
The authors represent the system by a very simple overall

2 H30+ + CO3 3 H2O + CO2
and they assume that the instantaneous rate, measured as the
rate of disappearance of the hydronium ion, is proportional
to the remaining surface area, a, of the tablets and the hydro-
nium ion concentration, with the order of the latter as yet
Nevertheless, the fact that the material balances and the
equilibrium between the acetic acid and the acetate, as well
as the dissolved CO2 were not considered in their model,
would lead to the nonsense conclusion that the CaCO3 and
acetic acid are almost not consumed during the experiment
reported, which is, obviously, contrary to reality.
According to the previous considerations, we suggest to
our students the following model to explain the dissolution
(and neutralization) of antacid tablets:
1. Reactions involved:

2 H30 + CaCO3 ---CO2 T +3 H20 + Ca2+

H2CO3 +CaCO3 Ca2 + 2 HCO3
2. The pH of the solution and the acid concentration are the result
of the equilibrium of the non-reacted acetic acid, the acetate anion
(formed as a consequence of the reaction between the acetic acid
and the CaCO, of the tablet), and the carbonic acid, which is in
equilibrium with the carbon dioxide of the surrounding atmosphere
and the bicarbonate anion. The concentration of these species are
Summer 1999

related by the corresponding chemical equilibrium, and the acid
dissociation of the HCO3 can be neglected, since the pH of the
reaction medium is, during all the experiment, low enough so as to
assure the practical absence of CO3 anions. To calculate the
instantaneous ratio of the tablet, we use the same equation, shown
by Sander and Sommerfeld.
3. We represent the rate of dissolution by the rate of disappearance
of calcium carbonate (dNgdt), and obviously

1 dN I d[H30] n4
V dt 2 dt ka' IH k 2C03

-k1an' H30]n2 -k2an


k2=k 2[HCO;]

and can be considered as constant, since the concentration of
dissolved carbonic acid has been considered as constant.
In order to estimate the validity of the model, the previous
equation must be numerically integrated in combination with
the equations corresponding to the equilibrium of the acetic
and carbonic acids, which requires the knowledge of the
concentration of all the species involved at any time. To
solve this problem, we use the charge balance and the mate-
rial balance for the acetic acid and the acetate. We have
implemented several improvements to the model in order to
obtain a better fit, and only the consideration of a time delay
in the response of the pH meter produces a significant im-
provement of the fitting.
Nevertheless, the main objective of this letter is to bring
attention to the type of models in our undergraduate labora-
tories; excessive simplifications can only be appropriate when
the simplified model is able to explain adequately the whole
experiment. In this case, a very simple model was applied,
but was incapable of explaining the concentration variations
of the different chemical species with time. On the other
hand, the increasing knowledge of our students about differ-
ent computer tools permits application to more complex
models, with more complex equations, while keeping the
time for data reduction and analysis-thus allowing teachers
to emphasize the importance of developing better and more
complete models.
Antonio Marcilla, Maribel Beltrdn,
Amparo G6mez
Universidad de Alicante
Apdo. de correos n 99
03080 Alicante, Spain

1. Sanders, A.A., and J.T. Sommerfeld, "A Laboratory Experi-
ment on Combined Mass Transfer and Kinetics," Chem.
Eng. Ed., 23(2), 86 (1989) 0

M 1 curriculum



Four Simple Experiments

University of Cape Town Rondebosch, Western Cape,

his paper describes the part played by four simple
experiments in a new approach to introducing stu-
dents to chemical engineering. Instead of the tradi-
tional introduction through a course in material and energy
balances in the second year of study, a first-year course was
introduced in 1995 in which students are exposed to some of
the basic concepts in chemical engineering. This course was
part of a major revamping of our curriculum aimed at reduc-
ing the overload on students, facilitating the transfer of knowl-
edge from science to engineering, providing a better grasp of
physical phenomena, and improving the motivation of fresh-
men.[m The course runs for the full academic year, with half
of it being the introduction to chemical engineering (taught
by the author) and the other half modeling and computing
(taught by staff from the mathematics department). The two
halves run in parallel throughout the year.
The paper will describe the introduction to chemical engi-
neering part of the course, with particular reference to the
role played by the experiments, the objectives of the experi-
ments, how they were developed, implementation issues, an
evaluation against the objectives, how two of them are mod-
eled, and finally a brief evaluation of the experiments in the
context of the course.

The course starts developing the concepts needed as a
basis for the study of chemical engineering. After much
grappling to identify the essential core of what makes up
chemical engineering, I came to the conclusion that we
function at three different levels:
At the one extreme, the first level is the systems level, in
which overall structures and inter-relationships between
components of systems are considered.
At the other extreme, the third level is the micro level of the
fundamental processes occurring in the systems we work

7701 South Africa

with. Again, after reflection, it was clear that there are
essentially only four such fundamental processes, which
occur on their own or in combination depending on the
system: mass transfer, heat transfer, momentum transfer, and
reaction kinetics.
SIn between these two levels is the second level, in which we
design the equipment in which these fundamental processes
occur. Here we need to make use of empirical correlations
because it is not possible to predict exactly what will happen
In this course I therefore sought to cover all three levels
in a way that would help students develop these concepts
as far as possible, given that they are just starting their
university studies.
The first level is handled by dealing with the structure of
chemical processes and showing how this is implemented in
practice for some particular processes. The course starts
with the manufacture of ammonia as the first example. Two
visits to chemical plants (an ammonia factory and a marga-
rine/soap factory) help to consolidate this in addition to
exposing students to industrial equipment.
The students are given a number of designs to do that
expose them both to the design process and to the use of
correlations as part of that process, which caters to the sec-
ond level. The first design is a straightforward design of a
cake factory. The second design is the sizing of an absorber

Duncan Fraser has been lecturing at UCT
since 1979. He holds the degrees of BSc and
PhD, both from UCT. He has taught a wide
range of courses from first year to fourth year,
including mass and energy balances, thermo-
dynamics, transport phenomena, solid-fluid
operations, optimization, process control, and
design. His primary research interests are in
engineering education and process synthesis.

Copyright ChE Division of ASEE 1999

Chemical Engineering Education

that involves setting up three equations for its solution and
the calculation of diffusivities in both gas and liquid phases
using correlations. This problem is specifically given to help
students experience the sort of technical problems they will
face in chemical engineering. The third design is part of a
project on open-ended problem solving and deals with im-
proving the energy recovery system on a plant.
The third level receives the most extensive coverage. First,
at the start of the course, students are introduced to this level
by discussing what happens in particular sections of the
ammonia process (such as the catalytic ammonia reactor and
the carbon dioxide absorber) at a micro level. The process of
diffusion, which they know from physics, is used as a plat-
form to introduce the concept of mass transfer and also the
basic equation for transfer processes
(Transfer rate per unit area) = (driving force)/(resistance) (1)
They are subsequently introduced to heat transfer and,
toward the end of the course, briefly to momentum trans-
fer and the analogy between momentum, heat, and mass
The four experiments were designed to give the students
hands-on experience with the four fundamental phenomena.
They are run toward the end of the first semester. Subse-
quently, at the start of the second semester (when the stu-
dents have just learned differential equations in mathemat-
ics), the experiments are modeled using a shell-balance ap-
proach, and the solutions of the models are fitted to the
experimental data.
The course also aims to begin developing the basic skills
needed in chemical engineering. One of these is unit conver-
sion, which students must master to pass the course. Another
is modeling, which is covered in both parts of the course.
The modeling of the experiments is important in this
regard. It also has a key role in creating links with sci-
ence and helping students to transfer knowledge from
science to engineering.

In framing the objectives for these experiments, I was
guided by the principles of fun, simplicity, quickness, and
low cost as espoused by chemical engineering educators in
recent years.12-71 The objectives of the experiments were:
1. To introduce students to the four fundamental processes of
chemical engineering given above.
2. To provide hands-on exposure to these processes in a way
that would help subsequent development of theoretical
3. To move students from the known to the unknown, using
familiar equipment and concepts to introduce them to
unfamiliar engineering equipment and concepts.
4. To be fun to perform, giving students a sense of the
enjoyment of doing chemical engineering.
5. To be performed within a limited time by first-year
Summer 1999

engineering students who are not familiar with experimental
6. To be performed by the large number of students in the
course within afew weeks, so that all students have had
exposure to them by the time the modeling is to be under-
7. To be inherently safe, given the inexperience of the students,
and not use materials or procedures that could be harmful.
8. To not be too costly (multiple sets of apparatus being
9. To be easily assembled because of the short time available
for building the rigs and the pressure on the departmental
10. To use robust equipment, so as to withstand the treatment
likely to be meted out by inexperienced students.
11. To be readily stored away so they do not occupy laboratory
space for the large proportion of the year during which they
are not used.
12. To be easily transported, so they can readily be used by
other institutions within the Western Cape region in which
the University of Cape Town is situated.

There were 160 students in the course the first time it was
conducted-we decided to group them in pairs for the ex-
periments. Two afternoons of three hours duration were
available each week. This meant that, if each experiment
could be performed within an hour and a half, then two
experiments could be done each afternoon. In order for all
the pairs to be able to perform each of four experiments (one
for each of the fundamental processes), five sets of apparatus
were required so they could all be done in four weeks. This
then set the time limit for each experiment at one-and-a-half
hours. This meant that measurements had to be made on the
spot-lengthy analytical procedures were excluded.

The process by which each of the experiments was devel-
oped will be described in turn. This is to illustrate the seren-
dipitous nature of such a creative exercise and to encourage
others to try something similar.

Heat Transfer: Coffee Cup Cooling
This experiment arose out of a class discussion con-
cerning heat transfer and the effect of lowering the driv-
ing force for cooling a hot cup of coffee by adding cold
milk; some students felt that the increase in contact area
would offset the decrease in driving force. It is, of course,
a classical example.[891
The students are asked to determine: if you want a cup of
coffee to be as hot as possible after five minutes, is it better
to put the milk in immediately or at the end of the five
minutes? The rate of cooling of the coffee is determined by
measuring the temperature using a hand-held digital ther-
mometer. Measurements are made on coffee with and with-
out milk, and also with the cup covered and/or exposed to a
fan. The students are also encouraged to drink the coffee

As an adjunct to this experiment, they are also asked to
perform two heat balances on a kettle while it is heating up
from cold to the boiling point and while it boils for five
minutes. A digital wattmeter is used for measuring the power
input to the kettle and a digital scale for weighing.

Mass Transfer: Dissolution of Suckers
The germ of the idea for this experiment came from Sensel
and Myers.1101 They dissolved particles of sourball candy in
an agitated system and then dried and weighed them to work
out the rate of dissolution. In order to model the dissolution
and to make for easier measurements, I thought of using a
round sweet that could be suspended in water. The answer to
this came when I was out with my daughters, buying some of
the equipment for these experiments. They bought some
round suckers on sticks. When we got home I suddenly
realized that this was exactly what I needed! I immediately
placed one in some cold water to see how long it would take
to dissolve. In twenty minutes it shrank from a diameter of
25 mm to 15 mm, which was just the right time scale. It was
not too fast for accurate measurements of the diameter to be
taken, but it was fast enough to allow testing of other condi-
tions as well, such as stirring or the effect of warm water (all
of which would increase the rate of dissolution), within the
total time available.
The experiment was formulated accordingly. Magnetic
stirrers were used for stirring and vernier calipers for mea-
suring diameter. As it happened, these suckers had sherbet
cores, so there was no point in dissolving them too far. The
students were therefore instructed to go ahead and eat them
when they reached a certain size!

Reaction Kinetics: Cooking Potatoes
This was the one I struggled with the most. How could I
find a reaction that the students could see happening right
before their eyes? Then, I read the comment "Consider bak-
ing a potato" at the end of the paper on model development
by Barton,18' and I suddenly remembered a demonstration
that one of my colleagues, Geoff Hansford, had done for
school children: he had cooked potatoes for different lengths
of time, cutting them open to reveal how far the cooking had
progressed. This suited my purpose ideally.
The students are given three sets of potatoes (small, me-
dium, and large) and are given different lengths of time for
cooking each of them. A vernier caliper is used to measure
the diameter of the whole potato and the uncooked por-
tion (the interface between the cooked and uncooked
potato is very distinct).
Momentum Transfer: Fluid Flow through Thin Tubing
I felt that momentum transfer is the most difficult of these
four concepts, so I did not use the term with the students,
simply referring to it as a fluid flow experiment. I wanted the
students to experience the pressure that is needed to make a

fluid flow through a pipe. I set up a series of pipes (thin
tubes, actually). The fluids were chosen for their wide range
of viscosities: water (1 cP), ethyl alcohol (1,2 cP), isopropyl
alcohol (2,23 cP), a 50% water-glycerol mixture (6,3 cP),
and ethylene glycol (23 cP). The density range is not as high
as I would have liked, from 789 to 1130 kg/m3 (bearing in
mind that in laminar flow the pressure drop for flow through
a pipe is independent of density).
For each fluid there were three tubes of nominal size 1/4",
3/16", and 1/8". A large medical syringe of 60-ml capacity
was used to suck the fluid from a reservoir into the tube and
then to force it out again. A tee-piece was used to join the
syringe, a pressure gauge, and the tube. The students had to
time the discharge of a certain volume through the tube
and measure the pressure for this flow. This was used to
verify the Hagen-Poiseuille law (AP = 32 gLv / d2, where
I. is viscosity, L is pipe length, v is fluid velocity, and d
is pipe diameter).

The equipment for these experiments was all purchased
and assembled within a fortnight. The apparatus worked
well, as would have been anticipated, apart from leaks in the
tee-pieces of the fluid-flow rigs.
One problem encountered was with the pressure gauges.
The ones originally used were only meant for positive pres-
sures, and this meant that they were damaged when sucking
the fluid into the syringes, especially in the lines with the
thin tubes and the higher viscosity fluids. The gauges were
therefore all replaced by pressure-vacuum gauges.
In this experiment you also have to be careful not to over-
pressurize the system or the flexible tubing connecting the
syringe to the tee-piece comes off the end of the syringe,
which is slightly conical. Another problem arose with the
heated stirrers-any sugar solution spilled on them tended
to carbonize, so they have to be cleaned carefully each
time they are used.

The experiments will now be evaluated against each of the
objectives listed earlier.
1. They introduced students to each of the four fundamental
2. They provided hands-on exposure to the processes. Students
at the end of their studies rated them on average as 4.1 on a
scale of 1 to 5 in terms of helpfulness.
3. The experiments used familiar equipment and concepts
(coffee cups, kettles, a fan, cooling, suckers, dissolution,
potatoes, pots, hot plates, cooking, syringes, water, ethyl
alcohol, antifreeze, flow) as well as unfamiliar equipment and
concepts (digital thermometers and wattmeters, heat transfer,
vernier calipers, magnetic stirrers, mass transfer, reaction
kinetics, pressure gauges, metal tubes, isopropyl alcohol,
Chemical Engineering Education

4. Students appeared to enjoy doing the experiments and tackled
them with great enthusiasm.
5. Each of the experiments was readily completed in one-and-a-
half hours by a pair of students.
6. A class of 160 was able to perform the experiments in four
sessions of one-and-a-half hours per week over four weeks.
7. The experiments were all safe, apart from the boiling kettle,
which is no more dangerous than what is done routinely in
the home and was used to bring home the danger of live
steam. The fluids were specifically chosen with safety in
mind-all are in common use and are safe unless ingested in
large quantities.
8. Five sets of equipment for all the experiments were purchased
for roughly $6,000.
9. The equipment was F e fl
Figure 1. pi
all purchased and 4.0
assembled within x /
two weeks. /7'
10. The equipment has 3. / ,' /'
lasted well. The /, -
only problems E 2.0 *
have been failure '
of the digital' -"
thermometers and M 1.0
wattmeters (care
also had to be
0.0 i
taken to remove 0 10 20
the batteries of Time (mn)
these items
between use).

11. Five sets of apparatus were able to be stored in five standard
laboratory cupboards.
12. The equipment is readily transported and has been used by
other institutions in the area.

Clearly, all of the objectives were met. The timing was
also amazing-without planning it, earlier in the week in
which we started the experiments the students were taught
how to read a vernier scale in physics. Students also com-
mented that the fluid flow experiment helped them to appre-
ciate the Bernoulli equation taught in physics.

A number of important features of the experiments are
exploited in discussion of the modeling. The first of these is
the importance of physical observations. For example, in
still water the bottom of the sucker dissolves away more
rapidly than the top. Close observation reveals that there is a
downward convection current of concentrated sugar solution
below the sucker. This does not appear to affect the top half
of the sucker, so it is still valid to assume diffusion in
modeling the dissolution.
Another aspect is the variability of real systems. The suck-
ers are neither completely round nor all exactly the same
size. The potatoes are certainly not all the same shape, and
within each size class there is also considerable size varia-
Summer 1999

tion. Some potatoes are also non-uniform inside.
The data for these experiments also brings out the impor-
tance of how a problem is represented for meaningful inter-
pretations to be made. In both the sucker and potato experi-
ments it is not helpful to look at the final radius when
making comparisons when the initial radii are different. As
soon as the data is presented as differences in radii, however,
clear trends emerge.
In the following paragraphs I will deal with the modeling
of the sucker dissolution and the potato cooking. I am able to
start this section of the course shortly after the students have
been taught differential equations in mathematics, thus pro-
viding motivation
for the mathematics
tion of Suckers
they are being taught
Cod Unstirred by showing that it is
SColdSrred at2 needed in chemical
A Cold.e,,at4 engineering.
X Hot, Unstired
X Hot Stirred at 2 Sucker
HotStirredat4 Dissolution
------ Linear (Cold, Unstirred)
Linear (Cold Stirred at 2) This is modeled as
--- Linear(HotUnstirred) diffusion of dis-
S- - Linear (Cld, Stirred at 4) solved sugar from
Linear(Hot Strred at2) the surface of the
40 -Linear (Hot Stirred at 4)
sucker into the sur-
rounding water. The
rate of diffusion into
the water is equated to the rate of shrinkage of the sucker. It
is assumed that the bulk concentration of the sugar in the
water does not change significantly. This yields the follow-
ing straightforward differential equation in which the rate of
change of radius with time is a negative constant:

dr kAC (2)

where k is the mass transfer coefficient, AC is the concen-
tration difference between the surface of the sucker and the
bulk water, and ps is the density of the sucker.
Solution of this differential equation gives a linear de-
crease of sucker radius with time, provided the term in the
brackets is constant (the only variable in this term that will
change with time is AC, but on checking the change is
minimal and may be neglected):

ri -r= t (3)
I Ps )
Figure 1 shows the fitting of this model to six sets of
experimental data, obtained at two different temperatures
and three different stirrer speeds. The slope of the straight
line includes two sets of variables, one being k, the mass
transfer coefficient (which is a function of the rate of stir-
ring), and the other (AC/ps), the concentration difference



Given that there are six sets of data, we can use the slopes
fitted to the experimental data to solve for the unknown
values of k and ( AC / ps) by regression, as shown in Table 1.
The absolute values of the variables are not important, but
we can draw conclusions from their relative values. The
mass transfer coefficient, as expected, is a nonlinear func-
tion of stirrer speed and the major variable in the other
group, the equilibrium concentration of sugar, approximately
doubles from the cold to the warm water.

Potato Cooking
In order to model this situation, a number of simplifying
assumptions have to be made. The first is that the potatoes
can be taken to be spherical. The next is that the rate of
cooking is determined by the rate at which heat arrives at the
cooking interface. This is used in conjunction with the as-
sumption that all the heat transferred to the interface is used
for the cooking reaction (this is based on the heat of reaction
being much larger than sensible heat effects). I also assume
that the driving force for heat transfer is constant-measure-
ments of the temperatures of the outside of the potato and the
cooking interface show that they stay constant at 980C and
65'C, respectively (these measurements were
suggested by my twelve-year-old daughter!).
In developing the differential equation for T
this system, you need an expression for con-
duction through a spherical shell. This is Sucke
readily derived as part of the analysis. This, Stirring: n
plus all the assumptions mentioned above, K 0.
leads to a differential equation that is a func-
tion of the outside radius and the radius of Temp: f
uncooked potato at any particular time: (AC / s) 1.

Figure 2. Potato Cooking Rate

18 -

- 14-
E 12


0 4-



dr1 4 4jikMAT 1
dt AHRp ) r2 1
ro rio

where r, is the outer potato diameter, r, is the radius of the
uncooked potato, k is the potato thermal conductivity, M is
the potato molar mass, AHR is the potato molar heat of
reaction, p is the potato density, and AT is the temperature
difference between the outside of the potato and the cooking
This equation is readily solved analytically, giving a cubic
relation between the uncooked radius and time:

1 l 3 1 2 1 2
3 ro 2 6r

4 7kMAT 0

If the assumptions are valid, then the term in the square
bracket would be constant. This equation is therefore solved
for this term and it is evaluated from the experimental data
for the outside and interface radii at different times. The
results are shown in Table 2, and this term is
found to be about the same for all points,
E 1 except for the single data point at one minute
s for and the longest time for each size. This
solution justifies the use of the assumptions made,
over all but the initial and final phases of
0.5 0.287 the experiments.
0.175 0.287
Figure 2 shows the resulting analytical so-
371C lution compared with the actual data. As one
1.978 would expect, only the points in Table 2 that

5 10 15 20 25
Time (mins)

Chemical Engineering Education
Chemical Engineering Education

r Diss

one s


were out of line do not match the predictions.
The deviation for the last data point in each size
is probably due to the assumption of a constant
temperature at the cooking interface breaking
down as the center of the potato is reached.
This exercise illustrates how one can derive a
model on the basis of a fairly gross simplifica-
tion of a situation, and also use it to make mean-
ingful predictions, even though one cannot di-
rectly measure the characteristics of the process,
such as the heat of reaction of the potato.

The course as a whole was evaluated by ques-
tioning students in the second year and the fourth
year. A free-form questionnaire was used in both
instances. In the second evaluation, students were
also asked to rate each of the main aspects of the
course. These two methods were used to obtain
both what had left an impression on the students
and the relative value they perceived in all the
aspects of the course.
When asked to give the most useful features
of the course, roughly two-thirds of them men-
tioned unit conversion (69% after one year and
64% after three years). In addition, the experi-
ments were mentioned by 31% after one year
and 56% after three years (this increase seems
significant). In both evaluations, no other topic
came close to these. In the first instance, they
were also asked to mention the most confusing
aspect of the course, and 16% felt the experi-

Evaluation of Constant Term for Potato

Small 1 1.70 5.33
2 3.70 11.60
3 4.80 12.16
4 5.50 11.94
5 7.75 17.37
Medium 3 4.35 11.08
6 6.90 12.90
9 7.80 10.25
12 11.55 15.06
Large 5 5.65 11.13
10 8.70 12.54
15 10.65 11.89
20 15.55 15.89

Average (of values between 10 and 13) 11.72

Summer 1999

ments had been confusing.
The overall helpfulness of the course was rated as 3.0 on a scale of 1 to 5
after one year, and 3.5 after three years. After three years, the two highest
ratings of course components were unit conversion (4.9) and the experi-
ments (4.1), followed by transfer processes (3.9), plant visits (3.9), and the
modeling of the experiments (3.8).
Unit conversion and the experiments (plus the related modeling and
transfer processes) were consistently the most significant aspects of the
course for the students. The increased rating of the experiments after three
years points to the long-term impact that they had.

The experiments described in this paper perform the crucial role of
introducing first-year students to four key fundamental physical phenom-
ena occurring in the majority of chemical engineering processes. They also
serve as a basis for exposing the students to modeling of real phenomena.
This was a very exciting part of this new course, which is an important
basis for the new curriculum we have developed at the University of Cape
Town. It has also given students something to refer back to when they
encounter the theory that uses these phenomena later in their studies.
Full details of the experiments may be obtained by e-mailing the author

I wish to acknowledge the help received from God, the Creator of the
universe, both in giving me the creativity needed to generate these ideas
and for placing the correct material in my path at just the right time. I also
wish to thank my two younger children, Andrew and Ann (ages 14 and 12
at the time) and our friend Brett Melville (age 17 at the time), for helping
me perform the experiments to get the data for modeling and for their keen
observations, patience, and ideas for extra measurements to take.

1. Fraser, D.M., "A New First-Year Programme for Engineers at the Univer-
sity of Cape Town," Proc. Fourth World Conference on Engineering Educa-
tion, St. Paul, MN; Vol. 2, 160 (1995)
2. Holland, W.D., and J.C. McGee, "An Interesting and Inexpensive Modeling
Experiment," Chem. Eng. Ed., 27, 150 (1993)
3. Ryan, J.T., R.K. Wood, and P.J. Crickmore, "An Inexpensive and Quick
Fluid Mechanics Experiment," Chem. Eng. Ed., 17, 140 (1993)
4. Fee, C.J., "A Simple but Effective Fluidized-Bed Experiment," Chem. Eng.
Ed., 28, 214 (1994)
5. Van Wie, B.J., J.C. Poshuta, R.D. Greenlee, and R.A. Brereton, "Fun Ways
to Learn Fluid Mechanics and Heat Transfer," Chem. Eng. Ed., 28, 188
6. Nirdosh, I., and M.H.I. Baird, "Low-Cost Experiments in Mass Transfer,"
Parts 1 and 2, Chem. Eng. Ed., 30, 50 and 142 (1996)
7. Palanki, S., and V. Sampath, "A Simple Process Dynamics Experiment,"
Chem. Eng. Ed., 31, 64 (1997)
8. Barton, G.W., "Model Development and Validation: An Iterative Process,"
Chem. Eng. Ed., 26, 72 (1992)
9. AIChE Extra, "Coffee Cools More Quickly If You Wait to Add the Cream," 4
10. Sensel, M.E., and K.J. Myers, "Add Some Flavor to Your Agitation Experi-
ment," Chem. Eng. Ed., 26, 156 (1992) 1

Cooked Thickness

Size Time


Random Thoughts...


North Carolina State University Raleigh, NC 27695

I If a doctor, lawyer, or dentist had 40 people in his office at one time, all of whom had different needs, and some of
whom didn't want to be there and were causing trouble, and the doctor, lawyer, or dentist, without assistance, had to
treat them all with professional excellence for nine months, then he might have some conception of the classroom
teacher's job.
Donald D. Quinn

3 Thoroughly to teach another is the best way to learn for yourself
Tryon Edwards

I You do not really understand something unless you can explain it to your grandmother.
Albert Einstein

3 It is noble to be good, and it is nobler to teach others to be good-and less trouble!
Mark Twain

3 The task of the excellent teacher is to stimulate "apparently ordinary" people to unusual effort. The tough problem is
not in identifying winners: it is in making winners out of ordinary people.
K. Patricia Cross

( I am not impressed by the Ivy League establishments. Of course they graduate the best: it's all they take, leaving to
others the problem of educating the country. They will give you an education the way the banks will give you money,
provided you can prove to their satisfaction that you don't need it.
Peter De Vries
Richard M. Felder is Hoechst Celanese Pro-
fessor of Chemical Engineering at North Caro- 3 One mark of a great educator is the ability to lead
lina State University. He received his BChE
from City College of CUNY and his PhD from students out to new places where even the educator has
Princeton. He has presented courses on chemi- never been.
cal engineering principles, reactor design, pro-
cess optimization, and effective teaching to vari- Thomas Groome
ous American and foreign industries and insti-
tutions. He is coauthor of the text Elementary
Principles of Chemical Processes (Wiley, 1986). 3 When you teach well, it always seems as if 75% of the
students are above the median.
copyright ChE Division of ASEE 1999 Jerome Bruner
196 Chemical Engineering Education

3 If at first you do succeed, try to hide your astonishment.
Source unknown

0 Picture yourself in France in a cave with prehistoric drawings on the wall. These drawings tell a story and were
perhaps the first use of technology for educational purposes. Now, thousands of years later, professors are still
drawing on walls!
Bruce Finlayson

3 The best learners...often make the worst teachers. They are, in a very real sense, perceptually challenged. They
cannot imagine what it must be like to struggle to learn something that comes so naturally to them.
Stephen Brookfield

3 The vanity of teaching often tempteth a man to forget he is a blockhead.
George Savile

3 Football combines the two worst elements of American society: violence and committee meetings.
Herb Childress

0 University politics are vicious precisely because the stakes are so small.
Henry Kissinger

0 The teachers who get "burned out" are not the ones who are constantly learning, which can be exhilarating, but
those who feel they must stay in control and ahead of the students at all times.
Frank Smith

0 When Pablo Casals reached ninety-five, a young reporter asked him a question: "Mr. Casals, you are ninety-five
and the greatest cellist who ever lived. Why do you still practice six hours a day?" Casals answered, "Because I
think I'm making progress.

0 Don't say you don't have enough time. You have exactly the same number of hours per day that were given to Helen
Keller, Louis Pasteur, Michelangelo, Mother Teresa, Leonardo da Vinci, Thomas Jefferson, and Albert Einstein.
H. Jackson Brown, Jr.

0 Ninety-five percent of this game is half mental.
Yogi Berra

I can't give you a brain, but I can give you a degree.
The Wizard of Oz

'See also "Speaking of Education," Chem. Engr. Ed., 27(2), 128 (1993)

All of the Random Thoughts columns are now available on the World Wide Web at and at

Summer 1999

f, curriculum



Texas A&M University College Station, TX 77843-3122

accident statistics for 1989 from the Accidental Re-
lease Information Program (ARIP) of the U.S. En-
vironmental Protection Agency'" are shown in Fig-
ure 1. These statistics cover catastrophic and unplanned
releases of chemicals into the atmosphere. They underline
the fact, however, that a large number of accidents and
catastrophic releases occur because of design flaws, wrong
equipment specifications, and lack of or disregard for oper-
ating and maintenance procedures. The boardroom perspec-
tive on the cause of these accidents and what to do about
them varies, but many believe that safety in the process
industry is of primary importance and is critical to the
industry's continuing "license to operate."
The total number of process plant accidents cannot be
accurately estimated because of underreporting, but the num-
ber is large and many people, both workers and the public,
are adversely affected by the accidents. For example, in
1991 the National Response Center received over 16,300
calls reporting the release or potential release of hazardous
chemicals.[21 Another study131 analyzed the EPA's Emergency
Response Notification System database of chemical acci-
dent notifications and found that from 1988 through 1992,
an average of nineteen accidents occurred each day, i.e.,
more than 34,500 accidents involving toxic chemicals oc-
curred over the five-year period. The promulgation of the
Toxic Release Inventory Reporting requirements'41 as part of
the Clean Air Act Amendments of 1990 led to the submis-
sion of toxic release information that clearly delineated the
number and extent of toxic chemical releases and their po-
tential impact on the public and on the environment. The
university plays a critical role in changing this situation.
Change in population demographics, increasing aware-
ness of process plant hazards, and above all, the continuing
threat of a chemical catastrophe continue to provide the
impetus for governments to develop legislation for eliminat-
* Corresponding author.

ing or minimizing the potential of such accidents. Interna-
tional efforts include the Seveso Directive covering mem-
bers of the European Community. Other nations have simi-
lar laws, such as the Sedesol guidelines in Mexico for
performing process risk audits, and the post-Bhopal acci-
dent-prevention law in India. The World Bank has devel-
oped guidelines for identifying and controlling hazards,
and the International Labor Organization has developed a
code of practice for preventing major accidents.
In 1990, the U.S. Congress enacted the Clean Air Act
Amendments (CAAA), which directed the Occupational

M. Sam Mannan is Associate Professor of Chemical Engineering at Texas
A&M University and Director of the Mary Kay O'Connor Process Safety
Center. He received his BS degree from the Engineering University in
Dhaka, Bangladesh, and his MS and PhD degrees from the University of
Oklahoma. His research interests include process safety and risk manage-
ment, quantitative risk assessment, reactive chemistry, and fate and trans-
port modeling of chemical releases.
Aydin Akgerman is the Chevron II Professor of Engineering in the Chemi-
cal Engineering Department at Texas A&M University. During his career
he has taught at Bogazici University and Ege University (in Istanbul and
Izmir, Turkey, respectively), and at Texas A&M University. He has also
worked as the R&D Manager at Cimentas Izmir Cement Plant.
Rayford (Ray) G. Anthony is the C.D. Holland Professor of Chemical
Engineering and Head of Chemical Engineering at Texas A&M University.
He received his BS degree from Texas A&M University and his PhD from
the University of Texas. His research interests include development of
catalysts and modeling catalytic reactors, and mathematical modeling of
multiphase reactors.
Ron Darby is Professor of Chemical Engineering at Texas A&M Univer-
sity. He holds a PhD degree from Rice University and has been at Texas
A&M since 1965. His primary research interests are flow of complex fluids,
two-phase flows, viscoelastic and non-Newtonian fluids, slurries and sus-
pensions, and process safety.
Philip T. (Toby) Eubank is Professor of Chemical Engineering at Texas
A&M University. He received his BS degree from Rose-Hulman Institute of
Technology and his PhD from Northwestern University. His research inter-
ests are in the thermo-physical properties of fluids and fluid mixtures plus
electrical discharge machining.
Kenneth R. Hall is the GPSA Professor of Chemical Engineering and
Director of the Thermodynamics Research Center at Texas A&M Univer-
sity. During his career, he has worked with AMOCO and ChemShare and
has taught at the University of Virginia and Texas A&M University.
Copyright ChE Division of ASEE 1999
Chemical Engineering Education

Safety and Health Administration (OSHA) and the Environ-
mental Protection Agency (EPA) to develop standards for
reducing the frequency and severity of chemical plant acci-
dents. In keeping with the congressional mandate, OSHA
promulgated the Process Safety Management (PSM) rule,
intended to protect workplace employees. Similarly, EPA
promulgated its risk-manage-
ment program rule in 1996 to 6
protect the public and the envi-
ronment. In the United States, s 5
federal agencies are not the only t
government regulators active in 40 -
the chemical accident preven- t
tion arena. Several states have ) 30
empowered their health, safety, -
and environmental agencies to Q,20
create regulations requiring
companies to establish and prac- 2 10
tice specific programs to im-
prove safety. Equipment Miscellam

Laws and regulations are I osI requenm c
logical reactions to cata- Figure 1. U.S. Environme
strophic process plant acci- tics on Accidental Release
dents. But can the mere pro-
mulgation and enforcement of
laws and regulations actually Non-conducting hos
affect the frequency and se-
verity of process plant acci-
dents? The philosophical is-
sue is that we can only regu-
late something for which we Steel
have knowledge and under- Nozzle
standing. For example, OSHA
process safety management
regulations require facilities to
develop and implement man- ump
agement of change procedures.
That is, before a process Figure 2. Static elec
change is implemented, engi- proce,
neers must evaluate the change and ensure that it is techni-
cally sound and cannot result in a hazardous situation. The
evaluation could consist of a hazard and operability (HAZOP)
study conducted by a multidisciplinary team using some
HAZOP software available in the marketplace.
For the process shown in Figure 2, consider the addition of
an organic A with a certain thermal conductivity to the glass-
lined reactor. During the original design, engineers made
necessary calculations to ensure that the voltage and ignition
energy caused by static electricity did not exceed the danger-
ous limits of 350 V and 0.1 mJ, respectively. Above these
limits there exists a potential for spark and possible fire and
explosion. But in response to market demands for product
specifications, the plant is planning a switch to organic B,

the only difference in thermophysical properties being a
slight increase in thermal conductivity. The calculations now
show that the voltage and ignition energy caused by the
static electricity exceeds the dangerous limits. The most
important question is whether we make this determination as
part of an after-the-fact accident investigation or as part of
the management of change
process. If we choose the lat-
ter, we must understand the
gravity of the problem and
take appropriate corrective
and remedial measures. These
measures may include instal-
lation of additional ground-
ing, control of flow rate to
reduce static electricity, and
relaxation (hold time to al-
low for charge reduction).
In addition to the above is-
pset Bypass Fire sues, the issues of inherent
s of Releases safety in process design,

ntal Protection Agency statis-
Information Program, 1989.

:tricity and the impact of
ss changes.

equipment selection, and op-
erating and maintenance pro-
cedures depend to a large ex-
tent on a fundamental under-
standing of the underlying sci-
ence and application of those
principles to the problem at
hand. For example, in design
and construction of a poly-
ethylene plant that uses a large
amount of flammables at very
high pressures and tempera-
tures, the inherent hazard is
that any accident has the po-
tential of releasing large quan-
tities of the flammable, which
because of the thermodynam-
ics can likely flash and form

an aerosol. While "bells and whistles" can be added after the
fact to make the process extrinsically safer, the comprehen-
sive education and research approach suggested in this paper
equips the engineer to come up with an intrinsically safer
process during the design and construction phase. Some
solutions may include intensification, substitution, attenua-
tion, or limitation of effects. These concepts, while avail-
able in some literature,"5-7] are not covered adequately in
chemical engineering instruction and research.
What can we do to fix or reduce the extent of the funda-
mental problem described here? The challenge is how to
create a culture in which consideration of process safety
issues is second nature, driven by a total understanding of
the underlying engineering, process chemistry, and other

Summer 1999





factors. While regulations, plant policies and procedures,
and industry standards accomplish much, universities must
play a significant role in addressing this challenge. The role
of engineers has changed dramatically, and as a result, uni-
versities must provide an integrated
engineering education that equips
engineers not only with the classi- The challenge
cal fundamental subjects (thermo- culture in which
dynamics, fluid mechanics, reac-
tion kinetics), but also provides process safety
them with an understanding of pro- nature, dri
cess safety engineering and how understanding
they can use their knowledge of
fundamental engineering subjects engineering, p
to make the process plant safer. and oth
The need is not the establishment
of a new discipline, but one of fine
tuning the engineering curriculum. To this end, Texas A&M
University established the Mary Kay O'Connor Process
Safety Center, an industrially sponsored Center of Excel-
lence, to produce engineers trained in process safety and to
provide industry with the research base it needs to compete
successfully in the global marketplace.
The Center charter is to broaden the scientific and engi-
neering knowledge base of industry and to educate engineers
and scientists in the field while striving to achieve techno-
logical breakthroughs necessary to reach ambitious long-
term, systems-level engineering goals. Its mission includes
bringing together researchers from diverse industrial, aca-
demic, and governmental laboratories whose work can con-
tribute to the development of process safety issues that can
have a far-reaching impact on the chemical processing in-
dustry. The Center also has the responsibility of outreach to
industry, to other universities and educational institutions,
and to the public as a whole. A program at the Michigan
Technological University bears certain similarities inso-
far as the educational and research component is con-
cerned. As illustrated by the following discussion, how-
ever, the Center programs span not only education and
research, but also include training, service, information
dissemination, and symposia.
The Center has also recently established a dialogue with
the Center for Chemical Process Safety (CCPS) in order to
coordinate activities for accomplishing mutually desirable
process safety goals. While CCPS also focuses on making an
impact on process-safety-related issues, the Center's goals
and objectives are much deeper. They include shifting the
paradigm to safety being second nature and incorporating
process safety into the curriculum. The Center programs and
activities are meant to complement and enhance the CCPS
efforts. Success toward these goals at Texas A&M Univer-
sity is a first step in this process. Future plans include activi-
ties to encourage similar initiatives at other institutions. Based

is l

on the availability of funds, these activities could include
joint projects or grants to encourage teaching and research in
the areas of process safety. The Center is working on several
proposals that could lead to joint projects with other univer-
sities under grants provided by
government agencies.
low to create a
consideration of GOALS OF AN

by a total The first step in accomplishing
te the goals and objectives from the
the underlying university perspective is for edu-
less chemistry, cators to recognize that process
factors. safety should be integrated into a
comprehensive instructional and
research program. For example, is
it appropriate for educators to teach process design courses
without adequately covering concepts of inherently safe de-
sign and other process safety concepts? Is a course on reac-
tion engineering complete without due treatment of runaway
reactions, the causes of such reactions, and what role an
engineer might have in preventing them? Finally, offering
the opportunity for students to take specific process safety
engineering courses is critical in the integrated approach.
Research should be directed toward developing safer pro-
cesses, equipment, procedures, and management strate-
gies that will minimize losses within the processing in-
dustry. The goals of an integrated approach span a large
spectrum of issues focused toward programs and activi-
ties that encourage safety as second nature. The goals
cover four broad areas: instruction; information storage,
retrieval, and analysis; service; and research. Some of
the general goals include
Marshaling all the resources of the university that can be
applied to process safety and risk management, advertising
these capabilities, and bringing these resources together to
solve complex problems that require multidisciplinary teams
Developing the capability to respond quickly and effectively
to the research needs of other organizations
Attracting outstanding faculty, researchers, practitioners,
and students to participate in process safety research
programs and activities
Sponsoring or participating in safety-related events such as
symposia and design contests
Serving as a role model in good safety practicesfor other
institutions and within the University
Of these general goals, attracting outstanding faculty, re-
searchers, practitioners, and students is by far the most criti-
cal. The extent of the problem is illustrated by the fact that
because of industry initiatives and regulatory requirements,
process safety engineering and associated technologies have
become an essential feature of all chemical processing de-
sign and operations. But almost all universities lack effec-
Chemical Engineering Education

tive teaching and research programs to support the needs of
industry. This situation can be changed only by putting in
motion a cycle that irrevocably changes the paradigm. For
example, we could produce several chemical engineering
PhDs per year with specialization in process-safety engi-
neering, and they could then go on to teach at other universi-
ties or conduct beneficial research in solving process safety
problems. Thus, the courses they teach (including classical
engineering courses) would contain a comprehensive ap-
proach including consideration of all process safety issues.
In addition, their research would definitely include the solu-
tion of many process safety problems.

The educational programs of the Center are based on a
three-pronged approach. First, to establish a series of under-
graduate and graduate courses dedicated specifically to pro-
cess safety engineering. Second, to act as a catalyst for
incorporating process safety problems into existing courses
such as design, reaction kinetics, and thermodynamics. Third,
to sponsor training of engineering faculty through participa-
tion in continuing-education short courses covering process
safety. The overall goals in education include
Improving knowledge and awareness of process hazards and
safety for faculty, students, engineers and other professionals,
plant workers, public safety personnel, transportation
workers, and the public
Developing state-of-the-art educational tools, undergraduate
and graduate courses, and continuing-education programs
Producing engineers with a good education in safety

The current program of the Center includes an interdisci-
plinary, elective course in process safety engineering that is
cross-listed between chemical engineering and safety engi-
neering programs and has been taught for the last three
years. The course is one of the most popular electives in the
department. A graduate counterpart of the course has also
been developed and taught.
The Center promotes the use of SACHE modules within
traditional chemical engineering courses. In addition, to in-
crease faculty awareness, the Center sponsors participation
in continuing education short courses on process safety. The
intent is to provide information on state-of-the-art safety
technologies as well as to encourage faculty to use these
courses as opportunities to update process safety elements in
the traditional chemical engineering courses. To date, vari-
ous faculty have attended the following courses:

Engineering Design for Process Safety
Tools for Making Acute Risk Decisions
Methods for Sizing Pressure-Relief Valves
Fundamentals for Fire- and Explosion-Hazards Evaluation
Use ofHAZOP Studies in Process-Risk Management
Human-Error Evaluation and Human-Reliability Analysis
Summer 1999

for Chemical Process Systems
Safe Automation of Chemical Processes
Consequences of Vapor Cloud Explosions, Flash Fires, and
Vapor Cloud Dispersion Modeling

The Center has begun an aggressive program to provide
continuing-education courses to practitioners in industry.
The intent is to provide training at outreach locations in a
format that allows attendees to take the short courses with-
out having to travel long distances and with minimal disrup-
tions in their work schedule. We started with a 13-course
syllabus at two campuses: Texas A&M University System
Galveston campus and the Texas Engineering Extension
Service Pasadena training facility. The courses, taught by
both industry and university experts, meet Monday and Tues-
day from 8am to 5pm. Continuing education credits are
provided for all short courses and attendees may choose to
take structured series of courses and receive certificates of
attendance for a specific program.

The Center has future plans calling for continued growth
and expansion of the efforts already underway. Several ad-
vanced-level courses on process safety and associated tech-
nologies are being developed. They can be taught by a
multidisciplinary team of instructors and offered at multiple
campuses through distance-learning technology. Some of
the courses under consideration for development include
Mechanical integrity of process plants (potential teaming
between chemical and mechanical engineering departments)
Advanced topics in safety and environmental management
(potential teaming between chemical engineering, industrial
engineering, and chemistry departments)
Quantitative risk assessments (potential teaming between
chemical engineering, statistics, and business administration

In the continuing education program, the Center plans to
add appropriate courses as necessary, but the ultimate objec-
tive is to move from the current campus-oriented offerings
to an interactive distance-learning system. Texas A&M Uni-
versity has already implemented distance-learning curricula
in the industrial engineering department. The Center intends
to collaborate with industrial engineering to develop dis-
tance-learning course modules for both graduate courses and
continuing education courses.
Within the next few years, MS and PhD graduates in
chemical engineering will be finishing their degree pro-
grams with emphases in process safety engineering. The
degree programs for these graduates will include
Traditional core chemical engineering courses
Additional process-safety-specific courses
MS or PhD theses addressing the solution of an engineering
problem related to process safety

One of the main causes of process safety incidents in the
chemical processing industry is the lack of access to neces-
sary information and data. The problem is threefold: first, in
many cases the information does not exist; second, even
when some information and data are available, accuracy and
credibility are questionable; and third, when information
exists, it is not well organized or easily accessible. Thus, in
the area of information storage, retrieval, and analysis, the
Center's goals include
* Gathering and storing information related to chemical
process safety, including case histories, equipment and
human reliability
Developing computer databases and user interfaces to
provide easy access to and analysis of this information
Analyzing the information andpublicizing the results

The heart of the Center work is its library, which includes
books, articles, reports, journals, and other documents fo-
cusing on engineering aspects of process safety (e.g., relief
systems, dispersion modeling, safe design) as well as the
social, economic, and behavioral aspects of process safety
incidents and natural disasters. Various software programs
are also available. The holdings are cataloged in a computer-
ized bibliographic database. The library catalog is available
on-line on the Center website, enabling web browsers to
search the library materials for specific publications.
The Center publishes the Centerline three times a year. It
contains technical and research issues of interest in the field
of process safety and risk management. It is also available
on the Web site ( The site
provides information on process safety-issues, publications,
and other items of interest for process-safety and risk-man-
agement topics. It also allows individuals, companies, and
organizations to browse actively and to acquire information
on process-safety-related subjects. Access is free and allows
the user to conduct interactive searches and provides compu-
tations, analyses, and calculations. The site contains infor-
mation on research, technical papers and reports, access to
the library database, regulations, frequently asked questions,
access to software, links to other appropriate sites, electronic
Centerline issues, and announcements for symposia, semi-
nars, and short courses. The site is updated regularly to
provide new items and state-of-the-art techniques to users.
Future plans for information storage, retrieval, and analy-
sis include development of computer databases and user
interfaces to provide easy access and analysis of process-
safety-related information. For example, one item under con-
sideration is development of an incident-history database
with fuzzy search capability. This effort can expand to de-
velop an interactive teaching module providing Web-based
training. Also, efforts are underway to establish a Process
Safety Newsgroup (PSN) that would provide an open forum

for exchange of ideas and questions for personnel involved
in the process-safety and risk-assessment fields. The pur-
pose of PSN would be to facilitate the exchange of ideas and
information among U.S. and international public- and pri-
vate-sector organizations about prevention of, preparations
for, recovery from, and/or mitigation of risk associated with
catastrophic accidents in chemical processing facilities.
Another current project is analysis of accident databases to
pinpoint specific causes and to determine areas of needed
research. The intent is to use the results to determine areas of
critical need and to focus efforts on those areas. At this time,
the project consists of analyzing portions of the EPA Acci-
dental Release Information Program to develop a strategy of
how these databases can be used to improve process safety.
As new information is compiled and research results be-
come available, the Center will disseminate them as widely
as possible. In many cases, it may be necessary to publish
monographs, research papers, and guidelines. The changing
environment and needs of industry dictate, however, that we
consider advanced electronic media such as CD-based publi-
cations and internet communication.

The mission of a university and its faculty includes pro-
viding service to industry and society. The changing nature
of the chemical engineering profession necessitates that we
take a closer look at how we provide this service. Universi-
ties and faculty are remiss if they do not play an adequate
role in ensuring public safety. Another issue is that a large
number of process plants exist that are either owned by small
companies or are so-called "mom-and-pop" operations. An
accident from such a small facility has the potential of severe
consequences and can damage the whole industry's "license
to operate" just as does an accident in a large plant. The
larger facility probably has resources, training, and equip-
ment to either prevent the accident in the first place or to
respond to the consequences if it does occur, however, while
the small facility probably lacks proper awareness, training,
and information. Thus, the Center's goals include
Providing service to small and medium enterprises,
government agencies, institutions, local emergency planning
committees, and others to evaluate and minimize risks
Providing independent accident investigation and analysis
services to industry and government agencies, particularly
for those accidents that suggest new phenomena or complex
In the area of service to small business, the Center seeks
collaborative efforts with government agencies (both state
and federal), professional and trade organizations, and in-
dustry. Another area of interest for the Center is accident
investigation. The Center objective in looking at accidents is
fourfold; first, identifying multiple accidents that may ex-
hibit common phenomena; second, finding accidents that
Chemical Engineering Education

suggest new phenomena related to basic research or funda-
mental issues; third, providing independent third-party evalu-
ation, peer review, or critique of accident investigations
conducted by government agencies; and fourth, researching
accident investigation techniques and issuing research re-
ports with recommendations for the best possible accident-
investigation techniques. Development of software and tools
for accident investigation is also an area of interest.

The overall goals of the research program aim at improv-
ing safety in process plants by identifying the greatest risks
and then developing inherently safer processes and designs,
developing best-practice databases, and solving problems
identified by industry. The research goals of the Center
Systematically identifying the greatest risk in terms of severity
of consequences and probability of occurrence and prioritiz-
ing them
Systematically identifying projects that could be undertaken
by the Center and would most effectively address the risks
identified by risk analysis
Developing safer process schemes for the most common and
most hazardous processes; developing design concepts for
implementing such processes
Developing devices, systems, and other means for improving
safety of chemical operations, storage, transportation, and
use by prevention or mitigation
Improving meansfor predicting and analyzing the behavior of
hazardous chemicals and the systems associated with them
Current research activities include a reactive systems re-
search and teaching laboratory established for evaluating the
reactivity of chemicals and mixtures of chemicals, and to
obtain data needed to size relief systems for runaway reac-
tions. A reactive systems screening tool (RSST) exists and
operating procedures have been prepared based on two base-
case runaway reactions: methanol and acetic anhydride (tem-
pered) and hydrogen peroxide (gassy). The RSST studies
can be used for initial reactivity characterization and vent
sizing, as well as for a laboratory experiment in the under-
graduate unit operations laboratory.
Another research project underway is "Two-Phase Vis-
cous Flow Through Safety Relief Valves." Phase I of the
project includes a survey of the literature and evaluation of
state-of-the-art procedures for relief-valve sizing in two-
phase flow, verification of various theoretical models by
experimental data, and recommendation of design practices
for viscous two-phase flow through safety-relief valves. Phase
II involves experimental design for Phase III, which is the
experimental phase. The program includes CFD numeri-
cal computation and prediction of two-phase flow through
safety relief-valves.
Other research includes "Post-Release Transport and Fate
Summer 1999

of Toxic Chemicals and Their Mixtures." The objective is to
develop mathematical models that accurately represent the
transport and fate of chemicals as well as their mixtures
resulting from process-plant accidents. Some computer mod-
els are available that can be used to make these calculations,
but many problems are associated with their application,
including problems in handling polar substances and mix-
tures. Our objective in conducting this research is to address
this problem and to develop an approach that can be applied
consistently and uniformly.
Future plans for research include establishment of a state-
of-the-art reactive chemicals laboratory. The Center has ac-
quired and installed an Automatic Pressure Tracking Adia-
batic Calorimeter (APTAC) for reactive screening of chemi-
cal reaction compounds and mixtures. The APTAC can be
used for thermal analysis of solid or liquid chemicals or for
gas/liquid, liquid/liquid, gas/solid, and liquid/solid mixtures.
It can obtain time-dependent kinetic data and temperature
and pressure profiles for both open and closed systems. It
can also be used for process simulation of batch and semi-
batch reactions, fire exposures, emergency relief venting,
and physical-properties measurement. The resulting in-
formation can identify potential hazards and tackle key
elements of process safety design such as emergency
relief systems, effluent handling, process optimization,
and thermal stability.
The Center Steering Committee from time to time evalu-
ates proposals and ideas regarding future research projects.
Depending on the situation, funding for these projects is
sought from external sources or from internal Center funds.
For example, some of the projects currently under consider-
ation are
Corrosion-induced fatigue failure for moving parts
(correlation between corrosion rates, failure frequency, and
intensity of movement)
Methodologies on inherent process safety
Comprehensive database for equipment and component
failure rates in the chemical industry
Incident history database
Data integrity and compilation during engineering projects
Human factors research
Thermodynamic data for specific mixtures (e.g., 30% oleum)
Flammability limits and explosivity limits (both experimen-
tal and correlation)
Use of computational fluid dynamics to evaluate damage to
facilities based on knowledge of gas concentrations in
cloud, confinement, and dynamic response of structures
Passive explosive suppression in compartments for offshore
Safety protective data and linkage with fire-school activities
Fire suppression with environmentally friendly chlorofluo-
Continued on page 209

f 1% curriculum

Experiments With



The University of Sydney Sydney, New South Wales 2006, Australia

Engineering education traditionally places initial em-
phasis on "exposition," followed by "application,"
within the domain of a specific course. Subject mat-
ter is programmed so that the general and inclusive ideas of
the discipline are presented first, followed by progressive
differentiation in terms of detail and specificity. Often, in the
early years of engineering education, exposition and appli-
cations phases dominate the curriculum, while integration
receives scant attention. With the current urgency to provide
a well-rounded learning experience, skills in addition to
conventional engineering abilities are being stressed at all
levels in academia, including the early formative years when
the science content dominates the curriculum.
A typical engineering teaching plan includes
LI Exposition of scientific principles
El Exposition of engineering principles
El Acquisition of practical and theoretical skills
E Application of acquired skills and knowledge to solve com-
plex problems
Appropriate assessment and feedback then follow. A sys-
tems analogy version of traditional education in terms of the
stimulus, response, and feedback process is that the input of
teaching material, the input from students, the course goals,
and the outcomes all feed into the course itself (see Figure 1).
An important detail missing from this unidimensional ap-
proach relates to the multidimensional, distributed, and in-
teractive nature of learning (drawing from the systems anal-
ogy) and consequently the complex multivariable structure
of the educational process. Further, most of the relevant
application phase usually occurs toward the end of the de-
gree curriculum (in the form of design and thesis), often long
after the principles have been taught. If we recognize that all
aspects of learning should be interrelated, then educators
need to explore appropriate integrative learning tools to avoid

excessive fragmentation of curriculum and to foster interac-
tion among the participants.1131
This article highlights the dangers of excessive fragmenta-
tion in course presentation, especially in the early stages of
engineering education. A lack of relationship between courses
could also create rigid compartmentalization of knowledge.
Our attempts to encourage cooperative learning and integra-
tive reconciliation (systems analogy) between courses will
be discussed.

Early engineering courses rich in scientific content at-
tempt to introduce key principles and tools so that more
detailed and differentiated material may follow and provide
the scaffolding for further learning. But it is common to find
that little effort has been spent in creating links between the
courses. For effective problem solving, it is necessary to
have an integrated cognitive structure for flexible retrieval

Vincent Gomes received his MEng and PhD
degrees from McGill University and his BTech
degree in chemical engineering from IIT. He is
currently a Senior Lecturer in the Department of
Chemical Engineering and Program Manager
of the Australian Key Center for Polymer Col-
loids at the University of Sydney. His research
interests include polymer engineering, sorption-
reaction, and pollution prevention.

Tim Langrish received his BE degree in chemi-
cal and process engineering in New Zealand
and his DPhil at the University of Oxford. After
working as a Research Fellow with the Sepa-
ration Processes Service at Harwell Labora-
tory, he returned to the University of Canter-
bury in New Zealand and subsequently took a
position at the University of Sydney. His re-
search interests include drying technology and
fluid mechanics.

Copyright ChE Division of ASEE 1999

Chemical Engineering Education

and application of acquired tools. Thus, the convenient prac-
tice of minutely segregating a discipline into courses and
sub-courses is often insufficient for deep learning. Courses
offered in parallel or in sequence, served in bite-size chunks
for ease of digestion, often appear confined within water-
tight compartments that serve their specific purpose, with
little time left for forays into neighboring territories. Such
methods may be responsible for eliciting student responses
such as "...that was taught in fluid mechanics...are we sup-
posed to know it for heat transfer?"
Studies145" show that the early formative years in engineer-
ing education are crucial in en-
gendering a professional attitude
and weaning students away from GOALS
their high school attitudes. The
first- and second-year students INPUT COr
face a range of diverse subjects,
involving various faculties, with
seemingly minimal connections
between them. Students fail to
see any relationship between the Figure 1. Systems c
early courses and their chosen curricuL
professional discipline. Ideas,
concepts, and applications are commonly presented in the
context of a particular course without recognition of courses
taught in parallel. This practice frequently results in sub-
stantial segregation between courses without relational
mapping and may foster a disposition toward rote memo-
rizing, which in turn results in loss of associative learn-
ing in the initial stages.
A message that needs to be emphasized early on is that as
a body of knowledge, chemical engineering has structure
and form, is built on fundamental laws, concepts, empirical
observations, and data that we believe are self-consistent.
Our discipline is not an unrelated collection of a few thou-
sand equations put together to solve problems in a "cook-
book" manner. Students must be guided to avoid missing the
forest for the trees. They often do not see any relationship
between concurrent courses, but view them as isolated hurdles
to be overcome sequentially, and they perceive that lessons
learned in a particular subject will not come under rigorous
tests within the framework of a different course. This im-
plies that the provision of a "road map" of the discipline,
showing links between the different courses and how they fit
in, may be advisable. This aspect is currently being explored
in our department.
For efficient delivery of a large body of knowledge, mini-
mal overlap tends to exist between syllabi of different courses.
As a result, students continue to solve unidimensional, spe-
cialized problems tailored for a specific course. The post-
ponement in training to solve multifaceted problems contin-
ues for the major part of undergraduate education. Thus,
crucial connective links between different courses may re-
Summer 1999


um e

main hidden for the majority of students. The consequence is
a distorted view of real-life problems that have been mas-
saged in view of a specific course syllabus.
Meaningful learning is by definition relatable and
anchorable to established ideas in the cognitive structure.16]
Thus, it is not uncommon to find that the structure of the
discipline is unclear in the early stages of engineering educa-
tion, sometimes resulting in lack of motivation and interest.
It is often only in the final stages of a degree program that
advanced students have opportunities for integrating their
learning through thesis and design work involving challeng-
ing problems.
DMIN Rigid compartmentalizing
implies that equations and
OUTCOME methods memorized to solve
S-- typical problems for respective
courses would not be flexibly
available for solving "open-
ended, real-life" problems. A
review or refresher material
ogy of unidimensional
Sxperiencwithin a course may tempo-
rarily help retrieve "lost infor-
mation." Thus, integration is
important not only within the confines of a course (intra-
course), but also in relation to the discipline and courses
conducted in parallel. Therefore, following a detailed expo-
sition and application phase within separate courses, a mecha-
nism for integration of the subject matter is desirable.

In order to encourage course integration, three years ago
we started the practice of collaborating with other instructors
within the department to devise a single joint project that
would count for both courses. The final project is nor-
mally carried out in small groups, often with minor tech-
nical variations for each group, thus encouraging cross-
fertilization within and between the groups without du-
plication. The presentation of results in formal reports,
and often orally, provide further opportunity to improve
verbal and writing skills.
At the University of Sydney, we offer Material and Energy
Balances and Process Case Studies during the first year.
During the second year, Chemical Engineering Computation
is offered as a sophomore problem-solving course, involving
nonlinear equations, interpolation, least-squares, and numeri-
cal calculus, while the parallel Fluid Mechanics and Heat
and Mass Transfer courses focus on the fundamental prin-
ciples of corresponding transport processes and on equip-
ment design. The remaining courses during the first and
second years involve other faculties. We decided to set joint
end-of-course projects that would highlight the lessons learned
in each of the engineering courses during the first and sec-
ond years, and also to combine elements from each course to

solve a complex engineering problem.
Our course schedule permitted setting a joint project on
early process design (Flowsheet of a Bio-Refinery) for Ma-
terial and Energy Balances and Process Case Studies during
the first year, and on fluid flow analysis/optimization for
Fluid Mechanics and Chemical Engineering Computation
during the second year. Our objectives were
[E To attempt integration between two courses
[E To solve a nontrivial problem
EL To provide early analysis and synthesis experience
El To encourage team effort

Typically, the classes were divided into four- or five-
member groups composed of weak, average, and strong
students (based on their grade-point average). The groups

were drawn from concurrently offered courses; thus, there
were often a few students (6-8) who were not part of one or
the other course. These students were distributed into groups
so that there were at least two members taking both courses.
The management and task allocation were left for the group
members to arrange. Initial instructions to help the teams
function effectively were provided through preparation
sessions in which professional roles and responsibilities
were discussed and group responsibilities were produced
in written form.

Opportunities were provided for the group members to
meet during the tutorial hours and on their own. Every group
member was asked to assess their peers' efforts (on a scale
of 0-10) and to indicate the actual contribution of each group
member in a section on "who did what" attached to the final


In order to specify appropriately sized flowmeters for
monitoring the water flowrates in the cleaning-water distribu-
tion network at the Waikikamukau Dairy Cooperative, you are
required to estimate the flowrate in each pipe in the system,
shown diagrammatically in the figure. You must use a
systematic technique for this problem.

The total flowrate of 0.4 ms enters the system at point A.
Cleaning equipment draws off the indicated flows at points
C, E, F, G, H, and I.
Ignore the effects of differences in elevation. Water may
be assumed to have a kinematic viscosity of 10 6m's-', and
all the pipes are made of carbon steel.
You must use both a spreadsheet and a FORTRAN
program to carry out the iterative calculations, but you
must first do a hand calculation for one iteration of your
method. You must then include a printout of your
spreadsheet for this case and the results of the FORTRAN
program to validate your computer code. Record the time
that you spend on each approach (spreadsheet and
program) and comment on the ease of use of each
approach for this application.
You must document the spreadsheet fully and clearly, and
also include an attachment specifying the formulae used in
the cells together with the order of calculation.
For the FORTRAN program, each step in the program
must be clearly documented, and you must attach an
explanation of the order of calculations (a flow chart may
be a useful way of doing this). The formulae used at each
step of the calculation must also be specified.
Relationships between the friction factor, the Reynolds
number, and the relative roughness are given in Perry's
Chemical Engineers/Handbook and in Coulson and
Richardson, Volume 1. It is your responsibility to use
formulae appropriate for the range of Reynolds numbers
and data given.
While your submissions should be tidy, handwritten
attempts will not be penalized compared with typewritten
ones, provided that the material can be readily assessed.
You must present a summary page at the front of the
submission, giving your group number and a table showing
the following results for each pipe:

0.4 m/4s Pipe Data
@ @ Pipe Length Diameter
A B C No. (m) (m)
0.04 m3/s 1 100 0.3
m3/s 2 100 0.3
D E @ F 3 125 0.4
4 125 0.2
0.01 m3/s 0.11 m3/s 5 125 0.3
S6 100 0.3
7 100 0.2
G G H I 8 250 0.25
S9 250 0.3
0.08 m3/s 0.08 m3s 0.08 m3/s 10 250 0.2
11 100 0.3
12 100 0.2
1. Direction of flow
2. Average velocity
3. Volumetric flowrate
4. Pressure drop
SYou must explain all your calculations, document your FORTRAN program
and your spreadsheet, and hand in printouts of your spreadsheet, your
program, and the results from the program.

Analyses and Discussion Topics
1. Why is more than one iteration necessary in order to get a converged
solution, even with a problem involving only two pipes?
2. Why do problems with many pipes (and "loops" of pipes) take more
iterations to converge than a two-pipe problem?
3. What would happen if all the pipe diameters were doubled? Would the
flowrates be the same in all the pipes? If not, why not?
4. Is your final solution sensitive to the initial guess?
5. If all the flowrates into and out of the system were increased by 10%, would
all the flowrates within the system also increase by 10%?
6. How much different would the flowrates be (for the original inlet and outlet
flowrates) if the pipes were completely smooth?
7. Suggest globe valves for flow regulation and check valves for back-flow
prevention at appropriate points. If a centrifugal pump is installed for
supplying water from a tank and the outgoing flows are connected to
specified equipment, show how these affect your calculations.

206 Chemical Engineering Education

Chemical Engineering Education


submission. The lecturers and tutors monitored the indi-
vidual group members during the tutorial, and class hours
were assigned for the project. At the conclusion of the project,
team members were questioned on the technical aspects of
the project such that team members were individually held
responsible for the entirety of the project content. This input
formed the basis for awarding a final mark to an individual
student. The final mark was the team project grade modified
by up to twenty percent according to the results of the peer-
effort assessments and the interviews such that the average
mark for the individuals in the group was the same as the
team mark. Evidence whether the group process was work-
ing was noted during the tutorials themselves.
Apart from the initial instructions on how to function as a
team, support was offered to help teams function effectively
if it was needed. Otherwise the groups were encouraged to
gain autonomy and to promote internal communication with
a minimum of staff intervention. The students approached
the problem through internal coordination, sharing of com-
mon difficulties and insights, encouraging weaker stu-
dents, fostering a sense of responsibility, accountability,
and creation of memorable situations. The average dura-
tion of the shared project was about four weeks before
the termination of the semester and the projects were
typically assessed at fifteen percent of the total course

marks for each of the courses.
An example of one of our problem statements is summa-
rized in Table 1. The problem on pipe network analysis
using fluid mechanics principles, was intended to bridge the
gap between two parallel courses: one based on program-
ming and numerical methods, and the other oriented toward
engineering science. The main tasks for the problem in Table
1 were

[E To formulate the design equations to be solved
E To determine methods for solving the sub-problems and the
overall problem
E1 To plan a set of modules using a spreadsheet and a program-
ming language
[ To debug and execute the programs
E To modify the programs to test different conditions
Table 2 provides a guideline on the approach and consid-
erations in solving the problem posed in Table 1.

The exercise of "shared course projects" has now been
carried out over a three-year period for the first- and second-
year courses mentioned above. The problems selected, in-
corporating components for each course, were relatively
large in order to provide a greater challenge than those from


Solution and Programming Issues (students normally provide full
programs and spreadsheets with model equations employed)
Carry out hand calculation for a single step using the Hardy-
Cross Method (i.e., assign signed flow directions to the pipe
segment flows and set the sum of flow rates in each loop as
Choose the guessed values of flow rates in all pipe segments by
mass balances.
Include the Colebrook (or similar) and Hagen-Poiseuille
equations for calculating pipe friction losses. Check for flow-
regime and use the appropriate equations.
Include pipe expansion-contraction losses at each junction and
the pipe fitting losses.
Use a local Newton or bisection method for solving the implicit
Colebrook or similar equation.
Develop a flow chart showing the sequence of computations.
Input the program in modules; test and debug the program at
each stage.
Use comments to make programs readable.
Carry out one computer iteration and compare with hand
Implement full iterative computation for all loops. Modify and
test various cases.
Print out results.

Summary of Analyses and Discussions (students normally provide
numerical answers and analyses)
1. Iterative solution is necessary in order to obtain a converged
solution because, even though mass is conserved for an initial
guess, the sum of head loss around each circuit is non-zero. The

converged solution will ensure that mass is conserved and a zero
net head loss is obtained within each of the flow circuits.
2. A problem with multiple pipes requires more iterations because
in a multiple-pipe circuit, some pipes participate in more than
one circuit, and consequently head and flow corrections from
more than one circuit need to be applied until the flow rates are
3. If all pipe diameters are doubled in size, the converged flow rates
in each pipe would approximately remain the same. The small
differences in converged flows are due to the fact that the head
losses computed are not linear with respect to the pipe diameters.
4. The final solution should be insensitive to the initial guess. The
solution should be unique.
5. If all the flow rates into and out of the system were increased by
10%, all the other flow rates in the system would increase by
approximately 10%. Small discrepancies are due to the
nonlinearity of the equations.
6. If the pipes were all smooth, friction factors would be less for a
given Reynolds number; hence the head losses would be less.
Thus, flow rates would be slightly different because the head loss
dependence is nonlinear with respect to pipe roughness.
7. The connection of a supply pump and of outgoing flows will
require setting up head-loss equations in an outer loop and
incorporating the flow resistances of the associated piping and
connected devices. After the inner-loop flow computations are
balanced, calculations must be carried out for the outer loop,
including the pump performance characteristics, to satisfy the
constraints provided. Any flow excess or deficit must be
balanced along with the inner-loop flow distributions and solved

Summer 1999 20

a single course. A side benefit of the shared project was that
the combined time and effort provided by the lecturers and
tutors for the two courses helped optimize the total input,
and no additional time allocation was needed.
The students found the problems challenging and stimu-
lating, but not overwhelming. The groups functioned better
than satisfactory in achieving their objectives and few con-
flicts were noted. Conflicts related to communication were
resolved through encouraging the students to make contact
by phone and/or e-mail. Lack of responsibility by any indi-
vidual was penalized through a lowering that individual's
grade. The fact that group members needed to cooperate for
a common goal engendered a sense of belonging and a
degree of independence and responsibility throughout the
project. Our experience indicates that the majority of the
teams functioned without major complaints, knew what the
other team members were doing, and displayed a satisfac-
tory level of understanding.
The levels of difficulty were not the same for the projects.
For some projects the students had to select an appropriate
computing tool. In such cases, about two-thirds of the groups
opted for the spreadsheet, while a third opted for FOR-
TRAN. For the project described above, no such choice was
provided, and the groups were required to set up both a
spreadsheet and a program. The groups typically divided
the tasks among themselves to concentrate on particular
aspects of the problem and then combined efforts on the
more difficult parts.
The student experiences indicated varied styles of learning
and degrees of expertise in using these tools. A majority of
the students were found to be competent in using one of the
techniques, while only about a third was proficient in using
both. The exercise also highlighted the strengths and weak-
nesses of the tools used. The spreadsheet was faster in pre-
senting results and was relatively easier to debug but it
provided less flexibility and was tedious for variable defini-
tion and usage. Developing a program required greater disci-
pline and effort, but was more flexible for experimenting
and when it worked, it was more satisfactory in terms of the
learning experience.
A majority of the students noted that the project usually
took on a life of its own. They also said that it helped them
gain an understanding of how their fellow students think and
work: "I not only got to know how I think and solve
problems, but I also realized the false steps taken and
assumptions made by my partners." The students also
gained first-hand experience on how to cope with dead-
line pressures and peer review.
The students spent more time on the project than they
anticipated, expressed satisfaction on completing the project,
and considered it a memorable experience. Feedback from
course evaluations showed improved satisfaction with the

courses (about ten to fifteen percent greater satisfaction rat-
ings were noted). Carry-over from the experience to subse-
quent years was noted in the form of better appreciation of
the principles learned from computation and fluid mechan-
ics. Apart from this improvement in carry-over to subse-
quent years, anecdotal evidence from written course assess-
ments suggests that more students learned more indi-
vidual course material than before and better understood
the relationship between the integrated courses than pre-
vious students who experienced a more compartmental-
ized approach to teaching.
Traditionally, early engineering courses are taught with-
out invoking a major project, sometimes involving only a
minor project with limited time and resources. With pooled
resources and more challenging projects to offer, the mecha-
nism described has the potential of achieving better results
than the traditional approach. In addition, teaching and learn-
ing tend to be more rewarding and enjoyable due to the
higher degree of interaction between participants with effi-
cient use of resources.
The exercise provided a mechanism to synergize greater
curriculum integration, minimize compartmentalization,
strengthen cross-disciplinary learning, and reduce student
anxiety in meeting submission deadlines for two different
courses. The distributed learning process also encour-
aged variations in approach and ways of learning, such
that the students were not subjected to a single pre-
scribed mode throughout. The format of these projects
could also be derived from a larger research project to
enhance the challenge and to permit the direct flow of
research work into teaching.

Tackling challenging problems with group-based learning
can foster deep learning and understanding within the disci-
pline during the formative stages of education. The integra-
tive learning experience can help in the following aspects:
1 It provides students with the option of being involved in
structuring their own learning experience
1 Teachers act as mentors, facilitators, and resource persons
rather than as dispensers of information
1 Teachers discipline and interrupt students less and are less
constrained by a lack of time
1 Students develop both initiative and the skills needed to work
cooperatively with their peers
0 The format assists in developing communication skills
0 The projects encourage and enable students to critically
evaluate their own and each other's work
D Students develop confidence in tackling challenging
1 Less able students have the opportunity to reach the
competency level of their peers

Chemical Engineering Education


Through positive intervention in encouraging reconcilia-
tion between courses, we may avoid the ill effects of com-
partmentalizing courses and help integrate the acquired
knowledge of our discipline. Research on cooperative learn-
ing is summed up succinctly by Wells, et al.:171 " achieve
most effectively the educational goal of knowledge con-
struction, schools and classrooms need to become communi-
ties of literate thinkers engaged in collaborative enquiries."

1. Aronson, E., The Jigsaw Classroom, Sage, Beverly Hills, CA
2. Doise, W., and G. Mugny, The Social Development of Intel-
lect, Pergamon Press, New York, NY (1984)
3. Johnson, D.W., Educational Researcher, 10, 5 (1981)
4. McInnis, C., R. James, and C. McNaught, First Year on
Campus, Australian Government Publishing Service,
Canberra, Australia (1995)
5. Dalziel, J.R., and M. Peat, in Improving Learning, C. Rust,
ed., p. 272, Oxford Centre for Staff and Learning Develop-
ment, Oxford (1998)
6. Ausubel, D.P., J.D. Novak, and H. Hanesian, Educational
Psychology: A Cognitive View, Holt, Rinehart, and Winston,
New York, NY (1978)
7. Wells, G., G.L.M. Chang, and A. Maher, in Cooperative
Learning: Theory and Research, S. Sharan, ed., p. 1, Praeger,
New York, NY (1992) O

Continued from page 203.
Best-practice databases (e.g., for an ethylene plant, what
controls, procedures, and training are adequate)
Methodology to determine time-concentration effects of
various toxic materials and combination of these materials
Computational methods for determining fire resistance of
structural components in process facilities

Universities solve problems identified by researchers or
industry. For an applied engineering field such as process
safety engineering, the problems are usually identified by
industry. The approach is to develop effective mechanisms
for getting industry input and then taking a multidisciplinary
approach to solve the problem. To address the latter, the
Center has assembled a highly qualified team of experts who
have international reputations in fields ranging across reac-
tion engineering, inherently safe design, numerical analysis,
system and equipment reliability, applied probability, orga-
nizational structure and planning, non-destructive evalua-
tion, experimental fracture mechanics, materials testing, risk
assessment, exposure assessment, cost-benefit analysis, and
other areas of expertise.
The vehicle used to identify problems is based on two

Sununer 1999

factors: first, the Center actively seeks input from industry in
identifying process safety engineering problems that the Cen-
ter can help solve, and second, an annual symposium "Be-
yond Regulatory Compliance: Making Safety Second Na-
ture" is a vehicle to generate ideas and to identify problems.

In response to the changing role of chemical engineering,
chemical engineering departments must adjust and modify
their approach to education and research. The education
must include a comprehensive exposure to core courses inte-
grated with process-safety problems as well as a limited
number of specific process safety engineering courses. Chemi-
cal engineering departments must also produce an appropri-
ate number of MS and PhD graduates whose degree pro-
grams are focused on process safety engineering problems.
Also, to help our graduate students transition into industry,
the research we conduct should help industry in a practical
and immediate manner. This can be ensured by seeking
adequate input from industry as well as other stakeholders.
Public perception of the process industry is significantly
affected by process plant accidents. The significant societal
role played by industry is largely overlooked when cata-
strophic accidents occur. The best way to change that per-
ception is through adoption of proactive programs by both
industry and universities.

The authors acknowledge the support provided by the
Mary Kay O'Connor Process Safety Center and the Chemi-
cal Engineering Department at Texas A&M University. They
also thank colleagues at Texas A&M University for discus-
sions and research that assisted in the evolution of the ideas
and concepts presented herein.

1. Accidental Release Information Program, US Environmen-
tal Protection Agency, Washington DC (1989)
2. A Review of Federal Authorities for Hazardous Materials
Accident Safety: Report to Congress Section 112(r)(10) Clean
Air Act as Amended, US Environmental Protection Agency,
Washington, DC (1993)
3. Accidents Do Happen: Toxic Chemical Accident Patterns in
the United States, National Environmental Law Center and
United States Public Interest Research Group, Washington,
DC (1994)
4. Toxic Chemical Release Reporting, Community Right-to-
Know, Code of Federal Register, 40 CFR 372, US Environ-
mental Protection Agency, Washington, DC (1991)
5. Guidelines for Engineering Design for Process Safety, Cen-
ter for Chemical Process Safety, American Institute of Chemi-
cal Engineers, New York, NY (1993)
6. Kletz, T., Lessons from Disaster: How Organizations Have
No Memory and Accidents Recur, Gulf Publishing Com-
pany, Houston, TX (1993)
7. Kletz, T., Learning From Accidents in Industry,
Butterworths, London, UK (1988) 3

a classroom



An Example Debottlenecking Problem

and Outcomes Assessment Tool

West Virginia University Morgantown, WV 26506-6102

Chemical engineering educators are searching for out-
comes assessment measures to incorporate as as-
sessment-plan components in order to satisfy the
requirements of ABET EC 2000. Student and alumni ques-
tionnaires are always a staple, but an assessment plan should
not rely too heavily on these self-assessment instruments.
Faculty evaluation instruments are also necessary.
All chemical engineering programs have capstone experi-
ences such as the unit operations lab and chemical process
design. In these courses, students are expected to apply
knowledge learned earlier in the curriculum to solve com-
plex problems. The capstone design experience presents an
excellent opportunity for outcomes assessment since it re-
quires that material from several classes be synthesized and
applied. It can provide detailed information on what seniors
have learned in their earlier classes. In this case, student
assessment of the capstone design experience is being used
as a program-assessment measure. In particular, the techni-
cal content of the capstone design experience can provide

Joseph A. Shaeiwitz received his BS from the
University of Delaware and his MS and PhD from
Carnegie Mellon University. His professional in-
terests are in design, design education, and out-
comes assessment. He is co-author of the text
Analysis, Synthesis, and Design of Chemical Pro-
cesses, published by Prentice Hall in 1998.

Richard Turton received his BSc from the Uni-
versity of Nottingham and his MS and PhD from
Oregon State University. His current research
interests are focused in the area of fluidization
and its application to the coating of pharmaceu-
tical products. He is co-author of the text Analy-
sis, Synthesis, and Design of Chemical Pro-
cesses, published by Prentice Hall in 1998.
Copyright ChE Division of ASEE 1999

data on EC 2000, Criterion 3, outcomes a, c, and e (ability to
apply knowledge of mathematics, science and engineering;
ability to design a system, etc.; ability to identify, formulate,
and solve engineering problems).'"
One key to using the capstone design experience for out-
comes assessment is the measurement method. For over
twenty-five years, seniors in chemical engineering at West
Virginia University have been required to do a series of
projects in the two-semester senior design course, to submit
a written report, and to defend their results to an audience of
at least two faculty. A typical defense lasts one hour, with a
fifteen- to twenty-minute presentation followed by a ques-
tion-and-answer session. Students do these projects and de-
fend them individually, which is a unique feature of our
curriculum. The question-and-answer period is tantamount
to an individual tutorial. Students get immediate feedback
on their work and faculty can determine in great detail the
level of each student's understanding of and ability to apply
fundamental principles. Historically, students were not permit-
ted to ask questions of anyone while doing this assignment, but
in recent years, they have been permitted to buy consulting
from faculty for a minor grade deduction. This system ensures
that students ask only well-formulated questions and that they
do not try to "nickel and dime" a solution from faculty.
Oral examinations like this have advantages and disadvan-
tages as an outcomes assessment measure.121 The advantages
include an ability to measure student learning in great detail
through follow-up questions. Faculty can learn how and why
students obtain their results and develop an understanding of
students' thought patterns. This makes it easier to determine
if a reasonable result was obtained by accident from a series
of unreasonable procedures. Additionally, the immediate stu-
dent feedback is an excellent learning experience. Oral and
written communication skills are also developed. The major
disadvantages to this method are the faculty time required
Chemical Engineering Education

and the potential for student intimidation.
In this paper, the production of acetone from isopropyl
alcohol (IPA) is used as an example. The assignments are
described, followed by a brief summary of the issues in-
volved in the problem's solution. Then, a typical series of
questions asked of students and how learning is assessed
through the responses to these questions is discussed.

Figure 1 is a process flow diagram for the production of
15,000 tonne/y acetone from IPA. Most of the world's sup-
ply of acetone is produced as a by-product of reacting cumene
to phenol via the cumene hydroperoxide process. Acetone
used for pharmaceutical applications, however, is sometimes
produced from IPA due to the requirement of zero aromatic
impurities. The problem assigned is one of debottlenecking.
As will be discussed later, the ability to use this process for
assessment purposes is independent of whether a traditional
process design, debottlenecking, or troubleshooting is in-
volved in the assignment.
The assignment scenario is that a company has designed
this process to produce 15,000 tonne/y of acetone and equip-
ment has already been ordered. The process was designed
assuming an 8,000-hour year, but it has now been learned
that the process is to produce the desired yearly amount of
acetone in 6,000 hours, allowing the equipment to be used to
produce another product for the remainder of the year. There-
fore, a method to scale the process up by 33% must be found at

minimum equipment cost, particularly for special-ordered equip-
ment that cannot be returned to the vendor for replacement.
This problem was assigned in two parts. The first part was
to analyze the process up to T-401, the acetone scrubber, and
the second assignment was to implement heat integration
between the reactor effluent and the reactor feed (more de-
tails later) and to analyze the second distillation column, T-
403. Students were given equipment specifications, some
design calculations, and stream and utility flow tables. Much
of this information is available elsewhere,[3,41 and interested
faculty can contact either of the authors for additional infor-
mation. It should be noted that prior to these assignments,
our students receive significant instruction on analysis of
performance problems, i.e., problems in which the equip-
ment and input is specified and where the outlet conditions
must be determined.15'

A brief summary of the debottlenecking problem is pre-
sented here. This information in incomplete and descriptive
in nature. It is presented to provide background for the
discussion on assessment.
System Pressure Drop The details of the problem state-
ment make it clear that the ideal scale-up situation is for the
input to the separation vessel, V-402, to be at the same
temperature and composition as in the original design, just at
a higher flowrate. This fixes the pressure entering the vessel.
It is stated that pressure drop in the pipes is negligible;

V-401 P-401 A/B E401 R-401 E-402 E-403 P-402 A/B H-401 V402 T401 T-402 E-404 V-403 E405 P-403 A/B P-404 A/B T-403 E406 V-404 E-407 P-405 A/B E408
IPA IPA Feed IPA IPA Reactor Trm Reactor Reactor Phase Acetone AcetoneAceone Acetone Acetone Acetone IPA IPA IPA IPA IPA IPA Waste
Feed Pumps Feed Reactor Effluent Cooler Heater Fumace Separator Scrnbber Column Overhead Reflux Reboller Reflux Column Column Overhead Relux Reboiler Reflux Water
Drum Vaporizer Cooler Pumps Condenser Drum Pumps Pumps Condenser Drum Pumps Cooler

Rpnenature C

Figure 1. Process flow diagram for acetone production from isopropyl alcohol.
Stream numbers refer to Stream Table in Reference 3.
Summer 1999 211

therefore, at the increased flowrate (assuming incompress-
ible flow) the pressure drop through certain pieces of equip-
ment increases by a factor of 1.332. For gas flows, the effect
of pressure on density and its effect on the pressure drop can
also be included, but a trial-and-error solution is required. At
the specified scale-up, the pressure drop in the fluidized bed
reactor is constant. The result is that the front end of the
process is pressurized relative to the original design. Each
piece of equipment has a maximum allowable working pres-
sure that must be checked at the scaled-up design.
Feed Pump A pump curve and a curve showing the net
positive suction head required by the pump (NPSHR) curve
are provided for P-401 A/B. The system curve must be
plotted with the pump curve to determine if the maximum
allowable flowrate has been exceeded. If so, remedies such
as running both pumps in parallel (and ordering another
spare) or attempting to exchange these pumps for ones gen-
erating more head are possible. If the former solution is
chosen, it must be determined if there is sufficient NPSH
available for the new suction-side flow.
Heat Exchanger E-401 The effluent from this heat ex-
changer is saturated vapor. The steam temperature, and hence
the steam pressure, must be increased to accommodate the
increased flow. Since the outlet pressure increases, the outlet
temperature also increases.
Reactor The reaction is endothermic. In the reactor, en-
ergy is supplied by molten salt heated in the fired heater. The
fired heater only has 10% additional capacity. The simple
solution is to purchase an additional fired heater. A more
elegant solution is to use the reactor effluent at 3500C to
preheat the reactor feed, which can lower the heat duty on
the fired heater, even at scaled-up conditions. The fluidized
bed has about 50% inert filler, so the fraction of active
catalyst can be increased to handle the increased throughput.
But the amount of additional active catalyst required is much
less than 33% since the space velocity decreases at the
increased reactor pressure.
Molten Salt Loop The performance of the molten salt
loop must be analyzed correctly to determine the molten salt
temperatures entering and leaving the reactor at scaled-up
conditions. Both the energy balance and the design equation
for the reactor heat exchanger must be solved simultaneously.
The two temperatures plus the flowrate of molten salt are
unknown. One may be set to solve for the other two. In
practice, the flowrate would be controlled and the tempera-
tures would respond to changes in flowrate.
Heat Exchangers E-402, E-403, and E-408 At the new
inlet conditions, the outlet conditions must be determined
for these three heat exchangers. There is a restriction that
cooling water and refrigerated water flowrates can only be
increased by 20% due to velocity considerations.
Tower. T-401 The original specification is 1-in ceramic
Raschig rings. The tower will flood at 33% increased through-

put of both gas and liquid. One solution is to change the
packing to 1.5-in ceramic Raschig rings, 1-in Berl Saddles, or
1-in Intalox Saddles. All of these have lower packing factors,
and the saddles have similar interfacial areas per unit packing,
which would presumably lead to similar mass transfer rates.
Tower T-403 and Peripheral Equipment The tower will
flood at 33% scale-up. There are three possible solutions.
Because this tower has a small diameter, the trays have been
designed as a module to drop into the vessel's shell, so the
number of trays can be easily increased if the tray spacing is
decreased. This permits the reflux ratio to be decreased and
avoids flooding, which is an example of the trade-off be-
tween the number of stages and the reflux ratio. But the
effect of decreased tray spacing on tray efficiency should be
considered. The pressure of the column can be increased if a
pump is added after T-402. Increasing the pressure increases
the vapor densities, decreasing the vapor velocity and avoid-
ing flooding. Some combination of increased pressure and
decreased reflux ratio provides a satisfactory solution.
Perhaps the best solution is just to decrease the reflux
ratio. The distillate is a near azeotropic mixture of IPA and
water. The original design, as illustrated in a McCabe-Thiele
diagram given to students, has more trays than necessary in
an attempt to get closer than necessary to the azeotrope.
Decreasing the reflux ratio to avoid flooding only reduces
the top IPA mole fraction from 0.65 to 0.64! Once the reflux
ratio is determined, the reboiler and condenser performance
must be analyzed to determine the new outlet conditions.
Also, the reflux pump must be analyzed. For cases involving
an increase in overhead liquid flow, there may be insuffi-
cient NPSH for pump P-405 A/B, but the original design
uses very small diameter (0.5 in) suction and discharge lines.
Increasing the diameter of these lines to 0.75 or 1 inch easily
lowers the friction since the pressure drop is inversely pro-
portional to d5.
It should be noted that all aspects of basic chemical engi-
neering are included in this project. This is desirable when a
process such as this is used for program assessment.

Three scenarios of faculty-student interaction during ques-
tioning are presented as examples of how projects such as
this one can be used for outcomes assessment. All of these
scenarios are paraphrased actual responses from several stu-
dents. The reader should observe how the student receives
immediate feedback on results presented.
The first example is the absorber, T-301. The student has
presented the solution of increasing the water rate by 33% to
handle the same increased rate of gas to be scrubbed. The
student also suggests changing the packing from 1-in ce-
ramic Raschig rings to 1.5-in ceramic Raschig rings because
the decrease in packing factor allows the column to remain
below flooding. Consider the following exchange between
Chemical Engineering Education

student and professor.
Professor Why did you increase the waterflowrate by 33%?
Student To maintain the same liquid-to-gas ratio so I could
get the same separation.
Professor Why did you go to 1.5-in Raschig rings?
Student Because the packing factor is smaller. This lowers
the y-position ordinatee) on the flooding graph
enough so the column will not flood at the
increased gas flowrate.
Professor What about the interfacial area of the new
Student I really did not think about that.
Professor Well, let's think about it now. What happens to the
interfacial area?
Student (stumbles around for an answer)
Professor What has a smaller surface area per unit volume-
a bed packed with sand or a bed packed with
marbles ?
Student Marbles. So, I guess the surface area decreases
with larger Raschig rings.
Professor Will this have any effect on the absorber?
Student Yes, it will have an effect.
Professor OK. Will it help or hurt the separation?
Student It will probably decrease the separation.
Professor Correct. So, what would you now have to do to
maintain the desired separation ?
Student Well, I would increase the water rate more.
Professor Would this cause the column to flood?
Student I'm not sure since I did not do this calculation.
Professor Well, what is the trend?
Student Increasing the liquid rate would increase the x-
position abscissaa) on the flooding graph, which
moves the column towardflooding.
Professor OK. Let's assume that flooding again becomes a
problem. What else could you do to maintain the
desired separation without increasing the water
rate ?
Student (stumbles around for an answer)
Professor Let me ask the question differently. What else can
you change to make the separation easier? What
will increase the affinity of the acetone for the
Student Oh. The pressure and temperature could be


In what direction ?
Let's see. Lower temperature and higher pressure
favor the liquid phase.

Clearly, this student understands most everything one would
expect a student to understand about absorbers, but the pre-
sentation of the student's solution alone does not reveal this
fact. It only becomes clear as a result of the question-and-
answer session. When this problem was assigned, increasing
the size of the Raschig rings was the most common solution.
Very few students proposed using larger Berl or Intalox
Summer 1999

saddles, which have similar interfacial areas to small Raschig
rings. Upon questioning, the better students immediately
understood the problem and responded as illustrated above.
When students have trouble answering a question, as in the
case above on packing area and other ways to maintain the
desired separation, the question is always rephrased in such
a way as to provide a hint for the student.
This type of faculty-student dialog can reveal situations in
which a student arrives at a good solution without fully
understanding the reasons why it is a good solution. The
following is an example from a solution to scale up T-403.
Student (proposes lowering the reflux ratio in T-403)
Professor How did you arrive at the solution of only lowering
the reflux ratio?
Student I did the simulation on Chemcad and found that I
could lower the reflux ratio without really affecting
the distillate or bottom mole fractions.
Professor Based on what you learned in separations, does
this make sense?
Student I didn't think about it. I assumed the simulation
results were correct.
Professor They may well be correct, but we need to under-
stand why. So, does it make sense that lowering the
reflux ratio with the same feed and the same
number of trays does not affect the outlet concen-
Student No, I would expect the separation to be worse.
Professor So, what is special about this case that allows the
separation to be maintained at the lower reflux
The discussion now continues as the student is shown the
McCabe-Thiele diagram, which was provided with the
assignment but apparently ignored. This reveals that the
original column was overdesigned. There are several
stages approaching the azeotrope that provide very little
incremental separation. Therefore, fewer stages at the top
or lowering the reflux ratio do not appreciably affect the
distillate concentration.

Once again, only the question-and-answer session reveals
that a correct solution was presented without in-depth analy-
sis, perhaps without a detailed understanding of the reason
why the solution was correct. Both situations illustrated
above are examples of how student learning can be assessed
while students are simultaneously provided with individual
feedback on their work. It is a win-win situation.
The following is an example of dialog when an incorrect
solution is presented. In this case, the student has attempted
to draw the system curve on the pump curve graph (which is
provided) for the reflux pump, P-405 A/B, to determine if
the pump has sufficient head to handle the increased over-
head liquid flowrate for a solution that involves replacing
the existing trays while maintaining the same reflux ratio.
Student (Presents Figure 2; claims that doubling the
diameter of the suction and discharge lines is not

sufficient to operate at the scaled-up conditions and
suggests purchasing a new pump with a more
favorable pump curve or running both pumps in
parallel and purchasing another spare.)
Professor I do not understand your pump and system curve
analysis. Please explain it to me.
Student The pump curve was supplied. I plotted the system
curve. Since the desired flowrate is larger than the
point at which the two curves intersect, the existing
pump does not supply sufficient head at the desired
Professor From what we did in class, does it make sense that
there is so little effect of pipe diameter?
Student I didn't think about that. I did the calculation just
like we did it in class, and this is what I got.
Professor Let's try to analyze this in more detail. What
relationship does the system curve represent?
Student (stumbles around, cannot generate the desired
Professor The system curve has an intercept. What does this
represent physically ?
Student Oh. Isn't that the static pressure difference?
Professor For this case, yes. Now, what else causes pressure
Student Friction.
Professor And, what part of the curve represents the
frictional pressure drop?
Student (stumbles around for an answer)
Professor What is frictional pressure drop most significantly
dependent upon?
Student Velocity.
Professor Where is velocity represented on the graph?
Student Ummm. Oh. It is in the flowrate on the x-axis.
Professor OK. So how is frictional pressure drop related to
flowrate or velocity?
Student It goes with velocity squared.
Professor OK. So how is this shown on the graph?
Student It is in the parabolic shape of the graph.
Professor OK. So we now know that the intercept of the graph
is the static pressure change, and the curvature of
the graph is related to the frictional loss. So, let's
look at the frictional pressure drop. Let's pick the
point on the original (0.5-in) system curve for your
scaled-up flowrate. What happens to this point if
the diameter of the suction and discharge lines are
Student It should be lower on the y-axis.
Professor Which you show on this graph. However, how
much lower should it be?
Student Well, this is what I got.
Professor If you increase the pipe diameters, what does that
do to the friction?
Student (stumbles around for an answer)
Professor What is the relationship for frictional pressure
drop? Do you remember it?

Student (Writes the equation AP= 2pfLeqV2 / d on the

board, perhaps with some assistance. Most students
know the square relationship on velocity and the
inverse relationship on d, but not all can remember
all of the other terms.)
Professor So, what happens to the frictional pressure drop if
the diameter is, for example, doubled?
Student It is half the original value. This is what my graph
Professor Yes, that is what your graph shows, but are you
sure that you have the correct relationship? Does
anything else in that equation change if the
diameter is doubled?
Student Oh. The velocity decreases. I guess I forgot to
consider that.
Professor By how much does it decrease:
Student (Figures out from m = pAv that velocity is
inversely proportional to d4, so that the frictional
pressure drop is inversely proportional to d5.
Assistance and coaching may be required.)
Professor So, if the diameter is doubled, by how much does
the frictional pressure drop decrease?
Student Let's see. By a factor of two to the fifth. That's 32.
Professor So, if the frictional pressure drop decreases by a
factor of 32, how does this affect the graph?
Student The y-axis value decreases by a factor of 32.
Professor Are you sure? Remember the intercept.
Student Oh. The difference between the intercept and the y-
value decreases by a factor of 32.
Professor So, what does that do to the system curve?
Student It will be almost flat. So I guess the existing pumps
will work after all if the pipe diameters are
This exchange is an example of the tutorial nature of the
interaction. An erroneous result is analyzed, via careful ques-
tioning, to lead the student to a correct result. Through
questioning and coaching, the student "independently" dis-
covers the error made and determines the correct result.

original scaled-up
operating operating
condition condition
pump curve

S1 in diameter
Student solution

volumetric flowrate

Figure 2. Sketches of pump and system curves for P-405.
Solid curves are student result; dashed curve is correct
calculation for larger pipe diameter.
Chemical Engineering Education

Assessment results from this exercise are used in several
different ways, all of which "close the loop" on the assess-
ment process. The one-hour presentation and question pe-
riod provide students with immediate feedback. After all of
the presentations have been completed, class time is devoted
to project review. One or two of the best projects are pre-
sented. Faculty review the problem, noting areas where better
solutions could have been presented. Follow-up problems are
usually assigned. Sometimes these are assigned only to indi-
viduals, i.e., to students who did not do them correctly on the
project. In the case of this acetone problem, the heat integration
option was ignored by most student on the first project. There-
fore, it was assigned specifically on the second project.
An assessment report following each project is also pre-
pared and circulated to all faculty. It describes the project,
what types of solutions were expected and what types were
actually submitted. Areas where a significant number of
students did well are pointed out. For example, if a majority
of students responded to questions about T-401 as the stu-
dent in the example did, this would be specifically stated.
Areas where a significant number of students were found to
be deficient are also pointed out-if a number of students
did not think about the meaning of process simulator results,
simply accepting the results on faith, or if a significant
number made the error regarding frictional losses, this would
be specifically cited. In these cases, remedies to ensure that
future students are not deficient in the same area are suggested.
Faculty are expected to respond to the suggestions. Do they? In
general, our faculty do because of our culture supporting these
projects and due to the pressure we all feel not to have material
we taught show up as being deficient on these projects.

Outcomes assessment using oral presentations of capstone
projects can be implemented by making only minor changes
in how typical design classes are run. First of all, it is not
necessary to use a performance (debottlenecking or trouble-
shooting) problem such as the one described here, although
such problems lend themselves to this type of assessment
process. Since our students enter the senior year having
already completed a process design during their sophomore
and junior years,161 they are prepared for this type of assign-
ment. Asking probing questions in a typical capstone design
project can yield the same type of assessment information.
The best questions to ask are "why" and "what if." For
example, ask why the column was designed for a specific
reflux ratio. Was it chosen ad hoc, or was it based on an
optimization of the trade-off between number of stages and
reflux ratio? What if scale-up is required in the future?
Similarly, why were the reactor temperature, pressure, and/
or conversion chosen at the specified values? Were they
merely convenient values? Or, was the selectivity analyzed
Summer 1999

to determine conditions that maximize profit?
It is also not necessary for students to do projects individu-
ally for the presentations to be used for assessment purposes.
To implement this in a group of 3-5 students, interim progress
reports (which can be informal) are suggested. Students can
make a brief presentation to either a professor or a TA (who
would need some training in what to look for and how to ask
questions), and the students would then be expected to re-
spond to questions. Questions should be directed to indi-
vidual group members to avoid domination by one person.
The assumption should be that any student is prepared to
respond to any question, not just to the material presented by
that student. If a student is unable to respond, then another
student can be chosen or the question could be answered by
a volunteer. Assessment information would be gathered and
students would get feedback on their project while it is in
progress, which would probably improve the final product.
A project review is also desirable to close the assessment
loop. This should be done after all presentations have been
completed, preferably after all project reports have been graded.

Performance problems such as the debottlenecking prob-
lem illustrated here are a rich opportunity for outcomes
assessment, as are process design problems. Asking "why"
and "what if' type questions probes students' understanding
of fundamental principles. The oral presentation format pro-
vides students with immediate feedback, closing one feed-
back loop. Another way to close the assessment loop is by
project assignment review in class and/or follow-up assign-
ments. Feedback to faculty regarding students' ability to
apply the principles they are expected to understand closes
another feedback loop. The only real disadvantage is the
investment in faculty time for the oral presentations. If it is
believed that outcomes assessment and EC 2000 will result
in increased faculty time devoted to the undergraduate cur-
riculum, a key choice is how to invest this time. Questioning
students in oral presentations of capstone projects is one
potentially beneficial way to invest that time.

1. Engineering Criteria 2000, 3rd. ed., ABET, Inc., Baltimore, MD,
December (1997)
2. Prus, J., and R. Johnson, "A Critical Review of Student Assess-
ment Options," in Assessment and Testing: Myths and Realities,
T.H. Bers and M.L. Miller, eds., New Directions for Community
Colleges, No. 88, Jossey-Bass, San Francisco, CA, 69 (1994)
3. Turton, R., R.C. Bailie, W.B. Whiting, and J.A. Shaeiwitz, Analy-
sis, Synthesis, and Design of Chemical Processes, Prentice Hall,
Upper Saddle River, NJ, App. B 3 (1998)
4. Website for our majors:
5. Ibid, Reference 3, Section 3
6. Bailie, R.C., J.A. Shaeiwitz, and W.B. Whiting, "An Integrated
Design Sequence: Sophomore and Junior Years," Chem. Eng. Ed.,
28, 52 (1994) 0





For First-Year ChE Students

Northeastern University Boston, MA 02115

Batch and semicontinuous operations are often used

within the chemical process industries. For example,
the pharmaceutical, food, consumer products, and
pulp and paper industries regularly use batch processing.
Also, many chemicals are made on a semicontinuous basis.
According to a recent AIChE survey of the entry-level job
market, 60% of recent chemical engineering graduates are
entering industries that use either batch or semicontinuous
operations.'" Yet the engineering curriculum devotes less
than 10% of its instructional time to batch or semi-con-
tinuous operations. Further, if chemical engineers are
making an impact in the area of industrial batch and
semicontinuous processes, then it seems logical to intro-
duce chemical engineering students to those operations
early in their education.

The University of Nottingham is located in the Midlands
of Great Britain about 120 miles north of London. The total
University enrollment is approximately 18,000 undergradu-
ate and 3,500 graduate students. Chemical engineering is
part of the School of Chemical, Environmental, and Mining
Engineering. The School's enrollment is about 400 students,
divided between courses in chemical, environmental, and

* Address: University of Nottingham, Nottingham,
Nottinghamshire, Great Britain NG7 2RD
" Students arrive at Nottingham after a 2-year curriculum
similar to US freshman and portions of a sophomore college-level
curriculum. BEng and BS are similar curricula. MEng is similar
to a MS Chem Eng.; however, the equivalent thesis component,
listed as a research or design project, leans more toward
application and advancement of current engineering practices.


mining engineering, undergraduate (3-year BEng and 4-year
MEng**), and postgraduate (taught MSc and research MPhil,
PhD). In Great Britain, most university entrants have A-
level, which is equivalent to two years of preparation beyond
the high school degree.
As shown in Figure 1, the chemical engineering laboratory
at the University of Nottingham is impressive even by United
States' standards. The operating laboratory floor space is
128 by 32 feet (4,096 ft2) and has a 27-foot-high ceiling to
accommodate tall equipment. There are five additional labo-
ratories (approximately 1,000 ft2 each) located on both sides
of the main laboratory.
The University of Nottingham laboratory has a substantial
support staff, with one chief technician in charge of four

Ronald Willey holds a BS from the University of New Hampshire and a
PhD from the University of Massachusetts (Amherst). He joined the North-
eastern faculty in 1983. His teaching centers around the unit operations
laboratory and his interests include integration of process safety into the
chemical engineering curriculum. He is a registered engineer in the Com-
monwealth of Massachusetts.
Anthony Wilson holds BSc and PhD degrees in chemical engineering
from the University of Nottingham. With industrial and consulting experi-
ence in process control and batch process engineering, and active re-
search in both fields, he coordinates the school's research in computer-
aided process engineering.
Warren Jones holds BSc and PhD degrees in chemical engineering from
the University of Nottingham. He has a wide-ranging interest in both front-
end process and detailed plant design, developed initially through nine
years of experience with a major engineering and construction company.
Teaching responsibilities include several plant design courses and engi-
neering thermodynamics.
John Hills holds MA and PhD degrees from the University of Cambridge.
He worked in industrial R&D and taught in Africa before coming to
Nottingham. His research interests are in gas-liquid reactors and multiphase
flow, and he currently teaches chemical reactor design and chemical ther-

Copyright ChEDivision of ASEE 1999

Chemical Engineering Education

machinists, one electronics technician, and two full-time laboratory technicians who prepare
experiments and watch students during lab operations. Two or three graduate students are
also present during any particular laboratory day. Typically, ten to fifteen groups, composed
of two or three students each, are in the laboratory on laboratory days (Tuesday and Thursday
afternoons). Typical hours are from 2:00 to 4:30 P.M., but several experiments take longer.
The laboratory experience encompasses the first three years of Nottingham's chemical
engineering course. Usually, experiments during a particular semester follow the chemical
engineering lecture class schedule.121 First- and second-year students spend approximately
five sessions per semester in the laboratory, and the third-year students spend two sessions on

Figure 1. Nottingham's unit operations laboratory.

process control experiments. Those students who elect to take the BEng spend an additional
five sessions working on a large unit operations experiment such as vacuum distillation,
crystallization, liquid-liquid extraction, or filtration. These comprehensive experiments serve
as term projects, and generally, one or two full-time faculty are present in the laboratory
during a session. During the spring semester, the third-year MEng students are in the
laboratory in place of the BEng for the term projects.
As mentioned above, experimental work is done continuously throughout the student's
university career. The first-year students do simple experiments such as flow through pipes,
flow through orifices, flow through pipe systems, velocity profiles, measurements of heat
transfer coefficients, and mass/energy balances. These experiments are done concurrently
while the students are taking lecture courses in fluid mechanics, heat transfer, and mass/
energy balances. For example, the sequential controlled-batch plant experiment is performed
by first-year students who are also taking the basic chemical-process principles course. The
purpose of this brief summary is to allow the reader to make a rough comparison between
U.S. and UK degree schemes. Grose provides a broader perspective on engineering educa-
tion in the UK.[31

There is renewed interest in introducing students to engineering concepts through hands-
on experience. Many efforts, with good reason, are at bench scale. One example is the work
Summer 1999

... students

are able





on a


similar to

that which

they will

encounter in





such a rig

has no



of Hesketh,[41 which includes the reverse engineering of a coffee maker
(conceived while a post doc in Great Britain). Another effort is the ongoing
work by Perna and Hanesian151 at the New Jersey Institute of Technology.
They have taken several freshman engineering groups through a set of instru-
mentation/fluid-mechanics experiments with excellent success. But, again,
these experiments are primarily bench scale.
The experiment described in this paper is on a larger scale. It is based
around a 400-liter (about 100 gallons) vessel and uses steam at significant
pressure (100 psig). Thus, students are able to experience process equipment
on a scale similar to that which they will encounter in industry. Observing and
operating such a rig has no bench-scale substitute.

The plant schematic is shown in Figure 2. The metering Tank M is filled
with Feed A by Pump P and discharged into process Tank T via Valve A.
High- and low-level sensors are provided for both Tanks M and T. Tank T is
only partially filled by A, so Feed B is added until a Hi position is reached. At
the same time, the agitator starts, and once full, the contents of the tank are
heated by steam to 55C and left for 10 minutes. The tank is then cooled to
400C and the contents are discharged. Other sensors involved include those
for temperature and valve position (open or shut).
The first step in the system design (discussed in more detail later) is to
specify the sequence of actions required and then to identify the conditions
necessary for a particular action. For example, Pump P operates if Tank M is
low and Valve A is closed. To successfully manually operate or automate any
batch process, it is very important to fully appreciate the process sequence
logic-otherwise valves will be in the wrong position and pumps will be left
on when they should be off.
Students were given two objectives for this experiment:

Figure 2. Sequential batch processing experiment schematic.


Process Sequence

1. Accurately measure quantity of expensive
feed stock A charged to the reactor (which is
initially empty) from the metering tank.
2. The agitator starts and the reactor is filled to a
preset volume with cheaper reactant B.
Meanwhile, the metering tank refills, ready for
the next cycle.
3. Low-pressure steam is admitted to the jacket
surrounding the reactor, and heating stops
when the desired "high" temperature is
4. Stirring is continued alone for a set reaction
5. Cooling water passes through a coil in the
reactor and stops when the desired "low"
temperature is reached.
6. Reactor empties and stirrer stops.
7. Cycle completed ready to start at (1) again.

Operating Check-List

1. Tank M at Hi level (high-level
panel light is on)
2. Tank T at Lo level (low-level
panel light is on)
3. Valve A is closed (panel light is
4. _x Mains water (i.e., B) supply is
5. Valve B is closed (panel light is
6. _x_ Steam supply available
7. Valve C is closed (panel light is
8. _x_ Cooling water supply available
9. Valve D is closed (panel light is
10. Pump P is off (no sound)
11. Valve E is closed (panel light is
12. Valve F is closed (panel light is
13. x Lower supply reservoir filled
with A and ready

Signature of Group Member Verifying the
If any of these conditions are not correct,
notify Prof Willey or a technician before

Chemical Engineering Education




T ~

To obtain experience with a sequentially con-
trolled batch plant and to gain appreciation of the
advantages resulting from automated operation.
To perform heat balances on the heating and
cooling operations that form part of the batch
To help students understand the process sequence, they
are provided with Table 1, which can be read while they
follow the automated operation on the control panel. To
reinforce the importance of having all components in the

Figure 3. LabTech Vision"" screen created for automatic
control. A mouse pointer is used to initiate the On/Off

correct mode before starting a batch, they also receive a
Check-List (see Table 2).

This sequence is implemented in two ways. First, the
experiment is done under automatic control; the students
simply initiate the experiment by moving the on-off switch
located on a computer monitor (see Figure 3) to "on," using
a mouse. In the second experiment, students implement the
sequence and record data manually. They use thermometers,
stop watches, and toggle switches located on the computer
monitor, as shown in Figure 4.

Figure 4. LabTech VisionT screen created for manual con-
trol. Students use a mouse pointer to initiate the various
On/Off switches in correct sequence.

Summary of Steps Involved in Sequential Batch Experiments


SFill Tank M with Feed A

Proceed on
Condition That...
* Valve A closed

Initiated by

STank M low

Ended by Action
* Tank M high Open Valve E
SStart Pump P

* Discharge Tank M to Tank T Tank M high Start: Push Button Tank M low Open Valve A
Valve F closed
Valve E closed
Tank T low
Pump P off
* Add Feed B Tank M discharged into T Tank T high Open Valve E
* Start agitator Close Valve A
Start agitator
* Heat to 55"C Tank T high Temp at 550C Open Valve C
Close Valve B
* Wait for the duration specified Temp 550C Timer Start Timer
Close Valve C
* Cool to 400C Timer Temp at 400C Open Valve D
*Discharge Valve A closed Temp at 400C Tank T low Open Valve F
Valve B closed Close Valve D
* Process Comolete Tank T low Start of new batch Light warning lan

* Stop agitator
SClose Valve F



Summer 1999


.I t U,.ll.i l i i4 U CU ...... LU I. -V.&VI.y s C.g
The biggest challenge is programming
LabTech Control[16 to operate a batch pro-
cess. It has the capability to read analog
signals, to record data to diskette, and to
display data on computer monitors. It also
has the ability to perform PID and on-off
The analog signals acquired are four level
sensors, 0.2 volts when low and 2.5 volts
when high. Signals are also acquired from
thermocouples reading the batch tempera-
ture and the cooling water temperature. The
signals sent out are all digital (Hi or Lo).
They control the agitator (on or off), the
opening of Valve A (which drains Tank M),
Valve B (which controls the admission of
Feed B to Tank T), Valve C (which controls
the admission of steam for heating), Valve
D (which controls the admission of cooling
water), Valve E (which controls the refilling
of Tank A), and Valve F (which controls the
draining of Tank T). These valves have to be
turned on in the correct sequence for the
batch experiment to operate correctly. In
LabTech this is done by using two stages
triggered by the proper conditions. For ex-
ample, Stage 1 for Valve A is a 1-Hz stage
of 1 second. It is triggered open when the
students switch the on-off switch (located
on the computer monitor) to "on" by using
the mouse. Stage 2 for Valve A is a 1-Hz
stage of 7200 seconds (the experiment takes
about 3600 seconds) triggered on after the
Lo-Level sensor in Tank M indicates that
Tank M is empty. When Tank M is sensed
empty, Valve A closes and stays closed
for the duration of the experiment. Valve
B is triggered open (Stage 1) by the same
Lo-Level sensor. It is triggered closed
(Stage 2) when the water reaches the Hi-
Level sensor in Tank T. Table 3 gives the
sequence of events programmed through
LabTech Control using essentially two
stages-one to initiate and one to termi-
nate the desired action.
Desired batch temperatures are read from
a data file that is set up beforehand which
contain the desired batch temperatures. In
this case, the initial set temperature is 20C
(ambient) (a LabTech Stage immediately
reading at frequency of 1 Hz and on for only

0.5 second, thus only one point is read). When Tank T becomes full, the high
temperature (550C) setpoint is read when triggered by a Hi signal received
from the Hi-level sensor located in Tank T. After the tank reaches this
temperature, another LabTech Stage with a frequency of 1 Hz over the hold
time (in Hz) is used to read the same temperature (550C), followed by the
lower cooler dump temperature (400C).

For the first run, data are acquired to ASCII data files by using LabTech
Control data acquisition capabilities. Information recorded is: time of sam-
pling event, temperature of Tank T contents, and the outlet temperature of the
cooling water. These are acquired at a frequency of 0.0167 Hz (or every
minute). For the second run, data are acquired manually by recording tempera-
ture every minute into a laboratory notebook. In groups that comprise three
students, one student controls the sequence of the experiment while another

(mins) (mins)


Valve A Opens 0.15 0.00
Valve A Closes, Valve B Opens + Mixer 1.25 1.08
Valve B Closes, Valve C & E Opens 2.67 2.28
Valve E Closes 3.92 3.92
Temp. reached set point 55 C 13.13 13.33
End of Holding Time (10 Mins), Valve D Opens 23.15 23.37
Valve D Closes, Valve F Opens 32.90 37.50
Valve F Closes 42.55 46.50
END 42.57 46.50

Figure 5. Event times reported by a student for the two experiments:
automatic and manual control.


0 5 10 15 20 25 30 35 40 45

Figure 6. Temperatures, C, acquired by Lab Tech as a function of
elapsed times, mins, as presented in a student's report.
Chemical Engineering Education

student records temperature data, and the third student collects steam conden-
sate, calibrates the rotameter, and monitors cooling water flow through the
Figures 5 and 6 are results taken from a student report. Figure 5 shows
timed events for the two methods of operation. In this case, the student noted
in his report that the automatic control was faster (by about 10%) and there-
fore manual control is less efficient, and that over time this would equate to a
10% decrease in production rate.
Figure 6 shows a very smooth temperature profile acquired by LabTech
Control and later plotted by the student using Microsoft Excel. Series 1
represents the batch temperatures (in Tank T), and Series 2 represents the
cooling water outlet temperature. We see that it took about ten minutes to heat
the batch to 620C (set point was 550C-on/off control was used in this case)
and that the batch held at this temperature for the required ten minutes.
Cooling followed.
It is interesting to see how close the cooling water exit temperature ap-
proached the batch temperature. These first-year students had not had a
course in heat transfer at this stage, so they did not recognize how efficient the
cooling coil inside the vessel performed. They did note that the plots of
temperature acquired automatically were smoother compared to temperatures
acquired manually. Figure 7 shows a set of data collected manually as read
from thermometers by students. The "noise" observed in this particular figure
is comparably low for manually acquired data typically collected by students.
Students also made observations about automatic-versus-manual control.
One student noted that when the plant is operated under automatic control,
operators are free to do other vital jobs. He also noted that running the rig
remotely (over the Internet!) would maintain a safe distance for dangerous
The students were also required to do energy balances around the reactor
for both the heating and cooling operations. For the heating cycle, students
typically reported a heating efficiency of about 45% (calculated as [Heat









Tank Temperature


i Cooling Wat r



0 5 10 15 20 25
Elapsed Time, Minutes

30 35 40

absorbed by water]/[Heat released by condens-
ing steam]). Two explanations exist. One, that
the apparent steam condensate collected is large
because it included hold-up from previously
condensed steam; the second explanation,
which most students mentioned, was that the
stainless reactor itself had thermal capacity
and also required heating. Heat losses also
exists, but are relatively small in comparison.
What did we discover during the first few
runs? We missed telling the students to record
the inlet cooling water temperature; this has
now been included in the procedures. Hind-
sight is 20/20.

The experiment provides a worthwhile edu-
cational experience for relatively inexperienced
students. In particular, the advantages of auto-
mated operation are demonstrated. Further, stu-
dents are able to practice their IT skills and to
apply basic energy-balance techniques.

The authors acknowledge Thomas Holgate
and Tracy Wong, first-year students at the Uni-
versity of Nottingham, whose data were used
as examples in this paper. The authors also
acknowledge the assistance of Fred Anderton
in wiring the circuits required to set up the
automation of the sequential control unit.
Prof. Willey acknowledges Northeastern
University for permission to do a sabbatical
at the University of Nottingham during the
fall of 1997.

1. Graham, E. Earl, AIChExtra, 4, Sept (1998)
2. Jones, W.E., "Basic Chemical Engineering
Experments," Chem. Eng. Ed., 27, 52 (1993)
3. Grose, T.K., "A Yankee Engineer in Queen
Elizabeth's Court," ASEE Prism, 28, March
4. Hesketh, R.P., "Wake-Up to Engineering!"
Chem. Eng. Ed., 30, 210 (1996), and R.P.
Hesketh and C.S. Slater, "Hands-On Fresh-
man Engineering at Rowan University,"
AIChE 1998 National Meeting, Session 175-
The Freshman Experience; November (1998)
5. Perna, A.J., and D. Hanesian, "The NJIT
Freshman Experience: A Historical Perspec-
tive," AIChE 1998 National Meeting, Ses-
sion 175-The Freshman Experience; No-
vember (1998)
6. LabTech User's Guide, Laboratory Technolo-
gies Corporation, 400 Research Drive,
Andover, MA (1999) 3

Figure 7. Temperatures as a function of elapsed time as recorded by a
student group during the manual run.
Summer 1999

* |


e Oclassroom



University of Nebraska Lincoln, NE 68588-0126

Substantial writing assignments are required in an in-
creasing number of undergraduate technical courses.
They are usually intended to give the students prac-
tice in formal written communication, which will prob-
ably be an important part of the jobs most of them will
choose after graduation.
With these jobs in mind, the assignments tend to concen-
trate on teaching the students to write formal reports and
polished memos, two common forms of professional writ-
ing. Not all of the important professional writing that stu-
dents will do in future jobs will be in these commonly
emphasized formats.
This article discusses a semester-long writing assignment,
the class-related logbook, that concentrates on teaching the
process of using regular writing as a powerful professional
tool apart from the formal documents required on the job.
The assignment itself will be discussed in the format it has
taken after three successful years of implementation in
the first junior-level transport class for chemical engi-
neering majors where class sizes ranged from fourteen to
thirty-three students.
The need for this innovative logbook assignment evolved
partly from the changing undergraduate engineering curricu-
lum. Engineering students once took more hours of under-
graduate laboratory courses and were required to keep lab
notebooks. These notebooks were graded according to rigid
rules concerning completeness, organization, and clarity.
Currently, the trend in many schools is away from these
laboratory courses, for a variety of reasons including ex-
pense, safety, intensity of instructional resources, and
logistics of fitting all the current requirements into shrink-
ing credit-hour limits.
While some parts of the laboratory experience can be
replaced by computer simulations, writing as a tool of orga-
nization, planning, and discovery that was inherent in good

laboratory notebooks seems to have fallen from the curricu-
lum. The class-related logbook revives this use of writing as
a technical tool.
There are also sound pedagogical reasons for including
writing assignments in addition to the typical formal reports
and memos. Writing can be a powerful tool for information
processing: for assimilation, for organization, for clarifica-
tion, for analysis, and for synthesis. In short, all levels of
higher-order cognitive learning (as outlined in Bloom's Tax-
onomy,'11 for example), can be more efficiently achieved
with good personal writing skills as tools.
Since we want our students to be lifelong learners, this
habit of using writing as a problem-solving tool should be
taught and practiced in conjunction with their classroom
experiences, just as their other intellectual tools such as
calculus, computer literacy, and the ability to produce for-
mal documents are.
The idea of class-related logbooks sounds like a simple
and laudable way to develop valuable personal writing hab-
its early in the undergraduate career, but there are certain
pitfalls to successful implementation of the practice. This
article includes ways of avoiding three of the most com-
mon pitfalls: unrealistic professorial expectations, inad-
equate assignment design and presentation, and ineffec-
tual grading practices.

Jennifer I. Brand is Assistant Professor in the
ChE Department at the University of Nebraska-
Lincoln. She received her PhD in chemical en-
gineering at the University of California, San
Diego. and her BSE and MSE from the Univer-
sity of Michigan. She has worked as a chemical
engineer in private industry and at Oak Ridge
National Laboratory. She has taught graduate
and undergraduate courses in thermodynam-
ics, transport processes, biotechnology, airpol-
lution, and semiconductor processing.

Copyright ChE Division of ASEE 1999

Chemical Engineering Education

The class-related logbook assignment consists of three
required sections: a journal, chapter outlines, and reference
pages. (Students have the option of adding other sections for
their own use, but seldom do.) Each section has an indi-
vidual function and format. When combined, the whole log-

book is an integrated study tool
demonstrating the interrelation-
ship among different compo-
nents. The logbook is graded
and makes up a little under ten
percent of the student's grade,
enough to insure that they do
the assignment but not enough
that they overemphasize it to
the detriment of their other
learning tasks.
Table 1 (next page) shows an
actual class handout given to
the students explaining the as-
signment. This written defini-
tion of the logbook assignment
includes general guidelines for
informal professional writing,
instructions for each of the sec-
tions, and examples of accept-
able journal entries. Students
are advised in writing, via this

... writing as a to
planning, and di
inherent in good la
seems to have
curriculum. The cl
revives this use ofw
This article discuss
writing assignment
logbook, that c
teaching the proce
writing as a power
apart from the f
required o

document, that while the

logbook is theirs, it will be a "semipublic document," and
parts of it may be shared in a professional manner with the
class as a whole or with other academicians. They are ver-
bally assured that this public sharing will not include the
explicit identification of individuals without their consent
and any requests for confidentiality will be honored.
The purpose of the journal is threefold: to encourage ha-
bitual writing as an organizational tool, to teach the use of
chronological records as measurements of progress and as
indicators of patterns, and to give students a regular opportu-
nity to communicate with the instructor concerning the course
and their progress. Anything relevant to the class can be
included: their insights, their questions on procedure or con-
tent, their personal methods or circumstances that legiti-
mately influence their class performance or their learning
experiences. Honesty is required (although they are not re-
quired to bare their souls). Special assignments are required
throughout the semester to help structure the journal, as will
be discussed below. Instructor feedback on the journal is the
key to success in developing the good writing and communi-
cation habits the journal is designed to foster.
The purpose of the chapter outlines is to demonstrate to
the students the strengths of this traditional and effective
study tool-one that seems to be falling by the wayside in
popularity without being replaced by anything that has been
Summer 1999

demonstrated to be as effective and as simple. Students who
stay current on their chapter outlines will read the assign-
ments in a timely manner and will usually read more thor-
oughly. Instructor feedback on these outlines is a valuable
tool in helping them hone their abilities to prioritize and to
understand how others have organized material. The out-
lines themselves are useful tools for studying for tests and
for taking open-book tests.
l of o, The reference pages are
ol f org z on, the students' own organiza-
scovery that was tional products, tailored to
boratory notebooks their individual needs. These
fallen from the pages contain the highlights
zss-related logbook of the key concepts pre-
ng as a technical sented in the course, as well
Sas any useful information
. from other courses. A typi-
;es a semester-long cal set of reference pages
t, the class-related might include a page of im-
oncentrates on portant dimensionless num-
s of using regular bers, with defining equations
Sof using regas well as a few words about
ul professional tool the physical meaning and
normal documents applications, a page or two
in the job. of named equations and key
definitions, and a page of
"reminders" with a few
facts from a math course or useful variables such as the
viscosity of air and water at standard conditions. There
may also be a brief annotated bibliography or a list of
particularly useful reference sources.
Unlike the journals, which are chronological, and the out-
lines, which are dictated by someone else's organizational
style, the reference pages are an opportunity for the students
to learn how they can most effectively organize material for
themselves. Supplemental material (usually from previous
classes) is often included in the best examples of these
pages. They are also useful tools for studying and taking
tests. The reference pages are usually quite compact, and
therefore the students will find them to be the most useful
summary of the course. They are often used to review key
concepts when students take subsequent courses or when
they study for comprehensive exams, such as profes-
sional licensing exams.

Pitfall I: Unrealistic Professorial Expectations
One of my colleagues tried using my logbook assignment
and gave up a few weeks into the semester. He found that
the good students will keep good logbooks, but similarly,
the students who didn't do the other assignments kept
bad logbooks if they kept them at all. He concluded that
the logbooks weren't serving his purpose of making the


What? You will be required to keep a logbook for this class. The logbook will be yours, so you have some freedom in what you put in it and how. It is, however, a
"semipublic document," so there are some requirements and guidelines. I will read it on a regular basis and we may share parts of it formally in class. I may make
copies of part or all of some of the logbooks for use in the future.
Why? The purpose of this logbook is communication. It is an ongoing written progress report of your work. It will let me know how the class is progressing; it will let you
see your progress and may help you in seeing patterns to accelerate or ease that progress. Logbooks are also powerful tools in the real world (see Clifford Stoll's
The Cuckoo's Egg, a true thriller where a physicist cracks an international computer spy ring and convicts the spies because his logbooks are used as
documentation in court).
How? A three-ring binder with dividers?

Guidelines and Requirements
It must be legible and understandable.
The writing should be professional but not necessarily polished. Be clear, concise, specific, and accurate. Be natural, but not too relaxed. (Do not use language you
would not use in front of a significant other's parents the first time you meet them, for example.) Grammar need not be perfect (Contractions and sentence fragments
are okay.) Active voice is preferred.
It should be scrupulously honest. (Recall that honest does not necessarily mean exhaustive. Everything you say should be true, but you needn't say everything.) You
are not being graded on how easily or independently you pick up material outside of class or whether or not you like the subject or class.
Every entry should be dated
The logbook will have at least three sections: the joural, the chapter outlines, and the reference pages.

The journall This is the most free-form section. There will be some specific questions and topics assigned for discussion, and these must be addressed in a
professional manner in the journal, within the time frames specified. They should also be clearly marked. In addition to these assignments, anything that relates to this class
belongs in this section. It should not contain irrelevant thoughts, but tangential thoughts are okay. How you study for the class, what works well, what doesn't, why and
how it works (or doesn't) can go in; with whom you study is okay; why you couldn't study (not excuses, but reasons); what was good or bad or frustrating or boring or
interesting or hard or easy about the day's class or homework problems or reading assignment (The more specific details, the more useful such entries will be for both you
and me.); a supplementary source you found particularly useful; a connection with something in another class or some personal experience; a worked-out example or
scratched notes on problem approaches might be included. You will find it useful to summarize periodically. An example of a journal is attached.
Chapter Outlines This is your personal annotated index of the important parts of the textbook. Construct it in a way that is useful to you. You may want to
emphasize points that were difficult for you, de-emphasizing the "intuitively obvious" subjects, even if those subjects are treated extensively in the book. Cross-reference
ideas and equations to pages in the text. You may also want to cross-reference pages in your notes or other materials. A page or two should be enough to hit the important
ideas in each chapter. Keep in mind that properly constructed outlines can be very useful study guides and invaluable for taking open-book tests.
Reference Pages This section will contain things like a bibliography, the list of named equations, key concepts, charts, etc. Organize this section in a way that is
logical to you, not just by chronological order in book or course notes. This section, too, can be invaluable for taking open-book tests.

A Sample journall (Based on Student Journals in a Fluid Mechanics Class)
9/15-worked this afternoon with Scott C. and Jason B. for two hours on problem 3-15. Brick wall! After dinner, reviewed lecture notes for 9/9 while waiting for the
9/16-No clue what that lecture meant! When can you assume what for boundary conditions? It seems so arbitrary. How do you know where to put the origins? And,
of course I could have answered any other class question today, except the one she asked me! How does she always pick the part of the reading I didn't get to to ask me?
ASSIGNED DISCUSSION "Write directions for designing and making an Egyptian water clock. The explanations of what you are doing and why should
be clear, accurate, and concise, and easily understood by a bright 12th grade science class. Sketches and equations may be useful." [...Two pages of
discussion omitted from logbook example...]
9/17-No way am I going to torture myself with that stupid stuff on a Friday night!
9/20-I just know there will be a quiz tomorrow. Always is when we have big assignments due in thermo. Will it be vocabulary or math? I only have time to study one!
Scott and Jason came over and we worked for about four hours on 3-11. Another hour just getting the details on 3-15. Too much work! And where are the physical
properties at that temperature? I just want to stop pollution, not write differential equations. So I'll review vocabulary lists from Chapter 1-3 (in my reference pages).
9/21-1 guessed right. Vocabulary quiz. I think I did okay. At least I finally have the difference between continuity and continuum down. Karyn said she and Skylar
only needed three hours for the whole homework set. They used White's book (on reserve in the library) because it has a good summary of when to apply boundary
conditions and the viscosities and densities are in Perry's! Wish I'd known. Jim asked how to know when to use which boundary conditions. "Experience" was the
answer-so what do we do, take the class five times? Still, when we broke into small groups to do b.c.'s for the examples, I got most of them right by the end of group time.
Small group work seems like an effective way to "experience" stuff like this.
9/22-So if there is a no-slip condition at a wall, and you can get non-symmetric velocity profiles with plates moving at different conditions, is this why you get those
divots in the cake mix next to the beaters? And, come to think of it, why does the cake batter always want to climb up the beaters? And why don't all the bubbles coalesce
during baking? (Baked a cake for Chris's birthday tonight.)
ASSIGNED DISCUSSION "Record all valves you use for a twenty-four hour period. Include a description of the valves, what they were used for, what
flow rates, what kind of valves they were, what conditions of service they see, what they are made of, special design features, or anything else relevant.
Evaluate the selection of the particular valves for the particular applications. Give preferred alternatives, if any." [...Valve list and discussion omitted from
logbook example...]

224 Chemical Engineering Education

students come to class prepared for lecture. This is not an
unexpected result.
A class-related logbook is not a panacea for student learn-
ing or attitude. Its purpose is to help the student develop
learning tools not specifically emphasized in other parts of
the curriculum. For instance, assigning chapter outlines will
not make the incorrigible student read the chapter, nor will it
be the only reason the good student does a reading assign-
ment. The main virtue of the outline assignment is that it
shows the student a proven way of organizing challenging
new material as a step to learning it. Whether or not this
particular material challenges this particular student is irrel-
evant. When the student eventually does encounter challeng-
ing material, he should have various tools, such as outlining,
in place so that the new content, not the concurrent develop-
ment of the learning tools to master the material, will be the
student's task at hand. The reasonable expectation here was
not that all students would come to class prepared because
logbooks were assigned, but that students who kept log-
books would improve their learning skills.
Another reasonable expectation for the instructor is that
communication within the class will be greatly improved,
especially if the instructor grades the journals encourag-
ingly, as discussed below. The positive benefits of good
student-instructor communication hardly need extolling here.
The instructor will have great quantities of class-related
information from the students for a surprisingly low in-
vestment of his time. In addition, that time can be sched-
uled at his convenience.

Pitfall II: Inadequate Presentation of the Assignment
To the students, the idea of "free-form" writing in a tech-
nical class sounds vague, confusing, and since it is a
graded assignment, more than a little frightening. The
instructor has three powerful tools to overcome student's
discomfort: explanation, examples (modeling), and spe-
cific, focused mini-assignments. The handout in Table 1
demonstrates the use of each.
Collecting and commenting on the assignment frequently
at the beginning of the term, as well as continued modeling
throughout the first weeks of class, is a valuable use of time.
As the students become more comfortable with the assign-
ment, their anxiety will subside and the logbook will take
very little class time the rest of the semester. The modeling
in the initial part of the class usually is most successful if it
draws on good examples from other students' logbooks, as
well as from class projects (it takes about ten minutes to
demonstrate how to outline a chapter from scratch). The
mini-assignments are usually designed to integrate previ-
ously covered material and real-life experiences or to sum-
marize and integrate recently studied material.
After a few summarization assignments, students often
begin summarizing periodically on their own, an indication
Summer 1999

that their writing tools are developing and becoming a habit.
A particularly useful integrative summary is asking the
students to predict what will be on an upcoming exam.
Another is to ask them to design a flowchart for problem
solving, based on their own problem-solving processes.
Comparing their flowcharts from the beginning and the
end of the term can be most instructive for both the
students and the instructor.
Pitfall III: Grading and Instructor Feedback
Grading must reflect the process, not the product, espe-
cially early in the semester and in grading the journals. Some
specific grading guidelines for the three parts of the logbook
follow. In general, students should be allowed the freedom
to make mistakes and should not be penalized for originality.
Having said that, leaving mistakes uncorrected or not revis-
iting erroneous logic by the next time the logbook is re-
viewed is a legitimate reason for lowering grades. Some of
the more specific detailed assignments may entail special
grading criteria such as completeness and accuracy, which
are mentioned at the time of the assignment.
Logbooks are graded on a ten-point scale each time they
are collected, which is twice during the first three weeks
of the term and then irregularly at two-to-four-week in-
tervals thereafter. The journals are worth six to eight of
the ten points. Chapter outlines and the reference pages
are worth two points each early in the term, later decreas-
ing to one point each.
Since chapter outlines and reference pages are good tools
for studying and test taking, the students usually need little
grade incentive to do well on those sections after the first
test. The journals are more heavily rewarded by grades to
encourage the students to develop the habit of making the
intellectual efforts required to produce good journal entries.
The journal should be graded encouragingly, especially
early in the semester. This assignment is strange to students,
who are usually more accustomed to worrying about "what
the professor wants" rather than how to acquire the lifelong
learning skills the journal promotes. Since this journal is to
encourage the habit of informal professional writing, it should
be rewarding, not intimidating.
The actual value of the early grades for journals should be
based on whether specific assignments are carried out and
whether a good faith effort is being made to keep a journal
according to the guidelines. As the semester progresses,
expectations for the journal entries will be raised. Later in
the term, grades may reward reasonable attempts for seeking
answers, not just wondering about things. By the end of the
term, the students should be able to attempt answering, at an
appropriate level, most questions that they pose for them-
selves. These attempted answers should be reasonable, logi-
cal, and accurate, at a level appropriate to the students'
Continued on oase 231.


M laboratory


For The Fluid-Mechanics and Heat-Transfer

Laboratory Class

Universidade do Porto Rua dos Bragas 4050-123 Porto, Portugal

Fluid mechanics and heat transfer are important sub-
jects in undergraduate courses in chemical engineer-
ing, and surely there is no danger of overemphasizing
the importance of performing simple illustrative experiments
that the students can fully comprehend. A wealth of demon-
strative experiments are available commercially in kits, but
they tend to be expensive and leave the user in some form of
dependence on special spare parts in case of breakage.
The experiments described in this paper are cheap to build
and rely on materials and instruments readily available in
most engineering departments. The equipment needed is

Manuel A. Alves graduated in chemical en-
gineering from the University of Oporto in
1995 and immediately began teaching as a
Demonstrator in the Chemical Engineering
Department there. He became a Teaching
Assistant in 1996. His research interests
are in fluid dynamics and applied thermo-

Alexandra M.F.R. Pinto graduated in chemi-
cal engineering from the University of Oporto in
1983, received her PhD from the same univer-
sity in 1991, and is now an Assistant Professor.
She has taught courses in heat and mass trans-
fer and ChE Laboratories, and her research
interests are in fluidized bed combustion and in
the hydrodynamics of multiphase flows.

Jooo R.F. Guedes de Carvalho graduated in
chemical engineering from the University of
Oporto in 1971 and received his PhD from the
University of Cambridge in 1976. He is Profes-
sor of Chemical Engineering at the University
of Oporto, and his research interest are in
multiphase flow and associated problems of
mass and heat transfer.

Copyright ChE Division of ASEE 1999

An electrical oven
A digital millivoltmeter
Two thermocouples
A fan
A viscometer
Two ball valves
Plastic beakers
An anemometer
and some pieces of metal and nylon rods and acrylic tubing
that can be machined in half a day in a rudimentary work-

Most students will be familiar with, or will easily under-
stand, a wetted-wall column, but from our experience, few
will have come across the concept of cylindrical bubbles.
Yet, these are easily formed during continuous bubbling
in narrow bubble columns if the gas flowrate is increased
sufficiently, and also in vertical boiler tubes if the heat-
ing rate is high.
An easy introduction to cylindrical bubbles is afforded by
means of a simple experiment in which a long and narrow
acrylic tube is initially filled with water to within a few
centimeters of the top. A stopper is then used to close the
tube, before turning it upside down. A cylindrical bubble
will be seen rising up the tube (see Figure la), and its
velocity, U, is easily determined by timing the rise along a
given height. If the experiment was repeated with tubes of
different diameters, D, it would be seen that

U = 0.345 (gD)12 (1)
where g is the acceleration due to gravity. (The same type of

Chemical Engineering Education

experiment can be performed to show that U is independent
of bubble length.)
Equation 1 is valid for cylindrical bubbles in liquids of
low-to-moderate viscosity, which according to Wallis111 cor-
responds to the criterion

Nf = (gD3)12/v > 300
where N, is the dimensionless inverse viscosity and is the
kinematic viscosity of the liquid.
The experiment we propose may be seen as a variation of
the one described above. If a cylindrical tube is completely
filled with liquid and a stopper is used to close it at the top,
and if the stopper at the bottom of the tube is removed, a
growing gas slug will be seen rising up the tube core while
liquid will continuously discharge at the bottom along the
wall (see Figure Ib).
The volumetric balance of gas and liquid flowing through
any cross section of the tube requires that

Q=-(D-28)2U (2)
where Q is the volumetric flowrate of liquid running down
along the tube wall, and 8 is the thickness of the liquid film.
If 8 / D << 1, the curvature of the liquid film can be ne-
glected and Nusselt's analysis for film flow
is known to give1m1

q g 3 (3)
where q = Q / ED is the liquid flowrate per
unit wetted perimeter. This equation is de-
duced for laminar flow, which is normally
observed when

4q U(D- 25)2
Re = < 1500 (4)
v vD
where Ref is the film Reynolds number.
Substitution of Eq. (3) into Eq. (2) leads

4 gD 83
U (5)
3v (D- 2)2
and with U from Eq. (1), we get
53 8 v
(D 2)2 3.86 (gD)1/2 (6)
It should be remembered that this equa-
tion is valid only if Nf > 300 and Ref < 1500.
For given values of D and g, Eq. (6) is
shown to relate 8 with v.
In film flow, a more general relationship
between the dimensionless film thickness,
4, and the film Reynolds number is ob-
tained if the definition of Re, is substituted
Summer 1999

in Eq. 3,

6 N21 /3
= -N3 = 0.909 Ref (7)
(valid only for laminar flow) as pointed out by Wallis.111

Experimental Work
A 1.5-m length of 19-mm i.d. acrylic tube is adapted to
one side of a 3/4" ball valve, the other side of which is fitted
to one end of a 1-m length of the same tube, and fitted with
another ball valve at the other end. The resulting column is
aligned vertically above a plastic bucket, as shown in Figure
2, and a stopper is placed at the bottom before filling the
column completely with liquid (a detailed drawing of a
nylon adapter used to connect the tube with the valves is
shown in Figure 2).
The valve at the top is then closed and the stopper at the
bottom is removed to let a growing gas slug form and rise up
the tube. The slug will be allowed to rise freely until its nose
is some 0.3 to 0.4 m above the ball valve, at which time the
valve will be suddenly closed and a plastic beaker placed
(simultaneously) right under the column. The liquid col-
lected in the beaker will then be that making up the film
running down the tube wall over the length H, measured

(or 32 mm)


Figure 2. Experimental setup.



(or 11/4")


Air in
Liquid out

Figure 1. (a) Slug rising in a
closed vertical tube; (b) film
flow around a growing slug.

between the bottom of the column and the ball valve. The
thickness, 8, f the liquid film is determined from the volume
of liquid collected, V, through

V = H[D2 -(D 2 )2 (8)

It is a simple matter to repeat the experiment with liquids
covering a range of viscosities (we used glycerol solutions
with viscosities up to 15x103 kg/ms). A 32-mm i.d. column
can also be used, with 1 1/4" ball valves (note that the
internal diameter of the ball valve must be exactly the same
as that of the column). Measurement of slug velocity is also
simple through timing of the passage of its "nose" between
two marks about 1m apart.

Results and Discussion
Results obtained by our students are plotted in Figures 3
and 4, and the agreement between theory and experiment
can be seen to be excellent (average deviation in 8 less than
50 9Im, or about 5%).
Although the experimental technique is rather crude, our
experimental points fall closer to the theory than those re-
ported in Figure 11.8 of Wallis.111

Pedagogical Comments
The study of laminar flows is an important part of a fluid
mechanics course. The two most common experimental il-
lustrations are laminar flow in a tube (Poiseuille's formula)
and free settling of a sphere in a viscous liquid (Stoke's law).
Film flows are an important class of laminar flows (e.g., in
lubrication, wetted-wall columns, and filmwise condensa-
tion), but they are not normally illustrated experimentally.
This experiment provides a vivid illustration of the theory of
laminar film flow and, as an additional bonus, it combines it
with a very simple analysis of two-phase flow

10 -


Eq.(6) D=19nmm
Exptl. D=19 mm
---- Eq.(6) D=32mm
O Exptl. D=32 mm
0.1 I I I
1 2 5 10 20
vxl10 (m2/s)

Figure 3. Film thickness as a function of the kinematic
viscosity for D=- 9mm and D=32mm. Comparison between
Eq. (6) and experimental results obtained by students.

The actual demonstration is deceptively simple for the
students. They only have to remove a cork and a few seconds
later close a valve while simultaneously placing a beaker
under the column. But from our experience, the interpreta-
tion of the results-namely, the interplay between upwards
gas flow and downwards liquid flow, (with volume conser-
vation (!))-is very instructive. Invariably the students are
amazed when they find the close agreement between experi-
mental and theoretical values of 8.

A metal rod (typically 0.15m to 0.25m long and 15mm to
30mm in diameter) is initially heated in an oven to around
900C, and then suspended from two thin wires with its axis
horizontal, as shown in Figure 5. A thin sheathed thermo-
couple is then introduced into a hole with a diameter only
slightly larger than the thermocouple and drilled near the
axis of the rod. This thermocouple is connected to a refer-
ence thermocouple immersed in an ice-water mixture and to
a mV meter, from which values of e.m.f. are read at regular
time intervals (of between 30s and 90s, depending on the
cooling rate of the rod). More "advanced" options are the
use of thermocouple compensation and direct data logging
on a computer.
In the natural convection experiment, the rod is allowed to
cool in still air, whereas in the forced convection experiment
an electrical fan is used to blow the air in a direction perpen-
dicular to the axis of the rod. An anemometer is then needed
to measure the velocity of the air near the rod.


Equation (7)
Exptl. D=19mm
0 Exptl. D=32 mm

S- o ... 0--


100 1000 3000

Re,= U (I 2D)2

Figure 4. Effect of Reynolds number on dimensionless
film thickness. Comparison between experimental points
obtained by students and Eq. (7).
Chemical Engineering Education

Data Treatment

If a heat transfer coefficient, h, is defined, the cooling law
for the rod is
-mCp = hA(T To) (9)

where m is the mass of the rod, with external area A and
specific heat capacity Cp. The temperature of the rod at time
t is T, and To is the temperature of the air far from the rod,
taken to be invariable in time. If the variable 0 = T To is
defined, integration of Eq. (9) from t=0, for which time
0 = i (= Ti To), leads to

In9= n09i -(hA/mCp)t (10)

where hA/mCp has been taken to be constant over the time
interval considered. Equation (10) suggests a representation

mV meter Ice/water
Figure 5. Diagram of experimental setup.

of the experimental data as 0 = Oi vs t in a semilog plot, and
in Figure 6 the data obtained in our lab with a Dural rod
(L=0.250m, d=0.0250m) cooling in still air are reproduced.

Natural Convection
It is important in this part of the experiment to use well-
polished rods that are not covered by a thin oxide layer, so
that the contribution of radiative transfer becomes negli-
gible. A careful look at Figure 6 shows a slight upward
curvature in the alignment of the experimental points;
this is because h decreases with the temperature of the
rod, for natural convection. If the data in Figure 6 are
sliced into a number of successive time intervals, with
each containing a number of experimental points, it is
possible to determine the best-fit straight line for each
time interval, and from the corresponding slope the value
of h is found.
The values of h obtained in this way are represented in
Figure 7, and they are seen to compare extremely well with
the correlation given by Holman,[21 which for atmospheric
air at moderate temperatures reduces to

h = 1.32 ( / d)/4 (11)

with h in W/m2K, 0 in K, and d (the diameter of the rod) in

Forced Convection
For the conditions in our experiments, the heat transfer
coefficient for cooling under forced convection has a negli-
gible dependence on the temperature of the rod and as a
result log 0 / 0i vs. t plots as a straight line for each value of
air velocity.
In order to compare the values of h obtained from these
lines with predictions from Fand's correlation[21

0 1000 2000 3000
t (s)

4000 5000



1000 10000
OlD (K/m)

Figure 7. Heat transfer coefficients in
natural convection.

Figure 6. Cooling curve for natural convection
(dural rod, d=0.0250m, L=0.250m).
Summer 1999

Nu = (0.35 + 0.56 Re052) Pr 0.3 (12)
it is necessary to measure the velocity with which the air
approaches the cylinder, u,. In Eq. (12), Nu = hd I k is the
Nusselt number, Re = pu-d / p the Reynolds number, and
Pr = Cp~ / kPrandtl's number. All the physical properties
are for air at mean film temperature (= (T + To) / 2). In our
experiments, where transient cooling is observed, values of
Nu, Re, and Pr could be evaluated at different times (in each
experiment we evaluated these dimensionless groups near
the beginning and near the end of the cooling period, to
obtain two points on the plot in Figure 8).

Results and Discussion
Our students performed experiments with three different
metal rods (see Table 1), and the data obtained were used to
organize the plots in Figures 7 and 8. It can be seen that the
experimental points fall quite close to the correlations sug-
gested in the literature, even though the experimental tech-
nique is rather crude.

Pedagogical Comments
Students like this work for a variety of reasons. (1) They
have an opportunity to successfully test the theory of tran-
sient heat transfer for lumped parameter systems. (2) They
obtain individual values of the heat transfer coefficient, which
(some are surprised to see) compare well with those given by
available correlations. (3) If two rods of the same size are
used, one made of duralumin and the other made of copper,
the latter cools more slowly under otherwise similar condi-
tions, due to its higher heat capacity. For the weaker stu-
dents, it is intriguing to see that the same heat transfer
coefficients are obtained for both rods. (4) The effect of
temperature difference on the heat transfer in natural convec-
tion is also brought out vividly. (5) The importance of radiative
heat transfer is brought to evidence if two equal rods are used
in the natural convection experiment, of which one has been
allowed to oxidize so as to lose its shiny appearance.

A area for convection (m2)
C specific heat capacity (J/kgK)
D column internal diameter (m)
d diameter of rod (m)
g acceleration of gravity (m/s2)
H length between bottom of column and ball valve (m)
h heat transfer coefficient (W/m2K)
k thermal conductivity (W/mK)
L length of rod (m)
m mass of rod (kg)
N, dimensionless inverse viscosity (-)
Nu Nusselt number evaluated at mean film temperature (-)

Figure 8. Dimensionless heat transfer coefficients in
forced convection.

Characteristics and Properties of the Metal Rods
Used in the Experiments

d(m) L(m) m(kg) Cp'*(J/kg.K)
Copper (1) 0.0158 0.250 0.440 383.1
Copper (2) 0.0199 0.250 0.693 383.1
Duralumin 0.0250 0.250 0.334 883.0
(*) Ref. 2.

Pr Prandtl number evaluated at mean film temperature (-)
Q volumetric flowrate of the liquid (m3/s)
q liquid flowrate per unit wetted perimeter (m2/s)
Re Reynolds number evaluated at mean film temperature (-)
Re, film Reynolds number (-)
t time (s)
T temperature of rod (K)
To ambient temperature (K)
U velocity of cylindrical gas bubble (m/s)
uo air velocity (m/s)
V collected volume (m3)
6 film thickness (m)
p dynamic viscosity (kg/ms)
v kinematic viscosity of the liquid (m/s)
0 temperature difference, T-To (K)
0i value of 0 at t=0 (K)
p density (kg/m3)
Sdimensionless film thickness (-)

1. Wallis, G.B., One-Dimensional Two-Phase Flow, McGraw-
Hill, New York, NY, Chap 10-11 (1969)
2. Holman, J.P., Heat Transfer, 8th ed., McGraw-Hill Interna-
tional Edition, Chap 6-7 (1997) O
Chemical Engineering Education



Continued from page 225.

educational achievements. For example, specific assignments,
such as summaries, can be graded more objectively and
rigorously, for completeness and accuracy.
Instructor comments, rather than grades, are the real key to
making the journal a valuable learning tool for enriching
professional writing and organizational skills. Comments
that praise curiosity, originality, insight, and innovative or-
ganization will increase students' efforts in journal writing.
Comments that demonstrate the instructor's desire for open
communication will make the journals more useful for the
instructors. Also, instructors should address specific con-
cerns raised in the journals, referring students to the text or
other references rather than just answering questions. In this
way, students will rely on their own initiative, rather than the
instructor, in seeking answers to questions that have solu-
tions within their capability to discover.
Copious instructor comments might seem to imply a huge
investment in time, but experience has shown the opposite.
Logbooks can be rapidly read. I often collect and read twenty
journals (but not the reference pages or outlines) while proc-
toring fifty-minute class exams. The chapter outlines and
reference pages generally need little correction after the first
week or two of class, because the writing assignments are so
concrete that engineering students have little trouble with
them. (For a large class, these two sections might even be
assigned to a teaching assistant for grading.) If term projects
are required, the time spent on grading logbooks also pays a
time dividend at the end of the term. Students produce better,
more organized, and more easily graded projects when they
have journal feedback throughout the semester on their writ-
ing and on their developing projects.
One technique that is useful in minimizing instructor ef-
fort while insuring copious useful instructor feedback is to
write group remarks for common questions, problems, or
insights. Instead of the instructor writing "Does this RE-
ALLY apply in the transition region?" in ten or twenty
journals, he or she can distribute these written general re-
marks in class and the students can incorporate them into the
individual journals. In addition to saving time, this method
also helps the students realize they were not the only ones
who had this particular insight or concern or error. This
sort of realization bolsters student confidence in the re-
wards of journal writing and increases their enthusiasm.
Journal entry quality seems to be directly correlated with
confidence and enthusiasm.
The chapter outlines are the most objectively graded sec-
tion. Format is generally up to the student, but it is evaluated
for usefulness as a study tool. For example, the outlines must
have proper page number annotation to serve as open-book
tools. The outlines should be a page or two long, highlight-
Summer 1999

ing and referencing the key concepts of the chapter. Long
passages copied from the text receive lower grades, as
they do not demonstrate the same ability to distinguish
the main from the supporting, incidental, or supplemen-
tal ideas and facts. Content is graded rigorously on com-
pleteness and accuracy.
It is usually obvious from the content whether or not the
students are really doing the intellectual work of reading and
outlining, rather than just copying the table of contents or
chapter headings. Such superficial efforts should not be re-
warded with good grades. Again, modeling the concepts, by
either composing the outline of the first chapter in class or
by handing out examples of good student outlines early in
the term will let the students know what is expected of them
and improve their ability in outlining.
The reference pages are the student's personal study guides
and need little grading. Instructor remarks can be limited to
suggestions on possible omissions. Grades are generally quite
liberal, but two behaviors should result in lowered grades for
this section. The first is inaccuracies and errors that are
pointed out by the instructor and are left uncorrected. The
second is incorrect or inadequate referencing of sources,
again, if left uncorrected.

For the past few years, I have had positive experiences
integrating writing and communication into undergraduate
chemical engineering classes by means of logbooks, which
include journals, chapter outlines, and reference pages. Al-
though journals are a well-established pedagogical tool in
many arenas, from music composition classes to English
classes to teaching-methods workshops, there are definite
challenges in using the device well in a technical course.
With proper structuring, the logbooks have been a posi-
tive communication and organizational tool for the students.
They report that they like keeping the logbooks and find
them useful as well. They feel that their opinions and con-
cerns are being heard and they especially enjoy the tan-
gible feeling of progress that looking over the journal
and the summary assignments gives them. They learn
and practice valuable writing skills not usually required
and graded in the classroom.
The instructor also benefits from the logbooks, which
provide an effective, convenient, and efficient manner of
communicating with the students and monitoring and en-
hancing their learning experiences.

1. Bloom, B.S., ed., Taxonomy of Educational Objectives: The
Classification of Educational Goals: Handbook 1. Cognitive
Domain, David McKay Company, New York, NY (1959) O

j"f laboratory



Using a Tank-Tube Viscometer

Tuskegee University Tuskegee, AL 36088

At Tuskegee University we have added a new labora-
tory experiment for the fluid mechanics and trans-
port phenomena laboratory (the Unit Operations I
Laboratory) that investigates the effects of both water con-
centration and temperature on the viscosity of aqueous glyc-
erol solutions. The experiment is designed as an extension of
the fluid mechanics course (offered to sophomore students),
the engineering mathematics course (for juniors), and the
transport phenomena course (offered to seniors).
We offer the course twice a year and it has an average
student enrollment of twelve. The students are usually di-
vided into three groups. The objectives of the course are to
engage each student in active participation and experimenta-
tion and to have them analyze statistically the experimental
data with the aid of a computer, to perform necessary calcu-
lations for tables and figures, and to prepare a written report
using a word processor.
In this paper we will describe several experiments for
measuring the viscosity of aqueous glycerol solutions using
a tank-tube viscometer. Measuring the viscosity of highly
viscous liquids with the tank-tube viscometer is easier than
using other types of viscometers. This inexpensive viscom-
eter generates numerous reproducible viscosity data of highly
viscous aqueous glycerol solutions under given experimen-
tal conditions. The tank-tube viscometer consists of a large-
diameter reservoir and a long, small-diameter, vertical tube.
Fabrication of the tank-tube viscometer is inexpensive
since it does not need ancillary equipment such as a high-
pressure pump, a pressure transducer, or an accurate flow
meter. Our viscosity experiment provides an opportunity for
students to apply mathematical and computational skills to
analyzing statistically experimental data and to write a re-
port using computer software. Mathematical and computa-

tional skills are learned through the mathematics courses as
well as the basic engineering courses that are offered to our
freshman and sophomore students. Our experiment also fa-
miliarizes students with the concept of viscosity of highly
viscous Newtonian fluids, which they learned in the lecture
class of the fluid mechanics course.
The main objective of the experiment is to demonstrate the
effects of water concentration as well as temperature on
viscosity by applying experimental data of accumulated
amounts of aqueous glycerol solutions at various drain dura-

K. C. Kwon is Professor of Chemical Engineer-
ing at Tuskegee University. He received his BS
from Hanyang University, Seoul, Korea, his MS
from University of Denver, and his Ph.D from
Colorado School of Mines. His industrial experi-
ence includes five years as a process engineer
at the synthetic fuel division of Gulf Oil Com-
pany, Tacoma, Washington. His research inter-
ests include reaction kinetics, coal conversion,
adsorption separation, metal oxide sorbents and
transport properties.
S. Pallerla is Assistant Professor of Chemical
Engineering at Tuskegee University. He received
his BS from Osmania University, Hyderabad,
India, his MS from the Indian Institute of Tech-
nology-Madras and his PhD from Auburn Uni-
versity. His research interests include environ-
mental biotechnology, bioprocessing, kinetics,
adsorption and pulp and paper engineering.

Sanjeev R. Roy is working in Chemical/Environ-
mental Engineering at Tuskegee University. He
received his BS from Birla Institute of Technol-
ogy, India, and his MS from Tuskegee Univer-
sity. He has worked as Executive Engineer for
Oil & Natural Gas Corporation India for nine
years. His research interests includes innovative
approach to system design.


Copyright ChE Division of ASEE 1999
Chemical Engineering Education

The experiment is designed as an extension of the fluid mechanics course
(offered to sophomore students), the engineering mathematics course (for juniors),
and the transport phenomena course (offered to seniors).

tions to a newly developed viscosity equation for the
fabricated tank-tube viscometer. The viscosity equation
was developed under the assumptions that both the quasi-
steady-state approach and the negligible friction loss due
to a sudden contraction between the reservoir tank and
the tube are valid.

Fluid mechanics is the study of forces and motions in
fluids.["] Treatment of fluid flow required understanding the
physical properties of a fluid that affects the motion. The
two most important fluid properties are density and viscos-
ity, and the most important physical property of a fluid from
the point of view of the study of fluid mechanics is the
viscosity21 A fluid is a substance that undergoes continuous
deformation when subjected to a shear stress; the resistance
offered by a real fluid to such deformation is called its
viscosity. All fluids have viscosity. This property causes
friction.t31 The viscosity of a Newtonian fluid is constant if
static pressure and temperature are fixed.
A chemical engineer is concerned with the transport of
fluids from one location to another by pumping fluids through
pipes over long distances from storage to reactor units.[41
Many intermediate products are pumped from one unit op-
eration to another, and raw materials such as natural gas and
petroleum products may be pumped very long distances to
domestic or industrial consumers.
These industrial processes require determination of the
pressure drops in both the pipeline and the individual units
themselves, evaluation of the power required for pumping,
and estimation of the most economical sizes of pipes and
measurement of the flow rates. The viscosity value of a
particular fluid flowing through pipes and individual equip-
ment is an essential property in designing these industrial
Viscosity values of a Newtonian fluid can be calculated

S(H+L ( gR4p
Ih+L) 8pR2LJ

if levels of a liquid in a reservoir tank of a tank-tube viscom-
eter151 at different drain durations are known. Equation (1) is
developed based on the assumptions that both the quasi-
steady-state approach and the negligible friction loss due
to a sudden contraction between the reservoir tank and
the tube are valid.
The change in the level of a highly viscous liquid in the
Summer 1999

reservoir tank at a given drain duration is very small, how-
ever. Often it is difficult to read the change in the level of the
liquid in the reservoir. Hence, the liquid level at a given
duration of time is described in terms of accumulated amounts
of the liquid drained from the reservoir tank (see Eq. 2). A
mass balance of the liquid around the reservoir tank of the
tank-tube viscometer produces

h =H -- (2)

We then obtain

( m ( gR4p
-n I- g t (3)
-nl (H + L)R2p 8R2L) (3)
by substituting the "h" value from Eq. (2) into Eq. (1). The
left-hand side of Eq. (3) contains accumulated amounts of
the liquid drained from the tank-tube viscometer, while
the right-hand side contains drain durations. Therefore,
Eq. (3) allows us to calculate the liquid viscosity if the
total mass of the liquid out of the reservoir at a given
drain duration is known.
The left-hand side of Eq. (3) is denoted as Y, as shown by

( m
Y = -(nl -H + LR 2 (4)
(H + L)rR
The Y values of Eq. (4) are easily calculated by substituting
both the amounts of aqueous glycerol solutions drained from
the tank-tube viscometer and their density values. Densities
of aqueous glycerol solutions were obtained with a densim-
eter. The "H" value in Eq. (3) is the same as the height of the
reservoir when the reservoir tank is filled with a liquid. In
this laboratory experiment, the reservoir is filled with aque-
ous glycerol solutions to avoid measuring the initial level of
aqueous glycerol solutions in the reservoir. Otherwise, a
possible experimental error source will be added to the ex-
periment by measuring an initial level of aqueous glycerol
solutions in the reservoir.
Equation (3) can be simplified to obtain

( ( gR4p )
1(H + L)xR2p ) 8 R2Lt (5)

if the length of the vertical tube of the tank-tube viscometer
is relatively longer than the initial level of a liquid in the
reservoir and if amounts of the liquid drained from the
reservoir are relatively small.
The viscosity of liquids decreases with increasing tem

perature. An approximate empirical observation for the tem-
perature dependency of viscosity for liquids is described by

p = Ae(B/RT) (6)
where A and B are empirical constants. This equation can be
used with viscosity data for interpolation or modest extrapo-
lation.161 Since liquids are essentially incompressible, the
viscosity of liquids is not affected by pressure.[71
An average velocity equation for liquid flow in the vertical
tube of the tank-tube viscometer is described as a function of
accumulated amounts of aqueous glycerol solutions drained
at a given drain duration. Combining the average velocity

v pgR 2 m
vm H+L--1 (7)
8uL E R2P (J

with Eq. (3) gives

V (R ( m m 1
v R i2 H+L-H 1, (8)
vm n-RJ [ 7 R2p(H +L) nR 2 R t (8)

A wide variety of viscometers (capillary, glass-tube, rota-
tional, falling-ball, cup, and oscillatory"81) is available for
measuring viscosity. An inexpensive viscometer, the so-
called tank-tube viscometer, was fabricated for the course
(see Figure 1). It consists of a cylindrical reservoir and a
long vertical tube. The radius and the height of the reservoir
are 2.5531 cm and 14.7 cm, respectively. The radius and
length are 0.1637 cm and 73.8 cm, respectively.
The reservoir is made of a transparent Plexiglas pipe,
whereas the tube of the viscometer is made of stainless steel.
The vertical tube is connected at the bottom of the reservoir.
Aqueous glycerol solutions are chosen to test the fabricated
tank-tube viscometer since glycerol is completely soluble in
water, very viscous, and does not inflict any health haz-
ards.r91 The bottom end of the vertical tube is initially closed
with a rubber bulb and aqueous glycerol solutions are fed to
the reservoir. An electronic balance placed beneath the bot-
tom end of the vertical tube is used to measure accumulated
amounts of aqueous glycerol solutions drained from the
reservoir at a given drain duration.

The tank-tube viscometer is set up by placing an electronic
balance beneath the bottom end of its vertical tube in a
constant-temperature chamber, as shown in Figure 1. The
cylindrical reservoir tank is filled with an aqueous glycerol
solution that is allowed to flow through the vertical tube.
When the vertical tube is filled with the aqueous glycerol
solution by evacuating air from it, its bottom end is closed
with a rubber bulb to stop the flow of solution. After closing

... we will describe several experiments
for measuring the viscosity of aqueous glycerol
solutions using a tank-tube viscometer.... This
inexpensive viscometer generates numerous
reproducible viscosity data of highly viscous
aqueous glycerol solutions under given
experimental conditions.

Resevoir 2R1
Tank - ,




..... .............

Figure 1. Schematic diagram of a tank-tube viscometer.

its bottom end, the cylindrical reservoir tank is again filled
with the aqueous glycerol solution. Consequently, measure-
ment of the initial level of the solution in the reservoir is not
necessary since the it is equal to the height of the reservoir
itself. The height of the reservoir is incorporated into Eq. (3).
An empty receiving beaker is placed on the electronic
balance and tared. Reading accumulated amounts of aque-
ous glycerol solutions off the LSD digital indicator of the
electronic balance starts when the rubber bulb from the
bottom end of the tube is removed to allow the solution to
flow into the beaker.
Accumulated amounts of aqueous glycerol solutions
drained from the viscometer at random drain durations are
read off the electronic balance, using a stopwatch. After the
tank-tube viscometer is rinsed with distilled water for the
next experiment, the reservoir tank is dried with paper tow-
els and the vertical tube is dried with acetone.

Experimental data of accumulated amounts of an aqueous
glycerol solution drained from the reservoir of a tank-tube
viscometer at various drain durations are obtained by using
an electronic balance and a stopwatch. Several experimental
data for aqueous glycerol solutions were obtained under
controlled experimental conditions, such as concentrations
Chemical Engineering Education

of water in aqueous glycerol solutions and temperatures of
aqueous glycerol solutions.
Each group of students obtained four different viscosity
values of aqueous glycerol solutions with four different wa-
ter contents at a controlled temperature during a 3-hour
laboratory session. These experiments provide opportunities
for students to learn the effects of water concentrations in
aqueous glycerol solutions on viscosity values. Experimen-
tal results (performed at three different temperatures by three
groups) are gathered and plotted after a complete rotation of
this laboratory experiment to each group. These experi-
mental results also provide opportunities for students to
learn effects of temperature of aqueous glycerol solu-
tions on viscosity values.
Personal computers, loaded with a FORTRAN program
and Microsoft Professional Office, are used to process sev-
eral series of experimental data of amounts of aqueous glyc-
erol solutions drained from the reservoir at various drain
durations. These data are applied to Eqs. (3) and (8) to
calculate viscosity values and average velocity values with
the aid of the FORTRAN program. Figures 2 through 8 are
plotted using Microsoft Excel.

0 045
> a 0025
Seat glycerol
002 3.44 w% water
0.0155 34 w% water
0.01 x8.08 w% water
0 200 400 600 800 1000 1200 1400 1600 1800
Drain Duration, s
Figure 2. Values of the left-hand side (LHS) of Eq. (3) at
various drain durations and 270C.

0 2 4 6 8 10 12 14 16
Concentration of Water, w%

Figure 3. Comparison of viscosity values of aqueous
glycerol solutions at various temperatures from the
experiment with those from the literature.
Summer 1999

A slope of the best-fit straight line passing through the
origin of the rectangular coordinates is obtained through the
linear least-squares method. A viscosity value is calculated
by substituting a density value of aqueous glycerol solu-
tions, the diameters of both the reservoir tank and the verti-
cal tube, and the length of the vertical tube into the slope
value (see Eq. 4). Consequently, these computations provide
an opportunity for students to process experimental data
with the aid of personal computers loaded with necessary
computer software.
Students also obtain the slope values of the best-fit straight
lines, their correlation coefficients, and their viscosity val-
ues using their hand calculators, rather than personal com-
puters, when preparing a model calculation section of a
laboratory report. The computer-generated data were used to
examine whether or not their hand calculations were correct.
As a result, they were able to enhance their computational
skills. Computations by hand calculators also help students
solve the written problems of a final examination since the
answers to its problems are obtained with hand calculators.
A detailed derivation of Eqs. (3), (5), and (8) is discussed
in the senior transport phenomena class, using the math-
ematical skills learned from the mathematics courses. Equa-
tion (3) is derived assuming that the quasi-steady-state ap-
proach and the negligible friction loss due to a sudden con-
traction between the reservoir tank and the tube are valid.
Validity of the quasi-steady-state assumption is discussed in
the transport phenomena class and the chemical reaction
engineering classes by presenting the experimental results
on viscosity of aqueous glycerol solutions. Deviation range
of viscosity values obtained from the approximate equa-
tion (Eq. 5) from those from Eq. (3) is discussed in the
engineering mathematics class by presenting the experi-
mental results of this experiment.
Students prepare a report using a word processor. A typi-
cal laboratory report includes several sections, such as an
abstract, an introduction, theory, the experimental set-up,
experimental procedures, calculations, results, discussion,
and conclusions. Figures and tables generated from a labora-
tory experiment and a schematic diagram on an experi-
mental set-up are also include in the report. Figures and
the schematic diagram must be drawn with the aid of
computer software, whereas the calculation section should
be handwritten.

Left-hand side values of Eq. (4) are obtained with accumu-
lated amounts of an aqueous glycerol solution drained from
the reservoir at various drain durations, which are plotted
against drain durations (see Figure 2). The slope of the best-
fit line is obtained through the linear least-squares method.
The viscosity value is calculated from the slope of the best-
fit line by substituting the density value of the aqueous

glycerol solution as well as the sizes of the viscometer into
the right-hand side of Eq. (3). The sizes of the viscometer
include the radius of the reservoir and the diameter and the
length of the vertical tube.
The slope of this plot increases with increased concentra-
tions of water in the aqueous glycerol solution. This obser-
vation shows that the viscosity of aqueous glycerol solutions
decreases with increased concentrations of water. A good
linear relationship between Y values and drain durations
(see Figure 2) may indicate that the assumptions made in
developing the viscosity equation for a tank-tube viscometer
are valid.
Viscosity values of aqueous glycerol solutions obtained
from this experiment are compared with those from the
literature at various concentrations and temperatures (see
Figure 3). The values obtained from this experiment are in
agreement with those from the literaturet10] over the range of
water concentration and temperature explored, with an aver-
age deviation of 3.8% (see Figure 3). These observations
also suggest that the validity of the assumptions made in
developing the viscosity equation for a tank-tube viscometer
are justified.
Viscosity values of aqueous glycerol solutions are plotted
against concentrations of water in aqueous glycerol solu-
tions at various temperatures (Figure 3). Viscosity values
decrease drastically at relatively low concentrations (below
8 wt %) of water in aqueous glycerol solutions, while viscos-
ity values decrease moderately at relatively high concentra-
tions (above 8 wt %) of water.
Viscosity values are a strong function of temperature at
relatively low concentrations (below 8 wt %) of water in
aqueous glycerol solutions, whereas viscosity values are a
moderate function of temperature at relatively high concen-
trations (above 8 wt %).
A series of viscosity values of neat glycerol at various
temperatures is applied to Eq. (6) to find the Arrhenius
relationship between viscosity values and temperatures of
aqueous glycerol solutions (see Figure 4). A very good
Arrhenius relationship

S= 1.4618 x 10-8 e(7439/T) (9)

is obtained over the temperature range of 15 to 27'C with a
correlation coefficient value of 0.99.
Each viscosity value of aqueous glycerol solutions at its
random drain duration is calculated using Eq. (3). This vis-
cosity value is plotted against its accumulated amount of
aqueous glycerol solution drained from the viscometer at
random drain durations (see Figure 5). Viscosity values ap-
pear to be independent of accumulated amounts of aqueous
glycerol solution drained at various drain durations. These
results show that the solutions are Newtonian fluids and
reproducibility of viscosity values of aqueous glycerol solu-

tions obtained from the viscosity equation of the tank-tube
viscometer appear to be excellent.
Viscosity values of aqueous glycerol solutions calculated
with the approximate equation (Eq. 5) are compared with
those from Eq. (3). Deviation percentages of viscosity val-
ues obtained from Eq. (5) in reference to those from Eq. (3)



1/T, K'


Figure 4. Effects of temperature on viscosity of
neat glycerol.

0 20 40 60 80 100 120 140
Amount of Solution, g
Figure 5. Experimental viscosity values of aqueous
glycerol solutions plotted against accumulated amounts
of solution drained at 24.9C.

500. ,1143w%water
400 x2041w%water


0 50 100 150 200 250
Amount of Soluton, g

Figure 6. Deviation percent of viscosity values
obtained from Eq. (5) in comparison with
those from Eq. (3) at 24.9C.

Chemical Engineering Education

y= 3.95E+65x2 E'
R2 = 9.87E-01

o 0w % water (neat glyceroD
S11.43w %water
S16.48 w % water
x 20.41 w %water

Sp- p--e--
----B---0----- ----- -----B -----o --o-

are evaluated at random drain duration. Deviation percent-
ages increase with amounts of aqueous glycerol solutions
drained from the tank-tube viscometer (see Figure 6) with a
deviation range of 0 to 5%. Deviation percentages appear to
be independent of water concentrations in aqueous glycerol
solutions over the water concentration range from 0 to 20%.
The maximum deviation error is 8.56% when the reservoir
tank filled with aqueous glycerol solutions is completely
Average velocities for flow of aqueous glycerol solutions
in the vertical tube of the tank-tube viscometer are calcu-
lated at 24.90C, using Eq. (8). Average velocities of aqueous
glycerol solutions increase with increased water concentra-
tions (see Figures 7 and 8). The standard deviation of aver-
age velocities decreases with decreased water concentra-
tions over the drain duration range of 10 to 250 seconds. The
mean values of average velocities and their standard devia-
tions for the 20.41-, the 16.48-, and the 11.43-wt%-water
glycerol solution, and the neat glycerol are 10.250.17 cm/s,
6.0610.16 cm/s, 2.490.11 cm/s, and 0.4680.024 cm/s,
respectively. These observations may indicate that validity
of the quasi steady-state approach is justified for the deriva-
tion of the viscosity equation of the tank-tube viscometer.

1200 .
1000 -e--2041 w%water
9 00 B-. 1648w% water
800 -- 11 43w%water
70 700 1 neat gccerol
60 00 .

3 00
1 00
0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300
Drain Duration, s

Figure 7. Average velocity of aqueous glycerol solutions
in the vertical tube of the tank-tube viscometer against
various drain durations at 24.9C.

Figure 8. Average velocity against various water concen-
trations in aqueous glycerol solutions at 24.9C.
Suunmer 1999

An inexpensive tank-tube viscometer was fabricated to
determine viscosity of highly viscous aqueous glycerol solu-
tions, and a viscosity equation of a tank-tube viscometer was
developed to calculate the viscosity of the solutions. This
experiment introduces chemical engineering students to the
concept of viscosity of a Newtonian fluid in fluid mechanics.
It also provides an opportunity for the students to carry out
experiments for the acquisition of experimental data, to ap-
ply their mathematical and computational skills as well as
statistical analysis to interpreting experimental data with the
aid of computer software, to survey the literature on viscos-
ity, and to write a laboratory report using a word processor.

cp centipoise
g acceleration of gravity
h level of a liquid in a reservoir tank at a drain duration t
H initial level of a liquid in a reservoir tank or height of a
reservoir tank when the reservoir tank is filled
L length of a vertical tube
LHS left-hand side values of Eq. (3)
m accumulated amount of a liquid drained at t
R inside radius of a cylindrical reservoir tank
R inside radius of a vertical tube
t drain duration
T temperature of aqueous glycerol solutions, K
vm average velocity of a fluid flow in a vertical tube
Y left-hand side value of Eq. (3)
p viscosity of aqueous glycerol solutions
P density of aqueous glycerol solutions

The authors thank Drs. Nader Vahdat and Dennis Liken
for comments and suggestions on this paper.

1. Denn, M.M., Process Fluid Mechanics, Prentice-Hall (1980)
2. Nevers, Noel D., Fluid Mechanics for Chemical Engineers,
2nd ed., McGraw-Hill, New York, NY (1991)
3. McCabe, W.L., J.C. Smith, and Peter Harriott, Unit Opera-
tions of Chemical Engineering, 5th ed., McGraw-Hill, New
York, NY (1993)
4. Coulson, J.M., and J.F. Richardson, Chemical Engineering,
3rd ed., Vol. 1, Pergamon Press, Oxford (1984)
5. Bird, R. Byron, Warren E. Stewart, and Edwin N. Lightfoot,
Transport Phenomena, John Wiley & Sons (1960)
6. Geankoplis, Christie J., Transport Processes and Unit Op-
erations, 2nd ed., Allyn and Bacon, Boston, MA (1983)
7. Brodkey, Robert S., and Harry C. Hershey, Transport Phe-
nomena: A Unified Approach, McGraw-Hill, New York, NY
8. Perry, J.H., and C.H. Chilton, Chemical Engineers' Hand-
book, 5th ed., McGraw-Hill, New York, NY (1973)
9. Kirk-Othmer Concise Encyclopedia of Chemical Technology,
3rd ed., John Wiley & Sons, New York, NY (1985)
10. Dean, John A., Lange's Handbook of Chemistry, 4th ed.,
McGraw-Hill, New York, NY (1992) 0

class and home problems

The object of this column is to enhance our readers' collections of interesting and novel
problems in chemical engineering. Problems of the type that can be used to motivate the student
by presenting a particular principle in class, or in a new light, or that can be assigned as a novel
home problem, are requested, as well as those that are more traditional in nature and that
elucidate difficult concepts. Manuscripts should not exceed ten double-spaced pages if possible
and should be accompanied by the originals of any figures or photographs. Please submit them to
Professor James O. Wilkes (e-mail:, Chemical Engineering Depart-
ment, University of Michigan, Ann Arbor, MI 48109-2136.




University of Mississippi University, MS 38677-9740

Electrically heated tubular flow reactors have been
commonly used for rate measurement in research
laboratories. Questions often arise concerning the
effect of velocity profile, axial and radial molecular diffu-
sions, and axial temperature distribution on the measured
conversion and estimated rate. Analysis of reactors of this
type can be a fruitful area for homework for a reaction
engineering course. Two home problems are offered in this
article; both are aimed at measurement of reaction rate of
nitrogen oxide and char with a flow reactor of commonly
adopted size.
The first problem justifies the use of conversion data from
a laminar flow reactor for rate estimation. In a laminar
tubular flow system without reaction, the concentration of a
gaseous species, C, can be described to a two-dimensional
dispersion model, i.e.

ac a2c (a2C 1 a ( r2 )ac
a=DM 2 +DM I ---ar2 r I- u 1 (1)

where D, represents the molecular diffusivity of the species
of interest, and u, is the maximum centerline velocity. The
analyses by Taylor,"21 Aris,t31 and Hunt[4] demonstrated that,
under a specific criterion, Eq. (1) reduces to an axial-dis-
persed, plug-flow model, or

aC a2C aC
t Ez u (2)

where u is the mean fluid velocity and Ez is the effective
axial dispersion coefficient, or the sum of molecular
diffusivity and the apparent diffusivity contributed by the
laminar velocity profile, i.e.
Ez = DM + (3)
48 DM

Equations (2) and (3) have been derived assuming that the
radial mixing is great enough compared to the longitudinal
convective mixing to ensure a uniform cross-sectional con-
centration. Mathematically, this criterion implies a small
radial Peclet number, Per, or

Wei-Yin Chen is Associate Professor of Chemi-
cal Engineering at the University of Mississippi.
His teaching and research interests have been
in reaction engineering and mathematical mod-
eling. He received a PhD in Chemical Engineer-
ing from the City University of New York, an MS
in Chemical Engineering from the Polytechnic
Instutute of New York, an MS in Applied Math-
ematics and Statistics from the State University
of New York at Stony Brook, and a BS in Chemi-
cal Engineering from Tunghai University.

Copyright ChEDivision of ASEE 1999

Chemical Engineering Education

uR --
uR 16 L

where L is the length of the tube.
For a laminar-flow reactor involving a first-order chemi-
cal reaction and back mixing, the full description of the
reacting system involves an additional reaction term in Eq.
(1). Wisslert15 demonstrated that the Taylor-Aris model re-
mains valid, i.e., Eqs. (3) and (4) remain the basis for

ac 2C ac
S= E, u- + kC (5)
at _z 2 z *
provided that the reaction rate, k, satisfies the condition

< 1 (6)
(3.82 DM)

At steady state, Eq. (5) has been solved by Wehner and
Wilhelm.'6 For small Ez/(uL), their solution after dropping
higher-order terms in its series expansion becomes

X = ex-kt + (kt)2 E (7)

where X is the conversion. The second factor in the above
exponential term represents the deviation of the dispersion
model, Eq. (5), from a plug-flow reactor.
The second problem investigates the effect of axial tem-
perature variation on the estimation of intrinsic rate. In a
tubular reactor heated by a single heating element of uni-
form resistance, reacting gas travels successively through
heating, isothermal, and cooling sections, and rate estima-
tion requires analysis of conversions in all three stages.
Using an experimentally measured wall temperature distri-
bution and a heat-transfer algorithm established by Sellars,
et al., 17 it is possible to calculate the average gas temperature

0 1200
800 -
0- 400
0 2 4 6 8 10 12 14 16
o Measured wall temp. --- Calculated gas temp.
Temp. profile for rate const- Curve fit of wall temp.

Figure 1. Measured wall temperature, calculated gas tem-
perature based on Sellars, et al.,7' and simplified three-
zone temperature profile for kinetic analysis.
Summer 1999

over a particular cross-sectional area of the tubular flow (see
Figure 1). The predicted distribution indicates that, for a
reactor equipped with a 30.5-cm-long heating element, each
nonisothermal section has about the same length as the iso-
thermal section. Since the gas travels slower and has a longer
residence time in these nonisothermal sections, the rate esti-
mation should take into consideration the conversions in
these two sections. Problem 2 addresses this issue and seeks
to recover the intrinsic rate of NO/char reaction and mass-
transfer limitations. To simplify the calculation, the gas tem-
peratures in the two nonisothermal sections are approxi-
mated by two constant heating-rate profiles. It is also dem-
onstrated that, while this type of analysis traditionally re-
quires extensive effort on trial and error, it can be conve-
niently solved now by contemporary software such as


An electrically heated tubular reactor with an I.D. of 1.91
cm and a heated length of 30.5 cm is used for measuring the
rate of NO/char reaction. Products of the reaction are N2,
CO, and CO,. The feed at 1 atm contains 1000 ppm of NO
and char particles in a helium base. Using the heat transfer
algorithm of Sellars, et a.,[7] and an experimentally mea-
sured temperature distribution, the axial gas temperature
distribution is calculated (see Figure 1); in the central seg-
ment of the reactor the gas reaches 11000C for about 10 cm.
Well-dispersed char particles in the reacting gas are fed
vertically downwards into the reactor at 250C at a volumet-
ric flow rate of 2000 cm3/min. Data collected from this
reactor indicate that the NO disappearance rate is first order
with respect to the NO concentration, and the NO conver-
sions are always below 95% if a proper char feeding rate is
1. Using the Taylor-Aris11'1 criterion, determine if an
axial-dispersed, plug-flow model is appropriate for the
approximation of gas concentration in a laminar flow
system similar to that discussed above but containing no
chemical reaction. If so, determine the effective
2. Using the Wissler"1 criterion, determine if the axial-
dispersed, plug-flow model (Eq. 5) can be used for
estimating the rate of the NO/char reaction.
3. Using the solution of Wehner and Wilhenlm,/6 conclude
if it is pertinent to use the plug-flow reactor model
without dispersion for estimating the reaction rate.

1. A laminar flow system without chemical reaction can be
approximated by an axial-dispersed, plug-flow model when
Eq. (4) is satisfied or the radial Peclet number, Pe,(=uR/D,),
is small. At 1100C, the flow rate and average residence

time of gas in the isothermal section of the reactor are 153.6
cm3s-' and 0.2 s, respectively. The diffusivity of NO in He
can be estimated, based on molecular theory,t81 as

AB r=-rAT- (8)

(PcAPcB)3(TcATcB)2 M a

DAB cm2sec-'
P atm
a 2.745 x 104
b 1.823

This formula suggests DAB = 13.7 cm2s' at 11000C. By
resorting to the Chapman-Enskog kinetic theory (see Prob-
lem 2), DAB = 8.1 cm2s'-. Variations of diffusivity in this
range do not change the conclusions to be discussed below.
Based on the system parameters obtained above, the radial
Peclet number at 1100'C and dimensionless term on the
right-hand side of Eq. (4) are estimated as

- = D =M 0.288
Per Ru

and 0.00597

Thus, the Taylor-Aris criterion, Eq. (4), is satisfied, and by
resorting to Eq. (3), the effective axial dispersion coefficient
22 2
Ru cm
Ez = D + -= 17.1 (10)
48 DM s

2. Wissler's criterion gives the range of k for adopting the
dispersion model, Eq. (5), as

(3.8)2 D
(3.8)' DM

or k < 216s

This condition corresponds to an extremely fast rate or a
nearly complete conversion, specifically

[NO] n
1- X=- [N = e-kt_=kt 0-19 (12)

Since NO conversions from the reactor have always been
below 95%, Wissler's criterion is satisfied.
3. At a high conversion, e.g., X=0.95, and a reaction time
t=0.2s, Eq. (12) gives the rate from a plug-flow reactor
without dispersion, or k=15.0 s-'.The reciprocal value of the
effective axial Peclet number is 1/Pe, = Ez/uL = 3.38 x 102.
This small number implies only a small deviation from the
assumption of zero dispersion. Specifically, resorting to the
solution of Wehner and Wilhelm, or Eq. (7), we obtain the
rate under the influence of dispersion, k = 16.9 s-' Thus, the
assumption of zero dispersion leads to a maximum error of

about 13% at X = 0.95; this error is lower at lower conver-
sions (4.4% at X=0.70 and 2.5% at X = 0.50).


While Problem 1 justifies the use of a plug-flow reactor
model for rate estimation, this problem seeks to study the
effect of the two nonisothermal sections of the flow on
measuring the intrisic rate of NO reaction with char.
1. Assuming the reaction is important only in the 10-cm
isothermal region and the reaction is free from internal and
external mass-transfer limitations, derive a design equation
that relates reaction rate to experimentally observed NO
conversion, internal surface area of char, char feeding rate,
and residence time. The first-order reaction rate of NO on
char surface can be expressed as
rNo = -kAPNo (13)
rNO rate of NO formation, moles s-' (g of char)'
A specific, internal surface area of char, m2g-
PNO partial pressure of NO, atm
k rate constant, moles s- 'm atm-1

2. Redo Part 1 of the problem assuming the conversions in
the two nonisothermal regions of the reactor are not negli-
gible and can be approximated by two constant heating-rate
profile (see Figure 1).
3. Assuming the internal mass-transfer limitation cannot be
ignored, estimate the frequency factor, activation energy,
Thiele modulus, and effectiveness factor in a three-zone
reactor as described in Part 2. For a reactive char derived
from lignite under mild pyrolysis conditions, the following
data are given:
conversion at 11000C: 0.875
conversion at 1000C: 0.815
mean particle radius: 0.0064 cm
char internal surface area, A: 255 m2g-'
char feeding rate, W,: 1.067 x 103 g min-'
bulk density of char: 1.2 g cm-3
pore volume: 0.07 cm3g-'
4. Estimate the external mass-transfer limitation.

1. Conversion in the Isothermal Region A molar balance
over a small section of tubular reactor yields191
FNO + dFNo = FNO + (rNo)dWt (14)

where W, is the char weight in the isothermal region and FNO
is the molar flow rate of NO. By the definition of conver-
Chemical Engineering Education

sion, X, we obtain

dFNo = d[FNo,in( X)] = -FNOndX = +FNO,ind PNo I (15)
+ PNNOj .

Substituting Eqs. (13) and (14) into Eq. (15) and integrating

n( X) = en NO -kAWtPNO, (16)
NO,in FNO,in
Furthermore, the molar flow rate of an ideal gas mixture can
be expressed as

FNOin NO,in v (17)
P 2.445 x104

where the constant in the denominator denotes the specific
volume of an ideal gas at 250C in cm3 mole ', and v is the
total inlet volumetric flow rate measured at 250C in cm3s-.
The char weight in the reactor, Wt, in Eq. (16) can be
expressed in terms of char feeding rate, W|, through the
relation Wt=W,t,. where W, is in g s-' and th the residence
time in the isothermal section in s. Substituting the relation
and Eq. (17) into Eq. (16), with P=1 atm and v=33.33 cm3s-',
we obtain

n(1 X) = -2.445x 104kAWlth =-734 x 10kAWth (18)
This expression can be used in the estimation of reaction
rate, k, if only the conversion in the isothermal region is
2. Conversion in the Nonisothermal Regions For an ideal
gas, the gas residence time t, in s, can be expressed as a
function of temperature and distance traveled
a 1
dt = -dz = 25.62 1 dz (19)
a cross-sectional area of the reactor tube, cm2
G gas volumetric flow rate, cm3s-1
z longitudinal distance of gas traveled in the tubular reactor,
T gas temperature in the tube, K

During the heating and cooling periods, the estimated gas
temperature is approximated by two linear temperature pro-
files, as shown in Figure 1. Specifically, since the gas tem-
perature varies from 700 to 11000C in 11.9 cm, the heating
rate can be characterized as dT/dz=+33.51 C/cm. Substitut-
ing this expression into Eq. (19), we obtain

dt 0.765 dT (20)
From a material balance, i.e., Eqs. (13) through (15), we

dPNo -kAPNOindW (21)
Substituting Eqs. (17) and (20) into Eq. (21) and integrating
the resultant expression over the heating, isothermal, and
cooling sections, we obtain

( E Ea
(n(l- X) = 733.6 0.-765 WAkoe RTdT thWAkoe T

O0.765 RE
0 WIAkoe RTdT (22)

where ko and E, are the preexponential factor and the activa-
tion energy. The two unknowns in Eq. (22), ko and E,, can be
estimated by implementing it twice for two sets of tempera-
ture/conversion data. Numerically, this set of two equations
can be conveniently solved with MathCad. A temperature at
which the reaction is slow, e.g., 6000C, can be chosen as the
lower integration limit. This procedure can be repeated for
data collected over a wide range of temperatures so that
linear regression can be conducted and the average k0 and Ea
can be obtained.
3. Internal Mass Transfer Limitation To investigate the
extent of internal mass transfer limitation, the observed
Arrhenius rate in the last section can be considered the
product of the intrinsic surface reaction rate and the effec-
tiveness factor, i1, i.e., from Eq. (22)

0Ea Ea
fn(1-X)=733.6 5 WArlkoe RdT-thWArkoe RT

+ 0.765 RT)
-J-T WAkoe RTdTj (23)

The effectiveness factor is a function of the Thiele modulus,
,[0I ] i.e.,

Tw t-h Thie (24)
0 tanh ) (24)
where the Thiele modules is defined as

R akAppD

R radius of char particles, cm
a molar volume of an ideal gas, 2.445x104, cm3 mole-'
Dff effective diffusion coefficienty, cm2s-1
Pp bulk density of char, g cm3

The reciprocal of the effective diffusivity can be considered

Summer 1999

a linear combination of the resistances contributed by the
Knudsen and the bulk diffusivity['0o

-=1 -+ (26)
Deff Dk,eff D12,eff

Dk,eff = 19400| | (27)
0.0018581.5 MI+M2
SMM2 (28)
D12,eff P 2 (28

Dkff Knudsen diffusion coefficient for a porous solid, cm2s-
Di2efr bulk diffusion coefficient of species 1 in species 2, cm2s-'
0 particle void fraction
tm tortuosity factor based on the mean pore radius, assumed 2
M,,M2 molecular weights of diffusing molecules, MNo=30,
M molecular weight of the gas medium
P pressure, 1 atm
QD the "collision integral," a function of kBT / 12, dimen-
e,o force constant of the Lenard-Jones potential function, E in
g cm2s -, a in A
kB Boltzmann constant, 1.38x10 16 g cm2s 2K

Equations (26) through (28) allow calculation of the effec-
tive diffusivity. The pore fraction can be considered a prod-
uct of the pore volume and bulk density. Thus, 0 = pore
volume density = 0.084. From Bird, et al., 8S the parameters
for the transport properties can be estimated as
EHe kB=10.2K, He = 2.576 A
No/kB=119K, a N =3.47A
0 He-NO = 1/2( a He+ G NO) = 3.023 A
He-NO = (He ENO )05 = 34.84 ksg cm2 sec-2
kBT/HCe-NO = 23.83
OD = 0.6776
Substituting these constants into Eqs. (26) through (28), we
D12,ff = 8.127 cm2s-l
Dkff = 3.98 x 10-5 cm2s- for lignite char at 1100C
Df = 3.98 x 10-5 cm2s- for lignite char at 1100 C
These results show that Knudsen diffusion controls the over-
all diffusion rate.
Equations (23) through (25) can be solved simultaneously
by MathCad for the four unknowns nI, 0, ko, and Ea. Equa-
tion (23) is implemented twice for two sets of temperature/
conversion data in each calculation. The results are: Ti =0.872,
4=1.53, ko=0.07 moles s 'm2atm-, and Ea=3.80x103 cal mole-'.
The results suggest that internal mass transfer limitation is

not negligible in the reacting system. This procedure can be
repeated for data collected over a wide range of tempera-
tures so that linear regression can be conducted and the
average ko and Ea can be obtained.
4. External Transfer Limitation The overall effectiveness
factor is defined as



F=C g (30)

is an index of external mass-transfer limitations; a large
concentration gradient indicates large mass-transfer resis-
tance. In the above expression, Cg=NO concentration in the
mainstream of gas flow, in moles cm-3, and C,=NO concen-
tration at the particle surface, in moles cm 3. Under steady-
state conditions, the mass-transfer rate of NO through the
boundary layer equals the reaction rate, i.e.,

kcS(Cg -C,)=(-r)Wp (31)

k mass transfer coefficient, cm s'
S external surface area of a single particle, cm2
-r reaction rate, moles g-'s-
W weight of a single particle, g

NO reduction on the char surface has been expressed as
(-r)= TkAPNo (32)
Assuming the gas is ideal, we have

PNO =CRT (33)
and the weight of a single char particle can be expressed as
1 3
Wp pp d3p (34)

Substituting Eqs. (31-34) into Eq. (30), we obtain

-CgCs rlkARTppTrd3
F= P (35)
Cs 6kcS

Assuming that the particles are entrained in the flow, the
mass-transfer coefficient can be estimated based on the
Frossling correlation with the Sherwood number (Sh)=2,

D12Sh 2D2 (36)
dp dp

where D12=8.127 cm2s-l (see Eq. 28). The particles have a
mean diameter of 0.0128 cm; therefore, from Eq. (36),

k = 1.27 x103cms-1 (37)
Substituting this value and the parameters discussed in the
Chemical Engineering Education

last section into Eq. (35), we obtain

F= 34.34 lk

and, from Eq. (29)

n= T (39)
1+34.34 lk

In the parameter-recovery process, Eq. (39) is added into
the four-equation set discussed in Part 3, and the resultant
five unknowns have been solved by MathCad (see Figure 2).
In fact, results from the four-equation model can be used as
initial trial values for the five-equation model to warrant
convergence and save trial time. For this particular case, we
obtain nr=0.765, 0=2.294, k0=0.156 moles s lm-atm ,
Ea=3.80xl03 cal mole-', and Q=0.389. The value of f is

Wt 1.066710-3
R =0.0064
p =1.2


In( 1 x) =-733.57A 0.2 Wt.kO .- e 1.987T

In( 1 x2)=-733.57A.0.2.Wt-kO0 \ e 1 987T2

E 0.5
T1 T2
S2.445104-kOe 2 .Ap

xl =0.875
rl 0.865
pv 0.084

smaller than that of T1, suggesting that external mass transfer
limitation cannot be ignored for the reactive char chosen in
this investigation. At flame temperatures higher than 11000C,
the reaction rate will be higher and the extent of internal and
external mass transfer limitations are expected to be even

Figure 2 also contains an estimation of the contributions
from the nonisothermal regions, which have the same order
of magnitude as the conversion in the isothermal region,
emphasizing the importance of including the conversions in
the two nonisothermal regions for rate estimation.


The author wishes to thank the Pittsburgh Energy Tech-

x2 = 0.815 T1 = 1373.15
S1.6 E = 3800
S= 0.7
T1 + T2 0.5
Def 19400(p.pv)2 2
2-A-p 30

0.727336557 WtA 1kO.e 987T
T dT
3 .873.15

2733.57 0.765WtAO-kOe 1.97T
J 873.15

3 / 1 1 ,
Stanh(. ) *

1.987(T1 + T2)
1 +- rl kOe 2 J34.34

Contribution of nonisothermal
region on conversion:

kO 0.156
S 0.765
solution of the
five unknowns: I 2.294
E 3.802-103
S 0.389

-733.57A .0.2Wt.kO. T.e 1.9871



0.76Wt- ke .dT = -2.903

2 1373.15


Figure 2. Solution of reaction parameters by MathCad. The conversions are
assumed to be governed by both internal and external mass transfer limita-
tions and reaction in a three-zone tubular reactor.

nology Center, U.S. Department of En-
ergy, for financial support under grant
DE-FG22-93PC93227. Mr. Lin Tang pro-
vided valuable assistance to the data col-
lection and calculation during the course
of this study.

1. Taylor, G., "Dispersion of Soluble Mat-
ter in Solvent Flowing Slowly Through
a Tube,"Proc. Roy. Soc., A219, 186 (1953)
2. Taylor, G., "Conditions Under Which
Dispersion of a Solute in a Stream of
Solvent Can Be Used to Measure Mo-
lecular Diffusion," Proc. Roy. Soc., A225,
3. Aris, R., "On the Dispersion of a Solute
in a Fluid Flowing Through a Tube,"
Proc, Roy. Soc., A235, 67 (1956)
4. Hunt, B., "Diffusion in Laminar Pipe
Flow," Int. J. Heat Mass Transfer, 20,
393 (1977)
5. Wissler, E.H., "On the Applicability of
the Taylor-Aris Axial Diffusion Model to
Tubular Reactor Calculation," Chem.
Eng. Sci., 24, 527 (1969)
6. Wehner, J.F., and R.H. Wilhelm, "Bound-
ary Conditions of Flow Reactor," Chem.
Eng. Sci., 6, 89 (1956)
7. Sellars, J.R., M. Tribus, and J.S. Klein,
"Heat Transfer to Laminar Flow in a
Flat: The Graetz Problem Extended,"
Trans. ASME, 78, 441 (1956)
8. Bird, R.B., W.E. Stewart, and E.N.
Lightfoot Transport Phenomena, John
Wiley & Sons, New York, NY; p. 505
9. Chan, L.K., "Kinetics of the Nitric Ox-
ide: Carbon Reaction under Fluidized
Bed Combustor Conditions," D.Sc. Dis-
sertation, Chemical Engineering, MIT,
Boston, MA (1980)
10. Satterfield, C.N., Mass Transfer in Het-
erogeneous Catalysis, MIT Press, Cam-
bridge, MA (1970) p

A =255
T2 =1273.15
kO =0.07

find (kO, r E, )

Contribution of isothermal region on conversion:

Summer 1999

FM classroom



Universitat Rovira i Virgili de Tarragona 43006 Tarragona, Catalunya, Spain

he rapid social and economic changes in the world
today, triggered by the revolution in communication
technologies, is leading to a concept of the "global
engineer.""11 Institutional efforts have already begun that
will induce and accelerate change in engineering education
toward this goal.'2-61 A good example is the ABET EC 2000
criteria in the United States.[3] This concern for a change in
higher education also exists in the European Union.[2]
The key concern is how engineering schools and depart-
ments can obtain and maintain the proactive involvement
of faculty in such a change.
The purpose of the paper is to analyze the key factors in
achieving involvement and active participation of faculty in
the conception and implementation of new and effective
teaching strategies. We will analyze the impact that a bench-
mark in cooperative learning implemented at the introduc-
tory level in the chemistry program at Tarragona had on
faculty involvement fifteen years ago. We will then go on to
identify factors responsible for the decline of the professors'
interest in effective teaching when the new five-year ChE
undergraduate program was introduced in 1993 and will
describe the teaching strategies presently underway to re-
cover the holistic approach to ChE education. These specific
and open-ended learning experiences managed by profes-
sors and students working together in close collaboration
include, at the first year of undergraduate education, devel-
opment of design projects directed by fourth-year students.

Chemical engineering studies were introduced at Tarragona
in 1977 as an Industrial (ICh) option in the five-year chemis-
try program. Students accessed this ICh option after three
years of compulsory education in science and mathematics.
Chemical engineering was introduced in the third-year course
"Fundamentals of Chemical Engineering." The ICh option
included a total of five yearly courses in ChE-one in the

fourth year and four in the fifth year-together with three
yearly courses in advanced chemistry. The "Fundamentals
of Chemical Engineering" course was attended by all stu-
dents enrolled in the chemistry program. The syllabus in-
cluded two parts: macroscopic mass and energy balances,
and transport phenomena and fluid mechanics.
Chemistry students decided whether to choose the ICh
option based on their success in the introductory third-year
chemical engineering course. In the early eighties, the num-
ber of students enrolled in the ICh option showed a system-
atic tendency to decrease, and enrollment reached a mini-

Francesc Giralt is Professor of Chemical Engineering at the University
Rovira i Virgili of Tarragona. He received his BSCh from the Institut
Quimic de Sarria (Barcelona), his BSChE from the University of Barcelona,
his MBA from the ICT (Barcelona), his MASc and PhD from the University
of Toronto, and his ScD from the University of Barcelona. His research
interests range from experimental and computational transport phenom-
ena to reactor engineering and artificial intelligence.
Joan Herrero is Associate Professor of Chemical Engineering at the
University Rovira i Virgili of Tarragona, where he has taught for the past
decade. He received his BSChE and ScD degrees from the University of
Barcelona. His research interests are in transport phenomena and com-
putational fluid dynamics.
Magda Medir is Associate Professor of Chemical Engineering and Sci-
ence Education at the University Rovira i Virgili of Tarragona. She re-
ceived her BSCh from the Institut Quimic de Sarria, her BSChE from the
University of Barcelona, her MASc from the University of Toronto, and her
ScD from the University of Barcelona. She directs the Chemical Educa-
tion for Public Understanding Program in Spain and her research is in the
areas of science and engineering education.
Francesc X. Grau is Associate Professor of Fluid Mechanics at the
University Rovira i Virgili of Tarragona, where he has taught for the past
fifteen years. He received his BSChE and his ScD from the University of
Barcelona. His research interests focus on experimental and computa-
tional fluid dynamics and heat/mass transfer, with applications to environ-
mental and industrial flow problems.
Joan R. Alabart has been teaching Project Management and Total
Quality Management at the University Rovira i Virgili of Tarragona for the
last five years. He received his BSCh and PhD degrees from the Univer-
sity of Barcelona and his MBA fron ESADE. He is currently an assessor
for the European Quality Award and practices as a consultant in TQM. His
research interests include the application of TQM principles in the teach-
ing and learning processes and the modeling of organizations.

Copyright ChE Division ofASEE 1999
Chemical Engineering Education

mum of four students from a total of forty candidates in
1983. This situation could have ultimately led to the elimina-
tion of the ICh option. The faculty then decided to focus
attention on the undergraduate program and to adopt a holis-
tic and professional approach to education. It is worth noting
that none of the nine faculty members involved in this deci-
sion was yet a full professor in 1983 and that seven
of these positions were temporary Assistant Pro-
fessorships, heavily dependent on teaching needs.
Thus, the motivation to change learning strategies pur
and course development was high. [this
The first faculty decision was to abandon tradi-
tional lecturing so that different learning and teach- Ga
ing styles[7-10] could be implemented in the third- th
year course. In particular, the strategies and ac- fac
tions that were field tested had the following char- ac,
acteristics: in
Cooperative learning"111 was introduced anc
because students learn more by doing part
things than by simply hearing or seeing.'91 of fi
Professors were free to experiment and
evaluate different learning strategies, with
colleagues and students providing direct
Contents and the corresponding classwork pl
were structured as a set of activities of
carried out cooperatively by students2'31" efJ
during three hour-long sessions. lec
Students decided the objectives of each st"r
class session within the framework of the
syllabus of the course. In other words, they
decided what to do next and how to do it.
Students developed a final design project during the
second semester of the course.
A strategy of continuous evaluation, including self-
assessment, was introduced in some activities to
enhance students' involvement in the class sessions.
Implementation of the above characteristics into the "Fun-
damentals of Chemical Engineering" course favored the adop-
tion of metacognitive learning strategies.[14-l7] For example,
students had to comprehend the demands of a given task and
to respond to them during development of the different class-
room activities. Students were responsible for their own
learning and learned how to learn, i.e, they recognized and
controlled the learning opportunities and became aware of
their own learning activities. This experience also helped
them to develop other competencies such as creative think-
ing, work interdependence with personal accountability, in-
dividual responsibility, self-esteem, communication skills,
and sharing the values of the organization. Students also
developed the desire to learn continuously and to grow
professionally in a multidisciplinary environment, not
Summer 1999

i pap
is to
e kej

only in technical and scientific areas but also socially
and in humanities.
The decision to adopt a student-centered educational
model,1"7 together with cooperation from students in this
educational effort, had a remarkable impact on the student
body as a whole. The percentage of students choosing the
ICh option reached, and even surpassed, the levels
before the enrollment crisis, finding a stable plateau
around 75%. Another relevant effect was an increase
of in the performance of students while at the univer-
er sity and after graduation. Our graduates became the
preferred choice of employers at the chemical and
e petrochemical sites located near Tarragona, and to-
ny day, many of them occupy positions of responsibil-
in ity in various world-class companies.

culty in
ew and

The third-year course also induced professors to
experiment with different teaching strategies. Teach-
ing the course jointly with more experienced profes-
sors became the training policy until 1993 when the
new five-year undergraduate chemical engineering
program was implemented. During the ten-year pe-
riod from 1983 to 1993, seven professors out of the
eleven that constituted the faculty participated in the
initiative, and two-thirds of those actively involved
in the initiative exported the model of cooperative
learning to other courses taught in the last two
years of the ICh option. Therefore, the prospects
for continuous improvement of ChE education in
Tarragona were excellent just before the new pro-
gram started in 1993.

The New ChE Program The new ChE Program gener-
ated high social expectations in Catalonia in 1993. This,
together with the favorable evaluation of our ICh graduates
by industry, conveyed a perception among faculty that ev-
erything related to academia would improve continuously,
by itself, as a result of the previous momentum. The follow-
ing factors were considered particularly relevant for the ini-
tial success of our new undergraduate offer and academic
* The design and deployment ofa ChE undergraduate
program attractive to students, employees, and the
administration. The new studies were modeled after the
most successful programs in North America and
Europe, with input from world-class chemical compa-
nies and a reliance on our previous educational
The segregation of a Department of Mechanical
Engineering from the original Chemical Engineering
Department, and the creation of a School of Chemical
Engineering (ETSEQ) to provide the necessary visibil-
ity for the unique engineering educational project. A

quality program was simultaneously launched to
continuously improve the organization.
The hiring of young professors educated abroad and
interested in excellence in research, undergraduate
education, and transfer of technology to industry. The
ChE department at Tarragona was the first in Spain to
hire citizens from other European countries as profes-
Transformation of the graduate school to offer an
international doctoral program in chemical engineer-
The assignment of thefirst-year engineering science
courses to the most experienced professors, with the
purpose of reproducing throughout the academic
organization of the ETSEQ, the student-centered
approach to ChE education achieved in the old ICh
The design of the new program started in 1987, five years
before the Spanish government established chemical engi-
neering as a separate undergraduate program and degree. A
semester organization was adopted, with 55% of credits
corresponding to regular classroom hours and the remaining
45% to laboratories, modeling and computer simulations,
research and development projects, or internships in indus-
try. The main characteristics are summarized in Table 1.
Students learn physics, chemistry, mathematics, and engi-
neering science during a first two-year cycle of undergradu-
ate education and apply this fundamental knowledge to real
problems, with emphasis on process downscaling. The ob-
jective of the third year is process upscaling because it is the
core of engineering practice and technology development.
The last two years bridge fundamental education with pro-
fessional practice. This academic organization allows the
adoption of Bloom's taxonomy of educational objectives.1181
Students advance from comprehension to analysis and syn-
thesis of the essential scientific and engineering concepts in
a smooth and continuous manner within each year of study
and over the five years of the program.
The start of the new program in the fall of 1993 signifi-
cantly increased the teaching load in chemical and mechani-
cal engineering, and new faculty positions were opened to
citizens of the European Union. Since that time, all new
faculty positions have been advertised in international engi-
neering journals. This new employment policy was later
complemented with a continuing-education program aimed
at all faculty. It consisted of seminars, courses, and work-
shops conducted by specialists from North America and
Europe on topics such as learning strategies, cooperative
learning, communication skills, project management, and
quality in education. Most of these on-the-job training ac-
tivities are also open to staff and to graduate students. Courses
in Spanish and the Catalan language are also offered by the
university to help new faculty.

Lessons To Be Learned The above undergraduate ChE
program features the most widespread methods of learning
activities applied in chemical engineering education, with
project-oriented capstone courses, short stays in university
research laboratories, and industrial placement. Neverthe-
less, the first evaluation carried out in 1995, two years after
the start of the new program, showed student retention of
only 60%, lower than expected. In addition, most teaching
had reverted to lecturing.
The return to traditional lecturing by both new faculty and
the professors who had previously experienced cooperative
learning was taken very seriously by the heads of the two
departments involved because the success of cooperative
learning and of any other innovative educational method can
be fully attained only when they are applied continuously
and systematically in an undergraduate program. We cannot
expect to move from the professor-directed and professor-
centered model of education toward student-centered in-
struction171 just by endowing a few scattered courses with
active learning. An undergraduate chemical engineering pro-
gram should offer individual students the possibility of ex-
periencing their own approaches to learning, to develop
engineering and cognitive skills, and thus to become fully
accountable for, and have ownership in, all organizational
processes that are relevant to corporate life or to commercial
organizations." '6 Moreover, deficiencies accumulated in such
areas during undergraduate education are difficult to address
after graduation by continuous on-the-job training, since
corporations are reluctant to take unnecessary risks and dedi-
cate resources to nonspecific training programs.
There are several factors that explain the return to tradi-
tional education at ETSEQ after the new program was imple-
SFaculty lost focus in education due to the extra effort

Summary of Contents, Present ChE Program

First and Second Years
Introduction to ChE and Process Downscaling
Chemical Engineering and Physics (20%)
*Chemistry (9%)
*Mathematics (6%)
*Elective courses in science, mathematics and engineering (2.5%)
Elective courses in arts and social sciences (2.5%)

Third and Fourth Years
Process Upscaling and Creative Management
*Chemical Engineering (25%)
Elective courses in science, mathematics, and engineering (10%)
Elective courses in arts and social sciences (5%)

Fifth Year
Professional Practice and Internship in Industry
*Intership in industry, R&D, and final project (15%)
*Elective courses in science and engineering (2.5%)
Elective courses in arts and social sciences (2.5%)

Chemical Engineering Education

required by implementation.
The student/faculty ratio increased without augment-
ing the population of candidates and the standard of
selected students. Under these circumstances there is
a natural tendency to do whatever is more usual and
easier, which in education means lecturing.
The success inherent to implementation of the new
program could have contributed to the perception
among faculty that there was no further need to
procure social recognition and support.
The shift of faculty interests toward research instead
of excellence in education, caused by a governmental
promotion policy based solely on research productiv-
The tremendous increase in the number of professors
in chemical and mechanical engineering, together
with the incorporation of professors from other
departments to teach mathematics, chemistry,
economics, and computer science, made coordinating
and sharing experiences more difficult. Also, it was
hard to maintain the notion that faculty must act as a
team"9" and service the department through educa-
tion and research, much as we expect our students to
work and cooperate efficiently in teams.
The incorporation of young professors with a
different organizational culture was perceived as a
loss of identity by some faculty, and a natural fear of
change appeared in the organization. It should be
noted that despite the interest of new faculty in
education, they did not receive any training of
effective teaching until 1996.

Nevertheless, consideration and discussion among faculty
of all these factors was not sufficient to adopt corrective
actions and to substitute a significant percentage of lecturing
with other more efficient teaching strategies in 1995. As a
consequence, the former holistic approach of ChE education
experienced in the ICh option'[2'131 was re-examined to iden-
tify the key elements that made this previous organization
successful and sustainable. Then new elements were consid-
ered in view of the situation created by the new program, and
finally, corrective actions were taken. The following section
presents these analyses and the corrective actions, in addi-
tion to the results obtained.

Key Elements The above considerations led to the fol-
lowing synthesis of key elements to sustain cooperative learn-
ing and faculty involvement:

SThe holistic approach adopted in the old ICh option
was a shared learning opportunity for both students
and professors, with everyone assuming responsibili-
Summer 1999

* Professors experienced different learning strategies
and became confident about effective teaching.
Students felt they were a part of the organization and of
the decision-making process, with the only limitations
being imposed by the syllabus of the course.
Professor and students alike were actively and person-
ally involved in all classroom activities.

In fact, these factors implied incorporation of the Kolb
learning cycle[91 into the professor's training process, in a
learning-to-learn hands-on classroom experience. The pro-
fessors involved in the old ICh option were capable of gener-
ating new concepts from observing the students working in
teams as they advanced in the series of classroom activities
that constituted the course. From this new evidence, both
professors and students created additional actions and learn-
ing experiences. Thus, the learning cycle advanced naturally
in the course and improved the metacognitive skills to learn-
ing[14-171 of all involved. Also, a deep approach to learningg19
was favored. The experimental character of the course,
blended by the holistic approach to engineering education,
and a decision-making process shared by professors and
students constituted the core of the above four key elements.
Once the key factors of the previous experience had been
identified, the question was how to implement cooperative
learning and the holistic approach in the new program. How
could lectures and classroom activities scattered over several
one-semester courses taught by senior and junior staff from
different departments be integrated into one single educa-
tional effort? How could we reintroduce the Kolb learning
cycle into the classroom so that professors and students
become confident in their everyday learning-to-learn experi-
ence? How could faculty recover the motivation to act as a
team interested in both education and research?
Holistic Approach Within the Academic Organization *
Two fundamental actions were taken during the academic
years of 1995 through 1998 to reintroduce the holistic ap-
proach to student-centered engineering education [2'131 and to
recover the sense of teamwork among faculty. One-semester
design projects, carried out by teams of first-year students
led by junior or senior students, were introduced, and a
continuous-improvement quality program was begun. In what
follows, only the introduction of the design projects is pre-
sented and discussed.
The first-year design project had the following objectives:
To redeploy cooperative learning in the organization
as a whole.
To enhance students' responsibilities in the decision-
making processes.
To include open-ended problems where professors do
not have the only solution.
To integrate knowledge; students should be immersed

in a multidisciplinary and humanistic educational
To use faculty as a team; professors should be encour-
aged to cooperate and innovate.
To enhance junior and/or senior students' role in the
educational organization; students should learn to
lead a project, to share the culture of the organization,
and to perform professional tasks while at the univer-
To teach ChE students to think and to work as engi-
neers and to use the highest cognitive levels of Bloom's
taxonomy from the beginning of their education.
To use it as a test for evolution toward a new organiza-
tion of the curriculum in Tarragona; the test should
lead to a student-centered, sustainable educational
To foster synergetic learning and metacognition where
students help other students to learn; junior and/or
senior students also learn when coaching first-year
From an organization point of view, execution of the project
required severing class hours from the participating courses
so the total teaching load per semester was not increased.
The scope of the project had to be concrete, but open enough
to allow the eventual integration of almost all subjects and to
secure the interest of the professors. Three trials were carried
out during the second semester of the first year over the
period from 1995 to 1998. Table 2 summarizes the academic
organization under which these trials were executed.
The first project dealt with the design of a craft oven and
integrated only the first-year courses of transport phenom-
ena and fluid mechanics (see Table 3). These projects were
directed by third-years students as part of their work in the
unit operations laboratory. The second trial integrated the
transport phenomena laboratory course with teams of first-
years students led by fourth-year students enrolled in project
management and management practice courses. The project
dealt with the catalytic conversion of the lactose contained

in a water effluent from a dairy factory into glucose and
galactose. The teams presented their results during two poster
sessions at the end of the semester. The third test involved
four first-year courses, with the addition of numerical meth-
ods, and two fourth-year courses, as shown in Table 2. A
total of 145 first-year students, organized into groups of five,
were directed by 29 fourth-year students to develop the
preliminary design of a low-density polyethylene plant.
The third trial received good evaluations from all. Also,
faculty members were asked to participate in the poster
presentations to get acquainted with the new project system.
As a result, the implementation scheduled for 1998-99 will
cover a total of 13 courses and 16 faculty members (see
Tables 2 and 3). Three hours per week will correspond to
face-to-face teamwork of first- and fourth-year students in
the classroom. During the remaining hours assigned to the
project, first-year students will work on their own but with
access to the professors for consultation.

Summary of Student-Directed Design Projects

First Trial Second Trial Third Trial Implementation
topic Craft oven Lactose Polyethylene Industrial Waste
Recovery Plant Reactor Treatment Plant
hours/week 1 2 4 6
% of project in
first year 20 40 50 50
coach 3rd-yr student 4th-yr student 4th-yr student 4th-yr student
subjects involved
in first year 2 3 4 8
subjects involved
in fourth year 1 2 2 5
involved 3 5 8 16
teams 21 23 29 35
students/team 5-6 5-6 5 4
Third-year subject

First- and Fourth-Year Subjects in Current Program (Hours/Week)

First year First semester First year Second semester Fourth year First semester Fourth year Second semester
Algebra 3 Statistics -3 Process Manufact. & Control Lab 8 Chemical Process Design 4
Calculus 6 Transport Phenomena 4 (2+2') Electives 8 Economy & Industrial Organization 4
Physics 6 (4+2') Fluid Mechanics 4 (2+2') Project Management 4 (2+2') Project Management Practice I 4 (1+3')*
Chem. Eng. Funds. 4 (2+2') Transport Phenomena Lab 7 (6+1') Project Management Practice I 4(1+3')' Environmental Technology 4 (2+2')
Physical Chemistry 4 (3+1') Numerical Methods 3 (2+1") Convective Heat/Mass Transfer 4 (2+2') Electives 12
Inorganic Chemistry 4 (3+1') Analytical Chemistry 4

* hours/week assigned to design project
t hours/week offace-to-face teamwork

4R Chemical Engineering Education

The progression experienced in the degree of professors'
involvement has encouraged ETSEQ to test the same organi-
zation in the second year of undergraduate education. This
trial will be carried out with the participation of 7 subjects
and 6 professors. The second-year preliminary design project
will emphasize process downscaling and will be organized
around the two-semester chemical engineering laboratory.
A direct benefit from the experience is that we have repro-
duced, on a larger scale, the model of education that was
successful in the old ICh option. Professors become spon-
sors of cooperative learning and perform as consultants, and
their curiosity is enhanced because they are faced with real
and interesting problems. As a consequence, we are involv-
ing professors to a higher degree than before, and students
are learning what chemical engineering is all about from the
first year of their undergraduate education. From an educa-
tional point of view, the project experience is a good ex-
ample of learning synergism; first- and fourth-year students
cooperate in their own instruction and are learning together.
First-year students (with the help of fourth-year students) act
as team leaders and coaches, apply knowledge to a real
problem, develop their competency and skill, and learn how
to learn and how to cooperate with peers. Fourth-year stu-
dents integrate into the academic organization by assuming
specific responsibilities and by acting as project managers.
They reinforce their own learing-to-learn strategies when
confronted with a teaching responsibility beyond simple tu-
toring experiences. The opportunity to re-examine funda-
mental subjects or topics, to define strategies to help first-
year students overcome their learning difficulties, and to
manage a real project until completion are also invaluable
professional assets related to team management and organi-
zational behavior.

Altogether, our experience indicates that professors have a
natural tendency to teach by lecturing since that is how they
were taught. The yearly design projects compel professors,
otherwise isolated in teaching their courses, to cooperate
with other faculty and with students. This new holistic ap-
proach represents a benchmark where professors can con-
tinuously experience effective teaching and innovation in
education. The strategies have also contributed to their per-
sonal and professional development.
The current study shows the benefits of incorporating
student-centered learning into classroom instruction, the dif-
ficulties of doing so, and the ease of reverting to more
traditional and less effective approaches when emphasis on
effective teaching is relaxed. Also, it is shown that the holis-
tic approach, with first-year students working in projects led
by junior or senior students, can be easily extended to full-
scale curriculum improvement throughout an academic or-
ganization. The outburst of creativity and participation that
was attained fifteen years ago at Tarragona has been revived.
Summer 1999

This paper was made possible through the collaboration of
students and the support received from the School of Chemi-
cal Engineering, in particular from its Director, Professor
Francesc Castells. The authors would also like to acknowl-
edge the involvement of Professors J. Bonet, I. Cuesta, A.
Fabregat, J. Font, C. Garcia, A. Garijo, R.M. Gilabert, J.
Grifoll, A. Mackie, F. Medina, D. Montan6, J. Renau, J.
Sueiras, and L. Vega in the implementation of the first- and
second-year projects. Their encouragement and suggestions
have been an invaluable asset during the trial process of the
first-year projects.

1. Buonopane, R.A., "Engineering Education for the 21st Cen-
tury: Listen to Industry!" Chem. Eng. Ed., 31(3), 166 (1997)
2. "Quality and Relevance: The Challenge to European Educa-
tion, Unlocking Europe's Human Potential," Industrial Re-
search and Development Advisory Committee of the Euro-
pean Commission (1994)
3. Phillips, W., "Ensuring Engineering Excellence," ASEE
Prism, p. 40 (March 1997)
4. Becker, P., "Education Standards Key for Engineering Stu-
dents," Mech. Eng., p. 42 (June 1995)
5. Evans, J.R., "What Should Higher Education be Teaching
About Quality?" Quality Progress, p. 83 (August 1996)
6. Weinstein, L.B., J.A. Petrick, and P.M. Saunders, "What
Higher Education Should Be Teaching About Quality-But
Is Not," Quality Progress, p. 91 (April 1998)
7. Felder, R.M. and R. Brent, "Navigating the Bumpy Road to
Student-Centered Instruction," College Teaching, 44(2), 43
8. Felder, R.M., "How About a Quick One?" Chem. Eng. Ed.,
26(1), 18 (1992)
9. Stice, J.E., "Using Kolb's Learning Cycle to Improve Stu-
dent Learning," Eng. Ed., p. 291 (1987)
10. Felder, R.M., "Reaching the Second Tier: Learning and
Teaching Styles in College Science Education," J. College
Sci. Ed., 23(5), 286 (1993)
11 Felder, R.M., and R. Brent, "Cooperative Learning in Tech-
nical Courses: Procedures, Pitfalls, and Payoffs," Report to
the National Science Foundation, ERIC Document Repro-
duction Service No. ED 377 038 (1994)
12. Giralt, F., M. Medir, H. Thier, and F.X. Grau, "A Holistic
Approach to ChE Education: Part 1. Professional and Issue-
Oriented Approach," Chem. Eng. Ed., 28(2), 122 (1994)
13. Giralt, F., A. Fabregat, X. Farriol. F.X. Grau, and M. Medir,
"A Holistic Approach to ChE Education: Part 2. Approach at
the Introductory Level," Chem. Eng. Ed., 28(3), 204 (1994)
14. Nisbet, J., and J. Shucksmith, Learning Strategies, Routledge
& Keagan Paul, London (1986)
15. Schmeck, R.R., Learning Strategies and Learning Styles,
Plenum Press, New York (1988)
16. Royer, J.M., Cisero, C.A., and M.S. Carlo, "Techniques and
Procedures for Assessing Cognitive Skills, Rev. of Ed. Res.,
63, 201(1993)
17. Sobral, T.D., "Improving Learning Skills: A Self-Help Group
Approach," Higher Ed., 33, 39 (1997)
18. Bloom, B.S., Taxonomy of Educational Objectives: 1. Cogni-
tive Domain," Longman, New York (1984)
19. Rudd, W.G., and C.M. Pancake, "The Emerging Faculty as
Team," J. Eng. Ed., p. 39. January (1997) 0

S classroom



At the University of Ottawa

University of Ottawa Ottawa, Ontario, Canada KIN 6N5

he undergraduate chemical engineering curriculum
at the University of Ottawa includes an introductory
course in process dynamics and control. Taught to
third-year students during their winter semester, it consists
of three hours of lectures per week. A typical class has
twenty-five students.
As part of a University initiative to incorporate computer
technology in the classroom, I received a grant to develop a
computer-based version of this course. Using this funding,
an undergraduate student from our co-op program, Alain
Turenne, and I constructed a series of computer-based
modules and interactive simulators to allow independent
study of the course material.
With the core material now available in a self-paced, com-
puter-mediated format, I was able to rethink how I managed
the lecture hours. I had from time to time incorporated
collaborative, in-class problem solving sessions in my lec-
tures, whereby students worked together in small groups to
solve problems related to that day's topic. I decided to make
this student-driven activity the focus of all of the lecture
periods. This paper describes how I combined computer-
assisted learning with collaborative learning in this course
and presents both students' and instructor's impressions re-
garding the effectiveness of the approach.

The computer-mediated portion of the course consisted of
a series of nine modules. The material for the modules
(Table 1) was drawn largely from Thomas Marlin's text,r'
with additional material taken from other standard textbooks
in the field.[2-4' Each module provided a condensed review of
its topic, although it included more detail than one might
expect from lecture notes alone.
The modules employed interactive text, graphics, and ani-
mation to present the core material of the course. We built
these using Asymetrix, Inc.,'s authoring software, Multime-
dia Toolbook. Toolbook provides a powerful, object-based

graphical framework for producing computer-based training
software for Microsoft operating systems (Windows 3.x,
Windows 95, and Windows NT 4). Beginning with an empty
window, we added various components (such as buttons,
text fields, images, animation, etc.) to create a page within
the module. These components were then scripted to re-
spond to keyboard and mouse events (such as button clicks),
imparting to the page its interactive qualities (see Figure 1).
Finally, the nine course modules were linked through a
graphical menu (see Figure 2).
To supplement the Toolbook modules, we constructed
four dynamic simulators that students later used to explore
the effects of process-model parameters and controller-
tuning parameters on system dynamics. We built these using
Delphi, an object-oriented, visual programming language
based on Pascal and produced by Inprise (formerly Borland)
Corporation. The Delphi development environment is simi-
lar to that of Toolbook; programmers add visual components
to empty windows and then write the requisite code for these
components as well as for the numerical routines. The final
simulators included a graphical user interface through which
students could adjust model parameters and view plots of the
process response (see Figure 3). Further, students could run
these simulators directly from the modules.
We designed and constructed the modules and simulators
over a one-year period prior to introducing them into the
course, after which time we installed them on a PC network

An Associate Professorof Chemical Engineer-
ing at the University of Ottawa, David Taylor
received his BASc in Engineering Science at
the University of Toronto and his PhD in
Chemical Engineering from the University of
British Columbia. His research focuses on tis-
sue engineering, process modeling, and com-
puter simulation. He is also keenly interested
in distributed and computer-mediated learn-

Copyright ChE Division of ASEE 1999
Chemical Engineering Education

Module Content

1. Introduction to Process Control

2. Mechanistic Modeling

3. Empirical Modeling

4. Analyzing Process Dynamics

5. The Feedback Control Loop

* definitions
* principal control components
* feedback vs. feedforward control
* calculating control benefits
* modeling methodologies
* defining a modeling approach
* developing the model
* motivation for empirical modeling
* procedural approach to empirical modeling
* statistical model building
* dynamics of linear Ist- and 2nd-order systems
* SISO and MIMO systems
* Laplace domain for linear systems
* transfer functions and block diagrams
* frequency domain for linear systems
* the feedback loop
* process elements and instrumentation
* block diagrams revisited


6. PID Controllers

7. Stability Analysis

8. Tuning PID Controllers

9. Digital Control and Filtering

* control performance measures
* the feedback loop revisited
* proportional mode
* integral mode
* derivative mode
* PID control

* stability and process control
* stability criterion
* Routh analysis
* direct substitution method
* frequency response analysis
* considerations and criteria for tuning
* Ciancone correlations
* Ziegler-Nichols correlations
* issues of fine tuning
* digital feedback control algorithms
* signal filtering
* valve control and failure modes

located in the engineering building at the University
of Ottawa. The classroom housing these PCs was
designated a "quiet room" so that the students could
study the modules at their convenience. In addition,
I constructed a course web site from which the
students could download the modules to run from
home. The site also contained supplementary course
material and a Java applet for retrieving marks on-

A typical single semester course in chemical en-
gineering at the University of Ottawa consists of
three lecture hours per week; these are normally
delivered in two ninety-minute sessions. In the pro-
cess control course, however, I combined the two
weekly lectures into a single, three-hour session.
While one would expect a lecture of this length to
tax even those with ironclad concentration, it proved
well suited to the collaborative, problem-based ses-
sions used in this course.
Each week the students were given an
assignment that was due two days before
the next lecture. In addition, they were
assigned a module (or portion thereof) to
review for the following week. Each stu-
dent was asked to submit, together with
his/her assignment, a review sheet that
highlighted any confusion with the sub-
ject matter contained in that week's mod-
ule. The review sheet also contained space
for the student to provide feedback re-

Summer 1999

Figure 1. A screen capture of a page from one of the course
modules. The pop-up box is animated.

Figure 2. A
capture of

garding the design of the module.
Prior to each weekly session, I would look over the review
sheets, noting the areas of concern raised by the students. I
would then design a set of short questions (normally requiring
no more than fifteen minutes to complete) and a brief fif-
teen-minute lecture that focused on the current module's
content and addressed those problem areas identified by the
The three-hour session would start with the prepared lecture,
after which time students would move into small groups. (These
groups were formed at the beginning of the term.) Each group
would be assigned the same in-class problem and given a short
period of time to work on it. During this time I would move
through the class and provide limited guidance where needed. I
would then randomly select one group from the class to present
its solution to the problem. While no mark was given for "right"
answers, the gfoup was evaluated on its understanding of the
problem and its ability to formulate a solution method. In this
way the group presentation served as a springboard for class
discussion regarding the problem and its solution. A typical three-
hour lecture would include several of these exercises.

The Students' Perspective
Since the standard course evaluations issued by the Univer-
sity of Ottawa do not directly address matters concerning course
delivery modes, I undertook my own student evaluation ap-
proximately half-way through the term. Assessing the students'
reaction to the new teaching style at this stage in the course also
permitted me time to make any changes that seemed necessary
from the students' perspective.
In addition to the standard questions appearing on the Univer-
sity of Ottawa's form, the evaluation included three questions
relating to the new teaching style. The first of these read
Compared to other lecture styles that I have experienced I find
the approach in this class to be..."
and offered five choices from excellent to very poor for the
student designation. The second question was

Question I



10a 2
5 4 3 2 1

Question 2

Ifind the time spent on in-class problem solving to be...
with choices of "Very Helpful" to "A Total Waste" on a
scale of five. The final question
Ifind the computer-based modules as learning aids ...
offered the same five choices listed under the first question.
The thirty anonymous responses to these three questions
are presented in Figures 4a, b, and c, respectively. Overall,
the students preferred this form of lecture to the standard,
passive approach. As the responses to question 2 demon-
strate, students particularly enjoyed the opportunity to apply
their problem-solving skills in a structured, professor-medi-
ated format. At least one student also saw value in having to
present a solution to the rest of the class, as (s)he noted in the
following comment:
Although I do not particularly like speaking in front of the
class, I find presenting assignments and in-class problems
useful. It forces me to keep up with the course material so that
I know what I'm talking about when presenting.
Of course, not all students saw it the same way. Another said
I think in-class teaching should be more emphasized since it
benefits the whole class, rather than in-class problems which

Figure 3. A screen capture of one of the dynamics
simulators accompanying the course modules.

Question 3
aS n


5 4

3 2 1

Figure 4. Student responses to questions 1, 2, and 3 of the midterm evaluation.

152 Chemical Engineering Education


usually benefit the group doing the problem.
(I found this remark puzzling, since all groups were required
to complete each question and since no group knew ahead of
time who would be presenting the solution to the class.)
The overall response to question 3, though favorable, sug-
gests that students were somewhat less enthusiastic about
having to review the modules each week. The weekly feed-
back regarding the modules' structure and content was gen-
erally positive, leaving me to suspect that some students
simply objected to having to spend time reviewing them
outside of class. (It is worth noting that the average module
would require less than three hours to complete, implying
under 2.5 review hours per week for the course.) Further,
since they could not print out the modules' content directly,
students had to generate their own course notes. Not surpris-
ing, several students stated on the evaluation sheet that they
would have preferred the option to print out the module con-
tent. But this limitation was deliberate, since I felt that the note-
taking was valuable from a pedagogical perspective.

The Professor's Perspective

Admittedly, the combined collaborative, computer-medi-
ated learning methodology used here required more work
outside of the classroom: customizing each weekly session
to address student comments, maintaining the course web
site, and updating the course modules. But I did not find that
any of these tasks required an inordinate amount of time.
The most significant changes for me as instructor lay in
the classroom. First, the class became student-centered rather
than professor-centered. This altered my role significantly
from one of lecturer to one of facilitator-it also raised the
students' level of interest during the lecture. The students
seemed to quickly adapt to their new, and more prominent,
role; in fact, the overall mood in the classroom was very
positive. (It is worth noting that none of the students com-
plained about the length of the weekly sessions, even though
these sessions were twice as long as the standard lecture.)
Attendance was also exceptionally good. Whether this was
due to increased interest among the students in the course, or
due to the fact that students were graded on their participa-
tion in the problem sessions, or attributable in varying de-
grees to both, is unknown. But the atmosphere within the
class was significantly better, compared to other courses that
I have taught using a more formal lecturing approach. Stu-
dents in this class were far more inquisitive, eager to ask
questions relating not only to the in-class problems, but to
broader issues of process control. They were also far less
hesitant to seek clarification during the lecture.
Clearly, this form of collaborative learning does not need
to be supplemented with computer-based learning. The value
of the modules, to my thinking, lay with their ability to
present the conventional course material more effectively
than paper handouts. In particular:
Summer 1999

1. Being interactive, the modules engaged the students in the
learning process (this is particularly true of the simulators,
which allowed the students considerable freedom to explore
the causal relationships in feedback control loops).
2. They allowed students to process information in a nonlinear,
and consequently more flexible, fashion. For example,
hyperlinks in the modules give students the choice of either
delving further into a topic or continuing to the next one,
without compromising the flow of the overall presentation.
3. They integrated several media forms (text, graphics, and
animation) to present the subject matter, thereby offering
various perspectives on the same topic.

The modules therefore provided added incentive for stu-
dents to review the material ahead of time, which in turn
contributed to the success of the classroom sessions.

Collaborative learning is well established as an effective
method for teaching engineering students. For example,
Felder included in-class, small-group problem solving as
part of a longitudinal study of student learning styles. 51 In
that paper, he notes that students' evaluations were "consis-
tently and overwhelmingly positive" and that their perfor-
mance was significantly better. Of course, Felder's study
incorporated cooperative learning, which goes well beyond
the small group sessions employed here. But improved stu-
dent attitudes that he observed are consistent with this study
as well.
Finally, the teaching technique applied here requires that
students undertake more independent study than might be
expected with conventional lecturing. Computer-mediated
delivery of the course material provides an attractive alterna-
tive to notes in this case, offering new approaches for stu-
dents to assimilate both theoretical concepts and their rami-
fications in practical engineering problems.

The author wishes to acknowledge Mr. Alain Turenne, whose
work during the design and construction of the computer-based
modules is greatly appreciated. The development of these modules
was made possible through a grant from the University of Ottawa.

1. Marlin, T.E., Process Control Designing Processes and Con-
trol Systems for Dynamic Performance, McGraw Hill (1995)
2. Seborg, D.E., T.F. Edgar, and D.A. Mellichamp, Process
Dynamics and Control, Wiley (1989)
3. Smith, C., and A.B. Corripio, Principles and Practice of
Automatic Process Control, Wiley (1985)
4. Stephanopoulos, G., Chemical Process Control: An Intro-
duction to Theory and Practice, Prentice-Hall (1984)
5. Felder, R.M., "A Longitudinal Study of Engineering Stu-
dent Performance and Retention: IV. Instructional Methods
and Students' Responses to Them," J. Engr. Ed., 84(4), 361
(1995) 3


, Oclassroom



University of New Hampshire Durham, NH 03824

EconExpert is a software package for capital cost esti-
mation, primarily intended for use by chemical engi-
neering students in plant design, and can be run on
any Unix platform. The system prompts the user for various
input, such as equipment category, equipment type, equip-
ment sub-type, material of construction, and operating pres-
sure, and then calculates the bare module cost of the equip-
ment. EconExpert is also programmed to calculate the grass-
roots capital if the user enters information on all relevant
equipment in a plant. It also provides the user with help.

EconExpert is an interactive software package for capital
cost estimation. The system was developed in CParaOPS5
and uses 'C' external functions. It uses cost data from A
Guide to Chemical Engineering Process Design and Eco-
nomics[11 and is thus a useful supplement to the text. The cost
data in the text are expressed graphically in the form of
charts, and according to the author, "the charts are accurate
enough for preliminary design estimates and are certainly
adequate for classroom work." In this software package, the
plots are represented as polynomial equations, and these
equations are stored as 'C' functions. If the purchase cost
is a function of more than one variable, a multiple regres-
sion technique is used to fit the data. The output from the
routine results in a polynomial equation that gives the
closest fit to the data.
OPS5, developed at Carnegie Mellon University in 1978
by Charles Forgy,t2] is a language especially developed for
rule-based expert systems. CParaOPS5 is a parallel version
of OPS5, written in 'C' programming language. One advan-
tage of CParaOPS5 is that it has parallel processing capabili-
ties in addition to the ability to run on a uniprocessor ma-
chine. It also allows the user to write external functions in
'C' language, which is an added advantage for calculation-

intensive applications since OPS5 is not good in mathemati-
cal computing. Another advantage is portability of the pro-
gram. CParaOPS5 converts the OPS5 program to 'C' lan-
guage, and this 'C' language program can be compiled on
any 'C' compiler without needing CParaOPS5.
In addition to the above capabilities, another important
feature of CParaOPS5 is that it allows the database to be
included in a separate file. The advantage is that the database
can be accessed and modified by the user at any time.
CParaOPS5 also allows the input to be read from a file and
the output to be written to files, which makes it easier for the
user to supply data and to obtain results.

The system consists of a user interface and an extensive
knowledge base. This knowledge is represented in OPS5
rules and 'C' external functions. The OPS5 rules mainly
ascertain the type of equipment and desired material of con-
struction, and also provide help. In addition, the OPS5 rules
contain information on the 'control strategy.' The 'C' func-

P.T. Vasudevan is Associate Professor of
Chemical Engineering at the University of New
Hampshire. He has taught a number of courses
that include process control, biochemical engi-
neering, reaction kinetics, mass transfer and
polymer engineering. His research interests are
in the area of catalysis and biocatalysis.

Deepak Agrawal graduated with a MS degree
in Chemical Engineering from the University of
New Hampshire in the area of Expert Systems.
He is a management consultant helping leading
global companies with strategic and tactical is-
sues in the areas of supply chain, operations,
and product development

Copyright ChE Division of ASEE 1999

Chemical Engineering Education

tions contain information on cost estimation, information on
the minimum and maximum sizes available, and information
on other key factors.
The system consists of several levels. At the upper level,

basic information is sought from the user.
After it is supplied, the system moves to the
next level to gather more detail information.
The levels are


Level I
Level H
Level III

Starts execution of the
program and asks about cost
Asks about equipment category
Asks about equipment type
Asks about equipment sub-type
and calculates purchase costs

Level IV Asks about material of
construction and other
information, calculates various
factors, and uses these factors
to calculate bare module
factor and bare module cost
Final Level Adds bare module cost of the
equipment to the total bare
module cost, asks if the user
wants to cost one more
equipment. If "yes," then goes
to Level I-otherwise
calculates total capital cost
and ends the session.

The structure
described more
* Start

of the software package is
fully in the following sec-

Once the program is started, it prints a wel-
come message, initializes the variables, and asks for the cost
index. After the cost index is supplied, the information is
stored for the duration of the session, avoiding the need to
seek this information while costing several pieces of equip-
The cost index is asked by the rule "ask-cost-index," which
in turn calls an external 'C' function. This function asks for
the cost index, accepts the input, and returns the value to an
OPS5 rule.

* Level I
In this level, the user is prompted to choose the equipment
category of interest to the user. There are 18 broad catego-
ries of equipment such as heat exchangers, pumps, mixers,
etc. When the user supplies the desired information, the
system checks the answer and transfers control to the next
level. The validity of the input is checked by another rule,
and if the answer is valid, the system moves to the next level
Summer 1999

The sys
prompts th
for various
such as equ
type, mate
pressure, a
calculates t
module cos
if the user

in the hierarchy. If the answer is invalid, the system prompts
the user again for a different input.
* Level II
After the equipment has been selected, the system moves
to this level to obtain additional information.
There is one rule for every category, and each
tem rule contains information on the type of equip-
ment. For example, under "mixers" there are
Ge user
several types available. The system needs to
input, ascertain the user's interests and hence it asks
pipment about the desired "type." When the system
ir receives the input, it validates it by using
t type rules written for checking invalid inputs.
it sub- These rules are written in such a way that
Srlo- they can be shared by various categories. Af-
rial of ter the system gets the correct information, it
In, and goes to the next level to seek additional input
ing from the user.
d then
he bare 0 Level III
t of the After the system receives input on the type
oef of equipment, it queries the user about the
en."sub-type" desired. For example, sub-type
t is also could be items such as "stuffing box" or "me-
ned to chanical seal." When this information is ob-
e the tained, it requests further information relat-
capital ing to that particular equipment, such as size,
area, etc., in order to calculate the purchase
enter cost. This is done by an external function
n on all written in 'C.' This function has information
rut on the minimum and maximum sizes avail-
t in a able for the equipment. If the size specified
t. by the user is less than the minimum size
available, it increases the size to the mini-
mum size; if the specified size is larger than
the available maximum size, EconExpert splits the unit into
smaller equal-sized units. This is a simplistic approach, but
this rule can be easily changed as more knowledge becomes
available. The relationship between purchase cost and size is
expressed in the form of equations, using 'C' functions.
After the functions calculate the purchase cost, the system
moves to the next level.
Certain types of equipment do not possess a "sub-type." In
these cases, the tasks of Levels II and III are combined; that
is, after determining the type of equipment, the system di-
rectly proceeds to the cost-calculation step.

Level IV
The cost of a piece of equipment depends on the construc-
tion material. There are factors that take into account the
effect of construction material on the equipment cost. In this
level, the system asks the user for a preferred material of
construction and then calculates the appropriate factor. This

level also has information for validating the input.
In some cases, the cost of equipment varies depending on
the operating pressure. This is taken into consideration by
employing a "pressure factor." Additionally, the system has
information on several other factors such as superheat, cor-
rosion, etc. These factors are also calculated in this level.
When the system has determined all these factors, the pur-
chase cost is multiplied by an overall factor, obtained by
multiplying all the relevant factors, to give the bare module
cost. After the system has calculated the bare module cost of
the equipment, it moves to the last level.

* Final Level
In this level, the system keeps track of the total bare
module cost of the plant by adding the total bare module cost
of each piece of equipment. The system thus asks the user
about costing new equipment, and if the answer is in the
affirmative, the system goes to Level I again. If not, it
calculates the total module cost of the plant and the grass
roots capital. Once the system has calculated the grass roots
capital of the plant, it prints the results and the session ends.

* Help Facility
The system also provides help if requested. When the user
prompts the system for help, it lists the different types of
equipment available in a particular category. If the user asks
for help in Level I, for instance, the system will print the list
of available categories and will then prompt the user for
input. The system provides appropriate help at each level
and contains rules for transferring control between any level
and the help facility, and vice versa. These rules are written
in a way that can be shared by all categories.

After development of EconExpert, the system's perfor-
mance, reasoning, and knowledge were tested and validated.
It was used by all seniors in the spring of 1998, and the bugs
reported by them have all been corrected. The program was
tested to check for the validity of its reasoning and also for
its ease of use and aesthetic appeal.
We have used the software in other courses, such a bio-
chemical engineering and thermodynamics. Thus, the soft-
ware aids students not only in the senior capstone design
class, but also in other courses where a quick economic
evaluation of a chemical process plant or unit is required.
The software has helped instructors in different courses to
obtain relative-cost data. Students have found the program
to be quite useful since it provides them with a quick means
of determining the cost of equipment to obtain total capital
cost, and in making economic evaluations of different pro-
cesses or competing technologies.
The software is available on the University mainframe

computers as well as Linux boxes, and students thus have
ready access to it. Students have been uniform in their praise
of the software since it is so simple to use and they can
access it from anywhere in the University or from their
homes. The software was used in the spring 1999 semester
by seniors taking design and by juniors taking thermody-
namics. Instead of spending an inordinate amount of time in
obtaining cost data from the plots (which required careful
interpolation, especially since the plots are rather small), the
students are able to focus on the design problem itself and to
quickly evaluate different flowsheets (technologies) or dif-
ferent options (equipment types) within a given process.
Thus the pedagogical value of the package is quite high.
The system was tested by having users try it and noting
their responses. EconExpert was tested for all the equipment
it contains. The bare module cost supplied by the system is
in agreement with values reported in the literature. The
system is successfully carrying out all the tasks such as
checking size, providing help, validating input, etc. The
system also has the flexibility to be continuously updated as
data become available in the future.

EconExpert is a software package for capital cost estima-
tion. It is intended for use by students in chemical engineer-
ing and has a number of features, including
3 The movement of the system is efficient since it
moves from a node at one level to a node at the next
level, until it tracks down the right equipment. The
way this system is written, it finds information about
the equipment using very little computer time.
3 It checks for the minimum and maximum sizes
available. If the desired size is out of range, it
knows what to do.
3 It has a help facility to assist the user.
3 It uses 'C'functions extensively since the design
process is highly computational. The calculation
part is encoded in 'C' functions to increase the
efficiency of the system.
3 The system has the capability of checking user input
at every level.
3 The executable code runs on any Unix platform.
Readers may send e-mail to the author
( for more information about
the software and its availability.

1. Ulrich, G.D., A Guide to Chemical Engineering Process De-
sign and Economics, John Wiley & Sons, New York, NY
2. Forgy, C.L., OPS5:A User's Manual, Technical Report CMU-
CS-81-135, Department of Computer Science, Carnegie
Mellon University, Pittsburgh, PA (1981) 0

Chemical Engineering Education


on the

at c c/

Full Text

xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EXVFO1KNJ_W2J3VQ INGEST_TIME 2012-02-17T16:57:42Z PACKAGE AA00000383_00143