Chemical engineering education

BriceDCarnahan
Dand

^^^^^^^ByCJames 0. WilkesB

featurearticle

ACKNOWLEDGEMENT

DEPARTMENTAL SPONSORS

The following 160 departments contribute to the support of CEE with bulk subscriptions.

If your department is not a contributor, write to
CHEMICAL ENGINEERING EDUCATION
c/o Chemical Engineering Department University of Florida Gainesville, FL 32611-6005
for information on bulk subscriptions

University of Akron
University of Alabama
University of Alberta
University of Arizona
Arizona State University
University of Arkansas
Auburn University
Ben Gurion University of the Negev
Brigham Young University
University of British Columbia
Brown University
Bucknell University
University of Calgary
University of California, Berkeley
University of California, Davis
University of California, Irvine
University of California, San Diego
University of California, Santa Barabara
California Institute of Technology
California State Polytechnic Institute
California State University
Carnegie-Mellon University
Case Western Reserve University
University of Cincinnati
Clarkson University
Clemson University
Cleveland State University
University of Colorado
Colorado School of Mines
Colorado State University
Columbia University
University of Connecticut
Cork Regional Technical College
Cornell University
Dartmouth College
University of Dayton
University of Delaware
Drexel University
University of Edinburgh
University of Florida
Florida Institute of Technology
Florida State/Florida A&M University
Georgia Institute of Technology
University of Houston
Howard University
University of Idaho
University of Illinois, Chicago
University of Illinois, Urbana
Illinois Institute of Technology
University of Iowa
Iowa State University
Johns Hopkins University
University of Kansas

Kansas State University
University of Kentucky
Lafayette College
Lakehead University
Lamar University
Laval University
Lehigh University
Loughborough University
Louisiana State University
Louisiana Technical University
University of Louisville
University of Maine
Manhattan College
University of Maryland
University of Maryland, Baltimore County
University of Massachusetts
University of Massachusetts, Lowell
Massachusetts Institute of Technology
McGill University
McMaster University
McNeese State University
University of Michigan
Michigan State University
Michigan Technical University
University of Minnesota
University of Minnesota, Duluth
University of Mississippi
Mississippi State University
University of Missouri, Columbia
University of Missouri, Rolla
Montana State University
University of Nebraska
University of New Hampshire
University of New Haven
New Jersey Institute of Technology
University of New Mexico
New Mexico State University
North Carolina A & T University
North Carolina State University
University of North Dakota
Northeastern University
Northwestern University
University of Notre Dame
Technical University of Nova Scotia
Ohio State University
Ohio University
University of Oklahoma
Oklahoma State University
Oregon State University
University of Ottawa
University of Pennsylvania
Pennsylvania State University
University of Pittsburgh
Polytechnic Institute of New York

Princeton University
Purdue University
Queen's University
Rensselaer Polytechnic Institute
University of Rhode Island
Rice University
University of Rochester
Rose-Hulman Institute of Technology
Rowan College
Rutgers, The State University
San Jose State University
University of Saskatchewan
University of Sherbrooke
University of South Alabama
University of South Carolina
South Dakota School of Mines
University of South Florida
University of Southern California
University of Southwestern Louisiana
State University of New York, Buffalo
Stevens Institute of Technology
University of Sydney
Syracuse University
Technion-Israel Institute of Technology
University of Tennessee
Tennessee Technological University
University of Texas
Texas A & M University, College Station
Texas Tech University
University of Toledo
Tri-State University
Tufts University
University of Tulsa
Tuskegee Institute
University of Utah
Vanderbilt University
Villanova University
University of Virginia
Virginia Polytechnic Institute
University of Washington
Washington State University
Washington University
University of Waterloo
Wayne State University
West Virginia Graduate College
West Virginia Institute of Technology
West Virginia University
Widener University
University of Wisconsin
Worcester Polytechnic Institute
University of Wyoming
Yale University
Youngstown State University

EDITORIAL AND BUSINESS ADDRESS:
Chemical Engineering Education
Department of Chemical Engineering
University of Florida Gainesville, FL 32611
PHONE and FAX : 352-392-0861
e-mail: cee@che.ufl.edu

EDITOR
T. J. Anderson

ASSOCIATE EDITOR
Phillip C. Wankat

CONSULTING EDITOR
Mack Tyner

MANAGING EDITOR
Carole Yocum

PROBLEM EDITORS
James 0. Wilkes and Mark A. Burns
University of Michigan
LEARNING IN INDUSTRY EDITOR
William J. Koros
University of Texas, Austin

PUBLICATIONS BOARD --
CHAIRMAN
E. Dendy Sloan, Jr.
Colorado School of Mines

PAST CHAIRMEN *
Gary Poehlein
Georgia Institute of Technology
Klaus Timmerhaus
University of Colorado

MEMBERS *
Anthony T. DiBenedetto
University of Connecticut
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
Angelo J. Perna
New Jersey 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

Chemical Engineering Education

Volume 30 Number 3 Summer 1996

> EDUCATOR
162 Brice Carnahan and James 0. Wilkes
of the University of Michigan

> ESSAY
168 Evolution for Chemical Engineers, E. N. Lightfoot

> LEARNING IN INDUSTRY
174 Industry, Academe, and Government: Building a New Relation-
ship, James A. Trainham, Arnold M. Eisenberg

> CLASS AND HOME PROBLEMS
180 "An Ode to That Distillation Tower" and Other Poetry: A
Creative Writing Assignment,
Gregory L. Rorrer

> SURVEY
184 The Chemical Engineering Curriculum-1994,
Ronald N. Occhiogrosso, Banta Rana
190 Teaching Colloid and Surface Phenomena-1995,
Donald R. Woods, Darsh T. Wasan

> RANDOM THOUGHTS
188 If You've Got It, Flaunt It: Uses and Abuses of Teaching
Portfolios, Richard M. Felder, Rebecca Brent

> CURRICULUM
198 Integrating New Separations Technologies into the
Undergraduate Curriculum,
Pamela M. Brown
220 Comparison of GAMS, AMPL, and MINOS for Optimization,
Xueyu Chen, Krishnaraj S. Rao, Jufang Yu, Ralph W. Pike

> CLASSROOM
204 Implementation of Multiple Interrelated Projects Within a Senior
Design Course, John T. Bell
210 Wake-Up to Engineering! Robert P. Hesketh
214 ChE Applications of Elliptic Integrals,
Peter W. Hart, Jude T. Sommerfeld
228 Problem-Centered Teaching of Process Control and Dynamics,
Paul Lant, Bob Newell

> 173 Book Review
0 183 Stirred Pots

CHEMICAL ENGINEERING EDUCATION (ISSN 0009-2479) is published quarterly by the Chemical Engineering
Division, American Society for Engineering Education, and is edited at the University of Florida. 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 1996 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, which body assumes no responsibility for them. Defective copies replaced
if notified within 120 days of publication. Write for information on subscription costs and for back copy costs and
availability. POSTMASTER: Send address changes to CEE, Chemical Engineering Department., University of Florida,
Gainesville, FL 32611-6005.

Summer 1996

e educators

Brice Carnahan

and

James 0. Wilkes

of The University of Michigan

JIM WILKES: THE EARLY YEARS

Jim was born in Southampton, England, in 1932. During the
Second World War, his hometown was badly bombed by the
Germans from 1939 onwards (an incendiary bomb landed on
his house but failed to ignite), and he was soon evacuated to
live in Shropshire with his mother and grandmother for the
duration of the war, until 1945.
Shropshire-on the Welsh border-was, and is still, a very
quiet county, little frequented by overseas visitors. Its rolling
hills are prime sheep country, and it is immortalized in A.E.
Housman's A Shropshire Lad, which refers to four of the
villages well known to Jim: "Clunton and Clunbury,
Clungunford, and Clun/Are the quietest places under the sun."
As a scholar of Emmanuel College, Jim obtained his
bachelor's degree in chemical engineering from the University
of Cambridge in 1955. The English-Speaking Union awarded
him a King George VI Memorial Fellowship to the University
of Michigan, where he received his master's degree in 1956.
He and his wife Mary Ann were married in St. Andrew's
Church, Ann Arbor, in 1956.
Jim returned to England for a four-year stint as a faculty
member at the University of Cambridge, coming back to Michi-
gan in 1960 to study for his PhD with Stuart Churchill. His
dissertation, "Finite-Difference Computation of Natural Con-
vection in an Enclosed Rectangular Cavity," was published in
1963. He has been a faculty member at the University of
Michigan since 1960.
Copyright ChE Division ofASEE 1996

Chemical Engineering Eduction

It is unusual for a single issue of CEE to feature two chemical engineering educators, but
Professors Brice Carnahan and Jim Wilkes have worked closely together for the past thirty-six
years and have shared several achievements during that period. They have also made individual
impacts of their own in chemical engineering education and research. These two men also present
an interesting contrast of personalities-Brice always bubbling over with good humor, very active
in professional societies, enjoying sunshine vacations, urban settings, and being with the crowd;
Jim being more reserved, devoting much of his energy internally at the University of Michigan and
enjoying vacations in remoter (and often colder) regions with his wife, Mary Ann.

BRICE CARNAHAN: THE EARLY YEARS

Brice was born in New Philadelphia, Ohio, in 1933, the lowest birth-
rate year in the 20th century, thus guaranteeing small classes from
kindergarten through college (and also making it easier to get ahead!).
In his first appearance in public print in 1939, Brice appears in a local
newspaper photograph as a member of Mrs. Dennison's kindergarten
kazoo band, the first of many bands/orchestras in which he played
clarinet (badly) during his New Philadelphia years. Thus began a life-
long interest in music and in the latest and best (and loudest) electronic
sound equipment.
His high-school chemistry teacher, Lila Helmick, was a strong influ-
ence on Brice and helped him obtain scholarship offers from two nearby
(but far enough away to escape small-town boredom) "big-city" engi-
neering schools-Carnegie and Case. He chose Case Institute of Tech-
nology and received his BS and MS degrees in 1955 and 1956, respec-
tively. As part of his scholarship/fellowship support from General Dy-
namics Corporation, Brice worked for several summers at the atomic
submarine plant, principally on design and testing of very compact
atmosphere control equipment.
As an extension of his interest in things nuclear, Brice's doctoral
research was on radiation-induced cracking of paraffins, under the su-
pervision of Joe Martin. His conclusion: this is a very expensive way to
crack hydrocarbons. At one point during his experimental work, he
managed to contaminate an entire engineering building with a weak
gamma-emitting silver nuclide of 270 day half-life, after which he was
known as the "silver kid." No doubt this led to his prematurely white
hair and a preference for non-experimental work!
Between 1959 and 1965, Brice worked closely with Professor Donald
Katz, first as technical director of the Ford Foundation project, "Com-
puters in Engineering Education," and then as Associate Director of a
follow-on NSF project, "Computers in Engineering Design Education."
Since 1960, with brief (sabbatical leave) stints as a visiting faculty
member at the University of Pennsylvania, Imperial College, and the
University of California at San Diego, he has been on the faculty of the
chemical engineering department at Michigan.

Summer 1996

DON KATZ'S INFLUENCE
In 1959, Professor Donald L. Katz (then chairman of
chemical engineering at the University of Michigan) fore-
saw the tremendous impact that computing would have on
engineering practice. He convinced the Ford Foundation to
support a feasibility study of broad-scale integration of com-
puter use into the undergraduate engineering curricula. In a
three-year period, over 200
faculty from nine engineer-
ing disciplines and 65 engi-
neering schools participated
in the various activities of
the Michigan project; they
jointly produced many use-
ful reports that were widely
distributed to other faculty.
Brice's first contact with
the Ford Foundation project
occurred in the summer of
1959 when Don offered him
a full-time job with the
project. Brice's acceptance -
put his doctoral thesis "on
hold" and delayed his PhD
by "an unconscionable Jim and Brice with the mai
number of years." But he
never regretted the deci-
sion-it provided opportunities that he would not have oth-
erwise had and steered him toward an academic career that
has brought him much pleasure. The principal recommen-
dations of the Ford project were to
Train faculty to use computers
Provide "free" time-shared computing services to
all students
Require a computer-programming course
Teach numerical and optimization methods
Integrate computing assignments into all engineer-
ing, science, and design courses
Stress design-like (now called "open-ended")
problems throughout the curriculum
Most of these recommendations are still on the mark-
thanks in large part to Don Katz's foresight and to Brice's
attention to detail, hard work, and ability to clearly and
directly communicate essentials to others.

A PROFESSIONAL
LIFETIME OF COLLABORATION
Numerical Methods Nationally, Brice and Jim are
probably best known for their coauthorship of Applied Nu-

ny b

medical Methods. The venture was conceived in typical
style by Don Katz, who suggested near the end of the Ford
Foundation project that Brice and Jim write up "a few
notes" on numerical methods for computers. They were
joined by mathematician Professor H.A. Luther from Texas
A&M University. After eighteen months of very hard work,
a paperback preliminary edition of "ANM" was published
locally in 1964; it contained eight chapters, 790 large (8 1/2
by 11) pages, and 47 com-
pletely documented com-
puter programs illustrating
the various techniques. It
also included a significant
h appendix on the "MAD"
(Michigan Algorithm De-
coder) language (an Algol
60 derivative), which was

Grams. A hardcover edi
tion of just over 600 pages
(again in a large format),
illustrated with 40 FOR-
,- TRAN programs, was fi
nally published by John
Wiley & Sons, in 1969,
ooks they have coauthored. and was very popular na-
tionally for the following
twenty years.
Freshman Computing For various extended periods
since 1967-and continuously since 1981-Brice and Jim
have been responsible for organizing and supervising the
freshman engineering digital-computing courses at the Uni-
versity of Michigan. The enterprise has grown in magnitude
and complexity, to the point where it has occupied about
half of their professional time for the past decade. These
courses are now taught, very successfully, by an all-student
cadre of instructors. Typically, about 1,100 students enroll
each year in about 30 sections of four different courses;
over the years, Brice and Jim have directly impacted per-
haps 30,000 University of Michigan freshmen through these
courses. Very frequently-sometimes annually-they have
updated their two books for use in these freshman courses,
the most recent titles being FORTRAN for the Macintosh
and IBM PS/2 (1994) and The Macintosh, the PC, and Unix
Workstations: Operating Systems and Applications. (1995).
In all, there have been 27 different editions of these two
texts or their predecessors, some of which are shown in the
photograph above.

BRICE'S INTERESTS AND ACCOMPLISHMENTS
Since coming to the University of Michigan forty years
ago, Brice has been at the forefront of computers and com-
puting, particularly in chemical engineering. His initial and

Chemical Engineering Eduction

extensive collaboration with Don
Katz firmly established his inter-
est in computing, numerical
methods, and process design and
simulation.
As an outgrowth of the Ford
Foundation project, Brice pre-
sented a famous (at Michigan,
anyway) and highly popular se-
ries of six two-hour evening lec-
tures on computers and program-
ming, first in "MAD" and later
in FORTRAN; one memorable
lecture was given in a Batman
costume to compensate for a time
conflict with the premier hour of
the Batman television series.
These evolving lectures were at-
tended each term by about 300 Since coz
students, faculty, staff, and lay University of
persons who needed a quick, non- years ago, Brici
credit introduction to computers forefront of c
and programming. The series be- computing,
gan in 1960 and lasted over a ch e
quarter century, well into the PC chemical en
era. In the chemical engineering in th
department, he mainly teaches 1970s and
numerical methods and com- Brice and a c
puter-aided process design, with assistants devel
an occasional foray into sopho- sponsorship, soi
more-level material and energy p
balances. computer-base
chemical
Brice's research interests and
those of his doctoral students
have focused on algorithm design and software develop-
ment for computer-aided process modeling, particularly for
dynamic process simulation. He is currently working on
decomposition, numerical, and coordination algorithms suit-
able for solution of large-scale dynamic process models in
distributed-memory parallel computing environments, and,
with Professor Costas Kravaris, on the potential of the ap-
proach for distributed model-based control.
In the late 1970s and early 1980s, Brice and a cadre of
student assistants developed, under NSF sponsorship, some
of the earliest computer-based courseware for chemical en-
gineers. His MicroCACHE software, consisting of execu-
tive routines for module authoring and presentation, and
several instructional models for numerical methods and
flowsheeting, was originally developed for the Apple II
personal computer and later converted for use on the IBM
PC. The MicroCACHE work was followed in the mid-to-
late 1980s by development of the more powerful

MicroMENTOR system software and
courseware, which is currently be-
ing used at Michigan as the principal
delivery vehicle for networked ac-
cess, control, delivery, and statistics-
gathering for all IBM PC-based soft-
ware used by students in the chemi-
cal engineering department at Michi-
gan (including the newest Michigan
instructional modules developed un-
der the direction of Michigan Pro-
fessors Fogler and Montgomery and
distributed by CACHE).
Brice is currently chairman of the
department's graduate committee,
a position in which he has served
for sixteen years. In this connec-
tion, he hosts a very popular party
at his house every March for re-
cruiting prospective graduate stu-
dents who are visiting the depart-
ment. In the Engineering College,
he was elected by college faculty
to the College Executive Commit-
tee for a four-year term (1979-
1983) and served from 1983 to
1993 as a member of the Executive
Committee of CAEN, the large and
versatile Computer-Aided Engi-
neering Network at Michigan.

ursewarejor On the national scene, Brice was a
ineers. founding member and the first in-
terim chairman of the CACHE (Com-
puter Aids for Chemical Engineer-
ing Education) Corporation-in fact, the organizational meet-
ing for CACHE, called by Brice and Warren Seider of the
University of Pennsylvania, was held in Ann Arbor in 1969.
In CACHE, he subsequently served as vice-chairman and
chairman (1974-1975) and is currently very active as a
board member and as CACHE publications chairman, posi-
tions he has held since 1970. As publications chair, he has
overseen production of nearly all of CACHE's major docu-
ments, including preparation and distribution of the Pro-
ceedings of ten International AIChE/CACHE conferences
in the last decade. He has held elected AIChE positions
leading to the chairmanship of the CAST (Computer and
Systems Technology) Division in 1981-1982, and has been
a member of the editorial board of Computers and Chemi-
cal Engineering since 1978.
Brice has received numerous citations for his dynamic
style of teaching and service, including the Engineering
Class of 1938 Distinguished Service Award (1963), the

Summer 1996

ning to the
Michigan forty
e has been at the
computers and
particularly in
gineering....
e late
early 1980s,
adre of student
oped, under NSF
me of the earliest
* -j,

d CO
eng

Jim was a pioneer in the numerical solution of partial differential equations, both by finite-difference
and finite-element methods, and his research interests have always been in that area. He has chaired
or cochaired the committees of twenty-one doctoral students ... he has [also] always been interested in
church organs and has served on numerous committees for doctoral organ students at the University of
Michigan. He often finds "historical performance
correctness" boring, preferring organ recitals that
incorporate a few tuneful selections and are
imaginative in their use of varied tone colors.

University of Michigan Outstanding Achievement Award
(1968), and awards from the University of Michigan Engi-
neering College for Excellence in Teaching (1983) and
Excellence in Service (1993). At the national level, his
leadership in computing for chemical engineers has been
recognized by the AIChE Computing in Chemical Engi-
neering Award (1980), the Detroit Engineering Society
Chemical Engineer of the Year Award (1987), and the
ASEE Chemical Engineering Lectureship Award (1991).
For the last of these, he presented a lecture at the Toronto
ASEE meeting in 1990, with two fascinating themes-an
outline of the development of computers and computing
over the previous fifty years, and not only educational uses
of computers over the same time period but also predic-
tions of future trends and developments (several of which
have already transpired!). This ASEE lecture was pub-
lished in the Fall 1991 and Winter 1992 issues of Chemical
Engineering Education.
Brice is an avid reader, especially of nonfiction, and has
a keen interest in world affairs, politics, education, and
travel. He is especially interested in the far East, and in the
past two years he and Jim have each taught two month-
long intensive graduate courses at the new College of
Petroleum and Petrochemical Technology at the
Chulalongkorn University in Bangkok.

JIM'S INTERESTS AND ACCOMPLISHMENTS
Jim was a pioneer in the numerical solution of partial
differential equations, both by finite-difference and finite-
element methods, and his research interests have always
been in that area. He has chaired or cochaired the commit-
tees of twenty-one doctoral students, the great majority of
whom have also engaged in experimental work in tandem
with their numerical studies. Topics studied have ranged
from two-phase flow, measurement of turbulent velocity
fluctuations, natural convection, reservoir engineering,
metal casting, and many aspects of polymer processing.
His two current doctoral students are working on paint-
leveling and injection-molding problems sponsored by the
General Motors Corporation.
Jim is most at home in the classroom, where he teaches
fluid mechanics and numerical methods. Occasionally,

Jim at the console of the 1891 "Father" Willis organ
in Blenheim Palace.

and only on April 1st, he demonstrates how dimensional
analysis can be used to estimate the speed of a dinosaur
by measuring its fossilized footprints. He has also re-
cently developed (with colleague Pablo LaValle) a fine
first undergraduate laboratory, with many experimental
projects that go beyond the traditional fare. He has been
recognized many times for his dedicated classroom teach-
ing, being a first recipient in 1980 of the College of
Engineering's newly instituted Engineering Excellence
in Teaching Award. In 1987 he received the highest Uni-
versity of Michigan award for classroom teaching-the
Amoco Good Teaching Award-and was named an Arthur F.
Thumau Professor from 1989-1992, an appointment that is
based largely on undergraduate teaching evaluations.
Jim was department chairman at Michigan from 1971

Chemical Engineering Eduction

to 1977 and Assistant Dean for Admissions in the Col-
lege of Engineering from 1990 to 1994. In the Engineer-
ing College, he was elected to the Executive Committee
for the period from 1985 to 1989. On the national and
international scene, he has been coeditor since 1989 of
the "Class and Home Problems" section of Chemical
Engineering Education, and since 1973 he has been As-
sociate Editor for the U.S.A. of Chemical Engineering
Research & Design (the Brit-
ish equivalent of the AIChE
Journal). Last year he was
elected (in a contested elec-
tion!) as Water Commissioner
of the village where he lives.
Jim has extensive interests
outside the university. Since
visiting Clungunford Church in
Shropshire in 1943 with his
neighbor, Graham Jukes, he
has always been interested in
church organs and has served
on numerous committees for
doctoral organ students at the
University of Michigan. He of-
ten finds "historical perfor-
mance correctness" boring,
preferring organ recitals that
incorporate a few tuneful se-
lections and are imaginative in
their use of varied tone colors.
One of his "heroes" was the Jim and his w
late Virgil Fox, an American perei
organist par excellence, who
could inspire vast audiences of people who were other-
wise little interested in classical organ performance. Jim
has an Allen digital-computer organ in his home.
Another source of inspiration was Professor Terence
Fox, who founded the chemical engineering department
at Cambridge in 1946. Fox was a shy but brilliant man
who knew what was important and who brought the de-
partment to preeminence before his untimely death in
1964. He was instrumental in bringing Kenneth Denbigh,
John Davidson, Peter Danckwerts and others into the
department. Danckwerts subsequently wrote an appre-
ciation of Fox's talents, saying, accurately, "Fox did no
research and published nothing." How times have
changed-today, Terence Fox's resume would be tossed
aside and he would stand no chance of being hired, let
alone of receiving tenure!
As an amateur organist, Jim received his performance
diploma, Associate of the Trinity College of Music (Lon-
don), in 1951, and his Service-Playing Certificate from

ife A
nia

the American Guild of Organists in 1981. He is a mem-
ber both of the American Guild of Organists and of the
Winchester & District Association of Organists in En-
gland. About once a year, he gives "popular-science"
lecture/demonstrations on how organ pipes work (Kelvin/
Helmholtz instabilities have to be simplified for lay audi-
ences!), the most recent being an invited presentation to
the 1995 National Convention of the Organ Historical
Society. He gives occa-
sional recitals, the most
recent being to an en-
thusiastic audience in a
packed church-back in
Clungunford in 1995, on
the occasion of the 100th
anniversary of the instal-
lation of their organ.

In 1995, Jim wrote and
published a profusely il-
lustrated 160-page book,
Pipe Organs of Ann Arbor,
which describes about sev-
a henty-five instruments in
the city's churches, col-
leges and universities, resi-
dences, and cinema-and
S even in a funeral parlor.
He is also working on
Sp two other books: Fluid Me-
lary Ann in their chanics for Chemical En-
lgarden gineers and Place-Names
of Hampshire and the Isle
of Wight. The latter was written in a beautifully illus-
trated manuscript of about 1,000 pages by his grandfa-
ther, Alfred Oscroft, in the two decades before his death
in 1939. It traces the origins of the names of all the
villages, hamlets, towns, etc., in Hampshire, many of
which have Anglo-Saxon roots. The cross-checking of
all the references, many of which are at least 100 years
old, will take much time, not to mention learning the
rudiments of the Anglo-Saxon language. Related to this
endeavor, he is a member of the English Place-Name
Society.
In addition to music and writing, Jim's hobbies include
hiking in North Wales and in the American West (he has
visited Zion National Park eight times and always enjoys
walking up to the West Rim), tennis and table tennis,
gardening, and reading. Most recently, he has read Mar-
tin Gilbert's Churchill, David McCullough's Truman,
Doris Kearns Goodwin's biography of the Roosevelts
during World War II, No Ordinary Time, and is just
beginning George Blake's No Other Choice. 0

Summer 1996

Roessay

EVOLUTION

FOR CHEMICAL ENGINEERS

E. N. LIGHTFOOT
University of Wisconsin Madison, WI 53706-1691

his essay is written to suggest that a type of thinking
described below under the term evolutionary dynam-
ics is a key component of chemical engineering that
should be given formal recognition in a variety of our pro-
fessional activities. These include education of our students,
recruiting of faculty, and even the direction of research.
Moreover, there is available a large and rapidly growing
reservoir of literature upon which we can draw for incorpo-
rating evolutionary concepts into our profession, and it is
important to note that some academic researchers have al-
ready begun to implement these ideas.""' We may in fact be
lagging behind some sister disciplines in this regard, and the
utility of evolutionary dynamics may be particularly impor-
tant for industry and government."5'
The basic premise behind the above suggestion is that the
primary activities of chemical engineers are either to invent
new concepts, processes, and equipment, or to improve ex-
isting ones. Since true de novo developments are rare, both
types of activities may be viewed as evolutionary, and the
term evolutionary dynamics seems appropriate to describe
the rates at which they proceed. So defined evolution may be
seen as related to but distinct from design, and in many
ways deserving of a higher conceptual priority; evolu-
tionary considerations provide the primary impetus for
design efforts even as the needs of the designer provide
the primary justification for engineering science and other
descriptive disciplines.
The recognition of evolutionary dynamics is both timely
and important for at least two reasons. The first is that we
live in an era of rapid and unpredictable changes, most of
which are beyond our control, and the ability for both indi-
viduals and social groups to evolve rapidly in some effective
sense is therefore of critical importance. The second is that
the dynamics of evolution are surprisingly complex in detail,
and it is only recently that tools and concepts needed for
their effective understanding have become available. Se-

elected examples of these tools and concepts are introduced
immediately below, and applications specific to chemical
engineering education are introduced in the last section.

4 BACKGROUND >

- Biological Evolution
Often lost in a fog of bewildering chemical and physi-
ological detail is the central fact that modem biotechnology
is built squarely and consciously on information theory and
that the great complexity of the biological world is in turn
the result of evolutionary dynamics, most probably driven
by a simple objective function: preservation of information
represented by chains of simple organic compounds, the
nucleotides generally known as DNA. In fact, elaboration of
genetic information theory predated the discovery of its
chemical basis, and a successor development, evolutionary
theory, is now ahead of experiment in its turn.
Moreover, as biologists are forced increasingly to deal
with enormous complexity, there is growing pressure to
develop sophisticated hierarchical models that will increas-
ingly make the systems analysis used by engineers look
rather primitive. Individual organisms, even microorgan-
isms and mammalian cells, are already more complex than
large chemical plants in terms of mass flows and control
strategy. One can already see sketched out a spectrum of

Ed Lightfoot was born and raised in suburban
Milwaukee and obtained both his BChE and PhD
degrees from Cornell University. After three years
of process development at Chas. Pfizer, he joined
the University of Wisconsin chemical engineering
department, and except for leaves he has re-
mained there since that time. He is still teaching,
though he formally retired in October of 1995. His
interests have centered around mass transport
with an emphasis of biological applications.

Copyright ChE Division ofASEE 1996

Chemical Engineering Eduction

complexity from relatively short nucleotide chains or genes
and the proteins produced by them to gene equivalents, such
as the "memes" of Dawkins,"' and on to large social groups
and organized bodies of knowledge.
These aspects of biology are steadily becoming more quan-
titative and systematic, and they are much more easily un-
derstood by chemical engineers than such classic sciences as
biochemistry and molecular biology where the non-expert
quickly becomes drowned in masses of detail and special-
ized notation. Moreover, I believe that they are also far
more important for most of us.

- Basic Questions

At first sight, the very existence of evolution is
counterintuitive. How can successively more complex life
forms arise in a dissipative universe, and is such a tendency
to increasing order inevitable? These basic questions have
been addressed by a great many eminent scientists, of which
the best known is perhaps Jacques Monod.1" But for many
engineers the clearest and most satisfactory answers are
provided by Manfred Eigen"' and his co-workers, on the
basis of information theory combined with Darwinian selec-
tion. Eigen shows that biological evolution depends upon
errors in replication of DNA and that there is an optimum
error rate. No evolutionary change can occur in the absence
of error, but too high an error rate can overwhelm the pro-
cess of natural selection and lead to degeneration.
For such simple structures as small viruses, error rates are
small enough to permit development of well-adapted organ-
isms, but the scales are tipped toward degradation as the
number of nucleotides in the organism DNA increases. Eigen
and Schuster"9 have been remarkably successful in estimat-
ing the maximum gene size permitting effective simple natu-
ral selection, and they have proposed a more complex mecha-
nism, "hypercycles," for organisms with larger genes.
The energy source for evolution is environmental degra-
dation of free energy, and it is found that the entropy genera-
tion needed to produce even so complex a structure as a
large mammal is not excessive.
Almost as puzzling as existence is the remarkable speed of
evolution, shown for example in our current difficulties with
the AIDS virus and the development of bacterial resistance
to antibiotics. Contrary to general perception, evolution is
fast-and ubiquitous. Eigen shows, again for very simple
organisms, that this speed results partly from heterogeneity
within apparently homogeneous species. He points out that
there is always a multidimensional distribution of genetic
content about the dominant or "wild" form, and that environ-
mental changes result in a rapid redistribution of frequency.
Such adaptation is particularly rapid for sexually reproduc-
ing organisms where combinations totally unsuited to a pre-
existing set of conditions are continually arising through

very large numbers of random binary combinations of pa-
rental genes. This is a particularly important point for non-
biological evolution, as we shall see below. In one sense
important to us as parts of vulnerable ecosystems, nature is
very wasteful: individuals and whole species are continually
sacrificed in the development of better adapted forms.

- Non-Biological Models

No well-substantiated models for natural selection in com-
plex organisms yet exist, and direct experimentation is at
best difficult. But analysis of non-biological model systems
has provided some provocative and stimulating insights.
Among these are the suggestions of Kauffman .1" that Dar-
winian theory must be extended. He suggests a three-tiered
approach:
Recognize and delineate the spontaneous sources of order,
the self-organizing properties of complex systems. as an
essential complement to the disorder postulated by Darwin
as the sole source of evolution.
Understand how such self-ordering properties permit,
enable, and limit the efficacy of natural selection.
Understand which properties of complex systems confer on
them their ability to adapt and evolve.
Kauffman's texts are characterized by the posing of a
great many seminal questions and by attractive but as yet
unproved possible answers. Among the most important is
his suggestion that living organisms, or their genes, are
parallel distributed regulatory networks operating on the
edge of chaos. His first text""' is the more complete, but
the second"" is by far more accessible for newcomers to
this field.
Prominent in Kauffman's developments is the concept of
fitness landscapes, which describes the evolutionary fitness
of organisms as functions of determining factors such as
amino acid content of enzymes. These in turn are used to
describe the counterbalancing of evolutionary driving forces
with the degrading effects of DNA replication errors and can
in principle be used to determine both the limits of achiev-
able fitness and the most attractive search routines across the
fitness landscapes. They can also be used, again in principle,
to describe co-evolution in ecosystems, a major problem in
evolutionary dynamics. Moreover, his ideas are readily ap-
plicable to non-biological systems.
The work of Holland"'- and others and the concept of
self-ordering supplement and extend Kauffman's arguments,
and a variety of auxiliary ideas appear to be important.
Chaos theory and nonlinear dynamics are obviously among
them, but the current arguments over the relative merits of
holistic versus reductionist thinking (see for example Refer-
ence 14) may contribute significantly as well.
Already these non-biological models provide highly use-
ful insight and show for example that evolution does not

Summer 1996

always produce very high degrees of "fitness." Fitness is
itself a difficult term to pin down, as are "adaptability" and
the even more vaguely defined "evolvability."

> Empirical Approaches and Hierarchical Modeling

At the moment, the suggestions of Kauffman and others
must be viewed as interesting but unproved hypotheses, and
we must usually settle for empiricisms based on study of a
variety of systems, from small biological structures through
whole organisms to ecosystems of varying complexity. More-
over, as the complexity of the system under study increases,
both the precision and reliability of available models de-
creases. The more complex situations are often the most
important, however, in chemical engineering as well as in
biology, and here the biologists may be ahead of us. As a
group, they have learned to work at a great many different
hierarchical levels, even as individual researchers tend to be
highly specialized. Global syntheses are still rare and highly
incomplete, but a great variety of useful disciplines (e.g.,
various aspects of ecology and sociobiology) has emerged.
Fortunately, many useful generalizations are available,
and those dealing with very small ecosystems are of particu-
lar interest to academics; most of us operate within small
and relatively isolated groups. Examples include aca-
demia itself relative to the larger world of chemical engi-
neering, groups of researchers in highly specialized fields,
and academic departments.
It is thus important to note that diversity within any given
ecosystem is a stabilizing factor that also increases ecosys-
tem productivity-and that small systems such as isolated
islands tend to be very poor in numbers of species; they
simply cannot hold a highly diverse system. Moreover, natu-
ral selection within a small system tends to produce highly
specialized species that cannot survive contact with a larger
and more competitive world. The flightless birds of New
Zealand and other island systems have fared poorly on con-
tact with rats and other invading organisms, but supreme
opportunists such as coyotes have thrived in fast-changing
circumstances. Moreover, the highly specialized species of
isolated systems may cease to evolve at an appreciable rate
in their protected and stable environments once the acces-
sible "niches" have been filled.
Another very important aspect is that of co-evolution. This
field is of considerable potential importance to engineers; all
of our work is done within the context of dynamic interac-
tive environments.

I Useful Similarities

All of the above discussion would be of relatively little
utility to chemical engineers were it not for the fact that non-
biological evolutionary processes, from the development of
social systems and industries to the refinement of such "spe-

cies" as chromatographic columns or oil refineries, share
many of the key features of biological evolution. This point
of view was discussed in philosophical terms by Dawkins16'
is repeatedly expressed by Kauffman, and is analyzed with
great enthusiasm and exhaustive detail by Dennet."5' This
last text is not as scholarly as that of Kauffman, but it is more
down to earth and accessible. In many ways it is the starting
point for the remaining discussion here. But there are now
very large numbers of books and shorter analyses dealing
with generalizing evolution theory in a wide variety of envi-
ronments (e.g., References 12 and 13) and even to the phi-
losophy of evolution."6'

4 APPLICATIONS AND CASE STUDIES >
The first priority is to recognize evolutionary dynamics as
a key aspect of engineering and then to review our activities
in the light of this new concept. The primary goal of such a
review should be improving our synthetic, as opposed to
analytic, abilities.
At a more detailed level we should take a new look at
departmental structures and hiring policies. Here, review of
current efforts of this type in other fields should prove help-
ful. A representative example is the application of Darwin-
ian models for corporate change.1"'
Introduction of evolutionary ideas into our curricula is
important, but it must follow faculty development. The tried
and true method of exploring new ideas at the research level
is the classic means of such development, and it must be
given major emphasis.

N Research

Much is already being done in biology, and the Pro-
ceedings of the National Academy of Science has a sec-
tion devoted to evolution in nearly every issue. Evolu-
tionary dynamics has proven an important aspect of the
AIDS problem."71
More recently, engineers have been using either biological
evolution or mimicking it is useful ways. John Yin has been
studying phage evolution for some years and is now seeking
such mundane but important applications as remediation of
metal contaminated soil.r21 Alex Zehnder has found that
evolutionary processes in wild environments can produce
hardy organisms capable of detoxifying previously resistant
substances."4' Here, success is achieved by transfer of en-
zyme producing genes between unrelated bacteria to provide
new and complex detoxification complexes. This evolution-
ary approach has a major advantage over conventional ge-
netic engineering in producing organisms capable of surviv-
ing in sewage streams. loannis Androulakis'" has developed
what are called genetic algorithms to speed process design.
Combinatorial chemistry"'4 is a natural subject for such

Chemical Engineering Eduction

analysis, and the evolutionary improvement of enzymes"8
may prove of general engineering interest.
At a more philosophical level, evolutionary researchers
such as Kauffman may be close to answering basic philo-
sophical questions as to why research and development are
even feasible-and perhaps help solve the vexing problems
as the economic establishment of research directions. We
should join with them.

Faculty Hiring and Departmental Organization

It appears clear that hiring, career development, and inter-
actions with outside influences all need a harder look.
Recruitment of new members is of primary and immediate
importance. Faculty hiring policy has great long-range im-
pact, is very hard to rectify once hiring decisions have been
made, and is now made rather casually. We seem to be quite
faddish as a profession, both as to specific technical fields
and to the approach candidates take to them. Moreover, it is
abundantly clear that we cannot hire enough individuals into
any department to adequately cover all important aspects of
chemical engineering.
Each of our departments is a tiny ecosystem, isolated to a
significant degree and trying to survive and prosper in a
tough world. Most of us are opting for narrow experts in
"hot" fields who can bring in substantial sums of research
money in competition with literally hundreds of like-minded
competitors. Few are thinking very far ahead or very deeply
about long-range problems. Finally, a large-scale wastage of
individuals and whole ecosystems, characteristic of biologi-
cal evolution, is highly undesirable for social "organisms"
even though it is presently quite common in the United
States. A major goal of social evolution should be to miti-
gate the iron laws of biological evolution.
I would suggest that highly specialized individuals with
narrow interests are unlikely to be good bets for making the
changes that will prove necessary for survival, and that a
"fine-grained" personnel structure characterized by such spe-
cialists can make cross-disciplinary interactions in a small
group inadequate for development of a strong department. It
will also result in inadequate coverage of our wide-ranging
profession. This is already being recognized at leading busi-
ness schools interested in restructuring industrial concerns,
and ongoing work in the area may be pertinent to our discus-
sion." In fact, engineering science may not be a good pri-
mary focus today, and certainly not for all departments. It
appears more likely that we need a mix: experts in important
core areas to deal with the increasing complexity of modem
science and technology; careful organizers to maximize effi-
ciency of our operations; and carefully selected generalists
to supply the "glue" and inspiration for change.
Generalists with wide-ranging interests and good educa-
tional backgrounds in the engineering sciences may be an

especially good bet right now. They can provide bridges
between specialists, extramural as well as intradepartmental,
and between academics and industrial engineers. They can
also provide the "noise" that may be needed to keep evo-
lutionary trends vigorous. More important, they tend to
be the optimistic opportunists who typically respond most
quickly and effectively to new circumstances. Time and
again specialists have proven excessively conservative
and resistant to change.
We must also rethink departmental structures and priori-
ties. The present intense concentration on immediate sur-
vival will produce few deep or long-range thinkers, and it
will reduce the possibilities for informal "multi-brain" inter-
actions that could be so valuable for rapid evolution of ideas
and concepts. Such interactions are the equivalent of multi-
sexual reproduction and can lead to extremely rapid genera-
tion of new ideas. Excessive survival stresses also severely
limit the kind of unstructured reflection known to stimulate
creativity. Our present modus operandi is unlikely to pro-
duce the major evolutionary changes needed to meet long-
term environmental stresses effectively.
The development of close external contacts must again
receive the high priority of past years. Modern means of
communication can certainly be used more extensively,
but there seems to be no adequate substitute for face-to-
face contacts.
Current pressures for submitting faculty to highly struc-
tured schedules is a formula for evolutionary disaster. The
chief administrative goals of our university are to increase
faculty productivity in narrowly focused ways: increased
contact hours of formal instruction, more service to soci-
ety, and more research funding. These are highly unreal-
istic unless accompanied by as yet unidentified ways to
increase efficiency.
Immediate priority must, however, go to increasing the
efficiency of funding and of conducting our fundamental
activities; money is clearly one analog of the free energy that
drives evolution, and all successful organisms are highly
efficient energy transducers. Success in these activities may
in fact help to achieve the above administrative goals, but we
must go one step at a time.
These last are not newly discovered problems, and they
need no special elaboration here. But they do need continued
restatement, and they are an important part of evolutionary
dynamics. Departments of chemical engineering will un-
doubtedly survive in the face of present administratively
imposed pressures, but they may end up like the lycopodium
and horse tails of Wisconsin forests: insignificant remain-
ders from a glorious carboniferous past.

1 Curricula and Training of Engineers

Curriculum modification is clearly near the top of the

Summer 1996

priority list, and it is important to begin with what we
have. Increased emphasis on process invention in our
introductory courses is promising, and it appears likely
that much of evolution dynamics will be found to parallel
design of engineering systems. A careful comparison of
biological evolution with design strategy may well prove
beneficial to both fields.
It does seem time to give a trial course on evolution,
probably as an elective at the graduate level, and this should
begin with the relatively advanced area of biological evolu-
tion. If possible, the first should be a highly interactive
course, preferably given jointly with biologists. Much re-
mains to be done before a realistic organization is achieved,
but it is possible to sketch out a rough outline:

Introduction to Evolutionary Dynamics
for Chemical Engineers

A. Biological Evolution

1. Basic definitions"91
Information theory and evolution"8'
Mechanistic bases of evolution dynamics
Origins of variability
Driving force and objective function
Selection
Quasi-species
Organizational levels
Complexity [8s.. ..
Fitness and fitness landscapes 10,11,19]
2. Evolution and adaptation in simple organisms: theory
and experiment
Simple replicators; small viruses
More complex replicators; hypercycles
Bacterial adaptation
3. Evolution of more complex systems
Overview of the origins of species
Comparison of the Cambrian and Permian
evolutionary explosions
Stasis and radiation
Evolution of ecosystems and effects of isolation

B. Evolution in Engineering

1. Introductory remarks
Definitions and scope of discussion
Foundations: are there coherent theories for non-
biological evolution?

Bases of non-biological evolution
Parallels to mutation
Driving forces for change
Selection
Organization: types and levels
2. Historical perspective
Major evolutionary spurts (tentative listing)
The western world
antiquity
renaissance
industrial revolution
China, Japan, others
The modem world
Chemical Engineering: selected examples
3. Search for a new synthesis: interaction of science,
technology, politics, and business

REFERENCES
1. Androulakis, I.P., and V. Venkatasubramanian, Computers
Chem. Engen., 15(4), 217 (1991)
2. Yin, John, "Metal Recovery by In Vitro Selection," Biotech.
Bioeng., 45(5), 458 (1995)
3. Yin, John, J. Inorg. Biochem.,accepted for publication in
1996
4. Zehnder, A., "Molecular Mechanism of Bacterial Adaptation
to Degradation of Chlorinated Organic Compounds," sym-
posium Louis Pasteur et l'Industrie aux XXI siecle, l'Institut
Pasteur, Marnes-la-Coquette-Paris, 25-28 Sept. (1995)
5. Gouillart, F.J., and J.N. Kelly, Transforming the Organiza-
tion: Reframing Corporate Direction, Restructuring the Com-
pany, Revitalizing the Enterprise, Renewing People, McGraw-
Hill, New York, NY (1995)
6. Dawkins, Richard, see for example The Selfish Gene, 2nd
ed., Oxford (1989)
7. Monod, Jacques, Hazard et la Necessitd, Editions du Seuil
Paris (1970); Chance and Necessity, Knopf (1971); Vintage
paperback (1972)
8. Eigen, Manfred, Stufen zum Leben, Piper, Miinchen (1987);
English edition, Steps Toward Life, Oxford (1992)
9. Eigen, M., and P. Schuster, The Hypercycle A Principle of
Natural Self-Organization, Springer (1979)
10. Kauffman, Stuart, The Origins of Order: Self-Organization
and Selection in Evolution, Oxford (1993)
11. Kauffman, Stuart, At Home in the Universe: The Search for
the Laws of Self-Organization and Complexity, Oxford (1995)
12. Holland, John, Adaptation in Natural and Artificial Sys-
tems, U. Michigan Press (1975)
13. Holland, John, Hidden Order, Addison-Wesley (1995)
14. Combinatorial Chemistry, a review in C&E News, pg. 28 (12
Feb. 1996)
15. Dennet, D.C., Darwin's Dangerous Idea, Simon and Schuster
(1995)
16. Brandon, R.N., Concepts and Methods in Evolutionary Biol-
ogy, Cambridge (1996)
17. Nowak, M.A., et al., "Antigenic Oscillations and Shifting
Immunodominance in HIV-1 Infections," Nature, 375, 606
(15 June 1955)
18. Davis, M.M., "Evolving Catalysts in Real Time," Science,
271, 1078(1996)
19. Keller, Evelyn Fox, and Elisabeth A. Lloyd, Keywords in
Evolutionary Biology, Harvard (1992) 0
Chemical Engineering Eduction

r1 M book review

BIOREA ACTION ENGINEERING PRINCIPLES
by Jens Nielsen and John Villadsen
Published by Plenum Press, 233 Spring Street, New York,
NY 10013-1578; $79.50

Reviewed by
James C. Liao
Texas A&M University

To non-practitioners, biochemical reactions appear to be
nebulous, formidably complex, and even a bit magical. To
students and practitioners of biochemical engineering, bio-
chemical reactions remain too unpredictable to warrant quan-
titative and theoretical analysis. However, no one denies that
bioreaction systems must obey the fundamental laws of chem-
istry and physics, and that given sufficient information,
bioreaction systems can be mathematically modeled. The
question is whether we know enough now to model the
bioreaction systems, and given the information available
today, how can mathematical models help us. The authors of
Bioreaction Engineering Principles have taken a positive ap-
proach to highlight the contribution of mathematical analysis
and to prepare students for future developments in this area.
Although it is uncommon to teach bioreactions from theo-
retical and mathematical viewpoints (an approach that is
commonly adopted in chemical reaction engineering), there
is no reason why bioreactions cannot be subjected to math-
ematical rigor. With such a philosophy in mind, the authors
have provided a mathematical treatment for every aspect of
bioreaction systems. The result is a clear and logical intro-
duction to bioreaction systems with useful examples and
stimulating problems. This book is one of the few texts, if
not the only one, attempting to carry the instructional ap-
proach and philosophy of chemical reaction engineering to
bioreaction systems. Although the book is mathematically
oriented, the authors showed "a deep respect for the wonder-
ful complexity of microbial reactions," making the volume
highly relevant to modern microbial biotechnology.
For chemical engineers, the book is an excellent introduc-
tion to the subject of microbial reaction systems. All the
intracellular reactions are introduced with mass and energy
balances in mind, making chemical engineers feel quite at
home. For students without a mathematical background, how-
ever, the book is a little intimidating: matrices, vectors,
integrals, and lots of Greek letters. The teacher will have lots
of coaching to do. Given the plethora of biochemistry and
microbiology textbooks that aim toward students without a
mathematical background, this book provides a unique and
useful view at the other end of the spectrum.
After an introductory chapter, the book begins with vari-

ous mechanisms of nutrient transport and major metabolic
pathways. Instead of the typical metabolic maps and mo-
lecular mechanisms seen in biochemistry texts, it empha-
sizes stoichiometry, overall reactions, and energy and mass
balances. The authors introduce mathematical representa-
tion of flux and elemental balances, often under-appreciated
in the area of biotechnology. The analysis is rigorous and
involves very few assumptions. The equations provide a
basis for further analysis of reaction rates. This chapter also
discusses the energetic of anaerobic and aerobic processes,
which are important considerations in bioreactor systems. With
a little touch of thermodynamics, this chapter provides a start-
ing point for biochemical engineers to take a serious look at
energy balance and the energetic aspects of biosystems.
Chapter three deals with metabolic flux analysis, meta-
bolic control analysis, and identification of measurement
errors, topics of significant scientific and practical interests.
The discussion gives a clear introduction to the methodol-
ogy. Mathematically inclined students will find the discus-
sion concise and precise-others may need more time to
digest the equations. The examples here are the best tutors.
The authors took the time to digest all current literature in these
areas and present a cohesive view of the methodology with
some nice ideas in examples and problems. Chapters 2 and 3
are perhaps the most unique features of the book compared to
other similar titles in biochemical engineering.
With a strong basis of intracellular reaction analysis, the
book then goes into modeling of cell growth and morphol-
ogy. A general mathematical formulation is first presented
as a framework for discussion. Kinetics of cell growth, struc-
tured and unstructured, and population balances based on
cell number are then discussed with sufficient details. The
general formulation may seem meaningless for beginners,
but with some understanding of the system, it offers an
overall picture of the problem under investigation. Again,
the authors designed excellent examples and problems for
illustration and practice.
The last part of the book is the application of hard-core
chemical engineering to bioreactors: mass transfer, interfa-
cial and bubble behavior, batch reactors, continuous stirred
tank reactors, plug-flow reactors, mixing, and scale-up. For
chemical engineering students, these chapters offer good ex-
amples to learn mass transfer and reactor design in an uncon-
ventional area-biotechnology. For biotechnologists, follow-
ing the equations may be difficult in the beginning. With the
help of examples, the task becomes much easier. Furthermore,
simply going through the discussion will gain a useful picture
of engineering approaches to biotechnology problems.
In summary, this is an excellent book dedicated to bioreaction
engineering. With increased understanding of cellular and in-
tracellular functions, it is a timely addition to the textbooks
available in biochemical engineering. The book set the founda-
tion for systematic and rigorous modeling in this area. 0

Summer 1996

e M learning in industry

This column provides examples of cases in which students have gained knowledge, insight, and
experience in the practice of chemical engineering while in an industrial setting. Summer interns and
coop assignments typify such experiences; however, reports of more unusual cases are also welcome.
Description of analytical tools used and the skills developed during the project should be emphasized.
These examples should stimulate innovative approaches to bring real world tools and experiences
back to campus for integration into the curriculum. Please submit manuscripts to Professor W. J.
Koros, Chemical Engineering Department, University of Texas, Austin, Texas 78712.

INDUSTRY,

ACADEME, AND GOVERNMENT

Building a New Relationship

JAMES A. TRAINHAM, ARNOLD M. EISENBERG
E.I. duPont de Nemours Co., Inc. PO Box 80357 *

very business is under increasing pressure to achieve

outstanding financial results. At the same time, how-
ever, achieving those results is becoming ever more
difficult. The reduction of international trade barriers com-
bined with the appearance of strong, technology-based re-
gional players has resulted in both increased competition
and reduced profit margins. To compete in this new global
marketplace, almost every large company in almost every
industry has found it necessary to right-size or restructure
their organization, or to re-engineer their work practices.
Although chemical industry research and development
(R&D) spending is growing modestly, an increasing portion

James A. Trainham has been the Director,
Engineering Research and Development for the
DuPont Company since 1992. He holds a BS
and PhD in Chemical Engineering from the
University of California, Berkeley, and a MS in
Chemical Engineering from the University of
Wisconsin, Madison.

Arnold M. Eisenberg is Manager of Operations
and Strategic Planning for Engineering Research
and Development for the DuPont Company. Dur-
ing his twenty-six year career, he has held a
variety of assignments in research, manufactur-
ing, process design, computer-aided engineer-
ing, and management. He holds a BS and MS in
Chemical Engineering from Drexel University.
Copyright ChE Division ofASEE 1996

Wilmington, DE 19880-0357

of R&D budgets is being dedicated to short-term technical
support of existing businesses and environmental compli-
ance. Most companies have reduced the amount of their
R&D budgets dedicated to exploratory or long-range re-
search at the same time the U.S. government is slashing both
its defense and nondefense related R&D spending. For many
of us, these were painful but necessary changes directed at
reducing our costs and increasing our global competitive-
ness. Now, we must look to the future to improve the value
our companies deliver to the customers and stockholders.
The chemical industry's traditional approach of doing es-
sentially all of its own R&D must yield to a new paradigm in
which the talents and resources of academe and government
will be leveraged to produce results while containing costs.
Some of what government spends on R&D should be chan-
neled into areas of research that will have a long-term effect
on improving the competitiveness of the chemical industry.
Together, industry, academe, and government must unleash
the pent-up power of our organizations and turn them loose
to create uncommon value in the marketplace-a sustainable
value that will provide an economic foundation for sustain-
able growth into the twenty-first century.
A new partnership between industry, academe, and gov-
ernment could provide a foundation upon which the value-
creation process could be revitalized. In this paper, we
will report on DuPont's recent experiences in establish-
ing a new type of partnership between government, in-
dustry, and academe.
Chemical Engineering Eduction

GROWTH IN THE
US CHEMICAL INDUSTRY

During the last decade, the U.S.
chemical industry has steadily
grown in terms of volume of prod-
uct shipped and exported, but com-
petitive pressures have steadily
eroded prices, resulting in the dol-
lar value of those shipments grow-
ing at a 1% annual rate, as shown
in Figure 1. This is in stark contrast
to the decades following World
War II that were benchmarked by
explosive growth fueled by the de-
velopment and commercialization
of synthetic polymers.

During the last decade, however,
margins have eroded and profitabil-
ity is at the mercy of the gross world
product (the sum of the gross do-
mestic products of the developed
and developing countries). When
the global economy is growing, in-
dustry returns are reasonable; when
it's not, industry often does not earn
the cost of capital. This is not a
formula for long-term success. In
R&D, this has meant that a larger
share of the R&D dollar goes to
customer support and regulatory
expense while less of it supports
development of new product chem-
istry and manufacturing processes.

Compared to defense-related in-
dustries such as aerospace and elec-
tronics, the chemical industry has
received a very small portion of gov-
ernment R&D money even though
it has been a major and consistent

140%

130,.

12" .

0 11_ 0
100"., 0
10 .

90.,

| Volume M Sales (Constat 1984 $s)

Figure 1. U.S. chemical industry shipments and constant
dollar sales indexed to 1984 (Source: 1995 CMA Data Book).

6.5%-

Figure 2. Chemical industry R&D funding as a percent of
sales. (Source: 1995 CMA Data Book).
6.0%

5.5%1

5.0%1

4.5%

Figure 3. NSF funding of areas of interest to the chemical

net exporter. At DuPont, for ex-
ample, current R&D expense is
roughly the same today as it was a decade ago, without any
adjustment for inflation, and all companies have cut employ-
ment to reduce costs, yet have seen little real growth. Under-
lying this trend is the simple fact that while volumes are up
modestly, selling prices continue to erode at roughly half the
rate of inflation. The net result is that there has been little
real growth in total revenue and that growth has barely kept
place with inflation.
The chemical industry is in the process of a major corpo-
rate transformation as it responds to this new environment.
We have worked to meet this global challenge and to be-
Summer 1996

come more cost-competi-
tive. Staying competitive
is good-it is essential-
but it will not create real,
sustainable growth, and
growth is critical both to
industry and to our na-
tional economy. Compa-
nies that create value in
the marketplace prosper
and grow; they create jobs
and opportunities for their
employees; they provide
products and services that
help people live better and
more comfortably; they
make a contribution to so-
ciety. Those companies
that fail to create value
wither and die.
If value creation is fun-
damental to business suc-
cess, then what is value?
We believe that all lasting
value is created by new
technology. If R&D is es-
sential to sustaining the
value-creation process,
how is the chemical indus-
try funding its R&D ac-
tivities? Overall, R&D
funding has increased
from just over 4.6% of
sales in 1984 to 6% in
1994 (see Figure 2).
The National Science
Foundation (NSF) is a key
source of academic R&D
funding. While NSF fund-
ing for materials research

industry. has increased significantly
during the last decade,
funding for basic research in chemistry has, in constant
dollars, increased only marginally, and chemical engineer-
ing funding has actually decreased (see Figure 3). This has
had a major impact on the chemical industry since new
chemistry is the engine that drives growth, and chemical
engineering is the route through which value is captured. In
the maturing chemical industry, new chemistry and engi-
neering technology will become even more important as the
low-cost, high-quality producers dominate the marketplace.
Industry, academe, and the federal laboratories have each
developed a certain character as they worked to fulfill what

has been their traditional role in the R&D community. This
character can be summarized by the strengths and weak-
nesses of these respective entities in carrying out their mis-
sion. Tables 1-3 summarize those strengths and weaknesses
as these organizations function to create value in the market-
place through the development and commercialization of
new technology. Since funding sources have, for the most
part, driven research priorities, industry, academe, and the
federal laboratories have remained separate and distinct enti-
ties, with limited interaction.

INDUSTRY'S ROLE
The chemical industry itself has been the traditional source
of chemical technology of commercial importance. Histori-
cally, the chemical industry has worked on major, propri-
etary developments without direct collaboration with either
government or academe. The collaborations that did exist
were focused on support of enabling technologies.
With significant research budgets dedicated to the devel-
opment of new chemistry and the processes needed to manu-
facture the products resulting from this new chemistry, this
traditional approach to research worked well; but as re-
search expenditures dedicated to new product and pro-
cess development shrank, innovation suffered. The result
has been a dearth of major new products and nearly
stagnant growth rates.
The historical role of the chemical industry in conducting
its own proprietary research has resulted in a matrix of
strengths and weaknesses of these research organizations, as
can be seen in Table 1. The chemical industry has developed
a significant capability to develop and commercialize new,
high-value products given the ideas and the adequate techni-
cal and financial resources to do so. Recognizing that R&D
budgets will remain under continuing pressure, the chemical
industry must return to a balanced R&D portfolio that in-
cludes a focused fundamental R&D effort, one that lever-
ages the capabilities of academe and government to gain
maximum benefit at an affordable cost.

ACADEME'S ROLE
Academe has been the traditional source of fundamental
scientific knowledge. Generally unconstrained by the need
to produce commercial success, it has been able to focus on
developing fundamental scientific knowledge and to work
on issues of academic interest, independent of their commer-
cial value. The result of academe's independence of com-
mercial success was the development of extraordinary capa-
bilities in the growth of fundamental science, summarized in
Table 2. Academe also gained the reputation of being
unresponsive to industry's needs and slow to respond to
specific requests, especially if those requests did not also
include copious funding.
With government funding of research and development

TABLE 1
R&D Strengths and Weaknesses of Industry
Strengths Weaknesses
* Owns the problem High cost
* Knows data needs Resources may not be available
* Has the resources when they are needed
* Knows the materials Cannot afford state-of-the-art
* Knows how to handle equipment in every area
hazardous materials safely Limited focus
* Can move quickly Reduced emphasis on
fundamental research

TABLE 2
R&D Strengths and Weaknesses of Academe

Strengths Weaknesses
* Outstanding fundamental Limited financial resources
research capabilities Sometimes unresponsive
* Lower cost Limited ability to manage
* Innovative and creative hazardous materials
approaches Uncertain continuity
* At or near the leading edge of Potential loss of proprietary
technology information
* Centers of expertise
* Source of future talent

TABLE 3
R&D Strengths & Weaknesses of Federal Laboratories

Strengths
* Highly skilled resource base
* State-of-the-art equipment
* Outstanding continuity
* High degree of specialization
* Outstanding fundamental
research capabilities

Weaknesses
* Uncertain and variable funding
strategies
* Slow to respond to urgent needs
* Proprietary information protection
* High cost

coming under harsh scrutiny, it is likely that money from
these sources will be, in the future, much less than it has
been in the past. To continue supporting the research infra-
structure in academe, new funding sources and structures
will be required. The new paradigm for industrial re-
search funding could have a major effect on academe. To
take advantage of this opportunity, academe has been
and must continue to look for new, innovative ways to
leverage its capabilities into research areas of commer-
cial importance. New alliances with industry are neces-
sary for both to prosper.

FEDERAL LABORATORIES' ROLE
Federal laboratories have been a nontraditional source of
commercial technology, but, recently, one of increasing im-
portance. They have some of the most capable, specialized,
and talented people available in the world in addition to
state-of-the-art facilities, capabilities industry cannot afford
to replicate. The strengths and weaknesses of the federal
laboratory system are summarized in Table 3. Until recently
Chemical Engineering Eduction

there has been little incentive for the federal laboratories
to collaborate with industry in developing products and
processes of commercial importance, but with recent
changes in both law and funding strategies, this situation
is rapidly changing.
Cooperative research agree- Advisory
ments, funds-in agreements (funds -Board
from industry to government), and Information
the Advanced Technology Pro- Flow V
gram are recent examples of gov-
ernment and industry cooperation.
Unfortunately, the government's
push to balance the federal budget Company Company2
has put these programs at risk. Like
their industrial counterparts, some Figure 4. The trc
government leaders are willing to
mortgage tomorrow by cutting fundamental research today.

MEETING THE CHALLENGE:
GOVERNMENT, ACADEME, AND INDUSTRY

To meet this challenge, government, academe, and indus-
try must form a new partnership designed to kick-start growth
and revitalize the industry. The traditional view of the roles
of these three entities shows each pursuing research directed
at their limited view of the world. There are many problems
with this view: there is little collaboration, and much compe-
tition; everyone is competing for the same, shrinking pool of
R&D dollars; the focus is on getting money, not getting
results of commercial importance; there are clear duplica-
tions and voids; and all too often, solutions are looking for
problems instead of problems finding solutions.
Together, government, academe, and industry need to use
their strengths and minimize their weaknesses to develop the
strongest research alliance possible and to deliver results of
both scientific importance and commercial worth. In some
cases, this may require redefining the traditional way they
work together through new alliances and consortia. To use
the unique strengths of industry, academe, and the federal
laboratories, they need to focus on research of commercial
interest, with industry assuming a leading role in the partner-
ship. Proprietary right must be maintained by the sponsoring
company which can realize a competitive advantage by get-
ting the best people with the best equipment working on the
most important problems and producing exceptional results
in a very short time.

REDEFINING THE CONSORTIA
Many universities sponsor special-interest consortia that
provide a focal point for companies with common technol-
ogy interests. The companies benefit by sharing the cost of
developing and leveraging information, while the university
receives a much-needed revenue stream to fund their re-
search efforts. This usually does not give companies access

Summer 1996

Spon
Univ

Corn

iditi

to many of the key academic experts in a particular field.
Individual universities find themselves competing with each
other for the limited funds available instead of collaborating
to leverage their collective expertise in a given field to the
mutual benefit of the compa-
asoring nies they seek to serve. Com-
'ersity J panics can derive competitive
s Flow advantage from these consor-
tia only if they can apply the
knowledge developed in a
unique way since all mem-
bers share equally in the in-
formation developed by the
pany 3 Company 4 Company 5
ny3 Company4 Company university-sponsored consor-
tia. This traditional consortia
onal consortia model. is pictured in Figure 4.

By stating this limitation, we do not imply there is not
great value in these consortia. For enabling technologies,
those needed to run a business efficiently but whose applica-
tion does not provide competitive advantage, these consortia
allow cost and idea sharing. For higher-risk areas of interest,
they permit companies to pool their resources, thus minimiz-
ing the cost of developing leading-edge technology. The
sponsors of these consortia can still gain competitive advan-
tage by applying the results of this research more effectively
than do other members.
These consortia usually have an advisory board composed
of representatives from both the university and the sponsor-
ing companies. Consortia priorities are decided by a voting
majority of this advisory board; thus, a new research pro-
gram requires consensus of the advisory board. One mem-
ber, with a narrow focus leading, perhaps, to a new product
or process, cannot always get the needed work done under
the auspices of the consortia. A member may also be reluc-
tant to discuss concepts with the other consortia members,
fearing that doing so may compromise any competitive ad-
vantage such a development may offer.
A key feature of the traditional consortia is the flow of
money and information. Money flows from many compa-
nies to the sponsor of the consortia (usually a single univer-
sity, although there are some multi-university sponsored
consortia). The sponsor then performs or coordinates the
research, compiles the results, and distributes the informa-
tion back to the sponsoring companies. Although led by an
advisory board, day-to-day operations of the traditional con-
sortia are managed by the sponsoring university.

FORMING A NEW PARTNERSHIP
Recently, several companies have developed a new, re-
verse consortia model (see Figure 5) in which the sponsoring
company, rather than the university, is at the core of the
consortia. In this model, one or more companies sponsor
the consortia and engage those universities and govern-

ment laboratories having the needed expertise. The focus
is, in general, more narrow than in the traditional consor-
tia and is usually directed at, but not limited to, the
development of specific product and process science and
the technology needed.
Unlike the traditional consortia, the reverse consortia is
formed to accomplish a specific purpose, and strategic direc-
tion is defined and controlled by the sponsoring company or
companies. Participating organizations are not selected based
on their willingness to contribute money, but on their spe-
cific expertise in the research area of interest. The composi-
tion of these contributing organizations may change as pro-
gram goals are accomplished. Performance against estab-
lished goals becomes a criterion for continuing participation.
Like the traditional consortia, money flows from the corpo-
rations to the research institutions and information flows to
the paying companies.
Since the sponsoring companies control the consortia, the
developed technology can, and often does, remain propri-
etary. Also, since sponsorship is restricted, potential com-
petitors can be excluded. The net result is that this new
consortia model provides companies with the ability to en-
gage the best research minds to achieve important business
results and still build a competitive advantage. Concurrently,
specialized research equipment resident in academe or at
government laboratories can be leveraged to meet the busi-
ness need. This new model melds together the best of each
organization to form an entity of great strength and vitality
with only a few weaknesses, as can be seen in Table 4.
DuPont has established several of these reverse consortia.
Each is targeted at a specific goal (e.g., improvement of
existing asset productivity, development of engineering pro-
cess control principles from analysis of biocontrol mecha-
nisms, etc.). Potential participants (including professors and
their students) are invited to submit research proposals that
are then upgraded interactively until they are either accepted
or rejected. Although the final decision rests with DuPont,
consortia members collectively upgrade these proposals to
meet the stated goals. DuPont then manages the projects and
works with participating members on project milestones,
timing, and resource requirements.
One of these reverse consortia, shown in Figure 6, is for
the development of an exciting software integration tool
called the Prosight Engineering Workbench. The Prosight
development is a low-risk, high-return effort that requires
many skills not resident in DuPont. We have formed a re-
verse consortia to acquire those skills and accelerate the
product development.
We are developing Prosight in conjunction with Microsoft,
Hyprotech, Interna, the University of Massachusetts, Carnegie
Mellon University, and the University of Edinburgh. We
envision Prosight as a tool our engineers will use to integrate

Figure 5. A reverse consortia model.

TABLE 4
Strengths and Weakness of the New Consortia Model

Strengths
* Sponsor owns both the problem and the
results of the research
* Sponsor understands both the commercial
needs and the materials
* The best research and development minds
can be employed to work on the problem
* Access to leading edge and highly
specialized technology and state-of-the-art
equipment
* The ability to get the right talent assigned
to the program and to change the mix of
assignments as the program progresses
* More rapid completion of the program
* Potentially lower cost than "in-house"
development

Weaknesses
* None of consequence
identified

data and models from many different sources, facilitating
the rapid incorporation of new and sophisticated model-
ing tools developed by academe or industry and making
them almost immediately available to our process engi-
neers and scientists.
This example of the new consortia model is producing
remarkable results. In just eighteen months the Prosight
Engineering Workbench moved from concept to first re-
lease-a remarkable achievement. Without the new consor-
tia model, this development would have surely taken consid-
erably longer and cost significantly more. Based on the
initial success of the Prosight Engineering Workbench, dis-
cussions are underway with other chemical companies, and
we anticipate that this effort will grow to a global, multi-
company consortia in the very near future.
DuPont has not been the only beneficiary of this effort.
Our university partners have adopted part of the product of
this effort as a teaching tool to more effectively connect their
instructional programs to industrial needs. Members of the
university staff have coauthored papers with other consortia
members, and students have had the opportunity to develop
solutions to current, high-value industrial problems. This
mutually beneficial relationship works because industry taps
the talent of academe while, simultaneously, academe con-
nects their efforts to important industrial problems.
Chemical Engineering Eduction

Advisory A

Participating
Organizations

U. Mass CMU U. Edin Hyprotec

Figure 6. The DuPont Process Synthesis and Optimiza-
tion Consortia-Prosight Engineering Workbench
Development

01 ONR
A $'s Flow
Information ..$ Flw
Flow Participating
SOrganizations

Purdue U. LSU U. of Illinois

Figure 7. Neurobiology: Process Control University
Consortium

A second example of the reverse consortia is neurobio-
logical control. This grew out of another industry-academe
relationship. Young prospective faculty members spent a
year in industry before starting their teaching careers. This
gave them an opportunity to develop a better understanding
of industry and industrial research, to building industrial
relationships that can last a career, and to be introduced to
problems, separate from their thesis work, that could start
them on a whole new area of research.
From this activity came the idea for another DuPont-
sponsored consortia-the Neurobiology: Process Control
University Consortium as shown in Figure 7. Unlike the
Prosight Engineering Workbench consortia, this is a high-
risk program that receives significant financial support from
the Office of Naval Research (ONR). Its objective is to
develop and use control systems based on neurobiological
models (e.g., the body's control of blood pressure) for com-
mercial applications. By forming a cooperative consortia
with ONR and academe, DuPont is able to minimize its risk
while taking advantage of the results of this speculative
research effort. If successful, this activity could lead to new
and innovative ways of controlling industrial processes that
could have applicability to problems far removed from the
chemical process industry.
Summer 1996

OTHER FORMS OF COLLABORATION
While the new consortia model provides an unique struc-
ture for extracting value from govemment-academe-indus-
try collaborations, it is not the only approach. For decades,
many companies, DuPont included, have had so-called "Year
in Industry" programs that allowed professors to spend their
sabbaticals working in an industrial research environment-
these proved to be mutually beneficial relationships since
both academe and industry benefited from gaining fresh
insights into the way research could be conducted.
More recently, we have used these programs to provide
specialized talent on focused research programs. As an out-
growth of this activity, we recently invited several graduate
students to do their thesis work with us at the DuPont Ex-
perimental Station. Some of them worked on mutually agreed
upon research and development programs upon which they
based their graduate dissertations. The graduate students
received firsthand industrial research experience while the
company gained the services of young, energetic, talented
people who brought with them novel approaches to our
R&D needs. Ultimately, the students may also benefit by
receiving an offer for full-time employment.
Several students not only made a significant and immedi-
ate contribution to our development needs, but they also
went on to extend their research after returning to the univer-
sity. Several visiting professors have continued their rela-
tionship with DuPont by providing ongoing consultation and
by directing their graduate students into areas of research
that have commercial significance to DuPont.
This effort also permitted visiting professors and graduate
students to interact with both industrial engineers and pro-
fessors and students from other universities. These joint
efforts have resulted in ongoing working relationships that
strengthened their individual research and fostered value for
each other and the benefits of collaborative teamwork. In-
stead of viewing each other as competitors, members of this
new consortia strive to achieve a common goal, competing
only to achieve a higher quality of thought and result.

SUMMARY
The global competitive environment, combined with a
change in funding of research and development in industry,
academe, and government necessitates significant changes
in the way these research organizations work with each
other. The industry-sponsored consortia has been used with
great success at DuPont and may form the model for other
such relationships. To improve the competitive position of
the U.S. chemical industry, we must keep looking for inno-
vative ways to capture exceptional value in the marketplace
from our limited research investment. Increasing the dialog
between industry, academe, and government, and identify-
ing areas of mutual interest and potential collaboration, is
essential for improving global competitiveness. O

TW 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 which
elucidate difficult concepts. Please submit them to Professor James 0. Wilkes (e-mail:
wilkes@engin.umich.edu) or Mark A. Bums (e-mail: mabums@engin.umich.edu), Chemical
Engineering Department, University of Michigan, Ann Arbor, MI 48109-2136.

"AN ODE TO

THAT DISTILLATION TOWER"

And Other Poetry

A Creative Writing Assignment -

GREGORY L. RORRER
Oregon State University Corvallis, OR 97331

Writing in the chemical engineering undergraduate
curriculum generally assumes the form of formal
reports in laboratory and capstone design courses,
but it is generally accepted that writing assignments should
be more frequently assigned and integrated throughout the
curriculum. Short writing assignments outside of the stan-
dard report format include cover memos for homework sets
or open-ended special projects, laboratory safety briefs, or
ethical-issues essays of chemical engineering interest.
Informal writing assignments can also promote student
learning of engineering concepts. For example, Felder'" rec-
ommends in-class writing to "define a concept in your own
words." The active process of expressing an idea or concept

in writing helps the student to work through problems
with understanding. In other words, writing is learning.[2]
There is generally a much lower "activation energy" as-
sociated with informal writing assignments, as content is
valued over mechanics.
I wanted to make a short, informal writing assignment that
would serve three purposes: It should 1) reinforce chemical
engineering concepts relevant to the course material, 2) pro-
mote creating thinking, and 3) put a smile on the faces of the
serious-minded students in my class. Toward this end, I
chose a poem format. The problem statement, samples of
poetry written by the students, and a few comments on how
well the assignment worked out follow.

(PROBLEM STATEMENT)

The writing task was assigned with the last homework set
near the end of the term in a senior-level mass transfer
operations course. The assignment stated:
"Artistic literary works such as parables or poems offer a
way to communicate abstract ideas or concepts that

Copyright ChE Division ofASEE 1996

Chemical Engineering Eduction

Gregory L. Rorrer is an Associate Professor
of Chemical Engineering at Oregon State
University. He received his BS degree from
the University of Michigan in 1983 and his
PhD from Michigan State University in 1989,
both in chemical engineering. His research
interests are in biochemical engineering with
current emphasis in algal cell culturing sys-
tems.

otherwise would be difficult to understand. In this last
homework assignment, I want you to write a short poem (in
any format) that attempts to communicate a mass transfer
operations related concept to someone with a very basic
technical background, say a sophomore in chemical
engineering. I will look at these personally and evaluate the
work based on the level of thought put into it. Don't wait
until the last minute to do this last homework assignment-
creativity requires a clear head."
The assignment was also read aloud in class to gauge the
student's reaction to it. The room was filled with laughter
and a few groans, but the class as a whole seemed very
receptive to this unconventional writing assignment.

(STUDENT POETRY SAMPLES

* Sample 1 (untitled)

Packing and trays, packing and trays,
Less volume to area, less distillation days.
Valves and saddles, valves and saddles,
Better mixing, less dripping, ChemE's have time to
babble.
Sieves and rings, sieves and rings,
Less reflux and boilup, accountants do sing.
Intalox and caps, intalox and caps,
Less fumes and waste, EPA drops their bats.
Packing and trays, packing and trays,
More calculations, but getting well paid.

* Sample 2 "Crude Technology"

There once was a mixture of crude
that splurged from a hole, so good.
But it can't be used for squat,
because its volatility is shot!
So rectification is a must,
or the company will go bust!
Feed crude to the tower,
insides filled with trays that shower.
Distill that crude solution...
Heavies flow to the bottom,
lighties rise to the top.
Heat loads on the tower,
more distilled feed is profit by the hour.

* Sample 3 "An Ode to that Distillation Tower"

Ode to that distillation tower,
With all its mighty separation' power.
Takin' one little stream of this and that,
and making' two streams of mainly this or that.
But don't go thinking' it's all just touchy-feely,
'cause the rules are spelled out by McCabe and Thiele.
And adiabatic is how she's gotta run,
otherwise no one's goin' to have any fun.
Now with all these rules you're ready to distill,
And with the instructor's help you'll get your fill.

* Sample 4 (untitled)

The ascension of purity is finite in steps,
unless one is faced with azeotropic effects.
Breaking through can be attained,
and in Treybal this process is well explained.
So fire up that tower and get on with the show,
but be careful with reflux to control cash flow.

* Sample 5 (untitled)

There once was a ChemE named Joe,
Who raised the reflux ratio.
The column did flood,
Now Joe's name is mud,
And he runs the tower no mo'.

COMMENTARY
When the students turned in the poem writing assign-
ment along with the rest of their homework, several asked
me to read the poems out loud. I considered the request,
but silently read through all of the poems first. I then
selected five poems that I thought the class might enjoy
and at the beginning of the next class, I read them to the
class under the lecture topic "Poem Time." I did not
acknowledge the student authors, to protect those who
might feel embarrassed about disclosing their work.
Ollis'3' claims that reading poetry aloud from estab-
lished literary works illustrates to students how ideas can
be presented with brevity. I noticed that the students
were very attentive during the five minutes of Poem
Time. This suggests the ChE-inspired poetry, if used
sparingly but effectively, can be a unique way to bring

Summer 1996

home ChE concepts to students.
I used two simple criteria to evaluate the student work: 1)
did the topic illustrate some concept relevant to the course?
and 2) was there an attempt to put some thought into the
work? Every student except one composed one poem, and
some even composed two! Overall, I was impressed with the
level of humor and the clever use of language that the stu-
dents put into their poems. By framing the poem assignment
to illustrate a mass transfer operations concept, students
attempted to use analogies to explain technical concepts, and
in so doing exercised creativity and higher-order thinking
skills. Above all, however, an "affective objective," described
in Bloom's Taxonomy,4' may have been attained. The as-
signment was perceived as unique and fun by the students.
Therefore, their attitude toward the subject area may have
been positively affected by the assignment, which in turn
would stimulate sustained interest in the subject area.
Students in engineering generally appreciate a diversity of
activities in their coursework experiences.`56' A little levity is
sometimes needed in senior-level courses where engineering
students are burdened with the pressures of career decisions,
difficult course material, and time-consuming projects. In

this regard, timing a short poem writing assignment near the
end of the term lifted the students' spirits a little and put a
smile on this instructor's face as well.

ACKNOWLEDGMENT
The author acknowledges a grant from the Writing Inten-
sive Curriculum Program at Oregon State University that
supported the development of this writing assignment and
the preparation of this paper.

REFERENCES
1. Felder, R.M., "Any Questions?" Chem. Eng. Ed., 28(3), 174
(1994)
2. Fulwiler, T., "Writing: An Act of Cognition," in C.W. Griffin
(ed.), New Directions for Teaching and Learning: Teaching
Writing in All Disciplines, No. 12, Jossey-Bass, San Fran-
cisco, CA, 15 (1982)
3. Ollis, D.F., "The Other Three R's: Rehearsal, Recitation,
and Argument," Chem. Eng. Ed., 27(1), 30 (1993)
4. Bloom, B.S., Taxonomy of Educational Objectives: The Clas-
sification of Educational Goals. Book 1: The Cognitive Do-
main, D. McCay, New York, NY (1984)
5. Wankat, P.C., and F.S. Oreovicz, Teaching Engineering,
McGraw Hill, New York, NY (1993)
6. Felder, R.M., and R. Brent, "Getting Started," Chem. Eng.
Ed., 29(3), 166 (1995) 7

Letters to the Editors

of the "Class and Home Problems" Column

Dear Sirs:
I am writing to you regarding the article "Distillation
Column Performance," by J.A. Shaeiwitz, in Chemical En-
gineering Education, 29(4), pages 240-243 (1995). The prob-
lem is interesting in that it sets out to examine operation of
an existing piece of equipment rather than designing a new
unit (which is the most common form taken by many chemi-
cal engineering exercises). However, considerable care is
needed with such problems if the wrong conclusion is not to
be reached.
In this problem, there are two aspects that really need
further consideration.
A. Tray Performance In many distillation services,
small reductions in feed rate will allow pro-rata reductions
in all other flows and their related heat-exchanger duties.
However, as the reduction approaches 35% of the original
throughput, weeping will become significant for sieve trays
and mass-transfer performance starts to decline-that is, the
required separation is not achieved. Further feed-rate reduc-
tions will not permit corresponding reductions in heat loads;
the heat input must be maintained to produce sufficient

vapor flows to limit weeping (obviously, the condenser duty
and liquid flows will follow). In summary, at low through-
puts, the column must be artificially loaded and energy-
efficient operation is not possible.
The exact amount of turndown possible depends on where
the original 100% point lies in the sieve-tray operating enve-
lope, and the important point to note is that it is unsafe to
assume that halving the feed rate allows one to pro-rate
down all flows and duties without detailed consideration. If
feed rate reductions larger than 30-40% are likely to be
required on many occasions, the designer should specify
valve trays.
B. Condenser Operation Most condensers are designed
with cooling water flowing in the tubes at a velocity of 1.5 to
2.0 m/s; the very minimum velocity suggested is 1.0 m/s.
Generally, a maximum cooling-water return temperature of
450C is used. Both of these parameters are based on operat-
ing experience and are intended to limit heat-exchanger foul-
ing and corrosion. In the proposed solution, a velocity well
below 1.0 m/s will result if the cooling water is reduced by
Chemical Engineering Eduction

65%. This, combined with a cooling-water return tempera-
ture of 510C, i.e., 60C above the suggested maximum, will
lead to severe tube-side problems if extended operation is
undertaken in this mode. It is perhaps worth observing that a
reduction in cooling-water velocity from 1.5 to 1.0 m/s de-
fines the practical turndown of a condenser, and this is
broadly in agreement with the limit of energy-efficient turn-
downs as discussed in A above.

W.E. Jones,
Chemical Engineering Dept.
University of Nottingham
Nottingham, England

Author's Response
Dear Sirs:
I thank Professor Jones for his interest in the problem
titled "Distillation Column Performance." His observations
regarding tray performance and condenser operation are cor-
rect, and the assumptions made in this regard should have
been clearly stated.
When this problem is assigned to students, the purpose is
to demonstrate the interrelationship between a distillation
column and the required heat exchangers. The problem, as
presented, demonstrates that neither can be analyzed in iso-

lation from the other.
Professor Jones' observations suggest an extension of this
distillation column performance problem, illustrating the rich-
ness of open-ended problems. After solving the distillation
problem as in the paper, the problem with tray performance
and condenser operation could be pointed out to students. They
would then be asked to suggest alternatives for compensating
for tray performance and condenser operation limitations.
Numerous alternatives exist, and the new assignment
would be an excellent creativity exercise. One alternative is
to replace equipment. Valve trays and small-diameter con-
denser tubes could be installed. Another alternative is to
maintain the original boil-up rate from the reboiler, or just
increase the boil-up rate from the scaled-down value enough
for the trays and condenser to operate correctly. This option
also requires an increased reflux ratio, which should result in
a better separation. If a better separation were not desired,
the feed location could be moved, equipment permitting, to
reduce the separation.
Consideration of these two alternatives might lead to a
discussion of the economics of replacing equipment versus
changing operating conditions.
Joseph A. Shaeiwitz
Dept. of Chemical Engineering
West Virginia University
Morgantown, WV 26506-6102

- re] M stirred pots

To the Editor:
A while ago I downloaded from the Internet a program called Karma Manager, which makes anagrams of any word or
phrase you input. It determines all possible sets of words that can be made by rearranging the letters of whatever you type
in (ignoring spaces), and it returns each set to you in a list. After typing in a few names and finding little (six entries for
my name, the most interesting being kavaa kid of Ed"), I entered "thermodynamics" and observed over 10,000 anagrams
emerge! Karma Manager merely presents the sets of words, without ordering them in a way that might make sense. I
didn't have enough free time to look at them all, but here are some of the interesting ones I found.

dim men try chaos
consider my math
charm in modesty
my romantic shed
its my amen chord
my sham doctrine
my hindmost care
hamster in my cod
shy dormant mice

sir, end thy comma
cram into my shed
emit many chords
so I mend my chart
many cords hit me
my thin comrades
do me in my starch
decant or shammy
my damn sore itch

some rancid myth
mystic harm done
them micron days
dim men crash toy
I deny most charm
some thin mad cry
Oh stem racy mind
short icy madmen
my son came third

had my nice storm
sad men cry to him
scorn media myth
scare my hot mind
macho men sit dry
most handy crime
not my cider mash
shy men or dim cat
me and my ostrich

Karma Manager (which itself is an anagram) can be obtained by going to the Web site http://www.shareware.com and
searching on "Karma."
David A Kofke SUNY at Buffalo

Summer 1996 18.

Survey

THE CHEMICAL ENGINEERING

CURRICULUM-1994

RONALD N. OCCHIOGROSSO, BANITA RANA
Manhattan College Riverdale, NY 10471

he most recent survey of the chemical engineering
undergraduate curricula, conducted every five years
since 1957, was made by the Undergraduate Curricu-
lum Subcommittee of the Education Projects Committee of
the American Institute of Chemical Engineers (AIChE) in
the summer of 1994.1" A questionnaire that closely corre-
sponds with ABET/AIChE categories was sent to the 158
chemical engineering departments listed with AIChE; sixty-
three departments responded, and the survey results are based
on those responses.
The spreadsheets that contained the data collected were
prepared using Quattro Pro (Borland Int'l.) to assist in ana-
lyzing the survey results. Table 1 summarizes the responses
received. As shown, seventeen Canadian schools were sent
the questionnaire, but only one responded. This school's
data was difficult to translate into reasonably related US
numbers, so it was not included in the report. The data
available from US schools was reported at the Department
Heads' Meeting of the AIChE annual meeting in Miami
Beach, Florida, in November of 1995.

Ronald N. Occhiogrosso is Assistant Profes-
sor of Chemical Engineering at Manhattan Col-
lege. He received his BS from Manhattan Col-
lege (1983), his MS from Notre Dame (1985),
and his PhD from Johns Hopkins (1987), all in
chemical engineering. Teaching and research in-
terests include supercritical fluid technology, poly-
mer science and engineering, advanced separa-
tion technology, and SPC.

Banita Rana is currently working for Allee
King Rosen & Fleming, Inc., an environmen-
tal and planning consulting company in Man-
hattan, New York. She obtained her MS in
chemical engineering from Manhattan College
and her BE from the University of Roorkee,
India.

TABLE 1
Summary of Responses Received

United States Canada
Total Schools Surveyed ................................ 158.................. 17
Total Responses Received ............................... .................... 1
Percent Responses .......................................... 39.9% ............. 5.9%
Percent Responses Overall....................... 36.6%

SURVEY RESULTS AND DISCUSSION
The semester credit hours required for the Bachelor's De-
gree remains almost the same as it was in the previous 1989
survey (when ninety-two schools responded). Figure 1 shows
that the trend seems to have stabilized in the low 130s. The
detailed information on the spreadsheet, however, indicates
that semester hours actually range from 115 to 145. The
lower bound has increased only slightly since the last sur-
vey. More than 80% of the departments require 125 to 140
semester hours, with only six having fewer than 125 and five

140 _

138 _

136 -

134 -

132

130
1950 1960 1970 1980 1990 2000

Figure 1. Total semester hours required for under-
graduate chemical engineering degree.

@ Copyright ChE Division ofASEE 1996

Chemical Engineering Eduction

[This survey] was conducted by the Undergraduate Curriculum Subcommittee of the
Education Projects Committee of the American Institute of Chemical Engineers (AIChE) in the
summer of 1994. A questionnaire that closely corresponds with ABET/AIChE categories was sent to the
158 chemical engineering departments listed with AIChE; sixty-three departments responded,
and the survey results are based on those responses.

having more than 140. Most schools operate on the Nominal
Semester basis; few use a quarter system.
As in the previous survey, the average curricular area
distribution continue to be somewhat close to ABET/AIChE
requirements. A closer look at the individual departmental
requirements reveals a wide range (see Table 2). For ex-
ample, mathematics, which has an average value of 16.5
contact hours (semester credits), ranges from 12 to 22 hours.
Most of the departments' requirements fall within 15 to 18
hours. Expressed as a percentage, the mathematics require-
ment is 12.5%, equaling the AIChE requirement of 12.5%.
Similar traits are observed in other categories. For ex-

TABLE 2
Distribution of Course Work

AIChE 1981 1985 1989 1994
Curricular Area % Avg Avg Avg Avg
Mathematics beyond Trigonometry 12.5 13.6 12.7 12.4 12.5
Basic Sciences 25.0 24.3 25.4 24.8 24.1
(Incl. Advanced Chemistry) (12.5) (11.7) (12.8) (12.3) (11.9)
Engineering Sciences/Design 37.5 37.3 37.2 39.7 39.7
Humanities/Social Sciences 12.5 16.1 15.1 13.5 14.6
Other 12.5 8.7 9.7 9.6 8.9
Total Percent 100.0 100.0 100.0 100.0 100.0
Total Credit Hours 133.4 131.4 132.8 132.8

35
30
25
20
15
10
5

0 1957 1961 1968 1972 1976 1981 1985 1989 1994

Figure 2. Chemistry content exclusive of advanced
chemistry.

Summer 1996

ample, "engineering science/design" and "humanities/social
sciences" continue to increase their share and move away
from ABET/AIChE requirements at the expense of other
categories that show a downward movement.
Some changes can be observed within the categories. Math-
ematics, which used to be predominantly calculus and dif-
ferential equations, show that calculus has maintained its
dominance, but differential equations has lost some of its
share to analytical geometry and advanced calculus.
The remaining credits still demonstrate wide diversity.
The popularity of linear algebra has increased to 26 depart-
ments (41%) requiring the course, compared to 20 depart-
ments (22%) in 1989. Partial differential equations show
reduced popularity. Many departments require a math-
ematics elective.
In the fundamentally important "basic science" category,
introductory physics and chemistry continue to maintain
diversity and dominate the credit hours. But ten departments
reported modem physics, five listed biology, and four indi-
cated that other basic sciences are required. A comparison of
these numbers to previous survey results indicates that the
popularity of these courses is decreasing. The advanced
chemistry requirements showed a drop to 11.95% from the
AIChE requirement of 12.5%, but, as usual, showed a wide
range of 10.67 to 21.0 hours. The total chemistry contents, as
indicated by Figure 2, has maintained a stable trend.
The engineering science and design requirements in-
creased to 39.9% from their 39.7% value in the 1989
survey. The statics course has suffered a significant drop
in popularity, with nearly 68% of the departments offer-
ing the course in 1989 and only about 59% in 1994. But
dynamics and/or mechanics of materials maintain their
positions, with approximately one-quarter of the depart-
ments requiring the courses.
Another startling change is seen in the introduction to
electrical engineering courses, which suffered a major drop
in popularity from 65% of departments requiring the courses
in 1989 to only 48% in 1994; material science maintained its
position at about 46%.
The chemical engineering component constitutes 65% of
the engineering category, a drop of 5% from the previous
survey. Although transport phenomena and unit operations
do overlap in course content to a considerable extent, they
suggest a difference in focus. The number of departments

0 % of curriculum

*I no of hrs

requiring unit operation theory decreased significantly from
about 74% to 63%, but departments requiring unit opera-
tions laboratory is almost 94%.
In the case of transport phenomena, theory courses were
reported by nearly 80% of the departments; the laboratory
component was reported by only half of that number. Mass
transfer is offered by 58%, and process control and process
dynamics were reported by approximately 86% and 57% of
the departments, respectively. Reactor design is required by
three-quarters of the departments.
Regarding electives, twenty-one specific electives and a
broad "other" choice were included in this category of the
questionnaire. The results are given in Table 3. Biochemical,
polymers, and the environmental electives continue to be the
top three, with approximately 49% of the departments offer-
ing them. There has been a shuffling of positions between
other electives. Equipment and energy related areas, such as
natural gas and fuel, are still in the lower end.
The cultural content (which includes the humanities and
the social sciences) has managed to break its declining trend
of the past three surveys and is approximately equal to that
of the 1985 survey (see Figure 3); it never actually reached
the ABET minimum value of 12.5%. Interestingly, the range
of credit hours required has narrowed, with the low end
moving up from 6 to 13 hours and the high end moving
significantly down from 55.3 to 37.36 hours. The high actu-
ally moved from 42% of the program to 28%, while the low
end shifted from 5% to 10%.
The fifth and final major section of the questionnaire was
classified as "other" and included diverse course offerings.
The communication category formed a major portion of this
section, but it has shown fluctuation over the years. In accor-
dance with this fluctuating trend, it decreased from 90% to
slightly less than 80%. As one would normally expect,
computer programming (which is in this fifth, "other"
category) was another course required by a significant
number of the departments. Figure 4 shows results of
responses for this category.
Table 4 depicts an average program that could be used
for comparison. The information provided in this table
might be useful to a school starting up a chemical engi-
neering program.
Table 5 indicates that the average department reported 8%
foreign undergraduate students, but 48% foreign graduate
students. There are on the average 7.04 full professors,
2.59 associate professors, 1.83 assistant professors, and
0.62 full-time equivalent other faculty in the 63 schools
that responded to the survey.
About 14% of salaries are obtained from other than gen-
eral educational funds, and there are about 0.59 faculty posi-
tions available on the average. The number of faculty posi-
tions for the 63 reporting schools is about 37, but closer

TABLE 3
Elective Offerings

Elective #Depts %

1. Biochemical
2. Polymers
3. Environmental
4. Transport Phenomena
5. Applied Mathematics
6. Control
7. Biomedical
8. Design
9. Mass Transfer
10. Reactors
11. Electrochemistry

34 54.0
29 46.0
29 46.0
11 17.5
16 25.4
13 20.6
13 20.6
8 12.7
8 12.7
6 9.5
10 15.9

Elective #Depts %

12. Petroleum 5 7.9
13. Catalysts 6 9.5
14. Paper 3 4.8
15. Nuclear 4 6.3
16. Coal 6 9.5
17. Energy 4 6.3
18. Equipment 1 1.6
19. Food 2 3.2
20. Fuel 2 3.2
21. Natural Gas 2 3.2
22. Others 27 42.9

25-

20

15

10

5

0
19571961 19681972 198119851989

E % of curriculum

no of hrs

Figure 3. Cultural content.

100

90

80

70

60

50
50 19571961 1968197219761981 198519891994

Figure 4. Communications (% of schools offering).

Chemical Engineering Eduction

TABLE 5
Summary of Student and Faculty Information

Total Avg.

Students (ChE)
Fraction, Non-U.S. Undergraduate
Fraction, Non U.S. Graduate

Faculty (Number of)
Full-time professors
Full-time associate professors
Full-time assistant professors
Full time equivalent, other teaching staff
% of salaries funded from other than
general education funds
Number of full-time faculty positions
open (tenure track )

443.63 7.04
163.33 2.59
115.00 1.83
39.3 0.62

N/A 13.99

37.00 0.59

TABLE 4
Average Program Abstract

Course Hours
Analytical Geometry ............... .................... ...2.94
C alculus ................................... .................................. 8.3 1
Differential Equations ...................................................3.21
General Physics .............................. .........................7.67
General Chemistry ................................... ................... 7.69
Physical Chemistry .......................... ........................6.36
Organic Chemistry ......................................7.10
Other Chemistry ............................... ....................... 3.98
S tatics ........ ..................... ....................... .................... 3.06
Electrical Engineering ..................... .......................... 3.77
Material Science ............................. ......................... 3.94
Fluid Mechanics ............................. .......................... 2.73
Heat Transfer .................................................. .................. 2.56
Material and Energy Balances....................................... 3.61
Thermodynamics .............................. ........................ 4.22
Reaction Engineering .................................................... 1.88
Transport Phenomena ....................... ....................... 3.88
Mass Transfer ........................................ ..................... 2.97
Unit Operations ................................... .................... 3.20
L laboratory ........................................ ....................... 3.74
Process Control ............................... ......................... 2.36
Design ............... .................................................. 4.93
ChE Electives ................................ ......................... 5.93
Humanities ..................................... ........................ 9.72
Social Science ........................................ ....................... 6.99
Communications ..................................... ....................... 6.82
Computer Programming ......................... ..................... 2.58
Other ........................... ................................................ 6.62
Total 132.77

Summer 1996

understanding of the variation among departments necessi-
tates a review of the entire information contained in the data
received in all of the 63 responses.
The spreadsheets have been made available for all par-
ticipating departments. For others who are interested, the
spreadsheet will be made available upon request made in
writing to the authors.

CONCLUSIONS
C Sixty-three departments of chemical engineering
(out of 158 schools that were solicited) completed
the most recent survey conducted by the Under-
graduate Curriculum Subcommittee of the AIChE
Education Projects Committee. The number of
credits requirement for a BS degree in chemical
engineering ranges from 115-145 on the ABET
semester basis.
C The average number of credits required for a BS
degree in the U.S. has remained almost the same
at about 133 credits for the past twenty-two years.
C The chemistry content remained approximately
the same for the past ten years, while the cultural
content appears to have fluctuated the most for
the past twenty years. The number of schools
offering communications has seemed to decrease
since 1985, although it had initially increased
since 1976.
C The results in Table 2 indicate that the distribu-
tion of course work has remained fairly constant
for the past thirteen years. There has been a slight
increase in math courses beyond trigonometry,
although it had decreased slightly and fairly
steadily until 1989. Basic sciences has decreased
slightly and fairly steadily for the past nine years;
engineering science has increased in a similar
fashion.
C The results in Table 3 indicate that biochemical
electives are offered at the highest percentage of
the schools.
C Table 4 provides an average program abstract of
the course offerings and indicates little change
since the last survey, performed in 1989.1

ACKNOWLEDGMENTS
We would like to thank Dr. Deran Hanesian and Dr. Angelo
Perna of NJIT for providing useful information and insight
that went into the preparation of this article.

REFERENCES
1. Coulman, G.A., "The Chemical Engineering Curriculum,"
Chem. Eng. Ed., 23(4), 184 (1990). Note; this reference con-
tains a bibliography (seven related references) for this paper's
topic. 0

Random Thoughts...

IF YOU'VE GOT IT, FLAUNT IT

Uses and Abuses of Teaching Portfolios

RICHARD M. FIELDER, REBECCA BRENT*
North Carolina State University Raleigh, North Carolina

A memo from the Provost appears in all faculty mail
boxes one morning, announcing that from now on
every candidate for tenure and promotion must sub-
mit a teaching portfolio along with the usual research docu-
mentation. Faculty reaction is swift and divided, even though
no one understands exactly what is being required or why.
Some professors see the requirement as an indication that
the administration is finally starting to take teaching seri-
ously, others view it as just another drain on their time that
won't accomplish anything useful and could hurt them. Ei-
ther viewpoint could turn out to be correct, depending on
how the portfolio program is handled.
A teaching portfolio is a collection of materials that docu-
ment a professor's teaching goals, strengths, and accom-
plishments. It contains
EO Self-generated material (e.g., a teaching philosophy
statement; representative syllabi, instructional objec-
tives, handouts, assignments, and tests; descriptions of
educational innovations and evaluations of their effec-
tiveness; textbooks and education-related papers pub-
lished; instructional software developed; teaching work-
shops and seminars presented or attended).
E Teaching products (e.g., graded assignments, tests, and
reports; scores on standardized tests; student publica-
tions or presentations on course-related work).
E Information generated by others (e.g., summaries of
student, alumni, and peer evaluations; honor and awards;
reference letters).
Some items may be mandated, others may be included at the
professor's option.
Portfolios have been used to document college teaching
performance beginning in Canada in the 1970s, and their use

*Address: School of Education, East Carolina University,
Greenville, North Carolina

has become increasingly widespread since the 1991 publica-
tion of The Teaching Portfolio by Peter Seldin.u" Despite
abundant evidence that their use improves teaching,121 the
required inclusion of teaching portfolios in promotion and
tenure dossiers is often viewed with faculty skepticism. This
attitude may prove to be justified, as poorly designed or
implemented portfolio programs are likely to have a mini-
mal impact on institutional teaching quality and a negative
impact on faculty morale. In the remainder of this column,
we extract ideas from Seldin',21 on ways to avoid the pitfalls
and make portfolio programs effective.

What is the purpose of the teaching portfolio? A
portfolio can be used for summative evaluation (to evaluate
teaching performance and provide a rational basis for pro-
motion and tenure decisions and teaching award selections)
or formative evaluation (to help identify and correct teach-
ing problems). What goes in the portfolio depends on which
function is intended. For summative evaluation, the portfo-
lio should include some mandated items like a teaching

^- Richard M. Felder is Hoechst Celanese Pro-
fessor of Chemical Engineering at North Caro-
lina State University. He received his BChE from
City College of CUNY and his PhD from
Princeton. He has presented courses on chemi-
S cal engineering principles, reactor design, pro-
cess optimization, and effective teaching to vari-
ous American and foreign industries and institu-
tions. He is coauthor of the text Elementary
Principles of Chemical Processes (Wiley, 1986).
Rebecca Brent is Associate Professor of Edu-
cation at East Carolina University. She received
her BA from Millsaps College, her MEd from
Mississippi State University, and her EdD from
Auburn University. Her research interests include
applications of simulation in teacher education
and writing across the curriculum. Before joining
the faculty at ECU, she taught at elementary
schools in Jackson, Mississippi, and Mobile, Ala-
bama. She received the 1994 East Carolina Uni-
versity Outstanding Teacher Award.

Copyright ChE Division ofASEE 1996

Chemical Engineering Eduction

philosophy and a summary of student ratings and some
optional items that reflect on teaching performance and edu-
cational scholarship (e.g., student products, descriptions of
teaching innovations, and reference letters from alumni or
colleagues). For formative evaluation, the choice of content
is entirely up to the professor and the focus should be on
problem areas. The same portfolio should not be used for
both functions.

How should summative portfolios be designed and evalu-
ated? Three key requirements for effective portfolios are
relevance (the elements selected for evaluation must be clearly
linked to established criteria for effective teaching), reliabil-
ity (ratings from different evaluators should be reasonably
similar), and practicality (portfolios should be well orga-
nized, not too long, and easy to evaluate). While the ideal
portfolio structure may vary considerably from one institu-
tion to another and from one discipline to another, the fol-
lowing design procedure is broadly applicable:
1. Select categories that will be used to define the quality
of a professor's teaching performance (e.g., course
design, instructional delivery, content expertise, de-
velopment of new instructional methods and materi-
als), and assign relative weights to each category.
2. Formulate an objective set of questions addressing
each category (e.g., questions for the course design
category might include, "Are the instructional objec-
tives appropriate and consistent with the candidate's
teaching philosophy and with institutional or depart-
mental goals?" "Are the assignments and tests consis-
tent with the objectives?")
2. Specify required portfolio materials that will help
provide meaningful answers to the questions.
Once a summative portfolio has been prepared, several people
should independently examine it, rate each category using a
predefined system (e.g., 0=poor, 5=outstanding), calculate a
weighted average rating, attempt to reconcile widely diver-
gent evaluations, and finally provide a collective rating.

What is the point of the teaching philosophy statement?
* The philosophy statement enables portfolio evaluators to
judge how well institutional goals and generally accepted
criteria for good teaching are reflected in the professor's
objectives, and the remaining portfolio contents can then be
used to assess how well the objectives are being met. Good
teaching is clearly being done when appropriate goals have
been chosen and the portfolio contents demonstrate success
in achieving them. Moreover, simply reflecting on why we
do what we do in the classroom is likely to improve our
teaching, even if the portfolio preparation goes no further.

How should new professors be assisted with portfolio
preparation? There should be no secret about what

constitutes an outstanding portfolio and what constitutes an
acceptable one. Discipline-specific model portfolios, like
the illustrative ones given by Seldin,m2' should be shown to
professors at the outset of the process, and faculty colleagues
or campus teaching consultants should be available as port-
folio mentors to offer guidance and support. The mentors do
not have to be in the same disciplines as the professors they
are helping, but they should clearly understand the evalua-
tion criteria used in those disciplines.

How should a portfolio program be initiated and institu-
tionalized? Seldin cautions, repeatedly and emphatically,
that a portfolio program developed by administrators and
imposed on the faculty will probably not achieve its objec-
tives, and suggests several ways to promote institutional
acceptance.12 Administrators at all levels (department, school,
and institution), in collaboration with the faculty, should set
clear standards for both outstanding teaching and acceptable
teaching, and they should publicize the portfolio evaluation
criteria so that faculty members are clear about institutional
expectations. The program should be pilot-tested on volun-
teers, including some of the most prestigious teachers and
researchers on the faculty, before an attempt is made to
institutionalize it. The administration should support portfo-
lio development workshops and mentorships, e.g., by pro-
viding release time or other compensation for the workshop
leaders and mentors.

Perhaps most importantly, the administration should dem-
onstrate by actions as well as words its commitment to take
portfolios seriously when making personnel decisions. If
professors with strong teaching portfolios are treated the
same as professors with strong research records in promo-
tion and tenure decisions, faculty acceptance is likely to
follow and the portfolio program has a good chance of
working. Conversely, if professors with strong teaching port-
folios and weak research records are denied tenure while
others with weak teaching portfolios and strong research
records get it, faculty acceptance will almost certainly be
unattainable and the portfolio program is likely to fail.
This synopsis hardly does justice to the wealth of models
and tips Seldin offers for portfolio preparation and evalua-
tion. Anyone thinking about implementing a portfolio pro-
gram should study the references and, if possible, attend a
Seldin workshop. The potential impact of the program on
teaching quality justifies doing whatever it takes to get it
right the first time.

REFERENCES
1. Seldin, Peter, The Teaching Portfolio: A Practical Guide to
Improved Performance and Promotion/Tenure Decisions,
Anker Publishing Company, Inc., Bolton, MA (1991)
2. Seldin, Peter, Successful use of Teaching Portfolios, Anker
Publishing Company, Inc., Bolton, MA (1993) 0

Summer 1996

rMSe survey

TEACHING COLLOID AND

SURFACE PHENOMENA

-1995-

DONALD R. WOODS, DARSH T. WASAN*
McMaster University Hamilton, Ontario, Canada

Variety. Variety in topic, in emphasis, and in approach-
that's what we found from a 1995 survey of how
colloids and surface phenomena is taught today. This
is really not surprising, though, because of the variety in the
topic itself: surfaces and interfaces; surfaces separate any two
phases. So the applications can be gas-liquid, gas-solid, liquid-
liquid, liquid-solid, and solid-solid boundaries. The materials
that reside in fluid surfaces-surfactants-represent unique
species with interesting behaviors such as micellization, liquid
crystals, cosurfactants, and/or microemulsions. Applications
abound. Surface phenomena is an integral part of water and
waste-water treatment, physical separations, catalysis, poly-
mer production, mineral processing, ceramics, and biomedical
systems. Surface phenomena has growing applications in mass
transfer, fluid mechanics, heat transfer, homogeneous phase
separations, and reaction engineering.
So, how is this material taught to today's professionals?
Rarely! In our survey, sent to 180 chemical engineering depart-
ments in the United States and Canada, only nineteen schools

Don Woods is a professor of chemical engineer-
ing at McMaster University. He is a graduate of
Queen's University and the University of Wis-
consin. His teaching and research interests are
in surface phenomena, plant design, cost esti-
mation, and developing problem-solving skills.

Darsh Wasan received his BS from the Univer-
sity of Illinois, Urbana, and his PhD from the
University of California, Berkeley, both in chemi-
cal engineering. He has spent his entire profes-
sional career at the Illinois Institute of Technol-
ogy, where he has held virtually every academic
and administrative post.

* Address: ChE Department, Illinois Institute of Technology,
Chicago, IL 60616

reported that they teach at least one course. Five schools teach
two or more courses (Princeton, Carnegie-Mellon University,
University of Washington, University of Minnesota, and
McMaster Unitersity). Except for a required junior-level course
in a ceramic engineering program, only eleven schools offer
this as a senior elective.
In this paper we will illustrate how the courses are taught,
with an emphasis on context and with varying emphasis on the
content (properties, phenomena, theory, practical and experi-
mental). Ideas will be presented on how to demonstrate and
measure the phenomena. Resources will be given. Methods for
teaching courses will be summarized, and we will close by
giving ideas about future developments.

CONTEXT: CORE FUNDAMENTALS
Sometimes surface phenomena is presented in the context of
environmental engineering, biomedical engineering, particle
processing, catalysis, ceramic or materials engineering; some-
times the central theory of surface phenomena is given with
little discussion of applications. Some instructors focus on the
theory, some blend applications with theory, and a few use the
applications as the focus, with the theory being learned for the
purpose of designing a device or a process.
Part of the difficulty in offering an applications approach is
the lack of available design data. Some instructors bring in
applications through research and consulting. Those that offer
courses focusing on the context independent fundamentals
include Radke(14), Israelachvili(19), Jacobson(20a),
DiMilla(20b), Prieve(20e), Evans(75a,75b), Saville(105a),
Russel(105b), Miller(110), Ploehn(116), Slattery(128),
Zollars(143), Thies(144), and Berg(145b).* Others place more
emphasis on adsorption at gas-solid surfaces (Fort(139)) and
on catalysis (Ko(20c)). Some instructors work in the context of
* Numbers in parentheses are the numbers assigned to entries in
the "Summary of Responses" list appearing in the Appendix to
this article.

@ Copyright ChE Division ofASEE 1996

Chemical Engineering Eduction

solid processing (whether it be transportation, separation, or reaction): Scheiner(3),
Tiller(45), Chiang(102b), and Nicholson(159c), and on coating: Tallmadge(38). Some
instructors describe a variety of applications that are not focused on any particular
industry or unit operations: Wasan(49), Bike(73), Phule(102a), Botsaris(134), and
Woods(159b). A few instructors focus on polymers: Anderson(20d) and Pelton(159a).
Ratner(145a) has some biomedical applications while environmental applications are
given by Dold, et al.(159d).

SURFACE PROPERTIES, THEORY,
SURFACE PHENOMENA, PRACTICAL APPLICATIONS
Being able to measure surface properties is seen as a key by many instructors. It helps
measure the required data .. and the process of measuring the data helps the under-
standing of the phenomena. Some instructors provide a laboratory course to comple-
ment the theoretical courses, while others blend the laboratory experiments into the
theory course. One course uses the laboratory to drive the learning. Some teachers
emphasize the experimental methods in this courses, and others place little
emphasis on how to measure properties. Some use demonstrations of measure-
ment techniques or of the phenomena.
To capture more of the flavor of the courses presented, we classify the approaches
according to six extremes. These are based on how much emphasis seems to be placed
on
1. The theory of surface and colloidal properties (e.g., surface tension) called "theory-
property. Instructors here would provide mathematical descriptions and derivations.
2. The properties and the theory/application of how to measure (e.g., surface tension and
how to measure) called "description-measurement properties." With this approach,
instructors might describe the theory, correlations, and methods to experimentally
measure surface tension.
3. The measurement of surface properties (e.g., experimental measurement of surface tension
via Wilhelmy plate) called "measurement-properties." Teachers might demonstrate how to
experimentally measure the property, they might ask students to estimate values for the
surface tension from experimental data, or students might perform a laboratory.
4. The theory of the behavior of surfaces and colloidal systems (e.g., theoretical definition,
estimation, and modeling of surface tension) called "theory-behavior." Professors might
provide mathematical descriptions and derivations.
5. Modeling and understanding the behavior of surfaces and colloidal systems (e.g., surface
tension variation and Marangoni behavior) called "description-behavior." Instructors in
this course would emphasize the phenomena that occur because of surface tension:
wetting, Marangoni behavior, capillarity, and fingering. They might illustrate the practical
applications, such as breakup of drops, prilling for fertilizer production, drop size in
emulsion polymerizers, gas bubble diameters in aeration basins, and ink-jet printing.
6. The applications of the behavior (e.g., explain the faulty performance of a solvent
extraction unit because the wrong phase is the dispersed phase) called "application-
behavior." This course could be presented by lecture, demonstration, videotapes, or labs.
Instructors might expect numerical calculations of the practical applications; this usually
requires the addition of engineering practice (such as information about mixing character-
istics and the dependence of drop size on the Weber number).
In general, rarely are courses or texts given that are strictly "theory-properties." Many
texts (and courses) are blends of theory of behavior combined with theory and descrip-
tion of measurement techniques for properties. For example, Hiemenz's book"' is about
half and half of these combinations. Applications are listed occasionally, but are not
emphasized (hence, we would code or describe courses given with this approach as
being "theory, behavior, properties"). Contrast this with the Evans and Wennerstrom
text."2' They place introductory emphasis on experimental measurement of the properties

In this paper we

will illustrate

how the courses

are taught, with

an emphasis on

context and with

varying emphasis

on the content

(properties,

phenomena,

theory, practical

and

experimental).

Ideas will be

presented on how

to demonstrate

and measure the

phenomena.

Resources will be

given. Methods

for teaching

courses will be

summarized, and

we will close by

giving ideas

about future

developments.

Summer 1996

TABLE 1
Experiments

Topic Elaboration Labs Demonstrations

Surface tension measurements DuNouy ring - - - - - - - - Jacobson(20a); Berg(145b) - - -Fort(139); Wasan(49)
Wilhelmy plate- - - - - - - - Zollars(143); Berg(145b) - - - Fort(139); Woods(159b); Wasan(49)
Drop weight -- - - - - - - Zollars(143); Berg(145b)
Sessile drop - - - - - - - - Berg(145b)
Maximum bubble pressure - - - - Zollars(143)

Surface pressure of insoluble monolayers - - - - - Myristic acid, Berg(145b) - - - Fort(139); Jacobson(20a)
Equilibrium contact angles Equilibrium for liquids - - - - - on polymers, Jacobson(20a)- - -Fort(139); Woods(159b); Wasan(49)
Wetting properties - - - - - plasma-treated polypropylene
Berg(145b)
Dynamic advancing/retreating - - - Zollars(143); Berg(145b)

Coefficient of friction Jacobson(20a)
Particle size measurement By photo counting - - - - - - - - - - - - - - - Woods(159b)
By Coulter counter - - - - - Jacobson(20a) - - - - Woods(159b)
By centrifugation - - - - - - Zollars(143); Berg (145b)
By QELS - - - - - - - - Zollars(143)
By sedimentation - - - - - - - Berg(145b)

Particle/surface characterization By SEM - - - - - - - - - Partch(24) - - - - - Jacobson(20a)
By TGA - - - - --- - - Partch(24)
By ESCA - - - - - - - -- - - - - - - - - Ratner(159a)
By STM - - - - - - Jacobson(20a)
By TEM--- - --------- Partch(24)

Particle preparation Submicron spherical silica particles - - Partch(24)
Coating w/polymers via in-situ polymerization- Partch(24)

Molar mass of polymers Light scattering - - - - - - - Zollars(143)
Electrophoretic mobility - - - - - - - Jacobson(20a); Zollars(143)
Electrophoresis - -- - - Dold(159d); Berg(145b)
Surface viscosity Of monolayer via deep channel - - - Berg(145b); Wasan(49)
Viscosity Effect of particulates on - - - - - Zoliars(143)
CMC determinations By surface tension - - - - - - Jacobson(20a)
By conductance - - - - --- Jacobson(20a); Berg(145b)
By dye titration - - - - - - - Berg(145b)
Surface adsorption or surface area
determination by BET - - - - - - - Jacobson(20a); Berg(145b)

Adsorption from solution - - - - - - - - - - Jacobson(20a); Berg(145b)

Colloid stability Jar test for clay removal - - - - Jacobson(20a); Dold(159d); Berg(145b)
and electrolyte addition - - - - - Berg(145b)
Soap bubbles Drainage, equilibrium angles - - - - Berg(145b)
Stability ----------------- ------ -------- Wasan(49)
Flotation Separation by preferential wetting- - - Berg(145b)

Emulsions Preparation and testing, HLB - - Berg(145b)
Stability - -------- -- - - ----- Wasan(49)
Deep bed filtration - - - - - - Dold(159d)

Surface filtration - - - - - - - Dold(159d)

Marangoni effects During mass transfer via Schlieren optics -Berg(145b)

Solubilization of dyes by aqueous
surfactants ------------- - Jacobson(20a)
Adsorptive bubble fractionation of dye - - - - - - - - Jacobson(20a); Berg(145b)
Scanning tunnel microscopy - - - - - - - - - - Jacobson(20a)

192 Chemical Engineering Eduction

and focus more on the phenomena and some applications
(code, theory-behavior).
Given the limitations of this classification, we have tried to
illustrate where most of the courses reside. This is based on our
knowledge of the texts and the descriptions, course outlines,
and exams submitted by the respondees to our questionnaire.
More theoretical courses seem to be offered by Prieve(20e),
Saville(105a), Russel(105b), and Slattery(128). A mix of theory,
behavior description, and property measurements seems to be
given by Radke(14), Israelachvili(19), DeMilla(20b),
Anderson(20d), Ploehn(ll116), Fort(139), and Thies(144).
Slightly more emphasis on behavior description is used by
Ko(20c), Bike(73), Evans(75a,b), Miller(l 10), Berg(145b), and
Pelton(159a), while slightly more emphasis on instrumentation
and measurement is given by Zollars(143) and Ratner(145a).
Heavier emphasis on phenomena and application is given by
Phule(102a), Botsaris(134), Shaeiwitz(149), Wasan(49), and
Nicholson(159c). Increased emphasis on practical applications
is given in courses by Tiller(45), Chiang(102b), and

Woods(159b). A blend of practical application and experimen-
tal measurement is given by Dold, et al. (159d), and laboratory
courses are given by Berg (145b), Jacobson(20a),
Nicholson(159c), and Partch(24).

LABORATORIES, DEMONSTRATIONS
Laboratories are offered as separate courses to complement
other courses: Jacobson(20a). Others have a required labora-
tory component where the student teams must complete some
experiments: Berg(145b) and Zollars(143). This is a mixture of
property measurement and phenomena demonstrations. The
list of experiments is given in Table 1. Other programs
(Evans(75a,b)) use laboratory experiments and demonstrations
as side enrichment via CD ROM or other media. Some teach-
ers emphasize analytical instruments and have combinations of
theory, demonstrations, and laboratory visits: Zollars(143) and
Ratner(145a), while others have demonstrations and perhaps
visits to laboratories: Wasan(49), Fort(139), Berg(145b), and
Woods(159b). Some demonstrations that instructors have used
are given in Table 2.

TABLE 2
Demonstrations

Concepts

Young-LaPlace equation
Surface tension

Contact angles

Marangoni behavior
Phase separation

Structure in suspensions

Characteristics of emulsions
Radii of curvature
Surfactants
Particle interactions in
colloidal suspensions
Film rheology

Wetting and spreading of
oil at air/water surface

Demonstrations

* Uneven sized soap bubbles on ends of a tube
* Equilibrium angles for intersecting soap films
* Floating loop of string on water with soap touching center of loop
* Floating razor blade
* Demonstrations from C.V. Boys
* Soldering of copper pipe
* Mixing cocoa in milk
* Oil spill cleanup and "herders"
* Spreading from liquid lenses and solid crystals
* Wave damping effect caused by spread films
* Ethanol, water, water plus 10x CMC value of surfactant on glass
of overhead projector
* Floating needle or razor blade plus soap touches surface

* Water plus drops of ethanol
* Slide with water and polydispersed clay particles
* Gold sol plus 2% gelatin plus salt addition; vary sequence of addition

* Monodispersed, highly charged latex suspension; shine layer through
slide to show inner structure
* Oil/water and water/oil emulsions in office supply device
* Glass of water and model of surface with normals
* Film balance and LB films

* Particle structure formation in colloidal suspensions quantified by
light diffraction

* Interference colors produced during the drainage of vertical foam
film. Each color indicates the local film thickness.

* Interference patterns in reflected light produced from the crude oil
layer at air/water interface.

Used by...

Berg(145b); Woods(159b)
Berg(145b); Woods(159b)
Woods(159b)
Nikolov and Wasan(49)
Berg(145b)
Woods(159b)
Woods(159b)
Woods(159b)
Fort(139)
Fort(139)
Nikolov and Wasan(49)

Nikolov and Wasan(49)

Nikolov and Wasan(49)
Nikolov and Wasan(49)
Lyklema

Nikolov and Wasan(49)

Nikolov and Wasan(49)
Woods(159b)
Fort(139); Berg(145b)

Wasan(49)

Wasan(49)

Wasan(49)

Summer 1996 19i

RESOURCES
Many instructors have created their own notes. Table 3 lists
the major texts used. In addition, some excellent videotapes are
available; they are listed in Table 4. Theo Overbeek's lectures,
videotapes, problems, and answers provide another rich source
of information and have been used by Botsaris(134),
Pelton(159a), and Woods(159b).
The American Filtration Society has been so concerned about
the lack of undergraduate courses in the areas of particle pro-
cessing that over the past five years they have held educational
workshop-conferences to bring together academia and indus-
try to design texts in four topics: particle science (particle
characterization and surface phenomena); flow through porous
media; particle fluid mechanics and transportation; and fluid-
particle separations. Some background is given by Ennis, Green,
and Davies."3' Two sets of notes have been completed by the
unique combination of industrialists and academicians: Par-
ticle Science (or Surface Engineering) and Flow Through Po-
rous Media. In 1996, the notes on Fluid-Particle Separations
should be available. For more details, contact S. Chiang(102a).
The University of Minnesota has and is preparing a series of
modules, distributed on MAC-based computers and eventually
on CD ROM. The modules developed so far are on forces (4
modules) and micelles (1 module). Complementing these will

be laboratories, demonstrations, and problems that will allow
one to use the material via problem-based or cooperative learn-
ing. The general overall themes of the modules are surfaces,
colloids, polymers, forces, fluids, and heat transfer. The mate-
rial is similar to a series of books being developed by Evans
and Davis (System Mechanics of Interfaces). The team devel-
oping this approach includes Karl Smith, an international au-
thority on cooperative and problem-based learning.

TEACHING APPROACHES
Although the lecture currently is the preferred instructional
style, there are some interesting and novel teaching approaches.
Three schools use initial interest surveys and modify the cur-
riculum to match the student's interests: Bike(73), Ploehn( 116),
and Ko(20c). Self-study is used by Slattery(128). Others bring
active learning into the classroom through cooperative learn-
ing activities, "guided-design," and in-class problem solving:
Chiang(102b), Shaeiwitz(149), and Woods(159b). Woods, for
example, shifted from lectures to the Osterman feedback lec-
ture system with a resulting increase in both student's marks in
the course and student ratings of the course. In this format, the
50-minute lecture is divided into two 20-minute mini-lectures
separated by a 10-minute cooperative calculation or discussion
activity. During this time, the instructor circulates through the
class to monitor how well the students have understood the

TABLE 3
Texts Used
Text (# of respondees using the text)
Own Notes (7) Hunter161 (1) Edwards, et al.[1'0 (2)
Hiemenz"l (4) Woodruff and Delchar71 (1) Shaw" I (1)
Russel, Saville, and Schowalter141 (3) Miller and Neogi181 (1) Adamson121 (1)
Evans and Wennerstrom[21 (2) Everettl91 (1) Walls1131 (1)
Israelachvili'51 (2) Slattery[ 4] (1)

TABLE 4
Videotapes
Topic Application For more...
Trefethan's "Surface Tension in
Fluid Mechanics".................................... Surface tension and Marangoni ........................................ Encyclopedia Britannica
Shell research .............................................. M aragoni behavior....................................... .................. W oods(159b)
Shell carburetor ........................................... Stability.......................................... -
Shell electrostatic explosions ...................... Electrostatic behavior ......................... ........................ Woods(159b)
Berg research ................................................. Marangoni roll cells; side and top views ............................. Berg(145b); Woods(159b)
Hickman research .......................................... Vapor recoil; Marangoni .................................................. Palmer(1 11); Woods(159b)
Brimacombe research ................................. Marangoni.......................................... ..... ................. Woods(159b)
W asan research ........................................... Coalescence ....................................... ............................ W asan(49)
.................................................................... Particle-particle impact on coalescence......................... Wasan(49)
....................................................................... Computer simulation of coalescence and separations ......... Wasan(49)
Woods research........................................... Coalescence ...................................... ....................... Woods(159b)
H artland ....................... ............................ C oalescence ....................................................... .... -

94 Chemical Engineering Eduction

material. If comprehension is lacking, the instructor can then
use the following 20-minute period to elaborate and correct
misconceptions. Pelton(159a) uses selected published papers
as the driving mechanism for student learning, while the ap-
proach used by Dold, et a/.(159d) is to identify a piece of
equipment to be designed, provide the students with a sample
of the feed, and ask them to measure the pertinent properties
and use the information to complete the equipment selection
and design. Nicholson(159c) uses two plant visits, laborato-
ries, and self-directed learning in his approach, and Wasan(49)
uses video-conferencing. Several approaches use an industrial
"process" for a focal point; Scheiner(3) uses the Bayer process;
and Dold, et al.,(159d) use an industrial waste-water treatment
process. The resources being produced by the team at the
University of Minnesota (Evans,75) will be of great assistance
to help us move to active, cooperative learning.

IDEAS ABOUT FUTURE TRENDS
Our hope is that surface phenomena will become a main-
stream, curricular requirement for all programs. In review-
ing the AIChE conference programming trends, we note that
in the 1960s, surface phenomena tended to have about three
sessions per conference. They were attended by researchers
dedicated to this specialized topic. At the 1995 Miami Beach
meeting, however, surface phenomena papers were presented
in about 30% of the conference sessions. Indeed, all physical
systems studied by engineers have surfaces and boundaries.
The more we learn about those surfaces, the better will be
our ability to predict what happens as material passes through,
reacts, or interacts with the surface. What still remains to be
done is to develop surface phenomena as a cohesive, core
fundamental subject for our undergraduate programs.
Of all the courses currently given, only Nicholson's(159c)
is required at the junior level. His course is characterized as
having plant visits, laboratory measurements, and practical
applications. Things that we might do to bring surface phe-
nomena into the mainstream of undergraduate chemical en-
gineering and to recruit students for our graduate programs
might include:
1. Using surface phenomena as the topic for communication
courses and projects.
2. Including surface phenomena projects and activities in the
laboratory program, as is done at Clarkson.
3. Developing a course on the practical engineering applica-
tions of surface phenomena or surface engineering and make
this required in the junior year. This will need the practical
applications flavor that is broad; it also needs data to allow
us to do practical problems.

SUMMARY
The responses to the survey (sent to about 180 chemical
engineering departments) reveal that about twenty schools
currently give at least one course in colloid and surface

phenomena; five schools offer two or more courses. The
courses tend to focus on the foundational theory; a few
courses include applications, and some teach surface phe-
nomena in courses on the environment, particle processing,
separations, and mineral processing.
Surface properties and their measurement is an important
theme for many respondents. Imaginative combinations of
laboratory courses and demonstrations enrich some of the
programs. The breadth of the subject is reflected in the many
different approaches taken in teaching it. The rich set of
practical applications of surface phenomena is illustrated
by the wide range of examination questions and prob-
lems assignments used.
There is no dominant and popular text. Most instructors
use their own set of notes (or textbooks that they have
written). A rich variety of films, videotapes, demonstrations,
and self-study tapes are available. A new development is the
computer modules being developed by the University of
Minnesota and the course notes prepared on "particle sci-
ence" by the American Filtration Society.
In methods of teaching the course, most use a lecture
format with active learning; cooperative learning approaches
are used in several schools.
Extensive cross-referencing has been used in presenting
the results so that those interested can follow up on some of
the many ideas used.

REFERENCES
1. Hiemenz, P.C., Principles of Colloid and Surface Chemistry, M.
Dekker, New York, NY (1977)
2. Evans, D.F., and H. Wennerstrom, The Colloidal Domain: Where
Physics, Chemistry, Biology, and Technology Meet, VCH Publishers,
New York NY (1994)
3. Ennis, B.J., J. Green, and R. Davies, "The Legacy of Neglect in the
U.S. Particle Technology," Chem. Eng. Prog., p 41, April (1994)
4. Russel, W.B., D.A. Saville, and W.R. Schowalter, Colloidal Disper-
sions, Cambridge University Press, New York, NY (1992)
5. Israelachvili, J.N., Intermolecular and Surface Forces, 2nd ed., Aca-
demic Press (1992)
6. Hunter, R.J., Introduction to Modern Colloid Science, Oxford Uni-
versity Press, Melbourne (1993)
7. Woodruff, D.P., and T.A. Delchar, Modern Techniques in Surface
Science, Cambridge University Press, New York, NY
8. Miller, C.A., and P. Neogi, Interfacial Phenomena, Marcel Dekker,
New York, NY (1985)
9. Everett, D.H., Basic Principles of Colloid Science, Royal Society of
Chemistry, London (1988)
10. Edwards, D.A., H. Brenner, and D.T. Wasan, Interfacial Transport
Processes and Rheology, Butterworth-Heineman, Boston, MA (1991)
11. Shaw, D.J., Introduction to Colloid and Surface Chemistry, 3rd ed.,
Butterworths, London (1980)
12. Adamson, A.W., Physical Chemistry of Surfaces, 5th ed., Wiley-
Interscience, New York, NY (1990)
13. Walls, J.M., Methods of Surface Analysis, Cambridge University
Press, New York, NY (1990)
14. Slattery, J.C., Interfacial Transport Phenomena, Springer Verlag,
New York, NY (1990)

Summer 1996

S APPENDIX

Summary of Responses
The number indicates the University as listed in the 1995 Index of Schools in the AIChE Faculty Directory
Key: J-Junior Required; S-Senior course; SL-Senior Lab; SE-Senior Elective; G-Graduate; AG-Advanced Graduate

3. Alabama, B.J. Scheiner, Hydrometallurgy (S,G) Text: own notes, half on hydrometallurgy and half on surface phenomena.
Integrates the ideas around the Bayer process. Style: lecture, question, pass out activities to get students to think about why things
happen.
14. California, Berkeley, C.J. Radke, Applied Surface and Colloid Chemistry (G) Text: own notes, capillary hydrostatics and
dynamics, capillary thermodynamics, colloids and electrical phenomena. Style: lecture
19. California, Santa Barbara, J. Israelachvili, Colloids and Surfaces Text: Israelachvili
20a. Carnegie Mellon University, A.N. Jacobson, Experimental Colloid and Surface Science (SE,G) Text: Hiemenz; laboratory
complements lecture course with experimental techniques. Complete 9 out of 12 experiments plus demonstrations in film balance
and SEM. Style: laboratory
20b. Carnegie Mellon University, P. DiMilla, Physical Chemistry of Colloids and Surfaces
20c. Carnegie Mellon University, E. Ko, Surfaces and Adhesion (G) Text: Woodruff and Delchar; gas-solid interactions with half on
principles and half on experimental techniques. Style: surveys students ahead of time and then sets the course content; replaced
exams with review paper or a research proposal with oral presentations.
20d. Carnegie Mellon University, J. Anderson, Physical Chemistry of Macromolecules (SE,G) Text: Young and Lovell; general
concepts, chemical synthesis of polymers, polymers in solution and bulk polymers. Style: lecture
20e. Carnegie Mellon University, D. Prieve, Colloid Science (G) Text: Russel, et al.; light and its application to colloids, Brownian
motion, diffusion in a force field (sedimentation), flocculation, electrostatics, double-layer forces, electrodynamics of continue,
slow Brownian flocculation, electrokinetic phenomena. Style: lecture
24. Clarkson, R. Partch (SL) Lab: preparation and characterization of aerosols
38 Drexel, J. Tallmadge, Interfacial Phenomena (SE) Text: own notes; half on fundamentals of basic phenomena, half on coating and
two-phase flow. Style: lecture
45a. Houston, F. Tiller, Theory and Practice of Solid-Liquid Separation (SE,G) Text: own notes on the theory and practice of solid/
liquid separation, particle characterization, flocculation, slurry properties, cake formation. Style: lecture
49. Illinois Institute of Technology, D.T. Wasan, Interfacial and Colloidal Phenomena with Applications (SE,G) Text: Edwards, et al.;
surface tension, contact angles, adhesion, wetting and spreading; adsorption and micellization; surface rheology, colloid stability;
thin liquid films and emulsions and foams; rheology of dispersions and electrophoresis and electrokinetic phenomena. Style: video-
conferencing with demonstrations, videotapes and three or four labs.
73. U. Michigan, S. Bike, Colloids and Surfaces (G) Text: Hunter; thermodynamics of surfaces, preparation and characterization of
colloids, electrochemical double layer, van der Waals forces, DLVO, polymeric stabilization and flocculation, transport, associa-
tion colloids, applications; Style: interest survey, class presentations, critical review of articles and emphasis on applications
75a. Minnesota, F. Evans, Colloidal Domain (G) Text: Evans and Wennerstrom; solutes and solvents, monolayers, double layer,
micelles, forces in colloidal systems, bilayers, polymers, colloidal stability, colloidal sols, phase equilibria, macro and microemulsions.
Style: lecture, unique computer modules
75b. Minnesota, F. Evans, Fundamentals of Surface Phenomena (G) Text: Evans and Wennerstrom
102a. Pittsburgh, P.P. Phule, Principles of Surfaces and Colloids (SE,G) Text: own notes; particulate surface and interfacial area; surface
tensions, energy; wetting, adhesion, adsorption, gas-solid, liquid solid; forces between particles and DLVO; processing fine
particles/emulsions; polycrystalline materials; experimental techniques of surface analysis Style: lecture with interdisciplinary
focus that attracts materials science, chemical engineering, chemistry, physics, and pharmacy students; take-home exam
102b.Pittsburgh, S. Chiang, Fluid Particle Processing and Separation (SE,G) Text: own notes; about a third on particle characterization
and surface phenomena Style: lecture plus cooperative learning plus project
105a. Princeton, D.A. Saville, Colloidal Dispersions I (G) Text: Russel, et al.; experimental foundations and theory Style: lecture with
tutorials
105b.Princeton, W. Russel, Colloidal Dispersions II (G) Text: Russel, et al.; experimental foundations and theory Style: lecture with
tutorials
110. Rice, C. Miller, Interfacial Phenomena (G) Text: Miller and Neogi; half on fundamentals of interfacial tension, contact angles and
surfactants together, and half on flow and transport at interfaces with a little on colloidal stability Style: lecture, term paper
116. South Carolina, H.J. Ploehn Colloids and Interfaces (SE,G) Text: Israelachvili and Everett; historical perspectives, interfacial
thermodynamics, capillarity and wetting, adsorption and monolayers, surface, micelles and self-assembly; intermolecular forces;
colloidal stability; Brownian motion; radiation scattering techniques; transport phenomena; phase behavior of concentrated
systems Style: interest survey, lecture with projects, and oral presentations

196 Chemical Engineering Eduction

128. Texas A&M, J.C. Slattery, Advanced Interfacial Phenomena (AG) Text: Slattery Style: self-study
134. Tufts, G.D. Botsaris, Surface and Colloid Chemistry (SE,G) Text: Shaw plus notes; I Fundamentals attractive and repulsive forces
between particles, electrokinetics, stability and flocculation, surfactants, micellization and adsorption, wetting, curved interfaces,
nucleation, capillarity and surface tension gradients; II Applications emulsions, concentrated suspensions and slurries, separation
processes, drying of coatings, and foams Style: lecture plus series of fascinating practical-case problems; Overbeek's videotapes
available
139 Vanderbilt, T. Fort, Surfaces and Adsorption (SE,G) Text: own notes; adsorption, wetting, detergency, flow through porous media.
Style: lecture enriched by films and slides from past research; demonstrations of experimental methods of measuring surface
tension and contact angle, the film balance, and techniques for making Langmuir Blodgett films, spreading from liquid lenses and
solid crystals and wave damping effects of spread films; videotapes
143. Washington State, R. Zollars, Interfacial Phenomena (SE,G) Text: Hiemenz; emphasis on molecular basis for interfacial forces and
the macroscopic phenomena that result and on the latest analytical techniques (QELS, Proton correlation spectroscopy; field flow
fractionation; STM and AFM); basic concepts and measurements; molar mass; sedimentation and diffusion; solution thermody-
namics; viscosity and light scattering; interfacial phenomena; surface tension; adsorption from solution; adsorption by a solid
surface; surfactant structures; colloidal phenomena; flocculation, electrostatic and electrokinetic behavior Style: lecture with
significant laboratory group work
144. Washington U., C. Thies, Principles of Surface and Colloid Chemistry (S) Text: Hiemenz; nomenclature, powder technology,
sedimentation and diffusion equilibrium; viscosity; osmometry; light scattering; surface tension; porosimetry; adsorption from
solution; adsorption at gas-solid surfaces and surface area determinations; electrical double layer and flocculation phenomena
Style: lecture
145a.U. Washington, B.D. Ratner, Surface Analysis (SE,G) Text Walls; practical course on how to measure the nature of solid surfaces
with emphasis on ESCA and how to interpret and quantify data Style: lecture plus real-world data that the students analyze (a
ladybug's wing in 1995); surfaces, energy interactions with matter, vacuum systems, ESCA, SIMS, contact angles, auger
apectroscopy, scanning, tunnelling microscopy, SEM, TEM, EDXA, vibrational spectroscopies (IR, SERS, IETS, EELS), applica-
tions in biomedical and microelectronics Style: lecture plus student projects and visit to instrumental lab
145b.U. Washington, J.C. Berg, Surface and Colloid Science Laboratory (SE,G) Text own notes "Surface and Colloid Science";
capillarity, capillary hydrostatics; solid-liquid interactions; interfacial thermodynamics (adsorption, self-assembly); colloids;
electrical properties of interfaces (double layers, DLVO, kinetics of aggregation, electrokinetics) capillary hydrodynamics (Marangoni
effects, Gibbs elasticity) Style: lectures, small demonstrations; laboratories with self-complete handouts; over twenty experiments
available with each student (working in pairs) doing four experiments; videotapes
145c.U. Washington, B. Rogers, Surface Science
149. West Virginia, J.A. Shaeiwitz, Interfacial Phenomena (SE,G) Text Hiemenz; intermolecular and interparticle forces, interfacial
tension, wetting, adsorption, colloids and sedimentation, sedimentation versus diffusion, colloid thermodynamics; viscosity of
suspensions; charged interfaces, double layers, DLVO, coagulation kinetics; stabilization and flocculation by polymers; electroki-
netic phenomena; application to particle pollution control; surfactants; micellization; emulsions, microemulsions; detergency;
surfactant adsorption and applications; surfactant-based separations Style: lecture plus active learning plus project; emphasis on
problem solving; applications
159a. McMaster, R. Pelton, Polymer Colloids (G) Text: Hunter, and Evans and Wennerstrom; colloid stability; colloid (latex) character-
ization; surface chemistry (surface tension, thermodynamics of interfaces and capillarity); surfactants (characterization and
properties) Style: assigns published papers as the mechanism for learning; students orally present summaries of findings
159b.McMaster, D. Woods, Colloids, Surfaces, and Unit Operations (SE,G) Text: own notes; when is surface phenomena important
(particle characterization, thin films and surfactants), surface tension with two surfaces, interactions of three surfaces, variation in
surface tension with temperature, pressure and concentration; attractive forces between surfaces, adsorption, adsorption of ions;
implications for two surfaces DLVO and rate; adsorption of polymers Style: problem-based with Osterman feedback lecture with
in-class problem solving; applications oriented; demonstrations, videotapes
159c.McMaster, P. Nicholson, Materials Processing I(J) Text: Adamson plus own notes; introduces powders and powder-liquid systems
and applies fundamentals to mineral processing and slip synthesis; comminution, grinding theory, and methods of powder
synthesis; particle statistics, measurement of particle size and surface area; mixing and packing of particles; surface chemistry of
suspensions; flocculation, deflocculation and ion-exchange; oxide structure and surface charge; clays, ion-exchange, suspension
stability, dilatancy, thixotropy and EDP; mineral flotation and elutriation, and process mass balances Style: cooperative self-
directed learning and active learning with lectures; two plant trips, experimental laboratories
159d.McMaster, P. Dold, A. Robertson, D. Woods Environmental Laboratories (G) Text: own notes; student teams do five experiments
to provide data to size/design water or waste water treatment facility; Topics flow measurement, coagulation/flocculation,
activated sludge, rotary vacuum filtration (surface filtration) and deep bed filtration Style: mini-lecture introduction; samples
supplied and students learn theory on a need-to-know basis; run experiments, interpret data, and size equipment 0

Summer 1996 197

r, curriculum

INTEGRATING NEW

SEPARATIONS TECHNOLOGIES INTO

THE UNDERGRADUATE CURRICULUM

PAMELA M. BROWN
Stevens Institute of Technology Hoboken, NJ 07030

Chemical engineering educators strive to prepare their
students for a professional career that may well ex-
tend forty years into the future. One way to meet this
demand is to introduce emerging separations technologies
into the undergraduate curriculum. This increases the value
of the students' undergraduate education since they acquire
knowledge in subjects that practicing engineers may not be
familiar with and which may become important during their
professional lives.
One method of introducing new separations technologies
is to develop problems using processes developed at the U.S.
Bureau of Mines. Its Office of Technology Transfer pub-
lishes information on processes that have been developed on
a laboratory scale and that are available for licensing. This
information is in the public domain, and enough data is
provided to perform scaleup calculations. Three problems
have been developed using this approach. The first, "Pilot
Plant to Leach Platinum from Catalytic Converters," was
presented in this journal."l In it, Joe Agman, Jr., owns a
chemical plant that recovers silver from used photographic
material. He is interested in diversifying and hires a student
to design and test a pilot plant to learn more about leaching
platinum from used catalytic converters. The problem was
first assigned in a reactor design course at Stevens Institute
of Technology several months before it was announced that
the process had been licensed and commercialized.[2'3' The
students in the class were proud to know that they had

IPamela Brown is a Visiting Assistant Professor
of Chemical Engineering at Stevens Institute of
Technology. Her research interests include sepa-
n i rations and crystallization.
Copyright ChE Division ofASEE 1996

tackled a problem practicing engineers were working on.
Two additional problems and their solutions are being
presented in this paper. The problems are presented in a
personal format. Additional information of industrial signifi-
cance is included in the problems to create an interesting and
believable scenario.

Problem 1
Plutonium Recovery from Wastewater Using
Metalloprotein Affinity Metal Chromatography"hJ

This problem requires mass balances for scaleup of a
process to recover plutonium from wastewater generated at
a former nuclear weapons production facility, using a state-
of-the-art chromatography technique. It can be assigned in
an introductory chemical engineering or separations course.

Imagine you like to live dangerously. You enjoy sky div-
ing and driving race cars. As a student, you always waited
until the night before exams to start studying. You have just
accepted employment as a site remediation engineer at the
560-square-mile Hanford nuclear site located in south-cen-
tral Washington. Plutonium for nuclear weapons was pro-
duced here from 1943 to 1987, resulting in 1100 waste sites.
Highly radioactive waste was stored in tanks, but between
1946 and 1966, low-radiation-level liquid waste was inten-
tionally discharged to the soil. As a result, there is a 150-
square-mile plume of hazardous chemicals and radionuclides,
and billions of cubic meters of contaminated soil. More than
60 million gallons of highly radioactive waste have accumu-
lated in 177 tanks. Sixty-eight of the single-shell tanks have
or are suspected of leaking (double-shell tanks were used
starting in 1968). Little documentation is available.[4' While
most people would be nervous about working in this envi-
ronment, you are pleased by the apparent job security-

Chemical Engineering Eduction

One method of introducing new separations technologies is to develop problems using
processes developed at the U.S. Bureau of Mines. It... publishes information on processes that
have been developed on a laboratory scale and that are available for licensing. This information is in
the public domain, and enough data is provided to perform scaleup calculations.

there is enough radioactive waste to last your whole career!
Your first assignment is to design a pilot plant to study the
feasibility of recovering plutonium from aqueous waste
streams. You will be scaling up a process, metalloprotein
affinity metal chromatography, developed at the University
of Alabama and funded by the U.S. Bureau of Mines.
Background Affinity chromatography is a separation
technique where a solution passes through a packed bed
filled with a porous stationary solid. The material to be
separated is adsorbed (attached) to the solid, while the re-
mainder of the solution passes through the column. To re-
cover the material, solvent conditions are altered so that the
separated material desorbs from the solid. Typical solid sup-
ports are characterized by large surface areas and include
silica, alumina, polymers, and carbohydrates such as cellu-
lose and Sepharose. The adsorption properties of all these
solid supports can be modified by bonding different mol-
ecules, or ligands, to their surface. Many researchers in this
area use Sepharose because it is commercially available and
has a successful history."5'
Metalloproteins are biological molecules that selectively
and stoichiometrically bind to metal ions under certain con-
ditions. This selectivity is the result of millions of years of
biological engineering due to evolution. Changes in pH, salt
concentration, etc., can cause the metal ions to be released.
The metalloprotein transferring is involved in Fe" transport
in living organisms. It is found in blood serum, milk, and
eggs. Transferrin preferentially binds to the ferric ion, but
will also bind to Cry, Cu2, Mn2+, Co3', Cd2", Zn2', Ni2+,
numerous trivalent lanthanides (including holmium), Th",
and Pu4+ (tetravalent plutonium ion, form of plutonium found
in contaminated water). Transferrin is available commer-
cially as conalbumin--egg white transferrin.67'1
In this problem, the transferring is covalently bonded to the
porous solid support Sepharose. The solid is loaded into a
column, and a solution containing plutonium ions passes
through the column. The plutonium ions are preferentially
and stoichiometrically adsorbed by the transferring, which
has been immobilized onto the solid. A flow diagram of the
adsorption step is presented in Figure la. To recover the
plutonium from the column, a solution with a low pH is
passed through the column, causing the plutonium to desorb.
A flow diagram of the desorption step is presented in Figure
lb. The plutonium is thus removed from the initial solution
and concentrated using metalloprotein affinity metal chro-
matography. The column can be repeatedly reused.
Summer 1996

Procedure The metalloprotein transferring was first
immobilized to the solid support, CNBr-activated Sepharose
B, purchased from Pharmacia Biotech Inc. The
manufacturer's recommended procedure was followed.,6,7'
Specifically, 2.0 gm of transferring was immobilized on 15
gm of CNBr-activated Sepharose B.
In laboratory-scale feasibility experiments, holmium was
recovered from solutions rather than plutonium, for safety
considerations. For scaleup, assume 1 mole of plutonium is
adsorbed for every 1 mole of holmium salt adsorbed. It was
found that this column could adsorb the holmium found in
50 mL of a solution into which 45.3 gm Ho(NO3)35H20 was
dissolved.
A buffered solution at a lower pH (pH=4) was used to

feed solution
containing plutonium

Figure 1
(a) Flow diagram for adsorption of plutonium
(b) Flow diagram for desorption of plutonium using low
pH solution.

Packed Column -
transferrin covalently bonded
to solid Sepharose support

) plutonium free solution
concentrated plutonium solution

Packed Column -
transferrin covalently bonded
to solid Sepharose support

desorb and recover the holmium. All the holmium was des-
orbed into 0.8 mL of this solvent.

Assignment You will be designing a pilot plant to
concentrate the plutonium solutions found in the single-shell
tanks. The volume of solution in these tanks is 530,000 to
1,000,000 gallons. You wish to process 40-gallon batches of
the solution. Assume the feed concentration is 16.3 gm239Pu4+/
L. A schematic diagram of the system is presented in Figure 2.

How many grams of transferring should be immobi-
lized onto the solid support?
How many grams of solid support CNBr-activated
Sepharose B will you need?
) Suppose 1 gm of 239Pu desorbs into 0.5 mL of low-pH
solvent. For a tank initially containing 1,000,000
gallons, estimate the final volume. The density of
plutonium is 17.14 gm/cm3.

Solution

0 Since plutonium and holmium both bond stoichio-
metrically to the transferring, the grams of transferring need is

(2 gm transferrin)(440) (16.3 gm Pu 3.783 L 40gal)
(45.3gmHo(N03)3 5H20)(239)Y L gat
= 201 gm transferring (1)

Note the molecular weight of the holmium salt is 440 and the
atomic mass of the plutonium ion is 239.

* The mass of CNBr-activated Sepharose needed is

S--gm rse (201 gm transferring = 1504 gm (2)
S2 gm transferring )

O The concentrated volume of a tank that was initially
1,000,000 gallons will be the volume due to the solution
and the volume of the plutonium, assuming perfect solu-

low pH solution

Figure 2. Schematic of adsorption column.

tion behavior:

0.5mL
gm Pu )

SmL )( L )(16.3gmPu3.785L)/ 6
17.14gmPu 1000 mL j L gal)1

= 39,850 gallons (3)

A substantial reduction in volume is achieved.

Problem 2
Design of a Novel Froth Flotation System
for Coal Purificatione'7

The second problem requires some of the calculations
necessary to scale up a process to recover coal fines from an
aqueous slurry. It can be assigned in an introductory chemi-
cal engineering, senior design, or separations course. Indus-
trial applications offroth flotation include the initial concen-
tration of copper, lead, zinc, molybdenum, phosphate, pot-
ash, nickel, fine coal, and other mineral commodities from
ores, de-inking of paper for recycling, and wastewater treat-
ment.12"' Froth flotation is a separation technique that is
mentioned in commonly used undergraduate chemical engi-
neering texts,'1"14' but is not usually studied in detail.

You are employed at a mid-sized company involved in
coal cleaning. In your area of the plant, coal is recovered
from an aqueous slurry containing finely sized solids. These
solids are 82 wt.%coal, 18 wt.%ash. The coal is purified and

Rotating Dish

Figure 3. Agitair flotation system'"

Chemical Engineering Eduction

concentrated plutonium solution

Packed Column -
transferrin covalently bonded
to solid Sepharose support

plutonium free solution

Froth Overflow -

recovered by froth flotation. Froth flotation is a separation
technique for separating solids or oils in an aqueous solution
based on differences in hydrophobicity. In this process, the
feed stream is agitated and air is bubbled through the solu-
tion. A froth forms that rises to the surface. The hydrophobic
coal particles tend to collect in the froth (bubble-to-particle
attachment), while the more hydrophilic ash tends to stay in
solution. The froth is then separated from the solution. A
typical flotation unit is the Agitair flotation machine shown
in Figure 3. Air is supplied through the pipe in the center of
the vessel and is dispersed through the rotating disk at the
bottom. The froth overflows at the top of the vessel.
One disadvantage of froth flotation is that it tends to be a
slow process, with the bubble-to-particle attachment the rate
limiting step. One way to increase the rate of bubble-to-

SValuable Mineral
o Waste Rock
0 Bubble

Figure 4. Schematic diagram of the rapid flotation system."2'

Figure 5. Schematic diagram of a three-stage continuous rapid
flotation system.112'

Summer 1996

particle attachment is to increase the agitation, but this can
dislodge the coal from the bubbles, reducing the yield. Be-
cause the bubble-to-particle attachment and the bubbles ris-
ing to the surface through the solution (called pulp) occur in
a single unit, optimizing one step tends to hurt the other.
The plant owner would like to increase capacity, but space
limitations present a problem. It is for this reason that you
have been assigned the task of designing a froth flotation
pilot plant capable of processing 50 gpm of liquid feed (not
including air) to study a potentially more efficient system for
froth flotation. Background information is given below.
Novel Froth Flotation System The U.S. Bureau of
Mines has recently developed a rapid froth flotation system
that separates the flotation into two discrete units."2'1517 This
system allows optimization of both the bubble-to-particle
attachment and the bubble-solution separa-
tion. It has been shown to be eighteen times
o_ t faster than conventional froth flotation on a
Q laboratory scale.

A schematic of the process in shown in
Figure 4. An in-line mixer is used for bubble-
to-particle attachment. An in-line (or mo-
tionless) mixer is a generic term for a mixer
with no moving parts. It consists of a pipe
containing baffles that cause turbulence and
hence mixing when liquids or gases pass
through them. Downstream of the in-line
mixer is a shallow-depth separator. In this
unit the bubbles with the coal fines attached
rise to the surface and froth forms. After the
froth overflows, it is collected and the puri-
fied coal is recovered. This shallow-depth
separator allows the bubbles to rise and the
froth to overflow.
Process Description"2 A slurry from a
coal cleaning operation and a bubble slurry
(bubbles plus water) each enter the in-line
mixer through opposite ends of a T-shaped
fitting. Before mixing, the coal slurry is con-
ditioned with 1 gm fuel oil per kg ore, for
three minutes. The bubble slurry is gener-
ated by mixing air and water in a conven-
tional flotation cell. The water is pretreated
with 0.1 gm frothing agent, methyl isobutyl
carbinol (MIBC), per kg ore. As mentioned
previously, the resulting slurry is 5 wt.%
solids, of which 82 wt.% is coal. The pH is
adjusted to 8.2. The solution then enters
three flotation units in series. The froth con-
taining the coal concentrate is collected, and
the tailings are sent to waste.
Scale-Up Data"5' Experiments per-
201

Slurry O %

Bubble-Particle Attachment
Unit

Ore Feed .. concentrate
Bubble Feed

tailings

-Ore Feed age
Bubble Feed

tailings

Bubble Feed -)

formed on a laboratory scale (5-29 L/min) using a three-
stage flotation unit (see Figure 5) recovered 93 wt.% of the
coal. At optimum conditions, the recovered product was
91.9 wt.% coal, 8.1 wt.% ash.'91 Scaleup was found to be
dependent on three parameters. These parameters and their
optimal values are

1. Mixing intensity of the in-line mixture (4.9 watts per 1 L
liquid feed/min)
2. Air to solids ratio (1.5 mL air (STP)/gm ore)
3. Bubble residence time; assume the surface area in all
three flotation units is equal and the depth is 5.4 cm. The
rate of flotation follows first-order kinetics. The percent-
age of recovery is given by
% Recovery = [1 exp(-kt,)] (100%) (4)
where k = 4.13 min-' is the first order flotation rate
constant, and t, is the total residence time of a particle in
the flotation system (sum of all three stages). The resi-
dence time is determined by dividing the total volume of
the flotation system by the flowrate.

Assignment You are to design a pilot plant to study froth
flotation using the process developed at the U.S. Bureau of
Mines. The flowrate of feed is to be 50 gpm, and you are to
recover 93 wt.% of the coal in the feed. The purity of the
recovered solids is expected to be 91.9 wt.%. In order to
accomplish this, please complete the following tasks:

Mass Balances
Q) Calculate the optimum flowrate of air (STP) and the
overall total flowrate (air plus solution). Assume the
specific gravity of the ore is 1.6.
Calculate the expected yield of coal per 100 gal. of feed
and the expected yield of solids.

In-Line Mixer Design
Estimate the pressure drop and the power requirements
in the in-line mixers when the intensity of mixing is 4.9
watts per 1 L liquid feed/min of feed and the flowrate is
50 gpm of liquid feed.

Shallow-Depth Separator Design
( Calculate the bubble residence time, t,.
) Calculate the total volume required for the shallow-
depth separators.
G) Calculate the total surface area required for the shal-
low-depth separators.

Solution

* The optimum flowrate of air is 1.5 mL(STP)/gm. ore.
The flowrate of liquid feed is 50 gpm and it is 5 wt.% ore.

First, the mass of ore per volume of feed must be found.
Basis: 100 gm feed, containing 5 gm of ore.
The densities of water and ore are 1.0 and 1.6 gm/mL,
respectively. The volume of water, VH20, and ore, Vore, may
thus be taken as

VH, =(95 gm){ mL =95 mL

Vore =(5 gm) r = 3.124 mL
(1.6 gm)

The mass of ore per volume liquid feed is thus
5 gm 0.051 gm ore=51gm/L (6)
95+3.125 mL mL
The optimum volumetric flowrate of air is thus

(50 gpm 3.785 L (51 gm ore) 1.5 mL air(SPT) 1L
0o gpm) g-- gm -
gal l L ) gm ore I1000 mL,

=14.5 Lpm (STP)air (7)

The overall total volumetric flowrate, Qtotal, is the sum of
the flowrates of the coal slurry and bubble slurry:

Qtotal = (50gpm) 3.785L +14.5 Lpm= 203 Lpm (8)
Sgal )
This calculation neglects any changes in the volume of the
gas due to pressure.

* The expected yield of coal is 93 wt.% of the coal in the
feed. The feed ore is 82 wt.% coal. The expected yield of
coal is thus

(50 gpm) 3.785 L)51 gm ore)(0.82)(0.93) 1lb
gal L 454 gm
16.2 lb coal (9)
min
The solid product is 91.9 wt.% coal. The total weight of
recovered solids is thus
(16.2 b/min
1 17.6 lb / min (10)

0 Since the pressure drop in the in-line mixer is the power
supplied divided by the flowrate, determination of the in-
line mixer simply involves manipulating the units of the

Chemical Engineering Eduction

given mixing intensity:

(4.9 W) ( Nm ( 14.7 psi (60 sec( 1000 L
(lLfeed/min)sec-w 1.01325x105N/m2r min ) m3 )

=42.7 psi (11)

The power supplied by the mixer is the product of the power
per volume times the volumetric flowrate. The power sup-
plied is thus

(4.9 W) (50 gpm liquid feed) 1 = 927 W (12)
(IL feed /min) ( gal )

OTo calculate the bubble residence time, use the flotation
rate constant and the required recovery of 93%. Rearranging
Eq. (4) to solve for t, gives

[tn (100)/(100 % recovery)] [in (100)/(100-93)]
tr k-- -- 0.644 mmin
Sk 4.13
(13)

0 The total volume required in the shallow-depth separa-
tors is equal to the product of the overall total volumetric
flowrate and the bubble residence time:

System =Qtotal tr = (203 Lpm)(0.644 min)= 131 L (14)

O Since the depth of the separators, h,
required surface area, A, is

(131L)(1000cm3 )( m )2
V (131L L 100cm)
h 5sysem4 L
h 5.4cm

is 5.4 cm, the

=2.4m2 (15)

CONCLUSION
Two problems for the undergraduate curriculum, using
new separations techniques, have been presented using
processes developed at the U.S. Bureau of Mines. Al-
though the U.S. Bureau of Mines was closed last Febru-
ary due to budget cutbacks, other U.S. government agen-
cies are actively involved in developing new technolo-
gies for commercial development, and data is in the pub-
lic domain. This is a rich source of information for devel-
oping state-of-the-art problems. The other agencies in-
clude the U.S. Department of Agriculture and the Envi-
ronmental Protection Agency.

Summer 1996

To obtain copies of government-owned patents and
patent applications, call (202) 260-7510.

ACKNOWLEDGMENTS
I would like to acknowledge Jay Panditaratne, BE, for his
assistance in preparing some of the figures in this article.

REFERENCES

1. Brown, P.M., "Design of a Pilot Plant to Leach Platinum
from Catalytic Converters," Chem. Eng. Ed., 28(4), 266
(1994)
2. Rosenzweig, M.D., "Update," Chem. Eng. Prog., p. 14,
Dec (1994)
3. "AIChE Extra," supplement to Chem. Eng. Prog., p. 8,
April (1995)
4. Campbell, J.A., et al., "Organic Analysis at the Hanford
Nuclear Site," Analytical Chem., 66(24), 1208A (1994)
5. Beitle, Robt. R., Jr., Asst. Prof., Dept of Chemical Engi-
neering, University of Arkansas, 3202 Bell Engineering
Center, Fayetteville, AR; personal communication
6. Donald, S., K. Spires, and J. Vincent, "Potential for
Decontamination of Plutonium-Containing Solutions
Using Transferrin Metalloprotein Affinity Chromatog-
raphy," J. of Inorg. Biochem., 21 (1994)
7. Spires, K., and J.B. Vincent, "Transferrin Metalloprotein
Affinity Metal Chromatography," J. Chem. Tech. Biotech.,
62, 373 (1995)
8. Perry, R.H., and D.W. Green, Eds, Perry's Chemical En-
gineers' Handbook, 6th ed., McGraw-Hill, New York, NY
(1984)
9. Austin G.T., Shreve's Chemical Process Industries, 5th
ed., McGraw-Hill, New York, NY (1984)
10. Levin, M.A., and M.A. Gealt, Biotreatment of Industrial
and Hazardous Waste, McGraw-Hill, New York, NY
(1993)
11. Wentz, C.A., Hazardous Waste Management, McGraw-
Hill, New York, NY (1989)
12. Susko, F.J., and C.E. Jordan, "Dilution Considerations for
the Rapid Flotation of Coal," in Advances in Filtration
and Separation Technology, Vol. 6, American Filtration
Society, 462 (1992)
13. King, C.J., Separation Processes, 2nd ed., McGraw-Hill,
New York, NY (1980)
14. Felder, R.M., and R.W. Rousseau, Elementary Principles
of Chemical Processes, 2nd ed., John Wiley & Sons, New
York, NY (1986)
15. Jordan, C.E., and F.J. Susko, "Rapid Flotation Using a
Modified Bubble-Injected Hydrocyclone and a Shallow-
Depth Froth Separator for Improved Flotation Kinet-
ics," Minerals Eng., 5(10-12), 1239 (1992)
16. Susko, F.J., and C.E. Jordan, "Modeling the Rapid Flo-
tation of Coal," Society for Mining, Metallurgy, and Ex-
ploration, Inc., AIME preprint, 93 (1993)
17. Hood, G.D., and C.E. Jordan, "In-Line Static Mixer Rapid
Flotation System for Improved Flotation Kinetics," Min.
& Metallur. Proc., 10(4), 182 (1993)
18. Ross Engineering, Inc., 32 Westgate Blvd., Savannah,
Georgia 31405-1475 0

S SOclassroom

IMPLEMENTATION OF MULTIPLE

INTERRELATED PROJECTS WITHIN A

SENIOR DESIGN COURSE

JOHN T. BELL
University of Michigan Ann Arbor, MI 48109-2136

apstone design courses typically involve many groups
of students working on identical design projects.
This approach leads to fierce competitiveness for
limited resources such as library materials, computer re-
sources, instructor feedback, and innovative ideas. At the
same time, employers are looking for "team players" who
can work cooperatively with other employees for the overall
good of the company. This standard approach to process
design instruction also yields a large number of similar re-
ports, which can be tedious to evaluate. Another difficulty
encountered in many capstone design courses is the wide
variety of (ABET required) topics covered, which leaves
many students wondering how they are all related and what
relevance each has to the overall design process.
This semester, a novel approach was investigated wherein
each design group was assigned the study of a different
production process within the petrochemical industry. The
projects were interrelated through feeds and products, just as
different production facilities are interconnected within a
large chemical processing complex. Students completed
midterm reports that analyzed different aspects of their
process and produced a final report that encompassed
their full semester's work.
The use of different projects for each group greatly re-
duced the competitive demand for limited resources and

John T. Bell teaches chemical engineering at
the University of Michigan, where he is also con-
ducting research into the applicability of virtual
reality to chemical engineering. His chemical
engineering degrees include a BS from Georgia
Tech, a MS and PhD from the University of
Wisconsin-Madison, and a DEA from institutee
du Genie Chimique in Toulouse, France. He
aalso holds a MS in computer science from UW-
Madison. His official home page is http://
www.engin.umich.edu/dept/cheme/bell.html.
Copyright ChE Division ofASEE 1996

provided the instructional staff with a more interesting vari-
ety of reports to evaluate. The design projects also served to
tie together the different course topics by serving as a focal
point upon which to apply each major topic as it was cov-
ered. The relationships between projects caused students to
take interest in other groups' work, and in some cases inter-
group cooperation was achieved.

THE COURSE
The course in which this procedure was developed is the
first semester of a two-semester senior plant-design sequence.
Due to a number of scheduling restrictions, many students
are allowed into the course without having completed their
courses in separations, heat and mass transfer, or reactor
design. This course also suffers from the common practice
of putting all ABET requirements that do not fit anywhere
else in the curriculum into the capstone design sequence.'"
As a result, the course delivers a wide variety of design
related material to students of varying backgrounds. Some
of the major topics covered in the first-semester course in-
clude ethics, safety, economics, metallic crystal structures,
phase diagrams, materials of construction, pressure vessel
codes, and environmental issues, all considered from the
point of view of the design engineer. Students apply these
topics to the development of original designs in the second
semester of the sequence, which is normally taken during
their final semester.
A major complaint that students have expressed about
this course in past years is that it is a collection of
miscellaneous topics having little apparent relationship
to each other or to the semester design project. Another
problem with previous years' projects has been that stu-
dents tend to wait until the last two weeks of the semester
to begin working on them, leading to sleep deprivation
and strained nerves as 150 students descend upon the

Chemical Engineering Eduction

finite resources of the engineering library and computing
center just before the project deadline.

THE PROJECT
Two major goals of this year's design project were to
provide a central focal point that would tie
together the myriad topics covered in the
course and to provide a vehicle for students ... each
to apply the material covered in class to dem- was assig
onstrate mastery of important concepts as each of a
topic is completed. Another goal was to focus product
heavily on the analysis level of Bloom's tax- wit
onomy of educational objectives."' petrochen
During the first week of class, students were The pri
assigned to groups and each group was as- interrelc
signed an industrially important chemical that feeds and
would serve as their focal point for the se- as differ
mester. Their first assignment was to conduct facil
a thorough literature search to gather the in- interconn
formation and background knowledge that large
they would need during the rest of the semes- process
ter. Later, as each major course topic was Student
completed, students handed in midterm re- midterm
ports that analyzed their process from the analyze
point of view of the topic just completed. aspects oJ
A final comprehensive report at the end of and proc
the semester was naturally commenced by rep
compiling the five midterm reports into encompa,
five sections of a large complete report. semesl
The benefit of this approach is that it forced
the students to work on their project continuously all
semester, and by the end of the semester their projects
were 80-90% completed.

The Chemical Processes
The chemical processes assigned to the students were not
chosen randomly; they were chosen so that every group's
production process would be related to at least one other
group's process through common feeds and products. The
basis for these interrelated groups was a series of charts in
Chemical Origins and Markets1'3 showing the production
relationships between key products of the petrochemical
industry, and the PhD thesis work of the course instructor.4'
Forty groups were subdivided into sections based on deriva-
tives of ethylene, propylene, n-butane, butylene, and ben-
zene as shown in Figure 1 (next page). The chemicals as-
signed to the students are shown in bold face, with the group
number given in parentheses in one location of the chart for
each assigned chemical. The multiple instances of several
chemicals in Figure 1 illustrate the variety of production meth-
ods available for most chemicals. The unassigned chemicals
show students where their chemical fits within the petrochemi-
cal industry and in relation to the other students' projects.
Summer 1996

des
ned
liff
ion
hin
ica
ojec
ited
prc
nt
itie
ecte
che
ng
s co
rel
d d
'tht
1uc
ort
sse
ter'

Midterm Reports
The students were asked to complete five midterm re-
ports regarding their assigned chemical's production pro-
cess, covering aspects of background, economics, mate-
rials of construction, safety, and environmental concerns
as described below.
Background The first midterm assign-
ign group ment, dealing with background information,
the study was designed to send students into the li-
erent brary to find as much information as pos-
process sible concerning the production processes
the used to manufacture their chemical. The re-
I industry. search that they conducted for this report
'ts were then provided them with the information they
through would need for the rest of the semester's
ducts, just work. In addition to production methods, the
production students were also asked to report on the
s are industrial significance of their chemical, what
9d within a industrial and consumer products were pro-
amical duced from their chemical, the feedstocks
complex. used to produce their chemical, the economic
impleted role their chemical played in the global
orts that economy (imports, exports, and trade pat-
ifferent teams and any other information that was
eir process significant or interesting. The purpose be-
ed a final hind this was to illustrate the importance of
that their chemicals and to heighten student in-
I their full terest in the overall project.
s work. Economics The first major topic that the
class covered was economics in process de-
sign, specifically the estimation of process equipment costs,
capital investment costs, and manufacturing costs.'5' One
week after completing the material on economics, the stu-
dents handed in their second midterm reports, which ana-
lyzed their processes from an economics standpoint. Stu-
dents were specifically asked to demonstrate their mastery
of the economics material by estimating the equipment, in-
vestment, and production costs for their process. A serious
hindrance to this evaluation was a lack of sufficient informa-
tion in the literature to accurately determine equipment sizes
or even to identify all of the correct processing equipment.
Students were therefore given a list of wild assumptions that
they were allowed to make, for the purposes of this assign-
ment only. Due to the highly inaccurate nature of these
equipment-sizing assumptions,* the results for the econom-
ics midterm reports were completely unreliable. They did,
however, allow students to exercise their cost estimation
skills, which was the point of the exercise. Surprisingly
enough, at least half of the class was within an order of

*Examples: All unspecified distillation towers are 50 feet high, 10
feet diameter, and contain 25 trays. Unspecified reactors are
5000-gallon stirred tanks; storage tanks hold 30 days supply of
feed or products.

Ethylten Ethylene Diclhoride( 1) Vinyl Chloride( 2) 1,1,1 Tri-ChloroEllthane
-1,1,2-TriChloroEthane
-Trichloroethylene
Perchloroethylene
EthyleneAmines EthyleneDiamine
Ethylene Oxide( 3) Ethylene Glycol( 4)
-DiEthylene Glycol
-TriEthylene Glycol
-TetraEthylene Glycol
-MonoEthanolAmine( 5) EthyleneDiamine( 6)
-EthylBenzn Styrene
-Acetaldehyde( 7) ---- Acetic Acid(8) Vinyl Acetate
--Acetic Anhydride( 9) l-Acetic Anhydride
-Propionaldehyde( 10) --- r-n-Proponal
--Propionic Acid
Ethyl Chloride

Propylene--- -Acrylonitrile( 11) ----- Adiponitrile( 12) -lHexaMethyleneDiarine
-AcrylAmide
Propylene Oxide
-Propylene Chlorohydrin( 13) -Propylene Oxide( 14) -- Propylene Glycol( 15)
I-Allyl Alcohol
L- IsoPropanolAmines
-Cumene
-n-Butyraldehyde( 16) -y n-Butanol
2-EthylHexanol
-n-Butyric Acid( 17)
S-iso-Butyrladoehyd IsoButand
L IsoButyric Acid
-looproponal(18) -Acetone Bisphenol-A
-Acrylic Acid( 19) Ethyl Acrylate
n-butyl Acrylate
I-Methyl Acrylate
-Acrolein( 20) Acrylic Acid

Q) n-Butane- -Acetic Acid
-Methyl Ethyl Ketone
-Maleic Anhydride( 21) Fumaric Acid
.-Malic Acid
--Malathion
-Butadiene
S- Formic Acid( 22) Oxalic Acid
Propionic Acid( 23)
Butyric Acid
-Methanol
S- Ethyl Acetate
-Methyl Acetate

Butylene nButeneo-Butanol( 24) ----- -Methyl Ethyl Ketone(25)
Sesec-Butyl Acetate( 26)
Butsdiene( 27) Adiponitrile
1-Butene( 28) ---- Valeraldehyde( 29)
1Butylene Oxide( 30)
Q L-Iso-Butyne- Methyl tert-Butyl Ether

Benzene- -EthylBenzne( 31) Styrene( 32) Styrene Oxide
-Cumene( 33) Phenol( 34) Bisphenol-A( 35)
Cyclohexanone
Aniline
C lAdipic Acid( 36)
Acetone( 37 ) Bisphenol-A
Cyclohexarn -Cyclohexanol( 38 ) -Adipic Acid
LCyclohexanone( 39) Adipic Acid
0 -NitroBenzene Aniline( 40)
Chloro Benzene

Chemical Engineering Eduction

magnitude of the published price per pound of their chemi-
cal, as listed in Chemical Marketing Reporter."1
Selection of Materials The next major topic covered by
the class was selection of materials for chemical production
service.5781 This topic included coverage of corrosion mecha-
nisms, mechanical strength, high and low temperature ef-
fects, chemical attack, alloying properties, machinability,
and cost. Besides the traditional coverage of metals, some
attention was also given to alternate materials such as poly-
mers (both plastics and rubbers), concrete, refractory brick,
ceramics, wood, glass, and glass-lined steel. Upon comple-
tion of this topic, students prepared a third midterm report
analyzing their process from a materials-of-construction view-
point. Students were asked to first identify all process condi-
tions that would have a significant impact on the selection of
materials and then to determine the appropriate materials)
of construction for their production process. Constraints were
imposed of no more than five materials for the construction
of the entire plant, including up to three primary materials
for the majority of construction, plus secondary materials for
special purposes. Students were also asked to evaluate how
their materials selection would affect their economic analy-
ses, without going back and recalculating any costs. Al-
though the more logical approach would be to select materi-
als first and then perform the economic analysis, the impact
of material choices on the cost estimation is emphasized by
performing the steps in the wrong order. Students were later
asked for similar judgment evaluations regarding design
changes made for safety and environmental reasons.
Safety Following selection of materials, the class re-
ceived two weeks instruction on safety in chemical process
design-specifically, one week on fires and explosions and
one week on hazards evaluation."9' Students then prepared a
midterm report analyzing the safety and hazards of their
chemical production processes. These reports started with
identification of the chemicals and process conditions that
were cause for particular safety concern. Information gath-
ered from Materials Safety Data Sheets on the world wide
web was particularly useful for this portion of the semester
project. The students performed a sample HAZOP analysis
of one portion of their process and concluded with recom-
mendations for precautions to be taken to properly handle
the safety concerns that had been identified. In many cases,
this assignment required students to study safety-related ma-
terial that was not specifically covered in class.
Environment The final major topic covered was envi-
ronmental issues in process design. The material covered in
class included nine major environmental regulations* that
apply to the chemical processing industry,"0" industrial meth-
ods for processing solid, liquid, and gas waste streams, and
methods of designing processes to minimize the amount and

* TSCA, CERCLA, RCRA, CWA, CAA, EPCRKA, PPA, OSHA,
and FIFRA.
Summer 1996

toxicity of waste generated. The midterm assignment for this
topic asked students to analyze and reduce the environmen-
tal impact of their production processes, first by identifying
all potential sources of environmental concern and then by
making recommendations regarding process modifications.
The recommendations were to consider both design adjust-
ments prior to plant construction and modifications appro-
priate to existing plants.
Final Report At the end of the semester, students were
asked to submit a final comprehensive report on their as-
signed chemical. Naturally, most groups started these final
reports by compiling the five midterm reports into five sec-
tions and correcting the errors from their earlier work. They
were also expected to assemble the whole into a cohesive
unit and to add any material that they felt was necessary for
complete coverage of the subject.

Summary Sheets
Each midterm report included an unfastened, single-page
summary sheet. Ungraded copies of the first summary sheets
(background) were compiled into a large hallway display so
that students could see the interrelationships between the
assigned processes and the rest of the petrochemical indus-
try. This display also served to inform other students and
faculty of the projects being conducted by the plant design
students. Copies of the background, safety, and environmen-
tal summary sheets were distributed to all students in the
class so that everyone could gain some understanding of the
chemical production processes being studied by their peers.
The economics and materials summaries were not distrib-
uted because there were not enough differences between
groups for the students to gain appreciably from viewing
their peers' work, and in the case of the economics reports,
the lack of sufficient design details made the results of the
analyses highly questionable.

Poster Presentations
Departmental interest in the activities of the design class
developed during the semester. Also, some students in the
class expressed regret that the hard work they were perform-
ing would never be seen by anyone other than the graders.
Because logistics prevented the use of oral presentations in
this particular class, we decided to display the students'
work in the form of a poster presentation in the corridors of
the chemical engineering department. The choice of venue
was due both to the space requirements for 38 posters and to
address student concerns that no one would bother to view
student posters during the last hectic week of the semester.
Some students expressed concerns that poster production
would require a lot of time at the end of the semester and that
the experience would only benefit the small number of stu-
dents who were planning to attend graduate school. There-
fore, several steps were taken to increase the value of the
poster display for all students. First, we pointed out that the

preparation of effective visual aids is an important skill in
engineering, whether presented in a report, a poster, or a
transparency, and that many of the same skills are required
in any case. Second, each group of students was given the
choice of either preparing a simple poster for homework
credit only or producing a more elaborate poster that would
also count for up to 20% of their final project report. Third,
engineers from nearby chemical companies were invited to
judge the posters, with prizes (1995 CACHE CD-ROMs)
awarded to the best entries. The industrial judges were cho-
sen to appeal to those students who were in the job market
by giving the students a chance to discuss their work with
the industrial contacts.
There were several unplanned benefits of the poster dis-
play, one of which was the chance for sophomores and
juniors to learn something about the petrochemical industry
and to see how their engineering skills might eventually be
used in industry. Another benefit was the positive impres-
sion the display made on a number of departmental visitors,
who expressed appreciation for the students' work.

LOGISTICAL ISSUES
Group Assignments The assignment of students to
groups can be conducted in a variety of ways.'"" In past
years, students were allowed to choose their groups, which
led to a concentration of experience within certain groups
(all the students who had taken reactor design together re-
formed themselves into plant design groups, leaving the
remaining groups with no reactor design background). This
semester, students were allowed to request their group
assignments, but the instructional staff made the final
assignments, with the criteria that each group have a
certain minimum background and that no group have an
excessively skewed GPA.
Group Participation In any group project situation
there is the potential problem of students who do not per-
form their share of the work, or conversely, students who
take over and do not allow their partners to contribute appre-
ciably. For this project there is the added temptation of
students dividing the midterm reports among group mem-
bers and then working individually rather than collectively.
The latter approach would be acceptable if the work were
divided fairly, except for the fact that each student would
then learn only one portion of the course material rather than
the broader coverage that is desired. To ensure a complete
understanding by all students, questions were placed on the
exams that required them to be familiar with all aspects of
their project, including portions completed by their partners.
Teaching Assistants In order to evenly distribute the
supervisory responsibilities of the four teaching assistants
(TAs) assigned to this course, the class was conceptually
divided into four sections based on principal derivatives of
ethylene, propylene, butylene/butane, and benzene, as shown

in Figure 1. Each section had a particular TA assigned as the
primary source of assistance for the groups within that sec-
tion. Students were asked to first seek assistance from the
TA assigned to their chemical and then seek further assis-
tance from an alternate TA or the course instructor if they
still had unresolved questions. In this manner, each TA was
responsible for understanding no more than ten (related)
production processes, while the primary instructor oversaw
the activities of all the groups.
Report Grading Grading thirty-eight midterm reports
every two to three weeks is too much work for any one
person to reasonably handle. Neither is it fair to have differ-
ent students graded by different graders. Therefore, the mid-
term grading was shared on a rotating basis, with the course
instructor grading the first (background) reports, and differ-
ent TAs grading the other midterm reports. The final semes-
ter reports were graded by the instructor while the TAs
graded the final exams.
Personalized Assignments The first project assignment
was personalized using the form letter capabilities of a popu-
lar word processor and data taken from the class roster
spreadsheet. Each assignment included the individual
student's name, group number, and assigned chemical wher-
ever appropriate in the document. The problem with this
technique was that it took an unacceptable amount of class
time to hand out the assignments to individual students, as
well as requiring a long time for the computer to print 155
assignments. As a compromise, later assignments were per-
sonalized by groups, with five stapled copies of the group
assignments being handed out to each group.

STUDENT RESPONSE
The University of Michigan employs a course evaluation
system similar to that used by many universities, in which
students rank various aspects of the course on a scale from
one to five at the end of the semester.""2 The year that this
design project was first implemented, 21 out of 25 questions
showed an increase in student rankings from the previous
year. The average ranking of all 25 questions rose from 3.31
to 3.71. For the questions specifically related to the design
project, the rankings rose even more dramatically, from 2.97
to 3.87. The lowest ranking increased from 2.48 to 3.02, and
the highest ranking rose from 3.85 to 4.08.
Of the rankings that decreased, one of the questions dealt
with the amount of work required for the credit received.
This ranking decreased slightly, from 3.83 to 3.76. But an-
other question, dealing specifically with the amount of work
required for the design project, increased its ranking from
2.96 to 3.71. Students were apparently more satisfied with
the project workload, but slightly less satisfied with the
overall workload for the course than in the previous year.
The other three questions that showed declining rankings
involved the assignment of grades, with the average of the
Chemical Engineering Eduction

three questions declining from 3.78 to 3.47. One cause of
this lowered ranking is believed to be student frustration
caused by the changing of graders for each midterm report.
Students felt that although they worked hard to address the
weaknesses pointed out by each TA, they would just be
rebuked for something different on the next report. Another
contributing factor to student dissatisfaction with grade as-
signment involved some regrading of the first exam, which
was totally unrelated to the design project.
Student responses on open-ended questions also show an
appreciation for the design project and an increased appre-
ciation for the class as a whole. Although there were no
specific questions regarding the design project during previ-
ous years, several students addressed the topic anyway, all
negatively. The general consensus of the previous year was
that the course as a whole was disjointed and the design
problem was completely irrelevant to the topics being
studied. Some students indicated that they had not learned
anything and still did not understand the point of the
course at semester's end.
The year that this design project was implemented, a spe-
cific question was added to the open-ended form requesting
student evaluation of the design project. Overall response
was highly favorable, with positive responses outnumbering
negative ones by four to one.
The negative comments were primarily from students who
dislike group work of any kind and from a few students who
felt the workload was too high, particularly when a midterm
report would happen to fall due the same week as other
assignments. The poster contest also drew criticism from
some students who felt that it was a lot of extra work with no
educational benefit and that it had no relevance to their
future careers in industry. It should be noted, however, that
the poster presentation was the only component of the
semester's workload that was not announced at the begin-
ning of the semester.

CONCLUSIONS AND RECOMMENDATIONS
The design project format outlined in this paper has been
highly effective in providing focus for a highly disjointed
course, and has been an interesting educational experience
for both students and their instructors. End-of-semester stu-
dent tensions were still high, as they probably always will be
in senior design courses, but there was much less frustration
expressed regarding competitiveness for limited resources.
Student evaluations of the course improved significantly,
especially for the questions relating to the design project
portion of the course.
There are, however, some areas for improvement. There
should be a clear, well-defined set of report-grading criteria,
used by all graders and clearly understood by all students.
(Those criteria could adapt from one report to the next, so
long as they are well understood by all concerned.) The
Summer 1996

poster display adds a definite benefit to the course and should
prove more palatable to students if it is announced at the
beginning of the semester. The safety and environmental
topics are not identical, but they are similar enough that they
could be combined into a single assignment. Equality of
effort in a group project is a serious concern, but one that is
common to all group activities. This approach does entail a
lot of work for all concerned, but it is also more interesting
and more educational for both the students and the instruc-
tional staff than the traditional approach.

ACKNOWLEDGMENTS
Implementation of this project would not have been pos-
sible without the help of teaching assistants Sanjeev Majoo,
Dieter Schweiss, Hetal Patel, and Mike DiBattista, who
handled the ethylene, propylene, butane/butylene, and ben-
zene sections, respectively. Grateful acknowledgment is also
given to Ravi Dixit of Dow Chemical Company and to Tom
Pakula of Marathon Oil Company for their assistance in
judging the poster contest and to Peter Rony for furnishing
the CACHE CD-ROMs awarded as prizes. Thanks are
also due to Jim Ottaviani, Leena Lalwani, and rest of the
University of Michigan engineering library staff for the
invaluable assistance they provided for both myself and
the "swarm of locusts" that descended on their library
every two to three weeks.

REFERENCES
1. Felder, Richard, "We Hold These Truths to be Self-Evi-
dent," Chem. Eng. Ed., 25(2) (1991)
2. Bloom, Benjamin S., Taxonomy of Educational Objectives:
The Classification of Educational Goals. Handbook I: Cog-
nitive Domain," David McKay Company, New York, NY
(1956)
3. McCaleb, Kirtland E., Chemical Origins and Markets, 6th
Ed., Stanford Research Institute, Menlo Park, CA (1993)
4. Bell, John T., "Modeling of the Global Petrochemical Indus-
try," PhD Thesis, University of Wisconsin, Madison, WI
(1990)
5. Peters, Max S., and Klaus D. Timmerhaus, Plant Design
and Economics for Chemical Engineers, 4th ed., McGraw-
Hill, New York, NY (1991)
6. "Chemical Prices," Chemical Marketing Reporter, various
issues
7. Van Vlack, Lawrence H., Elements of Materials Science and
Engineering, 6th Ed., Addison-Wesley (1989)
8. Kirby, Gary N., "How to Select Materials," Chem. Eng., 3
November (1980)
9. Crowl, Daniel A., and Joseph F. Louvar, Chemical Process
Safety: Fundamentals with Applications, Prentice Hall (1990)
10. Lynch, Holly, "A Chemical Engineer's Guide to Environ-
mental Law and Regulation," National Pollution Preven-
tion Center for Higher Education, Ann Arbor, MI, April
(1995)
11. Brickell, James L., David B. Porter, Michael F. Reynolds,
and Richard D. Cosgrove, "Assigning Students to Groups
for Engineering Design Projects: A Comparison of Five Meth-
ods," J. of Eng. Ed., July (1994)
12. Felder, Richard M., "What Do They Know, Anyway. 2. Mak-
ing Evaluations Effective," Chem. Eng. Ed., 27(1), (1993) 0

pMM classroom

WAKE-UP TO ENGINEERING!

ROBERT P. HESKETH*
The University of Tulsa Tulsa, OK 74104-3189

he work an engineer does is a mystery to many
people. Engineers will try to explain their work to
non-engineers by giving an example of a typical
problem that they have solved, but the explanation frequently
includes a description of an engineering process and the
equipment contained in the process. So, despite the engineer's
enthusiasm in giving his explanation, the non-engineer often
leaves the conversation as puzzled as he was to begin with.
This paper presents a different approach by using a device
that everyone is familiar with: a coffee machine. Just about
every home has one on the kitchen countertop. While other
appliances or equipment could be used to demonstrate engi-
neering concepts, they are less accessible to the non-engi-
neer. For example, a home heating and cooling system would
be an excellent subject, but it is usually hidden away from
view and parts of it extend both under the floor and through
the ceilings. Coffee machines, on the other hand, can easily
be cut open for closer examination and are inexpensive (you
can pick one up at a yard sale for next to nothing).
The coffee machine (see Figure 1) embodies principles
from several engineering disciplines. Chemical and mechani-
cal engineers design the heaters, the condensers, and the
systems for multiphase transport of fluids, and they fabricate
plastic and glass components. Leaching organic compounds
from coffee beans uses principles from mass transfer, unique
to chemical engineering, while automation requires concepts
from electrical, mechanical, and chemical engineering. Fi-
nally, engineering decisions are required to select the com-
ponents of a system and place them within an affordable,
compact unit that can be easily operated by the consumer.
The coffee machine embodies examples of at least eight
unit operations, as can be seen in Figure 1: tank drainage
through a one-way valve; tubular heater; upward two-phase
flow in pipes; condenser; flow distribution and bypass; leach-
ing and filtration; and particle size reduction. Underlying
these unit operations are fundamental principles of engineer-
ing and engineering science such as fluid flow (both single
and two-phase), heat transfer, thermodynamics ("engineer-
ing science" and equilibrium), mass transfer, particle size
* Currently at Rowan College, Glassboro NJ 08028-1701
Copyright ChE Dwision ofASEE 1996

Robert Hesketh is an assistant professor at
the University of Tulsa. He received his BS
(1982) at the University of Illinois-Urbana and
his PhD (1987) from the University of Dela-
ware. His research interests are in the areas of
combustion, fluidization, and multiphase flows,
and he teaches freshman engineering, mass
transfer, and reactor design. At the University
of Cambridge, England, he conducted
postdoctoral research in fluidization (and
cleaned many plugged coffee machines!).

distribution, surface area, and general and organic chemis-
try. Additional considerations such as strength of materials,
engineering economics, electronics, and circuits are involved.
The chemical engineering department at the University of
Tulsa uses a coffee machine demonstration to introduce high
school students to engineering concepts. Also, for the past
four years, a coffee machine demonstration has been used at
university recruitment functions and at Engineering Week,
and at an NSF Young Scholars summer program it is used to
introduce the Young Scholars to a series of engineering
laboratory experiments (described later in this article). It
could be used in other summer programs, such as the Sum-
mer Institute at the University of Nevada."' In recruitment
activities, where a large number of students visit the depart-
ment, an abbreviated version (20-30 minutes) of the demon-
stration is given, and it is also used in engineering classes
such as mass transfer. Northwestern University uses a coffee
machine example for their freshman engineering class.t2'
NOTE; In the following example I use technical terms for
the benefit of the readers on CEE. But in an actual demon-
stration, I would eliminate the use of words such as leach-
ing, condenser, thermocouple, etc.
COFFEE MACHINE DEMONSTRATION
For this demonstration, a coffee machine is altered to
make all of the components visible to the students. The back
and top are cut out and replaced with clear plastic sheets. In
addition, the bottom plate is removed. The riser tube that
connects the tubular heater to the condenser (see Figure 1) is
replaced with clear plastic tubing. These alterations enable
the students to observe the two-phase flow and steam con-
densing as coffee is made. At the start of the demonstration,
I set up a funnel stand with at least four funnels, which
require filter paper and several receiving flasks. I also have

Chemical Engineering Eduction

available roasted coffee beans and a coffee grinder.
The demonstration begins by introducing the students to
the basic fundamentals related to the coffee machine's op-
eration. I explain that the engineer must have a working
knowledge of basic and engineering science just to begin
designing the device. I introduce humor whenever possible
and make a mess on the table.
I show the students a bag of gourmet coffee beans and ask
them, "How do you make coffee?" The usual response is to
"add water to the beans." So, in a humorous vein, I add cold
water to the coffee beans and ask if anyone would like to
share my "gourmet coffee." Continuing in this interactive
mode, I make "coffee" of widely ranging strengths and after
each step I ask "Who would like to drink my gourmet cof-
fee?" The steps I use are listed in Table 1.
Throughout this sequence, I add discussions of several
subjects of engineering science.
For example, the thermodynam-
ics topic of phase equilibrium is
examined. I present it with the *
question, "How hot can you heat -
water?" Following their re- **
sponses I ask, "How could you F
get the water hotter than 1000C *S *.
or 2120F?" Usually there is no
response to this question, and I
ask them to think about how a oe a t
pressure cooker works. I show o 0
them a P-T phase diagram of
pure water and illustrate that at one-way valve
higher pressures water boils at
higher temperatures. Other sub-
jects that can be introduced are
given in Table 2. Fiur,,, 1. Schema

tubular
with tw
tirt nf

At this point in the demonstration I have made a mess
on the table and observe that "it would be nice to have
this process contained in one unit." I tell the students
they are now engineers and that we will design a coffee
machine, relating the principles of basic engineering sci-
ence to the design.
I begin with the water reservoir. The first questions are:
What size? Where should it be located with respect to the
other components? The next question involves how the wa-
ter should move from the reservoir to the heater. To help the
students answer these questions, I show them a coffee ma-
chine on which the back wall of the water reservoir has been
replaced by a clear plastic sheet. Examples from around the
community, such as water towers and pumping stations, can
be given to demonstrate fluid flow.
The next step is the heat exchanger. Open-ended questions
such as what energy source
should be used to heat the wa-
I ter (electric AC or DC, coal,
nser and Distributor natural gas, solar energy, etc.)
Bypass lever are discussed. Based on avail-
lacing ngand ability, electricity is chosen as
leaching and
filtration the energy source, and I show
them the tubular heater at the
bottom of the coffee machine.
Then there are questions on size,
fluid flow rates, and the desired
outlet temperature of the wa-
ter. I also show them the one-
1 timer a way valve at the inlet of the
timer and
Ster switch heater that prevents liquid and
Gases from flowing back into
heater AC Power the reservoir.
o-phase flow
n rnffpp mnrhine I then ask, "How do we get

TABLE 1

Student instructions Action Result
1. Add water to the coffee Add cold water to the coffee beans. Clear liquid
2. Grind the coffee beans For dramatic/humorous effect, add ground coffee to a funnel without a filter paper present. Dispersed coffee grounds in water
Pour cold water over the ground coffee and watch the grounds go into the receiving flask.
3. Use filter paper Separate the coffee grounds from the cold water using filter paper. Slightly colored water
4. Use hot water Make coffee. Brown colored liquid

TABLE 2
Courses in the Basic Sciences Related to the Coffee Machine

Course Topic Comment
General chemistry Solubility ...................................................... Effect of water temperature on solubility.
Organic chemistry Organic chemicals ........................................ The "brown stuff' and caffeine. The concept that everything is made of chemicals and the notion
that chemicals are always bad as being ludicrous.
Thermodynamics Boiling points ............................................... The effect of pressure on the boiling point. P-T phase diagram of water.
Mass transfer Surface area and particle size reduction............. The concept of increasing surface area on the leaching of chemicals from coffee beans.

Summer 1996 211

Conde

water to flow uphill?" The students' usual response is a
pump-this leads to a discussion of economics since the
addition of a pump would raise the price of the coffee ma-
chine by about $100. In some instances, students suggest
that if all of the water were converted to steam, a pump
would not be necessary-but I point out that this would
require larger heaters and condensers than are currently be-
ing used, again involving additional expense. One creative
response has been to place the reservoir and heater above the
coffee filter so that the water will drain by gravity. This leads
to a discussion of the space limitations and the need for
compact designs when marketing a product.
Upon starting the coffee machine, the students are able to
observe two-phase flow upward through the clear plastic
tube into the condenser. They see that the tubular heater has
three functions: it warms the brewed coffee directly above
the heater; it heats the water; and it provides the driving
force for fluid to flow uphill, similar to a thermosiphon unit.
The demonstration shows the students that there are often
many solutions to a single problem, but the best solution is
often the cheapest.
The condenser at the top is demonstrated by replacing the
opaque plastic with a clear plastic sheet. The students dis-
cuss how much of the water must be boiled to move the
liquid to the top of the machine. This can be discovered by
performing experiments in which the amount of steam in the
riser is varied and the total liquid flow rate is measured.
This, in turn, introduces the question of what fluid flow rate
is needed for proper operation of the leaching unit. Would
the maximum fluid flow rate flood the condenser or leaching
unit and cause dangerously hot water to flow out of the
machine? The students can see that each unit within the
coffee machine is interrelated; outputs from one unit are
inputs to another unit. In addition, aspects of safety in engi-
neering design are considered.
The next observable unit operation is unique to the field of
chemical engineering: leaching. The need for a distributor is
introduced by asking, "What happens if all the water flows
down one side of the coffee grounds?" Again, questions of
filter size and shape are discussed since they determine both
the amount of coffee grounds that can be loaded and the
residence time of the water in the coffee grounds. Demon-
strations of the effect of particle size and bed height on fluid
flow rates can be given using marbles and sand particles in
several of the funnels.
Many coffee machines have a lever that adjusts the strength
of the coffee. How is this achieved? Typical student re-
sponses suggest the examples of particle size, water tem-
perature, and contact time of the water with the coffee par-
ticles, but none of those methods are used. Instead, the
strength of the coffee is altered by having a portion of the
water bypass the coffee grounds and pass directly into the
receiving vessel. This is achieved by using a lever and slide
212

that allows water to flow through a hole located on the
perimeter of the distributor plate. Water flowing through this
hole passes between the filter paper and the plastic filter
support. This device produces the same effect as diluting
your coffee by adding hot water to your cup.
The next aspect of the design is to determine the materials
of construction for the coffee machine. Several options for
each of the components are discussed, as well as the glass
coffee pot. We bring in aspects of strength of materials,
temperature limitations, corrosion, and cost of materials.
Finally, we discuss ways in which the process can be
automated. This includes adding timing circuits and ends
with expensive options such as stereos and robotics. Again,
basic aspects of marketing and economics are discussed.
A summary of the courses and topics related to the coffee
machine are given in Table 3. Comparing Tables 2 and 3
show that at least nine courses in the engineering curriculum
are introduced to the student through this demonstration.

RECRUITMENT ACTIVITIES
The coffee machine is excellent for recruitment activities.
A typical recruiting session includes the coffee demonstra-
tion, a tour of the undergraduate computer and unit opera-
tions laboratories, and research laboratory demonstrations.
The tour of the research laboratory demonstrates the linkage
between education, research, and industry as well as illus-
trating undergraduate research opportunities.
Typically, one of the best laboratories to demonstrate
chemical engineering principles is a flow visualization labo-
ratory. It contains many examples of familiar devices. For
example, everyone drives a car with a catalytic converter,
but they have not seen it. A brief review of how a catalytic
converter is made and how it works is given, followed by a
demonstration of how the small 1-mm square channels are
coated with catalyst. Using a high-speed video camera, the
students can observe the two-phase flow coating process.3-4"
In another experiment, the relationship of fluid mechanics to
the performance of a heat exchanger fin is demonstrated by
using advanced laser Doppler anemometry systems.

TABLE 3
ChE Courses Related to the Coffee Machine
Course Principle
Fluid Mechanics.................... tank drainage, two-phase flow, flow through a bed
of particles and filter paper
Heat Transfer ........................ design of heaters and condensers
Unit Operations ..................... one-way valve; size-reduction equipment, filtration
Mass Transfer........................ leaching evaporation and flow distributors
Properties of Materials.............. materials of construction
Circuits or Advanced Physics ... timers and switches
Economics............................. cost of engineering design and construction of a
coffee machine

Chemical Engineering Eduction

In both of the above experiments, the students can relate
the principles of two-phase flow and heat transfer of the
coffee machine to industrial processes that affect their daily
lives. This tour usually ends by capturing the image of a
student volunteer on video, and then the image is digitized
and patterns are enhanced with a dazzling display of colors.
This laboratory is usually a highlight of the tour!

NSF YOUNG SCHOLARS SUMMER PROGRAM
The Young Scholars program is a hands-on summer camp
to introduce engineering to students entering the 9th and
10th grades. The experiments in this camp are designed to
stimulate the students' interest in the fields of science and
engineering by involving a device that is familiar to them:
the coffee machine. Each of these experiments is designed to
be challenging, but not overwhelming, to the student. In
summary, these activities

Demonstrate the role of laboratory experiments in the
engineering decision-making process
Show the interrelationship of engineering and science
required for the design and fabrication of a single product
Give stimulating and challenging experiments that relate the
laboratory experiments to a consumer product with which
most students are familiar.

In these experiments, we use equipment from the
undergraduate and graduate laboratories. A selection of ex-
periments that have been used in previous Engineering Sum-
mer Camps is given in Table 4.
We also discuss the chemical composition of a coffee
bean, the roasting process, and decaffeination using methyl-

ene chloride and supercritical fluids. We have taken scan-
ning electron micrographs of coffee beans and filter paper at
various stages of brewing.

SUMMARY
We have used the coffee machine for undergraduate re-
cruitment and for our engineering summer camp, and I have
also used it as part of a demonstration day in the mass
transfer class. Using dynamic simulators (such as HYSYS),
the coffee-making process could also be modeled as a
short design project in a senior class. The coffee ma-
chine, familiar to everyone, is an effective tool for moti-
vating students in engineering.

ACKNOWLEDGMENT
I would like to thank Martin Abraham, John Henshaw,
Ramon Cerro, Christi Patton, and Brenda McLaury for
their helpful suggestions. Funding for the Engineering
Summer Camp is through the NSF Young Scholars Pro-
gram ESI-9255956.

REFERENCES
1. Bayles, T.M., and F.J. Aguirre, "Introducing High School
Students and Science Teachers to Chemical Engineering,"
Chem. Eng. Ed., 26(1), 24 (1992)
2. Miller, W.M., and M.A. Petrich, "A Novel Freshman Class
to Introduce ChE Concepts and Opportunities," Chem. Eng.
Ed., 25(3), 134 (1991)
3. Kolb, W.B., A.A. Papadimitriou, R.L. Cerro, D.D. Leavitt,
and J.C. Summers, "The Ins and Outs of Coating Monolithic
Structures," Chem. Eng. Prog., February, p. 61 (1993)
4. Thulasidas, T.C., M.A. Abraham, and R.L. Cerro, "Bubble-
Train Flow in Capillaries of Circular and Square Cross
Section," Chem. Eng. Sci., 50(2), 183 (1995) O

TABLE 4
Engineering Laboratory Experiments

Principle
Particle Size Analysis

Extraction of Coffee

Heat Transfer

Fluid Flow

Timer construction

Digital Signals & Robotics
Fracture of Materials
Polymer Chemistry
Organic Chemistry

Computer-Aided Process Control

Experimental Goals
Establish the relationship of grinding time and grinder type to the size
of coffee grounds produced while exploring techniques for analyzing
particle size. Examine relationship between particle size, pressure drop,
and fluid flowrate

Determine the effect of water temperature, particle size, and filter type
on the strength of coffee produced

Examine conduction, convection, and radiation. Determine the effect
of insulation on heat loss

Determine effect of tube length, tube diameter, and liquid height on tank
drainage time and the length of a free jet. Simulate the riser in the
coffee machine using gas phase introduced at bottom of vertical tube.
Construct a timing device to turn a circuit on and off

Examine digital control and automation

Examine and compare the strength of glass, metal and plastics
Examine the production of polymers used in making plastics

Measure the concentration of caffeine as a function of time in a
percolator coffee machine

Investigate liquid level control using a computer

Engineering Equipment
Optical microscope, sieves, coffee grinders, stopwatch,
and funnels

UV spectrophotometer, electronic balance, grinder, stop-
watch, coffee machines, filter paper

Thermocouples, insulation, rods, heaters, and mixers

Tanks, tubes, measuring tape, compressed gas, rotameter,
graduated cylinders, stopwatch

Electrical circuit components: transistor, potentiometer, re-
sistors, LED, capacitors, peizoelectric disk, circuit board

Oscilloscope, robotic cars with paper card reader

Mechanical testing equipment
Molds, polymers

HPLC, percolator, and stopwatch

Tanks, valve, actuator, pressure transducer and computer

Summer 1996 21-

classroom

ChE APPLICATIONS

OF ELLIPTIC INTEGRALS

PETER W. HART,* JUDE T. SOMMERFELD
Georgia Institute of Technology Atlanta, GA 30332

Elliptic functions and elliptic integrals remain a mys-
tery to most chemical engineers-students, profes-
sors, and practitioners alike. Undoubtedly, this lack
of familiarity derives from the classical absence of any sig-
nificant general applications of these tools within the prac-
tice of chemical engineering. This situation is slowly chang-
ing, however, with recent developments in the area of fluid
mechanics, particularly in relation to safety considerations.
Thus, the purpose of this article is to present a brief exposi-
tion of the nature and genesis of elliptic functions and inte-
grals, followed by a summary of some of their applications,
with particular emphasis on chemical engineering problems.

ORIGIN OF ELLIPTIC FUNCTIONS
The fundamental elliptic functions actually derive from
the analytical solution"' to the parabolic partial differential
equation describing unsteady-state heat conduction in one
direction (z) through a flat plate n units thick. The initial
condition on the temperature for this problem is assumed to
be a Dirac function at the midplane of the plate ( z = n / 2).
The boundary conditions for the spatial variable (at z = 0
and at z = n) may be either of two such conditions com-

Jude T. Sommerfeld is professor in the School
of Chemical Engineering at Georgia Tech. He
received his BChE from the University of Detroit
and his MSE and PhD degrees, also in chemical
engineering, from the University of Michigan.
His industrial and academic experience has been
primarily in the area of computer-aided design,
and he has published over 100 articles in this
and other areas.

Peter W. Hart received his BS in Chemical Engi-
neering and Pulp and Paper Technology and his
MS in Chemical Engineering from the University
of Maine. His PhD in Chemical Engineering is
from Georgia Institute of Technology. He is cur-
rently working on pulping, bleaching, chemical
recovery, and environmental process develop-
ment, optimization, and improvements.
* Address: Westvaco Corporation, PO Box 118005, Charleston,
SC 29423-8005.

only invoked: 1) the two face temperatures are maintained
at a constant value, or 2) the two faces of the plate are
perfectly insulated, so that no heat transfer occurs at these
two boundaries.
The analytical solutions to this problem may then be recast
in terms of what are known as theta functions.12 These latter
are typically written as 0i(z), where i = 1,2,3,4 and 0 z 7 t.
The three fundamental elliptic functions are then defined as
various ratios of theta functions [0i (0), i (z)] and are denoted
by sn(u), cn(u), and dn(u). The parameters z and u are related
as follows: z = u /[03 (0)]2. A whole host of new elliptic func-
tions then derive from these three fundamental elliptic func-
tions, e.g., ns(u), cs(u), nc(u), sc(u), dc(u), sd(u), etc., as well as
a wide variety of mathematical expressions similar to trigo-
nometric identities. Lastly, the various elliptic integrals are
then defined in terms of these elliptic functions.

FUNDAMENTAL ELLIPTIC INTEGRALS
Perhaps a more straightforward manner in which to intro-
duce the subject of elliptic integrals, however, is to describe
one of the first problems that most likely led to their devel-
opment. Thus, consider an ellipse, with its center at the
origin of x-y coordinates (as in Figure 1), described by
2 2
x -1 (1)
a2 b2
where the lengths of its semi-major and semi-minor axes are
given by a and b, respectively. What then is the value of its
perimeter P (or periphery or circumference)? In the special
case of a circle with a = b = r, the area (A) and circumference
(C) are readily computed as nr2 and 2mr, respectively. Simi-
larly, the area of an ellipse is readily determined from the
calculus as nab, but the evaluation of its perimeter (P) is not
so simple. Specifically, this latter quantity must be obtained
by integration of the differential length of arc (ds) over the
entire periphery of the ellipse.
For this purpose, it is convenient to convert x and y in Eq.

Copyright ChE Division ofASEE 1996

Chemical Engineering Eduction

(1) to parametric form, e.g., to functions of the angular
parameter 6:

x = a sin 0

y = bcos9

where, as also indicated by Figure 1, 0 represents the eccen-
tric angle measured from the minor axis b. We recall the
definition of a differential length of arc as

ds= (dx)2 +(dy)2 (3

and let s here denote the arc length parameter measure
clockwise along the curve from the end of the minor axi,
Then, in terms of the angular parameter e,

ds = (a2 cos2 6 + b2sin2 0) dO

d
S.

Figure 1. Sketch of an ellipse for determination of the
value of its perimeter P.

TABLE 1
Fundamental Elliptic Integrals (of the First, Second, and I

Kind Incomplete
0
1. F(k, 0)= dO
S1- k2sin2

2. E(k, )= f -k2 sin2 6 de

0d

3. II(k,n, )= dO 2 20
S(I + n sin2 0) 1-k2 sin2

where k = modulus of the elliptic integral
= amplitude of the elliptic integrals
n = parameter in elliptic integrals of the third kind

Taking advantage of symmetry, it is clear that the total
perimeter P of the ellipse is given as four times the perimeter
of one quadrant, e.g., from 6 = 0 to 6 = n /2. Thus, after
replacing cos2 0 with (I sin2 o), we have
r/2 __2
P=4a l-e2 sin 20d (5)
0
) as the expression for the perimeter of an ellipse. In Eq. (5),

e-- (6,

and is known as the eccentricity of the ellipse. More com-
(4) only, this quantity is referred to as the modulus k of the
integral appearing in Eq. (5), which in turn is known as the
complete (because of the fixed upper limit of rt / 2) elliptic
integral of the second kind, generally denoted as E(k). An
incomplete elliptic integral of the second kind

E(k,) = l-k2 sin2 dO (7
0
has a second angular argument 0 and obviously corresponds
to incomplete integration (( < 7t / 2) about the arc of the first
quadrant in Figure 1.
The integral of Eq. (7) is one elliptic integral of three
fundamental types. It can be shown"3' that any integral of the
form

I = R(x, x)dx (8

where X is a cubic or quartic in x and R denotes a rational
function, can, by suitable linear transformations and reduc-
tion formulae, be expressed as the sum of a finite number of
elementary integrals plus elliptic in-
tegrals of these three fundamental
rhird Kinds) types. These types, in both incom-
plete and complete form are summa-
rized in Table 1.

There exist in the mathematical lit-
erature"4'5] extensive compilations of
the transformations necessary to per-
form any integration involving the
elliptic integrals associated with a
given problem. Similarly, there are
numerous handbooks6"81 that tabulate
numeric values of elliptic integrals
to aid in the actual computations as-
sociated with such a problem.

PHYSICAL APPLICATIONS
Before proceeding on to technical
applications of elliptic integrals
closely associated with chemical en-

Complete

1- k2 sin2

n/2
E(k)= fl-k2 sin 0 d
0

k n/2

1 + nsin20 1-) k2 sin2 0

Summer 1996

gineering practice, we choose to summarize briefly some of
the earliest physical problems whose solutions incorporate
elliptic integrals. Most of these are of a mechanical nature.'9'
One of the early practical problems involving elliptic inte-
grals pertains to determination of the oscillation period T of
a pendulum of length L swinging through a circular arc. The
solution of the ordinary differential equation describing this
situation yields the expression12'3'9'

T = K(k) (9)

where g is the acceleration due to gravity. The modulus k of
the elliptic integral in Eq. (9) is given by

k = sin (a / 2) (10)

Here, h represents the height of the maximum point to which
the pendulum swings above its rest point, while a is the
angular amplitude of the pendulum oscillations (correspond-
ing to the height of this maximum point h).
Numerous other applications of elliptic integrals include
characterization of planetary orbits under forces of attrac-
tion,'21 determination of the torque exerted by a mechanical
brake,'9 and calculation of electrical current flow in a con-
ducting plate.'2' And, of course, there is the natural geometric
extension of computing the surface area of an ellipsoid. The
general equation for the latter is
x2 2 z2
a-- + b-2 + c 2
where a > b > c. It can be shown'23' that the surface area of
such an ellipsoid in the general case is

S = 2 +c2 +tab Cos2 vU][F(u, k)] + sin2 ][E(u, k)]} (12
sinu I1 [
wherein the additional parameters u and k are defined as
1-c2 2 = sin2 v (13)

1-c2/b2 =k2sin 2 (14)
Simpler formulas (not requiring elliptic integrals) result in
the special cases of 1) an oblate spheroid, for which a = b
(and hence k = 1), and 2) a prolate spheroid, for which b = c
(and hence k = 0). These various expressions for the surface
areas of ellipsoids lead somewhat into the topic of applica-
tions of elliptic integrals in chemical engineering. Thus,
from mass transfer studies,"0' for example, it is known that
liquid droplets, such as are formed as the dispersed phase in
liquid-liquid extraction, are often ellipsoidal in shape and
their area is directly related to the rate of mass transfer.

CHEMICAL ENGINEERING APPLICATIONS
Most known applications of elliptic integrals in chemical
engineering derive from fluid mechanics. A simple such

application'9' which readily comes to mind is determination
of the hydraulic radius (ratio of flow area to the wetted
perimeter) for a pipe of elliptical shape, where a value for the
perimeter of the elliptical cross-section is clearly required.
Other early applications of elliptic integrals from fluid me-
chanics include derivation of the capillary curve for a fluid
enclosed between two parallel vertical plates"'9 and determi-
nation of the complex velocity potential for steady irrota-
tional flow of liquid in two dimensions."'
Perhaps one of the more practical early uses of elliptic
integrals is found in the case of liquid flow across weirs-
traditionally more in the province of civil engineering but,
with the recent advent of multifarious environmental con-
cerns, often also employed by chemical engineers as mea-
suring tools. Thus, classical civil engineering texts""' 12 present
flow formulas for the more popular types of weirs, including
rectangular and triangular (or V-notch weirs). While not
employed extensively in this country (as they are in Europe),
however, circular weirs for the measurement of liquid flow
rates in open channels, such as ditches, flumes, and troughs,
have the advantage that the crest can be turned and beveled
with precision in a lathe. Moreover, this weir crest does not
have to be leveled, and hence the point of zero flow is
readily determined.
From the Bernoulli equation, the volumetric flow rate q as
a function of the crest height h across a circular weir with a
diameter of D, as depicted in Figure 2, is given by the
integral equation
h
q = 2 Cw 2 (D- z)z(h z)dz (15)
0
where C. is a weir discharge coefficient, accounting prima-
rily for friction losses, much like an orifice discharge coeffi-
cient in closed channel flow measurement. In a 1957 paper,
Stevens"" found the analytical solution, incorporating ellip-
tic integrals, for Eq. (15) to be of the form

q 4Cw 2gD5/2 [2(1-k2 +k4)E(k)- (2-k2)(-k2)K(k)] (16)
q 15

Figure 2. Open channel flow across a circular weir.

Chemical Engineering Eduction

The modulus k of the elliptic integrals appearing in Eq.
(16) is merely equal to Vh/D). In his paper, Stevens also
examined hundreds of experimental data points on water
discharge rates from circular weirs. These data went back to
the beginning of this century and were taken over the entire
range of h/D from 0 to 1 on circular weirs up to three feet in
diameter. An average value of the discharge coefficient C. of
0.59 was determined from his analysis of these data.
Stevens' results were subsequently adapted to the problem
of determining liquid overflow rates through circular open-
ings in process and/or storage tanks."4" Equation (16) thus
applies equally to the problem of computing such discharge
rates through circular apertures (or short discharge pipes),
given the size of the opening and the liquid level therein.
Indeed, Stevens"3' first became interested in this problem in
conjunction with measuring the flow rate through a short
pipe from a fishway into a power canal. In Reference [14],
an approximate representation of Eq. (16), invoking the
concept of relative volatility from vapor-liquid equilib-
ria, was also developed and presented. Lastly, it comes
as no surprise that this equation for the liquid flow rate
across a circular weir is really just a special case for flow
across an elliptical weir."51
The drainage of process vessels of many different shapes,
such as cylindrical, spherical, and conical, represent conven-
tional calculus problems, solutions to which have long been
known.'16' To be sure, with the recently heightened interest in
chemical process hazard analysis in addition to environmen-
tal issues, many of these drainage (or efflux) formulas have
also appeared in recent textbooks on process safety.""7 It has
been recently found that elliptic integrals (like Bessel func-
tions in heat transfer) have a way of recurring in many fluid
efflux problems with macroscopic circular geometries.
Thus, consider the problem of gravity drainage of a hori-
zontal annulus, W units long, such as might be represented
by the shell side of a double-pipe heat exchanger (see Figure

Figure 3. Cross-section of a horizontal circular annulus.
Summer 1996

3). The inner and outer radii of this annulus are denoted by r,
and r,, respectively, while the drainage occurs through an
aperture with a cross-sectional area of A, located along the
bottom center line of the annulus. A constant value for the
orifice discharge coefficient (e.g., Co = 0.61) is assumed.
Expressions for the drainage times required for the top and
bottom thirds of this annulus (volumes I and III, respec-
tively, in Figure 3) are readily obtained from earlier results
for conventional horizontal circular cylinders."'6 But the drain-
age time requirement for the middle volume (t,,) of this
annulus (that is, from the level of h = r, + r, down to h = r, r,)
is given by an expression incorporating elliptic integrals"8'

4W {(r2 r,)3/2 rj)3/2
3CoAo r22g

-2(r2 + r)'/2[r2E(k)-(r2 r)K(k)]} (17)

where the modulus k in this case is given by

k=- 2r (18)
k r2 + rl
The more general expression for partial drainage of this
middle volume (II) of a horizontal annulus (i.e., from some
intermediate elevation and/or down to some other intermedi-
ate elevation, both within this middle volume) is consider-
ably more complicated and specifically incorporates incom-
plete elliptic integrals of the first and second kinds."8
Until recently, most fluid efflux analyses pertained to in-
tentional drainage from an opening at the bottom of a vessel.
But now, because of increasing concerns about safety and
loss prevention in the process industries, there exists a need
for accurate formulas to compute fluid discharge and vessel
emptying rates for an opening at an arbitrary elevation. Such
a need may arise in analyzing an accident scenario resulting
from a moving vehicle, e.g., a forklift truck or an automated
guided vehicle (AGV), being driven into the side of a vessel.
Such analytical formulas were originally presented by
Crowl"9' for spherical and vertical cylindrical vessels.
Subsequently, the following expression was developed1t20
for the time t required for drainage of a horizontal cylindrical
vessel, with a diameter of D and W units long, from an
arbitrary initial liquid level of h, through a hole with a cross-
sectional area of A, and located at an equally arbitrary eleva-
tion of h0,

4 W {=[(D 2 ho)E(O, k) + hoF(, k)]
t 3CoA0o 2g

+(2ho+h D) (D-h h) (19)

A sketch of this configuration is shown in Figure 4. The
parameters of the incomplete elliptic integrals in Eq. (19) are

555- W

= sin D(h, --ho)
(D-ho)hi

k D-h,
SD

In this case, if the time required for the liquid level to fall from
an initial elevation of h, to some intermediate elevation h, (or,
equivalently, to discharge a given amount of material) is de-
sired, two successive applications of Eqs. (19) and (20) can be
employed for this purpose.

Recent interest has also arisen in the problem of exhausting
process vessels through drain piping systems.121 Thus, the case
of pipeline drainage of horizontal cylindrical tanks also re-
quires elliptic integrals.122' Such a configuration is presented in
Figure 5. In this instance, one is interested in the time required
to drain the contents of a horizontal cylindrical vessel with a
diameter of D and a length of W through a drain pipe system
with an inside diameter of d, attached at the bottom center line
of the vessel. This drain piping system originates at an eleva-
tion of h. units above the datum plane and has an equivalent
length of L. Fully developed turbulent flow through this system
is assumed, with a constant Moody friction factor of f.
With these assumptions, the resulting analytical solution'22' to
this problem again incorporates (in the general case of incom-
plete drainage of a partially filled vessel) the incomplete ellip-
tic integrals of the first F(0, k) and second E(0),k) kinds. The
latter collapse down to their complete form for the special case
of complete drainage of a completely filled horizontal circular
cylinder through a drain piping system. Saturator troughs in the
shape of horizontal semi-elliptical cylinders are employed ex-
tensively in the textile finishing industries. Not surprisingly,
the solution to the problem of determining drainage times for
such troughs through a piping system also invokes elliptic
integrals.'23'

CONCLUSION
In this article, we have addressed the subject of elliptic

integrals, including their origins and definitions. Early
scientific applications of elliptic integrals, primarily from
the physics area, were briefly summarized. Then, a num-
ber of such applications in chemical engineering, most of
which are relatively recent in origin, were described (see

Figure 5. Sketch of a horizontal circular cylindrical
tank with drain piping.

TABLE 2
Summary of Technical Problems with
Elliptic Integral Solutions

Problem Reference(s)
Physics Problems
A rea of an ellipse ........................................................ [2,9]
Period of oscillation for a swinging pendulum........... [2,3,9]
Torque exerted by a mechanical brake .......................... [9]
Motion of a whirling chain or skipping rope.............. [2,3,9]
Area of the surface of an ellipsoid .............................. [2,3]
Planetary orbits under laws of attraction ....................... [2]
Current flow in a rectangular conducting plate ................ [2]
Electrostatics of a parallel plate capacitor ..................... [2]

Chemical Engineering Problems
Hydraulic radius of an elliptical pipe ............................ [9]
Capillarity between two parallel vertical plates ............... [9]
Steady irrotational liquid flow in two directions .............. [3]
Fluid flow across circular weirs or openings............. [13,14]
Fluid flow across elliptical weirs or openings ............. [15]
Bottom drainage of horizontal annuli .......................... [18]
Efflux from punctured horizontal cylinders ................ [20]
Drainage of horizontal cylinders through piping....... [22,23]

Chemical Engineering Eduction

KI1

l ~ ca

Figure 4. Horizontal circular cylindrical vessel with a
puncture hole in its side and resulting liquid drainage.

Table 2). Most of the chemical engineering applications of
elliptic integrals to date have been in the fluid mechanics area.

NOMENCLATURE
A = surface area formed by the liquid level in a tank
Ao = cross-sectional area of flow opening
a = length of semi-major axis of an ellipse
b = length of semi-minor axis of an ellipse
C = circumference of a circle; length of chord formed by a
liquid level
Co = orifice discharge coefficient
Cw = weir discharge coefficient
c = length of third semi-axis of an ellipsoid
D = diameter of a circular tank or weir
d = diameter of a circle
do = diameter of flow opening
E = elliptic integral (incomplete or complete) of the second
kind

e = eccentricity of an ellipse = [a2 -b2 1/2/a

F = incomplete elliptic integral of the first kind
g = acceleration due to gravity
H = variable elevation of the liquid level in a tank above the
outlet of drain piping
h = maximum elevation of a swinging pendulum above its
rest point; variable elevation or height of the liquid
level in a tank
h, = initial elevation or height of the liquid level in a tank
h0 = elevation of a tank bottom above the outlet of drain
piping
I = general integral of Eq. (8)
K = complete elliptic integral of the first kind
k = modulus of elliptic integrals; parameter in calculation of
ellipsoidal surface areas, defined in Eq. (14)
L = equivalent length of piping
n = parameter of elliptic integrals of the third kind
P = perimeter of an ellipse
q = volumetric flow rate
R = rational function of x and -X in Eq. (8); radius of a

circus
r = radii
S = surf
s = leng
T = peri

lar tank or weir
is of a circle
ace area of an ellipsoid
th of arc
od of oscillation for a swinging pendulum

t = time
u = argument of elliptic functions
V = fluid volume
v = linear velocity
W = length of a horizontal cylinder
X = cubic or quartic function of x in Eq. (8)
x = arbitrary independent variable of integration; horizontal
coordinate
y = vertical coordinate
z = thickness of a flat plate
eek Letters
a = angular amplitude of oscillation of a pendulum
0 = amplitude of elliptic integrals
v = parameter in calculation of ellipsoidal surface areas,

defined in Eq. (13)
fl = elliptic integral (incomplete or complete) of the third
kind
it = number pi (3.14159...)
9 = theta function; angular argument of elliptic integrals

REFERENCES

1. Geankoplis, C.J., Transport Processes and Unit Operations,
2nd ed., Prentice-Hall, Englewood Cliffs, NJ (1983)
2. Lawden, D.F., Elliptic Functions and Applications, Springer-
Verlag, New York, NY (1989)
3. Bowman, F., Introduction to Elliptic Functions with Appli-
cations, English Universities Press Ltd., London, England
(1953)
4. Byrd, P.F., and M.D. Friedman, Handbook of Elliptic Inte-
grals for Engineers and Physicists, 2nd ed., Springer-Verlag,
Berlin (1971)
5. Gradshteyn, I.S., and I.M. Ryzhik, Tables of Integrals, Se-
ries, and Products, Academic Press, New York, NY (1980)
6. Spiegel, M.R., Mathematical Handbook of Formulas and
Tables, (Schaum's Outline Series), McGraw-Hill, New York,
NY (1968)
7. Jahnke, E., and F. Emde, Tables of Functions, 4th ed.,
Dover Publications, New York, NY (1945)
8. Handbook of Chemistry and Physics, 38th ed., Chemical
Rubber Publishing Co., Cleveland, OH (1956)
9. Reddick, H.W., and F.H. Miller, Advanced Mathematics for
Engineers, 3rd ed., Wiley, New York, NY (1955)
10. Wellek, R.M., A.K. Agrawal, and A.H.P. Skelland, "The
Shape of Liquid Drops Moving in Liquid Media," AIChE J.,
12,854(1966)
11. Streeter, J.C., Fluid Mechanics, 4th ed., McGraw-Hill, New
York, NY (1966)
12. Vennard, J.K., and R.L. Street, Elementary Fluid Mechan-
ics, 5th ed., Wiley New York, NY (1976)
13. Stevens, J.C., "Flow Through Circular Weirs," Proc. ASCE,
J. Hydraulics Div., 83(HY6), Paper 1455 (1957)
14. Skelland, A.H.P., and J.T. Sommerfeld, "A Simple Equiva-
lent to an Elliptic Integral Expression for Liquid Overflow
Rates from Tanks," Tappi J., 73(8), 177 (1990)
15. Sommerfeld, J.T., and M.P. Stallybrass, "Flow Equations
for Parabolic and Elliptical Weirs," J. Envtl. Sci. Health, in
press
16. Foster, T.C., "Time Required to Empty a Vessel," Chem.
Engrg., 88(9), 105 (1981)
17. Crowl, D.A., and J.F. Louvar, Chemical Process Safety: Fun-
damentals with Applications, Prentice-Hall, Englewood
Cliffs, NJ (1990)
18. Hart, P.W., and J.T. Sommerfeld, "Expressions for Gravity
Drainage of Annular and Toroidal Containers," Proc. Safety
Progr., in press
19. Crowl, D.A., "Liquid Discharge from Process and Storage
Vessels," J. Loss Prev. Process Ind., 5, 73 (1992)
20. Sommerfeld, J.T., and M.P. Stallybrass, "Elliptic Integral
Solutions for Fluid Discharge Rates from Punctured Hori-
zontal Cylindrical Vessels," J. Loss Prev. Process Ind., 6, 11
(1993)
21. Loiacono, N.J., "Time to Drain a Tank with Piping," Chem.
Engrg., 94(13), 164 (1987)
22. Sommerfeld, J.T., and M.P. Stallybrass, "Elliptic Integral
Solutions for Drainage of Horizontal Cylindrical Vessels
with Piping Friction," Ind. Eng. Chem. Res., 31, 743 (1992)
23. Sommerfeld, J.T., and M.P. Stallybrass, "Elliptic Integral
Solutions for Drainage of Saturator Troughs Through Pip-
ing," Amer. Dyestuff Reporter, 80(10), 20 (1991) O

Summer 1996

Gr

16 'Ocurriculum

COMPARISON OF

GAMS, AMPL, AND MINOS

FOR OPTIMIZATION

XUEYU CHEN, KRISHNARAJ S. RAO, JUFANG YU, AND RALPH W. PIKE
Louisiana State University e Baton Rouge,, AL 70803

Optimization of plant operations and process design
requires maximizing a profit function subject to a
plant model that can involve thousands of con-
straint equations. The mathematical programming modeling
languages of GAMS and AMPL were developed to alleviate
many of the difficulties associated with the development
and solution of large, complex mathematical program-
ming models like these and to allow direct formulation
and solution on a computer. They have problem formula-
tion in a language very similar to the mathematical state-
ment of the optimization problem.
The modeling language GAMS (General Algebraic Mod-
eling System) was developed at the World Bank to facilitate
the solution of multi-sectoral economy-wide models111 where
FORTRAN programs had been previously used. The model-
ing language AMPL (A Modeling Language for Mathemati-
cal Programming) was developed at AT&T Bell Laborato-
ries for communication applications.[21 These two languages
offer an efficient and effective way to solve mathemati-
cal programming problems at the expense of learning
another programming language. Both languages have
similar construction, and AMPL is interactive and use
separate model and data files.
GAMS appeared in 1988, is now in version 2.25, and has a
number of linear, mixed integer linear, nonlinear, and mixed
integer nonlinear solvers, including MINOS, CONOPT,
CPLEX, DICOPT, LAMPS, XA, and OSL, among others.t31
AMPL appeared in 1993 and includes the solvers MINOS,

Xueyu Chen is a PhD candidate in chemical engineering at Louisiana
State University.
Krishnaraj S. Rao received his MS degree from Louisiana State Univer-
sity in computer engineering and is currently with Intel Corporation in Palo
Alto, California.
Jufang Yu is a PhD candidate in industrial engineering at Louisiana State
University.
Ralph W. Pike is the Paul M. Horton Professor of Chemical Engineering
at Louisiana State University.
Copyright ChE Division ofASEE 1996

XA, and OSL, with others to become available.141 Both have
mainframe, workstation, and PC versions, and they have
student editions that can solve small problems (about 300
constraint equations). The manual is the same for all ver-
sions, and licensing fees are comparable.
GAMS has been used to solve chemical engineering opti-
mization problems, and Grossmannt5" has edited a CACHE
Design Case Studies Series with a number of typical prob-
lems for use in optimization courses. Also, we have used
GAMS and AMPL in research and instruction and have
found them to be valuable tools that can be used to solve a
range of optimization problems. Consequently, we offer here
a brief comparison of GAMS, AMPL, and MINOS to assist
those who would like to take advantage of this new approach
for solving mathematical programming problems.
Prior to GAMS and AMPL, codes like MINOS were used
to solve large linear and nonlinear programming problems.
MINOS (Modular In-Core, Non-Linear Optimization Sys-
tem) is a widely used nonlinear programming solver that
was developed in the System Optimization Laboratory of the
Department of Operations Research at Stanford University.
It is described as a FORTRAN-based computer system that
solves large-scale linear and nonlinear optimization prob-
lems.J61 Two files are needed to solve linear programs. One a
MPS (IBM-Mathematical Programming System) file, is re-
quired for all problems to define the names of all variables
and constraints and to specify the bounds and initial values
for variables. The other is a SPECS (Specifications) file that
sets various run-time parameters.
For nonlinear programming problems, two additional FOR-
TRAN subroutines, FUNOBJ and FUNCON, are required.
The nonlinear parts of the objective function are provided in
a FORTRAN subroutine FUNOBJ, and the nonlinear con-
straints are defined by the subroutine FUNCON. The sub-
routine FUNOBJ calculates values of the nonlinear part of
the objective function and as many gradients as possible.
The subroutine FUNCON is used to evaluate the nonlinear

Chemical Engineering Eduction

TABLE 1
GAMS Program for Problem P1

$TITLE Example Problem
* Define the variables in the optimization problem
VARIABLES X,Y;
POSITIVE VARIABLES X,Y;
*Specify the values of constants in the problem
PARAMETER CT/ value.../;
PARAMETER DT / value.../;
PARAMETER Al / value.../;
PARAMETER A2 / value.../;
PARAMETER A3 / value.../:
* Define the objective function and constraints
EQUATIONS OBJFUN
CONI
CON2;
OBJFUN.. TCOST=E=F[X] + CT*X+DT*Y;
CONI..f[X] + Al*Y=E=Bl;
CON2.. A2*X + A3*Y=E=B2
*Impose the bounds on the variables
X.UP = U;
Y.UP = U;
X.LO = L;
Y.LO =L;
*Specify the equations included by model 'Example"
MODEL Example/all/;
* Give the solve statement
SOLVE Example USING NLP MINIMIZING TCOST;
* Display the optimal solution
DISPLAY X.L, Y.L;

constraints and as many elements of the Jacobian matrix as
possible. The current version of MINOS is 5.4, which added
a callable subroutine feature to version 5.3.
GAMS 2.25 is described as a high-level language that
makes concise algebraic statements of mathematical pro-
gramming models in a language that is relatively easy to
read and write and hence is easy to understand and imple-
ment.t" Further, the advantages of GAMS over FORTRAN
solvers like MINOS are described as providing a computer
language for compact representation of large and complex
models, allowing changes to be made in model specifica-
tions simply and safely, having unambiguous statements of
algebraic relationships, and permitting model descriptions
that are independent of solution algorithms.
A GAMS program is a collection of statements in the
GAMS language. These statements consist of the sentences
that define data structures, initial values, and data modifica-
tions and of equations that provide relationships among the
variables. When problems contain matrices and vectors, sets
and indices are used to express these statements in a concise
form. The program calls on an adapted version of a solver,
such as MINOS, that is controlled by a number of default
parameters or "options" similar to the SPECS file in MINOS.

TAB

a. AMPL Model Fil

# Input the bounds for the variab
var X>=L, <=U;
var Y>=L, <=U;
# Define the names of the consta
param CT;
param DT;
param A
param A2
param A3
# Define the objective function
minimize obj: F[X] + CT*X + D
# Define the constraint equation
subject to CONI; f[X] + Al*Y
subject to CON2: A2*X + A3*

b AMPL Data File

# Input the values of the constant
param CT :=value....;
param DT :=value....;
param Al :=value....;
param A2 :=value....;
param A3 :=value....;

minimize F(x)+cTx+dT y
subjectto f(x)+AIy=b
A2x+A3y=b2
l<(x,y)

objective function
nonlinear and linear
equality constraint s
variable bounds

where the vectors (c, d, bl, b2, 1, u) and the matrices (A1, A2,
and A3) are constants, where F(x) is a smooth scalar func-
tion, and where f(x) is a vector of smooth functions.161 P1 is a
linear programming problem if x is zero. The objective
function gives a measure of the profit or cost of the operation
of a plant, and the constraint equations represent material
and energy balances, rate equations, equilibrium relations,
demand for product, availability of raw material, etc.
The GAMS and AMPL statements are given in Tables 1
and 2 for the mathematical programming problem PI with
the parameters and variables as scalars. The AMPL model
file is in Table 2a and the data file is in Table 2b. As can be
seen in Tables 1 and 2a,b, the modeling language representa-
tions are similar to the mathematical statements for problem
P1. Both start by defining variables and parameters and then
follow with the objective function and constraints. GAMS

Summer 1996

AMPL has essentially all of the
LE 2 features of GAMS but is more flex-
ible and interactive. The process and
e for Problem P1 economic models can be input in
segments; debugging and running the
bles in the optimization problem optimization can be done with the

results viewed. In GAMS, a model
ants in the problem file has to be edited, and this file is
run in a separate step.
In summary, GAMS and AMPL
modeling languages act as a bridge
between mathematical programming
of the problem problems and FORTRAN solvers for
)T*Y: problem formulation, and they can
s of the problem apply different solvers to an optimi-
=Bl zation problem. Also, both have a
resolve phase that uses bound tight-

for Problem P1 ening procedures and variable sub-
stitutions to reduce the number of
ts in the problem constraints and variables. On the
other hand, FORTRAN solvers pro-
vide experienced modelers with
more flexibility in setting run-time
parameters, which is important for
large and complicated problems.

GAMS AND AMPL STATEMENTS
OF THE OPTIMIZATION PROBLEM
Both linear and nonlinear programming problems can be
expressed in the following standard mathematical form used
by MINOS:

has the values of parameters with their definitions, and
AMPL has the values of parameters in a data file. These
programs are easy to read, and they can be checked by
people other than the modeler.
A nonlinear fuel oil allocation optimization problem
by Karimi from the CACHE compilation of GAMS
modelst15 is given in the appendix with the GAMS,
AMPL, and MINOS codes and solutions. This is a
representative illustration for the comparison of these
three methods. In the next section, results are given
for comparisons of eleven small standard engineering
optimization problems. Copies of the GAMS, AMPL,
and MINOS codes for these problems are available
by sending an e-mail request to
chepik@lsuvm.sncc.lsu.edu

COMPARISONS OF
STANDARD OPTIMIZATION PROBLEMS
A comparison was made among GAMS, AMPL, and
MINOS to evaluate their capability of solving eleven
standard engineering optimization problems. These in-
cluded two linear and nine nonlinear programming prob-
lems given by Grossmann,'51 Pike,171 Hock and
Schittkowski,151 and Schittkowski.191 A brief description
of each problem is given in Table 3, and a summary of
the optimization results is given in Table 4. The perfor-
mance of these three programs was evaluated by com-
paring the number of major and minor iterations, the
number of superbasic variables left at the optimum, and
the number of function calls.
In a major iteration of the optimization algorithm, the
nonlinear constraints are linearized at a point to give a
set of linearized constraints. A major iteration is a step
between the linearizations of the nonlinear constraints.
The minor iterations are steps of the simplex or reduced
gradient method that search for the feasible and optimal
solution based on these linearized constraints. For lin-
early constrained problems, only minor iterations take
place. For nonlinearly constrained problems, both major
and minor iterations are required, and minor iterations
take place between the successive linearizations of the
nonlinear constraints. The number of major and minor
iterations, especially for nonlinear problems, strongly
depends on the initial values and bounds on the vari-
ables, the expressions for constraint equations, and the
run-time parameters.
In the reduced gradient algorithm, the total of n vari-
ables are separated into a set of m basic variables, where
m is the number of constraints and (n-m) nonbasic or
independent variables. The superbasic variables are sub-
set of the nonbasic variables that can profitably be
changed.1"1 At the first feasible point, all nonbasic vari-
ables away from their bounds are chosen as superbasic,

PROBLEM

Refinery Scheduling
LP
9 variables
4 eq., 8 ineq. constraints

Petroleum Refinery
LP
33 variables
21 eq., 16 ineq. constraints

Fuel Allocation
NLP
8 variables
2 eq., 6 ineq. constraints

Optimization of Sulfur Content
NLP
10 variables
5 eq., 2 ineq. constraints

Alkylation Process Optimization
NLP
10 variables
7 eq. constraints

Chemical Equilibrium I
NLP
12 variables
4 eq. constraints

Chemical Equilibrium II
NLP
10 variables
3 eq. constraints

Heat Exchanger Network
Configuration NLP
15 variables
13 eq., 16 ineq. constraints

A Multi-Spindle Autom. Lathe
NLP
10 variables
I eq., 14 ineq. constraints

Optimization of Linear Objective
Function & Quad. Constraints NLP
15 variables
10 ineq. constraints

Optimization of Nonlinear Objective
Function & Quad. Constraints-NLP
7 variables
2 eq., 3 ineq. constraints

DESCRIPTION

A refinery produced gasoline, heating oil, jet fuel, and
lube oil from limited amount of 4 different crudes. The
objective was to maximize the profit per week by increas-
ing product sales and reducing the operating and purchase
costs of crude (Karimi in [5])

The objective of this simple, yet non-trivial problem was
to find the optimum operating conditions for a refinery
that maximized profit. It had three process units, each
having several input and output streams, and it had four
product streams.[7

A two-boiler turbine-generator, using a combination of
fuel oil and blast furnace gas (limited amount) was used to
produce power. The objective was to minimize the con-
sumption of fuel oil required to generate a specified amount
of power. The fuel requirements were expressed as a
quadratic function of the generated power. (Karimi in [5])

Three streams having different sulfur intents were com-
bined to form two products having specifications on the
maximum sulfur content. The objective was to maximize
profit subject to linear and bilinear product and quality
constraints.i'l

A reactor and fractionator system was used with four
feeds to produce alkylate. The objective was to maximize
a profit function that included the cost of feed and recycle
and sale of product. (Biegler in [5])

The objective was to find the equilibrium composition of
a mixture often chemical species by minimizing the Gibbs
free energy subject to elemental balance constraints. This
was done by varying the composition of the mixture to
arrive at the optimal point. (Karimi in [5])

The objective was to find the equilibrium composition by
minimizing the Gibbs free energy subject to three elemen-
tal balances. 11

The objective was to identify the minimum cost for a
utility network configuration for a specified combination
of process stream matches. (Yee and Grossmann in [5])

The optimization of a multi-spindle automatic lathe was
to minimize a nonlinear objective function subject to fif-
teen generalized polynomial constraints.

This optimization problem was to minimize a linear ob-
jective function subject to ten quadratic constraints. 1

This optimization problem was to minimize a general
nonlinear objective function subject to two quadratic and
three linear constraints. 5l

Chemical Engineering Eduction

TABLE 3
Description of Standard Optimization Problems

222

TABLE 4
Comparison of Solutions for Standard Optimization Problems
with MINOS, GAMS and AMPL

Problem

No. of Iterations Superbasic Var No of Function Obj. Function
Solver Major Minor at Opt Calls Value

Refinery Scheduling MINOS 4 -$3.4x106/WK
LP GAMS 7 -$3.4xl06/WK
9 variables AMPL 5 $3.4xl0'/wK
4 eq., 8 ineq. constraints

Petroleum Refinery MINOS 32 $702,000
LP GAMS 26 $702,000
33 variables AMPL 26 -- $702,000
21 eq., 16 ineq. constraints

Fuel Allocation MINOS 7 15 1 29 4.681 ton/hr
NLP GAMS 10 33 1 73 4.681 ton/hr
8 variables AMPL 7 15 1 47 4.681 ton/hr
2 eq., 6 ineq. constraints

Optimization of Sulfur Content MINOS 14 24 0 86 -750 units
NLP GAMS 14 27 0 70 -750 units
10 variables AMPL 14 24 0 68 -750 units
5 eq., 2 ineq. constraints

Alkylation Process Optimization MINOS 14 19 1 76 $1.154.43/day
NLP GAMS 16 131 1 750 $1,154.43/day
10 variables AMPL 13 40 1 206 $1.161.34/day
7 eq. constraints

Chemical Equilibrium I MINOS 26 7 75 -43.38
NLP GAMS 26 7 76 -43.49
12 variables AMPL 26 7 72 -43.49
4 linear eq. constraints

Chemical Equilibrium II MINOS 39 7 111 -47.76109
NLP GAMS 21 7 45 -47.76109
10 variables AMPL 31 7 90 -47.76109
3 linear eq. constraints

Heat Exchanger Network MINOS 6 8 0 180 $56,825.83
Configuration NLP GAMS 8 78 0 22 $56,825.83
15 variables AMPL 19 29 0 172 $56,825.83
13 eq., 16 ineq. constraints

A Multi-Spindle Autom. Lathe MINOS 5 24 0 116 -4,430.088
NLP GAMS 4 8 0 22 -4,430.088
10 variables AMPL 4 12 1 78 -4,430.005
1 eq., 14 ineq. constraints

Optimization of Linear Objective MINOS 12 117 13 292 -1,840.00
Function & Quad. Constraints NLP GAMS 12 200 8 339 -1,840.00
15 variables AMPL 12 119 11 296 -1,840.00
10 ineq. constraints

Optimization of Nonlinear Objective
Function & Quad. Constraints-NLP
7 variables
2 eq., 3 ineq. constraints

MINOS
GAMS
AMPL

-37.413
-37.413
-37.413

Summer 1996

and a variable will leave the superbasis
if it hits a bound or becomes basic.
During the iterations, nonbasic vari-
ables are allowed to enter the
superbasis before the beginning of
each line search, provided their re-
duced gradients are significantly
large. The number of superbasic
variables left in the solution at the
optimal point indicates the number
of nonbasic variables whose opti-
mal values are not on the bounds.
The number of function calls is the
number of times that subroutines
FUNOBJ and FUNCON have been
called to evaluate the nonlinear objec-
tive function and nonlinear con-
straints.161 The number of functions
calls to nonlinear objective and con-
straint equations is a measure of the
computational effort required to reach
the optimum.161
For the two linear programming
problems, the values of the optimum
obtained by GAMS, AMPL, and
MINOS were the same as shown in
Table 4. The only difference was in
the number of iterations that each took
to reach the optimal solution. This dif-
ference probably came from the varia-
tions of default initial values and
bounds on the variables specified by
the three programs.

As shown in Table 4, there were
differences in the number of iterations,
superbasic variables left at the opti-
mum, and function calls for the solu-
tions of the nine nonlinear problems.
For six of the nine nonlinear optimi-
zation problems, the same optimal so-
lution was located by the three meth-
ods without providing starting points.
Also, the optimal solutions were sen-
sitive to the starting points of the vari-
ables for two of the problems because
of the nonlinearities in the objective
function and constraints as described
below. These two problems proved to
be a challenge for the methods, and
typical difficulties were encountered
in obtaining the solution of nonlinear
optimization problems.
For the alkylation process optimi-

zation, the values of the objective function at the optimum
were the same for GAMS and MINOS ($1,154.43/day),
which was the same as Grossmann's151 result. But AMPL
gave a slightly better optimal value ($1,161.34/day). This
optimal solution had been reported by the original author of
the problem, Liebman, et al."12 Grossmann claimed the dif-
ference between the optimal results from his GAMS solution
and Liebman's solution was likely due to different default
tolerances in MINOS. Also, we have shown that this prob-
lem has multiple optimal solutions, and several local maxima
have been found by giving different starting points. In the
absence of a specified starting point, MINOS executed the
problem by setting the variables to zero or to a bound (if it
was specified) that was closest to zero and exited when an
optimum was located. Without good starting points for
most of the variables, MINOS was unable to reach the
final maximum objective value. But GAMS found the
optimal solution with only one variable initialized, and
AMPL was able to reach the final optimal solution with-
out the initialization of any variable.
The multi-spindle automatic lathe problem minimized a
nonlinear objective function subject to ten nonlinear con-
straints. For this optimization problem, GAMS successfully
located the global optimal solution from different starting
points, or even without specifying a starting point. MINOS
and AMPL could locate the correct global optimal solution
only when a starting point close to the global optimal solu-
tion was given. Otherwise, some sub-optimal solutions were
found. Also, when this problem was solved using GAMS
with the CONOPT solver, re-scaling of variables and con-
straints was required-otherwise the problem could not be
solved. When a starting point close to the global optimal
solution was specified for the three methods, GAMS and
MINOS found the same optimal value (-4,430.088), but
AMPL located a slightly higher value (-4,430.005). This
illustrates the need for starting points close to the optimum
and scaling of variables and constraint equations.
In Table 5, measures of the computation efficiency are
given by the total number of iterations, superbasic variables
left, and function calls for the eleven problems. MINOS took
fewer iterations and function calls than GAMS and AMPL
in total and for most problems. This may be significant for
large, complicated problems. But creating the MPS file and
FORTRAN subroutine for MINOS is time consuming and
prone to errors. These drawbacks for MINOS may supplant
its advantage. For example, some of these optimization prob-
lems were assigned to students for homework in an optimi-
zation course. A few students solved the problems using
MINOS in the time allotted, while all found optimal solu-
tions by AMPL and GAMS. Also, they reported that GAMS
and AMPL were easier to use than MINOS when starting
with no experience with these methods.
All of the problems required well-scaled variables and

TABLE 5
Comparison of the Computation Efficiency for Eleven
Optimization Problems with MINOS, GAMS, and AMPL

Total of major Total of minor Total of superbasic Total
iterations iterations variables left function calls
MINOS 62 317 32 1011
GAMS 75 610 -27 1593
AMPL 81 377 31 1255

constraint equations. Scaling is performed by multiplying
factors to have the variables and constraints close to a
magnitude of one.J' Scaling is key to obtaining optimal
solutions for problems with widely varying values of the
variables and constraint equations. The users manuals
describe procedures for scaling.

SUMMARY
Programming and solving standard optimization problems
showed that GAMS, AMPL, and MINOS are all effective,
and they release modelers from programming optimization
algorithms. The comparisons showed that optimization prob-
lems are relatively easy to program in GAMS and AMPL,
and they offer a choice of solvers and have a resolve phase
to reduce model size. In addition, AMPL has features of
separate model and data files, flexible output, and options to
run batch operations. GAMS provides a comprehensive out-
put summary that is very helpful in detecting model errors,
and it is interfaced with more solvers than AMPL now.
MINOS could be more robust than GAMS and AMPL, but
programming is more difficult. In addition, this is an active
area for developments; Floudas describes MINOPT,1"31 an
automated mixed-integer nonliner optimizer. Also, GAMS
has been extended to use the APROS technique to connect
the NLP and MILP in the decomposition of MINLP (Paules
and Floudas in [5]).

REFERENCES
1. Brooke, A., D. Kendrick, and A. Meeraus, GAMS: A User's
Guide, Release 2.25, The Scientific Press, San Francisco,
CA (1992)
2. Fourer, R., D.M. Gay, and B.W. Kernighan, AMPL: A Mod-
eling Language for Mathematical Programming, The Scien-
tific Press, San Francisco, CA (1993)
3. Meeraus, A., General Algebraic Modeling System, GAMS
Development Corp., Washington, DC (1994)
4. Kernigham, B.W., personal communication (1994)
5. Grossmann, I.E., Ed., Chemical Engineering Optimization
Models with GAMS: CACHE Process Design Case Studies
Series, CACHE Corp., Austin, TX (1991)
6. Murtagh, B.A., and M.A. Saunders, MINOS 5.4 User's Guide,
Technical Report SOL 83-20R, Systems Optimization Labo-
ratory, Department of Operations Research, Stanford Uni-
versity, Stanford, CA (1993)
7. Pike, R.W., Optimization for Engineering Systems, Van
Nostrand Reinhold Company, Inc., New York, NY (1986)
8. Hock, W., and K. Schittkowski, Test Examples for Nonlin-
ear Programming Codes, Springer-Verlag, New York, NY

Chemical Engineering Eduction

(1981)
9. Schittkowski, K., More Test Examples for Nonlinear Pro-
gramming Codes, Springer-Verlag, New York, NY (1987)
10. Floudas, C.A., and I.E. Grossmann, "Algorithmic Approaches
to Process Synthesis: Logic and Global Optimization," Fourth
Int. Conf. on Founds. of Computer-Aided Prog. Design,
CACHE, American Institute of Chemical Engineers, New
York, NY (1995)
11. Drud, A., "CONOPT: A GRG Code for Large Sparse Dy-
namic Nonlinear Optimization Problems," Math. Program-
ming, 31, 153 (1985)
12. Liebman, J., L. Lasdon, L. Schrage, and A. Waren, Model-
ing and Optimization with GINO, Scientific Press, Palo
Alto, CA (1984)
13. Floudas, C.A. Nonlinear and Mixed-Integer Optimization,
Oxford University Press, New York, NY (1995)

APPENDIX

A FUEL ALLOCATION OPTIMIZATION PROBLEM

This is a simple, nonlinear, allocation optimization given
in the CACHE compilation of GAMS models by Karimi.15'
The problem statement has a two-boiler, turbine-generator
combination producing a minimum power output of 50 MW,
as shown in Figure 1 (next page). Fuel oil and blast furnace
gas (BFG) are to be used, and 10 fuel units per hour of BFG
are available. A minimum amount of fuel oil is to be pur-
chased to produce the required power from the two genera-
tors. The amount of fuel used, F, in tons per hour for fuel oil

TABLE Al

() GAMS Code for Fuel Allocation Optimizationt51
$TITLE Power Generation via Fuel Oil

* Define index sets
SETS G Power Generators /genl*gen2/
F Fuels/oil,gas/
K Constants in Fuel Consumption Equations/0*2/;

*Define and Input the Problem Data
TABLE A(G,F,K) Coefficients in the fuel consumption equations
0 1 2
genl.oil 1.4609 .15186 .00145
genl.gas 1.5742 .16310 .001358
gen2.oil 0.8008 .20310 .000916
gen2.gas 0.7266 .22560 .000778;
PARAMETER PMAX(G) Maximum power outputs of generators
/GEN1 30.0, GEN2 25.0/;
PARAMETER PMIN(G) Minimum power outputs of generators
/GEN1 18.0, GEN2 14.0/;
SCALAR GASSUP Maximum supply of BFG in units per h
/10.0/
PREQ Total power output required in MW
/50.0/;
*Design optimization variables
VARIABLES P(G) Total power output of generators in MW
X(G,F) Power outputs of generators from specific fuels
Z(F) Total Amounts of fuel purchased
OILPUR Total amount of fuel oil purchased;
POSITIVE VARIABLES P, X, Z;
* Define Objective Function and Constraints
EQUATIONS TPOWER Required power must be generated
PWR(G) Power generated by individual generators
OILUSE amount of oil purchased to be minimized
FUELUSE(F) Fuel usage must not exceed purchase;

TPOWER..
PWR(G)..
FUELUSE(F)..
OILUSE..

SUM(G, P(G))=G=PREQ;
P(G)=E=SUM(F, X(G,F));
Z(F)=G=SUM((K,G),a(G,F,K)*X(G,F)**(ORD(K)-1));
OILPUR=E=Z("OIL"):

* Impose Bounds and Initialize Optimization Variables
* Upper and lower bounds on P from the operating ranges
P.UP(G) = PMAX(G);

P.LO(G) = PMIN(G);
*Upper bound on BFG consumption from GASSUP
Z.UP("gas") = GASSUP;
* Specify initial values for power outputs
P.L(G)=.5*(PMAX(G)+PMIN(G));
* Define model and solve
MODEL FUELOIL/all/;
SOLVE FUELOIL USING NLP MINIMIZING OILPUR;
DISPLAY X.L, P.L, Z.L, OILPUR.L;

GAMS Solution for Fuel Allocation Optimization
MODEL STATISTICS

BLOCKS OF EQUATIONS
BLOCKS OF VARIABLES
NON ZERO ELEMENTS
DERIVATIVE POOL
CODE LENGTH
GENERATION TIME
EXECUTION TIME

4 SINGLE EQUATIONS
4 SINGLE VARIABLES
16 NON LINEAR N-Z
5 CONSTANT POOL
81
=0.220 SECONDS
-0.280 SECONDS VERID MW2-00-051

SOLVE SUMMARY

MODEL FUEL OIL
TYPE NLP
SOLVER MINOS5
**** SOLVER STATUS
**** MODEL STATUS
**** OBJECTIVE VALUE

OBJECTIVE OILPUR
DIRECTION MINIMIZE
FROM LINE 54
1 NORMAL COMPLETION
2 LOCALLY OPTIMAL
4.6809

EXIT OPTIMAL SOLUTION FOUND
MAJOR ITNS, LIMIT 10 200
FUNOBJ, FUNCON CALLS 0 73
SUPERBASICS 1
INTERPRETER USAGE 0.00
NORM RG / NORM PI 2.532E-10
VARIABLE X.L Power outputs of generators from specific fuels
OIL GAS
GENI 10.114 19.886
GEN2 3.561 16.439
VARIABLE P.L Total power output of generators in MW
GEN1 30.000, GEN2 20.000
VARIABLE Z.L Total Amounts of fuel purchased
OIL 4.681, GAS 10.000
VARIABLE OILPUR.L = 4.6809 Total amount of fuel oil purchased

Summer 1996

TABLE A2

) AMPL Model file for Fuel Allocation Optimization

set G;
set F;
set K;

param COEFF{G, F, K} >=0;
param PMAX (g in G)};
paramPMIN (ginG);
param J {k in K};

var P(g in G} >=PMIN[g],,<=PMAX[g);
varX{ginG,finF} >=0;
var Z{(finF) >-0;

minimize purch_oil{f in F): Z["oil"];
subject to TPWR: sum (g in G} P[g]>=50;
subject to PWR {g in G): sum (fin F) X[g,f]=P[g];
subject to FUELUSE {f in F): sum (k in K, g in G) COEFF[g, f, k]*X[g, f]**J[k]=Z[f];
subject to BFG (fin F): Z["gas") <=10;

S() AMPL Data file for Fuel Allocation Optimization

setG:=genl gen2;
set F:=oil, gas;
set K:-0, 1,2;

param COEFF:=

[genl,*,*]: 0
oil 1.4609
gas 1.5742
[gen2, *,]: 0
oil 0.8008
gas 0.7266

1
0.15186
0.16310
1
0.20310
0.22560

2 :=
0.001450
0.001358
2 :=
0.000916
0.000778

param: PMAX PMIN:=
genl 30 18
gen2 25 14;
param: J:=
0 0
1 1

I AMPL Solution for Fuel Allocation Optimization

MINOS 5.4:

EXIT-optimal solution found

No. of interations 15
No. of major interations 7
Penalty parameter .0
No. of calls to funobj 0
No. of superbasics 1 N
No of basic nonlinears 3

P[*] :=genl 30
X :=genl gas 19.8857
gen2 gas 16.4388
Z[*]:= gas 10

Objective value
Linear objective
00100 Nonlinear objective
No. of calls to funcon
orm of reduced gradient
Norm rg / Norm pi

4.6808895430E+00
4.6808895430E+00
0.0000000000E+00
47
1.350E-08
9.610E-09

gen2 20;
genl oil 10.1143
gen2 oil 3.56123;
oil 4.68089;

ru oilGenerator1
Fuel Oil -*
S50MW
Blast Furnace
Gas (BFG) ---- 1
Gas (BG) Generator 2

Generator Fuel Type a, a a
I Fueloil 1.4609 0.15186 0.001450
1 BFG 1.5742 0.16310 0.001358
2 Fueloil 0.8008 0.20310 0.000916
2 BFG 0.7266 0.22560 0.000778

Figure 1. Diagram and parameters for fuel
allocation optimization."s

or units per hour for BFG is a quadratic function of the
power produced, X, in MW, i.e.,

F=ao + aX + a2X2

where the regression parameters ao, at, and a2 are listed in
Figure 1 for the two fuels and the two generators. Also, the
ranges of operation for generators one and two are (18, 30)
MW and (14, 25) MW respectively.

The optimal solution will determine the minimum
amount of fuel oil to be purchased and its distribution
between the two generators. If Fi is the amount of fuel
type j (j=l for fuel oil and j=2 for BFG) used by genera-
tor i (i=1,2), then Xii is the corresponding power gener-
ated. If Zt is the total amount of fuel oil purchased for the
two generators, Z2 is the total usage of BFG for the two
generators, and P, is the power generated by generator I,
then the problem can be stated as:

Minimize: Z1
2
Subjectto: Xaijo+aijlXij +aij2Xi

Xil+Xi2-Pi =0 fori=1,2
PI +P2 50
0 18 14

This problem has eight variables and two equality and six
inequality constraint equations.

The input files for this problem in GAMS, AMPL, and
MINOS are given in Tables Al, A2, and A3. The model
statements are similar in GAMS and AMPL, and AMPL
has separate model and data files. But the files for MINOS
are more complicated, as shown in Table A3a,b, the
MINOS MPS and SPC files. The output files are given in
Table Alb for GAMS, Table A2c for AMPL, and Table
A3d for MINOS, and all three found the same optimal
fuel allocation.
Chemical Engineering Eduction

TABLE A3

( MINOS MPS File for Fuel Allocation Optimization

NAME FUELOIL
ROWS
L OILAMT
L GAS AMT
E GENTI
E GENT2
G POWER
N PUR_OIL
COLUMNS
X11 GENTI 1.0
X12 GENT1 1.0
X21 GENT 2 1.0
X22 GENT2 1.0
Zl PUROIL 1.0
Z2
P1 GENT1 -1.0 POWER 1.0
P2 GENT2 -1.0 POWER 1.0

DEMAND POWER 50.0
UPBOUND01 Z2
UPBOUND01 P1
LOBOUND01 PI
UPBOUND01 P2
LOBOUND01 P2
FR INITIAL PI
FR INITIAL P2
ENDATA

MINOS SPC (Specifications) File for Fuel Allocation
Optimization

BEGIN FUEL OIL (NLP problem)
*
* To Minimize the Consumption of Fuel Oil for Fuel Oil Allocation

Problem Number 11
Minimize
Rows 20
Columns 30
Elements 50

MPS file 10

Print level 1 *0
Print frequency I
Summary frequency 1

Nonlinear constraints
Nonlinear Jacobian Var
Nonlinear Objective Var

)K for small problems

2
6
0

Scale Option 2
END FUELOIL PROBLEM

@ Funcon Subroutines for Fuel Allocation Optimization

PROGRAM MINOS
IMPLICIT DOUBLE PRECISION (A-H, O-Z)
PARAMETER (NWCORE=30000)
DOUBLE PRECISION Z(30000)
CALL MINOS 1(Z,NWCORE)
END

SUBROUTINE TCON (MODE, M, N, NJAC, X, F, G, STATE, NPROB, Z, NWCORE)
IMPLICIT DOUBLE PRECISION (A-H, O-Z)
DOUBLE PRECISION X(N), F(M), G(M,N), Z(NWCORE)
COMMON /M1FILE/IREAD, PRINT, ISUMM
COMMON /M8DIFF/DIFINT(2),GDUMMY,LDERIV,LVLDIF,KNOWNG(2)
F(l)=1.4609 + (0.15186*X(1))+ (0.001450*(X(1)**2))
+ + 0.8008 + (0.20310*X(3)) + (0.000916*(X(3)**2)) X(5)
F(2) = 1.5742 + (0.16310*X(2)) + (0.001358*(X(2)**2))
+ + 0.7266 + (0.22560*X(4)) + (0.000778*(X(4)**2)) X(6)
G(1,1) = 0.15186 + (2.0*(0.001450)*X(1))
G(1,3) =0.20310+( l 'i iiar' \i .'1 1
G(1,5) =-1.0
G(2,2) = 0.16310 + (2.0*(0.001358)*X(2))
G(2,4) =0.22560 + ini"' \ ,
G(2,6) = -1.0
RETURN
END

O MINOS Solution for Fuel Allocation Optimization

EXIT optimal solution found

FUELOIL
No. of iterations 15 Objection value 4.6808896266E+00
No of major interations 7 Linear objective 4.6808896266E+00
Penalty parameter .00100 Nonlinear objective 0.OOOOOOOOOOE+00
No. of calls to funobj 0 No. of calls to funcon 29
No. of superbasics I Norm of reduced gradient 9.160E-07
No. of basic nonlinears 4 Norm rg / Norm pi 9.176E-08
No. of degenerate steps 0 Percentage .00
Norm of x (scaled) 3.148E+00 Norm of pi (scaled) 9.983E+00

COLUMN STATE ACTIVITY OBJ GRADIENT LOWER LIMIT UPPER LIMIT REDUCED GRADNT

10.11428
19.88572
3.56123
16.43877
4.68089
10.00000
30.00000
20.00000

.00000
.00000
.00000
.00000
1.00000
.00000
.00000
.00000

.00000
.00000
.00000
.00000
.00000
.00000
18.00000
14.00000

NONE
NONE
NONE
NONE
NONE
10.00000
30.00000
25.00000

.00000
.00000
.00000
.00000
.00000
-.83456
-.02843
.00000

Summer 1996

r

r, M classroom

PROBLEM-CENTERED TEACHING

OF

PROCESS CONTROL AND DYNAMICS

PAUL LANT, BOB NEWELL
The University of Queensland Queensland 4072, Australia

It has been our experience that undergraduate process
engineering students generally find dynamics and pro-
cess control conceptually difficult, perceive it as periph-
eral, find it difficult to integrate into their degree program,
and as such, tend to find it more of a chore than fun to learn!
In this paper we will introduce a new, problem-based
approach to teaching undergraduate dynamics and control
and will emphasize its effectiveness in integrating this ma-
terial into the process engineering curriculum. We also
hope to convey our enthusiasm for this approach, which
we have found to be tremendously rewarding for both
lecturer and tutors.
The subject introduces the dynamics and control of pro-
cesses by performing a series of exercises and design studies
on a selected process flowsheet covering basic instrumenta-
tion, synthesis of control schemes, modeling and simulation
of process units, feedback (PID) and feedforward controller
design, and discrete event control systems. The approach
places a greater emphasis on creativity in the areas of control
system synthesis and design. The students clearly acquired
greater confidence and competence than they did in previous
years. Student feedback was dominated by concerns about
group dynamics, and it is evident that group dynamics has a
significant impact on student learning. This is a difficult
problem to overcome, as problem-based learning inherently
requires group work and group interaction.
Experiences, observations, and difficulties encountered in
the introduction of this approach will be highlighted in this
paper, with modifications and recommendations suggested.

OUR "PROBLEM"
Dynamics and control is a compulsory subject taught in
the third year of the chemical engineering, environmental
engineering, and mineral processing degree programs at The
University of Queensland. The subject was scheduled for

three contact sessions per week (5 hours) for a 13-week
semester. The student workload (including class time) should
be approximately ten hours per week.
Upon completion of the course, students should be able to
Describe the architecture, components, and cost of instrumen-
tation and control systems
Synthesize control structures for process flowsheets
Develop mechanistic models of and simulate, relatively simple
unit operations
Design simple feedback controllers and feedforward compen-
sators
Design discrete event control systems
In previous years, the subject was taught in discrete
modules, consistent with the above description. Each
module was evaluated by the use of assignments (indi-
vidual) and quizzes. All contact hours were with the
whole class, as either conventional lecture or tutorial
sessions (where the lecturer presents the problem and
then works through its solution).

Paul Lant is a Lecturer in Chemical Engineering
at The University of Queensland. He received a
MEng in Chemical and Process Engineering
(1987) and a PhD (1991) from the University of
Newcastle upon Tyne (England). Current research
interests include modeling and control of biologi-
cal wastewater treatment processes and struc-
tural controllability.

R.B. Newell is a Senior Lecturer in Chemical
Engineering at The University of Queensland.
He received his PhD from the University of
Alberta. He also has a Dip. Ed in Tertiary Educa-
tion from Monash University. Current interests
include modeling and control of waste treatment
processes, combined fuzzy and deterministic
control, and the development of software tools.

Copyright ChE Division ofASEE 1996
Chemical Engineering Eduction

[No

While this approach was partially successful in achieving
the subject objectives, our observations were that
Students failed to integrate modules. Once a module was
finished, it was forgotten.
Small problems (individual assignments) failed to integrate the
subject within their course (mineral processing, environmental
or chemical engineering).
The structured nature of the assessment tended to prevent
students pursuing their own problems.
There was a clear lack of i.,i. in dynamic process
modeling.
In summary, students find this material conceptually diffi-
cult and generally fail to recognize how dynamics and con-
trol relates to their other core subjects; as such, this subject is
not normally seen as an enjoyable experience. Furthermore,
as the trend is for larger, more heterogeneous classes, it was
clear that the teaching strategy for the subject required revi-
sion. As a consequence, the approach outlined in this paper
was introduced in the second semester of 1994.

OUR STRATEGY
In an attempt to better integrate this subject into the pro-
cess engineering curriculum, we decided to revolve the sub-
ject around a single process and to base the student learning
on problems associated with that process. That is, the stu-
dent learning was to be problem driven and learner centered.
The subject matter, in terms of the objectives and material,
was unchanged. The class (65 students) was split into groups
of four or five, and the groups were divided into the three
engineering disciplines-mineral processing, environmental
engineering, and chemical engineering. The objectives were
addressed by assigning a series of major tasks for each
group. The problems were stated so as to move the students
through the several stages of control structure synthesis and
control system design. The problems were integrated via the
one process, with each group selecting their own process.
Each group operated as a 'consulting' company and was
required to cost their time. This was an indirect way of
assessing and controlling student workload.
The class was scheduled for three contact sessions per
week (Monday, Wednesday, and Friday). Since our objec-
tive was to orient the teaching around the problems, each
week we intended to introduce and discuss concepts and
analytical tools that the students were at that time trying to
use for their process.
The Monday class was essentially a lecture (to the entire
class) that attempted to provide the students with the tools
they would need to progress with their problem. Small ex-
ample problems were used for demonstration.
The second period each week (Wednesday) was the most
critical contact time. The class was divided into three smaller
classes (consisting of four groups each). Our objective in

these sessions was to assist the groups in implementing the
material (that had been presented on Monday) for their spe-
cific process problem; For ease of discussion, this session
will be called the tutorial session.
The tutorials were facilitated by postgraduate students and
were tightly structured. While we are aware that this is not
ideal for an orthodox problem-based course, it was neces-
sary due to time and resource constraints. The lecturer and
tutors met prior to each session.
On average, the tutorial sessions began with a short review
of the lecture material and proceeded to outlining what was
required within the session. Because each group was study-
ing a different process, it was important for the groups to
present their work to the other groups-this was an impor-
tant part of the learning process. Marks were not allocated
for tutorial attendance, but attendance was high (90-95%).
The Friday period was used for a 'standard' lecture to the
whole class. The aim of this session was to review the
work performed in the tutorial session and to address
specific problems and questions raised by the students.
Due to time constraints, this session was sometimes used
for additional lectures.

FOUR WEEKS IN THE LIFE ...
We must admit to feeling somewhat challenged to ad-
equately describe the experiences and feelings of students in
this class. We will attempt to guide you through the first four
weeks of the subject-our objectives, and the students reac-
tions to lectures, tutorials, and problems.

Week 1
Lecture Hello!
Introduce resource materials. The major resources
used were a subject study guide, a process control
textbook (Seborg, et al.[ 1]), a MATLAB software
package, and a PID controller tuning experiment.
Clarify the approach to teaching the subject. Why are
we teaching in this way?
Students are separated into groups of 4-5 and in-
structed to "select a process" to study. The only
guidance provided was that there should be approxi-
mately 10-20 units, multiple phase unit operations, and
recirculating inventories.
Tutorial No formal tutorial session. All groups are invited to
meet their tutor and discuss process selection
Reaction Students tend to display a lot of interest in this first
week. They are confronted with a different approach
for learning, and most are genuinely supportive.
Most groups will have no difficulty in selecting a
process.

Week 2
Lecture Subproblem 1 is handed out (see the Appendix)
Introduction to mass and energy inventory control (the
basic tools for addressing subproblem 1).

Summer 1996

Tutorial Each group presents their process to the rest of their
tutorial class (each class consists of four student
groups). We strive to emphasize the importance of
understanding their process at this early stage.
Reaction At this stage, the students are starting to feel a little
concerned-they have a problem that they do not
entirely understand, and they feel frustrated.

Week 3
Lecture The lecturer demonstrates mass and energy inventory
control loop pairing through several examples of unit
operations.
Tutorial Each group presents a control system design for one
unit on their flowsheet.
Reaction Panic! The report is due in one week; the students can
now define the problem and realize what is required.

Week 4
Lecture No formal lecture.
No formal tutorial, although the students are encour-
aged to privately consult with their tutor.
Reaction The first report is submitted.

The subject is taught via four subproblems. Table 1 sum-
marizes each problem in terms of our objectives. An ammo-
nium nitrate process is employed to provide an example of
specific outcomes for each problem (see the Appendix).
The problems are the major form of assessment (group
reports). A system of peer assessment was adopted for the
problems.12 Upon submission of a group report, each student
was required to assess the effort of his or her colleagues via
an assessment form that was handed out to the students (see
Table 2). The responses for each group are compiled and an

average-effort rating for the group is obtained. Each indi-
vidual mark is then obtained by
Individual mark = Group mark (Individual effort rating/Group effort rating)
We also included two pieces of individual assessment: a
quiz on dynamic modeling and a final examination. The
reasons for doing this were to reduce student concerns over
the peer assessment, to address our concerns about our
ability to assess students via group projects and peer assess-
ment, and to enable a comparison of performance with
previous years.
The group project was the major focus, however, and the
quiz and examination were restricted to assessing individual
understanding of the group-project activities.

IS THIS AN EFFECTIVE APPROACH FOR
TEACHING DYNAMICS/PROCESS CONTROL?
Formal subject evaluation, via student questionnaires, was
performed by The University of Queensland Tertiary Educa-
tional Institute. The subject ratings (1-7; 7 high) for 1994
and 1995 were 4.9 and 4.6. The ratings for the previous
years, prior to the subject change, were 5.2 and 4.7, respec-
tively. Student feedback was dominated by group dynamics;
an important outcome in itself. A summary of students com-
ments follows.

"Group projects are an excellent idea. However, there is a
problem with some people who do not pull their weight. "
"Group work sucks-in industry if you don't work properly
you get fired. At Uni if you don't work properly, everyone gets
shafted!"

TABLE 1
Objectives for Each Problem and Example of Resulting Outcome

Our Objectives
Develop an understanding of the process *
Synthesize a control system structure Develop
an appreciation of control system architecture *
Determine basic instrumentation costs Prepare a
P&I diagram Develop an appreciation of the
interaction between design and control Develop
project management skills.

Dynamic model synthesis Linearization of
nonlinear model Perform step-test identification
* Dynamic simulation Perform sensitivity
analysis.

Subproblem 1
.-. Control Structure
Synthesis

Subproblem 2
4 Dynamic Modeling and
Simulation of One Unit

Design and tune PID controllers Design a (static Subproblem 3
and dynamic) feedforward compensator Analyze -> 'Simple" Controller
control system performance. Design
I

Develop an understanding of discrete event
control strategies as opposed to all previous work, ->
which was on a continuous process.

Ammonium Nitrate Process Outcomes
22 control loops were specified to control the mass and energy
inventories DCS architecture was recommended Quality control
was specified for the ammonium nitrate product and both waste
-> streams P&I diagram showing basic control loops with sensors
and actuators Preliminary control system costing Discussion of
design/control interaction.

The loop reactor was modeled as a CSTR and evaporative separator
in series. The model consisted of 10 ODEs and 20 algebraic
equations. The reactor was simulated in MATLAB, with step
responses and sensitivity analyses performed. The effect of various
design options was also investigated.

P, PI, and PID controllers were evaluated for reactor temperature
-> and pressure control. Yuwana-Seborg, ISE and ITAE tuning
formulae were investigated. A feedforward regulator was
implemented for nitric acid feed flow disturbances.

Subproblem 4 GRAFCET diagram for the start-up and shut-down of the reactor.
Discrete Event
Systems

S30 Chemical Engineering Eduction

"Include more control practicals"
"Group work is very frustrating!"
"Flowsheets should be selected to be of equal difficulty."
"Group work was very difficult when you have one dominant
group member. I suppose it comes down to group dynamics
and my problem of not talking about my problems with other
group members."
"Make groups have a maximum of 4."
"Provide more support for groups struggling with their
models."
It is clear that working in groups polarized student opin-
ion. When teaching this subject the second time (second
semester 1995), we placed more emphasis on group dynam-
ics and introduced the students to the problems experienced
in the previous year in the naive hope that they might learn
from previous mistakes. Figure 1 clearly illustrates that this
was far from successful. This is a difficult obstacle to over-
come, as problem-based learning inherently requires group
work and group interaction. We have yet to resolve this
problem satisfactorily.
It is also apparent from the feedback that some students

TABLE 2
Assessment Form

NAME: Paul Lt GROUP MEMBERS
Bob Marc Lisa
Project mgt. and organization 2 5 4
Writing & compiling report 2 4 4
Data gathering and lit. survey 2 5 5
TOTAL (out of 15) 6 14 13

Minimal Satisfactory Outstanding
Contribution Contribution Contribution
1 3 5
3 5

Figure 1. Questionnaire response to the statement, "I
enjoyed doing the group project."
Sunmmer 1996

were uncomfortable with the open-ended nature of the sub-
ject and had gained little appreciation of why we adopted a
problem-based approach:
"Don't be so slack . use more of the lecture time available
to teach us."
"Do not be so lazy. If you are allocated lecture times, use
them!"
"When you are teaching things to people for the first time, they
have to be explained very thoroughly."
It is a sad reflection on our broader educational system that
intelligent, 20-year old, engineering undergraduates are un-
comfortable with ill-defined problems, threatened by some-
thing new, and fail to accept responsibility for their own
learning. If anything, this fortifies our belief in this ap-
proach. But it is clear that we need to expend more effort in
gradually introducing the students to the subject.
How well does this approach address the driving forces for
change? We shall address each in turn.
To what extent did this approach integrate dynamics and
control into the degree program? This was the single
most important aspect of this subject formulation. Students
were forced to think about dynamics and control within the
framework of the whole process. It was incredibly reward-
ing to see students actively considering control and design
issues simultaneously.
Did the subject address the different demands of different
groups of students? The group cases enabled students to
learn by employing control and modeling skills on a process
of direct interest to them. The processes investigated were
extremely varied and included:
Mineral Processing Groups: Updraught lead sintering Lead-
zinc concentrator Lead concentrator
Environmental Engineering Groups: SO,/NOx Flue Gas
Cleanup Wastewater treatment Combined cycle power gen-
eration Brewing
Chemical Engineering Groups: Ammonium nitrate Whey
fermentation to ethanol Formaldehyde Carbon tetrachloride *
Sugar milling
We believe that the scope of the problems investigated would
only be achieved by adopting this type of problem-based
approach.
Is this class more competent, and confident, with process
control and dynamics? The work submitted was of a very
high standard (for what were 'average' classes). Significant
improvement over previous years was observed. The moti-
vation and commitment of the students was high, as re-
flected in the tutorial attendance and well-presented reports.
Tutorial attendance was not compulsory, and yet was in
excess of 90%.

CONCLUSION
While it is always difficult to obtain an absolute measure
231

50
45
40
35
30
BS 25
20
15
10
5
0

a, a

, owf
C0 C O)
W 3

B1995|

1 1994

2o< <
U)

of 'improvement' in a subject (due to the lack of a control),
we are confident that this approach serves to emphasize and
enhance key process control skills. The problem-based ap-
proach to teaching dynamics and control presents students
with a real, yet ill-defined, challenge. Creative skills, such as
design and synthesis, are emphasized. Furthermore, it is also
amenable to larger, more heterogeneous classes, which ap-
pears to be an inevitable trend in Australia.
For anyone interested in using this approach, we offer
several recommendations for consideration:
Restrict groups to 3 or 4 people.
Do not underestimate the negative effect of group
dysfunctionality. As such, it is critical to pay significant
attention to group dynamics and project management (review
and discussion sessions during the semester).
Use mixed tutorial sessions to encourage interaction. In our
case, we mixed mineral processing, environmental engineer-
ing, and chemical engineering groups in one tutorial group.
Dynamic model synthesis and simulation tends to be a
difficult conceptual step for most students. It is important,
therefore, that this particular subproblem be tightly con-
trolled by the lecturer and tutors.
Do not attempt to use this approach without adequate
resources-in particular, sufficient good tutors. The role of
the tutors cannot be understated. It is important that they are
aware of their role and that they are competent of facilitating
and guiding their groups through the subject. Should the
tutors be 'experts' in the field? This question has raised
significant debate in the broader field of problem-based
learning. But when faced with tight time and resource
constraints (we cannot afford to have a ratio of one tutor per
group of four students), we believe that expert tutors are a
necessity.

Finally, while we must admit that the open-ended n
of the problems provides lecturer and tutors with more
lenges and is unquestionably more resource
intensive, our brief experience indicates that it
is a more rewarding and fun approach for teach-
ing dynamics and process control.

ACKNOWLEDGMENTS
We would like to thank Professor Peter Lee
for his input into the planning of this subject,
and also the 'guinea-pig' postgraduate tutors
whose lives we severely interfered with for
four months; a special thank you to Marc
Steffens, Ian Ramsay, Andrew Schroder, Lisa
Hopkins, and Damien Batstone.

REFERENCES
1. Seborg, D.E., T.F. Edgar, and D.A.
Mellichamp, Process Dynamics and Control,
Wiley & Sons, Brisbane (1991)
2. Conway, R., A. Kember, A. Sivan, and M.
Wu, "Peer Assessment of an Individual's Con-

nature
chal-

tribution to the Group Project," Assess. and Eval. in Higher
Ed., 18(1), 45 (1993)

APPENDIX)

Example Problem
The Stamicarbon process for the manufacture of ammonium
nitrate is representative of the size and complexity of the problems
chosen (see Figure 2).
Subproblem 1 Your group is to act as a consultant to Multinat
Pty Ltd. Multinat is the contractor responsible for designing and
constructing PROCESS. Multinat has subcontracted the process
control system design to you. Multinat is performing the project
management.
In order to coordinate all subcontractors, Multinat requires the
following information in your report: number and type of control
loops; instrumentation (sensing elements, controllers, and final con-
trol elements); and costing.
Multinat is not familiar with process control. It is, therefore,
imperative that you can justify your recommendations. Your report
must include a description of the process, with particular emphasis
on the process operating objectives and constraints (what are they?).
This initial contract with Multinat is worth $10,000. It costs your
organization $100/hour for labor (it is important that you accu-
rately record, and cost, your time). That is, each 1-hour meeting of
your team of 4 people costs $400. It is, therefore, important that
each meeting is efficient, with tasks clearly defined and allocated.
You must identify what the tasks are, who will perform them, and
by when (an action plan). You should include a memo to your
manager stating the cost of the study.
You are aware that Multinat will require further control work to
be performed on this project. The objective for your project team,
therefore, is to generate a report good enough to win future con-
tracts, while also maximizing the profit to your company. Do not
miss any opportunity to impress Multinat. Comment on any areas
where design modifications may be beneficial. Offer alternatives
when possible. O

I.5B PRIMARY ---- I

Figure 2. Ammonium Nitrate Process (A selected case study.)
Chemical Engineering Eduction

LOOP REACTOR SEPARATOR CONDENSER AM4ONIA SCRBBER

VAPOUR

R INTERMEDIATE
STORAGE
.//--, \. / C.W

on, '""g" "oTH

AUTHOR GUIDELINES

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.

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 consecutively.

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 biographi-
cal 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

We've changed our name,

but not our commitment

Mobil Research and Development Corporation has
changed its name to Mobil Technology Company (MTC).
This is more than just a name change, however. It is part
of a corporate-wide reorganization and consolidation of
research, development and engineering activities into one
division reporting to Mobil's Executive Committee. MTC is
lead by Mike Ramage, who is both its President and the
Chief Technology Officer of Mobil Corporation. The eleva-
tion of technology at Mobil reflects its long-term strategic
importance to our company's future.
"Our new organization unites Mobil's technology
functions from research and development to application
and deployment," said Dr. Ramage. "With many different
disciplines working together, they will generate the techno-
logical synergy we need to boost Mobil's future growth."
Downstream R&D is now located in Paulsboro, New
Jersey at the Mobil Refining and Chemical Technology
Center. All upstream R&D is centered in Dallas, Texas at
the Mobil Exploration and Producing Technology Center.
oaas Technical service, engineering design and construction
activities are located in both Paulsboro and Dallas. A new
Strategic Research Center, with personnel assigned to
both sites, has been formed to consolidate long-range
and exploratory research activities for all of Mobil's busi-
nesses.
The new MTC brings its core strengths and a wide
range of technical services to bear on all of Mobil's busi-
ness opportunities. This close alignment of technology
with new and existing business initiatives will aid rapid and
efficient technology transfer. Mobil remains committed to
research for the long term and will build strategic alliances
with universities, National Laboratories and other techno-
logically strong companies.
While we have streamlined our organization, we have
maintained our core strengths in catalysis, lubrication,
sboro seismology, geology and engineering (reaction, process,
facilities and reservoir). We will remain a world-class sci-
ence and technology company.

Mobil. Technology Company
3225 Gallows Road, Fairfax, Virginia 22037

Mobil Corporation 1996

http://www.mobil.com

	Front Cover
	Table of Contents
	Brice Carnahan and James O. Wilkes...
	Evolution for chemical enginee...
	Book review
	Industry, academe, and government:...
	"An Ode to that Distillation Tower"...
	Stirred pots
	The chemical engineering curriculum...
	If you've got it, flaunt it: Uses...
	Teaching colloid and surface phenomena...
	Integrating new separations technologies...
	Implementation of multiple interrelated...
	Wake-up to engineering!
	ChE applications of elliptic...
	Comparison of GAMS, AMPL, and MINOS...
	Problem-centered teaching of process...
	Back Cover

MILLISECOND	CLASS.METHOD	MESSAGE
0	sobekcm_page_globals.constructor
0	sobekcm_page_globals.constructor	Application State validated or built
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.constructor	Navigation Object created from URI query string
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.display_item	Retrieving item or group information
0	sobekcm_page_globals.get_entire_collection_hierarchy	Retrieving hierarchy information
0	sobekcm_assistant.get_entire_collection_hierarchy
0	cached_data_manager.retrieve_item_aggregation
0	cached_data_manager.retrieve_item_aggregation	Found item aggregation on local cache
0	item_aggregation_builder.get_item_aggregation	Found 'all' item aggregation in cache
0	system.web.ui.page.page_load (ufdc.page_load)
0	sobekcm_page_globals.constructor.on_page_load
0	html_echo_mainwriter.add_style_references	Adding style references to HTML
0	html_echo_mainwriter.add_text_to_page	Reading the text from the file and echoing back to the output stream
97	html_echo_mainwriter.add_text_to_page	Finished reading and writing the file


UFDC Home	<%MYSOBEK%> \| Help \| RSS