Chemical engineering education ( Journal Site )

Material Information

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


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


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

Record Information

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

Full Text


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

T. J. Anderson

Phillip C. Wankat

Carole Yocum

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

E. Dendy Sloan, Jr.
Colorado School of Mines

Gary Poehlein
Georgia Institute of Technology
Klaus Timmerhaus
University of Colorado

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

Chemical Engineering Education

Volume 34

Number 1

Winter 2000

2 Oklahoma State University, R. Russell Rhinehart

8 Don Green, of the University of Kansas, Prisella J. Adams

14 Introduction to a Series, Richard M. Felder
16 Part 1. A Vision for a New Century, Armando Rugarcia, Richard M. Felder,
Donald R. Woods, James E. Stice
26 Part 2. Teaching Methods that Work, Richard M. Felder, Donald R. Woods,
James E. Stice, Armando Rugarcia
40 Particle Dynamics in Fluidization and Fluid-Particle Systems:
Part 1. Educational Issues, Liang-Shih Fan
56 A Feed-Effluent Heat Exchanger/Reactor Dynamic Control Laboratory
Experiment, William L. Luyben

62 Some Pitfalls with Citation Statistics, Ignacio E. Grossmann

48 Toward Technical Understanding: Part 4. General Hierarchy Based on the
Evolution of Cognition, J.M. Haile

66 All in a Day's Work, Richard M. Felder, Rebecca Brent

68 Application of a Heat Pump: A Feasibility Study,
Ljubica Matijasevic, Eduard Beer

74 Vapor-Liquid Equilibria in the Undergraduate Laboratory,
S. Wrenn, V. Lusvardi, G. Whitmyre, D. Buttrey
80 An Introductory ChE Laboratory Incorporating EC 2000 Criteria,
Stuart H. Munson-McGee
90 A Maxwell-Stefan Experiment, Pedro Taveira, Paulo Cruz, Addlio Mendes

86 The Annual ChE Symposium at Caregie Mellon, Timothy D. Power

94 A Dimensional Equation from Environmental Engineering, Keith B. Lodge

> 65,88
> 55, 89

Letter to the Editor
Book Reviews
Positions Available

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 2000 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. Periodicals Postage Paid at Gainesville, Florida.

Winter 2000


Chemical Engineering at...



W It V t 4 Theta Pond, on the
path between
Classes and

V Design groups.
Left to right:
Seniors Jill Petersen,
Melissa Hayes,
Sally Gerhold,
STing Chung, Mike Hill,
SCeand T.J. Crowell
collaborating on
process design

Oklahoma State University Stillwater, OK 74078-5021

quality in education has been a defining value in the School of
Chemical Engineering at Oklahoma State University (OSU) since
its inception in 1917. Developing students who will ultimately
beneheir community has a different tone from simply challenging
students on their ability to memorize facts and calculation procedures.
While competency in the fundamentals of chemical engineering science
and methodology remains important, developing people who are creative
and effective within a team environment, who manage their own growth,
and who make good things happen is the goal at OSU.
The school also has a mission to develop new knowledge and tools for
engineering. This vision guides the graduate research program where
students and faculty work closely together to both create and discover.
Students' first-place awards in national team design championships,
their outstanding student-chapter awards, national scholarships, and per-
formance on the Fundamentals of Engineering Exam are all testaments to
the fact that human resource development is working at OSU.
Copyright ChE Division of ASEE 2000
2 Chemical Engineering Education

"I continue to be impressed with the collegiality of the faculty," says Dr. Karen High,
whose specialty is optimization. This is a theme expressed by many of the faculty. In
Jan Wagner's words, "What attracted me to OSU was the professionalism of the faculty
throughout the college. They pull together." Jan's interest is safety and management
of change. Kinetics expert Gary Foutch recalls the "candor, fun, mentoring, and
support of the senior faculty. Decisions are made for the best interest of the
School, not for individuals."
Mutual support among the faculty expresses the value that all benefit when any one
individual has success. And whether cooperation is on research or on continuous
improvement, the synergism and balance from partnerships makes the final result
better. Collaboration is a value: it is a way of interacting with other people, and it carries
over from the faculty to the undergraduate and graduate programs as well. Respect
between faculty and students, and respect between students is important to the maximi-
zation of student learning and research performance.
Even while struggling to make sense of lab data, senior Melissa Hays recalls that
"OSU followed through with a personal interest in me when I was seeking a college,
and throughout my time here, I have always felt that the chemical engineering faculty
care for and respect the students." Jake Dearman, also a senior, adds, "The main reason
that I came here was because the campus was friendly and caring. Faculty want to see
the students grow personally and professionally, and that atmosphere is conducive to
learning." To integrate students into the program, all engineering students enroll in
ENGR 111, where they are led by a faculty mentor, learn about engineering and about
the values and attitudes promoted at OSU, and are encouraged to participate in school
Chemical engineering is a difficult curriculum with a significant challenge-it must
prepare students to use a complicated technology for the good of society. At OSU,
students work in teams to help each other solve complex problems and to encourage
each other through the personal ups and downs of the program. The faculty's open-door
policy affirms the importance of the student's learning activity. Even the arrangement
of faculty offices-suites off of one reception area-engenders a common team mis-
sion: "All for one, one for all."
The activities of the Student Chapter of the American Institute of Chemical Engineers
and of Omega Chi Epsilon (the ChE honorary) also enhance student/faculty interac-
tion. Events include formal speaker meetings, drop-in pizza dinner, golf tourna-
ments, picnics, and service activities that integrate students and faculty. In fact, the
AIChE Student Chapter has earned a National Outstanding Chapter Award for
three of the past four years.
But it's not just a party. While it is fun, it is also hard work.

OSU was named "Best College Buy" by Institutional Research and Evaluation, Inc.,
in their Student Guide to America's 100 Best College Buys: 1999.
"As a former OSU student, I was familiar with the quality of the program and its
potential for growth. I knew that I'd have many talented colleagues with whom I could
associate in my research area of thermodynamics. I also felt a sense of loyalty and
commitment to OSU," reflects Rob Robinson, Regents Professor, Amoco Chair,
and former School Head. Long before the new EC 2000 criteria, the School
established a tradition of continuous assessment, including feedback from an
Industrial Advisory Committee, and experienced frequent revision of the priorities
that guided its goals and objectives.

Chronology of Events
School of Chemical Engineering
Oklahoma State University

1890 Oklahoma Agricultural
and Mechanical College
(A&M) established
1907 Oklahoma Territory
granted Statehood
1909 First chemical
engineering course
within Department of
1917 Chemical Engineering
BS Degree plan granted
1921 First BS ChE graduates


AIChE Student Chapter

1947 Chemical Engineering
moves from Chemistry
to the Division of
1948 Policy: Each engineer-
ing student will take a
course from each
engineering program
1951 ChE Ph.D. program


First ChE Ph.D. granted
A&M becomes
Oklahoma State

1964 Omega Chi Epsilon
(Mu Chapter) ChE
Scholastic Honorary
1964 Industrial Advisory
Committee established
1966 Phillips Lecture Series
in ChE Education
1984 Pre-medical BS option
1993 Environmental
Engineering BS option

Winter 2000

While competency in the fundamentals of chemical engineering
science and methodology remains important, developing people who are
creative and effective within a team environment, who manage their own growth, and
who make good things happen is the goal at OSU.

The department has been very fortunate to have
the Phillips Petroleum Company as a co-sponsor
of the Phillips Lecture Series in Chemical Engi-
neering Education, an annual lecture at Oklahoma
State University now in its 33rd year. Speakers
have established themselves as leaders in educa-
tion, and reprints of their lectures are distributed
nationally. "My first awareness of Oklahoma State
University," recounts Russ Rhinehart, School
Head and Bartlett Chair, "came as a result of the
lecture pamphlets. I was preparing for a new ca-
reer in academe, and the stimulating lectures gen-
erated respect for the School."
Of course, the program is ABET accredited.
"Unit Operations Lab is fun!"-the often-heard
sentiment describes the students' team appren-
ticeship on distillation, reaction, heat exchange,
refrigeration, absorption, and fluid-flow processes.
In the classroom, the concepts are isolated and idealized
students can learn the fundamentals of chemical proc
behavior, but since the behavior is neither ideal nor isola
in real process equipment, students are also taught how
deal with such complexity before they are released to pr
tice their profession. With professors taking the role
"coaches" and students the role of team "players," the U
Operations Laboratory is a place where active mentor
shows the students how to apply the fundamentals to the r
world. Features include "360" evaluations, student own
ship of experimental design, and conversion of lab findii
into business decisions.
Process Design and Economics is also a practice-orien
course. Two full-time professors manage the course a
coach the student teams. The Celanese Chemical Compa
provides an annual design challenge, and their engine
evaluate the students' oral and written presentations as w
as the technical work.

"Balancing fundamentals with practice, as well as resea
with education were important attributes to me when I v
looking for a school," says School Head Russ Rhineh&
Emeritus Professor and heat-exchanger expert Ken Bell nol
"The program has always emphasized the practicality
technology in both teaching and research." The career ex
rience of chemical engineering faculty at OSU averages f
years of full-time practice. The result is that they understand

Unit Operations: Tamika Killian, Chad Smith, Mike
Dickenson, and Sally Gerhold as operators
on a computer-controlled distillation.

ess how engineers need to work and they integrate that knowl-
ted edge into their classes.
to This balance can be credited to former Head, and Emeritus
ac- Professor Robert N. Maddox, who initiated the Phillips edu-
of national lectures as well as the Industrial Advisory Commit-
[nit tee (IAC) in the '60s. The role of the IAC is to "provide
ing advice and council to the faculty of the School of Chemical
eal Engineering. All areas of operation will be open to criticism
er- and suggestion." Those values remain important today.
ngs This experience also shapes the graduate research pro-
gram (where about half of the funding is industrial), which
ted includes three industrially sponsored consortia. Gary Foutch,
ind Regents Professor and Kerr-McGee Chair, leads a group
my concerned with ultrapure water. "The Ultrapure Water Group
ers is a consortium that has existed for eight years now. The
tell eight companies in the group represent nuclear power, mi-
croelectronics, domestic water, and the Navy nuclear pro-
gram. The primary research focus is to develop accurate
models that will predict the performance of contaminant
rch removal from water where the concentrations of interest are
vas at the parts-per-trillion level. Students in the group frequently
art. have the opportunity to work in the industry as co-ops."
:es, Centers are created to enhance industrial and academic
of research-and-development collaborations. The newest is the
pe- Measurement and Control Engineering Center (MCEC), a
ive joint collaboration with the University of Tennessee, Knox-
ind ville, the US National Science Foundation, and twenty com-
Chemical Engineering Education

panies to explore and guide the practicable application of
advanced techniques for process automation. OSU partici-
pants include four professors and five graduate students. Led
at OSU by Dr. Karen High, it includes sponsorship of work
by Russ Rhinehart in nonlinear control and management
automation, Rob Whiteley in fault detection, Gary Yen
(Electrical and Computer Engineering) in computer per-
ception, and Karen's work in optimization in process
design and control.

Jan Wagner and Marty High lead the downhole corrosion
consortium. One product from this consortium is a Windows
software package that can predict the location and rate of
corrosion in both sweet and sour natural gas wells. The
project not only aims to solve the practical problem of corro-

sion in the petroleum industry, but also involves the funda-
mental study of fluid mechanics, thermodynamics, corrosion
kinetics, and mass transfer.

Certainly, one of the leading indicators of either industrial
or academic success is the student's ability to understand
and apply fundamentals. Evidence of this competence is
demonstrated by student performance on the Fundamentals
of Engineering (FE) Exam, administered nationally by the
NCEES. It is a comprehensive test of competence in engi-
neering material, and passing it is the first step toward be-
coming registered as a professional engineer. For the past
five years, OSU chemical engineering students have consis-
tently shown a 92% pass rate on the exam. Nationally, the
pass rate for chemical engineering graduates is 83%. OSU's

Faculty Oklahoma State University School of Chemical Engineering

Gary L. Foutch. Ph.D.. P.E.. Kerr-McGee Chair and Regents Professor
Ph.D., Chemical Engineering, University of Missouri-Rolla, 1980
Ultrapure Water Processing, Reaction Kinetics and Reactor Design
National Fulbright Scholarship Selection Committee

Khaled A.M. Gasem. Ph.D.. R.N. Maddox Professor
Ph.D., Chemical Engineering, Oklahoma State University, 1986
Thermodynamic and Transport Studies, Equilibrium Calculation Algo-
rithms, Process Design and Simulation
Integrated Petroleum Environmental Consortium co-founder and liaison
to the US Congress

James E. Halligan. Ph.D.. Professor. System CEO. OSU President
Ph.D., Chemical Engineering, Pennsylvania State University
People, Development

Karen A. High. Ph.D.. Associate Professor
Ph.D., Chemical Engineering, Pennsylvania State University, 1991
Optimization, Kinetic and Reactor Modeling, Environmental-Based Pro-
cess Design and Optimization, Parallel Optimization
Director AIChE Division 10 (Computing and Systems Technology)

Martin S. High. Ph.D.. P.E.. Associate Professor
Ph.D., Chemical Engineering, Pennsylvania State University, 1990
Equations of State for Polymer Solutions, Diffusion of Polymer Mol-
ecules, Downhole Corrosion Rates

Arland H. Johannes. Ph.D.. P.E.. Professor
Ph.D., Chemical Engineering, University of Kentucky, 1977
Interfacial Mass Transfer, Mathematical Modeling, Heterogeneous Ca-
talysis and Energy Conversion Systems, Hazardous Waste Disposal,
Fluidized Bed Reactors

Randy S. Lewis. Ph.D.. Associate Professor
Ph.D., Chemical Engineering, Massachusetts Inst. of Technology, 1995
Biomass Conversion to Liquid Fuels, Drug Delivery, Artificial Organs,
Past National Student Chapters Committee Chair of the AIChE, Member
of the National AIChE Career and Education Operating Council

R. Russell Rhinehart. Ph.D.. Edward E. Bartlett Chair and School Head
Ph.D., Chemical Engineering, North Carolina State University, 1985
Nonlinear and Statistical Methods in Process Optimization and Control

Editor-in-Chief of ISA Transactions, General Chair of the 2002 Ameri-
can Control Conference

Robert L Robinson Jr PhD PE Amoco Chair and Re ents P r

Ph.D., Chemical Engineering, Oklahoma State University, 1964
Equilibrium Behavior, Design of Solvents for Extractive Distillation,
Equation-of-State Representations, Adsorption
Director of the National Technological University ChE program, Fellow

Alan Tree. Ph.D.. Associate Professor. Interim Associate Dean for Research
Ph.D.. Chemical Engineering, University of Illinois, 1990
Polymer Science and Engineering, Flow-Induced Crystallization, Melt

Jan Wagner. Ph.D.. P.E.. Professor
Ph.D., Chemical Engineering, University of Kansas, 1976
Process Safety, Process Design, Phase Equilibria and Equations of State
Member AIChE/CCPS SACHE Committee and editor ofSACHE News

James Robert Whiteley. Ph.D.. P.E.. Associate Professor
Ph.D., Chemical Engineering, Ohio State University, 1991
Advanced Process Control, Artificial Intelligence Applications in the
Process Industries

Kenneth J. Bell. Ph.D.. P.E. Regents Professor Emeritus
Ph.D., Chemical Engineering (Physics minor), Univ. of Delaware, 1955
Fluid Dynamics, Heat Transfer, Heat Exchangers
AIChE Fellow, 1978 AIChE Donald Q. Kern Award. 1999AIChE HT&EC
Division Award

Robert N. Maddox. Ph.D.. Sc.D.. P.E. Leonard F. Sheerar Distinguished

Prnfessnr Fmeritu.r. HeparI Fmerituia

Ph.D., Chemical Engineering, Oklahoma State University, 1955
Gas Processing
Sc.D. honors causa University of Arkansas 1991; 1985 GPA Hanlon
Award; 1994 AIChE Founders Award

Marvin M. Johnson. Ph.D.. Adjunct Professor
Ph.D., Chemical Engineering, University of Utah, 1956
Kinetics and Catalysts
AIChE Fellow; 1985 National Medal of Technology Recipient; 1994
National Academy ofEngineering inductee

Winter 2000

4 Edmond Low Library, center of the
OSU campus.

Students, faculty, and alumni enjoy a pre-game
picnic. University President Jim Halligan talks
with Professors Karen and Marty High. V

sustained 92% pass rate is one clear indication of excellence
in both students and faculty.
Fundamentals are important, but engineers have to inte-
grate fundamentals, apply them, and devise a solution that
meets all of the criteria for goodness. Perhaps one indication
of excellence here is OSU's record of accomplishment in the
National AIChE Student Design Contests. In the five-year
history of the team contest, Oklahoma State student teams
have won first place twice. Since 1953, they have also re-
ceived many honorable mentions and first-place awards for
the individual design contest.
As stated earlier, collegiality was an initial attraction for
Dr. Jan Wagner, but he goes on to say, "The most enjoyable
thing now is the work ethic and values of the students. We
have high expectations, and the students meet them." Profes-
sor "AJ" Johannes seconds this; "Warm weather, low taxes,
and friendly, honest people were the initial attraction to
Oklahoma State. But, it is the hard-working honest students
that make every day a joy."

The School emphasizes basic research, and here, too, there
is a practice-relevant component. "I was attracted to OSU by
the expertise in processing, process development, and the
practice orientation in phase behavior," says Khaled Gasem,
Maddox Professor. One of OSU's premier graduate research
programs has been that of Professors Khaled Gasem and
Rob Robinson in their experimental findings and equation-
of-state development for the thermodynamic behavior of
multicomponent phase equilibrium. Their U.S. Department
of Energy-funded research on carbon dioxide sequestering
in coal beds has the dual purposes of reducing CO, in the air
while enhancing natural gas production. They are also de-
veloping techniques for using molecular simulation software
to "design" specialty solvents for extractive distillation, and
have demonstrated industrial success. Their work has a strong
experimental component and employs about a dozen gradu-
ate and undergraduate students.
Randy Lewis, Associate Professor, has collaborative bio-

technology research projects with several departments, each
with experimental credibility. He is exploring novel routes
to use bacteria to generate liquid fuels from renewable re-
sources, developing mathematical models for analysis of
drug delivery, analyzing immune response molecules and
their role in the dysfunction of an artificial pancreas, and
developing biomaterials that control local nitric oxide levels
and prevent platelet adhesion on implanted materials. Randy's
primary funding has come from the National Institutes of
Health and the Department of Energy. He is also the advisor
to the award-winning Student Chapter, one of the prime
reasons for its success.
Can you make a bridge smart enough to take heat from the
ground to prevent freezing when the surface is wet? "Yes,"
according to Associate Professor Rob Whiteley, who is col-
laborating with civil, mechanical, and technology professors
on a DOT-sponsored project. Rob and his students are re-
sponsible for the artificial intelligence that reacts to weather
forecasts and underground conditions and manages the
ground-source heat pumps. Rob is a regular winner of col-
lege and university teaching awards.
Martin High ("Dr. Marty") and Alan Tree have a shared
interest in the fundamentals of polymer kinetics and thermo-
dynamics. "The long, chain-like nature of polymer mol-
ecules allows them to crystallize under certain flow condi-
tions, resulting in structures with superior mechanical prop-
erties. Part of our NSF funding has been used to support
experimental and theoretical efforts that extend classical
crystallization theory to account for flow-induced behavior,"
says Alan, presently Interim Associate Dean for Research.
In addition, the team of High, Tree, and High are exploring
the applicability of strong-acid polymer catalysts for the
synthesis of liquid organic molecules from gases.
Chemical Engineering Education


--- ... '- --


Research on particle properties and size development is
led by Professors "AJ" Johannes and Gary Foutch. They are
using computational fluid dynamics programs and kinetic
models of interfacial mass transfer and chemical reaction to
improve the design of reactors.
We have forty graduate students, averaging about four per
faculty member, and each faculty member has externally
sponsored research. All graduate students are supported, and
60% are PhD candidates. The faculty backgrounds and re-
search interests are shown in Table I.

"I sought a place where I could build a future for both my
career and my family. OSU gave me collaborative faculty,
and Stillwater gave me a great place to live," says biomedi-
cal expert Randy Lewis. He notes that the quality of life in
Stillwater, a town with about 35,000 permanent residents,
has the flavor of a small town but is only a little over an hour
from either Tulsa or Oklahoma City. With a temperate climate
and located in the rolling hills of "green country," there are
four balanced seasons and many local recreational opportuni-
ties. Ken Bell adds, "The beauty of the campus originally
attracted me to OSU, and the recent upgrades in the Student
Union gardens and Theta Pond have made it even better."
There are 250 undergraduates who have declared chemi-
cal engineering as their major. About one third of them are
female, about 15% are from minority groups, and about 20%
are from outside of the state. These statistics allow everyone
to have an extended family of diverse individuals. Upper-
level class sizes are about 35; thus, with a faculty size of
eleven, the student-to-faculty ratio is relatively low.
With about 19,000 students on the Stillwater campus,
OSU is large enough to offer a broad variety of classes and a
wide range of student activities: Big XII athletics, cultural
events, international student societies, intramural events, rec-
reational outings, and study groups. The "engineering floors"
in the dorms provide both social and academic support ac-
tivities for the students.

For freshman and Phillips Scholar Christy Petersen, "Op-
portunities made the program attractive." She should know.
Her older sister is a senior, and her oldest sister and brother-
in-law are recent graduates, all from chemical engineering at
Oklahoma State. "The program prepares students for a broad
range of careers. Carl is in law school, Tracie is enjoying her
industry assignments, and Jill is looking at medical school.
Within chemical engineering we have the bioengineering,
premedical, and environmental options. And, while on cam-
pus, there are many activities that allow us to explore our
potential." Hosting the 1999 National AIChE Annual Stu-
dent Conference in Dallas, Texas, was a major activity that
engaged about 50 of our students.
Winter 2000

One of the special features of the College of Engineering,
Architecture, and Technology is that all students are in a
"common" curriculum for their first two years. Once they
qualify academically, they are admitted into the School that
offers a degree in the major of their choice. Courses in the
common engineering curriculum include computer program-
ming, statics, thermodynamics, fluid dynamics, strength of
materials, electrical circuits, and materials science. Faculty
from separate Schools team-teach most of these courses, and
their content is controlled by the faculty from all disciplines
that require the courses. This means that chemical engineer-
ing students get a multidisciplinary perspective in their gen-
eral engineering courses, and as a result, they have subse-
quent career flexibility. Course management by an oversight
committee maintains a balanced structure, curriculum con-
tent, and high academic standards. This broad and quality-
monitored experience contributes to diverse career opportu-
nities for the students.
Students have to qualify for acceptance into the School of
Chemical Engineering. Making the "grades" in the first two
years is a prerequisite to taking the upper level chemical
engineering courses. It is a significant challenge, and the feel-
ing is that the student-faculty family is an important contribu-
tor for promoting academic scholarship and encouragement.
Undergraduates can be involved in research if they wish to
explore outside of the classroom. Recently, Nellie Bruce, a
junior, received a Wentz Scholarship to develop an artificial
kidney experiment that will be used for the newly developed
biomedical engineering course. Nellie is the fourth of her
siblings to choose engineering at OSU and the third in chemi-
cal engineering. Dipesh Pokharel received a Wentz Scholar-
ship to study the effects of nitric oxide (NO) on smooth
muscle-cell growth. Dipesh developed the experimental ap-
paratus that delivers controlled and predictable amounts of
NO to a culture dish containing smooth muscle cells. Cell
growth is monitored during NO exposure. Phoebe Brown,
recipient of a national Morris Udall Scholarship, is evaluat-
ing a method for automatic identification of steady state in
multivariable processes. Since most processes generate noisy
data, the method is statistically based. Phoebe is testing the
multivariable method on data from the Unit Operations Lab
distillation column.

Chemical engineering at OSU uses the same textbooks as
other universities. It is accredited by the same rules. It has
the same best students. It has the same inadequate operating
budget. It has the same web and library access. So-what is
the reason for its success? If not magic, then it must be a
shared commitment to human development and a focus on
getting results in the real world.
We invite you to enjoy a visit to our web pages at http:// 1

F =1 educator




of the


of Kansas


University of Kansas Lawrence, KS 66045
When Don Green entered graduate school in the
late 1950s, he didn't know that he would emerge
with a lifelong friendship and professional rela-
tionship that would affect how chemical engineers the world
over do their jobs. Perry's Chemical Engineers' Handbook-
relied on by practicing engineers and engineering students
alike-has a history few people know. Fewer still under-
stand the role Don Green has had in continuing the Perry
family legacy.
"I was Bob Perry's first PhD student when he was a
chemical engineering faculty member at the University of
Oklahoma," recalls Green, who is the Deane E. Ackers
Distinguished Professor of chemical and petroleum engi-
neering at The University of Kansas. "John Perry [Bob's
father] edited the first three editions, and Bob assumed
editorship after his father's death in 1953. While at OU, Bob
edited the fourth edition of the handbook, and I assisted him
in various ways, including working on the kinetics section."
The Green and Perry families continued their friendship
after both men left OU. In 1977, Perry was consulting and
living in London while continuing his work with the hand-

book, and he pressed Green into service again.
"Bob invited me to be a section editor on the sixth edition
and had a plan that I would join him as coeditor for the
seventh edition, targeted for publication about ten years after
the sixth," Green said. "Bob had two sons, neither of whom
had careers related to engineering. He had inherited the
book from his father, and I think he wanted to pass the
leadership along to a person who was close to him and
almost like family."
Green accepted the invitation, but before the two friends
were able to do much work, Bob Perry was killed instantly in
a tragic car accident in London while crossing the street on
foot. Perry's widow, Gail, and publisher of the handbook,
McGraw-Hill Book Co., asked Green to assume the editorship
of the sixth edition, and he accepted. Green engaged James
O. Maloney, KU emeritus professor of chemical and petro-
leum engineering and former department chairman of 19
years, to support the project as assistant editor. The sixth
edition sold approximately 190,000 copies and is probably
more widely distributed across the world than any other
chemical engineering book.

Copyright ChE Division of ASEE 2000

Chemical Engineering Education

1 "Don [comes] across
as a caring and
thoughtful person,
one who [embodies] the
philosophy that the
university experience can
transform the lives
of students ....
[He] is an excellent
role model for other
faculty in that he does
everything so well:
teaching, scholarship.
leadership, and service.
And he excels in all of
these areas while
maintaining a calm,
cool demeanor." 4

Today, Perry's Chemical Engineers' Handbook
is in its seventh edition, with Don Green as editor
and J.O. Maloney as associate editor. Published in
1997, it has sold approximately 37,500 copies in its
first two years and was recognized by the Library
Division of the American Society for Engineering
Education as the Best Reference Book of 1998.
Bob Perry's decision to pass his father's legacy
on to Green was well-founded. Don Green is an
inspired educator and esteemed research engineer
who holds genuine concern for the success and
happiness of his students and colleagues. He is a
well-known researcher in oil reservoir technology,
a Fellow in the American Institute of Chemical
Engineers, and an ABET accreditation visitor for
the Society of Petroleum Engineers (SPE). He is a
Distinguished Member of the SPE, a reviewer of
technical articles for the SPE Journal, and has been
an SPE Distinguished Lecturer. Green is chairman
of his department at KU and past chair of the Asso-
ciation of Petroleum Engineering Heads. He is a
sports enthusiast and a faculty athletic representa-
tive to the Big XII Conference and the National
Collegiate Athletic Association. But his role of pro-

fessor as teacher is the role Green loves best. During a 35-year teaching
career, he has inspired countless students with his enthusiasm, dedica-
tion, judgment, and good nature.
"Don made his class subject matter interesting and enjoyable," says
former student Bill Weisenborn, now with Conoco, Inc. "He could
explain all topics in such a way that we wanted to continue to learn and
improve ourselves."
Carlos Rocha, former graduate student and now a project manager
with Jacobs Engineering in Hamilton, Ohio, finds that he still calls
upon advice gained in Green's classes. "Don Green provided me with
an example of how a person should behave in our competing and
demanding society," Rocha says. "As a project manager for a very large
engineering consulting firm, sometimes I get caught on decisions that I
think are overwhelming. How can I make the right decision-do the
right thing? I go back to what I learned from Don: If I am not uncom-
fortable telling my family the decision I made and its potential conse-
quences, then I made the best decision I could. He taught me ethics as
an engineer and as a human being."

Green was born and raised in Tulsa, Oklahoma, when the city bore
the nickname of "Oil Capital of the World." Oil derricks and office
buildings of the world's largest petroleum companies neighbored the
American Legion ball fields that Green played on as a boy. In a
community with an industry based in petroleum, Green never ques-
tioned having a career in oil. He decided during high school to become
a petroleum engineer.
Green spent his freshman year at Oklahoma State University before
transferring to the University of Tulsa. His love for baseball won him
the varsity team's third-base position for two years, and he was selected
as second-team All-American. In 1954, Green and Patricia Polston,
who was studying nursing, were married. The next year, Green gradu-
ated with a BS in petroleum engineering and went to work as an
engineer trainee for Gulf Oil Co., in Tulsa.
Having participated in an Air Force ROTC program as an under-
graduate, Green was called into active duty after five months at Gulf
Oil. Point of duty: Suffolk County AFB, Long Island, New York.
"Being a Midwesterner, the East Coast was the last place in the world
I wanted to be stationed," Green says, "but it turned out to be a great
location and assignment."
Green was made petroleum supply officer for the base, a position of
responsibility that honed his management skills and gave him the
maturity to think clearly about the future. He settled on returning to
school for graduate study in chemical engineering, and in 1958 he and
Pat traveled back to Oklahoma. Green entered the master's program at
the University of Oklahoma under Professor Richard Huntington. He
continued at OU as a doctoral student under Professor Perry.
Still pursuing his interest in petroleum engineering, Green chose heat
transfer in porous media as his doctoral research topic. He considers his
excellent experience as a doctoral student to have been a strong influ-
ence on his decision to become a professor. Mentors that influenced

Winter 2000

him most were Perry, Cedomir Sliepcievich, and
Jack Powers.
Green finished his doctorate in 1962 and joined Con-
tinental Oil Company. As a research engineer in the
company's Petroleum Production Research Division in
Ponca City, Oklahoma, Green worked with computer
simulations of oil reservoirs. Although the research
was challenging, Green's graduate school experience
remained in his thoughts.
"The lifestyle the professors were leading seemed
exciting to me," Green says. "I liked the idea of being
in a university setting where teaching, research, and
scholarship activities were combined."
After two years in industry, Green made the move to
being a chemical and petroleum engineering educator
when he was hired by KU in 1964 as an assistant
professor. Because the chemical and petroleum engi-
neering department at KU offered undergraduate and
graduate degrees in both areas, Green and the job were
a perfect fit. He was promoted to associate professor in
1967 and to full professor in 1971. He was named the
Conger-Gabel Distinguished Professor in 1982, an
award that requires a demonstrated record of sustained
excellence in undergraduate education in addition to
excellence in research and professional service. And in
1995, Green was honored with the Deane E. Ackers
Distinguished Professorship, a post he still holds.
He is in his second term as department chairman and
has won teaching awards on the average of one every
other year. With his KU colleague Paul Willhite, Green
is co-director of a program in oil reservoir technology
that is recognized as a model for technology transfer.

4 Don (left) and his brother-in-law Bob Willett at
the top of Long's Peak in Rocky Mountain National
Park, Colorado, after scaling the 14,256-ft. moun-
tain. (Photo courtesy of Don Green, 1994)
V C.S. Howat, III (left) congratulates Don on receiving the
1987 HOPE Award while chancellor Gene Budig (right)
looks on. Howat and Green are both on the faculty
of the KU Chemical and Petroleum Engineering
Department. (Photo by R. Steve Dick, 1987)

From its location in the state's eastern corner, The University of
Kansas has long been considered by state residents, whether for
better or worse, as a liberal institution. The political atmosphere of
the Vietnam era only added to that reputation. While student
demonstrations at KU did not escalate to the levels of violence at
other universities across the country, KU's image statewide was
somewhat tarnished.
In 1973, then-Chancellor Archie Dykes pushed the university to
develop interactive programs with the state. Green and Willhite took
a close look at Kansas oil producers, who at the time were between a
rock and a hard place: The country was enduring an energy crunch,
yet within Kansas bedrock lay inaccessible oil resources. The pro-
ducers, who were mostly small, independent operators without engi-
neering staffs or direct access to research facilities, found them-
selves unable to use the enhanced oil recovery technology being
developed primarily by major oil companies. Green and Willhite
responded with a concept for a program that would help the state's
independent oil producers take advantage of technology research
that no single producer could afford to fund.
The Tertiary Oil Recovery Project (TORP) received state funding
in 1974, with Green and Willhite as co-directors. It has been funded
every year since and last fiscal year received more than $790,000 in
research support from the State of Kansas. Since 1979, the U.S.
Department of Energy has provided more than $8.5 million in sup-
port of research and technology transfer. Major companies such as
AMOCO, Phillips Petroleum, and Marathon Oil have also been
supporters. The project's objectives are to conduct research related
to enhanced oil recovery processes, to provide technical assistance

Chemical Engineering Education

to industry, and to educate students and operators in tertiary
oil recovery and reservoir management. TORP is tied closely
to the CPE department and to the KU School of Engineering.
Over the years, about 90 graduate students have conducted
research through TORP leading to their PhD or MS degree.
The project has also provided research opportunities for
undergraduate students.
Green and Willhite have published numerous articles and

reports related to oil recovery, and they are
frequent presenters at technical meetings. In
1998, they published Enhanced Oil Recovery
through the Society of Petroleum Engineers.
The book is a comprehensive text of advanced
oil-recovery processes and is applicable for
use in senior or graduate courses in petro-
leum engineering.
Over the years, TORP researchers have ex-
plored thermal recovery processes, micellar-
polymer flooding, carbon dioxide miscible dis-
placement, reservoir computer simulation, and
in situ permeability modification using gelled
polymer systems.

participant in technology transfer and because the project
has been so effective in fulfilling its objectives, the Depart-
ment of Energy used TORP as a model when it increased
funding of tech-transfer programs across the country in 1994.

TORP is one of several research programs that involve
KU's chemical and petroleum engineering students. The

"I strongly feel
that what is right
for our program
is a balance
teaching and
research that
gives each a
_.__LaJ__L __ H

A major, current emphasis is development wel
of gelled polymer technology, which aims to Gree
improve volumetric sweep efficiency in oil- think
recovery displacement processes such as wa-
terflooding. In such a displacement process, mypri
the injected fluid often flows between the in- as C.
jection wells to production wells in a "short
circuit" because of high permeability zones or
fractures in the reservoir. As a result, most of faCUll
the oil is bypassed and not contacted by the appro
injected fluid, resulting in poor displacement feas
efficiency. Additionally, the injected fluid must
be recirculated or disposed of, which can be
costly. Gelled polymer technology involves in-
jecting a gel system into the "thief" zone where it reacts to
form a gel and thereby seals off the zone. Fluid subsequently
injected will be forced into other parts of the reservoir,
thereby improving efficiency.
TORP researchers also are working on implementing a
field trial to recover oil from Kansas reservoirs through
applications of supercritical carbon dioxide. They believe
the process, which has been used successfully in west Texas,
has the potential to revitalize the Kansas oil industry. To
gain an understanding of the potential impact the field trial
may have, approximately 10 billion barrels of unrecovered
oil are estimated to lie within Kansas' borders.
The success of TORP at transferring viable technologies
into the private sector has gained the program recognition by
the Independent Petroleum Association of America and by
the Department of Energy. Because TORP was an early

-n says. "I
that one of
imary roles
hair is to
port the
ty in every
private and
ible way.

department has had a rotating chair since
1964, and since then Green has held the post
twice, including his current appointment.
Green credits Willhite, the department's pre-
vious chair, with building a well-rounded
faculty and with strengthening the research
programs. Both are areas Green continues
to emphasize.
"I strongly feel that what is right for our
program is a balance between teaching and
research that gives each a comparable weight-
ing," Green says. "I think that one of my
primary roles as chair is to support the fac-
ulty in every appropriate and feasible way.
This means providing the resources needed
for teaching, assisting in the development of
research programs, and working to see that
they are recognized for accomplishments."
The result is a cohesive department with a
diverse faculty, strong undergraduate and
graduate programs, and well-endowed schol-
arships that attract some of the brightest stu-
dents entering the university. A recent re-
view of the university by the Kansas Board
of Regents provided data showing that the
average ACT score of freshmen entering
the CPE program was the highest of any
program at KU.

Compared with their peers nationally, KU's CPE students
remain strong. In the last fifteen years, KU students have
won more awards in the National Student Design Competi-
tion sponsored by the AIChE than any institution in the
country. The record is due in large part to Sharp Teaching
Professor Colin S. Howat III, who teaches chemical engi-
neering design, as well as to the outstanding quality of
the student body.
Student achievements in the last two years have brought
the department further national recognition. In 1998, three
chemical engineering students were awarded Barry S.
Goldwater scholarships for outstanding academic achieve-
ment and research. The honor, considered the premier award
of its type, is given annually to undergraduates who excel in
engineering, natural sciences, or mathematics. An institution
may nominate only four students. Interestingly, that year a

Winter 2000

KU physics student also won a Goldwater Scholarship, so
the university as a whole was batting four for four.
One of the chemical engineering recipients, Larissa Lee,
went on to win a Churchill Foundation Scholarship to study
in England. Lee pursued the department's pre-medical op-
tion, one of three program options that Green and CPE
faculty developed to enable undergraduate students to
tailor their studies according to their plans for graduate
school or industry. The options are pre-medical, bio-
medical, and environmental.
The program options, combined with strong scholarships,
have attracted greater numbers of women students to the
CPE program. In the 1997-98 school year, nearly 46% of
undergraduate degrees in chemical engineering at KU were
awarded to women. The average throughout the School of
Engineering, which has ten undergraduate engineering pro-
grams, was 19.2%. Nationally, the figure for women earning
undergraduate degrees in engineering was 18.7% according
to the American Association of Engineering Societies, Inc.
As chair, Green has worked to maintain excellence in
teaching and research innovations by building faculty
strength. The department recently hired three new faculty,
two of whom are women, to make a total of thirteen mem-
bers. Bala Subramaniam, Conger-Gabel Distinguished Pro-
fessor and assistant department chair, notes that Green is the
type of leader who considers all viewpoints, yet arrives at
decisions without delay. "Don's impressive record of ac-
complishments in education, research, and service enables
him to command the respect and trust of his faculty as chair,"
Subramaniam says. "I've found Don to be a very open-minded
leader, and he has a knack for developing consensus."
Marylee Southard, an alumna of the chemical engineering
program and now an associate professor, says Green encour-
ages everyone to go as far as possible to reach their own
potential. "He is the ultimate cheerleader," she says, offering
as example Green's support when both educators were in
contention for a university-wide teaching award. "We stood
out on the football field at halftime on a typical sunny, chilly
November day, waiting for the winner to be announced. Don
commented to me that he hoped I'd win it; that he'd
already won this once before, and I deserved it. When it
happened, I believed him."
Subramaniam echoes Southard's observations. "Don be-
lieves that an important role of a chair is to facilitate the
professional development of every faculty member. He does
this fairly and effectively. I have been most impressed by his
genuine concern to see every one of his faculty succeed and
be suitably rewarded for their efforts."
Southard considers Green's most significant achievement
to be a consistency of excellence that affects how he mentors
both students and faculty. "He has lived what he advises us
'young Turks' to do," Southard says. "Don is committed to

Green also has been involved with the
KU Athletic Department [and] serves] as the
university's Faculty Athletic Representative to
the Big XII Conference and the NCAA [where he]
deals with legislative issues such as academic
requirements, rule-waiver requests,
eligibility concerns, and budgets.

doing the best at all facets of his job. He is tireless in his work
and completely prepared for meetings and classes every day."

Subramaniam says that whether Green is teaching an in-
troductory class to freshmen or a graduate-level special-
topic seminar, he prepares his lecture with the audience in
mind and delivers the lecture with the same enthusiasm and
clarity. "This is just a reflection of Don's positive attitude
toward, and respect for, students at all stages,"
Subramaniam says. "He treats them all alike-as junior
colleagues, as he likes to say."
Tom Edgar, chaired professor of chemical engineering at
the University of Texas, says his interaction with Don Green
while a student at KU in the late 1960s was a major factor in
his decision to become an educator. "Don came across as a
caring and thoughtful person, one who embodied the phi-
losophy that the university experience can transform the
lives of students," Edgar says. "Don is an excellent role
model for other faculty in that he does everything so well:
teaching, scholarship, leadership, and service. And he excels in
all of these areas while maintaining a calm, cool demeanor."
"Those of us who go through cycles of burnout and high
energy wish there were a pill that could give us his drive,"
Southard says. But the answer to Green's energy lies with
his mentor, Bob Perry. Green says, "Bob used to tell me,
when things were not going all that well because of adminis-
trative or bureaucratic problems, 'Thank God for the stu-
dents. It's because of the students that we're here.' That
quote accurately reflects my feelings."
His sense of humor no doubt has helped Green keep his
energy level high, particularly when he uses it in classroom
settings to ease tension and let students know that he, too, is
human. Rocha remembers Green as capable of taking jokes
and handling unforeseen situations well. "In one of his
courses, he gave us (the students) the wrong data to use in a
computer problem. He showed up in the classroom with a
sign hanging from his neck that read 'Stupid.'" Rocha con-
tinues, "He taught me that we all make mistakes, even a
professor, and that it makes the situation much better for
everybody if the mistake is acknowledged and everyone
moves forward together."
Over the years, Green has won numerous teaching awards,
but the award of which he is most proud is the university-

Chemical Engineering Education

wide Honor to Outstanding Progressive Educator, or
HOPE, award. The award selection is made annually by
members of the senior class.
Green was the initial sponsor of the precursor to the KU
student chapter of the Society of Women Engineers, and he
has involved himself in other campus programs related to
students and teaching. In the 1970s, Green and his colleague,
Floyd Preston, worked with a group of African American
students to develop a
program to recruit,
support, and mentor
Preston and Green
were the first sponsors
of the group, and over
the years it has evolved
into a well-established
program that incorpo-
rates four minority en-
gineering groups. It
now has a full-time ad-
viser and serves ap-
proximately eighty stu-
dents at any one time.
Green also has been
involved with the KU
Athletic Department,
serving two terms on (Left to Right) Don, his dat
the athletic board. In holding granddaughter
1996, Chancellor Rob- Patrick, and
ert Hemenway invited
Green to serve as the
university's Faculty Athletic Representative to the Big XII
Conference and the NCAA. In the post, Green deals with
legislative issues such as academic requirements, rule-waiver
requests, eligibility concerns, and budgets.
Several years ago Don took part in a university-wide dis-
cussion of teaching. The result was a committee that focused
on ways of improving teaching and undergraduate educa-
tion. Green chaired the committee for several years, during
which it looked at issues such as colleague-to-colleague
mentoring, improving classroom physical facilities, augment-
ing audio-visual equipment in classrooms, holding teaching
colloquia, and initiating teaching awards. The committee
evolved into the Center for Teaching Excellence at KU and
now has its own facility and dedicated staff.

Don Green may not allow his competitiveness to come
through when teaching, but he is only holding it in check.
An avid handball player since his days in graduate school,
he rarely lets anything interfere with his games and often
plays three times a week with a small group of friends.

Winter 2000


"We feel about handball as does Sarge of the 'Beetle
Bailey' comic strip," Green says. "According to Sarge,
handball is the only real court game and racquetball is for
wimps. At least, that's the kidding I give to our students,
who tend to play racquetball."
Green and his three sons, all KU graduates, are enthusias-
tic about KU sports, especially KU basketball. Guy, the
oldest son, is an environmental engineer with the U.S. Corps
of Engineers, Michael
is Assistant U.S. At-
torney in the Western
District of Missouri,
where he prosecutes
drug cases, and Patrick
is a medical doctor and
S 1 cardiologist in Ft.
Collins, Colorado. Pat
and Don are blessed
t c with two grandchil-
f dren, ages seven and
four, who were born
to Guy and his wife,
When his sons were
young, Green served
as Little League coach,
and the family often
er-in-law Aina, his son Guy went campin and hik-
ka, his wife Pat, his son ing together, some-
son Michael. thing they still enjoy.
The family shares a va-
cation home near Estes
Park, Colorado. Green has climbed Long's Peak, in Rocky
Mountain National Park, three times-twice with Pat. The
mountain, known as a "fourteener" by climbers, has an el-
evation of 14,256 feet and an elevation gain of 4,850 feet.
Green plans to go to the top again soon. "When you're
hiking, you're challenging yourself and you get to see beau-
tiful country," he says. "Being outdoors is wonderful, and
there's nothing like being in the mountains."
The drive to challenge oneself and others from a point of
respect while at the same time enjoying life is a gift Don
shares with those around him. Marylee Southard, who first
met Don in 1972 when she was a senior in high school,
characterizes him best. "Don said he has no recollection of
any childhood pivotal point or transforming cataclysm that
suddenly gave him drive and optimism," she says. "I believe
these traits are an inborn part of Don Green. He is a Mid-
westerner with a decent work ethic and a love for his stu-
dents and for working with them. We who have been
mentored by Don are his legacy, and we hope that his
attitude and energy have become part of us as we talk and
work with him." 0

SSpecial Feature Section


Introduction to a Series

North Carolina State University Raleigh, NC 27695

In the Spring 1990 semester, I spent a most enjoyable sabbatical semester at Georgia Tech,
where I worked with Ron Rousseau on the initial stages of the revision of Elementary
Principles of Chemical Processes. At the same time, I was wading through a mountain of
books and papers on cognitive psychology, educational psychology, and science and engineer-
ing education, building up my background for a longitudinal study of engineering education
for which the NSF had just provided funding.
The research I was immersed in led me to several observations. First, there was a lot of stuff
out there in the literature, some of which I found particularly relevant to my courses and my
students. Second, few engineering professors would ever have the time or inclination to wade
through all of it in search of something they could use. It occurred to me that as long as I was
going through the exercise of distilling the literature, it might be useful to my colleagues if I
shared the fruits of my labors. It also occurred that it would make little sense for me to do it
alone, since I knew of other engineering educators who had thought about these issues far
more than I had and had a much deeper knowledge of the literature.
At that point I conceived of a series of survey articles in Chemical Engineering Education,
coauthored by highly knowledgeable educators with me riding their coattails. Among the most
knowledgeable chemical engineering educators I knew at the time-and still among the most
knowledgeable-were (in alphabetical order) Armando Rugarcia of the Universidad
Iberoamericana in Mexico, Jim Stice of the University of Texas, and Don Woods of McMaster
University in Canada. I invited them to participate and was delighted when all three accepted.
The North American quartet got to work immediately.
Then life happened.

Copyright ChE Division of ASEE 2000
4 Chemical Engineering Education

Future of Engineering Education ]

Armando became Rector of his university, Jim started running all over the country
giving teaching workshops, and Don became a self-contained book-of-the-month club
as his problem-based learning approach became an international paradigm for effective
instruction. Also, owing to the incessant time demands of the book revision, the longitu-
dinal study, and my own teaching workshops, I became the worst offender of all. But at
length we picked it up again, thanks mostly to Don's unflagging energy and initiative,
and the series finally came into existence.

The first two papers follow in this issue, and the remaining four will appear in
subsequent issues. The first paper sets the stage and previews the structure of the series,
so I won't do so here. I will just say that it has been a privilege and pleasure to work with
such outstanding educators and good friends as my coauthors. I have been inspired by
their ideas for many years. I hope their enthusiasm and love of their work comes through
in these papers and inspires the readers in the same way. 0

Richard M. Felder is Hoechst
Celanese Professor (Emeri-
tus) of Chemical Engineering
at North Carolina State Uni-
versity. He received his BChE
from City College of New York
and his PhD from Princeton.
He has presented courses on
chemical engineering prin-
ciples, reactor design, process
optimization, and effective
teaching to various American and foreign industries
and institutions. He is coauthor of the text Elementary
Principles of Chemical Processes (Wiley, 2000).

James Stice is Bob R. Dorsey
Professor of Engineering
(Emeritus) at the University of
Texas at Austin. He received
his BS degree from the Univer-
sity of Arkansas and his MS and
PhD degrees from Illinois Insti-
tute of Technology, all in chemical engineering. He has
taught chemical engineering for 44 years at the Univer-
sity of Arkansas, Illinois Tech, the University of Texas,
and the University of Wyoming. At UT he was the direc-
tor of the Bureau of Engineering Teaching and initiated
the campus-wide Center for Teaching Effectiveness,
which he directed for 16 years.

Donald R. Woods is a pro-
fessor of chemical engineer-
ing at McMaster University. He
is a graduate of Queen's Uni-
versity and the University of
Wisconsin. He joined the fac-
ulty at McMaster University in
1964 after working in industry,
and has served as Department
Chair and as Director of the
Engineering and Management
program there. His teaching and research interests
are in surface phenomena, plant design, cost esti-
mation, and developing problem-solving skills.

Armando Rugarcia gradu-
ated from the Universidad
Iberoamericana (UIA) in
1970 and went on to earn
his MS in chemical engineer-
ing from the University of
Wisconsin in 1973 and his
Doctorate in Education from
West Virginia University in
1985. He has been a full-
time professor of engineering at UIA since 1974
and was chair of the Chemical Engineering Depart-
ment there from 1975 to 1980. He was also Director
of the Center for Teaching Effectiveness at UIA
from 1980 until 1986. He has written four books on
education, one on process engineering, and more
than 130 articles.

Winter 2000

Special Feature Section


Part 1. A Vision for a New Century

Armando Rugarcia Iberoamericana University Puebla, Mexico
Richard M. Felder North Carolina State University Raleigh, NC 27695-7905
Donald R. Woods McMaster University Hamilton, Ontario, Canada L8S 4L7
James E. Stice University of Texas Austin, TX 78712-1062

When we walk into an arbitrarily chosen engi-
neering classroom in 2000, what do we see? Too
often the same thing we would have
seen in 1970, or 1940. The professor stands at
the front of the room, copying a derivation num
from his notes onto the board and repeating un
aloud what he writes. The students sit pas- admnis
sively, copying from the board, reading, work-
ing on homework from another class, or day- profe
dreaming. Once in a while the professor asks a question
question: the student in the front row who feels of t
compelled to answer almost every question may engine
respond, and the others simply avoid eye con- tradition
tact with the professor until the awkward mo- taught,
ment passes. At the end of the class, students have begin
are assigned several problems that require them alternt
to do something similar to what the professor even those
just did or simply to solve the derived formula about ai
for some variable from given values of other fe
variables. The next class is the same, and so is
the next one, and the one after that. tr
the way
There are some differences from thirty years will w
ago, of course. The homework assignments fwl
require the use of calculators instead of slide cmm
rules, or possibly computers used as large cal-
culators. The math is more sophisticated and will le
graphical solution methods are not as likely to with insuJ
come up. The board is green or white, or maybe to pur
an overhead projector is used. Nevertheless, res,
little evidence of anything that has appeared
in articles and conferences on engineering education in
the past half-century can be found in most of our class-
rooms and textbooks.

In recent years, however, there have been signs of change."'
Engineering professors have increasingly begun to read the
education literature and to attend ASEE confer-

bers of
trators and
ssors ...
the viability
ie way
ering has
nally been
and some
n to explore
e who know
r that
they teach
require a
tment that
ave them
fflcient time
sue their

ences and teaching workshops, and some have
attempted to adopt new approaches in their teach-
ing. A number of factors are responsible for this
increased interest in effective teaching in engi-
neering schools. Growing numbers of parents,
taxpayers, and legislators have read graphic de-
scriptions of the de-emphasis of undergraduate
education at major universities2'" and have be-
gun to raise embarrassing questions with uni-
versity administrators. Corporations and employ-
ers have frequently and publicly complained
about the lack of professional awareness and
low levels of communication and teamwork
skills in engineering graduates13-61 and about the
failure of universities to use sound management
principles in their operations. [7'8
These rumblings have been heard by the U.S.
Accreditation Board for Engineering and Tech-
nology (ABET), which now proposes to hold
engineering schools accountable for the knowl-
edge, skills, and professional values engineer-
ing students acquire (or fail to acquire) in the
course of their education. Starting in 2001, En-
gineering Criteria 2000 will be implemented as
the standard for accreditation. Thereafter, all
U.S. engineering programs will have to demon-
strate that besides having a firm grasp of sci-

ence, mathematics, and engineering fundamentals, their
graduates possess communication, multidisciplinary team-
work, and lifelong learning skills and awareness of social

Copyright ChE Division of ASEE 2000

Chemical Engineering Education

Future of Engineering Education I
<_______________________________________ __ ___________________-

and ethical considerations associated with the en-
gineering profession. 91
These driving forces and personal convictions
about the importance of education in the academic
mission have led increasing numbers of university
administrators and professors to question the vi-
ability of the way engineering has traditionally
been taught, and some have begun to explore alter-
natives. Most, however, are unsure of what the
alternatives are to the traditional methods, and even
those who know about alternatives fear that trans-
forming the way they teach will require a full-time
commitment that will leave them with insufficient
time to pursue their research.
Our goal in this paper and in the five that follow
it is to offer some tools to engineering professors
who wish to become better teachers and to univer-
sity administrators who wish to improve the qual-
ity of teaching at their institutions. This paper at-
tempts to define in some detail the challenges cur-
rently facing engineering education. The second
article will survey teaching methods that have re-
peatedly been shown to improve learning; the third
will elaborate on methods that help students de-
velop critical skills; the fourth will examine effec-
tive ways to prepare the professoriate to learn and
implement the new methods; the fifth will propose
methods of assessing and evaluating teaching ef-
fectiveness; and the sixth will explore possible
modifications in the university incentive-and-re-
ward structure that will enable the desired changes
to occur on a systemic level.

A system of education is closely woven into the
fabric of the society within which it operates. Be-
fore examining new ways to train engineers, we
might do well to anticipate some characteristics of
the society within which the engineers we are train-
ing will function. We are writing from the perspec-
tive of Mexican, American, and Canadian cultures,
but we feel that the trends can be generalized to a
broad range of developed and developing nations.
We see seven features of the coming century that
will pose challenges to future engineers.
> Information: Proliferating In 1989, 10,000
volumes were required just to list the titles of
all the books that had been published, and
roughly 6,000 scientific articles were published
every day." The number of documents avail-

able has since tripled, and there is every
indication that the rate of growth will be
sustained, if not increased. Moreover, the
flood of information will wash right up to
the engineer's fingertips through the
internet, virtual environments, and CD-
ROM discs that can each hold up to one
million pages of text.
- Technological Development: Multi-disci-
plinary In the early part of this century,
engineering practice could be classified
along disciplinary lines (although not to the
extent that university curricula would have
had us believe). The body of knowledge
that constituted the working arsenal of, say,
a chemical engineer, was well-defined and
distinct from that which characterized a
mechanical or electrical engineer or a chem-
ist or physicist. The situation now is much
more complex: for example, engineers of
all types are finding themselves faced with
a need to know electronics and/or biochem-
istry. The key to better technological devel-
opment lies in cooperation among the pre-
viously separate disciplines to attack prob-
lems that have no recognizable disciplinary
- Markets: Globalized In the future, indus-
tries that cannot compete in the interna-
tional market are unlikely to survive in
the domestic market. Succeeding inter-
nationally requires cultural and economic
understanding no less than technological
- The Environment: Endangered Produc-
ing more in order to earn more will no
longer be the sole paradigm of industry.
The threats to quality of life resulting from
unrestrained environmental depredations
and the depletion of nonrenewable resources
are sources of growing concern, even within
industry. In addition to quality and produc-
tivity, industry will require that profitabil-
ity be achieved within a context of not harm-
ing people or their habitat. Increasingly, in-
dustries are adopting "The Natural Step"
process (TNS) or an equivalent to guide
their decision making about the global use
of the world's resources. 1" 21 The four prin-
ciples of TNS are
1. Substances extracted from the earth's



to better



lies in







to attack


that have





Winter 2000

Special Feature Section

crust (such as oil, fossil fuels, metals, and other min-
erals) must not systematically accumulate in the eco-
sphere. That is, the rate of mining from the earth's
crust must not occur at a pace faster than the ex-
tracted species can be redeposited and reintegrated
into the earth's crust.
2. Substances produced by society must not systemati-
cally increase in the ecosphere. That is, synthetic
substances must not be produced at a rate faster than
they can be broken down and integrated into natural
3. The physical conditions for productivity and assimi-
lation within the ecosystem cannot be systematically
diminished. Forests, wetlands, prime agricultural land,
natural plants, and animals cannot be systematically
4. Since resources are limited, basic human needs must
be met with the most resource-efficient methods avail-
able. Industrialized nations cannot use the resources
to create luxuries while the basic needs of people in
underdeveloped nations are not being met.

- Social Responsibility: Emerging Technology is respon-
sible for much of what we value about our society and our
way of life, but it must also take responsibility for the
threats to public health and the depletion of nonrenew-
able natural resources that now endanger that way of life.
The historical thrust of technological development has
been to increase consumption and profit; we are falling
well short of where we should be in our ability to provide
adequate health care, efficient public transportation, af-
fordable housing, and quality education for all citizens.
We are not bridging the gap between the technologically
advanced societies and those that do not have even the
basic means for survival. While the origins of many of
these problems may be political rather than technologi-
cal, it is up to scientists and engineers to participate in the
decision-making processes to a greater extent than ever
before. We have obligations to inform ourselves and the
rest of the population about the potential social conse-
quences of the decisions that are made, to judge whether
the implementation of decisions is consistent with the
objective of technology to improve our well-being for
citizens of the world (as outlined in TNS principle #4),
and to take appropriate action or choose inaction, depend-
ing on the outcome of the judgment. Acceptance of this
social responsibility by industry and individual engineers
is a necessary step for the survival of our society in the
next century. A corporate culture consistent with the four
principles of TNS, or equivalent, is needed.
0 Corporate Structures: Participatory Companies in dif-
ferent societies are moving toward structures that allow

for greater participation of individuals in the decision-
making process. Quality circles, small-group planning,
and troubleshooting sessions with joint participation by
management, technical, and operational staff are increas-
ingly common. Layers of middle management have been
eliminated, with much of the decision-making power be-
ing transferred downward to a broader spectrum of the
corporate body. Individual employees are acquiring to an
increasing extent the right to take part in decisions that
relate to their jobs and to assume responsibility for the
consequences of those decisions.
> Change: Rapid Changes of a magnitude that not long
ago would have taken years now occur on a time scale of
months or weeks, as anyone who purchased a computer
over one year ago realizes. Curricula that attempt to re-
main current with industrial practice by continually pro-
viding courses in the "new technology" are likely to be
ineffective. By the time the need is identified, the courses
developed, and the students trained, the new technology
has changed. The education that succeeds will be the one
that facilitates lifelong learning, equipping students with
the skills they will need to adapt to change.

What can we say about the individuals needed to function
as engineers in the society whose technological characteris-
tics we have just outlined? Their profiles may be conve-
niently sketched in terms of three components: (1) their
knowledge-the facts they know and concepts they under-
stand; (2) the skills they use in managing and applying their
knowledge, such as computation, experimentation, analysis,
synthesis/design, evaluation, communication, leadership, and
teamwork; (3) the attitudes that dictate the goals toward
which their skills and knowledge are directed-personal
values, concerns, preferences, and biases. Knowledge is the
data base of a professional engineer; skills are the tools used
to manipulate the knowledge in order to meet a goal dictated
or strongly influenced by attitudes.
In its early years, engineering education did a good job of
transmitting knowledge to engineering students, and it might
be argued that it facilitated the development of skills and
promoted values in ways appropriate for the time. Until
about thirty years ago, most engineering professors had ei-
ther worked in industry or consulted extensively, and the
facts and methods that constituted the knowledge base of the
engineering curriculum were by and large those that the
students would need in their careers. The tasks most engi-
neers were called upon to perform involved mostly routine
and repetitive calculations. Engineering students developed
and sharpened the requisite skills by working through nu-
merous laboratory exercises and industry-designed case stud-
ies and by participating in cooperative industrial work-study
Chemical Engineering Education

Future of Engineering Education

The circumstances facing practicing engineers today are considerably different from those of the past,
and the circumstances of the future will be even more different. Significant changes in
engineering education will be required if we are to meet the needs of our
graduates in preparing them for the challenges of the coming century.

programs and practice schools. The primary values of engi-
neering practice at the time were functionality and profit. A
good process was one that did what it was supposed to do in
as profitable a manner as possible. Both the engineering
curriculum and the faculty reinforced these values.
The circumstances facing practicing engineers today are
considerably different from those of the past, and the cir-
cumstances of the future will be even more different. Sig-
nificant changes in engineering education will be required if
we are to meet the needs of our graduates in preparing them
for the challenges of the coming century. Let us consider in
somewhat greater detail the knowledge, skills, and values
that will be necessary for engineers to deal successfully with
the challenges raised in the previous section.

4 Knowledge N
The volume of information that engineers are collectively
called upon to know is increasing far more rapidly than the
ability of engineering curricula to "cover it." Until the early
1980s, for example, most chemical engineering graduates
went to work in the chemical or petroleum industry. Now
they are increasingly finding employment in such nontradi-
tional (in chemical engineering) fields as biotechnology,
computer engineering, environmental science, health and
safety engineering, semiconductor fabrication technology,
and business and finance. To be effective across this broad
spectrum of employment possibilities, our graduates should
understand concepts in biology, physics, toxicology, fiscal
policy, and computer and software engineering that are well
beyond the range of the traditional chemical engineering
curriculum. Many who work in companies that have interna-
tional markets will also need to be conversant with foreign
languages, which have been phased out of both undergradu-
ate and graduate engineering curricula in recent decades. At
the same time, the work done by any one engineer tends to
occupy a relatively narrow band in the total spectrum of
engineering knowledge. Unlike their counterparts of several
decades ago, today's engineering students may never be
called upon to work with basic elements of the traditional
curriculum such as phase equilibria, thermodynamics, sepa-
rations, reactions, and process design.
For these reasons, structuring a four-year, or even a five-
year, engineering curriculum that meets the needs of most
engineering students appears to be an increasingly elusive
goal. One solution is to abandon the traditional one-size-fits-
all curriculum model and instead to institute multiple tracks
Winter 2000

for different areas of specialization, relegating some tradi-
tionally required courses to the elective category.""1 Design-
ing such tracks and keeping them relevant is a challenging
task, but it can be and is being done at many institutions.
No matter how many parallel tracks and elective courses
are offered, however, it will never be possible to teach engi-
neering students everything they will be required to know
when they go to work. A better solution may be to shift our
emphasis away from providing training in an ever-increas-
ing number of specialty areas to providing a core set of
science and engineering fundamentals,'[4] helping students
integrate knowledge across courses and disciplines,"'51 and
equipping them with lifelong learning skills.[16',71 In other
words, the focus in engineering education must shift away
from the simple presentation of knowledge and toward the
integration of knowledge and the development of critical
skills needed to make appropriate use of it.

4 Skills >
The skills required to address the challenges to future
engineers raised in the first section may be divided into
seven categories: (1) independent, interdependent, and life-
long learning skills; (2) problem solving, critical thinking,
and creative thinking skills; (3) interpersonal and teamwork
skills; (4) communication skills; (5) self-assessment skills;
(6) integrative and global thinking skills, and (7) change
management skills. From another perspective, ABET Engi-
neering Criteria 2000 requires that future graduates of ac-
credited programs should possess
(a) an ability to apply knowledge of mathematics,
science, and engineering
(b) an ability to design and conduct experiments, as
well as analyze and interpret data
(c) an ability to design a system, component, or
process to meet desired needs
(d) an ability to function on multidisciplinary teams
(e) an ability to identify, formulate, and solve engi-
neering problems
(f) an understanding of professional and ethical
(g) an ability to communicate effectively
(h) the broad education necessary to understand the
impact of engineering solutions in a global/
societal context

(Special Feature Section

(i) a recognition of the need for and an ability to
engage in lifelong learning
(j) a knowledge of contemporary issues
(k) an ability to use the techniques, skills, and modern
engineering tools necessary for engineering
In the following paragraphs we will suggest the parallels
between our proposed classification of skills and the ABET

Independent Learning, Interdependent Learning, and
Lifelong Learning Skills
(EC 2000 Outcomes a, d, e, and i)
Most students enter college as dependent learners, relying
on their instructors to present, organize, and interpret knowl-

... the focus in engineering
education must shift away from the
simple presentation of knowledge and
toward the integration of knowledge and the
development of critical skills needed to make
appropriate use of it.

edge. A model developed by Perry"19 describes the shift
many students undergo from being dependent learners to
independent learners to interdependent learners. Perry's model
includes nine levels, of which levels 2 to 5 characterize most
college students.[19-21]
In Perry's model, dependent learners tend to be dualists
(Level 2). In the dualist picture of the world, every point of
view is either right or wrong, all knowledge is known and
obtainable from teachers and texts, and the students' tasks
are to absorb what they are told and then demonstrate having
done so by repeating it back. A significant part of our re-
sponsibility as instructors is to move students from the de-
pendent stance to being independent learners, who realize
that all knowledge is not known and different points of view
may come in shades of gray rather than being either black or
white, and that their task is to acquire knowledge from a
variety of sources and subject it to their own critical evalua-
tion. Students at this level (which roughly corresponds to
Level 4 of Perry's model) should be able to identify the
pertinent factors and issues that affect a given situation, see
the situation from a variety of perspectives, recognize what
they need to know to resolve the situation, acquire the perti-
nent knowledge they do not already possess, and apply their
knowledge to achieve a successful resolution. They should
further be able to elaborate their knowledge so that future
recall and application will be easy. Evidence suggests that

some, but by no means all, students attain this level of
development by the time they graduate.[21-231
But the instructor's job does not end at this point. Students
should be helped to go beyond independent learning to in-
terdependent learning, recognizing that all knowledge and
attitudes must be viewed in context; that getting information
from a variety of sources is more likely to lead to success
than relying on a narrow range of sources and viewpoints,
and that the peer group can be a powerful learning resource.
These attitudes are characteristic of Level 5 on the Perry
scale. Students routinely work with peers to identify key
resources and to step through the superabundance of avail-
able information to identify what is really important, formu-
late learning objectives and criteria, assess the extent to
which they can believe what they read, and learn from and
communicate newly acquired information to others. In work-
ing with others, the students learn to recognize their own
learning styles, strengths, and weaknesses, and to take ad-
vantage of the synergy that comes from people with a diver-
sity of backgrounds and abilities working together toward a
common goal.1241
When students leave the university and enter the work
world, they can no longer count on teachers, textbooks, and
lectures to tell them what they need to know to solve the
problems they are called upon to solve. The only resources
they have access to are themselves and their colleagues. If
we help them to become independent learners, developing
and relying on their own reasoning ability rather than ac-
cepting information presented by others at face value, and
interdependent learners, using the strength of the group to
compensate for and overcome their own limitations, we will
be equipping them with the lifelong learning skills they will
need for success throughout their postgraduate careers.

Problem Solving, Critical Thinking, and Creativity
(EC 2000 Outcomes a, b, c, e, and k)
Some authors125'26' identify critical and creative thinking
as core skills that are applied to problem solving, while
others[23'27-32] define problem solving as the primary skill,
with critical and creative thinking as components.
Norman1331 questions whether "general" problem-solving
skills exist without subject context. Be all of that as it
may, to be considered effective problem solvers, our
students should be able to draw upon a wide range of
analytical, synthetic, and evaluative thinking tools,
problem-solving heuristics, and decision-making ap-
proaches. When given a problem to solve, they should be
equipped to identify the goal and put it in context;
formulate a systematic plan of attack that incorporates a
suitable blend of analysis, synthesis, evaluation, and
problem-solving heuristics; locate sources of information;
identify main ideas, underlying assumptions, and logical

Chemical Engineering Education

Future of Engineering Education

fallacies, and evaluate the credibility of the identified
sources; create numerous options, and classify and
prioritize them; make appropriate observations and draw
sound inferences from them; formulate and implement
appropriate measurable criteria for making judgments;
develop cogent arguments in support of the validity or
plausibility of a hypothesis or thesis; generate new
questions or experiments to resolve uncertainties; and
monitor their solution process continuously and revise it if

Interpersonal/Group/Team Skills
(EC 2000 Outcomes d, g, andf)
The image of the isolated engineer, working in solitary
splendor on the design of a bridge or amplifier or distilla-
tion column, probably never was realistic. Engineering is
by its nature a cooperative enterprise, done by teams of
people with different backgrounds, abilities, and responsi-
bilities. The skills associated with successful teamwork-
listening, understanding others' viewpoints, leading
without dominating, delegating and accepting responsibil-
ity, and dealing with the interpersonal conflicts that
inevitably arise-may be more vital to the success of a
project than technical expertise. Being aware of others'
needs and taking them into consideration when making
decisions-the essence of teamwork-is surely a prere-
quisite to functioning professionally and ethically,
regardless of how these terms are interpreted, and is
consequently a necessary condition for the fulfillment of
EC 2000 Outcome f.

Communication Skills
(EC 2000 Outcomes d, g, and h)
The teamwork necessary to confront the technological
and social challenges facing tomorrow's engineers will
require communication skills that cross disciplines,
cultures, and languages. Engineers will have to communi-
cate clearly and persuasively in both speaking and writing
with other engineers and scientists, systems analysts,
accountants, and managers with and without technical
training, within their company and affiliated with multina-
tional parent, subsidiary, and client companies, with
regulatory agency personnel, and with the general public.
Like all the other skills mentioned, effective communica-
tion is a skill that can be taught, but doing so requires a
conscious effort from those who design curricula.

Assessment and Self-Assessment Skills
(EC 2000 Outcomes d, f and i)
Gibbs1351 suggests that "whoever owns the assessment,
owns the learning." The more we can empower students
to assess accurately the knowledge and skills of others

Winter 2000

Engineering is by its nature
a cooperative enterprise, done by
teams of people with different backgrounds,
abilities, and responsibilities. The skills
associated with successful teamwork...
may be more vital to the success of
a project than technical expertise.

and their own knowledge and skills, the more effective
and confident they will become as learners. Moreover, as
professionals all of our graduates will receive perfor-
mance reviews and many will administer them to others.
Developing assessment skills could be an important
component of their preparation for professional practice.

Integration of Disciplinary Knowledge
(EC 2000 Outcomes a-e and h-k)
Chemical engineering students get used to solving
problems within the narrow context of individual courses.
They solve thermodynamics problems in the thermody-
namics course and heat transfer problems in the heat
transfer course, often never recognizing that the two
subject areas are intimately related. As professionals, on
the other hand, chemical engineers rarely solve "thermo-
dynamics problems" or "heat transfer problems." Rather,
they solve problems, drawing on knowledge from
thermodynamics and heat transfer and economics and
safety engineering and environmental science and any
other discipline that pertains. Doing this well requires
both generic problem-solving skills and integrated and
structured knowledge of the engineering curriculum.114'361
Thermodynamics and heat transfer should be seen as
related applications of the law of conservation of
energy and not as separate, self-contained subjects
taught at different times by different instructors using
different textbooks.

Managing Change
(EC 2000 Outcomes d, f h, j, and k)
The one certainty about engineering in the coming
decades is that it will change, because everything else will
change. The growth of technology will lead to rapid
product obsolescence and a decreasing need for engineers
to perform the tasks that occupied most of them for most
of this century, and also to a growth in nontraditional job
markets for engineers, especially in the international
arena. Industries that lack the capacity to adapt and
change to shifting markets and new technologies will
not survive, and successful engineers will be those
who can manage change, especially when change is
thrust upon them.

SSpecial Feature Section

with few
become totally
dependent on
research funds
to support
most of their
has dictated the
establishment of
achievement as
the primary
criterion for
advancement up
the faculty
ladder, and the
potential for
as the primary
criterion for
tend to put
minimal effort
into teaching
so that they
can concentrate
on research,
which they view
correctly) as
the key to
their career

4 Attitudes and Values
Vesilind1371 says that the most lasting effect
of education on students is the maturation of
their values and ethical sense. Essays on this
subjecti38"44 suggest that engineers should be
inculcated with the values of willingness to
cooperate, concern for the preservation of the
environment, coequal commitment to qual-
ity and productivity, and involvement in ser-
vice to others.
The fallacious assumption of those who de-
signed the engineering curricula of the past
half-century seems to have been that including
several humanities courses should be sufficient
to produce responsible and ethical engineers.
The failure of the engineering curriculum to
address attitudes and values systematically has
had unfortunate consequences. Engineers often
make decisions without feeling a need to take
into account any of the social, ethical, and moral
consequences of those decisions, believing that
those considerations are in someone else's pur-
view. By default, the decisions have conse-
quently become the exclusive province of
economists and politicians, who lack the abil-
ity to predict or evaluate their consequences.
The social penalties discussed in the intro-
ductory section have been the results of this
development. EC 2000 Outcomes f (an under-
standing of professional and ethical responsi-
bility), h (the broad education necessary to un-
derstand the impact of engineering solutions in
a global/societal context), and j (a knowledge
of contemporary issues), and in part, i (a recog-
nition of the need for lifelong learning), arose
from a perceived need to correct the situation.

In the traditional approach to teaching, the
professor lectures and assigns readings and well-
defined convergent single-discipline problems,
and the students listen, take notes, and solve
problems individually. Alternative pedagogi-
cal techniques have repeatedly been shown to
be more effective and much more likely to
achieve the objectives set forth in the preced-
ing section. Among those techniques are coop-
erative (team-based) learning, inductive (dis-
covery) learning, the assignment of open-ended
questions, multidisciplinary problems and prob-
lem-formulation exercises, the routine use of

in-class problem-solving, brainstorming, and
troubleshooting exercises, and other methods de-
signed to address the spectrum of learning styles
to be found among students in every class.145-47]
The superiority of the alternative methods at
achieving desired cognitive and affective educa-
tional outcomes has been demonstrated in thou-
sands of empirical research studies124'45-491 and is
heavily supported by modem cognitive science.1501
Nevertheless, straight lecturing and convergent
problems continue to predominate in engineering
courses at most institutions. A substantial num-
ber of engineering professors are still unaware of
alternative educational methods, and many who
are aware of them choose not to incorporate them
into their approach to teaching. There are several
likely reasons for this inertia, aside from the in-
evitable human resistance to change.
Modem universities have, with few exceptions,
become totally dependent on research funds to
support most of their functions, including educa-
tional and administrative functions only margin-
ally related to research. This circumstance has
dictated the establishment of research achieve-
ment as the primary criterion for advancement up
the faculty ladder, and the potential for research
achievement as the primary criterion for faculty
hiring. In consequence, many young faculty mem-
bers either have little interest in doing high-qual-
ity teaching or would like to do it but feel that
they cannot afford to invest the necessary time.
Individuals in both categories tend to put mini-
mal effort into teaching so that they can concen-
trate on research, which they view (generally
correctly) as the key to their career success. More-
over, most professors begin teaching without so
much as five minutes of training on how to do it.
Even those who are genuinely concerned about
their students and would like to be effective teach-
ers automatically fall back on straight lecturing,
which is the only instructional strategy most of
them have ever seen.
Another obstacle to change is the fear of loss of
control. Lecture classes in which student involve-
ment is essentially limited to passive observation
(perhaps broken by occasional questioning) and
out-of-class problem solving is safe: the profes-
sor is in almost complete control of what happens
in class. On the other hand, it is hard to predict
what might happen in a student-centered class.
Digressions may occur, making it difficult to stay
with the syllabus, and the discussion may wander

Chemical Engineering Education

Future of Engineering Education

into areas where the professor is not all that comfortable.
Perhaps worst of all, the students simply may not buy into
the program, remaining indifferent, uncooperative, or per-
haps hostile in their refusal to get involved in the planned
activities.51'521 Like any other skill, directing student-cen-
tered classes is an ability that can be learned and im-
proves with practice. Unless some training is provided
and feedback given on initial efforts, however, profes-
sors courageous enough to try the new teaching methods
are likely to become discouraged, give up, and revert to
straight lecturing.
In short, no matter how effective they may be, the new
approaches to teaching will not automatically replace the old
approach. The university administration must take steps to
establish a suitable climate for change before any significant
change can take place.

As imposing as the obstacles to change may be, we do not
believe they are insuperable, and indeed several things are
happening that are conducive to change.'1' As noted at the
beginning of this article, legislatures and industry have been
exerting increasing pressure on universities to pay more
attention to the quality of their undergraduate teaching
programs, and growing competition for a shrinking pool
of applicants for engineering school has provided further
impetus for change.
In the United States, the new ABET criteria were devel-
oped in response to these stimuli, and the knowledge that in
a short time they will be used to evaluate all engineering
programs is substantially increasing the pressure to change.
Moreover, major support for educational reform has come
from the National Science Foundation Division of Under-
graduate Education and the NSF-sponsored Engineering Edu-
cation Coalitions. This support has led to the emergence of a
large and rapidly growing number of innovative programs
and instructional methods and materials in the past decade,
as a perusal of recent issues of the Journal of Engineering
Education makes abundantly clear.
Finally, since both the National Science Foundation and
ABET insist on accountability, both traditional and innova-
tive instruction are being subjected to serious assessment
and evaluation. The presence of hard evidence to support
claims of improvement in learning should make it easier to
disseminate educational reforms to the skeptical mainstream
engineering professoriate.

The changes that will move engineering education in the
desired directions may be grouped into four categories: (1)
revisions in engineering curriculum and course structures;

(2) implementation of alternative teaching methods and as-
sessment of their effectiveness; (3) establishment of instruc-
tional development programs for faculty members and gradu-
ate students; and (4) adoption of measures to raise the status
of teaching in society and in institutional hiring, advance-
ment, and reward policies. In the next paragraphs, we will
propose questions that should be addressed in each of these
categories. The remaining papers in this series will be de-
voted to suggesting answers.

Engineering Curricula and Courses
* What is the appropriate balance between "fundamen-
tals" and "applications"? Should individual courses
stress one of these or the other, or should the two be
integrated within courses? Should the flow within a
course or curriculum generally proceed from funda-
mentals to applications (deductive presentation,
expository teaching) or from applications tofunda-
mentals (inductive presentation, discovery learning,
problem-based learning)?
What steps can be taken to integrate class material
across courses and disciplines, so that engineering
students become accustomed to thinking along
interdisciplinary lines in their approach to problem
solving? How can "clusters of concepts" be presented
systematically throughout the curriculum?
How should the development of critical skills-those
we outlined in this paper, and the overlapping set
defined in ABET Engineering Criteria 2000-be
facilitated in the curriculum? How much should be
done within core engineering courses and how much
should be relegated to specialized courses in such
things as communication and ethics?

Teaching Methods
What forms of in-class activities, homework assign-
ments, laboratory exercises, and testing and grading
policies and procedures, have been found most
effective at increasing knowledge and critical skills
and at promoting and reinforcing positive professional
What is an appropriate balance between teacher-
centered and student-centered instruction? Between
cooperative and individual learning? Between active
experimentation and reflective observation? Between
abstract concepts and concrete information? Between
routine drill and high-level thinking problems, and
between convergent (closed-ended) and divergent
(open-ended) problems? How can these balances be
achieved in practice?
How can students be motivated to be self-directed

Winter 2000

Special Feature Section

learners? How can they be helped to overcome the
resistance many of them feel to approaches that make
them take more responsibility for their own learning?
*How might we overcome faculty reluctance to try
something new in the classroom?

Instructional Development
What material should instructional development
("teacher training") programs cover? How much
should be generic, and how much should be specific to
Should the programs be mandatory or optional for
faculty members? For graduate teaching assistants?
For all PhD candidates?
What do instructional development programs cost?
How can they be financed?
How do the different types of programs (seminars,
workshops, courses) compare in effectiveness at
improving teaching? In cost effectiveness?

Faculty Hiring, Advancement, and Rewards
Does the requirement that every engineering professor
be a disciplinary researcher to enjoy full departmental
citizenship have a logical basis? Does it improve a
university's teaching program? Its research program?
Who will teach engineering practice in the coming
years as the number of engineering professors with
industrial experience continues to shrink? Who will
write undergraduate textbooks? Advise undergradu-
ates? Teach design? Keep the undergraduate labora-
tory running and periodically modernize it? Can
adjunct professors fill these roles? Should they?
Who will develop innovative and effective teaching
methods in the future, do the research to validate them,
and help other faculty members implement them?
Is it possible to assure that every engineering depart-
ment has at least a few individuals who can perform
the preceding tasks with dedication and skill? Can
engineering education survive without such individu-
als? What incentives, rewards, and policies will be
required to hire and keep them on our faculties? Can
their presence be maintained without completely
overturning the current financial structure of the
university, which depends so heavily on research

We have described many concerns and trends in this pa-
per. The key idea is that traditional instructional methods

will probably not be adequate to equip engineering gradu-
ates with the knowledge, skills, and attitudes they will need
to meet the demands likely to be placed on them in the
coming decades, while alternative methods that have been
extensively tested offer good prospects of doing so.

We are grateful to B.W. Baetz (McMaster University),
John O'Connell (University of Virginia), Angelo Perna (New
Jersey Institute of Technology), Tom Regan (University of
Maryland), Antonio Rocha (Instituto Technol6gico de
Celaya), D.E. Roy (McMaster University), and Wallace
Whiting (University of Nevada-Reno) for helpful reviews of
a draft of this paper.

1. Felder, R.M., "The Warm Winds of Change," Chem. Eng.
Ed., 30(1), 34 (1996)
2. Smith, P., Killing the Spirit: Higher Education in America,
Viking, New York, NY (1990)
3. Employability Skills Profile: What are Employers Looking
For? The Conference Board of Canada, Ottawa (1993)
4. Bradford School of Technical Management, Managerial
Skills and Expertise Used by Samples of Engineers in Brit-
ain, Australia, Western Canada, Japan, The Netherlands,
and Norway, University of Bradford, Technical Report TMR
152, Bradford, UK (1984)
5. Evers, F., J. Rush, and I. Berdrow, The Bases of Compe-
tence: Skills for Lifetime Learning and Employability,
Jossey-Bass, San Francisco, CA (1998)
6. Sparkes, J.J., "Quality in Engineering Education," Engi-
neering Professor's Conference, Occasional Paper #1, De-
partment of Mechanical Engineering, University of Surrey,
Guilford, Surrey, UK (1989)
7. White, J.A., "TQM: It's Time, Academia!" Education News,
3(2), 1, Nov. (1990)
8. Felder, R.M., and R. Brent, "How to Improve Teaching
Quality," Quality Management J., 6(2), 9 (1999)
9. Felder, R.M., "ABET Criteria 2000: An Exercise in Engi-
neering Problem Solving," Chem. Eng. Ed., 32(2), 126 (1998)
10. Meneses, E., "El Sistema UIA en el Umbral del Siglo XXI,"
Umbral, XXI(1), 2, Oct. (1989)
11. Examples of companies applying the principles outlined in
The Natural Step are given at the sites

12. Nattrass, B., and M. Altomare, The Natural Step for Busi-
ness: Wealth, Ecology and Evolutionary Corporation, New
Society Publishers, Gabriola Island, B.C., Canada (1998)
Available on-line at
13. Felder, R.M., "The Future ChE Curriculum: Must One Size
Fit All?" Chem. Eng. Ed., 21, 74 (1987)
14. "Knowledge Structure," series of articles in Chem. Eng.
Ed., 27(2), 78 (1993)
15. Prausnitz, J.M., "Integration of Knowledge is the Key to

Chemical Engineering Education

Future of Engineering Education

Versatility," Chem. Eng. Prog., p.7, Jan (1988)
16. Prausnitz, J.M., "Chemical Engineering and the Other Hu-
manities," Chem. Eng. Ed., 32(1), 14 (1998)
17. Longworth, N., and W.K. Davies, Lifelong Learning, Kogan
Page, London, UK (1996)
18. Details about EC 2000 can be found at the ABET Web site

19. Perry, Jr., W.G., Forms ofIntellectual and Ethical Develop-
ment in the College Years, Holt, Rinehart and Winston,
New York, NY (1968)
20. Felder, R.M., "Meet Your Students: 7. Dave, Martha, and
Roberto," Chem. Eng. Ed., 31(2), 106 (1997). Available on-
line at
21. Fitch, P., and R.S. Culver, "Educational Activities to Stimu-
late Intellectual Development in Perry's Scheme," 1984
ASEE Annual Conf Proc., American Society for Engineer-
ing Education, Washington, DC, 712 (1984)
22. Woods, D.R., Problem-Based Learning: How to Gain the
Most from PBL, Waterdown (1994). Distributed by
McMaster University Bookstore, Hamilton, Ont., Canada,
with additional guidebooks available from
23. Woods, D.R., et al., "Developing Problem Solving Skills:
The McMaster Problem Solving Program," J. Eng. Ed.,
86(2), 75, April (1997). Additional details about the MPS
program may be obtained from

24. Johnson, D.W., R.T. Johnson, and K.A. Smith, Active Learn-
ing: Cooperation in the College Classroom, 2nd ed., Interac-
tion Book Co., Edina, MN (1999)
25. Swartz, R.J., and D.N. Perkins, Teaching Thinking: Issues
and Approaches, Midwest Publications, Pacific Grove, CA
26. Paul, R., Critical Thinking: What Every Person Needs to
Survive in a Rapidly Changing World, 2nd ed., Foundation
for Critical Thinking, Santa Rosa, CA (1992)
27. Bransford, J.D., and B.S. Stein, The IDEAL Problem Solver,
W.H. Freeman and Co., San Francisco, CA (1984)
28. Hayes, J.R., Complete Problem Solver, 2nd ed., Lawrence
Erlbaum Associates, Mahwah, NJ (1989)
29. Fogler, H.S., and S.E. LeBlanc, Strategies for Creative Prob-
lem Solving, McGraw-Hill, New York, NY (1995)
30. Wicklegren, W.A., How to Solve Problems, W.H. Freeman
and Co., San Francisco, CA (1974)
31. Woods, D.R., "Problem Solving in Practice," in What Re-
search Says to the Science Teacher: Vol. 5. Problem Solving,
National Science Teachers Association, Washington, DC
32. Stice, J.E. (ed.), Developing Critical Thinking and Prob-
lem-Solving Abilities, New Directions in Learning and
Teaching Series, No. 30, Jossey-Bass, San Francisco, CA
33. Norman, G.R., "Problem-Solving Skills, Problem Solving,
and Problem-Based Learning," Med. Ed., 22, 279 (1988)
34. Kurfiss, J.G., Critical Thinking: Theory, Research, Prac-
tice, and Possibilities, ASHE-ERIC Higher Education Re-

port No. 2, Association for the Study of Higher Education,
Washington, DC (1988)
35. Gibbs, G., The A to Z of Student-Focused Teaching Strate-
gies, Educational Methods Unit, Oxford Polytechnic,
Headington, UK (undated)
36. Glaser, R., "Education and Thinking: The Role of Knowl-
edge," Amer. Psychologist, 55, 2 (1984)
37. Vesilind, P.A., "Rules, Ethics, and Morals in Engineering
Education," Eng. Ed., 78, 289 (1988)
38. Mancillas, W.R., and E.A. Sisson, "Cheating Among Engi-
neering Students," Chem. Eng. Ed., 23, 16 (1989)
39. Mingle, J.O., "Inchemacy: An Ethical Challenge for Chemi-
cal Engineers," Chem. Eng. Prog., p. 17, Aug (1989)
40. Morgan, R.P., "Educating Scientists and Engineers: The
View of OTA," Eng. Ed., 79, 17 (1989)
41. Rugarcia, A., "Metodologia de la Ensefianza de las Ciencias
y la Ingenieria," DIDAC, University Iberoamericana,
Primavera, p. 1 (1986)
42. Rugarcia, A., "El Profesional de Ingenieria Quimica ante
los Retos del Futuro," Revista del IMIQ, p. 37, Nov-Dec
43. Rugarcia, A., "Desarrollo de la Creatividad en la Formaci6n
de Ingenieros," Ed. Quimica, p. 1, Jan (1991)
44. Rugarcia, A., "Mitos y Cre6ncias en la Docencia de la
Ingenieria," Revista del IMIQ, p. 42, May/June (1991)
45. McKeachie, W.J., Teaching Tips: Strategies, Research, and
Theory for College and University Teachers, 10th ed.,
Houghton Mifflin, Boston, MA (1999)
46. Wankat, P., and F.S. Oreovicz, Teaching Engineering,
McGraw Hill, New York, NY (1993). Available on-line at

47. Felder, R.M., and L.K. Silverman, "Learning and Teaching
Styles in Engineering Education," Eng. Ed., 78(7), 674
48. Felder, R.M., G.N. Felder, and E.J. Dietz, "A Longitudinal
Study of Engineering Student Performance and Retention.
V. Comparisons with Traditionally Taught Students," J.
Eng. Ed., 87(4), 469 (1998). Available on-line at
49. Springer, L., M.E. Stanne, and S. Donovan, "Effects of
Small-Group Learning on Undergradutes in Science, Math-
ematics, Engineering, and Technology: A Meta-Analysis,"
Rev. Ed. Res., 69(1), 21 (1999). Available on-line at

50. Haile, J.M., "Toward Technical Understanding: Part 1. Brain
Structure and Function," Chem. Eng. Ed., 31(3), 152 (1997);
"Part 2. Elementary Levels," Chem. Eng. Ed., 31(4), 214
(1997); "Part 3. Advanced Levels," Chem. Eng. Ed., 32(1),
51. Benvenuto, M., "In the Age of Interactive Learning, Some
Students Want the Same Old Song and Dance," Chronicle
of Higher Ed., p. 89, June 4 (1999)
52. Felder, R.M., and R. Brent, "Navigating the Bumpy Road
to Student-Centered Instruction," Col. Teach., 44(2), 43
(1996). Available on-line at
public/Papers/Resist.html> O

Winter 2000

SSpecial Feature Section


Part 2. Teaching Methods that Work

Richard M. Felder North Carolina State University Raleigh, NC 27695-7905
Donald R. Woods McMaster University Hamilton, Ontario, Canada L8S 4L7
James E. Stice University of Texas Austin, TX 78712-1062
Armando Rugarcia Iberoamericana University Puebla, Mexico

Deficiencies in engineering education
have been exhaustively enumerated
in recent years. Engineering schools
and professors have been told by countless pan-
els and blue-ribbon commissions and, in the
United States, by the Accreditation Board for
Engineering and Technology that we must
strengthen our coverage of fundamentals; teach
more about "real-world" engineering design and
operations, including quality management;
cover more material in frontier areas of engi-
neering; offer more and better instruction in
both oral and written communication skills and
teamwork skills; provide training in critical and
creative thinking skills and problem-solving
methods; produce graduates who are conver-
sant with engineering ethics and the connec-
tions between technology and society; and re-
duce the number of hours in the engineering
curriculum so that the average student can com-
plete it in four years.0'

S.. even if nothing
new is added
to the existing
confining it to
four years will be
almost impossible
unless more
efficient and
effective ways to
cover the material
can be found....
The reality is
that better teaching

This is an impressive wish list-especially when the last
item is included-that cannot possibly be fulfilled using the
approach to educating engineers that has predominated in
the past fifty years. If, for example, courses continue to be
confined to single subjects (heat transfer in one course,
thermodynamics in another, environmental engineering in
another, technical writing in another, etc.), it will take a six-
or seven-year curriculum to produce engineers who have the
desired proficiency in the fundamentals and are conversant
with methods of modern engineering practice, culturally
literate, and skilled in communication. Moreover, if students
are assigned only well-defined convergent problems, they
will never gain the skills needed to tackle and solve chal-
lenging multidisciplinary problems that call for critical judg-

ment and creativity. Finally, even if nothing
new is added to the existing curriculum, con-
fining it to four years will be almost impossible
unless more efficient and effective ways to
cover the material can be found.
The reality is that better teaching methods
exist. The literature in general education, tech-
nical education, and educational psychology is
replete with methods that have been shown to
facilitate learning more effectively than the tra-
ditional single-discipline lecturing approach.
Unfortunately, these developments have so far
had relatively little impact on mainstream en-
gineering education. Although their content has
changed in some ways and the students use
calculators and computers instead of slide
rules, many engineering classes are taught
in exactly the same way that engineering
classes in 1960 were taught.
The purpose of this paper is to offer alterna-

tives. The instructional methods to be described have been
chosen to meet the following criteria:
* They are relevant to engineering education.
Many innovative instructional methods have been
developed for nontechnical courses and emphasize
free discussion and expressions of student opinions,
with minimal teacher-centered presentation of
information. We believe that involvement of students
is critical for effective classroom learning; however,
much of the basic content of engineering courses is
not a matter of opinion. Educational approaches that
emphasize process exclusively to the detriment of
content will not be considered.
They can be implemented within the context of the

Copyright ChE Division of ASEE 2000

Chemical Engineering Education

Future of Engineering Education

ordinary engineering classroom.
An instructional approach based entirely on, say, self-
paced computer-assisted instruction might be
extremely effective-at least for some students-but
it might also require a specialized network of
workstations that could cost an institution several
million dollars to purchase and set up. Such programs
will be left off the list. The techniques we describe
can be implemented in regular classrooms and
laboratories with no tools or devices beyond those
routinely available to all engineering instructors.

Most engineering professors should feel reasonably
comfortable with them after a little practice.
It is conceivable, for example, that getting students to
role-play molecules in a reactive gas would teach
them more about the dynamic behavior of a given
system than would a standard lecture. Some instruc-
tors find methods like this useful and can manage to
pull them off; still, it is safe to say that most engi-
neering professors would never contemplate doing
anything like that in their classes. Such methods will
not be included in our list of recommendations.

They are consistent with modern theories of learning
and have been tried and found effective by many
The literature is full of articles by professors who
have tried new methods and written about the results.
But the validity of a method must remain suspect if
the only evidence on its behalf is one person's
testimony that "I tried this and liked it and so did the
students." The methods to be given are consistent
with results of theoretical and/or empirical studies in
the cognitive and educational psychology literature,
and they have each been implemented successively in
engineering classes by independent investigators.

This paper surveys some (but by no means all) instruc-
tional methods that meet these criteria. Several excellent
references describe other techniques and summarize the sup-
porting research.12'4


Instructional objectives are statements of what students
should be able to do to demonstrate their mastery of course
material and desired skills. They contain a stem specifying
the point at which the mastery should occur, followed by one
or more phrases describing the expected behavior, with each
phrase beginning with an action verb. For example

When this chapter has been competed, the student should be able to
define the variables in the ideal gas equation ofstate in terms a high
school senior could understand, calculate the value of any one of
the variablesfrom given values of the other three, estimate the error
in the calculated values, and outline the derivation of the ideal gas
equation from the kinetic theory of gases.
The common stem of the four objectives in this paragraph is
"When this chapter has been completed." An alternative
stem might be "In order to do well on the next test." The
phrases that define the objectives begin with the verbs de-
fine, calculate, estimate, and outline. Other acceptable verbs
include list, identify, explain (without using jargon), predict,
model, derive, compare and contrast, design, create, select,
optimize, and many others.
The behavior specified in an instructional objective must
be directly observable by the instructor and should be as
specific and unambiguous as possible. For this reason, verbs
such as know, learn, understand, and appreciate are unac-
ceptable. These are critically important goals, but they are
not directly observable. For example, if an instructor states
that her goal is for her students to understand the first law of
thermodynamics, she might be asked how she will know
whether or not they do. She would then list the things she
would ask them to do to demonstrate their understanding.
The items on the list would constitute the instructional
objectives associated with the specified goal. If there
could be any possible doubt about whether or not an
objective has been met, metrics should be included in the
defining statement.
Instructional objectives may involve skills that cover a
broad spectrum of complexity and difficulty. The book Tax-
onomy of Educational Objectives (Cognitive Domain) de-
veloped by Bloom and colleagues1101 defines a hierarchy of
six levels:
1. Knowledge-repeating memorized information
2. Comprehension-paraphrasing text; explaining
concepts in jargon-free terms
3.Application-applying course material to solve
straightforward problems
4.Analysis-solving complex problems; developing
process models and simulations; troubleshooting
equipment and system problems
5. Synthesisdesigning experiments, devices, processes,
and products
6. Evaluation-choosing from among alternatives and
justifying the choice; optimizing processes; making
judgments about the environmental impact of engineer-
ing decisions; resolving ethical dilemmas

Levels 1 through 3 are commonly known as "lower-level
skills" and Levels 4 through 6 are "higher-level skills."
Most undergraduate engineering courses focus on Level-3

Winter 2000

SSpecial Feature Section

skills: an analysis of one four-year engineering program
showed that 2345 out of 2952 problems assigned (79%)
were Level 3 or lower.1"' On the other hand, probable de-
mands on engineering graduates in the coming decades and
many of the new ABET accreditation criteria (Engineering
Criteria 2000) involve skills at Levels 4 through 6.[11
Write instructional objectives for a course (or a section of
a course) that encompass both knowledge of content and
mastery of the skills you wish the students to develop. At all
levels of the engineering curriculum-including the first
year-include some higher-level problem-solving skills (e.g.,
multidisciplinary analysis, design, critical thinking) and the
"soft" skills (e.g., oral and written communication, team-
work, social and ethical awareness) specified in EC 2000.
Make the objectives as detailed and specific as possible;
rather than simply saying that the student should be able to
"design a chemical plant," list all the different things the
student will be expected to do (look up, estimate, calculate,
create, analyze, select, explain) when designing the plant.
Make class exercises, homework assignments, and tests con-
sistent with the objectives. Give the objectives to the stu-
dents to use as study guides.

Once formulated, instructional objectives reveal which
course topics are most important and deserve the greatest
coverage, and which involve little else than memorization
and thus merit only cursory attention or possible elimination
from the curriculum. Objectives enable instructors to design
consistent homework assignments that provide practice in
all of the desired skills and tests that assess mastery of the
skills. They make ideal study guides for the students; the
more explicit you are about what you want the students to be
able to do, the more likely they will be to succeed at doing
it." 21 The objectives provide an excellent outline of the course
content, for instructors teaching the course for the first time
as well as instructors of subsequent courses. Finally, the
instructional objectives for all departmental courses collec-
tively reveal gaps and redundancies in the curriculum and
provide an excellent curriculum overview to accreditation
visitors, especially if homework assignments and tests closely
follow the objectives.

Instructors often start a course by presenting totally new
material without putting it in any context. They make no
attempt to relate the material to things students already know
about from their own experience or from prior courses, nor
do they preview how it will be needed to solve problems of

Once formulated, instructional
objectives reveal course topics that are
most important and deserve the greatest
coverage, and which ones involve little else
than memorization and thus merit only cursory
attention or possible elimination
from the curriculum.

the types the students will encounter later in the curriculum
or in professional practice. These instructors are pursuing
what might be called the "Trust Me" approach to education
(as in "Trust me-what I'm teaching you may seem point-
less now, but in another year, or perhaps in four years, you'll
see why you needed it.").
Begin teaching each course and each new topic within it
by describing the physical and chemical phenomena to be
studied and the types of problems to be solved, using ex-
amples familiar to the students if possible. Discuss several
realistic situations in which engineers and scientists are re-
quired to understand the phenomena and solve the problems.
A good way to begin is to divide the class into groups of
three or four and have the groups generate as many examples
as they can think of in a brief period of time, adding your
own to supplement whatever they come up with. For ex-
For the next two weeks, we're going to be discussing characteristics
of afluidflowing through a pipe. In groups of three, come up with as
many situations as you can that involve this subject-three people
talking, one writing down the ideas. You have one minute-go!
Give them the allotted time (or a little more if they seem to
need it), then stop them and collect the ideas, listing them
without criticism. At least some of the groups are almost
certain to come up with home plumbing, irrigation, oil and
coolant flows in engines, municipal water and sewer flows,
flow of body fluids, and a variety of industrial examples.
Supplement their list with your own. You might then con-
Ok, you're now engineers designing a piping system to move fluid
from a storage tank to a reactor at a specified rate. What will you
need to know or figure out? Same groups, two minutes-go!
It may occur to some of the groups that they will need to
know the density and viscosity of the fluid, the distance from
the tank to the reactor, whether the fluid is corrosive or
dangerous in some way, the pipe material (aluminum, cop-
per, stainless steel, plastic), and costs of piping, pumps, and
power, and they will have to determine the pipe diameter,
the required valves, fittings, and flow meters, the kind of
pump to use, the size of the pump, and the path of the
system. Give hints if necessary, and add items to their list.
Spending ten minutes on such an exercise at the beginning of
Chemical Engineering Education

[ Future of Engineering Education )

a new topic can go a long way toward motivating the stu-
dents to pay attention to what takes place in the subsequent
two or three weeks.
The flow of information in the presentation of course
material should generally follow that of the scientific method:
begin with induction, proceeding by inference from specif-
ics (facts, observations, data) to generalities (rules, theories,
correlations, mathematical models), and then switch to de-

duction, using the rules and models to generate
additional specifics (consequences, applications,
Our goal in teaching is to get information and
skills encoded in our students' long-term memo-
ries. Cognitive research tells us that we learn
new material contextually, fitting it into existing
cognitive structures,['3-51 and new information
that cannot be linked to existing knowledge is
not likely to be retained. Moreover, once infor-
mation is stored in long-term memory, cues are
required for us to recall and use it. Linking the
new material to familiar material provides a
natural set of cues.
The motivational and learning benefits of pro-
viding context, establishing relevance, and teach-
ing inductively are supported throughout the lit-
erature on cognitive and educational psychology
and effective pedagogy."15'161 Ramsden and
Entwistle[121 note the motivational effective-
ness of "vocational relevance," and the same
authors show that establishing relevance is one
of the factors that induces students to adopt a
"deep" (as opposed to superficial) approach to

fore going into the details can provide the concrete experi-
ence that starts the learning cycle.


being cc

The problem
with introducing
[is that it is]
not firmly
grounded in
the student's
and experience
... the new
material is
not linked to
structures and
so is
to be
transferred to

Inductive teaching (wherein the information flow gener-
ally proceeds from specifics to generalities) takes several
forms in the literature, variously known as discovery learn-
ing, inquiry learning, problem-based learning, just-in-time
learning, and the case-study method. Problem-based learn-
ing (PBL), which involves students working in teams on
projects built around realistic problems, has been exten-
sively discussed and shown to be effective in science, engi-
neering, and medicine.118-231 (This approach will be treated in
greater detail in the next paper in this series.)
The literature on learning styles also supports the recom-
mendations in this section.24"331 Kolb127-29] suggests "teaching
around the cycle," starting with a concrete experience, docu-
menting observations, creating an abstract model, and then
experimenting and testing the model. This cycle has been
used to design a college-wide instructional program in engi-
neering.[30-311 Establishing the relevance of new material be-
Winter 2000

ial in engineering courses may be categorized as
ncrete (facts, observations, experimental data, ap-
plications) or abstract (concepts, theories, math-
ematical formulas, and models). Most engineer-
ing courses contain material in each category,
but the balance varies considerably from one
course to another and from one instructor to
another in a given course.
In recent decades, the balance between the two
categories in the engineering curriculum has been
shifting toward abstraction. The old courses on
industrial processes and machinery have been
largely replaced with courses that emphasize math-
ematical expressions of fundamental scientific
principles. While this movement may have ini-
tially had the effect of correcting an imbalance, it
has proceeded to an extent that has negative con-
sequences for many students. The problem with
introducing abstraction that is not firmly
grounded in the student's knowledge and ex-
perience has been described in the preceding
section; the new material is not linked to exist-
ing cognitive structures and so is unlikely to
be transferred to long-term memory.
Balance concrete and abstract content in the
presentation of all engineering courses. Most
courses currently contain a reasonable level of
abstraction, so the challenge is generally to pro-

vide sufficient concrete material for those who need it. Some
suggestions for doing so follow:
* Do everything listed under the category of establishing
relevance in the preceding section.
Intersperse concrete illustrations and applications throughout
theoretical developments rather than waiting until the final
formulas have been derived. When possible, tie the examples
back to the "real-world" systems and situations introduced in
the motivating introduction to the subject.
When illustrating how formulas and algorithms are applied,
use numbers rather than algebraic variables in at least the first
example. The greater the level of generality of the theory, the
greater the need for specificity in the examples. Some
students-specifically, sensing learners-understand "5" at a
level that they may never understand "x".125,32'331
Provide visual illustrations and demonstrations of course-
related material when possible. Most students get a great deal

Special Feature Section .

more out of visual information than verbal
information (written and spoken words and
mathematical formulas).1251 Show pictures,
sketches, schematics, plots and flow charts, and
computer simulations of process equipment and
systems. Take the class to the local boiler house
and point out pumps, flowmeters, boilers, heat
exchangers, refrigeration units, and turbines.
Bring demonstrations into class, such as those
described by Wood134' for heat transfer and
Kresta[351 for fluid mechanics.
SNever venture too far from the realm of
experimentation. In abstract subjects such as
thermodynamics and process control, for
example, it is easy for the students to drown in
an alphabet soup of variables that bear no
apparent relationship to anything one can
measure in a laboratory or plant (e.g., entropy,
free energy, and transfer functions). It is
important to remember that the ultimate goal of
all theories is to correlate data from measure-
ments on physical systems and to predict the
outcomes of future measurements. As each
abstract variable is introduced, provide examples
of how it could be determined experimentally
and how values of measured variables can be
predicted from known values of the abstract
variables, and give such problems as homework
assignments. Once the students have manipu-
lated a given variable or function in a variety of
contexts, its meaning can be assumed to be
anchored in memory, but in the absence of such
examples and exercises no such assumption can
be made.
Just as overemphasizing mathematical formu-
lations of course principles works against the
sensing learner, overemphasizing facts and com-
putational algorithms and shortchanging concep-
tual understanding works against intuitive learn-
ers.1331 (This concrete/abstract imbalance is also
not in the sensors' best interests, but it is less
likely to make them uncomfortable.) Engineer-
ing students are not generally overloaded with
spare time. If they can get away with memoriz-
ing problem solutions without understanding or
questioning the underlying concepts and meth-
ods, many will do it.1171
One way to help students gain a deeper under-
standing of course material is to ask questions
that require such an understanding, first in class
problems and homework and then on tests. For
Equation (8-34) in the textbook is presented with only
a sketchy explanation of where it comes from. Derive
it, starting with Eq. (8-5).
In Monday's handout there are a number of sugges-








when possible.


students get a

great deal

more out of



than verbal


(written and

spoken words








plots and flow

charts, and


simulations of



and systems.

tions to "prove" or "verify" some statement or result.
At least one of them will show up on the next test. I
won't go over them unless asked. (Or, I'll go over
them during my office hours, but only if you
demonstrate that you've attempted them yourself.)
Explain what a vapor pressure is in terms a high
school senior could understand.
Why do you feel comfortable in 20'C air and freezing
in 200C water? Your explanation should involve
several concepts introduced in this course.
Make up and solve a problem related to the material
just covered.136'37 The problem must be original, but
you can get ideas and help from one another and from
me. Start simply the first time you do this in class, and
gradually build in more depth. For example,
-Make up but don't solve a problem involving
Raoult's law.
-Make up and solve a problem involving Raoult's
Make up and solve a problem involving Raoult's
law. If your problem is straightforward (given
this, calculate that) and there are no mistakes,
you'll get a "C"; to earn full credit the problem
should involve a realistic situation.
-Make up and solve a problem that involves both
Raoult's law and what you covered during the
last two weeks of your organic chemistry course.
You may not get many good problems the
first time or two you do exercises like these, but
if you provide feedback and give examples of
successful efforts, many students will surprise
you (and themselves), both with the quality
of their problems and by how thoroughly they
learned the material in the course of the ex-
As noted in the previous section, a good way
to achieve concrete/abstract balance is to "teach
around the cycle."'26-311 When presenting a new
concept, start with a physical demonstration or
real-world example, model the results, test the
model through active experimentation, and ex-
plore its implications. You might also find it
worthwhile to have students measure their own
learning styles and talk about the implications.
The more they understand their own prefer-
ences, the more they can capitalize on the
strengths of their preferred styles and work
to build their capabilities in their less-pre-
ferred styles. Felder and Soloman's Index of
Learning Styles'381 and Keirsey's Tempera-
ment Sorter1391 are accessible on-line and easy
to use for this purpose.

Piaget'40] suggests that human capabilities
evolve in stages, beginning with the sensory-
Chemical Engineering Education

Future of Engineering Education ]

motor stage (up to age 2) and proceeding through pre-opera-
tional (ages 4 through 7) and concrete operational (about 7
to 12) stages to the formal operational stage. Concrete
operational thinkers can think logically in terms of objects,
but have difficulty replacing objects by symbols. They can
acknowledge different viewpoints and cause-effect logic,
but they have trouble generalizing through verbal or propor-
tional reasoning. Formal operational thinkers can replace
objects with symbols, generalize and work with abstract
concepts, use verbal and proportional reasoning, and derive
cause-effect relationships from results of experiments.
Piaget stated that the shift from concrete operational to
formal operational thinking should occur by age 12; but
more recent observations suggest that many first-year col-
lege students have not yet made it. Williams and Cavallo, 421
working with freshmen in physics courses, found that most
of their subjects were concrete operational, incapable of
grasping abstract concepts that were not firmly embedded in
concrete experience. By including concrete examples in our
teaching and explicitly showing how they can be general-
ized, we can help students make the shift from concrete to
formal operational thinking.[431
Learning-style differences also provide justification for
establishing a good concrete/abstract balance in every engi-
neering course.[24-26.3233] Sensing learners tend to be practical
and methodical; intuitors tend to be imaginative and quick-
thinking. Sensors are more comfortable with concrete infor-
mation (facts, data, "real-world" phenomena) than with ab-
stractions (theories, concepts, and models), and the converse
is true of intuitors. Both sensing and intuitive learners make
excellent engineers, although they tend to gravitate to differ-
ent specialties. Sensors make excellent experimentalists and
production engineers; intuitors do well in design and theo-
retical research and development, and both types may be-
come excellent managers and administrators. Industry and
academia need individuals with both type preferences.
Most engineering undergraduates are sensors, while most
engineering professors are intuitors.144'51 Most intuitive pro-
fessors, and even many of the sensing professors, teach in an
intuitor-oriented manner, emphasizing theories, mathemati-
cal models, and abstract prose to students who respond best
to concrete examples, well-established problem-solving pro-
cedures, and material that has a clear connection to the "real
world" (a classic sensor's phrase). This mismatch has sev-
eral unfortunate consequences for the sensing learners. Faced
with an incessant barrage of material that seems remote and
abstract, they have difficulty absorbing the material, become
bored in class, tend to do poorly on tests (frequently running
out of time on them) and tend to get lower grades in engi-
neering courses than their intuitive counterparts, even though
both types do equally well as practicing engineers.
Making courses overwhelmingly abstract is also a disser-
Winter 2000

vice to the intuitors. Even if they intend to go on to
graduate school and research careers, they need to
strengthen their sensing skills (observation of and atten-
tion to details, careful methodology, replication of mea-
surements and calculations), and they will not do so if
they are not challenged to do so in their courses.


In the traditional approach to higher education, the profes-
sor dispenses wisdom in the classroom and the students
passively absorb it. Research indicates that this mode of
instruction can be effective for presenting large bodies of
factual information that can be memorized and recalled in
the short term. If the objective is to facilitate long-term
retention of information, however, or to help the students
develop or improve their problem-solving or thinking skills,
or to stimulate their interest in a subject and motivate them
to take a deeper approach to studying it, instruction that
actively involves students has consistently been found
more effective than straight lecturing.[2,'346'471 The chal-
lenge is to involve most or all of the students in produc-
tive activities without sacrificing important course con-
tent or losing control of the class.

Several times during each lecture period, ask the students
to form into groups of 2 to 4 and give them brief exercises
that last anywhere from 30 seconds to 3 minutes. The exer-
cises may involve answering questions of the type instruc-
tors routinely ask the class as a whole, or they may call for
problem solving or brainstorming. For example,
Outline a strategy for solving the problem just posed.
Draw a flowchart (schematic)for the process just described.
Think of as many practical applications as you can of this (system,
device, formula).
Get started on the solution of the problem and see how far you can
get with it in two minutes.
What is the next step in the derivation?
Complete this calculation.
Prove or verify this result.
Suppose you carry out experimental measurements and the results
fail to agree with the theoretical formula we just derived. Think of as
many possible explanations as you can.
What questions do you have about this material?

The groups should generally be given a short time to re-
spond-long enough to think about the question and to
begin to formulate an answer, but not necessarily to work
out complete solutions.
Vary the format of these exercises to prevent their becom-
ing as tedious and ineffective as straight lecturing. Assign
some to pairs, some to groups of three or four, and some to

SSpecial Feature Section

individuals. Sometimes ask students to work on a problem
individually, and then compare their answers with a partner
("think-pair-share"). Sometimes give a rapid succession of
such exercises, and sometimes lecture for 10-15 minutes
between exercises.
To maximize the likelihood that most or all of the students
will be actively involved and that they will remain on task,
call on several individuals or groups to give their re-
sponses when the allotted time has elapsed. If you only
call for volunteers to share responses, the students will
know that the answer will eventually be forthcoming and
will have no incentive to participate in the activity-and
many will not; but if they know that any one of them
could be called on, fear of embarrassment will induce
most of them to do the work so they will be ready with
something if they are chosen.
Active learning methods make classes much more enjoy-
able for both students and instructors. Even highly gifted
lecturers have trouble sustaining attention and interest
throughout a 50-minute class. After 10-20 minutes in most
classes, the students' attention starts to drift, and by the end
of the class boredom is rampant. Even if the instructor asks
questions in an effort to spark some interest, nothing much
happens except silence and avoidance of eye contact. Tests
of information retention support this picture of what hap-
pens in terms of recall: immediately after a full lecture,
students were able to recall about 70% of the content
presented in the first ten minutes but only 20% of the
content of the last ten minutes.121
When active learning exercises are interspersed through-
out a lecture, the picture changes. Once a class accustomed
to group work gets started on a problem, the classroom
atmosphere is transformed: discussions, arguments, and oc-
casional laughter can be heard, all sounds of learning taking
place. Even students who may not be doing much talking are
engaged in thinking about the question at hand instead of
just mechanically transcribing notes. Just five minutes of
such activities in a 50-minute class can be enough to keep
the students attentive for the remaining 45 minutes of lectur-
ing. Many references offer specific suggestions for incor-
porating active learning exercises in the classroom.146-501
Felder151,521 and Woods153" discuss the implementation of
active learning in large classes, and Felder1511 discusses
how to incorporate active learning without sacrificing
content coverage.
Several authors have developed more formal active learn-
ing activities. One is "TAPPS" (thinking-aloud pair problem
solving), an activity where pairs of students take turns work-
ing their way through a problem solution;1541 another is the
"Osterman feedback lecture," where two 20-minute mini-
lectures are separated by a ten-minute activity, the latter
usually being a short problem that requires the students to

have learned certain material before class;[18] and still an-
other is "team learning," a more formal cooperative learning
structure where student teams work on structured learning
projects in every class session.'551 All of these techniques
require more time and training to implement than the brief
turn-to-your-neighbor exercises described previously, but
the potential return in depth of learning is greater.

Literature supporting the notion that active, student-cen-
tered learning is superior to passive, teacher-centered in-
struction is encyclopedic.[3'14'46-48] People acquire knowledge
and skills through practice and reflection, not by listening to
others telling them how to do something. Straight lecturing
may succeed at promoting short-term factual recall, but ac-
tive approaches have consistently been shown to be superior
for promoting long-term retention of information, compre-
hension, problem-solving skills, motivation to learn, and
subsequent interest in the subject. Active learning is one of
the seven, evidence-based recommendations for improving
learning summarized by Chickering and Gamson,[561 and the
active learning exercises described above also provide prompt
feedback, another of the recommendations.


Cooperative learning (CL) is an instructional approach in
which students work in teams on a learning task structured to
have the following features:[481
Positive independence. There must be a clearly defined group goal
(complete the problem set, write the lab report, design the process)
that requires involvement of every team member to achieve. If
anyone fails to do his or her part, everyone is penalized in some
Individual accountability. Each student in the team is held respon-
sible for doing his or her share of the work and for understanding
everyone else's contribution.
Face-to-face promotive interaction. Although some of the group
work may be parceled out and done individually, some must be done
interactively, with team members providing one another with
questions, feedback, and instruction.
Appropriate use of interpersonal and teamwork skills. Students
should be helped to develop leadership, communication, conflict-
resolution, and time-management skills.
Regular self-assessment of team functioning. Teams should
periodically be required to examine what they are doing well together
and what needs improvement.
Cooperative learning exercises may be performed in or out
of class. Common tasks for CL groups in engineering are
completing laboratory reports, design projects, and home-
work assignments in lecture courses. Only one problem set
or report is handed in by a group, and one group grade is
assigned to the project-but adjustments for individual team
citizenship (or lack thereof) can and should be made. Pre-
examination group study sessions can also be set up to
Chemical Engineering Education

Future of Engineering Education

meet out of class, with bonus points being awarded to
members of groups for which the team average test grade
exceeds a specified value.

The following suggestions are based on material in Johnson,
Johnson, and Smith,[481 Felder and Brent,'57'58 and Millis and
- Explain to students what you are do-
ing and why. As with in-class active In a mixed-al
learning methods, cooperative home- weaker stud
work may not be welcomed enthusi- seeing how
astically by all students. Some regard study and app
it as a game the instructor is playing at and the st
their expense or an experiment with a s
them as the guinea pigs, and some usually g
may complain that the instructor is understanding
not doing his or her job (which they through the
see as lecturing to them on everything explain th
they will need to know for the tests), phenomena
Felder and Brent"60' discuss the origin every I
and forms of student resistance to ac-
tive and cooperative learning and sug-
gest strategies for defusing and even-
tually overcoming the resistance. On the first day, twenty
minutes spent giving some of the reasons for using the
approach (e.g., it prepares students to function in the
environment in which engineers work) and explaining the
proven educational benefits to students (e.g., higher grades
and lower dropout rates) can go a long way toward over-
coming the resistance. Another option is to run a mini-
workshop on managing change."18'9"
Assign some or all homework to teams of 3-4 students. In
teams of two, one person tends to dominate and there is
usually no good mechanism for resolving disputes, and in
teams of five or more someone is usually left out of the
process. Collect one assignment per group.
Form the groups yourself. Considerable research shows
that instructor-formed teams on average function better
than self-selected teams. When students self-select groups,
the top students often find one another and form groups,
leaving the weak students to shift for themselves, which is
unfair. Also, good friends find each other, leading to
situations where their teammates are never fully inte-
grated into the team. Particularly in the freshman and
sophomore years, when most attrition from the curricu-
lum occurs, under-represented minorities (including
women) should not be isolated in teams. The ideal team is
heterogeneous in ability (which we will say more about
shortly), with team members who have common interests
and common blocks of time when they can meet outside
class. SAT or ACT scores or grades in prerequisite courses

Winter 2000

'g 0
ir a

can be used as measures of ability, or a diagnostic test
given early in the course can be used for the purpose of
forming teams.
Form teams that are heterogeneous in ability level. The
members of a team of only weak students are obviously at
a disadvantage (although sometimes they might do sur-
prisingly well), and the members of a uniformly strong
team may choose to divide up the home-
work and to communicate only cursorily
ty group, the with one another. Neither group receives
s gain from the full benefits of cooperative learning.
er students In a mixed-ability group, the weaker stu-
ter students
dents gain from seeing how better stu-
ich problems, dents study and approach problems, and
er students the stronger students usually gain deeper
deeper understanding of the subject through their
f the subject attempts to explain the material, a phe-
ittempts to nomenon familiar to every professor.
material, a Assign team roles that rotate with
familiar to each assignment. Three indispensable
lessor. roles are the manager (organizes the as-
signment into subtasks, allocates respon-
sibilities, and keeps the group on task),
the recorder (writes the final report or problem solution
set, or for large projects, assembles the report), and the
checker proofreadss and corrects the final report before it
is submitted). Other roles that may be performed sepa-
rately or combined with one of the preceding roles in-
clude group process monitor (makes sure that every team
member contributes and that all contributions are acknowl-
edged by the others, verifies that every team member
understands each part of the completed assignment) and
the skeptic (plays the role of devil's advocate, suggests
alternative possibilities, keeps the group from leaping to
premature conclusions). Only the names of the students
who actually participated should appear on the solution,
with their team roles for that assignment identified. In a
lecture course, the roles should rotate with each assign-
ment so that a student cannot repeat as (say) manager
until every other team member has held that position.
Promote positive interdependence. Assign roles. Provide
only one set of materials and require only one team prod-
uct. Provide specialized training to individual team mem-
bers on different aspects of the project that they must then
bring back to the group effort (this technique is known as
"jigsaw" in the cooperative learning literature). Give bo-
nuses on tests to groups when the team average exceeds
80 (or some other specified value). Randomly select one
member of each group to present a problem solution or
report on a specific aspect of the project and give every-
one in the group the grade earned by that individual. If
you use the last strategy (which also promotes individual

SSpecial Feature Section

accountability), tell the students well in advance that you
plan on doing so, but do not provide much advance
notice of which students will present on which parts of
the assignment.
Get teams to assess how well they are functioning. Peri-
odically ask the students to spend five to ten minutes at
the end of their work session assessing their performance,
identifying their strengths, and setting goals for improve-
ment.[19,62'63] A summary of the assessment might be in-
cluded with the group problem solution or in individual
journals on the group process.
Consider doing some testing of pairs or groups. One
mechanism is to administer and score an individual test
and then to allow CL teams to retake the test (perhaps as a
take-home exam) to earn additional points. The advan-
tage of this procedure is that most students will achieve a
deeper understanding of how to solve all the test prob-
lems; the disadvantage is that it requires more grading.
Dekker and Stice[641 recommend giving tests to pairs of
students as an alternative to individual tests and offer
ideas for structuring such tests.
- Do not re-form groups too often. A team should remain
together for at least a month in order to evolve through the
"form, storm, norm, and perform" evolution of team de-
velopment. If students know that they will only have to
remain in a team for two or three weeks, they will have
little incentive to confront and overcome the interpersonal
problems that commonly arise in team development. If,
however, they know they are going to be together for a
longer period of time, they are forced to deal with the
problems by establishing norms, developing strategies for
coping creatively with conflict, and taking advantage of
and valuing individual talents and learning styles.
0 Provide an escape mechanism for teams having severe
difficulties. Roughly halfway through the semester, an-
nounce that you will dissolve all of the teams and form
new ones, except that a team may stay together if each
member sends a note to the instructor expressing a desire
to do so. Typically, all but the most highly dysfunctional
teams elect to remain together, and the problem students
in the groups that dissolve often change their behavior in
their new groups. Consider instituting mechanisms for
teams to fire uncooperative students and for individuals to
quit uncooperative teams when all other avenues (includ-
ing instructor intervention) have been exhausted and prior
warnings have been given.[58'
- Do not assign course grades on a curve. If students rec-
ognize that by helping someone else they could be hurting
themselves (as is the case when grades are curved), they
may be inclined to avoid cooperation, making it less
likely that the benefits of cooperative learning will be
realized. On the other hand, if they are guaranteed a given

grade if they meet a specified standard (for example, a
weighted average grade of 88 or better for an A), they
have every incentive to help their teammates.
> Start small and build. If you have never used cooperative
learning and you are not working with a colleague who is
experienced in this approach, you might consider begin-
ning on a relatively small scale, with several assignments
done by groups and the rest done individually. Once you
gain confidence, increase the level of your involvement to
a point that feels comfortable to you. When problems
arise, remember to consult references on cooperative learn-
ing for ideas about how to deal with them.
Most engineering is done cooperatively, not individually,
and technical skills are often less important than interper-
sonal skills in getting the job done. In survey after survey,
representatives of industry place communication and team-
work at the top of their lists of desirable skills for new
engineering graduates. If teamwork is such a critical part of
what engineers do, surely engineering schools should pro-
vide some guidance in how to do it.
Cooperative learning may be the most thoroughly re-
searched instructional method in all of education, and a vast
and still rapidly growing body of research supports the ef-
fectiveness of the approach.148,57,59'65-68] Studies have shown
that compared to students taught traditionally (that is, prima-
rily with lectures and individual homework), cooperatively
taught students tend to have better and longer information
retention, higher grades, more highly developed critical-
thinking and problem-solving skills, more positive attitudes
toward the subject and greater motivation to learn it, better
interpersonal and communication skills, higher self-es-
teem, lower levels of anxiety about academics, and, if
groups are truly heterogeneous, improved race and gen-
der relations. Another benefit is that when homework is
done cooperatively, there are three to four times fewer
assignments to grade.
Felder, et al., 58.681 report on a longitudinal study compar-
ing the conventional instructor-centered approach with an
alternative approach that combined all of the methods rec-
ommended in this paper. Students experiencing the alter-
native approach outperformed students experiencing the
conventional approach in their academic performance,
development of higher-level thinking skills, retention in
chemical engineering, and attitudes toward their educa-
tional experience.
A variety of factors account for the observed benefits of
cooperative learning. Weaker students working individually
are likely to give up when they get stuck; working coopera-
tively with stronger students to assist them, they keep going
to completion. Many strong students tend to do the minimal
Chemical Engineering Education

I Future of Engineering Education I

work required to complete the assignment, which may not
require deep understanding of concepts; when faced with the
task of explaining and clarifying material to weaker stu-
dents, they often find gaps in their own understanding and
fill them in. Students working alone may tend to delay
completing assignments or skip them altogether; when they
know others are counting on them, they are often driven to
do the work on time.


Although we might wish it were otherwise, for many of
our students tests are the primary motivation to study. The
students may attend every class and complete all the as-
signments, but it is their preparation for the tests that
determines the breadth and depth of their learning. The
burden is on the instructor to make the tests challenging
enough to push each student to learn to the greatest ex-
tent of which he or she is capable.
But, just as tests can motivate students to learn at a deep
level, they can also lead to student demoralization and hos-
tility (both of which correlate with poor performance) if they
are perceived by the students as being unfair. The two most
common types of tests in this category are tests that are too
long and tests that contain surprises-problems with twists
unlike anything the students have seen before and problems
that call for skills that were never taught in class or required
on homework assignments.
Some students-sensing learners on the Myers-Briggs Type
Indicator and the Felder-Silverman Learning Styles Model24-
2632'33]--work more systematically and slowly than the intui-
tive learners who are their counterparts. On tests, the sensors
read and reread problem statements, often taking a relatively
long time to formulate their problem-solving strategies and
checking their calculations carefully. This methodical ap-
proach will make many of them excellent engineers and
experimental scientists, but it frequently leads to their run-
ning out of time on long tests. Nothing infuriates students
more than studying hard and being well prepared for a test,
and then getting a low grade because they lacked sufficient
time to demonstrate their understanding. A student who
gets a "D" on a one-hour test that he or she could have
gotten an "A" on if two hours had been allowed, deserves
the "A"; students who do not understand the material at
an "A" level will not earn an "A" on the test, regardless
of how much time they are given.
Students also resent surprises on tests. The functions of
tests are to motivate and help students to learn what the
instructor wants them to learn and to enable the instructor to
assess the extent to which they have succeeded in doing so.
When students understand the material for which they have
been prepared but do poorly because they cannot figure out a
Winter 2000

"tricky" problem on the spot, they see themselves (right-
fully) as having been cheated by the instructor.
Thinking and problem-solving skills-and speed in prob-
lem solving, for that matter-are only developed through
practice and feedback: testing students on skills they have
not had an opportunity to practice is unfair. There is neither
empirical evidence nor logic to support the argument that
long and tricky tests assess students' potential to be suc-
cessful engineers or help students become better problem
solvers. This does not mean that we should construct
easy tests, which do not motivate students to learn at a
deep level. It is rather to set the bar high, but to teach in a
manner such that all students who have the ability to
meet the challenge can do so.
- Give the students instructional objectivesfor each test in
the form of a study guide. ("In order to do well on this
test, you should be able to... ") Make the list comprehen-
sive and challenging. Include objectives that involve all
of the basic types of calculations the students should be
able to perform, concepts they should be able to explain
without using jargon, formulas they should be able to
derive, derivations they should be able to explain step-by-
step, familiar phenomena that they should be able to
interpret in terms of course concepts, and anything else
you might call on them to do on the test. 51
- When writing the test, consult the instructional objectives
and make sure that 10-15% of the test covers the more
challenging material in the study guide (which will allow
discrimination between the A-level and B-level students).
If the students have the study guide at least a week before
the test-and preferably longer than that-and the ob-
jectives provide the basis of the test construction, there
will be no surprises. The test will be just as challeng-
ing, or more so, than it would otherwise have been,
except that now the challenge is to the students' con-
ceptual understanding rather than to their speed or
puzzle-solving ability.
- Always work a test out yourself from scratch when you
have finished writing it, timing how long it takes to do it.
This burdensome exercise is the only way to discover the
overspecified and underspecified problems, the erroneous
or ambiguous problem statements, the numerical calcula-
tions that take large amounts of time but show very little
about conceptual understanding, and the appropriateness
or inappropriateness of the level of difficulty of the entire
test. The alternative is for these problems to show up
when the test is being given, which leads to disasters of
the type all instructors and students have experienced and
do not wish to experience again.
- Minimize speed as a factor in performance on tests. For

SSpecial Feature Section

quantitative problem-solving tests, you should be able to
work out the test in less than one-third of the time the
students will have to do it, and if the test is particularly
difficult or involves many numerical calculations, a one-
fourth rule might be more appropriate. If it takes you
longer than that, either find a longer time slot in which
to administer the test or consider eliminating ques-
tions, presenting some formulas instead of requiring
derivations, and asking for solution outlines rather
than complete calculations.
- Do not test skills that students have not had a chance to
practice. Don't make all homework problems straightfor-
ward calculations and then put deep analysis questions on
the test. Don't require numerical solutions on all home-
work problems and then ask students for qualitative solu-
tion outlines on the test. Don't give students problems
with extraneous data on the test unless the students have
worked on similar problems in the homework. If picking
important material from long readings is a skill you want
your students to develop, give them training and practice
in it-don't just tell them that they are responsible for
everything in their 500-page text and make them guess
what you plan to ask them to do. If you think ability to
solve quantitative problems quickly is an important skill
(it is generally not that important in engineering practice),
then give the students training and practice in speed-
solving in class and on the homework before you make it
a primary criterion for doing well on the tests.
Even if you curve grades, if the average is in the 50-60
range or below, consider the possibility that it was a poor
test or that you did a poor job of preparing the students
for it. If you decide that either is the case, consider adding
a fixed number of points to each student's grade to
bring the top grade or the average grade to a value of
your choosing. Alternatively, if most students missed
the same problem, announce a quiz for the following
week that will be a variation of that problem and add
the results to their test grades.
Education should not be viewed as a mystery religion.
There is no pedagogical value in making students guess
what they are supposed to know and understand or in testing
them on skills in which they have received no training.
When students know explicitly what is expected of them
(whether it be straightforward or high-level or ill-defined
problem solving, critical or creative or multidisciplinary think-
ing, or anything else) and they are given practice and feed-
back in the specified skills, the odds that they will be able to
meet the expectations go up. Even though the tests may be
harder, the average student performance will be better than it
would have been if the tests were exercises in speed and
guessing ability, student morale and motivation will increase,

and the students who get low grades will be much more
inclined to take responsibility for their poor performance
than to blame the test or the instructor.


The social environment in a class-the nature and quality
of interactions between the students and the instructor and
among the students-can have a profound effect on the
quality of learning that takes place in the class.156'70-75] In his
monumental study, What Matters in College,170] Alexander
Astin found that the quality of interactions between students
and instructors in and out of class was the factor that corre-
lated most highly with almost every positive learning and
attitude outcome he considered. If students believe that an
instructor is concerned about them and has a strong desire
for them to learn the course material, the effects on their
motivation to learn and their attitudes toward the course, the
subject, and the instructor can be profound. The suggestions
that follow are all known to instill such a belief. We suggest
that you consider all of them and try to adopt the ones with
which you feel comfortable.

- Learn the students' names. Taking the trouble to learn
names and use them in and out of class conveys a sense
of respect for the students as individuals. Their motiva-
tion to do well in your course is likely to increase consid-
erably once they realize that you know who they are. Use
place cards or seating charts, take and label photographs
of the class, or ask students to bring in photocopies of
their student identification cards or drivers licenses and
use them to help you learn the names quickly.
N Make yourself available. Announce office hours and keep
them; if you have to miss them, announce it in advance
and schedule replacement hours if possible. Encourage
students to contact you during your office hours or by e-
mail, perhaps insisting that they do so at least once during
the first two weeks of the course. Come to class a few
minutes early to answer any questions the students may
have or just to chat.
> If you use nontraditional methods such as cooperative
learning, explain how what you are doing has been shown
to lead to improved learning and/or improved prepara-
tion for their careers. References given in this paper
(e.g., Felder and Brent1601) provide supportive material for
such explanations.
- Celebrate the students' achievements. When a class does
well on a test or you get a number of creative solutions to
homework problems, offer commendation. When your
students win awards or write articles in the school paper,
Chemical Engineering Education

I Future of Engineering Education I

congratulate them publicly.
- Collect periodic feedback and respond appropriately to
it. Collect midterm evaluations, using either simple, open-
ended questions (What has helped you learn in the course?
What has detracted from your learning? What changes
would improve the course for you?) or a more formal
instrument, such as a Course Perceptions Questionnaire.175'
Periodically collect "minute papers": at the end of a
class, have individual students or pairs take a minute or
two to write (anonymously) the one or two main ideas
presented in the lecture and the muddiest point or con-
cept. Use the responses to monitor how the class went
and to plan the next class. In large classes, use
ombudspersons-class representatives who report to you
periodically about how well the teaching and learning is
going. Regardless of the feedback mechanism chosen,
summarize the most common suggestions, share them
with the class, accept those you can, and explain why you
cannot accept the others.
N Let students participate in learning and performance
assessment. Give choices on assignments (e.g., problem
sets or projects) and tests (e.g., solve any three of the
following four problems). Have students critique one
another's drafts of assignments or lab reports before the
final versions are turned in to you. Let them create poten-
tial examination questions, and use one of them on the
actual exam. Have them assess their own performance
and the performance of their colleagues in team-based
projects.[61] Let them contract for the relative weighting
of the term work and the final examination.[9,76,771
> Maintain a sense of respect for the students, individually
and collectively. Avoid belittling or sarcastic remarks
about their responses to questions, performance on tests,
behavior in class, or anything else. If you are disap-
pointed with any or all of them, express your disappoint-
ment calmly and respectfully. Avoid comments that in-
volve the slightest trace of disparagement or stereotyping
directed at students of a particular race, gender, or sexual
orientation, or with students who are disabled in any way.
If you fail to follow this recommendation, doing every-
thing else recommended in this paper may not be enough
to salvage the class.
The term "caring" or its synonym "concern" show up in
virtually every published study of what students consider to
be effective teaching. In a review of nearly 60 studies of
students' descriptions of effective teachers, Feldman[781 found
eight core characteristics in most lists: concern for students,
knowledge of subject, stimulation of interest, availability,
encouragement of discussion, ability to explain clearly, en-
thusiasm, and preparation. Factor analysis of rating scales
show four generic factors across disciplines: skill (ability to
Winter 2000

communicate), rapport (empathy, concern for students), struc-
ture (class organization, course presentation), and load
(workload).'791 No matter what your teaching style may be-
flashy or congenial or scholarly-if students believe you
care about them, most will be motivated to learn what you
are teaching. If you convey a sense of not caring, then no
matter how brilliantly or entertainingly you lecture, far fewer
will be so motivated.

We have discussed a wide variety of teaching techniques
that have been repeatedly shown to be effective in engineer-
ing education. The techniques are variations on the follow-
ing main themes:
1. Formulate and publish clear instructional objectives.
2. Establish relevance of course material and teach
3. Balance concrete and abstract information in every
4. Promote active learning in the classroom.
5. Use cooperative learning.
6. Give challenging, but fair, tests.
7. Convey a sense of concern about students' learning.
We do not claim that our suggestions constitute a compre-
hensive list of proven effective teaching methods. Such a list
would be encyclopedic and would be comprehensive only
until the appearance of the next issue of any journal on
education. We also do not claim that adopting all of the
suggestions will guarantee that all students in a class will
perform at a high level or even that they will all pass. The
performance of an individual student in a class depends on a
staggering variety of factors, many of which are out of the
instructor's control; moreover, an instructor who sets out
to implement all of the suggestions in this paper is likely
to be overwhelmed in the attempt and to end by imple-
menting none of them.
Our hope is that readers will consider all of the sugges-
tions in the paper in light of their teaching styles and person-
alities and attempt to adopt a few of them in the next course
they teach, and then perhaps a few more in the course after
that. While we cannot predict the extent to which the tech-
niques will succeed in achieving the instructors' objectives,
we can say with great confidence that their use will improve
the quality of learning that occurs in those classes.

Writing formal instructional objectives and using active
and cooperative instructional methods offers a good pros-
pect of equipping your students with the knowledge and
skills you wish them to develop.

(Special Feature Section

We are grateful to Robert Hudgins (University of Water-
loo), Jorge Ibafiez (Universidad Iberoamericana-Mexico
City), John O'Connell (University of Virginia), Tom Regan
(University of Maryland), Antonio Rocha (Instituto
Tecnologico-Celaya), Heather Sheardown (McMaster
University), and Phil Wood (McMaster University) for
helpful reviews of this paper.

1. Rugarcia, A., R.M. Felder, J.E. Stice, and D.R. Woods,
"The Future of Engineering Education: I. A Vision for a
New Century," Chem. Eng. Ed., 34(1), 16 (2000)
2. McKeachie, W.J., Teaching Tips: Strategies, Research, and
Theory for College and University Teachers, 10th ed.,
Houghton Mifflin, Boston, MA (1999)
3. Wankat, P., and F.S. Oreovicz, Teaching Engineering,
McGraw-Hill, New York, NY (1993). Available on-line at

4. Stice, J.E., ed., Developing Critical Thinking and Prob-
lem-Solving Abilities, New Directions in Learning and
Teaching, No. 30, Jossey-Bass, San Francisco, CA (1987)
5. Stice, J.E., "A First Step Toward Improved Teaching,"
Eng. Ed., 66(5), 394 (1976)
6. Felder, R.M., and R. Brent, "Objectively Speaking," Chem.
Eng. Ed., 31(3), 178 (1997). Available on-line at
7. Gronlund, N.E., How to Write and Use Instructional Ob-
jectives, 5th ed., Macmillan, New York, NY (1994)
8. Mager, R.F., Preparing Educational Objectives, Fearon
Publishers, San Francisco, CA (1962)
9. Popham, W.J., and E.L. Baker, Establishing Instructional
Goals, Prentice Hall, Englewood Cliffs, NJ (1970)
10. Bloom, B.S., and D.R. Krathwohl, Taxonomy of Educa-
tional Objectives. Handbook 1: Cognitive Domain, Addison-
Wesley, New York, NY (1984)
11. Woods, D.R., "How Might I Teach Problem Solving?" in
Developing Critical Thinking and Problem-Solving Abili-
ties, J.E. Stice, ed., New Directions for Teaching and Learn-
ing, Jossey-Bass, San Francisco, CA (1987)
12. Ramsden, P., and N.J. Entwistle, "Effects of Academic
Departments on Students' Approaches to Studying," Brit-
ish J. of Ed. Psychology, 51, 368 (1981)
13. Haile, J.M., "Toward Technical Understanding: Part 1.
Brain Structure and Function," Chem. Eng. Ed., 31(3),
152 (1997); "Part 2. Elementary Levels," Chem. Eng. Ed.,
31(4), 214 (1997); "Part 3, Advanced Levels," Chem. Eng.
Ed., 32(1), 30 (1998)
14. Bruning, R.H., G.J. Schraw, and R.R. Ronning, Cognitive
Psychology and Instruction, 3rd ed., Merrill (1999)
15. Glaser, R., "Education and Thinking: The Role of Knowl-
edge," American Psychologist, 55, 2 (1984)
16. Woods, D.R., "Three Trends in Teaching and Learning,"
Chem. Eng. Ed., 32(4), 296 (1998)
17. Felder, R.M., Meet Your Students. III. Michelle, Rob, and
Art," Chem. Eng. Ed., 24(3), 130 (1990). Available on-line
at lic/Columns/Objectives.html>
18. Woods, D.R., Problem-Based Learning: Helping Your Stu-
dents Gain the Most from PBL, Woods Publishing,
Waterdown (1997). Distributed by McMaster University

Bookstore, Hamilton, Ontario, Canada, and available on-
line at
19. Woods, D.R., Problem-Based Learning: How to Gain the
Most from PBL, Woods Publishing, Waterdown (1994). Dis-
tributed by McMaster University Bookstore, Hamilton,
Ontario, Canada
20. Barrows, H.S., and R. Tamblyn, Problem-Based Learning,
Springer, New York, NY (1980)
21. Engel, C.E., "Not Just a Method But a Way of Learning,"
Chapter 2 in The Challenge of Problem-Based Learning,
D.J. Boud and G. Feletti, eds., Kogan Page, London, UK
22. Schmidt, H.G., "Problem-Based Learning: Rationale and
Description," Medical Ed., 17, 11 (1983) and "Foundations
of Problem-Based Learning: Explanatory Notes," Medical
Ed., 27, 422 (1993)
23. Coles, C., "Is Problem-Based Learning the Only Way?" Chap-
ter 30 in The Challenge of Problem-Based Learning, D.J.
Boud and G. Feletti, eds., Kogan Page, London, UK (1991)
24. Felder, R.M., and L.K. Silverman, "Learning and Teaching
Styles in Engineering Education," Eng. Ed., 78(7), 674 (1988)
25. Felder, R.M., "Reaching the Second Tier: Learning and
Teaching Styles in College Science Education," J. College
Sci. Teach., 23(5), 286 (1993). Available on-line at
26. Felder, R.M., "Matters of Style,"ASEE Prism, 6(4), 18 (1996).
Available on-line at users/f/felder/public/Papers/LS-Prism.htm>
27. Kolb, D.A., Experiential Learning: Experience as the Source
of Learning and Development, Prentice Hall, Englewood
Cliffs, NJ (1984)
28. Stice, J.E., "Using Kolb's Learning Cycle to Improve Stu-
dent Learning," Eng. Ed., 77, 291 (1987)
29. McCarthy, B., The 4MAT System: Teaching to Learning
Styles with Right-Left Mode Techniques, EXCEL Inc.,
Barrington, IL (1986)
30. Harb, J.N., R.E. Terry, P.K. Hurt, and K.J. Williamson,
Teaching Through the Cycle: Application of Learning Style
Theory to Engineering Education at Brigham Young Uni-
versity, BYU Press, Provo, UT (1995)
31. Harb, J.N., S.O. Durrant, and R.E. Terry, "Use of the Kolb
Learning Cycle and the 4MAT System in Engineering Edu-
cation," J. Eng. Ed., 82, 70 (1993)
32. Lawrence, G., People Types and Tiger Stripes, 3rd ed., Cen-
ter for Applications of Psychological Type, Gainesville, FL
33. Felder, R.M., "Meet Your Students. I. Stan and Nathan,"
Chem. Eng. Ed., 23(2), 68 (1989). Available on-line at /
34. Wood, P.E., "Bring the Real World into the Classroom,"
Canadian Chem. News, March (1999)
35. Kresta, S., "Hands-On Demonstrations: An Alternative to
Full-Scale Lab Experiments," J. Eng. Ed., 87(1), 7 (1998)
36. Felder, R.M., "On Creating Creative Engineers," Eng. Ed.,
77(4), 222 (1987)
37. Felder, R.M., "The Generic Quiz: A Device to Stimulate
Creativity and Higher-Level Thinking Skills," Chem. Eng.
Ed., 19(4), 176 (1985)
38. Felder, R.M., and B.A. Soloman, Index of Learning Styles.
Available on-line at users/f/felder/public/ILSpage.html>
39. Keirsey, D., Keirsey Temperament Sorter. Available on-line
40. Piaget, J., and B. Inhelder, The Psychology of the Child,

Chemical Engineering Education

Future of Engineering Education j

Basic Books, New York, NY (1975)
41. Piaget, J., "Intellectual Evolution from Adolescence to Adult-
hood," Human Develop., 15, 1 (1972)
42. Williams, K.A., and A.M.L. Cavallo, "Reasoning Ability,
Meaningful Learning and Students' Understanding of Phys-
ics Concepts," J. College Sci. Teach., 24(5), 311 (1995)
43. Stonewater, J.K., and B.B. Stonewater, "Teaching Problem
Solving: Implications From Cognitive Development Re-
search," AAHE Bulletin, 7, February (1984)
44. Godleski, E.S., "Learning Style Compatibility of Engineer-
ing Students and Faculty," FIE '84 Proceedings, IEEE/ASEE,
45. Rosati, P., R.K. Dean, and S.M. Rodman, "A Study of the
Relationship Between Students' Learning Styles and In-
structors' Lecturing Styles," IEEE Trans. in Ed., 31(3), 208
46. Bonwell, C.C., and J.A. Eison, Active Learning: Creating
Excitement in the Classroom, ASHE-ERIC Higher Educa-
tion Report No. 1, George Washington University, Washing-
ton, DC (1991)
47. Sutherland, T.E., and C.C. Bonwell, Using Active Learning
in College Classes: A Range of Options for Faculty, Jossey-
Bass, San Francisco, CA (1996)
48. Johnson, D.W., R.T. Johnson, and K.A. Smith, Active Learn-
ing: Cooperation in the College Classroom, 2nd ed., Interac-
tion Book Co., Edina, MN (1999)
49. Felder, R.M., "How About a Quick One?" Chem. Eng. Ed.,
26(1), 18 (1992). Available on-line at unity/lockers/users/f/felder/public/Columns/Quickone.html>
50. Felder, R.M., "It Goes Without Saying," Chem. Eng. Ed.,
25(3), 132 (1991). Available on-line at
51. Felder, R.M., "FAQs II," Chem. Eng. Ed., 33(4), 276 (1999)
52. Felder, R.M., "Beating the Numbers Game: Effective Teach-
ing in Large Classes," 1997 ASEE Ann. Conf Proc., ASEE,
June (1997). Available on-line at unity/lockers/users/f/felder/public/Papers/Largeclasses.htm>
53. Woods, D.R., "Handling the Large Class: Ways to Deal With
a Potentially Big Problem," J. College Sci. Teach., 20(5), 297
54. Whimbey, A.E., and J.J. Lochhead, Problem Solving and
Comprehension: A Short Course in Analytical Reasoning,
Franklin Institute Press, Philadelphia, PA (1982)
55. Michaelson, L.K., R.H. Black, and L.D. Fink, "What Every
Faculty Developer Needs to Know about Learning Groups,"
in L. Richlin (ed.), To Improve the Academy: Resources for
Faculty, Instructional, and Organizational Development,
New Forums Press, Stillwater, OK (1996)
56. Chickering, A.W., and Z.F. Gamson, "Seven Principles for
Good Practice in Undergraduate Education," AAHE Bulle-
tin, p.3, March (1987)
57. Felder, R.M., and R. Brent, Cooperative Learning in Techni-
cal Courses: Procedures, Pitfalls and Payoffs, ERIC Docu-
ment Reproduction Service, ED 377038 (1994). Available
on-line at public/Papers/Coopreport.html>
58. Felder, R.M., "A Longitudinal Study of Engineering Stu-
dent Performance and Retention. IV. Instructional Methods
and Student Responses to Them," J. Eng. Ed., 84(4), 361
(1995). Available on-line at lockers/users/f/felder/public/Papers/long4.html>
59. Millis, B.J., and P.G. Cottell, Jr., Cooperative Learning for
Higher Education Faculty,American Council on Education
and the Oryx Press, Phoenix, AZ (1998)
60. Felder, R.M., and R. Brent, "Navigating the Bumpy Road to

Winter 2000

Student-Centered Instruction," College Teaching, 44(2), 43
(1996). Available on-line at lockers/users/f/felder/public/Papers/Resist.html>
61. Kaufman, D.B., R.M. Felder, and H. Fuller, "Peer Ratings
in Cooperative Learning Teams," 1999 ASEE An. Conf.
Proc., ASEE (1999). Available on-line at
62. Heller, P., R. Keith, and S. Anderson, "Teaching Problem
Solving Through Cooperative Grouping, Part 1: Group Ver-
sus Individual Problem Solving," Amer. J. of Physics, 60(7),
63. Heller, P., and M. Hollabaugh, "Teaching Problem Solving
Through Cooperative Grouping: Designing Problems and
Structuring Groups," Amer. J. of Physics, 60(7), 637 (1992)
64. Dekker, D., and J.E. Stice, "Pairs Testing," FIE '98 Proc., 14
65. Springer, L., M.E. Stanne, and S. Donovan, "Effects of Small-
Group Learning on Undergraduates in Science, Mathemat-
ics, Engineering, and Technology: A Meta-Analysis," Rev. of
Ed. Res., 69(1), 21 (1999). Available on-line at>
66. Goodsell, A.M., M. Maher, and V. Tinto (eds.), Collaborative
Learning: A Sourcebook for Higher Education. National Cen-
ter of Postsecondary Teaching, Learning, and Assessment,
Pennsylvania State University, University Park, PA (1992)
67. Kadel, S., and J.A. Keehner (eds.), Collaborative Learning:
A Sourcebook for Higher Education, Vol. 2, National Center
on Postsecondary Teaching, Learning, and Assessment,
Pennsylvania State University, University Park, PA (1994)
68. Felder, R.M., G.N. Felder, and E.J. Dietz, "A Longitudinal
Study of Engineering Student Performance and Retention.
V. Comparisons with Traditionally Taught Students," J.
Eng. Ed., 87(4), 469 (1998). Available on-line at
69. Felder, R.M., "Tips on Quantitative Tests," Emphasis on
Teaching and Learning, N.C. State University, 7, Novem-
ber (1997). Available on-line at lockers/users/f/felder/public/Papers/testingtips.html>
70. Astin, A.W., What Matters in College: Four Critical Years
Revisited, Jossey-Bass, San Francisco, CA (1993)
71. Brink, R.W., Educational Innovations in Canada and the
United States, Dept of Electrical Engineering, University of
Twente, Enschede, The Netherlands (1989)
72. Ramsden, P., Learning to Teach, Routledge, London, UK
73. Ramsden, P., "Student Learning and Perceptions of the
Academic Environment," Higher Ed., 8, 411 (1979)
74. Entwistle, N.J., and P. Ramsden, Understanding Student
Learning, Croom and Helm, London, UK (1983)
75. Ramsden, P., The Lancaster Approaches to Studying and
Course Perceptions Questionnaire: Lecturer's Handbook,
Oxford Polytechnic, Headington, Oxford, UK (1983)
76. Woods, D.R., R.R. Marshall, and A.N. Hrymak, "Self-As-
sessment in the Context of the McMaster Problem-Solving
Program," Evaluation and Assessment in Higher Ed., 12(2),
107 (1988)
77. Brown, G., and M. Pendleberry, Assessing Active Learning,
Parts 1 and 2, CVCP Universities; Staff Development and
Training Unit, University House, Sheffield, UK (1992)
78. Feldman, K.A., "The Superior College Teacher from the
Students' View," Research in Higher Ed., 5, 43 (1976)
79. Cross, K.P., "On College Teaching," J. Eng. Ed., 82(1), 9
(1993) 0


Award Lecture ...

Chemical Engineering Division, ASEE
1999 Union Carbide Award Lecture



Part 1. Educational Issues*
The Ohio State University Columbus, OH 43210

It is indeed a great honorfor me to be the recipient of the
1999 ASEE Chemical Engineering Division's Union Car-
bide Lectureship Award. I am particularly honored to be
included among the outstanding educators who have re-
ceived this award in the past.

Particle technology is not well covered in the chemical
engineering curriculum, but it is an important area for
chemical engineers from both industrial and academic
perspectives. In this lecture I will specifically discuss par-
ticle dynamics and will use fluidization and fluid-particle
systems as examples to illustrate the importance to chemical
engineers of knowing the particle dynamics of these sys-
I would like to define the scope of my lecture as follows:
1. Observations on educational issues in particle dynamics.
2. Comparison of the mechanics of flow of solids and fluids.
3. Sample subjects of interdisciplinary nature.
4. Sample subjects of pertinence to chemical engineers.
5. Computational fluid dynamics of particulate systems.
In Part 1, appearing here, I will discuss points 1 through 3. I
will discuss points 4 and 5 in Part 2, which will be published
in the next issue of CEE.

One of the most important fluid-particle applications in
the chemical and petrochemical industries is FCC (fluid
catalytic cracking) systems.m" In North America alone, there
are 120 to 135 FCC units in operation, with each processing
40,000 to 50,000 barrels of gas oil per day to generate olefin
gas, gasoline, diesel, and heavy cycle gas oil. Approximately
0.08 kg of catalyst are consumed for each barrel of gas oil
processed. Figure la shows a photograph of a commercial
FCC system comprised of a riser reactor and a catalyst
regenerator. In the schematic diagram shown in Figure lb,
gas oil is fed into the bottom of the riser in contact with high-
temperature catalyst particles recycled from the regenerator
to the riser. The gas oil is evaporated, carrying catalyst
particles along with it throughout the riser, where cracking
reactions take place. The product of the reactions is then sent
to the fractionator; the spent catalyst particles are stripped by
steam and recycled back to the regenerator.
The solids in this system are processed with gas in various
forms or modes. For example, within the riser, gas and solids
are in the dilute pneumatic transport mode, whereas within
the regenerator, gas and solids are in the dense, turbulent

L.S. Fan is Distinguished University Professor and Chairman of the Department of Chemical Engineering at the Ohio State
University. His expertise is in fluidization and multiphase flow, powder technology, and particulates reaction engineering.
Professor Fan is the U.S. editor of Powder Technology and a consulting editor of the AIChE Journal and the International
Journal of Multiphase Flow. He has also authored or co-authored three books, including the textbook Principles of Gas-Solid
Flows (with Chao Zhu; Cambridge University Press, 1998), in addition to 240 journal articles and book chapters, and has
edited nine symposium volumes.
Professor Fan is the principal inventor (with R. Agnihotri) of a patented process, "OSCAR," for flue gas cleaning in coal
combustion and is the Project Director for the OSCAR commercial demonstration, funded at $8.5 million as Ohio Clean Coal
Technology, currently taking place at Ohio McCracken power plant on the Ohio State University campus.
He has served as thesis advisor for two BS, twenty-nine MS, and forty-two PhD students at Ohio State, and is a Fellow of the
American Association for the Advancement of Science.
Part 2 of this lecture will appear in the Spring 2000 issue of CEE.
Copyright ChE Division of ASEE 2000
40 Chemical Engineering Education

Figure 1.
FCC systems for
gas-oil cracking:
(a) Photograph of
Equilon (formerly
Texaco) commer-
cial FCC unit plant
at Los Angeles
(kindly provided
by F. Bavarian of
(b) Schematic of
the system.

fluidization mode. In the catalyst recycle loop, there are dense-
phase standpipe solids flow and inclined pipe solids flow.
Solids processing is also involved in chemical synthesis,
such as the production of polyethylene/polyolefin. Using
the UNIPOL process[21 as an example (see Figure 2), cata-
lyst, cocatalyst, monomer, and comonomer are introduced
into a turbulent fluidized bed, where polymerization reac-
tions take place. In the reactor, polymer particles grow in
size through chain reactions and eventually reach the final
polyethylene size average of 600p.m.
Solids processing is also involved in a number of other
industries. For example, it is used in physical operations
such as powder coating, drying, and mixing. It is also used
in energy and environmental systems, e.g., coal combustion
and gasification, and incineration of solid wastes, and in
metallurgical and mineral processing such as titanium diox-
ide production. In biological systems it is used, for ex-
ample, in ethanol fermentation. Overall, solids processing
systems are responsible for well over $100 billion of the
chemical and petrochemical market economy annually.
Many students take part in industrial internships and co-
op programs. These students and many engineering gradu-
ates employed in industry frequently find themselves in-
volved in solids processing, for which they have not been
well prepared through regular course work. If we examine
the typical undergraduate educational material pertaining to
solids processing, we will note that this material is often
limited to single-particle behavior (drag, terminal velocity,
heat and mass transfer), fixed beds, catalytic and non-cata-
lytic fluid-particle reaction kinetics, and particulate reac-
tion engineering. In the latter, solids particles are often
treated the same way as gases or liquids. As a result, the
unique characteristics of particle mechanics are not intro-
Winter 2000

duced into the analysis of particulate reaction systems. In-
deed, very little is discussed on core topics relating to solids
processing and particle technology such as
> Particle characterization
Particle formation
Size enlargement and agglomeration
1 Comminution and attrition
> Tribology, friction, and interparticle forces
> Fluidization and multiphase flow
> Solids flow, handling, and processing
> Powder mechanics and slurry rheology
> Colloids and aerosols

The importance of particle technology education and re-
search was brought to the attention of the industrial and
academic communities through the perseverance of such
organizations as the American Filtration Society and the
Particle Technology Forum of the AIChE in the early 1990s.
Subsequent articles in Chemical Engineering Progress
(CEP)[3] and Chemical Engineering Education (CEE)[4] have
contributed to increasing awareness of this topic. As a result

flue gas


Many [co-op students and] engineering graduates employed in industry find themselves involved in
solids processing, for which they have not been well prepared through regular course work. If we
examine the typical undergraduate educational material pertaining to solids processing, we will
note that this material is often limited to single-particle behavior, fixed beds, catalytic and
non-catalytic fluid-particle reaction kinetics, and particulate reaction engineering.

of the Fluid-Particle Processes Workshop at the ASEE Sum-
mer School for Chemical Engineering Faculty (Snowbird,
Utah, 1997), the following articles and reprints related to
particle technology education were published in the spring
issue of CEE in 1998:
"Teaching Fluid-Particle Processes: A Workshop Report," R.H.
Davis, L.-S. Fan
"Industrial Perspective on Teaching Particle Technology," R.D.
Nelson, Jr., R. Davies
"Particle Technology Concentration at NJIT: An NSF-CRCD
Program," R.N. Dave, I.S. Fischer, J. Luke, R. Pfeffer, A.D.
"CFD Case Studies in Fluid-Particle Flow," J.L. Sinclair
"Experiments, Demonstrations, Software Packages, and Videos for
Pneumatic Transport and Solid Processing Studies," G. Klinzing
"Undergraduate Teaching in Solids Processing and Particle
Technology: An Academic/Industrial Approach," G.G. Chase, K.
"Particle Science and Technology Educational Initiatives at the
University of Florida," A.E. Donnelly, R. Rajagopalan

The authors of the articles above are heavily engaged in
fluid-particle education and research, and therefore the ar-
ticles are most pertinent to the point of the present discus-
sion. Recently, federal-, state-, and/or industry-funded re-
search and education centers were formed (e.g., NSF/ERC
in Particle Science and Technology at the University of
Florida, NJIT/Rutgers State Program on Particle Technol-
ogy, and the Ohio Board of Regents Universities Consor-
tium on Fine Particle Processing). New web sites (e.g., http:/
/, new instructional modules (e.g., "In-
troduction to the Principles of Size Reduction of Particles by
Mechanical Means," by Klimpel151), introductory textbooks
(e.g., Introduction to Particle Technology, by Rhodes'61),
advanced textbooks (e.g., Principles of Gas-Solid Flows, by
Fan and Zhu'71), and CD-ROMs (e.g., "Laboratory Demon-
strations in Particle Technology," by Rhodes and Zakhari'18)
have been published as a result of the growing interest and
acknowledged importance of particle mechanics.
I would now like to present some problems that I have
noted in my experience in teaching fluid-particle systems
that are confusing to students. I will give two examples.

- Log-Normal Distribution

There are three distribution functions that are commonly
used to describe particle size distributions, i.e., normal dis-
tribution, log-normal distribution, and Rosin-Rammler dis-
tribution. The log-normal distribution is particularly confus-
ing to students.

In examining the log-normal distribution, it is noted that it
can be expressed in two different equations, depending on
whether the random variable is d (Eq. 1) or Ind (Eq. 2), as
given by

N (d)= I nd -lnd,,
fN(d)= 1- ex 2 od5
2tadid 2 odl

<\ 11 I lnd-lnd5012!
fN(lnd)= exp -nd- d (2)
*V27ITdl 2 dl

Here, d is the diameter of the particle, d5o is the median
diameter, and 7dl is defined as ln(d4/d,,,), where d84 is the
diameter for which the cumulative distribution curve has the
value of 0.84. Further, Eq. (2) is the usual expression for the
normal distribution with the random variable taken as Ind.
The values of Ind5s and adl in Eq. (1) or (2) represent the
mean and standard deviation of the lnd distribution, but
they are not representative of those of the d distribution.
Students are often confused by the two forms of Eqs. (1) and
(2) and by the fact that the arithmetic mean of the Ind
distribution is not equal to the arithmetic mean of the d
distribution, nor does it equal the natural log of the arith-
metic mean of the d distribution; likewise the standard de-
viation of the Ind distribution is not equal to that of the d

Liquid <-

bed liqui




r dh

differential press

.- h

total pressure
|Liquid measurement

Liquid-solid fluidized bed

Figure 3. Pressure-drop measurement for a
liquid-solid fluidized bed.
Chemical Engineering Education

> Dynamic Pressure Drop
When a manometer is used to measure pressure drop of a
suspended particle flow, e.g., a liquid-solid fluidized bed,
the pressure drop is measured by (pm -P )gdh where Pm is
the density of the manometer fluid and dh is the level differ-
ence between the manometer fluid and the liquid in the
fluidized bed, as shown on the left-hand side of the mano-
meter arrangement in Figure 3. The pressure drop measured
in this manner is known as the differential pressure drop.
Students most often question whether the differential pres-
sure drop represents the dynamic pressure drop, APd, or the
total pressure drop APt, as defined by
APt = P,t P2,t (3)
APd = AP -pgH (4)

A force balance on the manometer fluid yields

PI,t +Pig(h; +dh)=P2, +pgH+plgh; +pgdh (5)
Rearranging Eq. (5) gives

APt-pHg = (pm p,)gdh (6)
Thus, we have

APd =(Pm P )gdh (7)
That is, it measures the dynamic pressure drop.
The total pressure drop can be measured with the mano-
meter open to the atmosphere, as shown in the right-hand
side of the manometer arrangement in Figure 3. Frequently,
pressure transducers are used for pressure-drop measure-
ments. In this situation, it is essential that proper calibration
of the transducer be made so that it reflects the correct type

of pressure drop being measured.
Correct identification of either type of pressure drop is
important, as they are often used to calculate the volume
fraction of the particle, es, or liquid, e1, in the bed through
the relationships

AP = (esPs +E,p)gH (8)

APd = s(p Pl)gH (9)


Phenomenologically, particles and fluids are similar in
that they both can flow, but there are distinct differences
between their mechanics of flow. For example, particles and
fluids respond to stress differently. Solids can transfer shear-
ing stresses under static conditions, while liquids cannot
transfer shearing stresses without flowing. For a slow mo-
tion of solids, the shear stress varies with the normal stress
rather than with shear rate. For a liquid flow, the opposite is
true. Solid particles can be consolidated by their cohesive
strength induced by internal friction, while there is no inter-
nal friction in liquids to sustain their consolidation. There-
fore, solid particles can form a heap with a non-zero angle of
repose, whereas liquids lie flat under static conditions.
When particles or liquids are placed vertically in a pipe
with both ends of the pipe open, the wall shear stress can
provide the predominant support for the particle weight.
Therefore, to maintain solid particles stationary in the pipe,
only a small force needs to be applied to the bottom of the
pipe.91' This is not the case for a liquid. The coherence of the

Figure 4.
Particle jet
(alumina, 75 pm)
formed from a
(3.1 mm in
in an air-fluidized
(from Martin and
reproduced with

Winter 2000

particles can also be demonstrated by a particle
jet '01 when formed from a nozzle in a gas-solid
fluidized bed of 75 tm alumina particles, as
shown in Figure 4. The coherence of the jet is
clearly seen in the figure. The liquid jet from a
liquid column is less coherent.
Despite the long history of practice in particle
mixing, the phenomena involved remain fasci-
nating and puzzling. Taking the particle band
formation in a rotating tube with a binary mix-
ture of particles as an example, a single band or
multiple bands would form transiently in a bi-
nary mixture of certain particle sizes, weight ra-
tios, and particle volume fractions in the tube.
Figure 5 shows single-band and double-band for-
mation of small glass beads (dp = 0.15 mm, col-
ored white, 30 vol% in the particle mixture) in
the presence of large glass beads (dp = 1.0 mm,
colored red, 70 vol% in the particle mixture)
with a rotation speed of 60 RPM. The volume
fraction of the particle mixture in the tube is
70%. The tube is of 3.1 cm ID and 36 cm in
length. The single band forms first and disap-
pears, and then a double band forms. Although
some theories, e.g., percolation theory, have been
used to explain the particle migration phenom-
enon, so far there is no overwhelmingly convinc- Figure 5. Illustrations of particle-band formation.
ing mechanistic explanation of the complex phe-
nomenon exhibited by such a simple experiment.

Probably the most effective way to impart knowledge Hopper Hopper
of such an interdisciplinary topic as particle mechanics to
students is through several required chemical engineer-
ing courses including fluid mechanics, heat and mass
transfer, and reaction engineering, in which clear distinc-
tions can be made between particle and fluid behavior.
Once students are vested with the background knowl- Standpipe Standpipe
edge relating to particle mechanics, they can choose to
take technical-elective courses in chemical engineering
or other engineering disciplines that cover some specific
aspects of particle-related subjects. In the following, I
have chosen hopper and standpipe systems as examples
to illustrate the relevance of powder mechanics to a
fundamental understanding of powder flow in these sys-
tems. These examples are introduced so that students
will be familiar with subjects of an interdisciplinary na-
ture. (a) (b)
Hopper and standpipe flows can be demonstrated with
a simple experiment. Figure 6 shows a photograph of a Figure 6. A device showing mass-hopper flow,
device partially filled with table salt. The device has funnel-hopper flow, and standpipe flow.
hoppers connected by a standpipe. Figures 6a and 6b
44 Chemical Engineering Education

show two different hopper flow patterns. For hoppers with a
small apex angle (Figure 6a), solid particles flow downward
uniformly across the whole cross section, forming a flat
surface that is known as a mass-flow hopper. For hoppers
with a large apex angle (Figure 6b), solid particles at the
central location flow faster than those at the wall, forming a
funnel-shaped free surface, known as a funnel-flow hopper.
The standpipe flow in the figure shows moving bed transport
followed by suspension transport of solids with a larger
moving bed region for the mass-flow hopper than for the
funnel-flow hopper.
The onset of powder motion in a hopper is due to stress
failure in powders. Hence, the study of hopper flow is closely
related to understanding the static stress distribution in a
hopper.[l2 The local distributions of static stresses of pow-
ders can only be obtained by solving equations of equilib-
rium. From stress analyses and suitable failure criteria, the
rupture locations in granular materials can be predicted. As a
result, the flowability of granular materials in a hopper de-

Figure 7. Stress components in a plane-strain
problem in the x-z plane.



Figure 8. Equilibrium of stress components on
a differential element.
Winter 2000

pends on the internal stress distributions determined by the
geometry of the hopper and the material properties of the
Stress analysis of solid materials is a typical subject for
engineering mechanists or soil mechanists in civil engineer-
ing, but chemical engineers also need to be familiar with the
subject in order to be able to quantify moving-bed transport
flow of solid particles. Here, students need to learn the Mohr
Circle for plane stresses and the Mohr-Coulomb failure cri-
terion, which can be illustrated as follows.

* Mohr Circle for Plane Stresses
We consider stresses on a point represented by a cubic
differential element. For simplicity, we examine only the
stresses acting on a plane, say the x-z plane, of a cube, as
shown in Figure 7. The stress tensor, expressed in Cartesian
coordinates, takes the form

ax 0 Txz
T= 0 (y
tzx 0 z1

where the o' s are normal stresses compressivee stresses are
considered as positive), and the t' s are shear stresses. From
Hooke's law and assuming no displacement in the y-direc-
tion, we have

Ty = v(o, + z) (10)

where v is Poisson's ratio. In addition, from the conserva-
tion of angular momentum, T is found to be a symmetric
tensor so that Tx = zx. Thus, the plane stress tensor in Eq.
(9) depends solely on ox, oz, and xz. As shown in Figure
8, the force balance on the differential element results in the
stress relationships on the BC plane, as given by

a = G, cos2 P+ z sin2 P+ 2Tx, sin P cos P
TTxz (cos2- sin2 P)+ (a -,x)sin Pcos (11)

where p is the angle between the normal of the BC plane
and the x-axis. From Eq. (11), two perpendicular planes can
be found on which the shear stress vanishes (i.e., T = 0). The
directions of these planes are known as the principal direc-
tions and the corresponding normal stresses as the principal
stresses. The angle for the principal directions, Ppr, is deter-
mined from Eq. (11) as

pr = tan- 2xz (12)
2 (ox -Cz)

which yields the principal stresses as

oxG z+ Iox 2
,3 a + + ( 2 z (13)
2 2)

If the principal directions are taken as the x- and z-axes,
i.e., ox =o1; oz =3, Eq. (11) reduces to

Ox =01; z = 53

G=O| Cos2 P+o3 sin2

S=(a3-, )sin cos3 (14)

which is the equation of a circle in (o-T coordinates as
shown in Figure 9. This circle is known as the Mohr circle.
The direction and the magnitude of stresses on any plane can
be determined graphically from the Mohr circle. As shown
in Figure 9, the normal stresses on the principal planes are of
maximum or minimum values.
9 Mohr-Coulomb Failure Criterion and Coulomb Powder
The most common failure criterion for granular materials
is the Mohr-Coulomb failure criterion. Based on this crite-
rion, the material fails along a plane only when a critical
combination of normal and shear stresses exists on the fail-
ure plane. This critical combination, known as the Mohr-
Coulomb failure criterion, is given by

t=c+ tan i (15)
where c is the cohesion defined as the resistance of the
material to shear under zero normal load and is a result of the
intermolecular cohesive forces, frictional forces, and other
forces acting on the material, and ir is the angle of internal
friction of the material, which corresponds to the maximum
static friction condition as the bulk solids start to slide on
themselves at the state of incipient failure.
The Mohr-Coulomb failure criterion can be recognized as
an upper bound for the stress combination on any plane in
the material. Consider points A, B, and C in Figure 9. Point
A represents a state of stresses on a plane along which
failure will not occur. On the other hand, failure will occur
along a plane if the state of stresses on that plane plots a
point on the failure envelope, e.g., point B. The state of
stresses at point C cannot exist since it lies above the failure
envelope. Since the Mohr-Coulomb failure envelope charac-
terizes the state of stresses under which the material starts to
slide, it is usually referred to as the yield locus, YL.
A rigid-plastic powder that has a linear yield locus is
called a Coulomb powder. Most powders have linear yield
loci, although in some cases nonlinearity appears at low
compressive stresses. The Mohr-Coulomb failure criterion
underlies a basic principle that quantifies important hopper
design variables such as
* Critical major principal stress in a stable arc, 0a (see
Figure 10)
* Hopper (half) apex angle, (p, (see Figure 10)
* Minimum outlet dimension of the hopper
Industrial accidents do occur frequently in hopper flow
due to the failure of hopper operators to recognize the stress

m A.

r'~ ,, .X _
g,(0;, < )

Figure 9. Mohr circle and Mohr-Coulomb
failure envelope.

Figure 10. Arching (or doing) at the
hopper outlet.

behavior acting on a stable arc (Figure 10), which blocks the
solids flow.

* Standpipe Flows
Standpipe flow refers to the downward flow of solids with
the aid of gravitational force against a gas pressure gradient.
Gas flow is in the upward direction with respect to the
downward-flowing solids; relative to the wall, the actual
direction of the flow of gas can be either upward or down-
ward.I31 Solids fed into a standpipe are often from hoppers,
cyclones, or fluidized beds. A standpipe can be either verti-
cal or inclined, and its outlet can be simply an orifice or can
be connected to a valve or fluidized bed. There can be
aeration along the side of the standpipe. Frequently, the
following assumptions are used in the analysis of a vertical
standpipe flow:
Both solids and gas are regarded as a pseudocontinuum
throughout the standpipe system.
Motions of solids and gas are steady and one-dimensional
(in the axial direction).
Solids can flow in either moving-bed mode or dilute-sus-
pension mode. Solids stresses among particles and between
particle and pipe wall are considered in the moving bed
Chemical Engineering Education

span ofihearch

* -a.

flows but neglected in the dilute suspension flows.
The gas can be regarded as an ideal gas, and the transport
process is isothermal.
The cylindrical coordinate system selected for the standpipe
is shown in Figure 11. For a one-dimensional steady motion
of solids, the momentum equation of the particle phase can
be written as

du dpz 2t
Pp (1- )up P =pg(1-)- + FD (16)
dz dz Rs

where a is the volume fraction of the gas phase; Opz is the
normal stress of solids, Tpw is the shear stress of solids at the
pipe wall, FD is the drag force per unit volume, and R, is the
radius of the standpipe. As solids flow can be in either a
moving packed bed mode or a suspension transport mode,
Eq. (16) can be simplified as
For a moving packed bed mode, a is constant.
For a suspension transport mode, Gpz is negligibly
The general momentum balance for the gas phase can be
expressed as

dp= -F (17)
where p is the pressure.
For a simple standpipe system, different flow patterns of
steady flow may exist, depending on the ranges of opera-
tional parameters of the system. This phenomenon is known
as steady-state multiplicity.1 41 The steady-state multiplicity
is considerably more complicated when gas aeration takes
place from the side of a standpipe for controlling the solids
flow rate. Such gas aeration is common in industrial opera-

I would like to conclude this part of my lecture with the
following thoughts:
Particle technology as exemplified byfluidization and
fluid-particle systems is an important interdisciplinary
area. Chemical engineers play a key role, as there are
many industrial applications in the chemical process
Education in particle technology is important from both
the industrial and the academic perspectives. Significant
progress in education and research in this area has
been made recently, such as increased textbook publica-
tions and increased industrial recruiting of U.S.-
educated graduates-but much remains to be done.
The most effective way to introduce particle technology
materials to chemical engineering students is through
such existing required courses as transport phenomena
Winter 2000

Figure 11. Coordinate
system for one-
standpipe flow.

and reaction engineering.


This lecture is dedicated to the
memory of Professor Shao-Lee
Soo of the University of Illinois,
Urbana. I am grateful to Prof.
Fernando Muzzio for insightful
discussions on powder mixing and
band formation, and to Dr. Fashad
Bavarian for providing the FCC
unit photograph used in Figure 1
and the information concerning
commercial operation of FCC
units. I am also indebted to Prof.
Jack Zakin and my research group
members, Dr. Jianping Zhang, Mr.
D.-J. Lee, Mr. Brian McLain, Mr.
Will Peng, and Mr. Guoqiang
Yang, who have provided con-
structive feedback in the prepara-
tion of this lecture material.

1. King, D., "Fluidized Catalytic Crackers: An Engineering
Review," in Fluidization VII (O.E. Potter and D.J. Nicklin,
eds.) p. 15, Engineering Foundation (1992)
2. Jenkins, III, J.M., R.L. Jones, and T.M. Jones, "Fluidized
Bed Reaction Systems," U.S. Patent 4,543,399 (1985)
3. Ennis, B., J. Green, and R. Davies, "Particle Technology:
The Legacy of Neglect in the U.S.," Chem. Eng. Prog., 90, 32
4. Nelson, R.D., R. Davies, and K. Jacob, "Teach 'Em Particle
Technology," Chem. Eng. Ed., 29, 12 (1995)
5. Klimpel, R., Instructional Module, "Introduction to the Prin-
ciples of Size Reduction of Particles by Mechanical Means,"
NSF/ERC of University of Florida, Gainesville, FL (1997)
6. Rhodes, M., Introduction to Particle Technology, John Wiley
& Sons, Chichester, England (1998)
7. Fan, L.-S., and C. Zhu, Principles of Gas-Solid Flows, Cam-
bridge University Press, New York, NY (1998)
8. Rhodes, M., and A. Zakhari, CD-ROM, Laboratory Demon-
strations in Particle Technology, Monash University,
Melbourne, Australia (1999)
9. Klinzing, G., "Pneumatic Transport and Solid Processing
Studies," Chem. Eng. Ed., 31, 114 (1998)
10. Martin, P.D., and J.F. Davidson, "Flow of Powder Through
an Orifice from a Fluidized Bed," Chem. Eng. Res. Des., 62,
162 (1983)
11. Fayed, M.E., and L. Otten, eds., Handbook of Powder Sci-
ence and Technology, Chapman and Hall, New York, NY
12. Nedderman, R.M., Statics and Kinematics of Granular Ma-
terials, Cambridge University Press, Cambridge (1992)
13. Knowlton, T.M., "Solids Transfer in Fluidized Systems," in
Gas Fluidization Technology, D. Geldard, ed., Wiley, New
York, NY (1986)
14. Chen, Y.-M., S. Rangachari, and R. Jackson, "Theoretical
and Experimental Investigation of Fluid and Particle Flow
in a Vertical Standpipe," I&EC Fund., 23, 354 (1984) 1

e learning


Part 4. A General Hierarchy

Based on the Evolution of Cognition*

Clemson University Clemson, SC 29634-0909

As their principal role, institutions of higher learning
are to develop and extend those high-level cogni-
tive skills that people need to function productively
in modern society.t" Such skills include complex abstract
thought, logical and mathematical reasoning, synthesis and
analysis, and the ability to recognize and apply patterns,
generalizations, theories, and schema to solve problems.
Such skills are developed by immersing students in a com-
munity whose members explicitly attempt to pass those skills
to other segments of the society. This difficult job is at-
tempted only at institutions of higher education. But though
we, as institutions, have been at this job for centuries, we
still do not have effective methods for accomplishing it.
In previous papers in this series, I presented a hierarchy of
technical understandings2'31 based on my experience in try-
ing to help students learn and on our current knowledge of
the structure and function of the human brain.[41 I will refer
to this as a special hierarchy of understandings.
But in addition to using observations of college students
and brains to obtain evidence for how learning occurs, we
can also pursue other routes. For example, Merlin Donald
studied the evolutionary history of culture from apes to
homo sapiens sapiens to show how high-level cognitive
skills probably developed.'51 And in another study, Kieran
Egan used mental growth in youngsters as the basis for a
theory of how humans learn.[6] Both these studies result in
cognitive hierarchies. That by Egan contains five levels of
human understandings: somatic, mythic, romantic, philo-
sophic, and ironic. I will refer to this as a general hierarchy
of understandings.

* Part 1, "Brain Structure and Function," CEE, Vol. 31(3), 152
(1997); Part 2, "Elementary Levels," CEE, Vol. 31(4), 214 (1997);
and Part 3, "Advanced Levels," CEE, Vol. 32(1), 30 (1998).

In this general hierarchy, it is the philosophic level that
encompasses the critical thinking skills required of engi-
neers. However, we cannot immediately begin instruction at
the philosophic level, because the special and general hierar-
chies are not merely sequential, but integrative: in such
models, mastery at any level requires assimilation, reorgani-
zation, and generalization of understandings gained at lower
levels. Hence, unless students have attained adequate facility
with somatic, mythic, and romantic thinking, they cannot
progress beyond a superficial level of philosophic under-
standing. Unfortunately, most students now entering engi-
neering schools in the U.S. are ill-prepared to develop tech-
nical understandings at the philosophic level. Moreover, vari-
ous strategies currently in vogue for addressing this prob-
lem-such as problem-based learning, discovery-based learn-
ing, group work, and web-based learning-are primarily
attempts to exercise thinking at the philosophic level. As
such, they fail to meet student needs at lower levels in the
hierarchy and therefore they are generally not as effective as
they could be. For some students, such learning exercises
are, in fact, counterproductive.
As engineering instructors, we are masters of philosophic
understanding, and we naturally want to teach what we do
best. But many engineering students are not prepared to
enter into philosophic modes of instruction. If those students
are not properly prepared, then philosophic instruction is
largely frustrating, and such students fail to develop the
skills we want them to have: ability to solve novel problems,
ability to extract meaning from data, ability to develop tech-
nical narratives that are well-reasoned and convincing, abil-
J.M. Haile, Professor of Chemical Engineering at Clemson University,
is the author of Molecular Dynamics Simulation, published by John
Wiley & Sons in 1992 and is the 1998 recipient of the Corcoran Award
from the Chemical Engineering Division of ASEE.

Copyright ChE Division of ASEE 2000

Chemical Engineering Education

ity to exercise sound engineering judgment. The question is,
can we do anything about it?


Before identifying levels of human under-
standings, we consider the demarcation between th
human and animal cognition. In animals, the and
highest levels of cognitive skills are found in hierar
chimps and the great apes. Beyond the instinc-
tive and procedural-habits characteristic of all not
animals, chimps and great apes are masters of sequel
the moment; they can contrive creative solu- integr
tions to problems as they arise. For example, such
they can combine available objects in new ways
and they can use available objects as tools to
achieve goals. Further, some individual apes level
have been taught a subset of American Sign asSin
Language.171 Donald refers to these achieve- reorga
ments as episodic learning. These kinds of
achievements are remarkable; nevertheless, they
are limited to the current situation-animals
live in the present. They do not plan for the
future. For example, they do not make tools of unden
their own. Although they may have used an gained
object repeatedly as a tool, they do not set it le
aside for future use. Even though they may
learn some sign language, they have never
made original contributions to their vocabu-
lary, much less created a grammar. In short, animals with
the most highly developed cognitive skills appear inca-
pable of abstract thought.


The first step beyond episodic learning is prelanguage and
relies on the sense of touch to gain and convey understand-
ing. For engineers, its important characteristics are tactile
learning, toolmaking, and communication by manual ges-
tures and body motions. Donald calls this mimetic learning,
but we follow Egan and call it somatic understanding. At the
somatic level, we have already taken a decisive step away
from episodic learning and into abstract thought. Thus, the
touching and manipulating of objects, which is characteristic
of tactile learning, seem to aid the human mind in learning to
create abstract images. We conjecture that mastery at the
somatic level is a prerequisite for later facility with highly
abstract thought. Thus, Newton was an accomplished ex-
perimentalist before he wrote the largely theoretical
Principia,81 Gibbs designed gears and brakes for railway
cars before he developed the abstract thermodynamics of
phase equilibria,191 and (to stretch the point only slightly)
Einstein worked with concrete inventions submitted for patent

before he developed the theory of relativity.'1t1 The close
interdependence of manual dexterity and abstract mental
processing has been emphasized in a book by Frank Wil-
son;"i' similarly, the connections between manual skills and

e special
chies are
ntial, but
native: in
ryat any
1 at lower

engineering talents have been emphasized in
an article by Petroski.'21
Toolmaking, which reverses tactile learning,
is the attempt to convert abstract images into
concrete objects. Toolmaking is taught in mas-
ter-apprentice relations with little verbal com-
munication; the instruction relies heavily on
gestures, physically realized procedures, and
concrete trial-and-error strategies. An en-
hanced remnant of this somatic mode of in-
struction serves as the basis for today's
graduate education.
Somatic modes of communication rely on
manual gestures and body motions-obvious
abstractions employed to convey ideas and re-
lations among concrete objects and situations.
There is a growing body of evidence to support
Donald's position that human language evolved
from manual gestures."1 Further, somatic forms
of communication are still employed in the per-
forming arts, in sign languages for the handi-
capped, and in signals used by referees and
umpires in sporting events.

An instructive example of somatic learning
has been documented in a recent article published in this
journal;4"" as a student, S. Godiwalla found herself frustrated
by instructors who consistently presented engineering sub-
jects at high levels of abstraction. She needed to see the
pumps, valves, and fittings that were being represented sym-
bolically in lectures; to understand, she needed to handle the
objects, look inside them, take them apart. Her response was
to find a technician who could help her convert abstract
symbols into concrete reality. It is germane to note that Ms.
Godiwalla was a double major in chemical engineering and
dance; thus, we have strong evidence for a student function-
ing at the somatic level.


In evolutionary terms, mastery of somatic skills serves as a
foundation for creating abstract names for concrete things,
then language, and then names for abstract things such as
virtue, patience, and deceit. Understandings at this level are
characterized by oral traditions, such as myths and epic
poetry, and so they can be called mythic understandings. In
an earlier paper in this series,2' I discussed the power that
primitive people attributed to names. That power becomes
extended and generalized when myths are used to explain

Winter 2000

Once a culture has established an oral tradition, it may proceed to
further levels of abstraction by creating graphic images for objects, situations,
and events. Such pictures, hieroglyphs, and other graphic devices may be followed by
creation of symbols for numbers, an alphabet, and writing.

how the world works. At the mythic level, understandings
are developed and conveyed through stories: oral structures
composed of an introduction that establishes a conflict, an
internally consistent narrative line, and a conclusion that
resolves the conflict. Egan reminds us that Carl Sagan and
Richard Feynman were both masters at presenting technical
material in narrative forms.161
To establish such narratives, storytellers usually create
conflicts based on binary opposites: good vs. bad, strong vs.
weak, industrious vs. lazy. For us as sophisticated instruc-
tors, this is a simple-minded way to view the world; further,
it leads to two-valued logic systems that are not merely
wrong, but dangerous.'15 (For example, "Never trust anyone
over 30." "All Democrats are liberals." "People who can't
do, teach.") Nevertheless, binary opposites are effective for
introducing new ideas, and they allow us to develop narra-
tive lines that conclude with discussions of engineering judg-
ment. In technical material, binary opposites rarely occur,
but the same advantages can be obtained by appealing to
binary alternatives; for example, we might introduce chemi-
cal processes as either batch or continuous, instruments as
either digital or analog, and pumps as either centrifugal or
positive displacement. The degree to which such a pair fails
to cover all possibilities would be left for later discussions at
higher levels of understanding.


Once a culture has established an oral tradition, it may
proceed to further levels of abstraction by creating graphic
images for objects, situations, and events. Such pictures,
hieroglyphs, and other graphic devices may be followed by
creation of symbols for numbers, an alphabet, and writing.
Note that graphic devices and writing involve abstractions
identified at the mythic level combined with manual dexter-
ity developed at the somatic level. This particular combina-
tion of manual and mental abilities may have prevented
some cultures from converting their oral traditions into writ-
ten language. Thus, some cultures remained at the mythic
level, while others developed graphic expression without
adding a written language. Graphics, numbers, and writ-
ing bring a richness and flexibility that is missing from
the somatic and mythic levels; however, these advan-
tages come at the price of greater difficulty in attaining
mastery at this level.

It is a command of graphic symbols and writing that
characterizes romantic understanding, so called because the
explanatory stories of mythic understanding are converted
into stories driven by human needs and aspirations. One
aspect of romantic understanding is an emphasis on bounds-
on the limits of human performance. Thus, at the romantic
level, we focus on the highest building, the longest bridge,
the fastest car, the most powerful rocket engine, and the
smallest (nanoscale) motor. The seven wonders of the an-
cient world were all made by man.
To illustrate the human context of a technical topic, let us
consider using the romantic mode for introducing the second
law of thermodynamics. In so doing, we would not merely
introduce such abstractions as entropy, irreversible processes,
and heat engines, we would place those abstractions within
the context in which they were invented: the needs driven by
the industrial revolution occurring in Europe in the early
1800s. To humanize the discussion, we could discuss the
personal histories of such figures as Sadi Carnot in France,
Rudolf Clausius in Germany, and William Thomson (later
Lord Kelvin) in Britain, whose efforts culminated in a for-
mal statement of the second law.
A second aspect of romantic instruction is that material is
not presented in a linear sequence; rather, the presentation
emphasizes salient points and ignores details. To illustrate,
Egan uses the metaphor of map-making. If we were to take a
romantic approach to mapping a country, we would not
proceed systematically from one coordinate to the next; in-
stead, we would locate the prominent features-the moun-
tains, lakes, rivers, canyons, gorges, and cities. Adding de-
tails involves understandings beyond the romantic. We find
it convenient to extend this metaphor by referring to the
"object" defined by this romantic activity as the conceptual
landscape for a topic.
A third aspect of the romantic mode is the uncovering of
interesting and unexpected connections. For example, Sadi
Carnot's work on heat engines was influenced by the inter-
ests of his father, Lazare Carot, who was minister of war
under Napoleon in 1800 and minister of the interior during
Napoleon's Hundred Days in 1814. When war erupted be-
tween France and Britain in 1792, France faced a possible
shortage of pencils. It was Lazare Carot who commis-
sioned Nicolas-Jacques Cont6 to develop a process for mak-
ing high-quality pencil lead from low-quality graphite.1161 By
about 1794, Cont6 had succeeded in inventing the crayons

Chemical Engineering Education

Conte, which are essentially our "lead" pencils. Thus, the
second law of thermodynamics is circuitously connected to
an instrument that contributed to writing and, hence, to the
spread of romantic understandings.
Connections are often interesting because they are
counterintuitive or amusing. For example, modem textbooks
routinely use the second law to prove that there can be no
perpetual motion machine. But it is amusing to note that
Sadi Carnot reversed the logic: he deduced the second law
from the assumption that perpetual motion machines cannot
exist.['71 Such connections serve as themes for popular es-

we force all students to construct plots on computers, for
somatic contact with the data is replaced by representations
and manipulations at higher levels of abstraction.


To our knowledge, all human cultures developed somatic
knowledge and mythic traditions and some developed ro-
mantic learning, but few developed philosophic understand-
ings; in fact, we know of only one such culture-the ancient
Greek. At the philosophic level, the graphic tools and writ-

says written by James Burke
and now regularly published
in Scientific American.
Besides bounds, connec-
tions, and human interest,
the romantic level invokes
pictorial symbols: diagrams,
flowsheets, plots, and other
figures that are characteris-
tic of engineering. At first
blush, there may seem to be
little to say about these de-
vices-they are taken for
granted in both engineering
education and practice-but
for this very reason, they
may be easily misused in
teaching. First note that al-


Figure 1. The x-y plot is a graphic device characteristic of
those deployed at the level of romantic understanding.
Nevertheless, such plots are late inventions in human his-
tory, coming long after romantic, philosophic, and ironic
understandings were fully developed. The first x-y plot was
apparently the musical staff, created by a Benedictine choir-
master during the Middle Ages."'1

though plots are romantic devices, they may invoke interpre-
tations at other levels of understanding. For example, some
plots can be interpreted in terms of the performance of
equipment; this appeals to somatic understanding. On other
plots, a curve might be interpreted in narrative terms as
illustrating a response to conflicts or competition between
variables; this appeals to mythic understandings. Still other
plots may be interpreted as expressing relations among terms
and quantities in equations; this appeals to philosophic un-
derstandings, as discussed in the next section. Thus on enter-
ing the romantic mode of learning, a student may readily
understand some plots, but have difficulty with others.
Second, note that interpreting an existing plot usually
involves lower levels of understanding than those used in
creating the plot. For example, the first x-y plot was, appar-
ently, the musical staff created by Benedictine monks during
the Middle Ages;[8' on the staff, pitch (frequency) of each
note is plotted on the ordinate, while time runs along the
abscissa, as shown in Figure 1. The musical staff is a
graphic-a romantic-evice created by philosophic think-
ing for use by mythic performers. Nevertheless, creation of a
plot can involve somatic elements that are beneficial to some
students; they gain understandings by manually transform-
ing a table of data onto graph paper. This benefit is lost when

ten language mastered at the
romantic level may enable de-
velopment of higher-order
thinking skills: inductive and
deductive logic, inferential
reasoning, analysis and syn-
thesis, critical thinking, cre-
ation of theoretical constructs,
and generalizations. These
abstractions relate, simplify,
and extend knowledge gained
at lower levels in the hierar-
Our experience implies that
the transition from romantic
to philosophic understandings
is a difficult one; in fact, in-
dividuals do not seem able to

make the complete transition by themselves.[61 That is, to
progress beyond a superficial level of philosophic under-
standing, an individual must reside in a community of philo-
sophic and ironic thinkers and learn from them. This is the
principal role of higher education in our society,1'61 although
the role is poorly understood by most students, many admin-
istrators, and some faculty.
To achieve technical understandings at the philosophic
level, we rely heavily on mathematical logic using equa-
tions. An equation is a romantic construct: a collection of
graphic symbols arranged to show relations among quanti-
ties and ideas. But even at this romantic level, many students
have difficulty distinguishing equations from formulae: for-
mulae are means for converting numbers into other numbers
(such is the use of the quadratic formula), while equations
are means for expressing relations. Of course, most equa-
tions can also be used as formulae, but their real import lies
in relating ideas, not numbers.
The use of equations in developing mathematical chains of
logic, however, is not a romantic activity, but rather a philo-
sophic one; examples include proofs, derivations, and the
deductions routinely employed in problem solving. Such
activities are highly abstract and require substantial sophisti-

Winter 2000

cation on the part of the student. As instructors, our tendency
is to underestimate the somatic, mythic, and romantic skills
that students must have mastered before they can manipulate
equations productively at the philosophic level. As Marvin
Minsky has noted, "it takes years to become proficient at the
language of mathematics."''1[
In addition to mathematics, we have a host of other de-
vices for developing and conveying philosophic understand-
ings; examples include problem-solving strategies, operat-
ing procedures, technical reports, computer programs, gen-
eralized patterns (such as the unit operations), and general-
ized theories (such as occur in transport phenomena). Such
philosophic devices are routinely explored and exploited in
our teaching and in this journal, so there is no need to
belabor them here.
Philosophic understandings develop from systematic ex-
plorations of a subject's conceptual landscape. In such ex-
plorations, we seek justification for the prominent features
identified at the romantic level; further, we seek to expose
the logical connections-the details- that relate the promi-
nent features. But such an exploration soon overwhelms us
with the innumerable details that establish the often-com-

plex web of connections among
important points. To maintain con-
trol over the material, we seek sim-
plifications via overriding patterns,
theories, schema, and generaliza-
tions that organize our knowledge
into structures that are useful; in
the words of Mach, we seek
economy of thought.120'
The reorganization of knowl-
edge into abstract and economical
structures is the characteristic ac-
tivity of learning at the philosophic
level. Following Vygotsky, we can
divide this activity into four

1. Conceptualization, which is
the creation or recognition of
a concept that arises from
observing concrete situations.
For example, placing a pan of
water on a hot stove might
lead us to the concept of heat
as an explanation for the
observed temperature rise.
Conceptualization may take
place at mythic, romantic, or
philosophic levels; often, it
incorporates features from all

2. Transference, which is the use of the concept to solve
problems in concrete situations other than the one that
inspired conceptualization. Thus, continuing with our
example, when we place the pan of water in a refrigera-
tor, we might again use the concept of heat, now to
explain the fall in temperature.
3. Generalization, which is creation of an abstract interpre-
tation of the concept, independent of any concrete object
or situation. Thus we might eventually generalize the
concept of heat to the more abstract notion of energy:
heat is a form of energy that "crosses" system bound-
aries. Exploration of the generalized abstraction might
lead us to generalized rules; for example, whenever the
net effect of a process is to add energy to a system, we
expect temperature to rise.
4.Extension, which occurs whenever we recognize
concrete situations, unlike those in conceptualization
and transference, to which the abstract form of the
concept can be applied. For example, we place ethanol
in an insulated vessel and then do work on it. We
understand that the temperature will rise because we
have added energy, even though no heat crossed the


Engineering Engin
Education Pract

Transference Extension




Figure 2. Psychological studies of learningJ211 and
neurological studies of brain function 2-41 confirm
that student understandings of abstractions de-
velop in a bottom-up learning strategy from con-
crete situations to abstract concepts. Thus, it is
counterproductive to attempt to teach conserva-
tion of energy by confronting students first with
the generalized energy balance. However, we ap-
ply abstractions in a top-down fashion, from ab-
stract notion to concrete situation. Thus in prob-
lem solving, students should be taught to start
with the generalized energy balance and then pro-
ceed deductively.

The articulation of these steps
helps us recognize a possible pit-
fall when using standardized tests
to assess student progress. It is
relatively "easy" to drill students
in conceptualization and transfer-
ence, so they can perform well on
standardized tests, but without the
ability to generalize and extend
what they know, such students
remain confined to a rather su-
perficial level of philosophic un-
Note that, as illustrated in Fig-
ure 2, we develop understandings
of abstractions by instructing in a
bottom-up mode: concrete situa-
tion to abstraction. But, we apply
abstractions in a top-down mode:
abstraction to concrete situation.
These two strategies, bottom-up
for engineering education and top-
down for engineering practice,
were deduced from Vygotsky's
psychological studies of language
acquisition in children;1211 but we
emphasize that they are consis-
tent with our earlier deductions
about proper learning strategies

Chemical Engineering Education


based on the current understanding of brain function.11-31
Note also that as students develop and practice these skills,
they often find extension, the transition from abstract to
concrete, to be just as difficult as generalization, the transi-
tion from concrete to abstract.1211
Finally, we must emphasize the dangers that are inherent
in the power of philosophic understandings: the command of
knowledge and economy of thought provided by patterns
and generalizations can easily seduce any of us into self-
We are particularly susceptible to self-deception at two
levels of philosophic development. One occurs at the novice
level, where the student's knowledge base is small, so nearly
any theory or generalization can organize and explain situa-
tions and events.161 The mild form of this disease leads to
overconfidence: the student considers his understanding com-
plete, so filling in details is considered to be an unnecessary
waste of effort. More severe cases lead to mental stagnation,
prejudice, and antisocial behavior.
The second window of susceptibility comes with mastery
of philosophic understanding of a particular, well-defined
and usually narrow, portion of a discipline. Though the
domain of knowledge may be small, it still requires years of
effort to master, so that although success is a true accom-
plishment, it may induce self-deception manifested as hu-
bris. A common symptom is the expectation that the pat-
terns, generalizations, and organizing principles found in the
restricted domain must apply to other domains; if they do
not, then those other domains are deemed unimportant
and can be ignored. Thus, we have scientists who treat
humanists with disdain, and humanists who treat scien-
tists with contempt. Such narrowly trained experts can
pose considerable dangers to a society, as was empha-
sized long ago by Ortega y Gasset.122'


If we are able to avoid or overcome self-deception, and if
we gain sufficient facility and experience with manipulating
knowledge at the philosophic level, then we may come to
realize that even the power of philosophic understanding is
limited. Any real situation is so complex that it is, at best,
only incompletely described by our abstractions, theories,
and generalizations; in fact, many real situations are not
described by any of our hard-won theoretical constructs.
Such realizations may drive us to a level of understanding
that Egan calls ironic.161
One aspect of ironic understanding is a proper perspective
on models; all our attempts to describe and explain reality
are merely models. At the somatic level, we use the human
body in our first crude attempts to model. At the mythic
level, the myths themselves serve as modeling devices.'[5 At

the romantic level, graphics and writing allow us to revise
the simple models of myths into more elaborate structures.
At the philosophic level, technical thinking is dominated by
mathematical models; at this level, we think we know much.
The transition to the ironic level starts when we realize we
still know very little.
As engineering instructors we are probably more comfort-
able than most with the roles that models assume in contrib-
uting to and limiting our understandings. As engineers we
routinely justify the use of a particular model in a given
situation by the a posteriori observation that it solved the
problem. "Whatever works" is laden with ironic overtones.
Nevertheless, engineering students have considerable diffi-
culty in recognizing models, in accepting their limitations,
and in selecting the appropriate model for a given situation.
For many students, "whatever works" is a cop-out rather
than a signal of subtle sophistication.
Another aspect of a properly developed ironic understand-
ing is an underlying sense of humor. To have successfully
completed the transition from the romantic to the philo-
sophic level, to have spent years in mastering a discipline at
the philosophic level, and then to realize that one still knows
little-such progression must drive an individual to either
despair or to humor. To react with humor is to recognize and
accept the irony of our lot.
More generally, the ironic thinker is sensitive to anoma-
lous situations that fail to adhere to the usual philosophic
patterns and theories. Such thinkers display considerable
insight in attaching abstract interpretations to concrete phe-
nomena, flexibility in manipulating concepts, and judgment
in combining models with formal theories. Ironic thinkers
are comfortable with multiple solutions, the lack of solu-
tions, ambiguity, uncertainty, and doubt.
It is probably too much to expect that in four years we can
bring many engineering undergraduates to even an opera-
tional understanding at the ironic level; nevertheless, we can
sow seeds for future growth. In our instruction, we can
continually emphasize the roles and limitations of models,
and we can give students exercises that force them to select
the model most appropriate for a given situation-such exer-
cises develop engineering judgment. To illustrate that many
situations have no single "right" answer, we can confront
students with open-ended problems; further, any prob-
lems having multiple solutions allow us to illustrate the
consequences of manipulating a situation to achieve dif-
ferent objectives.
Finally, we can exploit humor as an instructional device.
Elsewhere I have speculated about the probable relations
between humor and creativity.'3' Here it is appropriate to
twist an observation of Minsky's:'19' at the philosophic level,
engineering instruction is essentially the humorless activity
of using mathematical logic to establish connections, but

Winter 2000

ironic instruction contains a
humorous element that re-
laxes constraints and allows
the mind to seek unconven-
tional connections. Both
modes of instruction are
needed to start students to-
ward understandings at the
ironic level.


In this section we point out
that the hierarchy of techni-
cal understandings, intro-
duced previously,12-41 corre-
sponds to the general hierar-
chy,'16 which is described in
the previous sections. In fact,
the technical hierarchy is a

Special Hierarchy General Hierarchy

V. Ironic (models, exceptions, limitations)

IV. Philosophic (logical reasoning)
7. Creating extensions -Extensions (abstract to concrete)

6. Making connections Generalizations (concrete to abstract)
5. Posing Problems \
5. Posing Problems Transference (concrete situations)
4. Solving Problems/

3. Recognizing Patterns m. Romantic (graphics, writing, bounds)

2. IdentifyingElements Mythic (oral traditions)
1. Making Conversation/

I. Somatic (tactile learning, gestures)

Figure 3. The hierarchy of technical understandings
(left) introduced previously2-4' is a special case of the
more general hierarchy (right) developed by Egan.1"

subset of the more general one; this is illustrated in Figure 3.
The technical hierarchy begins, at its most elementary
level, with making conversation, and it continues with ar-
ticulation of definitions that identify conceptual elements.
These activities are fundamental to the oral traditions char-
acteristic of mythic understanding, for conversation leads to
storytelling, and both conversation and storytelling reveal
the need for a language composed of words having com-
monly accepted definitions.
The third level in the technical hierarchy is pattern recog-
nition; at this level we attach meanings, rather than mere
definitions, to a concept by relating it to other concepts. The
pattern formed in this way defines the conceptual landscape,
which is a product of romantic understanding.
The fourth and fifth levels of the technical hierarchy in-
volve problem solving and problem posing. These are the
principal activities that constitute transference of concepts
among concrete situations in philosophic understanding.
Making connections, at the sixth level of technical under-
standing, is the same as the philosophic exercise of general-
izing concepts from concrete situations to abstract ones.
Finally, at the seventh level of technical understanding,
creating extensions is the philosophic activity of applying
abstractions to different concrete situations. Thus, we have a
close and satisfying correspondence between the technical
hierarchy and the more general one.
In the next paper in this series we will discuss how the
general hierarchy can be applied to engineering education.

It is a pleasure to thank Pro-
fessor J. P. O'Connell of the
University of Virginia and Pro-
fessor K. Egan of Simon Fraser
University for offering con-
structive criticism on an early
draft of this paper.

1. Haile, J.M., "Universi-
ties...Why?" Chem. Eng. Ed.,
33(4), 288 (1999)
2. Haile, J.M., "Toward Tech-
nical Understanding. II. El-
ementary Levels," Chem.
Eng. Ed., 31, 214 (1997)
3. Haile, J.M., "Toward Tech-
nical Understanding. III. Ad-
vanced Levels," Chem. Eng.
Ed., 32, 30 (1998)
4. Haile, J.M., "Toward Tech-
nical Understanding. I.
Brain Structure and Func-
tion," Chem. Eng. Ed., 31,

5. Donald, M., Origins of the Modern Mind, Harvard Univer-
sity Press, Cambridge, MA (1991)
6. Egan, K., The Educated Mind, University of Chicago Press,
Chicago, IL (1997)
7. Savage-Rumbaugh, E.S., Ape Language: From Conditioned
Response to a Symbol, Columbia University Press, New
York, NY (1986)
8. Westfall, R.S., Never at Rest: A Biography of Isaac Newton,
Cambridge University Press, Cambridge, UK (1980)
9. Gibbs, J.W., The Early Work of Willard Gibbs, Henry
Schuman, New York, NY (1947)
10. Pais, A., 'Subtle is the Lord...'The Science and Life of Albert
Einstein, Clarendon Press, Oxford, UK (1982)
11. Wilson, F.R., The Hand, Pantheon Books, New York, NY
12. Petroski, H., "Work and Play," Am. Sci., 87, 208 (1999)
13. Corballis, M.C., "The Gestural Origins of Language," Am.
Sci., 87, 138 (1999)
14. Godiwalla, S., "What is Inside that Black Box and How Does
It Work?" Chem. Eng. Ed., 32, 306 (1998)
15. Hayakawa, S.I., and A.R. Hayakawa, Language in Thought
and Action, 5th ed., Harcourt Brace & Co., San Diego, CA
16. Petroski, H., The Pencil, A.A. Knopf, New York, NY (1998)
17. Carnot, S. Reflexions on the Motive Power of Fire, translated
and edited by R. Fox, Lilian Barber Press, New York, NY
18. Crosby, A.W., The Measure of Reality, Cambridge Univer-
sity Press, Cambridge, UK (1997)
19. Minsky, M., The Society of Mind, Simon and Schuster, New
York, NY (1986)
20. Mach, E., The Science of Mechanics, 6th ed., Open Court
Publishing, La Salle, IL (1960)
21. Vygotsky, L.S., Thought and Language, M.I.T. Press, Cam-
bridge, MA (1962)
22. Ortega y Gasset, J., The Revolt of the Masses, Norton, New
York, NY (1994) 0

Chemical Engineering Education

] book review

Separation Process Technology
By Jimmy L. Humphrey and George E. Keller II
Published by McGraw-Hill, New York, 408 pages includ-
ing index, $54.95 (1997)

Reviewed by
Phillip C. Wankat Purdue University

This is a book of tips, industrial applications, practical
flow sheets, new developments, and some theory, including
empirical expressions for industrially important separations
in the Chemical Process Industry (CPI). It contains a signifi-
cant amount of useful information gathered by the authors
during their years in consulting and industry-I took 19
pages of hand written notes while reading it.
The separations considered are those that are commer-
cially important within the CPI, with an emphasis on separa-
tion at the molecular level. Thus, distillation and its variants,
extraction, absorption/stripping, adsorption, and membrane
separation processes are covered in detail. The sections on
distillation, absorption/stripping, and adsorption appear to
be most authoritative. Mechanical separations such as set-
tling, cyclones, centrifuges, filtration and magnetic separa-
tors are not included. Processes that have some importance in
the CPI or are important in other industries, such as flotation,
crystallization, electrophoresis, and chromatography (except
for simulated moving bed systems), are also not covered.
Although not a textbook (the authors assume the reader
understands basic theories and there are no homework prob-
lems), this book is an excellent resource for professors who
teach separations. The authors discuss the latest commercial
advances and present a variety of applications; many tables
and figures collect useful information. Examples can be
used to provide a practical, industrial flavor to lectures.
There are useful ideas for the size ranges employed in indus-
try for different separations. The presentation is also particu-
larly good on rules of thumb that predict when to use par-
ticular separations. For example, the authors compare air
separation alternatives such as adsorption versus cyrogenic
distillation versus membranes. They clearly keep economics
in mind throughout the book. They note that the value of the
exponent in the "six-tenths" power rule explains why distil-
lation scales up well but doesn't scale down well, and why
membrane separations do the opposite. They discuss new
separation devices such as membrane contractors for absorp-
tion/stripping. Other examples include the extensive discus-
sions on new trays and structured packing.
For all its strengths, this book does have a few weak-
nesses. Some of the flow sheets (e.g., Figures 2.51, 2.59, and
5.22c) are clearly missing streams since they are not possible
Winter 2000

Use CEE's reasonable rates to advertise
Minimum rate, 1/8 page, $100
Each additional column inch or portion thereof, $40



The Department of Chemical Engineering invites applications for a
tenure-track faculty position at the Assistant Professor level. A PhD is
required and applicants must have an outstanding record of research
accomplishments and a strong interest in undergraduate and graduate
teaching. The successful candidates are expected to teach core chemical
engineering undergraduate and graduate courses, develop a research
program, collaborate with other faculty, and be involved in service to the
university and the profession. Applications from women and minorities
are encouraged. Interested persons should submit a detailed curriculum
vitae including academic and professional experience, a list of peer-
reviewed publications and other technical papers, and the names, ad-
dresses, and telephone numbers of three or more references to: Chair-
man, Department of Chemical Engineering, The University of Texas at
Austin, Austin, TX 78712-1062. The University of Texas is an Equal
Opportunity/Affirmative Action Employer.

as drawn. The selection of topics is occasionally mystifying
(e.g., the new rate analysis of distillation columns is pre-
dicted to eventually replace equilibrium analysis, but the
former is ignored while the equilibrium staged analysis is
explained in detail). There are a few places where the book is
a core dump instead of being fully digested. For example, on
p. 143 a figure and table from different sources presenting
extractor efficiency data don't agree, but the authors do not
comment on this discrepancy. In addition, some of the prac-
tical wisdom, such as "Although HTU is more rigorous,
HETP is used more frequently," may not sit well with pro-
fessors. These are really minor points considering the prac-
tical information and useful features in this book.
A wealth of practical information not available in text-
books is included. For example, the authors state that com-
posite polymer membranes may have the dense layer sepa-
rate from the support layer if they are back flushed. This
ruins the membrane and should normally be avoided. The
authors have collected a number of empirical equations that
give useful shortcuts. Examples include estimates of HETP,
the flooding pressure drop from the packing factor, the mini-
mum L/D, and the reboiler heat duty.
Additional useful features include tables of resources for
VLE data and predictions; an annotated list of suppliers of
simulation programs; a large number of references; an in-
dex; and appendices that include detailed design of distilla-
tion, lists of equipment suppliers, terminology for mem-
brane and membrane processes, and conversion factors.
Overall, this is a very good book. Every chemical engi-
neer who teaches, researches, or uses these separation pro-
cesses should have ready access to it. O

efI laboratory




Lehigh University Bethlehem, PA 18015

Exposing students to real process control instrumenta-
tion and to real process equipment in laboratory ex-
periments is of great pedagogical and motivational
importance. Simulations lack the feel, touch, smell, and sound
of a real chemical plant. All chemical engineering under-
graduate curricula should have some experimental dynamics
and control component.
Most control experiments used in undergraduate laborato-
ries are quite simple because they are intended to reinforce
basic ideas and principles learned in the theory part of the
course. Single-loop control of level, temperature, or pressure
is typical, with conventional PID algorithms used in digital
controllers. Many papers have been published over the last
four decades describing several types of laboratory experi-
ments and course objectives and approaches. Recent papers
include the work of Bequette, et al.,I'l and Lennox and Brisk.121
While these simple processes are necessary for initial ex-
periments, they do not expose students to processes that are
more challenging and that are commonly encountered in the
chemical industry. More complex experiments are of great
help in improving students' understanding of process-con-
trol basics. This paper describes one such process, the feed-
effluent heat exchanger/reactor system. The apparatus is
fairly simple and is safe to operate, and the investment in
process equipment and instrumentation is modest.
The experiment was developed as part of the Lehigh Inter-
disciplinary Controls Laboratory, which has been in opera-
tion for almost a decade.'31 The laboratory is an elective
course for chemical, electrical, and mechanical engineering
seniors. Experiments in the three disciplines are performed
by interdisciplinary teams. Students perform three very ba-
sic experiments during the first half of the semester.
During the last half they perform two more advanced

experiments. The laboratory is operated two afternoons a
week for three hours. Faculty supervision is provided by
all three departments.

The need to preheat the feed to a tubular reactor to some
minimum inlet temperature is one of the important features
of tubular reactors that distinguishes them from continuous
stirred-tank reactors in which a minimum feed temperature
seldom exists. With a tubular reactor, if the feed temperature
is too low, the reactor will "quench" (move to a low-conver-
sion steady state). Feed preheating can be done using a
steam-heated heat exchanger or a fired furnace, depending
on the temperature level required. Cooling of the reactor
effluent is usually required, and this can be done by steam
generation or using cooling water. The use of indepen-
dent utility streams for preheating and cooling makes the
control problem very easy because there is no interaction
(see Figure 1A).
This arrangement, however, is quite inefficient from a
capital-investment and energy standpoint. Separate heating
and cooling heat exchangers are required, which increases
capital investment in heat-transfer area. The need to have
reasonable temperature differential driving forces in both

William L. Luyben earned degrees in chemical
engineering from Penn State (BS 1955) and
Delaware (PhD 1963). His industrial experi-
ence includes four years with Exxon, four years
with DuPont, and three decades of consulting
with chemical and petroleum companies. He
has taught at Lehigh University since 1967 and
has participated in the development of several
innovative undergraduate courses, from the in-
troductory course in mass and energy balances
through the capstone senior design course and
an interdisciplinary controls laboratory.

Copyright ChE Division of ASEE 2000

Chemical Engineering Education

More complex experiments are of great help in improving
students' understanding of process-control basics. This paper describes one
such process, the feed-effluent heat exchanger/reactor system. The apparatus is fairly
simple and is safe to operate, and the investment in process equipment
and instrumentation is modest.

Preheated reactor
Cold feed feed efl--ent
Hiter -- Reactor c-.

Figure 1A. Independent heating and cooling.
1B. FEHE Process.

Figure 2. Process flowsheet.

Tin T. out
SHeating Element


SCR ---TC -

i--r Tout=ci+KR(Ti-c2e) ---

Figure 3. Simulate adiabatic exothermic reactor.

heat exchangers reduces the thermodynamic efficiency
of the process.
In a large number of industrial applications, the hot reactor
effluent is used to preheat the cold reactor feed. The result-
ing decrease in heat-transfer area means lower capital in-
vestment and a more efficient process in terms of steady-
state operation. Figure 1B shows a feed-effluent heat ex-
changer coupled with an adiabatic exothermic reactor. The
heat of reaction produces a reactor effluent temperature that
is higher than the temperature of the feed stream to the
reactor. Therefore heat can be recovered from the hot stream
leaving the reactor.
This feed-effluent heat exchanger (FEHE) configuration
results in significant dynamic control problems, however.
This is one of the classic examples of the interaction and
conflict between design and control. The control objective is
to control the reactor inlet temperature by manipulating the
bypass flow of cold material around the heat exchanger.
These FEHE systems have been used for many years in the
chemical and petroleum industries. Papers describing the
dynamics and control issues date back almost four decades,14,'5
with several recent studies appearing in the literature.16-101
The importance of the FEHE configuration in the chemical
industry is indisputable.

Figure 2 gives a schematic of the experiment. Air from a
90-psig air supply flows through a pressure regulator that
provides a constant pressure air source (50 psig). A manual
valve sets the total air flow through the process. The pres-
sure drop through the system is designed to be about 5 psi.
Thus the pressure drop over this manual valve gives choke
flow through the valve (Pin > 2 Pot,). This means that the
flowrate is independent of the downstream pressure, so the
total flow is constant for any downstream conditions.
About 25% of the air passes through the tube side of a heat
exchanger in which it picks up heat. The remainder of the
cold air is bypassed around the heat exchanger and mixed
with the hot air leaving the heat exchanger. The split-ranged
control valves in both the bypass line and the heat-exchanger
exit line are positioned by a temperature controller, which
controls the reactor inlet temperature. Since the safe failure
mode is bypassing cold gas to the reactor, the bypass valve is
air-to-close and the heat-exchanger valve is air-to-open.

Winter 2000

The total air stream after mixing enters a heater that is
used to add heat during startup and for openloop testing of
the individual components of the system. Then the air stream
enters the "reactor." Because of safety, environmental and
cost-of-raw-material concerns, we do not use real chemicals.
An exothermic adiabatic tubular reactor is simulated by us-
ing a vessel with an electric heater. The instrumentation
looks at the inlet temperature and adjusts power to raise the
exit temperature to the desired level. The reactor gain (i.e.,
how much the reactor exit temperature changes for a given
change in the inlet temperature) is set to simulate a typical
change in a chemical reaction rate with temperature. The
instrumentation to achieve this is shown in Figure 3. The
gain in the computing relay can be adjusted to give the
desired reactor gain. The hot gas leaving the reactor passes
through the shell side of the FEHE and a rotameter and is
vented to the atmosphere.

The power controllers in the heater and reactor are 110
volt SCR units driving the type of electrical heating element
used in heat guns (commonly used for paint removal). They

Equipment List


Heating elements
Temperature indicator
Heat exchanger

Control valves
Throttle valve
I/P transducers
Air supply regulator
4-way valves

Manual valves
Integral-orifice flow transmitter
Instrument air regulator
Panel lights and switches

SCR controller
Control console
Software for recorder
Piping and fittings
Computing relay

Flow switch
Relief valve
Air filter and drier

Flow indicator
Temperature transmitter
Support frame

Total Equipment Cost

Description Cost (/unit No.

See Fig. 9 1,100 2
30 2
450 1
500 1
700 1

=%trim, 1/2" trim 400 2
125 1
500 2
150 1
125 3

125 3
1,400 3
150 1
multi-trend, 8 inputs 3,500 1
100 3

2,100 2
1,800 1
300 1

reactor gain













Figure 5. Normal mode.
Chemical Engineering Education

Figure 4. Equipment layout.

have a power output of 600 watts. Table 1 gives a list of
equipment. Iron-constantan thermocouples are used for tem-
perature measurements at numerous locations in the process.
Integral-orifice differential pressure transmitters and a rota-
meter are used for flowrate measurements. Figure 4 gives
details of the piping and valving. Temperatures and flowrates
are recorded on strip charts on the control panel and logged
in a data-acquisition computer. The reactor-inlet tempera-
ture controller, the heater controller, and power switches for
the heater and reactor are also located on the panel.
The piping and valving are designed to have two modes of
1. Normal mode: Airflow is as described above (see Figure 5).
2. Test Mode: Airflow is split between two loops as shown in
Figure 6. In the first loop, airflows through the heater,
reactor, and shell side of the heat exchanger and is vented.
In the second loop, airflows through the tube side of the heat
exchanger and is vented. Changes in the power to the heater
produce changes in the temperatures in and out of the
reactor and in the temperature of the air leaving the tube
side of the heat exchanger. This data can be used to obtain
approximate transfer functions for the reactor and for the
heat exchanger.
The split-ranged control valves are 0.5 inch with equal-
percentage trim. The air-to-open valve in the heat exchanger
line is biased, as shown in Figure 7, so that it is wide open

Figure 6. Test mode.

when the controller output signal is at 50%. This keeps the
pressure drop through the valves fairly constant for any
controller output signal for a constant total flowrate of air,
which is set at about 8 SCFM. To protect against damage to
the heating elements, a low-flow switch cuts power to the
heating elements in the heater and reactor if the air flowrate
drops below 6 SCFM. This flow switch protects these ele-
ments in both the normal and test mode configurations.
Typical steady-state operating temperatures are:
inlet air, 70F
heat exchanger tube-side exit, 130F
mixed-gas, 115F
reactor inlet, 115F
reactor exit, 180F
heat exchanger shell-side exit, 135'F.
Heat-transfer area in the heat exchanger is 2.5 ft2.

Openloop Experimental Data Collection
Normal Mode
1. Position the valves so the gas flow is in the normal
operation configuration (Figure 5).
2. Open the supply air line and adjust valve V3 to get 8
SCFM on the rotameter. Turn on the main power
3. Calibrate the flow transmitters in the bypass and heat
exchanger lines by varying the reactor inlet tempera-
ture controller output (CO) from 0% to 50% with the
controller on manual. Read the total flow from the
rotameter and from the total flow meter. Make plots
of CO versus SCFM through each valve.
Test Mode
1. Position the valves so the gas flows can pass indepen-

Figure 7. Split-range control valves.

Winter 2000


Controller Output (%)




Controller Output (%)


dently through the reactor and the
heat exchanger in the test mode (see
Figure 6).
2. With the temperature controller in
manual, adjust the controller output
to 15%.
3. Turn on power to the heater and
adjust the heater controller to obtain
an exit temperature of about 1150F.
4. Turn on the power to the reactor. Let
the process come to steady state.
5. Calculate energy balances around
the heat exchanger and calculate the
overall heat-transfer coefficient U.
6. Make a step change in the heater
power and record the responses of
the reactor inlet temperature, the
reactor exit temperature, and the
heat exchanger tube-side exit
temperature. Use this data to
calculate the openloop transfer
function for the heat exchanger
relating TKx to Tou, and the openloop
transfer function for the reactor
relating T,, to T,u,.
7. Turn off the power to the heater and
reactor. Wait for two minutes with
air flowing to cool off the system.

Closedloop Experimental Data
1. Start up the system in the normal
operating mode with heater and
reactor power off. With the reactor
inlet temperature controller on
manual, set its output at 50%. Turn
on power to the heater and to the
2. When the reactor inlet temperature
reaches 950F, turn off the heater
power. Observe what happens to the
reactor inlet and exit temperatures.
3. Turn the power to the heater back
on. When the reactor inlet tempera-
ture reaches 120'F, turn off the
heater power. Observe what happens
to the reactor inlet and exit tempera-
4. Put the reactor inlet temperature
controller on automatic with K, = 5
(with a temperature transmitter span
of 1600F) and T, = 2 min, and with

CO KKK,2(z,s+ l)(zss+ ) PV
(z s + l)(rs + l)(rs + 1)- K,,,KR

Figure 8. Block diagrams of openloop coupled system.

Alumina Oxide RTC-60-2 Regular
Cotronics Corporation
3379 Shore Parkway Brooklyn NY.11235
NOTE:- Mold per Dwg.99030
NOTE:- Gnnd off ears to match element O.D.

P-03020-60 Heater Element 110VAC
Cole Parmer
Vernon Hills IL.60061-1844 5 BR Y PA.19539
DATE 4/17/99
DWG.NO. WE99028

Figure 9. Reactor/heater vessel

Chemical Engineering Education

a reactor inlet temperature setpoint of 115F.
5. Record the closedloop response of the system for
5'F changes in setpoint.
Theoretical Predictions
1. From the bypass and heat-exchanger flowrates and
temperatures, calculate the steady-state gain for the
transfer function relating Tin to the controller output
signal CO.
2. Assume this transfer function consists of a gain and
first-order lag. Use the same heater time constant
found above.
3. Using the three experimental openloop transfer
functions, predict the openloop response of the
coupled system.
4. Calculate the minimum controller gain that will
stabilize the coupled system.
5. Make a root locus plot and a Nyquist plot for the
coupled system.
6. Make root locus plots for PI controllers with the
values of TI = 1 min and T1 = 2 min.
7. Calculate the theoretical closedloop setpoint response
using the process openloop transfer functions and the
PI controller transfer function. Compare the predicted
response with the experimental response.

Figure 8 gives a block diagram of the individual compo-
nents in the openloop system. We assume that the dynamics
of the hot and cold stream mixing after the heat exchanger
are negligible, so the mixed-gas temperature Tm.x is related
to the flowrate of the bypass stream F, by the algebraic
Tmix(s) = KHX,2FB(s) (1)
Note that KHX,2 is negative since an increase in bypass flowrate
decreases Tmi. This means that the temperature controller
Gc() must increase the bypass flowrate when the temperature
increases. Since the bypass valve is air-to-close, its gain is
negative (increasing controller output signal CO decreases
bypass flow F,). Thus, the controller must have a positive
gain (reverse acting controller).
The transfer function relating bypass flow and controller
output is the valve transfer function, which we assume is just
a gain, FB/CO = Kv, and can be calculated from the experi-
mental valve calibration data.
The transfer function for the temperature transmitter, which
relates the controlled variable, Tn, to the signal to the con-
troller PV is

PV 100%
KT Span
Tin SpanoF

The mixed-gas temperature also depends on the tempera-
ture of the hot gas leaving the reactor (Tnot). The transfer
function relating T,,, to Tout is assumed to be a gain and first-
order lag

G s mix(s) HX (3)
Tout(s) THXS + 1
The reactor is assumed to have the simple openloop stable
transfer function with a negative pole at s = -1 / TR

G T out(s) KR
GR(s) = = (4)
R(s) n(s) T +RS + 1
These transfer functions can be combined, as shown in
Figure 8, to give the openloop transfer function of the coupled

G PV(s) KHX2KT Ky(TxS + l)(TRS +1)
CP(s) =CO(S) (THXS+)(TRS+ 1)(THs+ )-KHX,,KR


This equation shows that the coupled system is openloop
unstable if the product of the gains KHx.IKR is greater than
one. The heat exchanger gain KHx.I depends on the heat-
transfer area and the approach temperature differential on
the hot end of the process (the temperature difference be-
tween the entering hot stream and the exiting cold stream),
but it cannot be greater than unity. The reactor gain KR
depends on the heat of reaction, the temperature dependence
of the reaction rate, and the initial extent of conversion. In
the simulated reactor, this gain is set to be about four.
Root locus plots or Nyquist plots using a proportional
controller (Gc = K,) can be used to determine the mini-
mum value of controller gain that gives a stable closedloop
Note that the openloop transfer function of the coupled
system is net first order, i.e., the numerator polynomial is
second order and the denominator polynomial is third order.
This means that theoretically there is no maximum gain
(ultimate gain). Adding two first-order lags to account for
temperature measurement and valve lags gives a third-order
system, which has an ultimate gain and ultimate frequency.

KHX,2KTKv(THxS+l)(tRS+l) 1
(THXS + )(TRS + 1)(HS+ 1)-KHXKR ] (MS+ )(vs+ 1)

The experiment was constructed in late 1998 and operated
in the Interdisciplinary Controls Laboratory during the 1999
Continued on page 73.

Winter 2000

re, rankings




Carnegie Mellon University Pittsburgh, PA 15213-3890

he motivation for writing this article is to report on
an experience that we in the Department of Chemical
Engineering at Carnegie Mellon University had with
citation statistics. We believe it is worth sharing this experi-
ence, particularly in light of the recent article by Angus, et
al., "I who proposed alternative ways of measuring quality
for ranking chemical engineering departments. The main
idea in that article was to eliminate surveys and rely exclu-
sively on quantitative measures, with citations being one of
the major metrics. As we will describe in this article, great
care has to be exercised in gathering and interpreting these
data, as otherwise it is easy to obtain misleading conclu-
sions (see Centrar2l for a general discussion on problems
with citation analysis).

In 1992, Science Watch [3(2), pp. 1-8, April 1992] pub-
lished an article titled "Chemistry that Counts: The
Frontrunners in Four Fields." In that article, the table on
page 8 listed the following as the top six departments in
citations per paper during the period of 1984-1990:

Papers Citations Citations
# University 1984-90 1984-90 per paper
1 Carnegie Mellon University 98 670 6.84
2 Twente University of Technology 79 490 6.20
3 University of Wisconsin, Madison 106 629 5.93
4 University of Minnesota, Minneapolis 125 697 5.58
5 University of Texas, Austin 132 732 5.55
6 Massachusetts Institute of Technology 205 1134 5.53
The source used in that study was 58 dedicated journals of
chemical engineering (subsection of ISI Current Contents/
Engineering Technology and Applied Science).
On the other hand, according to the 1995 NRC Report,

Appendix Table P (p. 500), the ranking in terms of citations
per faculty for the five top U.S. departments and Carnegie
Mellon for the period of 1988-92 was:

# University
1 University of Minnesota, Minneapolis
2 Stanford University
3 University of Texas, Austin
4 University of California, Berkeley
5 Massachusetts Institute of Technology

41 Carnegie Mellon University


per Faculty

359 21.1

The source used was also the ISI Database and covered a
considerably larger, but unspecified, number of journals.
The studies covered different periods, 1984-1990 vs. 1988-
1992, as well as a different domain of journals. Neverthe-
less, it was clear that the number of citations reported in the
NRC Report for Carnegie Mellon seemed to be much lower.
In particular, the number of citations from the NRC study
(359) was one-half of the Science Watch study (670),
even though the NRC Report presumably covered a larger
number of journals.

--- Ignacio E. Grossmann is the Rudolph R. and
Florence Dean Professor of Chemical Engi-
neering and Department Head at Carnegie
Mellon University. He obtained his BS degree
his MS and PhD at Imperial College in 1975
and 1977, respectively, all in chemical engi-
neering. His research interests are in the areas
of process synthesis, energy integration, pro-
cess flexibility, planning andscheduling of batch
and continuous processes, and mixed-integer
and logic-based optimization.

Copyright ChE Division of ASEE 2000

Chemical Engineering Education

... great care has to be exercised in gathering and interpreting these

data, as otherwise it is easy to obtain misleading conclusions.

The above discrepancy prompted us to conduct an inde-
pendent study in the summer of 1996. We received two
databases from ISI containing the names and papers from
our faculty in the period 1981-1995. In the first database, no
biological science journals were included; in the second,
they were included. The numbers that we found were as
follows for the period 1988-1992 for Carnegie Mellon-
we included only active faculty (17 faculty, as in NRC
Report) during that period (retired or deceased faculty
were excluded):
a) Without biological sciences
Total citations 1241
Citations per faculty 73
b) With biological sciences
Total citations 2747
Citations per faculty 162
The reason for the large increase in citations with biologi-
cal sciences was that several papers were published by Rakesh
Jain in Cancer Research. For instance, one of his papers[31
had a total of 265 citations for the 1981-1995 period.
So, what can we conclude from the above numbers?
Even if we were to exclude the biological science journals
and remove 40% as an estimate for taking the data in 1996
rather than in 1993 (see point #3 below), the statistics are

Total citations
Citations per faculty
Therefore, compared to the
NRC numbers there was at
least a difference factor of two
in the number of citations. In
fact, if we consider the worst
case (only 745 citations),
Carnegie Mellon's rank
would have been 14, with 44
citations per faculty. In the
more realistic case of 1241
citations, our rank would
have been 6, with 73 cita-
tions per faculty. In both
cases, there is clearly a
rather large discrepancy
with the original rank of 41
for citations per faculty.
It should also be pointed
out that the number of publi-
cations per faculty reported
Winter 2000

30 1

for Carnegie Mellon in Appendix K (p. 286) is significantly
lower than it should be. That table reports 8.2 publications
per faculty compared to 14.9 from the ISI database (i.e., 254
publications and 17 faculty). Therefore, based on the count
of number of publications, only about one-half were consid-
ered in the NRC Report.
We contacted both the NRC and the ISI for clarification.
Based on their input, as well as on our experience in working
directly with the ISI database, we summarize below the
possible pitfalls that we identified with the citation statistics.

1. Misspelling of Names of Authors
This is a rather simple, but very critical, issue that we
found when requesting information from ISI. The two ex-
treme cases are 1) common names and 2) names that are
easily misspelled. For example, in our department the data
we received from John L. Anderson, our current Dean of
Engineering, contained a very large number of papers in
other areas. We had to manually separate the entries that
corresponded to our John Anderson because the database
did not allow simultaneous specification of both name
and affiliation.

At the other extreme, the name of the author of this manu-
script, Ignacio E. Grossmann, was initially misspelled with
one "n" (Grossman), and a similar difficulty occurred with
Andrew Gellman (Gelman).
As a consequence, we re-
ceived only a handful of cita-
tions in the initial request; the
ones with the misspelled last
names. Missing middle ini-
tials was another problem.

2. Domain of
Journals for Search
As the study in the 1992
Science Watch indicated, a
large number of journals was
excluded (compared to the
NRC Report) since the study
was confined to "chemical
engineering" journals. The
NRC, however, was also not
immune to problems. We
were told by its staff that only
certain disciplines were as-

81 82 83 84 85 86 87 88 89 90 91 92 93 94 95
Figure 1. Plot of citations versus time for IEEE, 69,
p. 1232 (1981)

sociated with each journal. This clearly means that depart-
ments with faculty publishing in the nontraditional disci-
plines were most probably penalized.

3. Timing of Measurement of Citations
Relative to Publication Times
In our study at Carnegie Mellon, we found the following
interesting observation: Most papers that have a significant
number of citations (say, greater than 50), achieve their
maximum number of citations between 4 to 6 years from
publication. An example of this is a paper by Arthur
Westerberg141 that had a total of 149 citations. As shown in
Figure 1, the maximum number of citations in this paper was
in 1986, five years after its publication. This trend held in
many of our papers. The implication is clear: Statistics for
papers that have been out for only 1 to 3 years will
probably miss more than 75% of the citations that such a
paper may receive.
4. Interpreting the Number of Citations
One of the most important issues when reporting the num-
ber of citations over a given time period is determining what
this exactly means. Intuitively, it may appear that it is the
total number of citations that were made to a given author
for that time period and for papers published prior to and
during that period. But if one uses an ISI database over a
given time period, the result obtained is only the number of
citations of papers published in that particular time period.
To give a specific example, consider the 1988-92 time
period that was used by NRC, measured in 1993. The total
number of citations would intuitively be the citations of
papers published before 1988-92 and during 1988-92. But
what one obtains from the ISI database is only the number of
citations of those papers published during the period 1988-
92. Therefore, according to point #3 above, for a 1988 paper
we pick up five years of the life of a paper, while for a 1992
paper we pick up only one year of its life. Aside from the
fact that this will be an inaccurate count that will greatly
underestimate the number of citations, it will be biased to-
ward papers that are cited earlier in their lifetime (i.e., papers
of immediate impact).
5. Variations of Citations by Areas
This is a well-known fact, but it deserves discussion. Let
us consider the two papers

1 Jain, R., "Determinants of Tumor Blood-Flow: A Review,"
Cancer Research (1988)
2 Fortescue, Kershenbaum, Ydstie, "Implementation of Self-
Tuning Regulators with Variable Forgetting Factors,"
Automatic (1981)

Up to 1995, paper #1 had 265 citations, while paper #2 had
195 citations. Based on point #3, we might say paper #1 may
still have some way to go to increase its number of citations.

Furthermore, it already has more citations than paper #2.
Should we then conclude paper #1 is more successful
than paper #2?
Consider the following fact: In Cancer Research the ex-
pected number of citations of any given paper is 163.6; in
Automatica it is only 15.9. (According to the ISI, the ex-
pected number of citations is the number of citations from
papers of that journal, divided by the number of papers in
that year.) If we divide the number of citations by the ex-
pected number of citations in the journal, one might argue
that paper #2 is ten times more successful!
Finally, a related issue in citation statistics is the "impact
score" of each journal, which often greatly varies by re-
search area and largely has to do with the size of its audi-
ence. The impact score is calculated by dividing the number
of citations in the past two years by the number of articles
published during the same period. Statistics reported (URL: in 1996 for some
journals where chemical engineers publish are


AIChE Journal
Industrial & Engineering Chemistry Research
Chemical Engineering Science
Canadian Journal of Chemical Engineering
Chemical Engineering Research & Design
Chemical Engineering Communications

25.466 Nature
22.067 Science






Pharmacological Reviews
Journal of Cell Biology
Journal of Clinical Oncology

Physics Letters A
Physics Letters B
Physical Review Letters

Journal of Chemical Physics
Journal of Catalysis
Surface Science

Journal of Fluid Mechanics
Colloids and Surfaces
Journal of Colloid and Interface Science
Environmental Science & Technology

Computers & Chemical Engineering
Mathematics of Operations Research
Mathematical Programming
Operations Research
European Journal of Operational Research
Chemical Engineering Education

Based on the above, it is clear that the impact factors are
quite uneven in different areas. Furthermore, it is interesting
to note that the Journal of Fluid Mechanics and Mathemati-
cal Programming, two journals that are notoriously difficult
for accepting papers, have rather low impact factors com-
pared to Nature and Science, which are also very selective.
These observations would suggest the need for some type
of normalization; e.g., 30 citations of an article in the
Journal of Fluid Mechanics could be considered a suc-
cess, while 30 in Nature would correspond to an average
paper in that journal.

This article has demonstrated, using the experience at
Carnegie Mellon with statistics on the numbers of citations,
that there are a number of important potential pitfalls in
compiling that type of information due to the great complex-
ity in gathering and interpreting the information. Our experi-
ence with Science Watch and the NRC Report suggest that
there is a pressing need for organizations that perform de-
partment rankings to carefully and rationally define mea-
sures of citations in order to avoid errors, misinterpretations,
and biases against certain research areas. While we do not
attempt to offer specific remedies, it would seem that the
following five general policies merit consideration:
1. Develop an identification number for authors to
avoid problems with misspellings and duplicate

2. Expand the domain of search to all areas to avoid
penalizing authors who publish outside of their
3. Consider normalizing citations according to the
impact score, to reduce discrepancies between
different research areas.
4. Ensure that the number of citations over a given
time period cover publications before and during
that time period.
5. Consult with departments to verify the statistics
before publishing them, to allow for possible
corrections of mistakes.
In order to implement these policies, close collaboration
with ISI and agencies performing the rankings is required. It
is also hoped that this article will temper the enthusiasm
of those who apply sophisticated analysis tools to ques-
tionable data, and thereby are likely to draw incorrect

1. Angus, J.C., R.V. Edwards, and B.D. Schulz, "Ranking
Graduate Programs: Alternative Measures of Quality,"
Chem. Eng. Ed., 33(1), 72 (1999)
2. Central, J.A., Reflective Faculty Evaluation, Jossey-Bass,
California (1993)
3. Jain, R., "Determinations of Tumor Blood-Flow: A Review,"
Cancer Research (1988)
4. Westerberg, Arthur, IEEE, 69, 1232 (1981) 0

leeletter to the editor

Dear Editor
Alves, et al.,1m are to be congratulated on their elegant
experiment on the drainage of liquids from vertical tubes of
diameter 19 and 32 mm. But their paper should have stipu-
lated that the diameter of tube used should be no less than
about 15 mm. The reason is that the simple equation for the
bubble velocity

U= 0.345(gD)05 (1)

does not apply for smaller tubes because of surface tension
effects. The relevant dimensionless group for such effects is
the Ebtvos number, which is given by

Eo=gD2Ap/o (2)
where D = tube internal diameter, Ap = density difference
between heavy and light phases, and a = surface or interfa-
cial tension.
At values of Eo less than about 50 in the case of low-
viscosity liquids, the velocity U is below the value predicted
from Eq. (1). Moreover, if Eo is below a critical value of
Winter 2000

3.37, corresponding to a tube diameter of about 5 mm for air/
water, the bubble will not rise at all.J21 Some years ago we
measured drainage rates from vertical tubes of between 6
and 10 mm diameter, with the bubble nose (meniscus) being
controlled at a constant position. 31 This technique enabled
the surface tension in gas-liquid and liquid-liquid systems to
be measured continuously.

M.H.I Baird, Professor
N.V. Rama Rao, Research Associate
McMaster University

1. Alves, M.A., M.F.R. Pinto, and J.R.F.G. de Carvalho, "Two
Simple Experiments for the Fluid-Mechanics and Heat-
Transfer Laboratory Class," Chem. Eng. Ed., 33, 226 (1999)
2. Bretherton, F.P., "The Motion of Long Bubbles in Tubes," J.
Fluid Mech., 10, 166 (1961)
3. Rao, N.V.R., and M.H.I. Baird, "Continuous Measurement
of Surface and Interfacial Tension by Stationary Slug
Method," Can. J. Chem. Eng., 61, 581 (1983) 0

Random Thoughts...


North Carolina State University Raleigh, NC 27695

It's a typical day in your class. As you lecture,
several students stroll in during thefirst 10 minutes of
the class and one arrives after 20 minutes. It is the
earliest she has arrived all semester.
a number of students are absorbed in the campus news-
two students are having an animated conversation,
punctuated by laughter. All heads around them are
turning to see what's going on.
> one student has his head back, eyes closed, and mouth
You are not thrilled by all this, but you're not sure what to
do about it.

We sometimes present this scenario in our teaching work-
shops and ask the participants to brainstorm possible re-
sponses to any of these behaviors-not just good responses,
but good, questionable, and terrible responses. Here are
typical suggestions.
1. Ignore it.
2. Lock the door.
4. Fall silent and wait.
5. Throw chalk.
6. Set off a firecracker.
7. Flap your arms and cluck like a chicken.
8. Ask a question.
9. Leave.
10. Set fire to the newspaper.
11. Talk to the offender outside class.
12. Review the rules.
13. Start an activity.
14. Throw the bums out.
15. "That looks like an interesting conversation over
there-why don't you share it with the rest of us?"
Next, we suggest that the best response depends on whether

the offending behavior is disruptive or non-disruptive-that
is, whether or not it distracts the class's attention from your
teaching-and whether it is a first offense or a recurring one.
Non-disruptive behaviors include sleeping (without snor-
ing), reading, or slipping into the back of the room late. You
may not like it-seeing students asleep drives some instruc-
tors crazy-but it is not distracting to the other students.
(Watching someone sleeping just doesn't have that much
entertainment value.) Disruptive behaviors include talking
or otherwise making noise, or coming in late and promenad-
ing ostentatiously up the aisle.
After making these distinctions between different offend-
ing behaviors, we tell the participants to get into groups of
three or four and try to reach consensus on the best response
for each category. We collect their nominations and then
propose ours. Sometimes several groups nominate our re-
sponses; often none do.
You might enjoy making your own nominations before we
tell you ours. In your opinion, what is the best way to deal
(a) a student sleeping in class whom you have never seen
sleeping before?
(b) a student who sleeps in almost every class session?
(c) two students talking and laughing who have not done
so before?
(d) two students talking and laughing who do so frequently?
First indicate what you would do in class when you ob-

Richard M. Felder is Hoechst Celanese Professor (Emeritus) of Chemical
Engineering at North Carolina State University. He received his BChE from
City College of CUNY and his PhD from Princeton. He has presented
courses on chemical engineering principles, reactor design, process opti-
mization, and effective teaching to various American and foreign industries
and institutions. He is coauthor of the text Elementary Principles of Chemi-
cal Processes (Wiley, 2000).
Rebecca Brent is an education consultant specializing in faculty develop-
ment for effective university teaching, classroom and computer-based
simulations in teacher education, and K-12 staff development in language
arts and classroom management. She co-directs the SUCCEED Coalition
faculty development program and has published articles on a variety of
topics including writing in undergraduate courses, cooperative learning,
public school reform, and effective university teaching.

Copyright ChE Division of ASEE 2000
Chemical Engineering Education

serve the offensive behavior, and then add what (if anything)
you would do outside class. Hint: One of our nominations is
not included in the 15 listed ones.

Best response to non-disruptive behavior
If you do anything in class to address a non-disruptive
behavior, you turn it into a disruptive one. Our suggestion
for what to do in class about a sleeping (or reading or
unobtrusively late) student is, therefore...nothing. If the stu-
dent is a first-time offender, forget about it. If you notice the
same student sleeping every period, you may continue to
ignore it, or if it seriously annoys you, you might express
your annoyance outside class and ask why he is doing it. If
he is bored, knowing that his sleeping bothers you may get
him to work harder at staying awake. On the other hand, if
he is holding down a 40-50 hour/week job while going to
school or is working the night shift, warn him that he could
be missing important information and then stop worrying
about it.
Sometimes someone suggests initiating a learning activity
to get students' attention. We are staunch believers in active
learning, but we want to use activities when they fit, not just
because we happen to see someone sleeping.

Best response to disruptive behavior
Ignoring disruptive behavior is not a viable option. If you
allow disruptions to proceed, they will become increasingly
widespread and frequent until the class is out of control.
Our nomination of the best response requires some pre-
liminary explanation. Speech communication experts tell
us that there are three categories of responses to objection-
able behavior: aggressive, passive (indirect), and assertive.
Yelling at students, throwing things at them, and throwing
them out of class are aggressive responses. Doing anything
non-aggressive other than clearly stating what you want is a
passive response. Calmly and clearly stating the problem
and asking for what you want is an assertive response.
Do aggressive responses work? In the short run, they
generally do. As an instructor, you hold a great deal of
power over the students: if you scream at them to shut up,
chances are they will. But while you may win the battle, you
are likely to lose the war. When you resort to aggression,
you effectively admit that the only way you can control your
class is to lose control of yourself. You will lose the respect
of the students, and the rest of the semester could be grim for
both you and them.

* We are indebted to Rebecca Leonard of the N.C. State Univer-
sity Department of Communication for the analysis that follows.

What about throwing the chalk or an eraser? Everyone has
stories-some fond, some bitter-about teachers they had or
knew about who used to do that sort of thing. That was then;
this is now. Can you say "law suit"?
Then there are passive responses. Ignoring those two
chattering students-the ultimate passive response-is clearly
a poor idea. Falling silent and waiting for them and other
noisemakers to quiet down themselves might work eventu-
ally, but it wastes valuable class time (especially in a large
class, where you might wait for a long time) and penalizes
the non-disruptive students as much as the few miscreants.
Locking the door penalizes chronic latecomers, but it also
penalizes the one-time offender who may have a perfectly
legitimate and unavoidable reason for being late.
Some professors argue for the ever-popular "Why don't
you share that joke with the rest of us?" That is, first of all, a
passive response. You are not asking for what you really
want: the last thing in the world you want is to know what
those two birds are twittering about. You know, and they
know, and the rest of the class knows, that your goal is
simply to embarrass them into quieting down. Will it work?
Again, probably in the short term, but once you resort to
sarcasm or anything else that has embarrassment as its ob-
jective you again lose respect that may be hard or impossible
to regain.
Which brings us to our nomination: the direct, assertive
response. Look in the direction of the offending students and
calmly say "Excuse me-that noise is disrupting the class.
Could you please keep it down?" They usually will. The
talkers may be mildly embarrassed but your primary objec-
tive was clearly not to embarrass them-it was simply to
quiet them down. You maintain control without having to
use aggression or sarcasm, and the students' respect for your
authority stays the same or increases.
Finally, what if you have to quiet down the same students
in several classes, or the same student keeps coming in late?
We propose doing the same thing we suggested for repeated
non-disruptive behaviors. Talk to the offenders outside class,
telling them that their behavior is offensive and must stop,
and then ask them why they're doing it. Regardless of what
they say, you will probably achieve your objective. In our
combined years of teaching, we have never had to do this
with a student more than once. Barring pathological cases,
neither should you.
Interestingly, the assertive response-simply asking the
offenders to stop doing what they're doing-is usually not
on the list of possibilities brought up during the initial brain-
storm. It's almost as if instructors don't know it's legal to do
it. It is legal. And it works. 0

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

Winter 2000

= class and home problems

The object of this column is to enhance our readers' collections of interesting and nove
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 nove
home problem, are requested, as well as those that are more traditional in nature and tha
elucidate difficult concepts. Manuscripts should not exceed ten double-spaced pages if possible
and should be accompanied by the originals of any figures or photographs. Please submit them t(
Professor James O. Wilkes (e-mail:, Chemical Engineering Depart
ment, University of Michigan, Ann Arbor, MI 48109-2136.


A Feasibility Study

University of Zagreb Zagreb, Croatia

istillation is one of the most widely used separation
processes in the chemical and allied industries.
Among possible distillation schemes with energy
saving, incorporation of a heat pump can be very efficient-
but any decision in a particular distillation problem should
be very carefully considered. First and foremost, a separa-
tion process must be feasible; i.e., it must have the potential
of giving desired results. The decision of whether to use a
heat pump or not depends on the economics of the process.
An established practice is to make a preliminary design
or feasibility study to decide whether to proceed with
detailed calculations.
Before digital computers, the economic design of multi-
component fractionating towers and other vapor-liquid con-
tacting devices was a tedious, difficult, and time-consuming
job. Various "short-cut" methods were developed to sim-
plify the task of designing multicomponent columns, and
those methods are still useful in preliminary design work,
where they can greatly reduce the number of calculations to
be made. With the advent of large-capacity, high-speed
digital computers and sophisticated calculation tech-
niques, however, many new and powerful programs for
process design have been developed.'" One of them is
ChemCAD III,'21 -a powerful and comprehensive chemi-
cal-plant simulation program.

* Polimeri d.o.o., Zagreb, Croatia

The initial design includes a prelimary flow diagram, ma-
terial and energy balances, determination of the equipment
size, and estimation of equipment costs and utility energy.
The preliminary flow diagram shows the arrangement of
basic individual units of equipment. The flowsheet com-
mands of ChemCAD III allow for a very simple creation of
the flowsheet. The flowsheet menu has two basic purposes:
1. To define the flowsheet; i.e., the unit operations and
their connectivity

-Ljubica MatijaJevic is an assistant professor
of plant design at the University of Zagreb,
Croatia. She received her BS, MS, and PhD
degrees in chemical engineering from the Uni-
versity of Zagreb. Her main interests include
plant design, mathematical modeling, and mass
and heat transfer.

Eduard Beer is a specialist for process de-
sign and development for Polimeri do.o.
Zagreb, Croatia. He has more than 35 years
of experience in development and design in
the petrochemical industry. He holds a BS
and MS in chemical engineering from the
University of Zagreb.

Copyright ChE Division of ASEE 2000

Chemical Engineering Education




2. To establish the calculation order of theflowsheet
The full simulation program is capable of carrying out rigor-
ous simultaneous heat and material balances and prelimi-
nary equipment design. Our example will describe the tech-
niques for solving multicomponent distillation problems of
an interconnected flowsheet.

Distillation with Vapor Recompression

Three arrangements of heat pumps are possible."-" In each
case, compression work is normally used to overcome the
adverse temperature difference, which precludes having the
condenser serve as the heat source for the reboiler in an
ordinary distillation column. If conditions are suitable, the
process fluid can be used as the working fluid for the heat
pump. In alternative arrangements, the process vapor is taken
from the top of the column, compressed, and fed to the
reboiler to provide heating. In distillation systems where the
heat pump is applicable,'14 the heat pump with direct over-
head vapor recompression proves to be the most economic
solution. The compressor is the "heart" of the system. The
ratio of compression is crucial to the power requirements,
and depends on

Apcom = Apcol + Apb + Apc + Aeq (1)
Relations between pressures and temperatures are given on
the p/T diagram shown in Figure 1.
The ratio of compression is

r 2 (Pt +Apcom) (2)
Pi Pt
A vapor compression heat pump applied to a distillation
column is shown in Figure 2.

Capital cost estimates for chemical process plants are of-
ten based on the purchased cost of the major equipment
items required for the process, the other costs being esti-
mated as factors of the equipment cost. Costs are correlated
against sizes of individual units. The relationship between
size and cost is given by Eqs. 3 through 14.[6'81 The equip-
ment-sizing facility inside ChemCAD III can be found under
the command describing each equipment type. The follow-
ing correlations for the base cost in carbon steel are given in
SI units and US dollars.
Pressure vessels

Cp = CFmFp 336.(3)

Cv = (a+bL)d'' (4)
Winter 2000

80 Tt 90 Tb 100 110 120 130 140
T, temperature, C

Figure 1. Diagram of p/Tfor overhead vapor

Figure 2. Distillation column with heat pump.

In the following two correlations, column cost represents the
sum of the costs of the shell and internals, either trays or

Cco = Cpv.vert + Cpack

C pc C0 dnV
Cpack packd Vpack 336.2

Heat Exchangers

Mass flow rate.

(5) Hp
(5) Qmass(H:Og) Hp

(6) Price of steam: $JO/t

Cost = 10 Qmass

0 I
Ce = C FF
exc exc p tex 336.2

Co = exp(a + b fnA)


o I
Cpu = CpFmFp
PU 418.3
COp =exp{a +nP[b+ nP(c+d nP)]}


o I
cm =C F
co com m 418.3

Co =exp(a+bn P)

Cem CF F I
em = CmtemFna n 1

C = exp{5.33+fn P[0.3+ nP(0.162-0.014n P)]} (14)


The word "utilities" is generally used for the ancillary
services needed in the operation of any production process.'81
The services in this example include electricity, steam for
process heating, and cooling water. The quantities required
can be obtained from the energy balances and the flowsheets.
The prices will depend on the primary energy sources and
the plant location.

Cooling water
Mass flow rate of water

mass(H20,,,) C AT

Price of cooling water: $0.02/mm

Qmass(HO,iq )
Cost = 0.02 Qm
P water

Electric power
(7) Cost = $0.07 / kWh

All costs of utilities are based on one year or 8,000 hours.
The base date for utilities energy is the start of 1987.31

A flow rate of 22,000 kg/h of liquid has the composition
(10) shown in Table 1. The mixture has to be distilled to give an
overhead product with 0.15 mole percent styrene and a bot-
tom product with 0.3 mole percent ethyl benzene. The nec-
essary information for using the ChemCAD III program
includes working pressures and temperatures along the col-
tri i umn and preliminary material balances. The column will

operate at a pressure of 0.23 bar at
the top and a temperature of 97 C at
the bottom. The pressure drop along
the column is 0.02 bar.
The preliminary flow diagram and
preliminary material balances are
shown in Figure 3 and Table 2. Fig-
ure 3 shows a standard column con-
figuration, i.e., conventional distilla-
tion that was modified to use a heat
pump with the stream transfer mod-

Feed Composition

Benzene 0.8
Toluene 1.2
Ethyl benzene 40.0
Styrene 58.0

ules (STM). These modules serve as an abstract block in the
flowsheet, and their output stream is exactly the same as
their input stream.

The feed rate is 22,000 kg/h. The mean molecular weight

Preliminary Material Balance

Feed Top Bottom
Comp kmol/h fraction kmol/h fraction kmol/h fraction
Benzene 1.680 0.008 1.680 0.019 0 0
Toluene 2.520 0.012 2.520 0.029 0 0
Etbenzene 84.000 0.400 83.750 0.950 0.250 0.002
Styrene 121.800 0.580 0.190 0.002 121.610 0.980
Total 210.000 1.000 88.140 1.000 121.860 1.000

Chemical Engineering Education

of the liquid mixture is

M = Y, x,Mi= 0.008 x 78.114 + 0.012 x 92.141 +
0.40 x 106.168 + 0.58 x 104.52 = 104.61

Thus, the molar flow rate of feed is 22,000/104.61 = 210.3


> Distillation Column
Type and size of packing: Intalox ceramic saddle (3 inch)
Height required for the specified separation: 86 m
Column diameter (capacity): 7.5 m

> Distillation Storage Tank
Conventional Distillation
Height (width), m 8.23
Diameter, m 2.74

Heat Pump

- Vapor-Liquid Separator

Height (width), m
Diameter, m

Heat Pumo

- Heat Exchanger: Heat duty and area
Heat duty Q, kJ/h (EO1, E05) -1.4 x 10O
Heat duty kJ/h (E02. E04) 1.4 x 10

Heat duty Q, kJlh (E03)
Area, m2 (EOI, E05)
Area, m2 (E02, E04)
Area, m2 (E03)


Heat Pump
-1.74x 107
3.73x 106
1.41 x 10o

Pumps: Pumping power
Distillation Heat Pump

Power, kW(PO1, P03) 100.7 2.5
Power, kW (P02, P04) 1.9 1.9

- Compressor

Compressor power: P=4,205 kW


> Conventional Distillation:
Purchased equipment cost

Equipment Items
Tower TO]
Pump POI
Pump P02
Vessel DO]
Condenser E01
Reboiler E02

> Utilities Cost

Cooling water
Electrical power

2 to I

Price: $

Cost: 1/year


, I Re iux STi

1^1 Z


P03 Top
29 3

P02 2 S7! FM 2
[] Bottom 9 to 32
1tto 1 to 3e S

Figure 3.


Winter 2000

-I El
"1D (J3

^ [s

> Distillation with Heat Pump: Purchased Equipment Cost

Equipment Items
Tower TOJ
Pump PO.
Pump PO'
Condenser EO,
Condenser EO.
Heater EO.
Compressor CO
Separator DO.
Storage tank DO.


> Utilities Cost
Cooling water
Electrical power

> Summary of Annual Costs

(amortized over ten yea,

Price: $


Cost: $/year

Conventional Heat
Distillation: $/vear Pump:$/vear

$253,630 $650,890




The purpose of this example is not to obtain a detailed
calculation, but to illustrate the methodology of a prelimi-
nary design of equipment and obtain a feasibility study. The
preliminary design includes shortcuts in calculations so that
the cost estimates of equipment probably contain moderate
errors. There are two alternative designs for the distillation
and the aim of this work is to find which of them is more
attractive economically. Both designs are feasible.
The results show that the application of the heat pump
with vapor recompression is more economically justified.
The purchased cost of the major equipment items required
for the process with the heat pump is 2.5 times higher
than for conventional distillation, but the energy costs
saved easily justifies the investment after one year's op-
eration of the equipment.
The above example shows the benefits of making a pre-
liminary calculation to prove whether or not to proceed with
detailed calculations.

a,b,c,d, coefficients in Eqs. (4) and (8)
A heat transfer area, m2
Co cost of column, $
Ccom cost of compressor, $
Com base cost of compressor, $
Cem cost of electromotor, $





base cost of electromotor, $
cost of exchanger, $
base cost of exchanger, $
cost of packing per unit volume, $/m3

base cost of packing per unit volume, $/m3
cost of pump, $

base cost of pump, $
cost of pressure vessel, $
cost of vertical vessel, $

base cost of pressure vessel, $

d diameter, m
Ap, vapor pressure difference between overhead and bottom
product at column bottom temperature, bar
Apc pressure difference between hot and cold side in
evaporator/condenser, bar
Apco, column pressure drop, bar
APeq equipment and pipes pressure drop, bar
Apcom growth pressure in compressor, bar
Fm material cost factor
Fna electromotor purpose cost factor
Fp pressure cost factor
Fm electromotor type cost factor
Frtx exchanger type cost factor
I inflation index
L length of column, m
P power, kWh
p, compressor suction pressure, bar
p, compressor discharge pressure, bar
Pb column bottom pressure, bar
pt column top pressure, bar
Q heat transferred per unit time, kJ/h
Qv flowrate, m3/h
r compression ratio
U overall heat transfer coefficient, kJ/m2K
Vpack volume of packing, m3
ATm mean temperature difference, the temperature driving
force, K
11p pump efficiency

1. Winkle, V., and W.G. Tood, "Optimum Fractionation Design
by Simple Graphical Methods," Chem. Eng., 20, 136 (1971)
2. ChemCAD III, Process Flow Sheet Simulator, User Guide,
Chemstations, Inc., Houston, TX (1995)
3. King, C.J., Separations Processes, 2nd ed., McGraw-Hill,
New York, NY (1980)
4. Beer, E., "Heat Pump in Distillation," Kem Ind., 41(8), 309
5. Meili, A., and A. Stuecheli, "Distillation Columns with Di-
rect Vapor Recompression," Chem. Eng., 94(2), 133 (1987)
6. Sef, F., and Z. Olujic, Projektiranje Procesnih Postrojenja,
SKTH, Zagreb (1988)
7. Sinnot, R.K., Chemical Engineering, Vol. 6, 2nd ed., Jordan
Hill, Oxford (1996)
8. Ulrich, G.D., A Guide to Chemical Engineering Process De-
sign and Economics, J. Wiley, New York, NY (1984) O
Chemical Engineering Education

Laboratory Experiment
Continued from page 61.

spring semester. Four groups of students ran the experi-
ment. Each group was in the laboratory collecting data
during three consecutive three-hour laboratory periods (a
week and a half). The theoretical analysis took about a
week. Reports were due at the end of the third week.
The initial experimental temperature data was inconsis-
tent and unreliable due to transmitter and recorder calibra-
tion problems. We also had to reduce the size of the control
valves and change the trim from linear to equal percentage.
Once these hardware bugs were worked out, the students
were able to collect and analyze good data. Energy balance
revealed significant heat losses because of the small scale of
the equipment. Additional insulation was added.
Although the digital data logger is convenient for storing
data for later plotting and for making calculations, an old-
fashioned multichannel analog strip-chart paper recorder
made it much easier to follow the dynamic transients as
they occurred.
The response of the students was quite positive. They
found the experiment both challenging and educational.
The chemical engineering students were particularly inter-
ested in this experiment because many of them were work-
ing in their design course on dynamic simulations of chemi-
cal processes that featured reactor/heat-exchanger systems.

The coupled system is openloop unstable, but whether the
temperature increases or decreases depends on the initial
conditions. At low-temperature initial conditions, tempera-
ture drops exponentially. At high-temperature initial condi-
tions, temperature increases exponentially. This is what
should occur in Steps 2 and 3 of the Closedloop Experimen-
tal Data Procedure. The mechanical equivalent of this phe-
nomenon is the trajectory of an inverted pendulum; it can
fall to either one side or the other, depending on the initial
position and velocity.
Runs with different reactor gains can be made to illustrate
how this parameter affects the rate of the openloop runaway
(or quench) and how it affects controller tuning.
Another modification is to run with different heat ex-
changer areas. This can be achieved by having two heat
exchangers in series (on both the tube and shell sides) that
can be valved in or out of the system.


The process and instrumentation hardware was designed
and constructed by Matthew (Scotty) Chalmers and Bill
Bechtel. Funds were provided by Ernst Benzien and the
Pennsylvania Equipment Grant Program, and instrumenta-

Winter 2000

tion equipment was provided by Honeywell.

F, flowrate through bypass
Gc controller transfer function
GCP openloop transfer function of coupled system
G heater openloop transfer function relating Tin to To.
G HX. heat exchanger transfer function relating Tout to Tn
GHX,2 heat exchanger transfer function relating Tout to F,
GR reactor openloop transfer function relating Tun to Tou,
Kc controller gain
KH gain of heater openloop transfer function relating Tn to
KHx. gain of heat exchanger transfer function relating Tou to
KHx.2 gain of heat exchanger transfer function relating Tout to
KR gain of reactor openloop transfer function relating Tin to
Ku ultimate gain
TH time constant of heater openloop transfer function
relating Tin to Tmix [min]
THX.I time constant of heat exchanger transfer function relating
Tout to T [min]

THX,2 time constant of heat exchanger transfer function relating
Tou to F, [min]
TR time constant of reactor openloop transfer function
relating Tin to Tout [min]
T, controller reset time [min]

1. Bequette, B.W., K.D. Schott, V. Prasad, V. Natarajan, and
R.R. Rao, "Case Study Projects in an Undergraduate Pro-
cess Control Course," Chem. Eng. Ed., 32(3), 214 (1998)
2. Lennox, B., and M. Brisk, "Network Process Control Labo-
ratory," Chem. Eng. Ed., 32(4), 314 (1998)
3. Johnson, S.H., W.L. Luyben, and D.L. Talhelm, "Under-
graduate Interdisciplinary Controls Laboratory," J. of Eng.
Ed., 84 (2), 133 (1995)
4. Douglas, J.M., J.C. Orcutt, and P.W. Berthiaume, "Design
and Control of Feed-Effluent Exchanger-Reactor Systems,"
Ind. Eng. Chem. Funds., 1, 253 (1962)
5. Anderson, J.S., "A Practical Problem in Dynamic Heat Trans-
fer," The Chem. Engr., 97 (1966)
6. Silverstein, J.L., and R. Shinnar, "Effect of Design on the
Stability and Control of Fixed Bed Catalytic Reactors with
Heat Feedback. 1. Concepts," Ind. Eng. Chem. Proc. Des.
Dev., 21, 241 (1982)
7. Tyreus, B.D., and W.L. Luyben, "Unusual Dynamics of a
Reactor/Preheater Process with Deadtime, Inverse Response
and Openloop Instability," J. Proc. Cont., 3, 241 (1993)
8. Jones, W.E., and J.A. Wilson, "An Introduction to Process
Flexibility. 1. Heat Exchange," Chem. Eng. Ed., 31, 172
9. Luyben, W.L., "Internal and External Openloop Instabil-
ity," Ind. Eng. Chem. Res., 37(7), 2713 (1998)
10. Luyben, W.L., B.D. Tyreus, and M.L. Luyben, Plantwide
Process Control, McGraw-Hill Book Co., New York, NY (1999)

ME laboratory



University of Delaware Newark, DE 19716

Separation and purification processes account for 70 to
90% of equipment and energy costs in modem chemi-
cal plants,'" with distillation being among the most
expensive.121 Moreover, phase equilibrium measurements and
thermodynamic modeling are key elements of chemical pro-
cess design, and it is therefore not surprising that vapor-
liquid equilibria (VLE) receives heavy emphasis in the un-
dergraduate chemical engineering curriculum. Students typi-
cally learn the fundamentals of VLE and other phase equi-
libria as part of their thermodynamics course work, but
many programs rely on a VLE laboratory exercise to rein-
force and extend the knowledge gained in the classroom.
The classical undergraduate experiment involves measur-
ing the VLE behavior of a binary mixture, often by direct
compositional measurement at constant temperature and pres-
sure. This requires charging a still with the binary system of
interest and allowing the system to equilibrate, after which
one records the temperature and pressure inside the still and
obtains samples of both phases for compositional analysis.
Common analytical techniques (e.g., gas chromatography
and refractive index) are time-consuming, allowing acquisi-
tion of just a few data points during a typical laboratory
session and precluding a meaningful data reduction. Stu-
dents therefore miss an opportunity to practice the thermo-
dynamic modeling skills that are crucial to process design.
In a recent paper, Campbell and Bhethanabotla addressed
the pitfalls of using direct compositional analysis in the
undergraduate laboratory and recommended the total-pres-
sure method as an alternative approach to acquiring VLE
data.[31 This method, in which students record pressure iso-
thermally as a function of (known) liquid composition over a
wide composition range, eliminates the need for composi-
tional analysis. Instead, vapor phase compositions are calcu-
lated by employing a suitable activity coefficient model in

Currently at Drexel University, Philadelphia, PA 19104-2875
2 Currently at DuPont Central Research & Development,
Experimental Station, Wilmington, DE 19880-0262

accordance with the method of Barker.141 A counterpart of
the total pressure method is to record isobarically the boiling
temperature for multiple liquid compositions, an approach that
has been in use at the University of Delaware for many years.
An important issue in any undergraduate laboratory exer-
cise is the effective use of limited time, and time is certainly
a key consideration in the industrial practice of thermody-
namics where project deadlines dictate a need for the rapid
acquisition of accurate VLE data.151 Although the total-pres-
sure method is adequate in this regard, we found that signifi-
cant improvements are possible via the method of infinite
dilution, isobaric ebulliometry. This method allows acquisi-
tion of data points every fifteen minutes and can be used with
a modest number of data points near the compositional limits.
Given a set of four ebulliometers, groups of three students
are able to measure the infinite dilution VLE behavior of two
binary systems during a single laboratory session. If the two
binary systems share a common component, then students
can model the VLE behavior of a ternary system by combin-

Steven P. Wrenn is Assistant Professor of Chemical Engineering at
Drexel University. He received his BS from Virginia Tech (1991), and his
MchE (1996) and his PhD (1999) from the University of Delaware, all in
chemical engineering. He served as a Teaching Fellow in Chemical Engi-
neering at the University of Delaware in Fall 1997. His research interests
include the study of biological colloids with applications to human physiol-
ogy and disease.
Victor S. Lusvardi is a research engineer working in Central Research
and Development at the DuPont Experimental Station. He received his BS
from the University of Illinois in 1990 and his PhD from the University of
Delaware in 1997, both in chemical engineering. His professional interests
include heterogeneous catalysis, reaction engineering, and surface sci-
George Whitmyre is the University of Delaware Chemical Engineering
Laboratory Coordinator. He upgrades experiments in undergraduate labo-
ratories and facilitates research safety activities. He earned his BS in
zoology from Pennsylvania State University and his MS in entomology and
applied ecology from the University of Delaware.
Douglas J. Buttrey is Associate Professor of Chemical Engineering at the
University of Delaware, where he has been since 1987. He received his
BS in biology from Wayne State University (1976) and his MS in chemistry
(1978) and PhD in physical chemistry (1984) from Purdue University. His
research interests include phase equilibrium studies of complex oxide
materials, heterogeneous catalysis, and electronic materials.
Copyright ChE Division of ASEE 2000
Chemical Engineering Education

ing their experimental results with data for a third binary
mixture taken from the literature. This paper describes the
implementation of infinite dilution ebulliometry in the un-
dergraduate laboratory and its application to ternary VLE.


Ebulliometry allows measurement of VLE by the prin-
ciple of liquid-phase and vapor-condensate recirculation,
and Cottrell was the first to use an ebulliometer for under-
graduate instruction (in 1910).12 The Cottrell design, which
used a thermal lift pump, was later modified by
Swietoslawski. Although alternative designs abound, the
original Swietoslawski design remains the standard in
ebulliometry (for a review, see Halal61).
Ebulliometers operate exceptionally well with dilute solu-
tions and allow accurate determination of infinite dilution
activity coefficients, y-. The practical definition, in our
VLE context, of an infinitely dilute solution is a mixture for
which the temperature changes linearly with mole fraction,
and the limits of infinite dilution are specific to a given
mixture. Typical limits for infinite dilution range from less
than a fraction of a mole percent to several mole percent.
Thus, infinite dilution ebulliometry refers to the isobaric
measurement of VLE temperatures in a narrow composition
range near the limit of infinite dilution rather than over a
wide composition range. Working near infinite dilution
limits the number of data points necessary, and the use of
small aliquots avoids the need to drain and refill systems,
thus conserving time.
The isobaric, infinite dilution technique involves boiling a
pure liquid (solvent) of known weight and adding a very
small (typically less than 0.5 mole percent), weighed amount
of a second fluid solutee). The addition of solute alters the
boiling point, and one records the new boiling temperature at
a fixed pressure after the system reaches equilibrium. By
recording the equilibrium boiling temperatures at constant
pressure for several successive aliquots of solute, one ob-
tains the dependence of boiling temperature on solute mole
fraction within and slightly beyond the limit of infinite dilu-
tion. This information is sufficient to calculate activity coef-
ficients at infinite dilution according to



where yolute is an activity coefficient at infinite dilution, P
is the total pressure, Pvap is the vapor pressure, T is the
absolute temperature, and x is mole fraction, where sub-
scripts denote a given component. All qualities involving the
vapor pressure are evaluated at the pure solvent boiling
temperature using a readily available vapor pressure correla-
Winter 2000

tion (e.g., Antoine, Reidel, or Harlecher-Braun)171 The infor-
mation obtained from the ebulliometric experiment is merely
the limiting value of the partial derivative

(aT / Ox olute )p,x ,,o1

and computation of the liquid phase activity coefficients at
infinite dilution from Eq. (1) is straightforward.
Activity coefficients at infinite dilution become useful
when one considers the criteria for VLE, namely the equality
of component fugacities among phases at uniform tempera-
ture and pressure. It is customary to represent the liquid
phase fugacity with an activity coefficient (as opposed to an
equation of state) so that under conditions of low total pres-
sure the equilibrium criterion for any component, i, becomes

xiYi(T,P,xi)Pivap (T)= yP (2)
where x, and y, denote liquid- and vapor-phase mole frac-
tions, respectively, yi is the liquid-phase activity coeffi-
cient, Piva(T) is the pure-component vapor pressure, and P
is the total-system pressure. Pressure remains constant, the
liquid-phase mole fraction is calculated by a mass balance,
and vapor pressures are calculated in the manner described
above. The only unknowns in Eq. (2) are the activity coeffi-
cient and the vapor-phase mole fraction. When activity coef-
ficients are known, one can easily calculate the composition
of the equilibrium vapor phase; the challenge is to evaluate
the activity coefficients.
Most activity coefficient models (e.g., Wilson, van Laar,
two-constant Margules) for binary systems contain two sys-
tem-dependent parameters, the values of which are generally
unavailable. As an example, the binary van Laar model may
be written

A21 x2
n Ti = A12X AI + A21x2

n y2 = A21 + A12x
A12X1 + A21X2)

where subscripts 1 and 2 denote the two binary components
and the parameters A12 and A21 are unknown. Under condi-
tions of infinite dilution, Eq. (3) simplifies to

en y7 = A12
en y7 = A21 (4)

where yT is the activity coefficient of component i when
component i is present as the (infinitely dilute) solute. Thus,
the infinite-dilution technique provides a means of estimat-
ing the two model parameters, A12 and A21, which can then
be used to calculate activity coefficients and the equilibrium
vapor-phase mole fraction for any liquid composition. This
technique can be implemented rapidly, and the values of y7

determined experimentally lead to fairly reasonable predictions of VLE over the
entire range of binary compositions.81] Azeotropic temperatures may be slightly
over- or underestimated, but generally the composition of an azeotrope, if
present, is well targeted.
Moreover, the technique allows estimates of ternary VLE when the procedure is
applied to two additional binary systems with a common component. A second
mixture, made of components 1 and 3, leads to the binary van Laar activity coeffi-
cient parameters AI3 and A31. Similarly, a third mixture comprising components 2
and 3 gives the binary van Laar activity coefficient parameters A23 and A32. The six
binary parameters enable calculation of activity coefficients in a ternary system
comprising components 1, 2, and 3 according to the ternary van Laar model[91

S+ A21 A3 2
2 2X+3 1 13) +X2X3 A1 2 3 A23

Al2 A13

A similar expression for eny2 is obtained by interchanging subscripts 1 and 2 in Eq.
(5) and for ny3 by interchanging subscripts 1 and 3.

Our undergraduate VLE laboratory is equipped with four ebulliometers to allow
simultaneous measurement of four activity coefficients (i.e., two binary mixtures).
Ideally, six ebulliometers would be used for
studying the ternary system, but preserving the
division of labor among three students during
a single laboratory session (i.e., 3-4 hours)
dictates a practical limit of four ebulliometers. 6 m. 0.D. tubing. 1 I
The ebulliometers were constructed to our
specifications by A.A. Pesce, Inc., and are a
slight modification of the original o o. j.
Swietoslawski design.r61 Figure 1 shows the
primary components of the ebulliometers: the n 'cl on Pen
boiler (A), the Cottrell pump (B), the 2- mm tophon
thermowell (C), and the condenser (D).
Each ebulliometer is packed with fiberglass Themnow
insulation, secured in a wooden housing, and
mounted on a steel frame. The network of
T 6 mm 0.0
ebulliometers, which is set on casters for ease comin ,
of transport, is positioned beneath a central
fume hood for ventilation (see Figure 2). The
boilers are supplied separately by four circu-
155 mm -
lating hot-water baths (Fisher Scientific model B oilr o to io .
9101), and each thermowell is equipped with a (m wh Sap broken 1
particles fused to In
platinum RTD (Omega model PR-13-2-100-1/ urtce.)
8-12-E) and digital display (Omega model t No. 30 nichrom. win
DP41-RTD). The circulating baths are control-
lable to within 0.1 C, and the precision of the mm telon e Ip..
RTDs is 0.010C. The condensers are supplied by
a central cooling system that consists of a 15- mNT
gallon polyethylene tank (Nalgene model 14100)
and a centrifugal pump (Cole-Parmer model Figure 1. Schematic diagram of a

7021). The cooling medium is an ice/
water bath that is maintained at about
2C. Ambient pressure is measured to
a resolution of 0.1 mm Hg with a
centrally located mercury barometer.
Operation of the ebulliometer in-
volves charging a known mass of sol-
vent (~60 mL) to the boiler, which
can be accomplished by simply pour-
ing the solvent from a tared beaker.
The solvent is then heated to its boil-
ing point via the circulating hot water
bath. When boiling commences, ris-
ing vapors within the Cottrell arms
entrain small droplets of liquid so that
a two-phase flow impinges upon the
thermowell. The liquid phase falls un-
der the influence of gravity, whereas
the vapors continue to rise into the
condenser. Condensed vapors flow to
the bottom of the thermowell, where
they mix with the descending liquid
phase, and all liquid returns to the boiler.
An important point to make with
students is that by its very nature the

modified Swietoslawski ebulliometer.
Chemical Engineering Education


Spiral glass beads on
outside of thelloell
Cooled Condenate
S mm O.D.

ebulliometric experiment represents a steady-state process
when stabilized. The intimate contacting of vapor and liquid
in the Cottrell pumps, however, assures achievement of local
equilibrium; boiling temperatures of pure solvents measured
with the ebulliometer agree to within 0.050C of the accepted
equilibrium values.
When steady state is reached, a small aliquot (100-1000
LL) of solute is added via a gas-tight syringe through an
injection port on the thermowell. This method for loading
solute is necessary to avoid errors in weighing, since the
solute mass must be known to within +0.005 g to obtain
values of (aT / axsolute )p,x lutO with less than 1% error. This
is less of a concern for the solvent mass, which requires a
precision nominally one-hundred fold less than that of the
solute. A new steady state is reached shortly after injecting,
and the new steady-state temperature is recorded on a plot of
temperature versus (liquid phase) solute mole fraction. This
process is repeated until the steady-state temperature pro-
file becomes nonlinear, indicating that the range of infi-
nite dilution has been exceeded.

Students are assigned one ternary system to be modeled on

Figure 2. Experimental apparatus used in the under-
graduate laboratoryfor VLE studies.
Winter 2000

the basis of the three associated binary systems. The students
are required to obtain the two activity coefficients at infinite
dilution (one for each component) for each of two of the
binary systems. They obtain infinite dilution activity coeffi-
cients for the third binary system from the Dechema Data
Series, and we encourage students to obtain the original refer-
ences cited therein.1101 Moreover, we require that students fa-
miliarize themselves with the Material Safety Data Sheets
(MSDS) for each chemical that they will be using and to
understand the implications of the information when perform-
ing the experiment. A wide variety of solvents can be selected,
but we typically choose a combination of binary pairs that
includes at least one azeotrope. Table 1 is a listing of the
ternary solvent systems used in recent years at the University
of Delaware. Note that several of these reagents, in particular
methanol and chloroform, require careful consideration in view
of safety issues associated with handling.

To illustrate the infinite dilution technique, we refer to an
experiment from the Spring 1997 semester in which students
modeled the behavior of the ternary system acetone-methyl
acetate-methanol. Students made measurements on the binary
systems acetone-methanol and methyl acetate-methanol, and
results for the acetone-methanol binary will be shown. The
raw data obtained from a single ebulliometer are provided in
Figure 3. Temperature was measured continuously as a func-
56.20 Figure 3. Raw data
Injection #1 obtained from the
I isobaric experi-
56.00 ment. The boiling
Injection #2 temperature in an
/ Iebulliometer filled
55.80 with acetone is re-
Injection #3 corded as a func-

55.60 Injection#4 aliquots of solute
Snj. #5 methanol are
S / added. Arrows indi-
55.40 cate the times at
P = atm. which aliquots were
S injected, and hori-
55.20 zontal bars denote
0 20 40 60 80 100 120 the ensuing steady-
Time, minutes state temperatures.

Table 1
Solvent Systems Used in
Ternary VLE Experiments

1997 Acetone Methyl Acetate Methanol"
1997 Acetone Methanol" Chloroform
1996 Isopropanol Cyclohexane Ethyl Acetate
1996 t-Butanol Acetone Hexane
1995 Ethanol Methyl Ethyl Ketone 2-Propanol
1995 Ethanol Acetone Ethyl Acetate

tion of time, and values were recorded every three min-
utes. Arrows in the figure denote solute injections, and
plateaus in the profile of temperature versus time indi-
cate the ensuing steady states. Under these conditions 56.0(
the temperature at the thermowell is taken to be that of
the equilibrium state as discussed earlier. 55.8(
A more useful form of the results in Figure 3 is a plot
of equilibrium boiling temperature versus solute mole 55.6(
fraction, and this is depicted in Figure 4. The plot exhib-
its curvature as cumulative injections begin to exceed 55
the limit of infinite dilution. Fitting the data with a 2nd-
order polynomial, the initial slope gives the desired
result (aT/ xMeOH )pxMH _0, which is used to com- 55.2
pute the infinite dilution activity coefficient, YMeoH'
The value of y1eoH = 1.8 was obtained using the three-
constant Antoine vapor pressure correlation and leads to
a value for the van Laar activity coefficient parameter of
Al2 = 0.59.
Constructing plots similar to Figures 3 and 4, in which acetone is the
solute, provides all the information necessary to estimate the full VLE
behavior for the acetone-methanol binary pair. Students use that infor-
mation to perform bubble-point calculations and generate a T-x-y
diagram (Figure 5). The lines in Figure 5 show the acetone-methanol
VLE behavior as predicted by the van Laar activity coefficient model
when using the experimentally determined values of A|2 and A,,. The
highlight of Figure 5 is the identification of an azeotrope at a methanol
mole fraction of 0.22, and students recognize the negative impact this
will have on distillation. Although the partial VLE data obtained from
the infinite dilution experiment are sufficient to estimate the entire
VLE behavior, there is no guarantee that the predicted VLE is correct.
Inaccuracies could stem from the infinite dilution measurements or
from the choice of activity coefficient model that may not be suitable
for the combination of solvents being studied. Students separate these
effects by first comparing experimental data with accepted literature
values within the range of infinite dilution. This allows a check of the
ebulliometric technique itself in that it provides an estimate of the
errors in the slopes (aT/ xsolute )px, that are used to predict the
VLE at intermediate compositions.

Figure 4. Deter-
mination of infi-
nite dilution ac-
tivity coefficients.
Steady-state tem-
peratures from
Figure 3 are plot-
ted as a function
of methanol liq-
uid mole fraction.
The initial slope
is used to com-
pute the infinite
dilution activity
coefficient for
methanol via Eq.

0.0 0.2 0.4 0.6 0.8 1.0
Methanol Mole Fraction
Figure 5. T-x-y diagram showing the VLE behavior for
the acetone-methanol binary system. Lines represent
predictions of the van Laar activity coefficient model,
based on results from the infinite dilution ebulliometric
experiment. Symbols denote literature data taken from
the DECHEMA series9' (open=vapor, filled=liquid). The
minimum reveals an azeotrope at a methanol mole
fraction of 0.22 and a boiling temperature of 55.3 C.

Students then test model suitability by calculating the
complete VLE behavior using a variety of activity coef-
ficient models. We require students to compare the
predictive capabilities of models that assume random
mixtures (e.g., two-constant Margules or van Laar) with
at least one model that accounts for local compositional
correlations due to differences in solute-solute, solvent-
solvent, and solute-solvent interactions and hence as-
sumes non-random mixtures (e.g., Wilson, NRTL, and
TK-Wilson[21'). Note that the NRTL model requires a
third parameter (i.e., the pre-factor, A), which cannot be
determined from this experiment but which tends to fall
in the range 0.2 the acetone-methanol VLE behavior as determined by
various authors, and students are required to compare

0.0 0.2 0.4 0.6 0.8 1.0
Methanol Liquid Mole Fraction

Figure 6. x,y dia-
gram showing the
VLE behavior of the
binary system. Pre-
dictions from the
ebulliometric ex-
periment, using the
van Laar model, are
shown as a solid
line. Symbols repre-
sent literature data
(same as in Figure
5). The presence of
an azeotrope is in-
dicated by intersec-
tion with the 45

Chemical Engineering Education

Figure 7. Isotherms in the ternary system
acetone-methyl acetate-methanol. Contours
of constant temperature are plotted as a func-
tion of liquid-phase mole fractions. The iso-
therms converge in a "bulls-eye" fashion to
reveal a ternary azeotrope at a composition
of 6 mole% acetone, 62 mole% methyl ac-
etate, and 32 mole% methanol. The azeo-
trope is minimum boiling at a temperature


o o o o o Acetone
'o "-3 O

Figure 8. Contours of K,=1 in the ternary
system acetone-methyl acetate-methanol.
Contours of the partition coefficient, K,, are
plotted for each component in the case where
each partition coefficient is one. Thus, the
mole fraction of any component is the some
in both phases along the contour for that
component. The intersection of two contours,
along the edge in which the mole fraction of
the third component is zero, represents a
binary azeotropic composition. The intersec-
tion of all three contours indicates the pres-
ence of a ternary azeotrope at a composition
of 7 mole% acetone, 61 mole% methyl ac-
etate, and 32 mole% methanol.

the predictions of each activity coefficient model with the literature studies.
Any variation in the goodness-of-fit between models, as compared with
both the infinite dilution data obtained by the students and the literature data
at intermediate compositions, requires students to consider which model
best captures the physical differences (e.g., size or polarity) between solvent
and solute molecules.
Another way in which students check their results with literature studies is
in the form of an x,y diagram (see Figure 6), identical to those used to
construct McCabe-Thiele diagrams when sizing distillation columns. Lines
again represent model predictions, and symbols denote literature studies. It
is pleasing that the azeotrope obtained in Figure 6 matches that of Figure 5
as it should, but it is not uncommon (although it is not the case here) for
predicted azeotropes to agree more favorably with literature compositions
than with temperatures. Over the years we have found that the ebulliometric
method leads to azeotropes that are nearly always correct in composition,
but not necessarily correct in temperature, and this is true regardless of
which activity coefficient model is used.
We attribute the above phenomenon to the presence of systematic errors
that affect measurements in all four ebulliometers to nearly the same extent.
This implies that variations in performance among the individual
ebulliometers are not the dominant source of systematic errors. Thus, the
slopes that are measured at infinite dilution for a given binary system are
either both greater than or both less than the accepted values, and the
magnitudes of the deviations are similar. Seldom is the case in which the
error of one slope is positive and the other negative, which would of course
skew the predicted azeotrope composition. This is yet another way in which
students distinguish between errors in the measurements and model suitability.
Having analyzed the experimental errors and model validity in this way,
students must then decide if the predicted VLE behavior is acceptable or if a
complete VLE study is warranted. Thus, students learn to weigh the need
for accuracy against limited resources, a lesson that will serve them well in

The culmination of experimental and modeling efforts is the generation of
ternary-phase diagrams. Using the four binary activity coefficient param-
eters they determined experimentally, in addition to two taken from the
literature, students perform ternary bubble-point calculations with an
activity coefficient model of their choice to create plots like those in
Figures 7 and 8. Both are triangular diagrams, in which each apex
denotes a pure component, and any point within the triangle represents
a particular liquid-phase composition.
The series of points in Figure 7 correspond to various isotherms, and the
"bulls-eye" pattern indicates the presence of a ternary azeotrope. In this case
the azeotrope occurs at a composition of 6% acetone, 62% methyl acetate, and
32% methanol (all mole %) and boils at 53.90C. This azeotrope is therefore
minimum boiling, although it is possible to obtain maximum boiling and
saddle-point azeotropes with other ternary systems, or no azeotrope at all.
Identification of azeotropes is a key factor when considering distillation,
and students address the feasibility of separation by plotting contours of
constant K, for each component, where K, (the distribution coefficient) is
Continued on page85


Winter 2000





New Mexico State University Las Cruces, NM 88003

In developing a laboratory course sequence for chemical
engineering undergraduates, it is necessary to define
overall course objectives as well as objectives for indi-
vidual experiments. This would correspond to defining the
overall course objectives and the objectives of each lecture
for a traditional lecture-based course. In the past four years
in this journal alone, over twenty articles"' 23 have appeared
describing new and innovative individual experiments. But
objectives of the course as a whole and how they are to be
defined have received less attention.124'29]
The new ABET EC 2000130] explicitly requires that engi-
neering departments develop in their students "the ability to
design and conduct experiments as well as analyze and inter-
pret data." Additionally, these same students must be able to
"function on multidisciplinary teams," and "communicate
effectively." It is incumbent on the department to document
that the students have these abilities. A logical place to
explicitly incorporate the development of these skills into an
undergraduate curriculum is within the laboratory sequence.
Here, we can not only develop the statistical experimenta-
tion and communication skills, but we can also document the
progress of students in these critical areas. In addition, we
can use a continuous feedback loop to revise and improve
the experiments as we receive input from our alumni, advi-
sory boards, and recruiters concerning the effectiveness and
suitability of the courses for the employability of our students.
With consensus from our department's Industrial Advi-
sory Board, we undertook a comprehensive review of our
entire laboratory sequence almost two years ago. This re-
view identified that our students needed to improve their
understanding of the abstract concepts of experimental de-
sign and data analysis and be given more opportunities to
practice these skills in the laboratory. Therefore, we devel-
oped a four-course sequence: one lecture course (which was
new to the curriculum) and three laboratory courses (which
were in the curriculum but were extensively modified) of

increasing complexity, that integrated experimentation with
statistical concepts and engineering science and design. These
courses are summarized below:
Chemical Engineering Data Analysis A 3-credit, second-
semester Sophomore course covering the theoretical aspects
of experimental design and data analysis.
Process Instrumentation Laboratory A 2-credit, first-
semester Junior laboratory introducing the students to
measurement techniques, statistical analysis of engineering
data, report writing, and oral presentations in small teams.
Transport Operations Laboratory A 2-credit, second-
semester Junior laboratory in thermodynamics and heat,
mass, and momentum transport where teams of students
measure transport coefficients using statistically designed
experiments and report their results both in writing and
Unit Operations Laboratory A 2-credit, first-semester
Senior laboratory where small teams of students characterize
the performance of several unit operations and use their
results in solving design problems. Written and oral reports
are required.
In this paper, the Process Instrumentation Laboratory, which
was completely redesigned with new experiments, data analy-
sis, and reporting requirements, is described in detail. By
carefully selecting and designing the experiments and the
organization of the course, it was possible to have the stu-

Stuart Munson-McGee, Professor of Chemi-
cal Engineering at New Mexico State Univer-
sity, received his BS in Chemical Engineering
from the University of Washington and his PhD
from the University of Delaware. His research
interests include advanced materials process-
ing and separation sciences.
Copyright ChE Division of ASEE 2000
Chemical Engineering Education

dents meet several course objectives, including
Conducting engineering experiments using varied experi-
mental designs
Analyzing experimental data using several statistical
Using different measurement methods
Exposing the students to a variety of engineering phenomena
Developing the student's written and oral presentation skills

The course met for 3 hours twice each week for 16 weeks.
The first four weeks were spent in 1-hour lectures reviewing
statistical design of experiments and data analysis. Also

* Instant feedback was provided for the oral reports by both the
class and the instructor. The presenter's group members were
required to identify at least one thing about the presentation
they thought was excellent and one that needed improve-
ment. The instructor and the other students provided
additional comments to the presenter as soon as the presenta-
tion was finished. This allowed all students to hear positive
comments as well as areas for improvement on 9-12
presentations in a single afternoon.
* During the poster sessions, faculty and visiting industrial
scientists and engineers were invited to review the posters
and quiz the presenters about their work. Feedback was given
immediately to the students concerning their presentation as
well as their poster design.

included in this introductory section were lec-
tures on laboratory safety, right-to-know train-
ing, laboratory notebook keeping, report prepa-
ration, and oral presentations.
The final twelve weeks covered the actual ex-
perimentation, analysis, reporting, and presenta-
tion phase of the course. This phase was divided
into three blocks of four weeks. For each block,
the students were divided into teams of 3-4 stu-
dents, and each team conducted three experi-
ments. At the conclusion of the first two experi-
ments, the students submitted individual memo-
randum reports (a 2-3 page report suitable for
submission to a technical manager, plus 5-10
pages of attachments documenting the proce-
dure and data analysis and answering questions
specific to the experiment). One member of each
team also gave a five-minute oral presentation.
At the conclusion of the third experiment in
each block, the team submitted a formal report
and the students who had not done an oral report

In this paper,
the Process
which was
with new

is descri
in deta

did individual poster presentations of their results. New teams
were formed at the beginning of each block with the same
procedure for experimentation and reporting. Thus, each
student submitted six memorandum and three formal re-
ports and conducted two oral and one poster presentation
during the semester.
Several additional aspects of the course organization are
worth mentioning:

Two days prior to each experiment (typically after the oral
reports), the students were given an hour in the laboratory to
review the experimental set-up.
Each experiment had to be conducted in the allotted time (3
hours); any group not finished within that time received a
zero grade for the experiment.
Reports were graded for both technical content and for
composition, grammar, readability, and conciseness. The
faculty instructor was responsible for the technical content
while a professional technical writer evaluated writing-
related issues.

The students enrolled in this course typically
had completed
* Engineering Data Analysis and Experimental
* Mass and Energy Balances
* Transport Operations I: Fluid Flow
* Chemical Engineering Thermodynamics I:
Engineering Thermodynamics
* Differential Equations
* Freshman Chemistry
* Freshman Composition

and In addition, the students had completed, as part
ng of Freshman chemistry courses, the equivalent of
ents, a 2-credit general chemistry laboratory course, so
they had not yet covered the engineering science
ibed background for many of the experiments. Thus, in
[l. the descriptions of the experiments and the data
analysis, it was necessary to either provide the
missing information (i.e., the theoretical descrip-
tion) or to provide the appropriate references.

The experiments for the laboratory were selected and de-
signed with the following objectives:
Each experiment had to be completed in the allotted time (3
Each experiment had to produce a sufficient number of data
points (depending on the design, this required between 8 and
30 points per experiment) to allow statistical analysis to
justify conclusions drawn by the students.
Each experimental design (which included fractional
factorials, Graeco-Latin squares, blocking, nested, and
mixtures designs) should be used at least twice during the
The experimental conditions had to be easily changed so no
two groups performed exactly the same experiment.
Different fields of engineering science were to be explored.

Winter 2000

Some of the experiments had to explore topics that had not
been covered extensively by their prior classroom experience
as preparation for lifelong learning.
For some of the experiments, the students were required to
rely only on the statistical analysis of the data to develop their
conclusions because of a lack of an engineering science de-
scription of the phenomena. But when a suitable engineering
science description was available, the students were required
to statistically validate the mathematical expressions using
their data.
When applicable, students were required to use ASTM stan-

In addition, there were the following constraints:

Minimal use of hazardous or dangerous chemicals.
Minimal cost of the individual experiments and, where pos-
sible, use of existing facilities and instrumentation.

A short description of each of the nine experiments can be

found in Table 1. The balance of design and topics provided
good coverage of the topics and designs within the con-
straints of the experiments (see Table 2).


The most common reaction from students was that the
most difficult portion of the course was analysis of their
data. At the beginning of the semester, most of the students
viewed data analysis as a cookbook task that could be done
with little thought. Throughout the semester, the students
repeatedly asked what the answer should be and how they
should get it. Only toward the end of the semester did most
students begin to realize that data analysis was a process of
discovery and that their data would have to lead them to the
answer. Of course, the consensus was that this was much
more difficult and time consuming than they had planned
and that waiting until the last moment to conduct the analy-
sis ensured that they would not finish in time.

Descriptions of Experiments

Specific gravity of aqueous solutions The specific gravity of
mixtures of water, salt, and sugar were measured using a hydrometer.
Since the maximum solubility of both solids was about 5% by weight,
the simplex-centroid mixtures design was constrained to
0.95 solution. The data anlaysis required the students to develop a
statistically significant polynomial expression for the specific gravity
and plot contours of constant specific gravity on triangular graph
paper. By changing the solutes, the experimental factor space can be
altered, which changes the data analysis.
Heat transfer from fins The effects of four factors on the convec-
tive heat transfer coefficient from fins were determined by measuring
the end-face temperature on 16 different fins of various geometry and
materials, as dictated by a 4x4 Graeco-Latin square design. A
nonlinear least-squares analysis allowed the students to determine the
best-fit convective heat transfer coefficients for the top and side of the
fins. The end temperatures calculated with these coefficients were
compared to those measured to determine if any of the factors
affected the difference between the measured and calculated
temperatures, i.e, the students were required to statistically validate
the underlying engineering science. To alter this experiment, we have
changed the bath temperature and could use a fan to change the
convective heat transfer coefficient.
Efflux time from a baffled tank Various baffle configurations were
added to a gravity-drained tank to study their effect on drain time.
Length and diameter of the exit pipe were also varied as dictated by a
3x3 Graeco-Latin square design. A simple ANOVA was used to
determine the factors that significantly affected the efflux time. By
changing the variable assignment in the design, a different experiment
Absorption by activated carbon Blue food coloring was absorbed
from aqueous solutions of various strengths by a commercial
activated carbon. Factors examined in the 25'' fraction factorial design
included amount of solution, concentration of food coloring in
solution, the contact time, the ratio of carbon to solution, and the
mixing speed. ANOVA was used to determine the significant factors.
Many factors can be changed in this experiment, e.g., type of carbon
or colorant, temperature of the solution, etc., to create different

Acid neutralization A three-component, constrained simplex-centroid
mixtures design was used to select the compositions for ten solutions of
vinegar and two commercial antacids. Solution pH was measured using
a digital pH meter. ANOVA and linear least squares to determine a
statistically significant polynomial fit of the data were used, and then
contours of constant pH were plotted on triangular graph paper. By
changing the brand of antacids, this experiment can be changed.
Frictional losses in pipes The Fanning friction factor was calculated
for laminar and turbulent flow in PVC and copper pipes of various
diameters based on pressure drop measured using an inclined
manometer. Due to time considerations, a balanced incomplete
blocking design was used to select the factor space combinations to be
tested. Linear regression allowed the students to determine if the
Hagen-Poiseuille law was valid.
Rotameter calibration A blocking design, using the operator as the
blocking factor and rotameter reading as the independent factor, was
used to determine the experimental space to create a calibration curve
for a salt-water solution in a rotameter. ANOVA was used to identify
the significant factors and linear regression was used to develop a
calibration curve and a 95% confidence interval for the predicted
values. This experiment was changed by altering the density of the fluid
used in the rotameter.
Efficiency of a parallel-plate exchanger A 24' fractional factorial
was used to evaluate the efficiency of a simple parallel-plate heat
exchanger (custom designed and manufactured for this course) using
the inlet temperatures and flow rates as the independent factors. The
students had to calculate the overall resistance to energy transfer for
both the cold and hot sides, determine if they were affected by any of
the factors, and decide whether or not the two coefficients were
statistically different. By changing the number of plates in the
exchanger and the thickness of the plates, the experiment could be
Viscosity of aqueous solutions The effect of a proprietary food
thickener on the apparent viscosity of aqueous solutions as a function of
shear rate and thickener concentration was measured using a rotating
spindle viscometer. A two-level nested design was used to determine
the factor space combinations to be tested, and ANOVA was used to
analyze the data. Changing the concentration and type of thickener
changed the experiment from group to group.

2 Chemical Engineering Education

The second most common reaction was that the experi-
ments were relatively simple to conduct and that they could
easily be accomplished in the allotted time. Having both the
in-lab preview and oral presentations by other students greatly
facilitated this efficiency. But there were some problems
with completing the experimental design (i.e., completely
specifying all the trials, including replicates, that would
be done) prior to beginning the experiments. In several
cases this meant that the students failed to conduct a
sufficient number of experiments to conduct a satisfac-
tory analysis of their data.
The students also appreciated the fact that the experiments
were always ready to run. Thanks to the help of an outstand-
ing teaching assistant and staff engineer, the experiments
were turned on and warmed up before the students arrived in
the lab; the students did not have to wait for water baths to
heat or for instrumentation to warm up before they were
ready to begin. The teaching assistant and staff engineer
were available to answer questions during the lab and to help
solve equipment problems that arose (which happened about
once every other week). The students truly appreciated the
willingness to help and approachability of both individuals.
Little comment was made by students regarding the use of
a technical editor to assist in grading the written reports. The
editor commented on the marked improvement of the writ-
ing as the semester progressed, however. Having to write six
memorandum reports and three formal reports gave the stu-
dents ample opportunity to improve-the average writing
grade increased by nearly 5 points (out of a possible 20) over
the course of the semester. The students also made little com-
ment about the oral and poster presentations. Again, grades

Summary of Experiments, Experimental Designs, and Engineering
Topics Covered in the First Laboratory Course

Engineering Topic
Heat Mass Momentum Chemical Physical
Transfer Transfer Transfer Reaction Properties

Efflux Time
From a
Baffled Tank

Fractional Parallel Plate Adsorption
Factorial Heat by Activated
Exchanger Carbon
Constrained pH of Specific
Mixtures Aqueous Gravity
Blocking Frictional
Losses in Pipes
Nested Viscosity
of Aqueous

Winter 2000


Heat Transfer
From Fins

significantly improved during the semester-the average grade
on the initial oral reports was ten points (on a 50-point scale)
lower than the average grade on the final oral reports.
The most frustrating aspect of the course for many stu-
dents was the different backgrounds in statistics of the stu-
dents. In addition to the statistics course offered in the de-
partment, other courses were accepted as satisfying the course
prerequisite. Most of the other courses did not have the same
emphasis on data analysis as the departmental course and
instead focused on probability and combinatorial theory.
Students who had taken the departmental course often found
themselves teaching the other students how to conduct the
data analysis and interpret their data. Although this was
probably a great learning experience for the students, they
resented the time it required for what to them was no return.

In general, the laboratory worked extremely well consid-
ering it was the first time the course was offered in this
manner. From the instructional side, the following lessons
were learned (or, in some instances, released):

* To compliment the experiments, the initial phase of the
course needed to focus more on how to develop a design
so that the proper factor space combinations and
replicates would be tested. In the lab manual, the
experimental design was specified, but the details were
left for the student to determine, which they did not
always complete prior to the experiments.
* To improve the written and oral communication skills,
more time needed to be devoted to reviewing the
structure and organization of technical communica-
tion during the initial phase of the course. In conjunc-
tion with this, reviewing document design aspects
would also be warranted.
To assist students who had taken a non-departmen-
tal experimental statistics course, grouping them
together and reviewing the design and analysis
techniques weekly assisted in reducing both the
intra- and inter-group variability.

* To provide sufficient time for data analysis, the
laboratories should be conducted on Thursdays,
with the reports due on the following Tuesday.
Initially, the labs were done on Tuesdays, with
reports due Thursday-leaving insufficient time to
conduct the analysis.
* To enhance the quality of the formal reports,
students need to cover a broader scope of material
than the experiments for the memorandum reports.
For example, the mixtures experiments could
involve a fourth component or the evaluation of
the heat exchanger could include the effects of the

number of plates.
To ensure that the students can complete the experi-
ments in time, a hands-on teaching assistant is abso-
lutely necessary.
To reduce student frustration at having to work for an
extended period of time with an under-achieving lab
partner, groups need to be reformed randomly and
frequently. Having each student work in three groups
over the semester seemed to avoid intra-group prob-

In addition to addressing the lessons above, the following
recommendations are also suggested:

To cover more chemical engineering science (in
particular, chemical reactions and kinetics), a greater
breadth of experiments is needed.
To further improve the writing skills, it would have
helped if report writing would have included revising
some of the reports until all structure, organization, and
grammatical problems were corrected.

Without the energy and expertise of Jim Anthony and
James Autry, this laboratory would not have been as suc-
cessful as it was. Jim Anthony, the departmental engineer,
was responsible for building and assembling the experi-
ments. James Autry, my teaching assistant, was responsible
for making sure that the experiments were operating every
week and answering questions about experimental proce-
dures and laboratory safety. I am also thankful for the com-
ments made by one of the reviewers about the value of
exploring topics not yet covered in courses.

1. Lennox, Barry, and Michael Brisk, "Network Process Con-
trol Laboratory," Chem. Eng. Ed., 32(4), 314 (1998)
2. Conlee, Thomas D., Helen C. Hollein, Charles H. Gooding,
and Stewart C. Slater, "Ultrafiltration of Dairy Products as
a ChE Laboratory Experiment," Chem. Eng. Ed., 32(4), 318
3. Delgado, P., A. Kasko, J. Nappi, and R. Barat, "An Experi-
ment in Applied Optics: Determination of the Kinetics of
the Oxidation of an Organic Dye," Chem. Eng. Ed., 32(3),
174 (1998)
4. Hellgardt, K., and G. Shama, "A Simple Method for Deter-
mining the Specific Heat of Solids," Chem. Eng. Ed., 32(3),
190 (1998)
5. Baird, M.H.I., and I. Nirdosh, "Low-Cost Experiments in
Mass Transfer: Part 4. Measuring Axial Dispersion in a
Bubble Column," Chem. Eng. Ed., 32(3), 198 (1998)
6. Luke, June, and N. Lawrence Ricker, "Unit Operations Lab:
Mass Transfer and Axial Dispersion in a Reciprocating-
Plate Liquid Extraction Column," Chem. Eng. Ed., 32(3),
202 (1998)
7. Nirdosh, I., L.J. Garred, and M.H.I. Baird, "Low-Cost Ex-
periments in Mass Transfer: Part 3. Mass Transfer in a
Bubble Column," Chem. Eng. Ed., 32(2), 138 (1998)

8. Rodriguez, J.M., F. Henriquez, and A. Macias-Machin, "A
Simple Experiment for Mass Transfer," Chem. Eng. Ed.,
32(2), 142 (1998)
9. Powers, Susan E., and Stefan J. Grimberg, "Experiments
Illustrating Phase Partitioning and Transport of Environ-
mental Contaminants," Chem. Eng. Ed., 32(1), 40 (1998)
10. Farooq, Shamsuzzaman, "An Undergraduate Experiment
on Adsorption," Chem. Eng. Ed., 32(1), 76 (1998)
11. Fordon, Keith, B., Antonio M. Vincitore, and Selim M.
Senkan, "An Experiment in Combustion," Chem. Eng. Ed.,
31(4), 236 (1997)
12. Lauterbach, J., S. White, Z. Liu, G.M. Bodner, and W.N.
Delgass, "A Novel Laboratory Course on Advanced ChE
Experiments," Chem. Eng. Ed., 31(4), 260 (1997)
13. Gerrard, Mark, Mark Hockborn, and Jason Glass, "An Ex-
periment to Characterize a Consolidating Packed Bed,"
Chem. Eng. Ed., 31(3), 192 (1997)
14. Priore, Brian, Shawn Whitacre, and Kevin Myers, "Being
Dynamic in the Unit Operations Laboratory: A Transient
Fluidized-Bed Heat Transfer Experiment," Chem. Eng. Ed.,
31(2), 120 (1997)
15. Kresta, Suzanne M., Andre Koenig, and Murray R. Gray,
"Choosing an Optimum Feedstock for Yeast Production: A
Design-Oriented Senior Laboratory Experiment," Chem. Eng.
Ed., 31(1), 22 (1997)
16. Anklam, Mark R., Robert K. Prud'homme, and Bruce A.
Finlayson, "Ion Exchange Chromatography Laboratory: Ex-
perimentation and Numerical Modeling, Chem. Eng. Ed.,
31(1), 26 (1997)
17. Campbell, Scott W., and Venkat R. Bhethanabotla, "Rapid
Determination of Vapor-Liquid Equilibria: An Undergradu-
ate Laboratory Exercise," Chem. Eng. Ed., 31(1), 34 (1997)
18. Palanki, Srinivas, and Vishak Sampath, "A Simple Process
Dynamics Experiment," Chem. Eng. Ed., 31(1), 64 (1997)
19. Robinson, Ken K., and Joshua S. Dranoff, "A Laboratory
Experiment that Enhances Environmental Awareness,"
Chem. Eng. Ed., 30(2), 98 (1996)
20. Nirdosh, I., and M.H.I. Baird, "Low-Cost Experiments in
Mass Transfer: Part 2," Chem. Eng. Ed., 30(2), 142 (1996)
21. Nirdosh, I., and M.H.I. Baird, "Low-Cost Experiments in
Mass Transfer: Part 1," Chem. Eng. Ed., 30(1), 50 (1996)
22. Ignacio-Zubizaretta, J., and Gabriel Pinto, "An Ancient
Method for Cooling Water Explained by Mass and Heat
Transfer," Chem. Eng. Ed., 29(2), 96 (1995)
23. Pendse, Ajit V., and John R. Collier, "Polymer Processing:
For the Undergraduate Unit Operations Laboratory," Chem.
Eng. Ed., 29(2), 120 (1995)
24. Abu-Khalaf, Aziz M., "Getting the Most Out of a Laboratory
Course," Chem. Eng. Ed., 32(3), 184 (1998)
25. Miller, Ronald L., James F. Ely, Robert M. Baldwin, and
Barbara M. Olds, "Higher-Order Thinking in the Unit Op-
erations Laboratory," Chem. Eng. Ed., 32(2), 146 (1998)
26. Newell, James A., Douglas K. Ludlow, and Steven P.K.
Sternberg, "Development of Oral and Written Communica-
tion Skills Across an Integrated Laboratory Sequence,"
Chem. Eng. Ed., 31(2), 116 (1997)
27. Stubington, John F., "Quality in Teaching Laboratories,"
Chem. Eng. Ed., 29(3), 186 (1995)
28. Middleberg, A.P.J., "Laboratory Projects: Should Students
Do Them or Design Them?" Chem. Eng. Ed., 29(1), 34 (1995)
29. Davies, W.A., and T.A.G. Langrish, "Putting Commercial
Relevance Into the Unit Operations Laboratory," Chem.
Eng. Ed., 29(1), 40 (1995)
30. "Engineering Criteria 2000," Accreditation Board for Engi-
neering and Technology, Inc., 111 Market Place, Suite 1050,
Baltimore, MD (1998) 0

Chemical Engineering Education

Vapor-Liquid Equilibria
Continued from page 79.
defined by K,=y,/x,. Of special interest is a plot of contours
for K,=l, as shown in Figure 8. Since points at which K,=l
indicate that the mole fraction of a particular component is
identical in both phases, any point at which the three indi-
vidual K,=1 contours intersect defines a ternary azeotrope.
Thus, Figure 8 contains a ternary azeotrope at a composition
of 7% acetone, 62% methyl acetate, and 31% methanol, in
good agreement with Figure 7 and indicating self-consis-
tency in the students' calculations. Certainly, the ternary plot
should also predict any binary azeotropes. For example, the
intersection of K,=I contours for acetone and methanol,
along the edge in which the methyl acetate mole fraction is
zero, reveals an acetone-methanol binary azeotrope at a
methanol mole fraction of 0.2. This result is in excellent
agreement with that of Figures 5 and 6.

A technical detail to consider when planning the experi-
ment outlined in this paper is the requirement for constant
pressure. Whereas the data reduction (specifically, Eq. 4)
assumes constant pressure, the actual system pressure is
subject to atmospheric changes. Occasionally, students may
face the challenge of performing the experiment as a storm
front approaches and the atmospheric pressure changes ap-
preciably. It is therefore imperative that students be able to
handle pressure fluctuations, either during data reduction by
modification of Eq. (4) to include a pressure dependence or
(preferably) by correcting the measured boiling points for
changes in pressure during the lab session. Even if the cor-
rection turns out to be negligible, good engineering practice
requires that this be tested rather than assumed. We therefore
view systematic pressure variation as a fortunate event be-
cause it affords students the opportunity to think through the
aberration and account for the effect.
Finally, a remaining consideration that may be of critical
concern is one of cost. Price quotes for the main components
in our system are $800 (Swietoslawski ebulliometer), $1900
(hot water circulator), $650 (digital temperature display),
$120 (pump), $125 (HDPE tank), and $130 (platinum
RTD).13] The capital cost for a simple VLE experiment
involving a single ebulliometer is therefore approximately
$3700 (plus associated piping), although time limitations
would likely limit such an experiment to a single binary
system. Reproducing the ternary VLE experiment we have
described requires a capital expenditure on the order of $15,000.

Our undergraduate VLE experiment has evolved over sev-
eral decades and has been in its current form the past five
years. We believe that it is unique because it allows genera-
tion of ternary phase behavior from a single afternoon of
Winter 2000

data collection.1141 The lab is therefore educational in two
very important general aspects. One is the technical training
that the lab provides, since students demonstrate a greatly
improved understanding of phase behavior and sharpen their
modeling skills as a direct result of the laboratory work.
Another is the practical lesson students learn; that time and
money are important considerations when planning any ex-
periment, and reasonably accurate data can often be ob-
tained without elaborate measurement techniques.
We invite you to visit our VLE website at

The authors wish to thank Professor Jon Olson and Dr.
Larry Dodd for contributions in the early stages of develop-
ment of this experiment.

1. Sandler, S.I., Models for Thermodynamic and Phase Equi-
libria Calculations, Marcel Dekker, Inc., New York, NY
2. Malanowski, S., "Experimental Methods for Vapor-Liquid
Equilibria. Part I. Circulation Methods," Fluid Phase
Equilib., 8, 197 (1982)
3. Campbell, S.W., and V.R. Bhethanabotla, "Rapid Determi-
nation of Vapor-Liquid Equilibria: An Undergraduate Exer-
cise," Chem. Eng. Ed., 31(1), 34 (1997)
4. Barker, J.A., "Determination of Activity Coefficients from
Total Pressure Measurements," Austral. J. Chem., 6, 207
5. Olson, J.D., "Measurement of Vapor-Liquid Equilibria by
Ebulliometry," Fluid Phase Equilib., 52, 209 (1989)
6. Hala, E., Vapor-Liquid Equilibrium, 2nd ed., Pergamon
Press, Oxford (1967)
7. Smith, J.M., H.C. Van Ness, and M.M. Abbott, Introduction
to Chemical Engineering Thermodynamics, 5th ed., McGraw-
Hill Book Company, New York, NY (1996)
8. Sandler, S.I., Chemical and Engineering Thermodynamics,
2nd ed., John Wiley & Sons, New York, NY (1989)
9. Strictly speaking, Eq. (5) is a truncated version of the ter-
nary van Laar model. Eq. (5) neglects all terms of third and
higher order in the volume fractions that appear in the
Wohl expansion for the excess Gibbs free energy.
10. Gmehling, J., and U. Onkne, "Vapor Liquid Equilibrium
Data Collection" Chemistry Data Series, Vol. 1, Part 1,
DECHEMA, Frankfurt/Main (1977)
11. Ethanol is an adequate substitute if safety considerations
preclude the use of methanol.
12. Tester, J.W., and M. Modell, Thermodynamics and Its Ap-
plications, 3rd ed., Prentice-Hall, Inc., Upper Saddle River,
NJ (1997)
13. Prices quoted in Spring, 1999
14. The lab operates on a four-week cycle. During week 1 stu-
dents view the equipment and complete a pre-laboratory
homework assignment. The experiment is performed by five
teams (each with three students) during week 2, where each
team is given one afternoon session (typically 3-4 hours) to
obtain data. Students prepare a preliminary technical re-
port during the third week that addresses only the binary
VLE data and modeling. The ternary system is handled
during week 4, in which students prepare a final technical
report that addresses the feasibility of separating the three
chemicals. O





Carnegie Mellon University Pittsburgh, PA 15213

Over the past few years, chemical engineering de-
partments in several universities around the country
have begun to hold annual graduate students re- s)
search symposia. This tradition began in 1979 at Carnegie th
Mellon when the Chemical Engineering Graduate Students
Association (ChEGSA) organized the first Annual Chemical
Engineering Symposium. Since that time, each symposium
at Carnegie Mellon has been organized by graduate students
for their colleagues and each has been funded entirely from it
industrial sponsorship, with all funds being raised by the
students themselves. Carnegie Mellon's Twentieth An-
nual Symposium was held in October of 1998, and so it
seems fitting to briefly review the significance of the
symposium on this anniversary.

In 1979, Department Head Tomlinson Fort suggested hold- o
ing an annual symposium, to be organized by ChEGSA. He re
felt the symposium would promote better communication
skills among graduate students and provide a forum in which
to exchange research ideas, both within the department and si
with industry. It has been with that objective in mind that the




Keynote speaker, Professor John Perkins,
describes a flowsheet during his talk.

rmposium has continued and enjoyed great success over
e past twenty years.
Industrial participation commenced with the Second An-
ual Symposium and has become increasingly important
very year since then. Because the Symposium is funded
itirely through donations made by industrial participants,
allows graduate students to refresh their contacts with
dustry and to learn more about current industrial needs
d concerns. In addition, it is a useful way for the students
Learn important networking skills and to keep in contact
ith alumni. Furthermore, the symposium gives industrial
participants an opportunity to learn more about current de-
artmental research and provides an excellent means for
em to meet graduate students and to get to know them
itside the artificially constrained atmosphere of a formal
cruiting process.
For many alumni and industrial participants, the sympo-
um provides a first point of contact for those who may be
eir future colleagues. For this reason, a resume book is
)mpiled and distributed to all of the industrial participants.

Each year since 1984, the symposium has included a
eynote address from a researcher and lecturer of interna-
onal standing. (A full list of all the keynote speakers over
e years is given in Table 1.) The purpose of the keynote
Dress is to promote better presentation skills among the
udents through the example set by an outstanding speaker.

Copyright ChE Division of ASEE 2000

Chemical Engineering Education

Timothy D. Power is a PhD student in the
Department of Chemical Engineering at
Carnegie Mellon University. He received his
BE degree in 1997 from University College,
Dublin, Ireland, and began his graduate work
at Carnegie Mellon in the fall of that year. As
the 1998 Vice-President of the ChEGSA,
Timothy was responsible for organizing the
Annual Symposium described in this paper.
He is currently working with Professor David
Sholl in the area of molecular simulations.

In 1998, the keynote speaker was Professor John D. Perkins,
Head of the Department of Chemical Engineering and Chemi-
cal Technology at Imperial College, London. His talk,
"Trends in Process Systems Engineering," was received
with great interest, as it touched both on the history of
process systems engineering and recent trends in design and
control integration.
In recent years, the symposium has been held in a confer-
ence room on campus in mid-October. Since it takes place in
mid-week, classes for all graduate students are cancelled for
those days. Over the two-day period of the symposium, the
PhD students in the department give approximately thirty-
five presentations. While a handful of second-year students
usually participate, most of the presentations are made by
the third-, fourth-, and fifth-year students.

The range of research topics covered
spans all primary areas of specialization
specifically: bioengineering, complex
fluids, environmental engineering,
process systems engineering, and
solid-state materials. Speakers are al-
lotted fifteen-minute time slots for
their presentations.

Since maximal industrial participa-
tion tends to occur on the first day,
priority for time slots on that day is
given to students in the final year of
their studies. A luncheon is also held
on the first day of the symposium and
participating students, industrial at-
tendees, and faculty are all invited,
providing further opportunities for in-
teraction and conversation. In addi-
tion, the poster session held that
evening is accompanied by a wine-
and-cheese reception where there
are further prospects for contact.


The ChEGSA symposium provides
a unique means for students to de-
velop the skills they will need for fu-
ture success. In almost any career, it
is essential to be able to present one's
work to others and to argue the merits
of one's case. It is not easy to deliver
a short presentation to an audience
with diverse interests, and practice is
the best way to become comfortable
and confident with making presenta-
tions. To this end, a panel of aca-
demic professors and industrial attend-

in the symposium
at Carnegie Mellon,

ees adjudicate each student's talk with the intention provid-
ing feedback to the students and improving their communi-
cation skills. The speaker with the highest evaluation re-
ceives the Geoffrey D. Parfitt Memorial Award. This award,
established by ChEGSA, honors the memory of Dr. Parfitt, a
Professor of Chemical Engineering at Carnegie Mellon who
passed away unexpectedly in 1985. There are, in addition,
two awards given to the second- and third-highest ranked
students, as well as two honorable mentions.
Awarding the students for their performance has proven
valuable-it provides a tangible incentive for students to
deliver high-quality presentations. Over the years, many who
have won awards at the symposium have gone on to pursue
very successful careers, i.e., among others, John Walz (1991
and 1992), Yale University; Christodoulos A. Floudas (1983
and 1985), Princeton University; Annette Jacobson (1987),
Carnegie Mellon University; Marco Duran (1984), Exxon

Corporation; James

Cuthrell (1985), Shell; Paul Bowman
(1986 and 1987), Arco Chemical.
The awards are presented the
week following the symposium
at a banquet organized by
ChEGSA. A list of the award win-
ners of the 1998 symposium is
given in Table 2.

The symposium is also crucially
important in assisting new first-year
graduate students in selecting a the-
sis advisor. The symposium takes
place midway through the first se-
mester and (at Carnegie Mellon) the
thesis advisor is usually not selected
until late in this semester. By at-
tending talks given by students of
various faculty, the symposium pro-
vides a valuable means for first-
year students to learn more about
the specifics of the research they
can expect to do if they work with a
certain professor.

The continued success of the sym-
posium can be attributed to the
many benefits derived from hold-
ing such an event. In the first in-
stance, the symposium internally
benefits the Chemical Engineering
Department. Graduate students are
given an opportunity to learn more
about the work in which their peers
are engaged, and the opportunities
for exchanging ideas and other feed-

Winter 2000

Keynote Speakers

1984 Ed Cussler, University of Minnesota
1985 Dan Luss, University of Houston
1986 George Keller, Union Carbide
1987 Alexis Bell, University of California, Berkeley
1988 Eduardo Glandt, University of Pennsylvania
1989 Robert Anderson, Monirex Systems, UOP Inc.
1990 Michael Shuler, Cornell University
1991 Michael Doherty, University of Massachusetts
1992 John O'Connell, University of Virginia
1993 Elizabeth Dussan, Schlumberger Doll
1994 Joe Pekny, Purdue Universitv
1995 Doug Lauffenberger, Massachusetts Inst. of Tech.
1996 Mark Barteau, University of Delaware
1997 Alice Gast, Stanford University
1998 John D. Perkins, Inperial College, London

1998 ChE Symposium Award Winners

Geoffrey D. Parfitt Award (Overall)
EI Scott A. Guelcher: Advisor, John L. Anderson
Investigating the Mechanism ofAggregation of
Colloidal Particles During Electrophoretic Deposition
Symposium Awards
[1 Celia N. Cruz; Advisor, Spyros N. Pandis
The Effect of Organic Coatings on the Cloud
Condensation Nuclei Activity of Inorganic Aerosol
E[ Stephen J. Vinay, III: Advisor, Myung S. Jhon
A Study ofMulti-Particle Dynamics in
Triboelectrostatic Systems
Honorable Mentions
E Hector Yeomans: Advisor, Ignacio E. Grossmann
A Disjunctive Programming Method for the Synthesis
of Heat Integrated Distillation Sequences
El Timothy D. Power: Advisor, David S. Sholl
Theoretical Studies of the Adsorption of Chiral
Molecules onto Chiral Metal Surfaces


back are substantial. In addition, the department greatly
benefits from the opportunity to refresh contacts with
Since it is entirely the responsibility of students, actually
organizing the symposium is a valuable experience in and of
itself. Its organization is generally the responsibility of just
one student, with assistance from fellow ChEGSA officers.
There is, of course, a considerable time investment required
from the individual concerned. In addition to the logistics of
accepting abstracts, allotting time slots for speakers, orga-
nizing flights for the keynote speaker, etc., there is also a
considerable fund-raising element involved. As a conse-
quence, competence in several areas is needed to success-
fully coordinate the event, including communication and
negotiation abilities, delegating skills, fund raising, and re-
source allocation. Time-management skills are crucial, since
the event needs to be planned while the organizer continues
to pursue research, attend classes, and attends to teaching-
assistant duties.
Typically, about $8,500 is required just to cover the basic
costs of the symposium. Apart from the obvious costs such
as the luncheon and travel expenses and honorarium for the
keynote speaker, there are additional expenses that include
the cost of coffee and refreshments, postage, audio-visual
equipment rental, etc. All of the funding to cover these costs
is derived from the donations of industrial sponsors (who
donate $500 or more) and contributors (who donate $100-
That the symposium has been a truly valuable event at
Carnegie Mellon is without question. As long as it continues
to serve its purpose, it requires and deserves continued strong
support from all who participate, including students, faculty,
and particularly industrial sponsors and contributors, whose
exceptional generosity has been more than appreciated
through the years.

Many thanks to my fellow ChEGSA officers for their
help in organizing the symposium in 1998. Also, thanks
are due to Professor David Sholl, Professor Ignacio
Grossmann, and Amanda Utts for their help in writing
this paper.
Thanks also must go to the 1998 industrial sponsors:
Air Products and Chemicals, Inc., ALCOA, Amoco
Chemical Corporation, ARCO Chemical, Aspen Tech-
nology, Inc., Bayer, BOC, Dow Chemical, Dow
AgroSciences, Dupont, The Goodyear Tire & Rubber
Company, Lubrizol, Merck & Company, Mitsubishi
Chemical America, Monsanto Company, PPG Industries,
and Simulation Sciences Inc. Industrial contributors for
1998 were Coca-Cola Company, International Paper,
Johnson & Johnson, McKinsey, Mobil, Schlumberger,
Sony Chemical, and Westinghouse. 0

e letter to the editor

To the Editor:
I have just looked through the Fall, 1999, issue of Chemical
Engineering Education-the well-known graduate educa-
tion issue. I noticed a number of advertisements in the gradu-
ate education section that have photographs of people in
laboratories who do not have proper personal protective
equipment. In particular, they lack proper safety glasses.
I can assure you that our industrial friends will notice this
problem. It is also contrary to a number of articles that have
appeared in CEE discussing proper safety culture in labora-
Several years ago I received an award from the Chemical
Manufacturers' Association. The CMA requested photo-
graphs with me and my students in the laboratory. The cover
letter stated that photos without proper personal protective
equipment would not be accepted. I would like to suggest
that CEE do the same.

Dan Crowl
Michigan Tech

Editor's Note: We agree with the comments and encourage
each advertising university to take note of this breach of
laboratory safety procedures when reviewing their adver-
tisements next year.

[, books received

Tailored Polymeric Materials for Controlled Delivery Systems, edited by
lain McCulloch and Shalaby W. Shalaby; Oxford Uiversity Press, 198
Madison Avenue, New York NY 10016; 322 pages, $15 (1998)
Oxford Dictionary ofBiochemistry and Molecular Biology, Oxford Uiversity
Press, 198 Madison Avenue, New York NY 10016; 739 pages, $60 (1997)
Design of Devices and Systems, 3rd edition, by William H. Middendorf and
Richard H. Engelmann; Marcel Dekker, Inc. 270 Madison Ave., New
York, NY 10016-0602; 584 pages, $69.75 (1998)
New Methods in Computational Quantum Mechanics, edited by I. Prigogine
and Stuart A. Rice; Wiley, 605 Third Avenue, New York, NY 10158; 813
pages, $54.95 (1997)
Organotin Chemistry, by Alwyn G. Davies; Wiley, 605 Third Avenue,
New York, NY 10158; 327 pages, $180 (1997)
Hydrocarbon Resins, by R. Mildenberg, M. Zander, and G. Collin; Wiley,
605 Third Avenue, New York, NY 10158; 180 pages, $140 (1997)
Solvent-Free Polymerizations and Processes: Minimization of Conven-
tional Organic Solvents, edited by Timothy E. Long and Michael O. Hunt;
Oxford University Press, 198 Madison Ave., New York, NY 10016; 292
pages, $110 (1999)
Fluid Dynamics and Transport of Droplets and Sprays, by William A.
Sirignano; Cambridge University Press, 40 West 20th St., New York, NY
10011-4211; $80 (1999)
Chemical Engineering Education

book review

Engineering Flow and Heat Exchange,
Revised Edition
by Octave Levenspiel
Plenum Press, New York and London (1998)

Reviewed by
Gabriel I. Tardos

This is the first revised edition of this book, first published
in 1984. Professor Levenspiel should be commended for
producing such an excellent text, written specifically for
engineering students. The book is a pleasure to read and
offers several amusing problems, all stated in the language
of students, with explanations and examples they can easily
understand. Very few texts in engineering can make such a
claim. I have used this text exclusively since 1992 in my
teaching of unit operations to chemical engineering students.
The material is broad enough, however, to also be used in
mechanical engineering, and perhaps in civil engineering
courses as well, to teach flow and heat transfer.
Students (especially undergraduates) tend to sell used text-
books once they finish a subject and pass their final exami-
nation. I found, with great pleasure, that Engineering Flow
and Heat Exchange was not one of those books; seniors use
it in their design courses and many graduates keep the book
as a reference. This is obviously due to the wealth of
information in the book and the ease with which the infor-
mation can be retrieved and used. Inclusion of compressible
and non-Newtonian fluid flow in the fluid-mechanics sec-
tion and direct-contact heat exchangers in the heat-exchang-
ers section is a substantial achievement and significantly
adds to the usefulness of the text.
One example of the book's unique approach to explaining
a complex concept through humor and straightforward, easy-
to-understand language is illustrated by how Professor
Levenspiel explains the concept of equivalent average slurry
density in the problem "Counting Canaries Italian Style."
The "slurry" consists of canaries flying in the air inside a
closed container. Measuring the pressure before and after
the canaries are airborne, and using the Bernoulli equation,
gives the change in density and therefore the number of
"particles" (birds). Ingenious!
As already mentioned, the book is divided into a section
on fluid mechanics and a section on heat transfer. The first
part includes basic equations for isothermal flowing systems
in Chapter 1, and as an example, flow of incompressible
Newtonian fluids in pipes and around solid immersed ob-
jects in Chapters 2 and 8, respectively. Unlike other similar
texts, the theory is kept short and the assumption is that the
Winter 2000

student has taken a prior course in fluid mechanics. It is
assumed, for example, that the student is familiar with the
concept of the Fanning friction factor.
Chapters 3 and 4 address compressible flow of gases
(through material taken mostly from thermodynamics) and
low pressure, "molecular" flows. Here the concept of "mo-
lecular slip" is introduced.
Chapter 5 contains, as mentioned above, concepts and
problems of non-Newtonian flow explained in a direct and
simple-to-understand fashion. The student is reminded that,
in general, this complex fluid can be treated as Newtonian
with an additional term and all that is required is to find the
correction due to the non-Newtonian behavior. Since most
fluids in industrial practice are non-Newtonian, the intro-
duction of this material is, I think, crucial. Furthermore,
rheometry to measure non-Newtonian behavior is also pre-
sented in detail.
Part one of the book also contains chapters of flow in
porous media and in fluidized beds. They are also well
written, with many examples and actual industrial applica-
tions both solved and presented as homework problems.
The second part of the book, on heat transfer and heat
exchanger design, is also enlightening, crisp, and well con-
structed. Chapters 9, 10, 12, and 13 contain the usual mate-
rial on different forms of heat transfer, combined heat trans-
fer, and two-fluid heat exchanger design. Here again, it is
assumed that the student has taken a previous introductory
course in heat transfer since familiarity with, for example,
the Nuselt number is required. The material in Chapters 11,
14, and 15 contains unsteady heating and cooling and design
of direct-contact exchangers and regenerators-material usu-
ally not covered in standard texts. The second part ends
(Chapter 16) with a set of recommended problems involving
material contained in the book, keeping in mind practical,
industrially relevant applications.
There is an extended Appendix with very useful informa-
tion such as transformation of units, some material proper-
ties, dimensionless groups, and values of more important
parameters such as heat transfer coefficients in different
geometries. The text also comes (available to the instructor)
with a set of solutions to the problems in each chapter, with
every second problem being solved. The problems in the last
chapter (16) all have solutions. The illustrations in the book
are inspired and clear, while the nomograms, mostly for heat
transfer calculations, are up-to-date and easy to use.
Over all, this is an excellent book, written with the heart.
The reader can visibly appreciate this. It should be a perma-
nent fixture on the bookshelf of any engineer who studied or
uses fluid flow and heat transfer in his work. O

5 1 lIaboratory


University of Porto 4099 Porto Codex, Portugal

Fickian mass transport is deeply rooted in the culture
of many engineering institutions, universities, and
companies. The mathematical equation that describes
Fick's law is simple and intuitive, but it is only valid for
binary mixtures or for diffusion of diluted species in a
multicomponent mixture, in the absence of electrostatic or
centrifugal force fields.'
The Maxwell-Stefan equation provides a better and a
more general approach. To show its relevance, while keep-
ing the mathematical treatment simple, we propose a ter-
nary mass diffusion transport experiment and its simulation.
The simulator was developed to solve the Maxwell-Stefan
equation for multicomponent isobaric and isothermal sys-
tems and is readily available on the web (
The concepts presented in this work are particularly suited
for both undergraduate and graduate chemical engineering
students provided that they are familiar with the first and
second Fick laws.

The Maxwell-Stefan equation for an isothermal and iso-
baric multicomponent system, where only pressure forces
act, is (a simple derivation of this equation is presented in
the appendix)1' 3

Xi d nii xjNi -xiNj
91T dz j c CtiJ

where x, is the i-solute molar fraction, T is the absolute
temperature, 9 is the ideal gas constant, pi is the molar
chemical potential, z is the axial coordinate, ct is the total
molar concentration, Ni is the molar flux of the species i
with respect to a fixed referential, and Z i is the Maxwell-
Stefan i,j diffusivity, with Z ij = ji .-2)
For ideal gases, Eq. (1) becomes

1 dxi xjNi-xiNj x jJ -xiJj
SRT dz Pi j Pij
jii jzi

where P, is the total pressure and Ji is the molar flux of the
species i with respect to the mixture molar average velocity
Ji =N-xi, Nk
For ternary systems, the Maxwell-Stefan equations are

dz ct AB CtAC
dz ct AB CT2 BC
dXA +dB +dx (5'
dz dz dz

Equation (2) is nonlinear and therefore when solving diffu-
sion problems it is more practical to revert it to the Fickian

(J)= -ct [D] x) (6
where (J) and (x) are (n-l)-component column vectors of
diffusion fluxes and molar fractions, respectively, and
[D] is the (n-l)x(n-l) Maxwell-Stefan diffusion coeffi-
cient matrix
[D]=[B]-' (7:
and [B] is the diffusion coefficient inverse matrix that is
obtained from Eq. (2)

Pedro Taveira is a Ph.D. student in Chemical Engineering at the Univer-
sity of Porto, Portugal. He received his degree in Chemical Engineering
from the same University in 1997. His research interests are in gas
separation using membrane technology.
Paulo Cruz is a first-year Ph.D. student in Chemical Engineering at the
University of Porto, Portugal, where he received his degree in Chemical
Engineering in 1998. His research interests are multicomponent mass
transport and sorption in porous solids and membranes.
Adelio Mendes received his degree and Ph.D. in Chemical Engineering
from the University of Porto, Portugal, where he is currently Assistant
Professor. He teaches Chemical Engineering Laboratories, Separation
Processes, and Numerical Methods. His main research interests include
membrane and sorption gas separations.
Copyright ChE Division of ASEE 2000
Chemical Engineering Education

n ( 1 1
Bij = xi ik Bj(i. ) = xi i, j= 1,2,3,-..n-1
in k=1 1k ^ij ij Z in

The Maxwell-Stefan diffusivities can be estimated by the Chapman-Ens
Eq. (9)[41

T3 1
T -+--
= 5.9543 x 10-24 2
PTo iD,ij

where Gij is collision diameter, ODij is the collision integral, and M, is
molecular mass of species i (all the units are in SI). The collision par
eters can be found, for instance in Bird, et al.14'

The experiment described below shows the limitations of Fick's equa
and introduces the Maxwell-Stefan equation for multicomponent diffus
The setup is shown in Figure 1.
Typically, two tanks of about the same volume are connected by a
mm internal diameter pipe (1/4" nominal diameter) that is 15.3 cm l1
Tank A has a 45.2-cm3 volume, while tank B has a 41.5-cm3 volume.
on/off valve divides the pipe at the middle and has about the same
internal diameter. A set of needle valves, on/off valves, and a pres
transducer are available to fill up the tanks. All the valves are mad
stainless steel by Whitey. The pressure transducer (Druck model PDCRI
has an 0-70 kPa absolute pressure range with 0.5% F.R.S. precision.
analysis is made at a high frequency by a mass spectrometer (Dataq
from Spectramass, UK) connected to one tank at a time. The n
spectrometer sample probe is made of a 1-m long fused silica coil
with 50 Ltm internal diameter. After two hours the total pres
changes less than 1% due to this mass withdraw when the in
pressure is 40 kPa (absolute pressure). A vacuum pump evacuates
system to a pressure below 0.3 kPa. All data are recorded on a c
puter every 100 seconds.
Students evacuate both tanks and then fill them with equimolar bi

Vacuum -------~ --- He

------ N2

Mass spectrometer P- C02

N2+He N2+CO2

[1] Needle valve
e Two-way on/off valve
I< Threeway switchingvalve
(p) Pressure transducer

Figure 1. Sketch of the experimental setup.

gas mixtures with the help of the pressure trans-
(8) ducer. They should fill both tanks with nitrogen
at the same time, up to a pressure of 20 kPa
(absolute pressure), and then add helium to
kog tank A and carbon dioxide to tank B, up to a
total pressure of 40 kPa. At the end, both
tanks must have the same total pressure. When
changing the feed gas, the filling circuit should
be evacuated-otherwise the residual gas will
(9) enter with the new feed gas. Nitrogen is the
common component in both tanks.
After filling the tanks, students are asked to
the start the data acquisition software, to switch
ram- on the mass spectrometer, to read the tem-
perature, and to open the switching valve con-
necting the tanks. Helium diffuses from tank
A to tank B, and carbon dioxide diffuses from
tion tank B to tank A. The total pressure difference
;ion. between the tanks should be negligibly small,
implying no viscous flow and so equimolar
4.3- diffusion.
ong. The diffusion constant is approximately in-
An versely proportional to the total pressure. A total
pipe pressure of 40 kPa allows students to complete
sure the diffusion experiment within the three-hour
e of laboratory session.
921) The experiment should be performed twice,
Gas switching the tanks' contents on the second run
uad, in order to record the concentration history of
rass both tanks. This can be done in two consecutive
umn classes of three hours each by two different stu-
sure dent groups. The two groups, working as a team,
itial should exchange their results and draw the nitro-
gen molar fraction curves as a function of time.
om- Then they can simulate their experimental sys-
tem with the available simulator and comment
nary on the results.

After opening the connecting valve, the gas
mixtures in the tanks enter in contact (see Figure
1). The connecting pipe mass balance can be
written as141

S- =0 (10)
az at
Since the total pressure gradient between the tanks
can be neglected, there is no viscous flow, and
therefore the total flux, N,, is zero. Introducing
Eqs. (6) and (7) into the mass balance, the fol-
lowing expression is obtained for constant tem-
perature and pressure:

Winter 2000

a(N) (c) a ct[D](x + (x)
+z at aL Jaz ] az
a[B] 8(x) 82(x) a(x)
ct[B]- z [B] C +ct

where the matrix product is not commutative. In the cas
three-component diffusion, a two-component matrix e
tion (Eq. 11) must be considered along with the total f
Nt=0. Assuming that the diffusion time constant inside
tanks is much smaller than the diffusion time constant in
pipe, the tanks can be considered as completely stirred.
mass balance of the complete system (Figure 1) is

VA ,

ANi 0 =0

for tank A, and

VA z= A Ni, =0

for tank B. L is the pipe length. 0
A fortran program, using an MS
Excel interface, was written to solve
this problem. The program is also 0
available on the web> 0 0
for remote simulation. It can be ap- L
plied to mixtures of 3 to 7 compo-
u- 0
nents. For solving the partial differ-
ential equations, along with the or- o
dinary differential equations, the 0
package FORSIM VI'5 was used.
At the beginning the nitrogen con- 0
centration is the same in both tanks
and therefore, according to the Fick
equation, nothing should happen to
it. Figure 2 shows the concentration
curves for the three gases in both
tanks. As can be seen, the concen- Figm
tration of nitrogen starts to decrease heliu
in tank B and to increase in tank A! curvi
Why? Helium and carbon dioxide tank
seem to behave as Fickian gases: lines
helium concentration decreases in

e of

Helium binary coefficients are high and it readily moves
from tank A to tank B while carbon dioxide moves slowly
from tank B to tank A. Nitrogen should balance these ef-
fects, and so it first moves with carbon dioxide from tank B
to tank A, to balance the very fast helium, and then returns to
tank A. Students are asked to internalize this picture in
opposition to the one given by Fick's law, where mass
transport is viewed as depending only on each component
concentration gradient.

the The mass spectrometer allows for an almost continuous
The concentration measurement. If not available, a different ex-
periment can be performed using a gas chromatograph. Two
samples can be collected from each tank using a syringe. In
this case, the tanks' total pressure should be 1 atm or more,
(12) to allow sampling. The samples can be collected at the
highest and lowest nitrogen partial pressures, and the time at
which this happens can be estimated from the simulation

re 2. Experimental an
'm (0), and carbon di.
es as a function of tim
A and closed symbol
represent the simulate

tank A and carbon dioxide in tank B until equilibrium is
reached. The step-like behavior of some experimental points
in Figure is related to the mass spectrometer resolution.
The diffusion coefficients of the three gas pairs at 40 kPa
and 200C are

SHe-N2 ='N, He=1.7076x10-4 m2/S

He-CO CO,-He = 1.4172x10- m2 /s

N2_Co, o,-N = 0.3688 x10-4 m2 /

Fick's equation is "intuitive"
and deeply rooted in the culture
of many engineering institutions.
While very simple, it is only valid
for binary systems or multicom-
ponent diluted systems. To change
the Fickian culture and internal-
ize a new feeling in the diffusion
area, we propose to the students a
ternary diffusion lab exercise. The
gases considered (helium, carbon
dioxide, and nitrogen) are neither
dangerous nor expensive. The ex-
perimental setup is also inexpen-
sive, provided that a mass spec-
trometer is available.
The experiment is very simple

Simulated nitrogen ( and can be easily performed in 3
d sim ulated nitrogen (N),
oxide (*) molar fraction hours. It also strongly demon-
e. Open symbols refer to states the inaccurate results that
Is to tank B. The solid Fick's equation can lead to under
ted results, some circumstances. The simula-
tion program that supports this ex-
periment allows students to play at home with different
systems, helping them to gain a new feeling for multicompo-
nent diffusion mass transport.

The authors wish to thank Professors Carlos Costa and Ferao
Magalhaes for their careful review of this manuscript.

A connecting pipe cross-section area (m2)
Chemical Engineering Education





.1 -- -- -

0 1500 3000 4500 6000 7500
TIME (s)

c. molar concentration of species i (mol m )
c, total molar concentration (mol m 3)
DAB Fickian A,B diffusivity (m2s ')
DAM Fickian diffusivity of species A in a mixture (m2s ')
SAB Maxwell-Stefan A,B diffusivity (ms ')
F molar force (N mol-')
J molar flux of species i with respect to the mixture molar
average velocity (mol m'-s-)
(J) (n-1)-component column vector of diffusion fluxes (mol
L connecting pipe length (m)
M molecular mass of species i (kg mol ')
N, molar flux of species i with respect to a fixed referential
(mol m 2s1)
N, total molar flux with respect to a fixed referential (mol
m-2s )
p partial pressure of species i (Pa)
Pt total pressure (Pa)
91 ideal gas constant (J mol 'K )
t time (s)
T absolute temperature (K)
VA,V, volume of tanks A and B, respectively (m3)
u velocity of species i (ms-')
x molar fraction of species i dimensionlesss)
(x) (n-1)-component molar fractions vector
z axial coordinate (m)
Greek Letters
pi molar chemical potential (J mol)')
oij collision diameter (m)

QD,ij collision integral dimensionlesss)

1. Krishna, R., and J.A. Wesselingh, "The Maxwell-Stefan Ap-
proach to Mass Transfer," Chem. Eng. Sci., 52, 861 (1997)
2. Taylor, R., and R. Krishna, Multicomponent Mass Transfer,
John Wiley & Sons, New York, NY (1993)
3. Wesselingh, J.A., and R. Krishna, Mass Transfer, Ellis
Horwood, New York, NY (1990)
4. Bird, R.B., W.E. Stewart, and E.N. Lightfoot, Transport
Phenomena, John Wiley & Sons, New York, NY (1960)
5. Carver, M., D. Stewart, J. Blair, and W. Selander, Forsim
VI, Chalk River Nuclear Laboratories, Ontario, Canada

Derivation of the Maxwell-Stefan Equation
(Based on References 1, 2, and 3)
Consider a pipe filled with an isobaric binary gas mixture, com-
ponents A and B. When moving, component A exerts a force on B.
This drag force should be proportional to the molar concentrations
of B and A and to the relative velocity of both components
FocACB(UA -UB) (Al)
where c, is the i-component molar concentration, u, is the i-solute
velocity referred to a fixed referential, and F is the drag force. On
the other hand, when considering an infinitesimal pipe slice, the
pressure exerted by component A on the imaginary left boundary is
PAIz and on the right boundary is PAIz=z+dz, where PA is the partial
pressure of A. The partial pressure gradient -dpA/dz is the driving
Winter 2000

force for component A to move inside the pipe and should be
balanced by the drag force



Calling ) AB / 91T to the proportional factor, we obtain

dz AB /9iT
For ideal gas mixtures, Eq. A3 simplifies to

dz AB



For ternary mixtures, we must add an additional drag force term to
the right-hand side of Eq. A4 to account for A-C interactions

dz AB AC
and for multicomponent systems

dxi n XiXj ui -Uj
dz j



The Maxwell-Stefan equation is usually written in terms of molar
fluxes: N,=c,ui. Replacing the velocities in Eq. A6 by molar fluxes,
we obtain

dx xNj -xjNi
dz c 'Th
j=1 t' i


The driving force is better represented by the chemical potential

dxi xi d(9T Cnxi) xi dpi
dz 9iT dz 9T dz
and introducing this result into Eq. A7, we obtain

xi dp-i xiNj -xjNi xiJj -xjJ
9iT dz c i j= c ij
jti j-i



This is the usual form of the Maxwell-Stefan equation for isobaric
and isothermal systems, where only pressure forces are present.
For binary mixtures, the Maxwell-Stefan equation reduces to the
Fick equation

JA = -ctDAB dA


and the Maxwell-Stefan diffusivity is the same as the Fickian
binary diffusivity. For multicomponent systems, the Fick equation
is written as



where DAM is the diffusion coefficient of A in the multicomponent
mixture. While the Maxwell-Stefan diffusion coefficients can be
considered essentially constant with the composition, the Fickian
diffusivity cannot, even for ideal gases and equimolar diffusion.[41 O




University of Minnesota Duluth, MN 55812-2496

generally, we prefer to work with equations that are
formulated to be independent of any particular sys-
tem of units. This is not always convenient to do,
and we use many equations that are valid only for a particu-
lar system of units. The undergraduate encounters these
throughout the chemical engineering curriculum. The ap-
pearance of "ge" in many equations in texts of American
origin prompts the reader to the fact that the units are En-
glish of the American variety. Students meet examples of
dimensional equations into which values of variables must
be entered with particular units; examples of these are em-
pirical correlations for heat and mass transfer coefficients
given in the text of McCabe, et al.1"
The need to find empirical correlations has probably been
the prime source of dimensional equations. In environmental
engineering and science, many empirical correlations in-
volve a relationship between the quantity of interest and the
octanol-water partition coefficient, Ko. This coefficient was
originally used by medicinal chemists12'31 interested in find-
ing correlations for equilibrium and transport properties of
chemicals within living systems. Many useful correlations
involving the octanol-water partition coefficient4'l5 exist;
these include measures of toxicity, measures of accumula-
tion of chemicals by organisms as well as molecular proper-
ties such as water solubility, Henry's law constants, molar
volumes, and measures of a chemical's surface area.
We wish to discuss the correlation that is widely used for
doing an order-of-magnitude estimation of the distribution
of hydrophobic chemicals in aquatic systems. It is
Koc _=axK (1)
Here, Koc is the sediment, or soil, sorption coefficient and a
is a constant, whose values are typically 1.0161 and 0.6.17' A

more general form that is often used is
log K =c = a-log Kow +b
The definitions of the coefficients are

K = C

Kow = Coct
y ^Ct,
'^o ^-

Here, Cs is the mass of chemical per unit mass of dry
sediment or soil, Foc is the fraction of organic carbon in the
dry sediment or soil, Cq is the mass of chemical per unit
volume of aqueous phase, and Co,t is the mass of chemical
per unit volume of octanol. The common use of the organic-
carbon normalized distribution coefficient, K,, for nonionic
organic chemicals arose out of the work of Karickhoff and
coworkers;171 they measured distribution coefficients, using
substrates containing various fractions of organic carbon,
and demonstrated that the normalized form is essentially
independent of the substrate type. Similar observations, in-
volving an organic matter basis (to be discussed later), had
been made earlier.18-101 Recent texts15s"' contain more de-

Keith Lodge is Assistant Professor of Chemi-
cal Engineering at the University of Minne-
sota in Duluth. He was educated in the United
Kingdom, obtaining his BSc from the Univer-
sity of Warwick and his PhD from the Univer-
sity of Sheffield. He teaches laboratory
courses, thermodynamics, heat transfer,
computational methods, reactor design, and
process control. Properties of hydrophobic
organic compounds are his principal research

Copyright ChE Division ofASEE 1999

Chemical Engineering Education

tailed descriptions.
Equations (1) and (2) are usually used in a tacitly dimen-
sional way with the chosen units being SI units; generally,
values of a are given without units, K,, is unitless, and Koc
has units of L/kg or mL/g. Our purpose here is to ask, "What
is a dimensionally consistent form of these equations?" We
wish to show there is pedagogical value in answering this. In
our view, the dimensional inconsistency arises from the fact
that the concentration bases are different. The concentration
of the chemical in sediment or soil, C,, is defined as the mass
of chemical per unit mass of dry sediment or soil and the
concentration of the chemical in octanol is defined as the
mass chemical per unit volume of octanol. So, to obtain a
dimensionally consistent equation we should ensure that we
have the same composition basis. To do this, we imagine
octanol as a sample of soil or sediment.
By definition

Coct = (5)
where m is the mass of chemical in the volume of octanol
Vot. Using the density of octanol, poct, and the mass of
octanol, Mot, we transform this equation to

Coct = Poct = CsPoct (6)
We recognize the term m/Mot, the mass of chemical per unit
mass of octanol, is equivalent to C,. Keeping in mind the
form of Eq. (3), we write

Coct =CsPoct = (FocPoct) (7)
where Foc, is the fraction of organic carbon in octanol.
Dividing through by the aqueous concentration of the chemi-
cal, we obtain the equation

Koc- Kw (8)
and this is dimensionally consistent. Recognizing that all the
carbon in octanol is organic, we calculate the fraction of
organic carbon in octanol from the relative atomic and mo-
lecular masses; Foc = 0.738. The density of octanol1 21 at
20'C is 0.827 g/mL. So, at 200C,

Koc = 1.638Kow = Koct or logK' = logKo +0.214 (9)
This transformation gives the octanol-water partition coef-
ficient on the same composition basis as the sediment, or
soil, sorption coefficient. It is still the octanol-water partition
coefficient, but expressed on a different basis. To emphasize
this, we now designate it as Kot. This is a very unusual way
of expressing compositions; the transformations between

concentrations, in terms of molarity or molality, and mole
fractions are much more familiar to us.
So, given a value of the octanol-water partition coefficient
on its normal basis, we can calculate it on an organic carbon
basis. The question now is, "To what extent does the octanol-
water partition coefficient on the organic carbon basis corre-
spond to the measured sediment, or soil, sorption coeffi-
cient?" If the organic carbon in the sediment behaves identi-
cally to octanol, then we expect the relationship in Eq. (9) to
hold. What is observed? From experimental data, many work-
ers have developed dimensional relationships with the gen-
eral form of Eq. (2), in which values of a and b are deter-
mined by linear regression. To make the essential point here,
we consider only the relationship developed in a recent
comprehensive review131 in which Baker and coworkers de-
veloped selection criteria and critically reviewed the avail-
able measurements. For 1.7 < log K.o < 7.0, using data for
72 chemicals, they found

a = 0.903 + 0.034

b = 0.094 0.142

r2 =0.91

We wish to compare Eq. (9) with this result.
The dimensionally consistent relationship that we derived,
Eq. (9), however, requires a=l. Using the data in the re-
view,131 we applied a regression model1141 in which we forced
a to be unity. We obtain

log Koc = log K,, (0.29 0.05)

r2 = 0.90 (10)

In Figure 1, we have plotted the data, the regression line (Eq.

1.0 2.0 3.0 4.0 5.0 6.0 7.0

log K,

Figure 1. The units of Koc are L/kg. Data were taken from
Ref. 13. The dashed line corresponds to the regression line,
Eq. (2), with a=0.903 and b=0.094. The continuous line
corresponds to Eq. (10) in the text.

Winter 2000

2 with the values of a and b given above), and the line
corresponding to Eq. (10). From Eqs. (9) and (11), we find

Ko (observed, Eq. 10) =0.31
K (ct (oc tan ol like behavior, Eq. 9)

In other words, experimental values of Koc are about one-
third of the values expected if the sediment or soil organic
carbon were to have the same partitioning properties as
The conclusion hinges on an appreciation of how to ex-
press compositions in various ways; the way here for soil or
sediments is peculiar to environmental work and is a useful
exercise for students to work out for themselves. Another
closely related example is the earlier uses8-10] of the soil
sorption coefficient on an organic matter basis. This is de-
fined as

Kom- (12)

Here, Fom is the fraction of organic matter. We may derive an
expression for the octanol-water partition coefficient on an
organic-matter basis, K ,ct following the same steps as be-
fore. The result is

K- K ow (13)

Octanol is all organic "matter," so Fo, = 1, and we obtain
Koct = 1.209Ko or log Kt = logKow +0.082 (14)
om ow om oW
This may be compared to the approximate experimental
relationship found between the soil sorption coefficient on
an organic-matter basis and the octanol water partition coef-
ficient,[41 Ko = 0.4 K.o. Here again, we may conclude that
the experimental values of Kom are about one-third of the
values expected if the sediment or soil organic matter were
to have the same partitioning properties as octanol. In con-
trast to octanol, the fraction of organic matter in soils is
about twice the fraction of organic carbon."5t'O"11 The mea-
surement of the fraction of organic carbon is now easier, and
so the use of the organic-carbon basis is now more prevalent.
We think it is important for the student to recognize when
an equation is dimensional, and it is often not immediately
obvious. The answer to the question, "Why is the sediment,
or soil, partition coefficient less than the octanol-water parti-
tion coefficient?" is a useful entry point into a discussion of
the structure of sediment, or soil, particles (a heterogeneous
solid system) and the nature of adsorption. This is in contrast
to distribution of a chemical between two essential homoge-
neous liquid phases, as represented by the octanol-water
partition coefficient.

1. The composition of a phase is usually described by the
mole fractions of the various components. Why is it
impractical to describe the composition of a soil or
sediment in terms of mole fractions?
2. Equation (10) is a dimensional equation in which the
units are SI. What is the equivalent equation in English
3. Ten milligrams of naphthalene is added to a container
that contains 10 g of sediment (dry wt), 50 mL of water,
and 5 mL of octanol. The system is allowed to reach
equilibrium. What are masses of naphthalene in the
sediment, water, and octanol at equilibrium? The sedi-
ment contains 5% organic carbon and the octanol-water
partition coefficient for naphthalene is about 2000.

We acknowledge support provided by the U.S. EPA with
cooperative agreements CR-813504 and CR-817486, and by
the U.S. Air Force with award number F49620-94-1-0401.
The ideas presented here were developed during experimental
work done under these grants; they should in no way be taken
to represent the opinions of the grantors.

1. McCabe, W.L., J.C. Smith, and P. Harriott, Unit Operations in
Chemical Engineering, McGraw-Hill, New York, NY (1993)
2. Leo, A., C. Hansch, and D. Elkins, "Partition Coefficients and
Their Uses," Chem. Rev., 71(6), 525 (1971)
3. Hansch, C., A. Leo, and D. Hoekman, Exploring QSAR, Ameri-
can Chemical Society, Washington, DC (1995)
4. Sandler, S.I., Chemical and Engineering Thermodynamics,
3rd ed., John Wiley & Sons, Inc., New York, NY (1999)
5. Schwarzenbach, R.P., P.M. Gschwend, and D.M. Imboden,
Environmental Organic Chemistry, 1st ed., Wiley, New York,
NY (1993)
6. Di Toro, D.M., "A Particle Interaction Model of Reversible
Organic Chemical Sorption," Chemosphere, 14(10), 1503 (1985)
7. Karickhoff, S.W., D.S. Brown, and T.A. Scott, "Sorption of
Hydrophobic Pollutants on Natural Sediments," Wat. Res., 13,
8. Lambert, S.M., "Omega, a Useful Index of Soil Sorption Equi-
libria," J. Agr. Food Chem., 16(2), 340 (1968)
9. Briggs, G.G., "Molecular Structure of Herbicides and Their
Sorption by Soils, Nature (London), 223, 1288 (1969)
10. Briggs, G.G., "Theoretical and Experimental Relationships
Between Soil Adsorption, Octanol-Water Partition Coefficient,
Water Solubilities, Bioconcentration Factors, and the Parachor,"
J. Agr. Food Chem., 29, 1050 (1981)
11. Valsaraj, K.T., Elements of Environmental Engineering, 1st
ed., CRC Press, Boca Raton, FL (1995)
12. Daubert, T.E., Physical and Thermodynamic Properties of Pure
Chemicals, Hemisphere Publishing Corporation, New York,
NY (1989)
13. Baker, J.R., J.R. Mihelcic, D.C. Luehrs, and J.P. Hickey, "Evalu-
ation of Estimation Methods for Organic Carbon Normalized
Sorption Coefficients," Water Environ. Res., 69, 136 (1997)
14. Synergy Software, KaleidaGraph. Ver. 3.09, Synergy Software,
Reading, PA (1996) J
Chemical Engineering Education

Full Text

xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E49TL8U6F_CVA5V2 INGEST_TIME 2012-02-17T16:26:59Z PACKAGE AA00000383_00145