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
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
sobekcm - AA00000383_00039
lcc - TP165 .C18
ddc - 660/.2/071
System ID:

Full Text



4Zdo f 9 7.2 4w a4d 2ed4f4e,:




Paper. You might think it's almost
too simple to be interesting. News-
papers, napkins, towels, cups. Simple.
But have you heard about the paper
that's used as the core of high-impact
plastic laminates? Or the pre-lubri-
cated communication paper tape for
data-processing equipment? Or the
specialty papers with complex coat-
ings for keeping packaged foods fresh?
Or the 3-ply paperboard made with
polyethylene film, virgin pulp
and recycled pulp which
together create an
excellent moisture
And did you know
about our chemical
products that are used

in environmental systems for air and
water purification? Or our specialty
chemicals used to make products in
almost any industry you name?
And about the research involved in
all this? The physical, organic and in-
organic chemistry? The physics, math-
ematics and biology? The mechanical,
chemical and electrical engineering?
At Westvaco, making paper is a
whole lot more than just making paper.
And that's why we're interested
in innovative, naturally
Curious people who
can see the future in
paper the way we do.
With them, we think
we can help shape the
F- 7 future for everyone.

299 Park Avenue
New York, N.Y. 10017

An Equal Opportunity Employer

Department of Chemical Engineering
University of Florida
Gainesville, Florida 32601

Editor: Ray Fahien
Associate Editor: Mack Tyner
Business Manager: R. B. Bennett
(904) 392-0861
(904) 392-0881

Advertising Representatives:
Publications Board and Regional
SOUTH: Charles Littlejohn
Chairman of Publication Board
Clemson University
Clemson, South Carolina 29631
CENTRAL: James H. Weber
University of Nebraska
Lincoln, Nebraska 68508
WEST: William H. Corcoran
California Institute of Technology
Pasadena, California 91109
University of Houston
Houston, Texas 77004
EAST: Miguel Hnatow
New York University
New York, N.Y. 10453
G. Michael Howard
University of Connecticut
Storrs, Connecticut 06268
George D. Keeffe
Newark College of Engineering
Newark, New Jersey, 07102
Thomas W. Weber
State University of New York
Buffalo, New York 14214
NORTH: J. J. Martin
University of Michigan
Ann Arbor, Michigan 48104
NORTHWEST: R. W. Moulton
University of Washington
Seattle, Washington 98105
D. R. Coughanowr
Drexel University
Philadelphia, Pennsylvania 19104
Stuart W. Churchill
University of Pennsylvania
Philadelphia, Pennsylvania 19104
University of California
Santa Barbara, California 93106
Secretary's Department
E. I. du Pont de Nemours
Wilmington, Delaware 19898

Chemical Engineering Education

Feature Articles
4 4 eot ial Csoiee4i4% 4,wa,*d l, ectiae :
Process Synthesis
Dale F. Rudd
30 The Polymer Program at Caltech
R. E. Cohen and N. W. Tschoegl
16 Applications of Heterogeneous Catalysis
Dan Luss and J. T. Richardson
4 The Educator
Steve Whitaker
8 Departments of Chemical Engineering
14 Views and Opinions
Freshman Engineering-A Student
D. L. Hunter
18 Curriculum
Network Planning and the ChE Curriculum
R. C. Cunningham and J. T. Sommerfield
22 The Classroom
The Interphase Catalytic Effectiveness
Germain Cassiere and J. J. Carberry
38 Modern Analysis Techniques with the APL
Co Pham and Leonce Cloutier
28 Book Review
33 Problems for Teachers
Analog Simulation of the Dispersion of
Atmospheric Pollutants
C. W. Miller and T. W. Cadman
40 The Laboratory
Design of Process Control Using Frequency
Response and Analog Simulation
J. F. Paul

CHEMICAL ENGINEERING EDUCATION is published quarterly by the Chemical
Engineering Division, American Society for Engineering Education. The publication
is edited at the Chemical Engineering Department, University of Florida. Second-class
postage is paid at Gainesville, Florida, and at DeLeon Springs, Florida. Correspondence
regarding editorial matter, circulation and changes of address should be addressed
to the Editor at Gainesville, Florida 32601. Advertising rates and information are
available from the advertising representatives. Plates and other advertising material
may be sent directly to the printer: E. 0. Painter Printing Co., P. 0. Box 877,
DeLeon Springs, Florida 32028. Subscription rate U.S., Canada, and Mexico is $10 per
year, $6 per year mailed to members of AIChE and of the ChE Division of ASEE,
and $4 per year to ChE faculty in bulk mailing. Write for prices on individual
back copies. Copyright @ 1973. Chemical Engineering Division of American Society
for Engineering Education, Ray Fahien, Editor. The statements and opinions
expressed in this periodical are those of the writers and not necessarily those of the
ChE Division of the ASEE which body assumes no responsibility for them. Defective
copies replaced if notified within 120 days.




The jollowian companies has e donated

unct foo ke o dpofw4 t CHEMICAL ENGINEERING EDUCATION d4ain" f973:




h 41owin sf 103 depaodents hkae

coatAduled to the dp2fl o4 CHEMICAL ENGINEERING EDUCATION in f973

University of Akron
University of Alberta
Arizona State University
University of Arizona
University of Arkansas
Auburn University
Brigham Young University
Bucknell University
University of Calgary
University of California (Berkeley)
University of California, Davis
University of California (Santa Barbara)
California Institute of Technology
Case-Western Reserve University
Clarkson College of Technology
Clemson University
University of Colorado
Colorado School of Mines
Cooper Union
Cornell University
University of Delaware
University of Detroit
Drexel University
Ecole Polytech, Canada
University of Florida
Georgia Institute of Technology
University of Idaho
University of Illinois (Urbana)
Illinois Institute of Technology
Iowa State University
University of Iowa
University of Kansas
Kansas State University
University of Kentucky
Laval University

Lehigh University
Loughborough University (England)
Louisiana Polytechnic Institute
Louisiana State University
University of Louisville
McGill University
McMaster University
McNeese State University
University of Maine
Manhattan College
Massachusetts Institute of Technology
University of Massachusetts
University of Michigan
Michigan State University
University of Mississippi
University of Missouri, Rolla
Montana State University
New York University
University of New Mexico
University of New Brunswick
University of New Hampshire
University of New South Wales
North Carolina State University
Northwestern University
University of Notre Dame
Nova Scotia Technical College
Ohio University
University of Oklahoma
Oklahoma State University
Oregon State University
University of Ottawa
University of Pennsylvania
Pennsylvania State University
University of Pittsburgh
Polytechnic Institute of Brooklyn

Princeton University
University of Puerto Rico
Purdue University
University of Quebec
Queen's University
Rensselaer Polytechnic Institute
University of Rhode Island
Rutgers-The State University
State University of N. Y. at Buffalo
Stevens Institute of Technology
Syracuse University
Technion, Israel
Tennessee Technological University
University of Tennessee
Texas A&I University
University of Texas at Austin
Texas A&M University
University of Toledo
Tufts University
University of Tulsa
University of Utah
Vanderbilt University
University of Virginia
Virginia Polytechnic Institute
Wayne State University
Washington University
University of Washington
University of Waterloo
West Virginia University
University of Wisconsin
Worcester Polytechnic Institute
University of Wyoming
Yale University
University of Windsor

TO OUR READERS: If your department is not a contributor, please ask your
department chairman to write R. B. Bennett, Business Manager, CEE, Depart-
ment of Chemical Engineering, University of Florida, Gainesville, Fla. 32601.
Bulk subscription rates at $4/yr each with a $25.00 minimum for six or
fewer subscriptions. Individual subscriptions are available to ASEE-CED and
AIChE members at $6 yr.


n letters

Experimentation and Modeling
Sir: The article on kinetics of yeast growth (Chem. Eng.
Education, 6, 134-7 (1972)) should not go unchallenged,
for the lesson it teaches is not what designers of the
experiments have intended. Oxygen or some constituent
other than sugar must have limited growth of the yeast,
despite apparent agreement of the data with a simple
One can overlook the unexplained failure of Beer's
Law to describe turbidity of a yeast suspension or what
one hopes is a typographical confusion between corn sugar
and cane sugar. (If the latter were indeed used, consistent
results for sugar analysis would have been difficult even
with a detailed recipe!) Of more consequence is a failure
to recognize that the experiment is not at all "patterned
after the commercial process for growing yeast," (which
employs a low concentration of sugar so as to avoid
catabolite repression of respiratory enzymes-the kinetic
behavior can be complex).
Results of the published experiment showing departure
from exponential growth after only five hours, when the
sugar content was still 15 gm/liter, should have alerted
the designers to some other rate-limiting constituent like
oxygen. Having called attention to such possibilities in
presenting theoretical background, the author neglected
to perform even the simple test of trying oxygen-enriched
air once the rate of growth declined. More elegant of
course would have been to use a dissolved oxygen probe.
The data suggest that after about five hours growth be-
came linear, a very nice example of mass transfer limita-
tion of a kinetic process.
The sought-for relationship between growth rate and
sugar concentration cannot be found with any accuracy
by taking slopes off an integral curve as proposed. The
correct concentration for half-maximal rate (1/bs) is so
small that in any properly conducted experiment the rate
curve remains exponential until just before complete ex-
haustion of sugar when it bends over sharply to zero.
Initial rate data were used by Monod; others have used
continuous backmix reactors to obtain the desired relation-
Except for the nagging question of how fractions of
a volume per cent of yeast can be measured in 12-ml
centrifuge tubes, the experiment itself is a good one. It
provides a first-class example of the hazard of overzealous
fitting of kinetic data to an a priori mathematical model
-a hazard fully emphasized in Levenspiel's text.
R. K. Finn
Cornell University

Anderson Replies
Sir: Professor Finn has restated one of the conclusions
reached by several of the students who have run the ex-
periment: Another factor may partially limit the growth
of the yeast. Those students received top marks, and with
a lab report written without jumping to conclusions
Professor Finn might also receive top marks. In fact, data
of other groups have yielded lower values for (1/b.), but

in all cases the yeast stop growing when the sugar runs
out. A simple check on the effect of the oxygen supply
can be obtained by varying the air feed rate. Regarding
corn vs. cane read corn but buy dextrose from a labora-
tory supply house. Of course, "patterned after" does not
imply exact duplication of a commercial process but
merely similarity. The use of 12-ml centrifuge tubes for
measuring yeast cell volume is quite simple and gives
surprisingly reproducible results as shown by sample
The objectives and requirements of research investi-
gations and undergraduate experiments differ in several
important respects. In the former the limitations of time
and complexity are not near as severe as in the latter.
The time available in an undergraduate course laboratory
precludes the use of both initial rate measurements and
continuous backmix reactors as suggested by Finn. The
data obtained in most batch experiments are adequately
represented by the accepted model, which includes the
effects of many constituents other than sugar. The model
is certainly not an a priori model since it results from
the "shrewd scientific detective work" (1) of a number
of workers over a period of years.
J. B. Anderson
Yale University

(1) 0. Levenspiel, "Chemical Reaction Engineering,"
Wiley & Sons, Inc., New York, 1962, p. 31.

Tenn Tech Is Accredited

Sir: The chemistry survey reported by Dr. James Cobb
in the summer 1972 issue of Chemical Engineering Edu-
cation may leave the impression that the Department of
Chemical Engineering at Tennessee Technological Uni-
versity is not presently accredited. Without worrying
about the merits of achieving accredited status I would
like to point out that this department, formed as a de-
partment in the fall of 1966, was inspected in the spring
of 1970 and accredited by action of the AIChE and ECPD
in the summer of 1970. I see other universities in the
list that may have similar comments.
While I have this opportunity let me express my
appreciation for the efforts you and your associates put
into Chemical Engineering Education. I find it both in-
teresting and useful.
John C. McGee, Chairman
Tennessee Technological University

Steam Table Correction
Sir: For those chemical engineering teachers of thermo-
dynamics who are accustomed to using the little steam
table booklets prepared by Combustion Engineering, Inc.,
the most recent edition contains an erroneous enthalpy
value that might cause a few anxious moments in class
for the unwarned. Whereas most of us are familiar with
the convention of assigning the value zero to the en-
thalpy and entropy of "saturated" liquid water at 320F,
the latest edition of "Steam Tables" shows the disquiet-
ing value of 0.0179 Btu/lbm for the enthalpy hf at this
(Continued on page 52)



Serae hote&k 01


IN CARMEL-BY-THE-SEA there once lived a black-
smith who fashioned hot iron into wondrous
objects while a son watched wide-eyed and took
note of each step as drill stock was fashioned
into candlestick and sheet iron into weather
vane. But the Golden Bear from Berkeley beck-
oned and an artisan was lost from the diminish-
ing ranks of "smithies" in order to swell the
ranks of another Smith. So it was that Steve
Whitaker began his undertaking as a chemistry
major at Berkeley when that campus was known
for its grassy glades, the pleasures of Strawberry
Canyon, and a clear view of the now disappear-
ing San Francisco skyline. A persuasive upper
classman, "Skip" Scriven, effected the change to
chemical engineering while also demanding par-
ticipation in the Sunday afternoon volleyball
game at Bowles Hall. Ever since, these two ac-
tivities have been interfering with each other,
one always diminishing the success of the other.
In 1954 Steve Whitaker started his graduate
studies at the University of Delaware working
with Professor R. L. Pigford on the problem of
interfacial resistance to gas absorption. The re-

This article was submitted to CEE by J. M.
Smith of the Chemical Engineering Department
of the University of California at Davis.

Underneath lay the perplexing
question of how to incorporate
a rigorous treatment of fluid
mechanics into the chemical
engineering program . and move
logically to a problem solving
technique with which the practical
problems of engineering could
be solved with confidence.

sults of that study have remained anonymous,
but the view of ubiquitous Bob Pigford was in-
spiring and Professor Whitaker's long-time in-
terest in interfacial phenomena can be traced
to that early encounter with the subject. When
life in the laboratory became too gruesome, Scriv-
en was always waiting with the Delaware Engi-
neers volleyball team, ready to sally forth to do
battle with the Four Corners Tuesday Night
Recreational Volleyball Club. Lee Brown offered
fantastic sailing on the Chesapeake Bay and was
brave enough to let novice Whitaker handle the
main in a howling October gale. After an hour
of clinging to the side of Lee's capsized boat,
Whitaker was relegated to crewing for Mary
Wilkenson who was fast enough at the tiller to
offset the many blunders made by her crew.
After completing graduate work in 1958
Steve Whitaker moved to the nearby Du Pont
Experimental Station and began working in the
area of fluid mechanics while attending Jim Car-
berry's weekly seminars on casualty and mod-
ern man the study of an evolutionary society.
Although his PhD thesis dealt with the use of
frequency-response techniques to study inter-
facial mass transfer, Whitaker's graduate school
experience left him with the impression that fluid
mechanics was the weakest link in the chemical
engineers training, and his research at the Engi-
neering Research Laboratory provided an op-
portunity to study this subject. A confused
attitudee prevailed until visiting lecturer Bob Bird


provided Steve
the opportunity
to renew his
acquaintance with
the High Sierra.


provided one of his inimitable multi-colored pres-
entations, and the elements of transport phe-
nomena were unfolded across a sixty foot black-
board. Bird's ideas were in marked contrast to
the pragmatic attitude at Du Pont where engi-
neers struggled to describe the vagaries of the
real world in terms of a finite number of pre-
viously derived equations. A more attractive
alternative was lurking in Bird's comments;
namely the ability to analyze any and all real
processes in terms of the finite set of laws of
continuum physics.
A brief discussion with George Bankoff at
an AIChE meeting in Mexico City eventually led
to an appointment on the faculty at Northwest-
ern University in 1961, and there Whitaker's
understanding of fluid mechanics began to take
on some recognizable form. This was due in no
small measure to the necessity of explaining the
subject to the undergraduates, and to endless
discussions with John Slattery. Interest in sur-
face phenomena, generated by the work with
Pigford, led to a series of research efforts on the
fluid mechanics of interfacial phenomena, while
the persistent nagging by Slattery about the
inadvisability of using Darcy's law for two-
dimensional flows led to a permanent interest in
the subject of transport processes in multi-phase

L IFE AT NORTHWESTERN was not all academic,
and dream of glory nurtured by earlier vic-
tories with the Bowles Hall Chargers and the
Delaware Engineers, led Professor Whitaker to
join the much-traveled (10,000 miles per year)
Chicago volleyball team. While holding records
for travel, the Chicago team reached in vain
for a national championship. Always finishing
among the top ten teams in the country yielded

but a mediocre thrill. Sailing on Lake Michigan
provided some sport in the off season, and as co-
owner with George Brown of an ageless and
nameless racing sloop a more definitive com-
petitive position was assumed, i.e. dead last.
Underneath these activities lay the perplex-
ing question of how to incorporate a rigorous
treatment of fluid mechanics into the chemical
engineering program, and how to structure an
approach to engineering analysis that would
"begin at the beginning," i.e. the fundamental
postulates of continuum physics, and move logic-
ally to a problem solving technique with which
the practical problems of engineering could be
solved with confidence. The crux of the matter
seemed to be to break away from the luxury of
using previously derived equations, and build
directly on the fundamental postulates in their
most elementary form.* Calculus, linear algebra,
and vector analysis, all standard lower division
subjects, would provide the tools if they could
only be put to use.

. . dream of glory nurtured by earlier victories
with the Bowles Hall Chargers and the Delaware
Engineers, led Steve to join the much-traveled
Chicago volleyball team.

- -_ -- .. ... ... __.
Lunchtime activity on the UCD sandcourts.
In 1964 Professor Whitaker returned to Cali-
fornia to join Joe Smith's new chemical engineer-
ing program at the University of California at
Davis. The unstructured nature of the new Col-
lege of Engineering provided the opportunity
of teaching both fluid and solid mechanics in the

In this respect Bertrand Russell once noted, "The
habit of simply assuming results, once one is persuaded
they are true, rather than trying to prove them, has all
the advantages of thievery over honest toil."


Steve and Suzanne plus
Suzy, Collin Walter, Larry,
Lynn and Jimmy on a
Sunday afternoon outing.

common core courses. A few (very few) polite
conversations regarding course content and level
were always held, but angry debate would more
accurately describe the discussions about the
common core courses. A heavy dose of "slings
and arrows" will usually clear ones head, and
Professor Whitaker's thoughts on fluid mechan-
ics were crystallized in the book, Introduction
to Fluid Mechanics, which had its beginnings at
Northwestern. Although the common core courses
have essentially disappeared, Professor Whita-
ker's interest in the undergraduate program has
not waned, and an introductory text on heat
transfer is currently being prepared.

Other climbing activities led to an assault on 21,000
ft Mt. Ausangati in Peru and a 28 hour off-route
climb of the East face of Mt. Whitney.

THE RETURN TO CALIFORNIA also provided Steve
Whitaker the opportunity to renew his ac-
quaintance with the High Sierra where many a
boyhood summer was spent hiking and fishing;
however, the population explosion in California
had flooded his favorite haunts and heavier loads
had to be carried higher and further to avoid the
crowds. Charlie Sleicher provided some relief
with excursions to the Northwest and the spec-
tacular Ptarmigan Traverse. Other climbing
activities led to an assault on 21,000 ft Mt.Aus-
angati in Peru (failure occurred at 19,000 ft
when the tent blew apart), and a 28 hour off-
route climb of the east face of Mt. Whitney.
Springtime will usually find Steve Whitaker
sneaking out of his office early on Fridays,
squeezing in three-day week ends with his wife
Suzanne at Yosemite Valley where the "big walls"
are contemplated, but avoided in favor of the

smaller and less strenuous climbs that abound
in the valley. The Whitaker boys, Larry and
Jimmy, have participated in some small climbs,
while the girls, Lynn and Suzy, must wait a few
summers before they can join their brothers on
the rock. Suzanne views this activity benignly
from below, but leads the way when it's time
to take the high country trails.
In 1964 Davis was a town without a volley-
ball team, and like a true devotee Steve Whitaker
organized and coached the UCD Volleyball Club
during its early years. After 16 straight defeats
at the hands of southern California teams, he
recognized that coaching was not his game and
he threw his talents in with a group of misfits
from several northern California cities known
as "Friends of Fred." Friends indeed they were
not, and the team was known far and wide for
its internal strife if not for its ability to win
matches. Personal opinions were put aside for the
1970 National AAU Masters (over 35) Volley-
ball Championships and the team at last managed
to bring home the first place medal. In 1972
everyone was two years older and fourth place
was the best that "Friends" could do in the same
While national competition is a thing of the
past, the local players are still easy pickings.
Fran Woods and Steve Whitaker have never lost
a coed doubles match on the UCD volleyball sand
courts, and lunch time will usually find Jim Hur-
ley (Physics) and Steve Whitaker tenaceously
holding the number one court against an array
of frustrated undergraduates. That too must
pass, but there will always be a bright-eyed
chemical engineering student intent on master-
ing the laws of physics for some good purpose,
and there is much to be done in that domain. D


Beneath this soft and warm exterior,
there lies a heart of plastic.

So tar, it's only a valve. Eight-year-old
Janet Hemandez has one.
It may not be long before a whole
working heart will be made out of plastic.
Men in plastics research at Union
Carbide are working on the almost im-
possible job of designing plastics com-
patible with the body.
Their most crucial job is making an
ultra-thin polypropylene fabric for lining
the inside of the heart. A fabric coated
with parylene that will allow human tis-
sue to grow into and around it to keep
blood from clotting.
A plastic heart isn't the only part of
the body we're working on. Maybe some-
day there will be a little plastic in all of us.
Right now, we've got you surrounded

by our plastics. We were in plastics be-
fore most people knew the word. We
make more plastics than anyone else. We
haven't scratched the surface yet.
Why is a great big company like Union
Carbide so concerned about a little bit of
plastic for the body?
Beneath our corporate exterior, there
beats a heart.

270 Park Ave., New York, N.Y. 10017




This paper was submitted by the Chemical Engi-
neering faculty, University of Texas, Austin.

How did chemical engineering get started at Texas?
The parallelism that exists between the De-
partment of Chemical Engineering at The Uni-
versity of Texas at Austin and the hydrocarbons
processing industry of the state is not accidental.
Each has complemented the other since before
World War I, when the late Dr. E. P. Schoch
foresaw the need to apply his profound knowledge
of chemistry to develop uses for the crude oil
which production was increasing daily. As a re-
sult of Dr. Schoch's direct effort, a degree pro-
gram leading to the Bachelor's Degree in Chemical
Engineering was authorized by the Regents early
in 1916 and the first two degrees were awarded
in June 1919. Official recognition was given to the
new Department through the changing of Dr.
Schoch's title from "Professor of Physical Chem-
istry" to "Professor of Chemical Engineering and
Physical Chemistry."
For almost 20 years Dr. Schoch was the only
permanent faculty member, but as the student
load became heavier he was given assistance by
the appointment of full time instructors on a year
to year basis. In 1938 Chemical Engineering was
established as a full Department in the College
of Engineering with Dr. E. P. Schoch as Chair-
man. Among the early faculty members were Drs.

J. S. Swearington, John Griswold, W. A. Cunning-
ham, and K. A. Kobe. Physically, the Depart-
mental activities were housed within the Chemis-
try Building until a new building for Chemical
Engineering was authorized and constructed dur-
ing the years immediately preceding the World
War II. The building, containing 33,000 sq. ft. of
floor space is devoted exclusively to office and
laboratory usage.
The accrediting program of the AIChE was
suspended nationally early in World War II, but
this Department was given provisional accredit-
ing in 1943, and after the war full accreditation
was received and has been retained since that time.

How big is your department?
Prior to 1935 only 130 bachelors degrees had
been granted with an average of 8 graduating
seniors per year, but this number jumped to 24
in 1935 and reached a pre-war peak of 54 in 1939.
With the onset of the war, the normal under-
graduate activities slowed very substantially, but
picked up again in 1947 when 80 degrees were
awarded and reached a peak in 1948 with 105
Bachelor's Degrees. Since that time the average
annual number of B.S. graduates has been ap-
proximately 60. Total undergraduate enrollment
is currently about 350 full time students. Gradu-
ate enrollment increased steadily after 1948 until
1965, reaching a level of 50 to 60 full time
students per year.


How does your undergraduate curriculum compare
to that of other ChE departments?
Chemical Engineering curricula among the
various schools have many similar features. We
require the usual freshman courses in the areas
of mathematics, physics, chemistry, drawing, etc.,
but in addition recently introduced a PSI course
(see below) in computer programming for enter-
ing freshmen. One feature of this course, in addi-
tion to the development of programming skills,
is the motivation of the students toward chemical
engineering by means of a series of talks by Pro-
fessor Schechter, the departmental chairman.
After the freshman year the typical student
takes a series of core courses in chemical engi-
neering consisting of material and energy bal-
ances, transport phenomena, thermodynamics,
unit operations, kinetics, plant design, and lab-
oratory. In addition, he takes courses in mathe-
matics, chemistry, some engineering, physics,
English, government, and history. He also takes
eight elective courses generally grouped into
blocks of similar content, such as

Computer applications, process analysis, control
Materials engineering and chemistry
Management and management decision making
Environmental improvement
Chemical processing
Biomedical engineering

It takes the average student four years plus a
semester to complete the undergraduate program.
Cooperative students spend three semesters work-
ing in industry in addition to their academic work.

What type of laboratory courses do you have?
A distinct upswing in the quality of our lab-
oratory courses and facilities has taken place in
the last three or four years. We believe that
modern methods of data acquisition, real time
analysis, and graphical display can be of value in
making the undergraduate laboratory experience
more meaningful for students. To implement this
philosophy we have assembled an 8K mini-com-
puter, A/D and D/A converters, magnetic tape
drives, teletype, and CRT display, and intercon-
nected them to the main university computer (see
Table 1). The system is used in a variety of ways
in the experiments from presenting tutorial ma-
terial to the calculation of final results. The com-
puter is incorporated into the procedure in a
manner that improves student's understanding of
the experiment while not isolating him from it.

R. S. Schechter, Department Chairman

The three main undergraduate laboratory courses
in which the system is used are transport phe-
nomena, process control, and the unit operations
In general the work assigned in these labora-
tory courses has the principal objective of bring-
ing students into actual contact with the physical
world. Students are expected to see the corres-
pondence between terms in an equation and the
corresponding physical quantities; to make valid
approximations while avoiding unwarranted as-
sumptions; and establish that the theory they
have learned does describe real world systems. To

Table 1. Characteristics of the Department of Chemical
Engineering Computer System
I. Nova Mini-computer
A. 16 bit word length
B. 8 K words of memory
C. Integer arithmetic
D. 1.2 microsecond memory cycle time
E. 4 accumulators
II. Data Acquisition
A. 12 channel multiplexer
B. Binary-gain range amplifier
C. Analog-to-digital converter
III. Analog Output on 4 Channels
IV. Peripherals
A. Computek graphics terminal
B. Dual deck magnetic tape cassette unit
C. ASR 33 teletype
V. Communications Interfaces (for communication with
XDS-930, Sigma 5, and CDC-6600)
A. 110 bits/sec,. to teletype
B. 110 bits/sec,. to Bell type 103A data set
C. 1200 bits/sec,. to Bell type 202 data set
D. Optional 2400 bit/sec,. interface to the Computek


realize these objectives, laboratory equipment has
been designed and built for appropriate experi-
ments some of which involve extensive computa-
tion. Efficient techniques have been implemented
to acquire data, put it in the form having the
greatest significance, and present the information
to the student during the laboratory period.

Is it true that you have a tutoring service for under-
graduates via the telephone?
Yes we do. Professor Rase realized several
years ago that an undergraduate student who has
difficulty in solving a homework problem often
can resolve the difficulty with the help of a key
suggestion. The telephone tutoring service enables
the student to call a phone number at any hour
of the day or night and receive hints that assist
in the solution of homework assignments. Indivi-
dual tutorial assistance in person or by phone can
be routinely obtained from tutors on duty Monday
through Friday from 3 P.M. to 5 P.M. Based
on the questions asked each day, the tutors record
or tape hints and suggestions for these homework
problems that seem to cause students difficulty.
The recorded material is activated when a student
dials the proper number. If the student does not
understand the information presented on the first
reading, he can hang up and dial again. Another
call re-initiates the tape.

What is the student evaluation of the program?
Almost every student who has used it believes
it is very helpful, especially when credit on a
homework problem is important.

What improvements are planned for the system in
the future?
We are working on a way to computerize the
system so that after the student dials the main
number, he can dial an extension corresponding
to the course number and receive a message for
that course without having to listen to the whole

Do you use the Personalized System of Instruction
to any extent?
The College of Engineering at the University
of Texas at Austin has been very active in de-
veloping applications of the Keller Plan, also
known as the Personalized System of Instruction
(PSI). PSI is distinguished by an initial careful
analysis by the teacher of what the students are



to learn in his course. Once the terminal and
intermediate objectives are fixed, he then divides
the course into units, each containing a reading
assignment, study questions, collateral references,
study problems and any necessary introductory
or explanatory material. The student studies the
units sequentially at the rate, time and place he
prefers. When he feels that he has completely
mastered the material for a given unit, a proctor
gives him a "readiness test" to determine if he
is ready to pass on to the next unit. The proctor
is a student who has been carefully chosen for
his mastery of the course material. The student
must make a grade of 100 on the "readiness test,"
and if he does not, he is told to restudy the unit
more thoroughly. He receives a different test form
each time he appears to be tested. All students
who demonstrate mastery of all course units re-
ceive a grade of A.
Lectures are given at stated intervals during
the course to students who have completed a speci-
fied number of units and can therefore understand
the material to be covered. The students who
qualify for a lecture are not required to attend
them, and the lecture material is not covered on
any examination.

Is PSI an improvement over the lecture-recitation
system of instruction?
The Keller Plan has worked well for us. Stu-
dents like the flexibility that it gives them. Most
(80 to 90 percent) prefer it to the more conven-
tional methods of instruction; they report that it
taught them how to study, and our results indi-
cate that they learn more and learn it better.
Next year there are plans to develop eight
new PSI courses in the College of Engineering,
including Material and Energy Balances and Pro-


D. M. Himmelblau with Data Processing Equipment

cess Analysis and Simulation in the Chemical
Engineering Department.
There are some problems, of course. Students
who are poorly prepared are often forced to drop
a course (they probably would have flunked in a
conventional course), and those who haven't de-
veloped self-discipline are prone to procrastinate.
And very little data are available on whether the
PSI student's better grasp of course materials at
the end of a course results in better long-term
mastery. We hope that our continuing study of
the method will throw some light on these matters.

Do you have a big graduate program?
We have a relatively large graduate study
body, with 50 to 60 full time students being en-
rolled each year. About one-half are working to-
ward the doctoral degree and the other half to
ward the M.S. In the last ten years 72 students
completed their Ph.D. degrees in our department.

What do you do in your M.S. program?
In the future we actually will have two types
of M.S. programs, one with a thesis and the other
without; the latter program was just introduced.
A student can complete the requirements for the
M.S. in one year but the average student takes
about a year plus a semester. A wide variety of
courses are available both within and outside of
the Ch.E. department. We do not encourage stu-
dents to specialize in any one area of course work
but instead try to provide them with as broad a
background as possible so that their future job
potential will not be too constrained. The research
completed for the M.S. degree is not nearly as
extensive as that for the Ph.D.

How does your Ph.D. program operate?
One of the main goals of our doctoral program,
as in other universities, it to provide a student
with the opportunity to carry out an independent
study under the general supervision of an expert,
his professor. Our faculty take pride in their close
personal contact with their graduate students and
we feel that this interaction is a big contributor
to high student morale. As with the M.S. students,
doctoral students are not expected to limit their
course work solely to the area of their research,
and are encouraged to broaden their interests. We
require no specific count of courses or hours com-
pleted for a Ph.D. so that the doctoral program
is completely flexible and can nicely fit in with
the student's needs and interests. We do re-
quire successful completion of preliminary written
qualifying examinations and a foreign language
test of some type before a student can become a
doctoral candidate. However, the bulk of the stu-
dent's time, beyond the M.S. requirements, is
spent in his research.

What types of research programs are underway?
We have a full time faculty of 14 and conse-
quently have a broad spectrum of on-going re-
search projects. Because of proximity of the Uni-
versity to the vast chemical and petroleum de-
velopments on the Gulf Coast of Texas, an active
interest is maintained in the fundamental prob-
lems of a wide variety of process plants. The
Department has earned a reputation for excellence
in research in materials, separation processes,
polymers, fluid properties, surface and aerosol
physics, catalysis and kinetics, automatic control,
process simulation and optimization, and biomedi-
cal engineering. A summary of the individual
faculty and their research interests is as follows:

James R. Brock
Aerosol physics and chemistry, nucleation processes, co-
agulation and condensation, deposition and filtration of
particles, particle charging, the surface properties of
particles, and quantitative human ecology as related to
environmental problems of air and water pollution.
William F. Bradley
Crystal structure analysis and crystal chemistry, in par-
ticular studies of natural fine-grained mineral and the
associations of organic matter with fine-grained minerals.
Tom F. Edgar
Process identification, optimization and control, energy
systems engineering.


D. R. Paul (right) with Carl Locke in polymer lab

Robert P. Popovich
Physiological transport parameters in a patient's artificial
kidney system; extracorporeal treatment of blood to
alleviate major disease states; enzymatatic detoxification
of endogeous and exhogeneous toxins (artificial liver);
pathogenesis and treatment of atheroscelerosis.

David M. Himmelblau
Process analysis and simulation, and system analysis, opti-
mization, stochastic modeling.

Joel 0. Hougen
Process dynamics and control system design.

John J. McKetta, Jr.
Phase equilibria, thermodynamic properties and high
pressure P-V-T.

Donald R. Paul
Polymer physics and chemistry. In particular the thermo-
dynamics and transport properties of polymers in bulk
and solution with applications to polymer processing.
Howard F. Rase
Reaction kinetics and catalysis with particular emphasis
on catalyst geometry and specificity, new catalyst develop-
ment and also enzyme model catalysts, process and chemi-
cal reactor design techniques.

Robert S. Schechter
Surface transport phenomena, surface viscosity, elasticity
and diffusion, hydrodynamic stability, and the application
of acids in oil reservoirs.

Hugo Steinfink
Crystal chemistry, magnetic, electrical, and optical prop-
erties of rare earth compounds, crystal chemistry of
silicate-organic complexes.

James E. Stice
/- Computer assisted instruction, research in engineering
| teaching.

'I Matthew Van Winkle
Azeotropic and extractive distillation, solvent requirements
and prediction of vapor liquid equilibria for multicom-
ponent solvent-containing systems. Studies of efficiency
in perforated tray columns including entrainment and
frothing characteristics related to efficiency.

Eugene H. Wissler
Properties of aerosol particles, aerosol beams, simulation
of the human thermal system, and non-Newtonian fluids.

aff e(*A

..- \


Members of the Chemical Engineering faculty
are not only recognized leaders in their individual
areas of expertise, but have been called on to
take positions of leadership in University ad-
ministrative functions, as well as in the AIChE
and other professional organizations, in civic,
and other activities of a broad nature. Dr. Mc-
Ketta is a member of the National Academy of
Engineering. -




Mathematical Methods in
Chemical Engineering,
Volume II, First-Order
Partial Differential
Equations with Applications
both of the University of
An extended treatment of
first-order equations that
includes a wide variety of
applications, particularly for
those who are not
mathematicians. Examples are
drawn from a considerable
range of scientific and
engineering disciplines. Topics
covered include: the
characteristics and craft of
constructing models dominated
by convective phenomena; the
differences between linear,
quasi-linear, and nonlinear
equations; the traditional
methods of the Laplace
transform and the connection
with the calculus of variations;
some of the ideas of weak
solutions; an introduction into
the work of Lax and others;
the equations of
chromatography; reducible
systems; and wave and
discontinuous solutions.
January 1973, approx. 416
pp., (013-561092-3), $16.50

For further information on any
of these titles, please write:
Robert Jordan
Dept. J-579
College Division
Prentice-Hall, Inc.
Englewood Cliffs
New Jersey

Chemical Engineering


Dynamic Behavior of
University of Rochester
Provides a complete and
systematic study of the
analysis of unsteady state
phenomena. Emphasizes
physical interpretation of
dynamic linear and nonlinear
responses as well as methods
of analysis. Covers both
distributed parameter and
lumped systems. Classes of
system models prepare reader
to generalize readily to other
systems and to diverse fields of
application. Over 100 detailed
examples illustrate the
fundamentals and varieties of
applications. Over 200 figures
present results graphically.
Appendices include vector and
matrix manipulation and
Laplace transform pairs for
easy reference.
1972, 590 pp.,
(013-221242-0), $18.95

Electrochemical Systems
JOHN NEWMAN, University
of California, Berkeley
Unified framework for the
analysis of problems in
electrochemical systems.
Clearly develops the
fundamentals of
thermodynamics, electrode
kinetics, and charge and mass
transport, with a view toward
their application to a wide
variety of electrochemical
engineering problems.
Emphasizes aspects which can
be treated quantitatively,
especially problems of scale up.
Introduces the beginning
student to the development,
design, and operation of
electrochemical syntheses,
processes, energy conversion
and storage devices, and
corrosion. Problems at the end
of each chapter clarify and
extend the theories developed.
Numerous examples are
worked out in the text.
January 1973, approx. 440
pp., (013-248922-8), $18.95



I Nh

views and opinions


University of Colorado
Boulder, Colo. 80302

Too often faculty members neglect students'
"non-academic deficiencies" in designing their
freshman engineering courses. Unfortunately,
these problems can be bigger "roadblocks" than
any academic deficiency. Adjustment to the Uni-
versity environment and learning the "rules of
the game" is more important than any basic
course. Thus, a department interested in success
of its students would be well advised to devote
some effort to helping students solve their ad-
justment problems. This effort, coupled with the
difficult problem of what to present to freshman
chemical engineers gave rise to a novel course
at the University of Colorado.
The course was designed (with a little help
from his friends) and is taught by Dr. R. Curtis
Johnson, Department Chairman. It makes use of
undergraduate teaching assistants and the entire
department faculty. The purposes of the course
are to answer students "quasi-academic" ques-
tions about the University and the Department,
to provide fundamentals useful in other courses,
and to identify other student problems, academic
and non-academic, so that appropriate solutions
may be sought.
The "medium" of the course is environmental
awareness. The course tries to emphasize the
relationship of the engineer to society. The text,
Population/Resources/Environment, by Paul R.
and Anne H. Ehrlich is supplemented by re-
prints from Chemical and Engineering News.
The topics are loosely defined to encourage the
students to discuss the topics with the senior
assistants, the faculty members and among them-
selves. Interaction and discussion rather than
presentation of specific material is stressed.
The course carries two credit hours and
meets once per week for two hours. The first
hour is devoted primarily to lectures and testing.
During the second hour the class is split into
groups of about ten. One senior assistant is re-
sponsible for each group. The time is spent dis-

cussing the topic for the week with a member
of the faculty. Each week the group discusses
a different topic with a different member of
the faculty. This affords each student an oppor-
tunity to meet every faculty member on a per-
sonal basis. Table 1 presents the activities of the
classes taught in the fall of 1971.
It is important to all freshman chemical
engineers to find out what a chemical engineer
really does. This information is communicated
in many ways. The AIChE film The Chemical
Engineer is an excellent primer. Two field trips
give the students a better idea. One trip to a
research facility, such as the National Bureau
of Standards, and one to an operating plant are
usually scheduled. Of course, contact with the
faculty, older students and the regular student
AIChE chapter speakers aid in the communica-
tion process.
Insuring that students are familiar with the
college, its facilities, and the services it offers
aids the adjustment process. Brief tours are
conducted so that students can become familiar
with important facilities such as the Chemical
Engineering Laboratory, the Computing Center,
the Engineering Library and the Administrative
Offices. The tours are designed to be instruc-
tional. University personnel are on hand to see
that the students are familiar with the available
facilities. In addition, student leaders are invited
to speak to the class to inform the students of
services provided by the student government
and to encourage them to become more involved.
Setting up appropriate counseling sessions is
facilitated by having all the freshmen in one
class. Little counseling is done in class, but aca-
demic counseling sessions are scheduled with the
department chairman. Senior students are given
a class period to present a student viewpoint of
the curriculum and the faculty. Additional coun-
seling sessions are scheduled with faculty, stu-
dent, or administrative counselors for freshmen
who so desire.
In addition to a general environmental aware-
ness, the course tries to impart some basic skills


Adjustments to the University environment and
learning the "rules of the game" are more important
than any basic course.

TABLE 1 Activity Charts for ChE 130, 1971.


David L. Hunter is presently "taking it easy" in
Boulder, Colorado. He earned BS degrees in ChE and
Business, University of Colorado, 1972 and hopes to at-
tend law school and work in environmental law.

that will be valuable in both future courses and
other endeavors. Basic sessions are conducted in
laboratory safety and fire fighting. Extensive
instruction is given in the use of the slide rule.
Basic math relationships and flow diagrams are
covered in homework assignments and exams.
Each student is required to prepare a paper
on a topic of his choice in the environmental
area. If the student wishes to continue his re-
search on the project he can obtain one hour of
independent study credit by expanding the
project during the following semester.
The response of freshmen at the end of the
semester was somewhat mixed. Most freshmen
were probably not able to appreciate the unique-
ness of the course. The course and the professor
received high ratings in the formal student
evaluation of teaching that is conducted each
semester. The course seemed to reach its target
students. Most of the students who were en-
thusiastic about the course were, in the opinion
of the senior assistants, students that were
having scholastic difficulty. Some students who
are doing well in other courses felt that they
wanted more technical material.
Students who had taken the same course
previously had much higher opinions of the
course. The "non-academic" aspects of the course
were particularly appreciated.
The senior assistants who have participated
in the program during its three years of exist-
ence are unqualifiedly enthusiastic about the
course, from both the standpoint of their own
experience and what they feel the freshmen have

1st Hour
All Groups
R. C. Johnson

2 Reading Test
College Counselor
3 Reading Results
and talks by
Student AIChE
4 Safety lecture and

2nd Hour
Small Groups
Photos and Introduction

Tours of ChE Lab
Eng. Library
Tours of Numerical
Analysis Center &
Computer facilities

Safety lecture
and Demonstration

5 Slide Rule Instruc- Topic: Pesticides
tion, J. F. Carley Paul Barrick

6 Slide Rule

7 Slide Rule

8 Slide Rule

9 Slide Rule
10 Pre-advising
R. C. Johnson

Topic: Chemical Mutagens
Igor Gamow
Topic: Population
W. B. Krantz
Topic: Air-Water Pollution
Frank Kreith
Exam Followup
Senior Assistants
Topic: Environmental
L. F. Brown

11 Academic Advising Topic: Colo. Environment,
by Student Advisors M. S. Peters

12 Slide Rule
Make-up Exam

Topic: Mathematic
W. F. Ramirez

13 Cryogenics, Topic: Peter Principle,
K. D. Timmerhaus G. J. Maler

14 Field Trip:

Field Trip

National Bureau of Standards

Final Exam

Final Exam

Supplemental Readings
1 ) Report of the Secretary's Commission on Pesticides
and their relationship to Environmental Health, U.S.
Department of Health, Education, and Welfare,
Washington, D. C., December, 1969.
2) Sanders, Howard J., Chemical and Engineering News,
"Chemical Mutagens, The Road to Genetic Disaster?",
Vol. 47, May 19, 1969.
(Continued on page 17)



4 She'd Gewae



University of Houston
Houston, Texas 77004

Chemical Engineering Departments provide a
useful service to the industrial community by
offering short courses on topics of current inter-
est. Many universities have found this to be so
profitable a utilization of faculty and facilities
that available courses proliferate during the sum-
mer months. The most valuable courses are those
which are tailor-made to suit the needs of po-
tential attendees and not the predilections of the
available faculty.
The Chemical Engineering Department of the
University of Houston organized two short course
on the applications of heterogeneous catalysis
during the summers of 1971 and 1972. The pro-
gram was so successful that essentially the same
course is being repeated this year. Other short
courses in catalysis have been presented before
at Rice University, Lehigh, Berkeley and Dela-
ware. However, our program is sufficiently dif-
ferent from these others so that it is worthwhile
to outline the philosophy, organization and logis-
tics of the course.
There are two types of practitioners in cataly-
sis-the researchers and the users. A short course
should be organized from the beginning with one
or the other of these groups in mind. This course
was aimed at the wide audience of users. These
individuals have usually a considerable practical
experience in a relatively narrow area but do not
have a systematic and comprehensive coverage of
the theory of catalysis and catalytic reactor de-
sign. Moreover they usually have little knowledge
of the principles and techniques used by others.
We decided upon a "how to" theme-how to
select, prepare, characterize, test and use a cata-
lyst. The program was built around these five
points with emphasis on practical implications
and applications of current theory and know how
and not on building an in-depth understanding
of the latest research developments.
In order to present a comprehensive coverage
of the subject and to expose the students to a

variety of viewpoints we selected as instructors
two industrial and three academic experts. Each
lecturer was instructed on the material to be in-
cluded in his section in order to avoid gaps in the
coverage and to prevent repetitious discussions
of certain material. A comprehensive set of notes
for classroom use was collected from each. These
were reproduced, bound and distributed to each
attendee before the start of the course. The class
schedule for 1972 is shown in Table 1.

(Summer 1972)
1. Catalyst Selection (0.5 day) J. T. Richardson
A. Introduction to catalysis
B. Theories of catalysis
C. Prediction of catalytic properties
D. Catalytic process development
2. Catalyst Preparation and applications (1 day) C. L.
Thomas Sun Oil Company (retired)
A. General principles
B. Preparation and use of non-supported catalysts
C. Preparation of catalyst supports
D. Preparation of supported catalysts
E. Proven catalysts and catalytic processes
3. Catalyst Characterization (0.5 day) J. T. Richardson
A. Common mechanical properties
B. Measurements of surface area, pore size distribu-
tion, acidity, metal surface area, effective diffusi-
vity and dispersion
C. Survey of instrumental techniques
4. Kinetics, Mechanisms and Experimental Reactors (0.5
day) J. W. Hightower Rice University
A. Theory of adsorption
B. Langmuir Hinshelwood kinetics
C. Discussion of some examples
D. Comparisons of various experimental reactors
E. Practical suggestions
5. Diffusional Effects (0.5 day) D. Luss
A. Isothermal inter and intra-particle diffusional dis-
guise of activity and kinetics
B. Effect of external and internal temperature gradi-
C. Diffusional disguise of selectivity
D. Experimental methods of determining these effects
6. Poisoning of Catalysts (0.5 day) J. T. Richardson and
D. Luss
A. Mechanisms of sintering, poisoning, and fouling
B. Determination of mechanisms by single pellet


James T. Richardson received his PhD ('56) degree
in solid state physics from Rice University. Following a
year of postdoctoral research in low temperature physics,
he joined Humble Oil and Refining Company (later Esso
Research and Engineering) at Baytown Texas. During
the next thirteen, years, his research activities included
all phases of fundamental and applied heterogeneous
catalysis, with emphasis on correlation between catalyst
properties and kinetics. In 1969 he joined the ChE De-
partment of the University of Houston. Dr. Richardson
is one of the organizers and a past chairman of the
Southwest Catalyst Society. (left photo)
Dan Luss received his BS and MS from the Technion-
Israel and his PhD from the University of Minnesota.
His areas of specialization are analysis and control of
chemical reactors and heat and mass diffusional intrusion
in catalytic reactor behavior. He received the Allan P.
Colburn award of the AIChE in 1972.

C. Effects of diffusion on poisoning and regeneration
D. Shell progressive non-catalytic gas solid reactions
7. Catalytic Emission Control Devices for Automobiles
(0.5 day) J. W. Hightower Rice University
A. Federal standards and testing procedures
B. Engine parametric effects on emissions
C. Chemistry of control systems
D. Choice of possible catalysts
E. Cold start up, poisoning and melting problems
8. Aspects of Catalytic Reactor Engineering (0.5 day)
D. Luss
A. Effect of poisoning on operation of catalytic re-
B. Regeneration of cokes packed reactors
C. Design aspects of exothermic reactors
D. Sensitivity and stability
9. Industrial Applications (0.5 day) V. W. Weekman, Jr.
Mobil Research and Development Company
A. Modelling techniques of complex processes
B. Modelling of catalytic cracking kinetics
C. Modelling of a regenerator for catalytic cracking
D. Modelling of catalytic mufflers

An important part of organizing such a course
is to attract a sufficient number of attendees. The
publicity for our course consisted primarily of a
brochure, outlining the aims and details of the

course, which was mailed to local members of
the ACS and AIChE, and to members of the
Catalysis Society of North America. This wais
followed by individual letters to various indus-
trial research organizations, catalyst manufac-
turers, etc. This procedure was found to be very
effective and in 1972 we had to reject 21 appli-
cants due to restrictions on the maximal size of
the class (40 students).
The students, representatives of a wide range
of industries, responded most favorably to this
format. Classes were informal with valuable con-
tributions being made by many of the group.
Using suggestions of the 1971 participants, we
expanded in 1972 the discussions of proven cata-
lysts and methods of catalyst preparation as well
as of problems related to the design of catalytic
mufflers for automobiles.
Short courses similar to this may, no doubt,
be organized in many other fields of chemical
engineering. The needs of industry for further
education are continually changing. Courses of
this nature help to fill the gap between formal
university curricula and the demands of our pro-
fession. In addition, exposure to groups such as
this is a rewarding and revitalizing experience
for any faculty member. L

HUNTER: (Continued from page 15)
3) Kiefer, David M., Chemical and Engineering News,
"Population," Vol. 46, October 14, 1968.
4) Levitt, Arnold, Chemical and Engineering News,
"Pollution; Causes, Costs, Controls," Vol. 47, June 9,
5) Cleaning Our Environment (Supplement), "The
Chemical Basis for Action," American Chemical
Society, Washington, D. C., 1971.
6) "Second Interim Report," Colorado Environmental
Commission, December, 1971.
7) Forrester, Jay W., Technology Review, "Counter-
intuitive Behavior of Social Systems," Vol. 73, Num-
ber 3, January, 1971.
8) Peter, Laurence J., and Hull Raymond, The Peter
Principle, Bantam Books, New York, N. Y. 4th
Printing, 1970.


I would like to acknowledge the work and
suggestions of the other twelve senior assistants
during the past three years; Don Hedden, Craig
Farmer, Bob Powell, John McMartin, Jim Rob-
erts, Craig Miller, Ed Heubach, Bill Barker,
Chris Harris, Craig Runyan, John Hayes and
Steve Hammill. OD


~Ici~r aluum


Georgia Institute of Technology
Atlanta, Georgia 30332

FOR MANY YEARS now industry has found
the Critical Path Method (CPM) to be a con-
siderable aid in planning and scheduling activities.
This method determines the shortest period of
time required for the completion of a system of
activities in the following manner. Each activity
is listed with (1) its activity number, (2) its time
required for completion (activity duration), (3)
those activities which must be completed before
starting the activity in question (preceding activi-
ties), (4) those activities which cannot be started
before the completion of the activity in question
(succeeding activities), and (5) any other con-
straints which are found necessary to be placed
on the particular activity, e.g., activity start or
finish times.
Once the activities have been listed with this
information, the activities may be graphed by
activity as in Figure 1 wherein the ordinate has
some arbitrary units of time. Those activities
having no preceding activities are listed by ac-
tivity number at the extreme left and the succeed-
ing activities are placed on the graph further to
the right according to the time scale. Preceding
and succeeding activities are connected with lines
to show their relationships with other activities.
The time element should never be forgotten with
respect to activity duration and specified activity
start or finish times.
Once all of the activities have been included
on this graph, the activity finding itself to the
extreme right is the last activity in a series of
activities which define the critical path. The re-
maining critical path activities can be found by
following the lines connecting activity to activity
which exhibit no time lag (slack) from the finish
of one activity to the start of the next. When this
procedure has been completed, the critical path
has been defined. If the system of activities is to
be completed in the shortest possible time, these
critical path activities must be completed in a
continuous manner, one right after another with
no interruption.

1 2 3 4
Fig. 1. Sketch of a CPM diagram. Critical path activities (101,
102, 201, 302, 401) are connected by heavy lines.

Now that the critical path has been delineated
all the activities can be listed with their earliest
start times, latest finish times and slacks such
that there will be no increase in the critical path
and therefore no increase in the time necessary
to complete the entire system of activities. This
list of earliest start times, latest finish times and
slack offers a framework around which a schedule
can be built.
With the advent of computers, programs have
been written which can aid in scheduling a system
of activities by using CPM. The necessary infor-
mation is the same as that which is required to
complete the CPM diagram as described above,
but the time and effort required to obtain the
necessary earliest start time and latest finish time
for each activity is greatly reduced.
laid, it is time to introduce the primary pur-
pose of this article, namely, CPM as an aid in
formulating curricula. The question might be
asked why such a method is necessary; it can be
seen today that universities are changing their
curricula at an ever increasing rate, and at the
same time complexities are being introduced when
departments offer new options to better meet the
personal needs and desires of individual students.
Keeping this in mind, a tool such as CPM could
greatly alleviate much of the toil involved with
adapting curricula.


Robert C. Cunningham is a field service engineer with
the Babcock & Wilcox Co. and is based in their Denver
office. He received his BSChE from Georgia Tech in 1972.
He had co-op experience with the Savannah River Plant
of the AEC in Aiken, S. C. The work described in this
paper was performed as an undergraduate special project.
Jude T. Sommerfeld received his BSChE from the Uni-
versity of Detroit and MS and PhD from the University
of Michigan. His activities include teaching, research and
consulting in the areas of reactor design and computer
applications. He has had 10 years of engineering and
management experience with BASF-Wyandotte Corp.,
Monsanto Co., Parke, Davis & Co. and Ethyl Corp., and
is a member of AIChE, ACS, and ISA. He is also a
registered professional engineer in Georgia. (left photo)

The method for reducing curriculum informa-
tion as it appears in a university catalog so that
it can be accepted by a CPM computer program
is quite easy. The individual courses are assigned
activity numbers by which the computer can rec-
ognize them. An activity duration whether on a
semester, quarter or any other time basis is given
to each course. The prerequisite courses fill the

preceding activities category and the courses for
which this particular course is a prerequiste fall
into the succeeding activities category. If a certain
course is required to be taken in a specified year
of study or before a specified time in a program
of study, the constraints can be accepted by the
CPM program by listing a start time or finish
time for that course.


04010402 0403

ORUl 0002




1 2 3 4 5 6 7 8 9 111

Fig. 2. CPM diagram of chemical engineering curriculum. Critical path activities are connected by heavy lines.



With this information the CPM program de- The CPM diagram is helpful in testing time feasibility

termines an earliest start time, latest finish time of proposed curricula and in offering a model which

and slack for each course in the particular cur- can answer some of the questions dealing with

riculum. All of the courses which comprise the curricula practicality.

critical path are fixed as to when they must be

Table I. Accelerated Chemical Engineering Curriculum (Eleven Quarters).

Qtr. Activity Cr. Prereq. Earliest latest
No. Number Course Name lHrs. Activities Start Start Slack

1 501 Calculus I
1 301 General Chemistry I
1 101 Humanities I
1 1101 Physical Training I
1 1001 Eng. Graphics I
1 601 Chem. Eng. Orientation'

2 502 Calculus II
2 302 General Chemistry II
2 201 Social Science I
2 102 Humanilties II
2 1102 Physical Training II
2 1002 Eng. Graphics II

3 503 Calculus III
3 303 General Chemistry III
3 602 Material Balances
3 202 Social Science II
3 103 Humanities III
3 1103 Physical Training III

4 504 Calculus IV
4 401 Physics I
4 603 Energy Balances
4 203 Social Science III
4 104 Humanities IV
4 1104 Physical Training IV

5 505 Calculus V
5 402 Physics II
5 801 Statics
5 1201 Elective I
5 105 Humanities V
5 1105 Physical Training V

6 506 Differential Equations
6 307 Physical Chemistry I
6 604 Computers in Chem. Eng.
6 403 Physics III
6 106 Humanities VI
6 1106 Physical Training VL

7 310 Physical Chemistry Lab.
7 308 Physical Chemistry II
7 605 Transport Phenomena 1
7 304 Organic Chemistry C
7 701 Electric Fields-Circuits
7 802 Material Science

8 309 Physical Chemistry III
8 608 Stagewise Operations
8 606 Transport Phenomena II
8 305 organicc Chemistry II
8 306 Organic Chemistry Lab
8 1202 Elective II

9 616 Thermodynamics I
9 607 Momentum-Heat Transfer
9 613 Chemical Kinetics
9 610 Chem. Eng. Literature
9 702 Elem. Electronics
9 1203 Elective III
9 204 Social Science IV

10 617 Thermodynamics II
10 614 Chem. aEng. Economics
10 611 Pro cess Instrumentation
10 609 Mass Transfer
10 1204 Elective IV
10 205 Social Science V

11 618 Comprehensive Prob.
11 615 Chemical Plant Design

11 612 Polymer Science
11 901 General Metallurgy
11 1205 Elective V
11 206 Social Science VI

5 1
5 I
3 1
1 1
3 1


5 501 2
5 301 2
3 1
3 101 2
1 1101 2
3 1001 2

5 502 3
4 302 3
3 302,502 3
3 201 2
3 102 3
_! 1102 3

5 503 4
5 502 3
3 602 4
3 202 3
3 103 4
_1 1103 4

5 504 5
5 401 4
3 401,504 5
3 1
3 104 5
.1 1104 5

5 505 6
3 303,402,505 6
3 505,602 6
5 402 5
3 105 6
1 1105 6

3 307 7
3 307 7
4 506,604 7
3 303 4
3 402,506 7
3 505,801 6

3 308 8
4 605 8
4 605 8
3 304 5
2 304 5
3 1201 2

3 309,608 9
3 606 9
3 305,309,608 9
1 304,307,605 8
3 701 8
3 1202 3
3 203 4

3 616 10
3 608 9
3 606,702 9
4 606 9
3 1203 4
3 204 5

3 613,614,617 11
3 609,611,613 11
3 305,309, 608 9
3 309 9
3 1204 5
3 205 b

1 0
3 2
6 5
6 5
10 9
11 10

2 0
4 2
6 5
7 5
7 5
11 9

3 0
5 2
5 2
7 5
8 5
8 5

4 0
4 1
11 7
8 5
9 5
9 5

5 0
5 1
9 4
7 6
10 5
10 5

6 0
6 0
6 0
11 6
11 5
11 5

11 4
7 0
7 0
8 4
8 1
10 4

8 0
8 0
9 1
9 4
11 6
8 6

9 0
11 2
10 1
11 3
9 1
9 6
9 5

10 0
10 1
10 1
10 1
10 6
10 5

11 0
11 0

11 2
11 2
11 b
11 5

Table II. Typical Chemical Engineering Curriculum (Twelve Quarters).

Qtr. Activity Cr. Prereq. Earliest Latest
No. Number Course Name Hrs. Activities Start Start Slack

1 501 Calculus I
1 301 General Chemistry I
1 101 Humanities I
1 1101 Physical Training I
1 601 Chem. Eng. Orientation
1 1201 Elective I

2 502 Calculus II
2 302 General Chemistry II
2 102 Humanities II
2 1001 Eng. Graphics I
2 1102 Physical Training II

3 503 Calculus III
3 303 General Chemistry III
3 103 Humanities III
3 1002 Eng. Graphics II
3 1103 Physical Training III
3 1202 Elective II

4 504 Calculus IV
4 401 Physics I
4 602 Material Balances
4 104 Humanities IV
4 1104 Physical Training IV

5 505 Calculus V
5 402 Physics II-
5 603 Energy Balances
5 105 Humanities V
5 1105 Physical Training V

6 506 Differential Equations
6 403 Physics III
6 604 Computers in Chem. Eag.
6 106 Humanities VI
6 1106 Physical Training VI

7 605 Transport Phenomena I
7 304 Organic Chemistry I
7 307 Physical Chemistry I
7 801 Statics
7 201 Social Science I
7 1203 Electives III

8 606 Transport Phenomena II
8 305 Organic Chemistry II
8 308 Physical Chemistry II
8 310 Physical Chemistry Lab
8 802 Material Science
8 202 Social Science II

9 607 Momentum-Heat Transfer
9 608 Stagewise Operations
9 306 Organic Chemistry Lab
9 309 Physical Chemistry III
9 203 Social Science III
9 701 Elec. Fields Circuits

10 616 Thermodynamics I
10 612 Polymer Science
10 609 Mass Transfer
10 610 Chem.Eng. Literature
10 702 Elem. Electronics
10 204 Social. Science IV

11 617 Thermodynamics II
11 613 Chemical Kinetics
11 611 Process Instrumentation
11 614 Chem.Eng. Economics
11 1204 Elective IV
11 205 Social Science V

12 618 Comprehensive Prob.

12 615 Chemical Plant Design
12 901 General Metallurgy
12 206 Social Science VI
12 1205 Elective V

5 1
5 1
3 1
1 1
S 1


5 501 2
5 301 2
3 101 2
3 1
I1 1101 2

5 502 3
4 302 3
3 102 3
3 1001 2
1 1102 3
-3 1201 2

5 503 4
5 502 3
3 302,502 3
3 103 4
_ 1103 4

5 504 5
5 401 4
3 602 4
3 104 5
1 1104 5

5 505 6
5 502 5
3 505,602 6
3 105 6
_! 1105 6

4 506,604 7
3 303 4
3 303,402,505 6
3 401,504 5
3 1
3 1202 3

4 605 8
3 304 5
3 307 7
2 307 7
3 505,801 6
.3 201 2

3 606 9
4 605 8
2 304 5
3 308 8
3 202 3
_3 402,506 7

3 309,608 9
3 305,309,608 9
3 606 9
1 304,307,605 8
3 701 8
3 203 4

3 616 10
3 305,309,608 9
3 606,702 9
3 608 9
3 1203 4
3 204 5

3 613,614,617 11
3 614,616,802 11
3 309 9
3 205 6
3 1204 5

2 1
4 3
7 6
7 6
12 11
8 7

3 1
5 3
8 6
11 10
8 6

4 1
6 3
9 6
12 10
9 6
9 7

5 1
5 2
6 3
10 6
10 6

6 1
6 2
12 8
11 6
11 6

7 1
12 7
7 1
12 6
12 6

8 1
9 5
7 1
10 5
7 6
10 7

10 2
10 5
8 1
12 5
11 5
8 6

12 3
9 1
12 7
9 1
9 6
9 2

10 1
12 3
11 2
12 4
10 2
10 6

11 1
11 2
11 2
11 2
11 7
11 6

12 1

12 1
12 3
12 6
12 7


started or completed; however, the remainder of
the courses can be taken at any time within the
constraints of the earliest start and latest finish
times. In order to assist in selecting definite start
dates for these courses which are not part of the
critical path, various programs, e.g., RAMPS (Re-
source Allocation and Multiple Project Schedul-
ing) *, can be used to devise the final schedule.
These programs distribute the various courses
within the limits of their earliest start and latest
finish times such that there will be a nearly con-
stant number of credit hours per school term
throughout the entire program of study.
A typical undergraduate chemical engineering
curriculum is used as an example to illustrate the
concept of CPM as a scheduling aid. Figure 2 is
a CPM diagram of this curriculum wherein the
activity numbers represent various courses (the
two activities at the extreme left and the two
activities at the extreme right, 0001, 0002, and
3002, are dummy activities and are included only
to facilitate the operation of the particular CPM
program which was used). This figure illustrates
the role prerequisites play in scheduling courses.
Table I shows the final schedule which was
derived from the CPM program. The earliest start
times, latest finish times and slacks were provided
by the CPM program enabling the critical path
to be defined and also giving the shortest time for
completion of the entire program of study. The
courses which do not lie on the critical path and
are accompanied by slacks were distributed manu-
ally, and the final product is a schedule which
breaks the individual courses down into school
The actual schedule for this curriculum as it
might have appeared in a university catalog is
listed in Table II and consists of twelve quarters.
The information obtained from the CPM program
showed that the curriculum could be completed
in an accelerated program of eleven quarters in
which no more than 20 credit hours are required
per quarter.
Referring once again to Figure 2 it is seen
that the courses have been distributed among the
eleven quarters of the accelerated program as they
appear in Table I. This type of CPM diagram can
be helpful not only in testing feasibility of the
time considerations of proposed curricula but also
in offering a model which can answer some of the
questions dealing with curricula practicality. El
*Kurzeja, J. T., Hydrocarbon Processing, April, 1965,
p. 171.


The 1st edition of

Levenspiel was adopted

at 145 colleges and



there |

a second


Look for the changes that make the
2nd edition even better than the 1st-
160 brand new problems
completely new chapter on systems of deactivating
expanded and thoroughly rewritten chapter on flow
modeling and diagnosis of ills of operating equip-
new single design procedures for packed bed and
fluidized bed and reactors
Plus all the features that made
the 1st edition a success-
easy to understand, requiring no sophisticated math
all the most recent thought, not available anywhere
else in a textbook
stress on principles and decision-making
over 330 problems
extensive illustrations

by Octave Levenspiel, Oregon State University.
1972 578 pp $16.95

For a complimentary copy, contact your Wiley representative or,
write to T. R. Poston, Dept. 1182, N.Y. office. Please include
course title, enrollment and present text.
Prices subject to change without notice
605 Third Avenue, New York, N.Y. 10016 IIlIim
In Canada: 22 Worcester Road, Rexdale, Ontario iic edma



Activity, Yield and Non-isothermality
GERMAIN CASSIERE and intrinsic ral
University of Notre Dame Insofar
Notre Dame, Ind. 46556 multipathed
diffusion of
An essential component of any course in can be affe
Chemical Reaction Engineering is the Diffusion- desired pro(
Catalytic Reaction topic. In an introductory reactive coi
course, it is deemed imperative that the student ratio of twc
gain an insight into the key features which char-
acterize the behavior of a reaction network in Isothermal In
which diffusion of heat and/or mass affects the
observed, global rate of catalytic reaction. In sum, The mo
how does heat and mass transport intrusions the problem
affect: with both ex
For a flat p
activity for various reaction orders?
Yield/selectivity for multipathed reaction
Observed reaction order and activational with b.c.
energy? and finally
how can diffusional intrusions be detected the solution
in terms of laboratory observables? lytic effective
We acknowledge two regions of diffusional
transport in the typical porous solid catalyzed
reaction: is

1) Intraphase; diffusion of heat and mass
within the porous catalyst with simultane-
ous reaction.
2) Interphase: Diffusion of heat and mass
through boundary layers surrounding the
catalyst pellet or extrusion. Reaction occurs
in series with interphase (external) dif-
The global (observed) rate is that phrased in
terms of observables. In general, external surface
and internal concentrations and temperatures are
beyond direct measurement. So a global rate, Ro is
to l Xg(C)'
where kx is a rate coefficient based upon observed
rate, concentration Co and order a. The intrinsic,
surface catalytic rate is, of course, a complex
function of surface concentrations and tempera-
ture. Thus the concept of catalytic effectiveness
(Thiele) was invoked and elaborated to relate the

;e function to the observed functional-

as one or more steps of a complex,
, reaction network can be altered by
heat and/or mass, yield/selectivity
acted. By yield we mean the rate of
iuct formation relative to that of key
isumption. Selectivity is merely the
i-point yields.

iter-intraphase Effectiveness and Yield
dern undergraduate readily handles
Ls of isothermal first order reaction
eternal and internal diffusional events.
v d2C = koC
x = L C) = -D
x = 0 dC/dx = 0
to which is in terms of overall cata-

SkoCo (2)
"o o



k L
= L / -- = Tniele Modulus; (di) = Mass Blot .umber
The diffusion affected yield for consecutive re-
action is also readily analyzed by the undergradu-
ate for isothermal conditions.2,4
However, for nonlinear kinetics and non-iso-
thermal environments, analytical solutions are
not to be found, and the more sophisticated ap-
proximate mathematical treatments are usually
beyond the undergraduate, or if within his grasp,
the very journey through that mathematical
jungle may cause the student to lose sight of the
physics of the problem and the engineering conse-
quences. A simple alternative exists requiring no
more than the use of a slide rule, whereby key
features of the more complex internal diffusion-


catalyzed oxidation of CO at CO concentrations
above 1%.5 The solutions are:


n = 1/2

n- 2

n 1-+ Da

2 (Da +Da2)

[ 4D -1 2

J. J. Carberry was a native of Brooklyn, N.Y., and
graduated from University of Notre Dame (BS, MS'51),
and Yale University (PhD'57). His experience includes
several years as Senior Research Engineer at the duPont
Experimental Station, NSF Senior Fellow at Cambridge
University ('65-'66), Visiting Professor at University of
Naples ('71) and faculty member at Notre Dame since
1961. He received the 1968 Yale Engineering Association
Award for Advancement of Pure and Applied Sciences
and he is a member of the Working Committee on US-
Soviet Cooperation in Catalysis. (left photo)
Germain Cassiere is from Shreveport, La. and at-
tended the University of Notre Dame, graduating magna
cum laude with a BSCHE in May 1972. He is a member
of Tau Beta Pi and was employed during previous sum-
mers with the chemical process division of Universal Oil
Products in Shreveport. Presently, he is attending Louisi-
ana State University Medical School in New Orleans.

reaction problem become manifest. We simply
deal with the external (interphase) problem.

Isothermal External Diffusion-Reaction

In steady state reaction of nth order over a
flat nonporous catalytic plate, equating mass
transport to surface reaction, we have
k a(C C) = ko C (4)
Dividing by kgaCo, where f = C/Co

ko Cn-1 j
1 -f fu Daon.f

k C n-1
Duo = k Damkohler Number
External isothermal effectiveness is
Actual Surface rate at C ( )
Rate if C = (

k C n 0 o
Equation (5) is easily solved for n = 1, 1/2,
2 and -1. The negative first order reflects ab-
normal kinetics (i.e. rate increases with conver-
sion). Such a case is found in transition metal

S1-1 i -= 2 no solution exists (10)
1 + V 1 4 Da for Da > 1/4
These external isothermal effectiveness factors
are displayed in Figure 1 on log-log co-ordinates.
The similarity to internal effectiveness behavior
for linear and nonlinear kinetics is evident. We
learn by this simple arithmetic manipulation that
(a) The greater the reaction order, the great-
er the diffusional taxation for a given
value of Dao.
(b) An effectiveness greater than unity is
possible if an adsorbable species inhibits
the rate.
(c) At large values of the modulus, all 71
values approach the reciprocal of the
Now the global rate is, of course,
Ro = f koCon
At large values of Dao., n I/DaO

o *-- C kaC
o oCn-i

Fig. 1. Isothermal External Catalytic Effectiveness for
Reaction Order n


so the experimental rate coefficient k. is kga and
the reaction order changes from n to unity. Acti-
vational energy changes from its true value when
71 = 1 to a small value characteristic of a mass
transfer coefficient kga.
The students readily realize however that Fig.
1, while instructive, is of little use unless ko is
known a priori to permit computation of Dao.
They soon learn that Fig. 1 is easily re-expressed
in terms of observables. Dividing both sides of
equation (11) by kgaCo
'7 _o C n-1
k C ka o =rDa (12)
Hence by replotting Fig. 1 in terms of n vs
] Da., we secure Fig. 2 which permits determina-
tion of 77 from the observed rate Ro, bulk con-
centration Co and a calculatable transport co-
efficient kga. Fig. 2 also teaches that the ultimate
observed rate is that of bulk mass transport, i.e.,
at 71 Dao, = 1. Aris* points out that by eq'ns (7-10)
and (5) Figure 2 obeys the equation:
n = (1 + n Dao ).

Isothermal Yield/Selectivity

A multipathed network, such as

is obviously a combination of consecutive and
simultaneous reaction. Consider consecutive re-
A B- 4 G
In the absence of mass diffusional limitations,
surface concentrations equal those in the bulk
stream, so for linear kinetics
SR = kAo (13)
RB = kA k2o
Dividing, we secure the point yield
dB k2 Bo (14)
dA k I TA
Anticipating diffusion, then
k a(A A) = klA (15)

kea(B Bo) = k1A k2B (16)
and solving for B, then the ratio of rates (yield)
Solving for A in (15), substituting in (16)
do1 1 k2 (1 + Da) B7)
dA B + Da2 k (1 + Da) Ao
Y 2 k A(18)

*Aris, R. Personal Communication Sept. 1972.

C Dao
17 NO Ckga

Fig. 2. Isothermal External Catalytic Effectiveness in terms of
Observables for orders n
When Bo = 0, initial point yield is
SA 1 (19)
dA I + Da2
or selectivity is, when Bo = 0
a __1L = '_ (20)
dC Da2 k2
Thus with mass diffusion, the survival of B
depends upon the ratio of the escape rate from
the surface to its rate of destruction on the
For simultaneous reaction

A k I

t order

Selectivity, s = = =A-
With no diffusional gradient

S (kl/k2 I A
Isothermal selectivity alteration is then
sis = A
I -A )"

As A < Ao, we readily see that when A < Ao
(diffusion intrusion)
a = f no affect upon selectivity
a > 8 selectivity for B declines
a < 8 selectivity for B improves
This finding follows from Fig. 1: The reaction
of highest order is most taxed by mass diffusion.




Non-isothermal Effectiveness and Yield

Here the analysis in terms of external effec-
tiveness is truly fruitful since the corresponding
internal problem is only solved rigorously by
numerical means even for linear kinetics.6,7,8'9
The non-isothermal external effectiveness is
The linear case is quite easily handled in terms
of observables as follows:*
*In view of equations (5) and (25), the non-isothermal
external effectiveness for any order is
S- (1 I Da)

k the rate coefficient as the surface tempera-
ture is, relative to ko

k = k. exp 1 T (26)

where C = E/RT and t = T/T

also C = 1 and Da = Dao exp - 1
C 1 + Da


S+ Da exp[. E -ij

rearranging (1 1 Da.) exp - 1

Note that 1 =D. is the
0 g





We now evaluate t by heat balance
ha(T Tr) = (-Al) R,,
we divide both sides by kga Co To

go g 0
Invoking the j factor analogy
g (s)2/3 h- (pr)2/3


t 1 + T Da1 (Le) 2/3 (30)

[(-Al) oI 2/3
t 1 + Dao where $ = T_ (Le)2
L po
and Le is the Lewis number D/a.
The student now chooses a value of e, and /3
as fixed. Then for a series of 71 Dao values (say
0.001 to the limiting value of unity), he or she*
computes t and by equation (29) the 71 -,q Dao
values are obtained. This simple procedure is re-
*Yes, Virginia, there is a co-ed Notre Dame

peated at other positive exothermicc) and nega-
tive endothermicc) value of P3 for the fixed e.
Results are shown in Figs. 3 and 4 for E of 20
and 10, and a range of 8 values.

E =10
t 10

Co kga
Fig. 3. Non-Isothermal External Catalytic Effectiveness in terms of
Observables-First Order, E = 10

E_ = 20

7 Dao =Cko

Fig. 4. Non-Isothermal External Catalytic
Observables-First Order,

Effectiveness in terms of
E = 20

We see that the chief characteristics of the
non-isothermal intraphase (internal) effective-
ness factor are displayed by the very easily cal-
culated external effectiveness behaviour:



(a) ] values much greater than unity are
found for exothermic reactions.
(b) the Arrhenius number, e is more import-
ant than /3 in determining q and
(c) at high values of the modulus 17 falls well
below unity.
Yield/selectivity in a non-isothermal atmos-
phere is readily treated. In equations (17) and
(23), the ratio k,/k2 at surface temperature ap-
pears. This ratio relative to its value at bulk
temperature To, is

kl/k) o

e(E E2)
Ac = -1 2 RT
A valuable qualitative insight into non-iso-
thermal yield/selectivity trends is secured by the
student by consideration of Fig. 5.
E1 < 2 El E2
kl/k 1
k/k2)o E > E2 El < E 2
t < L -1 t > 1
endothermic exothermic
Fig. 5. Yield/Selectivity Trends in Non-Isothermal Atmosphere.

In the light of Fig. 5, consider non-isothermal
selectivity in simultaneous reaction, equation (23)
0 1k2 [ ]U-A 1 (23)
w (k 1/k2). IA 0
From a knowledge of the sign of AE and the
thermochemistry of the reactions, the student can
predict whether mass and heat transfer limita-
tion will enhance or tax selectivity, or over a
range of conditions the heat transport limitation
may just compensate the taxation of mass trans-

Transient Isothermal Effectiveness
The analysis of transient external effective-
ness is also simple yet instructive. Consider a case
where reactant is exposed to the catalytic surface
at r = 0. A first order reaction is assumed under
isothermal conditions. What is the 7r-T relation-
koa((o C) = kC +

Dao + dr


n = C/C and 6 = k ae
P p

the desired solution is
Da + exp
(1 + Da) exp u
Which is more conveniently expressed in terms
of i7 at any time relative to 71s, at final steady-
S i ss exp 6
This equation teaches that the dimensionless
group kg a T determines the rate of approach to
steady-state. In the case of the internal effective-
ness in transient, approach to steady-state is gov-
erned by an analogous dimensionless group

where D is internal diffusivity and L the pellet

Practical Applications
While the primary purpose of the development
of external effectiveness is its instructive value
in shedding light on the internal problem, it should
not be forgotten that there exists some quite
important catalytic systems involving gaseous re-
action over nonporous catalysts. For example,
NH, is oxidized to NO and HCN is synthesized
from NH3, 02 and methane over platinum alloy
wire matrices. Methanol may be oxidized to for-
maldehyde over nonporous silver. Such systems
are susceptible to analyses in terms of the external

The diffusion-reaction problem is profitably
viewed in terms of external mass and heat trans-
port affected activity and selectivity in a simple
analytical fashion with readily calculated results
which bear very fruitful analogy to features re-
vealed by the often much more difficult-to-solve
internal diffusion-reaction problems. El
1. Thiele, E., Ind. Eng. Chem. 31, 916 (1939).
2. Wheeler, A. in Catalysis (P. H. Emmet, Ed.). Vol.
2 Reinhold, N. Y. 1955.
3. Weisz, P .B., Z. Physik. Chem. N.F. 11, 1 (1957).
4. Carberry, J. J., Chem. Eng. Science 17, 675 (1962).
5. Tajbl, D. G., Simons, J. and Carberry, J. J., Ind. Eng.
Chem. (Fund,) 5, 171 (1966).
6. Carberry, J. J., A.LCh.E. J. 7, 350 (1961).
7. Tinkler, J. D., and Metzner, A. B., Ind. Eng. Chem.
53, 663 (1961).
8. Weisz, P. B., and Hicks, J. S., Chem. Eng. Science
17, 265 (1962).
9. Hutchings, John and Carberry, J. J., A.I.Ch.E. J. 12,
30 (1966).
10. Hutchings, John, Ph.D. Thesis, Chem. Eng., Univ. of
Notre Dame (1968).


He's locked into

childhood by a disease that's

already licked.

What happened was measles. ...,-
Common measles. And what's
tragic is that it should never have x,
happened at all.
To most people, measles is
simply a childhood nuisance. But
statistics don't bear them out.
During the height of the mea-
sles season, 10,000 children are
stricken every three days. 60 are
hospitalized, 10 develop inflamma-
tion of the brain, 3 become men-
tally retarded. And one dies. 1 1
There's just no excuse for this
disastrous waste. Since 1965, the _.
measles vaccine developed by Dow
has more than proved its worth.
And the cost of immunization is
low compared to the consequences
of the disease, the staggering
expenditure in medical care and
the enormous number of school
days missed.
But after several years of dra-
matic decline, measles is now gal-
loping back. Because even the
best preventive is powerless if peo-
ple refuse to use it.
The answer is not more of
our vaccine. There's already plenty
of that. It's community awareness
of the threat measles poses to our
children. And community action to
stop the disease in its tracks.
At Dow, we're concerned with
more than chemistry. We're con-
cerned with life. And despite our
imperfections, we're determined to
share its promise. Wisely.
For a booklet on measles vac-
cination and your children, write to: ..-
The Dow Chemical Company, .'. "
Midland, Michigan 48640. .- "


I. book reviews

Chemical Plant Simulation, Crowe, Hamielec,
Hoffman, Johnson, Shannon, Woods, xiv+368 pp.,
Prentice-Hall, Englewood Cliffs, N. J., 1971.
Process engineers have long used computers
to provide refined, detailed studies of individual
process equipment. But the real virtue of large,
high-speed computers is the ability to classify, file,
and analyze masses of data. With a sufficiently
powerful computer system, we are able to analyze
complete processes. This attractive prospect has
led to many attempts to build up special systems
for chemical analysis. In a recent survey, I was
able to identify 31 distinct systems; many more
undoubtedly exist in proprietary situations. Most
of these are directed towards "simulation,' but
this differs greatly from the stochastic simula-
tion of the systems engineers. For chemical
plants, simulation means the generation and cal-
culation of the many material, heat, and pres-
sure balances needed to relate the flows, tempera-
tures, pressures, and compositions in a complex
process. Simulation can be applied to the material
balance studies of a process under design; it can
be used to provide cases for scheduling an oper-
ating process; or it can be used for control studies
of an existing plant.
Of course, a simulation really evaluates only
one operating condition or one design. The job
of the process engineer is to make decisions; for
him, simulation is just a first step, he must
explore many possible solutions, changing the
simulation variables through changes in input
data. This is really optimization, but optimization
systems seem to be either too specialized or un-
able to accommodate efficiently the general build-
ing-block units of the process simulators. In any
event, the recent widespread use of simulation
systems in industry indicates that these systems
do perform useful tasks. At the very least they
provide an improved, up-to-date way of assem-
bling and retaining the data needed for complex
processes. To be efficient and competitive, all
process engineering groups today must be facile
in their use. Thus, the young engineer's first
process assignment often involves computer simu-
How do we take account of this in the chem-
ical engineering curriculum? One solution,-fully
demonstrated in a number of universities in the
U.S. and Canada, is to have students work
realistic process simulations using a well-

maintained computer system. The group at Mc-
Master University, headed for many years by
Prof. Ab Johnson, built up an expertise in this
area through joint cooperative effort of stu-
dents and faculty in several full-scale simu-
lation studies. This effort was based on the
PACER system first developed at Purdue by
Prof. Paul Shannon and his student, Mosler, in
1964. The Important results of this collaboration
are now available in a well-constructed and care-
fully edited book. The preliminary paper-bound
version has been in use at McMaster and else-
where since 1969. With so many tables, figures,
and charts, the new edition is much easier to use.
Of course, the use of simulation studies in
the undergraduate curriculum does require time.
The student should not be asked to make such
studies if it detracts from his ability to think
about the process. If his major concern becomes
one of getting the cards punched in the right
places, rather than thinking about the effect of
the numerical input values on the properties and
flows, then the simulation study is not really
worthwhile. Adequate time must be provided to
evaluate carefully the significance of the process
behavior underlying the simulation. However,
if the system is properly used, simulations sup-
ply a good idea of the nature of real processes.
Simple examples may give the student a clear
understanding of process structure, but may also
mislead him, through oversimplification of com-
plex relationships.
The book is oriented to, and illustrated with,
a specific system, PACER (and in particular the
McMaster version, MACSIM), but it is a suit-
able introduction to many of the general ideas
and principles of simulation. In spite of differ-
ences in programming and data handling, all of
these simulation systems rely on the same basic
1. The existence of general-purpose equip-
ment modules interrelated by streams.
2. The use of equipment vectors to store the
particular characteristics of the equipment in-
volved in a particular process.
3. The use of stream vectors to describe the
properties of the streams, and more generally,
of the information flow.
4. The characterization of the information
flowsheet in terms of a process matrix.
5. The use of numerical iteration to calcu-
late systems with recycle.
6. The ability to develop and use in the


simulation system general-purpose prediction
routines or data libraries for physical and chem-
ical properties.
Many of these concepts are fundamentally
simple and are illustrated in elementary courses
in stoichiometry and thermodynamics through
simple, clearly described examples. However,
computer application to large, practical problems
results in system complexity and in multiplicity
of detail. It is difficult to impress the undergrad-
uate with the nature of this complexity and at
the same time keep him from complete confusion.
The straightforward how-to-do-it approach in
this text is a good solution. Careful study of the
text, with a simultaneous application to a real-
istic example project should give the student a
good appreciation of the value of computers in
process analysis.
The first two chapters provide an elementary
introduction of the basic ideas, using simple
process flows and simple equipment modules for
mixing, separation, etc. In these chapters, there
is a description of the recycle problem and its
solution by iteration, and a brief review of tech-
niques for analyzing process structure. Chapter
3 introduces PACER and describes the many
control elements needed for this specific system.
Even though the reader may be using some other
simulation system, the specific description sup-
plied may be the best way to understand clearly
the nature and extent of control information re-
quired by large systems.
The central portion of the book is a detailed
description, section by section, of the classic Mc-
Master example, the contact sulphuric acid plant
simulation. After a general description of the
plant (Chapter 4), the two major decisions about
modelling are described (Chapter 5). This is one
of the most important chapters in the book since
it emphasizes the way in which the engineer can
control the sophistication, adaptability, and ac-
curacy of the simulation at the expense of great-
er development time, more knowledge of basic
data, and greater demands on the computer. Then,
(Chapters 6-10), the various types of equipment
modules are developed in some detail, using those
for the sulphuric acid plant as basic examples.
Here, the need for various levels of sophistication
is described and illustrated. In spite of the spe-
cific nature of the examples, the discussion covers
all the major types of equipment. Moreover, the
problems at the end of each chapter are often
oriented to other systems which have been
studied by the McMaster group. In Chapter 11,

the authors describe the use of the simulation,
and the results obtained in some particular stud-
ies of the sulphuric acid plant. With a thorough
study of these chapters, the reader should be
able to apply PACER to other chemical processes
with which he is familiar. And he would prob-
ably find it helpful even with another simulation
system. Often, the users' manual is written for
skilled or experienced engineers. Here, the de-
scriptions and illustrations are definitely orient-
ed to use as a text, or basic reference. In addition
to the bare details, the book also includes a full
description of some of the tricks which are es-
sential for efficient simulation of real processes.
For example, the sulphuric acid plant simulation
uses a special pressure module to handle the
pressure balances in the system without excessive
information recycle.
In the closing chapters, several more general
ideas are covered. Chapter 12 briefly describes
optimization, using small-dimension direct-search
techniques. Of course, optimization is the only
efficient way to make decisions in large, multi-
variable problems, yet the need for optimization
is not always recognized. In the survey mentioned
earlier, only CHEOPS (which does not take ad-
vantage of the new developments in computers)
and RPMS (which is limited to linear models)
are truly oriented to optimization. The other sys-
tems, like PACER, require excessive duplication
in the computation of each case. Optimization
calls for several thousand cases in a typical direct
search of ten to fifteen variables. Early versions
of PACER are much too clumsy and slow to make
such computations efficiently, since each case is
treated as a complete simulation. There has re-
cently been some recognition of this handicap,
and it is possible that some of the proprietary
systems now short-cut duplicated computation.
But the description in Chapter 12 is little more
than an introduction to basic ideas which can
only be applied in a limited way to PACER
Chapter 13 is much more successful and im-
portant. This is a critique and summary of the
strategy of tackling a simulation. It includes a
detailed and complicated flowchart of simulation
procedures which is well worth careful study.
The authors are among the most experienced
users of simulation. Their comments about their
past experience are intelligent and significant.
Finally, Chapter 14 provides a look ahead to
possible future developments in the use of simu-
(Continued on page 32)



California Institute of Technology
Pasadena, California 91109

The California Institute of Technology (Cal-
tech) is a small private school with a total en-
rollment of about 1500 students, about half of
whom are graduates. The faculty consists of
about 250 professors of all ranks, about 130
visiting professors and associates, and some 240
research (postdoctoral) fellows. With this un-
usual student-faculty ratio classes at Caltech
are generally small, interpersonal relations are
fostered, and course programs and research ac-
tivities tend to be closely interwoven.
These typical Caltech tendencies are reflect-
ed in the Polymer Program in Chemical Engi-
neering. Two lecture courses and a laboratory
course are offered as part of the ChE program.
The research group in the Polymer Laboratory
currently comprises six graduate students, a
visiting scientist, and three research fellows,
representing five different countries. The poly-
mer activities of the Polymer Research Section
of Caltech's Jet Propulsion Laboratory are
brought to the campus by its head, Dr. R. F.
Landel, and Member of the Technical Staff, Dr.
A. Rembaum. Both serve as part-time Lecturers
in Chemical Engineering. A series of weekly
seminars keeps the group in touch with advances
being made in the study of polymers outside of
the Caltech community.

A basic course, Polymer Science, is offered
for graduate students and for those undergrad-
uates who have had a sufficient background in
physical and organic chemistry. The first term's
lectures, presented by A. Rembaum, cover poly-
mer chemistry. Topics discussed include: the
nature and classification of polymers, methods of
synthesis, polymerization kinetics and molecular
weight distribution, copolymerization and cross-
linking. The second and third terms' lectures
are given by N. W. Tschoegl. During the second
term, attention is focused on the physical char-

acterization of polymers by solution methods
and physical methods in bulk. A detailed treat-
ment of polymer properties is the subject of
the third term which includes a discussion of
the principles of polymer technology. Through-
out the course the emphasis is on an under-
standing of polymer properties in terms of poly-
mer structure.
During the third term, students can elect to
take the Polymer Science Laboratory course. This
course has been developed by R. E. Cohen on a
trial basis over the past two years and was offer-
ed for credit for the first time in 1971-72. The
laboratory course acquaints students with a
selection of techniques employed in the synthesis
and characterization of polymeric materials. The
techniques have been chosen for their practical
importance and instructional value. The student
first synthesizes a polymer, following the kinetics
of the free radical polymerization reaction. He
then characterizes his reaction product by de-
termining number average and viscosity average
molecular weights and the glass transition tem-
perature. Finally, he studies the mechanical
properties of his polymer by carrying out stress
relaxation measurements. A laboratory manual
with a discussion of the procedures and back-
ground material has been prepared for the course.
Polymer Science is essentially an introduc-
tory course, although the presentation is fairly
advanced, being on the level of P. J. Flory's
Principles of Polymer Chemistry. The course
alternates every second year with a more special-
ized course on the Mechanical Behavior and Ulti-
mate Properties of Polymers, open to graduate
students. Students enrolling in the latter course
are expected to have completed studies equiva-
lent to the Polymer Science course. The course
begins with an introduction to the theory of vis-
coelastic behavior. The discussion centers on
material functions and their interconversion,
model representation, time-temperature equiva-


Sti Ei X
Robert E. Cohen obtained his BSChE degree with
distinction from Cornell University and his ('72) PhD
from Caltech. He is currently a Research Fellow at
Caltech, and will take up a similar position in the De-
partment of Engineering Science at Oxford University
in England in the Fall of 1972. In the Fall of 1973 he
will be an Assistant Professor at MIT. (right photo)
Nicholas W. Tschoegl obtained the degree of BSc
(Hons.) in Physical Chemistry for the New South Wales
University of Technology in Sydney, Australia, and the
('58) PhD degree from the same institution which is
now the University of New South Wales. He spent two
years each at the University of Wisconsin and Stanford
Research Institute before coming to Caltech in 1965.

lence, and the molecular theories of polymer be-
havior. During the second term consideration is
given to the mechanical behavior of various
polymer systems including amorphous, crystal-
line, and cross-linked polymers, copolymers, elas-
tomers, filled and plasticized systems, blends and
melts, and to the elements of large deformation
theory. The first two terms' lectures are given
by N. W. Tschoegl. The third term, presented
by R. F. Landel, is devoted to a discussion of the
phenomenology and the molecular and statistical
theories of rupture in polymeric materials.
Throughout the course special attention is given
to controlling molecular parameters.
In accordance with Caltech's small enroll-
ment, a typical class in the polymer courses con-
sists of about ten students. Thus an informal
atmosphere and close interpersonal contacts are
insured. Each student prepares an individual
term paper on a topic related to the course once
during this academic year. These papers are
presented in form of a lecture to the class and
the research group in a common seminar. Grad-
uate teaching assistants (drawn from the re-
search group) are given opportunities for pre-
paring and presenting lectures within the frame-
work of each course.

Research activities in the Polymer Labora-
tory are directed toward obtaining an under-
standing of the molecular basis of the engineer-
ing properties of rubbery materials (elastomers).
The majority of the research projects is of an
experimental nature, dealing with the study of
elastomeric materials through the measurement
of various mechanical properties. Theory is de-
veloped as needed as an integral part of the re-
search activity.
Mechanical tests with both transient and dy-
namic loading patterns are used to evaluate the
properties of a variety of elastomers at atmos-
pheric pressure, in isothermal and isochronal
modes of testing. A new rheometer with a wide
modulus, frequency, and temperature range has
been developed. Two Instron testers are avail-
able for large deformation studies.
The Polymer Laboratory is uniquely equipped
for the study of pressure effects on polymer
properties. Pressurized testing equipment is be-
ing used to study the pressure dependence of
mechanical properties, failure criteria for filled
and unfilled elastomers in uniaxial and biaxial
deformation under superposed hydrostatic pres-
sure, and the internal energy contribution to
rubber elasticity.
With the help of A. Rembaum, some synthe-
sis is carried out in conjunction with the research
effort. The capability of synthesizing particular
polymers as part of the mechanical properties
research distinguishes the Caltech program from
most polymer programs in other ChE Depart-
ments. The laboratory has a high-vacuum ap-
paratus which is used for the anionic polymeri-
zations of block copolymers. The synthesis of
special elastomers containing a known amount
of chains tagged at both ends by a heavy atom
is expected to lead to X-ray scattering studies
of the end-to-end separation of chains as a func-
tion of temperature and strain. Ample character-
ization equipment is available so that the molecu-
lar structure of newly synthesized materials
can be determined.
A good deal of the effort is concerned with
the properties of novel elastomers. Thus, triblock
copolymer elastomers (thermoplastic rubbers)
are used as model substances for the elucidation
of properties which are difficult to study in con-
ventional elastomers. These studies have led to
the development of a theory for the superposition
of time and temperature effects in thermorheo-



With this unusual student-faculty ratio, classes at Caltech are generally small, interpersonal relations are fos-
tered, and course programs and research activities tend to be closely interwoven.

logically complex two-phase polymer systems.
A study was also made of terminal chains and
their entanglements in triblock-diblock blends
in which the proportion, length, and distribution
of the terminal chains could be controlled. A
study of rubbers consisting of two interpenetrat-
ing networks is expected to begin in the Fall of
Research activities are not restricted to the
Polymer Laboratory. Thus, within Chemical Engi-
neering, Professors W. H. Corcoran and N. W.
Tschoegl collaborate in an effort to assess the
effect of mechanical deformations on diffusion
through membranes. A joint effort on segmental
motion by NMR techniques is being planned
with Professor R. W. Vaughan. Outside of Chem-
ical Engineering, the Polymer Laboratory has
maintained close contact with Professor W. G.
Knauss's group in Aeronautics whose interests
lies in the field of crack propagation and frac-
ture in elastomers. Close liaison is also main-
tained with the Materials Science group at the
Institute. Because of its small size, such inter-
departmental contacts are easy to maintain at
In the past, both Materials and Chemistry
students have received Ph.D. or M.S. degrees
for thesis work done in the Polymer Laboratory.
Chemistry and Chemical Engineering form a
single division at Caltech, and the polymer
courses attract an increasing number of chemistry
students. Chemistry undergraduates receive cred-
it for undergraduate research performed in the

During the academic year, the polymer group
meets weekly for an informal luncheon seminar
held in Professor N. W. Tschoegl's office. These
seminars have the character of a special topics
course for which, however, no credit is given.
Approximately half of the seminar speakers
come from outside Caltech. An effort is made
to invite speakers who do research which is dif-
ferent from that of the group or who can sup-
plement the expertise of the lecturers presenting
the regular courses. Since it is recognized that
many of the group's graduates will go into
industry, a special effort is made to invite speak-

ers from industrial R&D laboratories to acquaint
group members with the directions and values
of high-level industrial research and develop-
ment work. Through the seminars the members
of the group meet the many well-known polymer
scientists who regularly pass through Caltech.
The seminars contribute to the broadening of
the group's concept of polymer research beyond
the study of mechanical properties and rubbers.
The seminars also serve as a forum for
periodic reporting by members of the group on
the progress of their research. A graduate stu-
dent is expected to present an average of three
to four talks per year and in this manner gains
valuable experience in the art of oral expression
and communication. Every group member is ex-
pected to be sufficiently familiar with everybody
else's research to be able to interact and such
interaction is encouraged. This had led to in-
tensive mutual reinforcement of interest and to
cohesion within the group, and has sensitized
group members to the demands and advantages
of the team work so often essential to successful
work in industry.
The seminar program has been very success-
ful and has attracted interest from other groups
at Caltech and also from outside the Campus. A
major advantage of the seminars is, however,
that they are normally restricted to a small
group of about 15 people to encourage free-
wheeling and uninhibited discussion. E

BOOK REVIEW: (Continued from page 29)
lation: the extension to design; the relationship
of process simulation to the stochastic simula-
tion of the systems engineer; and the strong de-
pendence of simulation systems on new hardware
In summary, this book is recommended as a
basic source reference on simulation systems.
The reader should find it valuable either as an
undergraduate text or for self-study. With its use
he should be able to understand the fundamental
elements of a simulation, and have less confusion
as to the application of simulation systems to
large complex processes.


0 l problems for teachers

Analog Simulation of the

Dispersion of

Atmospheric Pollutants

University of Maryland
College Park, Md. 20742

The problem is to devise an analog program
for simulation of the dispersion of atmospheric
pollutants. Beginning with a commonly accepted
dispersion model, which is analytical in nature,
show the steps involved in the analog simula-
tion of the vertical concentration profile at suc-
cessive points downwind of a line source. This
problem can be used as a very effective demon-
stration or as an introductory problem at various
stages of a course on analog simulation.

The most commonly used models of atmos-
pheric dispersion from continuous sources are
the Gaussian Plume Models (D. B. Turner,
Workbook of Atmospheric Dispersion Estimates,
U. S. Department of HEW, 1967). For an in-
finite line source such as might be used to simu-
late automotive emissions on a freeway, the
model has the following analytical form
q( -- exp xpr ( (+)
Y7T sin, 1L 2 L 2 z j J
H is the effective height at which the emis-
sions occur (meters).
z is the vertical distance above the ground
u is the wind speed (meters/sec).
is the angle between the source line and
the wind direction (limited to D> 45)
q is the rate of emission per unit length
(grams/ (meter sec)).
X is the concentration of the pollutant
(grams/cubic meter).
o-z is called the vertical dispersion coefficient

o-, has been empirically related to the distance, x
meters, from the line source to the point of in-
terest and to the stability condition of the at-
mosphere. Suggested correlations are given by
Turner. As approximations to these correlations:
oz 0.023 x (2)
when the turbulent nature of the atmosphere is
minimal to
l = 0.14 x (3)
when the atmosphere is very turbulent.
Models very similar in nature are also pre-
sented by Turner for point and instantaneous
source emissions.

Analog Simulations
Using the model described above, two sit-
uations will now be considered for analog simu-
lation. The first is the simulation of the vertical
concentration profile at a fixed distance down-
wind assuming, q, , o-z, u, and H are fixed. The
second is the simulation of the vertical profile
in an automated manner as the distance down-
wind increases linearly.

* Vertical Profile at Fixed Downwind Distance
To develop a model suitable for analog simu-
lation, the basic dispersion model is first written
X -- (a + f) u(4)
/FT sin4 o u

(Z+H) NQ ---- + REF
Fig. 1. Vertical Profile Generation


Using Equations (7) and (8), the analog
diagram presented in Figure 1 can be used to
compute X as a function of z for fixed q, 4, u, 0'z,
and H. Table 1 below details the variables and


Charles L. Miller received his BS in engineering from
PMC Colleges in 1968 and his MS in ChE from the Uni-
versity of Maryland in 1971. He is currently working
toward his PhD in ChE at the University of Maryland. He
is a member of the AIChE.
Theodore W. Cadman received his BS, MS, and PhD
from Carnegie Mellon University. Since 1965 he has been
engaged in teaching and research in the Department of
Chemical Engineering at the University of Maryland. His
primary interests are in process simulation. He is the
faculty advisor for the student AIChE and a member of
Analog Hybrid Educational Users Group, AIChE, ACS
and IST. (right photo).


a = xp a = exp cz+ ] (5,6)

The variables a and p are then differentiated
with respect to z to yield the following differen-
tial equations for the purposes of analog solu-

dz 2-- (7)

=(+H) ( 8)
dz 2

By reference to Equations (5) and (6), the ap-
propriate initial conditions are found to be

a (z = 0) = exp a = B(z-0) (9)

(a+ ,)/
Fig. 2. Vertical Profiles for Various H.

Amp 1
Amp 2
Amp 3
Amp 4
Amp 5
Amp 6
Amp 7
Amp 8
Pot 1
Pot 2
Pot 3
Pot 4
Pot 5
Pot 6
Pot 7
Pot 8
Pot 9
Pot 10
Pot 11

Before Scaling
- a
- 8
- a(z-H)
- B(z+H)
- (z-H)
- (z+H)


After Scaling
- a/a*
- 8/i*
- a(z-H)/(a*(z-H)*)
- (za+H)/(B*(+H)*)
- (z-H)/(z-H)*
- (z+H)/(z+H)*
a(0) la*
(z-H)*/(o B)
(z+H)*/(a2 B)
H/(z+) *
8"/( +B)*

parameters of the circuit before and after scal-
ing. For convenience in this table a superscript *
is used to indicate the maximum magnitude and
B is used to represent the time scaling factor.
Using the values given in Table 2, the sample
results given in Figure 2 were obtained by vary-
ing H from 0 to 3 o-.


Parameter Value

a 1

z 10 a
(z-H) 10 a

(z+H)* 20 a

(c+8) 2

C 10
As indicated in Figure 2, at low volues of H
the maximum concentration occurs at ground
level. In this context, it is worth noting that the
model assumes complete reflection of the pol-
lutant at the ground. At higher values of H,
the maximum concentration occurs at H, the
profile becomes more symmetric, and the maxi-
mum value of (a+8) approaches 1/o- as a be-


comes increasing dominant in the sum. It may
be further observed by reference to Equations
7, 8, and 9 that z and H may be related to o-z as
suggested in Figure 2 in order to obtain a gen-
eral solution valid for o-z.

e Spread of Profile in Downwind Direction

To simulate the dispersion as a function of
the downwind distance, it is necessary to change
ao- as suggested by the approximations in Equa-
tions 2 and 3. If an analog capable of two speed
integration, possessing a moderate amount of
logic, is available, the simulation of this down-
wind dispersion can be automated to yield an
effective demonstration. In essence, the automa-
tion procedure is to use a slowly continually op-
erating analog circuit to generate -z, o-(0), and
P3(0) for use in the more rapidly repetitive op-
erating circuit given in Figure 1. Pots 1 and 2
are, of course, eliminated and Pots 3 and 4 must
be replaced by dividers to handle a varying o-,.
In addition, a final divider must be used to com-
pute the (a + p ) lo-, terms in x as indicated by

.- -REF X + REF X/-- REF

-E X -x2/2 X3/6
-REF -

0 202 6a3

-z- (=0)= 3(Z=0)


Fig. 3. Generation of Parameters

Equation 4.
Assuming o-, = ax where a is a constant and
noting that

ac(z-) 8(z=0) = exp 2 ,
2 a 2
either x or ao- can be used as the independent
variable on the slowly operating circuit. As
indicated in Figure 3, o-, and 0-22 are obtained by
integration and a(z = 0) = /3(z = 0) are de-
termined using
d a(z=O)1 at2 o
d (z0) (10)
Initial conditions on Equation (10) are chosen

at the lowest non-zero value of x to be simu-
lated. Note that x = 0 can not be assumed be-
cause it is a singular point in the model. Again
referring to Figure 3, the generated values of o-.,
o-Z2, a(z = 0), and /8(z = 0) are fed to the re-
petitive operating circuit which computes X. If
the secondary circuit operates very slowly, ao-
and a-22 will change neglible during any one re-
petitive run and thus can be used directly in the
primary circuit. If the secondary circuit is fairly
rapid, a pair of track and store units can be used
to interface the two circuits and maintain o-, and
and a-'2 constant during each repetitive run. By
plotting X + x versus z the dispersion of the pol-
lutant as a function of the downwind direction
can be automatically obtained. Figure 4 presents
several example profiles obtained in this manner.
X is chosen to have a maximum value of 0.1 at
a distance of 100 meters downwind. The result-

100 200 500 400
X + 103X
Fig. 4. Dispersion Downwind of Source.

ing dispersion downwind at 200, 300, and 400
meters is given.

The analog simulation of the dispersion of
atmospheric pollutants has been briefly examined
in this paper. It has been shown that the prob-
lem can be used as an effective demonstration
in order to examine the essential features of at-
mospheric dispersion. In addition, the problem
presents several more difficult scaling problems
for the student. The analytical nature of the
solution provides, however, a ready check of the
simulation and thus the simulation can prove
to be an effective student problem at several
stages of an analog course. El


McGraw-Hill Presents

of the United States Naval Academy. 1973, 480
pages, $17.00
A balanced approach between theory and analy-
sis/application of that theory is presented for all
three modes of heat transfer. A thorough de-
velopment of the methods for formulating mathe-
matical models in terms of non-dimensional para-
meters is stressed. Well documented, interactive
computer programs, written in the BASIC pro-
gramming language, are an integral part of the

DAVID R. GASKELL, University of Pennsyl-
vania. 1973, 500 pages (tent.), $19.50 (tent.)
This new text provides a systematic illustration
of the application of modern thermodynamics to
the determination of equilibria in metallurgical
systems. After discussing the basic laws and
introducing the necessary thermodynamics func-
tions, application is made to increasingly com-
plex systems in the sequence, reactions between
gases, reactions between gases and pure con-
densed phases, reactions between gases and con-
densed solutions reactions in condensed solutions
and electrochemical reactions.

WILLIAM L. LUYBEN, Lehigh University. Mc-
Graw-Hill Series in Chemical Engineering. 1973,
558 pages, $18.50
Professor Luyben has devoted his book to pre-
senting only useful, state-of-the-art, applications-
oriented tools and techniques most helpful for
understanding and solving practical dynamics
and control problems in chemical engineering
systems. Written for the undergraduate student,
this text offers a unified, integrated treatment of
mathematical modeling, computer simulation, and
process control.

PROJECT, State University of New York. 1973,
400 pages (tent.), $7.50 (tent.). A Casebook of
Problems will be available.
new kind of science for all students in a text
that is fun to read, easy to understand, and bril-
liant in concept. Unveiled is a humanistic view
of technology and change stressing "technological
literacy" and the interaction of society, nature,
and the man-made world. The reader is offered
a look inside technology today including informa-

tion systems, modeling, communication with ma-
chines, and decision making. Students from all
disciplines are challenged to ask questions about
science and technology and then guided into de-
veloping the skills necessary to relate technology
to their own experience and communities.
Thermodynamic and Transport Properties
of Fluids
both of the University of Florida. McGraw-Hill
Series in Chemical Engineering. 1973, 496 pages,
With an emphasis on applications, APPLIED
STATISTICAL MECHANICS is an introduction
to the various ways in which statistical thermody-
namics and kinetic theory can be applied to sys-
tems of chemical and engineering interest. Pre-
sented is a fundamental, up-to-date treatment of
statistical-mechanics with primary interest focus-
ed on molecular theory as a basis for correlating
and predicting physical properties of gases and
liquids. Material on recent theoretical approaches
such as perturbation theory and the statistical-
mechanics of nonspherical molecules is included.
JOHN R. SIMONSON, The City University,
London. McGraw-Hill Series in Mechanical Engi-
neering. 1973, 244 pages (tent.), $10.50 (tent.)
With the guidance of this step-by-step introduc-
tion to the subject of heat transfer, the student
is exposed to almost every aspect taught at the
undergraduate level in engineering courses. In
addition to the British units utilized, the new S.I.
units are incorporated into the illustrative ex-
amples within the text. For the industrial reader,
the treatment covers the transfer unit approach
to heat exchanger design and the electrical analog
network approach to radiation problems, as well
as some of the latest empirical equations of con-
vective heat transfer.
TONER, both of Princeton University. 1973, 500
pages (tent.), $12.95 (tent.)
Unique in chemical engineering literature is the
treatment of degrees of freedom for material and
energy balances. Either chemical or physical pro-
cessing elements are handled in a unified manner.
The authors have included the first law of thermo-
dynamics, unsteady state mass and energy bal-
ances, and all pertinent physical chemistry re-
quired. The modular organization of material
offers the instructor a wide choice for his par-
ticular syllabus.

McGraw-Hill Book Company

A Hotbed Of Engineering Ideas

DITSWORTH, both of Arizona State University.
1972, 415 pages, $15.50
Offers a unified treatment of fluid mechanics
utilizing the student's background in thermo-
dynamics and dynamics. Vectors are employed
to formulate those physical laws which pertain
to continuum fluid mechanics and a large number
of illustrated problems is included.

E.R.G. ECKERT, University of Minnesota and
ROBERT M. DRAKE, JR., Vice-President, Com-
bustion Engineering, Windsor, Connecticut. Mc-
Graw-Hill Series in Mechanical Engineering.
1972, 750 pages, $21.50
Acquaints the senior or graduate level student
with the fundamentals of heat and mass transfer.
Unlike most other books at this level, this single
volume covers conduction, convection, and ra-

of Purdue University. 1972, 496 pages, $16.50
Provides the student with a basic knowledge of
the mechanisms by which momentum, heat, and
mass are transferred and introduces the design
and use of equations for momentum, heat, and

DAVID M. HIMMELBLAU, University of Texas,
Austin. 1971, 416 pages, $18.50
Containing the latest information on existing
methods of nonlinear programming, this advanced
text is an excellent sourcebook for engineers
working in optimization, operations research, and
systems engineering.

HEAT TRANSFER, Third Edition
JACK P. HOLMAN, Southern Methodist Uni-
versity. 1972, 496 pages, $13.50
This elementary text offers a brief and concise
treatment of all phases of heat transfer. New
features include chapters on environmental prob-
lems, emphasis on numerical techniques in con-
duction problems and an increase in text ex-

University of South Florida. 1972, 714 pages,

A systematically organized treatment of theory
and application proceeding from an analysis of
the simplest extended surface through design
and application of mass-producible components
for complex systems.

ROBERT SIEGEL, Lewis Research Center,
NASA and JOHN R. HOWELL, University of
Houston. 1971, 832 pages, $18.50
To establish a solid foundation for future study,
the authors have examined the fundamentals in
detail. The work deals generally with radiative
behavior of materials, radiation between surfaces,
and gas radiation.

both of Tennessee Technological University. 1971,
704 pages, $18.50
Assuming a basic knowledge of differential equa-
tions and vector calculus, this text offers a mod-
ern, integrated treatment for the study of heat,
mass, and momentum transfer.

JOHN C. SLATTERY, Northwestern University.
1971, 704 pages, $19.50
An integrated introduction for the first-year
graduate student which offers original treat-
ment of four subjects often taught separately:
fluid mechanics, thermodynamics, heat transfer
and mass transfer.

PHILIP A. THOMPSON, Rensselaer Polytechnic
Institute 1971, 688 pages, $17.50
Useful to students and professionals in diverse
fields, this text emphasizes the fundamentals of
compressible-fluid motion instead of its tech-

Second Edition
J. M. SMITH, University of California, Davis.
1970, 544 pages, $16.50
Acquaints the student with the tools necessary
to design new chemical reactors and predict the
performance of existing ones. The principles of
kinetics and reactor design are developed.
Prices are subject to change without notice.

1221 Avenue of the Americas, New York, N.Y. 10020




Laval University
Quebec, 10, Que., Canada

Since the last decades, the discipline of Chemi-
cal Engineering has been subjected to a radical
change with the now popular entry of Computer
Science and its related techniques. In every field
of Chemical Engineering, such as Transport
Phenomena, Process Control, Chemical Reactor
Design ... a knowledge of Computer Science and
its applications is evidently necessary. Applica-
tions of computer even permit the establishment
of various special techniques of analysis to Chemi-
cal Engineering: Optimization,1 Simulation,2
Mathematical Modelings 3'4 and Computer Aided
Design.5 6
Although everybody recognizes the advantages
of these modern techniques in Chemical Engi-
neering, the incorporation of courses in these sub-
jects often raises some problems for both the
graduate students and the instructor.
Indeed, graduate students in Chemical Engi-
neering seem reluctant to select, say, a course on
Optimization of Chemical Processes, for the
simple reason that they are aware that, to succeed
in such a graduate-level course, they should have
already passed or must follow various other
courses: Computer Programming, Applied Mathe-
matics, Numerical Analysis, Operational Research
. . On the other hand, for the instructor, a
common problem he must face is how to present
such a wide-range but necessary background and
the main content of the course under a rather
heavy constraint: the limited time of a semester
course with 2 or 3 hours weekly. Numerous in-
structors will share with the authors the regret-
ful sentiment of lacking the time to attain their
selfish desire: seeing their students being able
to apply the shown techniques to a chemical engi-
neering problem, rather than obtaining generally
a vague notion of theoretical principles. In view
of diminishing the difficulties mentioned, we, at
Laval University, have tried to introduce a gradu-

ate course on the Modern Analysis Techniques
for Chemical Engineering using extensively the
advantages of the APL System, which we report
here in this article.

The purposes of the course are two-fold: to
give chemical engineering graduates an opportu-
nity to be familiar with APL, one of the most
scientific and advanced programming languages,
and at the same time to get them acquainted with
modern analysis techniques.
The organization of the course depends heavily
on the APL system,7 9 hence a brief description
of it is recommended. An APL system comprises
a central computer and an indefinite number of
typewriter-like terminals. A certain number of
these remote terminals may be simultaneously
linked to the computer, therefore easing the wide
use of this computing tool. The user's instruc-
tions and programs can be stored in the center
and copied for other use at any moment by some
special instructions. As for the language itself,
first defined by K.E. Iverson,7 it is found to be
the most convenient for mathematical and engi-
neering computing purposes due to its logic and
compactness. Moreover, the system's possibility
of an interactive approach in problem solving
makes it very attractive to the beginners. From
the educational point of view, the characteristic
advantages of the system have been exploited to
give formal theorems demonstration and engi-
neering applications."

The course has been tested for the past 3
years at Laval University, Canada; after each
session, the results were analyzed with hopes of
making changes that would better suit the compre-
hension level and motivations of the "new-genera-


Co Pham was educated in chemical engineering at Laval
University, Quebec, Canada, (BS, MS). Currently, he is
a research fellow in process modeling and control and a
lecturer in applications of the APL System. Apart from
his PhD thesis on mass transfer and axial mixing, he is
also author of numerous articles. His research activities
include optimization, process control, industrial engineering,
unit operations and biomedical engineering.
L. Cloutier is Professor of Chemical Engineering at
Laval University, Quebec, Canada and Associate Editor of
the AIChE Journal. He received the doctorate in Applied
Sciences from Laval University where he is acting now
as Chairman. His fields of specialization are mixing, pro-
cess control and reactor design.

tion" Chemical Engineering Graduates. The tests
showed that good results were obtained by a well-
balanced content of scientific programming and
applied analysis techniques.
As a first part of the course, background on
Computer Programming, especially on Time-
Sharing Computing system, was given. After
that, the APL language itself was showed to the
students with particular reference to the scientific
and engineering aspects of the system. In addition
to formal lectures, students can learn the language
by programmed lessons and do exercises which
have been pre-registered in the system. This
proves to be very helpful since it permits students
to learn at their own rate while, by practicing
directly with the system, they get acquainted quite
easily with the rigid logic of a computer. As a
consequence, the time of formal lectures was de-
voted more to the special topics students usually
await from an instructor, such as the charac-
teristics of the system, fundamentals of the tech-
niques and particular hints for engineering stu-
dents rather than some "dull" specific rules to
write a program with an existing computing sys-
After the first part, one can be assured that
trainees can, from now on, master this useful
tool of computing and are ready to see the analy-
sis techniques in themselves with not much diffi-
culties from the programming point of view. As
for the second part of the course, some special-
chosen modern techniques of analysis varying
from Statistical Experimental Planning, Numeri-
cal Mathematics, Linear Programming to Simu-
lation and Optimization were presented and it
was particularly shown how they can be applied
to solve chemical engineering problems in practice.

As a result of our experience, it was found

that the organization of this course greatly im-
proved the student's comprehension, which is
usually difficult to attain in such a short period
due to the rather wide-range nature of the
covered subjects. Indeed, by using the pre-written
programs of the system, students and instructor
are free of the tedious work of having to re-write
the computer program for the studied analysis
technique. For example, when discussing optimal
strategy for allocation of materials in chemical
plants, we can apply directly pre-written pro-
grams related to transport and distribution prob-
lems12 and concentrate much more on the mathe-
matical philosophy of the technique and the chemi-
cal engineers' considerations. Trainees are better
motivated when they can solve in applying the
shown techniques, some real problems met in
practice. As a short-term result, it was found that
the course was also an immediate help to our
graduate students confronting current research
problems. Numerical computing, statistical tests,
decision making . in nearly all our graduate
dissertation thesis since the last years had been
performed by the APL language rather than
previously used cumbersome ones.
One of the reasons for this overwhelming vic-
tory of the APL language might be its scientific
and unified approach, which is very welcomed in
the engineering field.
As a counter-balance of all these advantages,
some drawbacks were observed in our adopted
way of course approach: firstly, the course re-
quired in fact a rather good background on mathe-
matical analysis, and secondly its future utilities,
especially for the industry-minded graduates,
were handicapped by the rare presence of APL
systems outside of the campus.

The course served a definite need in introduc-
ing the modern analysis techniques to chemical
engineering graduates. The course's approach en-
abled the student to gain a broad mastery of com-
puter programming and some modern techniques
to solve engineering problems.
From an instructor's point of view, the fact
that his students can solve, store their solutions
in the system and discuss and exchange after-
ward with his classmates, constitutes a real moti-
vation for the continuing efforts to improve the
course. Moreover, with the more and more widely
spread use of the APL system in engineering
(Continued on page 43)


I laboratory


Using Frequency Response

And Analog Simulation Techniques

San Fernando Valley State College
Northridge, California 91324

T HE TIME REQUIRED to design control systems
by analog simulation can be significantly re-
duced if initial approximations are made using
frequency response methods. The tendency for
most engineers is to use only one of the above
techniques, the one with which they are the most
The object of this study is to present a method
which will reduce the time involved in designing
a control scheme for a process control system.
The method itself combines the use of frequency
response methods (Bode diagrams) and time do-
main methods (analog computer simulation) to
give the most efficient design technique.

T O ILLUSTRATE THE METHOD, let us consider a
chemical process system described by the
linear model shown in Fig. 1. The system to be
studied is a spray dryer used to process an emul-
sion to obtain a powder. Control may be effected
by regulating the feed to the diffuser vanes. The
air heater consists of two non-interacting transfer
lags of time constant 100 seconds each. The
drum behaves as three transfer lags of time
constant 12.5 seconds each, and one distance
velocity lag of 2 seconds. A distance-velocity lag
of 3 seconds exists between changes in air tem-
perature at the heater and its appearance at the

L p Heater

Fig. 1. Linear Model of Chemical Process System.

diffuser vanes. The detecting element has a trans-
fer lag of time constant 1 second. We will neg-
lect the time constant of the control valve. The
block diagram of the system is given in Fig. 1.
Note that any controller designed should
reduce the steady-state error to a step input at
R to an arbitrarily small value, and at the same
time minimize the effect of the load of less than
10% and have reasonable stability.
For a type zero system excited by a step in-
put the steady-state error is defined as follows
e A
ss 1 + K ()
where A magnitude of step, K, = position
error constant defined as lim G (s) as s -- o.
This steady-state error is equal to R-B. What
we are actually interested in is R-C. However,
since there is unity gain in the feed-back path
B = C in the steady-state. For our system
Kp = Kc which is the controller gain. Thus the
steady-state specification will be met by values
of K, greater than 9. Let us now choose K, = 10.
The Bode diagram of the system including
controller gain, using straight line approxima-
tions, is given in Fig. 2. Using proportional con-
trol only we would have a slope of -3 at cross-


\ I
\ I

Fig. 2. Bode Diagram of System by straight line approximation.


01 01

W RadlSzi

Joseph F. Paul received his BS and MS degrees in
ChE from Washington University and has obtained ad-
vanced study at University of Southern California. He
formerly was employed by Electronic Associates, Inc.,
TRW Computer Co., and Monsanto Chemical Co. His
areas of interest include applied mathematics, computer
science, and process control.

over which indicates the system would be un-
stable. However, the presence of dead-time terms
requires that the phase angles be computed inde-
pendently rather than from the magnitude curves
using a tangent scale.
The crossover frequency occurs at 0.175 rad/
sec. This is only an approximation however since
we are using straight line approximations for
the Bode diagram. The simulation on the analog
computer will give us the true results for the
model but the information from the Bode dia-
gram is valuable in that it gives us approximate
values and enables us to go to the computer where
the optimal solution can be obtained by minor
adjustment of the parameters.
We will now check the stability of the system
using proportional control only, again by tab-
ulating the phase angle contributions at the
crossover frequency of all time constants and
dead time terms.
3 tan-' (.175) (12.5) = 3 tan- (2.185) = -196.20
tan-' (.175) (1) = tan-1 (.175) = -9.90
2(.175)180/3.14 = -20.0
The phase margin is -46.1' and the system
is clearly unstable.
In order to stabilize the system we must
bring the slope of the curve at crossover up to
a value of -1. This cannot be accomplished by
a single lag-lead network (three mode controller)
since the slope must be decreased by two. We
will try a double lead network (two-proportional
plus derivative controllers in series). The general
form of the transfer function for this control

system is given as

K (1 + T s)2
(1 + --
where y = 10, T,, = controller constant, and K, =
To check for stability we must select a value
of T,, and compute the phase margin. Once again
we stress that this will only be an approximation.
For a first try, let T, 7.1. This is shown in
Fig. 2. This gives a crossover frequency of 0.27
rad/sec from which we can compute the phase
Tabulating the phase angle contributions we
3 tan- (.27) (12.5) = 3 tan-1(3.38) = -220.5'
tan-l(.27) (1) = tan-l(.27) = -15.1
2 tan- (.27)(.71) = 2 tan-'(.193) = -21.80
2 tan- (.27) (7.1) = 2 tan-'(1.93) = +125.2'
2(.27)180/3.14 = -30.9
This gives a phase margin of 16.9. We will
make a second try in an attempt to increase the
phase margin. We will let Ta = 5.7 which is also
shown in Fig. 2. These values of rd are selected
by choosing a breakpoint on the curve.
Once again tabulating the phase angle contri-
butions we obtain

tan-' (.175) (12.5)
tan-'(.175) (1)
tan- (.175) (.57)
tan-1 (.175) (5.7)

= 3 tan-1(2.185)
= tan-1 (.175)
= 2 tan- (.10)
- 2 tan- (1.0)
(2) (.175) 180/3.14


This gives a phase margin of 32.50 which is
a safer margin of stability. A look at the calcu-

Poo = .1000,
Po, = .4500,

Fig. 3. Dead Time for Drum.

Pol = .5000, Pos = .2322,
Po, .5000, P07 = .5000.

These integrators are run at times ten normal speed.


lations shows that moving the breakpoint in
either direction will decrease the phase margin.
If we make the breakpoint smaller we have a
higher crossover frequency which yields a small-
er phase margin, and if we make it larger we
are back to the situation of a crossover slope of
-3. Thus we are at the point of maximum relative
stability. Therefore the transfer function for the
compensated system is given as

10(1 + 5.7s) 2e-2s (3)
(1 + .57s)2(1 + 12.5s)3(1 + s)
THE NEXT STEP WILL BE to simulate the system
on the analog computer. For this purpose the
EAI-680 analog computer was utilized. The first
phase of the study was to check the frequency

Fig. 4. Drum.

= 0.8000,
= 0.8000,
= 0.8000

Pi 0.8000,
P2o 0.8000.

Pi7 = 0.8000,
P21 = 0.1000,

Fig. 5. Double Lead Controller.
P30 = Pa =- 0.1750.
These integrators are run at ten times normal speed.

. . although simulation is the only method which
yields exact solutions for models . ., the use of . .
the straight line Bode diagram gives ball park
estimates . .

response by developing a Bode diagram. The
next phase was to study the dynamic operating
response to changes in set point and load vari-
ables, and attempt to optimize the design by
varying K, and y. The dead time was simulated
by a modified fourth order Pade delay circuit.
The complete set of analog diagrams is given
in the Figs. 3, 4, 5, 6, 7, and 8.

Fig. 7. Air Heater. Fig. 8. Sine Wave Generator

P,2 0.1000,
P21 = 0.1000,
p., 0.1000,
P2,s 0.1000.

P1o 0.1750,
P1- = 0.1000,
P2 =- 0.1750.

These integrators are run at ten times normal speed.
FROM APPROPRIATE CURVES, the open-loop fre-
quency response of the system was determined
and compared to the Bode diagram in Fig. 2.
The true frequency response of the model of
the system was obtained by generating a sine
wave over a range of frequencies and imposing
it as an input to the set point of the system. Both
the input and output of the system were recorded
on a strip chart recorder. From the input and
output curves the magnitude ratio and phase
angle can be measured as functions of frequency.
The results are plotted in Fig. 9. A comparison
with Fig. 2 shows the error from the straight

Fig. 6. Dead Time for Air Heater.
P0o = 0.3333, P42 = 0.0667, P41 = 0.1558,
P,, = 0.3000, P., = 0.3333, P5s = 0.3333.
These integrators are run at ten times normal speed.


- s.10

Fig. 9. Actual Bode Diagram of system.


.1 -

line approximations. However, from Fig 9 we
see that the crossover is about the same fre-
quency as in Fig. 2. The phase shift at this fre-
quency can be computed from the strip chart.
In this case the phase shift was -170 giving a
phase margin of +10'. This indicates that the
response of the system will be stable but highly
The next step was to test the steady state error
due to a step disturbance at the set point. A step
of one volt was applied and the steady state error
was calculated as 0.098 volt which is within the
specification of 10% The strip chart recording
of the test is shown in Fig. 10.

S.i . .




Fig. 10. Transient response of system to a step change in set point.
0.05 volts/div, 2mm/sec, Ke = 10.

A step disturbance was also applied at the
load variable. This produced the same steady state
error. This was expected since the steady state
value B (t) is the same as C (t).
The overshoot can be measured from Fig. 10
and is about 60% which is somewhat high. A
run was made with Ke = 9 which reduced the
overshoot to 55% but gave a steady state error
of 10% which is just on the limit of the specifi-
cation. A third run was made with K, = 8
which gave an overshoot of 50% and a steady
state error of 12.5%. It was felt then that the
original value of K, = 10 was the best com-
Several more runs were made using this value
of K, and varying Tr in both directions. In all
cases the overshoot was worse as predicted from
the original Bode diagram.
A last set of runs was made in an attempt
to optimize y. For the first run K, and Td were
kept the same and y = 15 was used. This gave
an overshoot of 60% which was the same as y =
10. A value of y 5 was used and this gave an

overshoot of 80%. Therefore the original value
of y was retained.

T HAS BEEN PROVEN by this experiment that
although simulation is the only method which
yields exact solutions for models used for control
system design, the use of approximations such
as the straight line Bode diagram gives ball park
estimates for the controller constants which are
valuable starting points for the simulation study.
Otherwise, one must hunt at random on the
computer until the optimum values are found,
which can be time consuming and wasteful. In
this particular problem the values found by
the approximate technique turned out to be the
optimum values from the computer study. El

APL: (Continued from page 39)
education circles, it is also possible to share the
resulting programs with other educators having
similar interests. L-

1. Himmelblau, D.M., "A Course on the Optimization of
Large Scale Systems." Chem. Eng. Education, 5, No.
4, pp 196 (1971).
2. Beamer, J.H., "Statistical Analysis and Simulation,"
Chem. Eng. Education, 5, No. 4, pp 192 (1971).
3. Franks, R.G.E., "Mathematical Modeling in Chemical
Engineering," John Wiley & Sons, Inc., N.Y. (1967).
4. Kabel, R.L., "Mathematical Modeling," Chem. Eng.
Education, 5, No. 4, pp 184 (1971).
5. Carnahan, B., Seider, W.D. and Katz, D.L., "Computers
in Engineering Design Education, Vol. II, Chemical
Engineering," The University of Michigan, College of
Engineering (1966).
6. Westerberg, A.W., "Computer Aided Process Design,"
Chem. Eng. Education, 5, No. 4, pp 180 (1971).
7. Iverson, K.E., "A Programming Language," John
Wiley, New York (1962).
8. Katzan, Harry Jr., "APL Programming and Computer
Techniques," Van Nostrand (1970).
9. Falkoff, A.D. and Iverson, K.E., "APL/360: User's
Manual," I.B.M. (1968).
10. Hatcher, W.S., and Rethier, P.E., "Une application
du language APL au problhme de demonstration de
th6oremes par ordinateur," in "Colloque APL," In-
stitut de Recherche d'Informatique et d'Automatique,
Paris, France (1971).
11. de Vahl, Davis G., and Holmes, W.N., "The use of
APL in Engineering Education," in "Colloque APL,"
Institute de Recherche d' Informatique et d'Automati-
que, Paris, France (1971).
12. Smillie, K.W., "Statpack 2: An APL Statistical Pack-
age," Department of Computing Science, The Univer-
sity of Alberta, Edmonton, Alberta, Canada.


197.2 4alwed ieckwe


University of Wisconsin
Madison, Wisconsin 53706

In the words of Webster, synthesis is "the
combining of often diverse conceptions into a co-
herent whole" and analysis is "an examination
of a complex, its elements and their relations."
Synthesis refers to the more inventive aspects of
engineering and analysis to the more scientific.
Both are required in the development of industrial
Since World War II engineering education has
moved strongly towards analysis, with the intro-
duction of courses which analyse individual pro-
cess operations and phenomena. Transport phe-
nomena, unit operations, process control, thermo-
dynamics and other engineering science courses
greatly strengthened engineering education by
showing how things are and how they work.
Unfortunately, there was not a parallel de-
velopment in the teaching of synthesis. The teach-
ing of how things ought to be rather than how
they are. This deficiency has been recognized for
years, but the remedy awaited the development of
sufficiently general principles about which to or-
ganize educational material.
The dominance of engineering science reveals
a natural tilt in the educational landscape towards
course material which possesses a natural organi-
zation. At the University, well organized and
easily taught material rises to the surface.
A course in transport phenomena has the
natural organization of the equations of change,
thermodynamics has the first, second and third
laws, unit operations can be organized about the
various processing operations and so it is with
the courses which dominate. This kept synthesis
in the background in spite of the important role
synthesis plays in the practice of engineering.
Methods of synthesis were not well organized and
easily taught, and for this reason synthesis could
not rise to the surface.
In the late 1960's and early 1970's, research
in process synthesis established the broad out-
lines of this field and it became apparent that a
careful interlacing of synthesis and analysis is a

proper way to approach process development. In
these theories of process development, each syn-
thesis step defines an analysis problem the solu-
tion of which provides data required for further
synthesis steps. Further, it became apparent that
this organization of alternating synthesis and
analysis steps begs the development of educational
material for the early stages of engineering edu-
In this report we examine the development of
a first course in engineering in which synthesis
and analysis are taught simultaneously. Elemen-
tary new principles of process synthesis are com-
bined with the classic analysis techniques of
material and energy balancing. Emphasis is on
the development of process technology rather than
on the analysis of existing processes.

Research in process synthesis has been ade-
quately reviewed elsewhere and it would be a
distraction here to go into any great detail.*
However, a brief discussion of the broad ways of
thinking found in the research literature is neces-
sary to set the stage for the educational develop-
ments. This, then, is a most superficial review of
the emerging research leading to principles for
the development of the process flow sheet. Three
approaches to process synthesis are apparent;
problem decomposition, evolution and optimiza-
In process synthesis by optimization a cam-
bined design is proposed which is known to be
redundant, containing process equipment and
material flows in a greater number and diversity
than is reasonable. However, somewhere hidden
within the combined process design is an eco-
nomically reasonable process. The desired process
flow sheet is exposed by the methods of mathe-

* J. E. Hendry, D. F. Rudd, J. D. Seader, AIChE J.
(to appear)


Dale F. Rudd was born, raised and educated in Minne-
sota, receiving a Ph.D. degree in Chemical Engineering
from the University of Minnesota in 1959. Most of his
professional life has been spent at the University of
Wisconsin where he is now Professor of Chemical Engi-
neering. His research and teaching efforts have had a
substantial influence on the chemical engineering pro-
fession in the United States and abroad. He has been
recognized as an innovative and creative teacher, author,
and lecturer.
Specializing in research on methods of process design,
Professor Rudd has been a frequent consultant and lec-
turer in industry. In addition to contributions to the
research literature, he is co-author of Strategy of Process
Engineering and the forthcoming Process Discovery. In
his books he has integrated modern mathematical con-
cepts, a knowledge of industrial processes, and recent
advances in econometrics to produce a novel treatment
of process design and analysis.
Professor Rudd is the recipient of awards from the
Canadian Institute of Chemistry, the AIChE, and the
Mexican Institute of Chemical Engineers. He serves on
the editorial boards of AIChE Journal, International
Chemical Engineering, and Chemical Engineering Com-

matical programming which trim away all but
the flow sheet sought. This approach to design
synthesis has been quite successful in specific
problems such as heat exchanger network syn-
In process synthesis by evolution one begins
with a reasonable design, a base design, which
may not be the economically optimal one. Methods
are then developed to detect the weak parts of
the design, pointing to reactors which ought to
be replaced, separation operations which ought
to be reordered and so forth. From the base de-
sign there evolves better process designs.
Process synthesis by problem decomposition
takes a completely different look at the problem,
leading to the natural organization sought in the
teaching of process synthesis. In decomposition a
synthesis problem is decomposed into a sequence

Science aims to analyze natural phenomena,
and engineering aims to synthesize
preferred situations from
natural phenomena.
of smaller and simpler problems which when
solved generate the flow sheet for the original
process development problem. The success of this
approach depends on the accuracy with which
the simpler problems can be identified. J. J. Siirola
and G. J. Powers are largely responsible for the
development of this approach.
In brief outline, flow sheet synthesis by prob-
lem decomposition occurs by the solution to these
basic problems. First the chemical reaction path
is established. This is the sequence of reactions
which best transforms the raw materials into the
products on the industrial scale. The second prob-
lem is one of species allocation in which a mapping
of material flow is proposed from raw material
and reaction site sources to product, waste and
reaction site destinations. During species alloca-
tion the easiest set of separation problems are
sought. The third problem is the selection of the
physical and chemical phenomena which best ac-
complish the separation problems which arose
from the species allocations. The final problem is
task integration in which the several separate
reaction and separation phenomena are integrated
by the reuse of energy and material.
In the following section we give a glimpse of
the detailed approach to process synthesis by
problem decomposition, and emphasize the teach-
ing of these methods as the first course in engi-
We now focus attention on undergraduate
education in engineering, which in recent history,
has a strong emphasis on the analysis of specific
processing phenomena and operations. Our con-
cern is to complement these courses with an in-
troductory course which shows how these specific
fields fit into the larger plan of process develop-
ment. Along with this orientation, certain basic
methods of engineering synthesis and analysis
are taught.
We have had several years of classroom ex-
perience with this approach at the freshman and
sophomore level. In mid 1973 Prentice-Hall Inc.
will publish Process Synthesis by D. F. Rudd,
G. J. Powers and J. J. Siirola.* The chapters in
this book are now reviewed.
*D. F. Rudd, G. J. Powers, and J. J. Siirola, Process
Synthesis, Prentice-Hall, Englewood Cliffs N.J. (1973).


Engineering of Process Systems

First the students become acquainted with a
bit of the history of processing to see how chem-
istry is used, how materials are separated and
how economy is reached by the reuse of material
and energy. In class we discuss the recovery of
nitrogen from the atmosphere for use as a ferti-
lizer, the conversion of vegetable protein into
meat analog foods for human consumption and
the treatment of sewage at the South Lake Tahoe
plant, the most advanced sewage treatment facility
in the world.
On their own, as home problems, the students
prepare a short history of important process
innovations. One project is to report on the events
which led from Sir Alexander Fleming's discovery
of penicillin to the massive process development
campaign to provide enough penicillin to accom-
pany the troops during Eisenhower's landing in
Europe. A second project begins with this beer
recipe from "Hints to Brewers," 1702.

"Thames water, taken up about Greenwich at low
tide, when it is free of all the brackishess of the sea
and has in it all the fat and sullage from this great
city of London, makes a very strong drink. It will of
itself ferment wonderfully, and after its due purga-
tion and three times stinking, it will be so strong that
several sea commanders have told me that it has often
fuddied their murriners"

and traces the development of the activated sludge
process for sewage treatment.
The students get a feel for processing, and
discover the role that the engineer plays in sup-
plying society's needs for food, water and other
material items.

Reaction Path Synthesis

The first major subproblem in process syn-
thesis involves the selection and analysis of the
chemistry of processing. The chemistry links the
products of processing with cheap and readily
available materials, and generates wastes and by-
products with which we must contend. In this
course we do not teach chemistry: we teach the
assessment of chemical change.
For example, given these reactions which lead to vinyl

C2H2 + HC1 -> C,2HC1
C2H4 + C12-> C2H4C12
C2H1C12 --> C2HC01 + HCI
2HC1 + 02 + CH- -> CH4C01, + H,20

. . research in process synthesis established the
broad outlines of this field and it became apparent
that a careful interlacing of synthesis and analysis
is a proper way to approach process development.

the students learn to select the reaction sequences which
require for each two molecules of vinyl chloride,

a) two molecules of ethylene, one of chlorine and
oxygen. This reaction path is to generate no by-
products other than water.
b) one molecule of ethylene, one of acteylene and one
of chlorine. No by-products are allowed for this

The students also learn the basic principles of eco-
nomic screening of the reaction paths they synthesize.
The idea is that the products must be more valuable than
the reactants if a reaction path is to be of any commercial
Home problems include an analysis of the chemistry
involved in these areas: the recycle of pickling liquor
waste, acrylonirite from propylene, titanium by chlorina-
tion, waste recycle in the solvay process, sodium bicarbon-
ate by the solvay chemistry, superphosphate fertilizer,
leaching copper ores, caustic manufacture the leBlanc
process, soap manufacture, nitrogen fixation, fire retard-
ant production, urea synthesis, and so forth.
In summary, the students are acquainted with
the chemistry of processing, synthesize alternate
reaction paths from given chemistry and perform
early economic screening of the chemistry.

Materials Balancing and Species Allocation

Once the chemistry of processing is at hand,
the problem arises of supplying the reactor feeds
and disposing of the reactor effluents. This leads
to the next major subproblem in process synthesis,
species allocation. The purpose of species alloca-
tion is to identify the routes that the several
species involved ought to take to support the
chemistry and at the same time lead to the simp-
lest separation problems. Given the needs of the
chemistry, we allocate species for easy separation.
As a trivial example, suppose we are to manufacture
sulfur dioxide by the direct oxidation of sulfur. Air and
sulfur are the reactant sources and pure sulfur dioxide is
needed as the product. A cold inert gas is needed to lower
the reaction temperature. Figure 1 shows two allocations,
one recycling sulfur dioxide as the inert and the other
using nitrogen as the inert. The one allocation leads to
the need to separate nitrogen from oxygen, a difficult
problem and the other leads to the separation of sulfur
dioxide from nitrogen, an easy problem. This illustrates
how the engineer assesses the nature of separation prob-
lems and uses this information to impose the proper
material flow on the emerging process flow sheet.


Allocation I


sulfur Ng waste

Allocation 2

N2 waste

Fig. 1. Alternate Species Allocations During Sulfur Oxidation.

Before anything can be taught about species
allocation as a technique in process synthesis, the
students must have skill in material balancing.
We spend the bulk of this chapter on material
balancing, and it is only towards the end of the
chapter that the idea is introduced that the real
purpose of material balancing is to identify those
allocations which involve easy separations.
Typical problems on material balancing and
species allocation involve sugar refining, oil and
meal from seed, soluble coffee powder production,
phosphate rock benification, city refuse compost-
ing, fish liver extraction, glycerine desalting, and
the manufacture of styrene.

Separation Technology
Species allocation is directed towards the
identification of the easiest separation problems
which have to be solved to support the process.
This leads to the need for an understanding of
the means by which materials are separated from
each other. One seeks to identify the ways in
which the materials differ, and develop equipment
to exploit the proper differences. In this chapter
we examine some of the separation processes,
mainly to get an understanding of how the equip-
ment works.
Equipment separating solids from solids are examined,
along with equipment which accomplish separations based
on volatility differences, and solubility differences. For
example we examine the Nowak-Othmer scheme for sepa-
rating the redmud waste from the Beyer Process, which

is a mixture of Al203, Fe20 SiO,, TiO2, Na20 and
CaO, by taking advantage of the differences in chloride
forming affinity of the oxides, differences in volatility of
the metal chlorides, and differences in water solubility.
All of this orientation in separating technology is sup-
ported by material balance calculations and provides the
background required to select the basis of separation,
discussed in the next chapter.
Typical problems include municipal solid waste recycle,
nut meat separation, activated carbon from saw dust,
cranberry draining, copper sulfate leaching, desalting sea
sand, and the Caban-Chapman process of mercury re-

Strategy of Task Selection
Having established the chemistry of proces-
sing, outlined the flow of material through the
process, and obtained an understanding of the
general means of separation, the students are
prepared for the study of the heuristic principles
of task selection. We examine the ways in which
separations ought to be performed to achieve
efficient processing.
Techniques are presented for examining the
ordered lists of physical and chemical properties
of materials to detect the best plan of separation.
Table 1 shows the species present in the reactor
effluent in the Sinclair-Koppers 500 million pounds
per year light olefin process; the species ordering
is according to volatility, the property to be ex-
ploited to separate the methane and hydrogen
for use as fuel, the ethylene and the propylene as
products, the ethane and propane for recycle, and
the heavies for processing into motor fuels. What
principles lead to the discovery of a sequence of
distillations to accomplish the separation?

Table 1. Reactor Effluent Ranked According to Volatility.
(Sinclair-Koppers 500 million lbs/yr ethylene process)
Boiling point
Species Amount Boiling point difference
Hydrogen H2 18% -2530C 920C
Methane C1 15 -161 57
Ethylene C2 24 -104 15
Ethane C2 15 89 41
Propylene C3 14 48 6
Propane C, 6 42 42
Heavies C+4 8 70
To be separated into: 1-Hydrogen-methane; 2-Ethylene;
3-Propylene; 4-Ethane-propane; 5-Heavies.

In this chapter we establish the following
general principles
1. Of the many differences which may exist between
the source and destination of a stream, differences
involving composition dominate. Select the separa-
tion tasks first. This heuristic is based on the idea
that attention should not focus on differences in


pressure, size, temperature, and other bulk prop-
erties until after the means of separation has been
2. When possible reduce the separation load by stream
splitting and blending. Here we reduce as much as
possible the amount of separation to be performed,
as is done in the blending of gasolines to each
desired product specifications.
3. All other things being equal, aim to separate the
more plentiful components early. This reduces the
load on the down stream units.
4. Remove the corrosive and hazardous material early.
The remaining separations can then be done in less
elaborate equipment.
5. The difficult separations are best saved for last.
The extremely expensive separations should not be
performed in the presence of material which need
not be there.
6. All other things being equal, shy away from sepa-
rations which require the use of species not norm-
ally present in the processing. This favors separa-
tions driven by the addition or removal of energy,
such as distillation, over those driven by the addi-
tion and removal of material not normally present
in the processing, such as extraction.
7. Avoid excursions in temperature and pressure, but
aim high rather than low. In distillation favor the
removal one-by-one of the more volatile components.
This rule derives from the high costs of low tem-
perature and low pressure operation compared to
pressure and temperature operation.
8. Favor the removal of products from the least harsh
environment. For example, the final products in a
distillation sequence ought to come from the top
of the towers, rather than the bottoms where de-
gredation may occur.

The application of these heuristics to a large
extent identifies the separation sections of indus-
trial processes. For example, the student examin-
ing Table 1 would arrive at these conclusions.

a) the boiling points of propane (420 C) and propy-
lene (48C) are very close, as are the boiling points
of ethane (89C) and ethylene (104C). Thus
the C, splitter and the C, splitter ought to be last
to place the most difficult separations in that posi-
b) to get at the C2 and C3 fractions species more
volatile and less volatile must be removed. We
favor the removal of the more volatile methane
and hydrogen to raise the temperature of coolant
needed to drive subsequent distillations. This will
reduce utilities costs substantially.
c) the ethylene and propylene ought to be tower top
products to insure the removal of degradation pro-

These heuristics explain the gross features of the
Sinclair-Koppers process the flow sheet of which is shown
in Figure 2.
This kind of heuristic reasoning leads to an under-
standing of the separation sections of a wide variety of

cLC4 L--c

Fig. 2. Sinclair-Koppers Separation Flowsheet.
Typical home work problems involve the processing
of cheese whey, crushed rock screening, high purity beryl-
lium from the oxide, detergent manufacture, allyl chloride
production, the development of screening circuits and an
examination of many other industrial processes.

Task Integration

The skelton of the process has been discovered
when the processing tasks were defined in the
previous chapters. Many of these tasks involve
the addition and removal of energy, and these
are costly operations. The last subproblem of pro-
cess synthesis introduced is task integration where
we free the process as much as possible from the
purchase of heating and cooling service. We seek
to have one task in the process drive other tasks,
by the integration of their operation.
Before task integration can be attempted, the
elementary principles of energy balancing must
be presented. We show how the heat energy added
to a system and the work done by a system are
related to the enthalpy changes which occur in the
material passing through the system. Then, meth-
ods of estimating the enthalpy change caused by
temperature change, phase change, and chemical
change are presented. This gives the quantitative
background required for the understanding of the
energy management principles to be discussed.
We show how energy management principles
are used to supply heat to endothermic reactions,
to remove heat from exothermic reaction, and to
utilize the energy of product recovery. For ex-
ample, the endothermic reaction of calcining lime-
stone can be driven by matching it with an exo-
thermic reaction, the combustion of coal: the be-


l.A' ~


6ives students the

tools theq need for

aunulsis and design.

Russell/Denn shows students the essential in-
terrelationships between experiment, modeling
and engineering design. And it presents all three
in a consistent, logical approach.
First, Russell/Denn presents isothermal react-
ing liquid systems to show the direct relation-
ship between bench-scale experiment and
reactor design.
Next it introduces the rate of reaction, rate of
mass transfer, and its application.
Then it analyzes isothermal systems and the rate
of heat transfer.

Russell/Denn shows students how to use chem-
istry and elementary calculus to solve problems
from the start.
And every time Russell/Denn introduces a phy-
sical situation it begins with an experiment and
data to compare model behavior and experi-
mental evidence.
The results?
Students develop mathematical descriptions for
various situations-a skill they must have to in-
corporate analysis into design.

By T. W. Fraser Russell and Morton M. Denn, both of University of Delaware
1972 502 pages $16.95

For a complimentary copy, contact your Wiley representative or write to T. R. Poston,
Dept. 1181, N.Y. office. Please include course title, enrollment and present text.

605 Third Avenue, New York, N.Y. 10016. In Canada: 22 Worcester Road, Rexdale, Ontario


price subject to change without notice

ginner can calculate how much coal is needed per
ton of limestone processed, obtaining a more com-
plete picture of raw material costs in lime pro-
The synthesis of networks of heat exchangers
to recover sensible energy is discussed using the
simple but approximate criteria of avoiding heat-
ing and cooling with external sources of these
expensive utilities.
Typical problems include freezing of fish fillets,
gas desulfurization, energy storage in checker
brick regenerators, cost of sugar evaporator op-
eration, and processing of junk cars by freeze

Fresh Water by Freezing

The scope of each of the preceding chapters was
limited in turn to a specific aspect of process synthesis.
Such limitations are necessary and desirable to present a
compact package of knowledge which can be digested with-
out too much difficulty. In practice no field is as clean
and orderly as textbooks tend to describe it to be. This is
certainly true of the field of process synthesis.
For one thing, few, if any, real problems are equally
balanced among the several topics presented in the earlier
chapters. Some problems are dominated by the chemistry,
others by task sequencing still others by a critical separa-
tion, and many by the integration of energy and material
use. Further, an experimental program usually parallels
process synthesis for rarely is sufficient information to
be found in the literature on phenomena to be exploited
in a new area of processing. Finally, the early discovery
stages of engineering are not sufficiently detailed to arrive
at just one process, and several alternatives survive to
undergo the detailed engineering studies necessary to
determine economic and engineering feasibility.
Thes2 things are now brought into focus by tracing
the development of a process to obtain fresh water from
brackish water. We risk the appearance of contrived
spontaneity by condensing the years of work of a number
of engineers into a scenario of process discovery. How-
ever, it is important for the reader to come close to the
act of discovery, even if the discovery is only one simu-
lated by the authors.
This chapter begins with the observation that the ice
crystals formed in brackish water are' salt free, and ends
with processes now commercially available for the pro-
duction of fresh water. This engineering problem is
dominated by task integration problems, for it is the
cost of energy which dominates the water cost. Figure 3
is a schematic drawing of the vacuum freezing-vapor
compression process, the synthesis of which is a prob-
lem in task integration discussed in this chapter.

Detergents from Petroleum

In this chapter we apply the principles of process
synthesis to the development of part of the technology to
convert crude oil into detergents. No new ideas are pre-
sented, we apply what we already know. Our attention



1. Vaporization of sea water at triple point causes
ice formation.
2. Vapor condenses on clean ice surface to form fresh
3. 5% of fresh water used to wash rising piston of ice.
4. Ice-brine slurry pumped to bottom of ice wash
Fig. 3. Vacuum Freezing-Vapor Compression Process.
is focused on the problem of converting a kerosene fraction
of the crude oil into an intermediate material, a chlori-
nated hydrocarbon, which then fits into the large cam-
paign of detergent production.
Given fragmentary conversion data on these reactions
CloH_, + Cl2 -> C10H.,Cl + HC1
v-ith the side reaction
C1oH21Cl + C, -> CloH20C1, + HC1
We seek the technology to produce monochlorodecane eco-
nomically on a commercial scale. Dominating this process
synthesis problem are problems in species allocation and
task selection,.
To give some idea of the tenor of this chapter, we
sketch out part of the development of the final process
flow sheets. Figure 4 shows five ways of separating the
reactor effluents shown in Table 2 into an HC1 waste,
C1,-decane recycle, monochlorodecane product, and dichlo-
rodecane waste. How might the beginning student select
among these?


C12- DEC

~__] MCD
d) [ HCI



__ C,1- DEC

Fig. 4. Alternative Separations.


The dominance of science oriented courses in undergraduate education is caused by the very important
role analysis has in engineering . The absence of the organizing influence of basic principles of synthesis
is possibly the major reason why analysis tended to dominate education.

Table 2. Effluent from Photochlorination Reactor

Moles Boiling point

0.95 2150C
0.05 241
4.0 174
trace -34
1.05 -85

Applying the task selection heuristics
Plans a and c lead to low amounts of material pro-
Plans a and b lead to low cooling costs by the early
removal of volatile material
Plans a and c reduce processing costs by minimizing
the amount of material processed during the most
costly separation.
Plans a and b reduce costs by removing corrosive
materials early
Plans a and c remove the MCD on a distillate rather
than bottoms product
Hence, by applying these simple rules it appears that
structure a is favored for a number of reasons. However,
much more synthesis must be performed to develop better
processing alternatives.
Figure 5 shows eight flowsheets which are synthesized
using the principles developed in this early course. Figure
6 shows the economic analysis against which these pro-
cesses are compared.
Flowsheet 2, which has the separation sequence for the
last two columns in Flowsheet 1 reversed, is less attrac-
tive than 1 for all values of conversion. The decane is
the largest component in the reactor effluent, and is also
the lowest boiling component leaving the phase separator.
The failure to remove decane early causes the increased
cost in this case.
Flowsheet 3 involves the use of decane as a solvent to
remove chlorine from the vapors leaving the phase separa-
tor. Since the chlorine not converted in the reactor is re-





^ DEC ____



Fig. 5. Flowsheets synthesized using the principles. (Reproduced
by permission of Chem. Eng. Prog. 68, No. 9. pp. 91-92, 1972)

Fig. 6. Venture Profit Surfaces for Monochlorodecane Processes.

cycled, this flowsheet is not nearly as sensitive to con-
version or chlorine cost as are Flowsheets 1 and 2. The
maximum profit for Flowsheet 3 is not as high as the
optimum for Flowsheet 1, but it is much less sensitive to
the level of conversion. If uncertainty in the reactor design
is large, the selection of Flowsheet 3 may be wise. The
addition of a separator to recover unconverted chlorine
adds to the equipment and operating costs but gives much
greater flexibility in operation.
Flowsheet 4, which utilizes the reverse separation se-
quence of Flowsheet 1, is decidedly less attractive. The
repeated processing of the low boiling components, par-



Cit CL2

,C DICoco

H CL k2 CL2'H 0E- !


OEC 1-- -J --- CD C L,



Hydrogen chloride



ticularly HC1, causes very high equipment and utility
costs. The loss of chlorine, as in Flowsheets 1 and 2
makes the profit for this process dependent on reactor
conversion and chlorine cost.
Flowsheet 5 shows the same independence of conver-
sion as Flowsheet 3. The profit is less due to solvent costs
and the additional separator.
A profit nearly as high as that for Flowsheet 3 is ob-
tained for Flowsheet 6. Flowsheet 6 utilizes water to
remove hydrogen chloride from the vapors leaving the
phase separator. Wet chlorine returning to the reactor
inlet is dried by contacting concentrated sulfuric acid.
The low cost of the solvent, water, and the fact that the
water need not be regenerated make this alternative very
attractive. However, the corrosive nature of the hydrogen
chloride-water mixture and the sulfuric acid-water mix-
ture require more expensive materials of construction.
The danger of water entering the reactor must also be
considered. If the reactor and separators are constructed
of carbon steel, any traces of water in the presence of
the hydrogen chloride reaction product would lead to
rapid failure of these units due to corrosion.
Flowsheet 6 has a higher profit than Flowsheet 3 at
low conversion. This because of the large amounts of
decane solvent required in Flowsheet 3 for low chlorine
Flowsheets 7 and 8 are the special cases when con-
version is complete. Flowsheet 7 has the advantage of
only requiring two separators and no recycle. The high
conversions using stoichiometric feed ratios lead how-
ever, to the production of large amounts of DCD. Hence,
the profit for this alternative is low.
Flowsheet 8 represents the situation when large ex-
cesses of decane are utilized. Complete conversion of
chlorine is achieved, hence no separation is required for
its recycle. A small amount of DCD appears in the
product. These advantages are offset by the costs asso-
ciated with the separation and recycling of large amounts
of decane.
In summary, three regions exist when comparing the
eight flowsheets over wide ranges of conversion. At low
conversions process Flowsheet 6, which utilizes water as
a means to separate HCI and chlorine, is best. At inter-
mediate conversion levels, process Flowsheet 3 is best.
Process 3 utilizes the feed and recycle decane as a solvent
for absorbing chlorine from the waste hydrogen chloride
stream. For high levels of conversion, process Flowsheet
1, which discards unconverted chlorine, is optimal.


An essential difference between the aims of
education in science and in engineering ought not
be lost. Herbert A. Simon states in the Sciences
of the Artificial
Historically and traditionally, it has been the task of
the science disciplines to teach about natural things:
how they are and how they work. It has been the
task of the engineering schools to teach about arti-
ficial things: how to make artifacts that have the de-
sired properties and how to design. . Everyone
designs who devises courses of action aimed at chang-
ing situations into preferred ones.

Science aims to analyse natural phenomena, and
engineering aims synthesize from natural phen-
omena preferred situations. For example, while
it may be sufficient for a scientist to develop an
understanding of the phenomena of ice formation,
the engineer ought to be concerned with use of his
knowledge in the development, say, of an artifact
which economically can produce fresh water from
the sea by freezing.
The dominance of science oriented courses in
undergraduate education is caused by the very
important role analysis has in engineering, and is
also caused by the availability of well organized
and carefully planned text material in these areas.
The natural organization of methods of analysis
lends itself to course development and text prepa-
The absence of the organizing influence of
basic principles of synthesis is possibly the major
reason why analysis tended to dominate education.
It remains to be seen if the course described will
bring the proper balance between analysis and
synthesis in engineering education. E

LETTERS: (Continued from page 3)
A little checking has revealed that the non-zero value
is the result of a reformulation of steam table data pub-
lished by the ASME in 1967.1 Recognizing that a stable
liquid-vapor equilibrium can't exist below the triple point
temperature of 32.018F, the reference conditions were
changed so as to assign the value zero to the entropy
and internal energy of the saturated liquid phase at the
triple point. This, of course, forces the associated enthalpy
to be (Pv) units greater or about 0.0003 Btu/lbm. One
can certainly estimate the conditions for the metastable
vapor-liquid equilibrium at precisely 32'F through simple
extrapolation and obtain values of about -0.0181 Btu/lbm
and -0.0178 Btu/lb,,,, respectively, for the internal energy
and enthalpy of the liquid phase The error in the Com-
bustion Engineering tables, then, is the omission of the
minus sign together, perhaps, with out spelling out the
convention not stating (as does the reference above) that
the first entry in the saturation table corresponds rigor-
ously to a metastable state. At least one recently published
text2 on chemical engineering thermodynamics has copied
this error, but the new edition of "Steam Tables" by
Keenan, Keyes, et al,3 does note both the negative values
and metastability below the triple point.
Kenneth R. Jolls
Iowa State University

1. "Thermodynamic and Transport Properties of
Steam," American Society of Mechanical Engineers (1967).
2. Balzhiser, R. E., M. R. Samuels and J. D. Eliassen,
"Chemical Engineering Thermodynamics," Prentice-Hall,
Inc (1972).
3. Keenan, J. H., F. G. Keyes, P. G. Hill and J. G.
Moore, "Steam Tables," John Wiley & sons, Inc. (1969).








They're Sun Oil recruiters.
You might meet one of them
face-to-face in a campus job
If you do, they'll tell you straight
about the petroleum industry, Sun
Oil Company and job opportunities.
Any other way, and you took a
job with Sun, you probably wouldn't
stay very long.
Take Ernie (he's black). He'll tell
you being black doesn't make very
much difference at Sun. He could
suggest that sometimes it helps,
citing the recent action of placing 25
million dollars of group insurance
with an all-black company.
Sun's not color-blind, or any

other kind of blind. It's just that in
our work, color or race or religion or
sex (within certain limitations)
doesn't have anything to do with it.
Then about the industry. Sure,
there's some pollution, but we're
working and spending like sin to
minimize it. We're also necessary.
Without gas, oil and grease for fuel,
power and lubrication America
would be in a mess.
Another point. Our industry is
responsible for nearly 1 /2 million
jobs. That's a lot of opportunity.
So-if you've got the urge to
get into petroleum, and want to
check out one of the industry's
exciting, progressive companies,

see Earl Pearce, Ernie Harvey,
Joe Pew or Harley Andrews. Your
Placement Director will know when
they'll be on campus.
For more information, or a copy
of our Career Guide, write SUN OIL
COMPANY, Human Resources Dept.
CED, 1608 Walnut Street,
Philadelphia, Pa. 19103.

An Equal Opportunity
Employer M/F


The energy to keep straining toward your chosen goal-and even
as you attain it, look forward to the ones beyond.
The energy to explore, evaluate, create, bring needed changes.
Energy to burn, figuratively-that wealth possessed by the
young, in mind no less than body.
Energy to burn, literally, because ideas-freedom, equality, well-being,
conservation of our natural environment-must be turned into
realities-food, shelter, warmth, access, economic independence
and the physical means to accomplish our goals.
Atlantic Richfield is an energy company-in all these ways. One of the
nation's thirty leading industrial corporations, and one of the
ten companies producing most of our energy needs. A company that
is forward-looking in management. Imaginative in organization and
operation. Open to fresh thinking. Responsible in outlook.
While our specific requirements continually change, we typically
offer opportunities to financial and systems analysts, accountants,
auditors, engineers, geologists, geophysicists, sales representatives,
agronomists and programmers.
We invite your interest. See our representative on campus or your
Placement Director.

AtlanticRichfieldCompany *
An equal opportunity employer M/F.

Full Text