Citation
Chemical engineering education

Material Information

Title:
Chemical engineering education
Alternate Title:
CEE
Abbreviated Title:
Chem. eng. educ.
Creator:
American Society for Engineering Education -- Chemical Engineering Division
Publisher:
Chemical Engineering Division, American Society for Engineering Education
Publication Date:
Frequency:
Quarterly[1962-]
Annual[ FORMER 1960-1961]
quarterly
regular
Language:
English
Physical Description:
v. : ill. ; 22-28 cm.

Subjects

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

Notes

Citation/Reference:
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:
01151209 ( OCLC )
70013732 ( LCCN )
0009-2479 ( ISSN )
AA00000383_00025 ( sobekcm )
Classification:
TP165 .C18 ( lcc )
660/.2/071 ( ddc )

UFDC Membership

Aggregations:
Chemical Engineering Documents

Downloads

This item has the following downloads:


Full Text





ceia engineering educaio







CHEMICAL ENGINEERS

GET TOTALLY

INVOLVED IN A

TOTAL ENGINEERING

ENVIRONMENT

AT ESSO


Esso Research and Engineering ultimate goal is the same as that cations and Technical Services,
Company, the principal technical of the university; namely the ex- Process Engineering, Project De-
affiliate of Standard Oil Company tension of knowledge and the bet- sign or Process Selection and
(N. J.), provides research and en- terment of the human condition Economics, the Chemical Engi-
gineering services to 250 world- through long-term fundamental neer serves with his professional
wide affiliates with assets of over and applied research, and the peers. He learns from them; he
thirteen billion dollars, accomplishment of immediate ob- teaches them. But he advances
The Chemical Engineer plays a jectives through the economical as far as his own talents take
vital role in helping us meet these design and operation of plants him, wherever his interests lead
vast responsibilities. But most and equipment. him; either in a technical or ad-
important to him, he functions in Whether he possesses a B.S., an ministrative capacity.
an environment as dedicated as M.S., or a PhD., and whether he Total involvement in a total
that of the university Chemical works in Product/Process Re- chemical engineering environ-
Engineering department. For our search and Development, Appli- ment. That's Esso.
For full details on the opportunities available, contact:
Dr. P. H. Watkins, Employment Coordinator, Dept.
ESSO) ESSO RESEARCH AND ENGINEERING COMPANY
P.O. BOX 175, Linden, New Jersey 07036
An Equal Opportunity Employer (M/F)









Chemical Engineering Education


EDITORIAL AND BUSINESS ADDRESS
Department of Chemical Engineering
University of Florida
Gainesville, Florida 32601


VOLUME 3, NUMBER 3


Departments


Editor: Ray Fahien 107 Report to our Readers
Editor: Ray Fabian


Associate Editor: Mack Tyner

Business Manager: R. B. Bennett


Publications Board and Regional
Advertising Representatives:

CENTRAL: James Weber
Chairman of Publication Board
University of Nebraska
Lincoln, Nebraska 68508
WEST: William H. Corcoran
California Institute of Technology
Pasadena, California 91109
SOUTH: Charles Littlejohn
Clemson University
Clemson, South Carolina 29631
SOUTHWEST: J. R. Crump
University of Houston
Houston, Texas 77004
EAST: Robert Matteson
College Relations
Sun Oil Company
Philadelphia, Pennsylvania 19100
E. P. Bartkus
Secretary's Department
E. I. du Pont de Nemours
Wilmington, Delaware 19898
Peter Lederman
Brooklyn Polytechnic Institute
Brooklyn, New York 11201
NORTHEAST: James M. Douglas
University of Massachusetts
Amherst, Massachusetts 01002
NORTH: J. J. Martin
University of Michigan
Ann Arbor, Michigan 48104
NORTHWEST: R. W. Moulton
University of Washington
Seattle, Washington 98105
UNIVERSITY REPRESENTATIVE
J. A. Bergantz
State University of New York
Buffalo, New York 14200
PUBLISHERS REPRESENTATIVE
D. R. Coughanowr
Drexel University
Philadelphia, Pennsylvania


108 Letters

110 The Educator
Dean Max Peters


124 Views and Opinions
The Dilemma of Innovating Societies,
A. G. Frederickson

126 The Classroom
Transport Phenomena Equations of Change
V. J. Lee

118 The Laboratory
A Microcatalytic Tracer Experiment, R. W.
Neumann, S. E. Riffle, S. T. Swenson, and
J. W. Hightower.

128 Book Review

142 Problems for Teachers

150 Departments of Chemical Engineering
University of Florida,
Ray Fahien

160 Division Activities
154 The Curriculum
Flexible Curricula Can Be Strong,
Ray Fahien, Mack Tyner, R. A. Keppel.

Feature Articles
113 Optimization Applications and Limitations,
R. R. Hughes

130 A Self-pacing, Auto-graded Course,
G. D. Shilling.

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 DeLand, 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., 137 E. Wisconsin
Ave., DeLand, Florida 32720. Subscription rate U.S., Canada, and Mexico is $10 per
year to non-members of the ChE division of ASEE and $6 per year to members.
Individual copies of Vol. 2 and 3 are $8 each.


SUMMER 1969


SUMMER 1969


















LOCATIONS HAVING
CURRENT OPENINGS


Olin
MAJOR PRODUCTS
PRODUCED


DISCIPLINE
REQUIREMENTS


TYPE OF WORK
PERFORMED


Chlor-Alkali Products
Ammonia Process Development,
Augusta, Ga. Phosphates Design, Maintenance,
Brandenburg, Ky. Urea Planning, Scheduling,
Charleston, Tenn. Nitrogen ChE Production, Sales,
Joliet, III. Acids ME Production, Sales,
CHEMICALS Lake Charles, La. Hydrazine IE Accounting,
-Inorganic Little Rock, Ark. Petrochemicals Chemistry Marketaing,
-Organic & McIntosh, Ala. Insecticides Accounting Financial Analysis,
Specialty New Haven, Conn. Pesticides Business Adm. Distribution,
-Agricultural Niagara Falls, N.Y. Polyurethane Transportation Project Engineering
Pasadena, Texas Carbon Dioxide Marketing (Plant Startup &
Rochester, N.Y. Animal Health Construction),
Saltville, Va. Products Research Engineering,
Automotive Chemicals Technical Service
Other derivatives

Alumina ChE
Burnside, La. Aluminum IE Manufacturing
METALS Chattanooga, Tenn. Aluminum Extrusions ME Production
-Aluminum Gulfport, Miss. Aluminum Sheet, Plate, Metallurgy Sales
-Brass Hannibal, Ohio Coils Met. Engineering Mae -
-Ormet, Corp. East Alton, II1. Brass Fabricated Parts Accounting Maintenance
New Haven, Conn. Sheet & Strip Brass Business Adm. Finance
Sedalia, Mo. Roll Bond Ind Tech. Metals R&D
Wire & Cable Ind. Mgmt.

Carbonizing Paper Marketing
Fine Printing Papers ChE Process Engineering
FOREST PRODS, West Monroe, La. Specialty Paper Chemistry Plant Engineering
PAPER & FILM Pisgah Forest, N.C. Products Pulp & Paper Research & Dev.
-Olinkraft, Inc. Covington, Indiana Cigarette Paper & Tech. Statistician
-Ecusta Filters IE Systems Engineering
-Film Cellophane ME Production
Kraft Bags Mathematics Management
Kraft Paper Business Adm. General IE
Kraftboard Cartons Accounting Design and
Corrugated Containers Development
Olinkraft Lumber Accounting


East Alton, III.
New Haven, Conn.
Marion, III.
Kingsbury, Ind.


Sporting Arms
Ammunition
Powder Actuated tools
Smokeless Ball
Powders
Solid Propellants
Safety Flares
Franchised Clubs


Ind. Tech.
IE
ME
Mathematics
ChE
Accounting
Business Adm.
Marketing
Personnel Mgt.
Physics
Ind. Mgmt.


Production Control
Purchasing
Manufacturing
Plant Engineering
Sales
Financial Analysis.
Personnel
Marketing
R&D


PRODUCT
GROUP


WINCHESTER-
WESTERN










REPORT to our Readers
Chemical Engineering Education has now
published six issues and completed one and a half
years of publication at the University of Florida.
It is both surprising and gratifying to the editors
that the publication continues to be well-received
by the profession. Although we have expected
our luck to run out, each issue seems to bring
forth new commendatory letters and comments
and there seems to be general agreement that the
publication is filling an important need. We very
much appreciate these expressions of support,
but we realize that we are still amateurs in the
publishing business. We cannot compete with
commercial ventures or professional journals that
are published by large societies and supported by
dues.
During the past year we have again had an
excellent response from our colleagues who have
submitted manuscripts for publication. We have
not been able to accept all papers submitted to us
(rejected papers have usually been negatively re-
viewed, but a few have been deemed inappropri-
ate for our journal.) Our editorial policy in each
issue has been either to emphasize some particu-
lar theme or to try to achieve a balance among
articles. The latter means not only that we try
to have an article for every recurring department
(e.g. Chemical Engineering Classroom), but also
that we try to have represented in the issue the
various technical areas of the profession. We also
try to achieve balance among articles published
in terms of type of school (public, private, gradu-
ate-oriented, undergraduate-oriented, etc.) and
geographical location. While we have appreciated
the opportunity to publish papers from some well-
known people in the profession, we do not in any
way discriminate against lesser-known but cap-
able people from the undergraduate-oriented
schools. We feel our journal is for the profession
as a whole and not just an elite group or "in-
group" of any kind.
The above policies also apply to the selection
of our featured departments and featured educa-
tors. Here we have generally had an excellent
response from people whom we have asked to
write articles. However, there were two schools
who were unable to submit a department article
for this issue. As a result we have been forced to
get together a last minute article on our own de-
partment-even though we have assiduously
tried to avoid using articles from our own de-


apartment in the journal. (Apparently we have
been reasonably successful in not playing up our
department, since a good number of visitors have
commented on the journal without knowing we
were publishing it!)
In order to survive it is necessary that we
receive the financial support of departments, ad-
vertisers, and industrial donors. We have appre-
ciated the fact that, due to the efforts of Joe
Bergantz, nearly 100 departments are now con-
tributing to CEE! However, last fall, the pros-
pects for advertisements and donations seemed
so gloomy, that we reduced our number of pages
by eight. Since then, I am happy to report, the
yeoman efforts of Professor Weber and the Pub-
lication Board have been paying off. It now looks
like our combined industrial advertising and do-
nation income will be within a few hundred
dollars of last year's figure and could even be
more than last year! Even brighter, however,
are the prospects for the success of George Bur-
net's suggestion that the fall issue go to seniors
interested in graduate work and that we seek
departmental ads on graduate programs for the
fall issue. Professor Bergantz states that now
about 30 departments have indicated that they
intend to buy a total of $3000 of such advertising
of which the increased costs of printing additional
pages and 2000 more copies will be about half
that amount.
Incidentally, since the Fall issue of CEE will
carry paid advertisements from ChE depart-
ments and will go to seniors interested in gradu-
ate work, we felt the editorial content should em-
phasize basic areas of instruction and research in
graduate studies. Accordingly prominent chem-
ical engineering scholars will write on their grad-
uate courses in certain important areas (such as
control and optimization, thermodynamics, kinet-
ics, applied mathematics, particular systems,
etc.). Each article will outline a graduate course
and indicate important areas of research in the
field. In this manner we hope to encourage more
seniors to continue their intellectual growth in
graduate school.
R.W.F.
P.S. The Publications Board recently approved,
effective January 1970, a charge to ChE Division
members of $6/yr. each for subscriptions. Bulk
subscriptions to departments will be $4/yr. each
with a minimum charge of $25. We expect to
reach more faculty members this way and also to
generate additional income.


SUMMER 1969








NOTES TO AUTHORS:

P.S. We have a few words of instruction for fu-
ture authors. Because of the limited amount of
space available, the articles, papers or reports
submitted to CEE should be concise, lucid and
also brief. Follow nomenclature of standard text-
books or write equations or formulas clearly. Use
consistent units of measurement and give dimen-
sions for all terms.


Assume your reader has some expertise in the
field and minimize the amount of historical back-
ground included. Avoid tables and graphs which
involve duplication or unnecessary data. Fre-
quently a graph or a few typical results may be
substituted for a lengthy table.
Two copies of a paper are sufficient for review.
After a paper has been accepted for publication,
the author should send the editor a short biog-
raphy and photo of himself to use with the
article. M.T.


ACKNOWLEDGMENTS

The /mUaowUim campaW"ie^ htcdw Jaiwpao4 d heowa
eafiesUtorf Cdoa-iAw d&WiMw de 'paji ye"' dphaaih

do4Ua&OH4 in &eM 4 ad aude4&Ui.

C. F. BRAUN AND COMPANY
DOW CHEMICAL COMPANY
MALLINCKRODT CHEMICAL COMPANY
MONSANTO COMPANY
MINNESOTA MINING AND MANUFACTURING COMPANY
OLIN MATHIESON CHEMICAL COMPANY
THE PROCTER AND GAMBLE COMPANY
STANDARD OIL (INDIANA) FOUNDATION
THE STAUFFER CHEMICAL COMPANY


from the READERS

Correction from Lih
Sir: The beginning of the article on Stu Churchill (CEE
Spring 1969) clearly illustrates what the chemical engi-
neering teacher has to do these days. You have to stand
on your head to catch the attention of students. Perhaps
this is why Professor Churchill has been so successful
and has had to engage in all sorts of athletic activities
to keep it up.
The Japanese (and Chinest as well) character for
HYO (leopard) is upside down.
Marshall M. Lih
Catholic University

Statistical Study
Sir: We have made a study which attempts to relate
mathematically the number of staff members of profes-
sional rank required in a chemical engineering depart-


ment to the numbers of bachelors, masters and doctors
graduated per year. No similar study could be found
in the literature.
One purpose of this study was to analyze the relation-
ship between the number of persons of professorial rank
required in a chemical engineering program and the num-
bar of students to be graduated per year at various degree
levels. In light of the rapid expansion currently taking
place in most universities it is important that this rela-
tionship be understood in order that intelligent admin-
istrative and educational decisions and forecasts may be
attempted. In the present study this relationship is
analyzed only for chemical engineering programs since
this case was of immediate concern to the authors. The
main problem involves estimating how much of the total
variability in the number of professors of chemical en-
gineering from university to university is due to the
different numbers of degrees granted, and how much is
due to "other factors" such as
Different emphasis on research activities
Different policies concerning the amount of admin-
istrative work to be performed by the professors


CHEMICAL ENGINEERING EDUCATION









Different professor-student ratios
Different student attrition or "drop-out" rates
Different amounts of teaching done by non-profes-
sional staff members
Different numbers of students per class
Henceforth, these "other factors" will be referred to as
educational, administrative and research policies.
If it is assumed that the departments of chemical
engineering in our American colleges and universities
have the same educational, administrative and research
policies, the following equation will hold.
P = A0 + A1B + A2M + A3D (1)
where P is the number of full-time teaching schedules
required in the professorial ranks (total of Assistant,
Associate and Full), B, M, and D are, respectively, the
number of bachelor's, master's and doctoral degrees
granted per year, and A1, A2 and A0 represent respec-
tively the additional numbers of professors needed for
each additional bachelor's master's or doctoral degree
granted per year.
However, as would be expected, deviations from Eq. 1
are observed, presumably because universities do not
have the same educational, administrative and research
policies and are at considerably different stages of devel-
opment. In order to estimate the proportion of the total
variation due to the different numbers of degrees granted
and how much is due to the difference in educational,
administrative and research policies, the technique of
least squares was used.
Data from accredited chemical engineering depart-
ments of 97 universities for the academic year 1964-1965
were gathered from Chemical Engineering Faculties of
Canada and the United States, which is compiled annu-
ally by the Chemical Engineering Projects Committee of
AIChE. For each of the 97 universities, the number of
professorial schedules in chemical engineering was ob-
tained along with the number of bachelor's, master's
and doctor's degrees granted. By calculations performed
on an IBM 7040 digital computer, the least squares
estimates of A0, A1, A2, and A3 were found to be 3.40,
0.06, 0.15 and 0.43 respectively, yielding the equation
P = 3.40 + 0.06B + 0.15M + 0.43D (2)
Thus we can estimate that for every additional pro-
fessor employed, schools can grant, on the average, an
additional 16-2/3 bachelor's degrees, or 6-2/3 masters
degrees or 2-1/3 doctors degrees per year and conversely.
It should be emphasized that these figures are estimates
of average behavior and that the situation in any one
school may depart markedly from these estimated
averages.
The multiple correlation coefficient was found to be
0.79. Accordingly, one may estimate that 0.62 (which is
0.792) of the total variation in the number of professorial
schedules in chemical engineering from university to uni-
versity may be explained by the different numbers of de-
grees granted while the residual 0.38 (which is 1.00-0.62)
of the variation may be attributed to the different edu-
cational, administrative and research policies prevailing.
A second objective of the present study was to deter-
mine changes with time in the coefficients such as those
calculated and presented in Eq. 2 in the light of the very
rapid educational growth rates and rapid changes in
student degree objectives currently occurring on the


American educational scene.
Data for 70 departments of chemical engineering for
the year 1962-63 had been analyzed and reported in a
previous study.*
The least squares equation for these data had been found
to be
P = 2.2 + 0.10B + 0.1M + 0.45D (3)
It is interesting to speculate on the results of the two
studies. A multiple correlation coefficient of 0.83 obtained
from the 1962-63 data against the value of 0.79 in the
current study indicates a reasonable degree of year-to-
year stability despite the relatively rapid changes cur-
rently occurring in Academia. The large change from
2.2 to 3.4 in the constant terms of Eqs. 3 and 2, may
indicate that in the 1964-65 period, considerably more
professorial time was allotted to administrative and
other non-teaching activities than in the 1962-63 period.
We come next to the coefficients of B. If we assume a
Gaussian distribution, the difference between the coeffi-
cients of B in equations 2 and 3 does not (at the 10%
level) turn out to be statistically significant. However,
since there is no a priori basis for assuming a Gaussian
distribution, the observed difference may be significant.
It is interesting to note that if the difference is signifi-
cant in fact, then it is in the expected direction; i.e.,
downward, indicating that fewer professors were required
per bachelor degree granted during the 1964-65 period
than were required during the 1962-63 period. The trend
towards larger lecture sections requiring fewer professors
per student as well as an increase in instruction by
graduate assistants has certainly been observed during
recent years. Finally, with regard to the coefficients of
the M and D terms in Eqs. 2 and 3 one can only say
that the agreement is indeed striking and beyond what
one might expect. To determine to what degree the above
conjectures are correct would of course require direct
corroboration in depth from the institutions involved.
At any rate, the implied stability of the results of these
studies is gratifying. It indicates that similar analyses
in the future may be desirable and useful.
Because of radical differences in the procedures,
policies and manpower requirements of graduate and
undergraduate programs, it was decided, as a third ob-
qective of this study, to try to analyze separately a group
of schools heavily oriented toward graduate work as
distinct from a group that is not. To distinguish between
them the following purely arbitrary criteria were set up.
For each institution the ratio of the combined number'
of master's and doctoral degrees (in ChE) to the number
of undergraduate degrees was determined. If the ratio
was greater than 0.4, or if there were five or more doc-
toral degrees granted in ChE for the year the school was
considered to be graduate-inclined. On this basis the
97 schools subdivided into 40 graduate-inclined and 57
undergraduate-oriented schools.
In both groups, a least squares analysis was per-
formerd as in the case for all 97 schools with the follow-
ing results:

*Schmidt, A. X. and Pfeffer, R., CEE, p. 13 (October
1965).
(Continued on page 143.)


SUMMER 1969








educator



MAX PETERS


University of Colorado
If there is a better way of educating engi-
neers, Max S. Peters, dean of the College of En-
gineering at the University of Colorado, will be
in pursuit of it. "Finding a better way" could be
the most fitting description characterizing his
performance, whether inside or outside the class-
room, the administrative halls of higher educa-
tion, the laboratory, or the smoke-filled commit-
tee rooms in which he is such a driving force.
Something of the vigor and hardihood of the
Ohio-Pennsylvania early American certainly is
evident in Max Peters. He was born in 1920 in
Delaware, Ohio, and received his early education
in State College, Pennsylvania, and at Penn State
University. He earned the PhD degree in chemi-
cal engineering at Penn State in 1951, perform-
ing research on vacuum distillation.
By the time he was awarded his degree he
had already worked as a production supervisor
in a wartime powder plant and had served with
distinction in the 10th Mountain Division of
Italy. For two postwar years he was in charge
of all technical work for Treyz Chemicals in Cook
Falls, New York.
Colleagues and students alike are apt to ad-
dress him as Max. They appreciate his capacity
for hard work and the sense of humor that is
ready to break to the surface in the midst of
serious considerations. His door is always open
to his graduate students. As one of them puts it,
"I'm amazed that a man with so many responsi-
bilities can be so available."
The responsibilities he has are the result of
his diligent research for "a better way." He
heads a college enrolling nearly 2000 under-
graduate and some 300 graduate students, with a
faculty numbering around 125, housed in a new
$8.5 million educational facility which he him-
self worked hard to make a reality. In addition,
he guides engineering programs at two off-
campus CU Centers, at Denver and Colorado
Springs.
The story of the remarkable growth of the
College of Engineering at the University of Colo-


Max Peters, newly appointed in 1962 as dean of the College
of Engineering at the University of Colorado, enthusiastically
pushed forward plans for the proposed Engineering Center,
dedicated in May, 1966.
rado since Max Peters arrived on campus in July
of 1962 is well known to engineering educators
across the country. He hadn't been in office long
before the building priority for the proposed
Engineering Center moved up from eleventh to
first. Then he went to work with the University
administration to acquire $7.2 million from the
state legislature for construction of the Engi-
neering Center the largest sum ever requested
from the people of Colorado. He further exerted
his persuasive and organizing powers to acquire
a supplementary $1.325 million from the Na-
tional Science Foundation.
The Engineering Center at CU can truthfully
be said to have been inspired in great part by the
enthusiasm and imagination of Max Peters. At
ground-breaking ceremonies for the Center in
1964 he revealed the scope of his expectations
when he described the Center as "a major mile-
stone in the forward progress of making
the State a major industrial and scientific
center."
He tackles the challenge at its grassroots -
the high school level with the same bold en-
thusiasm. "Is there anything wrong," he asks,
"with being old fashioned and strongly encour-
aging our high school students to start preparing
themselves while in high school"


CHEMICAL ENGINEERING EDUCATION









"To be what you can be, you must first and
foremost decide what you want to be."

Many of his colleagues and students have
heard him say, "Engineering is a tremendously
exciting and rewarding career!" and in a di-
versity of situations he proclaims vehemently, "It
is time that some of us decide to speak out."
Max showed signs of "speaking out" in his
first academic position, as assistant professor,
then professor, and then divisional head of chem-
ical engineering at the University of Illinois.
There he began the characteristic pattern: to
examine and re-examine the curriculum to expose
its weaknesses, identify its strength, and take
action to improve it. At Illinois he recognized
the need for more extensive chemical engineering
kinetics study and introduced a course to fill the
void.
His first book, Elementary Chemical Engi-
neering, (McGraw Hill, 1954) was written at
Illinois to fill a gap in engineering education for
students of other disciplines such as mechanical
engineering and chemistry. The text has been
especially valuable in foreign countries where
teachers were not capable of using standard
texts for chemical engineer majors.
In his second book, Plant Design and Econom-
ics for Chemical Engineers, (McGraw Hill, 1957)
Max Peters tackled another need that of grad-
uate students who went into design work in the
chemical industry with relatively little back-
ground in plant design problems and their solu-
tions. Plant Design has been adopted by more
than half of the chemical engineering curricula
across the country. A completely revised edition
(1968), written with Klaus D. Timmerhaus, as-
sociate dean of the CU College of Engineering,
provides greater depth in optimization and eco-
nomic evaluation.
Both books are considered classics in that
they speak to engineers of all disciplines and to
scientists and industrial managers who have no
formal educational background in chemical
engineering.
It was at Illinois that Dean Peters' drive to
improve educational standards carried him into
the local chapter of the AIChE as faculty repre-
sentative. Years of committee activity brought
him in 1968 to the presidency, where he com-
mitted himself wholeheartedly to the quest for
improved engineering education and true profes-
sionalism by AIChE members, and to the pro-


When Max Peters takes part in Fun & Games at the E-Days picnic
only his plaid shirt distinguishes him from the students.
session's obligation to assist society in the solu-
tion of its problems. His improvements in the
professional society, (among them he originated
the popular Free Forums) are well known to
AIChE members. Through the voice of AIChE
Max has supported the concept of the chemical
engineering degree instead of the general engi-
neering degree as the first professional degree.
As head of the chemical engineering division
at Illinois, Max Peters recognized that chemical
engineering students engage in a wide scope of
activities including economics, technical services,
laboratory research. Accordingly, he introduced
a flexibility into the undergraduate curriculum
that allowed the student to substitute advanced
mathematics, physics, and chemistry for more
conventional courses. His incorporation of
fundamental engineering sciences into the under-
graduate program has been copied by chemical
engineering departments in many institutions
throughout the United States.
It was natural for Max to strive toward rais-
ing the standards of excellence in the College of
Engineering at the University of Colorado. With
Dean W. L. Everitt of Illinois he initiated the
Bi-University Institutional Liaison for Develop-
ment (BUILD) program for experimentation and
development of faculty innovative ideas between
the two universities. Now concluding its fourth
year of support by the Kettering Foundation,
BUILD has implemented exchanges for profes-
sional development that have involved at one
time or another every faculty member of the CU
College of Engineering.
Max is a forceful and articulate committee
member, as many of his colleagues have learned.
As chairman of the CU proposal committee for
the National Science Foundation Scientific De-
velopment program, he helped bring $3.75 million


SUMMER 1969








to the University of Colorado, one million of
which want into electrical, aerospace, and me-
chanical engineering programs in the College of
Engineering. As a result, the College has
strengthened its programs in control theory,
solid state physics, computer logic, fluid mechan-
ics, applied mechanics and mathematics. Growth
in these areas is being watched with interest by
other institutions.
Behind this vigorous activity Max holds a
philosophy that has deeply affected graduate
study and research throughout the College. Grad-
uate students, he holds, should be actively in-
volved in research programs under the direction
of faculty members. Faculty members who are
engaged to fill needs in research areas must be
good teachers. Funding for this program has
increased during the past five years from less
than $200,000 to approximately two million
dollars.
In 1961, one PhD degree was awarded by the
University of Colorado to a student in engineer-
ing. In 1968, 30 engineering students earned the
degree. PhD degrees have been made available
in aerospace engineering sciences and in mechan-

"Is there anything wrong," he asks, "with being
old fashioned and strongly encouraging our high
school students to start preparing themselves while
in high school?"

ics in addition to all the other departments, with
the exception of a new department, Engineering
Design and Economic Evaluation, which offers
the master's degree.
It's easy for graduate students to discuss
their research problems with Max Peters; in fact,
he is himself making a significant contribution
in the laboratory. His research studies in kinet-
ics, particularly on nitrogen oxides and pentaery-
thritol, have resulted in increased understanding
of reaction mechanisms and in the chemical en-
gineer's ability to design reactors. He personally
directs graduate work in allied studies.
He gives his students as much freedom as
possible in their projects, only outlining the over-
all view and the goal. He believes that students
learn more from doing a thing wrong than from
doing it right the first time. This freedom to
experiment encourages his students to be creative
and analytical.
They recognize his qualities as a teacher. "He
took the complex and broke it down into simple
integral parts," one of his students comments,


As usual, Max Peters won the Dean's Challenge Race at the
E-Days picnic in May, 1968.
or, "He built up the complicated theories of
chemical engineering by starting with easily un-
derstood building blocks of knowledge."
As a teacher, characteristically Max looks
for better ways to explain points and ideas; he
designs quizzes and exams to test a student's com-
prehension rather than his memory of equations
and data. The secret of his ability is simple: he
is truly interested in each of his students as a
person. Because of this, he is able to instill in
his students the desire to be successful in study
and experimental work. Typically, his plant de-
sign students have repeatedly won or placed near
the top in the national AIChE Student Contest
Problems.
Students have discovered their Dean is a for-
midable contestant on the ski slopes and the En-
gineers' Days races. They hail him as champion
of faculty-student slalom race at the CU Winter
Carnival, and know him as an accomplished
figure-skater. Every year Dean Peters has won
the Dean's Challenge Race at the E-Days picnic.
He wears his more impressive honors with
modesty. This spring he was elected a member
of the National Academy of Engineering the
highest professional distinction that can be con-
ferred upon an American engineer. He is cited
this June by the American Association of Cost
Engineers for his "continuing contributions to
the field of cost engineering education." He was
recently named chairman of the President's Com-
mittee on the National Medal of Science.
In 1957 Max Peters received the George West-
inghouse Award from the American Society for
Engineering Education for outstanding teaching.
He has been active in the ASEE for nearly ten
years. In 1962 he served as chairman of the


CHEMICAL ENGINEERING EDUCATION








The secret of his ability is simple-he is truly
interested in each of his students as a person.

Chemical Engineering Division and was for six
years a member of the long range planning
committee of ASEE.
He is active on the air pollution committee of
the U. S. Department of Health, Education, and
Welfare, which he serves as consultant. He is
consulting editor for the McGraw Hill Chemical
Engineering Series, and is the author of many
technical articles.


Anyone who knows Max knows him as the em-
bodiment of a belief he has expressed to high
school students and to AIChE members: "To be
what you can be you must first and foremost
decide what you want to be." Since he will never
be satisfied with things as they are, but must al-
ways seek new and better answers, it follows that
Max Peters is not only dean, teacher, chemical
engineer, researcher, and innovator. He is per-
haps first of all a student a student of educa-
tion.


OPTIMIZATION


APPLICATIONS AND LIMITATIONS**


RICHARD R. HUGHES*
Shell Development Company
Emeryville, California

THE OPTIMIZATION MODEL
To begin with, optimization requires a formal
description of the problem. The elements in-
volved and their relationship are indicated in
Fig. 1. First, the problem must be isolated by
a formal description of the "state-of-nature" and
the problem premises. This is often the most
difficult part of the problem. A sound treatment
requires an assessment of whether the solution
will answer the question posed and whether all
significant variable elements are included within


Figure 1.


* Present Address: Univ. of Wisconsin, Madison, Wis.
** Presented at the Los Angeles ASEE meeting June
19-22, 1968.


Optimization implies logical, even formal, decision
making, i.e., the selection, for a set of decision variables,
of the best attainable (and allowable) values for a
designated objective. To successfully accomplish opti-
mization of practical non-trivial problems, two major
requirements must be met. First, we must have access
to computers (normally large digital computers), and we
must be able to use them. This, of course, implies opti-
mization of a mathematical model describing the prob-
lem; the second requirement is that this model must be
the simplest possible one for the job at hand. The calcu-
lation will be extremely repetitive; and any but the
simplest possible model will require excessive computa-
tion and make it uneconomic to use optimization.
In this paper, I first describe my concept of an opti-
mization model. Then I propose guidelines for formula-
tion and simplification of such models. Finally I offer a
few remarks on limitations and complications of the opti-
mization approach. My comments are based on several
years of study and practical application of optimization,
-by myself and many colleagues, to problems in
chemical engineering, process design, and operations re-
search. Most of the rules given are not hard and fast
limitations but merely express my observations of diffi-
culties we have encountered.

the system. Obtaining an optimum scale of manu-
facture at a fixed sale price is absurd if the scale
affects the sales price. With the state of nature
established we then identify the decisions we are
still free to make. As functions of these decisions,
we describe the payoff value, and formulate the
necessary restrictions which dictate limitations
on the problem, legal, physical, economical,
political, etc. These restrictions limit the freedom
of action of our decisions, but there is usually
some variability left. By optimizing, we take ad-


SUMMER 1969








vantage of this variability to obtain the best pos-
sible payoff value.
When formulated as a mathematical model,
the problem has the form indicated in Fig. 2.
The state of nature and problem premises are

I Parameters, pI
Parameters, p I


Constraints, g I I Objective Function, p


I

\ Variables, x
4 II
\ --_ ....

Optimizing Algorithm
L---------------J-
Figure 2.

described by a vector of parameters, p, which in-
cludes prices, sales figures, coefficients of correla-
tions, estimated technical values, and all the
many other numbers which must be used to
quantify the problem. The payoff value is for-
mulated as an object function, 4, whose value is
determined by the decision variable vector, x,
and the parameter vector, p. The various limita-
tions are described as a vector of constraint func-
tions, g, each element of which must be non-
negative for an allowable or feasible solution.
Once the problem is described in this format, we
can optimize it by using an optimizing algorithm
to adjust the variables, x, to obtain the best value
of 4) within the limitations of the constraints.
Mathematically, this can be written as follows:
Max {(P,x) I g(p,x) > 0}-- *(p) = ((p,x*)

(1)
The first part of this equation is read: "find the
maximum over the variable space x of the func-
tion 0 of p and x subject to the non-negativity
of the elements of the vector g, which are func-
tions of p and x." The resulting maximum or
"optimum," 4*, is a function of the parameter
vector p, which describes the particular case
which has been optimized. Corresponding to 4*
there are one or more points, x*, in variable space
x, where 4 = 4*. These values are usually the
most important part of the result; they indicate
the optimum or best choice of variables. Impli-


citly, it is clear that the optimum choice, x*, is
a function of the parameter vector p.

OPTIMIZATION ALGORITHMS
For further information as to the details of the opti-
zation algorithms, ample literature exists. For elemen-
tary introduction, there is a good book by Baumol on
Economic Theory and Operations Analysis3 and a some-
what more mathematical treatment by Carr and Howe5.
In the limited but important field of linear programming,
Dantzig's book6 is the fundamental authority; it is
sound and intelligible, but a little long. The book by
Gass10 on the other hand is quite elementary; for an
intermediate level, that of Hadley13 is probably best.
Hadley's second book14 extends the treatment to non-
linear programming, with emphasis on systems of many
dimensions. For those interested in these more-mathe-
matically oriented problems, the collection edited by
Graves and Wolfell is a good review of the state of the
art in 1963. However, for many chemical engineering
problems, Wilde's recent book on "Optimum Seeking
Methods"31 is more directly applicable. Two recent col-
lections also supply useful hints on this latter type of
problem: the CEP Symposium Series volume on "Opti-
mization Techniques",4 and the book edited by Lavi and
Vogl.19 A last general reference, we cite the recent re-
view by Wilde,32 which includes 74 references of recent
work.
In much of our work, we have used optimum-
seeking methods as described in the later refer-
ences listed above. Our particular versions are
described briefly by Singer29; one of them (the
Maze Method) is described in more detail by
Mugele.22 We have also found the MAP method
of Griffith and Stewart12 to be generally applica-
ble and quite powerful. Finally, for the problems
to which they apply, Rosen's methods, Gradient
Projection25 and Partition Programming26, 27,
have been quite successful. With these and many
more methods available, further extensive work
on mathematical programming or optimization
algorithms does not seem worthwhile for the
engineer. However, the method chosen must be
a suitable one for the problem at hand and the
computing equipment available; with large prob-
lems, the performance of the algorithms depends
strongly on the particular configuration of the
computer.

OPTIMIZATION PROBLEMS
Most optimization problems of interest to a
chemical engineer can be fit into one of four
categories: (1) process design, (2) operations
scheduling, (3) process control, and (4) equip-
ment design.


CHEMICAL ENGINEERING EDUCATION









1. Process Design
In many respects, this is the most important
category. Problems of this type range from de-
tailed process design (to chose the best possible
configuration of many process details) to quick
and rather generally formulated process evalua-
tion (to provide a general pattern of possible
profitabilities of a proposed new process or
product). To illustrate the features of process
design optimization, we use the example shown
in Fig. 3. Although this example was developed
independently, it is quite similar to the one pre-
sented by Williams and Otto33 and studied by
DiBella and Stevens.7




Solvent





By- Product

Figure 3.
Our example considers the reaction of a pure hydro-
carbon with a solvent to produce a desired volatile prod-
uct. The product reacts with excess solvent to produce
an insoluble and undesirable sludge. For simplicity, we
use a simple, continuous-flow stirred-tank reactor, stirred
well enough to permit assumption of uniform composition
and temperature throughout the liquid volume. The de-
sired temperature is maintained by a steam coil, and
evaporation is suppressed by using nitrogen pressure.
The product is fed through a heat exchanger/cooler sys-
tem to a flash valve, where it is flashed to a low pressure
and fed into a separator drum. In this drum, the desired
product evaporates, while the sludge settles and is drawn
off in the bottom liquid layer. Most of the liquid is re-
cycled back through the heat exchanger to the stirred-
tank reactor.
The figure identifies the decisions of optimization
variables selected for this process-design example. In
the reactor, these are the temperature T, pressure P,
molal concentration of solvent C, and the fractional
conversion of hydrocarbon per pass f. In the heat
exchanger, we select the temperature to which the recycle
is heated TR, and, in the cooler, the temperature rise in
the cooling water AC. Finally, in the separator drum we
choose the temperature Tp and pressure P, for the flash
separation, and the residence time T, in the liquid; the
latter is needed to predict the degree of separation of the
insoluble by-product. Note that these are design-type
variables. We are not selecting as prime decisions the
sizes of the equipment; they are an outgrowth of the
process decisions which are made. Consider the fractional
conversion f, as typical. Fig. 4 shows how our specified
objective, the manufacturing cost of the desired product,
depends on the conversion f, for fixed values of all the


other variables. Note the non-linearity, in fact discon-
tinuity, of the objective function with respect to this
variable. The breaks in the curve correspond to breaks
in the available reactor sizes, the number of parallel
passes through the heat exchanger system, and so forth.
Important discontinuities of this type should be included
and an optimization technique selected which permits
their inclusion. In less-important cases, the breaks can
be smoothed over, by replacing the appropriate cost
curves with smoothed approximations. In our example,
we have done this with the pump cost; no attempt is
made to account for discrete sizes of pumps which must
actually be used.


5400

Manufacturing Cost
50 MM Ib/yr of Product

S5200 -

Optimum


5000

0 10 20 30
Conversion of Hydrocarbon (% per Pass)
Figure 4.

2. Operations Scheduling
In addition to designing process units, chemi-
cal engineers are often involved in questions of
planning and scheduling their operation. Shifts
in prices, demands, and all the other problem
premises almost always make operation different
than was projected for the design. Moreover, in
the scheduling of operations, no account need be
taken of capital expenses. These have already
been committed, and we are merely concerned
with operating cost or profit. One example of
large-scale use of large optimization models for
scheduling is in the planning of oil refining oper-
ations, both in single refineries and in multi-
refinery complexes. Treatment of this by non-
linear methods is described by Ornea and El-
dredge.24
In our simple example, consider the effect of chang-
ing operations in a fixed system with a specified reactor
volume, heat exchange capacity, etc. Some variables will
remain the same; for example, the temperature and
pressure in the reactor will continue to be control vari-
ables. However, other variables will be replaced by new
ones; instead of specifying conversion, we will specify
the flows of the feed and recycle streams. In Fig 5 con-
tours of the objective are plotted as a function of these
two scheduling variables. Other variables (the tempera-
tures, pressures, and molal concentration of solvent) are
held at constant values. In addition to the objective


SUMMER 1969








Operating Economics


100 z00
Recycle (GPM)

Figure 5.


function contours, we indicate various constraints coming
from the physical specifications of the equipment: the
capacities of reactor, heat exchanger, and pump, the mini-
mum velocity in the heat exchanger to keep the sludge
suspended, and the minimum production rate required.
Note that within the central, feasible region the problem
is very nearly linear. The optimum point occurs at the
upper-right hand corner, where both the reactor, and the
heat exchanger/cooler are operating at capacity. The
near linearity of this example is illustrative of the fact
that, in scheduling, we very often can obtain an accept-
able description of the system in a linear programming
formulation; this tremendously simplifies the optimiza-
tion itself.

3. Process Control
In spite of their differences, these first two
problems both involve only steady-state analysis.
This is not true of the third type of problem, the
control problem. Here we must account for some
dynamic affects, even though it may be possible to
omit consideration of short-term dynamic affects
if we limit our model to the longer-period control
problem. For example, our objectives often con-
cern only temperature and concentration dynam-
ics; then the model can neglect the effect of pres-
sure waves or liquid-level fluctuations, as long as
the specified control instruments maintain aver-
age pressures and levels at the desired point, over
the period of the temperature or concentration
fluctuations. For our simple example, the flow
sheet illustrating the control problem appears in
Fig. 6. The variables are pretty much those that
are used for the scheduling operation, but we
must now incorporate, in the equations describ-
ing the model, time derivatives showing the dy-
namic effects. In general this means that a con-
trol model formulated with the same degree of
technological complexity as a process design


model will be a much more complicated model
involving more complex mathematics and more
y difficult optimization. Fortunately for industry,
the precise optimization of the control model is
usually less important economically than optimi-
zation of the process design. Thus, it is usually not
desirable to do as technologically complete a job
on the control model as one does with the process
design model. One complication of the economics
for a control model is that the objective function
Almost invariably involves an extension in time.
city We are interested in costs or profits expressed

300 400 as an average over a long period, and we will
invariably have means of evening out our uneven
operations, by the use of storage tanks or the
ability to delay delivery. This makes it extremely
difficult to formulate a true economic objective


in terms of an immediate control variable.
(Continued on page 134.)


Figure 6.


CHEMICAL ENGINEERING EDUCATION







4~ -~


-.7

J,


There are more than 100 billion
barrels of potential new oil on the
North American continent. But it
will have to be dug-not pumped-
out of the ground. It's in the form of
low-grade hydrocarbon solids. Yet,
the world needs so much more

AtlanticRichfieldCompany
making things happen with petroleum energy


oil in years to come that Atlantic
Richfield is already working on
ways to extract it and get it moving.
Projects like this take imagination
and fresh viewpoints. The kind that
come from young innovators like
yourself. We need you-and your


kind of ideas-to keep making
great things happen. Talk to our
interviewer when he's on your
campus. Or write to: Mr. G. 0.
Wheeler, Manager Professional
Recruitment, 717 Fifth Avenue,
New York, N.Y. 10022.


! ^

*s sfl









S[1 laboratory


A MICROCATALYTIC TRACER EXPERIMENT


RALPH W. NEUMANN*
STEVE E. RIFFLEt
STEPHEN T. SWENSON**
JOE W. HIGHTOWER
William Marsh Rice University
Houston, Texas 77001

Because of the rapid development of new
analytical techniques and the increasing demands
on students' time through expanded curricula,
it has become necessary to streamline laboratory
experiments to include as many of these tech-
niques as possible in the shortest period of time.
This report describes one of the experiments
aimed at pursuing this goal in our senior chemi-
cal engineering laboratory and presents some
results obtained by this year's students.
A large fraction of all industrial reactions
are catalytic, and one of the most active areas in
industrial research concerns the development of
more active and selective catalysts for specific
reactions. Although many of the early techno-
logical advances which revolutionized the petro-
leum industry before World War II were the
result of empirical observations, the significant
advances by such men as Sabatier,' Langmuir,2
Taylor,3 Ipatieff,4 Emmett,5 and others have
helped to change the application of catalysis from
an art into a science.
One of the standard catalytic activity tests
in the petroleum industry involves the dealkyla-
tion of cumene (isopropylbenzene). This reaction
seemed the logical choice for our studies over a
standard silica-alumina cracking catalyst in a
microcatalytic reactor for the following reasons:
It is essentially a "clean" reaction, i.e., the only sig-
nificant products are propylene and benzene. There is
little poisoning, which means the same catalyst can be
used from day to day without reactivation. Reactant
and product compounds are easily separated by GLC
and are amendable to isotopic tracer investigations in a
mass spectrometer. Cumene dealkylation has b2en used

*Present Address: Fluor Corporation, Houston, Texas
**Present Address: Enjay Chemical Co., Baytown,
Texas
tResearch technician.


Joe W. Hightower earned the PhD at Johns Hopkins
University working under Prof. P. H. Emmett. He was
a Postdoctoral Fellow at Queen's University, Belfast, Ire-
land and was a Research Fellow at Mellon Institute
until he joined the staff of Rice University in 1967.
Steve Riffle is a research technician and Ralph Neumann
and Stephen Swenson were fifth year ME students at
Rice University.

as a test reaction for diffusion studies6 and for investiga-
tions of active sites on zeolite catalysts.7 Cumene, an
intermediate in the production of phenol and acetone, is
an important commercial compound. Research involving
this compound is currently under way at Rice. The entire
experiment can be carried out in a reasonably short
length of time.

EXPERIMENTAL
A microcatalytic reactor,'9 involves combina-
tion of a flow reactor with a gas chromatograph,
Fig. 1. A helium carrier gas stream flowed con-
tinuously at 10 psig and about 100 cc/min
through the reference side of a standard Gow-
Mac thermal conductivity detector and then
through a small packed catalyst bed containing
a centered thermocouple well. Pulses of reactant
could be injected by means of a 10 p1 hypodermic
syringe through a rubber septum injection port
A. The reactant was carried over the catalyst
where it reacted, and the reaction products were
swept immediately into the analyzing column, a
six-foot coil of 1/4" copper tubing packed with
silicone oil on firebrick. The separated products
passed through the sample side of the Gow-Mac


CHEMICAL ENGINEERING EDUCATION















HELIUM -- .l Iv L

CELL -
ISILICONE
COLUMN
INJ INJ


FURNACE PRODUCT TRAP

CATALYST i LN2
TUBE
Fig. 1.-Schematic diagram of microcatalytic reactor and product traps.

detector where they caused an imbalance in a
Wheatstone bridge which was recorded as a peak
on a 10 mV strip chart recorder whose chart
speed was 2 min/in. Calibration was effected for
each compound by injection beyond the catalyst
bed at injection port B. The injection ports,
detector, and column were all enclosed in a tran-
site box whose temperature was maintained near
1150C by means of a Variac which supplied
power to the heating elements.
The separated products could be collected
individually in a trap thermostated at -195C
(liquid nitrogen temperature) for subsequent
analysis at low electron voltage in a CEC 21-104
medium resolution mass spectrometer. Helium
was removed from the sample trap by evacua-
tion at -1950C.
Pellets of commercial Houndry M-46 silica-
alumina (12.5% alumina) were ground and col-
lected between standard 20-60 mesh sieves to
give particles which varied from 250 to 800 mi-
crons in diameter. Half a gram of this material
was loosely packed to a depth of one cm between
glass wool plugs in the 1.5 cm OD Pyrex reactor.
The catalyst surface area was 270 m2/g.
Initial activation was accomplished by heat-
ing the catalyst in flowing 02 at 5300C for one
hour to burn off carbonaceous residues, and the
catalyst was then cooled in flowing helium to the
295-3650C reaction temperature range. Follow-
ing this pretreatment, the catalyst retained a
reproducible activity level for days without fur-
ther reactivation. The temperature of the elec-
trical resistance furnace around the reactor was
controlled simply by a Variac.


RESULTS
The entire experiment was designed to cover
three 3-hour laboratory periods. We have found
it most effective when each group consisted of
from three to six students. Our seniors were
divided into five groups, with each group coming
in on a different afternoon during the week to
perform the same part of the experiment. This
meant the whole experiment lasted three weeks.
Period I, Introduction During the first period, the
objectives, techniques, and mathematical analysis of the
experiment were described. Each student then practiced
making benzene injections through injection port B until
he obtained reproducible peaks on the GLC. Finally,
each measured his peak areas with a planimeter until
his measurements were reproducible.
Period II, Activation Energy A typical microcatalytic
reaction spectrum is shown in Fig. 2 Besides the unde-
Qa rt ------ --__________ .


70 -


60 -


50 -


40 -


I I I I I I
6 5 4 3 2 I 0
RETENTION TIME, MINUTES
Fig. 2. Typical chromatogram obtained from dealkyation of cumene
over a silica-alumina catalyst in a microcatalytic reactor.
alkylated cumene, the only significant peaks observed
were those of the products propylene and benzene. The
small peak just before the cumene peak may represent
a trace of the dehydrogenation product a-methylstyrene.
Conversions were determined from planimeter meas-
urements of areas under the cumene (C) and benzene (B)
peaks from the equation
Fractional conversion = B/(B+C) (1)
The two compounds were assumed to have similar molar
sensitivities.
Although the reaction is certainly much more com-
plicated than this, 6,10 for simplicity it was assumed to
follow first order kinetics with no reverse reaction of
products. Under the conditions used, equilibrium conver-


SUMMER 1969


20 -










sion was greater than 99%, i.e., the reaction was essen-
tially irreversible. The first order irreversible rate equa-
tion was transformed into one involving fractional con-
version, x, and integrated to give
In 1 --= Aexp(-E/RT)t (2)
1 x
which can be written
In (ln 1 ) + l n At (3)
1 x RT
Since the pre-exponential factor A and the contact
time t in the microcatalytic experiment are assumed to
be essentially invariant with temperature, the activation
energy E can be determined from the slop of a plot of In
(In 1 ) versus 1/T. Data from five sets of
1 x
experiments on five different days by 13 seniors are shown
collectively in Fig. 3; a least squares fit gives an apparent
activation energy of 14.3 0.9 kcal/mole. The sample
size was 2g/1 cumene at all temperatures in the region
of 295 to 3650C.

In (in ])

U1 0II


X






0
gl -


Fig. 3. Compilation of data from 13 seniors showing temperature
dependence of cumene dealkylation over silica-alumina in a micro-
catalytic reactor. The apparent Arrhenius activation energy is 14.3
kcal/mole.
Period III, Deuterium Isotopic Tracers Measurements"1
by exchange with D2 have shown that freshly activated
silica-alumina contains about 4 x 1020 H atoms/g. These
atoms have acidic properties and may provide Bronsted
active sites on which the dealkylation reaction occurs.
The purpose of this part of the experiment is to demon-
strate participation of these atoms in several different


reactions which may occur.
When 2 /1 pulses of benzene were passed over the
catalyst at 3400C, there was apparently no chemical
reaction, as only the benzene peak was observed in the
GLC spectrum. Similarly, when perdeuterio benzene
(C6D6) was injected, only one peak was observed. How-
ever, mass spectral analysis of that benzene peak showed
that extensive exchange had occurred between the cata-
lyst's H atoms and the hydrocarbon's D atoms.12 Fig. 4


do dI d2 d3 d4 d5 d6
ISOTOPIC SPECIES
Fig. 4. Deuterium distribution in benzene after exchange of a pulse
of C6D. with H atoms on the catalyst.
shows the relative amounts of product benzene molecules
which contained from 0 to 6 D atoms. All peaks were
corrected for naturally occurring C13; fragmentation
involving loss of one or more H atoms was negligible
under the low voltage mass spectrometer operating con-
ditions used.
Seven more identical 2 /1 pulses of CD6 were then
passed in succession every 10 minutes over the catalyst.
The products were trapped and analyzed mass spectrally;
the results are given in Table I. The last column showing
the atoms exchanged/molecule was calculated from the
equation
6
Atoms Exchanged/Molecule = Z (6-i) di/100
i = 0 (4)
where di is the percent of molecules containing i deute-
rium atoms. As the pool of available H atoms on the
surface became diluted with D atoms as a result of
exchange with each successive pulse, the amount of
measurable exchange decreased from pulse to pulse
(see Fig. 5).
From the number of benzene molecules injected in
each 2 /1 pulse and the average number of atoms ex-
changed (or titratedd") per molecule, it was possible
to determine the total number of surface H atoms which
were exchanged in all eight pulses. Such a cumulative


CHEMICAL ENGINEERING EDUCATION


4-- 364 C


/" --. 298 C










3.00

2.70


2.40
UJ
:2.10
-J
0

1.80
Q
Z 1.50

o
1.20


4 .90


I 2 3 4 5 6 7 8 9 10 II 12 13 14 15
PULSE NUMBER
Fig. 5. Average number of hydrogen atoms exchanged/molecule
during successive passage of several pulses of benzene CoDe and
cumene over a silica-alumina catalyst in a microcatalyst reactor.
plot is shown in Fig. 6. A large fraction (about 75%)
of the total H atoms originally present on the half gram
sample underwent exchange during passage of these
eight pulses of C6D6.
With the catalyst now in a partially deuterated state,
seven pulses of cumene were passed in succession over it
at 340'C, and the conversion was constant at about 50%
dealkylation. All three products were individually trapped
and analyzed for pulse 9, but for the remaining pulses
only the benzene peak was trapped and analyzed; the
results are given in Table I. It is apparent from the
results of pulse 9 that exchange was extensive in the
undealkylated cumene as well as in the reaction products.
In fact, to a rough approximation all the H atoms in all
the hydrocarbons essentially equilibrated with the D
atoms from the catalyst. For this to have been strictly
true, the benzene and propylene (each has 6 hydrogen
atoms) should have had the same number of D atoms/
molecule, and cumene (12 hydrogen atoms) should have
had doube that amount using this assumption, and basing
the calculation on the number of cumene molecules added
and on the benzene mass spectral analysis, the cumulative
number of D atoms recovered from the catalyst could
be determined. The atoms exchanged/molecule were
calculated from the equation
6
Atoms Exchanged/Molecule = Y idi/100 (5)
i = 0
and the cumulative plot for pulses 9 through 15 is shown
in Fig. 6. Most of the D atoms exchanged into the cata-
lyst from the first eight CD,, pulses were recovered in


15
14
-"13
212
x






6
(D 10
59


(7
0


4
23
2


I 2 3 4 5 6 7 8 9 10 II 12 13 14 15

PULSE NUMBER
Fig. 6. Cumulative of hydrogen atoms exchanged between catalyst
and hydrocarbon during passage of successive pulses of benzene C6D6
and cumene over a silica-alumina catalyst in a microcatalytic reactor.

the hydrocarbon products during cumene dealkylation in
the last seven pulses.

DISCUSSION
These microcatalytic tracer cumene dealkyla-
tion experiments over a silica-alumina catalyst
are well suited for a senior chemical engineering
laboratory. In a single integrated experiment in-
volving three laboratory periods, the students are
introduced to a wide range of concepts and tech-
niques including catalysis, kinetics, gas chroma-
tography, product trapping, vacuum systems, iso-
topic tracers, and mass spectrometry. None of
the chemicals is very expensive, and the micro-
catalytic reactor (excluding the recorder and
potentiometer) can be built for less than $350.
In our own department the mass spectrometer
from the catalysis research laboratory was made
available for these experiments. A research
assistant was in charge of the mass spectral
analyses, but the students themselves performed
all other parts of the experiment.
The use of stable isotopic tracers has demon-
strated that what appeared to be a relatively
simple heterogeneous catalytic reaction in fact
involves quite a complicated mechanism. This
certainly invalidates the naive assumption of
first order kinetics. Furthermore, since there was
a temperature dependent peak broadening due to
adsorption as each pulse was passed over the


SUMMER 1969









Table I
Isotopic Composition of Products in Microcatalytic Tracer Experiments


Pulse
Pulse Injected
No. ____


Measured d
S0


d d d4 d5 d
1 2 3 4 5 6


o'I #D atoms
d d d9 molecule
7 8 9 _


B,d6
B,d6
B,d6
B,d6

B,d6
B,d6
B,d6
B,d6
C,d0




C,d 0
C,d0
C,d0
C,d0
C,d0
C,d0


B 1.7 7.3 16.7 24.1 24.4 17.6 8.2
B 0.4 2.3 8.4 18.9 27.9 26.6 15.5
B 0 0.9 4.7 14.2 26.9 32.6 20.7
B 0 0.5 2.3 9.4 23.4 36.0 28.4

B 0 0.1 1.5 7.1 21.0 37.2 33.1


0 0 1.2 6.5 20.1


37.6 34.6


B 0 0 0.6 4.0 16.0 37.3 42.1
B 0 0 0.5 2.8 13.0 35.5 48.2
P 1.2 9.6 22.6 29.5 23.4 11.0 2.7
B 6.2 18.1 25.9 23.1 14.4 7,7 4.6
C 3.8 13.8 22.6 21.8 15.3 9.4 6.1 4


B 25.7 36.7 23.9 9.9 2.8 0.6


0.4


B 43.8 37.0 15.1 3.6 0.5 0 0
B 53.3 33.7 10.5 2.0 0.4 0.1 0


B 63.3 29.5 6.5 0.7


0 0 0


B 72.0 23.7 3.9 0.4 0 0 0
B 79.1 18.8 2.1 0 0 0 0


catalyst, the assumption of constant contact time
at various temperatures is also invalid. These
two factors were mainly responsible for the ap-
parent activation energy being much lower than
that13 reported in the literature over similar
catalysts in a steady state flow reactor.
Although microcatalytic reactors certainly
are not the best suited systems for kinetic meas-
urements, they are extremely useful for isotopic
tracer studies for several reasons:
Tracer compounds are expensive, and only very small
samples need to be used in this system. It is possible to
study "initial" interactions between reactants and surface
after only a relatively few hydrocarbon molecules have
contacted the catalyst. The method is reasonably fast.
With very few changes, this system can be
modified to study the reaction under steady state
flow conditions. Kinetic comparison between the
microcatalytic and steady state flow systems can
be made to investigate the role of diffusion in the
reaction. Furthermore, the GLC is sufficiently
versatile that it can be used without modification
for other analyses.
The authors acknowledge grants from E. I. du Pont de
Nemours and the Petro-Tex Chemical Corporation which


- 2.522
- 1.866
- 1.523
- 1.227
- 1.070
- 1.021
- 0.837
- 0.719
- 3.081
- 2.629

.2 2.1 0.9 3.235
- 1.308

- 0.800
- 0.628
- 0.446
- 0.327
- 0.230


provided funds for development of these experiments.
They are also grateful for helpful suggestions from other
members of the Chemical Engineering faculty.

REFERENCES
1. P. Sabatier, "Catalysis in Organic Chemistry," tr. by
E. Emmet Ried, Van Nostrand Co., New York, N. Y.
2. I. Langmuir, J. Am. Chem. Soc. 40, 1361 (1918).
3. H. S. Taylor, Adv. Catal 1, 1 (1948).
4. V. N. Ipateff, Catalytic Reactions at High Pressures
and Temperatures, MacMillan Co., New York, N. Y.
(1937).
5. P. H. Emmett, "Catalysis," Vols. I-VII, Reinhold
Publishing Corp., New York, N. Y. (1954).
6. C. D. Prater and R. M. Lago, Adv Catal. 8, 293
(1956).
7. J. T. Richardson, J. Catal. 9, 182 (1967).
8. R. J. Kokes, H. Tobin, and P. H. Emmett, J. Am.
Chem. Soc. 77, 5860 (1955).
9. J. W. Hightower, H. R. Gerberich, and W. K. Hall,
J. Catal. 7, 57 (1967).
10. Y. Murakami, T. Hattori, and T. Hattori, J. Catal.
10, 123 (1968).
11. W. K. Hall, F. E. Lutinski, and H. R. Gerberich,
J. Catal. 3, 512 (1964).
12. J. W. Hightower and W. K. Hall, unpublished results.
13. W. F. Pansing and J. B. Mallor, Ind. Eng. Chem.,
Process Design and Development 4, 181 (1965).


CHEMICAL ENGINEERING EDUCATION


4-vo p~~ 4 u p o.o.u 1,







if YOU can


answer "YES" to any


of these


questions, Rohm and Haas


can make your life


interesting.


Do you enjoy being responsible
(and the rewards that go with it)
for developing a project and
carrying it through to completion ?
Do you like a job that gives free
rein to your imagination and
tests your skills ?
Do you like the satisfaction of
making a contribution to society
by improving a wide range of
industrial and consumer products
or by protecting natural resources
or improving farming efficiency ?
Would you like to work for a
company that is concerned with
problems of social responsibility
and encourages its employees to
be active in urban affairs?


We are not altruists. We are a
strong and growth oriented
chemical company that needs
engineers of all types-chemical,
mechanical, electrical and
industrial-who want to make a
practical contribution to
improving man's lot and at the
same time advance themselves in
the world of business. We have
doubled our sales in the past
10 years to the $400,000,000
level. We make some 2,500
chemical products including
synthetic resins, plastics, fibers,
pharmaceuticals and animal
health aids. We are on the move
and you can move with us into
positions of responsibility just
as rapidly as you show us you
have the ability.


There's just one thing-you'll be
expected to work hard-but
you'll be in good company with
the 13,000 other people that
make up Rohm and Haas. We
have six major manufacturing
locations in the United States
and producing subsidiaries in
15 foreign countries. We need
engineers in research, production
and marketing,
Write to Manpower &
Employment #7069. It could
be good for both of us.


ROHM
IHRRSKI
PHILADELPHIA, PENNSYLVANIA 19105


ROHM AND HAAS COMPANY
An equal opportunity employer.









IonC i


views and opinions I


THE DILEMMA OF INNOVATING SOCIETIES: Implications

A. G. FREDRICKSON
University of Minnesota
Minneapolis, Minn. 55455


In the opening pages of his book Ancient Europe,
Professor Stuart Piggott* introduces the useful concepts
of innovating and conserving societies. He writes (p.
17): "In the one group, technological developments in
the arts of peace and war must have been socially accept-
able and therefore encouraged; in the other, once a
satisfactory modus vivendi for the community within its
natural surroundings had been achieved, there seems to
have been no urgent need felt to alter the situation. Or
again, the cultural pattern devised might be too delicately
adjusted to the circumstances, and too rigidly conceived,
to be susceptible of modification by technological innova-
tion ." Professor Piggott's definition suggests two
reasons why a community might choose to become a con-
serving society, and we shall explore his second possibility
- that concerning the delicate adjustment of the society
to its environmental circumstances-in some depth. First,
however, it is necessary to say a few words about inno-
vating societies.

THERE IS NO doubt that American Society,
or at least its white, middle-class sub-society,
is an innovating society. Indeed, few would deny
that the scope, scale, and pace of its technological
innovations- for it was technological innova-
tion that Professor Piggott was thinking of when
he wrote his definition outstrip those of any of
its contemporary societies, including the Soviet
Union, Western Europe, or Japan. Many Amer-
icans take great pride in our position of leader-
ship in technological innovation, and most Amer-
icans would reject with scorn any suggestion
that the United States become a conserving so-
ciety. To find support for these statements we
do not even need to look beyond our own academic
cloisters: the phenomenal growth of scientific
and engineering research and education in the
universities over the past two decades is a clear
indication of the high regard for technological
innovation held by the decision-making and direc-
tion-determining segments of our society. The
only audible voice of dissent to the notion that
the United States must remain the leading inno-
vating society seems to come from a small but
vocal group of young activists who have deci-
* Stuart Piggott, "Ancient Europe," (Chicago: Aldine
Publishing Co., 1965).


Arnold G. Fredrickson has BS and MS degrees from
the University of Minnesota and a PhD from the Uni-
versity of Wisconsin. His principal research interest
involves the study of the dynamics of biological popula-
tions, with special emphasis upon the interactions of such
populations with their environment. In addition he is a
dedicated nature photographer and student of the botany
of Minnesota. His research interests coupled with his
avocation prompted the writing of this essay.

sively rejected the values of contemporary so-
ciety. Unfortunately, this voice is so inexperi-
enced, so often charged with passion, and so con-
centrated on highly specific issues, that it has
not called attention to the basic dilemma that
must bedevil all innovating societies.
This dilemma is suggested by Professor Pig-
gott's second explanation for the origin of con-
serving societies. Let us recall his words: ". .
the cultural pattern devised might be too deli-
cately adjusted to the circumstances (of its
natural surroundings) to be susceptible of
modification by technological innovation ."
That is to say, the environmental resources of
land, water, vegetation, minerals, energy, etc.
may be so limited that technological innovations
within the possibility of the restricted experience
and capability of the community would upset the
balance of the community's ecological situation
irrevocably. For instance, a member of a hunt-
ing society inhabiting a forested region might
conceive the idea that game could be driven from
its cover, by selective burning in the forest, out
into places where it could be easily captured.
Obviously, if the territory available to the tribe


CHEMICAL ENGINEERING EDUCATION


!










for Engineering Education

is limited, and if the tribe's rulers have a modi-
cum of foresight and survival instinct, they will
squelch this kind of innovation. A member of a
modern innovating society might suggest that
by further innovations (such as devices to con-
trol the area of burning or devices to defeat the
tribe's enemies in war so as to enable expansion
of its territory) the tribe could solve its urgent
and ever-present problem: that of feeding every-
one in it. Such a suggestion is not useful, how-
ever, because it ignores the point that the tribe
often does not welcome such innovations, even
when it is clear that they would be beneficial.
On the other hand, technological change is welcomed
in innovating societies and an enormous capacity for in-
novation is the hallmark of present-day societies of that
variety. Indeed, there may even be some individuals who
hold the innovating capacity of those societies to be in-
finite. Be that as it may, the rate at which innovations
can be made is limited, both by the total of experience
possessed by the society, and by the material resources
available to it. The first limitation on the rate of inno-
vation is perfectly elementary; every student of freshman
physics knows that Newton could not have invented the
radio because he lived before Maxwell was born. The sec-
ond limitation on the rate of innovation is also perfectly
obvious to anyone who chooses to think about it. We live
in a large but nevertheless definitely bounded biosphere.
It has only so much lebensraum, so much sustenance that
we can draw from it, so much space in which we can
dump our waste matter and energy, and so much ca-
pacity for self-regeneration. We are not going to farm
the moon to feed the starving billions of Asia nor can
we export the surplus population there, neither can we
etherally dispose of our pollution problems by packaging
waste materials and firing them off to the great incinera-
tor in the sky.
These considerations suggest that innovating
societies, no matter how ingenious they may be,
must eventually encounter environmental restric-
tions on uncontrolled innovation just as do con-
serving societies. Furthermore, when such re-
strictions become apparent, they do so at an
almost unbelievable level of complexity and scale,
and there is even the chance that irreparable
environmental "mistakes" will have been made.
Finally, though a primitative and conserving so-
ciety may have the bonds imposed by its environ-
ment and its own lack of technological capacity
released by contact with an innovating society,
there is no higher institution to which an ad-
vanced innovating society becoming acutely con-
scious of environmental restrictions can turn for
guidance.
This, then, is the basic dilemma of all innovat-


ing societies: On the one hand, their institutions
and ways of doing things are founded alike on a
belief in the efficacy of unfettered technological
innovation for social progress and a fear that
without such innovation, they will stagnate and
decay. Thus, they look on conserving socie-
ties with condescension or disdain. On the other
hand, they are confronted with the fact that con-
tinued uncontrolled innovation is incompatible
with the material basis of life on earth; it is
suicidal. Therefore, if innovating societies are
not to destroy themselves, they must adopt some
of the features of conserving societies; they must
attempt to strike a bargain with Nature instead
of simply exploiting or seeking to overpower
Her; the societies must seek a material state of
things that is more nearly steady than that to
which their beliefs and inclinations have accus-
tomed them.
TN WHAT SENSE MUST INNOVATING
societies become more like conserving socie-
ties if they wish to survive? Not, to be sure, in
the sense that they would forbid technological
innovations out of a satisfaction with the status
quo or out of a dogmatic adherence to traditional
ways of doing things. What is required, rather,
is a sense of proportion and priorities geared to
the real needs of man and to the hard facts of his
existence on this planet. Within such a set of
priorities, innovations for the real benefit of
humanity would be pressed with all possible
speed. But innovations that are simply frivolous
would be looked on with disfavor, and innova-
tions that are destructive of the environment
would be suppressed.
The practical problems of developing the re-
quisite priorities, a mechanism to ensure their
application, and means to review and revise them
as needs change, are severe. This is so in large
part because there exist factors within innovat-
ing societies and within our own innovating
society in particular that militate against any
slackening of the pace of technological innova-
tion or any attempt to control its direction. These
factors may be ill-defined or inarticulated but
they nevertheless seem to cater to or perhaps be
expressions of some deep-seated urges in our
society, and they give tremendous momentum to
the processes that they generate; they determine
the dynamic aspects of our society.
We do not mean to imply, of course, that the
dynamics of our society are determined solely by
(Continued on page 144)


SUMMER 1969









" classroom



TRANSPORT PHENOMENA

EQUATIONS OF CHANGE

V. J. LEE
University of Missouri
Columbia, Missouri 65201


Equations describing fluid motion and energy
transport have been derived either from the Eu-
lerian point of view of a stationary fluid element
of infinitesimal volume or from the Lagrangian
point of view of a macroscopic volume of fluid
in motion. In the former derivations, lengthy
mass, momentum, and energy balances are in-
volved. In the latter derivations, integral trans-
formation theorems and the Reynold's transport
theorem are needed.1,2 The transition from New-
tonian body mechanics to fluids mechanics is less
than direct in both of the two derivations.
This note presents a derivation of equations of
fluid motion and energy transport by considering
an infinitesimal fluid element. 8V, in motion. In
addition to formalistic simplicity, the derivation
exposes the conceptual continuity from the New-
tonian equation of "body" motion to the con-
tinuum motion of fluids.

I. THE RATE EQUATION OF VOLUME DILATION
Let the mass velocity of an infinitesimal vol-
ume element 8V be v. The rate of dilation of 8V
spanned by the vector v is

D (V)= (v n) dS (1)

where dS is a surface element, n is a unit vector
normal to dS. The integration is to be carried
out at time t, over all the surface of 8V, whose
coordinates xj are equal to xj (t) with j = 1, 2, 3.
By the divergent theorem,2 one has


VTv = lim 1 C (v n dS
8V-o0 -8 V- (vn) d
Hence equation (1) can be written as

D (8V) = 8V V v
Dt


Dr. Lee is an associate Professor of Chemical Engi-
neering at the University of Missouri, Columbia. He
was educated at the Ordnance Engineering College, Taipei,
Republic of China (Diploma Engineer), University of
Notre Dame (MS) and the University of Michigan
(PhD '63).
His interests include heterogeneous catalysis, reactions
kinetics, solid state and surface physics, thermodynamics,
transport phenomena and energy conversion.

equation (2) expresses directly that for an in-
compressible fluid
V v = 0 (4)
II. THE EQUATION OF CONTINUITY
The equation of continuity expresses the con-
cept that 8V is a closed system as to mass trans-
fer; i.e., a "body." Let p denote the density of
fluid, the law of mass-conservation gives
D
D- (p8V) = 0 (5)

Remembering 8V = 0, differentiating equation (5)
and combining it with equation (3), we obtain

Dp +p V v = 0 (6)
Dt
III. NEWTON'S EQUATION OF FLUID MOTION
In view of equation (5), we can regard 8V as
a "body" with mass p8V. Applying Newton's
second law of motion to the "body" we obtain

D (pVv) -ff pndS
Dt

Sff (n .) dS

-p8v V V (7)


On the left hand side of equation (7) is the rate
(3) of change of linear momentum. On the right hand
side the first and the second terms are summa-


CHEMICAL ENGINEERING EDUCATION








tion of forces acting on the "body" along the in-
ward normal of its surface due to pressure and
viscous tensors respectively. The last term is a
force acting on the "body" with mass p8V due to
potential field 4 in energy per unit mass. Noting
equation (5), we can rearrange equation (7) in
the form

SD (v) 1 pn dS
p Dt 8V f nJ

1f f (n. 7) dS
-p V (8)
Applying the divergent theorem and noting 8V is
infinitesimal, we obtain
D
p (v) =-V- (V'r) -pV4 (9)

Application of the integral divergent theorem to
a tensor r is infrequent in textbooks, but its proof
is not difficult2-3.

IV. THE ENERGY TRANSPORT EQUATIONS
The kinetic energy transport equation can be
directly obtained from equation (9) by noting


D D
v t (v)[ p Dt (1M v[2)


(10)


Hence the transport equation of kinetic energy is
D
p Dt (1/2 Ivl2) = V.Vp V I (V ")


D
P- (l 2+ ) = V *Vp V (V r) (15)

Now the total energy per unit mass consists of
kinetic energy, potential energy and internal en-
ergy, U, (per unit mass). This transport equation
of total energy can be obtained by an over all
energy balance on the fluid element 8V

Dt [p8V (v/21vl2 + + U)] = ff v (n p) dS


+ ff v. (n-r) dS + ffq-ndS


(16)


where the vector q denotes the rate of energy dis-
sipation per unit surface area of all forms of en-
ergy including heat flux as a major form. The
L.H.S. of equation (16) is the rate of decrease of
total energy. The R.H.S. of equation (16) are
respectively rate of work done by 8V against the
pressure, rate of work done by 8V against the vis-
cous friction and the rate of energy dissipation as
heat. Upon differentiation and combining with
equation (5) and then applying the integral di-
vergent theorem with 8V approaching to zero, we
obtain
D
p Q- / (1v2 + 4 + ) = V (pv)
V (rv)


- V7-q


(17)


Since r is a symetric tensor, it can be shown that


- p vV 4)


(11)


For the transport equation of potential energy,
one notes that 4 is a scalar point function, there-
fore


D
Dt 4 = + vV 4


(12)


For an energy conserving potential field, (e.g.,
the gravitational field), ) does not depend on
time explicitly. Equation (12) becomes
D- v-V (13)
Dt
Multiplying equation (13) by p, we obtain the
transport equation of potential energy
D)
p Dt -= p v*V (14)

Combining equations (11) and (14) we obtain


V [Tv]= v- (VrT) + (T'Vv)


(18)


Hence upon combining equations (15) and (17),
we obtain the transport equation for internal
energy
DU
P Dt =[-(rVv) Vq] -p(V-v) (19)

The terms in the square bracket are rate of heat
generation due to friction and rate of energy
transfer to the system mainly as heat respec-
tively. Consequently upon multiplication of equa-
tion (19) by (8V-At), it becomes of the form


8U = 8Q pSV


(20)


This equation is the familiar first law of thermo-
dynamics for a closed system (i.e., a "body"). In
view of the assumption leading to equation (5),
equation (20) confirms the self-consistency of the
derivation.


SUMMER 1969








NOTATION


v Mass velocity of fluid
8V Volume of an infinitesimal fluid element
dS An infinitesimal surface element
n A unit vector normal to dS
D
D Substantial derivative operator
Dt

V Del operator= x I

p Density of fluid, a point function of xi, x2,
x, and time t
(A A scalar potential function of x1, x,, x,
T The viscous tensor of a fluid
q Vector heat energy flux
v Magnitude of fluid velocity

REFERENCES
1. Bird, R. B., W. E. Stewart, and E. N. Lightfoot,
Transport Phenomena, John Wiley & Sons, New York,
1960, pp. 74-81 and pp. 311-7.
2. Sommerfeld, A. Mechanics of Deformable Bodies, Aca-
demic Press, London, 1964, Aris, R. Vectors, Tensors
and Basic Equations of Fluid Mechanics, Prentice Hall,
New Jersey, 1962, Chapters 3-6.
3. Spiegel, N. R., Vector Analysis, Schaum Co. New York,
1959, pp. 122-3, also Reddick H. W., and F. H. Miller,
Adv. Math. for Engineers, John Wiley & Sons, New
York, 1955, 3rd ed., pp. 350-4.




book reviews


An Introduction to the Engineering Research
Project
Hilbert Schenck, Jr.
McGraw-Hill Book Co.,
New York (1969)
After having directed many theses and over-
seen thesis direction for many years, this writer
has thoroughly enjoyed reviewing this small (178
pg. 5" x 8") book.
Intended to be an introduction to the engi-
neering research project, it moves swiftly from
the selection of a topic through all the major
steps to an expected acceptance of a finished
manuscript for publication. The author is relent-
less as he points out the foibles of faculty and
academic systems and is no less discerning as
he analyzes student "hang-ups" which would hin-
der the choice and early completion of a desirable


research job. The book is written in contempo-
rary style and should be comprehensible to both
the would-be-researcher and his director.
Analyzing the volume in more detail, the re-
viewer believes that "The Selection of a Project"
covers the field well but probably ascribes some-
what more than a normal amount of initiative
to a student. Unfortuately the conception of a
project more frequently falls on a faculty mem-
ber than on a student and therefore Chapter 2,
"Sources for Project Ideas" (25 pages), is far
too long. However the short and meaty "Project
Check Sheet" should be noted by everyone.
The chapter "Searching the Literature" at-
tacks the subject with clarity, vigor, and decision.
It quickly covers the usual but needed generaliza-
tions but follows them up with a well conceived
and highly possible case history.
How many times have projects failed for lack
of apparatus, time, or cost planning? Here is an
author who believes in these efforts as an inte-
gral part of the project. Indeed he stresses these
activities not only as highly desirable but even
mandatory if a real researcher and a satisfactory
project are to be produced. His tips are pertinent,
timely and frequently annoyingly discerning.
Unfortunately the author chooses to elaborate
next on his categories of research an area
which he could better have omitted for al-
though his discussion of "Digital Computer
Studies" is a good short approach to a long
problem, his "Pedagogical Studies" and "Design
and Systems Areas" are far below his overall
standards.
In his last two chapters on "Reports" and
"Journal Papers and Meeting Presentations" the
author has been appropriately and pleasantly
brief. He has obviously called upon many experi-
ences, both sad and glad, and has extracted an
essence which combines philosophy with prac-
ticability.
There is much in this book for new researchers
to learn before sad experiences can dishearten or
even remove them completely from the field, but
the book also may be a gage for a more experi-
enced researcher or research director to recheck
his effectiveness.
Surprisingly despite the "heavy" material
contained in this book, the style is light, friendly
and interesting; it is to be hoped that the experi-
mental project reports will be, too!
Gordon C. Williams
University of Louisville


CHEMICAL ENGINEERING EDUCATION








CHEMICAL ENGINEERS


When you GO with


TEXACO

you GO with the BEST!

. The BEST Facilities
The BEST Products
The BEST People
The BEST Opportunities


Yet, being the best makes it that much harder to
get better. With competition constantly at our heels,
it takes an aggressive and imaginative R & D team
to maintain Texaco's leadership role in the petro-
leum and petrochemical industry. An important part
of this team are Chemical Engineers... men like
yourself constantly searching for better ways. It
is through their efforts, as well as their professional
colleagues and an aggressive management team,
that Texaco stays out front.
You, too, can be part of this winning combination.
For Chemical Engineers with a B.S. or M.S.,
the professional and economic rewards of a Texaco
career in process and product development have


never been greater. Texaco's practice of basing pro-
motions on performance and ability afford you the
BEST opportunities for advancement, while enjoying
our congenial, shirt-sleeve atmosphere.
Texaco has immediate openings at its laboratories
in Beacon, N. Y., Richmond, Va., and Port Arthur
and Bellaire, Texas for qualified Chemical Engi-
neers who already are, or are in the process of be-
coming U. S. Citizens.
Interested candidates are invited to send their
resume to: W. R. Hencke, Texaco, Research & Tech-
nical Department, P. 0. Box 509, Beacon, N. Y.
12508. Texaco is an equal opportunity employer.









A SELF-PACING,


AUTO-GRADED COURSE*

G. DAVID SHILLING
University of Rhode Island
Kingston, R. I. 02881

INTRODUCTION
A two-credit course in reactor engineering,
required for senior students in Chemical Engi-
neering at the University of Rhode Island, has
for the past three years been operated in an
auto-graded mode (so called because the student
supposedly decides what grade he will get.) Be-
cause the senior classes are not large, only one
section of the course was scheduled each year,
and all the students participated in this experi-
mental program. Except for three graduate stu-
dents, the students have all been seniors. A wide
range of scholastic achievement levels has been
represented from students headed for gradu-
ate school, down to some on the verge of being
dismissed.
The goal of the experimentation was to find a way
out of a situation where the instructor met two hours a
week with a well-knit little group of tired students willing
to let him do all the talking and most of the work. It was
hoped that by using the auto-graded mode, which was
described in ASEE Publications, first by Norman Bala-
banian' and then by Roland Mischke2, the more ambi-
tious students could be released from the lock-step pace
and the dependence on the instructor; and by requiring
each student to take the initiative for his progress toward
a passing grade, the less ambitious students would be
confronted with some facts of life.

DESCRIPTION OF COURSE
The rules governing the operation of the
course are shown in Table I, the "Course Plan"
which was distributed to the students at the be-
ginning of the course. A student's progress
through the course is marked by his passing a
series of tests, and his course grade is deter-
mined by the number of tests he passes (eight
for an A, seven for a B, etc.). Each test "covers"
one chapter in the textbook by treating in detail
a set of problems illustrating the principles set

1N. Balabanian, "Removing Emphasis on Grades,"
J. of Eng. Ed., 54, No. 7 (March 1964).
2R. A. Mischke, "A Semitutorial Approach to Teach-
ing," J. of Eng. Ed., 56, No. 3 (November 1965).
*Presented at the Annual Meeting of ASEE, June
17-20, 1968.


G. David Shilling is a graduate of the University of
Delaware and the University of Wisconsin (PhD '50).
He taught at Kansas State and is now completing his
seventeenth year on the URI faculty. He developed an
interest in process control at summer institutes at Case
Institute and the University of Colorado and published
Process Dynamics and Control in 1963. Since his parti-
cipation in the ASEE's Programmed Learning Project
(1965-66), he has experimented with programmed in-
struction and self-paced and auto-graded courses.

TABLE I COURSE PLAN
Chemical Engineering 64 Fall 1967
The text book will be Chemical Reaction Engineering
by Octave Levenspiel.
The instructor will select and announce a set of prob-
lems for each chapter as follows: Chapters 2, 3, 5, 6, 7,
8, 14, 13.
A student will obtain credit for a chapter by scoring
90% or better on a written examination on the chapter.
Only then will he be eligible to take the examination on
the next chapter, in the order shown above.
Examinations will be limited to 55 minutes. They will
be given at 1 P.M. on Monday, Wednesday, and Friday,
between September 21 and January 13, school holidays
excepted.
The instructor will retain all copies of the examina-
tion papers and questions. A student may see and discuss
his examination paper when convenient.
A student's course grade will depend on the number
of chapters he passes as follows: eight, A; seven, B;
six, C; five, D.

forth there. These problems are much more
involved than a student could be expected to work
"from scratch" in an hour examination. But the
students are given the problems in advance and
have worked them and perhaps discussed the
solutions with the instructor and other students.
It is assumed that each student will choose
what course grade he will get and decide when he
will pass the required tests for that grade. To
help the student pace himself through the semes-
ter, he is given a copy of the "Experience Table"
for last year's class. The table shows that it is


CHEMICAL ENGINEERING EDUCATION









not unusual for a student to take a test two or
three times before achieving the high-quality
performance required for passing (90%). The
tests are so comprehensive that, in order to score
90%, a good mastery of the material covered in
the chapter is needed, even when the student has
taken exactly the same test a few days earlier and
afterwards discussed his errors with the instruc-
tor.
In the beginning of the experiment (Class of 1966)
class meetings were held as scheduled. But it soon became
apparent that the students were getting "out of step,"
so that too few students were finding any given topic
sufficiently pertinent to the work they were doing (or
putting off) to make class meetings fruitful. After six
weeks of regular class meetings with steadily decreasing
attendance (since no effort was made to coerce attend-
ance) no more such meetings were held, although some
extemporaneous lectures were given to small groups.
With the Classes of 1967 and 1968, class meetings were
discontinued as soon as the students voted to use the
auto-graded mode. The most efficient plan would prob-
ably be to hold class meetings for the first three or four
weeks, while the students are relatively "in step." But
the students find the no-class feature one of the big
attractions of this mode of operation. Some of the disad-
vantages of not having formal class meetings may not
have showed up in this experiment, because nearly all
the students were also enrolled in a laboratory course
which met six hours a week with the same instructor.
So there were many opportunities for student-instructor
consultation.
The advantages and disadvantages observed
in the operation of the course are discussed later.

COURSE EVALUATION
A course-evaluation questionnaire was filled
out by the students at the end of the course. The
returns were anonymous and it appeared that
the students felt quite free to express their feel-
ings and suggestions. In 1966 there were seven
returns from a class of ten, and in 1968, nine
from twelve. In 1967 the students got away be-
fore the questionnaire session could be arranged.
For the two classes covered, the total enrollment
was 22 and the number of questionnaires com-
pleted was 16.
Table 2 summarizes the answers obtained. To
encourage the students to think about each ques-
tion, the answer blanks on the questionnaire were
scrambled as to positive, neutral, and negative
reactions. In Table 2, the answers have been
rearranged for easier analysis.
The two classes are not tabulated separately
because they did not differ greatly in their opin-
ions. The Class of 1968 indicated a somewhat
more positive attitude: surer of mastery of the


TABLE II COURSE EVALUATION
QUESTIONNAIRE


Questions


Do you feel that the yes
material covered in this 8
course will be of value
to you in your career?
Did you find the course most of it
work interesting? 9
Did you feel that you most of it
have a good understand- 9
ing of the material?
How did the amount of too much
material covered com- 11
pare with other two-
credit courses?
What do you think of
the text book used? interesting


How did you feel about
this course plan at the
beginning of the se-
mester?
How would you feel
about it now?
How well do you think
this plan would work on
various class levels?

How well did the prob-
lems assigned cover the
important material in
the book?
How well did the exams
test your mastery of the
problem topics?


Answers
Number
perhaps
6


Suggested
Checking
no
2


some of it little
4 3
some little
7 0

average too little
4 1


easy to understand
well organized
too condensed
too varied
enthusiastic willing
5 8


enthusiastic willing
0 14
well poorly


Soph
Junior
Senior
well
5


0 12
1 5
7 3
pretty well
9


yes no
8 6
11 3
14 2
3 12
1 10
reluctant
3


reluctant
2
average


4
10
6
poorly
2


well pretty well poorly
5 8 3


material and of its career value, less critical of
the textbook, and not so impressed with the ex-
cessive amount of work demanded. On the other
hand, they felt more keenly the danger of "put-
ting-off" work, and were more critical of the
tests. This modest improvement in attitude, dis-
played by the third group compared to the first
group, may be in part due to more experienced
handling of the course by the instructor. Also, the
third group had two outstanding students and
a well-developed group spirit.

DISCUSSION
The amount of material covered in this course
varies from student to student. ("A" students
study more chapters than "B" students; "B"
students more than "C" students, etc.) because


SUMMER 1969









TABLE II (continued)
Check any of the statements below
which represent fairly closely your
experience in this course:


I spent a lot of time studying the text book on
the topics covered by the problems.
I read the whole chapter carefully.
I also read some chapters not covered.
I tended to ignore the book and get problem
solutions from other students.
I did the problems mostly on my own and
really understood them.
I tried to memorize the problem solutions
instead of understanding them.
I felt that having the course organized this
way saved time for me and let me do my
best.
I expected to pass more exams, but found the
latter chapters were too hard or took too
much time
It was easy to put off working on this course.
I would have done better if there had been
some deadlines at (monthly) intervals.
I didn't like having to plan my own work.
I felt that some students had an unfair
advantage.
The instructor didn't seem to care whether I
worked or not.
The instructor was not very helpful when
consulted
What factors hurt exam effectiveness?
too easily memorized
not enough time
asked wrong things
hard to interpret
students cheated
too mickey-mouse
poor surroundings
too much time allowed
Write-ins (one each):
had to be memorized
many topics not covered
had to memorize numbers
rather picayune
too much detail in correction


Number
Checking
12


this seems to be the simplest way to organize an
auto-graded course. This disadvantage would be
harder to accept if the course were not at the end
of the curriculum. On the other hand, compre-
hension level is maintained high for all students
("90%"). In the conventional, lock-step course,
there is a standard "coverage" and the students
are graded according to their comprehension
level. Actually, both plans have arbitrary limita-
tions: 90% is not perfect, and there are always
more topics which could profitably be included
in the standard coverage. For many a student,
the feeling that he has achieved high-level com-


mand of a topic would be a welcome change from
the feeling of failure or mediocrity he gets in
many of his courses. It is certainly more realistic
training for engineering practice to develop a
comfortable facility with a limited subject area
than to get a haphazard acquaintance with a
broader field. The coverage achieved in this
course is considered by the author to be excellent
for the two semester-hours credit given. Of the
34 students who have taken the course, only four
(grade D) have stopped short of consideration of
optimum-temperature progression in homogene-
ous reactors (C-level), while thirteen (A and B),
also got work in heterogeneous reactions.
Retention of skills developed is, of course, as import-
ant as developing and demonstrating these skills. As in
the usual educational situation, no measurement of reten-
tion was here attempted. However, the instructor was
sensitive to indications of retention observable during
a laboratory course the students take in the semester
following the kinetics course. The observations have been
favorable: when these students were assigned lab prob-
lems related to reaction kinetics, they showed quick
recall of relationships, procedures, and even some details.
Their attitude suggested confidence in their ability to
handle this subject matter. It is reasonable that reten-
tion of skills learned in this type of situation would be
relatively high, because the student has to "dig it out"
for himself, and because, before leaving each unit, he is
assured that he has a good grasp of the material.
The problems chosen by the instructor for the
students to work on need to be "comprehensive,"
i.e., requiring a good grasp of the entire subject
area to be "covered." It is not cricket to require
the student to show more on the test than he had
to do to solve the problems. On the other hand,
the problems do not have to be neat and limited;
the students have time to chew them over, seek
out additional data, and resolve ambiguities (as
in engineering practice). Some "old" problems
can be used, since it is not essential that each
student work every problem entirely on his own.
This year, about half the problems in each set
were new, and the rest were taken from previous
years' sets. Some students attempt to memorize
problem solutions borrowed from other students
or found in "files" left by earlier students. How-
ever, students have testified that this approach
is not successful. The problems are so involved
that it is very difficult to write a 90% test with-
out understanding the solution.
The tests need to be comprehensive enough
that the student must either work through the
problems himself or thoroughly study the solu-
tions he borrows. The questions do not have to
be pared-down to what an average student can


CHEMICAL ENGINEERING EDUCATION








reasonably be expected to work out in 55 minutes.
The student works on the problems before the
test and he can take the test a second or third
time if he has trouble assembling and organizing
his answers in 55 minutes. So, even ambiguities
in test questions, though to be avoided, do not
have tragic consequences.
Repeated tests are not composed in the fa-
miliar "sampling" mode. This would result in
students using information about the test ques-
tions as a guide to slighting important parts of
the topic to the "covered." A typical test is shown
in Table 3. Students are asked to show parts of

TABLE III TYPICAL TEST
Exam on Chapter 5- Fall 1967
1. For Problem 11 (see problem statement below),
derive the differential equation relating reactor volume
to fractional conversion, and show the computations of
the values of the constants used (in the integrated form
of the equation) to find k from the experimental data.
2. For problem 18, derive the differential equation
relating volume of reactor to fractional conversion, and
explain how you proceed to compute the volume of the
required reactor.
3. For Problem 19, derive the required equation and
show all computations for the volume of (only) the back-
mix reactor.
(Problem statements followed.)

some problem solutions in detail, while for others,
they are asked to describe how the problem is
solved. Students are often required to show the
source of a model equation. This they refer to as
"memorizing derivations," which they consider
unfair. They also find that they can best get
through a test in the limited time if they memo-
rize a few key numerical values. Although many
complaints result, it is doubtful that there is any
lasting resentment. Students are so used to
cramming for exams, that to memorize, ten min-
utes before a test, a few simple things is not
much strain. The requirement that a test failed
must be repeated in its entirety is the cause of
the most-often expressed student irritation.
The repeated-test feature produces a steady
stream of tests to be graded, and grading often
must be done with unusually high precision, in
order to decide between a 91 and 89 without seem-
ingly arbitrary or inconsistent. Of course, a test
paper which has two really bad flaws can be
checked off quickly, as can a well presented
repeat test by a student who nearly made it last
time and has been shown his error. It is even
more essential than in conventional courses that
tests be graded promptly within 24 hours at


the most. This may require some careful plan-
ning by the instructor, such as limiting the time
when tests are given so that he is free to correct
them immediately. No "final exam" was given in
this course, as there seemed to be no role for it.
The role of the instructor in an auto-graded course is
different from that in a conventional course. It is expected
that the student will think of the instructor as a source
of useful information. The instructor can promote this
attitude by giving out hints and checking students' prob-
lem solutions, as well as by explaining mistakes on tests.
This consultant role contrasts with the need a student
sometimes feels to "snow" his instructors (impress them
with his knowledge while not revealing gaps in his com-
mand of their subjects.) The instructor of an auto-graded
course is in a good position to find out what a student
needs help on, and what he can do on his own. In order
to use this opportunity effectively, the instructor needs
a firm grasp of his subject matter and the ability to listen
to students. When meeting a rapid succession of students
with questions on a wide range of topics, he has to
"shift gears" a lot. If they come in groups, he may find
himself operating in a "time-sharing" mode.
When the instructor finds himself treating
the same, often trivial, question over and over as
each student comes upon it, it occurs to him that
if he were giving a lecture course he would need
to discuss the point only once. He can minimize
his losses here by individualizing these encoun-
ters and using them to build up rapport. For
topics that are sure to cause trouble for a lot of
students, a mimeographed hand-out can be pre-
pared. With the excellent textbook used in this
course, only one such text supplement has been
prepared. (It deals with the question of changing
density of the reaction mass in a flow reactor.)
The instructor of an auto-graded course spends
more time "consulting" with students, but less
time preparing and delivering lectures (and won-
dering if he is "getting through" to the stu-
dents.)
The auto-graded course rather than promot-
ing competition between students, stimulates the
formation of study groups. Leading students get
considerable opportunity to help other students,
and strengthen their own learning in the process.
A certain amount of working together on prob-
lems is good training for engineering practice,
and in an auto-graded course, is not the threat
of the instructor that it sometimes is in conven-
tional courses.
The effect of this course format on a student's
motivation is of course difficult to generalize.
Ostensibly, the student is working for a grade
and whatever that means to him. Although this
does not sound very commendable, perhaps in the


SUMMER 1969








present-day college context it is not in any real
sense a regression. And, it would seem that it
would help a student focus his energy to know
that the grade he chooses is his when he demon-
strates the required learning. The students indi-
cated on the course-evaluation questionnaire that
they put more effort into the course than they
would expect to put into a two-credit course.
The pressure on the student in this course was rather
even and continuous, compared to conventional courses,
where there are sharp peaks before six-weeks exams and
low periods between them. This low-tension atmosphere,
while it was an advantage for some students, was the
most important factor limiting the general success of
the course. Many students put off working on the course
so much that they came out with C-level achievement
where they could, with wiser investment of their time,
have attained the A level. (Some of these students may
have been waiting for their leaders to move, and when
they did move, couldn't keep up.) To combat procrastina-
tion, the instructor supplied each student with an Ex-
perience Table at the beginning of the course, and offered
additional copies occasionally. He also drew attention to
the deadline established for the end of testing. (This
deadline was extended for one student because of illness
and for one who was working hard on a D.) More
effective ways of reducing student procrastination are
still being sought.
The self-pacing feature, which permits pro-
crastination, is on the other hand a strong mo-
tivating factor for some students. They are very
impressed by the prospect of finishing the course
well before the end of the semester which is
entirely possible and has been done b some stu-
dents.
Undoubtedly a lot is gained, in the way of good
feeling about the course, by the freedom from
weekly schedules, class attendance, and the final
exam. Of course, one can not say how much this


pays off in faster, more lasting learning. If most
courses were self-pacing, the advantages of nov-
elty would fall to the lecture-exam courses.
Ideally, the different courses in a curriculum
should be operated in a wide variety of formats,
each in an optimal way for its particular objec-
tives. Cheating is not an important factor in this
course, because the test questions are not secret
and tests failed are repeated without significant
penalty, nor is any advantage gained by not giv-
ing credit for help received. Ethical problems
are raised by a largely undesirable behavior pat-
tern known as "leaching," where an unconfident
student will attach himself to a leader, or one
with a channel to a leader, in order to obtain
more information about the problem solutions
than he is able to contribute. There is little the
instructor can do about such a situation. The
students tend to work out arrangements so that
all involved gain something.


CONCLUSIONS
All the students in this auto-graded, self-pacing
course demonstrated high-quality command of a reason-
able amount of chemical-reaction-engineering skills. There
were no failures, and only one early drop out. Forty
percent of the students received A or B grades for work
beyond the level considered satisfactory for the number
of credits given. There is every reason to believe that
their retention of this learning will be superior. The
amount of material covered varied with the student, and
would have been greater for some if the tendency to
procrastinate had been suppressed. Students generally
agreed that they worked harder than in most courses.
A theoretical advantage of the course is that the student
behavior encouraged (if not uniformly obtained) bore a
strong resemblance to that of a practicing engineer.


OPTIMIZATION R. R. HUGHES
(Cont'd from p. 116)

4. Equipment Design
The final category is really just good equip-
ment design. To obtain detailed pictures of the
makeup of packed bed reactors, the nature of
internal baffles in stirred tanks or the exact form
of heat exchanger bundles, a good designer must
optimize in terms of some minimal cost or maxi-
mum-performance criteria. Each type of equip-
ment requires its own special treatment for op-
timization, so that a general treatment of equip-
ment optimization is not really desirable. In
many cases, however, it may be possible for the
designer to make use of some of the optimization
algorithms.


MODEL FORMULATION
How then do we go about formulating a proc-
ess design model? The calculations normally
involve six distinct steps, once the desired deci-
sion variables have been chosen and the neces-
sary objective and constraint functions have been
identified.
e Stoichiometry. The heat and material balances for
all major pieces of equipment are normally involved in
any process design. According to the problem, it may be
desirable to make the material balance on a mole, weight,
or volume basis. In some cases, for example in certain
types of refinery problems, it may be possible or neces-
sary to treat the stream in total. But, normally, at least
a nominal set of components should be identified and
separately balanced.


CHEMICAL ENGINEERING EDUCATION








* Chemistry of Conversion. For conversion processes,
some sort of chemical description is needed. In a few
situations, chemical equilibria may be adequate. Gen-
erally, however, at least in some simplified form, there
must be a treatment of the chemical kinetics. Which-
ever treatment is used, the equation should apply through-
out the region of interest. If necessary, new constraints
should be formulated to eliminate areas of ignorance
with respect to the chemistry. If these constraints are
significant at the optimum point, it may be desirable to
do further developmental analysis work in order to am-
plify the kinetics, or the chemical equilibrium.
o Thermodynamics of Separation. For a separation
process or the separation units in a larger process, some
representation is needed of the thermodynamics of the
separation, the phase equilibria and volume and enthalpy
changes.
* Equipment Sizing. To relate the stoichiometry, chem-
istry and thermodynamics of the process to the actual
process plant, equipment sizing calculations are needed.
In some cases, these will be nothing more than arbitrary
rules; in others they will be capacity or performance
correlations. This step is often the most uncertain, and
may require formulation and reformulation as the region
of interest is identified.
* Capital Cost Estimates. Once the equipment is sized,
a cost estimate must be developed for it. Usually the
detailed cost estimating methods suitable for a final con-
tract bid are not necessary. Instead we need approximate
methods that show how costs vary with small changes
in equipment size. Happel's book15 contains many useful
tables and equations of the type needed.
* Economics and Accounting. The accounting equations
must be combined with capital cost amortization to prod-
uce an economic balance for the process. Normally the
objective takes the form of some rate-of-return, payoff
time, or the like. Happel's book15 describes some of the
mathematics involved, but the best review of the proper
economic objectives is given by Souders.30


SOURCES FOR EQUATIONS
To obtain the desired model equations we can
use several different sources. First (and often
overlooked) are the definitions, e.g., the molal
concentration of a component, the average en-
thalpy of a stream, etc. Then there are first
principles, like the conservation of mass, the laws
of thermodynamics, the formulation of economic
objectives and the like. The mass of chemical
engineering knowledge usually appears in the
form of established correlations, such as the de-
pendence of the Fanning friction factor on Rey-
nolds number and pipe roughness, the capacity
of packed bed contractors, the heat transfer coef-
ficients in contact with a fluid bed, and many,
many more. Finally, if general correlations can-
not be found with sufficient reliability to describe
the desired application, we must have recourse
to experiment. If the optimization study is di-


With (the many methods available), further extensive
work on optimization algorithms does not
seem worthwhile for the engineer.


rectly involved in the guidance of development
work, we must keep the experiment to the mini-
mum necessary to satisfy the desired goal of de-
sign optimization. At the same time, we must
keep in mind what the outcome of a successful
calculation might be, a detailed design of the
final plant. If this design will be called for imme-
diately upon obtaining a satisfactory result from
the evaluation process, then a short cut in experi-
ments may lead to slowing down design.

SIMPLIFYING THE MODEL
The optimizer must keep in mind that his
mathematical model should be as simple as is
consistent with the problem. At times, a very
complex model is needed, when a very precise
answer is desired and justified. But many process
evaluations and most preliminary process designs
can be done with a greatly simplified model. It
must be remembered that the final design ob-
tained from an optimization consists of a set of
design variables, the best possible set. Once
these have been identified, it is possible to pro-
duce a much more refined design which will pro-
vide all the necessary engineering detail as well
as checking the estimated objective function and
the specified constraints.
Here are a few points which can be checked
to see whether a model satisfies this goal of
simplification:
* Use estimates or "average values" whenever these are
adequate for the purposes of the problem. Just because
an engineering correlation exists for a given piece of
equipment does not justify inclusion of this correlation
in a model. For example, detailed correlation of heat
transfer coefficients versus heat exchanger design para-
meters and throughput is merely wasted, if the heat
exchanger does not play a crucial economic role in the
overall design. Here, it is much better to use merely an
average heat transfer coefficient, estimated from good
practice, and a simple cost estimation as a function of the
square feet of exchange surface required. Another
example is the use of approximate over-all absorption-
factor equations to represent performance of an absorber
peripheral to the main process, in preference to a detailed
tray-to-tray calculation, even if the latter is readily
available, along with the necessary vapor-liquid equilbria
to permit its use.
* Group like components in the material balance. Most
chemical processes contain enough chemicals to make the
identification and separate calculation of all components
difficult. Unless their separation is crucial to the process,


SUMMER 1969







isomers and other like groups of components should be
treated as single components. Nearly all chemical engi-
neering correlations of separation equipment involve
summations over the components, so the number of com-
ponents identified should be minimized. This applies also
to kinetic models. Here, each additional component in-
volves at least one additional kinetic constant, and often
several. The kinetic model for the process study is merely
a representation of the kinetics. It is not a true scientific
explanation of the chemistry. This should be kept in
mind while developing the model. If necessary, additional
constraints can be added to limit the region of applica-
bility of a given equation, and insure that undesirable
extrapolation does not occur.
* Use a good base case and consider marginal changes
from this base case. In many cases this technique will
lead to a much simpler model, since first or, at most,
second-derivatives are all that need be included. At other
times too broad-brush a treatment of marginal affects
around the base case may completely vitiate the study.
Finding the proper balance is part of the business of
being a good engineer.
* Correlate results of detailed study of units or sub-
units of the main problem. This ties in with the base case
method discussed above. Even when the base case ap-
proach is not valid for the problem as a whole, it may
be used for certain parts of the problem. For example, in
the treatment of an oil refinery, it may be possible to
describe the performance of the gas-recovery unit for
the catalytic cracker in terms of a few crucial composi-
tion variables and certain major decisions as to recovery
of key components. Then a series of detailed tray-to-tray
calculations could be used and correlated to predict
expected costs and predicted separation performance, in
terms of the key variables.
* Use the simplest acceptable equations to describe
directly-related experiments. If the optimization is being
used to guide development work, there will be directly-
related experiments which can be used to update the
model. In a sense, these experiments appear, to the model,
just like the results of separate studies described above.
Elaborate analysis of the experimental data is only neces-
sary if extrapolation is essential in order to produce the
desired optimization. In most cases, a simple response-
surface-type equation may serve the purpose. However,
some thought should be given to choosing the right form
of the variables; for example, a logarithmic variable
should be used wherever these are more significant physi-
cally than arithmetic ones. However, the statistical
significance of the experiments is rarely sharp enough
to allow greater than a second-order response surface.


USE OF PROBLEM STRUCTURE IN OPTIMIZATION
Optimization problems can often be handled
more easily by taking advantage of the structure
of the problem. In the first place, the mathe-
matical form of the resulting equations may play
a significant role in terms of the ease of optimiza-
tion. If these equations are all linear or can be
linearized without excessive distortion, the pow-
erful techniques of linear programming can be


used. If they are non-linear but continuous or
have, at most, a few discontinuities, non-linear
programming or optimum-seeking algorithms
may be used. However, if these discontinuities
are extensive or if the discrete nature of some
decision variables must be considered, then the
problem becomes much more difficult; in princi-
ple, it requires use of integer programming,
where perfectly general methods for large prob-
lems are not yet available.
On top of this mathematical structure is the
logical structure of the problem. Many linear
problems fit into the so-called transportation
model, which corresponds to the problem of find-
ing the minimal-cost policy to supply a number
of demands at varying locations, by a number of
different factories with different capacity limits.
This is one of a number of network problems
considered, among others, by Ford and Fulker-
son.8 Some integer problems can be fit into the
travelling-salesman or knapsack forms, for which
general methods of approach exist. Many process
problems occur in staged or cyclic form, and can
be subdivided by techniques described by Rudd
and Watson28 and by Aris, Nemhauser and
Wilde.2 Finally, problems that are sequential or
repetitive in nature, such as the multi-period
planning problem, can often be formulated, and
sometimes optimized, in a way that takes advan-
tage of this repetitive structure. The Partition
Programming algorithm described by Rosen21
and used by Ornea and Eldredge24 is readily
adapted to this sequential problem, or to the
natural partitioning which occurs in large-scale
scheduling problems.

CHEOPS A CHemical Engineering
OPtimization System

In an earlier paper18 we described a system of
programs termed CHEOPS, which takes advan-
tage of the general structure of process prob-
lems. Fig. 7, taken from this paper, shows how
a modern refinery falls into units with informa-
tion flow between units confined to the process
streams connecting the units. Even if a detailed
design of one of the units is considered, this
same sub-division is possible. Fig. 8 shows what
happens to the vacuum flasher unit if the individ-
ual pieces of the unit are considered as separate
process units. We identify the furnace-cyclone
combination, the secondary-deentrainment sec-
tion, the pitch cooling section, the heavy and light
flashed distillate condensations, and, finally, the


CHEMICAL ENGINEERING EDUCATION


































Figure 7. Typical Refinery Process Scheme. Blending of final products not shown.


steam ejector and gas system. Within each of
these units or sub-units the same repetitive cal-
culations exist. These can be separated into five
steps: (1) the setup; (2) material balance; (3)
heat balance; (4) constraint calculations; and
(5) cost estimate and profitability.
Fig. 9 indicates how CHEOPS is structured
to handle these consecutive calculations, using a
set of individual unit sub-routines which describe
the units in the actual process. CHEOPS will
operate with any of a number of optimization
algorithms, as long as they are structured in the
form given in Eq. 1. By following a few simple


GAS

D<})-


Figure 8.


rules, the sub-routines for each unit can be struc-
tured to supply the necessary answers to each
part of CHEOPS. Further details on this appear
in Table 1. This table and Figs. 7 9 are all taken
from our earlier publication.18

LIMITATIONS ON PROBLEM SIZE
Precise formulation of the size of problems
which can be conveniently or economically han-
dled by optimization is very difficult. For one
thing, such limitations depend greatly on the
complexity of the simulation. If extremely de-
tailed, complex solutions are desired, this will
lead to extensive sub-programs just to describe
the engineering. If such programs become too
large, they will exceed the available core in the
computer, which means that one must go to
multi-coreloads, with all the attendant bookkeep-

OPTIMIZATION PROGRAM

CHEOPS

INPUT MATERIAL CONSTRAINT OBJECTIVE OUTPUT
BALANCES FUNCTIONS FUNCTION OUTPUT


Figure 9.


SUMMER 1969











Table 1. Program Functions for CHEOPS.

Section Optimization Program CHEOPS Program UNIT Subroutine

Initialization Initializes optimization 1. Ioads data, which may include: 1. Calls INDEX subroutine of CHEOPS
algorithm a) Control for optimization to permit setting indices
b) Control for material balancing 2. Makes preliminary calculations
c) Control for objective function calculation independent of decision variables
d) Control for output (except for a possible dependence
e) General economic and cost data on their starting values)
f) Overhead cost parameters
g) Offsite capital parameters
h) Utility' cost parameters
i) .Tankage parameters
j) Prices for supplies and materials
k) Plant feed amount, prices, and properties
.1) Plant product amounts, prices, and properties
m) Component properties
n) Flow-diagram connections
o) Optimization variable identification, bounds
'and starting values
p) Constraint & equality identification
and tolerancest
q) Parameters for UNIT subroutines

2. Sets indices in'UNIT subroutines
3. 'Prints record of data and derived values
4. Sets controls for remaining calculations


Material and Sets new values of decision 1. Calls UNIT-subroutines for preliminary 1. Makes preliminary calculations
heat balance variables, and calls for calculations which depend on.decision
preliminary calculations at variables, but not on stream
a new point flows, characteristics, or
properties

2. Calls UNIT subroutines, checks flows, 2, Calculates flows, characteristics,
characteristics, and properties of and properties of "asked" streams
recycle streams, makes adjustments, (normally, the streams leaving
and re-calls UNIT subroutines until the unit)
model is material and heat balanced
3. Calls UNIT subroutines for utility 3. Calculates utility demands and
demands and other functions. (returns functions to be used in both
to #2 if utility balance calculation constraint and objective
affects process streams.) calculations


Constraint Sets index identifying Identifies unit number for constraint, Calculates constraint function
functions constraint function and calls appropriate UNIT subroutine
to be calculated, and
calls for constraint
calculation

Objective Calls for objective function 1. Calls UNIT subroutines in order Calculates the following
function calculation 2. C41culates and totals unit operating cost contributions:
and capital costs a) Supply And material use
3. Calculates raw-material costs,,and b) Supply and material inventory
product credits, as needed c) Operating labor
4. Totals utility demands and use, d) Repair and maintenance costs
and calculates utility capital e) Capital cost (with useful
and operating costs life, and tax depreciation
5. Calculates indicated objective life, When applicable)
function, including tankage, overhead,
offsites, etc.


Output Calls for output, identified Outputs results, as indicated by printout- 1. If called for equipment print-out,
as non-feasible, intermediate control data, tallies, and objective function calculates details not required for
feasible, optimal, or program value. Types of output variable are: cost calculations but of interest at
error a) Variables and constraint and objective point selected
functions 2. If called for UNIT print-out, outputs
b) Process evaluation summary results as programmed
c) Capital cost breakdown
d) Utility'use, demand, and costs
e) Materials and supplies summary
f) Process stream material balance
g) Process stream flows and properties
h) Equipment details (UNIT subroutines
are called first to calculate
additional results if desired)
1) UNIT printouts .(obtained by calling
UNIT subroutines)


ing problems. Hopefully, this situation will be
much improved by the third generation of com-
puters, once the necessary systems are fully
checked out. In the meantime, for the second
generation computer (IBM 7094 and the like),
we supply the general guidelines given in Table 2.
These indicate the size of usable problems in


term of the number of variables and constraints
which can be conveniently handled by the opti-
mization algorithms indicated. It is assumed that
the equations relating these functions and vari-
ables are of no more than ordinary complexity.
The algorithms considered are as follows:
* LP Several standard linear programming systems


CHEMICAL ENGINEERING EDUCATION









Table 2. Usable Problem Size for Optimization Algorithms.
LP Map GP PPNL DA
Tvype of Objective F L NL NL NL NL


Decision Variables
Linear
Non-Linear
Bounds
Constraints
Equations


104 500 100 600
100 60

20001 200 120
200 1000o 200 }2000


*No more than two active at any one time.
can handle a practically-unlimited number of decision
variables, and over 2,000 rows; the latter can represent
bounds on the variables, constraints on combinations of
variables, or qualities involving one or more variables.
Note that all functions have to be linear, although it
is possible to do a reasonably good job of representing a
few non-linear relationships by defining new variables
and constraints.
* MAP The technique described by Griffith and
Stewart12 has been used very satisfactorily for problems
with about 500 linear variables and 100 non-linear ones.
The number of rows which can be conveniently handled
is somewhat less than in LP, because the necessary step-
size limitations on the non-linear variables add additional
rows. However, the allowance of non-linearity permits
practically as good a representation of the system with
the smaller MAP formulation as with the large LP
version.
o GP or Gradient Projection.25 The usual form of
this algorithm (available through SHARE) can only
handle linear constraints or qualities, but will handle a
non-linear objective function. Although the limitation
on constraints and bounds is fairly small in the usual
programs, there is nothing inherent in the algorithm
which requires this; it merely represents a balance
between available core on the IBM 7094, and the com-
plexity allowed for the non-linear system.
* PPNL the non-linear version of Partition Pro-
gramming described by Ornea and Eldredge.24 This can
handle much larger systems and can handle up to 60
truly non-linear variables, each of which can have bounds.
The approximate size for the total number of linear
variables is 600 and the total number of rows in the
linear system is about 2000. Normally this system is
used where the linear problem can be partitioned still
further into smaller sub-problems, each of which is
handled individually (but automatically) during the
optimization.
* DA Deflected Assent29 is typical of many hill-
climbing methods. It is very limited as to number of
variables. However, it can handle extremely non-linear
or discontinuous objective functions, and can satisfac-
torily treat non-linear constraints, as long as no more
than two are active at any one time. A further advantage
of this system and of others similar to it is the exereme
compactness of the program itself. Use of DA may
permit a much larger simulation model without requiring
multi-core use. On the other hand, if the problem be-
comes nearly linear at the optimum, Deflected Assent
may behave very poorly.


Much more significant than the actual computation
time is the program development time.



Nothing is included in this table about in-
teger or discrete variables, since most algorithms
to handle such variables are still in the experi-
mental stage. The only general method that is
really foolproof as yet is that of combinatorially
going through all possible cases. This is quite
feasible, once it is set up automatically on a
computer, as long as the total number of cases
is not more than a few hundred. However, for
each case, it may be necessary to do at least a
brief optimization of the continuous variables.


TIME LIMITATIONS
Other limitations on the usable size of prob-
lems arise from the various time requirements.
First, of course, is computation time; this is not
really too serious as long as the above size limits
are met. Moreover, if the problem is important
enough, whatever its size, it is possible to run it,
- at least in a stepwise fashion. With a standard
method of computer center organization, how-
ever, it is usually best to stick to runs between
a few minutes and a few hours in length. Much
more significant than the actual computation
time for optimization is the program develop-
ment time. With CHEOPS, it is possible to de-
velop a useful process model, for example, a
model for optimizing the design of a chemical
plant requiring several million dollars of invested
capital, in a matter of 4 to 6 weeks. But a com-
plex of several plants or an oil refinery might
require months and months of effort on the part
of several programmers, even if they are given
easy access to the computer. Still another type
of time limitation concerns the acquisition of
necessary data, i.e., the values of the parameters.
Even if the model is being developed entirely on
the basis of past information, the acquisition,
assembling, and proper checking might take
nearly as much effort as the program develop-
ment. Finally, the processing of these data for
new optimization runs is, in itself, a major task.
For routine use of optimization models in analyz-
ing different cases, and different sets of premises,
or in routing scheduling operations of multi-plant
or multi-refinery systems, it is essential that an
organization be set up with a full recognition of
the data processing aspects.


SUMMER 1969









UNCERTAINTY AND ITS EFFECTS
In the above discussion we have been imply-
ing that the entire problem is deterministic; that
is, that we can make decisions, once and for all,
based on certain premises which are completely
fixed, and come up with a true optimum value
which will always hold. In many engineering
problems, this is a close approximation to the
true situation. However, as we begin to get
more and more involved with the marketing and
business aspects, we often get into situations
where uncertainties dominate the effect. These
uncertainties can be classified into four general
types:
1. The Form of the Model. Under this heading comes
the question of identifying the proper kinetics, choosing
the right equations to represent vapor-liquid equilibria,
selecting the right correlations for sizing equipment, and
so forth. The only real cure for this uncertainly is to get
the best model builders you can. The implication is that
building the model is an engineer's job, not a mathe-
matician's. Model-builders must be people with engi-
neering judgment and, preferably, with some physical
feel for the system.
2. Accuracy of Data. Once the form of the equations
has been settled, we must estimate or select numerical
values for the various parameters in the system. Some
of these may be relatively precise, but for others, there
may be distribution curves of values (probability dis-
tributions) about the average values. Where such in-
accuracies become significant, we must use simulation or
stochastic programming, as described below.
3. Forecasts of Future Conditions. Nearly all useful
problems involve future behavior of the weather, cus-
tomers, competitors, and so forth. Mechanically, we can
handle these forecasts with the same techniques with
which we handle uncertainties in data. However, philo-
sophically, they are different in type; we are assuming
what the future looks like. In reality, factors that do not
enter into our equations may come to bear before we
actually get the desired results.
4. Responsive Actions of Others In some cases even
the forecasts are inadequate. In a highly competitive
situation, where we are dealing with one or two com-
petitors, one or two major marketers for our supplies,
or one or two major customers for our product, it is
very dangerous to forecast future action. If our optimiza-
tion is to be of any value, we would expect to do some-
thing new, i.e., something we have never done before.
Except in very limited situations, it is impossible to
forecast what the response of a major competitor, sup-
plier, or customer is apt to be to this new action. For
this reason, we have studied the use of game-theory in
analyzing such problems,17 but with only limited success
to date.

OTHER APPROACHES TO MATHEMATICAL
MODELLING
Because these uncertainties dominate many
problems, an alternative technique has developed,


called Monte-Carlo Modelling or Simulation. In
reality, this merely handles uncertainties of types
2 and 3, which can be represented by replacing
each uncertain parameter by a probability dis-
tribution for the parameter values. With these
distributions as guides, we select enough random
cases to produce a truly average value, or pref-
erably, a distribution of values for the desired
objective. Many references describe this proce-
dure. The text of Naylor, et.al.,23 is one of the
newest general references, while that of Franks9
concentrates on chemical engineering problems.
Unfortunately, use of these techniques gen-
erally makes optimization unwieldly or even im-
possible. The usual procedure is to use case-
studies, where a few important variables are set
at two or three values. Fortunately, in many
problems where this approach is essential, opti-
mization is relatively unimportant, precise selec-
tion of optimum variable values is prevented by
the uncertainties. This is certainly true of the
"venture analysis" of Andersen' or the "risk
analysis" of Hertz.16 However, other problems
could profit by a combined approach, which has
been termed stochastic programming.20, 21
One pattern for this stochastic programming is
sketched in Fig. 10. The basic simulation of Figs. 1 and 2
appears at the left in Fig. 10. However, the parameters


L------------------------
Figure 10.
are no longer fixed, but selected by some Monte-Carlo
technique, from the given probability distributions. And
the optimizing algorithm responds, not to individual
values of the objective function, but to the expected value,
or some other property of the calculated probability dis-
tribution for the objective function. Although many re-
searchers are studying this and other formulations of
stochastic programming, useful, general, computation
systems are not available. Until they are, the engineer
must select his tool according to the problem, optimiza-
tion for deterministic, well-defined, many-decision prob-
lems, and stochastic simulation for highly-uncertain,
few decision problems. (Continued on page 158.)


CHEMICAL ENGINEERING EDUCATION








You won't just get your feet wet.


Standard Oil Company of California offers all
the experience you can soak up.
You'll start out facing practical situations and
using your academic knowl-
edge and skills to solve real
problems. You may even have
to improvise and develop
new approaches to specific
questions.
We rotate the assign-
ments of young professionals.
You will be able to work
with different groups of


experienced colleagues and sharpen your skills on
a variety of projects.
Talk with our representative when he comes
to your campus about the
opportunities we have for
you. Check your placement
office for more information or
write to: D. C. Reid, Coordi-
nator, Professional Employ-
ment, Standard Oil Company
of California, 225 Bush Street
I -Room 105, San Francisco,
California 94120.


Standard Oil Company of California
An Equal Opportunity Employer











r07.Mproblems for teachers

1. Submitted by Professor R. B. Bird, University
of Wisconsin.

Hydrostatic Pressure Distribution in In-
compressible Fluids. Consider a beaker of liquid
which, for all practical purposes, can be consid-
ered to be incompressible; let its density be po.
It is desired to obtain an expression for the pres-
sure in the liquid as a function of position. Take
the origin of coordinates to be at the liquid-air
interface, with the positive z-axis pointing away
from the liquid; let the pressure at the liquid-air
interface be p(0). A friend comes to you with
the following comments:
I. "By simplifying the equation of motion for an
incompressible fluid at rest, I get 0 = -dp/dz
pog; I can solve this and get p = p(0) pogz.
That seems reasonable the pressure increases
as one goes deeper and deeper into the liquid."
II. "But, on the other hand, the equation of state of
any fluid is p = p(p,T). If the system is isother-
mal,, then p = p(p). If, furthermore, the fluid is
incompressible p = p(po) = constant. This tells
me that the pressure is constant throughout the
field which I don't believe!"
Clearly your friends needs help. Explain.

2. Submitted by Professor Dave Chittenden, Uni-
versity of New Hampshire.

Computer Solution for the Adiabatic Flame
Temperature. Find the adiabatic flame tempera-
ture for combustion of the following natural gas
mixture: CH4, 86.6%; C2He, 7.9%; CH5, 2.7%;
C4H2o, 1.3%; N2, 1.5%. This dry gas is mixed
with 130% theoretical air which contains 0.043 lb
HO/lb dry air. The gas-air mixture enters the


burner at 5000K and 4 atmospheres pressure. Dis-
sociation of water and carbon dioxide in the flue
gases must be considered.

Solution:
To do this type of computation, a general computer
program in Fortran IV-G has been developed and tested
on an IBM 360 Model 40 computer. To use the program,
one keypunches a few IBM cards containing a description
of the problem, and then the data cards are submitted
along with a machine language program deck for proces-
sing.
The data needed for the calculation of the adiabatic
flame temperature are as follows:
1. Number of hydrocarbon species in the gas mixture
being burned.
2. Pressure in the burner.
3. Percent theoretical air (Must be equal to or greater
than 100%.)
4. Absolute humidity of the incoming air.
5. Number of moles of each hydrocarbon and of nitro-
gen in the fuel stream.
6. Heats of combustion at 2910K for all of hydrocar-
bon gases burned. (The heats to be used assume
that gaseous water is formed in standard state
reaction.)
7. Specific heats of all components in the range be-
tween 2910K and the flame temperature as a func-
tion of temperature.
8. Inlet reactant temperatures, which may be differ-
ent for each reactant.
When the adiabatic flame temperature is calculated, a
number of assumptions are made:
1. The process of combustion is adiabatic.
2. There is no secondary air. The hydrocarbons are
completely converted using only the primary air.
3. The only combustion products are carbon dioxide,
carbon monoxide, water, hydrogen, oxygen and
nitrogen. These products are at ths equilibrium
conversions controlled by the equilibrium constants
for the reactions
2H2 + 02 = 2H20
H2 + CO, = H,0 + CO


THEORETICAL FLAME TEMPERATURE (01A)
GENERAL CASE FOR NATURAL GAS MIXTURE


PRESSURE = 4.00 ATMOSPHERES
4 HYDROCARBONS)


DATA DESCRIBING THE REACTING
HEAT OF COMB. ENT MOLES
-0.189700E 06 0.866
-0.336732E 06 0.079
-0.484100E 06 0.027
-0.630620E 06 0.013
0.015
CAL/GMMOLE


PERCENT THEORETICAL AIR = 130.00
0.04300LB H20 PER LB DRY AIR

HYDROCARBONS IS TABULATED BELOW
MOLES C MOLES H2 ENT TEMP
1.000 2.000 500.000
2.000 3.000 500.000
3.000 4.000 500.000
4.000 5.000 500.000


DEGREES KELVIN


FORMULA OF HYDROCARBON OR EQUIVALENT HYDROCARBON
C 1.157 H 4.284
ENTERING MIXTURE BASED ON ONE MOLE ENTERING HYDROCARBON OR HYDROCARBON MIXTURE
MOLES OXYGEN 2.8964 MOLES NITROGEN 10.9109 MOLES WATER 1.0258


PERCENT COMBUSTIBLE GAS BY VOLUME =


6.32


CHEMICAL ENGINEERING EDUCATION


THE
FORMULA
CH4
C2H6
C3H8
C4HIO
N2


Input Data





Input VData











GENERAL FORM OF THE SPECIFIC HEAT EQUATION
CP = A + B*T + C*(T)**2 + D*(T)**3
CONSTANTS IN THE SPECIFIC HEAT EQUATIONS FOR THE REACTING HYDROCARBONS.
CP UNITS ARE CAL/GMMOLES-DEG. KELVIN
A B C D
CH4 0.47500E 01 0.30000E-02 0.33630E-06 -0.16450E-09
C2H6 0.94400E 00 0.37350E-01 -0.19930E-04 0.42200E-08
C3H8 -0.96600E 00 0.72790E-01 -0.37750E-04 0.75800E-08
C4H10 0.94500E 00 0.88730E-01 -0.43800E-04 0.83600E-08


THE DATA DESCRIBING ALL OTHER COMPOUNDS IS
HEAT OF DISSOC. ENT. TEMP A
CAL/GMMOLE KELVIN
0 0.57830E 05 500.00 0.832000E 01
* 0.67960E 05 500.00 0.770000E 01
0.0 500.00 0.664000E 01
0.0 500.00 0.673000E 01
0.0 500.00 0.673000E 01
0.0 500.00 0.673000E 01


TABULATED BELOW
B

-0.653000E-03
0.530000E-02
0.492000E-03
0.408000E-03
0.408000E-03
0.408000E-03


0.270000E-05
-0.8 30000E-06
0.319000E-06
0.486000E-06
0.486000E-06
0.486000E-06


-0.614500E-09
0.0
-0.740000E-10
-0.123400E-09
-0.123400[-09
-0.123400E-09


TOTAL MOLES IN THE EQUILIBRIUM MIXTURE
15.90913

FINAL MIXTURE COMPOSITION, MOLE PERCENT

C02 H20 CO H2 02 N2
7.22799 19.89369 0.04454 0.01823 4.23272 68.58292

PERCENT DISSOCIATION OF CARBON DIOXIDE = 0.61
PERCENT DISSOCIATION OF WATER = 0.09


ACTUAL K = 0.264E 04 COMPOSITION K = 0.265E 04


THEORETICAL FLAME TEMPERATURE = 2

4. The fuel contains only hydrocarbons and nitrogen.
The air contains exactly 79 mole % N2 and 21 mole
% 02.
5. The properties of the reactant and of the product
gases are calculated as mixtures of ideal gases.
6. The flame temperature must be between 25000K
and 19000K. This range includes the flame tem-
peratures of all major hydrocarbons likely to be
in a fuel gas.
The accuracy of the results obtained with this pro-
gram is limited only by the validity of the above assump-
tions and the accuracy of the input data. The program
user should try to insure that his specific heat data is


016.619 DEGREES KELVIN

valid for the temperature range of the problem he is
doing.
The results of computation for the above problem are
shown in the figure. The input data read from cards is
printed on the output as indicated. The other results
are calculated by the program except for the specific
heats of the output gases which are stored in the
program.
Further information about the program can be ob-
tained from Dr. David H. Chittenden, ChE Department,
University of New Hampshire, Durham, New Hampshire
03824.


LETTERS (Continued from p. 109)
For the 40 graduate-inclined schools
P = 4.07 + 0.00B + 0.17M + 0.51D (4)
and for the 57 undergraduate schools
P = 2.56 + 0.11B + 0.08M + 0.27D (5)
One immediately notices that in Eq. 4 the coefficient
of B is zero. This result may not be as outlandish as it
appears at first blush. It indicates that in graduate-
oriented programs the number of undergraduates may
not appreciably affect the number of full-time profes-
sorial schedules. One may speculate that this is the
result of a growing and perhaps regrettable practice of
relegating undergraduate instruction to persons without
professorial rank; i.e., instructors and graduate assist-
ants. This practice may be dictated by necessity in
rapid-growth situations.
If one compares the coefficients of D in Eqs. 4 and 5,
one notices that graduate-inclined schools require more


professional time per doctoral degree granted than do
undergraduate-inclined schools. The situation is reversed
in the case of bachelor's degrees.
In conclusion, we wish to remind the reader that no
great accuracy is claimed for this study. It represents
a first attempt to analyze the relationship between the
number of full-time professorial schedules and the num-
ber and kinds of degrees granted. Considering the nature
of the variables, it is indeed surprising that the indicated
degrees of correlation and stability exist. It would be
interesting to follow the study with future ones, not
only in chemical engineering but in other disciplines as
well.
A. X. Schmidt
Robert Pfeffer
Leonard Cohen
The City College of the City
University of New York


SUMMER 1969


Input Vata









THE DILEMMA OF INNOVATING SOCIETIES FREDERICKSON
(Continued from page 125.)


internal factors. Clearly, as a society we are sen-
sitive to and react to factors that arise outside
of our geographical boundaries. The cold war,
Vietnam, and the ABM are not strictly of our
own doing. However, here we want to consider
only those dynamics-determining factors that can
be considered to be internal, since it seems that
these are of greater significance than factors
that can be definitely identified as external.
The first of these internal factors is the
existence and propagation of what might be
called the Creed of Technology. This is the widely
held belief mentioned previously that innovation
per se is good and that technology is always pro-
gressive. It might be argued that no intelligent,
well-educated person really believes the Creed
of Technology. If that is so, why do so many
intelligent, well-educated persons act as if they
did believe it? And if such is true of intelligent,
well-educated persons, what beliefs are held by
simple or uneducated persons in this regard?
By questioning the belief that innovation is
always good, we are not thereby saying that it
is always bad. The point is simply that if a
technological innovation has a good side (as it
almost always does), it will more than likely
have a bad side as well. A case in point are those
technological innovations that have led to large-
scale agricultural use of fertilizers, herbicides,
and insecticides. Surely these have greatly in-
creased yields of crops and made possible a
minimal standard of living for the world's ex-
panding population. There is also no question
that such use has led to progressive and cumula-
tive deterioration and pollution of important sec-
tors of our environment. Again, it is certain that
these innovations have relieved the farmer of a
number of burdensome chores, while on the other
hand, it is possible to wonder if the farmer is
any happier for being so relieved.
Associated with the Creed of Technology is
the Cult of the Product. If the former teaches
that innovation of itself is a positive good, the
latter is the logical consequence of belief in that
teaching. If innovation of itself is good, then the
products of innovation be they good or serv-
ices must themselves be good and those who
deny this are either foolish or wholly mistaken.
"For," claim the votaries of this cult, "our pro-
duction of this thing or this service puts people


to work; it gives them useful labor in recompense
for which they can obtain the good things of
life. And our product itself satisfies a demand,
otherwise we would hardly be making it. There-
fore, it is patent on all counts that our product
is a good thing, and deserves the importance
that we attach to it."
Yes, but if your production pollutes the air
that society must breathe or the water that it
must drink, shall humanity at large share your
enthusiasm for your cult idol? If this production
necessitates tearing down and destroying the
things of natural or man-made beauty that enrich
life so much, shall the next generation of men
hold you blameless? Where in the Cult of the
Product is there any sense of balance?
A third factor, with roots perhaps deeper in
human nature than any other, is the Gospel of
Growth. This Good News is this: That the ex-
pansion of human activities (that is, the produc-
tion and consumption of goods and services),
institutions, and population is a wonderful thing,
a panacea for economic and even social ills;
without growth, society stagnates. With it, so-
ciety's possibilities are unlimited. Hence, indus-
trial or commercial enterprises are quick to
brand themselves as "growth companies," cham-
bers of commerce describe the community that
they represent as a "vigorous, growing area,"
and radio and television advertising puts out an
incessant barrage of propaganda designed to
convince the public of the sovereign merits of
industrial and economic growth. Placed before
us are bright visions of a high-energy society,
where man will control the weather or live under
a weather-excluding dome, the grass will not
have to be cut because it is artificial, the women
will always be slender, young, and beautiful, and
each of the kids can have a snowmobile for the
winter, a sports car for fall and spring, and a
speedboat for the summer. Or sometimes a dif-
ferent approach is taken; "If you don't grow
you're dead!" is the hellfire-and-brimstone way
to put it. And it is preached that way, appar-
ently with complete conviction that it applied to
human institutions and societies as it does to the
biological cycle of growth, development, senes-
cense, and death exhibited by individual living
organisms. This sort of preaching has won con-
verts, and it is difficult to escape the conclusion


CHEMICAL ENGINEERING EDUCATION









. death and decay are necessary for birth
and growth, and great Caesar's dust may
appear in an ear of wheat as in a
bunghole stopper.


that for a substantial portion of the most influen-
tial elements of our society, growth has become
an end in itself.
The biological aspect of the Gospel of Growth, viz.,
that population growth is an inherently good thing, is no
longer a very explicit part of the Gospel's teaching. The
social stresses and political and other problems brought
about by overpopulation are so evident in so many parts
of the world that it is no longer fashionable or expedient
to come right out and urge population growth. However,
the teaching that population growth is inherently good
is implicit in the Gospel of Growth. For if expansion of
the production of goods and services is a good thing,
then population growth must also be welcome, for besides
increasing per capital consumption, what other way is
there of increasing total consumption, the sine qua non
for increasing production?
F, AS SUGGESTED ABOVE, THE HELL-
fire-and-brimstone aspect of the Gospel of
Growth is based on the invalid analogy that the
development of a society follows the same rules
as the development of individual living organ-
isms, it is nevertheless permissible and even
necessary to view the development of societies
in biological terms; not in the sense of trying to
draw analogies, since these are always of doubt-
ful validity, but in the sense of trying to see what
restrictions biology places on such development.
For instance, American society is composed of a
biological population inhabiting a large but
strictly bounded living space. Very many other
biological populations, ranging from bacteria to
beef cattle, share this living space with us. All
of these populations man's included are
subject to biological, chemical, and physical laws
that cannot be suspended or amended by majority
votes of our legislatures. In that sense, man can
never "master" nature. In many cases, these
laws are not all known with precision or else
their application is too complicated for quantita-
tive prediction; in such cases, it is not possible
to predict what consequences human actions will
have on our environment. It is necessary to
mention this here, since the contrary belief -
that we can predict everything is quite com-
mon in our society, and is the essence of what we
shall call Technological Megalomania ( see be-
low).
Moreover, even if population growth is not an explicit
teaching of the Gospel of Growth, the current policies


of our society guarantee that the population will continue
to grow anyway. The demographer Judith Blake has
recently pointed out* that Americans of all classes con-
sider it desirable to have large (greater than two
children) families: the mean number of children con-
sidered desirable varies from 3.2 for high-income and
college-educated women to 3.6 or 3.7 for low-income or
grade-school-educated women. If that is what is considered
desirable, then that is what couples will strive to attain,
and so in fact the natility rate is much greater than that
required to achieve a stable population.
Professor Blake attributes the desire for large fami-
lies to "pronatalist policies" of our society, policies which
evolved primarily in response to the past need for a high
natality rate to counteract the then-prevailing high mor-
tality rate. She says (pp. 528 and 529) that these policies
"insure that just about everyone will be propelled into
reproductive unions, and [the female] half of the popula-
tion will enter such unions as a 'career' a life's work.
This rigid structuring of the wife-mother position builds
into the entire motivational pattern of women's lives a
tendency to want at least a moderate-size family. .
the wish for a family of a particular size relates
... to a need for more than one or two children if one is
going to enjoy 'family life' over a significant portion of
one's lifetime."
However, the predictions based on these laws
are quite clear even if not specific when they are
applied to a population that is growing unchecked
by any competing population, by disease, by in-
traspecific rivalries, or by lack of food: eventu-
ally one or more of these factors will exert them-
selves and bring about a check to population
growth. It is true that by very great expenditure
of effort an expenditure that we do not now
seem willing to make the density of popula-
tion at which such limiting factors come into
play can be raised. But how high can it be
raised? That is one of the things that cannot
be answered with precision. Concerning this,
the ecologist Lawrence Slobodkin writes* (pp.
3-4) : "How many men can the earth hold? We
must abandon all pretense of saving intact any
wilderness areas and consider that we will treat
the earth as a combined garden and factory; all
other species will either prove useful to man or
will be eliminated; they will either adjust to the
omnipresence of man or die. Answers to the
question are now merely guesses, ranging from
7 billion to 200 billion, the difference in the
estimates depending on how several subsidiary
questions are answered." And what would it be

* Judith Blake, "Population policy for Americans: Is the
government being misled?" Science, 164, 522-529 (1969).
* Lawrence B. Slobodkin, "Growth and Regulation of
Animal Populations," (New York: Holt, Rinehart, and
Winston, 1961).


SUMMER 1969








like to live in such a world? Again we quote
Slobodkin (p. 4) : "Implicit in this picture of the
future is a mental health problem: a world com-
pletely full of man and his activities could well
be a maddening place. There is an esthetic prob-
lem: the beauty of the winderness is a very real
thing. There is a political problem: a world full
of men would be highly regimented, a world of
an Aldous Huxley or Orwell phantasy." When
measured against this sober (and understated)
appraisal, the Gospel of Growth begins to lose
some of its appeal. And what about that aspect
of the Gospel calling for continual expansion
of man's activities and man's consumption, and
not just total expansion, either, but expansion
per person? This compounds the problems raised
by the population explosion and hastens the time
when some limiting factor beyond man's control
will appear.
If our (U.S.) population were stationary, and if our
per capital rate of consumption of goods and services
were stationary, both at their present levels, then we
could probably afford to indulge ourselves in the other
follies enumerated here. It seems perfectly reasonable
to assume that with existing technology, or with technol-
ogy within the reach of present knowledge, we could
provide the necessities of life and many luxuries in
addition, as well as a clean, esthetically-pleasing environ-
ment, for all of the 200,000,000 people who now live in
the United States. But of course, our population and our
per capital consumption are not stationary, and these are
facts of cardinal importance.
The Gospel of Growth is at once the most insidious
and the most dangerous of all the internal factors that
will be considered here. It is the most insidious because
its biological aspect is based on an eminently reasonable
postulate: that we must reproduce ourselves if human
life is to continue on the earth. It is the most dangerous
because it interacts with and is reinforced by all the
other factors that we are enumerating, and because by
itself, it is a sufficient condition for producing ecological
catastrophe. We need to take a long look at the Gospel
of Growth.
S TILL ANOTHER INTERNAL FACTOR
that determines the dynamics of our society
is that which gives rise to what my biomedical
librarian friend calls the Sandbox Syndrome.
She coined this term one day when we were try-
ing to get to downtown Minneapolis. That day
it seemed that our path was diverted at every
intersection by a huge hole dug in the earth for a
freeway, by some enormous piece of earth-
moving machinery blocking the street, or for
preparations to throw up a new skyscraper.
After the tenth detour, she exclaimed "Why,
this is the Sandbox Syndrome; dig, dig, dig all
the time!"


. the sandbox syndrome; dig, dig, dig .

In case there be any uncertainty about it, let
us place in more explicit if less expressive terms
that which Miss Bohn so aptly characterized:
We shall say that an individual, an organization,
or a society exhibits the Sandbox Syndrome if
his or its thinking tends to be determined by
technological considerations, if it is fixed upon
the size or the speed of machines or projects, or
if it habitually turns to digging in the dirt, rear-
ranging the landscape, and generally "improving
upon nature."
A case that was a classic example of the Sandbox
Syndrome appeared on TV a few weeks ago. An official
of one of the large airline companies was describing the
SST. Facts and figures concerning the weight, size,
thrust, speed, carrying capacity, length of runway neces-
sary for takeoff, etc., came spewing out of him as paper
comes spewing out of the University's CDC 6600 com-
puter. But never a word about sonic boom, release of
pollution high in the atmosphere where dispersal is very
slow, or the consideration that it may not be necessary
or even desirable to be able to fly from Minneapolis-St.
Paul to London in four hours. The whole thing was
reminiscent of a joke once told by a colleague: The pas-
sengers on the first commercial flight of the SST had
fastened their seat belts in preparation for blast-off and
the doors had been sealed. A voice came over the PA
system and announced that "This is a recording. The
doors of the craft have been sealed and the program for
takeoff initiated. In two hours we shall touch down in
London. This operation is completely computerized and
is not subject to human error. So relax, enjoy the flight,
and be assured that nothing can go wrong nothing can
go wrong nothing can go wrong. ."
The forms of the Sandbox Syndrome de-
scribed above are of a relatively primitive nature.
More advanced forms are sometimes manifested
and these should perhaps be differentiated from
the Sandbox Syndrome; we might call these
Technological Megalomania. Particularly strik-
ing here is a tendency to view the earth as some
sort of spacecraft, with the earth's human popu-
lation as its crew and all the rest of the earth's
populations as its life support system. The mis-
sion in which this super space vehicle is engaged
is tremendously exciting, if rather ill-defined.
Hence, we must manage the life support system
and harness it every organism of it to the
one task of supporting the crew. And of course
all of the crew must dedicate themselves to the
great task of completing the mission. Possibly
the crew might become somewhat restive in their
cramped quarters (even though these be scien-
tifically designed), but no doubt application of
psychology, genetic strain selection, and new in-


CHEMICAL ENGINEERING EDUCATION








It is absolutely necessary that innovating
societies think about why they innovate and
what the consequences of innovation are.



novations in crowd control will minimize any
disturbances.
Another striking manifestation of Techno-
logical Megalomania is a willingness to play the
Game of Environmental Russian Roulette. The
elements of this Game are first a technological
innovation magnificent in conception and gran-
diose in scope, second a considerable uncertainty
about the long-term environmental consequences
of implementing this conception but a possibility
that these consequences could be serious indeed,
and third a willingness to proceed with imple-
mentation anyway.

A STRANGE OPHTHALMIC DEFECT THAT
we can call Utilitarian Vision is still another
factor that directs technological innovation into
dangerous channels. Characteristic symptoms of
this defect are: Looking at a forest and seeing
piles of boards instead of trees, looking at an old
but well-kept neighborhood and seeing high-rise
apartment buildings instead of homes, and look-
ing at an unspoiled river valley and seeking a
power plant with a six hundred foot smokestack
instead of a place where future generations of
men can breathe and renew their spirit. Some-
times, the defect becomes so severe that the
boards, apartment buildings, and power plant of
the foregoing examples are distorted further into
dollars. When that symptom shows up, there is
little that anyone can do to correct the defect.
And please do not assume that because Utilita-
rian Vision has been illustrated by homely exam-
ples that strike most closely at the author's
heart that it does not operate on a much larger
scale.
Another internal factor of importance is the
Concept of the Convenient Society. If some of
the other factors that have been enumerated
arose far back in the past or are inherent aspects
of human behavior, the Concept of the Conveni-
ent Society is of recent origin. Indeed, to gauge
by the drumfire of advertising in its behalf, we
can probably infer that the principal financial
beneficiaries of the Concept are not perfectly
sure that it has taken complete root in the
thought patterns of the American people.


What is the Concept of the Convenient So-
ciety ? Simply this: That everyone ought to have
all of the conveniences that our technological
capacity can produce, and that that capacity
ought to apply its ingenuity to the fullest to
make things ever more convenient. And part of
convenience of course is easy of disposal when
an object is empty or worn out; we cannot worry
about what becomes of our conveniences when
we are through with them.
The basic trouble with the Concept of the
Convenient Society is not that convenience is
immoral or that we should go back to the "good
old days." Rather, it is the fact that the Concept
is totally at variance with the ecological concept
that in a limited, living world, all material things
must cycle if life is to continue indefinitely. To
put it in a different form, death and decay are
necessary for birth and growth, and great Cae-
sar's dust may appear in an ear of wheat as in
a bunghole stopper. The Concept of the Conveni-
ent Society either does not recognize the neces-
sity for the recycling of materials or if it does, it
ignores the necessity. Thus, a shiny new car
eventually ends up a rusty wreck and so our end-
less production of automobiles and concomitant
failure to reuse the worn-out ones leads to the
proliferation of auto junkyards. Again, a new
house in the suburbs becomes an old house in a
slum you can see this happening already and
few of our suburbs are more than twenty-five
years old so the better-educated, more affluent
people move on to a new suburb and start a new
cycle of development and decay; thus, the urban
blight spreads over the land. Perhaps the whole
thing can be epitomized by the Story of the
Aluminum Beer Can. It started out in the mind
of some ingenious innovator. Industry fashioned
it into a shiny vision of promise. Its production,
filling, and distribution provided useful work for
more than one deserving man. It gave pleasure
with convenience to someone else. And then it
ended up with glass bottles, plastic-coated milk
cartons, throw-away aerosol cans, aluminum
trays from TV dinners, and an endless variety
of other junk in an ever-growing and unholy
mountain that does not rust or rot, and whose
eventual disposition is or ought to be giving gray
hairs to the city fathers.

THIS ESSAY WOULD BE UNBALANCED UNLESS
we conclude the catalog of the foibles of our inno-
vating society with mention of the general ignorance of,
or indifference to, the full costs of technological innova-


SUMMER 1969








tion. Everyone wants the necessities of life, and everyone
also wants a greater or lesser share of the amenities of
life. Unfortunately, production of both necessities and
luxuries entails charges against the environment. It is
even more unfortuate that society at large has only a
dim awareness of the nature of these charges (or even
that they exist;) and blithely dismisses the thought of
charges with the assumption that they are not serious
or will somehow be paid by some unit of government or
industry. Of course, the charges sometimes become
blatantly obvious, as in the recent oil-drilling catastrophe
on the coast of California, and the public ire is then
aroused. This wrath tends to center on the offending
company or on the government official who permitted
the risk to be taken, and that is indeed proper. However,
fairness demands the remark that the public's own hands
are not entirely clean in this matter; after all, if no one
drove a car, would it be profitable to drill for oil off
the shore of California?
The trouble here is that no one has told so-
ciety at large what the environmental costs of the
innovating society are. The dissemination of such
information is desperately needed, for unless it
is available, one cannot see how a rational set of
priorities designed to balance man's needs against
environmental costs can be enforced. Hence, five
minutes of gibes at air pollution by Arthur God-
frey are no doubt worth one thousand essays like
this, but the author is compelled to write it none-
theless.
The set of factors that have been described
give to the dynamics of American Society some-
thing of the character of a branching chain reac-
tion. The creed of Technology and the Gospel of
Growth serve as the initiating reaction. Once
technology is initiated, it is propagated by the
Cult of the Product and the Concept of the Con-
venient Society. The Gospel of Growth gives
rise to branching reactions at all stages of the
process. And those defects in our thinking that
manifest themselves as the Sandbox Syndrome,
Technological Megalomania, Utilitarian Vision
and ignorance of the costs of the innovating so-
ciety, make it virtually impossible to introduce
any terminating reactions into the whole scheme.
We have now reached the point where we can
see that some terminating reactions are needed
if we are not to strangle ourselves. The air and
the water have become so badly polluted that the
situation is apparent to everyone. Hence, indi-
viduals, organizations, and industries of vision
and conscience have become concerned about
pollution control and the effects of pollution on
our environment and what is even more import-
ant, are acting on their concern. Thank God for
such; we are indebted to them. Nevertheless,


the efforts we have today are not nearly enough,
nor in the long run, do they strike at the heart
of the matter. Unless America changes some of
its basic attitudes the internal factors men-
tioned above it is hard to see how even very
great efforts to control pollution can be much
different from a rear-guard action. Hopefully,
they can keep us one jump ahead of the wolf
for some time, but what is really needed is some
way to chain the wolf and put him to work for us.
It is absolutely necessary for the continued existence
of a descent sort of human life on this earth, that inno-
vating societies think about why they innovate and what
the consequences of innovation are. We must recognize
that innovations are a means to an end and not the end
itself. We must try to set up some definite goals that
have the benefit of all mankind as their objective. We
can no longer rely on a vague faith in progress to take
care of our tomorrows. Only by setting up defined goals
will it be possible to develop priorities and institutions
that can guide the innovative genius of men onto paths
that will be truly, as opposed to superficially, beneficial.

I DO NOT THINK THAT WE ENGINEERING
educators have done very much that is useful
with regard to the considerations raised above.
True, we need and shall continue to need engi-
neers and scientists, and it is our business to
produce them. But what kind of engineers are
we turning out? It seems to me that much of
current engineering education serves to reinforce
belief in the Creed of Technology and the Gospel
of Growth, and all the rest of those factors,
factors with which the freshman engineering stu-
dent has already been partially equipped by his
parents, his schoolmates, and society in general.
Readers who have persevered this far and
who agree that I have described some real prob-
lems will probably wonder if I have any concrete
suggestions to offer for their solution or is it to
be simply "ecrasez 1' inflame an attitude all too
common these days. It seems to me that there are
some things we can do, and even if they are not
very original, I give them for what they are
worth.
The most immediate thing that we can do is
to consider the contents of our undergraduate
textbooks and courses. It is true that these do
not exhort budding engineers to go out and rape
the environment. But it is also true that they
do not suggest that they have any responsibility
to conserve it nor do they often state that tech-
nology imposes any stress on the environment. To
be aware that a problem exists is the prerequisite
for any attempt to solve the problem, and by judi-


CHEMICAL ENGINEERING EDUCATION









. the mystic destiny towards which innovating societies strive has lost some of the rosy tints of
paradise and taken on the more lurid aspects of purgatory.


cious choice of examples in engineering texts and
courses we could certainly point out to our stu-
dents what some of our real environmental prob-
lems are. Such examples serve another purpose
also: they can be genuine and challenging illus-
trations of basic engineering principles, ranging
from applications of the laws of thermodynamics
to problems of diffusion and convection that are
as advanced as anything in Birdfoot. Why don't
we try to put this sort of example into our texts
and courses? We don't have to expurgate as
Mother Goose must be expurgated; we merely
need to add things.
Another useful thing that we can do is to
make an attempt to recruit women students for
engineering. This would help to open up careers
for women other than or in addition to that of
wife and mother; it would be a removal of one of
the factors repressing antinatalist tendencies
existing but not active in our society. Removal
of factors repressing antinatalist tendencies is
Judith Blake's principal suggestion for inhibiting
population growth in the United States. By act-
ing on her suggestion, we would also tap a source
of talent and brainpower that has hardly been
touched by engineering.
The next thing that we can do is to see that
our students are at least exposed to courses in
environmental engineering. Does our school have
courses in air and water pollution control? Do
we encourage our students to take these courses ?
If we do not have such courses, what are we doing
to get them?
Finally, it is clear that an innovating society
will be able to resolve its basic dilemma only if it
can replace those attitudes and values that we
have named as the Creed of Technology, the
Gospel of Growth, etc., by attitudes and values
more in keeping with the long-term needs of man.
It is equally clear, at least to the author, that if
this is to be done at all, it must be done through
education. Hence, what about our own attitudes?
In our teaching of the details of technology, do
we ever stop to point out that technology is a
two-edged sword? Indeed, do all of us even be-
lieve that? Do we think it worthwhile for our
students to learn something besides science and
engineering or do we regard the "liberal educa-
tion" part of our curriculum as a necessary evil
forced upon us by the rest of the university? Do


we have any concern that our students should
realize that we live in a limited world, a micro-
cosm, whose living and non-living components
interact in an endless spectrum of ways, some
of them of extraordinary subtlety? Do we think
students should be aware that man is part of this
microcosm or are we so far gone into the last
stages of Technological Megalomania that we
teach or at least imply that he is above and be-
yond it? Have we swalloded the Concept of the
Convenient Society which is in essence that we
can do just as we damn well please with our
surroundings so that we pass this monstrosity
on to the next generation of engineers?
We are concerned these days with a decline
in the number, or at least of the relative number,
of students entering college who want to become
engineers or scientists. Surely there is no single
or simple explanation for this, but I do think
that we can gain partial insight into the phe-
nomenon by applying the analysis given above.
Eighteen year olds can observe and draw conclu-
sions as we can. Their observations may not be
as thorough, and their conclusions may not be
based on a very disciplined or experienced
thought process; nevertheless, the Creed of
Technology, the Cult of the Product, and all the
rest are fairly obvious aspects of our society,
and it is easy to see how they could become inex-
tricably entangled in the minds of youth with
science and engineering. The fact that these fac-
tors are faults of society as a whole rather than
of technology alone is irrelevant; society's de-
fects are manifested in a most striking way by
our runaway technology.
To close this essay, let us return to Professor Pig-
gott's opening chapter. He speaks (p. 18) of a "mystic
destiny" towards which innovating societies believe they
should strive by continual technological innovation, and
his quote from Herbert Spencer may be taken as the
canonical form of the Creed of Technology: "Progress is
not an accident but a necessity. It is a part of nature.
Evil tends perpetually to disappear." Unfortunately, the
stock of evil seems these days to increase even faster
than the population. Change is a part of nature but
that which is called Progress is often a retrogression, or
if it is of benefit to us, it would be a curse to the next
generation. Thus, the mystic destiny towards which
innovating societies strive has lost some of the rosy tints
of paradise and taken on the more lurid aspects of
purgatory.
I am indebted to Carol Urness for her con-
structive criticism of my original manuscript.


SUMMER 1969

























J department


THE GATORS GO

RAY FAHIEN

What should be the goals of a department of
chemical engineering? National prominence
through a strong graduate program or a quality
undergraduate program? An orientation toward
"engineering science" or an orientation toward
engineering practice? A large graduate program
or a large undergraduate degree production? A
PhD-oriented graduate program or a master's-
oriented graduate program? Should it espouse a
philosophy of service to the state, a philosophy
of service to the engineering profession, or to
a "community of scholars"?
When a department has a single objective,
the fulfillment of its goal demands a concentrated
effort in one direction. For example, a depart-
ment that is interested in undergraduate degree
production can hire faculty who are inspiring
teachers and who would also enjoy visiting high
schools to aid in recruitment; a department that
aspires to national prominence for the quality
of its research can hire faculty who have bril-
liant, creative minds and a personal desire to do
research and to publish their results. When the
goal of the department is singular, and when the
faculty and administration accepts the singu-
larity, the implementation of its goal can be
carried out smoothly and without conflict. De-
partments of great prominence can be developed
in this manner.


. a balanced department with multiple
objectives is desirable at the University of Florida .

In many cases, external factors, such as
whether it is a private or public institution, may
influence or even fix the goal of the department.
In some cases, it is more desirable (and even
necessary) for a department to have multiple
goals. For example, the composition of a tenured
faculty can, by its very nature, demand a divers-
ity of objectives; or the faculty may express an
objective opposed to that of the institution or
college as a whole (such as in the case of an
undergraduate-oriented faculty in a graduate-
oriented institution).
While many departments seek more than one
of the many objectives listed above, few of them
strive for excellence in all of them. But one de-
partment that, for the last four or five years, has
been attempting to do all of these is the Chemical
Engineering Department at the University of
Florida. We might therefore properly ask the
following questions:
What are the reasons for such a multiplicity
of objectives?
What kind of results have been achieved ?

JUSTIFICATION FOR BALANCE
A balanced department with multiple objec-
tives is desirable at the University of Florida
because of the following:
It is the only department of chemical engineering
in a state serving over six million people; it therefore
feels a broad responsibility to provide a diversified and
balanced program.


CHEMICAL ENGINEERING EDUCATION









a department must recognize both academic and industrial professional goals its achievements will never
be easily measured by quantitative indices an engineer is not merely a technical robot the goal
of the department, and that of the student it educates, must be the betterment of human society.


Its tenured faculty in 1964 was already of above-
average size; it was a diversified, heterogeneous and
capable group that was brought together to do sponsored
research under the Engineering and Industrial Experi-
ment Station and partly to teach. It consisted of several
people with degrees in chemistry (organic, biochemistry,
pharmaceutical, inorganic and physical), a pulp and paper
technologist, an authority on imbedding flowers in plas-
tics, an expert on asphalt technology and economics, a
world-famous fluorine chemist (and philosopher of science
and education), and an electrochemist doing over $100,000
a year of research (much of it classified) on thermal
batteries and fuel cells. Some of the faculty were inter-
ested only in undergraduate teaching, others only in
sponsored research, still others in both activities. Its
average age was 55 and there was only one assistant
professor in the group.
The new energetic and dynamic dean of the College
of Engineering from 1964-68 strongly encouraged the
development of the graduate program both from
the standpoint of increased enrollment, and also in terms
of quality of research and instruction.
It began participation in 1965 in an NSF Science
Development (or "Center-of-Excellence") Grant that
provided funds for bringing in new faculty and graduate
students to do fundamental research in chemical engi-
neering.
The diversity of faculty interests and back-
grounds plus the strong leadership of the dean
made it obligatory for the department to pursue
multiple objectives lest it suffer from internal
conflict among its faculty or from an external
gap between the objectives of the administration
and those of the faculty. Hence the objectives of
the department became balanced ones and diversi-
fied ones; both quality and quantity were needed
in the graduate and undergraduate program;
both theory and practice had to be emphasized;
both teaching and research had to be acknowl-
edged; both masters and PhD degrees had to be
offered; and both service to the State and Nation
as well as service to the engineering profession
and to the academic community had to be a part
of departmental philosophy.
To express these multiple goals in a cohesive
philosophy became a first task of the new chair-
man when he arrived in June 1964. A statement
of goals recognized the diversity of the chemical
engineering profession through its strong roots
in both chemistry and physics. This diversity
meant that a chemical engineering department
must recognize bifold professional goals: aca-
demic goals that strive for the advancement of
fundamental knowledge and industrial (or pro-


fessional) goals that have to do with the econom-
ical design and operation of plants that produce
consumer goods (or of substances that go into
making consumer products). "Just as the overall
aim of the University is to serve mankind," it
further stated, "so also the goal of the depart-
ment, and that of the student it educates, must be
the betterment of human society. For as a pro-
fessional man, an engineer is not merely a techni-
cal robot who responds passively and unquestion-
ingly to conformist pressures or to the commands
of others. Instead he must be aware of, and
deeply concerned with the social and political
problems of our times. He must have a high
sense of values and be capable of making deci-
sions with regard to principles and ideals derived
from these, rather than from narrow self-interest
or partisan group interest. In keeping with this
philosophy, the department should investigate
methods of establishing communications between
the 'two cultures' of technology and the humani-
ties."
Somewhat later the first annual report for
the "Center of Excellence" Grant stated that "the
goal of the chemical engineering department in
the Science Development Program is to strive
towards an excellence that is better expressed in
terms of the significance of its contributions to
scientific progress than by the volume of its
activity. Its achievements therefore will never
be easily measured by quantitative indices-by
numbers of students or faculty added, by the
dollars worth of equipment purchased, by
the number of papers in various journals, or
by the number of degrees granted. We believe
that the kind of excellence for which we strive
cannot readily be programmed, budgeted or al-
located on a yearly or semiannual basis. Nor can
it be fully accomplished in a time space of one
year or three years or even five years.
"But seeds can be planted. Morale can be
improved. Research ideas can be generated. New
approaches to engineering education can be tried.
A creative intellectual atmosphere can be devel-
oped. Bright, highly motivated people, both
young and old, can be added to a faculty. A new
life, energy, and enthusiasm can be breathed into
a faculty with unfulfilled goals and unrealized
potential.


SUMMER 1969








seeds can be planted .

GRADUATE ENROLLMENT TRIPLES
With the addition of a new chairman and
three other faculty members, the chemical en-
gineering department began its period of devel-
opment in 1964-65 one year prior to the award
of the "Center of Excellence" grant. Although
this development was greatly accelerated by the
award of the grant, it was partially retarded by
inadequate space and facilities. Until November,
1967, the department had been housed entirely
in a crowded World War II airplane hangar
which was shared with the Aerospace Engineer-
ing Department. Essentially no additional space
was available for graduate students and research
equipment, and faculty offices were not condu-
cive to the recruitment of prominent senior
faculty members. But in less than three years the
department could point to the following accom-
plishments :
Revision of graduate and undergraduate curricula.
Graduate enrollment nearly tripled increasing to 66.
Undergraduate degree production increased 50%.
Seven outstanding young faculty members with
excellent backgrounds were added, decreasing the
average age of the faculty from 55 to 44.
Sponsored research support increased over 50%.
Faculty research productivity in terms of papers
submitted and published increased several fold. Two
books were published and two others were started.
Although the above quantitative increases
may be startling, even more impressive were the
indications of improvements in the quality of its
graduate student body, its faculty, and its gradu-
ate program. In 1964, over half the 23 graduate
students were foreign students, and roughly half
were University of Florida graduates. The aver-
age Graduate Record Examination scores of that
group was 550 or slightly above average. How-
ever of the group of 28 students admitted in
Fall, 1967, all but two were graduates of Ameri-
can institutions other than the University of
Florida. The first group of 22 who accepted ap-
pointments had an average Graduate Record
Examination (Verbal-Quantitative Average)
score of 654 at least one standard deviation
higher than in 1964. (Such a score meant that
the average student was in the top 6-7% of the
senior students throughout the nation who took
the examination.)
Initially, the award of the "Center of Excel-
lence" grant made possible a shift in the research
emphasis of many of the older faculty members


to more fundamental areas of research and away
from the highly applied sponsored research proj-
ects previously emphasized under the Engineer-
ing and Industrial Experiment Station. As a
result of this heightened interest, new proposals
for fundamental research were written and nine
new projects were accepted for support by vari-
ous agencies such as the NSF, NIH, AEC, NASA
and DOD. Thus stimulated, the face value of
sponsored research nearly doubled and the an-
nual rate increased by over 50%.
The increase in outside research support
made available state funds for the addition of
new faculty beyond the two positions allocated in
the grant. (In addition to positions generated
through research, one state supported position
was obtained from the University.) The new
faculty added were not only graduates of leading
institutions; they were also generally among the
top students to complete PhD work at their in-
stitution over a period of years. (Four of them
had won NSF Fellowships in national competi-
tion.) Table 1 gives their backgrounds.

TABLE 1. FACULTY ADDITIONS SINCE 1964*


Name and
Ph.D. School


Area


A. W. Westerberg Computer-Aided


London


L. E. Johns, Jr.
Carnegie Tech
J. P. O'Connell
Cal. Berkeley

X. B. Reed, Jr.
Minnesota
A. D. Randolph
Iowa State
(Now at U.
Arizona)
D. W. Kirmse
Iowa State
K. E. Gubbins
London
R. W. Fahien
Purdue


Design


Other
Background
Control Data Cp.
Princeton U.
U. of Minnesota


Polymer Dynamics Dow Chemical
Cont. Mech.


Thermodynamics
Transport
Properties
Bioengineering
Appl. Math
Crystallization-
Particulate
Systems

Turbulence

Transport
Properties
Transport
Processes in
Reactors


Mass. Inst. Tech.
Pomona Coll.
Union Oil Co.
U.C.L.A.
Texas A & M
Amer. Potash
Spencer Chem. Co.
Colorado U.

Union Carbide
Oklahoma State
Florida U.
(Post Doc)
Ethyl Corp.
Iowa State U.
Missouri (Rolla)
Washington Univ.


*Ronald Gordon (Ph.D. expected from Princeton Univer-
sity) will join faculty in September 1969.
(Continued on page 157)


CHEMICAL ENGINEERING EDUCATION








would you like


to plan a plant


in Puerto Rico?


b


Too late, the plant is planned! In fact
construction is already beginning on
Sun Oil's new $125 million refinery
complex and harbor at Yabucoa.
But the project at Yabucoa is sim-
ply one indication of Sun on the move.
We're geared for growth and we need
people. Maybe you ?
Perhaps you'd like to work for the
company that also recently boomed
into the 2 billion dollar class through
the merger of Sun and Sunray DX;
that's pioneering a new fertilizer


plant on the island of Martinique;
that operates a new Computation
Center in Philadelphia; that spon-
sors winning teams and cars in major
road racing championships in the
United States and Canada-to men-
tion just a few exciting projects.
We need men and women to grow
with us and build the future. We have
openings in Exploration, Production,
Manufacturing, Research and De-
velopment, Engineering, Sales, Ac-
counting, Economics and Computer


Operations. Locations-Philadelphia,
Toledo, Tulsa, Dallas and many other
areas.
Write us for an appointment, write
for our book "Sunoco Career Oppor-
tunities Guide," or contact your Col-
lege Placement Director to see Sun's
representative when on campus.
SUN OIL COMPANY, Industrial Rela-
tions Department, CED, 1608 Walnut
St., Phila., Pa. 19103.
An Equal Opportunity Employer M/F


---.
-^<^->











C5BS3


curriculum 4, eat CEE ffepadmemd


FLEXIBLE CURRICULA CAN BE STRONG


RAY FAHIEN
MACK TYNER
R. A. KEPPEL
University of Florida
Gainesville, Florida

UNDERGRADUATE PROGRAM IS DIVERSIFIED

There is much truth in the old story about the
late G. G. Brown telling a questioner that "chemi-
cal engineering is what chemical engineers do."
When asked what chemical engineers do, his
reply was, naturally, that they do chemical en-
gineering! But, contrary to the belief of certain
narrow-minded persons, there is no single well-
defined job that each and every chemical engineer
does and which can be described as THE work
of "THE chemical engineer." For in practice
individual chemical engineers do numerous kinds
of jobs each requiring different talents, abilities,
and interests.
In spite of this great diversity of the profes-
sion, chemical engineering curricula have for
years consisted of a rigid set of courses that
every student was required to take. No consid-
eration was made of his future career objectives,
his personal interests, or his individual abilities.
Although the typical curriculum might include
perhaps two elective courses, these were often
gleefully squandered in subjects such as bait
casting, photography, or basket weaving. In view
of our special nature as the only chemical engi-
neering department in the State of Florida, we
felt that we have special obligations to provide
as broad a program as possible. Accordingly,
in 1965 we developed a curriculum that treated
each student as an individual-one whose indi-
vidual interests, talents, and career objectives
could be expressed through a selection of option
programs.
The considerations that were involved in de-
veloping this curriculum are as follows:
First, we reaffirmed the belief that an engineering cur-
riculum must be designed to prepare not only a broadly
educated person but also train a professional man who
could, upon graduation, do the kind of engineering work


that employers have associated with chemical engineer-
ing. Both of these educational objectives had to be ful-
filled in four years since a five-year program has never
been accepted by students, educators, or industry. How-
ever, it seemed neither possible nor desirable to devise a
single four-year program that would prepare the student
for all the various kinds of work that chemical engineers
do. For it seemed wrong to subject the practice-oriented
student to intensive theory and mathematics that he would
not use; and it seemed equally unproductive of human
talent to deprive the science-oriented student of that kind
of experience that would best prepare him for graduate
school and a career in research.
Second, we felt that it is essential that all programs
in a chemical engineering department include those
fundamental and core courses required for the practice
of chemical engineering and that no program should be
"watered down" so as to become an easy path to a cheap
degree.
Third, we considered it important that a student not
be unduly harmed by a wrong decision as to the program
he chooses to enter. Therefore we made the differences
between programs only about 10%. Actually since we
require 213 quarter hours (142 semester hours) for a
bachelor's degree, more leeway was possible than in
schools requiring much fewer hours.
Fourth, we felt that a student should not be forced
to make a decision until his senior year so that he is
experienced and mature enough to make a wise choice.
Fifth, we believed that the option programs should not
be so rigid that a student is prevented from substituting,
when his reasons are sound, certain courses in the options
for other courses.
The above general considerations were em-
ployed in the development of specific programs as
follows:
Chemical engineering science option. Throughout the
country a strong trend has developed toward a chemical


HIGH SCHOOL -
Pre-Engineering
Courses
JR COLLEGE i
CORE AREAS
Chemistry
Eng Science
Chem Eng


UNDERGRAD OPTIONS
ChE Science .
ChE Systems a.
Process Eng
Interdisciplinary "
Practice Options C ,
aduate a Industry
hool


Figure 1. Flow Sheet for Undergraduate Options.

CHEMICAL ENGINEERING EDUCATION










. we developed a curriculum that treated each student as an individual whose interests,
talents, and career objectives could be expressed through a selection of option programs.


engineering science program that primarily prepares a
student for graduate work and for a career in funda-
mental research and teaching. Although a large majority
of our students go into industry we felt that our depart-
ment should have such a science-oriented program avail-
able in order to provide (nevertheless) an alternative
for the student who might otherwise go into the engineer-
ing science program available in the College of Engineer-
ing. We call this program our chemical engineering
science option.
Operations, business, and technical sales options.
Many of our students obtain jobs in operations and
eventually in middle management where theoretical and
mathematical tools are not as much needed as by the
researcher. Programs for such students were designed
for industrial employment or further study in business,
marketing, or law. These were called the operations
options and business options. A modification is available
for students interested in careers in technical sales.

TABLE 1 CORE AREAS


TABLE 2A SUBJECTS STUDIED IN
APPLIED SCIENCE OPTIONS


$ BE .cw
k a
B m*
h ^ C o


Reactor Dynamics and Design
Math Models in ChE
Technical Electives
Process Optimization
Applied Molecular Theory
Advanced Process Design
Ind. and Systems Eng.
Applied Math Electives
Polymeric Materials
Process Economics


* *
* *
* *
*


*
* *


Tensor Fields and Fluid Dynamics


Quarter Credits in Option


22 22 22


Engineering
Core
Computer Model
Formulation*
Intro. to Elec.
Eng.
Statics

Strength of
Materials
Materials of
Engineering*
Engineering
Statistics


Chemistry
Core
Organic
Chemistry
Organic Chem.
Processing*
Physical
Chemistry
Instrumental
Analysis*


*Taught in ChE department.


Chemical
Engineering Core

Thermodynamics*

Transport Phenomena*

Chemical Kinetics*

Systems Analysis*

Solid-Fluid Systems*

Control Theory*
Stagewise Separations*
Cost Estimation*
Process Design*


Process engineering option. Many students actually
do not know what they want to do after graduation. As
a result, chemical engineering curricula have traditionally
attempted to produce a highly versatile chemical engi-
neer one that can easily start his career in any of
many work assignments. We felt a program of this type
should be retained. We called this option process engi-
neering but upgraded it by the addition of transport
phenomena, computer modeling and applied math courses.
Systems engineering option. The systems engineering
approach is as much a part of chemical engineering as
any other field of engineering, including that taught in
systems engineering departments. Chemical engineers
are bringing the fruits of automation to the process
industries through the effective use of computers, mathe-
matical models and processes, and advanced hardware to
the design and operation of chemical complexes. Conse-
quently we developed a computer-oriented systems engi-
neering option in order to fill this need and also as an


alternative to the student who might otherwise enroll
in the systems engineering program taught in the College
of Engineering.
Interdisciplinary options. Some students have a dif-
ficult time choosing between chemical engineering and a
"glamorous" field such as aerospace or nuclear engineer-
ing. At the same time the challenging problems of today
tend to be coupled interations between the application of
engineering principles and the socio-economic needs of our
society; e.g., pollution abatement, food production and/or
population control. The inter-disciplinary options permit
the chemical engineer with paralleled interests in other
fields to take his degree in chemical engineering while at
the same time studying 22 hours of approved courses
in related disciplines such as aerospace engineering,
environmental engineering, nuclear engineering, food
science, or biomedical engineering. The latter program
is approved for direct entrance into medical school.
Humanities or Liberal Studies. Many students today
are concerned about the social problems of our society,
about man's obligation to his fellow man and himself,


TABLE 2B INTERDISCIPLINARY OPTIONS


Chem-Aerospace
Propulsion
Aerodynamics
Chem-Electrical
Electronics
Control Systems

Chem-Food Science
Chem. Principles
Eng. Principles

Chem-Mechanical
Turbines and Jets
Refrigeration


Chem-Biomedical
Zoology
Biology
Chem-Environmental
Waste Treatment
Special Topics

Chem-Materials
Elect. Properties
Corrosion

Chem-Nuclear
Nuclear Tech.
Nuclear Chemistry


SUMMER 1969









TABLE 2C PRACTICE OPTIONS


Operations
Corrosion or
Electrochemical
Engrg
Polymeric
Materials
Management
Electives
Process
Economics
ChE Electives


Business & Sales
Report Writing
Speech Courses
Management
Electives
Marketing
Electives
Process
Economics
ChE Electives


Humanities, or
Liberal Studies
Political Science
Sociology
History
English
Philosophy
Religion
Foreign
Languages
Process
Economics
ChE Electives


about values and ethics, and about the meaning of life
itself. Therefore we developed a group of courses in the
humanities and social sciences that would permit an
engineering student to obtain a degree that was perhaps,
because of its science content, more of a "liberal studies"
program than that offered in the College of Art and
Sciences.

GRADUATE PROGRAMS IN SCIENCE AND SYSTEMS
While some graduate students intend to teach
and do basic research, many others are interested
in industrial careers in development and design.
Consequently the graduate program in the de-
partment was divided into three main areas:
(1) Chemical Engineering Science: transport phe-
nomena, fluid dynamics, thermodynamics, kinetics, micro-
structure of matter, and materials science; (2) Chemical
Engingeering Systems: chemical reaction engineering,
process control, process dynamics, optimization, separa-
tions processes; and (3) Interdisciplinary Chemical Engi-
neering: energy conversion and fuel cells, polymer sci-
ence, microelectronics, process economics, and bioengi-
neering.

Master of Engineering with Project
There are four graduate programs in the
department. These are: 1) the 45 quarter-hour
Master of Science in Engineering program with
thesis;2) the 50 quarter-hour Master of Engi-
neering pre-PhD program; 3) the 50 quarter-
hour Master of Engineering terminal program
with a project (which can involve a design, a cost
analysis, an experimental investigation, a com-
puter study, or a technical report) ; 4) the PhD
program. Transfer between programs is possible
within limits.

PhD Program Requires Research Proposal

This program includes a written examination
(which may be waived), see below, a research


proposal defended orally, an oral examination,
and a final examination. The research proposal
sets forth and describes an original research
problem and/or solution, which if carried
through, would represent a significant contribu-
tion to chemical engineering knowledge. The area
of the proposal may be the student's dissertation
subject only if he has taken the written exam-
ination.


Graduate Courses
An orientation examination is used to deter-
mine whether entering students require any pre-
liminary course work, before taking the six
required core courses. After three to five quarters
in residence, all Master of Engineering students
and nearly all PhD students are required to take a
written examination based on these core courses:

Models and Methods
Multidimensional and Discrete Systems
Thermodynamics of Reaction and Phase Equilibria
Fundamental Transport Phenomena
Process Dynamics 1 or Process Dynamics 2
Reactor Design and Optimization (Systems Program) or
Chemical Kinetics (Science Program)


During 1968 the following additional courses
were taught by department faculty:

Mathematical Methods in Chemical Engineering
Applied Field Theory
Computer Control of Processes
Optimization Techniques
Transport Properties and Irreversible Thermodynamics
Applied Statistical Mechanics
Statistical Thermodynamics
Interfacial Transport Phenomena
Rheology
Non-Newtonian Fluid Dynamics
Chemical Energy Conversion
Particulate Systems
Applied Fluid Dynamics
Process Engineering
Process Equipment Design
Process and Plant Design
Process Economy Analysis
Tensor Fields and Fluid Dynamics
Analytical Techniques for Eng and Scientists 1
Analytical Techniques for Eng and Scientists 2
Analytical Techniques for Eng and Scientists 3


RESULTS
Student reaction to these diversified programs
has been very good at both the undergraduate
and graduate levels.


CHEMICAL ENGINEERING EDUCATION











THE GATORS GO
(Continued from page 152)

"CENTER-OF-EXCELLENCE" GRANT
In August 1965, the University of Florida
was awarded a 4.2 million dollar Science Develop-
ment Grant by the National Science Foundation.
The Chemical Engineering Department was
among the seven participating departments in the
University. The proposal submitted by the Uni-
versity was entitled "Radiation, Kinetics, and the
Microstructure of Matter." The proposal stated
that the first objective of the College of Engi-
neering was "to improve the scientific base of ed-
ucation and research through increased emphasis
on the engineering implications of the microstruc-
ture of matter." It pointed especially to the de-
veloping technology of microelectronics as "only
one aspect of the very general field of microengi-
neering which aims to place a strong emphasis
upon the microscopic statistical view of nature
and to relate this to human needs."
In keeping with this philosophy, the department de-
fined and delineated the meaning of the phrase "micro-
structure of matter" from the standpoint of modern fun-
damental research in chemical engineering in terms of
the following connotations:

1. MOLECULAR. This approach involves the use of
a knowledge of statistical mechanics, molecular structure,
and molecular and kinetic theory (a) to predict rates of
chemical reaction either on catalyst surfaces or in homo-
geneous systems, (b) to predict adsorption rates, (c) to
predict thermodynamic properties and phase equilibria,
or (d) to predict transport properties such as diffusivity,
thermal conductivity, or viscosity.
2. PARTICULATE. This approach analyzes par-
ticulate systems in terms of their statistical properties
and the particle-continuum interaction. Such systems are
found in industrial crystallizers and also include aerosols,
mists, dispersions, and suspensions.
3. STATISTICAL. This approach is used to describe
turbulent transport processes for energy, mass, and
momentum in terms of elements in which fluctuations of
velocity and other properties occur.
4. CONTINUUM. The microscopic view of matter
can be thought of in terms of processes that occur at a
point in a continuum. The conservation laws for energy,
mass, and momentum can be expressed in terms of the
differential equations of change.

Knowledge of matter from these microscopic
points of view of course can be used in a given
engineering system to predict macroscopic quan-
tities such as the total energy or mass transport
or the total friction or drag in a system. This


information can be incorporated with modern
design and optimization techniques in the design
of an engineering system or a complete plant.

NEW BUILDING FOR DEPARTMENT
In the fall of 1967 the department was able
to move into a modern air-conditioned educa-
tional building containing 51,000 sq. ft. of re-
search and teaching facilities made possible by
a State bond issue and funds from the NSF grant.
We now have undergraduate teaching space for
modern laboratories in process measurements,
transport properties, instrumental process analy-
sis, unit operations, process transients and con-
trol theory, chemical reaction kinetics, and indi-
vidual special projects. Graduate research space
is available in process dynamics and computer
control, transport phenomena and properties, in
vivo transport studies, fluid dynamics and
rheology.

COMPUTER CONTROLLED LABORATORY
Modern computer facilities will permit one to
control any of several pieces of process equipment
in the unit operations laboratory. At present, a
distillation column is being tied to a remote IBM
1070 process control terminal which connects to
the IBM 360/65 campus computer via telephone
lines. We have designed and are building a special
interface between the process equipment and the
terminal which serves two major functions. First
of all it is a patch panel permitting any one of
several processes to be "patched" into the ter-
minal using special jacks and plugs. Its second
function results from the fact that one can
simulate most of the computer actions to the
process and all of the process responses to the
computer at the interface itself. One can thus
almost completely "debug" the computer soft-
ware without the process and to some extent
"debug" the process hookup without the com-
puter.
The remote computer terminal with interface
can tie to 40 analog inputs (low and high level),
30 digital inputs, 24 digital outputs, 10 pulse
motor outputs (which can operate in parallel),
a digital display, and a rotary switch input sta-
tion. The terminal's transmission rate is 66 char-
acters per second to and from the computer which
will permit about 4 random accesses per second or
about 20 analog to 60 digital sequential accesses
per second.
The software is written in Fortran and is


SUMMER 1969









quite modular permitting most of the essential
portions to be used in all processes. The depart-
ment also has two remote consoles for the IBM
360, a 60-amplifier Ease computer, and a WANG
Calculator.

OTHER ACCOMPLISHMENTS
During the past three years, three different
members of the faculty have won undergraduate
teaching awards; Professor Tyner, Professor
Gubbins, and Professor O'Connell. This year the
Sigma Xi research award went to a chemical
engineering graduate student and the Phi Kappa
Phi award for the outstanding student in the
University went to a chemical engineering junior.
Last year the faculty published 20 papers, had
14 others accepted, and submitted 14. Two books
were published, two accepted and two submitted.

CONCLUSION
If the goal of the department is an excellence
that is not measured by quantitative indices, the
above achievements are not in themselves suffi-
cient indication that excellence has been attained.
But they may indicate that the seeds of excellence
have indeed been planted and have germinated. If
these are now nurtured by additional support,
the progress of the department toward excellence
can continue not only in its research program,
not only in its instructional program, not only in
the achievement of each of its multiple objectives,
but also in the fulfillment of its ultimate aim:
the betterment of human society.

(Cont'd from p. 140)
OPTIMIZATION: R. R. Hughes
BIBLIOGRAPHY
1. Andersen, S. L., Chem. Eng. Prog., 57, No. 3, 80-83
(March, 1961).
2. Aris, R., G. L. Nemhauser, and D. J. Wilde, AIChE
J., 10 913-919 (Nov., 1964).
3. Baumol, W. J., "Economic Theory and Operations
Analysis," 438 pp., Prentice-Hall, Englewood Cliffs,
N. J. (1961).
4. Blakemore, J. W. and S. H. Davis, Jr., edit. "Op-
timization Techniques" Chem. Eng. Prog. Symp.
Series No 50, 60, (1964).
5. Carr, C. R., and C. W. Howe, "Quantitative Decision
Procedures in Management and Economics Deter-
ministic Theory and Applications" 383 pp., McGraw-
Hill, N. Y. (1964).
6. Dantzig, G. B., "Linear Programming and Exten-
sions" 625 pp., Princeton Univ. Press, Princeton,
N.J. (1963).
7. DiBella, C. W., and W. F. Stevens, I & EC Process
Des. and Dev., 4, 16-20 (Jan, 1965).


8. Ford, L. R., Jr., and Fulkerson, D. R., "Flows in
Networks" 194 pp., Princeton Univ. Press, Princeton,
N. J. (1962).
9. Franks, R. G. E., "Mathematical Modeling in Chemi-
cal Engineering" 285 pp., J. Wiley, N. Y., (1966).
10. Gass, S. I., "Linear Programming Methods and
Applications," 2nd edit., 250 pp., McGraw-Hill, N. Y.
(1964).
11. Graves, R. L., and P. Wolfe, edit., "Recent Advances
in Mathematical Programming," 347 pp., McGraw-
Hill, N. Y., (1963).
12. Griffith, R. E., and R. A. Stewart, Mgt. Science, 7,
379-382 (July, 1961).
13. Hadley, G., "Linear Programming," 520 pp, Addison-
Wesley, Reading, Mass. (1962).
14. Hadley, G., "Nonlinear and Dynamic Program-
ming" 484 pp., Addison-Wesley, Reading, Mass.
(1964).
15. Happel, John, "Chemical Process Economics," 291
pp., J. Wiley, N. Y. (1958).
16. Hertz, D. B., Harvard Bus. Rev., 42, No. 1, 95-106
(Jan.-Feb., 1964).
17. Hughes, R. R. and J. C. Ornea, "Decision-Making in
Competitive Situations," Paper, Panel Disc. 34, 7th
World Petr. Congr., Mexico City, (April, 1967).
18. Hughes, R. R., E. Singer, and M. Souders, "Machine
Design of Refineries," Proc. 6th World Petr. Con-
gress, Frankfurt/Main, Section VII, pp. 93-102,
(June, 1963).
19. Lavi, A. and T. P. Vogl, edit., "Recent Advances in
Optimization Techniques," 656 pp., Wiley, N. Y.,
(1966).
20. Mangasarian, 0. L., Mgt. Sci., 10, 353-359 (Jan.,
1964).
21. Mangasarian, 0. L., and Rosen, J. B. J. Opns. Res.
Soc. Am., 12, 143-154, (Jan.-Feb., 1964).
22. Mugele, R. A., "The Probe and Edge Theorems for
Non-Linear Optimization," in Lavi, A. and T. P.
Vogl, Ref. 19 above, pp. 131-144.
23. Naylor, T. H., J. L. Belintfy, D. S. Burdick and Kong
Chu, "Computer Simulation Techniques," xiii + 352
pp., J. Wiley, N. Y. (1966).
24. Ornea, J. C. and G. G. Eldredge, "Nonlinear Parti-
tioned Models for Plant Scheduling and Economic
Evaluation," Paper 4.15, AIChE/I ChemE Joint
Mtg, London, June, 1965.
25. Rosen, J. B., J. Soc. Ind. App. Math, 8, 181-217
(1960) 9, 514-532 (1961).
26. Rosen, J. B., Num. Math., 6, 250-260 (1964).
27. Rosen, J. B. and J. C. Ornea, Mgt. Sci., 10, 160-173
(Oct. 1963).
28. Rudd, D. and C. C. Watson, "Strategy in Process
Engineering," Preliminary Edit., J. Wiley, (1966).
29. Singer, E., Chem. Eng. Prog. Symp. Series No. 37,
58, 62-74 (1962).
30. Souders, Mott, Chem. Eng. Prog. 62, No. 3, 79-81
(March 1966).
31. Wilde, D. J., "Optimum Seeking Methods," 202 pp.,
Prentice-Hall, Englewood Cliffs, N. J. (1964).
32. Wilde, D. J., Ind. Eng. Chem., 57, No. 8, 18-31 (Aug.
1965).
33. Williams, T. J., and R. E. Otto, AIEE Trans 79,
(Comm and Elect.), 458-473 (Nov. 1960).


CHEMICAL ENGINEERING EDUCATION






The world of Union Oil

salutes the world

of chemical engineering


We at Union Oil are particularly indebted to the colleges
and universities which educate chemical engineers.
Because their graduates are the scientists who contribute
immeasurably to the position Union enjoys today:
The thirtieth largest manufacturing company in
the United States, with operations throughout
the world.
Union today explores for and produces oil and natural gas
in such distant places as the Persian Gulf and Alaska's
Cook Inlet. We market petroleum products and petro-
chemicals throughout the free world.
Our research scientists are constantly discovering new
ways to do things better. In fact, we have been granted
more than 2,700 U.S. patents.
We and our many subsidiaries are engaged in such
diverse projects as developing new refining processes,
developing new fertilizers to increase the food yield, and
the conservation of air and water.
Today, Union Oil's growth is dynamic.
Tomorrow will be even more stimulating.
Thanks largely to people who join us from leading
institutions of learning.
If you enjoy working in an atmosphere of imagination and
challenge, why not look into the world of Union Oil?
Growth...with innovation. Union Oil Company of California.



union











CHEMICAL ENGINEERING DIVISION ACTIVITIES


Seventh Annual Lectureship
Award to C. J. Pings

The 1969 ASEE Chemical Engineering Divi-
sion Lecturer is Dr. C. J. Pings of the California
Institute of Technology. The purpose of this
award lecture is to recognize and encourage out-
standing achievement in an important field of
fundamental chemical engineering theory or
practice. The 3M Company provides the financial
support for this annual lecture award.
Bestowed annually upon a distinguished engi-
neering educator who delivers the Annual Lecture
of the Chemical Engineering Division, the award
consists of $1,000 and an engraved certificate.
These were presented to this year's Lecturer, Dr.
C. J. Pings, at the Annual Chemical Engineering
Division Banquet held June 24, 1969 at the Penn-
sylvania State University. Dr. Pings spoke on
"A Chemical Engineer Looks at the Physics of
Simple Liquids." A paper based upon his lecture
will be published in an early issue of Chemical
Engineering Education.


PREVIOUS LECTURES

1963, A. B. Metzner, University of Delaware,
"Non-Newtonian fluids."
1964, C. R. Wilke, University of California,
"Mass transfer in turbulent flow." -
1965, Leon Lapidus, Princeton University, "As-
pects of modern control theory and applica-
tion."
1966, Octave Levenspiel, Illinois Institute of
Technology, "Changing Attitudes to Reactor
Design."
1967. Andreas Acrivos, Stanford University,
"Matched Asympototic Expansions."
1968, L. E. Scriven, University of Minnesota,
"Flow and Transfer at Fluid Interfaces."


BIOGRAPHIC SKETCH


Cornelius J. Pings was
born in Montana in 1929 and
entered the California In-
stitute of Technology in
1947 from which he received
a BS degree in Applied
Chemistry in 1951, an MS
degree in Chemical Engi-
neering in 1952, and a PhD
degree in Chemical Engi-
neering in 1955. He served
on the faculty at Stanford
University from 1955 to
1959 before returning to
Caltech where he is now
Professor of Chemical Engineering.
In research and scholarly activities his interests have
centered about the areas of applied chemical thermody-
namics and the physics and chemistry of liquids. In the
area of thermodynamics his work has led to important
improvements in the methods for quantitatively describ-
ing the displacement of chemical equilibria. His research
in liquid state physics and chemistry, which has been both
theoretical and experimental, has the long-range objec-
tive of fundamental elucidation of the liquid state. He
has developed one of the most extensive and best equipped
laboratories in the United States for the fundamental
study of fluids. Recent advances, deal with sound absorp-
tion at critical states, the structure of liquid argon, and
studies of intermolecular forces.
In addition to his research and teaching activities, he
has been active in faculty government and in student
affairs. He has served for the past year as chairman of a
special faculty committee on the aims and goals of Cal-
tech. Also active in civic affairs, Dr. Pings was named in
1968 to the Community Redevelopment Agency of the
City of Pasadena. He has served as a consultant to a
number of industrial firms and to the Department of
Defense.
Professor Pings served as Visiting Professor of Chemi-
cal Engineering at the University of Brazil in 1963. He
has received two Presentation Awards from the AIChE-
one from the 56th National Meeting in Houston (1963)
and the other from the 56th Annual Meeting in San Fran-
cisco (1965). He is editor of the Journal of Physics and
Chemistry of Liquids, and will serve in August 1969, as
Chairman of the Gordon Research Conference on the
Chemistry and Physics of Liquids.


CHEMICAL ENGINEERING EDUCATION



















ar










a a






























We hav im edat 00. ing 0o 0H MIA 0.MIA EN IE R 0a0 RJ deeo exi0s tin





Prouc 00elpm n Deatet an Aor An ne prouc lie navreyo ils



Co pn' expandig diesfcto proram SN CK in th foo lin an AIST N A ALAM,


*ep n ib t for for0.lS t ig de eopn a.0 m AI T R in th to a c Aie
.jating~~ Ardut fo xsig-sw l sptnilTecalne h rwhadteftr r n
-na t *.. 3.h ara offos3. id5akgn iied fr C E IA NGN E S a _J
)roducts~~~~ ~ ~ ~ ~ ~ and toac.Ryod oac *opnWntnSlm .C







Right now,Westvaco engineers
are revolutionizing paper for
printing, reprographics, computers,
clothing, packaging, disposables,
structures.
And chemicals for coatings,
pharmaceuticals, inks, tires,
soaps, and waxes. And adsorbents
for environmental controL
Your job:
Our next revolution.
See our campus representative. Or write to:
Roger Keehn, Westvaco Building, 299 Park Avenue, New York, N.Y. 10017


I West Virginia
Pulp and Paper
An Equal Opportunity Employer




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E1KTI062P_V1LPMD INGEST_TIME 2011-08-09T20:44:39Z PACKAGE AA00000383_00025
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EYH015OUX_N0MRSQ INGEST_TIME 2011-11-02T15:04:22Z PACKAGE AA00000383_00025
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

z Q u => Cl w c., z 0:: w w z c., z w 0:: 0 LL .... w 0 0 V) z < u ii w u. 0 z 0 V) Cl (!) z a:: w w z (!) z w .....I < u w ::c u SU MMER 1 969 CHE EDUCATOR OPTIMIZATION in practice .... .... hughes KINETICS: an auto gra d ed course . .. shilling DILEMMA : innovating societies fredrickson TRANSPORT : equation derivation ......... lee C H E L AB : microcatalytic experimen t at ri c e u n iv. neumann, riffle, swenson, hightower REPOR T: t o our readers ........... editors THE GA TORS GO : .............. florida .,(/ho, BIRD WILLIAMS ~ SCHMIDT PFEFFER COHEN CHITTENDEN

PAGE 2

Essa Research and Engineering Company, the principal technical affiliate of Standard Oil Company (N J ), provides research and en gineering services to 250 world wide affiliates with assets of over thirteen billion dollars. The Chemical Engineer plays a vital role in helping us meet these vast responsibilities But most important to him, he functions in an environment as dedicated as that of the university Chemical Engineering department For our CHEMICAL ENGINEERS GET TOTALLY INVOLVED IN A TOTAL ENGINEERING ENVIRONMENT ultimate goal is the same as that of the university; namely the ex tension of knowledge and the bet terment of the human condition through long term fundamental and applied research, and the accomplishment of immediate ob jectives through the economical design and operation of plants and equipment. Whether he possesses a B.S an M S., or a PhD., and whether he works in Product / Process Re search and Development, Appli AT ESSO cations and Technical Services, Process Engineering, Project De sign or Process Selection and Economics, the Chemical Engi neer serves with his professional peers He learns from them; he teaches them But he advances as far as his own talents take him, wherever his interests lead him; either in a technical or ad ministrative capacity. Total involvement ... in a total chemical engineering environ ment. That's Essa. For full details on the opportunities available, contact: Dr P. H Watkins, Employment Coordinator, Dept. ESSO RESEARCH AND ENGINEERING COMPANY P.O BOX 175, Linden, New Jersey 07036 An Equal Opportunity Employer (M / F)

PAGE 3

EDITORIAL AND BUSINESS ADDRESS Department of Chemical Engineering University of Florida Gainesville, Florida 32601 Editor: Ray Fahien Associate Editor : Mack Tyner Business Manager: R. B. Bennett Publications Board and Regional Advertising Representatives: CENTRAL: James Weber Chairman of Publication Board University of Nebraska Lincoln, Nebraska 68508 WEST: William H. Corcoran California Institute of Technology Pasadena, California 91109 SOUTH: Charles Littlejohn Clemson University Clemson, South Carolina 29631 SOUTHWEST: J. R. Crump University of Houston Houston, Texas 77004 EAST: Robert Matteson College Relations Sun Oil Company Philadelphia, Pennsylvania 19100 E. P. Bartkus Secretary's Department E. I. du Pont de Nemours Wilmington, Delaware 19898 Peter Lederman Brooklyn Polytechnic Institute Brooklyn, New York 11201 NORTHEAST: James M. Douglas University of Massachusetts Amherst, Massachusetts 01002 NORTH: J. J. Martin University of Michigan Ann Arbor, Michigan 48104 NORTHWEST: R. W. Moulton University of Washington Seattle, Washington 98105 UNIVERSITY REPRESENTATIVE J. A. Bergantz State University of New York Buffalo, New York 14200 PUBLISHERS REPRESENTATIVE D. R. Coughanowr Drexel University Philadelphia, Pennsylvania SUMMER 1969 Chemical Engineering Education VOLUME 3, NUMBER 3 Departments 107 Report to our Readers 108 Letters 110 The Educator Dean Max Peters 124 Views and Opinions SUMMER 1969 The Dilemma of Innovating Societies, A. G. Frederickson 126 The Classroom Transport Phenomena Equations of Change V. J. Lee 118 The Laboratory A Microcatalytic Tracer Experiment, R. W. Neumann, S. E. Riffle, S. T. Swenson, and J. W. Hightower. 128 Book Review 142 Problems for Teachers 150 Departments of Chemical Engineering University of Florida, Ray Fahien 160 Division Activities 154 The Curriculum Flexible Curricula Can Be Strong, Ray Fahien, Mack Tyner, R. A. Keppel. Feature Articles 113 Optimization Applications and Limitations, R.R. Hughes 130 A Self-pacing, Auto-graded Course, G.D. Shilling. CHEMICAL ENGINEERING EDUCATION is published quarterly by the Chemical Engineering Division, American Society for Engineering Educatfon. The publication is edited at the Chemical Engineering Department, University of Florida. Second-class postage is paid at Gainesville, Florida, and at DeLand, 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 seat directly to the printer: E. 0 Painter Printing Co., 137 E. Wisconsin Ave., DeLand, Florida 32720. Subscription rate U.S ., Canada, and Mexico is $10 per year to non-members of the ChE division of ASEE and $6 per year to member s Individual copies of Vol. 2 and 3 are $3 each. 105

PAGE 4

Olin PRODUCT LOCATIONS HAVING MAJOR PRODUCT S DISCIPLINE T Y PE OF WORK GROUP CURRENT OPENINGS PRODU C ED REQUIREMENTS PERFORMED Chier Alkali Products Ammon i a Process Development Augusta, Ga Phosphates Design, Maintenance, Brandenburg, Ky Urea Planning, Scheduling, Charleston, Tenn Nitrogen ChE Joliet, Ill. Ac i ds ME Production, Sales CHEMICALS Lake Charles, La Hydrazine IE Accounting, -Inorganic Little Rock, Ark Petrochemicals Chemistry Marketing -Organ i c & McIntosh, Ala Insecticides Accounting Financial Analysis, Specialty New Haven, Conn Pesticides Business Adm Distribution -Agricultural Niagara Falls N.Y Polyurethane Transportation Project Eng i neering Pasadena, Texas Carbon Dioxide Marketing (Plant Startup & Rochester, N.Y Animal Health Construction), Saltville, Va. Products Research Engineering, Automotive Chemicals Technical Service Other derivatives Alumina ChE Burns i de La Aluminum IE Manufacturing METALS Chattanooga, Tenn Aluminum Extrusions ME -Aluminum Gulfport, Miss. Alum!num Sheet, Plate, Metallurgy Production -Brass Hannibal, Ohio Coils Met. Engineering Sales -Ormet Corp East Alton, Ill. Brass Fabricated Parts Accounting Maintenance New Haven, Conn Sheet & Strip Brass Bus i ness Adm. F i nance Sedalia, Mo. Roll Bond Ind Tech Metals R&D Wire & Cable Ind. Mgmt. Carbonizing Paper ChE Marketing F ine Printing Papers FOREST PRODS, West Monroe la. Specialty Paper Chemistry Process Engineering PAPER &FILM Pisgah Forest, N C Products Pulp & Paper Plant Eng i neering -Olinkraft, Inc. Cigarette Paper & Tech. Research & Dev. -Ecusta Covington, Indiana IE Statistician Filters Systems Engineering -Film Cellophane ME Mathematics Production Kraft Bags Management Kraft Paper Business Adm General IE Kraftboard Cartons Accounting Design and Corrugated Containers Development Olinkraft Lumber Accounting WINCHESTEREast Alton, Ill. Sporting Arms Ind. Tech Produ c t i on Control WESTERN New Haven, Conn. Ammunition IE Purcha s ing Marion, Ill. Powder Actuated tools ME Manufacturing Kingsbury, Ind. Smokeless Ball Mathematics Plant Engineering Powders ChE S ales Solid Propellants Accounting Financial Analysis Safety Flares Business Adm Personnel Franchised Clubs Marketing Marketing Personnel Mgt. R&D Physics Ind. Mgmt. If you find this chart interesting, we're interested. F or additional information ab0ut Olin, please contac t your Placement Office or write Mr. Monte H. Jacoby, College Relations Officer Olin, 460 Park Avenue, New York, N Y 10022. Olin is a Plan for Progress comrany and an equal opportunity employer (M & FJ.

PAGE 5

REPORT to our Readers Chemical Engineering Education has now published six issues and completed one and a half years of publication at the University of Florida. It is both surprising and gratifying to the editors that the publication continues to be well received by the profession. Although we have expected our luck to run out, each issue seems to bring forth new commendatory letters and comments and there seems to be general agreement that the publication is filling an important need. We very much appreciate these expressions of support, but we realize that we are still amateurs in the publishing business. We cannot compete with commercial ventures or professional journals that are published by large societies and supported by dues. During the past year we have again had an excellent response from our colleagues who have submitted manuscripts for publication. We have not been able to accept all papers submitted to us (rejected papers have usually been negatively re viewed, but a few have been deemed inappropri ate for our journal.) Our editorial policy in each issue has been either to emphasize some particu lar theme or to try to achieve a balance among articles. The latter means not only that we try to have an article for every recurring department (e.g. Chemical Engineering Classroom), but also that we try to have represented in the issue the various technical areas of the profession. We also try to achieve balance among articles published in terms of type of school (public, private, gradu ate-oriented, undergraduate-oriented, etc ) and geographical location While we have appreciated the opportunity to publish papers from some well known people in the profession, we do not in any way discriminate against lesser-known but cap able people from the undergraduate-oriented schools. We feel our journal is for the profession as a whole and not just an elite group or "in group" of any kind. The above policies also apply to the selection of our featured departments and featured educa tors. Here we have generally had an excellent response from people whom we have asked to write articles. However, there were two schools who were unable to submit a department article for this issue. As a result we have been forced to get together a last minute article on our own de partment-even though we have assiduously tried to avoid using articles from our own deSUMMER 1969 partment in the journal. (Apparently we have been reasonably successful in not playing up our department, since a good number of visitors have commented on the journal without knowing we were publishing it!) In order to survive it is necessary that we receive the financial support of departments, ad vertisers, and industrial donors. We have appre ciated the fact that, due to the efforts of Joe Bergantz, nearly 100 departments are now con tributing to CEE However, last fall, the pros pects for advertisements and donations seemed so gloomy, that we reduced our number of pages by eight. Since then, I am happy to report, the yeoman efforts of Professor Weber and the Pub lication Board have been paying off. It now looks like our combined industrial advertising and do nation income will be within a few hundred dollars of last year's figure and could even be more than last year! Even brighter, however, are the prospects for the success of George Bur net's suggestion that the fall issue go to seniors interested in graduate work and that we seek departmental ads on graduate programs for the fall issue. Professor Bergantz states that now about 30 departments have indicated that they intend to buy a total of $3000 of such advertising of which the increased costs of printing additional pages and 2000 more copies will be about half that amount. Incidentally, since the Fall issue of CEE will carry paid advertisements from ChE depart ments and will go to seniors interested in gradu ate work, we felt the editorial content should e:n phasize basic areas of instruction and research in graduate studies. Accord i ngly prominent chem ical engineering scholars will write on their grad uate courses in certain important areas (such as control and optimization, thermodynamics, kinet ics, applied mathematics, particular systems, etc.). Each article will outline a graduate course and indicate important areas of research in the field. In this manner we hope to encourage more seniors to continue their intellectual growth in graduate school. R.W.F. P.S The Publications Board recently approved, effective January 1970, a charge to ChE Division members of $6 / yr. each for subscriptions Bulk subscriptions to departments will be $4 / yr. each with a minimum charge of $25. We expect to reach more faculty members this way and also to generate additional income. 107

PAGE 6

NOTES TO AUTHORS: P.S. We have a few words of instruction for fu ture authors. Because of the limited amount of space available, the articles, papers or reports submitted to CEE should be concise, lucid and also brief. Follow nomenclature of standard text books or write equations or formulas clearly. Use consistent units of measurement and give dimen sions for all terms. Assume your reader has some expertise in the field and minimize the amount of historical back ground included. Avoid tables and graphs which involve duplication or unnecessary data. Fre quently a graph or a few typical results may be substituted for a lengthy table. Two copies of a paper are sufficient for review. After a paper has been accepted for publication, the author should send the editor a short biog raphy and photo of himself to use with the article. M.T. ACKNOWLEDGMENTS C. F. BRAUN AND COMPANY DOW CHEMICAL COMPANY MALLINCKRODT CHEMICAL COMPANY MONSANTO COMPANY MINNESOTA MINING AND MANUFACTURING COMPANY OLIN MATHIESON CHEMICAL COMPANY THE PROCTER AND GAMBLE COMPANY STANDARD OIL (INDIANA) FOUNDATION THE STAUFFER CHEMICAL COMPANY from the READERS Correction from Lih Sir: The beginning of the article on Stu Churchill (CEE Spring 1969) clearly illustrates what the chemical engi neering teacher has to do these days. You have to stand on your head to catch the attention of students. Perhaps this is why Professor Churchill has been so successful and has had to engage in all sorts of athletic activities to keep it up The Japanese (and Chinest as well) character for HYO (leopard) is upside down. Statistical Study Marshall M. Lih Catholic University Sir: We have made a study which attempts to relate mathematically the number of staff members of profe s sional rank requi r ed in a chemical engineering depart 108 ment to the numbers of bachelors, masters and doctors graduated per year. No similar study could be found in the literature. One purpose of this study was to analyze the relation ship between the number of persons of professorial rank required in a chemical engineering program and the num b :'! r of students to be graduated per year at various degree levels. In light of the rapid expansion currently taking place in most universities it is important that this rela tionship be understood in order that intelligent admin istrative and educational decisions and forecasts may be attempted In the present study this relationship is analyzed only for chemical engineering programs since this ca s e was of immediate concern to the autho r s The main problem involves estimating how much of the total variability in the number of professors of chemical en gineering from university to university is due to the different numbers of degrees granted, and how much is due to "other factors" such as Different emphasis on research activitie s Different policies concerning the amount of admin istrative work to be performed by the professors CHEMICAL ENGINEERING EDUCATION

PAGE 7

Different professor-student ratios Different student attrition or "drop-out" rates Different amounts of teaching done by non-profes sional staff members Different numbers of students per class Henceforth, these "other factors" will be referred to as educationa l, administrative and research policies. If it is assumed that the departments of chemical engineering in our American colleges and universities have the same educational, administrative and research policies, the following equation will hold. P = A 0 + A 1 B + A 2 M + A 3 D (1) where P is the number of full-time teaching schedules required in the professorial ranks (total of Assistant, Associate and Full), B, M, and D are, respectively, the number of bachelor's, master's and doctoral degrees granted per year, and A 1 A 2 and A 0 represent respec tively the additional numbers of professors needed for each additional bachelor's master's or doctoral degree granted per year. However, as would be expected, deviations from Eq. 1 are observed, presumably because universities do not have the same educational, administrative and research policies and are at considerably different stages of devel opment. In order to estimate the proportion of the total variation due to the different numbers of degrees granted and how much is due to the difference in educational, administrative and research policies, the technique of least squares was used. Data from accredited chemical engineering depart ments of 97 universities for the academic year 1964-1965 were gathered from Chemical Eng ineerin g Faculti es of Canada and the United States, which is compiled annu ally by the Chemical Engineering Projects Committee of AIChE For each of the 97 universities, the number of professorial schedules in chemical engineering was ob tained along with the numb er of bachelor's, master's and doctor's degrees granted. By calculations performed on an IBM 7040 digital computer, the least squares estimates of A 0 Al' A 21 and A 3 were found to be 3 .40 0.06, 0.15 and 0.43 respectively, yielding the equation P = 3.40 + 0.06B + 0.15M + 0.43D (2) Thus we can estimate that for every additional pro fessor employed, schools can grant, on the average, an additional 16-2 / 3 bachelor's degrees, or 6-2/ 3 masters degrees or 2-1/3 doctors degrees per year and conversely. It should be emphasized that these figures are estimates of average behavior and that the situation in any one school may depart markedly from these estimated averages. The multiple correlation coefficient was found to b e 0.79. Accordingly, one may estimate that 0.62 (which is 0.79 2 ) of the total variation in the number of professorial schedules in chemica l engineering from university to uni versity may be exp lained by the different numbers of de grees granted while the residual 0.38 (which is 1.00-0.62) of the variation may be attributed to the different edu cational, administrative and research policies prevailing. A second objective of the pres e nt study was to deter mine changes with time in the coefficients such as those calculated and presented in Eq. 2 in the light of the very rapid educational growth rates and rapid changes in student degree objectives currently occurring on the SUMMER 1969 American educational scene. Data for 70 departments of chemical engineering for the year 1962-63 had been analyzed and reported in a previous study.* The least squares equation for these data had been found to be P = 2.2 + 0.10B + 0.lM + 0.45D (3) It is interesting to speculate on the results of the two studies. A multiple correlation coefficient of 0.83 obtained from the 1962-63 data against the value of 0.79 in the current study indicates a reasonable degree of year-to year stability despite the relatively rapid changes cur rently occuring in Academia. The large change from 2.2 to 3.4 in the constant terms of Eqs. 3 and 2, may indicate that in the 1964-65 period, considerably more professorial time was allotted to administrative and other non-teaching activities than in the 1962-63 period. We come next to the coefficients of B. If we assume a Gaussian distribution, the difference between the coeffi cients of B in equations 2 and 3 does not (at the 10 % level) turn out to be statistically significant However, since there is no a priori basis for assuming a Gaussian distribution, the observed difference may be significant It is interesting to note that if the difference is signifi cant in fact, then it is in the expected direction; i.e., downward, indicating that fewer professors were required per bachelor degree granted during the 1964-65 period than were required during the 1962-63 period. The trend towards larger lecture sections requiring fewer professors per student as well as an increase in instruction by graduate assistants has certainly been observed during recent years. Finally, with regard to the coefficients of the M and D terms in Eqs. 2 and 3 one can only say that the agreement is indeed striking and beyond what one might expect. To determine to what degree the above conjectures are correct would of course require direct corroboration in depth from the institutions involved. At any rate, the implied stability of the results of these studies is gratifying. It indicates that similar analyses in the future may be desirable and useful. Because of radical differences in the procedures, policies and manpower requirements of graduate and undergraduate programs, it was decided, as a third ob qective of this study, to try to analyze separately a group of schools heavily oriented toward graduate work as distinct from a group that is not. To distinguish between them the following purely arbitrary criteria were set up. For each institution the ratio of the combined number of master's and doctoral degrees (in ChE) to the number of undergraduate degrees was determined. If the ratio was greater than 0.4, or if there were five or more doc toral degrees granted in ChE for the year the school was considered to be graduate-inclined. On this basis the 97 schools subdivided into 40 graduate-inclined and 57 undergraduate-oriented schools. In both groups, a least squares analysis was per formerd as in the case for all 97 schools with the follow ing results: Schmidt, A. X. and Pfeffer, R., CEE, p. 13 (October 1965). (Continued on page 143.) 109

PAGE 8

Ci n I educator MAX PETERS University of Colorado If there is a better way of educating engi neers, Max S. Peters, dean of the College of En gineering at the University of Colorado, will be in pursuit of it. "Finding a better way" could be the most fitting description characterizing his performance, whether inside or outside the class room, the administrative halls of higher educa tion, the laboratory, or the smoke-filled commit tee rooms in which he is such a driving force. Something of the vigor and hardihood of the Ohio-Pennsylvania early American certainly is evident in Max Peters. He was born in 1920 in Delaware, Ohio, and received his early education in State College, Pennsylvania, and at Penn State University. He earned the PhD degree in chemi cal engineering at Penn State in 1951, perform ing research on vacuum distillation. By the time he was awarded his degree he had already worked as a production supervisor in a wartime powder plant and had served with distinction in the 10th Mountain Division of Italy. For two postwar years he was in charge of all technical work for Treyz Chemicals in Cook Falls, New York. Colleagues and students alike are apt to ad dress him as Max. They appreciate his capacity for hard work and the sense of humor that is ready to break to the surface in the midst of serious considerations. His door is always open to his graduate students. As one of them puts it, "I'm amazed that a man with so many responsi bilities can be so available." The responsibilities he has are the result of his diligent research for "a better way." He heads a college enrolling nearly 2000 under graduate and some 300 graduate students, with a faculty numbering around 125, housed in a new $8.5 million educational facility which he him self worked hard to make a reality. In addition, he guides engineering programs at two off campus CU Centers, at Denver and Colorado Springs. The story of the remarkable growth of the College of Engineering at the University of Colo110 Max Peters, newly appointed in 1962 as dean of the College of Engineering at the University of Colorado, enthusiastically pushed forward plans for the proposed Engineering Center, dedicated in May, 1966. rado since Max Peters arrived on campus in July of 1962 is well known to engineering educators across the country. He hadn't been in office long before the building priority for the proposed Engineering Center moved up from eleventh to first. Then he went to work with the University administration to acquire $7.2 million from the state legislature for construction of the Engi neering Center the largest sum ever requested from the people of Colorado. He further exerted his persuasive and organizing powers to acquire a supplementary $1.325 million from the Na tional Science Foundation. The Engineering Center at CU can truthfully be said to have been inspired in great part by the enthusiasm and imagination of Max Peters. At ground-breaking ceremonies for the Center in 1964 he revealed the scope of his expectations when he described the Center as "a major mile stone . in the forward progress of making the State a major industrial and scientific center." He tackles the challenge at its grassroots the high school level with the same bold en thusiasm. "Is there anything wrong," he asks, "with being old fashioned and strongly encour aging our high school students to start preparing themselves while in high school" CHEMICAL ENGINEERING EDUCATION

PAGE 9

"To be what you can be, you must first and foremost decide what you want to be." Many of his colleagues and students have heard him say, "Engineering is a tremendously exciting and rewarding career!" and in a di versity of situations he proclaims vehemently, "It is time that some of us decide to speak out." Max showed signs of "speaking out" in his first academic position, as assistant professor, then professor, and then divisional head of chem ical engineering at the University of Illinois. There he began the characteristic pattern: to examine and re-examine the curriculum to expose its weaknesses, identify its strength, and take action to improve it. At Illinois he recognized the need for more extensive chemical engineering kinetics study and introduced a course to fill the void. His first book, El ementary Chemical Engi neering, (McGraw Hill, 1954) was written at Illinois to fill a gap in engineering education for students of other disciplines such as mechanical engineering and chemistry. The text has been especially valuable in foreign countries where teachers were not capable of using standard texts for chemical engineer majors. In his second book, Plant D esi gn and Econom ics for Chemical Engin eers (McGraw Hill, 1957) Max Peters tackled another need that of grad uate students who went into design work in the chemical industry with relatively little back ground in plant design problems and their solu tions. Plant Design has been adopted by more than half of the chemical engineering curricula across the country. A completely revised edition (1968), written with Klaus D. Timmerhaus, as sociate dean of the CU College of Engineering, provides greater depth in optimization and eco nomic evaluation. Both books are considered classics in that they speak to engineers of all disciplines and to scientists and industrial managers who have no formal educational background in chemical engineering. It was at Illinois that Dean Peters' drive to improve educational standards carried him into the local chapter of the AIChE as faculty repre sentative. Years of committee activity brought him in 1968 to the presidency, where he com mitted himself wholeheartedly to the quest for improved engineering education and true profes sionalism by AIChE members, and to the proSUMMER 1969 When Max Peters takes part in Fun & Games at the E-Days picnic only his plaid shirt distinguishes him from the students. fession's obligation to assist society in the solu tion of its problems. His improvements in the professional society, (among them he originated the popular Free Forums) are well known to AIChE members. Through the voice of AIChE Max has supported the concept of the chemical engineering degree instead of the general engi neering degree as the first professional degree. As head of the chemical engineering division at Illinois, Max Peters recognized that chemical engineering students engage in a wide scope of activities including economics, technical services, laboratory research. Accordingly, he introduced a flexibility into the undergraduate curriculum that allowed the student to substitute advanced mathematics, physics, and chemistry for more conventional courses. His incorporation of fundamental engineering sciences into the under graduate program has been copied by chemical engineering departments in many institutions throughout the United States It was natural for Max to strive toward rais ing the standards of excellence in the College of Engineering at the University of Colorado. With Dean W. L. Everitt of Illinois he initiated the Bi-University Institutional Liaison for Develop ment (BUILD) program for experimentation and development of faculty innnovative ideas between the two universities. Now concluding its fourth year of support by the Kettering Foundation, BUILD has implemented exchanges for profes sional development that have involved at one time or another every faculty member of the CU College of Engineering. Max is a forceful and articulate committee member, as many of his colleagues have learned. As chairman of the CU proposal committee for the National Science Foundation Scientific De velopment program, he helped bring $3.75 million 111

PAGE 10

to the University of Colorado, one million of which want into electrical, aerospace, and me chanical engineering programs in the College of Engineering. As a result, the College has strengthened its programs in control theory, solid state physics, computer logic, fluid mechan ics, applied mechanics and mathematics. Growth in these areas is being watched with interest by other institutions Behind this vigorous activity Max holds a philosophy that has deeply affected graduate study and research throughout the College. Grad uate students, he holds, should be actively in volved in research programs under the direction of faculty members. Faculty members who are engaged to fill needs in research areas must be good teachers. Funding for this program has increased during the past five years from less than $200,000 to approximately two million dollars. In 1961, one PhD degree was awarded by the University of Colorado to a student in engineer ing. In 1968, 30 engineering students earned the degree. PhD degrees have been made available in aerospace engineering sciences and in mechan"Is there anything wrong," he asks, "with being old fashioned and strongly encouraging our high school students to start preparing themselves while in high school?" ics in addition to all the other departments, with the exception of a new department, Engineering Design and Economic Evaluation, which offers the master's degree. It's easy for graduate students to discuss their research problems with Max Peters; in fact, he is himself making a significant contribution in the laboratory. His research studies in kinet ic s, particularly on nitrogen oxides and pentaery thritol, have resulted in increased understanding of reaction mechanisms and in the chemical en gineer's ability to design reactors. He personally directs graduate work in allied studies. He gives his students as much freedom as possible in their projects, only outlining the over all view and the goal. He believes that students learn more from doing a thing wrong than from doing it right the first time. This freedom to experiment encourages his students to be creative and analytical. They recognize his qualities as a teacher. "He took the complex and broke it down into simple integral parts," one of his students comments, 112 As usual, Max Peters won the Dean's Challenge Rac e a t the E-Days picnic in May, 1968 or, "He built up the complicated theories of chemical engineering by starting with easily un derstood building blocks of knowledge." As a teacher, characteristically Max looks for better ways to explain points and ideas; he designs quizzes and exams to test a student's com prehension rather than his memory of equations and data. The secret of his ability is simple: he is truly interested in each of his students as a person. Because of this, he is able to instill in his students the desire to be successful in study and experimental work. Typically, his plant de sign students have repeatedly won or placed near the top in the national AIChE Student Contest Problems. Students have discovered their Dean is a for midable contestant on the ski slopes and the En gineers' Days races They hail him as champion of faculty-student slalom race at the CU Winter Carnival, and know him as an accomplished figure-skater. Every year Dean Peters has won the Dean's Challenge Race at the E-Days picnic. He wears his more impressive honors with modesty. This spring he was elected a member of the National Academy of Engineering the highest professional distinction that can be con ferred upon an American engineer. He is cited this June by the American Association of Cost Engineers for his "continuing contributions to the field of cost engineering education." He was recently named chairman of the President's Com mittee on the National Medal of Science. In 1957 Max Peters received the George West inghouse A ward from the American Society for Engineering Education for outstanding teaching. He has been active in the ASEE for nearly ten years In 1962 he served as chairman of the CHEMICAL ENGINEERING EDUCATION

PAGE 11

The secret ~f his ability is simple-he is truly interested in each of his students as a person. Chemical Engineering Division and was for six years a member of the long range planning committee of ASEE. He is active on the air pollution committee of the U. S. Department of Health, Education, and Welfare, which he serves as consultant. He is consulting editor for the McGraw Hill Chemical Engineering Series, and is the author of many technical articles. OPTIMIZATION Anyone who knows Max knows him as the em bodiment of a belief he has expressed to high school students and to AIChE members: "To be what you can be you must first and foremost decide what you want to be." Since he will never be satisfied with things as they are, but must al ways seek new and better answers, it follows that Max Peters is not only dean, teacher, chemical engineer, researcher, and innovator. He is per haps first of all a student a student of educa tion. APPLICATIONS AND LIMITATIONS** RICHARD R. HUGHES Shell D eve lo pment Company E meryvi ll e, California THE OPTIMIZATION MODEL To begin with, optimization requires a formal description of the problem. The elements in volved and their relationship are indicated in Fig. 1. First, the problem must be isolated by a formal description of the "state-of-nature" and the problem premises. This is often the most difficult part of the problem A sound treatment requires an assessment of whether the solution will answer the question posed and whether all significant variable elements are included within State-of-Nature and Problem Premises Restrictions Legal, Physical, Economic, Political etc Figure I. Payoff Value for Desired Objective ,:, Present Address: Univ. of Wisconsin, Madison, Wi s ** Presented at the Los Angeles ASEE meeting June 19-22, 1968. SUMMER 1969 Optimization implies logical, even formal, decision making, i.e., the selection, for a set of decision variables, of the best attainable (and allowable) values for a designated objective. To successfully accomplish opti mization of practical non-trivial problems, two major requirements must be met. First, we must have access to computers (normally large digital computers), and we must be able to use them. This, of course, implies opti mization of a mathematical model describing the prob lem; the second requirement is that this model must be the simplest possible one for the job at hand. The calcu lation will be extremely repetitive; and any but the simplest possible model will require excessive computa tion and make it uneconomic to use optimization. In this paper, I first describe my concept of an opti mization model. Then I propose guidelines for formula tion and simplification of such models. Finally I offer a few remarks on limitations and complications of the opti mization approach. My comments are based on several years of study and practical application of optimization, -by myself and many colleagues, to problems in chemical engineering, process design, and operations re search. Most of the rules given are not hard and fast limitations but merely express my observations of diffi culties we have encountered the system. Obtaining an optimum scale of manu facture at a fixed sale price is absurd if the scale affects the sales price. With the state of nature established we then identify the decisions we are still free to make. As functions of these decisions, we describe the payoff value, and formulate the necessary restrictions which dictate limitations on the problem, legal, physical, economical, political, etc. These restrictions limit the freedom of action of our decisions, but there is usually some variability left. By optimizing, we take ad113

PAGE 12

vantage of this variability to obtain the best pos sible payoff value. When formulated as a mathematical model, the problem has the form indicated in Fig. 2. The state of nature and problem premises are Constraints, g Objective Function, cp \ I \ I \ I \ I \ I \ I \ I \ Variables, x / \ I \ I ' / \ I I ,._\._ -_l_ _____ ../_-, : Optimizing Algorithm I L ______________ J Figure 2. described by a vector of parameters, p, which in cludes prices, sales figures, coefficients of correla tions, estimated technical values, and all the many other numbers which must be used to quantify the problem. The payoff value is for mulated as an object function, cf,, whose value is determined by the decision variable vector, x, and the parameter vector, p. The various limita tions are described as a vector of constraint func tion~, g, each element of which must be non negative for an allowable or feasible solution. Once the problem is described in this format, we can optimize it by using an optimizing algorithm to adjust the variables, x, to obtain the best value of cf, within the limitations of the constraints. Mathematically, this can be written as follows: M;x{cf,(p,x) I g(p,x) o} = cf,*(p) = cf,(p,x*) (1) The first part of this equation is read: "find the maximum over the variable space x qf the func tion cf, of p and x subject to the non-negativity of the elements o{ the vector g, which are func tions of p and x." The resulting maximum or "optimum," cf,*, is a function of the parameter vector p, which describes the particular case which has been optimized. Corresponding to cf,* there are one or more points, x*, in variable space x, where cf, = cf,*. These values are usually the most important part of the result; they indicate the optimum or best choice of variables. Impli114 citly, it is clear that the optimum choice, x is a function of the parameter vector p. OPTIMIZATION ALGORITHMS For further information as to the details of the opti zation algorithms, ample literature exists. For elemen tary introduction, there is a good book by Baumol on Economic Theory and Op e rations Analysis 3 and a some what more mathematical treatment by Carr and Howe 5 In the limited but important field of linear programming, Dantzig's book 6 is the fundamental authority; it is sound and intelligible, but a little long. The book by Gass 1 0 on the other hand is quite elementary; for an intermediate level, that of Hadley 13 is probably best. Hadley's seco nd book 14 extends the treatment to non linear programming, with emphasis on systems of many dimensions For those interested in these more-mathe matically oriented problems, the co ll ect ion edited by Graves and Wolfe 11 is a good review of the state of the art in 1963. However, for many chemical engineering problems, Wilde's recent book on "Optimum Seeking Methods" a 1 is more directly applicable. Two recent col lections also supply useful hints on this latter type of problem: the CEP Symposium Series volume on "Opti mization Techniques",4 and the book edited by Lavi and Vogl. 19 A last general reference, we cite the recent re view by Wilde, 32 which includes 74 references of recent work. In much of our work, we have used optimum seeking methods as described in the later refer ences listed above. Our particular versions are described briefly by Singer 20 ; one of them (the Maze Method) is described in more detail by Mugele. 22 We have also found the MAP method of Griffith and Stewart 12 to be generally applica ble and quite powerful. Finally, for the problems to which they apply, Ro sen's methods, Gradient Projection 25 and Partition Programming 26 27 have been quite successful. With these and many more methods available, further extensive work on mathematical programming or optimization algorithms does not seem worthwhile for the engineer. However the method chosen must be a suitable one for the problem at hand and the computing equipment available; with large prob lem s, the performance of the algorithms depends strong ly on the particular configuration of the computer. OPTIMIZATION PROBLEMS Most optimization problems of interest to a chemical engineer can be fit into one of four categories: (1) process design, (2) operations scheduling, (3) process control, and ( 4) equip ment design. CHEMICAL ENGINEERING EDUCATION

PAGE 13

1. Process Design In many respects, this is the most important category. Problems of this type range from de tailed process design ( to chose the best possible configuration of many process details) to quick and rather generally formulated process evalua tion (to provide a general pattern of possible profitabilities of a proposed new process or product). To illustrate the features of process design optimization, we use the example shown in Fig. 3. Although this example was developed independently, it is quite similar to the one pre sented by Williams and Otto 33 and stud ied by DiBella and Stevens.7 Product cw C0 r-~~-,( / ~,_-~ ~ _,
PAGE 14

O p era ting Eco n om i cs 20 Ne t R e tu r n M ini mu m M $/c d R e a c t o r V e lo c it y"" Capacit y ~ p.. in H X "tl "' "' r,. 10 2 OL..__....._ __ .,__ _._ __ ...._ __, __ ....... __ ....__. 0 10 0 z oo R ecyc l e ( GPM) Figure 5. 300 400 function contours, we indicate various constraint s coming from the physical specifications of the equipment: the capacities of reactor, heat exchanger, and putnp, the mini mum velocity in the heat exchanger to keep the sludge suspended, and the minimum production rate required. Note that within the central, feasible region the problem is very nearly linear. The optimum point occurs at the upper-right hand corner, where both the reactor, and the heat exchanger/cooler are operating at capacity Th e near linearity of this example is illustrative of th e fact that, in scheduling, we very often can obtain an accept able description of the system in a linear programming formulation; this tremendously simplifies the optimiza tion itself. 3. Process Control In spite of their differences, these first two problems both involve only steady-state analysis. This is not true of the third type of problem, the control problem. Here we must account for some dynamic affects, even though it may be possible to omit consideration of short-term dynamic affects if we limit our model to the longer-period control problem. For example, our objectives often con cern only temperature and concentration dynam ics; then the model can neglect the effect of pres sure waves or liquid-level fluctuations, as long as the specified control instruments maintain aver age pressures and levels at the desired point, over the period of the temperature or concentration fluctuations. For our simple example, the flow sheet illustrating the control problem appears in Fig. 6. The variables are pretty much those that are used for the scheduling operation, but we must now incorporate, in the equations describ ing the model, time derivatives showing the dy namic effects. In general this means that a con trol model formulated with the same degree of technological complexity as a process design 116 model will be a much more complicated model involving more complex mathematics and more difficult optimization Fortunately for industry, the precise optimization of the control model is usually less important economically than optimi zation of the process design. Thus it is usually not desirable to do as technologically complete a job on the control model as one does with the process design model. One complication of the economics for a control model is that the objective function almost invariably involves an extension in time. We are interested in costs or profits expressed as an average over a long period, and we will invariably have means of evening out our uneven operations, by the use of storage tanks or the ability to delay deliery. This makes it extremely difficult to formulate a true economic objective in terms of an immediate control variable .... u (Continued on page 134.) r~ \ --5--oic:l------f 0 1-, p.. r-1 I I 0 I I I I I I I I I I I I I I I 0 I L ___ T_ 7 I I L ii) > 0 t/l Figure 6. CHEMICAL ENGINEERING EDUCATION

PAGE 15

They're already planning to mine oil instead of pump it. Isn't that the kind of company you'd like to work for? There are more than 100 billion barrels of potential new oil on the North American continent. But it will have to be dug-not pumped out of the ground. It's in the form of low-grade hydrocarbon solids. Yet, the world needs so much more AtlanticRichfieldC ompany making things happen with petroleum energy oil in years to come that Atlantic Richfield is already working on ways to extract it and get it moving. Projects like this take imagination and fresh viewpoints. The kind that come from young innovators like yourself. We need you-and your kind of ideas-to keep making great things happen. Talk to our interviewer when he's on your campus. Or write to: Mr. G. 0. Wheeler Manager Professional Recruitment, 717 Fifth Avenue New York N.Y. 10022

PAGE 16

ti Na laboratory , A MICROCA T AL YTIC TRACER EXPERIMENT RALPH W. NEUMANN* STEVE E. RIFFLEt STEPHEN T. SWENSON** JOE W. HIGHTOWER William Marsh Rice University Houston, Texas 77001 Because of the rapid development of new analytical techniques and the increasing demands on students' time through expanded curricula, it has become necessary to streamline laboratory experiments to include as many of these tech niques as possible in the shortest period of time. This report describes one of the experiments aimed at pursuing this goal in our senior chemi cal engineering laboratory and presents some results obtained by this year's students. A large fraction of all industrial reactions are catalytic, and one of the most active areas in industrial research concerns the development of more active and selective catalysts for specific reactions. Although many of the early techno logical advances which revolutionized the petro leum industry before World War II were the result of empirical observations, the significant advances by such men as Sabatier,1 Langmuir, 2 Taylor,3 lpatieff, 4 Emmett,5 and others have helped to change the application of catalysis from an art into a science. One of the standard catalytic activity tests in the petroleum industry involves the dealkyla tion of cumene (isopropylbenzene). This reaction seemed the logical choice for our studies over a standard silica-alumina cracking catalyst in a microcatalytic reactor for the following reasons: It is essentially a "clean" reaction, i.e., the only sig nificant products are propylene and benzene. There is little poisoning, which means the same catalyst can be used from day to day without reactivation. Reactant and product compounds are easily separated by GLC and are amendable to isotopic tracer investigations in a mass spectrometer. Cumene dealkylation has b 2 en used *Present Address: Fluor Corporation, Houston, Texas **Present Address: Enjay Chemical Co., Baytown, Texas tResearch technician. 118 Joe W. Hightower earned the PhD at Johns Hopkins University working under Prof. P. H. Emmett. He was a Postdoctoral Fellow at Queen's University, Belfast, Ire land and was a Research Fellow at Mellon Institute until he joined the staff of Rice University in 1967. Steve Riffle is a research technician and Ralph Neumann and Stephen Swenson were fifth year ME students at Rice University. as a test reaction for diffusion studies 6 and for investiga tions of active sites on zeolite catalysts. 7 Cumene, an intermediate in th e production of phenol and acetone, is an important commercial compound. Research involving this compound is currently under way at Rice. The entire experiment can be carried out in a reasonably short length of time. EXPERIMENT AL A microcatalytic reactor 8 9 involves combina tion of a flow reactor with a gas chromatograph, Fig. 1. A helium carrier gas stream flowed con tinuously at 10 psig and about 100 cc/min through the reference side of a standard Gow Mac thermal conductivity detector and then through a small packed catalyst bed containing a centered thermocouple well. Pulses of reactant could be injected by means of a 10 ,l hypodermic syringe through a rubber septum injection port A. The reactant was carried over the catalyst where it reacted, and the rea(!tion products were swept immediately into the analyzing column, a six-foot coil of 1/4," copper tubing packed with silicone oil on firebrick. The separated products passed through the sample side of the Gow-Mac CHEMICAL ENGINEERING EDUCATION

PAGE 17

HELIUM SUPPLY OVEN AT 105 C /===~r===-=-~--------, DC POWER SUPPLY WHEAT STONE BRIDGE I +-tt---..._' SILICONE ,COLUMN I 1 INJ INJ I A B I .... .J j I 10/301' FURNACE I t PRODUCT 'fllAP CATALYST LN2 TUBE 10 MV. RECORDE fig. I -Schematic diagram of microcatalytic reactor and product traps. detector where they caused an imbalance in a Wheatstone bridge which was recorded as a peak on a 10 mV strip chart recorder whose chart speed was 2 min / in. Calibration was effected for each compound by injection beyond the catalyst bed at injection port B. The injection ports, detector, and column were all enclosed in a tran site box whose temperature was maintained near 115 C by means of a Variac which supplied power to the heating elements. The separated products could be collected individually in a trap thermostated at -195 C (liquid nitrogen temperature) for subsequent analysis at low electron voltage in a CEC 21-104 medium resolution mass spectrometer. Helium was removed from the sample trap by evacua tion at 195 C. Pellets of commercial Houndry M-46 silica alumina (12.5 % alumina) were ground and col lected between standard 20-60 mesh sieves to give particles which varied from 250 to 800 mi crons in diameter. Half a gram of this material was loosely packed to a depth of one cm between glass wool plugs in the 1.5 cm OD Pyrex reactor. The catalyst surface area was 270 m 2 /g. Initial activation was accomplished by heat ing the catalyst in flowing 0 2 at 530 C for one hour to burn off carbonaceous residues, and the catalyst was then cooled in flowing helium to the 295-365 C reaction temperature range. Follow ing this pretreatment, the catalyst retained a reproducible activity level for days without fur ther reactivation. The temperature of the elec trical resistance furnace around the reactor was controlled simply by a Variac. SUMMER 1969 RESULTS The entire experiment was designed to cover three 3-hour laboratory periods. We have found it most effective when each group consisted of from three to six students. Our seniors were divided into five groups, with each group coming in on a different afternoon during the week to perform the same part of the experiment. This meant the whole experiment lasted three weeks. Period I, Introduction During the first period, the objectives, techniques, and mathematical analysis of the experiment were described Each student then practiced making benzene injections through injection port B until he obtained reproducible peaks on the GLC. Finally, each measured his peak areas with a planimeter until his measurements were reproducible. Period II, Activation Energy A typical microcatalytic reaction spectrum is shown in Fig. 2 Besides the undef:,: (!) w :,: "'
PAGE 18

----------------------------~ --sion was greater than 99 % i.e., the reaction was essen tially ir reversible The first order irreversible rate equa tion was transformed into one involving fractiona l con version, x, and integrated to give ln 1 1 X which can b e wr itt e n Aexp( E /R T)t (2) 1 E ln (ln l x ) = RT + ln At (3) Since the pre-expon e ntial factor A and the contact time t in the microcatalytic exper iment are assumed to be esse ntiall y invariant with temperature, the activation energy E can b e determined from the slop of a plot of ln (ln 1 ) versus 1 / T. Data from fiv e sets o.f 1 X experiments on five different days by 1 3 seniors are shown collectively in Fig. 3; a least squares fit gives an apparent activation energy of 14.3 0.9 kcal/mole. The sample size was 2,l c um e ne at all temperatures in the region of 295 to 365 C. 1-,.. X 8l o"' :.., 0 ln(ln[~]> 0 9 c,, g +-364 c +-298 C Fig. 3. Compilation of data from 13 seniors showing temperature dependence of cumene dealkylation over silica-alumina in a micro catalytic reactor The apparent Arrhenius activation energy is 14 3 kcal/mole. Period III Deuterium Isotopic Tracers Measurements 11 b y exc han ge with D 2 have shown that fresh l y activated s ilica-alumina contains about 4 x 10 2 0 H atoms/g. These atoms have acidic properties and may provide Bronsted active sites on which the dealkylation reaction occurs The purpose of this part of the exper im e nt is to demon strate participation of these atoms in several different 12 0 r 2 actions which may occur. Wh en 2 ,l pulses of benzene were passed over the catalyst at 340 C, there was apparently no che mical react ion as only the benzene peak was observed in the GLC spectrum Similarly, when perdeuterio benzene (C 6 DG) was injected, only one peak was observed. How ever, mass spectra l analysis of th at benzene peak showed that extensive excha n ge had occurred between the cata l yst's H atoms and the hyd roc arbon's D atoms. 1 2 Fig. 4 30 25 Cl) UJ 0 UJ a. Cl) 20 (.) a: 0 b Cl) 15 J: (.) ct UJ 110 z UJ u a:: UJ a. 5 0 ,--r d 0 d I d 2 d 3 d 4 d 5 d 6 ISOTOPIC S PEC IES Fig 4 Deuterium distribution in benzene after exchange of a pulse of C o D with H atoms on the catalyst s how s the re l a tiv e amounts of product b enze n e molecules wh i ch conta in ed from 0 to 6 D atoms A ll peaks were corrected for naturally occ u rri n g C 1 3 ; fragmentation involving lo ss of one or more H atoms was negligible under the l ow voltage mass spectrometer operating con ditions used. Seven more identical 2 ,l pulses of C 6 D 6 were then passed in succession every 10 minutes over the cat alyst. The products were trapped and analyzed mas s spectrally; the results are g i ven in Table I. The last column show in g the atoms exchanged/mo l ecule was calculated from the eq uation 6 Atoms Exchanged / Molecule = }; (6 i) d;f 100 i O (4) where di is the percent of molecules containing i deute rium atoms As the poo l of available H atoms on the surface became diluted with D atoms as a result of exchange with each successive pulse, the amount of measurable exchange decreased from pu l se to pulse (see Fig. 5). F rom the number of benzene mo l ecu l es injected in each 2 ,l pulse and t h e average number of at om s ex changed (or "titrated") per molecule it was possib l e to determine the total numb er of surface H atoms which were exchanged in all eight pulses. Such a cumulative CHEMICAL ENGINEERING EDUCATION

PAGE 19

3 00 2.70 2.40 IIJ .J 32.10 IIJ .J 0 :::E 1.80 C IIJ 1 50 <( ::c 0 X w 1.20 Ul :::E Ill 0 ti .90 60 30 0 .__ .......... _.._....__,_....1_...____.___,__.._ ............ ....1_...___,_.__, I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 PULSE NUMBER Fig. 5 Average number of hydrogen atoms exchanged/molecule during successive passage of several pulses of benzene C o D o and cumene over a silica-alumina catalyst in a microcatalyst reactor plot is shown in Fig. 6. A large fraction (about 75 % ) of the total H atoms originally present on the half gram sample underwent exchange during passage of these eight pulses of C 6 D 6 With the catalyst now in a partially deuterated state, seven pulses of cumene were passed in succession over it at 340 C, and the conversion was constant at about 50 % dealkylation. All three products were individually trapped and analyzed for pulse 9, but for the remaining pulses only the benzene peak was trapped and analyzed; the results are given in Table I. It is apparent from the results of pulse 9 that exchange was extensive in the undealkylated cumene as well as in the reaction products. In fact, to a rough approximation all the H atoms in all the hydrocarbons essent iall y equilib rated with the D atoms from the catalyst For this to have been strictly true, the benzene and propylene (each has 6 hydrogen atoms) should have had the same number of D atoms/ molecule, .and cumene (12 hydrogen atoms) should have had doube that amount using this assumption, and basing the calculation on the number of cumene molecules added and on the benzene mass spectral analysis, the cumulative number of D atoms recovered from the catalyst cou ld be determined. The atoms exchanged/molecu l e were calculated from the equation 6 Atoms Exchanged/Molecule I id 1 / 100 (5) i = 0 and the cumulative plot for pulses 9 through 15 is shown in Fig. 6. Most of the D atoms excha n ged into the cata lyst from the first eight C 6 D 6 pulses were recovered in SUMMER 1969 15 14 2?~13 '212 X 6 II IIJ 10 9 0 8 7 0 ti 6 IIJ 5 0 ~4 a: 3 2 0 .__...__.__.._ ............ __. ........... _,__._...____.___,__i.._.._...i.--, 2 3 4 5 6 7 8 9 10 II 12 13 14 15 PULSE NUMBER Fig. 6 Cumulative of hydrogen atoms exchanged between catalyst and hydrocarbon during passage of successive pulses of benzene C o D o and cumene over a silica-alumina catalyst in a microcatalytic reactor the hydrocarbon products during cumene dealkylation in the last seven pulses. DISCUSSION These microcatalytic tracer cumene dealkyla tion experiments over a s ilica-alumina catalyst are well su ited for a se nior chemical engineering laborator y. In a si n g le integrated experiment in vo lvin g three laborator y periods, the students are int roduced to a wide range of co ncepts and tech niques including catalysis, kinetics, gas chroma tography, product trapping, vacuum systems, i so topic tracers, and mass spec trometr y. None of the chemicals is very expensive, and the micro catalytic reactor (excluding the recorder and potentiometer) can be built for le ss than $350. In our own department the mass spectrometer from the catalysis research laborator y was made available for these experiments. A research assistant was in c h arge of the mass spectra l a nal ys e s, but the s tudent s themselves performed all ot he r parts of t he experiment. The use of stabl e isotopic tracers has demon strat ed that what a ppeared to be a relatively s imple heterogeneou s catalytic reaction in fact involves quite a complicated mechanism. Thi s certainly inv alidat e s the naive assumption of first order kinetics. Furthermore, since there was a temperature dependent peak broadening due to adsorption as each pulse was passed over the 121

PAGE 20

T ab l e J Is otopic Composition of Products i n Mic r ocata l ytic Tracer Experiments -------~ I ~s~o~t= qpic C o mposition ( % ) #D at om s molecule Pul se _Q 9_ ._ 1 2 3 4 5 6 7 8 9 10 11 1 2 13 1 4 1 5 In j ected M ec:,s ured B ,d 6 B B ,d 6 B B, d 6 B B ,d 6 B B ,d 6 B B d 6 B B d 6 B B d 6 B C, d 0 p B C C,d 0 B C,d 0 B C, c1 0 B C, d 0 B C, d 0 B C d 0 B 1. 7 7. 3 16 .7 0 .4 2. 3 8.4 0 0. 9 4 .7 0 0. 5 2 3 0 0 1 1. 5 0 0 1. 2 0 0 0 6 0 0 0 5 1. 2 9 6 22 6 6 2 1 8 .1 25 9 3. 8 1 3 8 22 6 25 .7 36 .7 2 3 9 4 3 .8 37 0 1 5 .1 53 3 33 .7 10 5 6 3 3 29 5 6 5 72.0 2 3 .7 3 9 79 1 1 8 8 2 .1 catalyst, the assumption of constant contact time at various temperatures is also invalid. These two factors were mainly responsible for the ap parent activation energy being much lower than that 1 3 reported in the literature over similar catalysts in a steady state flow reactor. Although microcatalytic reactors certainly are not the best suited systems for kinetic meas urements, they are extremely useful for isotopic tracer studies for several reasons: Tracer compounds are expensive, and only very small samples need to be used in this system. It is possible to study "initial" interactions between reactants and surface after only a relatively few hydrocarbon molecules have c ontacted the catalyst. The method is reasonably fast. With very few changes, this system can be modified to study the reaction under steady state flow conditions. Kinetic comparison between the microcatalytic and steady state flow systems can be made to investigate the role of diffusion in the reaction. Furthermore, the GLC is sufficiently versatile that it can be used without modification for other analyses The authors acknowledge grants from E. I. du Pont de Nemours and the Petro-Tex Chemical Corporation which 122 24 .1 2 4 4 17. 6 8 2 2 522 1 8 9 27 9 26 6 15 5 1. 866 1 4 2 2 6 9 32 6 2 0 .7 1. 52 3 9 4 23 4 36 0 28 4 1. 227 7.1 21.0 3 7. 2 33 1 1. 070 6 5 20 .1 37.6 3 4 6 1. 021 4 0 1 6.0 3 7. 3 42 .1 0 837 2 8 1 3.0 35 5 48 :;: 0.71 9 29 5 23 4 1 1.0 2 7 3 081 2 3 1 1 4.4 7.7 4 6 2 6 2 9 2 1. 8 1 5 3 9 4 6 .1 4 2 2 .1 0 9 3 235 9. 9 2 8 0. 6 0. 4 1. 308 3. 6 0 5 0 0 0. 800 2 .0 0 4 0 1 0 0 62 8 0.7 0 0 0 0. 446 0. 4 0 0 0 0. 32 7 0 0 0 0 0. 230 provided funds for development of these expe r iments They are also grateful for helpful suggestions from other members of the Chemical Engineering faculty. REFERENCES 1. P. Sabatier, "Catalysis in Organi c Chemist r y ," tr by E. Emmet Ried, Van Nostrand Co., New Yo r k, N Y. 2. I. Langmuir, J A m C h em Soc 40, 1361 (1918) 3. H S. Taylor, Ad v Catal 1, 1 (1948). 4. V N. Ipateff, Catalytic Reactions at High Pr es sure s and Temperatures, MacMillan Co., New York, N Y (1937). 5. P. H Emmett Catalysis, Vol s I-VII, Reinhold Publishing Corp., New York, N Y. (1954). 6. C. D. Prate r and R. M. La g o A dv C atal 8, 2 93 ( 1956 ) 7 J. T. Richard s on, J. Catal 9, 182 ( 1967 ). 8 R. J Koke s, H. Tobin, and P. H. Emmett J. Am. C h em Soc. 77, 5860 (1955) 9. J. W Hightower H R. Gerb e rich and W. K. Hall, J Ca tal. 7, 57 (1967). 10 Y. Murakami, T. Hattori, and T. Hattori, J Cat al. 10 123 (1968). 11. W. K Hall, F. E. Lutinski, and H R. G e rb e rich, J Cata l. 3, 512 ( 1964). 12. J. W. Hightow e r and W. K. Hall, unpublished 1 es ult s 13 W. F. Pansing and J B Mallo r I nd Eng. C h em P r oc es s D e s i g n an d D eve lo pment 4, 181 (1965). CHEMICAL ENGINEERING EDUCATION

PAGE 21

if YOU can answer "YES" to any of these questions, Rohm and Haas can make your life interesting. Do you enjoy being responsible (and the rewards that go with it) for developing a project and carrying it through to completion ? Do you like a job that gives free rein to your imagination and tests your skills ? Do you like the satisfaction of making a contribution to society by improving a wide range of industrial and consumer products or by protecting natural resources or improving farming efficiency? Would you like to work for a company that is concerned with problems of social responsibility and encourages its employees to be active in urban affairs? RDHMD iHAAS~ PHILADELPHIA, PENNSYLVANIA 19105 We are not altruists. We are a strong and growth oriented chemical company that needs engineers of all types-chemical, mechanical, electrical and industrial-who want to make a practical contribution to improving man's lot and at the same time advance themselves in the world of business. We have doubled our sales in the past 10 years to the $400,000,000 level. We make some 2,500 chemical products including sy nthetic resins, plastics fibers, pharmaceuticals and animal health aids. We are on the move and you can move with us into positions of responsibility just as rapidly as you show us you have the ability There's just one thingy ou'll be expected to w ork hard-but you'll be in good company with the 13 000 other people that make up Rohm and Haas We have six major manufacturing locations in the United States and producing subsidiaries in 15 foreign countries. We need engineers in research, production and marketing, Write to Manpower & Employment # 7069. It could be good for both of us. ROHM AND HAAS COMPANY An equal opportunity employer ~

PAGE 22

iiN:lviews and opinions THE DILEMMA OF INNOVATING SOCIETIES: Implications A.G. FREDRICKSON Univ er sity of Minn e sota Minn e apolis, Minn. 55455 In the op e ning pa ges of hi s boo k Anci e nt Europ e Profes s or Stuart Piggott ':' introdu ces the useful concepts of innova ting and conserving soc ieties. H e write s (p. 17) : "In the one g roup, t ec hnological dev e lopm e nts in the arts of peace and war mu s t hav e be e n socially acc e pt able and th ere for e e ncouraged; in the other, once a satisfactory modus viv e ndi fo r the community within it s natural surroundings had be e n achiev e d, th ere see ms to have been no urgent need felt to alt er the si tuation. Or again, the cultural patt e rn d e vised might be too delicat e ly adjusted to th e circumstances, and too rigidly conceived, to be susceptible of modification by t ec hnological innova tion . Professo r Piggott's d e finition suggests two reasons why a community mi g ht choose to become a con serving society, and we shall ex plore his second possibility that concerning the delicat e adjustm e nt of th e s ociety to its environmental circumstances-in s ome d e pth. First, hc .ve ver, it is necessary to s ay a few words about inno vating societies. T HERE IS NO doubt that American Society, or at least its white, middle-class sub-society, is an innovating society. Indeed, few would deny that the scope, scale, and pace of its technological innovations for it was technological innova tion that Professor Piggott was thinking of when he wrote his definition outstrip those of any of its contemporary societies, including the Soviet Union, Western Europe, or Japan. Many Amer icans take great pride in our position of leader ship in technological innovation, and most Amer icans would reject with scorn any suggestion that the United States become a conserving so ciety. To find support for these statements we do not even need to look beyond our own academic cloisters : the phenomenal growth of scientific and engineering research and education in the universities over the past two decades is a clear indication of the high regard for technological innovation held by the decision-making and direc tion-determining segments of our society. The only audible voice of dissent to the notion that the United States must remain the leading inno vating society seems to come from a small but vocal group of young activists who have deci* Stuart Piggott, "Ancient Europ e ," (Chicago: Aldine Publishing Co., 1965) 124 Arnold G. Fredrickson has BS and MS degrees from the University of Minnesota and a PhD from the Uni versity of Wisconsin His principal research int eres t involves the study of th e dynamics of biologi c al popula tions, with special emphasis upon the interactions of s u c h populations with their environment. In addition he is a dedicated nature photog r apher and student of the botany of Minnesota His research interests coupled with his avocation prompted the writing of this essay. sively rejected the values of contemporary so ciety. Unfortunately, this voice is so inexperi enced, so often charged with passion and so con centrated on highly specific issues, that it has not called attention to the basic dilemma that must bedevil all innovating societies. This dilemma is suggested by Professor Pig gott's second explanation for the origin of con serving societies. Let us recall his words : ". . the cultural pattern devised might be too deli cately adjusted to the circumstances (of its natural surroundings) to be susceptible of modification by technological innovation . ." That is to say, the environmental resources of land, water, vegetation, minerals, energy, etc. may be so limited that technological innovations within the possibility of the restricted experience and capability of the community would upset the balance of the community's ecological situation irrevocably. For instance, a member of a hunt ing society inhabiting a forested region might conceive the idea that game could be driven from its cover by selective burning in the forest, out into places where it could be easily captured Obviously, if the territory available to the tribe CHEMICAL ENGINEERING EDUCATION

PAGE 23

for Engineering Education is limited, and if the tribe's rulers have a modi cum of foresight and survival instinct, they will squelch this kind of innovation. A member of a modern innovating society might suggest that by further innovations (such as devices to con trol the area of burning or devices to defeat the tribe's enemies in war so as to enable expansion of its territory) the tribe could solve its urgent and ever-present problem: that of feeding every one in it. Such a suggestion is not useful, how ever, because it ignores the point that the tribe often does not welcome such innovations, even when it is clear that they would be beneficial. On the other hand, technological change is welcomed in innovating societies and an enormous capacity for in novation is the hallmark of present-day societies of that variety. Indeed, there may even be some individuals who hold the innovating capacity of those societies to be in finite. Be that as it may the rate at which innovations can be made is limited, both by the total of experience possessed by the society, and by the material resources available to it. The first limitation on the rate of inno vation is perfectly elementary; every student of freshman physics knows that Newton could not have invented the radio because he lived before Maxwell was born. The sec ond limitation on the rate of innovation is also perfectly obvious to anyone who chooses to think about it. We live in a large but nevertheless definitely bounded biosphere. It has only so much lebensraum, so much sustenance that we can draw from it, so much space in which we can dump our waste matter and energy, and so much ca pacity for self-regeneration. We are not going to farm the moon to feed the starving billions of Asia nor can we export the surplus population there, neither can we etherally dispose of our pollution problems by packaging waste materials and firing them off to the great incinera tor in the sky. These considerations suggest that innovating societies, no matter how ingenious they may be must eventually encounter environmental restric tions on uncontrolled innovation just as do con serving societies. Furthermore, when such re strictions become apparent, they do so at an almost unbelievable level of complexity and scale, and there is even the chance that irreparable environmental "mistakes" will have been made. Finally, though a primitative and conserving so ciety may have the bonds imposed by its environ ment and its own lack of technological capacity released by contact with an innovating society, there is no higher institution to which an ad vanced innovating society becoming acutely con scious of environmental restrictions can turn for guidance. This, then, is the basic dilemma of all innovatSUMMER 1969 ing societies: On the one hand, th~ir institutions and ways of doing things are founded alike on a belief in the efficacy of unfettered technological innovation for social progress and a fear that without such innovation, they will stagnate and decay. Thus, they look on conserving socie ties with condescension or disdain. On the other hand, they are confronted with the fact that con tinued uncontrolled innovation is incompatible with the material basis of life on earth; it is suicidal. Therefore, if innovating societies are not to destroy themselves, they must adopt some of the features of conserving societies; they must attempt to strike a bargain with Nature instead of simply exploiting or seeking to overpower Her; the societies must seek a material state of things that is more nearly steady than that to which their beliefs and inclinations have accus tomed them. IN :W~AT SENSE MYST INNOVATING societies become more hke conserving socie ties if they wish to survive? Not, to be sure, in the sense that they would forbid technological innovations out of a satisfaction wi th the status quo or out of a dogmatic adherence to traditional ways of doing things. What is required, rather, is a sense of proportion and priorities geared to the real needs of man and to the hard facts of his existence on this planet. Within such a set of priorities, innovations for the real benefit of humanity would be pressed with all possible speed. But innovations that are simply frivolous would be looked on with disfavor, and innova tions that are destructive of the environment would be suppressed. The practical problems of developing the re quisite priorities, a mechanism to ensure their application, and means to review and revise them as needs change, are severe. This is so in large part because there exist factors within innovat ing societies and within our own innovating society in particular that militate against any slackening of the pace of technological innova tion or any attempt to control its direction. These factors may be ill-defined or inarticulated but they nevertheless seem to cater to or perhaps be expressions of some deep-seated urges in our socity, and they give tremendous momentum to the processes that they generate; they determine the dynamic aspects of our society. We do not mean to imply, of course, that the dynamics of our society are determined solely by (Continued on page 144) 125

PAGE 24

[iJ ;j a classroom TRANSPORT PHENOMENA EQUATIONS OF CHANGE V. J. LEE University of Missouri Columbia, Missouri 65201 Equations describing fluid motion and energy transport have been derived either from the Eu lerian point of view of a stationary fluid element of infinitesimal volume or from the Lagrangian point of view of a macroscopic volume of fluid in motion. In the former derivations, lengthy mass, momentum, and energy balances are in volved. In the latter derivations, integral trans formation theorems and the Reynold's transport theorem are needed. 1 2 The transition from New tonian body mechanics to fluids mechanics is less than direct in both of the two derivations. This note presents a derivation of equations of fluid motion and energy transport by considering an infinitesimal fluid element. av, in motion. In addition to formalistic simplicity, the derivation exposes the conceptual continuity from the New tonian equation of "body" motion to the con tinuum motion of fluids. I. THE RATE EQUATION OF VOLUME DILATION Let the mass velocity of an infinitesimal vol ume element av be v. The rate of dilation of av spanned by the vector v is gt (aV) = f J (v n) as (1) where dS is a surface element, n is a unit vector normal to dS. The integration is to be carried out at time t, over all the surface of av, whose coordinates XJ are equal to XJ (t) with j = 1, 2, 3. By the divergent theorem, 2 one has V v = lim aV O Jv ff (v n) dS Hence equation (1) can be written as D -( 8V) = av V v Dt 126 (2) (3) Dr. Lee is an associate Professor of Chemical Engi neering at the University of Missouri, Columbia. He was educated at the Ordnance Engineering College, Taipei, Republic of China (Diploma Engineer), University of Notre Dame (MS) and the University of Michigan (PhD '63). His interests include heterogeneous catalysis, reactions kinetics, solid state and surface physics, thermodynamics, transport phenomena and energy conversion equation (2) expresses directly that for an in compressible fluid V v = 0 (4) II. THE EQUATION OF CONTINUITY The equation of continuity expresses the con cept that av is a closed system as to mass trans fer; i.e., a "body." Let p denote the density of fluid, the law of mass-conservation gives D 75t (paV) = 0 (5) Remembering av -=I=0, differentiating equation (5) and combining it with equation (3), we obtain Dp + p V v = 0 Dt Ill. NEWTON'S EQUATION OF FLUID MOTION (6) In view of equation (5), we can regard av as a "body" with mass paV. Applying Newton's second law of motion to the "body" we obtain gt (p8Vv) =-ff pndS J J (n "T) as -paV V cp (7) On the left hand side of equation (7) is the rate of change of linear momentum. On the right hand side the first and Jhe second terms are summaCHEMICAL ENGINEERING EDUCATION

PAGE 25

tion of forces acting on the "body" along the in ward normal of its surface due to pressure and viscous tensors respectively. The last term is a force acting on the "body" with mass paV due to potential field cp in energy per unit mass. Noting equation (5), we can rearrange equation (7) in the form p __E__ (v) = 1 -JJ p n dS Dt av fJ~ J J (n r) dS -p '\l cp (8) Applying the divergent theorem and noting av is infinitesimal, we obtain D p ])t (v) = '\J ( '\J T) p '\J


PAGE 26

NOTATION v Mass velocity of fluid av Volume of an infinitesimal fluid element dS An infinitesimal surface element n A unit vector normal to dS D Dt Substantial derivative operator ( a a Del operator = --::i. c)X 1 uX 2 a! a ) p Density of fluid, a point function of X 1 X 2 X 3 and time t cf, A scalar potential function of X 1 X 2 X a T The viscous tensor of a fluid q Vector heat energy flux v Magnitude of fluid velocity R E FE RENCES 1. Bird, R. B., W. E. Stewart, and E. N. Lightfoot, Transport Phenomena, John Wiley & Sons, New York, 1960, pp. 74-81 and pp. 311-7. 2. Sommerfeld, A. Mechanics of Deformable Bodies, Aca demic Press, London 1964, Aris, R. Vectors, Tensor s and Basic Equations of Fluid Mechanics, Prentice Hall, New Jersey, 1962, Chapters 3-6 3. Spiegel, N R., Vector Analysis, Schaum Co. New York, 1959, pp 122-3, also Reddick H W., and F H Miller, Adv. Math. for Engineers, John Wiley & Sons, New York, 1955, 3rd ed., pp 350-4. [j n pl book rev i ews An Introduction to the Engin e ering Res e arch Project Hilbert Schenck, Jr. McGraw-Hill Book Co., New York (1969) After having directed many theses and over seen thesis direction for many years, this writer has thoroughly enjoyed reviewing this small ( 178 pg. 5 x 8 ) book. Intended to be an introduction to the engi neering research project, it moves swiftly from the selection of a topic through all the major steps to an expected acceptance of a finished manuscript for publication. The author is relent less as he points out the foibles of faculty and academic systems and is no less discerning a s he analyzes student "hang ups" which would hin der the choice and early completion of a desirable 128 research job. The book i s written in contempo rary style and should be comprehensible to both the would-be-researcher and his director. Analyzing the volume in more detail, the re viewer believes that "The Selection of a Project" covers the field well but probably ascribes some what more than a normal amount of initiative to a student. Unfortuately the conception of a project more frequently falls on a faculty mem ber than on a student and therefore Chapter 2, "Sources for Project Ideas" (25 pages), is far too long. However the short and meaty "Project Check Sheet" should be noted by everyone. The chapter "Searching the Literature" at tacks the subject with clarity, vigor, and decision. It quickly covers the usual but needed generaliza tions but follows them up with a well conceived and highly possible case history. How many times have projects failed for lack of apparatus, time, or cost planning? Here is an author who believes in these efforts as an inte gral part of the project. Indeed he stresses these activities not only as highl y desirable but even mandatory if a real re s earcher and a satisfactory project are to be produced. His tips are pertinent, timely and frequently annoyingly discerning. Unfortunately the author chooses to elaborate next on his categories of research an area which he could better have omitted for al though his discussion of "Digital Computer Studies" is a good s hort approach to a long problem, his "Pedagogical Studies" and "Design and Sy s tems Area s" ar e far below his overall standards. In his last two chapters on "Reports" and "Journal Papers and Meeting Presentations" the author has been appropriately and pleasantly brief. He has obviously called upon many experi ences, both sad and glad, and has extracted an essence which combines philosophy with prac ticability. There is much in this book for new researchers to learn before sad experiences can dishearten or even remove them completel y from the field, but the book also may be a gage for a more experi enced researcher or research director to recheck his effectiveness. Surprisingly despite the "heavy" material c ontained in this book, the style is light, friendly and interesting; it is to be hoped that the experi mental project report s will be, too! Gordon C. Williams University of Louisville CHEMICAL ENGINEERING EDUCATION

PAGE 27

CHEMICAL ENGINEERS When you GO with TEXACO you GO with the BESTI The BEST Facilities ... The BEST Products ... The BE ST People Yet, being the best makes it that much harder to get better. With competition constantly at our heels, it takes an aggressive and imaginative R & D team to maintain Texaco's leadership role in the petro leum and petrochemical industry. An important part of this team are Chemical Engineers ... men like yourself ... constantly searching for better ways. It is through their efforts, as well as their professional colleagues and an aggressive management team, that Texaco stays out front. You, too, can be part of th i s winning combination For Chemical Engineers with a B.S or M S. the professional and economic rewards of a Texaco career in process and product development have The BEST Opportunities never been greater. Texaco's practice of basing pro motions on performance and ability afford you the BEST opportunities for advancement, while enjoying our congenial, shirt-sleeve atmosphere. Texaco has immediate openings at its laboratories in Beacon, N. Y., Richmond, Va and Port Arthur and Bellaire, Texas for qualified Chemical Engi neers who already are, or are in the process of be coming U. S. Citizens Interested candidates are invited to send their resume to: W. R. Hencke, Texaco, Research & Tech nical Department, P. o. Box 509, Beacon, N. Y. 12508. Texaco is an equal opportunity employer

PAGE 28

A SELF-PACING, AUTO-GRADED COURSE* G. DAVID SHILLING University of Rhode Island Kingston, R. I. 02881 INTRODUCTION A two-credit course in reactor engineering, required for senior students in Chemical Engi neering at the University of Rhode Island, has for the past three years been operated in an auto-graded mode (so called because the student supposedly decides what grade he will get.) Be cause the senior classes are not large, only one section of the course was scheduled each year, and all the students participated in this experi mental program. Except for three graduate stu dents, the students have all been seniors. A wide range of scholastic achievement levels has been represented from students headed for gradu ate school, down to some on the verge of being dismissed. The goal of the experimentation was to find a way out of a situation where the instructor met two hours a week with a well-knit little group of tired students willing to let him do all the talking and most of the work. It was hoped that by using the auto-graded mode, which was described in ASEE Publications, first by Norman Bala banian1 and then by Roland Mischke 2 the more ambi tious students could be released from the lock-step pace and the dependence on the instructor; and by requiring each student to take the initiative for his progress toward a passing grade, the less ambitious students would be confronted with some facts of life. DESCRIPTION OF COURSE The rules governing the operation of the course are shown in Table I, the "Course Plan" which was distributed to the students at the be ginning of the course. A student's progress through the course is marked by his passing a series of tests, and his course grade is deter mined by the number of tests he passes ( eight for an A, seven for a B, etc.). Each test "covers" one chapter in the textbook by treating in detail a set of problems illustrating the principles set 1 N. Balabanian, "Removing Emphasis on Grades," J. of Eng Ed 54, No. 7 (March 1964). 2 R A. Mischke, "A Semitutorial Approach to Teach ing," J. of Eng. Ed., 56, No. 3 (November 1965). *Presented at the Annual Meeting of ASEE, June 17-20, 1968. 130 G David Shilling is a graduate of the University of Delaware and the University of Wisconsin (PhD 50). He taught at Kansas State and is now completing his seventeenth year on the URI faculty. He developed an interest in process control at summer institut es at Case Institute and the University of Colorado and published P r oc e ss Dynamics and Control in 1963. Since his parti cipation in the ASEE's Programmed Learning Project (1965-66), he has experimented with programmed in struction and self-paced and auto-graded courses. TABLE I COURSE PLAN Chemical Engineering 64 Fall 1967 The text book will be Ch emica l Reaction Engin eerin g by Octave Levenspiel. The instructor will select and announce a set of prob lems for each chapter as follows: Chapters 2, 3, 5, 6, 7, 8, 14, 13. A student will obtain credit for a chapter by scoring 90 % or better on a written examination on the chapter. Only then will he be eligible to take the examination on the next chapter, in the order shown above Examinations will be limited to 55 minutes. Th e y will be given at 1 P.M. on Monday, Wednesday, and Friday, between September 21 and January 13, school holidays excepted The instructor will retain all copies of the examina tion papers and questions. A student may see and discuss his examination paper when conv e nient. A student's course grade will depend on the number of chapters he passes as follows : eight, A; seven, B; six, C; five, D . forth there. These problems are much more involved than a student could be expected to work "from scratch" in an hour examination. But the students are given the problems in advance and have worked them and perhaps discussed the solutions with the instructor and other students. It is assumed that each student will choose what course grade he will get and decide when he will pass the required tests for that grade. To help the student pace himself through the semes ter, he is given a copy of the "Experience Table" for last year's class. The table shows that it is CHEMICAL ENGINEERING EDUCATION

PAGE 29

not unusual for a student to take a test two or three times before achieving the high-quality performance required for passing (90 % ). The tests are so comprehensive that, in order to score 90 % a good mastery of the material covered in the chapter is needed, even when the student has taken exactly the same test a few days earlier and afterwards discussed his errors with the instruc tor. In the beginning of th e experiment (Class of 1966) class meetings were held as scheduled. But it soon became apparent that the students were getting "out of step," so that too few students were finding any given topic sufficiently p er tinent to the work th ey were doing (or putting off) to make class meetings fruitful. Aft er six weeks of regular class meetings with steadily decreasing attendance (since no effort was made to coerce attend ance) no more such meetings were held, although some extemporaneous lectures were g iven to small groups. With th e Classes of 1967 and 1968, class meetings were discontinued as soon as the students voted to us e the auto-graded mode. The most efficient plan would prob ably be to hold class meetings for the first three or four weeks, while th e students are relatively "in step." But the students find the no class feature one of the big attractions of this mod e of operation. Some of the disad vantages of not having formal class meetings may not have showed up in this experiment, because nearly all the students were also e nrolled in a laboratory course which met six hours a week with the same instructor. So there were many opportunities for student-instructor consultation. The advantages and disadvantages observed in the operation of the course are discussed later. COURSE EVALUATION A course-evaluation questionnaire was filled out by the students at the end of the course The returns were anonymous and it appeared that the students felt quite free to express their feel ings and suggestions. In 1966 there were seven returns from a class of ten, and in 1968, nine from twelve. In 1967 the students got away be fore the questionnaire session could be arranged. For the two classes covered, the total enrollment was 22 and the number of questionnaires com pleted was 16. Table 2 summarizes the answers obtained. To encourage the students to think about each ques tion, the answer blanks on the questionnaire were scrambled as to positive, neutral, and negative reactions. In Table 2, the answers have been rearranged for easier analysis. The two classes are not tabulated separately because they did not differ greatly in their opin ions. The Class of 1968 indicated a somewhat more positive attitude: surer of mastery of the SUMMER 1969 TABLE II COURSE EVALUATION QUESTIONNAIRE Questions Do you feel that the material covered in this course will be of value to you in your career? Did you find the course work interesting? Did you feel that you have a good understand ing of the material? How did the amount of material covered com pare with other two credit courses? What do you think of the text book used? How did you feel about this course plan at the beginning of the se mester? How would you feel about it now? How well do you think this plan would work on various class levels? How well did the prob lems assigned cover the important material in the book? How well did th e exams test your mastery of the problem topics? Answers Suggested Number Checking yes perhaps no 8 6 2 most of it some of it littl e 9 4 3 most of it some littl e 9 7 0 too much average too littl e 11 4 1 ye s no interesting 8 6 easy to understand 11 3 well organized 14 2 too condensed 3 12 too varied 1 10 enthusiastic willing reluctant 5 8 3 en thusiastic willing reluctant 0 14 2 well poorly average Soph 0 12 4 Junior 1 5 10 Senior 7 3 6 well pretty well poorly 5 9 2 well pretty well poorly 5 8 3 material and of its career value, less critical of the textbook, and not so impressed with the ex cessive amount of work demanded. On the other hand, they felt more keenly the danger of "put ting-off" work, and were more critical of the tests. This modest improvement in attitude, dis played by the third group compared to the first group, may be in part due to more experienced handling of the course by the instructor. Also, the third group had two outstanding students and a well-developed group spirit. DISCUSSION The amount of material covered in this course varies from student to student. ("A" students study more chapters than "B" students; "B" students more than "C" students, etc.) because 131

PAGE 30

TABLE II (continued) Check any of the statements below which represent fairly closely your experience in this course : I spent a lot of time studying the text book on the topics covered by the problems. I read the whole chapter carefully. I also read some chapters not covered. I tended to ignore the book and get problem solutions from other students I did the problems mostly on my own and really understood them. I tried to memorize the problem solutions instead of understanding them. I felt that having the course organized this way saved time for me and let me do my best. I expected to pass more exams, but found the latter chapters were too hard or took too much time It was easy to put off working on this course. I would have done better if there had been some deadlines at (monthly) intervals. I didn't like having to plan my own work. I felt that some students had an unfair advantage. The instructor didn't seem to care whether I worked or not. The instructor was not very helpful when consulted What factors hurt exam effectiveness? too easily memorized not enough time asked wrong things hard to interpret students cheated too mickey-mouse poor surroundi n gs too much time allowed Write-ins (one each): had to be memorized many topics not covered had to memorize numbers rather picayune too much detail in correction Number Checking 12 8 2 5 11 5 5 3 12 6 2 5 8 3 6 4 2 2 2 1 0 0 this seems to be the simplest way to organize an auto-graded course. This disadvantage would be harder to accept if the course were not at the end of the curriculum. On the other hand, compre hension level is maintained high for all students ("90 % "). In the conventiorial, lock-step course, there is a standard "coverage" and the students are graded according to their comprehension level. Actually, both plans have arbitrary limita tions: 90 % is not perfect, and there are always more topics which could profitably be included in the standard coverage. For many a student, the feeling that he has achieved high-level com132 mand of a topic would be a welcome change from the feeling of failure or mediocrity he gets in many of his courses. It is certainly more realistic training for engineering practice to develop a comfortable facility with a limited subject area than to get a haphazard acquaintance with a broader field. The coverage achieved in this course is considered by the author to be excellent for the two semester-hours credit given. Of the 34 students who have taken the course, only four (grade D) have stopped short of consideration of optimum-temperature progression in homogene ous reactors (C-level), while thirteen (A and B), also got work in heterogeneous reactions. Retention of skills developed is, of course, as import ant as developing and demonstrating these skills. As in the usual educational situation, no measurement of reten tion was here attempted. However, the instructor was sensitive to indications of retention observable during a laboratory course the students take in the semester following the kinetics course. The observations have been favorable: when these students were assigned lab prob lems related to reaction kinetics, they showed quick recall of relationships, procedures, and even some details. Their attitude suggested confidence in their ability to handle this subject matter. It is reasonable that reten tion of skills learned in this type of situation would be relatively high, because the student has to "dig it out" for himself, and because, before leaving each unit, he is assured that he has a good grasp of the material. The problems chosen by the instructor for the students to work on need to be "comprehensive," i.e., requiring a good grasp of the entire subject area to be "covered." It is not cricket to require the student to show more on the test than he had to do to solve the problems. On the other hand, the problems do not have to be neat and limited; the students have time to chew them over, seek out additional data, and resolve ambiguities (as in engineering practice). Some "old" problems can be used, since it is not essential that each student work every problem entirely on his own. This year, about half the problems in each set were new, and the rest were taken from previous years' sets. Some students attempt to memorize problem solutions borrowed from other students or found in "files" left by earlier students. How ever, students have testified that this approach is not successful. The problems are so involved that it is very difficult to write a 90 % test with out understanding the solution. The tests need to be comprehensive enough that the student must either work through the problems himself or thoroughly st4dy the solu tions he borrows. The questions do not have to be pared-down to what an average student can CHEMICAL ENGINEERING EDUCATION

PAGE 31

reasonably be expected to work out in 55 minutes. The student works on the problems before the test and he can take the test a second or third time if he has trouble assembling and organizing his answers in 55 minutes. So, even ambiguities in test questions, though to be avoided, do not have tragic consequences. Repeated tests are not composed in the fa miliar "sampling" mode. This would result in students using information about the test ques tions as a guide to slighting important parts of the topic to the "covered." A typical test is shown in Table 3. Students are asked to show parts of TABLE III TYPICAL TEST Exam on Chapter 5Fall 1967 1. For Problem 11 (see problem statement below), derive the differential equation relating reactor volume to fractional conversion, and show the computations of the values of the constants used (in the integrated form of the equation) to find k from the experimental data. 2. For p r oblem 18, derive the differential equation relating volume of reactor to fractional conversion, and explain how you proceed to compute the volume of the required reactor. 3. For Problem 19, derive the required equation and show aU computations for the volume of (only) the back mix reactor. (Problem statements followed.) some problem solutions in detail, while for others, they are asked to describe how the problem is solved. Students are often required to show the source of a model equation. This they refer to as "memorizing derivations," which they consider unfair. They also find that they can best get through a test in the limited time if they memo rize a few key numerical values. Although many complaints result, it is doubtful that there is any lasting resentment. Students are so used to cramming for exams, that to memorize, ten min utes before a test, a few simple things is not much strain. The requirement that a test failed must be repeated in its entirety is the cause of the most-often expressed student irritation. The repeated-test feature produces a steady stream of tests to be graded, and grading often must be done with unusually high precision, in order to decide between a 91 and 89 without seem ingly arbitrary or inconsistent. Of course, a test paper which has two really bad flaws can be checked off quickly, as can a well presented repeat test by a student who nearly made it last time and has been shown his error. It is even more essential than in conventional courses that tests be graded promptly within 24 hours at SUMMER 1969 the most. This may require some careful plan ning by the instructor, such as limiting the time when tests are given so that he is free to correct them immediately. No "final exam" was given in this course, as there seemed to be no role for it. The role of the instructor in an auto-graded course is different from that in a conventional course. It is expected that the student will think of the instructor as a source of useful information. The instructor can promote this attitude by giving out hints and checking students prob lem solutions, as well as by explaining mistakes on tests. This consultant role contrasts with the need a student sometimes feels to "snow" his instructors (impress them with his knowledge while not revealing gaps in his com mand of their subjects.) The instructor of an au t o-graded course is in a good position to find out what a student needs help on, and what he can do on his own. In order to use thi s opportunity effectively, the instructor needs a firm grasp of his subject matter and the ability to listen to students. When meeting a rapid succession of students with questions on a wide range of topics, he has to "shift gears" a lot. If they come in groups, he may find himself operating in a "time-sharing mode. When the instructor finds himself treating the same, often trivial, question over and over as each student comes upon it, it occurs to him that if he were giving a lecture course he would need to discuss the point only once. He can minimize his losses here by individualizing these encoun ters and using them to build up rapport. For topics that are sure to cause trouble for a lot of students, a mimeographed hand-out can be pre pared. With the excellent textbook used in this course, only one such text supplement has been prepared. (It deals with the question of changing density of the reaction mass in a flow reactor.) The instructor of an auto-graded course spends more time "consulting" with students, but less time preparing and delivering lectures (and won dering if he is "getting through" to the stu dents.) The auto-graded course rather than promot ing competition between students, stimulates the formation of study groups. Leading students get considerable opportunity to help other students, and strengthen their own learning in the process. A certain amount of working together on prob lems is good training for engineering practice, and in an auto-graded course, is not the threat of the instructor that it sometimes is in conven tional courses. The effect of this course format on a student's motivation is of course difficult to generalize. Ostensibly, the student is working for a grade and whatever that means to him. Although this does not sound very commendable, perhaps in the 133

PAGE 32

present-day college context it is not in any real sense a regression. And, it would seem that it would help a student focus his energy to know that the grade he chooses is his when he demon strates the required learning. The students indi cated on the course-evaluation questionnaire that they put more effort into the course than they would expect to put into a two-credit course. The pressure on the student in this course was rather even and continuous, compared to conventional courses, where there are sharp peaks before six-weeks exams and low periods between them. This low-tension atmosphere, while it was an advantage for some students, was the most important factor limiting the general success of the course. Many students put off working on the course so much that they came out with C-level achievement where they could, with wiser investment of their time, have attained the A level. (Some of these students may have been waiting for their leaders to move, and when they did move, couldn't keep up.) To combat procrastina tion, the instructor supplied eac h student with an Ex perienc e Table at the beginning of the course, and offered additional copies occasionally. He also drew attention to the deadline esta blished for the end of testing. (This deadline was extended for one student because of illness and for one who was working hard on a D.) More effective ways of reducing student procrastination are still being sought The self-pacing feature, which permits pro crastination, is on the other hand a strong mo tivating factor for some students. They are very impressed by the prospect of finishing the course well before the end of the semester which is entirely possible and has been done b ysome stu dents. Undoubtedly a lot is gained, in the way of good feeling about the course, by the freedom from weekly schedules, class attendance, and the final exam. Of course, one can not say how much this OPTIMIZATION R. R. HUGHES (Cont'd from p. 116) 4. Equipment Design The final category is really just good equip ment design. To obtain detailed pictures of the makeup of packed bed reactors, the nature of internal baffles in stirred tanks or the exact form of heat exchanger bundles, a good designer must optimize in terms of some minimal cost or maxi mum-performance criteria. Each type of equip ment requires its own special treatment for op timization, so that a general treatment of equip ment optimization is not really desirable. In many cases, however, it may be possible for the designer to make use of some of the optimization algorithms. 134 pays off in faster more lasting learning. If most courses were self -p acing, the advantages of nov elty would fall to the lecture-exam courses. Ideally, the different courses in a curriculum should be operated in a wide variety of formats, each in an optimal way for its particular objec tives. Cheating is not an important factor in this course, because the test questions are not secret and tests failed are repeated without significant penalty, nor is any advantage gained by not giv ing credit for help received. Ethical problems are raised by a largely undesirable behavior pat tern known as "leaching," where an unconfident student will attach himself to a leader, or one with a channel to a leader, in order to obtain more information about the problem solutions than he is able to contribute. There is little the instructor can do about such a situation. The students tend to work out arrangements so that all involved gain something. CONCLUSIONS All the students in this auto-graded, self-pacing course demonstrated high-quality command of a reason able amount of chemical-reaction-engineering skills. There were no failures, and only one early drop out. Forty percent of the students received A or B grades for work beyond the level considered satisfactory for the number of credits given. There is every reason to believe that their retention of this learning will be superior The amount of material covered varied with the student, and would have been greater for some if the tendency to procrastinate had been suppressed Students generally agreed that they worked harder than in most courses. A theoretical advantage of the course is that the student behavior encouraged (if not uniformly obtained) bore a strong resemblance to that of a practicing engineer. MODEL FORMULATION How then do we go about formulating a proc ess design model? The calculations normally involve six distinct steps, once the desired deci sion variables have been chosen and the neces sary objective and constraint functions have been identified. o Stoichiometry. The heat and material balanc es for all major pieces of equipment are normally involved in any process design. According to the problem, it may be desirable to make the material balance on a mole, weight, or volume basis. In some cases, for example in certain types of refinery problems, it may be possible or n eces sary to treat the stream in total. But, normally, at least a nominal set of components should be identified and separately balanced. CHEMICAL ENGINEERING EDUCATION

PAGE 33

Chemistry of Conversion. For conversion processes, some sort of chemical description is needed. In a few situations, chemical equilibria may be adequate. Gen erally, however, at least in some simplified form, there must be a treatment of the chemical kinetics. Which ever treatment is used, the equation should apply through out the region of interest. If necessary, new constraints should be formulated to eliminate areas of ignorance with respect to the chemistry. If these constraints are significant at the optimum point, it may be desirable to do further developmental analysis work in order to am plify the kinetics, or the chemical equilibrium. o Thermodynamics of Separation. For a separation process or the separation units in a larger process, some representation is needed of the thermodynamics of the separation, the phase equilibria and volume and enthalpy changes. Equipment Sizing. To relate the stoichiometry, chem istry and thermodynamics of the process to the actual process plant, equipment sizing calculations are needed. In some cases, these will be nothing more than arbitrary rules; in others they will be capacity or performance correlations. This step is often the most uncertain, and may require formulation and reformulation as the region of interest is identified. Capital Cost Estimates. Once the equipment is sized, a cost estimate must be developed for it. Usually the detailed cost estimating methods suitable for a final con tract bid are not necessary. Instead we need approximate methods that show how costs vary with small changes in equipment size. Happel's book 1 5 contains many useful tables and equations of the type needed. Economics and Accounting. The accounting equations must be combined with capital cost amortization to prod uce an economic balance for the process. Normally the objective takes the form of some rate-of-return, payoff time, or the like. Happel's book 1 5 describes some of the mathematics involved, but the best review of the proper economic objectives is given by Souders. 30 SOURCES FOR EQUATIONS To obtain the desired model equations we can use several different sources. First (and often overlooked) are the definitions, e.g., the molal concentration of a component, the average en thalpy of a stream, etc. Then there are first principles, like the conservation of mass, the laws of thermodynamics, the formulation of economic objectives and the like. The mass of chemical engineering knowledge usually appears in the form of established correlations, such as the de pendence of the Fanning friction factor on Rey nolds number and pipe roughness, the capacity of packed bed contactors, the heat transfer coef ficients in contact with a fluid bed, and many, many more. Finally, if general correlations can not be found with sufficient reliability to describe the desired application, we must have recourse to experiment. If the optimization study is diSUMMER 1969 With (the many methods available), further extensive work on . optimization algorithms does not seem worthwhile for the engineer. rectly involved in the guidance of development work, we must keep the experiment to the mini mum necessary to satisfy the desired goal of de sign optimization. At the same time, we must keep in mind what the outcome of a successful calculation might be, a detailed design of the final plant. If this design will be called for imme diately upon obtaining a satisfactory result from the evaluation process, then a short cut in experi ments may lead to slowing down design. SIMPLIFYING THE MODEL The optimizer must keep in mind that his mathematical model should be as simple as is consistent with the problem. At times, a very complex model is needed, when a very precise answer is desired and justified. But many process evaluations and most preliminary process designs can be done with a greatly simplified model. It must be remembered that the final design ob tained from an optimization consists of a set of design variables, the best possible set. Once these have been identified, it is possible to pro duce a much more refined design which will pro vide all the necessary engineering detail as well as checking the estimated objective function and the specified constraints. Here are a few points which can be checked to see whether a model satisfies this goal of simplification: Use estimates or "average values" whenever these are adequate for the purposes of the problem. Just because an engineering correlation exists for a given piece of equipment does not justify inclusion of this correlation in a model. For example, detailed con-elation of heat transfer coefficients versus heat exchanger design para meters and throughput is merely wasted, if the heat exchanger does not play a crucial economic role in the overall design. Here, it is much better to use merely an average heat transfer coefficient, estimated from good practice, and a simple cost estimation as a function of the square feet of exchange surface required. Another example is the use of approximate over-all absorption factor equations to represent performance of an absorber peripheral to the main process, in preference to a detailed tray-to-tray calculation, even if the latter is readily available, along with the necessary vapor-liquid equilbria to permit its use. Group like components in the material balance. Most chemical processes contain enough chemicals to make the identification and separate calculation of all components difficult. Unless their separation is crucial to the process, 135

PAGE 34

isomers and other lik e groups of components should be treated as single components. Nearly all chemical engi neering correlations of separation equipment involve summations over the components, so the number of com ponents identified should be minimized. This applies also to kinetic models. Here, each additional component in volves at least one additional kinetic constant, and often several. The kinetic model for the process study is merely a representation of the kinetics. It is not a true scientific explanation of the chemistry. This should be kept in mind while developing the model. If necessary, additional constraints can be added to limit the region of applica bility of a given equation, and insure that undesirable extrapolation does not occur Use a good base case and consider marginal changes from this base case. In many cases this technique will lead to a much simpler model, since first or, at most, second-derivatives are all that need be included. At other times too broad-brush a treatment of marginal affects around the base case may completely vitiate the study Finding the proper balance is part of the business of being a good engineer Correlate results of detailed stu dy of units or sub units of the main problem. This ties in with the base case method discussed above. Even when the base case ap proach is not valid for the problem as a whole, it may be used for certain parts of the problem For example, in the treatment of an oil refinery, it may be possible to describe the performance of the gas -r ecovery unit for the catalytic cracker in terms of a few crucial composi tion variables and certain major decisions as to recovery of key components Then a series of detailed tray-to-tray calculations could be used and correlated to predict expected costs and predicted separation performance, in terms of the key variables. Use the simplest acceptable equations to describe directly-related experiments. If the optimization is being used to guide development work, there will be directly related experiments which can be used to update the model. In a sense, these experiments appear, to the model, just like the results of separate studies described above. Elaborate analysis of the experimental data is only neces sary if extrapolation is essential in order to produce the desired optimization. In most cases, a simple response surface-type equation may serve the purpose. However, some thought should be given to choosing the right form of the variables; for example, a logarithmic variable should be used wherever these are more significant physi cally than arithmetic ones. However, the statistical significance of the experiments is rarely sharp enough to allow greater than a second-order response surface. USE OF PROBLEM STRUCTURE IN OPTIMIZATION Optimization problems can often be handled more easily by taking advantage of the structure of the problem. In the first place, the mathe matical form of the resulting equations may play a significant role in terms of the ease of optimiza tion. If these equations are all linear or can be linearized without excessive distortion, the pow erful techniques of linear programming can be 136 used. If they are non-linear but continuous or have, at most, a few discontinuities, non-linear programming or optimum-seeking algorithms may be used. However, if these discontinuities are extensive or if the discrete nature of some decision variables must be considered, then the problem becomes much more difficult; in princi ple, it requires use of integer programming, where perfectly general methods for large prob lems are not yet available. On top of this mathematical structure is the logical structure of the problem. Many linear problems fit into the so-called transportation model, which corresponds to the problem of find ing the minimal-cost policy to supply a number of demands at varying locations, by a number of different factories with different capacity limits. This is one of a number of network problems considered, among others, by Ford and Fulker son. 8 Some integer problems can be fit into the travelling-salesman or knapsack forms, for which general methods of approach exist. Many process problems occur in staged or cyclic form, and can be subdivided by techniques described by Rudd and Watson 28 and by Aris, Nemhauser and Wilde.2 Finally, problems that are sequential or repetitive in nature, such as the multi-period planning problem, can often be formulated, and sometimes optimized, in a way that takes advan tage of this repetitive structure. The Partition Programming algorithm described by Rosen 2 1 and used by Ornea and Eldredge 24 is readily adapted to this sequential problem, or to the natural partitioning which occurs in large-scale scheduling problems. CHEOPS A CHemical Engineering OPtimization System In an earlier paper 1 8 we described a system of programs termed CHEOPS, which takes advan tage of the general structure of process prob lems. Fig. 7, taken from this paper, shows how a modern refinery falls into units with informa tion flow between units confined to the process streams connecting the units. Even if a detailed design of one of the units is considered, this same sub-division is possible. Fig. 8 shows what happens to the vacuum flasher unit if the individ ual pieces of the unit are considered as separate process units. We identify the furnace-cyclone combination, the secondary-deentrainment sec tion, the pitch cooling section, the heavy and light flashed distillate condensations, and, finally, the CHEMICAL ENGINEERING EDUCATION

PAGE 35

FUEL GAS PU R CHA S ED c, -.----1 --c ,-.-------..-.----i CR UDE OI L C RU DE 1---~ SR GASO. PL ATFO RME R FEED S PE CI AL 1--~P RO D UCTS c P OL Y G ASO nC '---~-.....-ALKYL ATE C C GASO H, c,N APHTH A ETC D I S TILLAT ION l------l -------1 = ::_ ::_ -:_::_:: PREF T O PS PLATF O RMER PLATFORMATE S RR HGO X H GO GAS + GASO V AC U U M TFD C A TA L YTIC FL A SH I N G --~--~ ... C RACKING DA OIL P IT CH CC NA P H T HA HY D RO. D ESU LFU R IZATION H Y DR O D IS T IL LATES PREF BOTT S DE AS PHALTIN G '--~ CCHGO '--c C XHGO AS P HA LT Figure 7 Typical Refinery Process Scheme. Blending of final products not shown. steam ejector and gas system Within each of these units or sub-units the s ame repetitive cal culations exist. The s e can be separated into five steps: (1) the setup; (2) material balance; (3) he a t balance; ( 4) c onstraint calculations; and (5) cost estimate and profitability. Fig. 9 indicates how CHEOPS is structured to handle these con s ecutive calculations using a set of individual unit s ub-routines which describe the units in the actual process. CHEOPS will operate with any of a number of optimization algorithms, as long as the y are structured in the form given in Eq. 1. B y following a few simple SRR ; ( ,------._ i~-~ ,,, ---,., : \ I I .~-r ) : I I \ I I I -..: : i : 'F'F-~ '-r---;..--~ "'. -.._ _ CRUDE_f_OR HX _ '\ _/ PITCH -t~REAKDOWN INTO SUB-UNITS Figure 8 SUMMER 1969 GAS / SW rules, the s ub-routines for each unit can be struc tured to supply the necessary answers to each part of CHEOPS. Further details on this appear in Tabl e 1. Thi s table and Figs. 7 9 are all taken from our earlier publication. 1 8 LIMITATIONS ON PROBLEM SIZE Precise formulation of the size of problems which can be conveniently or economically han dled b y optimization is very difficult. For one thing such limitations depend greatly on the complexity of the s imulation. If extremely de tailed, complex s olutions are desired, this will lead to extensive sub-programs just to describe the engineering. If s uch programs become too large, they will exceed the available core in the computer, which means that one must go to multi-coreloads with all the attendant bookkeepICHEoPS--1 I I I I I I Figure 9 137

PAGE 36

Table 1 Program Functions for CHEOPS. Section Initializati<>!l Optimization P r og r am Init-ialfz e s .optimization algo r ithm Material and sets new va l ues of -d ecision heat balance variab l es, and c-e.lls for p r eliminary ca l cu l ations at a n ew point Const r aint fun ctio ns Objective function Output Sets index identifying constraint function to be ca l cu l ated, and ca ll s for constraint" calculation Ca ll s for objective function ca l culation Ca ll s f o r output, identified a s n on feasib l e, intermediate feasib l e optima l, or p r ogram e rr o r CHEOPS Program 1. J:Dads data, wllkh may l n c lu de: a) Contro l for optimization b) Cont r o l fd r mate r ial ba l ancing c) Contro l fot objective funct i on calcul a't ion d) Cont r o l for out p ut e) Gene r a l ~conomic and Cost dat a : f) Overhead cost pa r ameters g) Off site ca p i. tal parameters h) Utilit/ cost parameters i) .Tankage paramete r s j) Prices for supplies and materials k) P l ant feed amount,;, prices, and pro perti es J.) Plant pr oduct amounts prices, and p roperties m), Component pI' o perties 11) F l ow-diagram connections o) Optimi2ation ~ va r iab l e identificatib n, b ounds and si;_a rting values p) Constraint & equality identificatio n and t o leran ces, q) Parame te rs for UN I T subrouti!les 2. Sets indices in'. UNIT subroutines 3 Prints record of data and derived value s 4 Sets contro l s for remainiilg ca l cu l ation s 1 Ca ll s UN I '!',. suliroutines fo r prelimina ry calculatiOns 2 Ca ll s UN I T subroutines, checks. flo w s cha r acteristics, and pro p erties of recycle streams, makes adjustments and r e-ca1.ls UNIT s uOroutines until lllOdel is material arid heat balanced 3 Ca ll s UN I T subroutines for utility deman ds aild ot her functions ( returns to l/2 if utility balance ca l cu l ation affects process streams ) I dentifies unit number for Constr~int, and cal l s a pp ropriate UNIT sub routine l Ca ~l s UNIT subroutines i n o r de r 2 c,.lculates, and totals unit operating and cal)i tel costs ). CalculatesraV:..material cosis, nd product credits, a s n eeded 4. Totals utility demands and use, and calcu l ates utility capital and operating costs 5 ,. Calculates indicated obje"ctive function, including tankage, ov.erhead, offsi tes etc Out p uts r esu lt s, as indicated by p r intou contro l data ta l lies, and objec t ive func t i on va l ue Types of out p ut va ri a bl e a r e: a) Va r iab l es and constraint and o b je ct i ve functions b) Proces s evalu~tion swrmary c) Capital cost b r ea)ldown d) Ut i lity 1 use, demand, and costs e ) Materials and su ppl ies sWTmBry f) Process st r eam mate r ial balance g) Pr ocess st:i-cam f l ows and p r ope rt ies h) Equipment details ( UNI T subroutines are called first to Calculate additiona l resu l ts if des ir ed) 1) UNIT printouts ( obtained by ca ll ing UNIT subroutines) UN I T Subro u tine l. Ca ll s I ND EX sub rou tin e of C HE O P S to permit setting i n d i ces 2 Makes p r e li mina ry ca l cu l ations independent o f decision va riabl es ( except fo r a p oss ibl e d ependen c e on t h ei r s t a rt i ng val u es) 1. Makes p re li minary ca l cu l ations \lhich de p en~ on dec i s i o n variab l es but not on s tr eam flows, cha r acte r is ti cs o r pr operties 2 _. C alcu l ates flovs, c harac t eri s t ics and pro pertie s of 11 a sk ed 11 s treams (no rmally, the streams leavini: the unit) 3 Calcu la t e s u t i li ty demands and functions to be u s e d in bo tll constraint and o b j ecti ve ca l culatio n s Ca l cu l a t es con s traint !un ction Ca l cu l ates the f o ll o wing cost contributions: a) Sup p ly a n d matertal use b) Su pp ly and ma t er ia l i n y ent o cy c) Operating labor d) Repai r and ma in t ena nc e cos t s e) Ca p ita l cos t ( with u s eful life, a n d t a x de prec ia tion life \i h e n a ppli c abl e ) 1. If ca ll ed fo r e q ui pmen t pr i nto ut, ca l culates detai l s not re q u i r ed f or cost ca l cu l atio n s but o f i n t er est a t po i nt se l ected 2 If ca ll ed fo r UNIT print-out, o utp u t s r esults as pr og ramned ing problems. Hopefully, this situation will be much improved by the third generation of com puters, once the necessary systems are fully checked out. In the meantime for the second generation computer (IBM 7094 and the like), we supply the general guidelines given in Tabl e 2 These indicate the size of usable problems in term of the number of variables and constraints which can be conveniently handled by the opti mization algorithms indicated. It is assumed that the equations relating these functions and vari ables are of no more than ordinary complexity. The algorithms considered are as follows : LP Several standard linear programming systems 138 CHEMICAL ENGINEERING EDUCATION

PAGE 37

Table 2. Usable Problem Size for Optimization Algorithms. LP Map GP PPNL DA Type of Objective F L NL NL NL NL Decision Variables Linear 104 500 100 600 40 Non-Linear 100 60 Bounds } } 200 120 80 Constraints 2000 1000} 200 } 2000 20 Equations No more th a n two a c t i v e at a ny one t i me can handle a practically-unlimited number of decision variables, and over 2,000 rows; the latter can represent bounds on the variables, constraints on combinations of variables, or equalities involving one or more variables. Note that all functions have to be linear, although it is possible to do a reasonably good job of representing a f e w non-linear relationships by defining new variables and constraints. MAP The technique described by Griffith and Stewart1 2 has been used very satisfactorily for problems with about 500 linear variabl e s and 100 non-linear ones. The number of rows which can b e conveniently handled is somewhat less than in LP, because the necessary s t e size limitations on the non-linear variables add additional rows. However, the allowance of non-linearity permits practically as good a representation of the system with the smaller MAP formulation as with the large LP version. o GP or Gradient ;Frojection. 25 The usual form of this algorithm (available through SHARE) can only handle linear constraints or equalities, but will handle a non-linear objective function Althou g h the limitation on constraints and bounds is fai r ly small in the usual programs, there is nothing inherent in the algorithm which r equires this; it merely r e pr e sents a balanc e betw e en available core on the IBM 7094, and the c om plexity allowed for the non-linea r system. PPNL the non-linear version of Partition Pro gramming described by O r nea and Eldredge. 24 This can handle much larger systems and can handle up to 60 truly non-linear variables, each of which can have bounds. The approximate size for the total number of linear variables is 600 and the total number of r ows in th e linear system is about 2000. Normally this system i s used where the linear problem can be partitioned still further into smaller sub-problems, ea c h of which i s handled individually (but automati'cally) during the optimization. DA Deflected Assent 29 is typical of many hill climbing methods. It is ve r y limited as to number of variables. However, it can handle extremely non-linear or discontinuous objective function s and can satisfac torily treat non-linear constraints, as long as no more than two are active at any one time. A further advantage of this system and of others similar to it is the exereme compactness of the program itself Use of DA may permit a much larger simulation model without requiring multi-core use. On the other hand, if th e problem b e comes nearly linear at the optimum, Deflected Assent may behave very poorly SUMMER 1969 Much more significant than the actual computation time ... is the program development time. Nothing is included in this table about in teger or discrete variables, since most algorithms to handle such variables are still in the experi mental stage. The only general method that is really foolproof as yet is that of combinatorially going through all possible cases. This is quite feasible, once it is set up automatically on a computer, as long as the total number of cases is not more than a few hundred. However, for each case, it may be necessary to do at least a brief optimization of the continuous variables. TIME LIMITATIONS Other limitations on the usable size of prob lems arise from the various time requirements. First, of course is computation time; this is not really too serious as long as the above size limits are met. Moreover, if the problem is important enough, whatever its size, it is possible to run it, at least in a stepwise fashion. With a standard method of computer center organization, how ever, it is usually best to stick to runs between a few minutes and a few hours in length. Much more significant than the actual computation time for optimization is the program develop ment time. With CHEOPS, it is possible to de velop a useful process model, for example, a model for optimizing the design of a chemical plant requiring several million dollars of invested capital, in a matter of 4 to 6 weeks. But a com plex of several plants or an oil refinery might require months and months of effort on the part of several programmers, even if they are given easy access to the computer. Still another type of time limitation concerns the acquisition of necessary data, i.e., the values of the parameters. Even if the model is being developed entirely on the basis of past information, the acquisition, assembling, and proper checking might take nearly as much effort as the program develop ment. Finally, the processing of these data for new optimization runs is, in itself, a major task. For routine use of optimization models in analyz ing different cases, and different sets of premises, or in routing scheduling operations of multi-plant or multi-refinery systems it is essential that an organization be set up with a full recognition of the data processing aspects. 139

PAGE 38

UNCERTAINTY AND ITS EFFECTS In the above discussion we have been imply ing that the entire problem is deterministic; that is, that we can make decisions, once and for all, based on certain premises which are completely fixed, and come up with a true optimum value which will always hold. In many engineering problems, this is a close approximation to the true situation. However, as we begin to get more and more involved with the marketing and business aspects, we often get into situations where uncertainties dominate the effect. These uncertainties can be classified into four general types: 1. The Form of the Model. Under this heading comes the question of identifying the proper kinetics, choosing the right equations to represent vapor-liquid equilibria, selecting the right correlations for sizing equipment, and so forth. The only real cure for this uncertainly is to get the best model builders you can. The implication is that building the model is an engineer's job, not a mathe matician's. Model-builders must be people with engi neering judgment and, preferably, with some physical feel for the system. 2. Accuracy of Data. Once the form of the equations ha s been settled, we must estimate or select numerical values for the various parameters in the system Some of these may be relatively precise, but for others, there may be distribution curves of values (probability dis tributions) about the average values Where such in accuracies become significant, we must use simulation or stochastic programming, as described below. 3. Forecasts of Future Conditions. Nearly all useful problems involve future b e havior of the weather, cus tomers, competitors, and so forth. Mechanically, we can handle these forecasts with the same techniques with which we handle uncertainties in data. However, philo sophically, they are different in type; we are assuming what the future looks like. In reality, factors that do not enter into our equations may come to bear before we actually get the desired results. 4. Responsive Actions of Others In some cases even the forecasts are inadequate. In a highly competitive situation, where we are dealing with one or two com petitors, one or two major marketers for our supplies, or one or two major customers for our product, it i s very dangerous to forecast future action. If our optimiza tion is to be of any value, we would expect to do some thing new, i.e., something we have never done before. Except in very limited situations, it is impossible to forecast what the response of a major competitor, sup plier, or customer is apt to be to this new action For this reason, we have studied the use of game-theory in analyzing such problems, 1 7 but with only limited success to date OTHER APPROACHES TO MATHEMATICAL MODELLING Because these uncertainties dominate many problems, an alternative technique has developed, 140 called Monte-Carlo Modelling or Simulation. In reality, this merely handles uncertainties of types 2 and 3, which can be represented by replacing each uncertain parameter by a probability dis tribution for the parameter values. With these distributions as guides, we select enough random cases to produce a truly average value, or pref erably, a distribution of values for the desired objective. Many references describe this proce dure. The text of Naylor, et.al., 23 is one of the newest general references, while that of Franks 9 concentrates on chemical engineering problems. Unfortunately, use of these techniques gen erally makes optimization unwieldly or even im possible. The usual procedure is to use case studies, where a few important variables are set at two or three values. Fortunately, in many problems where this approach is essential, opti mization is relatively unimportant, precise selec tion of optimum variable values is prevented by the uncertainties. This is certainly true of the "venture analysis" of Andersen 1 or the "risk analysis" of Hertz. 1 6 However, other problems could profit by a combined approach, which has been termed stochastic programming. 20 21 One pattern for this stochastic programming is sketched in Fig. 10. The basic simulation of Figs. 1 and 2 appears at the left in Fig. 10. However, the parameters Distr i butions for Parameters I I ,-----i-----~ ' : C a se S electi o n : ( M o nte Carlo o r : ,,' I oth e r Technique : ,' '--7--------~-.J ,,,' ',, r-""---....L----, ,-_,,._-L./~ .. _____ _, Expected V a lue a nd D is tr i bution for Objecti v e Function ', ',, ... -~'~'-------------Jf_,, Optimiz i ng Algorithm i ~-----------------------~ Figure 10. are no longer fixed, but selected by some Monte-Carlo technique, from the given probability distributions. And the optimizing algorithm responds, not to individual values of the objective function, but to the expected value, or some other property of the calculated probability dis tribution for the objective function Although many re searchers are studying this and other formulations of stochastic programming, useful, general, computation systems are not available Until they are, the engineer must select his tool according to the problem, optimiza tion for deterministic well-defined, many-decision prob lems, and stochastic simulation for highly-uncertain, few decision problems. (Continued on page 158.) CHEMICAL ENGINEERING EDUCATION

PAGE 39

You won't just get your feet wet. Standard Oil Company of California offers all the experience you can soak up You ll start out facing practical situations and using your academic knowl edge and skills to solve real problems You may even have to improvise and develop new approaches to specific questions. We rotate the assign ments of young professionals. You will be able to work with different groups of experienced colleagues and sharpen your skills on a variety of projects Talk with our representative when he comes to your campus about the opportunities we have for you Check your placement office for more information or write to: D. C. Reid, Coordi nator Professional Employ ment Standard Oil Company of California 225 Bush Street -Room 105 San Francisco California 94120. Standard Oil Company of California An Equal Opportunity Employer

PAGE 40

tin :I problems for teachers I 1. Submitt e d by Professor R. B. Bird, University of Wisconsin. Hydrostatic Pressure Distribution in In compressible Fluids. Consider a beaker of liquid which, for all practical purposes, can be consid ered to be incompressible; let its density be P o It is desired to obtain an expression for the pres sure in the liquid as a function of position. Take the origin of coordinates to be at the liquid-air interface, with the positive z-axis pointing away from the liquid; let the pressure at the liquid-air interface be p ( 0). A friend comes to you with the following comments: I. "By simplifying the equation of motion for an incompressible fluid at rest, I get 0 = dp / dz p 0 g; I can solve this and get p = p(0) p 0 gz. That seems reasonable the pressure increases as one goes deeper and deeper into the liquid." II. "But, on the other hand, the equation of state of any fluid is p = p (p,T). If the system is isother mal,, then p = p(p). If, furthermore, the fluid is incompressible p = p(p 0 ) = constant. This tells me that the pressure is constant throughout the field which I don't believe!" Clearly your friends needs help. Explain. 2. Submitted by Professor Dave Chittend en, Uni versity of New Hampshire. Computer Solution for the Adiabatic Flame Temperature. Find the adiabatic flame tempera ture for combustion of the following natural gas mixture: CH 4 86.6 % ; C 2 H s, 7.9 % ; C 3 H s 2.7 % ; C H 10 1.3 % ; N 2 1.5 % This dry gas is mixed with 130 % theoretical air which contains 0.043 lb H 2 0 / lb dry air. The gas-air mixture enters the burner at 500 K and 4 atmospheres pressure. Dis sociation of water and carbon dioxide in the flue gases must be considered. Solution: To do this type of computation, a general computer program in Fortran IV-G has been developed and tested on an IBM 360 Model 40 computer. To use the program, one keypunches a few IBM cards containing a description of the problem, and then the data cards are submitted along with a machine language program deck for proces sing. The data needed for the calculation of the adiabatic flame temperature are as follows : 1. Number of hydrocarbon species in the gas mixture being burned. 2. Pressure in the burner. 3. Percent theoretical air (Must be equal to or greater than 100 % .) 4. Absolute humidity of the incoming air. 5. Number of moles of each hydrocarbon and of nitro gen in the fuel stream. 6. Heats of combustion at 291 K for all of hydrocar bon gases burned. (The heats to be used assume that gaseous water is formed in standard state reaction.) 7. Specific heats of all components in the range be tween 291 K and the flame temperature as a func tion of temperature. 8. Inlet reactant temperatures, which may be differ ent for each reactant. When the adiabatic flame temperature is calculated, a number of assumptions are made : 1. The process of combustion is adiabatic. 2. There is no secondary air. The hydrocarbons are completely converted using only the primary air. 3. The only combustion products are carbon dioxide, carbon monoxide, water, hydrogen, oxygen and nitrogen. These products are at th e! equilibrium conversions controlled by the equilibrium constants for the reactions 2H 2 + 0 2 = 2H 2 0 H ~ + CO 2 = H 2 O + CO THEORETICAL FLAME TEMPERATURE IOlAl GENERAL CASE FOR NATURAL GAS MIXTURE 142 PRESSURE 4.00 ATMOSPHERES PERCENT THEORETICAL AIR 130.00 THE FORMULA CH4 C2H6 C3H8 C4Hl0 NZ 4 HYDROCARBON( Sl DATA DESCRlBING THE HEAT OF COMB. -0.1 89 700E 06 -0.336732E 06 -0.484100E 06 -0.630620E 06 CAL /GM MOLE REACTING E N T ,'10 LES 0.866 0.079 0.021 0.013 0.015 0.04300LB HZO PER LB DRY AIR HYDROCARBONS JS TABULATED BELOW MO LES C MC1LE S H2 ENT TEMP 1.000 2.000 500.000 2 .000 3.000 500 .000 3.000 4.000 500.000 4.000 5.000 500.000 DEGREES KELVIN FORMULA OF HYDROCARBON OR EQUIVALENT HYDROCARBON C 1 15 7 H 4. 2 84 ENTERING MIXTURE BASED ON O NE MOLE ENTERING HYDROCA RBON OR HYDROCAR B ON MIXTURE MOLES OXYGEN 2.8964 MOLES NITROGEN 10. 9109 MOLES WATER l.0258 PERCENT COMBUSTJBLE GAS BY VOLU M E = 6.32 ] Input Vata ] Input Vata CHEMICAL ENGINEERING EDUCATION

PAGE 41

GENERAL FOR M OF THE SPECIF{C HEAT EQUATION CP = A + B*T + C*ITl**2 + D*(Tl**3 CONS T ANTS fN THE SPECIFIC HEAT E Q UATIO N S F OR THE REACTING CP UNITS AR E CAL / GMMOLES~DEG~ KELV{N HYDROCARBONS. B C D C H 4 C2H6 C3H8 C4Hl0 A 0 47500E 01 0 94400E 00 -0 96600E 00 0. 94500E 00 0 30000E 02 0.33630E-06 -0.1645 0E 09 0.422DOE-0 8 0.75 B O O E08 0 83600E-08 ] 0 37350E-Ol -D .1 9930E-04 Input Va.ta. 0 72790E Ol -0 37750E-04 0 88730E Ol -D.43800 E -04 THE DATA DESCR!Bf NG ALL HEAT F D 1ss n c E N T. TEM P C A L/ G MMO L E KELVI N HZO 0 57830E 05 50 0 0 0 CO2 0 679 6 0E 05 500.00 HZ o o 500.00 02 o o 500.0 0 N2 o o 500 00 co o.o 500 00 0 THER CO MPOUNDS IS A O 8 32 OOOE 0 1 O 77000 0E Ol o 664 000[ 01 0 673000E 01 0 6 73 OOOE 01 O. 6 73 OOOE 01 TA BU LATE D BE L OW i3 -0 653DOOE -03 0.530000E 02 0 4920DOE-03 0 40 8 0 00E -03 0.408000E-03 0.408000E-03 C 0.2700 00E 05 D 8 3 0000E 06 0 3 1 'l00Dc 06 0.4 8600 DE-D 6 0 4 86000E -0 6 0 486000E-06 I) D 6 14 500 E 09 o o 0 74 000DE tO 0 12"3',0DE 09 -0 l 2 34 00 1=09 -o 12340 0E09 TOTAL MOLES IN T HE EQUILIBRIUM M IXTURE 15.90913 FI N AL MIXTURE CO M POSITION, MOLE PERCE N T CO2 7. 2 2799 H20 19.89369 co 0 04454 H2 0 0 1 823 PERCE N T DISSOCI A TION OF CA RBON DIOXIDE 0 61 PERCEN T O!SSOC!ATION OF WATER= 0 09 A C TUAL K = 0.264E 04 COMPOSITION K 0 265E 04 02 4.23272 N2 6 8 58292 TH EO RE T !CAL FLAME TEMPERATURE = 2016.619 DEGREES KELVI N 4 The fuel contains only hydrocarbons and nitrogen. The air contains exactly 79 mole o/o N 2 and 21 mole o/o 0 2 5 The properties of the reactant and of the product gases are calculated as mixtur e s of ideal gases 6. The flame temperature must be between 2500 K and 1900 K. This range includes the flame tem peratures of all major hydrocarbons likely to be in a fuel gas The accuracy of the results obtained with this pro gram is limited on l y by the validity of the above assump tions and the accuracy of the input data The program user should try to insure that his specific heat data is LETTERS J Continued from p 109 ) For the 40 graduate-inclined schools P = 4 07 + 0.00B + 0 17M + 0 51D ( 4) and for the 57 undergraduate schools P = 2.56 + 0 llB + 0.08M + 0 .2 7D (5) One immediately notices that in Eq 4 the coefficient of B is zero. This result may not be as outlandish as it appears at first blush It indicates that in graduate oriented programs the number of undergraduates may not appreciably affect the number of full-time profes sorial schedules. One may speculate that this is the result of a growing and perhaps r egre ttable practice of relegating undergraduate instruction to persons without professorial rank; i.e instructors and graduate assist ants This practice may be dictated by necessity in rapid-growth situations. If one compares th e c o e fficients of D in Eqs. 4 and 5, one notices that graduate-inclined schools require mor e SUMMER 1969 valid for the temperature range of th e problem he i s doing. The results of computation for the above problem are shown in the figure The input data read from cards is printed on the output as indicated. The other results are calculated by the program except for the specific heats of the output g as es which are stored in th e program. Further information about th e program can be ob tained from Dr David H. Chittenden, ChE Department, University of New Hampshire, Durham, New Hampshir e 03824. professional time per doctoral degree granted than do undergraduate-inclined schools. The situation is reversed in the case of bachelor's degrees In conclusion, we wish to re mind the reader that no great accuracy is claimed fo r this s tudy. It represents a first attempt to analyz e the relationship b e tw ee n the number of full-time profe ss orial schedules and the num ber and kinds of degrees granted. Considering the nature of the variables, it is ind ee d surprising that the indicated d egree s of correlation and stability exist It would b e interesting to follow the study with future ones, not only in chemical engineering but in oth er disciplines as well. A. X. Schmidt Robert Pf e ff er Leona r d Cohen Th e City College of the City University of New York 143

PAGE 42

THE DILEMMA OF INNOVATING SOCIETIES FREDERICKSON (Continued from page 125.) internal factors. Clearly, as a society we are sen sitive to and react to factors that arise outside of our geographical boundaries. The cold war, Vietnam, and the ABM are not strictly of our own doing. However, here we want to consider only those dynamics-determining factors that can be considered to be internal, since it seems that these are of greater significance than factors that can be definitely identified as external. The first of these internal factors is the existence and propagation of what might be called the Creed of Technology. This is the widely held belief mentioned previously that innovation p er s e is good and that technology is always pro gressive. It might be argued that no intelligent, well-educated person really believes the Creed of Technology. If that is so, why do so many intelligent, well-educated persons act as if they did believe it? And if such is true of intelligent, well-educated persons, what beliefs are held by simple or uneducated persons in this regard? By questioning the belief that innovation is always good, we are not thereby saying that it is always bad. The point is simply that if a technological innovation has a good side (as it almost always does), it will more than likely have a bad side as well. A case in point are those technological innovations that have led to large scale agricultural use of fertilizers, herbicides, and insecticides. Surely these have greatly in creased yields of crops and made possible a minimal standard of living for the world's ex panding population. There is also no question that such use has led to progressive and cumula tive deterioration and pollution of important sec tors of our environment. Again it is certain that these innovations have relieved the farmer of a number of burdensome chores, while on the other hand, it is possible to wonder if the farmer is any happier for being so relieved. Associated with the Creed of Technology is the Cult of the Product. If the former teaches that innovation of itself is a positive good, the latter is the logical consequence of belief in that teaching If innovation of itself is good, then the products of innovation be they good or serv ices must themselves be good and those who deny this are either foolish or wholly mistaken. "For," claim the votaries of this cult, "our pro duction of this thing or this service puts people 144 to work; it gives them useful labor in recompense for which they can obtain the good things of life. And our product itself satisfies a demand otherwise we would hardly be making it. There fore, it is patent on all counts that our product is a good thing, and deserves the importance that we attach to it." Yes, but if your production pollutes the air that society must breathe or the water that it must drink, shall humanity at large share your enthusiasm for your cult idol? If this production necessitates tearing down and destroying the things of natural or man-made beauty that enrich life so much shall the next generation of men hold you blameless? Where in the Cult of the Product is there any sense of balanc e ? A third factor, with roots perhaps deeper in human nature than any other, is the Gospel of Growth. This Good News is this: That the ex pansion of human activities (that is, the produc tion and consumption of goods and services) institutions, and population is a wonderful thing, a panacea for economic and even social ills; without growth, society stagnates With it, so ciety's possibilities are unlimited. Hence, indus trial or commercial enterprises are quick to brand themselves as "growth companies cham bers of commerce describe the community that they represent as a "vigorous growing area," and radio and television advertising puts out an incessant barrage of propaganda designed to convince the public of the sovereign merits of industrial and economic growth Placed before us are bright visions of a high-energy society, where man will control the weather or live under a weather-excluding dome, the grass will not have to be cut because it is artificial, the women will always be slender, young, and beautiful, and each of the kids can have a snowmobile for the winter, a sports car for fall and spring, and a speedboat for the summer. Or sometimes a dif ferent approach is taken; "If you don't grow you're dead!" is the hellfire-and-brimstone way to put it. And it is preached that way, appar ently with complete conviction that it applied to human institutions and societies as it does to the biological cycle of growth development, senes cense, and death exhibited by individual living organisms. This sort of preaching has won con verts, and it is difficult to escape the conclusion CHEMICAL ENGINEERING EDUCATION

PAGE 43

. . death and decay are necessary for birth and growth, and great Caesar's dust may appear in an ear of wheat as in a bunghole stopper. that for a substantial portion of the most influen tial elements of our society, growth has become an end in itself. The biological aspect of the Gospel of Growth, viz., that population growth is an inherently good thing, is no longer a very explicit part of the Gospel's teaching. The social stresses and political and other problems brought about by overpopulation are so evident in so many parts of the world that it is no longer fashionable or expedient to come right out and urge population growth. However, the teaching that population growth is inherently good is implicit in the Gospel of Growth. For if expansion of the production of goods and services is a good thing, then population growth must also be welcome, for besides increasing per capita consumption, what other way is there of increasing total consumption, the sine qua non for increasing production? I F, AS SUGGESTED ABOVE, THE HELL fire-and-brimstone aspect of the Gospel of Growth is based on the invalid analogy that the development of a society follows the same rules as the development of individual living organ isms, it is nevertheless permissible and even necessary to view the development of societies in biological terms; not in the sense of trying to draw analogies, since these are always of doubt ful validity, but in the sense of trying to see what restrictions biology places on such development. For instance, American society is composed of a biological population inhabiting a large but strictly bounded living space. Very many other biological populations, ranging from bacteria to beef cattle, share this living space with us. All of these populations man's included are subject to biological, chemical, and physical laws that cannot be suspended or amended by majority votes of our legislatures. In that sense, man can never "master" nature. In many cases, these laws are not all known with precision or else their application is too complicated for quantita tive prediction; in such cases, it is not possible to predict what consequences human actions will have on our environment. It is necessary to mention this here, since the contrary belief that we can predict everything is quite com mon in our society, and is the essence of what we shall call Technological Megalomania ( see be low). Moreover, even if population growth is not an explicit teaching of the Gospel of Growth, the current policies SUMMER 1969 of our society guarantee that the population will continue to grow anyway. The demographer Judith Blake has recently pointed out that Americans of all classes con sider it desirable to have large (greater than two children) families: the mean number of children con sidered desirable varies from 3 2 for high-income and college-educated women to 3.6 or 3.7 for low-income or grade-school-educated women. If that is what is considered desirable, then that is what couples will strive to attain, and so in fact the natility rate is much greater than that required to achieve a stable population. Professor Blake attributes the desire for large fami lies to "pronatalist policies" of our society, policies which evolved primarily in response to the past need for a high natality rate to counteract the then-prevailing high mor tality rate. She says (pp. 528 and 529) that these policies "insure that just about everyone will be propelled into reproductive unions, and [the female] half of the popula tion will enter such unions as a 'career' a life's work. This rigid structuring of the wife-mother position builds into the entire motivational pattern of women's lives a tendency to want at least a moderate-size family. . the wish for a family of a particular size . relates ... to a need for more than one or two children if one is going to enjoy 'family life' over a significant portion of one's lifetime." However, the predictions based on these laws are quite clear even if not specific when they are applied to a population that is growing unchecked by any competing population, by disease, by in traspecific rivalries, or by lack of food: eventu ally one or more of these factors will exert them selves and bring about a check to population growth. It is true that by very great expenditure of effort an expenditure that we do not now seem willing to make the density of popula tion at which such limiting factors come into play can be raised. But how high can it be raised? That is one of the things that cannot be answered with precision Concerning this, the ecologist Lawrence Slobodkin writes* (pp. 3-4) : "How many men can the earth hold? We must abandon all pretense of saving intact any wilderness areas and consider that we will treat the earth as a combined garden and factory ; all other species will either prove useful to man or will be eliminated; they will either adjust to the omnipresence of man or die. Answers to the question are now merely guesses, ranging from 7 billion to 200 billion, the difference in the estimates depending on how several subsidiary questions are answered." And what would it be Judith Blake, "Population policy for Americans: Is the government being misled?" Sci ence, 164, 522-529 ( 1969). Lawrence B. Slobodkin, "Growth and Regulation of Animal Populations," (New York: Holt, Rinehart, and Winston, 1961). 145

PAGE 44

like to live in such a world? Again we quote Slobodkin (p. 4) : "Implicit in this picture of the future is a mental health problem: a world com pletely full of man and his activities could well be a maddening place. There is an esthetic prob lem: the beauty of the winderness is a very real thing. There is a political problem: a world full of men would be highly regimented, a world of an Aldous Huxley or Orwell phantasy." When measured against this sober (and understated) appraisal, the Gospel of Growth begins to lose some of its appeal. And what about that aspect of the Gospel calling for continual expansion of man's activities and man's consumption, and not just total expansion, either, but expansion per person? This compounds the problems raised by the population explosion and hastens the time when some limiting factor beyond man's control will appear. If our (U.S.) population were stationary, and if our per capita rate of consumption of goods and services were stationary, both at their present levels, then we could probably afford to indulge ourselves in the other follies enumerated here. It seems perfectly reasonable to assume that with existing technology, or with technol ogy within the reach of present knowledge, we could provide the necessities of life and many luxuries in addition, as well as a clean, esthetically-pleasing environ ment, for all of the 200,000,000 people who now live in the United States. But of course our population and our per capita consumption are not stationary, and these are facts of cardinal importance. The Gospel of Growth is at once the most insidious and the most dangerous of all the internal factors that will be considered here. It is the most insidious because its biological aspect is based on an eminently reasonable postulate: that we must reproduce ourselves if human life is to continue on the earth. It is the most dangerous because it interacts with and is reinforced by all the other factors that we are enumerating, and because by itself, it is a sufficient condition for producing ecological catastrophe. We need to take a long look at the Gospel of Growth. S TILL ANOTHER INTERNAL FACTOR that determines the dynamics of our society is that which gives rise to what my biomedical librarian friend calls the Sandbox Syndrome. She coined this term one day when we were try ing to get to downtown Minneapolis. That day it seemed that our path was diverted at every intersection by a huge hole dug in the earth for a freeway, by some enormous piece of earth moving machinery blocking the street, or for preparations to throw up a new skyscraper. After the tenth detour, she exclaimed "Why, this is the Sandbox Syndrome; dig, dig, dig all the time!" 146 the sandbox syndrome; dig, dig, dig ... In case there be any uncertainty about it, let us place in more explicit if less expressive terms that which Miss Bohn so aptly characterized: We shall say that an individual, an organization, or a society exhibits the Sandbox Syndrome if his or its thinking tends to be determined by technological considerations if it is fixed upon the size or the speed of machines or projects, or if it habitually turns to digging in the dirt, rear ranging the landscape, and generally "improving upon nature." A case that was a classic example of the Sandbox Syndrome appeared on TV a few weeks ago. An official of one of the large airline companies was describing the SST. Facts and figures concerning the weight, size, thrust, speed, carrying capacity, length of runway neces sary for takeoff, etc., came spewing out of him as paper comes spewing out of the University's CDC 6600 com pute r But never a word about sonic boom, relea s e of pollution high in the atmospher e where dispersal is very slow, or the consideration that it may not be necessary or even desirable to be able to fly from Minneapolis-St. Paul to London in four hours. The whole thing was reminiscent of a joke once told by a colleagu e : The pas sengers on the first commercial flight of the SST had fastened their seat belts in preparation for blast-off and the doors had been sealed. A voice came over the PA system and announced that "This is a recording. The doors of the craft have been sealed and the program for takeoff initiated. In two hours we shall touch down in London. This operation is completely computerized and is not subject to human error So relax, enjoy the flight, and be assured that nothing can go wrong nothing can go wrong nothing can go wrong .... The forms of the Sandbox Syndrome de scribed above are of a relatively primitive nature. More advanced forms are sometimes manifested and these should perhaps be differentiated from the Sandbox Syndrome; we might call these Technological Megalomania. Particularly strik ing here is a tendency to view the earth as some sort of spacecraft, with the earth's human popu lation as its crew and all the rest of the earth's populations as its life support system. The mis sion in which this super space vehicle is engaged is tremendously exciting, if rather ill-defined Hence, we must manage the life support system and harness it every organism of it to the one task of supporting the crew. And of course all of the crew must dedicate themselves to the great task of completing the mission. Possibly the crew might become somewhat restive in their cramped quarters (even though these be scien tifically designed), but no doubt application of psychology, genetic strain selection, and new inCHEMICAL ENGINEERING EDUCATION

PAGE 45

It is absolutely necessary . that innovating societies think about why they innovate and what the consequences of innovation are. novations in crowd control will minimize any disturbances. Another striking manifestation of Techno logical Megalomania is a willingness to play the Game of Environmental Russian Roulette. The elements of this Game are first a technological innovation magnificent in conception and gran diose in scope, second a considerable uncertainty about the long-term environmental consequences of implementing this conception but a possibility that these consequences could be serious indeed, and third a willingness to proceed with imple mentation anyway. A STRANGE OPHTHALMIC DEFECT THAT we can call Utilitarian Vision is still another factor that directs technological innovation into dangerous channels. Characteristic symptoms of this defect are: Looking at a forest and seeing piles of boards instead of trees, looking at an old but well-kept neighborhood and seeing high-rise apartment bulidings instead of homes, and look ing at an unspoiled river valley and seeking a power plant with a six hundred foot smokestack instead of a place where future generations of men can breathe and renew their spirit. Some times, the defect bcomes so severe that the boards, apartment buildings and power plant of the foregoing example s are distorted further into dollars. When that symptom shows up, there i s little that anyone can do to correct the defect And please do not assume that because Utilita rian Vision has been illustrated by homely exam ples that strike most closely at the author s heart that it does not operate on a much larger scale. Another internal factor of importance is the Concept of the Convenient Society. If some of the other factors that have been enumerated arose far back in the past or are inherent aspects of human behavior, the Concept of the Conveni ent Society is of recent origin. Indeed, to gauge by the drumfire of advertising in its behalf, we can probably inf er that the principal financial beneficiaries of the Concept are not perfectly sure that it has taken complete root in the thought patterns of the American people. SUMMER 1969 What is the Concept of the Convenient So ciety? Simpl y this: That everyone ought to have all of the conveniences that our technological capacity can produce, and that that capacity ought to apply its ingenuity to the fullest to make things ever more convenient. And part of convenience of course is easy of disposal when an object is empty or worn out; we cannot worry about what becomes of our conveniences when we are through with them. The basic trouble with the Concept of the Convenient Society is not that convenience is immoral or that we should go back to the "good old days." Rather, it is the fact that the Concept is totally at variance with the ecological concept that in a limited, living world, all material things must cycle if life is to continue indefinitely. To put it in a different form, death and decay are necessary for birth and growth, and great Cae sar's dust may appear in an ear of wheat as in a bunghole stopper. The Concept of the Conveni ent Society either does not recognize the neces sity for the recycling of materials or if it does, it ignores the necessity. Thus, a shiny new car eventually ends up a rusty wreck and so our end less production of automobiles and concomitant failure to reuse the worn-out ones leads to the proliferation of auto junkyards. Again, a new house in the suburbs becomes an old house in a slum you can see this happening already and few of our suburbs are more than twenty-five years old so the better-educated more affluent people move on to a new suburb and start a new cycle of development and decay; thus, the urban blight spreads over the land. Perhaps the whole thing can be epitomized b y the Story of the Aluminum Beer Can. It started out in the mind of some ingenious innovator. Industry fashioned it into a shiny vision of promise. Its production, filling, and distribution provided useful work for more than one deserving man. It gave pleasure with convenience to someone else. And then it ended up with glass bottles, plastic-coated milk cartons, throw-away aerosol cans aluminum tray s from TV dinners, and an endless variety of other junk in an ever-growing and unholy mountain that does not rust or rot and whose eventual disposition is or ought to be giving gray hairs to the city fathers. T HIS ESSAY WOULD BE UNBALANCED UNLESS w e c on c lude the c atalog of the foibl es of our inno vating society with m e ntion of the ge n er al igno r ance of, o r indff e ren c e to the full c o s t s o f t e chnolo g i c al innova147

PAGE 46

tion. Everyone wants the necessities of life, and everyone also wants a greater or lesser share of the amenities of life. Unfortunately, production of both necessities and luxuries entails charges against the environment. It i s even more unfortuate that society at large has only a dim awareness of the nature of these charges (or even that they exist;) and blithely dismisses the thought of charges with the assumption that th ey are not serious or will somehow be paid by some unit of government or industry. Of course, the charges sometimes become blatantly obvious, as in the recent oil-drilling catastrophe on the coast of California, and the public ire is then aroused. This wrath tends to center on the offending company or on the government official who permitted the risk to be taken, and that is indeed proper. Howev e r, fairness demands the remark that the public's own hands are not entirely clean in this matter; aft e r all, if no o ne drove a car, would it be profitable to drill for oil off the shore of California? The trouble here is that no one has told so ciety at large what the environmental costs of the innovating society are. The dissemination of such information is desperately needed, for unless it is available, one cannot see how a rational set of priorities designed to balance man's needs against environmental costs can be enforced. Hence, five minutes of gibes at air pollution by Arthur God frey are no doubt worth one thousand essays like this, but the author is compelled to write it none theless. The set of factors that have been described give to the dynamics of American Society some thing of the character of a branching chain reac tion. The creed of Technology and the Gospel of Growth serve as the initiating reaction. Once technology is initiated, it is propagated by the Cult of the Product and the Concept of the Con venient Society. The Gospel of Growth gives rise to branching reactions at all stages of the process. And those defects in our thinking that manifest themselves as the Sandbox Syndrome, Technological Megalomania, Utilitairian Vision and ignorance of the costs of the innovating so ciety, make it virtually impossible to introduce any terminating reactions into the whole scheme. We have now reached the point where we can see that some terminating reactions are needed if we are not to strangle ourselves. The air and the water have become so badly polluted that the situation is apparent to everyone. Hence, indi viduals, organizations, and industries of vision and conscience have become concerned about pollution control and the effects of pollution on our environment and what is even more import ant, are acting on their concern. Thank God for such; we are indebted to them. Nevertheless, 148 the efforts we have today are not nearly enough, nor in the long run, do they strike at the heart of the matter. Unless America changes some of its basic attitudes the internal factors men tioned above it is hard to see how even very great efforts to control pollution can be much different from a rear-guard action. Hopefully, they can keep us one jump ahead of the wolf for some time, but what is really needed is some way to chain the wolf and put him to work for us. It is absolutely necessary for the continued existence of a descent sort of human life on this earth, that inno vating societies think about why they innovate and what the consequences of innovation are. We must recognize that innovations are a means to an end and not the end itself. We must try to set up some definite goals that have the benefit of all mankind as their objective. We can no longer rely on a vague faith in progress to take care of our tomorrows. Only by setting up defined goals will it be possible to develop priorities and institutions that can guide the innovative genius of men onto paths that will be truly, as opposed to superficially, beneficial. I DO NOT THINK THAT WE ENGINEERING educators have done very much that is useful with regard to the considerations raised above. True, we need and shall continue to need engi neers and scientists, and it is our business to produce them. But what kind of engineers are we turning out? It seems to me that much of current engineering education serves to reinforce belief in the Creed of Technology and the Gospel of Growth, and all the rest of those factors, factors with which the freshman engineering stu dent has already been partially equipped by his parents, his schoolmates, and society in general. Readers who have persevered this far and who agree that I have described some real prob lems will probably wonder if I have any concrete suggestions to offer for their solution or is it to be simply "ecrasez l' infame," an attitude all too common these days. It seems to me that there are some things we can do, and even if they are not very original, I give them for what they are worth. The most immediate thing that we can do is to consider the contents of our undergraduate textbooks and courses. It is true that these do not exhort budding engineers to go out and rape the environment. But it is also true that they do not suggest that they have any responsibility to conserve it nor do they often state that tech nology imposes any stress on the environment. To be aware that a problem exists is the prerequisite for any attempt to solve the problem, and by judiCHEMICAL ENGINEERING EDUCATION

PAGE 47

. . the mystic destiny towards which innovating societies strive has lost some of the rosy tints of paradise and taken on the more lurid aspects of purgatory. cious choice of examples in engineering texts and courses we could certainly point out to our stu dents what some of our real environmental prob lems are. Such examples serve another purpose also: they can be genuine and challenging illus trations of basic engineering principles, ranging from applications of the laws of thermodynamics to problems of diffusion and convection that are as advanced as anything in Birdfoot. Why don't we try to put this sort of example into our texts and courses? We don't have to expurgate as Mother Goose must be expurgated; we merely need to add things. Another useful thing that we can do is to make an attempt to recruit women students for engineering. This would help to open up careers for women other than or in addition to that of wife and mother; it would be a removal of one of the factors repressing antinatalist tendencies existing but not active in our society. Removal of factors repressing antinatalist tendencies is Judith Blake's principal suggestion for inhibiting population growth in the United States. By act ing on her suggestion, we would also tap a source of talent and brainpower that has hardly been touched by engineering. The next thing that we can do is to see that our students are at least exposed to courses in environmental engineering. Does our school have courses in air and water pollution control? Do we encourage our students to take these courses? If we do not have such courses, what are we doing to get them? Finally, it is clear that an innovating society will be able to resolve its basic dilemma only if it can replace those attitudes and values that we have named as the Creed of Technology, the Gospel of Growth, etc., by attitudes and values more in keeping with the long-term needs of man. It is equally clear, at least to the author, that if this is to be done at all, it must be done through education. Hence, what about our own attitudes? In our teaching of the details of technology, do we ever stop to point out that technology is a two-edged sword? Indeed, do all of us even be lieve that? Do we think it worthwhile for our students to learn something besides science and engineering or do we regard the "liberal educa tion" part of our curriculum as a necessary evil forced upon us by the rest of the university? Do SUMMER 1969 we have any concern that our students should realize that we live in a limited world, a micro cosm, whose living and non-living components interact in an endless spectrum of ways, some of them of extraordinary subtlety? Do we think students should be aware that man is part of this microcosm or are we so far gone into the last stages of Technological Megalomania that we teach or at least imply that he is above and be yond it? Have we swalloded the Concept of the Convenient Society which is in essence that we can do just as we damn well please with our surroundings so that we pass this monstrosity on to the next generation of engineers? We are concerned these days with a decline in the number, or at least of the relative number, of students entering college who want to become engineers or scientists. Surely there is no single or simple explanation for this, but I do think that we can gain partial insight into the phe nomenon by applying the analysis given above. Eighteen year olds can observe and draw conclu sions as we can. Their observations may not be as thorough, and their conclusions may not be based on a very disciplined or experienced thought process; nevertheless, the Creed of Technology, the Cult of the Product, and all the rest are fairly obvious aspects of our society, and it is easy to see how they could become inex tricably entangled in the minds of youth with science and engineering. The fact that these fac tors are faults of society as a whole rather than of technology alone is irrelevant; society's de fects are manifested in a most striking way by our runaway technology. To close this essay, let us return to Professor Pig gott's opening chapter. He speaks (p. 18) of a "mystic destiny" towards which innovating societies believe they should strive by continual technological innovation, and his quote from Herbert Spencer may be taken as the canonkal form of the Creed of Technology: "Progress is not an accident but a necessity. It is a part of nature. Evil tends perpetually to disappear." Unfortunately, the stock of evil seems these days to increase even faster than the population. Change is a part of nature but that which is called Progress is often a retrogression, or if it is of benefit to us, it would be a curse to the next generation. Thus, the mystic destiny towards which innovating societies strive has lost some of the rosy tints of paradise and taken on the more lurid aspects of purgatory. I am indebted to Carol Urness for her con structive criticism of my original manuscript. 149

PAGE 48

THE GATORS GO RAY FAHIEN What should be the goals of a department of chemical engineering? National prominence through a strong graduate program or a quality undergraduate program? An orientation toward "engineering science" or an orientation toward engineering practice? A large graduate program or a large undergraduate degree production? A PhD-oriented graduate program or a master's oriented graduate program? Should it espouse a philosophy of service to the state, a philosophy of service to the engineering profession, or to a "community of scholars"? When a department has a single objective, the fulfillment of its goal demands a concentrated effort in one direction. For example, a depart ment that is interested in undergraduate degree production can hire faculty who are inspiring teachers and who would also enjoy visiting high schools to aid in recruitment; a department that aspires to national prominence for the quality of its research can hire faculty who have bril liant, creative minds and a personal desire to do research and to publish their results. When the goal of the department is singular, and when the faculty and administration accepts the singu larity, the implementation of its goal can be carried out smoothly and without conflict. De partments of great prominence can be developed in this manner. 150 ... a balanced department with multiple objectives is desirable at the University of Florida ... In many cases, external factors, such as whether it is a private or public institution, may influence or even fix the goal of the department. In some cases, it is more desirable (and even necessary) for a department to have multiple goals For example, the composition of a tenured faculty can, by its very nature, demand a divers ity of objectives; or the faculty may express an objective opposed to that of the institution or college as a whole (such as in the case of an undergraduate-oriented faculty in a graduate oriented institution). While many departments seek more than one of the many objectives listed above, few of them strive for excellence in all of them. But one de partment that, for the last four or five years, has been attempting to do all o~ these is the Chemical Engineering Department at the University of Florida. We might therefore properly ask the following questions : What are the reasons for such a multiplicity of objectives? What kind of results have been achieved? JUSTIFICATION FOR BALANCE A balanced department with multiple objec tives is desirable at the University of Florida because of the following: It is the only department of chemical engineering in a state serving over six million people; it therefore feels a broad responsibility to provide a diversified and balanc e d program. CHEMICAL ENGINEERING EDUCATION

PAGE 49

a department must recognize both academic and industrial professional goals ... its achievements will never be easily measured by quantitative indices . an engineer is not merely a technical robot . the goal of the department, and that of the student it educates, must be the betterment of human society. Its tenured faculty in 1964 was already of above average size; it was a diversified, heterogeneous and capable group that was brought together to do sponsored research under the Engineering and Industrial Experi ment Station and partly to teach. It consisted of several people with degrees in chemistry ( organic, bioch e mistry, pharmaceutical, inorganic and phy sic al), a pulp and paper technologist, an authority on imbedding flo we rs in plas tics, an expert on asphalt technology and economics, a world-famous fluorine chemist (and philosopher of science and education), and an e l ectrochemist doing over $100,000 a year of research (much of it classified) on thermal batteri~s and fuel cells. Some of the faculty were inter ested only in und ergraduate teaching, others only in sponsored research, still others in both activities. Its average age was 55 and there was only one assistant professor in the group. The new energetic and dynamic dean of the Colleg e of Engineering from 1964-68 strongly encouraged the development of the graduate program both from the standpoint of increased e nrollment, and also in terms of quality of research and instruction. It began participation in 1965 in an NSF Science Development ( or "Center-of-Excellence") Grant that provided funds for bringing in new faculty and graduate students to do fundamental research in chemical engi neering. The diversity of faculty interests and back grounds plus the strong leadership of the dean made it obligatory for the department to pursue multiple objectives lest it suffer from internal conflict among its faculty or from an external gap between the objectives of the administration and those of the faculty. Hence the objectives of the department became balanced ones and diversi fied ones; both quality and quantity were needed in the graduate and undergraduate program; both theory and practice had to be emphasized; both teaching and research had to be acknowl edged ; both masters and PhD degrees had to be offered; and both service to the State and Nation as well as service to the engineering profession and to the academic community had to be a part of departmental philosophy. To express these multiple goals in a cohesive philosophy became a first task of the new chair man when he arrived in June 1964. A statement of goals recognized the diversity of the chemical engineering profession through its strong roots in both chemistry and physics. This diversity meant that a chemical engineering department must recognize bifold professional goals : aca demic goals that strive for the advancement of fundamental knowledge and industrial ( or proSUMMER 1969 fessional) goals that have to do with the econom ical design and operation of plants that produce consumer goods ( or of substances that go into making consumer products). Just as th e overall aim of the Univ ersi ty is to serve mankind," it further stated, "so also the goal of the d ep art ment, and that of th e student it e ducates, must b e th e betterment of human soci e ty. For as a pro fessional man, an engineer is not merely a techni cal robot who responds passively and unquestion ingly to conformist pressures or to the commands of others. Instead he must be aware of, and deeply concerned with the socia l and political problems of our times. He must have a high sense of values and be capable of making deci sions with regard to principles and ideals derived from these rather than from narrow self-interest or partisan group interest. In keeping with this philosophy, the department shou ld investigate methods of establishing communications between the 'two cultures' of technology and the humani ties." Somewhat later the first annual report for the "Center of Excellence" Grant stated that "the goal of the chemical engineering department in the Science Development Program is to str iv e towards an excellence that is better expressed in terms of the significance of its contributions to scientific progress than by the volume of its activity. Its achi evemen ts therefor e will never b e e asily measured by quantitativ e indices-by numbers of students or faculty added, by the dollars worth of equipment purchased, by the number of papers in various journals, or by the number of degrees granted. We believe that the kind of excellence for which we strive cannot readily be programmed, budgeted or al located on a yearly or semiannual basis. Nor can it be fully accomplished in a time space of one year or three years or even five years. "But seeds can be planted. Morale can be improved. Research ideas can be generated. New approaches to enginee ring education can be tried. A creative intellectual atmosphere can be devel oped. Bright, highly motivated people, both young and old, can be added to a faculty. A new life, energy, and enthusiasm can be breathed into a faculty with unfulfilled goals and unrealized potential. 151

PAGE 50

... seeds can be planted . GRADUATE ENROLLMENT TRIPLES With the addition of a new chairman and three other faculty members, the chemical en gineering department began its period of devel opment in 1964-65 one year prior to the award of the "Center of Excellence" grant. Although this development was greatly accelerated by the award of the grant, it was partially retarded by inadequate space and facilities. Until November, 1967, the department had been housed entirely in a crowded World War II airplane hangar which was shared with the Aerospace Engineer ing Department. Essentially no additional space was available for graduate st udent s and research equipment, and faculty offices were not condu cive to the recruitment of prominent senior faculty members. But in les s than three years the department could point to the following accom plishments : Revision of graduate and undergraduate curricula. Graduate enrollment nearly tripled increasing to 66. Undergraduate degree production increased 50%. Seven outstanding young faculty members with excellent background s were added, decreasing the average age of the faculty from 55 to 44. Sponsored research support increased over 50%. Faculty research productivity in terms of papers submitted and publi s hed increased several fold. Two books were published and two others were started Although the above quantitative increases may be startling, even more impressive were the indications of improvements in the quality of its graduate st udent body, it s faculty, and its gradu ate program. In 1964, over half the 23 graduate students were foreign students, and roughly half were University of Florida graduates. The aver age Graduate Record Examination scores of that group was 550 or sl ightly above average. How ever of the group of 28 students admitted in Fall, 1967, all but two were graduates of Ameri can institutions other than the University of Florida. The first group of 22 who accepted ap pointments had an average Graduate Record Examination (Verbal-Quantitative Average) score of 654 at least one standard deviation higher than in 1964. (Such a score meant that the average student was in the top 6-7 % of the senior students throughout the nation who took the examination.) Initially, the award of the "Center of Excel lence" grant made possible a shift in the research emphasis of many of the older faculty members 152 to more fundamental areas of research and away from the highly applied sponsored research proj ects previously emphasized under the Engineer ing and Industrial Experiment Station. As a result of this heightened interest, new proposals for fundamental research were written and nine new projects were accepted for support by vari ous agencies such as the NSF, NIH, AEC, NASA and DOD. Thus stimulated, the face value of sponsored research nearly doubled and the an nual rate increased by over 50 % The increase in outside research suppo rt made availab le state funds for the addition of new faculty beyond the two positions allocated in the grant (In addition to positions generated through research, one state supported position was obtained from the University.) The new faculty added were not only grad uate s of leading institutions; they were also genera ll y among the top students to complete PhD work at their in stitution over a period of years. (Four of them had won NSF Fellowships in national competi tion.) Table 1 gives their backgrounds. TABLE 1. FACULTY ADDITIONS SINCE 1964 Name and Ph.D. School Area A. W Westerberg Computer -Aid ed London Design Other Background Control Data Cp. Princeton U. U. of Minnesota L E. Johns, Jr. Polymer Dynamics Dow Chemical Carnegie Tech Cont. Mech. J. P. O'Connell Cal. Berkeley X. B. Reed, Jr. Minnesota A. D. Randolph Iowa State (Now at U. Arizona) D. W. Kirmse Iowa Stat e K. E. Gubbins London R. W. Fahien Purdue Thermodynamics Transport Properties Bioengineering Appl. Math CrystallizationParticulate Systems Turbulence Transport Properties Transport Processes in Reactors Mass. Inst. Tech Pomona Coll. Union Oil Co U.C.L.A. Texas A & M Amer. Potash Spencer Chem. Co. Co lorado U. Union Carbide Oklahoma State Florida U. (Post Doc) Ethyl Corp. Iowa State U. Missouri (Rolla) Washington Univ. '' Ronald Gordon (Ph.D expected from Princeton Univer sity) will join faculty in September 1969. (Continued on page 157) CHEMICAL ENGINEERING EDUCATION

PAGE 51

would you like to plan a plant in Puerto Rico ? Too late, the plant is planned In fact construction is already beginning on Sun Oil's new $125 million refinery complex and harbor at Yabucoa. But the project at Yabucoa is sim ply one indication of Sun on the move. We're geared for growth and we need people. Maybe you ? Perh aps you'd like to work for the company that also recently boomed into the 2 billion dollar class through the merger of Sun and Sunray DX ; that's pioneering a new fertilizer plant on the island of Martinique; that operates a new Computation Center in Philadelphia; that spon sors winning teams and cars in major road racing championships in the United States an d Canada-to men tion just a few exciting projects. We need men and women to grow with us and build the future. We have openings in Exploration, Production, Manufacturing, Research and De velopment, Engineering, Sales, Ac counting Economics and Computer Operations. Locations-Philadelphia, Toledo, Tulsa, Dallas and many other areas. Write us for an appointment, write for our book "Sunoco Career Oppor tunities Guide," or contact your Col lege Placement Director to see Sun's representative when on campus. SUN OIL COMPANY, Industrial Rela tions Department, CED, 1608 Walnut St., Phila., Pa. 19103. An Equal Opportunity Employer M/F

PAGE 52

[iJ ;j I curriculum ,q Neac CEE ~epevdmeni FLEXIBLE CURRICULA CAN BE STRONG RAY FAHIEN MACK TYNER R. A. KEPPEL University of Florida Gainesville, Florida UNDERGRADUATE PROGRAM IS DIVERSIFIED There is much truth in the old story about the late G. G. Brown telling a questioner that "chemi cal engineering is what chemical engineers do." When asked what chemical engineers do, his reply was, naturally, that they do chemical en gineering! But, contrary to the belief of certain narrow-minded persons, there is no single well defined job that each and every chemical engineer does and which can be described as THE work of "THE chemical engineer." For in practice individual chemical engineers do numerous kinds of jobs each requiring different talents, abilities, and interests. In spite of this great diversity of the profes sion, chemical engineering curricula have for years consisted of a rigid set of courses that every student was required to take. No consid eration was made of his future career objectives, his personal interests, or his individual abilities. Although the typical curriculum might include perhaps two elective courses, these were often gleefully squandered in subjects such as bait casting, photography, or basket weaving. In view of our special nature as the only chemical engi neering department in the State of Florida, we felt that we have specia l obligations to provide as broad a program as possible. Accordingly, in 1965 we developed a curriculum that treated each student as an individua"fr-one whose indi vidual interests, talents, and career objectives could be expressed through a selection of option programs. The considerations that were involved in de veloping this curriculum are as follows: First, we reaffirmed the belief that an engineering cur riculum must be designed to prepare not only a broadly educated person but also train a professional man who could, upon graduation, do the kind of engineering work 154 that employers have associated with chemical enginee r ing. Both of these educational objectives had to be ful filled in four years since a five-year program has never been accepted by students, educators, or industry. How ever, it seemed neither possible nor desirable to devise a single four-year program that would prepare the student for all the various kinds of work that chemical engineers do. For it seemed wrong to subject the practice-oriented student to intensive theory and mathematics that he would not use; and it seemed equally unproductive of human talent to deprive the science-oriented student of that kind of experience that would best prepare him for graduate school and a career in research. Second, we felt that it is essential that all programs in a chemical engineering department include tho3e fundamental and core courses required for the practice of chemical engineering and that no program should be "watered down" so as to become an easy path to a cheap degree. Third, we co nsidered it important that a student not be unduly harmed by a wrong decision as to the program he chooses to enter. Therefore we made the differences between programs only about 10 %. Actually since we require 213 quarter hours (142 semester hours) for a bachelor's degree, more leeway was possible than in schools requiring much fewer hours. Fourth, we felt that a student shoul d not be forced to make a decision until his senior year so that he is experienced and ma.ture enough to make a wise choice. Fifth, we believed that the option programs should not be so rigid that a student is prevented from substituting, when his reasons are sou nd, certain courses in the options for other courses. The above general considerations were em ployed in the development of specific programs as follows: Chemical engineering science option. Throughout the country a strong trend has developed toward a chemical HIGH JR CO SCHOOL. Pre-Engineering Courses LLEGE). CORE ARF.AS Chemistry Eng Science Chem Eng UNDERGRAD OPTIONS ChE Science ChE Systems Process Eng -----Interdisciplinary Practice Options Graduate Ind School ., t] Po ., a .... "' ., "' >, .c i... Po'"' a ., "' ::, '"Cl .c i:: OH j ustry Figure 1. Flow Sheet for Undergraduate Options. CHEMICAL ENGINEERING EDUCATION

PAGE 53

.. we developed a curriculum that treated each student as an individual whose interests, talents, and career objectives could be expressed through a selection of option programs. engineering science program that primarily prepares a student for graduate work and for a career in funda mental research and teaching. Although a large majority of our students go into industry we felt that our depart ment should have such a science-oriented program avail able in order to provide (nevertheless) an alternative for the student who might otherwise go into the engineer ing science program available in the College of Engineer ing. We call this program our ch emica l engineering science option Operations, business, and technical sales options. Many of our students obtain jobs in operations and eventually in middle management where theoretical and mathematical tools are not as much needed as by the researcher. Programs for such students were designed for industrial employment or further study in business, marketing, or law. These were called the operations options and busin es s options. A modification is availabl e for students interested in careers in technical sales. TABLE 1 CORE AREAS Engineering Core Computer Model Formulation Intro. to Elec. Eng. Statics Strength of Materials Materials of Engineering* Engineering Statistics Chemistry Core Organic Chemistry Organic Chem. Processing Physical Chemistry Instrumental Analysis ,:, Taught in ChE department. Chemical Engineering Core Thermodynamics ':' Transport Phenomena Chemical Kinetics Systems Analysis Solid-Fluid Systems Control Theory Stagewise Separations Cost Estimation Process Design Process engineering option. Many students actually do not know what they want to do after graduation. As a result, chemical engineering curricula have traditionally attempted to produce a highly versatile chemical engi neer one that can easily start his career in any of many work assignments. We felt a program of this type should b e retained. We called this option process en gi neerin g but upgraded it by the addition of transport phenomena, computer modeling and applied math cours es Systems engineering option. The systems engineering approach is as much a part of chemical e ngineering as any other field of engineering, including that taught in systems engineering departments. Chemical engineers are bringing the fruits of automation to the process industries through the e ffective use of computers, mathe matical models and processes, and advanced hardware to the design and operation of chemical complexes. Conse quently we developed a computer -o riented systems en gi neerin g option in order to fill this need and also as an SUMMER 1969 TABLE 2A SUBJECTS STUDIED IN APPLIED SCIENCE OPTIONS Reactor Dynamics and Design Math Models in ChE ,:, Technical Electives Process Optimization Applied Molecular Theory Advanced Process Design Ind. and Systems Eng. Applied Math Electives Polymeric Materials i, Process Economics * Tensor Fields and Fluid Dynamics * Quarter Credits in Option 22 22 22 alternative to the stud e nt who might otherwise enroll in the systems engineering program taught in the College of Engineering. Interdisciplinary options. Some students have a dif ficult time choosing between chemical engineering and a "glamorous" field such as aerospace or nuclear engineer ing At the same time the challenging problems of today tend to be coupled interations between the application of engineering principles and the socio-economic need s of our society; e.g., pollution abatement, food production and/or population control. The inter-disciplinary options permit the chemical engineer with paralleled interests in other fields to take his degr ee in chemical engineering while at the same time studying 22 hours of approved courses in related disciplines such as aerospace engineering, environmental engineering, nuclear engineering, food science, or biomedi c al engineering The latter program is approved for direct entrance into medical school. Humanities or Liberal Studies. Many stu den ts today are concerned about the social problems of our society, about man's obligation to hi s fellow man and him self, TABLE 2B INTERDISCIPLINARY OPTIONS Chem-Aerospace Propulsion Aerodynamics Chem-Electrical Electronics Control Systems C hern-Food Science Chem. Principles Eng. Principles Chem-Mechanical Turbines and Jets Refrigeration Chem-Biomedical Zoology Biology C hern -Enviro nmental Waste Treatment Special Topics C hern-Material s Elect. Properties Corrosion Chem-Nuclear Nuclear Tech. Nuclear Chemistry 155

PAGE 54

TABLE 2C PRACTICE OPTIONS Operations Corrosion or Electrochemical Engrg Polymeric Materials Management Electives Process Economics ChE Electives Business & Sales Report Writing Speech Courses Management Electives Marketing Electives Process Economics ChE Electives Humanities, or Liberal Studies Political Science Sociology History English Philosophy Religion Foreign Languages Process Economics ChE Electives about values and ethics, and about the meaning of life itself. Therefore we developed a group of courses in the humanities and social sciences that would permit an engineering student to obtain a degree that was perhaps, because of its science content, more of a "liberal studies" program than that offered in the College of Art and Sciences. GRADUATE PROGRAMS IN SCIENCE AND SYSTEMS While some graduate students intend to teach and do basic research, many others are interested in industrial careers in development and design. Consequently the graduate program in the de partment was divided into three main areas: (1) Chemical Engineering Science: transport phe nomena, fluid dynamics, thermodynamics, kinetics, micro structure of matter, and materials science; (2) Chemical Engingeering Systems: chemical reaction engineering, process control, process dynamics, optimization, separa tions processes; and (3) Interdisciplinary Chemical Engi neering: energy conversion and fuel cells, polymer sci ence, microelectronics, process economics, and bioengi neering. Master of Engineering with Project There are four graduate programs in the department. These are: 1) the 45 quarter-hour Master of Science in Engineering program with thesis ;2) the 50 quarter-hour Master of Engi neering pre-PhD program; 3) the 50 quarter hour Master of Engineering terminal program with a project (which can involve a design, a cost analysis, an experimental investigation, a com puter study, or a technical report) ; 4) the PhD program. Transfer between programs is possible within limits. PhD Program Requires Research Proposal This program includes a written examination (which may be waived), see below, a research 156 proposal defended orally, an oral examination, and a final examination. The research proposal sets forth and describes an original research problem and / or solution, which if carried through, would represent a signficant contribu tion to chemical engineering knowledge. The area of the proposal may be the student's dissertation subject only if he has taken the written exam ination. Graduate Courses An orientation examination is used to deter mine whether entering students require any pre liminary course work, before taking the six required core courses. After three to five quarters in residence, all Master of Engineering students and nearly all PhD students are required to take a written examination based on these core courses: Models and Methods Multidimensional and Discrete Systems Thermodynamics of Reaction and Phase Equilibria Fundamental Transport Phenomena Process Dynamics 1 or Process Dynamics 2 Reactor Design and Optimization (Systems Program) or Chemical Kinetics (Science Program) During 1968 the following additional courses were taught by department faculty : Mathematical Methods in Chemical Engineering Applied Field Theory Computer Control of Processes Optimization Techniques Transport Properties and Irreversible Thermodynamics Applied Statistical Mechanics Statistical Thermodynamics Interfacial Transport Phenomena Rheology Non-Newtonian Fluid Dynamics Chemical Energy Conversion Particulate Systems Applied Fluid Dynamics Process Engineering Process Equipment Design Process and Plant Design Process Economy Analysis Tensor Fields and Fluid Dynamics Analytical Techniques for Eng and Scientists 1 Analytical Techniques for Eng and Scientists 2 Analytical Techniques for Eng and Scientists 3 RESULTS Student reaction to these diversified programs has been very good at both the undergraduate and graduate levels. CHEMICAL ENGINEERING EDUCATION

PAGE 55

THE GATORS GO (Continued from page 152) "CENTER-OF-EXCELLENCE" GRANT In August 1965, the University of Florida was awarded a 4.2 million dollar Science Develop ment Grant by the National Science Foundation. The Chemical Engineering Department was among the seven participating departments in the University. The proposal submitted by the Uni versity was entitled "Radiation, Kinetics and the Microstructure of Matter." The proposal stated that the first objective of the College of Engi neering was "to improve the scientific base of ed ucation and research through increased emphasis on the engineering implications of the microstruc ture of matter." It pointed especially to the de veloping technology of microelectronics as "only one aspect of the very general field of microengi neering which aims to place a strong emphasis upon the microscopic statistical view of nature and to relate this to human needs." In keeping with this philosophy the department de fined and delineated the meaning of the phrase "micro structure of matter" from the standpoint of modern fun damental research in chemical engineering in terms of the following connotations: 1. MOLECULAR. This approach involves the use of a knowledge of statistical mechanics molecular structure, and molecular and kinetic theory (a) to predict rates of chemical reaction either on cataly s t surfaces or in homo geneous systems (b) to predict adsorption rates (c) to predict thermodynamic properties and phase equilibria, or (d) to predict transport properties such as diffusivity, thermal conductivity, or viscosity. 2. PARTICULATE. This approach analyzes par ticulate systems in terms of their statistical properties and the particle-continuum interaction. Such systems are found in industrial crystallizers and also include aerosols, mists, dispersions, and suspensions. 3. STATISTICAL. This approach is used to describe turbulent transport processes for energy, mass and momentum in terms of elements in which fluctuations of velocity and other properties occur. 4. CONTINUUM. The microscopic view of matter can be thought of in terms of processes that occur at a point in a continuum The conservation laws for energy, mass, and momentum can be expressed in terms of the differential equations of change. Knowledge of matter from these microscopic points of view of course can be used in a given engineering system to predict macroscopic quan tities such as the total energy or mass transport or the total friction or drag in a system. This SUMMER 1969 information can be incorporated with modern design and optimization techniques in the design of an engineering system or a complete plant. NEW BUILDING FOR DEPARTMENT In the fall of 1967 the department was able to move into a modern air-conditioned educa tional building containing 51,000 sq. ft. of re s earch and teaching facilities made possible by a State bond issue and funds from the NSF grant. We now have undergraduate teaching space for modern laboratories in process measurements, transport properties, instrumental process analy sis, unit operations, process transients and con trol theory, chemical reaction kinetics, and indi vidual special projects. Graduate research space is available in process dynamics and computer control, transport phenomena and properties, in v i v o transport studies, fluid dynamics and rheology. COMPUTER CONTROLLED LABO RA TORY Modern computer facilities will permit one to control any of several pieces of process equipment in the unit operations laboratory. At present, a distillation column is being tied to a remote IBM 1070 process control terminal which connects to the IBM 360 / 65 campus computer via telephone lines. We have designed and are building a special interface between the process equipment and the terminal which serves two major functions. First of all it is a patch panel permitting any one of several processes to be "patched" into the ter minal using special jacks and plugs. Its second function results from the fact that one can simulate most of the computer actions to the process and all of the process responses to the computer at the interface itself. One can thus almost completely "debug" the computer soft ware without the process and to some extent "debug" the process hookup without the com puter. The remote computer terminal with interface can tie to 40 analog inputs (low and high level), 30 digital inputs, 24 digital outputs, 10 pulse motor outputs (which can operate in parallel), a digital display, and a rotary switch input sta tion. The terminal's transmission rate is 66 char acters per second to and from the computer which will permit about 4 random accesses per second or about 20 analog to 60 digital sequential accesses per second. The software is written in Fortran and i s 157

PAGE 56

quite modular permitting most of the essential portions to be used in all processes. The depart ment also has two remote consoles for the IBM 360, a 60-amplifier Ease computer, and a WANG Calculator. OTHER ACCOMPLISHMENTS During the past three years, three different members of the faculty have won undergraduate teaching awards; Professor Tyner, Professor Gubbins, and Professor O'Connell. This year the Sigma Xi research award went to a chemical engineering graduate student and the Phi Kappa Phi award for the outstanding student in the University went to a chemical engineering junior. Last year the faculty published 20 papers, had 14 others accepted, and submitted 14. Two books were published, two accepted and two submitted. CONCLUSION If the goal of the department is an excellence that is not measured by quantitative indices, the above achievements are not in themselves suffi cient indication that excellence has been attained. But they may indicate that the seeds of excellence have indeed been planted and have germinated. If these are now nurtured by additional support, the progress of the department toward excellence can continue not only in its research program, not only in its instructional program, not only in the achievement of each of its multiple objectives, but also in the fulfillment of its ultimate aim: the betterment of human society. (Cont'd from p. 140) OPTIMIZATION: R. R. Hughes BIBLIOGRAPHY 1. Andersen, S. L., Ch em E n g P ra g., 57, No. 3, 80-83 (March, 1961). 2. Aris, R., G. L Nemhauser, and D. J. Wilde, AI C hE J. 10 913-919 (Nov., 1964). 3. Baumol, W. J., "Economic Theory and Operations Analysi s ," 438 pp., Prentice-Hall, Englewood Cliffs, N. J. (1961). 4 Blakemore, J. W. and S. H. Davis, Jr edit. "Op timization Techniques" Ch e m. Eng. Prag. Symp. S erie s No 50, 60, (1964). 5. Carr, C. R., and C. W. Howe, "Quantitative Decision Procedures in Manag e ment and Economics Deter ministi c Theory and Applications" 383 pp., McGraw Hill, N. Y (1964). 6. Dantzig G. B., "Linear Programming and Exten sions" 625 pp., Princeton Univ. Press, Princeton, N. J. (1963). 7. DiBella, C W., and W F. Stevens, I & EC Proc e ss D e s an d D ev ., 4, 16-20 (Jan, 1965). 158 8. Ford, L. R., Jr and Fulkerson, D. R., "Flow s in Networks" 194 pp., Princeton Univ. Press, Princeton, N. J. (1962). 9. Franks, R. G. E "Mathematical Modeling in Chemi cal Engineering" 285 pp., J Wiley, N Y., (1966) 10 Gass, S. I., "Linear Programming M e thod s and Applications," 2nd edit., 250 pp McGraw-Hill, N. Y (1964). 11. Graves, R. L., and P. Wolfe, edit., "Recent Advances in Mathematical Programming," 347 pp., McGraw Hill, N. Y., (1963). 12. Griffith, R. E., and R. A. Stewart, Mgt. Sci e nc e 7, 379-382 (July, 1961). 13. Hadley, G., "Linear Programming," 520 pp, Addison Wesley, Reading, Mass (1962). 14. Hadley, G., "Nonlinear and Dynamic Prog r am ming" 484 pp., Addison-W e sl e y, Reading, Mas s (1964). 15. Happel, John, "Chemical Pro c ess Economics," 291 pp., J. Wiley, N. Y. (1958) 16. Hertz, D. B., Ha rv a r d B us R ev 42 No. 1, 95-106 (Jan.-Feb 1964) 17. Hughes, R. R. and J. C. Ornea, "Decision-Making in Competitive Situations," Paper, Panel Disc 34, 7th World Petr. Congr., Mexico City, (April, 1967). 18. Hughes, R. R., E Singer, and M. Souders, "Machin e Design of Refineri e s," Proc. 6th World Petr. Con gress, Frankfurt / Main, Section VII, pp. 93-102, (June, 1963). 19. Lavi, A. and T. P. Vogl, edit., "Rec e nt Advanc e s in Optimization Techniques," 656 pp Wiley, N. Y., (1966). 20. Mangasarian, 0. L., Mgt. S c i., 10, 353-359 (Jan 1964) 21. Mangasarian, 0. L., and Rosen, J. B J O pn s R e s Soc. Am., 12, 143-154, (Jan.-Feb 1964). 22. Mugele, R. A., "Th e P r obe and Edg e Theorems for Non-Linear Optimization," in Lavi, A and T. P. Vogl, Ref. 19 above, pp. 131-144. 23. Naylor, T. H J. L. B e lintfy, D. S Burdick and Kon g Chu, "Computer Simulation Techniques," xiii + 352 pp., J. Wiley, N. Y. (1966) 24. Ornea, J. C. and G. G Eldredg e "Nonlinear Parti tioned Models for Plant Scheduling and Economic Evaluation," Paper 4.15, AIChE / I ChemE Joint Mtg, London, June, 1965. 25. Ros e n, J B., J. Soc. Ind. A pp Math, 8, 181-217 (1960) 9, 514-532 (1961). 26. Rosen, J B., Num. Math., 6, 250-260 (1964) 27. Rosen, J. B. and J. C. Ornea, Mgt. Sci. 10, 160-173 (Oct. 1963). 28. Rudd, D and C. C. Watson, "Strategy in Process Engineering," Preliminary Edit., J. Wiley, (1966). 29. Singer, E., Ch em Eng. P r ag. Symp. S e ri e s No. 37, 58, 62-74 (1962). 30. Souders, Mott, Ch e m. Eng. P r ag. 62, No. 3, 79-81 (March 1966) 31. Wilde, D. J., "Optimum Seeking Methods," 202 pp., Prentice-Hall, Englewood Cliffs, N. J (1964) 32 Wilde, D. J., Ind. Eng Ch e m. 57, No 8, 18-31 ( Aug. 1965). 33. Williams, T. J., and R. E. Otto, AIEE T r a n s 79, (Comm and Elect.), 458-473 (Nov. 1960). CHEMICAL ENGINEERING EDUCATION

PAGE 57

The world of Union Oil salutes the world of chemical engineering We at Un i on Oil are particularly indebted to the colleges and universities wh i ch educate chemical eng i neers. Because their graduates are the scientists who contr i bute i mmeasurably to the pos i tion Union en j oys today : Th e thirtieth largest rnanufacturing company in t h e United States, wit h operations throughout t h e world. Uni on to day explores for and produces oi l and natural gas i n such distant places as the Persian Gulf and Alaska s Cook Inlet. We market petroleum products and petro chemicals throughout the free world Our research scientists are constantly d i scovering new ways to do things better. In fact we have been granted more than 2,700 U S patents. We and our many subsidiaries are engaged in such diverse projects as developing new refining processes develop1ng new fert i l i zers to i ncrease the food yield and the conservation of air and water. Today Union Oil s growth is dynamic Tomorrow will be even more stimulating Thanks largely to people who join us from leading i nstitutions of learning. If you enjoy working in an atmosphere of imagination and challenge why not look into the world of Union Oil? Growth ... with innovation. Union Oil Company of California

PAGE 58

CHEMICAL ENGINEERING DIVISION ACTIVITIES Seventh Annual Lectureship Award to C. J. Pings The 1969 ASEE Chemical Engineering Divi sion Lecturer is Dr. C. J. Pings of the California Institute of Technology. The purpose of this award lecture is to recognize and encourage out standing achievement in an important field of fundamental chemical engineering theory or practice. The 3M Company provides the financial support for this annual lecture award. Bestowed annually upon a distinguished engi neering educator who delivers the Annual Lecture of the Chemical Engineering Division, the award consists of $1,000 and an engraved certificate. These were presented to this year's Lecturer, Dr. C. J. Pings, at the Annual Chemical Engineering Division Banquet held June 24, 1969 at the Penn sylvania State University. Dr. Pings spoke on "A Chemical Engineer Looks at the Phy sics of Simple Liquids." A paper based upon his lecture will be published in an early issue of Chemical Engineering Education. PREVIOUS LECTURES 1963, A. B. Metzner, University of Delaware, "Non-Newtonian fluids." 1964, C. R. Wilke, University of California, "Mass transfer in turbulent flow!' 1965, Leon Lapidus, Princeton University, "As pects of modern control theory and applica tion." 1966, Octave Levenspiel, Illinois Institute of Technology, "Changing Attitudes to Reactor Design." 1967. Andreas Acrivos, Stanford University, "Matched Asympototic Expansions." 1968, L. E. Scriven, University of Minnesota, "Flow and Transfer at Fluid Interfaces." 160 BIOGRAPHIC SKETCH Cornelius J. Pings was born in Montana in 1929 and entered the California In stitute of Technology in 1947 from which he received a BS degree in Applied Chemistry in 1951, an MS degree in Chemical Engi neering in 1952, and a PhD degree in Chemical Engi neering in 1955. He served on the faculty at Stanford University from 1955 to 1959 before returning to Caltech where he is now Professor of Chemical Engineering. In research and scholarly activities his interests have centered about the areas of applied chemical thermody namics and the physics and chemistry of liquids. In the area of thermodynamics his work has led to important improvements in the methods for quantitatively describ ing the displacement of chemical equilibria. His research in liquid state physics and chemistry, which has been both theoretical and experimental, has the long-range objec tive of fundamental elucidation of the liquid state. He has developed one of the most extensive and best equipped laboratories in the United States for the fundamental study of fluids. Recent advances, deal with sound absorp tion at critical states, the structure of liquid argon, and studies of intermolecular forces. In addition to his research and teaching activities, he has been active in faculty government and in student affairs. He has served for the past year as chairman of a special faculty committee on the aims and goals of Cal tech. Also active in civic affairs, Dr. Pings was named in 1968 to the Community Redevelopment Agency of the City of Pasadena. He has served as a consultant to a number of industrial firms and to the Department of Defense. Professor Pings served as Visiting Professor of Chemi cal Engineering at the University of Brazil in 1963. He has received two Presentation Awards from the AIChE one from the 56th National Meeting in Houston (1963) and the other from the 56th Annual Meeting in San Fran cisco ( 1965). He is editor of the Journal of Physics and Chemistry of Liquids, and will serve in August 1969, as Chairman of the Gordon Research Conference on the Chemistry and Physics of Liquids. CHEMICAL ENGINEERING EDUCATION

PAGE 59

R. J. Reynolds Tobacco Company has one of the country's best Development Centers .. and we want CHEMICAL ENGINEERS! We have immediate openings for CHEMICAL ENGINEERS in the Tobacco Products Division of the Product Development Department Our Product Development Department is relatively new, being established in 1965 primarily due to the Company's expanding diversification program. The Product Development Department has the r esponsibility for formulating, developing and eval Jating products for existing, as well as potential Tiarkets in the areas of foods, allied packaginq Jroducts and tobacco. CHEMICAL ENGINEERS at RJR develop exi s ting lines in the Company's food and tobacco product s and work on new product lines in a variety of fields These include such leaders as; HAWAIIAN PUNCH, CHUN KING, PATIO, COLLEGE INN and FILLER SNACKS in the food line and WINSTON, SALEM, CAMEL, PRINCE ALBERT and MADEIRA MIXTURE in the tobacco line The challenge, the growth and the future a r e un limited for CHEMICAL ENGINEERS a t R J . Reynolds Tobacco Company, Winston Sal e m, N. C

PAGE 60

Right now, Westvaco engineers are revolutionizing paper for printing, reprographics,computers, clothing, packaging, disposables, structures. And chemicals for coatings, pharmaceuticals,inks, tires, soaps,and waxes. And adsorbents for environmental control Yourjob: Our next revolution. See our campus representative. Or write to: Roger Keehn, Westvaco Building, 299 Park Avenue New York, N. Y. 10017 West Virginia [I] Pulp and Paper An Equ a l Op po rtunit y Emplo y er