A GRADUATE COURSE IN POLYMER PROCESSING

STANLEY MIDDLEMAN
University of Massachusetts
Amherst, MA 01003

WHILE THE polymer-related industries hold

a most significant position among chemical
industries today, and while a large fraction of both
B.S. and advanced degree holders in Chemical
Engineering today find employment in some area
of polymer production or fabrication, a relatively
small number of ChE departments provide an op-
portunity for substantial course and research work
in this field. Through its association with the
Polymer Science and Engineering (PSE) Depart- Stanley Middleman is a Professor of Chemical Engineering at the
ment of the University of Massachusetts, gradu- University of Massachusetts, Amherst, where he holds joint appoint-
ate students in ChE at U. Mass can elect from a ments in Chemical Engineering, and in Polymer Science and Engineer-
large selection of polymer-related courses which ing. His major research interests are in the areas of fluid dynamics,
make up the Ph.D. curriculum of the PSE Depart- especially rheology and polymer processing. He is the author of three
books: "The Flow of High Polymers," "Transport Phenomena in the
Cardiovascular System," and "Fundamentals of Polymer Processing."
TABLE 1 His baccalaureate and doctoral degrees were both earned at The Johns
Major Course Offerings of the Hopkins University.
Polymer Science and Engineering Department
501 Introduction to Polymer Science
502 Polymer Science Lab such course: Polymer Processing.
503 Polymer Synthesis Lab To put the course in the perspective of a
589. Chemistry of Macromolecules broader program, Table 1 shows the total course
670 Applied Polymer Science
offerings of the PSE Department. The Polymer
720 Viscoelasticity
721 Polymer Microscopy and Morphology Lab Processing course is required of all PSE students,
731 Polymer Properties who normally take it in their second year of grad-
733 Polymer Reactions Induced by Stress uate study. ChE gradute students may elect the
734 Degradation and Stability of Polymers course at any time. Indeed, the nature of the
735 Interaction of Radiation with Matter course is such that senior ChE majors can and do
736 Applied Spectroscopy of Polymers
737 Polymer Reactor Engineering take Polymer Processing as a technical elective. A
740 Magnetic Resonance of Polymers typical class "mix" is 20-25 PSE graduate stu-
742 Biopolymers dents, 6-10 ChE graduate students, and several
790 Organic Polymerization Reactions
792 Polymer Rheology
793 Polymer Processing_
798 Physical Chemistry of High Polymers The student must learn that while
799 Physical Chemistry of High Polymers II there can be clearly incorrect models, most
engineering processes allow for many levels of
ment. Many of these courses cover engineering modeling which, while not incorrect, do differ in
aspects of the polymer field, and are taught by sophistication, ease of application, and detail
members of the ChE faculty who hold joint ap- of correspondence to reality.
pointments with PSE. This article focuses on one

164 CHEMICAL ENGINEERING EDUCATION

 models of a process. Much of undergraduate edu-
cation is occupied with learning to solve problems
Paramount to developing a which have correct solutions. Thus it is clear that
facility with modeling is examining the derivative of sin x is cos x, and there is no
the correspondence of a model to reality. room for debate about this. In this Polymer Proc-
Thus, the models are compared to essing course I try to emphasize the concept of
industrial or laboratory data to
the maximum extent possible. modeling physical phenomena. The student must
learn that while there can be clearly incorrect
models, most engineering processes allow for many
levels of modeling which, while not incorrect, do
seniors in the ChE department. In addition, we differ in sophistication, ease of application, and
often have a few (2-4) industrial employed en- detail of correspondence to reality. The task in
gineers who take the course. this course is to provide sufficient experience,
Table 2 shows the major chapter headings of through discussion and problem solving, that the
the text used in this course, a book which I wrote student develops some facility and confidence in
myself in response to needs I felt as a teacher for formulating a model that is appropriate to the
an appropriate text in this area. The book reflects goal at hand. It is very difficult to convince stu-
a certain philosophy of education in this field dents that in some cases the best model to use is
which I have developed from my experiences in one that is so simple and sloppy that it appears, at
teaching, directing research, and consulting in the first thought, that one might be ridiculed for even
field of polymer fluid dynamics. entertaining its use.
In many ways I regard this course to be one in Another feature of this course arises from the
applied fluid dynamics, w-ith an emphasis on flow unusual "mix" of student backgrounds that I must
processes dominated by viscous effects. As such, it deal with. The bulk of the students are in the PSE
provides a useful follow-up to the usual under- program, and two-thirds of them are graduates of
graduate course in fluid dynamics, which offers a Chemistry programs. Thus, these students have
broad coverage of fluid flow analysis with, usually, seen no fluid dynamics, no transport phenomena,
minimal depth of study of problems involving
highly viscous and non-newtonian fluids. In addi-
tion, the chapter on Heat and Mass Transfer (it- In many ways I regard this course
self one hundred pages in length) provides an op- to be one in applied fluid dynamics, with
portunity for reinforcement of the elements of an emphasis on flow pressures dominated by
convective transport phenomena usually touched viscous effects. As such, it provides a useful follow-up
on briefly in an undergraduate course. to the usual undergraduate course in fluid dynamics,
which offers a broad coverage of fluid flow analysis
Another goal of this course is the development with, usually, minimal depth of study of problems
of the student's skill in formulating engineering involving highly viscous and non-newtonian fluids.
TABLE 2
Contents of "Fundamentals of Polymer Processing," by
Stanley Middleman (Published by McGraw-Hill, 1977) in many cases no mechanics at all. They have had
a science education, they believe in science, and
1. Polymer Processing
2. Modeling Philosophy
for them the concept of crude, approximate,
3. Continuum Mechanics models is often an alien, disturbing, and offensive
4. Dimensional Analysis in Design and Interpretation of concept. Thus, the course begins with discussion of
Experiments the philosophy of modeling and then turns to de-
5. Simple Model Flows velopment of the principles of mechanics as ap-
6. Extrusion
7. Calendering
plied to the dynamics of viscous flow. A brief dis-
8. Coating cussion of rheology is included here, since some of
9. Fiber Spinning the students will not have done the full semester:
10. Tubular Film Blowing course we offer in that area.
11. Injection Molding Following this introduction of fundamentl:1.1.
12. Mixing
13. Heat and Mass Transfer
concepts and tools, we turn to ex:amining models
14. Elastic Phenomena for a wide range of idealized flow situations and
15. Stability of Flows then spend the bulk of the semester on applying

FALL 1978 165

these models to a variety of polymer flow proc-
esses. Paramount to developing a facility with
modeling is examining the correspondence of a
model to reality. Thus, the models are compared to
industrial or laboratory data to the maximum ex-
tent possible. Below is a typical example from
Chapter 13 of the text, in which several models
:::,
z
.
are examined in the light of existing experimental
data. The assumptions inherent in the models are
reviewed, and then models which relax these as-
sumptions are developed.
Example: Comparison of measured and predicted Nusselt
numbers-Griskey and Wiehe present data for heat transfer
to molten polyethylene pumped through a 3/8-in heated
pipe. They present the data in terms of an "arithmetic av-
erage Nusselt number," shown plotted in Figure 1. Compare FIGURE 1. Data of Griskey and Weihe compared to
the data with theory. theory (Eq. 1, using n = 0.7). *
We begin by constructing the theoretical curve in terms
of the average Nusselt number Nua. For ver y small values
of U 'TTR 2 I aTL = wCPl kL the Leveqque solution holds, and however, that the models are subject to certain assumptions
since the extent of heat transfer is not great, we expect that which are not always met. In particular we have assumed:
Nua = Nu. If the Leveque equation is used for the local 1. The viscosity is independent of temperature.
Nusselt number, and if integration is carried out to obtain 2. The pipe wall is isothermal.
the average, the result is found to be 3. No viscous heat generation occurs.*

-
Nua =Nu= 1.61 (3n
-~ +-
4n
1) 113
( -4UR2
- L -)
aT
113 The example continues with an examination of
several additional models.
(1)
At present there is no laboratory experience
=
l, 75
( 3n +
4n
1) 1/3 ( wCP )1/3
kL
associated with this course. However, we are pres-
ently building a Polymer Process Fundamentals
It is much more tedious to carry out the same procedure Laboratory, and expect to integrate this facility
using the Graetz infinite series solution, and instead we ex- into the graduate teaching program.
amine_the limiting behavior at the extreme where the fluid
is almost completely heated to the wall temperature. Under
In summary, then, while this course is nom-
those conditions we find inally one in the area of polymer transport phe-
nomena, it serves several more general roles as
q = -wCP(<T> -T0 ) = -wCP(Tw-T0 )
well. Of greatest importance, I think, is the de-
and, it follows that velopment of the capacity to examine a process,
- 2wC think about it in simple physical terms, and then
Nna = 1Tk!, (2)
-produce a mathematical model of the process that
Figure 1 shows this asymptotic relation, as well as the represents the best compromise between simplicity
Leveque limit [Eq. 1] for n = 0.7 (the value noted by Gris- of solution and application, on the one hand, and
key and Wiehe). It is not very difficult to interpolate a correspondence to reality, on the other. A second
smooth curve between the two asymptotic limits.
The data of this example are seen to be in reasonably
goal, and an important one, is development of a
good agreement with the theory. Other sets of experimental coherent set of principles of transport phenomena
data, obtained with polymer solutions, also bear out the gen- (fluid dynamics, heat and mass transfer) as ap-
eral validity of the models presented above. We must recall, plied to the design and analysis of highly viscous,
often non-newtonian, systems. Finally, of course,
the major polymer processes are discussed and il-
A second goal, and an lustrated, thereby providing an introduction to an
important one, is development of a coherent area of engineering practice that is already of
set of principles of transport phenomena (fluid major importance, and that continues to grow. •
dynamics, heat and mass transfer) as applied to the
design and analysis of highly viscous, often
non-newtonian, systems.
*Reproduced by permission from Middleman, S., "Fun-
damentals of Polymer Processing", McGraw-Hill-1977.

166 CHEMICAL ENGINEERING EDUCATION