MIT OpenCourseWare

http://ocw.mit.edu

Electromechanical Dynamics

For any use or distribution of this textbook, please cite as follows:

Woodson, Herbert H., and James R. Melcher. Electromechanical Dynamics.

3 vols. (Massachusetts Institute of Technology: MIT OpenCourseWare).
http://ocw.mit.edu (accessed MM DD, YYYY). License: Creative Commons
Attribution-NonCommercial-Share Alike

For more information about citing these materials or

our Terms of Use, visit: http://ocw.mit.edu/terms
ELECTROMECHANICAL DYNAMICS

Part II: Fields, Forces, and Motion

 ELECTROMECHANICAL
DYNAMICS
Part II: Fields, Forces, and Motion

HERBERT H. WOODSON
Philip Sporn Professor of Energy Processing
Departments of Electrical and Mechanical Engineering

JAMES R. MELCHER
Associate Professor of Electrical Engineering
Department of Electrical Engineering
both of Massachusetts Institute of Technology

JOHN WILEY & SONS, INC., NEW YORK . LONDON * SYDNEY

To our parents
 PREFACE

Part II: Fields, Forces, and Motion

In the early 1950's the option structure was abandoned and a common core
curriculum was instituted for all electrical engineering students at M.I.T.
The objective of the core curriculum was then, and is now, to provide a
foundation in mathematics and science on which a student can build in his
professional growth, regardless of the many opportunities in electrical
engineering from which he may choose. In meeting this objective, core
curriculum subjects cannot serve the needs of any professional area with
respect to nomenclature, techniques, and problems unique to that area.
Specialization comes in elective subjects, graduate study, and professional
activities.
To be effective a core curriculum subject must be broad enough to be
germane to the many directions an electrical engineer may go professionally,
yet it must have adequate depth to be of lasting value. At the same time, the
subject must be related to the real world by examples of application. This
is true because students learn by seeing material in a familiar context, and
engineering students are motivated largely by the relevance of the material
to the realities of the world around them.
In the organization of the core curriculum in electrical engineering at
M.I.T. electromechanics is one major component. As our core curriculum
has evolved, there have been changes in emphasis and a broadening of the
topic. The basic text in electromechanics until 1954, when a new departure
was made, was Electric Machinery by Fitzgerald and Kingsley. This change
produced ElectromechanicalEnergy Conversion by White and Woodson,
which was used until 1961. At that time we started the revision that resulted
in the present book. During this period we went through many versions of
notes while teaching the material three semesters a year.
Our objective has always been to teach a subject that combines classical
mechanics with the fundamentals of electricity and magnetism. Thus the
subject offers the opportunity to teach both mechanics and electromagnetic
theory in a context vital to much of the electrical engineering community.
Our choice of material was to some extent determined by a desire to give
the student a breadth of background sufficient for further study of almost
any type of electromechanical interaction, whether in rotating machinery,
 Preface

plasma dynamics, the electromechanics of biological systems, or magneto-

elasticity. It was also chosen to achieve adequate depth while maintaining
suitable unity, but, most important, examples were chosen that could be
enlivened for the engineering student interested in the interplay of physical
reality and the analytical model. There were many examples from which to
choose, but only a few satisfied the requirement of being both mathe-
matically lucid and physically demonstrable, so that the student could "push
it or see it" and directly associate his observations with symbolic models.
Among the areas of electrical engineering, electromechanics excels in offering
the opportunity to establish that all-important "feel" for a physical phe-
nomenon. Properly selected electromechanical examples can be the basis for
discerning phenomena that are remote from human abilities to observe.
Before discussing how the material can be used to achieve these ends, a
review of the contents is in order. The student who uses this book is assumed
to have a background in electrostatics and magnetostatics. Consequently,
Chapter 1 and Appendix B are essentially a review to define our starting
point.
Chapter 2 is a generalization of the concepts of inductance and capacitance
that are necessary to the treatment of electromechanical systems; it also
provides a brief introduction to rigid-body mechanics. This treatment is
included because many curricula no longer cover mechanics, other than
particle mechanics in freshman physics. The basic ideas of Chapter 2 are
repeated in Chapter 3 to establish some properties of electromechanical
coupling in lumped-parameter systems and to obtain differential equations
that describe the dynamics of lumped-parameter systems.
Next, the techniques of Chapters 2 and 3 are used to study rotating
machines in Chapter 4. Physical models are defined, differential equations
are written, machine types are classified, and steady-state characteristics are
obtained and discussed. A separate chapter on rotating machines has been
included not only because of the technological importance of machines but
also because rotating machines are rich in examples of the kinds of phe-
nomena that can be found in lumped-parameter electromechanical systems.
Chapter 5 is devoted to the study, with examples, of the dynamic behavior
of lumped-parameter systems. Virtually all electromechanical systems are
mathematically nonlinear; nonetheless, linear incremental models are useful
for studying the stability of equilibria and the nature of the dynamical
behavior in the vicinity of an equilibrium. The second half of this chapter
develops the classic potential-well motions and loss-dominated dynamics in
the context of electromechanics. These studies of nonlinear dynamics afford
an opportunity to place linear models in perspective while forming further
insights on the physical significance of, for example, flux conservation and
state functions.
 Preface

Chapter 6 represents our first departure from lumped-parameter systems

into continuum systems with a discussion of how observers in relative motion
will define and measure field quantities and the related effects of material
motion on electromagnetic fields. It is our belief that de rotating machines
are most easily understood in this context. Certainly they are a good demon-
stration of field transformations at work.
As part of any continuum electromechanics problem, one must know how
the electric and magnetic fields are influenced by excitations and motion. In
quasi-static systems the distribution of charge and current are controlled by
magnetic diffusion and charge relaxation, the subjects of Chapter 7. In
Chapter 7 simple examples isolate significant cases of magnetic diffusion or
charge relaxation, so that the physical processes involved can be better
understood.
Chapters 6 and 7 describe the electrical side of a continuum electro-
mechanical system with the material motion predetermined. The mechanical
side of the subject is undertaken in Chapter 8 in a study of force densities of
electric and magnetic origin. Because it is a useful concept in the analysis of
many systems, we introduce the Maxwell stress tensor. The study of useful
properties of tensors sets the stage for later use of mechanical stress tensors
in elastic and fluid media.
At this point the additional ingredient necessary to the study of continuum
electromechanics is the mechanical medium. In Chapter 9 we introduce
simple elastic continua-longitudinal motion of a thin rod and transverse
motion of wires and membranes. These models are used to study simple
continuum mechanical motions (nondispersive waves) as excited electro-
mechanically at boundaries.
Next, in Chapter 10 a string or membrane is coupled on a continuum
basis to electric and magnetic fields and the variety of resulting dynamic
behavior is studied. The unifying thread of this treatment is the dispersion
equation that relates complex frequency w with complex wavenumber k.
Without material convection there can be simple nondispersive waves, cut
off or evanescent waves, absolute instabilities, and diffusion waves. The
effect of material convection on evanescent waves and oscillations and on
wave amplification are topics that make a strong connection with electron
beam and plasma dynamics. The method of characteristics is introduced as a
convenient tool in the study of wave propagation.
In Chapter 11 the concepts and techniques of Chapters 9 and 10 are
extended to three-dimensional systems. Strain displacement and stress-strain
relations are introduced, with tensor concepts, and simple electromechanical
examples of three-dimensional elasticity are given.
In Chapter 12 we turn to a different mechanical medium, a fluid. We
first study electromechanical interactions with inviscid, incompressible
 Preface

fluids to establish essential phenomena in the simplest context. It is here that

we introduce the basic notions of MHD energy conversion that can result
when a conducting fluid flows through a transverse magnetic field. We also
bring in electric-field interactions with fluids, in which ion drag phenomena
are used as an example. In addition to these basically conducting processes,
we treat the electromechanical consequences of polarization and magnetiza-
tion in fluids. We demonstrate how highly conducting fluids immersed in
magnetic fields can propagate Alfv6n waves.
In Chapter 13 we introduce compressibility to the fluid model. This can
have a marked effect on electromechanical behavior, as demonstrated with
the MHD conduction machine. With compressibility, a fluid will propagate
longitudinal disturbances (acoustic waves). A transverse magnetic field and
high electrical conductivity modify these disturbances to magnetoacoustic
waves.
Finally, in Chapter 14 we add viscosity to the fluid model and study the
consequences in electromechanical interactions with steady flow. Hartmann
flow demonstrates the effect of viscosity on the dc magnetohydrodynamic
machine.
To be successful a text must have a theme; the material must be inter-
related. Our philosophy has been to get into the subject where the student
is most comfortable, with lumped-parameter (circuit) concepts. Thus many
of the subtle approximations associated with quasi-statics are made naturally,
and the student is faced with the implications of what he has assumed only
after having become thoroughly familiar with the physical significance and
usefulness of his approximations. By the time he reaches Chapter 4 he will
have drawn a circle around at least a class of problems in which electro-
magnetic fields interact usefully with media in motion.
In dealing with physical and mathematical subjects, as we are here, in
which the job is incomplete unless the student sees the physical laws put to
work in some kind of physical embodiment, it is necessary for the thread of
continuity to be woven into the material in diverse and subtle ways. A
number of attempts have been made, to which we can add our early versions
of notes, to write texts with one obvious, pedagogically logical basis for
evolving the material; for example, it can be recognized that classes of
physical phenomena could be grouped according to the differential equation
that describes the pertinent dynamics. Thus we could treat magnetic diffusion,
diffusion waves on elastic continua, and viscous diffusion waves in one
chapter, even though the physical embodiments are entirely different.
Alternatively, we could devise a subject limited to certain technological
applications or cover superficially a wide range of basically unrelated topics
such as "energy conversion" under one heading. This was the preva-
lent approach in engineering education a decade or so ago, even at the
 Preface

undergraduate level. It seems clear to us that organizing material in a teach-

able and meaningful fashion is far more demanding than this. To confess our
own mistakes, our material went originally from the general to the specific; it
began with the relativistic form of Maxwell's equations, including the effects
of motion, and ended with lumped-parameter devices as special cases. Even
if this were a pedagogically tenable approach, which we found it was not,
what a bad example to set for students who should be learning to distinguish
between the essential and the superfluous! Ideas connected with the propaga-
tion of electromagnetic waves (relativistic ideas) must be included in the
curriculum, but their connection with media in motion should be made after
the student is aware of the first-order issues.
A meaningful presentation to engineers must interweave and interrelate
mathematical concepts, physical characteristics, the modeling process, and
the establishment of a physical "feel" for the world of reality. Our approach
is to come to grips with each of these goals as quickly as possible (let the
student "get wet" within the first two weeks) and then, while reinforcing what
he has learned, continually add something new. Thus, if one looks, he will
see the same ideas coming into the flow of material over and over again.
For the organization of this book one should look for many threads of
different types. We can list here only a few, in the hope that the subtle
reinforcing interplay of mathematical and physical threads will be made
evident. Probably the essential theme is Maxwell's equations and the ideas of
quasi-statics. The material introduced in Chapter 1 is completely abstract,
but it is reinforced in the first few chapters with material that is close to home
for the student. By the time he reaches Chapter 10 he will have learned that
waves exist which intimately involve electric and magnetic fields that are
altogether quasistatic. (This is something that comes as a surprise to many
late in life.) Lumped-parameter ideas are based on the integral forms of
Maxwell's equations, so that the dynamical effects found with lumped-
parameter time constants LIR and RC in Chapter 5 are easily associated with
the subjects of magnetic diffusion and charge relaxation. A close tie is made
between the "speed voltage" of Chapter 5 and the effects of motion on
magnetic fields, as described by field transformations in Chapters 6 to 14.
Constant flux dynamics of a lumped coil in Chapter 5 are strongly associated
with the dynamics of perfectly conducting continuous media; for example,
Alfvdn waves in Chapter 12.
Consider another thread of continuity. The book begins with the mathe-
matics of circuit theory. The machines of Chapter 4 are essentially circuits in
the sinusoidal steady state. In Chapter 5 we linearize to pursue lumped-
parameter ideas of stability and other transient responses and then proceed
to nonlinear dynamics, potential-well theory, and other approaches that
should form a part of any engineer's mathematical background. By the time
 Preface

the end of Chapter 10 is reached these ideas will have been carried into the
continuum with the addition of tensor concepts, simple cases of the method
of characteristics, and eigenvalue theory. The w-k plot and its implication
for all sorts of subjects in modern electrical engineering can be considered as
a mathematical or a physical objective. The ideas of stability introduced
with ordinary differential equations (exp st) in Chapter 5 evolve into the
continuum stability studies of Chapter 10 [expj(wft - kx)] and can be
regarded as a mathematical or a physical thread in our treatment. We could
list many other threads: witness the evolution of energy and thermodynamic
notions from Chapters 3 to 5, 5 to 8, and 8 to 13.
We hope that this book is not just one more in the mathematics of elec-
trical engineering or the technical aspects of rotating machines, transducers,
delay lines, MHD converters, and so on, but rather that it is the mathe-
matics, the physics, and, most of all, the engineering combined into one.
The material brought together here can be used in a variety of ways. It has
been used by Professors C. N. Weygandt and F. D. Ketterer at the University
of Pennsylvania for two subjects. The first restricts attention to Chapters
1 to 6 and Appendix B for a course in lumped-parameter electromechanics
that both supplants the traditional one on rotating machines in the electrical
engineering curriculum and gives the background required for further study
in a second term (elective) covering Chapter 7 and beyond. Professors C. D.
Hendricks and J. M. Crowley at the University of Illinois have used the
material to follow a format that covers up through Chapter 10 in one term
but omits much of the material in Chapter 7. Professor W. D. Getty at the
University of Michigan has used the material to follow a one-term subject in
lumped-parameter electromechanics taught from a different set of notes.
Thus he has been able to use the early chapters as a review and to get well
into the later chapters in a one-term subject.
At M.I.T. our curriculum seems always to be in a state of change. It is clear
that much of the material, Chapters 1 to 10, will be part of our required
(core) curriculum for the forseeable future, but the manner in which it is
packaged is continually changing. During the fall term, 1967, we covered
Chapters 1 to 10 in a one-semester subject taught to juniors and seniors.
The material from Chapters 4 and 6 on rotating machines was used selectively,
so that students had "a foot solidly in the door" on this important subject
but also that the coverage could retain an orientation toward the needs of all
the diverse areas found in electrical engineering today. We have found the
material useful as the basis for early graduate work and as a starting point
in several courses related to electromechanics.
Finally, to those who open this book and then close it with the benediction,
"good material but unteachable," we apologize because to them we have
not made our point. Perhaps not as presented here, but certainly as it is
 Preface

represented here, this material is rich in teaching possibilities. The demands

on the teacher to see the subject in its total context, especially the related
problems that lie between the lines, are significant. We have taught this
subject many times to undergraduates, yet each term has been more enjoyable
than the last. There are so many ways in which drama can be added to the
material, and we do not need to ask the students (bless them) when we have
been successful in doing so.
In developing this material we have found lecture demonstrations and
demonstration films to be most helpful, both for motivation and for develop-
ing understanding. We have learned that when we want a student to see a
particular phenomenon it is far better for us to do the experiment and let
the student focus his attention on what he should see rather than on the
wrong connections and blown fuses that result when he tries to do the
experiment himself. The most successful experiments are often the simplest-
those that give the student an opportunity to handle the apparatus himself.
Every student should "chop up some magnetic field lines" with a copper
"axe" or he will never really appreciate the subject. We have also found that
some of the more complex demonstrations that are difficult and expensive
to store and resurrect each semester come through very well in films. In
addition to our own short films, three films have been produced professionally
in connection with this material for the National Committee on Electrical
Engineering Films, under a grant from the National Science Foundation, by
the Education Development Center, Newton, Mass.
Synchronous Machines: ElectromechanicalDynamics by H. H. Woodson
Complex Waves I: Propagation, Evanescence and Instability by J. R.
Melcher
Complex Waves II: Instability, Convection and Amplification by J. R.
Melcher
An additional film is in the early stages of production. Other films that
are useful have been produced by the Education Development Center for
the National Committee on Fluid Mechanics Films and for the College
Physics Film Program. Of particular interest, from the former series, is
Magnetohydrodynamicsby Arthur Shercliff.
A book like this can be produced only with plenty of assistance. We
gratefully acknowledge the help we received from many directions and hope
we have forgotten no one after seven years of work. First of all we want
to acknowledge our students with whom we worked as the material developed.
They are the one most essential ingredient in an effort of this sort. Next we
want to thank Dr. S. I. Freedman, Professor H. H. Richardson, and Dr.
C. V. Smith, Jr., for their assistance in framing worthwhile approaches to
several of our key topics. In seven years we have had the help of many able
 Preface

teachers in presenting this material to students. Their discussions and advice

have been most useful. In this category we want particularly to mention
Professors H. A. Haus, P. L. Penfield, D. C. White, G. L. Wilson, R. Gal-
lager, and E. Pierson and Doctors J. Reynolds, W. H. Heiser, and A. Kusko.
Professor Ketterer, who has taught this material at M.I.T. and the University
of Pennsylvania, Professors C. D. Hendricks and J. M. Crowley, who have
taught it at M.I.T. and the University of Illinois, and Professor W. D. Getty,
who has taught it at M.I.T. and the University of Michigan, have been most
generous with their comments. Messrs. Edmund Devitt, John Dressier, and
Dr. Kent Edwards have checked the correctness of many of the mathematical
treatments. Such a task as typing a manuscript repeatedly is enough to try
the patience of anyone. Our young ladies of the keyboard, Miss M. A. Daly,
Mrs. D. S. Figgins, Mrs. B. S. Morton, Mrs. E. M. Holmes, and Mrs. M.
Mazroff, have been gentle and kind with us.
A lengthy undertaking of this sort can be successful only when it has the
backing of a sympathetic administration. This work was started with the
helpful support of Professor P. Elias, who was then head of the Department
of Electrical Engineering at M.I.T. It was finished with the active encourage-
ment of Professor L. D. Smullin, who is presently head of the Department.
Finally, and most sincerely, we want to acknowledge the perseverance of
our families during this effort. Our wives, Blanche S. Woodson and Janet D.
Melcher, have been particularly tolerant of the demands of this work.
This book appears in three separately bound, consecutively paged parts
that can be used individually or in any combination. Flexibility is ensured
by including with each part a complete Table of Contents and Index. In
addition, for convenient reference, Parts II and III are supplemented by brief
appendices which summarize the relevant material from the preceding chap-
ters. Part II, Chapters 7 to 10, develops interactions between moving media
and fields with simple mechanical models that illustrate the dynamics of
continuum electromechanical systems.
H. H. Woodson
J. R. Melcher
Cambridge, Massachusetts
January 1968