Preface

Preface
There are men whose head is full

Of nothing, if not their science;
There are the learned of all trades.
I tell you without pretension:
Rather than learning too much,
Go and learn what really matters.
Hernandez, J., 1872, Martn Fierro,
Editorial de la Pampa, Buenos Aires (current text
in the original Spanish taken from the 1982 edition
by Bruguera Publishers, Barcelona).1
The need to provide instructors and students with a textbook on the classical prin-
ciples and the modern methods of analysis, modeling and simulation of mechanical
systems gave rise to The Dynamic Response of Linear Mechanical Systems. I came
across this need myself when I was assigned, in the late eighties, the teaching of the
undergraduate Dynamics of Vibrations course at McGill Universitys Department of
Mechanical Engineering, while one of the instructors was on sabbatical. This was
an interesting challenge, as I had never taken an undergraduate vibrations course
as such. In fact, I came from the 5-year Electromechanical Engineering Program
at the National Autonomous University of Mexico (abbreviated UNAM, from its
name in Spanish), where the teaching of vibrations was included in the 1-year
course on Applied Mechanics; this course comprised both kinematics and dynamics
of machines. Vibrations being the last topic in the syllabus, the instructor usually
rushed through it, the final examination hardly including a question on vibration
dynamics. In my senior year the curriculum underwent a radical updating, with
courses offered in semesters. This change gave me the opportunity to take a one-
semester course on Electromechanical Energy Conversion, which was about my first
and only exposure to the discipline of dynamics of systems as an undergraduate.
1 Verses 69236928, translated by the author.
vii
viii Preface
My first task as a designated instructor of undergraduate vibrations was to search

for the right textbook. I was overwhelmed by the vast bibliography on the subject.
To my surprise, with rare exceptions, all textbooks I consulted observed the same
pattern, which was originally set by Thomsons classical book, first published in
1948.2 As I was familiar with system theory, given that this was the minor I
took in my graduate studies of Mechanical Engineering (M.Eng., UNAM) and
Applied Mechanics (Ph.D., Stanford), I tried to make the connection between these
two closely related disciplines. I knew Cannons book [2], which makes such a
connection, this book thus appearing as the right choice. Given the size of the book
and its unusually broad scope, however, the students were intimidated, which forced
me to look for an alternative, more focused textbook, as I continued teaching that
course for several years afterwards. I found other books that somehow integrated
system theory with the dynamics of vibration,3 but I thought that there was more
to it that was not as yet in the textbooks. For one thing, I did not like the idea of
having to solve a generalized eigenvalue problem to find the natural frequencies and
the natural modes of a multi-degree-of-freedom system, as the generalized problem
does not necessarily lead to real, non-negative eigenvalues. Inmans book had an
interesting idea along these lines that I decided to pursue.
I thus undertook an in-depth research on a more intuitive, yet rigorous approach
to the teaching of vibration and, for that matter, mechanical-system analysis, that
would (a) exploit the mathematical knowledge required from the students, as
spelled out in the prerequisites (linear algebra and basic engineering mechanics);
(b) connect the theory of vibrations with the more general theory of systems; and
(c) resort to well-known graphical techniques of solving engineering problems. This
is how I decided to take a departure from traditional approaches to the teaching of
vibration dynamics. Features of the book along these lines are listed below:
Modeling is given due attentionmost books appear to take modeling for
granted. In doing this, a seven-step procedure is introduced that is aimed at
structuring the rather unstructured process of formulating mathematical models
of mechanical systems. I always insist in the need for a creative approach to
modeling, as no two engineers will always come up with the same model of the
same situation. Modeling, while based on sound science, is also an art, that can
only be developed by practice.
A system-theoretic approach is adopted in deriving the time response of the linear
mathematical models of the systems under study. In this regard, first what is
known in system-theoretic terminology as the zero-input responsedubbed the
free response in vibration analysisis obtained; next, the zero-state response is
derived by relying on the impulse response of the system under analysis. The
time response of a system to arbitrary excitations is then naturally derived in
convolution form upon resorting to a black-box approach that is applicable to all
2 The book has gone through many revisions, e.g., [1], but it keeps its original basic contents.
3 These were Meirovitch [3] and Inman [4].
Preface ix
linear, time-invariant dynamical systems. The fundamental concept of linearity,

which entails superposition and additivity, is stressed throughout the book.
The modal analysis and the time response of two-degree-of-freedom (two-dof)
systems is eased by invoking the standard, symmetric eigenvalue problem. Of
the many authors of books on vibrations, only Inman seems to have made a point
on the benefits of a symmetric eigenvalue problem, as opposed to its generalized
counterpart, in which the matrix in question is not symmetric. My approach
requires obtaining the positive-definite square root of the 2 2 mass matrix,
which is eased by resorting to (a) facts from linear algebra pertaining to the
analytic functions of matrices and (b) a graphical method that relies on the Mohr
circle, an analysis tool that is learned in elementary courses on solid mechanics.
The concept of frequency matrix, first found in Inmans book, although not
by this name, is introduced here, which helps the student use the Mohr circle
in conducting the modal analysis of the systems at hand. This concept was
published in a tutorial paper [5].
The time response of first- and second-order mechanical systems is derived in a
synthetic, i.e., constructive manner. What this means is that I do not follow the
classical math-book approach, under which the general response is derived as
the sum of a general solution, with undetermined coefficients, and a particular
solution. The downside of this approach is that it does not take into account the
causality of dynamical systems. I exploit causality by deriving the general time
response as a sum of the zero-input and the zero-state responses. The former is
derived, for first- and second-order systems, by means of an infinite series; this is
obtained in turn by successively differentiating the mathematical model at hand,
and reducing every higher-order derivative to a multiple of the first derivative of
the variable of interest for first-order systems; for second-order systems, every
higher-order derivative is reduced to a linear combination of this variable and
its first derivative. By evaluating these derivatives at the initial time, numerical
values for the coefficients of the series expansion are derived in terms of the
initial values of the problem at hand.
The time response of n-dof systems with a positive-definite stiffness matrix
is introduced rather informally, by resorting to the intuitive notion that the
response of these systems should be formally analogous to that of single-degree-
of-freedom systems. The analogy is achieved by means, again, of the concept
of frequency matrix. The time response of n-dof undamped systems is then
informally derived by replacing the natural frequency of single-dof systems
with the frequency matrix of n-dof systems. The underlying informality is then
justified by proving that the time response thus obtained is indeed the integral of
the system of governing ordinary differential equations (ODEs). The foregoing
proof is conducted by resorting to a basic concept of system theory: the response
thus obtained verifies both the ODEs and the initial conditions.
x Preface
The time response of n-dof damped systems, unfortunately, does not allow
for a straightforward derivation similar to that applicable to their undamped
counterparts. In this light, the time response of these systems is done first by
simulation, then by means of the Laplace transform and the concept of impulse
response.
Examples and exercises rely on modern computational toolboxes for both numer-
ical and symbolic computations; the powerful capabilities of readily available
commercial software for plotting are fully exploited.
Great care has been taken in producing drawings of mechanical systems, so
as to convey the most accurate information graphically. This feature should be
appreciated by students and instructors, as inaccurate information in a technical
document invariably leads to delays in the completion of a task.
Emphasis is placed on the logic of computations, and so, wherever needed,
procedures that can be implemented with commercial software are included.
While novel techniques are introduced throughout the book, classical approaches
are given due attention, as these are needed as a part of the learning process.
A common trend in the literature on the field is to be highlighted: with the aim of
bringing computers into the teaching of vibration analysis, many a textbook includes
code to calculate the time response of the systems of interest. The problem here
is that this code is, more often that not, nothing but the verbatim casting of the
time response formulas in computer language, thereby doing away with the actual
possibilities offered by computing hardware and software. An alternative approach
found in the literature is the numerical integration of the underlying systems of
linear ordinary differential equations (ODEs) using a RungeKutta algorithm. While
there is nothing essentially wrong with this approach, the use of such algorithms is
an unnecessary complication. Indeed, RungeKutta methods are suitable for the
integration of nonlinear ODEs; they do not exploit the linearity of the systems
encountered in a first course on dynamics modeling, analysis, and simulation,
thereby complicating the issue unnecessarily. We depart from these practices by
resorting to the concept of zero-order hold and by casting the numerical integration
of the underlying mathematical models in the context of discrete-time systems.
The outcome is that the problem is reduced to simple operationsadditions and
multiplicationsof arrays of numbers, i.e., vectors and matrices.
The book is accompanied by some MapleTM worksheets that illustrate: (a) the
discrete-time response of single-, two-, and three-degree-of-freedom systems; and
(b) the use of the Mohr circle in the derivation of the time response of undamped
two-dof systems. The worksheets are available at the Springer Extras website: http://
extras.springer.com/.
A Solutions Manual that includes solutions to selected problems accompanies
this book; it is made available to instructors.
Before closing, I would like to stress the philosiphy behind this book: knowledge
being such a complex, experiential phenomenon [6]it cannot be downloaded,
contrary to popular beliefit cannot be reduced to a set of ad-hoc rules; it can be
Contents xi
transmitted, though, via its underlying principles. This is probably what the gaucho
Martn Fierro had in mind when giving wise advice to his son in the verses quoted
above.
For the completion of this manuscript and its supporting materials, many
people are to be acknowledged, from undergraduate interns to graduate assistants,
postdoctoral fellows and colleagues. In the first versions of the manuscript, as a set
of Lecture Notes, Robert Lucyshyn, Meyer Nahon, Abbas Fattah and Eric Martin,
then Ph.D. students, made valuable contributions both with rigorous criticism and
suggestions to improve the pedagogical value of the material. Svetlana Ostrovskaya
played a key role, first as a Ph.D. student and then as a Postdoctoral Fellow, as editor
and contributor to the material in general. In the last stages of the editing, Vikram
Chopra, an undergraduate intern at the timenow a Ph.D. studentdiligently
edited the whole manuscript and assembled the Solutions Manual. The version
out of which this manuscript was produced is due to the diligent work of Danial
Alizadeh, a Ph.D. student in my research group. The professional work behind
the figures is credited to Max A. Gonzalez-Palacios, who set the standards. These
individuals, and many others that unavoidably escape my memory, are given due
recognition in making this book a reality.
Montreal Jorge Angeles
References
1. Thomson WT (1993) Theory of vibration with applications. Prentice-Hall, Upper Saddle River
2. Cannon RH (1967) Dynamics of physical systems. McGraw-Hill Book Company, New York
3. Meirovitch L (1986) Elements of vibration analysis. McGraw-Hill Book Company, New York
4. Inman DJ (2007) Engineering vibration. Prentice-Hall, Upper Saddle River
5. Angeles J (1992) On the application of Mohrs circle to two-dof vibration analysis. A tutorial.
J Sound Vibr 154(3):556567
6. Russell B (1992) In: Eames ER (ed) Theory of knowledge. The 1913 manuscript. In collabora-
tion with Blackwell, K., Routledge, London/New York
http://www.springer.com/978-1-4419-1026-4

Preface

Uploaded by

Copyright:

Available Formats

Preface

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Preface

Uploaded by

Copyright:

Available Formats

Preface

There are men whose head is full

1 Verses 69236928, translated by the author.

My first task as a designated instructor of undergraduate vibrations was to search

linear, time-invariant dynamical systems. The fundamental concept of linearity,

Montreal Jorge Angeles

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.