1 Optimization & Anti-Optimization of Structures Under Uncertainty - Isaac Elishakoff PDF

Optimization and
Anti-Optimization of
Structures Under Uncertainty
P678 tp.indd 1 1/29/10 12:16:33 PM

This page intentionally left blank
Optimization and
Anti-Optimization of
Structures Under Uncertainty
Isaac Elishakoff
Florida Atlantic University, USA
Makoto Ohsaki
Kyoto University, Japan
Imperial College Press

ICP
P678 tp.indd 2 1/29/10 12:16:34 PM

A-PDF Merger DEMO : Purchase from www.A-PDF.com to remove the watermark
Published by
Imperial College Press
57 Shelton Street
Covent Garden
London WC2H 9HE
Distributed by
World Scientific Publishing Co. Pte. Ltd.
5 Toh Tuck Link, Singapore 596224
USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601
UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE
British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.
OPTIMIZATION AND ANTI-OPTIMIZATION OF STRUCTURES UNDER

UNCERTAINTY
Copyright 2010 by Imperial College Press
All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means,
electronic or mechanical, including photocopying, recording or any information storage and retrieval
system now known or to be invented, without written permission from the Publisher.
For photocopying of material in this volume, please pay a copying fee through the Copyright
Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to
photocopy is not required from the publisher.
ISBN-13 978-1-84816-477-2
ISBN-10 1-84816-477-7
Printed in Singapore.
January 28, 2010 16:28 World Scientific Book - 9.75in x 6.5in optimization
To Aron Lev, Chana Batsheva, Yehudit Lea, Menachem Mendel,

Moshe Chaim, and Tamar Devorah
I.E.
To Noriko, Takeshi, and Satsuki

M.O.
v
Preface
Uncertainties appear everywhere in the model. When using a mathe-

matical model, careful attention must be given to the uncertainties in
the model. (Feynman, 1988)
New methods for treating uncertainty will become important in vir-
tually all branches of mechanics. (Oden, Belytschko, Babuska and
Hughes, 2003)
Scientific progress usually can only be achieved by experts of different
disciplines working together. (Eschenauer and Olhoff, 2001)
The ethics of our profession today does not allow any design for a
structure without optimization. (Mungan, 2001)
Researchers in applied mechanics, whether in the field of aerospace, architecture,

civil, mechanical or ocean engineering, invariably adopt the either/or style. Namely,
they devote themselves either to linear or to nonlinear analysis of the structure they
are dealing with, they are engaged in analyzing it either in the elastic or in the
inelastic range; they deal either with its static or with its dynamic behavior.
Along the same lines, researchers classify themselves according to the determin-
istic or non-deterministic nature of the problems they are tackling. The former
stipulate that the geometry, loads, boundary conditions, material characteristics
involved in a problem are fully specified; they stress that they are concerned with
understanding the phenomenon at hand, and then providing methods of solution.
To achieve such a basic comprehension of what is going on, they make (often far-
reaching) assumptions: that the parameters involved are fully specified, that they
take on large or small values, etc. Such an approach, which may seem unjustifi-
able at first glance, often leads to breakthrough discoveries in understanding the
phenomenon under study.
The difference between the deterministic and non-deterministic approaches lies
in the fact that the deterministic design attempts to retain its determinism until the
very last stage of design, whereas the non-deterministic analysts maintain that there
vii
viii Optimization and Anti-Optimization of Structures under Uncertainty
are no two structures with identical material and geometric properties, boundary
conditions, or loading and, therefore, uncertainty analysis is called for.
In reality, this deterministic versus non-deterministic attitude is a fallacy. To
justify this undiplomatic statement, we need deeper insight into the essence of de-
terministic analysis. In it, after careful evaluation of displacements, strains, and
stresses (often with many significant digits since computer outputs consist of
numbers), the engineers compare the latter against the allowable stress; i.e., the ex-
perimentally measured yield- or ultimate-level multiplied by some fudge coefficient
called the safety factor (as in better safe than sorry). Use of this factor in text-
books on engineering mechanics is invariably motivated by insufficient knowledge
(dont we always have to make analytic assumptions of various sorts?), or by un-
certainty (yes, deterministic analysts do pay lip service to the property they claim
to neglect!) in loads, geometric characteristics, or scatter in material properties.
Thus, deterministic analysis, strictly speaking, is not as deterministic as ad-
vertised. Within it, uncertainty is introduced as dessert or, as in Bolotins words
(Bolotin, 1961), via the back door.
By contrast, in non-deterministic design, uncertainty figures as a legitimate
ingredient throughout the whole design process, specifically during the thinking,
analysis and design stages.
One may conclude, therefore, that non-deterministic analysis is more honest than
its so-called deterministic counterpart as it places its cards on the table from the very
beginning. As for the latter, in view of the above reasoning, it would be appropriate
to dub it pseudo-deterministic rather than deterministic. If experience enables
us to establish the precise value of a safety factor, then the pseudo-deterministic
unknowns are nothing less than the lumped, integral equivalent of the inherently
present uncertainty in engineering problems.
Naturally, the advent of the safety factor was at the time a major breakthrough
in design. Despite its obvious arbitrariness, it allowed us to put up a wall against
failure. At the same time, this dogmatism implies that if the actual stress turns out
to be below the allowable level, then there can be no failure, and the structure will
be safe throughout its exploitation.
But how can we establish the safety factor, except by experience or trial-and-
error (one could sarcastically ask at this stage: is this a trial or an error?). According
to Norton (2000), choosing the safety factor is often a confusing proposition for the
beginning engineer, and not only for the beginning one, we may add. According to
Bruhn (1975),
[the safety] factor is designed to arbitrarily account for items such

as material properties, manufacturing differences, since no two
parts can be made exactly the same; uncertainties in the loading
environment; and unknowns in the internal load and stress distri-
butions.
Preface ix
As we observe, even with the manner of explanation of the safety factor, this
concept is quite critical within the semi-deterministic framework for it utilizes ad-
jectives like arbitrary (Bruhn, 1975) and confusing (Norton, 2000).
Freudenthal (1968) emphasized:
It seems absurd to strive for more and more refinement of meth-

ods of stress analysis, if in order to determine the dimension of the
structural elements, its results are subsequently compared with a
so-called working stress, derived in a rather crude manner by divid-
ing the values of somewhat dubious material parameters obtained
in a conventional material test by a still more dubious empirical
number, called a safety factor.
Bolotin (1961) had written,
The values of safety factors, as well as closely associated values

of design loads and design resistances, were imposed and modified
mainly empirically, by way of generalization of long-term experi-
ence in exploitation of structures. Yet, as is seen from the essence
of the problem, there are in principle, also theoretical approaches
possible, with wide application of the theory of probability and
mathematical statistics.
We must immediately note that probability theory and mathematical statistics

are not the only avenues for the theoretical approach to the safety factors.
Non-deterministic analysis enables us to recast the safety factor approach in
terms of other quantities. For example, Freudenthal (1957) and Rzhanitsyn (1947)
pioneered an interpretation of the safety factor in terms of reliability based on the
probability theory and mathematical statistics, and introduced the so-called central
safety factor. (It was shown by Elishakoff (2004) that probabilistic mechanics offers
four possible interpretations of the safety-factor concept.)
Probabilistic analysis is not the only game in town, in the non-deterministic con-
text. While the probability densities of the random variables involved in it are often
not known, simpler characteristics like the sets within which the parameters vary
may be specified. Such a description of uncertainty (Bulgakov, 1942) is referred
to as an unknown-but-bounded one. In the West it was independently developed
by Boley (1966a) for static thermoelastic problems; and by Drenick (1968) and
Shinozuka (1970) for dynamic ones. Some of these and other results in applied me-
chanics were summarized by Ben-Haim and Elishakoff (1990), who also developed
some new material. They concentrated their attention on models of uncertainty in
which the variations were treated as convex sets, and yielded the most and least
favorable stresses, strains or displacements over such a convex set. This approach
has been extended by Elishakoff, Lin and Zhu (1994e) by combining it with proba-
bilistic analysis; and by Elishakoff, Li and Starnes (2001) via systematic comparison
x Optimization and Anti-Optimization of Structures under Uncertainty
with the probabilistic methodology.

Elishakoff (1994) interpreted the safety factor within unknown-but-bounded un-
certainty as the interval [nL , nU ], between lower L and upper U bounds rep-
resenting the ratios of the yield stress to the least and most favorable response,
respectively, in which case the yield stress can safely serve as the design parameter.
Likewise, Elishakoff and Ferracuti (2006a, 2006b) provided interpretation of the
safety factor within fuzzy sets. Here, too, as in probabilistic analysis, there are four
possible interpretations of the fuzzy safety factor.
This book is exclusively concerned with the unknown-but-bounded uncertainty.
Chapter 1 explains in detail why this approach was chosen. Here, we content
ourselves with noting that in the case of random variables with bounded support,
it is shown that probabilistic design and the unknown-but-bounded approach yield
coincident results, when the required reliability tends to unity. Accordingly the
second version was chosen as the less complicated of the two, in keeping with Albert
Einsteins sage precept: Everything should be made as simple as possible, but not
simpler. In the past decade or so, many papers were devoted to unknown-but-
bounded uncertainty in applied mechanics. Several books were published as well.
Regrettably, some of the authors, instead of concentrating on the safety factor as the
main tool of structural design and professional communication, turned to alternative
concepts or actually introduced new (and hardly needed) ones, thereby diverting
research activities from the really relevant problems (see the instructive title of the
book by Sokal and Bricmont (1998)).
We would like next to consider the optimization/anti-optimization problem, but
before proceeding further it would be instructive to reproduce the text of a letter
sent by Dr. Rudolf F. Drenick (2001), Professor Emeritus of Brooklyn Polytechnic
Institute, to Dr. Izuru Takewaki, Professor of Kyoto University:
You might be amused by the story of how I came to do the work

on the critical excitation. In (I think) 1965 I participated in a
Japan/U.S. workshop on applied stochastics. After two days of
lectures, one of the observers at our meeting made some comments
which I found very interesting. He said something like this. My
name is Ozawa and I work for the Tokyo Building Licensing Bureau.
I am forever visited by architects who show me plans of buildings
they plan to construct and I am supposed to tell them whether
their plans are good or bad, and I have never really known what to
tell them. Now I have listened to you gentlemen for two days and
I still dont know what to tell them. There followed a great deal
of discussion among the workshop participants but in the end the
discussion leader, Professor Bogdanov, said I am very sorry, Mr.
Ozawa, but we cant help you.
I went home from the workshop and began to think of how I could
Preface xi
deal with Mr. Ozawas problem. I concluded that any method I

should suggest to him would have to be, first of all, practical. I
also felt that it should be as distribution-free as possible. After
some thought the concept of the critical excitation came to me.
(The term critical excitation was not my idea. It was suggested
by Prof. Penzien.) My first results were largely distribution-free,
as you know, but they were not practical. They were much too
conservative to be useful.
This worst-case design was never adopted by the nuclear industry because of
its extreme conservatism, as Drenick informed one of us (I.E.). He, accordingly,
became very pessimistic about the likelihood of the engineering profession ever
adopting it. He was however very encouraging, and so kind as to write the foreword
to the book by Ben-Haim and Elishakoff (1990) in which he stated perhaps with
some measure of exaggeration: Their approach is novel and highly welcome. In my
opinion, it is inevitable that it, and its extensions, will dominate the future practice
of engineering.
In consequence of the above-mentioned book, it was realized by one of us (I.E.)
that in order to make this worst-case scenario research practical, it needed an infu-
sion of new blood. (It was realized that unknown-but-bounded uncertainty analysis
is a defacto as anti-optimization process, the reverse of searching for the best so-
lution.) It became also apparent that it is impractical to content oneself with the
least favorable static or dynamic response, or with the worst possible buckling load.
Rather a structure should be so designed as to minimize the first and maximize the
second in other words, adherence to the time-honored precept of making the best
out of the worst. It is gratifying to read that this sentiment is now shared by other
investigators, for example, Ben-Tal and Nemirovski (2002):
The worst-case oriented interval model of uncertainty looks too

conservative. Indeed, when the perturbations in coefficients of an
uncertain linear inequality are of stochastic nature, it is highly im-
probable that they will simultaneously take the most dangerous val-
ues.
These considerations eventually gave rise to the optimization and anti-

optimization terminology or better yet optimization with anti-optimization. If there
is an uncertainty, one should, as it were, participate in two dances concurrently,
doing both optimization and anti-optimization steps.
Such a hybrid approach would be less conservative than its worst-scenario coun-
terpart, hence is much more likely to gain credibility and acceptance among re-
searchers and practicing engineers.
Indeed, it is often said that if a researcher cannot explain his or her research
to a lay person, then that researcher does not yet fully understand the subject
xii Optimization and Anti-Optimization of Structures under Uncertainty
under discussion (Einstein, here too, comes handy: you do not really understand
something unless you can explain it to your grandmother).
Indeed, would not it be an engineering sin, as it were, not to improve the worst
scenario and leave it intact especially if one has to save expenditure? Only very
rich companies, or those in positions of exceptional responsibility, can afford exclu-
sive reliance on anti-optimization.
One argument against optimization is that if it is applied to a specific loading
condition, the result may prove unsatisfactory under a different condition. In such a
case, it suffices to add the latter and repeat the procedure for the new circumstances.
Alternatively, one can introduce uncertainty and assign bounds so as to obtain the
optimal solution under the worst-case scenario.
Another criticism concerns optimization against nonlinear buckling. As ex-
plained in detail in Chap. 5, optimization against buckling does not always yield
imperfection-sensitive structures, the resulting sensitivity being in fact lower than
that of a non-optimal design (Ohsaki, 2002c). However, the buckling load of a
perfect system is drastically increased by optimization, and its counterpart for an
imperfect system is little affected by sensitivity changes.
The combined optimistic-pessimistic approach can be characterized as follows:
optimists build ships; pessimists build lifeboats. Both are needed for safe sailing.
Recall the overly optimistic view on the Titanic, which was claimed to be so robust
that even God couldnt sink it! As a result, an insufficient number of lifeboats was
on board during its maiden voyage. This complacency cost over 1,500 lives.
We advocate a combination of healthy doses of optimism (represented by opti-
mization) and pessimism (anti-optimization).
The above discussion gave rise to the notion that uncertainty analysis is re-
ducible to one of the three vertices of the uncertainty triangle (Elishakoff, 1990,
1998b), shown below. These three approaches do not represent three non-interesting
magistrata!
Although the first paper co-authored by one of us (Ben-Haim and Elishakoff,
1989b) was presented at two conferences in 1988 and appeared in 1989, we had been
able, on an earlier occasion (Elishakoff, 1983, p. 42), to compare the probabilistic
and worst-case designs on a simple example included on the Statistical Methods
in Elasticity course at the Technion - Israel Institute of Technology since 1973.
The monograph by Ben-Haim and Elishakoff (1990) did not correlate probabilistic
and convex modelings. Only later, in the paper by Elishakoff, Cai and Starnes
(1994a), was a direct comparison made between probabilistic and anti-optimization
techniques on the nonlinear boundary-value problem (see also its generalization by
Qiu, Ma and Wang (2006a)). It turned out that if the probabilistic information is
available and the required reliability is not excessively high, anti-optimization may
be conservative. However, if near-unity reliability is required, the two approaches
tend to yield close or coincident results.
It was also realized that these seemingly competing approaches are nevertheless
Preface xiii
amenable to a certain measure of cooperation. An instructive example is the case

of the turbulent excitation experienced by the space shuttle weather protection
systems at the Kennedy Space Center. This excitation is inherently random and
must be analyzed probabilistically, but inadequate data is available on its cross-
spectral density. This shortage was remedied (Elishakoff, Lin and Zhu, 1994e) by
means of the anti-optimization approach, which yielded the least favorable mean-
square displacement of the weather protection panels.
In another combination, Kosko (1990), Maglaras, Nikolaidis, Haftka and Gudney
(1997), Chiang, Dong and Wong (1987), and Wood, Antonnson and Beck (1990)
watched the fuzzy-sets approach against the probabilistic one, theoretically and/or
experimentally. Kwakernaak (1978), Haldar and Reddy (1992), Savoia (2002), and
Moller and Beer (2004) combined the two approaches. Finally, Fang, Smith and
Elishakoff (1998) and Tonon, Bernardini and Elishakoff (2001) utilized all three
approaches for their analyses.
Bernardini (1999) and Tonon, Bae, Grandhi and Pettit (2006) claim that all
three methodologies stem from a single theory of random sets (see Matheron (1975)
and Dubois and Prade (1991)). Recently, Ben-Haim (2001) also adopted combined
probabilistic and non-probabilistic analyses.
As is clearly seen from the above, probabilistic and non-probabilistic analyses
of uncertainty are not irreconcilable and can complement each other when deemed
useful. Such reconciliation was in fact advocated by Elishakoff (1995b):
Convex modeling of uncertainty, and in general the set-theoretic

modeling of it (the constraints should not necessarily be treated
as convex), do not support probabilistic ideas. Convex modeling
rather complements both the probabilistic approach and the fuzzy
subsets based treatment.
Theory of
Probability and
Randon Processes
Anti-Optimization with
Fuzzy Sets Set Theoretical Aproach
(Least Favorable Responses)
Uncertainty triangle (Elishakoff, 1990).

xiv Optimization and Anti-Optimization of Structures under Uncertainty
One should constantly bear in mind that just as pseudo-deterministic analy-

sis incorporates uncertainty via a lumped number (namely the safety factor), the
probabilistic approach is partly deterministic, as not all parameters are considered
variable. Does one have to consider all parameters of a structure as uncertain,
thereby covering the general case? Some research establishments take this line and
present on their slides perhaps tens of random variables with arbitrary probabilistic
distributions to choose from! Its Excellency the Computer then provides the desired
output in seconds. Such excessive zeal is unjustifiable, seeing that no probabilistic
dependencies are either known or incorporated into the programs! As Babuska,
Nobile and Tempone (2005) emphasize,
whenever the input data belong to an infinite dimensional space

(they might be functions of position and/or time), their probabilis-
tic characterization must include knowledge of the cross-correlation
of the values that the data can take at different points in space and
time. In this case, the solution of a stochastic model becomes
quickly too costly.
Still, we should be thankful that the innocent number , for example, is not
declared a random variable with mean value 3.14... and zero variance! One should
refrain from considering all of them as random variables unless they have sufficiently
large variability to justify the non-deterministic approach. Likewise, one should
abandon the idea (attractive for beginners) of treating all parameters as uncertain,
even if some of them have sufficiently small variability.
In conclusion, one has to maintain a delicate and healthy balance in modeling a
system. Bolotin (1969), for example, is against both underestimation and overesti-
mation of uncertainty analysis. We hope that this book will serve as a beacon for
practicing engineers and researchers, and help them reevaluate their thinking and
practices on uncertainty analysis.
The authors are indebted to the Japan Society for Promotion of Science (JSPS)
established by the late Emperor Showa (Hiro-Hito) in 1932. The JSPS, long before
the modern trends of globalization, made it possible for non-Japanese researchers
to spend time in the Land of the Rising Sun with their host counterparts, thereby
enabling both sides to elevate their horizons through collaboration. I.E. is thankful
to Kyoto University which provided the hospitable atmosphere for the joint research.
We record our appreciation to the co-authors of some joint papers that found
their way into our monograph. These are Professor Dan Givoli of the Department of
Aerospace Engineering, Technion - Israel Institute of Technology, for his kind agree-
ment for us to reproduce the shortened version of the paper Stress concentration
at nearly circular hole with uncertain irregularities, by D. Givoli and I. Elishakoff,
1992; Professor Nobuhiro Yoshikawa of the Institute of Industrial Science, Univer-
sity of Tokyo for his kind agreement for us to reproduce the contents of the paper
Worst case estimation of homology design by convex analysis, by N. Yoshikawa,
Preface xv
I. Elishakoff and S. Nakagiri, 1998; Professor Alessandro Baratta and Professor

Giulio Zuccaro of the Department of Scienza delle Costruzioni of the University of
Naples Federico II, for their kind agreement to draw portions from two papers: A
generalization of the DrenickShinozuka model for bounds on the seismic response
of a single-degree-of-freedom systems, by A. Baratta, I. Elishakoff, G. Zuccaro and
M. Shinozuka, 1998, and Antioptimization of earthquake excitation and response,
by G. Zuccaro, I. Elishakoff and A. Baratta, 1998 (we also thank Hindawi Pub-
lishers for allowing us to reproduce the results and figures of the paper); Professor
Massimiliano Zingales of the Dipartimento di Ingegneria Strutturale e Geotecnica,
Università degli Studi di Palermo, for his kind agreement to draw materials from the
following joint papers: Contrasting probabilistic and anti-optimization approaches
in an applied mechanics problem, by I. Elishakoff and M. Zingales, 2003, Anti-
optimization versus probability in an applied mechanics problem: Vector uncer-
tainty, by M. Zingales and I. Elishakoff, 2000, and Hybrid aeroelastic optimization
and antioptimization, by M. Zingales and I. Elishakoff, 2001; Professor Kiyohiro
Ikeda of the Department of Civil Engineering, Tohoku University for his kind agree-
ment to include materials in the paper Imperfection sensitivity analysis of hill-top
branching with many symmetric bifurcation points, by M. Ohsaki and K. Ikeda,
2006; Dr. Jingyao Zhang of the Department of Architecture and Urban Design,
Ritsumeikan University, for his kind agreement for presenting summaries of the
papers Optimization and anti-optimization of forces in tensegrity structures, by
M. Ohsaki, J. Y. Zhang and I. Elishakoff, 2008, and Optimal measurement po-
sitions for identifying stress distribution of membrane structures using cable net
approximation, by J. Y. Zhang, M. Ohsaki and Y. Araki, 2004.
The authors would also like to thank Prof. Yoshihiro Kanno of the Department
of Mathematical Informatics, University of Tokyo, for checking the details of the
manuscript; Dr. Takao Hagishita of Mitsubishi Heavy Industries (former graduate
student of Kyoto University), Mrs. Juxi Hu of Beijing University of Aeronautics and
Astronautics, and Ms. Melanie Dolly of the IFMAFrench Institute of Advanced
Mechanics for typing some parts of the book; Ms. Shawn Pannel and Mr. Joshua
Kahn of Florida Atlantic University for redrawing the figures; Mr. Eliezer Goldberg
of the Technion - Israel Institute of Technology, for checking the preface and Chapter
10 of this book; Ms. Sarah Haynes of Imperial College Press for her assistance in
bringing the manuscript to its final form.
We appreciate the support from the National Science Foundation over several
years, as well as the National Center for Earthquake Engineering Research. Any
opinions, findings, and conclusions or recommendations expressed by this mono-
graph are those of the authors and do not necessarily reflect the views of these
agencies.
May 2009 Isaac Elishakoff, Makoto Ohsaki


Contents
Preface vii
1. Introduction 1
1.1 Probabilistic Analysis: Bad News . . . . . . . . . . . . . . . . . . 1
1.2 Probabilistic Analysis: Good News . . . . . . . . . . . . . . . . . 10
1.3 Convergence of Probability and Anti-Optimization . . . . . . . . 13
2. Optimization or Making the Best in the Presence of

Certainty/Uncertainty 17
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2 What Can We Get from Structural Optimization? . . . . . . . . 19
2.3 Definition of the Structural Optimization Problem . . . . . . . . 20
2.4 Various Formulations of Optimization Problems . . . . . . . . . 23
2.4.1 Overview of optimization problems . . . . . . . . . . . . 23
2.4.2 Classification of optimization problems . . . . . . . . . . 24
2.4.3 Parametric programming . . . . . . . . . . . . . . . . . 26
2.4.4 Multiobjective programming . . . . . . . . . . . . . . . . 28
2.5 Approximation by Metamodels . . . . . . . . . . . . . . . . . . . 30
2.6 Heuristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.6.1 Overview of heuristics . . . . . . . . . . . . . . . . . . . . 31
2.6.2 Basic approaches of single-point search heuristics . . . . 32
2.6.3 Simulated annealing . . . . . . . . . . . . . . . . . . . . 35
2.7 Classification of Structural Optimization Problems . . . . . . . . 36
2.8 Probabilistic Optimization . . . . . . . . . . . . . . . . . . . . . . 39
2.9 Fuzzy Optimization . . . . . . . . . . . . . . . . . . . . . . . . . 41
3. General Formulation of Anti-Optimization 47

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.2 Models of Uncertainty . . . . . . . . . . . . . . . . . . . . . . . . 49
3.3 Interval Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
xvii
xviii Optimization and Anti-Optimization of Structures under Uncertainty
3.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.3.2 A simple example . . . . . . . . . . . . . . . . . . . . . . 51
3.3.3 General procedure . . . . . . . . . . . . . . . . . . . . . . 52
3.4 Ellipsoidal Model . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.4.1 Definition of the ellipsoidal model . . . . . . . . . . . . . 57
3.4.2 Properties of the ellipsoidal model . . . . . . . . . . . . . 58
3.5 Anti-Optimization Problem . . . . . . . . . . . . . . . . . . . . . 61
3.6 Linearization by Sensitivity Analysis . . . . . . . . . . . . . . . 63
3.6.1 Roles of sensitivity analysis in anti-optimization . . . . . 63
3.6.2 Sensitivity analysis of static responses . . . . . . . . . . 64
3.6.3 Sensitivity analysis of free vibration . . . . . . . . . . . 67
3.6.4 Shape sensitivity analysis of trusses . . . . . . . . . . . . 68
3.7 Exact Reanalysis of Static Response . . . . . . . . . . . . . . . . 69
3.7.1 Overview of exact reanalysis . . . . . . . . . . . . . . . . 69
3.7.2 Mathematical formulation based on the inverse of the
modified matrix . . . . . . . . . . . . . . . . . . . . . . . 70
3.7.3 Mechanical formulation based on virtual load . . . . . . . 74
4. Anti-Optimization in Static Problems 77

4.1 A Simple Example . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.2 Boleys Pioneering Problem . . . . . . . . . . . . . . . . . . . . . 80
4.3 Anti-Optimization Problem for Static Responses . . . . . . . . . 85
4.4 Matrix Perturbation Methods for Static Problems . . . . . . . . 86
4.5 Stress Concentration at a Nearly Circular Hole with Uncertain
Irregularities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.5.2 An asymptotic solution . . . . . . . . . . . . . . . . . . . 91
4.5.3 A worst-case investigation . . . . . . . . . . . . . . . . . 97
4.6 Anti-Optimization of Prestresses of Tensegrity Structures . . . . 102
4.6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 102
4.6.2 Basic equations . . . . . . . . . . . . . . . . . . . . . . . 103
4.6.3 Anti-optimization problem . . . . . . . . . . . . . . . . . 107
4.6.4 Numerical examples . . . . . . . . . . . . . . . . . . . . . 109
5. Anti-Optimization in Buckling 113

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
5.2 A Simple Example . . . . . . . . . . . . . . . . . . . . . . . . . . 114
5.3 Buckling Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 116
5.4 Anti-Optimization Problem . . . . . . . . . . . . . . . . . . . . . 117
5.5 Worst Imperfection of Braced Frame with Multiple Buckling
Loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
5.5.1 Definition of frame model . . . . . . . . . . . . . . . . . . 118
Contents xix
5.5.2 Worst imperfection of optimized frame . . . . . . . . . . 120

5.5.3 Mode interaction . . . . . . . . . . . . . . . . . . . . . . 123
5.5.4 Worst-case design and worst imperfection under stress
constraints . . . . . . . . . . . . . . . . . . . . . . . . . . 126
5.6 Anti-Optimization Based on Convexity of Stability Region . . . 128
5.7 Worst Imperfection of an Arch-type Truss with Multiple Member
Buckling at Limit Point . . . . . . . . . . . . . . . . . . . . . . . 133
5.7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 133
5.7.2 Hilltop branching point of perfect system . . . . . . . . . 134
5.7.3 Imperfection sensitivity of hilltop branching point . . . . 135
5.7.4 Worst imperfection . . . . . . . . . . . . . . . . . . . . . 136
5.7.5 Worst imperfection of an arch-type truss . . . . . . . . . 139
5.8 Some Further References . . . . . . . . . . . . . . . . . . . . . . . 144
6. Anti-Optimization in Vibration 145

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
6.2 A Simple Example of Anti-Optimization for Eigenvalue
of Vibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
6.2.1 Anti-optimization for forced vibration . . . . . . . . . . . 148
6.3 Bulgakovs Problem . . . . . . . . . . . . . . . . . . . . . . . . . 149
6.4 Non-probabilistic, Convex-Theoretic Modeling of Scatter in
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . 150
6.4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 150
6.4.2 Basic equations for vibrating viscoelastic beam . . . . . . 151
6.4.3 Application to a simply supported beam . . . . . . . . . 155
6.4.4 Least and most favorable responses . . . . . . . . . . . . 157
6.4.5 Numerical examples and discussion . . . . . . . . . . . . 160
6.5 Anti-Optimization of Earthquake Excitation and Response . . . 163
6.5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 163
6.5.2 Formulation of the problem . . . . . . . . . . . . . . . . . 164
6.5.3 Maximum structural response . . . . . . . . . . . . . . . 166
6.5.4 Ellipsoidal modeling of data . . . . . . . . . . . . . . . . 166
6.6 A Generalization of the DrenickShinozuka Model for Bounds on
the Seismic Response . . . . . . . . . . . . . . . . . . . . . . . . 175
6.6.1 Preliminary comments . . . . . . . . . . . . . . . . . . . 175
6.6.2 Credible accelerograms . . . . . . . . . . . . . . . . . . . 180
6.6.3 Application . . . . . . . . . . . . . . . . . . . . . . . . . . 184
6.6.4 Discussion and conclusion . . . . . . . . . . . . . . . . . 188
6.7 Aeroelastic Optimization and Anti-Optimization . . . . . . . . . 190
6.7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 190
6.7.2 Deterministic theoretical analysis . . . . . . . . . . . . . 191
6.7.3 Stability analysis within two-term approximation . . . . 193
xx Optimization and Anti-Optimization of Structures under Uncertainty
6.7.4 Convex modeling of uncertain moduli . . . . . . . . . . 195

6.7.5 Anti-optimization problem: polygonal region
of uncertainty . . . . . . . . . . . . . . . . . . . . . . . . 197
6.7.6 Anti-optimization problem: ellipsoidal region
of uncertainty . . . . . . . . . . . . . . . . . . . . . . . . 200
6.7.7 Minimum weight design . . . . . . . . . . . . . . . . . . . 201
6.7.8 A numerical example . . . . . . . . . . . . . . . . . . . . 206
6.7.9 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 208
7. Anti-Optimization via FEM-based Interval Analysis 211

7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
7.2 Interval Analysis of MDOF Systems . . . . . . . . . . . . . . . . 212
7.3 Interval Finite Element Analysis for Linear Static Problem . . . 214
7.4 Interval Finite Element Analysis of Shear Frame . . . . . . . . . 217
7.4.1 Basic equations . . . . . . . . . . . . . . . . . . . . . . . 217
7.5 Interval Analysis for Pattern Loading . . . . . . . . . . . . . . . 221
8. Anti-Optimization and Probabilistic Design 227

8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
8.2 Contrasting Probabilistic and Anti-Optimization Approaches . . 229
8.2.1 Problem formulation . . . . . . . . . . . . . . . . . . . . 229
8.2.2 Probabilistic analysis . . . . . . . . . . . . . . . . . . . . 231
8.2.3 Uniformly distributed random initial imperfections . . . 232
8.2.4 Random initial imperfections with truncated exponential
distribution . . . . . . . . . . . . . . . . . . . . . . . . . 240
8.2.5 Random initial imperfection with generic
truncated distribution . . . . . . . . . . . . . . . . . . . . 244
8.2.6 Buckling under impact load: Anti-optimization by
interval analysis . . . . . . . . . . . . . . . . . . . . . . . 245
8.2.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 246
8.3 Anti-Optimization Versus Probability: Vector Uncertainty . . . 247
8.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 247
8.3.2 Deterministic analysis . . . . . . . . . . . . . . . . . . . . 248
8.3.3 Probabilistic analysis . . . . . . . . . . . . . . . . . . . . 250
8.3.4 Initial imperfections with uniform probability density:
Rectangular domain . . . . . . . . . . . . . . . . . . . . . 251
8.3.5 Initial imperfections with general probability density:
Rectangular domain . . . . . . . . . . . . . . . . . . . . . 262
Contents xxi
8.3.6 Initial imperfection with uniform probability density

function: Circular domain . . . . . . . . . . . . . . . . . 264
8.3.7 Initial imperfections as interval variables: Interval
analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
8.3.8 Initial imperfections as convex variables: Circular
domain . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
8.3.9 Concluding remarks . . . . . . . . . . . . . . . . . . . . . 272
9. Hybrid Optimization with Anti-Optimization under Uncertainty or

Making the Best out of the Worst 273
9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
9.2 A Simple Example . . . . . . . . . . . . . . . . . . . . . . . . . . 274
9.3 Formulation of the Two-Level OptimizationAnti-Optimization
Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
9.4 Algorithms for Two-Level OptimizationAnti-Optimization . . . 277
9.4.1 Cycle-based method . . . . . . . . . . . . . . . . . . . . . 277
9.4.2 Methods based on monotonicity or convexity/concavity . 278
9.4.3 Other methods . . . . . . . . . . . . . . . . . . . . . . . . 279
9.5 Optimization against Nonlinear Buckling . . . . . . . . . . . . . 280
9.5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 280
9.5.2 Problem formulation . . . . . . . . . . . . . . . . . . . . 281
9.6 Stress and Displacement Constraints . . . . . . . . . . . . . . . 288
9.7 Compliance Constraints . . . . . . . . . . . . . . . . . . . . . . . 291
9.7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 291
9.7.2 Optimization problem and optimization algorithm . . . 294
9.8 Homology Design . . . . . . . . . . . . . . . . . . . . . . . . . . 298
9.8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 298
9.8.2 Deterministic loading . . . . . . . . . . . . . . . . . . . . 300
9.8.4 Convex model of uncertain loading . . . . . . . . . . . . 303
9.8.5 Worst-case estimation . . . . . . . . . . . . . . . . . . . . 304
9.8.6 A numerical example of worst-case estimation of
homology design . . . . . . . . . . . . . . . . . . . . . . . 305
9.8.7 Concluding remarks . . . . . . . . . . . . . . . . . . . . . 306
9.9 Design of Flexible Structures under Constraints on Asymptotic
Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
9.9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 307
9.9.2 Definition of asymptotic stability . . . . . . . . . . . . . 308
9.9.3 Optimization problem . . . . . . . . . . . . . . . . . . . . 309
xxii Optimization and Anti-Optimization of Structures under Uncertainty
9.10 Force identification of prestressed structures . . . . . . . . . . . . 316

9.10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 316
9.10.2 Equations for self-equilibrium state . . . . . . . . . . . . 317
9.10.3 Formulation of identification error . . . . . . . . . . . . 318
9.10.4 Sensitivity analysis with respect to nodal coordinates . . 320
9.10.5 Optimal placement of measurement devices . . . . . . . 322
10. Concluding Remarks 327

10.1 Why Were Practical Engineers Reluctant to Adopt Structural
Optimization? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
10.2 Why Didnt Practical Engineers Totally Embrace Probabilistic
Methods? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
10.3 Why Dont the Probabilistic Methods Find Appreciation among
Theoreticians and Practitioners Alike? . . . . . . . . . . . . . . . 333
10.4 Is the Suggested Methodology a New One? . . . . . . . . . . . . 334
10.5 Finally, Why Did We Write This Book? . . . . . . . . . . . . . . 339
Bibliography 343
Index 387
Author Index 391
Chapter 1
Introduction
May the best yeve ever seen, Be the warst yell ever see. (Scottish
blessing)
It is not enough to do your best; you must know what to do, and then
do your best. (W. Edwards Deming)
Its only the man who can look at the same problem from many dif-
ferent aspects that will make a true leader. (Takao Fujisawa)
In this chapter, we explain why it is preferable to deal with set-theoretical rather

than probabilistic or fuzzy uncertainty. It is demonstrated that probabilistic analy-
sis with unbounded support is prone to high sensitivity of the results to the behavior
of the tail of the distribution. This bad news is balanced by good news: a proba-
bility distribution with bounded support yields the same results as the simple and
transparent anti-optimization technique. We argue that since there are apparently
no physical parameters with unbounded support, one has to resort to probability
densities with bounded ones or directly to the anti-optimization technique. This
provides, in a nutshell, the main justification for this book to deal with bounded
uncertainty and anti-optimization.
1.1 Probabilistic Analysis: Bad News
Apparently, the first study in which probabilistic methods were applied to structures
was conducted by Mayer (1926). Since then, probabilistic methods have developed
beyond the age of adolescence (Cornell, 1981).
Usually, probabilistic reliability studies involve assumptions on the probability
densities, whose knowledge regarding the relevant input quantities is central. Given
this, the probabilistic properties of the output quantities can be determined. As
Wentzel (1980) stresses, probabilistic methods are
frequently regarded as a kind of magic wand which produces in-
1
2 Optimization and Anti-Optimization of Structures under Uncertainty
formation out of a void. This is a fallacy; the theory of probability

only enables information to be transformed, and conclusions on in-
accessible phenomena to be drawn from data on observable ones.
Probabilistic methods must yield a central quantityreliability of the structure,

namely, the probability that the structure will perform its mission satisfactorily.
Modern society rightfully expects extremely high reliabilities and, consequently,
extremely small probabilities of failure.
The deterministic mechanical theories represent a cornerstone of probabilistic
mechanics. First, the phenomenon should be understood qualitatively; next, it
should be understood sufficiently well quantitatively. Then the output quantities
in their intricate dependence on their input counterparts are developed as a set of
equations, algebraic expressions, or numerical codes of varying complexity. At this
stage the parameters to be treated as uncertain variables or functions are identified.
The deterministic relation or numerical code then serves as a transfer function
determining the probabilistic characteristics of the output and the reliability of the
structure.
Our goal being extremely high reliabilities, it is immediately understood that
the deterministic relations or numerical codes must be of supreme accuracy, in
order to avoid a GIGO (garbage ingarbage out) situation. However, since each
deterministic relation is based on simplifying assumptions with a specified degree of
accuracy of their own, legitimate questions arise: Are less accurate tools suitable for
determining the extremely accurate desired probabilities of failure? What happens
to the reliability of structures designed on the basis of increasingly accurate theories?
No less central is the input data. In most cases, accurate information is un-
avoidable. Probabilistic analysts maintain that their theoretical analyses, presently
devoid of experimental inputs, still are of values; and once information becomes
available, it can be incorporated into a ready-made theoretical framework.
But in the meantime these analyses have taken on lives of their own: to make up
for the lacking information, researchers resort to such assumptions as specific pat-
terns of normal distribution, time-wise stationarity of a random process, space-wise
homogeneity of a random field, ergodicity of the process, or the Markov property.
Thus the questions What can go wrong with probabilistic methods? and To
use or not to use them? have become extremely relevant. Their various aspects
have been addressed in monographs by Ben-Haim and Elishakoff (1990) and Ben-
Haim (1996), and in papers, e.g., Elishakoff and Hasofer (1996). The need for a
major revision of our understanding of the role of probabilistic methods is stressed
in the monograph by Ben-Haim and Elishakoff (1990) and in the articles by Kalman
(1994) and Elishakoff (1995b).
In this section, we elucidate difficulties inherent in structural reliability due to
imperfect information, in conjunction with the ever increasing desire to allow only
extremely small probability of failure.
First of all, we address this cardinal question: Why does one need probabilistic
Introduction 3
methods? The reply to this question is best given in terms of the design of a
structure. In a deterministic design process, one requires
y (1.1)
where is an actual stress that is assumed to be positive in tensile state, and y is
an yield stress. Here, uncertainty enters the picture in the sense that we may not
know the loads precisely, or there may be imprecision in measuring the geometric
parameters of the cross-section, or we may have built an imperfect mechanical model
to describe the behavior of the structure. Accordingly, a required safety factor k req
is introduced, and Eq. (1.1) is replaced by
y
(1.2)
kreq
Once the structure is designed, one can introduce the actual safety factor
y
kact = (1.3)
max
where max is the maximum actual stress occurring in the structure. Thus the
design requirement can be formulated as kact kreq , or, in other words, the actual
safety factor should not be less than the required one.
Can one quantify that the actual safety factor of the uncertainty is not hidden,
but is directly introduced into the scene? Let us attempt to answer this question.
Let the random force p, acting on a tension-compression element (bar) with cross-
sectional area a, have a Weibull (1951) distribution with the following probability
distribution function " k #
p p0
FP (p) = 1 exp , k > 0, w > p0 , p p0 (1.4)
w p0
For p < p0 , FP (p) 0. Equation (1.4) is called the 3-parameter Weibull distribu-
tion, which is a generalization of the exponential distribution. It has been developed
originally to describe failure strength of metals. More recently it has been utilized
in connection with fracture, as well as life distribution of mechanical compounds.
For the special case p0 = 0, the distribution is called 2-parameter Weibull distri-
bution; in this case k = 1 yields the exponential distribution, whereas k = 2 is
associated with Rayleigh distribution. The shape parameter k < 1 is typical of
wearing phenomena, whereas k > 1 is typical of aging effects. The 3-parameter
Weibull distribution is useful in describing phenomena for which some minimum
value p0 exists for the random variable P , so that P takes on values greater than
or equal to p0 . In the following, we denote random variables by upper-case, and
lower-case notation is reserved for their possible values.
We are interested in the reliability of the structure, i.e., the probability of the
structure performing its intended mission satisfactorily. Such a performance is
identified with relationship holding Eq. (1.1) true. Thus, the reliability becomes:
R = Prob( y )
= Prob(P/a y )
(1.5)
= Prob(P y a)
= FP (y a)
Thus, we have
" k #
y a p 0
R = 1 exp (1.6)
w p0
How can we define the safety factor in the context of probabilistic design? One
natural way is to relate it to some characteristic load, say the average load. The
latter equals
E(P ) = p0 + (w p0 )(1 + 1/k) (1.7)
where ( ) is the Gamma function. The variance of the load is
V (P ) = (w p0 )[(1 + 2/k) 2 (1 + 1/k)] (1.8)
The central safety factor n is defined as the ratio of the yield stress to its average
counterpart E(P )/a:
y a y a
n= = (1.9)
E(P ) p0 + (w p0 )(1 + 1/k)
Let us design the structure probabilistically. Probabilistic design requires that
the reliability be not less than a codified value r:
Rr (1.10)
Thus, in view of Eq. (1.6), we obtain
" k #
y a p 0
1 exp r (1.11)
w p0
The design value of the cross-sectional area adesign is found from the equality R = r,
and reads
p0 + (w p0 )[ln 1/(1 r)]1/k
adesign = (1.12)
y
Substitution of adesign for a in Eq. (1.9) enables us to write the central safety factor
explicitly in terms of the codified required reliability r:
p0 + (w p0 )[ln 1/(1 r)]1/k
n= (1.13)
p0 + (w p0 )(1 + 1/k)
For the set of parameters w = 3p0 , k = 4, we have
1 + 2[ln 1/(1 r)]1/4
n= (1.14)
1 + 2(1.25)
Therefore, from Eq. (1.14), n = 1.2314 for r = 0.9; n = 1.3971 for r = 0.99;
n = 1.5082 for r = 0.999; n = 1.5942 for r = 0.9999; n = 1.6653 for r = 0.99999;
n = 1.7263 for r = 0.999999, etc. As we see, a probabilistic model allows us to
associate the required reliability directly with the safety factor.
Let us see now how a small error can affect the reliability calculations. Say that
the actual values are w1 , p1 and k1 , while those used in the analysis are w, p0 and
Introduction 5
k. The actual reliability Ract and the actual probability of failure Pf,act are related
as
Pf,act = 1 Ract (1.15)
The actual reliability is given by Eq. (1.6) with the actual values substituted:
" k #
y a p 1 1
Ract = 1 exp (1.16)
w1 p 1
and the actual probability by
" k #
y a p 1 1
Pf,act = exp (1.17)
w1 p 1
If the design of the structure has been performed using the values w, p0 and
k, the appropriate value of the cross-sectional area is given by Eq. (1.12). To
calculate the actual probability of failure corresponding to the design value adesign ,
we substitute the expression adesign into Eq. (1.17) and have
" k #
y adesign p1 1
Pf,act = exp (1.18)
w1 p 1
or
" k1 #
p0 p1 + (w p0 )[ln 1/(1 r)]1/k
Pf,act = exp (1.19)
w1 p 1
Let us consider some particular cases. In the simplest case p1 = p0 , w = w1 ,
but k1 6= k,
" k1 /k #
1
Pf,act = exp ln (1.20)
1r
Since r is the required reliability, 1 r is recognized in Eq. (1.20) as the allowed
probability of failure Pf,all . Thus, Eq. (1.20) can be rewritten as:
" k1 /k #
1
Pf,act = exp ln (1.21)
Pf,all
Let Pf,all = 106 . Then

Pf,act = exp[13.81551056k1/k ] (1.22)
This function of the ratio k1 /k is shown in Fig. 1.1. As is seen, when k1 /k = 1,
Pf,act = Pf,all = 106 (1.23)
is satisfied.
By contrast, when k1 6= k, the actual probability of failure may differ from the
allowed one. Remarkably, there can be a serendipitous situation, namely, when an
error in measurement of k1 may be of a favorable nature: for k1 /k > 1, the actual
0.00002
0.000015
Pf , act
0.00001
0.000005
0
0.9 0.95 1 1.05 1.1
k1 / k
Fig. 1.1 Actual probability of failure as a function of the ratio k1 /k (solid line); for k1 = k it
coincides with the required one (dotted line); for k1 /k > 1 it is less than the required one; for
k1 /k < 1 it may well exceed the allowed value, resulting in a detrimental state.
probability of failure is less than the allowable one. However, when k1 /k > 1, the
effect of a small error in evaluating k may be detrimental: If k1 /k = 0.95, the
actual probability is about five times the allowed one; if k1 /k = 0.93, the actual
probability is approximately ten times as large as the one which was permitted! We
conclude that the probability of failure is too sensitive a parameter to make do with
imprecise input characteristics, and that accurate determination of the probabilistic
characteristics of input must be an integral part of a rigorous probabilistic analysis.
Regarding the effect of a small deviation in the probability density on the relia-
bility estimate, we consider a bar with cross-sectional area a subjected to a load P
which is a random variable with the probability density fP (p), p being a possible
value of the load. The material of the bar is assumed to be perfectly elastic in com-
pression and has a yield stress in tension y . We also assume that the bar cannot
lose its stability, or undergo any other form of failure. To recapitulate, its reliability
is defined as the probability of the stress = P/a not exceeding the yield stress:
R = Prob( y ) (1.24)
Let us consider the situation in which the data is suggestive for the analyst to
assume the probability density of the load in the form of the log-normal variable

1 (ln p b)2
exp , for p > 0
fP (p) = pc 2 2c2 (1.25)

0, otherwise
where b and c characterize the probability density. The mean value E(P ) and the
variance V (P ) are expressed as

1 2
E(P ) = exp b + c
2 (1.26)
2 2
V (P ) = exp(2b + c )[exp(c ) 1]
Introduction 7
The reliability is thus given by

R = Prob(P y a)
= Prob(ln P ln y a)
(1.27)
1 ln y a b
= + erf
2 c
where
Z x
1
erf(x) = exp(t2 /2)dt (1.28)
2 0
since ln P is a normal variable with mean b and variance c2 . Let us assume now
that the actual probability density differs slightly from the true one and reads, for
p > 0:

() 1 (ln p b)2
fP (p) = exp {1 + sin[2(ln p b)]} (1.29)
pc 2 2c2
where is a constant belonging to an interval [1, 1]. For p < 0, the probability
()
density vanishes. It can be shown that fP (p) is indeed a probability density, non-
negative and satisfies the equality
Z
()
fP (p)dp = 1 (1.30)

To prove this property, we have to demonstrate that
Z
fP (p) sin[2(ln p b)]dp = 0 (1.31)

To do this, we make a substitution ln p = t. Thus, the integral to be calculated
reads
Z
1 (t b)2
I= exp sin[2(t b)]dt (1.32)
c 2 2c2
and further substitution t b = u leads to
Z
1 u2
I= exp 2 sin(2u)du (1.33)
c 2 2c
which vanishes since the integrand is an odd function of u. Thus, the function
defined in Eq. (1.29) represents a probability density of some random variable,
()
denoted by P () . Obviously, if = 0, fP is equal to fP as per Eqs. (1.25) and
(1.29). Stoyanov (1987, pp. 8991) demonstrates that for any k = 1, 2, . . . , we have
E(P () )k = E(P k ) (1.34)
()
i.e., the perturbed random variable P has the same moments as those of an un-
perturbed variable P .
Let us calculate now the true reliability associated with f () :
Z y a
()
R = Prob(P y a) = fP (p)dp (1.35)
0
0.02
0.015
Pf , act
0.01
0.005
0
1 0.5 0 0.5 1

Fig. 1.2 Relation between and Pf,act for c = 0.1, Pf,all = 0.01 (r = 0.99).
0.002
0.0015
Pf , act
0.001
0.0005
0
1 0.5 0 0.5 1

or
Z y a
1 (ln p b)2
R= exp {1 + sin[2(ln p a)]}dp (1.36)
0 pc 2 2c2
Introducing again the new variable of integration ln p = t, we reduce reliability to
Z ln y a
1 (t b)2
R= exp {1 + sin[2(t b)]}dt (1.37)
0 c 2 2c2
and with t b = u, we have
Z ln y ab
1 ln y a b 1 u2
R = + erf + exp 2 sin(2u)du (1.38)
2 c c 2 0 c 2 2c
Evaluation of this integral yields
r
1 ln y a b 2 2 c2
R = + erf + e
2 c c 2 2
(1.39)
2c2 i(b + ln[y a]) 2c2 + i(b + ln[y a])
erfi + erfi
2c 2c
Introduction 9
0.0002
0.00015
Pf , act
0.0001
0.00005
0
1 0.5 0 0.5 1

where erfi(z) is an imaginary error function, defined as (Wolfram, 1996, p. 745)
erfi(z) = erf(iz)/i (1.40)

where i = 1. One can visualize that the reliability calculations have been per-
(0)
formed with the probability density of the load set at fP (p).
Let required reliability be r, so that the design is performed using the expression
in Eq. (1.27). Figures 1.21.4, associated with r = 0.99, r = 0.999 and r = 0.9999,
respectively, depict the actual situation in which the control parameter describing
the deviation from reality does not vanish identically. Thus, the actual probability
of failure Pf,act varies with . Naturally, when = 0, Pf,act takes the value 1 r, or
0.01, 0.001 and 0.0001, respectively, in Figs. 1.21.4. The integrand in Eq. (1.36)
is an oscillatory function. Therefore, depending on the value of c, the value of
the integral is either positive or negative; hence, different behaviors are exhibited
in Figs. 1.21.4. The diagrams demonstrate the possibility of safe errors; i.e.,
deviation of the model from the reality may turn out to be on the safe side. Indeed,
if in Fig. 1.2 tends to 1 from above, the actual probability of failure tends to zero.
Likewise, in Fig. 1.3, associated with c = 0.2 and Pf,all = 103 , if tends to unity
from below, the actual reliability too approaches unity from below. Analogously, in
Fig. 1.4, with c = 0.2, Pf,all = 104 , and tending to unity from below, the actual
probability of failure tends to zero from above. It is prudent, however, to consider
what can go wrong when the theoretical model differs from reality. As Fig. 1.2
demonstrates, when tends to unity, the actual probability of failure becomes
double the allowable one. This implies, through the frequency interpretation of the
probability notion, that the number of unsuccessful performances may be double
the allowed. Nearly analogous situations are depicted in Figs. 1.3 and 1.4 when
tends to 1.
For a critical appraisal of methods to determine the possibility of failure, the
reader may like to consult, for example, the paper by Schueller and Stix (1987) and
the monograph by Elishakoff (2004).
1.2 Probabilistic Analysis: Good News
The somewhat discouraging picture painted above by no means justifies boycotting

the probabilistic approach. In fact, the latter makes it possible to substantiate
the mysterious fudge factor, or factor of safety. Taking an example from a mono-
graph (Elishakoff, 2004) on safety factors, consider an element subjected to a stress
modeled as a random variable.
In a similar manner as in previous section, random variables are denoted with
capital letters. Hence the random stress is denoted by with its realization being
denoted by lower-case notation . We also assume that the yield stress y is a
deterministic quantity. Both the actual and yield stresses can be treated as random
variables, but for greater convenience only one therein is so treated. We also adopt
the simplest possible assumption of a continuously distributed , i.e., a uniform
probability density

0, for < L , > U
f () = 1 (1.41)
, for L U
U L
where L and U are the lower and upper bounds of the stress.
Now, the reliability R, i.e., the probability that the actual stress does not exceed
the yield stress y
R = Prob( y ) (1.42)
equals

Z y
0, for y < L

y L
R= f ()d = , for L y U (1.43)
U L

1, for y > U
Let us concentrate on the range L y U . We first note that the mean
value E() of the actual stress equals
1
E() = (L + U ) (1.44)
2
whereas the standard deviation d is given by
U L
d = (1.45)
2 3
By Eqs. (1.44) and (1.45), the lower and upper bounds of the stress are expressed,
respectively, as

L = E() 3d
(1.46)
U = E() + 3d
Thus, the reliability is rewritten, in the range L y U as

y E() + 3d
R= (1.47)
2 3d
Introduction 11
At this juncture it is instructive to introduce the central safety factor as the following
ratio
y
n= (1.48)
E()
along with the coefficient of variation of the stress
d
= (1.49)
E()
Dividing the numerator and denominator of Eq. (1.47) by E(), we obtain

n 1 + 3
R= (1.50)
2 3
This formula relates reliability, safety factor, and the coefficient of variation of the
stress. Expressing the safety factor from Eq. (1.50), we have

1
n = 1 + 2 3 R (1.51)
2
Probabilistic design is based on the requirement
Rr (1.52)
where r is the required, codified reliability, and the required safety factor nr is
obtained from Eq. (1.51) by setting R = r:

1
nr = 1 + 2 3 r (1.53)
2
This is a simple formula connecting the required safety factor nr with the required
reliability r and coefficient of variation of the stress.
We can deduce several useful conclusions:
(i) The safety factor, so often criticized by practitioners and researchers alike, is
actually a powerful concept, which can be given a probabilistic interpretation.
(ii) Probability theory strips the mystery from the safety factor, which instead of
being assigned at the will of the designer, or out of the sky as it were, gains an
analytical framework for its determination.
(iii) If one can quantify the required reliability, say through legislation, one can
quantify the safety factor as well.
(iv) If the required reliability r is greater than 0.5, i.e., if we do not tolerate nearly
half of our products being defective (and, hopefully, we dont!), the required
safety factor exceeds unity. This is in agreement with all textbooks and designs,
where the required safety factor is in excess of unity.
(v) The higher the coefficient of variation of the stress, the higher the required
safety factor. This matches ones anticipation, as it should be. Moreover,
Eq. (1.53) tells us exactly how much the safety factor must be, depending on
the magnitude of uncertainty . Were the coefficient of variation negligibly
small, i.e., = 0+ , the safety factor would be nr = 1+ . This implies that
the truly deterministic design is valid only when tends to zero. This fact
clearly indicates that purely deterministic design is contained as a particular
case in probabilistic design.
Consider now a closely related example from a textbook (Elishakoff 1983,

1999) on the probabilistic theory of structures. Again, our objective is a tension
compression element (bar), in this case a uniform one subjected to random load P
which is uniformly distributed between the lower bound pL and the upper bound
pU . The bar must be designed so that its reliability R equals the required value r
or exceeds it. The design itself amounts to choosing the cross-sectional area a of
the bar. The reliability equals
R = Prob(P/a y ) = Prob(P y a) = FP (y a) (1.54)
where FP (p) is the probability distribution function of the load. Their reliability
formula maintains that the reliability of the bar equals the value of the probability
distribution of the load function computed at the product of the yield stress and
the cross-sectional area. Calculation of the reliability yields

0, for y a < pL
ap
y L
R= , for pL y a pU (1.55)

pU p L
1, for a > p y U
As noted, the range of P is pL y a pU . The design criterion stipulates

y a p L
r (1.56)
pU p L
The minimum required cross-sectional area ar corresponds to
y a r p L
r= (1.57)
pU p L
which leads to the needed cross-sectional area
(pU pL )r + pL
ar = (1.58)
y
We observe that when r tends to unity, the numerator tends to pU and the
design value ar approaches the value
pU
ar = (1.59)
y
This expression is also obtainable from the third alternative in Eq. (1.55) with unity
reliability being obtained for y a pU .
It is remarkable that the expression Eq. (1.58) can also be obtained without
any probabilistic analysis. Indeed, since the load varies in the interval [pL , pU ], the
stress also is an interval variable I = [pL /a, pU /a], which must be less than the
yield stress y . This requirement is satisfied if
pU
y (1.60)
a
which leads to the minimum required value of the area as given in Eq. (1.58). We
arrive at the elegant conclusion: The probabilistic and anti-optimization approaches
Introduction 13
lead to the same result! Not only is there no antagonism between them, but they
both tell us the same thing!
This implies that we can interchangeably utilize either of the two approaches!
Those who prefer the probabilistic sophistification, must be allowed (does anyone
need permission, really?) to continue development and usage of probabilistic me-
chanics. Those who prefer simplicity will stick to the anti-optimization technique.
Naturally, this begs the question: Is one of the approaches preferable to the
other? It appears that the anti-optimization approach, as the simpler of the two,
has a certain advantage. Moreover, as one of the respondents stressed in the poll
conducted by Elishakoff (2000b):
Engineers, especially those at the top, do not trust statements and

predictions of probabilistic character. They must make decisions
on important projects, large structures, and big investments, and
they prefer to be completely sure that their decisions are true. Of
course, proper education in the theory of probability and mathe-
matical statistics helps here, but there is a more profound, maybe
subconscious cause of such an attitude.
Most engineers are more happy to look at samples of the behavior
of a system, at its time signatures, than at probability density func-
tions, and cumulative distribution functions that all seem alike to
practicing engineers. In addition, they become suspicious when a
probabilist talks about probabilities of failure, say 3.14 106 and
all that.
In my own experience, it is expedient to show a set of samples of
the system behavior, in particular the worst sample, the best one
and an average or a typical one...
At the same time, it should be borne in mind that in its pure form, without
proper improvements in the structure via optimization, anti-optimization is likely
to yield ultra-conservative results.
This is why this monograph opts for a combination of anti-optimization and
optimization.
1.3 Convergence of Probability and Anti-Optimization
Another natural question may pop up: if the probability approach is so good as to
provide the same answer as its intellectual competitor, anti-optimization, why was
it necessary to talk, in the previous section, about it being (sometimes) bad?
It is important to have a reply to this question, because the variously named
non-probabilistic approaches are presented as sound alternatives to the probabilistic
description. It appears that this happens because probability theory often uses
random variables which stretch from minus infinity to plus infinity. Such is the case
with the best-known discussion the normal or Gaussian; do the exponential and
Gamma distributed random variables take on values from zero to infinity? Are any
of these random variables, as mental constructs, good for describing geometric or
material characteristics?
Whereas the exponential and Gamma distributions with their only positive val-
ues can be utilized for describing positive quantities, it is doubtful that there exists
any physical parameter that can take on values beyond physical limits. Likewise,
use of the normal distribution to describe positive quantities appears to be doubtful,
if not outright abnormal (pun intended). At best, these distributions are convenient
analytical approximations.
The above implies that engineers ought to use truncated distributions, i.e., dis-
tributions with bounded support. It is gratifying that awareness of this fact grows
amongst stochastic analysts. For example, Grigoriu (2006) mentions: All distri-
butions with bounded support in (0, ) are consistent with available information.
Ma, Leng, Meng and Fang (2004) write: Bounded random parameters ... are more
reasonable for engineering structures than the unbounded Gaussian random ones.
In their study, Minciarelli, Gioffre, Grigoriu and Simiu (2001) write: Unless oth-
erwise indicated, all uncertainty variables . . . will be assumed to have truncated
normal distribution.
Cai and Wu (2004) note:
When investigating a dynamical system under random excitation,

it is important that each existing process should be modeled prop-
erly to resemble its measured or estimated statistical and proba-
bilistic properties. In many cases, Gaussian distribution is assumed
for convenience of analysis. However, the range of Gaussian distri-
bution is unbounded; namely, there exists probability of having
very large values. This violates the very nature of a real physical
quantity, which is always bounded.
Likewise, Cai and Lin (2006) write: Physically realistic random procedures are
bounded, and they may deviate far from being Gaussian. At the end of their paper
these authors are even more forceful: Physically realistic stochastic processes must
be bounded. (see also Cai (2003)).
Ang and Tang (1975) realized this fact quite long time ago:
In engineering, the information may often have to be expressed in

terms of the lower and upper limits of the variable. For example,
when judgment is necessary it is often convenient and perhaps more
realistic to express judgmental information in the form of a range of
possibilities given the range of possible values of a random variable
and the underlying uncertainty may be evaluated by prescribing a
suitable distribution within the range.
Introduction 15
In their study, Simiu and Heckert (1996) analyzed the wind speed model and
fitted an approximate extreme value distribution. They concluded that the reverse
Weibull distribution is more appropriate than those of Gumbel or Frechet. Their
conclusion is supported by the physical fact that non-tornadic extreme winds are
expected to be bounded:
In our opinion, the analysis provided persuasive evidence that ex-

treme wind speeds are described predominantly by reverse Weibull
distributions, which unlike the Gumbel distribution have finite up-
per tail and lead to reasonable estimates of wind load factors.
They also stress
It is a physical fact that extreme winds are bounded. Their prob-

abilistic model should reflect this fact. To the extent that an ex-
treme value distribution would be a reasonable model of extreme
wind behavior, one would intuitively expect that the best fitting
distribution to have finite tail...
Kanda (1994) also showed that extreme winds are best fitted by distributions
with limited tails; see also Walshan (1994). Holmes (2002) notes that extreme winds
have a physical upper limit (which may differ for different storm types); hence,
they have a bounded distribution with data from nearly all stations in Australia
using various fitting methods. He reached a bounded generalized extreme value
(GEV) distribution in over 80% of the cases. He stresses: The approach ... in
substituting one third of the observation by unbounded, randomly generated
normal variables is not convincing.
Thus, the pragmatic engineer cannot agree with Lindley (1987) who states:
Probability is the only satisfactory description of uncertainty. Neither can one
agree with the notion implied by the title of Talebs 2001 book that we are fooled
by randomness. (Readers are also advised to read insightful articles on the different
sides of the issue by Klir (1989, 1994).)
Berleant and Goodman-Strauss (1998) used interval analysis to bound the re-
sults of arithmetic operations on random variables of unknown dependency. Thus,
randomness and boundedness may pragmatically interact with each other. Kim,
Ovseyevich and Reshetnyak (1993) and Elishakoff, Cai and Starnes (1994a) com-
pared probabilistic and anti-optimization approaches. Ferson and Ginzburg (1996)
advocate for application of different methods to describe ignorance and variability.
We opt in this book for the anti-optimization, or worst-scenario approach be-
cause it is simpler than the probabilistic one, not because we are against as Kosko
(1994) calls it the probability monopoly. Moreover, it yields the same results as
a more complicated probabilistic methodology (according to Hans Hoffmann, the
ability to simplify means to eliminate the unnecessary so that the necessary may
speak). We just add a spice to the approach: anti-optimized, worst-scenario re-
sults need to be optimized so as to yield cheaper, less voluminous designs. This is

why this book is fully dedicated to marrying these two concepts: optimization and
anti-optimization.
Chapter 2
Optimization or Making the Best in the

Presence of Certainty/Uncertainty
Make the best of things. (Russian proverb)
Good, better, best. Never let it rest until your good is better and
your better is best! (Anonymous)
We have to do our best and leave the rest to Providence. (Japanese

proverb)
In this chapter, we first address the motivation for optimization of structures in

the presence of certainty or in its absence. We try to provide some thoughts on
both the philosophical and pragmatic question: What can we get from structural
optimization? We also provide an overview of optimization problems, and various,
but as expected, not all methodologies. Then we provide a touch on probabilistic
(or stochastic) optimization, and complete the chapter with optimization via fuzzy
sets. Reviews of these deterministic and non-deterministic techniques provide a
glimpse into what will be the main scope of the book, namely, a specific non-
deterministic technique which uses neither the concept of randomness nor the idea
of fuzziness. This is done not because we feel that these approaches are without
merit: a whole volume can be written on each of them. By discussing stochastic
and fuzzy optimization techniques, we hope that the reader can have at least a
fragmentary picture of topics which will not be central in our subsequent exposition.
2.1 Introduction
In the design process of structures in various fields of engineering, the design vari-
ables such as the cross-sectional geometries, nodal locations, material properties,
etc., are modified by the designer to improve structural performance within the
limited cost for production and construction. Hence, the whole process of struc-
tural design may be regarded as that of seeking the best set of design variables
under given design requirements.
17
The first paper on structural optimization is ascribed to Galileo Galilei who in

1638 discussed the optimal shapes of beams. In this respect it is instructive to give
a podium to Venkayya (1993):
The notion of an optimum solution to an engineering problem is

intriguing and has been investigated for a long time. The strongest
cantilever beam in bending and constant shear as formulated by
Galileo Galilei was also an optimum design for minimum weight
under a uniform stress constraint. Galileos problem was probably
one of the earliest optimization problems.
The various types of minimum principles, such as minimum potential energy

and minimum kinetic energy, were extensively investigated in the 18th century to
be established as calculus of variation. The minimization problem for finding the
solutions governed by a minimum principle is called variational problem. The con-
strained variational problem called isoperimetric problem was solved by Leonhard
Euler (17071783) to show that a circle maximizes the interior area for a speci-
fied length of a closed smooth curve on a plane. The shape of a hanging cable
called a catenary that minimizes the potential energy was found by John Bernoulli
(16671748). Variational methods are applied in structural mechanics to find the
equilibrium states of an elastic conservative system based on the minimum principle
of total potential energy.
The origin of modern optimization based on computational methods is generally
credited to the development of linear programming during the period of World War
II. Since then, many problem formulations and algorithms have been developed to
establish the field called mathematical programming (Luenberger, 2003) in applied
mathematics, and operations research in management science.
Structural optimization is regarded as an application of optimization strategies
to the field of structural design (Arora, 2004; Haftka, G urdal and Kamat, 1992;
Vanderplaats, 1999). The paper by Prager and Taylor (1968) may be cited as the
first contribution on modern structural optimization, where the optimality criteria
are derived for several problems under constraints on linear buckling load, eigen-
value of vibration, external work against static loads called compliance, and plastic
collapse load.
For topology optimization problems, where the locations of members of framed
structures are optimized (see Sec. 2.7 for details), the so-called Michell truss
(Michell, 1904; Hemp, 1973) is considered to be the first contribution and is ex-
tensively cited for verification of recently developed numerical methods. However,
the paper by Dorn, Gomory and Greenberg (1964), which initiated the widely used
method called ground structure approach, may be cited as the first work of mod-
ern computer-based topology optimization. In the 1970s, the analytical approaches
based on optimality conditions were investigated (Rozvany, 1976). Since the 1980s,
with rapid development of computer technology, many research papers have been
Optimization or Making the Best in the Presence of Certainty/Uncertainty 19
published on computational methods and their industrial applications. Now, in

the 21st century, optimization methods have matured, and we can optimize real-
world structures with moderately large numbers of degrees of freedom and design
variables.
2.2 What Can We Get from Structural Optimization?
The basic strategy of structural design without resort to optimization algorithm

may simply be stated as follows:
Find the best design iteratively modifying the design variables by trial-and-error
process based on the experience of the designer, under requirements on the struc-
tural responses and mechanical properties, e.g., the stresses and displacements
against static/dynamic design loads, eigenvalues of free vibration, liner/nonlinear
buckling loads, etc.
However, the following questions may arise:
How can we define the best design?

What is the best strategy to modify the design variables if the current design is
not the best?
Optimization can answer these questions.

In the traditional design process, arbitrary initial (trial) values are given for the
design variables, structural analysis is carried out for evaluating the responses, and
the variables are modified intuitively if the design requirements are not satisfied;
i.e., there is no general procedure or guideline for design modification. Furthermore,
the design process is often terminated if all the requirements are satisfied, and no
effort is made to find better solutions.
On the other hand, in view of code-based design, where the upper and lower
bounds of the responses as well as the bounds of design variables are given by the
design code, structural optimization provides us with the following benefits:
A design satisfying all the constraints (feasible design) can be found automatically,
and efficiently, while simultaneously minimizing the objective function, such as
the total structural volume, if the problem is appropriately formulated so that
the set of feasible designs is nonempty.
The optimization tool helps the decision making of the designer, i.e., it is not an
automatic design tool that gives a negative impression to the structural engineers
and designers. If the optimal design is not acceptable to the designer, the upper-
level parameters, such as the geometry and material properties of the structure,
can be modified, or additional constraints can be assigned for the optimization
problem, to obtain a more reasonable optimal solution. Therefore, the designers
can spend more time on the synthetic jobs, rather than structural analysis, if
Design parameter
Design load
(geometry, topology)
Structural analysis Structural optimization
Response
(stress, displacement) Design variable
A step of conventional A step of optimization-based

design process design process
Fig. 2.1 Illustration of conventional and optimization-based design processes.
optimization tools are effectively used for decision making. The conventional
design process and optimization-based design process are illustrated in Fig. 2.1.
Optimization is very helpful for designing complex structures, for which even an
experienced designer cannot easily find a feasible design.
Even if the optimal solution cannot be used directly in design practice, the optimal
design gives insight into a better design.
If we start with a feasible and nearly optimal design found by an expert, the
design cannot be worse after optimization, and usually a better design can be
found.
Furthermore, in view of decision making,
The trade-off relation between the structural cost and responses can be made
clear, if optimization is carried out several times by modifying the input param-
eters such as the cost coefficients and upper bounds of responses.
The most positive ways of using optimization may be
To find new structural systems and shapes by optimization (Bendse and Sig-
mund, 2003).
To generate innovative structure or material that cannot be found without op-
timization, e.g., compliant mechanisms. Figure 2.2 shows an example of a com-
pliant bar-joint structure utilizing snapthrough behavior (Ohsaki and Nishiwaki,
2005).
2.3 Definition of the Structural Optimization Problem
The structural optimization problem can be formally formulated as

Minimize objective function (cost, weight, volume) subject to constraints on me-
chanical performances.
Input node Input load

(forced displacement)
Output node
Output
displacement
Fig. 2.2 Illustration of a compliant bar-joint system.
where subject to means under constraints on, and mechanical performances include
member stresses, nodal displacements, etc., for which the upper and lower bounds
are given.
The bounds for the mechanical performances are determined from the require-
ments given by the codes (regulations) in practical design process. The total struc-
tural volume (or weight) is usually assigned as the objective function, because it is a
critical requirement to reduce the weight of the aerospace and mechanical structures.
For the long-span structures in architectural and civil engineering, the reduction of
the self-weight generally leads to reduction of the design loads, and hence leads to
lower cost. It is also important that, for any acceptable definition of the objective
function, a solution satisfying all the constraints, called feasible design, is obtained
after optimization as discussed in Sec. 2.2.
If the concept of minimum weight is not under consideration, the following
alternative formulation may be valid:
Maximize mechanical performance subject to upper bound for cost (weight, vol-
ume).
Note that the two formulations stated above lead to similar optimal solutions if the
problem parameters are appropriately assigned.
Consider, for instance, the problem of minimizing the structural volume of a
truss under constraints on stresses and displacements against static loads. The de-
sign variables are the cross-sectional areas of the members. Let = (1 , . . . , m )>
and U = (U1 , . . . , Un )> denote the vectors of member stresses and nodal displace-
ments that are called state variables, respectively, where m is the number of mem-
bers and n is the number of degrees of freedom. All vectors are column vectors, and
a subscript is used for indicating a component of a vector throughout this book.
The vector of member cross-sectional areas are denoted by A = (A1 , . . . , Am )> ,
and let Li denote the length of the ith member.
The structural optimization problem under stress and displacement constraints

is formulated as
m
X
minimize A i Li (2.1a)
i=1
subject to iL i (A) iU , (i = 1, . . . , m) (2.1b)

UiL Ui (A) UiU , (i = 1, . . . , n) (2.1c)
ALi Ai AU i , (i = 1, . . . , m) (2.1d)
where the upper and lower bounds are denoted by the superscripts ( )U and ( )L ,
respectively.
In Problem (2.1), the stress of the ith member is written as a function of A only
as i (A). However, for a general statically indeterminate structure, the responses
are defined implicitly through the vector of nodal displacements U (A), which is
also a function of A. The responses may also depend explicitly on A, e.g., the axial
force Ni of member i is defined as i (U )Ai .
Next, we formulate the problem in a more general form. Let x = (x1 , . . . , xm )>
denote the vector of design variables representing the cross-sectional areas, nodal
coordinates, etc. Suppose the objective function such as the total structural volume
is defined as an explicit function of x as F (x). Then, the structural optimization
problem is formulated as follows:
minimize F (x) (2.2a)

I
subject to Gi (x, U (x)) 0, (i = 1, . . . , N ) (2.2b)
E
Hi (x, U (x)) = 0, (i = 1, . . . , N ) (2.2c)
where Gi and Hi are generally nonlinear functions of x and U (x), and N E and N I
are the numbers of equality and inequality constraints, respectively. If the design
variables x can take continuous values, then the structural optimization problem
(2.2) is classified as a nonlinear programming problem, for which various algorithms
such as sequential quadratic programming and method of feasible directions are
available. On the other hand, if the variables take integer values or are selected
from a list or a catalog of available values, then the problem is classified as a mixed
integer nonlinear programming problem with continuous state variables U (x) (see
Sec. 2.4.2 for a detailed classification of optimization problems).
The standard approach to nonlinear programming problem is based on the gra-
dient information of the objective and constraint functions. Suppose the vector U
of state variables is defined in a general form by the state equations such as the
equilibrium (stiffness) equations as
Zi (x, U (x)) = 0, (i = 1, . . . , n) (2.3)
Then the differential coefficients, which is called sensitivity coefficients, Uk /xi

of Uk with respect to xi can be computed from a set of n simultaneous linear
equations (Choi and Kim, 2004)

n
Zj X Zj Uk
+ = 0, (j = 1, . . . , n) (2.4)
xi Uk xi
k=1
The details of sensitivity analysis can be found in Sec. 3.6.

In order to apply nonlinear programming algorithms, the constraint function
Hi (x, U (x)) is regarded as a function of x only:
e i (x) = Hi (x, U (x))
H (2.5)
e i (x) are obtained by using the sensitivity
Then the differential coefficients of H
coefficients of U (x):
n
ei
H Hi X Hi Uk
= + , (j = 1, . . . , m) (2.6)
xj xj Uk xj
k=1
e i (x) = Gi (x, U (x)) are obtained similarly.

The differential coefficients of G
2.4 Various Formulations of Optimization Problems
2.4.1 Overview of optimization problems

If the objective and constraint functions are smooth and continuously differentiable
with respect to the design variables, the optimal solutions can be found by the
so-called gradient-based mathematical programming approach, in which analysis and
sensitivity analysis are consecutively performed and the design variables are modi-
fied in accordance with the optimization algorithm. However, this approach cannot
be used if the variables have integer values.
The global optimal solution for such problems can be found by methods of integer
programming such as the branch-and-bound method (Horst and Tuy, 1990), which
requires computational time of exponential function of the problem size. Heuristic
approaches have been developed to obtain approximate optimal solutions within
reasonable computational cost (Reeves, 1995) (see Sec. 2.6 for the details).
In this section, we classify optimization problems based on the continuity of
the functions and variables. Then the methodologies that are closely related to
anti-optimization, and optimization with anti-optimization, such as approximation
methods, multilevel optimization and multiobjective programming, are described.
Text books, e.g., Luenberger (2003), may be consulted for the well-developed meth-
ods for linear and nonlinear programming problems.
We first summarize the terminologies used in the problem formulation and algo-
rithms of optimization. Note that a minimization problem is assumed for simplicity.
Feasible region: A solution that satisfies all the constraints is called a feasible
solution, and the region of feasible solutions in the space of variables is called a
feasible region.
Local optimal solution: The solution that has the smallest objective value
among those in its neighborhood feasible solutions is called a local optimal solu-
tion.
Global optimal solution: The solution that has the smallest objective value
among those of all the feasible solutions is called a global optimal solution.
Active/inactive constraints: An inequality constraint that is satisfied with
equality is said to be active. The constraint is said to be inactive if it is sat-
isfied with strict inequality.
Convex set: Let x1 and x2 denote two solutions belonging to a set I. If x1 +
(1 )x2 is contained in I for any 0 1, then the set I is said to be a
convex set.
Convex function: Let x1 and x2 denote two solutions belonging to a convex set
I. If F (x1 ) + (1 )F (x2 ) F (x1 + (1 )x2 ) for any 0 1, then F (x)
is said to be a convex function.
Convex programming problem: If the objective function is convex, and the
feasible region is a convex set, then the optimization problem is called a convex
programming problem, for which the solution satisfying the conditions of local
optimality is also a global optimal solution.
Positive semidefinite matrix: An n n symmetric matrix S is positive semidef-
inite, if all the eigenvalues of S are non-negative, which is equivalent to the non-
negativeness of the bilinear form b> Sb for any n-vector b. Furthermore, S is
positive definite if b> Sb is positive for any b 6= 0.
2.4.2 Classification of optimization problems

The lower and upper bounds for the variable xi are denoted, respectively, by xLi
and xU
i . The optimization problem for minimizing the objective function F (x) is
generally formulated as

E
subject to Gi (x) = 0, (i = 1, . . . , N ) (2.7b)
I
Hi (x) 0, (i = 1, . . . , N ) (2.7c)
xLi xi xU
i , (i = 1, . . . , m) (2.7d)
The constraints (2.7d) are called side constraints, bound constraints, or box con-
straints, which are treated independently of the general inequality constraints (2.7c)
in most of the optimization algorithms.
If the variables take only integer values, the problem is called an integer pro-
gramming (IP) problem. Since the optimal combination of the variables is found
by IP, it is equivalently called combinatorial optimization problem. Among various
formulations of IP, the problem with 01 variables only is called a 01 programming
problem. A problem that has real and integer variables is called a mixed integer
programming (MIP) problem.
Consider next the case where all the variables can take real values. If F (x),
Gi (x) and Hi (x) are all linear functions of x, the problem is called a linear pro-
gramming (LP) problem, for which the global optimal solutions can be found by
well-established methods such as the simplex method and interior point method. If
at least one of F (x), Hi (x) and Gi (x) is a nonlinear function of x, the problem is
called a nonlinear programming (NLP) problem.
For LP and NLP, the Lagrangian for Problem (2.7) is given as
E I
N
X N
X
L(x, , ) = F (x) + j Gj (x) + j Hj (x) (2.8)
j=1 j=1
where = (1 , . . . , N E )> and = (1 , . . . , N I )> are the vectors of Lagrange mul-

tipliers corresponding to the equality and inequality constraints, respectively. Note
that the side constraints (2.7d) are included in the general inequality constraints
(2.7c) for simplicity.
Suppose the feasible region of Problem (2.7) is nonempty and has a sufficiently
smooth boundary. Then the necessary conditions to be satisfied at an optimal so-
lution, called the optimality conditions or KarushKuhnTucker (KKT) conditions,
are written as
NE
X Gj X Hj NI
F
+ j + j = 0, (i = 1, . . . , m) (2.9a)
xi j=1 xi j=1
xi
j 0, j Hj = 0, (j = 1, . . . , N I ) (2.9b)
with the constraints (2.7b) and (2.7c), where the conditions (2.9b) are called a
complementarity conditions.
Consider a simple problem called quadratic programming (QP) problem, where
the objective function is a quadratic function of x and we have only linear equality
constraints as
minimize x> Dx + c> x (2.10a)
>
subject to C x b = 0 (2.10b)
where b = (b1 , . . . , bn )> and c = (c1 , . . . , cm )> are constant vectors, C is an n m
constant matrix, D is an m m constant matrix, and we assume rank C < m.
If D is positive definite, then Problem (2.10) is called convex QP. From the
KKT conditions, the optimal solutions are explicitly written as
1
x = D 1 (C > + c),
2 (2.11)
= (CD1 C > )1 (2b + CD1 c)
In this book, we often formulate an anti-optimization problem of maximizing
(minimizing) a linear objective function under a quadratic constraint on x as
maximize c> x (2.12a)
>
subject to x Dx b (2.12b)
where b is a positive constant, and D is a positive definite weight matrix. The anti-
optimal solutions for Problem (2.12) are explicitly derived from the KKT conditions
as

1 1 CD 1 C
x= D c, = = (2.13)
2 2 b
where is the Lagrange multiplier for the constraint (2.12b), and = corre-
sponds to the optimal solution that minimizes c> x.
The optimization problem containing constraints such that the variable matrix is
positive semidefinite is called a semidefinite programming (SDP) problem (Wolkow-
icz and Vandenberghe, 2000; Ohsaki and Kanno, 2007). The objective function and
other constraints are linear in the standard form of SDP problem. Since the con-
straints of positive semidefiniteness of matrices include linear and convex quadratic
constraints, the SDP is an extension of LP and convex QP.
Let X O indicate that the symmetric matrix X is positive semidefinite. The
inner product X Y of the n n matrices X = (Xij ) and Y = (Yij ) is defined as
n X
X n
X Y = Xij Yij (2.14)
i=1 j=1
The standard SDP problem is formulated as

minimize C X (2.15a)
subject to Ai X = bi , (i = 1, . . . , m) (2.15b)
XO (2.15c)
where C and Ai are n n constant matrices, bi (i = 1, . . . , m) are scalars, and X
is the variable matrix.
Various structural optimization problems including the truss topology optimiza-
tion problem under frequency constraint (Ohsaki, Fujisawa, Katoh and Kanno,
1999), linear buckling constraint (Kanno, Ohsaki and Katoh, 2001), and robust
optimization problem (Ben-Tal and Nemirovski, 1997) can be formulated as SDP
problems.
2.4.3 Parametric programming

An optimization problem that has parameters in addition to the variables is called a
parametric programming problem (Ant onio, 2002; Gal, 1979). For a two-level opti-
mization problem formulated for structural optimization considering the worst-case
scenario, the lower-level anti-optimization problem is conceived as a parametric pro-
gramming problem, for which the design variables of the upper-level optimum design
problem are regarded as parameters (see Sec. 9.3 for the problem formulation).
Parametric programming approaches have been widely applied to structural op-
timization. For example, for a simple cross-sectional optimization of trusses, the
geometrical variables such as nodal coordinates are considered to be parameters.
Nakamura and Ohsaki (1988) presented a parametric programming approach for

generating a set of optimal trusses under multiple frequency constraints. Ohsaki
and Arora (1993) presented a general parametric programming approach to struc-
tural optimization.
In the field of structural optimization, the sensitivity of the optimal design
with respect to the problem parameter is called the optimum design sensitivity,
which has been extensively studied for application to two-level structural optimiza-
tion problems (Sobieszczanski-Sobieski, James and Dovi, 1985; Vanderplaats and
Yoshida, 1985; Barthelemy and Sobieszczanski-Sobieski, 1983; Bloebaum, Hajela
and Sobieszczanski-Sobieski, 1992).
Let p = (p1 , . . . , pN P )> denote a parameter vector, where N P is the number of
parameters. A parametric optimization problem is stated as
minimize F (x, p) (2.16a)
I
subject to Gi (x, p) 0, (i = 1, , N ) (2.16b)
E
Hi (x, p) = 0, (i = 1, , N ) (2.16c)
where the side constraints are included in the general inequality constraints (2.16b).
Let R(x, p) denote the vector of constraint functions consisting of the equality
constraints (2.16c) and the active inequality constraints in Eq. (2.16b) at the optimal
solution. The ith component of R is denoted by Ri . Then the KKT conditions for
Problem (2.16) are written as
NA
X Rj
F
+ j = 0, (i = 1, . . . , m) (2.17)
xi j=1 xi
where N A is the number of constraints satisfied in equality, and j is the Lagrange

multiplier for the constraint Rj = 0. Note that j 0 should be satisfied if Rj is
related to the active inequality constraint, whereas no restriction in sign exists for
j related to an equality constraint.
The design variables and the objective value at the optimal solution is conceived
as functions of p, because they can be obtained for each specified value of p. A
function of p only is denoted by a hat (b ). The derivatives of the optimal objective
value Fb(p) = F (b
x(p), p) with respect to pk is computed from
m
X F b
Fb F xi
= + , (k = 1, . . . , N P ) (2.18)
pk pk i=1 xi pk
Suppose that the active constraints in

bj (p) = 0, (j = 1, . . . , N A )
R (2.19)
at an optimal solution remain active for a small variation of pk , i.e.,
bj
R
= 0, (j = 1, . . . , N A ; k = 1, . . . , N P ) (2.20)
pk
Then the following relation is obtained from Eqs. (2.19) and (2.20):
m
Rj X Rj b xi
+ = 0, (j = 1, . . . , N A ; k = 1, . . . , N P ) (2.21)
pk i=1
x i p k
By multiplying b
xi /pk to the both sides of Eq. (2.17) and taking summation
over i, we obtain
A
m N m
X F b xi X X Rj b xi
+ j = 0, (k = 1, . . . , N P ) (2.22)
i=1
x i pk j=1 i=1
x i p k
The following relation is derived by multiplying j to Eq. (2.21) and taking sum-
mation over j corresponding to the equality constraints and the active inequality
constraints:
A
N
X NA m
Rj X X Rj b xi
j + j = 0, (k = 1, . . . , N P ) (2.23)
j=1
p k j=1 i=1
x i p k
From Eqs. (2.18), (2.22) and (2.23),

X Rj NA
Fb F
= + j , (k = 1, . . . , N P ) (2.24)
pk pk j=1 pk
is derived. We can see from Eq. (2.24) that the derivative of the optimal objective
value with respect to pk can be obtained without computing the derivatives of the
variables and the Lagrange multipliers with respect to the problem parameter.
Differentiation of Eq. (2.17) with respect to pk leads to

Xm NA
X 2 2 NA
X
2
F R j b
x l F 2 Rj
+ j + + j
xi xl j=1 xi xl pk xi pk j=1 xi pk
l=1
(2.25)
XNA bj
Rj
+ = 0, (i = 1, . . . , m; k = 1, . . . , N P )
j=1
x i p k
The derivatives of b1 , . . . ,
bN P and x
b1 , . . . , x
bm with respect to pk are computed
A
from a set of m + N linear equations (2.21) and (2.25), where the terms in the
parentheses of the first term on the left-hand-side of Eq. (2.25) consist of the Hessian
of the Lagrangian.
2.4.4 Multiobjective programming

So far, we have assumed that a single objective function is minimized or maximized
through optimization. However, in the practical situation of designing structures,
it is very natural to assume that we have more than one performance measure to
consider. Thus, the optimization problem turns out to be a multiobjective pro-
gramming problem that has more than one objective functions (Cohon, 1978).
F2
Pareto optimal set
F1
Fig. 2.3 Pareto optimal set in objective function space for N F = 2; gray region: feasible region.
Let F1 (x), . . . , FN F (x) denote the objective functions to be minimized. The

multiobjective programming problem is formulated as
minimize F1 (x), . . . , FN F (x) (2.26a)

I
subject to Gi (x) 0, (i = 1, , N ) (2.26b)
E
Hi (x) = 0, (i = 1, , N ) (2.26c)
where the side constraints are included in the inequality constraints (2.26b) for
simplicity.
Contrary to the single-objective problem, generally there is no solution that si-
multaneously minimizes all objective functions. Consider two feasible solutions x(1)
and x(2) satisfying all the constraints (2.26b) and (2.26c). Dominance of solution
is defined as
If Fi (x(1) ) Fi (x(2) ) for i = 1, . . . , N F and Fj (x(1) ) < Fj (x(2) ) for one of

j {1, . . . , N F }, then x(2) is said to be dominated by x(1) .
The Pareto optimal solution is defined as
If there is no feasible solution that dominates a feasible solution x , then x

is called a nondominated solution, noninferior solution, compromise solution or
Pareto optimal solution.
The set of Pareto optimal solutions is called a Pareto optimal set, or Pareto set for
brevity. The thick line in Fig. 2.3 illustrates the set of Pareto optimal solutions in
the objective function space for the case of N F = 2, where the gray region is the
set of feasible solutions. For example, the solution A is a Pareto optimal solution,
because there is no solution in the region surrounded by the two dotted lines, where
both of the two objective functions are improved.
The approaches to selecting the most preferred solution from the Pareto optimal
set are classified as a priori articulation of preference and a posteriori articulation of
preference (Marler and Arora, 2004). In the a priori approach, the objective func-
tions are combined or transformed to constraints to formulate a single-objective
problem. The most popular and convenient approach for this purpose is the lin-
ear weighted sum approach, where the objective functions F1 (x), . . . , FN F (x) are
combined as
F
N
X
F (x) = wi Fi (x) (2.27)
i=1
with positive weight coefficients w1 , . . . , wN F . This approach is suitable for the case
where the Pareto optimal solutions form a smooth convex curve or hyper-surface
in the objective function space, and the preference of the designers or the decision
makers are rather fixed. Note that the single-objective problem for minimizing F (x)
should be solved iteratively modifying the weight coefficients wi , if the solution is
not acceptable to the decision maker.
In the a posteriori approach, on the other hand, a set of Pareto optimal solutions
is first generated, and the most preferred solution is selected by the decision maker
from the set based on additional preference functions or trade-off relations among
the objective functions. Even for a problem with continuous variables and smooth
differentiable functions, an NLP approach is not suitable for the purpose of gener-
ating many Pareto optimal solutions with enough diversity, because it generates a
single optimal solution for the given set of problem parameters. The genetic algo-
rithm (GA) is one of the most suitable method for this purpose (Goldberg, 1989;
Ohsaki, 1995), because it is classified as a population-based approach that has many
solutions at each step of iteration called generation. On the other hand, the sim-
ulated annealing (SA) and tabu search (TS) are categorized as single-point-search
heuristics that have only one solution at each step of optimization (see Sec. 2.6 for
details of SA for single-objective problems). Recently, SA has been extended to mul-
tiobjective programming problems (Whidborne, Gu and Postlethwaite, 1997). TS
has also been shown to be very effective with multiobjective problems (Baykasoglu,
Owen and Gindy, 1999; Ohsaki, Kinoshita and Pan, 2007).
2.5 Approximation by Metamodels
One of the main difficulties in structural optimization for large-scale, real-world

structures is that substantial computational cost is required for function evalua-
tion (structural response analysis). Also for anti-optimization, structural analysis
should be carried out many times to find the worst-case scenario. Therefore, sev-
eral approximation methods such as response surface approximation (RSA) (Myers
and Montgomery, 1995; Venter, Haftka and Starnes, 1998), kriging method (Lee
and Jung, 2007; Sakata, Ashida and Zako, 2003), radial basis function (Park and
Sandberg, 1991) and artificial neural networks (Rosenblatt, 1962) have been used
for optimization of complex structures. Among them, RSA, also called the response
surface method (RSM), has been widely applied to structural optimization problems
due to its simplicity of implementation (Roux, Stander and Haftka, 1998).
Let F (x) denote the response quantity to be approximated by a quadratic func-
tion of the variables x = (x1 , . . . , xm )> as
m
X m X
X m
F (x) ' F a (x) = c0 + c i xi + cij xi xj + (2.28)
i=1 i=1 j=i
where c0 , ci , cij , . . . are the coefficients. Suppose the values of F (x(k) ) at the set
of data points x(k) (k = 1, . . . , N D ) are known. The coefficients are determined by
minimizing the square error e given as
D
N
X
e= (F a (x(k) ) F (x(k) ))2 (2.29)
k=1
The differential coefficients of e with respect to cij vanishes as

ND
X Xm Xm X m
e c0 + (k) (k) (k) (k) (k)
=2 c i xi + cij xi xj F (x(k) ) xi xj = 0 (2.30)
cij i=1 i=1 j=i
k=1
Similar equations are obtained for differentiation with respect to ci . Thus, the
coefficients are found by solving a set of linear equations based on the well-known
method of least squares.
Suppose the variables are normalized to the range 1 xi 1 (i = 1, . . . , m).
The sampling values for xi may be given by the two-level model (1, 1), three-level
model (1, 0, 1), etc. In the three-level model, however, the number of data points is
3m if the set of all possible combinations is to be considered. The number increases
exponentially as m is increased. Therefore, the method called design by experiment
(Myers and Montgomery, 1995) is effectively used to select the data points of a
moderately small number, while a priori estimation of the approximation error is
minimized to obtain the set of data points.
2.6 Heuristics
2.6.1 Overview of heuristics

Optimal designs of moderately large structures can be obtained easily by a nonlinear
programming approach, if the objective function and the constraints are continuous
functions of the design variables such as cross-sectional properties of a structure.
In the various fields of engineering, however, especially in structural design in civil
engineering, the design variables are often selected from the lists or catalogs of
the standard sections. Also for an anti-optimization problem, the vertices of the
feasible region of the parameters are often searched to find the worst-case scenario.
Therefore, in these cases, the anti-optimization problem is formulated as a combina-
torial problem. Furthermore, we have state variables such as nodal displacements,
which are generally nonlinear functions of the design variables. Thus, the problem
turns out to be a mixed-integer nonlinear programming (MINLP) problem (Floudas,
1995).
It is very easy to solve a combinatorial optimization problem by using the inte-
ger programming approaches, e.g., branch-and-bound method and branch-and-cut
method (Horst and Tuy, 1990), if the number of variables is small and the sub-
problem formulated by relaxing some integer variables to real variables is convex.
However, the computational cost increases as an exponential function of the prob-
lem size, and it is not possible to solve a practical problem within a practically
admissible computational time. Furthermore, for the structural optimization prob-
lem, the relaxed nonlinear programming problems are usually non-convex problems,
for which the global optimality is not guaranteed.
Recently, with the rapid development of computer hardware and software tech-
nologies, we can carry out structural analysis many times to obtain optimal solu-
tions. Another important point is that the global optimality need not be strictly
satisfied in the practical design process. Heuristic approaches (or heuristics for sim-
plicity) have been developed to obtain approximate optimal solutions within rea-
sonable computational time, although there is no theoretical proof of convergence
(Reeves, 1995). The most popular approach is the genetic algorithm (GA) (Gold-
berg, 1989), which can be categorized as a multipoint search or population-based
method that has many solutions at each iterative step called generation. Since com-
putational cost for evaluating the objective and/or constraint functions at each step
can be very large for structural optimization problems, a multipoint strategy may
not be appropriate especially for optimization/anti-optimization of large structures.
Therefore, single-point search heuristics such as simulated annealing (SA) (Aarts
and Korst, 1989) and tabu search (TS) (Glover, 1989) can be effectively used. In
the following, we present basic algorithms of single-point search heuristics.
2.6.2 Basic approaches of single-point search heuristics

Single-point search heuristics are based on local search (Aarts and Lenstra, 1997),
where the solution is consecutively updated to a neighborhood solution if it improves
the value of the objective function. The neighborhood solutions are generated
by slightly modifying the value of one or several variables. The constraints are
incorporated by penalty functions, or the solutions violating the constraints are
simply rejected. Since it is not always possible to find an optimal solution by
simple local searches, various heuristic approaches have been proposed to improve
the convergence property to a good approximate optimal solution, e.g., update to
a non-improving solution is allowed in some heuristics including SA.
2.6.2.1 Neighborhood solutions

The neighborhood solutions should be defined before introducing the strategies of
heuristics. Let J = (J1 , . . . , Jm )> denote the vector of integer variables, and denote
by J (k) the current solution at the kth iterative step of optimization or local search.
The neighborhood of J (k) is defined as the set of solutions for which the distance
from J (k) is sufficiently small.
The distance of two solutions J A and J B may be defined, e.g., as
m
X
D1 = |JiA JiB |
i=1
q (2.31)
D2 = (J A J B )> (J A J B )
D = max |JiA JiB |
i
where D1 is the Manhattan norm (L1 -norm), D2 is the Euclidean norm (L2 -norm),
and D is the Chebychev norm (L -norm). For an integer variable defined by a
binary bit string, the humming distance is defined by the number of different bits,
e.g., the humming distance between the binary numbers 1010110 and 1001101 is 4.
However, the definition of a neighborhood for heuristics is less strict than those
based on distances. For example, the following operations can be used for generating
a neighborhood solution:
Increase or decrease the value of one variable Ji by a specified amount.

Modify the values of a set of variables by a predefined rule.
Exchange the values of a pair of variables (Ji , Jj ).
2.6.2.2 Basic algorithm of single-point search heuristics

Let F (J ) denote the objective function to be minimized. The values corresponding
to the kth iteration is indicated by the superscript ( )(k) . We only consider the
inequality constraints Hj (J ) 0 (j = 1, . . . , N I ) for simplicity. The basic algorithm
of single-point search heuristics can be summarized as
Step 1 Assign the initial solution J (0) , and set the iteration counter k = 0.
Step 2 Randomly generate a neighborhood solution J of the current solution J (k) ,
or select the best solution J from randomly generated neighborhood solutions
based on the criterion defined by the values of the objective function F (J ) and
the constraint function Hj (J ).
Step 3 Update the solution as J (k+1) = J or reject J as J (k+1) = J (k) in
accordance with the given strategy.
Step 4 Output the best solution satisfying the constraints and terminate the pro-
cess, if the solution converged, or the stopping criteria are satisfied; otherwise,
let k k + 1 and go to Step 2.
We can use either specified or randomly generated values as the initial solution.
An optimal solution obtained by another optimization approach can be used, with
possible modification, as the initial solution; e.g., the nearest solution with integer
variables from the optimal solution with real variables can be used.
For a constrained optimization problem, the constraints can be incorporated to
the objective function by using the penalty function approach. For example, for
a problem with inequality constraints Hj (J ) 0 (j = 1, . . . , N I ), the following
penalty function can be used to penalize the solution with violated constraints:
I
N
X
Fe (J ) = F (J ) + j [max(Hj (J ), 0)]2 (2.32)
j=1
where j is a sufficiently large penalty coefficient. It is seen from Eq. (2.32) that
no penalty is given if the constraint is satisfied as Hj (J ) 0, and the penalty
proportional to the square of Hj (J ) is given if the constraint is violated. The
penalty function of this type is called the exterior penalty function.
On the other hand, the interior penalty function can be defined by the logarith-
mic barrier function as
NI
X
e
F (J ) = F (J ) j log(Hj (J )) (2.33)
j=1
where the penalty coefficient j should be sufficiently small. However, the interior
penalty function approach cannot be used for the case where a solution can be
infeasible during the search process, because log(Hj (J )) is not defined for Hj (J ) >
0. Therefore, only the feasible solutions should be searched, which is very difficult
for a heuristic approach to optimization of large-scale complex structures.
2.6.2.3 Greedy method

Heuristics can be classified into deterministic and probabilistic approaches. The
simplest deterministic approach is the greedy method. The procedure of the greedy
method for a simple structural optimization problem such as minimum weight design
under stress constraints is described as follows:
Step 1 Assign the initial solution J (0) that does not satisfy all the constraints
Hj (J ) 0 (j = 1, . . . , N I ), e.g., choose the smallest value for all variables for the
case where F (J ) is an increasing function, and H1 (J ), . . . , HN I (J ) are decreasing
functions. Set the iteration counter k = 0.
Step 2 Select a neighborhood solution J that most efficiently improves the per-
formance of the solution, and update the solution as J (k+1) = J . The ratio of
decrease of the maximum violated value of H1 (J ), . . . , HN I (J ) to the increase of
F (J ) can be used for the definition of the efficiency of improvement.
Step 3 Let k k + 1 and go to Step 2 if at least one of the constraints is violated;
otherwise, output the best solution satisfying all the constraints, and terminate
the process.
The approach that is the reverse of the greedy method, which starts from a solution
satisfying all the constraints and reduces the objective value consecutively, is called
the stingy method. The greedy method and the stingy method work fairly well if
F (J ) and Hj (J ) are monotonic functions of J. However, the algorithms strongly
depend on the initial solution, and will often reach a local optimal solution.
2.6.3 Simulated annealing

Instead of defining complicated rules for moving to a neighborhood solution, we
can use a probabilistic approach to improve the possibility of obtaining the global
optimal solution. The simplest probabilistic approach is the random search (Ohsaki,
2001). Although the random search is not efficient in view of computational cost
for reaching the global optimal solution, it is simple and very easy to implement.
Another approach that incorporates probabilistic process is the simulated an-
nealing (SA) that has been developed to prevent convergence to a local optimal
solution by allowing the move to a solution that does not improve the objective
value, where the probability of accepting such a non-improving solution is defined
by the amount of increase (for minimization problem) of the objective value. The
term simulated annealing comes from the fact that it simulates the behavior of the
metals in the annealing process. The penalty function approach is usually used for
constrained optimization problems. The basic algorithm of SA for a minimization
problem is given as follows (Aarts and Korst, 1989; Kirkpatrick, Gelatt and Vecchi,
1983):
Step 1 Randomly generate the initial solution J (0) . Initialize the temperature
parameter as T (0) = T0 , where T0 is the specified value, and set the iteration
counter k = 0.
Step 2 Let J denote a randomly generated neighborhood solution, and define
F = F (J ) F (J (k) ). If F < 0, let J (k+1) = J ; otherwise, accept J with
the probability P defined by the temperature parameter T (k) as

F
P = exp (k) (2.34)
T
Step 3 Decrease T (k) to T (k+1) based on the prescribed rule, e.g., the Metropolis
rule T (k+1) = T (k) , where is the specified parameter that is slightly less than
1.
Step 4 Set k k + 1 and go to Step 2 if the termination condition is not satisfied;
otherwise, output the best solution satisfying the constraints, and terminate the
process.
Optimum design
Shape optimization
Cross-sectional Topology Geometry
optimization optimization optimization
Fig. 2.4 Classification of optimization of trusses and frames with respect to design variables.
P P
Fig. 2.5 Topology optimization.
2.7 Classification of Structural Optimization Problems
There are several approaches to the classification of structural optimization prob-

lems. One of the traditional classifications for trusses and frames is as shown in
Fig. 2.4. The cross-sectional properties are optimized in cross-sectional optimization
(sizing optimization). The shape optimization includes geometry optimization (con-
figuration optimization) and topology optimization, where the nodal locations and
connectivity between nodes by members are optimized, respectively (Kirsch, 1989;
Nakamura and Ohsaki, 1992; Ohsaki, 1997; Ohsaki, Nakamura and Isshiki, 1998a;
Ohsaki and Katoh, 2005). Simultaneous optimization of geometry and topology is
also called layout optimization (Rozvany, 1997).
For continuum structures such as plates and shells, the optimization of external
and/or internal boundary shape is called shape optimization, and the optimization
allowing addition and removal of holes is called topology optimization (Bendse and
Sigmund, 2003; Eschenauer, 2003).
A process of topology optimization of a truss is illustrated in Fig. 2.5, where
the dashed lines indicate the removed members after optimization. This approach
of removing unnecessary members from the highly connected initial structure is
called the ground structure approach. Thus, topology optimization is conceived as
P P
Feasible region
Fig. 2.6 Geometry optimization.
Ceiling
Wall
Cable
H
Rod
P = 1000 N
W = 1000 mm
Fig. 2.7 A model of coat-hanger problem.
an extension of cross-sectional optimization.

Figure 2.6 illustrates the process of geometry optimization of a truss, where the
gray squares indicate feasible regions of the nodal locations, which are assigned
so that the shape does not change drastically. Note that the feasible regions of
the nodes should also be appropriately assigned to prevent the existence of too
short members or overlap of nodes called melting nodes that results in singularity
of the stiffness matrix, i.e., the topology does not change in the process of geometry
optimization. Therefore, simultaneous optimization of topology and geometry is
very difficult (Ohsaki, 1997).
In the following, effectiveness of topology optimization and geometry optimiza-
tion is demonstrated using a simple rod supported by a cable as shown in Fig. 2.7.
The objective is to design a structure to hang a object with the weight of 1000 N
at a place 1000 mm from the wall. This very small illustrative problem is called
the coat-hanger problem. We use mm and N for the units of length and force,
respectively, which are omitted for brevity.
The constraints for the stresses c and r of the cable and rod, respectively, are
( 10 4 )
8.0
Total structural volume

7.0
6.0
5.0
30 40 50 60 70
Angle of cable (deg.)
Fig. 2.8 Relation between the angle of the cable and the optimal total structural volume.
W W
H H
P P
(a) (b)
Fig. 2.9 Examples of topologies for H/W = 0.3.
given as
c 50, 50 r (2.35)
The objective function is the total structural volume V that is to be minimized.

Note that the cable and rod are modeled by the truss elements, and the buckling
of the rod is not considered. Let denote the angle between the cable and the rod.
Then the length of the cable is 1000/ cos , and the axial forces of the rod and cable
are 1000/ tan and 1000/ sin , respectively.
We first consider the sizing optimization problem, where is fixed. If we de-
crease the cross-sectional areas, then the structural volume decreases; however, the
absolute values of the stresses increase. Consequently, there exist optimal cross-
sectional areas under stress constraints, where c = 50 and r = 50 are satisfied,
i.e., the optimal solution of this statically determinate truss is fully stressed. Hence,
the optimal cross-sectional areas of the cable and rod are 20/ tan and 20/ sin ,
respectively, and the value of V of the optimal solution is obtained as

1 1
V = 2.0 104 + (2.36)
tan sin cos
Fig. 2.10 Optimal cross-sectional areas with four units for H/W = 0.2.
Table 2.1 Relation between the number of units and the optimal structural vol-
ume.
Number of units 2 3 4 5 6
Optimal structural volume (105 ) 1.620 1.533 1.530 1.560 1.607
If we can modify in addition to the cross-sectional areas, i.e., if we consider

geometry optimization, the relation between the angle and the optimal structural
volume is as shown in Fig. 2.8. As is seen, V takes the minimum at ' 55
However, in a practical situation of the coat-hanger problem, may not attain

55 due to the restriction on the location of the ceiling. Obviously, if the height
H is small, the configuration as shown in Fig. 2.9(a) is not recommended, and
a truss as illustrated in Fig. 2.9(b) will be better. Therefore, we next consider
the topology optimization problem, where all the members are supposed to consist
of truss element, and the upper bound for the absolute values of the stresses is
50. Since the truss in Fig. 2.9(b) is statically determinate, the optimal design for
specified topology can be obtained by assigning the cross-sectional areas so that the
truss is fully stressed, i.e., the absolute value of the stress of each member is equal
to the upper bound.
Optimal solutions have been found for the truss with the number of units 2,
3, 4, 5 and 6, respectively, for the fixed values of W = 1000 and the aspect ratio
H/W = 0.2. The relation between the number of units and the optimal structural
volume is listed in Table 2.1. As is seen, the structural volume takes its minimum
when the number of units is 4. The optimal cross-sectional areas for this case
are as shown in Fig. 2.10, where the width of each member is proportional to its
cross-sectional area. This way, the objective value can be effectively reduced by
optimizing the topology.
2.8 Probabilistic Optimization
Optimization of structures in the random environment was pioneered apparently

by Hilton and Feigen (1960). It attracted the attention of early investigators; the
reader may consult the works by Switzky (1964), Khachaturian and Haider (1966),
Moses and Kinser (1967), Vanmarcke, Diaz-Padilla and Roth (1972), Kapur (1975),
Parimi and Cohn (1975, 1978a, 1978b), Frangopol (1995, 1998), Frangopol and
Corotis (1996), and many followers thereafter. A review of current state-of-the-art
was given by several researchers. Hilton and Feigen (1960) defined the task of the
probabilistic optimization as the proportioning of probabilities of failure among

structural components in terms of a preassigned probability of failure of the entire
structure, such that the total structural weight is a minimum. The papers by
Pochtman and Khariton (1976) and Frangopol and Maute (2005) give a careful
glimpse of early and recent developments, respectively. Modern state-of-the-art
is summarized by Rackwitz and Cuntze (1987), Marti (2005), and Tsompanakis,
Lagaros and Papadrakakis (2007).
The problem of optimal design can be posed in one of the following two alter-
native forms:
Minimize objective function (usually cost, weight or volume) subject to the con-
straint on probabilistic measure on mechanical performance. Such a probabilistic
measure usually is taken as the overall probability of failure of the structure, i.e.,
Pf (A) Pfa (2.37)
where A = (A1 , . . . , Am )> is the vector of design variables, Pf (A) is the proba-
bility of failure, and Pfa is the acceptable probability of failure, the value of which
ought to be available in codes.
Maximize the reliability of the structure R(A)
R(A) = 1 Pf (A) (2.38)
which is the complement of the probability of failure to unity, subject to the
upper bound for the cost, weight or volume denoted by C(A). In other words,
this constraint reads
C(A) Ca (2.39)
where Ca is the allowable maximum value of C(A).
There are also two equivalent approaches: (a) reliability index approach, and (b)
performance measure approach (Frangopol and Maute, 2005).
Often the probabilistic optimization is posed as the problem of minimizing the
overall cost of the structure
C = C 0 + P f Cf (2.40)
where C0 is the initial cost of the structure, the product Pf Cf is the cost of the
failure, and Cf is the projected cost incurred by the failure (Sarma and Adeli,
2002).
As Gasser and Schueller (1997) note,
The decision making process may be improved when the risk of
failure in terms of expected failure consequence costs are balanced
with the structural costs. For this purpose, an objective function
should contain the following types of costs: the initial construction
cost, and the expected consequences related to partial and total
system failure.
In Eq. (2.40), Cf is a deterministic quantity. However, some researchers treat it

as a function depending on random variables. Then the reliability-based optimiza-
tion reads
m
X
minimize E[C(A, Y )] = C0 (A) + E[Cf,j (A, Y )] (2.41)
j=1
where Y is the vector describing random parameters of the system, and E[C] and
E[Cf ] are the mean values of C and the expected cost Cf due to failure with regard
to the jth criterion, respectively. In his review article, Schueller (1998) emphasizes
that ... irrespective of the route of optimization which is followed, its treatment
within efficient software environment is an indispensable requirement.
A pertinent question arises concerning the assignment of the acceptable proba-
bility of failure. Freudenthal (1956), anticipating the reliability-based optimization,
stressed the following about acceptable risk, which ought to be evaluated:
... on the basis of economic balance between the cost of increasing

safety and the cost of failure...
Hence, the cost and probability of failure should be accurately defined for reliable
formulations of probabilistic or reliability-based optimization (see further discussion
in Chap. 10).
2.9 Fuzzy Optimization
It appears that we ought to describe all (that are known to us, obviously) approaches
to optimization of structures, so as not to leave an impression that we include only
what we think is important in this field.
The first study on fuzzy sets was apparently pioneered by a British-American
philosopher, Max Black (19091988) over 70 years ago. He started by quoting the
famous British philosopher Bertrand Russell (18721970), who stated that Vague-
ness and accuracy are important notions, which are very necessary to understand.
Black (1937) introduces the essence of his argument as follows:
It is a paradox, whose importance familiarity fails to diminish, that

the most highly developed and useful scientific theories are ostensi-
bly expressed in terms of objects never encountered in experience.
The line traced by a draftsman, no matter how accurate, is seen
beneath the microscope as a kind of corrugated trench, far removed
from the ideal line of pure geometry. And the point planet of as-
tronomy, the perfect gas of thermodynamics, or pure species
of genetics are equally remote from exact realization.
Further,
While the mathematician constructs a theory in terms of perfect

objects, the experimental scientist observes objects of which the
properties demanded by theory are and can, in the very nature of
measurement, be only approximately true.
He further discusses the vagueness of symbols as a symptom of the degree of

deviation of the model language from the empirically discoverable linguistic habits
in the corresponding speech community.
Zadeh (1965) independently introduced fuzzy sets, and coined the term fuzzy.
In an insightful article Hirota (1991) explains:
The word fuzzy has the meaning vague, not clear, and is
used ... in contrast to the precision normally expected of mathe-
matics ... with fuzzy sets, a mathematical model can be made of
properties or quantities that are imprecisely defined, such as every-
day statements.
According to Kosko (1992), fuzzy theory holds that all things are matters of de-
gree ... Fuzzy theory reduces black-white logic and mathematics to special limiting
cases of gray relationships.
In applications of the fuzzy set theory, one first identifies the so-called universal
set (Klir, St. Clair and Yuan, 1997), which is a set that consists of all the members
that are of interest in the problem at hand. For example, if one is dealing with
classifying loads applied at a structure by various criteria, then the universal set is
composed of all the values that the loads can take on.
Every subset A of a set X can be uniquely represented by a function, called
characteristic function A (x), defined as

1, if x A
A (x) = (2.42)
0, if x /A
for any x X . Examples of characteristic functions are shown in Fig. 2.11. The
function c (x) in Fig. 2.11(a) corresponds to a crisp set or a classical non-fuzzy set
defined by 0 x x0 , where c (x) = 1 for 0 x x0 ; otherwise, c = 0. However,
membership of the load in a fuzzy set may allow some uncertainty; therefore, it is
a matter of degree (Kosko, 1992; Klir, St. Clair and Yuan, 1997). One can state
that the degree of membership constitutes the degree of compatibility of introduced
operating concepts with the fuzzy set. Figure 2.11(b)(d) illustrates three types
of characteristic functions, where the function value 0 F (x) 1 indicates the
corresponding value of x is possibly the member of the set; hence, the characteristic
function is also called the membership function.
For example, consider the uniform bar under tensile load, which is chosen as
variable x. Then X can be identified with the set of all nonnegative real numbers,
assuming the material strength of the bar is unbounded. Let N be a set of loads
that are real values and range from 10 kN to 20 kN. Then, if we use the framework
c F
1 1
0 x0 x 0 x0 x
(a) crisp set (b) fuzzy set (type 1)
F
F
1 1
0 xL xU x 0 xL xU x
(c) fuzzy set (type 2) (d) fuzzy set (type 3)
Fig. 2.11 Examples of characteristic functions.
of a non-fuzzy set, N is a subset of X with the characteristic function

1, if 10 x 20
N (x) = (2.43)
0, otherwise
However, if we use the concept of a fuzzy set, the characteristic function N (x) may
be defined, e.g., as shown in Fig. 2.11(c) with xL = 10 and xU = 20.
As is seen, the definition of a fuzzy set is a generalization of that of a classical
set; in the former, the characteristic function is allowed, as it were, to take any
value between zero and unity. In the latter, the value of the membership function
is either zero or unity, which follows the Greek philosopher Aristotles (384322
B.C.E.) maxim: Everything must either be or not be, whether in the present or in
the future. Fuzzy set theory follows Bertrand Russell (18721970) who stated in
his book (Russell, 1985) Everything is vague to a degree you do not realize till you
have tried to make it precise.
Fuzzy optimum design of structures and machining processes was apparently
pioneered by Dubois (1987), Rao (1987a, 1987b, 1987c), Wang and Wang (1984,
1985), and was studied extensively by Arakawa and Yamakawa (1993), M oller, Grab
and Beer (2000), Rao, Sundararaju, Prakash and Balakrishna (1992a, 1992b), Shih,
Chi and Hsiao (2003), Xu (1989), Yeh and Hsu (1990), and others. Zheng and
Lewis (1993) conducted optimization of grey systems.
Bernardini (1999) presented a multiobjective programming approach based on
fuzzy theory. Suppose we have N F (x) objective functions F1 (x), . . . , FN F (x), for
which the membership function is given as 1 , . . . , N F . Then the membership
function (x) for the intersection of the objective functions is given as

(x) = min i (x) (2.44)
i
Therefore, the multiobjective optimization problem is formulated as

Find max min i (x) (2.45)
x i
with appropriate deterministic constraints.

The optimization problem in a fuzzy environment can be formulated as the
following minmax problem with deterministic inequality constraints:
max min Dj (x) (2.46a)
x j=1,...,h+1
subject to Hi (x) 0, (i = 1, . . . , N I h) (2.46b)

where Dj is the membership function of fuzzy decision, and x is the design vari-
able vector. The fuzzy environment is represented by the objective function and h
constraints out of N I inequality constraints; the rest represent the crisp (non-fuzzy)
deterministic constraints. The fuzzy constraints are incorporated into the objective
function; therefore, we have h + 1 membership functions in the objective function.
We must not hide the fact that many researchers criticize the fuzzy-sets based
approach in strong terms. Here are several quotes (see Kosko (1993)):
Fuzzy theory is wrong, wrong and pernicious. What we need is more logical
thinking, not less. The danger of fuzzy logic is that it will encourage the sort
of imprecise thinking that has brought us so mush trouble. Fuzzy logic is the
cocaine of science. (William Kahan)
Fuzzification is a kind of scientific permissiveness. It tends to result in socially
appealing slogans unaccompanied by the discipline of hard scientific work and
patient observation. (Rudolf Kalman)
Fuzziness is probability in disguise. I can design a controller with probability
that could do the same thing that you could do with fuzzy logic. (Myron Tribus)
It appears noteworthy to quote Ditlevsen (1980):
... the sources of fuzzy information are non-objectivistic and non-

reproductive in their very nature like the process of perception in
the human brain.
Zadeh (1994) vehemently disagrees with these criticisms in his tellingly titled
paper Why the success of fuzzy logic is not paradoxical. The following definition
is proposed:
... the skeptics will find it hard to understand why they failed to
realize that fuzzy logic is a phase in a natural evolution of science
an evolution brought by the need to find an accommodation with
the pervasive impression of the real world.
In their paper Dubois and Prade (1993) try to overcome misunderstandings

between the proponents of the stochastic and fuzzy-sets based methodologies (see
also Dempster (1967); Blockley (1985), Bezdek (1994)). Kosko (1990) strongly
defends fuzzy sets:
... a hundred years from now, a thousand years from now... no

one... will believe that there was a time when a concept as simple,
as expressive as a fuzzy set met with such impassional denial.
To conclude this section, we note that researchers performed comparison of

fuzzy and probabilistic modelings. The interested reader may consult, for exam-
ple, with papers by Chiang, Dong and Wong (1987), Lodwick, Jamison and Russell
(2000), Oberguggenberger and Russo (2001), and Moens and Vandepitte, (2005,
2006). Maglaras, Nikolaidis, Haftka and Cudney (1997) performed a definitive
analytical-experimental comparison between these two methods. These two ap-
proaches were also contrasted by Sophie (2000). A fuzzy safety measure was intro-
duced by Blockley (1977), Brown (1979), Kam and Brown (1983), Brown, Johnson
and Loftus (1984), Shiraishi and Furuta (1985), Brown and Yao (1987), and others.
Safety factor in the fuzzy environment was introduced by Elishakoff and Ferracuti
(2006a, 2006b). Optimization of structures in fuzzy environment was studied by
Rao (1987a, 1987b, 1987c), Rao, Sundararaju, Prakash and Balakrishna (1992a,
1992b), Arakawa and Yamakawa (1992, 1993), and Ohta and Haftka (1997). Possi-
bility theory was employed by Utkin and Gurov (1996), Cremona and Gao (1997),
Gurov and Utkin (1999), Jensen (2001), Stroud, Krishnamurthy and Smith (2001),
Jensen and Maturana (2002), Mourelatos and Zhou (2005), and Du, Choi and Youn
(2006) among many others. Nikolaidis, Chen, Cudney and Haftka (2004) provided
a systematic comparison of probabilistic and possibilistic methodologies. Safety
analysis of structure under hybrid uncertainty was performed by Chakraborty and
Sam (2006). Some authors lump both concepts to yield fuzzy reliability (Cai, 1996)
or fuzzy randomness (M oller and Beer, 2004).
Fang, Smith and Elishakoff (1998) combined anti-optimization and fuzzy-sets
based analyses. Adali (1992) and Pantelides and Ganzerli (2001) compared these
two models. Neumaier (2004) introduced a new concept, namely, that of a cloud as
a mediator as it were, between the concept of a fuzzy set and that of a probability
distribution. He states A cloud is to a random variable more or less what as interval
is to a number.
Elishakoff and Li (1999) combined probabilistic and ellipsoidal analyses, whereas
Qiu, Yang and Elishakoff (2008b) presented a hybrid probabilistic-interval analysis
study. Wang, Qiu and Elishakoff (2008b) proposed a non-probabilistic set-theoretic
model for structural safety measure. Good (1995), Elishakoff (1995a), Guo and L u
(2003), and Qiu, Chen and Wang (2004b) discussed a concept of non-probabilistic
reliability. The book edited by Ross, Booker and Parkinson (2002) is devoted to
bridging the gap between fuzzy logic and probability applications. Natke and Ben-
Haim (1997), Elishakoff (1990, 1998c), Pal (1999), and Oberguggenberger (2004)
discuss various possible modeling techniques of uncertainty. Finally, a mention
should be made of random sets that combine the probabilistic and set-based models
(Matheron, 1975; Dubois and Prade, 1991; Bernardini, 1999; Tonon, 2004). Opti-
mization of uncertain structures in the context of the random sets was performed by
Tonon and Bernardini (1998). M oller and Beer (2008) studied optimization in con-
junction with non-probabilistic uncertainty modeling. Popular and highly readable
books on the fuzzy sets and logic are those by McNeill and Freiberger (1993), Kosko
(1993, 1999), and Sangalli (1998). Pertinent books that employ the fuzzy sets in
the structural analysis context are those by Blockley (1980), Yao (1985), Ayyub
(1998), Vick (2002), M oller and Beer (2004), Hanss (2005), and Fellin, Lessmann,
Oberguggenberger and Vieider (2005). Nikolaidis and Haftka (2001) revised theo-
ries of uncertainty for risk assessment when data are scarce. Tsompanakis, Lagaros
and Papadrakakis (2007) edited a definitive collection of papers on various methods
of uncertainty analysis in the optimization context. Nikolaidis, Ghiocel and Singhal
(2005) edited Engineering Design Reliability Handbook which provides much more
than the title promises; the handbook covers both the traditional probabilistic re-
liability as well as various non-probabilistic treatments of uncertainty.
Chapter 3
General Formulation of Anti-Optimization
We have to do the best we can. This is our sacred human responsi-

bility. (Albert Einstein)
By contrast to the well-known probabilistic approach, the guaranteed

approach does not require precious knowledge of probability distribu-
tions (which are seldom available in practical problem) and yields re-
liable estimates for the system behavior. (Felix L. Chernousko, 1999)
We are not suggesting that all computation should be carried out using
interval techniques but only that interval methods provide another set
of tools in applied mathematics... (Ramon E. Moore, 1979)
In this chapter, a general formulation of anti-optimization is described briefly. We

first provide a motivation, accompanied by a simple example of interval analysis
of a bar subjected to loads that vary within intervals. Pros and cons of interval
analysis are presented. Ellipsoidal analysis is described next, along with sensitivity
analysis and exact reanalysis of static response.
3.1 Introduction
Anti-optimization represents an alternative and a complement to traditional prob-

abilistic methods, and also is a generalization of the mathematical theory of the
interval analysis.
It is instructive, as it often is, to start with a story. This one is by Hayes (2003):
On February 25, 1991, a Patriot missile battery assigned to pro-

tect a military installation at Dhahran, Saudi Arabia, failed to
intercept a Scud missile, and the malfunction was blamed on an
error in computer arithmetic. The Patriots control system kept
track of time by counting tenths of a second; to convert the count
47
into full seconds, the computer multiplied by 1/10. Mathemat-

ically, the procedure is unassailable, but computationally it was
disastrous. Because the decimal fraction 1/10 has no exact finite
representation in binary notation, the computer had to approxi-
mate. Apparently, the conversion constant stored in the program
was the 24-bit binary fraction 0.00011001100110011001100, which
is too small by a factor of about one ten-millionth. The discrep-
ancy sounds tiny, but over four days it built up to about a third
of a second. In combination with other peculiarities of the control
software, the inaccuracy caused a miscalculation of almost 700 me-
ters in the predicted position of the incoming missile. Twenty-eight
soldiers died.
Of course it is not to be taken for granted that better arithmetic
would have saved those 28 lives. (Many other Patriots failed for
unrelated reasons; some analysts doubt whether any Scuds were
stopped by Patriots.) And surely the underlying problem was not
the slight drift in the clock but a design vulnerable to such minor
timing glitches. Nevertheless, the error in computer multiplication
is mildly disconcerting. We would like to believe that the mathe-
matical machines that control so much of our lives could at least
do elementary arithmetic correctly.
Two other examples belong to Moore (2006), a recognized founding father of

modern interval analysis:
(i) A physics problem at Lockheed (ca. 1960): Q: Is the strange behavior of the
computer model due to round-off errors? A: After converting the program to run
in interval arithmetic with outward rounding, it was determined that round-off
error was very small in the original program. Result: the physicist took another
look at the model equations and found that there was a missing term.
(ii) A long controversy between research groups at MIT and Caltech concerned
whether the observed behavior of computer simulations was due to roundoff error
of defects in the mathematical model, in the case of computer solutions of the
Birkhoff-Rota complex partial differential/integral equations modeling the onset
of turbulence in wind-shears. My graduate student Jeffrey Ely wrote a program
for variable-precision interval arithmetic with outward rounding, and finally by
using about 300 decimal place (nearly 1000 bits) in outwardly rounded interval
arithmetic, was able to settle the controversy. By carrying enough bits, the model
realistically determined the onset of turbulence at a reproducible, finite time after
the initial appearance of the wind-shear. I have always found it odd to suppose
that anything we want to compute can be done carrying only some fixed number
of digits or bits. Is 40 bits enough? 80? A thousand? It depends on what we are
trying to compute...
General Formulation of Anti-Optimization 49
The natural question arises: What is interval analysis? As Hayes (2003) informs
us, Interval analysis is not a new idea. Invented and reinvented several times, it
has never quite made it into the mainstream of numerical computing, and yet it
has never been abandoned or forgotten either. The first paper on interval analysis
apparently belongs to Rosalind Cicely Young (1931) who published a paper on an
algebra of many-valued quantities and gave rules for calculating with intervals. Paul
S. Dwyer (1951) described arithmetic with intervals, calling it range numbers. The
consequent three works appeared nearly simultaneously: by Mieczyslaw Warmus
(1956) in Poland, Teruo Sunaga (1958) in Japan, and Ramon E. Moore (1966) in
the United States.
3.2 Models of Uncertainty
There are several concepts for modeling uncertainty based on the non-probabilistic
approach. Ladevèze, Puel and Romeuf (2006) introduced the concept of lack of
knowledge to take into account uncertainty of the parameters and modeling er-
rors. They proposed a method for reducing errors through additional information,
and applied it to static and free vibration problems. Oden, Babuska, Nobile, Feng
and Tempone (2005) classified the errors of numerical analysis into modeling error,
discretization error and error due to uncertainty. Soize (2005) classified the un-
certainty into data uncertainty and model uncertainty. Data uncertainty concerns
the structural parameters defining geometry, stiffness, material properties, and so
on, which can be successfully represented by the parametric probabilistic approach.
Model uncertainty, on the other hand, is related to approximation in analysis and
the details that are actually unknown. They applied their methods to dynamic
problems. Probabilistic and non-probabilistic approaches to anti-optimization are
compared by Elishakoff and Zingales (2003). Various models of uncertainty are
reviewed by Elishakoff and Fang (1995).
Let a = (a1 , . . . , aN )> denote the vector of uncertain parameters, such as nodal
loads and material properties. The types of the feasible regions or the bounds of
the uncertain parameters are classified as follows (Qiu, M uller and Frommer, 2001):
The parameters are linearly bounded by the intervals:
aLi ai aU
i , (i = 1, . . . , N ) (3.1)
where aLi and aU

i are the specified lower and upper bounds for ai (see Sec. 3.3
and Chap. 7 for details of the interval analysis).
A quadratic bound is given by the ellipsoidal model:
XN 2
ai
1 (3.2)
i=1
gi
where gi is the specified positive parameter that represents the semi-axis of ai .

Equation (3.2) can be rewritten in a vector-matrix form:

a> W a 1 (3.3)
where W = diag(1/g1 , . . . , 1/gN ) is the weight matrix. Note that a has been
translated to have zero central values, and rotated from the original axes so that
W is diagonal with respect to a1 , . . . , aN (see Sec. 3.4 for details of the ellipsoidal
model).
For transient responses, the uncertain parameter ai (t) of time t has the envelope
bound:
aLi (t) ai (t) aU
i (t), (i = 1, . . . , N ) (3.4)
where the lower and upper bounds aLi (t)
and aU
i (t),
respectively, are the specified
deterministic functions.
An integral of a quadratic norm of ai (t) in the time interval [t1 , t2 ] is bounded as
Z t2
[ai (t)]2 dt , (i = 1, . . . , N ) (3.5)
t1
where is the specified upper bound.
The vector of Fourier coefficients i for a function ai in space- or time-domain is
bounded as
>
i W i i i , (i = 1, . . . , N ) (3.6)
where W i is the specified positive definite weight matrix, and i is the specified
upper bound.
3.3 Interval Analysis
3.3.1 Introduction
In the theory of structural analysis, we often encounter the situation that once the
problem is formulated and the governing differential equations are written down,
the exact solution is either unavailable or demands elaborate symbolic analysis
with special functions. In these circumstances, we usually resort to approximation
techniques like the RayleighRitz or Galerkin method. For the eigenvalues, some
enclosure methods are developed. Thus, intervals which contain the exact solution
of the formulated problem are to be found.
Archimedes, in the third century before the common era (circa 287212 B.C.E.)
derived such bounds for the transcendental number , showing that it belongs to
the interval
10 1
3 <<3 (3.7)
71 7
by approximating the circle with the inscribed and circumscribed 96-side regular
polygons.
In this section, we present an example of interval analysis of simple static prob-
lems.
[12.5, 14.7]
[2.32, 2.41]
[12.5, 14.7]
M = [mL , mU]
Fig. 3.1 A cantilever beam.
N1
N2
R N1
N2
Fig. 3.2 A uniform bar (type-1).
3.3.2 A simple example

In structural engineering, the intervals appear as the means of representing un-
certainty in the values of the physical quantities. We will now introduce basic
operations with interval variables.
Let us consider a cantilever beam as shown in Fig. 3.1, which is subjected to a
load that varies in the interval [12.5, 14.7] kN at the location that is also not known
exactly and varies in the interval [2.32, 2.41] m. Then the reaction moment is also
an interval quantity whose bounds [mL , mU ] kN m are
mL = 12.5 2.32 ' 28.9, mU = 14.7 2.41 ' 35.4 (3.8)
Next, we consider the uniform bar (type-1) as shown in Fig. 3.2, which is clamped
at its left end, and is subjected to two loads of uncertain magnitudes. Still, we can
state with confidence that the values that the loads can take lie within certain
bounds. Let the first load N1 lie between 17.4 kN and 21.5 kN, whereas the second
load N2 lies somewhere between 12.5 kN and 14.7 kN. The reaction R = N2 + N1
then lies in the interval
[12.5, 14.7] + [17.4, 21.5] = [12.5 + 17.4, 14.7 + 21.5] = [29.9, 36.2] kN (3.9)
The difference of interval variables appears if the direction of the load N1 is

reversed as shown in Fig. 3.3. Then the reaction turns out to be the difference of
N1
N2
R N1
N2
Fig. 3.3 A uniform bar (type-2).
two interval valued loads as

R = N2 N1 = [12.5, 14.7] [17.4, 21.5]
(3.10)
= [12.5 21.5, 14.7 17.4] = [9.0, 2.7]
Since the upper and lower bounds are negative, we conclude that the direction of
the force R should be reversed. Then it takes values in the interval [2.7, 9.0].
The division of two interval variables occurs, for example, in the following prob-
lem. Suppose the cross-sectional area A of the bar is also an interval variable
A = [0.05, 0.06] m2 (3.11)
Then the normal stress = N/A for the uniform bar (type-1) in Fig. 3.2 is also
an interval variable. The maximum stress U occurs when the loads N1 and N2
take on the maximum values 21.5 and 14.7, respectively, resulting in the maximum
axial force N U = 36.2 kN; and the cross-sectional area takes on the minimum value
AL = 0.05 m2 in the region between the clamped end and the cross-section where
the force N2 is applied:
NU 36.2 2
U = = = 724 kN/m (3.12)
AL 0.05
The minimum value L of the stress near the clamped end is achieved when the loads
take on their minimum values resulting in the minimum axial force N L = 29.9 kN,
whereas the cross-sectional area takes on the maximum value AU = 0.06 m2 :
NL 29.9 2
L = U = = 498 kN/m (3.13)
A 0.06
Thus, although the stress near the clamped end is uncertain due to the uncertainty
of the loads and the cross-sectional area, we are sure that it takes values in the
2
interval [498, 724] kN/m .
3.3.3 General procedure

Interval analysis is a set theoretical approach that is closely related to fuzzy the-
ory (Mullen and Muhanna, 1999) (see Sec. 2.9 for the details of fuzzy sets). The
principles of interval algebra is discussed in a number of books such as Alefeld
and Herzberger (1983), Ferson (2002), Hansen and Walster (1992), Keafott (1996),
Moore (1962), Moore, Kearfott and Cloud (2009), Neumaier (1990), and Rohn
(2006).
An interval X of a real number x is defined with the upper bound xU and the
lower bound xL as
X = [xL , xU ] = {x|xL x xU } (3.14)
The following operations are defined for the addition and subtraction of the variables
in the intervals X = [a, b] and Y = [c, d]:
Addition: X + Y = [a + c, b + d]
(3.15)
Subtraction: X Y = [a d, b c]
Note that the subtraction can be reduced to the sum as
X Y = X + (1) Y (3.16)
Likewise, the division of the non-interval variable x by an interval variable Y , pro-
vided that Y does not contain zero, is reduced to the multiplication
x/Y = x [1/y U, 1/y L ] (3.17)
In general, a single formula provides a definition of the four standard arithmetic
operations to intervals. Let us denote by any of the operations +, , and /
(Muhanna, Mullen and Zhang, 2005), and the division by an interval variable that
includes zero is prohibited. Then X Y is defined as
X Y = [min xi y j , max xi y j ]
(3.18)
for i {U, L}, j {U, L}, {+, , , /}
which is alternatively written as
[xL , xU ] [y L , y U ]
= [min(xL y L , xL y U , xU y L , xU , y U ), (3.19)
L L L U U L U U
max(x y , x y , x y , x , y )]
In other words, one has to compute the four possible combinations of the lower and
upper bounds. The next step is to choose the smallest of the four values as the
lower bound, and to take the largest of them as the upper bound. We are sure that
every possible combination of x y lies within these limits.
It must be stressed that the interval analysis processes information obtained for
input data to provide the interval for the output. Moore (2006) writes:
Suppose we have measured that input parameters fall within cer-
tain upper and lower limits, then we can use interval inputs for
them, and interval computation is designed for just that sort of
thing.
For example, if we measure a width as w = 7.2 0.1, length as
l = 14.9 0.1, and height as h = 405.6 0.02, then the volume
V = w l h is within the interval
([7.1, 7.3] [14.8, 15.0]) [405.4, 405.8]
(3.20)
= [42599.432, 44435.1] = 43517.266 817.834
If desired, we could use outward rounding to retain contain-

ment with numbers having only one digit after the decimal point,
and find the volume is contained in the interval of numbers,
[42599.4, 4435.1] = 43517.3 817.9.
All I am saying in the above example is that IF all we know about
w, l, and h is that their values lie in the intervals, THEN all we
can say about V is that it is in the interval obtained by the given
endpoint of the input variables w, l, and h. If we need V more
successfully, then we have to measure w, l and h more accurately.
A question arises if the definition of the four algebraic operations and their use
is sufficient to obtain tight bounds of a function of an interval variable. Consider
the function (Alefeld and Claudio, 1998; Alefeld and Mayer, 2000)
x
f (x) = for x 6= 1 (3.21)
1x
with x being an interval variable x = [2, 3]. The derivative of this function equals
1 2x
f 0 (x) = (3.22)
(1 x)2
It vanishes when x = 1/2. Moreover, this value is outside the interval [2, 3]. There-
fore, the extreme values of f (x) are reached at the endpoints. Namely, the lower
bound is f L = f (2) = 2 whereas the upper bound is f U = f (3) = 3/2. Let us
now calculate straightforwardly f (x) by replacing x with the interval [2, 3]:
[2, 3] [2, 3] [2, 3] [2, 3]
f ([2, 3]) = = = = = [3, 1] (3.23)
1 [2, 3] [1, 1] [2, 3] [1 3, 1 2] [2, 1]
As we see, the true interval of the function f (x) is included in the one obtained by
the direct evaluation as follows:
[2, 3/2] [3, 1] (3.24)
As we observe, overestimation has taken place.
Another simple example is the uniform bar (type-2) subjected to two loads N1
and N2 , which are equal in their magnitude N , but have opposite directions as shown
in Fig. 3.3. The reaction equals R = N N = 0. However, if N is an interval value
[N L , N U ], then the expression for the reaction R = N N , evaluated via direct
use of interval algebra, leads to R = [N L , N U ] [N L , N U ] = [N L N U , N U N L ].
Since N U > N L , the lower bound of the reaction is negative as N L N U < 0, and
the upper bound is positive as N U N L > 0. Thus, the true zero value is contained
in the interval result [N L N U , N U N L ]. If, for example, N L = 1, N U = 2, then
[N L N U , N U N L ] = [1, 1]. Such an overestimation in the interval estimation
is called catastrophic by Muhanna and Mullen (2001).
What is the remedy for such an error?
Interval arithmetic cannot recognize the multiple occurrence of the same vari-
able. It evaluates the function R = N N as a function of two variables, namely,
R = N1 N2 , instead of evaluating the interval as a function of a single variable

R = N N . In other words, interval arithmetic is testing N N as if evaluating
N1 N2 with N2 equal to but independent of N1 .
If possible, one has to take care with such cases. This dependency problem can
be eliminated if the reaction is rewritten as R = N (1 1). Likewise, in the example
of Eq. (3.21), f (x) = x/(1 x) can be rewritten as
x 1
f (x) = = for x 6= 0, 1 (3.25)
1x 1/x 1
Direct use of interval arithmetic to Eq. (3.25) leads to the correct result
1
= [2, 3/2] (3.26)
1/[2, 3] 1
The reason for the overestimation is the fact that the rules that are valid for the
arithmetic for real numbers do not hold for interval arithmetic. The distributivity
property for real numbers x(y + z) = xy + xz is not valid for interval variation. If,
for example,
X = [3, 3], Y = [1, 3], Z = [3, 1] (3.27)
then
X(Y + Z) = [3, 3]([1, 3] + [3, 1]) = [3, 3][2, 2] = [6, 6] (3.28)
but
XY + XZ = [3, 3][1, 3] + [3, 3][3, 1] = [9, 9] + [9, 9] = [18, 18] (3.29)
We can confirm that the interval number X(Y + Z) is contained in the interval
[18, 18]; however, they are not equal. For interval numbers, the following property
of subdistributity generally holds:
X(Y + Z) XY + XZ (3.30)
For real numbers, cancellation takes place, namely, x x = 0; likewise x/x = 1
for x 6= 0. However, for interval numbers, as we have seen, X X 6= 0; moreover
X/X 6= 1. Indeed, Y Y = [1, 3][1, 3] = [2, 2], and Y /Y = [1, 3]/[1, 3] = [1/3, 3].
For interval numbers, the subcancellation property reads
0 Y Y, 1 Y /Y (3.31)
For the example of Y = [1, 3], the interval number Y Y = [2, 2] is enclosing but
not equal to zero, whereas in the case of division, the interval Y /Y = [1/3, 3] is
enclosing the true number 1 but not equal to it.
The distance q of two intervals X = [xL , xU ] and Y = [y L , y U ] is defined as a
real number
q(X, Y ) = max{|xL y L |, |xU y U |} (3.32)
L U
The absolute value of an interval X = [x , x ] is defined as the distance of X from
0, i.e.,
|X| = q(X, 0) = max{|xL |, |xU |} (3.33)
The diameter, or width w of an interval X = [xL , xU ] is defined as

w(X) = xU xL (3.34)
and the radius is given as w/2. The following rules are valid for two interval num-
bers:
w(X Y ) = w(X) + w(Y ), w(XY ) max{|X|w(Y ), w(X)|Y |} (3.35)
Consider the following example of Alefeld and Claudio (1998). Let f (x) = xx2
and
X = [1/2 r, 1/2 + r] (3.36)
where r is a real number taking values in the range 0 r 1/2. The function
attains its maximum value 1/4 at x = 1/2. The minimum value is attained at the
lower and upper bounds of the interval, namely at x = 1/2 r, and equals 1/4 r 2 .
Thus, the exact range of the function f (x), denoted as R(f ; X) is
R(f ; X) = [1/4 r 2 , 1/4] (3.37)
For the interval evaluation, we obtain
f (X) = [1/2 r, 1/2 + r] [1/2 r, 1/2 + r][1/2 r, 1/2 + r]
(3.38)
= [1/4 2r r 2 , 1/4 + 2r r 2 ]
Thus, the distance between R(f ; X) and f (X) is
q(R(f ; X), f (X)) = max{|1/4 2r r 2 1/4 + r2 |, |1/4 + 2r r 2 1/4|}
= max{2r, 2r r 2 }
= 2r (3.39)
Now the diameter of the interval X is

1 1
w(X) = xU xL = +r r = 2r (3.40)
2 2
As we see, the distance between R(f, X) and f (X) equals 2r, or coincides with the
width of the interval X. Moore (1966) has demonstrated that, under reasonable
assumption, the following inequality holds for the general interval number X:
q(R(f ; X), f (X)) w(X), 0 (3.41)
Note that = 1 in the example above. The inequality signifies that the overesti-
mation of R(f ; X) by f (X) approaches zero linearly with respect to the diameter
of interval number X.
Interval analysis found many applications in applied mechanics. First applica-
tions are apparently by Dimarogonas (1993), Elishakoff (1994), Dimarogonas (1995),
and Dessombz, Thouverez, Laine and Jezequel (2003). Numerous other references
such as Chen, Qiu and Liu (1994), Chen, Qiu and Song (1995), Qiu, Chen and Jia
(1995b), Qiu, Chen and Elishakoff (1995a, 1996a, 1996b), Jensen (2000) should also
be mentioned.
3.4 Ellipsoidal Model
3.4.1 Definition of the ellipsoidal model

In this section, we summarize what was apparently the first method of generat-
ing the multi-dimensional ellipsoidal model by Schweppe (1973). This method was
referred to by him as unknown-but-bounded. Further definitive contributions were
made by Chernousko (1994, 1999) and others who dubbed it as a guaranteed ap-
proach. The ellipsoidal model of uncertainty was utilized in the theory of control by
Schweppe (1973), and in modeling geometric imperfections in structures by Ben-
Haim and Elishakoff (1990) (see also an essay by Elishakoff (1991b)).
Suppose we have a vector of uncertain parameters a = (a1 , . . . , aN )> and M
experimental data points a(r) (r = 1, . . . , M ). The ellipsoidal convex model assumes
that all these points belong to an ellipsoid
(a a0 )> W (a a0 ) 1 (3.42)
where a0 is the center of the ellipsoid, and W is a positive definite constant N N
weight matrix.
The best ellipsoid with minimum volume is achieved by rotating the axes of the
ellipsoid. Let b = (b1 , . . . , bN )> denote the new coordinate transformed from a by
an N N matrix T as
b = Ta (3.43)
where the transformation matrix T is the function of the rotation angles
1 , . . . , N 1 with respect to the axes a1 , . . . , aN 1 , respectively. T can be con-
structed by any pair of orthogonal vectors, e.g., the GramSchmidt orthogonaliza-
tion. By using T , the experimental points are transformed as
b(r) = T a(r) , (r = 1, . . . , M ) (3.44)
To obtain the ellipsoid, we first consider the N -dimensional box
(r)
bk b0k dk , (k = 1, . . . , N ; r = 1, . . . , M ) (3.45)
0
that contains all M experimental points. The center b = (b01 , . . . , b0N )>
and the
vector d = (d1 , . . . , dN )> are defined so that the volume of the box is minimized.
The ellipsoid is then defined to enclose the box as
N
!2
X (r)
bk b0k
1, (r = 1, . . . , M ) (3.46)
gk
k=1
where gk is the kth semi-axis of the ellipsoid. We assume that gk is nonzero and
bounded.
The following condition is to be satisfied so that the ellipsoid passes through the
corners of the box:
XN 2
dk
=1 (3.47)
gk
k=1
The volume Ve is given as

N
Y
Ve = C gk (3.48)
k=1
where C is a constant depending on the dimension N of the ellipsoid. By minimizing
Ve under constraint (3.47), we obtain (Pantelides and Ganzerli, 1997)

gk = Ndk , (k = 1, . . . , N ) (3.49)
Then the ellipsoid is written as
(b b0 )> D(b b0 ) 1 (3.50)
with

D = diag 1/g12 , . . . , 1/gN
2
(3.51)
If there exists a corner of the box without experimental point, the semi-axis is
further reduced to gk , where
v !2
u
u XN (r)
bk b0k
t
= max 1 (3.52)
r gk
k=1
Hence, the ellipsoid finally is given as
e b0 ) 1
(b b0 )> D(b (3.53)
where

e = diag 1/(g1 )2 , . . . , 1/(gN )2
D (3.54)
The volume Ve is further minimized by considering the angles 1 , . . . , N 1 as vari-
ables. For this purpose, any method of nonlinear programming can be used.
3.4.2 Properties of the ellipsoidal model

The ellipsoidal model has been extensively investigated and extended by Chernousko
(1994), Kurzhanski and Valyi (1997), and others. Hereinafter, we closely follow
Chernousko (1999); however, we use the positive-definite N N weight matrix W
instead of W 1 for consistency with other parts of this book.
Consider the N -dimensional ellipsoid E defined using the inner (scalar) product
as
(W (a a0 ), (a a0 )) 1 (3.55)
0
which is denoted as E(a , W ). Thus,
E(a0 , W ) = {a|(W (a a0 ), (a a0 )) 1} (3.56)
The ellipsoid in Eq. (3.56) depends on N (N + 1)/2 parameters in W and N
parameters in a0 . Consider in general an affine transformation in N -dimensional
space as
b = Ta + h (3.57)
where T is a nonsingular N N transformation matrix and h is a constant vector. If

a belongs to an ellipsoidal set, b too belongs to an ellipsoidal set. Solving Eq. (3.57)
to obtain a = T 1 (b h) and substituting it into Eq. (3.55) result in
(W (T 1 (b h) a0 ), (T 1 (b h) a0 ))
(3.58)
= ((T > W T 1 )(b (T a0 + h)), (b (T a0 + h))) 1
where T > = (T 1 )> . We observe that b belongs to the ellipsoid with the center
T a0 + h and weight matrix T 1 W T > . In other words,
b = T a + h E(T a0 + h, T 1 W T > ) (3.59)
which is written as
b E(b0 , D) (3.60)
0 0 1 >
with b = T a + h and D = T W T .
As we have seen in Sec. 3.4.1, we can make its axes parallel to the principal axes
of the ellipsoid (3.59) with
D = diag(1/g12 , . . . , 1/gN
2
) (3.61)
2
where gi is the ith semi-axis of the ellipsoid. If D = r I, where r > 0 is a scalar, and
I is the N N identity matrix, then the ellipsoid (3.56) is a ball in N -dimensional
space with center b0 and radius r.
If W is positive definite and its eigenvalues are bounded, then W 1 is also pos-
itive definite and its eigenvalues are bounded. However, sometimes one encounters
a degenerate ellipsoid for which W 1 is not positive definite but positive semidef-
inite. In this case, D 1 = diag(g12 , . . . , gN
2
) is positive semidefinite, and some of
the semi-axes of the ellipsoid are zero (gi = 0 in Eq. (3.61)), and its volume is also
equal to zero. A disk in the three-dimensional space represents an example for this
case.
The second case, namely, when W is not positive definite but positive semidef-
inite, indicates that some of the semi-axes of the ellipsoid are infinite (gi = ,
in Eq. (3.61)), with the volume of the ellipsoid also being unbounded. A cylinder
represents an example for this case. When W 1 approaches zero, i.e., D 1 O,
all semi-axes vanish, and we obtain a point E(a0 , W ) = {a|a = a0 }. Hereinafter
we concentrate on non-degenerate ellipsoids, i.e., we assume W is positive definite
and bounded.
Ellipsoid (3.56) can be transformed into the unit ball E(0, I) by the affine trans-
formation (3.57). Indeed, according to Eq. (3.59), such a transformation is obtained
if
T > W T 1 = I, T a0 + h = 0 (3.62)
These conditions are satisfied if
T = W 1/2 , T 1 = W 1/2 , h = W 1/2 a0 (3.63)
1/2 1/2
where W is a square root of W . The existence of non-singular W and W 1/2
is guaranteed if W is a positive definite matrix.
The following theorems hold (Chernousko, 1999):
Theorem 1: For an arbitrary convex set D in N -dimensional space, there exists

an ellipsoid E(a0 , W ) D such that E(a0 , (1/N 2 )W ) contains D, i.e.,
E(a0 , W ) D E(a0 , (1/N 2 )W ) (3.64)
Theorem 2: Let D be a convex set in N -dimensional space, which has the property
of central symmetry with respect to some point a0 . There exists an ellipsoid
E(a0 , W ) D such that E(a0 , (1/N )W ) contains D, i.e.,
E(a0 , W ) D E(a0 , (1/N )W ) (3.65)
The proof of Theorem 1 is given in the book by Leichtweiss (1980), whereas the
proof of Theorem 2 is given by John (1998). For approximating sets, Chernousko
(1999) suggests using inner ellipsoid E L of the maximum volume and outer ellipsoid
E U of the minimal volume. The following theorems (Zaguskin, 1958) are relevant:
Theorem 3: For any bounded set D in N -dimensional space, there exists a unique
ellipsoid E U of the minimum volume containing D as E U D.
Theorem 4: For any closed convex set D in N -dimensional space, there exists a
unique ellipsoid E L of the maximum volume contained in D as E L D.
The set of all points a = a1 + a2 such that a1 A, a2 B is called the direct

sum (or the Minkowski sum) of the sets A and B, and is denoted by C = A B.
We treat the following ellipsoids in N -dimensional Euclidean space:
E(a01 , W 1 ) = {a|(W 1 (a a01 ), (a a01 )) 1}
(3.66)
E(a02 , W 2 ) = {a|(W 2 (a a02 ), (a a02 )) 1}
Here a01 and a02 are the N -dimensional vectors of the centers of ellipsoids, and W 1
and W 2 are the N N positive definite matrices. The Minkowski sum of these
ellipsoids reads
a = a1 + a2 S E(a01 , W 1 ) E(a02 , W 2 ),
(3.67)
a1 E(a01 , W 1 ), a2 E(a02 , W 2 )
where the set S is bounded, closed, and convex, and in general is not an ellipsoid.
Chernousko (1980) posed the following pertinent problems:
Problem 1 Find the ellipsoid E(aU , W U ) of the minimal volume containing the
sum S of two ellipsoids (3.67), i.e., an ellipsoid such that S E(aU , W U ),
det W U min.
Problem 2 Find the ellipsoid E(aL , W L ) of the maximal volume contained in
the sum S of two ellipsoids (3.67), i.e., an ellipsoid such that E(aL , W L ) S,
det W L max.
Chernousko (1980) solved both problems. For the details, see Chernousko (1999,
p. 143). Chernousko (2000) writes:
Dynamical systems with unknown, uncertain, or perturbed pa-

rameters appear in numerous applications. It is important to ob-
tain bounds on the reachable sets of such systems, in other words,
bounds on all possible motions of these systems when they are sub-
jected to unknown bounded perturbations.
Such reachable sets were derived by Chernousko (1988, 1994, 1999, 2005).
The apparently first applications of ellipsoidal analysis to applied mechanics was
the work by Borkowski (1983). Ellipsoidal analysis methods have been considered by
Attoh-Okine (2002, 2004), Au, Cheng, Tham and Zheng (2003), Ben-Haim (1992,
1999), Calafiore and El Ghaoui (2004), Chernousko (1980), Elishakoff (1991b, 1991c,
1994, 1995b, 1998a, 1998c, 2000b, 2004), Elishakoff and Ben-Haim (1990b), El-
ishakoff and Colombi (1993), Elishakoff, Gana-Shvili and Givoli (1991), Elishakoff,
Li, and Starnes (2001), Elishakoff, Lin and Zhu (1994e), Elseifi and Khalessi (2001),
Kurzhanski (1977), Kurzhanski and Valyi (1992, 1996), Lindberg (1992a, 1992b),
Pantelides and Booth (2000), Pantelides and Ganzerli (2001), Pantelides and Tzan
(1996), Qiu (2003, 2005), Qiu, Chen and Wang (2004a), Qiu, Ma and Wang (2006a),
Qiu, M uller and Frommer (2004e), Schweppe (1973), Tzan and Pantelides (1996a,
1996b), Vinot, Cogan and Lallement (2003), Wang, Elishakoff and Qiu (2008a),
Yoshikawa (2002, 2003), Yoshikawa, Nakagiri and Kuwazuru (1998b), Zhu and El-
ishakoff (1996), and Zuccaro, Elishakoff and Baratta (1998).
Ben-Haim (1985) observed that ellipsoidal analysis is a particular case of convex
analysis. A first application of convex analysis to applied mechanics are apparently
the works by Moreau (1966, 1976, 1979) and Panagiotopoulos (1976, 1985). The
comprehensive presentation of convex analysis in the applied mechanics context was
made by Ben-Haim and Elishakoff (1990).
In closing this section, it must be noted that in recent work by Kreinovich,
Neumaier and Xiang (2008) a combination of interval and ellipsoidal uncertainties
is treated. Specifically, they consider the interesting case in which the actual values
of uncertain variables belong to the intersection of interval and ellipsoid.
3.5 Anti-Optimization Problem
The response property in question is denoted by f (a), which is supposed to give

the worst value if it is maximized with respect to the uncertain parameters a.
Uncertainty may exist in material parameters, cross-sectional geometry, external
loads, and so on. If an ellipsoidal bound is given for a = (a1 , . . . , aN )> , the anti-
optimization problem is formulated as
maximize f (a) (3.68a)
XN 2
ai
subject to 1 (3.68b)
i=1
gi
where gi is the specified positive parameter that represents the semi-axis of the
parameter ai . Note that a has been appropriately transformed so that the ellipsoid
has its center at the origin 0 and the direction of each semi-axis coincides with a
coordinate axis.
The anti-optimal solution of Problem (3.68) can be found from the stationary
condition of the Lagrangian Q defined by
" N 2 #
X ai
Q(a, ) = f (a) + 1 (3.69)
i=1
gi
where 0 is the Lagrange multiplier. The stationary condition of Q with respect
to a leads to
Q f 2ai
= 2 = 0, (i = 1, . . . , N ) (3.70)
ai ai gi
The details of the Lagrange multiplier approach can be found in Sec. 2.4.2.
If f (a) is a linear function of a, or has been linearly approximated with respect
to a as
XN
f (a) = ci a i (3.71)
i=1
with the coefficient ci , then ai /gi is expressed from Eq. (3.70) as

ai gi ci
= (3.72)
gi 2
Since the objective function is linear and the feasible region of the variables a
is convex, the anti-optimal solution exists at the boundary of the feasible region.
Therefore, the constraint (3.68b) is satisfied in equality as
XN 2
ai
1 =0 (3.73)
i=1
gi
Incorporation of Eq. (3.72) into Eq. (3.73) results in
v
uN
1 uX
= = t (gi ci )2

(3.74)
2 i=1
Note that = corresponds to the minimum value of f (a). This way, the
anti-optimal solution can be obtained by a simple arithmetic operation.
If f (a) is a nonlinear function, then only locally anti-optimal solution can be
found by using a method of nonlinear programming. The global anti-optimality is
guaranteed when f (a) is a concave function.
Alternatively, the bound constraints can be assigned to the anti-optimization
problem as
maximize f (a) (3.75a)
subject to aLi ai aU
i , (i = 1, . . . , N ) (3.75b)
where aLi and aU i are the lower and upper bounds for ai .
Problem (3.75) is a linear programming problem, if f (a) is a linear function of
a. For this simple problem with bound constraints only, the solution is found at
ai = aLi or aUi depending on the sign of the coefficient for ai in f (a). For the case
where f (a) is a nonlinear function, the globally anti-optimal solution can be found
by a nonlinear programming approach, if f (a) is concave. On the other hand, if f (a)
is a convex or monotonic function of a, then the globally anti-optimal solution can
be found by enumerating the vertices of the feasible region defined by Eq. (3.75b)
(see Secs. 4.6 and 5.6 for examples of anti-optimization by vertex enumeration).
If ai 0 (i = 1, . . . , N ), the bounds for a are generalized as
" N # p1
X ai p
1, (p = 1, . . . , ) (3.76)
i=1
gi
Note that p = 2 corresponds to the ellipsoidal bound, and p = corresponds to

Eq. (3.75b) with aLi = 0 and aUi = gi (i = 1, . . . , N ) (see Eq. (2.31) in Sec. 2.6 for
various definitions of the norm). Special cases with N = 2 are discussed in Adali,
Lene, Duvaut and Chiaruttini (2003) (see also Kurzhanski (1977), Kurzhanski and
Valyi (1992, 1996, 1997), and Kurzhanski and Varaiya (2002)).
3.6 Linearization by Sensitivity Analysis
3.6.1 Roles of sensitivity analysis in anti-optimization

Sensitivity analysis is mathematically defined as a process of computing the deriva-
tives of the solution to governing equations with respect to the parameters defining
the equations.
In the field of structural analysis and design, sensitivity analysis has been devel-
oped as a tool for evaluating the derivatives (gradients or sensitivity coefficients) of
the structural responses with respect to variation of the design or shape parameters
(Haug, Choi and Komkov, 1986; Choi and Kim, 2004). Sensitivity analysis also
serves as a basic tool in various fields of mathematics and engineering, e.g., inverse
problem and stability analysis.
The procedures of sensitivity analysis for optimization are classified as follows
by the type of variables:
Design sensitivity analysis to find the sensitivity coefficients of the responses with
respect to the sizing design variables such as the cross-sectional areas of trusses
and thicknesses of the plate elements.
Shape sensitivity analysis to find the sensitivity coefficients of the responses with
respect to the shape design variables such as the nodal locations of the frames
and the plates discretized to finite elements.
The sensitivity coefficients of the responses with respect to the uncertain param-
y 4
3
x
Y 2 Node 2
1
X i
Node 1
Fig. 3.4 Coordinates, displacement numbers and node numbers of a plane truss member.
eters are to be computed in the process of anti-optimization. The role of sensitivity

analysis in anti-optimization is two-fold:
Linearly approximate the responses using sensitivity coefficients to formulate a
convex anti-optimization problem.
Evaluate the gradients of the objective and constraint functions with respect to
uncertain parameters to carry out anti-optimization using a nonlinear program-
ming approach.
The sensitivity coefficients can also be obtained by the perturbation equations
(see Sec. 4.4 for the static perturbation). However, in the field of numerical struc-
tural optimization, sensitivity equations are derived by differentiation of the gov-
erning equation as presented in this section.
3.6.2 Sensitivity analysis of static responses

In order to present the basic ideas of sensitivity analysis, the direct differentiation
method for static responses of plane trusses is briefly introduced. The global co-
ordinates (X, Y ), local coordinates (x, y), local node numbers, local displacement
numbers, and angle i of the member i to the X-axis are defined in Fig. 3.4.
The displacement vector uli = (ul1 , ul2 , ul3 , ul4 )> and the nodal force vector
l
f i = (f1l , f2l , f3l , f4l )> of member i with respect to the local coordinates are trans-
formed, respectively, to the displacement vector ugi = (ug1 , ug2 , ug3 , ug4 )> and the
nodal force vector f gi = (f1g , f2g , f3g , f4g )> with respect to the global coordinates by
the transformation matrix T i as
ugi = T i uli , f gi = T i f li (3.77)
where

cos i sin i 0 0
sin i cos i 0 0
Ti =
0
(3.78)
0 cos i sin i
0 0 sin i cos i
Let Ai and Li denote the cross-sectional area and the length of the ith mem-
ber, respectively. The 4 4 member stiffness matrix k li with respect to the local
coordinates is given as
Ai E >
kli = dd (3.79)
Li
where
d = (1, 0, 1, 0)> (3.80)
and E is Youngs modulus.
From Eqs. (3.77) and (3.79), the member stiffness matrix k gi with respect to the
global coordinates is given as
Ai E > >
kgi = T dd T i (3.81)
Li i
and kgi of all the members are assembled to construct the n n global stiffness
matrix K, where n is the number of degrees of freedom of the structure.
Let m denote the number of members. For design sensitivity analysis, where
the cross-sectional areas are considered as design variables, the stiffness matrix is
a function of A = (A1 , . . . , Am )> , which is written as K(A). If the self-weight is
considered, the nodal load vector P = (P1 , . . . , Pn )> is also a function of the design
variable vector A, which is written as P (A). The displacement vector U (A) =
(U1 , . . . , Un )> is obtained by solving the stiffness equation as
K(A)U (A) = P (A) (3.82)
In the following, the argument A is omitted for brevity.
Differentiation of Eq. (3.82) with respect to Ai leads to
K U P
U +K = (3.83)
Ai Ai Ai
Then the sensitivity coefficient of U with respect to Ai is obtained from
U P K
K = U (3.84)
Ai Ai Ai
Therefore, the sensitivity coefficients of P and K are needed to compute those of
the displacements.
By differentiating Eq. (3.81) with respect to Ai , we obtain the explicit relation
kgi E > >
= T dd T i (3.85)
Ai Li i
Hence, for a truss, the sensitivity coefficient of K with respect to Ai is easily
obtained by an arithmetic operation, because K is an assemblage of k gi . Sensitivity
of P is also easily computed, when the self-weight is considered, because P is an
explicit function of A. Note that K is decomposed to a product of triangular and
diagonal matrices by Cholesky decomposition in the process of computing U from
Eq. (3.82). Hence, the computational cost of solving Eq. (3.84) is very small even
for the case where the sensitivity coefficients with respect to all m design variables
A1 , . . . , Am are to be computed.
U1 U2
P1 P2
A1 A2
Fig. 3.5 A 2-bar structure.
Consider, for example, a 2-bar structure as shown in Fig. 3.5 subjected to axial
forces P1 and P2 at nodes 1 and 2, respectively. The displacements of nodes 1 and
2 are denoted by U1 and U2 , respectively. The two bars have the cross-sectional
areas A1 and A2 , and the same length L. Then Eq. (3.82) is written as

E A1 + A2 A2 U1 P1
= (3.86)
L A2 A2 U2 P2
which is easily solved for U1 and U2 as
(P1 + P2 )L (P1 + P2 )L P2 L
U1 = , U2 = + (3.87)
A1 E A1 E A2 E
Consider A1 as the design variable, and suppose that P1 and P2 are independent
of A1 . Equation (3.83) is then written as

U1

E 1 0 U1 E A1 + A2 A2 A1
0
+ = (3.88)
L 0 0 U2 L A2 A2 U2 0
A1
By incorporating Eq. (3.87) into Eq. (3.88), we obtain
U1 U2 (P1 + P2 )L
= = (3.89)
A1 A1 A21 E
which agrees with the direct differentiation of U1 and U2 in Eq. (3.87) with respect
to A1 .
The sensitivity coefficients can also be computed by the finite-difference ap-
proach. Let Ai denote the m-vector whose ith component is Ai and the other
components are 0, where Ai is the small variation of Ai . The approximate sensi-
tivity coefficients by the central finite-difference approach is given as
U U (A + Ai ) U (A Ai )
' (3.90)
Ai 2Ai
The approximate sensitivity coefficients can also be found by the forward finite-
difference approach
U U (A + Ai ) U (A)
' (3.91)
Ai Ai
or the backward finite-difference approach
U U (A) U (A Ai )
' (3.92)
Ai Ai
Although these finite-difference approaches are very simple, the number of response
analysis (solution process of Eq. (3.82)) for sensitivity evaluation with respect to
m design variables is m for the forward or backward finite-difference approach and
2m for the central finite-difference approach. Therefore, it is seen that the compu-
tational cost for sensitivity analysis is drastically reduced by direct differentiation
of the governing equations.
3.6.3 Sensitivity analysis of free vibration

We next consider the sensitivity analysis of eigenvalues and eigenmodes of free
vibration.
Let M (A) denote the n n mass matrix of the structure, which is a function of
the design variable vector A. The eigenvalue problem of free vibration is formulated
as
Kr = r M r , (r = 1, . . . , n) (3.93)
where r and r are the rth eigenvalue and eigenmode, respectively, which are
functions of A. The eigenmode r is ortho-normalized by
>
r M s = rs , (r, s = 1, . . . , n) (3.94)
where rs is the Kronecker delta. By differentiating both sides of Eq. (3.93) with
respect to Ai , we obtain
K r r M r
r + K = M r + r r + r M (3.95)
Ai Ai Ai Ai Ai
Premultiplying > r to both sides of Eq. (3.95), using Eqs. (3.93), (3.94), and
the symmetry of M and K, we obtain

r > K M
= r r r (3.96)
Ai Ai Ai
Note that K and M are explicit functions of Ai . Therefore, when the eigenvalue
r and eigenmode r are known after solving the eigenvalue problem (3.93) with
Eq. (3.94), the sensitivity coefficients of eigenvalues are computed from Eq. (3.96)
by simple arithmetic operation without resort to the sensitivity coefficients of the
eigenmodes.
Differentiation of Eq. (3.94) with respect to Ai leads to
M r
>r r + 2> r M =0 (3.97)
Ai Ai
From Eqs. (3.95) and (3.97),

r M K
r
K r M M r Ai r Ai Ai
r = (3.98)
>
r M 0 1 > M
r r
Ai 2 Ai
is derived. Therefore, the sensitivity coefficients of each pair of eigenvalue and
eigenmode can be found from Eq. (3.98), which is a set of n + 1 simultaneous linear
equations.
3.6.4 Shape sensitivity analysis of trusses

We next consider variation of nodal locations to compute the shape sensitivity
coefficients. A plane truss member as shown in Fig. 3.4 is used, for simplicity, also
in this section.
The 4 4 member stiffness matrix kgi with respect to the global coordinates is
given as Eq. (3.81). Since Li , sin i and cos i depend on the nodal coordinates, the
sensitivity coefficients of these values are needed to compute the shape sensitivity
coefficients of kgi .
Let (X1 , Y1 ) and (X2 , Y2 ) denote the global coordinates of nodes 1 and 2, re-
spectively, of member i. Then Li , sin i and cos i are given as
p
Li = (X2 X1 )2 + (Y2 Y1 )2 (3.99)
X2 X 1 Y2 Y 1
cos i = , sin i = (3.100)
Li Li
Note that (cos i , sin i )> is a unit vector directed from node 1 to 2.
Rewriting Eq. (3.99) as
L2i = (X1 X2 )2 + (Y1 Y2 )2 (3.101)
and differentiating it with respect to X1 and X2 , respectively, we obtain

Li 1
= (X2 X1 ) = cos i ,
X1 Li
(3.102)
Li 1
= (X2 X1 ) = cos i
X2 Li
where Eq. (3.100) has been used. The sensitivity coefficients of cos i are derived as

cos i 1 Li sin2 i
= 1 cos i =
X1 Li X1 Li
(3.103)
cos i 1 Li sin2 i
= 1 cos i =
X2 Li X2 Li
The sensitivity coefficients of sin i with respect to Y1 and Y2 are derived in a similar
manner.
Finally, the sensitivity coefficients of T i are easily obtained from Eq. (3.78), and
those of kgi with respect to the nodal coordinates; e.g., X1 are computed, as follows,
by differentiating Eq. (3.81):
kgi Ai E Li > > Ai E > > T i
= 2 T dd T i + T dd
X1 Li X1 i Li i X1
> (3.104)
T i Ai E
+ dd> Ti
X1 Li
which are assembled to the total structure and simply incorporated into Eq. (3.84)
to compute the shape sensitivity coefficients of the displacements.
Uj
reanalysis
sensitivity analysis
Ai Ai +Ai
Fig. 3.6 Estimation of displacement by sensitivity analysis and reanalysis.
3.7 Exact Reanalysis of Static Response
3.7.1 Overview of exact reanalysis

The differential coefficients of the responses with respect to the design variables are
computed by design sensitivity analysis as presented in Sec. 3.6. Then the responses
such as nodal displacements can be linearly estimated by the sensitivity coefficients
as shown in the straight tangent line in Fig. 3.6 that illustrates the variation of
the displacement component Uj with respect to the increment Ai of the design
variable Ai . On the other hand, the method for accurately estimating the responses
after modification of the variable by the specified amount is called exact reanalysis
as illustrated by the arrow in Fig. 3.6. Note that the term reanalysis does not mean
that responses are simply computed again after modification of the variable, but
that accurate responses are found by computationally more inexpensive manner
using the information of the structure before modification.
In an anti-optimization problem, the solution often exists at the boundary of the
feasible region of the uncertain parameters, if the objective and constraint functions
have monotonicity and/or convexity properties; see, e.g., Sec. 9.6. In this case, the
anti-optimal solution can be found by vertex enumeration of the feasible region, and
the methods of exact reanalysis can be effectively used.
There are several approaches to the approximate or exact reanalysis of responses
of structures subjected to static loads. Melosh and Luik (1968) presented a simple
approach for the reanalysis of trusses. Argyris and Roy (1972) presented a more
general approach allowing changes in the number of degrees of freedom of displace-
ments and support conditions. Kirsch and Rubinstein (1972) discussed convergence
properties of the iterative process of reanalysis. Kirsch (1994) presented an efficient
reanalysis method combining the reduced basis method and the expansion method.
His method has been shown to lead to exact responses of trusses (Kirsch and Liu,
1995), and is applicable to any finite dimensional structures. Application to shape
modification and topological changes has also been demonstrated (Kirsch and Liu,
1996). Kirsch (2000) extended his method to the combined reanalysis for nonlinear
static analysis and eigenvalue analysis.
The virtual distortion method has been developed for static and dynamic analy-
ses of trusses (Putresza and Kolakowski, 2001). Deng and Ghosn (2001) developed
a pseudo force method for nonlinear analysis and reanalysis. Bickford (1987) pre-
sented a refined higher-order perturbation method for eigenvalues and eigenvectors
of a generalized eigenvalue problem. Chen, Yang and Lian (2000) compared several
reanalysis methods for eigenvalue analysis. Chen, Wu and Yang (2006) presented
a reanalysis method for eigenvalue analysis allowing large parameter modification.
Makode, Corotis and Ramirez (1999) presented a pseudodistortion method for static
nonlinear analysis of trusses.
From a mathematical point of view, structural reanalysis can be regarded as
a process of finding the inverse of the stiffness matrix of the modified structure.
Calculation of the inverse of a modified matrix has been discussed in many fields
of engineering and mathematics. Developments in general form of inverting modi-
fied matrices are summarized in the review article by Henderson and Searle (1981).
However, it is important to note, for application to the static analysis problem of
structures, that the inverse of the stiffness matrix of the initial structure is not
usually known; i.e., the matrix is only decomposed to a product of triangular and
diagonal matrices in the process of structural analysis. Kavlie, Graham and Powell
(1971) developed a stiffness-based method for computing the displacements of the
modified design based on the ShermanMorrisonWoodbury (SMW) formula (Sher-
man and Morrison, 1950; Woodbury, 1950). Akg un, Garcelon and Haftka (2001)
extended the SMW formula to the nonlinear reanalysis problem.
Ohsaki (2001) showed that the displacements of the modified structures are
found without computation of the inverse of the stiffness matrix of the initial system;
i.e., only the Cholesky decomposition is necessary. We present below a method of
exact reanalysis for the static responses of trusses.
3.7.2 Mathematical formulation based on the inverse of the modi-

fied matrix
The equations for exact reanalysis is first formulated as a purely mathematical
process of solving a set of linear equations with respect to a modified matrix.
Consider a truss subjected to static nodal loads P . Let n denote the number of
degrees of freedom, and K denote the nn stiffness matrix. The nodal displacement
vector U 0 of the initial (reference) system before modification is computed from
KU 0 = P (3.105)
The matrix K is a function of the vector A = (A1 , . . . , Am )> of the cross-sectional

areas, where m is the number of members. Since the components of K of a truss
are linear functions of Ai , K is written as

m
X
K= Ai K i (3.106)
i=1
where K i is the nn stiffness matrix for unit cross-sectional area of the ith member.
Note that the member properties are defined by the matrices and vectors of the size
n for the convenience of presenting the matrix operation. Computation is carried
out, however, by element-size matrices and vectors.
Consider the case where the cross-sectional area of the kth member is increased
by Ak . Equation (3.105) for the modified truss is formulated as
(K + Ak K k )U 1 = P (3.107)
The objective here is to find the displacement vector U 1 of the modified truss for
the given load vector P . In the following, the superscripts ( )0 and ( )1 denote the
values of the initial and the modified structures, respectively.
There have been a variety of studies on computing the inverse of the modified
matrix from the known inverse of the initial matrix. For our purpose, however,
those formulas cannot be directly used, because the inverse of K is not usually
known; i.e., K is only decomposed for computing U 0 . Therefore, a more explicit
formula is needed for computing U 1 of the modified truss.
Let N = (N1 , . . . , Nm )> denote the vector of axial forces. The relation between
Ni and U 0 is written by using an n-vector ci as
0
0
Ni0 = Ai c>
i U (3.108)
Let E and Li denote Youngs modulus and the length of the ith member, and define
vector bi as
Li
bi = ci (3.109)
E
Then the relation between Ni0 and the n-vector of equivalent nodal forces F 0i of the
ith member is written as
F 0i = Ni0 bi (3.110)
Hence, bi represents the nodal load vector corresponding to unit axial force of the
ith member. The matrix K i is defined by using ci and bi as
K i = b i c>
i (3.111)
Note from Eqs. (3.109) and (3.111) that the rank of K k is 1. The inverse of the
modified matrix K + Ak K k , when rank K k = 1, is obtained from (Henderson
and Searle, 1981)
1 1
(K + Ak K k ) = K + Ak bk c> k
Ak K 1 bk c> 1 (3.112)
kK
= K 1 > 1
1 + Ak ck K bk
By post-multiplying P to Eq. (3.112), the following relation is derived:

0
Ak K 1 bk c>
kU
U1 = U0 1 (3.113)
1 + Ak c>
kK bk
where Eqs. (3.105) and (3.107) are used.
Let superscript k denote a value corresponding to the initial truss subjected to
the load vector bk that generates the unit axial force in member k. Then Eq. (3.113)
is rewritten as
Ak U k c>
kU
0
U1 = U0 k
1 + Ak c>
kU (3.114)
Ak E0k
= U0
1 + Ak Ek
k
where Eq. (3.108) is used, and k is the strain of member k. Note that U k and k k
can be computed through simple arithmetic operation, because K has been already
decomposed in the process of computing U 0 . This way, U 1 can be obtained without
using K 1 . The strain of the member with null cross-sectional area may also be
obtained from the displacements of the nodes connected to the member. Therefore,
Eq. (3.114) can be used for addition or removal of a member between two existing
nodes.
Next, we consider a case where the cross-sectional areas of q ( 2) members
are modified simultaneously. It is assumed without loss of generality that Ak (k =
1, . . . , q) are modified. The stiffness matrix of the modified truss is formulated using
Eqs. (3.109) and (3.111) as
q
X q
X Ak E
K+ Ak K k = K + bk b>
k
Lk (3.115)
k=1 k=1
= K + BDB >
where the ith column of the n q matrix B is bi , and D is a q q diagonal matrix
defied as
D = diag(A1 E/L1 , . . . , Aq E/Lq ) (3.116)
The inverse of the modified stiffness matrix is written as (Henderson and Searle,
1981; Sherman and Morrison, 1950; Woodbury, 1950)
(K + BDB > )1 = K 1 K 1 B(D 1 + B > K 1 B)1 B > K 1 (3.117)
By post-multiplying P to Eq. (3.117), the following relation is derived:
U 1 = U 0 K 1 B(D 1 + B > K 1 B)1 DB > U 0
(3.118)
= U 0 U (D1 + B > K 1 B)1 y
where the ith component of the q-vector y is 0i Li , and the ith column of the n q
matrix U is U i .
A temporary variable vector z is introduced as

z = (D 1 + B > K 1 B)1 y (3.119)
which is found by solving a set of q linear equations
(D 1 + B > K 1 B)z = y (3.120)
which is simply rewritten as
(D 1 + Y )z = y (3.121)

where the (i, j)-component of Y is equal to ij Lj . Therefore, the inverse matrix
1
K need not be explicitly computed also for this case. Finally, Eqs. (3.119) and
(3.121) lead to
U 1 = U 0 U z (3.122)
It may be easily seen that Eq. (3.114) for the modification of a single variable is a
special case of Eqs. (3.121) and (3.122) for multiple variable modification.
As an illustrative example, consider the 2-bar structure in Fig. 3.5 in Sec. 3.6.2.
Let P1 = 0, P2 = 1, E = 1, and L = 1, for simplicity. The matrices and vectors for
the design A1 = A2 = 1 are derived as

2 1 1 1 1
K= , K 1 = , U0 = (3.123)
1 1 1 2 2
Suppose A1 is increased to 2; i.e., A1 = 1. Then we have

1 0 3 1
K1 = , K + A1 K 1 = ,
0 0 1 1

1 1 1 1
(K + A1 K 1 ) = , (3.124)
2 1 3

1 1 1 1
b1 = , c1 = , U1 =
0 0 2 3
From Eq. (3.112), we obtain

1 1 1 1 1 1 0 1 1
(K + A1 K 1 )1 =
1 2 1+1 1 2 0 0 1 2

1 1 1 1 1
= (3.125)
1 2 2 1 1

1 1 1
=
2 1 3
which agrees with the result in Eq. (3.124). Furthermore, we obtain

1 1 1 1 1 1
U 1 = , 1
1 = 1, U 1
= = (3.126)
1 2 2 1 2 3
which also agrees with the result in Eq. (3.124).
Ak , Nkk
1 1
Ak , Nk
Fig. 3.8 Initial truss subjected to virtual

k
Fig. 3.7 Initial truss subjected to P . load P .
Ak , Nk + k Nkk Ak+ Ak , Nk + k Nkk+ k

k k
Fig. 3.9 Initial truss subjected to P and Fig. 3.10 Modified truss without virtual
k
virtual load k P . load.
3.7.3 Mechanical formulation based on virtual load

We next investigate the reanalysis process through mechanical formulations using
the virtual loads due to modification of the cross-sectional area of a member.
Consider again the case where the cross-sectional area Ak of member k of a
truss is to be modified. The initial truss subjected to the nodal load vector P is
illustrated in Fig. 3.7 for a 7-bar truss. Figure 3.8 shows the state where the load
k
vector P representing the unit virtual axial force of member k is applied to the
k k k k
truss. The axial force vector against P is denoted by N = (N 1 , . . . , N m )> .
Suppose the axial force of member k under P increases as the result of modifi-
cation of Ak to Ak + Ak . The axial force of member i of the initial truss under the
k
sum of P and the additional load k P corresponding to the incremental virtual
axial force k of member k is given as
k
Nik = Ni + k N i (3.127)
k
which is illustrated in Fig. 3.9. The strain ki corresponding to Nik + k N i is given
as
Li k
ki = (Ni + k N i ) (3.128)
Ai E
k
Note that the virtual load k P does not actually exist, and it should be supplied
as the result of an increment of the axial force due to the modification Ak of the
cross-sectional area of member k. Therefore, the deformation against the set of
k
loads P and k P shown in Fig. 3.9 should be the same as that of the modified
truss without the additional virtual loads. Hence, the pair of loads are transformed
back to the axial force k as shown in Fig. 3.10, and we obtain the following relation
from the equivalence of the strain of member k:
Lk k Lk k
(Nk + k N k ) = (Nk + k N k + k ) (3.129)
Ak E (Ak + Ak )E
which results in
Ak Nk
k = k
(3.130)
Ak N k Ak
k
Since N k is the axial force of member k against a pair of loads corresponding
to the unit axial force in member k, the following relation holds for a statically
indeterminate truss:
k
1 < N k < 0 (3.131)
and the denominator of Eq. (3.130) is positive for Ak Ak ; i.e., the value
of k for Ak + Ak 0 that includes removal of member k is obtained from
Eq. (3.130). Accordingly, the displacements, axial forces, strains, etc., can be com-
k
puted by applying the loads P + k P to the initial truss with cross-sectional areas
A. It is easily observed that k in Eq. (3.128) is equal to the coefficient for U k
in Eq. (3.118) derived from the formula of the inverse of a modified matrix. Hence,
the mathematical and mechanical formulations lead to the same result.
k
For a statically determinate truss, N k = 1 and the strain of member k of the
modified truss is obtained from Eq. (3.128) as
1 k
kk = (Nk + k N k )
Ak E
1
= (Nk k ) (3.132)
Ak E
Nk
=
(Ak + Ak )E
which corresponds to the fact that the axial force is independent of the cross-
sectional areas.
As an illustrative example, consider again the 2-bar structure in Fig. 3.5 in
Sec. 3.6.2, with the same parameter values in the example in Sec. 3.7.2. The cross-
sectional areas of the initial truss is A1 = A2 = 1, and A1 is increased to 2. Then
we have

1 1 1 1 1
P = , N1 = N2 = 1, N 1 = 1, 1 = = (3.133)
0 1+1 2
and the strains of the modified truss are obtained from Eq. (3.128) as
1 1
11 = 1 = , 12 = 1 (3.134)
2 2
which is consistent with the displacements in Eq. (3.125) obtained by the mathe-
matical formulation.
In closing this section, it should be noted that the exact reanalysis method can
be very effectively used for enumerating the responses of the trusses corresponding
to the vertices of feasible regions of the uncertain cross-sectional areas. Direct
application of the reanalysis method to truss topology optimization can be found
in Ohsaki (2001).
Chapter 4
Anti-Optimization in Static Problems
Hope for the best but prepare for the worst. (English proverb)
If you think the worst, you wont be far wrong. (Spanish proverb)
The worst is not always certain but its very likely. (French proverb)
If anything can go wrong, it will, and at the worst possible time.
(Popular version of Murphys law)
In this chapter, we first present a simple static anti-optimization problem so that the
analytically minded researcher may follow it with pen, pencil, and of course with the
head. Later we explore the apparently first static anti-optimization study by Boley
(1966a) devoted to the worst temperature distribution through the thickness of the
plate, and we present several anti-optimization problems including the evaluation
of the worst possible hole that produces the maximum stress concentration factor.
The worst distribution of prestress of a tensegrity structure is found by vertex
enumeration of the feasible region of the uncertain parameters.
4.1 A Simple Example

Consider a 2-bar truss as shown in Fig. 4.1, where H = 1, W1 = 3, W2 = 1. The
cross-sectional area and Youngs modulus of the members are denoted by A and E,
respectively. The nodal loads and displacements in (x, y)-directions are denoted by
P = (P1 , P2 )> and U = (U1 , U2 )> , respectively, which are related by
1
U1 = [(4 + 2)P1 + (4 + 6)P2 ]
AE(2 + 3)
(4.1)
1
U2 = [(4 + 6)P1 + (4 + 3 2)P2 ]
AE(2 + 3)
which are simply written as
U1 = c > >
1 P , U2 = c2 P (4.2)
77
P2 , U2
P1 , U1
H
x
W1 W2
y
Fig. 4.1 An asymmetric 2-bar truss (W1 6= W2 in general).
with

1 4 + 2 1 4 + 6
c1 = , c2 = (4.3)
AE(2 + 3) 4 + 6 AE(2 + 3) 4 + 3 2
We assume uncertainty in P as
P = P0 + a (4.4)
0
where P = (P10 , P20 )> >
is the nominal load vector and a = (a1 , a2 ) represents the
uncertainty.
Suppose, for simplicity, U2 is chosen as the performance measure, and our pur-
pose is to find the worst value of a that maximizes U2 , which is assumed to be
positive. Consider the case where the bound of a is given by the quadratic inequal-
ity constraint
a> a D (4.5)
as the simplest ellipsoidal model presented in Sec. 3.4, where D is a prescribed
value. Then the anti-optimization problem is formulated in the following form of a
quadratic programming problem:
maximize U2 (a) (4.6a)
>
subject to a a D (4.6b)
It is seen from Eqs. (4.2) and (4.4) that U2 is a linear function of a. Therefore, the
inequality constraint (4.6b) is satisfied with equality at the anti-optimal solution,
which is found explicitly as follows.
The Lagrangian Q of Problem (4.6) is defined as
Q(a, ) = U2 (a) + (D a> a) (4.7)
where 0 is the Lagrange multiplier. The stationary condition of Q with respect
to a leads to
U2
2ai = 0, (i = 1, 2) (4.8)
ai
from which we obtain
1
a= c2 (4.9)
2
Anti-Optimization in Static Problems 79
a2
Anti-optimal solution
1 1 a1
Fig. 4.2 Illustration of anti-optimal solution by quadratic programming.
where Eqs. (4.2) and (4.4) have been used. Since the constraint (4.6b) is satisfied
with equality at the anti-optimal solution, the Lagrange multiplier is obtained as
follows by incorporating Eq. (4.9) into Eq. (4.5), and using 0:
r
1 c>
2 c2
= = (4.10)
2 D
Note that = corresponds to the optimum (minimum) value of U2 .
If EA = 1 and D = 1, for simplicity, then c2 = (0.4155, 2.2086)>, and the
anti-optimization problem is formulated as
maximize U2 = 0.4155a1 + 2.2086a2 (4.11a)
subject to a21 + a22 1 (4.11b)
From Eq. (4.11a), a2 is written in terms of U2 and a1 as
a2 = 0.4528U2 + 0.1881a1 (4.12)
The gray area in Fig. 4.2 is the feasible region defined by Eq. (4.11b), and each
dashed line represents the solution with a constant value of U2 . The anti-optimal
solution is found as a feasible solution that attains the maximum intercept of the
a2 -axis. Since = 1.1237 is obtained from Eq. (4.10), the worst load-case is found
from Eq. (4.9) as a = (0.1849, 0.9828)>.
If the bounds are given for a by an interval
aL a a U (4.13)
with the lower bound aL and upper bound aU , the anti-optimization problem turns
out to be a linear programming problem:
maximize U2 (a) (4.14a)
L U
subject to a a a (4.14b)
a2
Anti-optimal solution
(1, 1) (1, 1)
a1
(1, 1) (1, 1)
Fig. 4.3 Illustration of anti-optimal solution by linear programming.
For the case EA = 1, D = 1 and aU L

i = ai = 1 (i = 1, 2), Problem (4.14) is
explicitly written as
maximize U2 = 0.4155a1 + 2.2086a2 (4.15a)
subject to 1 ai 1, (i = 1, 2) (4.15b)
The gray area in Fig. 4.3 is the feasible region defined by Eq. (4.15b), and each
dashed line represents the solution with a constant value of U2 . The anti-optimal
solution is obtained as the intersection of the feasible region and the dashed line with
maximum intercept of the a2 -axis. Therefore, the anti-optimal solution is attained
at the vertex (1, 1) of the feasible region, which can be easily obtained in view of
signs of the coefficients of ai in the linear expansion of the objective function U2 .
4.2 Boleys Pioneering Problem
Boley (1966a,b) pioneered a study on the bounds of the thermoelastic stresses and
deflections in a beam or a plate. Its derivations will be reproduced here briefly as
the first work of anti-optimization.
The normal thermoelastic stress in a beam of arbitrary cross-section, e.g.,
rectangular cross-section as shown in Fig. 4.4, is given by the formula (Boley and
Weiner, 1960)
PT MT y
= ET (y, z) + + (4.16)
A I
where T (y, z) is the temperature, which is a function of the coordinates (y, z) in
the cross-section, PT is the thermal force, MT is the thermal moment, defined,
respectively, as Z Z
PT = ET dA, MT = ET ydA (4.17)
A A
y y
c
x z
c
Fig. 4.4 A beam with rectangular cross-section.
and A is the cross-sectional area, I is the moment of inertia about the z-axis, E
is Youngs modulus, and is the thermal coefficient. Although Eq. (4.16) is valid
whether principal axes are used or not, we will assume that (y, z) is the principal
coordinate system for simplicity.
In his paper, Boley (1966b) poses the following question: What is the maximum
value of that can be caused in a given beam by arbitrary temperature distribu-
tions? He presumes that the temperature T (y, z) is bounded in the interval
Tm T (y, z) TM (4.18)
where Tm and TM are the constants such that TM Tm . The problem is to
determine the constant k such that
|| k(Tm + TM ) (4.19)
for all possible temperature distributions. The coefficient k has a physical sense of
the factor of proportionality between the maximum possible stress and the maxi-
mum temperature excursion Tm + TM .
The following question arises: Why is it useful to know the coefficient k? It
turns out that the temperature excursion can often be straightforwardly estimated,
especially in problems in which the body surface temperature is known. There-
fore, Eq. (4.19) permits the easy estimation of the maximum normal stresses. For
simplicity, we will deal with the beam of the rectangular cross-section as shown in
Fig. 4.4, which was studied by Boley (1966b) in an earlier paper, presented at a con-
ference in 1964. Accounting that the product E is constant, we rewrite Eq. (4.16)
as follows:
Z Z
() 1 1 3 1
= T () + T (0 )d0 + T (0 )0 d0 (4.20)
E 2 1 2 1
where = y/c with c being the half-depth of the beam as shown in Fig. 4.4. For a
special case

T0 , for 0 < < 0 +
T = 2 2 (4.21)
0, elsewhere
the stress, denoted as 0 (; 0 ), is obtained as

0 (; 0 ) 1 + (1 + 30 ), for 0 < < 0 +
= 2 2 2 (4.22)
ET0 0, otherwise
The function 0 (; 0 ) is a Green function or a fundamental solution. Then, for the
arbitrary temperature distribution, one may write
Z
() 1 1
= T () + T (0 )(1 + 30 )d0 (4.23)
E 2 1
with an observation of the fact that () vanishes identically if and only if the
temperature is the linear function of the y-coordinate, i.e., if T = c0 + c1 . We
now realize that in the realistic problems the exact temperature distribution along
the cross-sectional area is not known. In Boleys (1966b) words, an approximate
temperature distribution is available with a known bounded error Te , such that
Tm Te () TM (4.24)
We are interested in finding the maximum stress that Te () can cause. We consider
the case 0 < < 1, without loss of generality. The examination of the integral in
Eq. (4.23) reveals

1

0 for 1 < 0 < 1

3 if < 1
1 3
(1 + 30 ) = 0 for < 0 < 1 (4.25)

3

1
0 for all if 0
0
3
Boley (1966b) observes:
Now, clearly the largest possible positive stress that can occur is that
causes by a temperature which is at each point of the same sign (and
equal to its largest positive or negative value, as the case may be) as the
quantity (1 + 30 ).
Therefore, the expression for Te in the integrand of Eq. (4.23) ought to be

1

Tm for 1 < 0 < 1

3 if < 1
1 3
Te (0 ) = TM for < 0 < 1 (4.26)

3

1
T for all if 0
M 0
3
whereas T () is set equal to Tm in the non-integral term. To find the maximum
negative stress, one has to interchange all TM and (Tm ) in Eq. (4.26). The final
result turns out to be the same in both cases. Thus,

1 1 1
||max 2 + 3 + , for < 1
= 4 3 3 (4.27)
E(Tm + TM ) 1, 1
for 0
3
with a symmetric result for negative values of . It represents the maximum value,
either positive or negative, which can be expected in the stress. A comparison of
Eqs. (4.19) and (4.26) reveals that

1 1 1
2 + 3 + , for < 1
kmax = 4 3 3 (4.28)

1, 1
for 0
3
A less accurate inequality for kmax is

3 , for 1 < || < 1

kmax 4 3
1 (4.29)

1, for 0 ||
3
or even
4
kmax (4.30)
3
The above formulas allow us to estimate the maximum stress that can develop
in a beam of rectangular cross-section under any temperature distribution that
satisfies the inequality (4.18). If the temperature is a positive interval value, such
that
0 T Tmax (4.31)
then
4
|| ETmax (4.32)
3
Boley (1966b) also derived a more refined bound for the case when the tem-
perature distribution is symmetric or antisymmetric with respect to y. For the
symmetric case, T () = T (), and Eq. (4.23) turns out to be
Z 1

= T () + T (0 )d0 (4.33)
E 0
The following formula for the maximum stress is obtained
||max E(Tm + TM ) (4.34)
Thus, in this case
kmax 1 (4.35)
instead of the inequality kmax 4/3 in Eq. (4.30). Clearly, by gaining additional
information on the symmetry in distribution of the temperature, a smaller upper
bound is derived as expected.
The situation is analogous if the information is obtainable on the antisymmetry
of the temperature distribution, T () = T (). Then the stress becomes
Z 1

= T () + 3 T (0 )0 d0 (4.36)
E 0
Instead of Eq. (4.34), we obtain

3
||max ETM 1 + || (4.37)
2
Since in this case Tm = TM , the inequality (4.18) can be rewritten as follows:
TM T () TM , T () = T () (4.38)
Boley (1966b) derives the following final expression for this case:
5
||max ETM (4.39)
2
In order to compare with Eq. (4.27), Eq. (4.39) is rewritten as
5
||max E(2TM ), T () = T () (4.40)
4
Therefore, for the antisymmetric case, the inequality for kmax results in
5
kmax (4.41)
4
for which the upper bound is again smaller than its counterpart 4/3 in Eq. (4.30).
In the second paper, Boley (1966a) derived the expression for kmax for the gen-
eral case of the cross-section. For example, for the solid circular cross-section and
arbitrary temperature distribution

1 11 15 1 1
kmax + + sin1 ' 1.428 (4.42)
2 16 4
For the triangular cross-section one arrives at

27
kmax ' 1.6875, (4.43)
16
whereas for the thin circular tube

2 3
kmax + ' 1.218 (4.44)
3
In his paper, Boley (1966b) also derived the expression for maximum axial and
transverse deflections. It should also be noted that in earlier papers, Boley (1963,
1964) and Wu and Boley (1965) dealt with the upper and lower bounds for the
solution of melting problems.
To the best of our knowledge, the pioneering works of Boley remained largely
unknown and thus unappreciated despite the fact that they appeared in the prime
journals.
4.3 Anti-Optimization Problem for Static Responses
The general formulations presented in Chap. 3 are readily used for simple static
responses such as nodal displacements and member stresses of trusses.
Pantelides and Ganzerli (1997) found anti-optimal solutions of trusses consider-
ing uncertainty in the direction and magnitude of nodal load vector using ellipsoidal
bounds on parameters as
XN 2
ai
1 (4.45)
i=1
gi
where a = (a1 , . . . , aN )> is the vector of uncertain parameters, and the value of
the semi-axis gi is defined by the minimum volume method as shown in Sec. 3.4.
The static response Z(a) that represents nodal displacement or member stress is
linearly estimated from the nominal values a0 = (a01 , . . . , a0N )> with ai = ai a0i
as
XN
Z(a0 )
Z(a0 + a) = Z(a0 ) + ai (4.46)
i=1
ai
using the sensitivity analysis in Sec. 3.6.2 or the perturbation approach in Sec. 4.4.
The worst parameter values are obtained from the stationary conditions of the
Lagrangian of the anti-optimization problem as demonstrated in Sec. 4.1.
The elastic responses such as member stress and nodal displacement are often
chosen as the objective function to be maximized. The compliance defined as the
external work (twice of the strain energy) is also often taken as the global per-
formance measure (see Secs. 9.69.8 for examples of anti-optimization for static re-
sponses considering stress, compliance and homology design, respectively, which are
demonstrated as the lower-level problem of hybrid optimizationanti-optimization).
Cherkaev and Cherkaev (2003) defined principal compliance as the maximum value
of the compliance against the admissible loading scenarios. Barbieri, Cinquini and
Lombardi (1997) used second-order approximation for anti-optimization of trusses.
Other studies on static anti-optimization include papers by Alotto, Molfino and
Molinari (2001), Attoh-Okine (2002, 2004), Babuska, Nobile and Tempone (2005),
Banichuk (1975, 1976), Banichuk and Neittaanm aki (2007), Barbieri, Cinquini and
Lombardi (1997), Ben-Haim (1992, 1999), Ben-Tal and Nemirovski (1998, 2002),
Chen, Lian and Yang (2002), Chen and Ward (1990), Cho and Rhee (2003, 2004),
Corliss, Foley and Keafott (2007), Dzhur, Gordeyev and Shimanovsky (2001), El-
ishakoff, Gana-Shvili and Givoli (1991), Hlav acek (2002a, 2007), Hlavacek, Chle-
boun and Babuska (2004), Kanno and Takewaki (2006b,c, 2007a,b), K oyl
uo
glu,
C akmak and Nielsen (1995), Kulpa, Pownuk and Skalna (1998), Liu, Chen and
Han (1994), Lombardi (1995, 1998), Lombardi and Haftka (1998), Lombardi, Mar-
iani and Venini (1998), McWilliam (2001), Mullen and Muhanna (1999), Muhanna
and Mullen (2001, 2005), Neumaier and Pownuk (2007a), Popova, Datcheva, Iankov
and Schanz (2003, 2004), Qiu (2003), Qiu, Chen and Song (1996c), Qiu, Chen and
Wang (2004a), Qiu, Ma and Wang (2004d), Qiu, Wang and Chen (2006b), and
Skalna (2003, 2006).
Anti-optimization in the context of comparison of alternative structural models
was proposed by Haftka and his co-workers; see, e.g., the paper by Gangadharan,
Nikolaidis, Lee and Haftka (2003) and Lee, Haftka, Griffin, Watson and Sensmeier
(1994).
4.4 Matrix Perturbation Methods for Static Problems
In this section, we follow Qiu, Chen and Song (1996c) for the perturbation method
to solve the static problem in the matrix, or finite element setting
KU = P (4.47)
where K is the stiffness matrix, P is the load vector, and U is the displacement vec-
tor. The perturbation theory deals with the behavior of a structure that is subjected
to a small perturbation in its variables. The perturbation may occur in the stiffness
matrix as K + K and/or in the load vector, which takes the form P + P . We
are interested in the change of behavior of the system under perturbations. The
perturbation terms K and P will represent for us the ever-present uncertain-
ties. We suppose that the solution U = K 1 P of the unperturbed structure are
known. Note that Eq. (4.47) is solved without computing K 1 ; however, the form
U = K 1 P is used for the expression purpose. Note that the sensitivity analysis
in Sec. 3.6.2 is equivalent to the first-order perturbation.
The solution in the presence of perturbation is denoted as U new . Equation (4.47)
is replaced by
(K + K)U new = P + P (4.48)
For sufficiently small , U new is sought as a series
U new = U + U = U + U (1) + 2 U (2) + + n U (n) + (4.49)
Equation (4.48) becomes
(K + K)(U + U ) = P + P (4.50)
Grouping terms in powers of , we obtain the following set of equations:
KU (1) = P KU (4.51a)
KU (2) = KU (1) (4.51b)
..
.
KU (n) = KU (n1) (4.51c)
We derive the terms U (j) of series (4.49) as follows:

U (1) = K 1 (KU P ) (4.52a)
1
U (2) = K KU (1) (4.52b)
..
.
U (n) = K 1 KU (n1) (4.52c)
1
If the norm of the matrix K K is less than unity, or, more rigorously, if and
only if the spectral radius of K 1 K is less than unity, Eq. (4.49) represents a
convergent series. This result is considered by Deif (1986, 1991).
We now proceed in studying Eq. (4.47) subject to inequalities
KL K KU, P L P P U (4.53)
or componentwise,
L U
Kij Kij Kij , PjL Pj PjU (4.54)
L U L
where P and P are the lower and upper bounds for P , and K is the stiff-
ness matrix composed of the lower bounds of components Kij , whereas K U is its
counterpart composed of the upper bounds of components. The uncertainty is rep-
resented by Eq. (4.54), indicating that we possess practical information, in the form
of bounds alone. Equation (4.53) can be replaced by the statement
K K I, P P I (4.55)
where K = [K , K ] is a real symmetric interval matrix and P = [P , P U ] is
I L U I L
an interval vector. Thus, Eq. (4.47) is replaced by

KIU = P I (4.56)
Equation (4.55) can be rewritten in the following form:
K c K K K c + K, P c P P P c + P (4.57)
where K c and K contain the elements
c 1 L U 1 U L
Kij = (Kij + Kij ), Kij = (Kij Kij ) (4.58)
2 2
with similar expressions valid for P c and P . We assume hereafter that every
matrix contained in K I is nonsingular.
We employ the central interval notation, following Alefeld and Herzberger
(1983), Deif (1991), and Moore (1979):
(K c + K I )U = P c + P I (4.59)
where
K I = [K, K], P I = [P , P ] (4.60)
It is remarkable that Eq. (4.59) is analogous to Eq. (4.50) in its form. Qiu, Chen
and Song (1996c) pose the following question: If each element Kij of K, and each
component Pi of P , are allowed to exhibit perturbation around mid-points (denoted

c
by Kij and Pic ) within certain bounds, that is
c c
Kij Kij Kij Kij + Kij , Pic Pi Pi Pic + Pi (4.61)
then what would the expected variable in U be? The answer to this inquiring is
provided by Eqs. (4.52ac); namely,
U I(1) = (K c )1 (K I U c P I ) (4.62a)
U I(2) = (K c )1 K I U I(1) (4.62b)
..
.
U I(n) = (K c )1 K I U I(n1) (4.62c)
Hence,
U L = U L(1) + 2 U L(2) + + n U L(n) + ,
(4.63)
U U = U U 2 U n U
(1) + U (2) + + U (n) +
From Eq. (4.49), we deduce that

U I = [U L , U U ] = U c + U I (4.64)
with the following notations
KcU c = P c, (4.65a)
L c U c
U = U U , U = U + U (4.65b)
Qiu, Chen and Song (1996c) considered a 5-bar truss as shown in Fig. 4.5 with
two pin-supports, i.e., the total number of degrees of freedom is 4. Youngs modulus
2
and the length are fixed at E = 2.1 1011 N/m and L = 1.0 m, respectively. The
cross-sectional areas of the bars 14 are considered to have a deterministic value
1.0 102 m2 . The cross-sectional areas of diagonal bars 5 and 6 are interval
variables AI = [1.0 102 , 1.02 102 ] m2 . Likewise, the external load parameter
P is treated as an interval variable P I = [2.0998 104 , 2.1002 104 ] N. Then the
interval load vector is given as

[2.0988, 2.1002]
[4.1996, 4.2004]
PI =
[3.1403, 3.1497] 10
4
(4.66)
[5.2495, 5.2505]
where the displacement numbers are indicated in Fig. 4.5. The interval stiffness
matrix reads

[4.26, 4.27] [1.03, 1.01] [3.5, 3.5] [0, 0]
[1.03, 1.01] [3.97, 4.0] [0, 0] [0, 0]
KI = [3.5, 3.5]
108 (4.67)
[0, 0] [4.26, 4.27] [1.01, 1.03]
[0, 0] [0, 0] [1.01, 1.03] [3.97, 3.97]
2P
1 4
P 2 (2) 3
1.5P
(1) (5) (6) (3)
0.8 L
y
x
(4)
0.6 L
Fig. 4.5 A 5-bar truss.
The interval displacements after a single perturbation turn out to be
L U
U(1),1 = 0.00087, U(1),1 = 0.00093,
L U
U(1),2 = 0.00032, U(1),2 = 0.00034,
L U (4.68)
U(1),3 = 0.00091, U(1),3 = 0.00097,
L U
U(1),4 = 0.00033, U(1),4 = 0.00031
L U
where U(1),i and U(1),i are the lower and upper bounds for the first-order term
of Ui , respectively. As is seen, the uncertainty in interval displacements is very
c L U
small. Indeed, the central value U(1),1 = (U(1),1 + U(1),1 )/2 = 0.00090, whereas
U L
the uncertainty parameter U(1),1 = (U(1),1 U(1),1 )/2 equals 0.00003, with the
c
deviation parameter U(1),1 /U(1),1 being 0.03.
Other examples of interval analysis of structures were studied by Chen, Lian
and Yang (2002), Dimarogonas (1995), K oyl
uo
glu, C akmak and Nielsen (1995),
Kulpa, Pownuk and Skalna (1998), McWilliam (2001), Mullen and Muhanna (1999),
Neumaier and Pownuk (2007a, 2007b), Qiu, Chen and Song (1996c), Qiu and Gu
(1996), Qiu, Chen and Wang (2004a), Qiu, Ma and Wang (2004d), Qiu, Wang and
Chen (2006b), Rao and Berke (1997), and Wang and Li (2003). Ellipsoidal modeling
was employed by Kanno and Takewaki (2006a), Rao and Majumder (2008), Qiu,
Ma and Wang (2006a), Qiu, M uller and Frommer (2001), and Qiu, Wang and Chen
(2006b). Comparison of static response estimation by interval and ellipsoidal models
was conducted by Qiu (2003).
4.5 Stress Concentration at a Nearly Circular Hole with Uncertain

Irregularities
4.5.1 Introduction
Stress concentrations in the vicinity of holes of various profiles have been treated
in a number of publications. For a pertinent bibliography one may consult Neuber
(1958), Peterson (1974), and Savin (1961). Traditionally, the stress concentration
factors (SCF) around holes of ideal shapes have been considered. However, in prac-
tice the manufacturing process of structural parts inevitably produces imperfections
which may affect the actual value of the SCF. Irregularities in ideally shaped holes
have been modeled using a probabilistic description by Palmov (1964). The profile
of the hole as a random function was also treated by Sayles (1982). Lomakin and
Sheinin (1970, 1974) obtained stress concentration at a boundary of a randomly
inhomogeneous elastic body.
Lomakin (1968) developed a double perturbation scheme for a hole with a rapidly
oscillating boundary. The boundary was characterized as a random stationary Gaus-
sian function with zero mean and with a specified autocorrelation function. The
stress concentration factor was determined, as suggested by Palmov (1964), from
the condition that < E( ) holds in the region 0 2 with specified
probability. Here is the circumferential stress at the boundary (which turns
out to be a random function too), and E( ) is its mathematical expectation. If
one has a complete probabilistic description of the irregularity of the hole bound-
ary, Lomakins approach is an attractive method to examine its effect on the stress
concentration. Yet in many cases only a limited amount of information on the
irregularities may be available.
In this section, we present the results in Givoli and Elishakoff (1992) for stress
concentration at the boundary of a nearly circular hole in an infinite elastic plane
under uniform radial tension at infinity. We consider a realistic situation, where we
do not possess the full probabilistic information, but only have a partial knowledge
with regard to the shape of the imperfection around the hole boundary. We define
an admissible imperfection profile as one which is bounded in some sense, and which
may be required to have (depending on the boundedness restriction of the profile)
a finite number of harmonics in its Fourier expansion. Then we find the worst-case
possible imperfection profile that gives, among all admissible profiles, the maximum
SCF.
The technique used here to obtain the SCF for a circular hole with arbitrary ir-
regularities is the boundary perturbation method. Some early works employed other
methods to treat holes of arbitrary shapes (Ishida and Tagami, 1959; Averin, 1964);
however, those methods are less amenable for the anti-optimization performed as a
second step.
In Sec. 4.5.2, we find a first-order asymptotic approximation of the SCF at the
boundary of a nearly circular hole. The results are compared with the exact solution
Fig. 4.6 Unit vectors on the ideal circular boundary B and on the imperfect boundary B .
of Tunguskova (1970) for the specific geometry of a hypotrochoid. In Sec. 4.5.3,

we solve the anti-optimization problem and find the worst admissible imperfection
profile and the corresponding SCF. Some results are presented for typical numerical
data.
4.5.2 An asymptotic solution

Consider an infinite isotropic elastic plane with a nearly circular hole. We denote the
boundary of the hole B , as shown in Fig. 4.6, to indicate that it is an -perturbation
from an ideal circle denoted by B. The domains outside B and B are denoted by
and , respectively. The boundary of the hole B is assumed to be traction free.
We describe this boundary in polar coordinates by
r () = R + h() (4.69)
where R is the radius of the circle B, r () is the value of the radial distance r
of a point on B from the center of B, is a small parameter, and h() is a 2-
periodic function. It is assumed that h() possesses a sufficient number of continuous
derivatives, according to the requirements of the asymptotic expansion below, and
to the fact that is small with respect to h/R, h0 /R, etc. Here a prime indicates
differentiation with respect to .
At infinity, a uniform radial tensile stress is prescribed. Radially non-uniform
loading (e.g., uniaxial tension) may also be treated with exactly the same technique
as described below. However, the algebraic manipulations become considerably
more complicated in this case.
The unit vectors er and e for B, and n and t for B , are defined in Fig. 4.6.
The problem to find the stress distribution in is stated as
ij,j = 0 in (4.70a)
rr = , r = 0 at r (4.70b)

Ti = ij nj = 0 on B (4.70c)
with a standard notation and summation convention. In Eq. (4.70c), nj is the jth
component of the unit vector n normal to B and pointing outside .
We shall solve the problem (4.70a)(4.70c) using a perturbation method. A

general treatment of perturbation methods for eigenvalue problems with distorted
boundaries has been given by Morse and Feshbach (1953). Van Dyke (1975) used
the method to solve Laplaces equation outside a slightly distorted circle with
h() = R sin2 in Eq. (4.69). Parnes and Beltzer (1986) used it to solve elas-
ticity problems in circular domains with eccentric loading and in elliptic domains.
We shall use a similar method, but with no restriction on the function h() in
Eq. (4.69); namely, we shall find an asymptotic solution valid for any small per-
turbation from the perfect circular hole. We write the stress ij as a perturbation
series:
(0) (1) (2)
ij = ij + ij + 2 ij + (4.71)
Substituting ij in Eq. (4.71) into Eqs. (4.70a) and (4.70b), and setting the coeffi-
cients of different powers of equal to zero, we conclude that
(k)
ij,j = 0 in , (k = 0, 1, . . . ) (4.72a)
(0) (0)
rr = , r = 0 at r (4.72b)
(k) (k)
rr = r =0 at r , (k = 1, 2, . . . ) (4.72c)
The derivation of asymptotic boundary conditions on B from Eq. (4.70c) is more
complicated. First, the tractions in Eq. (4.70c) are defined with respect to the
normal n to B whereas the asymptotic expressions for the tractions on B have to
be given with respect to the normal er , as shown in Fig. 4.6. The components nr
and n in the directions er and e , respectively, are found to be
(h0 )2
nr = 1 2+ O(3 ) (4.73a)
2R2
h0 h0 h
n = + 2 2 + O(3 ) (4.73b)
R R
where O(n ) denotes the term of order n ; see Givoli and Elishakoff (1992) for
details. This agrees with Lomakins result (Lomakin, 1968). Now Eq. (4.70c) in
polar coordinates reads
Tr = rr nr + r n = 0
on B (4.74)
T = r nr + n = 0
Using Eqs. (4.71), (4.73a) and (4.73b) in Eq. (4.74), we obtain two equations in-
(k) (k)
volving rr (r , ) and r (r , ). However, these equations still cannot be equated
with respect to powers of to zero since r is also a function of . Therefore, we
(k) (k)
first expand ij (r , ) in a Taylor series around ij (R, ):
(k) (k) (k) 1 (k)
ij (r , ) = ij (R, ) + hij,r (R, ) + 2 h2 ij,rr (R, ) + (4.75)
2
After substituting Eq. (4.75) in the two equations obtained by Eq. (4.74), we can
equate with respect to powers of to obtain the desired boundary conditions on the
circle B. The resulting boundary conditions are written as follows up to the second
order:
(0)
rr (R, ) = 0 (4.76a)
0
(1) h (0) (0)
rr (R, ) = r hrr,r (4.76b)
R
(2) h0 (1) (1) 1 (h0 )2 (0)
rr (R, ) = r hrr,r h2 rr,rr
(0)
+
R 2 2R2 rr
hh0 (0) hh0 (0)
+ r,r 2 r (4.76c)
R R
(0)
r (R, ) = 0 (4.76d)
0
(1) h (0) (0)
r (R, ) = hr,r (4.76e)
R
2 h0 (1) (1) 1 (0) (h0 )2 (0)
r (R, ) = hr,r h2 r,rr +
R 2 2R2 r
0 0
hh (0) hh (0)
+ ,r 2 (4.76f)
R R
In Eqs. (4.76a)(4.76f), all the functions on the right-hand-side are evaluated at
r = R.
Note that Eqs. (4.76a), (4.76d) and (4.76e) can be used to simplify Eqs. (4.76b),
(4.76c) and (4.76f) as
(1) (0)
rr (R, ) = hrr,r (4.77a)
0 2
h (0) (1) 1
(2)
rr (R, ) = hrr,r h2 rr,rr
(0)
(4.77b)
R 2
2 h0 (1) (1) 1 (0) hh0 (0) hh0 (0)
r (R, ) = hr,r h2 r,rr + ,r 2 (4.77c)
R 2 R R
Equations (4.72a)(4.72c) and (4.76a)(4.76f) define a sequence of problems in do-
main . In what follows we solve the first two problems in this sequence; namely,
we find a first-order asymptotic approximation. The zeroth-order problem is ax-
isymmetric and is easily solved as

(0) R2
rr = 1 2 (4.78a)
r

(0) R2
= 1 + 2 (4.78b)
r
(0)
r = 0 (4.78c)
Using this solution as well as Eqs. (4.72a), (4.72c), (4.76b) and (4.76e), we conclude
(1)
that ij should satisfy
(1) (1)
rr = r = 0 at r (4.79a)
(1) 2
rr (R, ) = h() (4.79b)
R
(1) 2 0
r (R, ) = h () (4.79c)
R
(1)
The solution for can be written in the form
Z
(1) 2 X 2
= [Brn (r, , 0 )h(0 ) + Bn (r, , 0 )h0 (0 )]d (4.80)
R n=0 0
where Brn and Bn are the kernels; see Givoli and Elishakoff (1992) for details.
Now Eqs. (4.78c) and (4.80) provide us with an approximate solution on the
circular boundary B. We use Eqs. (4.71) and (4.75) again to obtain a solution on
the actual boundary B :
(0) (0) (1)
ij (r , ) = ij (R, ) + [h()ij,r (R, ) + ij (R, )] + O(2 ) (4.81)
which leads to an expression for on B . However, the SCF is defined via the
circumferential stress tt on B rather than via ; see Fig. 4.6 for the difference
between t and e . To find tt , we transform the stresses according to the formula
tt = rr sin2 + cos2 r sin 2 (4.82)
where is the angle between er and n, and cos = nr , sin = n . Then, from
Eqs. (4.73a), (4.73b), (4.78c) and (4.80)(4.82), the stress concentration factor
along B is obtained as
X Z 2
tt (r , )
' =1 h() + [Brn (R, , 0 )h()
2 2 R n=0 0
(4.83)
n 0 0 2
B (R, , )h ()]d + O( )
which leads to
" Z 2 #
1X
= 2 1 g() + n cos n( )g()d (4.84a)
n=0 0
h()
g() = (4.84b)
R
1
0 = , n = 2n 1 for n 1 (4.84c)
2
It is easy to verify that if g() is constant, corresponding to an ideal circle, then
= 2. Moreover, even when g() = cos or g() = sin , then Eq. (4.84a) yields
the constant value = 2. This implies that, for these boundary profiles, the effect
of the deviation from the ideal circle on the SCF does not manifest itself in the
first-order perturbation term.
We now compare our result to that of Tunguskova (1970), who presented an exact
expression for the SCF around a hole in the shape of a hypotrochoid (Lockwood,
1961). In cartesian coordinates, this curve is defined by the parametric equations
x = R(cos t + cos kt), y = R(sin t sin kt) (4.85)

Fig. 4.7 A hypotrochoid with k = 7 and = 0.01.
where k is a non-negative integer and 0 < k < 1. The polar coordinates on this
curve can easily be written as a function of the parameter t, and be expanded in
powers of as
sin(k + 1)t
tan = tan t + O(2 ) (4.86a)
cos2 t
r = R + R cos(k + 1)t + O(2 ) (4.86b)
One also can see that cos(k + 1)t = cos(k + 1) + O(). Comparing Eq. (4.86b) with
Eq. (4.69) and using Eq. (4.84b), we obtain
g() = cos(k + 1) (4.87)
Figure 4.7 describes a hypotrochoid with k = 7 and = 0.01.
Tunguskovas exact stress concentration factor around the hole is
2(1 2 k 2 )
= (4.88)
1 2k cos(k + 1)t + 2 k 2
with a maximum value of
2(1 + k)
max = (4.89)
t 1 k
On the other hand, we obtain as follows from our asymptotic solution (4.84a)
after substituting the function g() given in Eq. (4.87) and using the orthogonality
property of the trigonometric functions:
= 2 [1 + 2k cos(k + 1)] (4.90)
with a maximum value
max = 2(1 + 2k) (4.91)
t
Note that Eqs. (4.90) and (4.91) can be obtained from Eqs. (4.88) and (4.89),
respectively, by using Taylor series expansion up to the first order in .
In Fig. 4.8, we compare the exact solution (4.89) (solid line) and the asymptotic
solution (4.90) (dashed line) for k = 7 and = 0.01. The agreement between the
Fig. 4.8 A comparison between Tunguskovas exact SCF (solid line) and the present asymptotic
SCF (dashed line) for the hypotrochoid of Fig. 4.7.
Fig. 4.9 Relative error in the maximal asymptotic SCF as a function of k.
two solutions is excellent. An easy calculation shows that the relative difference
between the maximum values of and is
max max 2(k)2
= (4.92)
max 1 + k
This relative error as a function of k is plotted in Fig. 4.9. With the choice of
k = 7 and = 0.01, the relative error amounts to only 0.9%. Increasing to
0.035 we have an error of about 10%. This may raise the question of whether
our first-order asymptotic solution (4.84a) is satisfactory. We claim that in many
practical situations, it is satisfactory, since k is very small. We are interested in
small irregularities, and an irregularity characterized by = 0.01 is quite usual in
practice. As for the value of k, we confine ourselves to the case where k is of order
10 at the most. If k is large, say 100, we may use Lomakins asymptotic solution
(Lomakin, 1968), which is valid only for this case of a rapidly oscillating boundary
profile. In this sense, our solution and that of Lomakins are complementary. For the
review and numerous references therein on asymptotic methods of the boundarys

form, the reader can consult the review by Guz and Nemish (1987).
4.5.3 A worst-case investigation

Now we shall define a set of admissible functions g(), namely, a set of admissible
hole profiles. Then we shall find the admissible profile that gives the maximum SCF
at a given point on the boundary of the hole.
It is first assumed that an admissible profile has only a finite number M of
important Fourier components in the angle , and that all higher harmonics can be
neglected. Thus, g() is written as
M
X
g() = (Am cos m + Bm sin m) (4.93)
m=0
We also assume that the irregularities are bounded in the L2 sense, namely,
Z 2
g 2 ()d (4.94)
0
where is a given constant. The set of admissible profiles g() consists of those
functions which are of the form Eq. (4.93) and satisfy Eq. (4.94).
We find g() which maximizes in Eq. (4.84a) at a given angle = 0 , which
will turn out later to be arbitrary. Substituting Eq. (4.93) in Eq. (4.84a) and setting
= 0 result in
XM
(0 )
I' =1+ 2(m 1)(Am cos m0 + Bm sin m0 ) (4.95)
2 m=1
In addition, Eqs. (4.93) and (4.94) give the inequality constraint
M
X
2A20 + (A2m + Bm
2
) (4.96)
m=1
which is written in the form of an equality constraint
M
X
J ' 2A20 + (A2m + Bm
2
) + p2 =0 (4.97)
m=1

where p is considered as a new unknown. Now the problem is to find the coefficients
Am and Bm that will maximize I subject to the constraint J = 0.
This problem is easily solved by the Lagrange multiplier method. The unknowns
are A0 , Am , Bm (m = 1, 2, . . . , M ), p and the Lagrange multiplier . The necessary
conditions for extremum are

(I + J)
= 2(m 1) cos m0 + 2Am = 0, (m = 1, 2, . . . ) (4.98a)
Am
(I + J)
= 2(m 1) sin m0 + 2Bm = 0, (m = 1, 2, . . . ) (4.98b)
Bm
(I + J)
= 2p = 0 (4.98c)
p
(I + J)
= 4A0 = 0 (4.98d)
A0
From Eqs. (4.98a) and (4.98b), we deduce that 6= 0, and therefore from Eq. (4.98c)
p = 0; namely, the constraint (4.96) is active, and A0 = 0. Also, from Eqs. (4.98a)
and (4.98b),
1
Am = (m 1) cos m0 , (m = 1, 2, . . . ) (4.99a)

1
Bm = (m 1) sin m0 , (m = 1, 2, . . . ) (4.99b)

Using these expressions in Eq. (4.97) with p = 0, we obtain
B
2 = M (4.100)

where
M
X
M (M 1)(2M 1)
BM = (m 1)2 = (4.101)
m=2
6
Then, Eqs. (4.99a) and (4.99b) are reduced to
(m 1) cos m0
Am = /)1/2 , (m = 1, 2, . . . ) (4.102a)
(BM
(m 1) sin m0
Bm = /)1/2
, (m = 1, 2, . . . ) (4.102b)
(BM
Note that we chose the negative root of 2 in Eq. (4.100), because in order to obtain
a maximum for I, Am and Bm must have the same signs as cos m0 and sin m0 ,
respectively.
Equations (4.93), (4.101), (4.102a) and (4.102b) define the admissible worst-case
profile. The corresponding maximum stress concentration factor is found from
Eq. (4.95) to be
r !
BM
(0 ) = 2 1 + 2 (4.103)

As expected, the maximum does not depend on the angle 0 . The SCF around
the hole boundary is given by (cf. Eq. (4.95))
" M
#
X
() = 2 1 + 2(m 1)(Am cos m + Bm sin m) (4.104)
m=2
Fig. 4.10 The worst-case profile g() for the parameters M = 10 and = 10 4 .
Fig. 4.11 The worst-case profile g() for the parameters M = 5 and = 10 4 .
It is interesting
R 2 to note that the average of around the boundary of the hole,
(1/2) 0 ()d, is always 2, i.e., it is equal to the SCF for a circle.
Figure 4.10 is the graph of the profile function g() for the parameters M = 10
and = 104 . The angle 0 was chosen to be . Any other choice for 0 would just
shift the plot horizontally. The maximum value of g, namely, the maximum value
of imperfection divided by R, is 0.015 at 0 . Figure 4.11 shows the corresponding
distribution of around the hole boundary. The maximum SCF is 2.38, namely,
19% higher than the SCF associated with the ideal boundary. Figure 4.12 is the
graph of g() for the parameters M = 5 and = 104 . Here, the maximum value
of g is 0.010, and the maximum corresponding SCF is now 2.12, or 6% larger than
that for a circle.
It is instructive to compare the SCF obtained by the worst profile with the one
obtained by a hypotrochoid, as found by Tunguskova (1970). The maximum stress
Fig. 4.12 The distribution of the SCF around the worst-case profile.
concentration factor max

h for a hypotrochoid is given in Eq. (4.91) as max
h = 2+4k.
To compare this with the worst-case stress concentration factor max , we set the
maximum number of harmonics to M = k +1 (cf. Eq. (4.87)). Also, the appropriate
bound in Eq. (4.94) is calculated from Eq. (4.87) to be
Z 2
= g 2 ()d = 2 (4.105)
0
Then, from Eqs. (4.101) and (4.103), we rdeduce that
k(k + 1)(2k + 1)
max = 2 + 4 (4.106)
6
max
Comparing Eq. (4.91) with Eq. (4.106), it is easy to verify that always max
h ,
and that the equality holds only if k = 1; namely, M = 2. This means that among
all the holes whose profiles are described by at most two harmonics with the specified
norm , the hypotrochoid is the least maximum one. Substituting the numerical
values = 0.01 and k = 7, as in Fig. 4.7, we get maxh = 2.30 and max = 2.47. In
other words, the hypotrochoid and the worst profile yield SCFs which are larger by
15% and 23.5%, respectively, than that for a circle.
One may wonder whether it is also possible to find the best-case profile, for
which the stress concentration factor would be minimal. Since the average of
around the nearly circular boundary according to our asymptotic analysis is always
2, there is no profile within the set considered, which gives a SCF smaller than
2 at all the points on the boundary. Therefore, the optimal profile in a general
sense is a circle. However, an interesting (although maybe academic) question is:
What is the smallest that one can achieve within the set of admissible profiles (as
defined above)? The answer is obtained by repeating the analysis while choosing
the positive root of 2 in Eq. (4.100) instead of the negative root. The minimum
is found to be r !
BM
(0 ) = 2 1 2 (4.107)

For M = 10 and = 104 , the minimum SCF is 1.62. The corresponding profile
is the same as the one in Fig. 4.10, but with an opposite sign. The maximum
obtained for this profile is 2.30.
Up to this point, the set of admissible imperfection profiles g() was defined
via the bound constraint (4.94) and the requirement that g() has a finite number
of harmonics. The latter requirement has several deficiencies. First, it is a very
strong requirement which causes many otherwise reasonable profiles to be excluded.
Second, it may be impractical for the designer to determine how many harmonics
are needed in the analysis. Third, and perhaps most important, the SCF result
obtained above (see Eqs. (4.101) and (4.103)) grows without limit as the number
of harmonics M increases. This is expected since Eq. (4.94) does not pose any
restriction on the curvature of the profile, and so the worst-case profile tends to a
saw-type function as M tends to infinity.
To overcome these deficiencies, we now replace Eq. (4.94) by a constraint on the
curvature of the profile, i.e.,
Z 2
2
(g 00 ()) d (4.108)
0
where is a given positive number. This enables us to go again through the deriva-
tion while totally eliminating the requirement that g() has a finite number of
harmonics. The consequence of this derivation for the maximum SCF is found to
be (cf. Eq. (4.103)) r

=2 1+2 (4.109)

where
X
(m 1)2
= = 0.3230 . . . (4.110)
m=2
m4
Thus, with the constraint of the form Eq. (4.108), we are able to sum the series
in Eq. (4.109), whereas the series in Eq. (4.101), which is based on the constraint
(4.94), does not converge as M tends to infinity.
Both constraints (4.94) and (4.108) can be viewed as special cases of the general
constraint Z
2
[0 g 2 + 1 (g 0 )2 + 2 (g 00 )2 + + L (g (L) )2 ]d (4.111)
0
where i are the given real constants. The maximum SCF in this general case is
again given by Eq. (4.109), where the expression (4.110) for is replaced by
X
(m 1)2
= (4.112)
+ 1 m2 + 2 m4 + + L m2L
m=2 0
The sum in Eq. (4.112) converges if and only if at least one of the coefficients i for
i 2 is nonzero. Thus, in order that no requirement on the number of harmonics of
g() would be needed, the second derivative and/or higher derivatives of g() must
be constrained.
The generalization of this section to the orthotropic case was given by Gana-
Shvili (1991).
Fig. 4.13 A tensegrity structure; thick line: strut, thin line: cable.
4.6 Anti-Optimization of Prestresses of Tensegrity Structures
4.6.1 Introduction
A tensegrity structure consists of cables and struts that can transmit only tensile
and compressive forces, respectively, and it is stiffened by prestresses in the self-
equilibrium state. Figure 4.13 illustrates a simple example of a tensegrity structure,
where thick and thin lines represent struts and cables, respectively. The term tenseg-
rity was invented by Buckminster Fuller as a contraction of tension and integrity,
although the inventor of the structural system is under controversy (Fuller, 1975;
Lalvani, 1996). By the rigorous definition, a tensegrity structure is free-standing
without any support, and the struts are connected by continuous cables and do not
have contact with each other (Motro, 1992).
Since the tensegrity structure is unstable in the absence of prestresses, the shape
and the stability at the self-equilibrium state strongly depend on the member forces
(Zhang and Ohsaki 2006; 2007). Therefore, there exist many difficulties in the
design of shape and forces of tensegrity structures, and many methods have been
presented for this process.
Tensegrity structures usually have several modes of self-equilibrium forces that
are governed by a set of linear equations with respect to the member forces, and
the member forces are defined as a linear combination of the self-equilibrium force
modes. Additional requirements exist for the member forces, because the cables
and struts can transmit only tensile and compressive forces, respectively. Further-
more, the upper bounds should be given for the cable forces so that the stress
remains within the yield stress divided by the safety factor. A bound for compres-
sive stress should also be given for a strut to prevent buckling. Since these bound
constraints are expressed as linear inequalities with respect to the coefficients of the
self-equilibrium force modes, the feasible region of the coefficients forms a convex
region.
The performance of the force design defined by the feasible or admissible set of
coefficients should be determined based on the mechanical properties such as the
eigenvalues of the tangent stiffness matrix and the displacements against external
loads. However, the member forces are very sensitive to the imperfection or variation
of the geometrical parameters such as the unstressed lengths of members and nodal
locations. Therefore, the anti-optimal solution of self-equilibrium forces is defined as
the member forces that minimize the performance measures under bound constraints
of the forces.
If those performance measures have monotonicity and/or convexity properties
with respect to the member forces, the anti-optimal solutions can be found by
searching the vertices of the feasible region of the member forces. In this section, we
present an anti-optimization method of the forces of tensegrity structures (Ohsaki,
Zhang and Elishakoff, 2008a). The lowest eigenvalue of the tangent stiffness matrix
is considered as the performance measure (Ohsaki, Zhang and Ohishi, 2008b). The
properties of the objective functions are investigated, and the anti-optimal solution
is found by the enumeration of the vertices of the convex feasible region.
4.6.2 Basic equations

4.6.2.1 Equilibrium equations
The members of the tensegrity structures consist of cables and struts connected by
pin joints. The self-weight is neglected, and no external load exists at the initial
self-equilibrium state.
Let m and ns denote the numbers of members and nodes, respectively. The
m ns connectivity matrix is denoted by C s , and the (k, p)-component of C s is
s
denoted by Ck,p . If member k is connected by nodes i and j (i < j), the kth row of
s
C is defined as

1 for p = i
s
Ck,p = 1 for p = j (4.113)

0 for other cases
The nodes are classified into n free nodes and nf fixed nodes (supports); i.e.,
n = n + nf . Suppose the nodes are numbered such that the free nodes precede the
s
fixed nodes. Then C s is divided into the m n matrix C and the m nf matrix
C f corresponding to the free and fixed nodes, respectively, as
C s = (C, C f ) (4.114)
Let n-vectors x, y, z and nf -vectors xf , y f , z f denote the nodal coordinate

vectors of free and fixed nodes in x-, y- and z-directions, respectively. We assume a
tensegrity structure in the three-dimensional space for the presentation of general
formulations. The coordinate difference vectors of the members are denoted by h x ,
hy and hz for x-, y- and z-directions, respectively, which are calculated by
hx = Cx + C f xf
hy = Cy + C f y f (4.115)
z f f
h = Cz + C z
and the m m coordinate difference matrices H x , H y and H z are defined as

H x = diag(hx )
H y = diag(hy ) (4.116)
z z
H = diag(h )
In the following, the components of vectors and matrices are indicated by subscripts
as hx = (hxi ), H x = (Hij
x
). The length matrix L = (Lij ) is diagonal, and its ith
diagonal term is given as
q
Lii = (Hiix )2 + (Hiiy )2 + (Hiiz )2 , (i = 1, . . . , m) (4.117)
Let s = (s1 , . . . , sm )> denote the vector of member forces. In the state of self-
equilibrium, the equilibrium equation of a pin-jointed structure in three-dimensional
space is written as (Schek, 1974)
Ds = 0 (4.118)
where the 3n m equilibrium matrix D is defined by

C>Hx
D = C > H y L1 (4.119)
C>H z
4.6.2.2 Self-equilibrium forces

Let r denote the rank of D. Then the equilibrium equation (4.118) has q = m r
set of self-equilibrium force modes, which are found, as follows, by the singular value
decomposition of D.
The nonzero eigenvalues of D > D are denoted by i (i = 1, . . . , r). Then the
singular value decomposition of D is written as
D = SR> (4.120)
where

diag(1 , . . . , r ) O
= (4.121)
O O
and the diagonal terms of the 3n m rectangular matrix are called singular
values of D. The m m matrix R and the 3n 3n matrix S satisfy the following
orthogonality conditions:
R> R = RR> = I m
(4.122)
S > S = SS > = I 3n
where I m and I 3n are the m m and 3n 3n identity matrices, respectively.
By postmultiplying R to Eq. (4.120) and using Eq. (4.122), we obtain
DR = S (4.123)
The column vectors Ri (i = r +1, . . . , m) of R corresponding to zero singular values

satisfy the condition of self-equilibrium force mode as
DRi = 0 (4.124)
By denoting g i = Ri+r (i = 1, . . . , q), the self-equilibrium force vector s satis-

fying Ds = 0 is given as the linear combination of g i as
s = 1 g 1 + + q g q
(4.125)
= G
where = (1 , . . . , q )> is the coefficient vector, and G = (g 1 , . . . , g q ) is the matrix

of the self-equilibrium force modes.
Let b>
i denote the ith row of G. The components of s are written as
si = b >
i , (i = 1, . . . , m) (4.126)
The self-equilibrium equation with respect to member forces s in Eq. (4.118) can be
rewritten in the following form with respect to the nodal coordinates (Zhang and
Ohsaki, 2006):
Ex + E f xf = 0
Ey + E f y f = 0 (4.127)
f f
Ez + E z = 0
where E is called a force density matrix that is given as
E = C > diag(L1 s)C

(4.128)
E f = C > diag(L1 s)C f
4.6.2.3 Tangent stiffness matrix

The tangent stiffness matrix K is expressed as the sum of the linear stiffness matrix
K E and the geometrical stiffness matrix K G as
K = KE + KG (4.129)
with
K E = DKD > , K G = I 3 E (4.130)
where I 3 is the 3 3 identity matrix, and the ith diagonal component K ii of the
diagonal matrix K is the stiffness of the ith member; i.e., K ii = Ai Ei /Lii where Ai ,
Ei and Lii are the cross-sectional area, Youngs modulus and the length of member
i, respectively.
4.6.2.4 Lowest eigenvalue of tangent stiffness matrix

In the following discussions on stiffness, the rigid-body motions are assumed to be
constrained, i.e., the structure has 3n-degree-of-freedom with n free nodes. Let r
(1 2 3n ) and r denote the rth eigenvalue and eigenvector of K,
respectively, which are defined by
Kr = r r , (r = 1, . . . , 3n) (4.131)
The eigenvector r is ortho-normalized by
>
r s = rs , (r, s = 1, . . . , 3n) (4.132)
where rs is the Kronecker delta. The lowest eigenvalue 1 after constraining
the rigid-body motions is minimized as the performance measure in the anti-
optimization problem defined in Sec. 4.6.3.
By differentiation of Eq. (4.131) for r = 1 with respect to i , and using
Eq. (4.132), we obtain
1 K
= >1 1 (4.133)
i i
The details of eigenvalue sensitivity analysis can be found in Sec. 3.6.3. Incorpora-
tion of Eq. (4.129) into Eq. (4.133) leads to
1 (K E + K G )
= > 1 1 (4.134)
i i
Since K E is independent of i , we obtain
1 K G
= >1 1 (4.135)
i i
Furthermore, since K G is a linear function of , as is seen from Eqs. (4.127) and
(4.130), K G /i is a constant matrix. Therefore, if 1 is almost constant in the
feasible region, then 1 is a monotonic function of .
Concavity of 1 with respect to at a self-equilibrium state can be shown in a
similar manner as the proof for the lowest eigenvalue of tangent stiffness matrix; cf.
Sec. 5.6.
4.6.2.5 Compliance against external load

Suppose small external loads P = (P1 , . . . , P3n )> are applied to the structure. The
nodal displacements U = (U1 , . . . , U3n )> are linearly estimated by the tangent
stiffness matrix as
KU = P (4.136)
In the field of structural optimization, the external work, which is called compliance
is often used as the performance measure. The compliance W , defined as follows,
is to be minimized to obtain the stiffest design against the specified loads:
W = U >P
(4.137)
= U > KU
which is maximized in the anti-optimization problem in Sec. 4.6.3.

By differentiating Eq. (4.136) with respect to i , the sensitivity coefficient of U
with respect to i is written as
U
= K 1 K i U (4.138)
i
The details of sensitivity analysis of static responses can be found in Sec. 3.6.2.
Using the form W = U > KU in Eq. (4.137), the sensitivity coefficient of the com-
pliance is given as
W K U
= U> U + 2U > K
i i i
(4.139)
K
= U > U
i
Since K E is independent of i , we obtain
W K G
= U > U (4.140)
i i
Therefore, if U is almost constant in the feasible region, then W is a monotonic
function of .
Convexity of compliance can be proved in several manner (Prager, 1972; Svan-
berg, 1984, 1994). Stolpe and Svanberg (2001) presented the following simple proof.
Let G(, U ) be defined as

1
G(, U ) = 2 P > U U > K()U (4.141)
2
Then, for every ,
W () = max G(, U ) (4.142)
U
Suppose the extensional stiffness of member i is a concave function (including linear

function) of , which is valid for the tensegrity structures. Then, for each fixed U ,
G(, U ) is convex in . Since a function defined as a pointwise maximum of a
collection of convex functions is convex (Boyd and Vandenberghe, 2004), W () is
a convex function.
4.6.3 Anti-optimization problem

Consider uncertainty in member forces at the self-equilibrium state due to the er-
rors in unstressed (initial) lengths of members or change in member lengths after
construction resulting from relaxation of high-tensioned cables. Since the member
forces s should satisfy the self-equilibrium equation (4.118), the errors of s can-
not be distributed independently. Note that the change in member forces leads to
change in nodal locations, which is assumed to be very small in the following; i.e.,
the equilibrium matrix D is fixed. Therefore, the vectors g i of the self-equilibrium
force modes are fixed, and the variation of the member forces is investigated in the
space of the coefficients .
We first design the nominal value s0 = G0 of the self-equilibrium forces. The
lower and upper bounds sL and sU , respectively, are appropriately given for s0 to
assign the constraints as
sL s 0 s U (4.143)
where the signs of the forces for cables and struts are to be considered; i.e., the
lower bounds for the cables and the upper bounds for the struts are 0.
The vertices of the feasible region satisfying the constraint (4.143) are enumer-
ated in the space of . Denote by 0 the mean value of at all the vertices called
the center of the feasible region (Ohsaki, Zhang and Ohishi, 2008b), for brevity,
which is conceived as the nominal value of the uncertain coefficient vector for the
self-equilibrium force modes. Note that the nominal value need not be the center
of the feasible region, and can be defined arbitrarily, e.g., based on the preference
of the designer.
Next we assign the range of uncertainty of s as an interval
s0 s s s0 + s (4.144)
where si is the maximum possible increase or decrease of si . Note that the range
is given for the force vector s, although the independent parameters for the forces
are .
In the following anti-optimization problem, the deviation of the forces from the
desired values is chosen as one of the performance measures. Then the deviation e
of from the desired value 0 is defined as
e = ( 0 )> G> G( 0 ) (4.145)
which is a convex function of .
We consider the following three performance measures so that the anti-
optimization problem is formulated as minimization of a concave function, or max-
imization of a convex function, and the anti-optimal solution exists at a vertex of
the feasible region defined by linear inequalities:
(1) Lowest eigenvalue 1 of K after constraining the rigid-body motions, which is a
concave function of , and is to be minimized in the anti-optimization problem.
(2) External work called compliance against the specified load after constraining
the rigid-body motions, which is a convex function of , and is to be maximized
in the anti-optimization problem.
(3) Deviation e of the forces from the desired value, which is a convex function of
, and is to be maximized in the anti-optimization problem.
For example, the anti-optimization problem of minimizing 1 is formulated as
minimize 1 () (4.146a)
subject to s0 s s s0 + s (4.146b)
s = G (4.146c)
Other problems for maximizing W and e, respectively, are formulated similarly.

Then the anti-optimal solution is found by vertex enumeration of the feasible region
as follows:
Step 1 Specify the geometry, topology, and material property of the tensegrity
structure.
Step 2 Construct the equilibrium matrix D from Eq. (4.119) and compute the self-
equilibrium force modes g 1 , . . . , g q by the singular value decomposition (4.123).
Step 3 Assign bound constraints (4.143) for the nominal value of the self-
equilibrium forces.
Step 4 Generate the list of the vertices of the feasible region of satisfying (4.143).
Step 5 Find the center 0 of the feasible region that is regarded as the nominal
value of .
Step 6 Specify s, and generate the list of the vertices of the feasible region of
for the anti-optimization problem satisfying (4.144).
Step 7 Assign the support conditions, and compute the tangent stiffness matrix
K from Eq. (4.129) at each vertex.
Step 8 Compute the lowest eigenvalue 1 from Eq. (4.131), the compliance W from
Eq. (4.137), and the force deviation e from Eq. (4.145) at each vertex of the
feasible region satisfying Eq. (4.144), then find the three vertices, each of which
corresponds to the anti-optimal solution that minimizes 1 , maximizes W , or
maximizes e.
In the following numerical example, the vertices of the feasible region are enu-
merated by cdd+ (Fukuda, 1999; Avis and Fukuda, 1996) based on the efficient
procedure called reverse search. cdd+ can enumerate the vertices and the asso-
ciated active constraints of the region defined by linear inequality and equality
constraints.
4.6.4 Numerical examples

Consider a tensegrity grid as shown in Fig. 4.14 that consists of the unit cell as shown
in Fig. 4.15, where the thick and thin lines are struts and cables, respectively, and
the node numbers are shown in Fig. 4.14(a). This grid was first presented by Motro
(2003). The tensegrity grid is constructed by consecutively assembling the unit cell
in x- and y-directions.
Let t and c denote the numbers of rows (x-direction) and columns (y-direction)
of the struts, respectively; i.e., there are t + 1 struts in each column and c + 1 struts
in each row. Therefore, the structure has 2tc + t + c struts and ns = 2(tc + t + c)
nodes. The total number of members is m = 7tc + 5t + 5c 4. The rank of K E
after constraining rigid-body motions is 1; i.e., this kind of structure has only one
infinitesimal mechanism irrespective of the values of t and c.
The structure in Fig. 4.14 has three and four struts in x- and y-directions,
22
400
5 10 9
300
20 17 18
14
21
4
3 y 200
7
8
12
19 15 16 13 100
1 6 0
2
0 100 200 300
11 x
(a) diagonal view (b) plan view
Fig. 4.14 Tensegrity grid.
Fig. 4.15 Unit cell of the tensegrity grid.
respectively; i.e., t = 3 and c = 2. There are m = 63 members and ns = 22 nodes

in total. The x- and y-coordinates (mm) of the nodes are shown in Fig. 4.14, and
the height of the grid is 100 mm. The axial forces of the horizontal members are
equal to 0 at the self-equilibrium state. Hence, these members are called bars.
2
Youngs modulus is E = 200 103 N/mm for all members. The cross-sectional
areas are 50 mm for struts and bars, and 5 mm2 for cables. In the following,
2
the units of length and force are mm and N, respectively, which are omitted for
brevity. The rank r of the equilibrium matrix D is 59. Therefore, the structure
has q = 63 59 = 4 self-equilibrium force modes denoted by g 1 , . . . , g 4 with the
coefficients = (1 , . . . , 4 )> .
The lower bound sLi and the upper bound sU i for the axial forces of the struts
and cables, respectively, are 10000 and 10000. Therefore, the maximum absolute
values of the strains are 0.001 for the struts and 0.01 for the cables in the process
of designing the nominal values of the self-equilibrium forces. The vertices of the
feasible region have been enumerated by cdd+ (Fukuda, 1999; Avis and Fukuda,
6.8
Eigenvalue
6.6
6.4
6.2
6
0 20 40 60 80 100 120 140
Vertex number
Fig. 4.16 Lowest eigenvalue of the tangent stiffness matrix of each solution corresponding to the
vertex of the feasible region.
5.92
5.915
Compliance
5.91
5.905
5.9
0 20 40 60 80 100 120 140
Vertex number
Fig. 4.17 Compliance against the specified load of each solution corresponding to the vertex of
the feasible region.
1996) to find 74 vertices. The mean values are calculated for the coefficients of all
the vertices to obtain the nominal value 0 = (0.4198, 0.5832, 3.2567, 0.4294)>.
The maximum and minimum forces of the cables of the nominal self-equilibrium
forces are 8507.8 and 1359.6, respectively, whereas those for the struts are 3847.6
and 8290.3, respectively. The value of 1 at the center is 6.4812. The lowest
eigenvalue 1 of K is positive at all the vertices after constraining the rigid-body
motions; i.e., the structure is stable at any set of self-equilibrium forces in the
feasible region for designing the nominal values.
The worst-case scenario under constraint (4.144) is next investigated. Let si =
500 for all members. Figure 4.16 shows the values of 1 at the 138 vertices of the
2600
2400
Force deviation
2200
2000
1800
1600
0 20 40 60 80 100 120 140
Vertex number
Fig. 4.18 Force deviation of each solution corresponding to the vertex of the feasible region.
region defined by Eq. (4.144). The worst value of 1 is 6.0948, which is about 94%
of the nominal value 6.4812.
The compliance W under static loads is next investigated. The vertical loads of
10 are given asymmetrically in the negative direction of z-axis at nodes 3, 12 and
20 indicated in Fig. 4.14(a). To exclude the rigid-body motions of the structure,
the displacements are constrained at nodes 11 and 19 in x-direction, at nodes 6 and
11 in y-direction, and at nodes 6, 10, 11, 18 and 19 in z-direction. Note that the
loads are assumed to be sufficiently small to investigate the stiffness against small
disturbance by linear analysis. The values of compliances at the vertices of the
feasible region are plotted in Fig. 4.17. As is seen, the compliance does not strongly
depend on the forces.
Finally, the deviation e of forces from the nominal value is plotted in Fig. 4.18.
As is seen, the deviation is widely distributed in the region of uncertainty; however,
the distribution depends on the definition of s. Although we have considered
only three performance measures, the method based on vertex enumeration is very
effective for the case where many performance measures are to be evaluated in the
process of force design.
Here it is worthwhile to mention that some investigators have conducted studies
on determining the worst possible stochastic responses due to incompletely known
probabilistic characteristics of the input quantities. These works include those of
Abbas and Manohar (2005, 2007), Chen, Lian and Yang (2002), Deodatis, Graham
and Micaletti (2003a, 2003b), Deodatis and Shinozuka (1989, 1991), Iyengar (1970),
Iyengar and Dash (1976), Iyengar and Manohar (1987), Papadopoulos, Deodatis and
Papadrakakis (2005), Sarkar and Manohar (1998), Shinozuka (1987), and Wall and
Deodatis (1994).
The references mentioned in Sec. 6.6.4, pertaining to earthquake engineering,
also fall into this category.
Chapter 5
Anti-Optimization in Buckling
Uncertainties and optimization are two major considerations on struc-

tural design. (Royset, Der Kiureghian and Polak, 2003)
Any intelligent fool can make things bigger and more complex... It
takes a touch of genius and a lot of courage to move in the opposite
direction. (Albert Einstein)
In this chapter, we first deal with a simple buckling anti-optimization problem. We

next treat the anti-optimization problem by minimizing the buckling load. The no-
tion of worst imperfection is introduced for a braced frame with attendant multiple
buckling loads, including the problem of mode interaction. The worst-case design
and worst imperfection are treated via anti-optimization under stress constraints.
We show how to utilize the convexity of the stability region. Finally, we end the dis-
cussion with worst imperfection of an arch-type truss associated with the multiple
member buckling at limit point.
5.1 Introduction
It is well known that the buckling loads of thin-walled structures such as plates
and cylindrical shells are drastically reduced in the presence of initial geometrical
imperfection, and the magnitude of reduction of the buckling load strongly depends
on the shape of the imperfection. Therefore, extensive research has been carried
out for anti-optimization for finding the worst imperfection, which minimizes the
buckling load under constraint on the norm of the imperfection.
In the general theory of stability of elastic conservative systems, the buckling is
characterized as a critical point of equilibrium state, which is classified to limit
point and bifurcation point. The details of the classification of critical points
may be consulted in Thompson and Hunt (1973, 1984). Worst imperfections have
been found for various types of critical points including unstable-symmetric bifur-
cation point (Ikeda and Murota, 1990), stable-symmetric bifurcation point (Ohsaki,
113
K1
K2 L2
Fig. 5.1 A barspring structure.
2002b), multiple bifurcation points (Ho, 1974; Peek and Triantafyllidis, 1992), and
hilltop branching point (Ikeda, Oide and Terada 2002; Ohsaki and Ikeda 2007a,
2007b). Since the buckling loads generally coincide as a result of optimization
against buckling (Masur, 1984; Ohsaki, 2000; Olhoff and Rasmussen, 1977), it is
very important to investigate worst imperfection for coincident critical points.
In the course of finding the worst imperfection, an anti-optimization problem
is formulated and solved explicitly or numerically. For the structures where the
prebuckling deformation is negligible for the evaluation of the buckling load, anti-
optimization problems can be formulated in a simple form of linear buckling analysis
(de Faria and de Almeida, 2003a).
Lindberg (1992a, 1992b) investigated the problem with multimodal dynamic
buckling with convex bounds in the Fourier coefficients of the imperfection. Pan-
telides (1996a) investigated worst imperfection in geometrical and material proper-
ties for stiffening elements using the ellipsoidal bounds. Takagi and Ohsaki (2004)
defined worst imperfection using the eigenvalues of the tangent stiffness matrix. El
Damatty and Nassef (2001) found worst imperfection of a ribbed shell by using a
genetic algorithm.
Consider a rigid vertical bar with length L as shown in Fig. 5.1 supported by a
horizontal spring with extensional stiffness K1 and a rotational spring with stiffness
K2 L2 . The bar is subjected to a vertical load P , and the rotation of the bar is
denoted by . The dotted and solid lines in Fig. 5.1 are the shapes before and after
deformation, respectively.
The total potential energy is written as
1 1
= K1 (L sin )2 + K2 L2 2 P L(1 cos ) (5.1)
2 2
Anti-Optimization in Buckling 115
The stationary condition of with respect to leads to the equilibrium equation

1
K1 L2 sin 2 + K2 L2 P L sin = 0 (5.2)
2
which is satisfied along the fundamental path = 0. For 6= 0, we have from
Eq. (5.2) the equation for the bifurcation path

P = K1 L cos + K2 L (5.3)
sin
Further differentiation of with respect to gives the criticality condition
K1 L2 cos 2 + K2 L2 P L cos = 0 (5.4)
c
from which the buckling (critical) load P , which is also called the bifurcation load
in this case, is obtained as
P c = (K1 + K2 )L (5.5)
The bifurcation load P c can be also obtained by taking the limit 0 for Eq. (5.3)
as the intersection between the fundamental path and the bifurcation path.
Uncertainty is introduced in the stiffness as
K1 = K10 + a1 , K2 = K20 + a2 (5.6)
where K10 and K20 are the nominal values and a1 and a2 are the uncertain parame-
ters.
Consider a simple case where the bound of a = (a1 , a2 )> is given by the
quadratic constraint as a special case of an ellipsoidal bound as
a> a D (5.7)
where D is the upper bound for the square of the magnitude of the uncertainty.
Then the buckling load is a function of a, and the anti-optimization problem is
formulated as
minimize P c (a) (5.8a)
>
c
It is seen from Eqs. (5.5) and (5.6) that P is a linear function of a1 and a2 .
Therefore, the inequality constraint (5.8b) is satisfied by equality at the anti-optimal
solution, and the solution to Problem (5.8) is found from the stationary condition
of the Lagrangian as (cf. Sec. 2.4.2)
1
a 1 = a 2 = L2
2
r (5.9)
1 2L4
= =
2 D
where is the Lagrange multiplier. Note that = corresponds to the optimal
solution that maximizes P c (a). For the case where L = 1 and D = 1, the anti-
optimal solution is computed as a1 = a2 = 0.7071 with = = 0.7071.
Consider the case where the bounds are given for a by interval as
aL a a U (5.10)
with the lower bound aL and upper bound aU . Since P c is a linear function of a1
and a2 , as observed from Eqs. (5.5) and (5.6), the anti-optimization problem turns
out to be a linear programming problem
minimize P c (a) (5.11a)
L U
The anti-optimal solution exists at a vertex of the feasible region, and the obvious
anti-optimal solution is found as (a1 , a2 ) = (aL1 , aL2 ), because P c (a) is a monotoni-
cally increasing function of a.
5.3 Buckling Analysis
Consider an elastic conservative structure subjected to quasistatic proportional

nodal loads P parameterized by the load factor and the constant vector p of
load pattern as P = p; i.e., the structure has potential energy and the direction
of the dead load P is fixed. Let K denote the tangent stiffness matrix (stability
matrix) at an equilibrium state undergoing large deformation.
The eigenvalue problem of K is formulated as
Ki = i i , (i = 1, . . . , n) (5.12)
where i is the ith eigenvalue numbered in nondecreasing order as 1 2
n , i is the eigenvector associated with i , and n is the number of degrees of
freedom. The eigenvector is ortho-normalized as
>
i j = ij , (i, j = 1, . . . , n) (5.13)
where ij is the Kronecker delta.
The buckling load factor c is defined as the smallest positive load factor satis-
fying the criticality condition 1 = 0, as
Kc1 = 0 (5.14)
where c1
is called the buckling mode. There are many approaches for finding the
buckling load of a structure undergoing large deformation (Riks, 1998). However,
in this book, the details of buckling analysis are omitted, and we assume that the
buckling load can be found numerically within necessary accuracy.
Consider a structure for which the deformation before buckling, called the pre-
buckling deformation, is sufficiently small; e.g., columns under axial compression
and building frames under vertical loads. In this case, the axial forces or stresses
in each member or element of the structure is almost proportional to , and K is
expressed by the sum of the linear stiffness matrix K E that is independent of and
the geometrical stiffness matrix K G = K G0 that is proportional to , where K G0
is a constant matrix defined by p. Hence, the linear buckling load factor b , which
is often called simply the linear buckling load, is found as the smallest positive value
of satisfying
(K E + b K G0 )b = 0 (5.15)
where b is the linear buckling mode. Note that the linear eigenvalue problem
(5.15) generally has negative buckling load factors, because the structure may buckle
against the loads applied to the reverse direction and the symmetric matrix K G0 is
indefinite. However, the direction of the load is generally fixed, and we can assume
that the buckling load factor takes a positive value. The ith positive smallest value
of b is called the ith linear buckling load factor; however, we are interested in the
lowest positive buckling load only.
5.4 Anti-Optimization Problem
The buckling load factor represents the safety against static loads when plastic
failure is not considered. Therefore, in the following, the linear buckling load factor
is minimized in an anti-optimization problem. As demonstrated with the simple
example in Sec. 5.2, the anti-optimization problem can be formulated with either
the ellipsoidal or linear bound of the uncertain parameter vector a representing
variability of load, shape, material property, etc. For the ellipsoidal bound with
weight matrix W , the anti-optimization problem is formulated as
minimize b (a) (5.16a)
>
subject to a W a 1 (5.16b)
The details of the ellipsoidal model can be found in Sec. 3.4.
The anti-optimal solution to Problem (5.16) can be obtained explicitly if the
buckling load is linearized using sensitivity coefficients. Sensitivity analysis of the
eigenvalue of vibration has been presented in Sec. 3.6.3 as the simple case of the
eigenvalue sensitivity. For linear buckling load, differentiation of Eq. (5.15) with
respect to ai leads to
K E b b b K G0 b b
+ KE + K G0 b + b + b K G0 =0 (5.17)
ai ai ai ai ai
Since K E is positive definite and the buckling load b is positive, b is normalized
in view of Eq. (5.15) as
b> K G0 b = 1 (5.18)
Premultiplying b> to Eq. (5.17) and using Eqs. (5.15) and (5.18), we obtain the
sensitivity coefficient

b K E K G0
= b> + b b (5.19)
ai ai ai
It is seen by comparing Eq. (5.19) with Eq. (3.96) in Sec. 3.6.3 that the sensitivity
coefficients of the buckling loads can be computed in a similar manner as those of the
eigenvalues of free vibration. However, the geometrical stiffness matrix K G0 for the
unit load factor corresponding to the load pattern vector p implicitly depends on a i
through the axial forces or stresses against p. Therefore, sensitivity analysis of the
responses against the static load p is needed to compute the sensitivity coefficients
of the linear buckling load (see Sec. 3.6.2 for sensitivity analysis of static responses).
When the bounds of the uncertain parameters are given by interval, and b (a) is
linearized with respect to a, the anti-optimization problem turns out to be a linear
programming problem. For the case without linearization of b (a) and with linear
constraints for a, the anti-optimal solution can be found using vertex enumeration
of the feasible region as shown in Sec. 5.6.
5.5 Worst Imperfection of Braced Frame with Multiple Buckling

Loads
5.5.1 Definition of frame model

Imperfection sensitivity of multiple buckling loads has been extensively investigated
to demonstrate that imperfection sensitivity increases as the result of the coinci-
dence of buckling loads (Thompson and Hunt, 1984). However, Ohsaki (2002a)
showed that optimization of a braced frame that results in multiple buckling loads
does not lead to any enhancement of imperfection sensitivity. In this section, the
worst imperfection is investigated for a braced frame with multiple (coincident)
linear buckling loads.
Consider a 3-story 3-span braced plane frame as shown in Fig. 5.2 subjected to
nodal proportional loads P in vertical direction at all the nodes connecting the
beams and columns, where P = 1.0 104 N. In the following, the units mm and N
of the length and force, respectively, are omitted for brevity.
The braces are modeled by the truss elements. The intersecting pair of braces are
not connected at the centers, and Euler buckling of the braces is ignored. Sandwich
cross-sections are assumed for columns and beams; i.e., the second moment of inertia
I and the section modulus Z are functions of the cross-sectional area A as
I = r2 A, Z = rA (5.20)
where r is the distance between the flange and the center axis of the member. In
the following examples, the value of r is 80.0 for the interior columns in the first
story, and 60.0 for the remaining beams and columns. Youngs modulus is 2.0 105
for all members.
The standard assumption of rigid floor is used, i.e., the nodes in the same floor
has the same horizontal displacement. Each column is divided into two beam-
column elements, and the Green strain is used for the strain-displacement relation.
Since the prebuckling deformation of a frame subjected to vertical loads is very
y
P P P P
5000 9 10 11 12
5000 5 6 7 8
6000 1 2 3 4
4000 4000 4000
Fig. 5.2 A 3-story 3-span braced plane frame.
small, the buckling load is evaluated by linear buckling analysis, whereas the buck-
ling load of an imperfect structure is evaluated by nonlinear path-tracing analysis,
for which the displacement increment approach is used. The effect of yielding is not
considered here to investigate the elastic properties of the frame, thereby simplifying
the anti-optimization process.
Let i () denote the maximum absolute value of the stress of the flange at the
two ends of the ith beam-column element. The maximum value of i () throughout
the frame is denoted by m () as
m () = max i () (5.21)
i
The maximum load factor m () is defined, as follows, so that m () reaches the

specified upper bound m :
m = max{ | m () m } (5.22)
The purpose of this section is summarized as follows:
(i) The worst imperfection is investigated for a braced frame, where the norm of
the imperfection is defined by the absolute value of the inclination of each story
and the out-of-straightness of each column; i.e., these geometrical imperfections
are regarded as the uncertain parameters for formulating the anti-optimization
problem. The maximum load m is defined by the stresses as Eqs. (5.21) and
(5.22) at the two ends of the elements.
(ii) The worst imperfection mode of the nodal locations found by geometrically
nonlinear path-tracing analysis is compared with the buckling mode, and the
effect of modal interaction on imperfection sensitivity is evaluated using the
Table 5.1 Cross-sectional areas of the optimal solution.
1st story 2nd story 3rd story
External column 5838.64 4886.90 1999.45

Internal column 6139.51 4930.91 2060.32
External beam 1500.00 1500.00 1500.00
Internal beam 1500.00 3102.45 1500.00
External brace 44.42 0.0 8.59
Internal brace 22.56 76.69 5.30
third- and fourth-order differential coefficients of the total potential energy with
respect to the generalized displacements in the directions of buckling modes.
(iii) The worst-case design and the corresponding worst imperfection that maximize
the increase of the maximum stress are found by linear analysis under specified
loads, and compared with the worst imperfection obtained by the nonlinear
path-tracing analysis.
5.5.2 Worst imperfection of optimized frame

The optimal braced frame for minimizing the total structural volume is first found
under constraint on the linear buckling load factor with the lower bound = 100,
where the optimization library DOT Ver. 5 (VR&D, 1999) is used. Note that the
purpose here is to obtain a frame with multiple linear buckling loads as a result
of optimization. Therefore, the properties of the optimal braced frame are not
discussed. Using the symmetry condition of the frame with respect to the y-axis,
the members are classified to those in internal and external spans in each story.
Hence, the number of design variables is 18, including beams, columns, and braces.
Table 5.1 shows the optimal cross-sectional areas. Note that the beams in the 1st,
2nd, and 3rd stories refer to those in the 2nd floor, 3rd floor, and roof, respectively.
The four lowest linear buckling loads are 99.99, 100.00, 100.01, and 111.88, i.e., the
three buckling loads are almost equal to at the optimal solution. The coincident
buckling modes of the three multiple buckling loads are shown in Fig. 5.3(a)(c);
i.e., the buckling loads corresponding to two sway buckling modes (Modes 1 and 3)
and one non-sway buckling mode (Mode 2).
The imperfections are given for the x-coordinates of the nodes including those at
the centers of the columns. Let R denote the upper bound of the inclination Ri (i =
1, . . . , f ) of the ith story, as illustrated in Fig. 5.4, where f is the number of stories
that is equal to 3 for this example. The local imperfection (out-of-straightness of
the column) is defined by the additional inclination i of the elements in the ith
column as shown in Fig. 5.4, and its upper bound is denoted by . Note again that
each column is divided into two elements.
The number of columns, which is 12 for this example, is denoted by mc . Since
the locations of all the nodes, including the centers of columns, are defined by
(a) Mode 1 (b) Mode 2 (c) Mode 3
Fig. 5.3 Buckling modes of optimal solution under linear buckling constraint.
Ri
Fig. 5.4 Definition of imperfection parameters Ri and i .
their original locations and the imperfection vectors R = (R1 , . . . , Rf )> and =
(1 , . . . , mc )> , R and are regarded as the uncertain parameters in the anti-
optimization problem. The maximum load m of the imperfect frame in Table 5.1
is defined by the stress constraints (5.21) and (5.22), as the load factor when the
maximum stress m computed by path-tracing analysis for specified R and reaches
the upper bound. Hence, the anti-optimization problem is stated as
minimize m (R, ) (5.23a)

subject to |Ri | R, (i = 1, . . . , f ) (5.23b)
c
|i | , (i = 1, . . . , m ) (5.23c)
The maximum stress m at buckling of the optimal perfect frame is 511.37.

Therefore, the upper-bound stress of imperfect frames has been set as a slightly
smaller value m = 510.0, because the maximum stress of an imperfect frame may
rapidly increase at the load level slightly below the buckling load of the perfect
frame. The upper bounds for the story inclination and the out-of-straightness of
the column are given as R = = 1/500, which is a practically acceptable value
(Kim and Chen, 1996).
Fig. 5.5 Worst imperfection of the optimal frame.
Since the maximum loads of imperfect systems are highly nonlinear functions
of the imperfection parameters R and , it is very difficult to obtain the anti-
optimal solution by a gradient-based nonlinear programming approach. Therefore,
the worst imperfection is searched for at the upper and lower bounds of the column
out-of-straightness i = 1/500, whereas the three cases Ri = 1/500, 0, 1/500 are
searched for story inclination. Hence, the anti-optimization problem is formulated
as a mixed integer nonlinear programming problem.
From f = 3, mc = 12, and symmetry of the frame, the total number of solutions
is 33 212 /2 = 55296. Therefore, it is possible to find the anti-optimal solution
by the enumeration process. Incremental path-tracing analysis has been carried
out for each set of imperfection parameters to find the maximum load. The worst
imperfection found by this enumeration is shown in Fig. 5.5 and in the first row
of Table 5.2. In solution of the table, the first three columns give the story incli-
nation R1 , R2 , R3 , and the remaining show the out-of-straightness of the columns
1 , . . . , 12 , where the column numbers are given in Fig. 5.2. The signs + and
correspond to 1/500 and 1/500, respectively. As is seen from Fig. 5.5, the worst
imperfection is similar to a sway mode (Mode 1 in Fig. 5.3) and is asymmetric with
respect to the y-axis.
The maximum load of the imperfect system with worst imperfection is 76.067,
which is about 76% of the maximum load 100 of the perfect system. The relations
between m and the load factor obtained by path-tracing analysis for the perfect
frame and the frame with worst imperfection are plotted in Fig. 5.6. Note that no
drastic decrease due to geometrical nonlinearity is observed in the imperfect system.
For a larger frame with more members, it is very difficult to find the anti-optimal
solution by enumeration. Therefore, in order to present a general method, we next
find approximate anti-optimal solutions by simulated annealing (see Sec. 2.6.3 for
details of the method). The initial temperature parameter is 0.02, and the temper-
ature is multiplied by 0.99 at each iterative step. The iteration is terminated after
500 steps. The results from the randomly generated ten different initial solutions
are listed as SA-1, . . . , SA-10 in Table 5.2. It is seen that a good approximate
Table 5.2 Optimization results by enumeration and simulated annealing.
Method Maximum load Solution

R1 R2 R3 1 2 3 4 5 6 7 8 9 10 11 12
Enumeration 76.067 + + + + + + + + +
SA-1 76.306 0 + + + + + +
SA-2 76.216 + + + + + + + + +
SA-3 76.154 + + + + + + + + +
SA-4 76.306 0 + + + + + +
SA-5 76.454 0 + + + + + +
SA-6 76.638 + + + + +
SA-7 77.397 + + + + +
SA-8 76.995 + + 0 + + + + + + +
SA-9 76.398 + + + + + + + +
SA-10 76.763 + + + + + +
100
Perfect
80
Load factor
60 Imperfect
40
20
0
0 0.5 1 1.5
Maximum stress
Fig. 5.6 Relations between maximum stress m and load factor of the perfect and imperfect
systems.
solution with m = 76.154 has been found by the SA with random start, where
the total number of analysis is 500 10 = 5000, which is less than 1/10 of the
enumeration.
5.5.3 Mode interaction

It is well known that the maximum load of a structure with coincident buckling
loads may be drastically reduced in the presence of small imperfection due to the
interaction of buckling modes (Thompson and Hunt, 1974; Thompson and Supple,
1973).
In order to investigate the effect of mode interaction, the imperfection in the
Table 5.3 Relation between coefficients (c1 , c2 , c3 ) and the maxi-

mum load m .
(c1 , c2 , c3 ) m (c1 , c2 , c3 ) m (c1 , c2 , c3 ) m
(1, 0, 0) 80.44 (0, 1, 0) 79.37 (0, 0, 1) 78.93

(0, 1, 0) 79.42 (1, 1, 0) 79.65 (1, 1, 0) 79.89
(1, 0, 1) 79.84 (1, 0, 1) 79.20 (0, 1, 1) 78.73
(0, 1, 1) 78.71
linear combination of the eigenmodes corresponding to the zero eigenvalues of the

tangent stiffness matrix at the critical point of the perfect system should be consid-
ered. However, these eigenmodes are almost the same as those of the linear buckling
modes in Fig. 5.3(a)(c), because the prebuckling deformation of the perfect system
is negligibly small. Therefore, we use the linear buckling modes for investigation of
the effect of mode interaction.
Let c1 , c2 and c3 denote the coefficients for the three lowest linear buckling
modes c1 , c2 and c3 , respectively, and their linear combination is defined as
c = c1 c1 + c2 c2 + c3 c3 (5.24)
where ci (i = 1, 2, 3) are normalized by c> c

i i = 1.
Denote by i the inclination of the ith column element corresponding to c . The
inclination of the column element from the vertical axis is then defined such that
1/500 for i > 0 and 1/500 for i 0, and the nodal locations are computed from
the base to the top by summing up the inclinations of column elements multiplied
by the element lengths. This way, the imperfection that has i = 1/500, 0, or
1/500 is generated by assigning the set (c1 , c2 , c3 ). Note that the inclination may
be different for the columns in the same story, for simplicity.
The maximum load m computed by path-tracing analysis for various values
of (c1 , c2 , c3 ) are listed in Table 5.3. Note again that the buckling load of the
perfect system is 100. As is seen, the maximum load does not strongly depend on
the imperfection pattern, and no mode interaction is observed; e.g., m = 79.65 for
(c1 , c2 , c3 ) = (1, 1, 0) corresponding to the imperfection in the intermediate direction
between Modes 1 and 2 is even greater than m = 79.37 for the imperfection
(c1 , c2 , c3 ) = (0, 1, 0) in the direction of Mode 2.
The effect of mode interaction is investigated more accurately below using the
differential coefficients of the total potential energy (Thompson and Hunt, 1984;
Ohsaki, 2002a). It is known that the interaction in the third-order differential
coefficients of the total potential energy, called third-order interaction, may exist
in a semi-symmetric system (Huseyin, 1975; Ohsaki and Ikeda, 2007a), where a
symmetric and an asymmetric bifurcation points coincide as is the case with this
braced plane frame.
Note that the bifurcation corresponding to Mode 2 in Fig. 5.3 is asymmetric,
although the mode shape is symmetric with respect to the y-axis. The asymmetry
exists in the potential energy . Let U c denote the value of the displacement
vector at the critical point of the perfect system. For Mode 2, the deformation
corresponding to the displacement vector U = U c + c2 and U = U c c2 have
different values of . However, the asymmetry is very small, because the maximum
load for the imperfections (c1 , c2 , c3 ) = (0, 1, 0) and (0, 1, 0) are almost the same
as those shown in Table 5.3.
Let ,i denote the differential coefficient of with respect to the generalized
displacement in the direction of ci . The third- and fourth-order differential coeffi-
cients are computed as
,222 = 8.7 104 , ,112 = 1.3 104 , ,233 = 6.6 104 ,
,1111 = 3.9 102 , ,2222 = 5.1 102 , ,3333 = 8.5 101 , (5.25)
,1122 = 2.9 102 , ,1133 = 5.7 102 , ,2233 = 3.3 102
where the vanishing terms are not shown, e.g., 111 = 333 = 0 due to symme-
try (see Ohsaki and Ikeda (2007a) for details of computation of the higher-order
differential coefficients, where the symbolic computation software package Maple 9
(Maplesoft, 2004) has been extensively used).
To investigate the effect of third-order interaction, let q denote the generalized
coordinate in the direction of the asymmetric bifurcation mode c2 :
U = U c + qc2 (5.26)
Then can be expanded as
1 1 1
= c + ,2 q + ,22 q 2 + ,222 q 3 + ,2222 q 4 (5.27)
2 6 24
where c is the value of at the critical point.
Since ,2 = 0 and ,22 = 0 are satisfied from the equilibrium and the criticality
conditions, the leading term of governing the symmetry is ,222 . Suppose q = 100
corresponding to maximum out-of-straightness of the column about 1/100 in the
direction of c2 , which is larger than the expected value in design practice. In this
case, the orders of the third- and fourth-order differential coefficients ,222 and
,2222 , respectively, are 102 and 106 . Therefore, the third-order term is negligibly
small compared to the fourth-order term, and the asymmetry of c2 is very small;
hence, the third-order interaction can be neglected.
The fourth-order interaction should also be investigated for this almost sym-
metric system. Since the fourth-order interaction terms ,1122 , ,1133 and ,2233
among the three modes are all positive and their absolute values are in the same
order as those of 1111 , ,2222 and ,3333 , no increase of imperfection sensitivity
will be observed for this frame (Thompson and Hunt, 1984).
In order to further investigate the effect of mode interaction, maximum loads of
imperfect systems of non-optimal frames are computed. If the cross-sectional areas
of the braces of the optimal frame in Table 5.1 are replaced by 0, i.e., if we remove
the braces, the lowest buckling mode is a sway mode, and the three lowest buckling
loads are 22.53, 35.89 and 53.43, respectively, which do not coincide. The maximum
stress at buckling is 115.76. The worst imperfection is computed with m = 115.0,

i = 1/500, and Ri = 1/500, 0, 1/500, by enumeration to find the corresponding
maximum load is 15.38. Therefore, the reduction of the maximum load due to the
worst imperfection is about 32%, which is larger than the 24% of the optimal frame
with coincident buckling. Note that the worst imperfection in this case is similar
to the sway mode (Mode 1 in Fig. 5.3). In conclusion, the coincidence of buckling
loads does not enhance imperfection sensitivity of a braced frame.
5.5.4 Worst-case design and worst imperfection under stress con-

straints
Since the effect of geometrical nonlinearity on the maximum load defined by stress
constraint of an imperfect frame with practically acceptable proportion is rather
small as demonstrated in Sec. 5.5.3, the worst imperfection may be successfully
found by evaluating the maximum load under stress constraints by linear analysis.
In order to investigate the effect of imperfection on the maximum stress, consider
a pin-supported column with cross-sectional area A under axial load N . Suppose
an imperfection (out-of-straightness) is given at the center of the column. The
normal stress 0 under axial load N is computed as
N
0 = (5.28)
A
Since the bending moment at the center is N , the maximum stress due to bending
is
N
1 = (5.29)
Z
where Z is the section modulus.
Suppose the column consists of a sandwich cross-section, and denote by r the
distance between a flange and the center axis of the column. Then Z is given as
Z = Ar, and the ratio of bending stress to the axial stress is obtained as
1
= (5.30)
0 r
i.e., increase ratio of stress due to imperfection is /r. Let and denote the angle
of out-of-straightness of the column due to the imperfection at the center and the
slenderness ratio of the column, respectively. Then the following relation holds:
1
= (5.31)
0 2
For example, for = 1/500 and = 100, the ratio is obtained as 1 /0 = 0.1.
Since both ends of the column are supported by beams when we consider buckling
of a frame, and is usually less than 100, the increase ratio due to imperfection
will be less than 10%. Therefore, we can conclude that the effect of geometrical
imperfection on the maximum stress is not very large.
Table 5.4 Cross-sectional areas of the worst-case design

for maximizing the stress ratio I / P evaluated by linear
analysis.
1st story 2nd story 3rd story
External column 5850.58 4896.46 2009.12

Internal column 6152.71 4941.28 2070.51
External beam 1504.06 1503.20 1502.42
Internal beam 3106.50 1501.88 1500.96
External brace 46.04 8.32 19.77
Internal brace 13.14 91.57 15.52
Fig. 5.7 Worst imperfection for maximizing the stress ratio I / P by linear analysis.
We next consider the worst-case design of the 3-story 3-span frame. Let P and
I
denote the maximum absolute value of the stresses of the perfect and imperfect
systems, respectively, obtained by linear analysis, where in general, I > P holds.
The worst imperfection is defined so as to maximize the ratio I / P due to the
imperfection defined by the vector consisting of the out-of-straightness of each
column element, which is assumed to be a continuous variable with an upper bound
of 1/500 and a lower bound of 1/500. The nodal locations are computed from the
base to the top by summing up the inclinations of column elements multiplied by
the element lengths. The vector A of the cross-sectional areas is also considered
as design variable vector to find the worst-case design, where the symmetry of
the frame is incorporated. The constraint is given for the linear buckling load as
b 100.
Anti-optimization to find the worst-case design is carried out by the optimiza-
tion software package IDESIGN Ver. 3.5 (Arora and Tseng, 1987), where sequential
quadratic programming is used. The cross-sectional areas of the worst-case design
are listed in Table 5.4, and the corresponding worst imperfection is shown in Fig. 5.7.
The value of the objective function I / P at = 100 is 1.0912; i.e., the maximum
stress increases about 9.1% due to the worst imperfection. Therefore, if the maxi-
100
80 Perfect
Load factor
60
Imperfect
40
20
0
0 0.5 1 1.5
Maximum stress
Fig. 5.8 Relations between maximum stress ratio m / P and load factor by path-tracing analysis
of the worst-case design.
mum stress is specified, the ratio of the maximum load of the imperfect system to
that of the perfect system is 1/1.0912 = 0.91642 due to the worst imperfection.
Geometrically nonlinear analysis has been carried out to verify the validity of
using linear analysis for anti-optimization. The maximum loads of the imperfect
system defined by the upper-bound stress m = P at = 100 found by geomet-
rically nonlinear analysis is = 81.189. Therefore, the reduction of the maximum
load is 19%, and the reduction ratio is magnified twice due to geometrical non-
linearity. Figure 5.8 shows the relation between the maximum stress ratio m / P
and the load factor of the perfect and imperfect systems. The initial stiffnesses are
almost the same, and a slightly nonlinear behavior is observed near the maximum
load factor.
5.6 Anti-Optimization Based on Convexity of Stability Region
In this section, we briefly summarize the anti-optimization method for linear buck-
ling loads of multi-parameter systems utilizing the convexity of the stability region
that is derived from the convexity of the reciprocal value of the linear buckling load.
Consider the case where uncertainty exists in the load vector P , which is pa-
rameterized by
XN
P = a i pi (5.32)
i=1
where a = (a1 , . . . , aN )> is the vector of uncertain parameters, which take the
role of coefficients for the load pattern vectors pi , and N is the number of loading
patterns.
Consider the situation where the total load vector P is multiplied by the load
factor , which is increased to find the buckling (critical) load factor c for each
specified parameter vector a. For this multi-parameter loading condition, the set of
critical points in the space of a forms a surface that is called the stability boundary
(Huseyin, 1975). The boundary degenerates to a point for the conventional case
of proportional loading with single load pattern. If the prebuckling deformation is
negligibly small, and the buckling load is estimated by linear buckling analysis, the
anti-optimization problem can be solved efficiently, as follows, utilizing the convexity
of the stability region (de Faria, 2000; de Faria and Hansen, 2001a, 2001b).
De Faria and de Almeida (2003a) presented an anti-optimization method of a
plate with variable thickness. Although they presented detailed investigation for
general cases, we briefly present the results below based on the concavity of the
reciprocal value of the buckling load.
We first derive the concavity property of the lowest eigenvalue of the tangent
stiffness matrix at an equilibrium state, which is assumed to be a linear function
of a. Let 1I and 1II denote the lowest eigenvalues of the tangent stiffness matrices
K(aI ) and K(aII ), respectively, for two sets of parameter values aI and aII . The
interpolation between K(aI ) and K(aII ) is given by the parameter a = aI + (1
)aII (0 1). The eigenvector associated with the lowest eigenvalue of K(a )
is denoted by . Based on the Rayleigh principle, the following relations hold:
> K I > K II
>
1I , 1II (5.33)
>
Therefore, the following inequality is derived for the lowest eigenvalue 1 of K(a ):
> [K I + (1 )K II ]
1 =
>
> K I > K II (5.34)
= >
+ (1 )
>
1I + (1 )1II
It is seen from Eq. (5.34) that the lowest eigenvalue 1 (a) of K(a) is a concave
function of the load parameters a. Hence, the feasible (stable) region S in the
space of a satisfying 1 0 is convex.
Consider next the linear buckling problem, where the tangent stiffness matrix
K(a) is divided into the linear stiffness matrix K E and the geometrical stiffness
matrix K G = K G0 . Note that K E is independent of P , and K G is a linear
function of P , hence, K G is a linear function of a. Then the problem of finding the
linear buckling load factor b is formulated as
(K E + b K G0 ) = 0 (5.35)
Suppose we are interested in the positive smallest buckling load factor and define
b as
1
b = (5.36)
b
Then b is the largest eigenvalue of the following problem, which is denoted by

max :
K G0 = b K E (5.37)
max
Convexity of can also be proved based on the Rayleigh principle. For the
two vectors of uncertain parameters aI and aII , and the interpolation parameter
(0 1), define a = aI + (1 )aII and denote by the linear buckling
mode for a = a . Then the following relations hold from the Rayleigh principle
and positive definiteness of K E :
> K IG0 maxI > K II
G

, maxII (5.38)
> K E > K E
where maxI and maxII are the values of max corresponding to a = aI and aII ,
respectively. Therefore, the following inequality is derived for max corresponding
to a = a :
> [K IG + (1 )K II
G ]

max =
> K E
> K I > K II (5.39)
= > G (1 ) > G
KE K E
maxI + (1 ) maxII
It is seen from Eq. (5.39) that the largest eigenvalue max (a) of Eq. (5.37) is a
convex function of a for b > 0.
On the other hand, for b < 0, the buckling load with the smallest absolute
value corresponds to the smallest eigenvalue min of Eq. (5.37), which can be shown
to be a concave function of a in a similar manner as the case of b > 0, using the
Rayleigh principle. If we define the linear buckling load as the buckling load with
the smallest absolute value, then its reciprocal value b is found from
b = max{ max , min} (5.40)
Since min is concave, min is convex, and b is defined as the maximum value
of the two convex functions, which is also convex. Therefore, the reciprocal value
of the linear buckling load factor is a convex function of the load pattern vector a
irrespective of the sign of b .
Suppose ai 0 (i = 1, . . . , N ), for simplicity, and consider the problem of finding
the worst-load scenario defined by the parameters a under the constraints
N
X
ai D, (5.41a)
i=1
ai 0, (i = 1, . . . , N ) (5.41b)
where D is a specified positive value. The objective of the anti-optimization here is
to find the minimum buckling load factor b associated with the set of load param-
eters satisfying Eq. (5.41). Therefore, the problem is converted to a maximization
P2
P1
H H
y
1.5
a2
0.5
0
0 0.2 0.4 0.6 0.8 1 1.2
a1
Fig. 5.10 Stability boundary (solid line) of the 2-bar truss; : anti-optimal solution.
problem of b (a). Since the feasible region defined by Eq. (5.41) is convex, the
anti-optimal solution that maximizes the convex function b (a) exists at a vertex
of the convex polyhedron (5.41).
Let V denote the set of vertices of the region (5.41). The anti-optimal solution
is found by solving the following problem:
maximize b (a) (5.42a)
subject to a V (5.42b)
The solution of Problem (5.42) can be found by vertex enumeration, if the number
of uncertain load parameters is moderately small.
For example, consider a 2-bar truss as shown in Fig. 5.9, where H = 1000 mm.
The two members consist of the same cross-section, where the cross-sectional area
is 300 mm2 , and the moment of inertia is 7.5 105 mm4 . Youngs modulus is
2
200 kN/mm . Each member is divided into two elements to consider member
buckling. In the following, the units of length and force are mm and kN, respectively.
0.03
0.025
0.02
0.015
0.01
0.005
0
1
0.8
0.6
0
0.2
a2
0.4
0.4
0.6 0.2
a1
0.8
10
Fig. 5.11 Reciprocal value of linear buckling load b = 1/b .
The horizontal load P1 and the vertical load P2 are parameterized as

P1 = 1000a1, P2 = 1000a2 (5.43)
The linear buckling load factor b is computed for various sets of (P1 , P2 ), and the
values of b P that define the stability boundary are plotted as shown in the solid
line in Fig. 5.10. Suppose the constraint is given as
a1 + a2 1.0, a1 0, a2 0 (5.44)
Then the maximum load factor is attained at P = (0, 1.0587)> indicated by in
Fig. 5.10, which is located at a vertex of the stability boundary.
Consider next the braced frame shown in Fig. 5.2 in Sec. 5.5. The cross-sectional
areas are 1500.0 for beams and columns, and 100.0 for braces. Other parameters
are the same as those in Sec. 5.5. Let P1 and P2 denote the vertical loads applied
at the exterior and interior nodes, respectively, which are parametrized by a 1 and
0.8 0.028
0.026
0.024
0.022
0.6 0.02
0.018
0.016
a2
0.014
0.012
0.4 0.01
0.008
0.006
0.004
0.2 0.002
0
0 0.2 0.4 0.6 0.8 1
a1
Fig. 5.12 Contour lines of the reciprocal value of linear buckling load b = 1/b .
a2 as
P1 = 10000a1, P2 = 10000a2 (5.45)
The reciprocal value b = 1/b of the linear buckling load is computed for each set
of (P1 , P2 ) for 0 a1 1 and 0 a2 1 to obtain the plot in Fig. 5.11, where
the contour lines are also plotted. It is confirmed from Fig. 5.11 that b is a convex
function of (a1 , a2 ).
Suppose the constraint is given as
a1 + a2 0.8, a1 0, a2 0 (5.46)
Then the feasible region is a triangle as shown in the thick straight lines in Fig. 5.12,
and the worst value of b is about 0.0187, which is attained at the vertex (a1 , a2 ) =
(0.8, 0).
5.7 Worst Imperfection of an Arch-type Truss with Multiple Mem-

ber Buckling at Limit Point
5.7.1 Introduction
It has been pointed out by many researchers that imperfection sensitivity generally
increases as a result of coincidence of bifurcation points. Thompson and Hunt
(1973) showed that the critical load of a latticed column drastically decreases due
to interaction of the critical loads associated with local and global buckling modes.
However, very large imperfection beyond the practically acceptable level for civil
engineering was considered therein. Ohsaki (2002a, 2002c) demonstrated that a
building frame with practical proportions does not have severe reduction of buckling
loads due to mode interaction (see Sec. 5.5 for details).
Thompson and Schorrock (1975) derived imperfection sensitivity laws for the
hilltop branching point, often abbreviated as hilltop point, that has a bifurcation
point at a limit point. The critical loads of imperfect systems are expressed as
a piecewise linear function of the imperfection parameter. Therefore, imperfec-
tion sensitivity is even decreased by the coincidence of critical points from the
imperfection-sensitive structure with a simple bifurcation point.
Ohsaki (2000) demonstrated that optimization of shallow shell-type truss against
buckling often leads to a hilltop point with multiple bifurcation. Ikeda, Oide and
Terada (2002) verified the imperfection sensitivity laws using random imperfections.
Ohsaki (2003b) derived imperfection sensitivity law for the case where an asymmet-
ric bifurcation point exists at a limit point. Ikeda, Ohsaki and Kanno (2005) derived
imperfection sensitivity laws for hilltop points of symmetric systems based on group
theoretic approach.
Ohsaki and Ikeda (2006) derived a piecewise linear imperfection sensitivity law
for the hilltop point with many symmetric bifurcation points. Imperfection sen-
sitivity of degenerate hilltop points was investigated by Ohsaki and Ikeda (2009).
The worst imperfection for hilltop points is presented in Ikeda, Oide and Terada
(2002) and Murota and Ikeda (1991). In this section, we verify the results in Ohsaki
and Ikeda (2006) using random imperfections. Interested readers may consult with
Ohsaki and Ikeda (2007a, Chap. 11) for details on random imperfection. Anti-
optimization problems are formulated to derive the worst imperfection. Numerical
examples are given for an arch-type truss.
5.7.2 Hilltop branching point of perfect system

Consider a finite-dimensional structure subjected to quasistatic proportional loads
P , which is defined by the load factor and the vector p of load pattern as P = p.
Let n denote the number of degrees of freedom, and define the nodal displacement
vector as U = (U1 , . . . , Un )> .
We consider an conservative linear elastic system that has the total potential
energy denoted by Ve (U , ). The stability matrix (tangent stiffness matrix) K is
given as !
2 Ve
K= (5.47)
Ui Uj
The ith eigenvalue i (i i+1 ) and eigenvector i are defined by
Ki = i i , (i = 1, . . . , n) (5.48)
where i is normalized as
>
i i = 1, (i = 1, . . . , n) (5.49)
Suppose 1 > 0 is satisfied at the initial state = 0. The first critical point as is
increased from 0 is defined by 1 = 0 with the critical load factor = c .
Consider the hilltop point where m 1 bifurcation points exist at a limit point;
i.e., m eigenvalues 1 , . . . , m vanish at the critical point. Let qi denote the gener-
alized displacement in the direction of ci as
n
X
c
U =U + qi ci (5.50)
i=1
where U c is the displacement vector at the critical point. The coordinates defined
by q1 , . . . , qm are called active coordinates, whereas those by qm+1 , . . . , qn are called
passive coordinates.
The increment of from c is denoted by as
= c (5.51)
>
Then the total potential energy is written in terms of q = (q1 , . . . , qn ) and as
V (q, ). In the following, differentiation of V with respect to qi is denoted by the
subscript ( ),i , and all variables are evaluated at the critical point without explicit
notation of the superscript ( )c . Differentiation of V with respect to is denoted
by the subscript ( ), .
5.7.3 Imperfection sensitivity of hilltop branching point

Let a = (a1 , . . . , aN )> denote the imperfection vector representing errors in nodal
locations, cross-sectional areas, etc. The vector a is defined by the value of the
perfect system (nominal value) a0 , imperfection parameter , and the vector d of
imperfection pattern as
a() = a0 + d (5.52)
Differentiation of the total potential energy with respect to is denoted by the
subscript V, as
XN
V
V, = di (5.53)
i=1
ai
The load increment for an imperfect system is defined by
= c0 (5.54)
where c0 is the critical load factor of the perfect system.
Hence, the total potential energy of an imperfect system is written as V (q, , ),
as a function of n + 2 variables , q1 , . . . , qn and . Suppose 1 , . . . , m1 corre-
spond to bifurcation modes, and m to limit point mode. The symmetry of the
total potential energy is characterized by the third-order differential coefficients with

respect to qi . Consider the case, where m 1 bifurcation points are individually
symmetric (Thompson and Hunt, 1984) as
V,ijk = 0, (i, j, k = 1, . . . , m 1) (5.55)
which is satisfied in the following numerical examples where 1 , . . . , m1 are the
member buckling modes of a truss. Note that the orthogonality conditions > i j =
0 (i 6= j) for 1 , . . . , m1 are relaxed to satisfy Eq. (5.55). Since V is symmetric
in the direction of bifurcation modes, and asymmetric to the limit point mode, we
have
V,ijm 6= 0, (i, j = 1, . . . , m 1; i 6= j) (5.56a)
V,imm = 0, (i = 1, . . . , m 1) (5.56b)
Furthermore, from the definition of limit point,
V,mmm 6= 0 (5.57)
is satisfied (see Ohsaki and Ikeda (2007a) for details of stability analysis of elastic
conservative system exhibiting hilltop branching point with multiple bifurcation
points).
The incremental load factor c at the critical point of an imperfect system is
obtained as (Ohsaki and Ikeda, 2006)
V,m 1 p
c = V,mmm Cm || (5.58)
V,m V,m
where
m1
X m1
X
Cm = V,ijm qei qej (5.59)
i=1 j=1
and qei is obtained by solving the following m 1 simultaneous linear equations:
m1
X
V,ijm qej + V,i = 0, (i = 1, . . . , m 1) (5.60)
j=1
Note that the sign of the limit point mode cm is defined so that V,m > 0 is
satisfied, and V,mmm Cm > 0 holds. As is seen from Eq. (5.58), the critical load of
an imperfect system is a piecewise linear function of the imperfection parameter ,
which means that the structure is not imperfection-sensitive even though multiple
bifurcation points exist at the critical point of the perfect system.
5.7.4 Worst imperfection

The worst imperfection pattern that minimizes the critical load for a specified norm
of imperfection is derived below. Let B denote the matrix for which (j, i)-component
is given as 2 V /qj ai . Define hk = V,k for brevity. Then we obtain from Eq. (5.53)
n N
!
X X V
hk = di ckj
j=1
q j i=1
a i (5.61)
= c>
k Bd, (k = 1, . . . , m)
where ckj is the jth component of ck . In the following, the component of a vector
is denoted by a subscript as h = (hi ), q e = (eqi ).
The vector h is divided into the components hB = (h1 , . . . , hm1 )> correspond-
ing to the bifurcation modes and hL = hm to the limit point mode. The vector q e
eB = (e
is similarly divided into q q1 , . . . , qem1 )> and qeL = qem . Let F B denote the
(m 1) n matrix for which the ith row is equal to c> i . Then, from Eq. (5.61),
we obtain
hB = F B Bd (5.62)
The (m 1) (m 1) matrix G is defined so that the (i, j)-component is equal
to V,ijm . Then, from Eq. (5.60),
eB = G1 F B Bd
q (5.63)
is derived. Hence, from Eq. (5.59), we have
Cm = d> B > F B> (G1 )> F B Bd (5.64)
For simplicity, we write
e = V,mmm B > F B> (G1 )> F B B
G (5.65)
Then Eq. (5.58) is written as
c>
m Bd 1 p >e
c (d) = d Gd || (5.66)
V,m V,m
Finally, the anti-optimization problem for finding the worst imperfection pattern
that minimizes c is formulated as
minimize c (d) (5.67a)
N
X
subject to d2i = 1 (5.67b)
i=1
where Eq. (5.67b) is the norm constraint for d.

We next investigate the imperfection in the direction of active coordinates; i.e.,
the imperfection pattern vector d is defined as a linear combination of the critical
eigenmodes ci (i = 1, . . . , m) of the perfect system as
m
X
d= ci ci (5.68)
i=1
where ci is the coefficient, and the following orthogonality conditions are satisfied
between a bifurcation mode and the limit point mode:
c> c
i Bm = 0, (i = 1, . . . , m 1) (5.69)
The accuracy of this assumption is investigated in the numerical examples. Note
that N = n holds in this case.
Let D B denote the n (m 1) matrix for which the ith column is ci , and
divide c into the components cB = (c1 , . . . , cm1 )> and cL = cm , corresponding to
the bifurcation modes and the limit point mode, respectively. By defining
>
e DB
G = D B> G (5.70a)
em = c> c
m Bm (5.70b)
and using Eqs. (5.68) and (5.69), Eq. (5.66) is rewritten as
em cL 1 B> B
c (c) = c G c || (5.71)
V,m V,m
Therefore, c defined in Eq. (5.58) is conceived as a function of c, and the

anti-optimization problem for minimizing c is formulated as
minimize c (c) (5.72a)

Xm
subject to c2i = 1 (5.72b)
i=1
where Eq. (5.72b) is the norm constraint for c.

In the process of minimizing c (c), the right-hand-side of Eq. (5.71) is divided
into terms corresponding to cL and cB , and we first consider the maximization
problem of cB> G cB under the constraint
cB> cB = , (0 1) (5.73)
Since G is a real symmetric matrix, the minimum of the second term in Eq. (5.71)

is obtained as (1/V,m ) ||, where is the maximum eigenvalue of G . Hence,
using Eq. (5.72b), is obtained by minimizing c () as
em 1
c () = 1 || (5.74)
V,m V,m
By differentiating Eq. (5.74) with respect to , the value of corresponding to the

worst imperfection is obtained as

= p (5.75)
2
em +
Then the anti-optimal solution of Problem (5.72) is given as

r
e2
cL = 2 m , c B = 2
g (5.76)
em + em +
where g is the eigenmode corresponding to the maximum eigenvalue of G , which

is normalized by g > g = 1.
p
H
3 4 4 3
4 4 4 4 4 4
2 2
1 1 x
L L L L L L L L
y
Fig. 5.13 An arch-type truss.
5.7.5 Worst imperfection of an arch-type truss

Imperfection sensitivity analysis is carried out for an arch-type truss as shown in
Fig. 5.13, where L = 1000, H = 400, and the height of the center node is 200. A
vertical load p is applied at the center node. Youngs modulus is 1.0, for simplicity,
and p = 0.001. In the following, the units of length and force are omitted.
The numbers in Fig. 5.13 indicate group numbers; i.e., the members are classified
into four groups. Let Ai and Ii denote the cross-sectional area and the second
moment of inertia of the members in group i, which are related as
Ii = i2 Ai (5.77)
where i is a parameter that is independent of Ai .
Each member is divided into four beam elements. Note that the members are
pin-jointed at the nodes. The shape of the member is approximated as piecewise
linear when considering the imperfection in the direction of eigenmodes. The Green
strain is used, and the equilibrium paths are traced by the displacement increment
method. A symbolic computation software package Maple 9 (Maplesoft, 2004) is
used for differentiation of equations.
The relation between and the vertical displacement v at the center node
is shown in Fig. 5.14 for the design (A1 , A2 , A3 , A4 ) = (100, 100, 1000, 300) and
(1 , 2 , 3 , 4 ) = (18.15, 18.05, 30.0, 100.0). In this case, a limit point is reached at
= 4.7681. The five lowest eigenvalues are plotted against v in Fig. 5.15, where
the solid line and other lines correspond to the limit point mode and bifurcation
modes, respectively. Note that the chained line has duplicate eigenvalues. The five
modes c1 , . . . , c5 are shown in Fig. 5.16, where c1 , . . . , c4 correspond to member
buckling modes, and c5 to limit point mode. All the modes are normalized by
Eq. (5.49).
If we consider the imperfection in the direction of the sum of the five eigenmodes
P5
as d = i=1 ci , the relation between and c is written as
c = 1.8393 102 0.22176|| (5.78)
and c = c + c0 is plotted in Fig. 5.17. Note that + in Fig. 5.17 indicates the
maximum load obtained by the path-tracing analysis. The maximum value of the
Load factor
3
0
0 20 40 60 80 100
Displacement
Fig. 5.14 Relation between load factor and vertical displacement at the center node.
0.00015
duplicate
0.0001
Eigenvalues
0.00005
limit point
0
0 20 40 60 80 100
Displacement
Fig. 5.15 Relation between eigenvalues and vertical displacement at the center node.
nodal imperfection corresponding to = 1 is 0.5, which is very small compared

with the dimension of the arch. Therefore, from Fig. 5.17, the piecewise linear law
Eq. (5.78) has a good accuracy if the imperfection parameter is sufficiently small.
Next, we carry out path-tracing analysis to compute the critical loads corre-
sponding to random imperfection of nodal coordinates. The imperfection subject
to normal distribution N(0, ) is given for the (x, y)-coordinates of the nodes. The
similar imperfection with normal distribution is given for the initial deflection (out-
of-straightness) of each member as shown in Fig. 5.18. The standard deviation
of the
imperfection at the center node is , whereas the standard deviation is / 2 for the
interior nodes that are 1/4 of the member length form the two nodes, respectively.
Mode 1 c1
Mode 2 c2
Mode 3 c3
Mode 4 c4
Mode 5 c5 (Limit point mode)
Fig. 5.16 Four member buckling modes c1 , . . . , c4 and limit point mode c5 .
4.775
4.770
Maximum load
4.765
4.760
4.755
0.02 0.04 0 0.02 0.04
Imperfection parameter
Fig. 5.17 Relation between c and for imperfection in the direction of sum of five modes; solid
line: linear estimation using sensitivity coefficient, +: critical load found by path-tracing analysis.
The worst imperfection in the space of buckling modes, obtained by solving

Problem (5.72), is
c1 = 0.0, c2 = 0.93362, c3 = 0.0
(5.79)
c4 = 0.32002, c5 = 0.16108
which corresponds to the shape in Fig. 5.19. Since strict orthogonality is not sat-
P5
isfied in the modes ci (i = 1, . . . , 5), the norm of d = i=1 ci ci is 1.0478, which
is not equal to 1. However, the absolute value of c> c
i Bm (i = 1, . . . , m 1)
c> c
is less than 1/100000 of the absolute value of m Bm ; i.e., the assumption of
orthogonality in Eq. (5.69) is satisfied.
/ 2

/ 2
Fig. 5.18 Definition of column imperfection.
Fig. 5.19 Worst imperfection in the space of buckling modes as the anti-optimal solution of
Problem (5.72).
4.8
4.75
Maximum load
4.7
4.65
4.6
4.55
0 0.5 1 1.5 2
Norm of imperfection
Fig. 5.20 Relation between c and imperfection norm for small random imperfection = 0.1;
solid and dotted lines: linear estimation using Problem (5.67) and Problem (5.72), respectively,
+: critical load found by path-tracing analysis.
The worst objective value of Problem (5.72) is 0.11231, which is divided by the
norm 1.0478 to obtain the sensitivity 0.10719 with respect to the imperfection pa-
rameter . The estimation by linear approximation using this sensitivity coefficient
is plotted by dotted line in Fig. 5.20, where is the result of path-tracing analysis
for various norms of imperfections in the direction of the worst imperfection. The
+ mark shows the value of c for small random imperfection corresponding to
= 0.1. Similar results for moderately large imperfection = 1.0 are shown in
dotted line in Fig. 5.20. As is seen, the critical loads of imperfect systems can be
linearly approximated by sensitivity coefficients with good accuracy in the range of
4.8
Maximum load
4.6
4.4
4.2
4
0 5 10 15 20
Norm of imperfection
Fig. 5.21 Relation between c and imperfection norm for moderately large random imperfection
= 1.0; solid and dotted lines: linear estimation using Problem (5.67) and Problem (5.72),
respectively, +: critical load found by path-tracing analysis.
Fig. 5.22 Worst imperfection in the general direction as the anti-optimal solution of Prob-
lem (5.67).
small imperfection. The solution of Problem (5.72) gives a good lower bound for
imperfections defined by a moderately smooth shape as shown in Fig. 5.18.
On the other hand, the worst imperfection mode in the general direction ob-
tained by solving Problem (5.67) is as shown in Fig. 5.22, which is more irregular
than the mode in Fig. 5.19. Therefore, in order to incorporate it into practical
situations, the imperfection mode should be smoothed by restricting the shape as a
combination of possible imperfection shapes during fabrication and manufacturing.
The sensitivity coefficient with respect to the imperfection parameter correspond-
ing to this worst imperfection is 0.25822. The solid lines in Figs. 5.20 and 5.21
are the critical loads by linear approximation using the sensitivity coefficient. The
result of path-tracing analysis is indicated by . As is seen, the worst imperfec-
tion defines a conservative lower bound of the critical load for the specified norm of
smooth imperfection shown in Fig. 5.18.
If we express the worst imperfection mode of Problem (5.67) by ci (i = 1, . . . , 5),
the coefficients are obtained as 0.0, 0.41847, 0.0, 0.10663, 0.071310 for which the
norm is 0.43769, far less than 1. Therefore, the worst mode has components of the
passive coordinates, and we can conclude that the worst imperfection cannot be
obtained by considering the active coordinates only.
5.8 Some Further References
For other pertinent references on treatment of uncertainty in buckling, anti-

optimization in particular, the reader may consult with the monographs by El-
ishakoff, Li and Starnes (2001) and by Qiu and Wang (2008) as well as papers by
Adali (1992), Adali, Elishakoff, Richter and Verijenko (1994a), Adali, Lene, Du-
vaut and Chiaruttini (2003), Adali, Richter and Verijenko (1995b, 1997), Baitsch
and Hartmann (2006), Ben-Haim (1993a, 1993b, 1995), Ben-Haim and Elishakoff
(1989b, 1990), Cho and Rhee (2003, 2004), Chryssanthopoulos and Poggi (1995),
Conrado, de Faria and de Almeida (2005a, 2005b), de Faria (2000, 2002, 2004, 2007),
de Faria and de Almeida (2003a, 2003b, 2004), de Faria and Hansen (2001a, 2001b),
Deml and Wunderlich (1997), Elseifi (1998), Elseifi, G
urdal and Nikolaidis (1999),
Elishakoff (1998a, 2000b), Elishakoff and Ben-Haim (1990a), Elishakoff, Cai and
Starnes (1994a), Elishakoff, Li and Starnes (1994d, 1996), Elishakoff, van Manen,
Vermeulen and Arbocz (1987), Kim and Sin (2001), Kogiso, Shao and Murotsu
(1996), Lindberg (1992a,1992b), Pantelides (1996a, 1996b), Qiu, Elishakoff, and
Starnes (1996d), Qiu (2005), Qiu, Ma and Wang (2006a), Qiu and Wang (2005),
Reitinger, Bletzinger and Ramm (1994), Zharii, Jih-Hua, Chi-Ti, Li-Wei and Pan-
telides (1996), and Wang, Elishakoff, Qiu and Ma (2009).
Chapter 6
Anti-Optimization in Vibration
It was the best of times, it was the worst of times... (Charles Dickens,
A Tale of Two Cities)
... singular praise it is to have done the best things in the worst
times... (Inscription in Harold Church, Staunton, for Sir Robert
Shirley, Baronet)
In this chapter, we start, as in previous chapters, with a simple example. Then we

review Bulgakovs problem, which apparently was the first problem in the literature
on anti-optimization in vibration. Then we deal with the effect of scatter in material
properties on the dynamic response of viscoelastic beams by treating the aeroelastic
optimization and anti-optimization.
6.1 Introduction
The eigenvalues of vibration and responses to dynamic loads play important roles
in the design process of various fields of engineering. For eigenvalue problems, the
uncertainty in structural parameters can be introduced in a similar manner as the
static problems presented in Chap. 4. However, for the responses to dynamic loads
such as seismic excitations, the uncertainty of loads in time domain should be con-
sidered. Thus, the anti-optimization problem becomes very difficult to formulate.
Chen and Wu (2004a) applied interval parameters to anti-optimization consid-
ering dynamic responses, where the nonlinear responses are approximated by sensi-
tivity analysis. The beams under uncertain excitation was studied by Sadek, Sloss,
Adali, and Bruch (1993). Anti-optimization for earthquake excitation was studied
by Baratta, Elishakoff, Zuccaro and Shinozuka (1998) and Zuccaro, Elishakoff and
Baratta (1998).
Modares, Mullen and Muhanna (2006) presented an interval formulation for the
anti-optimization of eigenvalues of vibration, where linear perturbation by positive
semidefinite matrix representing the uncertainty in elastic modulus is considered and
145
1 M0 2 M0
K1 K2
Fig. 6.1 A 2-bar lumped mass structure.
the monotonicity of the eigenvalue is used. Interval matrix problem for eigenvalue
analysis was presented by Qiu, Chen and Elishakoff (1995a, 1996b) (see Chap. 7 for
basic properties of interval matrices).
6.2 A Simple Example of Anti-Optimization for Eigenvalue of Vi-

bration
Let K and M denote the stiffness matrix and mass matrix, respectively, of a
finite-dimensional structure. The eigenvalue problem of undamped free vibration is
formulated as
Ki = i M i , (i = 1, . . . , n) (6.1)
where i and i are the ith eigenvalue and eigenvector, respectively, and n is the
number of degrees of freedom. i is orthonormalized as
>
i M j = ij , (i, j = 1, . . . , n) (6.2)
where ij is the Kronecker delta. An anti-optimization problem of free vibration is
formulated to find the worst-case scenario for minimizing the lowest eigenvalue.
Consider a 2-bar structure as shown in Fig. 6.1 that has lumped mass M0 at
nodes 1 and 2. The extensional stiffnesses of members 1 and 2 are denoted by K1
and K2 , respectively. The stiffness matrix K and mass matrix M are defined as

K1 + K2 K2 M0 0
K= , M= (6.3)
K2 K2 0 M0
The lowest eigenvalue and eigenvector are obtained as
q
1
1 = K1 + 2K2 K12 + 4K22
2M0
p (6.4)
M0 K1 + K12 + 4K22
1 =
2K2
where
q
2 = 2K12 + 8K22 2K1 K12 + 4K22 (6.5)
The sensitivity coefficients of the eigenvalue can be computed generally using the
formulations in Sec. 3.6.3. However, for this simple example, the sensitivity coeffi-
cients c1 and c2 are obtained by directly differentiating 1 with respect to K1 and
Anti-Optimization in Vibration 147
K2 as
!
1 1 1 K1
c1 = = p >0
K1 M0 2 2 K12 + 4K22
! (6.6)
1 1 2K2
c2 = = 1 p 2 >0
K2 M0 K1 + 4K22
Therefore, as expected, 1 is a monotonically increasing function of K1 and K2 .
Consider uncertainty in the stiffnesses as
K1 = K10 + a1 , K2 = K20 + a2 (6.7)
where K10 and K20 are the nominal values and a1 and a2 are the uncertain param-
eters. If the bound of a = (a1 , a2 )> is given as follows by the quadratic inequality
constraint
a> a D (6.8)
which is a simple form of an ellipsoidal bound, then the anti-optimization problem
is formulated as
minimize 1 (a) (6.9a)
>
Since 1 is a nonlinear function of a1 and a2 , the anti-optimal solution cannot be
obtained explicitly by applying the Lagrange multiplier approach, as was the case
for the static response with load uncertainty in Sec. 4.1.
Let c = (c1 , c2 )> , and denote by c0 the value of c at (K1 , K2 ) = (K10 , K20 ).
Then 1 (a) is linearized at a = 0; i.e., at (K1 , K2 ) = (K10 , K20 ), to reformulate the
problem to a linearized problem:
minimize 1 (0) + c0> a (6.10a)
>
In a similar manner as the simple example of static anti-optimization problem in
Sec. 4.1, the anti-optimal solution can be obtained as
r
1 1 c> 2 c2
a = c2 , = = (6.11)
2 2 D
Note that = corresponds to the optimal solution that maximizes 1 .
Consider the case, for simplicity, where M0 = 1, K10 = 2, K20 = 1, and D = 1.
Then Problem (6.10) is explicitly written as
minimize 0.5858 + 0.1464a1 + 0.2928a2 (6.12a)
subject to a21 + a22 1 (6.12b)
>
and the anti-optimal solution is found to be a = (0.4472, 0.8944) with =
= 0.1637.
If 1 (a) is not linearized, and the bounds are given for a by the interval
aL a a U (6.13)
with the lower bound aL and upper bound aU , then the anti-optimization problem
has a nonlinear objective function and the linear bound constraints as
minimize 1 (a) (6.14a)
L U
Since 1 is a monotonically increasing nonlinear function of a1 and a2 , the anti-
optimal solution for Problem (6.14) exists at a vertex of the feasible region, and the
obvious anti-optimal solution is found as a = aL .
6.2.1 Anti-optimization for forced vibration

For forced vibration of a linear elastic system, the anti-optimization problem is
formulated against the time-history responses (Elishakoff and Pletner, 1991). Con-
sider, for simplicity, a single-degree-of-freedom system subjected to forced base ac-
celeration ub as
+ C u + Ku = M u
Mu b (6.15)
where u is the relative displacement from the base, M is the mass, C is the damping
coefficient, K is the stiffness, and ( ) is the time derivative. Suppose fb = M u
b
is expanded in Fourier series in time domain with N modes as
N
X
fb (t) = ai sin i t (6.16)
i=1
where ai is the Fourier coefficient for the vibration mode with circular frequency
i . Let i (t) denote the response corresponding to the ith mode that is an explicit
function of t. Then the response is written as a linear function of ai as
N
X
u(t) = ai i (t) (6.17)
i=1
Consider the case where the displacement, velocity, and acceleration of the base
excitation is linearly bounded with the specified upper bound denoted by ( )U as
| U
ub | u b | u U
b , |u
U
b , |ub | ub (6.18)
Then the constraints are written explicitly with respect to a, and the anti-
optimization problem for finding the worst excitation for maximizing |u(t)| at a
discretized time step, or maximizing the maximum value of |u(t)|, |u(t)| or |
u(t)|
during the specified time period, is formulated as a linear programming problem.
If the ellipsoidal bound is given for a = (a1 , . . . , aN )> , the problem turns out to be
a quadratic programming problem.
The bounds in Eq. (6.18) can be more generally written as time-varying func-
tions:
| U
ub | u b | u U
b (t), |u
U
b (t), |ub | ub (t) (6.19)
Michalopoulos and Riley (1972) investigated a two-degree-of-freedom system and
found that the worst excitation tends to be a bang-bang type function in control
theory.
A global measure can also be used for defining the bound of the input force
(Tzan and Pantelides, 1996a). For example, an energy bound can be given as
Z T
[fb (t)]2 dt E U (6.20)
0
where E U is the specified upper bound, and T is the duration of the motion.
6.3 Bulgakovs Problem
In his pioneering paper Bulgakov (1946) introduces his problem as follows:

Damped linear systems with constant parameters cannot develop
unlimited forced oscillations if the exciting actions remain finite.
Under unfavorable circumstances, however, the disturbances be-
come considerable. The purpose ... is to calculate the upper bounds
of the deviations when the disturbing actions y (t) ( = 1, . . . , r)
satisfy the conditions |y (t)| lp (lp constant) but are otherwise
arbitrary one-valued continuous functions of the time t possessing
as many derivatives as necessary; points of discontinuity of the first
kind are admitted, but their number in any finite interval must also
be finite.
Bulgakov (1946) derives the solution of some set of dynamic equations of motion
in the form
Xr Z tXr
d
xj (t) = [Sj y (t) j y (0)] + j ( ) y (t )d (6.21)
=1 0 =1 d
where y (t) ( = 1, . . . , r) are excitations, and the expressions for Sj and j ( )

depend on the particular system under consideration.
We refrain from reproducing the expressions for j (t), Sj and j . According
to the author, the expression under the sum sign
attains its maximum value mj (t) if the absolute magnitude of
y (t ) retains the value l , whereas its sign changes so as to
make all elements of the integral and the two other terms positive.
In other words, the sign of y (t ) must be altered for = t1 , t2 , . . .
and possibly for = 0, t, the ti s being the successive positive roots
of the equation j (t) = 0.
This equation is not reproduced here. Bulgakov (1946) writes the maximum
maximorum as the sum of the convergent series:
Z t1 Z t2
d d
mj () = Sj (lp ) j ( ) (lp )d + j ( ) (+lp )d +
0 d t1 d
(6.22)
The signs must be chosen so that both terms will be positive. Bulgakov (1946)
expresses mj () as mj () = Aj l , with

Aj = Sj j (0) 2j (t1 ) + 2j (t2 ) (6.23)
where Aj has a physical sense of influence coefficients. Finally, the maximum de-
viation in the response xj (t) caused by all disturbing actions combined is expressed
in the following form:
r
X
Mj () = Aj l (6.24)
=1
Special cases of this problem have been discussed in his book (Bulgakov, 1939)
and paper (Bulgakov, 1940). Bulgakovs solution is referred to in the Russian liter-
ature as the method of accumulation of disturbances. Bulgakovs approach was ex-
tended by Balandin (1993), Bulgakov and Kozovkov (1950), Gnoenskii (1961), Ko-
rneev and Svetlitskii (1995), Krichevskii and Ulanov (1973), Mironovkii and Slaev
(2002), Naishul and Svetlitskii (1956), Nguyen (1962, 1965), Roitenberg (1958),
Svetlitskii (1958, 1967, 2003), Svetlitskii and Prokhorov (1970), and Svetlitskii and
Stasenko (1967). Analogous solutions were also developed in the West, especially in
the theory of control (see, e.g., Michalopoulos and Riley (1972), Sage (1968)) and
is referred to as the bang-bang optimal control problem.
Later works, in the context of free and forced vibrations, are also associated
with evaluation of bounds albeit without uncertainty aspects. The reader may like
to consult, for example, works by Hu and Schiehlen (1996), Sharuz (1996), Sharuz
and Krishna (1996), and Schiehlen and Hu (1997).
6.4 Non-probabilistic, Convex-Theoretic Modeling of Scatter in

Material Properties
6.4.1 Introduction
Behavior, vibration, and stability of viscoelastic structures have been dealt with in
a number of monographs (Bland, 1960; Christensen, 1982; Fl ugge, 1975). In these
studies, material properties of the structure have been fixed at some deterministic
parameters. However, it is well known that the viscoelastic properties of structures
exhibit a large scatter (Koltunov, 1976). This scatter is usually accounted for by
considering the material properties as random variables. Cozzarelli and Huang
(1971) and Huang and Cozzarelli (1972) were apparently the first investigators
to include material uncertainty in their analyses. Hilton, Hsu and Kirby (1991)
extended the elastic-viscoelastic analogies to the stochastic case due to random
linear viscoelastic material properties. Both Gaussian and beta distributions were
considered for modeling the uncertainty in the data.
In probabilistic analyses, the needed probabilistic information for analysis was
postulated as given. For example, Huang and Cozzarelli (1973) utilized a log-normal
density or a truncated log-normal density. However, extensive experimental data
are needed to substantiate the probability densities with regards to the data.
More often, the necessary data are simply lacking, or only partial information is
available about the parameters. In these circumstances, the usefulness of the results
of probabilistic modeling based on incomplete data may be questionable.
Shinozuka (1987) studied the response variability in the stochastic context. Up-
per bound results were derived in two cases in which the spectral density function of
the stochastic field has assumed limiting shapes. The importance of such response
variability studies is immediately understood if one recognizes that it is rather dif-
ficult to estimate experimentally the autocorrelation function or, equivalently, the
spectral density function of the stochastic variation of material properties.
In this section, we present the non-probabilistic, convex modeling (Ben-Haim
and Elishakoff, 1990; Elishakoff and Ben-Haim, 1990a) for dealing with material
uncertainty and attendant response variability for viscoelastic structures. In par-
ticular, the problem of forced vibrations of viscoelastic beams is studied. First, the
analytical solution by Inman (1989) is generalized for a deterministic set of vari-
ables, describing material properties. Next, these variables are treated as varying
in a solid ball in the four-dimensional space, thus modeling the scatter in material
properties. The response uncertainty is directly related to the material uncertainty.
The least favorable response needed for the design of the structure is determined.
6.4.2 Basic equations for vibrating viscoelastic beam

This section follows closely the study by Elishakoff, Eliseeff and Glegg (1994b).
Transverse vibration of a viscoelastic beam is governed by the following differential
equation:
Z t
4 u(x, t) 4 u(x, )
E0 I + I g(t ) d
x4 0 x4
(6.25)
u(x, t) 2 u(x, t)
+c + = f (x, t)
t t2
where E0 is Youngs modulus, c is the viscous damping coefficient, I is the moment of
inertia, u(x, t) is the transverse displacement, which is a function of axial coordinate
x and time t, g(t ) is the viscoelastic kernel, is the mass per unit length of the
beam, and f (x, t) is the transverse excitation.
Equation (6.25) is supplemented by appropriate boundary and initial conditions.
We shall seek the solution of the problem through separation of variables. Normal
modes n of the beam with constant Youngs modulus E0 , uniform cross-section,

and uniform mass density satisfy the equation
4 n 2 n
+
E0 I =0 (6.26)
x4 t2
Letting u = n (x) exp(in t) with the circular frequency n = n2 , the equation for
n reads
d4 n (x)
= n4 n
E0 I (6.27)
dx4
The normal modes satisfy the orthogonality condition
Z l
n (x)m (x)dx = mn ,
0
Z l
(6.28)
d4 n (x)
E0 I m (x)dx = n4 mn
0 dx4
where mn is the Kronecker delta, and l is the length of the beam. The solution of
Eq. (6.25) is written in the following form:

X
u(x, t) = an (t)n (x) (6.29)
n=1
where an (t) is the modal response, which is a function of time t. We substitute

Eq. (6.29) into Eq. (6.25), multiply by m , and integrate with respect to x over the
interval (0, l) to yield
X
d2 a n dan
2 (n , m ) + c (n , m ) + E0 Ian (t)(IV
n , m )
n=1
dt dt
Z t (6.30)
+I g(t )an ( )(IV
n , m )d = (f, m )
0
where IV
is the fourth-order derivative of n with respect to x, and (, ) is the
n
inner product defined by
Z l
(, ) = (x)(x)dx (6.31)
0
Because of the orthogonality property (6.28), we are left with
Z t
d2 a m dam (IV
m , m ) (IV , m )
2 +c + E0 I am + I g(t )am ( ) m am d
dt dt (m , m ) 0 (m , m )
(f, m )
= , (m = 1, 2, . . . ) (6.32)
(m , m )
or, more specifically,
Z t
d2 a m c dam 4 4 1 1
2
+ + m am (t) + m g(t )am ( )d = fm (t),
dt dt 0 E0 (6.33)
g(t) = 0 for t < 0
where
(f, m )
fm (t) = (6.34)
(m , m )
We assume zero initial conditions. Taking the Laplace transform of Eq. (6.33)
yields

c 4 1
s2 + s + m 4
Am (s) + m G(s)Am (s) = Fm (s) (6.35)
E0
We now follow Golla and Hughes (1985) and Inman (1989) and take the following
analytical form for G(s):
s2 + s
G(s) = 2 (6.36)
s + s +
Some restrictions will be imposed later on the values of the parameters , ,
and . Now, G(s) can be represented as

s
G(s) = 1 + 2 2 (6.37)
s + s + s + s +
so that the positive kernel function g(t) is given as
aeat bebt

g(t) = (t) + e(1/2)t sinh Dt (6.38)
ab D
where (t) is the Dirac delta function, and
2
a = + D, b = D, D = (6.39)
2 2 4
In addition, we assume that the discriminant D is nonnegative. Substitution of
Eq. (6.36) into Eq. (6.35) results in
4

2 c 4 m s2 + s 1
s + s + m + 2
Am (s) = Fm (s) (6.40)
E0 s + s +
Equation (6.40) corresponds to an equation with the fourth degree of s; hence, it
can be written as two second-degree equations of the form

1
m1 m2 2 c1 c2 k1 k2 Am (s) F m (s)
s + s+ = (6.41)
m2 m3 c2 c3 k2 k3 Zm (s) 0
where Zm (s) is an artificial variable; for the general issue of artificial variables, one
may consult studies by Ahrens (1990), McTavish (1988), McTavish, Hughes, Soucy
and Graham (1992), and Ottl (1987). The matrices in Eq. (6.41) are supposed to be
symmetric to simplify the analysis. The artificial variable Zm (s) is chosen in such a
way that the transfer functions in Eqs. (6.40) and (6.41) are coincident. Equation
(6.40) reads
"
4 4
4 c 3 4 c m 2 c 4 m
s + + s + m + ++ s + + m + s
E0 E0
# (6.42)
4 1 2
+ m am (s) = Fm (s)(s + s + )

Equation (6.41) becomes

Fm (s)
(m1 s2 + c1 s + k1 )Am + (m2 s2 + c2 s + k2 )Zm = (6.43a)

(m2 s2 + c2 s + k2 )Am + (m3 s2 + c3 s + k3 )Zm =0 (6.43b)
To eliminate Zm (s) from Eqs. (6.43a) and (6.43b), we first obtain from Eq. (6.43b)
m2 s 2 + c 2 s + k 2
Zm (s) = Am (s) (6.44)
m3 s 2 + c 3 s + k 3
We substitute Eq. (6.44) into Eq. (6.43a) to yield
[(m1 m3 m22 )s4 + (m3 c1 + m1 c3 2m2 c2 )s3
+ (m1 k3 + m3 k1 + c1 c3 c22 2k2 m2 )s2 + (c1 k3 + c3 k1 2k2 c2 )s
(6.45)
1
+ (k1 k3 k22 )]Am (s) = (m3 s2 + c3 s + k3 )Fm (s)

Comparison of Eqs. (6.42) and (6.45) shows that it is sufficient to choose parameters
as
m1 = m3 = 1, m2 = 0, c3 = , k3 = (6.46)
In addition
4
c m
, k1 = c22 + m
c1 = 4
+ (6.47)
E0
We are left with two additional equations
4 4
m m
c22 2k2 c2 = ( )
, c22 k22 = (6.48)
E0 E0
We consider the specific case in which the parameters , and are dependent.
In particular, we assume that = , yielding
c2
k2 = (6.49)
2
Then Eq. (6.48) takes the form
2 4
m
c22 = (6.50)
4 E0
which suggests that we should require that 2 > 4 for c2 to be real. This, however,
is implied by Eq. (6.39). Hence, we put, as is done in Inman (1989),
2 = 8 (6.51)
which is in the allowable range of variation. Equation (6.50) then leads to
r
2
c2 = (6.52)
E0 m
Bearing in mind Eq. (6.49), we obtain
r
2 2 2 4

k2 = m , k 1 = 1 + m , c3 = = 2 2 (6.53)
E0 E0
The function G(s) in Eq. (6.36) becomes

(s2 + 2s 2)
G(s) = (6.54)
s2 + 2s 2 +
Effectively we are left with two parameters, and . Substitution of k1 , k2 , k3 , c1 ,
c2 , and c3 into Eq. (6.41) results in
r
c 2
m
1 0 a
m E0 a m
+ r
zm
0 1 zm 2
m 2 2
E0
r (6.55)
4 2 2
1
(1 + 2 E0 )m
E0 m
+ r am = fm (t)
2 2 zm
m 0
E0
which should be solved for the specific excitation.
6.4.3 Application to a simply supported beam

Consider now the beam that is simply supported at its ends. The mode shapes and
the eigenvalues are
r m 4 E I
2 mx 4 0
m = sin , m = (6.56)
l l l
Consider first an excitation that is spacewise uniform and timewise harmonic as
f (x, t) = q0 exp(it) (6.57)
We will confine ourselves to the steady-state solution. Via Eq. (6.34) we obtain
fm (t) = Rm exp(it) (6.58)
where
2lq0
Rm = [1 (1)m ] (6.59)
m
A steady-state solution of Eq. (6.55) is sought in the form
am (t) = am exp(it), zm (t) = z m exp(it) (6.60)
Substituting Eq. (6.60) into Eq. (6.55) yields
"
2 c 4
am = + i + 1 + 2 m
E0
r r ! #1
2 2 2 1
+ i + Pm Rm (6.61a)
E0 m E0 m
z m = P m am (6.61b)
r r !
1 2 2 2
Pm = + i + (6.61c)
2 2 2i E0 m E0 m
The steady-state response in the middle cross-section is

( )
X am
m it
U (l/2, t) = Re [1 (1) ]e (6.62)
m=1
2
where Re denotes the real part.
Consider now a case of a general nonharmonic excitation. We seek the comple-
mentary solution of Eq. (6.55) in the following form:
am (t) = E1 ert , zm (t) = E2 ert (6.63)
Substituting Eq. (6.63) into Eq. (6.55) and requiring that E1 and E2 do not vanish
simultaneously, we arrive at the following equation:

4
c 3 4 2c
r + 2 2 + r + + 1+ m + 2 r2
E0
(6.64)
c 4 4
+ + 2 2 1 + m r + m = 0
E0
We first study the behavior of the roots. It can be shown, through use of the
RouthHurwitz criterion, that the equation
r 4 + 3 r 3 + 2 r 2 + 1 r + 0 = 0 (6.65)
has roots with negative real parts, if and only if
1 2 3 > 21 + 0 23 (6.66)
It can be verified that this condition is satisfied. It should be noted that in-
equality (6.66) is also satisfied for the vanishing viscous damping. This implies
that the viscoelastic behavior is introducing an additional damping to the struc-
ture. Since the roots of the characteristic equation have negative real parts, the
complementary solution vanishes when t . We are interested in the behavior
of the displacement after sufficient time has elapsed from the start. In this case,
the complementary solution can be discarded. For the excitation
f (t) = bt (6.67)
Eq. (6.34) yields

b 2I
fm (t) = hm t, hm = [1 (1)m ] (6.68)
m
The particular solution of Eq. (6.55) is given in the following form:
am (t) = S1 t + S2 , zm (t) = G1 t + G2 (6.69)
Substitution of am (t) and zm (t) in Eq. (6.69) into Eq. (6.55) results in
r !
hm 2 0 c
S1 = 4 , S2 = 2f 4
+ 8 hm ,
m E 0 m m
v (6.70)
r u !
2 hm hm u
t 3 4 2c
G1 = 2
, G2 = 2 + + 4
E0 m m E 0 E0 m
The solution for am (t) can be written as

hm 1
am (t) = 4 t 4 ceq (6.71)
m m
where ceq is an equivalent viscous damping, given as follows, representing the com-
bined effect of both genuine viscous damping and additional damping introduced
by the viscoelastic behavior:
r
4
2 m
ceq = c + e
c, e
c = 2 (6.72)
E0
Note that, as expected, for = = and tending to zero, the equivalent
viscous damping ceq reduces to the genuine viscous damping c, since the contribution
due to viscoelasticity vanishes. The non-dimensional response reads
1
am (t) = am (t) = K1 (K2 T 1)
e (6.73)
re
p
where re = I/A is radius of inertia, T = m t is non-dimensional time, and
hm ceq 2
K1 = 8
, K2 = m (6.74)
m re ceq
Finally, we obtain

X
U (l/2, t) = am (t) sin(m) (6.75)
m=1
6.4.4 Least and most favorable responses

Hereinafter, we will apply the methodology of anti-optimization for modeling un-
certainty in material properties of viscoelastic structure. We discuss the realis-
tic situation, assuming that we possess only some limited information on material
parameters , , and ; namely, that they belong to some convex sets. The
displacement in either Eq. (6.62) or Eq. (6.75) is denoted as
U (l/2, t) = f (, , , ) (6.76)
We expand function f (, , , ) in Taylor series and take into account only the
linear terms as
f (, , , ) ' f (0 , 0 , 0 , 0 ) + > (6.77)
where
>
f f f f
= 1 , 1 , 1 , 1 at (, , , ) = (0 , 0 , 0 , 0 )

> (6.78)
0 0 0 0 >
= , , , = (1 , 2 , 3 , 4 )
1 1 1 1
and 0 , 0 , 0 and 0 are the average values of parameters. In addition, parameters
1 , 1 , 1 and 1 are introduced for dimensionless formulation. In particular, the
2 2
following values have been taken: 1 = 1 N/m , 1 = 1 s1 , 1 = 1 N/m s,
1 = 1 s2 . The scatter is modeled as the vector belonging to some set Z(P )
taken as a four-dimensional solid ball given by
Z(P ) = {| > P 2 } (6.79)
where P is the radius of the sphere. It should be noted that the derivation of radius
P assumes that the parameters , , and are of similar order of magnitude
and that the variations of these values are also of similar magnitude. On the other
hand, McTavish (1988) gives a numerical example (taken from Bagley and Torvik
(1983)) that demonstrates that these parameters have multiple orders of magnitude
difference. However, this may be simply remedied by choosing 1 , 1 , 1 and
1 such that an approximate distribution may be represented by a sphere in the
appropriately transformed coordinates. We are interested in the maximum possible
value the function f (, , , ) may take in this solid ball. This will result in the
least favorable response the system may experience when the parameters vary in
the solid ball defined by Eq. (6.79). We will also be interested in the minimum
value of f (, , , ). This will be identified with the best possible response.
Let 1 and 2 denote the maximum and minimum of the function f (, , , )
within the solid ball, given in Eq. (6.79):
1 (P ) = max f (, , , ), 2 (P ) = min f (, , , ) (6.80)

Z Z
Since Z(P ) is convex, the linear function f (, , , ) reaches its maximum and
minimum values on the boundary; i.e., on the set
B(P ) = {| > = P 2 } (6.81)
We define the Lagrangian as follows, in view of Eq. (6.77):
L(, ) = f (0 , 0 , 0 , 0 ) + > + ( > P 2 ) (6.82)
and demand that

L
= + 2 = 0 (6.83)

yielding
1
= (6.84)
2
However, in view of the equality > = P 2 , we arrive at
p
> P
= , = p (6.85)
2P >
where the plus sign is associated with the maximum and the minus sign with the
minimum.
Therefore, the final results for the least favorable response (denoted as LF R)
and for the most favorable response (indicated as M F R) are
p
LF R = 1 (P ) = f0 + P > (6.86a)
p
M F R = 2 (P ) = f0 P > (6.86b)
where f0 = f (0 , 0 , 0 , 0 ). Equation (6.76) can be put in the following form, in
conjunction with Eq. (6.58):
1
f (, , , ) = (eit + eit ) (6.87)
2
where asterisk denotes complex conjugate. The function , with term m = 1,
retained is
B1 2q0 2l
1 = , B1 = ( 2 + i) (6.88)
C1
where
(E0 )I 4 c
4
C1 = + + 2
L
(6.89)
E0 I 4 c (E0 )I 4 c 2
+ + i + +
L L
In the following, all the functions and their derivatives are evaluated at
(, , , ) = (0 , 0 , 0 , 0 ). The derivatives of are calculated as

1 4q0 2 I 4 2
= 2 ( + i) (6.90a)
C1 l

4q0 i 2 2 E0 I 4 ic
= C 1 ( + i) + + (6.90b)
C12 l

1 4q0 i 4
= 2 ( 2 + i) I (6.90c)
C1 l

4q0 E0 I 4 ic
= 2 C1 ( 2 + i) 2 + + (6.90d)
C1 l
Equations (6.86a) and (6.86b) become
!1/2
0 2 2 2 2
LF R = f0 P + + + (6.91a)

!1/2
2 2 2 2
M F R = f0 + P + + + (6.91b)

In the case of the timewise linearly increasing excitation given in Eq. (6.67), the
least favorable response as a function of non-dimensional time T reads, in view of
Eq. (6.73), as
!1/2
am 2 am 2
LF R(T ) = am (T ) + P + (6.92a)

!1/2
am 2 am 2
M F R(T ) = am (T ) P + (6.92b)

LFR LFR
MFR MFR
(a) (c)
LFR
LFR
MFR MFR
(b) (d)
Fig. 6.2 Dependence of the least favorable response (LF R) and most favorable response (M F R)
(a) (a) (b) (a)
upon uncertainty radius P : (a) a0 = 34.35109 N/m2 , 0 = 7.8125105 s2 ; (b) a0 = 2a0 ,
(b) (a) (c) (a) (c) (a) (d) (a) (d) (a)
0 = 0 ; (c) 0 = 0 , 0 = 20 ; and (d) 0 = 20 , 0 = 20 .
For numerical values hm = m , we have

am 2 2 m am 2 m
= , = 2 (6.93)
E0 E0
6.4.5 Numerical examples and discussion

Consider first the numerical results for the case of the harmonic excitation. Fig-
ure 6.2(a) portrays the dependence of the least and most favorable responses on
the uncertainty radius P . The parameters are fixed at the values defined by the
superscript ( )(a) as
(a) 2 (a)
a0 = 34.35 109 N/m , 0 = 7.8125 105 c2 , m = 1,
(a) (a)
E0 = 68.9 109 N/m2 , (a) = 27 kg/m, (1 )4 = 3.0688 105 , (6.94)
I (a) = 104 m4 , l(a) = 3 m
= T norm of f

Fig. 6.3 Absolute value = > of the norm of gradient of the function f (, , , ) as a
function of parameters and .
The value of the viscous damping is fixed at c(a) = 5983.2 kg/(m s), corre-
sponding to the coefficient = c/(21 ) = 0.2. As is seen in Fig. 6.2(a), the
amplitude of vibration for the nominal structure is 1.424 103 . For the beam
without viscoelastic effects, the response of the structure evaluated through the use
of the expression
4q0
Y2 = (6.95)
|EI(/l)4 2 + ic|
is 1.979103 m3/2 . This is 28% more than the nominal response of the viscoelastic
beam. The least favorable response linearly increases with P , whereas the most
favorable response linearly decreases with P . At P = 1000, for example, LF R =
1.293 103 m3/2 ; i.e., about 9% larger than the nominal response. Figure 6.2(b)
(b) (a)
is associated with the same data as in Fig. 6.2(a) except that now a0 = 20 .
(c) (a)
In Fig. 6.2(c), the value 0 = 0 , but the nominal value of is twice as much
(c) (a)
as in Fig. 6.2(a), namely, 0 = 20 . In Fig. 6.2(d), the data are the same as in
(d) (a) (d) (a)
Eq. (6.94) except 0 = 20 and 0 = 20 .
Figure 6.3 depicts the behavior p of the absolute value of the gradient of the
function f (, , , ), namely, = > defining the least and most favorable
responses in Eqs. (6.86a) and (6.86b). As is seen for large values of and small
values of , the values of can be significant, contributing to larger values of the
least favorable response. On the other hand, for materials with small values of
and relatively large values of , the effect of uncertainty will be less pronounced.
Figure 6.4 is associated with the response to timewise linearly increasing excita-
tion. It should be noted that the scatter in the responses stems from the constant
term in Eq. (6.71). Remarkably, the viscoelastic beam and its elastic counterpart
share the same slope in the asymptotic response, when t . In these circum-
LFR
MFR
Fig. 6.4 Response variability region in the beam subjected to timewise linearly increasing exci-
(3) (3)
tation (0 = E0 , 0 = 10 s2 ); response variability region is hatched.
LFR
MFR
(4) (3) (4) (3)
tation (0 = 20 , 0 = 20 ); response variability region is hatched.
stances, however, use of the nonlinear governing equations is necessary, since the
response becomes too large to justify employing the linearized analysis. The data
(3) (3)
in Fig. 6.4 are the same as in Fig. 6.2(a) except 0 = E0 and 0 = 10 s2 ; in
addition, the uncertainty radius is fixed at P = 2. The upper curve represents
the LF R, whereas the lower curve represents the M F R. The hatched area is a
region of the response variability (or uncertainty). Equation (6.71) in conjunction
with Eqs. (6.77) and (6.88) was utilized for obtaining the M F R and the LF R. It is
noteworthy that the contribution of the viscoelastic effects to the equivalent viscous
damping ceq in Eq. (6.72) through ec is much more important than that of the purely
viscous damping for the numerical data chosen in Fig. 6.4. Therefore, the time shift
predicted by the second term in Eq. (6.71) is non-negligible. In Fig. 6.5, the data
(4) (3) (4) (3)
were changed to 0 = 20 and 0 = 20 with the same value of uncertainty
radius. As is seen, the response variability is reduced in comparison with Fig. 6.3.
LFR
MFR
(5) (3) (5) (3)
tation (0 = 0 /2, 0 = 0 /2); response variability region is hatched.
(5) (3)
In Fig. 6.6, the data are the same as in Fig. 6.2(a) except that now 0 = 0 /2
(5) (3)
and 0 = 0 /2.
The response uncertainty region is now increased. This is understandable since
the relative measure of uncertainty (RM U ) could conveniently be defined as
P
RM U = (6.96)
Q
where P is the radius of input uncertainty, and Q is the measure of the nominal
location of the uncertain vector as
s
2 2 2 2
0 0 0 0
Q= + + + (6.97)
1 1 1 1
where 1 , 1 , 1 and 1 are defined in Eq. (6.78).
With increasing RM U , one would anticipate the increase in the response un-
certainty. This qualitative observation is in agreement with the obtained results.
Moreover, the method presented in this section provides a rigorous way to quantify
the response variability, if scant information on variability is available.
6.5 Anti-Optimization of Earthquake Excitation and Response
6.5.1 Introduction
There is a variety of literatures on modeling earthquake excitations. Modern anal-
ysis is based on the recognition that this excitation is an uncertain process. In an
overwhelming majority of the studies, the uncertainty is modeled as a random pro-
cess, either stationary or nonstationary with various approximations and attendant
models. The works by Drenick (1970, 1977b), Drenick, Novomestky and Bagchi
(1989), Drenick and Yun (1979), Elishakoff and Pletner (1991), and Shinozuka
(1970) have been the exceptions in the literature dedicated to uncertainty modeling
the earthquakes, in the sense that they utilized an alternative, non-probabilistic

avenue. Drenick (1970, 1977b) used a constraint on the total energy, which the
earthquake is likely to develop at the certain site, as a description of uncertainty.
He used the CauchySchwarz inequality to determine the maximum response of the
system to such an excitation. In the opinion of several investigators such a bound
was too conservative (Drenick, 1993). Shinozuka (1970) has suggested characteriz-
ing the earthquake uncertainty by specifying an envelope of the Fourier amplitude
spectrum. Numerical calculations have demonstrated that the maximum response of
the structure predicted by this method is less than that predicted in Drenick (1970,
1977b), Drenick, Novomestky and Bagchi (1989) and Drenick and Yun (1979). El-
ishakoff and Pletner (1991) investigated the modification of the response prediction
when the global information on the excitation is increased. In particular, the max-
imum possible response, which the structure may develop, was evaluated under the
assumption that only the bound on base acceleration is known; then the maximum
response was modified under the assumption that, in addition to the base acceler-
ation bound, the bounds on base velocity and/or displacement were specified. In
this section, we resort to ellipsoidal modeling of earthquake excitation.
6.5.2 Formulation of the problem

Consider a single-degree-of-freedom linear system subjected to earthquake excita-
tion. The motion is governed by
x + cx + kx = m
m xb (6.98)
where x(t) is the relative displacement of the structure from the base, c is the
damping coefficient, k is the stiffness, m is the mass, and x b (t) is the acceleration of
the earthquake excitation. If the excitation force m xb (t) is known, the response
is given by the Duhamel integral
Z t
x(t) = [mxb ( )]h(t )d (6.99)
0
Here h(t) is the impulse response function given as
1
h(t) = exp(0 t) sin d t (6.100a)
md
p
d = 0 1 2 (6.100b)
p p
where 0 = k/m is the natural circular frequency, and = c/(2 k/m) is the
damping ratio. Let us represent the excitation as a series in terms of complete and
orthogonal deterministic set of functions {i (t)}:
X
x
b (t) = Ai i (t) (6.101)
i=1
where Ai is the coefficient, and {i (t)} satisfies the orthogonality condition
Z T
i (t)j (t)dt = 0 for i 6= j (6.102)
0
in interval (0, T ), where T is the duration of the earthquake.

If the excitation is known, the coefficients Ai are readily obtainable as follows, by
multiplying both sides of Eq. (6.101) by j (t), integrating the result in the interval
(0, T ), and using the orthogonality property (6.102):
Z T
1
Ai = 2 x
b (t)i (t)dt (6.103)
Vi 0
where
Z T
Vi2 = [i (t)]2 dt (6.104)
0
The response, given in Eq. (6.99), is rewritten as
X
x(t) = Ai i (t) (6.105)
i=1
where
Z T
i (t) = m i (t)h(t )d (6.106)
0
Let us suppose now that we possess some fragmentary information at a given
site, namely, we assume that the historical data is available on the accelerograms
(1) (m)
x
b , . . . , x
b , where the superscript denotes the serial number of the earthquake,
and m is the total number of accelerograms. Then, using the decomposition of type
(6.101) for each of the earthquake realization
X
(k) (k)
x
b (t) = Ai i (t), (k = 1, . . . , m) (6.107)
i=1
we arrive at m vectors of the excitation parameters
(1) (1)
A(1) = (A1 , . . . , AN )> ,
(2) (2)
A(2) = (A1 , . . . , AN )> ,
.. (6.108)
.
(m) (m)
A(m) = (A1 , . . . , AN )>
where N is the number of most significant terms in the series (6.101) and (6.105),
which are rewritten as
N
X N
X
x
b (t) = Ai i (t), x(t) = Ai i (t) (6.109)
i=1 i=1
Thus we replace m accelerograms by m points in an N -dimensional space. To this
end, to use the ellipsoidal modeling of these point-images of original accelerograms,
we model these points as belonging to an N -dimensional ellipsoid
A> W A 2 (6.110)
where W is a symmetric positive-definite matrix defining the shape of the ellipsoid,
and is a positive constant defining the size of the ellipsoid. The way of obtaining
the matrix W and constant will be discussed in Sec. 6.5.4.
6.5.3 Maximum structural response

We are interested in determining the maximum response of the structure at time
instant t. Mathematically the problem reads as follows:
N
X
maximize Ai i (t) = A> (t) (6.111a)
i=1
subject to A> W A 2 (6.111b)

We define the Lagrangian
L(t, ) = A> (t) + (2 A> W A) (6.112)
From the stationary conditions of L with respect to A and , we obtain
q
W 1 > W 1
Aworst = q , = (6.113)
2
> W 1
The worst response is obtained by substituting Eq. (6.113) into Eq. (6.101):
q
xworst (t) = A>
worst = > W 1 (6.114)
Once the basic excitation functions i (t) are chosen, the basic response function
i (t) are readily obtained from Eq. (6.106). Then, having the information on matrix
W and constant from the accelerograms, we find the maximum response by
employing Eq. (6.114).
6.5.4 Ellipsoidal modeling of data

6.5.4.1 Basic ideas
Consider any collection of data to be processed by the procedure in Sec. 6.5.3.
Assume that one record consists of N numerical parameters, so that any observation
j of the phenomena is fully described by a point Pj in N -dimensional Euclidean
vector space EN .
The problem treated in this section is to find the smallest (in some sense) ellip-
soid containing all the observed data; i.e., to set the matrix W in Eq. (6.110), and
the shift of the origin to the center of the ellipsoid. This problem is rather hard to
solve, if the ellipsoid width defining the minimum volume is required, the approach
leads to cumbersome optimizations that may be impractical. So the problem is
treated aiming at a simple, as far as possible, solution.
In this view, and for simplicity of graphic representation only, let us consider the
three-dimensional case as shown in Fig. 6.7; i.e., a collection of m points P1 , . . . , Pm
that are the recorded observations of the phenomenon to be included in the ellipsoid,
and let

x11 x21 xm1

x1 = ... , x2 = ... , . . . , xm = ... (6.115)
x1N x2N xmN
'
''
Fig. 6.7 Collection of m = 7 points representing the recorded observations, to be included in the
N = 3 dimensional ellipsoid.
be the coordinate vectors of the points, collecting any relevant parameters to identify
the characters of the phenomenon at the observation Pi in the reference frame with
origin O, which is given, e.g., as follows for N = 3:

1 0 0
(e1 , e2 , e3 ) with e1 = 0 , e2 = 1 , e3 = 0 (6.116)
0 0 1
The proposed procedure is conducted basically in two phases:
(A) Find the smallest parallelepiped P in EN containing all homothetic points Pi

(i = 1, . . . , m).
(B) Find the smallest ellipsoid in EN , also containing all points Pi (i = 1, . . . , m),
searching in the family of ellipsoids E that are homothetic to the ellipsoid E1
inscribed in P having the same center as P and a principal axis parallel to the
sides of P.
The second phase (B) is straightforward, as it will be shown later. Phase (A) is
founded on the solution of two auxiliary problems as follows.
6.5.4.2 Preliminary statements

Consider the following auxiliary problems:
Problem I : Given the set of m points P1 , . . . , Pm in EN , find the couple of parallel
hyperplanes, as shown in Fig. 6.8 for three-dimensional case, containing all the
points and having the minimum distance from each other. Let 0 and 00 be any
two parallel hyperplanes in EN with being the unit vector orthogonal to both; let
0 and 00 be the orthogonal distances of 0 and 00 , respectively, from the point
O that is assumed to be the origin of the reference frame as shown in Fig. 6.8 when
0 > 0 and 00 < 0. Any point Pi is internal to the strip included by 0 and 00 if
'
''
Fig. 6.8 Two parallel planes 0 and 00 ; is the unit vector orthogonal to both.
and only if
x> 0 >
i , xi
00
(6.117)
are satisfied simultaneously. By applying Eq. (6.117) for i = 1, . . . , m, one obtains
that the strip includes all points if and only if
X > 0 1, X > 00 1 (6.118)
where
>
1 x1
.. ..
1 = . , X = . (6.119)
1 x>
m
The total width of the strip is given by 0 00 . Therefore, the first problem denoted
by Problem I turns out to be
find SN 1 , 0 and 00 that minimize 0 00 (6.120a)
subject to X > 0 1 (6.120b)
> 00
X 1 (6.120c)
with SN 1 being the unit sphere in EN .
This problem is approached in two steps:
Step I.1: For the given SN 1
find ( 0 , 00 ) that minimize 0 00 (6.121a)
> 0
subject to X 1 (6.121b)
X > 00 1 (6.121c)
It is trivially given as
0 = max x> >
i = xr ,
00
= min x> >
i = xs (6.122)
i i
Step I.2: In order to find the smallest strip containing all points, consider the
following problem:
find 1 that minimizes 01 001 (6.123a)
> 01
subject to X 1 1 (6.123b)
> 001
X 1 1 (6.123c)
>
1 1 =1 (6.123d)
where 01 and 001 are implicitly related to 1 by Eq. (6.122). The above
problem is a nonlinear programming problem, mainly because of the constraint
>1 1 = 1.
Any constrained search procedure can be applied to find the optimal vector
1 . The simplest, and also effective, as tested in this investigation, is a random
search. The only difficulty lies in generating unit random vectors 1 s satisfying
the constraint (6.123d). The following procedure is suggested. Let be a generic
unit random vector and 1 the unit vector that minimizes 01 001 . Let ni
and si (i = 1, . . . , N ) be any random numbers in [0, 1], i = 2ni the random
angle in [0, 2] of and SGi the sign of the ith component of with

SGi = +1 if si 0.5
(6.124)
SGi = 1 if si > 0.5
The components of are chosen through the following steps:
(a) RN = 1,
(b) RN i = sin2 (p
i )RN i+1 (i = 1, . . . , N 1) and i = 2ni ,
(c) i+1 = SGi+1 Ri+1 Ri (i = N 1, N 2, . . . , 1).
Problem I is thus solved after Steps I.1 and I.2.
Problem II : Suppose 1 is found, and the first minimal strip is obtained, with width
10 100 = d1 . Let 1 , . . . , k (k < N ) be given unit vectors, mutually orthogonal,
i.e., >i j = ij , and consider the following Problem II:
find k+1 that minimize 0k+1 00k+1 (6.125a)

> 0k+1
subject to X k+1 1 (6.125b)
X > k+1 00k+1 1 (6.125c)
>
k+1 k+1 = 1 (6.125d)
>
i k+1 = 0, (i = 1, . . . , k) (6.125e)
Problem II has k constraints in addition to those in Problem I, and it can be
regarded as a generalization of Problem I, by making reference to SN k , which is
the vector subspace of EN orthogonal to 1 , . . . , k .
To generate SN k , one has to infer an orthogonal unit minimal base, say
( k+1
1 , . . . , k+1
N k ), of SN k . To this aim one can take recourse to the GramSchmidt
orthogonalization procedure as follows.
Let k1 , . . . , kN k+1 be the minimal orthogonal base of SN k+1 so that both

sets of unit vectors

> 1 > 1
.. ..
. .

> >
k1 k1
k k > k+1 >
B = ( 1 ) , B = k (6.126)
.. k+1 >
(1 )
.
..
(kN k )> .
( kN k+1 )> ( k+1
N k ) >
are orthogonal unit bases for the entire EN , where the first is a known base and the
second is known up to k . Note that
>
e1
1 ..
B = . (6.127)
e>
N
One has to infer B k+1 from B k . Let 0 = k , and in accordance with the Gram
Schmidt orthogonalization method, one sets
1 = 11 k + k2
2 = 12 k + 22 1 + k3
3 = 13 k + 23 1 + 33 2 + k4 (6.128)
..
.
N k = 1N k k + 2N k 1 + 3N k 2 + + kN k+1
where, by orthogonalization
11 = (k2 )> k
(k )>
12 = (k3 )> k , 22 = (3 )> 1
1 1
(k >
4 ) 1 ( k )>
13 = (k4 )> k , 23 = ( 1 )> 1
, 33 = ( 4 )> 2
2 2
.. ..
. .
(k >
N k+1 ) 1 (k >
N k+1 ) 2
1N k = ( kN k+1 )> k , 2N k = ( 1 )> 1
, 3N k = ( 2 )> 2
,
N k (k )>
N k = (
N k+1
)>
N k+1
N k+1 N k+1
(6.129)
After the orthogonal vectors i have been found, one can normalize k+1
i as
k+1 i
i = , (i = 1, . . . , N k) (6.130)
| i |
Note that each i for i > 0, and consequently each k+1 i , is orthogonal to j
(i = 1, . . . , k). In fact, every i turns out to be a linear combination of vector k
and vectors ki . All these vectors are assumed to have been built up orthogonal
to 1 , . . . , k in the previous step. Hence, one can write for suitable aij
NX
k+1
i = ai0 k + air kr (6.131)
r=1
and
NX
k+1
> >
i j = ai0 k j + air kr j = 0 (6.132)
r=1
Therefore, the problem (6.125) is reduced to

find k+1 SN k that minimizes 0k+1 00k+1 (6.133a)
> 0k+1
subject to X k+1 1 (6.133b)
>
X 1 00k+1 1 (6.133c)
>
k+1 k+1 =1 (6.133d)
where k+1 SN k means
N
X k
k+1 = ik k+1
i (6.134)
i=1
with
N
X k N
X k
(k+1 )> k+1 = ik rk ( k+1
i )> k+1
r (6.135)
i=1 r=1
where, by orthonormality of s,
N
X k
(k+1 )> k+1 = (ik )2 (6.136)
i=1
With the above relations, Problem II is reformulated as

find k SN k that minimizes 0k+1 00k+1 (6.137a)
N
X k
subject to k+1 = ik k+1
i (6.137b)
i=1
X > k+1 0k+1 1 (6.137c)

> 00k+1
X 1 1 (6.137d)
k> k
=1 (6.137e)
Thus, one recognizes that Problem II is reduced to Problem I in the vector space
EN k .
Then phase (A) of the main problem is articulated in N steps as follows:
Step A.1: Solve Problem I, and find 1 , 01 , 001 . After this step a first strip
between hyperplanes 0 and 00 is determined.
Step A.2: Solve Problem II with k = 2 in the form (6.137), and find k+1 , 0k+1 ,
00k+1 .
Step A.3,. . . , A.N : Iterate step A.2 for k = 3, . . . , N . After Step A.N, N cou-
ples of hyperplanes are individuated; each couple defines a minimal strip in EN
containing all m points.
By iterating these N steps, the minimal parallelepiped is found.
6.5.4.3 Search of the smallest ellipsoid containing all points

The smallest ellipsoid is assumed to have the same centroid as P and principal
axis parallel to the edges of P which are identified in the unit vectors 1 , . . . , N .
The diametral lengths are assumed to be proportional, respectively, to d1 = 01
001 , . . . , dN = 0N 00N .
Let x0 be the position vector of the centroid of P, and shift the origin of the
reference system to x0 . Thus, the data points are identified in the new reference
frame by
Ai = xi x0 , (i = 1, . . . , m) (6.138)
so that the ellipsoid is identified by
A> W A 2 (6.139)
In order to identify matrix W , let R be the rotation matrix yielding
>
i
..
RB 1 = . = B N (6.140)
>
N
In the frame of reference B N , the equation of the ellipsoid is given by
Xm
zi2 > 2
d 2 = Z QZ (6.141)
i=1 i
where zi are the coordinates of any point with respect of B N and
Q = diag(1/d21 , . . . , 1/d2N ) (6.142)
In the reference frame B 1 , one has the coordinates A of any point P related to the
corresponding coordinates in B N
P = B> > > >
1 A = BN Z = B1 R Z (6.143)
whence
A = R> Z, Z = RA (6.144)
Therefore, the equation of the ellipsoid is
Z > QZ = A> R> QRA 2 (6.145)
Table 6.1 Coordinates of points Pi (i = 1, . . . , 14) to be enclosed by a three-dimensional

ellipsoid.
P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 P13 P14
x1 100 300 700 400 200 400 0 100 300 700 400 200 750 1380
x2 500 550 450 470 500 500 500 200 500 800 600 800 1370 250
x3 800 800 600 600 400 400 200 200 100 100 900 900 1600 260
whence, in B 1 (having again shifted back to x0 )

A> W A 2 (6.146)
where W = R> QR. Thus, the matrix W has been produced.
Now the minimal value for 2 can be found by the sequence of unidimensional
search
2 = max A>
i W Ai (6.147)
i=1,...,m
Note that the above minimal ellipsoid may not be the absolute minimum, since it
is minimal under the assumption that it is related to the minimal parallelepiped.
The ellipsoid so found is in general included between the ellipsoid contained in P
and the one containing P.
6.5.4.4 Numerical application

A numerical application of the procedure proposed above is performed. In order to
show the results graphically, the example is carried out for N = 3.
Let us consider a set of points Pi (i = 1, . . . , m = 14) with the coordinates given
in Table 6.1. The unit vectors 1 , 2 , 3 in E3 , orthogonal to the couples of planes
that define the minimal strips containing all points P , are

0.205 0.824 0.527
1 = 0.670 , 2 = 0.510 , 3 = 0.538 (6.148)
0.713 0.242 0.657
The minimal distances between the planes are
d1 = 647, d2 = 981, d3 = 3240 (6.149)
Figure 6.9 shows the minimal parallelepiped P, the ellipsoid E1 contained in P,
and the principal planes of the ellipsoid. Figure 6.10 shows the sectional view of
the three principal planes. Figure 6.11 shows the ellipsoid E found by enlarging
homothetically E1 in order to include all points.
In conclusion, we would like to note that, Elishakoff, Wang and Qiu (2008)
and Wang, Elishakoff and Qiu (2008a) recently addressed the following question:
Which analysis is preferable, ellipsoidal modeling or interval analysis? They showed
that the type of analytical treatment that should be adopted for non-probabilistic
analysis of uncertainty depends on the available experimental data. The main idea is
Fig. 6.9 Minimal parallelepiped, ellipsoid contained in it, and principal planes of the ellipsoid.
Fig. 6.10 Sectional views of three principal planes of the ellipsoid.
based on the consideration that the maximum structural response predicted by the
preferred method ought to be minimal. Prior to analytical treatment, the existing
data ought to be enclosed by the minimum-volume hyper-cube of volume V1 that
contains all experimental data. The available data also have to be enclosed by the
minimum-volume ellipsoid V2 . If the response R(V1 ) based on V1 is smaller than the
response R(V2 ) based on V2 , then one has to prefer interval analysis. If, however,
R(V1 ) is in excess of R(V2 ), one is recommended to utilize ellipsoidal analysis. If
R(V1 ) and R(V2 ) equal each other, or are in close vicinity, then two approaches
are equally valid. For further details and examples of data for 7-bar planar truss
Fig. 6.11 Ellipsoid obtained by homothetic enhancement of ellipsoid 1 to include all points
of observation.
structure, or for 60-bar space truss structure, one is referred to the above papers.
6.6 A Generalization of the DrenickShinozuka Model for Bounds

on the Seismic Response
6.6.1 Preliminary comments

This section closely follows the article by Baratta, Elishakoff, Zuccaro and Shinozuka
(1998). It aims at establishing basic criteria for treatment of uncertainty in seismic
excitation through anti-optimization, with the purpose of setting up convex models
for forecasting the maximum possible response parameters of an SDOF structure
in the seismic environment.
Consider the simple linear SDOF shear frame under the action of a ground
acceleration a(t). The equation of motion reads
+ 20 u + 02 u = a(t)
u (6.150)
where u(t) is the relative displacement from the ground, a(t) is the ground accel-
eration, 0 is the natural circular frequency of the undamped system, and is the
damping ratio.
The initial conditions read
u(0) = 0, u(0)
=0 (6.151)
The solution, as well known, is given by
Z t
u(t) = a( )h(t )d (6.152)
0
where h(t) is the impulse response function defined as

1 exp(0 t) sin(d ) for t 0
h(t) = d (6.153)

0 for t < 0
p
where d = 0 1 2 . The basic idea, first suggested by Drenick (1970), is to
find the base accelerogram function yielding the maximum possible value of the
response u(t). Following Drenick, assume that a(t) has a finite duration, say T , and
a bounded energy E0 as
Z T
a2 (t)dt = kak2 = E02 (6.154)
0
Let P(E0 ) be the set of functions defined in interval (0, T ) and possessing the given
energy E02 . Then the problem reduces to
maximize u(t) = kukT = E0 HT , t (0, T ) (6.155a)
subject to a P(E0 ) (6.155b)
The following relation was proved by Drenick:
Z T Z T
HT2 = max h2 (t )d = h2 (t0 )d, t (0, T ) (6.156)
0 0
As T , the base accelerogram ac (t) yielding the maximum peak response, and
coined by Drenick (1973) in a later paper as the critical excitation, has a shape
coincident with the impulse response function reversed with respect to time, and is
given by
E0
ac (t) = h(t) (6.157)
H
The maximum peak response is given by
kuk = uc (0) = E0 H (6.158)
where uc (t) denotes the response due to acceleration ac (t). Some criticism was
expressed by Drenick (1973) himself, mainly about the circumstance that aseismic
design based on critical excitation could be ... far too pessimistic to be practical.
Yet Drenick (1973) investigated the ratio of the maximum response due to criti-
cal excitation, to the maximum response of actually recorded seismic accelerograms.
Response spectra were derived on the basis of both the critical excitation and a num-
ber of accelerograms from the set of recorded earthquakes, with attendant compar-
ison of results. It was shown that for the range of structures with natural period
from 0.5 to 1.2 sec and damping ratio in the range of 0.02 and 0.10, the ratio of
the critical to the experimental response was almost invariably in the neighborhood
of 2. By extending the class of possible accelerograms from only the recorded ones
to linear combinations of these, the above ratio decreases to the range of values
between 1.3 and 1.6.
In later papers, Baratta (1980, 1981) attempted to develop a modified proce-

dure to obtain the least favorable excitation, searching in the set produced by a
generator of artificial earthquake accelerograms. The model chosen for such a gen-
erator followed the procedure of Ruiz and Penzien (1969). The class of admissible
seismograms was constrained by a specified value of the maximum peak ground
acceleration (PGA). The generators parameters were calibrated on the basis of
four available accelerograms recorded at Tolmezzo (northeastern Italy) during the
earthquake of May 1976 (namely the earthquakes of 69 May 1976), so that the
class of admissible accelerograms was compatible with these samples. The compar-
ison of the least-favorable responses of the undamped structure with the envelope
of responses obtained through the processing of four recorded accelerograms, also
yielded over-shooting values of about 2 in the range of vibrational period of about
0.5 sec of the structure, to the value of 7 for more deformable structures.
The result appears to confirm that at least Drenicks critical response, if not the
critical excitation, should be in the neighborhood of some realizable excitation dur-
ing an earthquake, if the behavior of the structure is in the elastic range. Analogous
results were confirmed by Baratta and Zuccaro (1992) by re-evaluation of the same
seismic model. In this case, the analysis was carried out for 5% damping, and the
overshooting ratio turned out to be around 2 or 2.5, in a much wider range around
T = 0.5 sec. The extension of this range was probably due to the influence of damp-
ing. Again, Drenicks results may well be confirmed: Earthquake-type functions of
a given site may give responses that are comparable to those by critical excitation.
It must also be stressed that, by introducing the least-favorable earthquake in a
full probabilistic analysis of seismic hazard, the severity of the approach is quite
mitigated by the combination with uncertainties deriving from regional seismicity
(Baratta and Zuccaro, 1992). It is possible that inelastic excursions of the structure
can further reduce the gap between worst excitations and recorded ones.
From the above considerations, one can conclude that the original idea of crit-
ical excitation is worth pursuing further. This is why the study in this section is
presented to set up a convex model to predict the worst possible shapes of accelero-
grams acting on linear structures, following the early idea by Shinozuka (1970), that
some bound on spectral ordinates can be used to improve a priori knowledge of the
excitation. In this section, we modify and combine the approaches of Drenick (1970)
and Shinozuka (1970). We are interested in the maximum stress the structure can
attain by earthquake action at a given site, and assume that the structural response
does not exceed the appropriate linear threshold. Finally, we assume that the base
accelerogram belongs to a set of possible accelerograms, whose basis is composed
of accelerograms ai (t) (i = 1, 2, . . . ), so that any possible accelerogram a(t) can be
expanded in the linear series of the base accelerograms as
X
a(t) = ci ai (t) (6.159)
i=1
Due to practical numerical reasons, this series should be truncated to N , where N
is the number of retained terms. One simple example of such a space is the Fourier
expansion. As a result, every accelerogram can be represented by a linear combi-
nation of harmonic functions. Let us assume that the duration of the earthquake
is known, and that it is possible (from geophysics, for instance) to find a bound for
the energy of the excitation, as in Drenicks problem
Z T
a2 (t)dt E02 (6.160)
0
If the base functions ai (t) are orthogonal and all scaled to the same energy, every
function of the type (6.159) satisfying the constraint (6.160) will possess coefficients
ci satisfying the following inequality:
N
X
c2i 1 (6.161)
i=1
Any possible function that satisfies Eq. (6.160) is referred to as an admissible ac-
celerogram. Thus, admissible functions can be mapped in the Euclidean space of
N -dimensional vector c = (c1 , . . . , cN )> . Admissible accelerograms are contained
in the unit sphere S0 centered at the origin.
The values of the maximum and minimum displacements of the structure are
defined for a(t) varying within S0 . Let ui (t) denote the response function under
ai (t). The response under any a(t) expressed by Eq. (6.159) are given as follows
due to the linearity of system:
N
X
u(t) = ci ui (t) (6.162)
i=1
Thus, with matrix notation, the problem is set as follows for any fixed instant t in
(0, T ):
find max c> u(t) (6.163a)
c
>
subject to c c 1 (6.163b)
>
where u(t) = (u1 (t), . . . , uN (t)) . Since Problem (6.163) is to maximize a linear
function within a convex domain, the anti-optimal solution exists on the boundary
of S0 , and the problem is reformulated as
c
>
subject to c c = 1 (6.164b)
which differs from Problem (6.163) in that the inequality in Problem (6.163) is
replaced by an equality. Let us now introduce notations
r(c) = c> u, f (c) = c> c 1 (6.165)
At the solution point, with 0,
f = r for c> u max. (6.166)
N=200
N=100
N=50
N=25
t
Fig. 6.12 Influence of N and the number of harmonics on the bound (6.170).
and
f = r for c> u min. (6.167)
where is the Lagrange multiplier (see Sec. 2.4 for details).
Hence, with Eq. (6.164b), the optimal and anti-optimal solutions for the fixed
time instance t is given as
c(t) = (t)u(t) (6.168)
with
1
2 (t) = (6.169)
u(t)> u(t)
The maximum displacement at any instant t in (0, T ), is finally given by
q
umax (t) = umin (t) = u(t)> u(t) (6.170)
A numerical example has been carried out by assuming that admissible functions
are continuous. All ai (t)s were taken as harmonic functions, as in the Fourier
expansion:

A0 sin(i t) for i m
ai (t) =
A0 cos(i t) for m < i N
(6.171)
2i 2 2 2
i = , A0 = E0
T T
with N even, and m = N/2.
In Fig. 6.12, the bound (6.170) is plotted for different values of m = 25, 50,
100, 200 for t ranging from 0 to 4 sec. The figure shows how the refinement of
the functional space of the excitation makes the bounds increase progressively, and
a convergence is attained for m = 100. Figure 6.13 depicts a sample maximizing
function for T = 30 and m = 100. These bounds are of the type of Drenicks critical
excitation. It appears to be highly questionable that such functions can be accepted
as credible accelerograms.
u ENVELOPES OF RESPONSE IN (0, 38)

a)
b)
t
MAXIMAL ACCELEROGRAM AT T = 15
Fig. 6.13 Upper and lower bounds and sample maximal accelerogram for T = 30 sec: (a) bounds
(6.170); (b) umax of any harmonic component.
6.6.2 Credible accelerograms

It is possible to investigate how the above bound is affected by the presence of
additional information. Assume that earthquakes, in general, exhibit a nominal
power spectrum whose shape is constructed to fit the following expression:
f (; 1 , 1 , 2 , 2 ) = K[g(; 1 , 1 ) + g(; 2 , 2 )] (6.172)
with
2 + 42 2
g(; , ) = (6.173)
( 2 2 )2 + 42 2
where and are parameters governing the shape of the spectrum, and K is a
normalizing factor. These parameters are assumed to have assigned values for any
given site under examination. Note that Eq. (6.172) is nothing else but the sum
of two functions of the KanaiTajimi type; the superposition is introduced in order
to have a better approximation for the spectra of recorded earthquakes exhibiting
more than one dominant frequency.
Once this shape is assumed as the central spectrum, it is to be expected that
accelerograms should not differ from the central spectrum by more than a given
amount. Let c0 be the vector of combination coefficients fitting the central spec-
trum; i.e., with reference to a basis of the type (6.171):
c20,i = c20,i+m = f (; 1 , 1 , 2 , 2 ), (i m) (6.174)
We assume that through the analysis of the previous recorded earthquakes, it

is possible to qualify earthquakes and their variability, as accelerograms with given
norm of Eq. (6.160), but with distance from c0 not larger than a given amount, say
. Thus, the constraints on the coefficients are of the type
XN
c2i = 1 (6.175a)
i=1
N
X
(ci c0,i )2 2 (6.175b)
i=1
which should be considered in conjunction with the condition
N
X
c20,i = 1 (6.176)
i=1
expressing the fact that c0 yields an accelerogram of a given norm, realized by
assigning E0 a proper value. The introduction of this additional information im-
plies that one intends to deal with functions that are not far from earthquake-type
accelerograms.
The expected result is the maximum stress in a linear SDOF structure subjected
to spectrum-compatible earthquakes. The term spectrum-compatible indicates that
at the given site the seismic excitation is qualified by its distance (6.175b) from the
central spectrum.
The problem is similar to the one dealt with in Sec. 6.5, except the fact that
now the admissible set of excitations is given by the intersection of the unit sphere
S0 in Eq. (6.175a), representing the energy constraint, and the side sphere Sd in
Eq. (6.175b), namely, the spectrum-compatible sphere as shown in Fig. 6.14. The
problem of finding the maximum displacement at a certain time instance t reads:
c
subject to (c c0 )> (c c0 ) 2 (6.177b)

>
c c1 (6.177c)
2
where < 4, otherwise Sd includes S0 , and the problem reduces to the formulation
given in Problem (6.164). Problem (6.177) can be solved in two steps:
Step 1: Solve the simpler Problem (6.164) and find the anti-optimal solution c1 .
If c1 is contained in Sd , then it represents also the solution of Problem (6.177).
If c1 is on S0 but outside Sd , then no internal point to Sd exists where the anti-
optimality (optimality) condition (6.166) (respectively, condition (6.167)) holds,
and the solution must be searched for on the boundary of the intersection of S0
and Sd , as described in Step 2.
Step 2: Now the problem transforms into the following one:
c
subject to (c c0 )> (c c0 ) = 2 (6.178b)

>
c c1 (6.178c)
c2
Sd
c0
c1
So
Fig. 6.14 Set of admissible coefficients.
Note that the constraint concerning Sd can be written as

N
X N
X
(ci c0,i )2 = (c2i 2ci c0,i + c20,i ) = 2 (6.179)
i=1 i=1
Taking into account Eq. (6.176), we can state that
N
X 2
ci c0,i = 1 =a (6.180)
i=1
2
with
|a| < 1 (6.181)
As a consequence, the problem can be formulated as follows:
c
subject to c c = 1, c>
>
0 c=a (6.182b)
where all constraints appear in the form of equalities. In order to solve Prob-
lem (6.182), we consider
L(c, r1 , r2 ; t) = c> u(t) + r1 (c> c 1) + r2 (c>
0 c a) (6.183)
where r1 and r2 are Lagrange multipliers. We require that the gradient of the
Lagrangian with respect to c vanishes as
L = u(t) + 2r1 c + r2 c0 = 0 (6.184)
Hence,
1
c = c(t) = (r2 c0 + u(t)) (6.185)
2r1
Substituting Eq. (6.185) into the second constraint c>
0 c = a in (6.182b) for S0 , we
obtain
1
c> c = 2 (u(t)> u(t) + r22 c> >
0 c0 + 2r2 c0 u(t)) = 1 (6.186)
4r1
where one puts

2
c> >
0 u(t) = b, u(t) u(t) = 4U (t) (6.187)
>
Bearing in mind that we must have the solution satisfying c c = c>
0 c0 = 1, we
arrive at
4r12 r22 2r2 b 4U 2 (t) = 0 (6.188)
Substituting now Eq. (6.185) into the second constraint c>
0c = a in (6.182b) for Sd
with equality, we obtain
1
c>0c= (r2 + b) = a (6.189)
2r1
Hence
r2 = (2r1 a + b) (6.190)
Substituting the latter expression into Eq. (6.188), the following equation for r 1 is
obtained:
4(1 a2 )r1 + b2 4U 2 = 0 (6.191)
Therefore,
r
1 4U 2 b2
r1 = (6.192)
2 1 a2
Note that the expression under the root is positive. Indeed, 0 < a2 < 1 from
Eq. (6.181), and by elementary use of the CauchySchwarz inequality, we establish
b2 = (c> 2 > > >
0 u(t)) (c0 c0 )(u(t) u(t)) = u(t) u(t) = 4U
2
(6.193)
After r1 has been calculated, r2 is given by Eq. (6.190). Consider now the elements
of the Hessian
2 L(c, r1 , r2 ; t) 2r1 for i = j
= (6.194)
ci cj 0 for i 6= j
Thus, we verify that the matrix of second-order derivatives of the Lagrangian with
respect to the basic variables ci is diagonal. Moreover, it is positive definite if r1
is positive; then the solution corresponds to the local minimum of the objective
function c> u. Analogously, if r1 < 0, the matrix of second-order derivatives of the
Lagrangian is negative definite, and the solution corresponds to the local maximum
of the objective function.
In conclusion, the solution for the problem of the maximum response is given by
r
1 4U 2 b2
r1 = ,
2 1 a2
r2 = (2r1 a + b), (6.195)
1
c = c(t) = (r2 c0 + u(t))
2r1
Likewise, the solution for the minimum response is given by changing the sign of r1
as r
1 4U 2 b2
r1 = (6.196)
2 1 a2
with the same definition of r2 and c as in Eq. (6.195).
Table 6.2 Values of epicentral distance, local intensity, PGA, duration, and norm at
various sites.
Epicentral Local
Site distance intensity PGA Duration Norm
(km) (MCS) (g/10) (sec) (cm sec 3/2 )
Torre del Greco 80.1 7.0 0.59 52.9 79.46

Brienza 41.3 7.0 2.24 78.7 185.19
Sturno 34.8 6.0 2.25 70.7 284.53
Bagnoli Irpino 22.3 6.0 1.31 79.1 152.29
6.6.3 Application
In order to have a quantitative estimate of the proposed procedure, consider a
particular area, namely, the Campania region in southern Italy, where a number of
accelerograms have been recorded during the Campania-Lucania earthquake which
took place 23 November, 1980. The magnitude of the event was estimated to be
6.5, the epicentral intensity was set at 7.5 degree the MercalliCancaniSieberg
(MCS) intensity scale. The event was rather non-typical, mainly because of its
duration that was up to almost 2 min at some sites. It should be noted that only
the duration of the strong motion will be considered here, which varies from about
52 sec to more than 86 sec. Information on the local characters of ground motion
was derived from the direct inspection of the recorded accelerograms, limiting the
analysis, for simplicity, to only the NS component. The considered records are
summarized in Table 6.2. Recorded ordinates of all accelerograms are converted to
cmsec2 . All earthquakes are preliminarily reduced to the same energy norm, given
in Eq. (6.154) as the one recorded in Torre del Greco, the closest site to Naples, the
town with the greatest intensity in the region, so that every record possesses energy
E0 = 79.46 cm sec3/2 .
The analysis was carried out including m = 400 sine and cosine waves for Torre
del Greco and Sturno, while m is increased to 800 for the accelerograms recorded in
Brienza and Bagnoli Irpino that exhibit power spectra scattered on a wider range
of frequencies. A plot of a typical accelerogram with the approximation resulting
from the associated Fourier expansion and its power spectrum is given in Fig. 6.15.
Note that the plot of Fourier approximation in Fig. 6.15(b) looks almost the same
as the original accelerogram Fig. 6.15(a).
Let fij be the ith coefficient of the Fourier expansion of the jth accelerogram
(i = 1, . . . , m; j = 1, . . . , 4). The central distribution of coefficients, collected in the
vector c0 , were obtained for each of the four analyzed sites, through the following
steps:
Step 1: Normalize fij , for every j, to unit modulus through division by a factor
Fj = |f j |, with f j = (f1j , . . . , fmj )> , and let cij = fij /Fj .
Step 2: Determine, for every site, the best values of the parameters j and j that
PLOT OF RECORDED ACCELEROGRAM - TORRE DEL GRECO-NS
POWER SPECTRUM
t S
(a)
a*

(c)
t
(b)
Fig. 6.15 Torre Del Greco-Campania Earthquake of Nov. 23, 1980; (a) recorded accelerogram;
(b) Fourier approximation; (c) power spectrum.
give the optimal fit of the ordinates of the KanaiTajimi-type spectrum given in
Eq. (6.172) with the values of the spectrum corresponding to the normalized
coefficients cij from the recorded accelerograms on the corresponding sites.
Step 3: Calculate, for every value of , the target spectrum s0,ij , e.g., the spectrum
in Eq. (6.172), as follows:
m
X
s20,ij = f (; 1 , 1 , 2 , 2 ), s20,ij = 1 (6.197)
i=1
Step 4: Let
ci+m,j
ij = , (i = 1, . . . , m) (6.198)
cij
and, finally, calculate the central target spectra at the considered sites as follows:
s0,ij
c0,ij = q sign fij (6.199)
2
1 + ji
Step 5: Calculate, for every site, the quadratic scatter of the coefficients obtained
by directly processing the accelerogram and those calculated by the target spec-
trum
Xm
j2 = (cij c0ij )2 (6.200)
i=1
RECORDED SPECTRUM TORRE DEL GRECO-NS

OMB=9.920058 CSIB=.109198 OMC=14.54868 CSIC=.1705714
R2=.2608822
Fig. 6.16 Processed power spectrum from recorded accelerograms and fitted central spectrum, in
the location of Torre del Greco.
Table 6.3 Values of j2 , 1 , 1 , 2 , and 2 at various sites.
Site j2 1 (sec1 ) 1 2 (sec1 ) 2
Torre del Greco 0.261 9.92 0.109 14.55 0.171

Brienza 0.219 32.98 0.478 37.67 0.403
Sturno 0.186 17.94 0.347 3.77 0399
Bagnoli Irpino 0.257 5.15 0.142 28.64 0.762
In Fig. 6.16, the processed spectrum at Torre del Greco and the central spectrum
are plotted. In Table 6.3, the final values of j2 are listed illustrating that the
inequality j2 4 holds.
Assuming that the central spectrum is site-dependent, but that independence
holds for the quadratic scatter, it is possible to infer from the data that j2 ' 0.26
for the area under examination. After introducing this value in the results discussed
in Sec. 6.6.2, one obtains the bounds plotted in Figs. 6.17 and 6.18 for 0 = 10
100 sec1 .
It is also possible to perform a comparison with the previous, spectrum-free
bound, showing that the new bound is approximately half of the one constrained
only by the upper bound by Drenick.
The response displacement spectrum by the present procedure is finally calcu-
lated, for a value of the damping coefficient = 0.05. The comparison with the
envelope of the same spectra calculated with reference to the processed accelero-
grams shows that the present results, although better founded and less sensitive to
uncertain parameters than recorded accelerograms, are not too conservative, but
Sn (cm)
PRESENT BOUNDS
5 SPECTRA OF RECORDED EARTHQUAKES:
TORRE DEL GRECO
4 BRIENZA
STURNO
CALITRI
3
BAGNOLI IRINO
0(sec-1)
20 40 60 80 100
Fig. 6.17 Response spectra of displacement: Comparison of present bounds with the envelope of
spectra from recorded accelerograms normalized to the same norm as Torre del Greco.
Sn (cm)
SHINOZUKAS BOUND
DRENICKS BOUND
5
PRESENT BOUNDS
4 ENVELOPE OF RECORDED SPECTRA
0(sec-1)
20 40 60 80 100
XMMAX=5.62035 YMMAX=4.21342 YMMIN.02036
Fig. 6.18 Response spectra of displacement: Comparison of present bounds with Drenicks and
Shinozukas bounds.
yield a magnification by a factor of about two in the design forces, in the whole
range of natural periods of the structures; this result can be directly incorporated
into the safety factor.
Following Shinozukas (1970) model, and the assumption given in Eq. (6.169),
with the parameters presented in Table 6.3, one obtains the following bound for the
maximum displacement over the whole duration of the excitation:
Z
1
Iei = |H()|X()d (6.201)
2
where
1
H 2 () = (6.202)
(02 2 )2 + 42 02 2
and
X() = K[g(|1 , 1 ) + g(|2 , 2 )] (6.203)
with K chosen so that
Z
1
X 2 ()d = E02 (6.204)
2
From inspection of Figs. 6.17 and 6.18, it is possible to infer that Problem (6.164)
yields the bound in the norm substantially close to that obtained via the Drenicks
original approach, apart from the range of higher frequencies, due to the truncation
of the expansion (6.171) by a finite number of harmonics (say, for i > 50 sec1 ).
Moreover, the spectrum-compatible bound in Problem (6.177) is not significantly
different from Shinozukas bound, although it yields slightly closer bounds in the
range of low frequencies. The opposite happens in the range of higher frequencies.
The ratio of the ordinates of the spectrum-compatible bound to the correspond-
ing ordinates of the envelope of spectra related to actual seismic records, normalized
to the same norm as Torre del Greco, vary in the range 2.0 2.5.
6.6.4 Discussion and conclusion

In this section, an approach has been presented for response analysis of a linear
structure subjected to seismic loads, and for calculation of an upper bound of the
response based on the solution of a convex anti-optimization problem. Starting
with a specialized function space with a given basis, the anti-optimal combination
of the coefficients can be found analytically. The data required for the analysis
in this section are: (i) the estimated energy of the earthquake, (ii) an estimate of
the shape of the central power spectrum at the site, (iii) an upper bound on the
maximum distance of the power spectrum of realizable earthquakes at the site from
the central one.
A direct comparison has been performed with Shinozukas (1970) approach,
yielding at some cases very close results. It should be noted that in Shinozukas
approach, however, the admissible spectra are bounded from above by the central
spectrum, while the approach in this section includes earthquakes whose spectra
are allowed to exceed the central spectrum; note, moreover, that in Shinozukas
approach the bound involves only the steady-state part of the response, while in
the approach in this section, as in Drenicks (1970) one, the bound involves also the
transient response starting with homogeneous initial conditions.
It should be emphasized that the present convex modeling combines positive
features of the approaches developed by Drenick (1970) and Shinozuka (1970), and
considerably reduces the estimates of the maximum possible response.
One should stress that the real structures are generally not SDOF-systems where
the maximum stresses are governed by the maximum displacement of a single mode.
The generalization of this method to multi-degree-of-freedom realistic structures is
possible through casting it in a state-space form as a vector differential equation.
One should also note that the maximum stress at two distinct positions of the
structure will be produced by the different worst-case combinations of excitations.
Obviously, one is interested in the worst response at the critical location of the
structure. Determining such a location may be a non-trivial task. Then, if the
structure is represented as the multi-degree-of-freedom system, one must explore
the maximum responses of each of the masses, and the maximum of these maxima
will constitute a critical response.
Analogously, a distributed system must be discretized through a fine mesh and
the maximum responses at each nodal point must be determined. Then the location
with maximum response amongst critical responses at each node will constitute the
globally critical response. The attractive feature in the present method is the fact
that one does not guess the critical location, but rather determine it through a
discretization scheme combined with convex optimization procedure for each nodal
point of the mesh.
Note that the present approach is restricted to linear systems. The generalization
to realistic nonlinear systems appears to be necessary for practical applicability of
this method. (This appears to be doable if and when the funding agencies are
more receptive to high-risk high-payoff methodologies.) In this context, convex
models have been applied to the nonlinear structures through treating the nonlinear
transfer function between the uncertain quantities and the output of the nonlinear
transformation, as a numerical code utilized by Ben-Haim and Elishakoff (1989b)
and the automatic differentiation procedure. Another alternative is the use of the
nonlinear programming schemes. For the application of the latter method the reader
may consult the paper by Li, Elishakoff, Starnes and Shinozuka (1996).
To sum up, it appears that the time is ripe to critically revisit existing stochastic-
ity concept-based approaches in earthquake engineering and examine the possibility
of utilizing alternative approaches. One such alternative approach was attempted
in this section, although to an SDOF system in the linear setting. Related sub-
jects, although in different settings were pursued by Abbas and Manohar (2005,
2007), Ben-Haim, Chen and Soong (1996), Elishakoff (1991b), Elishakoff and Plet-
ner (1991), Ganzerli, Pantelides and Reaveley (2000), He and Zhang (1997), Iyengar
and Manohar (1987), Lignon and Jezequel (2006), Ma, Leng, Meng and Fang (2004),
Manohar and Sarkar (1995), Moustafa (2009), Pantelides and Tzan (1996), Philap-
pacopoulos (1980), Schmitendorf, Jabbari and Yang (1994), Srinivasan, Corotis and
Ellingwood (1992), Srinivasan, Ellingwood and Corotis (1991), Takewaki (2002a,
2002b, 2006b), Tzan and Pantelides (1996b), Wu and Soong (1996), Yoshikawa
(2002, 2003), Zhai and Xie (2007), and possibly others. Finally, we mention a
definitive monograph by Takewaki (2006a) and references therein. It generalizes
earlier investigations on convex modeling of earthquakes.
Shinozuka (1970) pioneered the determination of the maximum response in
stochastic settings. Further stochastic anti-optimization-based studies include those
by Deodatis, Graham and Micaletti (2003a, 2003b), Deodatis and Shinozuka (1989,
1991), Manohar and Sarkar (1995), Sarkar and Manohar (1998), Shinozuka (1987),
Shinozuka and Deodatis (1988), and others.
6.7 Aeroelastic Optimization and Anti-Optimization
6.7.1 Introduction
In this section, we apply the hybrid optimization and anti-optimization to aeroe-
lastic problems, closely following the article by Zingales and Elishakoff (2001).
Optimization of structures with aeroelastic constraints has been dealt with by
several authors. A partial list of works includes those by Ashley (1982), Bishop,
Eastep, Striz and Venkayya (1998), Librescu and Bainer (1983), Livne and Mineau
(1997), Plaut (1971), Pierson and Genalo (1977), Ringertz (1994), Shirk, Hertz
and Weisshaar (1986), and Turner (1982). However, the uncertainty in elastic
moduli or in the material properties in conjunction with the aeroelastic optimization
is a relatively new topic. It has been addressed in the probabilistic setting by
Kuttekeuler and Ringertz (1998) and Allen and Maute (2004). The stochastic finite
element method for reliability of plates in supersonic flow was introduced by Liaw
and Yang (1990). In the future one would anticipate an increased utilization of the
stochastic finite element method in conjunction with aeroelastic phenomena. On
the other hand, it is instructive to quote Shinozuka (1987):
... It is recognized that it is rather difficult to estimate experimen-

tally the auto-correlation function, or in the case of weak homo-
geneity, the spectral density function of the stochastic variation of
material properties. In view of this, the upper bound results are
particularly important, since the bounds derived ... do not require
knowledge of the auto-correlation function.
This observation may limit the applicability of probabilistic analysis to uncer-

tainty. In these circumstances one should look for alternatives to the notion of
stochasticity. As Livne (1997) mentions in his careful overview of the subject of
aeroelastic optimization,
... approaches for addressing uncertainty in design optimization of

structural systems are still in various stages of study and evalua-

tion ... Ranging from statistical methods in which the statistical
characteristics of uncertainties are known or assumed or methods
based on fuzzy logic, as well as methods which use bounds on the
system uncertainties to optimize for a worst case combination of
pre-assigned parameter and modeling errors all these methods
have not been compared yet in the context of aeroelasticity and
aeroservoelasticity of fixed wing airplanes.
In this section, we adopt an anti-optimization method that utilizes information
that is easier to obtain (than either a stochasticity or fuzzy-sets based approach);
namely, the region of variation of the elastic moduli. Because of this partial informa-
tion it is sensible to require only the fragmentary characterization of the response;
namely, the determination of the maximum and minimum values of the critical
velocity. Once the minimum of the velocity is determined, the structural parame-
ters are chosen so as to exceed some preselected velocity. Numerous examples are
elucidated to gain physical insights into the problem.
6.7.2 Deterministic theoretical analysis

Let us consider a BernoulliEuler beam with variable cross-section and nonhomo-
geneous elastic properties. The beam is subjected to a supersonic stream flow in
the direction of beam axis (x-direction). Youngs modulus E(x) of the beam is
represented by the following expression:
E(x) = 1 E1 + 2 E2 (6.205)
where E1 and E2 can be characterized as amplitudes, and 1 (x) and 2 (x) are
continuous non-dimensional functions governing the variation of the modulus E(x).
In this section, we abstain from specifying particular forms of the functions 1 (x)
and 2 (x), except noting that they should be chosen to satisfy the obvious physical
requirement E(x) > 0. The width b(x) of the beam and its depth h(x) are considered
as varying with respect to x:
b(x) = b0 b (x), h(x) = h0 h (x) (6.206)
where b0 and h0 are positive constants, and b (x) and h (x) are shape functions
governing the variation of b(x) and h(x), respectively. Moreover, b (x) and h (x)
equal unity at the coordinate origin; i.e., b (0) = h (0) = 1. Therefore, b0 and h0
represent the width and the depth of the cross-section at x = 0. The description
of the aeroelastic interaction between the transverse deflection w(x, t) of the beam
and the air pressure load will be performed by means of the piston theory (Ashley
and Zartarian, 1956; Ilyushin, 1956; Bolotin, 1963). The effect of the structural
and aerodynamic damping on w(x, t) will be excluded for the sake of simplicity.
The equation
of motion reads
2 2 w(x, t) 2 w(x, t) kp w
E(x)I(x) + A(x) + U =0 (6.207)
x2 x2 t2 c x
where A(x) is the cross-sectional area, I(x) is the moment of inertia, and is the
mass per unit length of the beam. The gas-flow interaction is represented by the
last term in Eq. (6.207), where k is the exponent of the polytropic thermodynamic
transformation, p is the pressure of the nondisturbed gas flow, c is the speed
of the shock waves in the nonperturbed airstream, and U is the relative velocity
between the gas flow and the system. The boundary conditions associated with the
simply supported beam read
2 w(x, t)
w(x, t) = = 0 at x = 0 and L (6.208)
x2
where L is the length of the beam. The initial conditions are
w(x, t)
w(x, t) = = 0 at t = 0 (6.209)
t
The governing equation for the deflection w(x, t) is not solvable exactly for arbitrary
choices of the functions 1 (x), 2 (x), b (x), and h (x). Hereinafter, the Bubnov
Galerkin approach will be utilized. We approximate the solution as
N
X
w(x, t) = i (x)fi (t) (6.210)
i=1
with N denoting the number of retained terms in the expansion, and i (x) designat-
ing known comparison functions satisfying all boundary conditions in Eq. (6.208).
We substitute Eq. (6.210) into Eq. (6.207). Naturally, the expression Eq. (6.210)
does not satisfy Eq. (6.207). The result of the above substitution, therefore, differs
from zero. We denote it by . We require that the inner product vanishes:
(, j ) = 0, (j = 1, . . . , N ) (6.211)
where
Z L
(, j ) = (x)j (x)dx (6.212)
0
This procedure leaves us with a set of ordinary differential equations:
N
X N
X N
X
kp
K f (t) + M f (t) + U N f (t) = 0, ( = 1, . . . , N ) (6.213)
=1 =1
c =1
The coefficients appearing in Eq. (6.213) are given as follows:
Z L
d2 d2 (x)
K = (x) 2 E(x)I(x) dx,
0 dx dx2
Z L
M = A(x) (x) (x)dx, (6.214)
0
Z L
d (x)
N = dx
0 dx
The solution of the system in Eq. (6.213) is furnished in the form
f (t) = A eit , ( = 1, . . . , N ) (6.215)
where represents the eigenfrequency. We arrive at an algebraic homogeneous

system of equations for A :
N
X N
X N
X
K A 2 M A + U N A = 0, ( = 1, . . . , N ) (6.216)
=1 =1 =1
To obtain a the nontrivial solution; namely,

N
X
A2j 6= 0 (6.217)
j=1
one requires the determinant of the coefficient matrix of the system in Eq. (6.216)
to vanish:
det[ 2 M + U N + K ] = 0 (6.218)
leading to the approximate eigenfrequency equation for .
6.7.3 Stability analysis within two-term approximation

Let us confine our analysis to the case where only two terms are retained in
Eq. (6.210)
A1 (K11 2 M11 + U N11 ) + A2 (K12 2 M12 + U N12 ) = 0,
(6.219)
A1 (K21 2 M21 + U N21 ) + A2 (K22 2 M22 + U N22 ) = 0
It is instructive to write Eq. (6.219) in terms of the circular natural frequencies of
the system 1 and 2 . First, as an auxiliary problem, consider the free-vibration
case
det[ 2 M + K ] = 0, (, = 1, 2) (6.220)
resulting in the biquadratic equation
4 c1 2 + c2 = 0 (6.221)
with the coefficients given by
K11 M22 + K22 M11 K12 M21 K21 M12
c1 = ,
det[K ]
(6.222)
det[K ]
c2 =
det[M ]
The solution is given as (Panovko and Gubanova, 1965)
q q
1 1
21 = (c1 c21 4c2 ), 21 = (c1 + c21 4c2 ) (6.223)
2 2
Because of Vietas theorem, c1 = 21 + 22 and c2 = 1 2 . Hence, Eq. (6.218) can
be rewritten in the following form:
4 + [( 21 + 22 ) + d1 U ] 2 + 1 2 + d2 U + d3 U 2 = 0 (6.224)
with terms di defined as
N11 M22 + N22 M11 N12 M21 N21 M12

d1 = ,
det[M ]
K11 N22 + K22 N11 K12 N21 K21 N12
d2 = , (6.225)
det[M ]
det[N ]
d3 =
det[M ]
Equation (6.215) suggests that once the frequency takes a complex value, the
dynamic instability called flutter occurs. The critical condition appears when the
discriminant
D = [U d1 + ( 21 + 22 )]2 4[21 22 + d2 U + d3 U 2 ] (6.226)
of the biquadratic equation (6.224) vanishes, yielding the flutter velocity

1
Ucr = 2 d2 (2 + 22 ) + 2d2
d1 4d3 1 1
q (6.227)
2 2 2 2 2 2 2 2
+ [d1 ( 1 + 2 ) + 2d2 ] (1 + 2 ) (d1 4d3 )
For values larger than Ucr , the discriminant takes a negative sign, and the roots are
complex conjugate. For the case of the beam simply supported at both ends, the
following set can be utilized for the comparison functions:
x
(x) = sin , ( = 1, 2, . . . ) (6.228)
L
which represent the exact mode shapes of the associated uniform beam. Note that
in these circumstances,
N11 = N22 = 0, N12 = N21 (6.229)
yielding
p
22 22 ( 22 22 ) det[M ]
Ucr = = (6.230)
2 d3 2N12
Observe that Eq. (6.230) formally coincides with expression (4.104) in the mono-
graph by Bolotin (1963) for the uniform beam. The difference lies in the definition
of the circular natural frequencies .
Recall that E1 and E2 are uncertain variables, leading to variability of the
flutter velocity. The following question arises: How should one take into account
the variability of the elastic modulus on Ucr ?
6.7.4 Convex modeling of uncertain moduli

In this section, the uncertainty in the elastic modulus E(x), represented in
Eq. (6.205), will be specified. In many recent investigations, the elastic modulus
is treated as a random variable and the problems are solved by the finite element
method necessitating the knowledge of the autocorrelation function of the elastic
modulus. However, such information is often if not always unavailable. In these
circumstances, one can act in one of the following ways: Either postulate that such
an autocorrelation function is known, or pick up (as is usually done) any allowable
autocorrelation function based on the assumption that once the autocorrelation
function becomes available it can be introduced into the analysis. Another way will
be to abandon this excessively demanding analysis and look for an approach that
does not postulate that uncertainty and probability are synonymous.
In this section, we use a non-probabilistic model of uncertainty. The two ampli-
tude parameters E1 and E2 are treated as uncertain variables, varying in a convex
domain C. The admissible uncertainty region of variation C as shown in Fig. 6.19
is then described by an inequality:
(E1 , E2 ) C (6.231)
The domain C for E1 and E2 will be assumed to lie entirely in the positive quadrant
of the plane (E1 , E2 ). Because of the variability of Ucr , it appears natural to evaluate
the worst critical velocity Ucr,worst , when E1 and E2 vary in C. Thus we are looking
for
Ucr,worst = min Ucr (E1 , E2 ) (6.232)
(E1 ,E2 )C
Note that the coefficient c1 in Eq. (6.222) is a linear function of Ej (j = 1, 2),

whereas the coefficient c2 is a quadratic function of them, that is,
c1 = 1 E1 + 2 E2 , c2 = 11 E12 + 12 E1 E2 + 22 E22 (6.233)
where
1 = M22 K111 + M11 K221 M12 K211 M21 K121 ,
2 = M22 K112 + M11 K222 M12 K212 M21 K122 ,
11 = K111 K221 K121 K221 , (6.234)
22 = K112 K222 K122 K212 ,
12 = K111 K222 + K112 K221 K121 K212 K122 K211
The expression for Kijl reads
Z L
d2 d2 i (x)
Kijl = j (x) 2 I(x)l (x) dx (6.235)
0 dx dx2
Bearing in mind Eqs. (6.223) and (6.233), we can rewrite Eq. (6.230) for the square
of the critical velocity as follows:
2
Ucr (E1 , E2 ) = d11 E12 + 2d12 E1 E2 + d22 E22 (6.236)
Fig. 6.19 Shape and location of the admissible uncertainty region C and of the 2 ellipses.
where
12 411 det[M ]
d11 = 2 ,
4N12
21 2 412 det[M ]
d12 = 2 , (6.237)
8N12
2 422 det[M ]
d22 = 2 2
4N12
One recognizes that the expression of the square of critical velocity is a quadratic
form in variables E1 and E2 . In matrix form, Eq. (6.236) can be rearranged as
2
Ucr = E > DE (6.238)
with

E1 d11 d12
E= , D= (6.239)
E2 d21 d22
Accounting for the physical meaning of the right-hand-side in Eq. (6.235), we can
conclude that the matrix D is positive definite. Therefore, Eq. (6.236), conveniently
rewritten in the form
2
E > DE Ucr =0 (6.240)
represents an ellipse in the plane (E1 , E2 ) with the center coinciding with the origin
of the coordinate system. We thus obtain a family of homothetic ellipses correspond-
2
ing to different values of Ucr if the latter is treated as an independent parameter.
2
Naturally, the larger values of Ucr correspond to larger semi-axes of the ellipse in the
2
plane. For small values of Ucr the ellipse in Eq. (6.236) and the region in Eq. (6.231)
are disjoint. For large values of it they have a common area. Therefore, there exist
2 2 2 2 2
two values of Ucr , denoted by Ucr,worst and Ucr,best with Ucr,worst < Ucr,best , for
which the two convex domains in Eqs. (6.231) and (6.236) share a single point. One
of the values corresponds to the case when one ellipse contains the other one en-
tirely, whereas the other point is associated with the case when they are on different
sides of the common tangent.
2 2 2
Values of Ucr falling outside of the interval [Ucr,worst , Ucr,best ] do not represent
feasible critical velocities, because the set (E1 , E2 ) will not satisfy Eq. (6.231). Note
2 2
that the two values of velocities, Ucr,worst and Ucr,best , respectively, represent the
smallest and the largest solutions allowed by Eq. (6.231). To determine these values,
we need to examine closely the relative spacing of the critical velocity ellipse and the
region of uncertain variation of the elastic modulus. According to the interpretation
of the parameters E1 and E2 as some effective elastic modulus, we shall confine our
analytical derivation to the arc of the ellipse in Eq. (6.236) contained in the first
quadrant of the plane (E1 , E2 ). It can be shown (see Zingales and Elishakoff (2001))
that, under some assumptions about the functions i (x), the critical velocity ellipse
in Eq. (6.240) is always oriented with its major axis in the second and fourth
quadrant of the plane (E1 , E2 ), as shown in Fig. 6.19.
2
Once the orientation of the ellipse, representing the critical velocity Ucr with
respect to the region C of the uncertain parameters E1 and E2 has been identified,
one can solve analytically the anti-optimization problem. The following sections
will be concerned with the solution of Problem (6.232) for various practical shapes
of the uncertainty region.
Before proceeding further, let us pose the following question: Why do we need
to consider various shapes of the uncertainty region? To answer this question, we
visualize that the results of the experimental measurements yield an ensemble of
functions Ei (x). These functions are then decomposed to sets of pairs (E1i , E2i ).
Each of these pairs forms a point in the plane (E1 , E2 ). We then are interested
in determining the region that contains all of these points. A convex hull of these
points is naturally such a set, forming a polygonal uncertainty region. Also, it
is easily visualized that different researchers and engineers may approximate the
uncertainty region via differing means. It makes sense, therefore, to treat different
possibilities.
6.7.5 Anti-optimization problem: polygonal region of uncertainty

In this section, we consider an uncertainty region C possessing a polygonal shape
with m vertices. The vertices Pi (i = 1, . . . , m) of the polygon have coordinates
Pi = (E1i , E2i ), respectively. We denote the vertex in the lowest left corner by P1 ,
the others are numbered counterclockwise as P2 , P3 , etc. as shown in Fig. 6.19.
Hereinafter we form a vector V of distances di of the vertices from the origin O:
q
V = (d1 , . . . , dm )> , di = E1,i
2 + E2
2,i (6.241)
The result of anti-optimization given by the joint points of the homothetically
inflated ellipse in Eq. (6.237) depends on the values of the design variables b0 and
h0 appearing in Eq. (6.206), in addition to uncertain variables E1 and E2 . This

dependence can be highlighted by rewriting Eq. (6.236) in the form
2
Ucr = b20 h60 2 (6.242)
with the positive coefficient
2 = d011 E12 + 2d012 E1 E2 + d022 E22 (6.243)
The coefficients d0ij
(i, j = 1, 2) in Eq. (6.243) are obtained from Eq. (6.236) by
formally letting b0 = h0 = 1. Because the critical velocity depends on the uncertain
parameters of elasticity Ei through the coefficient 2 , the anti-optimization problem
will focus on the determination of the extremal of the coefficient 2 , provided that
E1 and E2 vary in region C. Equation (6.243) represents a family of concentric
ellipses in the plane E1 and E2 obtained by varying the value of the coefficient
2 . The major axis passes through the second and fourth quadrants of the plane
(E1 , E2 ). As shown in Sec. 6.7.4 for the critical velocity (and hence 2 ), the ellipse in
Eq. (6.243) and the region of uncertainty C must share a common point. There are
two possibilities for such a point to exist. One, Pbest corresponds to the maximum
of 2 , whereas the minimum is associated with Pworst .
2 2
The extreme values min and max of the coefficient 2 are found by substi-
tuting the coordinates (E1,worst , E2,worst ) and (E1,best , E2,best ) of the points Pworst
and Pbest into Eq. (6.243). Bearing in mind Eq. (6.243), the expressions for the
2 2
maximum and minimum values of the critical flutter velocity Ucr,best and Ucr,worst
become
2
Ucr,best = b20 h60 max
2
(6.244a)
= b20 h60 (d011 E1,best
2
+ 2d012 E1,best E2,best + d022 E2,best
2
)
2 2 6 2
Ucr,worst = b0 h0 min (6.244b)
= b20 h60 (d011 E1,worst
2
+ 2d012 E1,worst E2,worst + d022 E2,worst
2
)
The expressions of the coordinates corresponding to the points Pworst and Pbest are
given in Zingales and Elishakoff (2001) for all the cases reported in Figs. 6.20(a)(c).
Comparing the ellipses in Fig. 6.19 with those in Fig. 6.20, we observe that
they possess different ratios of their respective semi-axes e = 1 /2 , referred to
as eccentricity e. This is because the functions 1 and 2 chosen in Fig. 6.19 are
1 = 1 + 5.3(x/L)2 + 0.5(x/L)3 and 2 = 1 + 2(x/L), whereas in Fig. 6.20 they
read 1 = 1 3(x/L)2 and 2 = 0.001 + 0.3(x/L). The shape functions j (x) for
Fig. 6.19 are positive in the interval
[0, L]. The function 1 (x) for Fig. 6.20 takes
negative values in the interval [L/ 3, L]. Still, the modulus of elasticity is a positive
quantity. Despite the function 1 (x) not being positive in the whole entire interval
[0, L], the major axis of the 2 ellipse lies in the second and fourth quadrants of
the plane (E1 ,E2 ). Numerical calculations show that the positiveness requirement
of functions 1 (x) and 2 (x) in the interval [0, L] leads to a large eccentricity e of
the 2 ellipse. The weaker condition that E(x) is positive but j (x) may not be
(a) (b)
(c)
2
Fig. 6.20 Regions of uncertainty: (a) rectangular region of uncertainty, max 2
and min ellipses,
(b) polygonal region of uncertainty with particular location of vertices P 3 and P6 , (c) singular
location for rectangular region of uncertainty; extremal that does not coincide with one of the
vertices.
results in lower values of e. The particular choice of 1 (x) and 2 (x) in Fig. 6.20
has been made to enable an easier visualization of the ellipses corresponding to
2 2
extremal values min and max .
The objective of the next section is to derive approximations of the admissible
region for E1 and E2 by a continuous smooth curve for analytical purposes.
(a)
Eq.(6.244),
(b)
Fig. 6.21 Ellipsoidal region of uncertainty: (a) ellipse, (b) ellipse corresponding to the extremal
values of the coefficient 2 .
6.7.6 Anti-optimization problem: ellipsoidal region of uncertainty

Let us assume that the region C is represented by an ellipse with the center E0 =
(E10 , E20 ):
(E1 E10 )2 (E2 E20 )2
2
+ 1 (6.245)
a b2
where a and b represent the semi-axes of the ellipse as shown in Figs. 6.21(a)
and (b). We are looking for the points Pworst = (E1,worst , E2,worst ) and Pbest =
(E1,best , E2,best ) that satisfy Eqs. (6.244a) and (6.244b) and correspond, respec-
tively, to the worst and the best critical velocities Ucr,worst and Ucr,best. In the
following, we denote the coordinates of the points Pworst and Pbest , respectively, as
2 2
(e1,worst , e2,worst ) and (e1,best , e2,best ). Then Ucr,worst , and Ucr,best are given by
2
Ucr,worst = (d011 e21,worst + 2d012 e1,worst e2,worst + d022 e22,worst )b20 h60 ,
(6.246)
2
Ucr,best = (d011 e21,best + 2d012 e1,best e2,best + d022 e22,best )b20 h60
2
The obtained solution for the anti-optimized critical flutter velocity Ucr,worst shall
be utilized in the next section to obtain the best possible value of a certain objective
function involving it.
6.7.7 Minimum weight design

The optimal design process involves the weight W of the beam
Z L
W = b0 h0 b (x)h (x)dx (6.247)
0
which is a monotonically increasing function of the design variables; namely, the
width b0 and the depth h0 at x = 0. The functions b (x) and h (x) governing the
shape of the cross-section of the beam have been introduced in Eq. (6.206).
We are interested in designing the system, i.e., determining the values of b0 and
h0 , that minimizes the weight W under constraint on the critical flutter velocity
2
Ucr whose value must be greater than the specified velocity U0 :
Ucr (b0 , h0 ) U0 (6.248)
Because Ucr is an interval variable [Ucr,worst , Ucr,best], fulfilment of the inequality
Ucr,worst (b0 , h0 ) U0 (6.249)
automatically leads Eq. (6.248) to be satisfied. Hereinafter we will deal with
Eq. (6.249). It is rewritten in a form that is more appropriate to the design by
utilizing the representation in Eq. (6.243). This yields
Ucr,worst = min b0 h30 U0 (6.250)
where min is defined in Eq. (6.244b). Let the manufacturing requirements demand
that the values b0 and h0 vary in a box :
b0L b0 b0U , h0L h0 h0U (6.251)
where b0L , b0U and h0L , h0U denote the bounds for the width and the thickness,
respectively. An additional constraint involving the design parameters b0 and h0
is concerned with the range of validity of the utilized mechanical theory. Because
we apply the BernoulliEuler theory, we specify an additional constraint over the
design variables as
b0 zh0 (6.252)
Fig. 6.22 Alternative location of the design box with respect to the constraint g 2 0.
Specifically, for the sake of determinacy, z is fixed at 10. The optimization problem
is then stated as
minimize W (b0 , h0 ) (6.253a)
subject to g1 (b0 , h0 ) = min b0 h30 U0 0 (6.253b)
g2 (b0 , h0 ) = 10h0 b0 0 (6.253c)
(b0 , h0 ) (6.253d)
The optimization problem (6.253) will be solved by employing geometrical con-
straints. Note that if in Eq. (6.253c) the entire region of variation is on the right side
of the line g2 (b0 , h0 ) = 0, then no acceptable solution exists. This case is represented
by the design region 1 drawn with a dotted line in Fig. 6.22. If, however, the region
is located on the left side of the line g2 (b0 , h0 ) = 0, the entire region is feasible
except for the additional constraint Eq. (6.253b) (see domain 3 in Fig. 6.22). If
the line g2 (b0 , h0 ) = 0 passes through the region, only the upper part (see the filled
part of the region 2 bounded by solid lines in Fig. 6.22) is feasible. We denoted
by Q1 and Q2 , respectively, the intersection points of the line g2 (b0 , h0 ) = 0 with
the region 2 that are closest and farthest from the origin O. Depending on the
location of the region 2 with respect to the origin, several possibilities arise. If the
line g2 (b0 , h0 ) = 0 crosses the edges AD and CD, the intersection points have the
coordinates
Q1 = (b0L , 1/10b0L), Q2 = (h0U , 1/10h0U) (6.254)
as shown in Fig. 6.22. Other possibilities of crossing the region by the line g2 = 0 are
elucidated in Figs. 6.23(a)(c), where the coordinates of the intersection points are
also indicated. To obtain the solution of the optimization problem in Eq. (6.253),
let us examine closely the critical velocity constraint g1 (b0 , h0 ) 0. If the curve
(a)
(b)
(c)
Fig. 6.23 Feasible regions: (a) and (b) feasible design region e for specified velocity U0 , (c)
alternative location of the design box with respect to constraints g 1 0 and g2 0 for specified
velocity U0 .
(a)
(b)
e (b) in corre-
Fig. 6.24 Solution of the optimization problem: (a) on the vertex of the region ,
spondence of points Q1 and Q2 in Eq. (6.254).
defined by
g1 (b0 , h0 ) = min b0 h30 U0 = 0 (6.255)
lies below the feasible region , then the constraints (6.253b) and (6.253c) are not
active for the specified value of the velocity U0 . In this case, the solution of the
optimization problem coincides with the closest vertex of the hatched region to the
origin O. This vertex, denoted hereinafter by Qopt , is either the point Q1 , as shown
in Figs. 6.22 and 6.23(b), or point A, as shown in Figs. 6.23(a) and (c), respectively,
if the straight line g2 (b0 , h0 ) = 0 intersects the edge AD or AB. However, if the
line in Eq. (6.255) is entirely above the hatched region , then no solution of the
optimization problem exists for the specified value of the velocity U0 . Thus, we
identify an interval [U0,min , U0,max ], where the specified velocity U0 must lie, so that
the inequality Ucr,worst U0 may hold. The extreme values U0,min and U0,max of
the admissible range of U0 are given by

U0,min = min b01 h301 (6.256a)
U0,max = max b02 h302 (6.256b)
where the values b0i and h0i (i = 1, 2) denote, respectively, the points in the feasible
domain with smallest and largest distances from the origin O, as represented in
Figs. 6.24(a) and (b). In Fig. 6.24(a) such extremal points are A and C, corre-
sponding to U0,min = 1.75 107 and U0,max = 3.65 108 , respectively. Hereinafter
we assume that the specified velocity belongs to the region U0 [U0,min , U0,max ].
Denote by e a region that is both feasible and satisfies the constraint on the
critical velocity Ucr,worst in Eq. (6.253b). The weight in Eq. (6.247) is a mono-
tonically increasing function of the variables b0 and h0 . Therefore, the solution
Qopt = (b0,opt , h0,opt ) of the optimization problem lies on the boundary of the re-
gion e closest to origin O.
To obtain the point Qopt on the boundary with minimum distance OQ, we again
resort to a geometric argument. Note that there exists a specific value of the distance
OQ such that the family of circles (denoted by the dotted line in Figs. 6.24(a) and
(b))
2
b20 + h20 = OQ (6.257)
and the curve g1 (b0 , h0 ) = 0 share a single point; that is, they have a common
tangent in point Q(b0 , h0 ) as shown in Figs. 6.24(a) and (b). The tangent line to
the curve g1 (b0 , h0 ) = 0 in a plane Ob0 h0 passing through the point Q = (b0 , h0 )
reads
4
dh0 min h0
h0 h 0 = (b0 b0 ) = (b0 b0 ) (6.258)
db0 h0 =h0 ,b0 =b0 3U0
where
1 " !#1 !1
dh0 db0 d U0 3U0
= = 3 = 4 (6.259)
db0 dh0 dh0 min h0 min h0
The tangent to the circle in Eq. (6.257) is represented by
2
b0 b0 + h0 h0 OQ = 0 (6.260)
At point Q the two straight lines in Eqs. (6.258) and (6.259) coincide. Therefore,
the coefficients of the variables b0 and h0 in Eqs. (6.258) and (6.259) must be
proportional, with proportionality constant denoted by :
4
b0 = min h0 (6.261a)
h0 = 3U0 (6.261b)
The solution reads
5
h0 h min
= , b0 = 0 (6.262)
3min 3U0
We substitute b0 into Eq. (6.256a) to obtain h0 :

! 41
3U0
h0 = (6.263)
min
The value of b0 is found by substituting Eq. (6.263) and the expression for in
Eq. (6.262) into Eq. (6.261a) as
1
3U0 4
b0 = (6.264)
min
We denote by R1 and R2 the intersection points of the curve g1 (b0 , h0 ) = 0 with
the boundaries of the region . e If the point Q with coordinates (b0 , h0 ), defined
in Eqs. (6.263) and (6.264), lies on the boundary of the feasible region as shown in
Fig. 6.24(b), i.e., is located on the arc R1 R2 of the curve g1 (b0 , h0 ) = 0, then the
values of b0 and h0 so obtained represent the solution of the optimization problem
(6.253). If, however, the point Q is outside the admissible region as shown in
Fig. 6.24(a), then the solution of the optimization problem is given by the vertex
e closest to the origin O.
of the region
6.7.8 A numerical example

In this section, we apply the present hybrid optimization and anti-optimization
procedure to the minimum weight design of a simply supported beam with length
L = 15 m. The functions i (x) in Eq. (6.205) governing the variation of the elastic
modulus E(x) along the axis of the beam are chosen in the form
x 2 x 4 x 5
1 (x) = 1 + 1 + 3 , 2 (x) = 2 (6.265)
L L L
where i (i = 1, 2, 3) are the coefficients given as 1 = 1.1, 2 = 5.4, 3 = 2.2.
The shape functions b (x) and h (x) in Eq. (6.206) governing, respectively, the
geometric shape of the width b(x) and of the height h(x) are
x x 2
b (x) = 1 + 1 , h (x) = 1 + 2 (6.266)
L L
where i (i = 1, 2) are the coefficients, which are given as 1 = 0.7, 2 = 0.73.
The uncertain coefficients E1 and E2 in Eq. (6.205) are assumed to vary in an
ellipsoidal region. The coordinates E10 and E20 of the center and the semi-axes
a and b of the ellipse are listed in Fig. 6.25(a). The mass density is set at =
2.7 106 kg s2 /cm4 , whereas the aeroelastic parameters in Eq. (6.213) are c =
3.3 104 m/s and k = 1.66. The values of the coefficients d0ij are d011 = 3.4 105 ,
d012 = 3.91 106 , and d022 = 5.5 106 , which yield a family of concentric ellipses
after substituting the values of the coefficients d0ij into Eq. (6.243). The generic
ellipse of the family is represented by the dotted line in Fig. 6.25(a). Equation
(6.247), used to determine the coordinates e1 and e2 , becomes
4 7.81 106 3 6.32 1013 2 + 1.11 1011 5.22 1023 = 0 (6.267)
(a)
(b)
Fig. 6.25 Application of the hybrid optimization and anti-optimization.
whose smallest solution is = 5.07 106 . The values of e1 and e2 are then
e1 = 1.06 105 and e2 = 2.98 105 . In Fig. 6.25(a), the ellipse corresponding to
the minimum 2 and tangent to the region of uncertainty in Pworst = (e1 , e2 ) is
depicted with a solid line. We assume that the design variables b0 and h0 are in the
region
6 cm b0 30 cm, 0.5 cm h0 2.5 cm (6.268)
represented in Fig. 6.25(b). The lines corresponding to the extremal of the
interval [U0,min, U0,max ] are depicted, respectively, with dashed and dot-dashed
lines. The interval where the constraint g1 (b0 , h0 ) 0 is active is represented
by [1.9 103 , 1.02 106 ]. We set U0 = 900 m/s.
The region e of variation of the design variables b0 and h0 is depicted as a
rectangle in Fig. 6.25(b). The values of b0 and h0 corresponding to Eqs. (6.263)

and (6.264) are listed in Fig. 6.25(b). It is seen that they do not belong to the
boundary of the region e as shown in Fig. 6.25(b). Therefore, the combination of
the design values b0 and h0 corresponding to the minimum of the weight W (b0 , h0 )
is found at the point Qopt = (6.0 cm, 1.79 cm). With the obtained values, the
minimum weight in this simple model is W = 38.7 kg, which is the smallest value
satisfying all the constraints imposed on the system.
6.7.9 Conclusion
We discussed the flutter of a beam with uncertain elastic modulus that is simply
supported at both ends in a stream of gas flow. The structural model has a variable
cross-section and variable elastic modulus along the axis of the beam to reflect the
realistic situation. We addressed the following question: How can the optimization
methods in the presence of the bounded uncertainty, which is neither probabilistic
nor fuzzy, be utilized? The parameters describing the elastic modulus were modeled
as variables belonging, respectively, to a polygonal hull of experimental points, or
to an ellipsoidal region. Hybrid optimization and anti-optimization proved to be an
effective tool for dealing with optimum design in the presence of uncertainty. We
first solved the problem to find an expression for the worst possible combination of
the uncertain parameters involved in one of the constraints of the design variables.
As a second step we optimized the design variables under the anti-optimized version
of one of the constraints. Note that the further studies on flutter problems in non-
probabilistic setting were conducted by Wang and Qiu (2009) and Khodaparasi,
Marques, Badcock and Mottershead (2009).
For additional references to anti-optimization in the vibration context, the reader

may consult, for example, the works by Ben-Haim (1994, 1998a, 1998b), Ben-Haim,
Fr
yba and Yoshikawa (1999), Ben-Haim and Natke (1992), Capiez-Lernout and
Soize (2008), Chen, Chu, Yan and Wang (2007), Chen and Lian (2002), Chen, Qiu
and Liu (1994), Chen, Qiu and Song (1995), Chen and Wu (2004a), Chen, Wu and
Yang (2006), Chen, Yang and Lian (2000), Craig, Stander and Balasubramanyam
(2003), Drenick (1968, 1970, 1973, 1977a, 1977b, 1993), Drenick, Novomestky and
Bagchi (1989), Drenick and Park (1975), Drenick, Wang, Yun and Philappacopou-
los (1980), Drenick and Yun (1979), Dyne, Hammond and Davies (1988), Elishakoff
(1991a), Elishakoff and Ben-Haim (1990b), Elishakoff and Duan (1994), Fr yba, Nak-
agiri and Yoshikawa (1998), Fryba and Yoshikawa (1997, 1998) Gao, Zhang, Ma and
Wang (2008), Hlav acek (2002b), Hlav
acek, Plesek and Gabriel (2006), Leng and He
(2007), Li, Elishakoff, Starnes, and Shinozuka (1996), Lignon and Jezequel (2006),
Modares, Mullen and Muhanna (2006), Mullen and Muhanna (1999), Nakagiri and
Suzuki (1997), Qiu, Chen and Elishakoff (1995a, 1996a, 1996b) Qiu, Chen and Jia
(1995b), Qiu, Elishakoff and Starnes (1996d), Qiu and Hu (2008), Qiu, Hu, Yang
and Lu (2008a), Qiu, Ma and Wang (2004c), Qiu, M uller and Frommer (2004e),
Qiu and Wang (2003, 2005, 2006), Sadek, Sloss, Adali, and Bruch (1993), Srini-
vasan, Corotis and Ellingwood (1992), Srinivasan, Ellingwood and Corotis (1991),
Venini (1998), Wang and Qiu (2009), Yamaguchi, Kogiso and Yamakawa (2007),
and Yoshikawa and Fr yba (1997).

Chapter 7
Anti-Optimization via FEM-based

Interval Analysis
Make the best of things. (Russian proverb)

... approaches based on the concept of critical excitation or
worst-case input are promising and seem to lead to rational
design concept. (Takewaki, 2002b)
In this chapter, we deal with anti-optimization in the context of the finite element
method, which is apparently the most universal method currently for structural
analysis. We first discuss the interval analysis of multi-degree-of-freedom systems.
Then we deal with interval finite element analysis of shear frame.
7.1 Introduction
In many real-life problems of structural engineering, neither the initial conditions,

external forces, nor the parameters of the constitutive relations used in procedures to
calculate response quantities can be perfectly described, because of lack of detailed
knowledge and also due to the inherent uncertainties. Therefore, there have been
many approaches presented for numerical anti-optimization based on the general
finite element method (FEM). Pantelides and Booth (2000) developed a numerical
optimization algorithm with ellipsoidal bounds of uncertainty, and applied it to
reinforced concrete structures and trusses. Muhanna, Mullen and Zhang (2005)
presented a penalty-based approach. Lanzi and Giavotto (2006) used a commercial
finite element analysis package called ABAQUS for buckling optimization. General
methodologies of anti-optimization in parallel computing environment is discussed
in Gurav, Goosen and van Keulen (2005a).
Presently, there are three paradigms to describe uncertainty: (a) theory of prob-
ability and random processes, (b) fuzzy sets, and (c) convex set-theoretical analy-
sis. These three corners of the uncertainty triangle (Elishakoff, 1990), as mentioned
in the preface of this book, can be combined in various manners. The hybrid
stochastic-fuzzy approach was developed by Hoh and Hagaki (1989). Convex mod-
211
eling of uncertainty was studied in the book by Ben-Haim and Elishakoff (1990), in
which a complete literature survey for continuous systems and several illustrations
for the applications in mechanics are presented. Elishakoff and Colombi (1993)
developed a hybrid stochastic-convex approach to solve some actual space-shuttle
problems.
Interval analysis is one of the tools for incorporating uncertainty in structural
analysis. We presented in Sec. 3.3 the basic properties of interval analysis using
simple mathematical and structural models. However, for practical design pro-
cess, response quantities, in the form of internal forces, stresses, displacements and
strains, are needed to evaluate the safety and serviceability of structures with many
degrees of freedom. Interval analysis for multidimensional structures is studied in
Muhanna and Mullen (2001) and Muhanna, Zhang and Mullen (2007). Moens and
Vandepitte (2005) summarized the interval finite element methods for static and
eigenvalue analyses. In this chapter, we present some results of FEM-based interval
analysis of static responses.
7.2 Interval Analysis of MDOF Systems
In this section, interval analysis of linear equations is summarized following the

notations of Muhanna and Mullen (2001) and Muhanna, Zhang and Mullen (2007).
Consider a set of linear equations formulated as
Ax = b (7.1)
>
where A = (Aij ) is a constant n n matrix, b = (b1 , . . . , bn ) is a constant vector,
and x = (x1 , . . . , xn )> is the variable vector.
Suppose the components of A and b are bounded by intervals as
ALij Aij AU
ij , (i, j = 1, . . . , n) (7.2a)
bLi bi bU
i , (i = 1, . . . , n) (7.2b)
L U
where the superscripts ( ) and ( ) indicate specified lower and upper bounds,
respectively.
The set of A and b satisfying the conditions (7.2a) and (7.2b), respectively,
are denoted by the interval matrix A and interval vector B. The interval values
are denoted by calligraphic letters throughout this chapter. Then the family of
equations (7.1) satisfying (7.2a) and (7.2b) are given as
Ax = b, (A A, b B) (7.3)
The solution set of Eq. (7.3) is denoted by S(A, B) that is not generally an interval
vector. In the following, we assume that all the matrices A satisfying the condition
(7.2a) are regular (nonsingular) so that S(A, B) is bounded. Let S(A, B) denote
the smallest interval components called hull of S(A, B). The hull inverse AH of the
matrix A is defined by
AH B S(A, B) (7.4)
Anti-Optimization via FEM-based Interval Analysis 213
i.e.,
AH B = {A1 b|A A, b B} (7.5)
On the other hand, the matrix inverse A1 of a regular interval square matrix
A is defined as
A1 {A1 |A A} (7.6)
Then the following relation holds:
AH B A1 B (7.7)
where the equality holds if A is diagonal.
The problem of finding S(A, B) is a combinatorial problem that cannot be
solved in a practically acceptable computational time for a large-scale problem.
Therefore, our purpose here is to find the narrowest possible interval vector X
satisfying
AH B = S(A, B) X (7.8)
for which many algorithms exist such as interval Gaussian elimination, that may
break down for a large-scale problem (Nickel, 1977), GaussSeidel iteration (Gay,
1982), and its generalization (Neumaier, 1982; Rump, 1992). The typical approach
utilizes a regular n n transfer matrix R for converting the problem Ax = b to a
problem of finding the fixed point g(x) = x of an operator g(x):
g(x) = x R(Ax b) = Rb + (I RA)x (7.9)
where I is the n n identity matrix (see Muhanna, Zhang and Mullen (2007) for
details of this approach).
As an example, consider the case when uncertainty exists only in some compo-
nents of A as
2 [1, 0] 1.2
A= , B= (7.10)
[1, 0] 2 1.2
By computing x for the four cases (A12 , A21 ) = (1, 1), (1, 0), (0, 1), and
(0, 0), we have

0.4 0.3 0.6 0.6
x = A1 b = , , , (7.11)
0.4 0.6 0.3 0.6
Therefore, AH B is obtained as
H [0.3, 0.6]
A B= (7.12)
[0.6, 0.3]
On the other hand, we can compute A1 for the four cases (A12 , A21 ) =
(1, 1), (1, 0), (0, 1), and (0, 0) as

1 2/3 1/3 1/2 1/4 1/2 0 1/2 0
A = , , , (7.13)
1/3 2/3 0 1/2 1/4 1/2 0 1/2
Therefore, the interval of A1 is obtained
as
1 [1/2, 2/3] [0, 1/3]
A = (7.14)
[0, 1/3] [1/2, 2/3]
and A1 B results in
[0.2, 0.8]
1
A B= AH B (7.15)
[0.8, 0.2]
Hence, the relation (7.7) is satisfied.
U1 U2
P1 P2
K1 K2
Fig. 7.1 A 2-bar structure.
7.3 Interval Finite Element Analysis for Linear Static Problem
As an illustrative example, consider a simple 2-bar structure as shown in Fig. 7.1

subjected to the loads P = (P1 , P2 )> . The extensional stiffnesses of bars 1 and 2
are denoted by K1 and K2 , respectively. The stiffness matrix K is given as

K1 + K2 K2
K= (7.16)
K2 K2
The displacement vector U = (U1 , U2 )> is found from

K1 + K2 K2 U1 P1
= (7.17)
K2 K2 U2 P2
The inverse of K is obtained as

1 1
K1 K1
K 1 =
1 K1 + K 2 (7.18)
K1 K1 K2
Consider the case where the loads have deterministic values P1 = 0, P2 = 1.
The displacements U = (U1 , U2 )> are explicitly obtained as
1 K1 + K 2
U1 = , U2 = (7.19)
K1 K1 K2
Suppose K1 and K2 are interval values K1 = [1.9, 2.1] and K2 = [0.9, 1.1]. Then
[1.9, 2.1] + [0.9, 1.1] [2.8, 3.2]
U2 = = = [1.21, 1.87] (7.20)
[1.9, 2.1] [0.9, 1.1] [1.71, 2.31]
for which the diameter is 0.66. However, this expression gives an overestimation of
the interval of U2 due to the multiple appearance of K1 and K2 in a component of
K 1 as discussed in Sec. 3.3. The exact interval can be found by rewriting U2 as
1 1
U2 = + (7.21)
K1 K2
from which we obtain
1 1
U2 = + = [0.48, 0.53] + [0.91, 1.11] = [1.39, 1.64] (7.22)
[1.9, 2.1] [0.9, 1.1]
and the diameter is reduced to 0.25.
For a general finite element analysis problem, the explicit form of the displace-
ment vector cannot be obtained. Muhanna, Mullen and Zhang (2005) noted that
the accuracy of the interval vector of the displacements depends on the order of the
sequence of arithmetic operations in the solution process by finite element analysis.
They presented an element-by-element approach to delay the operation involving
the coupling of the interval variables, where the penalty terms are used to ensure
the compatibility among the element displacements as described below.
In the standard procedure of finite element analysis, the element stiffness matri-
ces are assembled to formulate the global stiffness matrix. If we write the stiffness
matrix of member i of the 2-bar structure as

Ki Ki
Ki = (7.23)
Ki Ki
then the four components of the matrix are treated as independent variables in the
interval analysis, although they depend on Ki only. If we write K i as

Ki 0 1 1
Ki = (7.24)
0 Ki 1 1
then the number of independent variables in interval analysis is reduced to two, and
the variables exist in a diagonal matrix.
Another source of overestimation of the intervals of displacements is the coupling
of elements through the common nodes; e.g., the term K1 + K2 appears due to the
coupling of the bars at the center node 1 of the 2-bar structure in Fig. 7.1. The
Lagrange multiplier approach can be used to delay the coupling in finite element
analysis (Muhanna and Mullen, 2001). The analysis problem is formulated as a
minimization problem of the total potential energy
1
= U > KU U > P (7.25)
2
under constraints
CU = V (7.26)
where C and V are the specified matrix and vector, respectively, to represent the
compatibility conditions between the elements. Note that the structure of K in
Eq. (7.25) is different from that of the conventional global stiffness matrix; it has
a block-diagonal form of the element stiffness matrices. Accordingly, the nodal
displacements of different elements at the same node are regarded as independent
variables. Although the definitions and sizes of K, U , and P are different from
those for conventional analysis, the same notations are used for simplicity. Let n e
and m denote the number of degrees of freedom of each element and the number of
elements, respectively. Then the size of K is ng ng with ng = ne m, and U and
P also have ng components. For example, for the 2-bar structure in Fig. 7.1, U is
written as
(1) (1) (2) (2)
U = (U0 , U1 , U1 , U2 )> (7.27)
where the superscript ( )(i) denotes the value in the ith bar, and the fixed support
is denoted by node 0. The compatibility conditions are given as
(1)
U
0(1)
1 0 0 0 U 1 0
(2) = (7.28)
0 1 1 0 U1 0
(2)
U2
The stationary conditions for with incorporated constraints (7.26) using the
vector Q of Lagrange multipliers lead to the following system of equations:

K C> U P
= (7.29)
C 0 Q V
The reader may consult Sec. 2.4 for details of the Lagrange multiplier approach.
Suppose the interval form K of K is expressed as K = DS, where D is the
uncertain diagonal matrix, and S is a deterministic matrix as demonstrated in
Eq. (7.24) for an element stiffness matrix. Then the first equation of Eq. (7.29) is
written in an interval form
DSU = P C > Q (7.30)
where P is assumed to be deterministic, for simplicity. Using V = 0 for gen-
eral compatibility conditions without forced displacements, the interval form of the
second equation in Eq. (7.29) is written as
CU = 0 (7.31)
>
which is premultiplied by C and added to Eq. (7.30) to obtain
D(S + C > C)U = P C > Q (7.32)
Eq. (7.32) can be rewritten as
DHU = P C > Q (7.33)
>
with H = S + C C. Hence, U is obtained as
U = H 1 D1 (P C > Q) (7.34)
1
where D is the exact inverse of the diagonal interval matrix D.
Note that Q has the physical meaning as internal force vector. Therefore, for
a statically determinate structure, for which the internal loads do not depend on
the stiffnesses of the elements, the vector P C > Q is deterministic, correspond-
ing to the deterministic (nominal) value of Q. For the interval of displacement
vector, suppose D is formulated so that each interval stiffness parameter appears
only once. In this case, the exact interval U can be successfully computed from
Eq. (7.34), if the structure is statically determinate. However, for a statically in-
determinate structure, the use of deterministic (nominal) value of Q in Eq. (7.34)
leads to underestimation of the interval U. By contrast, the use of Eqs. (7.31) and
(7.32) for computing the interval Q results in overestimation of U. Muhanna and
Mullen (2001) presented an iterative approach to obtain a sharp interval hull for U.
5
K5 /2 K5 /2
4
K4 /2 K4 /2
3
K3 /2 K3 /2
2
K2 /2 K2 /2
1
K1 /2 K1 /2
Fig. 7.2 A 5-story shear frame.
7.4 Interval Finite Element Analysis of Shear Frame
7.4.1 Basic equations

A multi-degree-of-freedom shear frame, e.g., the 5-story shear frame as shown in
Fig. 7.2, is considered. The bending stiffness of the storey columns are taken as
uncertain variables. The bending stiffnesses are modeled using a set-theoretical
interval approach, where the uncertain variable is assumed to be strictly bounded
from below and above. This section summarizes the results of K oyl
uoglu and El-
ishakoff (1998) with modification of notations to be consistent with other parts of
this book.
Finite element models with cubic deterministic shape functions are applied to
discretize the uncertain displacement fields. This yields the 2 2 element stiffness
matrix K e of the shear beam element. When the interval approach is applied, the
22 interval stiffness matrix Ke of the eth element is obtained (K oyl
uoglu, C
akmak
and Nielsen, 1995). Note that, in the following, the stiffness of each story is modeled
by a shear beam element representing the sum of the stiffnesses of the two columns.
All components Ke,ij of Ke are bounded from below and above. The interval matrix
Ke is located symmetrically around a central matrix K ce with a radius matrix K e ,
i.e.,
ce K e , K
Ke = [K ce + K e ] (7.35)
The element stiffness matrices of the shear frame appear in the uncertain set of linear
equations, which is solved using Fuchs method (Fuchs, 1991) to obtain the nodal
displacements in closed form with respect to uncertain story stiffnesses, where the
element stiffness matrix is first diagonalized and then assembled accordingly. Hence,
the lower and upper bounds for the nodal displacements are functionally related to
those of the interval element stiffness matrices. Elishakoff, Ren and Shinozuka
(1997) applied Fuchs method to statically determinate and indeterminate beams
with random stiffness.
According to the BernoulliEuler beam theory, the displacement field v(x) in
the transverse direction of the eth element is related to the loading field p(x) by
[(EI)e (x)v(x),xx ],xx = p(x) (7.36)
where x is the coordinate along the beam axis, (EI)e (x) is the bending stiffness
field, and ( ),x denotes differentiation with respect to x. The assumptions and the
differential relationship of the BernoulliEuler beam theory are assumed to be valid
even when (EI)e (x) possesses some uncertainties. This means that (EI)e (x) is such
that these assumptions are not violated.
In what follows, the uncertain field (EI)e (x) is modeled using set-theoretical
interval approach, where (EI)e (x) is assumed to be bounded from below and above
with constant envelope as
(EI)Le (EI)e (x) (EI)U
e , (0 x Le ) (7.37)
where (EI)Le and (EI)U e are the lower and upper bounds for (EI)e (x), and Le
is the length of element e. The central value of the bending stiffness becomes
(EI)0e = [(EI)Le + (EI)Ue ]/2.
For a shear frame in Fig. 7.2, the rotation at each end of the column element
is fixed. Therefore, the uncertain displacement field V(x) of an element is approx-
imated as a linear combination of the uncertain vector of two-degree-of-freedom
transverse nodal displacements Ve multiplied by the independent vector of cubic
interpolation functions N (x) = (N1 (x), N2 (x))> :
V(x) = N > (x)Ve (7.38)
where
3 2 2 3 2
N1 = 1 3
x + 2 x3 , N 2 = 3 x2 2 x3 (7.39)
Le Le Le Le
Following the standard finite element methodology, the interval stiffness matrix
Ke of the beam element with (i, j)-component Ke,ij is evaluated as
Z Le
d2 Ni (x) d2 Nj (x)
Ke,ij = (EI)e (x) dx (7.40)
0 dx2 dx2
where (EI)e (x) is the interval bending stiffness field. Owing to the uncertainty in
the bending stiffness, all components of Ke are uncertain, yet bounded from above
and below, as follows, as a result of the interval model:
L U
Ke,ij Ke,ij Ke,ij (7.41)
Let Re be defined as
12(EI)e
Re = (7.42)
L3e
with its lower and upper bounds ReL and ReU , respectively. The lower bound K Le and
the upper bound K U e of K e are found, as follows, from minimization/maximization
of the integral in Eq. (7.40)
for each i and j independently
using
Eq. (7.37):
ReL ReU ReU ReL
K Le = , KU e =
(7.43)
U L L U
Re Re Re Re
We consider the eigenvalues 1 and 2 (1 < 2 ) of K e . The independent
computation for each Ke,ij in Eq. (7.43) gives a four-dimensional rectangular prism
L U
with sides defined by the coordinates Ke,ij and Ke,ij , within which K e lies for
any realization of (EI)e (x), and leads to conservative intervals compared to the
intervals where the interdependency is taken into account. For that reason, working
with Eq. (7.43) to obtain the bounds for the eigenvalues of the element interval
matrix leads to very conservative results. Moreover, the smallest eigenvalue that is
obviously equal to zero for any realization of a beam element without constraining
rigid-body motions would not become zero, but become an interval value containing
zero. Thus, the eigenvalues are first computed from the integral of Eq. (7.40)
directly, which leads to
Z Le
72
1 = 0, 2 = (EI)e (x) 6 2 2
dx (7.44)
0 L e (L e 4L e x + 4Le x )
Hence, the lower and upper bounds L2 and U 2 , respectively, on the second eigen-
value 2 can be obtained from minimization and maximization of 2 in Eq. (7.44),
respectively, as
L2 = 14ReL 12ReU, (7.45a)
U L
2 = 12Re + 14Re
U
(7.45b)
0
Then, the central value 2 and the radius 2 of 2 are computed as
L + U
02 = 2 2
= ReL + ReU
2 (7.46)
U L
2 = 13(Re Re )
We next evaluate the displacements of a general multi-story shear frame sub-
jected to static loads. Let K and P denote the global stiffness matrix and the
applied nodal loads, respectively. We assume that K is an interval matrix, but
P is a deterministic vector, for simplicity. Then the nodal displacement vector U
is also an interval vector denoted by U. One can write down the interval nodal
displacements as a function of interval eigenvalues as
U = K1 P = (Q1 )> J 1 Q1 P (7.47)
where J 1 is a diagonal matrix with 2/ 2 for all the diagonal
components, and
1 1
1 1

Q= (7.48)

1 1
1
For further details the reader may consult with K

oyl
uo
glu and Elishakoff (1998).
Then, the lower and upper bounds U L and U U of U can be computed from mini-
mization/maximization of U in Eq. (7.47) as
U L = min[(Q1 )> J 1 Q1 P ] (7.49a)
U 1 > 1 1
U = max[(Q ) J Q P] (7.49b)
In the process of minimization/maximization of Eqs. (7.49a) and (7.49b), subtrac-
tion and addition of intervals appear. It should be noted that when two intervals
are added or subtracted, the resulting interval has a central value which is the sum
or subtraction of the central values of the two intervals, and has a radius which
is the sum of the radii of these two intervals (see Sec. 3.3 for the basic properties
of interval arithmetic). Furthermore, interval arithmetic is basically used to find
bounds for the discrete values of the displacements at the nodes, and these bounds
for the shape functions could only become approximate bounds for the uncertain
displacement field. In order to find bounds for the uncertain displacement field,
one needs to consider an interval validation methodology for the initial differential
equation (7.36) (Alefeld and Herzberger, 1983; Neumaier, 1990).

Consider a 5-story shear frame as shown in Fig. 7.2. Let the height of each story be
1 and (EI)0e = 1/12 so that each story with deterministic properties has a stiffness
Re = 12(EI)0e /L3e = 1, and, from Eq. (7.46), the central value 02 of the eigenvalue
is equal to 2Re = 2. Moreover, for the 5-story shear frame, Q1 can be obtained
in closed form as

1 1 1 1 1
1 1 1 1

1
Q = 1 1 1 (7.50)

1 1
1
Four different loading scenarios are considered. The global load vector is chosen
from P 1 = (1, 1, 1, 1, 1)>, P 2 = (1, 2, 3, 4, 5)>, P 3 = (1, 1, 1, 1, 1)>, and P 4 =
(2, 1, 2, 3, 5)>, which are called Loadings 1, 2, 3, and 4, respectively. Under these
loads, the nodal displacements of the shear frame with deterministic properties are
computed, and the results are listed in Table 7.1.
Suppose the lower and upper bounds for the bending stiffness field are taken
such that ReL = 12(EI)Le /L3e = 0.97, ReU = 12(EI)U 3
e /Le = 1.05. Then Re =
0
12(EI)0e /L3e = 1.01. From Eq. (7.46), we obtain L2 = 0.98 and U 2 = 3.06. The
bounds for the global nodal displacements are given in Table 7.2, which covers all
possible displacements for realizations of the shear frame within the bounds on Re .
In addition to such a relation, non-dimensional quantities wU1 , . . . , wU5 , which are
called coefficients of interval uncertainty and defined as the ratio of the radius to the
Table 7.1 Global nodal displacements of the deterministic

shear frame.
Loading 1 Loading 2 Loading 3 Loading 4
U1 5 15 1 1
U2 9 29 1 0
U3 12 41 0 0
U4 14 50 2 2
U5 15 55 3 7
Table 7.2 Lower and upper bounds for the global nodal displacements of
the shear frame with interval stiffness properties.
U1 [3.27, 10.20] [9,80, 30.61] [0.65, 2.04] [2.04, 0,65]

U2 [5.88, 18.37] [18.95, 59.18] [0,65, 2.04] [1.39, 1.39]
U3 [7.84, 24.49] [26.80, 83.67] [1.39, 1.39] [1.39, 1.39]
U4 [9.15, 28.57] [32.68, 102.04] [5.47, 0,08] [-5.47, 0.08]
U5 [9,80, 30.61] [35.95, 112.24] [-7.51, -0.57] [15.67, -3.19]
Table 7.3 Coefficients of interval dimensions of the global

nodal displacements of the shear frame with interval stiffness
properties.
w U1 0.5149 0.5149 0.5149 0.5149

w U2 0.5149 0.5149 0.5149
w U3 0.5149 0.5149
w U4 0.5149 0.5149 1.0297 1.0297
w U5 0.5149 0.5149 0.8581 0.6620
absolute value of the central value, are obtained as shown in Table 7.3 for U1 , . . . , U5 .
Note that in Table 7.3 means that the coefficient cannot be computed, because
the central value vanishes.
7.5 Interval Analysis for Pattern Loading
In this section, we follow the basic ideas of interval analysis of beams and frames
subjected to pattern loads (K oyl
uoglu, C
akmak and Nielsen, 1995); however, the
notations are modified to be consistent with other parts of the book. Furthermore,
the concepts are presented in a simpler form, and only new numerical examples are
given.
Consider a BernoulliEuler beam subjected to a uniformly distributed load p,
which exists in the interval [p0 p, p0 + p]. Based on the conventional finite
element formulation with cubic displacement interpolation function, the element

stiffness matrix is derived in a standard manner, and the nodal load vector p of an
element corresponding to the uniformly distributed load p is written as

p/2
pL/12
p= p/2
(7.51)
pL/12
The element stiffness matrices and the element load vectors are assembled to
form the n n global stiffness matrix K and the global nodal load vector P ,
respectively, which are defined as interval matrix and vector:
K = [K 0 K, K 0 + K],
(7.52)
P = [P 0 P , P 0 + P ]
Bounds for the displacement vector U = (U1 , . . . , Un )> can be found, as follows, by
using the OettliPrager lemma (Neumaier, 1990), which states that
|K 0 U P 0 | K|U | + P (7.53)
where, e.g., |U | denotes the vector consisting of the absolute values |Ui | of the
components of U .
The sign of Ui , denoted by sign Ui , is defined as

1 for Ui > 0
sign Ui = 0 for Ui = 0 (7.54)

1 for Ui < 0
Suppose the sign of each component of Ui is known. In most of the engineering
problems, the direction of the load vector is fixed, therefore, this is a very reasonable
assumption. Let S be a diagonal matrix with the signs of Ui :
S = diag(sign Ui ) (7.55)
Then the following relations hold for the absolute values of Ui :
(K 0 S K)|U | P 0 P 0 (7.56a)
0 0
(K S K)|U | + P P 0 (7.56b)
Therefore, |Ui | can be conceived as independent variables, and the conservative
upper bounds (lower bounds) for |Ui | can be found by solving a linear programming
problem for maximizing (minimizing) |Ui | under constraints (7.56) and |Ui | 0
(i = 1, . . . , n). Note that we must solve linear programming problems 2n times
to obtain the upper and lower bounds of the absolute values of all displacement
components.
Consider next a pattern loading problem without uncertainty in the stiffness
matrix. The upper and lower bounds for the uniformly distributed load pe on
element e are denoted by pU L
e and pe , respectively. The ith component of the global
2.2
2.1
U2
2.0
1.9
1.8
0.9 1.0 1.1
U1
Fig. 7.3 Bounds of nodal displacements of the 2-bar structure in Fig. 7.1.
displacement vector against the uniformly distributed unit load on element e is

denoted by Die . Then, the maximum and minimum values UiU and UiL of Ui are
obtained from
Xm
U
Ui = max{Die pU e L
i , Di pi } (7.57a)
e=1
m
X
UiL = min{Die pU e L
i , Di pi } (7.57b)
e=1
where m is the number of elements.
On the other hand, an interval can be defined for the nodal load vector P so
that any realization of the set of distributed loads on m elements are contained in
the interval [P L , P U ] = [P 0 P , P 0 + P ], where P 0 is the nominal value and
P is the smallest radius. Then, the extremum values of Ui can be computed by
simple addition of extremum of the two terms as
Xn
UiU = 1
max{Kij (Pj0 + Pj ), Kij
1
(Pj0 Pj )} (7.58a)
j=1
n
X
UiL = 1
min{Kij (Pj0 + Pj ), Kij
1
(Pj0 Pj )} (7.58b)
j=1
For a frame structure, the total number of degrees of freedom is generally larger
than the number of members; i.e., n > m. The number of independent load pa-
rameters is m for Eq. (7.57), and n for Eq. (7.58). Therefore, Eq. (7.58) leads to
a more conservative (wider) interval than Eq. (7.57), as illustrated in the following
simple example.
We first demonstrate a simple interval algebra using the 2-bar structure in
Fig. 7.1, where the stiffnesses are fixed at K1 = 2, K2 = 1; i.e.,

3 1
K= (7.59)
1 1
p1 p2
1 (1) 2 (2)
2 4
1 3
Fig. 7.4 A cantilever beam consisting of two elements subjected to distributed loads.
The nominal values of the loads are P10 = P20 = 1, and the load uncertainty is given
by P1 = P2 = 0.1, which are independent of each other. Using Eq. (7.58), the
intervals of U1 and U2 are obtained as [0.9,1.1] and [1.8,2.2], respectively.
The displacements for the loads at the vertices of its feasible region (0.9,0.9),
(0.9,1.1), (1.1,0.9), (1.1,1.1) are obtained as (0.9,1.8), (1.0,2.1), (1.0,1.9), (1.1,2.2),
respectively. Since the nodal displacements are linear functions of the nodal loads,
the existable region of the nodal displacements (U1 , U2 ) is a convex region as indi-
cated in the gray area of Fig. 7.3, which verifies that the narrowest intervals of U 1
and U2 are [0.9,1.1], and [1.8,2.2], respectively.
Consider next the cantilever beam as shown in Fig. 7.4 consisting of two beam
elements, where the displacement numbers are defined in the lower figure. The
beam is subjected to uniformly distributed loads p1 and p2 on members 1 and 2,
respectively. Suppose the length L and the bending stiffness EI are 1, for simplicity,
for the two members. Then the inverse of the stiffness matrix is obtained as

2 3 5 3
1 3 6 9 6
K 1 = (7.60)
6 5 9 16 12
3 6 12 12
The load vectors corresponding to the nominal values p01 = p02 = 1 of distributed
loads are given as

6 6
0 1 1
, p2 =
0 1 1

p1 = (7.61)
12 0 12 6
0 1
When the radii p1 and p2 of uncertainty are equal to 0.1, the upper bound
P U = P 0 + P and the lower bound P L = P 0 P are obtained as

1.1000 0.9000
0.0167 0.0167
PU = L
0.5500 , P = 0.4500
(7.62)
0.0750 0.0917
Since all components of K 1 are positive, the upper and lower bounds U U and U L
are easily computed from Eq. (7.58) as

0.7959 0.6208
1.3167 1.0166
UU = L
2.2584 , U = 1.7416
(7.63)
1.5167 1.1499
We next use Eq. (7.57) for the same nominal values and radii of p1 and p2 . Then
from the displacements corresponding to the loads (p1 , p2 ) = (1.1, 1.1), (1.1,0.9),
(0.9,1.1), and (0.9,0.9), the following upper and lowers bounds are obtained:

0.7792 0.6375
1.2833 1.0500
UU = L
2.2000 , U = 1.8000
(7.64)
1.4667 1.2000
Therefore, narrower bounds are obtained by evaluating the uncertainty in the dis-
tributed loads; i.e., evaluation after converting to nodal loads leads to a conservative
bound, because the number of independent variables increases from two to four as
a result of conversion.
For additional references to anti-optimization in the finite element context, the

reader may consult, for example, the works by Chen and Yang (2000), Dessombz,
Thouverez, Laine and Jezequel (2001), Elishakoff (1998c), Guo and L u (2001),
Koyl
uo
glu and Elishakoff (1998), Matsumoto and Iwaya (2001), Muhanna and
Mullen (2002, 2005), Muhanna, Kreinovich, S olin, Chessa, Araiza and Xiang (2006),
Muhanna, Mullen and Zhang (2005), Muhanna, Zhang and Mullen (2007), Mullen
and Muhanna (2002), Nakagiri (1995, 1996), Nakagiri and Suzuki (1996, 1997,
1999), Nakagiri, Hoshi and Yamada (2003), Nakagiri and Yoshikawa (1996), Rao,
Chen and Mulkay (1998), Saxena (2003), Suzuki and Otsubo (1995), Vitali, Blanc,
Larque, Duplay, Morvan, Nakagiri and Suzuki (1999), and Zhang (2005).

Chapter 8
Anti-Optimization and Probabilistic

Design
The worst-case and probabilistic world should not be opposed, but

considered together. (Ben-Tal, El Ghaoui and Nemirovski, 2006)
Make the best out of the worst. (English proverb)
This chapter is devoted to systematic comparison of probabilistic and anti-

optimization approaches in the dynamic environment. We will consider impact
buckling of uniform columns possessing uncertain initial imperfections. A single
imperfection parameter is considered first, and then the methodologies and results
are extended to the case of multiple imperfection parameters.
8.1 Introduction
In Chapter 1, we have already found, albeit on a simple example of a bar subjected

to a bounded random load, that the worst-case scenario and probabilistic methods
yield coincident or close results. In this chapter, less transparent problems are
considered.
The probabilistic modeling of various kinds of uncertainties in the dynamic
buckling setting was dealt with by Goncharenko (1962), Budiansky and Hutchin-

son (1964), Lindberg (1965), Ariaratnam (1967), Kildibekov (1971), Amazigo and
Frank (1973), Amazigo (1974), Lockhart and Amazigo (1975), Maymon and Libai
(1977), Elishakoff (1978b), Bogdanovich (1987), and others. Convex modeling of
the dynamic buckling problems was facilitated by Ben-Haim and Elishakoff (1989a),
Elishakoff and Ben-Haim (1990a), Lindberg (1992a, 1992b), and Ben-Haim (1993a).
These two avenues of thoughts on uncertainty modeling, namely, the probabilistic
and anti-optimization ones, have not been compared in either of the above studies.
This will be the principal objective of the present chapter. The direct compari-
son of designs yielded by two alternative approaches sheds light on the possible
compatibility or incompatibility of these two alternative approaches.
According to Einstein, As far as the laws of mathematics refer to reality, they
227
are not certain, and as far as they are certain, they do not refer to reality. How
much more does this statement appear to apply to engineering, where numerous
uncertain factors are encountered almost in every problem? The pertinent question
arises of how to model this uncertainty. It is nowadays widely accepted that there
are three competing methods of describing uncertainty, as discussed in the preface of
this book; namely, the probabilistic method, the fuzzy-sets based approach, and the
method which is referred to as anti-optimization (Elishakoff and Ben-Haim, 1990a)
or as the guaranteed approach by Kurzhanski (1977) and by Chernousko (1994).
Each of these approaches has generated numerous research papers and monographs.
Yet the systematic comparisons among these three approaches are still missing.
Naturally, there may be a strong argument put forward against such a comparison,
in the first place. Indeed, these three approaches provide judgments of different
kinds: The probabilistic approach furnishes the reliability of the structure, namely,
the probability that the stresses, strains or displacements of the structure will not
take values over some safe threshold. The fuzzy-sets based approach yields the
safety margin via the concept of a membership function. Anti-optimization (or the
guaranteed approach) provides the least favorable, maximum displacements which
can be contrasted with the threshold values. If the maximum response does not cross
the threshold, we can guarantee that the failure will not occur. This corresponds to
the statement that if the least favorable, or an anti-optimized, response is below the
failure boundary, a successful performance will be secured. This particular thought,
namely, that three approaches provide different types of answers, precluded the very
attempt to compare them for a long time.
The first numerical comparison has been performed independently by Elishakoff,
Cai and Starnes (1994a) and Kim, Ovseyevich and Reshetnyak (1993) in different
but complementary contexts: The former study dealt with nonlinear static buck-
ling, whereas the latter investigated linear dynamic response. In the paper by
Elishakoff, Cai and Starnes (1994a), for certain regions of variations of the pa-
rameters, the design values of the buckling loads obtained via applications of two
alternative approaches were found to be extremely close to each other. In the paper
by Kim, Ovseyevich and Reshetnyak (1993), the design value corresponding to the
mean plus three times standard deviation, was adopted. Elishakoff and Li (1999)
studied a benchmark static problem, leading to direct contrast of the probabilis-
tic and non-probabilistic analyses of uncertainty. General understanding on the
limitation of probabilistic methods is summarized in Elishakoff (2000a). Langley
(2000) presented a unified framework of combined probabilistic and possibilistic
(deterministic) approaches.
In Secs. 8.2 and 8.3, probabilistic and anti-optimization approaches are compared
for impact buckling of uniform columns possessing single and multiple uncertain
initial imperfections following the studies by Elishakoff and Zingales (2000) and
Zingales and Elishakoff (2000), respectively.
Anti-Optimization and Probabilistic Design 229
Initial imperfection amplitude
Additional displacement amplitude
P(t) P(t)
h
w
L/2 Total displacement
Fig. 8.1 Description of the structural model.
8.2 Contrasting Probabilistic and Anti-Optimization Approaches
8.2.1 Problem formulation

Let hti denote a singularity function, namely, a unit step function defined as

0, t < 0
hti = (8.1)
1, t 0
where t is the time. The differential equation for the uniform column as shown in
Fig. 8.1 under axial impact load
P (t) = P hti (8.2)
reads
4w 2w 2w 2 w0
EI + P (t) + mJ = P (t) (8.3)
x4 x2 t2 x2
where x is the axial coordinate, E is Youngs modulus, I is the moment of inertia,
m is the mass density, J is the cross-sectional area, w0 (x) is the initial imperfection
constituting a small deviation of the initial shape of the unstressed column, and
w(x) is the additional transverse deflection of the axis of the column, so that
wT (x, t) = w0 (x) + w(x, t) (8.4)
represents the total displacement.
The boundary conditions for the column that is simply supported at both ends
read
w(x, t) = 0 at x = 0 and x = L (8.5a)
2
w(x, t)
= 0 at x = 0 and x = L (8.5b)
x2
where L is the length of the column. The initial conditions are
w(x, t)
w(x, t) = 0 and = 0 at t = 0 (8.6)
t
We confine ourselves to the case where the initial imperfection w0 (x) has a multi-
plicative representation
x
w0 (x) = h sin (8.7)
L
where h is the magnitude of initial imperfection. Following Elishakoff (1978a) we
will use the non-dimensional quantities
x w0 w P
= , = 1 t, u0 = , u= , = (8.8)
L Pcl
where is the non-dimensional axial coordinate, is the non-dimensional time,
u0 () is the non-dimensional initial displacement, u(, t) is the non-dimensional
additional displacement, and is the non-dimensional axial load. In Eq. (8.8)
2 r EI r
I EI 2
1 = , = , Pcl = (8.9)
L mJ J L2
where 1 is the fundamental circular natural frequency of the ideal column, i.e., of
a column with neither initial imperfections nor axial load, is the radius of inertia,
and Pcl is the classical Euler buckling load. The differential equation (8.3) reduces
to the following non-dimensional one:
2 2 2
4u 2 u 4 u 2 u0
+ + = (8.10)
4 2 2 2
We introduce the non-dimensional initial imperfection g as
h
g= (8.11)

The additional displacements are expressed in a separable form analogous to
Eq. (8.7), namely,
u(, ) = e() sin() (8.12)
where e() is a time-dependent function. Substituting Eqs. (8.7) and (8.12) in
conjunction with Eq. (8.11) into Eq. (8.10) yields the following ordinary differential
equation with respect to the function e():
d2 e()
+ (1 )e() = g (8.13)
d2
The solution satisfying the initial conditions (8.6) is obtained as
g

(cosh(r) 1), > 1

1

1 2
e1 () = g , =1 (8.14)

2
g

(1 cos(r)), <1
1
where
p
r= | 1| (8.15)
The non-dimensional total displacement z(, ) equals

1
wT (x, t) = () sin()
z(, ) = (8.16)

with () being a function of a non-dimensional time alone

g 1
() = cosh(r) , > 1, (8.17a)
1
g(2 + 2)
() = , = 1, (8.17b)
2
g 1
() = cos(r) , <1 (8.17c)
1
The next sections will deal with various uncertainty analyses.
8.2.2 Probabilistic analysis

We treat the uncertain initial imperfection magnitude as a random variable with
specified probability distribution function
FG (g) = Prob(G g) (8.18)
We are interested in finding the survival reliability of the structure, which is de-
fined in general form as the probability that the structure will perform its specified
mission. Herewith we adopt Hoffs criterion (Hoff, 1965):
A structure is in a stable state if admissible finite disturbances of initial
state of static or dynamic equilibrium are followed by displacements
whose magnitude remains within allowable bounds during the required
lifetime of the structure.
Within this definition, we introduce the admissible state as
d z(, ) c (8.19)
where c is the non-dimensional upper bound, and d is the non-dimensional lower
bound of the safety region. On the other hand, the inequalities
max z(, ) > c or min z(, ) < d (8.20)

mean unsafe regions, or, as defined by Hoff, as regions of impact buckling. Here-
inafter we denote random variables with capital letters, whereas the possible values
they take on are designated with lower-case notation. Reliability R() at a pre-
selected non-dimensional time is defined as a probability that the event described
in Eq. (8.19) takes place:
R() = Prob(d Z(, ) c) (8.21)
Define V () as
V () = Z(1/2, ) (8.22)
Then the reliability becomes
R() = Prob(d V () c) (8.23)
Note that V () is a random function of , and is represented as a product
V () = Ga() (8.24)
where G is a random variable and a() is a deterministic function as

1
a() = cosh(r) , > 1, (8.25a)
1
2 + 2
a() = , = 1, (8.25b)
2
1
a() = cos(r) , < 1 (8.25c)
1
Reliability, as per in Eq. (8.21), becomes
R() = Prob(V () c) Prob(V () d)

c d (8.26)
= Prob G Prob G
a() a()
since a() is a positive-valued function. In terms of probability distribution function
FG (g), Eq. (8.26) is rewritten as

c d
R() = FG FG (8.27)
a() a()
Equation (8.27) depends, in general, on the specific form of the probability distribu-
tion function FG . The influences of the specified probability distribution function
over the design of the geometric characteristic of the column will be the objective
of the next sections.
8.2.3 Uniformly distributed random initial imperfections

Consider first the case when the initial imperfection amplitude G for the sinusoidal
imperfection in Fig. 8.1 possesses a uniform probability density function over the
interval [1 , 2 ] and vanishes elsewhere. The probability distribution function FG (g)
then reads
1
FG (g) = (hg 1 i1 hg 2 i1 ) (8.28)
2 1
where hg i1 is a singularity function as

1 0, g<
hg i = (8.29)
g , g
The reliability of the column becomes

1 1 !
1 c c
R() = 1 2
2 1 a() a()
1 1 ! (8.30)
1 d d
1 2
2 1 a() a()
Hereinafter, for the sake of simplicity, we take 1 0, 2 0. Since the arguments
in the third and fourth terms in Eq. (8.30) are negative, they vanish. Therefore, we
have the following expression:
1 1 !
1 c c
R() = 1 2 (8.31)
2 1 a() a()
We are interested in designing the system; namely, choosing the radius of a circular
solid cross-section such that the reliability R() will be no less than some codified
reliability rc as
R() rc (8.32)
Consider first a sub-case when the applied axial load exceeds the classical buckling
load, so that
P P L2 M 4P L2
= = 2 = 4 > 1, M = 3 (8.33)
Pcl EI E
where M is a load-dependent parameter. By virtue of Eq. (8.25), the reliability
reads
1
1 c(1 4 /M )
R() = 1
2 1 cosh(r) 4 /M
1 (8.34)
1 c(1 4 /M )
2
2 1 cosh(r) 4 /M
Let us closely investigate this expression. If, for the pre-selected non-dimensional
time instant , the argument of the singularity functions in Eq. (8.34) satisfies the
inequality
c(1 4 /M )
< 1 (8.35)
cosh(r) 4 /M
then the reliability vanishes identically. On the other hand, if
c(1 4 /M )
> 2 (8.36)
cosh(r) 4 /M
then the reliability equals unity. In the intermediate range, i.e., for
c(1 4 /M )
1 2 (8.37)
cosh(r) 4 /M
the reliability varies between zero and unity. Generally, in engineering practice, the
codified reliability rc is not taken to be unity. Rather, it is a pre-selected quantity in
extremely close vicinity of unity. For rc tending to unity from below, the inequality
in Eq. (8.37) is valid. Hence the argument in the second term in Eq. (8.34) is
negative. Therefore, it vanishes, and we are left with
1
1 c(1 4 /M )
R() = 1 (8.38)
2 1 cosh(r) 4 /M
In order for the reliability to tend to unity from below, it is necessary and sufficient
that
c(1 4 /M )
lim = 2 (8.39)
d cosh(r) 4 /M
where d is the design value of the cross-sectional radius. For this we obtain the
following equation:
c(1 4d /M )
p = 2 (8.40)
cosh( M/4d 1) 4d /M
It should be noted that the non-dimensional time parameter is proportional to

the radius, through Eq. (8.9); i.e., is expressed using t = /1 :
" r #
1 2 E
= td (8.41)
2 L m
Equation (8.40) is a transcendental equation in terms of d , once the time instant t

at which reliability rc is desired, and the failure boundary c and other parameters are
fixed. Design process requires us to obtain the cross-sectional radius d as a function
of these quantities, as d = d (t, c, E, m, L) either analytically or numerically. It is
advantageous, however, to view Eq. (8.40) as a function of the failure boundary c
depending upon the radius d and the remaining parameters:
p
2 [cosh( M/4d 1) 4d /M ]
c(d , t, E, m, L) = (8.42)
1 4d /M
This relationship enables the design of the column graphically. One obtains a curve
c versus d with other parameters fixed. Once such a curve is drawn, the deter-
mination of the required cross-sectional radius d becomes a straightforward task:
One specifies the failure boundary c and reaches the corresponding radius in the
graph.
Two other sub-cases, associated with P = Pcl or P < Pcl , respectively, are dealt
with in a perfect analogy. For P =Pcl , the design value of the cross-sectional radius
is immediately expressed as d = 4 M . Then

2 4 E 2
c(d , t, E, m, L) = t M +2 (8.43)
2 L m
c = 0.5
0.8 1 = 0.1
2 = 0.6
0.6
R()
0.4
0.2
R = 0.133
0 1 2 3 4 5 6

Fig. 8.2 Reliability versus non-dimensional time; initial imperfections with uniform probability
density (P = 0.5Pcl ), Eqs. (8.25c) and (8.30).
c = 0.5
0.8 1 = 0.1
2 = 0.6
0.6
R()
0.4
0.2
0 0.5 1 1.5 2 2.5 3 3.5

density (P = Pcl ), Eqs. (8.25b) and (8.30).
whereas for P < Pcl this equation takes the following form:
p
2 [4d /M cos( M/4d 1)]
c(d , t, E, m, L) = (8.44)
1 4d /M
The results of sample calculations are presented in Figs. 8.28.4 portraying relia-
bility R() versus non-dimensional time for P < Pcl , P = Pcl , and P > Pcl , respec-
tively, for c = 0.5, 1 = 0.1, 2 = 0.6. In the following, we assume the mechanical
c = 0.5
0.8 1 = 0.1
2 = 0.6
0.6
R()
0.4
0.2
0 0.5 1 1.5 2

density (P = 2Pcl ), Eqs. (8.25a) and (8.30).
3.5
Probability of initial failure
3
2.5
fG (g) 2
1.5
0.5
0 0.2 0.4 0.6 0.8

g
Fig. 8.5 Geometrical representation of the reliability at the initial time instant.
and geometrical parameters of the column as L = 10.0 m, E = 2.1 106 kg/cm2 ,

m = 2.1 103 /9.81 kg s2 /cm4 . The axial load P is set at P = 3000 kg. As it is
seen from Fig. 8.2, R() does not assume a value of zero or unity. At time instant
zero, the reliability equals 0.8. This value is deducible directly; indeed, probability
density of the initial imperfections for g [0.1, 0.6] is
1
fG (g) = =2 (8.45)
0.6 0.1
as shown in Fig. 8.5. Hence, the area under the probability density curve between
Boundary c (cm)
3 II
1 I
c = 0.6
4
M = 3.86
0 2 4 6 8 10
Radius (cm)
Fig. 8.6 Design curve c = c(d ), t = 0.2 with G having a uniform probability density and unity
reliability requirement, Eq. (8.44).
the values c = 0.5 and the uppermost value that G may take, namely, 0.6, equals
2(0.6 0.5) = 0.2, which constitutes the probability of failure Pf () at the initial
time = 0 denoted by Pf (0) as shown in Fig. 8.5. This means that the reliability
curves in Figs. 8.28.4 have an initial value equal to R(0) = 1 Pf (0) = 0.8. This
probability is equivalent to the failure probability of the item at delivery.
For the actual load less than the classical buckling load, the variation of the total
displacement with time is harmonic. When approaches value /r, the absolute
value of the total displacement attains its maximum. At this time, the reliability
of the system reaches its minimum. Values of larger than /r yield a constant
value of reliability achieved at the latter time instant
R() = R(/r) for /r (8.46)
In Fig. 8.2 the value /r equals 4.1. Therefore, the reliability in Fig. 8.2 remains
constant and equals 0.15. Figure 8.6 shows the relation between the failure boundary
and the cross-sectional radius d with other parameters fixed. This curve enables us,
by specifying the failure boundary c, to obtain the smallest required cross-sectional
radius yielding the prescribed reliability for the prescribed external load. One must
observe that with values of c less than 2 = 0.6, it is not possible to design the
column in a manner so as to attain a unity or near-unity reliability. In fact, if
approaches infinity, the column gains an infinite frequency of vibration and the
displacement at the center remains fixed at the initial imperfection value. Naturally,
if the failure boundary happens to be less than or equal to 1 , i.e., the minimum
value that the initial imperfection can assume, there is no finite value of d that
will ensure extremely high reliability, since the structure buckles at the outset of
the experiment. Likewise, if c [1 , 2 ) the reliability that is arbitrarily close to
8
6 5
Boundary c
4
4
2
3
Time t (sec)
2 2
4
6 1
Radius (cm)
8
10
Fig. 8.7 Design surface c = c(d ), with G having a uniform probability density and unity relia-
bility requirement, Eq. (8.44).
unity cannot be obtained, since the non-zero probability of failure occurs. On the
other hand, when c 2 , the initial probability of collapse is absent, and the design
with near-unity reliability is feasible.
Figure 8.7 depicts the surface = (c, t) for a prescribed value of the external
load and unity reliability. With increase of the time interval [0, t], where safe perfor-
mance is required, the value of the failure boundary allowing unity reliability should
be greater. Naturally, values of d are larger than the ones obtained from such a
surface for the pre-selected time instant t corresponding to unity reliability. This
behavior is elucidated in Fig. 8.8, where the design curves are presented for various
values of the design times t1 = 0.3 sec, t2 = 0.8 sec, and t3 = 2.5 sec. Figure 8.9
represents the cross-section of the surface in Fig. 8.7 with the plane t = 0.5 sec
(solid line), contrasted with other design curves, which are obtained with smaller
values of the required reliability, namely, rc = 0.9 or 0.8, at the same time instant.
It is seen that, by requiring less reliability, the design value of the cross-sectional
radius decreases, as expected, for the same pre-selected value of the safety bound c.
We note that in Figs. 8.6 and 8.8 the abscissa axes are divided into two sub-
regions corresponding to different behaviors of the system. We denote the value 4 M
by dc . If the design radius d satisfies d dc , then the load ratio P/Pcl satisfies
P/Pcl 1, corresponding to region I. In this region, the displacement max z(, )
is an exponential function of the time t. The design relationships, conveniently
expressed as the failure boundary function depending on the variables d , t, E, m
and L as c(d , t, E, m, L), takes the form of either Eq. (8.42) or Eq. (8.43). The
6 Solid line t =0.3 sec
Boundary c (cm)
Dashed line t 0.8 sec
5
Dot-Dashed line t = 2.5 sec
4
2 II
1 I
0 2 4 6 8 10
Radius (cm)
Fig. 8.8 Design curves c = c(d ), with G having uniform probability density and unity reliability
requirement for different safe time instants, Eq. (8.44).
Solid line rc = 1
4 Dashed line rc = 0.83

Boundary c (cm)
Dot-Dashed line rc = 0.75
II
2
1 I
0 2 4 6 8 10
Radius (cm)
Fig. 8.9 Comparison of design curve, with G having uniform probability density and different
codified reliabilities, t = 0.5, Eq. (8.44).
argument of the hyperbolic function in Eq. (8.42), with the aid of Eq. (8.41), takes
the form
s p r
M (d )4 M 2 E
1 =t (8.47)
(d )4 d L m
By inspecting Eqs. (8.42) and (8.47), we conclude that the failure boundary function
in Eq. (8.42) is a monotonically decreasing function of the cross-sectional radius d
in region I as shown in Figs. 8.68.8. On the other hand, if d > dc ; i.e., region II
in Figs. 8.68.8, then the displacement function is represented by a non-monotonic
trigonometric function Eq. (8.17c). In this case, the function c(d , t, E, m, L) is
given in Eq. (8.44) with the attendant argument of the cosine function in Eq. (8.47).
Inspecting the expression of the function c(d , t, E, m, L), we note that its maximum
value is reached for the first time when the expression in Eq. (8.47) reaches . If,
for a combinations of design radius d and time t, the expression in Eq. (8.47) has
a value larger than , then the displacement in Eq. (8.17c) has already reached
its maximum value at a time t < t. Therefore, in order for the column to perform
satisfactorily with the required reliability until the specified time t, the trigonometric
function in Eq. (8.44) must acquire the value cos() = 1, and Eq. (8.44) becomes
d /M + 1
c = 2 (8.48)
d /M 1
Interestingly, an asymptotic behavior of the design curves may be observed from
Figs. 8.68.8. The asymptotic value of the cross-sectional radius d corresponds to
an infinite value of the first natural circular frequency of vibration 1 in Eq. (8.9).
In these circumstances the column behaves statically at every time instant t. Unity
reliability R() is achieved by setting the failure boundary c equal to the maximum
value of the initial imperfection g. The above discussion is confirmed analytically
by calculating the limit of the expression as
lim c = (2 1 )rc + 1 (8.49)
d
which, in case the specified reliability rc = 1 , will yield 2 as the final result.
8.2.4 Random initial imperfections with truncated exponential dis-

tribution
Assume that the probability density function differs from the uniform one, although
initial imperfection amplitude G varies in the interval [1 , 2 ], namely, G is supposed
to have a truncated exponential probability density function
fG = A exp(qg)[hg 1 i hg 2 i ] (8.50)
where
1
A= (8.51)
exp(q1 ) exp(q2 )
is the normalization coefficient, and h i is the singularity function defined in
Eq. (8.1).
The parameter q is a deterministic positive number characterizing the probabil-
ity distribution. For 1 = 0 and 2 tending to infinity, q is the reciprocal to the
mathematical expectation of the random variable G. The probability distribution

function, obtained by integrating Eq. (8.50), reads
FG (g) = A[exp(q1 ) exp(qg)]hg 1 i
(8.52)
A[exp(q2 ) exp(qg)]hg 2 i
In this new circumstance, Eq. (8.27) involves the probability distribution function
in Eq. (8.52). Hence,

c
R() = FG
a()

c
= A[exp(q1 ) exp(qc/a())] 1 (8.53)
a()

c
A[exp(q2 ) exp(qc/a())] 2
a()
Note that a() is given in Eq. (8.25); it depends upon the ratio of the external
load P to the classical buckling load Pcl . Figure 8.10 depicts the reliability function
for the case P/Pcl = 0.5 versus the non-dimensional time . It is seen that, as
approaches infinity, the reliability R() attains a constant non-vanishing value. In
Figs. 8.11 and 8.12 the reliability R() is portrayed for the cases P/Pcl = 1 and
P/Pcl = 2, respectively. In these cases, the reliability tends to zero when .
This behavior is explained through closer examination of Eqs. (8.25) and (8.53).
Inspecting Eq. (8.53), we find that a vanishing value of the reliability R() can be
obtained only if
c c
1 0 and 2 0 (8.54)
a() a()
In the case = P/Pcl = 0.5, as shown in Fig. 8.10, the function a() assumes
the expression in Eq. (8.25c), i.e., a harmonic dependence on the non-dimensional
time . We denote by m = /r, which is the value of the non-dimensional time
corresponding to the minimum of the term c/a(). We conclude that if both con-
ditions hold in Eq. (8.54) for m , then the reliability will vanish. If, instead,
the conditions in Eq. (8.54) are not fulfilled for the non-dimensional time in the
range 0 m , the reliability R() never attains zero. For the non-dimensional
time > m , the reliability R() maintains its minimum value achieved at = m .
The minimum non-vanishing value of the reliability can be analytically evaluated
by calculating

c +1
lim R() = lim FG = A exp(q1 ) exp qc (8.55)
a() 1
For 1 = 0.1, 2 = 0.6, c = 0.5 and q = 0.1, this value is 0.233 as shown in
Fig. 8.9. In the cases = P/Pcl = 1 and = P/Pcl = 2 of Figs. 8.10 and
8.11, respectively, a() is a monotonically increasing function with respect to the
non-dimensional time. Hence there always exists a value c such that both of the
conditions in Eq. (8.54) are satisfied for c , resulting in a vanishing limiting
c = 0.5
0.8 1 = 0.1
2 = 0.6
0.6
R()
0.4
R = 0.233
0.2
0 1 2 3 4 5 6

Fig. 8.10 Reliability versus non-dimensional time, with G having truncated exponential proba-
bility distribution (P = 0.5Pcl ), Eqs. (8.25c) and (8.53).
c = 0.5
0.8 1 = 0.2
2 = 0.6
0.6
R()
0.4
0.2
0 0.5 1 1.5 2 2.5 3 3.5

bility distribution (P = Pcl ), Eqs. (8.25b) and (8.53).
value of reliability R(). We are interested in the design of a column with near-
unity reliability. It may be found from Eq. (8.53) that the necessary and sufficient
condition to achieve this value reads
c
2 (8.56)
a()
where the function a() assumes one of the three expressions (a), (b) and (c) in
c = 0.5
0.8 1 = 0.1
2 = 0.6
0.6
R()
0.4
0.2
0 0.5 1 1.5 2

bility distribution (P = 2Pcl ), Eqs. (8.25a) and (8.53).
Solid line r0 = 1
6
Dashed line r0 = 0.83
5 Dot-Dashed line r0 = 0.75
Boundary c (cm)
3 II
2
I
1
0 2 4 6 8 10
Radius (cm)
Fig. 8.13 Comparison of design curves for different codified reliabilities, t = 0.5 with imperfections
possessing truncated exponential distribution, Eq. (8.44).
Eq. (8.25) depending upon the value of the external load P . Specifying the appro-
priate expression for a() and choosing the time interval [0, t] required for successful
performance, Eq. (8.56) reduces to either Eq. (8.42), Eq. (8.43) or Eq. (8.44), re-
spectively, for P > Pcl , P = Pcl or P < Pcl . Thus we conclude that, although
the imperfections possess an exponential distribution curve, the curve in Fig. 8.6
obtained for uniformly distributed imperfections can be used for the column design
with near-unity reliability.

Several design curves corresponding to different values of reliability are portrayed
in Fig. 8.12, all corresponding to the same pre-selected time instant t. An inspection
of Fig. 8.13 reveals that lower reliability demand corresponds to smaller values of
the cross-sectional radius d , for a given value of the failure boundary c. A natural
question arises: Is it possible that, for a near-unity reliability requirement, the same
design curve will be obtained for initial imperfections distributed with a completely
generic, but bounded, probability density function? The answer to this question
will be provided in the following section.
8.2.5 Random initial imperfection with generic

truncated distribution
Let us assume that the initial imperfection amplitude is a random variable with a
generic probability density function fG (g) defined as follows in the interval [1 , 2 ]
with 1 and 2 both positive:

AfG (g) for g [1 , 2 ]
fG (g) = (8.57)
0 elsewhere
where A is a normalization coefficient such that
Z
fG (g)dg = 1 (8.58)

and fG (g) is the generic probability density function defined in (, ). With the
aid of singularity functions, Eq. (8.57) can be written as

fG (g) = AfG (g)(hg 1 i hg 2 i ) (8.59)
The probability distribution function FG (g) is obtained through formal integration
of Eq. (8.59), reading
Z g Z g

FG (g) = fG (y)dy = AfG (y)dy (8.60)

The reliability R() of the column is given through Eq. (8.27) with FG (c/a())
represented by Eq. (8.60).
In order to design a column with an extremely high reliability requirement, it is
sufficient that the argument c/a() 2 , resulting in
Z c/a()
c
lim FG = lim AfG (g)[hg 1 i hg 2 i ]dg
c/a()2 a() c/a()2
Z c/a()

= lim AfG (g)[hg 1 i hg 2 i ]dg
c/a()2 1
Z 2

= AfG (g)dg (8.61)
1
=1
Necessity could be demonstrated as follows: If c/a() equals some e that is less

than 2 , then the reliability equals F (e
) that is smaller than unity. Hence it is
necessary that the argument c/a() approaches 2 to obtain unity reliability.
To sum up, in this section, we established that, for the imperfection amplitude
having either uniform distribution or truncated exponential distribution, the near-
unity reliability design requires that c/a() tends to 2 . Now, we observe that
the analogous conclusion holds without specifying the type of probability distri-
bution, except that its density is bounded between 1 and 2 . Thus, irrespective
of the specified character of the bounded density, the same design value of the
cross-sectional radius d is achieved. This independence of the design from the
probabilistic contents appears to be remarkable. It leads to the following question:
Since the obtained designs are independent of the particular form of the probabil-
ity density, may it be that the same result is obtainable without resorting to the
probability notion at all? The reply to this question is affirmative as will be shown
below.
8.2.6 Buckling under impact load: Anti-optimization by interval

analysis
Let us postulate that the initial imperfections constitute an uncertain variable,
although not a random one. Specifically, we assume that imperfections constitute
an interval variable
g = [1 , 2 ] (8.62)
with 1 and 2 both positive. The interval analysis is a mathematical tool dealing
with intervals rather than with numbers (see Sec. 3.3 for details of interval analysis).
In the present context, the analysis is straightforward, since the response function
(), as seen in Eq. (8.17), is a linear function of the initial imperfection amplitude
g. We define the functions 1 () and 2 () as the lower and upper envelopes of the
response, respectively. These constitute
the interval
responsefunction [()] as

1

[1 , 2 ] cosh(r) , >1

1

2 + 2
[()] = [1 (), 2 ()] = [1 , 2 ] , =1 (8.63)

2

1

[1 , 2 ] cos(r) , <1
1
Inspection of Eq. (8.63) reveals that the interval displacement function of non-
dimensional time is represented as a product of an interval variable [1 , 2 ] and
a deterministic function a() defined in Eq. (8.25). In order to design the column,
i.e., to guarantee its successful performance until the pre-selected time t, or its
non-dimensional counterpart , it is necessary and sufficient to bound the largest
possible displacement to satisfy the inequality
2 () c (8.64)
where c is the failure boundary. Bearing in mind Eq. (8.63) in conjunction with
Eq. (8.25), we obtain
a()2 c (8.65)
We note that the design condition in Eq. (8.65) coincides formally with Eq. (8.56),
when the required reliability approaches unity from below. This proves the equiva-
lence between the results furnished by interval analysis and probabilistic one, when
rc attains values extremely close to unity.
8.2.7 Conclusion
A one-dimensional impact-buckling problem has been studied in this section in
order to contrast the probabilistic and non-probabilistic analyses of uncertainty.
Simplicity of the example allows us not to lose sight of the forest for the trees of
analytical and numerical derivations. In particular, this example illustrates that
probabilistic analysis and non-probabilistic interval analyses are compatible with
each other. In probabilistic analysis, one postulates the knowledge of the probability
density function. However, such information is often unavailable. To model this
situation, we constructed two alternative models of probability density utilizing
a random variable with either uniform or non-uniform truncated density. Both
yielded the same value for the design variable, if the required reliability is extremely
close to unity. This result leads to the general conclusion that irrespective of the
density involved, if it extends between the lower possible value 1 and the upper
possible one 2 , the same designs will follow. This could be interpreted as good
news for the probabilists. Researchers in the field of stochasticity could argue
that the unavailability of the probability density, except the information on its
boundedness, may be immaterial. Thus any assumption on the probability density
will be equally efficient. If, however, the experimental data is available, then an
appropriate density should be utilized. We maintain that a design point at the
bound of the distribution provides a near-unity probability. This seems blatantly
transparent; yet this fact appears not to be emphasized by the probabilistic analysts.
Even though the knowledge of the probability density function is not needed in
interval analysis, one must face the issue on how to properly define bounds. Except
in special cases where there exist well defined bounds, the selection of the bounds
will likely rely on a decision, perhaps based on data analysis or physical arguments.
Remarkably, the same designs furnished by the probabilistic analysis are also
obtained through the use of the interval analysis. This suggests the following idea:
Since the initial imperfection is an uncertain value, and often no data is available
to justify the particular choice of the density, why not employ the simplest possible
approach? Interval algebra is indeed such a method. Although in this particular
case interval analysis may appear simple, it provides the same result as derived by
more sophisticated and algebraically lengthier probabilistic analysis. It is concluded,
therefore, due to engineering pragmatism, that in some circumstances the non-
probabilistic analysis of uncertainty may be preferable to the probabilistic one,

since it leads to the same results in a much simpler manner.
8.3 Anti-Optimization Versus Probability: Vector Uncertainty
8.3.1 Introduction
In Sec. 8.2, a direct comparison was performed between the widely popular prob-
abilistic methods and interval analysis, which represents a simplest version of the
anti-optimization technique. Immediately a natural question arises: How can we
compare these two competing approaches in the case of the vector uncertainty?
This problem is investigated in this section.
Whereas in a single-dimensional case uncertainty can be identified as a variable
belonging to an interval, the two-or-more-dimensional case presents more possibil-
ities. For example, in a two-dimensional case, the uncertain vector is identified
with two coordinates varying in a rectangular region, which enables one to use in-
terval analysis in a vector setting. It is intuitively understood that the narrowest
interval shall be chosen to characterize the one-dimensional uncertainty; for the
two-dimensional case one should seek a rectangle of minimum area so that the fur-
ther evolution of the system will be more closely bracketed. Yet, this may not be a
best representation of the available data whose scatter must be modeled. Indeed, in
some cases, the use of regions other than the rectangle may result in an even smaller
area enclosing all available data. The possibility arises, for example, of enclosing
the data by the minimum area ellipse, whose area may turn out to be smaller than
that of the minimum-area rectangle. Along these thoughts, in addition to interval
analysis, ellipsoidal modeling was developed (Schweppe, 1973; Chernousko, 1994;
Kurzhanski and Valyi, 1992) for uncertainty analysis (see Sec. 3.4 for details of
ellipsoidal model). Interestingly, these two lines of thoughts (on intervals and ellip-
soids) intersected in extremely few works, and essentially have been developed in
parallel, mostly without knowledge about the developments in the other field.
It appears that the ellipsoidal framework has some advantages over interval anal-
ysis in the sense that it deals with a smooth, convex boundary of the enclosed data
with associated straightforward analytical or numerical treatment. Yet the ellip-
soidal model may suggest that the components are functionally dependent. This
tacit assumption may be unjustified in some circumstances. Hence the independent
data may be better justified as enclosed by a rectangular region. Interval and ellip-
soidal modeling are particular cases of convex modeling (Ben-Haim and Elishakoff,
1990; Ben-Haim, 1996). In fact, convex description of uncertainty is richer than the
ellipsoidal one: In addition to the ellipsoids per se, it includes sets with functions
with envelope bounds, or those with bounded integral rectangles, or those with
bounded integral rectangles of its derivatives and so on.
The case where the uncertain variables do not belong to a convex set may be
a=5
P(t) P(t)
h2 h
1
Initial imperfections
L/2
L
Fig. 8.14 Description of the structural model.
dealt with using the methods of nonlinear programming (Li, Elishakoff, Starnes and
Shinozuka, 1996; Luenberger, 2003) (see Sec. 2.4.2 for details of nonlinear program-
ming). All these analyses share the main ideas of anti-optimization, namely, of the
desire to determine least favorable responses, in order to guarantee the successful
performance despite the presence of uncertainty.
In this section, the two-dimensional uncertain imperfections are considered in
the context of the column impact problem (Zingales and Elishakoff, 2000), whose
one-dimensional counterpart was discussed in Sec. 8.2.
8.3.2 Deterministic analysis

The differential equation describing the problem of the simply supported column,
as shown in Fig. 8.14, under axial impact load with initial imperfections is given
in Eq. (8.3) in Sec. 8.2. The same boundary and initial conditions are set in this
section, so that Eqs. (8.5) and (8.6) in Sec. 8.2 remain unchanged. Let the initial
imperfections be given as a sum of two sinusoidal terms
x ax
w0 (x) = f1 (x) + f2 (x) = h1 sin + h2 sin (8.66)
L L
where h1 and h2 are the amplitudes of f1 (x) and f2 (x), respectively. In Eq. (8.66),
the integer a > 1 represents the wave number in the variation of f2 (x). The same
non-dimensional parameters as adopted in Eqs. (8.8) and (8.9) in Sec. 8.2 are used
hereafter. The non-dimensional amplitudes of initial imperfections are given as
h1 h2
g1 =
, g2 = (8.67)

The non-dimensional differential equation governing the problem is formulated in a
similar manner as Eq. (8.10) in Sec. 8.2. The additional displacement is sought as
a superposition
u(, ) = u1 (, ) + u2 (, ) (8.68)
where the functions uj (, ) (j = 1, 2) are the solutions of the following equations:
(j)
4 uj 2 uj 2 uj d2 u 0
4
+ 2 2
+ 4 2
= 2 , (j = 1, 2) (8.69)
d 2
with the initial imperfections expressed as

(1) (2)
u0 () = g1 sin(), u0 () = g2 sin(a) (8.70)
The functions uj (, ) are represented in the separable forms as the initial im-
perfections in Eq. (8.70), namely,
u1 (, ) = e1 () sin(), u2 (, ) = e2 () sin(a) (8.71)
where e1 () and e2 () are time-dependent functions. Substitution of Eqs. (8.70)
and (8.71) into Eq. (8.69) yields ordinary differential equations for e1 () and e2 ()
d2 e1 ()
+ (1 )e1 () = g1 (8.72a)
d2
d2 e2 ()
+ a2 (a2 )e2 () = a2 g2 (8.72b)
d2
Satisfying the initial conditions ej = 0 and dej /dt = 0 at = 0 (j = 1, 2) yields
g1

(cosh(r) 1), > 1

1

g 1 2
e1 () = , =1 (8.73)

2

g1 (1 cos(r)),

<1
1
and
g2

(cosh(aq) 1), > a2

a2

1
e2 () = a2 g2 2 , = a2 (8.74)
2

g2

(1 cos(aq)), < a2
a2
where
p p
r = | 1|, q = | a2 | (8.75)
The non-dimensional modal displacements zi (, ) (i = 1, 2) are introduced as
(1)
z1 (, ) = u0 () + u1 (, ) = [g1 + e1 ()] sin(),
(2)
(8.76)
z2 (, ) = u0 () + u2 (, ) = [g2 + e2 ()] sin(a)
With the aid of Eqs. (8.71) and (8.72), we form the total displacement wT (x, t) or
its non-dimensional counterpart z(, ) as
1
z(, ) = wT (x, t) = z1 (, ) + z2 (, ) (8.77)

(j) (j)
By defining the non-dimensional displacement wT (x, t) = w0 (x) + wj (x, t) (j =
1, 2), we have
1 (1)
z1 (, ) = wT (x, t) = 1 sin(),
(8.78)
1 (2)
z2 (, ) = wT (x, t) = 2 sin(a)

where the functions j () (j = 1, 2) of the non-dimensional time alone read

g1 1 (1)
1 () = cosh(r) = g1 a1 (), >1 (8.79a)
1
g1 (2 + 2) (2)
1 () = = g1 a1 (), =1 (8.79b)
2
g1 1 (3)
1 () = cos(r) = g1 a1 (), <1 (8.79c)
1

g2 a2 (1)
2 () = cosh(aq) = g2 a2 (), > a2 (8.79d)
a2
g2 (a2 2 + 2) (2)
2 () = = g2 a2 (), = a2 (8.79e)
2
g2 a2 (3)
2 () = 2 cos(aq) = g2 a2 (), < a2 (8.79f)
a
Since there exist different analytical expressions for j () depending on the value of
, the total displacement possesses various analytical representations. In particular,
five different cases occur as follows:
(3) (3)
z(, ) = a1 ()g1 sin() + a2 ()g2 sin(a), < 1 (8.80a)
(2) (3)
z(, ) = a1 ()g1 sin() + a2 ()g2 sin(a), = 1 (8.80b)
(1) (3)
z(, ) = a1 ()g1 sin() + a2 ()g2 sin(a), 1 < < a2 (8.80c)
(1) (2) 2
z(, ) = a1 ()g1 sin() + a2 ()g2 sin(a), = a (8.80d)
(1) (1)
z(, ) = a1 ()g1 sin() + a2 ()g2 sin(a), > a2 (8.80e)
(i)
Note that, although there are five different cases, each of the parameters a 1 and
(j)
a2 has three separate expressions. If the initial imperfection amplitudes g1 and g2
are given deterministically, the design is performed so that the total displacement
does not exceed the threshold value c. Hereafter the cross-section of the column is
assumed to have a circular solid shape.
The main thrust of this section is an investigation of the effect of uncertainty in
initial imperfections on the resulting design values of the cross-sectional radius. If
g1 and g2 constitute uncertain variables, the output z(, ) will likewise represent
the uncertain function. Properties of the function z(, ) depend on the informa-
tion provided about the uncertain variables g1 and g2 . We will investigate several
alternative avenues of describing this vector uncertainty, either probabilistically or
without recourse to the stochasticity concept.
8.3.3 Probabilistic analysis

Let the amplitudes of the initial imperfections g1 and g2 constitute a random vector
with specified joint probability density function fG1 G2 (g1 , g2 ). Capital letters denote
the random variables, whereas lower-case notation is reserved for the possible values.
We are interested in finding the reliability of the system; namely, the probability
that the total displacement will remain in a safe region, in accordance with Hoffs
criterion given by Eq. (8.19) in Sec. 8.2.
The reliability is defined as the probability that the maximum displacement
Y = max Z(, ) remains in a safe region. For simplicity we assume that a is an
odd number. Hence
Y max Z(, ) = Z(1/2, ) = V1 () + V2 () (8.81)

Functions V1 () and V2 () are multiplicative random processes depending upon the

non-dimensional parameter
(i) (j)
V1 () = G1 a1 (), V2 () = G2 a2 (), (i, j = 1, 2, 3) (8.82)
and the reliability is given as
R() = Prob(d Y = V1 () + V2 () c) (8.83)
In view of Eq. (8.81), the reliability in Eq. (8.83) becomes
R() = Prob(Y c) Prob(Y d) (8.84)
Equation (8.84) can be written as
R() = FY, (c, ) FY, (d, ) (8.85)
where FY, is the probability distribution functions of random process Y :
(i) (j)
Y () = G1 a1 () + G2 a2 (),
(8.86)
(i, j) {(1, 1), (1, 2), (1, 3), (2, 3), (3, 3)}
Note that Y () represents a linear combination of two random variables with de-
terministic real-valued positive functions of non-dimensional time as coefficients.
The evaluation of the reliability, as stated in Eq. (8.84), needs the particularization
of the joint probability density function of the random variables G1 and G2 . We
will consider the cases in which G1 and G2 are either statistically independent or
dependent.
8.3.4 Initial imperfections with uniform probability density: Rect-

angular domain
Let the initial imperfections be independent random variables with a uniform prob-
ability density function in a rectangular domain as shown in Fig. 8.15, which is
defined as
1 1
fG1 G2 (g1 , g2 ) = [hg1 1 i hg1 2 i ]
2 1 2 1 (8.87)
[hg2 1 i hg2 2 i ]
where h i is the singularity function defined in Eq. (8.1) in Sec. 8.2. The varia-
tions of G1 and G2 are confined to the intervals G1 [1 , 2 ] and G2 [1 , 2 ],
2 2
f G 1 G2
1
0 1.5
g2
1
1
1.5
g1
2
Fig. 8.15 Uniform probability density function over a rectangular domain.
respectively, as shown in Fig. 8.15. Moreover, we assume, for the sake of simplicity,
that j > 0, j > 0 (j = 1, 2). Marginal probability densities read
1
fG1 (g1 ) = [hg1 1 i hg1 2 i ],
2 1
(8.88)
1
fG2 (g2 ) = [hg2 2 i hg2 2 i ]
2 1
with
fG1 G2 (g1 , g2 ) = fG1 (g1 )fG2 (g2 ) (8.89)
In order to perform the reliability analysis of the column, it is necessary to

find the probability distribution function FY, (y, ) of the random variable Y () in
Eq. (8.85). With the second term in Eq. (8.85) vanishing identically, we obtain
(j)
Z Z ca
2
(i)
()g2
a ()
R() = FY, (c, ) = 1
fG1 G2 (g1 , g2 )dg1 dg2 (8.90)

By substituting Eqs. (8.88) and (8.89) into Eq. (8.90) and performing integration,
the reliability function is obtained as

" #
1 1 1 (i) (j)
R() = (j) (i)
hc a 1 1 a 2 1 i 2
2 1 2 1 a2 a1
" #
1 1 1 (i) (j)
(j) (i)
hc a 1 1 a 2 2 i 2
2 1 2 1 a2 a1
" # (8.91)
1 1 1 (i) (j)
(j) (i)
hc a 1 2 a 2 1 i 2
2 1 2 1 a2 a1
" #
1 1 1 (i) (j)
(j) (i)
hc a 1 2 a 2 2 i 2
2 1 2 1 a2 a1
where

x2
2
hxi = , x0 (8.92)
0,2 x<0
Equation (8.91) suggests several useful conclusions regarding the characterization
(i)
of the reliability. At the initial time instance t = 0 (or = 0), we have a1 (0) =
(j)
a2 (0) = 1. We conclude that if the failure boundary c is set as
c 1 + 1 (8.93)
then the reliability vanishes, and no possibility of design associated with non-
vanishing reliability exists for the column subjected to non-zero values of the applied
load P . Figures 8.168.19 portray some interesting aspects that may be deduced
by investigation of the various terms in Eq. (8.91). We first observe that each one
of the four terms represents the area of the region created by the straight line
(i) (j)
c = a 1 g1 + a 2 g2 (8.94)
and an appropriate boundary of the rectangular domain of the initial imperfection
amplitudes g1 and g2 . Let us consider various straight lines passing through either
of the four corners of the boundary, and parallel to the line given in Eq. (8.94). The
plane (g1 , g2 ) is subdivided into five regions denoted by j (j = 1, . . . , 5) as shown
in Fig. 8.16. The broken line in Fig. 8.16 represents the line in Eq. (8.94). Four
dotted lines are parallel to it, and pass through either corner A, B, C, or D. In
the case where the line in Eq. (8.94) belongs to region 2 , the coordinates of the
intersection points A and B with the edges AD and AB, respectively, are given
by
! !
(j) (i)
c a 2 1 c a 1 1
A = (i)
, 1 , B = 1 , (j)
(8.95)
a1 a2
The distances AA and AB are
(j) (i)
c a 2 1 c a 1 1
AA = (i)
1 , AB = (j)
1 (8.96)
a1 a2
2.75
3 4 5
2.25
B C
1.75
g2 B*
A*
A D
1.25
1 2
.75
0.75 1 1.25 1.5 1.75 2 2.25 2.5
g1
Fig. 8.16 Geometrical representation of the first term in reliability expression, Eq. (8.97).
2.75
3 4 5
2.25
B*
l C* C
B
1.75
g2
A*
A D
1.25
1 2
.75
0.75 1 1.25 1.5 1.75 2 2.25 2.5
g1
Fig. 8.17 Geometrical representation of the second term in reliability expression, Eq. (8.100).
The area SAB A under the triangle AB A , denoted as a hatched region in

Fig. 8.16, equals
(i) (j)
1 (c a1 1 a2 1 )2
SAB A = (AA )(AB ) = (i) (j)
(8.97)
2 2a a 1 2
The reliability of the system is determined by multiplying the area in Eq. (8.97)
by the constant value of the probability density function (2 1 )1 (2 1 )1 .
In this case, only the first term in the bracketed expression in Eq. (8.91) is non-
vanishing. By inspection of Eq. (8.91), the coincidence between this first term and
the area SAB A is recognized.
Figure 8.17 deals with the case in which the straight line in Eq. (8.94) belongs to
3.25
B*
2.25
3 4 5
2.75
B C* C
g2 D*
1.75
A D A*
1.25 1 2
0.75
0.75 1 1.25 1.5 1.75 2 2.25 2.5
g1
Fig. 8.18 Geometrical representation of the third term in reliability expression, Eq. (8.103).
3.5
B*
2.5
3 4 5
g2
D*
B C C*
2
1.5 A*
A D
1 2
1 1.25 1.5 1.75 2 2.25 2.5
g1
Fig. 8.19 Geometrical representation of the fourth term in reliability expression, Eq. (8.104).
the region 3 . The shaded area ABC A , multiplied by the above constant value of
the probability density function, represents the reliability of the system. This area
is obtained as a difference between the area of the triangle AB A and the area of
the triangle BB C . The coordinates of the intersection point C are
!
(j)
c a 2 2
C = (i)
, 2 (8.98)
a1
whereas the expression of the coordinates of B coincides with Eq. (8.95). The side
lengths for the triangle AB A are then given by Eq. (8.96) and the ones of the
triangle BB C read
(i) (j)
c a 1 1 c a 2 2
BB = (j)
2 , BC = (i)
2 (8.99)
a2 a1
The area SBB C equals
(i) (j)
1 (c a1 1 a2 2 )2
SBEC = (BB )(BC ) = (i) (j)
(8.100)
2 2a a 1 2
and is recognized as the second term in Eq. (8.91). Note that the area of the triangle
AB A coincides with the first term of Eq. (8.91).
Analogously, the geometrical meanings of the third and fourth terms involved in
Eq. (8.91) may be deduced by examining Figs. 8.18 and 8.19, respectively. In fact,
Fig. 8.18 deals with the case that the broken line belongs to region 4 . In this case,
the reliability is given by the area ABC D D, and is obtained as the algebraic sum
SBC D DA = SAB A SBEC SDD A (8.101)
The first two terms of this expression have been already identified. In order to
obtain the last term in Eq. (8.101), we determine the coordinates of point D as
!
(i)
c a 1 2
D = 2 , (j)
(8.102)
a2
Note that the coordinates of the point A are given in Eq. (8.95). The area of the
triangle DD A is so obtained as
(i) (j)
1 (c a1 2 a2 1 )2
SDD A = (DD )(DA ) = (i) (j)
(8.103)
2 2a a 1 2
which coincides with the third term in Eq. (8.91). To determine the geometric
interpretation of the fourth term in Eq. (8.91), let us consider the case when the
broken line belongs to region 5 as shown in Fig. 8.19. The reliability is given as the
product of the area of the rectangle ABCD and the density (2 1 )1 (2 1 )1 .
The area itself equals (2 1 )(2 1 ). Hence, the reliability is unity as expected.
On the other hand, in order to identify the fourth term in Eq. (8.91), we represent
the area ABCD as the following algebraic sum:
SABCD = SAB A SBB C SDD A + SCC D (8.104)
The first three terms in Eq. (8.104) are given by their respective counterparts in
Eq. (8.91). The fourth term in Eq. (8.104) may be easily obtained by inspection of
Eqs. (8.98) and (8.102) as
(i) (j)
1 (c a1 2 a2 2 )2
SCC D = CC )(CD ) = (i) (j)
(8.105)
2 2a a 1 2
As is seen, each term in the reliability expression has an appropriate geometrical
meaning. Naturally, if the broken line lies in region 1 , then one immediately
c = 1.2(2+2) 1=2.39
1.2 = 0.5 2=2.0
1
0.8 5 4 3
R
0.6
0.4
0.2
4 3
0 1 2 3 4 5

Fig. 8.20 Reliability versus non-dimensional time, initial imperfections with uniform probability
density (D = [1.2, 2] [1.4, 2]), Eqs. (8.80a) and (8.91).
c = 1.2(2+2) 1=1.640
1.2 = 1 2=1.327
3=1.133
1
4=0.796
0.8 5 4 3 2 1
R
0.6
0.4
0.2
4 3 2 1
0 0.5 1 1.5 2 2.5

density (D = [1.2, 2] [1.4, 2]), Eqs. (8.80b) and (8.91).
deduces that the reliability vanishes. Likewise, if the broken line passes through
region 5 , the corresponding reliability is unity.
The results of sample calculations are portrayed in Figs. 8.208.24, with the
same mechanical and geometrical parameters used in Sec. 8.2, and the second term
of imperfection in Eq. (8.66) is given by a = 5, which is an odd number. The cases
(i) (j)
correspond to the five independent expressions for the coefficients a1 and a2 in
Eq. (8.80), which correspond, respectively, to < 1, = 1, 1 < < 25, = 25,
> 25. For a specified value of the failure boundary c and the different values of the
ratio P/Pcl , the reliability functions are depicted as functions of a non-dimensional
time . For the case < 1 in Fig. 8.20, buckling occurrence is not a certain event:
The structure may or may not buckle depending upon the parameters of the system.
Hence, the reliability does not necessarily tend to zero with the increase of time .
c = 1.2(2+2)
1.2 = 2
1
0.8 5 4 3 2 1
R
0.6
1=0.471
0.4 2=0.743
3=0.982
0.2 4=1.203
4 3 2 1
0 0.2 0.4 0.6 0.8 1 1.2 1.4

density (D = [1.2, 2] [1.4, 2]), Eqs. (8.80c) and (8.91).
c = 1.2(2+2) 1=0.0697
1.2 = 25 2=0.0550
3=0.0500
1 4=1.0351
0.8 5 4 3 2 1
R
0.6
0.4
0.2
4 3 2 1
0 0.02 0.04 0.06 0.08 0.1

density (D = [1.2, 2] [1.4, 2]), Eqs. (8.80d) and (8.91).
Instead, it gains a minimum value given by

R() = R(/r) for /r (8.106)
depending on the value c that delimits stable and unstable states. When the non-
dimensional time reaches the value /r, the displacement at the center of the column
will achieve its maximum value, corresponding to the least possible reliability. In
these circumstances, the value R(/r) represents the guaranteed minimum reliabil-
ity that the column may possess.
Our main objective is to design the system, i.e., to obtain the radius d of the
circular solid cross-section of the column so as to maintain a prescribed reliability
rc , up to a pre-selected time instant t, for a specified value of the failure boundary
c.
c = 1.2(2+2) 1=0.0680
1.2 = 26 2=0.0538
3=0.0489
1 4=1.0343
0.8 5 4 3 2 1
R
0.6
0.4
0.2
4 3 2 1
0 0.02 0.04 0.06 0.08 0.1

density (D = [1.2, 2] [1.4, 2]), Eqs. (8.80e) and (8.91).
10
8 Ma = 1.65
2
4
M = 3.68
6
Boundary c
4
I II III
2
Ma 2
4
M
0
0 2 4 6 8 10 12
Radius d (cm)
Fig. 8.25 Design curve c = c(d ) at t = 0.2 for uniform probability density function and unity
reliability requirement (P = 3000 kg, D = [1.2, 2] [1.4, 2]), Eq. (8.91).
It is advantageous, however, to view c as a function of the remaining parameters,

as c(d , t, E, L, m), which yields the failure boundary, for various combinations of
its arguments, corresponding to satisfactory performance with the specified relia-
bility rc . Thus, specifying the mechanical and the geometrical characteristics of the
column and the external load allows one to obtain design curve c(d ) and surface
c(d , t) shown in Figs. 8.258.27. The value of the cross-sectional radius d neces-
sary for a successful performance is obtained by means of these curves, specifying
the value of the failure boundary c. Note that the ratio = P/Pcl depends upon
6
3
Boundary c 4
4 2
0
Time t (sec)
2 1
4
Radius d (cm)
6
0
Fig. 8.26 Design surface c = c(d , t) for uniform probability density function and unity reliability
requirement (P = 3000 kg, D = [1.2, 2] [1.4, 2]), Eq. (8.91).
10
Solid Line rc = 1
Dashed Line rc = 0.9
8
Dotted Line rc = 0.8
Boundary c
4
I II III
2
0
0 2 4 6 8 10 12
Radius d (cm)
Fig. 8.27 Comparison of design curves; uniform probability density function and different codified
reliabilities at t = 0.5 (P = 3000 kg, D = [1.2, 2] [1.4, 2]), Eq. (8.91).
the cross-sectional radius as

P P L2 M 4P L2
= = 2 = 4, M = 3 (8.107)
Pcl EI E
It is immediately observed, in view of Eq. (8.91), that when
(i) (j)
lim a1 (, t)2 + a2 (, t)2 = c (8.108)
d
the reliability tends to unity from below.

Figure 8.25 depicts the failure boundary c versus the radius d for the prescribed
time t = 0.5 and unity reliability. As the dependence c = c(d ) has an asymptote at
c = 4, when 2 = 2 and 2 = 2, this numerical result can be derived from analytical
considerations. In fact, when d tends to infinity,
(3)
lim aj (d , t) = 1, (j = 1, 2) (8.109)
d
and Eq. (8.108) takes the following form:

c = 2 + 2 (8.110)
Since 2 + 2 = 4 in Fig. 8.25, the asymptote d is represented by c = 4. This
result is straightforward: an infinite value of the cross-sectional radius d yields an
infinite frequency of vibration of the column with J as observed in Eq. (8.9)
in Sec. 8.2. In these circumstances, the system will remain in the same position
represented by its initial imperfection alone.
Figure 8.26 portrays a design surface c = (d , t) allowing one to obtain, for a
specified time t and failure boundary c, the required radius d of the cross-sectional
area so that the reliability is rc . In Fig. 8.26, rc is taken to be unity. Figure 8.27
depicts design curves for various codified reliabilities rc = 0.8, 0.9, and 1.0. These
figures show that smaller values of design cross-sectional radius are needed when
less stringent reliabilities are imposed for a specified failure boundary, as it should
be.
Let us examine the dependence c = c(d ) in Figs. 8.25 and 8.27. Each figure
is composed of three different sub-regions. Values of the cross-sectional radius
(2)
d dc = (M/a2 )1/4 (Region I) lead to a load ratio a2 . Therefore, the
(i) (j)
functions a1 (d , t) and a2 (d , t) in the expression of the failure boundary c are
represented by Eqs. (8.79a) and (8.79d), respectively. In this case, the displacement
function max Z(d , t) is a monotonically increasing function with time t. A large
failure boundary c, combined with small values of the time interval t, where the
successful probabilistic performance is required, is necessary for high prescribed
reliabilities rc .
(2) (1)
In the case where dc < d < dc = M , the load ratio belongs to the open
interval (1, a2 ). The expression of the failure boundary c contains two functions: A
(3) (1)
trigonometric function for a2 (d , t) and an expression a1 (d , t) that is monoton-
ically increasing. Also for this case, large values of c along with small performance
times would guarantee the high reliability requirement. To obtain more insights on
the dependence upon time, we express the argument 1 of the cosine function in
Eq. (8.79f), in view of Eq. (8.107), as
p r
a2 4d M E
1 = t (8.111)
d L m
allowing one to have a better look at the influence of the pre-selected time t. In
(3)
fact, for the function a2 (d , t) in Eq. (8.79f) that attains its maximum for values
of d and t making 1 = , reliability reaches its minimum level. For values of
parameters leading to 1 > , the previously achieved minimum reliability is valid.

(1)
Values of d dc = (M )1/4 lead to a load ratio < 1; hence, the expression of
(j)
the failure boundary c(d , t) involves Eq. (8.79a) for a1 (d , t). Interestingly, if the
(3)
argument of the trigonometric function in the expression of a1 (d , t), bearing in
mind Eq. (8.107),
p r
4d M E
2 = t (8.112)
d L m
is larger than for a specified time t and d , the reliability achieved for 2 = is
(3)
retained. Hence, the cosine function in the expression of a1 (d , t) must be replaced
(3)
by 1, and the function a1 (d , t) becomes
(3) 1 + 4d /M
a1 (d , t) = (8.113)
1 4d /M
As may be observed from Fig. 8.27, different values of the required reliability rc
lead to appropriate asymptotic values. In fact, let us design the system for high
reliability requirement as shown in region 4 in Fig. 8.18. The expression of the
failure boundary as a function of reliability, with the aid of Eq. (8.82), is given by
q
(i) (j) (i) (j)
c = 2 a1 + 2 a2 2a1 a2 (1 rc )(2 1 )(2 1 ) (8.114)
Evaluation of the limit value of Eq. (8.115) for an infinite value of d reads
p
lim = 2 + 2 (1 rc )(2 1 )(2 1 ) (8.115)
d
Equation (8.115) allows us to obtain the asymptotic value of the failure boundary
function c(d , t) by specifying the reliability value.
The results of this section clearly demonstrate that for high values of required
reliability, the design radii d of the cross-section are extremely close to the values
obtained by non-probabilistic analysis. The following question can be asked: Do the
design values of the cross-sectional radius d depend upon the particular form of the
probability density function fG1 G2 (g1 , g2 )? One can anticipate, generally speaking,
that the designs depend on the probabilistic inputs. Yet, as it will be elucidated,
the design in the extremely high reliability region is practically independent of the
input probabilistic information.
8.3.5 Initial imperfections with general probability density: Rect-

angular domain
Let the initial imperfections possess a general probability density function in a
rectangular domain

fG1 G2 (g1 , g2 ) = AfG 1 G2
(g1 , g2 )[hg1 1 i hg1 2 i ]
(8.116)
[hg2 1 i hg2 2 i ]
located in the first quadrant of the plane (g1 , g2 ), where A represents a normalization
coefficient depending upon the specific expression of the probability density function

fG 1 G2
(g1 , g2 ) that extends over the entire plane.
The reliability expression with z = c substituted into Eq. (8.90), and taking into
account Eq. (8.116), read,
(j)
Z Z ca
2
(i)
g2
a1
R() = dg2 AfG 1 G2
(g1 , g2 )[hg1 1 i hg1 2 i ]

[hg2 1 i
hg2 2 i ]dg1
Z
(8.117)
= [hg2 1 i hg2 2 i ]dg2

(j)
Z ca2 g2
(i)
a1
AfG 1 G2
(g1 , g2 )[hg1 1 i hg1 2 i ]dg1

Denoting the inner integral in Eq. (8.117) by I(g2 ), Eq. (8.117) can be rewritten as
Z
R() = I(g2 )[hg2 1 i hg2 2 i ]dg2

Z 2
= I(g2 )dg2 (8.118)
1
(j)
Z 2 Z ca2 g2
(i)
a1
= dg2 AfG 1 G2
(g1 , g2 )[hg1 1 i hg1 2 i ]dg1
1
In a high reliability range as shown in region 4 in Fig. 8.18, c lies in the vicinity
of a critical value c given as
(i) (j)
c = a 1 2 + a 2 2 (8.119)
The highest reliability requirement will be obtained when the line in Eq. (8.94)
tends to pass through the point B. This corresponds to value c = c substituted in
Eq. (8.94). Thus we calculate the limit
(j)
Z 2 Z ca2 g2
(i)
a1
lim R() = lim dg2 AfG 1 G2
(g1 , g2 )[hg1 1 i hg1 2 i ]dg1
cc cc 1
(j)
Z 2 Z c a2 g2
(i)
a
= dg2 1
AfG 1 G2
(g1 , g2 )[hg1 1 i hg1 2 i ]dg1
1
Z 2 Z 2

= dg2 AfG 1 G2
(g1 , g2 )[hg1 1 i hg1 2 i ]dg1
1
(j)
Z 2 Z 2 +
a
2
(i) (2 g2 )
a
+ dg2 1
AfG 1 G2
(g1 , g2 )[hg1 1 i hg1 2 i ]dg1
1 2
(8.120)
2.5
A2
2 A1

G0
1.5
g2 K
D
0.5
0 1 2 3 4
g1
Fig. 8.28 Domain of integration of the probability density function (g 10 = 2, g20 = 1.5, K = 1.0),
Eq. (8.122).
The inner integral vanishes, because integration is carried out in the interval
(j) (i)
[2 , 2 + (a2 /a1 )(2 g2 )]. Since 1 g2 2 , the upper limit of integration is
greater than 2 . However, in the interval beyond 2 , the integrand in the square
parenthesis vanishes. Reliability is then given only by the first term in Eq. (8.120),
yielding
Z 2 Z 2

lim R() = lim dg2 AfG 1 G2
(g1 , g2 )dg1 = 1 (8.121)
cc cc 1 1
which represents the integral of the probability density function in its entire domain.
In this case of unitary reliability, the design value of the cross-sectional radius d is
found from Eq. (8.119).
8.3.6 Initial imperfection with uniform probability density func-

tion: Circular domain
This section deals with a different shape of the variable domain. The magnitudes
of the initial imperfections G1 and G2 are modeled as random variables belonging
to a circular domain on the plane (g1 , g2 ). The joint probability density function is
assumed to be uniform as
( 1
, for (g1 g10 )2 + (g2 g20 )2 K 2
fG1 G2 (g1 , g2 ) = K 2 (8.122)
0, elsewhere
where (g10 , g20 ) is the center of the circle, and K the radius. As may be seen in
Fig. 8.28, we assume, without loss of generality, that the circle is placed in the first
quadrant of the plane (g1 , g2 ). Also we assume that the straight line
(i) (j)
c = a1 ()g1 + a2 ()g2 (8.123)
(i)
is drawn in a generic position depending on the values of the function and a1 ()
(j)
a2 ().
According to Eq. (8.85), in order to find the reliability, it is necessary to integrate
the probability density function below the straight line in Eq. (8.122), or over the
hatched area in Fig. 8.28 as
(j)
ZZ Z Z ca2 g2
(i)
a1 1
R() = fG1 G2 (g1 , g2 )dg1 dg2 = dg1 dg2 (8.124)
D K 2
The double integral in the latter equation will be calculated with the aid of the
following polar coordinate system whose origin is at the center of the circle:
g1 = g10 + cos , g2 = g20 + sin (8.125)
where is the polar distance and is the polar angle. In this coordinate system,
the reliability becomes
ZZ

R() = 1 2
dd (8.126)
Dc K
where Dc is the non-hatched area in Fig. 8.28. Eq. (8.126) yields

1
[ sin()]
R() = 1 (8.127)
2
where = 1 2 is the angular difference of the phase angles 1 and 2 between
the polar axis G0 D and the polar directions G0 A1 and G0 A2 , respectively. The polar
phase angles 1 and 2 may be obtained by solving the following set of equations:
(i) (j) 1 (i) (j)
a1 cos + a2 sin = (c a1 g10 a2 g20 ) (8.128a)

2 (sin2 + cos2 ) = K 2 (8.128b)
where the first equation represents a straight line in the polar coordinate system,
and the latter is the equation of a circle.
Equations (8.128a) and (8.128b) have the solution
q
(i) (j) (i) 2 (j) 2 2
f a 1 + a 2 (a 1 ) + (a 2 ) f
1 = cos1 (i) (j)
(8.129a)
(a1 )2 + (a2 )2
q
(j) (j) (i) 2 (j) 2 2
f a 1 + a 2 (a 1 ) + (a 2 ) f
2 = cos1 (8.129b)
(i) 2 (j) 2
(a1 ) + (a2 )
with
1 (i) (j)
f= (c a1 g10 a2 g20 ) (8.130)
K
In case of R() tending to unity from below, we see from Eq. (8.127)
0 or 1 2 (8.131)
Hence, the square-root terms in Eq. (8.129) must gain
(i) (j)
(a1 )2 + (a2 )2 f 2 0+ (8.132)
Thus
q
(i) (j)
f (a1 )2 + (a2 )2 (8.133)
The geometrical meaning of the condition stated in Eq. (8.133) can be shown,
expressing the system in Eqs. (8.127a) and (8.127b) back in Cartesian coordinates.
Thus, bearing in mind Eqs. (8.125) and (8.133), we write Eq. (8.127) in the form
q
(i) (j) (i) (i) (j)
a1 g10 + a2 g20 a1 g1 + K (a1 )2 + (a2 )2
g2 = (i)
(8.134a)
a1
(g1 g10 )2 + (g2 g20 )2 = K 2 (8.134b)
Substituting Eq. (8.134a) into Eq. (8.134b), we obtain a quadratic equation for the
unknown g1
q0 (g1 )2 2g1 q1 + q2 = 0 (8.135)
with coefficients
!
(i) 2
a
q0 = 1 + 1
(j)

a2
q
(i) 2 (i) (i) 2 (j) 2
2(a ) g 10 2a 1 K (a 1 ) + (a 2 )
q1 = g10 + 1
+
(j) 2 (j) 2
(a2 ) (a2 )
! ! q
(i) 2 (i) 2 2a
(i)
(a
(i) 2
) + (a
(j) 2
)
a a 1 1 2
q2 = (g10 )2 + 1
(j)
(g10 )2 + 1
(j)
K2 + (j)
g10 K
a2 a2 (a2 )2
(8.136)
of which the discriminant q12 4q0 q1 is identically zero. Hence, the two coinciding
solutions of Eq. (8.135) read
(i)
(1) (2) a1 K
g1 = g1 = g10 + q (8.137)
(i) (j)
(a1 )2 + (a2 )2
This implies that the straight line in Eq. (8.134a) and the circle in Eq. (8.134b)
share one common point. Thus, we conclude that the straight line is tangent to the
circular domain of the initial imperfections. Eq. (8.133) can then be interpreted as
the condition for this situation.
1 1 = 1.628
2 = 2.490
rc = 0.9
0.8
rc = 0.7
0.6
rc
0.4
0.2
1 2
0 1 2 3 4 5 6

Fig. 8.29 Reliability versus angular difference, Eq. (8.127).
10
Solid Line rc = 1
Boundary c (cm)
Dashed Line rc = 0.95

8
Dotted Line rc = 0.9
4 I II III
2
0 2 4 6 8 10 12 14
Radius d (cm)
Fig. 8.30 Comparison of design curves for uniform probability density function over a circular
domain and different required reliabilities (t = 0.5, P = 3000 kg, g10 = 2, g20 = 1.5, K = 1.0),
Eq. (8.139).
In order to design the system with a prescribed required reliability rc , we must

solve Eq. (8.127) for the angular difference , with respect to rc substituted for
R. Figure 8.29 depicts the curve R = R(), allowing one to obtain the angular
difference given the value of the design reliability rc . For example, for rc = 0.9,
= 1.628.
Having found that corresponds to rc , we proceed to design the system. In
order to obtain the design curves c = c(d ) for specified time and pre-selected
7
6 5
Boundary c
5
4
4
3 3
t (sec)
2 2
4
6 1
Radius d (cm)
8
10
Fig. 8.31 Design surface c = c(d , t), uniform probability density function over a circular domain
and required unity reliability (P = 3000 kg, g10 = 2, g20 = 1.5, K = 1.0), Eq. (8.139).
reliability rc , one subtracts Eq. (8.128) from Eq. (8.129), leading to
= 2 1
q
(i) (j) (i) (j)
f a1 + a2 (a1 )2 + (a2 )2 f 2
= cos 1
(i) (j)
(a1 )2 + (a2 )2 (8.138)
q
(j) (j) (i) 2 (j) 2 2
f a 1 + a 2 (a 1 ) + (a 2 ) f
cos1
(i) (j)
(a1 )2 + (a2 )2
Once is determined from Fig. 8.29, one can calculate the value from Eq. (8.138).
The solution of Eq. (8.138) for f , in conjunction with Eq. (8.130), allows one to
obtain the required expression for the design curve c = c(d ), which reads
(i) (j)
c(d ) = a1 (d )g10 + a2 (d )g20
q (8.139)
K (i) (j)
+ (a1 (d ))2 (1 + cos()) + (a2 (d ))2
2
and is portrayed in Fig. 8.30 for specified time t = 0.5 and different values of rc .
Figure 8.31 represents the design surface for unity reliability as a function c =
c(d , t), allowing one to find the cross-sectional radius d for specified exploitation
time t and the failure boundary c.
8.3.7 Initial imperfections as interval variables: Interval analysis

Let the initial imperfection amplitude be represented by a vector interval variable
g1 = [1 , 2 ], g2 = [1 , 2 ] (8.140)
so that the additional displacement function at the center of the column is an

interval variable as well:
(i) (j)
z = [1 , 2 ] = [1 , 2 ]a1 () + [1 , 2 ]a2 () (8.141)
where 1 and 2 are the lower and upper bounds of z, respectively. Since g1 and g2
are positive, so are the additional displacements. Hence, in the safety requirement
condition d z c, only the condition z c is operative. The column design via
the anti-optimization technique identifies the worst possible reachable condition for
the uppermost bound 2 of the interval variable z. If 2 c, the system will remain
in the safe domain, otherwise it will fail. Expressing the argument of the functions
(i) (j)
a1 () and a2 () in terms of the cross-sectional radius d and the pre-selected
time t by means of Eq. (8.108) (see Sec. 8.2), we obtain the formal design relation
(i) (j)
2 = 2 a1 (d , t) + 2 a2 (d , t) = c (8.142)
Comparing Eq. (8.142) with Eq. (8.108), we conclude that the former matches the
design condition in case of unity reliability requirement given in Eq. (8.108). If
the uncertainty region coincides with the domain represented in Fig. 8.15, then the
design surface c(d , t) represented by Eq. (8.142) coincides with Eq. (8.77). Thus,
the anti-optimization method and the probabilistic one with the highest reliability
requirement lead to the same design values for the cross-sectional radius d .
8.3.8 Initial imperfections as convex variables: Circular domain

Let us consider the case when the information about the uncertain imperfection
magnitude is given as
(g1 g10 )2 + (g2 g20 )2 K 2 (8.143)
where (g10 , g20 ) is the center of the circle, and K is the radius as shown in Fig. 8.32.
The displacement at the center of column reads
(i) (j)
z = g1 a1 () + g2 a2 () (8.144)
where g1 and g2 are uncertain variables belonging to the set defined by Eq. (8.143),
which is convex.
The Lagrangian for the problem of maximizing z under constraint (8.143) is
given as
(i) (j)
= a1 g1 + a2 g2 + [(g1 g10 )2 + (g2 g20 )2 K 2 ] (8.145)
3
S
2.5
A
2
G0
1.5
g2 K
d
1 B
0.5
0 1 2 3 4
g1
Fig. 8.32 Initial imperfection amplitudes modeled by convex variables: anti-optimization design
(g10 = 2, g20 = 1.5, K = 1.0), Eq. (8.152).
where is the Lagrange multiplier (see Sec. 4.1 for details of the Lagrange multiplier
approach). Necessary conditions for the extremal read
(i)
= a1 + 2(g1 g10 ) = 0 (8.146a)
g1
(j)
= a2 + 2(g2 g20 ) = 0 (8.146b)
g2

= (g1 g10 )2 + (g2 g20 )2 K 2 = 0 (8.146c)

The solutions of Eqs. (8.146a)(8.146c), denoted by g 1 and g2 , corresponding
to maximum z read
(i)
Ka1
g 1 = g10 + q (8.147a)
(i) (j)
(a1 )2 + (a2 )2
(j)
Ka2
g 2 = g20 + q (8.147b)
(i) (j)
(a1 )2 + (a2 )2
Note that Eq. (8.147a) coincides with Eq. (8.137) which represents the abscissa
of the common point between the circular domain of the random variables in
Eq. (8.122) and the straight line in Eq. (8.123).
Let us investigate the geometrical meaning of Eqs. (8.146a)(8.146c) and the
results in Eqs. (8.147a) and (8.147b). Eliminating parameter from Eqs. (8.146a)
and (8.146b) leads to

(j)
a2 1 (i) (j)
g2 = g
(i) 1
+ (i)
(a1 g20 a2 g10 ) (8.148)
a1 a1
(j) (i)
This expression represents a straight line with slope a2 /a1 passing through the
center of the circle, which is identified in Fig. 8.32 with label S. Eq. (8.148) must be
solved in conjunction with Eq. (8.146c). Thus the solution of the anti-optimization
problem is found as the intersection point of the straight line S and the circle
represented by Eq. (8.146c). There are two intersection points in Fig. 8.32 denoted
by A and B, respectively, corresponding to the maximum and minimum of the
displacement function. In order to deal with the anti-optimization technique, one
must look for the maximum of the displacement function, namely, the point A in
Fig. 8.32, which is obtained in Eq. (8.147).
The design problem can be stated differently. Let us consider the expression
(j) q
a2 (i) (j) (i) 2 (j) 2
g2 = g
(i) 1
+ a1 g10 a2 g20 + K (a1 ) + (a2 ) (8.149)
a1
representing a line orthogonal to S and tangent to the circle at point A. One may
obtain the solution of the system in Eq. (8.146) by solving
(i) (j)
a1 g1 + a 2 g2 T = 0 (8.150a)
2 2 2
(g1 g20 ) + (g2 g20 ) = K (8.150b)
where the line in Eq. (8.150a) is perpendicular to the line in Eq. (8.149). The
parameter T is a free parameter that must be chosen so that the straight line in
Eq. (8.150a) constitutes a tangent to the circle in Eq. (8.150b). It may be deduced by
inspection of Eq. (8.147a) that the parameter T coincides with the failure boundary
value c.
The value of T is obtainable as follows by solving Eqs. (8.150a) and (8.150b):
q
(i) (j) (i) (j)
T = c = a1 (d , t)g10 + a2 (d , t)g20 + K (a1 (d , t))2 + (a2 (d , t))2 (8.151)
Rewriting Eq. (8.139) in Cartesian coordinates for unity required reliability r c as

q
(i) (j) (i) (j)
c = a1 (d , t)g10 + a2 (d , t)g20 + K (a1 (d , t))2 + (a2 (d , t))2 (8.152)
one immediately recognizes that the different models of uncertainty treated in

Sec. 8.3.6 and in the present one lead to the same analytical expression of the
failure boundary surface. Then, with the initial imperfection amplitude given in
the same region of variation, these different approaches result in the same design
surface.
8.3.9 Concluding remarks

As shown in this section, the two-dimensional interval analysis provides a design
that is a limiting case of the probabilistic design when the required reliability tends
to unity from below. Likewise, convex modeling yields a design to which the proba-
bilistic calculations tend, when the variation of uncertain variables is limited to the
circular domain. The natural question arises: Can the non-probabilistic method be
preferred to the probabilistic approach? In general, should the simpler method be
preferred to the more complex one?
It appears instructive to reproduce here three relevant quotations:
(i) The aim of science is to seek the simplest explanations of complex functions.
We are apt to fall into the error of thinking that the facts are simple because
simplicity is the goal of our quest. The guiding motto in the life of every
natural philosopher should be, Seek simplicity and distrust it. (Alfred North
Whitehead).
(ii) Everything should be made as simple as possible, but not one bit simpler.
(Albert Einstein)
(iii) Never use a long word where a short one will do. If it is possible to cut a
word, always cut it out. (George Orwell)
If one follows these recommendations, one concludes that interval analysis is the
simplest yet tenable method of describing uncertainty. It leads to the same answer
as the theory of probability, but without using long words, as Orwell put it; it
also appears that uncertainty modeling cannot be made one bit simpler, to borrow
Einsteins words, than the interval analysis. Yet, in accordance with Whitehead, we
do not discourage the distrust. Indeed, the following challenging question appears
to be in order: Can interval or ellipsoidal analyses describe first-excursion failure,
fatigue problems, wind loads, earthquake engineering and a long list of other issues
dealt with to some degree of success, recognized at least by some researchers, by
the theory of probability and random processes? This question remains yet to be
answered. The establishment of identity of designs furnished by the probabilistic
and non-probabilistic analyses shows that all the analytical roads lead to the same
results for the problem in question.
Note that hybrid probabilistic and non-probabilistic analysis of structures ap-
parently was initiated by Elishakoff and Colombi (1993) and Elishakoff, Lin and
Zhu (1994e) in the context of space shuttle application. The later works include
those by Elishakoff and Li (1999), Kreinovich, Xiang, Starks, Longpre, Ceberio,
Araiza, Beck, Kandathi, Nayak, Torres and Hajagos (2006), Adduri and Penmetsa
(2007), Qiu, Yang and Elishakoff (2008b, 2008c), and possibly others.
Chapter 9
Hybrid Optimization with

Anti-Optimization under Uncertainty or
Making the Best out of the Worst
Make the best of a bad bargain. (Japanese proverb)

Things turn out the best for people who make the best of the way
things turn out. (John Wooden)
A given form will be optimum if all failure modes which can possibly
intersect occur simultaneously under the action of the load environ-
ment. (Spunt, 1971)
In this chapter, as the main purpose of this book, hybrid approaches of optimization
with anti-optimization are presented for static and buckling problems. We first
present the framework of the problem and the methodologies for solving the two-
level optimization problem with anti-optimization. Several applications are shown
for buckling optimization, static problems considering stress, displacements and
compliance. The design methods for flexible structures are also shown, and finally,
an example of force identification of a tensegrity structure is presented.
9.1 Introduction
Structural design problems are usually formulated, or implicitly formulated, to min-

imize the structural cost under constraints on structural performance, defined by
the responses to design loads. In the practical design process, the design loads and
the bounds on the responses are considered to be deterministic, where the bounds
are decreased, or the design loads are increased by incorporating the safety factor.
However, it is desirable to explicitly incorporate uncertainties in the formulation of
the design problem. Hence, the responses or the design loads are obtained by solv-
ing anti-optimization problems, and the design problem turns out to be a hybrid
and a two-stage optimizationanti-optimization problem.
Elishakoff, Haftka and Fang (1994c) presented a basic concept of optimization
with anti-optimization. Kwak and Haug (1976) formulated a worst-case design
problem and developed a parametric programming approach. They applied it to
273
P2 , U2
P1 , U1
H
x
W1 W2
y
Fig. 9.1 An asymmetric 2-bar truss.
problems with loading uncertainty and a vibration isolator design with a specified
band of excitation frequency. Lanzi and Giavotto (2006) used a multiobjective
genetic algorithm for buckling optimization with uncertainty. El Damatty and
Nassef (2001) optimized a ribbed shell by using a genetic algorithm.
Structural optimization considering the worst-case scenario is closely related to
the so-called robust design. The purpose of robust design is to achieve the design
that is least sensitive to the uncertain parameters (Ben-Haim, Fr
yba and Yoshikawa,
1999; Au, Cheng, Tham and Zheng, 2003). Various methodologies have been de-
veloped based on the intuitively defined robustness function.
Structural optimization under uncertainty has been extensively studied in the
framework of reliability-based design (Palassopoulos, 1991; Kogiso, Shao, Miki and
Murotsu, 1997; Hajela and Vittal, 2006). There are many methods for reliability-
based structural optimization; see, for instance, Youn, Choi and Park (2003), Khar-
manda, Olhoff and El-Hami (2004), Cheng, Xu and Jiang (2006), and Kim and Choi
(2008). Optimization considering uncertainty can also be formulated by the fuzzy
set theory. Pantelides and Ganzerli (2001) showed that similar designs can be found
by fuzzy set and convex model (see Sec. 2.9 for fuzzy-set based optimization).
Consider again the asymmetric 2-bar truss as shown in Fig. 9.1, which has been

used in Sec. 4.1, subjected to static loads P1 and P2 , where H = 1, W1 = 3,
and W2 = 1. Let Ai and Li denote the cross-sectional area and the length of the
ith member, respectively. Youngs modulus is denoted by E. The horizontal and
vertical displacements are denoted by U1 and U2 , respectively.
We consider a simple optimization problem to find the design variable vector
A = (A1 , A2 )> for minimizing the total structural volume
V = A 1 L1 + A 2 L2 (9.1)
under displacement constraint
U2 U2U (9.2)
Hybrid Optimization with Anti-Optimization 275
where U2U is the specified upper bound for U2 , which is assumed to be positive.
The relation between the load vector P = (P1 , P2 )> and the displacement vector
U = (U1 , U2 )> is written as

2
U1 = [(A1 + 2 2A2 )P1 ( 3A1 2 2A2 )P2 ]
A1 A2 E(2 + 3)
(9.3)
2
U2 = [( 3A1 2 2A2 )P1 + (3A1 + 2 2A2 )P2 ]
A1 A2 E(2 + 3)
Uncertainty is introduced for P as
P = P0 + a (9.4)
0
where P = (P10 , P20 )> is the vector of nominal values, and a = (a1 , a2 )> is the
uncertain parameter vector. Since U2 depends on A and a, it is written as U2 (a, A).
We are interested in designing the truss, i.e., selecting the values of A1 and A2 ,
so that the displacement U2 under the worst possible load scenario satisfies the
constraint (9.2).
Let ab (A) denote the worst load scenario for maximizing U2 . Since the worst load
scenario depends on the design variable vector A, the value of U2 corresponding
b is denoted as U
to a b2 (A) = U2 (b a(A), A). The lower and upper bounds for Ai are
denoted by ALi and AU i , respectively. Then the optimization problem is formulated
as
minimize V = A 1 L1 + A 2 L2 (9.5a)
b2 (A)
subject to U U2U (9.5b)
ALi Ai AU
i , (i = 1, 2) (9.5c)
Consider the case where the bound of a is given by the quadratic constraint
(ellipsoidal bound) as
a> a D (9.6)
where D is the specified upper bound. Then the lower-level anti-optimization prob-
b2 (A) is formulated as
lem for finding U
find b2 (A) = max U2 (a, A)
U (9.7a)
a
subject to a> a D (9.7b)

Here, A is regarded as a parameter vector. Thus, optimization with the worst load
scenario is formulated as a two-level optimizationanti-optimization problem. In
this example, we can obtain the anti-optimal solution explicitly, as follows, in the
same manner as Sec. 4.1 by using the Lagrange multiplier approach:
1
b (A) =
a c(A) (9.8)
2
where 0 is the Lagrange multiplier, and c = (c1 , c2 )> is given as

6A1 4A2 3 2A1 + 4A2
c1 (A) = , c2 (A) = (9.9)
A1 A2 E(2 + 3) A1 A2 E(2 + 3)
Since the constraint (9.7b) is satisfied with equality at the anti-optimal solution, we
obtain r
1 c> c
= (9.10)
2 D
Hence, by incorporating Eqs. (9.8)(9.10) into Eqs. (9.3) and (9.4), the load
vector P = P 0 + ab(A) and the displacement U b2 (A) corresponding to the worst-case
scenario are written as explicit functions of A, and the two-level optimizationanti-
optimization problem can be reformulated as a single-level optimization problem.
However, in general, such an explicit reformulation is not possible. Therefore,
many algorithms for solving the two-level problems have been developed.
9.3 Formulation of the Two-Level

OptimizationAnti-Optimization Problem
The general formulations of the two-level optimizationanti-optimization problem

are presented below (Lombardi and Haftka, 1998; Lombardi, 1998).
Let x = (x1 , . . . , xm )> denote the vector of m design variables. The objective
function, such as the total structural volume, to be minimized is denoted by F (x).
Constraints are given as
Hj (x) 0, (j = 1, . . . , N I ) (9.11)
I
where N is the number of inequality constraints, which include the side (bound)
constraints of the design variables for a simple formulation of the problem. Note that
equality constraints are omitted also, for simplicity, because the design requirements
in practice are usually given in inequalities. The structural optimization problem
is stated as
subject to Hj (x) 0, (j = 1, . . . , N I ) (9.12b)
The various definitions of the problem can be found in Chap. 2.
Consider the case where uncertainty exists in the vector of N parameters
a = (a1 , . . . , aN )> defining the loads, geometry, etc., and Hj (x) is regarded as
a function of a and x as Hj (a, x). Then the constraints should be given for the
worst-case scenario (maximum possible value) H b j (x) of Hj (a, x). Then the upper-
level optimum design problem is to be reformulated with the maximum value H b j (x)
as
b j (x) 0, (j = 1, . . . , N I )
subject to H (9.13b)
b
where Hj (x) is obtained by solving the following lower-level anti-optimization prob-
lem:
find b j (x) = max Hj (a, x)
H (9.14a)
a
subject to Ck (a) 0, (k = 1, . . . , N C ) (9.14b)

Here, the feasible domain C for a is defined by the inequality constraints Ck (a) 0
(k = 1, . . . , N C ). Note that if a simple approach is used, Problem (9.14) should be
solved N I times at each step of modification of x in the upper-level problem (9.13).
Alternatively, the two-level problem can be formally written as

b j (x) = max Hj (a, x) 0, (j = 1, . . . , N I )
subject to H (9.15b)
aC
9.4 Algorithms for Two-Level OptimizationAnti-Optimization
9.4.1 Cycle-based method

As can be seen from the formulations of the upper-level problem (9.13) and the
lower-level problem (9.14), we have to solve a nested optimization problem. The
simplest method for this purpose is the cycle-based method that is summarized as
follows:
(0)
Step 1: Assign an initial value aj for a corresponding to each of the inequality
constraints Hj (a, A) 0 (j = 1, . . . , N I ), and set the iteration counter k = 0.
(k)
Step 2: Solve Problem (9.13) for fixed a = aj , for the jth constraint to obtain
the optimal solution x(k) .
(k)
Step 3: Fix x at x(k) and find anti-optimal solution aj of Problem (9.14) for
each response function Hj (a, x(k) ) (j = 1, . . . , N I ), and set k k + 1.
Step 4: Go to Step 2, if the stopping criteria for the cycle are not satisfied.
The readers can consult with Sec. 6 of Gurav (2005), Lombardi and Haftka (1998),
etc., for the details.
In this method, the lower-level anti-optimization problem should be solved many
times at each iterative step of design optimization, if constraints are given for many
response quantities as Hj (a, x) 0 (j = 1, . . . , N I ). Since the anti-optimization
problem for finding the worst value of Hj (a, x) can be solved simultaneously, the
use of parallel computation will be very effective for reducing the computational
cost (Gurav, Goosen and van Keulen, 2005a). Although this cycle-based approach
is simple and easy to implement, this approach does not rapidly converge if the
anti-optimal solution strongly or nonlinearly depends on x.
Lombardi and Haftka (1998) applied this algorithm to the minimum weight
design of a composite laminated structure with displacement constraints, where the
thickness and angle of each ply are chosen as design variables, and the uncertainty
is introduced for the load components, which are expressed by the multipliers ai
(i = 1, . . . , N ) for the load pattern vectors p1 , . . . , pN . The displacement vector
against pi is denoted by U i = (U1i , . . . , Uni )> , where n is the number of degrees of
freedom. Then the total displacement vector, considering the uncertainty in the
load vector, is given as a linear function of a as

N
X
U= ai U i (9.16)
i=1
The bounds for a are defined by the linear inequalities
bL Ba bU (9.17)
where B is a constant matrix, and bL and bU are constant vectors of the lower and
upper bounds. Then the anti-optimization problem for finding the worst value of
the displacement component Uj is formulated as
N
X
find bj (x) = max
U ai Uji (9.18a)
a
i=1
subject to bL Ba b U
(9.18b)
where Uj has been assumed to be positive for simplicity. Then the cycle-based
method is applied to find the optimal distributions of the thickness and ply-angle.
It has been observed from their numerical experiments that the convergence of
the cycle-based method is very good if a is bounded with linear inequalities as in
Eq. (9.18b).
9.4.2 Methods based on monotonicity or convexity/concavity

Consider the case where the bounds for a are given by the interval
aLi ai aU
i , (i = 1, . . . , N ) (9.19)
and no other constraint is given. If the objective function Hj (a, x) that is to be

maximized in the lower-level anti-optimization problem (9.14) is a continuous and
concave function of a, then the anti-optimal solution is the global maximum of a
nonlinear programming problem, which can be easily found using a gradient-based
approach. If Hj (a, x) is a continuous and monotonic function of a, the global anti-
optimal solution exists at a vertex of the feasible region defined by Eq. (9.19), which
can be found easily from the signs of differential coefficients of Hj (a, x) with respect
to a.
If Hj (a, x) is a convex function of a, the anti-optimal solution also exists at the
vertex of the feasible region. In this case, the solution can be found by enumerating
the values of Hj (a, x) at the vertices of the feasible region. Note that the solution
can exist on an edge or on a hyper-plane for a non-generic case where the differential
coefficient of Hj (a, x) with respect to a parameter ai vanishes at the anti-optimal
solution.
Let V denote the set of 2N vertices of the feasible region; i.e., ai (i = 1, . . . , N ) is
equal to either aLi or aUi for the parameter set in V. Then the two-level optimization
problems (9.13) and (9.14) are combined into a single-level minimization problem
as

I
subject to Hj (a, x) 0, (j = 1, . . . , N ; a V) (9.20b)
Although this problem formulation is simple, the total number of constraints (9.20b)
is 2N N I ; hence, the computational cost increases as an exponential function of
the number N of the uncertain parameters, if all the vertices of the feasible region
must be enumerated.
There are several studies based on this enumeration approach (Ganzerli and
Pantelides, 2000) (see Sec. 4.6 for anti-optimization of a tensegrity structure by
enumeration, and Sec. 9.6 for hybrid optimizationanti-optimization of a truss under
stress and displacement constraints).
9.4.3 Other methods

9.4.3.1 Approximation methods
To reduce the computational cost of solving the lower-level anti-optimization prob-
lem with large numbers of design variables and uncertain parameters, approximation
methods such as response surface method (Myers and Montgomery, 1995) can be
used (see, e.g., Gurav, Goosen and van Keulen (2005a) and Secs. 3 and 7 of Gurav
(2005) for details of this approach). In this method, the responses are approximated
by a polynomial of the parameters as described in Sec. 2.5. Then the lower-level
anti-optimization problem turns out to be a small optimization problem that is eas-
ily solved by a gradient-based nonlinear programming for maximizing a polynomial
objective function in a convex feasible region for a.
Lanzi and Giavotto (2006) solved a multiobjective optimization problem for
simultaneously minimizing the total structural volume and maximizing the collapse
load in the postbuckling range, using a genetic algorithm with approximation by
neural networks, response surface, radial basis function (Nakayama and Hattori,
2003) or kriging method (Sakata, Ashida and Zako, 2003).
9.4.3.2 Parametric programming approach

In the process of solving the upper-level design optimization problem (9.13), it is
b j (x) for the jth
effective to incorporate the sensitivity of the anti-optimal solution a
constraint with respect to the design variables x, rather than fixing a b j (x) at each
step of optimizing x. Therefore, the parametric programming approach presented
in Sec. 2.4.3 can be used to accelerate the convergence of the cyclic optimization
process.
This idea has been mentioned in Kwak and Haug (1976) and Gurav, Goosen and
van Keulen (2005a), but has not been actually implemented. Baitsch and Hartmann
(2006) used this approach for optimizing the truss under constraint on strain energy
(compliance) considering uncertain geometrical parameters.
9.5 Optimization against Nonlinear Buckling
9.5.1 Introduction
There has been extensive controversy over the effectiveness of optimization under
nonlinear buckling constraints. Thompson and Supple (1973) cautioned the danger
of optimization due to interaction of critical modes at buckling:
If there is severe imperfection-sensitivity associated with the

coupling then the optimization cast in this form may be invalid.
We conclude, therefore, that any optimization scheme which is to
be applied to a structural system displaying multi-mode buckling
must be so defined as to take into account the complete buckling
characteristics including the effects of unavoidable manufactur-
ing imperfections.
As one can see by careful reading of their comments, they did not unconditionally
reject the idea of optimization against buckling. However, the terms if and so defined
have been omitted for some reason by some researchers to state that optimization
against buckling is unconditionally meaningless. The safety against buckling of
structures can be effectively improved by optimization if uncertainties in geometry,
material, etc., are appropriately incorporated. Furthermore, optimization does not
always enhance imperfection sensitivity, and for some structures, sensitivity is even
decreased by optimization.
For plates and frames, for which the prebuckling deformation is negligibly small,
the buckling loads can be accurately estimated by linear buckling analysis. Angular-
ply laminas are extensively studied for optimization under buckling constraints with
uncertain initial imperfection (Adali, Richter and Verijenko, 1997). Laminated
plates are also optimized considering uncertainty in material properties (Adali, El-
ishakoff, Richter and Verijenko, 1994a) and external loads (Adali, Lene, Duvaut and
Chiaruttini, 2003).
The convexity of the stability region described in Sec. 5.6 can be effectively
used for optimization under the worst linear buckling load (de Faria and Hansen,
2001a,b). De Faria and de Almeida (2003b) presented the optimization method of
a plate with variable thickness. They extended their methodology to optimization
for fundamental frequency considering geometrical stiffness under the worst loading
scenario (de Faria and de Almeida, 2006).
For flexible structures such as spherical shells and latticed domes in civil and
architectural engineering, the effect of prebuckling deformation should be incor-
porated for accurate estimation of the buckling load. Ohsaki (2000, 2002c, 2005)
Ui
Fig. 9.2 Equilibrium path of a geometrically nonlinear structure.
optimized shallow space trusses to find that imperfection sensitivity does not always
increase as a result of optimization, but is often reduced. Furthermore, optimization
of a shallow shell-type structure often leads to a coincident critical point, called the
hilltop branching point, as a coincidence of bifurcation point(s) and a limit point,
which is not always imperfection-sensitive (Ohsaki and Ikeda, 2007a). Therefore,
the critical load considering its reduction due to initial imperfection can be effec-
tively increased by optimizing the perfect system.
However, to be more comfortable with optimization against buckling, we can
directly incorporate the reduction of the critical load due to imperfection into the
process of optimization (Ramm, Bletzinger and Reitinger, 1993; Reitinger, Blet-
zinger and Ramm, 1994; Reitinger and Ramm, 1995). In this process, the worst-
case scenario should be taken into account to evaluate the lower bound of the crit-
ical load for the given magnitude of imperfection. Since optimization may lead to
imperfection-sensitive structures with multiple (coincident) buckling loads, if some
bifurcation points coincide, the imperfection sensitivity of such structures has been
extensively investigated (Ohsaki, 2002a,b; Ikeda, Ohsaki and Kanno, 2005; Ohsaki
and Ikeda, 2006). Ohsaki, Uetani and Takeuchi (1998b) presented an optimization
method for symmetric systems considering reduction of the maximum load due to
the worst imperfection. El Damatty and Nassef (2001) optimized a ribbed shell by
using a genetic algorithm. Ohsaki (2003a) defined the worst load case of flexible
structures by the strain energy stored at the deformation with the specified norm,
and optimized trusses under constraint such that the initial undeformed state is
recovered after releasing the loads corresponding to the worst-case scenario.
9.5.2 Problem formulation

Consider an elastic structure subjected to a set of proportional loads P = p
parameterized by the load factor and the constant load pattern vector p. The
nodal displacement vector U nonlinearly depends on as shown in Fig. 9.2, where
Ui is the representative displacement component.

Perfect
Perfect
Imperfect Imperfect
Imperfect Imperfect
q q
(a) Unstable-symmetric bifurcation point (b) Stable-symmetric bifurcation point
Fig. 9.3 Classification of bifurcation points and associated equilibrium paths of perfect and im-
perfect systems; : bifurcation point.
Suppose the structure reaches a critical point at = c as is increased from 0

(see Sec. 5.3 for the definition of the critical point). Note that the critical load factor
c is also called the buckling load factor or simply the buckling load. It is known
that optimization against nonlinear buckling is very difficult if the buckling loads
coincide (Ohsaki, 2000, 2005). However, only the simple critical point is considered
in this section; i.e.,
0 = c1 < c2 cn (9.21)
where ci is the ith eigenvalue of the tangent stiffness matrix at the critical point,
and n is the number of degrees of freedom.
Let x = (x1 , . . . , xm )> denote the vector of design variables. The first critical
load c (x) on the fundamental path is the function of x. The optimization problem
of minimizing the total structural volume V (x) under nonlinear buckling constraint
is simply formulated as
minimize V (x) (9.22a)
c
subject to c (x) (9.22b)
xLi xi xU
i , (i = 1, . . . , m) (9.22c)
c
where is the specified lower bound for the critical load, and xLi and xU
i are the
lower and upper bounds for xi , respectively.
The major disadvantage of this formulation is that the possible reduction of
the maximum (critical) load due to initial imperfection is not considered. Fig-
ures 9.3(a) and (b) illustrate the equilibrium paths of the structures exhibiting
an unstable-symmetric bifurcation and stable-symmetric bifurcation, respectively,
where q represents the generalized displacement in the direction of the buckling
mode, or the generalized antisymmetric component of the displacements. As is

c

b
(q, x) Optimization
b
(q, x)
c

q q
Fig. 9.4 Stabilization of the bifurcation path by optimization.
seen, for the unstable-symmetric bifurcation, the critical point of an imperfect sys-
tem turns out to be a limit point, which may be far below the bifurcation load of
the perfect system. However, for the stable-symmetric bifurcation point, the critical
point disappears due to the presence of imperfection.
For structures exhibiting unstable-symmetric bifurcation point, the critical load
reduced due to the worst initial imperfection with specified bound may be used
instead of c in Eq. (9.22b). However, the maximum load cannot be defined by a
critical load for the case of stable-symmetric bifurcation. In this case, the maximum
load may be defined by the upper-bound constraints on the responses such as stresses
and displacements. For example, consider a truss and let = (1 , . . . , N m )>
denote the vector of member stresses, where N m is the number of members. The
maximum load can be defined with the upper and lower bounds denoted by iU and
iL , respectively, of i as
m = max{|iL i () iU , i = 1, . . . , N m } (9.23)

Hence, the optimization problem may be defined as follows under constraint on the
maximum load m (x):
m
subject to m (x) (9.24b)
xLi xi xU
i , (i = 1, . . . , m) (9.24c)
m
where is the specified lower bound for the maximum load (see Sec. 5.5 for anti-
optimization of the linear buckling load of a braced frame using this definition of
the maximum load).
The maximum load of a structure exhibiting stable bifurcation can be alterna-
tively defined by the displacement constraint (Ohsaki, 2002b). Constraints on the
antisymmetric components of displacements can also be given for the case of un-
stable bifurcation point, because the antisymmetric deformation rapidly increases
with decreasing load after reaching the bifurcation load (Pietrzak, 1996).
The optimization problem can be alternatively formulated under the condition
such that the bifurcation point is stable; i.e., the structure is stabilized by opti-
mization so that the perfect system has the bifurcation path (initial postbuckling
path) with increasing load factor as illustrated in Fig. 9.4. Design sensitivity anal-
ysis of the responses on the postbuckling path is presented by Godoy and Taroco
(2000). Schranz, Krenn and Mang (2006a) investigated the characteristics of the
bifurcation path based on Koiters perturbation approach (Koiter, 1945), and pre-
sented a numerical approach in Schranz, Krenn and Mang (2006b) to convert an
imperfection-sensitive structure into an imperfection-insensitive one by placing ad-
ditional supports. Bochenek (2003) investigated the effect of the stiffness of spring
support on the shape of the postbuckling path in the loaddisplacement space.
Let b (q, x) denote the load factor along the bifurcation path of the perfect
system, which is a function of the generalized antisymmetric component q of the
displacement. Note that q = 0 corresponds to the bifurcation point. The curvature
b (x) of the bifurcation path at the bifurcation point is defined by

b 2 b (q, x)
(x) = (9.25)
q 2 q=0
Then the optimization problem under curvature constraint is formulated as

c c
subject to (x) (9.26b)
b
(x) 0 (9.26c)
By solving Problem (9.24), the structure is stabilized as illustrated in Fig. 9.4.

Bochenek (1997, 2003) also presented a problem of converting an asymmetric
bifurcation point to a symmetric one, which seems to be difficult because the sym-
metry of the bifurcation point generally depends on the geometrical property of the
structure.
Bochenek (2003) formulated the following optimization problem, considering
extended local stability, as

c c
subject to (x) (9.27b)
b b
(qi , x) (qi+1 , x) 0, (i = 1, . . . , r) (9.27c)
where q1 , . . . , qr+1 satisfying
0 = q1 < q2 < < qr+1 (9.28)
are the specified values of generalized displacement in the direction of the bifurcation
mode. By enforcing the constraint (9.27c), the structure with a non-decreasing
bifurcation path is obtained, and the limit point (snapthrough) in the bifurcation
path is also avoided. Bochenek (2003) formulated similar problems considering
multiple buckling loads.
(5) (6) H
(12)
5 6
(3) (4) H
(9) (10)
(11)
3 4
(1) (2) H
(7) (8)
y
1 2
x
W W
Fig. 9.5 A 12-bar tower truss.

The effect of geometrical imperfection on the maximum load is investigated, and
stabilization by design modification is demonstrated for a simple truss.
Consider a 12-bar tower truss as shown in Fig. 9.5 subjected to a vertical load
P at the top node 7, where P = 1 kN for simplicity. The member number and
node number are indicated by numbers with and without parentheses, respectively.
The dimension of the structure is given as W = 250 mm and H = 1000 mm. The
intersecting pair of diagonal members are not connected at their centers. Youngs
2
modulus is 200 kN/mm , and yielding or buckling of members is not considered. In
the following, the units of force and length are kN and mm, respectively, which are
omitted for brevity.
Geometrical nonlinearity is incorporated, and the strain of a member is measured
from the member lengths before and after deformation. The displacement increment
method is used for tracing the equilibrium path. No iteration is carried out at each
step, but the unbalanced load is canceled at the subsequent step.
The properties of the equilibrium paths are first investigated for different designs
with the same total structural volume V = 1.0 106 . The design parameter is the
ratio of the cross-sectional area of the diagonal members to that of the remaining
members, i.e., the cross-sectional areas are scaled for each value of so that V =
1.0 106 is satisfied. The nonlinear buckling load c of the perfect system is 3678.0
4000
3000
Load factor
2000
1000
0
0 100 200 300 400
Horizontal displacement of node 7
Fig. 9.6 Relation between and U for = 0.01; solid line: = 1.0; dashed line: = 0.01.
for the truss with uniform cross-sectional area; i.e., = 1.0, and is 3842.1 for
= 0.01. Therefore, the truss with a given total structural volume has a larger
buckling load if the diagonal members have smaller cross-sectional areas than the
vertical and horizontal members.
An imperfection is given in the x-direction of the top node 7. Consider first the
case where = 0.01. The relation between P and the x-directional displacement
U of node 7 for = 1.0 is shown in the solid line in Fig. 9.6. As is seen, the
horizontal displacement suddenly increases as P approaches c (= 3678.0). Note
that P is a monotonically increasing function of U . For = 0.01, the equilibrium
path is plotted in a dashed line in Fig. 9.6. The load factor for = 0.01 is larger
than that for = 1.0 in the range of small U . However, gradually decreases
as U is increased after approaching the bifurcation load of the perfect system. In
fact, the value of for = 0.01 is smaller than that for = 1.0 for the range
U 160. Therefore, a stable bifurcation with smaller bifurcation load is preferred
to an unstable bifurcation with larger bifurcation load if the allowable displacement
is moderately large.
Figure 9.7 shows the equilibrium paths of imperfect systems for = 1.0, where
the solid and dashed lines correspond to the designs with = 1.0 and 0.01, respec-
tively. As is seen, the equilibrium path has a limit point for = 0.01, and the load
factor monotonically increases for = 1.0. The load factor for = 0.01 is smaller
than that for = 1.0 in the range U 185.
Optimal design is next found for = 1.0, where is chosen as the design
parameter. The allowable displacement is given as U U = 200; i.e., the maximum
load m is defined as the limit point load in the range U < U U or the load factor
m
at which U = U U is satisfied, and the specified maximum load factor is 4000.0.
The relation between the design parameter and the total structural volume for
m
satisfying = 4000.0 is plotted in Fig. 9.8. The optimal design that minimizes
the total structural volume exists at ' 0.027, and the corresponding optimal
4000
3000
Load factor
2000
1000
0
0 100 200 300 400
Fig. 9.7 Relation between and U for = 1.0; solid line: = 1.0; dashed line: = 0.01.
1400000
Total structural volume
1200000
1000000
800000
0 0.02 0.04 0.06 0.08 0.1
Design parameter
Fig. 9.8 Relation between design parameter and total structural volume.
objective value is about 9.0467 105 . This optimal solution with a small value of
agrees with the fact that a tower-type structure can be effectively stabilized by
adding very slender braces. Figure 9.9 shows the relation between and U for
designs satisfying m = 4000.0. As is seen, the maximum load is achieved at a
limit point for = 0.01 and at U = U U = 200 for = 1.0. Consequently, for the
optimal design = 0.027, has the maximum value at a limit point that satisfies
U = U U = 200.
4500
4000
Load factor
3500
3000
0 100 200 300 400
Fig. 9.9 Relation between and U for designs satisfying m = 4000.0 ( = 1.0); solid line:
= 0.027; dashed line: = 0.1; dotted line: = 1.0.
1 (1) 2 (2) 3
2.0 m
(5) (6)
(8) (7) (9)
(10)
(3) (4) 6
y 4 5 P2
P1 P3
x
2.0 m 2.0 m
9.6 Stress and Displacement Constraints
An example of optimization of trusses under stress and displacement constraints

is presented in this section considering the worst load scenario. Consider a 10-bar
truss as shown in Fig. 9.10. The examples below are basically the same as those in
Elishakoff, Haftka and Fang (1994c); however, the geometry and mechanical prop-
erties have been modified to use SI units. The total structural volume is minimized
under constraints on stresses and displacements. Let i denote the stress of the
ith member. The upper and lower bounds denoted by iU and iL , respectively,
2 2
are 0.2 kN/mm for all members except 0.6 kN/mm for member 9 defined in
Fig. 9.10. The external loads are P1 = 1.0, P2 = 4.0, P3 = 1.0 (kN). Youngs mod-
2
ulus is 200.0 kN/mm . The units for force and length are kN and mm, respectively,
Table 9.1 Cross-sectional areas and the structural vol-

ume of the optimal solution (deterministic loads).
Member number Ai (mm2 )

Case 1 Case 2
1 5.3536 6.0910
2 1.0000 1.0000
3 5.3536 5.3502
4 16.0000 15.3359
5 3.6464 4.9652
6 1.0000 1.0000
7 13.6421 13.6469
8 1.0000 1.0000
9 2.8284 7.5264
10 1.4142 1.0000
Total volume (mm3 ) 1.1812 105 1.3303 105
in the following examples.

The objective here is to minimize the total structural volume under stress and
displacement constraints. Optimization is carried out by SNOPT Ver. 7 (Gill, Mur-
ray and Saunders, 2002) based on the sequential quadratic programming algorithm.
The cross-sectional areas of all members are independent variables; i.e., we have ten
design variables. The lower bound ALi for the cross-sectional area Ai is 1.0 for all
members, and the upper bounds are not given.
Consider first the optimization under deterministic loads, where the loads P1 ,
P2 , and P3 are applied simultaneously as shown in Fig. 9.10. Optimal solutions
are found for two cases: without displacement constraints (Case 1), and with upper
bound U U = 5.0 for negative y-directional displacement U6 at node 6 (Case 2). The
optimal cross-sectional areas and the total structural volume are listed in Table 9.1.
As is seen, the members 2, 6, and 10 connected to node 3, and member 8 have small
cross-sectional areas. Since the displacement constraint is active at the optimal
solution for Case 2, it has a larger objective value than Case 1.
Consider next the case with uncertain loads. Let the nominal loads be defined
as P10 = 1.0, P20 = 4.0, P30 = 1.0. The lower and upper bounds PiL and PiU ,
respectively, for Pi are given as
PiL = 0.9Pi0 , PiU = 1.1Pi0 , (i = 1, 2, 3) (9.29)
Since the stresses and displacements are linear functions of the loads, the maximum
and minimum values can be found by enumerating the eight sets of loads P V k (k =
1, . . . , 8) at the vertices of the feasible region defined by
P1L P1 P1U , P2L P2 P2U , P3L P3 P3U (9.30)

Therefore, the optimization problem with the worst load scenario is formulated as
Xm
minimize V (A) = A i Li (9.31a)
i=1
subject to iL i (A, P V U
k ) i , (i = 1, . . . , m; k = 1, . . . , 8) (9.31b)
U6 (A, P V U
k ) U , (k = 1, . . . , 8) (9.31c)
ALi Ai , (i = 1, . . . , m) (9.31d)
where Li is the length of the ith member, and m is the number of members that is
equal to 10. In this formulation, we have 168 inequality constraints in addition to
the side constraints (9.31d) for Ai .
Alternatively, the maximum and minimum stresses and displacements for the
feasible set of loads defined by Eq. (9.30) can be found by combining the responses
against each load P1 , P2 , and P3 applied independently. Let ij denote the stress of
member i against the load Pj0 . From the definition (9.29) of load uncertainty, the
maximum imax and minimum imin of i are found as
X3
imax = (1 + 0.1 sign ij )ij ,
j=1
3
(9.32)
X
imin = (1 0.1 sign ij )ij
j=1
For example, if ij is negative, the maximum and minimum values of the stress of
the ith member to the jth load are (1 0.1)ij = 0.9ij and (1 + 0.1)ij = 1.1ij ,
respectively. Then the constraints (9.31b) can be written as
iL imin , imax iU , (i = 1, . . . , m) (9.33)
This way, the number of constraints is reduced to 21, including the displacement
constraint that can be defined in a similar manner as Eqs. (9.32) and (9.33).
Table 9.2 shows the optimal solutions without displacement constraint (Case 1)
and with displacement constraint (Case 2). It is seen by comparing the results in
Tables 9.1 and 9.2 that the optimal objective value significantly increases, especially
for Case 2, as a result of load uncertainty.
Table 9.3 shows the values of stresses at the eight vertices of the feasible region
of the optimal solution for Case 1, where U and L denote that the load takes
upper- and lower-bound values, respectively. For example, (U,L,U) corresponds to
the vertex defined by (P1 , P2 , P3 ) = (P1U , P2L , P3U ). The underline indicates that
the stress is equal to the lower or upper bound; i.e., the member is fully stressed.
As is seen, each member is fully stressed for at least one loading condition, except
members 8 and 9. Note that the cross-sectional area of member 8 is equal to the
lower bound and the stress constraint is satisfied with inequality. Since member 9
has a larger range of allowable stress, the stress is not equal to either the upper or
lower bound; i.e., the optimal solution under multiple loading conditions is not fully
stressed when the load uncertainty is considered and the bounds of the stresses are
not given uniformly.
Table 9.2 Optimal cross-sectional areas and the struc-

tural volume of the optimal solution (uncertain loads).
Member number Ai (mm2 )

Case 1 Case 2
1 6.0932 14.3939
2 1.0000 1.0000
3 8.7550 10.4616
4 18.2481 17.6119
5 4.1565 4.2834
6 1.0000 1.0000
7 14.8966 11.9314
8 1.0000 7.0584
9 3.3680 12.7440
10 1.4142 1.0000
Total volume (mm3 ) 1.3699 105 1.9009 105
Table 9.3 Stresses at vertices of the optimal solution (Case 1).
Member number (U,U,U) (U,U,L) (U,L,U) (U,L,L) (L,U,U) (L,U,L) (L,L,U) (L,L,L)
1 0.1958 0.1589 0.2000 0.1631 0.1930 0.1561 0.1972 0.1602

2 0.1919 0.1488 0.1992 0.1562 0.1927 0.1496 0.2000 0.1570
3 0.1363 0.1791 0.0478 0.0906 0.1572 0.2000 0.0687 0.1115
4 0.1914 0.2000 0.1479 0.1565 0.1914 0.2000 0.1480 0.1566
5 0.1960 0.1643 0.1882 0.1564 0.2000 0.1683 0.1921 0.1604
6 0.1919 0.1488 0.1992 0.1562 0.1927 0.1496 0.2000 0.1570
7 0.2000 0.1834 0.1976 0.1810 0.1827 0.1660 0.1802 0.1636
8 0.1319 0.0965 0.1679 0.1324 0.1076 0.0721 0.1434 0.1080
9 0.3813 0.3154 0.3783 0.3123 0.3810 0.3151 0.3779 0.3120
10 0.1918 0.1488 0.1992 0.1562 0.1926 0.1496 0.2000 0.1569
9.7 Compliance Constraints
9.7.1 Introduction
It is well known that the interval analysis for anti-optimization of structures under
bounded uncertainty leads to a too conservative result if the dependency among the
uncertain variables are not appropriately incorporated; see Sec. 3.3 for details. One
strategy to avoid this unrealistic situation is to reduce the number of independent
uncertain parameters using Fourier transformation or modal decomposition where
the bounds are to be given for the coefficients of the modes, or simply by linking the
variables (Elishakoff, 1999). However, in these approaches, explicit bounds cannot
be assigned for each variable. Therefore, bounds should be given as additional
linear constraints, because the coefficient itself has no physical meaning related to
uncertainty in external loads, material parameters, geometrical properties, and so
on.
Consider, for example, a finite-dimensional elastic structure subjected to earth-

quake loads. The standard approach to reduce the computational cost of the eval-
uation of the maximum elastic response may be the response spectrum approach
(Clough and Penzien, 1975), which utilizes modal decomposition of the responses.
After computing the maximum response of the each dominant mode, the max-
imum responses of structure can be found by the square-root-of-sum-of-squares
(SRSS) method or complete quadratic combination (CQC) method (Wilson, Der
Kiureghian and Bayo, 1982). However, in the practical situation of civil engineering,
dynamic analysis is not carried out for the design of low-rise building frames, and
the equivalent static loads are usually used for evaluating the maximum elastic and
elastoplastic responses by the so-called static pushover analysis; see, e.g., FEMA
(2000); Han and Chopra (2006). Therefore, in this section, an example is shown for
static anti-optimization with additional linear constraints on uncertain parameters.
Let U (t) denote the time-dependent vector of nodal displacements of a finite-
dimensional structure, which is decomposed to modal responses as
n
X
U (t) = ci (t)i (9.34)
i=1
where i is the ith eigenmode of undamped free vibration, ci (t) is the modal re-
sponse that is a function of time, and n is the number of degrees of freedom.
The equivalent static load P i corresponding to i is defined as
P i = i M i (9.35)
where M is the mass matrix, and i is the ith eigenvalue of undamped free vibra-
tion.
Suppose a single mode, say mode i, dominates in the response. For example, for
a building frame, the response can be evaluated with good accuracy by using the
first mode only. In this case, the maximum response U max
i corresponding to mode
i is computed from
KU max
i = cmax
i Pi (9.36)
where K is the stiffness matrix, and cmax
i is the maximum modal response obtained
from the displacement response spectrum, which is multiplied by the participation
factor of mode i (see Sec. 6.6 for examples of response spectra).
Let R denote the representative response, such as stress or strain of the specified
member or element, which is a linear function of U , as
R = b> U (9.37)
where b is a constant vector. The maximum absolute value of the response corre-
sponding to mode i against the equivalent static load cmax
i P i for seismic design,
called design load for simplicity, is obtained as
Rimax = |b> U max
i | (9.38)
Even when a few modes dominate in the response of a structure, only those
dominant modes can be used for constructing the design loads. Suppose we use the
three modes 1, 2 and 3, for simplicity, and assume the response R is positive for
the maximum responses of all the modes. The worst possible static load pattern is
then given as
P worst = cmax
1 P 1 + cmax
2 P 2 + cmax
3 P3 (9.39)
which corresponds to the absolute sum approach of the modal combination.

However, in Eq. (9.39), all the three modal responses are assumed to take their
maximum simultaneously at a certain time instance during vibration. Therefore,
Eq. (9.39) overestimates the design load and the corresponding responses. The
standard approach to avoid this overestimation is to use the SRSS combination.
For example, the maximum displacements are estimated as
q
UjSRSS = (U1,jmax )2 + (U max )2 + (U max )2 , (j = 1, . . . , n)
2,j 3,j (9.40)
max
where Ui,j is the jth component of U max
i . However, we can assign more explicit
constraints on cmax
i in Eq. (9.39) to give a set of design loads that consider uncer-
tainty in the combination of the coefficients; e.g.,
cL cmax
1 + cmax
2 + cmax
3 cU (9.41)
where cL and cU are the specified lower and upper bounds, respectively. More rig-
orous constraints can be incorporated based on the statistical data of maximum
responses of structures to a set of recorded seismic motions. Zhang, Ohsaki and
Uchida (2008) presented an approach to generating several design loads as combi-
nations of multiple modes.
The design load vector is alternatively defined by using only the most dominant
mode, say 1 , and adding the uncertain load deviation vector P dev for representing
the contribution of the higher modes:
P = cmax
1 P 1 + P dev ,
dev
(9.42)
Pj,L Pjdev Pj,U
dev
, (j = 1, . . . , n)
dev dev
where Pj,L and Pj,U are the specified lower and upper bounds for Pjdev , respectively.
However, also for this case, it is unrealistic that all Pjdev will take their maximum
or minimum at the same time instance during vibration. Therefore, we need to
incorporate additional constraints on Pjdev .
In the following, we present a general procedure for anti-optimization of static
responses with additional linear constraints. As an example, anti-optimization prob-
lem of a truss is formulated for maximizing the compliance (external work) under
uncertain static nodal loads. The optimum design problem under compliance con-
straint is solved by a simple heuristic approach called the greedy method.
9.7.2 Optimization problem and optimization algorithm

Let A = (A1 , . . . , Am )> and L = (L1 , . . . , Lm )> denote the vectors of cross-
sectional areas and the lengths of members of a truss. The objective function
to be minimized is the total structural volume defined as
V (A) = A> L (9.43)
The truss is subjected to static nodal loads P , for which the nodal displacement
vector U is found from
U = CP (9.44)
where C is the flexibility matrix; i.e., C = K 1 with K being the stiffness matrix.
A constraint is given for the compliance W , defined by
W (P , A) = U > P = P > CP (9.45)
which is a quadratic convex function of P , because C is positive definite for a stable

structure. The compliance is equivalent to the external work (twice of the strain
energy), and smaller compliance under specified loads leads to a stiffer structure.
For simplicity, the nodal loads are conceived as uncertain variables; i.e., let
a = P and N = n in the general formulation in Sec. 9.3. The anti-optimization
problem (lower-level problem) for finding the worst load scenario for maximizing
the compliance is formulated as
find c (A) = max W (P , A) = P > C(A)P

W (9.46a)
P
subject to b> s
i P di , (i = 1, . . . , N ) (9.46b)
PiL Pi PiU , (i = 1, . . . , n) (9.46c)
where bi is a constant vector, di is a constant, and N s is the number of additional

linear constraints. Note that C(A) depends on A which is the design variable
of the upper-level problem. Since C(A) is positive definite, Problem (9.46) is a
maximization problem of a convex function over a convex feasible region. Therefore,
the anti-optimal solution exists at a vertex of the feasible region.
The optimum design problem (upper-level problem) under constraint on com-
pliance corresponding to the worst load scenario is formulated as
minimize V (A) = A> L (9.47a)

c (A) W U
subject to W (9.47b)
ALi Ai AU
i , (i = 1, . . . , m) (9.47c)
where W U is the specified upper bound for W , and ALi and AU i are the lower and
upper bounds for Ai .
Since the constraint function of (9.47b) is not generally continuously differen-
tiable with respect to A, a simple heuristic approach called the greedy method is
used for design optimization (see Sec. 2.6 for details of heuristic approaches). The
design variable Ai is discretized as
Ai {jA | j = 1, . . . , N D } (9.48)
D
where A is the specified unit value of the cross-sectional area, and N is the
number of possible values of each cross-sectional area. The algorithm is summarized
as follows:
(0)
Step 1 Assign initial value Aj for Aj (j = 1, . . . , m), and set the iteration counter
k = 0.
Step 2 Find the anti-optimal solution P (k) and the corresponding worst value
c (A(k) ) of W , for A fixed at A(k) , by enumerating the values of W at all the
W
vertices of the feasible region defined by Eqs. (9.46b) and (9.46c).
Step 3 If Wc > W U , increase Aj (j {1, . . . , m}) by A for the member which
satisfies Aj < N D A and maximizes the ratio of the decrease of W c to the
c U
increase ALj of the structural volume. If W W , decrease Aj by A for
a member which satisfies Aj > A and minimizes the ratio of the increase of
c to the decrease of the total structural volume.
W
Step 4 Set k k + 1 and go to Step 3, if the termination condition is not satisfied.
Step 5 Output the best solution satisfying the constraints, and terminate the pro-
cess.

Consider a 15-bar arch-type truss as shown in Fig. 9.11 subjected to static
nodal loads. The nodal coordinates are listed in Table 9.4. Youngs modulus is
2
200.0 kN/mm . The units of length and force are mm and kN, respectively, which
are omitted for simplicity in the following.
The vertical loads at nodes 1, 2 and 3 are denoted by P1 , P2 and P3 , respectively,
which are uncertain variables. Uncertainty is directly introduced to the loads, in-
stead of using the nominal values and uncertain parameters also in this section, for
simple presentation of problem formulation and numerical results.
Linear constraints are given for P1 and P3 as
D P1 P3 D (9.49)
The upper bound D characterizes the asymmetry of the loads; i.e., the loads are
symmetric if D = 0. Let P denote the vector of nonzero components of the loads
as P = (P1 , P2 , P3 )> . The vertical displacements U1 , U2 and U3 at nodes 1, 2 and
3 are the linear functions of P .
The anti-optimization problem is formulated as
find c (A) = max W (P , A)
W (9.50a)
P
subject to D P1 P3 D (9.50b)
PiL Pi PiU , (i = 1, 2, 3) (9.50c)
5 (6) 6
(5) (7)
y (11) (12)
4 (10) (13) 7
2
(9) (2) (3) (14)
1 3
(8) P2
(15)
(1) P1 P3 (4)
8 9
x
Fig. 9.11 A 15-bar arch-type truss.
Table 9.4 Nodal coordinates (mm)

of the arch-type truss.
Node number x y
1 2000.0 1500.0
2 4000.0 2000.0
3 6000.0 1500.0
4 1000.0 1750.0
5 3000.0 2750.0
6 5000.0 2750.0
7 7000.0 1750.0
8 0.0 0.0
9 8000.0 0.0
which is a maximization problem of a quadratic convex function over a convex

region. Therefore, the anti-optimal solution exists at a vertex of the feasible region.
First, we investigate the properties of the anti-optimization problem by fix-
ing the cross-sectional areas of all members at 100.0 mm2 . The values of PiL
and PiU , respectively, are 10.0 and 10.0 (kN) for i = 1, 2, 3, which lead to
D 20. Therefore, for D < 20, we have 12 vertices (P1 , P2 , P3 ) = (10, 10, 10),
(10D, 10, 10), (10, 10, 10D), (10, 10, 10), (10+D, 10, 10), (10, 10, 10+
D), (10, 10, 10), (10 + D, 10, 10), (10, 10, 10 + D), (10, 10, 10),
(10 D, 10, 10), (10, 10, 10 D).
The following properties hold between the compliance W (P , A) and nodal loads
P for fixed cross-sectional areas:
P1: W (P , A) takes the maximum value at a vertex of the feasible region of P

defined by the constraints (9.50b) and (9.50c).
P2: W (P , A) does not change by the simultaneous reversal of the signs of the loads
as P P .
P3: W (P , A) monotonically increases if the absolute values of all the components
of P are monotonically and simultaneously increased.
Table 9.5 Values of compliance W at the vertices of the feasible

region, as well as the worst values and the corresponding worst
load patterns, for various values of D.
(P1 , P2 , P3 ) D=6 D = 12 D = 18
(10, 10, 10) 123.67 123.67 123.67

(10 D, 10, 10) 96.270 102.31 141.79
(10, 10, 10 + D) 56.704 79.699 136.14
(10, 10, 10) 67.150 67.150 67.150
Worst value 123.67 123.67 141.79

(10, 10, 10) (10, 10, 10) (2, 10, 10)
140
130
Compliance
120
110
100
P1*
90
-6 -4 -2 0 2 4 6 8 10
P1
Fig. 9.12 Variation of compliance against P1 for P2 = P3 = 10.0.
P4: From the symmetry of the truss, W (P , A) does not change by exchanging the
values of P1 and P3 .
From the properties P1 and P2, we can fix P2 at 10.0, because Eq. (9.50b)
does not include P2 . From P4, we can assume that P1 P3 . Then utilizing
the convexity of W with respect to P , the anti-optimal solutions are found by
enumerating the vertices (P1 , P2 , P3 ) = (10, 10, 10), (10D, 10, 10), (10, 10, 10+
D) and (10, 10, 10). Thus the number of vertices for which the responses are to
be evaluated is reduced from 12 to 4. The values of W at the vertices corresponding
to various values of D are listed in Table 9.5, along with the worst value among the
four vertices.
Variation of the compliance W is plotted in Fig. 9.12, for example, with respect
to P1 for P2 = P3 = 10.0. The value of W for P1 = 10 is 123.67, and W = 123.67 is
satisfied at P1 = P1 ' 5.833 as indicated in Fig. 9.12. Since P3 = 10.0, the lower
bound of P1 is 10.0 D from Eq. (9.50b). Hence, the anti-optimal value (worst load
scenario) of P1 corresponding to the maximum value of W is 10.0 for D < 15.833,
Fig. 9.13 Optimal cross-sectional areas of the arch-type truss.
and is 10.0 D for D 15.833.

We next find the optimal solution under constraints on the compliance corre-
sponding to the worst load scenario for D = 10. Let A1 , A2 and A3 denote the
cross-sectional areas of the lower chords, upper chords, and diagonals, respectively.
The design variable Ai is discretized as
Ai = jA, (i = 1, 2, 3; j = 1, . . . , 20) (9.51)
where A = 10. The worst load scenario is searched for at the vertices
(P1 , P2 , P3 ) = (10, 10, 10), (0, 10, 10), (10, 10, 0), and (10, 10, 10) for 20 3 = 8000
designs to find the exact optimal solution (A1 , A2 , A3 ) = (170, 90, 60), where
V = 2.8411 106 and W c = 98.391. The optimal solution is illustrated in Fig. 9.13,
where the width of each member is proportional to its cross-sectional area. Note
that the lower chords have large cross-sectional areas in the optimal solution. The
worst load for this optimal design is P1 = P2 = P3 = 10. Note that the number of
solutions to be enumerated can be reduced by using the current minimum value V
of the solution satisfying the compliance constraint. For example, in the process of
increasing A1 while fixing A2 and A3 , the process is terminated if V exceeds V ,
and A2 is next to be increased.
The optimal solution has also been found by the greedy method presented in
Sec. 9.7.2. The process is terminated in 30 iterations. Among the ten solutions
found from ten randomly generated initial solutions, seven solutions coincided with
the optimal solution (A1 , A2 , A3 ) = (170, 90, 60) found by enumeration. Therefore,
for this example, optimal solution can be easily found by the greedy method, for
which the total number of design evaluation is 30 10 = 300, that is far smaller
than the 8000 for the enumeration approach.
9.8 Homology Design
9.8.1 Introduction
A great deal of attention is paid to structure/control-integrated optimum design
for advanced structures, such as space structures, adaptive structures and smart
structures, which are required to fulfil complicated missions at high quality with
Advanced structures
Structure
High quality
of performance
Mechanism Control
Fig. 9.14 Concept of advanced structure.
the aid of active control systems. The concept of advanced structures is illustrated
in Fig. 9.14 (Kida and Komatsu, 1993; Thomas, Sepulveda and Schmit, 1992).
The precise control on geometrical properties of these structures is stringently
demanded for high quality performance, and a sophisticated structural design
methodology considering the interaction with an active control system and mech-
anism should be devised. Homology design can be a candidate for such structural
optimization of advanced structures, utilizing the concept of homologous defor-
mation, in which a prescribed geometrical property at a part of the structure is
maintained before, during, and after the deformation. The resolution of a huge and
precise radio telescope was greatly enhanced by realizing homologous deformation,
which kept, the shape of the reflector structure parabolic in spite of the deformation
caused by its own weight (Morimoto, Kaifu, Takizawa, Aoki and Sakakibara, 1982;
von Hoerner, 1967). The control cost of the active system to adjust the shape of the
parabolic reflector is greatly reduced and the resolution quality is ensured easily.
Two formulations based on the finite element method have been derived for ho-
mology design in static problems (Hangai, 1990; Yoshikawa and Nakagiri, 1990).
However, the formulations were given in a purely deterministic manner, where the
external loadings must be specified. The advanced structures are to be made very
flexible owing to the lightweight requirement. Thus, the homologous deformation
under specified loading appears to be easily disturbed by the uncertain fluctua-
tion of loading. The sensitivity of homologous deformation with respect to such
uncertainties should be estimated prior to the application of homology design.
In this section, we closely follow the results of Yoshikawa, Elishakoff and Nak-
agiri (1998a). The formulation of homology design with an illustrative numerical
example of an 11-bar truss is given, using finite element sensitivity analysis, prior
to the discussion on uncertain loading. The uppermost deck subjected to uniformly
distributed vertical load as nominal loading is kept both straight and horizontal,
before and after the deformation by homology design. The methodology for the es-
timation of the disturbed homologous deformation is developed based on the convex
model of uncertain loadings (Ben-Haim and Elishakoff, 1990; Elishakoff, Lin and
Zhu, 1994e), in which uncertain variables are assigned to discretized nodal forces
and the existence domain of the uncertain variables is confined within the convex
hull of hyperellipse. The worst case of homology design is estimated on the bound-
ary of the convex hull as the point that maximizes the error index. The latter is
defined as the square of the Euclidean norm of displacements error from objective
homologous deformation. The validity of the proposed method for the worst-case
estimation of the homology design is demonstrated in a numerical example of the
11-bar truss after homology design.
9.8.2 Deterministic loading

Consider the static deformation of the linear and elastic structure discretized by the
finite elements. The formulation for homology design under deterministic loading
is summarized first. The structural response is governed by the following stiffness
equation in matrix form after incorporation of the geometrical boundary condition:
KU = P (9.52)
where K is the n n stiffness matrix, U is the unknown displacement vector, P

is the external nodal load vector, and the number of degrees of freedom of the
discretized structure is denoted by n.
In the following, a simple iterative approach is presented without resort to a
sophisticated nonlinear programming algorithm. For this, we need a trial design to
satisfy the constraint of homologous deformation
H(U ) = 0 (9.53)
which consists of J equations with respect to the nodal displacements. The way of
representation is not unique even for the same homologous deformation. The success
of the objective homology design greatly depends on the manner of determining an
adequate trial design and neat representation of the homologous constraint. For
the sake of simplicity of discussion, we treat only linear homologous deformation,
the constraint for which is given by linear equations as follows with respect to the
nodal displacements:
CU = d (9.54)
where C is the constant J n matrix, and d is the constant vector of J components.

If we find an adequate trial design a priori, the nodal displacements obtained
by solving the stiffness equation (9.52) satisfy the homologous constraint in (9.54)
and there is no need to change the trial design. In general, however, it is difficult
to find the transparent homology design. Therefore, we are interested in changing
the trial design such that the objective homologous deformation is achieved under
specific loading. For the design change, we judiciously select structural parameters
p1 , . . . , pm and assign the design variables 1 , . . . , m in the form of
pi = pi (1 + i ), (i = 1, . . . , m) (9.55)
The effect of the design change is linearly approximated and the change of the nodal
displacement vector is expressed in the following linear form:
m
X
U =U+ U
i i (9.56)
i=1
where U is the displacement vector of the current design, and U i is the first-order
displacement sensitivity vector with respect to i , obtained by finite element sensi-
tivity analysis (see Sec. 3.6.2 for details of static sensitivity analysis). Substituting
the linearly approximated nodal displacement vector of Eq. (9.56) into the homolo-
gous constraint in Eq. (9.54), we obtain the governing equation of design variables
in the form
Xm
CU i i = d CU (9.57)
i=1
which can be rewritten in a matrix form as

G = b (9.58)
>
where G is a J m rectangular matrix, = (1 , . . . , m ) is the design variable
vector to be determined, and b = d CU is a constant vector. If m is larger than
J, the governing equation (9.58) has solutions, which in general, are not unique.
On the other hand, if J is larger than m, Eq. (9.58) scarcely has any solutions. The
prediction about the existence of solutions based on the size of the coefficient matrix
G mentioned above suggests that the increase of the number of design variables
enlarges the possibility of success in homology design.
The solution to Eq. (9.58) is efficiently handled with the MoorePenrose gener-
alized inverse (Rao and Mitra, 1971). The general solution of Eq. (9.58) is given
as
= G b + (I G G)h (9.59)
when the condition of solution existence given by
(I G G)b = 0 (9.60)
is satisfied, where G is the MoorePenrose generalized inverse of G, I is the J J
identity matrix, and h is an arbitrary vector. The first and second terms of the
right-hand-side of Eq. (9.59) are called particular and complementary solutions, re-
spectively. Equation (9.59) indicates that the way of changing the current trial
design to realize objective homologous deformation in general is not unique, where
arbitrariness creeps into the general solution of the design variable vector through
the arbitrary vector h. We employ only the particular solution to determine the
design variable vector for design change. The solution thus determined is the solu-
tion of least design change, because the Euclidean norm of design variable vector
becomes a minimum owing to the property of the MoorePenrose generalized
inverse.
1 2 3 4
6 1500 mm
5 3000 mm 7
Fig. 9.15 An 11-bar truss structure.
0.1 mm
Fig. 9.16 Deformed shape of the initial trial design.
The objective homologous deformation is not realized strictly by the above men-
tioned design modification, since the governing equation (9.58) is derived based on
the linear approximation of nodal displacements change. The deficient approxima-
tion can be compensated by the iterative update of the design variables, in which
the updated structure according to the solution of Eq. (9.58) is used as the trial
design for the next step. The iterative process is stopped when the homologous
constraint (9.54) is regarded to be satisfied, that is, the Euclidean norm of CU d
becomes sufficiently small.

An instructive example of homology design is demonstrated by using the 11-bar
truss illustrated in Fig. 9.15. The cross-sectional area of all the members is 100 mm 2
in the initial trial design, and Youngs modulus is 70.0 GPa. The structure is sub-
jected to vertical nodal loads that are consistent with a uniformly distributed ver-
tical load of 1.0 N/mm as nominal loading on the uppermost deck. The deformed
structure of the initial trial design is depicted in Fig. 9.16. The measure for displace-
ment is put at the upper right of Fig. 9.16. The maximum vertical displacement
turns out to be 0.133 mm in the initial trial design.
The objective homologous deformation is set so as to keep the uppermost deck
straight and horizontal before and after deformation. Hence, the homologous con-
straint equation is given as

v1 v m 0
v2 vm 0

v3 vm = 0 (9.61)
v4 v m 0
100 156 100
327 104 104 326

48.5 48.5
143 142
unit: mm2
Fig. 9.17 Realized homologous deformation and the final cross-sectional areas.
where vi is the vertical nodal displacement of node i, and vm is the average of vertical
nodal displacements on the uppermost deck, that is, vm = (v1 + v2 + v3 + v4 )/4.
The realized homologous deformation by changing the cross-sectional areas of all
the members is illustrated in Fig. 9.17. The vertical displacement of the uppermost
deck turns out to be 0.0516 mm. The manner of illustration of the deformation is
the same as that of Fig. 9.16. In Fig. 9.17, the cross-sectional area of the member
after homology design is indicated by the number attached to each member. The
increase in total weight is 58.7% in this example. Three update steps are needed
for the design variables to ensure a sufficiently small Euclidean norm of the vector
on the left-hand-side of Eq. (9.61); i.e., less than 0.001 mm in this case. One of
the transparent homology designs for this example is constructed so as to make the
cross-sectional area of vertical and outer horizontal members extremely large. This
method inevitably results in an extreme increase of the total weight.
9.8.4 Convex model of uncertain loading

The legitimate question arises of how sensitive the homology design is to uncer-
tain fluctuation of the loadings. The behavior of the structure is described by the
following stiffness equation in the state of the homology design:
KHU = P 0 (9.62)
where K H corresponds to the stiffness matrix of the structure experiencing homol-
ogous deformation at the specified nominal loading P 0 . The fluctuation of loading
is represented by introducing n uncertain variables ai assigned to n nodal force
components as
Pi = Pi0 (1 + ai ), (i = 1, . . . , n) (9.63)
The amplitude of variation (P10 a1 , . . . , Pn0 an )> from the nominal loading is assumed
to be unknown-but-bounded by a convex hull of a hyperellipse with respect to
uncertain variables given as
q 2 a> W a 1 (9.64)
where a = (a1 , . . . , an )> , and W is a symmetric and positive definite matrix that
defines the shape of the convex hull. The magnification coefficient q governs the
size of the convex hull. W and q are determined from the available information.
9.8.5 Worst-case estimation

In the context of convex analysis, the problem is posed as follows:
Find the most harmful (worst-case) amplitude of variation to maximize the dis-
tortion from objective homologous deformation.
The fluctuation of nodal displacement vector is expressed in the following linear
approximation form as the result of sensitivity analysis with respect to uncertain
variables for the structure governed by the stiffness equation (9.62):
Xn
U = U0 + U ai ai (9.65)
i=1
0
where U is the displacement under nominal loading after homology design, and
U ai is the sensitivity vector of U with respect to ai . Obviously the homologous
constraint in Eq. (9.54) cannot be satisfied under the uncertain fluctuation of load-
ing. The error in Eq. (9.54) after substitution of the fluctuating nodal displacement
vector expressed by Eq. (9.65) is defined by an error vector e, defined as follows,
from the exact homologous deformation:
e = CU d
n
!
X
0
=C U + U ai ai d
i=1
n
X
0
(9.66)
= (CU d) + CU ai ai
i=1
n
X
= CU ai ai
i=1
To evaluate the magnitude of the distortion from the objective homologous defor-
mation, we introduce the concept of error index, which can be defined by using the
components of the error vector. Here, we define the error index 2 as the square
of the Euclidean norm of the error vector:
2 = e> e = a> Da (9.67)
where D is an nn symmetric matrix defined by Eq. (9.66) using the rate of change
of nodal displacements with respect to the uncertain variables.
The worst case of uncertain variables, which gives the maximum value of the
error index, is searched for within the convex hull of Eq. (9.64). As the error index
is expressed in quadratic form with respect to the uncertain variables and becomes
a convex function, the anti-optimization problem turns out to be a maximization
problem of a convex function over a convex domain, and the point of the worst case
is located on the boundary of the convex hull. The search is carried out by the
Lagrange multiplier method employing the following Lagrangian :
= 2 (q 2 a> W a 1) (9.68)
where 0 is the Lagrange multiplier (see Sec. 4.1 for details of the Lagrange
multiplier approach). The uncertain variables in the worst case are so determined
as to satisfy the stationary condition of derived as

= 2(D q 2 W )a = 0 (9.69a)
a

= 1 q 2 a> W a = 0 (9.69b)

The stationary condition (9.69a) results in the eigenvalue problem formed by the
eigenvalue q 2 , which is rewritten as = q 2 and the eigenvector a. On the other
hand, the stationary condition (9.69b) gives the normalization condition for the
eigenvector. Generally, n eigenpairs are obtained by solving the eigenvalue problem
(9.69a) under the normalization condition (9.69b). In view of Eqs. (9.69a) and
(9.69b), the error index of Eq. (9.67) becomes

2 = a> Da = q 2 a> W a = = (9.70)
q2
as the magnification coefficient q is constant and the maximum value of the error
index, i.e., the worst case, is estimated by the maximum eigenvalue obtained
by solving Eq. (9.69a). The eigenvector corresponding to the maximum eigenvalue
gives the uncertain variable vector for the worst case.
9.8.6 A numerical example of worst-case estimation of homology

design
The 11-bar truss obtained from the homology design in Sec. 9.8.3 is employed in
this numerical example to evaluate the worst case of the distortion from the ob-
jective homologous deformation caused by uncertain fluctuation of the uniformly
distributed vertical load. Four uncertain variables are assigned to vertical load
components for nodes 1, 2, 3, and 4 in Fig. 9.15. The fluctuation of the uncertain
variables is bounded by hypersphere with radius 1.0, which is expressed by setting q
equal to 1.0 and W as an identity matrix in Eq. (9.64). The nodal load components
fluctuate from 0 to twice the nominal values in this case.
The result of the worst-case estimation obtained by solving the eigenvalue
problem of Eq. (9.69a) with Eq. (9.69b) is shown in Table 9.6. The error in-
dices estimated by the eigenvalue, i.e., 2 = /q 2 , are listed in the leftmost
column and corresponding uncertain variables, i.e., the eigenvector components,
are listed in the right columns. As is seen, the maximum eigenvalue is dupli-
cate, and the error index is maximized by two cases of uncertain variables, namely,
(a1 , a2 , a3 , a4 ) = (0.707, 0.0, 0.0, 0.707), called worst case 1, and (a1 , a2 , a3 , a4 ) =
(0.5, 0.5, 0.5, 0.5), called worst case 2. The actual error indices evaluated by the
analyses employing the nodal force components corresponding to the worst cases 1
and 2 turn out to be equal to 2.82 103 mm2 in both cases, because the nodal
Table 9.6 Estimated error index by convex analysis.
Error index 2 (103 mm2 ) Uncertain variables
a1 a2 a3 a4
2.82 0.707 0.000 0.000 0.707

2.82 0.500 0.500 0.500 0.500
1.94 0.000 0.707 0.707 0.000
0.00 0.500 0.500 0.500 0.500
0.1 mm
Fig. 9.18 Deformed structure in worst case 1.
0.1 mm
Fig. 9.19 Deformed structure in worst case 2.
displacements are linear functions of the nodal loads. The deformations of the 11-
bar truss in the two worst cases are illustrated in Figs. 9.18 and 9.19, respectively,
in a similar manner as Fig. 9.16.
9.8.7 Concluding remarks

A formulation of homology design based on finite element analysis has been re-
viewed. The change of displacements by design variables is approximated by means
of first-order finite element sensitivity analysis and the governing equation for de-
sign variables is derived in the matrix form with a rectangular coefficient matrix.
The equation is solved with the MoorePenrose generalized inverse to obtain the
solution with least design change. The illustrative numerical example of the 11-bar
truss demonstrates that the uppermost deck remains straight and horizontal even
after the deformation caused by a uniformly distributed vertical load.
The worst case of homologous deformation caused by uncertain fluctuation of
loadings is estimated by means of convex analysis. The uncertain variables are
assigned to the discretized nodal force components, and the fluctuation of the un-
certain variables is bounded within a convex hull. The worst case maximizing the
error index of homologous deformation is searched for on the boundary of the convex
hull by the Lagrange multiplier method. The numerical example using the 11-bar
truss structure shows the validity of the method by demonstrating the identification
of the worst case caused by uncertain variables confined within a hypersphere.
9.9 Design of Flexible Structures under Constraints on Asymptotic

Stability
9.9.1 Introduction
The approaches to designing structures against external loads are classified into two
categories: stiff structure and flexible structure. In the stiff structure approach, the
structure is designed so as to have enough stiffness and strength against the static
and dynamic loads, and the deformation under such loads should be small enough.
In some situations, however, the structure may be flexible and large deformation
may be allowed so far as no failure or yielding occurs and the undeformed initial
state can be recovered after releasing the external loads.
Contrary to mechanical parts and structures, the loading condition for struc-
tures in the field of civil engineering is highly unpredictable. Therefore, it is not
practically acceptable to design such structures by considering only static loads and
proportional loading conditions with fixed load patterns. The magnitude of the
input energy by dynamic motions such as earthquake excitations and wind loads
can be characterized and bounded by the total energy applied to the structure.
Hence, the strain energy that can be stored may be used as the performance mea-
sure of elastic structures in the civil engineering field. This way, the stability of the
structure under dynamic seismic or wind loads with specified energy is assured (see
Sec. 6.6 for anti-optimization with energy bounds of seismic excitation). Note that
the energy dissipation capacity may be an important performance measure also for
the case where inelastic responses are allowed (Ohsaki, Kinoshita and Pan, 2007;
Pan, Ohsaki and Tagawa, 2007).
This section summarizes the optimizationanti-optimization approach developed
by Ohsaki (2003a) to designing elastic flexible structures. The structure is designed
allowing large deformation under the external loads with specified energy. A con-
straint is given for the storable strain energy under the specified norm of deforma-
tion. Furthermore, the structure should recover the undeformed state after releasing
the external loads. The criterion for asymptotic stability is used for formulating the
optimization problem. An illustrative example is shown using a 2-bar truss, and the
performances of the optimum design is verified by carrying out dynamic analysis of
a 24-bar dome-type truss.
H(U)
Ui
Fig. 9.20 A quasi-convex function H(U ) that has a stationary point at the origin U = 0; U i : a
representative displacement component.
9.9.2 Definition of asymptotic stability

Consider a finite-dimensional elastic structure. The vector of nodal displacements is
denoted by U = (U1 , . . . , Un )> , where n is the number of degrees of freedom. In the
following, a component of a vector is indicated by a subscript. The performance of
the structure is defined by the strain energy H(U ) that can be stored in the structure
within the specified norm of the nodal displacements. Lagrangian formulation is
used for the strain-displacement relation allowing large deformation. The nodal
force vector F (U ) that is equivalent to the deformation U is defined by
H
Fj (U ) = , (j = 1, . . . , n) (9.71)
Uj
In order to ensure that the undeformed state U = 0 is recovered after releasing
the external loads, the condition of asymptotic stability is to be satisfied at U = 0.
There are several definitions for asymptotic stability of an elastic system. From the
definition of Liapunovs direct method (Salle and Lefschetz, 1961), if the undeformed
state is the only stationary point of H(U ) in a bounded domain containing the
origin U = 0, then it is asymptotically stable. Hence, H(U ) should be a strictly
quasi-convex function (Boyd and Vandenberghe, 2004) as shown in Fig. 9.20 that
has only one stationary point U = 0 (see Ohsaki (2003a) for details).
Let G(U ) be defined as
G(U ) = U > H(U ) (9.72)
where H is the gradient of H(U ). By using Eq. (9.71), Eq. (9.72) is rewritten as
G(U ) = U > F (U ) (9.73)
A strictly quasi-convex function H(U ) in a bounded domain of U , which has
only one stationary point at U = 0, is characterized by
G(U ) > 0 for U and U 6= 0 (9.74)
9.9.3 Optimization problem

The strain energy that can be stored in the structure is used as a performance
measure. Moreover, the structure should have enough flexibility. Therefore, the
lower bound is given for the worst (minimum) value of the strain energy at the
deformed state defined by the specified norm of displacements, and the undeformed
state should be recovered as the external loads are removed from the deformed state.
Let A denote the vector of design variables such as the cross-sectional areas
of the members of a truss. The objective function to be minimized is denoted by
V (A). The norm of the displacement vector is defined by
p
D(U ) = U > U (9.75)
The specified value for D(U ) is denoted by D; i.e., the feasible domain for
U is given by D(U ) D. A tilde is used as H(Ue , A) to indicate that U is the
variable vector, and A is regarded as a parameter vector, whereas a hat is used
to indicate a function of A only. The worst (minimum) value H(A) b e , A)
of H(U
among the displacements satisfying D(U ) = D should not be less than the specified
lower bound H.
b
Hence, H(A) is defined as the anti-optimal objective values of the following
problem:
find b
H(A) e , A)
= min H(U (9.76a)
U
subject to D(U ) = D (9.76b)
The undeformed initial state should be recovered by releasing the external loads
at any state with U in . Therefore, the condition (9.74) should be satisfied. Since
a strict inequality condition cannot be given for an optimization problem, Eq. (9.73)
is replaced by
e , A) 0
G(U (9.77)
Moreover, it is difficult to satisfy the equality constraint (9.76b) during the optimiza-
tion process. Therefore, the equality constraint is relaxed to inequality D(U ) D.
However, if the minimum value of G(U e , A) is simply searched for in a region
bounded by D(U ) D, an obvious and meaningless solution U = 0 will be found.
Therefore, the objective function G(U e , A) of the anti-optimization problem is re-
e
placed by G(U , A)/D(U ) to prevent convergence to U = 0.
The minimum value of G(U e , A)/D(U ) at the anti-optimal solution is denoted
b
by G(A), which is a function of the design variable vector A. The anti-optimization
problem is formulated as
e , A)
G(U
find b
G(A) = min (9.78a)
U D(U )
subject to D(U ) D (9.78b)
U1
h
U2
w w
b
If G(A) 0, then D(U ) > 0 leads to G(U e , A) 0 throughout the region
except U = 0, and H(U ) is a quasi-convex function that has the maximum value
at D(U ) = D. Since the minimum value of H(U ) along the boundary D(U ) = D is
b
H(A), which is obtained by solving Problem (9.76), H(U ) is a quasi-convex function
in the region including the origin U = 0 bounded by H(U ) H(A). b Therefore,
the undeformed state is recovered from any state U satisfying H(U ) H(A).b
Finally, the upper-level optimum design problem is formulated as
minimize V (A) (9.79a)

b
subject to H(A) H (9.79b)
b
G(A) 0 (9.79c)
where H is the specified lower bound of the strain energy. Suppose that the objective
function V (A) is an increasing function of A, which is a natural assumption for the
b
case, e.g., V (A) is the total structural volume. If the strain energy H(A) is also an
increasing function of A, e.g., for the case where A is the vector of cross-sectional
area of a truss, the constraint (9.79b) is satisfied in equality at the optimal solution.
In the following numerical examples, optimization is carried out by using a
standard gradient-based nonlinear programming with line search, where the sensi-
tivity coefficients are found by using the finite difference approach. The lower-level
anti-optimization problems (9.76) and (9.78) are solved at each step of solving the
upper-level optimum design problem (9.79).

An example of a 2-bar truss is first shown to illustrate the validity of the problem
formulation, and for verification of the optimization algorithm. Optimum designs
are next found for a 24-bar shallow truss, and the results are verified by carrying
out dynamic analysis.
The Green strain is used for computing the strain energy. The objective func-
tion to be minimized in the upper-level problem is the total structural volume.
Optimization is carried out by the library DOT Ver. 5.0 (VR&D, 1999), where
sequential linear programming is used.
Table 9.7 Optimization results of the 2-bar truss.
D 100.0 200.0 300.0
V (A) 3.8133 105 1.4896 105 1.1770 105

A 182.63 71.338 56.366
H(U I ) 2.0060 107 2.0060 107 2.0060 107

D(U I ) 100.00 200.00 300.00
D(U II ) 99.700 200.00 300.00
b
G(A) 3.2163 105 5.0150 104 0.0
U1I 0.0 0.0 0.0

U2I 100.00 200.00 300.00
U1II 0.0 0.0 0.0

U2II 99.700 200.00 300.00
9.9.4.1 A 2-bar truss

Consider a 2-bar truss as shown in Fig. 9.21, where w = 1000 mm and h = 300 mm.
The specified strain energy is H = 2.0 107 N mm. In the following, the units of
force and length are N and mm, respectively, which are omitted for brevity.
Assuming the symmetry of the truss, the cross-sectional areas of the two mem-
bers have the same value; i.e., A1 = A2 = A, and the number of design variables is 1.
The optimization results for D = 100, 200 and 300 are as shown in Table 9.7. Note
b
that the constraint H(A) H is satisfied with equality for all the cases within the
specified tolerance. Let U I and U II denote the anti-optimal solutions of lower-level
problems (9.76) and (9.78), respectively. Displacement components U1 and U2 are
defined as shown in Fig. 9.21; i.e., U I = (U1I , U2I )> , U II = (U1II , U2II )> .
Since h < w, U1I = 0 should be obviously satisfied to minimize the strain energy
for the specified displacement norm. Consequently, |U2I | = D is obtained from the
constraint D(U ) = D, which agrees with the results in Table 9.7. The inequality
constraint D(U ) D of Problem (9.78) is satisfied with equality for all the cases.
The minimum value G(A) b of Problem (9.78) is positive for D = 100 and 200.
Therefore, the undeformed state is asymptotically stable in the region defined by
the upper bound D of the displacement norm. For D = 300 (= h), G(A) b = 0, i.e.,
e
G(U , A) = 0, is satisfied at (U1 , U2 ) = (0, 300), because the two bars are colinear
at (U1 , U2 ) = (0, 300), and a statical equilibrium state is attained without external
load. If U2 > 300, the equilibrium point (U1 , U2 ) = (0, 600) without external loads
will be reached as the external load is released. Therefore, D = 300 is the upper
bound of D for the undeformed state to be asymptotically stable.
9.9.4.2 A 24-bar truss

Consider next a 24-bar truss as shown in Fig. 9.22, where the nodal coordinates are
listed in Table 9.8. The numbers with and without the parentheses in Fig. 9.22 are
13
(23) (24)
9 (20) 12
10 (21) 11 (22)
(14) (15) (17) (18)

(19)
(16)
(13)
6
(12) 7
8 x
(6) (9) (11)
(7) (8) (10)
(3) 3 (4) 4 (5)
2 5
(1) (2)
1
z
x
Table 9.8 Nodal coordinates (mm) of the 24-bar

truss.
Node Number x y z
1 0.0 5000.0 0.0

2 4330.0 2500.0 0.0
3 1250.0 2165.0 621.6
4 1250.0 2165.0 621.6
5 4330.0 2500.0 0.0
6 2500.0 0.0 621.6
7 0.0 0.0 821.6
8 2500.0 0.0 621.6
9 4330.0 2500.0 0.0
10 1250.0 2165.0 621.6
11 1250.0 2165.0 621.6
12 4330.0 2500.0 0.0
13 0.0 5000.0 0.0
the member numbers and node numbers, respectively. The specified strain energy
is H = 1.0 106 . The members are divided into three groups 1, 2 and 3, each of
which has the members {8, 9, 12, 13, 16, 17}, {4, 7, 10, 15, 18, 21}, {1, 2, 3, 5, 6, 11,
14, 19, 20, 22, 23, 24}, respectively, where the cross-sectional areas of the members
in group i have the same value Agi . Therefore, the number of design variables is
three.
Optimal solutions are found for various values of D. The optimization results
Table 9.9 Optimization results of the 24-bar truss.
D 50.0 100.0 180.0
V (A) 3.5598 107 1.0905 107 5.2154 106
Ag1 795.19 161.62 110.25

Ag2 370.69 168.13 165.88
Ag3 476.60 156.93 281.74
H(U I ) 1.0000 106 1.0000 106 1.0000 106

D(U I ) 49.978 99.956 179.92
D(U II ) 50.040 45.069 179.84
b
G(A) 5.6848 104 2.4705 104 13.121
Fig. 9.23 Displacement U I that has the minimum value of strain energy for specified displacement
norm D = 180.0; solid line: deformed shape; dotted line: undeformed shape.
are listed in Table 9.9. For D = 200.0, a solution satisfying G(A) b 0 could not
be found; i.e., the undeformed state cannot be recovered from the worst state cor-
responding to D = 200.0. The constraints on strain energy and the displacement
norm are satisfied in equality for three cases with D = 50.0, 100.0 and 180.0. G(A) b
has the minimum value on the boundary D(U ) = D for D = 50 and 180. For
b
D = 100, however, G(A) has the minimum at the point inside the region. The total
structural volume decreases as D is increased, which agrees with the intuitive obser-
vation that the amount of material may be reduced by allowing large deformation.
The cross-sectional areas of the members in group 1 that connect to the center node
have relatively large values if D is small. On the contrary, the cross-sectional areas
of the members in group 3 that connect to the supports have large values if D is
large.
The displacement U I at the anti-optimal solution of Problem (9.76) for D =
180.0 is shown in solid lines in Fig. 9.23 in real scale, where the dotted lines indicate
the undeformed shape. Note that U II at the anti-optimal solution of Problem (9.78)
has a similar shape as U I . It is observed from Fig. 9.23 that the vertical displace-
ment of the center node is very large with deformation of the members in group 1
only in this worst-case deformation.
Asymptotic stability of the undeformed state may be verified by one of the
following two approaches:
12000000
1000000
Strain energy
800000
600000
400000
200000
0
0 1 2 3 4
Time
Fig. 9.24 Time histories of strain energy from various initial conditions.
120
100
Norm of displacement
80
60
40
20
0
0 1 2 3 4
Time
Fig. 9.25 Time histories of norm of displacement vector from various initial conditions.
(i) Randomly assign an initial static equilibrium state with specified strain energy,
and carry out transient response analysis.
(ii) Assign initial velocities at the nodes so that the initial kinetic energy is equivalent
to the specified strain energy, and carry out transient response analysis.
Assuming that the target application of the proposed method is seismic design
of a latticed dome as shown in Fig. 9.25, the second approach is used for verification
purposes; i.e., the specified strain energy represents the maximum energy supplied
by the seismic motion.
Dynamic analysis for specified initial velocity has been carried out by using
MSC.Nastran Ver. 70.5 (Nastran, 2001). The solution type of Nastran is SOL129,
which is a transient response analysis with geometrical nonlinearity. The time
increment for integration with quasi-Newton iteration is 0.02 sec. Since the dome
structures in civil engineering usually have large masses attached to the nodes, the
member mass is neglected in the following analysis.
Initial velocities are given randomly only in z-direction of the internal nodes. A
random number ri is generated for each node, and let
N
X
ri
ri = sum
, rsum = rj (9.80)
r j=1
where N is the number of internal nodes, which is equal to 7 for this example.
Let mi denote the mass at the ith internal node. Then the initial velocity vi is
calculated from
s
XN
2H

vi = r i , M= (mj rj ) (9.81)
M j=1
In the following, mi = 100.0 kg for all the internal nodes.

Damping has not been considered in the problem formulation, in order to in-
vestigate the maximum values of strain energy and displacement norm that can
be reached. Obviously, dynamic response does not converge to the undeformed
state if no damping is included. It is very difficult, however, to estimate the cor-
rect damping matrix of the latticed dome, and also it is important to investigate
the maximum response that can be reached without damping, because damping
of structures in civil engineering is usually very small, and the real response with
damping is bounded by the response without damping.
Suppose mi for all the nodes are simultaneously scaled by (> 1). Then M

becomes M , vi is replaced by vi / , and the rth natural circular frequency r of

the truss is also scaled to r / , without changing the maximum strain energy, if
the damping is negligibly small. Hence, the values of mi are not important provided
that the ratios among mi are fixed.
Figures 9.24 and 9.25 show the time histories of strain energy and the norm
of displacement vector, respectively, through the vibration process of the optimal
solution for D = 100.0. The specified strain energy 1.0 106 can be reached if the
mode of vibration defined by the distribution of the initial nodal velocities is close
to one of the eigenmodes of free vibration, and velocities of all the nodes vanish
simultaneously at the state corresponding to the maximum strain. However, the
situation stated above cannot be realized as a result of randomly generated initial
excitation; hence, it is verified from Fig. 9.24 that the maximum value of the strain
energy is a little below the specified value. The maximum value of the displacement
norm for each case is far less than the specified value 100.0 as observed in Fig. 9.25,
because the displacements increase rapidly with a small increment of strain energy
near the final deformed shape of a flexible structure, and the displacements are
more sensitive than the strain energy against the variation of initial velocity.
9.10 Force identification of prestressed structures
9.10.1 Introduction
Assessment and maintenance of member forces are of great importance for pre-
stressed structures such as cable nets and tensegrity structures, which take advan-
tage of prestress to enhance their stiffness. Therefore, the distribution of forces
should be precisely identified and carefully adjusted in the construction process.
The member forces should also be monitored and maintained after construction
for the process of life-cycle management (see Sec. 4.6 for anti-optimization in force
design of tensegrity structures).
In this section, we present the methodology for the force identification of pre-
stressed pin-jointed structures based on measurements of the member forces and
nodal locations (Zhang, Ohsaki and Araki, 2004). An optimization problem is for-
mulated for determination of the optimal placement of the measurement devices
for the member forces, and a heuristic approach is presented for this combinatorial
optimization problem.
Because of the high ratio of stiffness to structural weight, prestressed structures
are considered to be the ideal structural form for lightweight structures, and they
attract much attention in theoretical studies as well as practical applications. We
mainly consider the prestressed structures, of which members are pin-jointed. Many
novel structural forms, such as cable nets, tensegrity structures, and tensile mem-
brane structures modeled by cable nets, fall into this category (Kanno and Ohsaki,
2005; Ohsaki and Zhang, 2006).
Since the prestressed pin-jointed structures are usually stabilized by the pre-
stresses in the state of self-equilibrium, their configurations and member forces are
highly interdependent (Kanno, Ohsaki and Ito, 2002; Vassart and Motro, 1999;
Zhang and Ohsaki, 2006). Furthermore, the distribution of member forces is of
great importance to the stiffness of the structures, because it has significant influ-
ence on the geometrical stiffness. In the design problem of this kind of structure,
member forces are usually properly assigned to satisfy the self-equilibrium equa-
tions, and more importantly, to have positive effect on its stiffness so as to have
higher capability of resisting external loads.
However, it is very difficult to achieve the member forces as expected in practical
applications, mainly due to (a) fabrication errors, (b) installation errors, (c) short
and long term creep of materials, and (d) some other unexpected environmental
influences such as temperature change, deformation of base and supporting struc-
tures, etc. Their capability to resist external loads may be significantly reduced or
even lost due to unexpected change of member forces. Moreover, some structures
make use of the interdependency between configurations and member forces to ac-
tively modify their configurations in order to satisfy certain requirements; see for
example the work by Shea, Fest and Smith (2002). Hence, it is extremely important
to accurately identify the current distribution of member forces.
Member forces can be obtained with high accuracy by measurement devices.

There have been many studies on the optimal placement of measurement devices for
conventional structures; see for example Abdullah, Richardson and Hanif (2001),
Pothisiri and Hjelmstad (2002), Sanayei, Onipede and Babu (1992), and Zavoni,
Iturrizaga and Esteva (1999). However, it is not practically acceptable to measure
the forces of all members due to the limitation in cost and availability of measure-
ment devices. Therefore, in this section, we present a method for optimal placement
of the limited number of measurement devices to minimize the errors in force iden-
tification.
9.10.2 Equations for self-equilibrium state

The equations for the self-equilibrium state are briefly summarized below for a
prestressed pin-jointed structure (see Sec. 4.6.2 for the details of derivations of the
equations).
It is assumed that the self-weight of the members are neglected, and the external
loads are applied only at the nodes. Therefore, the members of a prestressed pin-
jointed structure transmit only axial forces. Moreover, the cable members transmit
only tensile forces, whereas the struts usually transmit only compressive forces.
Pin-jointed structures consist of two types of nodes: fixed nodes (supports)
and free nodes (internal unconstrained nodes). Suppose that a structure has m
members, n free nodes and nf fixed nodes. Therefore, a structure in d-dimensional
space (d = 2 or 3) has dn degrees of freedom. In the following, we assume a
structure in three-dimensional space. The m (n + nf ) matrix C s that has 0, 1 or
1 components represents the connectivity of the nodes and members defined by
Eq. (4.113) in Sec. 4.6.2 (Kaveh, 2004).
For convenience, the fixed nodes are preceded by the free nodes in the numbering
sequence, so that the m (n + nf ) connectivity matrix C s can be partitioned into
two parts as
C s = (C, C f ) (9.82)
f f
where the m n matrix C and the m n matrix C are the connectivity matrices
corresponding to the free and fixed nodes, respectively.
Let x, y, z and xf , y f , z f denote the nodal coordinate vectors of free and fixed
nodes in x-, y- and z-directions, respectively. The coordinate difference vectors of
members hx , hy and hz in x-, y- and z-directions, respectively, are calculated by
hx = Cx + C f xf
hy = Cy + C f y f (9.83)
z f f
h = Cz + C z
Then the diagonal coordinate difference matrices H x , H y and H z in x-, y- and
z-directions, respectively, are defined as
H x = diag(hx ), H y = diag(hy ), H z = diag(hz ) (9.84)
In the following, the components of vectors and matrices are indicated by subscripts,
e.g., as hx = (hxi ), H x = (Hij
x
). The length matrix L = (Lij ) is diagonal and given
as
q
Lii = (Hiix )2 + (Hiiy )2 + (Hiiz )2 , (i = 1, . . . , m) (9.85)
Let s denote the vector of member forces. In the state of self-equilibrium, the
equilibrium equation of a pin-jointed structure is written as (Schek, 1974)
Ds = 0 (9.86)
where the 3n m matrix D is called equilibrium matrix, which is defined by
> x
C H
D = C > H y L1 (9.87)
C>H z
The self-equilibrium equation (9.86) will be extensively used to identify the force
distribution of the structure in the state of self-equilibrium, based on measurement
data of the nodal coordinates and the forces of some members. It is easily observed
from Eq. (9.86) that the measurement errors in nodal coordinates in D and some
components of member forces in s have significant influence on the identification
accuracy of the member forces that are not measured, because the self-equilibrium
equation has to be exactly satisfied.
Suppose that the structure consists of h independent modes of member forces
that satisfy the self-equilibrium equation (9.86), where
h = m rank(D) (9.88)
For prestressed structures, h 1 always holds so that there exist forces in the
self-equilibrium state. As will be discussed in the following sections, at least
hindependent components of member forces need to be measured so as to determine
the forces of the remaining members based on the self-equilibrium equation.
9.10.3 Formulation of identification error

9.10.3.1 Force errors
Let s = (s1 , . . . , sp )> denote the vector of the forces of p members, which are to be
measured, while the vector of the forces of the remaining members, which are to be
estimated, is denoted by e s = (es1 , . . . , semp )> . Equation (9.86) can be rewritten as
follows by classifying the member forces into the measured set s and the estimated
set e
s:
es = 0
D s + De (9.89)
where the 3n p matrix D and the 3n (m p) matrix De are constructed
by assembling the columns in D that correspond to the members in s and e
s,
respectively.
e is full-rank, i.e., rank(D)

If D e = m p, then the solution of e
s with least square
norm can be determined as
e e D s
s = D (9.90)
where ( ) denotes the MoorePenrose generalized inverse matrix. When D e has
e e
rank deficiency, i.e., rank(D) < m p, there exist m p rank(D) independent
modes of e s that satisfy Eq. (9.89), and Eq. (9.90) gives the solution that has the
minimum norm among the possible solutions. Hence, D e should be full-rank to have
accurate estimation of the member forces. Furthermore, h or more members should
be measured to exclude the h dependent columns in D from D e so that De can be
full-rank.
The nodal coordinates are combined as X = (x> , y > , z > )> . Suppose that there
exist measurement errors in X and the member forces s . Total differentiation of
Eq. (9.90) with respect to X and s leads to
3n
X X3n e
D D e s=0
Xi s + D s + Xi e
s + De (9.91)
i=1
Xi i=1
Xi
where Xi and s , respectively, are the measurement errors of nodal coordinates
and member forces to be measured, and e s is the resulting estimation error of
the forces of the remaining members. D /Xi and D/Xe i can be derived by
assembling the corresponding components in D/Xi ; see details in Sec. 9.10.4.
From Eqs. (9.90) and (9.91), es is written as
" 3n ! #
X D
e
D
es = (D)e
e D s Xi + D s
(D)
(9.92)
i=1
Xi Xi
which is simplified to

x
e
s = (H, G) (9.93)
s
where
H = (H 1 , . . . ,H i , . . . ,H 3n )
!
D e
D
e
H i = D e D e
D s (9.94)
Xi Xi
e D
G = (D)
By combining the measurement errors s and the estimation errors e s of mem-
s> )> , we have the following equation from Eq. (9.93):
ber forces as s = (s> , e

s O Ip X
s = = (9.95)
e s H G s
where I p is the p p identity matrix.
This way, the force error vector s of all members is formulated as a function
of measurement errors of nodal coordinates X and member forces s .
9.10.3.2 Identification error

Let ec and ef denote the parameters for defining the bounds of the measurement
errors X and s , respectively. The measurement errors of member forces and
nodal coordinates with different dimensions are transformed into dimensionless vari-
ables by dividing them by ec and ef , respectively. Hence, Eq. (9.95) is rewritten
as

O ef I p X/ec
s = Br = (9.96)
ec H ef G s /ef
where

O ef I p X/ec
B= , r = (9.97)
ec H e f G s /ef
When the sets of members, for which the forces are to be measured and to be
estimated, are specified, B is a constant matrix from the definition of H and G.
Hence, s is conceived as a function of r. To incorporate the worst-case scenario,
the worst
p value of r with unit norm is found to lead to the maximum Euclidean
norm (s)> s. Hence, the performance measure of the identification is defined
as the identification error E, which can be obtained by solving the following anti-
optimization problem:
q
find E = max (Br)> Br (9.98a)
r
q
subject to (r)> r = 1 (9.98b)
Note that E is equivalent to the 2-norm kBk2 of the matrix B (Horn and John-
son, 1990), which is equal to the square root of the largest eigenvalue max of the
symmetric matrix B > B; i.e.,
p
E = max (9.99)
Note that max is also defined as the maximum singular value of B.
When the forces of all members are to be measured, i.e., p = m, and the errors
in nodal coordinates are not considered, we have E = ef , because

> O O
B = (O, ef I m ), B B = (9.100)
O e2f I m
where I m is the m m identity matrix.
9.10.4 Sensitivity analysis with respect to nodal coordinates

Let xi denote the x-coordinate of free node i. The sensitivity coefficient of the
coordinate difference vector hx in x-direction with respect to xi is given as
hx (Cx) (C f xf )
= + (9.101)
xi xi xi
Since C and C f are constant, and the measurement error of coordinates of fixed
nodes is not considered, Eq. (9.101) becomes
hx
= Ci (9.102)
xi
where C i is the ith column of C.
From H x = diag(hx ), we have
H x
= diag(C i ) (9.103)
xi
Since H y and H z do not depend on the x-coordinates, their sensitivity coefficients
with respect to xi vanish as
H y H z
= 0, =0 (9.104)
xi xi
x y z
The sensitivity coefficients of H , H and H with respect to yi and zi can be
formulated similarly to Eq. (9.103) and Eq. (9.104) as

H x H x

=0
=0
yi

y
ziy

H H
= diag(C j ) , =0 (9.105)

y i
zi

z
z

H
H
=0 = diag(C k )
yi zi
From Eq. (9.85), we have
L2 = (H x )2 + (H y )2 + (H z )2 (9.106)
Let t represent a variable xi , yi , or zi . The sensitivity coefficient of L with respect
to t is derived as x y z
L 1 x H y H z H
=L H +H +H (9.107)
t t t t
Since the sensitivity coefficient of L1 with respect to t is
L1 L
= L2 (9.108)
t t
the sensitivity coefficient of L1 with respect to t is written as follows by substi-
tuting Eq. (9.107) into Eq. (9.108):

L1 H x H y H z
= L3 H x + Hy + Hz (9.109)
t t t t
For the structure with given topology, the connectivity matrix C is a constant
matrix. From the definition of the equilibrium matrix D in Eq. (9.87), its sensitivity
coefficient with respect to a variable t is derived as
x 1

> H 1 > x L
C t L + C H t

D 1
> H y 1 L
= C L + C>H y (9.110)
t
t t
H z 1 L1
C> L + C>H z
t t
Substituting H x /t, H y /t and H z /t in Eqs. (9.104) and (9.105) into

Eq. (9.109) and then to Eq. (9.110), we have the sensitivity coefficient D/t of
the equilibrium matrix D with respect to a nodal coordinate t.
9.10.5 Optimal placement of measurement devices

9.10.5.1 Problem formulation
In the process of force identification of a prestressed pin-jointed structure, the key
factors to be considered are: (a) accuracy of measurement, (b) number of measure-
ment devices, and (c) locations of measurement devices. We consider only the last
two factors, since the accuracy of the measurement depends on the property of the
measurement device.
In practical applications of force identification, the nodal coordinates can be eas-
ily measured with relatively small cost, while the measurement devices for member
forces are much more expensive. Hence, we assume that all nodal coordinates are
measured, whereas only some of the member forces are measured.
From the definition of identification error in Eq. (9.99), it is seen that the matrix
B depends on the set of measurement members to be selected even for the case
where the number of devices is kept constant. Therefore, our problem is to find the
optimal locations of measurement devices for member forces, that lead to minimum
identification error E with the fixed number p of devices.
Let J = {J1 , . . . , Jp } denote the set of p members, for which the forces are to
be measured. The variable J in the optimization problem is a parameter for the
anti-optimization problem (9.98) that is reformulated as
q
find b ) = max (B(J )r)> B(J )r
E(J (9.111a)
r
q
subject to (r)> r = 1 (9.111b)
b ) is the worst identification error corresponding to the specified set J .
where E(J
Then the optimum location problem for the measurement devices is formulated
as
minimize E(J b ) (9.112a)
subject to |J | = p (9.112b)
where |J | is the number of elements in J . Note that there is actually another
e has to be full-rank so as to have the least squares
inherent constraint: the matrix D
solution for the member forces to be estimated as in Eq. (9.90).
9.10.5.2 Solution process

Since the proposed problem (9.112) is a typical combinatorial optimization problem,
a heuristic approach called Simulated Annealing (SA) and the stingy method are
adopted (see Sec. 2.6.3 for the details of heuristic approaches).
In SA, the initial solution and the cooling schedule play critical roles in find-
ing the optimal solution with good accuracy. Hence, it is more reliable to start
the SA from a better initial solution. Zhang, Ohsaki and Araki (2004) developed
several simple heuristic approaches, including the stingy method, to investigate the
identification accuracy of force distribution of prestressed pin-jointed structures,
where only the measurement errors of member forces are considered. It has been
demonstrated that the stingy method has relatively good accuracy with small com-
putational cost. Therefore, the stingy method can be effectively used for producing
a good initial solution, rather than random solutions, to reduce the computational
cost as well as to improve the accuracy in SA.
The stingy method is a basic heuristic approach, based on the local search,
to combinatorial optimization problems. For the problem considered here, it starts
from a complete set of the possible measurement members, and successively removes
the member with least contribution to the objective function from the current set
of solutions, under the condition that the removal does not lead to an infeasible
e
solution violating the requirement on rank deficiency of D.
The stingy method used in this example is summarized as follows:
Step 0: Set J = {1, . . . , m}, i.e., the forces of all members are to be measured,
and p = m.
Step 1 Let Ep0 and Epj denote the identification errors by the sets I and I {j},
respectively. Find the member k in I that minimizes Epj Ep0 .
Step 2 If |J | = p, then terminate the process; otherwise, set I I {k} and
|J | |J | 1 accordingly, because it has the minimum contribution in I to
reduction of the identification errors and go to Step 1.
This way, we can find the p measurement members with the approximately
minimal identification error for the optimization problem (9.112), which is used as
the initial solution for SA.

Figure 9.26 shows a three-dimensional cable dome, consisting of 12 struts and 48
cables, which are shown in thick and thin lines, respectively (Ohsaki and Kanno,
2003). The nodes and supports of the structure are located on three circles with
different radii, 5.0 m, 10.0 m, and 15.0 m, and the four different elevations of the
circles are 1.0 m, 0.0 m, 1.0 m, and 2.0 m. Computation is carried out using
MATLAB (Borse, 1997).
The variance of measurement error of the nodal coordinate is set as ec = 1.0 mm,
and that of the member force is ef = 0.5 kN, which is about 1.7% of the largest
member force of the cables. In practical applications, it might be more convenient
to measure the forces in the cables rather than those in the struts, because the
measurement devices can be embedded into the cables more easily. Suppose that
10
10
15 10 5 0 5 10 15
(a) top view (b) diagonal view
2
1
0
1
15 10 5 0 5 10 15
(c) side view
Fig. 9.26 A three-dimensional cable dome.
(a) stingy method only (b) SA with stingy method

Fig. 9.27 Optimal locations of measurement devices for cables (dotted line) of the three-
dimensional cable dome found by the stingy method only, and the SA with stingy method for
generating the initial solution.
we have only 12 measurement devices, and the objective of this example is to find
12 cables, i.e., p = 12, to be measured so as to minimize the identification error.
The dotted lines in Fig. 9.27(a) show the cables to be measured, which are
determined by the stingy method only. The identification error in this case is
1.1988. As is seen, the members are not symmetrically located. We next use this
result as the initial solution for SA, and obtain the optimal locations of cables to be
measured as shown in the dotted lines in Fig. 9.27(b). The identification error in
this case is reduced to 1.0418, which is smaller than that of using the stingy method
only.
For additional references on hybrid optimization and anti-optimization, the reader

may consult, for example, the works by Adali, Richter and Verijenko (1995a), Adali,
Verijenko and Richter (1994b), Banichuk and Ivanova (2009), Chen and Wu (2004b),
Chen, Wu and Chen (2004), Cheng, Au, Tham and Zeng (2004), Cherkaev and
Cherkaev (1999, 2003), Cherkaev and Kucuk (1999), Dube and Hausen (2004),
de Faria (2000, 2002, 2004, 2007), de Faria and de Almeida (2003a, 2003b, 2004,
2006), de Faria and Hansen (2001a, 2001b), Fatemi, Mooij, Gurav, Andreykiv and
van Keulen (2005), Ganzerli and de Palma (2007), Ganzerli and Pantelides (2000),
Ganzerli, Pantelides and Reaveley (2000), Gu, Renaud, Batill, Brach and Budhiraja
(2000), Gurav (2005), Gurav, Goosen and van Keulen (2005a), Gurav, Kasyap,
Sheplak, Cattafesta and Haftka (2004), Gurav, Langhaar, Goosen and van Keulen
(2005b), Gurav and van Keulen (2005), Kocvara and Stingl (2006), Lombardi (1995,
1998), Lombardi and Haftka (1998), Lombardi, Mariani and Venini (1998), Natke
and Soong (1993), Pantelides and Booth (2000), Pantelides and Ganzerli (1997),
Pownuk (2000, 2004), van Keulen, Damveld, Elishakoff and Toporov (2001), Venini
(1998), Venter and Haftka (1996, 1999), and Zhang and Zhang (2000).

Chapter 10
Concluding Remarks
Would you tell me, please, which way I ought to go from here?
That depends a good deal on where you want to get to, said the
cat. (Lewis Carroll, Alice in Wonderland, Chapter 8)
Youve got to be very careful if you dont know where you are going,
because you might not get there. (Yogi Berra)
Optimal design became a rich and rewarding field of research. (Niord-
son, 2001)
In this chapter we try to summarize this book and some thoughts on the opti-
mization and probabilistic methods. It appears to us that total acceptance of the
ever-present uncertainty may infuse the old subject of optimization with the needed
new blood. The hybrid approach may help release both concepts from their con-
finement in the esoteric ivory tower and use them in the everyday quest for the
best and safest designs. To encourage further thought, the sections hereinafter
will be presented as questions in an attempt to provide replies in the form of first
approximations.
10.1 Why Were Practical Engineers Reluctant to Adopt Structural

Optimization?
Ever since structural optimization came into being, researchers have been expressing
doubts as to its justification and/or practicality. The background can be described
as follows:
(1) Why should we look for the best design if an acceptable, hence, a good one, can
be found? In short, adherence to the American maxim If it aint broke, dont
fix it was recommended.
(2) In his review article Ashley (1982) describes a poll conducted among 46 people
about the merits and demerits of structural optimization. The responses were
327
as follows:
an engineer experienced in civil and aeronautical structures: One of the
reasons that I stopped work in optimization was my dismay... that there
were so very few applications.
a Dean of Engineering who has known the field for over a quarter of a
century: I do not recollect any applications.
a foremost specialist on synthesis with aeroelastic constraints: I am sorry,
but I dont really have any...
a recently retired senior design engineer, describing events at his aerospace
company:
For 15 years I beat my head against a stone wall... The end was:
formal optimization techniques were never used in aircraft design
(even to this day!). The company was forced to use [them] in its
subsequent ICBM and space programs.
an interview with a distinguished European aerodynamicist regarding wing
design:
Numerical optimization techniques were being used to some ex-
tent in prefeasibility studies, but [he was] unable to produce any
specific or referenceable material.
Did the situation change in the past quarter century? Venkataraman and Haftka
(2004) address this topic. They write:
Yet, in 2004 structural optimization practitioners hear the same re-
frain from structural analysts that was heard in 1971; a single struc-
tural analysis takes several hours or even several days to run on my
computer; how can I even afford to think about optimization?
However,
... computing speed and storage capacity increased one million fold
at a rate of about 100-fold increase every decade. In 1971, when
structural optimization was still struggling to deal with any real-
life problems, a one-million factor increase in computational speed
would have seemed to solve all problems. After all, engineers were al-
ready analyzing complex structures with detailed finite element (FE)
models, and the ability to do one million analyses is sufficient for
most optimization problems.
Others concluded that the paradox of the continuing difficulty of applying
optimization to real-life problems is, of course, due to increasing complexity of
the structural models that analysts consider adequate for predicting structural
response.
Concluding Remarks 329
Venkataraman and Haftka ascribe this situation to the computerized

Parkinsons law. The classical law (Parkinson, 1959) states that the work done
expands to fill up all the available time. Its computerized counterpart (Thim-
bleby, 1993) manifests itself, in the words of Venkataraman and Haftka, by
software applications growing to fill up increased computer memory, processing
capacities and storage space. They further state:
In terms of structural analysis, it appears that the computer time re-

quired for an adequate structural analysis has been fixed at several
hours. Essentially, if a single analysis cannot be computed overnight
on the available computer, progress in terms of debugging models and
improving structures can become intolerably slow. So a requirement
that a working model can be analyzed in a few hours appears to be
the main restraint to the desire of the structural analyst to have a
high-fidelity structural analysis. In their paper on optimal design of
a sandwich composite fuselage section, Ley and Peltier (2001) discuss
how they had to limit the analysis model so as to be able to finish a
local optimization overnight.
Venkataraman and Haftka also stress the progress made by structural optimiza-
tion:
In spite of increased appetite of structural analysis for more complex

models and analyses, there is no question that structural optimiza-
tion did benefit from faster computers and more efficient analysis and
optimization algorithms. The turning point was in the mid-1980s. In
1981, Holt Ashley still lamented that optimization was used by re-
searchers but not by industry... Since then [1985], the increasing
number of optimization papers reflecting industrial applications at-
tests to the acceptance of structural optimization as a design tool...
one benefit of having faster computers is that engineering computa-
tions and challenging optimization are now routinely performed on
much cheaper desktop computers.
As Vanderplaats (1993) notes,
Structural optimization technology has now passed its thirtieth birth-

day. This technology has matured to the point that these methods
have been added to many commercial finite element codes... it is
clear that structural optimization technology may now be efficiently
applied to a wide range of design tasks... As advanced software be-
comes more widely available, it will be used to dramatically improve
design efficiency, as well as to aid in the allocation of scarce resources.
In an earlier paper, Vanderplaats (1991) writes:

Structural optimization technology has now matured to the point

that these methods have been added to most commercial finite el-
ement codes. Various methods are employed, ranging from simple
coupling between the analysis and optimization to reasonably sophis-
ticated use of approximation techniques.
The above paper concludes in an optimistic vein: ... it is clear that research
and development will continue to expand and to continue the improve[ment of]
the efficiency and reliability of the overall structural synthesis process.
(3) There has been a lot of controversy since the 1960s about the effectiveness of
optimizing structures against buckling. Thompson and Hunt (1974) come with
apparently the most powerful argument
the thesis that a process of optimization leads almost inevitably to
designs which exhibit the notorious failure characteristics often as-
sociated with the buckling of thin elastic shells. Here the idealized
perfect structure exhibits unstable and often compound branching
point and would fail by an explosive instability, while nominally per-
fect real structures containing inevitable initial imperfections fail at
scattered loads which can be quite considerably lower than that of
the idealization.
They also emphasize that
an optimum design is by its very nature imperfection-sensitive, since
any deviation from the idealized optimum design must yield a lower
failure load.
and conclude as follows:
Perhaps we can close with the question expressed by Pugsley (1966).
Should the structures be designed to be just stable, or in some way
more stable than stable?
We arrive at the following conclusion: optimization may be a generator of
severe imperfection sensitivity for some structures like thin-walled shells, but
need not necessarily lead to an imperfection-sensitive structure as we have seen
in Secs. 5.5 and 5.7. Optimization against buckling is very effective if instability
phenomena are fully understood, as was clearly demonstrated by Ohsaki and
Ikeda (2007a).
(4) It appears to some investigators, and we tend to agree with them, that opti-
mization theory, in order to become practical, ought to embrace the notion of
uncertainty. Frangopol and Maute (2005) stress:
[A] major limitation of the deterministic optimization approach is
that uncertainties in the loads, resistances, and structural responses
are not included in the computation process. Therefore, irrespective
of the level of sophistification in the approach used in the compu-

tational process, the optimum solutions are predicated on idealized
deterministic assumptions. Consequently, the solutions are not com-
puted under realistic conditions.
Indeed, Thoft-Christensen (2006) writes:
Classical deterministic structural optimization has only in few cases

been seriously used in designing real engineering structures.
Pu, Das and Faulkner (1997) write:
Conventional optimization... is a deterministic process. It has been

observed that the optimum structures obtained through deterministic
optimization do not necessarily have high reliability. Hence it is likely
that the failure probability of the optimum structure does not satisfy
the reliability criteria. The optimization based on reliability concepts
will lead to more consistent safety in the structure system.
Frangopol and Maute (2005) stress:
... despite the demonstrated exponential growth in structural opti-

mization research, most of it has been cast in deterministic format.
In fact, a quick survey of about 2400 references (Burns, 2002) in
structural optimization shows that more than 95% are deterministic-
oriented.
Frangopol (1995) is a strong proponent of stochastic optimization, stating that

stochastical design based on both reliability and optimization represents the
ultimate goal in design under uncertainty. This gives rise to the following
question: If probabilistic (or stochastic) optimization is the remedy for the ills
of structural optimization, why dont theoreticians and practitioners alike jump
on the wagon? An even more cardinal question is discussed in the next section.
10.2 Why Didnt Practical Engineers Totally Embrace

Probabilistic Methods?
Reservations regarding the various uncertainty analyses are not unlike their counter-
parts regarding the optimization approaches. Currently, the most utilized methods
for incorporating uncertainty are the probabilistic and fuzzy-sets based methodolo-
gies.
The book by Ben-Haim and Elishakoff (1990) devotes its first chapter to the
pros and cons of probabilistic modeling, including numerous pertinent quotations
in Sec.1.3; these will not be repeated here. Elishakoff (2000b) published a result of
the poll that he conducted for 12 years among more than 40 engineering researchers
extremely active in probabilistic analysis of structures. The possible limitations of

the probabilistic methods are listed and discussed in some detail.
It appears that probabilistic design of structures, based upon its fundamental
notion of reliability, finds itself in a situation reminiscent of the state of structural
optimization. Complaints in this context are nearly the same.
In his Freudenthal Lecture at the ICOSSAR93 conference, Yang (1994) stated
in his concluding remarks: To date, the reliability method has not been widely
used as an analysis or design tool in industry.
Schueller (1996) emphasizes:
There appears to be no much disagreement, that, despite the

significant progress made so far, particularly over the last 30 years,
a large number of procedures are still in the stage of development,
i.e., have not yet passed academic stage.
Soong (1988) notes: The gap between research and practice, it appears, is
widening... Sexmith and Nelson (1969) state: In spite of many notable contri-
butions... probabilistic concepts have not found their way into practice or design
codes in anything but a superficial way.
Cohn (1994) emphasizes:
Along with the research optimism expressed by these publications,

some reservations are raised on the philosophical, theoretical and
practical implications of reliability-based design and optimization.
Templeman (1988) notes:
... truly safe structures can only be assured by design and fab-
rication procedures which thoroughly examine and accommodate
the real causes of structural failures. The probability of failure
approach forms only a small part of these...
Cohns (1994) verdict on stochastic optimization is quite peremptory: ...

Large-scale, practical use of reliability-based optimization is unlikely, and quite
unjustified, at least in the foreseeable future. By contrast, it should be noted that
reliability-based optimization is being extensively studied in the 21st century, and
some industrial applications can be found (Kim and Choi, 2008; Cheng, Xu and
Jiang, 2006; Kharmanda, Olhoff and El-Hami, 2004). It also should be mentioned
that Working Groups on Reliability and Optimization of Structural Systems belong-
ing to the International Federation for Information Processing (IFIP) conducts reg-
ular meetings since 1987. There are at least 12 volumes edited by Thoft-Christensen
(1987, 1988), Der Kiureghian and Thoft-Christensen (1990), Rackwitz and Thoft-
Christensen (1991), Thoft-Christensen and Ishikawa (1993), Rackwitz, Augusti and
Bori (1995), Frangopol, Corotis and Rackwitz (1996), Nowak and Szerszen (1998,
2000), Furuta, Dogaki and Sakano (2002), Maes and Huyse (2003), and Srensen
and Frangopol (2006). these volumes, except the latter one, are conveniently titles
Reliability and Optimization of Structural Systems.
10.3 Why Dont the Probabilistic Methods Find Appreciation

among Theoreticians and Practitioners Alike?
Probabilistic modeling involves design with allowance for the aspect of reliability,
defined as the probability of successful performance of the structure. This means
that the engineer envisages some measure of the probability of failure, which is unity
minus the reliability. The main proponent and developer of probabilistic methods,
Alfred Freudenthal (1961) states: ... the notion that a finite (no matter how small)
probability of failure or at least of unserviceability is repulsive to a majority of
engineers. The persistent question arises: Has the situation changed since this
observation was made over half a century ago?
Vladimir Vasilievich Bolotin (see Elishakoff (2000b)) responds in the following
manner in the poll:
Engineers, especially those at the top, do not trust statements and

predictions of probabilistic character. They must make decisions
on important projects, large structures and big investments, and
they prefer to be completely sure that their decisions are true.
Of course, a proper education in the theory of probability and
mathematical statistics helps here, but there is a more profound,
maybe subconscious, case of such an attitude.
Grandori (1991) writes: ... the concept of structural safety will not leave the
realm of metaphysics unless we devise a method for justifying the choice of risk
acceptability levels.
Kalman (1994) is even harsher:
I see enormous activity, seemingly aimless, I see fanatical devotion

to ideas and principles which have grown into a quasi-religion...
Probability is an intellectual construct. It does not exist in the real
world... It is not of scientific concern today because it exists only
in the self-interest of gamblers or the imagination of statisticians
or the mind-reading of philosophers. There was and is no such
thing as a probabilistic revolution in science...
Schueller (1996) responds with the following statement, expressing guarded op-
timism:
... it should be stated, that the consideration of uncertainties, in-

herent in the problems of the field of engineering mechanics, by
stochastic methods certainly has not revolutionized our profes-

sion, but there are very good reasons to expect that the change of
thinking will come about in a long-time evolution.
Likewise, Rackwitz (2001) asks:
... Cornell (1981) called structural reliability a healthy adoles-

cent. In 1990 structural reliability then certainly was an efficient
executive in all his virulence. But has this man in the meantime
just got more years, but no more wisdom?
It appears that this controversy, sometimes very hot, will continue in the fu-
ture. It is more pragmatic to pose a more practical question on how to model the
unavoidable uncertainty in engineering practice. Bolotin (see Elishakoff (2000b))
provides a useful insight:
In my own experience, it is expedient to show to an engineer a set

of samples of the system behavior, in particular the worst sample,
the best sample and an average or a typical one... In general, in my
opinion, a representative set of time histories obtained with nu-
merical simulation of the probabilistic model presents to engineers
more information than the final results of statistical treatment.
It is this very methodology of finding the most favorable (best) responses along
with the least favorable (worst) ones that is adopted in this monograph. Bolotins
idea is overlaid with optimization study. Indeed, it may be overly conservative to
use the worst possible response without optimization: in fact, Kharitonov (1997)
tellingly titles his paper Interval uncertainty structure: conservative but simple.
Optimization minimizes overdesign, thereby becoming a natural ally in looking for
the worst-case scenario, which was dubbed here and elsewhere as anti-optimization.
Therefore, this book deals with optimization (Chap. 2), anti-optimization (Chaps.
38) and their hybrid (Chap. 9).
10.4 Is the Suggested Methodology a New One?
Usually, researchers tend to characterize their approaches as new, or novel,

or even radically new. However, in full agreement with the wisest of men,
King Solomon, we feel that There is nothing new under the sun, and the anti-
optimization approach is not new. Ovid (43B.C.E.18C.E.) writes in his Metamor-
phoses: I see and approve better things, but follow worse. (VII, line 20)
Rustem and Howe (2002) lead into their book with a quote from William Shake-
speare (15641616): ... since the affairs of men rest still uncertain, Lets reason
with the worst that may befall. (Julius Caesar, Act 5, Scene 1)
Sniedovich (2008a) and Burgman (2008) include Shakespeares name in the titles
of their respective papers. Burgman (2008) notes:
We can again claim Shakespeares credentials as a decision theo-

rist. He tells us that it is better to be safe than sorry, that The
better part of valour is discretion. (Henry IV, Part 1, Act 5,
Scene 4)
Leonhard Euler (17071783) stressed that For since the fabric of the universe
is most perfect and the work of a most wise Creator, nothing at all takes place
in the universe in which some rule of maximum or minimum does not appear.
(Tikhomirov, 1990).
In theoretical and applied mechanics, the idea of minimization of the upper
bounds of mechanical characteristics apparently occurs first in the 1877 book by
J. W. Strutt (Lord Rayleigh), Theory of Sound, in which a method of determining
the natural frequencies of discrete and continuous structures was presented. The
method now bears his name. Specifically, he addressed the well-known problem of
the vibrating string. For the fundamental mode he assumed
n
2x
w(x, t) = 1 cos t (10.1)
L
where is the sought circular frequency, 2L is the length of the string, x is the axial
coordinate (he chose x = 0 at the midpoint of the string). In the end, Rayleigh
arrived at the following formula:
2(n + 1)(2n + 1) T
2 = (10.2)
2n 1 L2
where T is the tension in the string, and is the mass density. For n = 1, the formula
yields = 2 L2 /T = 12, instead of the exact value = 2 ' 9.8696; n = 2 yields
a better result, = 10, whereas a minimum , or in Leissas (2005) words the best
possible result from Eq. (10.2) is 9.8990 when n = ( 6 + 1)/2 = 1.72474.
Clearly, Rayleigh carried out minimization of the upper bound of the fundamen-
tal natural frequency. Apparently he was also the first to use (see p. 287 in his
book) the combined term maximumminimum condition. (A detailed description
of Rayleighs method can be found in almost any text on vibration, and a careful
paper on its history was published by Leissa (2005)). Use of a function containing a
power-law term with an undetermined exponent was resurrected by Schmidt (1985),
and reviewed extensively by Bert (1987). Note that minimization of the maximum
response under a constraint on structural weight is equivalent to that of structural
weight under a constraint on the response.
Sniedovich (2006, 2007, 2008a, 2008b, 2008c) attributes the modern terminol-
ogy of the maximin principle to Abraham Wald (19021950) as follows: Rank
alternatives by their worst possible outcomes: adopt the alternative the worst out-
come of which is at least as good as the worst outcome of the others. Moreover,
Walds Maximin Principle remains one of the modeling pillars of decision-making

under severe uncertainty, although he is not advocating Maximin as a panacea for
decision under uncertainty. And finally,
On the face of it, the Maximin paradigm seems to be something

of a second nature of many of us: its basic attitude to uncertainty
is summed up in the familiar, indeed widely held, maxim: When
in doubt, assume the worst!
Thus, the worst-case scenario design in mechanics appears to be contemporary

with Wald (1945, 1950). Bulgakovs and Boleys problem (see Secs. 4.2 and 6.3,
respectively) is a testimony to this.
The idea of minimizing the maximum response apparently first appeared in a
deterministic setting. As Manasyan (1962) notes Usually, the dynamic analysis of
an elastic body is reduced to static analysis via introducing the so called dynamic
coefficient, the latter being the absolute value of the ratio of the maximum dynamic
displacement and the maximum static one.
Manasyan (1962) proposes to minimize the dynamic coefficient; in other words,
the maximum dynamic displacement. Later on, Brach (1968) treats a class of simply
supported beams with constrained total mass and minimizes the maximum dynamic
displacement at midpoint. Plaut (1970) and Yau (1975) study optimal design of
a simply supported beam with given total mass, minimizing the upper bound of
its dynamic response. Fox and Kapoor (1970) develop a bounding technique for
deriving the approximate peak response of planar truss frames under shock loads;
this response is minimized by a feasible-direction technique of optimization. Adali
(1982) deals with minimum-maximum deflection and stress design of beams sub-
jected to harmonic excitation. A general minimax methodology is expounded in the
monographs by Boca (1982), Demyanov and Malozemov (1990), and Du and Parda-
los (1995). Minimization of maximum of some quantity, associated with an adverse
effect, is suggested also by Bellman (1962), Besharati and Azarm (2006), Chang,
Paez and Ju (1983), Chern, Dafalias and Martin (1973), Cherkaev and Cherkaev
(1999, 2008), Cherkaev and Kucuk (1999), Gonzales and Elliot (1995), Guretskii
(1969), Hakim and Fuchs (1996), Hilding, Klarbring and Pang (1999), Kalinowski
and Pilkey (1975), Komkov and Dannon (1987), Nishihara, Asami and Watanabe
(1999a, 1999b), Paez (1981), Papoulis (1970), Popplewell and Youssef (1979), Puig,
Saludes and Quevedo (2003), Sevin (1957), Sevin and Pilkey (1967), Smallwood
(1972b, 1972a, 1973, 1989), Tseng (2006), Westermo (1985), Witte (1974), Witte
and Smallwood (1974), and Youssef and Popplewell (1978, 1979).
In his correspondence with I.E., Sniedovich (2009) stresses:
I do not have any problem with the use of term anti-

optimization, provided that it is clear what it means. However,
it is important to relate it to the state of the art. Personally, I
am happy with the old terms worst case and best-case and
Maximin. I have been using them for more that 40 years.
Elishakoff (1999) anticipating, as it were, the above message, writes:
It will be no surprise to find that the idea of anti-optimization

in its various forms stems from the times immemorial. This an-
ticipation is strengthened by the fact that its main idea resonates
well with a universal wisdom Make the best out of the worst.
Here the following personal anecdote seems to be instructive. Some years ago
one of us (I.E.) lectured at a certain university; then, the idea of combined anti-
optimization and optimization was propageted. In the questions-and-answers pe-
riod, one faculty member said: I dont like the term anti-optimization, in fact, had I
gotten your manuscript for review, I would have rejected it, just for this term! The
answer was that one of the leading optimization experts, Professor R. T. Haftka, had
endorsed it. The faculty member responded, If Haftka has embraced this notion, I
feel and am defeated! It is pleasing to record that this term has been adopted by
many researchers around the world.
The above correspondence with Sniedovich helped us to spell out the need
to discuss its various brothers and sisters. We hope that in this section we
have correlated the various origins, counterparts and equivalents of the term
optimization/anti-optimization hybrid. As can be seen, applied mechanics pro-
vided development of ideas from various analyses, leading to terms worst-case.
least favorable, best-case, most favorable, minmax or anti-optimization with
a view to determining the worst-case scenarios. Then, combined optimization seeks
to find the minimum of the worst response, or the least worst response. Hence, it is
worthwhile to make comprehensive survey of the problem formulations and method-
ologies for optimization considering maximum (worst-case) responses. Although the
idea is known independently in several fields of engineering, this book presents all
the necessary knowledge and approaches, as well as the philosophy underlining the
hybrid concept, and establishes a new direction based on computer technology.
In Chap. 4 we analyze the worst, or anti-optimized, responses in a static setting,
whereas Chap. 6 deals with analogous vibration problem. In the buckling context,
worst-case or anti-optimization entails finding the worst (or least favorable) re-
sponse, i.e., the minimum buckling load. Combination with anti-optimization leads
then to maximization of the minimum buckling load. Since the FEM is the central
numerical method in applied mechanics, it made sense to discuss, in Chap. 7, how
the anti-optimization is performed in a discretized environment. The topic is under
active development and the final word has not been said about it, luckily for us
researchers!
Chapter 8 correlates the probabilistic and anti-optimization approaches, showing
perhaps to the disappointment of separationists (those who divide and conquer
the specific fields of research) and hopefully, to the satisfaction of unionists (those
who try to find similarities in seemingly different fields) that the two approaches
lead to close or even the same results. Our general conclusion is in line with the
title of the paper by Bai and Andersland (1994) Stochastic and Worst Case Sys-
tem Identification are not Necessarily Incompatible; moreover, Bai and Andersland
claim:
Clearly the two approaches [stochastic and worst case I.E. and
M.O.] are based on different philosophies. In the literature they are
usually treated differently and separately, we believe that they are
not necessarily incompatible, and, in fact, may be complementary.
Likewise, Elishakoff (1999) in his paper Are Probabilistic and Anti-Optimization

Methods Interrelated? claims:
This study does not aim to make a bombastic statement like that
of de Finetti (1974), Probability does not exist. We can take
the liberty of suggesting to consider two possible scenarios: either
probability does not exist or it does. Consider first the former case:
if probability does not exist, then obviously, anti-optimization
method is of utmost importance, and it must be utilized in de-
sign. Consider now the latter case: if probability exists, then for
the structure to be acceptable it must possess high, near-unity
reliability; for the case of bounded random variations, the prob-
abilistic method tends to one furnished by the anti-optimization.
Thus, irrespective of the existence or nonexistence of probability,
the anti-optimization method must be incorporated.
In this vein, in Sec. 1.3 we demonstrated the convergence of probabilistic and

anti-optimization. Systematic comparison of these seemingly contradictory ap-
proaches was conducted in Chap. 8, establishing the identity of designs developed
by the probabilistic and non-probabilistic analyses, constituting pacifying news: no
need to fight and argue. We may conclude this section by quoting from Minsky
(1987)
If you understand something in only one way, then you do not

really understand it at all... The secret of what anything means
to us depends on how we have connected it to all the other things
we know.
Likewise, Takao Fujisawa (see Sato (2006)) mentions: Its only the man who
can look at the same problem from many different aspects that will make a true
leader.
It is noteworthy that stochasticians are tending more and more to invoke the
fact of physically realizable random variables and processes. To review the cur-
rent thinking of the stochasticians and the interested reader may like to consult
the papers by Cai (2003), Cai and Lin (2006), Grigoriu (2006), Ma, Leng, Meng
and Fang (2004), and Simiu and Heckert (1996) already quoted and discussed in
Sec. 1.3. It appears therefore that the prediction of increased appreciation, of the
anti-optimization method in the future, is not premature.
The combination of the probabilistic and ellipsoidal analyses was facilitated in
works by Elishakoff and Colombi (1993), Elishakoff, Lin and Zhu (1994e), Elishakoff
and Li (1999), and Qiu, Yang and Elishakoff (2008b). In this context Drenick (1995)
wrote in the correspondence with one of us (I.E.)
You mentioned that the problems you liked the best had both
probabilistic and non-probabilistic aspects. I agreed with you be-
cause I felt that the treatment of such mixed problems might
not run into the same unyielding resistance that is encountered by
the pure non-probabilistic type.
10.5 Finally, Why Did We Write This Book?
We would like to end with the provocative question in the above subtitle. Indeed,
with the abundance of texts on optimization and/or uncertainty why bother with
another?
Our answer is: because it combines these two cardinal ideas, thereby ejecting
them from their respective ivory-tower spheres and providing a friendly environment
for the much needed communication and interaction. In such circumstances, the
existential question posed by Laville (2000), Shall we abandon optimization the-
ory? can be answered in the negative, and the equally important one by Mongin
(2000) Does optimization imply rationality? in the affirmative.
The thesis that the optimization/anti-optimization hybrid saves both optimiza-
tion and uncertainty analysis from pure academism is an exaggeration. In fact, the
reliability-based design optimization (RBDO) is an established field answering the
above need of marrying uncertainty and optimization. As Chateauneuf and Aoues
(2007) note:
A very intensive research activity is performed in the field of

RBDO solution methods.
Although significant progress is performed in developing ef-
ficient numerical methods, the approach to practical engineering
structures is still a challenge, knowing the complexity of realistic
industrial systems.
These authors also stress:
While the optimization problem is carried out in the space of de-

sign variables, the reliability analysis is performed in the space
of random variables, where a lot of numerical calculations are re-

quired to evaluate the failure probability. Consequently, in or-
der to search for the optimal structural configuration, the design
variables are repeatedly changed, and each set of design variables
corresponds to a new random variable space which they need to
be manipulated to evaluate the structural reliability at the point
(Murotsu, Shao and Watanabe, 1994). Because of the too many
repeated searches needed in the above two spaces, the computa-
tional time for such an optimization becomes the main problem.
Moreover,
... the progress in System Reliability-Based Design Optimization

(SRBDO) is relatively slow because it depends on the system reli-
ability analysis where the computational time, and the numerical
instability lead to many difficulties in the SRBDO formulation
which is mainly due to design variable changes at each iteration
of the optimization procedure.
Note that the overwhelming majority of works on RBDO and SRBDO deal in
independent random variables. The need to allow for interdependence in stochastic
analysis is discussed in numerous papers; the interested reader is referred to the
papers of Ferson, Ginzburg and Akcakaya (1995) and Ferson, Nelsen, Hajagos,
Berleant, Zhang, Tucker, Ginzburg and Oberkampf (2004).
Ganzerli and de Palma (2007) write:
The book, Convex Models of Uncertainty in Applied Mechanics

by Ben-Haim and Elishakoff, established the foundation for the
convex model theory and application in 1990. Since then, much
work has been done on convex models. Together with probabilis-
tic and fuzzy sets, convex models can be considered part of the
uncertainty triangle (Elishakoff, 1990). Convex models are espe-
cially useful for problems where data on the uncertain parameters
is scarce, as in the case of severe uncertainties. One of the most
widely used of the convex models is the uniform bounded model.
Here the convex set that encloses the uncertain parameter, has
the shape of a rectangle in two dimensions. The uniform bound
convex model has been implemented in structural design using a
technique called anti-optimization (Elishakoff, Haftka and Fang,
1994c).
We here already showed the contiguity of probability and anti-optimization.

Likewise, RBDO and the optimization/anti-optimization hybrid are expected to
yield similar designs. However, we do not advocate replacement of RBDO by the
hybrid. We think that they are complementary approaches capable of peaceful

coexistence, with a preference for the hybrid due to its conceptual and numerical
simplicity.
Throughout this book, by adopting a hybrid of optimization and anti-
optimization instead of a more complex RBDO, we here tried to follow the sage
advice of Rene Descartes (15961650): Remove all redundancies from the problem
and reduce it to the simplest elements.
The simple hybrid method may attract more deterministic engineers and re-
searchers to the topic of uncertainty especially those who, for some reason or
other, doubt or deny the existence of probability.
Please communicate to us your own ideas by e-mail: elishako@fau.edu and
ohsaki@archi.kyoto-u.ac.jp, or by airmail. We will try to respond to input!
Sorry if we here omitted to quote some pertinent work: it was unintentional!
And sorry for the excess of references, if any.
Enjoy the book! This is the reason why we undertook it: we felt that both
industry and academia needed it; no less a reason was our own pleasure!

Bibliography
Aarts, E. and Korst, J. (1989). Simulated Annealing and Boltzmann Machines: A Stochas-
tic Approach to Combinatorial Optimization and Neural Computing (Wiley, Chich-
ester, U.K.).
Aarts, M. and Lenstra, J. K. (eds.) (1997). Local Search in Combinatorial Optimization
(John Wiley & Sons, Chichester, U.K.).
Abbas, A. M. and Manohar, C. S. (2005). Reliability-based critical earthquake load mod-
els, Part I: Linear structures, Part II: Nonlinear structures, Journal of Sound and
Vibration 287, pp. 865882:883900.
Abbas, A. M. and Manohar, C. S. (2007). Reliability-based vector nonstationary random
critical earthquake excitations for parametrically excited systems, Structural Safety
29, pp. 3248.
Abdullah, M. M., Richardson, A. and Hanif, J. (2001). Placement of sensors/actuators
on civil structures using genetic algorithms, Earthquake Engineering and Structural
Dynamics 30, pp. 11671184.
Adali, S. (1982). Minimum-maximum deflection and stress design of beams under harmonic
excitation by mathematical programming, Engineering Optimization 5, 4, pp. 223
234.
Adali, S. (1992). Convex and fuzzy modelings of uncertainties in the optimal design of
composite structures, in P. Pedersen (ed.), Optimal Design with Advanced Materials
(Elsevier, Amsterdam, Netherlands).
Adali, S., Elishakoff, I., Richter, A. and Verijenko, V. E. (1994a). Optimal design of sym-
metric angle-ply laminates for maximum buckling load with scatter in material prop-
erties, in J. Sobieszczanski-Sobieski (ed.), Proc. 5th AIAA/USAF/NASA/ISSMO
Symposium on Multidisciplinary Analysis and Optimization, Vol. 2, pp. 10411045,
PaperAIAA944365CP.
Adali, S., Lene, F., Duvaut, G. and Chiaruttini, V. (2003). Optimization of laminated
composites subjected to uncertain buckling loads, Composite Structures 62, pp.
261269.
Adali, S., Richter, A. and Verijenko, V. E. (1995a). Minimum weight design of symmetric
angle-ply laminates under uncertain loads, Structural and Multidisciplinary Opti-
mization 9, 2, pp. 8995.
Adali, S., Richter, A. and Verijenko, V. E. (1995b). Non-probabilistic modelling and design
of sandwich plates subject to uncertain loads and initial deflections, International
Journal of Engineering Science 33, pp. 855866.
Adali, S., Richter, A. and Verijenko, V. E. (1997). Minimum weight design of symmetric
angle-ply laminates with incomplete information on initial imperfections, Journal of
343
Applied Mechanics 64, 1, pp. 9096.

Adali, S., Verijenko, V. E. and Richter, A. (1994b). Optimal design of symmetric angle-
ply laminates subjected to uncertain loads and initial imperfections, in W. R. Blain
(ed.), Proc. Computer Aided Design in Composite Material Technology IV (Wessex
Institute of Technology, Southampton, U.K.), pp. 3948.
Adduri, P. R. and Penmetsa, R. C. (2007). Bounds on structural system reliability in the
presence of interval variables, Computers & Structures 85, 56, pp. 320329.
Ahrens, R. (1990). Modal analyse fur linearviskoelastische Schwinger, Zeitschrift fur ange-
wandte Mathematik und Mechanik (ZAMM) 70, 4, pp. T60T63, (in German).
Akgun, M. A., Garcelon, J. H. and Haftka, R. T. (2001). Fast exact linear and non-linear
structural reanalysis and the ShermanMorrisonWoodbury formulas, International
Journal for Numerical Methods in Engineering 50, pp. 15871606.
Alefeld, G. and Claudio, D. (1998). The basic properties of interval arithmetics, its software
realizations and some applications, Computers & Structures 67, pp. 38.
Alefeld, G. and Herzberger, J. (1983). Introduction to Interval Computation (Academic
Press, New York).
Alefeld, G. and Mayer, G. S. (2000). Interval analysis : Theory and applications, Journal
of Computational and Applied Mathematics 121, pp. 421464.
Allen, M. and Maute, K. (2004). Reliability-based design optimization of aeroelastic struc-
tures, Structural and Multidisciplinary Optimization 27, 4, pp. 228242.
Alotto, P., Molfino, P. and Molinari, G. (2001). Optimization of electromagnetic devices
with uncertain parameters and tolerances in the design variables, OMPEL: The
International Journal for Computation and Mathematics in Electrical and Electronic
Engineering 20, 3, pp. 808812.
Amazigo, J. C. (1974). Dynamic buckling of structures with random imperfections, in
H. E. Leipholz (ed.), Stochastic Problem in Mechanics (University of Waterloo Press,
Waterloo, Canada), pp. 243254.
Amazigo, J. C. and Frank, D. (1973). Dynamic buckling of an imperfect column on non-
linear foundation, Quarterly of Applied Mathematics 31, pp. 19.
Ang, A. H.-S. and Tang, W. H. (1975). Probabilistic Concepts in Engineering Planning
and Design (John Wiley, New York), p. 388.
Antonio, C. A. C. (2002). A multilevel genetic algorithm for optimization of geometrically
nonlinear stiffened composite structures, Structural and Multidisciplinary Optimiza-
tion 24, pp. 372386.
Arakawa, M. and Yamakawa, H. (1992). A study on optimum design using fuzzy numbers
as design variables, Japan Society of Mechanical Engineers, Part C 91-1001A, pp.
17101715, (in Japanese).
Arakawa, M. and Yamakawa, H. (1993). A study on multi-criteria structural optimum
design using qualitative and fuzzy reasoning, Finite Elements in Analysis and Design
14, pp. 127142.
Argyris, J. H. and Roy, R. (1972). General treatment of structural modification, Journal
of Structural Division, ASCE 98, ST2, pp. 465492.
Ariaratnam, S. T. (1967). Stability of column under random loading, in G. Hermann (ed.),
Dynamic Stability of Structures (Pergamon Press, New York).
Arora, J. S. (2004). Introduction to Optimum Design, 2nd edn. (Academic Press, New
York).
Arora, J. S. and Tseng, C. H. (1987). IDESIGN Users Manual, Ver. 3.5, Tech. rep.,
Optimal Design Laboratory, The University of Iowa, Iowa City, IA.
Ashley, H. (1982). On making things the best Aeronautical uses of optimization, Journal
of Aircraft 19, 1, pp. 528.
Bibliography 345
Ashley, H. and Zartarian, C. (1956). Piston theory: A new aerodynamic tool for the
aeroelastician, Journal of the Aeronautical Sciences 23, pp. 11091118.
Attoh-Okine, N. O. (2002). Uncertainty analysis in structural number determination in
flexible pavement design a convex model approach, Construction and Building
Materials 16, 2, pp. 6771.
Attoh-Okine, N. O. (2004). Uncertainty analysis in surface micromachined force gauges:
Convex model approach, Journal of Aerospace Engineering 17, 1, pp. 4044.
Au, F. T. K., Cheng, Y. S., Tham, L. G. and Zheng, G. W. (2003). Robust design of
structures using convex models, Computers & Structures 81, 2829, pp. 26112619.
Averin, A. M. (1964). On the bending of a cantilever plate weakened by a hole of a general
shape, in Proc. 4th USSR Conference on Plates and Shells (American Academy of
Sciences Press, Erevan, Armenia), pp. 167170, (in Russian).
Avis, D. and Fukuda, K. (1996). Reverse search for enumeration, Discrete Applied Math-
ematics 65, 13, pp. 2146.
Ayyub, B. (1998). Uncertainty Modeling and Analysis in Civil Engineering (CRC Press,
Boca Raton, FL).
Babuska, I., Nobile, F. and Tempone, R. (2005). Worst case scenario analysis for elliptic
problem with uncertainty, Numerische Mathematik 101, 2, pp. 185219.
Bagley, R. L. and Torvik, P. J. (1983). Fractional calculus: A different approach to the
analysis of viscoelastically damped structures, AMA Journal 21, 5, pp. 741748.
Bai, E.-W. and Andersland, M. S. (1994). Stochastic and worst-case system identification
are not necessarily incompatible, Automatics 30, 9, pp. 14911493.
Baitsch, M. and Hartmann, D. (2006). Optimization of slender structures considering
geometrical imperfections, Inverse Problems in Science and Engineering 14, 6, pp.
623637.
Balandin, D. V. (1993). Accumulation of perturbations in linear and non-linear shock-
driven systems, Journal of Applied Mathematics and Mechanics 57, 1, pp. 2127.
Banichuk, N. V. (1975). Game problems in the theory of optimal design, in Proc. IUTAM
Symposium on Optimization in Structure Design (Springer Verlag, Berlin), pp. 111
121.
Banichuk, N. V. (1976). Minimax approach to the structural optimization problems, Jour-
nal of Optimization Theory and Applications 20, 1, pp. 111127.
Banichuk, N. V. and Ivanova, S. Y. (2009). Optimization problems of contact problems,
Mechanics Based Design of Structures and Machines 37, 2, pp. 143156.
Banichuk, N. V. and Neittaanm aki, P. (2007). On structural optimization with incomplete
information, Mechanics Based Design of Structures and Machines 35, 1, pp. 7596.
Baratta, A. (1980). Analisi dei dati e simulazione per la sicurita antisismica delle
costruzioni, in Proc. National Conference L Ingegneria Sismica in Italia (Udine,
Italy).
Baratta, A. (1981). Optimized earthquake simulation for high-reliability aseismic design,
in Proc. 3rd International Seminar on Reliability of Nuclear Power Plants (Paris).
Baratta, A., Elishakoff, I., Zuccaro, G. and Shinozuka, M. (1998). A generalization of the
DrenickShinozuka model for bounds on the seismic response of a single-degree-of-
freedom system, Earthquake Engineering and Structural Dynamics 27, pp. 423437.
Baratta, A. and Zuccaro, G. (1992). High-reliability aseismic design, in Proc. 10th World
Conference in Earthquake Engineering (Madrid).
Barbieri, E., Cinquini, C. and Lombardi, M. (1997). Shape/size optimization of truss struc-
tures using non-probabilistic description of uncertainty, in S. Hern andez and C. A.
Brebbia (eds.), Computer Aided Optimum Design of Structures V (Computational
Mechanics Publications, Southampton, U.K.), pp. 163172.
Barthelemy, J. and Sobieszczanski-Sobieski, J. (1983). Optimum sensitivity derivatives of

objective functions in nonlinear programming, AIAA Journal 21, pp. 913915.
Baykasoglu, A., Owen, S. and Gindy, N. (1999). A taboo search based approach to find
the pareto optimal set in multiple objective optimization, Engineering Optimization
31, pp. 731748.
Bellman, R. (1962). On minimizing the probability of maximum deviation, IRE Transac-
tion on Automatic Control 7, 4, p. 45.
Ben-Haim, Y. (1985). The Assay of Spatially Random Material (Kluwer Academic Pub-
lishers, Dordrecht, Netherlands).
Ben-Haim, Y. (1992). Convex models for optimizing diagnosis of uncertain slender obsta-
cles on surfaces, Journal of Sound and Vibration 152, pp. 327341.
Ben-Haim, Y. (1993a). Convex models of uncertainty in radial pulse buckling of shells,
Journal of Applied Mechanics 62, pp. 683688.
Ben-Haim, Y. (1993b). Failure of an axially compressed beam with uncertain initial de-
flection of bounded strain energy, International Journal of Engineering Sciences 31,
pp. 9891001.
Ben-Haim, Y. (1994). Fatigue lifetime with load uncertainty represented by convex model,
Journal of Engineering Mechanics 120, 3, pp. 445462.
Ben-Haim, Y. (1995). A note on convex models of uncertainty for small initial imperfec-
tions of non-linear structures, Zeitschrift f
ur angewandte Mathematik und Mechanik
(ZAMM) 75, pp. 901908.
Ben-Haim, Y. (1996). Robust Reliability in the Mechanical Systems (Springer, Berlin).
Ben-Haim, Y. (1998a). Reliability of vibrating structures with uncertain inputs, The Shock
and Vibration Digest 30, 2, pp. 106113.
Ben-Haim, Y. (1998b). Sequential tests based on convex models of uncertainty, Mechanical
Systems and Signal Processing 12, pp. 427448.
Ben-Haim, Y. (1999). Convex models of uncertainty: Applications and implications,
Erkenntnis 41, 2, pp. 139156.
Ben-Haim, Y. (2001). Info-Gap Decision Theory: Decisions under Severe Uncertainty
(Academic Press, San Diego, CA), (second edition, 2006).
Ben-Haim, Y., Chen, G. and Soong, T. T. (1996). Maximum structural response using
convex model, Journal of Engineering Mechanics 122, 4, pp. 325333.
Ben-Haim, Y. and Elishakoff, I. (1989a). Dynamics and failure of thin bar with unknown-
but-bounded imperfections, in D. Hui and N. Jones (eds.), Recent Advances in Impact
Dynamics of Engineering Structures, Vol. AMD-105 (ASME Press, New York), pp.
8996.
Ben-Haim, Y. and Elishakoff, I. (1989b). Non-probabilistic models of uncertainty in the
nonlinear buckling of shells with general imperfection: Theoretical estimates of the
knockdown factor, Journal of Applied Mechanics 56, pp. 403410.
Ben-Haim, Y. and Elishakoff, I. (1990). Convex Models of Uncertainty in Applied Mechan-
ics (Elsevier, Amsterdam, Netherlands).
Ben-Haim, Y., Fr yba, L. and Yoshikawa, N. (1999). Robust design of structures using
convex models, in B. Kumar and B. Topping (eds.), Computing Developments in
Civil and Structural Engineering (Civil-Comp Press, Edinburgh, U.K.), pp. 7176.
Ben-Haim, Y. and Natke, H. G. (1992). Diagnosis of changes in elastic boundary conditions
in beams by adaptive vibrational testing, Archive of Applied Mechanics (Ingenieur
Archiv) 62, pp. 210221.
Ben-Tal, A., El Ghaoui, L. and Nemirovski, A. (2006). Foreword: Special issue on robust
optimization, Mathematical Programming, Ser. B 107, pp. 13.
Ben-Tal, A. and Nemirovski, A. (1997). Robust truss topology optimization via semidefi-
Bibliography 347
nite programming, SIAM Journal on Optimization 7, pp. 9911016.

Ben-Tal, A. and Nemirovski, A. (1998). Robust convex optimization, Mathematics of Op-
erations Research 23, 4, pp. 769805.
Ben-Tal, A. and Nemirovski, A. (2002). Robust optimization methodology and applica-
tion, Mathematical Programming B92, pp. 453480.
Bendse, M. P. and Sigmund, O. (2003). Topology Optimization: Theory, Methods and
Applications (Springer, Berlin).
Berleant, D. J. and Goodman-Strauss, C. (1998). Bounding the results of arithmetic op-
erations of random variables of unknown dependency using intervals, Reliable Com-
puting 4, pp. 147165.
Bernardini, A. (1999). What are the random and fuzzy sets and how to use them for
uncertainty modelling in engineering systems? in I. Elishakoff (ed.), Whys and Hows
in Uncertainty Modelling (Springer, Vienna), pp. 63126.
Bert, C. W. (1987). Application of a version of the Rayleigh technique to problems of bars,
beams, columns, membranes, and plates, Journal of Sound and Vibration 119, 2,
pp. 317326.
Besharati, B. and Azarm, S. (2006). Worst case deterministic feasibility and multiobjective
robustness measures for engineering design optimization, International Journal of
Reliability and Safety 1, 1/2, pp. 4058.
Bezdek, J. (1994). Fuzziness vs. probability: Again (!?), IEEE Transactions on Fuzzy
Systems 2, pp. 13.
Bickford, W. B. (1987). An improved computational technique for perturbations of the
generalized symmetric linear algebraic eigenvalue problem, International Journal
for Numerical Methods in Engineering 24, pp. 529541.
Bishop, J. A., Eastep, F. E., Striz, A. G. and Venkayya, V. B. (1998). Influence of model
complexity and aeroelastic constraints on multidisciplinary optimization of wings,
Journal of Aircraft 35, pp. 784790.
Black, M. (1937). Vagueness: An exercise in logical analysis, Philosophy of Science 4, 4,
pp. 427455.
Bland, D. R. (1960). The Theory of Viscoelasticity (Pergamon, London).
Blockley, D. I. (1977). Analysis of structural failures, Proceedings of the Institution of Civil
Engineers 62, 1, pp. 5171.
Blockley, D. I. (1980). The Nature of Structural Design and Safety (Ellis Horwood, Chich-
ester, U.K.).
Blockley, D. I. (1985). Fuzziness, probability, and fril, in Proc. 4th International Conference
on Structural Safety and Reliability (Kobe, Japan).
Bloebaum, C. L., Hajela, P. and Sobieszczanski-Sobieski, J. (1992). Non-hierarchic system
decomposition in structural optimization, Engineering Optimization 19, pp. 171
186.
Boca, T. J. (1982). Characterization of conservatism in mechanical shock testing of struc-
tures, Ph. D. dissertation, Stanford University, Stanford, CA.
Bochenek, B. (1997). Optimization of geometrically nonlinear structures with respect to
both buckling and postbuckling constraints, Engineering Optimization 29, pp. 401
415.
Bochenek, B. (2003). Problem of structural optimization for post-buckling behavior, Struc-
tural and Multidisciplinary Optimization 25, pp. 423435.
Bogdanovich, A. E. (1987). Nonlinear Problems of Dynamics of Cylindrical Composite
Shells (Zinatne Publishers, Riga, Latvia), (in Russian).
Boley, B. A. (1963). Upper and lower bounds for the solution of a melting problem, Quar-
terly of Applied Mathematics 21, 1, pp. 111.
Boley, B. A. (1964). Upper and lower bounds in problems of melting and solidifying slabs,
Quarterly Journal of Mechanics and Applied Mathematics 17, pp. 253269.
Boley, B. A. (1966a). Bounds on the maximum thermoelastic stress and deflection in a
beam and plate, Journal of Applied Mechanics 33, 4, pp. 881887.
Boley, B. A. (1966b). Estimate of errors in approximate temperature and thermal stress
calculations, in H. Gortler and P. Sorger (eds.), Applied Mechanics, Proc. 11th Inter-
national Congress of Applied Mechanics, Munich, Germany, 1964 (Springer, Berlin).
Boley, B. A. and Weiner, J. H. (1960). Theory of Thermal Stresses (Wiley, New York).
Bolotin, V. V. (1961). Statistical Methods in Structural Mechanics (Gosstroiizdat,
Moscow), (First edition, in Russian, second edition, 1965).
Bolotin, V. V. (1963). Non-Conservative Problems in the Theory of Elastic Stability (Perg-
amon, New York), pp. 242257.
Bolotin, V. V. (1969). Statistical Methods in Structural Mechanics, Holden-Day Series
in Mathematical Physics 7 (Holden-Day, San Francisco, CA), (Translation of the
second, 1965 edition).
Borkowski, A. (1983). On application of ellipsoidal methods in rigid-plastic analysis and
design, in H. A. Eschenauer and N. Olhoff (eds.), Optimization Methods in Structural
Design (Wissenschaftsverlag, Mannheim, Germany), pp. 289294.
Borse, G. J. (1997). Numerical Method with MATLAB (International Thomson Publish-
ing).
Boyd, S. and Vandenberghe, L. (2004). Convex Optimization (Cambridge University Press,
Cambridge, U.K.).
Brach, R. M. (1968). Minimum dynamic response for a class of simply supported beam
shapes, International Journal of Mechanical Science 10, pp. 429439.
Brown, C. B. (1979). A fuzzy safety measure, Journal of Engineering Mechanics Division,
ASCE 105, 5, pp. 855872.
Brown, C. B., Johnson, J. L. and Loftus, J. J. (1984). Subjective seismic safety assessments,
Journal of Structural Engineering 110, 9, pp. 22122233.
Brown, C. B. and Yao, J. T. P. (1987). Resent civil engineering applications of fuzzy sets,
Mathematical Modelling 9, 6, pp. 491492.
Bruhn, E. F. (ed.) (1975). Analysis and Design of Missile Structures (Tri-Stable Offset
Company, Cincinnati, OH).
Budiansky, B. and Hutchinson, J. W. (1964). Dynamic buckling of imperfection-sensitive
structures, in H. G ortler and P. Sorger (eds.), Applied Mechanics, Proc. 11th Inter-
national Congress of Applied Mechanics (Springer, Berlin), pp. 636651.
Bulgakov, B. V. (1939). Applied Theory of Gyroscopes (Moscow), (in Russian).
Bulgakov, B. V. (1940). Fehleranhaefung bei Kreiselapparaten, Ingenieur-Archiv 11, pp.
461469, (in German).
Bulgakov, B. V. (1946). On the accumulation of disturbances in linear systems with con-
stant coefficients, Doklady Akademii Nauk SSSR LI(5), pp. 339342, (in Russian).
Bulgakov, B. V. and Kozovkov, N. T. (1950). On the accumulation of disturbances in linear
systems with variable coefficients, Prikladnaya Matematika i Mekhanika (Applied
Mathematics and Mechanics) 14, pp. 712, (in Russian).
Burgman, M. (2008). Shakespeare, Wald and decision making under uncertainty, Decision
Point 23, p. 10.
Burns, S. (ed.) (2002). Recent Advancements in Optimal Structural Design (ASCE Press,
Raton, VA).
Cai, G. Q. (2003). Generation of bounded stochastic process, in W. Q. Zhu, G. Q. Cai
and R. C. Zhang (eds.), Advances in Stochastic Structural Engineering (CRC Press,
Boca Raton, FL).
Bibliography 349
Cai, G. Q. and Lin, Y. K. (2006). An improved equivalent linearization procedure for

nonlinear system under bounded random excitations, Journal of Structural Control
and Health Monitoring 13, 1, pp. 336346.
Cai, G. Q. and Wu, C. F. (2004). Modeling of bounded stochastic processes, Probabilistic
Engineering Mechanics 19, 3, pp. 197203.
Cai, K.-Y. (1996). Introduction to Fuzzy Reliability (Springer, Berlin).
Calafiore, G. and El Ghaoui, L. (2004). Ellipsoidal bounds for uncertain linear equations
and dynamical systems, Automatica 40, pp. 773787.
Capiez-Lernout, E. and Soize, C. (2008). Design optimization with an uncertain vibroa-
coustic model, Journal of Vibration and Acoustics 130, pp. 02100210210018.
Chakraborty, S. and Sam, P. C. (2006). Probabilistic safety analysis of structures under
hybrid uncertainty, International Journal for Numerical Methods in Engineering 70,
4, pp. 405422.
Chang, F. C., Paez, T. L. and Ju, F. (1983). Least favorable response of inelastic struc-
tures, Tech. Rep. Technical Report CE-63(83) AFOSR-793-1, Department of Civil
Engineering and Department of Mechanical Engineering, The University of New
Mexico, Albuquerque, NM.
Chateauneuf, A. and Aoues, Y. (2007). Advances in solution methods for reliability-based
design optimization, in Y. Tsompanakis, N. D. Lagaros and M. Papadrakakis (eds.),
Structural Design Optimization Considering Uncertainties (Taylor & Francis, Lon-
don), pp. 217246.
Chen, R. and Ward, A. C. (1990). Generalizing interval matrix operations for design,
Journal of Mechanical Design 119, 11, pp. 6572.
Chen, S.-H. and Lian, H. D. (2002). Dynamic response analysis for structures with interval
parameters, Structural Engineering and Materials 13, 2, pp. 299312.
Chen, S.-H., Lian, H. D. and Yang, X. W. (2002). Interval static displacement analysis for
structures with interval parameters, International Journal for Numerical Methods in
Engineering 53, pp. 393407.
Chen, S.-H., Qiu, Z. P. and Liu, Z. S. (1994). Perturbation method for computing eigen-
value bounds in structural vibration systems with interval parameters, Communica-
tions in Numerical Methods in Engineering 10, 2, pp. 121134.
Chen, S.-H., Qiu, Z. P. and Song, D. (1995). A new method for computing the upper
and lower bounds on frequencies of structures with interval parameters, Mechanics
Research Communications 22, 5, pp. 431439.
Chen, S.-H. and Wu, J. (2004a). Interval optimization of dynamic response for structures
with interval parameters, Computers & Structures 82, 1, pp. 111.
Chen, S.-H. and Wu, J. (2004b). Interval optimization of dynamic response for uncertain
structures with natural frequency constraints, Engineering Structures 26, 2, pp.
221232.
Chen, S.-H., Wu, J. and Chen, M. D. (2004). Interval optimization for uncertain structures,
Finite Elements in Analysis and Design 40, pp. 13791398.
Chen, S.-H., Wu, X. M. and Yang, Z. J. (2006). Eigensolution reanalysis of modified
structures using epsilon-algorithm, International Journal for Numerical Methods in
Chen, S.-H. and Yang, X. W. (2000). Interval finite element method for beam structures,
Finite Elements in Analysis and Design 34, pp. 7588.
Chen, S.-H., Yang, X. W. and Lian, H. D. (2000). Comparison of several eigenvalue reanal-
ysis methods for modified structures, Structural and Multidisciplinary Optimization
20, pp. 253259.
Chen, W. Y., Chu, F. L., Yan, S. O. and Wang, K. Y. (2007). An interval parameter
perturbation method for predicting the natural frequency bounds of intelligent truss
structures with uncertain-but-bounded parameters, Key Engineering Materials 347,
pp. 569574.
Cheng, G., Xu, L. and Jiang, L. (2006). A sequential approximate programming strategy
for reliability-based structural optimization, Computers & Structures 84, pp. 1353
1367.
Cheng, Y. S., Au, F. T. K., Tham, L. G. and Zeng, G. W. (2004). Optimal and robust
design of doking blocks with uncertainty, Engineering Structures 26, pp. 499510.
Cherkaev, A. V. and Cherkaev, E. (1999). Optimal design for uncertain loading conditions,
in V. Berdichevsky, V. Jikov and G. Papanicolaou (eds.), Homogenization (World
Scientific, Singapore), pp. 193213.
Cherkaev, A. V. and Kucuk, I. (1999). Optimal structures for various loadings, in Proc.
3rd World Congress of Structural and Multidisciplinary Optimization (Buffalo, N.
Y.).
Cherkaev, E. and Cherkaev, A. V. (2003). Principal compliance and robust optimal design,
Journal of Elasticity 72, pp. 7198.
Cherkaev, E. and Cherkaev, A. V. (2008). Minimax optimization problem of structural
design, Computers & Structures 86, 1314, pp. 14261435.
Chern, J. M., Dafalias, J. F. and Martin, J. B. (1973). Structural design for bounds on
dynamic response, Journal of Engineering Mechanics Division, ASCE 99, pp. 261
270.
Chernousko, F. L. (1980). Optimal guaranteed estimates of uncertainties by means of
ellipsoids, IIII, Tekhnicheskaya Kibernetika Part I, issue 2, 311: Part II, issue 4,
311: Part III, issue 5, 511 (in Russian), (translated into English as Engineering
Cybernetics).
Chernousko, F. L. (1988). Estimation of the Phase State of Dynamic Systems (Nauka
Publishing House, Moscow), (in Russian).
Chernousko, F. L. (1994). State Estimation of Dynamic Systems (CRC Press, Boca Raton,
FL).
Chernousko, F. L. (1999). What is ellipsoidal modeling and how to use it for control and
state estimation? in I. Elishakoff (ed.), Whys and Hows in Uncertainty Modelling
(Springer, Vienna), pp. 127188.
Chernousko, F. L. (2000). Ellipsoidal bounds of reachable sets of dynamical systems with
matrices subjected to uncertain perturbations, Journal of Optimization Theory and
Application 104, 1, pp. 119.
Chernousko, F. L. (2005). Ellipsoidal state estimation for dynamical systems, Nonlinear
Analysis 63, pp. 872879.
Chiang, W. L., Dong, W. M. and Wong, F.-S. (1987). Dynamic response of structures with
uncertain parameters : A comparative study of probabilistic and fuzzy set models,
Probabilistic Engineering Mechanics 2, pp. 8291.
Cho, M. and Rhee, S. Y. (2003). Layup optimization considering free-edge strength and
bounded uncertainty of material properties, AIAA Journal 41, 11, pp. 22742282.
Cho, M. and Rhee, S. Y. (2004). Optimization of laminates with free edges under bounded
uncertainty subject to extension, bending and twisting, International Journal of
Solids and Structures 41, pp. 227245.
Choi, K. K. and Kim, N.-H. (2004). Structural Sensitivity Analysis and Optimization 1,
Linear Systems (Springer, New York).
Christensen, R. M. (1982). Theory of Viscoelasticity (Academic Press, New York).
Chryssanthopoulos, M. K. and Poggi, C. (1995). Stochastic imperfection modeling in shell
buckling studies, Thin-Walled Structures 23, 14, pp. 179200.
Bibliography 351
Clough, R. W. and Penzien, J. (1975). Dynamics of Structures (McGraw-Hill, New York).

Cohn, M. Z. (1994). Theory and practice of structural optimization, Structural Optimiza-
tion 7, pp. 2031.
Cohon, J. L. (1978). Multiobjective Programming and Planning, Mathematics in Science
and Engineering, Vol. 140 (Academic Press, New York).
Conrado, A. C., de Faria, A. R. and de Almeida, S. F. M. (2005a). Optimum design for
buckling of arbitrary shaped ribs under uncertain loadings, Aeronautical Journal
109, pp. 609618.
Conrado, A. C., de Faria, A. R. and de Almeida, S. F. M. (2005b). A procedure to optimize
aeronautical ribs subjected to several load cases, in Proc. 6th World Congress of
Structural and Multidisciplinary Optimization (Rio de Janeiro, Brazil).
Corliss, G., Foley, C. and Keafott, R. B. (2007). Formulation of reliable analysis of struc-
tural frames, Reliable Computing 13, 2, pp. 125147.
Cornell, C. A. (1981). Structural safety: Some historical evidence that is a healthy adoles-
cent, in T. Moan and M. Shinozuka (eds.), Structural Safety and Reliability (Elsevier,
Amsterdam, Netherlands), pp. 1951.
Cozzarelli, F. A. and Huang, W. N. (1971). Effect of random material parameters on
nonlinear steady creep solutions, International Journal of Solids and Structures 2,
pp. 14771494.
Craig, K. J., Stander, N. and Balasubramanyam, S. (2003). Worst-case design in head
impact crashworthiness optimization, International Journal for Numerical Methods
in Engineering 57, 6, pp. 795817.
Cremona, C. and Gao, Y. (1997). The possibilistic reliability theory: Theoretical aspects
and applications, Structural Safety 19, 2, pp. 173201.
de Faria, A. R. (2000). Buckling Optimization of Composite Plates and Cylindrical Shells:
Uncertain Loading Conditions, Ph. D dissertation, Institute of Aerospace Studies,
University of Toronto, Toronto, Canada.
de Faria, A. R. (2002). Buckling optimization of composite plates: uncertain loading com-
bination, International Journal for Numerical Methods in Engineering 53, 3, pp.
719732.
de Faria, A. R. (2004). Compliance minimization of structures under uncertain loadings,
Latin American Journal of Solids and Structures 1, pp. 363378.
de Faria, A. R. (2007). Prebuckling enhancement of beams and plates under uncertain
loading and arbitrary initial imperfections, Journal of the Brazilian Society of Me-
chanical Sciences and Engineering 29, 4, pp. 388395.
de Faria, A. R. and de Almeida, S. F. M. (2003a). Buckling optimization of plates with
variable thickness subjected to nonuniform uncertain loads, International Journal of
Solids and Structures 40, pp. 39553966.
de Faria, A. R. and de Almeida, S. F. M. (2003b). Buckling optimization of variable thick-
ness plates subjected to nonuniform uncertain loads, in XXIV Iberian Latin-America
Congress on Computational Methods in Engineering (Auto Preto/MG, Brazil).
de Faria, A. R. and de Almeida, S. F. M. (2004). Buckling optimization of variable thickness
composite plates subjected to nonuniform loads, AIAA Journal 42, 2, pp. 228231.
de Faria, A. R. and de Almeida, S. F. M. (2006). The maximization of fundamental
frequency of structures under arbitrary initial stress states, International Journal
for Numerical Methods in Engineering 65, pp. 445460.
de Faria, A. R. and Hansen, J. S. (2001a). Buckling optimization of composite axisym-
metric cylindrical shells under uncertain loading conditions, Journal of Applied Me-
chanics 68, 4, pp. 632639.
de Faria, A. R. and Hansen, J. S. (2001b). On buckling optimization under uncertain
loading combinations, Structural and Multidisciplinary Optimization 21, pp. 272

282.
de Finetti, B. (1974). Theory of Probability: A Critical Introductory Treatment, Vol. 1
(Wiley, London).
Deif, A. (1986). in Sensitivity Analysis in Linear Systems (Springer, Berlin), pp. 1620.
Deif, A. (1991). Advanced Matrix Theory for Scientists and Engineers (Abacus Press, New
York).
Deml, M. and Wunderlich, W. (1997). Direct evaluation of the worst imperfection shape
in shell buckling, Computer Methods in Applied Mechanics and Engineering 149,
14, pp. 201222.
Dempster, A. P. (1967). Upper and lower probability induced multivalued mapping, Annals
of Mathematics and Statics 38, pp. 325339.
Demyanov, V. M. and Malozemov, V. N. (1990). Introduction to Minimax (Dover, New
York).
Deng, L. and Ghosn, M. (2001). Pseudoforce method for nonlinear analysis and reanalysis
of structural systems, Journal of Structural Engineering 127, 5, pp. 570578.
Deodatis, G., Graham, L. and Micaletti, R. (2003a). A hierarchy of upper bounds on the
response of stochastic systems with large variation of their properties: Random field
case, Probabilistic Engineering Mechanics 18, 4, pp. 365375.
Deodatis, G., Graham, L. and Micaletti, R. (2003b). A hierarchy of upper bounds on
the response of stochastic systems with large variation of their properties: Random
variable case, Probabilistic Engineering Mechanics 18, 4, pp. 349364.
Deodatis, G. and Shinozuka, M. (1989). Bounds on response variability of stochastic sys-
tems, Journal of Engineering Mechanics 115, pp. 25432563.
Deodatis, G. and Shinozuka, M. (1991). Weighted integral method II: Response variability
and reliability, Journal of Engineering Mechanics 117, 8, pp. 18651877.
Der Kiureghian, A. and Thoft-Christensen, P. (eds.) (1990). Reliability and Optimization
of Structural Systems (Springer, Berlin).
Dessombz, O., Thouverez, F., Laine, J.-P. and Jezequel, L. (2001). Analysis of mechanical
systems using interval computations applied to finite element methods, Journal of
Sound and Vibration 239, 5, pp. 949968.
Dessombz, O., Thouverez, F., Laine, J.-P. and Jezequel, L. (2003). Structural design
of uncertain mechanical systems using interval arithmetic, in G. Gogu, D. Coute-
lier, P. Chedmail and P. Ray (eds.), Recent Advancements in Integrated Design and
Manufacturing in Mechanical Engineering (Kluwer Academic Publishers, Dordrecht,
Netherlands).
Dimarogonas, A. D. (1993). Interval analysis of vibrating systems, in Vibration of Me-
chanical Systems and the History of Mechanical Design (ASME Press, New York),
pp. 8995.
Dimarogonas, A. D. (1995). Interval analysis of vibrating systems, Journal of Sound and
Vibration 183, pp. 739749.
Ditlevsen, O. (1980). Formal and real structural safety: Influence of gross errors, in
P. Thoft-Christensen (ed.), Lectures on Structural Reliability (Institute of Building
Technology and Structural Engineering, Aalborg University, Denmark), pp. 121147.
Dorn, W., Gomory, R. and Greenberg, H. (1964). Automatic design of optimal structures,
Journal de Mecanique 3, pp. 2552.
Drenick, R. F. (1968). Functional analysis of effect of earthquakes, in Proc. 2nd Joint
United States-Japan Seminar on Applied Acoustics (Washington D.C.).
Drenick, R. F. (1970). Model-free design of aseismic structures, Journal of Engineering
Mechanics Division, ASCE 96, pp. 483493.
Bibliography 353
Drenick, R. F. (1973). Aseismic design by way of critical excitation, Journal of Structural

Division, ASCE 99, pp. 649667.
Drenick, R. F. (1977a). The critical excitation of nonlinear systems, Journal of Applied
Mechanics 44, E2, pp. 333336.
Drenick, R. F. (1977b). On a class of non-robust problem in stochastic dynamics, in B. L.
Clarkson (ed.), Stochastic Problems in Dynamics (Pitman, London), pp. 237255.
Drenick, R. F. (1993). Personal communication to Isaac Elishakoff.
Drenick, R. F. (1995). Personal communication to Isaac Elishakoff, September 3.
Drenick, R. F. (2001). Private communication to Izuru Takewaki, kindly provided by the
latter to us.
Drenick, R. F., Novomestky, F. and Bagchi, G. (1989). Critical excitation of structure, in
Wind and Seismic Effects, Proc. 12th Joint UJNR Panel Conference (U.S. National
Bureou of Standards Special Publication), pp. 133142.
Drenick, R. F. and Park, C. B. (1975). Comments on worst inputs and a bound on the
height peak statistics of a class of non-linear systems, Journal of Sound and Vibration
41, pp. 129131.
Drenick, R. F., Wang, P. C., Yun, C. B. and Philappacopoulos, A. J. (1980). Critical
seismic response of nuclear reactors, Journal of Nuclear Engineering and Design 59,
pp. 425435.
Drenick, R. F. and Yun, C. B. (1979). Reliability of seismic resistance prediction, Journal
of Structural Division, ASCE 105, pp. 18791891.
Du, D. Z. and Pardalos, P. M. (1995). Minimax and Applications (Springer, Berlin).
Du, L., Choi, K. K. and Youn, B. D. (2006). Inverse possibility analysis method for
possibility based design optimization, AIAA Journal 44, 11, pp. 26822690.
Dube, R. and Hausen, Y. S. (2004). Optimal design for structures subjected to many
load cases, in Proc. 10th AIAA/ISSMO Multidisciplinary Analysis and Optimization
Conference, Paper-AIAA-2004-4531 (Albany, NY).
Dubois, D. (1987). An application of fuzzy arithmetic to the optimization of industrial
machining processes, Mathematical Modelling 9, 6, pp. 461435.
Dubois, D. and Prade, H. (1991). Random sets and fuzzy interval analysis, Fuzzy Sets and
Systems 42, pp. 87101.
Dubois, D. and Prade, H. (1993). Fuzzy sets and probability: misunderstandings, bridges
and gaps, in Proc. 2nd IEEE Conference on Fuzzy Systems (San Francisco), pp.
10591068.
Dwyer, P. S. (1951). Linear Computations (John Wiley & Sons, New York).
Dyne, S. J. C., Hammond, J. K. and Davies, P. (1988). A method of finding an upper bound
for the response of structures for blast waves, in M. Petyt, H. F. Wolfe and C. Mei
(eds.), Recent Advances in Structural Dynamics (Institute of Sound and Vibration
Research, Southampton, U.K.), pp. 551559.
Dzhur, Y., Gordeyev, V. and Shimanovsky, A. (2001). Determining disadvantageous com-
binations of loads in analysis of linearly deformable systems, in H. Kunieda (ed.),
Theory, Design and Realization of Shell and Spatial Structures (International Asso-
ciation for Shell and Spatial Structures, Nagoya, Japan), pp. 3839.
El Damatty, A. A. and Nassef, A. O. (2001). A finite element optimization technique to
determine critical imperfections of shell structures, Structural and Multidisciplinary
Optimization 23, pp. 7587.
Elishakoff, I. (1978a). Axial impact buckling of columns with random initial imperfections,
Journal of Applied Mechanics 45, pp. 361365.
Elishakoff, I. (1978b). Impact buckling of thin bar via Monte Carlo method, Journal of
Elishakoff, I. (1983). Probabilistic Methods in the Theory of Structures (Wiley (Inter-

science), New York), second edition: Minneola, NY: Dover Publications, 1999.
Elishakoff, I. (1990). An idea of uncertainty triangle, The Shock and Vibration Digest 22,
10, p. 1, (editorial).
Elishakoff, I. (1991a). Convex versus probabilistic modelling of uncertainty in structural
dynamics, in M. Petyt, H. F. Wolfe and C. Mei (eds.), Structural Dynamics: Recent
Advances (Elsevier, London), pp. 321.
Elishakoff, I. (1991b). Essay on probability, convexity and earthquake engineering, in
Llngegneria Sismica in Italia, Proc. 5th Italian National Conference on Earthquake
Engineering, Vol. 2, pp. 907935.
Elishakoff, I. (1991c). Essay on reliability index, probabilistic interpretation of safety fac-
tor, and convex models of uncertainty, in F. Casciati and J. B. Roberts (eds.),
Reliability Problems: General Principles and Applications in Mechanics of Solids
and Structures (Springer, Berlin), pp. 237271.
Elishakoff, I. (1994). A new safety factor based on convex modelling, in B. M.Ayyub and
M. M. Gupta (eds.), Uncertainty Modelling and Analysis (Elsevier Science Publish-
ers, Amsterdam, Netherlands), pp. 145172.
Elishakoff, I. (1995a). Discussion on A non-probabilistic concept of reliability, Structural
Safety 17, pp. 195199.
Elishakoff, I. (1995b). Essay on uncertainties in elastic and viscoelastic structures: From
A. M. Freudenthals criticisms to modern convex modeling, Computers & Structures
56, pp. 871895.
Elishakoff, I. (1998a). How to introduce initial-imperfection sensitivity concept into design
2, Tech. Rep. M. Stein Vol., NASA CP, 206280.
Elishakoff, I. (1998b). Three facets of uncertainty, Computers & Structures 67, 1/3, pp.
12.
Elishakoff, I. (1998c). Three versions of the finite element method based on concepts of
either stochasticity, fuzziness, or anti-optimization, Applied Mechanics Reviews 51,
3, pp. 209218.
Elishakoff, I. (1999). Whys and Hows of Uncertainty Modeling: Probability, Fuzziness, and
Anti-Optimization, CISM Courses and Lectures, No. 388 (Springer, Vienna).
Elishakoff, I. (2000a). Possible limitations of probabilistic methods in engineering, Applied
Mechanics Reviews 53, 2, pp. 1936.
Elishakoff, I. (2000b). Uncertain buckling: Its past, present and future, International
Journal of Solids and Structures 37, 4647, pp. 68696889.
Elishakoff, I. (2004). Safety Factors and Reliability: Friends or Foes? (Kluwer Academic
Publishers, Dordrecht, Netherlands).
Elishakoff, I. and Ben-Haim, Y. (1990a). Dynamics of thin shell under impact with limited
deterministic information on initial imperfections, Structural Safety 8, pp. 103112.
Elishakoff, I. and Ben-Haim, Y. (1990b). Probabilistic and convex models of uncertainty in
structural dynamics, in Proc. 8th International Modal Analysis Conference (Orlando,
FL), pp. 14871492.
Elishakoff, I., Cai, G. Q. and Starnes, Jr., J. H. (1994a). Non-linear buckling of a column
with initial imperfection via stochastic and non-stochastic convex model, Interna-
tional Journal of Non-Linear Mechanics 29, 1, pp. 7182.
Elishakoff, I. and Colombi, P. (1993). Combination of probabilistic and convex models
of uncertainty when scarce knowledge is present on acoustic excitation parameters,
Computer Methods in Applied Mechanics and Engineering 104, 2, pp. 187207.
Elishakoff, I. and Duan, D. H. (1994). Application of the mathematical theory of interval
analysis for uncertain vibrations, in Y. M. Cuscheiri, S. A. L. Glegg and D. M.
Bibliography 355
Yeager (eds.), Proc. 1994 Conference on Noise Control Engineering (New York),
pp. 519524.
Elishakoff, I., Eliseeff, P. and Glegg, S. (1994b). Convex modelling of material uncertainty
in vibrations of viscoelastic structures, AIAA Journal 32, pp. 843849.
Elishakoff, I. and Fang, J. J. (1995). Diagnosis of local uncertain modification in the
boundary condition of a rectangular plate via convex modeling, Computer Methods
in Applied Mechanics and Engineering 124, 4, pp. 303319.
Elishakoff, I. and Ferracuti, B. (2006a). Four alternative definitions of the fuzzy safety
factor, Journal of Aerospace Engineering 19, 4, pp. 281287.
Elishakoff, I. and Ferracuti, B. (2006b). Fuzzy sets based interpretation of the safety factor,
Fuzzy Sets and Systems 157, 18, pp. 24952512.
Elishakoff, I., Gana-Shvili, J. and Givoli, D. (1991). Treatment of uncertain imperfec-
tions as a convex optimization problems, in L. Esteva and S. E. Ruiz (eds.), Proc.
6th International Conference on Applications of Stochastics and Probability in Civil
Engineering, Vol. 1 (Mexico City), pp. 150157.
Elishakoff, I., Haftka, R. T. and Fang, J. (1994c). Structural design under bounded
uncertaintyoptimization with anti-optimization, Computers & Structures 53, 6,
pp. 14011405.
Elishakoff, I. and Hasofer, A. M. (1996). Detrimental and serendipitous effect of human
error on reliability of structures, Computer Methods in Applied Mechanics and En-
gineering 129, pp. 17.
Elishakoff, I. and Li, Q. (1999). How to combine probabilistic and anti-optimization meth-
ods? in Whys and Hows of Uncertainty Modeling: Probability, Fuzziness, and Anti-
Optimization, CISM Courses and Lectures, No. 388 (Springer, Vienna), pp. 319339.
Elishakoff, I., Li, Y. W. and Starnes, Jr., J. H. (1994d). A deterministic method to pre-
dict the effect of unknown-but-bounded elastic moduli on the buckling of composite
structures, Computer Methods in Applied Mechanics and Engineering 111, 12, pp.
155167.
Elishakoff, I., Li, Y. W. and Starnes, Jr., J. H. (1996). Imperfection sensitivity due to
the elastic moduli in the RoordaKoiter frame, Chaos, Solitons & Fractals 7, 8, pp.
11791186.
Elishakoff, I., Li, Y. W. and Starnes, Jr., J. H. (2001). Non-Classical Problems in the
Theory of Elastic Stability (Cambridge University Press, Cambridge, U.K.).
Elishakoff, I., Lin, Y. K. and Zhu, L. P. (1994e). Probabilistic and Convex Modelling of
Acoustically Excited Structures, Studies in Applied Mechanics 39 (Elsevier, Amster-
dam, Netherlands).
Elishakoff, I. and Pletner, B. (1991). Analysis of base excitation as an uncertain function
with specified bounds on it and its derivatives, in Structural Vibration and Acoustics,
Proc. ASME Design Technical Conference (Miami, FL), pp. 177183.
Elishakoff, I., Ren, Y. J. and Shinozuka, M. (1997). New formulation of FEM for deter-
ministic and stochastic beams through generalization of Fuchs approach, Computer
Methods in Applied Mechanics and Engineering 144, 3-4, pp. 235243.
Elishakoff, I., van Manen, S., Vermeulen, P. G. and Arbocz, J. (1987). First-order second-
moment analysis of the buckling of shells with random initial imperfections, AIAA
Journal 25, 8, pp. 11131117.
Elishakoff, I., Wang, X. J. and Qiu, Z. P. (2008). Comparison of interval analysis and
convex analysis, in R. L. Muhanna and R. L. Mullen (eds.), Proc. 3rd International
Workshop on Reliable Engineering Computing (Georgia Institute of Technology, Sa-
vannah, GA), pp. 269288.
Elishakoff, I. and Zingales, M. (2000). Probability vs anti-optimization in an applied me-
chanics problem: Part I, scalar uncertainty, Journal of Applied Mechanics 67, pp.
472483.
Elishakoff, I. and Zingales, M. (2003). Contrasting probabilistic and anti-optimization
approaches in an applied mechanics problem, International Journal of Solids and
Structures 40, pp. 42814297.
Elseifi, M. (1998). A New Scheme for the Optimum Design of Stiffened Composite Panels
with Geometric Imperfections, Ph. D dissertation, Department of Aerospace Engi-
neering, Virginia Polytechnic Institute and State University, Blacksburg, VA.
Elseifi, M. A., G urdal, Z. and Nikolaidis, E. (1999). Convex/probabilistic models of un-
certainties in geometric imperfections of stiffened composite panels, AIAA Journal
37, 4, pp. 468474.
Elseifi, M. A. and Khalessi, M. R. (2001). A convex set based optimization algorithm
for structural probabilistic analysis, in 42nd AIAA/ASME/AHS/ASCE Structures,
Structural Dynamics, and Materials Conference and Exhibit.
Eschenauer, H. A. (2003). Optimization procedures in structural design, Applied Mechanics
Reviews 56, 4, pp. R19R26.
Eschenauer, H. A. and Olhoff, N. (2001). Topology optimization of continuum structures:
A review, Applied Mechanics Reviews 54, 4, pp. 331390.
Fang, J. J., Smith, S. A. and Elishakoff, I. (1998). Combination of anti-optimization and
fuzzy-set-based analysis for structural optimization under uncertainty, Mathematical
Problems in Engineering 4, pp. 187200.
Fatemi, J., Mooij, E., Gurav, S. P., Andreykiv, A. and van Keulen, F. (2005). Re-entry
vehicle design optimization with integrated trajectory uncertainty, in AIAA/CIRA
13th International Space Planes and Hypersonics Systems and Technologies Confer-
ence, AIAA-2005-3386 (Capua, Italy).
Fellin, W., Lessmann, L., Oberguggenberger, M. and Vieider, R. (eds.) (2005). Analyzing
Uncertainty in Civil Engineering (Springer, Berlin).
FEMA (2000). Recommended Seismic Design Criteria for New Steel Moment-Frame Build-
ings, FEMA-350, Federal Emergency Management Agency, Washington D.C.
Ferson, S. (2002). Ramas Risk Calc 4.0 Software: Risk Assessment with Uncertain Numbers
(Lewis Publishers, Boca Raton, FL).
Ferson, S. and Ginzburg, L. R. (1996). Different methods are needed to propagate igno-
rance and variability, Reliability Engineering and System Safety 54, 2, pp. 133144.
Ferson, S., Ginzburg, L. R. and Akcakaya, R. (1995). Whereof one cannot speak: when
input distributions are unknown, (Available at http://www.ramas.com.whereof.pdf).
Ferson, S., Nelsen, R. B., Hajagos, J. G., Berleant, D. J., Zhang, J., Tucker, W. T.,
Ginzburg, L. R. and Oberkampf, W. L. (2004). Dependence on probabilistic mod-
eling, DempsterShafer theory and probability bounds analysis, in Sandia Report
SAND 20043072.
Feynman, R. F. (1988). What Do You Care What Other People Think? (Bantham Books,
New York).
Floudas, C. A. (1995). Nonlinear and Mixed-Integer Optimization (Oxford University
Press, New York).
Fl
ugge, W. (1975). Viscoelasticity (Springer-Verlag, Berlin).
Fox, R. L. and Kapoor, M. P. (1970). Structural optimization in the dynamic response
regime: A computational approach, AIAA Journal 8, 10, pp. 17981804.
Frangopol, D. M. (1995). Reliability-based optimum structural design, in C. Sundararajan
(ed.), Probabilistic Structural Mechanics Handbook: Theory and Industrial Applica-
tions (Chapman & Hall, New York), pp. 352387.
Frangopol, D. M. (1998). Probabilistic structural optimization, Progress in Structural En-
Bibliography 357
gineering and Mechanics 1, 2, pp. 223230.

Frangopol, D. M. and Corotis, R. B. (1996). Reliability-based structural system optimiza-
tion, in F. Y. Cheng (ed.), Analysis and Computation (ASCE Press, New York), pp.
6778.
Frangopol, D. M., Corotis, R. B. and Rackwitz, R. (eds.) (1996). Reliability and Optimiza-
tion of Structural Systems (Boulder CO).
Frangopol, D. M. and Maute, K. (2005). Reliability-based optimization in civil and
aerospace structural systems, in E. Nikolaidis, D. M. Ghiocel and S. Singhal (eds.),
Engineering Design Reliability Handbook (CRC Press, Boca Raton, FL), pp. 24.1
24.32.
Freudenthal, A. M. (1956). Safety and the probability of structural failure, Transaction of
ASCE 121, pp. 13371397.
Freudenthal, A. M. (1957). Safety and safety factor for airframe, Technical Report 153,
NATO, Palaia de Chaillot, France.
Freudenthal, A. M. (1961). Fatigue sensitivity and reliability of mechanical systems, espe-
cially aircraft structures, in WAAD Technical Report, pp. 5363.
Freudenthal, A. M. (1968). Critical appraisal of safety factor criteria and their basic
concepts, in Preliminary Publication, Eighth Congress (International Association
of Bridge and Structural Engineers, New York), pp. 1328.
Fr
yba, L., Nakagiri, S. and Yoshikawa, N. (1998). Interval analysis of a beam on uncertain
foundation subjected to a single moving force, in Proc. Conference on Computational
Engineering and Science (Tokyo), pp. 849850.
Fr
yba, L. and Yoshikawa, N. (1997). The worst case estimation of the dynamic response
of a beam based on convex model of uncertainties formulation, in P. Anagonos-
foupoules, G. M. Carbmagns and C. A. Brebbia (eds.), Computational Methods and
Experimental Measurement (Computational Mechanics Publications, Southampton,
U.K.), pp. 223232.
Fr
yba, L. and Yoshikawa, N. (1998). Bounds analysis of a beam based on the convex model
of uncertain foundation, Journal of Sound and Vibration 212, 3, pp. 547557.
Fuchs, M. B. (1991). Unimodal beam elements, International Journal of Solids and Struc-
tures 27, 5, pp. 533545.
Fukuda, K. (1999). cdd+ Ver. 0.76 Users Manual, Tech. rep., Institute of Operation
Research, ETH-Zentrum, Zurich.
Fuller, R. B. (1975). Synergetics, Explorations in the Geometry of Thinking (Collier
Macmillan, London).
Furuta, H., Dogaki, M. and Sakano, M. (eds.) (2002). Reliability and Optimization of
Structural Systems (Balkema, Rotterdam, Netherlands).
Gal, T. (1979). Postoptimal Analyses, Parametric Programming, and Related Topics (Mc-
Graw Hill, New York).
Gana-Shvili, J. (1991). Solution of Optimization Problems of Structures with Uncertain
Parameters, M. Sci. Thesis, Department of Aeronautical Engineering, Technion
Israel Institute of Technology, Haifa.
Gangadharan, S. N., Nikolaidis, E., Lee, K. and Haftka, R. T. (2003). Antioptimization for
comparison of alternative structural models and damage detection, AIAA Journal
37, 7, pp. 857864.
Ganzerli, S. and de Palma, P. (2007). Genetic algorithms in structural optimum design
using convex models of uncertainty, in Y. Tsompanakis, N. D. Lagaros and M. Pa-
padrakakis (eds.), Structural Design Optimization Considering Uncertainties (Taylor
& Francis, London), pp. 549568.
Ganzerli, S. and Pantelides, C. P. (2000). Optimum structural design via convex model
superposition, Computers & Structures 77, 6, pp. 639647.

Ganzerli, S., Pantelides, C. P. and Reaveley, L. D. (2000). Performance-based design with
structural optimization, Earthquake Engineering and Structural Dynamics 29, 11,
pp. 16771690.
Gao, W., Zhang, N., Ma, J. and Wang, X. B. (2008). Interval dynamic response analysis
of structures with interval parameters, Journal of Mechanical Engineering Science:
Proc. Institution of Mechanical Engineering, Part C 222, C3, pp. 377385.
Gasser, M. and Schueller, G. I. (1997). Reliability based optimization of structural systems,
Mathematical Methods of Optimization 46, pp. 287307.
Gay, D. M. (1982). Solving interval linear equations, SIAM Journal on Numerical Analysis
19, 4, pp. 858870.
Gill, P. E., Murray, W. and Saunders, M. A. (2002). SNOPT: An SQP algorithm for large-
scale constrained optimization, SIAM Journal on Optimization 12, pp. 9791006.
Givoli, D. and Elishakoff, I. (1992). Stress concentration at nearly circular hole with un-
certain irregularities, Journal of Applied Mechanics 59, pp. 6571.
Glover, F. (1989). Tabu search: Part I. ORSA Journal on Computing 1, 3, pp. 190206.
Gnoenskii, L. S. (1961). On accumulation of disturbances, Prikladnaya Matematika i
Mekhanika (Applied Mathematics and Mechanics) 25, 2.
Godoy, L. A. and Taroco, E. O. (2000). Design sensitivity of post-buckling states including
material constraints, Computer Methods in Applied Mechanics and Engineering 188,
pp. 665679.
Goldberg, D. E. (1989). Genetic Algorithms in Search, Optimization, and Machine Learn-
ing (Addison-Wesley, Reading, MA).
Golla, D. F. and Hughes, P. C. (1985). Dynamics of viscoelastic structures: A time domain
finite element formulation, Journal of Applied Mechanics 52, pp. 897906.
Goncharenko, V. M. (1962). Application of Markov processes in statistical theory of shells
stability, Ukrainskii Matematicheskii Zhurnal (Ukrainian Mathematical Journal) 14,
2, pp. 198202, (in Russian).
Gonzales, A. and Elliot, S. (1995). Minimization of the maximum error signal in active con-
trol, ISVR Technical Memos, No. 758, Institute of Sound and Vibration, Southamp-
ton, U.K.
Good, I. J. (1995). Reliability always depends on probability, Journal of Statistical Com-
putation and Simulation 52, pp. 192193.
Grandori, G. (1991). Paradigms and falsification in earthquake engineering, Meccanica 26,
pp. 1721.
Grigoriu, M. (2006). Probabilistic models for material properties. Available at:
http://www.esc.sandiagov/VCWwebsite/PostConference/Statics/grigoriu.pdf.
Gu, X., Renaud, J. E., Batill, S. M., Brach, R. M. and Budhiraja, A. S. (2000). Worst-case
propagated uncertainty in multidisciplinary systems in robust design optimization,
Structural and Multidisciplinary Optimization 20, pp. 190213.
Guo, S.-X. and L u, Z.-Z. (2001). Interval arithmetic and static interval FEM, Applied
Mathematics and Mechanics 22, 12, pp. 13901396.
Guo, S.-X. and L u, Z.-Z. (2003). Comparison between the non-probabilistic and proba-
bilistic reliability methods for uncertain structure design, Chinese Journal of Applied
Mechanics 3.
Gurav, S. P. (2005). Uncertainty-Based Design Optimization of Structures with Bounded-
But-Unknown Uncertainties, Ph. D. dissertation, Delft, Netherlands: Delft Univer-
sity Press.
Gurav, S. P., Goosen, J. F. L. and van Keulen, F. (2005a). Bounded-but-unknown uncer-
tainty optimization using design sensitivity and parallel computing: Application to
Bibliography 359
MEMS, Computers & Structures 83, pp. 11341149.

Gurav, S. P., Kasyap, A., Sheplak, M., Cattafesta, R. T. and Haftka, R. T. (2004).
Uncertainty-based design optimization of a micro piezoelastic composite energy
reclamation device, in 10th AIAA/ISSMO Multidisciplinary Analysis and Optimiza-
tion Conference, AIAA-2004-4619 (Albany, NY).
Gurav, S. P., Langhaar, M., Goosen, J. F. L. and van Keulen, F. (2005b). Uncertainty-
based design optimization of MEMS structures using combined cycle-based alter-
nating anti-optimization, in Proc. 6th World Congress of Structural and Multidisci-
plinary Optimization (Rio de Janeiro, Brazil).
Gurav, S. P. and van Keulen, F. (2005). Bounded-but-unknown uncertainties in design
optimization by combining multipoint approximation and design sensitivities, Tech.
rep., Structural Optimization and Computational Mechanics Group, Faculty of De-
sign, Engineering and Production, Delft University of Technology, Delft, Nether-
lands.
Guretskii, V. V. (1969). Towards the problem of minimization of maximum deviation,
Proc. Leningrad Polytechnic Institute 307, pp. 1121, (in Russian).
Gurov, S. V. and Utkin, L. V. (1999). System Reliability under Incomplete Information
(Lubavitch Publishing, St. Petersburg, Russia), (in Russian).
Guz, A. N. and Nemish, Y. N. (1987). Perturbation method of the boundarys form in
mechanics of continua (review), Prikladnaya Mekhanika (Applied Mechanics) 23, 9,
pp. 329.
Haftka, R. T., G urdal, Z. and Kamat, M. P. (1992). Elements of Structural Optimiza-
tion: Solid Mechanics and Its Applications, 3rd edn. (Kluwer Academic Publishers,
Dordrecht, Netherlands).
Hajela, P. and Vittal, S. (2006). Optimal design in the presence of modeling uncertainty,
Journal of Aerospace Engineering 19, 4, pp. 204216.
Hakim, S. and Fuchs, M. B. (1996). Quasi-static optimal actuator placement with minimum
worst case distortion criterion, AIAA Journal 34, 7, pp. 15051511.
Haldar, A. and Reddy, R. K. (1992). A random fuzzy analysis of existing structures, Fuzzy
Sets and Systems 48, pp. 201210.
Han, S. W. and Chopra, A. K. (2006). Approximate incremental dynamic analysis us-
ing the modal pushover analysis procedure, Earthquake Engineering and Structural
Dynamics 35, pp. 18531873.
Hangai, Y. (1990). Shape analysis of structures, in Theoretical and Applied Mechanics,
Proc. 39th Japan National Congress for Applied Mechanics, Vol. 39 (University of
Tokyo Press, Tokyo), pp. 1128.
Hansen, E. and Walster, G. W. (1992). Global Optimization Using Interval Analysis (Mar-
cel Dekker, New York).
Hanss, M. (2005). Applied Fuzzy Arithmetic: An Introduction with Engineering Applica-
tions (Springer, Berlin).
Haug, E. J., Choi, K. K. and Komkov, V. (1986). Design Sensitivity Analysis of Structural
Systems (Academic Press, New York).
Hayes, B. (2003). A lucid interval, American Scientist 91, pp. 484488.
He, C. Y. and Zhang, J. H. (1997). Convex-model analysis of the seismic response, Applied
Mathematics and Mechanics (English Edition) 18, 10, pp. 941947.
Hemp, W. S. (1973). Optimum Structures (Oxford University Press, Oxford, U.K.).
Henderson, H. V. and Searle, R. (1981). On deriving the inverse of a sum of matrices,
SIAM Review 23, pp. 5360.
Hilding, D., Klarbring, A. and Pang, J.-S. (1999). Minimization of maximum unilateral
force, Computer Methods in Applied Mechanics and Engineering 177, 34, pp. 215
234.
Hilton, H. H. and Feigen, M. (1960). Minimum weight analysis based on structural relia-
bility, Journal of Aerospace Science 27, 9, pp. 641652.
Hilton, H. H., Hsu, J. and Kirby, J. S. (1991). Linear viscoelastic analysis with random
material properties, in Y. K. Lin and I. Elishakoff (eds.), Stochastic Structural Dy-
namics 1: New Theoretical Developments (Springer-Verlag, Berlin), pp. 83110.
Hirota, K. (1991). Fuzzy concept very clear in Japan, Asahi Evening News, August 20,
Special Supplement .
Hlavacek, I. (2002a). Post-buckling range of plates in axial compression with uncertain
initial imperfection, Applied Mathematics 47, pp. 25444.
Hlavacek, I. (2002b). Worst scenario approach for elastoplasticity with hardening and
uncertain input data, Zeitschrift f
ur angewandte Mathematik und Mechanik (ZAMM)
82, pp. 671684.
Hlavacek, I. (2007). Von K arman equations with uncertain input data and worst scenario
method, Zeitschrift f ur angewandte Mathematik und Mechanik (ZAMM) 87, pp.
747761.
Hlavacek, I., Chleboun, J. and Babuska, I. (2004). Uncertain Input Data Problem and the
Worst-Case Scenario Method (North-Holland, Amsterdam, Netherlands).
Hlavacek, I., Plesek, J. and Gabriel, D. (2006). Validation and sensitivity study of an
elastoplastic problem using the worst scenario method, Computer Methods in Applied
Mechanics and Engineering 195, pp. 763774.
Ho, D. (1974). Buckling load of nonlinear systems with multiple eigenvalues, International
Journal of Solids and Structures 10, pp. 13151330.
Hoff, N. J. (1965). in G. Herrmann (ed.), Dynamic Stability of Structures (Pergamon Press,
New York), pp. 744.
Hoh, S. and Hagaki, H. (1989). Fuzzy-Bayesian analysis to structural reliability, in Proc.
ICOSSAR89, San Francisco (ASCE Press, New York), pp. 17711774.
Holmes, J. D. (2002). Discussion of Classical extreme value model and prediction of
extreme wind, by Janos Galambos and Nicholas Macri, Journal of Structural En-
gineering 125, 7, pp. 792794.
Horn, R. A. and Johnson, C. R. (1990). Matrix Analysis, reprint edn. (Cambridge Univer-
sity Press, Cambridge, U.K.).
Horst, R. and Tuy, H. (1990). Global Optimization (Springer-Verlag, Berlin).
Hu, B. and Schiehlen, W. (1996). Eigenvalue, frequency response and variance bounds of
linear damped systems, European Journal of Mechanics, A/Solids 15, pp. 617646.
Huang, W. N. and Cozzarelli, F. A. (1972). Steady creep in a beam with random material
parameters, Journal of the Franklin Institute 294, 5, pp. 323338.
Huang, W. N. and Cozzarelli, F. A. (1973). Damped lateral vibration in an axially creep-
ing beam with random material parameters, International Journal of Solids and
Huseyin, K. (1975). Nonlinear Theory of Elastic Stability (Noordhoff, Leyden, Nether-
lands).
Ikeda, K. and Murota, K. (1990). Critical initial imperfection of structures, International
Journal of Solids and Structures 26, 8, pp. 865886.
Ikeda, K., Ohsaki, M. and Kanno, Y. (2005). Imperfection sensitivity of hilltop branching
points of systems with dihedral group symmetry, International Journal of Non-linear
Mechanics 40, pp. 755774.
Ikeda, K., Oide, K. and Terada, K. (2002). Imperfection sensitive strength variation of crit-
ical loads at hilltop bifurcation point, International Journal of Engineering Science
40, pp. 743772.
Bibliography 361
Ilyushin, A. A. (1956). The law of plane sections at high supersonic velocities, PMM;
Journal of Applied Mathematics and Mechanics 20, pp. 733755, (in Russian).
Inman, D. (1989). Vibration analysis of viscoelastic beams by separation of variables and
modal analysis, Mechanics Research Communications 16, 4, pp. 213218.
Ishida, M. and Tagami, S. (1959). On the tension of an infinite strip containing a hole of
an arbitrary profile, in Proc. 9th Japan National Congress for Applied Mechanics,
pp. 5154.
Iyengar, R. N. (1970). Matched inpurs, Report 47, Series Y, Center of Applied Statistics,
Purdue University, West Lafayette, IN.
Iyengar, R. N. and Dash, P. K. (1976). Random vibration analysis of stochastic time-
varying systems, Journal of Sound and Vibration 45, pp. 6989.
Iyengar, R. N. and Manohar, C. S. (1987). Nonstationary random critical seismic excita-
tions, Journal of Engineering Mechanics 113, 4, pp. 529541.
Jensen, H. (2000). An interval analysis approach to deal with earthquake response of uncer-
tain systems, in Proc. 12th World Conference on Earthquake Engineering (Auckland,
New Zealand).
Jensen, H. (2001). Structural optimal design of systems with imprecise properties: a pos-
sibilistic approach, Advances in Engineering Software 32, 12, pp. 937948.
Jensen, H. and Maturana, S. (2002). A possibilistic decision support system for impre-
cise mathematical programming problems, International Journal of Production Eco-
nomics 77, 2, pp. 145158.
John, F. (1998). Extremum problems with inequalities as subsidiary conditions, in
K. Friedrichs, O. Neugebauer and J. J. Stocker (eds.), Studies and Essays Presented
to R. Courant on His 60th Birthday (Wiley, New York).
Kalinowski, A. and Pilkey, W. D. (1975). Design for incompletely prescribed loading,
Journal of Engineering Mechanics Division, ASCE 101, 4, pp. 505510.
Kalman, R. E. (1994). Randomness reexamined, Modeling, Identification and Control 15,
3, pp. 141151.
Kam, T.-Y. and Brown, C. B. (1983). Updating parameters with fuzzy entropies, Journal
of Engineering Mechanics 109, 6, pp. 13341343.
Kanda, J. (1994). Application of an empirical extreme value distribution to load models,
Journal of Research at the National Institute of Standards and Technology 99, pp.
413420.
Kanno, Y. and Ohsaki, M. (2005). Minimum principle of complementary energy for non-
linear elastic cable networks with geometrical nonlinearity, Journal of Optimization
Theory and Application 126, 3, pp. 617641.
Kanno, Y., Ohsaki, M. and Ito, J. (2002). Large-deformation and friction analysis of
non-linear elastic cable networks by second-order cone programming, International
Kanno, Y., Ohsaki, M. and Katoh, N. (2001). Sequential semidefinite programming for
optimization of framed structures under multimodal buckling constraints, Interna-
tional Journal of Structural Stability and Dynamics 1, 4, pp. 585602.
Kanno, Y. and Takewaki, I. (2006a). Confidence ellipsoids for static response of truss
with load and structural uncertainty, Computer Methods in Applied Mechanics and
Kanno, Y. and Takewaki, I. (2006b). Robustness analysis of trusses with separable local
and structural uncertainties, International Journal of Solids and Structures 93, 9,
pp. 26962669.
Kanno, Y. and Takewaki, I. (2006c). Sequential semidefinite program for maximum robust-
ness design of structures under load uncertainty, Journal of Optimization Theory and
Application 130, 2, pp. 265287, also Kyoto University, BGE Research Report 04-05,
2004, Available at http://www.archi.kyoto-u.ac.jp/bge/RR/.
Kanno, Y. and Takewaki, I. (2007a). Maximum robustness design of trusses via semidefi-
nite programming, in Y. Tsompanakis, N. D. Lagaros and M. Papadrakakis (eds.),
Structural Design Optimization Considering Uncertainties (Taylor & Francis, Lon-
don), pp. 471498.
Kanno, Y. and Takewaki, I. (2007b). Worst-case plastic limit analysis of trusses under
uncertain loads via mixed 0-1 programming, Journal of Mechanics of Materials and
Kapur, K. C. (1975). Optimization in design by reliability, AIIE Transactions 7, 2, pp.
185192.
Kaveh, A. (2004). Structural Mechanics: Graph and Matrix Methods, 3rd edn. (Research
Studies Press, Somerset, U.K.).
Kavlie, D., Graham, H. and Powell, G. H. (1971). Efficient reanalysis of modified struc-
tures, Journal of Structural Division, ASCE 97, ST1, pp. 377392.
Keafott, R. B. (1996). Interval computations: Introduction, uses, and resources, Euromath
Bulletin 2, 1, pp. 95112.
Khachaturian, N. and Haider, G. S. (1966). Probabilistic design of determinate structures,
in Proc. Specialty Conference (ASCE, New York), pp. 623641.
Kharitonov, V. (1997). Interval uncertainty structure: conservative and simple, in H. G.
Natke and Y. Ben-Haim (eds.), Uncertainty: Models and Measures (Akademie Ver-
lag, Berlin), pp. 231243.
Kharmanda, G., Olhoff, N. and El-Hami, A. (2004). Recent developments in reliability-
based design optimization, in Proc. WCCM VI in conjunction with APCOM04
(Tsinghua University Press & Springer-Verlag, Beijing).
Khodaparasi, H. H., Marques, S., Badcock, K. J. and Mottershead, J. E. (2009). Estima-
tion of flutter boundaries in the presence of structural uncertainty by probabilistic
and fuzzy methods, in N. Maia (ed.), Proc. International Conference on Structural
Engineering Dynamics (Ericeira, Portugal).
Kida, T. and Komatsu, K. (1993). Vibration control of space structures, Science of Ma-
chine 45, 12, pp. 12731279, (in Japanese).

Kildibekov, I. G. (1971). Nonlinear vibrations of circular cylindrical shells under axial
loading and acoustic pressure, in W. V. Bolotin and A. Chyras (eds.), Problemi
Nadezhnosti v Stroitel noi Mekhanike (Problems of Reliability in Structural Me-
chanics) (Vilnius, Lithuania), pp. 5863.
Kim, C. and Choi, K. K. (2008). Reliability-based design optimization using response
surface method with prediction interval estimation, Journal of Mechanical design
130, 12, Paper No. 121401.
Kim, S.-E. and Chen, W.-F. (1996). Practical advanced analysis for braced steel frame
design, Journal of Structural Engineering 122, 11, pp. 12661274.
Kim, T.-U. and Sin, H.-C. (2001). Optimal design of composite laminated plates with
the discreteness in ply angles and uncertainty in material properties considered,
Computers & Structures 79, pp. 25012509.
Kim, Y. V., Ovseyevich, A. I. and Reshetnyak, Y. N. (1993). Comparison of stochastic and
guaranteed approaches to the estimation of the state of dynamic systems, Journal
of Computer and System Sciences International 31, 6, pp. 5664.
Kirkpatrick, S., Gelatt, C. D. and Vecchi, M. P. (1983). Optimization by simulated an-
nealing, Science 220, 4598, pp. 671680.
Kirsch, U. (1989). Optimal topologies of truss structures, Applied Mechanics Reviews 42,
pp. 223239.
Bibliography 363
Kirsch, U. (1994). Efficient sensitivity analysis for structural optimization, Computer Meth-
ods in Applied Mechanics and Engineering 117, pp. 143156.
Kirsch, U. (2000). Combined approximation: A general reanalysis approach for structural
optimization, Structural and Multidisciplinary Optimization 20, pp. 97106.
Kirsch, U. and Liu, S. (1995). Exact structural reanalysis by a first-order reduced basis
approach, Structural Optimization 10, pp. 153158.
Kirsch, U. and Liu, S. (1996). Structural reanalysis for general layout modification, in
Proc. 6th AIAA/NASA/ISSMO Symposium on Multidisciplinary Analysis and Op-
timization (AIAA), pp. 11421149.
Kirsch, U. and Rubinstein, M. F. (1972). Structural reanalysis by iteration, Computers &
Klir, G. J. (1989). Is there more to uncertainty than some probability theorists might have
us to believe? International Journal for General Systems 15, pp. 347378.
Klir, G. J. (1994). On the alleged superiority of probabilistic representation of uncertainty,
IEEE Transaction on Fuzzy Systems 2, pp. 2731.
Klir, G. J., St. Clair, U. H. and Yuan, B. (1997). Fuzzy Set Theory: Foundations and
Applications (Prentice Hall, Upper Saddle River, NJ), pp. 50,63,75.
Kogiso, N., Shao, S., Miki, M. and Murotsu, T. (1997). Reliability analysis and reliability-
based design of composite laminated plate subjected to buckling, in Proc. 38th
AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials
Conference and Exhibit, Vol. Part 2 (Kissimee, FL).
Kogiso, N., Shao, S. and Murotsu, T. (1996). Sensitivity analysis of buckling load of
the simply supported laminated plate, Transaction of JSME 62A, pp. 834841, (in
Japanese).
Koiter, W. T. (1945). On the Stability of Elastic Equilibrium, Ph. D. dissertation, Delft Uni-
versity of Technology, Delft, Netherlands, english Translation, NASA, TTF-10833,
1967.
Koltunov, M. A. (1976). Creep and Relaxation (Visshaya Shkola Publishing House,
Moscow), (in Russian).
Komkov, V. and Dannon, V. (1987). Design optimization for structural or mechanical
systems against the worst possible loads, Acta Applicandae Mathematicae 8, pp.
3764.
Korneev, A. B. and Svetlitskii, V. A. (1995). Analysis of nonstationary dynamic processes
in the presence of mixed information on random perturbation, Journal of Machinery
Manufacture and Reliability 3, pp. 612.
Kosko, B. (1990). Fuzziness vs. probability, International Journal of General System 17,
pp. 211240.
Kosko, B. (1992). Neural Networks and Fuzzy Systems: A Dynamical Systems Approach
to Machine Intelligence (Prentice Hall, Englewood Cliffs, NJ), p. 3.
Kosko, B. (1993). Fuzzy Thinking: The New Science of Fuzzy Logic (Hyperion, New York).
Kosko, B. (1994). The probability monopoly, IEEE Transaction on Fuzzy System 2, 1, pp.
2233.
Kosko, B. (1999). The Fuzzy Future: From Society and Science to Heaven in Chip (Har-
mony Books, New York).
Kocvara, M. and Stingl, M. (2006). The worst-case multiple load FMO problem revisited, in
M. P. Bendse, N. Olhoff and O. Sigmund (eds.), IUTAM Symposium on Topological
Design Optimization of Structures, Machines and Materials: Status and Perspectives
(Springer, Berlin), pp. 403411.
K
oyluoglu, H. U., C akmak, H. S. and Nielsen, S. R. K. (1995). Interval algebra to deal
with pattern loading and structural uncertainties, Journal of Engineering Mechanics
121, pp. 11191157.

K
oyluoglu, H. U. and Elishakoff, I. (1998). A comparison of stochastic and interval finite
elements applied to shear frames with uncertain stiffness properties, Computers &
K
oyluoglu, U. and Elishakoff, I. (1998). A comparison of stochastic and interval finite
elements applied to shear frames with uncertain stiffness properties, Computers &
Structures 67, 1-3, pp. 9198.
Kreinovich, V., Neumaier, A. and Xiang, G. (2008). Towards a combination of interval
and ellipsoidal uncertainty, Computational Technologies 13, 6, pp. 516.
Kreinovich, V., Xiang, G., Starks, S. A., Longpre, L., Ceberio, M., Araiza, R., Beck, J. L.,
Kandathi, R., Nayak, A., Torres, R. and Hajagos, J. G. (2006). Towards combining
probabilistic and interval uncertainty in engineering calculations: Algorithms for
computing statistic under interval uncertainty, and their computational complexity,
Reliable Computing 12, 6, pp. 471501.
Krichevskii, I. Y. and Ulanov, G. M. (1973). About a generalization of theory of accumu-
lation of disturbances, Prikladnaya Matematika i Mekhanika (Applied Mathematics
and Mechanics) 37, pp. 794802, (in Russian).
Kulpa, Z., Pownuk, A. and Skalna, I. (1998). Analysis of linear mechanical structures with
uncertainties by means of interval methods, CAMES 15, 4, pp. 443447.
Kurzhanski, A. B. (1977). Control and Observation under Condition of Uncertainty
(Nauka, Moscow), (in Russian).
Kurzhanski, A. B. and Valyi, I. (1992). Ellipsoidal techniques for dynamic systems: Control
synthesis for uncertain systems, Dynamic Control 2, pp. 87111.
Kurzhanski, A. B. and Valyi, I. (1996). Ellipsoidal Calculus for Estimation and Control
(Birkh auser, Boston).
Kurzhanski, A. B. and Valyi, I. (1997). Ellipsoidal Calculus of Estimation and Control
(Birkh auser, Boston).
Kurzhanski, A. B. and Varaiya, P. (2002). Ellipsoidal techniques for reliability analysis,
Part 1: External approximation, Optimization Methods and Software 17, 2, pp.
177206.
Kuttekeuler, J. and Ringertz, U. T. (1998). Aeroelastic tailoring considering uncertainties
in material properties, Structural Optimization 15, pp. 157162.
Kwak, B. M. and Haug, E. J. (1976). Optimum design in the presence of parametric
uncertainty, Journal of Optimization Theory and Applications 19, 4, pp. 527546.
Kwakernaak, H. (1978). Fuzzy random variables I: Definitions and theorems, Information
Science 15, 1, pp. 129.
Ladevèze, P., Puel, G. and Romeuf, T. (2006). Lack of knowledge in structural model
validation, Computer Methods in Applied Mechanics and Engineering 195, pp. 4697
4710.
Lalvani, H. (1996). Origins of tensegrity: views of Emmerich, Fuller and Snelson, Interna-
tional Journal of Space Structures 11, 1, pp. 2755.
Langley, R. S. (2000). Unified approach to probabilistic and possibilistic analysis of uncer-
tain structures, Journal of Engineering Mechanics 126, 11, pp. 11631172.
Lanzi, L. and Giavotto, V. (2006). Post-buckling optimization of composite stiffened pan-
els: Computations and experiments, Composite Structures 73, pp. 208220.
Laville, F. (2000). Shall we abandon optimization theory? The need for bounded rational-
ity, Journal of Economic Methodology 7, 3, pp. 395426.
Lee, J., Haftka, R. T., Griffin, O. H., Watson, L. T. and Sensmeier, M. D. (1994). Detecting
delaminations in a composite beam using anti-optimization, Structural Optimization
8, 23, pp. 93100.
Bibliography 365
Lee, T. H. and Jung, J. J. (2007). Kriging metamodel based optimization, in J. S. Arora

(ed.), Optimization of Structural and Mechanical Systems (World Scientific, Singa-
pore).
Leichtweiss, K. (1980). Konvexe Mengen (VEB Deucher Verlag der Wissenschaft, Berlin),
(in German).
Leissa, A. W. (2005). The historical bases of the Rayleigh and Ritz methods, Journal of
Sound and Vibration 287, pp. 961978.
Leng, H. N. and He, I. Q. (2007). Computing eigenvalue bounds of structures with
uncertain-but-non-random parameters by a method based on perturbation theory,
Communications in Numerical Methods in Engineering 23, 11, pp. 973982.
Ley, R. and Peltier, A. (2001). Optimal sizing of a composite sandwich fuselage component,
Vehicle Design 25, 12, pp. 89114.
Li, Y. W., Elishakoff, I., Starnes, Jr., J. H. and Shinozuka, M. (1996). Prediction of natural
frequency and buckling load variability due to uncertainty in material properties by
convex modeling, Fields Institute Communication 9, pp. 139154.
Liaw, D. G. and Yang, T. Y. (1990). Reliability of initially compressed uncertain laminated
plates in supersonic flow, AIAA Journal 29, 1, pp. 5260.
Librescu, L. and Bainer, L. (1983). Aeroelastic optimization of orthotropic rectangular flat
panels, Optimal Control Applications and Models 4, pp. 193198.
Lignon, S. and Jezequel, L. (2006). A robust approach for seismic damage assessment,
Computers & Structures 85, 12, pp. 414.
Lindberg, H. E. (1965). Impact buckling of thin bar, Journal of Applied Mechanics 32,
pp. 312322.
Lindberg, H. E. (1992a). Convex models for uncertain imperfection control in multimode
dynamic buckling, Journal of Applied Mechanics 59, 4, pp. 937945.
Lindberg, H. E. (1992b). Evaluation of convex modeling for multimode dynamic buckling,
Journal of Applied Mechanics 59, 4, pp. 929936.
Lindley, D. V. (1987). The probability approach to the treatment of uncertainty in artificial
intelligence and expert systems, Statistical Science 2, pp. 1724.
Liu, Z. S., Chen, S.-H. and Han, W. Z. (1994). Solving the extremum of static response for
structural systems with unknown but bounded parameters, Computers & Structures
50, pp. 557561.
Livne, E. (1997). Integrated areoservoelastic optimization: Status and direction, AIAA
Paper 971407.
Livne, E. and Mineau, D. (1997). Panel flutter constraints: Analytic sensitivities and
approximations including planform shape design variables, Journal of Aircraft 34,
pp. 558568.
Lockhart, D. and Amazigo, J. C. (1975). Dynamic buckling of externally pressurized im-
perfect cylindrical shells, Journal of Applied Mechanics 42, pp. 316320.
Lockwood, E. H. (1961). A Book of Curves (Cambridge University Press, Cambridge,
U.K.), p. 148.
Lodwick, W., Jamison, D. and Russell, S. (2000). A comparison of fuzzy, stochastic and
deterministic methods in linear programming, in T. H. Whalen (ed.), Proc. 19th In-
ternational Conference of the North American Fuzzy Information Processing Society
NAFIPS (Atlanta, GA), pp. 321325.
Lomakin, V. A. (1968). Concentration of stresses near a surface with rapidly oscillating
irregularities, Prikladnaya Mekhanika 4, 2, pp. 18, (English translation: Soviet
Applied Mechanics, No. 2, pp. 14, 1968).
Lomakin, V. A. and Sheinin, V. I. (1970). Statistical characteristics of stress fields in a
randomly inhomogeneous elastic plane, Mechanics of Solids 5, 4, pp. 110115.
Lomakin, V. A. and Sheinin, V. I. (1974). Stress concentration at the boundary of a

randomly inhomogeneous elastic body, Mechanics of Solids 9, 2, pp. 6570.
Lombardi, M. (1995). Ottimizzazione Strutturale: Metodo a due Passi e Applicazioni a
Strutture in Materiale Composito, Ph. D. dissertation, Polytechnic of Milan, Italy,
(in Italian).
Lombardi, M. (1998). Optimization of uncertain structures using non-probabilistic models,
Lombardi, M. and Haftka, R. T. (1998). Anti-optimization technique for structural design
under load uncertainties, Computer Methods in Applied Mechanics and Engineering
157, 1-2, pp. 1931.
Lombardi, M., Mariani, M. and Venini, P. (1998). Optimization of uncertain layered com-
posites using the LMI method, in Proc. Australian Conference on Structural Opti-
mization, ACSO 98, pp. 157164.
Luenberger, D. G. (2003). Linear and Nonlinear Programming (Kluwer Academic Pub-
lishers, Boston).
Ma, X. P., Leng, X. L., Meng, G. A. and Fang, T. (2004). Evolutionary earthquake response
of uncertain structure with bounded random parameter, Probabilistic Engineering
Mechanics 19, 3, pp. 239246.
Maes, M. and Huyse, L. (eds.) (2003). Reliability and Optimization of Structural Systems
(Balkema, Rotterdam, Netherlands).
Maglaras, G., Nikolaidis, E., Haftka, R. T. and Cudney, H. H. (1997). Analytical
experimental comparison of probabilistic methods and fuzzy set based methods for
designing under uncertainty, Structural Optimization 13, 23, pp. 6980.
Makode, P. V., Corotis, R. B. and Ramirez, M. R. (1999). Nonlinear analysis of frame
structures by pseudodistortions, Journal of Structural Engineering 125, 11, pp.
13091317.
Manasyan, A. A. (1962). A law of systems loading that guarantees minimum dynamic
coefficient, Mechanics of Solids 4, pp. 178180, (in Russian).
Manohar, C. S. and Sarkar, A. (1995). Critical earthquake input power spectral density
function models for engineering structures, Earthquake Engineering and Structural
Dynamics 24, pp. 15491566.
Maplesoft (2004). Maple 9 Programming Guide.
Marler, T. and Arora, J. S. (2004). Survey of multi-objective optimization methods for
engineering, Structural and Multidisciplinary Optimization 26, 6, pp. 369395.
Marti, K. (2005). Stochastic Optimization Methods (Springer, Berlin).
Masur, E. F. (1984). Optimal structural design under multiple eigenvalue constraints,
International Journal of Solids and Structures 20, pp. 211231.
Matheron, G. (1975). Random Sets and Integral Geometry (Wiley, New York).
Matsumoto, M. and Iwaya, E. (2001). Interval finite element analysis of structural systems
with uncertainties (static analysis of truss structures with geometric uncertainties),
Transactions JSME 67A, 662, pp. 15761582, (in Japanese).
Mayer, M. (1926). Die Sicherheit des Bauwerke und Ihre Berechnung nach Grenzkr aften
anstatt nach Zulgassigen Spannungen (Springer, Berlin), (in German).
Maymon, G. and Libai, A. (1977). Dynamics and failure of cylindrical shells subjected to
axial impact, AIAA Journal 15, pp. 16211630.
McNeill, D. and Freiberger, P. (1993). Fuzzy Logic: The Revolutionary Computer Tech-
nology That Is Changing Our World (Touchstone, New York).
McTavish, D. J. (1988). The mini-oscillator technique: A finite element method for the
modeling of linear viscoelastic structures, Tech. Rep. Report No. 323, Institute for
Aerospace Studies, University of Toronto, Toronto, Canada.
Bibliography 367
McTavish, D. J., Hughes, P. C., Soucy, Y. and Graham, W. B. (1992). Prediction and mea-
surement of modal damping factors for viscoelastic space structures, AIAA Journal
30, 5, pp. 13921399.
McWilliam, S. (2001). Anti-optimization of uncertain structures using interval analysis,
Melosh, R. J. and Luik, R. (1968). Multiple configuration analysis of structures, Journal
of Structural Division, ASCE 94, ST11, pp. 25812596.
Michalopoulos, C. D. and Riley, T. A. (1972). Response of discrete linear systems to forcing
functions with inequality constraints, AIAA Journal 10, 8, pp. 10161019.
Michell, A. G. M. (1904). The limits of economy in frame structures, Philosophical Maga-
zine Sect. 6 8, 47, pp. 589597.
Minciarelli, F., Gioffre, M., Grigoriu, M. and Simiu, E. (2001). Estimation of extreme
wind effects and wind load factors: influence of knowledge uncertainties, Probabilistic
Engineering Mechanics 16, 4, pp. 331340.
Minsky, M. (1987). The Society of Mind (Simon and Schuster, New York).
Mironovkii, L. S. and Slaev, V. A. (2002). Optimal test signals as a solution of the gener-
alized Bulgakov problem, Automation and Remote Control 63, 4, pp. 568577.
Modares, M., Mullen, R. L. and Muhanna, R. L. (2006). Natural frequencies of a structure
with bounded uncertainty, Journal of Engineering Mechanics 132, 12, pp. 1363
1371.
Moens, D. and Vandepitte, D. (2005). A survey of non-probabilistic uncertainty treatment
in finite element analysis, Computer Methods in Applied Mechanics and Engineering
194, pp. 15271555.
Moens, D. and Vandepitte, D. (2006). Recent advances in non-probabilistic approaches
for non-deterministic finite element analysis, Archives of Computational Methods in
M
oller, B. and Beer, M. (2004). Fuzzy Randomness: Uncertainty in Civil Engineering and
Computational Mechanics (Springer, Berlin).
M
oller, B. and Beer, M. (2008). Engineering optimization under uncertainty: Capabilities
of non-traditional models, Computer Methods in Applied Mechanics and Engineering
86, 1112, pp. 10241041.
M
oller, B., Grab, W. and Beer, M. (2000). Fuzzy structural analysis using -level opti-
mization, Computational Mechanics 26, pp. 547565.
Mongin, P. (2000). Does optimization imply rationality? Syntheses 124, 12, pp. 73111.
Moore, R. E. (1962). Interval arithmetic and automatic error analysis in digital computing,
Ph. D. dissertation, Stanford University.
Moore, R. E. (1966). Interval Analysis (Prentice-Hall, Englewood Cliffs, NJ).
Moore, R. E. (1979). Methods and Applications of Interval Analysis (SIAM, Philadelphia,
PA).
Moore, R. E. (2006). Introductory remarks on reliable engineering computing, Reliable
Computing 12, pp. 405408.
Moore, R. E., Kearfott, R. B. and Cloud, M. J. (eds.) (2009). Introduction to Interval
Analysis (SIAM, Philadelphia, PA).
Moreau, J. J. (1966). Quadratic programming in mechanics: One-sided constraints, SIAM
Journal of Control 4, pp. 153158.
Moreau, J. J. (1976). Application of convex analysis to the treatment of elastoplastic
systems, in P. Germain and B. Nayroles (eds.), Applications of Methods of Functional
Analysis to Problems in Mechanics (Springer, Berlin).
Moreau, J. J. (1979). Application of convex analysis to some problems of dry friction, in
H. Zorski (ed.), Trends in Applications of Pure Mathematics to Mechanics, Vol. 2
(Pitman, London).
Morimoto, M., Kaifu, N., Takizawa, Y., Aoki, K. and Sakakibara, O. (1982). Homology de-
sign of large antenna, (Mitsubishi Electric Corporation), pp. 495502, (in Japanese).
Morse, P. M. and Feshbach, H. (1953). Methods of Theoretical Physics (McGraw-Hill, New
York), pp. 10521060.
Moses, F. and Kinser, D. E. (1967). Optimum structural design and failure probability
constraints, AIAA Journal 6, 6, pp. 11521158.
Motro, R. (1992). Tensegrity systems: the state of the art, International Journal of Space
Structures 7, 2, pp. 7583.
Motro, R. (2003). Tensegrity Structural Systems for the Future (Butterworth-Heinemann,
Oxford, U.K.).
Mourelatos, Z. P. and Zhou, J. (2005). Reliability estimation and design with unsufficient
data based on possibility theory, AIAA Journal 43, 8, pp. 16961705.
Moustafa, A. (2009). Critical earthquake load inputs for multi-degree-of-freedom inelastic
structures, Journal of Sound and Vibration 325, 3, pp. 532544.
Muhanna, R. L., Kreinovich, V., S olin, P., Chessa, J., Araiza, R. and Xiang, G. (2006).
Interval finite element methods: New directions, in REC 2006: Reliable Engineering
Conference (Georgia Institute of Technology, Savannah, GA).
Muhanna, R. L. and Mullen, R. L. (2001). Uncertainty in mechanics problems interval
analysis approach, Journal of Engineering Mechanics 127, pp. 557566.
Muhanna, R. L. and Mullen, R. L. (2002). Short review for the latest development
of interval FEM, Available at: http://www.gtrep.gatech.edu/rmuhanna/IFEM-
review.pdf.
Muhanna, R. L. and Mullen, R. L. (2005). Interval methods for reliable computing, in
E. Nikolaidis, D. Ghiocel and S. Singhal (eds.), Engineering Design Reliability Hand-
book, pp. 12.112.24.
Muhanna, R. L., Mullen, R. L. and Zhang, H. (2005). Penalty-based solution for the in-
terval finite-element methods, Journal of Engineering Mechanics 131, 10, pp. 1102
1111.
Muhanna, R. L., Zhang, H. and Mullen, R. L. (2007). Interval finite elements as a basis
for generalized models of uncertainty in engineering mechanics, Reliable Computing
13, 2, pp. 173194.
Mullen, R. L. and Muhanna, R. L. (1999). Bounds of structural response for all possible
loading combinations, Journal of Structural Division, ASCE 125, 1, pp. 98126.
Mullen, R. L. and Muhanna, R. L. (2002). Efficient interval methods for finite element
solution, in J. N. Almhana and V. C. Bhavsar (eds.), High Performance Computing
Systems and Applications (IEEE Press, New York), pp. 161168.
Mungan, I. (2001). Structural engineering and structures from antiquity to the present,
in Proc. IASS Symposium 2001 (International Association for Shell and Spatial
Structures, Nagoya, Japan), pp. 13.
Murota, K. and Ikeda, K. (1991). Critical imperfection of symmetric structures, SIAM
Journal on Applied Mathematics 51, 5, pp. 12221254.
Murotsu, Y., Shao, S. and Watanabe, A. (1994). An application to reliability-based opti-
mization of redundant structures, Structural Safety 16, pp. 133143.
Myers, R. H. and Montgomery, D. C. (1995). Response Surface Methodology: Process and
Product Optimization using Design Experiments (Wiley, New York).
Naishul, A. B. and Svetlitskii, V. A. (1956). Determination of region of possible solutions
of the system of linear differential equations, Journal of Applied Mathematics and
Mechanics 20, 1, pp. 144147, (in Russian).
Nakagiri, S. (1995). A note on finite element interval analysis (Part 1: Linear stiffness
Bibliography 369
equation), Seisan-Kenkyu, Monthly J. Inst. Industrial Sci., The University of Tokyo,

Tokyo 47, 12, pp. 597600.
Nakagiri, S. (1996). A note on finite element interval analysis (Part 2: Linear, undamped
eigenvalue problem), Seisan-Kenkyu, Monthly J. Inst. Industrial Sci., The University
of Tokyo, Tokyo 48, 1, pp. 912.
Nakagiri, S., Hoshi, Y. and Yamada, T. (2003). Interval estimation of time-history re-
sponse without use of direct time integration, in K. J. Bathe (ed.), Proc. 2nd MIT
Conference on Computational Fluid and Solid Mechanics, pp. 509512.
Nakagiri, S. and Suzuki, K. (1996). Interval estimation of eigenvalue problem based on
finite element sensitivity analysis of stiffness estimation, Transactions JSME 62A,
603, pp. 24352439, (in Japanese).
Nakagiri, S. and Suzuki, K. (1997). Interval estimation of eigenvalue problem based on
finite element sensitivity analysis and convex method, JSME International Journal
Series A, Solid Mechanics and Materials Engineering 40, 3, pp. 228233.
Nakagiri, S. and Suzuki, K. (1999). Finite element interval analysis of external loads iden-
tified by displacement input with uncertainty, Computer Methods in Applied Me-
chanics and Engineering 168, pp. 6372.
Nakagiri, S. and Yoshikawa, N. (1996). Finite element interval estimation of convex model,
in Proc. 7th ASCE EMD/STD Joint Specialty Conference on Probabilistic Mechanics
and Structural Reliability (ASCE Press, New York), pp. 278281.
Nakamura, T. and Ohsaki, M. (1988). Sequential optimal truss generator for frequency
ranges, Computer Methods in Applied Mechanics and Engineering 67, 2, pp. 189
209.
Nakamura, T. and Ohsaki, M. (1992). A natural generator of optimum topology of plane
trusses for specified fundamental frequency, Computer Methods in Applied Mechanics
and Engineering 94, 1, pp. 113129.
Nakayama, H. and Hattori, A. (2003). Incremental learning and forgetting in RBF net-
works and SVMs with applications to financial problems, Lecture Notes in Artificial
Intelligence, Vol. 2773 (Springer-Verlag, Berlin).
Nastran (2001). MSC.Nastran Users Manual, Ver 70.5.
Natke, H. G. and Ben-Haim, Y. (eds.) (1997). Uncertainty: Models and Measures
(Akademie Verlag, Berlin).
Natke, H. G. and Soong, T. T. (1993). Topological structural optimization under dynamic
loads, in S. Hernandez and C. A. Brebbia (eds.), Computer-Aided Optimum Design
of Structures III: Optimization of Structural System and Application (Computational
Mechanics and Publications, Southampton, U.K.), pp. 6778.
Neuber, M. (1958). Kerbspannungslehre: Grundlagen f ur genaue Festgkeitsberechnung mit
Ber ucksichtigung von Konstruktionsform und Werkstoff (Springer, Berlin), (in Ger-
man).
Neumaier, A. (1982). Overestimation in linear interval equations, SIAM Journal on Nu-
merical Analysis 24, 1, pp. 207214.
Neumaier, A. (1990). Interval Methods for Systems of Equations (Cambridge University
Press, Cambridge, U.K.).
Neumaier, A. (2004). Clouds, fuzzy sets, and probability intervals, Reliable Computing 10,
pp. 249272.
Neumaier, A. and Pownuk, A. (2007a). Linear systems with large uncertainties with ap-
plication to truss structures, Reliable Computing 13, 2, pp. 149172.
Neumaier, A. and Pownuk, A. (2007b). Worst case analysis of mechanical structures by in-
terval methods, http://www.mat.univie.ac.at/neum/papers.html s0122 and s000.
Nguyen, T. B. (1962). Towards solving some problems of dynamic programming with the
aid of electronic modelling, Automatics and Telemechanics 23, 9, pp. 11301139, (in
Russian).
Nguyen, T. B. (1965). On accumulation of disturbances, Izvestiya Akademii Nauk SSSR
Mekhanika (Communications of the USSR Academy of Sciences, Mechanics) 4, pp.
97100.
Nickel, K. (1977). Interval-analysis, in D. A. H. Jacobs (ed.), The State of the Art in
Numerical Analysis (Academic Press, London).
Nikolaidis, E., Chen, S., Cudney, H. and Haftka, R. T. (2004). Comparison of probability
and possibility for design against catastrophic failure under uncertainty, Journal of
Mechanical Design 126, pp. 386394.
Nikolaidis, E., Ghiocel, D. M. and Singhal, S. (eds.) (2005). Engineering Design Reliability
Handbook (CRC Press, Boca Raton, FL).
Nikolaidis, E. and Haftka, R. T. (2001). Theories of uncertainty for risk assessment when
data is scarce, International Journal for Advanced Manufacturing Systems 4, 1, pp.
4956.
Niordson, F. I. (2001). Early numerical computations in engineering, Applied Mechanics
Reviews 54, 6, pp. R17R19.
Nishihara, O., Asami, T. and Kumamoto, H. (1999a). Design optimization of dynamic
vibration absorber for minimization of maximum amplitude magnification factor:
Consideration of primary system damping by numerical exact solution, Dynamics &
Design Conference 1999, B, pp. 365368, (in Japanese).
Nishihara, O., Asami, T. and Watanabe, S. (1999b). Exact expressions of a dynamic
vibration absorber for minimization of maximum amplitude response: Case 2, hys-
teretic damping absorber, Dynamics & Design Conference 1999, B, pp. 361364,
(in Japanese).
Norton, R. L. (2000). Machine Design (Prentice-Hall, Upper Saddle River, NJ).
Nowak, A. S. and Szerszen, M. M. (eds.) (1998). Reliability and Optimization of Structural
Systems (University of Michigan, Ann Arbor, MI).
Nowak, A. S. and Szerszen, M. M. (eds.) (2000). Reliability and Optimization of Structural
Systems (University of Michigan, Ann Arbor, MI).
Oberguggenberger, M. (2004). Introductory remarks: Mathematical models of uncertainty,
Zeitschrift f
ur angewandte Mathematik und Mechanik (ZAMM) 84, pp. 661665.
Oberguggenberger, M. and Russo, F. (2001). Fuzzy, probabilistic and stochastic modelling
of an elastically bedded beam, in G. de Cosman, T. L. Fine and T. Seidenfell (eds.),
Proc. 2nd International Symposium on Imprecise Probabilities and Their Applica-
tions (Shaker Publishing BV, Cornell University, Ithaca, NY), pp. 293300.
Oden, J. T., Babuska, I., Nobile, F., Feng, Y. and Tempone, R. (2005). Theory and
methodology for estimation and control of errors to modeling, approximation, and
uncertainty, Computer Methods in Applied Mechanics and Engineering 194, pp.
195204.
Oden, J. T., Belytschko, T., Babuska, I. and Hughes, T. Y. R. (2003). Research direc-
tions in computational mechanics, Computer Methods in Applied Mechanics and
Ohsaki, M. (1995). Genetic algorithm for topology optimization of trusses, Computers &
Ohsaki, M. (1997). Simultaneous optimization of topology and geometry of a regular plane
truss, Computers & Structures 66, 1, pp. 6977.
Ohsaki, M. (2000). Optimization of geometrically nonlinear symmetric systems with coin-
cident critical points, International Journal for Numerical Methods in Engineering
48, pp. 13451357.
Bibliography 371
Ohsaki, M. (2001). Random search method based on exact reanalysis for topology opti-
mization of trusses with discrete cross-sectional areas, Computers & Structures 79,
6, pp. 673679.
Ohsaki, M. (2002a). Imperfection sensitivity of optimal symmetric braced frames against
buckling, International Journal of Non-Linear Mechanics 38, pp. 11031117.
Ohsaki, M. (2002b). Maximum loads of imperfect systems corresponding to stable bifur-
cation, International Journal of Solids and Structures 39, pp. 927941.
Ohsaki, M. (2002c). Structural optimization for specified nonlinear buckling load factor,
Japan Journal of Industrial and Applied Mathematics 19, 2, pp. 163179.
Ohsaki, M. (2003a). Optimum design of flexible structures under constraints on strain
energy and asymptotic stability, Computer Methods in Applied Mechanics and En-
gineering 192, pp. 44874496.
Ohsaki, M. (2003b). Sensitivity analysis of an optimized bar-spring model with hill-top
branching, Archive of Applied Mechanics 73, pp. 241251.
Ohsaki, M. (2005). Design sensitivity analysis and optimization for nonlinear buckling
of finite-dimensional elastic conservative structures, Computer Methods in Applied
Ohsaki, M. and Arora, J. S. (1993). A direct application of higher order parametric pro-
gramming techniques to structural optimization, International Journal for Numeri-
cal Methods in Engineering 36, pp. 26832702.
Ohsaki, M., Fujisawa, K., Katoh, N. and Kanno, Y. (1999). Semi-definite programming
for topology optimization of trusses under multiple eigenvalue constraints, Computer
Methods in Applied Mechanics and Engineering 180, pp. 203217.
Ohsaki, M. and Ikeda, K. (2006). Imperfection sensitivity analysis of hill-top branching
with many symmetric bifurcation points, International Journal of Solids and Struc-
tures 43, 16, pp. 47044719.
Ohsaki, M. and Ikeda, K. (2007a). Stability and Optimization of Structures Generalized
Sensitivity Analysis, Mechanical Engineering Series (Springer, New York).
Ohsaki, M. and Ikeda, K. (2007b). Worst imperfection of an arch-type truss with multiple
member buckling at limit point, Journal of Structural and Constructional Engineer-
ing, Architectural Institute of Japan , 611, pp. 7177, (in Japanese).
Ohsaki, M. and Ikeda, K. (2009). Imperfection sensitivity of degenerate hilltop branching
points, International Journal of Non-linear Mechanics 44, pp. 324336.
Ohsaki, M. and Kanno, Y. (2003). Form-finding of cable domes with specified stresses by
using nonlinear programming, in Proc. IASS-APCS Symposium 2003 (International
Association for Shell and Spatial Structures, Taipei).
Ohsaki, M. and Kanno, Y. (2007). Semidefinite programming for engineering applications,
in J. S. Arora (ed.), Optimization of Structural and Mechanical Systems (World
Scientific, Singapore), pp. 541567.
Ohsaki, M. and Katoh, N. (2005). Topology optimization of trusses with stress and local
constraints on nodal stability and member intersection,, Structural and Multidisci-
plinary Optimization 29, pp. 190197.
Ohsaki, M., Kinoshita, T. and Pan, P. (2007). Multiobjective heuristic approaches to
design of frames with standard sections, Earthquake Engineering and Structural Dy-
namics 36, 11, pp. 14811495.
Ohsaki, M., Nakamura, T. and Isshiki, Y. (1998a). Shape-size optimization of plane trusses
with designers preference, Journal of Structural Engineering 124, 11, pp. 13231330.
Ohsaki, M. and Nishiwaki, S. (2005). Shape design of pin-jointed multistable compliant
mechanism using snapthrough behavior, Structural and Multidisciplinary Optimiza-
tion 30, pp. 327334.
Ohsaki, M., Uetani, K. and Takeuchi, M. (1998b). Optimization of imperfection-sensitive

symmetric systems for specified maximum load factor, Computer Methods in Applied
Ohsaki, M., Zhang, J. Y. and Elishakoff, I. (2008a). Optimization and anti-optimization
of forces in tensegrity structures, in Proc. IASS Symposium 2008 (International
Association for Shell and Spatial Structures, Acapulco, Mexico).
Ohsaki, M., Zhang, J. Y. and Ohishi, Y. (2008b). Force design of tensegrity structures by
enumeration of vertices of feasible region, International Journal of Space Structures
23, 2, pp. 117126.
Ohsaki, M. and Zhang, T. Y. (2006). Stability conditions of prestressed pin-jointed struc-
tures, International Journal of Non-Linear Mechanics 41, pp. 11091117.
Ohta, Y. and Haftka, R. T. (1997). Optimization of FRP laminated plates
under uncertainty by fuzzy-based and genetic algorithms, in Proc. 38th
AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials
Conference and Exhibit, Vol. Part 2 (Kissimee, FL), pp. 13251330.
Olhoff, N. and Rasmussen, S. H. (1977). On single and bimodal optimum buckling loads of
clamped columns, International Journal of Solids and Structures 13, pp. 605614.
Ottl, D. (1987). Beschreibung linear-viskoelastischer schwingungssysteme mit hilfe von
inner variablen, Zeitschrift f
ur angewandte Mathematik und Mechanik (ZAMM) 67,
4, pp. T122T124, (in German).
Paez, T. L. (1981). Conservatism in least favorable response analysis and testing, in The
Shock and Vibration Bulletin, Vol. 51 (Naval Research Laboratory, Wasington D.C.),
pp. 93109.
Pal, N. R. (1999). On quantification of different facets of uncertainty, Fuzzy Sets and
Systems 107, pp. 8191.
Palassopoulos, G. V. (1991). Reliability-based design of imperfection sensitive structures,
Journal of Engineering Mechanics Division, ASCE 117, 6, pp. 12201240.
Palmov,
V. A. (1964). An elastic plane with hole of random shape, Trudy Leningradskogo
Politekhnicheskogo Instituta , 235, (in Russian).
Pan, P., Ohsaki, M. and Tagawa, H. (2007). Shape optimization of H-beam flange for
maximum plastic energy dissipation, Journal of Structural Engineering 133, 8, pp.
11761179.
Panagiotopoulos, P. D. (1976). Convex analysis and unilateral static problems, Ingenieur
Archiv 45, pp. 5568.
Panagiotopoulos, P. D. (1985). Inequality Problems in Mechanics and Applications: Convex
and Nonconvex Energy Functions (Birkh auser, Boston).
Panovko, Y. G. and Gubanova, I. I. (1965). Stability and Oscillations of Elastic Systems,
Paradoxes, Fllacies and New Concepts (Consultants Bureau, New York), pp. 215
231.
Pantelides, C. P. (1996a). Buckling and postbuckling of stiffened elements with uncertainty,
Thin-Walled Structures 26, 1, pp. 117.
Pantelides, C. P. (1996b). Stability of elastic bars on uncertain foundation using a convex
model, International Journal of Solids and Structures 33, 9, pp. 12571269.
Pantelides, C. P. and Booth, B. C. (2000). Computer-aided design of optimal structures
with uncertainty, Computers & Structures 74, pp. 293307.
Pantelides, C. P. and Ganzerli, S. (1997). Optimal design of structures under uncertain
loads using convex models, in S. Hern andez and C. A. Brebbia (eds.), Computer
Aided Optimum Design of Structures V (Computational Mechanics Publications,
Southampton, U.K.), pp. 173182.
Pantelides, C. P. and Ganzerli, S. (2001). Comparison of fuzzy set and convex model
Bibliography 373
theories in structural design, Mechanical Systems and Signal Processing 15, 3, pp.
499511.
Pantelides, C. P. and Tzan, S.-R. (1996). Convex models for seismic design of structures
1: Analysis, Earthquake Engineering and Structural Dynamics 25, 9, pp. 927944.
Papadopoulos, V., Deodatis, G. and Papadrakakis, M. (2005). Flexibility-based upper
bound on the response variability of simple beams, Computer Methods in Applied
Mechanics and Engineering 194, 1216, pp. 13851404.
Papoulis, A. (1970). Maximum response with input energy constraints and matched filter
principle, IEEE Transaction on Circuit Theory CT-17, 2, pp. 175182.
Parimi, S. R. and Cohn, M. Z. (1975). Optimal criteria in probabilistic structural design,
in Z. Mr oz and A. Sawczuk (eds.), Proc. IUTAM Symposium on Optimization in
Structural Design (Springer Verlag, Berlin).
Parimi, S. R. and Cohn, M. Z. (1978a). Optimal solutions in probabilistic structural design:
Part I, theory, 2, 1, pp. 4772.
Parimi, S. R. and Cohn, M. Z. (1978b). Optimal solutions in probabilistic structural design:
Part II, applications, 2, 1, pp. 7392.
Park, J. and Sandberg, I. W. (1991). Universal approximation using radial-basis-function
networks, Neural Computation 3, pp. 246257.
Parkinson, C. N. (1959). Parkinsons Law on the Pursuit of Progress (John Murray, Lon-
don).
Parnes, R. and Beltzer, A. I. (1986). A higher-order boundary perturbation method for
asymmetric dynamic problems in solids I. general formulation, International Jour-
nal of Solids and Structures 22, 11, pp. 11771187.
Peek, R. and Triantafyllidis, N. (1992). Worst shapes of imperfections for space trusses with
many simultaneously buckling modes, International Journal of Solids and Structures
29, 19, pp. 23852402.
Peterson, R. E. (1974). Stress Concentration Design Factors, Charts and Relations Useful
on Making Strength Calculations for Machine Parts and Structural Elements (Wiley,
New York).
Philappacopoulos, A. J. (1980). Critical excitation for linear and non-linear structural
systems, Ph. D. dissertation, Polytechnic Institute of New York.
Pierson, B. L. and Genalo, L. J. (1977). Minimum weight design of a rectangular panel
subjected to a flutter speed constraint, Computer Methods in Applied Mechanics and
Pietrzak, J. (1996). An alternative approach to optimization of structures prone to insta-
bility, Structural Optimization 11, pp. 8894.
Plaut, R. H. (1970). On minimizing the response of structures to dynamic loading, Journal
of Applied Mathematics and Physics 21, pp. 10041006.
Plaut, R. H. (1971). Structural optimization of a panel flutter problem, AIAA Journal 9,
1, pp. 182184.
Pochtman, Y. M. and Khariton, L. E. (1976). Optimal design of structures with reliabil-
ity taken into account, in Stroitelnaya Mekhanikai Raschet Sooruzhenii (Structural
Mechanics and Structural Analysis), No. 6, pp. 815, (in Russian).
Popova, E. D., Datcheva, M., Iankov, R. and Schanz, T. (2003). Mechanical models with
interval parameters, in K. Guerlebeck, L. Hempel and C. Koenke (eds.), Proc. 16th
International Conference on the Applications of Computer Science and Mathematics
in Architecture and Civil Engineering (Bauhaus University, Weimar, Germany).
Popova, E. D., Datcheva, M., Iankov, R. and Schanz, T. (2004). Sharp bounds for strains
and stresses of uncertain mechanical models, in I. Lirkov et al. (ed.), Lecture Notes
in Computer Science 2907 (Springer, Berlin), pp. 262269.
Popplewell, N. and Youssef, N. A. N. (1979). A comparison of maximax response estima-

tion, Journal of Sound and Vibration 62, pp. 339352.
Pothisiri, T. and Hjelmstad, K. D. (2002). Strategy for finding a near-optimal measurement
set for parameter estimation from modal response, Journal of Sound and Vibration
257, 1, pp. 89106.
Pownuk, A. (2000). Optimization of mechanical structures using interval analysis, Com-
puter Assisted Mechanics and Engineering Sciences 7, 4, pp. 699705.
Pownuk, A. (2004). Efficient method of solution of large scale engineering problems with
interval parameters based on sensitivity analysis, in R. L. Muhanna and R. L. Mullen
(eds.), Proc. NSF Workshop on Reliable Engineering Computing (Savannah, GA),
http://www.gtsav.gatech.edu/rec/recworkshop/presentations/presentations.html.
Prager, W. (1972). Conditions for structural optimality, Computers & Structures 2, pp.
833840.
Prager, W. and Taylor, J. E. (1968). Problem of optimal structural design, Journal of
Pu, Y., Das, P. K. and Faulkner, D. (1997). A strategy for reliability-based optimization,
Engineering Structures 19, 3, pp. 276282.
Pugsley, A. G. (1966). The Safety of Structures (Edward Arnold, London).
Puig, V., Saludes, Y. and Quevedo, J. (2003). Worst-case simulation of discrete linear
time-invariant dynamic systems, Reliable Computing 9, 4, pp. 251290.
Putresza, J. T. and Kolakowski, P. (2001). Sensitivity analysis of frame structures: Vir-
tual distortion method approach, International Journal for Numerical Methods in
Qiu, Z. P. (2003). Comparison of static response of structures using convex models and in-
terval analysis method, International Journal for Numerical Methods in Engineering
56, 12, pp. 17351753.
Qiu, Z. P. (2005). Convex models and interval analysis method to predict the effect on
buckling of composite structures, Computer Methods in Applied Mechanics and En-
gineering 94, pp. 21752189.
Qiu, Z. P., Chen, S.-H. and Elishakoff, I. (1995a). Natural frequencies of structures with
uncertain but nonrandom parameters, Journal of Optimization Theory and Appli-
cations 86, 3, pp. 669683.
Qiu, Z. P., Chen, S.-H. and Elishakoff, I. (1996a). Bounds of eigenvalues for structures with
an interval description of uncertain-but-non-random parameters, Chaos, Solitons &
Fractals 7, 3, pp. 425434.
Qiu, Z. P., Chen, S.-H. and Elishakoff, I. (1996b). Non-probabilistic eigenvalue problem
for structures with uncertain parameters via interval analysis, Chaos, Solitons &
Fractals 7, 3, pp. 303308.
Qiu, Z. P., Chen, S.-H. and Jia, H. G. (1995b). The Rayleigh quotient iteration method
for computing eigenvalue bounds of structures with bounded uncertain parameters,
Computers & Structures 55, 2, pp. 221227.
Qiu, Z. P., Chen, S.-H. and Song, D. (1996c). The displacement bound estimation for
structures with an interval description of uncertain parameters, Communications in
Numerical Methods in Engineering 12, pp. 111.
Qiu, Z. P., Chen, S.-H. and Wang, X. J. (2004a). Criterion on the non-probabilistic robust
reliability of structures, Yisual Lixue Xuebao (Chinese Journal of Computational
Mechanics) 21, 1, pp. 16, (in Chinese).
Qiu, Z. P., Chen, S. Q. and Wang, X. J. (2004b). Criterion of the non-probabilistic robust
reliability for structures, Chinese Journal of Computational Mechanics 21, 1, pp.
16, (in Chinese).
Bibliography 375
Qiu, Z. P., Elishakoff, I. and Starnes, Jr., J. H. (1996d). The bound set of possible eigen-
values of structures with uncertain-but-non-random parameters, Chaos, Solitons &
Fractals 7, 11, pp. 18451857.
Qiu, Z. P. and Gu, Y. (1996). Extension of convex models and its improvement on the
approximate solution, Acta Mechanica Sinica 12, 4, pp. 349357.
Qiu, Z. P. and Hu, J. X. (2008). Two non-probabilistic set-theoretical models to predict
the transient vibration of cross-ply plates with uncertainty, Applied Mathematical
Modelling 32, 12, pp. 28722887.
Qiu, Z. P., Hu, J. X., Yang, J. L. and Lu, Q. S. (2008a). Exact bounds for the sensitivity
analysis of structures with uncertain-but-bounded parameters, Applied Mathematical
Modelling 32, 6, pp. 11431157.
Qiu, Z. P., Ma, L. H. and Wang, X. J. (2006a). Ellipsoidal-bound convex model for the
non-linear buckling of a column with uncertain initial imperfections, International
Journal of Non-linear Mechanics 41, 8, pp. 919925.
Qiu, Z. P., Ma, Y. and Wang, X. J. (2004c). Comparison between non-probabilistic interval
analysis method and probabilistic approach in static response problem of structures
with uncertain-but-bounded parameters, Communications in Numerical Methods in
Qiu, Z. P., Ma, Y. and Wang, X. J. (2004d). Interval analysis method for vibration of com-
posite laminate with uncertain parameters, Journal of Beijing University of Aero-
nautics and Astronautics 30, 7, pp. 682685.
Qiu, Z. P., M uller, P. C. and Frommer, A. (2001). Ellipsoidal set-theoretic approach for
stability of linear state-space models with interval uncertainty, Mathematics and
Computers in Simulation 57, 1-2, pp. 4559.
Qiu, Z. P., M uller, P. C. and Frommer, A. (2004e). The new non-probabilistic criterion of
failure for dynamical systems based on convex methods, Mathematical and Computer
Modelling 40, 12, pp. 201215.
Qiu, Z. P. and Wang, X. J. (2003). Comparison of dynamic response of structure with
uncertain-but-bounded non-probabilistic interval analysis, International Journal of
Solids and Structures 40, 20, pp. 54235439.
Qiu, Z. P. and Wang, X. J. (2005). Two non-probabilistic set-theoretical models for dy-
namic response and buckling failure measures of bars with unknown-but-bounded
initial imperfections, International Journal of Solids and Structures 42, 34, pp.
10391054.
Qiu, Z. P. and Wang, X. J. (2006). Interval analysis method and convex models for im-
pulsive response of structures with uncertain-but-bounded external loads, Acta Me-
chanica Sinica/Lixue Xuebao 22, 6, pp. 265276, (in Chinese).
Qiu, Z. P. and Wang, X. J. (2008). Set-Theoretic Convex Methods for Problems in Struc-
tural Mechanics with Uncertainties (Science Press, Beijing), (in Chinese).
Qiu, Z. P., Wang, X. J. and Chen, J. (2006b). Exact bounds for the static response set of
structures with uncertain-but-bounded parameters, International Journal of Solids
and Structures 43, pp. 65746593.
Qiu, Z. P., Yang, D. and Elishakoff, I. (2008b). Combination of structural reliability and
interval analysis, Acta Mechanica Sinica 24, pp. 6167.
Qiu, Z. P., Yang, D. and Elishakoff, I. (2008c). Probabilistic interval reliability of structural
systems, International Journal of Solids and Structures 45, pp. 28502860.
Rackwitz, R. (2001). Reliability analysis : A review and some perspectives, Structural
Safety 23, pp. 365395.
Rackwitz, R., Augusti, G. and Bori, A. (eds.) (1995). Reliability and Optimization of
Structural Systems (Chapman & Hall, London).
Rackwitz, R. and Cuntze, R. (1987). Formulations of reliability-oriented optimization,

Engineering Optimization 11, pp. 6976.
Rackwitz, R. and Thoft-Christensen, P. (eds.) (1991). Reliability and Optimization of
Structural Systems (Springer, Berlin).
Ramm, E., Bletzinger, K.-U. and Reitinger, R. (1993). Shape optimization of shell struc-
tures, Bulletin of International Association for Shell and Spatial Structures 34, 2,
pp. 103121.
Rao, C. R. and Mitra, S. K. (1971). Generalized Inverse of Matrices and Its Applications
(John Wiley & Sons).
Rao, S. S. (1987a). Description and optimum design of fuzzy mechanical systems, Journal
of Mechanisms, Transmission, Automatic Design 109, pp. 126132.
Rao, S. S. (1987b). Multiobjective optimization of fuzzy structural systems, International
Rao, S. S. (1987c). Optimum design of structures in a fuzzy environment system, AIAA
Journal 25, 12, pp. 16331636.
Rao, S. S. and Berke, L. (1997). Analysis of uncertain structural systems using interval
analysis, AIAA Journal 35, 4, pp. 727735.
Rao, S. S., Chen, L. and Mulkay, E. (1998). Unified finite element method for engineering
systems with hybrid uncertainties, AIAA Journal 36, 7, pp. 12911299.
Rao, S. S. and Majumder, L. (2008). Optimization of aircraft wing for gust loads: Interval
analysis-based approach, AIAA Journal 46, 3, pp. 723732.
Rao, S. S., Sundararaju, K., Prakash, B. G. and Balakrishna, C. (1992a). Fuzzy goal
programming approach for structural optimization, AIAA Journal 30, 5, pp. 1425
1432.
Rao, S. S., Sundararaju, K., Prakash, B. G. and Balakrishna, C. (1992b). Multiobjective
fuzzy optimization techniques for engineering design, Computers & Structures 42,
1, pp. 3744.
Reeves, C. (1995). Modern Heuristic Techniques for Combinatorial Problems (McGraw-
Hill, New York).
Reitinger, R., Bletzinger, K.-U. and Ramm, E. (1994). Shape optimization of buckling
sensitive structures, Computing Systems in Engineering 5, pp. 6575.
Reitinger, R. and Ramm, E. (1995). Buckling and imperfection sensitivity in the optimiza-
tion of shell structures, Thin-Walled Structures 23, pp. 159177.
Riks, E. (1998). Buckling analysis of elastic structures: a computational approach, in
Advances in Applied Mechanics, Vol. 34 (Academic Press, San Diego, CA), pp. 1
76.
Ringertz, U. T. (1994). On structural optimization with aeroelastic constraints, Structural
Rohn, J. (2006). A handbook of results on interval linear problems, Available at:
http://www.cs.cas.cz/rohn/publist/handbook.zip.
Roitenberg, Y. N. (1958). On accumulation of disturbances in nonstationary linear systems,
Doklady Akademii Nauk SSSR (Proceedings of the USSR Academy of Sciences) 121,
2, pp. 221224, (in Russian).
Rosenblatt, F. (1962). Principles of Neurodynamics: Perceptrons and the Theory of Brain
Machines (Spartan Books, Washington D.C.).
Ross, T. J., Booker, J. M. and Parkinson, W. J. (eds.) (2002). Fuzzy Logic and Probability
Applications: Bridging the Gap (SIAM, Philadelphia, PA).
Roux, W. J., Stander, N. and Haftka, R. T. (1998). Response surface approximation for
structural optimization, International Journal for Numerical Methods in Engineering
42, pp. 517534.
Bibliography 377
Royset, J. O., Der Kiureghian, A. and Polak, E. (2003). Algorithms for variability-based
optimal design, in S. Jendo and K. Doloinski (eds.), Reliability-Based Design and
Optimization, Lecture Notes 16, Institute of Fundamental Technological Research
(Warsaw, Poland), pp. 211240.
Rozvany, G. I. N. (1976). Optimal Design of Flexural Systems (Pergamon Press).
Rozvany, G. I. N. (1997). Aims, scope, basic concepts and methods of topology opti-
mization, in G. I. N. Rozvany (ed.), Topology Optimization in Structural Mechanics,
CISM Courses and Lectures, No. 374 (Springer, Vienna).
Ruiz, P. and Penzien, J. (1969). Probabilistic Study of the Behavior of Structures During
Earthquakes (Earthquake Engineering Research Center, University of California,
Berkeley).
Rump, S. M. (1992). On the solution of interval linear systems, Computing 47, pp. 337
353.
Russell, B. (1985). The Philosophy of Logical Atomism (Open Court, LaSalle, IL).
Rustem, B. and Howe, M. (2002). Algorithms for Worst-Case Design and Applications to
Risk Management (Princeton University Press, Princeton, NJ).
Rzhanitsyn, A. R. (1947). Determination of Safety Factor in Construction, Vol. 8 (Stroi-
telnaya Promyshlennost (Building Industry)), (in Russian).
Sadek, I. S., Sloss, J. M., Adali, S. and Bruch, Jr., J. C. (1993). Nonprobabilistic mod-
elling of dynamically loaded beams under uncertain excitations, Mathematical and
Computer Modelling 18, pp. 5967.
Sage, A. P. (1968). Optimum Systems Control (Prentice Hall, Englewood Cliffs, NJ).
Sakata, S., Ashida, F. and Zako, M. (2003). Structural optimization using Kriging ap-
proximation, Computer Methods in Applied Mechanics and Engineering 192, pp.
923939.
Salle, J. L. and Lefschetz, S. (1961). Stability by Liapunovs Direct Method (Academic
Press, New York).
Sanayei, M., Onipede, O. and Babu, S. R. (1992). Selection of noisy measurement locations
for error reduction in static parameter identification, AIAA Journal 30, 9, pp. 2299
2309.
Sangalli, A. (1998). The Importance of Being Fuzzy and Other Insights from the Border
between Math and Computers (Princeton University Press, Princeton, NJ).
Sarkar, A. and Manohar, C. S. (1998). Critical seismic vector random excitations for
multiplly supported structures, Journal of Sound and Vibration 212, 3, pp. 525
546.
Sarma, K. and Adeli, H. (2002). Life-cycle cost optimization of steel structures, Interna-
tional Journal for Numerical Methods in Engineering 55, pp. 14511462.
Sato, M. (2006). The Honda Myth: The Genius & His Wake (Vertical, Inc., New York).
Savin, G. N. (1961). Stress Concentration Around Holes (Pergamon, Oxford, U.K.).
Savoia, M. (2002). Structural reliability analysis through fuzzy number approach, with
application to stability, Computers & Structures 80, 12, pp. 10871102.
Saxena, V. (2003). Interval Finite Element Analysis for Load Pattern and Load Combina-
tion, M. Sci. Thesis, Georgia Institute of Technology, Atlanta, GA.
Sayles, R. S. (1982). The profile as a random process, in T. R. Thomas (ed.), Rough
Surfaces (Longman, London), pp. 91118.
Schek, H. J. (1974). The force density method for form finding and computation of general
networks, Computer Methods in Applied Mechanics and Engineering 3, pp. 115134.
Schiehlen, W. and Hu, B. (1997). Amplitude bounds of linear vibration responses,
Zeitschrift f
ur angewandte Mathematik und Mechanik (ZAMM) 76, pp. 453454.
Schmidt, R. (1985). Towards resurrecting the original Rayleigh method, Industrial Math-
ematics 35, 1, pp. 6973.

Schmitendorf, W. E., Jabbari, F. and Yang, J. N. (1994). Robust control techniques for
buildings under earthquake excitations, Earthquake Engineering and Structural Dy-
namics 23, pp. 539552.
Schranz, C., Krenn, B. and Mang, H. A. (2006a). Conversion from imperfection-sensitive
into imperfection-insensitive elastic structures I: Theory, Computer Methods in Ap-
plied Mechanics and Engineering 195, pp. 14221457.
Schranz, C., Krenn, B. and Mang, H. A. (2006b). Conversion from imperfection-sensitive
into imperfection-insensitive elastic structures II: Numerical investigation, Computer
Methods in Applied Mechanics and Engineering 195, pp. 14581479.
Schueller, G. I. (1996). Engineering applications to stochastic mechanics achievements
and prospects, in A. Naess and S. Krenk (eds.), Proc. IUTAM Symposium on Ad-
vances in Nonlinear Stochastic Mechanics (Kluwer Academic Publishers, Dordrecht,
Netherlands), pp. 383402.
Schueller, G. I. (1998). Structural reliability recent advances, in N. Shiraishi, M. Shi-
nozuka and Y. K. Wen (eds.), Structural Safety and Reliability (Balkama, Rotterdam,
Netherlands), pp. 335.
Schueller, G. I. and Stix, R. (1987). A critical appraisal of methods to determine failure
probabilities, Structural Safety 4, pp. 293309.
Schweppe, F. C. (1973). Uncertain Dynamic Systems (Prentice Hall, Englewood Cliffs,
NJ).
Sevin, E. (1957). Min-max solutions for the linear mass-spring system, Journal of Applied
Sevin, E. and Pilkey, W. D. (1967). Minmax response problems of dynamic systems and
computational solution techniques, Shock and Vibration Bulletin 36, 5, pp. 6976.
Sexmith, R. G. and Nelson, M. I. (1969). Limitations on application of probabilistic con-
cepts, ACI Journal 66, 10, pp. 823828.
Sharuz, S. M. (1996). Bounded-input bounded-output stability of damped linear strings,
Journal of Sound and Vibration 191, pp. 993998.
Sharuz, S. M. and Krishna, L. G. (1996). Bounded displacement of a damped non-linear
string, Journal of Sound and Vibration 193, pp. 11151121.
Shea, K., Fest, E. and Smith, I. F. C. (2002). Developing intelligent tensegrity structures
with stochastic search, Advanced Engineering Informatics 16, 1, pp. 2140.
Sherman, J. and Morrison, W. J. (1950). Adjustment of an inverse matrix corresponding
to changes in one element of a given matrix, Annals of Mathematical Statistics 21,
pp. 124127.
Shih, C. J., Chi, C. C. and Hsiao, J. H. (2003). Alternative -level cuts methods for
optimum structural design with fuzzy resources, Computers & Structures 81, pp.
25792587.
Shinozuka, M. (1970). Maximum structural response to seismic excitation, Journal of
Engineering Mechanics Division, ASCE 96, pp. 729738.
Shinozuka, M. (1987). Structural response variability, Journal of Engineering Mechanics
Division, ASCE 113, 6, pp. 825843.
Shinozuka, M. and Deodatis, G. (1988). Response variability of stochastic finite element
systems, Journal of Engineering Mechanics 114, pp. 499519.
Shiraishi, N. and Furuta, H. (1985). Assessment of structural durability with fuzzy sets, in
NSF Workshop on Civil Engineering Application of Fuzzy Sets (Purdue University,
West Lafayette, IN).
Shirk, M. H., Hertz, T. H. and Weisshaar, T. A. (1986). Aeroelastic tailoring: Theory,
practice, and promise, Journal of Aircraft 23, 1, pp. 618.
Bibliography 379
Simiu, E. and Heckert, A. (1996). Extreme wind distribution fails: A peak over threshold
approach, Journal of Structural Engineering 122, 5, pp. 539547.
Skalna, I. (2003). Methods for solving systems of linear equations of structural mechanics
with interval parameters, Computer Assisted Mechanics and Engineering Science
10, pp. 281293.
Skalna, I. (2006). A method for outer solution of systems of linear equations depending
linearly on interval parameters, Reliable Computing 12, 2, pp. 107120.
Smallwood, D. O. (1972a). An extension of transient vibration technique using least favor-
able responses, Tech. Rep. Report SC-RR-72 0735, Sandia National Laboratories,
Albuquerque, NM.
Smallwood, D. O. (1972b). A transient vibration test technique using least favorable re-
sponses, Tech. Rep. Report SC-DR-72 0897, Sandia National Laboratories, Albu-
querque, NM.
Smallwood, D. O. (1973). A transient vibration test techniques and the method of least
favorable responses, in The Shock and Vibration Bulletin, Vol. 43 (Naval Research
Laboratory, Wasington D.C.), pp. 151164.
Smallwood, D. O. (1989). Personal communication to Isaac Elishakoff, March 23.
Sniedovich, M. (2006). Eureka! info-gap is worst case analysis (maximum) in disguise!
Working Paper No. MS-2-06, Department of Mathematics and Statistics, The Uni-
versity of Melbourne, Australia.
Sniedovich, M. (2007). The art and science of modeling decision-making under severe
uncertainty, Decision Making in Manufacturing and Services 1, 12, pp. 111136.
Sniedovich, M. (2008a). From Shakespeare to Wald: Modeling worst-case analysis in the
face of severe uncertainty, Decision Point 22, pp. 89.
Sniedovich, M. (2008b). The mighty Maximin, Working Paper No. MS-02-08, Depart-
ment of Mathematics and Statistics, The University of Melbourne, available at:
http://maximin.moshe-online.com.
Sniedovich, M. (2008c). Walds maximin model: a treasure in disguise! Journal of Risk
Finance 9, 3, pp. 287291.
Sniedovich, M. (2009). Personal communication, January 10.
Sobieszczanski-Sobieski, J., James, B. and Dovi, A. (1985). Structural optimization by
multilevel decomposition, AIAA Journal 23, pp. 17751782.
Soize, C. (2005). A comprehensive overview of a non-parametric probabilistic approach of
model uncertainties for predictive models in structural dynamics, Journal of Sound
and Vibration 288, 3, pp. 623652.
Sokal, A. and Bricmont, J. (1998). Intellectual Impostures (Profile Books, London), pub-
lished in the USA as Fashionable Nonsense.
Soong, T. T. (1988). Stochastic structural dynamics: Research versus practice, in S. V. Sri
Namachchivaya, H. H. Hilton and Y. K. Wen (eds.), Stochastic Structural Dynamics
(University of Illinois, Urbana-Champaign, IL), pp. 289297.
Sophie, Q. C. (2000). Comparing Probabilistic and Fuzzy Set Approaches for Design in the
Presence of Uncertainty, Ph. D. dissertation, Virginia Polytechnical Institute and
State University, Blackburg, VA.
Srensen, J. D. and Frangopol, D. M. (eds.) (2006). Advances in Reliability and Optimiza-
tion of Structural Systems (Taylor & Francis, London).
Spunt, L. (1971). Optimum Structural Design (Prentice-Hall, Englewood Cliffs, NJ).
Srinivasan, S., Corotis, R. and Ellingwood, B. (1992). Generation of critical stochastic
earthquakes, Earthquake Engineering and Structural Dynamics 21, pp. 275288.
Srinivasan, S., Ellingwood, B. and Corotis, R. (1991). Critical base excitations of structural
systems, Journal of Engineering Mechanics 117, 6, pp. 14031442.
Stolpe, M. and Svanberg, K. (2001). An alternative interpolation scheme for minimum com-
pliance topology optimization, Structural and Multidisciplinary Optimization 22, pp.
116124.
Stoyanov, J. (1987). Counterexamples in Probability (Wiley, Chichester, U.K.).
Stroud, W. J., Krishnamurthy, T. and Smith, S. A. (2001). Probabilistic and possi-
bilistic analyses of the strength of a bonded joint, in Proc. 42nd AIAA/ASME/
ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference and
Exhibit, Paper-AIAA-2001-1238 (Seattle, WA).
Strutt, J. W. (1877). Theory of Sound (Macmillan, London), (reprinted Dover Publica-
tions, New York, 1945, pp. 112113).
Sunaga, T. (1958). Theory of interval algebra and its application to numerical analysis, in
RAAG Memoirs, Vol. 2 (Gakujutu Bunken Fukyu-kai, Tokyo), pp. 2944, (Available
at http://www.cs.utep.edu/interval-comp/sunaga.pdf).
Suzuki, K. and Otsubo, H. (1995). The structural optimization under uncertainty, in Com-
putational Mechanics 95, Theory and Applications, Proc. International Conference
on Computational Engineering Science (Hawaii).
Svanberg, K. (1984). On local and global minima in structural optimization, in E. Atrek,
R. H. Gallagher, K. M. Ragsdel and O. C. Zienkiewicz (eds.), New Directions in
Structural design, Vol. 34 (John Wiley & Sons, Chichester, U.K.), pp. 327341.
Svanberg, K. (1994). On the convexity and concavity of compliances, Structural Optimiza-
tion 7, pp. 4246.
Svetlitskii, V. A. (1958). Determination of initial data of a perturbed motion, in G. A.
Nikolaev (ed.), Raschety na Prochnost v Mashinostroenii (Strength Analysis in Me-
chanical Engineering), pp. 234245, (in Russian).
Svetlitskii, V. A. (1967). Determination of maximally possible reactions in dynamic sys-
tems under excitations that are bounded by their modulus, Inzhenernyi Zhurnal
Mekhanika Tverdogo Tela (Engineering Journal of Mechanics of Solids) 3, pp. 161
165, (in Russian).
Svetlitskii, V. A. (2003). Statistical Dynamics and Reliability Theory for Mechanical Sys-
tems (Springer, Berlin).
Svetlitskii, V. A. and Prokhorov, V. N. (1970). Vibrations of a beam under the moving
random force that is bounded by its modulus, Izvestiya VUZOV, Mashinostroenie
(Proceedings of the Institution of Higher Learning, Mechanical Engineering) 6, pp.
2529, (in Russian).
Svetlitskii, V. A. and Stasenko, I. V. (1967). Beam on elastic foundation under the loading
that is bounded in modulus, Prikladnaya Mekhanika (Applied Mechanics) 3, 7, pp.
6569, (in Russian).
Switzky, H. (1964). Minimum weight design with structural reliability, in Proc. AIAA 5th
Annual Structural and Materials Conference, pp. 316322.
Takagi, J. and Ohsaki, M. (2004). Design of lateral braces for columns considering critical
imperfection of buckling, Int. J. Structural Stability and Dynamics 4, 1, pp. 6988.
Takewaki, I. (2002a). Robust building stiffness design for variable critical excitations, Jour-
nal of Structural Engineering 128, 12, pp. 15651574.
Takewaki, I. (2002b). Seismic critical excitation method for robust design: A review,
Journal of Structural Engineering 128, 5, pp. 665672.
Takewaki, I. (2006a). Critical Excitation Methods in Earthquake Engineering (Elsevier,
Amsterdam, Netherlands).
Takewaki, I. (2006b). Probabilistic critical excitation method for energy input rate, Journal
of Engineering Mechanics 132, 9, pp. 9901000.
Taleb, N. (2001). Fooled by Randomness: The Hidden Role of Chance in the Markets and
Bibliography 381
Life (Random House, New York).

Templeman, A. B. (1988). Some reservations about reliability-based design, in D. M. Fran-
gopol (ed.), New Directions in Structural System Reliability (University of Colorado),
pp. 265276.
Thimbleby, H. (1993). Computerized Parkinsons law, Computer Control Engineering 4,
5, pp. 197198.
Thoft-Christensen, P. (ed.) (1987). Reliability and Optimization of Structural Systems
(Springer, Berlin).
Thoft-Christensen, P. (ed.) (1988). Reliability and Optimization of Structural Systems
(Springer, Berlin).
Thoft-Christensen, P. (2006). Risk based structural optimization, in P. Thoft-Christensen
(ed.), Papers, Aalborg University, Vol. 3: 19901994 (Aalborg, Denmark), pp. 679
682.
Thoft-Christensen, P. and Ishikawa, H. (eds.) (1993). Reliability and Optimization of Struc-
tural Systems (Elsevier, Amsterdam, Netherlands).
Thomas, H. L., Sepulveda, A. E. and Schmit, L. A. (1992). Improved approximations for
control augmented structural synthesis, AIAA Journal 30, 12, pp. 171179.
Thompson, J. M. T. and Hunt, G. W. (1973). A General Theory of Elastic Stability (John
Wiley, New York).
Thompson, J. M. T. and Hunt, G. W. (1974). Dangers of structural optimization, Engi-
neering Optimization 1, pp. 99110.
Thompson, J. M. T. and Hunt, G. W. (1984). Elastic Instability Phenomena (Wiley,
Chichester, U.K.).
Thompson, J. M. T. and Schorrock, P. A. (1975). Bifurcation instability of an atomic
lattice, Journal of Mechanics of Physics and Solids 23, 1, pp. 2137.
Thompson, J. M. T. and Supple, W. J. (1973). Erosion of optimum designs by compound
branching phenomena, Journal of Mechanics of Physics and Solids 21, 3, pp. 135
144.
Tikhomirov, V. M. (1990). Stories about Maxima and Minima (Mathematical Association
of America, Providence, RI).
Tonon, F. (2004). On the use of random set theory to bracket the results of Monte Carlo
simulations, Reliable Computing 10, pp. 107137.
Tonon, F., Bae, H.-R., Grandhi, R. V. and Pettit, C. L. (2006). Using random set theory
to calculate reliability bounds for a wing structure, Structure and Infrastructure
Tonon, F. and Bernardini, A. (1998). A random set approach to the optimization of un-
certain structures, Computers & Structures 68, pp. 583600.
Tonon, F., Bernardini, A. and Elishakoff, I. (2001). Hybrid analysis of uncertainty: Proba-
bility, fuzziness and anti-optimization, Chaos, Solitons & Fractals 12, pp. 14031414.
Tseng, W. K. (2006). Minimization of maximum acoustic pressure over space and fre-
quency, Proc. Institution of Mechanical Engineering, Part C 220, 4, pp. 447455.
Tsompanakis, Y., Lagaros, N. D. and Papadrakakis, M. (eds.) (2007). Structural Design
Optimization Considering Uncertainties (Taylor & Francis, London).
Tunguskova, Z. G. (1970). Stress concentration around a hole with periodic irregularities,
Vestnik Moskovskogo Universiteta (Proceedings of Moscow University) , 4, pp. 127
133, (in Russian).
Turner, M. J. (1982). Optimization of structure to satisfy flutter requirement, AIAA Jour-
nal 7, 1, pp. 233239.
Tzan, S.-R. and Pantelides, C. P. (1996a). Convex models for impulsive response of struc-
tures, Journal of Engineering Mechanics 122, 6, pp. 521529.
Tzan, S.-R. and Pantelides, C. P. (1996b). Convex models for seismic design of struc-
tures 2: Design of conventional and active structures, Earthquake Engineering and
Structural Dynamics 25, 9, pp. 945963.
Utkin, L. V. and Gurov, S. V. (1996). A general formal approach for fuzzy reliability
analysis in the possibility context, Fuzzy Sets and Systems 83, pp. 203213.
van Dyke, M. (1975). Perturbation Methods in Fluid Mechanics (Parabolic Press, Stanford,
CA), pp. 1314.
van Keulen, F., Damveld, H., Elishakoff, I. and Toporov, V. (2001). Bounded-but-unknown
uncertainties in design optimization, in Proc. AIAA Structural Dynamics Conference
(CD-ROM), AIAA Paper 2001-1242.
Vanderplaats, G. N. (1991). Efficient structural optimization techniques, in Computer
Application in Civil and Building Engineering, Proc. 4th International Conference
(Kozo System Inc, Tokyo), pp. 151158.
Vanderplaats, G. N. (1993). Thirty years of modern structural optimization, Advances in
Engineering Software 16, pp. 8188.
Vanderplaats, G. N. and Yoshida, N. (1985). Efficient calculation of optimum design sen-
sitivity, AIAA Journal 23, pp. 17981803.
Vanderplaats, G. N. (1999). Numerical Optimization Techniques for Engineering Design
(Vanderplaats Research & Development, Inc., Colorado Springs, CO).
Vanmarcke, E. H., Diaz-Padilla, J. and Roth, D. A. (1972). Reliability-based optimum
design of simple plastic frames, Tech. Rep. R72-46, Department of Mechanical En-
gineering, Massachusetts Institute of Technology.
Vassart, N. and Motro, R. (1999). Multiparametered formfinding method: application to
tensegrity systems, International Journal of Space Structures 14, 2, pp. 147154.
Venini, P. (1998). Robust control of uncertain structures, Computers & Structures 67, pp.
165174.
Venkataraman, S. and Haftka, R. T. (2004). Structural optimization complexity: What
has Moores law done for us? Structural and Multidisciplinary Optimization 28, pp.
375387.
Venkayya, V. B. (1993). Introduction: Historical perspective and future direction, in M. P.
Kamat (ed.), Structural Optimization: Status and Promise (AIAA Press, Washing-
ton D.C.), pp. 110.
Venter, G. and Haftka, R. T. (1996). A two species genetic algorithm for designing
composite laminates subjected to uncertainty, in Proc. 37th AIAA/ASME/ASCE/
AHS/ASC Structures, Structural Dynamics, and Materials Conference and Exhibit:
Part 3 (Salt Lake City, UT).
Venter, G. and Haftka, R. T. (1999). Two species genetic algorithm for design under
worst case condition, in Proc. 3rd World Congress of Structural and Multidisciplinary
Optimization (Buffalo, N. Y.), (Also Evolutionary Optimization (Internet Journal),
2(1), 119).
Venter, G., Haftka, R. T. and Starnes, Jr., J. H. (1998). Construction of response surface
approximation for design optimization, AIAA Journal 36, 12, pp. 22422249.
Vick, S. G. (2002). Degrees of Belief: Subjective Probability and Engineering Judgement
(ASCE Press, Reston, VA).
Vinot, P., Cogan, S. and Lallement, G. (2003). Approche non-probabiliste de fiabilite base
sur les mode les convexes, Mecanique & Industries 4, 1, pp. 4550, (in French).
Vitali, F., Blanc, G., Larque, P., Duplay, J., Morvan, G., Nakagiri, S. and Suzuki, K.
(1999). Finite element interval analysis of external loads identified by displacement
input with uncertainty, Computer Methods in Applied Mechanics and Engineering
168, 1, pp. 6372.
Bibliography 383
von Hoerner, S. (1967). Homologous deformation of tiltable telescopes, Journal of the

Structural Division, ASCE 93, ST5, pp. 461485.
VR&D (1999). DOT Users Manual, Ver 5.0 (Colorado Springs, CO).
Wald, A. (1945). Statistical decision functions which minimize the maximum risk, The
Annals of Mathematics 46, 2, pp. 265280.
Wald, A. (1950). Statistical Decision Function (Wiley, New York).
Wall, F. J. and Deodatis, G. (1994). Variability response functions of stochastic plane
stress/strain problems, Journal of Engineering Mechanics 120, 9, pp. 19631982.
Walshan, D. (1994). Getting the most form your extreme wind data: A step-by-step guide,
Journal of Research at the National Institute of Standards and Technology 99, pp.
399412.
Wang, D. and Li, J. (2003). A reliable approach to compute the static response of un-
certain structural system, Yisual Lixue Xuebao (Chinese Journal of Computational
Mechanics) 20, 6, pp. 662669.
Wang, G. Y. and Wang, W. Q. (1984). Fuzzy optimum design of structures, Computational
Structural Mechanics and Application 2, pp. 6773, (in Chinese).
Wang, G. Y. and Wang, W. Q. (1985). Fuzzy optimum design of structures, Engineering
Wang, X. J., Elishakoff, I. and Qiu, Z. P. (2008a). Experimental data have to decide
which of the nonprobabilistic uncertainty descriptions convex modeling or interval
analysis to utilize, Journal of Applied Mechanics 75, pp. 04101810410188.
Wang, X. J., Elishakoff, I., Qiu, Z. P. and Ma, L. H. (2009). Comparison of probabilistic
methods and two non-probabilistic methods for uncertain imperfection sensitivity
of a column on a nonlinear mixed quadratic-cubic foundation, Journal of Applied
Mechanics 76, 1, p. 011007.
Wang, X. J. and Qiu, Z. P. (2009). Nonprobabilistic interval reliability analysis of wing
flutter, AIAA Journal 47, 3, pp. 743748.
Wang, X. J., Qiu, Z. P. and Elishakoff, I. (2008b). Non-probabilistic set-theoretic model
for structural safety measure, Acta Mechanica 198, pp. 5164.
Warmus, M. (1956). Calculus of approximations, Bulletin de lAcademie Polonaise des
Sciences 4, 5, pp. 253257, also available at http://www.cs.utep.edu/interval-
comp/warmus.pdf.
Weibull, W. (1951). A statistical distribution of wide applicability, Journal of Applied
Wentzel, E. C. (1980). Operations Research: Problem, Principles, Methodology (Nauka,
Moscow), (in Russian).
Westermo, B. D. (1985). The critical excitation and response of simple dynamic systems,
Journal of Sound and Vibration 100, 2, pp. 233242.
Whidborne, J. F., Gu, D. W. and Postlethwaite, I. (1997). Simulated annealing for multi-
objective control system design, IEE Proceedings of Control Theory and Applications
144, 6, pp. 582588.
Wilson, E. L., Der Kiureghian, A. and Bayo, E. P. (1982). A replacement of the SRSS
method in seismic analysis, Earthquake Engineering and Structural Dynamics 13,
pp. 112.
Witte, A. F. (1974). Experimental shock tests on the SPARTAN missile guidance section
using shock spectrum and least favorable response methods, Tech. Rep. Report K-
74-654(R), Kaman Sciences Corporation, Colorado Springs, CO.
Witte, A. F. and Smallwood, D. O. (1974). A comparison of shock spectra and the least
favorable response techniques in a transient test program, in Proc. 20th Annual
Meeting, Institute of Environmental Sciences (Washington D.C.), pp. 1629.
Wolfram, S. (1996). The Mathematica Book, 3rd edn. (Cambridge University Press, New
York).
Wolkowicz, R. and Vandenberghe, L. (eds.) (2000). Handbook of Semidefinite Programming
Theory, Algorithms, and Applications (Kluwer, Dordrecht, Netherlands).
Wood, K. L., Antonnson, E. K. and Beck, J. L. (1990). Representing imprecision on engi-
neering design comparing fuzzy and probabilistic calculus, Research in Engineering
Design 1, 34, pp. 187203.
Woodbury, M. A. (1950). Inverting modified matrices, Memorandom Report 42, Statistical
Research Group, Princeton, NJ.
Wu, T. S. and Boley, B. A. (1965). Bounds in melting problems with arbitrary rates of liq-
uid removal, Technical Report 24, Department of Civil Engineering and Engineering
Mechanics, Columbia University, New York.
Wu, Z. and Soong, T. T. (1996). Design spectra for actively controlled structures based
on convex models, Engineering Structures 18, 5, pp. 341350.
Xu, C. (1989). Fuzzy optimization of structures by the two-phase method, Computers &
Yamaguchi, T., Kogiso, N. and Yamakawa, H. (2007). Optimal interplanetary trajectories
for impulsive deflection of potentially hazardous asteroids under velocity increment
uncertainties, Transactions of the Japanese Aerospace Sciences 55, pp. 432438.
Yang, J. N. (1994). Application of reliability methods to fatigue, quality assurance and
maintenance, in G. I. Schueller, M. Shinozuka and J. T. P. Yao (eds.), Structural
Safety and Reliability (Balkama, Rotterdam, Netherlands), pp. 320.
Yao, J. T. P. (1985). Safety and Reliability of Existing Structures (Pitman, Boston, MA).
Yau, W. (1975). Optimal design of simply supported beams for minimum upper bound of
dynamic response, Journal of Applied Mechanics 41, 1, pp. 301302.
Yeh, Y. C. and Hsu, D. S. (1990). Structural optimization with fuzzy parameters, Com-
puters & Structures 3, 4, pp. 575580.
Yoshikawa, N. (2002). Convex approach to estimate the worst excitation, in H. A. Mang,
F. G. Rammerstorfer and J. Eberhardsteiner (eds.), Proc. 5th World Congress on
Computational Mechanics (Vienna).
Yoshikawa, N. (2003). Structural integrity analysis via convex model of uncertainty for
worst-case ground motion caused by earthquakes, Journal of Society of Material
Science of Japan 52, 1, pp. 1015.
Yoshikawa, N., Elishakoff, I. and Nakagiri, S. (1998a). Worst case estimation of homology
design by convex analysis, Computers & Structures 67, pp. 191196.
Yoshikawa, N. and Fr yba, L. (1997). Interval analysis of dynamic railway track on random
foundation, in Y. Naprstek and J. Minsther (eds.), Dynamics and Vibrations, Proc.
Engineering Mechanics, Vol. 2, pp. 227232.
Yoshikawa, N. and Nakagiri, S. (1990). Homology design of frame structure by finite el-
ement method, in Theoretical and Applied Mechanics, Proc. 42nd Japan National
Congress for Applied Mechanics, Vol. 42 (University of Tokyo Press), pp. 4351.
Yoshikawa, N., Nakagiri, S. and Kuwazuru, O. (1998b). Structural synthesis for worst case
mitigation based on convex set of uncertainty, in N. Shiraishi, M. Shinozuka and
W. K. Wen (eds.), Structural Safety and Reliability (Balkema, Rotterdam, Nether-
lands), pp. 693696.
Youn, B. D., Choi, K. K. and Park, Y. H. (2003). Hybrid analysis method for reliability-
based design optimization, Journal of Mechanical Design 125, 2, pp. 221232.
Young, R. C. (1931). The algebra of many-valued quantities, Mathematische Annalen 104,
pp. 260290, (Available at http://www.cs.utep.edu/interval-comp/young.pdf).
Youssef, N. A. N. and Popplewell, N. (1978). A theory of the greatest maximum response
Bibliography 385
of linear structures, Journal of Sound and Vibration 56, pp. 2133.

Youssef, N. A. N. and Popplewell, N. (1979). The maximax response of discrete multi-
degree-of-freedom systems, Journal of Sound and Vibration 64, 1, pp. 115.
Zadeh, L. A. (1965). Fuzzy sets, Information and Control 8, pp. 338353.
Zadeh, L. A. (1994). Why the success of fuzzy logic is not paradoxical, in IEEE Expert
Intelligent Systems and Their Applications, pp. 4345.
Zaguskin, V. L. (1958). On circumscribed and inscribed ellipsoids of the external value,
Uspekhi Makmaticheskikh Nauk (Progress in Mathematical Sciences) 13, 6, pp. 89
93, (in Russian).
Zavoni, E. H., Iturrizaga, R. M. and Esteva, L. (1999). Optimal instrumentation of struc-
tures on flexible base for system identification, Earthquake Engineering and Struc-
tural Dynamics 28, pp. 14711482.
Zhai, C.-H. and Xie, L.-L. (2007). A new approach of selecting real input ground mo-
tions for seismic design: the most unfavorable real seismic design ground motions,
Earthquake Engineering and Structural Dynamics 36, pp. 10091027.
Zhang, H. (2005). Nondeterministic linear elastic finite element analysis: An interval ap-
proach, Ph. D dissertation, Georgia Institute of Technology.
Zhang, J. Y. and Ohsaki, M. (2006). Adaptive force density method for form-finding
problem of tensegrity structures, International Journal of Solids and Structures 43,
1819, pp. 56585673.
Zhang, J. Y. and Ohsaki, M. (2007). Stability conditions for tensegrity structures, Inter-
national Journal of Solids and Structures 44, 11-12, pp. 38753886.
Zhang, J. Y., Ohsaki, M. and Araki, Y. (2004). Optimal measurement positions for iden-
tifying stress distribution of membrane structures using cable net approximation, in
Proc. 3rd China-Japan-Korea Joint Symposium on Optimization of Structural and
Mechanical Systems (CJK-OSM3) (Kanazawa, Japan), pp. 547552.
Zhang, J. Y., Ohsaki, M. and Uchida, A. (2008). Equivalent static loads for nonlinear
seismic design of spatial structures, in Proc. 14th World Conference on Earthquake
Engineering, Paper No. 05-01-0110 (Beijing).
Zhang, Z. Y. and Zhang, J. H. (2000). A two-strength-parameters-bounds convex model
using both modelling and numerical calculation, Chinese Journal of Applied Me-
chanics 17, 1, pp. 8590, (in Chinese).
Zharii, O. Y., Jih-Hua, C., Chi-Ti, H., Li-Wei, L. and Pantelides, C. P. (1996). Stability
of elastic bars on uncertain foundations using a convex model, International Journal
of Solids and Structures 33, 9, pp. 12571269.
Zheng, Y. and Lewis, R. W. (1993). Optimization concept of grey systems, Applied Math-
ematical Modelling 17, 7, pp. 388397.
Zhu, L. P. and Elishakoff, I. (1996). Hybrid probabilistic and convex modelling of excitation
and response of periodic structure, Mathematical Problems in Engineering 2, 2, pp.
143163.
Zingales, M. and Elishakoff, I. (2000). Anti-optimization versus probability in an applied
mechanics problem: Vector uncertainty, Journal of Applied Mechanics 67, 3, pp.
472484.
Zingales, M. and Elishakoff, I. (2001). Hybrid aeroelastic optimization and antioptimiza-
tion, AIAA Journal 39, 1, pp. 161165.
Zuccaro, G., Elishakoff, I. and Baratta, A. (1998). Antioptimization of earthquake excita-
tion and response, Mathematical Problems in Engineering 4, pp. 119.

Index
accelerogram convex hull, 303

admissible, 177 critical excitation, 176
credible, 179 critical point, 281
autocorrelation function, 195 coincident, 281
criticality condition, 115, 116, 125
ball, 59 cycle-based method, 277
BernoulliEuler beam, 191, 218, 221 cylinder, 59
bifurcation
point, 135 dependency problem, 55
stable, 283 design by experiment, 31
stable-symmetric, 282 design load, 273
unstable-symmetric, 282 Dirac delta function, 153
bound disk, 59
ellipsoidal, 115, 147, 275 distance
energy, 149 Euclid, 33
interval, 49 humming, 33
BubnovGalerkin approach, 192 Manhattan, 33
buckling
Euler load, 230 earthquake
linear, 117 epicentral intensity, 184
linear analysis, 117, 280 spectrum-compatible, 181
mode, 282 eigenvalue problem, 67
linear, 117 eigenmode, 67, 292
nonlinear, 280282 eigenvalue, 67, 292
element-by-element approach, 215
calculus of variation, 18 ellipsoid
CauchySchwartz inequality, 183 principal axis, 59
compliance, 294 semi-axis, 57
constraint ellipsoidal model, 57, 61, 78, 85, 117, 166
active, 24 enumeration, 122
bound, 24, 62, 80 vertex, 63, 109, 131, 278, 297
box, 24 equilibrium path
equality, 22 bifurcation, 115, 284
inactive, 24 fundamental, 115, 282
inequality, 22 postbuckling, 284
side, 24 error
387
identification, 320 initial, 229

index, 304 parameter, 135
round-off, 48 pattern, 135
vector, 304 random, 140
exact reanalysis, 69 worst, 136, 281
interval
feasible design, 20 analysis, 49, 50, 212, 245
feasible region, 23 arithmetic, 54
finite element method, 211 coefficient, 221
stochastic, 190 Gaussian elimination, 213
fixed point, 213 matrix
flexible structure, 307 radius, 217
flutter velocity, 194 number
force identification, 316 absolute value, 55
forced vibration, 148 addition, 53
Fourier diameter, 56, 214
expansion, 184 distance, 55
series, 178 distributivity property, 55
transformation, 291 division, 53
frame multiplication, 53
braced, 118 overestimation, 214
shear, 217 radius, 56, 219
function subcancellation property, 55
characteristic, 42 subdistributity property, 55
concave, 62, 129, 278 subtraction, 53
convex, 24, 107, 278, 294 width, 56
impulse response, 164, 176
membership, 42 KarushKuhnTucker condition, 25
monotonic, 278 kinetic energy, 314
quasi-convex, 308 Kronecker delta, 67, 152
singularity, 229, 232
step, 229 lack of knowledge, 49
fuzzy theory, 42 Lagrange multiplier, 25, 62, 115, 147, 158,
166, 179, 215, 275, 304
Gamma function, 4 Lagrangian, 25, 62, 78, 115, 269
gas-flow interaction, 192 Laplace transform, 153
generalized displacement, 135 Liapunovs direct method, 308
GramSchmidt orthogonalization, 169 limit point, 135, 283
greedy method, 34, 294, 298 load factor, 116, 281
ground structure approach, 36 bifurcation, 115
guaranteed approach, 57, 228 buckling, 115, 116, 282
critical, 115, 282
heuristics, 31 linear buckling, 117
hilltop branching point, 134, 281 multiple buckling, 118
homologous deformation, 299 local search, 32
homology design, 298
hyperellipse, 303 matrix
hypersphere, 305 connectivity, 103, 317
flexibility, 294
imperfection interval, 87
Index 389
mass, 67 prebuckling deformation, 116

negative definite, 183 prestress, 316
perturbation, 86 probabilistic method, 227
positive definite, 24, 57, 59, 183, 196, probability
294, 303 exponential, 3
positive semidefinite, 24, 26, 59 joint, 250
square root, 59 log-normal, 6
stability, 116, 134 marginal, 252
tangent stiffness, 105 normal, 140
transformation, 57, 59 Rayleigh, 3
weight, 57 truncated exponential, 240
Minkowski sum, 60 uniform, 10, 232
modal decomposition, 292 Weibull, 3
mode interaction, 123, 124 programming
MoorePenrose generalized inverse, 301, convex, 24
319 integer, 24
linear, 25, 79, 116, 148
norm multiobjective, 28
Chebychev, 33 nonlinear, 22, 25
Euclidean, 301, 320 parametric, 26, 279
quadratic, 25, 78, 148
OettliPrager lemma, 222 semidefinite, 26
optimal solution sequential linear, 310
global, 24 proportional load, 281
local, 24 pseudo force method, 70
Pareto, 29 pseudodistortion method, 70
optimality condition, 25
optimization random search, 35
fuzzy, 41 rank, 318
geometry, 36 deficiency, 319
layout, 36 Rayleigh principle, 129
probabilistic, 39 reliability, 3, 231, 251
shape, 36 reliability-based design, 274
sizing, 36 response
topology, 18, 36 least favorable, 158
two level, 276 most favorable, 158
optimum design sensitivity, 27 response spectrum approach, 292
orthogonality condition, 164 response surface method, 31, 279
out-of-straightness, 120 robust design, 274
outward rounding, 54 RouthHurwitz criterion, 156
parallel computation, 277 safety factor, 273

parallelepiped, 167 actual, 3
peak ground acceleration, 177 central, 4
penalty function, 34 required, 3
exterior, 34 self-equilibrium
interior, 34 force, 104
polytropic thermodynamic state, 317
transformation, 192 sensitivity
possibilistic approach, 228 imperfection, 135
coefficient, 22, 65, 117, 146 stability

imperfection, 280 asymptotic, 307
linear buckling load, 117 boundary, 129
sensitivity analysis, 63, 117, 301, 304, 320 static pushover analysis, 292
compliance, 107 stiffness matrix
design, 63, 65 geometrical, 116, 129
direct differentiation method, 64 tangent, 116, 129, 282
eigenvalue, 67 strain energy, 308
finite-difference approach, 66 stress concentration factor, 90
shape, 63, 68 supersonic stream flow, 191
tangent stiffness matrix, 106
set tensegrity structure, 102, 316
convex, 24 total potential energy, 124, 134, 215
crisp, 42
fuzzy, 42 uncertainty
Pareto optimal, 29 data, 49
ShermanMorrisonWoodbury formula, 70 model, 49
simulated annealing, 35, 122 unknown-but-bounded, 57, 303
slenderness ratio, 126
solution Vietas theorem, 193
complementary, 301 virtual distortion method, 70
dominance, 29 viscoelastic kernel, 151
neighborhood, 33
particular, 301 worst load scenario, 275, 298
spectrum worst load-case, 79
central, 180 worst-case design, 273
KanaiTajimi, 180 worst-case scenario, 320
power, 180
response, 180
response displacement, 186
Author Index
Aarts, E. 32, 35 Au, F. T. K 325

Abbas, A. M. 112, 189 Averin, A. M. 90
Abdullah, M. M. 317 Avis, D. 109, 110
Adali, S. 45, 63, 144, 280, 325, 336 Ayyub, B. 46
Adduri, P. R. 272 Azarm, S. 336
Adeli, H. 40
Ahrens, R. 153 Babu, S. R. 317
Akcakaya, R. 340 Babuska, I. vii, xiv, 49, 85
Akgun, M. A. 70 Badcock, K. J. 208
Alefeld, G. 52, 54, 56, 87, 220 Bae, H-R. xiii
Allen, M. 190 Bagchi, G. 163, 164, 208
Alotto, P. 85 Bagley, R. L. 158
Amazigo, J. C. 227 Bai, E-W. 338
Andersland, M. S. 338 Bainer, L. 190
Andreykiv, A. 325 Baitsch, M. 144, 279
Ang, A. H-S. 14 Balakrishna, C. 43, 45
Antonio, C. A. C. 26 Balandin, D. V. 150
Antonnson, E. K. xiii Balasubramanyam, S. 208
Aoki, K. 299 Banichuk, N. V. 85, 325
Aoues, Y. 339 Baratta, A. 61, 145, 175, 177
Araiza, R. 225, 272 Barbieri, E. 85
Arakawa, M. 43, 45 Barthelemy, J. 27
Araki, Y. 316, 323 Batill, S. M. 325
Arbocz, J. 144 Baykasoglu, A. 30
Argyris, J. H. 69 Bayo, E. P. 292
Ariaratnam, S. T. 227 Beck, J. L. xiii, 272
Arora, J. S. 18, 27, 30, 127 Beer, M. xiii, 43, 45, 46
Asami, T. 336 Bellman, R. 336
Ashida, F. 30, 279 Beltzer, A. I. 92
Ashley, H. 190, 191, 327 Belytschko, T. vii
Attoh-Okine, N. O. 61, 85 Ben-Haim, Y. ix, xixiii, 2, 57, 61, 85,
391
144, 151, 189, 208, 212, 227, 228, Chakraborty, S. 45

247, 274, 299, 331 Chang, F. C. 336
Ben-Tal, A. xi, 26, 85, 227 Chateauneuf, A. 339
Bendse, M. P. 20, 36 Chen, G. 189
Berke, L. 89 Chen, J. 86, 89
Berleant, D. J. 15, 340 Chen, L. 225
Bernardini, A. xiii, 43, 46 Chen, M. D. 325
Bert, C. W. 335 Chen, R. 85
Besharati, B. 336 Chen, S.-H. 45, 56, 61, 70, 8589, 112,
Bezdek, J. 45 145, 146, 208, 209, 225, 325
Bickford, W. B. 70 Chen, S. Q. 45
Bishop, J. A. 190 Chen, W-F. 121
Black, M. 41 Chen, W. Y. 208
Blanc, G. 225 Cheng, G. 274, 332
Bland, D. R. 150 Cheng, Y. S. 61, 274, 325
Bletzinger, K.-U. 144, 281 Cherkaev, A. V. 85, 325, 336
Blockley, D. I. 45, 46 Cherkaev, E. 85, 325, 336
Bloebaum, C. L. 27 Chern, J. M. 336
Boca, T. J. 336 Chernousko, F. L. 47, 5761, 228, 247
Bochenek, B. 284 Chessa, J. 225
Bogdanovich, A. E. 227 Chi, C. C. 43
Boley, B. A. ix, 77, 8084 Chi-Ti, H. 144
Bolotin, V. V. viii, ix, xiv, 191, 194 Chiang, W. L. xiii, 45
Booth, B. C. 61, 211, 325 Chiaruttini, V. 63, 144, 280
Borkowski, A. 61 Chleboun, J. 85
Borse, G. J. 323 Cho, M. 85, 144
Boyd, S. 107, 308 Choi, K. K. 23, 45, 63, 274, 332
Brach, R. M. 325, 336 Chopra, A. K. 292
Bricmont, J. x Christensen, R. M. 150
Brown, C. B. 45 Chryssanthopoulos, M. K. 144
Budhiraja, A. S. 325 Chu, F. L. 208
Budiansky, B. 227 Cinquini, C. 85
Bulgakov, B. V. 149, 150 Claudio, D. 54, 56
Burgman, M. 335 Clough, R. W. 292
Cogan, S. 61
Cai, G. Q. xii, 14, 15, 144, 228, 339 Cohn, M. Z. 39, 332
Cai, K.-Y. 45 Cohon, J. L. 28
Calafiore, G. 61 Colombi, P. 61, 212, 272, 339
Capiez-Lernout, E. 208 Conrado, A. C. 144
Cattafesta, R. T. 325 Corliss, G. 85
C
akmak, H. S. 85, 89, 217, 221 Cornell, C. A. 1, 334
Ceberio, M. 272 Corotis, R. 190, 209
Author Index 393
Corotis, R. B. 39, 70 Duplay, J. 225

Cozzarelli, F. A. 151 Duvaut, G. 63, 144, 280
Craig, K. J. 208 Dwyer, P. S. 49
Cremona, C. 45 Dyne, S. J. C. 208
Cudney, H. 45 Dzhur, Y. 85
Cudney, H. H. 45
Cuntze, R. 40 Eastep, F. E. 190
El Damatty, A. A. 114, 274, 281
Dafalias, J. F. 336 El Ghaoui, L. 61, 227
Damveld, H. 325 El-Hami, A. 274, 332
Dannon, V. 336 Eliseeff, P. 151
Das, P. K. 331 Elishakoff, I. ixxiii, 2, 9, 10, 12, 13, 15,
Dash, P. K. 112 45, 46, 49, 56, 57, 61, 85, 90, 92, 94,
Datcheva, M. 85 103, 144146, 148, 151, 163, 164, 173,
Davies, P. 208 175, 189, 190, 197, 198, 208, 209, 211,
de Almeida, S. F. M. 114, 129, 144, 212, 217, 218, 220, 225, 227, 228, 230,
280, 325 247, 248, 272, 273, 280, 288, 291, 299,
de Faria, A. R. 114, 129, 144, 280, 325, 331, 333, 334, 337340
325 Ellingwood, B. 190, 209
de Finetti, B. 338 Elliot, S.J. 336
de Palma, P. 325, 340 Elseifi, M. 144
Deif, A. 87 Elseifi, M. A. 61, 144
Deml, M. 144 Eschenauer, H. A. vii, 36
Dempster, A. P. 45 Esteva, L. 317
Demyanov, V. M. 336
Deng, L. 70 Fang, J. 273, 288, 340
Deodatis, G. 112, 190 Fang, J. J. xiii, 45, 49
Der Kiureghian, A. 113, 292 Fang, T. 14, 189, 339
Dessombz, O. 56, 225 Fatemi, J. 325
Diaz-Padilla, J. 39 Faulkner, D. 331
Dimarogonas, A. D. 56, 89 Feigen, M. 39
Ditlevsen, O. 44 FEMA 292
Dong, W. M. xiii, 45 Feng, Y. 49
Dorn, W. 18 Ferracuti, B. x, 45
Dovi, A. 27 Ferson, S. 15, 52, 340
Drenick, R. F. ix, x, 163, 164, 176, Feshbach, H. 92
177, 189, 208, 339 Fest, E. 316
Du, D. Z. 336 Feynman, R. F. vii
Du, L. 45 Floudas, C. A. 32
Duan, D. H. 208 Fl
ugge, W. 150
Dube, R. 325 Foley, C. 85
Dubois, D. xiii, 43, 45, 46 Fox, R. L. 336
Frangopol, D. M. 39, 40, 330, 331 Goodman-Strauss, C. 15

Frank, D. 227 Goosen, J. F. L. 211, 277, 279, 325
Freiberger, P. 46 Gordeyev, V. 85
Freudenthal, A. M. ix, 41, 333 Grab, W. 43
Frommer, A. 49, 61, 89, 209 Graham, H. 70
Fr
yba, L. 208, 209, 274 Graham, L. 112, 190
Fuchs, M. B. 217, 336 Graham, W. B. 153
Fujisawa, K. 26 Grandhi, R. V. xiii
Fukuda, K. 109, 110 Grandori, G. 333
Fuller, R. B. 102 Greenberg, H. 18
Furuta, H. 45 Griffin, O. H. 86
Grigoriu, M. 14, 339
Gabriel, D. 208 Gu, D. W. 30
Gal, T. 26 Gu, X. 325
Gana-Shvili, J. 61, 85, 101 Gu, Y. 89
Gangadharan, S. N. 86 Gubanova, I. I. 193
Ganzerli, S. 45, 58, 61, 85, 189, 274, Guo, S-X. 225
279, 325, 340 Gurav, S. P. 211, 277, 279, 325
Gao, W. 208 G
urdal, Z. 18, 144
Gao, Y. 45 Guretskii, V. V. 336
Garcelon, J. H. 70 Gurov, S. V. 45
Gasser, M. 40 Guz, A. N. 97
Gay, D. M. 213
Gelatt, C. D. 35 Haftka, R. T. 18, 30, 31, 45, 46, 70, 85,
Genalo, L. J. 190 86, 273, 276, 277, 288, 325, 328, 340
Ghosn, M. 70 Hagaki, H. 211
Giavotto, V. 211, 274, 279 Haider, G. S. 39
Gill, P. E. 289 Hajagos, J. G. 272, 340
Gindy, N. 30 Hajela, P. 27, 274
Ginzburg, L. R. 15, 340 Hakim, S. 336
Gioffre, M. 14 Haldar, A. xiii
Givoli, D. 61, 85, 90, 92, 94 Hammond, J. K. 208
Glegg, S. 151 Han, S. W. 292
Glover, F. 32 Han, W. Z. 85
Gnoenskii, L. S. 150 Hangai, Y. 299
Godoy, L. A. 284 Hanif, J. 317
Goldberg, D. E. 30, 32 Hansen, E. 52
Golla, D. F. 153 Hansen, J. S. 129, 144, 280, 325
Gomory, R. 18 Hanss, M. 46
Goncharenko, V. M. 227 Hartmann, D. 144, 279
Gonzales, A. 336 Hasofer, A. M. 2
Good, I. J. 45 Hattori, A. 279
Author Index 395
Haug, E. J. 63, 273, 279 Isshiki, Y. 36

Hausen, Y. S. 325 Ito, J. 316
Hayes, B. 47, 49 Iturrizaga, R. M. 317
He, C. Y. 189 Ivanova, S. Y. 325
He, I. Q. 208 Iwaya, E. 225
Heckert, A. 15, 339 Iyengar, R. N. 112, 189
Hemp, W. S. 18
Henderson, H. V. 7072 Jabbari, F. 190
Hertz, T. H. 190 James, B. 27
Herzberger, J. 52, 87, 220 Jamison, D. 45
Hilding, D. 336 Jensen, H. 45, 56
Hilton, H. H. 39, 151 Jezequel, L. 189, 208, 225
Hirota, K. 42 Jia, H. G. 56, 209
Hjelmstad, K. D. 317 Jiang, L. 274, 332
Hlavacek, I. 85, 208 Jih-Hua, C. 144
Ho, D. 114 John, F. 60
Hoff, N. J. 231 Johnson, C. R. 320
Hoh, S. 211 Johnson, J. L. 45
Holmes, J. D. 15 Ju, F. 336
Horn, R. A. 320 Jung, J. J. 30
Horst, R. 23, 32
Hoshi, Y. 225 Kaifu, N. 299
Howe, M. 334 Kalinowski, A. 336
Hsiao, J. H. 43 Kalman, R. E. 2, 333
Hsu, D. S. 43 Kam, T.-Y. 45
Hsu, J. 151 Kamat, M. P. 18
Hu, B. 150 Kanda, J. 15
Hu, J. X. 209 Kandathi, R. 272
Huang, W. N. 151 Kanno, Y. 26, 85, 89, 134, 281, 316, 323
Hughes, P. C. 153 Kapoor, M. P. 336
Hughes, T. Y. R. vii Kapur, K. C. 39
Hunt, G. W. 113, 118, 123125, 133, Kasyap, A. 325
136, 330 Katoh, N. 26, 36
Huseyin, K. 124, 129 Kaveh, A. 317
Hutchinson, J. W. 227 Kavlie, D. 70
Keafott, R. B. 52, 85
Iankov, R. 85 Khachaturian, N. 39
Ikeda, K. 113, 114, 124, 125, 134, 136, Khalessi, M. R. 61
281, 330 Khariton, L. E. 40
Ilyushin, A. A. 191 Kharitonov, V. 334
Inman, D. 151, 153, 154 Kharmanda, G. 274, 332
Ishida, M. 90 Khodaparasi, H. H. 208
Kida, T. 299 Laine, J.-P. 56, 225

Kildibekov, I. G. 227 Lallement, G. 61
Kim, C. 274, 332 Lalvani, H. 102
Kim, N-H. 23, 63 Langhaar, M. 325
Kim, S-E. 121 Langley, R. S. 228
Kim, T-U. 144 Lanzi, L. 211, 274, 279
Kim, Y. V. 15, 228 Larque, P. 225
Kinoshita, T. 30, 307 Laville, F. 339
Kinser, D. E. 39 Lee, J. 86
Kirby, J. S. 151 Lee, K. 86
Kirkpatrick, S. 35 Lee, T. H. 30
Kirsch, U. 36, 69, 70 Lefschetz, S. 308
Klarbring, A. 336 Leichtweiss, K. 60
Klir, G. J. 15, 42 Leissa, A. W. 335
Kogiso, N. 144, 209, 274 Lene, F. 63, 144, 280
Koiter, W. T. 284 Leng, H. N. 208
Kolakowski, P. 70 Leng, X. L. 14, 189, 339
Koltunov, M. A. 150 Lewis, R. W. 43
Komatsu, K. 299 Ley, R. 329
Komkov, V. 63, 336 Li, J. 89
Korneev, A. B. 150 Li, Q. 45, 228, 272, 339
Korst, J. 32, 35 Li-Wei, L. 144
Kosko, B. xiii, 16, 42, 4446 Li, Y. W. ix, 61, 144, 189, 208, 248
Kocvara, M. 325 Lian, H. D. 70, 85, 89, 112, 208
K
oyluo
glu, H. U. 225 Liaw, D. G. 190
K
oyluo
glu, U. 217, 220 Libai, A. 227
Kozovkov, N. T. 150 Librescu, L. 190
Kreinovich, V. 61, 225, 272 Lignon, S. 189, 208
Krenn, B. 284 Lin, Y. K. ix, xiii, 14, 61, 272, 299, 339
Krichevskii, I. Ya. 150 Lindberg, H. E. 61, 114, 144, 227
Krishna, L. G. 150 Lindley, D. V. 15
Krishnamurthy, T. 45 Liu, S. 69
Kucuk, I. 325, 336 Liu, Z. S. 56, 85, 208
Kulpa, Z. 85, 89 Livne, E. 190
Kumamoto, H. 336 Lockhart, D. 227
Kurzhanski, A. B. 58, 61, 63, 228, 247 Lockwood, E. H. 94
Kuttekeuler, J. 190 Lodwick, W. 45
Kuwazuru, O. 61 Loftus, J. J. 45
Kwak, B. M. 273, 279 Lomakin, V. A. 90, 92, 96
Kwakernaak, H. xiii Lombardi, M. 85, 276, 277, 325
Longpre, L. 272
Ladevèze, P. 49 Lu, Q. S. 209
Author Index 397
L
u, Z-Z. 225 Mitra, S. K. 301
Luenberger, D. G. 18, 23, 248 Modares, M. 145, 208
Luik, R. 69 Moens, D. 45, 212
Molfino, P. 85
Ma, J. 208 Molinari, G. 85
Ma, L. H. xii, 61, 89, 144 M
oller, B. xiii, 43, 45, 46
Ma, X. P. 14, 189, 339 Mongin, P. 339
Ma, Y. 86, 89, 209 Montgomery, D. C. 30, 31, 279
Maglaras, G. 45 Mooij, E. 325
Majumder, L. 89 Moore, R. E. 4749, 52, 53, 56, 87
Makode, P. V. 70 Moreau, J. J. 61
Malozemov, V. N. 336 Morimoto, M. 299
Manasyan, A. A. 336 Morrison, W. J. 70, 72
Mang, H. A. 284 Morse, P. M. 92
Manohar, C. S. 112, 189, 190 Morvan, G. 225
Maplesoft 125, 139 Moses, F. 39
Mariani, M. 85, 325 Motro, R. 102, 109, 316
Marler, T. 30 Mottershead, J. E. 208
Marques, S. 208 Mourelatos, Z. P. 45
Marti, K. 40 Moustafa, A. 189
Martin, J. B. 336 Muhanna, R. L. 5254, 85, 89, 145, 208,
Masur, E. F. 114 211213, 215, 216, 225
Matheron, G. xiii, 46 Mulkay, E. 225
Matsumoto, M. 225 Mullen, R. L. 5254, 85, 89, 145, 208,
Maturana, S. 45 211213, 215, 216, 225
Maute, K. 40, 190, 330, 331 M
uller, P. C. 49, 61, 89, 209
Mayer, G. S. 54 Mungan, I. vii
Mayer, M. 1 Murota, K. 113, 134
Maymon, G. 227 Murotsu, T. 144, 274
McNeill, D. 46 Murotsu, Y. 340
McTavish, D. J. 153, 158 Murray, W. 289
McWilliam, S. 85, 89 Myers, R. H. 30, 31, 279
Melosh, R. J. 69
Meng, G. A. 14, 189, 339 Naishul, A. B. 150
Micaletti, R. 112, 190 Nakagiri, S. 61, 208, 225, 299
Michalopoulos, C. D. 149, 150 Nakamura, T. 27, 36
Michell, A. G. M. 18 Nakayama, H. 279
Miki, M. 274 Nassef, A. O. 114, 274, 281
Minciarelli, F. 14 Nastran 314
Mineau, D. 190 Natke, H. G. 208, 325
Minsky, M. 338 Nayak, A. 272
Mironovkii, L. S. 150 Neittaanm aki, P. 85
Nelsen, R. B. 340 Panovko, Y. G. 193

Nelson, M. I. 332 Pantelides, C. P. 45, 58, 61, 85, 114, 144,
Nemirovski, A. xi, 26, 85, 227 149, 189, 190, 211, 274, 279, 325
Nemish, Y. N. 97 Papadopoulos, V. 112
Neuber, M. 90 Papadrakakis, M. 112
Neumaier, A. 45, 52, 85, 89, 213, 220, Papoulis, A. 336
222 Pardalos, P. M. 336
Nguyen, T. B. 150 Parimi, S. R. 39
Nickel, K. 213 Park, C. B. 208
Nielsen, S. R. K. 85, 89, 217, 221 Park, J. 30
Nikolaidis, E. 45, 46, 86, 144 Park, Y. H. 274
Niordson, F. I. 327 Parkinson, C. N. 329
Nishihara, O. 336 Parnes, R. 92
Nishiwaki, S. 20 Peek, R. 114
Nobile, F. xiv, 49, 85 Peltier, A. 329
Norton, R. L. viii, ix Penmetsa, R. C. 272
Novomestky, F. 163, 164, 208 Penzien, J. 177, 292
Peterson, R. E. 90
Oberguggenberger, M. 45, 46 Pettit, C. L. xiii
Oberkampf, W. L. 340 Philappacopoulos, A. J. 189, 208
Oden, J. T. vii, 49 Pierson, B. L. 190
Ohishi, Y. 103, 108 Pietrzak, J. 283
Ohsaki, M. xii, 20, 26, 27, 30, 3537, Pilkey, W. D. 336
70, 76, 102, 103, 105, 108, 113, 114, Plaut, R. H. 190, 336
118, 124, 125, 134, 136, 280283, Pletner, B. 148, 163, 164, 189
293, 307, 308, 316, 323, 330 Plesek, J. 208
Ohta, Y. 45 Pochtman, Y. M. 40
Oide, K. 114, 134 Poggi, C. 144
Olhoff, N. vii, 114, 274, 332 Polak, E. 113
Onipede, O. 317 Popova, E. D. 85
Otsubo, H. 225 Popplewell, N. 336
Ottl, D. 153 Postlethwaite, I. 30
Ovseyevich, A. I. 15, 228 Pothisiri, T. 317
Owen, S. 30 Powell, G. H. 70
Pownuk, A. 85, 89, 325
Paez, T. L. 336 Prade, H. xiii, 45, 46
Pal, N. R. 46 Prager, W. 18, 107
Palassopoulos, G. V. 274 Prakash, B. G. 43, 45
Palmov,
V. A. 90 Prokhorov, V. N. 150
Pan, P. 30, 307 Pu, Y. 331
Panagiotopoulos, P. D. 61 Puel, G. 49
Pang, J.-S. 336 Pugsley, A. G. 330
Author Index 399
Puig, V. 336 Rzhanitsyn, A. R. ix

Putresza, J. T. 70
Sage, A. P. 150
Qiu, Z. P. xii, 45, 49, 56, 61, 8589, Sakakibara, O. 299
144, 146, 173, 208, 209, 272, 339 Sakata, S. 30, 279
Quevedo, J. 336 Salle, J. L. 308
Saludes, Y. 336
Rackwitz, R. 40, 334 Sam, P. C. 45
Ramirez, M. R. 70 Sanayei, M. 317
Ramm, E. 144, 281 Sandberg, I. W. 30
Rao, C. R. 301 Sangalli, A. 46
Rao, S. S. 43, 45, 89, 225 Sarkar, A. 112, 189, 190
Rasmussen, S. H. 114 Sarma, K. 40
Reaveley, L. D. 189, 325 Sato, M. 338
Reddy, R. K. xiii Saunders, M. A. 289
Reeves, C. 23, 32 Savin, G. N. 90
Reitinger, R. 144, 281 Savoia, M. xiii
Ren, Y. J. 218 Saxena, V. 225
Renaud, J. E. 325 Sayles, R. S. 90
Reshetnyak, Y. N. 15, 228 Schanz, T. 85
Rhee, S. Y. 85, 144 Schek, H. J. 104, 318
Richardson, A. 317 Schiehlen, W. 150
Richter, A. 144, 280, 325 Schmidt, R. 335
Riks, E. 116 Schmit, L. A. 299
Riley, T. A. 149, 150 Schmitendorf, W. E. 190
Ringertz, U. T. 190 Schorrock, P. A. 134
Rohn, J. 52 Schranz, C. 284
Roitenberg, Ya. N. 150 Schueller, G. I. 41, 332, 333
Romeuf, T. 49 Schweppe, F. C. 57, 61, 247
Rosenblatt, F. 30 Searle, R. 7072
Roth, D. A. 39 Sensmeier, M. D. 86
Roux, W. J. 31 Sepulveda, A. E. 299
Roy, R. 69 Sevin, E. 336
Royset, J. O. 113 Sexmith, R. G. 332
Rozvany, G. I. N. 18, 36 Shao, S. 144, 274, 340
Rubinstein, M. F. 69 Sharuz, S. M. 150
Ruiz, P. 177 Shea, K. 316
Rump, S. M. 213 Sheinin, V. I. 90
Russell, B. 43 Sheplak, M. 325
Russell, S. 45 Sherman, J. 70, 72
Russo, F. 45 Shih, C. J. 43
Rustem, B. 334 Shimanovsky, A. 85
Shinozuka, M. ix, 112, 145, 151, 163, Switzky, H. 39

164, 175, 177, 187190, 208, 218,
248 Tagami, S. 90
Shiraishi, N. 45 Tagawa, H. 307
Shirk, M. H. 190 Takagi, J. 114
Sigmund, O. 20, 36 Takeuchi, M. 281
Simiu, E. 14, 15, 339 Takewaki, I. 85, 89, 190, 211
Sin, H-C. 144 Takizawa, Y. 299
Skalna, I. 85, 86, 89 Taleb, N. 15
Slaev, V. A. 150 Tang, W. H. 14
Smallwood, D. O. 336 Taroco, E. O. 284
Smith, I. F. C. 316 Taylor, J. E. 18
Smith, S. xiii, 45 Templeman, A. B. 332
Sniedovich, M. 334336 Tempone, R. xiv, 49, 85
Sobieszczanski-Sobieski, J. 27 Terada, K. 114, 134
Soize, C. 49, 208 Tham, L. G. 61, 274, 325
Sokal, A. x Thimbleby, H. 329
Song, D. 56, 8589, 208 Thoft-Christensen, P. 331
Soong, T. T. 189, 190, 325, 332 Thomas, H. L. 299
Sophie, Q. C. 45 Thompson, J. M. T. 113, 118, 123125,
Soucy, Y. 153 133, 134, 136, 280, 330
Spunt, L. 273 Thouverez, F. 56, 225
Srinivasan, S. 190, 209 Tikhomirov, V. M. 335
St. Clair, U. H. 42 Tonon, F. xiii, 46
Stander, N. 31, 208 Toporov, V. 325
Starks, S. A. 272 Torres, R. 272
Starnes, Jr., J. H. ix, xii, 15, 30, 61, Torvik, P. J. 158
144, 189, 208, 209, 228, 248 Triantafyllidis, N. 114
Stasenko, I. V. 150 Tseng, C. H. 127
Stingl, M. 325 Tseng, W. K. 336
Stix, R. 9 Tucker, W. T. 340
Stolpe, M. 107 Tunguskova, Z. G. 91, 99
Stoyanov, J. 7 Turner, M. J. 190
Striz, A. G. 190 Tuy, H. 23, 32
Stroud, W. J. 45 Tzan, S.-R. 61, 149, 189, 190
Sunaga, T. 49
Sundararaju, K. 43, 45 Uchida, A. 293
Supple, W. J. 123, 280 Uetani, K. 281
Suzuki, K. 208, 225 Ulanov, G. M. 150
Svanberg, K. 107 Utkin, L. V. 45
Svetlitskii, V. A. 150
S
olin, P. 225 Valyi, I. 58, 61, 63, 247
Author Index 401
van Dyke, M. 92 Weisshaar, T. A. 190

van Keulen, F. 211, 277, 279, 325 Wentzel, E. C. 1
van Manen, S. 144 Westermo, B. D. 336
Vandenberghe, L. 107, 308 Whidborne, J. F. 30
Vandepitte, D. 45, 212 Wilson, E. L. 292
Vanderplaats, G. N. 18, 27, 329 Witte, A. F. 336
Vanmarcke, E. H. 39 Wolfram, S. 9
Varaiya, P. 63 Wong, F.-S. xiii, 45
Vassart, N. 316 Wood, K. L. xiii
Vecchi, M. P. 35 Woodbury, M. A. 70, 72
Venini, P. 85, 209, 325 Wu, C. F. 14
Venkataraman, S. 328 Wu, J. 145, 208, 325
Venkayya, V. B. 18, 190 Wu, T. S. 84
Venter, G. 30, 325 Wu, X. M. 70, 208
Verijenko, V. E. 144, 280, 325 Wu, Z. 190
Vermeulen, P. G. 144 Wunderlich, W. 144
Vick, S. G. 46
Vinot, P. 61 Xiang, G. 61, 225, 272
Vitali, F. 225 Xie, L-L. 190
Vittal, S. 274 Xu, C. 43
von Hoerner, S. 299 Xu, L. 274, 332
VR&D 120, 310
Yamada, T. 225
Wald, A. 336 Yamaguchi, T. 209
Wall, F. J. 112 Yamakawa, H. 43, 45, 209
Walshan, D. 15 Yan, S. O. 208
Walster, G. W. 52 Yang, D. 45, 272, 339
Wang, D. 89 Yang, J. L. 209
Wang, G. Y. 43 Yang, J. N. 190, 332
Wang, K. Y. 208 Yang, T. Y. 190
Wang, P. C. 208 Yang, X. W. 70, 85, 89, 112, 208, 225
Wang, W. Q. 43 Yang, Z. J. 70, 208
Wang, X. B. 208 Yao, J. T. P. 45, 46
Wang, X. J. xii, 45, 61, 85, 86, 89, Yau, W. 336
144, 173, 208, 209 Yeh, Y. C. 43
Ward, A. C. 85 Yoshida, N. 27
Warmus, M. 49 Yoshikawa, N. 61, 190, 208, 209, 225,
Watanabe, A. 340 274, 299
Watanabe, S. 336 Youn, B. D. 45, 274
Watson, L. T. 86 Young, R. C. 49
Weibull, W. 3 Youssef, N. A. N. 336
Weiner, J. H. 80 Yuan, B. 42
Yun, C. B. 208 Zhang, J. Y. 102, 103, 105, 108, 293, 316,

323
Zadeh, L. A. 42, 44 Zhang, N. 208
Zaguskin, V. L. 60 Zhang, T. Y. 316
Zako, M. 30, 279 Zhang, Z. Y. 325
Zartarian, C. 191 Zharii, O. Y. 144
Zavoni, E. H. 317 Zheng, G. W. 61, 274
Zeng, G. W. 325 Zheng, Y. 43
Zhai, C-H. 190 Zhou, J. 45
Zhang, H. 53, 211213, 215, 225 Zhu, L. P. ix, xiii, 61, 272, 299, 339
Zhang, J. 340 Zingales, M. 49, 190, 197, 198, 228, 248
Zhang, J. H. 189, 325 Zuccaro, G. 61, 145, 175, 177

1 Optimization & Anti-Optimization of Structures Under Uncertainty - Isaac Elishakoff PDF

Uploaded by

Document Information

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

1 Optimization & Anti-Optimization of Structures Under Uncertainty - Isaac Elishakoff PDF

Uploaded by

Copyright:

Available Formats

Optimization and

P678 tp.indd 1 1/29/10 12:16:33 PM

Imperial College Press

P678 tp.indd 2 1/29/10 12:16:34 PM

British Library Cataloguing-in-Publication Data

OPTIMIZATION AND ANTI-OPTIMIZATION OF STRUCTURES UNDER

To Aron Lev, Chana Batsheva, Yehudit Lea, Menachem Mendel,

To Noriko, Takeshi, and Satsuki

Uncertainties appear everywhere in the model. When using a mathe-

Researchers in applied mechanics, whether in the field of aerospace, architecture,

viii Optimization and Anti-Optimization of Structures under Uncertainty

[the safety] factor is designed to arbitrarily account for items such

It seems absurd to strive for more and more refinement of meth-

Bolotin (1961) had written,

The values of safety factors, as well as closely associated values

We must immediately note that probability theory and mathematical statistics

x Optimization and Anti-Optimization of Structures under Uncertainty

with the probabilistic methodology.

You might be amused by the story of how I came to do the work

deal with Mr. Ozawas problem. I concluded that any method I

The worst-case oriented interval model of uncertainty looks too

These considerations eventually gave rise to the optimization and anti-

xii Optimization and Anti-Optimization of Structures under Uncertainty

amenable to a certain measure of cooperation. An instructive example is the case

Convex modeling of uncertainty, and in general the set-theoretic

Uncertainty triangle (Elishakoff, 1990).

xiv Optimization and Anti-Optimization of Structures under Uncertainty

One should constantly bear in mind that just as pseudo-deterministic analy-

whenever the input data belong to an infinite dimensional space

I. Elishakoff and S. Nakagiri, 1998; Professor Alessandro Baratta and Professor

May 2009 Isaac Elishakoff, Makoto Ohsaki

This page intentionally left blank

2. Optimization or Making the Best in the Presence of

3. General Formulation of Anti-Optimization 47

xviii Optimization and Anti-Optimization of Structures under Uncertainty

4. Anti-Optimization in Static Problems 77

5. Anti-Optimization in Buckling 113

5.5.2 Worst imperfection of optimized frame . . . . . . . . . . 120

6. Anti-Optimization in Vibration 145

xx Optimization and Anti-Optimization of Structures under Uncertainty

6.7.4 Convex modeling of uncertain moduli . . . . . . . . . . 195

7. Anti-Optimization via FEM-based Interval Analysis 211

8. Anti-Optimization and Probabilistic Design 227

8.3.6 Initial imperfection with uniform probability density

9. Hybrid Optimization with Anti-Optimization under Uncertainty or

xxii Optimization and Anti-Optimization of Structures under Uncertainty

9.10 Force identification of prestressed structures . . . . . . . . . . . . 316

10. Concluding Remarks 327

In this chapter, we explain why it is preferable to deal with set-theoretical rather

1.1 Probabilistic Analysis: Bad News

frequently regarded as a kind of magic wand which produces in-

2 Optimization and Anti-Optimization of Structures under Uncertainty

formation out of a void. This is a fallacy; the theory of probability

Probabilistic methods must yield a central quantityreliability of the structure,

4 Optimization and Anti-Optimization of Structures under Uncertainty

Let Pf,all = 106 . Then

6 Optimization and Anti-Optimization of Structures under Uncertainty

The reliability is thus given by

8 Optimization and Anti-Optimization of Structures under Uncertainty

where erfi(z) is an imaginary error function, defined as (Wolfram, 1996, p. 745)

erfi(z) = erf(iz)/i (1.40)

10 Optimization and Anti-Optimization of Structures under Uncertainty

1.2 Probabilistic Analysis: Good News