\documentstyle[IEEEtran]{article}

\tolerance 10000
\begin{document}

\title{Convex Modeling - A Generalization Of Interval Analysis For
Nonprobabilistic Treatment Of Uncertainty}
\author{Isaac Elishakoff}

\maketitle

\begin{abstract} 
This paper reviews some new developments is convex modeling of
uncertainty in structural approaches to engineering. Convex modeling, on
one hand, represents an alternative and complement to traditional
probabilistic methods, and on the other hand, it is a generalization of
the mathematical theory of interval analysis.
\end{abstract}

\auffil{The author is with the Center for Applied Stochastics Research,
Department of Mechanical Engineering, Florida Atlantic University, Boca
Raton, FL 33431-0991. This study has been supported by the National
Science Foundation under award No. MSS-9215698 (Dr. R. C. Chong --
Program Director). Any opinions, findings, and conclusions or
recommendations expressed in this publication are those of the author
and do not necessarily reflect the views of the National Science 
Foundation.}

\section{Probabilistic Models}

A structure can be modeled as a system subject to influences,
disturbances, excitations, and so forth, designated as {\it inputs}. We
are usually interested in the response of the system: its dynamic
displacements, stresses, or strains designated as {\it outputs.} In
formulating the probabilistic model, we first identify the random
variables and/or functions, which can be the characteristics of the
inputs and/or of the system itself. To perform the analysis, we need to
know the joint probability density function of these variables. Very
often the experimental information on the probability densities is
lacking, and the researchers are making unsubstantiated and unverified
assumptions on the density and distribution functions.
The probabilistic analyst in actuality claims: ``Give me the joint
probability densities of random variables involved, and I will calculate
the probability of the mechanical system performing its designated
mission with desired accuracy!" This reminds us of the well-known
statement by Archimedes: ``Give me a firm spot on which to stand, and I
will move the Earth". It should be borne in mind that, as Wentzel
\cite{Wentzel 1988} notes, probabilistic models are ``...viewed to be a
sort of magic wand yielding information from nothing, i.e., from total
ignorance. Those who think so are under a misapprehension since
probability theory is used but to transform data on observed phenomena
to infer the behavior of those which cannot be observed."

\section{Convex Modeling}
Very often, the information on uncertain variables is fragmentary.
Indeterminacy about the uncertain variables involved could be formulated
in terms of these variables belonging to certain sets, such as:
\begin{itemize}
\item[a)]The uncertain parameter $x_j$ is bounded: $$\underline{a}_j\le
x_j\le\overline{a}_j,$$ where $\underline{a}_j$ and $\overline{a}_j$ are
lower and upper values, respectively, that the variable $x_j$ can take.
\item[b)]The uncertain function has envelope bounds:
$$\underline{x}(t)\le x(t)\le\overline{x}(t),$$ where 
$\underline{x}(t)$ and $\overline{x}(t)$ are deterministic functions
which delimit the range of variation of $x(t)$. 
\item[c)]The uncertain function has an integral square bound:
$$\int x^2(t)\,dt\le a.$$
\item[d)]The Fourier coefficients $A$ of the uncertain function $x(t)$
belong to an ellipsoidal set $$A^TWA\le a^2,$$ where $W$ is the weighing
matrix.
\end{itemize}
All above models represent convex sets. As is seen, the information
contents of a convex model is usually a bound or constraint which
defines an infinite set of values of functional forms which an uncertain
variable may assume. Let us consider a set-theoretic approach to
modeling the uncertainty of a time dependent macroscopic vector function
$f$. Let $\Gamma$ be a set of vector-values functions. For a positive
integer $n$, consider the set of functions:
$$F_n=\{f\,|\,f(t)={1\over n}\sum_{i=1}^n g_i(t)\ {\rm for}\
g_i\in\Gamma, i=1,2,...,n\}.$$
$F_n$ is the set of $n-$fold averages of vector functions in $\Gamma$.
It is known \cite{Artstein 1974} that as $n\to\infty$, the sequence of
sets $F_n$ converges to the convex hull of $\Gamma.$ Or, in physical
terms, if a macroscopic time-dependent vector $f(t)$ is formed as the
superposition of numerous microscopic time-varying events $g_i(t)$
chosen from a set $\Gamma$, then the set of all such functions $f(t)$
will tend to be convex, regardless of the structure of the set $\Gamma$.

The convex modeling of uncertainty produces:
\begin{itemize}
\item  maximum, or least favorable
response of the structure, and 
\item
minimum, a best favorable response. 
\end{itemize}
It was applied to:
\begin{itemize}
\item vehicles moving on the rough surface with bounded terrain
\cite{Elishakoff Ben-Haim 1990,Elishakoff Cai Starnes 1994};
\item stress concentration in elastic sheets with bounded irregularities
\cite{Givoli Elishakoff 1992};
\item optimization of trusses \cite{Elishakoff Haftka Fang 1994},
\item and several other problems.
\end{itemize}

\section{Convex Modeling Represents a Generalization of Interval
Analysis}
Indeed, type a) of the above four constraints is a representation of an
interval variable. For this type of convex models, one can utilize the
ample knowledge accumulated in interval analysis:
\begin{itemize}
\item 
Elishakoff and Duan  \cite{Elishakoff Duan 1994} have utilized interval
analysis to deal with interval masses and stiffnesses of the system.
\item Analogously, Qiu et al \cite{Qiu et al 1992,Qiu et al 1994} developed
interval results for natural frequencies, whereas 
\item K\"oyl\"uoglu et al \cite{K et al 1994} developed interval
finite element analysis.
\end{itemize}

The lecture presents some recent developments of convex (and, in
particular, of interval) analysis of structures to analyze uncertainty,
and future research needs. 

\begin{thebibliography}{99}

\bibitem{Artstein 1974}
Z. Artstein, ``On the calculus of closed set-valued functions", {\it
Indiana University Mathematics Journal}, 1974, Vol. 24, pp. 433--441.

\bibitem{Ben-Haim Elishakoff 1989} 
Y. Ben-Haim and I. Elishakoff,
``Dynamics and failure of a thin bar with unknown but bounded
imperfections", In: D. Hui and N.~Jones (eds.), {\bf Recent advances in
impact dynamics of engineering structures}, 1989, AND-Vol. 105, AD-Vol.
17, ASME, N.Y., 1989, pp. 89--96.

\bibitem{}
Y.~Ben-Haim and I. Elishakoff,
``Non-probabilistic models of uncertainty in the buckling of shells with
general imperfections: theoretical derivation of the knockdown factor",
{\it Journal of Applied Mechanics}, 1989, Vol. 56, pp. 403--410.

\bibitem{}
Y. Ben-Haim and I. Elishakoff, {\bf Convex models of uncertainty in
applied mechanics}, Elsevier Publ., Amsterdam, 1990. 

\bibitem{}
Y. Ben-Haim and I. Elishakoff, ``Convex models of vehicle response to
unknown but bounded terrain'', In: H, Chung (ed.), 
{\it Proc. of the ASME Pressure Vessels
and Piping Conference}, Honolulu, Hawaii, 1989, pp. 81--88; extended
version: {\it Journal of Applied Mechanics}, 1991, Vol. 58, pp. 354--361.

\bibitem{}
F. L. Chernousko, {\bf State Estimation for Dynamic Systems}, CRC Press,
Boca Raton, 1994. 

\bibitem{}
I. Elishakoff, ``Convex versus probabilistic modeling of uncertainty
in structural dynamics'', In: M. Petyt, H. F. Wolfe, and C. Mei
(eds.), {\bf Structural Dynamics: Recent Advances}, Elsevier Applied
Science Piblishers, Lonodn, 1991, pp. 3--21.

\bibitem{}
I. Elishakoff, ``Some questions in engineering eigenvalue problems in
natural sciences'', In: J. Albrecht, L. Collatz, P. Hagedorn, and W.
Velte (eds.), {\bf Numerical Treatment of Eigenvalue Problems}, Intl.
Series of Numerical Mathematics, Vol. 96, Birkh\"auser Publishers,
Basel, Switzerland, 1991, pp. 71--107. 

\bibitem{}
I. Elishakoff, ``Essay on probability, convexity, and earthquake
engineering'', {\it Proceedings of the 5th Italian National Conference
on Earthquake Engineering ``L'Ingegneria Sismica in Italia''}, 1991,
Vol. 2, pp. 907--935. 

\bibitem{}
I. Elishakoff, ``Essay on reliability index, probabilistics
interpretation of safety factor, and convex models of uncertainty'',
In: F. Casciati and J. B. Roberts (eds.), {\bf Reliability Problems:
General Principles and Applications in Mechanics of Solids and
Structures}, Springer Verlag, 1991, pp. 237--271.

\bibitem{}
I. Elishakoff, ``Essay on uncertainties in elastic and viscoelastoc
structures: from A.~M.~Freudental's criticisms to modern convex
modeling'', {\it Intl. Journal of Computers and Structures}, 1995 (to
appear). 

\bibitem{Elishakoff Ben-Haim 1990}
I. Elishakoff and Y. Ben-Haim, ``Probabilistic and convex models of
uncertainty in structural dynamics'', {\it Proceedings of the 8th
Intl. Modal Analysis Conference}, Orlando, 1990, pp. 1487--1492.

\bibitem{}
I. Elishakoff and Y. Ben-Haim, ``Dynaimcs of a thin cylindrical shell
under impact with limited deterministic information on its initial
imperfections'', {\it Journal of Structured Safety}, 1990, Vol. 8, pp.
103--112.

\bibitem{}
I. Elishakoff, G. Q. Cai, and J. H. Starnes, Jr., ``Non-linear
buckling of a column with initial imperfections via stochastic and
non-stochastic, convex models'', {\it Intl. Journal of Non-Linear
Meahcnics}, 1994, Vol. 29, pp. 71--82. 

\bibitem{Elishakoff Cai Starnes 1994}
I. Elishakoff, G. Q. Cai, and J. H. Starnes, Jr., ``Probabilistic and
convex models of uncertainty in buckling of structures", 
In: G.~I.~Schu\"eller, M. Shinozuka, and J. Yao (eds.), {\it Proceedings of
the 6th Intl. Conference on Structural Safety and Reliability}, Balkema
Publ., Rotterdam, 1994, Vol. 2, pp. 761--766. 

\bibitem{}
I. Elishakoff and P. Colombi, ``Combination of probabilistic and 
convex models of uncertainty when scarce knowledge is present on
acoustic excitation parameters'', {\it Computer Methods in Applied
Mathematics and Engineering}, 1993, Vol. 104, pp. 187--209. 

\bibitem{}
I. Elishakoff and P. Colombi, ``Ideas of probability and convexity
combined for analyzing parameter uncertainty'', 
In: G. I.
Schu\"eller, M.~Shinozuka, and J. Yao (eds.), {\it Proceedings of
the 6th Intl. Conference on Structural Safety and Reliability}, Balkema
Publ., Rotterdam, 1994, Vol. 2, pp. 109--113. 

\bibitem{Elishakoff Duan 1994}
I. Elishakoff and Dehe Duan, ``Application of the mathematical theory
of interval analysis for uncertain vibrations'', In: J. M. Cuschieri,
S.~A.~L.~Glegg, and D. M. Yeager (eds.), {\it Proceedings of the 1994
National Conference on Noise Control Engineering}, N.Y., 1994, pp.
519--524.

\bibitem{}
I. Elishakoff, P. Eliseeff, amd S. Glegg, ``Convex modeling of
material uncertainty in vibrations of a viscoelastic structure'', 
{\it American Institute of Aeronautics and Astronautics (AIAA)
Journal}, 1994, Vol. 32, pp. 843--849. 

\bibitem{}
I. Elishakoff and J. J. Fang, ``Diagnosis of local modifications in
the boundary conditions of a rectangular plate'', {\it Computer
Methods in Applied Mechanics and Engineering}, 1995 (to appear). 

\bibitem{}
I. Elishakoff and J. J. Fang, ``A new safety factor based on convex
modeling'', {\it Journal of Structural Safety}, 1995 (to appear). 

\bibitem{}
I. Elishakoff, Yosi Gana-Snvili, and D. Givoli, ``Treatment of
uncertain imperfections as a convex optimization problem'', In: L.
Esteva and S. E. Ruiz (eds.), {\it Proceedings of the Sixth Intl.
Conference on Applications of Statistics and Probability in Civil
Engineering}, Mexico, 1991, Vol. 1, pp. 150--157. 

\bibitem{Elishakoff Haftka Fang 1994}
I. Elishakoff, R. T. Haftka, and J.~J.~Fang, ``Structural design under
bounded uncertainty optimization with antioptimization", {\it Intl.
Journal of Computers and Structures}, 1995 (to appear).

\bibitem{}
I. Elishakoff, C. Q. Li, and M. Shinozuka, ``Hybrid probabilistic and
convex modeling for robust reliability analysis of structures'', {\it
Intl. Journal of Solids and Structures}, 1995 (to appear). 

\bibitem{}
I. Elishakoff, Y. W. Li, and J. H. Starnes, Jr., ``A deterministic
method to predict the buckling locad variability due to uncertain
elastic moduli'', {\it Computer methods in Applied Mechanics and
Engineering}, 1994, Vol. 111, pp. 155--167.

\bibitem{}
I. Elishakoff, Y. K. Lin, and L. P. Zhu, {\bf Probabilistics and
convex modeling of acoustucally excited structures}, Elsevier Science,
Amsterdam, 1994. 

\bibitem{}
I. Elishakoff and B. Pletner, ``Analysis of base excitation as an
uncertain factor with specific bounds on it and its derivatives'', In:
T.~C.~Huang et al (eds.), {\bf Structural Vibrations and Acoustics},
ASME Press, 1991, pp. 177-184. 

\bibitem{}
J. J. Fang and I. Elishakoff, ``Convex identification of boundary
conditions by the finite element method'', {\it Journal of Shock and
Vibration}, 1995 (to appear). 

\bibitem{Givoli Elishakoff 1992}
D. Givoli and I. Elishakoff, ``Stress concentration at a nearly
circular hole with uncertain irregularities'', ASME Paper
91--WA/APM--22, {\it Journal of Applied Mechanics}, 1992, Vol. 59, pp.
65--71.

\bibitem{}
D. Givoli, I. Elishakoff, and Y. Stavsky, ``A boundary-perturbation
finite-element method for plane elasticity problems'', {\it Computer
methods in Applied Mehcanics and Engineering}, 1992, Vol. 96, pp.
45--63. 

\bibitem{K et al 1994}
H. U. K\"oyl\"uoglu, A. S. Cakmak, and S.~R.~K.~Nielsen,
{\bf Applications of interval mapping in mechanics for structural
uncertainties and pattern loading}, Department of Building Technology
and Structural Enginering, Aaborg University, Ppare No. 121, 1994. 

\bibitem{}
Y. W. Li, I. Elishakoff, J. H. Starnes, Jr., and M. Shinozuka,
``Prediction of natural frequency and buckling load variability by 
convex modeling'', {\it Fields Institute Communications}, American
Math. Society, 1995 (to appear).

\bibitem{Qiu et al 1992}
Z. P. Qiu, S. H. Chen, and J. X. Na, ``The Rayleigh quotient method
for computinfg eigenvalue bounds of vibrationsl systems with interval
parameters'', {\it Acta Mehcanica Sinica}, 1993, Vol. 6, pp. 309--318. 

\bibitem{Qiu et al 1994}
Z. P. Qiu, S. H. Chen, and I. Elishakoff, ``Natural frequenices of
structures with uncertain but nonrandom parameters'', {\it Journal of
Optimization Theory and Applications}, 1995 (to appear). 

\bibitem{}
F. C. Schweppe, {\bf Uncertain Dynamical Systems}, Prentice Hall,
Englewood Cliffs, NJ, 1973.

\bibitem{Wentzel 1988}
E. Wentzel, {\bf Operations Research}, Mir Publishers, Moscow, 1988,
p. 28.

\end{thebibliography}

\end{document}