ICIAM 95 July, 3 -- July, 7 PRELIMINARY SCHEDULE *********************************************************************** Geometric Modeling with Free Form Surfaces Organizer: K. H{\"o}llig, Universit{\"a}t Stuttgart, Germany Summary: Nonuniform rational B-Splines (NURBS) have become a widely accepted standard in geometric modeling. A variety of algorithms and approximation methods exist and have been implemented in commercially available software. Nevertheless, a fundamental unsolved problem remains: The modeling of smooth surfaces of arbitrary topology. \smallskip In this minisymposium several new techniques for modeling surfaces over irregular control nets are presented. Moreover, as an industrial application, methods for constructing blending surfaces are discussed. Day: July 3, 1995 Time: 10.00 K. H{\"o}llig, Universit{\"a}t Stuttgart, Germany: Some current issues in surface modeling H.-P. Seidel, Universit{\"a}t Erlangen, Germany; R. Pfeifle, Universit{\"a}t Erlangen, Germany: Polar forms and triangular B-splines S. Steiner, Mercedes Benz AG, Germany: Generating B-spline blendings that fit to given boundary strips J. Peters, Purdue University, USA: Spline surfaces from irregular control meshes U. Reif, Universit{\"a}t Stuttgart, Germany: Modeling smooth free form surfaces of arbitrary genus by degenerated B{\'e}zier patches *********************************************************************** Nonlinear Iteration and Verification of Solution Organizer: T. Yamamoto, Ehime University, Japan Summary: In scientific computing, much attention has recently been paid to a problem of determining the quality of computed results. \smallskip This minisymposium focuses on the state-of-the-art in verification methods for solution of discrete/continuous nonlinear problems. \smallskip The speakers will present recent theoretical and computational results on this subject. All presentations will begin with a tutorial overview. Day: July 3, 1995 Time: 15.30 G. Alefeld, Universit{\"a}t Karlsruhe, Germany: On the verification of solutions of nonlinear systems A.J. Frommer, Bergische Universit{\"a}t-GH Wuppertal, Germany: Asynchronous iterations for enclosing solutions M. Kashiwagi, Waseda University, Japan: An all solution algorithm with guaranteed accuracy using interval analysis S. Oishi, Waseda University, Japan: Numerical existence theorems for solutions of nonlinear boundary value problems for ODE M.T. Nakao, Kyushu University, Japan: Numerical verification methods for solutions of nonlinear elliptic equations with applications to nondifferentiable problems T. Yamamoto, Ehime University, Japan: Split nonsmooth equations in $\R ^n$ and verification of solution *********************************************************************** Information-Based Complexity Organizers: E. Novak, Universit{\"a}t Erlangen, Germany H. Wo\'zniakowski, Columbia University, USA and University of Warsaw, Poland Summary: Problems in science and engineering often have continuous models. Information-based complexity studies the intrinsic difficulty of approximate solutions of continuous problems defined in spaces of functions, usually of several variables. Many multivariate problems are intractable in the worst case deterministic setting, i.e., their complexity grows exponentially with the number of variables. One of the main subjects of this minisymposium is the study of whether intractability can be broken by randomization or by settling for an average or probabilistic assurance. \smallskip Other subjects include noisy information and adaption. Applications of complexity results to a number of different areas will also be presented. Day: July 4, 1995 Time: 09.30 H. Wozniakowski, Columbia University, USA: An introduction to information-based complexity S. Heinrich, Universit{\"a}t Kaiserslautern, Germany: Monte Carlo algorithms and complexity theory A. Keller, Universit{\"a}t Kaiserslautern, Germany: Quasi-Monte Carlo methods in computer graphics G. Wasilkowski, University of Kentucky, USA; K. Ritter, Universit{\"a}t Erlangen, Germany: Integration and $L_2$-approximation: average case setting with isotropic Wiener measure for smooth functions K. Ritter, Universit{\"a}t Erlangen-N{\"u}rnberg, Germany: Some average case results for nonlinear problems Day: July 4, 1995 Time: 15.30 E. Novak, Universit{\"a}t Erlangen-N{\"u}rnberg, Germany: Error bounds for adaptive methods P. Math{\'e}, WIAS, Germany: Complexity issues of sampling L. Plaskota, University of Warsaw, Poland: Complexity of problems with noisy information K. Frank, Universit{\"a}t Kaiserslautern, Germany: On the complexity of local solution of multivariate integral equations H. Wozniakowski, Columbia University, USA: Complexity of multivariate problems with applications to path integrals *********************************************************************** Aerodynamic Optimization and Drag Reduction Organizer: A. Nastase, RWTH Aachen, Germany Summary: A modern flying configuration (FC) must perform its mission with minimal fuel consumption in order to be competitive, economic and ecological. \smallskip For the aerodynamic point of view the necessary drag reduction can be obtained by shape optimization and by boundary layer control. \smallskip There is a great difference between the {\bf aerodynamical optimal design} and the {\bf computational fluid dynamics}. \smallskip The aerodynamic optimal design gives the possibility of the insight design i.e. {\bf to modify the shape} of the flying configuration in order to obtain a shape of minimum aerodynamic drag, under some constraints. The computational fluid dynamics can accurately {\bf determine the flow over a given shape} of the flying configuration without to be able to modify it in order to attain a certain goal. \smallskip Several mathematical strategies for the performing of the global aerodynamic optimization of the shape of the flying configurations under taking into consideration of the multipoint and multidisciplinary design are here presented, discussed and compared. Day: July 5, 1995 Time: 15.30 I.-M. Navon, Florida State University at Tallahassee, USA: Large scale unconstrained and constrained optimization methods and their applications A. Nastase, RWTH Aachen, Germany: The drag reduction via aerodynamic global shape optimization C. Poloni, Universita di Trieste, Italy; G. Mosetti, Universit{\`a} di Trieste, Italy: Aerodynamic shape optimization by means of a hybrid genetic algorithm J.E. Chacksfield, British Aerospace Defence, United Kingdom: Mathematical optimization techniques and their impact on the aircraft design scenario Day: July 7, 1995 Time: 15.30 J. Periaux, Dassault Aviation, France; H.K. Chen, B. Mantel, B. Stoufflet, Dassault Aviation, France; A. Dervieux, J.-M. Male, N. Marco, INRIA, Sophia Antipolis, France: Optimal shape design with and without derivatives for aerodynamic applications P.-A. Lambert, SNECMA Villaroche Test Centre, France; J.-L. Lecordix, SNECMA, Moissy Cramayel, France; V. Braibant, Ecole Nationale Sup{\'e}rieure de Cachan, France: Constrained optimization of nacelle shapes in axisymmetrical Euler flow P. Chaviaropoulos, Center of Renewable Energy Sources, Greece: Optimization of stall regulated horizontal axis wind turbine rotors T. Labrujere, National Aerospace Laboratory, The Netherlands; C. Hendriks, National Aerospace Laboratory, Amsterdam, The Netherlands: A comparison of deterministic and probabilistic approaches to aerodynamic multi-point inverse airfoil design *********************************************************************** Controllability and Stabilizability of Distributed Systems Organizers: J.-P. Puel, Universit{\'e} de Versailles, France E. Zuazua, Universidad Complutense de Madrid, Spain Summary: This minisymposium is devoted to the controllability and stabilizability of Distributed Systems described in terms of Partial Differential Equations (PDE). Their analysis from a control theoretical point of view is relevant in many applications: flexible structures, robots, noise reduction in aeronautics, ... \smallskip Recently research in this area has been very active and important advances have been made. Various methods and analytical techniques have been developed to address the diversity of problems that the control and stabilization of PDE presents, most of them being specific to the infinite-dimensional character of the underlying dynamics. \smallskip Recent results in this area will be presented. Day: July 4, 1995 Time: 09.30 J.-M. Coron, Ecole Normale Superieure de Cachan, France: On the controllability of $2-D$ incompressible perfect fluids C. Fabre, Ecole Polytechnique, France: Insensitizing controls for parabolic equations G. Lebeau, Universit{\'e} de Paris-Sud, France: Micro-analyticity and control for the wave equation E. Zuazua, Universidad Complutense de Madrid, Spain; S. Micu, Universidad Complutense de Madrid, Spain: Analysis and control of a model of noise reduction L. Robbiano, Universit{\'e} de Paris Val de Marne, France: Carleman inequalities and control M. Tucsnak, Ecole Polytechnique, France: Control of some elastic structures by means of piezoelectric actuators *********************************************************************** Mathematical Finance Organizers: P. Embrechts, ETH Z{\"u}rich, Switzerland P. Wilmott, Oxford University, United Kingdom Summary: Though the subject `Mathematical Finance' can be traced back to the turn of the century (through Bachelier's work), it was mainly 70 years later that financial options were actually traded on an official exchange. The year 1973 marks the opening of the Chicago Board Options Exchange, as well as the appearance of the fundamental Black-Scholes paper on option pricing. Over the recent years, financial products have been created with rather exotic payoffs. Examples include path-dependent options, catastrophic insurance futures, quantos etc.. The latter for instance relate to products denominated in one currency, but paid out in another. Invariably, these new products lead to various interesting questions in applied mathematics. Examples of such questions will be discussed. Day: July 4, 1995 Time: 09.30 P. Wilmott, Oxford University, United Kingdom: to be announced C. Alexander, University of Sussex, United Kingdom: to be announced C. Kl{\"u}ppelberg, ETH-Zentrum Z{\"u}rich, Switzerland: Modelling time series with infinite variance C. Hipp, Universit{\"a}t Karlsruhe (TH), Germany: Financial risk management and insurance business *********************************************************************** Applied Mathematics and Theoretical Mechanics in Ship and Ocean Engineering Organizer: O.M. Faltinsen, The Norwegian Institute of Technology, Norway Summary: An historical overview from Euler until today of the interaction of applied mathematics and mechanics with ships are presented. Hydrodynamic and structural aspects of ships and offshore structures are discussed. This includes resistance and manoeuvring of ships, flow around foils and propulsion units, hydroelasticity, wave induced motions and loads on ships and offshore structures. The presence of the free water surface represents a challenge in developing reliable numerical method. Nonlinearities due to body motions and free surface waves as well as interactions between the free surface and a viscous wake are discussed. Day: July 5, 1995 Time: 15.30 O. Faltinsen, The Norwegian Institute of Technology, Norway: The recent role of mathematical methods in the design of high speed ships U.P. Bulgarelli, INSEAN, Italy; P. Bassanini, Universit{\`a} di Roma "La Sapienza", Italy: On the application of intensive computational methods on the prediction of flows about ships B. Molin, Ecole Sup{\'e}rieure d'Ing{\'e}nieurs de Marseille, France; R. Cointe, Paris, France: The computational prediction of wave loadings on large ocean structures - a success for theory T. Miloh, University of Tel Aviv, Israel: Optimal ship collision avoidance - a differential game approach K.V. Rozhdestvensky, St. Petersburg State Marine TU, Russia: Matched expansions method in hydromechanics of lifting bodies M. Tulin, University of California at Santa Barbara, USA: The interaction of applied mathematics and mechanics with ships - an historical overview *********************************************************************** New Algorithms for High Performance Computing Organizer: R. Schreiber, RIACS, CA, USA Summary: The minisymposium features talks by developers of state-of-the-art algorithms in several central topics in scientific computing. The new algorithmic developments to be discussed are in most cases motivated by developments in computer architecture, particularly the emerging highly parallel, distributed memory model of computing. \smallskip The areas covered include iterative solution of large sparse linear systems (Freund, Plassmann), direct solution of large sparse linear systems (Rothberg, Gilbert), mapping unstructured meshes to distributed memory machines (Leland, Gilbert, Plassmann), matrix computation libraries for advanced architectures (Dongarra) and the use of high performance architectures for difficult applications (Freund, Plassman). Day: July 5, 1995 Time: 09.30 P. Plassmann, Argonne National Laboratory, USA; L.A. Freitag, Argonne National Laboratory, USA; M.T. Jones, University of Tennessee, USA: The efficient parallel solution of linear systems arising from unstructured meshes E. Rothberg, Supercomputer Systems Division, USA: A scalable approach to sparse Cholesky factorization J.R. Gilbert, Xerox Palo Alto Research Center, USA: Geometric spectral mesh partitioning --or-- Solving nonsymmetric sparse linear systems on high-performance computers J. Dongarra, The University of Tennessee, USA: Library software and tools for numerical linear algebra R. Freund, AT{\&}T Bell Laboratories, USA: Iterative methods in device and circuit simulation R. Leland, Sandia National Laboratories, USA: Mapping unstructured meshes to distributed memory machines *********************************************************************** Models and Methods of Stochastic Geometry and Spatial Statistics Organizer: D. Stoyan, Technische Universit{\"a}t - Bergakademie Freiberg, Germany Summary: Stochastic geometry studies models of random geometrical structures such as point processes, fibre processes or random sets. Spatial statistics provides the necessary methods for fitting models and parameter estimation. Important fields of application are materials science and structural geology. Four lectures present methods for materials characterization: one discusses general statistical strategies, two measurement problems and one a random growth model. A further lecture uses similar methods, namely marked point processes in a Bayesian framework, for evaluation of petroleum reservoirs. Finally, methods for the statistical analysis of samples of particles are discussed. Day: July 6, 1995 Time: 15.30 A.F. Karr, National Institute of Statistical Sciences, USA: Microstructural design of materials: key issues and statistical strategies J. Ohser, TU Bergakademie Freiberg, Germany: Random Minkowski measures and digitalization H. Omre, Norwegian Institute of Technology - NIT, Norway: Marked point models in evaluation of petroleum reservoirs I. Saxl, Academy of Sciences of the Czech Republic, Czech Republic: Particle arrangement by distance and polygonal methods: application in materials science D. Stoyan, TU Bergakademie Freiberg, Germany: Statistical analysis of random particles T. Liebling, Ecole polytechnique f{\'e}d{\'e}rale de Lausanne, Switzerland: $3D$ dynamic power diagram polycrystal growth model *********************************************************************** Semiconductor Device Simulation Organizers: H. Gajewski, WIAS Berlin, Germany A. Arnold, Technische Universit{\"a}t Berlin, Germany Summary: Modeling and numerical simulation of semiconductor devices is of primal interest for modern high--technology industries. The scope of the minisymposium is to discuss current results in mathematical modeling and numerical simulation of semiconductor devices including nanostructures. Accordingly, topics to be adressed by the speakers are: new aspects in phenomenological device models, kinetic models, quantum mechanic models. Special attention is paid to new numerical tools.\\ Device simulation is based on sytems of nonlinear partial differential equations arising in many branches of applied mathematics. Because of its technological relevance, this topic has received much attention and achieved valuable results in the last years. Day: July 4, 1995 Time: 15.30 L.L. Bonilla, University Carlos III, Spain: Theory, asymptotics and numerics of current instabilities in semiconductor superlattices A. Arnold, Technische Universit{\"a}t Berlin, Germany; C. Ringhofer, Arizona State University, Tempe, USA: Numerical coupling of kinetic and hydrodynamic semiconductor models C. Ringhofer, Arizona State University, USA: Extension of the hydrodynamic model equations for the simulation of transport in semiconductor materials T. Kerkhoven, University of Illinois, USA: Fast 3-dimensional Schr{\"o}dinger-Poisson computations for nanostructure G. Albinus, WIAS Berlin, Germany: Carrier transport and self-consistent thermal behaviour K. G{\"a}rtner, ETH-Zentrum Z{\"u}rich, Switzerland; H. Gajewski, WIAS Berlin, Germany: On the discretization of van Roosbroeck's equations with magnetic field *********************************************************************** Modeling Thin Structures in Mechanics -- Asymptotic and Finite Element Analysis Organizers: C. Schwab, University of Maryland, Baltimore, USA M. Suri, University of Maryland, Baltimore, USA Summary: Plates, beams and shells (including laminated structures are basic to computational engineering mechanics. They are usually based on hierarchic families of plate or shell {\it models} obtained from three-dimensional problems by {\it dimension-reduction}. Finite element discretizations must deal with {\it shear} and {\it membrane locking} and accurately resolve {\it boundary layers}. We discuss locking-free {\it mixed} and {\it high order} finite elements for plates and shells, asymptotic analysis of plate and shell models, a-posteriori estimation of the {\it dimension reduction error} and {\it adaptive model selection}. {\it Hierarchic modeling of thin, laminated composites} will also be addressed. Day: July 6, 1995 Time: 09.30 D.N. Arnold, Pennsylvania State University, USA: Locking free finite elements for shells J. Pitk{\"a}ranta, Helsinki University of Technology, Finland; J. Piila, Helsinki University of Technology, Finland: On the asymptotics of shell deformations B.A. Szabo, Washington University, USA: Hierarchic models of laminated composites E. Stein, Universit{\"a}t Hannover, Germany; S. Ohnimus, Universit{\"a}t Hannover, Germany: Integrated model -- and -- solution adaptivity in the $FEM$-analysis of plates and shells R. Stenberg, Helsinki University of Technology, Finland: Stabilized plate bending elements M. Suri, University of Maryland Baltimore County, USA; C. Schwab, University of Maryland, Baltimore, USA: to be announced *********************************************************************** Mathematics in Electronic Industry Organizers: P. Rentrop, Technische Hochschule Darmstadt, Germany A. Gilg, Siemens AG M{\"u}nchen, Germany Summary: Numerical simulation is a necessary tool in electronic industry. In this minisymposium we will concentrate on the simulation chain {\it process simulation -- device simulation -- electric circuits simulation.}\\ The process simulation treats the local manufactoring steps of integrated circuits. Etching, implantation or oxidation are typical simulation tasks. The mathematical problem leads to free boundary value problems for systems of partial differential equations.\\ In the device simulation the characteristics of MOSFETs should be computed. The drift-diffusion equations or more refined models are the mathematical base. Due to different time scales adaptive techniques for mesh refinement are essential.\\ An electric circuits simulator (a typical package count up to 150.000 lines of FORTRAN code) must be able to simulate circuits up to several thousands of basic electrical elements: like MOSFETs, capacitors, resistors. The mathematical model leads to an implicit system of nonlinear ordinary differential equations.\\ In the simulation chain the circuit simulator uses the MOSFET characteristics from device simulation via a Table model. The dotation profiles, which are typical results of the process simulation, define the input of the device simulation. Changes in the manufactoring process can be used for the electric circuits performance. Day: July 7, 1995 Time: 09.30 R.H.W. Hoppe, Technische Universit{\"a}t M{\"u}nchen, Germany; R. Hiptmair, Technische Universit{\"a}t M{\"u}nchen, Germany; B. Wohlmuth, Technische Universit{\"a}t M{\"u}nchen, Germany: Mixed finite elements and their application in semiconductor device simulation A.M. Anile, Universit{\'a} di Catania, Italy: Hydrodynamical models in device simulation W.H. Schilders, Philips Research Laboratories, The Netherlands: Numerical methods for semiconductor device simulation M. G{\"u}nther, Technische Universit{\"a}t M{\"u}nchen, Germany; P. Rentrop, TH Darmstadt, Germany: The differential-algebraic index concept in electric circuit simulation Q. Zheng, Siemens AG M{\"u}nchen, Germany; U. Wever, Siemens AG, M{\"u}nchen, Germany: Circuit simulation on workstation cluster *********************************************************************** Progress in Computational Geometry Organizer: H. Imai, University of Tokyo, Japan Summary: Computational geometry treats geometric problems from a unified standpoint based on algorithm design and complexity theory. Geometric problems arise in many fields such as combinatorial optimization, computer graphics, image processing, mesh generation, robotics, geographical information processing, etc., so that algorithms developed in computational geometry may have big impact in such large fields. \smallskip This minisymposium features several invited talks on computational geometry which have strong interactions with other fields related to SIAM. Topics covered in this minisymposium are numerical issues in implementing geometric algorithms, enumeration paradigm, good mesh generation, geometric clustering, software packages of computational geometry, etc. Day: July 5, 1995 Time: 09.30 K. Sugihara, University of Tokyo, Japan: Numerical issues and practical implementations in computational geometry D. Avis, McGill University, Canada: Large convex hull problems T. Tokuyama, IBM Tokyo Research Laboratory, Japan: Computational-geometric approach to transportation problems T.-S. Tan, National University of Singapore, Republic of Singapore: Optimal triangulation problems Day: July 6, 1995 Time: 09.30 T. Masada, University of Tokyo, Japan: An output size sensitive algorithm for the enumeration of regular triangulations S. N{\"a}her, M.-Luther Universit{\"a}t Halle-Wittenberg, Germany: LEDA: a platform for combinatorical and geometric computing H. Imai, University of Tokyo, Japan; M. Inaba, University of Tokyo, Japan: Geometric clustering with applications *********************************************************************** Differential-Algebraic Equations Organizers: E. Hairer, University of Geneva, Switzerland C. Lubich, Universit{\"a}t T{\"u}bingen, Germany Summary: When the flow of a differential equation is constrained to a manifold we speak of a differential-algebraic equation (DAE). Such problems arise in a variety of applications, e.g., mechanical systems, simulation of electrical networks, chemical reaction kinetics and control engineering. \smallskip For the numerical treatment of DAEs standard ODE-methods can often be adjusted, but new efficient algorithms have been designed for special DAEs. The construction of new methods and the analysis of their stability and convergence properties complement and require each other. Recent research on such problems will be the subject of the minisymposium. Day: July 7, 1995 Time: 15.30 U.M. Ascher, University of British Columbia, Canada: Stabilization for the shape-from-shading problem S.L. Campbell, North Carolina State University, USA; E. Moore, Western Illinois University, USA; Y. Zhong, North Carolina State University, USA: Constraint preserving integrators for unstructured higher index DAEs C. F{\"u}hrer, DLR, Germany; Arevalo, Universidad Simon Bolivar, Caracas, Venezuela; G. S{\"o}derling, Lund University, Sweden: Generalizing multistep methods for index-2 differential algebraic equations V. Mehrmann, Technische Universit{\"a}t Chemnitz-Zwickau, Germany; P. Kunkel, Carl von Ossietzky Universit{\"a}t, Oldenburg, Germany: New discretization methods for general, linear, higher index DAE's with variable coefficients L.R. Petzold, University of Minnesota, USA: DAEs and the stability of moving mesh systems of partial differential equations W.C. Rheinboldt, University of Pittsburgh, USA; P.J. Rabier, University of Pittsburgh, USA: Discontinuous solutions of semilinear differential algebraic equations *********************************************************************** Blow Up for Parabolic Equations Organizer: M. Herrero, Universidad Complutense de Madrid, Spain Summary: The purpose of this Minisymposium is to provide an updated account of recent work on singularity formation for solutions of equations and systems of parabolic type. These last constitute a field of active research in view, among other things, of their relevance as models in Continuum Mechanics. \smallskip Concerning the contents of the Session, topics to be discussed include \smallskip \item{1.} Reaction -- Diffusion systems of interest in combustion theory, \item{2.} Geometrical aspects of evolution of surfaces in space, \item{3.} Identification of critical blow--up parameters, \item{4.} Characterization of blow--up patterns and their stability properties. Day: July 7, 1995 Time: 09.30 M.A. Herrero, Universidad Complutense de Madrid, Spain: Blow up for parabolic equations: an overview M. Escobedo, Universidad del Pa{\'i}s Vasco, Spain: Critical exponents in parabolic blow up problems A.A. Lacey, Heriot-Watt University, Scotland: Blow-up for parabolic systems J.J.L. Vel{\'a}zquez, Universidad Complutense de Madrid S/N, Spain: Singularity analysis for mean curvature flow A. Bressan, Universita di Trieste, Italy: Stable and unstable blow-up patterns *********************************************************************** Boundary Variation Organizer: J.P. Zol{\'e}sio, CNRS, France Summary: The modelling of many non cylindrical evolution problems, of coupled fluid structures, of steady and dynamical free boundary problems is realized via a relaxed formulation of the problem in which the domain and its boundary need not be smooth. In order to relax the usual manifold structure of the boundary we developed an intrinsic geometry and generalized curvature.... Applications are presented, in particular, a new shell analysis, capacitary analysis for existence, new free boundary conditions for steady Navier-Stokes free boundary problems, control of free boundaries in hydrodynamics and the shape derivative of the spectrum of a shell.\\ This minisymposia should be an introduction to intrinsic geometry and its application to control theory. Day: July 3, 1995 Time: 10.00 M.C. Delfour, Universit{\'e} de Montr{\'e}al, Canada: New intrinsic models for shells D. Bucur, Ecole des Mines de Paris, France: Existence and stability of the optimum in shape optimization R. Dziri, Ecole des Mines de Paris, France: Shape sensitivity analysis: an application to a free boundary problem F. Desaint, CNRS - INLN, France: Shape derivative of all eigenvalues of a fluid-shell problem J.-C. Aguilar, CNRS - INLN, France: Shape control of an hydrodynamical wake Y. Guido, Ecole des Mines de Paris, France: Shape hessian for a nondifferentiable variational free boundary problem *********************************************************************** Harnessing of Chaos Organizer: M. Yamaguti, Ryukoku University, Japan Summary: This minisymposium provides an opportunity to discuss possibilities of applications of deterministic chaos. Since deterministic chaos is not only a profound concept in science but also ubiquitously observed in many real-world systems, it is closely related to industries, engineering, social sciences and so on. In fact, various interesting studies are now in progress for possible applications of deterministic chaos. Day: July 3, 1995 Time: 15.30 M. Yamaguti, Ryukoku University, Japan: Harnessing of Chaos M. Hasler, Swiss Federal Institute of Technology, Switzerland: Secure communication by synchronized chaos H. Kawakami, Tokushima University, Japan; T. Ueta, University of Tokushima, Japan: Synchronization and destabilitation of chaotic systems K. Aihara, The University of Tokyo, Japan: Computation by chaotic neural networks K. Suyama, Columbia University, USA; Y. Horio, Tokyo Denki University, Japan: Engineering applications of chaotic neural networks with integrated switched-capacitor chaotic neurons *********************************************************************** Functional Differential Equations and Their Applications Organizers: V.B. Kolmanovskii, Moscow Institute of Electronics and Math., Russia G.E. Kolosov, Moscow Institute of Electronics and Math., Russia Summary: During the last two decades, functional differential equations (FDEs) have spread at a rate never experienced before. The main reason for this is a huge variety of applications in different branches of industry, automatic regulation, mechanics, physics, medicine, neural networks, economics etc., where FDEs appear as a natural tool to model evolution phenomena because both measurements of time evolving variables and their dynamics usually contain some delays. Complicated behaviour that can arise in a hereditary system characterizes the modern area of research. Along with this is the problem of how such dynamical richness could be used in a technologically useful way. The scope of the minisymposium is to discuss some modern results of the theory of FDEs as well as the latest applications. Also, we propose to discuss the possibility of introducing FDEs in education processes. Day: July 4, 1995 Time: 09.30 J. Belair, Universit{\'e} de Montr{\'e}al, Canada: Stability analysis of a second-order differential equation with delayed negative feedback L. Berezansky, Ben Gurion University of the Negev, Israel: A new approach in the oscillation theory of linear functional differential equations K. Burrage, The University of Queensland, Australia: Parallel algorithms using waveform techniques for large IVPs K. Cooke, Pomona College, USA: Delay models of respiratory systems M. Falcone, Universit{\`a} di Roma La Sapienzia, Italy: Approximation of optimal control problems for delayed eqations E. Fridman, Tel Aviv University, Israel: Decoupling transformation of singularly perturbed systems with small delays and its applications Day: July 5, 1995 Time: 15.30 V.B. Kolmanovskii, Moscow Institute of Electronics and Math., Russia: Stability of nonlinear functional differential equations (FDEs) and applications G. Kolosov, Moscow Institute of Electronics and Math., Russia: Exact solution of an optimal control problem for dynamic systems with delay E. Litsyn, Bar-Ilan University, Israel; M. Drakhlin, College of Judea and Samaria, Israel: Stability of infinite systems of linear functional-differential equations X. Mao, University of Strathclyde, Scotland: Exponential stability in mean square of neutral stochastic differential functional equations J. Milton, The University of Chicago Hospitals, USA: Modeling the pupil light reflex with delay differential equations Day: July 6, 1995 Time: 09.30 I. Nechaeva, McGill University, Canada: Stability and oscillating properties of differential delay equations driven by white and coloured noise P. Nistri, Univ. Firenze, Italy; J.W. Macki, University of Alberta, Canada; P. Zecca, Universit{\`a} di Firenze, Italy: Periodic solutions to differential equations with delay H.J. Pesch, Technische Universit{\"a}t Clausthal, Germany: Numerical solution of delay differential equations: survey and applications M. Sambandham, Morehouse College, USA: Neural methods for boundary value problems P. Zecca, University of Firenze, Italy: Functional differential inclusions in Banach spaces Day: July 6, 1995 Time: 15.30 V.V. Obukhovskii, Voronezh State Pedagogical University, Russia: On some properties of semilinear functional differential inclusions in Banach spaces M.I. Kamenski, Voronezh State University, Russia: On the averaging principle for degenerate ordinary differential equations A. Pelster, Universit{\"a}t Stuttgart, Germany; W. Wischert, A. Wunderlin, Universit{\"a}t Stuttgart, Germany; M. Olivier, J. Groslambert, Centre National de La Recherche, Besancon, France: Application of synergetic concepts to nonlinear dynamical systems with time delay M. Schanz, Universit{\"a}t Stuttgart, Germany; H.-J. Freund, H. Hefter, P. Tass, Heinrich-Heine Universit{\"a}t D{\"u}sseldorf, Germany; A. Wunderlin, Universit{\"a}t Stuttgart, Germany: Sinusoidal forearm tracking with delayed visual feedback -- theory and experiment M. Schanz, Universit{\"a}t Stuttgart, Germany; A. Pelster, Universit{\"a}t Stuttgart, Germany: The first oscillatory instability in the Verhulst model with time delay *********************************************************************** Materials Theory Organizers: F. Ziegler, Technische Universit{\"a}t Wien, Austria O.T. Bruhns, Ruhr-Universit{\"a}t Bochum, Germany Summary: The main goal in the interdisciplinary field addressed by the minisymposium is to reach a fruitful combination of the latest physical-chemical oriented scientific findings of material-science, the mathematical methods of pattern recognition and functional analysis and of the thermodynamically based constitutive equations of continuum solid-mechanics. In addition, aspects of the computational mechanics of the constitutive equations of high-tech materials has become a matter of increasing attention. Hence, phenomena of composite materials are also addressed.\\ Currently, research in materials sciences is being restructured to account for new experimental and computational developments. Hence, phenomenologically derived constitutive relations are to be based on micro- or meso-mechanical models and measurements. New metallurgical findings like the intermetallics are a challenge and new material models are required to predict the mechanical behaviour in their engineering applications.\\ ICIAM 95 provides a forum to share the challange of new mathematical approaches in materials science. All those interested in the applications of mathematics to solid mechanics and materials science are invited to discuss open problems. Day: July 7, 1995 Time: 09.30 J.P. Boehler, Universite Joseph Fourier Grenoble 1, France; Th. Dietl, University Joseph Fourier, Grenoble, France; T. Keller, University Joseph Fourier, Grenoble, France: On homogenized transient cross-linking and temperature fields in polymer concrete structures O.T. Bruhns, Ruhr-Universit{\"a}t Bochum, Germany: Influence of phase transformation processes on the stress and inelastic deformation behaviour P.A. Fotiu, Technische Universit{\"a}t Wien, Austria: Wave propagation in elastic-plastic polycrystals P. Haupt, Gesamthochschule Kassel, Germany: On the thermodynamic representation of rate-independent material behaviour J.R. Willis, University of Cambridge, United Kingdom: Static and dynamic nonlinear response of composites Day: July 7, 1995 Time: 15.30 S. Schl{\"o}gl, Montanuniversit{\"a}t Leoben, Austria: Micromechanical modelling of Ti-Al intermetallics - a crystal plasticity concept and experiments M. Berveiller, Universit{\'e} de Metz, France: Thermomechanical behaviour of solids during martensitic phase transformation D. Besdo, Universit{\"a}t Hannover, Germany: to be announced E. Steck, Technische Universit{\"a}t Braunschweig, Germany: Micromechanical development and identification of a stochastic constitutive model R. Gerdes, Technische Universit{\"a}t Braunschweig, Germany; F. Thielecke, TU Braunschweig, Germany: Micromechanical development and identification of a stochastic constitutive model *********************************************************************** Recent Advances in Computational Mechanics on Parallel Machines Organizer: G. Yagawa, University of Tokyo, Japan Summary: Computational mechanics is one of the most important and popular fields of applied mathematics and mechanics and includes the finite element method, the finite difference method and the boundary element method. However, conventional computer mechanics only has the capability of solving much smaller scale problems than what industry requires in reality. To overcome this difficulty, the most promising technology seems to be the use of various types of parallel environments, which include the massively parallel computers and the work stations in cluster. The proposed minisymposium is to be a forum to discuss this hot issue among international speakers active in this field, particularly in the field of fluid and solid mechanics. Day: July 4, 1995 Time: 15.30 I.S. Doltsinis, University of Stuttgart, Germany: Coupling of distinct operators in parallel processing T.E. Tezduyar, University of Minnesota, USA: Parallel finite element computation of $3D$ compressible and incompressible flows N. Satofuka, Kyoto Institute of Technology, Japan: Parallel computing of Navier-Stokes and Boltzmann equations using methods of lines G. Yagawa, University of Tokyo, Japan: A high performance finite element method with massively parallel processors D.R.J. Owen, University of Wales, United Kingdom: Some recent advances in the parallel finite element simulation of metal forming problems *********************************************************************** Analysis of Nonlinear Ill-Posed Problems with Applications to Parameter Estimation Organizer: O. Scherzer, Johannes-Kepler-Universit{\"a}t Linz, Austria Summary: An important class of nonlinear operator equations $$F(x)=y$$ are parameter estimation problems. In these problems the model is known in principle but the unknown parameters have to be determined from observations of the system. \smallskip In this minisymposium different methods are applied for the stable solution of parameter estimation problems, and are compared both from a theoretical point of view and from practical aspects, so as to make evident the advantages and shortcomings of different techniques. \smallskip Efficient numerical solvers for nonlinear problems require the derivative of $F$. By explicitly calculating the Fr\'echet--derivative for concrete examples, possibilities for calculating derivatives of related problems are demonstrated. Day: July 5, 1995 Time: 15.30 J. Blum, Universit{\'e} de Grenoble, France; H. Buvat-Dousteyssier, Grenoble, France: Identification of the current density profile in a tokamak B. Hofmann, Technische Universit{\"a}t Chemnitz-Zwickau, Germany: Decomposition cases of nonlinear ill-posed problems O. Scherzer, Johannes-Kepler-Universit{\"a}t Linz, Austria: Convergence analysis of some iterative methods for the solution of nonlinear ill-posed problems K. Kunisch, Technische Universit{\"a}t Berlin, Germany: Numerical methods for parameter estimation problems F. Hettlich, Universit{\"a}t Erlangen-N{\"u}rnberg, Germany: Regularization of inverse scattering problems by iterative methods S.I. Kabanikhin, Institute of Mathematics, Russia: Convergence analysis of Newton-Kantorovich, Picard and Caratheodory algorithms in hyperbolic inverse problems *********************************************************************** Mathematical Models in Epidemiology Organizers: V. Capasso, Universit{\'a} di Milano, Italy P. van den Driessche, University of Victoria, Canada Summary: Recently there has been rapid growth in the use of mathematics to model infectious disease transmission. This minisymposium focuses on deterministic models giving rise to nonlinear dynamical systems, and statistical models. Challenging mathematical analysis, often supplemented with numerics, is needed to determine the disease dynamics. Speakers address important, current questions relating either to a class of diseases (e.g., fatal diseases, sexually-transmitted diseases) or to a specific disease (e.g., influenza, pertusis, schistosomiasis). This minisymposium brings together experts active in mathematical epidemiology, and appeals to a wider audience of applied mathematicians wanting to learn about this exciting interdisciplinary area. Day: July 5, 1995 Time: 15.30 V. Andreasen, Roskilde University, Denmark; S.A. Levin, Princeton University, USA; J. Lin, Washington College, Chestertown, USA: Models of epidemiology and natural selection of influenza A A.D. Barbour, Universit{\"a}t Z{\"u}rich, Switzerland: Mean worm burden and schistosomiasis modelling C. Castillo-Chavez, Cornell University, USA: The effect of behavioral changes and treatment on the dynamics of sexually-transmitted diseases including HIV/AIDS H.W. Hethcote, University of Iowa, USA: Mathematical modeling of pertussis epidemiology K.-P. Hadeler, Universit{\"a}t T{\"u}bingen, Germany: Optimal vaccination strategies in the treatment of infectious diseases Day: July 6, 1995 Time: 15.30 A. Pugliese, Universit{\`a} di Trento, Italy; M.L. Damaggio, Centro di Ecologia Alpina, Italy: Host-macroparasite models with variable aggregation M. Kretzschmar, Rijksinstituut voor Volksgezondheid en Milieuhygiene, The Netherlands; M. Morris, Columbia University, New York, USA: Modelling the influence of concurrent partnerships on the spread of sexually transmitted diseases C. Lef{\`e}vre, Universit{\'e} Libre de Bruxelles, Belgium: Final size of collective (SIR) epidemic models H. Thieme, Arizona State University, USA: Aspects of dispersal in metapopulation models *********************************************************************** Control Problems for Distributed Parameter Systems with Applications Organizer: I. Lasiecka, University of Virginia, Charlottesville, USA Summary: This minisymposium will focus on control and optimization problems for dynamic processes formulated in the context of distributed parameter systems. Particular emphasis will be placed on nonlinear models. The speakers will present an overview of current research and methodologies in various areas of modern control theory including topics such as: \item{(i)} Controllability/stabilizability of systems arising in continuum mechanics. \item{(ii)} Optimality conditions/maximum principle for nonlinear models. \item{(iii)} Shape optimization/shape design and free boundary optimization problems in elastic structures. \smallskip Applications of the theoretical results to areas such as space structures, robotics, petroleum industry, hydromagnetism, chemical processes will be also discussed. Day: July 4, 1995 Time: 09.30 F. Abergel, Universite de Paris Sud, France: Critical point conditions for optimal free surfaces E. Casas, Universidad de Cantabria, Spain: Pontryagin's principle for state-constrained control problems governed by parabolic equations E. Feireisl, Academy of Sciences of the Czech Republic, Czech Republic: Global in time behavior of hyperbolic equations with distributed damping A. Khapalov, Oregon State University, USA: Some aspects of the approximate controllability properties of the semilinear distributed parameter systems J. Sokotowski, PAN, Poland: Optimal design methods in PDE's Day: July 4, 1995 Time: 15.30 T. Swobodny, Wright State University, USA: Analysis of a nonlinear control problem arising from a model of hydromagnetodynamics V. Valente, I.A.C. del CNR, Italy: Exact controllability of thin elastic shells J.-P. Zol{\'e}sio, CNRS - INLN, France: Eulerian extremality of the action in dynamical free boundary problems I.M. Lasiecka, University of Virginia, USA; R. Triggiani, University of Virginia, USA: Uniform stabilization of spherical shells by using moments and shears applied on the boundary *********************************************************************** Mathematical Theory of Semiconductors Organizers: P.A. Marcati, University of L'Aquila, Italy P. Markowich, Technische Universit{\"a}t Berlin, Germany Summary: The trend toward an increasing miniaturization of modern semiconductor devices forces one to analyze certain physical situations where the Drift-Diffusion equations, today a standard tool in device simulation, are no longer valid. A more realistic approach can be obtained going back to kynetic models or deriving a more accurate set of macroscopic equations: the Hydrodynamic approximations. This situation opens a wide new variety of challenging mathematical problems which are expected to provide a new impulse both to the theory and to the applications. The analysis of shock waves, of ill-posed boundary value problems, the study of very singular perturbation problems, of relaxation processes, of semiclassical limits and more generally the analysis of quantum effects are only a few of the possible objects of investigation and of the important questions posed by the applications. Day: July 7, 1995 Time: 15.30 I.M. Gamba, New York University, USA: Steady fluid-poisson systems: theoretical results in 2-dimensional geometries C. Schmeiser, Technische Universit{\"a}t Wien, Austria; A. Zwirchmayr, Technische Universit{\"a}t Wien, Austria: Expansion methods for the Boltzmann equation and domain decomposition for semiconductor device simulation R. Natalini, CNR, Italy; P. Marcati, Universit{\`a} degli Studi del l'Aquila, Italy: Hydrodynamic models for semiconductors F. Poupaud, University of Nice, France: Wigner series and semiclassical limit in a crystal A.M. Anile, Universit{\'a} di Catania, Italy: An improved hyrodynamical model for carrier transport in semiconductors A. Unterreiter, Technische Universit{\"a}t Berlin, Germany: Stationary quantum hydrodynamic models *********************************************************************** Mathematical Modelling of Piezoelectric Drive Systems Organizers: P. Hagedorn, Technische Hochschule Darmstadt, Germany J. Wallaschek, Universit{\"a}t-GH Paderborn, Germany Summary: Piezoelectric drive systems, like e.g. the travelling wave ultrasonic motor are characterized by high torque at low rotational speed, simple mechanical design and good controllability. Compared to electromagnetic actuators, the torque per volume ratio can be higher by an order of magnitude. This is the main reason why piezoelectric drive systems have replaced electromagnetic motors in autofocus systems and several other industrial applications. Since the energy conversion is based on the piezoelectric effect instead of electromagnetic interaction, many challenging modelling issues have arised in the design of these novel mechatronic drive systems. In this minisymposium, the most important topics in this area will be addressed. Day: July 5, 1995 Time: 15.30 H.P. Sch{\"o}ner, Daimler Benz Forschung, Germany: Piezoelectric drive systems: their history, applications and mathematical challenges O.Y. Zharii, Kiev University, Ukraine: Nonclassical mixed boundary-value problems of elastodynamics arising in mathematical modelling of ultrasonic motors L.P. Levin, Worcester Polytechnic Institute, USA; R. Ludwig, Worcester Polytechnic Institute, MA, USA; P. Hagedorn, Technische Hochschule Darmstadt, Germany: A $2D$ finite element simulation of Piezoelectrically driven bending in the stator of a traveling wave ultrasonic motor H. Irschik, Johannes-Kepler-Universit{\"a}t Linz, Austria: The role of the transition matrix for flexural vibration of beams with distributed piezoelectric actuators J.W. Krome, Universit{\"a}t-GH Paderborn, Germany; J. Wallaschek, Universit{\"a}t-GH Paderborn, Germany; J. Maas, Universit{\"a}t-GH Paderborn, Germany: Models for the electro-mechanical interaction of the stator of a Piezoelectric ultrasonic motor and its power supply *********************************************************************** Modelling, Analysis and Control of Transmission Problems Organizer: G. Leugering, Universit{\"a}t Bayreuth, Germany Summary: Transmission problems appear in various different areas of application, like continuum and quantum mechanics, geophysics, accustics etc.. Typically, a connected graph or a set of surfaces (manifolds) with interfaces are given, together with some differential operators defined on these structures. One is interested in the transient behaviour of the state variables governed by the corresponding PDEs. A variety of mathematical problems are at issue: concise mathematical modelling, existence, uniqueness and regularity of solutions to the corresponding (nonlinear) transmission problems, spectral analysis, asymptotic behaviour, controllability and stabilizability, as well as numerical analysis and simulation. \smallskip The speakers cover a wide range of applications from structural mechanics to quantum mechanics and, thereby, focus precisely on the main issues mentioned above. Day: July 6, 1995 Time: 09.30 J.E. Lagnese, Georgetown University, USA: Controllability of systems of interconnected elastic plates S. Nicaise, Uni. de Valenciennes et du H. Camb{\'e}sis, France: Singularities for interface problems J. von Below, Universit{\"a}t T{\"u}bingen, Germany: Dynamical interface and node transition in ramified media with diffusion F. Ali Mehmeti, Technische Hochschule Darmstadt, Germany: The tunnel effect G. Leugering, Universit{\"a}t Bayreuth, Germany: Reverberation and control for 1-d elastic networks *********************************************************************** Numerical Algorithms for Air Pollution Models Organizers: J. Verwer, CWI, Amsterdam, The Netherlands Z. Zlatev, NERI, Roskilde, Denmark Summary: Air pollution modelling has shown substantial growth in the last decade. Subjects vary from smog-prediction to climate research on the expected long term change of atmospheric chemical composition due to ever increasing emission of pollutants. Mathematical models for predicting atmospheric transport, chemistry and exchange of trace constituents are essential for this research. Generic models consist of huge systems of partial differential equations of the advection-diffusion-reaction type, supplemented with a variety of subgrid scale, physical parameterizations. Besides data access and physical knowledge, efficient algorithms and computer capacity are and will remain a critical factor for resolution. This minisymposium will address new developments in the analysis and implementation of numerical algorithms particularly designed for large-scale air pollution models. Issues include numerical advection, stiff chemistry solvers, operator splitting, method of lines, high performance computing and parallelism. Day: July 6, 1995 Time: 15.30 Z. Zlatev, National Environmental Research Institute, Denmark: Three-dimensional version of the Danish Eulerian model M.Z. Jacobson, Stanford University, USA: Application of a sparse-matrix, vectorized Gear-type code (SMVGEAR) in an air pollution modeling system O. Knoth, Institut f{\"u}r Troposph{\"a}renforschung, Germany; R.Wolke,Institue f{\"u}r Tropossph{\"a}renforschung,Leipzig, Germany: Time integration of advection-diffusion-reaction equations by explicit-implicit methods W. Hundsdorfer, Centrum voor Wiskunde en Informatica, The Netherlands: Numerical modelling of global transport and chemistry C.J. Aro, Lawrence Livermore National Laboratories, USA; G.H. Rodrigue, Lawrece Livermoore National Laboratory, Livermore, USA: Preconditioned time differencing in an atmospheric 3-D chemical-radiative-transport model J.G. Verwer, Centrum voor Wiskunde en Informatica, The Netherlands; J.G. Blom, CWI, Amsterdam, The Netherlands: Algorithm and software aspects of an implicit-explicit method for $3D$ atmospheric transport-chemistry problems *********************************************************************** Applied and Theoretical Problems in Graph Drawings Organizer: R.D. Ringeisen, Old Dominion University, Norfolk, Virginia, USA Summary: Graph or network drawing problems are of increasing interest to mathematicians, who find their fascination intrinsic, and to applied mathematicians and computer scientists who see applications in such fields as layout or algorithm design. It is noteworthy that theoreticians may be unaware of the needs of those working on applications and vice versa. Techniques vary from computational to purely theoretical, with neither alone being sufficient. This minisymposium chooses problems encompassing both theoretical and applied fields. The speakers represent this diversity and have as one goal to demonstrate the techniques, both developed and needed, to attack specific kinds of problems. Day: July 3, 1995 Time: 10.00 R.D. Ringeisen, Old Dominon University, USA: Adding and deleting crossings in graphs P. Eades, University of Newcastle, Australia: Edge crossings in layered drawings of graphs B. Richter, Carleton University, Canada: Crossing numbers of products of cycles H. Harborth, Technische Universit{\"a}t Braunschweig, Germany: Crossing regular drawings of graphs I.G. Tollis, The University of Texas at Dallas, USA: Algorithms and bounds for orthogonal drawings of graphs *********************************************************************** Numerical Solution of Differential-Algebraic Equations Organizers: L. Petzold, University of Minnesota, USA S. Campbell, North Carolina State University, USA Summary: Many physical processes are naturally described as systems of differential-algebraic equations (DAEs). These types of equations occur in the modelling of electrical networks, flow of incompressible fluids, control, chemical process simulation, multibody systems simulation, molecular dynamics and many other applications. \smallskip In recent years much work has been devoted to the study of the mathematical structure of DAEs and to the analysis and development of numerical methods for DAEs. Software has been developed which can solve many of these problems routinely. In this minisymposium, we will focus on numerical methods for the solution of some particularly challenging classes of DAE systems: stiff DAEs, constrained Hamiltonian systems, and oscillating DAEs. Day: July 3, 1995 Time: 15.30 C. Lubich, Universit{\"a}t T{\"u}bingen, Germany: Stiff differential-algebraic equations S. Reich, Konrad-Zuse-Zentrum f. Informationstechnik, Germany: Numerical integration of constrained Hamiltonian systems with applications to molecular dynamics simulations E. Hairer, University of Geneva, Switzerland: On the integration of Hamiltonian systems with holonomic constraints R. M{\"a}rz, Humboldt-Universit{\"a}t Berlin, Germany; R. Winkler, Humboldt-Universit{\"a}t Berlin, Germany; R. Lamour, Humboldt-Universit{\"a}t Berlin, Germany: Numerical treatment of nonlinear oscillation phenomena in differential algebraic equations *********************************************************************** Applications of Geometric Methods in Nonlinear Control Theory Organizer: W.F. Shadwick, University of Toronto, Waterloo, Ontario, Canada Summary: The application of techniques from the geometry of exterior differential systems has recently led to new insight into non-linear control systems. This symposium will focus on applications to problems from feedback and dynamic feedback linearization, aircraft flight control and robotics. Day: July 4, 1995 Time: 15.30 P. Crouch, Arizona State University, USA: Constrained variational problems and optimal control M. Fliess, CNRS - Supelec, France: Differentially flat control systems R.B. Gardner, University of North Carolina, USA: A pseudogroup isomorphism between control theory and the calculus of variations R. Murray, California Institute of Technology, USA: Control of robots through Goursat normal forms W. Sluis, University of Cambridge, United Kingdom: Dynamic feedback linearization of control systems *********************************************************************** Mathematical and Computational Models for $2D$ Turbulence Organizer: R. Piva, Universit{\`a} di Roma "La Sapienza", Italy Summary: The study of $2D$ decaying turbulence has been numerically investigated by using spectral methods and, more recently, with the Contour Surgery algorithm. These approaches enable one to analyse the decay in moderate Reynolds number flows and in the limit for viscosity going to zero, respectively. In the two cases the number of vorticity structures during the decay behaves in a quite different way: it is rapidly decreasing in time in the spectral simulations, where any interaction between two vortices leads to a single one, through a fast dissipation of small scales. On the contrary, by the Contour Surgery, which is practically an inviscid algorithm, the small scale vortices are retained and we observe a tendential growth, at least initially, of the vortex number. \smallskip The decay in moderate Reynolds numbers flows has also been investigated by using simplified models, as the Vortex Model. To model the viscous decay, the Hamiltonian dynamics of the vortices, that are roughly approximated by using circular patches, is coupled with a dissipative rule that mimics the merging. Namely, two vortices, when below a critical distance, lead to a single one with an instantaneous interaction in which the self-induced part of the kinetic energy is conserved but the enstrophy is dissipated, according with the diffusion of small scale structures in viscous flow. By means of this merging rule coupled with a Vortex Model, the time scaling of the vortex number (and, as a consequence of the enstrophy) is recovered with an exponent very close to the one obtained by spectral simulations. \smallskip The decay rates in viscous and inviscid flows are very different and the understanding of the physical behaviour for growing Reynolds numbers is still an open question. Day: July 4, 1995 Time: 09.30 R. Piva, Univ. "Lasapienza", Italy; G. Riccardi, Universit{\`a} di Roma "La Sapienza", Italy: to be announced D. Dritschel, DAMPT, United Kingdom: to be announced G. Carnevale, SCRIPPS, USA: to be announced X. Jimenez, Universidad Politechnica, Spain: to be announced R. Benzi, Univ. Roma "Tor Vergata", Italy: to be announced M. Farge, Ecole Normale Superieure, France: Coherent structures education in two-dimensional turbulent flows using wavelet packets *********************************************************************** Identification Problems Organizer: G. Talenti, University of Florence, Italy Summary: Typical identification problems arise in partial differential equations, when one has to detect a coefficient or a domain from overspecification of boundary data. These problems are of interest in engineering, medicine and geosciences. Day: July 5, 1995 Time: 15.30 M.S. Vogelius, Rutgers University, USA: to be announced G.A. Uhlman, University of Washington at Seattle, USA: to be announced F. Natterer, Universit{\"a}t M{\"u}nster, Germany: Numerical solution of the inverse problem of acoustics G. Alessandrini, Universita di Trieste, Italy: to be announced G. Nakamura, Science University of Tokyo, Japan: to be announced *********************************************************************** Hyperbolic Conservation Laws with Source Organizer: R. Natalini, IAC "M. Picone", Roma, Italy Summary: Hyperbolic conservation laws with source (or balance laws) arise in many fields of mathematical physics: reacting gas flows (including combustion), nozzles, compressible Euler equations with spherical symmetry or gravitation, viscoelasticity, electrons flows in semiconductors. \smallskip Although the general theory of balance laws is not yet fully developed, it is possible to study the combined effects of the convection and the source terms for particular models. \smallskip In the scalar case and for weakly coupled systems some singular phenomena occur: blow--up, istantaneous expansion of the support, asymptotic instability. \smallskip For gas--dynamics and elasticity models some results have been established about the existence of weak solutions and their relaxation limits. Day: July 5, 1995 Time: 09.30 G.-Q. Chen, Northwestern University, USA: Hyperbolic systems of conservation laws with stiff relaxation terms C. Sinestrari, Universit{\'a} di Roma "Tor Vergata", Italy; R. Natalini, IAC-CNR, Roma, Italy; A. Tesei, Universit{\`a} di Roma "Tor Vergata", Italy: Asymptotic behaviour and blow-up of solutions of conservation laws with source B. Hanouzet, University of Bordeaux I, France; R. Natalini, Instituto per le Applicazioni del Calcolo "M. Picone", Italy; A. Tesei, Universit{\`a} degli Studi di Roma, Italy: Asymptotic behaviour of solutions for weakly coupled quasilinear systems H. Holden, Universitetet i Trondheim, Norway; N.H. Risebro, University of Oslo, Norway: Conservation laws with a random source *********************************************************************** Dynamics of Partial Differential Equations Organizer: G.R. Sell, University of Minnesota, USA Summary: The theory of dynamical systems has become a major tool for the study of the long-time behavior of solutions of partial differential equations, and especially for such equations in higher space dimension. For example, a number of papers have recently appeared addressing the issues of finite dimensional strucutres, including global attractors and inertial forms, for the Navier-Stokes equations and for partly dissipative systems in 3D. New related results concerning the dimensions of the attractors, the dynamics on the attractor, and the dynamics in physical domains with small aspect ratio are also of interest. While this minisymposium will focus on the more theoretical issues of the topic, the developments and methodology described above are important for such applications as: the development of new algorithms for computational mechanics and fluid dynamics, turbulence in fluid mechanics, and for oceanic models used in global climate modeling. Day: July 6, 1995 Time: 09.30 C. Foias, Indiana University, USA: Dimension of attractors for the Navier-Stokes equations G. Raugel, Universite de Paris-Sud et CNRS, France: Navier-Stokes and Euler equations on thin domains M. Kwak, Chonnam National University, Korea: Finite dimensional structure for Navier-Stokes equations I. Kukavica, University of Chicago, USA; C. Foias, Indiana University, Chicago, USA: New results on the determining nodes for partial differential equations Z. Shao, Millersville University, USA; G.R. Sell, University of Minnesota, Minneapolis, USA: Inertial manifolds for partly dissipative reaction diffusion systems in higher space dimensions G.R. Sell, University of Minnesota, USA: Global attractors for the Navier-Stokes equations in $3D$ *********************************************************************** Super-Resolution Problems: Theory and Applications Organizer: P. Barone, CNR, Roma, Italy Summary: Loosely speaking, a super-resolution problem (SRP) consists in recovering a function from partial knowledge of its Fourier transform. For example, for space-limited functions, whose Fourier transform is given over a finite frequency band, the problem consists in extrapolating the Fourier transform outside the finite band. This problem can be reduced to the solution of an integral equation of convolution type. This exemplifies the connection of SRP's with ill-posed problems. Many methods exist for solving SRP's in different contexts. However, they are not completely understood from the theoretical point of view. Specifically, quantitative statements about the true limit of resolution, as a function of the noise, developed up to now, seem to be quite conservative, with respect to the results obtained in practice. Day: July 6, 1995 Time: 15.30 R. March, IAC, CNR, Italy; P. Barone, CNR, Roma, Italy: On the super-resolution properties of Prony's method M. Bertero, Universit{\'a} di Genova, Italy: Super-resolution and the ill-posedness of out-of-band extrapolation C. De Mol, Universit{\'e} Libre de Bruxelles, Belgium: How to beat the Rayleigh limit? D.L. Donoho, University of California at Berkeley, USA: to be announced F. Prodi, Alenia, Italy; A. Farina, ALENIA, Roma, Italy: Application of superresolution to $2D$ ISAR images E.R. Pike, King's College London, United Kingdom; U. Brand, King's College London, UK; R. Coulam, King's College London, UK; J. Grochmalicki, King's College London, UK: Super-resolution in three-dimensional fluorescent laser scanning microscopy *********************************************************************** Value Function and Sufficient Conditions in Deterministic Nonlinear Control Organizers: P. Cannarsa, Universit{\`a} di Roma "Tor Vergata", Italy H. Frankowska, CEREMADE-Universit{\'e} Paris-Dauphine, France Summary: Nonlinear optimal control problems originated from a large number of applied research areas, such as aerospace and electrical engineering, theoretical economics, finance, population dynamics, theoretical immunology, etc. \smallskip The value function is a traditional tool to solve such control problems. In fact, both necessary and sufficient conditions for optimality can be derived from its properties. \smallskip Most classical results were obtained under the assumption that the value function is smooth. However, it is known that this assumption is very restrictive. Typically, it is equivalent to the uniqueness of optimal trajectories. For this reason, it does not allow treatment of genuinely nonlinear problems. \smallskip In recent years, various approaches based on nonsmooth analysis have been developed to treat general cases when the value function is not differentiable. On one hand, this effort made rigorous the classical theory. On the other hand, it discovered new phenomena, essentially due to nonsmoothness. \smallskip The purpose of this minisymposium is to present different approaches to such issues as sufficient conditions for optimality, regularity of the value function and construction of optimal trajectories. \smallskip Our goal is to address an audience of both specialists in control theory and applied mathematicians, interested in getting a clear picture of modern achievements in this area. \smallskip All presented talks concern results that were obtained only recently in full generality. Therefore, this seems to be the best time for such a symposium to take place. Day: July 7, 1995 Time: 15.30 A. Bressan, Universita di Trieste, Italy; B. Piccolo, S.I.S.S.A., Trieste, Italy: A generic classification of optimal stabilizing feedbacks in the plane C.I. Byrnes, Washington University, USA: Nonlinear regulation P. Cannarsa, Universit{\'a} di Roma "Tor Vergata", Italy; C. Sinestrari, Universit{\`a} di Roma "Tor Vergata", Italy: Optimality conditions for the minimum time problem with a general target H. Frankowska, Universit{\'e} Paris-Dauphine, France; N. Caroff, University of Bordeaux II, France: Conjugate points, optimality conditions and regularity of the value function in nonlinear optimal control V. Zeidan, Michigan State University, USA: Second order optimality conditions for optimal control problems *********************************************************************** Advances in Dynamic Programming Organizer: I. Capuzzo-Dolcetta, Universit{\`a} di Roma "La Sapienza", Italy Summary: The area addressed by the minisymposium is that of optimal control of deterministic and stochastic dynamical systems with a finite number of degrees of freedom. Relevant applications are traditionally found in engineering, economics and, more recently, in finance.\\ The common approach adopted by the speakers is that of dynamic programming in its more updated and mathematically rigorous versions. This approach leads to the analytical and numerical study of non smooth (sometimes, even non continuous) solutions of the Hamilton - Jacobi - Bellman partial differential equation satisfied by the value function. An important role in this context has been played in the last ten years by the theory of viscosity solutions. \smallskip The presentations scheduled in the minisymposium are aimed at giving an account of some recent advances in the direction of implementing the dynamic programming approach to the design of optimal feedbacks for a variety of optimal control problems motivated by applications. Day: July 7, 1995 Time: 09.30 M. Bardi, Universit{\`a} di Padova, Italy: Dynamic programming for optimal control problems and differential games with discontinuous value function M. Falcone, Universit{\`a} di Roma La Sapienzia, Italy; P. Lanucara, Universit{\`a} di Roma "La Sapienza", Italy: Parallel algorithms for Hamilton-Jacobi equations H. Ishii, Chuo University, Japan: Approximation of optimal controls via viscosity solutions J.L. Menaldi, Wayne State University, USA; M. Robin, Centre National d'Etudes Spatiales, Paris, France: Ergodic control of reflected diffusions with jumps I. Capuzzo-Dolcetta, Universit{\`a} di Roma La Sapienzia, Italy: Approximate viscosity solutions of Hamilton-Jacobi-Bellman equations *********************************************************************** Numerical Lower Index Differential Algebraic Systems Organizers: R. M{\"a}rz, Humboldt-Universit{\"a}t Berlin, Germany K. Strehmel, MLU Halle-Wittenberg, Germany Summary: Differential algebraic systems arising in various fields of industrial applications (electric circuit simulation, controlling chemical reactions etc.) are typically those of lower index, and their data are not very smooth. This minisymposium is devoted to numerical methods which can be directly applied to those kind of differential algebraic systems. \smallskip Although numerical integration methods for differential algebraic systems having index 1 and 2 have been topical for years, only little attention has been paid to the asymptotical behaviour of both the continuous solution and their numerical approximation.\\ However, an asymptotically correct numerical modelling is known to be the crucial point, even in solving regular differential equations numerically. \smallskip In particular, the asymptotical stability behaviour of integration methods is studied in the first two lectures. \smallskip The following two papers are devoted to transfer methods which allow to describe the correct solution subspaces very well.\\ The next lecture shows how to take advantage of special information on subspaces given, in particular, in the case of regularized problems.\\ Finally, on the last contribution, the dichotomic behaviour of differential algebraic systems is studied with the objective to improve boundary value problem solvers. Day: July 3, 1995 Time: 10.00 J. Wensch, Martin-Luther Universit{\"a}t Halle-Wittenberg, Germany; K. Strehmel, R. Weiner, MLU Halle-Wittenberg, Germany: Stability investigations for index 2 systems M. Hanke, Humboldt-Universit{\"a}t Berlin, Germany; R. M{\"a}rz, Humboldt-Universit{\"a}t Berlin, Germany: On asymptotics in case of DAEs K. Balla, Hungarian Academy of Sciences, Hungary: Transfer of boundary conditions for DAEs T. Petry, Humboldt Universit{\"a}t Berlin, Germany: A stable realization of lower index DAE transfer methods A.R. Rodriguez Santiesteban, Humboldt Universit{\"a}t Berlin, Germany; M. Hanke, Humboldt-Universit{\"a}t Berlin, Germany: Solution subspaces and operator splitting methods R. England, Open University Milton Keynes, United Kingdom; R. Lamour, Humboldt Universit{\"a}t Berlin, Germany: Numerical boundary value problems in lower index DAEs *********************************************************************** Knowledge Through Images for the Cultural Heritage Protection Organizer: L. Moltedo, CNR, Roma, Italy Summary: The Cultural Heritage protection represents one of main cultural and strategic interests for many European Countries. The images constitute a very important knowledge set on the art-works, to be used for multiple useful investigations. For instance, we can deal with images both for studying and representing deterioration processes of materials and paintings and for designing and managing the automation of the restoration process. Coding methods of multidimensional images allow interaction, archiving and transmission. The fidelity of image representation can also assure an efficient local and remote fruition, by using new telematics networks. Generally speaking, the knowledge through images for the Cultural Heritage protection requires the development of methodologies and techniques for the analysis and synthesis of images. Mathematical and computational aspects related to the above methodologies are presented. Day: July 3, 1995 Time: 15.30 V. Cappellini, Universit{\`a} degli Studi di Firenze, Italy: Art-work image processing and transmission L. De Floriani, Universit{\`a} di Genova, Italy; D. Marini, Universit{\`a} di Milano, Italy; R. Scopigno, Istituto CNUCE-CNR, Pisa, Italy: Reconstruction, modelling and visualization of art objects for restoration and presentation L. Moltedo, IAC-CNR, Italy; O. Salvetti, IEI-CNR, Pisa, Italy: A generalized model for texture coding and processing in applications of art conservation A. Guidazzoli, CINECA, Casalecchio di Reno-Bologna, Italy; M. Forte, CINECA, Casalecchio di Reno-Bologna, Italy: $3D$ virtual landscape navigation: a tool for monitoring and safeguarding archaeological areas B. Eberhardt, Universit{\"a}t T{\"u}bingen, Germany: Castel del Monte -- from the idea to the movie T. Grunert, Universit{\"a}t T{\"u}bingen, Germany: Putting the Stoeffler celestial globe into the computer A. Gagalowicz, INRIA Rocquencourt, France: Use of analysis/synthesis techniques for image restoration *********************************************************************** Design of Real and Simulated Experiments in Industrial Product Optimization Organizer: G. Morra, Fiat Research Center, Torino, Italy Summary: In the last years the use of design of experiments (D.O.E.) methodologies has become more and more extensive in industries. Experience has confirmed that well planned experiments can give good results in short times and at low costs. \smallskip On the other hand, numerical simulation is now a fundamental tool for designers in industry: the idea to substitute real experiments with virtual experiments is very appealing due to the enourmus time-saving that computer experiments could provide. \smallskip The minisymposium will focus on the use of D.O.E. for real and simulated experiments. The connection with optimization of performances of products will be stressed: industrial cases from automotive and aeronautic industies as well as methodological aspects from universities will be presented. \smallskip The main goal of the minisymposium is to allow a comparison of different experiences: the audience could be prevalently constituted by researchers in the field of industrial statistics and optimization. Day: July 4, 1995 Time: 09.30 T. Felici, L.E.M.T.A. - C.N.R.S., France; O. Sero-Guilaume, LEMTA URA CNRS 875, Vandoeuvre, France: Global optimization algorithm using an evolutionary strategy: application to optimal aerospace design G. Pistone, Politecnico di Torino, Italy: Algebraic algorithms in simulated experiments G. Iuculano, University of Florence, Italy: Experimental design in industries A.E. Bourguignon, SNECMA, France: Design of experiments for numerical simulation H.P. Wynn, City University, United Kingdom: Search and experiment for large systems R. Fontana, Fiat Research Center, Italy: Design of experiments at Fiat Research Centre *********************************************************************** Stochastic Methods in Nonlinear PDE for Physics and Applications Organizer: M. Pulvirenti, Universit{\`a} di Roma "La Sapienza", Italy Summary: Stochastic processes play an increasingly important role in the analysis of nonlinear phenomena which are usually described by partial differential equations. The interplay between stochastic and analytical methods in approaching nonlinear problems is certainly a current and interesting topic in the field of applied mathematics and attracts the efforts of many researchers. The organization of a minisymposium in this direction has the purpose of reviewing some recent results in the field of Mathematical Physics and applications. Day: July 6, 1995 Time: 09.30 A. De Masi, University of L'Aquila, Italy: to be announced R. Esposito, Univ. Roma "Tor Vergata", Italy: to be announced C. Grahm, Ecole Polytechnique, France: to be announced H. Spohn, Universit{\"a}t M{\"u}nchen, Germany: to be announced T.H. Yau, Courant Institute, USA: to be announced L. Tubaro, Universit{\'a} di Trento, Italy: to be announced *********************************************************************** Scalable High Performance Libraries for Large Scale Scientific Computing Organizers: A. Murli, Universit{\`a} di Napoli "Federico II", Italy P. Messina, California Institute of Technology, Pasadena, USA Summary: In the last few years, many advances have been made in the development of parallel algorithms for scalable architectures. However, one of the main difficulties for the scientific community in effectively using these architectures is the lack of software libraries (both mathematical software and graphics routines). Presently only few routines, mainly of linear algebra, are available (see the ScaLAPACK project). \smallskip The aim of this minisymposium is to give a perspective of the development of Scalable High Performance Libraries for large scale scientific computing. The speakers will describe recent results in the field of mathematical software for advanced architectures in some of the main areas of Computational Mathematics (linear algebra, multidimensional quadrature, ..). The minisymposium is addressed to all scientists interested in solving scientific problems on advanced architectures. Day: July 5, 1995 Time: 15.30 J. Demmel, University of California at Berkeley, USA: The scalapack linear algebra library for distributed memory machines J. Dongarra, The University of Tennessee, USA: Software tools and templates for developing a software library S. Hammarling, The Numerical Algorithm Group Ltd, United Kingdom: Development of a parallel library for distributed memory machines P.C. Messina, California Institute of Technology, USA: System functions in support of mathematical software libraries on parallel computing environment J.N. Lyness, Argonne National Laboratory, USA: Guidelines for parallel software in numerical quadrature *********************************************************************** Optimal Stopping and Burn-in Problems Organizer: F. Spizzichino, Universit{\`a} degli Studi di Roma "La Sapienza", Italy Summary: The subject of the Minisymposium will be the following problem of applied probability: \smallskip Consider N similar units which are prone to {\it early failures} and which are simultaneously tested in a duration experiment; provided complete statistical data from the experiment are avalaible, when are these units to be considered out of the {\it infant mortality} period? This problem is referred to as the burn-in problem. Under given cost structure this can be shown to be equivalent to an optimal stopping problem for a suitably defined continuous time stochastic process. \smallskip However, a purely analytical study seldom leads to explicit solutions of problems of optimal stopping. \smallskip The scope of the Minisymposium is to favour the cross-fertilization on the burn-in problem between probabilists and reliability engineers, in order to make it possible to combine the general theory of optimal stopping with statistical features (e.g., symmetry or monotonicity properties) specific to the burn-in problem at hand. \smallskip Further problems of interest originate from the necessity of providing appropriate probabilistic characterizations of situations of infant mortality for interdependent life-times. Day: July 5, 1995 Time: 09.30 H.W. Block, University of Pittsburgh, USA: Burn-in F.A. Boshuizen, Erasmus University Rotterdam, The Netherlands; J.M. Gouweleeuw, National Aerospace Laboratory NLR, The Netherlands: Optimal stopping problems for semi-Markov processes C.A. Clarotti, ENEA, Italy: Optimal burn-in of not-ageing pieces of software U. Jensen, Universit{\"a}t Stuttgart-Hohenheim, Germany; G. Heinrich, Universit{\"a}t Stuttgart-Hohenheim, Germany: Bivariate lifetime distributions and optimal replacement J. Mi, Florida International University, USA: On burn-in and its applications G.J. Shanthikumar, University of California, USA: Optimality of burn-in for Schur-constant lifetimes *********************************************************************** Nonclassical Spectral Problems in Fluids, Magnetohydrodynamics and Quantum Mechanics Organizers: H. Langer, Technische Universit{\"a}t Wien, Austria R. Mennicken, Universit{\"a}t Regensburg, Germany Summary: Nonclassical spectral problems frequently occur in applications, e.g. in the theory of fluids, magnetohydrodynamics and quantum mechanics. Typical examples are spectral problems for operators defined by systems of differential operators of mixed order and/or mixed type which may be elliptic, Douglis-Nirenberg elliptic or even non-elliptic. Problems of this type are closely related to boundary eigenvalue problems with polynomial or rational dependence on the eigenvalue parameter $\lambda$. \smallskip The lectures of this minisymposium will present a survey on the methods and results which have recently been developed. The results concern the completeness, minimality and basis properties of eigenfunctions and associated functions, half-range completeness and basis properties, characterisations of the essential spectra, criteria on the eigenvalue accumulation at the boundary of the essential spectra, as well as stability theorems. \smallskip The presentations include applications to various concrete problems such as to the Hain-Luest equation in magnetohydrodynamics or to the Orr-Sommerfeld equation in hydrodynamics. Day: July 4, 1995 Time: 09.30 A. Lifschitz, University of Illinois at Chicago, USA: Spectral problems of hydrodynamics and magnetohydrodynamics R. Mennicken, Universit{\"a}t Regensburg, Germany: On the essential spectrum of some matrix operators M. Faierman, University of the Witwatersrand, South Africa: A boundary eigenvalue problem occuring in magnetohydrodynamics M.S. Agranovich, Moscow, Russia: Spectral properties of nonselfadjoint elliptic integral operators of potential type on smooth and nonsmooth curves and surfaces Day: July 4, 1995 Time: 15.30 H. Langer, Technische Universit{\"a}t Wien, Austria: Spectral properties of a class of rational operator valued functions A. Motovilov, Dubna Joint Institute for Nuclear Research, Russia: Removal of the resolvent-like dependence on the spectral parameter from interactions V. Adamian, Odessa University, Ukraine; H. Langer, Technische Universit{\"a}t Wien, Austria; R. Mennicken, Universit{\"a}t Regensburg, Germany: Absolutely continuous spectral components of some selfadjoint matrix operators H. Schmid, Universit{\"a}t Regensburg, Germany: to be announced B. Pavlov, The University of Auckland, New Zealand; I. Antoniou, Solvay Institute Bruxelles, Belgium; S. Fedorov, The University of Auckland, New Zealand: Time operators in general Lax-Phillips scheme A. Trefethen, Cornell University, USA; L. Trefethen, Cornell University, Ithaca, USA, P. Schmid, University of Washinton, Seattle, USA: Spectra and pseudospectra for pipe Poiseuille flow Day: July 5, 1995 Time: 09.30 C. Tretter, Universit{\"a}t Regensburg, Germany: Spectral analysis for linear differential operator pencils $N-\lambda P$ A.A. Shkalikov, Moscow State University, Russia: Boundary value problems containing a spectral parameter in the boundary conditions A. Dijksma, University of Groningen, The Netherlands: Hamiltonian systems, Titchmarsh-Weyl coefficients, and models H. Gail, Universit{\"a}t Regensburg, Germany: Boundary value problems for first order systems of ODE's with boundary conditions depending polynomially on the eigenvalue parameter S. Naboko, St. Petersburg University, Russia: Operator-valued Nevanlinna functions and their applications in spectral analysis V. Pivovarchik, Civil Engineering Institute of Odessa, Ukraine: Mechanical systems in presence of gyroscopic dissipative and nonconservative positional forces and corresponding operator pencils *********************************************************************** Nonlinear Dynamics and Chaos in Engineering Organizers: J.M.T. Thompson, University College London, United Kingdom S. R. Bishop, University College London, United Kingdom Summary: Engineers and applied scientists are increasingly responding to the revolution in nonlinear dynamics which has uncovered the full complexity inherent in the equations of motion of macroscopic mechanical systems. The unpredictability implied by the chaotic motions and fractal basin boundaries of simple "deterministic" systems is but one feature which nicely epitomises the field. The ubiquitous new phenomena necessitate a change in emphasis away from the classical reliance on perturbation and averaging methods towards the use of computational techniques employing the powerful geometrical concepts developed by mathematicians. There will be two expository general lectures, and other contributions are planned to cover current research in mechanical and industrial applications. In terms of the ICIAM classification, the contributions will relate to {\it ordinary differential equations} and {\it dynamical systems} and {\it solid mechanics} and {\it nonlinear oscillations}. Day: July 3, 1995 Time: 15.30 F. Pfeiffer, Technische Universit{\"a}t M{\"u}nchen, Germany: Mechanical systems with dry friction and impacts J.M.T. Thompson, University College London, United Kingdom: Danger of unpredictable failure due to indeterminate bifurcation D.H. Van Campen, Eindhoven University of Technology, The Netherlands; E.L.B. van de Vorst, TNO Centre for Mechanical Engineering, Delft, The Netherlands: Finite element techniques for nonlinear engineering dynamics S.R. Bishop, University College London, United Kingdom: Topological methods for parametrically excited systems Day: July 6, 1995 Time: 09.30 H. Troger, Technische Universit{\"a}t Wien, Austria: Normal form analysis of non-conservative mechanical systems B. Lagemann, Technische Universit{\"a}t Hamburg-Harburg, Germany; E. Kreuzer, Technische Universit{\"a}t Hamburg-Harburg, Germany: On the cell mapping method for systems with multiple degrees of freedom G. Rega, Universita dell'Aquila, Italy: Nonlinearity and chaos in the dynamics of hanging cables *********************************************************************** Functional Equations with Rescaling - Analysis, Numerics and Applications Organizer: A. Iserles, University of Cambridge, United Kingdom Summary: Functional equations with rescaling are a subject of increasing interest in applied analysis and in computation. Typical are the {\it pantograph equation} ${\bf y}'(t)=A{\bf y}(t)+B{\bf y}(qt)+C{\bf y}'(qt)$, ${\bf y}(0)={\bf y}_0$, and the {\it dilation equation} ${\bf y}_{n+1}=A{\bf y}_n+B{\bf y}_{\lfloor qn\rfloor}$, ${\bf y}_0$ given, where $A,B,C$ are square matrices and $q>0$. Such equations have been used in many applications, from number theory to astronomy, from collection of current by electric locomotives to wavelets. They often display intricate behaviour, inclusive of fractal attractors and of finitely-supported smooth solutions. \smallskip The minisymposium presents a wide range of topics: delay-differential ODEs (G. Derfel, A.N. Sharkovsky), partial functional-differential equations (F. Vogl), wavelet analysis (J. Lagarias), computation of integro-differential equations with proportional delays (Y. Liu) and applications to hydrodynamics (S. Mol\-cha\-nov). Day: July 6, 1995 Time: 09.30 G. Derfel, Ben Gurion University, Israel: Functional equations with rescaling in approximation theory J. Lagarias, AT{\&}T Bell Laboratories, USA: Dilation equations Y. Liu, University of Cambridge, England; A. Iserles, University of Cambridge, United Kingdom: Integro-differential equations and generalized hypergeometric functions A.N. Sharkovsky, Ukrainian Academy of Sciences, Ukraine: Universal phenomena for some difference-differential equations F. Vogl, Technische Universit{\"a}t Wien, Austria: On the Cauchy problem for a class of linear partial FDEs with nonbounded delay *********************************************************************** Nonlinear Ill-Posed Problems: Theory, Numerical Methods and Applications Organizer: A. Yagola, Moscow State University, Russia Summary: Many problems of science, technology and engineering are posed in the form of an operator equation of the first kind with the operator and right part only approximately known. Often such problems turn out to be ill-posed. The theory of solving linear ill-posed problems is advanced greatly today. It is not so in the case of nonlinear ill-posed problems, since the direct transfer of linear theory to the nonlinear case is not possible in general. The possible topics for consideration in the minisymposium are the following: \item{1)} Tikhonov's approach with aposteriori choice of regularization parameter; \item{2)} Iterative methods including the case of operator equations without regularity conditions; \item{3)} Numerical implementation; \item{4)} Applications. Day: July 7, 1995 Time: 09.30 A. Bakushinskii, Institute of Systems Analysis, Russia: Iterative methods for the solution of nonlinear operator equations without conditions of regularity A. Leonov, Moscow Institute of Engineering Physics, Russia: Optimal mathematical design of technical systems and nonlinear ill-posed problems A. Neubauer, Johannes-Kepler-Universit{\"a}t Linz, Austria: On iterative methods for solving nonlinear ill-posed problems A. Yagola, Moscow State University, Russia: Nonlinear ill-posed problems in applications V. Trifonenkov, Moscow Physical Engineering Institute, Russia: Some problems of interpretation of electron microscopy data *********************************************************************** Kinetic Theory Organizers: C. Bardos, C.M.L.A., Cachan, France D. Levermore, University of Arizona at Tucson, USA B. Perthame, University of Arizona at Tucson, USA Summary: The kinetic equations were introduced by Boltzmann at the beginning of this century. Their range of application may be much wider than Boltzmann may have foreseen. In fact, they turn out to be the natural tool to connect, at the level of mathematical physics, the microscopic and the macroscopic description of phenomena. In the meantime, it turned out to be the right tool to describe the qualitative properties of any rarified gas of particles. \smallskip Important theoretical progress, including in particular the work of R. Di Perna and P.L. Lions, has been achieved recently and in the meantime the number of fields of application, ranging from fluid mechanical computations for the reentry in the atmoshere of space vehicles to the description of microscale semiconductors, has expanded exponentially. \smallskip This Minisymposium will try to reflect this evolution. It includes both examples of applications and the basic approach of multiscale analysis, which is intimately associated with this theory. Day: July 6, 1995 Time: 15.30 H. Spohn, Universit{\"a}t M{\"u}nchen, Germany: Derivation of the Boltzmann equation for microspace dynamic T. Paul, University of Paris-Dauphine, France: Perivation of the Vlasov equation through semi classical analysis of linear and nonlinear Schr{\"o}dinger equations J. Sommeria, Lyon Laboratoire de Physique, France: Reducation toward statistical equilibrium state in $3d$ turbulence F. Bouchut, Universit{\'e} d'Orleeans, France; G. Bonnaud, C.E.A., Villeneuve St. Georges, France: Numerical Simulation of relativistic plasmas in hydrodynamic regime C.D. Levermore, University of Arizona, USA: Moment equation for macroscopic ??? of kinetic equations B. Nicolaenko, Arizona State University, USA: Shock paefiles as fluid limiting Boltzmann equations *********************************************************************** Mathematical Techniques for Financial Analysis Organizer: A. Roma, Universit{\`a} di Siena, Italy Summary: The minisympossium addresses some typical issues in the valuation of financial assets, with emphasis on derivative assets and assessment of interest rate risk. Numerical techniques are presented for the solution of asset pricing problems. Theoretical modelling of the real and nominal term structure of interest rates is discussed with application to bond portfolio management. Empirical estimation of term structure models is discussed. Day: July 7, 1995 Time: 15.30 E. Platen, The Australian National University, Australia: Application of stochastic numerical methods in finance M. De Felice, Universit{\`a} di Roma "La Sapienzia", Italy: Asset - liability management models for insurance portfolios C. Mari, Universit{\`a} "G. D'Annunzio" di Chieti, Italy: Arbitrage based valuation of contingent claims on the term structure K.L. Rindell, The Swedish School of Economics, Finland: Equity option models C. Pacati, Universit{\`a} di Perugia, Italy: Term structure estimation methods *********************************************************************** The Dynamics of Thin Fluid Films Organizer: M. Bertsch, Universit{\`a} di Roma "Tor Vergata", Italy Summary: The lubrication approximation for thin liquid films on liquid or solid surfaces leads to nonlinear PDE's taking into account surface tension, liquid-solid interaction and/or gravity. Modeling questions are how to describe film rupture (which forces are dominant?) and the dynamics near a contact line (is the lubrication approximation still valid? does one need slip conditions or an experimental law at the interface?). Many fundamental mathematical questions are closely related to the modeling aspects (strong molecular forces may lead to ill-posed problems; which free boundary conditions at the contact lines lead to well-posed problems and which are the physically relevant ones?). Day: July 7, 1995 Time: 15.30 G.I. Barenblatt, University of Cambridge, England: Dynamics of thin liquid films - new nonlinear problems F. Bernis, Universidad Autonoma de Madrid, Spain: Viscous flows and fourth order nonlinear degenerate parabolic equations E. Dibenedetto, Universit{\`a} di Roma "Tor Vergata", Italy: A singular evolution equation arising in thin film dynamics A. Bertozzi, The University of Chicago, USA: The lubrication approximation for thin viscous films R. Dal Passo, Universit{\`a} di Roma "Tor Vergata", Italy: Thin film dynamics and fourth order PDE's *********************************************************************** Theory of Discretization Methods Organizers: P. Anselone, Oregon State University, Corvallis, USA R. Ansorge, Universit{\"a}t Hamburg, Germany B. Silbermann, Technische Universit{\"a}t Chemnitz-Zwickau, Germany Summary: During the past few decades, a number of investigators have developed general frameworks for the approximate solution of problems of physical interest. This minisymposium will focus on discretization methods for partial differential equations, pseudodifferential equations, and integral equations. Topics to be discussed include Galerkin methods and quadrature methods, for example, as applied to boundary integral equations. The presentations will strike a balance between theory and applications. Thus, the discussion of discretization methods will be broad enough to have a significant range of applications, and the applications presented will be specific enough to address practical considerations. Day: July 3, 1995 Time: 10.00 S. Pr{\"o}{\ss}dorf, WIAS Berlin, Germany; W. Dahmen, RWTH Aachen, Germany; R. Schneider, TH Darmstadt, Germany: Multiscale methods for pseudodifferential equations J. Elschner, WIAS Berlin, Germany; I. G. Graham, University of Bath, UK: Quadrature methods for Symm's integral equation on polygons I.H. Sloan, University of New South Wales, Australia: High order recovery and computable error bounds from finite element calculations R.D. Grigorieff, Technische Universit{\"a}t Berlin, Germany: High order spline Petrov-Galerkin methods with quadrature W. Wendland, Universit{\"a}t Stuttgart, Germany: On Galerkin methods G. Vainikko, Tartu State University, Estonia: Fast solvers of boundary integral and pseudodifferential equations *********************************************************************** Programming Environments for Scientific Computing Organizers: C.P. Ullrich, Universit{\"a}t Basel, Switzerland J.W. von Gudenberg, Universit{\"a}t W{\"u}rzburg, Germany Summary: Programming environments for scientific and engineering applications should combine all aspects of reliable computations on different computer architectures. Software implementation requires powerful programming tools providing, for example, a wide variety of data structures with operations in mathematical notation. More reliable calculations are possible by self-validating numerical algorithms which base on easy access to rounding control and sometimes on extensive use of accurate dotproduct operations. Combining floating-point algorithms and computer algebra systems can even improve the reliability of computations. \smallskip The minisymposium is intended to give insight into recent developments in programming environments for reliable computations and to report experiences with these. Day: July 4, 1995 Time: 09.30 C.P. Ullrich, Universit{\"a}t Basel, Switzerland: Scientific programming language concepts S.M. Rump, Technische Universit{\"a}t Hamburg-Harburg, Germany; R. Lindemann, Technische Universit{\"a}t Hamburg-Harburg, Germany: C extension for scientific computing B.M. Verdonk, University of Antwerpen, Belgium; A. Cuyt, B. Naudts, J. Verelest, University of Antwerp, Belgium: On the integration of software tools for scientific computation B. Philippe, INRIA-IRISA, France; J. Erhel, INRIA/IRISA, Rennes, France; N. Mallejac, CEA, Villeneuve Saint Georges, France: AQUARELS: a problem-solving environment for numerical quality -- review of the project and first use on realistic problems -- H. Burkhart, Universit{\"a}t Basel, Switzerland: Software engineering techniques for parallel computing in science and engineering J.W. Wolff von Gudenberg, Universit{\"a}t W{\"u}rzburg, Germany: Structure of a C++ library for parallel accurate linear algebra *********************************************************************** Function Analytic Methods in Shapes, Boundaries, and Interfaces Organizer: M.C. Delfour, Universit{\'e} de Montr{\'e}al, Canada Summary: Shape analysis and optimization have received a lot of attention over the past two decades among engineers and mathematicians. Much of the emphasis has been on numerical computations and sensitivity analysis. Yet what was really needed is not more computing power but more theoretical work. To-day a large body of literature is available to deal with problem formulations and existence issues. This has significantly broadened the scope of Shape Analysis leading to both smooth and nonsmooth geometric modelling and optimization of physical and technological systems. Bridges have been thrown with older tools such as the theory of minimal surfaces, and in new directions such as free and moving boundary problems and related areas where the geometry is the central variable: shapes, boundaries, interfaces, transition phenomena, etc... Day: July 3, 1995 Time: 15.30 M.C. Delfour, Universit{\'e} de Montr{\'e}al, Canada: Issues and perspectives in geometric modelling, identification, optimization and control L. Bronsard, Mc Master University, Canada: On the existence of a three-layered minimizer for a variational problem with a three-well potential G. Buttazzo, Universit{\'a} di Pisa, Italy: Some shape optimization problems with volume penalization S.K. Mitter, Massachusetts Institute of Technology, USA: Problems and issues in image processing J.-P. Zol{\'e}sio, CNRS - INLN, France: Bounded total curvature sets *********************************************************************** Recent Developments in Direct and Iterative Methods for Sparse Systems of Equations I Organizer: S. Hammarling, NAG Ltd, Oxford, United Kingdom Summary: This minisymposium investigates recent trends in the solution of large sparse systems of linear equations. It is particularly concerned with developments to exploit the power of modern high performance computers. \smallskip Both direct and iterative methods of solution are considered, and all the talks consider issues of parallelism. There are talks on Krylov methods, Cholesky factorization on a distributed memory machine and preconditioning \smallskip This minisymposium is a companion to minisymposium II of the same title. Rather than have one minisymposium devoted to direct methods and one devoted to iterative methods, we aim to discuss both direct and iterative methods in each of the two minisymposia with the aim of promoting discussion and interaction. \smallskip The solution of large sparse systems of equations is important in a wide range of industrial applications and these minisymposia should appeal both to those working in the area of large sparse systems and to those who need to solve such systems as part of their applications. Day: July 4, 1995 Time: 15.30 I.S. Duff, Rutherford Appleton Laboratory, United Kingdom: Block iterative solution of sparse linear equations W.D. Gropp, Argonne National Laboratory, USA: Executable templates for single source uni- and parallel processor implementations of Krylov methods R. Schreiber, NASA Ames Research Center, USA; E. Rothberg, Silicon Graphics Inc., Mountain View, USA: Efficient distributed memory sparse Cholesky factorization H.A. van der Vorst, Utrecht University, The Netherlands; T.F. Chan, University of California at Los Angeles, USA: Preconditioning for sparse linear equations *********************************************************************** Recent Developments in Direct and Iterative Methods for Sparse Systems of Equations II Organizer: I. Duff, RAL, Didcot, England Summary: This minisymposium investigates recent trends in the solution of large sparse systems of linear equations. It is particularly concerned with developments to exploit the power of modern high performance computers. \smallskip Both direct and iterative methods of solution are considered. There are talks on domain decomposition and partitioning schemes, the use of multigrid techniques involving both direct and iterative solvers, and the direct solution of sparse systems on distributed memory computers. \smallskip This minisymposium is a companion to minisymposium I of the same title. Rather than have one minisymposium devoted to direct methods and one devoted to iterative methods, we aim to discuss both direct and iterative methods in each of the two minisymposia with the aim of promoting discussion and interaction. \smallskip The solution of large sparse systems of equations is important in a wide range of industrial applications and these minisymposia should appeal both to those working in the area of large sparse systems and to those who need to solve such systems as part of their applications. Day: July 6, 1995 Time: 09.30 A. Pothen, Old Dominion University, USA: Partitioning and ordering methods D.E. Keyes, Old Dominion University, USA: Newton-Krylov-Schwarz methods: interfacing sparse linear solvers with nonlinear applications C. Douglas, IBM-Watson Research Center, USA: Sparse methods for abstract multilevel solvers on serial and parallel computers R.H. Bisseling, Utrecht University, The Netherlands: Sparse matrix computations on bulk synchronous parallel architectures *********************************************************************** Mathematical Problems in Glass Processing Organizer: R. Mattheij, Eindhoven University of Technology, The Netherlands Summary: Glass processing is an important area of industry which poses many challenging problems for scientists and engineers. The objective of this minisymposium is to focus on the role of mathematics for understanding various aspects of the processes, where modeling, analysis and numerical methods have become and indispensable tool. Topics will include models for thin sheets and fibers, float glass production, thin films in relation to blowing of bottles and fibers. Mathematically the problem consists of solving the creeping flow equations (Stokes) which also occur in a number of related fields (like coalescence in viscous granular compacts), whether or not coupled with an energy equation. Of special interest are some heat transfer problems. One of the speakers will discuss methods to model the effects of radiative heat transfer and its numerical consequences. \smallskip The symposium should be of interest both to engineers and industrial mathematicians. The industrial aspects involved make it very appropriate for ICIAM~95. Day: July 5, 1995 Time: 09.30 G. Vorst, Eindhoven University of Technology, The Netherlands: An integral formulation to simulate the viscous sintering of a two-dimensional lattice of periodic unit cells P.D. Howell, Oxford University, England; B.W. van de Fliert; J.R. Ockendon, Oxford University, U.K.: Mathematical models for slender viscous sheets and fibres D.M. Burley, University of Sheffield, United Kingdom: Problems arising in the computation of the flow of Molten glass during a pressing operation W. Potze, Philips Research Laboratories, The Netherlands; C. Aldridge, Eindhoven University of Technology, The Netherlands: Radiative heat transfer in an infinitely long circular cylindrical tube *********************************************************************** Modeling Challenges in Polymer Processing Organizer: A.N. Hrymak, Mc Master University, Canada Summary: The polymer processing industry is increasingly using CAD/CAM/CAE tools and techniques. The underlying applied mathematics involve gridding methods, free and moving surfaces, mixed type partial differential equations and various methods for discretization (finite volumes, finite elements, boundary elements, spectral element etc.). Each speaker will present the state-of-the-art in their respective area of modeling with emphasis on the features that are of special interest to the industrial and academic applied mathematics community. Each speaker will pose an example challenge problem to the audience, which is beyond current simulation and modeling techniques, but is a significant industrial problem. Day: July 5, 1995 Time: 15.30 M.J. Crochet, Universit{\'e} Catholique de Louvain, Belgium; G. Georgiou, University of Cyprus, Kallipoleos, Cyprus: Modeling the extrusion of viscoelastic fluids: new challenges S.-F. Shen, Cornell University, USA: Fountain effect simulation in highly obstructed cavities J. Vlachopoulos, Mc Master University, Canada: Computer simulation of thermoforming and blowmolding P.A. Tanguy, Ecole Polytechnique, Canada: Recent progress in the modelling of viscous batch mixing *********************************************************************** Wavelet Methods in Applied Analysis and Mechanics Organizer: K. Hackl, Technische Universit{\"a}t Graz, Austria Summary: When attempting to solve numerically a PDE one has to approximate the functions involved in one way or another. Wavelets as approximating functions prove especially effective when dealing with extremely localized or singular phenomena, two quite different examples being turbulence or plastic flow. They can provide an alternative to methods involving Fourier expansions but also to finite element or multigrid techniques or they can be used to adopt these to specific situations. \smallskip Due to the hierarchic nature and rapid convergence of wavelet expansions it is often possible to implement in a natural way adaptive schemes for wavelet based methods leading to very efficient algorithms. The minisymposium will present the mathematical concepts as well as their computational implementations and results. Applications to fluid and solid mechanics should demonstrate the power and usefulness of the new methods. Day: July 6, 1995 Time: 15.30 X. Zhou, Rice University, USA; A. Rieder, Universit{\"a}t des Saarlandes, Saarbr{\"u}cken, Germany; R. O. Wells, Rice University, Houston, Texas, USA: A robust wavelet multilevel solver for anisotropic equations K. Hackl, Technische Universit{\"a}t Graz, Austria: A wavelet based elastoplastic beam model J.E. Weiss, AWARE Inc., USA: Wavelets, turbulence and boundary value problems for partial differential equations S. Jaffard, CERMA, France: Some examples of wavelet methods in functional analysis and the study of turbulence L. Wei, ZhongShan University, China: Wavelet solutions to the plane boundary value problems of quasistatic linear thermoelasticity *********************************************************************** Dynamics of Tethered Satellite Systems Organizer: H. Troger, Technische Universit{\"a}t Wien, Austria Summary: Tethered satellite systems, that is, satellites in orbit, joined by a thin flexible cable (up to 100 km length) are a promising future industrial space technology which has had a -- not completely successful -- first test in August 1992 during the Atlantis shuttle flight. \smallskip The mathematical formulation results in a stiff set of nonlinear partial differential equations for the tether and nonlinear ordinary differential equations for the satellites, with time dependent coefficients if the tether length is varied. \smallskip Engineering aspects concerning the modelling and the formulation of the equations of motion and mathematical aspects concerning their theoretical and computational treatment (by three different numerical methods) will be given. The theoretical mathematical analysis relates to the stability analysis of relative equilibria by means of the reduced energy momentum method and to the occurence of chaotic motions due to impacts. Day: July 5, 1995 Time: 09.30 H. Troger, Technische Universit{\"a}t Wien, Austria; W. Steiner, Technische Universit{\"a}t Wien, Austria: Applications of tethered satellite systems, modelling and derivation of the equations of motion V. Beletsky, Russian Academy of Sciences, Russia: Regular and chaotic motions of a two satellite tether system A. Misra, McGill University, Canada: Nonlinear vibrations of low-tension orbiting tethers M. Pasca, Universit{\'a} di Roma "La Sapienzia", Italy; F. Vestroni, Universit{\'a} di Roma "La Sapienza", Italy; A. Luongo, Universit{\'a} de L'Aquila, Italy: Stability and bifurcations of transversal motions of an orbiting string with a longitudinal control force W. Seemann, Technische Hochschule Darmstadt, Germany; P. Hagedorn, Technische Hochschule Darmstadt, Germany: Analysis of a tethered satellite with a new type of deployment and retrieval mechanism M. Krupa, University of Groningen, The Netherlands: Stability analysis of relative equilibria of a tethered satellite system by means of the energy momentum method W. Steiner, Technische Universit{\"a}t Wien, Austria; A. Kuhn, Technische Universit{\"a}t Wien, Austria; J. Zemann, Technische Universit{\"a}t Wien, Austria: Simulation of the motions of a tethered satellite system by means of three different numerical methods *********************************************************************** Dynamics of Fluid Conveying Tubes Organizer: H. Troger, Technische Universit{\"a}t Wien, Austria Summary: Besides their technical importance (e.g. fire hoses) fluid conveying tubes can be considered as a mechanical model problem to demonstrate the efficiency of the mathematical methods of local and global bifurcation theory. The solutions of the governing system of nonlinear partial differential equations can easily be tested experimentally and they cover a great variety of motions reaching from regular to chaotic ones. In addition by simple natural parameter variations different system symmetries ($O(2)$ and $D_n$) and all different codimension 2 bifurcations of vectorfields can be produced. \smallskip Experimental, theoretical and numerical aspects of local and global bifurcation theory explaining regular and chaotic tube motions are given. Global bifurcation aspects concern the numerical calculation of wiggly heteroclinic orbits for an articulated tube and the existence and stability of heteroclinic cycles. Day: July 7, 1995 Time: 09.30 H. Troger, Technische Universit{\"a}t Wien, Austria; A. Steindl, Technische Universit{\"a}t Wien, Austria: The significance of fluid conveying tubes as a model problem in bifurcation theory M. Yoshizawa, Keio University, Japan: Nonlinear lateral vibration of a hanging flexible pipe conveying fluid A. Bajaj, Purdue University, USA; C. Folley, Purdue University, USA: Three-dimensional dynamics of a continuous cantilever tube conveying pulsatile flow A.R. Champneys, University of Bristol, United Kingdom: Nonlinear dynamics of articulated pipes conveying fluid A. Steindl, Technische Universit{\"a}t Wien, Austria: Heteroclinic cycles in the dynamics of a fluid conveying tube with symmetric elastic support *********************************************************************** Mathematical Models for Hysteresis in Fatigue and Friction Organizer: M. Brokate, Universit{\"a}t Kiel, Germany Summary: Models for friction often explicitly incorporate hysteresis. Fatigue is connected to hysteresis since the local stress-strain hysteresis loop constitutes a basic event relevant for damage assessment. The rainflow counting method, widely used in the automobile industry, exploits this relationship for damage estimation. \smallskip Mathematics provides a framework to discuss wellposedness, stability and energy dissipation, and addresses the complications stemming from the globally nonsmooth character and piecewise structure of hysteresis. This forms a basis for system design, control, and numerical approximation. Particular needs and open problems arise when vector quantities determine the hysteresis. \smallskip We put emphasis on rate independent processes. Day: July 7, 1995 Time: 15.30 M. Fr{\'e}mond, CNRS-LCPC, France: Hysteresis and dissipative phenomena in solid mechanics J. Macki, University of Alberta, Canada: Periodic solutions to ordinary differential equations with hysteresis term P.-A. Bliman, INRIA, France; T. Bonald, INRIA, France; M. Sorine, INRIA, France: Hysteresis operators and tyre friction models. Application to vehicle dynamic simulations I. Rychlik, University of Lund, Sweden: Extremes, rainflow cycles and damage functionals in continuous random processes M. Brokate, Universit{\"a}t Kiel, Germany; K. Dre{\ss}ler, Tecmath GmbH, Kaiserslautern, Germany; P. Krejc{\'i}, Academy of Sciences, Prag, Czech Republic: Hysteresis operators, constitutive laws and rainflow counting M.A. Krasnosel'skii, Voronezh State University, Russia: New problems in systems with hysteresis *********************************************************************** Numerical Methods for Time-Dependent PDEs Organizer: U.M. Ascher, University of British Columbia, Canada Summary: The quest for numerical methods for the efficient integration of time-dependent PDEs began at the earliest days of the computer era, and simple, well-known methods are classroom material. Recently, however, some interesting advances have been made in several less usual directions, in an effort to obtain proven and efficient methods for more difficult or time-consuming problems, and to analyze the errors generated in less straight-forward discretizations. The purpose of this minisymposium is to explore these directions, both theoretically and practically. \smallskip Among the various approaches that have been recently investigated are both unusual variants of the method of lines (such as implicit-explicit methods or waveform methods) and methods which combine stepping in time and in space (such as various characteristics-based methods). Such methods and their properties will be discussed by the speakers. Day: July 3, 1995 Time: 10.00 J. Janssen, Katholieke Universiteit Leuven, Belgium; S. Vandewalle, Katholieke Universiteit Leuven, Belgium: Accelerated waveform relaxation methods for the parallel solution of time-dependent partial diffential equations S. Ruuth, University of British Columbia, Canada: Implicit-explicit methods for time-dependent PDEs S. Ta'asan, Carnegie-Mellon University, USA: to be announced B. Wetton, University of British Columbia, Canada: Analysis of the spatial error of finite difference methods for viscous incompressible flow *********************************************************************** Mathematical Developments on Wave Problems Organizer: G.M. Iooss, Universit{\`e} de Nice, France Summary: It is intended to confront different schools and popularize recent important results on the existence and stability of various types of waves, and on the relevance of certain models with respect to the physical original problem. \smallskip It is worthwhile to clarify what is mathematically proved, with respect to what is physical or numerical evidence. A special emphasis will be made on how relevant is the adoption of dynamical systems techniques, and on progresses in understanding some famous model equations as for instance on the water wave problem and its related simplified models. The status of the justification of the Ginzburg-Landau equation for hydrodynamical stability problems in extended domains will be one of the subjects. \smallskip Intended audience: PDE in unbounded domains - Fluid Mechanics, Dynamical Systems (conservative or not). Day: July 3, 1995 Time: 15.30 J.L. Bona, Ecole Normale Superiure de Cachan, France; J.-C. Saut, Universit{\'e} de Paris - Sud, Orsay, France: Two-way propagation of water waves A. Mielke, Universit{\"a}t Hannover, Germany: Mathematical treatment of sideband instabilities R.L. Pego, University of Maryland at College Park, USA: Spectral stability of solitary waves J.F. Toland, University of Bath, United Kingdom: The existence of infinitely many multi-humped solutions of the capillary-gravity wave problem Day: July 4, 1995 Time: 09.30 G. Schneider, Universit{\"a}t Hannover, Germany: Description of waves by modulation equations -- rigorous results M. Weinstein, The University of Michigan, USA: Asymptotic stability of nonlinear bounded states G.M. Iooss, I.U.F. Universit{\'e} de Nice, France: Codimension-$2$ singularity in $2D$ capillary-gravity wave problems *********************************************************************** Dynamical Systems with Symmetry and Cosymmetry Organizer: V.I. Yudovich, Rostov State University, Russia Summary: Dynamical systems of physics and engineering often appear to be degenerate from the point of view of the general theory. The usual reason for degeneration is symmetry. The other reason, namely the cosymmetry, was found recently for PDE systems of plane filtrational convection of fluid in porous media, the systems of classical dynamics with symmetric potential energy, hydrodynamical problems, MHD and chemical kinetics. \smallskip The main problems are following: \item{i)} New systems with symmetries and cosymmetries; \item{ii)} Numerical and analytical investigation of dynamical systems; \item{iii)} Effects of interactions of symmetries and cosymmetries, intersections of bifurcations. Day: July 7, 1995 Time: 09.30 V.I. Yudovich, Rostov State University, Russia: Cosymmetry and dynamical systems G. Zaslavsky, New York University, USA: Symmetry, quasisymmetry and chaos D. Lyubimov, Perm Sate University, Russia; Bratsun. Perm State University, Russia: Effect of cosymmetry on bifurcations at convection in porous medium N. Petrovskaya, Rostov State University, Russia; V.V. Kolesov, Rostov State University, Russia; S.N. Ovchinnikova, Rostov State University, Russia; V.I. Yudovich, Rostov State University, Russia: Onset of chaos through intersections and bifurcations in Couette-Taylor flow B. Karas{\"o}zen, Middle East Technical University, Turkey: Hamiltonian's with symmetry: numerical studies Day: July 7, 1995 Time: 15.30 V. Govorukhin, Rostov University, Russia: Computer experiments in cosymmetric models V. Tsybulin, Rostov State University, Russia; V.I. Yudovich, Rostov State University, Russia: Attractors of quadratic mapping of plane: computer experiment and analytical treatment M. Golubitsky, University of Houston, USA; B. Dionne, University of Ottawa, Canada; I. Stewart, University of Warwick, UK: Coupled cell systems with internal symmetries P. Chossat, Universit{\`e} de Nice Sophia-Antipolis, France; F. Guyard, Universit{\'e} de Nice Sophia-Antipolis, Valbonne, France: Symmetry-forced heteroclinic cycles in the sperical Benard problem *********************************************************************** Geometric Methods in the Theoretical and Numerical Modelling of Structured Continua Organizers: K. Hutter, Technische Hochschule Darmstadt, Germany B. Svendsen, Technische Hochschule Darmstadt, Germany C. Sansour, Technische Hochschule Darmstadt, Germany\\ A. Bertram, BAM Berlin, Germany Summary: A wide variety of different materials are characterized in materials science and mechanics by possessing what is referred to as a "structure" or "microstructure". Examples include dislocations, fractures, texture, pores, material inhomogeneity, electromagnetic interactions and so on. All of these models are based in one way or another on an extension of the classical material body concept to account for the type of structure present. Since this extension generally involves some kind of {\it non-Euclidean} space whose geometry reflects that of the structure involved, differential geometric methods are essential to the formulation of models for such materials. \smallskip The mini-symposium is intended to address different aspects of the applications of geometric methods in the modelling and finite element simulation of structured continua. Lectures in theory, applications, and numerics will be presented by well-known researchers working actively in these areas. Day: July 3, 1995 Time: 15.30 A. Bertram, BAM, Germany; M. Kraska, TU Berlin, Germany: Geometrical description of materials with elastic range M. Brocato, Ecole Nationale des Ponts et Chauss{\'e}es, France; Ph. Tamagny, A. Ehrlacher, Ecole Nationale des Ponts et Chauss{\'e}es, Noisy Le Grand, France n: On the relation between continua with microstructure and a 'deep space' model for polycrystals G. Capriz, Universit{\'a} di Pisa, Italy: Invariance and balance in structured continua G. del Piero, Instituto di Ingeneria, Italy: Structured stress tensors for continua M. Elzanowski, Portland State University, USA: Continous distribution of defects in material bodies with microstructure Day: July 6, 1995 Time: 09.30 M. Epstein, University of Calgary, Canada; G.A. Maugin, Universit{\'e} P{\&}M Curie, Paris, France: Geometrical material structure of finite-strain elasticity and anelasticity V.I. Levitas, Universit{\"a}t Hannover, Germany: Averaged description of phase transformations in inelastic materials J. Makowski, Ruhr-Universit{\"a}t Bochum, Germany: Some aspects of mechanical and electro-magnetic interactions W. Muschik, Technische Universit{\"a}t Berlin, Germany; H. Ehrentraut, C. Papenfu{\ss}, S. Blenk, TU Berlin, Germany: Structural roots of different phenomenological theories of liquid crystals Day: July 7, 1995 Time: 09.30 D.R. Owen, Carnegie-Mellon University, USA: Derivation of balance laws for continua undergoing disarrangements P. Roug{\'e}e, E.N.S. de Cachan, Universit{\'e} Paris VI, France: Micro-structure elements of continuous media C. Sansour, Techn. Hochschule, Germany: On geometric aspects in the theory and finite element computation of structured continua with application to the Cosserat continuum B. Svendsen, Technische Hochschule Darmstadt, Germany: Fibre-bundle model for structured continua *********************************************************************** Diffusive Phase Transitions Organizers: J. Sprekels, WIAS Berlin, Germany P. B{\'e}nilan, Universit{\'e} de Franche-Comt{\'e}, France Summary: Diffusive phase transitions occur in almost all fields of science and play a fundamental role in many technological processes. Among these are the casting of high quality steels, the growth of semiconducting crystals, the production of magnetic materials and the spinodal decomposition of binary alloys, to name only a few. \smallskip The minisymposium deals with mathematical models for different classes of diffusive phase transitions, namely \item{$\bullet$} spinodal decomposition (temperature-dependent models of Cahn-Hilliard type), \item{$\bullet$} liquid-solid transitions (models of Stefan and phase-field type), \item{$\bullet$} austenite-pearlite-martensite transitions in eutectoid carbon steels (models using laws from metallurgy like Scheil's additivity rule). \smallskip For these classes of models, qualitative and quantitative results will be presented which have been established in the past few years. Day: July 3, 1995 Time: 15.30 N. Kenmochi, Chiba University, Japan: Large-time behaviour for non-isothermal models of phase transitions P. Laurencot, Universit{\'e} de Nancy I, France: Global solutions to a Penrose-Fife phase-field model under flux boundary conditions for the inverse temperature P. Colli, Universit{\`a} di Pavia, Italy; J. Sprekels, WIAS Berlin, Germany: Nonlinear Stefan problems obtained as asymptotic limits of a Penrose-Fife model D. H{\"o}mberg, WIAS Berlin, Germany: Mathematical models for the phase transitions in eutectoid carbon steels *********************************************************************** Invariant Imbedding Methods in Direct and Inverse Scattering Organizers: L. Fishman, Iowa State University, Ames, USA J.P. Corones, Iowa State University, Ames, USA Summary: This minisymposium addresses both direct and inverse scattering problems of applied and industrial interest. The mathematical modeling generally divides between effective one-dimensional formulations incorporating increasingly complex physics and irreducible multidimensional formulations of relatively simple environments. Key ideas and methods from wave field splitting and invariant imbedding provide the unifying theme in the analysis and computation of the above problems in both the time and frequency domains. Extensive numerical results and exact solution cases for the effective one-dimensional problems, and the progress in the development of the multidimensional algorithms will be discussed in detail. Day: July 3, 1995 Time: 10.00 M.V. de Hoop, Schlumberger Cambridge Research, England; A.J. Haines, Institute for Geological and Nuclear Sciences, Wellington, New Zealand: Aspects of wave field decomposition in inhomogeneous and anisotropic media G. Kristensson, Lund University, Sweden: Wave splitting techniques for hyperbolic systems F.A. Grunbaum, University of California, USA: Diffuse tomography G. Zhang, Chinese Academy of Sciences, P. R. of China: Initialization of two-point boundary value problems S. He, Royal Institute of Technology (KTH), Sweden: Some numerical aspects for time-domain acoustic inverse problems in $\R ^3$ using wave-splitting approach Day: July 5, 1995 Time: 15.30 L.P. Nizhnik, Ukrainian Academy of Sciences, Ukraine: Inverse scattering problem for hyperbolic equations L. Fishman, Iowa State University, USA: Wave field splitting, invariant imbedding, and phase space methods in direct and inverse frequency domain wave propagation modeling S. Str{\"o}m, Royal Institute of Technology (KTH), Sweden; S. He, Royal Institute of Technology, Sweden; J. Lundstedt, Royal Institute of Technology, Sweden: Multiparameter reconstruction for an inhomogeneous $LCRG$ transmission line G.R. Wickham, University of Manchester, United Kingdom: Wave splitting across arbitrary surfaces and wavefront asymptotics for the reflection operator Day: July 7, 1995 Time: 15.30 V.H. Weston, Purdue University, USA: On the wave-splitting, layer-stripping approach to the $3-D$ inverse problem D.J. Wall, University of Canterbury, New Zealand; I. {\"U}berg, Lund University, Sweden; G. Kristensson, Lund University, Sweden: Propagation of transient waves on a time-varying transmission line - A paradigm for nonlinear problems M.I. Belishev, Russian Academy of Sciences (POMI), Russia: On a canonical realization of dynamical systems with boundary control (BC-method) A. Karlsson, Lund University of Technology, Sweden: Wave propagators for transient waves in one-dimensional media *********************************************************************** Vortex Dynamics Organizer: E. Krause, RWTH Aachen, Germany Summary: Five contributions relating to vortex dynamics are proposed. They include the computation of the motion of slender vortex filaments; of the motion of a vortex in an initially stable layer of thermally stratified viscous fluid with a free surface; the computation of sound generated by two vortex rings with different core size; the numerical simulation of shock-induced breakdown on a delta wing, and the numerical investigation of the transition from bubble to spiral type breakdown. The results demonstrate that the complex flow structure of the vortices, their time-dependent variations and their migrations can reliably be simulated. Day: July 5, 1995 Time: 09.30 R. Klein, RWTH Aachen, Germany; O.M. Knio, John Hopkins University, Baltimore, MD 21218, USA; L. Ting, New York University, NY 10012, USA: Accurate numerical computation of stretched, high-Reynolds slender numbers vortices in three space dimensions R. Peyret, University of Nice - Sophia Antipolis, France; T. Pages, Universit{\'e} de Nice, France; T. Pasquetti, Universit{\'e} de Nice, France: Interaction of vortices with free surfaces K. Ishii, Nagoya University, Japan; S. Adachi, Institute of Computational Fluid Dynamics, Tokyo, Japan; K. Kuwahara, Institute of Computational Fluid Dynamics, Tokyo, Japan: Vortex Sound O.A. Kandil, Old Dominon University, USA; C. Liu, NASA Langely Research Center, Hampton, VA 23665-5225, USA: Shock-induced vortex breakdown on a delta wing W. Althaus, RWTH Aachen, Germany; E. Krause, RWTH Aachen, Germany: Bubble- and spiral-type breakdown *********************************************************************** Inverse Problems in Structural Dynamics Organizers: G.M.L. Gladwell, University of Waterloo, Canada H.G. Natke, Universit{\"a}t Hannover, Germany Summary: Inverse vibration problems are a subset of parameter identification problems for which there is a mechanical vibration model linking the parameters to the data, the behaviour. Even if the vibration system is linear, the functional relation which expresses the behaviour, e.g. the frequency response, in terms of the parameters is highly nonlinear. The problem of finding the effect of the parameters on the data is one of sensitivity. \smallskip Inverse problems pose peculiar difficulties: insufficient and inaccurate data; non-unique and/or non-existent solutions; ill-posedness which will, for instance, mean that small changes in data may lead to large changes in the solution. The study of inverse problems is the search for methods and algorithms which can cope simultaneously with all these difficulties. Day: July 6, 1995 Time: 09.30 F. Vestroni, Universit{\`a} di Roma "La Sapienzia", Italy; D. Capecchi, University of Rome La Spienza, Italy: Damage detection in vibrating beam structures using experimental modal data and finite element models C. Davini, Universit{\`a} degli Studi di Udine, Italy: Damage localization in a $3$-dim truss B. Piombo, Politecnico di Torino, Italy; A. Fasana, Politecnico di Torino, Italy: Identification of mechanical systems by deconvolution techniques A. Fregolent, Universit{\`a} degli Studi Roma "La Sapienzia", Italy; A. Sestieri, Universit{\'a} degli Studi di Roma "La Sapienza", Italy: Force identification from noise-polluted vibration measurements D.J. Inman, Virginia Polytechnic Institute, USA: Symmetry in inverse eigenvalue problems for mechanical systems Day: July 6, 1995 Time: 15.30 Y. Ben-Haim, Technion-Israel Institute of Technology, Israel; Y. Freund, Israel Institute of Technology, Haifa, Israel: Selectively sensitive identification of connectivity matrices of linear elastic systems G. Lallement, Universit{\'e} de Franche-Comt{\'e}, France: Identification of external forces in the frequency domain: reduction of ill conditioning J.E. Mottershead, University of Wales Swansea, United Kingdom; M.I. Friswell, University of Wales Swansea, UK; G.H.T. Ng, University of Wales Swansea, UK; Y. Zhang, University of Wales Swansea, UK: On the application of eigenvalue sensitivity data in structural dynamics model updating J.E. Mottershead, University of Wales Swansea, United Kingdom; M.I. Friswell, University of Wales, Swansea, UK; H. Ahmadian, University of Wales, Swansea, UK; G.H.T. Ng, University of Wales, Swansea, UK: Finite element modelling of joints and boundary conditions M. Tanaka, Shinshu University, Japan; T. Matsumoto, L. Huang, Sinshu University, Nagano, Japan: Computer simulation for active control of vibration in beams and framed structures Day: July 7, 1995 Time: 09.30 G. Bruckner, WIAS, Germany: A decomposition method for the numerical solution of ill-posed linear first kind integral equations G.M.L. Gladwell, University of Waterloo, Canada; H. Ahmadian, University College of Swansea, Wales: The modelling of joints for structural vibration Y.M. Ram, University of Adelaide, Australia: Inverse problems for a vibrating rod: Continuous versus discrete models A. De Stefano, Politecnico di Torino, Italy; R. Ceravolo, Politecnico di Torino, Italy; D. Sabia, Politecnico di Torino, Italy: Neural network approaches in structural identification and defect detection S. Handrock-Meyer, WIAS, Germany: Ill-posed problems and wavelets *********************************************************************** Seismic Inverse Scattering Organizer: G.C. Herman, Delft University, The Netherlands Summary: The inverse scattering problem of determining the earth's structure from seismic exploration is a research topic of great interest and activity. There are different approaches to the seismic inverse problem. One of them is the imaging approach, being a (usually asymptotic) procedure for directly expressing subsurface parameters in terms of seismic measurements. Another (and more recent) approach is the iterative minimization of a data misfit function. \smallskip Despite the fact that much progress has been made recently, many challenges still remain because the underlying problem is nonlinear, the available data is incomplete and the true complexity of the propagation problem is only approximately modeled. \smallskip The speakers will review a number of research topics of current interest -- comparisons between different methods, development of techniques that efficiently account for small-scale variations in the earth's structure, robust asymptotic imaging methods that do not break down in the presence of caustics and the use of asymptotic methods for accelerating the rate of convergence of the nonlinear seismic problem. Day: July 4, 1995 Time: 09.30 S. Gray, Amoco Production Research, USA: True-amplitude seismic imaging: three inverse scattering methods N. Bleistein, Colorado School of Mines, USA: Analysis of a continuum analog of the WRW form of the Berkhout approach to modeling, migration and inversion C.P.A. Wapenaar, Delft University of Technology, The Netherlands: Acoustic one-way wave theory in highly heterogeneous media and its application in acoustic modelling and imaging A. ten Kroode, Shell Research, The Netherlands; D.-J. Smith, Shell Research, The Netherlands; A.R. Verdel, Shell Research, The Netherlands: Linearized inverse scattering in the presence of caustics W.W. Symes, Rice University, USA; C. Nolan, Rice University, USA: Waveform inversion for velocities and reflectivities in the presence of caustics A.G. Sevink, Delft University of Technology, The Netherlands; G.C. Herman, Delft University of Technology, The Netherlands: Nonlinear asymptotic seismic inversion *********************************************************************** Nonlinear Oscillations Organizers: E. Kreuzer, Technische Universit{\"a}t Hamburg-Harburg, Germany W. Wedig, Universit{\"a}t Karlsruhe Summary: Computer simulations of nonlinear dynamic systems show a high complexity of behaviour documented by rich scenarios of bifurcations from periodic motions into unpredictable chaos. Moreover, there is a close relationship between random vibrations and chaotic behaviour in the sense that there is no strict boundary between random processes and chaotic motions. In order to understand these phenomena, analytical and numerical methods of investigations have been developed to obtain more sophisticated and powerful descriptions of local and global bifurcations. \smallskip Recent developments utilize the theory of invariant measures and control sets. The interaction of both numerical and analytical methods allows to characterize the global bifurcation behaviour of the nonlinear systems and to quantify the stability of bifurcated solutions by the associated Lyapunov exponents. Day: July 3, 1995 Time: 10.00 F. Colonius, Universit{\"a}t Augsburg, Germany; W. Kliemann, Iowa State University, USA: Multistability for stochastic models and control sets E. Kreuzer, Technische Universit{\"a}t Hamburg-Harburg, Germany: Nonlinear oscillations of a crane ship P.C. M{\"u}ller, Bergische Universit{\"a}t-GH Wuppertal, Germany: Stability of nonlinear descriptor systems H. Troger, Technische Universit{\"a}t Wien, Austria; A. Steindl, Technische Universit{\"a}t Wien, Austria: Nonlinear oscillations in symmetric systems W.V. Wedig, Universit{\"a}t Karlsruhe, Germany: Invariant measures of perturbed dynamical systems V. Wihstutz, University of North Carolina at Charlotte, USA: Stabilizing versus destabilizing classes of noise *********************************************************************** Shape Memory Alloys: Modelling, Mathematical Analysis, Numerical Simulation and Applications Organizer: K.-H. Hoffmann, Technische Universit{\"a}t M{\"u}nchen, Germany Summary: Shape memory alloys belong to the class of intelligent materials. Depending on temperature they show plastic or elastic behaviour. It is characteristic for these solids that they can be permanently deformed at a low temperature and recover their original shape simply by heating. Moreover, educated alloys show two different shapes at low or high temperature, respectively. These phenomena are due to austenite - martensite phase transitions, which can be induced by thermal treatment or mechanical forces. The stress-strain curves of shape memory alloys show temperature dependent hysteresis effects. \smallskip Different approaches to simulate the behaviour of these alloys use statistical mechanics, Landau theory, phasefield models or microscopic models. The resulting constitutive equations have to take into account the coupling effects which yield systems of nonlinear partial or sometimes ordinary differential equations. These systems are solved, e.g., by finite element methods. \smallskip Optimization problems occur when shape memory material is used to suppress vibration or noise. Modelling the shape or position of the material yields optimization problems as well. The comparison of some models with experimental results leads to the problem of identifying parameters. \smallskip There are many applications for shape memory alloys, for example in medicine, the aerospace industry, electrical engineering and robotics. Day: July 4, 1995 Time: 15.30 J. Melcher, DLR, Germany; E.J. Breitbach, German Aerospace Research Establish, Braunschweig, Germany; R. Lammering, German Aerospace Research Establish, Braunschweig, Germany: Shape memory materials as multifunctional actuators and sensors in adaptive structural systems P. Colli, Universit{\`a} di Pavia, Italy: Remarks and results on the Fr{\'e}mond model for shape memory alloys M. Fr{\'e}mond, CNRS-LCPC, France: A $3D$ predictive theory of shape memory alloy evolutions D. Kinderlehrer, Carnegie-Mellon University, USA: Computational hysteresis in magnetic and magnetoelastic systems G. W{\"o}rsching, Technische Universit{\"a}t M{\"u}nchen, Germany: Numerical simulation of shape memory alloys J. Sprekels, WIAS, Germany: Mathematical models of stress -- and deformation controlled experiments in shape memory alloys *********************************************************************** Oxidation and Dopant Diffusion Processes in Semiconductor Devices Organizer: K.-H. Hoffmann, Technische Universit{\"a}t M{\"u}nchen, Germany Summary: Generally speaking, the aim of process simulation is the prediction of dopant distribution and of layer structure at the end of the production process of microelectronic circuits. In particular, oxidation and dopant diffusion processes are an interesting field of research because of the difficult modelling problems they pose. \smallskip\noindent There is a host of mathematical models of various degrees of complexity and completeness, which give rise to complicated free boundary value problems associated with nonlinear partial differential equations. Some of those have already been treated analytically, proving existence and uniqueness results of solutions as well as considering their asymptotic behaviour. \smallskip\noindent The numerical simulation in two or three space dimensions is still limited due to the enormous difficulties of automatic grid handling of the time dependent domains and doping profiles. Other difficulties are the complex nonlinearities inherent in the problems and also the great numbers of parameters appearing in the equations which can not be measured directely. \smallskip\noindent In the current minisymposium we want to present a variety of models describing different process steps in microchip fabrication and want to demonstrate a series of numerical simulations occuring in typical applications. We also want to discuss mathematical and analytical results and to examine numerical methods which are typical in that area. Another aim is an exchange of ideas between industry and research, aiming to explore new manufacturing methodes and new modelling ideas. Day: July 5, 1995 Time: 15.30 W. Klein, Siemens AG, Germany: Industrial application of semiconductor process simulation J.R. King, University of Nottingham, United Kingdom: Asymptotic results for impurity diffusion in compound semiconductors J. Lorenz, Fraunhofer Institut f{\"u}r Integrierte Schaltungen, Germany: Physical models in semiconductor process simulation and resulting numerical requirements W. Merz, Technische Universit{\"a}t M{\"u}nchen, Germany: Theoretical aspects in the process-simulation N. Strecker, ETH-Z{\"u}rich, Switzerland: Applications and numerical aspects in process simulation *********************************************************************** Industrial and Applied Mathematics in South America Organizer: M.A. Raupp, Universidade de Sao Paulo, Brazil Summary: The minisymposium will be devoted to industrial and applied mathematics being developed in South America, covering areas having significant activity nowadays: \smallskip oil reservoir simulation, seismic imaging, combustion modelling, crack detection, portfolio optimization, among others. \smallskip The minisymposium will create opportunities, in discussing case studies of mathematical modelling of real life problems, for mathematicians to exchange experiences of value in the context of the developing world. Day: July 5, 1995 Time: 09.30 M.F. Tom{\'e}, Universidade de Sao Paulo, Brazil: Numerical simulation of unsteady Newtonian and non-Newtonian free surface flows C.R. Maliska, Federal University of Santa Catarina, Brazil; A.F.C. Silva, A.R. Cunha, M.A. Livraman, Federal University of Santa Catarina, Brazil: A $3D$ numerical method with boundary-conforming discretization for petroleum reservoir simulation M. Tygel, Universidade Estadual De Campinas, Brazil; P. Hubral, Universit{\"a}t Karlsruhe, Germany; J. Schleicher, Universit{\"a}t Karlsruhe, Germany: Amplitude preserving $3-D$ seismic reflection imaging C.E. D'Attellis, Centro de C{\'a}lculo Cient{\'i}fico (CNEA-CAC), Argentina; J.L. Mancilla Aguilar, Universidad de Buenos Aires, Argentina; R. A. Garcia, Universidad de Buenos Aires, Argentina: Nonlinear control systems: trajectory tracking with bounded controls, state constraints and observers D. Marchesin, Instituto de Matem{\'a}tica Pura e Aplicada, Brazil; J. Glimm, State University of New York, Stony Brook, USA; A.C. Grossi, State University of New York, Stony Brook, USA: Front tracking for tracer flow in petroleum reservoirs A.D. Loula, LNCC/CNPq, Brazil: Finite element solution of tracer injection simulation Day: July 5, 1995 Time: 15.30 L.P. Franca, University of Colorado at Denver, USA; C. Farhat, University of Colorado at Boulder, USA: Unusual stabilized finite element methods L.R. Sinay, National Space Research Institute, Brazil: Asymptotic-modal solutions of the Sivarhinsky-type equations D.A. Tarzia, FCE - Universidad Austral, Argentina: Some new results on the one phase supercooled Stefan problem A. Pignotti, FUDETEC, Argentina; E. Altschuler, FUDETEC, Buenos Aires, Argentina; J. Paiuk, TECHINT, Buenos Aires, Argentina: Monte Carlo simulation of crack detection in steel pipes M. Thompson, Universidade Federal do Rio Grande do Sul, Brazil: Approximate inertial manifolds for rotating internal dilute viscoelastic flows J.M. Stern, Universidade de Sao Paulo, Brazil: Critical-point, a software for portfolio optimization Day: July 6, 1995 Time: 15.30 M.A. Raupp, Universidade de Sao Paulo, Brazil: Simulation of heat conduction in liquid Helium J.R. Claeyssen, Universidade Federal do Rio Grande do Sul, Brazil: Response functions in time and frequency domain M. Lentini, CESMa Universidad Simon Bol{\'i}var, Venezuela; J.C. D{\'i}az, University of Tulsa, USA; A. Gersztenkorn, Tulsa Research Center, USA: Shifted preconditioning for the QMR method P.R. de Avila Zingano, Universidade Federal do Rio G. do Sol, Brazil: Diffusion waves of viscous conservation laws C.C. Gonzaga, Federal University of Rio de Janeiro, Brazil: Large steps path following algorithms for linear programming and linear complementarity problems J.W.C. Sotil, Universidad Nacional de Ingeniera, Peru: Approximation of quasi-geostrophic states and global attractor estimates *********************************************************************** Inverse Problems in Rheology Organizers: R. Anderssen, CSIRO, Canberra, Australia A.R. Davies, University of Wales, UK Summary: In industrial rheological applications, much decision-making centres around answering questions about the properties of the non-Newtonian fluid under examination. One first collects experimental information such as shear- and extensional-dependence on viscosity, viscoelastic response, and normal stress behaviour. However, the mathematical models, which relate such indirect measurements back to the properties of the fluid, correspond to inverse problems of one form or another. This is a direct reflection of the fact that one can only determine the properties of a fluid, such as viscosity, via measurements performed on that fluid. Consequently, the formulation and solution of inverse problems plays a central role in the mathematical modelling, analysis and control of rheological processes. Day: July 6, 1995 Time: 09.30 K. Walters, University of Wales, United Kingdom: The role and importance of inverse problems in rheology J. Honerkamp, Universit{\"a}t Freiburg, Germany; C. Friedrich, Universit{\"a}t Freiburg, Germany; J. Weese, Philips GmbH Hamburg, Germany: Inverse problems in the recovery of material properties of synthetic polymers A.R. Davies, University of Wales, United Kingdom: Dual inverse problems in the determination of shear viscosity D.W. Mead, University of California at Santa Barbara, USA: Calculation of molecular weight distributions from linear viscoelastic properties of linear flexible polymers R.S. Anderssen, CSIRO, Australia: Parameter identification in rheology Y.L. Yeow, University of Melbourne, Australia; J.C. Munoz, University of Philippines, Quezon City: Regularization methods in capillary viscometry - Application of maximum entropy method *********************************************************************** Free and Moving Boundary Problems in Hydrodynamics Organizers: P.J. Zandbergen, University of Twente, The Netherlands O.H.G. Mahrenholtz, Technische Universit{\"a}t Hamburg-Harburg, Germany Summary: In many engineering problems in marine hydrodynamics and offshore technology one has to deal with free surface flows. This is especially related to the fluid flow of the upper ocean layer and to gravity driven water waves. Due to the time-dependent and nonlinear boundary conditions on the free surface the mathematical modeling and numerical treatment of these free and moving boundary problems are subject to restrictions. Large deformations and arbitrarily shaped domains require effective and reliable numerical methods. Some typical examples of 2-D and 3-D free surface flow problems are discussed, with emphasis on practical aspects and an outlook on challenges in free surface flow problems. Day: July 6, 1995 Time: 15.30 P.J. Zandbergen, University of Twente, The Netherlands; P.J.F. Berkvens, J. Broeze, E.F.G. van Daalen, P.C.A. de Haas, University of Twente, The Netherlands: Computation of hydrodynamic loads on a bottom-mounted surface piercing cylinder: application of a time-domain boundary element method for nonlinear gravity waves E. Fontaine, Bassin d'Essais des Car{\`e}nes, France; R. Cointe, Bassin dssais des Car{\`e}nes, France: Non-linear high Froude number slender ship theory H.C. Raven, Maritime Research Institute Netherlands, The Netherlands: A raised-panel method for the solution of the nonlinear ship wave-resistance problem T. Vinje, Norwegian Contractors, Norway: Ringing - a hydrodynamic challenge to designers of offshore structures J. Skourup, Danish Hydraulic Institute, Denmark; H.A. Sch{\"a}ffer, Danish Hydraulic Institute, Horsholm, Denmark: Active absorption in boundary element modelling of nonlinear water waves C. Haack, Technische Universit{\"a}t Hamburg-Harburg, Germany: Multiple connected floating structures in free surface flow R.W. Yeung, University of California at Berkeley, USA: Three-dimensional interaction surface waves with a cylindrical structure in a viscous fluid *********************************************************************** Mathematics in Physiology Organizer: P. Lory, Universit{\"a}t Regensburg, Germany Summary: The last years have seen a growing interest in mathematical approaches to the understanding of physiological functions. The increasing interest in mathematics arises naturally because of the need to provide an integrated explanation of experimental data. Purely heuristic interpretations can be tested only by mathematical models. Consequently, the way towards a complete understanding of physiological phenomena is often characterized by an iterative process between experiment and mathematical simulation. The minisymposium presents a selection of recent work in this spirit. A group of contributions focuses on renal function, including the tubuloglomerular feedback pathway, the countercurrent system and acid-base balance. Other speakers present results on tumour angiogenesis, in vivo NMR spectroscopy and biofluid dynamics in the context of swimming microorganisms and of blood flow in large arteries. Day: July 3, 1995 Time: 10.00 H.E. Layton, Duke University, USA; E.B. Pitman, State University of New York at Buffalo, USA; L.C. Moore, State University of New York at Stony Brook, USA: Dynamics of the tubuloglomerular feedback pathway K.S. Jensen, Dansk Bioprotein A/S, Denmark: Complex dynamical phenomena in a system of coupled nephrons E.B. Pitman, State University of New York at Buffalo, USA; H.E. Layton, Duke University, USA: A dynamic numerical method for models of renal tubules R.P. Tewarson, State University of NewYork at Stony Brook, USA: Efficient computational algorithms for kidney modeling R. Mejia, National Institutes of Health, USA; M.A. Knepper, National Institutes of Heath, USA: Examples of the use of mathematical models in renal physiology: urine concentration to acid-base balance Day: July 4, 1995 Time: 15.30 X.-Y. Luo, University of Leeds, United Kingdom; T.J. Pedley, University of Leeds, U.K.: Flow and instability in collapsible tubes G. Rappitsch, Technische Universit{\"a}t Graz, Austria; K. Perktold, Technische Universit{\"a}t Graz, Austria: Numerical simulation of pulsatile convective diffusion in large arteries M.A.J. Chaplain, University of Bath, United Kingdom: The mathematical modelling of tumour angiogenesis and tumour invasion M.v. Kienlin, Universit{\"a}t W{\"u}rzburg, Germany: Localized nuclear magnetic resonance NMR spectroscopy of arbitrarily shaped volumes in biomedical research *********************************************************************** Mathematics in the Textile and Clothing Industry Organizer: H. Ockendon, Oxford University, United Kingdom Summary: The textile and clothing industries offer the opportunity for a wide range of mathematical applications. This minisymposium will include work on models of polymer extrusion, fibre formation and stretching and spin bath dynamics. The modelling of liquid polymer crystals in fabric manufacture is a more recent development and even more tentative is the idea of using neural nets to model yarns. There are also a number of optimisation problems in the clothing industry including 'lay planning' for the optimal use of fabric. Day: July 7, 1995 Time: 09.30 J. Greenberg, University of Maryland at Baltimore County, USA; Y. Demay, Universit{\'e} de Nice, France: A simple model of the melt fracture instability Y. Demay, Universit{\'e} de Nice, France: Stability of industrial fibre and film processing P. Wolfenden, O.C.I.A.M., Courtaulds, Coventry, United Kingdom: The uses of mathematics in the textile industry R. Rennell, Shirley House, United Kingdom; P.J. Maris, British Textile Technology Group, UK: New applications of mathematics to yarn formation H. Ockendon, Oxford University, United Kingdom: Problems in fibre processing *********************************************************************** Discretely Supported Continuous Structures under Moving Loads Organizers: K. Knothe, Technische Universit{\"a}t Berlin, Germany R. Bogacz, Polish Academy of Sciences, Warsaw, Poland Summary: Due to new transportation systems (high speed railway vehicles and magnetic elevated vehicles) the subject of the minisymposium has drawn increasing attention in the last years. \smallskip Time domain solutions based on FE approximations, though very time consuming, are best suited for systems with nonlinearities and irregularities. On the other hand, efficient frequency domain solutions and wave propagation solutions have been developed, whereby it is difficult to include nonlinearities or irregularities or even arbitrary transient loads. \smallskip The aim of the minisymposium is mainly to define the range of validity of the various solution procedures and possibly collate the results of some benchmark-tests. Day: July 3, 1995 Time: 15.30 A. Igeland, Chalmers University of Technology G{\"o}teborg, Sweden: Time domain solution of the dynamic interaction between railroad structures and moving loads J.J. Kalker, Delft University of Technology, The Netherlands: A method of mixed time and frequency domains for discretely supported rails subject to transient loads T. Krzyzynski, Polish Academy of Sciences Warszawa, Poland; K. Popp, Universit{\"a}t Hannover, Germany: On the travelling wave approach for discrete-continuous structures under moving loads S. M{\"u}ller, Technische Universit{\"a}t Berlin, Germany; K. Knothe, TU Berlin, Germany: Response of periodically supported tracks to moving railway vehicles - a generalization of Jezequel's solution M. Meywerk, Technische Universit{\"a}t Braunschweig, Germany; E. Brommundt, TU Braunschweig, Germany: Floquet stability analysis of an oscillator moving along a cyclic periodically suspended string Day: July 5, 1995 Time: 09.30 W. Zhai, Southwest Jiatong University Chengdu, P. R. of China: Application of new fast time stepping methods to discretely supported tracks under moving vehicles I. Zobory, Technical University of Budapest, Hungary; V. Zoller, Technical University of Budapest, Hungary; Z. Z{\'a}bori, Technical University of Budapest, Hungary: Time domain analysis of a railway vehicle running on a discretely supported continuous rail model at a constant velocity I.S. Vorobev, Ukrainian Academy of Sciences at Charkov, Ukraine: Bending-torsional vibrations of structures under travelling forces Z. Cai, Royal Military College of Canada, Canada; G.P. Raymond, Queen's University, Kingston, Ontario, Canada: Dynamic response of railway track represented by generalized support spring and exact dynamic stiffness of rail beam H. Ilias, Technische Universit{\"a}t Berlin, Germany; B. Ripke, Technische Universit{\"a}t Berlin, Germany: Nonlinear vehicle-track interaction in consideration of vertical and lateral dynamics *********************************************************************** Elliptic Equations with Nonlinear Boundary Conditions Organizers: E. Wegert, Technische Universit{\"a}t - Bergakademie Freiberg, Germany L. von Wolfersdorf, Technische Universit{\"a}t - Bergakademie Freiberg, Germany Summary: The minisymposium is devoted to boundary value problems for (linear) elliptic equations with nonlinear boundary conditions as well as to related boundary integral equations (especially singular integral equations).\\ Besides its theoretical interest this class of problems is also relevant in several fields of applied mathematics, such as potential flow through and around porous media, vibration and nonlinear waves, heat transfer, $H^{\infty}$-optimization and free boundary value problems. Special emphasis is on problems which are 'strongly' nonlinear, i.e., which are not a small perturbation of linear problems. \smallskip The aim of the minisymposium is to present methods (functional analytic, complex, topological), theoretical results (existence, uniqueness and bifurcation of solutions), applications and numerical algorithms. Day: July 7, 1995 Time: 15.30 M.A. Efendiev, Freie Universit{\"a}t Berlin, Germany: Homotopy invariant and noninvariant degree for pseudodifferential operators P. Junghanns, Technische Universit{\"a}t Chemnitz-Zwickau, Germany: On the numerical solution of nonlinear singular integral equations M. Plum, Technische Universit{\"a}t Clausthal, Germany: Numerical enclosures for solutions of nonlinear boundary value problems E. Wegert, TU Bergakademie Freiberg, Germany: Nonlinear boundary value problems for holomorphic functions *********************************************************************** Application of Interior Point Methods to Large Scale Optimization Problems Organizers: F. Potra, The University of Iowa, USA H. G. Bock, Universit{\"a}t Heidelberg, Germany Summary: The field of interior point methods has been the most active research area in mathematical programming during the past decade. Over 1500 paper have been written and a number of excellent codes have been developed. The goal of the minisymposium is to give an overview of the state of the art of interior point based software and of the most significant large scale optimization problems solved by such codes. Day: July 3, 1995 Time: 15.30 P.T. Boggs, National Institute of Standards and Techn., USA; J.W. Tolle, A. Kearsley: Solving large scale nonlinear programming problems with an SQP algorithm based on an interior-point QP solver K.O. Kortanek, The University of Iowa, USA; X. Xu, Y.Ye, University of Iowa, Iowa City, USA: An interior-point algorithm for solving primal and dual geometric programs F.A. Potra, The University of Iowa, USA: Interior point methods for image reconstruction problems M.G.C. Resende, AT{\&}T Bell Laboratories, USA: Interior-point methods for solving large scale network flow optimization R.J. Vanderbei, Princeton University, USA: Interior point methods for large scale convex optimization M.C. Steinbach, Universit{\"a}t Heidelberg, Germany: Structured interior point SQP methods in optimal control *********************************************************************** Modeling, Analyses and Computations for Superconductivity Organizer: M.D. Gunzburger, Virginia Tech, Blacksburg, USA Summary: Superconductivity has enjoyed renewed interest due to the recent discovery of materials that retain superconducting properties at temperatures above the boiling point of nitrogen. The mathematical modeling and analyses and the computational simulation of superconducting phenomena has been part of this renewal. In this minisymposium, the speakers will present recent results concerning all of these aspects of mathematical research into superconductivity. New, macro-scale models that promise to be of great utility in the design of superconducting devices will be discussed, as well as refinements and analyses for existing mezo-scale models that can be used for studying basic phenomena. Computational algorithms and the results of numerical simulations using both types of models will also play an important role in some of the talks. Day: July 3, 1995 Time: 15.30 S.J. Chapman, University of Oxford, United Kingdom: Models for vortices in type-II superconductors Q. Du, Michigan State University, USA: Analysis and simulation of vortex motion and vortex pinning J. Peterson, Virginia Tech., USA: Computational simulation of time-dependent superconducting phenomena L.S. Hou, York University, Canada; M.D. Gunzburger, Virginia Tech, Blacksburg, USA; S.S. Ravindran, North Carolina State University, Raleigh, USA: Optimal control for the Ginzburg-Landau equations D. Phillips, Purdue University, USA: Qualitative features of solutions for superconductivity Q. Tang, University Sussex, United Kingdom; S. Wang, Indiana University at Bloomington, USA: Long time behaviour of a Ginzburg-Landau system describing superconductivity *********************************************************************** Modern Applications of Complex Analysis Organizer: Y.E. Hohlov, Steklov Mathematical Institute, Moscow, Russia Summary: The minisimposium presents a survey of recent applications of complex analysis to the wide range of boundary value problems arising in applied and computational mathematics. It focuses on powerful tools of complex analytic and geometric methods such as Riemann-Hilbert boundary value problems and parametric representations of families of conformal mappings by means of the Loewner-Kufarev differential equations. New results on classical solvability of different evolutionary problems, construction of explicit solutions, geometrical behaviour of solutions based on the methods will be highlighted. Effective approaches for construction of numerical conformal mappings will be presented. Day: July 3, 1995 Time: 10.00 H. Begehr, Freie Universit{\"a}t Berlin, Germany: Complex analytic methods for partial differential equations V.Y. Gutlyanskii, Ukrainian Academy of Sciences, Ukraine; Yu.E. Hohlov, Steklov Mathematical Institute, Moscow, Russia: New applications of dynamical complex analysis to numerical conformal mappings Y.E. Hohlov, Steklov Mathematical Institute, Russia: Complex analysis and free boundary problems: trends and developments S.D. Howison, University of Oxford, United Kingdom: Complex variables in codimension-two free boundary problems Day: July 6, 1995 Time: 09.30 J.R. King, University of Nottingham, United Kingdom: Corners development in moving boundary problems S. Tanveer, The Ohio State University, USA: Time-evolving bubbles in 2D Stokes flow A.I. Aptekarev, Keldysh Institute of Applied Mathematics, Russia; L.R. Volevich, Keldysh Institute, Moscow, Russia; E.P. Kazandjan, Keldysh Institute, Moscow, Russia: A numerical method of conformal mapping of a multi-connected domain *********************************************************************** Numerical Methods in Constrained Multibody Dynamics Organizers: U.M. Ascher, University of British Columbia, Vancouver, Canada R. von Schwerin, Universit{\"a}t Heidelberg, Germany Summary: Numerical methods for constrained multibody dynamics are important tools in computer aided engineering for complex applications like vehicle system dynamics or robotics. The models usually consist of rigid bodies connected by joints with incorporated mechatronical parts, leading in general to linearly-implicit DAEs with additional dynamics and discontinuities caused by underlying discrete effects from, e.g., contact or controllers. Conventional industrial approaches ignore the specific topological structures of multibody systems and interactions between discrete and continuous model components. Thus, current research focusses on discrete-continuous simulation and sensitivity analysis with emphasis on exploitation of inherent structures, all of which the speakers will discuss. Day: July 4, 1995 Time: 15.30 T. Alishenas, Kungl Tekniska H{\"o}gskolan, Sweden: Velocity stabilization of mechanical index-3 DAE's M. Arnold, Universit{\"a}t Rostock, Germany: Numerical problems in the dynamical simulation of wheel-rail systems R. von Schwerin, Universit{\"a}t Heidelberg, Germany; M. Winckler, Universit{\"a}t Heidelberg, Germany: Some aspects of sensitivity analysis in vehicle system dynamics B. Simeon, Technische Hochschule Darmstadt, Germany; P. Rentrop, TH Darmstadt, Germany: A projection oriented formulation of multibody system dynamics with application to a wheel suspension model A. Sj{\"o}, Lund University, Sweden: Numerical aspects in contact mechanics simulation - a case study on a mathematical model for a roller bearing R.J. Spiteri, University of British Columbia, Canada; U.M. Ascher, University of British Columbia, Vancouver, Canada; D. Pai, University of British Columbia, USA: Numerical solution of differential-algebraic ineqalities arising in constraint multibody systems J. Yen, Army High Performance Com. and Res. Center, USA; L. Petzold, University of Minnesota, USA: Numerical methods for real-time vehicle simulation *********************************************************************** Towards the Understanding of Chemical Pattern Formation Organizer: M. Mimura, University of Tokyo, Japan Summary: Pattern formation in chemical systems often appears in a variety of contexts, with implications not only for chemistry but also biology and other fields as well. Reaction and diffusion, sometimes with advection, generate a remarkable variety of patterns, waves and dynamics of interfaces. \smallskip Towards the understanding of chemical pattern formation, the aim of the organizer is to discuss the current research on spatio and/or temporal pattern formation and to find out to what extent theories and experiments in this field move along the same lines, in other words, how far experimental results can be explained by existing mathematical models. In order to account for the interdisciplinary nature of this theme, 6 speakers are invited from such diverse fields as mathematics, scientific computing, chemistry, biochemistry and biology. Each topic could be discussed in reasonable depth. The emphasis is on investigating important questions from chemical and biological pattern formation view points, though this may often lead to new and exciting applied mathematics. Day: July 4, 1995 Time: 09.30 R. Kobayashi, Ryukoku University, Japan; T. Ochta, Y. Hayase, Ochanomizu University, Tokyo, Japan: Self-organized pulse generator in a reaction diffusion system Y. Nishiura, Hiroshima University, Japan: Nonlocal effects on phase separation dynamics W. Alt, Universit{\"a}t Bonn, Germany: Modelling of protrusion patterns at the cell periphery T. H{\"o}fer, University of Oxford, United Kingdom; P.K. Maini, University of Oxford, UK: Cellular pattern formation during dictyostelium aggregation P. De Kepper, Domaine Universitaire Bordeaux I, France: Stationary Turing patterns and travelling waves in a chemical system S.C. M{\"u}ller, MPI f{\"u}r molekulare Physiologie, Germany: Excitation waves under external control *********************************************************************** Numerical Optimization in PDE -- Case Studies in Applications Organizers: V. Schulz, Universit{\"a}t Heidelberg, Germany F. Tr{\"o}ltzsch, Technische Universit{\"a}t Chemnitz-Zwickau, Germany Summary: Many physical processes, like heat transfer in metals or the flow of gas in turbine machines are modeled by PDE's. \smallskip Although efficient numerical simulation methods are available for some classes of PDE's, there is a lack of efficient optimization methods, especially for practical problems from industry. Due to the high number of variables in PDE-discretizations, standard optimization problem solvers fail. In contrast, the exploitation of the special structures of the PDE-model leads to efficient and reliable optimization methods. \smallskip In this minisymposium theoretical results and practical experiences following this approach are discussed in case studies of application problems. Day: July 4, 1995 Time: 15.30 M. Heinkenschloss, Virginia Polytechnic Institute, USA: Numerical solution of heat conduction problems with constraints G. Landl, Johannes-Kepler-Universit{\"a}t Linz, Austria; W. Grever, Johannes Kepler Universit{\"a}t, Linz, Austria: Optimal control of cooling in the continuous casting and hot rolling of steel with varying production speed R. Lezius, Technische Universit{\"a}t Chemnitz-Zwickau, Germany; F. Tr{\"o}ltzsch, Technische Universit{\"a}t Chemnitz-Zwickau, Germany: Numerical treatment of the controlled cooling of steel profiles J. Roche, Universit{\'e} de Nancy I, France; J. Sokolowski, University of Nancy I, France: Numerical Newton's like algorithms in shape identification for elliptic systems P. Neittaanm{\"a}ki, University of Jyv{\"a}skyl{\"a}, Finland; T. R{\"a}is{\"a}nen, University of Jyv{\"a}skyl{\"a}, Finland; D. Tiba, Romanian Academy, Bucharest, Romania: On the approximation of some ill-posed problems V.H. Schulz, Universit{\"a}t Heidelberg, Germany: Numerical optimization of the cross-sectional shape of turbine blades *********************************************************************** Industrial Application of Optimization Methods Organizers: C.T. Kelley, North Carolina State University, Raleigh, USA E.W. Sachs, Universit{\"a}t Trier, Germany Summary: The speakers of this minisymposium address the successful use of optimization algorithms in industrial applications. Areas of applications are in design optimization, control of heat conduction problems, control of fluid flow problems, and parameter identification problems. Many of these problems cannot be solved by standard optimization algorithms because of the size of the problem or the degree of nonlinearity involved. Therefore special algorithms have to be developed to successfully solve these problems. In a discussion of these methods the speakers will address issues like proper globalization strategies, useful implementations of iterative methods for the subproblems and use of sparsity to overcome the size of some problems. It is also illustrated how the numerical results at an early stage can lead to fruitful feedback in the modelling phase. Other applications require a careful analysis of local convergence properties of the algorithms. Day: July 6, 1995 Time: 09.30 H. J{\"a}ger, Universit{\"a}t Trier, Germany; E. Sachs, Universit{\"a}t Trier, Germany: Applications of optimization methods in optimal control of industrial furnaces C.T. Kelley, North Carolina State University, USA; J.W. David, Mech. {\&} Aero. Eng., Raleigh, USA: Implicit filtering and applications K. Kunisch, Technische Universit{\"a}t Berlin, Germany: Applications in heat conduction processes H.F. Walker, Utah State University, USA: Applications of Newton-Krylov methods *********************************************************************** Probabilistic Numerical Methods for PDE's Organizer: D. Talay, INRIA, Sophia Antipolis, France Summary: The objective of this minisymposium is to present recent numerical and theoretical advances concerning Monte Carlo methods and stochastic particles methods for deterministic PDE's, with particular emphasis on nonlinear PDE's, such as Burger's equation, Navier-Stokes equation, quasilinear parabolic equations, PDE's coming from genetics models. \smallskip Accurate estimates on the convergence rates, variance reduction techniques and illustrative numerical experiments will be discussed. Day: July 6, 1995 Time: 15.30 N. Newton, University of Essex, United Kingdom: Variance reduced Monte Carlo methods for PDEs C. Graham, Ecole Polytechnique, France; S. Meleard, Universit{\'e} Paris, France: to be announced M. Bossy, INRIA, France: A stochastic particle method for one-dimensional nonlinear PDE's of Burgers type D. Chevance, INRIA, France: Discretisation of Pardoux-Peng's backward stochastic differential equations *********************************************************************** Extended Surface Heat Transfer Organizer: G.E. Tupholme, University of Bradford, United Kingdom Summary: Current research and design of extended surfaces, i.e. fins, which are used in heat exchangers throughout the process industries, still often consider mathematically decoupled descriptions of the assembly. \smallskip Effects such as non-uniform heat transfer coefficients, property variations, radiation, optimization of fin mass, and transient situations are appearing in mathematical and engineering journals. It is timely to review this area of heat transfer and to consider the complete multi-dimensional coupled system. \smallskip This minisymposium will provide a unique platform to launch the impetus to accept this challenge and layout the necessary studies that should be undertaken. Day: July 5, 1995 Time: 15.30 P.J. Heggs, UMIST, England: Do the current engineering aspects of extended heat transfer comply with the mathematical representation? B. Sunden, Lund Institute of Technology, Sweden: Two-dimensional coupled conduction and forced convection of a rectangular fin in a confined space H. Kalman, Ben-Gurion University of Negev, Israel: Transient temperature response and performance of fins D.B. Ingham, Leeds University, England; P.J. Heggs, UMIST, Manchester, UK: Transient modelling of extended surface heat transfer P.J. Heggs, UMIST, England; D.B. Ingham, University of Leeds, UK: Two-dimensional considerations of steady state heat flow through extended surface assemblies G.E. Tupholme, University of Bradford, United Kingdom; A.S. Wood, University of Bradford, UK: Mathematics of performance indicators *********************************************************************** Thermic Contact with Friction Organizer: H. Gl{\"a}ser, Technische Universit{\"a}t - Bergakademie Freiberg, Germany Summary: Subjects of the minisymposium represent following problems: \item{--} thermic contact analysis and accesses \item{--} constitutive contact problems, question of their modelling \item{--} computer field programs and applications \smallskip This complex field of topics includes scientific interest and novel questions and possibilities of engineering applications concerning researchers as well as users dealing with the fields of activity like mathematics (functional analysis, numerical methods etc.), mechanic, thermodynamic, material research, metal forming processes etc. \smallskip A priori, geometric and physically nonlinear well as nonconservative problems are in the centre of discussion. \smallskip Possibilities to their modelling, satisfying the physical, mathematical and engineering restrictions, will be discussed. For example, the formulation of the kind of thermo--mechanical coupling and the heat transfer across the contact surfaces, changing their micro-- and macro--shape by deformation, should be an interesting question. The evaluation of the classical friction laws and others contact laws, considering the flexibility of the contact surfaces, including the identification of the constitutive parameters by experiments or numerical simulation constitute a further aspect for discussion. Numerical methods are of special interest solving nondifferantiable functions resulting from the general formulation of mechanical unilateral restrictions in the frame work of the convex analysis. \smallskip The development of effective solutions plays an important role in the realisation of contact problems by field programs (FEM, BEM etc.). This solution algorithm should depend on the kind of therm--mechanical coupling and questions of the discretization. First examples of solution showed the output capacity of the changed and proved therm--mechanic contact model. Day: July 7, 1995 Time: 15.30 P. Wriggers, Technische Hochschule Darmstadt, Germany: On thermomechanical contact including Coulomb friction R. Hupach, TU Bergakademie Freiberg, Germany: Untersuchung der sich als Folge der Dissipation beim plastischen Kontakt entwickelnden Temperaturfelder bei Anlaufvorg{\"a}ngen O.H.G. Mahrenholtz, Technische Universit{\"a}t Hamburg-Harburg, Germany; Dipl.-Ing. Brozozwski: Reibungsanalyse beim Walzen mit Breitung C. Bederna, Universit{\"a}t Hannover, Germany; E. Doege, Universit{\"a}t Hannover, Germany: Analyse der lokalen Kontaktparameter in der Warmmassivumformung *********************************************************************** Nucleation, Memory Effects and Crystal Growth in Polymers Organizer: A. Fasano, Universit{\`a} di Firenze, Italy Summary: Solidification of molten polymers is an extremely complex process leading to the formation of a partially amorphous product with crystalline aggregates having a rather complicated internal structure. Such a process can be influenced by many factors, including the previous thermal history of the sample. There is also a sharp experimental evidence of memory effects, since in a melting-solidification cycle crystals tend to reappear in the same location. \smallskip The purpose of this minisymposium is to present some advances in the modelling of nucleation and growth of crystals in molten polymers, illustrating both experimental procedures and mathematical aspects with particular emphasis on nucleation and memory effects. Numerical calculations and computer simulations will also be presented. Day: July 7, 1995 Time: 09.30 G.C. Alfonso, University of Genova, Italy: Melt history effects on polymer crystallization kinetics. Experimental evidence and theoretical considerations D. Andreucci, Universit{\`a} di Roma, Italy; A. Fasano, Universit{\`a} di Firenze, Italy; M. Primicerio, Universit{\`a} di Firenze, Italy; R. Ricci, Universit{\`a} di Milano, Italy: Mathematical modelling of solidification of polymers G. Eder, Universit{\"a}t Linz, Austria: Crystallization kinetic equations incorporating surface and bulk nucleation processes H. Janeschitz-Kriegl, University of Linz, Austria; G. Eder, University of Linz, Austria: Model experiments in polymer crystallization posing novel mathematical problems *********************************************************************** Numerical Treatment of Inverse Problems Organizers: M. Hanke, Universit{\"a}t Karlsruhe, Germany A. Kirsch, Universit{\"a}t Erlangen - N{\"u}rnberg, Germany Summary: Many problems in the applied sciences lead to inverse problems. They are illposed in the sense that the solutions do not depend continuously on the data. Therefore, naive approximations give useless results due to measurement errors in the given data. Various approaches have been developed to regularize, i.e., stabilize such problems. \smallskip It is the aim of this minisymposium to report on very recent developments of specific algorithms for selected linear and nonlinear inverse problems including the theoretical analysis of their performance. Day: July 3, 1995 Time: 15.30 C.R. Vogel, Montana State University, USA: Fast total variation-based methods in image reconstruction P. Maa{\ss}, Universit{\"a}t Potsdam, Germany: Wavelet projection methods for inverse problems J.G. Nagy, Southern Methodist University, USA: Aspects of preconditioning conjugate gradient iterations for discrete ill-posed problems F. W{\"u}bbeling, Westf{\"a}lische Wilhelms-Universit{\"a}t M{\"u}nster, Germany: Finite difference schemes for solving the inverse scattering problem B. Hoffmann, Universit{\"a}t G{\"o}ttingen, Germany: Newton methods in impedance tomography M. Hanke, Universit{\"a}t Karlsruhe, Germany: A conjugate gradient type approach for inverse heat equations W.W. Symes, Rice University, USA: Regularization and duality in inverse problems for wave velocity *********************************************************************** Assessing the Accuracy and Reliability of Numerical Solutions Organizer: S. Hammarling, NAG Ltd, Oxford, United Kingdom Summary: This minisymposium is concerned with a central problem in scientific computing, that of assessing the accuracy of solutions obtained by numerical methods on computers using floating point arithmetic. The speakers in this minisymposium will review different methods and approaches to solving this problem. \smallskip The first speaker will look at an approach called Qualitative Computing, the second speaker will look at the state of the art in rounding error analysis, the third speaker will offer a critique of the probabilistic approach to estimating roundoff error, and the fourth speaker will review the methods of interval analysis. Day: July 5, 1995 Time: 09.30 F. Chaitin-Chatelin, Universit{\'e} Paris IX Dauphine and CERFAS, France: Qualitative computing and the robustness of numerical methods to high nonnormality J. Demmel, University of California at Berkeley, USA: Modern error analysis in the LAPACK linear algebra library W. Kahan, University of California, USA: The improbability of probabilistic estimates for roundoff G. Mayer, Universit{\"a}t Rostock, Germany: Interval methods for the assessment of numerical accuracy *********************************************************************** Mathematical Modelling of Industrial Fluid Flows - Aims and Limitations Organizers: W. Schneider, Technische Universit{\"a}t Wien, Austria F. Bark, KTH Stockholm, Sweden Summary: Computing fluid flows has become common practice in industry. The aims of the computations vary, however. Industries with a tradition in applying fluid mechanical tools, e.g. for designing aircrafts, aim at accurate predictions of flows that are basically understood. On the other hand, there are more recent fields of applications, e.g. casting of metals, where the main purpose of the modelling is to analyze the process. \smallskip Though various software packages are now available, satisfying industrial needs is, in general, not an easy task. Often an extensive analysis is required to keep computational efforts within certain limits, make a very complex flow problem tractable, or deal with time and length scales of different orders of magnitude. Further difficulties may arise from instabilities and non-uniqueness, or from downstream boundary conditions. There is also the problem of applying turbulence models to types of flow for which the models have been neither designed nor tested. \smallskip Based on examples from various branches of industry, it is intended to give a survey on both the aims and the limitations of predicting fluid flows of industrial relevance. Day: July 6, 1995 Time: 15.30 P. Bedrikovetsky, Cenpes, Ciade Universitaria, Brazil; K.T. Potsch, OMV-ID-Lap, Wien: Mathematical models for oil recovery H. Keck, Sulzer Hydro, Escher Wyss A.-G., Switzerland: The unsteady, turbulent 3D flow in a pump. Deviations between numerical and experimental results as a function of the boundary conditions K. M{\"o}rwald, VOEST- ALPINE Industrieanlagen GmbH, Austria: Free surface flows and phase transition in continuous casting of steel H. Sobieczky, DLR, Germany: Mathematical models for generic aerodynamics in four dimensions J. Unger, Technische Hochschule Darmstadt, Germany: Mixed diffusion and free convection problems W. Schneider, Technische Universit{\"a}t Wien, Austria: Introduction to the modelling of industrial fluid flows *********************************************************************** Structural Optimization: Multiple Objectives and Random Data Organizers: H. Eschenauer, Universit{\"a}t Siegen, Germany K. Marti, Universit{\"a}t der Bundeswehr M{\"u}nchen, Germany Summary: Having the very efficient and nearly universally accepted tools of finite element analysis methods to determine the structural behaviour, one of the remaining important goals of engineering activities is to optimize technical designs, structural assemblies and structural components. In order to increase the reliability and efficiency of the optimized design, multiple objectives and random data must be taken into account. Hence, optimization techniques based on stochastic optimization, vector optimization and parallel concepts are considered for design improvements; furthermore, applications to the optimization of advanced materials, optimal design of robots/flexible robots and structural optimization of redundant space frame structure are given. Day: July 5, 1995 Time: 09.30 R. Rackwitz, Technische Universit{\"a}t M{\"u}nchen, Germany: Structural reliability and optimization M. Grauer, Universit{\"a}t Siegen, Germany; H. Eschenauer, Universit{\"a}t Siegen, Germany: Multiobjective decision-making using parallel concepts F. Pfeiffer, Technische Universit{\"a}t M{\"u}nchen, Germany: Design optimization of a 6-legged walking machine P. Pedersen, Technical University of Denmark, Denmark: Material optimizations - an engineering view J. Reinhart, Universit{\"a}t der Bundeswehr, Germany: Discussion of stochastic effects on optimal fibre reinforced structures C. Haubach, Universit{\"a}t der Bundeswehr M{\"u}nchen, Germany: Flexible robots - an example of stochastic structural optimization *********************************************************************** Stochastic Optimization Organizers: K. Marti, Universit{\"a}t der Bundeswehr M{\"u}nchen, Germany W. R{\"o}misch, Humboldt Universit{\"a}t Berlin, Germany Summary: Applied optimization problems in industrial engineering and economics often contain data (e.g., parameters of the material, costs, production factors, demand, etc.) which are not known, but have to be considered as random variables with certain probability distribution. Illustrative examples are: the optimal design of mechanical structures, the optimization of flexible robots, financial planning, electricity planning under uncertainty etc. The two main approaches in stochastic optimization are: \item{(i)} the introduction of expected optimal costs for the compensation of (stochastic) constraints and \item{(ii)} to satisfy constraints involving stochasticity with certain reliability. \smallskip Both approaches lead to nonlinear programs including multidimensional integrals over (often) nonsmooth (implicitly defined) integrands in the objective or constraints. \smallskip The idea of the minisymposia consists in presenting recent progress in modelling, theory, numerical methods, software and applications of stochastic optimization. Day: July 3, 1995 Time: 10.00 K. Frauendorfer, University of St. Gallen, Switzerland: Stochastic multistage programming in financial decision making F. Fantauzzi, ITALTEL and Milano University, Italy; A.A. Gaivoronski, ITALTEL and Milano University, Italy: Stochastic gradient decomposition with application to the planning of high speed telecommunications networks K. Marti, Universit{\"a}t der Bundeswehr M{\"u}nchen, Germany: Differentiation of probability functions by using probability density estimators J. Mayer, Universit{\"a}t Z{\"u}rich, Switzerland; P. Kall, Universit{\"a}t Z{\"u}rich, Switzerland: Modeling support to stochastic linear programming: the solver interface S. Qu, Universit{\"a}t der Bundeswehr M{\"u}nchen, Germany: An efficient open-loop strategy for robotic systems under stochastic uncertainty W. R{\"o}misch, Humboldt-Universit{\"a}t Berlin, Germany: Power dispatch under uncertainty via stochastic programming R. Schultz, Konrad-Zuse-Zentrum f{\"u}r Informationstechnik Berlin, Germany: An algorithm for two-stage stochastic integer programs *********************************************************************** Mechanical Systems with Symmetry: Dynamics and Bifurcation Organizers: J.E. Marsden, University of California at Berkeley, USA T. Ratiu, University of California at Santa Cruz, USA J. Scheurle, Universit{\"a}t Hamburg, Germany Summary: The use of symmetry and bifurcation in the analysis of mechanical dynamical systems continues to be an active area of research. The use of symmetry, for example by means of reduction or normal form procedures, has had much success in the area of stability and bifurcation of relative equilibria and this will be one of the themes of the minisymposium. Relative equilibria are solutions whose dynamic orbits in phase space are contained in orbits of the symmetry group of the system. Examples of current research topics are resonance bifurcations and instabilities induced by double bracket dissipation. These ideas are also currently being merged with singular reduction, and are being applied to control theory in topics such as locomotion generation in nonholonomic mechanical systems that occur in, for example, robotics. Also, symmetries of invariant sets will be an issue of the minisymposium. Day: July 4, 1995 Time: 09.30 A. Bloch, The University of Michigan, USA: The geometry of nonholonomic mechanical systems with symmetry M. Golubitsky, University of Houston, USA; J.E. Marsden, University of California at Berkeley, USA; I. Stewart, University of Warwick, UK; M. Dellnitz, Universit{\"a}t Hamburg, Germany: Bifurcation of periodic solutions in Hamiltonian systems T.S. Ratiu, University of California at Santa Cruz, USA; A.Bloch, University of Michigan, Ann Arbor, USA; P.S. Krishnaprasad, University of Maryland, College Park, USA; J.E. Marsden, U.of California, USA: Dissipation induced instabilities J.K. Scheurle, Universit{\"a}t Hamburg, Germany; J.E. Marsden, University of California, Berkeley, USA: Pattern evocation in mechanical systems S.R. Wiggins, California Institute of Technology, USA: N-pulse homoclinic orbits in perturbations of resonant Hamiltonian systems: the energy-phase method *********************************************************************** Nonlinear Phenomena in Plates and Shells Organizer: R. Schmidt, Bergische Universit{\"a}t-GH Wuppertal, Germany Summary: The development of structural theories and associated computational methods is a field of permanent intense research because of the need of structural elements to work to the limit of their load carrying capacity, the increasing use of advanced structural materials and the advances in modelling material behaviour even far from the linear regime. The present Minisymposium continues those held at the First and Second ICIAM and deals with recent advances concerning modelling and computational analysis of geometrically and/or physically nonlinear plate and shell structures. The problems considered include the development of new nonlinear theories, FE-methods, stability and post-buckling analysis, large amplitude vibrations, composite laminated structures and structures with inelastic material behaviour. Day: July 5, 1995 Time: 15.30 P. Klosowski, Politechnika Gdansk, Poland; J.N. Reddy, Texas A{\&}M, USA; R. Schmidt, Bergische Universit{\"a}t - GH Wuppertal, Germany: Nonlinear transient analysis of composite laminates undergoing moderate rotations F.-A. Emmerling, Universit{\"a}t der Bundeswehr M{\"u}nchen, Germany: Nichtlineare Biegung von Torusschalen J.-B. Tritsch, EUDIL, Laboratoire de M{\'e}canique de Lille, France; D. Weichert, Laboratoire de M{\'e}canique de Lille, Villeneuve, France: Behaviour of thin-walled structures under variable thermo-mechanical loads J. Makowski, Ruhr-Universit{\"a}t Bochum, Germany; J. Chroscielewski, Ruhr-Universit{\"a}t Bochum, Germany; H. Stumpf, Ruhr-Universit{\"a}t Bochum, Germany: Irregular shell structures -- analytical and computational problems *********************************************************************** Mathematics and the World of Patents Organizer: A.H.P.v.d. Burgh, Delft University of Technology, The Netherlands Summary: As may be known, mathematical principles or methods can not be patented. However in a society with an increasing knowledge revealing problems of increasing complexity the role of mathematics in patent development and applications can no longer be ignored.\\ In this minisymposium three speakers will throw light, from different points of view, on the role of mathematics in the field as mentioned.\\ The first speaker will elucidate the experiences of the European Patent Office with mathematics in patent applications. Moreover, a vision on the most desirable role which mathematics should play in patent applications in the near future will be discussed.\\ The second speaker will report on his experiences in using mathematics for the development of special patents.\\ The last speaker will indicate how existing patents can be used as a source of inspiration for research at universities.\\ The final part of the symposium will be organized as a forum discussion open to those who have attended the minisymposium. Day: July 6, 1995 Time: 09.30 A.S. Holzwarth, European Patent Office, Germany: Patentability of inventions involving mathematical methods C.P.M.J. Baggen, Philips Research Laboratories, The Netherlands: On the use of mathematics by inventors A.H. van der Burgh, Delft University of Technology, The Netherlands: Patents as a source of inspiration for applied mathematicians *********************************************************************** Optimization with Validation Organizers: S.M. Rump, Technische Universit{\"a}t Hamburg-Harburg, Germany A. Frommer, Bergische Universit{\"a}t-GH Wuppertal, Germany Summary: The minisymposium will deal with numerical methods for optimization which deliver guaranteed bounds for the optimal value of the objective function and/or the minimizers. The methods can be applied to a large class of (potentially non-differentiable or discontinuous) optimization problems. \smallskip As opposed to 'classical' numerical optimization methods, these methods use information on the whole domain rather than on a finite number of isolated points. For certain classes of problems these algorithms are competitive (with respect to speed) to well-established non-validating algorithms. Day: July 7, 1995 Time: 09.30 S. Berner, Bergische Universit{\"a}t-GH Wuppertal, Germany: Parallel validated global optimization T. Csendes, Jozsef Attila University, Hungary; D. Ratz, University of Karlsruhe, Germany: Subdivision direction selection in interval methods for global optimization C. Jansson, Technische Universit{\"a}t Hamburg-Harburg, Germany: An expansion principle in validated global optimization D. Ratz, Universit{\"a}t (TH) Karlsruhe, Germany: Improved techniques for gap-treating and box-splitting in interval Newton Gauss-Seidel steps K. Madsen, The Technical University of Denmark, Denmark: Real versus interval methods for global optimization *********************************************************************** Parallel Algorithms for Dense and Banded Matrices Organizer: A.J. Frommer, Bergische Universit{\"a}t-GH Wuppertal, Germany Summary: Large linear systems and eigenvalue problems with dense or banded matrices arise in many applications. To solve them on a parallel computer one needs specifically adapted parallel algorithms. Most of the current parallel machines have processors with a memory hierarchy and use message passing. To obtain maximum performance one therefore needs an adequate distribution of the data on the processors and has to use block (BLAS3) operations on the individual processors. \smallskip The proposed minisymposium addresses these topics for the typical problems in numerical linear algebra. The contributions will concentrate on two major issues: The rearrangement of established serial algorithms in order to obtain good parallel efficiency and new parallel numerical methods for those problems, where serial algorithms cannot be parallelized efficiently, especially for banded matrices. Day: July 7, 1995 Time: 15.30 P. Arbenz, ETH Zentrum Z{\"u}rich, Switzerland; W. Gander, ETH-Z{\"u}rich, Switzerland: A survey of direct parallel methods for banded linear systems P. Fiebach, Bergische Universit{\"a}t-GH Wuppertal, Germany: Block algorithms for triangular linear systems S. Hofmann, Bergische Universit{\"a}t-GH Wuppertal, Germany: Efficient parallel algorithms for the singular value decomposition B. Lang, Bergische Universit{\"a}t, Germany: Reduction of banded matrices to bidiagonal form G. Mayer, Universit{\"a}t Rostock, Germany; G. Alefeld, Universit{\"a}t Karlsruhe, Germany; I. Lenhardt, Universit{\"a}t Karlsruhe, Germany: Multisplitting methods for banded matrices X. Sun, Duke University, USA: Orthogonal bandreduction and tridiagonalization techniques *********************************************************************** Equivariant Dynamical Systems Organizers: M. Dellnitz, Universit{\"a}t Hamburg, Germany R. Lauterbach, WIAS Berlin, Germany Summary: Dynamical Systems are models to describe the evolution of real world systems. Many such systems possess symmetry properties, either due to a symmetric experimental setup or to other geometrical features of the underlying problem. Important problems where the theory has been applied are the Taylor--Couette problem, coupled oscillators, buckling problems and reaction diffusion systems arising in chemical or biological applications. The typical behaviour of a dynamical system crucially depends on the type of underlying symmetry. Important mathematical tools have been developed to help understand this phenomenon: reduction methods combined with local bifurcation analysis, normal form theory, orbit space reductions and, more recently, a group theoretic description of complicated dynamical behavior, e.g. characterization of symmetry groups of attractors. Due to the presence of a group action equivariant dynamical systems become very complex. On the other hand the theoretical tools use properties of group actions to reduce this complexity. The minisymposium aims to provide an overview of recent developments in this field and to show, by way of examples, how these apply to interesting problems. Day: July 3, 1995 Time: 10.00 P. Chossat, Universit{\`e} de Nice Sophia-Antipolis, France: The convective dynamo problem from a symmetry-breaking bifurcation point of view M. Dellnitz, Universit{\"a}t Hamburg, Germany: Symmetry breaking bifurcations of chaotic attractors M.J. Field, University of Houston, USA: The structure of symmetric attractors S.v. Gils, University of Twente, The Netherlands; M. Krupa, Rijksuniversiteit Groningen, The Netherlands; V. Tchistiakov, University of Twente, The Netherlands: The dynamics of coupled Josephson junctions M. Golubitsky, University of Houston, USA: Symmetry and chaos: patterns on average R. Lauterbach, WIAS Berlin, Germany: Forced symmetry breaking *********************************************************************** Interface Dynamics: Singular Perturbation and Geometric Evolutions Organizer: L. Bronsard, McMaster University, Hamilton, Canada Summary: Phase boundary motion arises in singular perturbation problems such as the Allen-Cahn equation in materials science, reaction-diffusion processes in chemical reactions and superconductors. Various methods are being developed to solve these problems and have already led to deep results: an example is the convergence of the Allen-Cahn equation to the motion by mean curvature of the limiting interfaces, which model antiphase boundaries in binary alloys. So far very little is known about physical systems with more than two phases or for models such as reaction-diffusion processes or eutectic growth. We shall address some of these problems. Day: July 5, 1995 Time: 09.30 S. Angenent, University of Wisconsin, USA: Examples of singularities in mean curvature flow M.V. Schatzman, Universite de Lyon, France: Asymptotic expansions for $u_t-\varepsilon ^2 \Delta u + 2u \bigl ( \vert u\vert ^2-1)=0$ P. Sternberg, Indiana University, USA; J. Rubinstein, Technion Haifa, Israel: Results on vortices in the Ginzburg-Landau system B.E. Stoth, Universit{\"a}t Bonn, Germany; L. Bronsard, McMaster University, Hamilton, Canada: The singular limit of a reaction-diffusion process F.L. Reitich, North Carolina State University, USA; H. Mete Soner, Carnegie Mellon University, Pittsburg, PA, USA: Three-phase boundary motions under constant velocities: the vanishing surface tension limit *********************************************************************** Turbulence Modelling - Research and Industrial Aspects Organizer: W. Rodi, Universit{\"a}t Karlsruhe, Germany Summary: Recent years have seen great advances in the calculation of practically relevant turbulent flows, and there has been much research in the area of developing improved turbulence models, extensive testing of these models, and their application to a large variety of flow problems in industry. Five renowned experts will review the recent developments in methods for calculating turbulent flows, including both the large-eddy simulation technique and statistical turbulence models, efforts in testing the various models in order to check their suitability for practical applications, and the role as well as the potential and limitations of the models for solving industrial flow problems. Day: July 3, 1995 Time: 15.30 J.H. Ferziger, Stanford University, USA: Prospects for large-eddy simulation of complex flows K. Hanjalic, Delft University of Technology, The Netherlands: Recent developments in turbulence modelling M. Leschziner, UMIST, United Kingdom: Collaborative validation initiatives for turbulent-flow CFD: does the outcome justify the effort? G. Scheuerer, Advanced Scientific Computing GmbH, Germany: Relevance of turbulence modelling for industrial flow calculations B. Aupoix, ONERA/CERT, France; J. Cousteix, G. Chevalier, CERT/ONERA, Toulouse, France: Aeronautical applications of turbulence models *********************************************************************** Polynomial System Solving: New Research Directions Organizer: L. Gonzalez-Vega, Santander, Spain Summary: Systems of polynomial equations arise in many technological and scientific areas. Solutions are usually sought by using numerical methods which themselves suffer from inherent drawbacks leading one to consider a symbolic approach giving exact representations of the solutions. The purely symbolic approach however is of limited use when numerical solutions are sought: it is better to consider the symbolic approach as being complementary to the numerical one and a mixed strategy integrating both computations in unified algorithms should be the winning strategy. \smallskip This minisymposium is devoted to presenting the new research directions in the area of symbolic polynomial system solving including the description of some applications and the discussion of the mixing of the symbolic and numerical approaches. Day: July 6, 1995 Time: 15.30 C. Traverso, Universit{\'a} di Pisa, Italy: Trends in symbolic polynominal system solving: the PoSSo project D. Lazard, Universit{\'e} Paris IV, France: Recent advances in Grobner solving T. Recio, University of Cantabria, Spain: Applications of symbolic solving: robot kinematics, scientific visualization and simulation H.M. M{\"o}ller, Fernuniversit{\"a}t Hagen, Germany: Symbolic $+$ numerical solving = seminumerical solving Day: July 7, 1995 Time: 09.30 L. Gonzalez-Vega, University of Cantabria, Spain: Recent advances in real solving F. Rouillier, Universite de Rennes 1, France: Software for symbolic polynomial system solving: the PoSSo solver J. Marchand, Ecole Polytechnique, France: The PoSSo test suite examples *********************************************************************** Time-Dependent Ginzburg-Landau Equations in Superconductivity and Turbulence Organizer: P. Tak{\'a}c, Washington State University, USA Summary: Time-dependent Ginzburg-Landau equations describe the temporal evolution of a pattern formation in {\it superconductivity} (phase transition from normal to superconducting state) and {\it turbulence} (instability phenomena in fluid dynamics). The following mathematical and physical problems will be addressed: Existence, uniqueness and analyticity of solutions, validity of the Ginzburg-Landau equations by computational (superconductivity) and rigorous analytical methods (fluid dynamics) and the large-time asymptotic behavior of the dynamical system generated by such equations. Also, numerical approximations and simulations of the dynamics on the attractor will be presented. A short video will show the vortex dynamics, such as vortex pinning, creation and annihilation. Day: July 4, 1995 Time: 09.30 P. Bollerman, Rijksuniversiteit Utrecht, The Netherlands: Validity of the Ginzburg-Landau approximation in Navier-Stokes type equations A. Doelman, Rijksuniversiteit Utrecht, The Netherlands: On the existence of localised structures in a perturbed Ginzburg-Landau equation H.G. Kaper, Argonne National Laboratory, USA: Numerical simulation of vortex dynamics in superconductors E.S. Titi, University of California at Irvine, USA; C. Foias, Indiana University, Bloomington, USA: Determining nodes, inertial manifolds and finite difference schemes for the complex Ginzburg-Landau equation D.J. Wollkind, Washington State University, USA: to be announced P. Tak{\'a}c, Washington State University, USA: Existence, uniqueness and analyticity of $L^p$-integrable solutions to semilinear parabolic systems *********************************************************************** Shape Optimization and Parameters Identification Organizer: M. Masmoudi, CNRS-Universit{\'e} Paul Sabatier, Toulouse, France Summary: The recent developments in the numerical simulation of physical phenomena (computational mechanics, electromagnetics, ...) for industrial applications has been spectacular. Usually, the results of the simulations are only used to update the design parameters. The goal of Shape Optimal Design (SOD) is to determine, in the most afficient manner, the best parameters of the studied structure. \smallskip Currently there are two important areas of research in this field: \item{$\bullet$} theoretical results concerning the existence and regularity of optimal solutions are few, and difficult to obtain, \item{$\bullet$} until some years ago SOD worked only with a user-defined topology, defined by the designer. New methods based on the theory of homogenization and composite material optimization provide spectacular results for the design of static linear structures. The generalization to other situations is an open problem. \smallskip In order to introduce SOD into industry, it is necessary to lower the cost of current SOD implementations. There are at least two means to achieve this objective: \item{$\bullet$} the fictitious domain method which reduces the dimension of the moving geometry. \item{$\bullet$} automatic differentiation and high order derivatives. \smallskip This minisymposium will give an overview on the previous mentioned topics and present the future trends of research in this field. Day: July 6, 1995 Time: 09.30 O. Pironneau, Universit{\'e} de Paris VI, France: An overview on shape optimization methods G. Francfort, Universit{\'e} Paris 13, France; G. Allaire, Commissariat {\`a} l'Energie Atomique; France: Shape optimization by the Homogenization method R. Glowinski, University of Houston, USA: Shape optimum design and fictitious domain method J. Haslinger, Charles University Praha, Czech Republic: Fictitious domain approach in optimal shape design problems A. Griewank, Technische Universit{\"a}t Dresden, Germany: Automatic differentiation of algorithms *********************************************************************** Interval Systems -- Theory and Algorithms Organizers: G. Alefeld, Universit{\"a}t Karlsruhe, Germany G. Mayer, Universit{\"a}t Rostock, Germany Summary: If the coefficient matrix $A$ of a linear system $Ax=b$ is known to vary between $A_1$ and $A_2$ and correspondingly the right hand side between $b_1$ and $b_2$ one is interested in the set $S$ of all possible solutions. \smallskip The minisymposium will focus on the description and on effective methods for the inclusion of $S$. A series of new results will be presented especially in the case of a symmetric matrix $A$. Day: July 6, 1995 Time: 15.30 J. Rohn, Charles University, Czech Republic: Solving linear interval equations G. Heindl, Bergische Universit{\"a}t-GH Wuppertal, Germany: Some inclusion results based on generalized versions of the Oettli-Prager-theorem G. Mayer, Universit{\"a}t Rostock, Germany; G. Alefeld, Universit{\"a}t Karlsruhe, Germany: On the solution set of symmetric interval systems C. Jansson, Technische Universit{\"a}t Hamburg-Harburg, Germany: A verification algorithm for symmetric systems with interval data *********************************************************************** Optimal Control in Aerospace Engineering Organizers: R. Bulirsch, Technische Universit{\"a}t M{\"u}nchen, Germany R. Callies, Technische Universit{\"a}t M{\"u}nchen, Germany Summary: The minisymposium reflects the state of the art in the application of infinite dimensional optimization to open and closed loop control problems in aerospace engineering. Emphasis lies on the solution of practical industrial problems in aerospace. Various real-life optimal control problems are treated with their elaborate mathematical models being close to reality. \smallskip For these very complicated problems indirect methods turn out to be the right choice when highly accurate optimal solutions are to be computed and various constraints, inner point and jump conditions are to be handled in a mathematically correct way. Not only the optimal trajectories are calculated, but also different approaches are presented for the robust and optimal feed-back control of these solutions. Day: July 3, 1995 Time: 10.00 V.H. Schulz, Universit{\"a}t Heidelberg, Germany: A direct RSQP method for path planning of satellite mounted robots K. Chudej, Technische Universit{\"a}t M{\"u}nchen, Germany: Extended necessary conditions for challenging state-constrained optimal control problems in aerospace engineering W. Grimm, Universit{\"a}t Stuttgart, Germany: Optimal control of a capsule in the super-/subsonic region B. Kugelmann, Technische Universit{\"a}t M{\"u}nchen, Germany: Parallel computation of feedback controls in aerospace engineering H.J. Pesch, Technische Universit{\"a}t Clausthal, Germany: Applications of optimal control and differential game theory in aerospace engineering M. Dinkelmann, Technische Universit{\"a}t M{\"u}nchen, Germany; G. Sachs, Technische Universit{\"a}t M{\"u}nchen, Germany: Optimum three-dimensional hypersonic cruise for stage separation *********************************************************************** Application of Numerical Techniques in Fluid Mechanics Organizer: F. Durst, Universit{\"a}t Erlangen-N{\"u}rnberg, Germany Summary: In recent years, advancements in numerical methods and rapid developments in parallel computing have led to new ways to numerically solve fluid flow problems as they exist in many fields of engineering. Close collaborations of mathematicians, computer scientists and engineers have resulted in highly efficient computer codes that are readily applicable to study complex flow fields. This will be demonstrated in the five presentations of the proposed ICIAM-Minisymposium and will be discussed in the panel session which follows the presentations. In this way, contributions of the audience to the subject are invited in order to provide an overview of the state-of-the-art of numerical fluid flow studies. Different aspects of the usage of high-performance scientific computing in fluid mechanics are outlined in the presentations. Day: July 4, 1995 Time: 15.30 M. Breuer, Universit{\"a}t (TH) Karlsruhe, Germany; W. Rodi, Universit{\"a}t Karslruhe, Germany: Large-eddy simultation of internal and external flows P. Bontoux, Marseille, France: Sprectral simulation of oscillatory convection K. Naitoh, Nissan Research Center, Japan; Y. Takagi, Nissan Research Center, Kunio Kuwahara, Japan: Cycle-resolved computation of turbulent premixed-flames in engines M. Sch{\"a}fer, Universit{\"a}t Erlangen-N{\"u}rnberg, Germany; F. Durst, Universit{\"a}t Erlangen-N{\"u}rnberg, Germany: Simulation of complex flows with parallel algorithms M. Meinke, RWTH Aachen, Germany; E. Krause, RWTH Aachen, Germany: Simulation of wake flows *********************************************************************** Mathematical Modelling in the Paper Industry Organizer: P. Neittaanm{\"a}ki, University of Jyv{\"a}skyl{\"a}, Finland Summary: The symposium focuses on mathematical models and methods relevant to understanding the papermaking process and the properties of paper. \smallskip Several contributions are devoted to analysing multiple flows in a porous medium with applications to the wet pressing, to the drying and to the copying behaviour of paper. Another main line in the symposium is the study of random media. Mechanical and acoustic properties of random media are studied using, in particular, cellular automates and Monte Carlo methods. Day: July 3, 1995 Time: 10.00 K. Hiltunen, University of Jyv{\"a}skyl{\"a}, Finland: Coupled flow between saturated deformable porous medium and open domain in paper forming R. Pietik{\"a}inen, University of Jyv{\"a}skyl{\"a}, Finland: Modelling of coupled heat and mass transfer in copying paper J.P. H{\"a}m{\"a}l{\"a}inen, Technical Research Centre of Finland, Finland: Modelling and simulation of fluid flows in a paper machine headbox M. Karlsson, VTT Energy, Finland; H. Lepom{\"a}ki, VTT Energy, Finland; M. Luoma, Lappeenranta University of Technology, Finland; L. Nystr{\"o}m, Lappeenranta University of Technology, Finla: Identification of drying parameters for paper drying model M. Karlsson, VTT Energy, Finland; P. Rajala, Jyv{\"a}skyl{\"a} Technical Research Centre, Finland: A drying model for coated paper with respect of printing properties Day: July 5, 1995 Time: 15.30 M. Kellom{\"a}ki, University of Jyv{\"a}skyl{\"a}, Finland; J. {\AA}strom, University of Jyv{\"a}skyl{\"a}, Finland: Elastic waves in random fiber networks A. Koponen, University of Jyv{\"a}skyl{\"a}, Finland: Lattice gas automata and flow through porous media V. R{\"a}is{\"a}nen, Center for Scientific Computing Ltd., Finland; M.J. Alava, Helsinki University of Technology, Finland; R.M. Nieminen, Helsinki University of Technology, Finland: Mechanics of random fiber networks V.J. Rivkind, University of St. Petersburg, Russia; M. Karlsson, VTT Energy, Finland; O. Timofeev, St. Petersburg TU of Plants Polymers, Russia: Mathematical and numerical modeling of paper drying with supercritical moisture content *********************************************************************** Numerical Methods for Maxwell's Equations Organizers: A. Bachelot, Universit{\'e} Bordeaux, Talence, France P. Joly, Le Chesnay, France Summary: The purpose of this Minisymposium is to present the state of the art in the field of numerical modelisation for Maxwell's equations. The stress is put upon a broad range of problems including temporal and harmonic propagation phenomena, scattering problems as well as gratings, electromagnetic wave guides,... For these topics, the recent advances shall be described from both the points of view of fine theoretical analysis and numerical implementation. The talks should be of interest to a broad audience of mathematicians and engineers. Day: July 5, 1995 Time: 15.30 P. Monk, University of Delaware, USA: High order edge finite element methods for Maxwell's equations F. Collino, INRIA, France: Absorbing boundary conditions in finite domains T. Abboud, Ecole Polytechnique, France: Diffraction of electromagnetic waves by plane and curved gratings V. Lange, CEA/CESTA, France: Retarded potentials method for the Maxwell equations A.d.l. Bourdonnaye, INRIA, France: Integral equations and high frequency scattering C. Hazard, ENSTA, France: Regularization of Maxwell's equations for diffraction problems *********************************************************************** Dynamic Optimization in Robotics Organizers: F.L. Chernousko, Russian Academy of Science, Moscow, Russia M. Steinbach, Universit{\"a}t Heidelberg, Germany Summary: Increasing competititon and quality standards as well as economic reasons impose high demands on speed and precision of robot maneuvers, both in industry and sciences. In order to meet these requirements in practice one must obtain solutions to optimal design, calibration, off-line and feed-back control problems, calling for sophisticated dynamic optimization strategies and algorithms. The session covers the full range of these interrelated mathematical problems, discussing new theoretical approaches and numerical algorithms. Practical requirements such as collision avoidance and safety restrictions are addressed; the real-life applications presented include mobile robots as well as manipulators in car manufacturing and in space. Day: July 6, 1995 Time: 15.30 J.-C. Samin, Universit{\'e} Catholique de Louvain, Belgium; P. Fisette, B. Raucent, University of Louvain-la-Neuve, Belgium: Contribution to the identification of dynamic parameters of robot manipulators with closed loops N.N. Bolotnik, Russian Academy of Sciences, Russia; G.V. Kostin, F. L. Chernousko, Institute for Problems in Mechanics, MoMoscow, Russia: Optimization of walking robot parameters F.L. Chernousko, Russian Academy of Sciences, Russia: Suboptimal feedback control of manipulation robots H.G. Bock, Universit{\"a}t Heidelberg, Germany; M.C. Steinbach, Universit{\"a}t Heidelberg, Germany: Optimal trajectory planning for robots on the factory floor V.H. Schulz, Universit{\"a}t Heidelberg, Germany: Optimal paths for satellite mounted robots R.W. Longman, Columbia University, USA: The set of possible learning control laws for maximum precision robot path tracking *********************************************************************** Integrable Dynamical Systems and Their Applications Organizers: D. Blackmore, New Jersey Institute of Technology, USA V.K. Mel'nikov, Joint Institut for Nuclear Research, Moscow, Russia A.K. Prikarpatsky, Ukrainian Academy of Sciences, L'viv, Ukraine Summary: We propose to bring together experts in integrable dynamical systems and related fields to present their latest research results and discuss their work among themselves and others who are interested in this important discipline. \smallskip New results, techniques and applications of nonlinear dynamical systems will be addressed. Some of the application areas covered will be biology, fluid mechanics and optics. \smallskip Among the research trends to be discussed are harmonic maps in discrete systems, parametric integrability and restricted integrability. Some of the new approaches covered will be advances in the IST, the Bogoliubov hierarchy and the gradient-holonomic algorithm. The main advantage of these methods is their generality. A goal of the minisymposium is to present these methods in ways useful to engineers and industrial researchers. Day: July 4, 1995 Time: 15.30 D.L. Blackmore, New Jersey Institute of Technology, USA; J. Kappraff: Integrable discrete dynamics and Fibonacci sequences I.D. Chueshov, University of Kharkov, Ukraine: On a description of long-time behaviour of dissipative perturbations of infinite dimensional Hamiltonian systems W. Strampp, Gesamthochschule Kassel, Germany; F. Steuernagel: On the $\tau$ -function of the Kaup-Broer system D. Papageorgiou, New Jersey Institute of Technology, USA: Stability of core-annular flows: the Kuramoto-Sivashinsky equation and modifications C.C. Lim, Rensselaer Polytechnic Institute, USA: Whiskered tori in vortex dynamics V.K. Mel'nikov, Joint Inst. for Nuclear Research Moscow, Russia: Integrable dynamical systems with sources Day: July 5, 1995 Time: 09.30 A.K. Prikarpatsky, Ukrainian Academy of Sciences, Ukraine: Geometric Models of the Blackmore's swept volume - dynamical systems and their integrability V.V. Gafiychuk, Ukrainian Academy of Sciences, Ukraine; I.A. Lubashevskii, Academy of Technol. Sciences of Russian Federation, Moscow, Russia: The projection dynamics of high dissipative systems M.M. Prytula, L'viv State University, Ukraine; J. Tavantzis, New Jersey Institute of Technology, Newark, USA: Parametric integrability of the nonuniform dynamical systems within Prikarpatsky's gradient-holonomic algorithm Y.M. Sidorenko, University of L'viv, Ukraine: Multicomponent integrable reductions of $(2+1)$-dimensional integrable systems I.L. Nizhnik, Ukrainian National Academy of Sciences, Ukraine: The solitons attraction law R. Samuliak, State University "L'viv Polytechnic", Ukraine: Integrability of the Higgs type nonlinear dynamical systems N. Euler, Ukrainian Academy of Sciences, Ukraine: Painleve test and exact solutions of differential equations *********************************************************************** Flow in Reactive Porous Media and Pollution Organizers: P. Knabner, Universit{\"a}t Erlangen-N{\"u}rnberg, Germany C.J. van Duijn, Delft University of Technology, The Netherlands Summary: This minisymposium is concerned with various aspects of modelling and mathematical analysis of flow and transport processes in porous media such as soils or aquifers. These problems have a considerable importance for environmental protection and remediation. Numerical simulation has become a strong engineering tool, but has to be based on proper modelling and complemented by a qualitative analysis of the models. These are interesting mathematical problems, as due to the strongly heterogeneous microscopic structure of the medium, the transfer of microscopic models to a macroscopic, observable scale is not obvious and often leads to new, strongly nonlinear models in terms of systems of nonlinear, possibly degenerate partial differential equations. Day: July 7, 1995 Time: 09.30 A. Mikelic, Universit{\'e} Lyon - 1, France: Recent results in modelling of the flow through porous media using homogenization R.E. Grundy, University of St. Andrews, United Kingdom: Large time profiles for some contaminant transport models E. Priesack, GSF-Institut f{\"u}r Boden{\"o}kologie, Germany: A micro-structure model for microbial growth in aggregated soils M.A. Peletier, Delft University of Technology, The Netherlands; C.J. van Duijn, Tu Delft, The Netherlands: An adsorption model with a time-dependent isotherm C.J.v. Duijn, Delft University of Technology, The Netherlands; P. Knabner, Universit{\"a}t Erlangen, Germany: Free boundary problems from crystal dissolution in porous media flow *********************************************************************** Usage and Characteristics of Tangent Linear and Adjoint Models in Geosciences Organizer: A. Griewank, Technische Universit{\"a}t Dresden, Germany Summary: Like other complex simulation efforts, the modeling of atmospheric or oceanographic circulations involves the adjustment of initial and boundary conditions as well as other model parameters to fit empirical data. This optimization and tuning task requires accurate sensitivity information, which is usually obtained by augmenting the computer models with so-called tangent linear and adjoint codes. General purpose software tools for generating these emhancements have been developed, but they tend to be less efficient than hand-coded models. \smallskip The speakers in this symposium will report on the use of tangent linear or adjoint models in metereology and oceanographty as well as the development of suitable differentiation software. Day: July 7, 1995 Time: 15.30 C.H. Bischof, Argonne National Laboratory, USA: Experiences with automatic differentiation in atmospheric modelling N. Rostaing-Schmidt, INRIA, France: Examples of tangent linear and adjoint codes of atmospheric and oceanographic models produced using Odyssee A. Rhodin, GKSS Research Center, Germany; U. Callies, GKSS Research Center, Geesthacht, Germany; D.P. Eppel, GKSS Research Center, Geesthacht, Germany: An object-oriented approach to program a meteorological model and its adjoint in Fortran 90 F.-X. Le Dimet, Universite Joseph Fourier Grenoble 1, France: Sensitivity and predictability studies in the geosciences *********************************************************************** Domain Decomposition Methods Organizers: U. Langer, Johannes-Kepler-Universit{\"a}t Linz, Austria Y. Kuznetsov, Moscow, Russia R. Glowinski, University of Houston, Texas, USA\\ O. Widlund, New York University, USA Summary: Finite element, finite volume, finite difference and many other mesh discretization methods for problems with partial derivatives lead to large scale systems of linear and nonlinear algebraic equations. These systems are very sparse and often very badly conditioned. These systems are hard to solve by direct and classical iterative methods, especially on modern computers. \smallskip In this framework, domain decomposition methods are very attractive for both numerical and applied mathematicians as well as people working in scientific computing. To construct an efficient solver, original large scale problems are reduced to a set of smaller subproblems by domain decomposition while treating interfaces between them by well developed iterative procedures like PCG and GMRES. At the proposed minisymposia up-to-day results on design, theoretical investigation, implementation algorithms (especially for parallel computers) as well as advanced applications of domain decomposition methods will be presented. The minisymposia collects leading experts in the field from many countries. Day: July 5, 1995 Time: 15.30 S. Nepomnyaschikh, Russian Academy of Science, Russia: Preconditioning operators for elliptic problems A. Meyer, Technische Universit{\"a}t Chemnitz-Zwickau, Germany: Application of fast parallel solution methods in continuum mechanics D. Maksymilian, Warsaw University, Poland: to be announced B. Khoromskij, Joint Institute for Nuclear Research, Russia; G. Schmidt, WIAS, Berlin, Germany: Fast iterative substructuring solvers for biharmonic Dirichlet problem on polygonal domains Day: July 6, 1995 Time: 09.30 R. Rannacher, Universit{\"a}t Heidelberg, Germany: A domain-splitting method for nonstationary convection diffusion problems A. Quarteroni, Technical University of Milan, Italy: to be announced Y.M. Laevsky, Russian Academy of Sciences, Russia; S.V. Gololobov, Novosibirsk State University, Russia: Explicit-implicit domain decomposition methods for parabolic equations V.V. Shaidurov, Russian Academy of Sciences, Russia: Some implementation of the cascadic methods *********************************************************************** Numerical Methods for Compressible Flows Organizers: D. Kr{\"o}ner, Universit{\"a}t Freiburg, Germany R. Jeltsch, ETH Z{\"u}rich, Switzerland B. Van Leer, The University of Michigan at Ann Arbor, USA Summary: Modern upwind techniques are necessary for succesful numerical simulations of compressible flows. Usually real problems appear in three space dimensions, in particular in complex geometries and therefore additional effort for improving the mathematical tools has to be made. During the last years many new and effective methods have been developed: modern multidimensional schemes beside the classical fractional step algorithms, higher order finite volume methods on unstructured grids for complex geometries, error estimators to perform local mesh refinement and coarsening, grid alignement and multigrid accelaration. In a final panel discussion we shall discuss the advantages and shortcomings of the various discretizations for multidimensional Euler schemes. Day: July 6, 1995 Time: 15.30 A. Szepessy, Kungl Tekniska H{\"o}gskolan, Sweden: Adaptive finite element methods for system of conservation laws D. H{\"a}nel, Universit{\"a}t Duisburg, Germany: Finite-volume schemes with grid alignment M. Fey, California Institute of Technology, USA: Multidimensional Euler scheme S. Ta'asan, Carnegie-Mellon University, USA: Canonical variables and multigrid methods for conservation laws Day: July 7, 1995 Time: 09.30 M. Wierse, Universit{\"a}t Freiburg, Germany: Upwind finite volumes schemes in timedependent geometries T.H. Sonar, DLR G{\"o}ttingen, Germany; A. S{\"u}li: A dual graph norm refinement indicator for finite volume approximations of the Euler equations H. Deconinek, K{\'a}rman Institute for Fluid Dynamics, Belgium: Fluctuation distribution scheme *********************************************************************** Education in Scientific Computing Organizer: G. Golub, Stanford University, USA Summary: With the advent of powerful computers, including parallel processors and work stations, the methodology of scientific computing represents an important component of engineering and science. It is important to provide advanced training to students with a variety of backgrounds in this area. The education of students working in this area will by necessity be highly interdisciplinary. Applied mathematics, numerical analysis and computer science serve as a unifying theme to this discipline but applications areas will play a vital role, too. We will discuss various programs that have been designed and we shall examine future directions. Day: July 4, 1995 Time: 09.30 G.H. Golub, Stanford University, USA: to be announced H.A. van der Vorst, Utrecht University, The Netherlands: to be announced S. Norsett, Universitetet i Trondheim, Norway: to be announced M. Griebel, Technische Universit{\"a}t M{\"u}nchen, Germany; C. Zenger, Technische Universit{\"a}t M{\"u}nchen, Germany: Computational fluid dynamics -- an interdisciplinary course in scientific computing *********************************************************************** Domain Optimization Organizers: K. Kunisch, Technische Universit{\"a}t Berlin, Germany J. Sokolowski, Vandoeuvre Les Nancy, France Summary: Shape optimization concerns itself with the optimal choice of geometrical domains for systems described by partial differential equations. Applications include structural mechanics, electromagnetics, fluid mechanics and free boundary value problems. Since such problems are nonconvex and highly nonlinear the analysis requires an appropriate formulation which allows the study of existence and optimality conditions as well as the development of efficient numerical schemes. Contributions to this minisymposium will include the recent development of the general theory of shape Hessians and Newton-like methods applied to nonlinear systems of partial differential equations. Day: July 5, 1995 Time: 09.30 M.C. Delfour, Universit{\'e} de Montr{\'e}al, Canada: Curvatures and skeletons in shape optimization M.D. Gunzburger, Virginia Tech., USA; J. Burkhardt, J. Peterson, Interdisciplinary Center for Applied Mathematics, Blacksburg, USA; H. Kim, Seoul National University, Korea: Shape optimization problems for the Navier-Stokes equations J. Haslinger, Charles University Praha, Czech Republic: Fictitious domain approach in free boundaries K. Ito, North Carolina State University, USA: Shape optimization problems in flow control S. Stojanovic, University of Cincinnati, USA: Optimal control of free boundaries J. Simon, Universit{\'e} Blaise Pascal, France: Loss of regularity due to normal variations of a domain Day: July 7, 1995 Time: 15.30 S. Ta'asan, Carnegie-Mellon University, USA: One shot methods for domain optimization J.-P. Zol{\'e}sio, CNRS - INLN, France: Structure of shape hessian using lie brackets M. Pierre, Universit{\'e} de Nancy 1, France; E. Rouy, University of Nancy, France: A tridimensional inverse shaping problem A.M. Khludnev, Lavrentyev Institute of Hydrodynamics, Russia: Identification methods for cracks in elastic plates P. Neittaanm{\"a}ki, University of Jyv{\"a}skyl{\"a}, Finland; D. Tiba, University of Jyvaskyla, Finland: Fixed domain approach in optimal shape design *********************************************************************** Multidisciplinary Design Optimization Organizer: N. Alexandrov, Hampton, Virginia, USA Summary: Multidisciplinary Design Optimization (MDO) has recently emerged as a field of research and practice that has brought together many previously disjointed disciplines and tools of engineering and applied mathematics. MDO is a methodology for the design of complex engineering systems, such as aircraft and other mechanisms, the behavior of which is determined by interacting subsystems. This technology is also applicable to parameter identification and control. It is important to any industry where design of complex systems takes place. The potential of MDO for improving the design process and reducing the manufacturing cost of complex systems is widely recognized by the engineering community. Day: July 4, 1995 Time: 15.30 J.E. Dennis, Rice University, USA: Nonlinear programming approaches to MDO H.G. Bock, Universit{\"a}t Heidelberg, Germany: Boundary value problem methods for large-scale nonlinear programming J.P. Schl{\"o}der, Universit{\"a}t Heidelberg, Germany: Parallel methods for parameter estimation problems R.D. Braun, NASA Langely Research Center, USA; I.M. Kroo, Stanford University, USA: Use of collaborative optimization in a multidisciplinary enviroment N. Alexandrov, NASA Langely Research Center, USA: Experience with nonlinear programming algorithms in multidisciplinary design optimization *********************************************************************** Subspace Iteration Methods Organizers: H.v.d. Vorst, University of Utrecht, The Netherlands O. Nevanlinna, Helsinki University of Technology, Finland Summary: In many scientific and industrial applications we are faced with the problem of solving very large (non)linear systems or large linear eigenproblems. A modern approach to these problems is to restrict the solution space to some well chosen subspace and to solve the given problem for a projected system with respect to this subspace. Guided by residual information one then tries to expand the subspace in order to further improve the current solution. \smallskip In nonlinear problems the subspace has to be changed from time to time in order to account for the nonlinearity of the problem, e.g., the subspace is chosen in agreement with the Jacobian as in Newton's method. \smallskip For eigenvalue problems one has to keep the dimension of the subspace reasonably small and the question arises which information from the current subspace can be discarded in favor of new search directions. These issues will be the main topic of our minisymposium in relation with problems from industry (circuit simulation) and science (MHD problems). Day: July 3, 1995 Time: 15.30 G. Degrez, K{\'a}rman Institute, Belgium; E. Issman: Solving compressible flow problems with matrix free type methods W.H. Schilders, Philips Research Laboratories, The Netherlands: Nonlinear solvers based on Krylov methods for circuit simulation D.C. Sorensen, Rice University, USA: Implicitly restarted Arnoldi/Lanczos methods A. Booten, Centrum voor Wiskunde en Informatica, The Netherlands: Jacobi-Davidson methods for generalized MHD-eigenvalue problems *********************************************************************** New Monte Carlo Approaches for Numerical Solution in High-Dimensional Problems of Mathematical Physics Organizers: G. Mikhailov, Novosibirsk, Russia K.K. Sabelfeld, Novosibirsk, Russia M.H. Kalos, Cornell University Ithaca, New York, USA Summary: New stochastic models attract an increasing attention in many areas. However rigorous analysis of such models is not simple, since it requires deep knowledge in the theory of probability and statistical modelling by computers, and in the relevant fields of Natural Sciences. The high-dimensional problems of mathematical physics is the area where the new stochastic models are developed and tested. \smallskip Significant applications already exist -- we mention few of them: random walk algorithms for solving boundary value problems of electrostatics, elasticity theory, linear and non-linear diffusion and kinetic Boltzmann equations, heat and mass transfer and turbulence. Day: July 5, 1995 Time: 09.30 G.A. Mikhailov, Academy of Sciences, Russia: New Monte Carlo methods with estimating derivatives M.H. Kalos, Cornell University, USA: Monte Carlo Methods and parallel computers K.K. Sabelfeld, Academy of Sciences, Russia: A probabilistic representation for systems of elliptic equations D. Talay, INRIA, France: Numerical solution of stochastic differential equations J. Spanier, Blaisdell House, USA: Monte Carlo algorithms in linear radiative transfer N. Newton, University of Essex, United Kingdom: Numerical solution of non-linear diffusion equations W. Wagner, WIAS Berlin, Germany; S. Rjasanow, Universit{\"a}t Kaiserslautern, Germany: Stochastic systems of weighted particles approximating the spatially inhomogeneous Boltzmann equation *********************************************************************** High Performance Computing: A Current View and Future Directions for Selected Disciplines Organizer: G. Astfalk, Convex Computer Corp., Richardson, USA Summary: High performance computing and applied mathematics are not separable. In this minisymposium we have speakers to present, for several different disciplines within the realm of applied mathematics, the current advantage and usage of high performance computing. We also indicate what the future requirements in each of these disciplines are for both computing and mathematics. \smallskip Related to applied mathematics on high performance computers is the notion of the parallel programming and knowledge of the machine architectures. Speakers will address these topics as well. Day: July 3, 1995 Time: 10.00 G. Astfalk, Convex Computer Corp, USA: High performance computing and applied mathematics: expectations for the future K. Burrage, The University of Queensland, Australia: A parallel implementation of a deflation algorithm for solving large linear systems of equations C. Douglas, IBM-Watson Research Center, USA: High performance computing and numerical simulation of laminar diffusion flames with finite rate chemistry S.A. Zenios, University of Cyprus, Cyprus: High-performance computing in financial modeling Day: July 7, 1995 Time: 15.30 W.D. Gropp, Argonne National Laboratory, USA: Why should software be called parallel? R.J. Hanson, Visual Numerics Inc., USA: A view of distributed mathematical software using Fortran 90 H.D. Simon, NASA Ames Research Center, USA: Seven years of parallel computing at NAS (1987-1994): what have we learned? T. Sterling, Goddard Space Flight Center, USA: HPC architecture requirements for earth and space science applications *********************************************************************** Homogenization Methods and their Applications in Mechanics of Composite Materials and Lattice Structures Organizers: N.S. Bakhvalov, Moscow University, Russia G. Panasenko, Moscow State University, Russia U. Hornung, Universit{\"a}t der Bundeswehr M{\"u}nchen, Germany Summary: Composite materials and lattice structures are widely adopted in the modern engineering, especially spacecraft and aircraft industries. The homogenization method describe the physical processes taking place in such non-homogeneous media. Nowadays the new homogenization schemes containing two or more small parameters have been developed: multi-component homogenization method, two scale convergence, lattice homogenization etc. These methods are applied to problems of structural optimization and optimization of form. These homogenization models describe some new physical effects. Therefore they are actual for the industrial and applied mathematics. Day: July 3, 1995 Time: 15.30 N. Bakhvalov, Moscow University, Russia: Effective properties of irregularly reinforced materials A. Bourgeat, Universite Jean Monnet de Saint-Etienne, France: Computation of effective permeability for randomly heterogeneous porous media D. Cioranescu, Universit{\'e} Pierre et Marie Curie, France; D.G. Griso, Universit{\'e} Pierre et Marie Curie, Paris, France: Asymptotic behaviour of thin tall structures M.E. Eglit, Moscow University, Russia: Averaged equations for statics and dynamics of incompressible elastic media with pores U. Hornung, Universit{\"a}t der Bundeswehr M{\"u}nchen, Germany: Monte-Carlo simulations for stochastic homogenization G.P. Panasenko, Moscow University, Russia: Homogenization of lattice structures: L-convergence J. Saint Jean Paulin, University of Metz, France; I. Charpentier, University of Metz, France: Theoretical and numerical study of thin reticulated structures *********************************************************************** Fictitious Domain Methods and Applications Organizers: R. Glowinski, University of Houston, Texas, USA Yu. Kuznetsov, Russian Academy of Sciences, Moscow, Russia J. Periaux, Dasssault Aviation, St. Cloud, France Summary: The fictitious domain methodology consists in a replacement of an original differential or mesh problem in a complex shape domain by a similar problem in a simpler shaped domain.The main advantage of this approach is an opportunity to indroduce very efficient separable and multilevel preconditiners depending on particular properties of operators and meshes. Within the proposed minisymposium a number of lectures on theoretical and numerical aspects of fictitious domain methods as well as on their applications to solving some challenging scientific and industrial problems in fluid dynamics,electromagnetism and acoustics will be presented. Day: July 4, 1995 Time: 15.30 R. Glowinski, University of Houston, USA; M. O. Bristeau, INRIA, Le Chesnay, France; J. Periaux, Dassault Aviation, St Cloud, France: Fictitious domain/exact controllability methods applied to the scattering of electromagnetic waves by coated obstacles Y. Kuznetsov, Russian Academy of Sciences, Russia: Mathematical foundations of fictitious domain methods J. Periaux, Dassault Aviation, France; R. Glowinski, University of Houston, Texas, USA; T.-W. Pan, University of Houston, Texas, USA: Fictitious domain/domain decomposition methods applied to the unsteady Navier-Stokes flow simulation N. Bakhvalov, Moscow University, Russia: Fictitious domain methods for solving elasticity problems Day: July 5, 1995 Time: 09.30 H. Kawarada, Chiba University, Japan; H. Fujita, Meiju University, Kawasaki, Japan; H. Kawahara, Chiba, Japan: Distribution theoretic approach to fictitious domain methods for the Neumann problems A. Bespalov, INRIA-Rocquencourt, France: Fictitious domain method for solving the Maxwell equations M. Dryja, Warsaw University, Poland: Additive Schwarz methods for elliptic problems with discontinuous coefficients on irregular subtructures O.B. Widlund, New York University, USA: Domain decomposition methods for higher order finite approximations of elliptic problems *********************************************************************** Analysis of Microstructures and Advanced Materials Organizers: S. M{\"u}ller, Universit{\"a}t Freiburg, Germany J.M. Ball, Heriot-Watt University, Edinburgh, United Kingdom D. Kinderlehrer, Carnegie-Mellon University, Pittsburgh, USA \\ M. Luskin, University of Minnesota, USA Summary: The behaviour of materials is strongly influenced by their microscopic structure. In recent years there has been considerable progress to develop mathematical tools to describe and analyze such microstructures. This effort involves diverse mathematical disciplines such as the calculus of variations, partial differential equations, functional analysis and differential geometry as well as a strong interaction between mathematics and material science. It has already found applications to the design of advanced materials such as shape memory materials or materials with giant magnetostriction. At the same time it lead to new and challenging mathematical problems. \smallskip The minisymposium will focus on recent research in the following three areas: Prediction of microstructures by minimization, dynamics of microstructures and hysteresis and the computation of microstructures. Day: July 3, 1995 Time: 10.00 J.M. Ball, Heriot-Watt University, United Kingdom: Local minimizers and phase transformations M. Chipot, Universit{\'e} de Metz, France: Computation of microstructures C. Collins, The University of Michigan, USA: to be announced A. DeSimone, Universita di Roma II, Italy: Characterization of the macroscopic response of magnetostrictive materials via microstructural analysis I. Fonseca, Carnegie-Mellon University, USA: Nonconvex variational problems and relaxations techniques Day: July 3, 1995 Time: 15.30 G. Friesecke, Universit{\"a}t Freiburg, Germany: Dynamic energy minimization and hysteresis D. Kinderlehrer, Carnegie-Mellon University, USA: Magnetoelastic interactions M. Luskin, University of Minnesota, USA; P. Kloucek, University of Minnesota, USA: The computation of the dynamics of the martensitic transformation S. M{\"u}ller, Universit{\"a}t Freiburg, Germany: Microstructures with multiple scales Day: July 4, 1995 Time: 09.30 P. Pedregal, Universidad Complutense de Madrid, Spain: Remarks about the computation of microstructure F.L. Reitich, North Carolina State University, USA; O.P. Bruno, Georgia Institute of Technology, Atlanta, USA; P.H. Leo, University of Minnesota, Minneapolis, USA: Nonlinear homogenized energies and the overall behavior of shape memory polycrystals V. Sver{\'a}k, University of Minnesota, USA: Quasiconvexity and the continuum theory of microstructure N.J. Walkington, Carnege Mellon University, USA: Approximation of nonconvex variational problems *********************************************************************** Mathematical Modelling of Mining and Mineral Processing Organizers: A. Fitt, University of Southampton, United Kingdom N. G. Barton, CSIRO Division of Mathematics and Statistics, Australia Summary: Numerous physical problems arise in the mining and mineral processing industries for which mathematical models may give insight and understanding into possible solutions. This minisymposium considers one such problem in geothermal mining, one in rock blasting and two in mineral processing. The modelling required will be discussed and the rich variety of mathematical methods used to solve the equations will be presented. Mathematical topics will include particle methods, singular integral equations, models for granular flow and nonlinear diffusion. Day: July 4, 1995 Time: 15.30 N.G. Barton, CSIRO, Australia; P.W. Cleary, CSIRO, Clayton, Australia: Use of particle methods to design mineral processing equipment A.D. Fitt, University of Southampton, England: Crack propagation in a geothermal energy reservoir C.-H. Li, CSIRO, Australia: Continuum methods for granular flow in mineral processing C.P. Please, University of Southampton, England: Gas propagation in cracks from cylindrical boreholes during blasting in mines *********************************************************************** Random Dynamical Systems Organizers: L. Arnold, Universit{\"a}t Bremen, Germany N.S. Namachchivaya, University of Illinois at Urbana, USA Summary: The subject of the minisymposium is the theory of dynamical systems under randomness and its applications, mainly to nonlinear engineering systems. \smallskip There has been dramatic theoretical progress, mainly based on the multiplicative ergodic theorem, in the areas of stochastic stability, invariant manifolds, normal forms, stochastic bifurcation theory, stochastic structural stability, stochastic chaos and attractors, classification, etc. \smallskip Parallel to this we witness a growing interest of the applied sciences in random systems, and the results available today allow a much deeper insight and better analysis. \smallskip The minisymposium presents six speakers who are leading experts from engineering and mathematics. They will try to give an up-to date account of the theory and its applications. Some of the talks will have survey character. Day: July 5, 1995 Time: 15.30 R.Z. Khasminskii, Wayne State University, USA: Stability index for stochastic systems S.T. Ariaratnam, University of Waterloo, Canada: Stochastic delay differential equations P. Baxendale, University of Southern California, USA: Bifurcation theory for stochastic differential equations H. Crauel, Saarbr{\"u}cken, Germany; F. Flandoli, Scuola Superiore, Pisa, Italy: Attractors for random dynamical systems N. Sri Namachchivaya, University of Illinois at Urbana-Champaign, USA; Y. Liang, University of Illinois at Urbana-Champaign, USA; H. Van Roessel, University of Alberta, Edmonton, Canada: Random dynamical systems in engineering L. Arnold, Universit{\"a}t Bremen, Germany: Recent progress in the theory of random dynamical systems *********************************************************************** Control of Complex Systems Organizers: F. Colonius, Universit{\"a}t Augsburg, Germany W. Kliemann, Iowa State University, Ames, USA Summary: The speakers -- mathematicians and engineerings -- will address a wide range of concepts and problems in the interplay between control and dynamical systems. The focus will be on problems, for which deeper insight into the dynamical behavior, like invariant manifolds or bifurcations, allows a better understanding of controlled or excited systems in the engineering sciences. In particular, stabilization procedures with time varying feedback, numerical stabilization algorithms, disturbance attenuation in $H$-infinity theory, analysis of non autonomous control systems via skew product flows, global bifurcation phenomena leading to chaotic motion in parametrically excited systems with symmetry, and the theoretical and experimental tracking of unstable motions via insight into the underlying geometry will be discussed. Day: July 6, 1995 Time: 09.30 J.-M. Coron, Ecole Normale Superieure de Cachan, France: Feedback stabilization of controllable systems L. Gr{\"u}ne, Universit{\"a}t Augsburg, Germany: Numerical stabilization of bilinear control systems D. Flockerzi, Universit{\"a}t W{\"u}rzburg, Germany: Integral manifolds in the $L_2$-gain analysis of nonlinear systems I. Schwartz, Naval Research Laboratory, USA: Geometric control and tracking of nonlinear dynamical systems Day: July 6, 1995 Time: 15.30 R. Johnson, Universit{\`a} di Firenze, Italy: Nonautonomous control processes and topological dynamics G. H{\"a}ckl, Iowa State University, USA: Numerical computation of orbits N. Sri Namachchivaya, University of Illinois at Urbana-Champaign, USA; N. Malhotra, University of Illinois at Urbana-Champaign, USA: Global bifurcations in nonlinear systems with reversible symmetry W.P. Dayawansa, University of Maryland, USA; S.L. Samelson, SUNY - New Paltz, NY, USA: Asymptotic stabilization of systems with symmetries undergoing Hopf bifurcation *********************************************************************** Accuracy Issues in Eigenvalue Problems Organizers: J.L. Barlow, Pensylvania State University, University Park, USA K. Veselic, Fernuniversit{\"a}t Hagen, Germany Summary: In recent years, significant progress has been made in understanding how to solve eigenvalue problems more accurately. The speakers in this minisymposium will discuss some of this progress. The main application of this work is the development of robust software for eigenvalue problems. The efforts described here have had great influence on the eigenvalue codes in the LAPACK project. \smallskip There are two important thrusts to these efforts. The first is to find clases of matrices which are "well behaved". These are matrices for which structured relative changes in the entries of the matrices cause only relative changes in the eigenvalues. We also expect that the invariant subspaces associated with clusters of eigenvalues will be very stable. The second thrust is to find algorithms that compute the eigenvalues and eigenvectors to the expected accuracy. \smallskip This minisymposium will demonstrate that efforts of solve eigenvalue problems as accurately as possible are progressing well, but significant open questions remain. Day: July 7, 1995 Time: 09.30 J.L. Barlow, The Pensylvania State University, USA: Accuracy issues in eigenvalue problems I.C.F. Ipsen, North Carolina State University, USA; S.C. Eisenstat, Yale University, New Haven, CT, USA: Relative perturbation results for non-Hermitian eigenvalue problems Z. Drmac, Fernuniversit{\"a}t Hagen, Germany; I. Slapni\^car, University of Split, Croatia; K. Veselic, Universit{\"a}t Hagen, Germany: Generalised eigensolutions by Jacobi methods R.C. Mathias, College of William and Mary, USA: Fast accurate methods for eigenvalues and singular values M. Gu, University of California, USA; J.W. Demmel, D. Inderjit, University of California, Berkeley, USA: Solving eigenvalue problems using divide-and-conquer and inverse iteration techniques V.K. Fernando, Numerical Algorithm Group, United Kingdom; B.N. Parlett, University of California, Berkeley, USA: Orthogonal eigenvectors with clustered eigenvalues *********************************************************************** Automatic Differentiation and New Approaches in Large-Scale Optimization Organizer: C. Bischof, Argonne National Laboratory, USA Summary: Large-Scale optimization techniques are being used in multidisciplinary design optimization and inverse problems, for example, driven by the need to shorten the engineering design cycle. As a result, complex simulation codes are being embedded as objective functions or constraints in numerical optimization frameworks, which typically require first- or second-order derivative information about these "functions". Automatic differentiation (AD) is a technique, which allows one to transform computer models written in languages such as Fortran or C into codes for the computation of derivatives, based on the fact that every computation, as complicated as it may be, is at its heart only a (potentially very long) sequence of elementary arithmetic operations. Employing advanced compiler technology or novel language features such as operator overloading, automatic differentiation can be implemented efficiently to produce tools which can deal with arbitrarily complex codes. Unlike divided differences (DD), AD does not suffer from truncation error, and delivers derivatives which are accurate to machine precision. Experience with recent tools has also shown that derivative codes generated via AD can outperform DD by considerable margins. \smallskip The talks in this minisymposium illustrate the impact of AD on large-scale optimization. Jorge More's talk shows how AD impacts the development of software for the solution of large-scale optimization problems. John Dennis' talk illustrates the challenges in the context of multidisciplinary optimization. Andreas Griewank's talk shows how information available via AD can improve numerical optimization approaches for conjugate-gradient based matrix-free methods and nonsmooth problems, and Peter Eberhard presents an environment for the optimization of complex multibody systems. Day: July 3, 1995 Time: 15.30 J.J. Mor{\'e}, Argonne National Laboratory, USA; A. Bouaricha, Argonne National Laboratory, USA: Enviroments for large-scale optimization J.E. Dennis, Rice University, USA: Derivatives for multidisciplinary optimization A. Griewank, Technische Universit{\"a}t Dresden, Germany: Using curvature information in large-scale optimization P. Eberhard, Universit{\"a}t Stuttgart, Germany: Analysis and optimization of complex multibody systems using advanced sensitivity analysis methods *********************************************************************** Tools for the Automatic Differentiation of Computer Programs Organizer: C. Bischof, Argonne National Laboratory, USA Summary: The computation of derivatives plays a central role in numerical computing, for example in numerical optimization or the solution of differential equations. Derivatives are also used in the sensitivity analysis for the validation of computer models. Automatic differentiation (AD) is an approach for augmenting computer codes with statements for the computation of derivatives. AD is applicable to codes of arbitrary length, and can compute derivatives of arbitrary order that are exact up to machine precision. The elementary ideas of AD have been around for over 30 years, but recently, the need to obtain derivatives of "functions" defined by large simulation codes for such purposes as optimal simulation-based design or data assimilation techniques has spurred the development of AD tools. AD techniques are at the heart of a new field which we call "computational differentiation", which, in addition to the essentially "black-box" functionality of AD tools concerns itself with the the development of strategies that exploit user knowlegde about code structures, and the interactions between AD and numerical paradigms. \smallskip This minisymposium tries to provide an introduction to the basic AD principles, and provide an overview of publicly available AD tools for languages such as Fortran77, Fortran90, C or C++. In an half-hour talk Andreas Griewank, will provide an introduction to AD, and point to some recent developments in computational differentiation. Thereafter, developers of AD tools will highlight the functionality and salient features of their tools in short presentations. A technical report summarizing the session as well as a web page will be made available as well. Day: July 5, 1995 Time: 15.30 A. Griewank, Technische Universit{\"a}t Dresden, Germany: An introduction to automatic differentiation software M. Berz, Michigan State University, USA: to be announced S.A. Brown, University of Hertfordshire, England: to be announced A. Carle, Rice University, USA: to be announced D.M. Gay, AT{\&}T Bell Laboratories, USA: to be announced Day: July 7, 1995 Time: 09.30 R. Giering, Max-Planck Institut f{\"u}r Meteorologie, Germany: Automatic adjoint code generation K. Kubota, Chuo University, Japan: A Fortran precompiler for automatic differentiation and estimates of rounding error J. Reid, Rutherford Appleton Laboratory, United Kingdom: to be announced N. Rostaing-Schmidt, INRIA, France: to be announced J. Utke, Technische Universit{\"a}t Dresden, Germany: to be announced *********************************************************************** Domain Decomposition Methods and Their Applications in Fluid and Structural Mechanics Organizers: X.-C. Cai, University of Colorado at Boulder, USA B. Smith, Argonne National Laboratory, USA Summary: Domain decomposition is a very powerful technique for the parallel numerical solution of partial differential equations. In additional to its mathematically provable optimal convergence properties, it also has a wide range of applications. In this minisymposium, we present some new development in domain decomposition methods and the focus is on some important applications in the areas of computational fluid dynamics and structural mechanics. Day: July 6, 1995 Time: 15.30 R.G. Melvin, Boeing Computer Services, USA; C.C. Ashcraft, M.B. Bieterman, C.L. Hilmes, W.P. Huffman, F.T. Johnson, D.P. Young, The Boeing Company, Seattle; D.A. Keyes, Old Dominion: A two level acceleration scheme applied to full potential flow computations T. Mathew, University of Wyoming, USA: Domain decomposition methods for convection dominated elliptic problems and parabolic problems C. Farhat, University of Colorado at Boulder, USA; Po-Shu Chen, University of Colorado at Boulder, USA: Towards the ultimate domain decomposition method for solid and structural mechanics M.D. Tidriri, NASA Langely Research Center, USA; D.E. Keyes, Old Dominion University, Norfolk, USA: Newton-Krylov-Schwarz methods for compressible flows D. Knoll, Idaho National Engineering Laboratory, USA: Domain decomposition methods for systems of convection-diffusion-reaction equations *********************************************************************** Iterative Methods for Large Linear Systems of Equations Organizers: D. Calvetti, Stevens Institute of Technology, Hoboken, USA L. Reichel, Kent State University, USA Summary: Many problems in science and engineering require the solution of large linear systems of equations. The storage requirements of direct methods and recent developments in computer architecture make it attractive to apply iterative solution methods for a large number of problems. The development and analysis of iterative methods for the solution of symmetric and nonsymmetric linear systems continues to be an active area of research. It is the purpose of this minisymposium to present an overview of recent developments in iterative methods. Day: July 3, 1995 Time: 10.00 M.H. Gutknecht, ETH Z{\"u}rich, Switzerland: Recent progress on Lanczos-type methods T. Chan, University of California, USA: The composite step family of nonsymmetric conjugate gradient methods L. Reichel, Kent State University, USA: Adaptive Richardson iteration for nonsymmetric linear systems H.F. Walker, Utah State University, USA; P. N. Brown, Lawrence Livermoore National Laboratory, USA: GMRES on (nearly) singular systems Day: July 7, 1995 Time: 15.30 D. Calvetti, Stevens Institute of Technology, USA: Iterative methods for ill-conditioned linear systems R. Freund, AT{\&}T Bell Laboratories, USA: Iterative solution of symmetric indefinite linear systems arising in large-scale optimization problems B. Fischer, Medizinische Universit{\"a}t L{\"u}beck, Germany: On solving the Stokes problem by polynomial iteration methods H. Elman, University of Maryland, USA: Effects of Reynolds number on convergence of iterative methods for the Navier-Stokes equations *********************************************************************** New Developments in the Mathematical Theory of Capillarity Organizers: P. Concus, University of California at Berkeley, USA R. Finn, Stanford University, USA Summary: Following a quietus of over 50 years, the theory of capillarity has sprung to life during the last two decades, with new mathematical developments inspired by applications in space technology, physical and biological sciences, and engineering. These developments are using new mathematical techniques deriving largely from the achievements of BV theory, geometric measure theory, and the (associated) modern calculus of variations. \smallskip The minisymposium will present a spectrum of original modern mathematical contributions on a common range of topics, by mathematicians with varying backgrounds and points of view. Day: July 6, 1995 Time: 09.30 C. Bandle, Universit{\"a}t Basel, Switzerland: Isoperimetric inequalities for the capillary problem M.-F. Bidaut-Veron, Universit{\'e} de Tours, France: Hypersurfaces with prescribed mean curvature in Euclidean or Minkowsky spaces E. Miersemann, Universit{\"a}t Leipzig, Germany: A new numerical method for solving the capillary tube problem T. Suzuki, Ehime University, Japan; K. Narukawa, Naruto University of Education, Japan: Oscillatory theorem and pendent liquid drops D. Siegel, University of Waterloo, Canada; K.E. Lancaster, Wichita State University, Kansas, USA: Bahaviour of discontinuous capillary surfaces at a corner *********************************************************************** The Eigenvalue and Inverse Eigenvalue Problems for Second Order Systems Organizers: B. Datta, Northern Illinois University, DeKalb, USA D. Inman, Virginia Polytechnic Institute, Blacksburg, USA Summary: The systems of second order differential equations arise in a wide variety of practical applications, the most important one being the vibration of structures. \smallskip The associated eigenvalue problems here are generalized and quadratic eigenvalue problems. though the problems are classical, because of practical interests, an active research is still being pursued in all areas: theory, methods, and applications. Two emerging of current research are development of nonmodal approaches for stability and feedback stabilization, and constructions of mass and stiffness matrices using the knowledge of partial spectrum. The speakers of this session, who are well-known experts in vibration engineering, computational and applied mathematics, will speak on the latest developments both on direct and inverse problems, from theoretical, engineering, and computational viewpoints. The topic is interdisciplinary, blending applied mathematics, engineering, and computational mathematics. The session, thus, will be very appropriate for ICIAM. The talks will be of interest to applied mathematicians, computational scientists, and research and practicing engineers. Day: July 3, 1995 Time: 15.30 G.M.L. Gladwell, University of Waterloo, Canada; A. Morassi, Universit{\'a} deglie Studi di Udine, Italy: Isospectral rods D.J. Inman, Virginia Polytechnic Institute, USA; A. Kress, Virginia Polytechnic Institute, Blacksburg, USA: Eigenstructure assignment for 2nd order systems via inverse methods B.N. Parlett, University of California, USA: Frequency response plots: a new approach K. Veselic, Fernuniversit{\"a}t Hagen, Germany: Passive control of linear systems Day: July 4, 1995 Time: 09.30 A. Guran, The Catholic Univerity of America, USA: to be announced Y.M. Ram, University of Adelaide, Australia: The construction of a symmetric tridiagonal quadratic pencil from spectral data B.N. Datta, Northern Illinois University, USA; E.K. Chu, Monash University, Australia: Robust pole assignment for second-order systems A. Ruhe, Chalmers University of Technology, Sweden: Computing nonlinear eigenvalues by spectral transformation Arnoldi *********************************************************************** Theoretical Models for Advanced Semiconductor Devices Organizer: P. Hagan, Los Alamos National Laboratory, USA Summary: There are two main approaches to predicting the electrical behavior of advanced semiconducting devices. The first is using reduced dimensionality models, such as the drift-diffusion or hydrodynamic models. Although these models can be solved rapidly on workstations, their predictions have proven to be unreliable for the generation of advanced devices currently being developed. The second is using Monte Carlo methods to simulate the much more fundamental Boltzmann equation for electrons in semiconductors. However, these simulation are exceedingly slow, often taking tens or hundreds of hours on modern supercomputers to simulate a single device. \smallskip We address these problems by using singular perturbation techniques to analyze the Boltzmann equation. From this analysis we derive a new set of reduced dimensionality models which remain accurate in the physical regimes appropriate to the new devices. We use similar techniques to analyze the underlying Schr\"odinger equation, and from this analysis we derive a Boltzmann equation which correctly accounts for the quantum nonlocality of the electrons. Day: July 4, 1995 Time: 09.30 P.S. Hagan, Los Alamos National Laboratory, USA: Asymptotically accurate models for advanced semiconducting devices J.R. Sobehart, Los Alamos National Laboratory, USA: The high-field semiconductor equations L.G. Reyna, IBM-Watson Research Center, USA: The energy group drift-diffusion equations D.E. Woodward, Southern Methodist University, USA: Electron nonlocality in semiconductors *********************************************************************** Solution Structure in Multi-Dimensional Riemann Problems Organizers: S. Canic, Iowa State University, Ames, USA W.B. Lindquist, SUNY - Stony Brook, USA Summary: From the classic work of the late 1950's on scalar non-linear hyperbolic equations to intense work on 2x2 systems of conservation laws over the last decade, the study of self-similar solutions has provided tremendous insight into the structure of solutions to hyperbolic equations in one space dimension. Continued interest in physical problems in such areas as gas dynamics, material properties, and flow in porous media has prompted similar active investigation of the solution structure in multiple spatial dimensioned conservation laws. As in one space dimension, the major tool employed is the study of Riemann problem solutions. This minisymposium will provide a view of the current directions and methods being developed to elucidate and study these structures. The session will touch upon numerical algorithms for exact solution of scalar equations, interesting singularities in 2x2 systems, spirals in the Euler equations, partially aligned systems, and large time behavior. Day: July 7, 1995 Time: 15.30 G.-Q. Chen, Northwestern University, USA: Some theoretical results on multi-dimensional problems B.L. Keyfitz, University of Houston, USA; S. Canic, Iowa State University; B.L. Keyfitz, University of Houston, , USA: Self-similar solutions of multi-dimensional conservation laws H. Nussenzveig Lopes, IMECC UNICAMP, Brazil; M.C. Lopes-Filho, IMECC-UNICAMP, Campinas, Brazil: Partially aligned systems and control theory N.H. Risebro, University of Oslo, Norway: An exact solution to an approximate 2D Riemann problem Y. Zheng, Indiana University, USA; T. Zhang, Academia Sinica, Beijing, P.R. of China: Spiral solutions of the 2D compressible Euler equations E. Zuazua, Universidad Complutense de Madrid, Spain; M. Escobedo, Universidad del Pais Vasco, Spain: Large time behavior for solutions of viscous conservation laws *********************************************************************** Valuation of Financial Instruments Organizers: P.S. Hagan, Los Alamos National Laboratory, USA L. G. Reyna, IBM-Watson Research Center, New York, USA Summary: Financial instruments, such as stocks, bonds, options, and derivatives are evaluated via mathematical models describing the underlying markets. Slightly more accurate methods of evaluating financial instruments can lead to dramatic trading advantages. Clearer insight into how market forces change the value of portfolios of these instruments lead to more adroit trading strategies. More accurate portfolio evaluations lead to better risk management and potentially to fairer taxation. \smallskip In this minisymposium we examine the range of mathematical models being used to evaluate instruments, portfolios, and manage risk. We also explore new models and modelling methods which may potentially lead to more accurate valuations and deeper insights into factors influencing the value of portfolios. Day: July 5, 1995 Time: 09.30 M. Avellaneda, New York University, USA: Transaction costs in derivative valuation L.G. Reyna, IBM-Watson Research Center, USA: Capital funding optimization G. Williams, Kalotay Associates, USA: Technology in finance: link between theory and practice C. Gomez, Los Alamos National Lab, USA: Applications of neural nets to interest rate derivatives *********************************************************************** Practical Implementation of Spectral Methods Organizer: A. Karageorghis, University of Cyprus, Nicosia, Cyprus Summary: The aim of the minisymposium is to bring together workers and present recent developments in the field of the numerical solution of partial differential equations with spectral methods. Emphasis will be given to the practical implementation of the methods for problems in computational fluid dynamics. In particular, spectral domain decomposition techniques for such problems in two- and tree-space dimensions will be discussed and their accurate and efficient implementation addressed. Day: July 5, 1995 Time: 15.30 C. Bernardi, Universit{\'e} de Paris VI, France; M. Azaiez, Universit{\'e} Paul Sabatier, Toulousr; M. Dauge, Universit{\'e} de Rennes; X. Maday, Universit{\'e} Pierre et Marie Curie, Paris, France: Spectral element methods for axisymmetric three-dimensional problems G. Karniadakis, Brown University, USA: Triangular spectral elements: algorithms and flow simulation M. Paprzycki, University of Texas of the Permian Pasin, USA; A. Karageorghis, University of Cyprus, Nicosia: High performance solution to the linear systems arising from the spectral decomposition methods T.N. Phillips, University of Wales, United Kingdom: to be announced J. Shen, Pennsylvania State University, USA: Efficient implementions of Spectral-Galerkin methods T. Tang, Simon Fraser University, Canada: Boundary layer resolving pseudospectral methods for singular perturbation problems *********************************************************************** A New Explicit Half-Range Expansion Technique: Construction, Justification, and Applications Organizer: M.M. Klosek, University of Wisconsin, Milwaukee, USA Summary: Recently, a new half-range expansion technique was developed to make possible, for the first time, explicit solutions to a class of transport problems in which external conditions can only be prescribed on half of the boundaries. In this minisymposium this explicit construction will be presented, along with its rigorous justification, and its application to solutions of problems from different areas of science: the classical Milne problem from neutron transport theory; Boltzmann equations with a differential (Fokker-Planck) scattering operator; problems in the area of activated rate processes; and problems in semiconductor physics. Day: July 3, 1995 Time: 10.00 P.S. Hagan, Los Alamos National Laboratory, USA: Half-range expansions and the solution of the generalized Milne problem P.T. McDonald, Ohio State University, USA: Zeta functions of Sturm-Liouville operators and rigorous justification of the explicit half range expansion technique M.M. Klosek, University of Wisconsin, USA: Level crossing problems for a class of diffusion processes J.R. Sobehart, Los Alamos National Laboratory, USA; P.S. Hagan, Los Alamos National Laboratory, USA: Determination of velocity overshoot in semiconductor devices via half-range resolution of the boundary layer *********************************************************************** Solution of Ocean Related Mathematical Problems Organizers: J.R. McLaughlin, Rensselaer Polytechnic Institute, Troy, USA M. Collins, Naval Research Laboratory, Washington D.C., USA Summary: The topics will cover both direct and inverse ocean related problems. They will include: wave propagation in a depth and range dependent ocean, solution of inverse problems where the seabottom is modeled as a viscoelastic medium, and wave propagation for acoustic, elastic, and poro-elastic media. Theoretical results as well as numerical solutions will be presented. Day: July 3, 1995 Time: 10.00 Y.Y. Lu, Rensselaer Polytechnic Institute, USA; J. McLaughlin, Rensselaer Polytechnic Institute, Troy, USA: The Riccati method for sound propagation in the ocean W.W. Symes, Rice University, USA; J.O. Blanch, Rice University, Houston, TX, USA: Viscoelastic inversion in layered and nonlayered media M.D. Collins, Naval Research Laboratory, USA; N.C. Makris, B.E. McDonald, Naval Research Laboratory, Washington, USA; W.L. Siegmann, Rensselaer Polytechnic Institute, Troy, USA: Nonseparable wave propagation in the ocean R. Gilbert, University of Delaware, USA: Direct and inverse acoustic problems in a shallow ocean *********************************************************************** High Performance Solution of Structured Linear Systems Organizer: M. Paprzycki, University of Texas, Odessa, USA Summary: Recent years can be characterized by a rapid growth of knowledge about high performance solution of structured linear systems (understood as: a collection of relatively small rectangular and/or triangular dense blocks that assemble a large sparse matrix). These systems result from various discretizations of ordinary and partial differential equations. Examples of such systems include, but are not limited to: block tridiagonal, block bidiagonal, almost block diagonal, arrowhead, banded linear systems. \smallskip The primary aim of the minisymposium is to summarize the state of knowledge about high performance solution to such systems, where high performance is understood very broadly (including high and RISC-based workstations, shared memory and distributed memory parallel computers as well as distributed multicomputers such as networks of workstations). Future research directions will be also outlined. Day: July 3, 1995 Time: 10.00 L. Brugnano, Universit{\`a} di Firenze, Italy; P. Amodio, Universit{\`a} di Bari, Italy: Stable parallel solvers for tridiagonal linear systems P. Amodio, Universit{\`a} di Bari, Italy; T. Politi, Universit{\`a} di Bari, Italy; M. Paprzycki, University of Texas, Odessa, USA: Survey of parallel methods for the solution of block bidiagonal systems G.L. Kraut, The University of Texas at Tyler, USA; I. Gladwell, Southern Methodist University of Texas at Dallas, USA: Preconditioned conjugate gradients for two-point boundary value problems with non-separated boundary conditions L. Johnsson, Thinking Machines Corp., USA: Solving structured linear systems on the connection machine A.H. Sameh, University of Minnesota, USA; M. Berry, University of Tennessee, W. Harrod, Cray Research Inc., USA: Parallel algorithms for structured sparse linear systems P. Stpiczynski, Maria Sklodowska Curie University, Poland; M. Paprzycki, University of Texas, Odessa, USA: Parallel solution of linear recurrence systems *********************************************************************** The Methods of Lines Organizers: S. Thompson, Radford University, USA W.E. Schiesser, Lehigh University, Bethlehem, USA Summary: The method of lines (MOL) has developed as a flexible and comprehensive methodolgy for the solution of mixed systems of ODEs/DAEs/PDEs. This development continues at a rapid pace, with emphasis on new ODE/DAE integration methods that can be applied to PDEs, particularly as implemented with new computer architectures. New and diverse MOL applications are also providing motivation for new MOL developments and enhancements. \smallskip This minisymposium covers recent MOL developments in five areas: \item{(a)} ODE/DAE integration algorithms and codes with special emphasis on applications to the MOL. \item{(b)} Spatial differentiation algorithms and routines, including adaptive grids. \item{(c)} Parallel MOL implementations. \item{(d)} Combined numerical and symbolic computation applied to MOL implementations. \item{(e)} MOL applications in science engineering. Day: July 4, 1995 Time: 15.30 J.M. Hyman, Los Alamos National Lab, USA: Adaptive mesh methods for partial differential equations G.D. Byrne, Illinois Institute of Technology, USA: Storage reduction techniques for time-dependent kinetics transport models W.F. Lawkins, Oak Ridge National Laboratory, USA: Parallel method of lines solutions on a distribut memory message passing machine S. Lee, Oak Ridge National Laboratory, USA: Krylov methods for initial-value problems in differential-algebraic equations E.G. Ng, Oak Ridge National Laboratory, USA; B. Peyton, Oak Ridge National Laboratory, TN, USA: Parallel sparse matrix factorizations L.R. Petzold, University of Minnesota, USA: Parallel solution and sensitivity analysis for large-scale differential-algebraic systems Day: July 5, 1995 Time: 09.30 J. Carroll, Dublin City University, Ireland; S. Stewart, Dublin City University, Ireland: A comparison of some adaptive space mesh solvers for the numerical solution of parabolic partial differential equations P. Saucez, Faculte Polytechnique de Mons, Belgium; A. Vande Wouwer, Laboratoire d'Automatique, Mons, Belgium W.E. Schiesser, Lehigh University, Bethlehem, PA, USA: An implementation of a simple adaptive gridding procedure within the method of lines R. Allen, University of Canterbury, New Zealand; J.H.L. Thompson, Aluminium Smelters Ltd., Invercargill, New Zealand: Modelling the startup of the aluminium reduction cell S.L. Campbell, North Carolina State University, USA; W. Marszalek, North Carolina State University, USA: Recent developments in ODE/DAE integrators and MOL problems I. Gladwell, Southern Methodist University, USA; M.V. Melander, Southern Methodist University, Dallas, TX, USA: Using RKSUITE in the method of lines T.R. Taha, University of Georgia, USA; W.E. Schiesser Lehigh University, Bethlehem, PA, USA: Numerical simulation of a family of fully nonlinear dispersive KdV-like equations Day: July 5, 1995 Time: 15.30 H.J. Halin, Swiss Federal Institute of Technology, Switzerland: The numerical method of lines revisted: finite difference approximations - do we need them A. Papadourakis, Rohm and Haas Company Bristol, USA: Dynamic response of storage tanks exposed to fire: modeling and parametric investigation A. Gerstlauer, Universit{\"a}t Stuttgart, Germany; S. R{\"a}umsch{\"u}ssel, S. Sviatnyi, R. Zimmermann, E.D. Gilles, M. Zeitz, Universit{\"a}t Stuttgart, Germany; W. Marquardt, RWTH Aachen, Germany: Symbolic preprocessing of systems of partial differential equations S.C. Chu, Radford University, USA; S. Thompson, Radford University, VA, USA; W.F. Lawkins, Oak Ridge National Laboratory, TN, USA; E. Ng, Oak Ridge National Laboratory, TN, USA: Parallel method of lines solutions of partial differential equations H. Haario, University of Helsinki, Finland; P. Neittaanmaki, University of Jyvaskyla, Finland; V. Rivkind, University of Jyvaskyla, Findland: Numerical solution methods for sharp fronts *********************************************************************** Mathematical Problems in Transonic Flow Organizer: D.W. Schwendeman, Rensselaer Polytechnic Institute, Troy, USA Summary: Transonic flows are characterized by having flow speeds at or near sonic. Such flows are of physical interest because they lie at the border between subsonic and supersonic flow and because such flows are typically very sensitive to perturbations in the flow conditions. Transonic flows are interesting mathematically because they are governed by nonlinear equations. Perturbations methods and other analytical methods are often used to reduce the mathematical problems to a simpler form to be handled ultimately by numerical methods. The speakers with discuss various topics involving transonic flows, both steady and unsteady, including the design of transonic airfoils and the formation and propagation of weak shocks. Day: July 7, 1995 Time: 09.30 P.L. Cook, University of Delaware, USA; J. Cole, Rennselaer Polytechnic Institute, USA; G. Schleiniger, University of Delware, Newmark, USA: Unsteady transonic flows D.W. Schwendeman, Rensselaer Polytechnic Institute, USA; M.C. Kropinski, Courant Institute of Mathematical Sciences, USA; J. Cole, Rensselaer Polytechnic Institute, USA: An analytical and numerical study of optimal critical airfoils E.G. Tabak, New York University, USA; R.R. Rosales, M.I.T., Cambridge, USA: Focusing of weak shock waves, nonlinear caustics, and the von Neumann paradox B.L. Keyfitz, University of Houston, USA: Admissibility criteria for transonic shocks R.R. Rosales, Massachusetts Institute of Technology, USA; C.M. Celentano, M.I.T., Cambridge, USA: Large amplitude nonlinear resonant acoustic waves without shocks *********************************************************************** Theoretical Immunology Organizers: L. Segel, Weizmann Institute, Rehovot, Israel R. de Boer, Utrecht, The Netherlands Summary: The immune system is of major importance in mammalian physiology, deriving considerable extra interest because immunological defects bring about disease. Theoretical research in immunology is relatively new and thus demands the full gamut of techniques from imaginative modeling of components of this highly complex system through analytical and numerical treatment of model equations to model interpretations that can be understood by biologist and physicians. \smallskip This minisymposium presents a selection of recent modeling efforts whose approaches range from exploitation of analogies with neural networks, through uses of probability and control theory to model affinity maturation, to a "super-phenomenological" approach to characterizing auto-immune disease. Day: July 6, 1995 Time: 09.30 A. McLean, University of Oxford, United Kingdom: Competition as a regulatory force in the immune system: experimental and mathematical models L.A. Segel, The Weizmann Institute of Science, Israel: Modelling auto-immune disease as a problem in applied mathematics R. de Boer, Theoretical Biology UU, The Netherlands: New functions for immune networks T.B. Kepler, North Carolina State University, USA; A.S. Perelson, Los Alamos National Laboratory, NM, USA: Modeling and optimization of populations subject to time-dependent mutation *********************************************************************** Estimation and Modeling of Nonstationary Stochastic Processes Organizer: K.S. Riedel, New York University, USA Summary: Many physical phenomena are approximately stationary in the sense that the system parameter are slowly varying with respect to the sampling rate. In some cases, the process can be modelled as a generalized ARMA process with temporally varying coefficients. In other cases, a fully nonparametric approach is necessary with a more general covariance structure is necessary. The most common of these models is the evolutionary spectrum {\it a la} Priestley. For these nonparametric models, the representation is generally not unique unless additive conditions are imposed. This minisymposium will examine both nonparametric and semiparametric models of the spectrum. Nonparametric models will be formulated in terms of the Weyl correspondence and reproducing kernel Hilbert spaces. Newer multiresolution models will also be examined. Day: July 3, 1995 Time: 15.30 P. Flandrin, CNRS-ENS Lyon, France; R.G. Baraniuk, Rice University, Houston, TX, USA; O. Michel, ENS Lyon, France: Information and distance measures for time-varying spectra W. Kozek, Universit{\"a}t Wien, Austria: On the underspread/overspread classification on nonstationary Random processes R. von Sachs, Universit{\"a}t Kaiserslautern, Germany: Nonlinear wavelet methods for time-dependent spectra of locally stationary time series K.S. Riedel, New York University, USA: Evolutionary spectra and WKBJ approximation P.-O. Amblard, CEPHAG, France: Higher-order statistics for nonstationary signals T. Subbarao, University of Manchester, United Kingdom: Analysis of nonstatioary and nonlinear time series *********************************************************************** Piecewise Convex Function Estimation Organizer: K.S. Riedel, New York University, USA Summary: In many applications, one needs to estimate an unknown function given noisy measurements. Traditional methods include smoothing splines and nonparametric kernel smoothing. Both classes of methods tend to either oversmooth the peaks in the unknown function or to have artificial wiggles. To eliminate these spurious relative extrema, a new class of methods have recently been introduced where the fit is constrained to have a small number of changes from convex to concave. In general, the number and locations of these change points are unknown and need to be estimated. Major research areas are model selecting convergence rates and improved numerical methods. Day: July 4, 1995 Time: 09.30 E. Mammen, Humboldt Universit{\"a}t Berlin, Germany: Estimation of functions with spatially inhomogeneous smoothness: an approached based on total variation penalties K.A.P.M. Willemans, University of Leuven, Belgium; P. Dierckx, Katholieke Universiteit Leuven, Belgium: Piecewise convex surface fitting M.B. M{\"a}chler, ETH Zentrum, Switzerland: Very smooth curve estimation via semiparametric penalty K.S. Riedel, New York University, USA: Duality, fast optimization, and model selection *********************************************************************** Orthogonal Polynomials and Spectral Methods Organizers: M.E. Muldoon, York University, North York, Canada A. Ronveaux, Facultes Universitaires, Namur, Belgium F. Marcellan, EPS Carlos III, Madrid, Spain Summary: There has been much recent research on orthogonality with respect to Sobolev inner products. The analysis of spectral problems for ordinary differential operators, the asymptotic estimates for the corresponding eigenfunctions as well as the location and properties of their zeros constitute some new directions in this research. These methods can be used in the design of pseudospectral methods for the discretization of parabolic partial differential equations as well as for the design of efficient algorithms for approximation by polynomials in the underlying Sobolev space. Day: July 4, 1995 Time: 15.30 Y. Maday, Universite Paris VI, France: Spectral methods for solving axisymmetric partial differential equations A. Iserles, University of Cambridge, England: Sobolev-orthogonal polynomials and pseudospectral methods for parabolic PDE's W. van Assche, Katholieke Universiteit Leuven, Belgium: Direct and inverse problems for Sobolev orthogonal polynomials M. de Bruin, Delft Univiversity of Technology, The Netherlands: Sobolev-orthogonal polynomials and their zeros *********************************************************************** Nonlinear Waves in Physiology Organizer: J. Sneyd, University of Canterbury, New Zealand Summary: Mathematical Biology is one of the fastest-growing areas in applied mathematics and has been the source of many interesting mathematical problems. Mathematical approaches are particularly valuable in the study of nonlinear spatio-temporal phemonema which are intractable to intuitive approaches. \smallskip Waves are observed in many areas of physiology. Some of the earliest work in the field was that of Hodgkin and Huxley who formulated a model for the propagation of an action potential in a nerve axon. More recently, there has been a great deal of work done on electrical wave propagation in cardiac tissue, and on the propagation of calcium waves. Since much of this work is based on reproducible data and measurable parameters, it provides the opportunity of making specific predictions which can contribute to experimental design and model improvement. \smallskip The minisymposium will present a variety of physiological problems, unified by the underlying wave behaviour. Speakers will include representatives from cardiology, calcium waves, acoustics, and applied mathematics. Day: July 5, 1995 Time: 09.30 J. Sneyd, University of Canterbury, New Zealand: Intracellular and intercellular calcium waves A.V. Panfilov, Utrecht University, The Netherlands: Modeling of re-entrant patterns in an anatomical model of the heart R. Chadwick, NIH/BEIP, USA: Actively-controlled wave propagation in a three-partition model of the cochlea J.A. Sherratt, University of Warwick, United Kingdom: Wound healing waves in the cornea *********************************************************************** Wavelets and Signal Processing Organizers: G. Strang, MIT, Cambridge, Massachusetts, USA M. K. Kwong, Argonne National Laboratory, Illinois, USA Summary: The advent of the information superhighway makes it possible to produce, transmit, and archive high-resolution and high-quality images. Medical imaging, 3D virtual reality, telephony, and telepresence are other areas that require large-scale images. As the demand grows, the amount of data to be treated and stored becomes enormous; and efficient means of data compresion and signal processing are sought. There is no doubt that wavelet theory has had great impact on the field of signal and image processing. Wavelet techniques can be used in almost all areas of image processing, including image compression, denoising, enhancement, hierarchical motion estimation, video encoding, and pattern recognition. New wavelets, such as multiwavelets and biorthogonal spline wavelets, continue to be constructed and new applications investigated. The minisymposium will survey some of the new developments in wavelet theory, efficient software implementation of discrete wavelet transform, and specific applications in image processing. Day: July 5, 1995 Time: 15.30 G. Strang, MIT, USA: Matrix theory for filter banks and wavelets M.K. Kwong, Argonne National Laboratory, USA: $W$-transforms and image processing W. Sweldens, University of South Carolina, USA: Second generation wavelets and signal processing H.G. Feichtinger, Universit{\"a}t Wien, Austria: Efficient algorithms for coherent non-orthogonal expansions P. Flandrin, CNRS-ENS Lyon, France; P. Abry, E. Chassande-Mottin, ENS Lyon, France: Sharply localized time-frequency and time-scale energy distributions C.E. Heil, Georgia Institute of Technology, USA: Existence and uniqueness of matrix refinement equations *********************************************************************** Singularly-perturbed Dynamical Systems: Theory and Applications Organizers: B. Braaksma, Limburgs Universitair Centrum, Diepnbeek, Belgium A. Doelman, Utrecht University, The Netherlands T. Kaper, Boston University, USA Summary: Singularly-perturbed dynamical systems arise as models in a diverse array of physical, chemical, and biological processes. These include traveling waves in reaction-diffusion systems, slow manifolds in atmospheric dynamics, resonances in mechanical systems, nonlinear beam dynamics in lasers, and pulse-propagation in nonlinear optics, to name just a few. \smallskip The mathematical techniques currently employed to study these singular perturbation problems vary widely. They range from the matched asymptotic expansion method to singularity analysis, as well as from dynamical systems approaches to numerical bifurcation studies and nonstandard analysis. \smallskip This pair of minisymposia presents a sampling of some of these mathematical techniques and the results they yield in the applications. Day: July 4, 1995 Time: 15.30 C.K.R.T. Jones, Brown University, USA: Geometric singular perturbation theory and multiple jump orbits B. Sandstede, WIAS Berlin, Germany: Bifurcations of homoclinic solutions in the FitzHugh-Nagumo system K. Schneider, WIAS, Germany; N.N. Nedlov, Lomonosov Moscow State University, Russia; A. Schnuppert, Hoechst AG, Frankfurt, Germany: Jumping behaviour of the reaction rate of fast bimolecular reactions Y. Kuznetsov, Vrije Universiteit Amsterdam, The Netherlands: Chaos in slow-fast 3D food chain Day: July 5, 1995 Time: 09.30 F. Dumortier, Limburgs Universitair Centrum, Belgium: Global blow-up and center manifolds E. Benoit, Universit{\'e} de La Rochelle, France: Asymptotic expansions of canards with poles. Application to the stationary unidimensional Schr{\"o}dinger equation B. Braaksma, Limburgs Universitair Centrum, Belgium: Canards and saddle-loop connections G. Mar{\'e}e, Landbouwuniversiteit Wageningen, The Netherlands: Periodic and state-dependent passage through pitchfork bifurcations T. Erneux, Vrije Universiteit Brussel, Belgium; D. Pieroux, Universit{\'e} Libre de Bruxelles, Belgium; V. Kovanis, A. Gavrielides, P.M. Alsing, Philipps Laboratory, USA: Mechanisms for period doubling bifurcation in a semiconductor laser Day: July 6, 1995 Time: 15.30 W. Eckhaus, Utrecht University, The Netherlands: to be announced R. Camassa, Los Alamos National Laboratory, USA: On the geometry of an atmospheric slow manifold A.R. Champneys, University of Bristol, United Kingdom: to be announced G. Kovacic, Rensselaer Polytechnic Institute, USA; T. A. Wettergren, Rensselaer Polytechnic Institute, Troy, USA: Homoclinic orbits in the dynamics of resonantly driven coupled pendula *********************************************************************** Dynamics of Periodic Structures Organizers: R. Bogacz, IPPT Warsaw, Poland K. Popp, Universit{\"a}t Hannover, Germany Summary: Periodic structures often occur in physics and engineering. These structures behave like bandpass filters. If energy dissipation is omitted, there is a sharp destinction between frequency bands exhibiting wave propagation without attenuating (passing bands) and those with attenuation and no propagation (stopping bands). The theory of wave propagation in periodic structures is well developed, cf. the classical book by Brillouin (1946). However, the results seem to be forgotten. Thus, there is a strong need to remind to old results, to show new developments and to demonstrate industrial applications. Attention will be paid to the influence of damping, to cyclic and finite periodic structures, to nonlinear as well as disordered periodic structures. The latter can lead to localization effects which are of great practical importance. Applications are shown mostly for mechanical structures, e.g. stiffened plates, periodic laminates, Maglev guideways, railway tracks, turbine disk assemblies. Day: July 6, 1995 Time: 15.30 D.J. Mead, University of Southampton, United Kingdom: Wave propagation in continuous periodic structures - historical background and recent developments - A.N. Guz, Ukrainian Academy of Sciences, Ukraine: Elastic waves in laminated periodic bodies with initial (residual) stresses A. Vakakis, University of Illinois at Urbana-Champaign, USA: Spatially localized oscillations in nonlinear periodic structures A.V. Metrikin, Russian Academy of Sciences, Russia; A.I. Vesnitsky, Russian Academy of Sciences, Russia: Transitional and steady-state vibrations of periodic and quasi-periodic structures, interacting with moving objects T. Krzyzynski, Polish Academy of Sciences Warszawa, Poland; R. Bogacz, Polish Academy of Sciences Warszawa, Poland; K. Popp, Universit{\"a}t Hannover, Germany: On the dynamics of cyclic and linear extended periodic structures with mistuning R. Bogacz, Polish Academy of Sciences, IPPT, Poland; K. Popp, University of Hannover, Germany: On the dynamics of cyclic and linear extended periodic structures with mistuning *********************************************************************** Numerical Methods for Some 3-D Problems of Mathematical Physics Organizer: G. Kobelkov, Moscow State University, Russia Summary: The present symposium is devoted to implementation of finite difference methods for solving difficult problems of mathematical physics, such as Navier-Stokes equations for incompressible and compressible viscous flow, elasticity theory eq uations with cracks, etc. All methods have been investigated from mathematical point if view (approximation, stability, convergence, etc.) and illustrated by numerical examples. In particular, it was managed to prove slight stability for finite differ ence scheme for compressible viscous flow, and solve the problem of asymptotic behavior of the 3-D elasticity theory equations solution at flat corner crack, what is quite important for applications. Day: July 6, 1995 Time: 09.30 G. Kobelkov, Moscow State University, Russia: Numerical methods for incompressible Navier-Stokes equation with high Reynolds number A. Sokolov, Moscow State University, Russia: On finite difference schemes for compressible Navier-Stokes equations V. Staroverov, Moscow State University, Russia: Numerical simulation of a solution of 3-D elasticity theory equations at flat corner crack N. Ardelian, Moscow State University, Russia: Iterative solution methods for the implicit difference schemes of MHD *********************************************************************** Numerical Treatment of Eigenvalue Problems Organizer: F. Goerisch, Technische Universit{\"a}t Braunschweig, Germany Summary: Day: July 3, 1995 Time: 10.00 W.N. Everitt, University of Birmingham, United Kingdom: Computing eigenvalues of singular Sturm-Liouville problems S.M. Rump, Technische Universit{\"a}t Hamburg-Harburg, Germany: Lower bounds to the smallest singular value of large matrices M. Plum, Technische Universit{\"a}t Clausthal, Germany: Enclosures for eigenvalues and the essential spectrum of the Orr-Sommerfeld equation H. Behnke, Technische Universit{\"a}t Clausthal, Germany: Inclusions for eigenvalue curves of parameter-dependent eigenvalue problems Z. Woznicki, Institute of Atomic Energy, Poland: The Garloff's conjecture is true F. Goerisch, Technische Universit{\"a}t Braunschweig, Germany: Bounds to eigenvalues of operators of the form $T^\ast T$ *********************************************************************** Nonlinear Modelling of Innovation Diffusion Organizer: H.K. Babovsky, WIAS Berlin, Germany Summary: The mathematical modelling of real-world phenomena requires a deep interaction of application sciences with mathematic analysis and numerics. One of the fields of nonlinear modelling emerging in recent years is that of innovation diffusion in socio-technical systems intending to model technological changes, to understand influence parameters and to optimally control these systems. \smallskip The minisymposium addresses various mathematical problems related to nonlinear modelling of innovation diffusion which are most relevant for a full understanding. It is intended as a link between theory-oriented mathematical work and a specific application. Some of the keywords are: Stability of nonlinear stochastic time-delay systems, synergetics, pathwise numerics of stochastic differential equations. Day: July 7, 1995 Time: 09.30 A. Wunderlin, Universit{\"a}t Stuttgart, Germany: Application of synergetics to innovation diffusion . Karmeshu, Jawaharlal Nehru University, India: Stochastic modelling of innovation diffusion M. Scheutzow, TU Berlin, Germany: Stochastic systems with time delay H. Schurz, WIAS Berlin, Germany: Numerical analysis of stochastic innovation diffusions *********************************************************************** Mathematical Topics in Combustion Organizers: H. Berestycki, Universit{\'e} Paris VI, France B. Matkowsky, Northwestern University, Evanston, USA Summary: Combustion waves are either supersonic (detonations) or decidedly subsonic (deflagrations or flames). This minisymposium focuses on the subsonic regime, considering mathematical modeling, analysis and computations for a variety of problems describing the behavior of flames, including flame structure, flame dynamics, flame patterns and flame instabilities. Day: July 4, 1995 Time: 15.30 J.M. Vega, Universidad Polit{\'e}cnica de Madrid, Spain; C. Martel, Universidad Polit{\'e}cnica de Madrid, Spain: The oscillatory instability of flames V.A. Volpert, Northwestern University, USA; A.P. Aldushin, Institute of Structural Macrokinetics, Chernogolovka, Russia; B.J. Matkowsky, Northwestern University, Evanston, IL, USA: Stoichiometric flames and their stability S. Kamin, Tel Aviv University, Israel; H. Berestycki, Universite Pierre et Marie Curie 4, Paris, France; G.I. Sivashinsky, Tel Aviv University, Ramat Aviv, Israel: Nonlinear dynamics and metastability in a Burgers type equation (for upward propagating flames) G. Joulin, URA. 193 au CNRS -ENSMA, France: Recent advances and open problems in the nonlinear theory of wrinkled premixed flames Day: July 6, 1995 Time: 15.30 A. Bonnet, University of Cergy-Pontoise, France; H. Berestycki, B. Larrouturou: The mathematical modelling of planar flames with complex chemistry J.-M. Roquejoffre, Ecole Polytechnique, France; L. Glangetas, Universit{\'e} de Paris IV, France: Bifurcations of travelling waves in the thermo-diffusive model for flame propagation D. Meink{\"o}hn, DLR, Germany: Phase transition concepts for heterogeneous combustion J.W. Dold, University of Bristol, United Kingdom: Diffusive and anti-diffusive motion of a flame front *********************************************************************** Vortex Sheet Initial Data for the Incompressible $2-D$ Euler Equations Organizers: H.J. Nussenzveig Lopes, IMECC UNICAMP, Campinas, Brazil Y. Zheng, Indiana University, Bloomington, USA Summary: Our concern is the rigorous analytical treatment of incompressible flows in 2-D with highly irregular initial data and related problems. The evolution of a vortex sheet trailing the edge of an airfoil in subsonic regime is an example. The problems addressed will be: qualitative behavior, specially the concentration of kinetic energy on small scales of the flow, and the existence of solutions. Researchers are developing a growing body of techniques from real analysis, which, together with the a-priori estimates, manage to control nonlinear functionals of weakly converging approximate solution sequences. One scientific concern here is understanding the onset of turbulence. Day: July 5, 1995 Time: 09.30 Y. Zheng, Indiana University, USA; A. Majda, Princeton University, Princeton, USA: Existence of global weak solutions to the one-component Vlasov-Poisson system with measures as initial data S. Wu, Northwestern University, USA; I. Vecchi, ETH-Zentrum, Z{\"u}rich, Switzerland: On $L^1$ vorticity for 2-D incompressible flow M. Bildhauer, Universit{\"a}t des Saarlandes, Germany: On the Hausdorff dimension of $n\times m$ concentration sets D. Chae, Seoul National University, Korea; N. Kim, POSTECH, Pohang, Corea: Axisymmetric weak solutions of 3-D Euler equations for incompresslble fluid flows Day: July 5, 1995 Time: 15.30 H. Nussenzveig Lopes, IMECC UNICAMP, Brazil; M.C. Lopes-Filho, IMECC - UNICAMP, Campinas, Brazil: Concentration sets for $2-D$ incompressible flow S. Schochet, Tel-Aviv University, Israel: Examples related to concentration-cancellation J.S. Soler, Universidad de Granada, Spain: On uniqueness of solutions for the $2D$ Euler equations *********************************************************************** F.E.M. and Numerical Analysis of Incompressible Navier-Stokes Equations Organizers: O. Pironneau, Universit{\'e} Pierre et Marie Curie, Paris, France V. Girault, Universit{\'e} Pierre et Marie Curie, Paris, France Summary: Four Navier-Stokes specialists will address different aspects of the numerical solution of flow problems. \smallskip Yves Achdou will present a domain decomposition method for flows around wing profiles, where the convection term is discretized along characteristics and the vorticity is computed by boundary elements. Brigitte M\'etivet will solve Navier-Stokes flows with varying density by a F.E.M., where the density and the velocity are discretized by polynomials of equal degree. Masahisa Tabata will discuss and compare two families of finite element approximations of axisymmetric Navier-Stokes flows. R\"udiger Verf\"urth will give {\it a posteriori} estimates for finite element approximations of steady or unsteady Navier-Stokes equations. Day: July 3, 1995 Time: 10.00 Y. Achdou, Ecole Polytechnique, France; O. Pironneau, Universit{\'e} Pierre et Marie Curie, Paris, France: A fast solver for Navier-Stokes equations using mortar finite element and boundary element methods B. M{\'e}tivet, Electricit{\'e} de France, France; C. Bernardi, Univ. Paris et M. Curie, France; B. Thomas, Electricit{\'e} de France, Clamart, France; X. Warin, Electricit{\'e} de France, Clamart, France: Solving Navier-Stokes equations with varying density M. Tabata, Hiroshima University, Japan: Finite element analysis of axisymmetric flow problems R. Verf{\"u}rth, Ruhr Universit{\"a}t Bochum, Germany: A posteriori error estimates for finite element approximations of stationary and non-stationary Navier-Stokes equations *********************************************************************** Computational Modelling of Multi-Physics Processes Organizer: M. Cross, University of Greenwich, London, United Kingdom Summary: Generally numerical techniques have been developed and used to model problems involving uncoupled physical phenomena such as solid mechanics, fluid flow, electromagnetics, surface contact forces, etc. Processes such as the casting of metals, electromagnetic melting and the structural response of buildings during a fire are truly multi-physics in nature where phenomena such as fluid flow and structural deformation are highly coupled. Modelling these types of problems involves developing novel and robust algorithms that can solve the resulting coupled non-linear discretised equations in an efficient manner. This symposium will address some of the current issues in this area of modelling. Day: July 3, 1995 Time: 15.30 M. Cross, University of Greenwich, United Kingdom; P. Chow, C. Bailey, N. Croft, J. Ewer, P. Leggett, K. Mc Manus, K.A. Pericleous, M. Patel, University of Greenwich, London, UK: PHYSICA - software environment for the modelling of multiphysics phenomena C. Bailey, University of Greenwich, United Kingdom; E. Galea, University of Greenwich, London, UK; A. Camroux, University of Greenwich, London, UK: Behaviour of building structures subjected to a thermal load from a fire K. Pericleous, University of Greenwich, United Kingdom; M. Hughes, University of Greenwich, London, UK; D. Cook, University of Greenwich, London, UK; M. Cross, University of Greenwich, London, UK: Computational magnetohydrodynamics - applications P. Chow, University of Greenwich, United Kingdom; C. Bailey, University of Greenwich, London, UK; M. Cross, University of Greenwich, London, UK; K. Pericleous, University of Greenwich, London, UK: The numerical modelling of multi-physics phenomena: the shape casting process K. Mc Manus, University of Greenwich, United Kingdom; M. Cross, University of Greenwich, London, UK: Unstructured mesh computational mechanics on DM parallel platforms *********************************************************************** Mathematical Methods in Plasma Physics Organizer: J.I. Diaz, Universidad Compultense de Madrid, Spain Summary: The magnetic confinement of a Plasma in Tokamaks and Stellarators geometries is a very riche source of new problems in Applied Mathematics since the sixties. Different models concerning the evolution problem, the equilibrium states and their stability will be considered for Tokamak and Stellarator devices. By using different arguments, the Magnetohydrodynamic system leads to a large family of nonlinear boundary value problems in partial differential equations including inverse type problems, free boundary problems, equations with non local terms, bifurcation problems and so on. Some statistics models will be also considered. Day: July 4, 1995 Time: 15.30 J. Blum, Universit{\'e} de Grenoble, France: Evolution of the plasma equilibrium in a Tokamak at the diffusion time scale J.M. Rakotoson, Universit{\'e} de Poitiers, France: Grad-Shafranov equations with nonlocal terms H. Tasso, Max-Planck-Institut f{\"u}r Plasmaphysik, Germany: Hamiltonians and statistics of continuous plasma models B. Saramito, Universit{\'e} Blaise Pascal, France: Stability and asymptotic states in plasma physics J.I. Diaz, Universidad Complutense de Madrid, Spain: Equilibrium two-dimensional models for helical axis stellarators *********************************************************************** Interactive Grid Generation: Methods and Tools Organizer: R.M. Spitaleri, IAC-CNR, Rome, Italy Summary: Any simulation technique that requires the solution of partial differential equations could have the critical problem of modelling geometrical complex solution domains. A wide spectrum of applications, from aircraft design to oceanography and automobile aerodynamics to meteorology, need powerful methods for generating grids. Although substantial progress has been made in grid generation advances have yet to be made in achieving new skills in the solution of problems. Main current directions of research deal with structured, unstructured and hybrid grid generation, adaptive techniques, multigrid algorithms. High level interactivity, driven by visualization, characterizes grid generation tools. The minisymposium will emphasize grid generation as a crucial point of the numerical modelling and present recent advances in development of methods and tools. Day: July 5, 1995 Time: 09.30 C. Dener, SAMTEK-ITC Ankara, Turkey: Recent advances in GEMS L. Formaggia, CRS 4, Cagliari, Italy: Three dimensional mesh generation by unstructured/hybrid grids V. Micacchi, IAC-CNR Rome, Italy; R.M. Spitaleri, CNR, Roma, Italy: Multiblock multigrid grid generation S. Paoletti, IBM ECSEC Rome, Italy: ENGAGE, a multi-block environment for grid generation *********************************************************************** New Mathematical Methods and Problems in Image Analysis Organizer: J.-M. Morel, Universite Paris-Dauphine, France Summary: Image analysis is an emerging discipline, still in course of formalization. Indeed, the sophistication of feasible methods has increased very fast with the growth of computing power. During the eighties, a stage has been attained where powerful mathematical tools in nonlinear analysis and geometry have been needed and developed. The minisymposium will present some samples of such developments, made by mathematicians and leading to rigorous theories and applications. Among the applicative themes which will be presented, we shall discuss image and movie restoration, shape recognition, texture analysis, image segmentation, with applications in industrial process control, seismology, criminology, etc... From the mathematical viewpoint, new variational methods, new partial differential equations and new nonlinear filtering and decomposition methods for signal and images will be introduced and mathematically discussed. Day: July 6, 1995 Time: 09.30 S. Osher, Cognitec Inc., USA; L. Rudin, Santa Monica, CA, USA: Applications of variational and PDE methods to real case solving in image and movie processing N. Saito, Schlumberger-Doll Research, USA; R.R. Coifman, Yale University, New Haven, CT, USA: Selection of best bases for classification and regression A. Bonnet, University of Cergy-Pontoise, France: The Mumford-Shah functional for image segmentation J. Weickert, Universit{\"a}t Kaiserslautern, Germany: Foundations and applications of nonlinear anisotropic diffusion filtering T. Cohignac, Universite Paris Dauphine, France: Invariant recognition of shapes and multiscale analysis F. Guichard, Universite Paris Dauphine, France; L. Moisan, Universite Paris Dauphine: Multiscale analysis of movies. Applications to depth recovery *********************************************************************** Boundary Element Methods: Mathematical Foundation Organizers: W.L. Wendland, Universit{\"a}t Stuttgart, Germany W. Hackbusch, Universit{\"a}t Kiel, Germany E.P. Stephan, Universit{\"a}t Hannover, Germany Summary: Although mathematical foundation and theory for boundary element methods is still in vivid development, many important fundamental issues led to a fast growing mathematical theory supporting these methods and providing new algorithmic developments. In this minisymposium on the mathematical foundation of boundary element methods a general overview of the current state of the art will be given, then new results on the efficient computation of the entries of the influence matrix will be presented. Since boundary element methods provide very efficient realizations of Poincar\'e--Steklov operators in domain decomposition, these methods led to highly efficient algorithms on massively parallelized computers, also in combination with finite element approximations. One of the most impressive developments concerns the data compression in combination with wavelet and wavelet--like approximations of the boundary integral equations involved. For two--dimensional problems, modified collocation and fully discrete Galerkin methods turned out to provide superconvergent approximation schemes. More recent developments in time--dependent boundary element methods open up a big variety of new future applications. Day: July 4, 1995 Time: 09.30 G.C. Hsiao, University of Delaware, USA: Applications of boundary element methods to problems in mechanics S.A. Sauter, Universit{\"a}t Kiel, Germany: On the efficient computation of singular and nearly singular surface integrals arising in 3D-BEM U. Langer, Johannes Kepler Universit{\"a}t Linz, Austria; M. Kuhn, Johannes Kepler Universit{\"a}t Linz, Austria: Parallel algorithms with boundary element methods C. Schwab, University of Maryland at Baltimore, USA: Wavelet-based BEM a-posteriori error estimation I.H. Sloan, University of New South Wales, Australia: Qualocation T. Ha-Duong, Universit{\'e} de Technologie de Compiegne, France: On boundary integral equations associated to scattering problems of transient waves *********************************************************************** Symbolic + Numeric = Scientific Computing Organizer: R.M. Corless, University of Western Ontario, London, Canada Summary: This minisymposium will explore the combination of numerical and symbolic computation in a modern scientific computing environment, by examples. A main sub-theme is "correct computation", so automatic error analyses and their use are expected to be part of each talk. \smallskip The topics of the talks range from the solution of DAE through proof techniques to the solution of polynomial systems; on the face of it, all more-or-less unrelated topics. However, the solutions of these problems are all facilitated by the use of computer algebra packages, interval techniques, and robust numerical techniques. It is the interplay between these that is the focus of this minisymposium. Day: July 7, 1995 Time: 15.30 S.M. Watt, IBM Thomas J. Watson Research Center, USA: System support for symbolic-numeric computation J.D. Pryce, Royal Military College of Science, United Kingdom; B. M. Brown, University of Wales, Cardiff, UK: Machine-verifiable proofs for numerical analysis R.M. Corless, University of Western Ontario, Canada; P.M. Gianni, B.M. Trager, S.M. Watt, IBM T.J. Watson Research Center, Yorktown Heights, NY, USA: Solving polynomial systems with inexactly known coefficients D.A. Cowey, Cranfield University, England: AD01 - a Fortran 90 automatic differentiation package, and its incorporation into the LANCELOT optimisation package S.S. Dooley, IBM - T. J. Watson Research Center, USA: "Flush the pipe!": how to write symbolic/numeric programs in $A^{{\#}}$ without interprocess communication *********************************************************************** Medicine and CAD Organizers: J. Hoschek, Technische Hochschule Darmstadt, Germany U. Weber, Orthop{\"a}dische Klinik der Freien Universit{\"a}t Berlin, Germany Summary: The Minisymposium will demonstrate how mathematical modelling and the use of computation methods can help doctors to improve the diagnosis, the operation planning and operation technique. In particular methods for optimal hip joint readjustment are being developed, the birth will be simulated to detect risks during the delivery, a training system based on Virtual Reality for arthroscopic operation of the knee will be introduced, the support of plastical operations and roboter assisted brain surgery can be demonstrated. Day: July 5, 1995 Time: 15.30 M. Eck, Technische Hochschule Darmstadt, Germany; J. Hoschek, Technische Hochschule Darmstadt, , Germany; U. Weber, Freie Universit{\"a}t Berlin, Germany: $3D$ planning of surgical hip joint readjustments A.W. Wischnik, Universit{\"a}t Heidelberg, Germany; K.J. Lehmann, Klinikum Mannheim; E. Nalepa, FH Darmstadt, Germany: Computer aided simulation of delivery by means of nuclear magnetic resonance imaging and finite element analysis W. M{\"u}ller, Fraunhofer Institut f{\"u}r Computergrafik Darmstadt, Germany; R. Ziegler, Fraunhofer Institut f{\"u}r Computergrafik, Darmstadt, Germany; M. G{\"o}bel, Fraunhofer Institut f{\"u}r Computergrafik, Darmstadt, Germany: Towards interactive realism -- the virtual reality arthroscopy training simulator A. Berger, Medizinische Hochschule Hannover, Germany; F. Siebert, Medizinische Hochschule Hannover, Germany; H.O. Shin, Medizinische Hochschule Hannover, Germany: Computer aided design supporting the plastic surgeon by CT-scan based three dimensional reconstruction J. Gilsbach, RWTH Aachen, Germany; G. Laborde, RWTH Aachen, Germany; A. Moesges, RWTH Aachen, Germany: Computer assisted localizer (CAL) in brain surgery C. Hagemann, Ipos Barendorf, Germany: Barlach CAD/CAM-System *********************************************************************** Mathematical Modelling of Road Traffic and Communication Networks Organizer: J. Norbury, University of Oxford, United Kingdom Summary: Network analysis is now being applied to two related areas of enormous social and financial significance, those of data communication and freeway/city road systems. The major aims are to predict large scale network behaviour in terms of local or microscale decision processes, and to control the system by identifying key parameters. \smallskip Over the past two decades computer simulations have increased in size and complexity; present large scale programs often have difficulty in relating parameter changes to system effects. New approaches involve analytical models which replace detailed simulation with simplified, but tractable traffic laws. This minisymposium focuses on these four areas: \item{$\bullet$} Hyperbolic conservation law on road networks. Practical route guidance systems developed from this theoretical approach. \item{$\bullet$} The modelling of road traffic at the microscopic urban level. \item{$\bullet$} Techniques in Markovian network analysis. \item{$\bullet$} Tesselation problems in radio frequency assignment. \smallskip The advantage of these models, with their analytical averaging/extrapolation of local results to the larger scale of the network, is that the relation betweeen system parameters and large scale behaviour becomes more transparent. The difficulties lie in the analytical extrapolation and that is where the breakthroughs are occuring. The meeting will be of interest to generalist and specialist alike. Day: July 6, 1995 Time: 15.30 J. Norbury, University of Oxford, United Kingdom: Long time behaviour of networks with local Markov processes R.A. Leese, Smith Institute, United Kingdom: Frequency assignment and tesselation problems for radio communication networks R. K{\"u}hne, Steierwald Sch{\"o}nharting und Partner GmbH, Germany: Practical traffic control based on continuum modelling R.E. Wilson, University of Oxford, United Kingdom: A new methodology for dynamic urban traffic flow B. Ycart, Laboratoire LMC - IMAG, France: Large scale Markovian network analysis H. Holden, Universitetet i Trondheim, Norway; N. H. Risbero, University of Oslo, Norway: Systems of hyperbolic conservation laws on networks - a model for traffic flow *********************************************************************** Boundary Element Methods: Applications Organizers: E. Stein, Universit{\"a}t Hannover, Germany H.A. Mang, Technische Universit{\"a}t Wien, Austria Summary: Boundary Element Methods (BEM) are now well known to be very efficient in many areas of technical applications but seemed long time to be practically restricted to linear problems. In the last years there were many new developments so that boundary element methods become now very attractive for more complicated especially nonlinear problems like elastoplasticity or viscoelasticity. \smallskip This minisymposium deals with the application of BEM in solid mechanics and solid-liquid interactions, respectively. For these problems new special boundary element formulations are developed, like BEM-FIS or a hybrid symmetric BEM. Efficient solvers for engineering problems like vibration analyses in liquid-filled tanks or frictional contact are shown. \smallskip Because FEM and BEM has complementary advantages and disadvantages the coupling of these two methods can lead to more effective results. Symmetric coupling formulations (symmetric Galerkin-BEM coupled with mixed finite elements and hybrid symmetric BEM) are derived leading to very good convergence properties demonstrated by examples. Day: July 7, 1995 Time: 09.30 H. Antes, Technische Universit{\"a}t Braunschweig, Germany; K. Latz, B{\"o}ger {\&} J{\"a}ckle Ing. GmbH, Henstedt-Ulzburg, Germany: Vibration analyses of liquid-filled tanks by a coupled boundary/finite element method U. Brink, Universit{\"a}t Hannover, Germany; E. Stein, University of Hannover, Germany: Coupling symmetric Galerkin-BEM with mixed finite elements of the Raviart-Thomas-type L. Gaul, Universit{\"a}t Stuttgart, Germany; C. Fiedler, Universit{\"a}t Stuttgart, Germany: A new hybrid symmetric boundary element method in elastodynamics A. Huesmann, Universit{\"a}t Erlangen-N{\"u}rnberg, Germany; G. Kuhn, Universit{\"a}t Erlangen-N{\"u}rnberg, Germany: Frictional contact problems - solution by means of discrete nodal constraints H.A. Mang, Technische Universit{\"a}t Wien, Austria; Z.-S. Chen, Technische Universit{\"a}t Wien, Austria; G. Hofstetter, Universit{\"a}t Innsbruck, Austria: A Galerkin BEM for determination of sound radiation and reflection of structural members *********************************************************************** Exponential Asymptotics Organizer: S. Tanveer, The Ohio State University, USA Summary: Exponential asymptotics have become of increasing interest due to the wide variety of physical problems where they arise. This minisymposium touches on some of these applications including spectral problems in quantum mechanics, travelling wave speed selection, oscillatory water wave amplitudes and steady state and stability problems in viscous displacement in a Hele-Shaw cell. In addition, recent fundamental findings on exponential asymptotics for solutions to higher than second order differential equations will be presented. Talks include a wide range, both of topics and techniques. Day: July 6, 1995 Time: 15.30 C.J. Howls, The Manchester Metropolitan University, United Kingdom: Exponentially asymptotic quantum billards R. O'Malley, University of Washington at Seattle, USA: Using exponential asymptotics to obtain travelling wave slutions to singularly perturbed boundary value prolems A.D. Wood, Dublin City University, Ireland: Stokes phenomena for high order differential equations T.R. Akylas, Massachusetts Institute of Technology, USA: On short-scale oscillatory tails of long-wave disturbances S. Tanveer, The Ohio State University, USA: Exponential asymptotics in Hele-Shaw displacement *********************************************************************** Asymptotics and 3 D-Visualization of Navier-Stokes Approximations Organizers: W. Borchers, Universit{\"a}t-GH Paderborn, Germany K. Pileckas, Universit{\"a}t-GH Paderborn, Germany R. Rautmann, Universit{\"a}t-GH Paderborn, Germany Summary: The recent progresses in the construction of flow machineries, car and aeroplane design are essentially based on the numerical solution of flow problems in complex 3 D-geometries. The known challenges in this field require even more efficient numerical methods. Therefore the minisymposium presents recent theoretical and numerical results for the Navier-Stokes equations of incompressible fluids, including asymptotic analysis for free boundary and mixed boundary value problems, for in- and outflow conditions, error estimates for discretization models, and parallelization methods for massiv parallel computer architectures as well as visualization technics to handle large flow data on high performance graphic computers. Special results on 3 D-flows will be shown in real time interactive visualization. Day: July 4, 1995 Time: 15.30 W. Borchers, Universit{\"a}t-GH Paderborn, Germany: Singularities for a mixed boundary value problem of Stokes equations S. Blazy, Universit{\"a}t-GH Paderborn, Germany: Preconditioning technics on massiv parallel architectures R. Bruckschen, Universit{\"a}t-GH Paderborn, Germany: 3 D particle tracing for real-time visualization U. Dralle, Universit{\"a}t-GH Paderborn, Germany: Finite element software engineering for parallel computersystems K. Pileckas, Universit{\"a}t-GH Paderborn, Germany: Asymptotic behaviour of solutions to Navier-Stokes equations and application to free boundary problems R. Rautmann, Universit{\"a}t Paderborn, Germany: Navier-Stokes approximations in high order norms *********************************************************************** Adaptive and Multilevel Methods for the Numerical Solution of Transonic Flow Organizer: M. Feistauer, Charles University Prague, Czech Republic Summary: The minisymposium is concerned with recent results obtained in the area of numerical simulation of transonic inviscid and viscous gas flow. The main emphasis is laid on suitable adaptive strategies for the automatic mesh refinement with the aim to obtain a highly accurate solution with the precise resolution of shock waves and boundary layers. Moreover, acceleration techniques for the improvement of the convergence of numerical processes to steady state solutions are treated. The contributions will cover problems described by the full potential transonic equation, compressible Euler equations and Navier-Stokes equations. Theory, numerical algorithms and computational results will be discussed. Day: July 5, 1995 Time: 09.30 M. Feistauer, Charles University Prague, Czech Republic: Numerical simulation of compressible viscous flow through cascades of profiles J. Felcman, Charles University Prague, Czech Republic: Adaptive methods for the solution of the Euler equations in elements of blade machines M. Hunek, The Czech Technical University Prague, Czech Republic; K. Kozel, The Czech Technical University Prague, Czech Republic: Acceleration techniques for computation of inviscid or viscous transonic flows G. Warnecke, Otto-von-Guericke-Universit{\"a}t Magdeburg, Germany: Error estimates and adaptive methods for transonic flow problems D. Kr{\"o}ner, Albert-Ludwig-Universit{\"a}t Freiburg, Germany: Gridrefinement and grid alignment for Euler equations in multidimensions *********************************************************************** Singularities in Solid Mechanics Organizers: V.G. Maz'ya, University of Link{\"o}ping, Sweden A.M. S{\"a}ndig, Universit{\"a}t Stuttgart, Germany Summary: The investigation of singularities of solutions of elliptic boundary value problems is essential for solid mechanics, especially, stress concentrations can occur if the bodies have edges or corners. In this situation standard numerical techniques may lose accuracy and a-priori information about the singularities of the solutions is useful.\\ During the last three decades a mathematical theory of boundary value problems in domains with non-smooth boundaries was developed. However, there still exists a certain gap between this theory and the demands of the applications. Besides. there are still open problems in the theory itself.\\ The minisymposium reflects the state of the art as well as new results in this field. The following topics wil be discussed: branching asymptotics of solutions near edges and corners, calculation of the coefficients in the edge asymptotics of elliptic systems, in particular, of the Lam\'e system, asymptotic analysis of elastic fields in multi-structures, non-linear operator differential equations with applications to the singularities of the solutions of non-linear partial differential equations and the treatment of nonlinear 3D crack problems using the BEM. Day: July 3, 1995 Time: 15.30 O. Huber, ABB Power Generation Ltd., Switzerland; G. Kuhn, Universit{\"a}t Erlangen-N{\"u}rnberg, Germany: Treatment of nonlinear 3D crack problems using the boundary element method M. Dauge, Universit{\'e} de Rennes 1, France; M. Costabel, Universit{\'e} de Rennes 1, France: Combined corner and edge singularities V.A. Kozlov, Link{\"o}ping University, Sweden; Maz'ya, University of Link{\"o}ping, Sweden: Asymptotics of solutions to nonlinear ordinary differential equations with operator coefficients A.-M. S{\"a}ndig, Universit{\"a}t Stuttgart, Germany; J. Rossmann, Universit{\"a}t Rostock, Germany: Formulae for the coefficients in the asymptotics of elliptic systems of second order near edges, applications to the Lam{\'e} system A.B. Movchan, University of Bath, United Kingdom: Asymptotic study of solutions of linear elasticity in a multi-structure *********************************************************************** Bifurcations and Their Applications Organizer: V. Trenogin, Moscow Institute of Steel and Fusions, Russia Summary: Branching and bifurcations theory successfully develops on account of its applications and arising from them new matematical problems. Note here the following new directions. \item{1)} Boundary value problems for ordinary and partial differential equations (DEq) where the unknown function belongs to one Banach space and boundary and initial values to other. Here impulse DEq, DEq in composite domains and bifurcations of their solutions are contained. \item{2)} Parameter continuation method, bifurcation points passage problem by optimal parameter selection. \item{3)} Wide using of group analysis methods, potentiality properties. \item{4)} New theorems about Hopf bifurcation and other bifurcation phenomena. \smallskip Applications are considered: waves on the fluid surface and inner waves; equilibrium loosing phenomena in elastic systems; bifurcations in biology and ecology; bifurcations in problems of chemical technology, theoretical physics. Day: July 4, 1995 Time: 09.30 V.A. Trenogin, Moscow Institute of Steel and Fusions, Russia: Certain problems of contemporary branching theory V.A. Trenogin, Moscow Institute of Steel and Fusions, Russia; N.A. Sidorov, Irkutsk University, Russia; B.V. Loginov, Ulyanovsk State Technical University, Russia: Bifurcation, potentiality, group-theoretical and iterative methods B.V. Loginov, Ulyanovsk State Technical University, Russia; V.A. Trenogin, Moscow Institute of Steel and Fusions, Russia: Group symmetry of bifurcation equation in dynamic branching B.V. Loginov, Ulyanovsk State Technical University, Russia; V.A. Trenogin, Moscow Institute of Steel and Fusions, Russia; P.A. Velmisov, Ulyanovsk State Technical University, Russia: Bifurcation and stability in some problems of continua mechanics P.A. Velmisov, Ulyanovsk State Technical University, Russia: Stability of viscoelastic bodies accounting aging and interaction with fluid or gas N.A. Sidorov, Irkutsk University, Russia; N.V. Ermilova, Irkutsk State University, Russia: Iteration methods in the neighbourhood of the branching point of nonlinear equations *********************************************************************** What do Car Parking, Space Robots and Air Traffic Control have in Common? Organizer: S.S. Sastry, University of California at Berkeley, USA Summary: This session is intended to give the conference a sense of the application of new mathematical techniques from exterior differential systems, Hamiltonian dynamics, reduction and control of systems on Lie groups to many important new applications. The applications span mobile robots (cars with many trailers), robots in space (satellites) and underwater (submarines), mechatronic systems with energy conservation and manufacturing. \smallskip Dominant themes in this session include the control of systems with nonholonomy, Hamiltonian reduction, control of left invariant control systems on Lie groups and symplectic mechanics techniques in control theory. The talk of Tilbury with emphasize the use of exterior differential systems in studying nonholonomic kinematics, that of van der Schaft will emphasize the difficulties in maintaining a symplectic structure in the control of Hamiltonian systems with nonholonomic constraints.\\ The talks of Li, Leonard and Sastry will be on the control of systems on Lie groups, each using different methods and highlighting different applications. Li will emphasize manufacturing, Leonard will use averaging techniques for the control of underwater vehicles and submarines. Sastry will use optimal control, Hamiltonian reduction for the problems of air traffic control. Day: July 6, 1995 Time: 09.30 D. Tilbury, University of Michigan, USA: Exterior differential systems and nonholonomic motion planning A.J. van der Schaft, University of Twente, The Netherlands; B. Maschke, Conservatoire National des Arts et Metiers, France: The Hamiltonian formulation of energy conserving physical systems: mathematical modeling and system theoretic properties Z. Li, Hong Kong Univ. of Science and Technology, Hong Kong: On geometric localization and its applications to manufacturing N.E. Leonard, Princeton University, USA: Periodic forcing of underactuated dynamical systems on Lie groups S.S. Sastry, University of California at Berkeley, USA: Optimal control of systems on Lie groups *********************************************************************** Voronoi Networks Organizers: D.J. Bell, UMIST, Manchester, United Kingdom G.A. Davies, UMIST, Manchester, United Kingdom Summary: Problems of quantification and modelling of spacial patterns exist in branches of mathematics, computation, science, medicine and engineering. In this symposium the use of Voronoi tesselations applied to a wide range of spacial problems will be considered. These include modelling the growth of tumours, the spacial distribution of galaxies, the shape and size distributions of crushed particles, the structure of porous membranes and foams and percolation problems on Voronoi lattices to describe problems in filtration. Day: July 5, 1995 Time: 09.30 G.A. Davies, UMIST, United Kingdom; D.J. Bell, UMIST, UK: A brief survey of natural patterns and distributions which may be described by Voronoi networks W. D{\"u}chting, Universit{\"a}t-Gesamthochschule Siegen, Germany; W. Ulmer, Max-Planck-Institute of Biophysical Chemistry, G{\"o}ttingen, Germany: Modelling of tumor growth and treatment L. Dunkley, University of Luton, United Kingdom; C. Blackburn, University of Luton, UK: The application of Voronoi tessellations in development of $3D$ stochastic models to represent tumour growth R. van de Weygaert, Max Planck Institut f{\"u}r Astrophysik, Germany: The application of Voronoi foams in Astrophysics G.A. Davies, UMIST, United Kingdom; J. Broughton, UMIST, UK; N.M. Jackson, UMIST, UK; R. Jafferali, UMIST, UK; D.J. Bell, UMIST, UK: The application of Voronoi tessellations in 2 and 3 space to simulate the structure of porous membranes D.J. Bell, UMIST, United Kingdom; P. Deckmyn, UMIST, UK; G.A. Davies, UMIST, UK: Percolating clusters on Voronoi lattices and the relationship to particle fouling on filters P. King, University of Utah, USA: Decomposed Voronoi lattices to model the size and shape distributions of crushed particles *********************************************************************** Trends in Hard- and Software for Scientific Computing Organizer: R. Jan{\ss}en, IBM Heidelberg, Germany Summary: This minisymposium intends to give researchers on new algorithms and methods for scientific computing, developers of applications codes, etc., a comprehensive view on trends in hard- and software for scientific computing. Four survey talks will answer the questions: \item{$\bullet$} what are the future hardware (processor/architecture) trends to be reflected in algorithm and software development to achieve efficient codes? \item{$\bullet$} what are the new programming environments and tools to be used for efficient development and testing of new methods? \item{$\bullet$} what are evolving standards and experiences in designing interfaces in mathematical software (from numeric to symbolic to graphic to text)? \item{$\bullet$} what are guidelines to develop new and enable existing applications for parallel architectures, including results of two major european efforts (EUROPORT and RAPS)? Day: July 3, 1995 Time: 15.30 W. Gentzsch, Fachhochschule Regensburg, Germany: Hardware for scientific computing S.M. Rump, Technische Universit{\"a}t Hamburg-Harburg, Germany: Development environments and tools for scientific computing software development S.J. Hague, Numerical Algorithm Group, United Kingdom: Interface design for mathematical software A. Sch{\"u}ller, GMD, Germany; C. Thole, GMD Bonn, Germany: Application development and enabling for parallel systems *********************************************************************** Mathematical and Physical Interpretation of Climate Models Organizer: J.F.B. Mitchell, Hadley Centre, England Summary: The possibility that human activity may have noticeable impact on climate has received increasing attention in recent years. There are considerable areas of uncertainty including the size of future changes in greenhouse gases and other factors affecting climate, the response of the highly non-linear climate system to radiative perturbations, and the detection of forced changes above the natural variations in the climate. This sysmposium considers some of the mathematical problems of modelling climate, detecting climate change and of dealing with uncertainty. Day: July 7, 1995 Time: 09.30 J.C.R. Hunt, Hadley Centre, United Kingdom: Anthropogenic climate change M.K. Davey, United Kingdom Meteorological Office, United Kingdom: ENSO predictability in simple and GCM models G.C. Hegerl, Max-Planck-Institut f{\"u}r Meteorologie, Germany; H. von Storch, Max-Planck-Institut f{\"u}r Meteorologie, Hamburg, Germany: Applying coupled ocean-atmosphere models for predicting and detecting climate change P.C. Young, Lancaster University, United Kingdom: Simplicity out of complexity in climate models: sensitivity to uncertainity and modal dominance *********************************************************************** Homoclinic and Heteroclinic Behavior of Continua Organizers: B. Fiedler, Freie Universit{\"a}t Berlin, Germany A. Mielke, Universit{\"a}t Hannover, Germany Summary: Dimension reduction relates the behavior of continua, described by the partial differential equations of fluids or elastic media, to low dimensional dynamics. Moreover, spatial structures may appear as "time" evolution, after reduction. Homoclinic and heteroclinic orbits in these spatial dynamical systems correspond to (traveling) solitary pulses and fronts in the continuum mechanical setting. Moreover, they are related to both, stability questions and to complicated dynamics or chaotic patterns. \smallskip The Mini-Symposium intends to ignite communication among speakers and audience from both sides: those deeply involved with the behavior of continua, and those familiar with the complications of homoclinic and heteroclinic dynamics. Day: July 3, 1995 Time: 10.00 D. Turaev, Russian Academy of Sciences, Russia: Dimension reduction for semi-local bifurcations W.-J. Beyn, Universit{\"a}t Bielefeld, Germany: Numerical computation and stability analysis of travelling waves C.K.R.T. Jones, Brown University, USA: Stability of solitary waves M.V. Schatzman, Universite de Lyon, France: The Thual-Fauve pulse, theory and computation M. Renardy, Virginia Tech., USA: Shocks and shock structures for surfactant spreading in thin films H. Kokubu, Kyoto University, Japan: Multiple homoclinic bifurcations in vector fields *********************************************************************** Theory and Applications of Neuronal Networks Organizer: R. Rojas, Martin Luther Universit{\"a}t Halle, Germany Summary: Neural networks can be defined as networks of primitive functions. They can be used to approximate unknown functions of which just some input-output examples are known. They can also be used to implement massively parallel systems capable of solving certain pattern recognition tasks. The main difference with other kind of parallel computing devices is their self-organizing ability, provided by a learning algorithm.\\ In this minisymposium we want to assess some of the latest advances in the theoretical understanding of neural networks. We want also to discuss some interesting applications in the field of robotics and biomolecular modeling. One of the main problems in the field of neural networks is the exact statistical validation of the output produced by the network. In the minisymposium we will discuss some ideas from the field of computational statistics that can provide a way out of this dilemma. Day: July 4, 1995 Time: 09.30 M. Pfister, Freie Universit{\"a}t Berlin, Germany: to be announced M. Eldracher, Technische Universit{\"a}t M{\"u}nchen, Germany: to be announced M. Schwerk, Technische Universit{\"a}t Berlin, Germany: to be announced *********************************************************************** Inverse Problems and Wavelets Organizers: D.L. Donoho, University of California at Berkeley, USA A.K. Louis, Universit{\"a}t des Saarlandes, Saarbr{\"u}cken, Germany Summary: The minisymposium is devoted to applications of wavelets in industrial problems. A central role is played by the impact on medical imaging and nondestructive evaluation. Two of the lectures treat the derivation of inversion formulas both for linear and nonlinear problems. If only a small number of data are available discrete versions of the mathematical models have to be studied. Fast and stable solutions of these ill-posed problems are presented using the separation of different frequency components of the solutions by the wavelet transform. Day: July 4, 1995 Time: 09.30 M. Holschneider, Institut de M{\'e}canique des Fluids de Marseille, France: Application of inverse wavelet transforms to Radon data A.K. Louis, Universit{\"a}t Saarbr{\"u}cken, Germany: Approximative inverse for nonlinear ill-posed problems P. Maa{\ss}, Universit{\"a}t Potsdam, Germany: Wavelet accelerated iteration schemes for inverse problems A. Rieder, Universit{\"a}t des Saarlandes, Germany: A multilevel iteration for linear ill-posed problems *********************************************************************** Adaptive Finite Element Methods Organizers: C. Johnson, Chalmers University of Technology, Sweden E. Stein, Universit{\"a}t Hannover, Germany Summary: New results on a posteriori error analysis and resulting error indicators for essentially non-linear problems in continuum mechanics are treated by leading authorities in the field, facing the objects reliability and efficiency for complex problems. Some key-points are: balancing of stability requirements and optimal convergence for time-dependent, esp. parabolic equations; increasing of computational efficiency by combining interpolation -- and solution-errors within Multigrid solvers for elliptic problems; error indicators for Prandtl-Reuss elasto-plasticity with critical comparisons of efficiency; adaptive mesh refinements for non-linear elastic multi-body contact mechanics; hierarchical multilevel ph-methods with local h-refinements.\\ The general scope is the integration of rigouros error analysis with effective solution algorithms for treating challenging engineering problems. Day: July 4, 1995 Time: 15.30 K. Eriksson, Chalmers University of Technology, Sweden: Adaptive time stepping with automatic stability and error control R. Becker, Universit{\"a}t Heidelberg, Germany; R. Rannacher, Universit{\"a}t Heidelberg, Germany: Adaptive error control for finite element multigrid solution of elliptic equations F.-J. Barthold, Universit{\"a}t Hannover, Germany; E. Stein, Universit{\"a}t Hannover, Germany: Error estimation and mesh adaptivity for elasto-plastic deformations including mesh coarsening P. Wriggers, Technische Hochschule Darmstadt, Germany; O. Scherf, Technische Hochschule Darmstadt, Germany: Adaptive finite element techniques for contact problems involving large elastic strains E. Rank, Universit{\"a}t Dortmund, Germany; R. Krause, Universit{\"a}t Dortmund, Germany: Adaptive multi-scale computations using standard finite element packages *********************************************************************** Pseudodifferential Equations and Problems of Anisotropic Elasticity Organizers: R.V. Duduchava, Academy of Sciences of Georgia, Republic of Georgia G.V. Jaiani, Tbilsi University, Republic of Georgia D.G. Natroshvili, Georgian Technical University, Republic of Georgia Summary: Althoug mathematical theory of elasticity for bodies with smooth boundaries is far developed, most problems, especially, for anisotropic bodies with non-smooth boundaries and cracks still remain unsolved or need further investigation. Recent developments in the theory of pseudodifferential equations with non-classical symbols (non-classical $\Psi$DOs) on manifolds with boundary provide a powerfool tool for investigating the solvability, a-priori smoothness and asymptotic behaviour of solutions to such problems in the vicinity of singularities of manifolds. Since the approach provides an equivalent reduction of boundary value problem to a corresponding boundary integral equation, investigations of solvability properties and behaviour of solutions, it accomplishes a solid theoretical basis for development of efficient boundary element methods for approximate solutions of problems which are within the interests of engineering applications. The same approach can be successfully applied to various static and dynamic problems of thermoelasticity, theory of elastic composites, diffraction of electro-magnetic waves, propagation of elastic waves, hydrodynamics, theory of plates and shells etc. Essential progress in these fundamental investigations may considerably improve the state of the art in engineering applications. Day: July 7, 1995 Time: 09.30 R.V. Duduchava, Academy of Sciences of Georgia, Republic of Georgia: The Wiener-Hopf method in crack and interface problems D.G. Natroshvili, Georgian Technical University, Republic of Georgia: Steady state oscillation problems in anisotropic elasticity G.V. Jaiani, Tbilisi State University, Republic of Georgia: Elastic bodies with non-smooth boundaries -- cusped plates and shells E.M. Shargorodsky, Academy of Sciences of Georgia, Republic of Georgia: Stationary Navier-Stokes equation in a domain exterior to an open surface L.B. Sigua, Academy of Sciences of Georgia, Republic of Georgia: Boundary value problems for Helmholtz vector equation -- non-smooth domains V.V. Kirvalidze, Academy of Sciences of Georgia, Republic of Georgia: Boundary value problems of hydrodynamics for noon-smooth domains