From nacomb@surfer.EPM.ORNL.GOV Sun Mar 10 12:06:40 1991 Return-Path: Received: from surfer.EPM.ORNL.GOV by CS.UTK.EDU with SMTP (5.61++/2.5.1s-UTK) id AA09386; Sun, 10 Mar 91 12:06:35 -0500 Received: by surfer.EPM.ORNL.GOV (5.61/1.34) id AA10847; Sun, 10 Mar 91 12:06:27 -0500 Date: Sun, 10 Mar 91 12:06:27 -0500 From: nacomb@surfer.EPM.ORNL.GOV (NA-NET) Message-Id: <9103101706.AA10847@surfer.EPM.ORNL.GOV> Comment: Submissions for NA News Digest, mail to na.digest@na-net.ornl.gov. Comment: Information about NA-NET, mail to na.help@na-net.ornl.gov. Comment: Comments about the NA-NET, mail to nanet@na-net.ornl.gov. Apparently-To: dongarra@cs.utk.edu Status: R NA Digest Sunday, March 10, 1991 Volume 91 : Issue 10 Today's Editor: Cleve Moler Today's Topics: Information About Finite Element Software Numerical Solution of Time Delayed Differential Equations Large Dimensional Dynamic Models SLICOT, Software Library for Control Theory IMACS International Symposium on Iterative Methods in Linear Algebra Multigrid Short Course ------------------------------------------------------- From: Shirley Pomeranz Date: Mon, 4 Mar 91 23:02:33 CST Subject: Information About Finite Element Software I would like to know of any current surveys or articles describing finite element software packages that are available. Anything comparing different packages and discussing the particular field(s) or engineering or mathematics discipline for which the software is intended would be of interest to me. Thank you for whatever information you can send my way. Shirley Pomeranz Assistant Professor Mathematics Department of Mathematical and Computer Sciences The University of Tulsa ------------------------------ From: Russell Van Gelder Date: 7 Mar 91 20:42:14 GMT Subject: Numerical Solution of Time Delayed Differential Equations I am trying to model an oscillatory biochemical process and I have run into difficulty due to time delays in the equations. Does anyone know of any public domain algorithms for numerically integrating first order differential equations with retarded arguments? The problem to be solved is an initial value problem. An adaptation of fourth order Runge-Kutta would be especially welcome - I have looked in most of the standard differential equation and numerical methods books, but to no avail. Russell Van Gelder vangeldr@cmgm.stanford.edu Stanford Center for Molecular and Genetic Medicine - CMGM Department of Biochemistry Stanford University School of Medicine Stanford, CA 94305 (415) 723-5871 ------------------------------ From: A. Scottedward Hodel Date: Fri, 8 Mar 91 13:17:20 CST Subject: Large Dimensional Dynamic Models I am working on algorithms for the solution of large-dimensional Lyapunov and algebraic Riccati equations. I have been testing these algorithms on ``artificially'' constructed problems that meet certain requirements, but I would like to test them on large order models of physical systems (typically large, flexible structures). I am attempting to build a collection of large dimensional dynamic system models of the form d/dt x = A x + B u y = C x To date, the largest system I have been able to obtain is roughly 150th order; they are typically half that size and are already in modal (block diagonal) form. If anyone has or is aware of system models of larger order, I'd very much like to obtain the system data. Thanks! Scotte ------------------------------ From: Paul Van Dooren Date: Mon, 04 Mar 91 10:05:06 N Subject: SLICOT, Software Library for Control Theory RELEASE OF THE NAG SLICOT LIBRARY The Subroutine Library in COntrol Theory (SLICOT) is a collaborative effort between the Numerical Algorithms Group (NAG) and the Benelux Working Group on Software (WGS). SLICOT is a library of FORTRAN 77 subroutines for control systems analysis and design and consists of 68 user-callable routines covering a wide range of applications areas within control theory. Great care has been taken to build SLICOT on reliable and robust algorithms. Wherever possible, numerical and statistical routines from the NAG Fortran library are used. The SLICOT library can be run on any machine using a Fortran 77 compiler. Source text (including example programs) is provided. The current NAG price lies around 500 UKPounds for academic institutes and 1000 UKPounds for indutry. A short description of the contents of the library is given below : Utility routines (2 routines): Printing real matrices and evaluating machine-dependent parameters Mathematical routines (20 routines): Basic linear algebra operations, linear equations, (total) least squares, eigenanalysis, balancing, transformations to diagonal bidiagonal and Hessenberg forms, Choleski decomposition, matrix exponential Transformation routines (18 routines): Transformations from state-space to various canonical forms, to transfer functions and polynomial matrix representation, from polynomial matrix form to transfer matrix and vice-versa, Markov parametrization from state-space and transfer matrix Analysis routines (11 routines): Reduction to minimal state space models, continuous to discrete, interconnection of subsystems, dual forms, invariant zeros, Hankel and Toeplitz expansion of multivariable sequences, deadbeat control Synthesis routines (9 routines): Pole placement, algebraic Riccati equations, Lyapunov and Sylvester equations, realisation methods, optimal control and spectral factorization Data Analysis (4 routines): Convolution and deconvolution, forward and inverse Fourier transform, anti-aliasing windowing Filtering (4 routines): Kalman filtering in different square root implementations SLICOT is also a continuing effort. The WGS is responsible for following the evolution of numerical algorithms in the area and tries to complete the library in this respect (possibly by inviting experts to contribute their software to the library). A 2nd release is currently being produced and will contain up to 20 new routines. In the future we also hope to be able to include a chapter of nonlinear methods and adaptive control. To obtain more information about SLICOT please reply to one of the following addresses : NAG c/o A. Brown WGS c/o R. Kool Wilkinson House Eindhoven University of Technology Jordan Hill Road Dept. Math. & Comp. Sc., DG 1.08 OXFORD P.O. Box 513 OX2 8DR NL-5600 MB Eindhoven United Kingdom The Netherlands Tel: +44 865 511245 Tel: +31 40 472374 Fax: +44 865 310139 Fax: +31 40 442150 NAGAAB@VAX.OXFORD.AC.UK WSCOLB@HEITUE5 (on BITNET) ------------------------------ From: Pieter de Groen Date: Tue, 5 Mar 91 14:07:10 +0100 Subject: IMACS International Symposium on Iterative Methods in Linear Algebra IMACS International Symposium on Iterative Methods in Linear Algebra. Brussels Free Universities (ULB & VUB), April 2nd - 4th, 1991. Aula VUB, Building Q, Pleinlaan 2 (Av. de la Plaine), B-1050 Brussels. SCOPE: The purpose of the symposium is to provide a forum for the presentation and the discussion of recent advances in the analysis and implementation of iterative methods for solving large linear sytems of equations and for determining eigenvalues, eigenvectors or singular values of large matrices. TOPICS: Matrix analysis: convergence acceleration - preconditioning - methods for nonsymmetric, singular and overdetermined systems - sparse eigenvalue problems Boundary value problems: multigrid methods - domain decomposition - spectral methods Implementation techniques: on vector processors - on multiprocessors - on massively parallel systems Software developments: for sparse linear systems - for sparse eigenproblems Mathematical applications: partial differential equations - systems theory - least squares problems INVITED SPEAKERS: A. van der Sluis O. Axelsson A. Yeremin H. van der Vorst D. Kincaid E. Wachspress F. Chatelin SPECIAL SESSIONS (30 papers) Coupled inner outer iteration methods, organized by O. Axelsson Numerical methods for the analysis of Markov models, organized by P. Courtois & G. Latouche. Iterative solution of unsymmetric systems, organized by H. van der Vorst. Spectral methods, organized by M. Deville & E. Mund. The Lyapunov equation, organized by E. Wachspress. Parallel and vector iterative methods, organized by D. Kincaid & C. Wu. Complex variable methods for solving non-positive definite linear systems, organized by M. Eiermann & W. Niethammer. Sessions of Contributed Papers (45 papers) Basic iterative methods Eigenvalue Problems Least squares and linear complementary problems Methods for nonsymmetric systems Mixed Hybrid Methods Semiconductor device equations Preconditioned conjugate gradients Software Developments Monte Carlo Methods. The symposium is sponsored by the Belgian Fund for Scientific Research, IBM and Honeywell. The reception in the Brussels Town Hall is made possible by the kind cooperation of the Brussels Mayor and Alderman. The organising committee, Robert Beauwens and Pieter de Groen. ------------------------------ From: Gaoming Yang Date: Tue, 5 Mar 91 13:47:22 -0700 Subject: Multigrid Short Course MULTIGRID SHORT COURSE Information & Schedule University of Colorado at Denver March 25-29, 1991 Principal Lecturer: Achi Brandt Supporting Lecturers: John Adams, William Briggs, Chaoqun Liu, Steve McCormick, John Ruge Short Course Chairman: Chaoqun Liu Purpose: To provide an understanding of the principles and procedures for multilevel methods, especially for partial differential equations, including new multilevel approaches in computational fluid dynamics. Registration Fee: $500 (regular), $150 (student). Includes books and other course materials, refreshments, and computer lab access. Topics: . Basic tutorial . Advanced methods for PDE's (including multigrid procedures for general systems, nonlinearity, ellipticity/nonellipticity, time dependence, inverse problems, indefiniteness, discontinuities, singularities, performance prediction/analysis, constrained optimization) . Adaptive techniques (MLAT, FAC) . Algebraic multigrid methods (AMG) . Computational fluid dynamics (including finite volume element methods, high Reynolds number flow, steady and unsteady Navier-Stokes equations, Steady and unsteady Euler equations, flow transition, and two-phase porous flow) . Introduction to non-PDE multilevel techniques (including integral and integro-differential equations, fast dense matrix multiplication, many-body interactions, direct solvers, large determinants, global discrete highly-nonlinear optimization, and multilevel Monte Carlo method in statistical physics). For Further Information, Please Contact: Chaoqun Liu Computational Mathematics Group University of Colorado at Denver 1200 Larimer Street, Campus Box 170 Telephone: (303) 556-4807 or (303) 556-4886 (e-mail: cliu@copper.denver.colorado.edu) (e-mail: cliu@cudnvr.denver.colorado.edu) ------------------------------ End of NA Digest ************************** -------