Friday, September 22

MS23
Computational Issues in Fully Implicit Solutions for Large-Scale Nonlinear Systems

10:30 AM-12:30 PM
Center City 1

Many applications give rise to mathematical models involving multiple equations in multiple unknowns that are nonlinearly dependent on each other. These systems have often been formulated in an operator split or time-lagged manner in order to develop a model numerically solvable within a reasonable amount of time and computer memory. However, with the advent of large, parallel computers, fully implicit formulations are now more tractable. The speakers in this minisymposium will present work in solving fully implicit formulations of large, nonlinear systems found in various applications. They will discuss the algorithmic and computational issues they are facing in order to develop solution methods that run on today's state-of-the-art computers and produce solutions in reasonable times.

Organizer: Carol S. Woodward
Lawrence Livermore National Laboratory, USA
10:30-10:55 Implicit Solution of Radiation-Diffusion Problems
Carol S. Woodward, Organizer; Peter N. Brown, Britton Chang, and Frank Graziani, Lawrence Livermore National Laboratory, USA
11:00-11:25 Computational and Experimental Study of Flames with Detailed Chemistry
Mitchell D. Smooke, Yale University, USA
11:30-11:55 Newton-Krylov-Multigrid Methods for Incompressible Navier-Stokes Equations
Michael Pernice, University of Utah, USA
12:00-12:25 Parallel Unstructured Fully Implicit Finite Element Simulations of Reacting and Non-Reacting Flows
John N. Shadid, Sandia National Laboratories, Albuquerque, USA

©2000, Society for Industrial and Applied Mathematics
Designed by Donaghy's Web Consulting
Created 6/14/00; Updated 6/14/00