DISTRIBUTED-MEMORY COMPUTING WITH THE LANGLEY AEROTHERMODYNAMIC UPWIND RELAXATION ALGORITHM (LAURA)

Citation
Cj. Riley et Fm. Cheatwood, DISTRIBUTED-MEMORY COMPUTING WITH THE LANGLEY AEROTHERMODYNAMIC UPWIND RELAXATION ALGORITHM (LAURA), Advances in engineering software, 29(3-6), 1998, pp. 317-324
Citations number
28
Categorie Soggetti
Computer Science Software Graphycs Programming","Computer Science Interdisciplinary Applications","Computer Science Software Graphycs Programming","Computer Science Interdisciplinary Applications
ISSN journal
09659978
Volume
29
Issue
3-6
Year of publication
1998
Pages
317 - 324
Database
ISI
SICI code
0965-9978(1998)29:3-6<317:DCWTLA>2.0.ZU;2-8
Abstract
The Langley Aerothermodynamic Upwind Relaxation Algorithm (LAURA), a N avier-Stokes solver, has been modified for use in a parallel, distribu ted-memory environment using the Message-Passing Interface (MPI) stand ard. A standard domain decomposition strategy is used in which the com putational domain is divided into subdomains with each subdomain assig ned to a processor. Performance is examined on dedicated parallel mach ines and a network of desktop workstations. The effect of domain decom position and frequency of boundary updates on performance and converge nce is also examined for several realistic configurations and conditio ns typical of large-scale computational fluid dynamic analysis. (C) 19 98 Published by Elsevier Science Limited. All rights reserved.