Application des methodes de decomposition de domaines au calcul parallele

Authors

  • Yves Escai Laboratoire de Mecanique de Rouen CNRS UPRESA 6104 Place Emile Blonde/ BPB 76131 Mont Saint-Aignan
  • Gilbert Touzot Laboratoire de Mecanique de Rouen CNRS UPRESA 6104 Place Emile Blonde/ BPB 76131 Mont Saint-Aignan

Keywords:

distributed computing, domain decomposition methods, direct solvers

Abstract

In this article, we study the use of distributed parallel processing for the resolution of linear static problems using the finite element method. In the first part, we briefly present some concepts and definitions of parallel processing. In the second part, the domain decomposition methods are introduced. Finally, we focus on a particular method where the interface problem is explicitely constructed and solved with a direct solver. Its parallelisation is presented and the performances obtained on a Cray T3D are analysed.

 

Downloads

Download data is not yet available.

References

[AGO 87] AGOSHKOV V.I., « Poincare-Steklov operators and domain decomposition

methods », Proceedings of the first international symposium on domain decomposition

methods for partial differential equations, Paris, 1987, Glowinski, Golub, Meurant,

Periaux, eds, SIAM, Philadelphia (1988).

[AUN 90] AUNAY S., Architecture de logiciel de modelisation et traitements distribues. These

de I'Universite de Technologie de Compiegne, 1990.

[BRA 89] BRAMBLE J.H., PASCIAK J.E., SCHATZ A.H., « The construction of preconditioners

for elliptic problems by substructuring IV », Mathematics of computation, vol. 53, num.

, 1989.

[BRA 93] BRA.UNL T., Parallel programming, Prentice-Hall International Editions, 1993.

[BRE 91] BREITKOPF P., KNOPF-LENOIR C., MORANCAY L., TOUZOT G., VAYSSADE M., New

Software Architecture: SIC, Communication a STRUCOM, novembre 1991.

[COS 93] CoSNARD M., TRYSTRAM D., Algorithmes et architectures paralleles, InterEditions,

[DON 91] DONGARRA J., DUFF I., SORENSEN D., VAN DER VORST H., « Solving linear

systems on vector and shared memory computers », SIAM.

[ESC 96] EsCAIG Y., TouzoT G., «Optimization od direct domain decomposition methods»,

Computer Assisted Mechanics and Engineering Sciences, 3, 1996.

[FAR 94] FARHAT C., Roux F-X., «Implicit parallel processing in strctural mechanics»,

Computational Mechanics Advances, 1994.

[FOS 95] FOSTER 1., Designing and building parallel programs, Addison-Wesley, 1995.

[GAL 94] GALLN AN K. & al., « Parallel algorithms for matrix computations », SIAM, 1994.

[HOC 92) HOCKNEY R., JESSHOPE C., Parallel Computers 2, Institut of Physics Publishing,

[HWA 84] HWANG K., BRIGGS F., Computer Architecture and Parallel Processing,

McGRAW-HILL, 1984.

[HWA 93] HWANG K., Advanced Computer Architecture, McGRAW-HILL, 1993.

[IRO 69] IRONS B.M., « A frontal solution program for finite element analysis », Int. Jour.

Num. Meth. Eng., Vol2, 1970, pp 5-32.

[KUM 94] KUMAR V., GRAMA A., GUPTA A., KARYPIS G., Introduction to parallel

computing, The Benjamin/Cummings Publishing Company, 1994.

[LET 90] LETALLEC P., DEROECK Y-H., VIDRASCU M., «Domain decomposition methods

in computational mechanics», Computational Mechanics Advances, Vol /, 2 NorthHolland,

-220.

[LET 94] LETALLEC P., Domain decomposition methods for large linearly elliptic three

dimensional problems. Rapport de recherche de l'INRIA n° 1182, 1990.

[LEW 92] LEWIS T., EL-REwiNI H., Introduction to parallel computing, Prentice-Hall

International Editions, 1992.

[LEW 93] LEWIS T., Fundations of parallel programming, IEEE Computer Society Press,

[LIO 87] LIONS P.L., Poincare-Steklov operators and domain decomposition methods,

Proceedings of the first international symposium on domain decomposition methods for

partial differential equations, Paris, 1987, Glowinski, Golub, Meurant, Periaux, eds,

SIAM, Philadelphia, 1988.

[ORT 89] ORTEGA J ., Introduction to parallel and vector solution of linear systems, Plenum

Press, 1989.

[QUA 90] QUARTERONI A., Domain decomposition method for the numerical solution of

partial differential equations. University of Minesota Supercomputing Institute Research

Report UMSI 90/246, 1990.

[ROU 89] Roux F-X., Methode de decomposition de domaine ~ !'aide de multiplicateurs de

Lagrange et application ~ Ia resolution en parall~le des equations de l'elasticite lineaire.

Th~se de l'Universite de Paris 6, 1989.

[STO 93] STONE H., High-Performance Computer Architecture, Addison-Wesley, 1993.

[TOU 89] TouzoT G., VAYSSADE M., Un syst~me de modelisation interactive. Rapport final,

Contrat MRT 86.E.0265.01., juin 1989.

Downloads

Published

1997-02-01

How to Cite

Escai, Y. ., & Touzot, G. (1997). Application des methodes de decomposition de domaines au calcul parallele. European Journal of Computational Mechanics, 6(5-6), 589–609. Retrieved from https://journals.riverpublishers.com/index.php/EJCM/article/view/3425

Issue

Section

Original Article