Premium
Parallel finite element solution of three‐dimensional Rayleigh–Bénard–Marangoni flows
Author(s) -
Carey G.F.,
Mclay R.,
Bicken G.,
Barth B.,
Swift S.,
Ardelea A.
Publication year - 1999
Publication title -
international journal for numerical methods in fluids
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.938
H-Index - 112
eISSN - 1097-0363
pISSN - 0271-2091
DOI - 10.1002/(sici)1097-0363(19990915)31:1<37::aid-fld954>3.0.co;2-s
Subject(s) - marangoni effect , computation , computer science , finite element method , node (physics) , scalability , computational science , domain decomposition methods , parallel computing , flow (mathematics) , algorithm , mechanics , topology (electrical circuits) , physics , convection , mathematics , acoustics , thermodynamics , database , combinatorics
A domain decomposition strategy and a parallel gradient‐type iterative solution scheme have been developed and implemented for the computation of complex three‐dimensional viscous flow problems involving heat transfer and surface tension effects. Special attention has been paid to the kernels for the computationally intensive matrix–vector products and dot products, to memory management, and to overlapping communication and computation. Details of these implementation issues are described together with associated performance and scalability studies. Representative Rayleigh–Bénard and microgravity Marangoni flow calculations and performance results on the Cray T3D and T3E are presented. Performance studies have been recently carried out and sustained rates above 50 gigaflops and 100 gigaflops have been achieved on the 512‐node T3E‐600 and 1024‐node T3E‐900 configurations respectively. The work is currently being extended to tightly‐coupled parallel ‘Beowulf‐type’ PC clusters and some preliminary performance results on this platform are presented. Copyright © 1999 John Wiley & Sons, Ltd.