Premium
A scalable dual‐primal domain decomposition method
Author(s) -
Farhat Charbel,
Lesoinne Michael,
Pierson Kendall
Publication year - 2000
Publication title -
numerical linear algebra with applications
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.02
H-Index - 53
eISSN - 1099-1506
pISSN - 1070-5325
DOI - 10.1002/1099-1506(200010/12)7:7/8<687::aid-nla219>3.0.co;2-s
Subject(s) - domain decomposition methods , discretization , scalability , partial differential equation , dual (grammatical number) , supercomputer , finite element method , mathematics , decomposition , iterative method , mathematical optimization , parallel computing , decomposition method (queueing theory) , scale (ratio) , domain (mathematical analysis) , computer science , construct (python library) , computational science , algorithm , mathematical analysis , discrete mathematics , structural engineering , art , ecology , physics , literature , database , biology , programming language , quantum mechanics , engineering
We blend dual and primal domain decomposition approaches to construct a fast iterative method for the solution of large‐scale systems of equations arising from the finite element discretization of second‐ and fourth‐order partial differential equations. We show numerically that our method is scalable with respect to the mesh size, the subdomain size, and the number of elements per subdomain. We apply it to the solution of several realistic structural mechanics problems, and report on parallel performance results obtained on an Origin 2000 system, as well as the ASCI Option Red supercomputer. Copyright © 2000 John Wiley & Sons, Ltd.