Computational & Technology Resources
an online resource for computational,
engineering & technology publications
Civil-Comp Proceedings
ISSN 1759-3433
CCP: 20
ADVANCES IN PARALLEL AND VECTOR PROCESSING FOR STRUCTURAL MECHANICS
Edited by: B.H.V. Topping and M. Papadrakakis
Paper III.3

An Implementation of a Generalized Lanczos Procedure for Structural Dynamic Analysis on Distributed Memory Computers

D.R. Mackay* and K.H. Law#

*Intel Corporation, Oak Ridge National Laboratory, Oak Ridge, United States of America
#Department of Civil Engineering, Stanford University, Stanford, United States of America

Full Bibliographic Reference for this paper
D.R. Mackay, K.H. Law, "An Implementation of a Generalized Lanczos Procedure for Structural Dynamic Analysis on Distributed Memory Computers", in B.H.V. Topping, M. Papadrakakis, (Editors), "Advances in Parallel and Vector Processing for Structural Mechanics", Civil-Comp Press, Edinburgh, UK, pp 93-105, 1994. doi:10.4203/ccp.20.3.3
Abstract
Structural dynamic analysis often involves the solution of the generalized eigenvalue problem. This paper describes a parallel implementation of the Lanczos method, which has rapidly become the preferred solution method for the eigenvalue problems. We review a parallel sparse matrix factorization scheme implemented on a distributed memory computer and propose a strategy to facilitate the solution of triangular systems. Furthermore, different parallel strategies for mass matrix-vector multiplication and how they can be implemented effectively in the Lanczos procedure are discussed. The Lanczos procedure implemented includes partial and external selective reorthogonalizations and spectral shifts. Numerical results on an Intel's iPSC/860 parallel computer are presented to examine the effectiveness of the implementation.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description