Premium
Application of Multipreconditioned Iterative Algorithms in Dual Domain Decomposition Methods for Structural Dynamics
Author(s) -
Leistner Michael,
Rixen Daniel,
Gosselet Pierre
Publication year - 2017
Publication title -
pamm
Language(s) - English
Resource type - Journals
ISSN - 1617-7061
DOI - 10.1002/pamm.201710134
Subject(s) - tearing , domain decomposition methods , dual (grammatical number) , measure (data warehouse) , simple (philosophy) , computer science , algorithm , decomposition , domain (mathematical analysis) , basis (linear algebra) , finite element method , minification , process (computing) , selection (genetic algorithm) , space (punctuation) , mathematical optimization , mathematics , data mining , artificial intelligence , engineering , structural engineering , mathematical analysis , mechanical engineering , chemistry , geometry , art , philosophy , literature , epistemology , organic chemistry , operating system
We apply a multipreconditioned domain decomposition method based on Finite Element Tearing and Interconnecting to linear structural dynamics with highly heterogeneous material properties. A recently published method to build and select the multiple directions is presented and applied. We hint at possible problems when localized phenomena are considered and present a new simple measure that extends the selection process in existing algorithms. We show numerical results and conclude how to effectively prevent a possible degeneration of the minimization space basis. (© 2017 Wiley‐VCH Verlag GmbH & Co. KGaA, Weinheim)