Premium
A normal modes technique to reduce the order of poroelastic models: application to 2D and coupled 3D models
Author(s) -
Dazel O.,
Brouard B.,
Groby J.P.,
Göransson P.
Publication year - 2013
Publication title -
international journal for numerical methods in engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.421
H-Index - 168
eISSN - 1097-0207
pISSN - 0029-5981
DOI - 10.1002/nme.4551
Subject(s) - poromechanics , substructure , reduction (mathematics) , inertia , set (abstract data type) , normal mode , basis (linear algebra) , model order reduction , decomposition , simple (philosophy) , mathematics , algorithm , computer science , porous medium , geometry , porosity , structural engineering , physics , engineering , classical mechanics , vibration , acoustics , projection (relational algebra) , philosophy , ecology , biology , epistemology , programming language , geotechnical engineering
SUMMARY A reduced‐order model for structures involving poroelastic materials is proposed in this paper. The approach is based on a separation of the solid and fluid phases of the porous material into separate substructures. For each individual substructure, a decoupled normal mode basis is considered, from which a set of vectors for the decomposition is selected. The preserved modes are completed by an additional family to correct for the influence of the static response of the non‐preserved. It is shown that the only neglected phenomenons in the model are the inertia of the non‐preserved modes and part of their intercoupling. The following three features render the proposed scheme computationally attractive: (i) real valued matrices are involved in the transformations; (ii) the assembly of complex, frequency dependent matrices is only performed at the stage of solving for a particular frequency; and (iii) the number of normal modes required are selected using a novel method. The computational efficacy is demonstrated, on a simple but realistic 3D case, through numerical results obtained using a reduced number of DOFs, showing a significant reduction of computational cost compared with traditional methods. Copyright © 2013 John Wiley & Sons, Ltd.