Premium
Component mode synthesis with constant mass and stiffness matrices applied to flexible multibody systems
Author(s) -
Gerstmayr J.,
Ambrósio J. A. C.
Publication year - 2007
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.2133
Subject(s) - linearization , multibody system , factorization , stiffness , mass matrix , deformation (meteorology) , matrix (chemical analysis) , diagonal , stiffness matrix , constant (computer programming) , rigid body , rotation (mathematics) , control theory (sociology) , mathematics , mathematical analysis , computer science , geometry , algorithm , structural engineering , physics , classical mechanics , engineering , nonlinear system , materials science , artificial intelligence , composite material , control (management) , quantum mechanics , nuclear physics , programming language , meteorology , neutrino
In flexible multibody systems, component mode synthesis is used to reduce the size of system matrices of the single bodies from millions to a few hundred, while maintaining the full coupling between body deformation and overall rigid body motion. The combination of clamped eigenmodes and static deformation modes results in a dense and non‐diagonal structure of the system matrices which is computationally costly. In the present paper, the conventional concept of the floating frame of reference is transformed to a description based on absolute coordinates and a corotational linearization of the deformation. The constant mass matrix and the corotated stiffness matrix need to be factorized only once for the whole simulation. In a planar example, component mode synthesis isapplied to this absolute coordinate formulation by using twice as many deformation modes as usual, while the solution of the overall system becomes less expensive due to efficient factorization. The size of the system matrices needs to be increased considerably for each axis of rotation. However, the computational complexity for the factorization reduces from cubic to quadratic. Numerical examples are presented in order to demonstrate the difference between standard procedures and the methodology now proposed. Copyright © 2007 John Wiley & Sons, Ltd.