The eigenvalue and eigenvector errors using model reduction techniques

摘要

Methods of reduced order model are widely applied in reduction of degree of freedom of large finite element models, the static condensation method used for fast computation of natural frequencies and mode shapes exactly in low frequency. The improve reduce system (IRS) method produces a reduced model which more accurate estimated the modal model of the full system, Rayleigh Ritz vector is a most general technique for finding approximations to the lowest eigenvalue and corresponding eigenvectors. In this paper a comparison between IRS, Ritz and condensation is presented, numerical example applied for three methods, IRS is strongly has agreement with FEM.