Fast dynamic solver for the design of bridge using a model order reduction

Mitsuteru Asai, Hiroki Iida, Norliyati M. Amin, Yoshimi Sonoda

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

A model order reduction via Krylov subspace has been utilized as a booster of the finite element analysis in this paper. In the seismic design of structures, low frequency responses up to about 20Hz should be accurately predicted. A stepwise iterative frequency response analysis is utilized to evaluate an effective reduced order to maintain the accuracy at the target frequency range. In the frequency analysis, simple error estimation is performed to check the comparison is supported by the Moment Matching (MM), which is one of the important approximation properties of the KS-MOR. In the numerical example, a 3D bridge model with different bearing system was solved to show performance of KS-MOR. One of the bearing system is classical steel hinge, the other is steel laminated rubber bearing. The KS-MOR can apply to the general dynamic structure analysis including the damping matri. The efficiency in the CPU time is about 20-30 times faster than the conventional FEM calculation.

Original languageEnglish
Title of host publicationProceedings of the 9th International Conference on Shock and Impact Loads on Structures
Pages225-234
Number of pages10
Publication statusPublished - 2011
Event9th International Conference on Shock and Impact Loads on Structures - Fukuoka, Japan
Duration: Nov 16 2011Nov 18 2011

Publication series

NameProceedings of the 9th International Conference on Shock and Impact Loads on Structures

Other

Other9th International Conference on Shock and Impact Loads on Structures
Country/TerritoryJapan
CityFukuoka
Period11/16/1111/18/11

All Science Journal Classification (ASJC) codes

  • Civil and Structural Engineering

Fingerprint

Dive into the research topics of 'Fast dynamic solver for the design of bridge using a model order reduction'. Together they form a unique fingerprint.

Cite this