Model Updating by Combining Substructure Energy Functions and Subset Selection

Haishan Wu,Limin Sun

DOI: https://doi.org/10.1007/978-1-4419-9831-6_8

2011-01-01

Abstract: Since a reliable finite element model has broad applications in structural response prediction, control, health monitoring and condition assessment, model updating techniques are often applied to improve the performance of an analytical model for its intended use. The sensitivity-based iterative model updating approach becomes preferable because of its physically meaningful results, however, the selection of updating parameters remains a difficult problem to handle. In this paper, a synthesized approach, which combined substructure energy functions and subset selection, was proposed recognizing that the former can be used to locate the area containing dominant modeling errors and the latter to ensure the least number of effective parameters be chosen to reduce the modal residue adequately. This synthesized method can potentially reduce the burden of calculation but have more credible results. Its effectiveness was demonstrated systematically by numerical simulation with encouraging results.