Skip to main navigation Skip to search Skip to main content

Structural finite element model updating by using response surfaces and radial basis functions

  • Linren Zhou
  • , Lei Wang
  • , Lan Chen
  • , Jinping Ou*
  • *Corresponding author for this work
  • South China University of Technology
  • University of Jinan
  • Dalian University of Technology

Research output: Contribution to journalArticlepeer-review

Abstract

Updating a finite element model of a large and complex structure involves a large number of updating parameters, some of which may be insensitive to structural measurements. Updating them wastes computing time and easily yields false-positive results. Model updating of response surfaces based on radial basis functions has previously been demonstrated to alleviate these problems. This study was designed to further validate the effectiveness and feasibility of the method for real-world civil engineering structures. The model updating procedure is illustrated in detail using a theoretical structure, a laboratory-scale bridge model, and an actual bridge. The key issues in selecting parameters and responses for updating are discussed. An objective function is proposed based on the differences between model predictions and measured responses. The alternatives of model updating solutions are discussed to obtain an optimal result. On the basis of the numerical simulation, experimental study, and real bridge application, the results indicate that this method works well and can easily be implemented in practice for updating models of complex bridges such as long-span, cable-stayed bridges.

Original languageEnglish
Pages (from-to)1446-1462
Number of pages17
JournalAdvances in Structural Engineering
Volume19
Issue number9
DOIs
StatePublished - Sep 2016
Externally publishedYes

Keywords

  • Bridges
  • Finite element models
  • Model updating
  • Radial basis functions
  • Response surface method

Fingerprint

Dive into the research topics of 'Structural finite element model updating by using response surfaces and radial basis functions'. Together they form a unique fingerprint.

Cite this