Candidate Reduced Order Models for Structural Parameter Estimation

This paper considers the reduction of high order models generated by a finite element analysis. The aim is to provide a low order model which retains the effect of parameter changes and so may be used to update the unknown physical structural parameters of the modelled system in the time or frequency domains. After reviewing the methods of structural parameter identification, the available order reduction algorithms and their properties are introduced. Modal truncation is determined to be the most suitable method and the standard algorithm is extended to accommodate unknown parameters. Finally the effect of the reduction process on the modelled receptances is demonstrated.