A Study on a Ranking Approach to Estimating Structural Damages

(Performance Evaluation for Multiple Damages)

Abstract

The damage identification analysis based on filtering theory has a potentially effectiveness as a global monitoring both for identifying the damage locations and for estimating the extent of damage. In the computation process, the natural frequencies are utilized as observation data, and then the inverse analysis are performed recurrently to obtain the estimate values of system parameters. However, it is difficult to experimentally measure a large number of natural frequencies accurately. The purpose of this study is to develop a ranking approach to estimating structural damages. In this method, it requires only a few lower natural frequencies which can be measured experimentally. Although the location and extent of damages cannot be determined, the damage potential rankings are obtained as minimum required information to detect damage locations. In this paper, for the multi damage estimation problem in a simple straight pipe structure, the validity of the developed algorithm is confirmed experimentally.