A Structural Damage Identification Procedure With Application To A Frame Structure With Bolted Joints

L Yu,M Link,Ss Law,Lm Zhang
1999-01-01
Abstract:A procedure is presented for damage identification of structures. It includes three parts : (1) sensor location determination, (2) mode shape expansion and (3) damage detection. At first sensor locations are selected according to the Fisher information matrix using analytical mode shapes of a refined Finite Element Model (FEM) of the structure, then measured mode shape data at the selected sensor locations (DOFs) are expanded to the unmeasured DOFs of the FE model by using three different methods. Finally the damage of structure is detected by means of the "Damage Vector (D.V.)/ Damage Angle (D.A.)" concept. The capability of this identification process is demonstrated in cases with or without noise on simulated test data and also by using real test data obtained from an experimental frame structure with bolted joints.
What problem does this paper attempt to address?