Abstract

Proposed modifications to two existing algorithms, based on mode shape, for locating of damage in a beam like structure model are presented in this paper. The Curvature Damage Factor (CDF) developed by Wahab and Roeck (1999) computes the change in mode shape curvature between two sets of mode vectors, that is, undamaged and damaged conditions. The local stiffness indicator (LSI) was developed by Ismail and Abdul Razak (2006) and based on the fourth derivative of the mode shape. The proposed modifications to the mathematical form of the original algorithms aim to improve its sensitivity and overcome its drawback. In order to verify the suitability and necessity for implementing the modification to the existing algorithms, eigenvalue analyses on a finite element model of a beam-like structure model were carried out and the eigenvectors for different cases were obtained. The proposed modified forms of the algorithms exhibited better sensitivity for detecting damage location in addition to the anomalies at the supports being eliminated. The modified algorithms are able to detect the damage wherever its location, applying even to cases of multi damage locations. It is also concluded that the modified algorithms have the sensitivity to detect the damage regardless of its severity. Moreover, procedures for elimination of anomalies have been proposed. The first procedure is based on eliminating the algorithms values at the degrees of freedom which match the nodes along the beam length and it has been applied to all the data presented in this paper. The second procedure is based on eliminating the damage algorithm’s values at the supports and this has been applied to the original form of the LSI. Finally, statistical anomalies elimination (SAE) procedure has been proposed and applied to the cases of anomalies along the entire beam length. The SAE elimination procedure has helped to improve the sensitivity by suppressing the anomalies along the beam length.

Key words: Curvature damage factor, modified curvature damage factor, local stiffness indicator, modified local stiffness indicator, statistical anomalies elimination.