Structural damage localization in beam-like structures using Multi-channel Empirical Mode Decomposition of random response

A lot of researchers have focused on vibration-based structural damage detection in last few decades. Most of proposed methods are based on modal parameters. Due to difficulties in performing modal tests, response- based methods have got much more attention. In this paper, a new technique for processing random vibrational response with the aim of structural damage localization is introduced. The technique is based on Multi-channel Empirical Mode Decomposition of random response of the structure. The main advantage of Multi-channel Empirical Mode Decomposition over traditional Empirical Mode Decomposition is its ability to extract consistent Intrinsic Mode Functions in the case of multi-channel measurements. Regarding the fact that a local stiffness reduction in a beam causes an abrupt change in the curvature of deformation, spatially distributed Intrinsic Mode Functions are employed for damage localization. The proposed method has been carried out on a finite element model of a damaged cantilever beam. To investigate the efficiency and robustness of the method, the effect of number of measurement points, damaged zone location as well as damage intensity have been studied in detail. The results in all cases are satisfactory.