Passive detection and localization of fatigue cracking in aluminum plates using Green's function reconstruction from ambient noise.

Recent theoretical and experimental studies have demonstrated that a local Green's function can be retrieved from the cross-correlation of ambient noise field. This technique can be used to detect fatigue cracking in metallic structures, owing to the fact that the presence of crack can lead to a change in Green's function. This paper presents a method of structural fatigue cracking characterization method by measuring Green's function reconstruction from noise excitation and verifies the feasibility of crack detection in poor noise source distribution. Fatigue cracks usually generate nonlinear effects, in which different wave amplitudes and frequency compositions can cause different nonlinear responses. This study also undertakes analysis of the capacity of the proposed approach to identify fatigue cracking under different noise amplitudes and frequency ranges. Experimental investigations of an aluminum plate are conducted to assess the cross-correlations of received noise between sensor pairs and finally to detect the introduced fatigue crack. A damage index is proposed according to the variation between cross-correlations obtained from the pristine crack closed state and the crack opening-closure state when sufficient noise amplitude is used to generate nonlinearity. A probability distribution map of damage is calculated based on damage indices. The fatigue crack introduced in the aluminum plate is successfully identified and oriented, verifying that a fatigue crack can be detected by reconstructing Green's functions from an imperfect diffuse field in which ambient noise sources exist locally.