Crack identification in aluminium plates using Lamb wave signals of a PZT sensor network

With an integrated active piezoelectric sensor network, a Lamb wave-based crack identification technique for aluminium plates was developed. Experimental results showed that the propagation of Lamb waves in aluminium plate-like structures is considerably complicated due to wave dispersion, material attenuation, boundary reflection, etc. In order to eliminate the diverse interference, a wavelet transform technique was applied to purify the acquired Lamb wave signals, and the characteristics of Lamb wave signals were extracted from the wave energy spectrum. A correlation function was further established, which helped identify the crack position based on a triangulation approach with the aid of a nonlinear least-squares optimization algorithm. Such an approach provides satisfactory results in locating the crack position in aluminium plates with cracks of 5 and 20mm in length.