Inspection of disbonds in multilayer dissimilar metal structure using lock-in thermography

This paper reports the characterization of disbonds between a steel plate and a lead plate on the surface of the lead using lock-in thermography. Based on the photothermal model, the excitation frequency of the bonding specimen with the steel plate and the lead plate is optimized. A lock-in thermography testing system is established to inspect the specimen. The Fourier transform method is used to process the infrared image sequences. To improve the signal-to-noise ratio (SNR), a fuzzy c-means (FCM) algorithm is used to process the phase matrix. The influence of number of clusters on the processing results is researched, and the optimal value of clusters number is obtained. The shearing phase technique is used to evaluate the size of the disbonds quantitatively. The measurement results of the artificial disbonds in the specimen show good agreement with the actual values. The results prove that FCM is effective in enhancing the SNR of the phase image, which makes it feasible for the quantitative determination of defect size by the shearing phase technique.