Enhanced damage imaging of a metallic plate using matching pursuit algorithm with multiple wavepaths

Abstract

Waveforms received by sensors resulting from multiple wavepaths overlap and are hard to interpret. Because of this difficulty, they are usually intentionally ignored, thereby only the first arrival of wave mode being used for damage localization. This article proposes an imaging algorithm for damage localization by incorporating multiple wavepaths using piezoelectric wafers affixed on a metallic plate. Matching pursuit (MP) algorithm to enhance image quality is adopted for separating each wave packet individually. MP algorithm is an adaptive time-frequency signal decomposition technique that matches the best-fit elementary atom functions from an overcomplete dictionary. This study proposes a new dictionary composed of atom functions that constitute possible wave packets propagated by an excitation of Hann-windowed toneburst. The proposed dictionary converges faster and separates individual wave packets more accurately than typical Gaussian based dictionaries. Simulated studies first confirm the performance of MP algorithm with the proposed dictionary in comparison with those using conventional non-adaptive time-frequency analysis as well as MP with heuristic Gaussian-based dictionaries. The results of this study validate the proposed algorithm that multiple wavepaths can localize the damage with three to four piezoelectric wafers versus typical approaches employing only primary scattered waves.