Acoustic emission signal analysis using wavelet synchrosqueezing transform for incipient damage detection

In this study, acoustic emission (AE) is employed to detect incipient damage in polycarbonate, while wavelet synchrosqueezing transform (WSST) enhanced time-frequency resolution, enabling accurate characterization and interpretation of AE signals. Quasi-static mode-I experiments are performed using tapered double cantilever beam (TDCB) specimens with a 25 mm pre-crack in displacement control mode. The test results reveal low-frequency AE signals (25–30 kHz) appeared during elastic loading, while high-frequency components (65–120 kHz) indicated localized yielding near the crack tip. Scattered frequencies emerged just before crack initiation due to fracture processes. In contrast to the continuous wavelet transform (CWT), WSST sharpens the time-frequency representation by limiting energy smearing. The WSST achieves higher time-frequency resolution by squeezing and reconstructing the CWT coefficients along the frequency axis. This improved frequency localisation enables the accurate tracking of transient features, particularly during localised yielding and incipient crack propagation, as observed in this study. To complement the experimental results, finite element analysis (FEA) is performed to simulate the stress and strain fields near the crack tip by using an elastic-perfectly plastic model in ABAQUS. The evolving plastic deformation near the crack tip shows good correlation with cumulative AE energy. This study offers a promising technique to detect incipient failure.