To prevent failures of operating mechanical components and structural members, it is important to investigate the condition of materials and structures by non-destructive inspection. Acoustic emission (AE) testing is a non-destructive evaluation technique that detects elastic waves generated by crack propagation or fracture inside materials using sensors, and evaluates the integrity or fracture behavior based on waveform analysis. In particular, for thin plate structures, AE waves propagate as Lamb waves whose propagation velocities depend on frequency, enabling long-distance monitoring. AE sensors are designed to have specific resonance frequencies to detect weak AE waves with high sensitivity. However, due to their resonance characteristics, the detected waveforms are significantly distorted according to the frequency characteristics, making it difficult to directly compare waveforms measured by different sensors. In this study, a time-dfrequency domain filtering method using inverse continuous wavelet transform based on theoretical group velocity dispersion of Lamb waves was developed to compensate for waveform distortion caused by the resonance characteristics of AE sensors. The proposed method was applied to experimental AE waveforms, and it was confirmed that waveform distortion due to sensor resonance could be reduced, enabling comparison of waveforms obtained from sensors with different resonance characteristics.
Yamaguchi et al. (Wed,) studied this question.