Improvement of internal defect detection accuracy using correlation processing with the emission waveform for noncontact acoustic inspection method

We have been studying the inspection of the tile outer wall using a noncontact acoustic inspection method with a sound source-mounted type UAV. However, if the position change of the sound source mounted on the UAV was large, the defect detection may not be successful. For this reason, a resonance judgement process was devised to determine whether or not the signal was actually vibrated by emitted sound waves, but there was a problem that the calculation load was rather heavy. Therefore, this time, as a faster method, a cross-correlation method between the emission waveform and the vibration waveform measured by the LDV was investigated. From the experimental results, it was clarified that using cross-correlation processing enables highly accurate defect detection even if the sound source position changes.