Acoustic Emission-Based Weld Crack In-situ Detection and Location Using WT-TDOA

Abstract

As one of the essential infrastructures in rockets, the aluminum alloy tanks are mostly produced by tungsten argon arc welding. Considering the alternating load and concentrated stress through the operation of tanks, weld cracks have become one of the most severe hazards that can compromise the normal production. Therefore, the in-situ detection and accurate location for weld cracks are of great importance. With the high sensitivity and nondestructive way of testing, acoustic emission (AE) detection is suitable for weld cracks’ detection and location of the aluminum alloy tanks. A time difference of arrival (TDOA) method based on wavelet transform (WT) is proposed in this paper to enable the accurate location of weld cracks at the beginning of its appearance. The research was carried out on an AE experimental platform based on a scale-down aluminum alloy tank of rockets. The propagation speed of AE signals on the sidewall of the tank was measured after its attenuation characteristics were analyzed first and the noise was restrained based on wavelet threshold denoising. According to the time-domain invariability of WT, the component that accounts for the largest share of energy in wavelet decomposition was extracted to represent the raw AE signal for signal source location. Location errors on height and angle were calculated after three times repeated experiments. The results of the study show that TDOA based on wavelet transform (WT-TDOA) outperformed the conventional TDOA significantly in terms of location accuracy.