Crack Formation Detection during Ultra-Fast Laser Beam Machining of Alumina Ceramic Based on Acoustic Emission Signals

Cracks are usually formed during ultra-fast laser beam machining of alumina ceramics. In this study, acoustic emission technology is adopted to monitor the cracks formation in the process. The root-mean-squares (RMS) and measured areas under rectified signal envelope (MARSE) of acoustic emission signals are analyzed. For detecting crack forming real-timely, a novel method based on continuous wavelet transform (CWT) is proposed. Firstly, the frequency distributions of acoustic emission signals are extracted with CWT. Then a 5-dimensional feature vector is constructed to describe the frequency distributions. Finally, support vector machine method is adopted to identify the crack formation. The results demonstrate that the identification accuracy of the proposed method achieves 95.8%.