Steel Shearing Acoustic Emission Signal Identification Using Wavelet Scattering On Micro Forming Irwan Setyanto, Gandjar Kiswanto, Sugeng Supriadi, Siska Titik Dwiyati
University Of Indonesia
Abstract
Real-time defect detection is crucial for advanced in-process quality inspection systems, particularly in addressing intricate and challenging cracks within small-scale products through automation. Acoustic emission (AE), a potent non-destructive technique, offers the capability for rapid crack initiation detection. To advance this detection capability in micro-scale manufacturing, our study investigated acoustic emissions generated during the micro-blanking process and try to recognized it. Using a proprietary 5 kN micro-forming machine, experiments were conducted on 0.1 mm SK-5. Comparison of various signal processing methods is done using Fast Fourier Transform (FFT) , Short-Time Fourier Transform (STFT) , and the wavelet scattering transform (WST). With similar result from STFT and WST, the latter offers a richer feature set that is invariant and stable against small shifts and noise, making it well-suited for AE signal classification tasks where the differences can be nuanced and involve complex time-frequency patterns.
Keywords: Acoustics Emission- micro forming- crack monitoring- energy saving
