Characterization of machine noise signals during L-PBF for online monitoring using gas-borne acoustic emission

Department of Mechanical and Mechatronic Engineering, Central University of Technology, South Africa, Free State, karabomoore@gmail.com, dkouprianoff@cut.ac.za, iyadroitsava@cut.ac.za, iyadroitsau@cut.ac.za

^* Corresponding author: karabomoore@gmail.com

Abstract

The metal laser powder bed fusion (L-PBF) technology uses a layer-by-layer manufacturing technique. During the build process, various acoustic emission (AE) signals are emitted by the machine components inside the build chamber: which include AE signals such as that of the movement of the build platform, the powder delivering system, the inert gas flow, and laser scanning. In this work, the machine AE signals recorded from a microphone are characterised, studied, and labelled as noise signals to provide insights for monitoring of defects such as cracks using the EOS M280 L-PBF system. The frequency and time domain features of the machine AE signals, such as the fast Fourier transform, root mean square and signal-to-noise ratio, were used to indicate the machine AE signals peak frequencies, loudness, and effect of the applied filter on the AE signals. It is also shown how that the data obtained can further be used for when selecting appropriate signal conditioning parameters for defect monitoring of the crack and delamination signals during the build process.