Acoustic Diagnostic of Laser Powder Bed Fusion Processes

Online monitoring of Laser Powder Bed Fusion is critical to advance the technology and its applications. Many studies have shown that the acoustic signal from the laser powder bed fusion process contains a large amount of information about the process condition. In this research, we used an acoustic system for the in-situ characterization of a wide variety of different single-track geometries. The internal acoustic system includes a microphone and accelerometer. The melting mode, cross-sectional shape and dimensions of Ti6Al4V single tracks at different process parameters are presented. We have established a correlation between track geometry, internal defects and acoustic signals. The parameters are varied and tested against the acoustic frequency measurements to determine the sensitivity. We determined the patterns of signal behaviour in the event of anomalies (spatter, balling, pores, undercut). The characteristic features of the process are traced to a commercial machine. Well described dataset with correlated monitoring data and signal tracks properties obtained and can be used for building classification model and quality prediction. All this is aimed at creating a database of experimental data that will be a key for LPBF digitalization and control, allowing real-time control of the process to optimize part quality and, more importantly, help with decision-making algorithms.