In-situ camera-based condition monitoring for wear classification and quantification on scanner protective windows in laser powder bed fusion (PBF-LB/M)

This work presents a camera-based condition monitoring system for scanner protective windows (SPWs) in laser powder bed fusion of metals (PBF-LB/M). Methods for classification of dust and fiber particles of an area equivalent diameter x A r e a ≥ 0.030 m m , scorch marks, and metal condensate contamination are described. To quantify the effect of metal condensate contamination on relevant laser beam parameters and correlate it with SPW condition monitoring images, the transmissivity of worn SPWs, wear-induced changes in measured beam caustics, and transient thermally induced effects on the power density distribution are investigated. In this work, SPW wear results in a transmissivity loss of up to 5.66 % and a peak intensity reduction of up to 56.4 % . Moreover, even moderate condensate coverage induces transient focus shifts lasting up to 17 s . In experiments with a dynamically deflected laser beam, the focus position shifts by up to 8.3 m m within only 1 s of exposure. Based on these measurements, SPW images are labeled and a critical metal condensate coverage is derived as an alert criterion. The proposed image processing and labelling thereby enables in-situ quantification of SPW wear progression during ongoing PBF-LB/M processes.