Abstract

In metal laser powder bed fusion (L-PBF), the manufacturing process thermomechanical entails interactions complex that yield deviations from the nominal geometry. While these deviations are typically unwanted, not all are necessarily detrimental to the functionality of the part. In this paper, we describe a multisensing approach, where layer-by-layer in process and post-process measurements are collected as an initial step to reliably discriminate harmful defects from neutral anomalies in a
manufactured part. Focus is particularly placed on in-situ measurements, with an account of initial results and the data fusion strategy.