Predicting optimal L-PBF printing location according to desired part quality: A data-driven methodology

The influence of print position on final quality cannot be neglected when parts are fabricated by laser-powder bed fusion (L-PBF) additive manufacturing technique. Several studies investigated this topic, focusing mainly on critical locations, but without mapping build platforms. Due to the availability of many L-PBF system architectures and processable materials on the market, it is imperative to define a simple, scalable, and easily adaptable mapping approach. This article thus proposes an innovative and scalable methodology to map the spatial performance of L-PBF printing platforms. The presented approach exploits experiments and natural-neighbor interpolation/extrapolation to map powder bed areas complying with two sets of dimensional and surface roughness tolerances. Regions where parts comply with desired design requirements are successively identified through classification. A demonstration is provided by applying the methodology to 316L stainless steel. The predictive maps in output reveal patterns consistent with known challenges in L-PBF additive manufacturing, underlining the importance of strategic positioning of critical components and the potential for adaptive process parameter optimization. Specifically, the predictive maps supply useful information for optimizing the print position reducing the need for post-processing. The suggested solution is practical for improving process efficiency and reducing waste and is simply adaptable to different material-machine combinations.