In-situ monitoring of heat propagations in electron beam powder bed fusion

In-situ sensing and monitoring techniques for characterizing heating and cooling patterns in powder bed fusion have attracted significant attention in the additive manufacturing community. Studies on thermal gradients and heat accumulation aim to detect local anomalies and predict the quality characteristics of fabricated parts. Most existing methods focus on spatial and temporal heat maps within the exposed layer area. However, heat dissipation into the surrounding powder is also critical. Indeed, variations in heat transfer between the part and the surrounding powder can adversely affect surface quality, dimensional accuracy, and microstructural uniformity of the final product. In electron beam powder bed fusion (EB-PBF), the heat transfer and dissipation in the pre-sintered powder shall be maintained as homogenous and uniform as possible to achieve the target quality and mechanical performance. We explore opportunities and methods in the field of in-situ monitoring of the EB-PBF thermal history. We also present a novel approach for in-line monitoring of the heat propagation beyond the exposed area. It integrates a spatio-temporal model of the thermal history with a non-linear low-dimensional learning technique to effectively distinguish between normal and anomalous heat propagation patterns. A case study on pure copper in EB-PBF is presented. Pure copper, with its contrast in thermal conductivity between the consolidated material and the surrounding powder, is prone to defect formation and underscores the need for advanced in-situ thermal monitoring approaches.