Towards material and process agnostic features for the classification of pore types in metal additive manufacturing

The manufacturing of metal parts via powder-bed fusion is often still facing quality issues due to microstructural porosity. Minimizing this porosity remains a priority and requires the optimization printing process parameters. While analysis printed using X-ray computed tomography can localize identify pore types (e.g. keyhole or lack-of-fusion pores), these be difficult across printer settings print materials. Therefore, there need for material agnostic approach. This work presents such an approach by considering set geometric features that do not differ considerably scenarios. These are then leveraged supervised type classification. distributions were analyzed in different materials under varying laser parameters, showing they behave generic way. For classification, it observed outperform other state-of-the-art classification single material, reaching up 93.0% accuracy. Additionally, accuracies 90.2% cross-material training on pores one validating another. results pave way general-purpose method usable conditions.