Effect of surface geometry on laser powder bed fusion defects

Abstract

Laser Powder Bed Fusion (L-PBF) is an additive manufacturing (AM) technology which has enabled the fabrication of highly complex structures and components. A major challenge in producing these high complexity structures is predicting where defects are likely to occur, and how severe those defects will be. In this study build geometry is linked to manufacturability, using quantifiable AM defects and metrics. This was achieved by fabricating a variety of unsupported, thin-walled samples with different curvatures using Al-10Si-Mg alloy. A combination of X-Ray computed tomographic (XCT) and digital stereomicroscopy were used to capturing positional data about the total geometric error, surface quality and roughness metrics (Ra and Rz). This dataset was used to create empirical model which predicts geometric error based on the local curvature, orientation and thickness of the computer-aided-design (CAD) geometry. Local inclination angle was observed to have the strongest effect on manufacturability, heavily impacting the geometric error and surface quality metrics. Sample shape defined by curvature was observed to be a secondary factor, which became important in areas of low inclination angle. The empirical model was applied in a case study of a cellular structure, the double gyroid, which successfully identified problematic regions of the structure. The approach of linking fundamental geometry features to defects shown in this study provides additional guidance in design for AM, with potential applications in topology optimisation, optimising part orientation to reduce support, and for design of other intricate components such as the internal conduits of heat exchange devices, where supports are not feasible to use.