Multi-layer thermo-fluid modeling of powder bed fusion (PBF) process

Abstract

In this study, a 3D thermo-fluid computational method has been developed and employed to simulate the laser powder bed fusion (PBF) process. The particular objective is to demonstrate the feasibility of multi-layer simulations of the laser PBF process and to discuss the potential and challenges of this approach. The approach includes using the discrete element method (DEM) to simulate particle spreading on a powder bed, as well as computational fluid dynamics (CFD) and heat transfer to simulate laser-powder/matter interactions, in a sequential manner to about 10 layers only with a small scan area, limited by currently available computational power. The simulation results can offer insight such as melt pool shapes and sizes, also solidified surface morphology along different build layers. The model also includes a surface tracking algorithm to account for the formation of voids and lack-of-fusion pores. In addition to achieving the main objective of this study, i.e., feasibility demonstration of this computational process, it is also noted that, for the case studied, the defects in a former layer may shrink, or even vanish, due to the extra thermal energy received from laser scanning of subsequent layers. This study has successfully demonstrated the sequential linkage between a discrete element method and a thermo-fluid model in a multi-layer deposition fashion and the experimental validation will be performed in future work. With significantly greater computational capabilities in the future, the developed method may potentially be utilized as a means to understand physical phenomena and select key process parameters for PBF fabrications of small features.