Calibration-Less Finite Element Model of the Laser Powder Bed Fusion Process

Abstract

The melt pool characteristics govern process stability and microstructure development during Laser Powder Bed Fusion (LBPF). The melt pool is strongly affected by the laser parameters and based on the melt pool depth-to-width ratio, three melting modes can be distinguished: conduction, transition, and keyhole mode. Finite element (FE) simulations have been extensively used to investigate the cyclic thermal history of the LPBF process and the effects of different process parameters on the part quality. However, the heat sources applied in such simulations can only accurately predict the melt pool behavior, and thus the temperature profile, when processing in conduction mode. As for the transition and keyhole regime modeling, an experimentally based heat source calibration is usually performed to make the simulated melt pool dimensions match the real values. In this study, a novel calibration-less finite element model is proposed to simulate the LBPF process across different melting regimes without the need for experimental input. The predicted melt pool dimensions are validated for a wide range of process parameters applied to three different alloys: Ti-6Al-4V, 316L austenitic stainless steel (316L), and 2507 super duplex stainless steel (SDSS). The average errors obtained on the melt pool width, depth, length, and aspect ratio are 9.7 %, 12.4 %, 14.5 %, and 14.3 % respectively. These results demonstrate how the proposed approach can capture the melt pool dimensions throughout different materials and melting modes without relying on any experimental input data.

Keywords: Additive manufacturing, laser powder bed fusion, melt pool, Finite Element Modeling, Calibration-less heat source