In recent years, owing to its capability to handle large deformations and complex material interfaces, the Smooth Particle Hydrodynamics (SPH) method has emerged as a promising numerical approach for modelling laser powder bed fusion (LPBF). However, most existing SPH-based models primarily focus on thermo-fluid coupling and are unable to capture the coupled thermo-fluid-microstructure evolution during LPBF. In this context, an integrated mesh-free computational framework is developed to simultaneously simulate melt pool dynamics and the solidification microstructure evolution. Following the model validation, the proposed framework is applied to simulate single-track LPBF of 316L stainless steel and Inconel 625. The simulations reveal that the balling phenomenon at the beginning of the laser scan can induce a local cooling direction that deviates from the global one, resulting in a corresponding deviation in grain growth direction. At the same laser energy input, a higher scanning speed can result in more pronounced balling at the beginning of the melt track, consequently producing larger columnar grains with greater inhomogeneity in the laser-grain angle. Upon full solidification of the melt pool, three melt track states, namely, balling, fluctuating, and continuous can be simulated using the developed framework. These simulated track states, together with melt pool profiles, show good agreement with three independent experimental observations reported in the literature. Overall, the results demonstrate that the proposed framework provides a promising computational tool for integrated thermo-fluid-microstructure analysis in LPBF.
Zhong et al. (Wed,) studied this question.