High Fidelity Modeling of Melt Pool Dynamics in Laser Powder Bed Fusion

Laser powder bed fusion (LPBF) part quality is tightly coupled to melt pool behavior, which governs porosity, lack of fusion, and keyhole defects. However, most process windows are still tuned by trial and error, with limited understanding of how process and modeling parameters jointly influence melt pool dynamics.

This presentation introduces a high-fidelity thermofluid model for LPBF that extends a hybrid viscosity continuum framework to resolve melt pool geometry and pore evolution under realistic processing conditions. The model incorporates temperature-dependent material properties and explicitly varies key modeling parameters such as critical solid fraction, dendritic arm spacing, and viscosity formulation, together with process parameters including laser power and scan speed. Three-dimensional simulations are performed, and the predicted melt pool dimensions are quantitatively compared to experimentally measured tracks.

The results demonstrate good agreement with measured melt pool width and depth while capturing the onset and evolution of pores within the melt pool. Sensitivity studies show that changes in critical solid fraction have a limited influence on overall melt pool dimensions but affect free surface morphology, whereas laser power and scan speed strongly modify melt pool shape, solidification behavior, and pore formation.

The talk will focus on how these insights can narrow process parameter search spaces, reduce porosity-related defects, and support physics-informed approaches to process planning and qualification in LPBF, independent of any specific machine or vendor.