A systematic investigation of the effects of process parameters on heat and fluid flow and metallurgical conditions during laser-based powder bed fusion of Ti6Al4V alloy

Additive manufacturing (AM) of metals faces a growing number of applications in different industries e.g. aerospace, medical, automotive, etc. Although metal AM outweighs current conventional production methods in some certain areas, the exact effect of processing conditions on the final quality and microstructure of the parts is still not well understood. An efficient way of understanding the effect of these processing conditions on a part's quality is via a calibrated and validated numerical model. Hence, in the current work a finite element model for analyzing the heat and fluid flow along with metallurgical conditions during Laser-based Powder Bed Fusion (L-PBF) of a titanium alloy has been developed and implemented in the commercial software code COMSOL Multiphysics. The thermal effect of the laser is modelled via a novel conico-Gaussian moving heat source, based on the concept of modified optical penetration depth. Analytical expressions for the geometrical distribution of the heat source are derived to obtain the heat source's effective depth. The model has been both verified and validated through mesh sensitivity analysis and comparison with experimental results. Furthermore, a detailed description about the role of the various driving forces for fluid flow has been given based on a thorough analysis using relevant dimensionless numbers. A systematic procedure to study the influence of neglecting the fluid flow inside the melt pool on the thermal field has also been devised. Moreover, a parametric study has been carried out to understand the effect of varying beam size and laser travel speed on heat and fluid flow conditions along with the final microstructures. The results show that changing the beam size or travel speed highly influences the grain sizes, dendritic growth directions and also the grain morphologies. To study the metallurgical conditions of the process, a microstructural sub-model has been developed. It is shown that by choosing different process parameters, one can manipulate the direction of the dendritic growth and change the grain sizes. Specifically, it is found that the overall effect of changing beam size on grain morphology is less pronounced than changing the travelling speed. (C) 2019 Elsevier Ltd. All rights reserved.