Residual stress development in selective laser-melted Ti6Al4V: a parametric thermal modelling approach

Abstract

High cooling rates within the selective laser melting (SLM) process can generate large residual stresses within fabricated components. Understanding residual stress development in the process and devising methods for in-situ reduction continues to be a challenge for industrial users of this technology. Computationally efficient FEA models representative of the process dynamics (temperature evolution and associated solidification behaviour) are necessary for understanding the effect of SLM process parameters on the underlying phenomenon of residual stress build-up. The objective of this work is to present a new modelling approach to simulate the temperature distribution during SLM of Ti6Al4V, as well as the resulting melt-pool size, solidification process, associated cooling rates and temperature gradients leading to the residual stress build-up. This work details an isotropic enhanced thermal conductivity model with the SLM laser modelled as a penetrating volumetric heat source. An enhanced laser penetration approach is used to account for heat transfer in the melt-pool due to Marangoni convection. Results show that the developed model was capable of predicting the temperature distribution in the laser/powder interaction zone, solidification behaviour, the associated cooling rates, melt-pool width (with 14.5% error) and melt-pool depth (with 3% error) for SLM Ti6Al4V. The model was capable of predicting the differential solidification behaviour responsible for residual stress build-up in SLM components. The model-predicted trends in cooling rates and temperature gradients for varying SLM parameters correlated with experimentally measured residual stress trends. Thus, the model was capable of accurately predicting the trends in residual stress for varying SLM parameters. This is the first work based on the enhanced penetrating volumetric heat source, combined with an isotropic enhanced thermal conductivity approach. The developed model was validated by comparing FEA melt-pool dimensions with experimental melt-pool dimensions. Secondly, the model was validated by comparing the temperature evolution along the laser scan path with experimentally measured temperatures from published literature.