Process Optimization By Multiscale Modeling To Minimize Residual Stress In Powder Bed Fusion

Abstract

Metal additive manufacturing often uses powder bed fusion (PBF), a renowned technology that selectively fuses metal powder particles in a bed using a laser or electron beam to create three dimensional objects. The metal powder exposed to the laser undergoes enormous temperature and phase change variations in a short period of time during PBF, resulting in undesired thermal stresses known as residual stresses. To quantify these stresses, the bridge curvature method (BCM) was applied. Multiscale modelling using adaptive coarsening was used to predict distortions based on experimentally validated models. Taguchi and Response Surface Method (TM and RSM) were used to minimize residual stress in stainless steel 316L. Based on optimal parametric results for minimal residual stress from part-scale simulation and statistical techniques, the parts were printed avoiding costly experiments. There was a minimum 8% error between optimized predicted and experimental results. The approach used was critical in lowering computational printing expense. The effects of individual parameters and their combinations in terms of energy density on residual stress were also analyzed. The relationship between residual stress, hatch spacing, scanning speed, and power in metal additive manufacturing can be characterized by an initial increase in residual stress, followed by a decrease as hatch spacing and scanning speed are increased, while power is also increased. The effect of beam diameter is very nominal and diminishes in comparison with energy density parameters.