Multi-physics simulation for predicting surface roughness of laser powder bed fused parts after laser polishing

Laser powder bed fusion (L-PBF) uses a controlled laser beam to melt specific regions of a metal powder bed in a layer-by-layer fashion to fabricate parts with an intricate geometry. However, due to the stochastic nature of the L-PBF process, many defects may occur during the build process, including distortion, porosity, and high surface roughness. A poor roughness of the upper surface is frequently associated with impaired mechanical properties and a lower corrosion resistance. Thus, laser polishing (LP) is commonly employed to smooth the surface of the component following the build process. The surface finish of the polished part is dependent not only on the initial morphology of the surface, but also the processing conditions employed in the polishing process (i.e., the laser power, scanning speed, and hatching space). The surface profile is also influenced by physical phenomena such as the surface tension force, recoil pressure, and Marangoni force. The present study thus proposes an integrated framework based on discrete element method (DEM) and computational fluid dynamics (CFD) simulations which takes account of all of these factors to predict the final surface morphology and roughness of L-PBF components following LP processing. The validity of the simulation model is confirmed by comparing the calculated mean surface roughness of the polished components ($S_a)$with the experimental values. It is found that the maximum error of the simulation results for different initial surface morphologies and LP processing conditions is less than 6.8 %.