Improving Surface Roughness of Additively Manufactured Parts Using a Photopolymerization Model and Multi-Objective Particle Swarm Optimization

Although additive manufacturing (AM) offers great potential to revolutionize modern manufacturing, its layer-by-layer process results in a staircase-like rough surface profile of the printed part, which degrades dimensional accuracy and often leads to a significant reduction in mechanical performance. In this paper, we present a systematic approach to improve the surface profile of AM parts using a computational model and a multi-objective optimization technique. A photopolymerization model for a micro 3D printing process, projection micro-stereolithography (P mu SL), is implemented by using a commercial finite element solver (COMSOL Multiphysics software). First, the effect of various process parameters on the surface roughness of the printed part is analyzed using Taguchi's method. Second, a metaheuristic optimization algorithm, called multi-objective particle swarm optimization, is employed to suggest the optimal P mu SL process parameters (photo-initiator and photo-absorber concentrations, layer thickness, and curing time) that minimize two objectives; printing time and surface roughness. The result shows that the proposed optimization framework increases 18% of surface quality of the angled strut even at the fastest printing speed, and also reduces 50% of printing time while keeping the surface quality equal for the vertical strut, compared to the samples produced with non-optimized parameters. The systematic approach developed in this study significantly increase the efficiency of optimizing the printing parameters compared to the heuristic approach. It also helps to achieve 3D printed parts with high surface quality in various printing angles while minimizing printing time.