Simultaneous consideration of relative density, energy consumption, and build time for selective laser melting of Inconel 718: A multi-objective optimization study on process parameter selection

Abstract

Additive manufacturing (AM) is regarded as sustainability friendly thanks to its significantly reduced production steps and simplified supply chain at the system level, but it is often not energy efficient at the process level. In this study, we propose that the optimization of AM process parameters should not only maximize build quality, but also minimize environmental impact and build time. A multi-objective optimization problem of maximizing relative density (RD), minimizing energy consumption and build time per unit volume (ev and tv respectively) is formulated and solved for selective laser melting (SLM) of Inconel 718 alloy. A high accuracy RD prediction model with R2 of 93.23% is obtained by fitting SLM experiment data from wide ranges of laser power (P), scanning speed (SS), hatch spacing (HS), and layer thickness (LT). The ev and tv models are created considering printing area utilization rate (U) based on laser scanning and powder coating process. The two approaches of Pareto front and global criterion method are adopted for solving the multi-objective optimization problem. It is found that the highest RD of close to 100% can be obtained at the expense of high ev and tv of close to 500 J/mm3 and 0.3 s/mm3, respectively, and about three times ev and tv are needed to increase RD from about 95% to 100% than from 85% to 95%. At the same U level, the solutions obtained from prioritizing ev and tv are consistent, while those from prioritizing RD show smaller process parameter values. The multi-objective optimization formulation provides an enabling tool for industry to obtain high quality SLM-ed parts with improved sustainability and efficiency at the process level. The methodology developed is applicable to selective laser melting of other materials, and such consideration can be readily expanded to other AM processes as well.