Development of ANN model for surface roughness prediction of parts produced by varying fabrication parameters

Abstract

Selective laser melting (SLM) is the most common additive manufacturing technique designed to fabricate functional parts with high accuracy. Depending on the desired properties, the process parameters for a given material need to be optimized for improving the overall reliability of the SLM devices. As all the process parameters are inter-dependent on each other, it is important to find an optimum value to suit the requirement and render the best build quality. This work primarily focuses on the effect of various process parameters such as laser power, scanning speed, and hatch spacing on the roughness of Inconel 718 parts fabricated on an EOS M290 machine. Statistical models of surface roughness are established to identify the relationship between the abovementioned process parameters. The capabilities developed in this study will permit a deep understanding of the process- property relationships in structural SLM components.