Process Parameter Optimization for Selective Laser Melting of 316L Stainless Steel Considering Multi Performance Metrics

Abstract

In the last decade, additive manufacturing has seen tremendous growth in various industrial applications. It showed better flexibility than conventional techniques for the manufacturing of small volume, complicated, and customized components. Selective laser melting, one of the most promising additive manufacturing technologies, can be utilized to build metal parts layer by layer within a powder bed system. However, as a new technique, it is still in its infancy, as the melted metallic parts' characteristics are not yet strong enough for many industrial uses. This research aims to improve multiple performance metrics of the selective laser melting process by optimizing the process parameters that have a very strong influence on the characteristics of the built parts. Process parameters including laser power, scan speed, hatch distance, and layer thickness are used to predict the relative density of 316L stainless steel specimens fabricated by selective laser melting. An extensive dataset is created by systematically combining experimental results from prior studies with experimental results from the current work. By utilizing promising machine learning techniques; Ridge Regression, Support Vector Regression, Kernel Ridge Regression, Polynomial Regression, predictive models have been developed and validated by 10-fold cross-validation to understand how process parameters affect the relative density. Several multi-performance metrics optimization models are developed to efficiently optimize process parameters with respect to some targeted properties of the final parts. The machine operating cost, a major portion of the total production cost of a part fabricated by selective laser melting, is quantified and modeled using different cost elements that are affected by the process parameters. The results from the optimization models provided a set of process parameters that result in a satisfactory level of the relative density and operating cost of a part. Further analysis on the optimization results included utilizing a robust optimization approach to enhance the stability of the system and performing a sensitivity analysis to understand the effect of varying some parameters on the relative density of the parts. The overall results obtained from this research can assist in achieving the best trade-off among different characteristics of built parts while substantially lowering the number of costly experimental trials.

Date of Award	Nov 2021
Original language	American English