A MULTI-OBJECTIVE OPTIMIZATION OF PROCESS PARAMETERS IN LASER-BASED POWDER BED FUSION FOR TI-6AL-4V: A STUDY ON DENSITY, SURFACE ROUGHNESS, AND MECHANICAL ANISOTROPY

Abstract - The process of laser-based powder bed fusion (LPBF) of titanium alloy (Ti-6Al-4V) poses challenges related to concurrently controlling porosity, poor surface finish (i.e., surface roughness) and mechanical anisotropy due to rapid melting and solidification dynamics. Whereas most research has focused on optimizing a single property, manufacturers today face the challenge of creating parts that have acceptable levels of all these properties, as they all interact with each other in such a way that there is often an inherent conflict between them. This study presents a multi-objective optimization framework to maximize the relative density, minimize the vertical surface roughness (Sa_Z) and minimize the mechanical anisotropy of as-built Ti-6Al-4V parts. A face-centered central composite design (DoE) was created and test parts created from this DoE were evaluated for density (by Archimedes principle), surface finish (by means of stylus profiling) and two sets of tensile properties in different orientations. Significant predictive models were developed using Response Surface Methodology (RSM) and ANOVA and then used in conjunction with an NSGA-II genetic algorithm to create a Pareto-optimal set of solutions. The results indicate the optimal processing conditions for creating Ti-6Al-4V parts, and they also illustrate how to effectively quantify the trade-offs associated with; processing at high densities yields higher Sa_Z values, whereas processing for isotropy yields lower relative densities. The derived Pareto frontier provides a decisive map for selecting process parameters based on application priority—enabling the choice of low-roughness parameters for biomedical implants or high-density parameters for structural components—thereby advancing the systematic, application-tailored qualification of LPBF Ti-6Al-4V.