Fuzzy process optimization of laser powder bed fusion of 316L stainless steel

This study deals with the fuzzy-based process optimization of 316L stainless steel components manufactured by Laser Powder Bed Fusion for high-performance applications. First, a systematic experimental plan was aimed at determining how the process input parameters, i.e., volumetric energy density and building orientation, affect density, ultimate tensile strength, hardness and roughness. Then, a fuzzy-based model, optimized through genetic algorithms, was developed and tested to find the best process window allowing the obtainment of the most performing mechanical properties as output. The use of the genetic algorithms concerned the identification of the optimal support of the fuzzy numbers at each membership level. The experimental results, when compared with a traditional annealed 316L stainless steel alloy, show an improvement of the mechanical properties, except for the roughness. The proposed fuzzy model shows the ability to replicate the experimental data with an increasing precision for increasing membership level, representing a new tool for understanding how much a modification at the input level can affect both the model precision and the process variability.