Taguchi–ANOVA optimization of LPBF processing parameters for porosity reduction in 316L stainless steel

Laser Powder Bed Fusion (LPBF) is widely used to fabricate 316L stainless steel components for aerospace, biomedical, and tooling applications; however, achieving consistently low porosity remains challenging due to coupled thermal conditions and melt-pool dynamics. This study applies a Taguchi – ANOVA-based statistical framework to a consolidated literature dataset comprising 40 experimentally reported LPBF builds of 316L stainless steel to examine the influence of key process parameters on porosity across multiple studies. Laser power and scan speed are treated as the primary controllable inputs, while volumetric energy density (VED) is employed as a derived descriptor of overall energy input. To enable balanced comparison across heterogeneous literature data, parameters are grouped into low, medium, and high levels using quantile-based discretisation. Signal-to-noise ratio analysis and ANOVA results indicate that scan speed exerts the strongest statistical influence on porosity, while variations in VED are associated with transitions between lack-of-fusion and keyhole-affected melting regimes. Within the analysed dataset, combinations involving moderate laser power and scan speed together with higher VED correspond to reduced porosity. Predicted-versus-measured comparisons demonstrate that the framework captures dominant porosity trends despite inter-study variability. Overall, the study provides a data-driven basis for interpreting and comparing reported LPBF processing windows for 316L stainless steel.