A multi-response optimization framework for fused deposition modeling of load-bearing components: Application to transtibial prosthetic sockets

Fused deposition modeling (FDM) of load-bearing polymer components involves a persistent trade-off between manufacturability and structural performance because process settings that improve production efficiency can also alter the effective structural section and the resulting stress field. This study develops a multi-response optimization framework for FDM-based load-bearing components, demonstrated through a transtibial prosthetic socket case. A Taguchi L9 design was implemented using three controllable factors at three levels, namely wall thickness (1–3 mm), filament material (PLA, ABS, and PP), and nozzle diameter (0.4–0.8 mm), applied to a constant socket geometry under a fixed slicing profile. Build time and mass were extracted from FlashPrint slicing, while screening-level mechanical performance was evaluated using specific strength and finite element outputs of von Mises stress and safety factor under standardized loading. Signal-to-noise analysis and one-way ANOVA were used to quantify factor effects, and desirability aggregation was used to identify the best multi-response compromise. The results reveal a clear factor hierarchy across responses. Nozzle diameter dominates build time, accounting for 55.40% of the variation, whereas wall thickness governs the structural screening metrics, explaining 99.98% of the variance in both von Mises stress and safety factor. Mass is controlled by a more balanced interaction among nozzle diameter, wall thickness, and material, while specific strength is governed primarily by wall thickness with meaningful secondary influence from nozzle diameter. The optimal compromise was obtained at 2 mm wall thickness, ABS, and a 0.8 mm nozzle, yielding the highest overall desirability (D = 0.753). The study establishes a practical screening-oriented workflow for local FDM socket fabrication and provides a transferable basis for multi-response optimization of other FDM-based load-bearing components.