Dual-Method Design Exploration for Parametric Sensitivity and Performance Optimization in CF-PA6 Additive Manufacturing

Material characteristics play a significant role in the engineering design process. Enhancing these properties through improved processing techniques and tighter process control is essential for achieving optimal outcomes. Fiber-reinforced additive manufacturing (FRAM) enables the production of composite parts with high mechanical performance, with Fused Filament Fabrication (FFF) being the most common AM method using fibre-reinforced polymer filaments. Among these, carbon fibre reinforced polyamide (CF-PA) is of particular interest due to its favourable strength-to-weight ratio and thermal stability.

The properties of short carbon fibre reinforced polyamide composites are highly sensitive to variations in process parameters. In this study, a dual-phase experimental strategy was employed to investigate the influence of key FFF process parameters on the performance of carbon fibre reinforced polyamide (CF-PA6) parts. Five control factors, print temperature (260, 265, 270, 275 °C), print speed (10, 20, 30, 40 mm/s), layer height (0.1, 0.18, 0.26, 0.35 mm), raster angle (0°, 30°, 60°, 90°), and infill percentage (70%, 80%, 90%, 100%) were examined using a structured Taguchi L16 orthogonal array to assess the main effects of each parameter. To complement this design and explore the intermediate design space between the individual factor levels more thoroughly, Latin Hypercube Sampling (LHS) was employed with additional 18 runs.

The comparative evaluation of Taguchi L16 and LHS results in this research enabled the extraction of both general trends and localized sensitivities across the focused parameter ranges. While the Taguchi method provides a highly efficient framework for identifying influential parameters and optimizing within defined levels, the LHS design offers finer resolution and captures variations within the gaps left by orthogonal designs. By analyzing both datasets for porosity and print time, this approach allows for a better understanding of parameter scalability, helps identify stable processing windows, and supports more informed decision-making in fiber-reinforced additive manufacturing. This integrated methodology thus strengthens process insight by balancing statistical efficiency with design space exploration.