Carbon fiber (Cf) reinforced filaments are increasingly used in fused deposition modeling (FDM) to produce functional components. However, selecting an appropriate material–infill combination remains challenging because mechanical performance, surface quality, and environmental burden do not improve simultaneously. This study presents a multi-criteria performance assessment of Cf reinforced filaments by quantifying the influence of infill pattern on overall behavior. These filaments are PC/Cf, PA12/Cf, PETG/Cf, and PPS/Cf. Six infill patterns (Grid, Honeycomb, Triangular, Gyroid, Concentric, and Cubic) were investigated. Tensile strength, flexural strength, surface roughness, and carbon emissions were evaluated jointly through min–max normalization and equal-weight aggregation to rank all material–infill combinations. In parallel, an undirected principal component analysis (PCA) was conducted on standardized raw variables to explore intrinsic multivariate relationships without reversing cost criteria. The trade-off ranking identified PC/Cf–Concentric as the optimum overall configuration with an overall score of 0.775, reflecting the most balanced compromise among the four metrics. The highest mechanical performance was achieved by PPS/Cf–Concentric, reaching 76.32 MPa in tensile strength and 98.95 MPa in flexural strength, while also providing the lowest measured surface roughness (Ra = 8.25 µm). However, this configuration exhibited the highest carbon burden (197.63 g CO 2 -eq), evidencing a clear performance–sustainability conflict. The minimum carbon emission was obtained for PC/Cf–Grid (131.75 g CO 2 -eq). PCA indicated that the global variability is strongly material-driven, with PC1 and PC2 explaining 79.2% and 19.3% of the total variance, while infill architecture acts as a secondary modifier that shifts the balance among mechanical, surface, and environmental metrics. The combined ranking–PCA framework offers a transparent route for selecting Cf reinforced filament–infill configurations when both performance and sustainability are considered.
Vural et al. (2026) studied this question.