Interpretable machine learning analysis of printing parameter effects on mechanical behavior of FDM-printed polylactic acid

Significant attention was given to the use of machine learning (ML) in fused deposition modeling (FDM) to optimize printing processes. In this study, an ML-based framework was developed for establishing interpretable quantitative relationships between printing parameters and the mechanical properties of polylactic acid (PLA). A polynomial ElasticNet model was employed to derive closed-form predictive equations with a strong predictive performance (R2 ≈ 0.87 for ultimate tensile strength and R2 ≈ 0.89 for elastic modulus). The layer height was identified as the dominant parameter. The proposed approach provides continuous, interpretable relationships to enable efficient process optimization in FDM.