Machine Learning-Enabled Optimization and Prediction of Mechanical Properties of 3D-Printed PLA Composites Filled with Rice Husk Biochar

This investigation focuses on rice husk biochar (RHBC) as a sustainable filler in a polylactic acid (PLA) matrix. This study employs optimization techniques, including central composite design (CCD) and analysis of variance (ANOVA), to systematically evaluate the effects of key 3D printing parameters such as filler content (0 wt.%, 10 wt.%, 20 wt.%), nozzle temperature (190 °C, 200 °C, 210 °C), orientation angle (0°, 60°, 120°), and fill pattern (hexagon, triangle, and 3D infill). Furthermore, machine learning models are used to predict the mechanical properties of PLA/RHBC composites from experimental data. The effects of these parameters on tensile strength, Young's modulus, and hardness were analyzed. The ANOVA results showed that filler content was the most influential factor for tensile strength and Young's modulus, contributing 36.47% and 73.25%, respectively, compared to pure PLA. For hardness, both filler content and nozzle temperature were key contributors, with a 44.08% improvement over pure PLA. Machine learning models, including multiple linear regression (MLR), K-Nearest Neighbors (KNN), Support Vector Machine (SVM), and Gradient Boosting, were used to predict the mechanical properties. Among these, Gradient Boosting achieved the best performance, with R² values of 97.79% for tensile strength, 98.79% for Young's modulus, and 96.8% for hardness. This study provides a robust framework that combines experimental analysis, statistical design, and machine learning to optimize RHBC as an eco-friendly filler for the development of PLA composites for adoption in the automotive, sports and aerospace industries.