Sustainable Materials & Feedstocks

Machine learning models are revolutionizing material selection by ranking eco-friendly polymers and composites based on performance metrics, enabling a balance between environmental sustainability and mechanical requirements.

Hassan et al., Composites Part C (2024) - Section 6: Material Selection & Formulation

Key Impact Metrics

Lowest RMSE
XGBoost outperformed 6 regressors for tensile strength prediction
0.28mm
Optimal layer thickness identified by ANN-GA hybrid
34%
Optimal infill density for ductility improvement

Highlighted Studies

Study Highlights
Pearson Heatmap Analysis
Factorial Experiments
Exposed how shell thickness, nozzle temperature, and layer height drive tensile strength for PLA/ABS blends.
XGBoost Regression
Comparative ML Study
Delivered the lowest RMSE and MAE among six regressors (linear, RF, AdaBoost, etc.) when predicting tensile strength.
ANN-GA Hybrid Optimization
Neural Network + Genetic Algorithm
Pinpointed 0.28mm layer thickness, 34% infill, and 222°C nozzle settings that improved ductility while minimizing build time and cost.

See also