Effect of Process Parameters on Mechanical Strength of 3D-Printed Polymer Parts Using Machine Learning-Based Prediction Models
- Version
- Download 8
- File Size 366.80 KB
- Download
Effect of Process Parameters on Mechanical Strength of 3D-Printed Polymer Parts Using Machine Learning-Based Prediction Models
Shani Singh (M.Tech)
Research Scholar, Mechanical Engineering
Shanince1@gmail.com
Abstract
This study develops machine learning models to predict the mechanical strength of fused deposition modeling (FDM) printed polymer parts using common process parameters. A dataset covering variations in layer height, infill density, wall thickness, thermal settings, deposition speed, and material type was used to model tensile and impact strength. Decision Tree, Random Forest, and XGBoost regressors were trained to evaluate how these factors influence mechanical performance. The models were assessed using R², RMSE, and MAE values. XGBoost provided the highest accuracy for both outputs, achieving an R² of 0.86 for tensile strength and 0.81 for impact strength. The results show that material type and infill density are the most influential parameters, while layer height and print speed negatively affect strength due to reduced interlayer bonding. Residual analysis confirmed the stability and generalization capability of the models. The findings demonstrate that machine learning offers a reliable method for predicting mechanical performance in FDM and can support optimization of print settings for improved part quality.
Keywords: FDM, tensile strength, impact strength, machine learning, process parameters, XGBoost