Scholarlink

<p>The present study investigates the tribological behaviour of polylactic acid (PLA) composites reinforced with rice husk biochar (RHBC) through an integrated approach combining experimentation, statistical optimization, and machine learning. PLA/RHBC composites with filler contents ranging from 10 to 20 wt% were fabricated using fused deposition modeling (FDM). A Box–Behnken design (BBD) within the framework of Response Surface Methodology (RSM) was employed to systematically examine the effects of key process parameters—printing angle, infill type, nozzle temperature, and filler content at three levels and across 24 experimental runs. Tribological performance was evaluated in terms of wear rate and coefficient of friction (COF) using a pin-on-disc setup under applied loads of 10, 20, and 30&#xa0;N. Analysis of variance (ANOVA) revealed that the applied load is the most significant factor influencing wear rate, whereas COF is predominantly governed by the combined effects of the printing parameters. Optical microscopy further indicated that incorporating RHBC reduces ploughing and improves interfacial stability. To enhance predictive capability, a machine learning framework incorporating Multiple Linear Regression (MLR), Extreme Gradient Boosting (XGBoost), and Artificial Neural Networks (ANN) was implemented. Among these, the ANN model demonstrated superior performance, achieving the highest prediction accuracy for wear rate with R² values of 0.9852, 0.9845, and 0.9891 at 10, 20, and 30&#xa0;N, respectively. A similar trend was observed for COF, where ANN again outperformed the other models, yielding R² values of 0.9892, 0.9880, and 0.9910 at the corresponding loads. The integration of RSM with machine learning enables efficient optimization of FDM parameters and accurate prediction of tribological performance. This highlights the suitability of RHBC/PLA composites for lightweight, wear-resistant applications such as bearings, bushings, and sliding components.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine learning and response surface methodology for optimization and prediction of tribological performance of PLA/rice husk biochar composites

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月6日

查看全文
  • Sundarasetty Harishbabu,
  • Amina Salhi,
  • Leila Jamel,
  • Borhen Louhichi,
  • It Ee Lee,
  • Santosh Kumar Sahu,
  • Qamar Wali

摘要

The present study investigates the tribological behaviour of polylactic acid (PLA) composites reinforced with rice husk biochar (RHBC) through an integrated approach combining experimentation, statistical optimization, and machine learning. PLA/RHBC composites with filler contents ranging from 10 to 20 wt% were fabricated using fused deposition modeling (FDM). A Box–Behnken design (BBD) within the framework of Response Surface Methodology (RSM) was employed to systematically examine the effects of key process parameters—printing angle, infill type, nozzle temperature, and filler content at three levels and across 24 experimental runs. Tribological performance was evaluated in terms of wear rate and coefficient of friction (COF) using a pin-on-disc setup under applied loads of 10, 20, and 30 N. Analysis of variance (ANOVA) revealed that the applied load is the most significant factor influencing wear rate, whereas COF is predominantly governed by the combined effects of the printing parameters. Optical microscopy further indicated that incorporating RHBC reduces ploughing and improves interfacial stability. To enhance predictive capability, a machine learning framework incorporating Multiple Linear Regression (MLR), Extreme Gradient Boosting (XGBoost), and Artificial Neural Networks (ANN) was implemented. Among these, the ANN model demonstrated superior performance, achieving the highest prediction accuracy for wear rate with R² values of 0.9852, 0.9845, and 0.9891 at 10, 20, and 30 N, respectively. A similar trend was observed for COF, where ANN again outperformed the other models, yielding R² values of 0.9892, 0.9880, and 0.9910 at the corresponding loads. The integration of RSM with machine learning enables efficient optimization of FDM parameters and accurate prediction of tribological performance. This highlights the suitability of RHBC/PLA composites for lightweight, wear-resistant applications such as bearings, bushings, and sliding components.