Scholarlink

<p>AlSi12CuNiMg alloy composites were synthesized via a modified stir-casting, incorporating various weight fractions of micro/ nano B<sub>4</sub>C (3/0.5, 5/0.5 and 7/0.5&#xa0;wt%). The wear performance was assessed through a full factorial design of experiments, in view of critical wear parameters, including applied load, incorporation of B<sub>4</sub>C particulates, sliding velocity and sliding distance. To predict the wear characteristics of AlSi12CuNiMg composites, three distinct machine learning (ML) algorithms such as Random Forest (RF), Artificial Neural Network (ANN) and K-Nearest Neighbour (KNN) were trained using experimental data. Herein, the ANN model outperformed in predicting the wear rate (R<sup>2</sup>—0.9574). While the RF model excelled in prediction of the COF (R<sup>2</sup>—0.9151). Moreover, the analysis of feature importance indicated that B<sub>4</sub>C reinforcement and load are the utmost important variables effecting on the estimation of wear rate and COF. </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

Prediction of wear behavior of AlSi12CuNiMg/Micro-Nano B4C composites by machine learning algorithms

  • 期刊论文

  • 发表日期：

    2025年7月10日

查看全文
  • Rama Prasad Gajjarapu,
  • Udaya Prakash Jayavelu,
  • Kishorekumar Paleti,
  • Arun Prasad Murali,
  • Dinesh Kumar Nagarajan

摘要

AlSi12CuNiMg alloy composites were synthesized via a modified stir-casting, incorporating various weight fractions of micro/ nano B4C (3/0.5, 5/0.5 and 7/0.5 wt%). The wear performance was assessed through a full factorial design of experiments, in view of critical wear parameters, including applied load, incorporation of B4C particulates, sliding velocity and sliding distance. To predict the wear characteristics of AlSi12CuNiMg composites, three distinct machine learning (ML) algorithms such as Random Forest (RF), Artificial Neural Network (ANN) and K-Nearest Neighbour (KNN) were trained using experimental data. Herein, the ANN model outperformed in predicting the wear rate (R2—0.9574). While the RF model excelled in prediction of the COF (R2—0.9151). Moreover, the analysis of feature importance indicated that B4C reinforcement and load are the utmost important variables effecting on the estimation of wear rate and COF.