Scholarlink

<p>The prediction of materials behaviour and optimizing processes has been made easier with the use of machine learning. This paper clearly describes the fabrication of Udimet 500 composites and the evaluation of their mechanical properties. A machine learning regression model based recommender system (RS-MLR) is proposed to predict values of tensile strength and hardness. Experimental findings are used to create and test the predictive model. TiC has positive effects on mechanical properties. It has been shown that tensile strength and hardness improved by 8%. With an R² increase of 11.6%, the proposed RS-MLR model is able to predict values with better accuracy than previous methods. Values that were predicted are in close proximity to the experimental findings verifying that the approach to predict material properties is valid. From machine learning RS-MLR method, it was observed that maximum tensile strength of 822&#xa0;MPa and hardness of 82HV. R squared value (0.96) of the RS-MLR is higher than the single MLR method (0.86).</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

Hybrid RS-ML framework for forecasting mechanical responses of TiC-Reinforced Udimet 500 composites

  • 期刊论文

  • 发表日期：

    2026年4月22日

查看全文
  • G Kumaresan,
  • M.S. Srinivasa Rao,
  • S Mohan Doss,
  • Rajendra Shimpi,
  • M Vijayakumar,
  • Pothamsetty Kasi V Rao

摘要

The prediction of materials behaviour and optimizing processes has been made easier with the use of machine learning. This paper clearly describes the fabrication of Udimet 500 composites and the evaluation of their mechanical properties. A machine learning regression model based recommender system (RS-MLR) is proposed to predict values of tensile strength and hardness. Experimental findings are used to create and test the predictive model. TiC has positive effects on mechanical properties. It has been shown that tensile strength and hardness improved by 8%. With an R² increase of 11.6%, the proposed RS-MLR model is able to predict values with better accuracy than previous methods. Values that were predicted are in close proximity to the experimental findings verifying that the approach to predict material properties is valid. From machine learning RS-MLR method, it was observed that maximum tensile strength of 822 MPa and hardness of 82HV. R squared value (0.96) of the RS-MLR is higher than the single MLR method (0.86).