Scholarlink

The modern industry sectors is in desperate need of prediction of the accurate mechanical properties that could help to enhance the performance of composite materials. This research is intended to know the tensile strength, flexural strength, impact strength, and hardness of Natural Fiber Composites that has been predicted using Linear Regression Algorithm. Because NFCs are known for their sustainability and lightweight nature, different fiber weight percentages have been taken viz., 0, 5, 10, and 15%. A study has been done on the tensile strength, flexural strength, impact strength, and hardness of composite materials with different compositions in the range from 0 to 16%. Measurements are done to evaluate the performance of the linear regression model with standard metrics: MAE, MSE, and the coefficient of determination, R2. Results are highly accurate for the predictions of flexural strength, MAE: 0.190, MSE: 0.054, R2: 0.9974; second was tensile strength (MAE: 1.452, MSE: 3.168, R2: 0.9673); third one was impact strength, MAE: 0.252, MSE: 0.096, R2: 0.9962; and last but not least, hardness MAE: 0.200, MSE: 0.060, R2: 0.9944. In its application, it shows an opportunity by which ML can be used to improve material properties and thus serves as a useful source of information in the use of NFCs in structural and other engineering applications.

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

Analysis of Enhanced Predictive Models on Mechanical Behavior in Natural Fiber Composites Using Linear Regression Algorithms

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • R. Alagulakshmi,
  • R. Ramalakshmi,
  • V. Arumugaprabu,
  • Geetha Palani

摘要

The modern industry sectors is in desperate need of prediction of the accurate mechanical properties that could help to enhance the performance of composite materials. This research is intended to know the tensile strength, flexural strength, impact strength, and hardness of Natural Fiber Composites that has been predicted using Linear Regression Algorithm. Because NFCs are known for their sustainability and lightweight nature, different fiber weight percentages have been taken viz., 0, 5, 10, and 15%. A study has been done on the tensile strength, flexural strength, impact strength, and hardness of composite materials with different compositions in the range from 0 to 16%. Measurements are done to evaluate the performance of the linear regression model with standard metrics: MAE, MSE, and the coefficient of determination, R2. Results are highly accurate for the predictions of flexural strength, MAE: 0.190, MSE: 0.054, R2: 0.9974; second was tensile strength (MAE: 1.452, MSE: 3.168, R2: 0.9673); third one was impact strength, MAE: 0.252, MSE: 0.096, R2: 0.9962; and last but not least, hardness MAE: 0.200, MSE: 0.060, R2: 0.9944. In its application, it shows an opportunity by which ML can be used to improve material properties and thus serves as a useful source of information in the use of NFCs in structural and other engineering applications.