Scholarlink

<p>The objective of this study is to investigate the impact of cutting parameters (depth of cut, feed rate, cutting speed, and cutting tool material) on the resultant cutting force (<i>F</i><sub>r</sub>), surface roughness (<i>R</i><sub>z</sub>), and cutting pressure (<i>K</i><sub>c</sub>) during the turning of EN-GJL-250 cast iron. An L54 experimental design was adopted with the following input factors and levels: type of tools (coated Si<sub>3</sub>N<sub>4</sub>, uncoated Si<sub>3</sub>N<sub>4</sub>), depth of cut (0.25, 0.5, 0.75 mm), feed rate (0.08, 0.14, 0.2 mm/rev), and cutting speed (260, 370, 530 m/min). To determine the contribution of each cutting parameter to the studied factors, an ANOVA analysis was conducted. Machine learning algorithms employed include the Levenberg–Marquardt backpropagation algorithm (LM), decision tree algorithm (DT), support vector machines algorithm (SVM), and Dragonfly Algorithm-optimized deep neural network (Da-DNN). These algorithms generated predictive models of the technological parameters, and their performance was compared and discussed. To optimize the cutting parameters, the desirability function method (DF) and the Multi-Objective Ant Lion Optimizer (MOALO) algorithm were used. Statistical analysis demonstrated the significant role of insert coating in improving surface roughness and reducing cutting forces and pressures. The findings from MOALO are promising for predicting and optimizing the turning process.</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

Statistical analysis and predictive modeling of cutting parameters in EN-GJL-250 cast iron turning: application of machine learning and MOALO optimization

  • 期刊论文

  • 发表日期：

    2025年2月25日

查看全文
  • Omar Reffas,
  • Haithem Boumediri,
  • Yacine Karmi,
  • Mohamed Said Kahaleras,
  • Issam Bousba,
  • Laouissi Aissa

摘要

The objective of this study is to investigate the impact of cutting parameters (depth of cut, feed rate, cutting speed, and cutting tool material) on the resultant cutting force (Fr), surface roughness (Rz), and cutting pressure (Kc) during the turning of EN-GJL-250 cast iron. An L54 experimental design was adopted with the following input factors and levels: type of tools (coated Si3N4, uncoated Si3N4), depth of cut (0.25, 0.5, 0.75 mm), feed rate (0.08, 0.14, 0.2 mm/rev), and cutting speed (260, 370, 530 m/min). To determine the contribution of each cutting parameter to the studied factors, an ANOVA analysis was conducted. Machine learning algorithms employed include the Levenberg–Marquardt backpropagation algorithm (LM), decision tree algorithm (DT), support vector machines algorithm (SVM), and Dragonfly Algorithm-optimized deep neural network (Da-DNN). These algorithms generated predictive models of the technological parameters, and their performance was compared and discussed. To optimize the cutting parameters, the desirability function method (DF) and the Multi-Objective Ant Lion Optimizer (MOALO) algorithm were used. Statistical analysis demonstrated the significant role of insert coating in improving surface roughness and reducing cutting forces and pressures. The findings from MOALO are promising for predicting and optimizing the turning process.