Scholarlink

<p>This study examines the influence of aluminum oxide (Al<sub>2</sub>O<sub>3</sub>) nanofluid concentration under minimum quantity lubrication (MQL) on the machinability of AISI 304 stainless steel. Cutting force and surface roughness were the primary response parameters, evaluated using a structured L27 orthogonal experimental design. Experiments tested three levels of cutting speed, feed rate, and nanoparticle concentration. Analysis of variance (ANOVA) showed that nanoparticle concentration and cutting speed were the most significant factors, markedly reducing both cutting force and surface roughness, while interaction effects were statistically insignificant. Complementing the experimental analysis, an artificial neural network (ANN) model, employing the Levenberg-Marquardt (LM) algorithm, was developed to predict responses based on process parameters. The ANN achieved excellent prediction accuracy, with R<sup>2</sup> values &gt; 0.998 and minimal mean squared error, demonstrating its ability to capture complex nonlinear relationships. The model generalized effectively across training, validation, and testing datasets, with residual and performance plots confirming convergence and reliability. The combined application of ANOVA and ANN provides a robust process analysis and optimization methodology. This integrated approach identifies key influencing factors and accurately predicts machining outcomes, thereby contributing to developing intelligent and sustainable strategies for machining difficult-to-machine materials.</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

Experimental and Predictive Analysis of Cutting Force and Surface Roughness in MQL-Assisted Turning of AISI 304 Using Varying Concentrations of Al2O3 Nanofluid

  • 期刊论文

  • 发表日期：

    2025年10月20日

查看全文
  • Mahip Singh,
  • Amit Rai Dixit,
  • Anuj Kumar Sharma

摘要

This study examines the influence of aluminum oxide (Al2O3) nanofluid concentration under minimum quantity lubrication (MQL) on the machinability of AISI 304 stainless steel. Cutting force and surface roughness were the primary response parameters, evaluated using a structured L27 orthogonal experimental design. Experiments tested three levels of cutting speed, feed rate, and nanoparticle concentration. Analysis of variance (ANOVA) showed that nanoparticle concentration and cutting speed were the most significant factors, markedly reducing both cutting force and surface roughness, while interaction effects were statistically insignificant. Complementing the experimental analysis, an artificial neural network (ANN) model, employing the Levenberg-Marquardt (LM) algorithm, was developed to predict responses based on process parameters. The ANN achieved excellent prediction accuracy, with R2 values > 0.998 and minimal mean squared error, demonstrating its ability to capture complex nonlinear relationships. The model generalized effectively across training, validation, and testing datasets, with residual and performance plots confirming convergence and reliability. The combined application of ANOVA and ANN provides a robust process analysis and optimization methodology. This integrated approach identifies key influencing factors and accurately predicts machining outcomes, thereby contributing to developing intelligent and sustainable strategies for machining difficult-to-machine materials.