Scholarlink

<p>This study aims to optimize the mechanical properties of graphene-enhanced jute fiber epoxy composites by simultaneously improving open-hole tensile strength (OHTS), elasticity modulus (EM), and failure strain (FS) using a hybrid machine learning approach. A central composite design (CCD) was employed to explore the effects of molding pressure (1–6&#xa0;MPa), molding temperature (140–190&#xa0;°C), and graphene content (0.1–2.0 wt.%) on the composite performance. An Artificial Neural Network (ANN) model trained on experimental data achieved high predictive accuracy (<i>R</i><sup>2</sup> = 0.92–0.94), capturing non-linear dependencies among processing parameters and mechanical responses. Multi-objective optimization using the Non-dominated Sorting Genetic Algorithm II (NSGA-II) integrated with ANN-predicted optimal conditions of 5.967&#xa0;MPa pressure, 140.176&#xa0;°C temperature, and 1.980 wt.% graphene. The predicted tensile strength (339.05&#xa0;MPa), modulus (20.19 GPa), and failure strain (FS) (0.0244&#xa0;mm/mm) closely matched experimental values with &lt; 9% error. SEM analysis revealed fracture mechanisms including fiber pull-out, delamination, and void formation, emphasizing the need for improved interfacial adhesion and graphene dispersion. This study establishes a robust ANN-NSGA-II framework for accelerating process optimization in bio-based nanocomposites, with potential scalability to diverse sustainable material systems.</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

Machine Learning-Guided Evolutionary Optimization of Compression Molding Parameters for Graphene-Enhanced Jute Fiber Composites

  • 期刊论文

  • 发表日期：

    2025年5月22日

查看全文
  • P. Hariharasakthisudhan,
  • M. Kandasamy,
  • K. Logesh,
  • K. Sathickbasha

摘要

This study aims to optimize the mechanical properties of graphene-enhanced jute fiber epoxy composites by simultaneously improving open-hole tensile strength (OHTS), elasticity modulus (EM), and failure strain (FS) using a hybrid machine learning approach. A central composite design (CCD) was employed to explore the effects of molding pressure (1–6 MPa), molding temperature (140–190 °C), and graphene content (0.1–2.0 wt.%) on the composite performance. An Artificial Neural Network (ANN) model trained on experimental data achieved high predictive accuracy (R2 = 0.92–0.94), capturing non-linear dependencies among processing parameters and mechanical responses. Multi-objective optimization using the Non-dominated Sorting Genetic Algorithm II (NSGA-II) integrated with ANN-predicted optimal conditions of 5.967 MPa pressure, 140.176 °C temperature, and 1.980 wt.% graphene. The predicted tensile strength (339.05 MPa), modulus (20.19 GPa), and failure strain (FS) (0.0244 mm/mm) closely matched experimental values with < 9% error. SEM analysis revealed fracture mechanisms including fiber pull-out, delamination, and void formation, emphasizing the need for improved interfacial adhesion and graphene dispersion. This study establishes a robust ANN-NSGA-II framework for accelerating process optimization in bio-based nanocomposites, with potential scalability to diverse sustainable material systems.