Scholarlink

<p>In this study, molecular dynamics (MD) simulations were employed to conduct tensile tests on twinned Cu-Ag alloys with varying twin boundary (TB) parameters (spacing: 1.22–6.10&#xa0;nm; orientation: 0°-360°) and grain boundary affected zone (GBAZ) segregation features. The deformation history, stress-strain response, and microstructural evolution throughout the tensile process were systematically analyzed. By integrating common neighbor analysis (CNA) and dislocation structure characterization, the regulatory effects of TB parameters and GBAZ segregation on the alloy’s mechanical behavior were elucidated. Results show that smaller TB spacing (e.g., 1.22&#xa0;nm) elevates the alloy’s stress level, while TB orientation exerts a periodic effect on the maximum stress—peaking at 0° and 90° (2.7&#xa0;MPa) and reaching minima at 45° and 135°. Subsequently, the MD simulation data were processed into a normalized structured dataset. Machine learning (ML) techniques were then applied: regression models (Random Forest, Linear Regression, Extra Trees, CatBoost) were trained on this dataset to predict the alloy’s mechanical properties, and a stacked ensemble framework was established to enhance the prediction accuracy. Among single models, Extra Trees performed best—achieving R²=0.604, RMSE = 0.104 for flow stress prediction and R²=0.662, RMSE = 0.072 for maximum stress prediction—exhibiting high accuracy especially in the elastic stage (ε &lt; 5%). This study provides a theoretical basis for the microstructural optimization and mechanical property prediction of twinned Cu-Ag alloys, and the proposed method exhibits extensibility to other nanostructured alloy 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

The effect of twin boundary spacing, orientation and separation of grain boundary affected zones on the mechanical properties of twinned Cu-Ag alloys

  • 期刊论文

  • 发表日期：

    2025年12月15日

查看全文
  • Yufan Shen,
  • Feng Zhang,
  • Mingjun Li,
  • Guo Li,
  • Dasheng Zhu,
  • Hao Su,
  • Wenchun Jiang

摘要

In this study, molecular dynamics (MD) simulations were employed to conduct tensile tests on twinned Cu-Ag alloys with varying twin boundary (TB) parameters (spacing: 1.22–6.10 nm; orientation: 0°-360°) and grain boundary affected zone (GBAZ) segregation features. The deformation history, stress-strain response, and microstructural evolution throughout the tensile process were systematically analyzed. By integrating common neighbor analysis (CNA) and dislocation structure characterization, the regulatory effects of TB parameters and GBAZ segregation on the alloy’s mechanical behavior were elucidated. Results show that smaller TB spacing (e.g., 1.22 nm) elevates the alloy’s stress level, while TB orientation exerts a periodic effect on the maximum stress—peaking at 0° and 90° (2.7 MPa) and reaching minima at 45° and 135°. Subsequently, the MD simulation data were processed into a normalized structured dataset. Machine learning (ML) techniques were then applied: regression models (Random Forest, Linear Regression, Extra Trees, CatBoost) were trained on this dataset to predict the alloy’s mechanical properties, and a stacked ensemble framework was established to enhance the prediction accuracy. Among single models, Extra Trees performed best—achieving R²=0.604, RMSE = 0.104 for flow stress prediction and R²=0.662, RMSE = 0.072 for maximum stress prediction—exhibiting high accuracy especially in the elastic stage (ε < 5%). This study provides a theoretical basis for the microstructural optimization and mechanical property prediction of twinned Cu-Ag alloys, and the proposed method exhibits extensibility to other nanostructured alloy systems.