Scholarlink

<p>The hot deformation behavior and constitutive modeling of the 4043 aluminum alloy were investigated using a Gleeble 3500 thermomechanical simulator. The tests were conducted with deformation parameters including temperatures of 300&#xa0;°C, 400&#xa0;°C, and 450&#xa0;°C; strain rates of 0.01&#xa0;s<sup>−1</sup>, 0.1&#xa0;s<sup>−1</sup>, and 1&#xa0;s<sup>−1</sup>; and an interpass holding time of 1&#xa0;s. This setup simulates the industrial “deformation—interpass holding—secondary deformation” sequence. The microstructural characterization by SEM and XRD revealed that the second-phase area fraction decreased from 8.55 to 6.92% with increasing temperature and decreasing strain rate. Based on the homogenized alloy, a back-propagation artificial neural network (BP-ANN) constitutive model and a strain-compensated Arrhenius equation constitutive model were developed. The comparative results demonstrate the better predictive capability of the BP-ANN model (R = 0.999, AARE = 0.201%) over the Arrhenius model (R = 0.996, AARE = 2.031%). Consequently, this work confirms that the BP-ANN model offers higher accuracy and more robust predictions across the parameter range for this alloy. Overall, the employed double-pass methodology overcomes the limitation of single-pass testing and provides a scientific foundation for process optimization.</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

Double Pass Hot Deformation Behavior, Microstructure Evolution and Constitutive Model Comparison of Al-Si Aluminum Alloy

  • 期刊论文

  • 发表日期：

    2026年4月1日

查看全文
  • Bojie Ding,
  • Wenhao Yang,
  • Hongfu Yang,
  • Shanju Zheng,
  • Mengnie Li

摘要

The hot deformation behavior and constitutive modeling of the 4043 aluminum alloy were investigated using a Gleeble 3500 thermomechanical simulator. The tests were conducted with deformation parameters including temperatures of 300 °C, 400 °C, and 450 °C; strain rates of 0.01 s−1, 0.1 s−1, and 1 s−1; and an interpass holding time of 1 s. This setup simulates the industrial “deformation—interpass holding—secondary deformation” sequence. The microstructural characterization by SEM and XRD revealed that the second-phase area fraction decreased from 8.55 to 6.92% with increasing temperature and decreasing strain rate. Based on the homogenized alloy, a back-propagation artificial neural network (BP-ANN) constitutive model and a strain-compensated Arrhenius equation constitutive model were developed. The comparative results demonstrate the better predictive capability of the BP-ANN model (R = 0.999, AARE = 0.201%) over the Arrhenius model (R = 0.996, AARE = 2.031%). Consequently, this work confirms that the BP-ANN model offers higher accuracy and more robust predictions across the parameter range for this alloy. Overall, the employed double-pass methodology overcomes the limitation of single-pass testing and provides a scientific foundation for process optimization.