Scholarlink

<p>Enhancing the solidification rate of eutectic alloys presents a significant challenge for researchers. Classical theories of solidification often fall short in addressing this need, as they do not adequately account for the structural properties of the material. In this study, we investigate the solidification behavior of a spherical Cu-Al alloy by incorporating a volumetric heat sink, which effectively utilizes the entire volume of the material—rather than just its surface area—for heat absorption and dissipation. The thermal properties of the mushy region are modeled as a linear combination of the solid and liquid phases, weighted by the solid fraction, which is assumed to vary linearly with temperature. The introduction of a volumetric heat sink is found to accelerate the solidification process, with numerical simulations based on Cu-Al alloy (5% Cu) data revealing the influence of different parameters on the freezing rate. The results indicate that convection significantly enhances the solidification rate, whereas an increase in heat sink strength leads to lower temperature fields and consequently accelerates the freezing process. Furthermore, an Artificial Neural Network (ANN), trained using Bayesian regularization, was employed to predict heat flux and phase evolution. The ANN predictions closely matched analytical results, demonstrating its effectiveness as a reliable surrogate model. These findings not only advance the fundamental understanding of eutectic alloy solidification but also highlight the potential of ANN-assisted modeling for real-time thermal management and intelligent control in industrial applications involving Cu-Al alloys.</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

Artificial Neural Network Enhanced Modeling of Convection Driven Heat Transfer During Volumetric Freezing of Binary Eutectic Systems

  • 期刊论文

  • 发表日期：

    2026年4月13日

查看全文
  • Smile Bansal,
  • Sapna Sharma

摘要

Enhancing the solidification rate of eutectic alloys presents a significant challenge for researchers. Classical theories of solidification often fall short in addressing this need, as they do not adequately account for the structural properties of the material. In this study, we investigate the solidification behavior of a spherical Cu-Al alloy by incorporating a volumetric heat sink, which effectively utilizes the entire volume of the material—rather than just its surface area—for heat absorption and dissipation. The thermal properties of the mushy region are modeled as a linear combination of the solid and liquid phases, weighted by the solid fraction, which is assumed to vary linearly with temperature. The introduction of a volumetric heat sink is found to accelerate the solidification process, with numerical simulations based on Cu-Al alloy (5% Cu) data revealing the influence of different parameters on the freezing rate. The results indicate that convection significantly enhances the solidification rate, whereas an increase in heat sink strength leads to lower temperature fields and consequently accelerates the freezing process. Furthermore, an Artificial Neural Network (ANN), trained using Bayesian regularization, was employed to predict heat flux and phase evolution. The ANN predictions closely matched analytical results, demonstrating its effectiveness as a reliable surrogate model. These findings not only advance the fundamental understanding of eutectic alloy solidification but also highlight the potential of ANN-assisted modeling for real-time thermal management and intelligent control in industrial applications involving Cu-Al alloys.