Scholarlink

<p>Foam glass ceramics were prepared using high-titanium blast furnace slag (HTBFS) and waste glass powder (WG) via a one-step sintering process that combined foaming and crystallization, with SiC used as the foaming agent. Experiments including x-ray diffraction (XRD), scanning electron microscopy (SEM), porosity measurement using an automatic true density analyzer, compressive strength testing, and thermal conductivity testing were conducted to evaluate the material’s structure and thermal properties. An ANSYS finite element model was developed to investigate the effects of porosity, average pore size, and crystallinity on thermal conductivity, with porosity identified as the dominant factor. At lower porosity, average pore size and crystallinity had a stronger impact on thermal conductivity, while their effects diminished as porosity increased. To reduce the computational cost of repeated simulations, a backpropagation (BP) neural network was developed to predict the thermal conductivity of foam glass ceramics. The model was trained using the ANSYS simulation data and validated using experimental results. With 9 hidden neurons, the BP network achieved a maximum relative error of 7%, demonstrating its potential as an efficient and accurate tool for predicting the thermal performance of foam glass ceramics.</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

Prediction Modeling for Thermal Conductivity of Foam Glass Ceramics Prepared by High-Titanium Blast Furnace Slag

  • 期刊论文

  • 发表日期：

    2026年2月25日

查看全文
  • Qiaoling Jiang,
  • Liangping Cai,
  • Sixuan Ke,
  • Keqin Feng

摘要

Foam glass ceramics were prepared using high-titanium blast furnace slag (HTBFS) and waste glass powder (WG) via a one-step sintering process that combined foaming and crystallization, with SiC used as the foaming agent. Experiments including x-ray diffraction (XRD), scanning electron microscopy (SEM), porosity measurement using an automatic true density analyzer, compressive strength testing, and thermal conductivity testing were conducted to evaluate the material’s structure and thermal properties. An ANSYS finite element model was developed to investigate the effects of porosity, average pore size, and crystallinity on thermal conductivity, with porosity identified as the dominant factor. At lower porosity, average pore size and crystallinity had a stronger impact on thermal conductivity, while their effects diminished as porosity increased. To reduce the computational cost of repeated simulations, a backpropagation (BP) neural network was developed to predict the thermal conductivity of foam glass ceramics. The model was trained using the ANSYS simulation data and validated using experimental results. With 9 hidden neurons, the BP network achieved a maximum relative error of 7%, demonstrating its potential as an efficient and accurate tool for predicting the thermal performance of foam glass ceramics.