Scholarlink

<p>Cement-based composites are commonly used in construction for their cost-effectiveness and ease of use; however, they exhibit poor high-temperature resistance, often leading to cracking, spalling, and strength loss, which limits their application in fire-exposed environments. To address these challenges, this paper presents a new Logarithmic Differential Convolutional Neural Network (LDCNN)approach to enhancing the mechanical properties (MP) of multi-scale fiber-reinforced cement-driven composites by incorporating reinforced fibers, steel–polyvinyl alcohol (PVA) fibers and calcium carbonate whisker (CW) that improve performance under high-temperature conditions and structural damage.Initially, the input data is collected from a fiber-reinforced cementitious composite subjected to high temperatures.The main objective is to improve fire resistance, bending strength, and compressive strength by boosting temperature resistance and stability at high temperatures.It is employed to forecast the mechanical strength of multi scale fiber-reinforced cement-based composites (MSFRC).The simulation outcomes were compared with existing techniques, including Levenberg–Marquardt back propagation (LMBP), Artificial Neural Network (ANN), and one-dimensional conventional neural network (1DCNN).Steel-PVA fibers-CW enhances strength after heating up to 900&#xa0;°C, with optimal volumes of 3% (flexural) and 2% (compressive). MSFRC shows better higher-temperature resistance than normal concrete, Reactive Powder Concrete (RPC), and Engineered Cement-drivenComposite (ECC). Strength peaks at 200&#xa0;°C (flexural) and 400&#xa0;°C (compressive). These results highlight the potential of MSFRC for structural applications in environments exposed to fire or extreme heat.</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

Enhanced mechanical properties of multi-scale fiber-reinforced cement-based composites using reinforced fibers under high temperature for structural damage

  • 期刊论文

  • 发表日期：

    2026年4月27日

查看全文
  • K. Venkatesan,
  • S. Sivalingam,
  • R. Ashok,
  • Srinivasan Rajaram,
  • G. B. Bhaskar

摘要

Cement-based composites are commonly used in construction for their cost-effectiveness and ease of use; however, they exhibit poor high-temperature resistance, often leading to cracking, spalling, and strength loss, which limits their application in fire-exposed environments. To address these challenges, this paper presents a new Logarithmic Differential Convolutional Neural Network (LDCNN)approach to enhancing the mechanical properties (MP) of multi-scale fiber-reinforced cement-driven composites by incorporating reinforced fibers, steel–polyvinyl alcohol (PVA) fibers and calcium carbonate whisker (CW) that improve performance under high-temperature conditions and structural damage.Initially, the input data is collected from a fiber-reinforced cementitious composite subjected to high temperatures.The main objective is to improve fire resistance, bending strength, and compressive strength by boosting temperature resistance and stability at high temperatures.It is employed to forecast the mechanical strength of multi scale fiber-reinforced cement-based composites (MSFRC).The simulation outcomes were compared with existing techniques, including Levenberg–Marquardt back propagation (LMBP), Artificial Neural Network (ANN), and one-dimensional conventional neural network (1DCNN).Steel-PVA fibers-CW enhances strength after heating up to 900 °C, with optimal volumes of 3% (flexural) and 2% (compressive). MSFRC shows better higher-temperature resistance than normal concrete, Reactive Powder Concrete (RPC), and Engineered Cement-drivenComposite (ECC). Strength peaks at 200 °C (flexural) and 400 °C (compressive). These results highlight the potential of MSFRC for structural applications in environments exposed to fire or extreme heat.