Scholarlink

<p>This study presents an integrated experimental and machine–learning–based investigation into the mechanical performance and carbonation behavior of M40-grade concrete under varying curing ages, temperatures, and relative humidity. A total of 48 concrete cube specimens were prepared and water-cured for 7 and 28 days, then subjected to controlled environmental exposure at 27 and 50&#xa0;°C with relative humidity levels of 40, 60, 80, and 100%. Non-destructive testing techniques, including Ultrasonic Pulse Velocity (UPV) and Rebound Hammer tests, were employed to assess concrete quality and estimate compressive strength, while carbonation depth was determined using the phenolphthalein indicator method. Experimental results indicate that higher relative humidity and extended curing significantly enhance UPV by improving hydration and microstructural densification. In contrast, elevated temperatures accelerate strength development but increase susceptibility to carbonation under low-humidity conditions. Carbonation depth consistently decreases with increasing relative humidity, while higher temperatures promote deeper carbonation penetration. To efficiently capture these complex nonlinear relationships, Artificial Neural Network (ANN) models were developed using temperature, curing age, and relative humidity as input parameters. The ANN models demonstrated excellent predictive performance, with high correlation coefficients (<i>R</i> &gt; 0.95) and low error metrics across training, testing, and overall datasets. Residual analysis further confirmed the unbiased and robust nature of the model predictions. The findings highlight the effectiveness of combining non-destructive testing with machine learning to accurately predict performance and assess the durability of concrete under realistic environmental conditions, offering a valuable tool for service-life evaluation and sustainable infrastructure design.</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

Analysis of NDT and carbonation depth in M40 concrete using experimental methods and artificial neural networks

  • 期刊论文

  • 发表日期：

    2026年7月14日

查看全文
  • Prashant Chavan,
  • Chaitanya Mishra,
  • M. R. Bendre

摘要

This study presents an integrated experimental and machine–learning–based investigation into the mechanical performance and carbonation behavior of M40-grade concrete under varying curing ages, temperatures, and relative humidity. A total of 48 concrete cube specimens were prepared and water-cured for 7 and 28 days, then subjected to controlled environmental exposure at 27 and 50 °C with relative humidity levels of 40, 60, 80, and 100%. Non-destructive testing techniques, including Ultrasonic Pulse Velocity (UPV) and Rebound Hammer tests, were employed to assess concrete quality and estimate compressive strength, while carbonation depth was determined using the phenolphthalein indicator method. Experimental results indicate that higher relative humidity and extended curing significantly enhance UPV by improving hydration and microstructural densification. In contrast, elevated temperatures accelerate strength development but increase susceptibility to carbonation under low-humidity conditions. Carbonation depth consistently decreases with increasing relative humidity, while higher temperatures promote deeper carbonation penetration. To efficiently capture these complex nonlinear relationships, Artificial Neural Network (ANN) models were developed using temperature, curing age, and relative humidity as input parameters. The ANN models demonstrated excellent predictive performance, with high correlation coefficients (R > 0.95) and low error metrics across training, testing, and overall datasets. Residual analysis further confirmed the unbiased and robust nature of the model predictions. The findings highlight the effectiveness of combining non-destructive testing with machine learning to accurately predict performance and assess the durability of concrete under realistic environmental conditions, offering a valuable tool for service-life evaluation and sustainable infrastructure design.