Scholarlink

<p>The growing demand for sustainable and chemically durable construction materials has increased interest in geopolymer concrete as an alternative to conventional cement-based systems. However, few studies have integrated durability assessment in aggressive environments with interpretable machine-learning prediction approaches. Therefore, this study investigates the mechanical performance and durability behaviour of geopolymer concrete subjected to acid and sulphate exposure under different alkaline molarities (12&#xa0;M and 14&#xa0;M) and curing temperatures (40&#xa0;°C, 60&#xa0;°C, and 80&#xa0;°C). Durability performance was evaluated using compressive strength, weight loss, expansion behaviour, and residual strength analysis. In addition, machine learning models, including Artificial Neural Networks (ANNs), Gaussian Process Regression (GPR), Support Vector Machines (SVMs), and Random Forests (RFs), were developed to predict durability behaviour. The experimental results showed that higher molarity and curing temperature significantly improved strength retention and reduced degradation under aggressive exposure conditions, with the 14&#xa0;M and 80&#xa0;°C mix exhibiting the best overall performance. Among the developed models, the GPR model demonstrated the highest predictive capability with strong agreement between experimental and predicted results. Microstructural investigations using SEM, XRD, XRF, and TGA confirmed the formation of dense N-A-S–H and C-A-S–H gel phases, reduced pore connectivity, improved thermal stability, and enhanced aluminosilicate reaction products at elevated curing conditions. The findings highlight the potential of combining geopolymer technology with machine learning-based predictive modelling to support the development of durable and sustainable construction materials for aggressive service environments. The findings confirm that elevated curing temperature significantly improves the geopolymerization efficiency, durability, and long-term stability of geopolymer concrete, thereby supporting its potential application as a sustainable construction material in aggressive environmental conditions.</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

Experimental and machine learning-based evaluation of durability and mechanical performance of geopolymer concrete under acid and sulphate exposure

  • 期刊论文

  • 发表日期：

    2026年6月26日

查看全文
  • Dipak Patare,
  • Pravin Nemade

摘要

The growing demand for sustainable and chemically durable construction materials has increased interest in geopolymer concrete as an alternative to conventional cement-based systems. However, few studies have integrated durability assessment in aggressive environments with interpretable machine-learning prediction approaches. Therefore, this study investigates the mechanical performance and durability behaviour of geopolymer concrete subjected to acid and sulphate exposure under different alkaline molarities (12 M and 14 M) and curing temperatures (40 °C, 60 °C, and 80 °C). Durability performance was evaluated using compressive strength, weight loss, expansion behaviour, and residual strength analysis. In addition, machine learning models, including Artificial Neural Networks (ANNs), Gaussian Process Regression (GPR), Support Vector Machines (SVMs), and Random Forests (RFs), were developed to predict durability behaviour. The experimental results showed that higher molarity and curing temperature significantly improved strength retention and reduced degradation under aggressive exposure conditions, with the 14 M and 80 °C mix exhibiting the best overall performance. Among the developed models, the GPR model demonstrated the highest predictive capability with strong agreement between experimental and predicted results. Microstructural investigations using SEM, XRD, XRF, and TGA confirmed the formation of dense N-A-S–H and C-A-S–H gel phases, reduced pore connectivity, improved thermal stability, and enhanced aluminosilicate reaction products at elevated curing conditions. The findings highlight the potential of combining geopolymer technology with machine learning-based predictive modelling to support the development of durable and sustainable construction materials for aggressive service environments. The findings confirm that elevated curing temperature significantly improves the geopolymerization efficiency, durability, and long-term stability of geopolymer concrete, thereby supporting its potential application as a sustainable construction material in aggressive environmental conditions.