Scholarlink

<p>This study investigates the durability and performance of cement mortar incorporating dimensional limestone crushed sand (DLCS) and dimensional limestone slurry (DLS) waste as a partial replacement for river sand. Fourteen mortar mixes with DLS and DLCS replacements of 20%, 30%, and 40% were evaluated for resistance to environmental conditions, including acid attack, sulfate attack, chloride ion penetration, and carbonation. Multiple Linear Regression (MLR), Random Forest Regression (RF), Support Vector Regression (SVR), and Artificial Neural Networks (ANNs) were used to predict the long-term durability of the mortar mixes. ANNs, a machine learning approach that mimics neural structures to model complex relationships in data, achieved the highest prediction accuracy, with R<sup>2</sup> values exceeding 0.90 across all degradation mechanisms. Feature importance analysis indicated that DLS/DLCS replacement percentage and exposure duration were the most influential factors affecting durability. Results revealed that mixes with 20–30% replacement exhibited superior performance in reducing strength loss and permeability while enhancing microstructural integrity. This study provides critical insights into optimizing cementitious materials for enhanced durability in environmentally challenging 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

Performance evaluation predictive modeling of cement mortar durability with dimensional limestone waste

  • 期刊论文

  • 发表日期：

    2025年6月22日

查看全文
  • Harshwardhan Singh Chauhan,
  • Kul Vaibhav Sharma,
  • Pradeep Kumar Gautam

摘要

This study investigates the durability and performance of cement mortar incorporating dimensional limestone crushed sand (DLCS) and dimensional limestone slurry (DLS) waste as a partial replacement for river sand. Fourteen mortar mixes with DLS and DLCS replacements of 20%, 30%, and 40% were evaluated for resistance to environmental conditions, including acid attack, sulfate attack, chloride ion penetration, and carbonation. Multiple Linear Regression (MLR), Random Forest Regression (RF), Support Vector Regression (SVR), and Artificial Neural Networks (ANNs) were used to predict the long-term durability of the mortar mixes. ANNs, a machine learning approach that mimics neural structures to model complex relationships in data, achieved the highest prediction accuracy, with R2 values exceeding 0.90 across all degradation mechanisms. Feature importance analysis indicated that DLS/DLCS replacement percentage and exposure duration were the most influential factors affecting durability. Results revealed that mixes with 20–30% replacement exhibited superior performance in reducing strength loss and permeability while enhancing microstructural integrity. This study provides critical insights into optimizing cementitious materials for enhanced durability in environmentally challenging conditions.