Scholarlink

<p>This study examines the bond strength between normal-strength concrete and steel reinforcement, incorporating pozzolanic materials to enhance bond behaviour through partial cement replacement. The pozzolanic materials studied include fly ash, silica fume, metakaolin, rice husk ash, ground granulated blast furnace slag, and waste glass powder. Experimental data from previous studies were collected and used to develop predictive models using four machine learning techniques: Gaussian Process Regression, Support Vector Machines, Artificial Neural Networks, and Random Forest Regression (RFR). These models were trained and tested on a dataset comprising 97 samples, evaluating the impact of key factors such as water-to-binder ratio, cement content, pozzolanic material content, aggregate content (sand and gravel), superplasticizer content, and curing time on bond strength. The study found that curing time and coarse aggregate content were the most influential parameters. The Random Forest Regression (RFR) model was determined to be the most accurate model for predicting bond strength, providing insights into the complex relationships between input parameters and bond performance. This work contributes to the growing body of knowledge on sustainable concrete design, offering practical implications for improving the durability and performance of concrete structures while reducing environmental impact.</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

Predicting Bond Strength Between Normal Concrete with Pozzolanic Materials and Steel Bars Using Machine Learning Techniques

  • 期刊论文

  • 发表日期：

    2025年7月28日

查看全文
  • Fkrat Latif Hamid,
  • Bedar Rauf Hassan,
  • Diar Fatah Abdulrahman Askari,
  • Dler Ali Ahmad,
  • Sardam Salam Shkwr,
  • Soran Abdrahman Ahmad

摘要

This study examines the bond strength between normal-strength concrete and steel reinforcement, incorporating pozzolanic materials to enhance bond behaviour through partial cement replacement. The pozzolanic materials studied include fly ash, silica fume, metakaolin, rice husk ash, ground granulated blast furnace slag, and waste glass powder. Experimental data from previous studies were collected and used to develop predictive models using four machine learning techniques: Gaussian Process Regression, Support Vector Machines, Artificial Neural Networks, and Random Forest Regression (RFR). These models were trained and tested on a dataset comprising 97 samples, evaluating the impact of key factors such as water-to-binder ratio, cement content, pozzolanic material content, aggregate content (sand and gravel), superplasticizer content, and curing time on bond strength. The study found that curing time and coarse aggregate content were the most influential parameters. The Random Forest Regression (RFR) model was determined to be the most accurate model for predicting bond strength, providing insights into the complex relationships between input parameters and bond performance. This work contributes to the growing body of knowledge on sustainable concrete design, offering practical implications for improving the durability and performance of concrete structures while reducing environmental impact.