Scholarlink

<p>In recent years, the construction industry has increasingly focused on sustainable materials, prompting the exploration of agricultural by-products like corncob ash (CCA) as additives in concrete production. This study aims to predict the compressive strength of CCA-blended concrete using advanced machine-learning techniques. A comprehensive dataset comprising 405 samples was compiled, capturing key parameters such as fine aggregate-to-binder ratio (F.Agg/B), coarse aggregate-to-binder ratio (C.Agg/B), CCA-to-binder ratio (CCA/B), water-to-binder ratio (W/B), and curing periods. The compressive strength values ranged from 8.17&#xa0;MPa to 78.6&#xa0;MPa, highlighting significant variability. Four machine learning models were employed: Extreme Gradient Boosting (XGB), Artificial Neural Networks (ANN), K-Nearest Neighbors (KNN), and Support Vector Regression (SVR). The XGB model exhibited the highest predictive accuracy, achieving a training R² of 0.98 and a test R² of 0.93, with Root Mean Square Error (RMSE) values of 1.93 (training) and 3.75 (testing). The ANN model followed closely with a training R² of 0.96 and RMSE values of 2.78 (training) and 4.23 (testing). Meanwhile, KNN and SVR models showed lower performance, with training R² values of 0.89 and 0.87, respectively. The SHAP analysis identified vital factors influencing compressive strength in corncob ash (CCA)-blended concrete. Notably, the CCA-to-binder ratio positively impacts strength, while the water-to-binder ratio negatively affects it. Extended curing periods enhance strength, providing valuable insights for optimizing concrete formulations using agricultural by-products. These findings underscore the potential of CCA for improving the sustainability of concrete while providing a foundation for future research to optimize concrete formulations. Ultimately, this work advances eco-friendly construction practices by utilizing agricultural waste.</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

Innovative use of corncob ash in concrete: a machine learning perspective on compressive strength prediction

  • 期刊论文

  • 发表日期：

    2025年2月4日

查看全文
  • Navaratnarajah Sathiparan

摘要

In recent years, the construction industry has increasingly focused on sustainable materials, prompting the exploration of agricultural by-products like corncob ash (CCA) as additives in concrete production. This study aims to predict the compressive strength of CCA-blended concrete using advanced machine-learning techniques. A comprehensive dataset comprising 405 samples was compiled, capturing key parameters such as fine aggregate-to-binder ratio (F.Agg/B), coarse aggregate-to-binder ratio (C.Agg/B), CCA-to-binder ratio (CCA/B), water-to-binder ratio (W/B), and curing periods. The compressive strength values ranged from 8.17 MPa to 78.6 MPa, highlighting significant variability. Four machine learning models were employed: Extreme Gradient Boosting (XGB), Artificial Neural Networks (ANN), K-Nearest Neighbors (KNN), and Support Vector Regression (SVR). The XGB model exhibited the highest predictive accuracy, achieving a training R² of 0.98 and a test R² of 0.93, with Root Mean Square Error (RMSE) values of 1.93 (training) and 3.75 (testing). The ANN model followed closely with a training R² of 0.96 and RMSE values of 2.78 (training) and 4.23 (testing). Meanwhile, KNN and SVR models showed lower performance, with training R² values of 0.89 and 0.87, respectively. The SHAP analysis identified vital factors influencing compressive strength in corncob ash (CCA)-blended concrete. Notably, the CCA-to-binder ratio positively impacts strength, while the water-to-binder ratio negatively affects it. Extended curing periods enhance strength, providing valuable insights for optimizing concrete formulations using agricultural by-products. These findings underscore the potential of CCA for improving the sustainability of concrete while providing a foundation for future research to optimize concrete formulations. Ultimately, this work advances eco-friendly construction practices by utilizing agricultural waste.