Scholarlink

<p>In this study, a bio-derived graphene-based nanosheet (GNS) synthesized from recycled organic waste was incorporated into cement mortars to enhance mechanical and physical performance. The effects of GNS on compressive strength, flexural strength, bulk density, and water absorption were systematically evaluated. The results showed significant mechanical improvements, with compressive and flexural strengths increasing by up to 37% and 56%, respectively, compared with the control mortar. In parallel, machine learning models – multilayer perceptron artificial neural networks (MLP-ANN), Extreme Gradient Boosting (XGBoost), and Light Gradient Boosting Machine (LightGBM) – were developed to predict compressive strength using experimental data at curing ages of 7, 14, and 28 days. Among the evaluated models, the ANN achieved the highest predictive accuracy, with <i>R</i><sup><i>2</i></sup> values of 0.92 and 0.90 for the training and testing datasets, respectively. SHapley Additive exPlanations (SHAP) were further employed to interpret the ANN predictions and quantify the influence of curing age, water-to-binder ratio, GNS dosage, and cement content. Overall, the findings confirm the potential of bio-derived GNS as a sustainable mortar additive and highlight the effectiveness of ANN-based models for performance prediction and mixture optimization.</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

Bio-derived graphene nanosheets for sustainable enhancement of cement mortar properties: experiments and machine learning insights

  • 期刊论文

  • 发表日期：

    2026年5月25日

查看全文
  • Tuong Dat Dinh Do,
  • Quoc-Bao Bui,
  • Minh Tuan Le

摘要

In this study, a bio-derived graphene-based nanosheet (GNS) synthesized from recycled organic waste was incorporated into cement mortars to enhance mechanical and physical performance. The effects of GNS on compressive strength, flexural strength, bulk density, and water absorption were systematically evaluated. The results showed significant mechanical improvements, with compressive and flexural strengths increasing by up to 37% and 56%, respectively, compared with the control mortar. In parallel, machine learning models – multilayer perceptron artificial neural networks (MLP-ANN), Extreme Gradient Boosting (XGBoost), and Light Gradient Boosting Machine (LightGBM) – were developed to predict compressive strength using experimental data at curing ages of 7, 14, and 28 days. Among the evaluated models, the ANN achieved the highest predictive accuracy, with R2 values of 0.92 and 0.90 for the training and testing datasets, respectively. SHapley Additive exPlanations (SHAP) were further employed to interpret the ANN predictions and quantify the influence of curing age, water-to-binder ratio, GNS dosage, and cement content. Overall, the findings confirm the potential of bio-derived GNS as a sustainable mortar additive and highlight the effectiveness of ANN-based models for performance prediction and mixture optimization.