Carbon footprint reduction and performance optimization of sustainable free cement concrete with eggshell powder and rice husk ash using machine learning | ||
| International Journal of Sustainable Development and Science | ||
| Volume 7, Issue 1, 2024, Pages 195-213 PDF (1.42 M) | ||
| Document Type: Original Article | ||
| DOI: 10.21608/ijsrsd.2024.396837 | ||
| Authors | ||
| Mahmoud Abdellatief1; Mohamed Abdellatief* 2; Ezzat Elfadaly1; Hassan Hamouda3 | ||
| 1Evaluation of Nat. Resources Dep., Environmental Studies and Research Institute, University of Sadat City, Egypt | ||
| 2Department of Civil Engineering, Higher Future Institute of Engineering and Technology in Mansoura, Egypt | ||
| 3Civil and Architectural Construction Department, Faculty of Technology and Education, Suez University, Egypt | ||
| Abstract | ||
| This article explores the performance and carbon footprint reduction of geopolymer mortar (GM) that incorporates eggshell powder (ESP) and rice husk ash (RHA) as sustainable alternatives to traditional binders. Using response Surface Methodology (RSM), ESP and RHA were added at volumetric percentages from 0% to 30% as partial replacements for GGBS. The experimental findings revealed that the inclusion of RHA and ESP significantly enhances compressive strength, particularly at optimal dosages, with the highest recorded strength reaching 48 MPa. RSM effectively predicted compressive strength values, aligning well with experimental data. Furthermore, machine learning models, including Gaussian Process Regression (GPR), Artificial Neural Networks (ANN), and Gradient Boosting (GB), were employed to analyze the compressive strength predictions, with GPR demonstrating superior accuracy. An ecological assessment indicated that using RHA and ESP can lower CO₂ emissions compared to traditional materials, thereby promoting more sustainable construction practices. Finally, the dataset of 606 compressive strength results validated the effectiveness of the GPR, ANN, and GB models, all showing high predictive accuracy (R² > 0.85), with the GPR model outperforming the others. | ||
| Keywords | ||
| Geopolymer mortar; Eggshell powder; Rice husk ash; Machine learning | ||
|
Statistics Article View: 324 PDF Download: 250 |
||
