Title:
Predicting Geopolymer UHPC Strength Using Machine Learning (Prepublished)
Author(s):
Kamran Aghaee and Kamal H. Khayat
Publication:
Materials Journal
Volume:
Issue:
Appears on pages(s):
Keywords:
compressive strength; ensemble machine learning; environment; geopolymer ultra-high-performance concrete; sustainability; ultra-high-performance concrete (UHPC)
DOI:
10.14359/51747873
Date:
6/11/2025
Abstract:
Ultra-high-performance geopolymer concrete (UHP-GPC) can exhibit high to exceptional strength. Given the importance of UHP-GPC’s mechanical properties, the prediction of its 28d compressive strength (f’c) remains insufficiently explored. This study predicts UHP-GPC’s f’c based on alkali-activated materials, sand, fiber volume, water-to-geopolymer binder, and alkali activator ratios. Advanced statistical modeling and a spectrum of ensemble machine learning (ML) algorithms, including random forest (RF), gradient boosting (GB), extreme gradient boosting (XGB), and stacking, are utilized to predict UHP-GPC’s strength. The derived models reveal the significance of fiber, slag, and sand as the most significant factors influencing the 28d f’c of UHP-GPC. All the ML models demonstrate higher precision in forecasting f’c of UHP-GPC compared to statistical modeling, with R2s peaking at 0.85. Equations are derived to predict the strength of UHP-GPC. This article reveals that UHP-GPC with superior mechanical properties can be designed for further sustainability.