| Summary: | Concrete, a fundamental building material renowned for its strength, durability, and adaptability, plays a pivotal role in global construction. However, traditional concrete production exacts a toll on energy consumption and environmental footprint, primarily through aggregate extraction and processing. To address these challenges and to promote sustainability and reduce waste, this study investigates the partial substitution of fine and coarse aggregates with crushed brick powder (CBP) and recycled concrete aggregate (RCA), respectively. The fresh, mechanical, durability properties of CBP-RCA concrete were evaluated through tests such as slump, compacting factor, water permeability, compressive strength (CS), splitting tensile strength (TS), and flexural strength (FS). Scanning Electron Microscopy (SEM) was used to analyze microstructural changes in selected mixes. Additionally, the study evaluates machine learning algorithms such as extreme gradient boosting (XG Boost), random forest (RF), and bagging model (BAG) for predicting the mechanical strength of concrete specimens. These models supported the experimental findings by identifying key influencing factors and enabling accurate predictions, with XG Boost yielding the highest performance based on R2 and lower error metrics such as RMSE and MAE. The results showed that replacing 20 % of fine aggregates with CBP and 30 % of coarse aggregates with RCA led to the most optimum strength gains, with CS and TS increasing by 8.24 % and 2.89 %, respectively, after 28 days. However, higher replacement levels negatively impacted workability and strength due to reduced packing density. This study highlights the potential of combining experimental methods with ML-based prediction for sustainable manufacturing of concrete with optimized performance.
|
|---|
