Experimental evaluation, machine learning prediction, and life cycle assessment of sustainable self compacting concrete incorporating limestone powder and calcined clay
摘要
This study investigates the performance of self-compacting concrete (SCC) incorporating limestone powder (LP) and calcined clay (CC) as partial cement replacements to develop a sustainable ternary binder system. Cement was replaced with LP (10–20%) and CC (5–15%) while maintaining constant binder content. Fresh, mechanical, durability, microstructural, environmental, and machine learning-based predictive analyses were conducted. All mixtures satisfied EFNARC requirements for SCC. The optimum mixture containing 15% LP and 10% CC (SCC-L15C10) exhibited the best overall performance, achieving a 15.4% increase in 90-day compressive strength, along with improvements of approximately 13% in split tensile strength and 12% in flexural strength compared with the control mix. Durability performance was significantly enhanced, with reductions of about 15% in water absorption, 23% in sorptivity, 35% in chloride permeability, 36% in sulfate expansion, and 52% in alkali silica reaction (ASR) expansion. SEM and XRD analyses confirmed the formation of a denser matrix with reduced calcium hydroxide and enhanced C–A–S–H and carboaluminate phases. Machine learning models including Artificial Neural Network (ANN), Random Forest (RF), and XGBoost accurately predicted compressive strength, with RF achieving the highest accuracy (R² = 0.992). Life cycle assessment (LCA) indicated up to 23.5% reduction in CO2 emissions, while SCC-L15C10 achieved an 18.5% reduction relative to conventional SCC. The findings demonstrate that LP-CC based SCC is a promising low-carbon alternative for sustainable high-performance concrete applications.