This study presents a comprehensive investigation into the optimization of hybrid fiber reinforced concrete (HFRC) containing basalt and steel fibers using Response Surface Methodology (RSM) and predictive modeling through Machine Learning (ML). The experimental program employed a Central Composite Design (CCD) with basalt fiber (BF) and steel fiber (SF) volume fractions ranging from 0.25% to 0.75% to evaluate workability and mechanical properties, including compressive, split tensile, and flexural strengths, as well as elastic modulus. Statistical analysis revealed significant interactions between BF and SF, with R 2 values between 0.85 and 0.97 indicating excellent model fit. Multi-objective optimization achieved an optimal mix of 0.42% BF and 0.75% SF, yielding compressive strength of 52.58 MPa and split tensile strength of 6.10 MPa with a composite desirability of 0.68. Experimental validation confirmed the RSM model accuracy within a 7% error margin. Furthermore, machine learning models Decision Tree, Random Forest, and XGBoost were trained on 250 data points to predict compressive strength. The Random Forest Regressor demonstrated the best predictive performance with R 2 = 0.84, RMSE = 2.75 MPa, and MAE = 2.18 MPa. SHAP analysis identified the water-cement ratio, BF, and SF as the most influential parameters affecting compressive strength. The integrated RSM–ML framework offers an efficient approach for mix design optimization and property prediction of HFRC, supporting data-driven development of high-performance and sustainable concrete. This study directly contributes to UN Sustainable Development Goal (SDG) 9 (Industry, Innovation and Infrastructure) by promoting advanced material optimization and SDG 11 (Sustainable Cities and Communities) through the development of durable, resource-efficient concrete, while also aligning with SDG 12 (Responsible Consumption and Production) by enabling optimized material usage and reduced experimental waste.
Vengadeshwari et al. (Fri,) studied this question.