Strength–Permeability Optimization of FA–MK–NS Blended Pervious Concrete Based on Response Surface Methodology

Abstract

Pervious concrete must balance mechanical capacity with interconnected voids required for drainage. A three-factor, three-level Box–Behnken design examined fly ash (FA; 10–20%), metakaolin (MK; 5–15%), and nano-silica (NS; 0.5–1.5%) at a fixed total binder content of 380 kg/m3. Compressive strength, water-accessible open porosity, and the apparent permeability coefficient ranged from 16.80 to 28.20 MPa, 14.07 to 24.50%, and 2.20 to 7.29 mm/s, respectively. Quadratic models for compressive strength and the apparent permeability coefficient were statistically adequate (R2 = 0.9907 and 0.9861); the porosity model showed significant lack of fit. The porosity model was retained only for local trend interpretation and excluded from optimization. In a researcher-defined design scenario that maximized strength while targeting an apparent permeability coefficient of 3.00 mm/s, desirability analysis selected a model-predicted compromise solution containing 14.038% FA, 12.556% MK, and 1.483% NS, with a predicted compressive strength and an apparent permeability coefficient of 28.233 MPa and 3.000 mm/s. The NS factor was close to the upper boundary, and the solution was not experimentally validated. Selected single-field SEM observations illustrated local differences among M5, M7, and M12, while selected-area EDS sum spectra provided only local elemental composition information. The results indicate potential for future pavement evaluation after independent validation, multi-field microstructural analysis, durability testing, and economic assessment.