Optimizing concrete mix proportions with zeolite, GGBS, and CDW: a data-driven approach integrating experimental analysis and machine learning models

The depletion of natural sand resources and the environmental impact of cement production necessitate sustainable alternatives in concrete manufacturing. This study evaluates the potential of zeolite, ground granulated blast-furnace slag (GGBS), and construction and demolition waste (CDW) as partial replacements for sand in concrete mix proportions. Experimental investigations revealed that the optimal mix proportion, identified as Mix Batch M4 (60% Sand, 20% Zeolite, 10% GGBS, and 10% CDW), achieved a compressive strength (CS) of 67.37 MPa, flexural strength (FS) of 6.80 MPa, split tensile strength (ST) of 5.61 MPa, and notable reductions in water absorption (WA) to 4.00% and drying shrinkage (DS) to 4.02%. Additionally, durability improvements included a 30% reduction in rapid chloride permeability and enhanced ultrasonic pulse velocity (UPV) and rebound hammer (RH) values. Advanced machine learning models were utilized to analyze and optimize the mix designs, integrating the Sparrow Search Algorithm (SSA) with models such as Extreme Gradient Boosting (XGB), Backpropagation Neural Network (BPNN), Support Vector Regression (SVR), Random Forest Regression (RFR), Gradient Boosting Regression (GBR), and LightGBM. The XGB model demonstrated the highest predictive accuracy with an R 2 of 1.000. Multi-objective optimization techniques, including Non-dominated Sorting Genetic Algorithm II (NSGA-II), Multi-Objective Particle Swarm Optimization (MOPSO), and Genetic Algorithm with Fuzzy models, were employed to refine mix proportions further, balancing mechanical properties, material sustainability, and environmental benefits. This study highlights significant reductions in natural sand consumption and waste generation while enhancing concrete performance. Practical implications include reduced environmental impact, improved resource efficiency, and the promotion of circular economy principles. These findings provide a pathway toward innovative and sustainable concrete solutions, aligning with global sustainability goals in the construction industry.