Eco-friendly concrete mix design incorporating steel slag, textile sludge, and polypropylene fibers through a hybrid experimental and machine learning approach

The rapid growth of industrial production has created massive quantities of steel slag and textile sludge that pose severe disposal and environmental challenges. This study develops an eco-friendly M35-grade concrete that integrates steel slag (SS), textile sludge (TS), and polypropylene (PP) fibers, supported by a hybrid experimental and machine-learning (ML) approach for mix optimization. Fine aggregates were partially replaced with SS (0–15%) and TS (0–20%), while PP fibers (0–1.0% by volume) were incorporated to enhance crack resistance and tensile strength. Experimental testing showed that 10% SS alone achieved the highest compressive strength of 50.58 MPa, representing a 16.7% increase over the control (43.36 MPa). The optimal multi-waste blend containing 15% SS, 10% TS, and 0.50% PP fibers reached 48.37 MPa compressive strength, 5.2 MPa flexural strength, and 7.4 MPa splitting tensile strength, along with about a 22% improvement in durability indices such as water absorption and abrasion resistance. The XGBoost model predicted concrete properties with high accuracy (R2 > 0.94 for all parameters), validating the data-driven optimization process. Feature-importance analysis revealed that SS primarily governs compressive strength, TS affects durability, and PP fibers enhance flexural and tensile behavior. This integrated experimental ML framework demonstrates that valorizing multiple industrial wastes, combined with fiber reinforcement, can produce structurally efficient and environmentally responsible concretes suitable for sustainable infrastructure.