Eco-Friendly Concrete Mix Design Using Steel Slag, Textile Sludge, and Polypropylene Fibers: A Hybrid Experimental and Machine Learning Approach
Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
The rapid growth of industrial production generates substantial solid waste, posing serious environmental challenges. Steel slag, textile sludge, and synthetic polymer residues are major disposal concerns. This study presents a novel triple-waste valorization strategy, converting these by-products into high-performance concrete blocks reinforced with polypropylene (PP) fibers. Fine aggregates were partially replaced with steel slag (0–15%) and textile sludge (0–15%). PP fibers were incorporated at up to 1.0% by volume to enhance tensile strength, crack resistance, and durability. M35-grade concrete mixes were prepared and evaluated for compressive, flexural, and splitting tensile strengths at 7, 14, and 28 days. Durability was assessed through water absorption and abrasion resistance tests. To complement experimental testing, machine learning models—Random Forest (RF) and Extreme Gradient Boosting (XGBoost)—were developed to accurately predict performance based on mix proportions, achieving high predictive accuracy with R² values above 0.94. Results identified an optimal mix containing 15% steel slag, 5% textile sludge, and 0.50% PP fibers. This formulation achieved a compressive strength of 48.37 MPa and flexural strength of 5.2 MPa, representing up to 18% improvement over the control mix. Water absorption decreased by approximately 22%, improving durability, while the inclusion of textile sludge slightly reduced density but significantly minimized environmental impact by diverting waste from landfills. This integrated experimental and AI-driven approach demonstrates that combining multiple industrial wastes with fiber reinforcement can produce sustainable, high-performance concrete. The study offers a replicable framework for eco-friendly construction and highlights the potential of machine learning to accelerate data-driven mix optimization, aligning structural performance with circular economy principles.