Sustainable use of red mud in concrete: Assessing mechanical strength, durability, and performance through machine learning models

Red mud that is produced as a residue of alumina from bauxite ore through the Bayer process is undesirable or has potential environmental problems due to its large volume and high pH. Red mud was incorporated into the concrete mixtures at 0%, 5%, 10%, 15%, and 20% to determine the amount that enhances the strength and durability of the concrete. Furthermore, four machine learning models, including MARS, MPMR, GMDH, and ENN, were used in the present work to assess the compressive strength of red mud concrete and compare their efficacies. The replacement of cement with red mud in the range of 10% to 15% has beneficial effects on the acid, sulfate, and chloride resistance, as well as on the compressive and flexural strength, of concrete. Replacing 10% increased the compressive strength, and replacing 15% increased the flexural strength and durability. Replacement above 15% resulted in a decrease in the durability and strength of the concrete, which suggested the necessity of careful optimization. Among all the formulated models, the MARS model is the best predictor of compressive strength prediction on the basis of performance indicators, Taylor diagrams and comprehensive measurement analysis. Finally, the research validates that using red mud in concrete can be a sustainable solution that will lead to a green construction approach with long-standing impacts on the construction industry.