Predictive Modelling of Micro-Reinforced Rubberized Concrete Using ANFIS

This study explores the performance of rubberized concrete with the inclusion of steel fibers through a neural network approach for the determination of its strength parameters. Rubberized concrete is made by partially replacing conventional aggregates with recycled rubber. The use of rubberized concrete reuses waste materials and improves some of the mechanical properties but typically at the reduction of compressive strength. To overcome this limitation, it introduces steel fibers to boost the strength and ductility of the materials. This research study investigates combine impacts of rubber particles and steel fibers on performance of concrete. A developed neural network model is adopted to predict the key parameters concerning compressive strength, flexural strength, and modulus of elasticity with variable ratios of rubber and steel fibers. The sample is trained on experimental data for a broad range of concrete mixes, thus allowing for good predictions and optimization of the material composition. Results showed that rubberized concrete with steel fibers could maintain adequate strength while offering improved flexibility and crack resistance, making it a viable alternative to conventional concrete in certain structural applications. This approach also showcases the potential of machine learning techniques in advancing sustainable construction materials.