Elastic modulus of steel fiber-reinforced concrete: application of optimized algorithms

Incorporating steel fibers into plain concrete has been shown to effectively improve the load-bearing capacity of structural elements. As of now, the impact of fibers on the fundamental mechanical characteristics of Steel Fiber Reinforced Concrete (SFRC) and the associated uncertainties remain inadequately characterized, and precise models for predicting their mechanical properties are lacking. This significantly restricts the applicability in actual buildings. The study investigates various factors affecting settlement, including the water-tocement ratio, sand-to-cement mass ratio, coarse aggregate-to-cement mass proportion, fiber reinforcement coefficient, fiber morphology parameter, coarse aggregate size, superplasticizer-to-cement mass ratio, and fiber tensile yield strength. This research evaluates three approaches for forecasting Ecy, emphasizing the use of a hybrid optimization approach, the Electric Eel Foraging (EEF), with Gradient Boosting regression (GB) and Random Forests regression (RF) methods. Based on the available data, it is likely that RF-EEF and GB-EEF will be able to accurately compute the E_cy. The GB-EEF approach exhibited a high level of functional dependability throughout the training and evaluation phases, as shown by the R^2 values of 0.9764 and 0.9822 . Findings of 0.9892 and 0.9965 , achieved utilizing the RF-EEF technique, were relatively close. Feature importance depicted the highest impact of coarse aggregates to cement ratio and fiber kinds on the target by 0.9961 and 0.995.