Data-Driven Modeling of Mechanical Properties of Fiber-Reinforced Concrete: A Critical Review

Fiber-reinforced concrete (FRC) is extensively used in diverse structural engineering applications, and its mechanical properties are crucial for designing and evaluating its performance. The compressive, flexural, splitting tensile, and shear strengths of FRCs are among the most important attributes, which have been discussed more extensively than other properties. The accurate prediction of these properties, which are required for design criteria, has been a challenge because of their high uncertainties. Statistical and empirical models have been extensively utilized. However, such models require extensive experimental work and can produce incorrect outcomes when there are complicated interactions among the qualities of concrete, the makeup of the blend, and the curing environment. To address this issue, machine learning (ML) methods have been increasingly applied in recent years to solve complex structural engineering problems. Predictive models can provide a strong solution for time-consuming numerical simulations and expensive experiments. This study explores the ML techniques applied in this context and provides a comprehensive analysis of artificial intelligence methods used for predicting the mechanical properties of FRCs. It also highlights the key observations, challenges, and future trends in this field. This study serves as a valuable resource for researchers in selecting accurate models that match their applications. It also encourages material engineers to become familiar with and employ ML methods to design FRC mixtures appropriately.