Calibration of group effect parameters through genetic algorithms for micromechanical modeling of UHPFRC

Abstract

Ultra-High Performance Fiber Reinforced Concrete (UHPFRC) is a cement-based composite with superior mechanical and durability properties compared to conventional concrete. The inclusion of short fibers in the brittle matrix can change the behavior to ductile and tough. Several experimental investigations on fiber content and orientation on the UHPFRC tensile properties have been reported in contrast to the few numerical studies. Moreover, most numerical researches consider each phase (fiber and matrix) modeled individually, resulting in complex models with a high computational cost. A constitutive model that allows homogeneous modeling accounting simultaneously for fiber content and orientation improves the simulation processing time for practical applications. Therefore, this paper presents a micromechanical-based model for UHPFRC tensile behavior considering fiber pullout, fiber content and orientation in the composite, and group effect. The group effect was considered in the constitutive model through the parameters ξ and β capable of considering the bond strength reduction and the change in the shape of the composite post-cracking curve due to the iteration between the fibers. The fiber spacing decreases as the fiber content increases, and their interaction cannot be neglected. If the group effect is not considered, the material response is overestimated. The parameters ξ and β were calibrated by the finite element model updating, which is a commonmethodto improve the correlation betweenthe finite element and experimental model, using Genetic Algorithms (GA). The parameters were dependent on fiber content and orientation. Besides, heterogeneous models were developed to compare the results. It was possible to obtain a constitutive model representing the UHPFRC tensile behavior based on several numerical simulations and present guidelines for the material simulation. The numerical results showed excellent agreement with the considered experimental results from the literature. From the parametric analysis, it is concluded that the interfacial parameters and the fiber content and orientation significantly influence the response of the UHPFRC.