A modeling strategy for the shear and flexural performance prediction of SFRC beams without stirrups accounting for the variability of properties

Abstract

AbstractThe behavior of steel fiber reinforced concrete (SFRC) is highly dependent on mix design, where the properties of its components, such as cementitious matrix and fibers, impact the physical and mechanical properties of the composite. Incorporating steel fibers into concrete modifies its physical properties, potentially leading to fiber segregation and resulting in a nonuniform distribution of the fibers within the material. In experimental studies, an increase in fiber volume has been found to change the failure mode of SFRC beams from shear to flexure, and numerical investigations have shown that randomness in fiber distribution can significantly influence the behavior of the composite. Therefore, this study aims to develop a numerical model that considers variations in material properties to simulate the change in the failure mode of SFRC beams due to variations in fiber volume. The failure and yielding parameters were initially established and integrated into the concrete damaged plasticity (CDP). Subsequently, the response of both steel and concrete to tension or compression stresses was ascertained. Finally, the appropriate meshes and finite elements were chosen for the simulation. In the final approach, two methods were employed to introduce randomness into the beam. One divided the beam into different vertical segments: 30, 45, and 90, and the other combined vertical and horizontal segments: 30 vertical with 4 horizontal, 45 vertical with 5 horizontal, and 90 vertical with 10 horizontal. A script was developed in Python to automatically insert the properties. The numerical model successfully captured the change in failure mode from shear to flexure resulting from fiber volume increase. Vertical and horizontal segments increased toughness due to crack deviation, better predicting cracking pattern and loading capacity for the highest fiber contents.