Torsional behavior of concrete beams reinforced with longitudinal Fiber reinforced polymers (FRP) bars and FRP transverse stirrups has not been well-defined in various international codes and design guidelines as evidenced from the discrepancies in design equations. Such discrepancies are due to the complex behavior of beams subjected to torsional stresses and the limited research available in the literature; especially for FRP reinforced beams. A non-linear finite element (FE) analysis is conducted using ATENA software to generate a numerical model to predict and verify the failure loads and strains of beams of a compiled database from literature. This paper presents a comparison between the results of the experimental data and analytical models in order to verify the efficiency of the proposed numerical model for predicting failure load/strains in longitudinal FRP bars, FRP stirrups, and rotation angle of the tested beams. The comparison showed good agreement between experimental and analytical models in ultimate torque with difference of 6.5 % and the model is used in a parametric study to investigate the effect of concrete compressive strength, longitudinal reinforcement ratio, and beam depth under pure torsion in addition to the effect of concrete contribution to the torsional resistance. The numerically computed cracking and ultimate torques are compared with results from theory of elasticity, skew-bending theory and code equations such as: ACI-318 and CSA-S806-12. Based on this study, CSA-S806 gives acceptable prediction of ultimate torque, while ACI code is found to be more conservative. The increase of concrete compressive strength can enhance the cracking and ultimate torque. The increase of longitudinal reinforcement ratio has an insignificant effect on ultimate torque but increases the twist capacity of Reinforced concrete (RC) beams. The increase of longitudinal reinforcement can improve also the twist ductility.
Read full abstract