Ultimate torsional resistance and failure modes of FRP-strengthened reinforced concrete members: A nonlinear design model

Abstract

Reinforced concrete (RC) structural members, especially flanged and spandrel beams, bridge girders, curved beams, and eccentrically loaded beams may be subjected to significant torsional loading. Despite extensive research on the strengthening of RC members, far too little attention has been paid to the torsional behavior of RC members strengthened with fiber-reinforced polymer (FRP) sheets. Therefore, a limited number of guidelines specified the design of externally bonded FRP systems in torsion. This study provides an important opportunity to advance the understanding of the torsional behavior of FRP-strengthened members by developing a nonlinear design model predicting the failure mode and ultimate torsional resistance of FRP-strengthened RC members subjected to torsion. Serious discussions and analyses of desired failure modes emerge by defining two balanced conditions between the components, and eight failure regions. The first balanced condition defined between FRP and concrete specifies the failure mode of the member, and the second balanced condition defined between steel and concrete is used for the analysis of steel reinforcement stresses. A clear design approach is proposed in this study to access the desired percentage of the externally applied FRP sheets based on the required torsional strength. Moreover, the ultimate stresses and strains of concrete struts, steel reinforcements, and FRP sheets are derived by incorporating the nonlinear behavior of concrete. To evaluate the validity of the design method, 70 FRP-strengthened specimens are collected from 11 previous studies; then, a comparative analysis is conducted among the experimental results, the estimations of the presented model, and the predictions of the commonly available guidelines for torsional strengthening. The excellent agreement between the test results and the presented model estimations confirms the validity and accuracy of the presented model.