Evaluation and modification of bend corner strength prediction models of FRP reinforcement

Abstract

The bend corner of Fiber-reinforced polymer (FRP) reinforcement has been shown significantly lower tensile strength than the straight portion. To date, various prediction models were developed for the bend corner strength. In this study, three types of six prediction models are reviewed and evaluated based on a most extensive database to date, including tests from 15 investigations from 1999 to 2020. The validities of the models for cross-section types and test methods are analyzed. The logarithmic and linear regression models are further modified. The macro-mechanical model provides the best predictions for the whole database with similar validity for test specimens with different cross-section types. The linear regression and logarithmic models show lower accuracy and inconsistency in cross-section types. The different test methods adopted in the literature show limited influence on the bend corner strength and the performance of the prediction models. The proposed cross-section geometry factor can effectively consider the cross-section types and the modified linear regression and logarithmic models could provide more accurate predictions. The model evaluations and modifications in this study could provide a more comprehensive understanding of the calculation philosophy of the bend corner strength of FRP reinforcement, which could be used for the future updates of shear design provisions for FRP reinforced concrete structures.