Effectiveness of Fiber Anchors in CFRP Flexural Strengthening of RC Girders

Abstract

This research presents an experimental program for flexural strengthening of reinforced concrete (RC) full-scale girders with externally bonded carbon fiber–reinforced polymer (CFRP) sheets secured with fiber anchors. The testing program consisted of five RC T-girders strengthened with identical CFRP sheets anchored with different numbers of CFRP fiber bundles. The main objective of this study is to address the effectiveness of the CFRP fiber anchors in controlling premature failure of the CFRP sheets due to debonding. To evaluate the experimental results, a simplified design model was introduced to predict the effective CFRP maximum strain limit associated with a given number of fiber anchors in the shear span. The test results show that using CFRP fiber anchors significantly delayed the failure of the CFRP sheet beyond the classical intermediate crack (IC) debonding failure mode. In addition, the presented design model agrees reasonably well with the experimental findings as it pertains to the various numbers of anchors. Nevertheless, an anchor efficiency factor is proposed herein to reflect the fiber anchor group effect based on the experimental findings. This paper presents an experimental program and a design model to capture the efficiency improvements of the response of carbon fiber–reinforced polymer (CFRP) strengthened full-scale reinforced concrete T-girders secured with increasing number of carbon fiber anchors along the shear span. Both the equivalent debonding strain and load capacity were improved owing to the use of increasing numbers of fiber anchors per shear span. This study showed that the ultimate failure of the girders can be controlled by the number of fiber anchors used to secure the CFRP sheets. Sectional analysis may not be strictly applicable once the progression of debonding starts. However, it is shown to capture a reasonably accurate measure of the maximum CFRP debonding strain without the need to resort to higher-order analysis that does not lend itself to design practice.