With increasing numbers of strength-deficient concrete infrastructure assets, strengthening and repair of concrete structures is becoming an issue of international importance. This study examines the bond behaviour of deep embedment (DE) glass fibre reinforced polymer (GFRP) and carbon FRP (CFRP) bars embedded into concrete prisms using an epoxy adhesive. The experimentally investigated parameters were the embedment length, FRP bar type and diameter, concrete compressive strength and hole diameter. The increase in embedded length enhanced the pull-out capacity for both GFRP and CFRP bar types. However, the bond strength and initial stiffness of the bond stress-slip curves decreased with the increase in embedded length. The specimens with DE CFRP bars had higher pull-out capacities and better bond performance than the corresponding specimens with DE GFRP bars. For the specimens with DE CFRP bars, the pull-out capacity increased with the increase in bar diameter and concrete strength but these two parameters did not affect the behaviour of the specimens with DE GFRP bars. The increase in hole diameter reduced the initial stiffness of the specimens with DE GFRP bars but affected neither the failure mode nor the failure loads. For the first time, this paper presents a mathematical 2 model for predicting the bond strength of DE FRP bars. The mathematical model was validated against experimental results and demonstrated to produce accurate predictions. It is envisaged that both the mathematical model and the experimental results will contribute to the development of future design guidelines for DE concrete shear strengthening.

This paper presents an overview of the deep embedment (DE) method – also referred to as the embedded through-section method – a promising technique for concrete shear strengthening. The aim of this work was to collate pull-out and beam test results with a view to identifying existing gaps in knowledge. A careful evaluation of pull-out test results showed that, except for the effect of embedded length, the bond performance of DE bars epoxy-bonded into concrete is not fully understood. Analysis of beam test results identified significant gaps in knowledge, particularly with regard to the effect of concrete strength, loading type, shear span-to-effective depth ratio, size effect and DE bar surface coating. Moreover, the analysis of beam test results revealed a novel interaction between steel tension and DE shear reinforcement. Finally, yet importantly, the paper summarises the research required for robust implementation of the DE technique.