Should FRP Be Bonded To Concrete

C. J. Burgoyne

Department of Engineering, University of Cambridge, Cambridge, UK

The question of whether it is right to bond tendons made of glass, aramid or carbon fibres to concrete has not yet been directly addressed. This paper discusses the various issues involved, and concludes that in many cases, these tendons should remain unbonded.

All the new materials which a high enough stiffness and a low enough creep show a linear elastic response right up to failure, with little or no ductility. This contrast with steel, even very high tensile steel, which shows a considerable reduction in stiffness at high loads. In a bonded beam, when cracks form on the tension face of the concrete, very high strains are generated across the crack. This allows the strain at the crack to reduce below its theoretical maximum value. In calculation, average steel strains are used, which ignore any local increase at the crack positions, but there are some controversial code rules which limit the (average) steel strain to less than the material can actually sustain.

When new materials are used, the local yielding mechanism is no longer available, and the concept of using average strains is no longer justified. In concrete reinforced with FRP, the whole strain capacity of the fibres is available, and it is unlikely that fibre failure will occur before the concrete strains become unacceptable. But in prestressed concrete, much of the fibre strain capacity is absorbed in the prestress, leaving a tendon very sensitive to high strains in the vicinity of cracks.