Externally unbonded post-tensioned CFRP tendons a system solution

Sammanfattning: The introduction of Fibre Reinforced Polymers (FRP) to the civil engineering market in the late 1980s resulted in the emergence of a range of new tools for rehabilitating and strengthening concrete structures. Strengthening using FRPs is typically accomplished using non-prestressed externally bonded FRPs. The technical and economic benefits of such strengthening could be further increased by prestressing the FRPs, especially when dealing with concrete structures. Prestressing concrete structures suppresses the appearance and growth of cracks in the serviceability limit state. This in turn increases the structure’s stiffness and resistance to degradation. Prestressing also increases the structure’s yield load but does not change its failure load relative to that of an analogous non-prestressed structure, provided that all other parameters are kept constant. In 2004, a pilot study was carried out at the Luleå University of Technology (LTU) to investigate the scope for using unbonded Carbon Fibre Reinforced Polymer (CFRP) strengthening systems, particularly those involving prestressing. In the early stages of this project, a number of difficulties were encountered in anchoring the CFRP rods to concrete structures: the conical wedge anchorages that were used tended to either cause premature failure of the rods or allowed the rod to slip out of the anchorage. It was therefore decided to study the mechanisms at work within these anchorages in more detail. The goal of the project was to develop a small, practical, reliable, and userfriendly anchorage for use in unbonded external CFRP strengthening systems. On the basis of a thorough literature review, which is described in Paper 1, it was concluded that despite the difficulties encountered, the conical wedge anchorages used with steel reinforcing rods were the most promising starting point for the design of a new anchorage for use with CFRPs. Importantly, the conical wedge anchorage can be made small in size and easy to mount while retaining a high degree of versatility; this is not true of bonded, sleeve, and clamping anchorages. Analytical and numerical models were used to investigate the distribution of radial stress within these highly pressurized anchorages. Paper 2 describes an evaluation of the capability of three types of models - an analytical axisymmetric model based on the thick-walled-cylinder-theory and two Finite Element (FE) models, one axisymmetric and one three-dimensional - to predict the behaviour of a conical wedge anchorage. It was concluded that the axisymmetric models were incapable of modelling the stress distribution within the anchorage with sufficient accuracy, and so 3D FE models were used exclusively in subsequent studies. Paper 3 describes the development of a new anchorage for CFRP rods. The design process involved conducting pull-out studies on a series of prototypes, in conjunction with computational studies using a basic FE model, to identify and understand the prototypes’ failure modes. Between the computational data and experimental results, a good understanding of the factors affecting the interaction between the CFRP rod and the anchorage was obtained. The new anchorage design employs a one-piece wedge which effectively incorporates the three wedges and the inner sleeve from more conventional wedge anchorages into a single unit. This increases the reliability and user-friendliness of the anchorage because it eliminates the need to check the alignment of individual wedges. The new design has been patented; the published Swedish patent is included in the thesis as Paper 6. The newly-developed anchorage was then incorporated into a prestressing system and its performance was evaluated using a series of test beams. In parallel with the planning of these tests, a series of pull-out tests was conducted using the new anchorage. The strain measurements obtained in these experiments were compared to predictions made using a new, more advanced FE model, and used to refine the design of the new anchorage. Paper 4 describes this new FE model, the most important parameters affecting anchorage behaviour, and the final anchorage design. Paper 5 focuses on the possibilities provided by the new anchorage. Tests were performed using seven three meter long concrete beams prestressed with external unbonded CFRP tendons. One beam was unstrengthened; the other six were strengthened in different ways, with different prestressing forces, initial tendon depths, and with or without the use of a midspan deviator for the tendons. The results of these tests were compared to those obtained using otherwise identical beams prestressed with steel tendons and to the predictions of an analytical beam model developed for use with steel tendons. These tests showed that the prestressing works as intended and that the behaviour of beams prestressed with external unbonded CFRP tendons is fully comparable to that of beams prestressed with steel tendons. It was also found that the predictions of the analytical model were in good agreement with experimental observations, although there were some differences between the measured and predicted tendon stresses. The development of a functional anchorage represents a fulfilment of the objectives laid out at the start of this project, and represents an important step towards the practical use of prestressed unbonded external CFRP tendons in strengthening concrete structures. However, a number of outstanding questions remain to be addressed. Little is known about the safety of this kind of system, and the benefits of using CFRP tendons should be quantified. Furthermore, there are a number of potential technical issues that must be addressed. These include the risk of creep-rupture in the CFRP, the effects of thermal contraction and expansion on the anchorage, and the scalability of the anchorage as the tendon diameter is increased. Finally, the long-term behaviour of the anchorage and prestressing systemshould be investigated.