| Resumo: | The rational use of natural, economic and social resources in order to ensure the sustainability and a long-term balance has become one of the largest global concerns. In the civil engineering field, the limited durability of steel reinforced concrete structures, especially in aggressive environments, and the high costs of the repair and maintenance operations have motivated the search for alternative materials and solutions to steel. One of these alternative reinforcements is the glass fibre reinforced polymer (GFRP) bars due to their immunity to corrosion, which is an important advantage when comparing to steel. However, several factors such as the novelty in the market, the high fabrication costs, the different design philosophies and the uncertainties of its behaviour with the concrete have been delaying the use of the GFRP bars in a larger scale. This thesis aims to contribute to the scientific knowledge of the GFRP reinforced concrete, as it studies its behaviour and design. The research work is mainly experimental and is based on a campaign with 24 full-scale reinforced concrete (RC) beams 4.30 m long and rectangular crosssection of 0.25 x 0.40 m2, divided into two groups with different purposes: - 18 beams to study the performance of different GFRP bar layouts as shear reinforcement; - 6 beams to assess the behaviour of a rehabilitation solution with GFRP bars to replace the deteriorated flexural steel reinforcement. The specimens of the first group were designed to fail due to shear with four different GFRP shear reinforcement solutions: 1) closed hoop GFRP stirrups, 2) two C shaped GFRP bars forming a stirrup, 3) two double headed GFRP bars and 4) two simple straight GFRP bars. Two shear reinforcement ratios with different spacing were also tested with the closed hoop GFRP stirrups. For each GFRP shear reinforcement layout, three different longitudinal stiffnesses were considered using steel and GFRP bars with different ratios. The beam specimens were tested until failure under a four point loading set-up and both the serviceability and the ultimate performance were analysed. The results were reported in terms of deflections, crack pattern, crack width, strains in the longitudinal and shear reinforcements, ultimate load capacity and failure modes. The different shear layouts were compared regarding their load carrying performance and their field implementation easiness. The design of the beams and their result predictions were made according to the existing guidelines and codes. It was concluded that the closed hoop stirrups and the C-stirrups were the most efficient and that the beams load capacity was highly underestimated by the GFRP codes. To improve the design formulas of these codes, different values for the limit strains and for the strut angle were proposed. The double headed bars as shear reinforcement were also efficient in the cases with higher longitudinal stiffness because it contributed to keep the integrity of the beam by exhibiting low deflections and crack widths. It was observed that a wide crack at the end of these bars highly compromises the anchorage function of the head. The solution of the simple straight bars was not effective because of the lack of anchorage length. The idea for the second group of beams was inspired on the RC structures with deteriorated bottom concrete due to the corrosion of the longitudinal steel reinforcement. Actually, no steel corrosion was considered in these specimens, but they were concreted in two phases to simulate the replacement of the deteriorated concrete, starting at the stage after its complete removal. The rehabilitation procedure consisted on the insertion of the longitudinal GFRP bars and the concreting of a new bottom layer in the beam. Two solutions with different GFRP longitudinal cross-section areas were designed according to the existing guidelines, one to restore the ultimate load capacity of the original beam, and the other to maintain the deflection of the original beam. The ends of the GFRP bars were conic heads to compensate their lower anchorage length. The rehabilitated beam specimens were subjected to 3 point bending tests until failure, and their service and ultimate behaviour were analysed. Results are presented in terms of deflection, crack pattern, mid-span crack width, reinforcement strains, ultimate flexural capacity and failure modes. It was concluded that this technique was effective for both the serviceability and ultimate limit states of the rehabilitated beam, as it was able to restore the deflection and the load capacity of the original beam, and that the existing GFRP design documents can be used. Although this was mainly an experimental research work, a simple but reliable two-dimensional finite element (FE) model was defined using ATENA software to simulate the tests, which helped to better understand some issues regarding the specimens behaviour and enabled to extrapolate some results of non-tested possibilities. The linear and nonlinear behaviour of all materials was adequately modelled by appropriate constitutive laws. Furthermore, numerical results were compared with the experimental results. Results show that, in general there was a good agreement between the overall modelling results and the experimental ones. The constructed models were able to predict the experimental behaviour in terms of ultimate capacity and load-deflection curves. Regarding the first group of beams, two additional stirrups spacing were modelled in order to clarify its influence in the shear capacity. It was simulated different longitudinal reinforcement ratios to assess its influence in the shear capacity. As a final remark, the results of the present work show that the use of GFRP bars is viable in RC structures, which contributes to more durable structures in long-term. This material can be used as longitudinal and shear reinforcement of new structures and as a rehabilitation solution to replace the corroded steel in deteriorated structures.
|
|---|