Strengthening of unbonded post-tensioned concrete systems using external FRP composites : experimental evaluation and analytical modeling

Abstract

Since its establishment in 2002, based on extensive research in the area of FRP strengthening, the ACI Committee 440 has been providing guidelines for using external FRP composites for flexural, shear and axial strengthening of reinforced concrete (RC) structural systems. It was only recently that the ACI Committee 440 (2008) provided guidelines involving the use of FRP technology for flexural strengthening of posttensioned prestressed concrete (PC) members. However, unfortunately, these guidelines covered only pretensioned or bonded post-tensioned PC members and excluded posttensioned members with internal or external unbonded tendon system. It is believed that the lack of experimental research on FRP strengthened unbonded PC members, together with the analytical difficulty in evaluating the strain/stress in the unbonded tendons at ultimate because of slip of the tendons relative to the surrounding concrete, are the two main reasons which hindered the development of guidelines for the analysis and/or design of FRP strengthened unbonded flexural members. The study presented in this report concentrated on experimental and analytical evaluation of the potential use of FRP laminates for strengthening unbonded posttensioned concrete members covering both slabs and beams. In the experimental part, twenty four full-scale simply supported beam and slab specimens reinforced with internal unbonded tendon system and strengthened using external FRP composites were tested. Additional twelve companion bonded prestressed concrete (PC) and reinforced concrete (RC) specimens were also tested for comparison. In addition to the type of structural system (unbonded PC, bonded PC, RC), the test parameters included area of the internal tension reinforcement, area of the external FRP reinforcement, span-todepth ratio of the member (slab, beam), and profile of the unbonded tendons. The test results showed that the use of FRP laminates increases the load capacity and postcracking stiffness of unbonded members. The increase in load capacity was accompanied with reduction in deformation capacity. One of the most important findings is that no distinct difference beyond expectation was observed between the flexural responses of FRP strengthened unbonded PC and those of bonded PC or RC systems. Therefore, provided a method is available for calculating the strains or stresses viii in the unbonded tendons at ultimate taking into account the effect of FRP reinforcement, the same guidelines reported by ACI Committee 440 for designing the FRP system for flexural strengthening of RC and bonded PC members can be extended for unbonded members. In the analytical part, a design oriented approach for evaluating the tendon stress at ultimate and the nominal moment capacity of FRP strengthened post-tensioned PC members with internal or external unbonded tendon system is developed. The approach is consistent with the procedure proposed in the ACI Committee 440 report for RC or bonded PC members and accounts for all the parameters that influence the nominal flexural capacity of unbonded PC members. These include area of the tension reinforcement (FRP, unbonded prestressing steel, reinforcing steel), span-to-depth ratio of the member, member continuity (simply supported, continuous), and loading pattern in continuous members. The accuracy of the proposed approach was verified, vis-a-vis the approach recommended by the ACI Committee 440 (2008) for RC and bonded PC members, by comparing with the test results of the experimental part of this investigation. Furthermore, using constitutive material models and a number of simplifying assumptions, a 3D non-linear finite element (FE) analysis was carried out for reproducing the test results of the various specimens tested in the experimental part of this investigation. These include the full load-deflection response and the responses of applied load versus strain in the prestressing steel and applied load versus strain in the FRP reinforcement. Despite little discrepancies, the FE analysis predicted the test results with remarkable agreement.