Date of Award
Doctor of Philosophy (PhD)
Intermediate crack (IC) induced debonding failure of Fiber Reinforced Polymer (FRP)-strengthened reinforced concrete beams starts at the tip of flexural/shear cracks within the shear span and propagates towards the FRP plate termination. In this study, experimental and numerical programs are performed to characterize and predict the failure mechanisms of IC debonding failure, and identify the key parameters affecting such failure. It is found that the bond-slip relation obtained from the pullout test does not represent the bond-slip relation of the FRP/concrete interface in the FRP-strengthened concrete beams, and it cannot be directly used for predicting the load capacity of the FRP-strengthened concrete beams. A mathematical and systematic method is also successfully established to predict the variation of the FRP-bonded concrete specimens' capacities. In the experimental program, the bond-slip behavior of the FRP/concrete interface is obtained by single shear pullout and beam tests. In the beam specimens, a notch at the mid-span of the beam represents the main flexural/shear crack. In order to study the sensitivity of the IC debonding failure to the location of the major flexural/shear crack, the notch is located at different locations along the shear span of the beam. In all beam specimens, a concrete wedge located at the edge of the notch detached with the FRP debonding failure. This phenomenon shows that the initiation of debonding is due to a diagonal crack formation close to the major flexural/shear crack inside the concrete. Numerical analyses are performed using the experimentally-obtained bond-slip relations to model the shear pullout and beam tests. The application of concrete damaged plasticity model in XFEM is proposed to model the constitutive behavior of concrete. The numerical analyses show that the boundary conditions of the concrete block at the loaded end play an important role in the resulting bond-slip relationship and stress state of the FRP/concrete interface in the pullout tests. According to the numerical analyses, the diagonal crack formation observed in the experiments is due to a local moment at the tip of the notch. This causes the difference in behaviors of beam specimens than pullout specimens. The variation of localized FRP stiffness and concrete strength is combined into a single parameter as the variation of the interfacial fracture energy. A systematic method using the concept of Brownian motion is successfully established to determine the range of the interfacial fracture energies and load carrying capacities of the FRP-bonded concrete specimens.