الفهرس 
Thesis layoutOnly 14 pages are availabe for public view
 |  | from | 141 |  |  |
| | | from | 141 | | |
Abstract
Strengthening with near-surface-mounted (NSM) fiber-reinforced-polymer (FRP) reinforcement has become a well-known technique, which provides a good bond between the FRP element and concrete. However, one of the most common failure modes of NSM-FRP strengthened beams is concrete cover separation (CCS). In this research, the shear flexural behavior of RC beams strengthened with fully and partially bonded NSM CFRP bars was studied. Two different bar configurations with straight and hooked ends were used. The hooked ends were used to act as end anchors, which might delay the CCS failure. The results indicated that the end anchoring was effective in delaying the CCS and increasing the ultimate carrying capacity. The ultimate load of the beams strengthened with two straight NSM CFRP bars increased by 166.2%, while that of the corresponding beam having end anchorage increased by 180.5% compared to the unstrengthened beam. On the other hand, unbonding the NSM bars along the mid-span zone slightly decreased the ultimate load compared with the fully bonded bars, however it slightly increased the beam deformability. Increasing the unbonded length shifted the failure mode from CCS to CCS and anchorage shearing off, which is not preferred from the point of view of the structural safety.
A numerical investigation using the non-linear finite element (FE) modeling was performed using ANSYS®. The developed FE model considered the nonlinear constitutive material properties of concrete, yielding of steel reinforcement, and bond slip of non-anchored NSM bars with adjacent epoxy surface. Progressive continuum damage mechanics (CDM) along with the fracture concepts were employed to simulate the damage initiation and propagation at the epoxy-concrete interface. A strain-based failure criteria was proposed to predict the CCS failure. The developed models were validated by comparing the numerical and experimental results in terms of load-deflection behavior, load-CFRP
ii
strain response, and failure modes. Overall, a good agreement was obtained with a Mean Absolute Percentage Error (MAPE) of 6.50% for the ultimate loads. The developed models were then used to study effects of extra parameters. The effect of the NSM CFRP bar length, the diameter of NSM bar, and compressive strength of concrete were evaluated.
With respect to the ultimate load, it was indicated that increasing the bar diameter had a great effect in increasing the ultimate capacity. Increasing the NSM bar length over 175 times the bar diameter did not have a significant effect neither on the ultimate load, or on the beam deformability.