Reinforced concrete (RC) has been extensively used for construction of building and bridge infrastructures for over a century, yet the corrosion of steel reinforcement may significantly limit the long-term performance of these structures. Instead of demolition and reconstruction, retrofitting and strengthening of the existing structures to extend their service life and upgrade their load-carrying capacity to accommodate the modern demand is often more cost efficient and practical. In this realm, fiber-reinforced polymer (FRP) composite materials have shown more potential than steel plates because they are corrosion-free, higher in strength and stiffness to weight ratios, lightweight and more durable. In the recent years, using the near-surface mounted (NSM) FRP composites to strengthen the deficient structures has attracted broad research attention, as a promising alternative to the externally bonded (EB) FRP method. However, research on the characteristics of bond between NSM FRP and concrete under fatigue loading is still lacking, and a comprehensive study focusing on the bond level is necessary prior to the practical implementation in the field. This research aims to significantly advance our understandings of the bond development and failure mechanism of NSM carbon FRP (CFRP) reinforcements in concrete under fatigue loading, as well as the influence on the local bond characteristics induced by the key parameters. A total of 84 NSM CFRP-to-concrete joint specimens, with a dimension of 350 × 300 × 150 mm, were constructed and tested under a single shear direct pull-out configuration, including 24 specimens under static loading and 60 specimens under fatigue loading. Investigated variables in the experiment include: different cross‐sectional shape (rod vs. strip), surface treatment of NSM reinforcement (roughened, sand‐coated, sand‐coated and spirally wound), adhesive type (one three-component epoxy and three two-component epoxies), concrete strength (two concrete batches) and fatigue load range (10-50%, 10-60%, and 10-70% of the corresponding static load-carrying capacity Pf). The local bond degradation was observed for all the specimens under fatigue loading. In general, the bond regions closer to the loaded end of the specimen were firstly developed to resist the applied fatigue load. With the gradual local bond degradation within these regions, more bond regions were further developed along the bond line toward the free end during the fatigue cycles. This phenomenon was accompanied by a peak local bond stress migration from the loaded end to the free end of the specimen. The specimen failed when the residual load-carrying capacity was unbale to balance the applied fatigue load.
The test results also indicated that the fatigue cycles changed the failure mechanism of CFRP rod specimens from concrete and epoxy breakage under static loading to interfacial debonding between FRP and adhesive by different extents, but the CFRP strip specimens had the same interfacial debonding failure mode in both loading cases. In addition, under the same 10-60%Pf fatigue load range, specimens using CFRP strip with sand-coated and spirally wound surface and three-component epoxy had a better resistance to fatigue degradation and longer fatigue life. Furthermore, a higher fatigue load range (i.e., 10-70%Pf) significantly shorten the fatigue life of specimens, but the concrete strength did not noticeably affect the fatigue bond behavior.
An analytical model adopting the finite element method (FEM) was proposed based on a trilinear local bond stress-slip law (representing the distinctive linear elastic, softening and debonding stages), and it provided good agreement with the experimental data under both static and fatigue loading cases. The model well captured the migration of the peak local bond stress toward the free end of the NSM FRP-to-concrete bonded joints during fatigue cycles, and it visualized the bond development and failure procedure because of the local bond degradation. Additionally, the parametric study based on the analytical model confirmed the importance of the local bond strength, as well as the positive effect induced by a higher residual bond strength ratio and a higher Young’s modulus of FRP reinforcements on the fatigue life of specimen.
To supplement the experimental study, a three-dimensional (3D) finite element (FE) modeling using the concrete damaged plasticity (CDP) model in ABAQUS was adopted to simulate both NSM CFRP rod and strip joint specimens under the static direct pull-out force. The concrete breakage failure of CFRP rod specimen and interfacial debonding failure of CFRP strip specimen were successfully reproduced in the analysis. The parametric study from the numerical simulation showed that a higher dilation angle and concrete strength could increase the load-carrying capacity of the specimen, but the influence caused by groove dimension was negligible.
To sum up, this research conducts a thorough investigation on the fatigue bond characteristics of NSM FRP-to-concrete joint specimens both experimentally and analytically. Model predictions in terms of pull-out force, distribution of FRP strain and local bond stress, and relative slip between FRP reinforcement and concrete are verified by the experimental study. The proposed strategy of FE modeling also provides an efficient alternative to the static direct pull-out test to study the bond performance of NSM CFRP reinforcements in concrete.