The occurrence of new FRP (fibre reinforced polymer) materials of high strength and low weight is related to developing of aerospace industry in the period of 50-s and 60-s years of last century. Meanwhile, civil engineering branch met new problems with maintenance and durability of their structures. The corrosion of steel bars in RC (reinforced concrete) structures caused severe damage to these objects in last decades. With the aim to prevent the occurrence of rust in RC structures, new type of reinforcing bars – FRP bars – were developed due to their non-corrosive nature. Moreover, first FRP RC structures showed good durability, life-cycle cost and safety performances.
A direct substitution of the FRP rebars for steel bars in reinforced concrete is not possible due to different characteristics of two materials and different manufacturing technologies. Nowadays there are considerable number of design codes and guidelines that treat the design procedure of FRP RC structures. However, they are still incomplete or very conservative due to insufficient knowledge concerning certain issues. Therefore, the application of this methodology is still very limited. There are unknown issues that are particularly related to durability aspects of FRP RC structures. One of them is the fatigue behaviour. Fatigue behaviour is especially important limit state for concrete bridge decks, but also for parking garages, naval objects or any other structure subjected to cyclic loading.
As one of the most important issues in RC design is the bond between concrete and reinforcing bar, this research mainly focuses on FRP bar/concrete bond. The bond is assessed under quasi-static and fatigue loading. Experimental campaign consisted of pull-out tests that were adopted to measure the effect of three parameters on the bond mechanical features: 1) reinforcing bar type, 2) thickness of the concrete cover, and 3) concrete mechanical properties. For cyclic tests, another parameter, the maximum load in the cycle, was introduced to estimate the fatigue life under different load levels. Quasi-static tests showed that concrete compressive strength influence on bond properties is much more pronounced than that of the concrete cover. GFRP (glass FRP) bars showed comparable bond strength to steel bars, especially in case of lower concrete cover. They developed their splitting crack at higher bond stress level and created less damage during debonding process. As to fatigue tests, two loading levels were adopted: 60% and 70% of the corresponding quasi-static bond strength. Under 60% loading level, almost all configurations of specimens reached runout after one million cycles. The ‘survived’ specimens were tested quasi-statically afterwards and showed undisturbed residual bond properties. 70% loading level resulted in failure of almost all bond configurations. The only configuration that had runout at one million cycles at 70% loading level was made of strongest concrete mixture (approx. 55 MPa) and GFRP bar with concrete cover of 20 mm.
Numerical simulation of quasi-static tests was performed subsequently. GFRP bar/concrete interface was modelled using the bond damage evolution approach based on bond stiffness degradation. Good matching of numerical and experimental results was obtained. Obtained results from bond level were applied on structural level (beam) to perceive and explain better its behaviour. The advantage of using proposed bond slip model in simulation of structural behaviour is visible at ultimate (failure) load phase.
Overall, present work represents a contribution to data and knowledge base of FRP RC structures. It represents one step forward in understanding these aspects of FRP RC design that are still of limited knowledge. Such research is necessary for improving safety and service life, as well as complete standardization of this technology, which represents the aspiration of modern ways of constructing.