Fiber-reinforced polymer (FRP) composites have received considerable attention for a while from researchers, structural engineers, and contractors for use in structural strengthening and repair applications around the world. High strength-to-weight ratios combined with superior environmental durability have made FRP composites a competitive alternative to the conventional strengthening methods using materials such as steel and concrete.

 

For many civil engineering repair applications, in particular, carbon fiber-reinforced polymers (CFRP) have proven to be very effective as externally-bonded reinforcement. Investigations on externally- bonded flexural strengthening started in the late 1980s. It was shown through numerous experimental and analytical studies that externally-bonded FRP composites can be applied to improve and repair structural performance criteria such as stiffness, load-carrying capacity, ductility, and even durability of various structural members including columns, beams, slabs, and walls. Flexural strengthening projects for bridge beams and slabs have been carried out in almost every state in the US.

 

As the technology is gaining ground, manufacturing companies are focusing on making the products cheaper, easier to use, and working with various organizations to make them more acceptable by building officials.

 

NRI prides itself in being a leader in pre-saturated composite systems for infrastructure use since its beginning in 1982. This passion for innovation has expanded to include structural strengthening solutions with NRI’s introduction of the first pre-saturated system in the market. Pre-saturated systems are externally applied systems which are manufactured in a controlled environment, ensuring high quality assurance. They are brought to the field in hermetically sealed pouches which when opened reacts with moisture and initiates curing of the composite. These systems offer many advantages including controlled and even fabric to resin ratio, less time in the field, and less manpower needed.

 

A recent research performed at University of Central Florida (UCF) looked at the performance of large scale concrete beams damaged prior to repair with FRP and compared strengthening capabilities between traditional epoxy FRP system and pre-saturated system. Both systems use the same type of unidirectional carbon fabric. Results show that both systems increase flexural capacity of concrete reinforced beams. In addition, data shows that while pre-saturated matrices are typically characterized by lower shear and tensile strengths, results demonstrate that the flexibility of the polyurethane matrix is advantageous in spreading the bond stresses over a larger area compared with epoxy composites, resulting in similar or better performance of this system in large scale tests.

 

EXPERIMENTAL PROCEDURE

 

Nine reinforced concrete girders were designed, constructed, and tested at a full scale at the University of Central Florida (UCF) structures lab. Standard four-point bend test were performed. The loads were applied mid-span with 12in separation to create a shear span of 45 inches. A servo-hydraulic actuation system was used to impose the load off of a steel load-frame as seen in Figure 1.

Figure 1: Beam test set-up

 At a theoretical concrete strength of 5000 psi and without any FRP retrofitting, the reinforced concrete beams were designed to allow the reinforcing steel to yield at a load of 13 kips and the concrete crushing to occur at 14 kips. The beams were designed to not fail in shear.

 

The damage loading protocol was generated based on the results from the control beam. The following presents a summary of the test procedure, which was divided into three phases.

  •  Phase 1: the control was tested under monotonic loading, providing the necessary data to obtain the damage loading protocol for the rest of the girders. The remaining 8 girders were damaged to different ductility levels shown in Table1.
  • Phase 2: The 8 girders were repaired, epoxy injected and retrofitted with the FRP composite systems: epoxy and PU based systems.
  • Phase 3: all 8 girders were tested monotonically to failure after retrofitting.

Damage Phase

The first control beam was tested without any FRP retrofitting. The load versus displacement response is shown in Figure 1. The global yielding of the girder occurred at a load of 12.4 kips at and a displacement of 1.2 inches. The concrete crushing in the compression zone occurred at a load of 16.9 kips and a displacement of 2.3 inches. The loading continued until a displacement of 4 inches was reached and one of the tension bars ruptured. Photos from the loading procedure are shown in Figure 2. The rupture to the tension bar is visible in the last picture.

Figure 2. Control girder: Load vs. Displacement

Figure 3. Control Test Photos

The damage phase loading protocol for the remaining 8 girders was generated based on the results obtained from the control beam test.

Repair Phase

The beams that were damaged up to ductility level 1 did not need any geometric restoration or compression zone restoration. Beam 8, which was damaged to ductility level 2, had minor (less than 0.5 in at the beam mid-span) residual deformation after it was unloaded. Therefore, it was not reverse- loaded to restore it geometrically. Beam 7, which was damaged up to ductility level 3, had more residual deformation after it was unloaded. It was geometrically restored by reverse-loading the beam (loading was applied while the beam was upside down).

 

The compression zones for beams 7 and 8 had to be repaired by grouting the areas where concrete spalling was significant to restore the original cross-sectional shape. A 5000 psi grout was used after chipping the cracked zones in the compression zone only.

 

The cracks in all 8 beams were epoxy-injected before FRP was applied. Photos from the epoxy- injection procedure are shown in Figure 5. After the epoxy injection was performed the beams bottom surfaces was prepared for FRP installation by grinding and cleaning the surface to expose the aggregate per the manufacturer’s specifications.

Figure 4. Epoxy-Injection of Beams

Figure 5. FRP Repair of Beams

After the repair phase, the beams were tested under monotonic loading. The results of the load versus displacement for the different systems are shown in Figure 9, Figure 10, and Figure 11. All failure modes were consistent. All the beams strengthened with the pre-saturated system debonded in the adhesive (primer) layer, while all the epoxy beams de-bonded by removing a thin layer of concrete (substrate). Photos from the damage phase are shown in Figure 12 and Figure 13.

Figure 9. Load vs. Displacement Behavior of Repaired Beams Plotted along with the Control

Figure 10. Load vs. Displacement: Epoxy CFRP Repair Systems

Figure 11. Load vs. Displacement: Pre-saturated system

Figure 12. Typical failure mode of the epoxy system

Figure 13. Typical failure mode of the pre-saturated system

Several observations can be made from the results. The most apparent is that, as with previous studies on FRP-repair concrete beams, there is a substantial increase in both the stiffness and the strength of the repair system. However, more interesting are the similarities in responses between the epoxy and pre-saturated composite systems as well as between the levels of damage. The only case where the peak load is lower than anticipated is in beam 7 which had more substantial plastic deformation and visible damage before repair.

CONCLUSIONS

The epoxy and pre-saturated FRP composite systems were equally effective in repairing the damaged beams. The stiffness for all beams was increased significantly. The ultimate strength of the repaired beams reached more than double that of the undamaged (non-FRP retrofitted) control. Even though the failure mode was different between the two systems, the global load displacement behavior of the pre-saturated and epoxy systems was very similar and comparable.

ACKNOWLEDGEMENTS

NRI would like to acknowledge the research work completed for this project by Dr. Kevin Mackie (PI) and Elie El Zghayar (GA)