Introduction
Eight beams of the third and fourth span of a forty-five year old concrete bridge were strengthened with the CFRP/Sprayed GFRP hybrid system that has been described in the previous chapters. The whole project consisted of three parts: installation of the CFRP/Sprayed GFRP system, load tests before and after the repair, and long-term
monitoring. In this chapter, only the first part will be detailed since it reveals more issues related to application of the hybrid system in reality, the second part is described in detail in the final project report (A.P. Michael et.al., 2003) and the third part is still an ongoing project.
Background
The University Boulevard Bridge (NBI # 724214) was built in 1957. It consists of 27 spans, with overall length of 290m and a total width of 12m (Figure 7-1). The bridge was constructed of cast-in-place reinforced concrete and is sitting over the Arlington River in Jacksonville, Florida. An inspection report(Lichtenstein, 2000) indicates that severe cracking, spalling and reinforcement corrosion exists in some beams.
Consequently, the structure was posted in order to reduce the maximum vehicle load permitted to cross the bridge (Figure 7-2).
Figure 7-1. University boulevard bridge (NBI # 724214).
Figure 7-2. Vehicle weight limit for University boulevard bridge.
Replacement of the existing structure with a new bridge is considered excessively expensive and, therefore, an economical but reliable alternative is necessary. Under these conditions the hybrid system of CFRP laminates and sprayed GFRP was proposed to strengthen the bridge in a demonstration project funded by the Innovative Bridge Research & Construction (IBRC) program from the Federal Highway Administration (FHA). The third span from the south end, the most severely damaged, was chosen for repair; in addition the adjacent fourth span was also chosen because of its proximity to the third span, which enables easier installation of the hybrid system and the monitoring afterwards. In total, eight beams (four beams per span) were strengthened. Load tests, before and after the repair, were conducted to evaluate its effectiveness.
Bridge Conditions Before Repair
Other than the inspection report (Lichtenstein, 2000), which reveals some
deficiencies of the bridge, the research group at the University of Florida responsible for the project took several trips to the bridge before the repair and checked the two spans chosen to be retrofitted. Concrete spalling and steel corrosion were observed. Some beams showed more severe damage than others. The following pictures, taken during the inspections, show several locations of serious damages.
Figure 7-3. A beam after removal of loose concrete cover.
Figure 7-4. Corroded reinforcement
Strengthening Design Load Test before Repair
A load test was carried out before the repair to observe the behavior of the two chosen spans under truck load, which helped determine the current mechanical properties of the beams to be strengthened. The final project report (Michael, et al. 2005) contains a detailed description of the load test. Since no concrete and steel samples were taken from the bridge and no original design data were available, the concrete strength was assumed to be 27 MPa and the steel yield strength to be 414 MPa. Then, a sectional analysis of the beams was conducted according to the ASSHTO code, mid-span deflections and strains were calculated and compared to the corresponding test data. The results indicated that the assumptions concerning the material properties were reasonable. Actually, the results even implied that the strength of the concrete and steel was higher to a certain amount than that assumed. Therefore, the load test revealed that the two spans were still in good condition despite the obvious concrete spalling and steel corrosion at some spots as the inspection found. This observation may be attributed to the conservativeness built into the original bridge design.
Based on the load test findings, the original goal of the project was changed from repairing the two deficient spans and increasing their strength to demonstrating the effects of the hybrid system in field applications (i.e. ease of application and stiffness increase). Therefore, the repair design followed common repair configurations and previous experience. Two Sika CFRP laminates were used and 13mm thick sprayed GFRP was proposed for its feasibility based on the technician’s experience.
Materials
Table 7-1 shows the materials properties of concrete, steel, CFRP and sprayed GFRP, as used in the design.
Table 7-1. Materials properties-bridge repair CFRP Sprayed Property
Laminate GFRP Concrete Steel
Tensile Strength
(MPa) 2800 108 27a 414
Elastic Modulus
(GPa) 165 12 25 200
Elongation at Break
(%) 1.69 1.32 0.3b 2c
Thickness (mm) 1.2 13 N/A N/A
Width (mm) 100 N/A N/A N/A
Notes: a – Compressive strength b – Tensile strain
c – Yield strain
The CFRP Laminates and sprayed GFRP are the same as those described in Chapter 2.
The values of the concrete and steel are chosen for the reasons discussed above.
Flexural Strength
The hybrid system was applied to eight beams of two spans. Figure 7-5 shows a section view of one strengthened span, where the outside face of the two exterior beams was not sprayed with GFRP due to environmental concerns that will be discussed later.
Figure 7-5. Cross section of strengthened bridge.
The design moment capacity of the strengthened beams can be calculated using the same procedure outlined in the previous chapter. Table 7-2 shows the results, both before and after strengthening.
Shear Strength
To calculate the shear strength of the strengthened beams, an assumption similar to that in the ACI code was used, i.e. the total shear strength is the addition of the
contributions from concrete, stirrups and sprayed GFRP. The contribution from CFRP laminates is negligible, which is comparable to the negligible effects of tensile
reinforcing steel to the shear strength of reinforced concrete.
Vn =Vc +Vs +Vsfrp (1)
where Vn is the nominal shear strength, and Vc, Vs, Vsfrp are the shear strength contributions of concrete, reinforcing steel and sprayed GFRP respectively.
From AASHTO,
Vc=0.0316β fc'bvdv (2)
s d
f
Vs Av y v(cotθ +cotα)sinα
= (3)
According to Boyd (2000), the contribution of sprayed GFRP can be considered equivalent to that of continuous stirrups. Thus,
Vsfrp =ϕsfrp2tsfrp fsfrphsfrp (4)
where tsfrp is the thickness of sprayed GFRP that is equivalent to the horizontal area of sprayed GFRP per unit length in the longitudinal direction. The term hf refers to the overall height of sprayed GFRP. The term ϕsfrpis a reduction factor applied to the shear contribution of sprayed GFRP. This factor is similar to that defined in the ACI code for shear reinforcement of FRP.
Applying a strength-reduction factor, as required by ACI318-99 and AASHTO, to the nominal shear strength gives:
φVn =φ(Vc +Vs +Vsfrp) (5)
The nominal shear strength thus obtained for a sprayed GFRP-reinforced concrete beam system should be no less than the required strength, i.e.
φVn ≥Vu (6)
The calculated nominal flexural and shear strength of the bridge girders before and after repair are compared in Table 7-2.
Table 7-2. Nominal strength of bridge girders.
Interior Beams Exterior Beams Shear Flexural Shear Flexural
(kN) (kN-mm) (kN) (kN-mm)
Before Repair 650 1917000 650 1698000
After Repair 2215 2130000 2272 1890000
Increase(%) +241.8 +11.1 +249.6 +11.3
The table shows that the hybrid system of CFRP laminates and sprayed GFRP can greatly increase the shear resistance of the beams, while the flexural resistance is only marginally improved. The potential of sprayed GFRP in shear strengthening is
outstanding, as only 13mm thick sprayed GFRP increased the shear capacity by more than 200%. Additionally, the thickness of spray GFRP can be varied due to the spraying process, resulting in a system that can be tailored to different design requirements.
Installation of CFRP-sprayed GFRP System Environmental Issues and Protection
The bridge sits on the Arlington River, which is the habitat of a number of protected or endangered species, such as manatees. The materials used for the repair, especially the resin and acetone (for cleanup), are considered hazardous chemicals.
Therefore, appropriate containment procedures were required to prevent these chemicals from contaminating the water under the bridge during the repair work. The major
hazardous material, resin, could either be spilled from a resin drum into the river or be over sprayed into the river during spray process. Thus, the containment plan needed to address these two concerns.
The actual containment scheme was composed of two parts: 1. Containment of resin drum. 2. Containment of overspray from the work platform. The picture below shows the actual setup for resin drum protection. Sitting in a truck, the resin drum is supported by a yellow plastic tray. The tray was specially manufactured to not only secures the resin drum in place but also is capable of collecting leaked resin in the case of resin spilling due to accidental impact on the resin drum.
Figure 7-6. Containment for resin drum.
Since the repair work had to be carried out underneath the bridge and there is water even at low tides, a work platform (Figure 7-7) was built. Scaffoldings were erected in the mud on both sides of the bridge, and aluminum beams spanned from one side to the other, then plywood was laid on the top of them to form a work platform.
To prevent overspray, a double containment system was used. The primary
protection was using plastic sheeting to enclose the work space on the platform (Figure 7- 8). The secondary protection was putting turbidity barrier in the water around the work platform (Figure 7-8). In case that the first barrier is breached, the second barrier can contain over sprayed resin within the restricted area and allow easy cleanup, thus
preventing hazardous chemicals from diffusing into the water system, which in turn could possibly harm the protected fauna and further contaminate the environment.
Figure 7-7. Work platform.
Figure 7-8. Turbidity barrier and plastic sheeting.
Thanks to this detailed containment system, no serious resin spilling, overspray or other chemical contamination occurred during the entire period of the repair work.
Repair Work
The repair work was carried out in July, 2003. A crew of three professors and five students from the University of Florida and one technician from the spray equipment manufacturer worked for almost two weeks to finish the installation of the hybrid system on the eight beams of two spans of the bridge. Since the bridge is situated on a key route in that area, the shutdown time was by necessity as short as possible. The bridge was closed from 8 am to 5 pm during each day for the work and open at other times.
To provide flat surfaces so quality repair could be achieved, loose concrete on some beams was removed and the beams were patched with mortar to restore the rectangular cross-section profile (Figure 7-9). Moreover, the beams were sandblasted to make surfaces rougher for better bonding between concrete and FRP.
Figure 7-9. Patched beam.
After the surface preparation was completed, installation of the hybrid system started. The installation was a two-step process: bonding CFRP laminates to the underside of the beams, and spraying GFRP onto the sides of the beams.
1. Installation of CFRP Laminates
The CFRP laminates used in the repair are the same as those used in the research presented in Chapter 2. Sika, the manufacturer, donated all of the CFRP laminates for the project and precut them to approximately 11 m, the length of the spans. Sika Dur30, a two-component epoxy adhesive, was used to bond the laminates to the concrete. To
ensure good bond strength between the laminates and the adhesive, the laminates were cleaned using rags saturated with acetone until all dust and debris on the surfaces of the laminates were removed (Figure 7-10). The adhesive was then applied to the clean and dry surfaces in constant thickness following the procedure specified by the manufacturer.
As a beam was strengthened with two laminates, each laminate close to beam edges, the underside of all the beams were first marked by straight lines so that the laminates could line up well with the beam edges. Four people held each laminate up against the beam soffit, the person at one end pressed it on while moving toward the other end. After a laminate was set in place the excessive adhesive was squeezed out, creating a flat laminate surface. Since everything was done by hand, the thickness of adhesive could only be kept approximately 5mm as the installed laminates were made as flat as possible.
Figure 7-11 shows two installed CFRP laminates on a beam.
Figure 7-10. Cleaning of CFRP laminate surface
Figure 7-11. Two installed CFRP laminates on a beam.
2. Installation of Sprayed GFRP
The spraying of GFRP was performed by an experienced technician from MVP, the manufacturer that donated the spraying equipment, resin and glass fiber. The spraying equipment and resin drums were stored in a truck (Figure 7-6) so it was convenient to move them around, to the bridge in the morning and away in the late afternoon after a day’s work was done. During working hours, the truck was parked on the bridge where the spraying was carried out under the bridge deck. An approximate 20m-long bundle of hoses was connected to the spraying equipment so that the spraying gun could reach the work platform while the spraying equipment remained in the truck.
The technician would first spray three or four passes of an approximate four-meter long spray section in an up-down wave fashion (Figure 7-12). Then two or three other people would help roll out the sprayed GFRP with metal rollers (Figure 7-13). Initial bonding between the resin and the concrete affects the number of passes that can be put
on at one time. Too many passes or too long a spray section would cause the whole sprayed mat sliding down under gravity. Given twenty or so minutes, the resin would harden, thus leading to a better bond of the resin to the concrete. After allowing the resin to cure for a certain amount of time, the whole work group came back and did another spray session. The sprayed mat was built up by doing a number of spray sessions until the desired thickness was reached. However, the thicker the sprayed mat was the longer the curing time for the resin would be so a stronger bond between the resin and the concrete and also between different sprayed layers could be formed to hold the heavier mat. The ultimate scenario is to wait until the resin hardens completely and then come back for additional sprays. But this would create a time-consuming process because complete hardening of the resin takes about 40~50 minutes in summer weather. Also, a large amount of heat released due to resin hardening from older layers to the surface would form air bubbles that try to escape through the newly sprayed layer, resulting in a
possible gap between the old and the new layer and a poor quality sprayed layer with air voids as well. 50mm fiber length was used in the spray. It was found by trial that this length produces a uniform spray layer while longer fiber usually piles up quickly after coming out of the gun, making the sprayed mat not only hard to roll out but also a hilly surface. On the other hand, short fibers make sprayed GFRP weaker as shown by
previous research (Boyd, 2000). Therefore, the 50mm fiber length became the optimized choice for spraying on the vertical concrete surfaces in the repair situation.
Figure 7-12. Spray of another layer on hardened GFRP.
Figure 7-13. Rollout of newly sprayed GFRP.
In the original design, an end portion of the underside of the beams was supposed to be sprayed so U-shaped anchorages form at both ends of the beams. Under the bridge, however, overhead spraying was found to be very difficult due to limited clearance between the beam soffit and the work platform. After several tries, it was decided to abandon the attempt and spray only the side surfaces of all the beams. Further, for the four exterior girders, only the inside faces were sprayed. This practice was to avoid spraying resin and fiber into the water because of the difficulty in providing an
appropriately contained work space on the narrow platform area projecting outside the bridge.
After the resin hardened, the excessive GFRP draping down over the edges of the beams was cut off to make a cleaner repair. At the conclusion of the spraying, all the surfaces of the sprayed GFRP were painted with an ultraviolet-proof coat. The work platform and surrounding area were cleaned up after the whole repair was done. It was noticed that there was no significant amount of resin and fiber in the water, indicating the effectiveness of the implemented containment plan.
151 CHAPTER 8