Strengthening of Scaled Steel-Concrete Composite Girders and Steel Monopole Towers with CFRP

Cost-effective rehabilitation and/or strengthening of steel structures currently demanded by the telecommunications industry and transportation departments. Rehabilitation is often required due to crosssection losses resulting from corrosion damage and strengthening may be required due to changes in the use of a structure. Current strengthening techniques, have several disadvantages including their cost, poor fatigue performance and the need for ongoing maintenance due to continued corrosion attack. The current research program makes use of new high modulus types of carbon fiber for strengthening steel structures. The research program, currently in progress, includes phases to determine the appropriate resin and adhesive for wet lay-up of carbon fiber reinforced polymer (CFRP) sheets and bonding of CFRP strips, respectively. Test results of three scaled monopoles showed significant stiffness increases prior to yield. A significant stiffness as well as ultimate strength increase was found for the first steel-concrete composite girder tested in the program. Surface preparation of the steel must be undertaken to enhance the formation of chemical bonds between the adherend surface and the adhesive. This requires a chemically active surface that is free from contaminants. The most effective means of achieving this is by grit blasting (Hollaway and Cadei, 2002). For the CFRP strip, it is usually desirable that the strip would be fabricated with a peel-ply on one or both surfaces. However, for the small amount of CFRP produced for this program it was not economical to manufacture the CFRP strips with peel plys. As such, the procedure recommended by Hollaway and Cadei (2002) was followed, whereby the strips were abraded on the side to be bonded with sandpaper and cleaned with a solvent, which was methanol in this study. 1.2.3 Previous Work Previous work has shown the effectiveness of the technique in improving the ultimate strength of steel-concrete composite girders, although little enhancement to the stiffness has been shown. Sen et al. (2001) strengthened steel-concrete composite girders that were initially loaded past the yield strength of the tension flange. Ultimate strength increases were possible, however stiffness increases were small particularly for the thinner of the two types of CFRP laminate strips studied. It was noted that even for these specimens, the increase in the elastic region of the strengthened members might allow service load increases. Tavakkolizadeh and Saadatmanesh (2003a) also noted considerable ultimate strength increases and insignificant elastic stiffness increases. Potential to increase the elastic stiffness increase by strengthening with many plys of CFRP strips was discounted, since as the number of plys increase, the efficiency for utilizing the CFRP decreased. However, for girders that simulated corrosion damage with notches of the tension flange, Tavakkolizadeh and Saadatmanesh (2003c) found that elastic stiffness increases were possible. Vatovec et al. (2002) performed tests on square tubular steel sections that were 152 mm in depth with a span of 3048 mm. After some early trials, the tubes were filled with concrete to prevent premature local buckling of the tubes. The reported increases of strength varied from 6 to 26 percent depending on the configuration and number of plys used. No meaningful difference in stiffnesses between the unstrengthened tubes and strengthened tubes could be found and it was claimed that strengthened steel elements could not develop the full ultimate tensile or compressive strength of the CFRP due to premature delamination. Current techniques for strengthening and rehabilitation of steel structures often require bolting or welding steel plates to the existing structure. Welding is often not desirable due to the poor fatigue performance of welded connections. In contrast, the fatigue performance of repairs made to cracked steel cross-girders by bonding with CFRP has been shown to be effective up to 20 million cycles (Bassetti et al. 2000). For notched tensile specimens subjected to fatigue loading, Gillespie et al. (1997) has shown that CFRP patches applied across the notch have the effect of reducing the stress concentration at the notch, thereby substantially increasing the life of the specimen due to the slower rate of crack propagation. This finding was confirmed for notched flexural specimens subjected to fatigue loading (Tavakkolizadeh and Saadatmanesh, 2003b). The durability of CFRP materials bonded to metallic surface has to be carefully considered due to the potential for galvanic corrosion to occur if three conditions are met: an electrolyte (such as salt water) must bridge the two materials, there must be electrical connection between the materials and there must be a sustained cathodic reaction on the carbon (Francis, 2000). Brown (1974) studied the corrosion of different metals connected to CFRP by adhesive bonding or bolting. For the specimens connected by adhesive bonding there was no accelerated corrosion attack. This behavior was claimed to be due to the insulating behavior of most structural adhesives in not allowing electrical contact between the two materials. Tavakkolizadeh and Saadatmanesh (2001) provided the most comprehensive study of galvanic corrosion between steel and CFRP to date. Thicker epoxy films between the steel and CFRP surfaces were shown to significantly slow the corrosion rate of steel. The proceeding study proposed placing a layer of non-conductive GFRP as an insulating layer between the steel and CFRP interface. However, Tucker and Brown (1989) have found that glass fibers placed within a carbon fiber composite result in the blistering of the composite by creating conditions favorable for the development of a strong osmotic pressure within the composite. Clearly, water being drawn within the bond line by osmotic pressure is not favorable for maintaining a durable bond. 1.3 Carbon Fiber Material The work presented in this paper makes use of two types of carbon fiber with properties given in Table 1. The high modulus carbon fiber used, was in the form of unidirectional tow sheets or CFRP laminate strips. These sheets had a width of 330 mm and are suitable when a wet lay-up process is necessary to conform to the exact surface configuration of the structure. The same fiber was also pultruded into unidirectional CFRP laminate strips using Resolution Performance Products Epon 9310 epoxy resin with Ancamine 9360 curing agent at a fiber volume content of 55 percent. These strips were expected to be more suitable for field applications where a greater degree of strengthening is required and flat uniform surfaces are available for bonding. An intermediate modulus fiber was also pultruded using the same epoxy and to the same fiber volume fraction. The properties of the CFRP strips, as determined by the manufacturer, are provided in Table 2. Table 1. Fiber properties for two types of fiber used in the ex-