Journal of Composites for Construction

This study investigates the performance of new bridge columns wrapped with fiber reinforced polymers (FRP) when exposed to aggressive environmental conditions. This has been accomplished through field monitoring and laboratory tests. As part of the field monitoring, temperature data were collected at various locations of bridge columns. In addition, visual inspection of two bridges was performed periodically for over a period of two years. No evidence of deterioration of the FRP wraps was detected during that period. Laboratory tests were performed to investigate how FRP wraps protect reinforced concrete columns from corrosion, and freeze–thaw laboratory tests were conducted to study the impact of temperature cycles on the mechanical behavior of FRP-wrapped columns. From the corrosion experimental tests, it was found that FRP provides excellent protection against aggressive agents (salty water or moisture) even when a single layer is used. Compression tests were conducted on specimens subjected to freeze–thaw cycles. It was found that minor thermal cycles have no effect on the performance of FRP-wrapped concrete specimens. However, for large thermal cycles, some degradation of ductility in the axial and the hoop directions was observed.

The Wright Laboratory Air Base Survivability Section has been studying the development, application, and effects of externally applied composite reinforcing materials. The strengthened facilities would be capable of surviving an air-blast load at relatively short stand-off distances (11–15 m). An agreement was reached with the Israeli officials to conduct full-scale explosive tests in Israel using 860 kg of TNT on structures that had been reinforced externally with composite reinforcing materials. The strengthening procedure employed in this study involved two types of material: (1) an autoclave-cured, three-ply, carbon fiber-epoxy laminate; and (2) a knitted biaxial E-glass fabric. The Air Base Survivability Section applied the composite materials in Israel after the facilities had been constructed using an epoxy adhesive to bond the composite materials to the concrete substrate. This provided a simple, effective, and quick method of retrofitting an existing structure. The free-field and reflected pressures and accelerations on the walls were measured. The results of these tests were considered successful, considering the fact that the externally reinforced walls suffered high displacements, yet did not fail. The pressure and impulse data indicate that both structures would have failed catastrophically without the externally applied composite reinforcing materials.

Two concrete structures and three air-entrained concrete (AEC) masonry walls were subjected to two, high explosive detonations. The concrete structures were placed at a stand-off distance such that medium damage was expected. The stand-off distance of the AEC-masonry walls was reduced on each successive detonation until breaching occurred. The two concrete structures retrofitted with composite materials were subjected to air-blast loading at a stand-off distance of approximately 14.6 m. The structures were constructed such that each long side of the structure contained a wall retrofitted with a composite material and a wall left bare as a control. Both concrete structures exhibited less residual displacement on the walls strengthened with composite materials than the bare control walls.

Although it is well known that moisture ingress in glass fiber reinforced polymers (GFRP) enhances the phenomenon of stress corrosion cracking in the fibers, and that this reaction is likely to proceed more rapidly at the weakest sites in the glass fiber surface, a fundamental law that would permit the valid extrapolation of stress rupture curves to long service lives is yet to be developed. As a result, design guidelines for glass fiber reinforced polymers components have been developed mainly on a prescriptive rather than of a performance basis. Based on the well established knowledge on the chemical behavior of glass and, in particular, that of glass flaws, a model that combines fracture mechanics, shear lag theory, and a probability model for flaw size is developed to describe the behavior of GFRP composites. The predicted results, although limited to rather idealized situations, are very encouraging. They suggest that, with only modest assumptions about material properties, it is possible to obtain mechanisms of GFRP breakdown, which correspond with observed experimental behavior.

Transportation departments have been using aluminum overhead sign structures since the 1950s. It is well documented that cracks develop in the welds between diagonal and chord members due to fatigue stresses from wind-induced vibration of the slender members. The cracks propagate to complete failure of the members, which can cause collapse of the truss and inflict injuries. The original design of overhead sign structures did not consider fatigue as a limit state. In addition, field welding of aluminum structures for any possible repairs is prohibited. A repair method for the cracked aluminum welded connections between diagonals and chord members using glass fiber reinforced polymer composites (GFRPs) is proposed. The static load carrying capacity of the welded connection, and the cracked connection repaired with GFRP composites are established. The paper describes the surface preparation of the aluminum tubular members, and the architecture and application sequence of the GFRP composite to retrofit the connection. Experimental results are presented from static tests of welded aluminum connections, welded aluminum connections retrofitted with GFRP composites, and new aluminum connections that depend only on GFRP composite elements for their strength. The results from monotonic static tests carried out on cracked welded specimens from actual sign structures show that the retrofitted connection with GFRP reinforcement achieved 1.17 to 1.25 times the capacity of the welded aluminum connection without any visible cracks. This result, and the minimal traffic disruption anticipated in the actual field application, makes this retrofit method a good candidate for implementation.

The strengthening of reinforced concrete structures by means of externally bonded fiber reinforced polymers (FRPs) is becoming an attractive technique for upgrading existing structures. Although previous laboratory investigations have shown that the bending capacities of beams can be increased considerably with this strengthening technique, premature failure by debonding of the FRP reinforcement can often limit its effectiveness. To gain insight into debonding phenomena, various experimental and analytical investigations of the behavior of bonded FRP-to-concrete joints have been carried out. However, such studies have generally been limited to monotonic (“static”) loading conditions. In this paper, we present results from an experimental investigation of bonded FRP-to-concrete joints under cyclic loading. First, we describe the experimental setup and test parameters. Next experimental results for the effects of cyclic loading on slip at the FRP–concrete interface, crack opening, and strain profiles along the bonded FRP joint are presented and discussed. A power-law expression for the so-called $“S-N”$ curves (cyclic stress ranges versus numbers of cycles to failure) is proposed, and the parameters in this expression are determined from the experimental data. The influence of various parameters such as bond length, bond width, and cyclic bond stress levels on fatigue behavior are discussed.

The civil engineering industry is constantly striving for ways to improve design and construction technologies to obtain more efficient solutions to engineering problems. It has long been recognized that, in the area of construction that utilizes reinforced concrete beams, the region of the beam below the neutral axis (NA) is wasteful of material. The only function that the low tensile strength concrete has in this situation is to locate the steel rebars and protect them from aggressive environments. This latter property is not completely fulfilled since the concrete will crack in the zone where there is tension in the loaded beam and will allow aggressive substances to attack the steel. It is proposed in this paper that the zone of a T cross-sectional beam below the NA is removed and replaced by an advanced polymer composite (APC) material. Two designs of a T beam were analyzed: one in which the failure mode is by buckling of the web and one by shear failure in the bond along the interface between the concrete and permanent APC shuttering. The test results have indicated that the failure mode for this type of design is buckling of the webs followed by concrete crushing and buckling of the permanent glass fiber-reinforced plastic shuttering.

This paper presents the results of an experimental investigation undertaken to evaluate the punching shear capacity of interior slab–column connections, strengthened using flexible carbon fiber-reinforced polymer (CFRP) sheets. Sixteen square $(670\times 670\mathrm{mm})$ slab–column connections with different slab thicknesses (55 and 75 mm) and reinforcement ratios (1 and 1.5%) were tested. Twelve specimens were strengthened using CFRP sheets and the remaining four specimens were kept as controls. Without strengthening, all specimens were designed to experience punching shear failure. The CFRP sheets were bonded to the tension face of the specimens in two perpendicular directions parallel to the internal ordinary steel reinforcement. The test results clearly demonstrate that using CFRP leads to significant improvements in the flexural stiffness, flexural strength, and shear capacity of beam–column connections. Depending on the content of the ordinary reinforcement, thickness of the slab, and area of CFRP sheet, the flexural strength increased between 26 and 73% and the shear capacity increased between 17 and 45%. The measured stress in the CFRP sheets at nominal strength varied between 22 and 69% of the ultimate tensile strength of the fibers. Comparison with available prediction equations showed that the punching shear capacity can be predicted with reasonable accuracy if the contribution of CFRP reinforcement to the increase in flexural strength is accounted for. On the other hand, the code design expressions were conservative in predicting the capacity observed in the tests.

The bond characteristics of four different types of carbon fiber reinforced polymer (CFRP) rebars (or tendons) with different surface deformations embedded in lightweight concrete were analyzed experimentally. In a first series of tests, local bond stress-slip data, as well as bond stress-radial deformation data, needed for interface modeling of the bond mechanics, were obtained for varying levels of confining pressure. In addition to bond stress and slip, radial stress and radial deformation were considered fundamental variables needed to provide for configuration-independent relationships. Each test specimen consisted of a CFRP rebar embedded in a 76-mm-(3 in.)-diam, 102-mm-(4 in.)-long, precracked lightweight concrete cylinder subjected to a constant level of pressure on the outer surface. Only 76 mm (3 in.) of contact were allowed between the rebar and the concrete. For each rebar type, bond stress-slip and bond stress-radial deformation relationships were obtained for four levels of confining axisymmetric radial pressure. It was found that small surface indentations were sufficient to yield bond strengths comparable to that of steel bars. It was also shown that radial pressure is an important parameter that can increase the bond strength almost threefold for the range studied. In a second series of tests, the rebars were pulled out from 152-mm-(6 in.)-diam, 610-mm-(24 in.)-long lightweight concrete specimens. These tests were conduced to provide preliminary data for development length assessment and model validation (Part II).

Computational modeling of the behavior of concrete reinforced with fiber reinforced polymer (FRP) bars requires models for the constitutive behavior of concrete and FRP and a model for their interaction, bond. This study focuses upon the application of an elastoplastic bond model to represent the behavior of the type A bar experimentally examined in Part I. By fully coupling the tangent and radial response, the model is sensitive to the stress state around the bar and can induce longitudinal cracking in the adjacent concrete. The model is calibrated and then applied to predict the splitting behavior of two types of specimens. Though the two specimens differ significantly, the estimated splitting loads fall within the experimental scatter. The model is then used to obtain preliminary design data for the needed cover thickness, development length, and transfer lengths. Computational models, when calibrated with limited experimental data, allow the behavior of several different specimens to be estimated. This approach is an effective means of obtaining preliminary design data and of planning further experimental verification.