Journal of Composites for Construction

Recent shear tests at Memorial University of Newfoundland that were conducted on beams reinforced with fiber-reinforced polymer (FRP) bars without stirrups showed that there is a relationship between the flexural cracking load and the shear capacity of the beams. Based on this relationship, a simple and unified equation for calculating the concrete contribution to the shear capacity of FRP-reinforced beams is proposed. The proposed equation accounts for the effect of shear span-to-depth ratio, axial stiffness of the longitudinal reinforcement, size of the member, and concrete strength. The prediction of the proposed equation has been evaluated by comparing the calculated shear capacity with the corresponding test results of 132 FRP-reinforced rectangular specimens without stirrups. The predictions of the shear capacity using the proposed equation are compared with those obtained using the shear equations in existing design codes and guidelines for FRP-reinforced beams in Japan and North America. It is found than the predictions of the proposed equation are more consistent than those obtained from the existing codes and guidelines.

The increased use of carbon fiber–reinforced polymer (CFRP) rods in prestressed concrete applications has been challenged by identifying a suitable anchor system. To overcome such a challenge, the design of a simplified anchor system composed of three wedges and a barrel, without a soft sleeve, is presented to duplicate the simplicity of the widely utilized anchor systems in steel strands. A numerical study and experimental verification of the simplified anchor system for a CFRP rod are presented. Three-dimensional (3D) finite-element modeling has been conducted on the anchor system consisting of isotropic steel wedges and barrel, in addition to an orthotropic CFRP rod. The wedges and barrel are modeled as elastoplastic materials of different hardness. The rod/wedge and wedge/barrel interfaces are simulated using surface-based contact models having different coefficients of friction. Two hardness levels [171 and 319 Vickers hardness numbers (VPN)] are considered for the wedges. When the softer wedges are utilized the ultimate tensile strength of the CFRP is attained. With the harder wedges the rod fails prematurely due to slippage followed by fracture. The performance of the anchor system using numerical modeling successfully describes the experimental results.

The use of fiber-reinforced polymer (FRP) reinforcement in structural engineering has attracted great attention due to high tensile strength, good fatigue performance, and inherent corrosion resistance. Engineered cementitious composite (ECC) is a class of high-performance cementitious composites with pseudo-strain-hardening behavior and excellent crack control. Substitution of concrete with ECC can avoid the cracking and durability problems associated with brittleness of concrete. In this paper, six FRP-reinforced ECC, or ECC/concrete composite beams with various longitudinal and transverse reinforcement ratios and ECC thicknesses, are tested in bending. According to the test results, FRP-reinforced ECC beams show much better flexural properties in terms of load-carrying capacity, shear resistance, ductility, and damage tolerance compared with FRP-reinforced concrete beams. For the FRP-reinforced ECC beam without stirrups, final failure occurs in shear. However, the ultimate load capacity and deformation are comparable to the FRP-reinforced concrete beams with properly designed stirrups, and the failure process is ductile due to the strain hardening behavior of ECC materials. For ECC/concrete composite beams, strategic application of ECC can lead to considerable increase of energy dissipation capacity. When a layer of ECC is placed in the tension zone, the crack width along the beam can be well controlled. High residual strength and stiffness of the composite beam can hence be obtained. In addition to experimental work, a theoretical model is proposed to predict the moment-curvature responses of FRP-reinforced ECC beams. Model results are found to be in good agreement with test data. Theoretical analysis is then conducted to illustrate the effect of reinforcement ratio, compressive strength, and thickness of ECC on the ultimate moment, curvature and ductility of beams.

This paper investigates the seismic behavior of RC shear walls strengthened by using carbon fiber-reinforced polymer (CFRP) composites. Three RC shear walls were tested: one control wall and two walls strengthened by fiber-reinforced polymer (FRP), using two different strengthening schemes. The walls were tested when subjected to a constant axial load along with synchronized cyclic moment and shear force at the top of the tested panel. The primary purpose of the FRP retrofit schemes was to increase the flexural and shear capacities of wall panels that experienced higher seismic demands than the design ones. The walls represent the sixth story panel of an eight-story RC, moderately ductile, shear wall designed according to the 2005 National Building Code of Canada. Reported shake table tests on two eight-story walls demonstrated the increase in the demand in higher story levels as a result of the effect of higher modes of vibration, which calls for increasing the capacity of the walls at a high floor level. In this study, the first wall was strengthened by using two layers of unidirectional CFRP sheet that were applied vertically and anchored to the top and bottom slabs, above which unidirectional horizontal C-shaped CFRP sheets were applied. The second wall was strengthened by applying cross-FRP bracings on the two sides of the wall. The three walls were tested under constant axial load and with increasing cycles of synchronized top moment and lateral load with a top moment-to-shear ratio of 2.75 up to failure. The strengthened walls showed satisfactory performance with improved flexural and shear strengths compared to the control wall. The two retrofitted walls showed different load and deformability capacities because of the nature of crack propagation and the orientation of the applied CFRP sheets.

Current American Concrete Institute (ACI) guidelines for the design of flexural RC members strengthened with externally bonded fiber-reinforced polymer (FRP) systems assign an additional partial strength reduction factor to the contribution of FRP, marking a deviation from ACI’s approach in building codes. This conservative method finds its rationale in the novelty and the higher variability of FRP, because of its nature as a material (compared to steel) and the conditions of its installation (externally bonded). Using the case of near-surface-mounted (NSM) FRP bars, this paper demonstrates that a single strength reduction factor can be formulated, while maintaining the same reliability and safety required in conventional RC members. Using a comprehensive test matrix of flexural members processed with a computerized Monte Carlo simulation technique, the probabilistic implications of strengthening RC beams and slabs with NSM FRP bars are investigated. The generated statistical data are employed to recommend revised strength reduction factors for flexural RC members strengthened with NSM FRP bars that eliminate the partial factor, and yet, provide a safety level equal to ordinary steel RC members.

This paper presents the evaluation of parameters describing bond action between fiber-reinforced polymer (FRP) and concrete in RC structures strengthened in flexure. The parameters were evaluated on the basis of fitting nonlinear finite-element results to experimental results from literature. Three-dimensional (3D) finite-element simulations were used to account for transversal effects without introducing any geometry-related correction coefficients in the material model. The parameters defining the material model, describing bond action, were initial stiffness, shear strength, and fracture energy of the FRP-concrete interface. Bilinear, trilinear, and exponential curve shapes were evaluated. The parameters were related to concrete tensile strength and shear stiffness of the adhesive. According to the analysis, the models provide a good estimation of ultimate load and strain distribution in FRP compared with test results. The results showed that whether the bond-slip curve was assumed to be bilinear, trilinear, or exponential has a minor effect.

Fiber-reinforced polymer (FRP) externally bonded to RC beams is commonly applied by hand layup. Vacuum-assisted resin transfer molding (VARTM), a novel application method in infrastructure, can produce FRP with more-consistent quality and uniformity. Additional steps are required, however, which can increase the application time and cost. This research will investigate the reduction of VARTM wet-out time achieved by sawing grooves into the concrete surface. Fiber-reinforced polymer U-jackets are applied by VARTM to beams with vertical grooves. The wet-out of the beams is recorded and timed. Results indicate that beams with 3.2-mm-deep grooves achieve 95% wet-out in an average of 200 s. This is only 22% of the time it takes to wet-out a beam with no grooves, an average of 911 s. A wet-out mechanism based on a modified Darcy’s law is presented to verify the results. Darcy’s law predicts that time savings increase exponentially with the length that the epoxy has to travel, which indicates that a significant amount of time could be saved on actual girders.

To delay or prevent debonding of externally bonded fiber-reinforced polymer (FRP) laminate from concrete substrate, a new carbon FRP (CFRP) anchor is developed. The monolithically built anchor comprises two legs and a wide head plate, with the legs inserted into adhesive-filled drilled holes in the concrete and the head plate bonded to the surface of the FRP laminate and the adjacent concrete. Full-scale reinforced concrete T-beams were strengthened in flexure with different amounts of CFRP laminate and tested in four-point bending. The anchor number and spacing were varied to find the most effective configuration. It was determined that the configuration involving nearly uniform spacing of the anchors along the laminate and the placement of the laminate strips between the anchor legs were most effective to allow beams with up to eight plies of the laminate to fail by rupture of FRP rather than debonding. When comparing the efficiency factor of the proposed anchor with those of other anchors reported in the literature, it was found to have the highest efficiency factor.

This paper presents the principle and the application of an innovative anchorage technique for prestressed carbon fiber–reinforced polymer (CFRP) strips in structural strengthening. Additionally, large-scale static loading tests of retrofitted concrete beams are shown. The gradient anchorage, based on the adhesive’s ability to undergo accelerated curing at high temperatures, consists of a purely concrete-adhesive strip connection without any mechanical devices, such as bolts or plates. In a first step, this study summarizes anchorage techniques presented in the literature and introduces the basic principles of the new method as well as the necessary components. In a second step, an application on a full-scale RC beam is explained in detail. A commercially-available CFRP strip is prestressed up to 0.6% prestrain and subsequently anchored by sequential epoxy-curing and force-releasing steps at both strip ends. Furthermore, uniaxial tensile tests on the epoxy adhesive and the CFRP strip are used for material characterization and to demonstrate the reinforcing materials’ integrity after the heating process. It appeared that prestress losses during the anchoring phase are negligible. The method allows much faster installation than conventional mechanical techniques and increases durability because no permanent steel elements are necessary. The material tests indicate no damage in the reinforcing CFRP strip as well as a sufficiently fast strength development of the adhesive after accelerated curing. Static loading tests on strengthened large-scale RC beams are presented and show the efficiency of a prestressed CFRP strip with gradient anchorage as a retrofitting technique. Finally, first long-term measurements over 13 years on a prestressed strip bonded to a concrete plate revealed small prestrain losses.

A bridge was damaged when a dump truck violated the height clearance limitation on Highway 401 in Ontario, Canada. The collision caused extensive damage to the AASHTO Type-III precast/prestressed bridge girders, which led to the closure of the two-lane bridge. Crack mapping showed extensive torsion-shear cracks between the girder quarter points, horizontal crack at the flange-web junctions, and spalled concrete at point of impact. Preliminary elastic testing on the girder established that the flexural capacity of the girder had not been significantly affected. As such, flexural strengthening was not necessary. Crack patterns and severity, followed by analysis, have shown that the girder is deficient in shear capacity. Therefore, the girder was strengthened for shear throughout its entire length using carbon fiber–reinforced polymer (CFRP) sheets. This paper presents a summary of the design and detailing of the elastic behavior test conducted before repair, the girder repair methodology, and results from proof load testing of the repaired girder. It was shown that the rehabilitated girder could sustain flexural live load demand. A field application was also carried out using the same rehabilitation technique on another impact-damaged bridge in Ontario. It was viewed as a major budget-saving project compared to the girder replacement alternative, because of the speed of rehabilitation and the minor traffic disruptions.

The interfacial bond parameters are quantified and modeled in this work through experimental and analytical studies. A database comprising 628 shear tests of externally-bonded fiber reinforced polymer (FRP) joints is built for model development. A newly derived closed-form solution is used for identification of bond parameters by matching the analytical solution with test results. Factors that affect the bond parameters are identified and configurations of the models are determined through analytical reasoning. Coefficients of the models are subsequently derived by regression analyses. It is found that the FRP-to-concrete width factor, which has been modeled as a function of FRP-to-concrete width ratio only in all existing bond strength models, should also be a function of concrete strength. A new model for the width factor as a function of both the width ratio and concrete strength is developed, which leads to more accurate and rational bond strength and fracture energy models.

In this paper, a three-dimensional finite-element (FE) analysis was carried out to study the effect of new variables on predicting the ultimate shear capacity of reinforced concrete (RC) beams strengthened with fiber-reinforced polymer (FRP) sheets. 55 specimens were analyzed by considering the effect of beam width, concrete strength, shear span-to-depth ratio, FRP thickness, and strengthening configuration (completely wrapped, U-jacketing, and side bonding). Experimental results of 274 beams collected from previous published work were analyzed to verify the accuracy of the proposed model. The results show that lateral strain along the top and the bottom of beams are affected by all these variables. This was not considered in previous studies. The results also indicate that the suggested model can calculate the shear capacity of RC beams strengthened with FRP sheets with higher accuracy than existing models, with coefficients of variation reaching 18.9% for side bonding, 17.0% for U-jacketing, and 18.3% for completely wrapped, respectively.

Pipelines extend thousands of kilometers for transport and distribution of oil, gas, and other chemical products. With the ever persistent challenges often faced with corrosion, relative rigidity, and other issues characteristic to steel pipes, the need to explore the use of new pipeline materials, such as glass fiber–reinforced polymers (GFRP), increases. The pipe-soil interaction and the longitudinal behavior of such pipes resulting from relative ground movements is poorly understood. In this study, a series of GFRP pipe bending experiments have been conducted on 115-mm-diameter and 1,830-mm-long GFRP pipes buried in dense sand. The pipe ends were pulled by two parallel cables attached to a spreader beam outside the test region, which was pulled by a hydraulic actuator. The study investigated the effect of laminate structure of pipe, including a cross-ply and angle-ply laminates, on the strength, deflections and failure modes, at different burial depth-to-diameter (H/D) ratios of 3, 5, and 7. Results were also compared with steel control pipes of comparable dimensions and pressure rating. The peak load was shown to increase as burial depth increases, and was generally associated with soil failure, except for the angle-ply pipe at $\mathrm{H}/\mathrm{D}=7$ that experienced a structural failure. At peak loads, the net deflections of GFRP pipes were 4–7.5 times those of the equivalent steel pipes, with the cross-ply pipes being stiffer than angle-ply pipes.

Glass fiber–reinforced polymer (GFRP) shear connectors provide much reduced thermal bridging in insulated concrete sandwich panels compared to steel connectors. In this study, 50 specimens with dimensions of $254\times 254\times 900\mathrm{mm}$ representing segments of a precast sandwich wall comprising two concrete wythes and a concrete stud surrounded by insulation foam have been tested in a double-shear configuration. Three types of GFRP connectors produced from available sand-coated and threaded rods were tested and compared to conventional steel and polymer connectors. GFRP connector diameters varied from 6 to 13 mm, and spacing varied from 80 to 300 mm. Both circular and rectangular cross sections were examined, along with various end treatments to compare with simple straight embedment. The shear strength of GFRP connectors, including the effect of friction between concrete and foam, ranged from 60 to 112 MPa, significantly higher than polymer connectors but lower than steel connectors. As the connectors bridged a small gab of insulation between concrete wythe and stud, their shear strength was lower than manufacturer-reported values. Varying the size, spacing, cross-section shape, or end treatment of connectors had insignificant effect on their strength. The connectors failed by longitudinal delamination then transverse shear but did not pull out of the concrete wythe. Adhesion bond between concrete and insulation was quite significant and contributed about 28%, but was quite random and variable. A model was developed to predict the strength of connectors and clearly demonstrated the reduction in strength as thickness of insulation layer increases.

This paper investigates the composite action and performance of rectangular concrete-filled glass fiber–reinforced polymer (GFRP) tubes in four-point bending. The study compares the performance of three different concrete-filled beam configurations with different levels of adhesive bonding of the concrete core to the GFRP tube. The three different configurations include (1) unbonded, (2) bonding of the flanges, and (3) bonding of the flanges and webs. The results show low composite action for the unbonded beam configuration up to full composite action for the completely bonded beam configuration. An analytical study is presented to provide insight into the failure mechanism for the fully bonded beam configuration and to predict ultimate capacity. For evaluation, the stiffness and strength of the fully bonded concrete-filled GFRP tube considered in this study is compared with a conventional steel reinforced concrete beam, concrete-filled steel tube, and concrete-filled GFRP tube with carbon FRP (CFRP) tension flange of the same dimensions.

This study looks into the mechanical behavior of concrete lightly confined by fiber-reinforced polymer (FRP) sheets made with glass fibers and by polypropylene fiber ropes (PPFRs). PPFRs have ultrahigh deformation at failure. The hybrid confining technique suggests applying the fiber rope on already cured FRP jackets as external unbonded reinforcement. No impregnation or gluing resins are necessary for the fiber rope (FR) application. The technique uses mechanical anchorage for the ends of the FR. Sixteen confined standard concrete cylinders, in two series and with different concrete qualities, were tested under repeated axial compression cycles of increasing displacement. They included confinement of one layer of glass FRP and FR confinement in different volumetric ratios. The investigation also compares columns confined only by glass fiber–reinforced polymer (GFRP) jackets or only by PPFR wrapping. The elaboration evaluates the axial stress versus axial and lateral strain behavior of the columns. Proper design of the hybrid confinement utilizes further the confining effects of the FRP sheet up to its multiple fracture. Then, FR ensures increased axial strain (higher than 5%) and dissipated energy of concrete. The load presents a temporary yet controlled loss, followed by an increase for further loading.

This paper proposes the equivalent strip width for stiffness-driven design of fiber-reinforced polymer (FRP) superstructures in slab-type bridges. To calculate the deflections for design, this paper first presents a tractable and accurate analytical solution for FRP superstructures under certain symmetry in accordance with classical laminate plate theory. The equivalent strip width can then be obtained by equating beam flexural deflections to the analytical solution. Through the application of the equivalent strip width, two-dimensional FRP superstructures can be simplified as one-dimensional Timoshenko beams. In accordance with the parametric studies from finite-element analysis, Timoshenko beam approximation with the equivalent strip width can predict the deflections at points of interest in practical design with sufficient accuracy. At the design stage of FRP superstructures, several material and section properties required by typical plate analysis remain unknown to the designers. This paper shows that the simplified equivalent strip width with proper assumptions about a few key parameters can avoid this practical inconvenience and facilitate design. A procedure for the design of FRP superstructures is proposed and a design example is shown.