De Lorenzis, and A. Nanni, “Bond Between Near Surface Mounted FRP Rods and Concrete in Structural Strengthening,” ACI Structures

The use of Near Surface Mounted (NSM) Fiber Reinforced Polymer (FRP) rods is a promising technology for increasing flexural and shear strength of deficient reinforced concrete (RC) members. As this technology emerges, the structural behavior of RC elements strengthened with NSM FRP rods needs to be fully characterized, and bond is the first issue to be addressed. Bond is of primary importance since it is the means for the transfer of stress between the concrete and the FRP reinforcement in order to develop composite action. The objective of this research program was to investigate bond between NSM FRP rods and concrete. Some of the factors expected to affect bond performance were addressed, namely: bonded length, diameter of the rod, type of FRP material, surface configuration of the rod, size of the groove. Results are presented and discussed in this paper. INTRODUCTION The use of Near Surface Mounted (NSM) Fiber Reinforced Polymer (FRP) rods is a promising technology for increasing flexural and shear strength of deficient reinforced concrete (RC) and prestressed concrete (PC) members. Advantages of using NSM FRP rods with respect to externally bonded FRP laminates are the possibility of anchoring the reinforcement into adjacent RC members, and minimal installation time (Alkhrdaji et al., 1999). Furthermore, this technique becomes particularly attractive for flexural strengthening in the negative moment regions of slabs and decks, where external reinforcement would be subjected to mechanical and environmental damage and would require protective cover which could interfere with the presence of floor finishes. 1 Ph.D. Candidate, Department of Innovation Engineering, University of Lecce, Via per Monteroni, 73100 Lecce, Italy. 2 Jones Professor of Civil Engineering, University of Missouri – Rolla, 65409 Rolla, MO. De Lorenzis, and A. Nanni, “Bond Between Near Surface Mounted FRP Rods and Concrete in Structural Strengthening,” ACI Structures Journal, Vol. 99, No. 2, March-April 2002, pp. 123-133. 2 Although the use of FRP rods for this application is very recent, NSM steel rods have been used in Europe for strengthening of RC structures since the early 50's. The earliest reference that could be found in the literature dates back to 1949 (Asplund, 1949). Nowadays, FRP rods can be used in place of steel and epoxy paste can replace cement mortar. The advantage is primarily the resistance of FRP to corrosion. This property is particularly important in this case due to the position of the rods very close to the surface, which exposes them to the environmental attacks. The method used in applying the rods is described as follows. A groove is cut in the desired direction into the concrete surface. The groove is then filled half-way with epoxy paste, the FRP rod is placed in the groove and lightly pressed. This forces the paste to flow around the rod and fill completely between the rod and the sides of the groove. The groove is then filled with more paste and the surface is leveled. Very limited literature is available do date on the use of NSM FRP rods for structural strengthening. Laboratory studies and field applications are reported in Alkhrdaji et al. (1999), Crasto et al. (1999), Hogue et al. (1999), Tumialan et al. (1999), Warren (1998), Yan et al. (1999), De Lorenzis (2000). Yan et al. (1999) performed experimental tests to characterize the bond strength of NSM FRP rods. As this technology emerges, the structural behavior of RC elements strengthened with NSM FRP rods needs to be fully characterized and bond is the first issue to be addressed. Bond is of primary importance since it is the means for the transfer of stress between the concrete and the FRP reinforcement in order to develop composite action. The bond behavior has influence on the ultimate capacity of the reinforced element as well as on serviceability aspects such as crack width and crack spacing. The objective of this experimental program was to investigate bond between NSM FRP rods and concrete. Some of the factors expected to affect bond performance were addressed, namely: bonded length, diameter of the rod, type of FRP material, surface configuration of the rod, size of the groove. A series of specimens was tested to assess the influence of each of the above mentioned factors on the bond behavior. DESCRIPTION OF SPECIMENS AND TEST MATRIX A beam pull-out test was adopted for this project, based on merits of test methods and previous work on bond between CFRP sheets and concrete (Nanni et al. 1995, Miller, 1999). The specimens were unreinforced De Lorenzis, and A. Nanni, “Bond Between Near Surface Mounted FRP Rods and Concrete in Structural Strengthening,” ACI Structures Journal, Vol. 99, No. 2, March-April 2002, pp. 123-133. 3 concrete beams with an inverted T-shaped cross-section, chosen to provide a larger tension area for concrete while minimizing the overall weight of the beam. A larger tension area for concrete was needed to increase the cracking moment resistance of the specimen. Furthermore, the position of the bonded length of the rod had to be chosen appropriately in order to prevent flexural cracking before bond failure. For different applied load levels, the maximum tensile stress in the concrete cross-section was computed as a function of the distance of the bonded length from the support. This distance was then chosen such that the maximum tensile stress in the concrete, at a load level corresponding to the envisioned bond failure load, was conveniently lower than the concrete modulus of rupture. The dimensions of the beams are given in Figure 1. The specimen had a steel hinge at the top and a saw cut at the bottom, both located at mid-span. The purpose of the hinge and saw cut was to allow control of the internal force distribution. During loading of the specimen, the saw cut caused a crack to develop at the center of the beam and extend up to the hinge. Therefore, the compressive force in the beam at mid-span was located at the center of the hinge and the internal moment arm was known and constant for any given load level above the cracking load. Each beam had a NSM FRP rod applied to the tension face and oriented along the longitudinal axis of the beam. One side of the beam was the test region, with the NSM FRP rod having a limited bonded length and being unbonded in the remaining part. The rod was fully bonded on the other side of the beam, to cause bond failure to occur in the test region. The variables examined in the experimental test matrix were the following: • bonded length. Four different bonded lengths were selected, equal to 6, 12, 18 and 24 times the diameter of the rod; • diameter of the rod. Rods No. 3 and No. 4, having nominal diameter 3/8 in. (9.5 mm) and 1/2 in. (13 mm) respectively, were examined; • type of FRP material. Both CFRP and Glass FRP (GFRP) rods were used; • surface configuration of the rod. For the CFRP rods, the effect of two different surface conditions, deformed and sandblasted, was examined; De Lorenzis, and A. Nanni, “Bond Between Near Surface Mounted FRP Rods and Concrete in Structural Strengthening,” ACI Structures Journal, Vol. 99, No. 2, March-April 2002, pp. 123-133. 4 • size of the groove. For some types of rods, specimens with three different sizes of the groove were tested. The groove width was maintained equal to the groove depth in all the tested specimens, so that the term “size” is used in the following to refer to either the groove depth or width. Concrete strength and type of epoxy were not varied, although they are significant parameters. Concrete with 4000-psi (27.6 MPa) nominal compressive strength and a commercially available epoxy paste were selected as representative systems. The average concrete strength, determined according to ASTM C39-97 on three 4-in. (102mm) diameter by 8-in. (203-mm) cylinders for each concrete batch, ranged from 3880 to 4100 psi (26.7 to 28.2 MPa). The mechanical properties of the epoxy paste, as specified by the manufacturer, were: 2000 psi (13.8 MPa) tensile strength (ASTM D638), 4% elongation at break (ASTM D638), 8000 psi (55.2 MPa) compressive yield strength (ASTM D695) and 400 ksi (2757 MPa) compressive modulus (ASTM D695). Tensile strength and modulus of elasticity of the CFRP deformed rods were determined from laboratory testing. The average values resulted to be 272 ksi (1875 MPa) and 15200 ksi (104.8 GPa) with a standard deviation of 6.9 ksi (47.6 MPa) and 700 ksi (4.8 GPa), respectively. The manufacturer specified for the GFRP No. 4 rods a tensile strength of 116 ksi (799 MPa) and a Young’s modulus of 6000 ksi (41.3 GPa). Properties of the CFRP No. 3 sandblasted rods were: 225 ksi (1550 MPa) tensile strength and 23900 ksi (164.7 GPa) Young’s modulus. Table 1 summarizes all specimens that were tested together with the designation used to identify them. Six specimens with GFRP No. 4 deformed rods were tested. The first series consisted of three specimens (G4D6a, G4D12a and G4D18a) with the same groove size (5/8 in., 16 mm) and three different bonded lengths (6, 12 and 18 db). The value of the groove size was chosen as the smallest possible from a practical standpoint. Subsequently, two more specimens were tested (G4D12b and G4D12c), characterized by a bonded length of 12 db and two different values of the groove size, namely, 3/4 in. (19 mm) and 1 in. (25 mm). After testing of these two specimens, a groove size of 1 in. (25 mm) was identified as the recommendable size for embedment of a No. 4 GFRP deformed rod. Therefore, an additional specimen was tested (G4D24c), with 1-in. (25-mm) groove and a bonded length of 24 db. Longer bonded lengths are more capable to represent the non-uniform interface conditions and to make negligible the unavoidable end effects. The same test matrix was followed for the twelve specimens with CFRP No. 3 deformed and sanblasted rods. The groove size reflected the use of a smaller diameter rod. De Lorenzis, and A. Nanni, “Bond Between Near Surface Mounted FRP Rods and Concrete in Structural Strengthening,” ACI Structures Journal, Vol. 99, No. 2, March-April 2002, pp. 123-133. 5 Four specimens with CFRP No. 4 sandblasted rods were tested, with a groove size of 3/4 in. (19 mm) and four different bonded lengths (6, 12, 18 and 24 db). No different groove sizes were investigated, since testing of the specimens with No. 3 sandblasted rods had shown no influence of the groove size on ultimate load and failure mode. PROCEDURE The epoxy paste was allowed to cure for at least 15 days (full cure time at room temperature) prior to testing of the beams. The beams were loaded under four-point bending with a shear span of 19 in. (483 mm). Each beam was instrumented with two LVDTs, one placed at mid-span to measure deflection of the beam and the other one used to monitor the slip of the rod at the end of the test region. Three or four strain gages were applied on the surface of the FRP rods within the bonded length and an additional one was applied in the unbonded region. No strain gages were applied to the specimens with a bonded length of six diameters. Testing was performed by loading the beam in displacement controlled mode until failure. RESULTS General Results The test results in terms of ultimate pull-out load, average bond strength and failure mode are summarized in Table 1. The expression “pull-out load” has been adopted to refer to the tensile load directly applied to the NSM rod after cracking of the beam, which could be computed with accuracy from the value of the external applied load as a result of the specimen configuration. The ultimate pull-out load has been also expressed as a percentage of the ultimate tensile load of the FRP rod, to give an idea of how efficiently the rods can be used when bond is the controlling factor. The failure mode is indicated for all the specimens in the last column. A detailed description of the failure modes is reported in the next section. The load vs. mid-span deflection curves were used to check the overall behavior of the specimen during loading. Three types of load vs. mid-span deflection behavior were observed, depending on the failure mode. For the specimens failed by splitting of the epoxy cover, the load vs. mid-span deflection curve appeared very similar to that shown in Figure 2-curve a, relative to specimen G4D12a. The instant of cracking is clearly visible at a load level of 2930 lbs (13.0 kN), after which the applied load starts increasing again until failure occurs in a sudden De Lorenzis, and A. Nanni, “Bond Between Near Surface Mounted FRP Rods and Concrete in Structural Strengthening,” ACI Structures Journal, Vol. 99, No. 2, March-April 2002, pp. 123-133. 6 fashion. When failure occurred at the interface between epoxy paste and FRP rod (pull-out), the load dropped to a value smaller than the cracking load and remained constant while deflection kept on increasing until the test was stopped (Figure 2-curve b). This behavior is due to the presence of friction between the rod and the surrounding paste, which keeps on resisting a certain amount of load even after the other bond-resisting mechanisms are lost. This phenomenon is clearly absent when splitting leads to the total or partial loss of the epoxy cover. Finally, specimen C3D12c experienced failure by cracking of the concrete surrounding the groove, that led to gradual dropping of the load, as shown in Figure 2curve c. Average bond stress vs. free-end slip diagrams were also plotted. Those of the specimens with GFRP rods are illustrated in Figure 3. For this type of rebar, an 18-diameter bonded length was enough to prevent the free end from slipping before bond failure. In the case of CFRP sandblasted rods, free-end slip was recorded prior to failure when the bonded length was 18 db, while no slip occurred in the specimen with 24-db bonded length. In the case of CFRP deformed rods, the rod's free end started slipping before failure even in the specimen with the longest bonded length. Failure Modes Specimens with Deformed Rods For all the specimens with deformed rods, with the only exception of specimen C3D12c, failure occurred by splitting of the epoxy paste in which the NSM rods were embedded, accompanied or not by cracking of the concrete surrounding the groove. During testing, a crackling noise revealed the progressive cracking of the epoxy paste, until the epoxy cover was completely split and the load suddenly dropped. Although this overall behavior was common to all the specimens, some differences could be observed as bonded length and groove size were changed. In specimens G4D6a, G4D12a, C3D6a and C3D12a, having the minimum groove size, failure occurred by splitting of the epoxy paste which disintegrated in very small pieces, while no damage was visible in the surrounding concrete and the deformations on the bar surface were also intact. Very similar was the aspect of specimens G4D18a and C3D18a, having the minimum groove size and a bonded length of 18 bar diameters. In this case, some damage in the concrete surrounding the rod was visible and, for GFRP rods, In the case of GFRP, the De Lorenzis, and A. Nanni, “Bond Between Near Surface Mounted FRP Rods and Concrete in Structural Strengthening,” ACI Structures Journal, Vol. 99, No. 2, March-April 2002, pp. 123-133. 7 deformations on the rod’s surface in the region close to the loaded-end had been partially sheared off. In specimens G4D12b, G4D12c and C3D12b, due to the larger groove depth, a higher level of damage was observed in the concrete at the sides of the groove (Figure 4). The epoxy paste, rather than disintegrating in small pieces, developed longitudinal and inclined cracks that propagated in the concrete. Also, since specimen G4D12c was able to sustain a load slightly higher than G4D18a, shearing-off of the rod’s lugs occurred also in specimen G4D12c. Specimen C3D12c experienced a different mode of failure. Inclined cracks propagated in the concrete surface at one side of the groove and led to a gradual dropping of the load. Visual inspection after failure revealed also the presence of inclined cracks in the epoxy close to the rod’s loaded end. This specimen had the largest value of the ratio cover thickness to rod diameter among all the specimens with deformed rods. The cover was thick enough to offer a higher resistance to splitting, so that the controlling failure mechanism shifted to cracking of the surrounding concrete. Finally, in specimens G4D24c and C3D24b, inclined and longitudinal splitting cracks formed in the epoxy paste and the inclined cracks propagated in the concrete surrounding the groove (Figure 5). The rod’s deformations were extensively sheared-off in specimen G4D24c and damaged in localized areas in specimen C3D24b. The failure mode by splitting of the epoxy cover is similar in its mechanics to splitting of the concrete cover for reinforcing rods embedded in concrete. The radial components of the bond stresses are balanced by circumferential tensile stresses in the epoxy cover which may lead to the formation of longitudinal splitting cracks (Tepfers, 1979). The load at which splitting failure develops is influenced by the surface characteristics of the rods, the tensile strength of the cover material and the thickness of the cover, which in turn depends on the depth of the groove in which the rods are embedded. Epoxy has typically a much higher tensile strength than the concrete, however, the cover thickness of NSM reinforcement is very low compared to that of reinforcing bars in concrete, which makes this mode of failure critical for NSM reinforcement. The concrete material surrounding the groove is also subjected to tensile stresses along inclined planes, and the tensile strength of the material may be eventually reached causing fracture along these planes. Whether failure in the concrete occurs before or after the formation of splitting cracks in the epoxy or even the complete fracture of the epoxy cover, depends on the groove size and the tensile strength of the two materials. De Lorenzis, and A. Nanni, “Bond Between Near Surface Mounted FRP Rods and Concrete in Structural Strengthening,” ACI Structures Journal, Vol. 99, No. 2, March-April 2002, pp. 123-133. 8 Internal cracks in the epoxy paste were observed in all the specimens after failure. These “secondary” cracks indicate the trajectories along which the compressive forces leave the surface of the bars and spread out into the surrounding material. Unlike in the case of reinforcing bars embedded in concrete, in NSM reinforcement these inclined cracks may form not only as secondary cracks in the internal surface of the epoxy paste, but also, due to the small cover thickness, as primary cracks that can lead to bond failure. Specimens with Sandblasted Rods Different failure modes were observed in the specimens with CFRP sandblasted rods. Specimens C3S6a and C4S6a failed by splitting of the epoxy cover. In specimen C3S6a, the epoxy paste disintegrated in small fragments, while the cover of specimen C4S6a broke up in two pieces along a longitudinal splitting crack. Inclined cracks in the epoxy underneath the rod and inclined white lines on the rod surface were visible tracks of the bond stresses. Specimens C3S12a, b and c all failed at the interface between epoxy and CFRP rod. This failure mode has been referred to as “pull-out” in Table 1. The degree of micro-deformation on the surface was not enough to provide mechanical interlocking and bond was primarily guaranteed by chemical adhesion and friction after onset of slip. As a result, the rod was pulled out as soon as adhesion was lost. No sign of damage was visible in the test side of these specimens after failure. All the other specimens experienced a mixed failure mode between the previous two. After failure, either the epoxy cover appeared partially damaged (C3S18a) or a longitudinal splitting crack has developed close to the rod’s loaded end (C4S12a). Effect of the Tested Variables Among the two different rod surface conditions examined in this experimental study, deformed and sandblasted, the former appeared to have a greater tendency to induce splitting failure, as expected. However, comparing the average bond strength of specimens with CFRP deformed and sandblasted rods having the same values of all the remaining parameters, it can be concluded that deformed rods are more efficient than sandblasted rods from the standpoint of the bond performance. De Lorenzis, and A. Nanni, “Bond Between Near Surface Mounted FRP Rods and Concrete in Structural Strengthening,” ACI Structures Journal, Vol. 99, No. 2, March-April 2002, pp. 123-133. 9 When failure was by splitting of the epoxy cover, increasing the groove size led to a higher bond strength. For specimens G4D12, the ultimate load increased 8% and 24% as the groove size increased from 5/8 in. (16 mm) to 3/4 in. (19 mm) and 1 in. (25 mm), respectively. For the specimens C3D12, the ultimate load increased 15% and 8% as the groove size increased from 0.5 in. (13 mm) to 3/4 in. (19 mm) and 1 in. (25 mm), respectively. The smaller increase in the second case corresponded to a different mode of failure, as previously discussed. As the groove size increases, the thickness of the epoxy cover increases, so offering a higher resistance to splitting. The ultimate load increases correspondingly, and failure may eventually shift from the epoxy to the surrounding concrete. For specimens C3S12, increasing the groove size did not influence the failure load, since pull-out was the controlling mechanism. However, in specimens with longer bonded lengths splitting cracks developed and accelerated pull-out failure. Therefore, it is expected that increasing the groove size would have been beneficial also for specimens with sandblasted rods. The ultimate load increased, as expected, with the bonded length of the rod. For the specimens with GFRP No. 4 deformed and CFRP sandblasted rods, the average bond strength was found to decrease as the bonded length increased. On the contrary, for specimens with CFRP No. 3 deformed rods, the average bond strength resulted approximately constant with the bonded length, which indicates an even distribution of bond stresses along the bonded length at ultimate. ANALYSIS OF RESULTS Elastic Modulus of the Rods After the mid-span cross-section had cracked, the specimen configuration allowed to compute the tensile load in the rod from the externally applied load using equilibrium. On the other hand, strain at the rod loaded end was monitored by an electric strain gage. Therefore, at each load level after cracking, the elastic modulus of the rod was calculated dividing the tensile load by the loaded-end strain and by the nominal cross-sectional area of the rod. As the crack reached the top hinge, the value of the elastic modulus computed as described stabilized. The average of all values in the stable range (i.e., variation less than 5%) is used as the “reference” elastic modulus. Once the “reference” elastic modulus had been computed, the tensile load in the rod became known at every load level, also De Lorenzis, and A. Nanni, “Bond Between Near Surface Mounted FRP Rods and Concrete in Structural Strengthening,” ACI Structures Journal, Vol. 99, No. 2, March-April 2002, pp. 123-133. 10 before cracking, from the values of elastic modulus and loaded-end strain. For the purpose of data analysis, it was decided to use, for each specimen, the “reference” elastic modulus pertaining to that specimen and not the grand average, so that, after stabilized cracking, the values of tensile load calculated by equilibrium from the external load coincided with those computed from the strain readings. The elastic modulus computed for the specimens in which the strain gage in the unbonded region was properly functioning up to failure is reported in Table 2, where it is compared with the value known by manufacturer specifications or experimental testing. It can be seen that reasonable agreement exists between the average of the two sets. Analysis of Strain Data The data from the strain gages was used to plot strain vs. location graphs. In these graphs, the strain in the rod along the bonded length is plotted for different values of the pull-out load. All points were obtained from the readings of the strain gages, except for the strain at the end of the bonded length, which was assumed to be equal to zero. From the strain-location data many useful information can be drawn. Equilibrium of a piece of rod of length dx (Figure 6), along with the assumption of linearly elastic behavior of the rod, leads to the following: dx x d E d x b b b ) ( 4 ) ( ε τ ⋅ ⋅ = (1) where db, Eb, εb are diameter, Young’s modulus and strain of the rod, respectively, while τ is the bond stress and x is the coordinate along the bonded length. Given that strain measurements are available at discrete points along the bonded length, and indicating with εbi the strain reading at the location expressed by the coordinate xi, (1) may be approximated as follows: i j bi bj b b j i x x E d x x − − ⋅ ⋅ = + ε ε τ 4 ) 2 ( (2) De Lorenzis, and A. Nanni, “Bond Between Near Surface Mounted FRP Rods and Concrete in Structural Strengthening,” ACI Structures Journal, Vol. 99, No. 2, March-April 2002, pp. 123-133. 11 Recalling the definition of slip as the relative displacement between reinforcement and parent material, and recalling also that: dx dub b = ε and dx due e = ε , (3) where ub and ue are the displacements of the FRP reinforcement and of the epoxy, respectively, it follows that: b e b dx ds ε ε ε ≅ − = (4) where the epoxy strain, εe, is assumed to be negligible when compared to the FRP strain, εb. Thus: ∫ + = x b dx x s x s 0 ) ( ) 0 ( ) ( ε (5) or, in the case of discrete strain readings,