Experimental Study on Eccentric Compressive Performance of Concrete Column Strengthened with CFRP Grid Reinforced ECC Matrix

Carbon-ﬁber-reinforced polymer (CFRP) grid and engineered cementitious composite (ECC) were combined in this study to strengthen concrete columns. The inﬂuences of the number of layers, the overlap length of CFRP grids, and the eccentricity on the bearing capacity and rigidity of reinforced concrete columns were determined. The results show that the principal failure of the reinforced column was debonding of external ECC from FRP grids at the compressive area, edges, or sides. Signiﬁcant enhancement in the ultimate bearing capacity and rigidity of eccentrically loaded columns was observed after they were externally reinforced by CFRP grids and ECC; such enhancement increased with the number of reinforced layers. Eccentricity made little diﬀerence to the enhancement rate of bearing capacity when the number of reinforced layers was the same. At diﬀerent eccentricities, the composite layers at the tensile area and the compressive area had diﬀerent contributions to the bearing capacity. An eﬀective bond and eﬃcient stress transfer could be ensured as long as the overlap length between CFRP grids reached 120mm.


Introduction
Fiber-reinforced plastic (FRP) has been widely used to strengthen concrete structures due to its numerous advantages, such as light weight, high strength, corrosion resistance, and durability. e most common strengthening method is external bonding of FRP sheets or plates with epoxy adhesive on the concrete members; however, there are still some issues to be addressed, such as aging and poor durability of the adhesive [1]. FRP debonding is the primary cause of failure of concrete structures strengthened by this method. Fabric-reinforced cementitious matrix (FRCM) reinforcement systems not only feature low construction requirements, better flexural-strengthening efficiency, higher cracking loads, and postcracking stiffness [2][3][4][5] but also can be bonded using inorganic materials. A variety of FRCM strengthening methods have been developed, including textile reinforced concrete/mortar (TRC/TRM) [6] and FRP grids reinforced cement mortar. When FRP grids are used to reinforce cement mortar, they often slip during the loading process due to insufficient cohesiveness between grids and cement mortar [7]. Furthermore, brittle fracture of cement mortar will occur prior to the failure of FRP grids due to its high brittleness. To increase both the reinforcement efficiency and service life of the structure, an appropriate method should be developed to enable coworking between the cementitious material and FRP grids.
Engineered cementitious composite (ECC) has good penetration resistance and carbonation resistance [8] and shows strong crack control capability and excellent ductility in the presence of an axial tension load [9]. It can provide a protective layer to reinforced concrete (RC) structures that improves their durability if ECC at a thickness of about 20-40 mm is used in structural strengthening. In addition, ECC can provide effective tensile resistance even at large tensile strain due to the bridging effect of the fibers [9,10], which enables it to work cooperatively with FRP grids. ere is high bond strength between ECC and FRP [11,12]; furthermore, ECC embeds in the rectangular holes of FRP grids and covers the grids, which effectively enhances the mechanical interlocking force between them and ensures the synergistic deformation and load-bearing of ECC and FRP grids. ere have been some studies on the flexural properties of RC beams strengthened by FRP grids and ECC composite [13][14][15][16][17][18]; however, few publications have reported the mechanical responses of RC columns strengthened by this system.
For the above reasons, CFRP grids and ECC were combined to strengthen RC columns in this study. e effectiveness of this strengthening method was examined by determining the influence of the number of layers, overlap length of CFRP grids, and eccentricity on the bearing capacity and rigidity of RC columns. On the basis of the experimental results, the enhancing mechanism of an FRP-ECC strengthened structure is discussed.

Materials.
e cement used for concrete and ECC was a P.O.42.5R ordinary Portland cement that complies with China code GB8076. Local river sand with a fineness modulus of 2.7 was used as the fine aggregate in the concrete, and locally available crushed gravel with a well-grade particle size of 5-31.5 mm was used as the coarse aggregate. e concrete was mixed using tap water, according to the mixing proportions indicated in Table 1. e designed cubic compressive strength of concrete was 35 MPa.
FRP grids were meshed through bidirectional carbon fiber bundles, as shown in Figure 1. Each grid was a square with a spacing of 20 mm. A carbon fiber bundle has a width of 2 mm and a thickness of 0.044 mm. Its physical properties are shown in Table 2.
e ECC was a composite consisting of the matrix and randomly distributed PVA fiber that served as the reinforcing material in a volume fraction of 2% [7]. e waterbinder ratio for the matrix was 0.3. e fine aggregate in the matrix is a type of local fine sand with an average size of 150 μm. A certain amount of fly ash, silica fume, and metakaolin was also added as the cementitious binder. e ultimate tensile strength and the strain corresponding to the ultimate stress were 3.99 MPa and 2.04%, respectively. e ECC material exhibited a strain-hardening characteristic, and the axial tensile stress-strain relationship can be divided into the following three stages: (1) elastic rising stage, in which the matrix did not crack; the first crack occurred at the end of the elastic stage; (2) strain-hardening stage, which was accompanied by growing strain, slightly rising stress, and steady propagation of cracks. When the first crack occurred but had not run through the specimen, the randomly distributed fiber would play a bridging role and create many microcracks around the first crack [19,20]; (3) strain softening stage, in which tensile strength dropped slowly after it peaked, and cracks expanded and concentrated, finally resulting in the fracture of the specimen.

Specimens.
Hollow RC columns were used in the test, as shown in Figure 2. e cross section, wall thickness, hollowed part, and total height of each specimen were 240 mm × 240 mm, 50 mm, 140 mm × 140 mm, and 1600 mm, respectively. Steel bars with a yield strength of 400 MPa were used in the eccentrically loaded columns. Sixteen longitudinal bars with a diameter of 10 mm were evenly distributed around the column wall, and stirrups with a diameter of 8 mm were arranged at a spacing of 150 mm. To ensure the bearing capacity of the bracket, four longitudinal bars with a diameter of 10 mm in the hollow column were curved into the bracket as longitudinal bars. In addition, three longitudinal bars with a diameter of 24 mm were placed in the bracket, stretching 150 mm into the column along the outer edge of the bracket. Stirrups were also more densely arranged, at a spacing of 80 mm, in the bracket ( Figure 2). e eccentrically loaded concrete columns were externally strengthened with CFRP grids and ECC. ree control specimens and seven strengthened specimens were prepared in the test to determine the influence of overlap length, number of CFRP grid layers, and eccentricity on the bearing capacity and rigidity of the RC structure.
e information of the specimen is summarized in Table 3.
All specimens were cast in wooden formworks. e specimens were demolded after 24 h and were wet cured in ambient conditions for one month before they were strengthened. During strengthening, the edges on the column were chamfered at a radius of 20 mm. e four sides of the column body were roughened to increase the bond strength between strengthening materials and concrete.
en, a layer of ECC with a thickness of 5 mm was applied to the cleaned surface of the column, followed by FRP grids on the surface of the ECC. A certain overlap length was reserved as per the designed parameter. Finally, another layer of ECC with the same thickness was plastered on the surface of FRP grids. e total strengthening thicknesses of specimens with two and three FRP grid layers were 15 mm and 20 mm, respectively. Specimens were wet cured for 28 days after strengthening.

Methodology.
All tests were performed using a YAW-10000F electrohydraulic servo testing machine developed by Zhejiang University. As shown in Figure 3(a), there was a spherical hinged support at the loading end of the testing machine; a cutter hinged support was placed at the lower support of the column to control the eccentricity of loading. e locations of the concrete strain gauges and dial indicators are shown in Figure 3(b). Concrete strain gauges were placed on the lateral side, and on the compressive and tensile areas of each column. Concrete strain and steel bar strain were collected using a static strain measurement system. A DJCK intelligent crack detector was used to observe the cracks throughout the test, which was carried out following the Standard for Test Method of Concrete Structures (GB50252-92). Monotonic multistage loading was applied at the beginning, with every increment not exceeding 10% of    ultimate load. When the load was approaching peak load, displacement loading was applied at a rate of 0.2 mm/min.

Results and Discussion
3.1. Failure Pattern. Unstrengthened hollow columns failed in the same way as typical RC columns under eccentric loading. Transverse cracks occurred at the tensile area and vertical cracks appeared at the compressive area, both of which expanded with the increase of the applied load. Finally, concrete crushing failure was observed at the compressive area, and the eccentrically loaded columns failed. Strengthened columns showed the same deformation trend as control columns at the beginning of loading. With the increase of load, transverse cracks occurred at the tensile area; the concrete at the compressive area expanded laterally, activating the FRP grid-ECC system to provide lateral confinement.
e FRP grid-ECC composite reinforcing layer bore increasingly higher pressure and tension as the load increased. Finally, the external ECC was stripped from the FRP grids, which caused a decline in the rigidity of the composite layer and the drop of bearing capacity of the structure.
e typical failure patterns are displayed in Figure 4. Table 4. Initial lateral rigidity was obtained using the equation of material mechanics based on the measured deflection. As shown in Table 4, the ultimate strength of the columns and their lateral rigidities improved significantly after the CFRP grid-ECC system was used to strengthen the columns. e ultimate bearing capacity and the lateral rigidity of strengthened columns increased with the increasing number of grid layers. At different eccentricities, there was no significant difference in the compressive strength enhancement rate, and the enchantment was about 27% when the columns were strengthened by two composite layers. Overlap length made little difference to the bearing capacity and rigidity of reinforced columns.

Influence of the Number of Grid Layers.
e load-lateral deflection curves of RC columns strengthened with different grid layers and ECC composite at the same eccentricity are shown in Figure 5. It can be seen that the peak load and rigidity of the columns increased with the number of grid layers. Two major factors can explain the bearing capacity enhancement: (1) tensile strain occurred at the tensile area of the column under eccentric load, causing the FRP grid-ECC composite layer to provide a tensile resistance [21]. In addition, compared with the unstrengthened column, the growing tensile force increased the section height at the compressive area. (2) Since all the columns failed as concrete crushing at the compressive area, the compressive strength of concrete is critical to the capacity of an RC column. e externally coated FRP grid-ECC composite layer can provide lateral confinement to the concrete at the compressive area. When the concrete strain at the compressive area reached a certain level, this lateral confinement would be activated [22,23]. As a result, the core concrete was subjected to a three-directional stress state, and its compressive property was significantly improved, [24,25] thereby increasing the bearing capacity of a column. e cross-sectional strain and stress distributions of strengthened RC columns are shown in Figure 6. e initial lateral rigidity of the section was impacted by the position of the neutral axis and the material property of the section. When columns were strengthened by FRP grid-ECC composite, the position of the neutral axis would be shifted. e FRP grid-ECC layer at the tensile area could be simplified as longitudinal bars. Both the increased reinforcement ratio and compressive property improvement of the confined concrete at the compressive area had an influence on the lateral rigidity of the strengthened column, which can be written as where E C ′ , T, and x c are the elasticity modulus of the confined concrete, the tensile force provided by the tensile side, and the height of the compressive area, respectively. e sectional rigidity corresponding to the elasticity moduli of the confined concrete, the steel bars at the tensile area, and the FRP grid-ECC composite layer can be determined by obtaining the height (x c ) of the compressive area. e derivative of sectional rigidity with respect to x c is    Table 4 that when the number of grid layers was the same, the sectional rigidity at the eccentricity of 140 mm was smaller than that at eccentricities of both 90 mm and 200 mm. In other words, section rigidity showed a decreasing-increasing trend with the increase of eccentricity. According to the relationship between lateral rigidity and the height of the compressive area shown above, the section rigidity at the eccentricity of 140 mm was close to the minimum value. erefore, the sectional rigidity of the reinforced column was enhanced only by about 25%.

Influence of Eccentricity.
e load-lateral deflection curves of strengthened and unstrengthened columns at different eccentricities are shown in Figure 7. For the column with small eccentricity, the ultimate bearing capacity increased by 30% to 794 kN after strengthening; the lateral deflection at the ultimate bearing capacity was 4.74 mm, and the initial lateral rigidity improved by 214.4%. By contrast, the ultimate bearing capacity of the column with large eccentricity increased by 29.3% to 340 kN after strengthening; the lateral deflection at the ultimate bearing capacity was 13.35 mm, and the initial lateral rigidity improved by 148.7%. It can thus be inferred that this strengthening method showed a significant effect on the lateral rigidity of both types of columns. e same strengthening method had a roughly equivalent effect on the bearing capacity of columns at different eccentricities.
When the column subjected to small eccentric compression reached its ultimate strength, the maximum tensile strain of ECC was 768.2 με. At this point, the FRP grid-ECC composite layer at the tensile area showed a low utilization ratio, the compressive area was high, and the lateral confinement provided by the FRP grid-ECC composite layer at the compressive area contributed greatly to bearing capacity enhancement. However, the maximum tensile strain at the tensile area reached 5881 με when the column subjected to large eccentric compression reached its ultimate strength.
e composite layer at the tensile area showed a high utilization ratio; however, the compressive area was lower, and the lateral confinement provided by the composite layer at the compressive area contributed less to the enhancement of bearing capacity. us it can be seen that although the tension provided by the FRP grid-ECC composite layer at the tensile area and the lateral confinement by the composite layer at the compressive area made different degrees of contribution to the enhancement of column bearing capacity at different eccentricities, their overall contribution is roughly the same.   Advances in Civil Engineering

Influence of Overlap Length.
e synergistic deformation of FRP grids and ECC depends mainly on the bond strength and mechanical interlocking force between them. ECC and FRP themselves have excellent bond performance. In this test, the rectangular frames of FRP grids were 20 mm × 20 mm in size. Due to the interlocking force of ECC in the rectangular holes between FRP grids and its coverage on the grids, the mechanical interlocking force between FRP and ECC was enhanced [7]. is ensured synergistic deformation between them. At different overlap lengths, the ultimate bearing capacities of eccentrically loaded columns differed less than 8%; the load-lateral deflection curve and load-maximum strain curve of the column basically coincided in the linear elastic stage (Figure 8). It can thus be concluded that the strengthening method is effective and the load can be efficiently transferred between structural materials as long as the overlap length reaches 120 mm.

Conclusions
When an eccentrically loaded column was strengthened with FRP grid-ECC composite layers, the principal failure mode of an RC column was debonding of external ECC from CFRP grids at the compressive area, sides, or edges. CFRP grids did not break and were well bonded within the inner ECC layer. Material utilization can be enhanced by properly increasing the thickness of the inner ECC layer and reducing that of the outer ECC layer. e bearing capacity and structural rigidity of an RC column increased with the increasing number of strengthened grid layers. e ultimate bearing capacity of columns with one, two, and three layers of CFRP grids was improved by 15.8%, 27.1%, and 38.9%, respectively. e lateral rigidity of a strengthened column was significantly enhanced when an FRP grid-ECC composite system was used in the reinforcement.
Eccentricity makes little difference to the enhancement of ultimate bearing capacity of the strengthened column. At different eccentricities, the tension provided by the FRP grid-ECC composite layer at the tensile area and the lateral confinement by the composite layer at the compressive layer made different degrees of contribution to the bearing capacity enhancement of columns when strengthened in the same way. Nevertheless, their overall contribution to the mechanical response was roughly the same.
ECC and FRP grids can be effectively bonded. e reinforcement is effective and the load can be efficiently transferred between structural materials as long as the overlap length reaches 120 mm.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest regarding the publication of this paper.