Behavior of RC columns strengthened with different CFRP systems under eccentric loading

Construction and Building Materials 25 (2011) 452–460

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Behavior of RC columns strengthened with different CFRP systems under eccentric loading Marc Quiertant ⇑, Jean-Luc Clement Université Paris-Est, Laboratoire Central des Ponts et Chaussées (LCPC), Paris, France

a r t i c l e

i n f o

Article history: Available online 20 September 2010 Keywords: Fiber-reinforced polymers Reinforcement Columns Concrete Eccentric load

a b s t r a c t The present study investigates the performance of eccentrically loaded columns externally strengthened with different carbon fiber-reinforced polymer (CFRP) systems. The 10 specimens were representativescale square columns made of normal-strength concrete with substandard (internal) reinforcement details that were designed to represent old building structural columns. Eight columns were upgraded by four types of commercially available systems of external reinforcement, using plates, unidirectional or bi-directional composite fabrics. It was considered necessary to get information on a wide spectrum of carbon fiber-reinforcement systems in order to provide a satisfactory set of experimental data for validating future suitable retrofitting design methods. Experimental results presented in this paper show that a significant improvement of the strength capacity, deformation capacity and ductility of columns can result of the CFRP application, but the observed gains strongly depend on the reinforcement systems. Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction Strengthening and rehabilitation of reinforced concrete (RC) structures by externally-bonded carbon fiber-reinforced polymer (CFRP) systems is now a commonly accepted and widespread technique. Typical applications deal with the strengthening of flexural members or the confinement of concrete under axial load. However, while practical execution and the majority of design problems are well documented [1–3], strengthening design of columns under flexure–compression loading is subject of ongoing research and development [4–10]. Although such loading configuration is considered by the Japan Building Disaster Prevention Association [11] and CNR-DT 200 [12], proposed design rules are not fully satisfactory since they do not clearly establish the portion of the flexural strengthening and the portion of the confinement in the behavior the strengthened element. Several investigations considering such loading have shown that external CFRP reinforcement is effective in improving a column’s capacity both in terms of strength and ductility [13,14] and of seismic resistance [15–17]. Nevertheless, it was experimentally demonstrated that the flexural deformation of the column reduces the retrofit efficiency of the fiber-reinforced polymer (FRP) jacket [18,19]. Moreover, studies conducted so far on external strengthening of concrete columns have mainly concentrated on retrofit systems ⇑ Corresponding author. Tel.: +33 140 43 53 22; fax: +33 140 43 53 43. E-mail address: [email protected] (M. Quiertant). 0950-0618/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2010.07.034

designed for confinement only without specific flexural strengthening [20–23]. For such strengthening configuration, the jacketing is achieved by saturating fiber wrap in special epoxy formulation that allows them to be easily wrapped around columns. This simple technique provides a passive confinement that has been proven to increase the compressive strength of concrete. However, the strengthening of flexural members by externally bonded FRP plates (prefabricated laminates) or fabrics to their tension face is a widespread technique (see for example Taerwe and Matthys [24]) that can be applied to columns when their load-carrying capacity must be maintained despite significant deformations [25]. Such flexural strengthening is commonly proposed by design department in charge of the drawing of the FRP retrofitting project to moderate effect of eccentric loads that may lead to a buckling moment in columns. As a combination of these two techniques, specifiers now propose to associate a flexural strengthening coupled with a confinement by wrapping. Such retrofitting is particularly adapted to structural concrete columns that are rarely perfectly axially compressed. A prior experimental investigation of Chaallal and Shahawy [13] demonstrated that the strength capacity of beam-columns improved significantly as a result of the coupled action of the longitudinal and the transverse weaves of the bi-directional composite fabric. To study this concept of coupled reinforcement, Quiertant et al. [26] have tested different types of commercially available systems combining (longitudinal) flexural reinforcement and (lateral) confinement. Various arrangements of plates, unidirectional and

453

M. Quiertant, J.-L. Clement / Construction and Building Materials 25 (2011) 452–460

bi-directional composite fabrics were investigated. Tested columns were designed to be representatives of bridge structural applications, and were eccentrically loaded up to failure. The main conclusion of this preliminary study was that the strength capacity and ductility of columns loaded eccentrically improved significantly thanks to CFRP application but a large difference in ductility was observed between reinforcement systems. As a continuity of this previous work, the experimental study described in this paper investigates the reinforcement effectiveness of the same combination of retrofitting systems (longitudinal and lateral reinforcement) applied to columns with inferior quality of concrete and insufficient ratio of transverse reinforcement which might be found in old constructions or structures affected by corrosion. In the present study, the considered loading is still the eccentric loading. As there is a lack of consensus as to which model is the most suitable for design of columns [27–29], the main target of this paper is to present a database established from the most important experimental result of this research program. This database, extended from results proposed in Quiertant et al. [26], aims at being used as templates for future validation of CFRP strengthening design methods. 2. Experimental procedure The experimental program consisted of testing 10 square columns under combined axial–flexural loading up to failure. The program comprised five groups of two identical specimens; a first group of two control columns (CC-a and CC-b) and four groups of similar columns but externally strengthened with four kinds of CFRP systems combining longitudinal and transverse reinforcement. The specimens were labeled as ESx-a and ESx-b for the two columns externally strengthened using the system labeled x. In a same group, the external reinforcement was the same for each specimen. Repeating the experiments twice was an experimental choice to increase the confidence level in the results.

3. Specimens 3.1. Details of columns The columns tested had a 200  200 mm2 square cross section and an overall height of 2500 mm. For all the specimens, a unique batch of self-compacting concrete was delivered by a local supplier. The specimens were cast in moulds with chamfered corners in order to avoid the premature fracture of the CFRP fabric due to kinking, and to enhance the confining effect of the wrap. All columns were reinforced longitudinally with four deformed rebars with 12 mm nominal diameter. The transverse reinforcement consisted of deformed rebars with 6 mm nominal diameter. For longitudinal and transverse reinforcement, steel with 500 MPa nominal tensile strength was used. The dimensions of columns and details of internal reinforcement are shown in Fig. 1. A low amount of internal reinforcement was planned to be representative of ancient building applications for which retrofitting should be necessary. 3.2. External strengthening Except for the two reference specimens, two layers of CFRP were bonded on columns. A flexural reinforcement was first achieved by a unidirectional composite (plate or sheet) bonded in the axial direction. Then each column was externally confined by transverse composite straps wrapped around the column. Such method, widely recognized, permits to exert a lateral pressure that increases strength and ductility of concrete in the axial direction [20]. The sheet was bonded as a continuous spiral on columns ES1-a and ES1-b (first strengthening method). The type 1 dry sheet was hand-laid with a winding angle (between transverse direction

70 60 252

160

2500

φ12

φ6

20

Fig. 1. Reinforcement details of tested columns (dimensions in mm).

of column and fill direction of the fabric) of approximately 20° and with an axial lap joint measuring approximately 20 mm. Analysis of the effect of winding angle is not developed here, but a study of the influence of wrap angle configuration is proposed in Parvin and Jamwal [30]. All remaining columns were wrapped with discontinuous straps with transverse lap joints measuring approximately 100 mm (Fig. 2) but without (or with no significant) axial lap joints. The value of 100 mm for the transverse lap joint is consistent with previous study concerning the transfer length of CFRP-to-concrete bonded joints (see for example [31]). Similarly to prior studies (see for example [19]), the ends of all externally strengthened columns were reinforced with one more transverse layer of FRP strap to prevent premature failure outside the test region. Structural analysis and resultant design was carried out by the authors while the strengthening of columns was accomplished by four different technical professional teams using their own procedures and products to ensure the representativeness of experimental results. It must be emphasized that the intent of the column’s strengthening design was to cover a wide range of strengthening

Flexural reinforcement

Confinement by wrapping (individual straps) Axial lap joint Transverse lap joint

Fig. 2. Principle of column’s external reinforcement (wrapping by discontinuous rings).

454

M. Quiertant, J.-L. Clement / Construction and Building Materials 25 (2011) 452–460

rates and systems. Based on this consideration, it is easy to understand that no mechanical or cost equivalence was targeted for the studied strengthening systems. The type 4 dry stretched sheets were saturated using a resin-impregnating machine before being laid to the columns. For other strengthening using dry sheet, CFRP was fabricated by the ‘‘wet lay-up” technique; that is, the dry sheets were placed on the surface of the column and then impregnated with epoxy resins. Prior to such laying of the sheets, adhesive was applied to column surfaces. The CFRP system manufacturer’s reported material properties are shown in Table 1. The type 1 woven sheet is a bi-directional fabric (70% of fibers are in wrap direction). Type 2, type 3 and type 4 sheets are unidirectional carbon fabrics. Weights of the dry carbon sheets are ranging from 610 g/m2 for the type 4 sheet to 200 g/ m2 for the type 2 sheet (respectively 225 g/m2 and 500 g/m2 for type 3 and type 1). The four types of carbon fiber fabric are field laminated using epoxy to form a carbon fiber-reinforced polymer used to externally confine the columns. In all cases, a unique epoxy formulation was used for saturant and adhesive. Type x plates and type y plates are both pultruded CFRP laminates. Widths of type x and type y plates are respectively 254 mm and 500 mm. The CFRP strengthening configuration of columns are summarized on Table 2.

3.3. Mechanical properties of concrete In order to evaluate the effectiveness of external reinforcement by a comparison with reference columns, it was necessary, first to

Table 1 Manufacturers reported CFRP system properties. Carbon fiber product

Thickness (mm)

Tensile modulus (GPa)

Tensile strength (MPa)

Type 1 woven sheet

–

Of fibers: 240– 221 Of CFRP: 105

Of fibers: 4900– 4510 Of the CFRP layer: 1400 –

Type 2 stretched sheet

Type 3 stretched sheet Type 4 stretched sheet

Type x plate Type y plate

Of one layer of CFRP: 0.43 Of the dry sheet: 0.117

Of fibers: 240

Of one layer of CFRP: 0.334 Of the dry sheet: 0.13

Of CFRP: 84 Of fibers: 230

Of the CFRP >1050 Of fibers >3500

– Of the dry sheet: 1

– Of fibers: 235

– Of fibers: 3450

Of one layer of CFRP: 1 Of a plate: 1.2 Of a plate: 1.2

Of CFRP: 62–70

Of the CFRP: 620–700 Of plate: 3000 Of plate >2800

cast a set of 10 similar RC columns, and second to check that mechanical properties of the columns concrete remained stable all through the experimental program. This material analysis was necessary due to the long duration of the experimental study as compared to the age of columns (the first column was tested at the age of 309 days, 138 days before the last column). Therefore a total of thirty standard 16  32 cm concrete cylinders were cast at different steps of the fabrication of columns. No significant difference on material properties measured by conventional tests was noticed between the samples. Moreover, experimental results demonstrated a small evolution of concrete during the period of tests. Consequently, average material properties hereafter listed are assumed to be representative of the concrete of tested columns: compressive strength = 40.1 MPa, modulus of elasticity = 27.2 GPa, Poisson’s ratio = 0.19 and tensile strength (determined by splitting tensile test) = 3.0 MPa.

4. Test setup In order to test representative full-scale columns, it was necessary to create a loading frame capable of bringing stout specimens to failure. Assumptions based on previous experience in testing confined cylinder [32] were made on the expected concrete strength when reinforced with two layers of CFRP materials. Then, the design of the testing frame was based on a global vertical load capacity of 4.4 MN applied by four annular hydraulic jacks (filled in parallel within the same servo-controlled closed loop) inserted in a closed frame made of struts and ties and fixed to the strong floor of LCPC structures laboratory (Fig. 3). Basically, a vertical load is applied by hydraulic jacks between the lower and middle plates. While the middle plate is fixed on the laboratory strong floor, a vertical displacement of the lower plate is resulting from the jacks thrust. Lower and upper plates are linked by ties. Then consecutively to the displacement of the lower plate, a displacement of the upper plate is generated, also directed downwards.

Upper plate Upper cap

Of plate: 180 Of plate >165

Column

Table 2 CFRP strengthening configuration of columns. Specimen series

Flexural reinforcement

CFRP wrapping material

Wrapping configuration

CC ES1

None Six type x plates on each side One layer of type 2 stretched sheet Two type y plates on each side One layer of type 4 stretched sheet

None One layer of type woven sheet One layer of type stretched sheet One layer of type stretched sheet One layer of type stretched sheet

None Continuous spiral Discontinuous rings Discontinuous rings Discontinuous rings

ES2 ES3 ES4

1 2 3 4

Lower cap Middle plate Jacks Lower plate Fig. 3. Test setup for combined flexure–compression.

455

M. Quiertant, J.-L. Clement / Construction and Building Materials 25 (2011) 452–460

(a)

(b)

Upper plate Ball joint Upper cap

Loading axis

Column axis

200 mm Eccentricity: 20 mm

Fig. 4. (a) Dimensions of the loading caps and (b) principle of load application device.

Steel diffusion elements (called caps, Fig. 4) were designed to receive the load applied by plates and to transmit that loading 20 mm eccentrically to the column, thus generating the combined flexure–compression load. It corresponds to a nominal deviation accounted for in French design rules for columns [33]. Moreover, this magnitude of eccentricity was chosen in order to place the entire column under compression during the first part of the test. This is a particularly inadequate loading for the externally-bonded flexural reinforcement that can lead to local debonding of composite material and/or to local axial buckling of fibers. However, when bending becomes sufficiently pronounced, one side of the column is in tension. The bearing zone of caps on plates is realized with partially spherical shapes of the cap thus forming a ball joint (Fig. 4). This kind of bearing ensures a free rotation of the column. Symmetrical bearing conditions are provided by the lower cap and upper cap. Particular care was taken of the experimental boundary conditions (design of cap, lubrication of the ball joint, connection between the column and the cap ensured by a high strength grout) and the effectiveness of the expected mechanical scheme was checked during the tests.

a

b

Side with larger compression (south)

475 mm

775 mm

Side with lower compression (north)

West side Up

Down

East side Three-element strain gauge rosette Lateral gauge Axial gauge Fig. 5. Details of the external instrumentation by strain gauges.

5. Loading program The load was increased monotonically up to 70% of the expected failure load, with a constant 1 kN/s loading rate. Then the jack displacement was used as the servo-control parameter (0.08 mm/s) that helps recording post-peak behavior, provided the failure is ductile enough. Due to an operator error, loading was the only control parameter during the test of specimen ES3-b.

extensive measurement program will be useful for calibration of future finite element modeling and definition of serviceability and ultimate limit states for design recommendations. Only the main results are presented in the following.

6. Structural monitoring

7. Test results

All specimens were instrumented using surface strain gauges both on the longitudinal and transverse direction on each face of the specimens (Fig. 5). Strain gauges were glued on concrete surface for control columns and on CFRP outer layer for other specimens. The strains on the internal steel reinforcement were also monitored using eight strain gauges bonded on longitudinal rebars and six gauges bonded on closed stirrups (Fig. 6). Steel and CFRP strains were respectively measured by 5 mm electric resistance strain gauges and 60 mm electric resistance strain gauges. The lateral deflection was recorded at seven locations (Fig. 7) by using linear variable differential transducers (LVDTs) and the axial contraction was measured by 2 LVDTs. The applied load was recorded with four load cells. Overall, 56 measurement channels helped describing the structural behavior of the columns. This

7.1. Validation of testing process and analysis of hypotheses Within the hypotheses of the strength of materials theory and considering a homogenized inertia for the RC column, it is possible to predict the axial strain at mid-height, on the surface of the sides with respectively larger and lower compression. With the same hypotheses, strains of the longitudinal reinforcement can be determined at mid-height. Compared with experimental data in the elastic part of the column behavior, results of this calculation permit to establish the validity of the loading setup (Figs. 8 and 9). In addition, the failure of most of the specimens near mid-height confirms the dominant role of elastic second order moment. The validity of the choice of magnitude of eccentricity, to obtain on one side the transition from compressive loading to tension,

456

M. Quiertant, J.-L. Clement / Construction and Building Materials 25 (2011) 452–460

6

2 10

1

3

14

5 9

4

8 12 13

7

11

Axial Load (kN)

Fig. 6. Details of the strain gauges bonded on steel.

1500 1250

350

KYOWA 50mm

1000

KYOWA 20mm

750

300

Elastic prediction

500

100 100

mid-span

Longitudinal rebar (west)

250

300

Longitudinal rebar (east)

0 0

350

1000

2000

3000

4000

5000

Microstrain Fig. 9. Load–axial strain curves on rebars near the side with larger compression (CC-b).

7.2. Failure mode

Axial Load (kN)

Fig. 7. Layout of instrumentation for lateral deflection measurement.

1250 1000

1600 1400

Axial Load (kN)

1500

Both externally strengthened columns and reference specimens (except CC-a) failed by crushing of the concrete and buckling of longitudinal rebars on the side with larger compression near the column mid-height, as designed (Fig. 11). The same mode of failure was observed for column CC-a but on section located higher (about 800 mm from the end of the column).

1200

750

1000

500

Elastic prediction

800

Longitudinal strain of concrete (gauge a)

250

Longitudinal strain of concrete (gauge b)

600

1500

400

0 0

500

1000

2000

2500

3000

3500

Microstrain

200

Fig. 8. Load–axial strain curves on side with larger compression (CC-b).

was checked. This evolution of the axial strain recorded on the side here called ‘‘Side with lower compression” (see Fig. 5) is illustrated in Fig. 10. Such transition is due to a second order moment growing with the deflection of the column.

Microstrain 0 -600

-400

-200

0

200

400

600

Fig. 10. Example of transition from compressive loading to tension (load–axial strain curves on north side of ES1-a).

M. Quiertant, J.-L. Clement / Construction and Building Materials 25 (2011) 452–460

457

Fig. 11. Typical failure of (a) control columns and (b) CFRP column.

7.3. Bearing capacity and deformation capacity

Fig. 12. Rupture of CFRP reinforcement.

For CFRP strengthened specimens, one could see the hoop fracture of fibers during the last loading stage. The failure of specimens by rupture of the CFRP jacket due to hoop tension (Fig. 12) is the most common mode of failure for FRP-confined concrete [28]. In this study, the CFRP jacket failure was always initiated at a corner of the column at the most compressed face and appeared later on the tensile face, during post-peak behavior, when flexure of the column was increased. Sound of snapping of the fibers could be heard near the failure load. The failure mode of presently reported CFRP strengthened columns is much less brittle than those described by Li and Hadi [14] for high-strength concrete columns.

A comparison of load–deflection curves between externally strengthened columns and the reference specimens is presented in Fig. 13 for each series of test (in this figure, axial contraction has negative value). Strength enhancement of columns (calculated as the ratio of the strength of externally strengthened specimens over the strength of control columns) and deformability enhancement (ratio calculated with maximum lateral deflections until loss of stability) are presented in Table 3. As a first remark, it can be emphasized that strength and deformability enhancement strongly depends on the strengthening system. Depending on the system, the strength enhancement varied from 0.98 to 1.30 (see Table 3). At present time, no definitive explanation is proposed to justify the ineffectiveness of the retrofitting system 2 (see Fig. 13 and Table 3), but it must be emphasized that the stretched sheet used in this system was the lightest (smaller amount of fibers) of all studied confining material. This result is consistent with the conclusion of Hadi [6] that suggested the need of a minimum confinement ratio to achieve significant structural gains. As described in the previous section, columns failed by crushing of the concrete and buckling of the longitudinal rebars on the side with larger compression. For externally strengthened columns, the two failure mechanisms are delayed due to the effectiveness of the confinement. In the case of pure compressive test, it was demonstrated that the confining effect that allowed the concrete in compression to achieve a higher strain [13] is still effective when the concrete core is entirely cracked [34]. At that stage, confined concrete behaves like a confined non-cohesive material. In the present study, the same phenomena locally occurred in the zone with larger compression. Then, the equilibrium of the column’s section is preserved as long as the transverse strain of the concrete core near the side with larger compression is lower than the ultimate strain of the bonded FRP strap. As a plastic hinge region was only observed at ultimate stage, it can be concluded that confined cracked zone was developed in a large height of the column located near the most compressive side, resulting in a lose of structural

458

M. Quiertant, J.-L. Clement / Construction and Building Materials 25 (2011) 452–460

1800

Axial Load (KN)

1600 1400 1200 1000

CC-a

800

CC-b

600 400

ES1-a

CC-a

CC-b

200

ES1-b

ES2-a

ES2-b

CC-a

CC-b

ES4-a

ES4-b

0

1800

Axial Load (KN)

1600 1400 1200 1000 800 CC-a

600

CC-b

400

ES3-a

200

ES3-b

0 -10

10

50 -10

30

10

Deflection (mm)

30

50

Deflection (mm)

Fig. 13. Load–deflection curves of all specimens (axial contraction is negative, lateral deflection is positive).

Table 3 Experimental results for strength and deflection. Specimen label

Load (kN)

CC-a CC-b

1150 1272

ES1-a ES1-b

1507 1649

ES2-a ES2-b

1262 1125

ES3-a ES3-b

1300 1544

ES4-a ES4-b

1482 1442

Max.

Average

Ultimate lateral deflection (mm) Max.

Ratio ESx/CCa

Average

1.00

6.43

1.30

1.00

15.98

0.98

2.48

26.57

1.17

4.13

1.21

1.00

6.45

1.49

5.99

1.39

4.86

1.12

8,90

2.06

4.68 5.03 16.27

2.53

30.11 48.78 1462

4.32

5,97 6,01

15.02 17.52 1422

Average

7.66 5.23

22.65 30.49 1194

Ratio ESx/CCa

4.25 4.40

18.40 13.56 1578

Ultimate axial contraction (mm) Max.

6.28 6.59 1211

a

Ratio ESx/CCa

7.18 10.62 39.44

6.13

Average ESx/average CC.

stiffness. For columns with an important confinement level (i.e. with high amount of CFRP confinement), this large zone of confined cracked concrete resulted in a significantly larger lateral deflection. 7.4. Deformation capacity and flexural stiffness Lateral deflection and axial contraction recorded for two loading steps are reported in Table 4. Considering results obtained for a loading of 500 kN, it can be noticed that the axial contraction is lowered by the FRP confinement whatever the confinement level is. Reported ratios are ranging from 0.73 to 0.89. This clearly indicates that the confinement produces a significant increase of the axial stiffness of the columns. This observation is also valid for

the later stage of loading reported in Table 4 (1100 kN) where calculated rates are ranging from 0.74 to 0.96. Considering results presented in Table 4, and comparing values of the flexural deflection obtained from each series for the same loading (500 kN or 1100 kN), it can then be concluded that the composite material has not contributed to significantly enhance the flexural stiffness of columns during the considered part of the loading. Moreover it is surprising that for the two loading considered, average deflection of reference columns is smaller than three of the four average deflections of strengthened columns. In design consideration, the longitudinal external strengthening is applied to reduce the curvature of the column and consecutive secondary moment, hence limiting the induced compression and the

459

M. Quiertant, J.-L. Clement / Construction and Building Materials 25 (2011) 452–460 Table 4 Deflection for loading steps of 500 and 1100 kN. Specimen

500 kN

1100 kN

Lateral deflection (mm) Average CC-a CC-b

1.21 0.65 0.93 0.73 0.88 0.81 1.68 2.12 1.90 2.27 1.46 1.86 1.71 1.74 1.72

ES1-a ES1-b ES2-a ES2-b ES3-a ES3-b ES4-a ES4-b a

Axial contraction (mm)

Ratea

Average 1.66 1.38 1.52 1.26 1.25 1.25 1.40 1.31 1.35 1.19 1.01 1.10 1.36 1.26 1.31

1.00

0.87

2.04

2.00

1.85

Ratea

1.00

0.83

0.89

0.73

0.86

Lateral deflection (mm) Average 5.11 3.24 4.18 2.94 2.99 2.97 5.59 8.24 6.92 6.50 4.93 5.72 4.84 4.97 4.91

Ratea

1.00

0.71

1.66

1.37

1.17

Axial contraction (mm) Average 3.86 3.36 3.61 2.82 2.78 2.80 3.37 3.54 3.45 3.00 2.38 2.69 3.07 3.02 3.05

Ratea

1.00

0.78

0.96

0.74

0.84

Average ESx/average CC.

The observed failure of specimens is the rupture of the CFRP jacket due to hoop tension. The maximum hoop strain in FRP recorded during test can be then considered as the most significant parameter for design consideration. The experimental values of lateral deformation for a loading near failure of columns are reported in Table 5. On the last section of this table, it is specified if the gauge was broken after the failure of the specimen. For CFRP columns it never happened, due to the location of gauges that were not on the corner of columns where the fracture of the external strengthening jacket was always initiated. Note that the lateral strain of CFRP at failure of specimens is much lower than the CFRP material ultimate tensile strain reported in Table 2. This was underlined by Yuan et al. [28].

characterized by a ratio of 1.30. Deformation capacity and ductility improvement was more distinctive than the gains in strength. It was demonstrated that externally confined column could undergo large deformation without rupture. This characteristic is important considering that an over-compressed structure could adapt to abnormal loads and before full redundancy, redistribute bending moments to other elements. Considering average lateral deflection of columns strengthened with the system 4 (ES4-a and ES4-b), a ratio of 6.13 was obtained for deformability enhancement (see Table 3). Nevertheless, the external strengthening efficiency was shown, in the condition of our study, to be strongly dependant of the chosen system. But, in view of the acceptable dispersion of results for the columns strengthened using the same system and the fact that the inspections of specimens before and after the tests have not revealed significant default, the authors conclude that the large difference observed on bearing capacity is representative of efficiency of the different systems. For the coupled strengthening methods tested, the expected strength enhancement should result from the combined action of the longitudinal and transverse strengthening. The experimental results recorded during the tests show that, when the transverse jacketing is strong enough, it increases the axial stiffness and enhances the compressive capacity of concrete through confinement action. Moreover, it is evident that the lateral pressure exerted by the straps also provides additional support against buckling of longitudinal rebars. However, contribution of flexural reinforcement to strength enhancement was not clearly established in the tested strengthening configuration. Considering that the failure of compressed concrete occurs prior to the tensile failure of FRP jacket that directly leads to the columns failure, it can be concluded that the ultimate state depends mainly on the behavior of the bonded FRP strap, notably on it ultimate strain. A unified model for confined concrete must then consider the rupture strain of CFRP straps. Experimental results presented in this paper can be used as templates for future validation of CFRP strengthening design methods.

8. Conclusions

Acknowledgments

Depending on the CFRP strengthening system (type of material and bonding process), significant increase in deformability and strength can be achieved for columns under combined flexuralcompressive loading. The maximum strength enhancement was

Experimental program has been carried out at the Structures Laboratory of the French Public Works Research Laboratory (Laboratoire Central des Ponts et Chaussées – LCPC) with the technical collaboration of Freyssinet, Vinci Construction, SIKA-France and

Table 5 Lateral strain on the side with larger compression. Specimen label

Lateral strain (microstrain)

ES1-a

4275 x

ES1-b

1531 1666 2566 3926 1902 2967 1411 1711 5840

ES2-a ES2-b ES3-a ES3-b x ES4-a ES4-b

1982 3990 3409 5660

Loading (kN)

Failure of gauge

1489 x 1649 1457 1230 1262 1125 697 1300 1269 1543 x 1482 1148 1442 1353

Yes Yes No No No No No No No No No No No No No No

concrete cracking on the tensile face. This kind of reinforcement mode was not here clearly demonstrated. 7.5. Test database for future design methods

460

M. Quiertant, J.-L. Clement / Construction and Building Materials 25 (2011) 452–460

VSL France. The research was supported by the LCPC but previously mentioned enterprises provided CFRP strengthening material. Their aid is gratefully acknowledged. References [1] ACI Committee 440.2R02. Guide for the design and construction of externally bonded systems for strengthening concrete structures. Michigan (USA): American Concrete Institute; 2002. [2] Fédération Internationale du Béton. Externally bonded FRP reinforcement for RC structures. Lausanne (Switzerland): Fib-Bulletin 14; 2001. [3] Hamelin P. AFGC: recommendations concerning repairing and strengthening of concrete structures by composite materials. In: Teng J-G, editor. Proc FRP composites in civil engineering, vol. 1; 2001. p. 165–75. [4] Tan TH, Nguyen NB. Flexural behavior of confined high-strength concrete columns. ACI Struct J 2005;102(2):198–205. [5] Hadi MNS. Behaviour of FRP wrapped normal strength concrete columns under eccentric loading. J Compos Struct 2006;72(4):503–11. [6] Hadi MNS. Comparative study of eccentrically loaded FRP wrapped columns. J Compos Struct 2006;74(2):127–35. [7] Hadi MNS. Behaviour of FRP strengthened concrete columns under eccentric compression loading. J Compos Struct 2007;77(1):92–6. [8] Hadi MNS. The behaviour of FRP wrapped HSC columns under different eccentric loads. J Compos Struct 2007;78(4):560–6. [9] Hadi MNS. Behaviour of eccentric loading of FRP confined fibre steel reinforced concrete columns. Constr Build Mater 2009;23(2):1102–8. [10] Rocca S, Galati N, Nanni A. Interaction diagram methodology for design of FRPconfined reinforced concrete columns. Constr Build Mater 2009;23(4):1508–20 [Sp Iss FRP Composites in Construction]. [11] Japan Building Disaster Prevention Association. Seismic retrofit design and construction guidelines for existing RC buildings and steel-encased RC buildings using continuous fiber reinforced materials; 1999. [12] National Research Council. Guide for the design and construction of externally bonded FRP systems for strengthening existing structures. CNR-DT 200/2004, Rome. [13] Chaallal O, Shahawy M. Performance of fiber-reinforced polymer-wrapped reinforced concrete column under combined axial–flexural loading. ACI Struct J 2000;97(4):659–68. [14] Li J, Hadi MNS. Behaviour of externally confined high-strength concrete columns under eccentric loading. Compos Struct 2003;62:145–53. [15] Colomb F, Tobbi H, Ferrier E, Hamelin P. Seismic retrofit of reinforced concrete short columns by CFRP materials. Compos Struct 2008;82(4):475–87. [16] Cheng CT, Yang JC, Yeh YK, Chen SE. Seismic performance of repaired hollowbridge piers. Constr Build Mater 2003;17(5):339–51. [17] Sause R, Harries KA, Walkup SL, Pessiki S, Ricles J. Flexural behavior of concrete columns retrofitted with carbon fiber-reinforced polymer jackets. ACI Struct J 2004;101(5):708–16. [18] Nanni A, Norris MS. FRP jacketed concrete under flexure and combined flexure–compression. Constr Build Mater 1995;9(5):273–81.

[19] Parvin A, Wang W. Behavior of FRP jacketed concrete columns under eccentric loading. ASCE J Compos Construct 2001;5(3):146–52. [20] Saadatmanesh H, Ehsani MR, Li MW. Strength and ductility of concrete columns externally reinforced with fiber composite straps. ACI Struct J 1994;91(4):434–47. [21] Demers M, Neale K. Confinement of reinforced concrete columns with fibrereinforced composite sheets – an experimental study. Can J Civ Eng 1999;26(2):226–41. [22] Rochette P, Labossiere P. Axial testing of rectangular column models confined with composites. ASCE J Compos Constuct 2000;4(3):129–36. [23] Lau KT, Zhou LM. The mechanical behaviour of composite-wrapped concrete cylinders subjected to uniaxial compression load. Compos Struct 2001;52:189–98. [24] Taerwe L, Matthys S. FRP reinforcement for concrete structures: state-of-theart. Proc IABSE Symp 1999:578–85. [25] Wu YF, Oehlers DJ, Griffith MC. Composite plated columns. In: Teng J-G, editor. Proc FRP composites in civil engineering, vol. 1; 2001. p. 767–74. [26] Quiertant M, Toutlemonde F, Clement JL. Combined flexure–compression loading for RC columns externally strengthened with longitudinal and transverse CFRP retrofitting. In: Proc concrete structures: the challenge of creativity fib symp. Avignon (France) 2004 [CD-Rom]. [27] Denton SR, Christie TJC, Powell SM. The development of design charts for column strengthening using FRP. Bridge management 4. London: Thomas Telford; 2000. p. 516–23. [28] Yuan XF, Lam L, Teng JG, Smith ST. FRP-confined RC columns under combined bending and compression: a comparative study of concrete stress-strain models. In: Teng J-G, editor. Proc FRP composites in civil engineering, vol. 1; 2001. p. 749–58. [29] Deniaud C, Neale KW. An assessment of constitutive models for concrete columns confined with fibre composite sheets. Compos Struct 2006;73(3):318–30. [30] Parvin A, Jamwal A. Performance of externally FRP reinforced columns for changes in angle and thickness of the wrap and concrete strength. Compos Struct 2006;73(4):451–7. [31] Chataigner S, Caron JF, Benzarti K, Quiertant M, Aubagnac C. Characterization of FRP-to-concrete bonded interface – description of the single lap shear test. Eur J Environ Civ Eng 2009;13(9):1073–82. [32] Clement JL, Verok K, Boulay C, Le Maou F. Retrofitting of reinforced concrete columns with composites: static compressive tests and modeling – creep behaviour. In: Proc computer methods in composite materials and structures CADCOMP. p. 53–61. [33] Association Française de Génie Civil. Réparation et renforcement des structures en béton au moyen des matériaux composites – recommandations provisoires. Bulletin scientifique et technique de l’AFGC; 2007. [34] Berthet JF, Ferrier E, Hamelin P. Compressive behavior of concrete externally confined by composite jackets. Part A: experimental study. Constr Build Mater 2005;19(3):223–32.