CONTRIBUTION OF EXTERNALLY APPLIED FRP COMPOSITES TO CONCRETE SHEAR TRANSFER

Saenz, N., Pantelides, C.P., and Reaveley, L.D.

Department of Civil and Environmental Engineering, University of Utah, Salt Lake City, U.S.A.

ABSTRACT

This investigation is concerned with the determination of the contribution of fiber-reinforced polymer (FRP) composites to concrete shear transfer. The FRP composite was applied externally to uncracked plain concrete push-off test units designed to fail along a known shear plane. The FRP composite used was carbon fiber fabric with epoxy resin and was applied perpendicular to the shear plane. Test units with three shear-to-transverse stress ratios were constructed to encompass various shear transfer applications. The experiments included several FRP composite wrapping schemes and four FRP reinforcement ratios. The shear strength contributed by the concrete-FRP shear friction interaction is found to be a function of the concrete-to-concrete shear friction coefficient and the effective FRP composite tensile strain. The shear transfer units reinforced with carbon FRP composites increased shear transfer capacity by a factor from 1.32 to 3.25 times that of the as-is concrete units. 1. INTRODUCTION Shear transfer concrete test units internally reinforced with steel stirrups were used to determine the horizontal design of precast shear connections; this research, in addition to other findings, was used to develop the “shear friction” hypothesis [1]. Initially cracked and uncracked steel reinforced concrete connections have been studied [2-4]; modified shear friction equations and boundary conditions for its applicability were proposed; cohesion and friction effects for uncracked connections and limits on ultimate shear strength were also introduced. The shear friction concept has been studied for high-strength reinforced concrete [5, 6]. Shear connections for wall panels have been studied using shear transfer units, where evaluation of the shear friction coefficient factor was of interest [7].

In Fiber Reinforced Concrete (FRC), steel or polypropylene fibers are mixed with concrete. The shear strength and ductility of high-strength FRC concrete using shear transfer units has been studied [8]; specimens with polypropylene fibers had a lower increase in the ultimate load but greater ductility compared to specimens with steel fibers. In another study, steel fiber FRC concrete increased the shear transfer capacity up to 60% of the concrete compressive strength [9]. The Iosipescu test, adopted from ASTM D5379 [10], has been used to determine the shear transfer strength of RC members with CFRP composites [11]. Shear transfer strength for concrete internally reinforced with Glass FRP (GFRP) composite stirrups has also been studied, where some plasticity was observed because of gradual delamination of the GFRP composite [12].

In the present research, CFRP composite strips are used to strengthen the concrete externally at a known failure plane to resist stresses in shear transfer. Thirty-six shear transfer units were built and tested with these objectives: (1) investigate the efficiency of CFRP wrapping scheme configuration; (2) investigate the influence of CFRP reinforcement ratio; (3) determine the influence of shear-to-transverse stress ratio on shear transfer, and (4) understand the fundamental behavior of CFRP composite external connections in shear transfer. 2. DESCRIPTION OF TEST UNITS The test units were designed to fail in shear at a known plane, as shown in Fig. 1; the transverse stress, σy, and shear stress, τxy, are:

bLPCy σσ = ;

bhPCxy ττ = (1)

where if uniform stress distribution is assumed for both cases. From Eq. (1), the shear-to-transverse stress ratio is:

1== τσ CC

hLk

y

xy ==στ

(2)

Three different shear-to-transverse stress ratios were considered. The concrete units had constant L = 305 mm, and b = 127 mm, with k = 1.26, 1.50, and 1.85, corresponding to h = 165 mm (Type I units), 203 mm (Type II units) and 241 mm (Type III units), respectively. To ensure failure of the units in the shear plane, reinforcing steel was placed away from the shear plane so no other mode of failure such as flexural, compression or bearing capacity might be exceeded. Figure 1 shows typical steel reinforcement for units with a k-ratio of 1.85.

Material Properties The test units were all cast in one batch along with concrete cylinders, which were tested at the same time as the test units so that hardening due to concrete aging would not be a variable. The concrete compressive strength, , ranged from 34 to 37 MPa. Mild steel reinforcement was used with an f

'cf

y = 410 MPa which was not placed at the shear failure plane as shown in Fig. 1, so it would not influence the results. A CFRP composite with an epoxy-resin matrix was used; the carbon fiber was a 0.610 kg/m2 high-strength, unidirectional fabric with these properties: filaments per tow = 12,000; tensile strength = 4.76 GPa; tensile modulus = 234 GPa; density = 1800 kg/m3; and elongation = 1.5%. A high-modulus, high-strength epoxy-resin was used with these properties after 72 hours of curing at 60°C: tensile strength = 72.4 MPa; tensile modulus = 3.16 GPa; density = 1157 kg/m3; and elongation at break = 4.8%.

The tensile material properties of the CFRP composite laminate were determined following ASTM D3039 [13]. The average mechanical properties of the CFRP composite laminate were: tensile strength ffu = 903 MPa; tensile modulus, Ef = 68 GPa; tensile strain εfu = 1.33%; and ply thickness ti = 1.00 mm. A high-strength, high-modulus structural adhesive was used to enhance

Figure 1. Typical shear transfer test unit setup and steel reinforcement for Type II units

bond between the concrete surface and CRFP composite laminate, with these properties: tensile strength = 37 MPa; tensile modulus = 2.83 GPa; elongation at break = 1.3%; modulus of rupture = 46.2 MPa; and shear strength = 25.5 MPa. A 1 mm thick layer was spread with a plastic spatula on the concrete surfaces where the CFRP composite would be applied.

Surface Preparation and Layup From previous experiments, it is known that the most effective surface preparation technique for this type of application is high-pressure water jet with a high-strength adhesive/epoxy primer [14]. The water stream is delivered at a constant rate at a high-pressure (280 MPa) to the concrete surface using a rotating manifold to expose the aggregate. A wet-layup procedure was used to apply the CFRP composite. To saturate the carbon fiber, a custom saturating machine was used to maintain a constant fiber volume ratio.

Test Setup The experiments were conducted as monotonic tests under a compressive load, P, as shown in Fig. 1, using displacement control at 0.127 mm/s. To prevent local compressive failure, a 6 mm thick high-density polyethylene plate was inserted inside the steel caps. Electronic data was recorded every 0.02 s; the following instruments were used in the data acquisition process: (a) Linear Variable Displacement Transducers (LVDT), and (b) strain gauges.

3. EXPERIMENTAL RESULTS Four different wrapping schemes were tested with the same CFRP reinforcement ratio and unit type. Units of Type II (k = 1.50) and a CFRP reinforcement ratio of 0.6% were used for comparing various wrapping schemes. The CFRP reinforcement ratio is defined as:

hbwt fi

f **∑=ρ (3)

where ti = ply thickness; wf = ply width; b = width of rectangular cross section; and h = shear plane height, as shown in Fig. 2(d). For each of the four wrapping schemes, two units were tested to investigate the influence of the wrapping scheme on shear strengthening efficiency: (1) Type IIA: Four-sided wrapped with three evenly distributed single layer 25 mm-wide strips, as shown in Fig. 2(a); (2) Type IIB: Four-sided wrapped with a single layer 75 mm-wide strip, as shown in Fig. 2(b); (3) Type IIC: Four-sided wrapped with a double layer 38 mm-wide strip, as shown in Fig. 2(c); and (4) Type IID: Two-sided wrapped with a single layer 75 mm-wide strip, as shown in Fig. 2(d).

Type IIA: Four-sided wrapped with three evenly distributed single layer strips The maximum shear load, Pmax, was 185.3 and 201.6 kN and the maximum horizontal slip at the shear plane was 4.04 and 2.46 mm for the two units. Figure 3(a) shows a typical unit at failure, where diagonal tension cracks developed. Debonding of the CFRP laminate at the shear plane was observed, which extended 76 mm on each side of the shear failure plane defined in Fig. 1.

Type IIB: Four-sided wrapped with single layer strip The maximum shear load, Pmax, was 168.8 and 195.4 kN and the maximum horizontal slip at the shear plane was 0.38 and 1.91 mm for the two units. Figure 3(b) shows the typical unit at failure. Both units developed diagonal tension cracks; debonding of the CFRP laminate at the shear plane was observed, which extended 102 mm on each side of the shear failure plane.

Type IIC: Four-sided wrapped with double layer strip The maximum shear load, Pmax, was 177.4 and 188.5 kN and the maximum horizontal slip at the shear plane was 1.52 and 1.75 mm for the two units. Figure 3(c) shows the typical unit at failure. Debonding of the CFRP laminate at the shear plane was observed, which extended 89 mm on each side of the shear failure plane.

Type IID: Two-sided wrapped with single layer strip The maximum shear load, Pmax, was 195.1 and 165.7 kN and the maximum horizontal slip at the shear plane was 1.19 and 2.13 mm for the two units. Figure 3(d) shows the typical unit at failure, caused by bond failure of the CFRP laminate. This type of failure was observed on both sides of each unit. The bond failure mechanism was brittle and hence, diagonal tension cracks were not clearly observed. Figure 4 summarizes the experimental results in terms of normalized shear stress and CFRP tensile strain normalized by the CFRP ultimate tensile strain. For units with the four-sided wrapped scheme, the tensile strain in the CFRP composite located at the shear plane reached 6985 µs at failure; the highest strain reached at the side of the unit (b = 127 mm) was 1100 µs. There is no evidence that additional shear strength was developed due to the four-sided wrapped concrete with CFRP composite, as shown in Fig. 4.

(a) (b)

(c) (d)

Figure 2. Test unit detail: (a) Type IIA; (b) Type IIB; (c) Type IIC; (d) Type IID

) c ) Figure 3. Typi

Figure 4. Norm

(a)

cal unit at failure: (a) Ty

0 .0 0

0 .0 5

0 .1 0

0 .1 5

0 .2 0

0 .2 5

0 0 .1

'c

u

fv

alized ultimate shear st

(b

pe IIA; (b) Type IIB; (c)

0 .2 0 .3 0 .4 0 .5

T yp e IIT yp e IIT yp e IIT yp e II

fufe εε

rength-CFRP efficiency

(

Type IIC; (d) Type IID

0 .6

ABCD

for varying wrapping sche

(d

mes

According to ACI 440 [15], the four-sided wrapping scheme is the most efficient FRP composite wrapping scheme and the two-sided is the least efficient. Based on the experimental results, two parameters are affected by the wrapping scheme: (1) Shear strength, and (2) CFRP tensile strain. ACI 440 [15] makes no distinction between shear strength and CFRP tensile strain efficiency as a function of the wrapping scheme. From Fig. 4, shear strength shows no significant dependence on the wrapping scheme used. Tensile strain levels in the CFRP reinforcement show some dependence on the wrapping scheme, which can be up to 19% higher for four-sided compared to two-sided wrapped units.

Tests with Varying FRP Composite Ratio For this series of tests, the two-sided wrapping scheme Type IID of Fig. 2(d), with a single CFRP layer on each face, was used. Four CFRP ratios defined in Eq. (3) were used: 0.3%, 0.6%, 0.9% and 1.2%. Two units for each CFRP ratio and each of the three shear-to-transverse stress ratios were tested; in addition, two units were tested for each of the three shear-to-transverse stress ratios without CFRP composite shear reinforcement. The two-sided wrapping scheme was chosen to investigate the influence of CFRP ratio on shear strengthening because it is simpler and provides essentially the same shear strengthening as the four-sided wrapping scheme.

4. CFRP COMPOSITE EFFECTIVE TENSILE STRAIN The CFRP composite is stressed after the concrete shear friction or horizontal slip capacity of the as-is units is exceeded. The unit fails in two successive stages: first, the imposed shear stress is controlled by the concrete alone until the concrete shear friction capacity is reached; second, the additional imposed shear stress is controlled by the CFRP composite acting as a clamping force, which induces additional aggregate interlock/shear friction until the bond between laminate and concrete fails. In all tests, the highest strain in the CFRP composite occurred at the shear failure plane, and varied approximately linearly from the shear failure plane (highest strain) to the edge (least strain) as shown in Fig. 5. The CFRP composite had sufficient length exceeding the required bond length. The tensile strain along the midheight of the CFRP composite in the horizontal direction was symmetric about the shear failure plane. Figure 6 shows the effective CFRP composite tensile strain, defined as the effective strain at failure, εfe, over the ultimate

0

1000

2000

3000

4000

-152 -114 -76 -38 0 38 76 114 152Distance from Center (mm)

Stra

in ( µ

s)

L1

L2

C3

S3

R3

R4

R5

R2

Pu = 301.6 kN

289

267

244.7

127 (As-Built Fail. Load)

Figure 5. Typical CFRP composite tensile strain distribution in the horizontal direction

0.4

0.5

0.6

feε

CFRP tensile straidefined as the CFultimate tensile stredifferent CFRP reicorresponds to fρ =(4) establishes an aUsing the least squ

0==fu

fef ε

εξ

5. SHEAR TRANFrom the experimeshear stress was coreached; and (b) thas a clamping forcbetween laminate aadditional shear fbetween ultimate normalized by the 7. From Fig. 7, usi

505.0uv = ρ

The first term in Eqinteraction, where 0

0

0.1

0.2

0.3

0.0 0.2 0.4 0.6 0.8 1.0

Type IType IIType III

fuε

*fuff fEρ

Figure 6. CFRP tensile strain efficiency

n, εfu, versus the normalized stiffness of the CFRP laminate; the latter is RP reinforcement ratio times the modulus of elasticity, Ef, divided by the ss, . The CFRP efficiency ranges from 0.18 to 0.48, and each of the four

nforcement ratios is distinguishable. The normalized stiffness value of 0.20

*fuf

0.3%, the smallest CFRP reinforcement ratio used in these tests. Equation verage CFRP strain-efficiency, which varies with CFRP normalized stiffness. ares method, the CFRP strain-efficiency is:

27.0

*26.−

⎟⎟⎠

⎞⎜⎜⎝

⎛

fu

ff

fEρ ; 20.090.0 * ≥

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛≥

fu

fff

Eρ (4)

SFER CAPACITY ntal results, the test unit was observed to fail in two successive stages: (a) the ntrolled by the concrete alone until the concrete shear friction capacity was e additional imposed shear stress was resisted by the CFRP composite acting e, which induced additional aggregate interlock/shear friction until the bond nd concrete failed. To determine the concrete shear friction strength and the riction strength provided by the concrete-CFRP interaction, the relation shear to concrete compressive strength, '

cu fv , versus CFRP stiffness concrete compressive strength, '*

cfuf ffρ , is calculated and presented in Fig. ng the method of least squares:

'* 117.0 cfuf ff + (5)

. (5) is the shear friction strength contributed by concrete-CFRP shear friction .505 is the shear friction interaction coefficient, and is the effective *

fuf fρ

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35

Type IType IIType III

R2 = 0.91

'c

u

fv

'*cfuf ffρ

Figure 7. Normalized ultimate shear stress versus normalized CFRP stiffness

CFRP composite tensile stress, which is the clamping stress provided by CFRP reinforcement. The second term in Eq. (5) is the concrete shear friction strength, where 0.117 is the component for bond and asperity shear. Type I units with a shear-to-transverse stress ratio k = 1.85, for a CFRP reinforcement ratio ranging from 0.3% to 1.2%, increased the shear transfer capacity by a factor from 2.12 to 3.25; for Type II units with k = 1.50, the increase was 1.32 to 2.08 times; and for Type III units with k = 1.26, the increase was 1.50 to 2.41 times. From Fig. 7, the upper bound shear stress for the shear transfer units tested of all three types was 0.28 f’c, regardless of the shear-to-transverse stress ratio.

6. CONCLUSIONS Shear transfer tests were carried out for initially uncracked concrete that was strengthened with varying CFRP composite wrapping schemes. There was no evidence that additional shear strengthening was developed due to a four-sided wrapped scheme compared to a two-sided scheme. However, up to 19% higher CFRP tensile strain efficiency was observed for four-sided wrapped compared to two-sided wrapped units. The maximum CFRP tensile strain at failure for the four-sided wrapped units varied from 18% to 53% of the ultimate tensile strain and was a function of its stiffness. The failure mode was debonding of the CFRP laminate; the wrapping scheme type did not affect the shear-to-horizontal slip relationship. Units with a two-sided wrapped scheme were tested for varying CFRP reinforcement ratios. The imposed shear stress was controlled by concrete alone until the concrete shear capacity was reached; the additional imposed shear stress was resisted by the CFRP composite acting as a clamping force, inducing additional aggregate interlock/shear friction until the bond between CFRP laminate and concrete failed. The CFRP composite was not significantly stressed for loads below the concrete shear capacity, so the principle of superposition between concrete shear friction capacity and shear friction capacity due to concrete-CFRP interaction can be applied.

The CFRP composite tensile strain distribution along the CFRP strip for the two-sided wrapped units was symmetric with respect to the shear plane, and was approximately triangular; in addition, the length of the CFRP strip was sufficient to develop up to 48% of its tensile capacity. The shear transfer units reinforced with CFRP composites having a reinforcement ratio of 0.3% to 1.2% increased shear transfer capacity by a factor from 1.32 to 3.25 times. The increase in shear transfer capacity was a function of the shear-to-transverse stress ratio. An upper bound for the shear capacity of shear transfer units was established as 28% of the concrete compressive strength, regardless of the shear-to-transverse stress ratio. The failure mechanism in all the tests, regardless of the wrapping scheme, was bond failure of the CFRP composite laminate.

ACKNOWLEDGMENTS

The writers acknowledge the financial support of the National Science Foundation under Grant No. CMS 0099792. The writers acknowledge in-kind support from Sika Corporation and Eagle Precast Inc.; they also acknowledge the assistance of graduate students at the University of Utah.

References

1. Birkeland, P. W. and Birkeland, H. W., “Connections in precast concrete construction”, ACI J. Proc., 63/3 (1966), 345-368.

2. Mattock, A. H., Hofbeck, J. A. and Ibrahim, I. O., “Shear transfer in reinforced concrete”, ACI J. Proc., 66/2 (1969), 119-128.

3. Mattock, A.H. and Hawkins, N. M., “Shear transfer in reinforced concrete-recent research”, PCI J., 17/2 (1972), 55-75.

4. Mattock, A. H., Li, W. K. and Wang, T. C., “Shear transfer in lightweight reinforced concrete”, PCI J., 21/1 (1976), 20-39.

5. Mattock, A. H., “Shear friction and high-strength concrete”, ACI Struct. J., 98/1 (2001), 50-59. 6. Kahn, F. L. and Mitchell, A. D., “Shear friction tests with high-strength concrete”, ACI Struct. J., 99/1

(2002), 98-103. 7. Foerster, H. R., Rizkalla, S. H. and Heuvel, J. S., “Behavior and design of shear connections for load

bearing wall panels”, PCI J., 34/1 (1989), 102-119. 8. Valle, M. and Büyüköztürk, O., “Behavior of fiber reinforced high-strength concrete under direct shear”,

ACI Mater. J., 90/2 (1993), 122-133. 9. Allos, A. E., “Shear transfer in fibre reinforced concrete”, Int. Conf. on Recent Developments in Fibre

Reinforced Cements and Concretes, Eds. R. N. Swamy and B. Barr, Elsevier Appl. Sci., New York, (1989), 146-156.

10. ASTM, “Standard test method for shear properties of composite materials by the V-Notched beam method”, D5379/D5379M-98, (1998), West Conshohocken, Pennsylvania.

11. Dolan, B. E., Hamilton III, H. R. and Dolan, W., “Strengthening with bonded FRP laminate”, ACI Conc. Int., 20/6 (1998), 51-55.

12. Burgoyne, C. and Ibell, T., “Use of fiber-reinforced plastics versus steel for shear reinforcement of concrete”, ACI Struct. J., 96/6 (1999), 997-1002.

13. ASTM, “Standard test method for tensile properties of polymer matrix composite materials”, D3039/D3039M-00e1, (2000), West Conshohocken, Pennsylvania.

14. Pantelides, C. P., Volnyy, V. A., Gergely, J. and Reaveley, L. D., “Seismic retrofit of precast concrete panel connections with carbon fiber reinforced polymer composites”, PCI J., 48/1 (2003), 92-104.

15. American Concrete Institute, “Guide for the design and construction of externally bonded FRP systems for strengthening concrete structures”, ACI Committee Report 440.2R-02, (2002), Farmington Hills, Michigan.