Experimental Studyof Reactive Powder Reinforced Concrete Beams Strengthened withCFRP for Critical Shear Zones

The behavior and shear strength characteristics of eight SFRHSC beams strengthened with CFRP strips subjected to combined bending and shear are studied in the present research (in addition to a 9 th control beam without CFRP strengthening). The studied variables were shear span to effective depth ratio (a/d) and the deep beam effect, the effect of end anchorage of the CFRP strips with the beams, and effect of the amount of wrapping (width and spacing of the CFRP strips).Tests show that the presence of end anchorage for the strips increases the shear capacity of the beams by 12%for beams with the same properties regardless to the compressive strength

, suggested the bond strength is proportional to �f ′ c .They showed that there is a bond development length of the CFRP strip beyond which no further increase in load force can be achieved.It has been found that total wrap or U-wrap with end anchor is the alternative solution for U-wrap if debonding is to be avoided (15) . However, total wrap is not practical from a constructability standpoint.Bousselham and Chaallal (16) , concluded that, the shear capacity gain due to the CFRP was greater for deep specimens than for slender specimens.

Experimental Program
The properties of the component materials, mixing procedures and curing used for the production of SFRHSCbeams and their control specimensare detailed below.The experimental details of the behavior and load carrying capacity of SFRHSC beams strengthened in their shear zone with CFRP are also presented.Some beams are tested to illustrate the effect of anchoring the end of CFRP strips.

Materials
The properties of cement, sand, reinforcing steel, steel fiber, silica fume, high range water reduction (HRWR) admixture, carbon fiber reinforced polymer CFRP, epoxy and the resulting SFRHSC used in this investigation are presented in this section.

Cement
Ordinary Portland cement Tasluja LAFARGE satisfying the experimental property No. 5 / 1984 (17) , and manufactured in Iraq was used throughout this work. The laboratory testing of this type of cement illustrates its physical properties as: The compressive Strength at 3 days and 7 days are 22.86 MPa and 25.46 MPa respectively.

Fine Aggregate
Natural sand from BAHR EL-NAJAF region in Iraq was used for concrete mixes of this study. The fine aggregate was passing from sieve No.40 (450 µm) maximum size; and remaining on sieve No.100 (150 µm). The main properties for the fine sand used are: The sulfate content is 0.47%, Specific gravity is 2.75, Absorption is 0.63%.

Silica Fume
Micro silica (MS) is a densified powder which is a highly active pozzolanic material (reactive powder).The chemical composition of micro silica used in this investigation is shown in Table (1). The micro silica used in this work conforms to the chemical and physical requirements of ASTM C1240-03 (18) as shown in Tables (2) and (3) respectively.The benefits that result from adding silica fume are related to changes in the microstructure of concrete. These changes result from two different but equally important processes:-I-Physical contribution; Adding silica fume brings millions of very small particles to a concrete mixture, it fills in the spaces between cement grains. II-Chemical contribution;When the concrete hydration releases calcium hydroxide. The silica fume reacts with this calcium hydroxide with the presence of water to form additional binder material (calcium silicate hydrate (C-S-H)).

Mixing Water
Ordinary tap water was used for mixing and curing all the concrete specimens used in this work.

High Range Water Reducing Admixture (HRWR)
A high performance concrete superplasticiser based on modified polycarboxylic ether manufactured and supplied by BASF ® under the commercial name Degussa GLENIUM 54 was used as a water reducing admixture with nominal dosage of 6 % of cement weight. GLENIUM 54 is almost free from chlorides and complies with ASTM C494 (19) Type A and F.

Steel Fiber
The steel fibers used in this test program were straight steel fibers manufactured by Bekaert Corporation. The fibers have the propertiessupplied by the manufacturer: Length, Diameter, Density, Tensile strength, Aspect ratio are 13 mm, 0.2 mm, 7800 kg/m 3 , 2600 MPa, 65 respectively.

Reinforcing Steel bars
Ukrainian reinforcing steel bars wereused to reinforce the flexural tension zone for the beams. The properties of the bars :Bar diameter (mm), Modulus of elasticity (GPa), Yield stress (MPa), Ultimate Stress (MPa) are 16, 245.03, 612, 727.9 respectively.

Steel plates
Mild steel plates (2mm thick) were used to make the end anchorage of CFRP strips. The properties of the plates Modulus of elasticity (GPa), Yield stress (MPa), Ultimate stress (MPa) are 200, 240, 340 respectively.

Carbon Fiber Fabric (CFRP)
The SikaWrap ® 230C is a unidirectional woven carbon fiber fabric used for strengthening the shear zones of the beams in the present work. All information related to this system is summarized in Table (4).

Adhesive (Sikadur-330) Properties
A two part, solvent free, thixotropic epoxy based impregnating resin / adhesive less viscous paste manufactured by Sika Company, was used for bonding the carbon fiber reinforced polymers laminate to the surface of reinforced concrete beam specimens.The properties of this resin which is taken from manufacturer's specification are tabulated in Table (5).

SFRHSC Mix Design
The dominant mix proportioning (by weight) for all beams and their reference specimens is 1:1:0.15 cement : fine sand : silica fume with water/cement ratio 0.24. 6 % by weight of cement of superplasticizer, and 2% volume fraction steel fibers.

Mixing and curing of Concrete Batches
The mixing of all the batches are in the following steps (21) : -The desired quantity of silica fume was mixed in dry state with the sand for 2 minutes.
-The cement was loaded into the mixer and mixed for another 2 minutes.
-The superplasticizer was dissolved in water and the solution was added to the rotary mixer and whole mix ingredients were mixed for about 5 minutes.
-Steel fibers were uniformly distributed into the mix slowly in 5 minutes during mixing process, and then the mixing process continued for additional 3 minutes. The hardened specimens were demolded after 24 hours. They were cured at about (50 -80) C o for 7 days in a fabricated water bath tank. After that the samples were left to be cooled at room temperature. Then all specimens were kept in a normal water tank up to 28 days age.

Beams Details
In this work nine shear tests of SFRHSC beams without steel stirrups and strengthened with CFRP strips are reported. Beams were designed to have extra strength in flexure to ensure shear failure. The beam details are presented in Fig.(1) and Table (6). For all beam specimens, the cross section was 100 mm wide and 180 mm in depth. The overall length was 1250 mm, with clear span 1150 mm.All beams (except beam B1 which is the control beam without any strengthening for shear) were strengthened using U-wrap CFRP strips which were bonded to the beams by epoxy and without stirrups. The strengthened beams were divided into two groups according to the anchorage of the CFRP strip end or not.

Strengthening with CFRP
The strengthening system used in this work comprised of fiber strips, epoxy, connectors (bolts) and small steel plates (anchorage). All the strengthening processes were done after 28 days of moist curing for the beam specimens. The strengthening process is detailed in (ref. 24 ) see Figs. (2).

Strengthening Schemes
Only the first specimen (B1) was kept without strengthening as control specimen, whereas the other eight beam specimens were strengthened with externally applied CFRP strips with or without end anchorage. The following different schemes illustrate the technique of this strengthening as shown in Figs. (3) to (6).

Test Measurement and Instrumentation
All specimens were tested as simply supported beams using two-point loading with shear span to effective depth ratio (a/d) equal to 1.5 and 3. Fig.(7) shows the details and instrumentation used for testing the beam specimens. The measurements includes:

6
• The central deflections were measured by using a dial gage of 0.01mm/div. sensitivity.
• A demec strain device was used on a length of 200 mm to measure the surface strains at the tension and compression zones. These demec points were fixed at specific positions as shown in Fig.(8).
• The strain of the CFRP U-wrap strips was measured at center of each strip with the 100 mm demec gage along the CFRP strip on the specimen side [see

Shear Strength results
Table (7) represents the shear strength at the appearance of diagonal crack (diagonal cracking loads) for SFRHSC beams without stirrups and strengthened in shear zones with CFRP, along with the ultimate shear failure loads, as measured during testing. In this study the diagonal cracking load is defined as the shear load at the time when the critical diagonal crack formed within the shear span propagated to the mid-depth of the beam. Also the failure types of the CFRP strips are shown in Table (7).
The effects of several parameters on shear behavior of SFRHSC beams containing shear strengthening with CFRP were studied in the present study were: 1. Effect of shear span to effective depth ratio (a/d). 2. Effect of the anchorage of the ends of CFRP strips with the beams. 3. Effect of the amount and distribution of the wrapping (width and spacing of the CFRP strips). The experimental results for each parameter are discussed below:

Effect of Shear Span to Effective Depth Ratio (a/d)
The diagonal cracking and ultimate shear failure loads of the tested SFRHSC beams decrease with the increase of (a/d) ratio. However, the values of the ultimate load in all tested beams with the variation of (a/d) ratio is much more pronounced than those of the diagonal cracking loads.
It can be clearly seen from Table (7) and Fig.(9) that by increasing the (a/d) ratio from 1.5 in B2 and B9 (deep beams) to 3 in B5 and B7 respectively, the diagonal cracking load decreased by 100 kN, 25 kN respectively, which represents a decrease ratio of 40 %, 14.28 % respectively. Also for the same increase in the (a/d) ratio the ultimate shear strength decreases by 175kN, 141kN respectively, which represents a decrease of 43.2 %, 39.6 % respectively. The test results for shear failure of the beams have shown that for (a/d) of 1.5, shear failure due to arch action (deep beam effect) is dominant; whereas in the region with (a/d) of 3 the beam action due to concrete teeth (beam action) is more effective.

Effect of End Anchorage of the CFRP Strips
Anchorage the ends of the CFRP strips decreases the required development length for the bonding with concrete, this is more effective for the limited depth girders (22) . In this work, the strengthening of beams in shear is with U-wrap CFRP strips, and with two methods, the first group is with end anchorage for beams B2, B3, B4 and B5, while the second is without anchorage, for beams B6, B7, B8 and B9 (see Table  6).

7
Table (7) clearly shows the increase in the ultimate load of beams due to the end anchorage of the strips. For example, B2 and B3 had a failure load increased by 49 kN and 10 kN respectively compared with B9 and B8 due to the strip ends anchorage. Also the end anchorage of the CFRP leads to increasing the diagonal cracking load for the same beams by 75 kN and 50 kN respectively.
It is obvious from the results of the tested beams, in case of using the mechanical fixing technique of CFRP (using the end anchorage) the overall length of the CFRP strip works as a link between the top zone (concrete compressive cord) and the bottom zone (main reinforcement tension cord) similar to the steel stirrups function. This is by preventing cracks widening and propagation to the compression zone, which leads to a more uncracked concrete sections.

Effect of Width and Spacing of the CFRP Strips
The type of failure of the CFRP strips depends on the values of the bond stress (produced between the FRP and the concrete) and its resultant bond force that are normal to the crack, and these values depend on the spacing and active bond area of FRP (active bond area = FRP effective width × FRP effective length) (23) . The failure shape and the design of the FRP depend on the quantity and distribution of CFRP strips used in shear strengthening or repairing of the beams (24) .
As it is mentioned in Table (6), the beams in this work were fabricated in two groups. The first group, beams B2, B5, B7 and B9 were strengthened with CFRP strips of 40 mm width with a spacing of 100 mm center to center. While in the second group, beams B3, B4, B6 and B8 were strengthened with CFRP strips of 75 mm width spacing at 200 mm.
It can be clearly observed from Table (7), that beams B4, B6 and B8 from the second group have larger ultimate loads than their analogous beams B5, B7 and B9 respectively from the first group, by 20 kN, 10 kN and 4 kN respectively. The results showed approximately equal values of diagonal cracking loads for these beams from both groups. this may be attributed to the delay of the response of CFRP strips to resist the stresses transported from the concrete.

Crack Patterns
Two types of cracks may be observed in the tested beams; the flexural cracks resulting from flexural tensile stresses at the region of the simple beam cross-section below the neutral axis, and the shear cracks which are formed as a result of the inclined or "principal" tensile stresses acting on the beam at regions of combined moment and shear (2) .
The type of failure that occurs depends mainly on the (a/d) ratio. In normal concrete beams, diagonal tension failure generally occurs when this ratio is between 2.5 and 6.0. While shear-compression failure generally occurs when this ratio is between 1.0 and 2.5 (deep beam effect). However, there is no distinct boundary between these two failure types and either type may occur when (a/d) ratio is in the range between 2.0 and 3.0 (25) .
The CFRP strips did not affect the basic mechanism, nor did they affect starting of cracking or the angle of diagonal compression which formed. The effect of the CFRP strips was noticed once the crack began to open up and propagate, at which time they began to influence the behavior of the member (22) . 8 Fig.(10) shows photographs of the crack patterns after the failure of the tested beams of this work. The numbers shown beside the crack indicate the load when the crack penetrated to that position.
The behavior of the control beam B1 specimen supports the idea which states that, the SFRHSC tends to be more slender (beam action is dominant) compared with NSC beams, through the mode of failure was flexural mode with the absence of shear reinforcement. However, its obvious from table (7) and Fig.(10) that CFRP strips did not effect the mode of failure of the specimens,especially when end anchorage was not used.In addition,concrete crushing in the compressive zones was dominant due to the high values of tensile reinforcement ratio and volume fraction of the steel fibers.

Load-Deflection Behavior
The load-midspan deflection curves of the tested SFRHSC beams are plotted in groups having beams differing in the parameters considered with the other variables being kept constant as shown in Figs. (11) to (14). The load deflection of the control beam B1 is plotted in the Figs. (13) and (14) to show the effect of the CFRP strengthening and the (a/d) ratio on the behavior of SFRHSC beams.Comparisons among the load-deflection relationships of all tested SFRHSC beams will be discussed and related to major characteristic items. These are the deflection characteristics at ultimate load, and the ductility ratio as shown in Table (8).

Effect of Shear Span to Effective Depth Ratio (a/d) on The Load -Deflection
The load-midspan deflection curves for SFRHSC beams, Figs. (11) and (12), show that for a given load level at early stages of loading, the deflection increases with increasing shear span to effective depth ratio (a/d). Also, it is evident from these figures and Table (8) that the final deflection at ultimate load decreases with the increase of (a/d) ratio. As shown by the results of SFRHSC beams with and without anchoring the CFRP strips, the increase in (a/d) ratio from 1.5 to 3 causes a decrease in the ultimate deflection by about 50.6% for beams with anchored 40 mm CFRP strips with 100 mm spacing, and 26.52% for beams with anchored CFRP of 75 mm width @ 200 mm spacing as shown in Fig.(11).
For beams without anchored CFRP strips the increase in (a/d) ratio from 1.5 to 3 causes a decrease in the ultimate deflection by about 51.45% for beams with CFRP of 40 mm width @ 100 mm spacing, and 13.1% for beams with CFRP of 75 mm width @ 200 mm spacing as shown in Fig.(12).It is clear from Table (8) that the ductility ratio decreases with increasing shear span to effective depth ratio (a/d). With varying (a/d) ratio from 1.5 to 3, the ductility ratio decreases by 55.34% and 57.93% for the beams with anchorage end CFRP of 40 mm @ 100 mm and of 75 mm @ 200 mm respectively. This ratio decreases by 68.84% and 68.38% for the beams without anchorage CFRP of 40 mm @ 100 mm and of 75 mm @ 200 mm respectively.

Effect of End Anchorage of the CFRP on The Load -Deflection
Regardless of the values of compressive strengths of the SFRHSC, it is observed that, the ends anchorage effect of the CFRP strips on the behavior of the tested beams, one can generally say that, the ends anchorage of the CFRP strips leads to a significant decrease in the ultimate deflections of the SFRHSC tested beams. However, Figs.(13), (14) and Table (  Fig. (13) shows that for the same (a/d)=1.5 and amount of strengthening with CFRP (the width and spacing of strips), there is an increase in the ultimate deflection of the beams at the ultimate load when the end anchorage exists. This increase is of about 14% and 5.3% for the beams strengthened with 40 mm CFRP strips spacing at 100 mm and of 75 mm @ 200 mm respectively. Also, Fig.(14) shows that, for (a/d) =3 an increase by 15.48% was observed in the ultimate deflection due to the end anchorage for the beams strengthened with CFRP of 40 mm @ 100 mm. While the same figure shows that the varying values of ultimate deflections due to the end anchorage is decreased by 12% for the beams strengthened with CFRP of 75 mm @ 200 mm. Despite that the overall behavior of the beam B4 had a proper line which gives less deflection values for all loading stages than the behavior of its analogous beam B6 and an increase in the ultimate load. This may be due to the difference in the nature of failure in the two beams.

Effect of Width and Spacing of the CFRP Strips
From observing Figs. (13) and (14), one can conclude that approximately in general, increasing the width and the spacing of the CFRP strips leads to increased the stiffness of the strengthened beams. However, with increasing width and spacing of the strips the ductility ratio increased by 13.4% and 8% for the beams with ends anchorage having (a/d) 1.5 and 3 respectively, while for the beams without anchorage the ratio increased by 6.92% for the (a/d = 1.5).

FRP Strain
The load-vertical FRP strains (in the direction of the fibers orientation) measured at the center of the strips are shown in Figs (15) to (22). The vertical strains of singlelayer of FRP strips are for the outside of the strips and there is no relation of these measured strains with the epoxy or the concrete under the strips. It is clear that the strain of the FRP increased at strips which crossed the diagonal cracks.Approximately in all beams the initial loading stages show compressive or zero strains for the strips, and when the cracks formed, the shear stresses were redistributed along the shear spans of the beams and the CFRP strips started to respond for this stress redistribution and generating the CFRP strains. However, the test results show that the first two strips near the supports have the major strains compared with the other strips.

CONCLUSIONS
1. Experimental tests of the eight SFRHSC beams strengthened by CFRP strips indicate that the presence of end anchorage for the strips increases the shear capacity of the beams by 12%. 2. The amount and distribution (width and spacing) of the CFRP strips show an influence on the behavior and shear capacity V of the SFRHSC beams. In the present work, there is an increase in the ultimate shear capacity by 8%, 4.4% and 1.1% for beams B4, B6 and B8 (having the wider width and larger spacing of strips) respectively. 3. Approximately, for all tested SFRHSC beams there is a small difference in the values of the diagonal cracking strength even for beam B1 without strengthening, this may be attributed to the retardation of the CFRP strips to response and resisting the tensile stresses transported from the concrete due to forming the micro diagonal cracks.

10
Despite the presence of deep beam effect (a/d< 2) in beam B9, the beam failed by flexure. This means that, the CFRP strips play a role similar to stirrups as vertical web reinforcement, while the steel fibers replaced the horizontal web reinforcement.

11
Supplied by the manufacturer