【正文】 pecimen SW42 were less than those of specimen SW41. The applied load vs. midspan deflection curves for beams SW41 and SW42 are illustrated in Fig. 7. It may be noted that specimen SW42 resulted in greater deflection when pared to specimen SW41. When paring the test results of series SW3 specimens to that of series SW4, the ultimate failure load of specimen SW32 and SW42 was almost the same. However, the enhanced capacity of specimen SW32 (a/d=3) due to the addition of the CFRP reinforcement was 101 kN, while specimen SW42 (a/d=4) was 161 kN. This indicates that the contribution of external CFRP reinforcement may be influenced by the ayd ratio and appears to decrease with a decreasing a/d ratio. Further, for both strengthened specimens (SW32 and SW42), CFRP sheets did not fracture or debond from the concrete surface at ultimate and this indicates that CFRP could provide additional strength if the beams did not failed by splitting. . Series SO3 Fig. 8 illustrates the failure modes for series SO3 specimens. Fig. 9 details the applied load vs. midspan deflection for the specimens. The failure mode of control specimen SO31 was shear pression. Failure of the specimen occurred at a total applied load of 154 kN. This load was a decrease of shear capacity by kN pared to the specimen SW31 due to the absent of the steel stirrups. In addition, the crack pattern in specimen SW31 was different from of specimen SO31. In specimen SW31, the presence of stirrups provided a better distribution of diagonal cracks throughout the shear span. In specimen SO32, strengthened with 50mm CFRP strips spaced at 125 mm, the first diagonal shear crack was observed at an applied load of 100 kN. The crack 7 propagated as the load increased in a similar manner to that of specimen SO31. Sudden failure occurred due to debonding of the CFRP strips over the diagonal shear crack, with spalled concrete attached to the CFRP strips. The total ultimate load was 262 kN with a 70% increase in shear capacity over the control specimen SO31. The maximum local CFRP vertical strain measured at failure in specimen SO32 was mm/mm (. 28% of the ultimate strain), which indicated that the CFRP did not reach its ultimate. Specimen SO33, strengthened with 75mm CFRP strips failed as a result of CFRP debonding at a total applied load of 266 kN. No significant increase in shear capacity was noted pared to specimen SO32. The maximumrecorded vertical CFRP strain at failure was mmymm (. 31% of the ultimate strain). Specimen SO34, which was strengthened with a continuous CFRP Uwrap (908), failed as a result of CFRP debonding at an applied load of 289 kN. Results show that specimen SO34 exhibited increase in shear capacity of 87, 10 and % over specimens SO31, SO32 and SO33, respectively. Applied load vs. vertical CFRP strain for specimen SO34 is illustrated in Fig. 10 in which strain gauges sg1, sg2 and sg3 were located at midheight with distances of 175, 300 and 425 mm from the support, respectively. Fig. 10 shows that the CFRP strain was zero prior to diagonal crack formation, then increased slowly until the specimen reached a load in the neighborhood of the ultimate strength of the control specimen. At this point, the CFRP strain increased significantly until failure. The maximum local CFRP vertical strain measured at failure was approxi mately mm/mm. When paring the results of beams SO34 and SO32, the CFRP amount used to strengthen specimen SO34 was 250% of that used for specimen SO32. Only a 10% increase in shear capacity was achieved for the additional amount of CFRP used. This means that if an end anchor to control FRP debonding is not used, there is an optimum FRP quantity, beyond which the strengthening effect is questionable. A previous study [11] showed that by using an end anchor system, the failure mode of FRP debonding could be avoided. Reported findings are consistent with those of other research [7], 8 which was based on a review of the experimental results available in the literature, and indicated that the contribution of FRP to the shear capacity increases almost linearly, with FRP axial rigidity expressed by ffE? ( f? is the FRP area fraction and fE is the FRP elastic modulus) up to approximately GPa. Beyond this value, the effectiveness of FRP ceases to be positive. In specimen SO35, the use of a horizontal ply over the continuous Uwrap (. 90176。) resulted in a concrete splitting failure rather than a CFRP debonding failure. The failure occurred at total applied load of 339 kN with a 120% increase in the shear capacity pared to the control specimen SO31. The strengthening with two perpendicular plies (. 90176。) resulted in a 17% increase in shear capacity pared to the specimen with only one CFRP ply in 90176。 and CFRP debonding. Furthermore, two limits on the contribution of CFRP shear were proposed. The first limit was set to control the shear crack width and loss of aggregate interlock, and the second was to preclude web crushing. Also, the concrete strength and CFRP wrap ping schemes were incorporated as design parameters. In recent study [9,10], modifications were proposed to the 1998 design approach to include results of a new study on bond mechanism between CFRP sheets and concrete surface [14]. In addition, the model was extended to provide the shear design equations in Eurocode as well as ACI format. Comparing with all test results available in the literature to date, 76 tests, the design approach showed acceptable and conservative estimates [10,13]. In this section, the summary of the design approach is presented. The parison between experimental results and the calculated factored shear strength demonstrates the ability of the design approach to predict the shear capacity of the strengthened beams. demonstrates the ab