【正文】 r in shear, it is not possible to attain the full strength of the FRP [7,13]. Failure is governed by either fracture of the FRP sheet at average stress levels well 11 below FRP ultimate capacity due to stress concentrations, debonding of the FRP sheet from the concrete surface, or a significant decrease in the post cracking concrete shear strength from a loss of aggregate interlock. Thus, the effective average CFRP stress is puted by applying a reduction coefficient, R, to the CFRP ultimate strength as expressed in Eq. (4). The reduction coefficient depends on the possible failure modes (either CFRP fracture or CFRP debonding). In either case, an upper limit for the reduction coefficient is established in order to control shear crack width and loss of aggregate interlock. . Reduction coefficient based on CFRP sheet fracture failure The proposed reduction coefficient was calibrated on all available test results to date, 22 tests with failure controlled by CFRP fracture [10， 13]. The reduction coefficient was established as a function of ffE? (where f? is the area fraction of CFRP) and expressed in Eq.(5) for ?ff E? GPa. . Reduction coefficient based on CFRP debonding failure The shear capacity governed by CFRP debonding from the concrete surface was presented [9,10]as a function of CFRP axial rigidity, concrete strength, effective depth of CFRP reinforcement, and bonded surface configurations. In determining the reduction coefficient for bond, the effective bond length, eL , has to be determined first. Based on analytical and experimental data from bond tests, Miller [14] showed that the effective bond length slightly increases as CFRP axial rigidity, ffEt , increases. However, he suggested a constant conservative value eL for equal to 75 mm. The value may be modified when more bond tests data bees available. After a shear crack develops, only that portion of the width of CFRP extending past the crack by the effective bonded length is assumed to be capable of carrying shear.[13] The effective width, feW , based on the shear crack angle of 45176。) similar to SO34. . Test setup and instrumentation All specimens were tested as simple span beams subjected to a fourpoint load as illustrated in Fig. 3. A universal testing machine with 1800 KN capacity was used in order to apply a concentrated load on a steel distribution beam used to generate the two concentrated loads. The load was applied progressively in cycles, usually one cycle before cracking followed by three cycles with the last one up to ultimate. The applied load vs. deflection curves shown in this paper are the envelopes of these load cycles. Four linear variable differential transformers (LVDTs) were used for each test to monitor vertical displacements at various locations as shown in Fig. 3. Two LVDTs were located at midspan on each side of the specimen. The other two were located at the specimen supports to record support settlement. For each specimen of series SW, six strain gauges were attached to three stirrups to monitor the stirrup strain during loading as illustrated in Fig. 1a. Three strain gauges were attached directly to the FRP sheet on the sides of each strengthened beam to monitor strain variation in the FRP. The strain gauges were oriented in the vertical direction and located at the section midheight with distances of 175, 300 and 425 mm, respectively, from the support for series SW3 and SO3. For beam specimens of series SW4 and SO4, the strain gauges were located at distance of 375, 500 and 625 mm, respectively, from the support. 3. Results and discussion In the following discussion, reference is always made to weak shear span or span of interest. . Series SW3 5 Shear cracks in the control specimen SW31 were observed close to the middle of the shear span when the load reached approximately 90 kN. As the load increased, additional shear cracks formed throughout, widening and propagating up to final failure at a load of 253 kN (see Fig. 4a). In specimen SW32 strengthened with CFRP (90176。） .This ply [. 0176。 Shear。 Carbon fiber reinforced polymer 1. Introduction Fiber reinforced polymer (FRP) posite systems, posed of fibers embedded in a polymeric matrix, can be used for shear strengthening of reinforced concrete (RC) members [1–7]. Many existing RC beams are deficient and in need of strengthening. The shear failure of an RC beam is clearly different from its flexural failure. In shear, the beam fails suddenly without sufficient warning and diagonal shear cracks are considerably wider than the flexural cracks [8]. The objectives of this program were to: 2 1. Investigate performance and mode of failure of simply supported rectangular RC beams with shear deficiencies after strengthening with externally bonded CFRP sheets. 2. Address the factors that influence shear capacity of strengthened beams such as: steel stirrups, shear spantoeffective depth ratio (a/d ratio), and amount and distribution of CFRP. 3. Increase the experimental database on shear strengthening with externally bonded FRP reinforcement. 4. Validate the design approach previously proposed by the authors [9]. For these objectives, 12 fullscale, RC beams designed to fail in shear were strengthened with different CFRP schemes. These members were tested as simple beams using a fourpoint loading configuration with two different a/d ratios. 2. Experimental program . Test specimens and materials Twelve fullscale beam specimens with a total span of 3050 mm. and a rectangular crosssection of 150mmwide and 305mmdeep were tested. The specimens were grouped into two main series designated SW and SO depending on the presence of steel stirrups in the shear span of interest. Series SW consisted of four specimens. The details and dimensions of the specimens design