Beam Group A  Beam Group A contains ihc heains strengthened at the bottom face only. Figure 6 to 11 show the  test results for these beams. The results of Beams H-50-1 and H-75-1 were very close to  those of H-50-2 and H-75-2, respectively, and hcncc, the discussions concerning these beams arc focused on the last two to avoid repetition. The ductility of each beam is observed by calculating the ductility index as the ratio between the deflection of the beam at failure and its deflection at yield.    Figure 6(a) shows the load-vcrsus-midspan deflection diagram for Beam C-l, in which the carbon fiber sheet was used for strengthening. The beam yielded at a load of 85.9 kN (19.3 kips) and failed at a load of 101.9 kN (22.9 kips) due to rupture of the carbon fiber sheet. From this figure, it should be noted that, although ductile behavior is experienced, only a 4% increase inthe yield load compared with that of the control beam was achieved. A ductility index of 2.15 was experienced. Figure 6(b) shows the load versus carbon fiber strain at mid-span.  Figure 7(a) shows the load-deflection response  for Beam C-2. This beam was strengthened using the pultrudcd carbon fiber plate. The beam showed no yielding plateau (1.0 ductility index) and had a sudden failure at 132.6 kN (29.8 kips) due to shear-tension failure at the end of the plate. Although an increase in load of 61% was obtained, the failure was brittle. Figure 7(b) shows the load versus carbon fiber strain at midspan. The maximum recorded strain of carbon fiber plate at failure was 0.33%, which indicates that 24% of the capacity of the plate was utilized.  The load-deflection response of Beam C-3 is shown in Fig. 8(a). Beam C-3 was strengthened by two layers of the carbon fiber fabric. The beam yielded at a load of 107.7 kN (24.2 kips) and failed by fabric debonding at a load of 134.4 kN (30.2 kips) before showing any significant yielding plateau similar to that of the control beam. A ductility index of 1.64 was achieved. From Fig. 8(b), it should be noted that the maximum recorded carbon fiber strain at failure was 0.67%, which indicates that approximately 48% of the fabric capacity was utilized.  Figure 9(a) shows the load-deflection response of Beam H-50-2. This beam was strengthened with developed I mm- thick hybrid fabric. A yield load of 97.9 kN (22.0 kips) was experienced (a 19% increase in yield load over that of the control beam). It should be noted from Fig. 9(b) that the fabric had a strain 

of    25  0.40% when the beam yielded. The beam  experienced a ductility index of 2.33 and failed by total rupture of the fabric at an ultimate load of I 14.8 kN (25.8 kips). Figure 9(c) shows the beam at failure.  Figure 10(a) shows the load-deflection response for Beam H-75-2. The beam was strengthened with 1.5 mm-thick developed hybrid fabric. The beam yielded at a load of 113.9 kN (25.6 kips) and exhibited a ductility index of 2.13 before  total failure occurred from the debonding of the fabric at an ultimate load of 130.8 kN (29.4 kips). It is noticed that, although final failure was from the debonding of the fabric, it happened after achieving a reasonable ductility. Figure 10(b) shows that the fabric had a strain of 0.35% when the beam yielded. Figure 10(c) shows the beam at failure.  Figure 11 and Table 5 compare the results from Beam Group A. The following arc observed:  

1. Beams C-1 and 11-50-2 exhibited relatively good ductile behaviors. Beam 11-50-2, however, showed a higher yield load than Beam C-l. This is because the developed hybrid fabric was designed so that it has a higher initial stiffness than the carbon fiber sheet; hence, it contributed to strengthening more effectively than the carbon fiber sheet before the steel yielded; 

2. Although the carbon fiber fabric had an ultimate load several times greater than the yield-equivalent load of the 1.5 mm-thick hybrid fabric, Beam H-75-2 showed a similar behavior to Beam C-3. up to its yield. Beam H-75-2, however, exhibited a reasonable yielding plateau, and Beam C-3 did not;  3.Relative to current carbon fiber strengthening materials, the developed fabric has a yield-equivalent strain that is close to the yield strain of steel. Although it is still higher, hybrid fabric strain values were close to its yield value when the beam yielded, which indicated that it yielded simultaneously with the steel. This is attributed in part to the fabric being installed on the outer surface of the beam, which undergoes more tensile strain than inner steel. As a result, the designed yield strain value of the fabric seems to be acceptable; and  4. While the use of a carbon fiber plate of a high load capacity (like the one used in Beam C-2) provided a high failure load, it also produced a brittle failure.   Beam Group B  The beams in this group were strengthened at the bottom face and also up 152 mm (6 in.) onboth sides. The results of this group are shown in Table 5 and Fig. 12 through 15. The results of Beams H-S50-1 and H-S75-1 were very close to those of H-S50-2 and H-S75-2, respectively, and hence, the discussions concerning these beams are focused on the last two to avoid repetition.  Figure 12(a) shows the load-versus-deflection response of Beam CS. This beam was strengthened using the carbon fiber sheet system. The beam yielded at a load of 99.2 kN (22.3 kips) due to the yielding of the steel. The increase in  yield load was 20%. The beam failed at an ultimate load of 123.3 kN (27.7 kips) due to compression failure of concrete at midspan. Figure 12(b) shows that the carbon fibers had a strain of 0.35% when the beam yielded, and hence contributed approximately 30% of its capacity at this stage of loading. The maximum recorded strain before beam failure was 1.0%. A ductility index of 2.04 was attained.