Cohesive Crack Model and Fracture Energy of Steel-Fiber-Reinforced-Concrete Notched Cylindrical Specimens

The most commonly used method of measuring the fracture energy, $G_F$ , is the method proposed by RILEM TC-50. Although this procedure is widely examined for plain concrete, its applicability to steel-fiber-reinforced concrete (SFRC) needs further evaluation. In this paper, in addition to the RILEM specimen of the three-point bend test, a cylindrical test specimen is also examined. All cylindrical-beam specimens had $300\text{-}\mathrm{mm}$ length, with $250\text{-}\mathrm{mm}$ span, and $150\text{-}\mathrm{mm}$ diameter. The same molds of standard compression tests were used to cast the SFRC specimens. Steel fibers were crimped and had an average length of $32\mathrm{mm}$ . Fibers had an average tensile strength of $600\mathrm{MPa}$ and the apparent modulus of elasticity of $200\mathrm{GPa}$ . Results obtained from this study were compared with other experimental results on notched beams. The results of the cylindrical specimens were very close to the results of the beams with rectangular cross section. The effect of the volumetric fraction of fibers in SFRC on the $G_F$ for three low-percent-volume fractions of steel fibers, 0.5, 1, and 1.5%, was examined. The values of $G_F$ increased about 11, 15, and 21 times when fiber volume fraction of 0.5, 1, and 1.5% were added to plain concrete, respectively. The cylindrical beams also could be used for the evaluation of existing materials by drilling the structure and obtaining cylindrical cores. By means of linear elastic fracture mechanics analysis it was found that, the minimum depth of initial notch in cylindrical specimen should be greater than 10% of the diameter to ensure geometrical unstable crack growth. Moreover, a new trilinear cohesive law, which predicts both the peak-load and the postpeak behaviors more accurately than the bilinear cohesive law, is proposed.