| In order to design the soil structures, we must consider the safety against seepage failure in the structures. After the failure of Teton Dam in 1976, hydraulic fracturing has been attracted a lot of attention, many experimental and analytical researches have been performed, but there are many unclear points such as mechanism and occurrence condition of hydraulic fracturing. Hydraulic fracturing is one of the major problems in embankment dams because it plays important role in the initiation of cracks in core zone in a dam body, but there are a few examples of experimental investigation using construction materials of embankment dams that are relatively large structures in soil structures that should be secured the safety against hydraulic fracturing. We must make the researches on hydraulic fracturing of embankment dams, because 37% of failures in the failures of embankment dams are caused by seepage failure and contraction of core width has been recently investigated in the construction. In this study, core material of rockfill dam under construction is used as material of laboratory tests. The maximum grain size, Dmax, is arranged to be 19mm and 2mm. To produce hydraulic fracturing in the laboratory tests, we use large size hollow cylindrical specimens with the external diameter of 30cm. We investigate the influences of Dmax and confining pressure, σ3, on the pressure at the hydraulic fracturing, pf.We find that it is very important to appropriately set the values of Dmax and the size of specimens in order to produce the phenomena of hydraulic fracturing in laboratory tests. It is difficult to enlarge the size of specimens in the laboratory tests, but it is also difficult to steadily produce the hydraulic fracturing when we use small size specimens. In this study, we find that the minimum size of external diameter of specimens is needed to be 15cm and large-size specimens tests constantly produce the hydraulic fracturing.There are linear relationships between pf and σ3, and pf is described as linear formula by σ3, pf = m×σ3 + n, with m and n as constant numbers that are decided by the test results, respectively. We find that, in the 30cm large size specimens tests, m of Dmax = 19mm is smaller than that of Dmax = 2mm but, in the 15cm middle size specimens tests, m of Dmax = 19mm is almost same as that of Dmax = 2mm. We can conclude that the number of m is at least 1.17 when the material causes the hydraulic fracturing in this study. |
