2001 Volume 41 Issue 3 Pages 45-55
A series of permeability tests on both high-quality undisturbed samples and reconstituted samples was performed using a large-scale triaxial cell. Based on the test results and discussion, the following conclusions were noted. (1) The coefficient of permeability in the horizontal direction is larger than that in the vertical direction. However, its difference is between 10% and 70% and not so large. (2) The coefficient of permeability decreases with increasing confining stress. The effect of the confining stress can be understood as the effect of the void ratio. And the small change of the void ratio due to consolidation only leads to a small change of the coefficient of permeability. (3) There is no clear correlation between the physical properties and the coefficient of permeability. (4) The coefficient of permeability of gravelly soils is almost the same as that of sandy soils, even though the 50% diameter of gravelly soils is about ten to a hundred times that of sandy soils. This result implies that the large size particles of gravelly soils have no significant effect on permeability characteristics of gravelly soils. (5) A new definition for determination of D10, D20 and fines content only from those soil particles with a diameter smaller than 2mm was introduced. A correlation similar to sandy soils can be seen between the coefficient of permeability and D10, D20, and fines content based on the new definitions. (6) The effect of the small size particles of the gravelly soils on the coefficient of permeability was found significant based on the test results using samples with a special blend of particle size and density reconstituted from undisturbed samples. (7) The coefficient of permeability of multi-layer gravelly soils, in the direction perpendicular to the sedimentation, was significantly affected by the lowest permeability. And the coefficient of permeability of the multi-layer in total agrees well with the theoretical estimation. (8) Although the data is limited, there was no significant different of coefficient of permeability between undisturbed and reconstituted samples. This result accords with that reported for sandy soils by Hatanaka et al. (1997). This result also means that the coefficient of permeability of gravelly soil is not affected by the soil fabric. As a result, the in-situ coefficient of permeability of gravelly soils can be well estimated for practical purposes from the reconstituted samples with the same gradation properties.
