ABSTRACT
Because of difficulties in testing and to save time most authors have conducted tests using rigid cell permeameters (Lafleur, 1999; Tomlinson & Vaid, 2000; Reddi et al 2000). Drawbacks come from the confinement characteristics which are uniaxial and from the existence of potential leakage flows between the sample and the cell. In order to reduce such flows, a rubber like diaphragm has been introduced between the rigid cell and the sample compacted within the cell (Kenney & Lau, 1985). It follows from this that the cell should preferably consists of a supple wall to prevent leakage and erosion of the sides of the sample on the one hand, and to control the sample confinement which affects soil structure, on the other hand. Some authors (Tomlinson & Vaid, 2000; Hameiri & Fannin, 2002) have tested samples consisting of glass spheres for ease of satisfying test repeatability. They, however, cannot account for cohesive soils and grain angularities. Yet, their employment capabilities make it possible to anticipate satisfactory results as to their use for phenomenological tests.
