ABSTRACT
Vertical deformations were recorded during a series of direct shear tests performed on Ottawa Sand. Since the initial void ratio is known, it is possible to determine the final void ratio, and thereby determine if the specimen experiences a net decrease or an increase in volume. A dense sand specimen tested in a direct shear test will typically compress at relatively small horizontal displacements and then expand at higher displacements. In the other hand, a loose sand specimen is expected to compress during shearing. It was observed that some specimens that were prepared to a void ratio of 0.60 (dense condition) behaved as loose specimens (contraction). This can be explained if it is considered that the critical void ratio is not a constant but a function of the effective normal stress. Since the effective stress varies with suction stress, and this varies with degree of saturation, it can be expected that as the effective normal stress is increased it might be enough to relocate the normal stress onto the loose (contraction) side of the critical state line. This can lead to the thinking that for unsaturated soils not only the normal stress but the mechanisms involved in the unsaturated regimes acting in the specimen affect the deformation behavior of the soil. Because soil strength depends partially on how much deformation has occurred under a particular loading condition, the influence of these results on soil strength should be considered for Geotechnical Engineering Projects. Then a series on shear strength testing on cohesive soils was performed in order to compare to the results observed on granular soils.
