Failure Modes of Sand in Undrained Cyclic Loading: Impact of Sample Preparation

This paper presents a systematic experimental investigation into the impact of specimen preparation on the cyclic loading behavior of saturated sand, including the deformation pattern, pore-water pressure generation, stress-strain relationship, and cyclic shear strength. Moist tamping and dry deposition were used in the laboratory to prepare sand specimens with distinct fabrics for cyclic triaxial tests under a range of conditions. It is found that the soil fabric formed by dry deposition can lead to unique failure modes different from those of moist-tamped samples in certain situations. These failure modes are hybrid in nature, characterized by a contractive response in the form of limited flow followed by cyclic strain hardening in the form of either cyclic mobility or plastic-strain accumulation. The hybrid nature of the failure patterns makes defining failure for liquefaction-resistance evaluation crucial; the conventional failure criteria based on a certain level of strain or pore-water pressure do not appear to properly represent the failure mechanism involved and may lead to a substantial overestimation of liquefaction resistance. The experiments reveal that the method used to reconstitute specimens or the soil fabric they form plays a role that is far more complicated than previously thought. Depending on the combination of relative density, confining stress, and degree of stress reversal in cyclic loading, a change of reconstitution method can have a marked or little effect on the nature of the response in terms of deformation pattern and failure mechanism; nevertheless, the two reconstitution methods always give significantly different liquefaction-resistance values under otherwise similar testing conditions.