Pore Water Pressure Development in Non-Uniform Cyclic Triaxial Tests

A variety of models have been developed over the past few years to predict pore pressure development in saturated cohesionless soils subject to seismic loading. These models are based primarily upon the results of uniform cyclic tests. Little data is available on pore pressure development during non-uniform loading. This report describes the results of ten non-uniform cyclic triaxial tests performed on Monterey #0 sand at a relative density of 54%.

The different series of tests are performed. Within each series, the magnitude and number of cycles of loading remains unchanged but the order in which the stress cycles are applied is varied. Results of these tests shows that the magnitude of pore pressure development depends upon the order of the cycles as well as their magnitudes. This finding is contrary to what is predicted using the Seed, Martin and Lysmer pore pressure generation model with a single, unique pore pressure generation curve.

Predictions of pore pressure development made using the Seed, Martin, and Lysmer pore pressure generation model do not agree well with the results of the laboratory tests. In this study, the theoretical predictions always underestimated pore pressure development. A modified model using stress-ratio dependent pore pressure generation curves is developed to ameliorate this deficiency. Predictions made using the stress-ratio dependent model are in substantial agreement with the laboratory test results. The primary discrepancy occurs when the applied cyclic stress ratio is less than the threshold value required to induce liquefaction in uniform cyclic tests.