Abstract
Two alternative approaches to the conventional e-log p method for determining yield stress are critically examined in this paper, and their advantages and limitations are identified. The yield stresses defined in both alternative approaches are also theoretically verified as identical for the oedometer test data. A comparison of the advantages and limitations of both alternative approaches suggests that the ln(1+e)-log p approach should be used to interpret the oedometer test data for natural clays, and the defined yield stress can be interpreted as being responsible for the critical state of structure collapse. Systematic analyses on the oedometer test data for undisturbed Ariake clays and Shimabara clay demonstrate that it is valid to use the ln(1+e)-log p approach for determining the yield stress and both the compression and swelling indices of natural clays. In addition, some analysis implications are also given for useful parameters, such as, the critical-state pore pressure parameter in predicting the undrained strength of overconsolidated clays and the compression index ratio in evaluating the aging effect of natural clays. Finally, Nagaraj's sample disturbance degree concept based on oedometer test data is extended to take into account both the effects of stress release and mechanical disturbance. The analysis results indicate that the defined degree of sample disturbance can reflect the combined extent of stress release and mechanical disturbance during sampling and handling.
