Yield states and stress–strain relationships in a natural plastic clay: Discussion

Abstract

The authors are to be congratulated for their very careful and well-performed tests on Winnipeg clays. The results add valuable data and further our knowledge of the yield behaviour of natural clays, and show a well-defined yield locus that can be normalized with respect to the vertical preconsolidation pressure o:,. The results show that the yield behaviour for Winnipeg clays, as for other natural clays, differs to some extent from that of the Cam-clay models. The establishment of the yield locus in this case involved an extensive triaxial testing program that can normally be carried out only for research purposes. Larsson and Sallfors (1981) have previously shown that a very good estimate of the yield locus can be obtained from the results of only two oedometer tests, one on an ordinary vertical sample and one on a horizontally trimmed sample. The method is based on the simple hypothesis that yield will occur when the preconsolidation pressure is exceeded in any direction or when the effective shear strength parameters are mobilized (see Fig. 1). This requires knowledge of the effective shear strength for stresses up to the preconsolidation pressure. In this range the effective shear strength forms a slightly curved envelope, which for many Scandinavian and Canadian clays can roughly be described by a straight line with +' = 30 and ct = 0, or in more detail, by a curve with the end points on the line 4' = 30 and c' = 0, and an average c' = 0 . 0 3 ~ : ~ in the intermediate stress range. The investigated Winnipeg clays differ from these clays in that the dominating clay mineral is smectite and FIG. 1. Simplified yield locus defined by the vertical preconsolidation pressure a:, the horizontal preconsolidation pressure DL,, and an effective angle of friction of 30. Standard plot for stress paths in triaxial tests.