CONSOLIDATION AND FAILURE STATE
Figures 10, 11 and 12 show the e - log pnet relationships at the states of post-consolidation, post-dehydration, the failure state and the ultimate state. The failure state is defined as the state at the peak point or the maximum q during shearing. The ultimate state is defined as the state at the maximum εs during shearing in this research. The reason why the ultimate state is used instead of the critical state is that the unsaturated specimen continues to dilate even on the final state of the experiment. In Fig. 10 the specimen of OCR(overconsolidation ratio)=2 was used on the experiment carried out under pnet =25kPa because the normally consolidated soil under p' =25kPa was too soft to make the specimen.
The inclination slightly becomes gentle at post-dehydration compared with NCL. It means that the volume change is large in low net stress during the dehydration. Although the difference between post-dehydration state and failure state cannot be recognized in low net stress, the failure states are a little below the dehydration state in high net stress because contractive behaviour appears during shearing. It is found from these figures that the specimen dilates after the failure in low net stress because the ultimate state is located above the failure state. The difference in the tendency of these figures between the case of s =100kPa, s =200kPa and s =400kPa can hardly be recognized except for having more dilative behaviour in the case of s =100kPa.
Figure 13 shows the relationships between q and pnet at the failure state. Although the failure line on certain matric suction is parallel to the critical state line of the saturated soil in higher net stress, which agrees with the existing failure criterion, the strength is lower than that of the failure criterion in lower net stress. Moreover, the effect of matric suction is almost nothing in lower net stress.