Experimental results are presented in this paper to show that a runaway type of pre-failure instability can occur for sand under dilatancy rate controlled conditions when an appropriate strain increment ratio, dεv/dε1, is imposed. This type of instability is similar to the runaway type of instability observed for very loose sand under undrained conditions. Whether a soil element will undergo pre-failure instability depends on the difference between the strain increment ratio of the soil obtained from drained test, under a specified effective confining pressure, (dεv/dε1)s, and the strain increment ratio imposed during the test, (dεv/dε1)i, rather than the absolute magnitude of (dεv/dε1)i. Based on the experimental data obtained in this study it was found that an instability line can be determined from a series of strain path tests conducted at different effective confining pressures but with the same dεv/dε1 by joining the peak points of the effective stress paths to the origin in the q–p′ stress space. This line is similar to the instability line obtained from undrained tests on loose sand. The instability tests under dilatancy rate controlled conditions indicate that the stress ratio at the onset of instability obtained in the instability tests coincide with the peak stress ratio line. This suggests that the peak stress line can be used to predict the onset of instability under dilatancy rate controlled conditions in the same way as the use of instability line to predict the onset of instability under undrained conditions.
展开▼
