ABSTRACT
Most rock slopes are inhomogeneous structures comprising anisotropic layers of rock characterized by different material properties, and they are often discontinuous because of jointing, bedding and faults. In rock slope stability analyses, the failure surface is often assumed to be predefined as a persistent plane or series of interconnected planes, where the planes are fitted to the surfaces based on the structural observation. Such assumptions are partly due to the constraints of the analytical technique employed (e.g. limit equilibrium method, the distinct element method, etc.) and can be valid in cases in which the response of single discontinuity or a small number of discontinuities is of critical importance on the stability. However, especially on a large scale slope, it is highly unlikely that such a system of fully persistent discontinuous planes exists a priori to form the failure surface. Instead, the persistence of the key discontinuities may be limited and a complex interaction between pre-existing flaws, stress concentration and resulting crack generation, is required to bring the slope to failure. In small engineered slopes, excavation gives significant changes in stress distribution in the slopes and may generate fully persistent planes keep propagating with stress re-distribution. Larger natural rock slopes seldom experience such a disturbance and have stood in relatively stable features over the period of thousands of years. This does not imply that in natural rock slopes a system of discontinuities may not be interconnected developing the portion of where the failure surface will be formed. Strength degradation may occur in rock mass with time-dependent manner and drive the slope unstable state. Thus, rock slope instability
problem requires the progressive failure modeling to drive the slope to catastrophic events.
