A new approach for persistence in probabilistic rock slope stability analysis

Abstract

Discontinuity persistence is defined as the fraction of area (or length) that is actually discontinuous, with reference to a discontinuity plane which is through the rock mass containing a combination of discontinuities and intact rock regions. Although persistence is one of the most significant discontinuity parameters in slope stability analysis, it is impossible in practice to measure the discontinuity area accurately in a field investigation. Therefore, several researches have carried out on the basis of different approaches such as numerical analysis and fracture mechanics. In this study, the persistence is considered as a random variable since the persistence is difficult to obtain in the field and subsequently involves uncertainty. In addition, while most previous stability analyses have assumed that discontinuity on the failure plane is fully persistent, the probability that the joint length is long enough to produce a rock block failure (or that the joint length is equal to or greater than maximum sliding dimension) is evaluated in this study. That is, the probability of failure obtained from the previous approach is a conditional probability on the premise that the discontinuity on the failure plane is fully persistent. This approach simply uses joint length data rather than the persistence value in the procedure of obtaining the probability of fully persistent joint. Later the probability that fully persistent joint exists is multiplied by the probability of slope failure which itself is based on the assumption that joints are fully persistent. Consequently, in order to overcome the limitation of a conservative analysis, assuming 100% joint persistence, the proposed approach suggested new persistence concept based on the discontinuity length information. In this study, the proposed concept applies to the practical example.