UNDERGROUND EXCAVATIONS IN JOINTED ROCKS
Excavation in rock masses requires a detailed knowledge of site geology, rock properties, hydrology, and other issues. A rock mass usually consists of intact rock separated by discontinuities. One of the most serious problems in tunnel excavation is the accidental falling of rock blocks that are formed by the intersection of the tunnel surface and discontinuities . Therefore, in the preliminary design stages of tunnel excavation, it is essential to give a reasonable estimation of unstable rock block volume and failure mode. In practice, the largest deterministic unstable rock volume is estimated for design purpose. However, the result may be too conservative and failure mode may vary. In this paper, a probabilistic approach is used in order to generate unstable block volume distribution. Fisher distribution is assumed for rock discontinuity data fitting and Fisher constant is employed to present the dispersion of each discontitnuity set. Both deterministic and probabilistic approach are carried out for unstale rock block volume determination. To account for the uncertainty reflected by the sample dispersion, discontinuity data are generated systematically around each mean discontinuity normal following Fisher distribution. The probability of rock block failure at a certain location and the largest possible block volume distribution are determined in a logical manner. The results show that deterministic can give an overall conservative estimation of the largest unstable block formed. However, the failure mode may vary due to joint cluster dispersion. Probabilistic approach shows more failure mechanism with different probability.