Back analysis of a large landslide in a flysch rock mass

Flysch is a sedimentary rock consisting of a rhythmic alternation of hard (limestone, sandstone, siltstone) and weak (marl, mudstone, claystone) layers. Because of the presence of layers with different physical properties, the mechanical characterization of heterogeneous rock masses such as flysch is a real challenge. Different methods have been proposed in the literature to characterize flysch, combining empirical classification indexes with laboratory tests. Most of these methods, however, were specifically designed for tunneling and underground excavations, and their applicability to slope stability problems is not yet fully investigated. In this study, we analyze a large landslide in a cretaceous flysch rock in order to compare the mobilized strength at failure with those predicted by the modified GSI method (Marinos and Hoek, 2001). The landslide occurred in the Savena River basin (Northern Apennines of Italy) on April 6, 2013, with a volume of about 3 million m3. Soon after the failure, geological, geotechnical, and geophysical investigations were carried out to detect the failure mechanism and define the landslide geometry. Back analyses of the failed slope were performed using both limit equilibrium and finite difference methods to estimate the in situ strength of the flysch. The results show that the mobilized rock mass cohesion is very low (c'≈ 20 ÷ 40 kPa) and that the modified GSI method can predict the in situ strength only assuming a disturbance factor D = 1. Moreover, the analysis shows that the linearization criteria proposed in literature to compute the equivalent Mohr-Coulomb parameters remarkably overestimate the rock mass strength.