DAMAGE-BASED LONG-TERM MODELLING OF PARAGLACIAL TO POSTGLACIAL EVOLUTION OF ALPINE ROCK SLOPES

Large mountain slopes in alpine environments undergo a complex long-term evolution from glacial to postglacial environments, through a transient period of paraglacial readjustment. During and after this transition, the interplay among rock strength, topographic relief, and morpho-climatic drivers varying in space and time can lead to the development of different types of slope instability, from sudden catastrophic failures to large, slow, long-lasting yet potentially catastrophic rockslides. Understanding the long-term evolution of large rock slopes requires accounting for the time-dependence of deglaciation unloading, permeability and fluid pressure distribution, displacements and failure mechanisms. In turn, this is related to a convincing description of rock mass damage processes and to their transition from a sub-critical (progressive failure) to a critical (catastrophic failure) character. Although mechanisms of damage occurrence in rocks have been extensively studied in the laboratory, the description of time-dependent damage under gravitational load and variable external actions remains still difficult. This PhD project aims at studying the long-term evolution of alpine rock slopes through their glacial, paraglacial and postglacial stages of evolution. In particular, I investigate the mechanisms driving the transition from a relatively undisturbed, deglaciating slope to a paraglacial rock slope affected by initial development of large slope instabilities, and finally to a mature, hydraulically coupled postglacial rockslide or Deep Seated Gravitational Slope Deformation (DSGSD). To do this, I developed a novel approach able to overcome the limitations of existing numerical modelling techniques and capture the long-term evolution of real, large rock slopes. In this perspective, starting from a time-dependent model conceived to reproduce laboratory rock deformation experiments by combining damage and time-to-failure laws, we pointed at reproducing both diffused and localized damage, meanwhile tracking long-term slope displacements from primary to tertiary creep stages. The adopted approach is completed by taking into account rock mass heterogeneity and property upscaling, time-dependent deglaciation and damage-dependent fluid pressure occurrence.