Dry-snow slab avalanches are the main cause of avalanche fatalities in mountainous regions. Their release is a multi-scale process which starts with the formation of a localized failure in a highly porous weak snow layer underlying a cohesive snow slab, followed by rapid crack propagation within the weak layer. Finally, a tensile fracture through the slab leads to its detachment. The dynamic process of crack propagation, which affects the size of avalanche release zones, is still rather poorly understood. To shed more light on this crucial process, we performed a series of flat field fracture mechanical experiments, up to ten meters long, over a period of 10 weeks from January to March 2019. These experiments were analyzed using digital image correlation to derive high-resolution displacement fields to compute dynamic crack propagation metrics. We then used a 3D discrete element method (DEM) to numerically simulate these experiments to investigate the micro-mechanics. Both in the experiments and in the simulations, we observed a stationary regime after several meters of crack propagation. The DEM simulations showed that in this regime crack propagation is driven by compressive stresses. A parametric DEM study showed that the elastic moduli of the slab and weak layer, as well as weak layer shear strength, are key variables affecting crack propagation. Our results also highlight that these mechanical parameters influence the propagation distance required to attain the steady-state regime. Finally, DEM simulations on steep slopes showed the emergence of a so-called supershear crack propagation regime, driven by shear stresses, in which crack propagation velocity becomes intersonic. These simulations were confirmed by preliminary experimental results obtained on a steep slope. Our experimental and numerical datasets provide unique insight into the dynamics of crack propagation and lay the foundation for comprehensive studies on the influence of snowpack mechanical properties on the fundamental processes of slab avalanche release.
Insight Numerics in:Flux v1.25 Crack
Highly-porous cohesive granular materials such as snow possess complex modes of failure. Apart from classical failure modes, they show microstructural failure and fragmentation associated with densification within a local, narrow zone. Therefore cracks may form and propagate even under compressive load ('anticracks', 'compaction bands'). Such failure modes are of great importance in a range of geophysical contexts. For instance, they may control the release of snow slab avalanches and influence fracturing of porous rock formations. In the snow context, specific failure mechanisms of the ice matrix and their interplay with the microstructure geometry of snow are still poorly understood. Recently, X-ray computed tomography images have provided insights into snow microstructure and capability of directly simulating its elastic response by the finite element method (FEM). However, apart from thermodynamically driven healing processes the inelastic post-peak behaviour of the microstructure is controlled by localized damage, large deformations and internal contacts. As a result of the well-known limitations of FEM to capture these processes we use Peridynamics (PD) as a non-local continuum method to approach the problem. Due to its formulation, (micro)cracks and damage are emergent features of the problem solution that do not need to be known or located in advance. In this contribution we perform unconfined displacement controlled high strain-rate uniaxial compression simulations of snow microstructures within a peridynamic framework. Computed tomography images of snow specimen serve as a simulation data base. The obtained results show a novel insight into local failure of snow and allow a better comprehension of the underlying failure mechanisms. This study contributes to improve non-local macroscopic constitutive models for snow for future applications. 2ff7e9595c
Comments