Micromechanical modeling of snow failure

Abstract

Abstract. Dry-snow slab avalanches start with the formation of a local failure in a highly porous weak layer underlying a cohesive snow slab. If followed by rapid crack propagation within the weak layer and finally a tensile fracture through the slab appears, a slab avalanche releases. While the basic concepts of avalanche release are relatively well understood, performing fracture experiments in the lab or in the field can be difficult due to the fragile nature of weak snow layers. Numerical simulations are a valuable tool for the study of micromechanical processes that lead to failure in snow. We used a three-dimensional discrete element method (3D-DEM) to simulate and analyze failure processes in snow. Cohesive and cohesionless ballistic deposition allowed us to reproduce porous weak layers and dense cohesive snow slabs, respectively. To analyze the micromechanical behavior at the scale of the snowpack (~ 1 m), the particle size was chosen as a compromise between a low computational cost and a detailed representation of important micromechanical processes. The 3D-DEM snow model allowed reproducing the macroscopic behavior observed during compression and mixed-modes loading of dry snow slab and weak snow layer. To be able to reproduce the range of snow behavior (elastic modulus, strength), relations between DEM particle/contact parameters and macroscopic behavior were established. Numerical load-controlled failure experiments were performed on small samples and compared to results from load-controlled laboratory tests. Overall, our results show that the discrete element method allows to realistically simulate snow failure processes. Furthermore, the presented snow model seems appropriate for comprehensively studying how the mechanical properties of slab and weak layer influence crack propagation preceding avalanche release.

Crack propagation speeds in weak snowpack layers

“…The PST system was generated using commercial DEM software; PFC3D (v5) (). For a detailed description of the DEM model for snow, and in particular how particle (and particle contact) parameters relate to macroscopic snow parameters, refer to Bobillier and others (2020). Recently, Bobillier and others (2021) modelled the PST design and showed that crack propagation in weak snow layers can be reproduced with their DEM model (Fig.…”

“…Bobillier and others (2020) presented a 3-D DEM model to reproduce failure in a layered snow sample. DEM, first introduced by Cundall and Strack (1979), is composed of a large number of discrete connected and interacting particles.…”

“…Performing experiments in avalanche terrain is challenging and often not possible due to safety concerns, so it is a good idea to apply numerical models to simulate this process. Recently, models based on the discrete element method (DEM; Gaume and others, 2015; Gaume and others, 2017; Bobillier and others, 2020, 2021) and the material point method (MPM; Gaume and others, 2018, 2019) were developed to investigate crack propagation in snow. Both methods were validated against PST experiments, showing their ability to reproduce crack propagation processes at the scale of a PST.…”

Microstructural controls of anticrack nucleation in highly porous brittle solids

“…As aforementioned, the advantage of this approach is that parameters of the Sticky Hard Spheres model (volume fraction and coordination number) can be directly evaluated based on X-ray tomographic images by matching correlation functions 31 . While several recent numerical studies have analyzed the mechanical response of porous brittle solids like snow based on the real samples’ microstructures 18 , 19 , 21 , it has been recently shown that simplified structures made of spherical particles can be used to reproduce accurately important features of snow mechanics in the brittle range for different processes: failure initiation 35 , 36 , crack propagation 9 , 10 , snowflake fragmentation 37 , wind blowing snow 38 , snow granulation 39 and avalanche impact pressures 40 . One the one hand, this simplification makes us loose important information about the microstructure.…”

A Depth-Averaged Material Point Method for Shallow Landslides: Applications to Snow Slab Avalanche Release

“…Such an avalanche releases due to the failure of a weak layer buried below a cohesive snow slab (Schweizer et al, 2016). Although a lot of progress has been done regarding the understanding of slab avalanche release processes, including the onset and dynamics of crack propagation (Gaume et al, 2013;Bobillier et al, 2021;Trottet et al, 2022), the evaluation of the size of avalanche release zones still remains a major issue. This difficulty impedes avalanche forecasting and hazard mapping procedures which rely on the avalanche release size as input.…”