Intersection Angle between shear fractures in Sea Ice
B. Tremblay, Damien Ringeisen, and Mathieu Plante
McGill University
Abstract
Sea ice models can reproduce the sea ice drift, the spatial and temporal scaling and the temporal intermittency of sea-ice deformation (Linear Kinematic Features) irrespective of the rheology used in the model (VP, EVP, EAP, MEB). While the same models can reproduce the lead density when they are run at sufficiently high resolution, the intersection angle between conjugate pairs of LKFs is systematically overestimated. Fracture in sea ice results in local weaknesses. Given that the alignment of fracture is not random – instead, it is aligned at some angle with respect to the main loading axis – a proper simulation of LKFs intersection angle is key for a proper representation of anisotropy (an emergent property) and the prediction of future deformation. There are two lead theories for the fracture angle in granular materials: the first entirely based on internal stresses imposed by external loading (wind and ocean current) and a specified yield curve; a second entirely based on post-fracture deformation and the alignment of grains (or floes) and the microscopic plane of deformation with respect to the macroscopic plane of deformation. In this presentation, we compare the intersection angle between conjugate pairs of LKFs simulated with sea ice models with normal or non-normal flow rules with the Mohr-Coulomb theory (applicable for normal or associated flow rule), Roscoe theory (applicable for non-normal flow rule), and Arthur theory (the average between the Roscoe and Mohr-Coulomb angle). Results from sea-ice models with a convex yield curve and a normal or non-normal flow rule are in perfect agreement with the theory (Mohr-Coulomb or Roscoe), but over-estimate the observed intersection angle. Models with a non-convex yield curve (e.g., the Mohr-Coulomb yield curve typically used for the deformation of granular material such as sea ice) are not in agreement with theory (Mohr-Coulomb or Roscoe) and still overestimate the intersection angle between conjugate pairs of fractures in both an MEB or VP framework when using an internal angle of friction derived from observation and assuming that MC theory holds.