The Evolution of Ice Fabrics: A Continuum Modelling Approach Validated Against Laboratory Experiments

Abstract:

We develop the first fully constrained continuum model, validated against experiments, to predict the evolution of a crystal fabric, such as ice, for any flow field or temperature. We use a mesoscopic continuum approach to model the evolution of a probability distribution function of c-axis orientations. The model assumes that ice deforms by dislocation creep with slip primarily along the basal plane. It incorporates the effects of rigid body rotation, grain-boundary migration and rotational recrystallization. We solve the system using a new spectral method, which is computationally highly efficient, and able to resolve fully the multiple dimensions of the problem (time, space and the two dimensions of orientation angle). By inverting the model against data from laboratory experiments in simple shear we establish the first values for the parameters controlling the importance of recrystallization processes as functions of temperature. We also provide constraints on the strain-rate dependence. Inverted parameters from simple shear are then applied to compression, and give excellent agreement with experimental results. Thus, the combination of the model, the spectral method and parameters as functions of temperature are able to give accurate and efficient predictions of ice crystal fabric evolution for general deformations, temperatures and strain rates. The model-solver, SpecCAF, can be extended to other polycrystalline materials such as olivine in the mantle.