Abstract
The crystallographic preferred orientations (CPO) of ice crystals that occur in glacial ice produce significant mechanical anisotropy. Understanding the effect of this is important in producing accurate mechanical models for ice streams and glaciers, as shear margins with strong CPOs are understood to facilitate ice stream flow and stability. It is implicitly understood that ice anisotropy is important and geophysical proxy and direct measurements of anisotropy are common. Furthermore, some numerical models attempt to incorporate anisotropic mechanics. However, there are almost no published experimental laboratory deformation data from natural ice samples that could be used to qualify mechanical anisotropy. We provide the first data from laboratory experiments on natural ice to qualify the effect of anisotropy on deformation kinematics and creep strength.
This thesis presents data showing the effect that the orientation of a strong preferred alignment of c-axes relative to compression direction has on strain symmetry, mechanical strength, fabric evolution and micro-structural characteristics of polycrystalline ice. Constant displacement rate experiments were conducted at -20℃ and 20 MPa of confining pressure at three different strain rates. Our experiments use ice from the Priestley Glacier shear margin, which is characterised by a very strong horizontal c-axis maxima sub-perpendicular to the shear plane. Cylinders cored parallel to the c-axis maximum deform two orders of magnitude slower under axial compression compare to those cored 45 degrees to the c-axis maximum, while cylinders cored perpendicular to the c-axis maximum deform one order of magnitude slower. Cylinders cored at 45 degrees and cylinders cored perpendicular to the c-axis maximum undergo approximate plane strain deformation while cylinders cored parallel undergo pure shear flattening. Our results suggest that ice with a strong c-axis maximum can constrain its own kinematics. We document
peak/yield stress exponent values approaching 3 for ice oriented for hard-slip, values greater than 4 for ice in easy-slip, and stress exponent values of 5 or greater for ice in approximate steady state flow. These experiments provide important constraints for modelling parameters, and indicate that Glen’s flow law, as an isotropic secondary creep flow law, is inappropriate for use in environments where ice is strongly anisotropic. The micro-structural data set produced from these experiments is vast, and holds great potential for future investigation of a variety of micro structural features. The CPOs after experiments show that samples oriented for easy-slip maintain a cluster of c-axes, but that these weaken and rotate consistent with deformation by sliding on the basal planes. CPOs in samples oriented for unstable hard-slip rotate to form two c-axis maxima reflecting lattice rotations needed to facilitate deformation. CPOs in samples oriented for stable hard-slip are weakened but have no substantial change in orientation. Several of the micro-structural features are anomalous for the deformation conditions: such as disconformities and low angle boundaries occurring with higher frequency at lower strain rates than higher strain rates. Our data set also highlights the importance of kink-banding as an anisotropic deformation mechanism. Over-all these experiments are a significant step in investigating the rheology of anisotropic natural ice in shear conditions, and, provide the first step towards producing an adequate flow law for the Priestley Glacier shear margin. Our data may also help with understanding the complex kinematics and resultant fabrics, micro and macro-structures that occur in grounding zones such as in Priestley.