Examining the effect of snow type on effective viscoplastic properties in micro-compression experiments through repeated load-relaxation cycles

It is well known that the snow type can affect the mechanical behavior during slow compression, which may indicate fundamental differences in the deformation mechanisms. To examine these differences, we performed consecutive loading-relaxation tests on three different snow types (rounded grains, depth hoar, and faceted crystals) at the same strain rate of approximately 10^-6 s^-1 using a micro-compression stage that allowed for X-ray tomography imaging before and after the experiment. By using consecutive loading-relaxation cycles, we were able to eliminate unavoidable structural transients that occurr during the first loading. This allowed us to study the stress-time data in the following cycles and probe the pure viscoplastic behavior of the intact ice matrix in the snow in the absence of microstructural changes. We could consistently analyze the stress-time data of all curves using an implicit, analytical solution of a non-linear Maxwell model for loading and relaxation. Our analysis showed that the estimated mechanical parameters were highly consistent between loading and relaxation and between consecutive cycles. We observed that the exponent n in Glen's law takes either high or low values depending on snow type: rounded grains with n=1.9 and depth hoar/faceted crystals with n=4.4. The transition from rounded grains to depth hoar/faceted crystals also appears consistent with an underlying influence of the optical equivalent diameter but clearly rules out a previously hypothesized dependence of n on volume fraction. In contrast, the effective compactive viscosity obtained from loading and relaxation had a dependency on volume fraction. Our results complement the understanding of how snow type and microstructure influence the mechanical behavior during slow compression, which we discuss in terms of potential transitions in dominant deformation mechanisms.