| Résumé | In discussing ice crushing in the brittle regime Gagnon (2016) had speculated that small Tyndall melt figures, functioning in conjunction with a previously observed thin squeeze-film slurry layer of melt and ice particles at the ice/structure interface, could play a role in the removal of ice from the interface. Here we present supporting evidence for the existence of the melt figures. Using high-speed imaging, small features were observed to rapidly appear and grow in size at the ice/surface interface during pressure spikes caused by spallation of ice in the contact zone during crushing experiments conducted at −10 °C. The observations correlate with data from a previous study of inwardly-propagating Tyndall melt figures that nucleate at the surface of ice, situated in an oil-filled pressure vessel, when subjected to rapid adiabatic pressure increments. That is, during the ice-crushing experiments the mean growth length of the features was 1.5 ± 0.6 mm, where associated ice/surface interfacial pressure spikes (from similar tests) were about 36 MPa. For comparison, the earlier melt-figure study yielded a growth length of ~ 1.7 mm, which is within reasonable agreement with the ice-crushing value for the same rapid pressure increment. Furthermore, the features had orientations that generally were aligned with the two diagonal axes that connected opposing corners of the four-sided pyramid-shaped samples, where confinement was greatest. Hence the internal stress, and associated superheating, in hard-zone ice at the crushing interface during pressure spikes was greatest along the two axes. The features tended to grow in these directions, where the most heat was available, so this suggests they were melt figures. An incidence where a twin pair of features exhibited an angular separation close to a crystal symmetry angle also suggests the features were melt figures. The orientation data generally indicate that orientations of melt figures in non-uniformly rapidly-stressed ice may differ from orientations in uniformly stressed ice. We conjecture that in the present study small melt figures form a tenuous liquid matrix of shallow depth (roughly 0.2 mm) in the hard-zone ice surface during pressure spikes. A unique erosive effect may occur as the material in the top portion of this weakened layer is sheared off by, and entrained in, the ambient viscous flow of slurry. |
|---|
