Author(s): Ian J. Jordaan; Jing Xiao

Linked Author(s):

Keywords: No Keywords

Abstract: Recent field experiments of ice indentation in the medium scale, with contact areas up 3 m2, have shown a progression of failure from largely creep, enhanced by damage and related microstructural changes, at lower rates (up to 3 mm/s) to a combination of spalls with zones of high-pressure loading areas, also highly damaged, at the higher rates (20-100 mm/s). Here, "damage" implies changes to the structure of ice in compression that lead to increased compliance or reduced strength. This includes microcracks and the formation of fine-grained, fragmented (comminuted) ice. Another distinct difference in the failure modes was observed. In the low-rate tests, large spalls occurred suddenly, after the elapse of several, up to 50, seconds. This crack would run several metres up to the ice surface creating a very large flake with dimensions of the order of many metres. By contrast, in the high-rate tests, which would typically last a few seconds or less, the spalls would be localised near the indenter. The piece sizes were of the order of a third of a metre. These results indicate a time-dependence in the crack propagation and the source of a scale effect related to the time under load. Computer simulations of the above processes have been carried out using finite elements. The constitutive modelling for the material response includes a nonlinear damage mechanics model for ice in compression and a fracture model for cracks in tensile zones. The model for fracture consists of an initial flaw placed in the ice sheet, which propagates at a critical value of the strain energy release rate (SERR). The computer simulations were implemented for ice sheets and for geometries similar to the medium-scale indentation tests. Very good agreement between force-time curves for the slower indentation rates was obtained; the associated distributed damage also was realistic. For ice sheets without spalls, damage would tend to be distributed throughout the beam. The effect of initial flaws was to cause spalls to propagate; these followed approximately the kind of contour that had been observed in the field. The spalls were found to be an essential part of high-speed indentation; a purely damagebased model would give loads that far exceed those measured. The result of spalling is to cause localisation of damage with high stress concentrations. The initial spalls are related to the contact geometry and high stresses associated with the initial rapid rise in stress. Boundary conditions for spall analysis are important.

DOI:

Year: 1992