Ice mechanical behavior is time-dependent, as has been known for many decades. But in many references, the attempt is made to use time-independent plasticity theory. The relevant analytical approach that accounts for time is viscoelastic theory. The need for this approach is made quite essential by study of microstructural changes that occur in ice under high stresses. In no case does there appear to be a clear yield condition, with flow occurring after a certain threshold value. Furthermore, the microstructural changes occurring under stress result in a highly significant enhancement of the creep rates. This results in a spatially varying viscoelastic response that is a function of prior stress history. The ice response is then a function of position resulting in a microstucturally modified layer in the region where compressive stress is applied. This can be deep or highly localized, depending on the loading rate. The most promising approach is that based on damage mechanics combined with viscoelasticity, using the thermodynamics of irreversible processes.
Ice is also prone to fracture, especially at high loading rates and under high stresses. This is basic to the notion of scale effect. Fracture processes are also time-dependent in viscoelastic materials, a phenomenon that needs to be explored further. Furthermore, failure often will take place in a random fashion, depending on the distribution of flaws in ice. This indicates strongly that a theory based on “weakest-link” hypotheses and probability theory is appropriate.
Finally, some aspects relevant to practical data analysis are discussed. These include measurement uncertainties of Molikpaq data, and geometric approximations of ice features, e.g. ridges as uniform beams.