STATISTICAL EFFECTS ON THE EVOLUTION OF COMPLIANCE AND COMPRESSIVE FRACTURE-STRESS OF ICE

The large scatter in the experimental data of ice is partly due to the difference in the microstructure. The micromechanical damage model of Wu and Shyam Sunder [Wu, M.S., Shyam Sunder, S., 1992. Elastic anisot ropy and micro-damage processes in polycrystalline ice, Part I. Theore tical formulation. Int. J. Fracture 55, 223-243.] is used to investiga te the statistical effects caused by the randomness in the microstruct ural properties such as: microcrack length, grain size and crystallogr aphic orientation, Probability density functions are constructed from experimental data to represent the distribution of parameters, and are used to predict the evolution of the compliance. A probability theory is used to formulate the probabilistic distribution of the failure st ress and the compliance caused by the distribution of the underlying m icrostructural parameters. The 10 to 90th percentiles of the failure s tress obtained from the model analysis reasonably match with the scatt er in the given experimental data. (C) 1997 Elsevier Science B.V.