Résumé
Static recrystallization and grain growth take place during high-temperature annealing following deformation at lower temperature, and they reorganize grain and subgrain boundaries to produce entirely new, strain-free grains. These processes influence the microstructure and mechanical behavior of the material by reducing the level of long-range internal stress field. The non-linear flow law of ice I (sub h) makes it an ideal analogue to model rock deformation, thus analyses of recrystallized ice textures contribute to understanding mechanisms which control the nucleation and grain growth deep in the Earth's interior. Water ice microstructures allow for the study of grain-scale processes including nucleation, grain growth, recovery and recrystallization. On the other hand, these mechanisms directly influence the behavior of terrestrial ice systems (e.g., glaciers and ice sheets) in response to deformation. Therefore, their understanding may help us to improve climatic signal interpretation and predictions, that are mostly based on ice dynamics. However, at present there are no data by which recovery, and strain energy-driven grain boundary migration rates in ice can be estimated. Here we present the results of annealing experiments of pre-deformed polycrystalline water ice together with microstructural analyses by means of Automatic Ice Texture Analyzer (AITA) and high-resolution electron backscatter diffraction (EBSD) crystal orientation maps. We performed constant load uniaxial compression creep experiments at -7 degrees C on columnar samples, composed of mm-sized polycrystals, to attain 4% deformation and we applied post-deformational heat treatment at -5 degrees C. We carried out a series of AITA and EBSD maps in representative time steps up to 25 hours of total annealing in order to study the microstructural evolution. In this contribution, we evaluate the stored energy distribution and the rate of grain boundary migration during post-dynamic and static recrystallization, that control strain energy-driven grain boundary mobility of water ice. We have used the state-of-art open source MTEX toolbox for quantitative texture analysis. MTEX is a scriptable MATLAB toolbox, which permits all aspects of quantitative texture analysis using the modern methods and anisotropic physical property calculations.