Substructure dynamics : In-situ annealing coupled with numerical modelling

Sammanfattning: The understanding of substructural behaviour during post-deformational annealing is key to interpreting rheological adjustments during tectonic change, and the processes which cause them. The focus of this study is to use coupled in-situ experimental techniques with numerical simulation to increase understanding of substructure dynamics in geological materials. 2D in-situ annealing experiments have been conducted in the Scanning Electron Microscope, using Electron Backscatter Diffraction (EBSD) to collect information about the crystallographic orientation of the surface. A single crystal of halite, pre-deformed under uniaxial compression at temperatures of ~450 ºC with a strain rate of 6.9x10-6s-1, and to a final strain of 0.165, was examined. Different temperature time-paths were investigated with temperatures between 280-470 ºC and durations of heating between 30 min and 6 h. EBSD maps were taken before, during and after heating. Behaviour during annealing was found to be temperature dependent and could be divided into three main phases of development. Subgrain boundaries could be divided into five categories based on behaviour during annealing, morphology and orientation.While the 2D experiments provided valuable information, it is impossible to rule out the potential influence surface effects may have on annealing behaviour. In order to verify the results of these experiments, a 3D X-ray diffraction experiment was conducted at the synchrotron in Grenoble, France. The experiment followed a similar heating procedure as that for the 2D experiments and was performed on the same sample. This newly developed technique allows non-destructive internal examination of the crystal. Preliminary results suggest that similar processes may be occurring as those observed in the 2D experiments. Full analysis and comparison to the 2D results should determine whether behaviour is truly accurate, or if the experiments show some surface effects.Development of a numerical model for the processes occurring during annealing has also been undertaken. The experiments were directly compared to the simulation in order to improve the model. A first attempt at modelling the behaviour in the 2D experiments applied a phase field method of gradient reduction. While the model did reproduce some of the results from the 2D experiments, including plateau-building it did not realistically replicate other features. To combat the limitations presented by this method a new model is being developed focusing on new techniques in dislocation density and burgers vector calculation. Information about dominant rotation axes will be directly input along with Euler angle orientation. This model should be able to reproduce behaviour observed in the experiments more accurately.