| Přispěvatelé: |
Tajčmanová, Lucie, Moulas, Evangelos, Stünitz, Holger, Schmalholz, Stefan M., Eglington, Tim |
| Popis: |
Metamorphic petrologist and structural geologist study rocks that have experienced deformation and simultaneously underwent phase transformations and mineral reactions. However, these two processes – deformation and mineral reactions - are often studied separately, without investigating their mutual feedback. Similarly to studies on natural samples, experiments are performed in both fields separately. As a consequence, the mutual effect of deformation on mineral reactions and transformation and vice versa remains unclear. This thesis explores both the chemical and mechanical feedback on the final microstructural and petrological observations by employing experimental and numerical techniques. The thesis is divided into three case studies that build upon each other: The initial aim was to bring an experimental evidence on how deformation and stress are related to phase transitions. To achieve this goal, experiments close to the phase transition between calcite and aragonite were performed in a Griggs rig and the experimental stress conditions were varied. The results show different localisation patterns for the formation of aragonite and distribution of grain-size within the sample. These patterns are compared to numerical modelling results resembling the experimental set-up. Based on this comparison, a correlation of local differential stress with grain size and local first principal stress or mechanical pressure (=mean stress) with phase transitions is documented. Importantly, the documentation of such stress and pressure variations in the experimental sample contradicts the conventional experimental assumptions on homogeneous distributions of these properties within the sample. So far, mechanically heterogeneous materials have never been systematically analysed with respect to petrological modifications during deformation. Based on the previous results, a new experimental design was developed. The new setting induces well-defined stress heterogeneities inside the sample. This is achieved by introducing a non-reactive, strong elliptical inclusion into a weak material. For such an inclusion system, analytical solutions for pressure distribution under differential stress exist. Additionally, we use the numerical model developed in the previous case, to resolve the stress and strain state of the whole sample and assembly. This experiment shows, that, when deformed, the mechanically-heterogeneous material shows a spatial distribution of phase transformation that is correlated with the spatial variation in principal stress and pressure. Mechanical properties (including flow laws) are derived from rock-deformation experiments under the assumption of sample homogeneity in stress, strain-rate and viscosity. In the previous two chapters, it is shown that these properties may spatially vary by order of magnitude, depending on the sample geometry. To account for these variations a new way of determining flow laws was explored. This new approach relies on experimental data-sets that are used for the conventional flow law derivation. Importantly it minimises simplifications regarding the geometry of experimental set-up during the derivation of the flow law. This approach is based on forward numerical modelling of the experimental set-up with different flow law parameters. The model results are constrained against experimental measurements to find a best-fit flow law. For two datasets of general shear experiments the stress exponents derived by using the geometry resolving model are higher than previous predictions (2 and 2.25 are increased by 0.75 to 2.75 and 3 respectively). After extrapolation to natural strain rates, this change of the stress exponent increases stress predictions by at least one order of magnitude. The newly derived stress exponents show a better fit with flow laws derived from coaxial tests (n = 2.8 - 4). This thesis contributes numerically and experimentally to the discussion on coupling of deformation and phase transitions. By exploring the effects of stress and strain variations inside the experimental sample, a new perspective is offered. The new observations significantly affect the interpretation of both, observed phase transition distributions and variations in mechanical properties. Adopting this new, locally resolved, perspective offers a unified explanation of previously contradicting interpretations. |
