Evolution of the statistical distribution of crystal orientations in time- and space-varying viscous flows
Abstract
Magmas and other viscously deforming fluids in the Earth frequently contain embedded crystals or other solid inclusions. These inclusions generally rotate about their own axis and, under certain conditions, align themselves in a direction dictated by the details of the flow. This rotational behaviour has been studied extensively for homogeneous flows. Here, we couple the crystal rotation dynamics with the fluid mechanical Navier-Stokes equations for the large-scale flow, thus allowing the analysis of crystal rotations in settings that are variable in both space and time. The solution is valid provided that the intercrystal spacing is sufficiently large to preclude interaction between crystals. Additionally, we derive an evolution equation for the probability density function (PDF) of crystal orientations based on the fundamental concept of conservation of generic properties in continuum mechanics. The resulting system of equations is extensively tested against previous analytical and numerical solutions. Given the focus on method validation, we limit the fluid mechanics to simple systems with analytical solutions for the velocity field. Even for the simple examples computed, all of which are characterized by fluid flow that is constant in time, the crystal orientation patterns are spatially complex and change in time. Pressure-driven flow in a channel results in coherent bands of crystal orientations with band thickness decreasing towards the channel walls. In corner flow constrained by two mutually perpendicular walls, the pattern of crystal orientations does not exhibit any significant similarity with the flow field. Given that there is no local one-to-one correspondence between the flow and the PDF pattern, a combined and larger-scale solution of the two systems is generally required. The simple flow examples shown demonstrate the viability of this new approach. Application to more complex flow geometries which may commonly occur in nature is deferred to future studies.
- Publication:
-
Geophysical Journal International
- Pub Date:
- August 2019
- DOI:
- 10.1093/gji/ggz174
- Bibcode:
- 2019GeoJI.218..773B
- Keywords:
-
- Numerical modelling;
- Probability distributions;
- Physics of magma and magma bodies