 |
| Titel |
Evolutionary models of the Earth with a grain size-dependent rheology |
| VerfasserIn |
Antoine Rozel, Gregor Golabek, Paul Tackley |
| Konferenz |
EGU General Assembly 2015
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 17 (2015) |
| Datensatznummer |
250106337
|
| Publikation (Nr.) |
 EGU/EGU2015-6002.pdf |
|
|
|
|
|
| Zusammenfassung |
| Thermodynamically consistent models of single phase grain size evolution have been
proposed in the past years [Austin and Evans (2007), Ricard and Bercovici (2009), Rozel et
al. (2011), Rozel (2012)]. Following the same physical approach, the mechanics of two-phase
grain aggregates has been formulated [Bercovici and Ricard (2012a)]. Several non-linear
mechanisms such as dynamic recrystallization or Zener pinning are now available
in a single non-equilibrium formulation of grain size distributions evolution. The
self-consistent generation of localized plate boundaries is predicted in [Bercovici and Ricard
(2012b)] using this model, but it has not been tested in a dynamically consistent
way.
Our preliminary results have shown that out of equilibrium grain size dynamics
leads to localization of deformation below the lithosphere rather than subduction
initiation. Yet this result was obtained assuming indealized conditions. We study
here, for the first time, the evolution of grain size in the mantle and lithosphere in
evolutionary models, starting from a just-frozen magma ocean until the present day
situation. Following complexities are considered in these models: melting, phase
transitions, compressible convection, and different pressure-temperature-dependent
composite rheologies in upper and lower mantles. We use a visco-plastic rheology in
which the viscous strain rate is obtained by summation of dislocation and diffusion
creep.
Pressure and velocity fields are solved on a staggered grid using a SIMPLER-like method.
Multigrid W-cycles and extra coarse-grid relaxations are employed to enhance the
convergence of Stokes and continuity equations. The grain size is stored on a large number of
tracers advected through the computational domain (a 2D spherical annulus), which prevent
numerical diffusion and allows a high resolution. We also describe the physical formalism
itself and derive a set of free parameters for the model. The results show that Normal growth,
dynamic recrystallization and phase transitions all have a strong effect on the average grain
size.
Â
Austin N.J. and B. Evans. Geology, 35:343–346, 2007.
Ricard Y. and D. Bercovici. J. Geophys. Res., 114(B01204), 2009.
Rozel A. et al. Geophys. Journ. Int., 184(2):719–728, 2011.
Rozel A. Geochem. Geophys. Geosyst., 13(10), 2012.
Bercovici D. and Y. Ricard. Phys. Earth Plan. Int., 202-203:27–55, 2012a.
Bercovici D. and Y. Ricard. Earth and Plan. Sci. Lett., 365:275–288, 2012b. |
|
|
|
|
|
|