University:
Widener University
Course:
CHEM-255 | Organic Chemistry I
Academic year:

2023
Views:

77

Pages:

10
Author:

Pablo Anthony

Development of anisotropic structure in the Earth’s lower mantle by solid-state convection Is there evidence for the localization of dislocation creep in the lowermost mantle? • As we discussed last class, LPO of (Mg,Fe)O may be a viable candidate mechanism for D’’ anisotropy. • But: LPO requires a “kick” from the diffusion creep regime (most of the lower mantle) to the dislocation creep regime. • This paper: implement a numerical model for deformation due to a downgoing slab. Is there evidence for this transition from the model? A quick review: diffusion creep vs. dislocation creep • DIFFUSION CREEP – – – – – – High T Small grain size Low stress Low strain rate Newtonian viscosity DOES NOT result in LPO or (macroscopic) seismic anisotropy – DOES erase previous LPO. • DISLOCATION CREEP – – – – – – Low T Large grain size High stress High strain rate Non-Newtonian viscosity DOES result in LPO and (macroscopic) seismic anisotropy (IF STRAINS HIGH ENOUGH, >100200%) Model Details (I) (Details mostly from McNamara et al. [2001]) Utilizes a composite rheology incorporating both diffusion & dislocation creep, with: Model Details (II) The physics: conservation of mass, momentum, and energy in the extended Boussinesq approximation. Solve w/ finite element code in 2D cylindrical geometry. Model Details (III) Fixed parameters: Play around with: lower mantle dislocation creep activation coefficient, viscosity magnitude, transition stress, strength of slab… “Acceptable” class of models has lower mantle dominated by diffusion creep. Calculating strain/LPO • Use strain as a proxy for development of LPO, and track strains only in regions dominated by dislocation creep. • Calculate Lagrangian finite strain (as a post-processing step) by time-integrating the DGT (deformation gradient tensor) for individual strain tracers. • When a tracer leaves the dislocation creep regime, diffusion creep destroys LPO. So, the main features of the model results are… • Deformation in slab is dominated by dislocation creep in a mantle otherwise dominated by diffusion creep… • Directly above CMB: high-magnitude, laterally-directed strain. Consistent feature of nearly all models. • Details of strain field are time-dependent and can be quite complicated. Integration with mineral physics data… • The authors consider (Mg,Fe)O to be a more likely candidate than (Mg,Fe)-pvskite. [(Mg,Fe)O is weaker phase & has higher intrinsic anisotropy.] • Steady-state (strains > 400-500%) LPO develops in (Mg,Fe)O; horizontal shear will result in VSH > VSV. Conclusions • Slabs are characterized by high stress, resulting in deformation dominated by dislocation creep. • Complicated strain fields result, but one consistent feature is a large degree of laterally directed strain directly above the CMB. • Combined with mineral physics experiments on (Mg,Fe)O, predict VSH > VSV anisotropy (consistent with seismological results for paleoslab regions). • LPO is a likely candidate mechanism for D’’ anisotropy in slab regions. • Although other processes may contribute to the formation of anisotropy, they are not required, and solidstate processes within a homogenous material may suffice.