The aim of the PhD project is to explore the interactions between processes active at different spatial (and temporal) scales during strain localization and develop coarse-graining techniques to define rheological laws able to produce self-consistently strain localization and, hence, simulate plate tectonics in geodynamical models.
The work will build on the results of the entire ERC RhEoVOLUTION team, that is composed by >15 researchers of varied backgrounds (Geology, Glaciology, Solid and Fluid Mechanics, Material Sciences, Applied Mathematics…), mostly based at Geosciences Montpellier.
Strain localization is the rule rather than the exception in the external layers of the solid Earth. Yet, modelling ductile (viscoplastic) strain localization, the dominant deformation mode in the lithosphere, in which deformation is concentrated in shear zones, remains a challenge >50 years after the scientific revolution that established Plate Tectonics as the main paradigm in Earth Sciences. Analysis of shear zones at various spatial scales in nature and experiments implies that strain localization stems from spatial heterogeneity in the mechanical behavior. This heterogeneity (and the associated strain localization) exists at all scales, is omnipresent at small ones, and evolves in response to the mechanical fields (stress and strain). In the ERC project RhEoVOLUTION, we posited that poor representation of this heterogeneity in mechanical behavior (rheology) of rocks and its evolution during deformation was the locking point for generating strain localization in geodynamical models. Thus, we developed a new approach to examine how strain localization may arise in rocks deforming by ductile processes. This approach associates a stochastic description of the mechanical properties of the medium with simple laws describing how these properties evolve in response to the resulting spatial variations in stress and strain rate. These models successfully produce, from an initially random rheological heterogeneity field, strain localization at scales 2-3 orders of magnitude larger than the characteristic length scale of the initial field. A few shear zones lengthen, coalesce, and widen, dominating the whole system, which evolves towards a new equilibrium, characterized by a bulk anisotropic softening. However, the formulation of these models is strongly based on our knowledge of the processes controlling the rock deformation at the grains scale (µm to cm). Direct extrapolation of the results to the plate tectonics scale (100s to 1000s of km) is therefore not possible.
The PhD will explore questions such as: Does the self-similarity of the strain localization structures – the shear zones (cf. figure) – from the mm to the 100s of km scales justify the use of a single numerical formalism to describe the rheology at all scales? May stochastic approaches, as those developed up to now in the project, be used to describe the submesh behavior of the system at all scales?