2D numerical modelling of fluid and melt percolation in the subduction zone

Subducting slab dehydration and resulting aqueous fluid percolation triggers partial melting in the mantle wedge and is accompanied with the further melt percolation through the porous space to the region above the slab. This problem is a complex coupled chemical, thermal and mechanical process responsible for the magmatic arcs formation and change of the mantle wedge properties. We have created a two-dimensional model of a two-phase flow in a porous media solving a coupled Darcy-Stokes system of equations for two incompressible media for the case of visco-plastic rheology of solid matrix. Our system of equation is expanded for the high-porosity limits and stabilized it for the case of high porosity contrasts. Melting process is implemented according to the model of Katz (2003) where melting degree is a function of pressure, temperature, composition and water content. We use a finite-difference method with fully staggered grid in a combination with marker-in-cell technique for advection of fluid and solid phase. We performed a comparison with a benchmark of a thermal convection in a porous media in a bottom-heated box to verify the interdependency of Rayleigh and Nusselt numbers with a theoretical one. We have demonstrated the stability and robustness of the algorithm in case of strongly non-linear visco-plastic rheology of solid including cases with localization of both deformation and porous flow along spontaneously forming shear bands. We have checked our model for the forming of localized porous channels under a simple shear stress (channelling instability). Current work includes implementation of non-liner viscous rheology and elaboration on the setup of self-initiating subduction. Later we plan to include solid elasticity and fluid/solid compressibility. Also we have developed a full complexity system of equations for visco-elastic case and currently are working on numerical realisation of it to verify our simplifying assumptions for the general model. Ultimate goal is to simulate in a realistic self-consistent manner fluid and melt generation and transport in subduction zones including fluid/melt focussing phenomena above slabs.