Modeling phase separation and phase change for magma ocean solidification dynamics

Abstract Just after accretion, the Earth's mantle was significantly molten by the heat dissipation due to large impacts and to the segregation of the core. The mineralogical observations and thermodynamics models of solid‐liquid equilibrium of silicates show that several types of crystallization may have happened at different depths in the mantle. Solids were probably formed first at the bottom of the lower mantle or at midmantle leaving two possible magma oceans, a shallow one and an abyssal one. Near the bottom of the mantle, the liquid phase might become denser than solids due to iron enrichment. In the shallow magma ocean, the crystallizing solid phase was denser and sank through the magma to settle and compact at depth. To understand these complex dynamics, we develop a two‐phase numerical code that can handle simultaneously convection in each phase and in the slurry, and the compaction or decompaction of the two phases. Although our code can only run in a parameter range (Rayleigh number, viscosity contrast between phases, Prandlt number) far from what would be realistic, we think it already provides a rich dynamics that illustrates what could have happened. We show situations in which the crystallization front is gravitationally stable and situations where the newly formed solids are gravitationally unstable and can snow across the magma. Our study suggests that the location of a density contrast between solid and magma must be considered of equal importance with that of the intersection between liquidus and isentrope for what concerns mantle solidification.