Turbulent convection in the Sun: modeling in unstructured meshes

TL;DR

This paper explores modeling solar turbulent convection using unstructured meshes with the CharLES solver, highlighting challenges and potential for more realistic simulations at higher resolutions.

Contribution

It demonstrates the application of unstructured mesh hydrodynamics to solar convection and assesses the feasibility of high-resolution simulations with current computational resources.

Findings

Coarse mesh fails to reproduce realistic convection patterns.

Finer mesh shows cooling effects due to thermal conduction.

Realistic simulations feasible with increased computational power.

Abstract

We adopted an unstructured hydrodynamical solver CharLES to the problem of global convection in the Sun. With the aim to investigate the properties of solar turbulent convection and reproduce differential rotation pattern. We performed simulations in two spherical shells, with 1.3 and 10 million cells. In the first, coarse mesh, the solution does not reproduce realistic convection, and is dominated by numerical effects. In the second mesh, thermal conduction leads to cooling of bottom layers, that could not be compensated by solar irradiance. More simulations in the 10M cells mesh should be performed to investigate the influence of transport coefficients and numerical effects. Our estimate of the code performance suggests, that realistic simulations in even finer grids could be performed for reasonable computational cost.