École Nationale Supérieure de
Techniques Avancées

Résumé

We present the extension of physics-based numerical diffusion diagnostics in advection schemes that were developed in oceanography. These diagnostics use a quadratic quantity (variance of a scalar distribution, kinetic energy density) as a measure of numerical diffusion, and separate transport and diffusion effects of numerical advection thanks to physical arguments of reversibility. We also consider the extension to astrophysical contexts by considering the compressibility of the flows. We present the first tests of these diagnostics in Implicit Large Eddy Simulations(ILES), using the MUSIC code, designed to solve compressible hydrodynamic convection equations in planets and stars. We then consider the internal structure dynamics in giant planets, which were found to be structured in layers of different chemical compositions thanks to new observations. We present first results of double-diffusive convection simulations using MUSIC, with a twofold objective : initiate a study of giant planets’ interior dynamics and consider a first application of the numerical diffusion diagnostics.