AtmoFlow: Thermo-electrohydrodynamic convection in the thermally driven spherical shell with differential rotation

In a geophysical point of view, large scale prediction of atmospheric flows become more and more important to forecast e.g. extreme weather conditions that are observed in recent days more frequent and may relate to the overall climate change.  The AtmoFlow experiment is a small scaled laboratory spherical shell to investigate such atmospheric flow fields in a miniaturized model of a planet. Besides this physical experiment, numerical simulations are performed to analyze the resulting convective patterns in more detail. The experiment is composed of two spherical shells, which can rotate independently. The temperature on the shell's surface can be defined as a heated equator and a cooled pole. To model the terrestrial gravitation, an electric potential is applied on a dielectric fluid confined between the shells. This so called dielectrophoretic force triggers the formation of buoyant patterns, and is in fact the artificial equivalent of terrestrial gravitation. 

The simulations are processed with a custom programmed solver in the OpenFOAM ecosystem. It covers all predefined rotation combinations starting with no rotation, solid body rotation and differential rotation. The latter are the latest results of the computational simulation campaign and used to investigate the influence of the artificial central force field to differential rotation. While differential rotation can cause the well known Taylor vortices, it has to be noted that the cell formation induced by the central dielectrophoretic force field maybe significantly extubated and thus new convection patterns may arise.  This in fact is the overall focus of this study to understand the underlying physical process of such pattern formations. The analyses focus first on the amount of convective heat that these patterns are able to transport, and second to quantify their shape and intensity via a spatial Fast Fourier Transformation to identify the most dominant structures. Finally, statistical moments will provide an estimation about the shape and location of vacillating patterns.