The overall driving mechanism of fluid flow in the inner Earth is convection in the gravitational buoyancy field of our planet. In particular there has been involved a lot of effort in theoretical prediction and numerical simulation of both the geodynamo [1-2], which is maintained by convection, and mantle convection, which is the main cause for plate tectonics [3-4]. All the joined research objectives use the methods of theory, numerical simulation and experiments. While the theory determines the basic concepts, the numerical simulation is able to check approximations and modelling concepts for a rich variety of parameters. Finally there are the experiments, which allow capturing all non-linear effects and associated instabilities without analytical and numerical simplifications.
With the rather simplified experiment ‘GeoFlow’ (Geophysical Flow Simulation) instability and transition of hydrodynamic convection in spherical shells are traced. In the specific focus here is the realization of a so called self-gravitating spherical shell set-up [1,5]. More precisely, the fluid motion in a gap between two concentric spheres is observed, with the inner spherical shell heated and the outer spherical shell cooled. A high voltage potential between the inner and outer spheres together with the use of a dielectric working fluid induces an electro-hydrodynamic force, which is in analogy to the gravitational buoyancy force inside the Earth. To reduce unwished directed gravity this experiment requires microgravity conditions. We refer to [6-8] for scientific background and application of this technique in spherical shell experiments.
The ‘GeoFlow I’ experiment was accomplished on the International Space Station's module COLUMBUS inside the Fluid Science Laboratory FSL. Special goal of that experiment was to capture the large-scale convection without as well as with rotation. For this spherical Rayleigh-Bénard convection the working fluid was silicone oil, for which kinematic viscosity is approximately constant. In contrast, for the second experiment, named ‘GeoFlow II’, we use an alkanole with pronounced temperature-dependent physical properties, e.g. viscosity and thermal volume expansion. This topic is of high interest in mantle convection studies. Thus the purpose of ‘GeoFlow II’ is to observe the basic properties of the flow in a non-linear regime, by achieving the maximal viscosity variation allowed with the hardware limitation inside the Fluid Science Laboratory.
 F. Busse, Phys. Fluids, 14, 4 (2002).
 G. Schubert, P. Olson, Treatise on Geophysics - Core Dynamics, Elsevier, 2009.
 M. Ogawa, Fluid Dyn. Res., 40 (2008).
 G. Schubert, D. Bercovici, Treatise on Geophysics - Mantle Dynamics, Elsevier, 2009.
 I. Yavorskaya, N. Fomina, Y. Balyaev,, Acta Astronaut., 11 (1984).
 J. Hart, G. Glatzmaier, J. Toomre, J. Fluid Mech., 173 (1986).
 B. Futterer, M. Gellert, T. von Larcher, C. Egbers, Acta Astronaut., 62 (2008).
 B. Futterer, C. Egbers, N. Dahley, S. Koch, L. Jehring, Acta Astronaut., 66 (2010).