Journal of Fluid Mechanics


Sheet-like and plume-like thermal flow in a spherical convection experiment performed under microgravity

B. Futterera1a2 c1, A. Krebsa1, A.-C. Plesaa3, F. Zaussingera1, R. Hollerbacha4, D. Breuera3 and C. Egbersa1

a1 Chair of Aerodynamics and Fluid Mechanics, Brandenburg University of Technology Cottbus, Siemens-Halske-Ring 14, 03046 Cottbus, Germany

a2 Institute of Fluid Dynamics and Thermodynamics, Otto von Guericke Universität Magdeburg, Universitätsplatz 18, 39106 Magdeburg, Germany

a3 Institute of Planetary Research, German Aerospace Center, Rutherfordstrasse 2, 12489 Berlin, Germany

a4 Institute of Geophysics, Earth and Planetary Magnetism Group, ETH Zürich, Sonneggstrasse 5, 8092 Zürich, Switzerland


We introduce, in spherical geometry, experiments on electro-hydrodynamic driven Rayleigh–Bénard convection that have been performed for both temperature-independent (‘GeoFlow I’) and temperature-dependent fluid viscosity properties (‘GeoFlow II’) with a measured viscosity contrast up to 1.5. To set up a self-gravitating force field, we use a high-voltage potential between the inner and outer boundaries and a dielectric insulating liquid; the experiments were performed under microgravity conditions on the International Space Station. We further run numerical simulations in three-dimensional spherical geometry to reproduce the results obtained in the ‘GeoFlow’ experiments. We use Wollaston prism shearing interferometry for flow visualization – an optical method producing fringe pattern images. The flow patterns differ between our two experiments. In ‘GeoFlow I’, we see a sheet-like thermal flow. In this case convection patterns have been successfully reproduced by three-dimensional numerical simulations using two different and independently developed codes. In contrast, in ‘GeoFlow II’, we obtain plume-like structures. Interestingly, numerical simulations do not yield this type of solution for the low viscosity contrast realized in the experiment. However, using a viscosity contrast of two orders of magnitude or higher, we can reproduce the patterns obtained in the ‘GeoFlow II’ experiment, from which we conclude that nonlinear effects shift the effective viscosity ratio.

(Received February 08 2013)

(Revised August 21 2013)

(Accepted September 19 2013)

(Online publication October 29 2013)

Key words

  • Bénard convection;
  • geophysical and geological flows;
  • nonlinear dynamical systems


c1 Email address for correspondence: