Rotating, hydromagnetic laboratory experiment modelling planetary cores

Abstract

This dissertation describes a series of laboratory experiments motivated by planetary cores and the dynamo effect, the mechanism by which the flow of an electrically conductive fluid can give rise to a spontaneous magnetic field. Our experimental apparatus, meant to be a laboratory model of Earth's core, contains liquid sodium between an inner, solid sphere and an outer, spherical shell. The fluid is driven by the differential rotation of these two boundaries, each of which is connected to a motor. Applying an axial, DC magnetic field, we use a collection of Hall probes to measure the magnetic induction that results from interactions between the applied field and the flowing, conductive fluid. We have observed and identified inertial modes, which are bulk oscillations of the fluid restored by the Coriolis force. Over-reflection at a shear layer is one mechanism capable of exciting such modes, and we have developed predictions of both onset boundaries and mode selection from over-reflection theory which are consistent with our observations. Also, motivated by previous experimental devices that used ferromagnetic boundaries to achieve dynamo action, we have studied the effects of a soft iron (ferromagnetic) inner sphere on our apparatus, again finding inertial waves. We also find that all behaviors are more broadband and generally more nonlinear in the presence of a ferromagnetic boundary. Our results with a soft iron inner sphere have implications for other hydromagnetic experiments with ferromagnetic boundaries, and are appropriate for comparison to numerical simulations as well. From our observations we conclude that inertial modes almost certainly occur in planetary cores and will occur in future rotating experiments. In fact, the predominance of inertial modes in our experiments and in other recent work leads to a new paradigm for rotating turbulence, starkly different from turbulence theories based on assumptions of isotropy and homogeneity, starting instead with inertial modes, which are the linear eigenmodes of any rapidly rotating fluid.