Optimizing the flow in a liquid sodium dynamo experiment

The Madison Dynamo experiment drives a turbulent flow of liquid sodium in a sphere in order to observe a MHD dynamo instability: An exponentially growing magnetic field at the expense of kinetic energy. Initial runs of the experiment observed intermittent bursts of the predicted magnetic mode, but no self-excited field was observed. It was found that turbulent fluctuations were producing large-scale magnetic fields that were a significant fraction of the magnitude of the fields induced by the mean flow. These turbulent-induced fields were solely detrimental, opposing the generation of the magnetic field produced by the mean flow. Baffles and vanes were added to the experiment to optimize the helical pitch of the mean flow and to remove the large-scale detrimental fluctuations. The observed drop in required motor power and a drop in specific measured magnetic response modes gives direct confirmation that these large detrimental eddies have been removed. A probe was developed to characterize the turbulence in the MDE after the baffles were installed and it was determined that the remaining turbulent EMF was mostly acting as an enhanced dissipation to the induced magnetic field. After these modifications, the induced magnetic field produced by the flowing sodium interacting with a seed magnetic field now closely matches laminar predictions. However, no self-excited field has been observed. A velocity inversion technique has been developed that compares internal and external field measurements with a predictive model and determines what the effective mean flow is in the experiment. Results from this velocity inversion give another metric on how optimized the flow profile is and also provide the most robust method of determining how close the experiment is to achieving a dynamo.