A turbulent, high magnetic Reynolds number experimental model of Earth's core

Abstract

We present new experimental results from the University of Maryland Three Meter Geodynamo experiment. We drive a fully turbulent flow in water and also in sodium at magnetic Reynolds number Rm = ΔΩ(r o − r i ) 2 ∕ , up to 715 (about half design maximum) in a spherical Couette apparatus geometrically similar to Earth's core. We have not yet observed a self-generating dynamo, but we study MHD effects with an externally applied axisymmetric magnetic field. We survey a broad range of Rossby number −68 < Ro = ΔΩ∕Ω o < 65 in both purely hydrodynamic water experiments and sodium experiments with weak, nearly passive applied field. We characterize angular momentum transport and substantial generation of internal toroidal magnetic field (the Ω effect) as a function of Ro and find a rich dependence of both angular momentum transport and Ω effect on Ro. Internal azimuthal field generation peaks at Ro = 6 with a gain as high as 9 with weak applied field. At this Rossby number, we also perform experiments with significant Lorentz forces by increasing the applied magnetic field. We observe a reduction of the Ω effect, a large increase in angular momentum transport, and the onset of new dynamical states. The state we reach at maximum applied field shows substantial magnetic field gain in the axial dipole moment, enhancing the applied dipole moment. This intermittent dipole enhancement must come from nonaxisymmetric flow and seems to be a geodynamo-style feedback involving differential rotation and large-scale drifting waves.

Rotating convective turbulence in Earth and planetary cores

“…Given accurate field models, it is essential to generate advanced models of the essential turbulent processes and dynamo physics occurring in liquid metals. Efforts are being made in this direction via laboratory experiments (e.g., Cabanes et al, 2014;Zimmerman et al, 2014;Ribeiro et al, 2015). To computationally access turbulent liquid metal dynamo action, numerical models will need to be run on tens to hundreds of thousands of cores.…”

Scaling Laws in Rayleigh‐Bénard Convection

“…For the geodynamo, the liquid iron outer core is characterized by Pm = 10 −6 , indicating a transitional Re ≈ 10 8 . Invariably, despite many heroic efforts, numerical simulations must substantially increase Pm to achieve dynamo action and laboratory experiments have yet to either produce dynamo action (Lathrop & Forest, 2011;Zimmerman et al, 2014) or sustain a large-scale dynamo (Berhanu et al, 2010). Thus, the dynamo process is highly complex and the intimate details such as the impact of field strength remain to be elucidated (Fauve & Petrelis, 2003;Tobias, 2019).…”

Dynamo generated by the centrifugal instability

“…Such spherical Taylor-Couette dynamo also opens interesting perpectives for experimental flows: MHD spherical Couette experiments ( [39,40]) are currently based on aspect-ratios similar to the Earth's core, thus forbidding the occurence of the centrifugal instability. Although the smallest Rm reported in this paper (Rm min ∼ 150) is still very challenging from an experimental point of view, one may imagine to use a smaller gap in these experiments, more prone to the generation of Taylor vortices in the equatorial plane.…”