[Paper Review] The Nonlinear Turbulent Dynamo

Jason Maron, S. C. Cowley|ArXiv.org|Nov 1, 2001

Solar and Space Plasma Dynamics4 references3 citations

TL;DR

This paper investigates the nonlinear turbulent dynamo in high magnetic Prandtl number plasmas using direct numerical simulations, showing that magnetic fields grow via a kinematic phase with a $k^{3/2}$ spectrum until magnetic energy approaches viscous-scale kinetic energy, after which backreaction slows growth until saturation when magnetic and kinetic energies are equal at the resistive scale. The results reveal fundamental differences between high- and low-Prandtl-number dynamos, with implications for galactic magnetic field amplification.

ABSTRACT

We simulate the evolution of an initially weak magnetic field in forced turbulence for a range of Prandtl numbers. The field grows exponentially with the Kulsrud-Anderson $k^{3/2}$ spectrum until the magnetic energy approaches the viscous-scale kinetic energy, where the magnetic forces then backreact on the velocity. Further growth proceeds more slowly until a saturated state is reached where the magnetic and kinetic energies are equal, and where the magnetic energy exists primarily at the resistive scale. We discuss the structure of this turbulence and the extrapolation of the results to astrophysically-large Prandtl numbers.

Motivation & Objective

• To understand the nonlinear evolution of magnetic fields in high-Prandtl-number turbulence, relevant to the interstellar medium and galaxy formation.
• To investigate how magnetic feedback on velocity fields alters dynamo growth beyond the kinematic regime.
• To determine whether the saturated state of the dynamo reaches equipartition between magnetic and kinetic energy, as observed in galaxies.
• To clarify the differences between high-Prandtl-number and low-Prandtl-number dynamo mechanisms, especially in astrophysical contexts.

Proposed method

• Direct numerical simulations of forced turbulence with varying magnetic Prandtl numbers ($P_r$) to model astrophysical conditions.
• Use of spectral methods with Fast Fourier Transforms (FFTs) to compute nonlinear terms in the MHD equations efficiently.
• Application of 3/2-dealiasing (truncating at $|s| < N/3$) for MHD terms and 1/6-dealiasing ($|s| < N/6$) for the Braginskii term to minimize aliasing errors.
• Implementation of a two-grid approach: $N^3$ for MHD terms and $(2N)^3$ for the Braginskii term to maintain accuracy while reducing computational cost.
• Use of random-phase transformations to compare coherent structures with randomized counterparts, preserving power spectra but removing coherent features.
• Validation of numerical accuracy by comparing results from smaller grids with those from larger grids, ensuring convergence.

Experimental results

Research questions

• RQ1 How does the magnetic field grow in high-Prandtl-number turbulence when nonlinear feedback becomes significant?
• RQ2 Does the saturated state of the dynamo achieve equipartition between magnetic and kinetic energy, as observed in galaxies?
• RQ3 What is the spectral structure of the magnetic field at saturation, and does it peak at the resistive scale as predicted?
• RQ4 How do the dynamics of high-$P_r$ dynamos differ from those in low-$P_r$ plasmas, particularly in the nonlinear regime?
• RQ5 To what extent do numerical artifacts like aliasing affect the simulation of complex terms such as the Braginskii viscosity?

Key findings

• The magnetic field grows with a $k^{3/2}$ spectrum during the kinematic phase, consistent with theoretical predictions, until magnetic energy reaches the viscous-scale kinetic energy level.
• At this point, magnetic forces begin to backreact on the velocity field, slowing further growth and marking the end of the kinematic phase.
• Further growth proceeds slowly until a saturated state is reached where magnetic and kinetic energies are equal, indicating equipartition.
• In the saturated state, the magnetic energy is concentrated at the resistive scale, not at the viscous or large-scale forcing scales.
• The simulation results show that high-Prandtl-number dynamos are fundamentally different from low-Prandtl-number ones, with distinct spectral and energetic characteristics.
• Dealiasing at $N/6$ for the Braginskii term is necessary for numerical accuracy, as $N/3$ leads to significant aliasing errors, especially in small-scale field structures.