Comparative Analysis of Hall Effect Implementations in Hall-Magnetohydrodynamics
Kazunari Iwasaki, Kengo Tomida · 2025
Read Comparative Analysis of Hall Effect Implementations in Hall-Magnetohydrodynamics by Kazunari Iwasaki, Kengo Tomida (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-14842.txt.
Introduction
Kazunari Iwasaki and Kengo Tomida compare ways to put the Hall effect into plasma simulations without blowing them up. Their paper from twenty twenty-five works in Hall magnetohydrodynamics, where a Hall electric field joins the standard set. A naive add-on triggers numeric instability. Two cures from the literature face off in their tests.
One cure reshapes a Riemann solver so whistler wave speeds set the signal speeds. The other adds a fourth-order hyper-resistivity term to damp the trouble. Wide test runs favor the hyper-resistivity route on accuracy. The gap grows when whistler waves tick faster than the physics of interest. A von Neumann stability study plus experiments fix a sound size for the added term. The paper gives plasma coders a tested default for Hall runs.
Catalog record
| CREATOR | Kazunari Iwasaki, Kengo Tomida |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 72 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-14842
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Comparative Analysis of Hall Effect Implementations in Hall-Magnetohydrodynamics Kazunari Iwasaki, Kengo Tomida Abstract There is no standard numerical implementation of the Hall effect, which is one of the non-ideal magnetohydrodynamic (MHD) effects. Numerical instability arises when a simple implementation is used, in which the Hall electric field is added to the electric field to update magnetic fields without further modifications to the numerical scheme. In this paper, several implementations proposed in the literature are compared to identify an approach that provides stable and accurate results. We consider two types of implementations of the Hall effect. One is a modified version of the Harten-Lax-van Leer method (Hall-HLL) in which the phase speeds of whistler waves are adopted as the signal speeds; the other involves adding a fourth-order hyper-resistivity to a Hall-MHD code. Based on an extensive series of test calculations, we found that hyper-resistivity yields more accurate results than the Hall-HLL, particularly in problems where the whisler-wave timescale is shorter than the the timescale of physical processes of interest. Through both von Neumann stability analysis and numerical experiments, an appropriate coefficient for the hyper-resistivity is determined. Athena++ [ 35 ] , numpy [ 38 ] , Matplotlib [ 19 ] I Introduction The Hall effect is one of the non-ideal magnetohydrodynamic (MHD) effects. It occurs in situations where electrons are coupled to magnetic fields, while ions are decoupled. The characteristics of the Hall effect differ between fully ionized and weakly ionized gases [ 33 ] . An intriguing property of the Hall effect is that it modifies the dispersion relation of linear MHD waves, leading to generation of whistler and ion-cyclotron waves. In the long-wavelength limit, these waves transition into Alfvén waves. The phase speed of whistler waves increases proportionally to the wavenumber, whereas that of ion-cyclotron waves remains constant at higher wavenumbers. Numerically implementing the Hall effect presents significant challenges, unlike other non-ideal MHD effects (Ohmic resistivity and ambipolar diffusion), which are relatively simple to implement. One reason is that the Hall effect causes waves to be dispersive. In particular, as whistler waves with shorter wavelengths propagate faster, grid-scale disturbances in magnetic fields can become significant under certain conditions. A simple implementation of the Hall effect involves adding the Hall electric field to the electric field to update the magnetic field without further modifications to the numerical scheme. Falle [11] conducted a von Neumann stability analysis of such a naive implementation and found it to be unconditionally unstable when using a first-order explicit time integrator (forward Euler integrator). The stability of this simple implementation depends on the accuracy of the time integrator [ 20 ] . While third-order time integrators are conditionally stable, second-order time integrators may lead to numerical instability. Lesur et al. [22] confirmed that a third-order Runge-Kutta (RK3) time integrator suppresses numerical instabilities in the nonlinear development of Hall-dominated magnetorotational instability using SNOOPY , an incompressible pseudo-spectral code. However, no studies have shown that the RK3 integrator, combined with the simple implementation, provides stable results when using Godunov-type schemes with the constrained transport (CT) methods [ 10 ] , which is a popular combination used in many MHD simulation codes, such as Stone et al. [35, Athena++] , Fromang et al. [12, RAMSES] , Mignone et al. [27, Pluto] , Collins et al. [7, Enzo] . Another issue with the simple implementation is that it does not provide numerical dissipation because the Hall electric field is oriented perpendicular to the electric current. This lack of numerical dissipation causes serious problems in magnetic reconnection driven by the Hall effect [ 24 , e.g.,] . Several numerical methods have been proposed to suppress numerical instabilities caused by the Hall effect. Tóth et al. [37] and Lesur et al. [22] suggested modifying the signal speeds in the Harten-Lax-van Leer [ 18 , HLL,] numerical fluxes by considering the phase speed of whistler waves to estimate the signal speeds of the two characteristics. This method is called Hall-HLL and is widely used in numerical simulations of star and planet formation [ 3 , 2 , 25 ] . Marchand et al. [26] proposed a modified Hall-HLL method to improve the conservation of angular momentum in collapsing dense molecular cloud cores. Alternative approaches for modifying time integration methods have also been proposed. Falle [11] demonstrated that an implicit method stabilizes Hall-MHD. Furthermore, Tóth et al. [37] showed that an implicit Hall-MHD scheme is stable even without modifications to the numerical fluxes. O’Sullivan & Downes [30] , O’Sullivan & Downes [31] found that the Hall-MHD stability is achieved when the magnetic field is updated using a dimensionally split method [ 1 , also see] . Another approach involves introducing artificial resistivity into the induction equation. For instance, hyper-resistivity is added to damp whistler waves with wavelengths comparable to the grid scale in numerical simulations of magnetic reconnection involving the Hall effect [ 4 , 23 , 5 , 40 , e.g.,] . However, an appropriate choice of the hyper-resistivity coefficient has not been thoroughly examined. Recently, Zier et al. [42] proposed a method in which the diffusion coefficient of Ohmic resistivity is artificially increased to stabilize the schemes. As mentioned earlier, various implementations of the Hall effect have been proposed in the literature. However, comprehensive comparisons of their stability and accuracy have not yet been conducted. In this paper, we compare the results of Hall-HLL [ 22 ] , a modified version of Hall-HLL [ 26 ] , and hyper-resistivity [ 4 ] . This paper is organized as follows: Section II reviews the basic properties of Hall-MHD and describe the implementations considered in this paper. Section III presents numerical experiments. II Basic Properties of Hall-MHD and Its Numerical Implementations II.1 Basic Equations The basic equations of non-ideal MHD are given by \frac{\partial\rho}{\partial t}+\frac{\partial\rho v_{i}}{\partial x_{i}}=0, (1) \frac{\partial\rho{v}_{i}}{\partial t}+\frac{\partial}{\partial x_{j}}(\rho v_{i}v_{j}+{\cal T}_{ij})=0, (2) \frac{\partial E}{\partial t}+\frac{\partial}{\partial x_{i}}\left\{Ev_{i}+{\cal T}_{ij}v_{j}+(\bm{E}_{\mathrm{ni}}\times\bm{B})_{i}\right\}=0, (3) and \frac{\partial\bm{B}}{\partial t}+\bm{\nabla}\times(-\bm{v}\times\bm{B}+\bm{E}_{\mathrm{ni}})=0, (4) where \rho is the density, \bm{v} is the velocity, \bm{B} is the magnetic field, P is the pressure, and E=\rho\bm{v}^{2}/2+P/(\gamma-1)+\bm{B}^{2}/8\pi represents the total energy per unit volume. {\cal T} denotes the stress tensor, given by {\cal T}_{ij}=\left(P+\frac{B^{2}}{8\pi}\right)\delta_{ij}-\frac{B_{i}B_{j}}{4\pi}. (5) \bm{E}_{\mathrm{ni}} represents the electric field resulting from non-ideal MHD effects, expressed as \bm{E}_{\mathrm{ni}}=\frac{4\pi}{c}\left(\eta_{\mathrm{O}}\bm{J}+\eta_{\mathrm{H}}\frac{\bm{J}\times\bm{B}}{|\bm{B}|}+\eta_{\mathrm{A}}\bm{J}_{\perp}\right), (6) \bm{J}=(c/4\pi)\bm{\nabla}\times\bm{B} is the electric current density, and \bm{J}_{\perp}=\bm{J}-(\bm{J}\cdot\bm{B})/|\bm{B}| is the \bm{J} components perpendicular to the local magnetic field direction. \eta_{\mathrm{O}} , \eta_{\mathrm{H}} , and \eta_{\mathrm{A}} correspond to the diffusion coefficients for Ohmic resistivity, the Hall effect, and ambipolar diffusion, respectively. In this study, Ohmic resistivity and ambipolar diffusion are not considered because the focus is on implementations that ensure stability and accuracy even with \eta_{\mathrm{O}}=\eta_{\mathrm{A}}=0 . II.2 Review of Properties of Linear Waves in Hall MHD Although linear Hall-MHD wave propagation tests are widely used to evaluate various methods, most studies focus primarily on incompressible waves propagating along the unperturbed magnetic field. Thus, it remains unclear whether these methods can accurately capture other types of linear waves. In this study, we assess performance of various methods on all linear waves in Section III.4 . This section provides a brief review of the physical properties of the linear waves in Hall-MHD [ 17 , e.g.,] . We consider a uniform, static gas with a density of \rho_{0} and a