Boris and Exponential Integrators in the Theory of Particles Interacting with Magnetic Turbulence
A. Shalchi · 2026
Read Boris and Exponential Integrators in the Theory of Particles Interacting with Magnetic Turbulence by A. Shalchi (2026) 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-2605-14065.txt.
Introduction
Shalchi studies exponential integrators and the Boris integrator for charged bits moving through magnetic turbulence. Such motion matters for fast bits in magnetized plasma, and test-particle runs explore it by building a turbulent field and solving the Newton-Lorentz equation step by step. The paper derives both the Rodrigues scheme and the famous Boris integrator step by step from exponential integrators. When the field is rebuilt fresh at each step, the two perform much alike overall, though theory favors Rodrigues for exactness.
Catalog record
| CREATOR | A. Shalchi |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 53188 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.14065
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Boris and Exponential Integrators in the Theory of Particles Interacting with Magnetic Turbulence A. Shalchi Abstract The interaction of electrically charged particles with magnetic fields is a fundamental problem in several areas of physics. An example is the motion of energetic particles through a magnetized plasma. The most accurate and reliable way to explore theoretically the interactions between particles and fields is via test-particle simulations. In such simulations one creates the turbulent magnetic field and solves the Newton-Lorentz equation numerically by employing an integration scheme. In the current article we discuss exponential integrators and derive systematically from this the Rodrigues scheme as well as the famous Boris integrator. For an approach where one creates the magnetic field anew at each time step, both integrators are overall comparable. In theory the Rodrigues approach should be more accurate due to the fact that the occurring matrix exponential is evaluated without further approximations. Practically, both methods provide very similar results. It is argued in the current article that a Rodrigues based integrator is a very strong alternative because for the specific problem discussed here, it does not require longer computing times. keywords diffusion – magnetic fields – turbulence 1 Introduction The motion of electrically charged energetic particles in a plasma is difficult to describe analytically due to complicated interactions between the particles and turbulent magnetic fields. However, this problem is fundamental and the knowledge of particle diffusion coefficients is important in several areas of research. The most prominent examples are studies of solar modulation and diffusive shock acceleration (see, e.g., Zank et al. (2000), Li et al. (2003), Li et al. (2005), Zank et al. (2006), Dosch & Shalchi (2010), Li et al. (2012), Ferrand et al. (2014), Hu et al. (2017), Qin et al. (2018), Shen & Qin (2018), Moloto & Engelbrecht (2020), Engelbrecht & Wolmarans (2020), Engelbrecht & Moloto (2021), Shen et al. (2021), and Ngobeni et al. (2022)). Ideally, one describes the motion of energetic particles through a plasma by using analytical theory. However, this task is difficult due to the nonlinearity of the problem (see Shalchi (2009), Shalchi (2020), and Effenberger et al. (2025) for reviews). Especially in analytical treatments of the transport, one has to distinguish between transport along and across the mean magnetic field. Parallel diffusion is controlled by pitch-angle scattering and, associated with this, a pitch-angle isotropization process. Perpendicular transport, on the other hand, depends on the structure of magnetic field lines because initially particles follow individual lines. As soon as the transverse complexity of the turbulence becomes manifested, particles leave their original field lines and the transport become diffusive. In the past it was often assumed that perpendicular transport is more difficult to describe analytically, but significant progress has been achieved during the past two decades (see, e.g., Matthaeus et al. (2003), Shalchi (2010), and Shalchi (2021)). An alternative to analytical theories is to perform test-particle simulations. In such simulations the turbulence is created artificially by using a certain spectrum in combination with random numbers. For a given turbulence model, particle orbits can then be obtained by solving the Newton-Lorentz equation numerically. Considering and averaging thousands of particle trajectories allows one to determine the diffusion coefficients in the different directions of space. The advantage of simulations compared to analytical theories is obvious: while theories require to make several assumptions which are often difficult to justify, the simulations are based on an exact (besides numerical inaccuracies) particle orbits. A disadvantage of the simulations is, of course, that they cannot be implemented in applications such as diffusive shock acceleration. In the past, test-particle simulations have been performed by numerous authors in the context of space physics and astrophysics (see, e.g., Qin et al. (2002a), Qin et al. (2002b), Qin et al. (2006), Zimbardo et al. (2006), Perri & Zimbardo (2007), Pommois et al. (2007), Perri & Zimbardo (2009a), Perri & Zimbardo (2009b), Tautz (2010), Zimbardo et al. (2012), Hussein et al. (2015), Hussein & Shalchi (2016), Heusen & Shalchi (2016), Ivascenko et al. (2016), Reichherzer et al. (2022), Snodin et al. (2022), and Els & Engelbrecht (2024)). In the aforementioned simulations different methods have been applied. For instance, there are grid-based simulations in which the turbulent magnetic field is computed over a whole grid and thereafter the particle orbits are determined. In other simulation codes the magnetic field is only computed at the current particle position. Both methods have advantages and disadvantages is terms of memory needs and run times. In either case, one has to solve the Newton-Lorentz equation numerically. It became clear over years that one of the best possible integration schemes to solve the equations of motion is the so-called Boris integrator (see Boris (1970)). In the current article we review this type of integration scheme and derive it systematically from exponential integrators. As alternative we explore the combination of such exponential integrators with the so-called Rodrigues formula (see Rodrigues (1840)). We compare both integration schemes with the aim to see which one is better suited for test-particle simulations. By doing this we focus on the very specific case of a purely magnetic scenario where particles interact with a constant mean magnetic field and a turbulent component. Furthermore, we consider simulations in which the magnetic field is computed anew at each particle position. We shall demonstrate that, for this special case, both integrator schemes, Boris as well as Rodrigues, are strong tools with a slight advantage of the latter approach due to more accuracy in the phase. A comprehensive comparison of relativistic particle integrators can be found in Ripperda et al. (2018). However, in the current article we focus on the very specific problem of particles interacting with a purely magnetic field which consists of a constant mean field and a turbulence component. Furthermore, the turbulent component is created in specific way as described later in this article. 2 Equations of Motion The fundamental equation describing the motion of an electrically charged particle in a purely magnetic system is the Newton-Lorentz equation \frac{d\vec{p}}{dt}=\frac{q}{c}\vec{v}\times\vec{B}(\vec{x}) (1) where we have used the relativistic momentum \vec{p}=m\gamma\vec{v} . In the latter relation, as well as in Eq. ( 1 ), we have used the particle charge q , the vacuum speed of light c , the particle velocity \vec{v} , the particle’s rest mass m , and the Lorentz factor \gamma . Furthermore, we assume that the total magnetic field is given by \vec{B}(\vec{x})=\delta\vec{B}(\vec{x})+B_{0}\hat{z} where \delta\vec{B}(\vec{x}) describes the magnetic fluctuations (the turbulence) and B_{0} is the constant mean field pointing in the z -direction. Note, here we have used cgs units rather than SI units . The three components of Eq. ( 1 ) can easily be determined and one gets \displaystyle\dot{v}_{x} \displaystyle= \displaystyle\Omega v_{y}+\Omega\left(v_{y}\frac{\delta B_{z}}{B_{0}}-v_{z}\frac{\delta B_{y}}{B_{0}}\right), \displaystyle\dot{v}_{y} \displaystyle= \displaystyle-\Omega v_{x}+\Omega\left(v_{z}\frac{\delta B_{x}}{B_{0}}-v_{x}\frac{\delta B_{z}}{B_{0}}\right), \displaystyle\dot{v}_{z} \displaystyle= \displaystyle\Omega\left(v_{x}\frac{\delta B_{y}}{B_{0}}-v_{y}\frac{\delta B_{x}}{B_{0}}\right). (2) In the latter equations of motion we have used the gyro-frequency \Omega=\frac{qB_{0}}{mc\gamma} (3) which is the angular velocity of the rotation if there was only a mean magnetic field B_{0} and no turbulence. Therefore, we also call \Omega the unperturbed gyro-frequency . Note, for negative q the gyro-frequency defined here becomes negative. This simply means that the particle reverses its direction of rotation. The three components listed via Eq. (