A structure-preserving multiscale solver for particle-wave interaction in non-uniform magnetized plasmas
Kun Huang, Irene M. Gamba, Chi-Wang Shu · 2025
Read A structure-preserving multiscale solver for particle-wave interaction in non-uniform magnetized plasmas by Kun Huang, Irene M. Gamba, Chi-Wang Shu (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-2505-20210.txt.
Introduction
Kun Huang, Irene Gamba, and Chi-Wang Shu present a 2025 solver for how particles and waves trade energy in magnetized plasma. Such trades matter for runaway electrons in fusion machines. Plasma waves can be treated as packets with particle-like traits, which yields a paired set of kinetic equations.
Two snags face solvers: uneven plasma makes the system high-sized with many time scales, and the math must keep mass, push, and energy exact. Their solver pairs a local method for the trade step with a path-averaging step for wave flow. Tests on a tube-shaped plasma show all guard laws hold.
Catalog record
| CREATOR | Kun Huang, Irene M. Gamba, Chi-Wang Shu |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 58 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-20210
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A structure-preserving multiscale solver for particle-wave interaction in non-uniform magnetized plasmas Kun Huang, Irene M. Gamba, Chi-Wang Shu Abstract Particle-wave interaction is of fundamental interest in plasma physics, especially in the study of runaway electrons in magnetic confinement fusion. Analogous to the concept of photons and phonons, wave packets in plasma can also be treated as quasi-particles, called plasmons. To model the “mixture” of electrons and plasmons in plasma, a set of “collisional” kinetic equations has been derived, based on weak turbulence limit and the Wentzel-Kramers-Brillouin (WKB) approximation. There are two main challenges in solving the electron-plasmon kinetic system numerically. Firstly, non-uniform plasma density and magnetic field results in high dimensionality and the presence of multiple time scales. Secondly, a physically reliable numerical solution requires a structure-preserving scheme that enforces the conservation of mass, momentum, and energy. In this paper, we propose a struture-preserving multiscale solver for particle-wave interaction in non-uniform magnetized plasmas. The solver combines a conservative local discontinuous Galerkin (LDG) scheme for the interaction part with a trajectory averaging method for the plasmon Hamiltonian flow part. Numerical examples for a non-uniform magnetized plasma in an infinitely long symmetric cylinder are presented. It is verified that the LDG scheme rigorously preserves all the conservation laws, and the trajectory averaging method significantly reduces the computational cost. 1 Introduction The interaction between charged particles and waves is a critical aspect of plasma behavior, which occurs across multiple time scales. The wave frequency in plasmas is extremely high. If one directly simulates the system using the Vlasov–Maxwell equations, the time step of the numerical method must be smaller than the wave period. In order to simulate waves within reasonable computational cost, they are modeled as quasi-particles (plasmons) under WKB approximation [ 4 ] , which have Hamilton’s equations of motion in the phase space. At the same time, their interaction with particles are modeled by the quasilinear theory [ 12 , 5 , 9 ] , which renders two additional collision-like integral-differential operators. It turns out that, such interaction results in diffusion of electrons and reaction of plasmons. For a comprehensive introduction of the model, we refer the readers to Chapter 23 of Thorne & Blandford’s book [ 11 ] . The two components of this model—the Hamiltonian flow of plasmons and the electron-plasmon interaction—give rise to two numerical challenges. The first challenge is on efficiency. In practice, we are primarily interested in how the electron distribution is affected by the electron–plasmon interaction, rather than the detailed phase-space distribution of the plasmons themselves. However, the Hamiltonian flow of plasmons involves the smallest time scale in the entire model. As a result, solving this flow directly would waste significant computational resources on information that is not of interest. Fortunately, the Hamiltonian of the plasmons is time-independent, which means their trajectories are fixed. This allows us to eliminate the Hamiltonian flow term from the equation through trajectory averaging. The trajectory average technique [ 3 ] has been successful in tackling problems with simple advection field lines. For example, the gyro-average of Vlasov equations, where the average operator is simply integration along circles. For more complicated problems, there is the ray-tracing method [ 1 ] , where sample trajectories are calculated with the Hamilton’s equation, and numerical averaging is performed on these trajectories. However, such approach is expensive, not conservative, and not compatible with finite element or finite difference solvers. To address these issues, we conducted an in-depth study of the trajectory-averaging method in the language of functional analysis. It turns out that, the trajectory-averaging operator is essentially an orthogonal projection onto the null space of the advection (Hamiltonian flow) operator [ 3 ] . Based on this perspective, we propose not to discretize the trajectory-averaging operator directly. Instead, we first discretize the null space of the advection operator and then construct the corresponding projection operator. The second challenge is on accuracy. As a reduced model stemming of the Vlasov-Maxwell system, our kinetic system of electrons and plasmons inherits the conservation properties, which relies on the delicate structure of the “collision” operators and the Poisson bracket associated to plasmon dynamics. The goal of this paper is to develop a numerical solver that preserves the structure of the underlying physical systems, particularly for highly magnetized plasmas with non-uniform density profiles. The solver is designed to respect the conservation of mass, momentum, and energy while capturing the kinetic evolution of both particles and waves (electrons and plasmons). In 2023, Huang et. al. [ 7 ] proposed a structure-preserving solver for quasilinear theory, in which a continuous Galerkin method was employed to discretize the electron diffusion operator. However, in practical applications [ 4 ] , the electron kinetic equation involves many additional terms, notably convection terms. To ensure compatibility with such more general models, we extend the idea in [ 7 ] and propose a novel discontinuous Galerkin version. This paper is organized as follows. In section 2 we describe the set up of the problem. After that we derive the weak form of trajectory-averaged equation in section 3 and the connection-proportion algorithm in section 4 . The conservative LDG method for quasilinear theory will be introduced in section 5 . Implementation and complexity analysis will be discussed in section 6 . Section 8 contains the numerical results. 2 The kinetic system for particles and plasmons Consider an infinitely long cylinder \Omega_{x}=(0,R)\times(0,2\pi)\times\mathbb{R} . At any given location (r,\phi,z)\in\Omega_{x} , there is a set of local orthonormal basis (\mathbf{e}_{z},\mathbf{e}_{r},\mathbf{e}_{\phi}) . Suppose a plasma is confined in the cylinder, embedded in a magnetic field along the z-axis, \mathbf{B}(\mathbf{x})=B(r)\mathbf{e}_{z} . We focus on the z,\phi -symmetric case for simplicity, which means any function g(\mathbf{x}) depends only on r . The momentum \mathbf{p} of a particle at (r,\phi,z) can be decomposed as \mathbf{p}=p_{z}\mathbf{e}_{z}+p_{r}\mathbf{e_{r}}+p_{\phi}\mathbf{e}_{\phi}\in\mathbb{R}_{p}^{3} . Define p_{\parallel}=\frac{\mathbf{B}}{|\mathbf{B}|}\cdot\mathbf{p} and p_{\perp}=\left|\left(I-\frac{\mathbf{B}}{|\mathbf{B}|}\otimes\frac{\mathbf{B}}{|\mathbf{B}|}\right)\cdot\mathbf{p}\right| . Then for highly magnetized plasma, the particle probability density function f(\mathbf{p},\mathbf{x},t)=f(p_{\parallel},p_{\perp},r,t) . The momentum \mathbf{k} of a plasmon (the wave vector of a wave packet) can also be decomposed as \mathbf{k}=k_{z}\mathbf{e}_{z}+k_{r}\mathbf{e_{r}}+k_{\phi}\mathbf{e}_{\phi}\in\mathbb{R}_{k}^{3} . Define q_{\phi}=k_{\phi}\cdot r , then (k_{r},q_{\phi},k_{z}) and (r,\phi,z) are a set of canonical coordinates. Given z,\phi -symmetry, the plasmon probability density function N(\mathbf{k},\mathbf{x},t)=N(k_{r},q_{\phi},k_{z},r,t) . The following system governs the evolution of particle pdf f(\mathbf{p},\mathbf{x},t) and plasmon pdf N(\mathbf{k},\mathbf{x},t) , \begin{cases}\partial_{t}f+v_{z}\partial_{z}f&=\nabla_{p}\cdot(D[N]\cdot\nabla_{p}f),\\ \partial_{t}N+\left\{\omega,N\right\}&=\Gamma[f]N,\end{cases} (2.1) where the quasilinear diffusion term and the reaction term accounts for electron-plasmon interaction [ 4 ] . The advection term v_{z}\partial_{z}f comes from the gyro-averaged Vlasov equation, it vanishes since f does not depend on z . In addition, the Poisson bracket \left\{\omega,N\right\}=\frac{\partial N}{\partial r}\frac{\partial\omega}{\partial k_{r}}-\frac{\partial\omega}{\partial r}\frac{\partial N}{\partial k_{r}} is a result of WKB approximation. Denote it as \mathcal{T}N , then obviously the operator \mathcal{T} is anti-symmetric. The transport of plasmons is fully determined by the Hamiltonian \omega(\mathbf{k},\mathbf{x}) , i.e. the dispersion relation of waves in plasma. For a detailed discussion of wave dispersion relation in homogeneous plasmas, we refer the readers to section 3.2 of [ 6 ] . As have been shown in [ 6 ] , given the plasma frequency \omega_{pe} and electron gyro-frequency \omega_{ce} , wave frequency \omega is a function of wave-vector \mathbf{k} . Note that the plasma frequency \omega_{pe} is proportional to the square root of electron number density n_{e} and the gyro-frequency