Leaking Outside the Box: Kinetic Turbulence with Cosmic-Ray Escape

Evgeny A. Gorbunov, Daniel Grošelj, Fabio Bacchini · 2025

Read Leaking Outside the Box: Kinetic Turbulence with Cosmic-Ray Escape by Evgeny A. Gorbunov, Daniel Grošelj, Fabio Bacchini (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-2503-03820.txt.

Introduction

Evgeny Gorbunov, Daniel Groselj, and Fabio Bacchini simulated cosmic ray escape from turbulent plasma in a 2025 paper. Standard boxes keep fast particles trapped, which distorts how acceleration settles. Their three-dimensional runs inject particles from an outside heat bath and let them diffuse away.

A steady state formed, with nonthermal particles reaching the theoretical energy ceiling. Plasma and magnetic pressures balanced, and that balance capped the acceleration rate. As the cold plasma's magnetization rose, spectra hardened and more of the spent energy left as cosmic rays. Past a threshold, escaping rays carried over half the dissipated energy. The method opens steady-state kinetic studies for many cosmic settings, from jets to galaxy clusters.

Catalog record

CREATOREvgeny A. Gorbunov, Daniel Grošelj, Fabio Bacchini
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE26 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-03820

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Leaking Outside the Box: Kinetic Turbulence with Cosmic-Ray Escape Evgeny A. Gorbunov, Daniel Grošelj, Fabio Bacchini Abstract We study particle acceleration in strongly turbulent pair plasmas using novel 3D Particle-in-Cell simulations, featuring particle injection from an external heat bath and diffusive escape. We demonstrate the formation of steady-state, nonthermal particle distributions with maximum energies reaching the Hillas limit. The steady state is characterized by the equilibration of plasma kinetic and magnetic pressures, which imposes upper limits on the acceleration rate. With growing cold plasma magnetization σ 0 \sigma_{0} , nonthermal power-law spectra become harder, and the fraction of energy channeled into escaping cosmic rays increases. At σ 0 ≳ 1 \sigma_{0}\gtrsim 1 , the escaping cosmic rays amount to more than 50% of the dissipated energy. Our method allows for kinetic studies of particle acceleration under steady-state conditions, with applications to a variety of astrophysical systems. Introduction. —Turbulence is believed to be a widespread mechanism for nonthermal particle acceleration in collisionless plasmas [ 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 ] . In recent years, kinetic plasma simulations have provided first-principles insight into the mechanisms of particle acceleration in turbulent astrophysical environments by uncovering the nature of particle dynamics over their entire acceleration history [ 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 ] . It was shown that particles can be rapidly injected from the thermal pool into the suprathermal population by nonideal electric fields near reconnecting current sheets [ 27 , 19 , 28 , 24 ] . The injection is followed by stochastic energization via interactions with turbulent structures [ 29 , 30 , 31 , 32 , 33 , 26 ] . This can be seen as a generalization of the second-order Fermi process [ 1 , 2 , 34 , 35 ] , which involves diffusion in energy space. However, most kinetic simulations to date face a significant shortcoming: they lack a genuine steady state, due to the fact that the simulation domains employed are of finite size and closed, which leads to a progressive accumulation of energy over time and to the pile-up of particles at the highest energies [ 17 , 36 , 37 ] . This presents challenges for connecting simulations and realistic astrophysical objects, which may convert a significant fraction of their available power into escaping cosmic rays. Moreover, it is known that power-law distributions emerge most naturally from a competition between particle acceleration and escape. In closed domains, the local trapping of particles provides an effective “escape” mechanism [ 36 ] , enabling the development of power-law distributions. However, without an energy sink to balance the turbulent forcing, such power-law distributions are only transient states, followed by progressive energy pile-up. Therefore, steady-state simulations, if achievable, could offer a convenient and sorely needed testbed for theoretical models. In this Letter, we develop an efficient method for achieving steady state in kinetic plasma-turbulence simulations in open domains, and we characterize the emerging steady-state statistics of particle acceleration and escape. As an example, we study large-amplitude 3D Alfvénic turbulence in electron–positron plasmas. The steady state is characterized by a near-equilibration of plasma kinetic and magnetic pressures. As a result, bulk turbulent motions remain subrelativistic, which limits the rate of stochastic particle acceleration. The particle distributions exhibit nonthermal power laws extending up to the system size (Hillas) limit, where the rate of diffusive particle escape reaches the theoretical maximum (set by the light-crossing time). We also show that strongly magnetized sources may convert a large fraction of the available power into escaping cosmic rays. In our simulations, the nonthermal energy fraction of escaping particles grows with the cold plasma magnetization \sigma_{0} (defined below) and saturates near 70% at \sigma_{0}\gtrsim 100 . Kinetic Leaky-Box Model. —We perform 3D simulations of kinetic plasma turbulence with particle injection and escape using the Particle-in-Cell (PIC) code Tristan-MP v2 [ 38 ] . Our model mimics an open turbulent box of side L=2l_{\rm esc} , coupled to an external heat bath at a fixed temperature T_{0} . Particles escape the magnetized turbulent accelerator by diffusing over a distance l_{\rm esc} perpendicular to the mean magnetic field \mbox{\boldmath{$B$}}_{0}=B_{0}\hat{\mbox{\boldmath{$z$}}} , while at the same time new thermal particles are introduced into the domain from the external heat reservoir. To implement escape, we introduce the particle variables \delta x and \delta y , which track the displacement of each particle from its initial location in x and y , respectively 1 1 1 Particles inserted at t>0 are initialized with \delta x=\delta y=0 (i.e., the reference point for the displacement is the particle’s own initial location). For particles in the box at t=0 we make an exception. Their initial \delta x and \delta y are preset to random numbers between -l_{\rm esc} and l_{\rm esc} , as if those particles had already dispersed from their point of origin. . A particle counts as escaped when \max(|\delta x|,|\delta y|)>l_{\rm esc} (see also [ 25 ] ). In order to keep the total number of particles fixed, we insert a new thermal particle (sampled from a Maxwellian at temperature T_{0} ) at the location of every escaping particle. The inserted particle has the same charge and mass as the escaping particle. Employing for simplicity an electron–positron composition ( m_{e^{-}}=m_{e^{+}}=m ) of mean number density n_{0} , we explore the effect of different cold plasma magnetizations \sigma_{0}=B_{0}^{2}/(4\pi n_{0}mc^{2}) on particle acceleration and escape. Several runs with different \sigma_{0}=\{0.2,1,4,10,20,80\} are performed. A periodic box of size L^{3} is initially filled with thermal particles at temperature T_{0} . We set T_{0}=mc^{2}[\sigma_{0}(2+\sigma_{0})]/[3+3\sigma_{0}] , which ensures that the mean energy of the injected particles \mathrm{E}_{0}\sim mc^{2}\sigma_{0} is close to the expected steady-state mean particle energy (see Eq. ( 1 ) below). To sustain turbulence, a Langevin antenna drives an external current [ 40 ] at the largest (box-scale) wavelengths with frequency \omega_{\rm ant}\approx 0.9\omega_{\rm A} , slightly off-resonant to the lowest Alfvén frequency of the box \omega_{\rm A}=(2\pi/L)v_{\rm A} . The Alfvén speed v_{\rm A}=c\sqrt{\sigma/(\sigma+1)} is defined here based on the initial hot magnetization \sigma=B_{0}^{2}/(4\pi w_{0}n_{0}mc^{2})=\sigma_{0}/w_{0} , where w_{0}=\mathcal{K}_{3}(mc^{2}/T_{0})/\mathcal{K}_{2}(mc^{2}/T_{0}) is the initial specific enthalpy and \mathcal{K}_{n}(z) are n -th order modified Bessel functions of the second kind. The decorrelation rate of the antenna is \gamma_{\rm ant}=0.5\omega_{\rm ant} . We focus on the strong-turbulence regime, where the magnetic perturbation driven by the antenna is comparable to the background magnetic field: \delta b\equiv\delta B/B_{0}\approx 1 . The spatial resolution of all simulations presented in this Letter is 960^{3} , and the physical box size L=480d , where d=[(1+2\sigma_{0})mc^{2}/(4\pi n_{0}e^{2})]^{1/2} is a reference hot-plasma skin depth based on the expected steady-state kinetic temperature (discussed below). We use 16 particles per cell in all runs 2 2 2 For a convergence check in particle number, see e.g. [ 19 ] . . The simulations are run for at least 20 light-crossing times t_{\rm cross}=l_{\rm esc}/c . Energy Balance in Steady State. — Our simulations presented in the following show that large-amplitude kinetic turbulence with particle escape settles into

Read the complete text