pith. sign in

arxiv: 2604.09032 · v1 · submitted 2026-04-10 · ⚛️ physics.plasm-ph

A Fully Electromagnetic Hybrid PIC-Fluid Model for Predictive Fusion Neutron Yield in Dense Plasma Focus

Yinjian Zhao , Zhe Liu , Qiang Sun , Qianhong Zhou , Guangrui Sun This is my paper

Pith reviewed 2026-05-10 16:45 UTC · model grok-4.3

classification ⚛️ physics.plasm-ph
keywords dense plasma focushybrid PIC-fluid modelneutron yieldD-D fusionelectromagnetic simulationplasma sheathpost-pinch dynamics
0
0 comments X

The pith

A hybrid electromagnetic PIC-fluid model for dense plasma focus devices produces a D-D neutron yield of 0.296e7 at 180 kA, matching the order of magnitude of fully kinetic simulations.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces a simulation framework that advances ions kinetically with particle-in-cell methods while treating electrons as a quasi-neutral fluid to model the full electromagnetic behavior in a dense plasma focus. This setup solves Maxwell's equations across plasma and vacuum regions and incorporates a generalized Ohm's law with resistive, electron pressure, and Hall contributions, updated via a predictor-corrector step for current density. The model reproduces the sequence of sheath formation, axial rundown, radial compression, and post-pinch expansion, with sheath front positions agreeing within 10 percent of fully kinetic benchmarks. Using a compact fit to the D-D fusion cross section, it calculates a total neutron yield of 0.296e7, which is comparable in scale to reported kinetic results at similar currents and nearly two orders of magnitude above earlier hybrid predictions.

Core claim

The fully electromagnetic hybrid simulation framework, with kinetic particle-in-cell ions, quasi-neutral electron fluid, generalized Ohm's law including resistive, pressure-gradient, and Hall terms, and predictor-corrector current density update, when applied to a non-hollow 180 kA DPF geometry, reproduces sheath formation, axial rundown, radial compression, and post-pinch expansion, and yields a total neutron output of 0.296e7 that is comparable in order of magnitude to fully kinetic results at similar currents.

What carries the argument

The hybrid simulation framework that advances ions kinetically via particle-in-cell while evolving electrons as a quasi-neutral fluid under the generalized Ohm's law and solves Maxwell's equations in both plasma and vacuum regions.

If this is right

  • The outer sheath front position matches fully kinetic benchmarks within 10 percent over the comparison interval.
  • The predicted neutron yield is nearly two orders of magnitude higher than earlier hybrid results at comparable currents.
  • The model self-consistently captures post-pinch expansion dynamics that contribute to neutron production.
  • Quantitative neutron yield forecasts become feasible without resolving full electron kinetics.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The framework could be applied to other pulsed-power geometries to estimate yields more efficiently than full kinetic runs.
  • Validation against measured ion temperature profiles in the post-pinch phase would strengthen in the electron fluid approximation.
  • Insights from the reproduced axial electric field evolution might inform design adjustments to enhance neutron output in compact devices.

Load-bearing premise

Treating electrons as a quasi-neutral fluid with the generalized Ohm's law and predictor-corrector update captures the essential kinetic ion motion and electromagnetic coupling for accurate neutron yield prediction after the pinch.

What would settle it

Experimental measurement of total neutron yield in a non-hollow 180 kA DPF matching the simulated geometry would directly test whether the predicted 0.296e7 value holds.

Figures

Figures reproduced from arXiv: 2604.09032 by Guangrui Sun, Qiang Sun, Qianhong Zhou, Yinjian Zhao, Zhe Liu.

Figure 1
Figure 1. Figure 1: The operational structure of a dense plasma focus device consists of distinct evolution stages, including the flashover stage (I), the rundown stage (II), the run in stage (III), the pinch stage (IV). 2. Axial rundown. The plasma sheath is accelerated down the length of the inner electrode by the self-generated azimuthal magnetic field through the j×B force. During this rundown phase the sheath sweeps up a… view at source ↗
Figure 2
Figure 2. Figure 2: Main time-stepping workflow of the hybrid ion-PIC/electron-fluid electromagnetic solver. The loop advances ion macroparticles (Boris), deposits charge/current, solves the Ohm-Ampère closure for the current density, updates fields by FDTD with Marder correction, and applies ion-ion collisions. The solving procedures are described below: (1) Applying the Boris’ leap-frog method[36] and using x n i , v n+1/2 … view at source ↗
Figure 3
Figure 3. Figure 3: Predictor-corrector scheme used to estimate the end-of-step current density. The method time-extrapolates J, performs a provisional ion update, rebuilds ne, Te and Ji , and then solves the generalized Ohm’s law to obtain a stable J n+1 . We set σ according to σ =    0, if ne/n0 ∈ [0,0.1), (ne/n0) 3 σ0, if ne/n0 ∈ [0.1,1), σ0, if ne/n0 ∈ [1,∞), (28) where n0 denotes the initial background plasma dens… view at source ↗
Figure 4
Figure 4. Figure 4: Staggered Yee grid in 2D-RZ. Densities/temperatures (n,T) are defined at grid nodes; Bθ at cell centers; (Er ,Ez) and (Jr ,Jz) at face midpoints [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Computational domain and boundary conditions for the LLNL-like DPF geometry. The left and the right boundary employs an axial PML; the upper boundary and shaded regions are perfect conductors. 3 Simulation setup 3.1 Grid and particles For the implosion to pinch phase in DPF, this paper establishes a particle PIC-electron fluid model based on the quasi-neutral assumption and develops hybrid simulation code.… view at source ↗
Figure 6
Figure 6. Figure 6: Initial conditions at the end of rundown. A ∼1 mm plasma sheath (orange) centered near z≈4.5 cm with ns,0 = 3.3×1023 m−3 is followed upstream by deuterium plasma number density of nn,0 = 6.7×1022 m−3 and by vacuum downstream. The computational domain extends to a maximum radius of Rmax = 1.56 cm and an axial length of Zmax = 10.46 cm, and is discretized on a uniform cylindrical grid with ∆r = ∆z = 0.02 cm … view at source ↗
Figure 7
Figure 7. Figure 7: Time evolution of ion number density from the hybrid simulation. The axial phase (top panels) shows sheath propagation along the anode surface, and the radial phase (bottom panels) shows the sheath wrapping around the anode tip and imploding toward the axis to form a dense pinch column [PITH_FULL_IMAGE:figures/full_fig_p011_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Ion temperature evolution in the hybrid simulation. Localized heating develops at the sheath front during the axial phase, and a compact hot region forms near the axis in the radial phase, where the maximum ion temperature reaches Tmax ≈ 2.45 keV. A second order median filter is applied to the ion temperature field before plotting in order to suppress cell scale particle noise while preserving the large sc… view at source ↗
Figure 9
Figure 9. Figure 9: Evolution of the azimuthal magnetic field during the (a) axial and (b) radial phases. The self generated magnetic field builds up around the anode surface and tip, reaching a peak value Bmax ≈ 37.6 T near the end of radial compression. Fig.9 shows the evolution of the azimuthal magnetic field around the anode during the discharge. At t = 0 ns the field is essentially zero everywhere, since the external cir… view at source ↗
Figure 10
Figure 10. Figure 10: Axial electric field Ez at representative times before, during, and after pinch. Strong bipolar fields form at the sheath front during rundown and compression, followed by an extended mainly positive field region in the post-pinch stage that drives beam like ion acceleration. indicating that the electric field is still mainly confined to the vicinity of the current sheath. As the discharge evolves to t = … view at source ↗
Figure 11
Figure 11. Figure 11: Sheath front position zmax as a function of time. Blue circles show the hybrid simulation result and orange markers indicate sheath positions inferred from LLNL fully kinetic simulations and measurements. Shaded bands distinguish axial and radial phases, demonstrating good agreement in sheath kinematics over the measured range. zmax /cm LLNL /ns Ours /ns Difference /% 5.12 40 38.21 -4.47 5.42 60 65.39 -8.… view at source ↗
Figure 12
Figure 12. Figure 12: Time evolution of the instantaneous neutron production rate dN/dt (solid blue, left axis) and cumulative yield N(t) (dashed red, right axis). The arrow marks the time when the plasma sheath first reaches the axis at r = 0, and the shaded region indicates the main pinch burst during which almost all neutrons are produced. Fig.12 shows that the neutron production rate remains negligible during the rundown p… view at source ↗
Figure 13
Figure 13. Figure 13: Time evolution of the discharge current IDPF and system voltage UDPF obtained from the hybrid simulation. The current between electrodes remains close to 180 kA during the axial phase, then experiences a small drop at the onset of radial compression and gradually decreases in the post-pinch stage. The voltage across the plasma remains relatively small in the early phase, but develops large oscillations of… view at source ↗
Figure 14
Figure 14. Figure 14: Sheath front position zmax(t) for the baseline run with ∆t = ∆t0, a refined time step run with ∆t = 0.5∆t0, and a refined spatial resolution run with ∆r = 0.5∆r0 and ∆z = 0.5∆z0, showing that the macroscopic sheath motion is insensitive to further refinement in space and time. are almost indistinguishable: sheath arrival times at several reference axial positions differ by less than 0.5%. For the refined … view at source ↗
Figure 15
Figure 15. Figure 15: Sheath-front position zmax(t) for parameter scans: (a) conductivity threshold nth, (b) Ohmic CFL factor Cσ , (c) Marder factor d, and (d) electron temperature closure Te = αTi . The trajectories are nearly identical, indicating weak sensitivity to these parameters. To assess whether the inferred sheath dynamics is robust with respect to the numerical and modeling choices in the hybrid formulation, we perf… view at source ↗
Figure 16
Figure 16. Figure 16: Time evolution of the ion number density when the electron-pressure gradient term is included in the generalized Ohm’s law. The sheath front propagates slightly faster, the pinch occurs earlier, and the high-density region becomes broader compared with the reference case without the electron-pressure gradient and Hall terms. and subsequent radial compression follow the same global sequence as in the refer… view at source ↗
Figure 17
Figure 17. Figure 17: High-speed ion macroparticles and the ray-like structures produced when the electron pressure gradient is applied everywhere. In the final simulations, the pressure-gradient term is limited to cells with ni > 1.2×1023 m−3 to avoid this effect while keeping the main plasma behavior. Fig.17 illustrates a numerical artefact that appears when the electron pressure gradient is evaluated throughout the entire d… view at source ↗
Figure 18
Figure 18. Figure 18: Time evolution of the ion number density when both the electron-pressure gradient and Hall terms are included in the generalized Ohm’s law. The Hall contribution further modifies the sheath dynamics, enhancing early-stage acceleration while amplifying small-scale structure and late-stage instability relative to the case without the Hall term. Fig.18 shows the corresponding ion number density evolution whe… view at source ↗
read the original abstract

While magnetic confinement fusion (MCF) and inertial confinement fusion (ICF) remain the primary routes toward controlled fusion, progress is still constrained by energy loss, plasma instabilities, and the cost and complexity of large-scale facilities. The Dense Plasma Focus (DPF) device presents a compact, pulsed-power-driven alternative for producing fusion-relevant conditions and neutron emissions. However, the quantitative prediction of neutron yield in DPF devices poses a significant numerical challenge, primarily due to the imperative of self-consistently resolving kinetic ion behavior, electromagnetic energy coupling, and vacuum field evolution. This complexity often impedes a definitive understanding of the underlying neutron production mechanisms. To address this, we develop a fully electromagnetic hybrid simulation framework: ions are advanced kinetically with particle-in-cell, electrons are a quasi-neutral fluid, and Maxwell's equations are solved in both plasma and vacuum. The generalized Ohm law includes resistive, electron pressure-gradient, and Hall terms, with a predictor-corrector update for current density. We apply the model to a non-hollow 180 kA DPF geometry similar to the LLNL configuration. The simulated ion density, ion temperature, and axial electric field reproduce sheath formation, axial rundown, radial compression, and post-pinch expansion. The outer sheath front position agrees with fully kinetic benchmarks within 10\% over the available comparison interval. With a compact fit to the D-D fusion cross section, the predicted total neutron yield is 0.296e7, comparable in order of magnitude to reported fully kinetic results at similar currents and nearly two orders of magnitude higher than earlier hybrid results.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript introduces a fully electromagnetic hybrid PIC-fluid model for Dense Plasma Focus (DPF) simulations, treating ions kinetically via PIC and electrons as a quasi-neutral fluid with a generalized Ohm's law including resistive, pressure gradient, and Hall terms, solved with Maxwell's equations in plasma and vacuum regions. Applied to a non-hollow 180 kA DPF similar to LLNL configuration, the model reproduces sheath formation, rundown, compression, and post-pinch expansion, with outer sheath front position agreeing within 10% of fully kinetic benchmarks. Using a compact fit to the D-D fusion cross section, it predicts a total neutron yield of 0.296e7, which is order-of-magnitude consistent with fully kinetic results at similar currents and significantly higher than prior hybrid models.

Significance. If the post-pinch ion distributions and fields prove accurate, the hybrid framework offers a computationally tractable route to neutron-yield predictions in DPF devices that improves on earlier hybrid results while retaining key electromagnetic and kinetic-ion physics. This could support parameter studies that remain prohibitive for fully kinetic codes.

major comments (2)
  1. [Abstract] Abstract: the central claim that the model is 'predictive' for fusion neutron yield rests on integrating a compact D-D cross-section fit over the simulated ion distribution, yet the only quantitative benchmark reported is 10% agreement on outer sheath front position; no error bars, ion velocity spectra, time-resolved fusion-rate histories, or post-pinch E-field comparisons against fully kinetic references are supplied, leaving the quantitative support for the 0.296e7 yield limited given the exponential sensitivity of the cross section above ~10 keV.
  2. [Model formulation] Model formulation (generalized Ohm's law and quasi-neutral closure): the electron fluid treatment with resistive + ∇p_e + Hall terms plus predictor-corrector current update is asserted to capture the electromagnetic coupling needed for accurate ion acceleration in the post-pinch phase, but the manuscript provides no test of whether non-thermal electron kinetics or inertia effects (suppressed by quasi-neutrality) alter the axial E-field or high-energy ion tail that dominate the yield integral.
minor comments (1)
  1. [Abstract] Abstract: the phrase 'compact fit to the D-D fusion cross section' is used without stating the functional form or fitted coefficients, which would improve reproducibility of the 0.296e7 result.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the thorough review and constructive comments. We address each major comment point by point below and indicate where revisions have been made to the manuscript.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim that the model is 'predictive' for fusion neutron yield rests on integrating a compact D-D cross-section fit over the simulated ion distribution, yet the only quantitative benchmark reported is 10% agreement on outer sheath front position; no error bars, ion velocity spectra, time-resolved fusion-rate histories, or post-pinch E-field comparisons against fully kinetic references are supplied, leaving the quantitative support for the 0.296e7 yield limited given the exponential sensitivity of the cross section above ~10 keV.

    Authors: We agree that the neutron yield validation rests primarily on the 10% agreement in outer sheath front position, a key metric for the compression dynamics where fusion occurs. The reported yield of 0.296e7 is computed directly from the simulated ion distribution via the D-D cross-section fit and matches the order of magnitude of fully kinetic results at comparable currents. We acknowledge the exponential sensitivity of the cross section and the lack of additional benchmarks (spectra, time histories, E-field) in the original text. In revision we have updated the abstract and added a limitations paragraph noting that the model targets order-of-magnitude predictive capability rather than precise quantitative forecasts, while retaining the improvement over earlier hybrid results. revision: partial

  2. Referee: [Model formulation] Model formulation (generalized Ohm's law and quasi-neutral closure): the electron fluid treatment with resistive + ∇p_e + Hall terms plus predictor-corrector current update is asserted to capture the electromagnetic coupling needed for accurate ion acceleration in the post-pinch phase, but the manuscript provides no test of whether non-thermal electron kinetics or inertia effects (suppressed by quasi-neutrality) alter the axial E-field or high-energy ion tail that dominate the yield integral.

    Authors: Quasi-neutrality is a deliberate closure that removes electron inertia and non-thermal kinetics to enable tractable simulations. The generalized Ohm's law (resistive, ∇p_e, Hall) together with the predictor-corrector current update is intended to retain the dominant electromagnetic coupling. The 10% agreement with fully kinetic benchmarks on sheath position indicates that the essential ion-acceleration physics is captured for this regime. We have added a paragraph in the revised manuscript explicitly discussing the possible influence of the suppressed effects on post-pinch E-field and high-energy tails, together with supporting references from the hybrid-model literature. revision: partial

Circularity Check

0 steps flagged

No significant circularity; yield computed from independent cross-section data and standard governing equations

full rationale

The derivation chain begins from Maxwell's equations, the generalized Ohm's law (resistive + ∇p_e + Hall terms), quasi-neutral electron fluid, and kinetic PIC ions, all standard and not fitted to the target neutron yield. The yield is obtained by integrating a compact fit to the external D-D fusion cross-section data over the simulated ion distribution; this step does not reduce the output to the model's inputs by construction. No self-citations, uniqueness theorems, or ansatzes imported from prior author work appear in the provided text. The reported 10% agreement on sheath-front position against fully kinetic benchmarks supplies an independent check, confirming the central prediction retains content beyond its inputs.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 0 invented entities

The framework rests on standard plasma physics approximations with one fitted element for yield calculation and no new postulated entities.

free parameters (1)
  • compact fit to D-D fusion cross section
    Used to compute neutron production rate from simulated ion density and temperature distributions.
axioms (2)
  • domain assumption Electrons form a quasi-neutral fluid
    Core approximation enabling fluid treatment of electrons while advancing ions kinetically.
  • domain assumption Generalized Ohm's law with resistive, electron pressure-gradient, and Hall terms governs current density
    Invoked for the predictor-corrector update of current in the electromagnetic solver.

pith-pipeline@v0.9.0 · 5599 in / 1587 out tokens · 42740 ms · 2026-05-10T16:45:17.578351+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

42 extracted references · 42 canonical work pages

  1. [1]

    Zylstra A, Hurricane O, Callahan D, Kritcher A, Ralph J, Robey H, Ross J, Young C, Baker K, Casey Det al.2022Nature601542–548

    work page

  2. [2]

    Matzen M K, Sweeney M, Adams R, Asay J, Bailey J, Bennett G, Bliss D, Bloomquist D, Brunner T, Campbell R eet al.2005Physics of Plasmas12

    work page

  3. [3]

    Zohuri B 2017Inertial confinement fusion driven thermonuclear energy(Springer)

    work page

  4. [4]

    Brezιnsek S, Dhard C P, Jakubowski M, König R, Masuzaki S, Mayer M, Naujoks D, Romazanov J, Schmid K, Schmitz Oet al.2021Nuclear Fusion62016006

    work page

  5. [5]

    Mather J 1964The Physics of Fluids7S28–S34

    work page

  6. [6]

    Fusion, Suppl

    Filippov N, Filippova T and Vinogradov V 1961Nucl. Fusion, Suppl

    work page

  7. [7]

    Bernstein M and Hai F 1970Physical Review Letters25641

    work page

  8. [8]

    Schmidt A, Tang V and Welch D 2012Physical review letters109205003

    work page

  9. [9]

    Lee S 2014Journal of Fusion Energy33319–335

    work page

  10. [10]

    Milanese M and Pouzo J 1978Nuclear Fusion18533

    work page

  11. [11]

    Lerner E J, Hassan S M, Karamitsos I and V on Roessel F 2017Physics of Plasmas24

    work page

  12. [12]

    Lerner E J, Murali S K, Shannon D, Blake A M and Roessel F V 2012Physics of Plasmas19

    work page

  13. [13]

    Cicuttin A, Crespo M, Gribkov V , Niemela J, Tuniz C, Zanolli C, Chernyshova M, Demina E, Latyshev S, Pimenov Vet al.2015Nuclear Fusion55063037

    work page

  14. [14]

    Bernstein M, Lee C and Hai F 1971Physical Review Letters27844

    work page

  15. [15]

    Inestrosa-Izurieta M J, Ramos-Moore E and Soto L 2015Nuclear Fusion55093011

    work page

  16. [16]

    Lerner E J, Krupakar Murali S and Haboub A 2011Journal of fusion energy30367–376 21 IOP PublishingJournalvv(yyyy) aaaaaa Authoret al

    work page

  17. [17]

    Lerner E J, Hassan S M, Karamitsos-Zivkovic I and Fritsch R 2023Journal of Fusion Energy427

    work page

  18. [18]

    Kiai S S, Adlparvar S, Sadeghi H and Fazelpour S 2025Scientific Reports1514081

    work page

  19. [19]

    Yousefi H, Haruki T, Sakai J, Lumanta A and Masugata K 2009Physics Letters A3732360–2363

    work page

  20. [20]

    Bennett N, Blasco M, Breeding K, Constantino D, DeYoung A, DiPuccio V , Friedman J, Gall B, Gardner S, Gatling Jet al.2017Physics of Plasmas24

    work page

  21. [21]

    Tang V , Adams M and Rusnak B 2010IEEE transactions on plasma science38719–727

    work page

  22. [22]

    Mejia F C, Milanese M, Moroso R and Pouzo J 1997Journal of Physics D: Applied Physics301499

    work page

  23. [23]

    Johnson D J 1974Journal of Applied Physics451147–1153

    work page

  24. [24]

    Gratton F and Vargas J 1983 Two dimensional electromechanical model of the plasma focusEnergy storage, compression, and switchingpp 353–386

    work page 1983

  25. [25]

    Schmidt A, Link A, Welch D, Ellsworth J, Falabella S and Tang V 2014Physical Review E89061101

    work page

  26. [26]

    Schmidt A, Welch D, Meehan B, Tang V , Halvorson C, May M and Hagen E 2014Physics of Plasmas 21

    work page

  27. [27]

    Mathuthu M, Zengeni T G and Gholap A V 2002IEEE transactions on plasma science251382–1388

    work page

  28. [28]

    Ay Y , Abd Al-Halim M A and Bourham M A 2015The European Physical Journal D69205

    work page

  29. [29]

    Li H, Li S, Jungman G and Hayes-Sterbenz A C 2017 Dense plasma focus modeling Tech. rep. Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)

    work page 2017

  30. [30]

    Beresnyak A, Giuliani J L, Jackson S L, Richardson A S, Swanekamp S B, Schumer J, Weber B V and Mosher D 2018IEEE Transactions on Plasma Science463881–3885

    work page

  31. [31]

    Jiang S, Higginson D, Holod I and Schmidt A 2019Physics of Plasmas26

    work page

  32. [32]

    Gribkov V , Van Oost G, Malaquias A and Herrera J 2006Nuclear Fusion46868

    work page

  33. [33]

    Thoma C, Welch D, Clark R, Rose D and Golovkin I 2017Physics of Plasmas24

    work page

  34. [34]

    Goyon C, Link A, Drury O and Schmidt A 2023

    work page 2023

  35. [35]

    Goyon C, Drury O, Youmans A, Angus J, Jiang S, Rocco S, Chapman S, Cooper C, McMahon M, Jibodu Aet al.2025Physics of Plasmas32

    work page

  36. [36]

    Delzanno G and Camporeale E 2013Journal of Computational Physics253259–277

    work page

  37. [37]

    Zhao Y , Cui C and Hu Y 2022Computational Particle Mechanics

    work page

  38. [38]

    Inan U S and Marshall R A 2011Numerical Electromagnetics The FDTD Method(Cambridge University Press)

    work page

  39. [39]

    Marder B 1987Journal of Computational Physics6848–55

    work page

  40. [40]

    Nanbu K 1997Physical Review E554642–4652

    work page

  41. [41]

    Shaw B, Chapman S, Cooper C, Goyon C, Angus J, Higginson D, Liu J, Mitrani J, Podpaly Y , Povilus Aet al.2018Journal of Applied Physics124

    work page

  42. [42]

    Novotny J, Cikhardt J, Cikhardtova B, Klir D, Kravarik J, Kubes P, Malir J, Munzar V and Rezac K 2023Physics of Plasmas30 22

    work page