Pith. sign in

REVIEW 4 major objections 4 minor 26 references

High order interpolation of magnetic fields with vector potential reconstruction for particle simulations

T0 review · 4 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Reconstructing the vector potential from Hermite-interpolated field data yields a globally divergence-free magnetic field with $C^m$ continuity, preserving the designed order of Runge-Kutta integrators in guiding-center and Poincaré…

desk verdict A sound Hermite-based vector-potential reconstruction for divergence-free field interpolation, but the high-order convergence claims from discrete data rest on unspecified stencils and there is no baseline comparison. read the letter →

arxiv 2501.01523 v1 pith:JPJGZH6D submitted 2025-01-02 math.NA cs.NAphysics.plasm-ph

Oleksii Beznosov , Jesus Bonilla , Xianzhu Tang , Golo Wimmer This is my paper · ORCID
classification math.NAcs.NAphysics.plasm-ph MSC 65D0565L0665N30
keywords vectorpotentialreconstructionHermiteinterpolationdivergence-freemagneticfieldsRunge-KuttaorderreductionguidingcenterequationsPoincarésectionmimeticfinitedifferenceelement
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper proposes an interpolation recipe for discrete magnetic fields whose result is exactly divergence-free everywhere, not merely in the discrete sense. Instead of interpolating $\mathbf{B}$ directly, the method Hermite-interpolates the cylindrical field components, analytically integrates these interpolants to reconstruct the vector potential $\mathbf{A}$, and then evaluates $\mathbf{B}=\nabla\times\mathbf{A}$. This gives a field with $C^m$ continuity, meaning derivatives through order $m$ are continuous, and with high-order accuracy. That smoothness matters because high-order Runge-Kutta integrators such as the Dormand-Prince 4(5) pair rely on Taylor expansion and lose their designed convergence order when the field has jumps in high derivatives. The paper demonstrates the benefit in guiding-center trajectory conservation and Poincaré section analysis using analytical, mimetic finite difference, and weakly divergence-free finite element data.

What carries the argument

The load-bearing identity is the gauge choice $A_Z=0$ with the rescaled components $\psi=R A_\phi$ and $\chi=R A_R$. Under this gauge the reconstruction reduces to two line integrals, $\psi(R,\phi,Z)=-\int_{Z_0}^{Z}R B_R\,dZ' + \int_{R_0}^{R}R' B_Z(R',\phi,Z_0)\,dR'$ and $\chi(R,\phi,Z)=\int_{Z_0}^{Z}R B_\phi\,dZ'$, evaluated analytically from Hermite interpolants of the field data. The interpolation basis is piecewise Taylor-trigonometric polynomials, so all derivatives and integrals are exact operations on the stored Taylor coefficients; consequently the final $\mathbf{B}=\nabla\times\mathbf{A}$ inherits $C^m$ continuity and has identically zero divergence rather than a small residual.

What would settle it

Take a weakly divergence-free finite element field whose true high-resolution reference is known in the same domain, reconstruct with increasing $m$, and check whether the $L^2$ error of $\mathbf{B}$ and the conservation errors of $p_\phi$ and $\mu$ follow the predicted orders; if the error saturates because of the projection step, especially below the X-point in a divertor geometry, the assumption fails.

Watch

Extended reading notes

Core claim

The central claim is that a globally solenoidal, high-order-continuous interpolant can be produced from discrete field data by reconstructing the vector potential in the gauge $A_Z=0$. With the rescaled variables $\psi=R A_\phi$ and $\chi=R A_R$, the components of $\mathbf{B}=\nabla\times\mathbf{A}$ reduce to two line integrals: $\psi$ is obtained by integrating $R B_R$ in $Z$ and then $R B_Z$ along the boundary plane, and $\chi$ by integrating $R B_\phi$ in $Z$. These integrals are evaluated analytically on the Hermite interpolants, so taking the curl is exact and the resulting field is divergence-free by construction with the declared $C^m$ smoothness. In the axisymmetric convergence test, $\psi$ and $B_R$ converge at order $2m$ while $B_Z$ converges at order $2m-1$. In guiding-center simulations, insufficient smoothness ($m<3$ on the mimetic finite difference field) degrades the Dormand-Prince 4(5) method's convergence, especially at low error tolerances.

Load-bearing premise

The load-bearing premise is that the input discrete field is regular and close enough to divergence-free that the line-integral reconstruction yields a vector potential near the true physical field; this is unquantified for weakly divergence-free finite element data, and the paper explicitly notes that errors can propagate in non-convex regions such as the divertor.

Editorial extensions

If this is right

  • Adaptive high-order ODE integrators such as Dormand-Prince 4(5) can run at their designed order when the reconstructed field has sufficient Hermite smoothness; the paper finds $m\geq 3$ is needed for the mimetic finite difference tokamak field, while $m=2$ sufficed for analytic fields.
  • Guiding-center invariants, toroidal canonical momentum and magnetic moment, are conserved tightly enough for long-term trajectory integration, with the paper's tests spanning $10^5$ to $10^8$ timesteps.
  • Weakly divergence-free finite element data can be projected onto an exactly solenoidal field, enabling meaningful Poincaré sections even when the source mesh is coarse and semi-regular.
  • Field-line integration for Poincaré sections becomes cheap because evaluating the reconstructed $\mathbf{B}$ only requires local polynomial and trigonometric operations at each integrator stage.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same gauge-and-integrate reconstruction should apply to any solenoidal vector field in cylindrical geometry, not just magnetic fields; nothing in the reconstruction equations is magnetohydrodynamic-specific.
  • The $m\geq3$ requirement suggests a practical quality check: measure the fourth-derivative jumps of the reconstructed field at cell boundaries and compare them with the integrator's truncation error, increasing $m$ or refining the grid when they dominate.
  • Because the paper notes that errors from outside the domain can propagate inward through the line integrals in non-convex regions such as the divertor, a natural extension is a domain-extension or multiple-reference-point scheme that contains those errors.
  • Compared with the dense-output and root-finding alternative the paper mentions, this method pays a one-time global preprocessing cost to avoid per-particle root finding; a direct wall-clock comparison on the same Poincaré workflow would be a useful follow-up.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper develops a method for reconstructing a globally divergence-free, C^m-smooth magnetic field from discrete field data. Taylor coefficients at primal grid points are used to build tensor-product Hermite interpolation in the poloidal plane and Fourier interpolation in the toroidal direction; the interpolated field components are then integrated analytically to construct the vector-potential variables psi = R A_phi and chi = R A_R in the gauge A_Z = 0, and B = curl A is evaluated by exact differentiation. The authors report (2m+2)-order convergence for psi and B_R (one order less for B_Z) on analytic fields, study conservation of p_phi and mu in relativistic guiding-center runs with a Dormand-Prince 4(5) integrator, and apply the method to Poincaré section analysis for mimetic finite-difference and finite-element data. The main claims are that the exactly divergence-free character improves long-time particle conservation and that insufficient Hermite order (m<3 for non-smooth numerical fields) degrades the Runge-Kutta order.

Significance. The vector-potential reconstruction itself is the strongest part of the paper: the gauge choice A_Z = 0 with psi = R A_phi and chi = R A_R turns the line integrals into simple polynomial integrations, and the derivation of Eqs. (40)-(43) is coherent and checkable. The method contains no fitted parameters, and the analytic convergence experiments in Fig. 3 are consistent with the expected Hermite rates once the text/caption discrepancy is resolved. This would be a practically useful tool for field-line tracing and guiding-center simulations, where divergence errors corrupt Poincaré sections and destroy long-term conservation. The demonstration that insufficient smoothness degrades RK convergence is also useful. However, the paper's reproducibility is currently blocked by the missing derivative stencils for m>2, and the quantitative value of the finite-element application is weakened by the absence of a comparison with the existing spline-based reconstruction of Finn and Chacón.

major comments (4)
  1. [Section III, Eqs. (8)-(10)] The manuscript's only recipe for constructing the Taylor coefficients from grid data provides û0, û1, and û2, but the experiments in Fig. 3 and the claims in Section VI require m=3 and m=4, for which û3 and û4 must be supplied to even define the C^m interpolant. The text also does not state the relation between the fine-grid spacing ∆x and the Hermite grid spacing h. If ∆x is merely a fixed fraction of h, the O(∆x^4) error in Eq. (9) becomes an O(h^5) error in the interpolant's first derivative data, which is incompatible with the reported 8th- and 10th-order convergence rates. The authors should either provide the higher-derivative stencils and the required ∆x(h) scaling, or state explicitly that the analytic convergence test in Section V.A used exact derivatives via Eq. (7) rather than the finite-difference pipeline.
  2. [Section V.A, text following Fig. 3] The prose says that 'psi and BR approximations converge at fifth order' and that 'BZ converges at an order decreased by 1,' while the caption of Fig. 3 says that psi and BR converge at 6th, 8th, and 10th order and that BZ converges at 5th, 7th, and 9th order. These statements should be reconciled, and the m values corresponding to the three convergence lines should be identified explicitly. This is part of the central accuracy claim and should not be left ambiguous.
  3. [Section VI, Eq. (64)] The local truncation error formula for a Runge-Kutta method near a third-derivative discontinuity is asserted without derivation or citation. The constants α and C, the regularity assumptions on f, and the meaning of the +/− notation are not defined, so the reader cannot verify the claimed O(∆t^4) degradation. Since this formula is used to explain the observed order reduction, it should be derived in an appendix or replaced by a precise reference.
  4. [Section V.C.2] The compatible finite-element Poincaré plots are presented visually, but no quantitative measure of the reconstructed field's accuracy is given. The paper itself notes that outside-domain and non-convex-region errors can propagate (Section V). For the k=0 case, where the equilibrium field B0 is known, a quantitative comparison of the reconstructed flux surfaces with the equilibrium field would support the claim that the routine 'effectively reproduces the expected magnetic field behavior.'
minor comments (4)
  1. [Section IV.C, Eq. (52)] The interval in Z is written as Z_j2 - h_R/2 < Z < Z_j2 + h_R/2; this should presumably use h_Z/2 rather than h_R/2.
  2. [Section IV, text after Eq. (34)] The word 'poloynomials' should be 'polynomials'.
  3. [Introduction / Section V] The introduction motivates Hermite interpolation as an alternative to the cubic-spline-based reconstruction of Ref. 12, but no numerical comparison with Ref. 12 is included; adding one on the same test cases would make the practical advantages concrete.
  4. [Section III, paragraph after Eq. (10)] The statement that the Taylor coefficients are 'evaluated with centered finite differences' is followed only by formulas for the first two derivative coefficients; the paragraph should either explain how higher derivatives are obtained or point to a reference containing the missing stencils.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the vector-potential reconstruction is derived from first principles, and the high-order convergence claims are validated against analytic benchmark fields rather than fitted to data.

full rationale

The derivation chain is self-contained. Section IV constructs ψ and χ by integrating the discrete/interpolated field components (Eqs. (42)-(43)) and evaluates B = ∇×A with the gauge AZ=0; the divergence-free property is a mathematical identity of the curl, not an output fitted to data. The convergence study in Section V.A uses an analytic field (Eq. (54)) with a fixed, parameter-free safety-factor profile, so the reported 6th, 8th, and 10th-order rates for ψ and BR are genuine external benchmarks rather than consequences of tuning. The Hermite interpolation error theory cited from [16] is independent prior work; the authors' own Hermite overset paper [17] is mentioned only for an optional boundary-condition extension, and the self-authored MHD/FE solvers [24] and [25] supply input data sets rather than justify the interpolation claim. There is no fitted parameter renamed as a prediction and no uniqueness theorem imported from the authors' prior work. The text does leave unspecified the higher-derivative stencils needed to reproduce m=3 and m=4 convergence from discrete data (only Eqs. (8)-(10) are given), and the weak-divergence finite-element projection error is not quantified; these are reproducibility and correctness caveats, not circularity. A score of 1 reflects the presence of minor background self-citations while the central claim remains independent and externally benchmarked.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim rests on standard Hermite interpolation theory and on the assumption that the input B data are smooth and nearly divergence-free. The only manually selected parameters are m, the Hermite grid coarsening ratio, and the finite-difference fine-grid spacing. None are fitted to match a target result. No new physical entities are introduced.

free parameters (3)
  • m: number of Hermite derivatives = m = 1 to 4 in experiments; m >= 3 needed for the mimetic FD field in the guiding-center test
    User-chosen smoothness and order parameter. The central claims about C^m continuity and convergence rates depend on m. It is selected by hand, not fitted to data.
  • Hermite grid coarsening factor (N=(nR-1)/4, M=(nZ-1)/4) = 4
    Convergence experiments use a Hermite grid four times coarser than the data grid. This ratio is chosen by the authors and affects the measured rates.
  • Fine-grid spacing Δx for finite-difference derivative estimates = not stated (fraction of h)
    Equations (8)-(10) use centered finite differences on a finer grid, but the ratio Δx/h and stencils for derivatives beyond second order are not specified. The reconstruction accuracy depends on these choices.
assumptions (4)
  • domain assumption The underlying continuous magnetic field is smooth enough to admit Taylor coefficients up to order m at grid points.
    Hermite interpolation and the finite-difference estimates (8)-(10) assume sufficient local regularity of B. Section III relies on this for error rates.
  • domain assumption The input discrete field is either exactly divergence-free or close enough that projecting it onto the solenoidal space via vector potential reconstruction gives a physically meaningful field.
    The derivation of C1 in Eqs. (40)-(41) uses ∇·B=0. For weakly divergence-free FE data the method acts as a projection whose accuracy is not quantified.
  • standard math Runge-Kutta order conditions require derivatives of the right-hand side up to a certain order, so C^m with m=3 is sufficient for Dormand-Prince 5 order.
    Invoked from Hairer et al. (ref 19) and used in Sections II and VI to interpret m<3 as causing order loss.
  • domain assumption The poloidal data is periodic in φ and can be represented with a discrete Fourier transform with l modes.
    Fourier interpolation in φ, Eqs. (17)-(18), assumes periodicity and that retained modes capture the field. Truncation error is not analyzed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of High order interpolation of magnetic fields with vector potential reconstruction for particle simulations." pith.science (2026). https://pith.science/paper/JPJGZH6D

@misc{pith2026250101523,
  author       = {Pith},
  title        = {Pith review of: High order interpolation of magnetic fields with vector potential reconstruction for particle simulations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JPJGZH6D}},
  note         = {Machine review of arXiv:2501.01523}
}
read the original abstract

We propose a method for interpolating divergence-free continuous magnetic fields via vector potential reconstruction using Hermite interpolation, which ensures high-order continuity for applications requiring adaptive, high-order ordinary differential equation (ODE) integrators, such as the Dormand-Prince method. The method provides C(m) continuity and achieves high-order accuracy, making it particularly suited for particle trajectory integration and Poincar\'e section analysis under optimal integration order and timestep adjustments. Through numerical experiments, we demonstrate that the Hermite interpolation method preserves volume and continuity, which are critical for conserving toroidal canonical momentum and magnetic moment in guiding center simulations, especially over long-term trajectory integration. Furthermore, we analyze the impact of insufficient derivative continuity on Runge-Kutta schemes and show how it degrades accuracy at low error tolerances, introducing discontinuity-induced truncation errors. Finally, we demonstrate performant Poincar\'e section analysis in two relevant settings of field data collocated from finite element meshes

Figures

Figures reproduced from arXiv: 2501.01523 by the authors.

Figure 1
Figure 1. FIG. 1: Convergence rates for 4th order (blue curves) and 5th order (red curves) solutions to the Dormand-Prince [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2: 9th order Hermite interpolation of analytically constructed magnetic field (54), and reconstructed poloidal flux [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3: Hermite interpolation convergence for analytic magnetic fields. Solid lines display the relative error arising when [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: FIG. 4: Guiding center equations conservation-convergence test. Average, relative conservation error of [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5: 9th order accurate Hermite interpolation of ITER magnetic fields, reconstructing the poloidal flux function and [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6: Meshes used to produce Poincaré plots. [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]
Figure 7
Figure 7. Figure 7: FIG. 7: Poincaré plots at different times using the proposed flux reconstruction of an ITER major disruption event [PITH_FULL_IMAGE:figures/full_fig_p013_7.png]
Figure 8
Figure 8. Figure 8: FIG. 8: Flux function for the toroidally averaged field and Poincaré plots at different times. [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]
Figure 9
Figure 9. Figure 9: FIG. 9: Mesh and space setup for finite-element based [PITH_FULL_IMAGE:figures/full_fig_p014_9.png]
Figure 10
Figure 10. Figure 10: FIG. 10: Poincaré plots for equilibrium magnetic field perturbed by a helical coil, for perturbation strengths [PITH_FULL_IMAGE:figures/full_fig_p015_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 19 canonical work pages

  1. [1]

    Zhang \ and\ author X

    author author M. Zhang \ and\ author X. Feng ,\ title title A comparative study of divergence cleaning methods of magnetic field in the solar coronal numerical simulation , \ 10.3389/fspas.2016.00006 journal journal Frontiers in Astronomy and Space Sciences \ volume 3 ( year 2016 ),\ 10.3389/fspas.2016.00006 NoStop

    arXiv 2016

  2. [2]

    author author G. T\'oth ,\ title title The b=0 constraint in shock-capturing magnetohydrodynamics codes , \ https://doi.org/10.1006/jcph.2000.6519 journal journal Journal of Computational Physics \ volume 161 ,\ pages 605--652 ( year 2000 ) NoStop

    arXiv

  3. [3]

    Dedner , author F

    author author A. Dedner , author F. Kemm , author D. Kröner , author C.-D. \ Munz , author T. Schnitzer , \ and\ author M. Wesenberg ,\ title title Hyperbolic divergence cleaning for the mhd equations , \ https://doi.org/10.1006/jcph.2001.6961 journal journal Journal of Computational Physics \ volume 175 ,\ pages 645--673 ( year 2002 ) NoStop

    arXiv

  4. [4]

    Lipnikov , author G

    author author K. Lipnikov , author G. Manzini , \ and\ author M. Shashkov ,\ title title Mimetic finite difference method , \ https://doi.org/10.1016/j.jcp.2013.07.031 journal journal Journal of Computational Physics \ volume 257 ,\ pages 1163--1227 ( year 2014 ) ,\ note physics-compatible numerical methods NoStop

    doi:10.1016/j.jcp.2013.07.031 2013

  5. [5]

    Lipnikov , author J

    author author K. Lipnikov , author J. Moulton , \ and\ author D. Svyatskiy ,\ title title A multilevel multiscale mimetic (m3) method for two-phase flows in porous media , \ https://doi.org/10.1016/j.jcp.2008.03.029 journal journal Journal of Computational Physics \ volume 227 ,\ pages 6727--6753 ( year 2008 ) NoStop

    work page doi:10.1016/j.jcp.2008.03.029 2008

  6. [6]

    Beir \ a o da Veiga , author V

    author author L. Beir \ a o da Veiga , author V. Gyrya , author K. Lipnikov , \ and\ author G. Manzini ,\ title title Mimetic finite difference method for the stokes problem on polygonal meshes , \ https://doi.org/10.1016/j.jcp.2009.06.034 journal journal Journal of Computational Physics \ volume 228 ,\ pages 7215--7232 ( year 2009 ) NoStop

    work page doi:10.1016/j.jcp.2009.06.034

  7. [7]

    author author L. Beir \ a o Da Veiga ,\ title title A mimetic discretization method for linear elasticity , \ 10.1051/m2an/2010001 journal journal ESAIM: Mathematical Modelling and Numerical Analysis \ volume 44 ,\ pages 231--250 ( year 2010 ) NoStop

    work page arXiv

  8. [8]

    da Veiga , author J

    author author B. da Veiga , author J. Droniou , \ and\ author G. Manzini ,\ title title A unified approach for handling convection terms in finite volumes and mimetic discretization methods for elliptic problems , \ 10.1093/imanum/drq018 journal journal IMA Journal of Numerical Analysis \ volume 31 ,\ pages 1357--1401 ( year 2010 ) ,\ http://arxiv.org/abs...

    work page doi:10.1093/imanum/drq018

Show all 26 references
  1. [9]

    Cangiani , author G

    author author A. Cangiani , author G. Manzini , \ and\ author A. Russo ,\ title title Convergence analysis of the mimetic finite difference method for elliptic problems , \ 10.1137/080717560 journal journal SIAM Journal on Numerical Analysis \ volume 47 ,\ pages 2612--2637 ( y...

    doi

  2. [10]

    Brezzi , author K

    author author F. Brezzi , author K. Lipnikov , author M. Shashkov , \ and\ author V. Simoncini ,\ title title A new discretization methodology for diffusion problems on generalized polyhedral meshes , \ https://doi.org/10.1016/j.cma.2006.10.028 journal journal Computer Methods...

    doi

  3. [11]

    author author K. Yee ,\ title title Numerical solution of initial boundary value problems involving maxwell's equations in isotropic media , \ 10.1109/TAP.1966.1138693 journal journal IEEE Transactions on Antennas and Propagation \ volume 14 ,\ pages 302--307 ( year 1966 ) NoStop

    1966

  4. [12]

    author author J. M. \ Finn \ and\ author L. Chac \'o n ,\ title title Volume preserving integrators for solenoidal fields on a grid , \ @noop journal journal Physics of Plasmas \ volume 12 ( year 2005 ) NoStop

    2005

  5. [13]

    author author C. J. \ McDevitt , author Z. Guo , \ and\ author X.-Z. \ Tang ,\ title title Avalanche mechanism for runaway electron amplification in a tokamak plasma , \ @noop journal journal Plasma Physics and Controlled Fusion \ volume 61 ,\ pages 054008 ( year 2019 ) NoStop

    2019

  6. [14]

    author author C. J. \ McDevitt \ and\ author X.-Z. \ Tang ,\ title title Runaway electron current reconstitution after a nonaxisymmetric magnetohydrodynamic flush , \ 10.1103/PhysRevE.108.L043201 journal journal Phys. Rev. E \ volume 108 ,\ pages L043201 ( year 2023 ) NoStop

    doi

  7. [15]

    Goodrich , author T

    author author J. Goodrich , author T. Hagstrom , \ and\ author J. Lorenz ,\ title title Hermite methods for hyperbolic initial-boundary value problems , \ @noop journal journal Mathematics of computation \ volume 75 ,\ pages 595--630 ( year 2006 ) NoStop

    2006

  8. [16]

    Appel\" o , author T

    author author D. Appel\" o , author T. Hagstrom , \ and\ author A. Vargas ,\ title title Hermite methods for the scalar wave equation , \ 10.1137/18M1171072 journal journal SIAM Journal on Scientific Computing \ volume 40 ,\ pages A3902--A3927 ( year 2018 ) ,\ http://arxiv.org...

    doi

  9. [17]

    Beznosov \ and\ author D

    author author O. Beznosov \ and\ author D. Appel \"o ,\ title title Hermite-discontinuous galerkin overset grid methods for the scalar wave equation , \ @noop journal journal Communications on Applied Mathematics and Computation \ volume 3 ,\ pages 391--418 ( year 2021 ) NoStop

    2021

  10. [18]

    Lekien \ and\ author J

    author author F. Lekien \ and\ author J. Marsden ,\ title title Tricubic interpolation in three dimensions , \ @noop journal journal International Journal for Numerical Methods in Engineering \ volume 63 ,\ pages 455--471 ( year 2005 ) NoStop

    2005

  11. [19]

    Hairer , author G

    author author E. Hairer , author G. Wanner , \ and\ author S. P. \ N rsett ,\ @noop title Solving Ordinary Differential Equations I: Nonstiff Problems ,\ edition 2nd \ ed.\ ( publisher Springer ,\ year 1993 ) NoStop

    1993

  12. [20]

    author author J. R. \ Dormand \ and\ author P. J. \ Prince ,\ title title A family of embedded runge-kutta formulae , \ @noop journal journal Journal of computational and applied mathematics \ volume 6 ,\ pages 19--26 ( year 1980 ) NoStop

    1980

  13. [21]

    author author B. Fornberg ,\ title title Generation of finite difference formulas on arbitrarily spaced grids , \ @noop journal journal Mathematics of computation \ volume 51 ,\ pages 699--706 ( year 1988 ) NoStop

    1988

  14. [22]

    Hagstrom \ and\ author D

    author author T. Hagstrom \ and\ author D. Appel \"o ,\ title title Solving pdes with hermite interpolation , \ \ ( organization Springer ,\ year 2015 )\ pp.\ pages 31--49 NoStop

    2015

  15. [23]

    author author J. R. \ Cary \ and\ author A. J. \ Brizard ,\ title title Hamiltonian theory of guiding-center motion , \ @noop journal journal Reviews of modern physics \ volume 81 ,\ pages 693--738 ( year 2009 ) NoStop

    2009

  16. [24]

    Jorti , author Q

    author author Z. Jorti , author Q. Tang , author K. Lipnikov , \ and\ author X.-Z. \ Tang ,\ title title A mimetic finite difference based quasi-static magnetohydrodynamic solver for force-free plasmas in tokamak disruptions , \ @noop journal journal arXiv preprint arXiv:2303....

    2023 arXiv

  17. [25]

    Bonilla , author J

    author author J. Bonilla , author J. N. \ Shadid , author X.-Z. \ Tang , author M. M. \ Crockatt , author P. Ohm , author E. G. \ Phillips , author R. P. \ Pawlowski , author S. Conde , \ and\ author O. Beznosov ,\ title title On a fully-implicit vms-stabilized fe formulation ...

  18. [26]

    author author H. M. \ Smith , author A. H. \ Boozer , \ and\ author P. Helander ,\ title title Passive runaway electron suppression in tokamak disruptions , \ 10.1063/1.4813255 journal journal Physics of Plasmas \ volume 20 ,\ pages 072505 ( year 2013 ) ,\ note \_eprint: https...

    doi

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.