Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

ThinCurr: An open-source 3D thin-wall eddy current modeling code for the analysis of large-scale systems of conducting structures

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read ThinCurr is an open-source boundary finite element code for 3D thin-wall eddy currents that compresses the dense inductance matrix to scale to whole-device tokamak models while matching established verification benchmarks.

desk verdict Solid, honest software/methods paper for thin-wall eddy current modeling; the new integration and open-source HODLR implementation are verified well, though the thin-wall approximation is only tested at one point in parameter space. read the letter →

arxiv 2412.14962 v2 pith:SVOR4SS2 submitted 2024-12-19 physics.plasm-ph

Christopher Hansen , Alexander Battey , Anson Braun , Sander Miller , Michael Lagieski , Ian Stewart , Ryan Sweeney , Carlos Paz-Soldan This is my paper · ORCID
classification physics.plasm-ph
keywords thin-wallapproximationeddycurrentsboundaryfiniteelementmethodhierarchicaloff-diagonallow-rankcompressionadaptivecrosstokamakdisruptionloadscurrentpotentialopen-sourceplasmamodeling
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

ThinCurr is an open-source solver for inductively coupled eddy currents in thin conducting structures, aimed at fusion devices where vacuum-vessel walls and other sheet-like conductors carry large transient currents during disruptions. The paper claims that a boundary finite element method with a surface current potential correctly captures these currents, and that hierarchical off-diagonal low-rank compression of the otherwise dense inductance matrix makes full-device models practical by cutting memory and setup cost from quadratic to nearly linear scaling. The implementation is verified through cross-code comparisons with an established thin-wall research code and a commercial finite-element package, including a tokamak disruption benchmark, with close agreement in currents, forces, and time scales. A reader interested in tokamak structural design should care because this is a freely available route to eddy-current loads on realistic 3D geometry.

What carries the argument

The central object is the boundary finite element formulation on surfaces: the current is written as $\mathbf{J}_s = \nabla\chi\times\hat{\mathbf n}$, the inductance matrix $L$ is assembled from double surface integrals of the Biot-Savart kernel, and the resistance matrix $R$ uses surface resistivity $\eta_s = \eta/t_w$, where $t_w$ is wall thickness. Topological correctness is carried by "hole" elements, which represent current loops around distinct homological cycles, and "closure" elements, which fix the gauge on closed surfaces; these are located automatically by a greedy homology-basis algorithm. The scalability claim is carried by the HODLR/ACA+ compression of $L$, supported by block-Jacobi preconditioning so that iterative solves retain nearly linear scaling.

What would settle it

Take a wall geometry like the paper's benchmark, with walls 2 to 4 cm thick and an 8.7 MA current quench whose exponential decay time is 1.385 ms, and compare ThinCurr's total currents and forces against a reference model that resolves the wall thickness across the same transient; if the disagreement exceeds the few-percent level the paper reports, the thin-wall approximation—not the solver implementation—is the failing link.

Watch

Extended reading notes

Core claim

The central claim is that representing each conducting structure as an oriented surface with a scalar current potential $\chi$, so that the surface current is $\mathbf{J}_s = \nabla\chi\times\hat{\mathbf n}$, and discretizing that potential with linear finite elements on an unstructured triangular mesh yields a correct and practical model of eddy currents in large-scale systems. The dense inductance matrix $L$ that couples every surface element is compressed with a hierarchical off-diagonal low-rank approximation whose far blocks are built by adaptive cross approximation, restoring nearly linear scaling in memory and time. The paper reports that eigenvalue, time-domain, and frequency-domain results match a community thin-wall code, and that disruption-induced currents and forces computed for a full tokamak vacuum-vessel model match a commercial volume-conducting finite-element solver, with all comparisons showing close agreement.

Load-bearing premise

The load-bearing premise is the thin-wall approximation: each structure is treated as a surface whose only thickness information is the surface resistivity, so currents flowing through the wall thickness are not resolved; if the electromagnetic skin depth is comparable to or smaller than the wall thickness during a transient, the model misses through-thickness current structure regardless of how accurately the matrices are built.

Editorial extensions

If this is right

  • Models with on the order of one hundred thousand elements fit in memory and solve in nearly linear time, bringing whole-device disruption load calculations onto workstation-class hardware.
  • Automatic detection of holes and closures lets users build solvable models directly from triangular meshes without manually specifying topological loops.
  • The same model supports time-domain, frequency-domain, and eigenvalue analyses, so design cycles can move from decay-time scoping to full transient loads without rebuilding the mesh.
  • Agreement with a commercial finite-element solver on a realistic tokamak disruption case supports using ThinCurr for design-cycle estimates of eddy-current forces on thin-walled structures.
  • Integration into workflows for resistive-wall mode stability, current reconstruction, and coil optimization is listed as ongoing work in the paper.

Reading between the lines

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

  • I infer that users should check skin depth before applying the code: whenever the transient's skin depth is comparable to wall thickness, the thin-wall representation is the likely error source even though the numerical method is sound.
  • The compression tolerance gives a tunable accuracy-versus-speed knob, so a natural extension is to verify convergence of integrated quantities such as total force, not just eigenvalue spectra, as the tolerance is tightened.
  • The method is geometry-general, so the same open-source machinery could serve eddy-current studies outside fusion, such as induced loads in accelerator vacuum chambers, electromagnetic forming, shielding design, or nondestructive evaluation of thin sheet structures.
  • Because the code is open source, community additions such as higher-order elements, curved triangles, and source/sink support at T-junctions are plausible next steps that would expand the class of representable geometries.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper presents ThinCurr, an open-source boundary finite element code for thin-wall eddy current modeling in fusion devices, developed within the Open FUSION Toolkit. The authors describe the mathematical formulation using a surface current potential, the construction of inductance and resistance matrices, the treatment of multiply connected geometries via automatic homology-based hole and closure detection, and three solution modes (time domain, frequency domain, and eigenvalue). A central contribution is the use of hierarchical off-diagonal low-rank (HODLR) compression with ACA+ to replace the dense inductance matrix, with scaling tests showing O(N log N) memory and solution time. Verification is performed by cross-code comparisons against VALEN (eigenvalue, time-domain, and frequency-domain tests) and against Ansys for a SPARC tokamak disruption quench. The paper concludes that the numerical implementation is correct and that the HODLR compression enables scalability to whole-device models, with all comparisons showing excellent agreement.

Significance. If the claims hold, ThinCurr is a useful, freely available tool for large-scale eddy current load calculations in fusion device design, and the paper provides a detailed and largely clear description of the numerical methods. The eigenvalue verification is a strength: the relative error between ThinCurr and VALEN decreases at the expected O(dx^2) rate across 100 modes and three geometries, indicating that the dense-matrix L and R assembly is correct. The scaling measurements for HODLR are also valuable and provide concrete performance data. The automatic homology-based hole detection is a practical usability contribution. However, two gaps currently limit the strength of the central claims: the HODLR/ACA+ path is not validated for accuracy, and the only volumetric-solver benchmark (Ansys) lacks quantitative error metrics and covers only a single operating point. These issues are fixable and do not appear to undermine the underlying method, but they need to be addressed before the paper can fully support its claims.

major comments (3)
  1. [Sec. 4, Figs. 8-10] The HODLR/ACA+ compression is characterized only in terms of memory and time scaling; no test demonstrates that the compressed inductance matrix or the resulting eigenvalues, time traces, or forces match the dense solution to the specified tolerance. A user of the code needs to know that the compressed solution is accurate, not merely fast. Please add a convergence study on a moderately sized model (e.g., ~10^4 elements) comparing dense and HODLR results for a few eigenvalues and a representative time-domain response, with errors reported as a function of the SVD/ACA+ tolerance. Without this, the claim that HODLR 'enables scalability to whole device models' is a performance claim without a corresponding correctness verification.
  2. [Sec. 5.2, Fig. 18] The Ansys benchmark is the only test against a volumetric solver, yet the paper reports no quantitative error metric for the toroidal current and force traces, relying instead on the statement 'excellent agreement.' Moreover, the single 8.7 MA quench has a 1.385 ms exponential decay time, for which the diffusion skin depth in the 2-4 cm thick walls can be comparable to the wall thickness. The paper should report the skin depth-to-wall thickness ratio for this case and discuss the expected range of validity of the thin-wall approximation for faster quenches or thicker sections. Adding a second Ansys comparison with a faster decay time (or otherwise different skin-depth regime) would substantially strengthen the claim that the approximation is appropriate for design-limiting events.
  3. [Secs. 5.1.2 and 5.1.3] The time-domain and frequency-domain comparisons against VALEN are presented only as overlaid plots with the qualitative statement 'excellent agreement.' The verification claim would be falsifiable and more convincing if the authors reported quantitative measures, such as the maximum or RMS relative error over the plotted sensor signals, or a normalized error norm. The eigenvalue comparison already has such metrics; the time- and frequency-domain tests should be held to the same standard.
minor comments (5)
  1. [Sec. 4] The acronym 'HODLR' is introduced as 'HOLDR' in the first paragraph of Section 4 ('...utilizes an Hierarchical Off-Diagonal Low-Rank (HOLDR) approximation...'); the correct acronym is used elsewhere and should be made consistent.
  2. [Sec. 5.1.3] In the text describing the frequency-domain test, 'ThinCurr used the preconditioned GRMES approach' should be 'GMRES'.
  3. [Sec. 4.1] The text reads 'Adaptive Cross-Appoximation+ (ACA+)'; the word 'Approximation' is misspelled.
  4. [Sec. 5.2] The sentence 'Although not plotted, a comparison of the local current densities at the midplane of the inner and outer VVs also shows similar agreement' makes an unverifiable claim; either add the plot or remove the statement.
  5. [Sec. 2.1.1, Eq. (6)] The adaptive quadrature order formula p = log(err)/log(1 - dmin/dmax) would benefit from a brief definition of 'err' (target error) and a note on the expected range of p, since the denominator is negative and the intended behavior may not be immediately clear to readers.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: ThinCurr's central validation is direct cross-code comparison against two independent codes, VALEN and Ansys, with no fitted input being renamed as a prediction.

full rationale

The paper's central claim is that ThinCurr correctly implements the thin-wall BFEM eddy-current model and that the HODLR/ACA+ compression enables scalable models. This claim is verified externally, not by construction. Section 5.1 compares ThinCurr to the independently developed VALEN code on eigenvalue, time-domain, and frequency-domain test cases, using matched geometry, resistivity, and drive conditions; the observed O(Δx^2) convergence of the eigenvalue comparison follows from the linear discretizations and is not obtained by fitting ThinCurr to VALEN. Section 5.2 compares ThinCurr to Ansys on a SPARC disruption benchmark, again with matched coil currents, wall thicknesses encoded in resistivity, and the same 8.7 MA quench input; total toroidal current and forces agree without any calibration parameter. The paper's self-citations to the Open FUSION Toolkit and prior fusion papers are infrastructure or motivation references, not the source of the validation results. The Sec. 1 limitation statement that 'thick-wall effects may be important in some configurations and events' is a modeling-validity caveat, not a circular derivation step: it identifies where the thin-wall approximation could fail, but ThinCurr's equations and verification do not assume the answer they purport to predict. No fitted parameter, self-citation chain, or definitional identity forces the reported agreement, so the circularity burden is not met.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central claim relies on standard magnetoquasistatic physics (Eq. 1), the thin-wall surface-current representation (Eq. 2), a manifold-mesh restriction (Sec. 2), low-rank compressibility of the inductance matrix (Sec. 4), and an unproved greedy-topology completeness for linked holes (Sec. 2.1.3). There are no data-fitted free parameters in the central claim.

assumptions (5)
  • domain assumption Thin-wall approximation: currents are represented as a surface current Js = grad(chi) x n with no through-thickness variation.
    Introduced in Sec. 1 and Eq. 2; if walls are thick compared with skin depth, the model omits physics.
  • domain assumption The triangulated surfaces are manifolds with no edge shared by more than two triangles.
    Stated in Sec. 2 before Eq. 2; T-shaped connections are disallowed, limiting geometric generality.
  • domain assumption The greedy homology algorithm produces a complete set of hole and closure elements for the target geometries.
    Sec. 2.1.3; for linked holes the authors state they do not prove general correctness.
  • standard math Off-diagonal blocks of the inductance matrix have rapidly decaying singular values and can be represented to tolerance by low-rank factors.
    Assumed in Sec. 4 as the basis of HODLR/ACA+; standard for integral-equation kernels but not proven in the paper.
  • standard math The magnetoquasistatic circuit equation d/dt(LI)+RI=V governs the eddy-current dynamics.
    Eq. 1; standard model in the cited eddy-current literature.

how reviews work

0 comments
Cite this review

Pith. "Pith review of ThinCurr: An open-source 3D thin-wall eddy current modeling code for the analysis of large-scale systems of conducting structures." pith.science (2026). https://pith.science/paper/SVOR4SS2

@misc{pith2026241214962,
  author       = {Pith},
  title        = {Pith review of: ThinCurr: An open-source 3D thin-wall eddy current modeling code for the analysis of large-scale systems of conducting structures},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SVOR4SS2}},
  note         = {Machine review of arXiv:2412.14962}
}
read the original abstract

In this paper we present a new thin-wall eddy current modeling code, ThinCurr, for studying inductively-coupled currents in 3D conducting structures -- with primary application focused on the interaction between currents flowing in coils, plasma, and conducting structures of magnetically-confined plasma devices. The code utilizes a boundary finite element method on an unstructured, triangular grid to accurately capture device structures. The new code, part of the broader Open FUSION Toolkit, is open-source and designed for ease of use without sacrificing capability and speed through a combination of Python, Fortran, and C/C++ components. Scalability to large models is enabled through use of hierarchical off-diagonal low-rank compression of the inductance matrix, which is otherwise dense. Ease of handling large models of complicated geometry is further supported by automatic determination of supplemental elements through a greedy homology approach. A detailed description of the numerical methods of the code and verification of the implementation of those methods using cross-code comparisons against the VALEN code and Ansys commercial analysis software is shown.

Figures

Figures reproduced from arXiv: 2412.14962 by the authors.

Figure 1
Figure 1. Example of longest-lived eddy current structure predicted [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. ThinCurr mesh and single-valued current potential [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Surface current corresponding to the single-valued current [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (11 more)
Figure 5
Figure 5. Figure 5: Automatic determination of boundary (blue/ [PITH_FULL_IMAGE:figures/full_fig_p004_5.png]
Figure 6
Figure 6. Figure 6: Example triangle fan (yellow), temporarily adding during [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]
Figure 7
Figure 7. Figure 7: ThinCurr mesh colored by block number for a single level of [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 9
Figure 9. Figure 9: Scaling of time to solution for a time-dependent solve with [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]
Figure 8
Figure 8. Figure 8: Scaling of the memory required for the L matrix (top) and [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 10
Figure 10. Figure 10: Relative time required for different phases of time￾dependent solve for the Direct and ACA+ approaches. Each case (bar stack) is normalized separately. As the problem size grows, inverting the system matrix (“Inv”) dominates computation time for the Direct approach. I…
Figure 12
Figure 12. Figure 12: Maximum (circles) and average (squares) relative error [PITH_FULL_IMAGE:figures/full_fig_p010_12.png]
Figure 14
Figure 14. Figure 14: Signals for the upper (blue) and lower (orange) flux [PITH_FULL_IMAGE:figures/full_fig_p010_14.png]
Figure 15
Figure 15. Figure 15: Current distribution (vectors and shading) of the real com [PITH_FULL_IMAGE:figures/full_fig_p011_15.png]
Figure 16
Figure 16. Figure 16: Real (top) and imaginary (bottom) components of the field [PITH_FULL_IMAGE:figures/full_fig_p011_16.png]
Figure 18
Figure 18. Figure 18: Comparison between Ansys and ThinCurr results for total [PITH_FULL_IMAGE:figures/full_fig_p012_18.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Electromagnetic System Conceptual Design for a Negative Triangularity Tokamak

    physics.plasm-ph 2025-01 conditional novelty 6.0 of 10

    A conceptual electromagnetic design shows that a compact 3 tesla, 0.75 megaamp negative triangularity tokamak can be built from existing copper magnet technology.

Reference graph

Works this paper leans on

53 extracted references · 38 canonical work pages · cited by 1 Pith paper

  1. [1]

    Battey, C

    A. Battey, C. Hansen, D. Garnier, D. Weisberg, C. Paz-Soldan, R. Sweeney, R. Tinguely, A. Creely, Design of passive and structural conductors for tokamaks using thin-wall eddy cur- rent modeling, Nuclear Fusion 64 (1) (2023) 016010. doi: 10.1088/1741-4326/ad0bcf

    work page doi:10.1088/1741-4326/ad0bcf 2023

  2. [2]

    V . Izzo, A. Battey, R. Tinguely, R. Sweeney, C. Hansen, Bound- ary condition effects on runaway electron mitigation coil model- ing for the SPARC and DIII-D tokamaks, Nuclear Fusion 64 (6) (2024) 066003. doi:10.1088/1741-4326/ad3c52. 12

    work page doi:10.1088/1741-4326/ad3c52 2024

  3. [3]

    Bettini, N

    P. Bettini, N. Marconato, M. F. Palumbo, S. Peruzzo, R. Specogna, R. Albanese, G. Rubinacci, S. Ventre, F. Villone, Numerical modeling of 3D halo current path in ITER structures, Fusion Engineering and Design 88 (6) (2013) 529–532, pro- ceedings of the 27th Symposium On Fusion Technology (SOFT- 27); Li ´ege, Belgium, September 24-28, 2012. doi:https: //do...

    doi:10.1016/j.fusengdes.2012.11.017 2013

  4. [4]

    Albanese, B

    R. Albanese, B. Carpentieri, M. Cavinato, S. Minucci, R. Pal- maccio, A. Portone, G. Rubinacci, P. Testoni, S. Ventre, F. Vil- lone, Effects of asymmetric vertical disruptions on ITER com- ponents, Fusion Engineering and Design 94 (2015) 7–21. doi: https://doi.org/10.1016/j.fusengdes.2015.02.034

    work page doi:10.1016/j.fusengdes.2015.02.034 2015

  5. [5]

    Battey, J

    A. Battey, J. Hanson, J. Bialek, F. Turco, G. Navratil, N. Logan, Simultaneous stabilization and control of the n = 1 and n = 2 resistive wall mode, Nuclear Fusion 63 (6) (2023) 066025.doi: 10.1088/1741-4326/accd81

    work page doi:10.1088/1741-4326/accd81 2023

  6. [6]

    Albanese, G

    R. Albanese, G. Rubinacci, Finite element methods for the so- lution of 3D eddy current problems, V ol. 102 of Advances in Imaging and Electron Physics, Elsevier, 1997, pp. 1–86. doi: https://doi.org/10.1016/S1076-5670(08)70121-6

    work page doi:10.1016/s1076-5670(08)70121-6 1997

  7. [7]

    H ¨olzl, P

    M. H ¨olzl, P. Merkel, G. T. A. Huysmans, E. Nardon, E. Strum- berger, R. McAdams, I. Chapman, S. G ¨unter, K. Lackner, Cou- pling JOREK and STARW ALL codes for non-linear resistive- wall simulations, Journal of Physics: Conference Series 401 (1) (2012) 012010. doi:10.1088/1742-6596/401/1/012010

    doi:10.1088/1742-6596/401/1/012010 2012

  8. [8]

    Bialek, A

    J. Bialek, A. H. Boozer, M. E. Mauel, G. A. Navratil, Modeling of active control of external magnetohydrodynamic instabilities 8 (5) 2170–2180. doi:10.1063/1.1362532

    doi:10.1063/1.1362532

Show all 53 references
  1. [9]

    URL https://www.ansys.com/

    ANSYS, Inc, Ansys © academic research mechanical. URL https://www.ansys.com/

  2. [10]

    Hansen, I

    C. Hansen, I. Stewart, D. Burgess, M. Pharr, S. Guizzo, F. Lo- gak, A. Nelson, C. Paz-Soldan, TokaMaker: An open-source time-dependent grad-shafranov tool for the design and modeling of axisymmetric fusion devices, Computer Physics Communica- tions 298 (2024) 109111. doi:https...

    2024

  3. [11]

    Hansen, D

    C. Hansen, D. Burgess, M. Pharr, john4255, josephwj, S. Guizzo, Openfusiontoolkit /openfusiontoolkit: v1.0.0-beta6 (Jun. 2025). doi:10.5281/zenodo.15603027

    2025 doi

  4. [12]

    Landreman, An improved current potential method for fast computation of stellarator coil shapes, Nuclear Fusion 57 (4) (2017) 046003

    M. Landreman, An improved current potential method for fast computation of stellarator coil shapes, Nuclear Fusion 57 (4) (2017) 046003. doi:10.1088/1741-4326/aa57d4

    2017 doi

  5. [13]

    E. Paul, M. Landreman, A. Bader, W. Dorland, An ad- joint method for gradient-based optimization of stellarator coil shapes, Nuclear Fusion 58 (7) (2018) 076015. doi:10.1088/ 1741-4326/aac1c7

    2018

  6. [14]

    Albanese, E

    R. Albanese, E. Coccorese, R. Martone, G. Miano, G. Rubi- nacci, Periodic solutions of nonlinear eddy current problems in three-dimensional geometries, IEEE Transactions on Magnetics 28 (2) (1992) 1118–1121. doi:10.1109/20.123880

    1992 doi

  7. [15]

    Miano, F

    G. Miano, F. Villone, A surface integral formulation of maxwell equations for topologically complex conducting domains, IEEE Transactions on Antennas and Propagation 53 (12) (2005) 4001–4014. doi:10.1109/TAP.2005.859898

    2005

  8. [16]

    Bettini, R

    P. Bettini, R. Specogna, A boundary integral method for com- puting eddy currents in thin conductors of arbitrary topology, IEEE Transactions on Magnetics 51 (3) (2015) 1–4. doi: 10.1109/TMAG.2014.2347894

    2015

  9. [17]

    Ferguson, X

    A. Ferguson, X. Zhang, G. Stroink, A complete linear discretiza- tion for calculating the magnetic field using the boundary el- ement method, IEEE Transactions on Biomedical Engineering 41 (5) (1994) 455–460. doi:10.1109/10.293220

    1994 doi

  10. [18]

    Van Oosterom, J

    A. Van Oosterom, J. Strackee, The solid angle of a plane trian- gle, IEEE Transactions on Biomedical Engineering BME-30 (2) (1983) 125–126. doi:10.1109/TBME.1983.325207

    1983

  11. [19]

    Cools, A

    R. Cools, A. Haegemans, Algorithm 824: Cubpack: a pack- age for automatic cubature; framework description, ACM Trans. Math. Softw. 29 (3) (2003) 287–296. doi:10.1145/838250. 838253

    2003 doi

  12. [20]

    S. E. Mousavi, N. Sukumar, Generalized du ffy transformation for integrating vertex singularities, Computational Mechanics 45 (2) (2010) 127–140. doi:10.1007/s00466-009-0424-1

    2010 doi

  13. [21]

    Erickson, K

    J. Erickson, K. Whittlesey, Greedy optimal homotopy and ho- mology generators, in: Proceedings of the Sixteenth Annual ACM-SIAM Symposium on Discrete Algorithms, SODA ’05, Society for Industrial and Applied Mathematics, USA, 2005, p. 1038–1046

    2005

  14. [22]

    URL https://cubit.sandia.gov/

    Sandia National Laboratories, Cubit ©. URL https://cubit.sandia.gov/

  15. [23]

    A. Dhar, V . Natarajan, A. Rathod, Fast algorithms for mini- mum homology basis, Discrete & Computational Geometry (Jul 2024). doi:10.1007/s00454-024-00680-8

    2024 doi

  16. [24]

    H. J. S. Smith, J. J. Sylvester, Xv. on systems of linear in- determinate equations and congruences, Philosophical Trans- actions of the Royal Society of London 151 (1861) 293–326. doi:10.1098/rstl.1861.0016

  17. [25]

    Pellikka, S

    M. Pellikka, S. Suuriniemi, L. Kettunen, C. Geuzaine, Homol- ogy and cohomology computation in finite element modeling, SIAM Journal on Scientific Computing 35 (5) (2013) B1195– B1214. doi:10.1137/130906556

    2013 doi

  18. [26]

    Landreman, S

    M. Landreman, S. Hurwitz, T. M. A. Jr, E fficient calculation of self magnetic field, self-force, and self-inductance for electro- magnetic coils (2023). arXiv:2310.12087

    2023 arXiv

  19. [27]

    Reimerdes, J

    H. Reimerdes, J. Bialek, M. Chance, M. Chu, A. Garofalo, P. Gohil, Y . In, G. Jackson, R. Jayakumar, T. Jensen, J. Kim, R. L. Haye, Y . Liu, J. Menard, G. Navratil, M. Okabayashi, J. Scoville, E. Strait, D. Szymanski, H. Takahashi, Measure- ment of resistive wall mode stabilit...

    2005

  20. [28]

    Karypis, V

    G. Karypis, V . Kumar, A fast and high quality multilevel scheme for partitioning irregular graphs, SIAM Journal on Sci- entific Computing 20 (1) (1998) 359–392. doi:10.1137/ S1064827595287997

    1998

  21. [29]

    J. D. Hanson, The virtual-casing principle and helmholtz’s the- orem, Plasma Physics and Controlled Fusion 57 (11) (2015) 115006. doi:10.1088/0741-3335/57/11/115006

    2015 doi

  22. [30]

    Okabayashi, J

    M. Okabayashi, J. Bialek, M. S. Chance, M. S. Chu, E. D. Fredrickson, A. M. Garofalo, R. Hatcher, T. H. Jensen, L. C. Johnson, R. J. L. Haye, G. A. Navratil, H. Reimerdes, J. T. Scov- ille, E. J. Strait, A. D. Turnbull, M. L. Walker, the DIII–D Team, Stabilization of the resis...

    2002 doi

  23. [31]

    Isernia, N

    N. Isernia, N. Schwarz, F. J. Artola, S. Ventre, M. Hoelzl, G. Ru- binacci, F. Villone, J. Team, Self-consistent coupling of jorek and cariddi: On the electromagnetic interaction of 3d toka- mak plasmas with 3d volumetric conductors, Physics of Plasmas 30 (11) (2023) 113901. d...

    2023 doi

  24. [32]

    Komzsik, The Lanczos method: evolution and application, SIAM, 2003

    L. Komzsik, The Lanczos method: evolution and application, SIAM, 2003

    2003

  25. [33]

    R. B. Lehoucq, D. C. Sorensen, C. Yang, ARPACK users’ guide: solution of large-scale eigenvalue problems with implic- itly restarted Arnoldi methods, SIAM, 1998

    1998

  26. [34]

    URL https://github.com/opencollab/arpack-ng

    ARPACK-NG. URL https://github.com/opencollab/arpack-ng

  27. [35]

    T. A. Davis, Algorithm 832: UMFPACK v4.3— an unsymmetric-pattern multifrontal method, ACM Trans. Math. Softw. 30 (2) (2004) 196–199. doi: 13 10.1145/992200.992206

    2004

  28. [36]

    J. W. Demmel, S. C. Eisenstat, J. R. Gilbert, X. S. Li, J. W. H. Liu, A supernodal approach to sparse partial pivoting, SIAM J. Matrix Analysis and Applications 20 (3) (1999) 720–755

    1999

  29. [37]

    Schenk, K

    O. Schenk, K. G ¨artner, W. Fichtner, Efficient sparse LU factor- ization with left-right looking strategy on shared memory mul- tiprocessors, BIT Numerical Mathematics 40 (2000) 158 – 176. doi:10.1023/A:1022326604210

    2000 doi

  30. [38]

    Y . Bao, Z. Liu, J. Song, Adaptive cross approximation algorithm for accelerating bem in eddy current nondestructive evaluation, Journal of Nondestructive Evaluation 37 (4) (2018) 68. doi: 10.1007/s10921-018-0521-1

    2018 doi

  31. [39]

    Y . Bao, Z. Liu, J. R. Bowler, J. Song, Multilevel adaptive cross approximation for e fficient modeling of 3D arbitrary shaped eddy current NDE problems, NDT & E International 104 (2019) 1–9. doi:https://doi.org/10.1016/j.ndteint.2019. 03.005

    2019 doi

  32. [40]

    F. Cau, A. G. Chiariello, G. Rubinacci, V . Scalera, A. Tambur- rino, S. Ventre, F. Villone, A fast matrix compression method for large scale numerical modelling of rotationally symmetric 3D passive structures in fusion devices, Energies 15 (9) (2022). doi:10.3390/en15093214

    2022 doi

  33. [41]

    Rokhlin, Rapid solution of integral equations of clas- sical potential theory, Journal of Computational Physics 60 (2) (1985) 187–207

    V . Rokhlin, Rapid solution of integral equations of clas- sical potential theory, Journal of Computational Physics 60 (2) (1985) 187–207. doi:https://doi.org/10.1016/ 0021-9991(85)90002-6

    1985

  34. [42]

    Bebendorf, S

    M. Bebendorf, S. Rjasanow, Adaptive low-rank approximation of collocation matrices, Computing 70 (1) (2003) 1–24. doi: 10.1007/s00607-002-1469-6

    2003 doi

  35. [43]

    Alman, V

    J. Alman, V . V . Williams, A Refined Laser Method and Faster Matrix Multiplication, TheoretiCS V olume 3 (Sep. 2024).doi: 10.46298/theoretics.24.21

    2024 doi

  36. [44]

    A. J. Creely, M. J. Greenwald, S. B. Ballinger, D. Brunner, J. Canik, J. Doody, T. F ¨ul¨op, D. T. Garnier, R. Granetz, T. K. Gray, et al., Overview of the SPARC tokamak, Journal of Plasma Physics 86 (5) (2020) 865860502. doi:10.1017/ S0022377820001257

    2020

  37. [45]

    A. C. Hindmarsh, Lsode and lsodi, two new initial value or- dinary di fferential equation solvers, SIGNUM Newsl. 15 (4) (1980) 10–11. doi:10.1145/1218052.1218054. URL https://doi.org/10.1145/1218052.1218054

    1980

  38. [46]

    The Numerical Algorithms Group Ltd, Oxford, UK, NAG Library Manual, Mark 26, available at https: //support.nag.com/numeric/nl/nagdoc_26/nagdoc_ fl26/html/frontmatter/manconts.html (2017)

    2017

  39. [47]

    Sweeney, A

    R. Sweeney, A. J. Creely, J. Doody, T. F ¨ul¨op, D. T. Garnier, R. Granetz, M. Greenwald, L. Hesslow, J. Irby, V . A. Izzo, et al., MHD stability and disruptions in the SPARC tokamak, Journal of Plasma Physics 86 (5) (2020) 865860507. doi:10.1017/ S0022377820001129

    2020

  40. [48]

    Riccardo, the SPARC Team, Disruptions loads in SPARC, Talk presented at the Second Technical Meeting on Plasma Disruptions and their Mitigation (2022)

    V . Riccardo, the SPARC Team, Disruptions loads in SPARC, Talk presented at the Second Technical Meeting on Plasma Disruptions and their Mitigation (2022). URL https://conferences.iaea.org/event/281/ contributions/24400/attachments/13000/20067/ Riccardo_SPARC_220719.pdf

    2022

  41. [49]

    Geuzaine, J.-F

    C. Geuzaine, J.-F. Remacle, Gmsh: a three-dimensional finite element mesh generator with built-in pre- and post-processing facilities, in: Proceedings of the fourth international conference on advanced computational methods in engineering, ACOMEN 2008, 2008

    2008

  42. [50]

    D. A. Humphreys, A. G. Kellman, Analytic modeling of ax- isymmetric disruption halo currents, Physics of Plasmas 6 (7) (1999) 2742–2756. doi:10.1063/1.873231

    1999 doi

  43. [51]

    A. R. Saperstein, R. A. Tinguely, R. S. Granetz, J. P. Levesque, M. E. Mauel, G. A. Navratil, Disruption halo current rotation scaling on Alcator C-Mod and HBT-EP, Physics of Plasmas 30 (4) (2023) 042506. doi:10.1063/5.0140867

    2023 doi

  44. [52]

    Landreman, B

    M. Landreman, B. Medasani, F. Wechsung, A. Giuliani, R. Jorge, C. Zhu, SIMSOPT: A flexible framework for stellara- tor optimization, Journal of Open Source Software 6 (65) (2021)

    2021

  45. [3525]

    URL https://doi.org/10.21105/joss.03525 14

    doi:10.21105/joss.03525. URL https://doi.org/10.21105/joss.03525 14

    doi

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.