Pith. sign in

REVIEW 6 minor 39 references

A Cloud-Accessible Open-Source Framework for the Electromagnetic Modelling of Applied Superconductors

T0 review · 0 major / 6 minor · reviewed 2026-07-13 · grok-4.5

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

Pith's one-line read An open, cloud-run finite-element residual for type-II superconductors matches commercial magnetisation loops to about 1%.

desk verdict Solid open-source H-formulation with real COMSOL agreement and Colab portability; methods paper, not a physics breakthrough. read the letter →

arxiv 2607.09572 v1 pith:YKCODFPU submitted 2026-07-10 cond-mat.supr-con math-phmath.MP

Yusen Guo , Alberto Paganini , Harold S. Ruiz This is my paper · ORCID
classification cond-mat.supr-conmath-phmath.MP
keywords Type-IISuperconductorsElectromagneticModellingH-formulationFiniteElementMethodCloudComputingNédélecElementsAdjointOptimization
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 introduces H-cloud: a transparent finite-element statement of the magneto-quasi-steady H-formulation for type-II superconductors. The magnetic field lives in a curl-conforming Nédélec space, the applied field is imposed as a tangential trace, the nonlinear E–J power law sits inside a fully implicit residual, and the same residual drives both forward solves and adjoint-based optimisation. The identical Python script runs on a laptop or in a free browser notebook. On the classical Bean cylinder it recovers the expected full-penetration scale; on a Bi2212-wire benchmark it reproduces independently built commercial magnetisation loops to roughly 1% loop error and under 1.5% peak error. The claim is that once the physics is written this way, superconducting simulation becomes portable, shareable, and ready for inverse design without black-box software.

What carries the argument

The fully discrete nonlinear residual F(H^{k+1}; ψ) formed by backward-Euler time stepping of the H(curl) weak form of Faraday’s law closed by the power-law E–J relation; this single object is the forward solver, the equality constraint for PDE-constrained optimisation, and the object differentiated by automatic adjoint tools.

What would settle it

Re-run the matched Bi2212 magnetisation benchmark with n→∞ (or a true critical-state formulation) and successively larger outer domains; if loop or peak errors against the commercial reference rise well above 1–1.5%, or the recovered full-penetration field drifts from the Bean value, the claim of rigorous validation collapses.

Watch

Extended reading notes

Core claim

A curl-conforming weak residual of the H-formulation, with explicit tangential applied-field boundary condition and power-law resistivity, can be coded once in a scripted finite-element language and executed identically on local machines and free cloud notebooks, matching commercial magnetisation results for a practical Bi2212 wire to about 1% while remaining open to adjoint optimisation.

Load-bearing premise

A finite-n power-law resistivity plus a finite outer air domain is close enough to the ideal critical-state / infinite-domain problem that recovering the analytical full-penetration field and 1% commercial agreement truly validates the formulation itself.

Editorial extensions

If this is right

  • Any researcher can share an exact superconducting magnetisation calculation as a single notebook that runs in a free browser without a commercial licence.
  • The same residual can be handed to automatic adjoint tools, turning forward H-formulation models into gradient-based inverse design of cloaks, cables, or flux-control devices.
  • Benchmark meshes, power-law parameters, and solver tolerances become visible and modifiable objects rather than GUI settings, enabling community-standard verification suites.
  • Modest free-cloud hardware already supports practical wire-scale runs; larger geometries can move to paid multi-core or GPU backends without reformulating the physics.

Reading between the lines

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

  • If the residual-as-constraint pattern generalises cleanly, the same notebook workflow could become a teaching and inter-lab reference for competing formulations (T–A, A–V, J-based) on identical meshes.
  • The 1% commercial match on a simple cylinder does not yet guarantee that adjoint-driven material designs will remain manufacturable; regularisation choices for the control variable will likely dominate practical inverse results.
  • Memory footprints reported for the fine mesh (≈0.5 GB) suggest that three-dimensional tape stacks or multi-turn coils will hit free-cloud CPU limits long before memory, so community adoption may hinge on open multi-core or GPU PETSc paths.
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

0 major / 6 minor

Summary. The manuscript introduces the H-cloud formalism: a curl-conforming H-formulation of magneto-quasi-steady type-II superconductivity, discretised with Nédélec elements and written as an explicit fully implicit residual (Eq. 13) in Firedrake/UFL/PETSc. The same residual is presented as the basis for forward simulation and for future adjoint/PDE-constrained optimisation. Verification is performed on the Bean cylinder magnetisation benchmark (n=96, slow AC drive), recovering the analytic full-penetration scale ba=8 (Table 1, Fig. 1–2). The method is then compared with an independently built COMSOL model of a Bi2212-inspired wire on three matched meshes; loop errors eM are ~0.8–1.4% and peak errors epeak <1.5% (Table 2), with the identical script producing the same solution in Google Colab at higher but hardware-consistent runtime. Code and data are deposited (Zenodo DOI).

Significance. If the reported agreement and portability hold, the paper supplies a transparent, licence-free, cloud-executable reference implementation of the standard H-formulation that commercial multiphysics packages do not expose at the variational level. Strengths that should be credited explicitly are: (i) the residual is written in full at the discrete level (Eq. 13) rather than hidden behind GUI templates; (ii) quantitative external validation against both Bean’s analytic penetration field (Table 1) and an independent COMSOL discretisation on matched meshes (Table 2); (iii) domain-truncation sensitivity is shown rather than assumed (Fig. 2); and (iv) the identical script is demonstrated to run in Colab with deposited code, which is rare and useful for benchmark sharing and teaching in applied superconductivity. The adjoint/optimisation pathway is only sketched (via prior cloak work [2]) and is not a delivered result of this manuscript, but the forward-solver contribution is self-contained and of clear practical value.

minor comments (6)
  1. Abstract and §3.2 state “approximately 1%” loop agreement; Table 2 shows eM rising from 7.75e-3 (coarse) to ~1.43e-2 (fine). A one-sentence clarification that the ~1% figure is an order-of-magnitude summary across the hierarchy (or quoting the fine-mesh value) would avoid any impression of overstatement.
  2. §3.2 notes that COMSOL uses adaptive time-stepping while H-cloud uses fixed Δt=1e-4 s, with COMSOL outputs sampled onto the same grid. A brief remark on whether residual or magnetisation differences are sensitive to this sampling choice would strengthen the comparison.
  3. Eq. (21) writes M = (l/2) ∫ r imes J dΩ; for the 2-D per-unit-length cylinder the factor l and the precise 2-D reduction of the moment integral should be stated explicitly so that Mp = (2/3π)Jc R^3 is recovered without ambiguity.
  4. Fig. 1 caption and the definition of ČBR = (μ0/4π)Jc R are clear, but a short note that Bp = 8 ČBR recovers the classical Bean cylinder result would help readers who work in SI units only.
  5. The adjoint section (§2.2, Eqs. 14–20) is motivational and correctly deferred; a single sentence in the conclusion reminding the reader that no inverse-design result is claimed in this paper would prevent over-reading of the optimisation language in the abstract.
  6. Minor typography: “F ramework” (section 2 heading), “disappearssatisfying” (§3.1), and inconsistent spacing around some equation references should be cleaned in production.

Circularity Check

1 steps flagged · score 1.0 of 10

No significant circularity: central claims validated against Bean analytics and independent COMSOL, with only non-load-bearing self-citation to deferred adjoint work.

  1. self citation load bearing [§2.2 (after Eq. 16) and §4 Conclusion]
    "take as reference our recently proposed method for designing functional magnetic cloaks at [2] ... Extensions to multi-objective inverse-design functionals, of the type explored in [2], are therefore deferred to future studies."

    Citation [2] is by the same author team and is invoked to motivate the adjoint capability of the residual. However the optimisation path is not executed or required for any numerical result in the present paper; the load-bearing claims rest solely on the Bean analytic scale and the independent COMSOL comparison. The self-citation is therefore present but non-load-bearing.

full rationale

The paper's derivation chain is the standard magneto-quasi-steady H-formulation (Faraday + Ampere + power-law E-J) written in weak form on H(curl) with Nédélec elements, fully-implicit residual (13), and tangential-trace BC (11). Forward verification uses the classical Bean cylinder full-penetration scale Bp = 8 (µ0/4π)Jc R and Mp (Table 1, Fig. 1-2) plus an independently constructed COMSOL model of a Bi2212-inspired wire on matched meshes (Table 2, eM ≈ 1 %, epeak < 1.5 %). Neither reference is defined by the authors' residual or fitted to their outputs; material parameters are taken from external literature. The sole self-citation [2] (authors' prior cloak-optimisation paper) appears only as motivation for the adjoint/PDE-constrained extension, which is explicitly deferred and not used to support the forward accuracy or portability claims. No self-definitional identities, fitted-inputs-as-predictions, uniqueness theorems, or ansatz smuggling occur. The framework is therefore self-contained against external benchmarks; residual circularity is limited to ordinary non-load-bearing self-citation of prior related work.

Assumptions & free parameters 4 free parameters · 5 assumptions · 1 invented entities

The central numerical claim rests on standard magneto-quasi-static Maxwell equations, the conventional power-law E–J constitutive model, Nédélec conformity, and a set of concrete numerical choices (n, domain size, time step, mesh hierarchy) that are stated rather than fitted to the COMSOL target. No new physical entities are postulated; free parameters are the usual material and discretisation knobs of the community.

free parameters (4)
  • power-law exponent n = 96 (Bean), 25 (Bi2212)
    Chosen by hand: n=96 to approach Bean critical-state limit; n=25 for the Bi2212 wire. Controls how closely the model approximates the pure critical-state inequality.
  • outer-domain truncation radius = 10R (wire); 1.25R–2.5R (sensitivity)
    Finite computational domain replaces the analytic condition at infinity; set to 10R for the wire benchmark and varied (1.25R–2.5R) in the sensitivity study.
  • time-step size Δt = 1e-4 s
    Fixed Δt=10^{-4}s (250 steps) for the Bi2212 comparison; must be small enough to preserve MQS and Newton convergence.
  • critical current density Jc and Ec criterion = Jc=6.64e9 A/m², Ec=1e-4 V/m
    Material parameters taken from literature for Bi2212; not fitted to the COMSOL magnetisation loop.
assumptions (5)
  • domain assumption Magneto-quasi-steady approximation: displacement current neglected, Faraday + Ampère close the system.
    Stated in §2.1; standard for applied-superconductor modelling at the frequencies considered.
  • domain assumption Power-law E–J constitutive relation ρ(J) = (Ec/Jc)(|J|/Jc)^{n-1} inside the superconductor.
    Eq. (5); community-standard regularisation of the critical-state model.
  • domain assumption Linear isotropic magnetic response μr=1 throughout the domain.
    §2.1; allows direct comparison with non-magnetic Bean analytics.
  • standard math Nédélec (edge) elements of the first kind are H(curl)-conforming and prevent spurious modes.
    §2.2; classical finite-element exterior-calculus result.
  • domain assumption Backward-Euler fully implicit time discretisation of the residual is stable for the stiff power-law problem.
    Eq. (13); standard choice, not proved here but widely used.
invented entities (1)
  • H-cloud formalism
    purpose: Name given to the combination of the explicit H(curl) residual, Firedrake/UFL implementation, and cloud-executable notebook workflow.
    Branding of an implementation stack rather than a new physical object; no independent ontological claim.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Cloud-Accessible Open-Source Framework for the Electromagnetic Modelling of Applied Superconductors." pith.science (2026). https://pith.science/paper/YKCODFPU

@misc{pith2026260709572,
  author       = {Pith},
  title        = {Pith review of: A Cloud-Accessible Open-Source Framework for the Electromagnetic Modelling of Applied Superconductors},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YKCODFPU}},
  note         = {Machine review of arXiv:2607.09572}
}
abstract

We present the H-cloud formalism, a cloud-accessible and open-source finite-element framework for electromagnetic modelling of applied superconductors. The proposed method expresses the nonlinear electromagnetic response of type-II superconductors in a curl-conforming discretisation based on N\'ed\'elec finite elements, where the tangential applied-field boundary condition, nonlinear E-J power law, and fully implicit time-discrete residual are stated explicitly at the variational level, all within a scripted Python finite-element workflow. The weak form is used as the basis for forward simulation and for extension to adjoint differentiation and PDE-constrained optimisation, while keeping the governing assumptions, boundary conditions, and solver structure fully visible to the user. The implementation is realised in Firedrake with UFL and PETSc-backed nonlinear solvers, allowing the identical script to run on local machines and in browser-accessible environments such as Google Colab without reformulating the problem. The method is verified on the canonical magnetisation benchmark of a cylindrical superconductor under Bean-like penetration conditions and then benchmarked against an independently constructed COMSOL model for a practical high temperature superconducting Bi2212 wire. Across matched mesh studies, the open-source workflow reproduces the commercial-reference magnetisation loops to within approximately \(1\%\) , with relative peak errors below 1.5%, while cloud execution preserves the same numerical solution at rather modest additional runtime considering the use of (freely available) reduced hardware resources. The proposed framework provides a rigorous, reproducible, and portable route for superconducting simulation, benchmarking, and future optimisation-led modelling of applied and functional superconductors, shareable and executable into open cloud environments.

Figures

Figures reproduced from arXiv: 2607.09572 by the authors.

Figure 1
Figure 1. Calculated magnetization loops within the H-cloud formalism for the cylindrical superconducting benchmark under AC transverse magnetic field of amplitude ba = 1, 2, . . . , 10 in units of BˇR = (µ0/4π)JcR. The magnetic moment is normalized by the analytical full-penetration value Mp. 3.1 Magnetisation benchmark For the cylindrical benchmark, when the AC magnetic field is applied along a single direction, say Hy, the… view at source ↗
Figure 2
Figure 2. Current density profiles at Ba = Bp = 8BˇR with the SC wire centred at (0, 0), for three surrounding boundary-domain sizes delimited by max(|x|, |y|) relative to (a) 1.25R, (b) 1.65R, and (c) 2.5R. The comparison shows that a close outer boundary artificially suppresses flux penetration, whereas a sufficiently large air domain recovers the expected fully penetrated state at the expected analytical solution. metallic… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 27 canonical work pages

  1. [2]

    Guo Y, Paganini A and Ruiz H S 2025Science Advances11eaea2468 URL https://doi.org/10.1126/sciadv.aea2468

    work page doi:10.1126/sciadv.aea2468

  2. [1]

    G¨ om¨ ory F, Solovyov M,ˇSouc J, Navau C, Prat-Camps J and Sanchez A 2012Science335 1466–1468 URLhttps://doi.org/10.1126/science.1218316

    work page doi:10.1126/science.1218316

  3. [3]

    MacManus-Driscoll J L and Wimbush S C 2021Nature Reviews Materials6587–604 URL https://doi.org/10.1038/s41578-021-00290-3

    doi:10.1038/s41578-021-00290-3

  4. [4]

    Coombs T A, Wang Q, Shah A, Hu J, Hao L, Patel I, Wei H, Wu Y, Coombs T and Wang W 2024Nature Reviews Electrical Engineering1788–801 ISSN 2948-1201 URL https://doi.org/10.1038/s44287-024-00112-y

    doi:10.1038/s44287-024-00112-y

  5. [5]

    Ruiz H S, Zhang X and Coombs T A 2015IEEE Transactions on Applied Superconductivity 251–5 ISSN 1051-8223 URLhttps://doi.org/10.1109/TASC.2014.2387115

    work page doi:10.1109/tasc.2014.2387115 2014

  6. [6]

    Brooks J M, Ainslie M D, Jiang Z, Pantoja A E, Badcock R A and Bumby C W 2020 Superconductor Science and Technology33035007 URL https://dx.doi.org/10.1088/1361-6668/ab6bfe

    work page doi:10.1088/1361-6668/ab6bfe 2020

  7. [7]

    Mataira R C, Ainslie M D, Badcock R A and Bumby C W 2019Applied Physics Letters114 162601 URLhttps://doi.org/10.1063/1.5085226

    work page doi:10.1063/1.5085226

  8. [8]

    Robert B C and Ruiz H S 2018Superconductor Science and Technology3135006 ISSN 0953-2048 URLhttps://iopscience.iop.org/article/10.1088/1361-6668/aaa823

    work page doi:10.1088/1361-6668/aaa823

Show all 39 references
  1. [9]

    Bad´ ıa-Maj´ os A, L´ opez C and Ruiz H S 2009Physical Review B80(14) 144509 URL https://doi.org/10.1103/PhysRevB.80.144509

    doi

  2. [10]

    Baghdadi M, Ruiz H S and Coombs T A 2014Applied Physics Letters104232602 ISSN 0003-6951 URLhttps://doi.org/10.1063/1.4879263

    doi

  3. [11]

    Ruiz H S, Bad´ ıa-Maj´ os A, Genenko Y A, Rauh H and Yampolskii S V 2012Applied Physics Letters100112602 URLhttps://doi.org/10.1063/1.3693614

    doi

  4. [12]

    Ruiz H S, H¨ anisch J, Polichetti M, Galluzzic A, Gozzelino L, Torsello D, Miloˇ sevi´ c-Govedarovi´ c S, Grbovi´ c-Novakovi´ c J, Dobrovolskiy O, Lang W, Grimaldi G, Crisan A, Badica P, Ionescu A, Cayado P, Willa R, Barbiellini B, Eley S and Bad´ ıa–Maj´ os A 2025Progress in ...

    2025 doi

  5. [13]

    Grilli F, Pardo E, Morandi A, G¨ om¨ ory F, Solovyov M, Zerme˜ no V M R, Brambilla R, Benkel T and Riva N 2021IEEE Transactions on Applied Superconductivity311–9 URL http://dx.doi.org/10.1109/TASC.2020.3013028 10 XXXXXXXXvv(2026) XXXXX Yusen Guoet al

    2020 doi

  6. [14]

    Wimbush S C and Strickland N M 2022Superconductor Science and Technology35024004 URLhttps://doi.org/10.1088/1361-6668/ac4172

    doi

  7. [15]

    Bean C P 1962Physical Review Letters8(6) 250–253 URL https://doi.org/10.1103/PhysRevLett.8.250

    doi

  8. [16]

    Shen B, Grilli F and Coombs T 2020IEEE Access8100403–100414 ISSN 2169-3536 URL https://doi.org/10.1109/ACCESS.2020.2996177

    2020 doi

  9. [17]

    Benkel T, Lao M, Liu Y, Pardo E, Wolfstadter S, Reis T and Grilli F 2020IEEE Transactions on Applied Superconductivity30ISSN 15582515 URL https://doi.org/10.1109/TASC.2020.2968950

    2020 doi

  10. [18]

    Zhang H, Zhang M and Yuan W 2016Superconductor Science and Technology30024005 URL https://doi.org/10.1088/1361-6668/30/2/024005

    doi

  11. [19]

    Ruiz-Alonso D, Coombs T and Campbell A M 2004Superconductor Science and Technology 17S305 URLhttps://doi.org/10.1088/0953-2048/17/5/042

    doi

  12. [20]

    Lahtinen V, Lyly M, Stenvall A and Tarhasaari T 2012Superconductor Science and Technology25115001 URLhttps://doi.org/10.1088/0953-2048/25/11/115001

    doi

  13. [21]

    Dular J, Harutyunyan M, Bortot L, Sch¨ ops S, Vanderheyden B and Geuzaine C 2021IEEE Transactions on Applied Superconductivity311–12 URL https://doi.org/10.1109/TASC.2021.3098724

    2021 doi

  14. [22]

    Arsenault A, Sirois F and Grilli F 2021IEEE Transactions on Applied Superconductivity31 1–9 URLhttps://doi.org/10.1109/TASC.2021.3073274

    2021 doi

  15. [23]

    Brambilla R, Grilli F, Nguyen D N, Martini L and Sirois F 2009Superconductor Science and Technology22075018 URLhttps://doi.org/10.1088/0953-2048/22/7/075018

    doi

  16. [24]

    Pardo E and Kapolka M 2017Journal of Computational Physics344339–363 URL https://doi.org/10.1016/j.jcp.2017.05.001

    2017 doi

  17. [25]

    Bad´ ıa-Maj´ os A and L´ opez C 2012Superconductor Science and Technology25104004 URL https://doi.org/10.1088/0953-2048/25/10/104004

    doi

  18. [26]

    Mitusch S K, Funke S W and Dokken J S 2019Journal of Open Source Software41292 URL https://doi.org/10.21105/joss.01292

    doi

  19. [27]

    Bean C P 1964Reviews of Modern Physics36(1) 31–39 URL https://doi.org/10.1103/RevModPhys.36.31

    doi

  20. [28]

    Karmakar D and Bhagwat K 2003Physica C: Superconductivity39820 – 30 ISSN 0921-4534 URLhttps://doi.org/10.1016/S0921-4534(03)01202-4

    doi

  21. [29]

    Ruiz Rondan H 2012Material Laws and Numerical Methods in Applied Superconductivity (Zaragoza, Spain: University of Zaragoza Press) ISBN 978-84-15770-29-9 URL https://doi.org/10.48550/arXiv.1203.2159

    doi arXiv

  22. [30]

    Ham D A, Kelly P H J, Mitchell L, Cotter C J, Kirby R C, Sagiyama K, Bouziani N, Vorderwuelbecke S, Gregory T J, Betteridge J, Shapero D R, Nixon-Hill R W, Ward C J, Farrell P E, Brubeck P D, Marsden I, Gibson T H, Homolya M, Sun T, McRae A T T, Luporini F, Gregory A, Lange M,...

    doi

  23. [31]

    Rathgeber F, Ham D A, Mitchell L, Lange M, Luporini F, Mcrae A T T, Bercea G T, Markall G R and Kelly P H J 2016ACM Trans. Math. Softw.43ISSN 0098-3500 URL https://doi.org/10.1145/2998441

    doi

  24. [32]

    Balay S, Abhyankar S, Adams M F, Benson S, Brown J, Brune P, Buschelman K, Constantinescu E M, Dalcin L, Dener A, Eijkhout V, Faibussowitsch J, Gropp W D, Hapla V, Isaac T, Jolivet P, Karpeev D, Kaushik D, Knepley M G, Kong F, Kruger S, May D A, McInnes L C, Mills R T, Mitchel...

    2025

  25. [33]

    Google LLC Google Colaboratory (Colab)https://colab.research.google.com/accessed: 2026-06-08

    2026

  26. [34]

    Arnold D N 2018Finite Element Exterior Calculus(Philadelphia, PA: Society for Industrial and Applied Mathematics)

  27. [35]

    Virtanen P, Gommers R, Oliphant T E, Haberland M, Reddy T, Cournapeau D, Burovski E, Peterson P, Weckesser W, Bright J, van der Walt S J, Brett M, Wilson J, Millman K J, Mayorov N, Nelson A R J, Jones E, Kern R, Larson E, Carey C J, Polat ˙I, Feng Y, Moore E W, VanderPlas J, L...

    doi

  28. [36]

    Ruiz H S and Bad´ ıa-Maj´ os A 2010Superconductor Science and Technology23105007–105007 ISSN 0953-2048 URL https://iopscience.iop.org/article/10.1088/0953-2048/23/10/105007

    doi

  29. [37]

    Barua S, Davis D S, Oz Y, Jiang J, Hellstrom E E, Trociewitz U P and Larbalestier D C 2021 IEEE Transactions on Applied Superconductivity311–6 URL https://doi.org/10.1109/TASC.2021.3055479

    2021 doi

  30. [38]

    Oloye T A, Matras M, Jiang J, Hossain S I, Su Y, Trociewitz U P, Hellstrom E E, Larbalestier D C and Kametani F 2021Superconductor Science and Technology34035018 URL https://doi.org/10.1088/1361-6668/abd575

    doi

  31. [39]

    Jiang J, Bradford G, Hossain S I, Brown M D, Cooper J, Miller E, Huang Y, Miao H, Parrell J A, White M, Hunt A, Sengupta S, Revur R, Shen T, Kametani F, Trociewitz U P, Hellstrom E E and Larbalestier D C 2019IEEE Transactions on Applied Superconductivity291–5 12

Pith tools

Reviewed July 13, 2026 · model on record in the stance chip above.