Pith. sign in

REVIEW 3 major objections 4 minor 19 references

Simulation of Rutherford Cable AC Loss and Magnetization with the Coupled Axial and Transverse Currents Method

T0 review · 3 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read A two-dimensional finite-element model coupled to circuit equations reproduces Rutherford cable coupling loss to within 5 percent at a fraction of the 3D cost.

desk verdict A solid single-geometry verification of the CATI extension to Rutherford cables, with a real speedup; needs a parameter sweep and a correction to the network-model check before publication. read the letter →

arxiv 2411.13347 v2 pith:YUA4OTMU submitted 2024-11-20 physics.acc-ph

Julien Dular , Fredrik Magnus , Erik Schnaubelt , Arjan Verweij , Mariusz Wozniak This is my paper · ORCID
classification physics.acc-ph
keywords RutherfordcableAClossinterstrandcouplingcurrentsCATImethodreducedordermodelfiniteelementmagnetizationsuperconductingacceleratormagnets
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

Rutherford cables are periodic arrays of transposed superconducting strands whose interstrand coupling currents dominate AC loss, but full three-dimensional finite-element simulations of them are too slow for routine design studies. This paper adapts the coupled axial and transverse currents (CATI) method to such cables: a 2D cross-section model coupled to circuit equations that encode the cable's periodicity. It claims that the 2D model reproduces the coupling currents and loss per cycle of a reference 3D simulation to within 5 percent at the loss peak and at the same characteristic frequency, while using roughly 17 times fewer degrees of freedom and running about 80 times faster. The point is to make accurate AC-loss and magnetization predictions for Rutherford cables cheap enough for accelerator-magnet design and protection studies.

What carries the argument

The CATI method: a 2D finite-element model of the cable cross-section that solves for axial strand currents using an $h$-$\phi$ formulation (a magnetic-field-based variational form), coupled through an electrical circuit to lumped contact resistances $R_c$ and $R_a$ for crossing and adjacent strand contacts. The periodicity length $\ell = p/N_s$ is the coupling step: after one length $\ell$, each strand has moved to its neighbor's position, so the circuit equations close the axial currents with the transverse currents flowing through the contact resistances. This decomposition splits the 3D problem into a cheap 2D magnetodynamic problem plus a small circuit, which is what carries the computational saving.

What would settle it

Measure the loss peak on a real Rutherford cable segment with independently measured contact resistances; if the peak frequency deviates from the predicted 3.5 Hz (i.e., time constant 45–47 ms) or the peak loss differs from the CATI prediction by more than 5 percent, the central claim is falsified.

Watch

Extended reading notes

Core claim

The paper shows that the interstrand coupling currents in a Rutherford cable can be represented by lumped resistors between adjacent and crossing strands, coupled to a 2D axial-current finite-element model through circuit equations that impose periodicity over one strand pitch, $\ell = p/N_s$. This circuit-finite-element coupling reproduces the current-density distribution, strand currents, and coupling loss of a full 3D $h$-$\phi$ finite-element model for frequencies up to about 1 kHz, with a loss peak at $f_c \approx 3.5$ Hz whose height differs from the 3D result by less than 5 percent. The associated time constant, $\tau_c = 45$ ms, also matches the network-model estimate of 47 ms. Above 1 kHz, skin and side-edge effects appear in the 3D model that the CATI method, which assumes purely axial strand currents and fields constant along the cable, does not capture.

Load-bearing premise

The cable's transposition is fully captured by the circuit coupling over one periodicity length, while the two-dimensional axial model treats strand currents as purely axial and fields as constant along the cable, ignoring the strand tilt angle; if longitudinal field variation or tilt matters at the frequencies of interest, the claimed accuracy breaks down.

Editorial extensions

If this is right

  • Below roughly 1 kHz, CATI reproduces crossing and adjacent coupling currents and their losses to within a few percent of a 3D simulation, making fast frequency sweeps of cable loss practical.
  • The loss peak frequency and time constant emerge from the method without tuning and match the network-model estimate of $\tau_c \approx 45$–47 ms, so the method captures the actual coupling-current dynamics.
  • Replacing a 1.1-million-degree-of-freedom, several-minutes-per-frequency 3D simulation with a 64,000-degree-of-freedom, roughly 3-second-per-frequency 2D simulation enables parameter scans over contact resistances and cable geometries.
  • Because the linear model reproduces the 3D current density with and without transport current, it can be used to study screening currents and field distortions in the coupling-dominated regime.
  • The authors identify the next step as coupling this cable model to nonlinear strand models that include filament-level dynamics, which would cover all loss contributions down to the filament level.

Reading between the lines

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

  • If the CATI cable model is coupled to a strand-level hysteresis and interfilament-coupling model, the combined tool could predict full AC loss and magnetization curves for real accelerator magnets without resolving the full 3D cable, speeding up cryogenic-load and quench-protection studies.
  • Because the method only assumes periodicity, the same circuit-finite-element coupling could apply to other periodic transposed conductors, such as twisted stacks or cable-in-conduit conductors, not just Rutherford cables.
  • The known breakdown above roughly 1 kHz, where the skin depth (4 µm at 10 kHz) becomes comparable to the strand size, suggests a natural extension: add a boundary-layer or tilt-angle correction to CATI to push its valid frequency range upward, or use CATI only in the coupling-dominated regime and switch to 3D for high-frequency eddy-current effects.
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 / 4 minor

Summary. The manuscript extends the coupled axial and transverse currents (CATI) reduced-order method from twisted composite strands to Rutherford cables. The cable is represented by a two-dimensional h-phi finite-element model of the cross-section for axial strand currents, coupled through circuit equations to lumped resistance elements for crossing and adjacent interstrand contacts, with transposition represented by a discrete permutation of strand ports after one periodicity length ℓ = p/Ns. The method is verified in the linear frequency domain against a three-dimensional h-phi finite-element model of one period of the same cable using the same contact-resistance values. For a 26-strand cable with pitch 0.1 m, the authors report matching current-density maps, coupling-current patterns, and loss-per-cycle curves; the maximum coupling loss differs by less than 5%, the peak occurs at fc ≈ 3.5 Hz in both models, and the computational cost is reduced from 1.1 million degrees of freedom and 4 minutes per frequency to 64 thousand degrees of freedom and about 3 seconds per frequency.

Significance. If the result holds, this is a practically useful reduction: interstrand coupling loss in Rutherford cables can be computed with a two-dimensional field model at a small fraction of the cost of a three-dimensional model, while retaining screening effects that network or continuum models omit. The paper's strengths are the direct side-by-side comparison with a full 3D finite-element reference, the frequency-resolved loss and current comparisons, the absence of any fitting of the loss curves, and the open-source implementation in Gmsh, GetDP, and FiQuS. The main limitations are that the reference is the authors' own 3D model built with the same contact resistances, and the verification covers only one geometry. I do not regard the model-to-model comparison as circular, because the 3D model includes strand tilt and continuous transposition that CATI deliberately approximates and because no loss quantity is tuned to force agreement; nevertheless, the independent confirmation via Eq. (4) is weakened by an arithmetic inconsistency noted below.

major comments (3)
  1. [Eq. (4), Table I] Inserting the printed values C = 1.65e-8 Ω s m^-1, p = 0.1 m, Ns = 26, and Rc = 20 μΩ into τc = C p (Ns^2 − Ns)/Rc gives 53.6 ms, not the printed 47 ms. The observed peak at fc ≈ 3.5 Hz corresponds to τc ≈ 45 ms, so the statement that the result 'also matches calculations from network models' is not supported by the numbers as written. Please correct either the constant, the parameter values, or the formula, and recompute the associated fc.
  2. [Section III-B, Figs. 4-5] The verification is performed with a single geometric and electrical configuration (Ns = 26, p = 0.1 m, Table I). The two core simplifications of CATI—fields assumed constant over one period ℓ and transposition represented only by an endpoint permutation—are not stress-tested by varying p, Ns, or the Ra/Rc ratio. Consequently, the conclusion that the method 'allows for detailed and fast analyses of arbitrary cable geometries' is broader than the evidence. Please add at least one additional configuration or explicitly restrict the claim to the verified parameter range.
  3. [Section III-B, mesh sizes] No mesh-convergence study is reported for either the two-dimensional CATI model or the three-dimensional reference. The statement that the 3D mesh uses elements of similar size to the 2D mesh is not sufficient to establish that the <5% loss agreement and the matching fc are converged values. A convergence check of the maximum coupling loss, the peak frequency, and the current patterns with respect to mesh refinement would substantiate the accuracy claim.
minor comments (4)
  1. [Eq. (4)] The quantity C is used in Eq. (4) but is never defined; please state its physical meaning and provenance even if it is only a fit constant from Ref. [8].
  2. [Section III-A] The surface areas Sc and Sa are only indicated in Fig. 2(c); a sentence giving their definitions or values would make the 3D contact-resistance assignment easier to reproduce.
  3. [Section II-A] The sentence 'The magnetic field is defined as a transverse vector field' should clarify that this transverse-field assumption and the neglect of strand tilt are modeling approximations, not exact properties of the cable geometry.
  4. [Fig. 5] The high-frequency curves in Fig. 5 extend to 10 kHz, but the text states that agreement is expected only for f ≲ 1 kHz; a vertical guideline or a note in the caption would help readers avoid overinterpreting the disagreement at the highest frequencies.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the CATI-to-3D comparison is a genuine numerical verification, not a reduction to inputs.

full rationale

The central claim is a cost-reduction verification: the CATI reduced-order model is compared against an independently solved 3D h-phi FE model on the same geometry and material parameters. The agreement in loss curves, current distributions, and peak frequency is not guaranteed by construction; the 3D model retains full three-dimensional fields and contact-surface resistivities, while CATI imposes axial currents and one-periodicity constant-field assumptions. No quantity used in the comparison is fitted to make CATI match the 3D result, and the peak-frequency comparison with the network-model formula in Eq. (4) uses a parameter-free expression from an external, though same-author, thesis, which does not include the target result among its assumptions. The paper itself flags the known breakdown above about 1 kHz due to skin and edge effects absent from CATI; that limitation undercuts accuracy claims but is not circularity. The only concern of note is a numerical inconsistency in Eq. (4), where inserting Table I values gives about 53.6 ms rather than 47 ms, but that is a correctness issue, not a circular-derivation issue. Overall, the derivation chain is self-contained as a numerical verification study.

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

The central verification rests on the linear ohmic model, the periodic geometry assumption, and uniform contact resistances; all are disclosed by the authors.

free parameters (4)
  • Crossing contact resistance Rc = 20 µΩ
    Hand-chosen input from Table I, used in both CATI and 3D models; not fitted to the comparison.
  • Adjacent contact resistance Ra = 10 µΩ
    Hand-chosen input from Table I; same value used in both models.
  • Strand resistivity ρs = 0.001 nΩ m
    Chosen to model strands as near-ideal ohmic conductors in the linear verification.
  • Coating matrix resistivity ρm = 100 nΩ m
    Chosen high enough to make the numerical coating domain negligible.
assumptions (5)
  • standard math Maxwell's equations in the magnetodynamic regime with linear constitutive laws
    Basis for both the CATI h-phi formulation (Eq. 2) and the 3D reference model (Eq. 3).
  • domain assumption Cable is periodic along ez with periodicity length ℓ = p/Ns and the transposition is captured by circuit connections
    Introduced in Sec. II and used to couple axial and transverse current models.
  • domain assumption Strand current density is purely axial and fields are constant over ℓ; tilt angle neglected
    Stated in Sec. II-A; this is the key reduction that makes the 2D model possible.
  • domain assumption Strands behave as ohmic conductors with very low resistivity for this verification
    Stated in Sec. I and II-A; excludes superconducting nonlinearity and filament effects.
  • domain assumption Interstrand contacts reduce to uniform lumped resistances Rc and Ra with uniform contact resistivity r = R S
    Used in Sec. II-B and III-A; edge contact surfaces approximated as rectangles.
invented entities (1)
  • Coating domain Ωm surrounding each strand
    purpose: Numerical construct to enable definition of currents with cohomology basis functions; assigned high resistivity ρm to stay negligible
    Introduced in Sec. II-A; it is a computational device, not a physical layer in the cable.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Simulation of Rutherford Cable AC Loss and Magnetization with the Coupled Axial and Transverse Currents Method." pith.science (2026). https://pith.science/paper/YUA4OTMU

@misc{pith2026241113347,
  author       = {Pith},
  title        = {Pith review of: Simulation of Rutherford Cable AC Loss and Magnetization with the Coupled Axial and Transverse Currents Method},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YUA4OTMU}},
  note         = {Machine review of arXiv:2411.13347}
}
read the original abstract

The coupled axial and transverse currents (CATI) method was recently introduced to model the AC loss and magnetization in twisted composite superconducting strands with low computational cost and high accuracy. This method involves two-dimensional finite element (FE) models coupled with circuit equations representing the periodicity of the strand. In this paper, we propose to adapt the CATI method to Rutherford cables, which are periodic structures made of transposed superconducting strands. We focus on reproducing the interstrand coupling currents flowing across contact resistances between the strands and we analyze the associated AC loss. We show that results of a reference three-dimensional FE model are accurately reproduced with a strongly reduced computational cost.

Figures

Figures reproduced from arXiv: 2411.13347 by the authors.

Figure 1
Figure 1. Illustration of the CATI method on a Rutherford cable with [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FE mesh for the 3D reference model with Ns = 26 strands (the mesh used for results of Section III is much finer). One periodicity length ℓ = p/Ns is modelled with a periodic mesh and periodic boundary conditions on the top and bottom surfaces. A. Reference solution The reference solution is obtained with a 3D FE model of the cable over one periodicity length ℓ = p/Ns . The mesh TABLE I PARAMETER VALUES. Number of st… view at source ↗
Figure 3
Figure 3. Comparison of the current density distribution in three situations. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (2 more)
Figure 5
Figure 5. Figure 5: Comparison of currents between the CATI method (solid lines) and 3D [PITH_FULL_IMAGE:figures/full_fig_p004_5.png]
Figure 4
Figure 4. Figure 4: Comparison of the loss per cycle and per unit length between the [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

19 extracted references · 19 canonical work pages

  1. [1]

    Quasi-3D magnetic field simulation of superconducting devices with translational symmetry,

    L. A. D’Angelo and H. De Gersem, “Quasi-3D magnetic field simulation of superconducting devices with translational symmetry,” IET Science, Measurement & Technology, vol. 15, no. 3, pp. 319–327, 2021

    work page 2021

  2. [2]

    H-phi formulation in Sparselizard combined with domain decomposi- tion methods for modeling superconducting tapes, stacks, and twisted wires,

    N. Riva, A. Halbach, M. Lyly, C. Messe, J. Ruuskanen, and V . Lahtinen, “H-phi formulation in Sparselizard combined with domain decomposi- tion methods for modeling superconducting tapes, stacks, and twisted wires,” IEEE Transactions on Applied Superconductivity , vol. 33, no. 5, pp. 1–5, 2023

    work page 2023

  3. [3]

    Superconducting magnets,

    M. N. Wilson, “Superconducting magnets,” Clarendon Press, United Kingdom, 1983

    work page 1983

  4. [4]

    External coil coupled loss induced quench (E-CLIQ) system for the protection of LTS magnets,

    T. Mulder, B. Bordini, E. Ravaioli, E. Schnaubelt, M. Wozniak, and A. Verweij, “External coil coupled loss induced quench (E-CLIQ) system for the protection of LTS magnets,” IEEE Transactions on Applied Superconductivity, vol. 33, no. 5, pp. 1–5, 2023

    work page 2023

  5. [5]

    Optimizing secondary CLIQ for protecting high-field accelerator magnets,

    E. Ravaioli, T. Mulder, A. Verweij, and M. Wozniak, “Optimizing secondary CLIQ for protecting high-field accelerator magnets,” IEEE Transactions on Applied Superconductivity, vol. 34, no. 5, pp. 1–5, 2024

    work page 2024

  6. [6]

    Theoretical behavior of twisted multicore superconducting wire in a time-varying uniform magnetic field,

    G. Morgan, “Theoretical behavior of twisted multicore superconducting wire in a time-varying uniform magnetic field,” Journal of Applied Physics, vol. 41, no. 9, pp. 3673–3679, 1970

    work page 1970

  7. [7]

    Coupling losses in finite length of supercon- ducting cables and in long cables partially in magnetic field,

    G. Ries and S. Tak ´acs, “Coupling losses in finite length of supercon- ducting cables and in long cables partially in magnetic field,” IEEE Transactions on Magnetics, vol. 17, no. 5, pp. 2281–2284, 1981

    work page 1981

  8. [8]

    Electrodynamics of superconducting cables in accelerator magnets.,

    A. P. Verweij, “Electrodynamics of superconducting cables in accelerator magnets.,” PhD thesis, Twente University, Enschede , 1997

    work page 1997

Show all 19 references
  1. [9]

    Influence of supercurrents on the stability of superconducting magnets,

    L. Krempasky and C. Schmidt, “Influence of supercurrents on the stability of superconducting magnets,” Physica C: Superconductivity , vol. 310, no. 1-4, pp. 327–334, 1998

    1998

  2. [10]

    Calculation of interstrand coupling losses in superconducting rutherford cables with a continuum model,

    L. Bottura, M. Breschi, and A. Musso, “Calculation of interstrand coupling losses in superconducting rutherford cables with a continuum model,” Cryogenics, vol. 96, pp. 44–52, 2018

    2018

  3. [11]

    Cou- pled axial and transverse currents method for finite element modelling of periodic superconductors,

    J. Dular, F. Magnus, E. Schnaubelt, A. Verweij, and M. Wozniak, “Cou- pled axial and transverse currents method for finite element modelling of periodic superconductors,” Superconductor Science and Technology , vol. 37, no. 9, pp. 1–18, 2024

    2024

  4. [12]

    Gmsh: A 3D finite element mesh generator with built-in pre-and post-processing facilities,

    C. Geuzaine and J.-F. Remacle, “Gmsh: A 3D finite element mesh generator with built-in pre-and post-processing facilities,” International journal for numerical methods in engineering, vol. 79, no. 11, pp. 1309– 1331, 2009

    2009

  5. [13]

    A general environ- ment for the treatment of discrete problems and its application to the finite element method,

    P. Dular, C. Geuzaine, F. Henrotte, and W. Legros, “A general environ- ment for the treatment of discrete problems and its application to the finite element method,” IEEE Transactions on Magnetics, vol. 34, no. 5, pp. 3395–3398, 1998

    1998

  6. [14]

    An open-source finite element quench simulation tool for superconducting magnets,

    A. Vitrano, M. Wozniak, E. Schnaubelt, T. Mulder, E. Ravaioli, and A. Verweij, “An open-source finite element quench simulation tool for superconducting magnets,” IEEE Transactions on Applied Superconduc- tivity, vol. 33, no. 5, pp. 1–6, 2023

    2023

  7. [15]

    STEAM: A hierarchical cosimulation framework for superconducting accelerator magnet circuits,

    L. Bortot, B. Auchmann, I. C. Garcia, A. F. Navarro, M. Maciejewski, M. Mentink, M. Prioli, E. Ravaioli, S. Schoeps, and A. Verweij, “STEAM: A hierarchical cosimulation framework for superconducting accelerator magnet circuits,” IEEE Transactions on Applied Supercon- ductivity...

    2017

  8. [16]

    Bossavit, Computational electromagnetism: variational formulations, complementarity, edge elements

    A. Bossavit, Computational electromagnetism: variational formulations, complementarity, edge elements . Academic Press, 1998

    1998

  9. [17]

    Dual magnetodynamic formulations and their source fields associated with massive and stranded inductors,

    P. Dular, P. Kuo-Peng, C. Geuzaine, N. Sadowski, and J. Bastos, “Dual magnetodynamic formulations and their source fields associated with massive and stranded inductors,” IEEE Transactions on Magnetics , vol. 36, no. 4, pp. 1293–1299, 2000

    2000

  10. [18]

    Homology and cohomology computation in finite element modeling,

    M. Pellikka, S. Suuriniemi, L. Kettunen, and C. Geuzaine, “Homology and cohomology computation in finite element modeling,” SIAM Journal on Scientific Computing , vol. 35, no. 5, pp. B1195–B1214, 2013

    2013

  11. [19]

    Finite-element for- mulations for systems with high-temperature superconductors,

    J. Dular, C. Geuzaine, and B. Vanderheyden, “Finite-element for- mulations for systems with high-temperature superconductors,” IEEE Transactions on Applied Superconductivity , vol. 30, no. 3, pp. 1–13, 2019

    2019

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.