Discretely structured magnetic flux concentrators

Xiaoting Feng; Yujun Shi

REVIEW 3 major objections 5 minor 10 references

Discretely structured magnetic flux concentrators

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The paper argues that discretely structured magnetic flux concentrators—arrays of thin high-permeability strips—can concentrate magnetic flux nearly as well as solid structures while using only about 13% of the material volume.

desk verdict Plausible simulation study with a useful design rule, but the quantitative claims need a mesh-convergence check before they can be trusted. read the letter →

arxiv 2506.01424 v1 pith:ZFJC3I2W submitted 2025-06-02 physics.app-ph

Yujun Shi , Xiaoting Feng This is my paper

classification physics.app-ph

keywords magneticfluxconcentratordiscretizedstructurehigh-permeabilitystripsfiniteelementsimulationdepletionzonematerialefficiencymetamaterialsensing

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 asks whether a magnetic flux concentrator (MFC) made of thin, widely separated high-permeability strips can guide and concentrate magnetic flux as effectively as a solid MFC. Using two-dimensional finite element simulations, it finds that discretizing a solid structure into thin strips spaced by a rule based on a single strip's depletion zone can match solid performance while cutting material volume to about a tenth. For a trapezoidal MFC, the best discrete design reaches 95% of the solid structure's average center-section flux density with about 13% of the material volume. The authors argue that in a true three-dimensional device, discretizing along two dimensions could cut material use by two orders of magnitude. The practical stake is lighter, cheaper flux concentrators for magnetic sensing, wireless tracking, and other applications.

What carries the argument

The central object is the isolated rectangular strip of high-permeability material (relative permeability μ = 100) in a uniform field parallel to its long edge. Its depletion zone is defined by the contour where the relative flux density magnitude falls to |B|/|B0| = √2/2, i.e., where the relative magnetic energy density drops below 1/2; L1 is the maximum distance from the strip surface to that contour. The discretization rule places unit strips with center-to-center spacing L1 along the short edge, so each strip sits in the depleted region of its neighbor. The rule works because L1/d grows as the aspect ratio d/L shrinks, so thinner strips leave more empty space between them while still depleting the same flux.

What would settle it

Build or simulate the Discrete-N11 trapezoidal MFC described in the paper (strips of relative permeability 100, aspect ratio 1/60, center-to-center spacing equal to the isolated strip's depletion width L1) and measure the average flux density at the central cross-section; the central claim fails if it does not reach about 95% of the solid design's value while using about 12.7% of the material volume. A second check is to recompute L1 with a different threshold for the depletion contour (e.g., |B|/|B0| = 0.5 instead of √2/2) and verify whether the optimal spacing and the reported savings remain roughly unchanged.

Watch

Extended reading notes

Core claim

The central claim is that spatially discretizing a magnetic flux concentrator does not destroy its function: an array of thin, high-permeability strips spaced so that each strip sits in the flux-depleted zone of its neighbor reproduces the flux concentration of the equivalent solid body. In two-dimensional simulations, a solid rectangle with aspect ratio about 1/3 can be replaced by two or three strips of aspect ratio 1/60 spaced at the isolated strip's depletion width L1, achieving 77–87% of the solid's average center and end cross-section flux density with only 10–15% of the material volume. For a typical trapezoidal MFC, the Discrete-N11 design attains 95% of the solid's average center-section flux density using 12.7% of the material volume. The paper also shows that a discrete MFC expels an interior source's field to the exterior nearly as well as a solid one, with the dense discrete design slightly outperforming the solid at distances near 20 source radii.

Load-bearing premise

The whole design rests on the assumption that spacing strips by the width L1 of the isolated strip's depletion zone—defined at the arbitrary contour |B|/|B0| = √2/2—is the right rule to reproduce a solid structure's flux concentration; that threshold and spacing are not optimized or validated against experiment.

Editorial extensions

If this is right

Discrete MFCs built from thin strips can serve as drop-in replacements for solid MFCs in applications where weight or material cost matters, at 10–13% of the material volume.
Denser discrete designs (more strips) close the performance gap with the solid structure, with Discrete-N11 reaching 95% of the solid's center-section flux density.
The spacing rule based on the isolated strip's depletion width L1 gives a simple, parameter-free recipe for designing discrete MFCs without numerical optimization.
If the one-dimensional savings carry over to two-dimensional discretization in 3D, material consumption could fall by about two orders of magnitude for volumetric devices.
The authors point to low-frequency alternating-field operation (up to tens of kilohertz) and to miniature magneto-mechanical resonators as the next steps, noting that discrete-MFC performance under AC fields is not yet tested.

Reading between the lines

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

The depletion-width rule is defined at an arbitrary threshold (√2/2 in field magnitude); choosing a different threshold would change L1 and likely shift the optimal spacing, so the reported material savings are not necessarily the best achievable.
The claim that two-dimensional discretization multiplies the savings is a dimensional extrapolation; a full 3D simulation of a discretized volumetric MFC would be needed to confirm the 'two orders of magnitude' figure.
The same spacing logic should transfer to other high-permeability geometries, such as cylindrical shells or rod arrays, suggesting a general design principle for lightweight magnetic metamaterials.
Because the discrete design leaves most of the volume as air or non-magnetic material, it may also reduce eddy-current losses under AC fields compared with a solid concentrator of the same envelope—an untested but plausible benefit.

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.

Desk Editor's Note

read the letter

Short take: this is a plausible simulation study with a simple, potentially useful design rule, but the headline numbers are not yet established because no one has shown the FEM results are converged.

What's actually new: Section II's finding that the depleted-zone width L1/d grows as the strip gets thinner, and the resulting spacing rule (L1 = 19.5d for d/L = 1/60), plus the quantitative figures for material savings (0.77 at 0.10 volume for two strips; 0.95 at 0.127 for the N11 trapezoid). The general idea of discrete high-permeability concentrators is already in refs. 6 and 7, and the paper says so; the new bit is the design heuristic and the parameter study. That is worth having.

The paper is honest: no parameters are fit to the target performance, the 3D two-order-of-magnitude claim is framed as an extrapolation, and the method (Elmer 2D FEM) is disclosed.

The soft spot is exactly the stress-test. Every quantitative number here is cooked in an FEM run, and there is no mesh-convergence study, no element order or solver tolerance stated, and no check against an analytic or independently computed magnetostatic case. The strips have sharp corners, so the field has singularities. The contour |B|/|B0| = sqrt(2)/2 that defines L1 is exactly the kind of mesh-sensitive quantity that can move with refinement. If L1 shifts, the spacing in every discrete design shifts with it. I don't doubt the qualitative result -- discrete strip arrays should concentrate flux with less material -- but the specific ratios are conditional until convergence is shown. Also the spacing threshold is arbitrary and not optimized, and there's no head-to-head comparison with the discrete metamaterial designs in refs. 6 and 7, which would help position the contribution. Minor typos (e.g., 'Collage' in the affiliation) don't matter.

Who it's for: applied magnetics people looking for lightweight concentrator designs. It deserves a serious referee: the idea is useful, the write-up is clear, and the missing numerics are fixable. If I were editing, I'd send it out and ask the referee to insist on a convergence check and a benchmark before acceptance. I wouldn't cite the specific ratios in my own work until those checks appear.

Referee Report

3 major / 5 minor

Summary. The manuscript proposes replacing solid magnetic flux concentrators (MFCs) with arrays of thin high-permeability strips separated by air gaps, and studies the concept via two-dimensional finite element simulations in the open-source solver Elmer. A single rectangular strip in a uniform field is shown to deplete the surrounding flux, and the paper defines a depletion width L1 as the maximum distance to the contour |B|/|B0| = sqrt(2)/2. Using L1 as the periodic strip spacing, the authors simulate discrete versions of a rectangular block and a trapezoidal MFC. They report that a discrete structure with eleven strips (Discrete-N11) reaches 0.95 of the solid structure's average center-section flux density with only 0.127 of the material volume (Figure 3f), and that coarser discrete versions reach 0.77–0.87 with 0.10–0.15 of the volume (Figures 2d and 3f). For an interior magnetic source, the discrete MFCs expel the field to the exterior with a far-field gain comparable to the solid MFC (Figure 4). The conclusion extrapolates that discretizing along two dimensions in a 3D device could reduce material use by two orders of magnitude.

Significance. The central idea is simple, plausible, and potentially useful for weight- and cost-sensitive applications such as miniaturized magnetic sensors, wireless tracking, and low-frequency field concentration. The paper has a clear design rule (L1 spacing) and uses transparency-friendly open-source software, and the performance values are computed after the design is fixed, so the comparison is not circular. However, every quantitative result—including L1, the normalized flux ratios, and the material-volume fractions—is an output of FEM simulations that are not verified by mesh-convergence or benchmark studies. Because the strip corners produce field singularities, the extracted contour and derived ratios could shift under mesh refinement. In addition, the two-orders-of-magnitude 3D material reduction is an extrapolation, not a simulation. If the numerical foundation is strengthened and the 3D claim is properly qualified, the work would be a solid contribution to applied magnetics.

major comments (3)

[Section II and Figure 1] The depletion width L1 is extracted from the contour |B|/|B0| = sqrt(2)/2 in a 2D FEM solution around a strip with sharp corners. The manuscript reports no mesh-convergence study, no element order, no solver tolerances, and no benchmark against an analytic magnetostatic solution. The value L1 = 19.5d, which sets the spacing of every discrete design in Sections III and IV, is exactly the kind of quantity that can shift with mesh refinement near a corner singularity. Please add a systematic mesh-refinement study, state the element type/order and solver settings, and validate the solver on a simple geometry with a known solution (e.g., a permeable cylinder in a uniform field) to demonstrate that the contour location and the derived L1 are converged.
[Section III and Figure 2] The design rule assumes that placing strips at intervals of L1—the depletion width of an isolated strip—recreates the flux concentration of a solid block. This is a heuristic: L1 is defined from one strip in isolation, and the paper does not test whether this spacing is near-optimal for an array, whether it depends on the number of strips, or whether it is robust to the chosen threshold |B|/|B0| = sqrt(2)/2. Since all discrete designs in Sections III and IV use this rule, its validity is load-bearing. Please include a parameter sweep over the spacing multiplier (around L1) and, if feasible, over the relative permeability mu_r, and show how the normalized center-section flux and material volume vary. This would establish that the reported 0.77–0.95 ratios are not an artifact of an arbitrary threshold.
[Conclusion and Abstract] The abstract and conclusion state that for three-dimensional structural devices, discretization along two dimensions 'has the potential to reduce material consumption by two orders of magnitude.' This is not supported by the simulations, which are 2D and discretize along only one spatial dimension. The 3D material saving does not necessarily multiply with the number of discretized dimensions, because the depleted zones around orthogonal strips can overlap and the mechanical support of thin sheets is not modeled. This claim should either be backed by a 3D simulation or explicitly labeled as an unverified extrapolation in the abstract and conclusion.

minor comments (5)

[Global] There are several typographical errors: the affiliation reads 'Collage of Physics and Electronics Engineering' (should be 'College'), the acknowledgment reads 'Research Proiect' (should be 'Project'), Section IV uses 'discretizie' and 'discretizied' (should be 'discretize' and 'discretized'), and the legend of Figure 3(e) has 'Soild' (should be 'Solid').
[Section IV.A] The text says 'Figure 4(f) shows the normalized values of the mean magnetic flux density magnitudes at the center cross-sections of the four MFCs,' but the referenced panel is Figure 3(f). Please correct the cross-reference.
[Figures 2 and 3] The 'material volume' reported in the normalized bar charts is computed in 2D, so it is actually an area fraction. Please state this explicitly in the figure captions or text to avoid confusion when comparing with the 3D extrapolation in the conclusion.
[Figure 1(c)] The axes of Figure 1(c) are not fully labeled in the caption; please clarify which curve corresponds to L1/d and which to L1/L, and specify whether the axes are logarithmic.
[Section IV.B] The statement that Discrete-N11 shows a 'slightly higher gain factor' than the solid MFC at y=20R is made without quantified uncertainty or a table of values. Please provide the numerical gain values for all structures, along with the convergence evidence requested in the major comments.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the discrete designs are fixed by a stated heuristic, and all performance ratios are computed from subsequent FEM runs rather than fitted to the target values.

full rationale

The paper's derivation chain is not circular. The central design rule is the depleted-zone width L1, defined in Section II as the maximum distance from a rectangular strip to the contour where |B|/|B0| = sqrt(2)/2. This quantity is extracted from a single-strip simulation and then used as a spacing rule for discrete structures in Sections III and IV. The headline results, such as Discrete-N11 reaching 0.95 of the solid structure's center-section flux density with 0.127 of the material volume, are computed after the discrete geometry is fixed. No parameter is fitted to the solid structure's performance or to the reported normalized ratios. The 0.77, 0.84, 0.95, etc. values are outputs of forward FEM simulations, not targets used to tune the design. The L1 rule is a heuristic motivated by the depleted-zone picture, but it does not constrain the simulated performance to match the solid structure by construction; the paper reports both under-performance and over-performance relative to the solid benchmark. There are no self-citations to prior work by the same authors, and the cited references to Navau et al. and Gleich et al. are contextual rather than load-bearing for the paper's central claim. The absence of a mesh-convergence study and of experimental or analytic benchmarks is a numerical-correctness concern, not a circularity concern, because the FEM computations are not used as both the input and the output of the same inference. The score is therefore 0.

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

The ledger is small but not empty. The key chosen numbers are the strip aspect ratio 1/60, the permeability 100, and the energy-depletion threshold used to set strip spacing. These are design choices, not fits to a target, so circularity burden is low. The main domain assumptions are the 2D infinitely long approximation and linear material response. No new physical entities are introduced.

free parameters (4)

strip aspect ratio d/L = 1/60
Chosen as a representative thin strip in Section II; results from this geometry set the spacing for all discrete designs.
relative permeability mu_r = 100
Assumed constant for all magnetic materials; not varied or experimentally characterized.
L1 spacing multiplier = 19.5 times strip width for d/L=1/60
Defined from the |B|/|B0| = sqrt(2)/2 contour and used as the center-to-center spacing in Figures 2 and 3.
energy-depletion threshold = relative energy density below 1/2 (|B|/|B0| below sqrt(2)/2)
Arbitrary criterion for defining the depleted zone L1; central to the discretization rule but not independently justified.

assumptions (4)

domain assumption 2D simulation with an infinitely long third dimension approximates 3D MFC behavior
Invoked in Section I to justify using MagnetoDynamics2D; 3D edge effects are ignored.
domain assumption Linear magnetostatics with constant relative permeability mu_r = 100
Assumed throughout Sections II-IV; no saturation or nonlinear B-H curve is included.
standard math Finite element solutions are numerically converged and accurate
No mesh convergence study is reported, yet all conclusions depend on FEM accuracy.
ad hoc to paper L1 periodic placement of strips recreates the flux concentration of a solid block
Used to build the Discrete-N designs in Sections III-IV; the threshold and spacing are chosen, not derived from a first-principles homogenization.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Discretely structured magnetic flux concentrators." pith.science (2026). https://pith.science/paper/ZFJC3I2W

@misc{pith2026250601424,
  author       = {Pith},
  title        = {Pith review of: Discretely structured magnetic flux concentrators},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZFJC3I2W}},
  note         = {Machine review of arXiv:2506.01424}
}

read the original abstract

Conventional magnetic flux concentrators (MFCs) are typically designed with solid structures, which may not be optimal for applications requiring lightweight design or material efficiency. In this study, we investigate the feasibility of spatially discretized MFCs in guiding and concentrating magnetic flux, through finite element simulations. Our results demonstrate that discretely structured MFCs can achieve performance comparable to that of solid counterparts while significantly reducing material usage. For instance, even with a non-optimized discretized design, discretization along a single dimension can reduce material usage by an order of magnitude. Moreover, for three-dimensional structural devices, discretization along two dimensions has the potential to reduce material consumption by two orders of magnitude. This may offer a notable advantage in applications where weight reduction and cost efficiency are of primary concern.

Figures

Figures reproduced from arXiv: 2506.01424 by the authors.

**Figure 2.** FIG. 2. (a, b, c) Distribution of magnetic flux around two discrete rectangular structures and their [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

**Figure 3.** FIG. 3. Magnetic flux concentration. (a, b, c, d) Distribution of magnetic flux around three [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: FIG. 4. Magnetic flux expulsion. (a, b, c, d) Distribution of magnetic flux around three discrete [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

10 extracted references · 10 canonical work pages

[1]

Leroy, C

P. Leroy, C. Coillot, A.F. Roux, and G.M. Chanteur. High magnetic field amplification for improving the sensitivity of hall sensors. IEEE Sensors Journal , 6(3):707--713, 2006

work page 2006
[2]

Clark Griffith, Ricardo Jimenez-Martinez, Vishal Shah, Svenja Knappe, and John Kitching

W. Clark Griffith, Ricardo Jimenez-Martinez, Vishal Shah, Svenja Knappe, and John Kitching. Miniature atomic magnetometer integrated with flux concentrators. Applied Physics Letters , 94(2):023502, 01 2009

work page 2009
[3]

Ilja Fescenko, Andrey Jarmola, Igor Savukov, Pauli Kehayias, Janis Smits, Joshua Damron, Nathaniel Ristoff, Nazanin Mosavian, and Victor M. Acosta. Diamond magnetometer enhanced by ferrite flux concentrators. Phys. Rev. Res. , 2:023394, Jun 2020

work page 2020
[4]

Kiyoshi, S

T. Kiyoshi, S. Choi, S. Matsumoto, T. Asano, and D. Uglietti. Magnetic flux concentrator using gd-ba-cu-o bulk superconductors. IEEE Transactions on Applied Superconductivity , 19(3):2174--2177, 2009

work page 2009
[5]

Zhang, S

Z.Y. Zhang, S. Matsumoto, S. Choi, R. Teranishi, and T. Kiyoshi. A new structure for a magnetic field concentrator using NbTi sheet superconductors. Physica C: Superconductivity and its Applications , 471(21):1547--1549, 2011. The 23rd International Symposium on Superconductivity

work page 2011
[6]

Magnetic energy harvesting and concentration at a distance by transformation optics

Carles Navau, Jordi Prat-Camps, and Alvaro Sanchez. Magnetic energy harvesting and concentration at a distance by transformation optics. Phys. Rev. Lett. , 109:263903, Dec 2012

work page 2012
[7]

Experimental realization of magnetic energy concentration and transmission at a distance by metamaterials

Jordi Prat-Camps, Carles Navau, and Alvaro Sanchez. Experimental realization of magnetic energy concentration and transmission at a distance by metamaterials. Applied Physics Letters , 105(23):234101, 12 2014

work page 2014
[8]

Available: https://www.elmerfem.org/blog/

Elmer—open source finite element software for multiphysical problems. Available: https://www.elmerfem.org/blog/

work page

Show all 10 references

[9]

A time-varying magnetic flux concentrator

B Kibret, M Premaratne, P M Lewis, R Thomson, and P B Fitzgerald. A time-varying magnetic flux concentrator. Journal of Physics D: Applied Physics , 49(33):335003, jul 2016

2016
[10]

Miniature magneto-mechanical resonators for wireless tracking and sensing

Bernhard Gleich, Ingo Schmale, Tim Nielsen, and Jürgen Rahmer. Miniature magneto-mechanical resonators for wireless tracking and sensing. Science , 380(6648):966--971, 2023

2023

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

Leroy, C

P. Leroy, C. Coillot, A.F. Roux, and G.M. Chanteur. High magnetic field amplification for improving the sensitivity of hall sensors. IEEE Sensors Journal , 6(3):707--713, 2006

work page 2006

[2] [2]

Clark Griffith, Ricardo Jimenez-Martinez, Vishal Shah, Svenja Knappe, and John Kitching

W. Clark Griffith, Ricardo Jimenez-Martinez, Vishal Shah, Svenja Knappe, and John Kitching. Miniature atomic magnetometer integrated with flux concentrators. Applied Physics Letters , 94(2):023502, 01 2009

work page 2009

[3] [3]

Ilja Fescenko, Andrey Jarmola, Igor Savukov, Pauli Kehayias, Janis Smits, Joshua Damron, Nathaniel Ristoff, Nazanin Mosavian, and Victor M. Acosta. Diamond magnetometer enhanced by ferrite flux concentrators. Phys. Rev. Res. , 2:023394, Jun 2020

work page 2020

[4] [4]

Kiyoshi, S

T. Kiyoshi, S. Choi, S. Matsumoto, T. Asano, and D. Uglietti. Magnetic flux concentrator using gd-ba-cu-o bulk superconductors. IEEE Transactions on Applied Superconductivity , 19(3):2174--2177, 2009

work page 2009

[5] [5]

Zhang, S

Z.Y. Zhang, S. Matsumoto, S. Choi, R. Teranishi, and T. Kiyoshi. A new structure for a magnetic field concentrator using NbTi sheet superconductors. Physica C: Superconductivity and its Applications , 471(21):1547--1549, 2011. The 23rd International Symposium on Superconductivity

work page 2011

[6] [6]

Magnetic energy harvesting and concentration at a distance by transformation optics

Carles Navau, Jordi Prat-Camps, and Alvaro Sanchez. Magnetic energy harvesting and concentration at a distance by transformation optics. Phys. Rev. Lett. , 109:263903, Dec 2012

work page 2012

[7] [7]

Experimental realization of magnetic energy concentration and transmission at a distance by metamaterials

Jordi Prat-Camps, Carles Navau, and Alvaro Sanchez. Experimental realization of magnetic energy concentration and transmission at a distance by metamaterials. Applied Physics Letters , 105(23):234101, 12 2014

work page 2014

[8] [8]

Available: https://www.elmerfem.org/blog/

Elmer—open source finite element software for multiphysical problems. Available: https://www.elmerfem.org/blog/

work page