Pith. sign in

REVIEW 2 major objections 4 minor 31 references

Homogenisation and spectral convergence of high-contrast convolution type operators

T0 review · 2 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read This paper establishes that high-contrast convolution-type operators homogenise to a two-scale limit, characterises the limit spectrum via an auxiliary beta-function, and proves that in the whole space the original spectrum converges…

desk verdict A significant and sound extension of high-contrast homogenisation to nonlocal convolution operators; the flagged typo is a misreading, and the only real caveat is that the abstract should advertise Assumption 2.2's scope restriction. read the letter →

arxiv 2507.02638 v1 pith:432E7WOU submitted 2025-07-03 math.AP

Mikhail Cherdantsev , Andrey Piatnitski , Igor Velcic This is my paper · ORCID
classification math.AP MSC 35B2745H9945M0545M1545P05
keywords homogenisationconvolution-typeoperatorhigh-contrastmediumspectralconvergencetwo-scaleGelfandtransformbeta-functionnonlocal
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

High-contrast convolution-type operators describe nonlocal interactions in periodic media where stiff and soft phases alternate. This paper adapts two-scale convergence to these nonlocal operators and obtains the homogenised limit operator in the whole space and in bounded domains with Dirichlet conditions. The main claim is a complete spectral picture: the spectrum of the limit two-scale operator is characterised by an auxiliary $\beta$-function built from the soft-phase operator, and the actual spectrum of the original operators converges, in the whole space, to a generally larger set whose extra part comes from quasiperiodic modes on the soft component. The paper also gives explicit rates for this convergence, controlled by the tail of the convolution kernel; in particular, the Hausdorff distance on any bounded interval is of order $\max\{h(\varepsilon), \varepsilon^{2/3}\}$. This is a spectral convergence result with rates for high-contrast nonlocal operators, showing that nonlocal interactions can produce additional limiting spectrum even when the soft inclusions are disconnected.

What carries the argument

The argument is carried by the scaled Gelfand transform, which fibers $A_\varepsilon$ into a family $A^\theta_\varepsilon$ on the torus, and by norm-resolvent approximation of these fibers with a homogenised operator $A^{h,\theta}_\varepsilon$ whose stiff part is the homogenised matrix $A_{\mathrm{hom}}$ built from correctors solving the stiff-cell problem. The limit two-scale operator $A$ is the sum of a homogenised stiff form and the periodic soft form $a^\#_{\mathrm{soft}}$, and its spectrum is described by the $\beta$-function $\beta(\lambda) = \lambda + \lambda^2\langle (A^\#_{\mathrm{soft}} - \lambda I)^{-1} 1_{Y_{\mathrm{soft}}}\rangle$, giving $\operatorname{Sp}(A) = \{\beta(\lambda) \in \operatorname{Sp}(A_{\mathrm{hom}})\} \cup \operatorname{Sp}(A^\#_{\mathrm{soft}})$. In the whole space the relevant soft object is the non-periodic operator $A_{\mathrm{soft}}$, whose spectrum accounts for quasiperiodic modes; the rate function $h$ comes from the tail $g(r) = \int_{|\xi|>r} a(\xi)|\xi|^2\,d\xi$. A new extension lemma, based on piecewise-constant extension by local averages, supplies the needed a priori bounds under only Assumption 2.2.

What would settle it

Compute, in one dimension with period-one microstructure and soft interval length $1/4$, the spectra $\operatorname{Sp}(A_\varepsilon)$ for a compactly supported even kernel satisfying Assumption 2.2 and compare $\operatorname{Sp}(A_\varepsilon)\cap[0,\Lambda]$ with $G = \{\beta(\lambda) \ge 0\} \cup \operatorname{Sp}(A_{\mathrm{soft}})$ as $\varepsilon \to 0$; finding a point of $G$ that is not approached by $\operatorname{Sp}(A_\varepsilon)$, or a distance that violates the claimed $C(\Lambda)\max\{h(\varepsilon), \varepsilon^{2/3}\}$ bound for a kernel with finite third moment, would disprove the characterisation.

Watch

Extended reading notes

Core claim

The central claim is the equality, for $S = \mathbb{R}^d$, $\lim_{\varepsilon\to 0} \operatorname{Sp}(A_\varepsilon) = G := \{\beta(\lambda) \ge 0\} \cup \operatorname{Sp}(A_{\mathrm{soft}})$, together with the quantitative bound $d_{H,[0,\Lambda]}(\operatorname{Sp}(A_\varepsilon), G) \le C(\Lambda)\max\{h(\varepsilon), \varepsilon^{2/3}\}$, where $h$ is determined by the decay of the convolution kernel and equals $t$ when the kernel has a finite third moment. In general the spectrum of the two-scale limit operator $A$ is only a subset of the limit spectrum, and the inclusion may be strict; the additional limiting spectrum is produced by quasiperiodic approximate eigenfunctions supported on the soft component, and it appears even for disconnected soft inclusions provided the kernel connects them through the stiff phase. For bounded domains the boundary layer spectrum is generally erratic, but for rectangular domains and $\varepsilon = 1/N$ the Hausdorff limit exists and equals the union of the spectra of soft-component operators attached to the vertices.

Load-bearing premise

The load-bearing premise is that the convolution kernel is positive on a ball large enough to connect nearby stiff regions through the nonlocal interaction ($r_a \ge 2r_0 + r_1$); without this connectivity the coercivity of the stiff-cell corrector problem and the extension estimates collapse, and with them the two-scale limit and the spectral characterisation.

Editorial extensions

If this is right

  • In the whole space, the spectrum of $A_\varepsilon$ converges in Hausdorff distance on bounded intervals to $G$, with a rate that is explicit and, for kernels with finite third moment, of order $\varepsilon^{2/3}$.
  • The spectrum of the two-scale limit operator is always contained in the limit spectrum, and the containment is strict for a robust family of high-contrast nonlocal operators; the extra spectrum is carried by quasiperiodic modes on the soft component.
  • Disconnected soft inclusions do not prevent extra limiting spectrum: the soft inclusions communicate when the kernel's support spans the stiff gaps, so the whole-space soft operator $A_{\mathrm{soft}}$ rather than the periodic $A^\#_{\mathrm{soft}}$ controls the limit.
  • In bounded rectangular domains with $\varepsilon = 1/N$, the limiting spectrum is the union of vertex soft spectra, showing that boundary-layer spectrum is stable for self-congruent microstructures along the boundary.
  • Norm-resolvent convergence of the fibered operators yields spectral convergence bounds; the same two-scale compactness result applies to Dirichlet problems without regularity of the phase interface.

Reading between the lines

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

  • A likely extension is the same spectral convergence programme for random statistically homogeneous high-contrast nonlocal media, where the extra spectrum should be governed by an ergodic analogue of $A_{\mathrm{soft}}$ rather than by quasiperiodic modes.
  • The $\varepsilon^{2/3}$ rate appears to come from the split $|\theta| \gtrless \varepsilon^{2/3}$; for kernels with finite third moment one would expect the sharper rate $O(\varepsilon)$ for the Hausdorff distance, matching the norm-resolvent bound $h(\varepsilon)=\varepsilon$.
  • The vertex-spectrum characterisation for rectangles suggests a general polytope rule: for any polytope whose boundary layer is self-similar along a subsequence, the boundary spectrum should be the union of spectra of soft operators on orthants at the vertices, including exterior and interior corners.
  • The boundary-layer spectrum cannot be removed by replacing the soft phase near $\partial S$ with the stiff phase unless the kernel support is smaller than the distance between soft inclusions; this predicts a critical kernel-support threshold for boundary-layer control.
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

2 major / 4 minor

Summary. The paper develops a homogenisation theory for high-contrast symmetric convolution-type operators with periodic microstructure. The authors adapt two-scale convergence to nonlocal convolution operators, obtaining homogenisation results in the whole space and in bounded Lipschitz domains with Dirichlet conditions. They introduce a two-scale limit operator, characterise its spectrum through a Zhikov-type beta-function, and study the limiting behaviour of the spectrum of the original operators. They show that the spectrum of the limit operator is contained in the limit of the spectra, that the reverse inclusion may fail, and they quantify the failure in the whole-space setting via a scaled Gelfand transform, obtaining norm-resolvent and spectral convergence rates. For rectangular domains and a discrete subsequence of the period, they provide a Hausdorff-limit characterisation in terms of vertex soft-component operators.

Significance. If the main results are correct, this is a substantial contribution to nonlocal high-contrast homogenisation. The paper extends the two-scale convergence method to convolution-type operators, provides a new and simple extension theorem for the stiff component under very mild geometric assumptions, and gives explicit spectral descriptions including cases where the limiting spectrum is strictly larger than the spectrum of the two-scale limit. The norm-resolvent estimates with explicit rates in Theorem 2.13 are a notable strength. The treatment is largely self-contained, with appendices supplying the needed extension, compactness, and two-scale convergence tools. The central limitation is that the main theorems require Assumption 2.2, a kernel-support/connectivity condition whose role is load-bearing and which is not advertised in the abstract. Within the stated assumptions, the argument appears coherent and the claims are plausible.

major comments (2)
  1. [Remark 2.8, Eq. (13); Section 4, Eq. (34)] As printed, the displayed representation of A#soft in Eq. (13) is inconsistent with the definition via the form (6). Since every element of L2#(Ysoft) vanishes on Y#stiff, the second term containing 1_{Y#stiff}(xi) is identically zero, so Eq. (13) reduces A#soft to multiplication by m. Likewise, in Eq. (34) the factor 1_{Y#stiff}(x) vanishes on the stated domain L2#(Ysoft), making A#,2soft zero. The correct formula should use 1_{Y#soft} in the second argument (and in the x-factor, or the domain should be used to suppress the x-indicator). This representation is used in Proposition 4.1 and in the spectral decomposition (33)-(34) that underlies Proposition 4.5 and Theorem 2.9, so it is load-bearing and must be corrected.
  2. [Abstract/Introduction; Assumption 2.2] The abstract and the first paragraph of the Introduction advertise the setting as high-contrast operators with integrable kernels in a periodic microstructure, without the support/connectivity condition ra >= 2r0+r1. This condition is genuinely load-bearing: it enters the coercivity of the stiff-cell corrector problem in Lemma 3.1 via inequality (22) and the extension Lemma A.5, and without it the central two-scale limit and the spectral characterizations in Theorems 2.5, 2.9, and 2.13 are not justified. The authors should state Assumption 2.2 in the abstract or otherwise clearly delimit the scope, and ideally add a remark on the degenerate regime in which the stiff-cell form is non-coercive.
minor comments (4)
  1. [Abstract and Introduction] The abstract contains a grammatical typo: 'a subset the limit' should read 'a subset of the limit'; the Introduction also contains 'the the quasiperiodic'.
  2. [Appendix A, proof of Lemma A.6] The phrase 'with m = 3' after the normalization m+r0 = 1 is confusing; since Assumption A.3(b) is stable under enlargement of m, the point should be stated explicitly to avoid the appearance of a contradiction.
  3. [Remark 2.8, Eq. (13)] The factor 1_{Ysoft}(y) in the second term of Eq. (13) is redundant because y is already in the domain of the operator; removing it would make the formula easier to compare with the form definition.
  4. [Section 6, Proposition 6.12] The constant R0 is used in Step 2 but is defined only in the preceding paragraph of the proof; a short definition in the proposition statement or at first use would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the spectral characterizations are derived, not assumed; the only caveats are a stated geometric restriction and self-citations that are not load-bearing.

full rationale

The paper's central objects—the correctors (8), the homogenized matrix (7), the soft-cell operator A#_soft, and the β-function (14)—are each defined from the problem data (kernel a, contrast coefficients Λ0,p, and the periodic geometry). Theorem 2.9 characterizes Sp(A) as {β(λ)∈Sp(A_hom)} ∪ Sp(A#_soft); this is a derived spectral relation, not an identity in which the conclusion is inserted into β. No parameter is fitted to the spectra being predicted; the error function h(t) in Theorem 2.13 is constructed from the tail of the kernel via (98), (102), (119)-(120), independent of the spectra. The approximation scheme follows [16], an external source with no overlapping authors, and the paper verifies hypotheses (H1)-(H4) in Section 6.2 rather than importing the conclusion. The cited results [11, 12] are either restated and proved (Lemma 5.8) or used as tools with the argument adapted and given (Proposition 4.1). Assumption 2.2 (ra ≥ 2r0+r1) is genuinely load-bearing for coercivity of the stiff cell problem (Lemma 3.1(c), Lemma A.5), and the abstract's phrase 'integrable kernels' does not advertise this connectivity requirement; but this limits scope, it is not circular. The strict-inclusion theorems (2.10, 2.12) construct Weyl sequences directly from A_soft/A^v_soft and do not rename the target spectrum. Overall the derivation chain is self-contained: the spectral limits are consequences of two-scale convergence, coercivity estimates, and explicit quasi-mode constructions, with no step reducing to its own output.

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

The central theorems are parameter-free. The constants appearing in the examples of Section 4.2 are illustrative choices for constructing special spectra, not inputs to the main results. No new physical entities, forces, or particles are introduced.

assumptions (4)
  • domain assumption Assumption 2.1: kernel a is nonnegative, even, bounded below by ca on a ball of radius ra, and a(·)(1+|·|^2) ∈ L1(Rd).
    Used throughout: the ellipticity radius gives coercivity of stiff-cell forms in Lemma 3.1; the second moment controls the homogenised matrix and the θ-expansions in Section 6.
  • domain assumption Assumption 2.2: ra ≥ 2r0 + r1, where r0, κ0 quantify the density of Y#stiff in balls and r1, k, N bound discrete paths inside Y#stiff connecting points in each cell.
    Provides connectivity of the stiff component through the convolution kernel; essential for coercivity (22), corrector existence, and the extension Lemma A.5.
  • domain assumption S is either Rd or a bounded open Lipschitz domain, and functions are extended by zero outside S (homogeneous Dirichlet condition).
    Defines the setup; used in two-scale convergence and boundary-layer analysis. Theorem 2.12 additionally restricts S to a rectangle and ε = 1/N with self-congruent boundary geometry.
  • standard math Standard analytic tools are accepted: Lax-Milgram, spectral theorem, Gelfand transform, Schur/Hilbert-Schmidt compactness tests, and the abstract resolvent scheme of [16] (hypotheses H1-H4).
    The paper proves adaptations of propositions from [16] but still relies on the general framework, as well as Pastukhova's Proposition 3.2 equating weak and strong two-scale resolvent convergence.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Homogenisation and spectral convergence of high-contrast convolution type operators." pith.science (2026). https://pith.science/paper/432E7WOU

@misc{pith2026250702638,
  author       = {Pith},
  title        = {Pith review of: Homogenisation and spectral convergence of high-contrast convolution type operators},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/432E7WOU}},
  note         = {Machine review of arXiv:2507.02638}
}
read the original abstract

The paper deals with homogenisation problems for high-contrast symmetric convolution-type operators with integrable kernels in media with a periodic microstructure. We adapt the two-scale convergence method to nonlocal convolution-type operators and obtain the homogenisation result both for problems stated in the whole space and in bounded domains with the homogeneous Dirichlet boundary condition. Our main focus is on spectral analysis. We describe the spectrum of the limit two-scale operator and characterize the limit behaviour of the spectrum of the original problem as the microstructure period tends to zero. It is shown that the spectrum of the limit operator is a subset the limit of the spectrum of the original operator, and that they need not coincide.

Figures

Figures reproduced from arXiv: 2507.02638 by the authors.

Figure 1
Figure 1. Function p defined on (0, 1)2 In this example the periodicity cell is Y = [0, 1) and the soft component is Ysoft = (0, 1 2 ). Choosing a(·) as in the previous example we have ˜a = 1 2 . Next we define the func￾tion p(y, ξ), see [PITH_FULL_IMAGE:figures/full_fig_p022_1.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

31 extracted references · 30 canonical work pages

  1. [16]

    Cooper, I

    S. Cooper, I. Kamotski, V. P. Smyshlyaev, Uniform asymptotics for a family of degenerating variational problems and error estimates in homogenisation theory, arXiv:2307.13151 (2023)

    arXiv 2023

  2. [1]

    Allaire, Homogenization and two-scale convergence, SIAM J

    G. Allaire, Homogenization and two-scale convergence, SIAM J. Math. Anal., 23(6)(1992), 1482–1518

    work page 1992

  3. [2]

    Allaire, C

    G. Allaire, C. Conca, Bloch wave homogenisation and spectral asymptotic analysis, J. Math. Pures Appl. , 77(2) (1998), 153–208

    work page 1998

  4. [3]

    Amaziane, A

    B. Amaziane, A. Piatnitski, E. Zhizhina, Homogenization of diffusion in high contrast random media and related Markov semigroups, Discrete and Continuous Dynamical Systems - B , 28(8) (2023), 4661–4682

    work page 2023

  5. [4]

    Arbogast, J

    T. Arbogast, J. J. Douglas and U. Hornung, Derivation of the double porosity model of single phase flow via homogenisation theory, SIAM J. Math. Anal. , 21 (1990), 823–836

    work page 1990

  6. [5]

    Homogenization of the stochastic double-porosity model

    E. Bonhomme, M. Duerinckx, A. Gloria, Homogenization of the stochastic double- porosity model, arXiv:2502.02847 (2025)

    work page Pith review arXiv 2025

  7. [6]

    Braides, A

    A. Braides, A. Piatnitski, Homogenization of random convolution energies, J. Lond. Math. Soc., 104 (2) (2021), 295–319

    work page 2021

  8. [7]

    Braides, A

    A. Braides, A. Piatnitski, Homogenization of quadratic convolution energies in period- ically perforated domains, Adv. Calc. Var., 15(3) (2022), 351–368

    work page 2022

Show all 31 references
  1. [8]

    Braides, V

    A. Braides, V. Chiado Piat, L. D’Elia, An extension theorem from connected sets and homogenisation of non-local functionals. Nonlinear Anal., 208 (2021), Paper No. 112316, 25 pp

    2021

  2. [9]

    Buˇ zanˇ ci´ c, K

    M. Buˇ zanˇ ci´ c, K. Cherednichenko, I. Velˇ ci´ c, J.ˇZubrini´ c,Spectral and evolution analysis of composite elastic plates with high contrast , J. Elast., 152 (2022), 79–177

    2022

  3. [10]

    Cherdantsev, K

    M. Cherdantsev, K. Cherednichenko, Bending of thin periodic plates, Calc. Var., 54 (2015), 4079–4117

    2015

  4. [11]

    Cherdantsev, K

    M. Cherdantsev, K. Cherednichenko, I. Velˇ ci´ c, High-contrast random composites: ho- mogenisation framework and new spectral phenomena, arXiv:2110.00395 (2024)

    2024 arXiv

  5. [12]

    Cherdantsev, K

    M. Cherdantsev, K. Cherednichenko, I. Velcic, Stochastic homogenisation of high- contrast media. Appl. Anal., 98(1–2) (2019), 91–117. 57

    2019

  6. [13]

    Cherednichenko, A

    K. Cherednichenko, A. V. Kiselev, I. Velˇ ci´ c, J.ˇZubrini´ c, Effective behaviour of critical- contrast PDEs: micro-resonances, frequency conversion, and time dispersive properties. II, Comm. Math. Phys. , 406(4) (2025), Paper No. 72, 72 pp

    2025

  7. [14]

    Cooper, Quasi-periodic two-scale homogenisation and effective spatial dispersion in high-contrast media, Calc

    S. Cooper, Quasi-periodic two-scale homogenisation and effective spatial dispersion in high-contrast media, Calc. Var. Partial Differential Equations, 57(3) (2018), Paper No. 76, pp. 33

    2018

  8. [15]

    Cooper, K

    S. Cooper, K. Cherednichenko, Resolvent estimates for high-contrast homogenisation problems, Arch. Rational Mech. Anal. , 219 (2016), 1061–1086

    2016

  9. [17]

    G. B. Folland. Real Analysis: Modern Techniques and their Applications , Second Edi- tion. John Wiley & Sons, 1999

    1999

  10. [18]

    P. R. Halmos, V. S. Sunder. Bounded Integral Operators on L2 Spaces. Springer-Verlag, 1978

    1978

  11. [19]

    Mielke, A

    A. Mielke, A. Timofte, Two-scale homogenisation for evolutionary variational inequal- ities via the energetic formulation, SIAM J. Math. Anal. , 39 (2) (2007), 642–668

    2007

  12. [20]

    Kondratiev, O

    Yu. Kondratiev, O. Kutoviy, S. Pirogov, Correlation functions and invariant measures in continuous contact model, Infin. Dimens. Anal. Quantum Probab. Relat. Top., 11(2) (2008), 231–258

    2008

  13. [21]

    Kondratiev, S

    Yu. Kondratiev, S. Molchanov, S. Pirogov, E. Zhizhina, On ground state of some non local Schr¨ odinger operators,Appl. Anal., 96(8) (2017), 1390–1400

    2017

  14. [22]

    Kondratiev, S

    Yu. Kondratiev, S. Pirogov, E. Zhizhina, A quasispecies continuous contact model in a critical regime, J. Stat. Phys. , 163(2) (2016), 357–373

    2016

  15. [23]

    S. E. Pastukhova, On the convergence of hyperbolic semigroups in variable Hilbert spaces, J. Math. Sci., 127(5) (2005), 2263–2283

    2005

  16. [24]

    Piatnitski, E

    A. Piatnitski, E. Zhizhina, Homogenization of biased convolution type operators. Asymptotic Analysis, 115(3-4) (2019), 241–262

    2019

  17. [25]

    Piatnitski, E

    A. Piatnitski, E. Zhizhina, Stochastic homogenisation of convolution type operators. J. Math. Pures Appl. ; 134 (2020), 36–71

    2020

  18. [26]

    Piatnitski, E

    A. Piatnitski, E. Zhizhina, Periodic homogenisation of nonlocal operators with a convolution-type kernel, SIAM J. Math. Analysis , 49(1) (2017), 64–81

    2017

  19. [27]

    Piatnitski, V

    A. Piatnitski, V. Sloushch, T. Suslina, E. Zhizhina, On operator estimates in homogeni- sation of nonlocal operators of convolution type. J. Diff. Equations , 352 (2023), 153– 188. 58

    2023

  20. [28]

    Piatnitski, E

    A. Piatnitski, E. Zhizhina, Homogenization of convolution type semigroups in high contrast media, Analysis and Mathematical Physics , 15 (2025), Paper No. 36

    2025

  21. [29]

    Jikov, S.M

    V.V. Jikov, S.M. Kozlov, O.A. Oleinik, Homogenization of Differential Operators and Integral Functionals. Springer-Verlag, Berlin, 1994

    1994

  22. [30]

    Zhikov, On an extension of the method of two-scale convergence and its applica- tions, Sb

    V.V. Zhikov, On an extension of the method of two-scale convergence and its applica- tions, Sb. Math., 191(7–8) (2000), 973–1014

    2000

  23. [31]

    Zhikov, Gaps in the spectrum of some elliptic operators in divergent form with periodic coefficients, St

    V.V. Zhikov, Gaps in the spectrum of some elliptic operators in divergent form with periodic coefficients, St. Petersburg Math. J., 16(5) (2004), 773–790. 59

    2004

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.