Pith. sign in

REVIEW 3 major objections 3 minor 2 cited by

Homogenization of the stochastic double-porosity model

T0 review · 3 major / 3 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read Stochastic double-porosity homogenization is proven under weak assumptions, and sharp two-scale error rates are established.

desk verdict Qualitative stochastic homogenization for double-porosity is a genuine breakthrough and the proof looks solid; the quantitative theorem is real but rests on a corrector-moment assumption that likely fails in 2D random media, so its advertised generality is overreaching. read the letter →

arxiv 2502.02847 v2 pith:DVVXML2L submitted 2025-02-05 math.AP math.PR

Elise Bonhomme , Mitia Duerinckx , Antoine Gloria This is my paper · ORCID
classification math.APmath.PR MSC 35B2735J1560H2574Q05
keywords double-porositymodelstochastichomogenizationhigh-contrastmediacorrectorstwo-scaleexpansionerrorestimatesresonantregimeextensionoperators
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

The paper claims to resolve two open problems in the homogenization of the stochastic double-porosity model, an elliptic problem with inclusions whose conductivity scales as the square of their size, the resonant regime used for metamaterials and fractured-rock flow. The first result proves qualitative stochastic homogenization under weak conditions that do not require inclusions to be uniformly bounded or separated, using only an extension property with possible loss of integrability and suitable elliptic regularity inside inclusions. The second establishes quantitative error estimates for the two-scale expansion: O(sqrt(ε)) on bounded Lipschitz domains and O(ε) on the whole space, in both periodic and random settings. A sympathetic reader would care because these are the first such rates and the first stochastic qualitative theory beyond well-separated inclusions.

What carries the argument

The argument rests on two structural assumptions: an extension operator H1(p) that continues functions defined outside the inclusions into them, possibly losing integrability, and a localized L^q elliptic regularity estimate H2(q) inside each inclusion. To overcome the loss of integrability, the paper introduces ε-dependent approximate correctors with a massive term ε², and uses Dirichlet correctors on D/ε together with the subadditive ergodic theorem to obtain almost-sure convergence. For the quantitative result, flux correctors σ and inclusion correctors θ are introduced, and the corrector moment bound (1.12) is used to absorb flux and boundary-layer terms via Young's inequality.

What would settle it

Take a stationary ergodic inclusion process satisfying H0, H1(2), H2(2), and uniform Poincaré constants, but with heavy-tailed inclusion sizes such that the inclusion corrector's second moment diverges; if the O(sqrt(ε)) error bound then fails or degrades, the quantitative claim as stated does not extend to that process.

Watch

Extended reading notes

Core claim

The paper establishes that the solution u_ε of the double-porosity problem (1.4) converges almost surely to ū + E[v](f − ū) weakly in L²(D), and that (1 − χ_ε)∇u_ε converges weakly to ā∇ū, where ū solves the homogenized equation (1.7) with homogenized matrix ā given by the soft-inclusion cell formula (1.8). This holds under Assumptions H0, H1(p), and H2(p′) for some 1 < p ≤ 2, covering inclusion processes that are not uniformly bounded or separated, such as random spheres with unbounded radii, long cylinders with weak separation, and subcritical percolation clusters. Under the stronger Assumption H3 and the corrector moment bound (1.12), the paper further proves that the two-scale expansion error is O(sqrt(ε)) on bounded Lipschitz domains and O(ε) on the whole space, which is new both in the periodic and random settings.

Load-bearing premise

For the quantitative rate, the load-bearing premise is that the corrector, flux corrector, and inclusion corrector all have uniformly bounded second moments (condition (1.12)); this is assumed, not proved, for general random media.

Editorial extensions

If this is right

  • Qualitative homogenization now holds for many random inclusion processes without uniform separation or boundedness, including Poisson inclusions and anisotropic cylinders.
  • The quantitative error estimates O(sqrt(ε)) on bounded domains and O(ε) on the whole space apply to both periodic and random double-porosity media, for the first time.
  • The extension to oscillatory data yields qualitative resolvent convergence, a starting point for spectral analysis of high-contrast operators.
  • The two-scale system (1.15) is identified as the natural candidate for describing the spectrum, with an O(ε) error bound under L² forcing.

Reading between the lines

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

  • The corrector moment bound (1.12) is the likely bottleneck for extending the quantitative rate to heavier-tailed inclusion processes; if the second moment of the inclusion corrector diverges, the O(sqrt(ε)) rate may fail or need refinement.
  • The abstract formulation in terms of H1 and H2 separates geometry from probability, suggesting the qualitative theorem may extend to other inclusion geometries not covered by the examples, as long as the two assumptions hold.
  • The use of massive approximate correctors to tame the resonant loss of integrability could be transferable to other high-contrast problems, such as elastic or Maxwell systems, where similar losses appear.
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 / 3 minor

Summary. This paper studies homogenization of elliptic double-porosity equations with random inclusions in the resonant high-contrast regime. The qualitative Theorem 1.1 establishes almost-sure convergence of u_epsilon to a limit characterized by the soft-inclusion homogenized matrix and the resonant cell function v, under extension and regularity assumptions H1(p) and H2(q) that allow inclusions that are not uniformly bounded or separated; Theorem 1.3 gives the corresponding corrector asymptotics. The quantitative Theorem 1.6 claims O(sqrt(epsilon)) (resp. O(epsilon)) two-scale expansion error on bounded Lipschitz domains (resp. on R^d) under Assumption H3 and a uniform corrector second-moment bound (1.12). The proofs use compactness via extension operators and elliptic regularity, an ergodic theorem for resonant cell problems, approximate Dirichlet correctors controlled by the subadditive ergodic theorem, and energy estimates involving flux and inclusion correctors. The appendix surveys sufficient conditions for H1 and H2, including new cases such as anisotropic and non-separated inclusions.

Significance. If the results are correct, the qualitative part is a substantial advance: it removes the uniform boundedness and separation conditions that were required in all previous stochastic treatments of the double-porosity model, and the proof strategy via approximate Dirichlet correctors and the subadditive ergodic theorem is original and well executed. The appendix also provides a useful survey of extension results for gradient fields. However, the quantitative claim as stated is not established for the generic two-dimensional random setting, because the central moment hypothesis (1.12) appears to be incompatible with the stationary-increment structure of soft-inclusion correctors. The paper deserves publication only after this load-bearing point is fixed, either by proving (1.12) under acceptable hypotheses, by restricting Theorem 1.6 to settings where such bounds are known, or by replacing pointwise corrector moments with large-scale averages and accepting the corresponding losses.

major comments (3)
  1. [Section 1.2, Eq. (1.12); Section 3, Step 2] Assumption (1.12) is load-bearing for Theorem 1.6, but it is not a harmless strengthening: in the standard two-dimensional stationary ergodic soft-inclusion model, the corrector phi of Lemma 1.2 has only stationary gradient, not a stationary potential. For any ergodic medium with short-range correlations and nonzero spectral density at zero, E|phi(x)-phi(0)|^2 grows logarithmically in |x|, so sup_x E|phi(x)|^2 is infinite. The proof after (3.6) uses exactly this quantity to bound terms such as epsilon^2 E int_D |phi(x/epsilon)|^2 |nabla^2 bar u|^2 and the analogous sigma and theta terms; with an infinite constant the displayed O(epsilon) estimate does not follow. The references cited after (1.12) provide bounds on gradients, large-scale averages, or spectral quantities, but not pointwise second moments of phi or sigma. Please either prove (1.12) under H3, restrict Theorem 1.6 to settings where such bounds are actually available (for example periodic inclusions, or d >= 3 with sufficient decorrelation), or replace pointwise corrector moments by large-scale averages with the corresponding logarithmic losses. As written, the advertised random-case quantitative result does not follow.
  2. [Appendix A.1.4, Step 2] The proof of Lemma 1.8(iv) depends on a 'trivial generalization' of [27, Lemma 3.14] that is not stated or proved. Since this is the key input for the advertised no-separation convex-inclusion case, the generalization should be formulated explicitly and proved, or the statement should be restricted to the cases covered by the provided argument. As written, this part of the qualitative applicability claim is not self-contained.
  3. [Appendix A.1.2 and A.1.5] The proofs of Lemma 1.8(ii) and (v) are sketches: (ii) leaves 'the remaining details to the reader' and (v) is attributed to Zhikov with details omitted. Since Theorem 1.1 is stated conditionally on H1(p), these appendix items are not logically necessary for the main theorem, but they are used in the examples to argue that the qualitative result covers many new models. For a journal proof, either expand these arguments or clearly mark them as a survey of existing results; in the latter case the newness of the corresponding examples should be stated more cautiously.
minor comments (3)
  1. [Abstract and Introduction] The abstract and introduction state that the paper solves two open problems; given the conditional nature of Theorem 1.6, the wording for the quantitative problem should be softened or accompanied by a clear statement of the additional moment hypothesis (1.12).
  2. [Section 1.2, Remarks 1.7(b)] The sentence 'This extends [14, Theorem 7.8] to the random setting' should be checked against the hypotheses of the two-scale system (1.15); if (1.12) is not assumed there, the claim needs justification.
  3. [Section 1.3, Lemma 1.8(iii)] In the proof of the anisotropic case, the local extension operators P_{n,pm} at the cylinder ends are asserted without details; please add a sentence explaining their construction or a reference, since this is the new part of the lemma.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: all results are conditional on explicit assumptions; the corrector-moment bound (1.12) is an input and not a disguised version of the conclusion.

full rationale

The derivation of the qualitative theorem (Thm 1.1) is self-contained modulo the explicitly stated geometric and regularity assumptions H0-H2 and classical tools (ergodic theorem, subadditive ergodic theorem, and the external reference [27] for soft-inclusion correctors). The homogenized data E[v] and bar{a} are defined through the cell problems (1.6) and (1.8), which do not involve the sequence u_eps; the proof then identifies the limit of u_eps as the solution of (1.7) via compactness, the resonant ergodic lemma, and Tartar's method. The quantitative theorem (Thm 1.6) is explicitly conditional: it assumes H3 and the corrector-moment bound (1.12), and the proof bounds the two-scale expansion error by these assumed moments together with regularity of the homogenized solution. The paper does not claim to derive (1.12) for random media; the sentences citing [20,21,3,15,5] are contextual ('we refer ... for similar corrector bounds'), not a proof step inserted into the theorem. Consequently, even if (1.12) fails for two-dimensional soft-inclusion correctors with stationary increments and divergent variance, that would narrow the applicability of Thm 1.6; it would not make the theorem circular. Self-citations are present ([5,15,19,20,21]) but none is load-bearing: corrector existence in Lemma 1.2 is attributed to the external classical reference [27], the flux and inclusion correctors of Lemma 1.5 are introduced by explicit PDEs, and (1.12) is a hypothesis rather than a consequence of the cited works. No fitted parameter is renamed as a prediction, and no uniqueness theorem from the authors' prior work is invoked to force the choice of ansatz. Verdict: no significant circularity; score 0.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The results are conditional on the explicitly stated geometric and stochastic assumptions H0-H3 and the corrector moment bound (1.12). No free parameters are fitted; the paper is a pure mathematics contribution. The new mathematical constructs (approximate correctors, flux correctors, inclusion correctors) are standard tools in homogenization, not postulated physical objects.

assumptions (6)
  • domain assumption Stationary ergodic inclusion process H0
    The stochastic model: inclusions are disjoint, open, connected, bounded with Lipschitz boundary, complement connected, nontrivial (Section 1.1, Assumption H0).
  • domain assumption Extension property H1(p)
    Abstract assumption that functions in H^1 outside inclusions extend to W^{1,p} with controlled gradient (Section 1.1, Assumption H1(p)).
  • domain assumption Elliptic regularity H2(q)
    Localized Calderon-Zygmund estimate inside inclusions (Section 1.1, Assumption H2(q)).
  • domain assumption Quantitative assumptions H3
    For Theorem 1.6: H1(2) with L2 control and uniform Poincare constants (Section 1.2, Assumption H3).
  • domain assumption Corrector moment bound (1.12)
    Assumed bounded second moments of correctors, flux correctors and inclusion correctors; not proved in the paper for general random media.
  • standard math Classical ergodic theorems and elliptic regularity theory
    Used for subadditive ergodic theorem (Step 2.1 of Theorem 1.1), ergodic theorem, Calderon-Zygmund estimates (Appendix A), maximum principle.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Homogenization of the stochastic double-porosity model." pith.science (2026). https://pith.science/paper/DVVXML2L

@misc{pith2026250202847,
  author       = {Pith},
  title        = {Pith review of: Homogenization of the stochastic double-porosity model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DVVXML2L}},
  note         = {Machine review of arXiv:2502.02847}
}
read the original abstract

This work is devoted to the homogenization of elliptic equations in high-contrast media in the so-called 'double-porosity' resonant regime, for which we solve two open problems of the literature. First, we prove qualitative stochastic homogenization under very weak conditions, which cover the case of inclusions that are not uniformly bounded or separated. Second, under stronger assumptions, we provide sharp error estimates for the two-scale expansion. The main difficulty is related to the loss of integrability of the control in the resonant zones.

Figures

Figures reproduced from arXiv: 2502.02847 by the authors.

Figure 1
Figure 1. This illustrates typical realizations for the different models constructed in Examples 1.10, for which the assumptions of the qualitative homogenization result hold [PITH_FULL_IMAGE:figures/full_fig_p011_1.png] view at source ↗
Figure 3
Figure 3. This illustrates the construction of local neighborhoods at a boundary point, say x∗ = 0 ∈ ∂I′ n [PITH_FULL_IMAGE:figures/full_fig_p031_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. Homogenisation and spectral convergence of high-contrast convolution type operators

    math.AP 2025-07 conditional novelty 7.0 of 10

    The limiting spectrum of high-contrast convolution-type operators is characterized via a beta function and generally contains extra contributions beyond the two-scale limit operator's spectrum.

  2. Quantitative estimates for high-contrast random media

    math.AP 2025-02 conditional novelty 6.0 of 10

    For elliptic equations with random well-separated holes, the paper proves stretched-exponential tails for the regularity radius and sublinear corrector growth under a multiscale spectral gap condition.

Reference graph

Works this paper leans on

36 extracted references · 29 canonical work pages · cited by 2 Pith papers

  1. [1]

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

    1992

  2. [2]

    Arbogast, J

    T. Arbogast, J. Douglas, Jr., and U. Hornung. Derivation of the double porosity model of single phase flow via homogenization theory.SIAM J. Math. Anal., 21(4):823–836, 1990

    work page 1990

  3. [3]

    Armstrong, T

    S. Armstrong, T. Kuusi, and J.-C. Mourrat.Quantitative stochastic homogenization and large-scale regularity, vol- ume 352 ofGrundlehren der Mathematischen Wissenschaften. Springer, Cham, 2019

    work page 2019

  4. [4]

    Quantitative estimates for high-contrast random media

    P. Bella, M. Capoferri, M. Cherdantsev, and I. Velčić. Quantitative estimates for high-contrast random media. Preprint, arXiv:2502.09493

    work page Pith review arXiv

  5. [5]

    Homogenization of active suspensions and reduction of effective viscosity

    A. Bernou, M. Duerinckx, and A. Gloria. Homogenization of active suspensions and reduction of effective viscosity. Preprint, arXiv:2301.00166. 12Views and opinions expressed are however those of the authors only and do not necessarily reflect those of the European Union or the European Research Council Executive Agency. Neither the European Union nor the ...

    work page Pith review arXiv

  6. [6]

    Bouchitté, C

    G. Bouchitté, C. Bourel, and L. Manca. Resonant effects in random dielectric structures.ESAIM Control Optim. Calc. Var., 21(1):217–246, 2015

    work page 2015

  7. [7]

    Bourgeat, A

    A. Bourgeat, A. Mikelić, and A. Piatnitski. On the double porosity model of a single phase flow in random media. Asymptot. Anal., 34(3-4):311–332, 2003

    work page 2003

  8. [8]

    Braides, V

    A. Braides, V. Chiadò Piat, and A. Piatnitski. A variational approach to double-porosity problems.Asymptot. Anal., 39(3-4):281–308, 2004

    work page 2004

Show all 36 references
  1. [9]

    Capoferri, M

    M. Capoferri, M. Cherdantsev, and I. Velčić. Eigenfunctions localised on a defect in high-contrast random media. SIAM J. Math. Anal., 55(6):7449–7489, 2023

    2023

  2. [10]

    Cherdantsev, K

    M. Cherdantsev, K. Cherednichenko, and I. Velčić. High-contrast random composites: homogenisation framework and new spectral phenomena. Preprint, arXiv:2110.00395

    arXiv

  3. [11]

    Cherdantsev, K

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

    2019

  4. [12]

    K. D. Cherednichenko and S. Cooper. Resolvent estimates for high-contrast elliptic problems with periodic coeffi- cients. Arch. Ration. Mech. Anal., 219(3):1061–1086, 2016

    2016

  5. [13]

    K. D. Cherednichenko, Y. Y. Ershova, and A. V. Kiselev. Effective behaviour of critical-contrast PDEs: micro- resonances, frequency conversion, and time dispersive properties. I.Comm. Math. Phys., 375(3):1833–1884, 2020

    2020

  6. [14]

    Cooper, I

    S. Cooper, I. Kamotski, and V. P. Smyshlyaev. Quantitative multiscale operator-type approximations for asymp- totically degenerating spectral problems. Preprint, arXiv:2307.13151

    arXiv

  7. [15]

    Duerinckx and A

    M. Duerinckx and A. Gloria. Quantitative homogenization theory for random suspensions in steady Stokes flow.J. Éc. polytech. Math., 9:1183–1244, 2022

    2022

  8. [16]

    Duerinckx and A

    M. Duerinckx and A. Gloria.On Einstein’s effective viscosity formula, volume 7 ofMemoirs of the European Math- ematical Society. EMS Press, Berlin, 2023

    2023

  9. [17]

    Duminil-Copin, A

    H. Duminil-Copin, A. Raoufi, and V. Tassion. Subcritical phase ofd-dimensional Poisson-Boolean percolation and its vacant set.Ann. H. Lebesgue, 3:677–700, 2020

    2020

  10. [18]

    Gilbarg and N

    D. Gilbarg and N. S. Trudinger.Elliptic partial differential equations of second order. Classics in Mathematics. Springer-Verlag, Berlin, 2001. Reprint of the 1998 edition

    2001

  11. [19]

    Gloria, S

    A. Gloria, S. Neukamm, and F. Otto. A regularity theory for random elliptic operators.Milan J. Math., 88(1):99–170, 2020

    2020

  12. [20]

    Gloria, S

    A. Gloria, S. Neukamm, and F. Otto. Quantitative estimates in stochastic homogenization for correlated fields. Anal. PDE, 14(8):2497–2537, 2021

    2021

  13. [21]

    Gloria and F

    A. Gloria and F. Otto. The corrector in stochastic homogenization: optimal rates, stochastic integrability, and fluctuations. Preprint, arXiv:1510.08290

    arXiv

  14. [22]

    Grisvard

    P. Grisvard. Le problème de Dirichlet dans l’espaceW 1 p. Portugal. Math., 43(4):393–398, 1985/86

    1985

  15. [23]

    M. Heida. Stochastic homogenization on perforated domains? - Extension operators.Networks and Heterogeneous Media, 18:1820–1897, 01 2023

    2023

  16. [24]

    Hornung, editor.Homogenization and porous media, volume 6 ofInterdisciplinary Applied Mathematics

    U. Hornung, editor.Homogenization and porous media, volume 6 ofInterdisciplinary Applied Mathematics. Springer- Verlag, New York, 1997

    1997

  17. [25]

    D. Jerison. Personal communication, 2024

    2024

  18. [26]

    Jerison and C

    D. Jerison and C. E. Kenig. The inhomogeneous Dirichlet problem in Lipschitz domains.J. Funct. Anal., 130(1):161– 219, 1995

    1995

  19. [27]

    V. V. Jikov, S. M. Kozlov, and O. A. Ole˘ ınik.Homogenization of differential operators and integral functionals. Springer-Verlag, Berlin, 1994. Traduit du russe par G. A. Iosif′yan

    1994

  20. [28]

    I. V. Kamotski and V. P. Smyshlyaev. Localized modes due to defects in high contrast periodic media via two-scale homogenization. J. Math. Sci. (N.Y.), 232(3):349–377, 2018

    2018

  21. [29]

    I. V. Kamotski and V. P. Smyshlyaev. Two-scale homogenization for a general class of high contrast PDE systems with periodic coefficients.Appl. Anal., 98(1-2):64–90, 2019

    2019

  22. [30]

    Maz’ya and J

    V. Maz’ya and J. Rossmann.Elliptic equations in polyhedral domains, volume 162 ofMathematical Surveys and Monographs. American Mathematical Society, Providence, RI, 2010

    2010

  23. [31]

    E. M. Stein.Singular integrals and differentiability properties of functions. Princeton Mathematical Series, No. 30. Princeton University Press, Princeton, N.J., 1970

    1970

  24. [32]

    V. V. Zhikov. Averaging of functionals of the calculus of variations and elasticity theory.Izv. Akad. Nauk SSSR Ser. Mat., 50(4):675–710, 877, 1986

    1986

  25. [33]

    V. V. Zhikov. Asymptotic problems connected with the heat equation in perforated domains.Mat. Sb., 181(10):1283– 1305, 1990

    1990

  26. [34]

    V. V. Zhikov. Problems of the continuation of functions in connection with averaging theory.Differentsial′nye Uravneniya, 26(1):39–50, 181, 1990

    1990

  27. [35]

    V. V. Zhikov. On an extension and an application of the two-scale convergence method.Mat. Sb., 191(7):31–72, 2000

    2000

  28. [36]

    V. V. Zhikov. Gaps in the spectrum of some elliptic operators in divergent form with periodic coefficients.Algebra i Analiz, 16(5):34–58, 2004. 38 E. BONHOMME, M. DUERINCKX, AND A. GLORIA (Elise Bonhomme) Université Libre de Bruxelles, Département de Mathématique, 1050 Brussel...

    2004

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.