Pith. sign in

REVIEW 4 major objections 4 minor 79 references

A single tunable weight turns higher-order homogenized wave equations into well-posed strain-gradient models for any periodic elastic medium.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-01 01:57 UTC pith:WEOB3ZJ2

load-bearing objection Solid well-posedness theorem for a Boussinesq-corrected strain-gradient elastodynamic model; the formal asymptotic validity of the effective equation is the main caveat. the 4 major comments →

arxiv 2607.25602 v1 pith:WEOB3ZJ2 submitted 2026-07-28 math.AP math-phmath.MPphysics.class-ph

Well-posed homogenized strain-gradient models for linear elastodynamics and elastostatics in arbitrary periodic media

R\'emi Cornaggia , Marc Bonnet , Giuseppe Rosi , Saad El Ouafa , Nicolas Auffray This is my paper
classification math.AP math-phmath.MPphysics.class-ph MSC 35B2774Q0574J0574B05
keywords homogenizationstrain-gradient elasticityelastodynamicstwo-scale asymptotic expansionBoussinesq trickwell-posednessreciprocity identitiesdispersion relations
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper establishes that the higher-order homogenized equations obtained by pushing two-scale asymptotic expansions to order ϵ² can be made mathematically sound for any periodic elastic medium in two or three dimensions. The key move is a sign-restoring 'Boussinesq trick': adding a tunable scalar weight ϑ to a Laplacian of the leading-order balance equation replaces the potentially sign-indefinite higher-order effective tensors with ϑ-dependent tensors, yielding a genuine strain-gradient elasticity (SGE) model whose strain and kinetic energy densities are positive definite. With ϑ above an explicitly computable threshold, the resulting transient wave equation has a unique solution for given initial data and forcing, and conserves energy when unforced. Reciprocity identities for the cell problems reduce the required computations to two static cell functions and one inertial cell function. The upshot is a practical, well-posed effective model that captures anisotropic and dispersive wave propagation in architectured materials at a fraction of the cost of resolving the microstructure, with dispersion errors of fourth order in the scale parameter.

Core claim

Central claim: the fourth-order PDE from second-order homogenization of elastodynamics, generically ill-posed due to wrong-sign higher-order tensors, can be recast as a well-posed strain-gradient (SGE) model. The Boussinesq trick adds a tunable weight ϑ times the Laplacian of the leading-order balance equation, producing tensors A2(ϑ)=ϑD2−C2+c2 and J2(ϑ)=ϑI2−B2+b2. For ϑ above a threshold from small eigenvalue problems, strain and kinetic energies are positive definite and Hille-Yosida gives a unique solution with conserved energy; statics follow the same route. Reciprocity identities compute all effective tensors from χ1, χ2, ζ2, skipping the third-order cell problems.

What carries the argument

The central mechanism is the Boussinesq trick: adding to the O(ϵ²) macroscopic balance equation a tunable scalar weight ϑ times the Laplacian of the leading-order balance equation (and, in dynamics, an additional identity obtained by applying a filtered differential operator to the leading-order equation). This converts the raw effective tensors C2 and B2 — which can be sign-indefinite and are responsible for ill-posedness — into ϑ-dependent tensors A2(ϑ)=ϑD2−C2+c2 and J2(ϑ)=ϑI2−B2+b2 with the symmetry and positivity needed for a valid SGE model. The threshold ϑ0 is obtained from the symmetric eigenvalue problems (8.2)/(8.6)/(8.8), and the reciprocity identities of Propositions 1 and 2 suppl

Load-bearing premise

The load-bearing premise is that the truncated two-scale expansion, together with the Boussinesq-trick modifications, yields an effective PDE with a true O(ϵ³) residual for full elasticity; the paper supplies no rigorous a posteriori error estimate for general 2D/3D elastic media, so if the residual is worse than O(ϵ³) for rough, high-contrast, or perforated cells, the well-posed model may be approximating the wrong asymptotic limit.

What would settle it

For a specific non-centrosymmetric high-contrast 2D cell, compute the residual term Oϵ in (3.39)–(3.40) from the computed cell functions and measure how it scales with ϵ; if the L² norm of the residual does not decay as ϵ³, the central asymptotic claim fails. Alternatively, run the well-posed SG(ϑ) model and a Floquet-Bloch reference for a 3D perforated cell and check whether the relative phase-velocity error follows the predicted O(ϵ⁴) trend; deviation would indicate that the formal model, though well-posed, is not the actual asymptotic model.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • For any periodic medium satisfying the minimal ellipticity and boundedness assumptions, the transient SGE effective wave equation has a unique solution for ϑ above the threshold, with energy conservation in the unforced case.
  • The effective dispersion relations come from a generalized Christoffel equation whose matrices are Hermitian positive definite for ϑ>ϑ0, and the phase-velocity expansion has a real O(δ²) correction; the O(δ) term vanishes except for non-centrosymmetric cells with a double leading-order eigenvalue.
  • Only the first and second static cell functions (χ1, χ2) and the second inertial cell function (ζ2) are needed to build the model; the third-order cell problems are not required for evaluating the effective tensors, cutting the 3D cell-solution count from 90 to 27.
  • The dynamic strain-gradient elasticity tensor differs from its static counterpart whenever the mass density is heterogeneous, so using a static SG tensor in dynamics requires the fourth-order-in-time term or the modified tensor A2(ϑ).
  • The well-posedness statement applies beyond this homogenization context to any SGE material whose operators satisfy the positivity and coercivity properties of Proposition 4.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • A testable extension is to choose ϑ just above the threshold, or to use a direction- or frequency-dependent weight, to recover some of the dispersion accuracy that the numerical examples show is lost when ϑ is taken larger than necessary.
  • Because the paper proves well-posedness but not a rigorous a posteriori error estimate for full elasticity, the practical validity for high-contrast, perforated, or 3D cells remains open; a direct residual computation for such cells would settle whether the O(ϵ³) claim holds.
  • The paper's outlook identifies boundary and interface conditions preserving O(ϵ³) accuracy as an open problem in higher dimensions; without such conditions, finite-domain simulations with the SGE model will require ad-hoc closures, limiting quantitative use.
  • The reciprocity-based reduction suggests a cheap computational pipeline — solve χ1, χ2, ζ2, then two eigenvalue problems — that could be embedded in topological optimization of microstructures for target dispersion, matching the paper's stated de-homogenization motivation.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper derives second-order two-scale homogenized models for linear elastostatics and elastodynamics in periodic media in arbitrary dimension (d=2,3). It develops reciprocity identities that give alternative expressions for the effective stiffness and inertia tensors, reducing the number of cell problems, and introduces a one-parameter ('Boussinesq trick') family of fourth-order-in-space effective PDEs. The main results are: (i) for a weight ϑ above an explicitly computed threshold, the homogenized strain-gradient elasticity (SGE) energy densities are positive definite; (ii) the corresponding transient wave equation (5.22) is well-posed in free space, proved by verifying maximal monotonicity and applying Hille-Yosida; and (iii) the dispersion relation of the SGE model reproduces the expected second/fourth-order asymptotic behavior in numerical tests on three 2D cells (square, hexagonal, and a non-centrosymmetric Z3 lattice). The well-posedness theorem is internally sound; the fragile part is the claim that the corrected SGE equation is the actual effective equation, which rests on a formal two-scale expansion truncated at O(ε^3) with no a posteriori error estimate supplied for full elasticity.

Significance. If the asymptotic validity were established, this would be a substantial contribution: it provides a systematic, microstructure-based construction of well-posed strain-gradient effective models for arbitrary periodic media, with a transparent, computable threshold for the tunable weight. The reciprocity reductions are practically valuable (e.g., 27 cell problems instead of 90 in 3D dynamics). The well-posedness proof for the corrected model is genuine and does not rely on fitting dispersion data; the effective tensors are computed from cell problems, and the Boussinesq weight is selected from explicit eigenvalue problems. The numerical experiments give honest evidence of the expected asymptotic rates and illustrate the ill-posedness of the uncorrected SG(0) model. However, the claim that the SGE equation is the homogenized model with an O(ε^3) residual is not proven for full 2D/3D elasticity; the manuscript itself defers field-convergence studies to future work. This is the main gap between the paper's title and what is demonstrated.

major comments (4)
  1. [§3.3, §4.3, Eqs. (3.39)–(3.40), (4.29)–(4.30)] The central quantitative claim is that U_ϵ satisfies the effective PDEs up to an O(ϵ^3) residual. This is asserted on the basis of the formal two-scale expansion (2.7)-(2.8) truncated at O(ϵ^3). No a posteriori error estimate is given for full 2D/3D elastodynamics; the cited [21] is leading-order only and [5,3] treat scalar or 1D problems. For rough, high-contrast, or perforated cells, the residual may fail to be genuinely O(ϵ^3). The manuscript itself acknowledges in §9 that a quantitative field-convergence study is still needed. This missing support is load-bearing because, without it, Proposition 5 proves well-posedness of an SGE-like PDE but not that this PDE is the effective equation of the microstructure.
  2. [§6.2 (perforated media) and §2.1] The paper claims the entire homogenization procedure carries over to periodically perforated media by replacing Y with Y^s and citing [22]. For the leading-order problem this is standard, but for second-order strain-gradient terms the effect of traction-free holes on the higher-order cell problems and on the O(ϵ^3) residual is not analyzed. Given that all three numerical examples are perforated, the absence of a rigorous justification for this extension weakens the 'arbitrary periodic media' claim exactly in the regime where the central residual estimate is most questionable.
  3. [§8.5, Proposition 5] The Hille-Yosida proof is internally consistent and a genuine contribution. However, the statement is restricted to the unbounded-domain problem in R^d with the particular variational definition of R_ϑ^{-1}. The paper correctly notes in §9 that boundary and interface conditions for the SGE model in finite domains remain open. This is a scope limitation rather than an error, but it should be clearly separated from the well-posedness claim in free space, which is fully proved.
  4. [§7 and §6.7, Proposition 6] The numerical validation covers only three two-dimensional, relatively smooth periodic cells. Proposition 6 establishes the O(δ^2) expansion of phase velocities under a formal small-δ assumption, but the agreement with Floquet-Bloch results is only empirical and is not used to validate the O(ϵ^3) residual. In particular, the paper does not test the model on rough, high-contrast, or genuinely three-dimensional cells, which are the cases where the missing error estimate is most relevant.
minor comments (4)
  1. [Title/Abstract] The running title contains a typo ('ELASTOST A TICS'); the abstract also renders 'R^d' as 'R d' in several places. Please proofread.
  2. [§8.5, surjectivity argument] In the proof of Proposition 5, the scalar parameter in (8.15) is written as 'some µ∈R^d'; it should be µ∈R. This is a typographical error and does not affect the argument.
  3. [Remark 18] The remark refers to 'Prop. 4(b)' for the positivity of the Christoffel eigenvalues, but the relevant property is the positive definiteness of the Fourier symbols Q_ϑ(iξ), R_ϑ(iξ), which is Prop. 4(c). Please correct the cross-reference.
  4. [§6.6] The discussion of the fourth-order-in-time equation is candid about the lack of well-posedness guarantees and the Ostrogradsky instability. This is useful context, but the paragraph could be shortened, as it is not used in the main line of the paper.

Circularity Check

0 steps flagged

No significant circularity: cell-problem-based effective tensors, explicit positivity construction for ϑ, Hille-Yosida proof, and external Floquet-Bloch/direct-simulation benchmarks make the derivation self-contained.

full rationale

The paper's central derivation is not circular. The effective tensors C0, C1, C2, B1, B2, β2 are computed by solving the explicit cell problems (3.6), (3.15), (4.6) and evaluating averaged expressions such as (3.12), (3.20), (3.27), (4.11), (4.17), (4.20); none of these quantities is fitted to the Floquet-Bloch or transient data used later for validation. The Boussinesq weight ϑ is introduced as a tunable parameter, and Propositions 3 and 4 prove existence of thresholds ϑe, ϑk via the explicit eigenvalue problems (8.6) and (8.8), so the well-posedness condition ϑ > ϑ0 is a constructed sufficient condition, not a fitted prediction. Proposition 5's well-posedness proof is a direct verification of the Hille-Yosida hypotheses (monotonicity computed in (8.14), surjectivity through a Lax-Milgram argument) and does not import any uniqueness or validity result from the authors' prior papers. The cited results for cell-problem solvability (Lemma 1, [14]) and leading-order ellipticity ([21, Thm. 10.11]) are classical external theorems. Self-citations such as [10,11,23,25,67] occur in contextual remarks, symmetry classifications, or consistency comparisons; none is load-bearing for the homogenization or well-posedness claims. The main limitation—the absence of a rigorous a posteriori error estimate for the O(ϵ^3) residual in full 2D/3D elastodynamics, acknowledged in the introduction and Section 9—is an accuracy/validity caveat, not a circularity. Because the numerical validation compares against independent Floquet-Bloch computations and direct microstructured simulations, the model is externally benchmarked. No step in the derivation chain reduces by construction to its inputs, so the circularity score is 0.

Axiom & Free-Parameter Ledger

1 free parameters · 8 axioms · 0 invented entities

No new physical entities are introduced. The only new construction is the mathematical weight ϑ and the corrective tensor choices D2, I2, c2, b2, which are part of the method rather than new physics.

free parameters (1)
  • Boussinesq weight ϑ = Chosen values: 0.16 (D4), 0.35 (D6), 0.05 (Z3); thresholds ˚ϑe, ϑe, ϑk listed in Table 1
    Introduced in Eq. (3.38)/(4.28) to enforce positivity and coercivity. It is constrained only to exceed eigenvalue thresholds (8.2)/(8.6)/(8.8), not fixed by the microstructure, so the effective SGE model is a one-parameter family rather than a unique prediction.
axioms (8)
  • standard math Solvability of cell problems via Fredholm alternative and elliptic variational formulation (Lemma 1)
    Section 2.4; used throughout to define χ1, χ2, ζ2 and to set up reciprocity.
  • standard math Reciprocity identity for cell solutions (Lemma 2)
    Section 2.4; the basis for Propositions 1 and 2.
  • standard math Hille-Yosida theorem, Lax-Milgram, Gronwall, Fourier/Parseval identities
    Sections 5.4 and 8.2-8.5; used to prove well-posedness and coercivity.
  • domain assumption Material regularity and ellipticity: ρ ∈ L∞, ρ ≥ ρm>0; C ∈ L∞ and ε:C:ε ≥ κm|ε|^2
    Section 2.1; minimal conditions for the cell problems and the homogenization procedure.
  • domain assumption Scale separation ϵ=ℓ/L≪1 and validity of the formal two-scale expansion (2.7)-(2.8)
    Section 2.2; the entire derivation is formal; no a posteriori error estimate is proved.
  • domain assumption Body force depends only on the slow variable, F(X,t)=F(x,t)
    Equation (2.10); used in the solvability conditions and effective forcing terms.
  • domain assumption Unbounded domain setting with compactly-supported test functions; no boundary/interface conditions needed
    Section 5.1; the paper states in Section 9 that boundary and interface conditions for finite domains remain open.
  • ad hoc to paper Boussinesq-trick identities (3.35), (4.24)-(4.25) can be added to the balance equations, and the residual is O(ϵ^3)
    Sections 3.3 and 4.3; the corrective terms and the weight ϑ are chosen inside the paper to enforce positivity, and other corrective choices are acknowledged in Remark 6.

pith-pipeline@v1.3.0-alltime-deepseek · 56367 in / 14319 out tokens · 136278 ms · 2026-08-01T01:57:45.310824+00:00 · methodology

0 comments
read the original abstract

This work develops well-posed homogenized strain-gradient models for linear elastostatics and elastodynamics in periodic media, with a primary focus on elastic wave propagation. Using the classical two-scale asymptotic expansion method, we carry out second-order periodic homogenization for media in $\mathbb{R}^d$ ($d = 2, 3$), with no restriction on the periodicity cell geometry or material distribution. Reciprocity identities applied to suitably chosen pairs of cell solutions provide alternative expressions for the effective stiffness and inertia tensors arising at the leading, first and second orders, substantially reducing the number of cell problems that must actually be solved. Since direct two-scale homogenization beyond leading order generically yields ill-posed effective operators, a Boussinesq-trick procedure is introduced, involving a tunable scalar weight, to recast the resulting fourth-order partial differential equation as a valid strain-gradient elasticity (SGE) model possessing the requisite symmetry, sign-definiteness and coercivity properties. These properties are then used, via the Hille-Yosida theorem, to establish the well-posedness of the corresponding transient initial-value and forced-response problems. Several practically relevant special cases are examined, including centrosymmetric cells, homogeneous mass density and homogeneous elasticity, each yielding simplified model structures. Numerical illustrations on three two-dimensional periodicity cells (square, hexagonal and a non-centrosymmetric chiral lattice) compare the resulting dispersion relations against reference Floquet-Bloch computations and assess transient wave propagation, demonstrating the model's capacity to capture anisotropic and dispersive effects beyond classical elasticity while preserving mathematical well-posedness.

Figures

Figures reproduced from arXiv: 2607.25602 by Giuseppe Rosi, Marc Bonnet, Nicolas Auffray, R\'emi Cornaggia, Saad El Ouafa.

Figure 1
Figure 1. Figure 1: Sketch of an unbounded heterogeneous periodic medium and its periodicity cell Y. Left: un￾bounded (perforated) domain Ω, made of replica of the cell Yℓ = ℓY and supporting the propaga￾tion of waves with large wavelength λ. Right: unit cell Y and associated lattice vectors (v1, v2) [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: D4, D6 and Z3 unit cells. The dimensions are given in [PITH_FULL_IMAGE:figures/full_fig_p028_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Reciprocal unit cells for square and hexagonal cells, with irreducible Brillouin zone (gray triangle). 7.1.1. Computations on the unit cell. For each choice of periodicity cells, the steps outlined in Section 6.1 are performed as follows: (1) Cell geometry and meshes are generated using the free GMSH mesher [46]. (2) Computations on the periodicity cell using the FreeFEM finite element platform [48], with … view at source ↗
Figure 4
Figure 4. Figure 4: From left to right: a zoom on the D4, D6 and Z3 meshes used for FEM computations [PITH_FULL_IMAGE:figures/full_fig_p029_4.png] view at source ↗
Figure 2
Figure 2. Figure 2: In the transient computations, the displacement magnitude is plotted. [PITH_FULL_IMAGE:figures/full_fig_p030_2.png] view at source ↗
Figure 5
Figure 5. Figure 5: D4 unit cell: dispersion diagram and dispersion branches of the LOE and SG models [PITH_FULL_IMAGE:figures/full_fig_p030_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: D4 unit cell: errors on the dispersion relations for all models, averaged on the propagation direction θ (here errors for branches at θ = 0, 5, . . . , 45◦ were computed), and up to κ = π/2ℓ, i.e. the middle of the Brillouin zone in the horizontal direction [PITH_FULL_IMAGE:figures/full_fig_p031_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: D4 unit cell: shear waves generated using a central frequency 0.36 MHz. Wavefronts (displace￾ment magnitude) at t = 0.20 µs, in the leading-order homogenized Cauchy medium (top left), the microstructure (top middle) and the SG(ϑ) medium (top right). Comparisons between the microstructured and homogenized solutions: LOE (bottom left) and SG(ϑ) (bottom right). with the presence of a nearly flat shear-wave br… view at source ↗
Figure 8
Figure 8. Figure 8: D6 unit cell: dispersion diagram and dispersion branches of the LOE and SG models. 10−2 5.10−2 10−1 3.10−1  = κ`/2π = δ/2π 10−5 10−4 10−3 10−2 10−1 Relative error First branch (S-like) 10−2 5.10−2 10−1 3.10−1  = κ`/2π = δ/2π 10−5 10−4 10−3 10−2 10−1 Second branch (P-like) LOE O( 2) SG(ϑ) SG(0) O( 4) [PITH_FULL_IMAGE:figures/full_fig_p032_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: presents the relative error on dispersion for both branches, averaged on the propagation direction θ ∈ [0, π/6]. Again second-order and fourth-order errors are observed for the LOE and SG models, respectively [PITH_FULL_IMAGE:figures/full_fig_p032_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: D6 unit cell: shear waves generated using a central frequency 0.11 MHz. Wavefronts (displace￾ment intensity) at t = 0.67 µs in the leading-order homogenized Cauchy medium (left), and in the microstructure and the SG(ϑ) medium, plotted side-by-side for comparison (right). successfully reproduces this anisotropic wave propagation, in agreement with the anisotropy predicted by the corresponding dispersion br… view at source ↗
Figure 11
Figure 11. Figure 11: Z3 unit cell: dispersion diagram and dispersion branches of the LOE and SG models. 5The situation is obviously completely different in R 3 [PITH_FULL_IMAGE:figures/full_fig_p033_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Z3 unit cell: errors on the dispersion relations for all models, averaged on the propagation direction θ (here errors for branches at θ = 0, 5, . . . , 30◦ were computed), and up to κ = π/2ℓ, i.e. the middle of the Brillouin zone in the horizontal direction [PITH_FULL_IMAGE:figures/full_fig_p034_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Z3 unit cell: shear waves generated using a central frequency 0.12 MHz. Wavefronts (displace￾ment magnitude) at t = 0.62 µs in the microstructured medium (left) and comparisons with the leading-order homogenized Cauchy medium (middle), and in the SG(ϑ) medium (right). model are consequently closer to those of the SG(0) model. Contrarily to the previous examples, both SG models provide dispersion curves sp… view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

79 extracted references

  1. [1]

    Abdulle and T

    A. Abdulle and T. Pouchon,Effective models and numerical homogenization for wave propagation in heterogeneous media on arbitrary timescales, Found. Comput. Math.20(2020), 1505–1547

    2020

  2. [2]

    Allaire, M

    G. Allaire, M. Briane, and M. Vanninathan,A comparison between two-scale asymptotic expansions and Bloch wave expansions for the homogenization of periodic structures, SeMA Journal73(2016), no. 3, 237–259

    2016

  3. [3]

    Allaire, A

    G. Allaire, A. Lamacz-Keymling, and J. Rauch,Homogenization of locally periodic wave systems for long times,https: //hal.science/hal-04718949/, 2024

    2024

  4. [4]

    Allaire and T

    G. Allaire and T. Yamada,Optimization of dispersive coefficients in the homogenization of the wave equation in periodic structures, Numer. Math.140(2018), 265–326

    2018

  5. [5]

    3, 295–336

    Gr´ egoire Allaire, Agnes Lamacz-Keymling, and Jeffrey Rauch,Crime pays; homogenized wave equations for long times, As- ymptotic Analysis128(2022), no. 3, 295–336

    2022

  6. [6]

    Amrouche and M

    C. Amrouche and M. Fontes,Biharmonic problem in exterior domains, C.R. Acad. Sci. Paris, s´ erie I338(2004), 121–126

    2004

  7. [7]

    Amstutz, S

    S. Amstutz, S. M. Giusti, A. A. Novotny, and E. A. de Souza Neto,Topological derivative for multi-scale linear elasticity models applied to the synthesis of microstructures, Int. J. Numer. Meth. Engrg.84(2010), 733–756

    2010

  8. [8]

    J. P. Aubin,Applied functional analysis (second edition), Wiley, 2000

    2000

  9. [9]

    Audoly and C

    B. Audoly and C. Lestringant,An energy approach to asymptotic, higher-order, linear homogenization, J. Theor. Comput. Appl. Mech. (2023)

    2023

  10. [10]

    Auffray, J

    N. Auffray, J. Dirrenberger, and G. Rosi,A complete description of bi-dimensional anisotropic strain-gradient elasticity, Int. J. Solids Struct.69–70(2015), 195–206

    2015

  11. [11]

    Auffray, Q

    N. Auffray, Q. C. He, and H. Le Quang,Complete symmetry classification and compact matrix representations for 3D strain gradient elasticity, Int. J. Solids Struct.159(2019), 197–210

    2019

  12. [12]

    Bacigalupo and L

    A. Bacigalupo and L. Gambarotta,Second-gradient homogenized model for wave propagation in heterogeneous periodic media, Int. J. Solids Struct.51(2014), 1052 – 1065

    2014

  13. [13]

    Bakhvalov and G

    N. Bakhvalov and G. Panasenko,Homogenisation: Averaging processes in periodic media, Springer Netherlands, 1989

    1989

  14. [14]

    Bensoussan, J

    A. Bensoussan, J. L. Lions, and G. Papanicolaou,Asymptotic analysis for periodic structures, North-Holland, 1978

    1978

  15. [15]

    Boutin,Microstructural effects in elastic composites, Int

    C. Boutin,Microstructural effects in elastic composites, Int. J. Solids Struct.33(1996), 1023 – 1051

    1996

  16. [16]

    Boutin and J

    C. Boutin and J. L. Auriault,Rayleigh scattering in elastic composite materials, Int. J. Engrg. Sc.31(1993), 1669–1689

    1993

  17. [17]

    Brezis,Functional analysis, sobolev spaces and partial differential equations, Springer, 2011

    H. Brezis,Functional analysis, sobolev spaces and partial differential equations, Springer, 2011

    2011

  18. [18]

    Cakoni, B

    F. Cakoni, B. B. Guzina, S. Moskow, and T. Pangburn,Scattering by a bounded highly oscillating periodic medium and the effect of boundary correctors, SIAM J. Appl. Math.79(2019), 1448–1474

    2019

  19. [19]

    Calisti, A

    V. Calisti, A. Leb´ ee, A. A. Novotny, and J. Sokolowski,Emergence of elastostatic strain-gradient effects from topological optimization, Eur. J. Mech. A/Solids100(2023), 104979

    2023

  20. [20]

    P. G. Ciarlet,The finite element method for elliptic problems, North Holland, 1980

    1980

  21. [21]

    Cioranescu and P

    D. Cioranescu and P. Donato,An introduction to homogenization, Oxford University Press, 1999

    1999

  22. [22]

    Cioranescu and J

    D. Cioranescu and J. Saint Jean Paulin,Homogenization of reticulated structures, Springer, 1999

    1999

  23. [23]

    Cornaggia and C

    R. Cornaggia and C. Bellis,Tuning effective dynamical properties of periodic media by FFT-accelerated topological optimization, Int. J. Numer. Meth. Engrg.121(2020)

    2020

  24. [24]

    Cornaggia and B

    R. Cornaggia and B. B. Guzina,Second-order homogenization of boundary and transmission conditions for one-dimensional waves in periodic media, Int. J. Solids Struct.188-189(2020), 88–102

    2020

  25. [25]

    Cornaggia and B

    R. Cornaggia and B. Lombard,A homogenized model accounting for dispersion, interfaces and source points for transient waves in 1d periodic media, M2AN Math. Model. Numer. Anal57(2023), 1413–1444

    2023

  26. [26]

    Cosserat and F

    E. Cosserat and F. Cosserat,Th´ eorie des corps d´ eformables, A. Hermann et fils, 1909

    1909

  27. [27]

    Craster, S

    R. Craster, S. Guenneau, M. Kadic, and M. Wegener,Mechanical metamaterials, Rep. Prog. Phys.86(2023), 094501

    2023

  28. [28]

    P. Curie,Sur la sym´ etrie dans les ph´ enom` enes physiques, sym´ etrie d’un champ ´ electrique et d’un champ magn´ etique, Journal de physique th´ eorique et appliqu´ ee3(1894), 393–415

  29. [29]

    Dautray and J

    R. Dautray and J. L. Lions,Mathematical analysis and numerical methods for science and technology: Volume 5: Evolution problems 1, Springer-Verlag, 1988

    1988

  30. [30]

    D. P. DiVincenzo,Dispersive corrections to continuum elastic theory in cubic crystals, Physical Review B34(1986), 5450. 42 R ´EMI CORNAGGIA, MARC BONNET, GIUSEPPE ROSI, SAAD EL OUAF A, AND NICOLAS AUFFRAY

    1986

  31. [31]

    Dumontet,Study of a boundary layer problem in elastic composite materials, M2AN Math

    H. Dumontet,Study of a boundary layer problem in elastic composite materials, M2AN Math. Model. Numer. Anal20(1986), no. 2, 265–286

    1986

  32. [32]

    Durand, A

    B. Durand, A. Leb´ ee, P. Seppecher, and K. Sab,Predictive strain-gradient homogenization of a pantographic material with compliant junctions, J. Mech. Phys. Solids160(2022), 104773

    2022

  33. [33]

    El Ouafa,Elastodynamic design and optimization of architectured materials, Ph.D

    S. El Ouafa,Elastodynamic design and optimization of architectured materials, Ph.D. thesis, Universit´ e Gustave Eiffel, 2023

    2023

  34. [34]

    V. A. Eremeyev,On well-posedness of the first boundary-value problem within linear isotropic toupin-mindlin strain gradient elasticity and constraints for elastic moduli, Z. Angew. Math. Mech.103(2023), e202200474

    2023

  35. [35]

    foundations and solids, Springer, 2012

    A Cemal Eringen,Microcontinuum field theories: I. foundations and solids, Springer, 2012

    2012

  36. [36]

    Fergoug, A

    M. Fergoug, A. Parret-Fr´ eaud, N. Feld, B. Marchand, and S. Forest,Multiscale analysis of composite structures based on higher-order asymptotic homogenization with boundary layer correction, Eur. J. Mech. A/Solids96(2022), 104754

    2022

  37. [37]

    J. Fish, W. Chen, and G. Nagai,Non-local dispersive model for wave propagation in heterogeneous media: multi-dimensional case, Int. J. Numer. Meth. Engrg.54(2002), 347–363

    2002

  38. [38]

    ,Non-local dispersive model for wave propagation in heterogeneous media: one-dimensional case, Int. J. Numer. Meth. Engrg.54(2002), 331–346

    2002

  39. [39]

    S. Fliss,Wave propagation in periodic media : mathematical analysis and numerical simulation, Habilitation ` a diriger des recherches, Universit´ e Paris Sud (Paris 11), 2019,https://hal.archives-ouvertes.fr/tel-02394976

    2019

  40. [40]

    Forest and K

    S. Forest and K. Sab,Finite-deformation second-order micromorphic theory and its relations to strain and stress gradient models, Math. Mech. Solids (2017), 108128651772084

    2017

  41. [41]

    Fratzl and R

    P. Fratzl and R. Weinkamer,Nature?s hierarchical materials, Prog. Mater. sci.52(2007), no. 8, 1263–1334

    2007

  42. [42]

    Gambin and E

    B. Gambin and E. Kr¨ oner,Higher-order terms in the homogenized stress-strain relation of periodic elastic media, Physica status solidi (b)151(1989), 513–519

    1989

  43. [43]

    Ganz and K

    A. Ganz and K. Noui,Reconsidering the Ostrogradsky theorem: higher-derivatives Lagrangians, ghosts and degeneracy, Clas- sical Quant. Grav.38(2021), 075005

    2021

  44. [44]

    M. G. D. Geers, V. G. Kouznetsova, and W. A. M. Brekelmans,Multi-scale computational homogenization: trends and challenges, J. Comput. Appl. Math.234(2010), 2175–2182

    2010

  45. [45]

    Germain,The method of virtual power in continuum mechanics

    P. Germain,The method of virtual power in continuum mechanics. part 2: Microstructure, SIAM J. Appl. Math.25(1973), 556–575

    1973

  46. [46]

    Geuzaine and J.-F

    C. Geuzaine and J.-F. Remacle,Gmsh: A 3-d finite element mesh generator with built-in pre-and post-processing facilities, Int. J. Numer. Meth. Engrg.79(2009), 1309–1331

    2009

  47. [47]

    J. A. Goldstein,Semigroups of linear operators and applications, Oxford University Press, 1985

    1985

  48. [48]

    Hecht,New development in freefem++, J

    F. Hecht,New development in freefem++, J. Numer. Math.20(2012), 251–265

    2012

  49. [49]

    Hu and C

    R. Hu and C. Oskay,Multiscale nonlocal effective medium model for in-plane elastic wave dispersion and attenuation in periodic composites, J. Mech. Phys. Solids124(2019), 220 – 243

    2019

  50. [50]

    Lakes,Elastic and viscoelastic behavior of chiral materials, Int

    R. Lakes,Elastic and viscoelastic behavior of chiral materials, Int. J. Mech. Sci.43(2001), 1579–1589

    2001

  51. [51]

    Laude,Phononic crystals, De Gruyter, 2015

    V. Laude,Phononic crystals, De Gruyter, 2015

    2015

  52. [52]

    Lazar and G

    M. Lazar and G. A. Maugin,Nonsingular stress and strain fields of dislocations and disclinations in first strain gradient elasticity, Int. J. Engrg. Sc.43(2005), 1157–1184

    2005

  53. [53]

    Lazar and G

    M. Lazar and G. Po,The non-singular green tensor of gradient anisotropic elasticity of helmholtz type, Eur. J. Mech. A/Solids 50(2015), 152–162

    2015

  54. [54]

    Lin and S

    Y.-H. Lin and S. Meng,Leading and second order homogenization of an elastic scattering problem for highly oscillating anisotropic medium, J. Elast.137(2019), 177–217

    2019

  55. [55]

    Maranganti and P

    R. Maranganti and P. Sharma,A novel atomistic approach to determine strain-gradient elasticity constants: tabulation and comparison for various metals, semiconductors, silica, polymers and the (ir) relevance for nanotechnologies, J. Mech. Phys. Solids55(2007), 1823–1852

    2007

  56. [56]

    R. D. Mindlin and N. N. Eshel,On first strain-gradient theories in linear elasticity, Int. J. Solids Struct.4(1968), 109–124

    1968

  57. [57]

    Mindlin,Micro-structure in linear elasticity, Arch

    R.D. Mindlin,Micro-structure in linear elasticity, Arch. Ration. Mech. Anal.16(1964), 51–78

    1964

  58. [58]

    G. Mou, B. Desmorat, R. Turlin, and N. Auffray,On exotic linear materials: 2d elasticity and beyond, Int. J. Solids Struct. 264(2023), 112103

    2023

  59. [59]

    Nazarenko, R

    L. Nazarenko, R. Gl¨ uge, and H. Altenbach,Positive definiteness in coupled strain gradient elasticity, Contin. Mech. Thermodyn. 33(2021), 713–725

    2021

  60. [60]

    5, 745–752

    S-A Papanicolopulos,Chirality in isotropic linear gradient elasticity, International Journal of Solids and Structures48(2011), no. 5, 745–752

    2011

  61. [61]

    Pavˇ siˇ c,Pais–Uhlenbeck oscillator and negative energies, Int

    M. Pavˇ siˇ c,Pais–Uhlenbeck oscillator and negative energies, Int. J. Geom. Methods Mod.Phys.13(2016), 1630015

    2016

  62. [62]

    A. V. Pichugin, H. Askes, and A. Tyas,Asymptotic equivalence of homogenisation procedures and fine-tuning of continuum theories, J. Sound Vibr.313(2008), 858 – 874

    2008

  63. [63]

    Polizzotto,Anisotropy in strain gradient elasticity: simplified models with different forms of internal length and moduli tensors, Eur

    C. Polizzotto,Anisotropy in strain gradient elasticity: simplified models with different forms of internal length and moduli tensors, Eur. J. Mech. A/Solids71(2018), 51–63

    2018

  64. [64]

    Poncelet, A

    M. Poncelet, A. Somera, C. Morel, C. Jailin, and N. Auffray,An experimental evidence of the failure of cauchy elasticity for the overall modeling of a non-centro-symmetric lattice under static loading, Int. J. Solids Struct.147(2018), 223–237

    2018

  65. [65]

    D. L. Portigal and E. Burstein,Acoustical activity and other first-order spatial dispersion effects in crystals, Phys. Rev.170 (1968), 673

    1968

  66. [66]

    Rosi and N

    G. Rosi and N. Auffray,Anisotropic and dispersive wave propagation within strain-gradient framework, Wave Motion63(2016), 120–134

    2016

  67. [67]

    ,Continuum modelling of frequency dependent acoustic beam focusing and steering in hexagonal lattices, Eur. J. Mech. A/Solids (2019), 103803

    2019

  68. [68]

    G. Rosi, N. Auffray, and C. Combescure,On the failure of classic elasticity in predicting elastic wave propagation in gyroid lattices for very long wavelengths, Symmetry12(2020), 1243

    2020

  69. [69]

    ,Elastic wave propagation in cubic non-centrosymmetric and chiral architectured materials: Insights from strain gra- dient elasticity, Int. J. Solids Struct.305(2024), 113059

    2024

  70. [70]

    G. Rosi, N. Bochud, V.-H. Nguyen, and N. Auffray,Identification of the constitutive parameters of strain gradient elasticity using guided wave propagation in architectured materials, Mech. Res. Commun.138(2024), 104292. WELL-POSED HOMOGENIZED STRAIN-GRADIENT MODELS FOR LINEAR ELASTODYNAMICS 43

    2024

  71. [71]

    L. Rosi, G. Placidi and N. Auffray,On the validity range of strain-gradient elasticity: a mixed static-dynamic identification procedure, Eur. J. Mech. A/Solids69(2018), 179–191

    2018

  72. [72]

    Sanchez-Palencia and A

    E. Sanchez-Palencia and A. Zaoui,Homogenization techniques for composite media, 1 ed., LNP0272, Springer, 1987

    1987

  73. [73]

    V. P. Smyshlyaev and C. D. Cherednichenko,On rigorous derivation of strain gradient effects in the overall behaviour of periodic heterogeneous media, J. Mech. Phys. Solids48(2000), 1325 – 1357

    2000

  74. [74]

    Thbaut, B

    M. Thbaut, B. Audoly, and C. Lestringant,Fixing non-positive energies in higher-order homogenization, J. Mech. Phys. Solids 203(2025), 106168

    2025

  75. [75]

    ,Asymptotic strain-gradient theory for one-dimensional continua, J. Mech. Phys. Solids206(2026), 106392

    2026

  76. [76]

    Tisseur and K

    F. Tisseur and K. Meerbergen,The quadratic eigenvalue problem, SIAM Review43(2001), 235–286

    2001

  77. [77]

    R. A. Toupin,Elastic materials with couple-stresses, Arch. Ration. Mech. Anal.11(1962), 385–414

    1962

  78. [78]

    T.-H. Tran, V. Monchiet, and G. Bonnet,A micromechanics-based approach for the derivation of constitutive elastic coefficients of strain-gradient media, Int. J. Solids Struct.49(2012), 783–792

    2012

  79. [79]

    Wautier and B

    A. Wautier and B. B.. Guzina,On the second-order homogenization of wave motion in periodic media and the sound of a chessboard, J. Mech. Phys. Solids78(2015), 382–404. Institut Jean Le Rond∂’Alembert, Sorbonne universit ´e, CNRS UMR 7190, 75252 Paris, France Email address:remi.cornaggia@sorbonne-universite.fr POEMS (CNRS-INRIA-ENSTA), Dept. of Applied Mat...

    2015