Pith. sign in

REVIEW 3 major objections 4 minor 40 references

Unfitted Nitsche's method for computing wave modes in topological materials

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read The paper proves that an unfitted Nitsche finite element method computes the bulk dispersion relation and edge modes of high-contrast photonic graphene with optimal $O(h^2)$ accuracy and no spectral pollution, with constants independent…

desk verdict The numerical method looks useful and the high-contrast experiments are convincing, but the central convergence proof relies on a false trace inequality and needs substantial revision. read the letter →

arxiv 1908.06585 v2 pith:VB5ODRX4 submitted 2019-08-19 math.NA cs.NA

classification math.NAcs.NA MSC 65N3065N2535P15
keywords Nitsche'smethodunfittedfiniteelementphotonicgraphenetopologicaledgestateseigenvalueapproximationFloquet-Blochtheoryhigh-contrastmediaspectralpollution
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 establishes that an unfitted Nitsche finite element method, built on a uniform mesh that ignores the material interfaces, can compute both the bulk dispersion relation and the topologically protected edge modes of photonic graphene when the material weight is discontinuous and highly contrasted. Its central theorem is an optimal convergence result: discrete eigenvalues and $L^2$-normalized eigenfunctions approximate the exact wave modes at order $h^2$ in the mesh size $h$, with constants that do not depend on where the interface cuts the mesh. The result matters because Fourier spectral methods, the standard tool in photonics, become unreliable or wrong at high contrast, while body-fitted meshes are costly to generate for honeycomb structures. The proof supplies the missing stability ingredient, a trace inequality on cut elements, and the numerical experiments confirm the predicted rate, reproduce the expected Dirac points and their opening, and identify localized edge modes inside a high-contrast slab.

What carries the argument

The load-bearing object is the unfitted Nitsche finite element space $V_h = V_{1,h} \oplus V_{2,h}$, made of two continuous piecewise-linear spaces on overlapping fictitious domains that are cut by the material interface, coupled through a Nitsche bilinear form with weighted average fluxes and a penalty term $h^{-1}\int_\Gamma [\![u_h]\!][\![v_h]\!]\,ds$. The key new ingredient is a trace inequality on a cut element that bounds the $L^2$ trace of a finite element function on one part of the element by $C h^2 |\Gamma_K|/|K_i|$ times the gradient energy over that same part, a one-sided refinement of earlier two-sided trace theorems; this makes the Nitsche bilinear form coercive and continuous in the mesh-dependent norm $|||\cdot|||_h = \|(\nabla+ik)\cdot\|_{0,\Omega_1\cup\Omega_2} + \left(\sum_{K\in T_{\Gamma,h}} h^{-1}\|\cdot\|^2_{0,\Gamma_K}\right)^{1/2}$. With stability in hand, the paper invokes the abstract spectral approximation theory of compact self-adjoint operators and duality arguments to convert the $O(h)$ energy-norm approximation of the solution operator into $O(h^2)$ convergence of eigenvalues and eigenfunctions.

What would settle it

Fix a high-contrast edge-mode configuration and recompute the isolated in-gap eigenvalue at the same mesh size for increasing truncation lengths, say $L=40,80,160$, while tracking whether the localized eigenfunction's amplitude at the artificial boundaries decays below round-off; any systematic drift of the eigenvalue with $L$, or a boundary-attached profile, would falsify the truncation assumption underlying the edge-state computation. A second check would be to refine the mesh and count eigenvalues in a spectral gap, which would falsify the no-pollution claim if spurious eigenvalues appear.

Watch

Extended reading notes

Core claim

The paper's discovery is that the unfitted Nitsche formulation, originally designed for elliptic interface problems, extends to the eigenvalue problems describing topological photonic materials while preserving optimal finite element accuracy and a pollution-free spectrum. For periodic bulk materials, the authors use a Floquet-Bloch transform, turning the quasi-periodic eigenproblem on the torus into a periodic interface eigenproblem, and discretize it on a uniform triangular mesh whose elements are cut by the material interface. For materials with a line defect, they truncate the infinite cylinder to a finite slab and impose periodic conditions along the edge and homogeneous Dirichlet conditions at the artificial boundaries, so that localized in-gap eigenfunctions are recognized as edge modes. The main estimates state $|E - E_h| \le C h^2 \|g\|_{2,\ast}$ and $\|g - g_h\| \le C h^2 \|g\|_{2,\ast}$ for eigenvalues and eigenfunctions, with the constant independent of mesh size and interface location; a companion result rules out spectral pollution. The numerical experiments verify the rate, show that the Fourier spectral method wrongly opens a gap and breaks symmetry at high contrast, and display edge eigencurves whose eigenfunctions are localized in the center of the computational domain.

Load-bearing premise

The edge-mode computation assumes that the true edge state decays fast enough across the truncated slab that imposing homogeneous Dirichlet boundary conditions at distance $L$ does not change the eigenvalues or wave profiles of interest; the paper relies on the localization property of the eigenfunction and provides numerical but not analytical control of this truncation error.

Editorial extensions

If this is right

  • Bulk dispersion relations for piecewise-constant, high-contrast photonic materials can be computed at optimal second-order accuracy on uniform meshes, with convergence constants independent of interface position.
  • The computed spectrum is pollution-free: for sufficiently small mesh size, no spurious eigenvalues appear away from the true spectrum, so in-gap features are genuine.
  • Edge eigencurves for a line-defect slab can be resolved by truncating the cylinder and using the same unfitted Nitsche discretization; the numerical examples isolate in-gap curves whose eigenfunctions are localized in the interior, marking topological edge states.
  • Fourier spectral methods are not a reliable baseline for discontinuous high-contrast material weights: the paper shows they wrongly open a gap and break spectral symmetry at jump ratio $J=30$, while the unfitted Nitsche method preserves the theoretical picture.
  • The method handles both the symmetric case with Dirac points and the Faraday-rotation case where the gap opens, because the stability analysis accommodates complex matrix-valued material weights.

Reading between the lines

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

  • A natural next step the paper does not take is to prove an $L$-dependent or exponential-decay bound for the eigenfunctions of the discontinuous high-contrast operator, which would turn the finite-slab truncation from a practical choice into a certified approximation.
  • Because the stability proof rests only on the one-sided cut-element trace inequality and coercivity of the Nitsche form, the method should extend to other interface eigenvalue problems, such as phononic crystals, elastic composites, or Schrodinger operators with singular potentials.
  • The comparison suggests a testable diagnostic for existing photonic codes: run a Fourier spectral solver and an interface-respecting finite element solver on the same high-contrast structure; discrepancies near the Dirac point indicate spectral resolution failure, not physics.
  • The observed optimal convergence at high contrast suggests that post-processing recovered gradients, which the authors mention as future work, could push eigenvalue accuracy beyond $O(h^2)$ at nearly no additional cost.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. This paper proposes an unfitted Nitsche finite element method for the eigenvalue problem (1.1) associated with photonic graphene. For bulk materials it applies Floquet-Bloch theory on a torus with a uniform unfitted mesh; for edge modes it truncates the infinite cylinder to a large rectangle with homogeneous Dirichlet conditions. The method uses two independent P1 spaces on the fictitious domains and Nitsche-type interface terms. The authors claim well-posedness, no spectral pollution, and optimal O(h^2) convergence for eigenvalues and L^2 eigenfunctions, and report numerical experiments showing second-order self-convergence and edge states.

Significance. If correct, the method would be practically valuable: body-fitted meshes are avoided, periodic boundary conditions are easy to impose, high-contrast piecewise-constant coefficients are handled, and the numerical comparison shows that Fourier spectral methods fail in this regime. The no-pollution statement and the claimed constants independent of interface location are also attractive. However, the central coercivity proof rests on a trace inequality that is false as stated, so the theoretical claims are not established. The edge-mode computation also lacks a rigorous error bound for the finite-domain truncation. Strengths include the clear physical setting, the explicit use of Babuska-Osborn theory, and the reproducible numerical comparisons.

major comments (3)
  1. [§3.1.2, Eq. (3.22), Appendix A] Lemma 3.4, inequality (3.22), is false. For φ_{i,h} ≡ 1 on an interface element K, the left-hand side equals |Γ_K| > 0 while the right-hand side is 0. The proof in Appendix A says it suffices to verify the inequality for nodal basis functions, but the inequality is not preserved under linear combinations: constants can cancel in the gradient while the trace remains O(1). Since Lemma 3.5 and Theorem 3.6 use (3.22) to bound the Nitsche flux terms and prove coercivity, the subsequent source approximation estimates (4.6)–(4.7), the no-pollution result (Theorem 4.3), and the eigenpair error bounds (Theorem 4.4) are not established. A correct trace estimate must contain an L^2 term of order h^{-1} on the right-hand side, and the proof must be reworked accordingly.
  2. [§3.1.1, Eq. (3.18)] The mass form b(φ_h,q_h) = ∫_Ω φ_h·q_h dx is not well-defined as written. Because V_h = V_{1,h} ⊕ V_{2,h}, a discrete function has two independent components on the overlap region Ω_{Γ,h} of the interface elements (Section 3.1.1), so the integrand φ_h·q_h is not single-valued there. The authors must define b as the sum over i=1,2 of the integrals over Ω_i (or an equivalent unambiguous counterpart). Without this, the generalized eigenvalue problem (3.16) and the claim that T_h is self-adjoint in the L^2 inner product are ambiguous.
  3. [§3.2, Eq. (3.39)] The passage from the infinite cylinder to the truncated domain Ω_{Σ,L} with homogeneous Dirichlet boundary conditions at τ_2 = ±L is justified only by 'thanks to the localization property of the eigenfunction.' The paper proves no exponential decay for edge states of the discontinuous, high-contrast operator, and no L-dependent error bound is provided. Consequently Theorem 4.4 applies only to the truncated eigenvalue problem (3.40)–(3.43), not to the physical point spectrum defined by (2.13)–(2.15); the numerical edge states are therefore not guaranteed to represent the infinite-cylinder problem.
minor comments (4)
  1. [§5.1.1 and §5.2.1] The convergence rates shown in Figures 5.1–5.2 and 5.5 are self-convergence rates between successive mesh sizes, not errors against exact eigenvalues. Without an accurate reference solution, these data confirm asymptotic consistency but not the claimed optimal order against the true solution.
  2. [§3.1.2, Eq. (3.17)] Please clarify the complex-conjugation convention in the form (3.17). The proof of (4.16) uses symmetry of a_h, which requires a Hermitian sesquilinear convention for complex-valued functions; the text calls a_h a bilinear form, which is inconsistent with that usage.
  3. [§4, Theorem 4.3] In the statement of Theorem 4.3, the phrase 'there are m eigenvalues E_h^1, E_h^1, ..., E_h^m' contains a duplicated superscript; it should read E_h^1, E_h^2, ..., E_h^m.
  4. [§5.2.2, Figure 5.7] The caption of Figure 5.7 omits the label '(b)' before the description of the 80th eigenfunction.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: convergence analysis is self-contained and no fitted quantity is recycled as a prediction.

full rationale

The derivation chain is self-contained. The method's convergence claims rest on (i) the Nitsche bilinear form and its coercivity, established in Theorem 3.6 via a trace inequality proved in Appendix A (Lemma 3.4), and (ii) the Babuska-Osborn spectral approximation framework, applied directly in Section 4 with error estimates (4.6)-(4.7) proved from interpolation estimates (3.34)/(3.62) and regularity (4.3). No numerical parameter is fitted and then renamed as a prediction; the convergence tests in Section 5 compare successive mesh refinements and benchmark against a spectral method, and are not used as inputs to the theorems. Prior self-citations ([14] superconvergent recovery, [22] Dirac point and edge-state theory) provide background or physical context, but the convergence proof does not invoke them as premises; the cited facts are also external archival results, not unpublished assertions by this paper's authors. The truncation of the infinite cylinder in Section 3.2 is an assumption justified by localization, not an input-output recycling. A possible defect in Lemma 3.4 (whether (3.22) holds for constant finite-element functions) would be a correctness issue in the proof, not circularity, because it does not make the conclusion equivalent to an input by construction.

Assumptions & free parameters 2 free parameters · 6 assumptions · 0 invented entities

The central h-convergence claim rests on standard periodic spectral theory, a geometric mesh assumption, piecewise H^2 regularity with stable extensions, and uniform ellipticity. The edge-mode claim additionally rests on an unquantified localization and truncation assumption. The two hand-chosen parameters are the Nitsche penalty and the truncation length; neither is fitted to reproduce a target result.

free parameters (2)
  • Nitsche stabilization parameter lambda_hat = unspecified; chosen sufficiently large
    Required to be above a threshold in Theorem 3.6 for coercivity. It is not fitted to output data, but its value is not quantified.
  • Truncation length L in the v2 direction = L=80 in the edge-mode tests
    The infinite cylinder is replaced by [-L,L] with zero boundary conditions; the accuracy of the computed edge modes depends on L being large, but no a priori bound or L-convergence test is provided.
assumptions (6)
  • standard math Floquet-Bloch theory: the spectrum of L_W on R^2 equals the union of band ranges over the Brillouin zone (Eq. 2.10).
    Standard spectral theory for periodic operators, invoked in Section 2.2 to reduce the bulk problem to a torus.
  • domain assumption Assumption 3.1: the interface intersects each interface element boundary exactly twice and each open edge at most once.
    Restricts interface geometry relative to the uniform mesh; used for trace inequalities and interpolation estimates.
  • domain assumption Piecewise H^2 regularity of the source problem, with H^2-stable extension operators (Eq. 4.3 and Eqs. 3.30-3.31).
    Cited to references [4,20]; needed for the O(h^2) interpolation and eigenvalue error estimates in Theorems 4.1 and 4.4.
  • ad hoc to paper The exact edge mode is localized strongly enough that homogeneous Dirichlet conditions at tau2=+/-L introduce negligible error.
    Stated in Section 3.2 as 'thanks to the localization property'; no decay rate or L-dependent error bound is proved.
  • domain assumption Uniform ellipticity of W: min(epsilon^2-gamma^2)>c0>0 (Section 2.1).
    Ensures the continuous and discrete problems are coercive; treated as a physical condition, not proved in the paper.
  • standard math Babuska-Osborn spectral approximation theory and Boffi's spectral pollution theorem (Section 4).
    Background framework used to convert finite element error estimates into eigenvalue and eigenfunction error estimates.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Unfitted Nitsche's method for computing wave modes in topological materials." pith.science (2026). https://pith.science/paper/VB5ODRX4

@misc{pith2026190806585,
  author       = {Pith},
  title        = {Pith review of: Unfitted Nitsche's method for computing wave modes in topological materials},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VB5ODRX4}},
  note         = {Machine review of arXiv:1908.06585}
}
read the original abstract

In this paper, we propose an unfitted Nitsche's method for computing wave modes in topological materials. The proposed method is based on Nitsche's technique to study the performance-enhanced topological materials which have strongly heterogeneous structures (e.g., the refractive index is piecewise constant with high contrasts). For periodic bulk materials, we use Floquet-Bloch theory and solve an eigenvalue problem on a torus with unfitted meshes. For the materials with a line defect, a sufficiently large domain with zero boundary conditions is used to compute the localized eigenfunctions corresponding to the edge modes. The interfaces are handled by Nitsche's method on an unfitted uniform mesh. We prove the proposed methods converge optimally, and present numerical examples to validate the theoretical results and demonstrate the capability of simulating topological materials.

Figures

Figures reproduced from arXiv: 1908.06585 by the authors.

Figure 3.1
Figure 3.1. Triangulation Th on fundamental cell. (a): Triangulation Th; (b): Triangu￾lation T1,h; (c): Triangulation T2,h. The elements of Th can be categorized into two different classes: regular elements and interface elements. An element τ is called an interface element if the interface Γ passes through K. The set of all elements that intersect the interface Γ is denoted by TΓ,h. Then, it is easy to see that TΓ,h =  K ∈ Th… view at source ↗
Figure 3
Figure 3. gives an illustration of [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 3.2
Figure 3.2. Plot of the interface ΓΣ,L with v2 being the x-axis and v1 being the y-axis. Let Tˆ h denote the uniform triangular partition of the computational domain ΩΣ,L. The mesh Tˆ h is generated by firstly dividing ΩΣ,L into 2LN2 sub-rhombuses with mesh size h = kv1k N and splitting each sub-rhombus into two triangles. Similarly, the elements in mesh Tˆ h can be classified as regular elements or interface elements. Let Tˆ i… view at source ↗
Figures from the paper (9 more)
Figure 5.1
Figure 5.1. Figure 5.1: Numerical errors for eigenvalue approximation: (a) [PITH_FULL_IMAGE:figures/full_fig_p019_5_1.png]
Figure 5.2
Figure 5.2. Figure 5.2: Numerical errors for eigenvalue approximation: (a) [PITH_FULL_IMAGE:figures/full_fig_p019_5_2.png]
Figure 5.3
Figure 5.3. Figure 5.3: Dispersion relations when J = 30: (a) Unfitted Nitsche’s Method with γ = 0; (b) Spectral Method with γ = 0; (c) Unfitted Nitsche’s Method with γ = 0.1; (d) Spectral Method with γ = 0.1 -0.01 -0.005 0 0.005 0.01 7.2 7.4 7.6 7.8 8 8.2 8.4 8.6 8.8 (a) -0.01 -0.005 0 0.0…
Figure 5.4
Figure 5.4. Figure 5.4: Comparison of the unfitted Nitsche’s method and the Fourier spectral method [PITH_FULL_IMAGE:figures/full_fig_p020_5_4.png]
Figure 5
Figure 5. Figure 5: a, we show the plot of first 85 eigencurves in term of [PITH_FULL_IMAGE:figures/full_fig_p021_5.png]
Figure 5.5
Figure 5.5. Figure 5.5: Numerical errors for eigenvalue approximation: (a) [PITH_FULL_IMAGE:figures/full_fig_p022_5_5.png]
Figure 5.6
Figure 5.6. Figure 5.6: Plot of the first 85 eigencurves for with [PITH_FULL_IMAGE:figures/full_fig_p022_5_6.png]
Figure 5.7
Figure 5.7. Figure 5.7: Plot of the module of the eigenfunctions: (a) The [PITH_FULL_IMAGE:figures/full_fig_p023_5_7.png]
Figure 5.8
Figure 5.8. Figure 5.8: Plot of the module of the eigenfunctions: (a) The [PITH_FULL_IMAGE:figures/full_fig_p023_5_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

40 extracted references · 40 canonical work pages

  1. [1]

    M J Ablowitz, S D Nixon, and Y Zhu , Conical diffraction in honeycomb lattices, Physical Review A, 79 (2009), p. 053830

  2. [2]

    M. J. Ablowitz and Y. Zhu , Nonlinear waves in shallow honeycomb lattices, SIAM J. Appl. Math., 72 (2012), pp. 240–260

  3. [3]

    Anna v arapu, M

    C. Anna v arapu, M. Hautefeuille, and J. E. Dolbow , A robust Nitsche’s formulation for interface problems, Comput. Methods Appl. Mech. Engrg., 225/228 (2012), pp. 44–54

  4. [4]

    Babuška, The finite element method for elliptic equations with discontinuous coefficients, Computing (Arch

    I. Babuška, The finite element method for elliptic equations with discontinuous coefficients, Computing (Arch. Elektron. Rechnen), 5 (1970), pp. 207–213. 24

  5. [5]

    Babuška and J

    I. Babuška and J. Osborn , Eigenvalue problems, in Handbook of numerical analysis, Vol. II, Handb. Numer. Anal., II, North-Holland, Amsterdam, 1991, pp. 641–787

  6. [6]

    Boffi , Finite element approximation of eigenvalue problems, Acta Numer., 19 (2010), pp

    D. Boffi , Finite element approximation of eigenvalue problems, Acta Numer., 19 (2010), pp. 1–120

  7. [7]

    Burman, S

    E. Burman, S. Claus, P. Hansbo, M. G. Larson, and A. Massing , CutFEM: discretiz- ing geometry and partial differential equations, Internat. J. Numer. Methods Engrg., 104 (2015), pp. 472–501

  8. [8]

    Chen and J

    Z. Chen and J. Zou , Finite element methods and their convergence for elliptic and parabolic interface problems, Numer. Math., 79 (1998), pp. 175–202

Show all 40 references
  1. [9]

    P. G. Ciarlet , The finite element method for elliptic problems, vol. 40 of Classics in Applied Mathematics, Society for Industrial and Applied Mathematics (SIAM), Philadelphia, PA,

  2. [10]

    Guo and X

    H. Guo and X. Yang , Gradient recovery for elliptic interface problem: II. Immersed finite element methods, J. Comput. Phys., 338 (2017), pp. 606–619

  3. [11]

    Body-fitted mesh, Commun.Comput

    ,Gradient recovery for elliptic interface problem: I. Body-fitted mesh, Commun.Comput. Phys., 23 (2018), pp. 1488–1511

  4. [12]

    Nitsche’s method, J

    , Gradient recovery for elliptic interface problem: III. Nitsche’s method, J. Comput. Phys., 356 (2018), pp. 46–63

  5. [13]

    H. Guo, X. Yang, and Z. Zhang , Superconvergence of partially penalized immersed finite element methods, IMA J. Numer. Anal., 38 (2018), pp. 2123–2144

  6. [14]

    H. Guo, X. Yang, and Y. Zhu , Bloch theory-based gradient recovery method for computing topological edge modes in photonic graphene, J. Comput. Phys., 379 (2019), pp. 403–420

  7. [15]

    F D M Haldane and S Raghu , Possible realization of directional optical waveguides in photonic crystals with broken time-reversal symmetry, Physical review letters, 100 (2008), p. 013904

  8. [16]

    Hansbo and P

    A. Hansbo and P. Hansbo , An unfitted finite element method, based on Nitsche’s method, for elliptic interface problems, Comput. Methods Appl. Mech. Engrg., 191 (2002), pp. 5537– 5552

  9. [17]

    Hou and X.-D

    S. Hou and X.-D. Liu , A numerical method for solving variable coefficient elliptic equation with interfaces, J. Comput. Phys., 202 (2005), pp. 411–445

  10. [18]

    S. Hou, P. Song, L. W ang, and H. Zhao , A weak formulation for solving elliptic interface problems without body fitted grid, J. Comput. Phys., 249 (2013), pp. 80–95

  11. [19]

    P. Hu, L. Hong, and Y. Zhu , Linear and nonlinear electromagnetic waves in modulated honeycomb media, Studies in Applied Mathematics, 144 (2020), pp. 18–45

  12. [20]

    R. B. Kellogg , On the Poisson equation with intersecting interfaces, Applicable Anal., 4 (1974/75), pp. 101–129. Collection of articles dedicated to Nikolai Ivanovich Muskhelishvili

  13. [21]

    A. B. Khanikaev, S. H. Mousa vi, W.-K. Tse, M. Kargarian, A. H. MacDonald, and G. Shvets, Photonic topological insulators, Nature materials, 12 (2013), pp. 233–239

  14. [22]

    J. P. Lee-Thorp, M. I. Weinstein, and Y. Zhu , Elliptic operators with honeycomb symme- try: Dirac points, edge states and applications to photonic graphene, Arch. Ration. Mech. Anal., 232 (2019), pp. 1–63

  15. [23]

    R. J. LeVeque and Z. Li , The immersed interface method for elliptic equations with discon- tinuous coefficients and singular sources, SIAM J. Numer. Anal., 31 (1994), pp. 1019–1044

  16. [24]

    Li, The immersed interface method using a finite element formulation, Appl

    Z. Li, The immersed interface method using a finite element formulation, Appl. Numer. Math., 27 (1998), pp. 253–267

  17. [25]

    Li and K

    Z. Li and K. Ito , The immersed interface method, vol. 33 of Frontiers in Applied Mathe- matics, Society for Industrial and Applied Mathematics (SIAM), Philadelphia, PA, 2006. Numerical solutions of PDEs involving interfaces and irregular domains

  18. [26]

    Z. Li, T. Lin, and X. Wu , New Cartesian grid methods for interface problems using the finite element formulation, Numer. Math., 96 (2003), pp. 61–98

  19. [27]

    T. Lin, Y. Lin, and X. Zhang , Partially penalized immersed finite element methods for elliptic interface problems, SIAM J. Numer. Anal., 53 (2015), pp. 1121–1144

  20. [28]

    L Lu, J D Joannopoulos, and M Soljačić , Topological photonics, Nature Photonics, 8 (2014), pp. 821–829

  21. [29]

    S H Mousa vi, A B Khanikaev, and Z W ang , Topologically protected elastic waves in phononic metamaterials, Nature communications, 6 (2015)

  22. [30]

    Nitsche, über ein Variationsprinzip zur Lösung von Dirichlet-Problemen bei Verwendung von Teilräumen, die keinen Randbedingungen unterworfen sind, Abh

    J. Nitsche, über ein Variationsprinzip zur Lösung von Dirichlet-Problemen bei Verwendung von Teilräumen, die keinen Randbedingungen unterworfen sind, Abh. Math. Sem. Univ. Hamburg, 36 (1971), pp. 9–15. Collection of articles dedicated to Lothar Collatz on his sixtieth birthday

  23. [31]

    C. S. Peskin, Numerical analysis of blood flow in the heart, J. Computational Phys., 25 (1977), pp. 220–252. 25

  24. [32]

    , The immersed boundary method, Acta Numer., 11 (2002), pp. 479–517

  25. [33]

    Y Plotnik, M C Rechtsman, D Song, M Heinrich, J M Zeuner, S Nolte, Y Lumer, N Malkov a, J Xu, A Szameit, Z Chen, and M Segev , Observation of unconventional edge states in ‘photonic graphene’, Nature materials, 13 (2014), pp. 57–62

  26. [34]

    M C Rechtsman, J M Zeuner, Y Plotnik, Y Lumer, D Podolsky, F Dreisow, S Nolte, M Segev, and A Szameit , Photonic floquet topological insulators, Nature, 496 (2013), pp. 196–200

  27. [35]

    Maksim Skorobogatiy and Jianke Yang , Fundamentals of photonic crystal guiding, Cam- bridge University Press, 2009

  28. [36]

    R Süsstrunk and S D Huber , Observation of phononic helical edge states in a mechanical ’topological insulator’, Science, 349 (2015), pp. 47–50

  29. [37]

    Trefethen , Spectral methods in MATLAB, SIAM, 2000

    Lloyd N. Trefethen , Spectral methods in MATLAB, SIAM, 2000

  30. [38]

    M Xiao, G Ma, Z Yang, P Sheng, Z Q Zhang, and C T Chan , Geometric phase and band inversion in periodic acoustic systems, Nature Physics, 11 (2015), pp. 240–244

  31. [39]

    Xie and Y

    P. Xie and Y. Zhu , Wave packet dynamics in slowly modulated photonic graphene, Journal of Differential Equations, 267 (2019), pp. 5775–5808. 26

  32. [2002]

    Reprint of the 1978 original [North-Holland, Amsterdam; MR0520174 (58 #25001)]

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.