Pith. sign in

REVIEW 6 minor 24 references

Stability and Convergence of Spectral Mixed Discontinuous Galerkin Methods for 3D Linear Elasticity on Anisotropic Geometric Meshes

T0 review · 0 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read This paper uses numerical experiments to claim that a spectral mixed discontinuous Galerkin method for 3D linear elasticity on anisotropic geometric meshes is stable up to the incompressible limit and converges exponentially in a natural…

desk verdict A sound, honest computational validation of the authors' own theory; the SVD inf-sup tool is genuinely useful, and the soft spots are reproducibility and a narrow set of manufactured tests. read the letter →

arxiv 1908.04647 v1 pith:W7LP5D26 submitted 2019-08-13 math.NA cs.NA

Thomas P. Wihler , Marcel Wirz This is my paper · ORCID
classification math.NAcs.NA MSC 65N3065N3565N12
keywords mixeddiscontinuousGalerkinmethodlinearelasticitypolyhedraldomainsgeometricedgemeshesinf-supstabilityexponentialconvergenceincompressiblelimitspectralmethods
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper makes the case, through extensive numerical experiments, that a spectral mixed discontinuous Galerkin method can solve the three-dimensional Lamé system of linear elasticity in polyhedral domains — shapes with edges and corners where solutions develop singularities — with no loss of stability. It focuses on anisotropic geometric meshes that refine toward edges, corners, and corner-edges, and on the behavior of discrete inf-sup constants as the Poisson ratio $\nu$ approaches $1/2$, the incompressible Stokes limit. The authors claim two things: the discrete stability constants remain robust under mesh refinement and in the incompressible limit, with a polynomial-degree dependence far milder than the proved $k^{-3/2}$ bound, and the method converges exponentially in a natural DG norm for solutions with edge, corner, and corner-edge singularities under analytic regularity assumptions. If these claims hold, the method is a practical high-order solver for 3D elasticity and Stokes problems, including nearly incompressible materials.

What carries the argument

The load-bearing mechanism is the discrete inf-sup condition (13): the assumption that the inf-sup constant of $B_h$ over the discrete spaces $V_h \times Q_h$ is bounded below by $\kappa k^{-\rho}$ with $\kappa$ and $\rho$ independent of the polynomial degree $k$, the number of refinement layers $\ell$, and the aspect ratio of anisotropic elements. This condition upgrades coercivity — which degenerates as $(1-2\nu) \to 0$ — into full inf-sup stability of the mixed form $a_h$ even at $\nu=1/2$, and it underlies the exponential convergence estimate (28). Computationally, the paper evaluates these constants as the smallest positive singular values of scaled system matrices, and builds the meshes from canonical geometric patches toward edges, corners, and corner-edges.

What would settle it

Recompute the discrete inf-sup constants for a geometric mesh family with a different refinement ratio or with polynomial degrees beyond $k=10$, or test the corner-edge example on a Fichera domain with traction boundary conditions; if the constants decay to zero with degree or the exponential error decay flattens to algebraic as $\nu \to 1/2$, the central claim is refuted.

Watch

Extended reading notes

Core claim

The central discovery is computational but with theoretical teeth: on anisotropic geometric edge meshes, the discrete inf-sup constant $\gamma_B$ of the bilinear form $B_h$ stabilizes after a few geometric refinement layers and shows a dependence on the polynomial degree $k$ that is considerably more optimistic than the theoretical $k^{-3/2}$ bound; likewise, the inf-sup constant $\gamma_a$ of the mixed form $a_h$ remains bounded away from zero uniformly as $\nu \to 1/2$, confirming that the DG scheme does not degenerate in the incompressible limit. Moreover, for manufactured solutions with edge, corner, and corner-edge singularities, and with the polynomial degree grown proportionally to the refinement level, the DG error in the norm (12) decays as $\exp(-b N^{1/5})$ or $\exp(-b N^{1/4})$, exhibiting the exponential convergence predicted by the theory. These experiments together support the conclusion that the spectral mixed DG method (8) is stable and exponentially convergent on geometric edge meshes for 3D linear elasticity.

Load-bearing premise

The entire argument rests on an assumed stability bound, the discrete inf-sup condition (13), which says the mixed system cannot become singular as the mesh is refined or the polynomial degree grows; the paper verifies this bound numerically but does not prove it beyond the exponent $3/2$ already established in the literature.

Editorial extensions

If this is right

  • The spectral mixed DG method (8) can be used as a reliable solver for 3D linear elasticity in polyhedral domains, including near-incompressible and exactly incompressible materials, without special locking treatment.
  • Exponential convergence in the DG norm is achievable for solutions with edge, corner, and corner-edge singularities when geometric refinement is paired with a polynomial degree that grows with the number of layers.
  • The measured $k$-dependence of the inf-sup constants indicates that the theoretical worst-case decay $k^{-3/2}$ (and $k^{-3}$ for the mixed form) is not sharp, so sharper stability bounds should be attainable.
  • The augmented Lagrange-multiplier formulation (16) offers a practical way to enforce the zero-mean pressure constraint and to compute inf-sup constants using standard local basis functions.

Reading between the lines

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

  • If the observed mild $k$-dependence of the inf-sup constants is proved rather than merely observed, the exponential convergence estimates for hp-DG methods on geometric meshes could be sharpened, yielding better cost predictions for high-accuracy 3D elasticity.
  • The same SVD-based constant computation could serve as a numerical pre-screen for stability of hp-adaptive DG variants (variable or anisotropic polynomial degrees), which the paper mentions as a future extension but does not analyze.
  • Since the method tolerates $\nu \to 1/2$ without deterioration, coupling it with the iterative Newton-DG approach mentioned in the conclusions could yield robust solvers for nonlinear hyperelasticity near the incompressible limit.
  • The canonical-patch experiments use manufactured solutions; a natural next test, not performed here, would be on a Fichera domain with traction boundary conditions, where the combination of corner and edge singularities is genuinely three-dimensional and the analytic regularity assumptions are harder to satisfy.
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

0 major / 6 minor

Summary. This paper presents a computational study of a spectral mixed discontinuous Galerkin method for the Lamé system of linear elasticity in mixed form on anisotropic geometric edge meshes in three-dimensional polyhedral domains. The authors recall the discrete inf-sup framework from their earlier work [24] and the assumption (13) on the form B_h, which was proved with rho = 3/2 in [19]. They then derive SVD-based characterizations (Propositions 4.3 and 4.4) for the discrete inf-sup constants of B_h and of the full mixed form a_h in the incompressible limit nu = 1/2, together with an augmented formulation (16) that avoids the zero-mean constraint on the pressure. Numerical experiments measure these constants on edge, corner, corner-edge, and Fichera-type geometric refinements, and show robustness with respect to mesh refinement and a k-dependence much milder than the theoretical bounds k^{-3/2} and k^{-3}. A second set of experiments with manufactured edge, corner, and corner-edge singularities and with smooth incompressible examples exhibits near-exponential convergence in the DG norm and robustness as nu approaches 1/2.

Significance. The paper's main contribution is a clean SVD-based procedure for evaluating discrete inf-sup constants of mixed DG forms, together with the first systematic numerical evidence that the inf-sup constants of the spectral mixed DG method are robust on anisotropic geometric edge meshes with respect to refinement and Poisson ratio, and that exponential convergence is achieved for solutions with edge and corner singularities. The derivations in Section 4 are mathematically sound given the inf-sup hypothesis (13), and the experimental design tests the theory rather than fitting constants. The absence of shipped code or data is a reproducibility limitation, but the algorithmic description is sufficiently detailed to be reimplemented in deal.II.

minor comments (6)
  1. [Section 5.1] The text states that the singular examples use nu in {1/8, 1/2, 3/8}, but the figures show nu = 1/8, 1/4, 3/8 and the formula p = -div u/(1-2 nu) is undefined at nu = 1/2; the set should be corrected.
  2. [Equation (28) and Figures 7-8] The notation 'exp(-b 5 sqrt(N))' in (28) is ambiguous; it should be typeset as exp(-b N^{1/5}) or exp(-b times the fifth root of N), and the relationship between this bound and the N^{1/4} axes used for the edge and corner cases in Figure 7 should be stated explicitly.
  3. [Section 5.1] The manufactured solutions do not satisfy the homogeneous Dirichlet condition u = 0 on the whole boundary; the paper explains that only u dot n = 0 is enforced and that the remaining boundary data are imposed through flux terms, but the precise modification of the right-hand side in (10) is not given, which would improve reproducibility.
  4. [Section 5.2, Example 5.2] The reference solution is computed only at k = l = 5; since the reported errors approach the level of this reference, a comparison against a finer reference or an estimate of the reference discretization error would make the claimed exponential convergence more conclusive.
  5. [Section 4.2, Eq. (19)] Equation (19) is not readable as typeset; the relation between M and N should be written clearly, for example M = 3((k+1)/k)^3 N.
  6. [General] No code or data files are provided; for a purely computational paper, making the deal.II scripts and mesh-generation routines available would substantially aid reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the numerical study tests externally proved theory rather than deriving predictions from fitted inputs.

full rationale

The paper is a computational validation of prior theory. The discrete inf-sup assumption (13) is taken from [19], where it was proved with rho=3/2, and the numerical SVD computations of gamma_B and gamma_a in Section 4.4 are independent probes of that condition, not fitted parameters later relabeled as predictions. Theorem 3.2 and the exponential convergence bound (28) are cited from the authors' earlier paper [24], but they are stated with explicit assumptions (discrete inf-sup condition, analytic regularity in A_{-1-beta} x A_{-beta}) and are not re-derived by fitting the experiments; the manufactured solutions in Section 5.1 are chosen to satisfy those assumptions and the observed errors confirm the bound. The equivalence reformulation in Section 4.1 is proven within the paper (Lemma 4.1, Proposition 4.2) and uses (13) only as an input hypothesis. No equation in the derivation chain reduces to its own output, and no fitted constant is renamed as a prediction. Self-citations occur, but they are not load-bearing in a circular sense; [24] provides an independent theorem rather than an unverified assertion unique to this paper.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central mathematical claims rest on the inf-sup Assumption (13) and on analytic regularity of the solution. Two hand-chosen numerical parameters (gamma = 10, sigma = 1/2) enter all experiments. No fitted parameters are used to obtain the reported convergence or stability behavior. The auxiliary variable ~r in the augmented system (16) is a Lagrange multiplier used for computational convenience, not a new physical or mathematical entity.

free parameters (2)
  • penalty parameter gamma = 10
    Chosen in all experiments (Section 4.4); the stability theory requires gamma sufficiently large and the numerical results depend on this choice.
  • mesh grading factor sigma = 1/2
    Chosen for the geometric refinements (Section 4.4); affects the element aspect ratios and the value of the inf-sup constants.
assumptions (4)
  • domain assumption Discrete inf-sup condition (13): there exist kappa > 0 and rho >= 0 independent of k, refinement level, and aspect ratio such that gamma_B >= kappa k^{-rho}.
    Used in Theorem 3.2 (Section 3.4); proved with rho = 3/2 in [19], numerically investigated in Section 4.4.1.
  • domain assumption Analytic regularity of the solution: (u,p) in A_{-1-beta}(Omega)^3 times A_{-beta}(Omega), the countably normed analytic spaces.
    Assumed in Section 5 based on the regularity theory of [6]; needed to apply [24, Theorem 6.2] and obtain bound (28).
  • domain assumption Weighted Sobolev regularity of the continuous solution (Proposition 2.1) in M^m_{-1-beta} spaces.
    Invoked from [6] (Section 2.2) to motivate the mesh grading; standard in the hp-FEM literature.
  • standard math Coercivity of A_h and well-posedness of the continuous mixed problem, including the standard Stokes inf-sup property for nu = 1/2.
    Assumed as background (Section 2.1), citing [5,8].

how reviews work

0 comments
Cite this review

Pith. "Pith review of Stability and Convergence of Spectral Mixed Discontinuous Galerkin Methods for 3D Linear Elasticity on Anisotropic Geometric Meshes." pith.science (2026). https://pith.science/paper/W7LP5D26

@misc{pith2026190804647,
  author       = {Pith},
  title        = {Pith review of: Stability and Convergence of Spectral Mixed Discontinuous Galerkin Methods for 3D Linear Elasticity on Anisotropic Geometric Meshes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/W7LP5D26}},
  note         = {Machine review of arXiv:1908.04647}
}
abstract

We consider spectral mixed discontinuous Galerkin finite element discretizations of the Lam\'e system of linear elasticity in polyhedral domains in $\mathbb{R}^3$. In order to resolve possible corner, edge, and corner-edge singularities, anisotropic geometric edge meshes consisting of hexahedral elements are applied. We perform a computational study on the discrete inf-sup stability of these methods, and especially focus on the robustness with respect to the Poisson ratio close to the incompressible limit (i.e. the Stokes system). Furthermore, under certain realistic assumptions (for analytic data) on the regularity of the exact solution, we illustrate numerically that the proposed mixed DG schemes converge exponentially in a natural DG norm.

Figures

Figures reproduced from arXiv: 1908.04647 by the authors.

Figure 1
Figure 1. Canonical geometric refinements (patches) towards edges, corners, and corner-edges (bottom) to be built into the macro mesh T 0 (top) by means of suitable affine transformations. as well as the full norm kvk 2 Mm β (Ω) = Pm k=0 kvk 2 Mk β (Ω); here, the operator D α denotes the partial derivative in the local coordinate directions corresponding to the multi-index α. The space Mm β (Ω) is the weighted Sobolev space o… view at source ↗
Figure 2
Figure 2. We clearly observe that the values of γB stabilize after some initial refinement steps, thereby underlining the robustness of the inf-sup constant of Bh with respect to the anisotropic geometric refinements. The asymptotic values are visualized in [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗
Figure 2
Figure 2. Inf-sup constant γB of the form Bh in case of geometrically refined edge, corner, corner-edge patches, as well as for a Fichera corner refinement, for different approximation degrees k. 2 4 6 8 10 k 0.2 0.22 0.24 0.26 0.28 0.3 0.32 0.34 B Edge Corner Corner-edge Fichera corner 2 4 6 8 10 k 10-1 B Edge Corner Corner-edge Fichera corner [PITH_FULL_IMAGE:figures/full_fig_p012_2.png] view at source ↗
Figures from the paper (9 more)
Figure 3
Figure 3. Figure 3: Stabilized values from [PITH_FULL_IMAGE:figures/full_fig_p012_3.png]
Figure 4
Figure 4. Figure 4: Inf-sup constant γa of the form ah in case of geometric edge, corner, and corner-edge refinements, with different approximation degrees k and ν = 1/2. We focus on the canonical geometric edge, corner, and corner-edge refinements from Section 3.1. As before, we first fi…
Figure 5
Figure 5. Figure 5: Stabilized values from [PITH_FULL_IMAGE:figures/full_fig_p014_5.png]
Figure 6
Figure 6. Figure 6: The mesh and the approximation degree during the first 4 refinement steps for the approximation of the solution with a corner-edge singularity. 2 4 6 8 10 12 14 16 10−3 10−2 10−1 100 101 N 1/4 log(DG error) Solution with an edge singularity ν = 1/8 ν = 1/4 ν = 3/8 2 4 …
Figure 7
Figure 7. Figure 7: Performance of the DGFEM for the solutions with an edge singularity (left) and a corner singularity (right). singularities, and simultaneously increase the approximation degree by one in each refinement step such that k ∼ ℓ, where ℓ is the number of layers. Since the s…
Figure 8
Figure 8. Figure 8: Performance of the DGFEM for the solution with a corner-edge singularity. 10 15 20 25 30 35 40 45 50 55 60 10−8 10−7 10−6 10−5 10−4 10−3 10−2 10−1 100 N 1/3 log(DG error) Divergence free solution ν = 0.49 ν = 0.499 ν = 0.4999 ν = 0.49999 ν = 0.5 [PITH_FULL_IMAGE:figur…
Figure 9
Figure 9. Figure 9: Example 5.1: Performance of the spectral DGFEM for different values of ν. For this example, we use a fixed uniform mesh consisting of 64 elements, and simply vary the uniform polynomial degree k. For this setup, since the solution (u, p) is analytic, we expect exponent…
Figure 10
Figure 10. Figure 10: The mesh and the approximation degree k for the first five refinement steps for the approximation of the reference solution. 1 2 3 4 5 6 7 8 9 10 10−3 10−2 10−1 100 101 102 Reference solution N 1/5 log(DG error) ν=0.49 ν=0.499 ν=0.4999 ν=0.49999 ν=0.5 [PITH_FULL_IMAG…
Figure 11
Figure 11. Figure 11: Example 5.2: Performance of the spectral DGFEM for different val￾ues of ν. we have validated our theoretical results from [24] for various canonical reference situations. In particular, we have performed a computational study on the inf-sup stability and the exponenti…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

24 extracted references · 24 canonical work pages

  1. [24]

    T. P. Wihler and M. Wirz, Mixed hp-discontinuous Galerkin FEM for linear elasticity and Stokes flow in three dimensions, Mathematical Models and Methods in Applied Sciences 22 (2012), no. 8, 1250016

    work page 2012

  2. [19]

    , Mixed hp-DGFEM for incompressible flows. II. Geometric edge meshes , IMA J. Numer. Anal. 24 (2004), no. 2, 273–308

    work page 2004

  3. [6]

    Dauge, M

    M. Dauge, M. Costabel, and S. Nicaise, Analytic regularity for linear elliptic systems in polygon s and polyhedra, Mathematical Models and Methods in Applied Sciences 22 (2012), no. 8, 1250015

    work page 2012

  4. [1]

    D. N. Arnold, F. Brezzi, B. Cockburn, and L. D. Marini, Unified analysis of discontinuous Galerkin methods for elliptic problems , SIAM J. Numer. Anal. 39 (2001), 1749–1779

    work page 2001

  5. [2]

    Babuˇ ska and B

    I. Babuˇ ska and B. Q. Guo, Approximation properties of the h-p version of the finite element method , Comput. Methods Appl. Mech. Engrg. 133 (1996), no. 3-4, 319–346

    work page 1996

  6. [3]

    Bangerth, R

    W. Bangerth, R. Hartmann, and G. Kanschat, deal.II – a general purpose object oriented finite element li brary, ACM Trans. Math. Softw. 33 (2007), no. 4, 24/1–24/27

    work page 2007

  7. [4]

    Bangerth, T

    W. Bangerth, T. Heister, G. Kanschat, et al., deal.II differential equations analysis library, technical refere nce, http://www.dealii.org

    work page

  8. [5]

    Brezzi and M

    F. Brezzi and M. Fortin, Mixed and hybrid finite element methods , Springer Series in Computational Mathe- matics, vol. 15, Springer, New York, 1991

    work page 1991

Show all 24 references
  1. [7]

    E. H. Georgoulis, E. Hall, and P. Houston, Discontinuous Galerkin methods on hp-anisotropic meshes. I. A priori error analysis , Int. J. Comput. Sci. Math. 1 (2007), no. 2-4, 221–244

    2007

  2. [8]

    Girault and P

    V. Girault and P. A. Raviart, Finite element methods for Navier–Stokes equations , Springer, New York, 1986. 18 T. P. WIHLER AND M. WIRZ

    1986

  3. [9]

    B. Q. Guo, The h-p version of the finite element method for solving boundary val ue problems in polyhedral domains, Boundary Value Problems and Integral Equations in Nonsmoo th Domains, Lecture Notes in Pure and Applied Mathematics, vol. 167, Dekker, New York, 1995, p p. 101–120

    1995

  4. [10]

    B. Q. Guo and I. Babuˇ ska, Regularity of the solutions for elliptic problems on nonsmo oth domains in R3. I. Countably normed spaces on polyhedral domains , Proc. Roy. Soc. Edinburgh Sect. A 127 (1997), no. 1, 77–126

    1997

  5. [11]

    , Regularity of the solutions for elliptic problems on nonsmo oth domains in R3. II. Regularity in neigh- bourhoods of edges , Proc. Roy. Soc. Edinburgh Sect. A 127 (1997), no. 3, 517–545

    1997

  6. [12]

    Houston, D

    P. Houston, D. Sch¨ otzau, and T. P. Wihler, An hp-adaptive mixed discontinuous Galerkin FEM for nearly incompressible linear elasticity , Comput. Methods Appl. Mech. Engrg. 195 (2006), no. 25-28, 3224–3246

    2006

  7. [13]

    Houston and T

    P. Houston and T. P. Wihler, An hp-adaptive Newton-discontinuous-Galerkin finite element a pproach for semilinear elliptic boundary value problems , Math. Comp. 87 (2018), no. 314, 2641–2674

    2018

  8. [14]

    Maday and C

    Y. Maday and C. Bernardi, Uniform inf–sup conditions for the spectral discretizatio n of the Stokes problem , Mathematical Models and Methods in Applied Sciences 09 (1999), no. 03, 395–414

    1999

  9. [15]

    Maday and C

    Y. Maday and C. Marcati, Regularity and hp discontinuous Galerkin finite element approximation of lin ear elliptic eigenvalue problems with singular potentials , Math. Models Methods Appl. Sci. 29 (2019), no. 8, 1585– 1617

    2019

  10. [16]

    Maz’ya and J

    V. Maz’ya and J. Rossmann, Elliptic equations in polyhedral domains , Mathematical Surveys and Monographs, vol. 162, American Mathematical Society, Providence, RI, 2 010

  11. [17]

    A. L. Mazzucato and V. Nistor, Well-posedness and regularity for the elasticity equation with mixed boundary conditions on polyhedral domains and domains with cracks , Arch. Ration. Mech. Anal. 195 (2010), no. 1, 25–73

    2010

  12. [18]

    Sch¨ otzau, C

    D. Sch¨ otzau, C. Schwab, and A. Toselli, Mixed hp-DGFEM for incompressible flows , SIAM J. Numer. Anal. 40 (2003), 2171–2194

    2003

  13. [20]

    Sch¨ otzau, C

    D. Sch¨ otzau, C. Schwab, and T. P. Wihler, hp-dGFEM for second-order elliptic problems in polyhedra I: Stability on geometric meshes , SIAM J. Numer. Anal. 51 (2013), no. 3, 1610–1633. MR 3061472

    2013

  14. [21]

    , hp-dGFEM for second order elliptic problems in polyhedra II: E xponential convergence , SIAM J. Numer. Anal. 51 (2013), no. 4, 2005–2035. MR 3073653

    2013

  15. [22]

    , hp-DGFEM for second-order mixed elliptic problems in polyhed ra, Math. Comp. 85 (2016), no. 299, 1051–1083

    2016

  16. [23]

    Sch¨ otzau and T

    D. Sch¨ otzau and T. P. Wihler, Exponential convergence of mixed hp-DGFEM for Stokes flow in polygons , Numer. Math. 96 (2003), 339–361

    2003

Pith tools

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