Pith. sign in

REVIEW 2 major objections 3 minor 46 references

Equivalence of local-and global-best approximations, a simple stable local commuting projector, and optimal $hp$ approximation estimates in $H(\mathrm{div})$

T0 review · 2 major / 3 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 Global and local best-approximation errors in H(div) are equivalent up to a constant, yielding hp-optimal convergence rates under minimal regularity.

desk verdict A strong paper that constructs the first local, commuting, minimal-regularity H(div) projector with p-optimal approximation; the only soft spot is the too-terse import of a Costabel–McIntosh bound in Lemma A.1. read the letter →

arxiv 1908.08158 v2 pith:NINIP3Q2 submitted 2019-08-22 math.NA cs.NA

Alexandre Ern , Thirupathi Gudi , Iain Smears , Martin Vohralík This is my paper · ORCID
classification math.NAcs.NA MSC 65N3065N1541A10
keywords H(div)approximationRaviart-Thomas-Nédélecelementslocal-bestglobal-bestcommutingprojectorhpestimatesfluxequilibrationminimalregularity
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 shows that, for an arbitrary vector field in H(div), the error of the best global approximation by H(div)-conforming piecewise polynomial Raviart-Thomas-Nédélec elements of degree p, with the divergence and boundary fluxes constrained, is comparable to the sum of independent elementwise best approximation errors with no such constraints. The comparison constant depends only on the space dimension, the shape regularity of the mesh, and the polynomial degree. The proof constructs a simple, local projector that commutes with the divergence and is stable in L2 up to an hp data-oscillation term, built by gluing elementwise local-best approximations through equilibrated flux reconstructions on vertex patches. From this equivalence, the paper derives hp-optimal approximation rates $h^{{min(s,p+1)}}$/(p+1)^s requiring only elementwise H^s regularity, with no logarithmic factors in p. It then applies these bounds to mixed and least-squares mixed finite element methods to obtain optimal a priori error estimates.

What carries the argument

The load-bearing object is the projector P_T^p defined in Definition 3.1: on each element, τ_T is the divergence-constrained Raviart-Thomas-Nédélec best approximation of v; over each vertex patch ω_a, an equilibrated flux σ_a is chosen by a constrained minimization (3.2); summing σ_a over all vertices yields an H(div)-conforming piecewise polynomial that commutes with the divergence. The argument that converts the elementwise minimizers into a global H(div) function is the patchwise flux equilibration bound of Lemma 4.4, and the p-robust passage from constrained to unconstrained local errors is Lemma A.1, which imports the result [16, Corollary 3.4].

What would settle it

On a fixed shape-regular simplex, compute the ratio between the divergence-constrained local-best error and the unconstrained one from Lemma A.1 for p=1 through 20 and for a collection of smooth fields; if the ratios grow without bound in p, the p-robustness the final theorem needs is false.

Watch

Extended reading notes

Core claim

The central claim is Theorem 3.3: for any v in H0,Γ_N(div,Ω), the global-best approximation error E_{T,p}(v) defined in (3.10), with divergence fixed to Π_T^p(∇·v) and normal flux constrained on Γ_N, satisfies [E_{T,p}(v)]^2 ≤ C Σ_K [e_{K,p}(v)]^2 ≤ C [E_{T,p}(v)]^2, where e_{K,p}(v) is the unconstrained elementwise best error (3.11). The upper bound is obtained by taking the global competitor to be a new projector P_T^p(v) whose per-element error is controlled by local-best errors over the element and its neighbours; the lower bound is immediate from the definitions. A second, one-sided bound with p replaced by p−1 and a constant independent of p is proved using an unbalanced patchwise construction; combining the two bounds gives the hp-optimal estimate (3.13) under only elementwise H^s regularity.

Load-bearing premise

The proof rests on the imported fact that, on one triangle or tetrahedron, forcing the divergence to a polynomial projection costs no more than a factor depending only on the simplex shape, not on the polynomial degree; if that factor secretly grows with degree, the p-optimal rates of Theorem 3.6 collapse.

Editorial extensions

If this is right

  • The equivalence makes global H(div) approximation error bounds reduce to elementwise computations, so the global error can be bounded by summing independent local problems on each element.
  • The projector P_T^p is a commuting local projection valid for all of H0,Γ_N(div,Ω), so it supplies a Fortin operator for mixed methods without any extra smoothness assumption.
  • For vector fields with only elementwise H^s regularity, the approximation rate is h^{min(s,p+1)}/(p+1)^s in the weighted H(div) norm, with no logarithmic factors in p and no global regularity requirements.
  • In mixed finite element methods, the flux error is exactly a constrained global-best error, so the new bounds give fully optimal hp a priori error estimates; least-squares mixed methods inherit the same rate for the flux plus a best-approximation term for the gradient.

Reading between the lines

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

  • The same equilibrated-flux construction should carry over to H(curl) and to tensor-valued H(div) settings, yielding analogous local commuting projectors; the paper does not treat these cases.
  • The one-sided p-robust bound suggests a concrete algorithmic prescription for hp-adaptive codes: compute local minimizers one degree lower than the target space to keep all constants independent of p.
  • A direct numerical test of Lemma A.1 on anisotropic elements would show whether the shape-regularity parameter alone controls the constant, or whether a more detailed geometric quantity enters; the paper assumes the former.
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

2 major / 3 minor

Summary. The paper establishes that, for arbitrary vector fields in H0,Gamma_N(div,Omega), the error of the global-best approximation by Raviart-Thomas-Nedelec finite element functions satisfying divergence and boundary constraints is equivalent, up to a generic constant, to the sum of elementwise local-best approximation errors without those constraints. The proof is constructive: the authors define a locally defined, commuting projector (Definition 3.1) built from elementwise local-best approximations and patchwise equilibrated flux reconstructions, prove its stability and approximation properties (Theorem 3.2), and derive the local-global equivalence (Theorem 3.3). They then prove an unbalanced, polynomial-degree-robust variant (Proposition 5.1) which, combined with the equivalence result, yields hp-optimal approximation rates with no logarithmic factors under only elementwise H^s regularity (Theorem 3.6). The last section applies these results to mixed and least-squares finite element methods for a model diffusion problem. The proofs are detailed and organized; the main external input is a locally constrained-unconstrained equivalence on a simplex (Lemma A.1) imported from Costabel-McIntosh [16, Corollary 3.4].

Significance. If the results hold, they settle a long-standing gap in the literature by providing a stable, local, commuting projector under minimal H(div) regularity, together with fully optimal hp-approximation estimates that require only elementwise minimal Sobolev regularity. The local-global equivalence in H(div) extends earlier scalar results of Veeser and others, and the applications to mixed and least-squares methods are immediate and useful. The construction based on equilibrated flux reconstructions is simple and elegant, and the proof structure makes the dependence of each constant transparent. The main achievement is the p-robust removal of the divergence constraint in Lemma A.1; this is also the only point where the manuscript relies on a deep external result without stating its precise content.

major comments (2)
  1. [Appendix A, Eq. (A.2); used in Lemma 4.6 and Proposition 5.1] The proof of Lemma A.1, which is the load-bearing p-robust constrained-unconstrained equivalence on a simplex, rests entirely on the sentence 'It follows from [16, Corollary 3.4] that there exists v_K in RTN_p(K) ...' without quoting the corollary or explaining how the constant in (A.2) is independent of the polynomial degree. This lemma is used in the proof of Lemma 4.6 (final step of the proof of (4.6)) and in the proof of Proposition 5.1 (via Lemma 5.3), and it is exactly the mechanism that removes the p-logarithm in Theorem 3.6. If the constant in [16, Cor. 3.4] had any hidden dependence on p, then the bounds (A.1), (3.12), and (5.1) would inherit that dependence and the claimed rate (3.13) would lose its p-optimality. The authors should quote the precise statement of [16, Corollary 3.4], verify that it applies to the spaces RTN_p(K) rather than to the spaces of the de Rham complex treated there, and show in detail that the resulting constant depends only on d and kappa_K, or else provide a self-contained proof of Lemma A.1.
  2. [Section 5.2, Step 2] In the proof of Theorem 3.6 for the case p > s, the bound [e_{K,p-1}(v)]^2 <= C { [h_K^s/p^s ||v||_{H^s(K)}]^2 + delta_{s<1} [h_K/p ||div v||_K]^2 } is stated with a constant depending only on s, d, and kappa_T. Since this is the step that achieves the fully p-robust constant in (3.13), the authors should make explicit which approximation result supplies a constant independent of p for the (p-1)-degree local best approximation; the cited elementary hp-bounds usually contain constants that may depend on p, and this point is central to the advertised p-optimality.
minor comments (3)
  1. [Section 2.1] The notation L2(Omega) is introduced as L2(Omega; R^d), but scalar-valued L2 spaces also appear throughout; this overloading is clear from context but could be noted explicitly.
  2. [Proof of Lemma 4.6] The jump notation /llbracket v_T /rrbracket is introduced within the proof; defining it in Section 2 alongside the face notation would improve readability.
  3. [Section 5.2, Step 1] The bound (5.8) cites [2, Lemma 4.1] for the hp-approximation estimate; since the constant is claimed to depend only on s, d, and kappa_T, a short justification of the elemental version of that lemma would be helpful.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the core projector construction, local-global equivalence, and hp estimates are derived from independent external stability and lifting theorems, not from the target bounds they establish.

full rationale

The paper's main chain is: define a patchwise equilibrated-flux projector from local best approximations (Definition 3.1); prove its approximation property using Lemma 4.4 (patchwise flux equilibration stability from [8, Theorem 7] and [30, Corollaries 3.3, 3.6, 3.8]) and Lemma A.1 (local constrained-unconstrained equivalence from [16, Corollary 3.4]); then Theorem 3.3 follows by taking the projector as a candidate in the global minimization; Theorem 3.6 combines Theorem 3.3 with Proposition 5.1. None of these steps presupposes the desired equivalence or rate. Lemma A.1 is the only place where the divergence constraint is removed on a simplex, and it imports an external, published theorem from Costabel-McIntosh, not a result of this paper and not a restatement of Theorem 3.3. The cited works [8,30] are independent published stability and extension theorems whose assumptions do not contain the target bounds. The shared divergence term in the definitions of E_{T,p} and e_{K,p} does not make the equivalence circular: the nontrivial inequality is the upper bound on E_{T,p} in terms of the local errors, and the converse is a direct consequence of the minimization definitions. The dependence of Lemma A.1's constant on only d and kappa_K is an imported hypothesis whose correctness is a matter of numerical-analysis risk, not circularity: if [16, Corollary 3.4] had hidden p-dependence, Theorem 3.6's p-optimality would fail, but that would be a false assumption, not a derivation that assumes its own conclusion. There is no fitted parameter renamed as prediction, no self-citation chain forcing the main choice, and no renaming of a known pattern. The proof is self-contained against external benchmarks, so the appropriate finding is no significant circularity.

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

The central claim rests on standard finite element and harmonic analysis tools plus two imported equilibrium and regularity results. No new physical entities and no data-fitted parameters appear.

assumptions (4)
  • domain assumption Omega is a bounded Lipschitz polygon or polyhedron and T is a conforming, shape-regular, simplicial mesh matching Gamma_D and Gamma_N.
    Section 2.1; all estimates depend on the shape-regularity parameter kappa_T and the boundary partition.
  • standard math Stable patchwise flux equilibration bounds of Lemma 4.4, proven in [8, Theorem 7] for 2D and [30, Corollaries 3.3, 3.6, 3.8] for 3D.
    Invoked in Section 4.3 to bound sigma_a minus I_T^p(psi_a tau_T); this is an imported published result from the a posteriori error literature.
  • standard math Costabel-McIntosh regularized Poincare operator bound [16, Corollary 3.4].
    Imported in Appendix A, Lemma A.1, to prove the p-robust constrained-unconstrained local equivalence; load-bearing for p-optimality.
  • standard math Standard hp approximation bounds for the L2 projection (Babuska-Suri [2, Lemma 4.1]) and Poincare inequalities.
    Used throughout Sections 4 and 5 to bound projection errors; assumed p-robust with constants depending only on d and kappa_T.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Equivalence of local-and global-best approximations, a simple stable local commuting projector, and optimal $hp$ approximation estimates in $H(\mathrm{div})$." pith.science (2026). https://pith.science/paper/NINIP3Q2

@misc{pith2026190808158,
  author       = {Pith},
  title        = {Pith review of: Equivalence of local-and global-best approximations, a simple stable local commuting projector, and optimal $hp$ approximation estimates in $H(\mathrmdiv)$},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NINIP3Q2}},
  note         = {Machine review of arXiv:1908.08158}
}
read the original abstract

Given an arbitrary function in H(div), we show that the error attained by the global-best approximation by H(div)-conforming piecewise polynomial Raviart-Thomas-N\'ed\'elec elements under additional constraints on the divergence and normal flux on the boundary, is, up to a generic constant, equivalent to the sum of independent local-best approximation errors over individual mesh elements, without constraints on the divergence or normal fluxes. The generic constant only depends on the shape-regularity of the underlying simplicial mesh, the space dimension, and the polynomial degree of the approximations. The analysis also gives rise to a stable, local, commuting projector in H(div), delivering an approximation error that is equivalent to the local-best approximation. We next present a variant of the equivalence result, where robustness of the constant with respect to the polynomial degree is attained for unbalanced approximations. These two results together further enable us to derive rates of convergence of global-best approximations that are fully optimal in both the mesh size h and the polynomial degree p, for vector fields that only feature elementwise the minimal necessary Sobolev regularity. We finally show how to apply our findings to derive optimal a priori hp-error estimates for mixed and least-squares finite element methods applied to a model diffusion problem.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 46 canonical work pages

  1. [1]

    Each H 1/2- stable projection yields convergence and quasi-optimality of adapt ive FEM with inhomogeneous Dirichlet data in Rd

    Aurada, M., Feischl, M., Kemetm ¨uller, J., Page, M., and Praetorius, D. Each H 1/2- stable projection yields convergence and quasi-optimality of adapt ive FEM with inhomogeneous Dirichlet data in Rd. ESAIM Math. Model. Numer. Anal. 47 , 4 (2013), 1207–1235

    work page 2013

  2. [2]

    The h-p version of the finite element method with quasi-uniform meshes

    Babuˇska, I., and Suri, M. The h-p version of the finite element method with quasi-uniform meshes. RAIRO Mod´ el. Math. Anal. Num´ er. 21, 2 (1987), 199–238

    work page 1987

  3. [3]

    Local flux reconstructions for standard finite element methods on triangular meshes

    Becker, R., Capatina, D., and Luce, R. Local flux reconstructions for standard finite element methods on triangular meshes. SIAM J. Numer. Anal. 54 , 4 (2016), 2684–2706

    work page 2016

  4. [4]

    A local regularization operator for triangular and quadrilateral finite elements

    Bernardi, C., and Girault, V. A local regularization operator for triangular and quadrilateral finite elements. SIAM J. Numer. Anal. 35 , 5 (1998), 1893–1916

    work page 1998

  5. [5]

    Quelques propri´ et´ es d’approximation des ´ el´ ements finis de N´ed´ elec, application ` a l’analyse a posteriori

    Bernardi, C., and Hecht, F. Quelques propri´ et´ es d’approximation des ´ el´ ements finis de N´ed´ elec, application ` a l’analyse a posteriori. C. R. Math. Acad. Sci. Paris 344 , 7 (2007), 461–466. 18

    work page 2007

  6. [6]

    A new H(div)-conforming p-interpolation operator in two dimen- sions

    Bespalov, A., and Heuer, N. A new H(div)-conforming p-interpolation operator in two dimen- sions. ESAIM Math. Model. Numer. Anal. 45 , 2 (2011), 255–275

    work page 2011

  7. [7]

    Mixed finite element methods and applications , vol

    Boffi, D., Brezzi, F., and Fortin, M. Mixed finite element methods and applications , vol. 44 of Springer Series in Computational Mathematics . Springer, Heidelberg, 2013

    work page 2013

  8. [8]

    Equilibrated residual error estimates are p-robust

    Braess, D., Pillwein, V., and Sch ¨oberl, J. Equilibrated residual error estimates are p-robust. Comput. Methods Appl. Mech. Engrg. 198 , 13-14 (2009), 1189–1197

    work page 2009

Show all 46 references
  1. [9]

    Equilibrated residual error estimator for edge elements

    Braess, D., and Sch ¨oberl, J. Equilibrated residual error estimator for edge elements. Math. Comp. 77 , 262 (2008), 651–672

    2008

  2. [10]

    Optimal error estimate for the div least-squares method with data f ∈L2 and application to nonlinear problems

    Cai, Z., and Ku, J. Optimal error estimate for the div least-squares method with data f ∈L2 and application to nonlinear problems. SIAM J. Numer. Anal. 47 , 6 (2010), 4098–4111

    2010

  3. [11]

    H., Stevenson, R., and Verani, M

    Canuto, C., Nochetto, R. H., Stevenson, R., and Verani, M. Convergence and optimality of hp-AFEM. Numer. Math. 135 , 4 (2017), 1073–1119

    2017

  4. [12]

    Comparison results of finite element methods for the Poisson model problem

    Carstensen, C., Peterseim, D., and Schedensack, M. Comparison results of finite element methods for the Poisson model problem. SIAM J. Numer. Anal. 50 , 6 (2012), 2803–2823

    2012

  5. [13]

    Medius analysis and comparison results for first-order finite element methods in linear elasticity

    Carstensen, C., and Schedensack, M. Medius analysis and comparison results for first-order finite element methods in linear elasticity. IMA J. Numer. Anal. 35 , 4 (2015), 1591–1621

    2015

  6. [14]

    H., and Winther, R

    Christiansen, S. H., and Winther, R. Smoothed projections in finite element exterior calculus. Math. Comp. 77 , 262 (2008), 813–829

    2008

  7. [15]

    Approximation by finite element functions using local regularization

    Cl´ ement, P. Approximation by finite element functions using local regularization. RAIRO Anal. Numer. 9 , R-2 (1975), 77–84

    1975

  8. [16]

    On Bogovski ˘ ı and regularized Poincar´ e integral operators for de Rham complexes on Lipschitz domains

    Costabel, M., and McIntosh, A. On Bogovski ˘ ı and regularized Poincar´ e integral operators for de Rham complexes on Lipschitz domains. Math. Z. 265 , 2 (2010), 297–320

    2010

  9. [17]

    Polynomial exact sequences and projection-based interpolation w ith application to Maxwell equations

    Demkowicz, L. Polynomial exact sequences and projection-based interpolation w ith application to Maxwell equations. In Mixed finite elements, compatibility conditions, and appli cations, D. Boffi, F. Brezzi, L. F. Demkowicz, R. G. Dur´ an, R. S. Falk, and M. Fortin, Eds., vol. 19...

    1939

  10. [18]

    H 1, H(curl) and H(div)-conforming projection-based interpo- lation in three dimensions

    Demkowicz, L., and Buffa, A. H 1, H(curl) and H(div)-conforming projection-based interpo- lation in three dimensions. Quasi-optimal p-interpolation estimates. Comput. Methods Appl. Mech. Engrg. 194 , 2-5 (2005), 267–296

    2005

  11. [19]

    Polynomial extension operators

    Demkowicz, L., Gopalakrishnan, J., and Sch ¨oberl, J. Polynomial extension operators. Part I. SIAM J. Numer. Anal. 46 , 6 (2008), 3006–3031

    2008

  12. [20]

    Polynomial extension operators

    Demkowicz, L., Gopalakrishnan, J., and Sch ¨oberl, J. Polynomial extension operators. Part II. SIAM J. Numer. Anal. 47 , 5 (2009), 3293–3324

    2009

  13. [21]

    Polynomial extension operators

    Demkowicz, L., Gopalakrishnan, J., and Sch ¨oberl, J. Polynomial extension operators. Part III. Math. Comp. 81 , 279 (2012), 1289–1326

    2012

  14. [22]

    Explicit error bounds in a conforming finite element method

    Destuynder, P., and M ´ etivet, B. Explicit error bounds in a conforming finite element method. Math. Comp. 68 , 228 (1999), 1379–1396

    1999

  15. [23]

    hp-adaptation driven by polynomial-degree-robust a posteriori error estimates for elliptic problems

    Dolejˇs´ı, V., Ern, A., and Vohral ´ık, M. hp-adaptation driven by polynomial-degree-robust a posteriori error estimates for elliptic problems. SIAM J. Sci. Comput. 38 , 5 (2016), A3220–A3246

    2016

  16. [24]

    Mollification in strongly Lipschitz domains with application to continuous and discrete de Rham complexes

    Ern, A., and Guermond, J.-L. Mollification in strongly Lipschitz domains with application to continuous and discrete de Rham complexes. Comput. Methods Appl. Math. 16 , 1 (2016), 51–75

    2016

  17. [25]

    Finite element quasi-interpolation and best approximation

    Ern, A., and Guermond, J.-L. Finite element quasi-interpolation and best approximation. ESAIM Math. Model. Numer. Anal. 51 , 4 (2017), 1367–1385. 19

    2017

  18. [26]

    Quasi-optimal nonconforming approximation of elliptic PDEs with contrasted coefficients and minimal regularity

    Ern, A., and Guermond, J.-L. Quasi-optimal nonconforming approximation of elliptic PDEs with contrasted coefficients and minimal regularity. HAL Preprint 01964 299, submitted for publication, 2018

    2018

  19. [27]

    Discrete p-robust H(div)-liftings and a posteriori estimates for elliptic problems with H −1 source terms

    Ern, A., Smears, I., and Vohral ´ık, M. Discrete p-robust H(div)-liftings and a posteriori estimates for elliptic problems with H −1 source terms. Calcolo 54 , 3 (2017), 1009–1025

    2017

  20. [28]

    Adaptive inexact Newton methods with a posteriori stopping criter ia for nonlinear diffusion PDEs

    Ern, A., and Vohral ´ık, M. Adaptive inexact Newton methods with a posteriori stopping criter ia for nonlinear diffusion PDEs. SIAM J. Sci. Comput. 35 , 4 (2013), A1761–A1791

    2013

  21. [29]

    Polynomial-degree-robust a posteriori estimates in a unified settin g for conforming, nonconforming, discontinuous Galerkin, and mixed discretizations

    Ern, A., and Vohral ´ık, M. Polynomial-degree-robust a posteriori estimates in a unified settin g for conforming, nonconforming, discontinuous Galerkin, and mixed discretizations. SIAM J. Numer. Anal. 53 , 2 (2015), 1058–1081

    2015

  22. [30]

    Stable broken H 1 and H(div) polynomial extensions for polynomial- degree-robust potential and flux reconstruction in three space dimensions

    Ern, A., and Vohral ´ık, M. Stable broken H 1 and H(div) polynomial extensions for polynomial- degree-robust potential and flux reconstruction in three space dimensions. Math. Comp. 89 , 322 (2020), 551–594

    2020

  23. [31]

    S., and Winther, R

    F alk, R. S., and Winther, R. Local bounded cochain projections. Math. Comp. 83 , 290 (2014), 2631–2656

    2014

  24. [32]

    A new error analysis for discontinuous finite element methods for line ar elliptic problems

    Gudi, T. A new error analysis for discontinuous finite element methods for line ar elliptic problems. Math. Comp. 79 , 272 (2010), 2169–2189

    2010

  25. [33]

    A comment on least-squares finite element methods with minimum regu larity assumptions

    Ku, J. A comment on least-squares finite element methods with minimum regu larity assumptions. Int. J. Numer. Anal. Model. 10 , 4 (2013), 899–903

    2013

  26. [34]

    Licht, M. W. Smoothed projections and mixed boundary conditions. Math. Comp. 88 , 316 (2019), 607–635

    2019

  27. [35]

    Luce, R., and Wohlmuth, B. I. A local a posteriori error estimator based on equilibrated fluxes. SIAM J. Numer. Anal. 42 , 4 (2004), 1394–1414

    2004

  28. [36]

    M., and Rojik, C

    Melenk, J. M., and Rojik, C. On commuting p-version projection-based interpolation on tetra- hedra. Math. Comp. 89 , 321 (2020), 45–87

    2020

  29. [37]

    Mixed finite elements in R3

    N´ ed´ elec, J.-C. Mixed finite elements in R3. Numer. Math. 35 , 3 (1980), 315–341

    1980

  30. [38]

    H., and Stamm, B

    Nochetto, R. H., and Stamm, B. A posteriori error estimates for the electric field integral equation on polyhedra. In Contributions to partial differential equations and applic ations, vol. 47 of Comput. Methods Appl. Sci. Springer, Cham, 2019, pp. 371–394

    2019

  31. [39]

    I., Carey, G

    Pehlivanov, A. I., Carey, G. F., and Lazarov, R. D. Least-squares mixed finite elements for second-order elliptic problems. SIAM J. Numer. Anal. 31 , 5 (1994), 1368–1377

    1994

  32. [40]

    A mixed finite element method for 2nd order elliptic prob- lems

    Raviart, P.-A., and Thomas, J.-M. A mixed finite element method for 2nd order elliptic prob- lems. In Mathematical aspects of finite element methods (Proc. Conf. , Consiglio Naz. delle Ricerche (C.N.R.), Rome, 1975) . Springer, Berlin, 1977, pp. 292–315. Lecture Notes in Math., Vol. 606

    1975

  33. [41]

    E., and Thomas, J.-M

    Roberts, J. E., and Thomas, J.-M. Mixed and hybrid methods. In Handbook of Numerical Analysis, Vol. II . North-Holland, Amsterdam, 1991, pp. 523–639

    1991

  34. [42]

    Commuting quasi-interpolation operators for mixed finite elements

    Sch¨oberl, J. Commuting quasi-interpolation operators for mixed finite elements. Tech. rep. ISC- 01-10-MATH, Texas A&M University, 2001

    2001

  35. [43]

    A multilevel decomposition result in H(curl)

    Sch¨oberl, J. A multilevel decomposition result in H(curl). In Multigrid, Multilevel and Multiscale Methods (EMG 2005) (2005), P. Wesseling, C. Oosterlee, and P. Hemker, Eds., EMG 2005

    2005

  36. [44]

    R., and Zhang, S

    Scott, L. R., and Zhang, S. Finite element interpolation of nonsmooth functions satisfying boundary conditions. Math. Comp. 54 , 190 (1990), 483–493

    1990

  37. [45]

    Approximating gradients with continuous piecewise polynomial funct ions

    Veeser, A. Approximating gradients with continuous piecewise polynomial funct ions. Found. Comput. Math. 16 , 3 (2016), 723–750

    2016

  38. [46]

    A posteriori error estimation techniques for finite element methods

    Verf¨urth, R. A posteriori error estimation techniques for finite element methods. Numerical Mathematics and Scientific Computation. Oxford University Press, Oxford, 2013. 20

    2013

Pith tools

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