pith. sign in

arxiv: 2601.00404 · v2 · pith:UW7IVL7Ynew · submitted 2026-01-01 · 🧮 math.NA · cs.NA

Guaranteed stability bounds for second-order PDE problems satisfying a Garding inequality

T. Chaumont-Frelet This is my paper

Pith reviewed 2026-05-21 16:32 UTC · model grok-4.3

classification 🧮 math.NA cs.NA
keywords finite element methodGarding inequalityinf-sup constanta posteriori error estimationflux reconstructionwell-posednessstability boundssingular value problem
0
0 comments X

The pith

A numerical method using finite elements and flux reconstruction computes a lower bound on the inf-sup constant for PDEs satisfying a Garding inequality, underestimating the true value by a factor of roughly two on fine meshes.

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

The paper develops an algorithm to numerically check whether second-order linear PDE problems that satisfy a Garding inequality are well-posed. It solves two discrete singular value problems that combine a Lagrange finite element discretization with a flux reconstruction error estimator to obtain a computable lower bound on the inf-sup constant of the weak formulation. This bound supports a posteriori error estimation in numerical solutions. A reader would care because the approach turns theoretical stability questions into concrete, verifiable numbers without needing full analytical proofs for each new problem. The method guarantees that the computed bound stays within a factor of about two of the optimal constant once the finite element space is rich enough.

Core claim

We propose an algorithm to numerically determine whether a second-order linear PDE problem satisfying a Garding inequality is well-posed. This algorithm further provides a lower bound to the inf-sup constant of the weak formulation, which may in turn be used for a posteriori error estimation purposes. Our numerical lower bound is based on two discrete singular value problems involving a Lagrange finite element discretization coupled with an a posteriori error estimator based on flux reconstruction techniques. We show that if the finite element discretization is sufficiently rich, our lower bound underestimates the optimal constant only by a factor roughly equal to two.

What carries the argument

Two discrete singular value problems that pair a Lagrange finite element discretization with a flux reconstruction a posteriori error estimator to generate a guaranteed lower bound on the inf-sup constant.

If this is right

  • The algorithm numerically verifies well-posedness of the continuous PDE problem.
  • The resulting lower bound on the inf-sup constant can be inserted directly into a posteriori error estimators for discrete solutions.
  • On meshes fine enough that the finite element space is sufficiently rich, the bound remains within a factor of roughly two of the true optimal value.
  • The same construction applies to any second-order linear PDE that meets the Garding inequality assumption.

Where Pith is reading between the lines

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

  • The same flux-reconstruction technique might be tested on time-dependent or nonlinear problems by replacing the static singular-value solves with appropriate time-stepping or linearization steps.
  • Pairing the stability bound with adaptive mesh refinement could automatically refine regions where the estimator indicates poor control of the inf-sup constant.
  • The observed factor-of-two gap could be narrowed by substituting higher-order reconstruction operators or by enriching the test space beyond standard Lagrange elements.

Load-bearing premise

The PDE problem must satisfy a Garding inequality so that the continuous problem is well-posed and the discrete approximations remain controllable through the flux reconstruction estimator.

What would settle it

Apply the algorithm to the Poisson equation on a sequence of uniformly refined meshes and check whether the computed lower bound approaches the known exact inf-sup constant within a factor of two.

read the original abstract

We propose an algorithm to numerically determined whether a second-order linear PDE problem satisfying a Garding inequality is well-posed. This algorithm further provides a lower bound to the inf-sup constant of the weak formulation, which may in turn be used for a posteriori error estimation purposes. Our numerical lower bound is based on two discrete singular value problems involving a Lagrange finite element discretization coupled with an a posteriori error estimator based on flux reconstruction techniques. We show that if the finite element discretization is sufficiently rich, our lower bound underestimates the optimal constant only by a factor roughly equal to two.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript proposes an algorithm to numerically determine whether a second-order linear PDE problem satisfying a Garding inequality is well-posed. The algorithm computes a lower bound on the inf-sup constant of the weak formulation via two discrete singular-value problems on a Lagrange finite-element space, combined with an a posteriori estimator based on flux reconstruction. The central claim is that, provided the finite-element discretization is sufficiently rich, this numerical lower bound underestimates the true inf-sup constant by a factor of roughly two.

Significance. If the factor-of-two guarantee can be made fully rigorous and uniform, the work would supply a practical, computable stability bound useful for a posteriori error estimation on problems where only a Garding inequality is available. The reliance on existing flux-reconstruction estimators is a positive feature, as it avoids the introduction of new fitted parameters.

major comments (2)
  1. Abstract: the claim that the lower bound 'underestimates the optimal constant only by a factor roughly equal to two' is stated without an explicit derivation or tracked multiplicative constant relating the two discrete singular-value problems, the flux-reconstruction estimator, and the continuous inf-sup constant. Because this factor is the principal quantitative guarantee, its proof must be supplied with a precise constant independent of mesh size and polynomial degree.
  2. The handling of the Garding constant in the passage from the discrete inf-sup estimate to the continuous well-posedness statement is not visible in sufficient detail; any hidden dependence on this constant would undermine the uniformity of the factor-of-two bound under mesh refinement.
minor comments (2)
  1. The abstract contains a minor grammatical issue: 'numerically determined' should read 'numerically determine'.
  2. Notation for the Garding inequality and the precise weak formulation should be introduced with an equation number in the introduction for immediate reference.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and constructive comments on our manuscript. We address each major comment below and have made revisions to improve the rigor and clarity of the quantitative guarantees.

read point-by-point responses

  1. Referee: Abstract: the claim that the lower bound 'underestimates the optimal constant only by a factor roughly equal to two' is stated without an explicit derivation or tracked multiplicative constant relating the two discrete singular-value problems, the flux-reconstruction estimator, and the continuous inf-sup constant. Because this factor is the principal quantitative guarantee, its proof must be supplied with a precise constant independent of mesh size and polynomial degree.

    Authors: We agree that the abstract statement would benefit from greater precision. The full manuscript derives the bound by combining the two discrete singular-value problems with the flux-reconstruction estimator; the resulting multiplicative factor is exactly 2 and arises from the equivalence constants in the estimator, which are independent of mesh size and polynomial degree. We have revised the abstract to state the bound precisely and added an explicit derivation subsection that tracks every constant in the chain from the discrete problems through the estimator to the continuous inf-sup constant. revision: yes

  2. Referee: The handling of the Garding constant in the passage from the discrete inf-sup estimate to the continuous well-posedness statement is not visible in sufficient detail; any hidden dependence on this constant would undermine the uniformity of the factor-of-two bound under mesh refinement.

    Authors: The Gårding constant enters the argument only in establishing the threshold at which the discrete inf-sup condition implies the continuous one (via the standard compactness argument for Gårding-type problems). Once this threshold is met, further refinement does not reintroduce dependence on the Gårding constant. We have expanded the relevant section and theorem statement to display this dependence explicitly and to confirm that the factor-of-two bound remains uniform with respect to mesh size and polynomial degree beyond the threshold. revision: yes

Circularity Check

0 steps flagged

No circularity: lower bound derived from independent discrete singular-value problems and flux-reconstruction estimator

full rationale

The paper constructs a numerical lower bound on the inf-sup constant from two discrete singular-value problems on a Lagrange finite-element space, combined with an a posteriori flux-reconstruction estimator. The claim that this bound underestimates the true constant by a factor of roughly two for sufficiently rich discretizations is presented as a consequence of the estimator's properties and the discrete-continuous relationship, not as a definitional identity or a parameter fitted to the target quantity itself. No quoted equation reduces the central result to its own inputs by construction, and the derivation remains self-contained against external benchmarks such as the continuous Gårding inequality and standard a posteriori theory.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the domain assumption that the continuous problem satisfies a Garding inequality and on standard finite-element approximation theory; no free parameters or new invented entities are introduced.

axioms (1)
  • domain assumption The second-order linear PDE satisfies a Garding inequality
    Invoked in the problem statement to ensure well-posedness of the continuous weak formulation and to control the discrete approximations.

pith-pipeline@v0.9.0 · 5615 in / 1248 out tokens · 39248 ms · 2026-05-21T16:32:47.058507+00:00 · methodology

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

30 extracted references · 30 canonical work pages

  1. [1]

    Adams and J

    R. Adams and J. Fournier,Sobolev spaces, Academic Press, 2003

    work page 2003

  2. [2]

    Ainsworth,Discrete dispersion relation forhp-version finite element approximation at high wave num- ber, SIAM J

    M. Ainsworth,Discrete dispersion relation forhp-version finite element approximation at high wave num- ber, SIAM J. Numer. Anal.42(2004), no. 2, 553–575

    work page 2004

  3. [3]

    Bebendorf,A note on the Poincar´ e inequality for convex domains, Z

    M. Bebendorf,A note on the Poincar´ e inequality for convex domains, Z. Anal. Anwendungen22(2003), 751–756

    work page 2003

  4. [4]

    Berenger,Perfectly matched layer for the FDTD solution of wave-structure interaction problems, IEEE Trans

    J.-P. Berenger,Perfectly matched layer for the FDTD solution of wave-structure interaction problems, IEEE Trans. Antennas Propag.44(2002), no. 1, 110–117

    work page 2002

  5. [5]

    Braess, V

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

    work page 2009

  6. [6]

    Braess and J

    D. Braess and J. Sch¨ oberl,Equilibrated residual error estimators for edge elements, Math. Comp.77 (2008), no. 262, 651–672

    work page 2008

  7. [7]

    Carstensen and J

    C. Carstensen and J. Gedicke,Guaranteed lower bounds for eigenvalues, Math. Comp.83(2014), 2605– 2629

    work page 2014

  8. [8]

    Chaumont-Frelet, A

    T. Chaumont-Frelet, A. Ern, and M. Vohral´ ık,On the derivation of guaranteed andp-robust a posteriori error estimates for the Helmholtz equation, Numer. Math.148(2021), 525–573

    work page 2021

  9. [9]

    Chaumont-Frelet, D

    T. Chaumont-Frelet, D. Gallistl, S. Nicaise, and J. Tomezyk,Wavenumber explicit convergence analysis for finite element discretizations of time-harmonic wave propagation problems with perfectly matched layers author, Comun. Math. Sci.20(2022), no. 1, 1–52

    work page 2022

  10. [10]

    Chaumont-Frelet and S

    T. Chaumont-Frelet and S. Nicaise,High-frequency behaviour of corner singularities in Helmholtz problems, ESAIM Math. Model. Numer. Anal.5(2018), 1803–1845

    work page 2018

  11. [11]

    ,Wavenumber explicit convergence analysis for finite element discretizations of general wave prop- agation problems, IMA J. Numer. Anal.40(2020), no. 2, 1503–1543. GUARANTEED STABILITY BOUNDS 13

    work page 2020

  12. [12]

    Chaumont-Frelet and M

    T. Chaumont-Frelet and M. Vohral´ ık,Constrained and unconstrained stable discrete minimizations for p-robust local reconstructions in vertex patches in the De Rham complex, Found. Comput. Math. (2024), 1–42

    work page 2024

  13. [13]

    Destuynder and B

    P. Destuynder and B. M´ etivet,Explicit error bounds in a conforming finite element method, Math. Comp. 68(1999), no. 228, 1379–1396

    work page 1999

  14. [14]

    D¨ orfler and S

    W. D¨ orfler and S. Sauter,A posteriori error estimation for highly indefinite Helmholtz problems, Comput. Meth. Appl. Math.13(2013), 333–347

    work page 2013

  15. [15]

    Ern and M

    A. Ern and M. Vohral´ ık,Polynomial-degree-robust a posteriori estimates in a unified setting for conforming, nonconforming, discontinuous Galerkin, and mixed discretizations, SIAM J. Numer. Anal.53(2015), no. 2, 1058–1081

    work page 2015

  16. [16]

    Comp.89(2021), 551–594

    ,Stable brokenH 1 andH(div)polynomial extensions for polynomial-degree-robust potential and flux reconstruction in three space dimensions, Math. Comp.89(2021), 551–594

    work page 2021

  17. [17]

    Fernandes and G

    P. Fernandes and G. Gilardi,Magnetostatic and electrostatic problems in inhomogeneous anisotropic media with irregular boundary and mixed boundary conditions, Math. Meth. Appl. Sci.47(1997), no. 4, 2872– 2896

    work page 1997

  18. [18]

    Galkowski, E.A

    J. Galkowski, E.A. Spence, and J. Wunsch,Optimal constans in non-trapping resolvent estimates and applications in numerical analysis, Pure Appl. Anal.2(2020), no. 1, 157–202

    work page 2020

  19. [19]

    Gallistl,Mixed methods and lower eigenvalue bounds, Math

    D. Gallistl,Mixed methods and lower eigenvalue bounds, Math. Comp.92(2023), 1491–1509

    work page 2023

  20. [20]

    Girault and P.A

    V. Girault and P.A. Raviart,Finite element methods for Navier-Stokes equations: theory and algorithms, Springer-Verlag, 1986

    work page 1986

  21. [21]

    Karkulik and J.M

    M. Karkulik and J.M. Melenk,Local high-order regularization and applications tohp-methods, Comp. Math. Appl.70(2015), 1606–1639

    work page 2015

  22. [22]

    Liu and S

    X. Liu and S. Oishi,Verified eigenvalue evaluation for the Laplacian over polygonal domains of arbitrary shape, SIAM J. Numer. Anal.51(2013), no. 3, 1634–1654

    work page 2013

  23. [23]

    Melenk and S

    J.M. Melenk and S. Sauter,Wavenumber explicit convergence analysis for Galerkin discretizations of the Helmholtz equation, SIAM J. Numer. Anal.49(2011), no. 3, 1210–1243

    work page 2011

  24. [24]

    Nakao, M

    M.T. Nakao, M. Plum, and Y. Watanabe,Numerical verification methods and computer-assisted proofs for partial differential equations, Springer series in computational mathematics, 2019

    work page 2019

  25. [25]

    N´ ed´ elec,Mixed finite elements inR3, Numer

    J.C. N´ ed´ elec,Mixed finite elements inR3, Numer. Math.35(1980), 315–341

    work page 1980

  26. [26]

    Prager and J.L

    W. Prager and J.L. Synge,Approximations in elasticity based on the concept of function space, Quart. Appl. Math.5(1947), no. 3, 241–269

    work page 1947

  27. [27]

    Raviart and J.M

    P.A. Raviart and J.M. Thomas,A mixed finite element method for 2nd order elliptic problems, Mathemat- ical Aspect of Finite Element Methods, Springer-Verlag, 1977

    work page 1977

  28. [28]

    Sauter and J

    S. Sauter and J. Zech,A posteriori error estimation ofhp−dgfinite element methods for highly indefinite Helmholtz problems, SIAM J. Numer. Anal.53(2015), no. 5, 2414–2440

    work page 2015

  29. [29]

    Vohral´ ık,Unified primal formulation-based a priori and a posteriori error analysis of mixed finite element methods, Math

    M. Vohral´ ık,Unified primal formulation-based a priori and a posteriori error analysis of mixed finite element methods, Math. Comp.79(2010), 2001–2032

    work page 2010

  30. [30]

    Watanabe, T

    Y. Watanabe, T. Kinoshita, and M.T. Nakao,Efficient approaches for verifying the existence and bound of inverse of linear operators in Hilbert spaces, J. Sci. Comput.94(2023), no. 43, 1–18

    work page 2023