Pith. sign in

REVIEW 4 minor 46 references

On positivity preservation of hybrid discontinuous Galerkin methods on hypergraphs

T0 review · 0 major / 4 minor · reviewed 2026-08-08 · deepseek-v4-flash

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

Pith's one-line read The paper proves which low-order HDG discretisations of the diffusion equation preserve solution nonnegativity — a nonobtuse mesh or a sufficiently large stabilisation parameter suffices — and shows by counterexample that many standard…

desk verdict Solid, useful paper: proves positivity preservation for low-order HDG on hypergraphs via M-matrix arguments, with honest counterexamples; the main soft spot is a broader closing claim and missing reproducibility scripts, not the math. read the letter →

arxiv 2502.07976 v2 pith:2JCC27RY submitted 2025-02-11 math.NA cs.NA

Petr Knobloch , Philip L. Lederer , Andreas Rupp This is my paper · ORCID
classification math.NAcs.NA MSC 65N3065N1265N06
keywords positivitypreservationhybriddiscontinuousGalerkinHDGmethodsdiffusionequationM-matrixdiscretemaximumprincipleRaviart–Thomaselementshypergraphs
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 asks when a numerical method for the stationary diffusion equation inherits a basic property of the continuous solution: nonnegative data must produce a nonnegative concentration, which matters whenever the unknown is a temperature or a concentration. The authors prove that a small family of low-order hybrid discontinuous Galerkin (HDG) discretisations on graphs, hypergraphs, and polyhedral domains preserves positivity in both the bulk unknown $u_h$ and the skeleton unknown $\lambda_h$. The sufficient conditions are explicit: either every hyperedge has no obtuse angles ($\mathbf{n}_{E|N}\cdot\mathbf{n}_{E|N'} \le 0$ for distinct faces), or the stabilisation parameter satisfies $\tau_E \ge |\partial E|/\int_E \kappa^{-1}\,\mathrm{d}x$. The mechanism is that the reduced skeleton system becomes an M-matrix, so all computed values are nonnegative, and local reconstruction formulas pass the property to the element interiors. The paper also constructs counterexamples showing that many otherwise standard HDG choices, including bilinear unknowns with linear skeleton unknowns, fail to preserve positivity.

What carries the argument

The load-bearing object is the reduced linear system satisfied by the skeleton unknown $\lambda_h$ after the element unknowns are eliminated, displayed as (4.4) for the lowest-order scheme and as (3.1)/(3.8) on graphs. Its matrix has entries $a_{NN'} = \sum_E |N||N'|(\bar\kappa_E \mathbf{n}_{E|N}\cdot\mathbf{n}_{E|N'} - \tau_E/|\partial E|)$ for distinct nodes $N, N'$, so the off-diagonal entries are nonpositive exactly when the nonobtuse angle condition (4.5) holds or the penalty satisfies (2.6). A square matrix of nonnegative type with vanishing row sums and nonnegative right-hand side is an M-matrix, which forces $\lambda_h \ge 0$; explicit local formulas such as (4.1b) then carry nonnegativity from the skeleton to the bulk unknown $u_h$. On graphs the same structure is seen most directly: the HDG method is provably identical to the standard finite difference stencil, whose matrix is tridiagonal with nonpositive off-diagonals.

What would settle it

Run the lowest-order LDG-H method of Theorem 2.4(3) on a single obtuse simplex with $\kappa$ constant, $f = 0$, and nonnegative Dirichlet data, with $\tau_E$ set exactly at the right-hand side of (2.6): any computed skeleton value below zero would refute the theorem. As a boundary probe of the square bound, repeat the computation on a square with $Q_1$-type flux at $\tau_E$ just below $2\kappa_E/h$: if $\lambda_h$ stays nonnegative for all nonnegative data there, the claimed threshold is not necessary.

Watch

Extended reading notes

Core claim

The central result, Theorem 2.4, characterises a hierarchy of positivity-preserving HDG methods for the diffusion equation on hypergraphs. On graphs, with edgewise constant diffusion, the HDG method coincides with a finite difference scheme whose matrix is an M-matrix: for $P_1/P_1$ spaces the skeleton variable and element means are nonnegative (and the whole solution is nonnegative if $f$ is constant per edge), while for the combinations $P_1/P_0$, $P_0/P_1$, and $P_0/P_0$ the pointwise solution is nonnegative. In any dimension, the method with piecewise-constant unknowns for the bulk, the flux, and the skeleton (LDG-H) preserves nonnegativity if each hyperedge satisfies the nonobtuse-angle condition $\mathbf{n}_{E|N}\cdot\mathbf{n}_{E|N'} \le 0$ for distinct faces, or if the stabilisation satisfies $\tau_E \ge |\partial E| / \int_E \kappa^{-1}$. The lowest-order Raviart–Thomas method (RT-H) is positivity preserving on nonobtuse simplices with $\tau_E = 0$, and on rectangles the $Q_1$, $P_1^2$, and $\mathrm{RT}_0$ flux choices are equivalent and preserve positivity once $\tau_E \ge C(\varrho)\kappa_E/\mathrm{diam}(E)$, with the square case reducing to the explicit value $\tau_E \ge 2\kappa_E/h$. Complementing these theorems, Section 5 exhibits explicit two-square configurations, with nonnegative Dirichlet data and $f = 0$, in which $P_1/P_1$, $Q_1/P_1$, $P_1/Q_1$, $Q_1/Q_1$, and $Q_1/P_0$ versions produce negative values, and the authors conclude that for most HDG methods positivity can be violated for any choice of $\tau$.

Load-bearing premise

The positivity conclusions are driven by one structural fact: after eliminating the element unknowns, the skeleton system's off-diagonal entries are nonpositive only if every hyperedge is nonobtuse or the stabilisation parameter satisfies the threshold $\tau_E \ge |\partial E|/\int_E \kappa^{-1}\,\mathrm{d}x$; where neither holds, the paper's own counterexamples show that positivity can fail, so no general guarantee exists outside this M-matrix regime.

Editorial extensions

If this is right

  • On a one-dimensional graph the $P_1/P_1$ HDG scheme with edgewise-constant diffusion is exactly a finite difference method, so its skeleton values and element means are nonnegative and, for edgewise-constant $f$, the whole approximate concentration is nonnegative.
  • The lowest-order LDG-H method ($P_0/P_0/P_0$) preserves positivity on arbitrary hypergraphs once $\tau_E \ge |\partial E|/\int_E \kappa^{-1}$, so it can serve as the consistent low-order scheme inside flux-corrected or limited transport constructions.
  • On nonobtuse simplices the RT0 hybrid method with $\tau_E = 0$ preserves positivity, and because hybridized and non-hybrid RT0 coincide, the classical lowest-order Raviart–Thomas method inherits the property.
  • For rectangular hyperedges the $Q_1^2$, $P_1^2$, and $\mathrm{RT}_0$ flux choices define one and the same method, so the entire rectangle analysis collapses to a single inequality in the aspect ratio.
  • The positivity-preserving penalty regimes cost convergence order: choosing $\tau_E = O(1/h)$ or $O(1)$ without the angle condition changes the predicted $L^2$ error rates for the flux and primal variables listed in the paper.

Reading between the lines

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

  • The counterexamples in Section 5 fail already at the skeleton level with $f = 0$ and piecewise-constant Dirichlet data, which suggests positivity is governed purely by the discrete Green's function of the reduced skeleton operator; compiling the minimal positivity-preserving $\tau_E$ per element shape would give a practical design chart for monotone HDG.
  • The threshold (2.6) is the harmonic conductance ratio $|\partial E|/\int_E \kappa^{-1}$, the same quantity that dictates monotonicity in two-point flux finite volume schemes; the analysis here suggests a quantitative transfer of these monotonicity ideas to hybrid high-order methods, as the authors anticipate.
  • The rectangle theorem assumes $\kappa$ constant on each element; since the proof only needs the sign of one coefficient in (4.15), a natural extension is to test whether the same threshold works for elementwise-varying $\kappa$, where the current argument would need a modified inequality.
  • The M-matrix structure actually yields inverse-positivity, not merely nonnegativity, so the graph results should extend to a full discrete maximum principle (bounds by the extrema of the boundary data), not just a sign-preservation statement; checking this on hypergraphs is a direct next question.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

0 major / 4 minor

Summary. The paper studies positivity preservation (nonnegativity of bulk and skeleton unknowns) for hybridized discontinuous Galerkin (HDG) discretizations of stationary diffusion equations posed on graphs, hypergraphs, and standard domains. The main result, Theorem 2.4, gives sufficient conditions under which λh and uh are nonnegative for several low-order space choices: P1/P1 and low-order combinations on graphs, P0/P0 with either a nonobtuse-angle condition or a large-enough penalty parameter on hypergraphs, lowest-order Raviart–Thomas spaces on nonobtuse simplices with τ=0, and rectangular-hyperedge methods with U=P0, M=P0 and Q∈{Q1², P1², RT0} provided τ is sufficiently large. The proofs proceed by eliminating cell unknowns and showing that the reduced skeleton system has nonpositive off-diagonal entries and vanishing row sums, hence is a nonsingular M-matrix after imposing Dirichlet data. Section 5 supplies explicit two-dimensional counterexamples for space choices or parameter ranges outside the sufficient conditions, and Section 6 reports numerical experiments that match the predicted thresholds, including the sharpness of the square threshold τ≥2κ/h. The paper does not claim a complete characterization of the intermediate-τ regime, and the conclusion's statement that 'most HDG methods' fail positivity is broader than what the counterexamples strictly establish.

Significance. If the results are correct, the paper fills a genuine gap: local mass conservation and positivity preservation are both desirable for diffusion discretizations, and existing positivity-preserving mixed/hybrid results were restricted to special two-dimensional meshes or to strongest-order cases. The paper's main strengths are that the positivity conditions are explicit and parameter-free in the sense that they involve only geometric properties of the mesh and known stabilization thresholds; that the M-matrix arguments are transparent and checkable; that the counterexamples and numerical experiments directly probe the sharpness of the conditions; and that the numerical experiments are reproducible with public software (HyperHDG and NGsolve). The constructive route of restoring positivity by over-penalization, at the price of convergence order, is a practically useful observation for flux-limited or convex-limited high-order schemes. I found no fitted parameters or circular reasoning in the proof of the main theorem: the cited works [25] and [47] supply the hypergraph framework and software, not the positivity conclusion.

minor comments (4)
  1. [Section 7] The final sentence states that 'for most HDG methods, the positivity can be violated for any choice of τ', but this is stronger than what Section 5 demonstrates: the counterexamples cover specific low-order space combinations, and some of them (for example the two-square example in Section 5.4) exhibit failure only for specific values or ranges of τ, not for all τ. I recommend qualifying this sentence to say that the tested low-order choices can fail, rather than asserting a general impossibility statement.
  2. [Section 6.3] The caption of Figure 11 says 'varying values of θ', but in Section 6.3 the angle parameter θ is fixed at 1.5 and the plotted quantity is the value of λh for varying τE; the caption should say 'varying values of τE'.
  3. [Section 6.1.2] The text 'suspect that for dimensions d ≥ 2, we need to ensure τ ≥ 6' is a numerical conjecture for d>2, not a consequence of Theorem 2.4; it should be explicitly labeled as a conjecture inferred from the experiments.
  4. [Various] There are several typographical errors: 'transferreed' in the Introduction, 'aspect ration' before equation (4.16), and 'be' instead of 'by' in the sentence preceding (2.6). These should be corrected in a final pass.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the positivity theorems are proved by M-matrix arguments from the HDG equations, with self-citations only supplying framework and software.

full rationale

The paper's central claim is Theorem 2.4, which gives sufficient conditions for nonnegativity of HDG solutions on graphs and hypergraphs. The proof chain is self-contained: for graphs, Theorem 3.1 and Theorem 3.2 show equivalence to finite-difference systems whose matrices have nonpositive off-diagonal entries and nonnegative row sums, hence are M-matrices after Dirichlet completion; for hypergraphs, Section 4.1 derives the skeleton system (4.4), verifies a_{NN'} ≤ 0 under either the geometric condition (4.5) or the penalty bound (4.6), and then applies the same M-matrix argument. The rectangular case in Section 4.3 derives explicit local formulas and reduces the sign condition to the quadratic inequality (4.16), yielding the threshold (2.7). No fitted parameter is introduced: the thresholds τ_E ≥ |∂E| / ∫_E κ^{-1} dx and τ_E ≥ C(ϱ)κ_E/diam(E) are a priori inequalities obtained by requiring nonpositivity of computed off-diagonal entries, not by calibrating to the solution. The cited works [25] and [47] by one of the authors provide the hypergraph modeling framework and the HyperHDG software, but the positivity conclusion, the M-matrix structure, and the counterexamples are derived within the paper from the discrete equations (2.4). The counterexamples in Section 5 give explicit nonnegative data with negative HDG solutions, so the sufficiency claims are not vacuous and the converse is honestly bounded. The uncharacterized intermediate-τ region is a limitation, explicitly acknowledged by the conditions being sufficient only, not a circularity. The concluding sentence that 'most HDG methods' fail positivity is broader than the proven cases, but it is not load-bearing for Theorem 2.4, which remains an internally consistent derivation from stated assumptions.

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

No quantities are fitted to data. The stabilization parameter tau is user-selected, not fit to data; the constant C(rho) for rectangles is existential, defined through inequality (4.16), and is not a fitted measurement. The derivations rest on standard Sobolev and M-matrix facts plus the hypergraph geometric model from [25].

assumptions (4)
  • standard math Existence and uniqueness of the weak solution and the equivalence of the primal and mixed-hybrid formulations, cited from [25] and [42], are used to define the continuous problem and to justify unique solvability of the HDG system.
    Section 2, before Theorem 2.3; the discrete analysis starts from this equivalence.
  • standard math M-matrix criterion: a nonsingular matrix with nonpositive off-diagonal entries and positive diagonal has an inverse with nonnegative entries, cited from [4, Corollary 3.13].
    Used in Lemma 3.4 and Sections 4.1 to 4.3 to conclude lambda_h at least zero from the skeleton system.
  • domain assumption The hypergraph domain model from [25]: hyperedges are polyhedra, hypernodes are (d-1)-dimensional faces, and flux balance holds at each hypernode.
    Section 2, item (2); the junction condition (2.5) and the positivity proof at junctions depend on this structure.
  • domain assumption Piecewise constant kappa and f, nonnegative Dirichlet data gD,h, and elementwise constant data conditions in Theorem 2.4 are assumed; variable coefficients inside elements are not covered.
    Theorems 2.4 and 3.1 to 3.2; local formulas (4.1) and (3.3) require elementwise constant kappa.

how reviews work

0 comments
Cite this review

Pith. "Pith review of On positivity preservation of hybrid discontinuous Galerkin methods on hypergraphs." pith.science (2026). https://pith.science/paper/2JCC27RY

@misc{pith2026250207976,
  author       = {Pith},
  title        = {Pith review of: On positivity preservation of hybrid discontinuous Galerkin methods on hypergraphs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2JCC27RY}},
  note         = {Machine review of arXiv:2502.07976}
}
read the original abstract

Hybrid finite element methods, particularly hybridized discontinuous Galerkin (HDG) methods, are efficient numerical schemes for discretizing the diffusion equation, which encompasses two main physical principles: mass conservation and positivity preservation. While the former has been extensively analyzed in the literature, this paper investigates the latter. We state a theorem that guarantees the positivity of both the bulk and skeleton approximations to the primary unknown (concentration) and provide counterexamples for nonpositive discretizations. The theoretical findings are confirmed by numerical experiments.

Figures

Figures reproduced from arXiv: 2502.07976 by the authors.

Figure 1
Figure 1. Local coordinate system for a rectangular hyperedge E and no￾tation for hypernodes. • ph = ek, k = 1, 2, yields q0 = − κE |E| Z ∂E λhnE dσ, (2.10) • ph = xiek, i, k ∈ {1, 2}, yields a k i = δik 12κE h 2 i u0 − 12κE |E| h 2 i Z N + k λhxi dσ − Z N − k λhxi dσ ! , • ph = x1x2ek, k ∈ {1, 2}, yields b k = ai 12κE h 2 k − 72κE h 3 i h 2 k Z N + k ∪N − k λhxi dσ, where i ̸= k, i ∈ {1, 2}. On the other hand, testing (2.4b)… view at source ↗
Figure 2
Figure 2. Hypergraph consisting of two squares E1, E2. 5.1. Counterexample for P1(Eh)×P1(Eh) 2 ×P1(Nh). Let us first consider the case when Eh consists of only one hyperedge E being a rectangle with edges parallel to the coordinate axes. Assume (5.1), i.e., ΓD = ∂E, and let gD(x, y) = (a(x − xE) + b)(c(y − yE) + d) ≥ 0 for all (x, y) ∈ E, (5.2) where (xE, yE) is the barycenter of E and a, b, c, d are real numbers. Consider th… view at source ↗
Figure 3
Figure 3. Counterexample for Q1(Eh) × P1(Eh) 2 × P1(Nh): values of λh (left) and uh (right). boundary ΓD is again assumed to consist of all boundary hypernodes so that the only hypernode which is not contained in ΓD is N12. Let us set gD(x, y) = ( (2(x − xE1 ) + 1)(2(y − yE1 ) + 1) for all (x, y) ∈ E1, (2(x − xE2 ) + 3)(2(y − yE2 ) + 1) for all (x, y) ∈ E2 \ N12. Then gD ∈ C(E1 ∪ E2) and gD ≥ 0. The definition of gD makes it … view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Counterexample for Q1(Eh) × Q1(Eh) 2 × P1(Nh): values of λh (left) and uh (right). 5.4. Counterexample for Q1(Eh) × Q1(Eh) 2 × P1(Nh). Like in the counterexamples 5.1 and 5.2, we consider a hypergraph consisting of the two unit squares depicted in [PITH_FULL_IMAGE:fig…
Figure 5
Figure 5. Figure 5: Counterexample for Q1(Eh)×Q1(Eh) 2 ×P1(Nh): values of uh for τ = 3. It can be verified that then both (2.4a) and (2.4b) are satisfied. Moreover, we obtain (qh|E1 · nE1 + τ (uh − λh))|N12 + (qh|E2 · nE2 + τ (uh − λh))|N12 = 1 2 − 5 2 (λ1 + λ2) + (1 − (2 + β)(λ1 − λ2)) (…
Figure 6
Figure 6. Figure 6: The value of λh at x1 = 1 (ordinate) for τ > 0 (abscissa) in the case U(E) = P0(E), Q(E) = P0(E) d and M(N) = P0(N) [PITH_FULL_IMAGE:figures/full_fig_p025_6.png]
Figure 7
Figure 7. Figure 7: The value of λh at x1 = 1 (ordinate) for τ > 0 (abscissa) in the case U(E) = P0(E), Q(E) = Q1(E) d and M(N) = P0(N). 6.1.3. Multilinear primal approximation: U(E) = Q1(E), Q(E) = P0(E) d and M(N) = Q1(N). Selecting multilinear functions for the primal and skeletal unkn…
Figure 8
Figure 8. Figure 8: The mean of λh at x1 = 1 (ordinate) for τ > 0 (abscissa) in the case U(E) = Q1(E), Q(E) = P0(E) d and M(N) = Q1(N) [PITH_FULL_IMAGE:figures/full_fig_p026_8.png]
Figure 9
Figure 9. Figure 9: The reference configuration on Ω and the mapped triangulation ˆ and domain for θ = 1. 6.2. Raviart-Thomas approximation on simplices: U(E) = P0(E), Q(E) = RT0(E) and M(N) = P0(N). As a reference domain we consider the unit hypercube Ω := (0 ˆ , 1)d and a structured tri…
Figure 10
Figure 10. Figure 10: The value of λh on the hypernode N for varying values of θ for d = 2 (top) and d = 3 (top and bottom). 6.3. The lowest-order case on quadrilaterals: U(E) = P0(E), Q(E) = P0(E) d and M(N) = P0(N). We use the same setting as in the previous example but now consider a qu…
Figure 11
Figure 11. Figure 11: The value of λh on the hypernode N for varying values of θ for d = 2. is selected. This choice of τ leads to positivity preservation also for a few other discretiza￾tions as the present paper shows. However, for most HDG methods, the positivity can be violated for any…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 46 canonical work pages

  1. [25]

    Partial differential equations on hypergraphs and networks of surfaces: Derivation and hybrid discretizations

    A. Rupp, M. Gahn, and G. Kanschat. “Partial differential equations on hypergraphs and networks of surfaces: Derivation and hybrid discretizations”. In: ESAIM: Math- ematical Modelling and Numerical Analysis 56.2 (2022), pp. 505–528

    work page 2022

  2. [47]

    Rupp and G

    A. Rupp and G. Kanschat. HyperHDG: Hybrid discontinuous Galerkin methods for PDEs on hypergraphs. published online. 2021. REFERENCES 33 Department of Numerical Mathematics, F aculty of Mathematics and Physics, Charles University, Prague, Czech Republic Email address: knobloch@karlin.mff.cuni.cz Department of Mathematics, University of Hamburg, Germany Ema...

    work page 2021

  3. [1]

    A nodally bound- preserving finite element method

    G. R. Barrenechea, E. Georgoulis, T. Pryer, and A. Veeser. “A nodally bound- preserving finite element method”. In: IMA Journal of Numerical Analysis 44 (2024), pp. 2198–2219

    work page 2024

  4. [2]

    Edge-based nonlinear diffusion for finite element approximations of convection-diffusion equations and its relation to algebraic flux-correction schemes

    G. R. Barrenechea, E. Burman, and F. Karakatsani. “Edge-based nonlinear diffusion for finite element approximations of convection-diffusion equations and its relation to algebraic flux-correction schemes”. In: Numer. Math. 135.2 (2017), pp. 521–545

    work page 2017

  5. [3]

    An algebraic flux correction scheme satisfying the discrete maximum principle and linearity preservation on general meshes

    G. R. Barrenechea, V. John, and P. Knobloch. “An algebraic flux correction scheme satisfying the discrete maximum principle and linearity preservation on general meshes”. In: Math. Models Methods Appl. Sci. 27.3 (2017), pp. 525–548. 30 REFERENCES

    work page 2017

  6. [4]

    Finite element methods respecting the discrete maximum principle for convection-diffusion equations

    G. Barrenechea, V. John, and P. Knobloch. “Finite element methods respecting the discrete maximum principle for convection-diffusion equations”. In: SIAM Rev. 66.1 (2024), pp. 3–88

    work page 2024

  7. [5]

    The discrete maximum principle for linear simplicial finite element approximations of a reaction-diffusion problem

    J. H. Brandts, S. Korotov, and M. Kˇ r ´ ıˇ zek. “The discrete maximum principle for linear simplicial finite element approximations of a reaction-diffusion problem”. In: Linear Algebra Appl. 429.10 (2008), pp. 2344–2357

    work page 2008

  8. [6]

    Discrete maximum principle for finite-difference operators

    P. G. Ciarlet. “Discrete maximum principle for finite-difference operators”. In: Ae- quationes Math. 4 (1970), pp. 338–352

    work page 1970

Show all 46 references
  1. [7]

    Maximum principle and uniform convergence for the finite element method

    P. G. Ciarlet and P.-A. Raviart. “Maximum principle and uniform convergence for the finite element method”. In: Comput. Methods Appl. Mech. Engrg. 2 (1973), pp. 17–31

    1973

  2. [8]

    A Personal Discussion on Conservation, and How to Formulate It

    R. Abgrall. “A Personal Discussion on Conservation, and How to Formulate It”. In: Finite Volumes for Complex Applications X—Volume 1, Elliptic and Parabolic Problems. Ed. by E. Franck, J. Fuhrmann, V. Michel-Dansac, and L. Navoret. Cham: Springer, 2023, pp. 3–19

    2023

  3. [9]

    The Continuous Galerkin Method Is Locally Conservative

    T. Hughes, G. Engel, L. Mazzei, and M. Larson. “The Continuous Galerkin Method Is Locally Conservative”. In: Journal of Computational Physics 163 (Sept. 2000), pp. 467–488

    2000

  4. [10]

    Proximal Galerkin: A Structure-Preserving Finite El- ement Method for Pointwise Bound Constraints

    B. Keith and T. M. Surowiec. “Proximal Galerkin: A Structure-Preserving Finite El- ement Method for Pointwise Bound Constraints”. In: Foundations of Computational Mathematics (2024)

    2024

  5. [11]

    Barrenechea, V

    G. Barrenechea, V. John, and P. Knobloch. Monotone Discretizations for Elliptic Second Order Partial Differential Equations . Cham: Springer, 2025

    2025

  6. [12]

    Bound-preserving flux limiting schemes for DG discretizations of conservation laws with applications to the Cahn–Hilliard equa- tion

    F. Frank, A. Rupp, and D. Kuzmin. “Bound-preserving flux limiting schemes for DG discretizations of conservation laws with applications to the Cahn–Hilliard equa- tion”. In: Computer Methods in Applied Mechanics and Engineering 359 (2020)

    2020

  7. [13]

    Locally bound-preserving enriched Galerkin methods for the linear advection equation

    D. Kuzmin, H. Hajduk, and A. Rupp. “Locally bound-preserving enriched Galerkin methods for the linear advection equation”. In: Computers & Fluids 205.104525 (2020), p. 15

    2020

  8. [15]

    Kuzmin, S

    D. Kuzmin, S. Lee, and Y. Yang. Bound-preserving and entropy stable enriched Galerkin methods for nonlinear hyperbolic equations . 2024

    2024

  9. [16]

    Monolithic convex limiting in discontinuous Galerkin discretizations of hyperbolic conservation laws

    H. Hajduk. “Monolithic convex limiting in discontinuous Galerkin discretizations of hyperbolic conservation laws”. In: Computers & Mathematics with Applications 87 (2021), pp. 120–138

    2021

  10. [17]

    Sparse invariant domain preserving discontinuous Galerkin methods with subcell convex limiting

    W. Pazner. “Sparse invariant domain preserving discontinuous Galerkin methods with subcell convex limiting”. In: Computer Methods in Applied Mechanics and Engineer- ing 382 (2021), p. 113876

    2021

  11. [18]

    Invariant-Domain-Preserving High-Order Time Stepping: I. Explicit Runge–Kutta Schemes

    A. Ern and J. Guermond. “Invariant-Domain-Preserving High-Order Time Stepping: I. Explicit Runge–Kutta Schemes”. In: SIAM Journal on Scientific Computing 44.5 (2022), A3366–A3392. REFERENCES 31

    2022

  12. [19]

    Invariant domain preserving discretization- independent schemes and convex limiting for hyperbolic systems

    J. Guermond, B. Popov, and I. Tomas. “Invariant domain preserving discretization- independent schemes and convex limiting for hyperbolic systems”. In: Computer Methods in Applied Mechanics and Engineering 347 (2019), pp. 143–175

    2019

  13. [20]

    Kuzmin and H

    D. Kuzmin and H. Hajduk. Property-Preserving Numerical Schemes for Conservation Laws. World Scientific, 2023

    2023

  14. [21]

    Limiters for high-order discontinuous Galerkin methods

    L. Krivodonova. “Limiters for high-order discontinuous Galerkin methods”. In: Jour- nal of Computational Physics 226.1 (2007), pp. 879–896

    2007

  15. [22]

    A vertex-based hierarchical slope limiter for p-adaptive discontinuous Galerkin methods

    D. Kuzmin. “A vertex-based hierarchical slope limiter for p-adaptive discontinuous Galerkin methods”. In: Journal of Computational and Applied Mathematics 233.12 (2010). Finite Element Methods in Engineering and Science (FEMTEC 2009), pp. 3077– 3085

    2010

  16. [23]

    A new perspective on flux and slope limiting in discontinuous Galerkin methods for hyperbolic conservation laws

    D. Kuzmin. “A new perspective on flux and slope limiting in discontinuous Galerkin methods for hyperbolic conservation laws”. In: Computer Methods in Applied Me- chanics and Engineering 373 (2021), p. 113569

    2021

  17. [24]

    Unified Hybridization of Discon- tinuous Galerkin, Mixed, and Continuous Galerkin Methods for Second Order Elliptic Problems

    B. Cockburn, J. Gopalakrishnan, and R. Lazarov. “Unified Hybridization of Discon- tinuous Galerkin, Mixed, and Continuous Galerkin Methods for Second Order Elliptic Problems”. In: SIAM J. Numer. Anal. 47.2 (2009), pp. 1319–1365

    2009

  18. [26]

    The maximum princi- ple violations of the mixed-hybrid finite-element method applied to diffusion equa- tions

    H. Hoteit, R. Mos´ e, B. Philippe, P. Ackerer, and J. Erhel. “The maximum princi- ple violations of the mixed-hybrid finite-element method applied to diffusion equa- tions”. In: International Journal for Numerical Methods in Engineering 55.12 (2002), pp. 1373–1390

    2002

  19. [27]

    A new mass lumping scheme for the mixed hybrid finite element method

    A. Younes, P. Ackerer, and F. Lehmann. “A new mass lumping scheme for the mixed hybrid finite element method”. In: International Journal for Numerical Methods in Engineering 67.1 (2006), pp. 89–107

    2006

  20. [28]

    A technique for improving the accuracy of quadrangular mixed finite elements for Darcy’s flow on heterogeneous domains

    A. Younes, P. Ackerer, S. Ahmed, and R. Bouhlila. “A technique for improving the accuracy of quadrangular mixed finite elements for Darcy’s flow on heterogeneous domains”. In: Computers & Fluids 39.2 (2010), pp. 189–196

    2010

  21. [29]

    A conservative and monotone mixed- hybridized finite element approximation of transport problems in heterogeneous do- mains

    M. Brera, J. Jerome, Y. Mori, and R. Sacco. “A conservative and monotone mixed- hybridized finite element approximation of transport problems in heterogeneous do- mains”. In: Computer Methods in Applied Mechanics and Engineering 199.41 (2010), pp. 2709–2720

    2010

  22. [30]

    An analysis of monotonicity conditions in the mixed hybrid finite ele- ment method on unstructured triangulations

    A. Mazzia. “An analysis of monotonicity conditions in the mixed hybrid finite ele- ment method on unstructured triangulations”. In: International Journal for Numer- ical Methods in Engineering 76.3 (2008), pp. 351–375

    2008

  23. [31]

    On Some Mixed Finite Element Methods with Numerical Integration

    S. Micheletti, R. Sacco, and F. Saleri. “On Some Mixed Finite Element Methods with Numerical Integration”. In: SIAM Journal on Scientific Computing 23.1 (2001), pp. 245–270

    2001

  24. [32]

    A discrete maximum principle for the weak Galerkin finite element method on nonuniform rectangular partitions

    Y. Liu and J. Wang. “A discrete maximum principle for the weak Galerkin finite element method on nonuniform rectangular partitions”. In: Numerical Methods for Partial Differential Equations 36.3 (2020), pp. 552–578. 32 REFERENCES

    2020

  25. [33]

    Discrete maximum principle for the P1-P0 weak Galerkin finite element approximations

    J. Wang, X. Ye, Q. Zhai, and R. Zhang. “Discrete maximum principle for the P1-P0 weak Galerkin finite element approximations”. In: Journal of Computational Physics 362 (2018), pp. 114–130

    2018

  26. [34]

    Discrete maximum principle for the weak Galerkin method on triangular and rectangular meshes

    H. Zhou, X. Wang, and J. Jia. “Discrete maximum principle for the weak Galerkin method on triangular and rectangular meshes”. In: Journal of Computational and Applied Mathematics 402 (2022), p. 113784

    2022

  27. [35]

    Discrete Maximum Principle for the Weak Galerkin Method for Anisotropic Diffusion Problems

    W. Huang and Y. Wang. “Discrete Maximum Principle for the Weak Galerkin Method for Anisotropic Diffusion Problems”. In: Communications in Computational Physics 18.1 (2015), 65–90

    2015

  28. [36]

    Bridging the hybrid high-order and hy- bridizable discontinuous Galerkin methods

    B. Cockburn, D. A. Di Pietro, and A. Ern. “Bridging the hybrid high-order and hy- bridizable discontinuous Galerkin methods”. In:ESAIM: M2AN 50.3 (2016), pp. 635– 650

    2016

  29. [37]

    Monotone finite vol- ume schemes for diffusion equations on unstructured triangular and shape-regular polygonal meshes

    K. Lipnikov, M. Shashkov, D. Svyatskiy, and Y. Vassilevski. “Monotone finite vol- ume schemes for diffusion equations on unstructured triangular and shape-regular polygonal meshes”. In: Journal of Computational Physics 227 (Nov. 2007), pp. 492– 512

    2007

  30. [38]

    Non-negative mixed finite element formulations for a tensorial diffusion equation

    K. Nakshatrala and A. Valocchi. “Non-negative mixed finite element formulations for a tensorial diffusion equation”. In: Journal of Computational Physics 228 (Oct. 2008), pp. 6726–6752

    2008

  31. [39]

    Monotone nonlinear finite-volume method for challenging grids

    M. Schneider, B. Flemisch, R. Helmig, K. Terekhov, and H. Tchelepi. “Monotone nonlinear finite-volume method for challenging grids”. In:Computational Geosciences 22 (Apr. 2018)

    2018

  32. [40]

    Analysis of the nonlinear scheme preserving the maximum principle for the anisotropic diffusion equation on distorted meshes

    Z. Sheng and G. Yuan. “Analysis of the nonlinear scheme preserving the maximum principle for the anisotropic diffusion equation on distorted meshes”. In: Science China Mathematics 65.11 (2022), pp. 2379–2396

    2022

  33. [41]

    Gilbarg and N

    D. Gilbarg and N. Trudinger. Elliptic partial differential equations of second order . Springer-Verlag, Berlin, 2001

    2001

  34. [42]

    Brezzi and M

    F. Brezzi and M. Fortin. Mixed and hybrid finite element methods . Springer-Verlag, New York, 1991

    1991

  35. [43]

    A projection-based error analysis of HDG methods

    B. Cockburn, J. Gopalakrishnan, and F. Sayas. “A projection-based error analysis of HDG methods”. In: Mathematics of Computation 79.271 (2010), pp. 1351–1367

    2010

  36. [44]

    A superconvergent LDG-hybridizable Galerkin method for second-order elliptic problems

    B. Cockburn, B. Dong, and J. Guzm´ an. “A superconvergent LDG-hybridizable Galerkin method for second-order elliptic problems”. In: Math. Comp. 77.264 (2008), pp. 1887– 1916

    2008

  37. [45]

    Mimetic Finite Difference Meth- ods for Diffusion Equations

    J. Hyman, J. Morel, M. Shashkov, and S. Steinberg. “Mimetic Finite Difference Meth- ods for Diffusion Equations”. In: Computational Geosciences 6.3 (2002), pp. 333–352

    2002

  38. [46]

    Boffi, F

    D. Boffi, F. Brezzi, and M. Fortin. Mixed finite element methods and applications . Vol. 44. Springer Series in Computational Mathematics. Springer, Heidelberg, 2013, pp. xiv+685

    2013

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.