An Extended Galerkin Analysis for Elliptic Problems

Jinchao Xu; Qingguo Hong; Shuonan Wu

REVIEW 3 major objections 3 minor 52 references

An Extended Galerkin Analysis for Elliptic Problems

T0 review · 3 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 The paper claims that a single four-variable discretization is uniformly stable in mesh size and penalty parameters, so most finite element and discontinuous Galerkin methods can be analyzed as special cases of it.

desk verdict A useful 4-field unified framework with two genuinely parameter-uniform inf-sup theorems, but the abstract's 'most methods' claim overshoots the proven parameter regimes—the standard O(1)-penalty equal-order HDG case is explicitly left unproved. read the letter →

arxiv 1908.08205 v2 pith:TLQYZI3L submitted 2019-08-22 math.NA cs.NA

Qingguo Hong , Shuonan Wu , Jinchao Xu This is my paper · ORCID

classification math.NAcs.NA MSC 65N3065N1265N15

keywords extendedGalerkinanalysisfour-fieldformulationinf-supconditiondiscontinuousmethodshybridizableweakmixedfiniteelementellipticboundaryvalueproblems

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 tries to establish that one four-variable discretization—interior solution and flux approximations $u_h$, $\boldsymbol{p}_h$, plus boundary corrections $\check u_h$, $\check p_h$—carries a complete, parameter-uniform stability theory for second-order elliptic problems. If that is true, most existing finite element and discontinuous Galerkin methods (continuous Galerkin, mixed, HDG, weak Galerkin, LDG, interior-penalty variants) are special cases of one formulation and share one quasi-optimal error analysis. The supporting theorems are two inf-sup conditions, one gradient-based and one divergence-based, whose constants do not depend on mesh size or stabilization parameters. The same framework also interpolates between $H^1$-conforming primal methods and $H(\mathrm{div})$-conforming mixed methods as the parameter $\rho$ tends to zero.

What carries the argument

The central object is the four-field bilinear form $\tilde a((\tilde{\boldsymbol p}_h,\tilde u_h),(\tilde{\boldsymbol q}_h,\tilde v_h))=a(\tilde{\boldsymbol p}_h,\tilde{\boldsymbol q}_h)+b(\tilde{\boldsymbol q}_h,\tilde u_h)+b(\tilde{\boldsymbol p}_h,\tilde v_h)-c(\tilde u_h,\tilde v_h)$, where $a$ contains the flux inner product and the $\tau^{-1}$ penalty on $\check p_h$, $c$ contains the $\eta^{-1}$ penalty on $\check u_h$, and $b$ couples the interior and trace variables either through $\nabla_h$ (3.11c) or through $\mathrm{div}_h$ (3.11d). This dual reading of $b$ is what generates the two theorems: the gradient reading drives Theorem 4.1 and the divergence reading drives Theorem 4.2, with test functions assembled from the stable-pair argument of [36] so that all constants are independent of $\rho$, $h$, and the stabilization parameters. The framework's reach comes from eliminating the trace variables in Section 6, which converts the four-field system into the equations of existing methods while preserving the corresponding uniform inf-sup condition.

What would settle it

Measure the inf-sup constant of the four-field form for the equal-order HDG choice $Q_h=\boldsymbol{Q}_h^k$, $V_h=V_h^k$, $\check V_h=\check V_h^k$, $\eta=\tau^{-1}/4=O(1)$ on a sequence of refined meshes; a bounded constant across refinements would show a stable, widely used method outside the range of the two theorems, while a growing constant would confirm the gap identified in Remark 6.1.

Watch

Extended reading notes

Core claim

The paper's central claim is that the four-field system (3.10), built from the special traces $\bar u_h=\{u_h\}$ and $\bar{\boldsymbol p}_h=\{\boldsymbol{p}_h\}_e$ and the penalties $\check p_h\approx\tau[u_h]_e$, $\check u_h\approx\eta[\boldsymbol{p}_h]$, is uniformly well posed under either of two sets of discrete-space assumptions. Theorem 4.1 asks that the scalar trace space $\check Q_h$ contain piecewise constants, that $\nabla_h V_h\subset \boldsymbol{Q}_h$, and that $\{\nabla_h V_h\}_e\subset\check Q_h$, with $\tau=(\rho h_e)^{-1}$ and $\eta\simeq\tau^{-1}$. Theorem 4.2 asks that $R_h=\boldsymbol{Q}_h\cap H(\mathrm{div},\Omega)$ form a stable mixed pair with $V_h$, that $\mathrm{div}_h\boldsymbol{Q}_h=V_h$, and that $\{\mathrm{div}_h\boldsymbol{Q}_h\}\subset\check V_h$, with $\eta=(\rho h_e)^{-1}$ and $\tau\simeq\eta^{-1}$. From either set of assumptions follows inf-sup stability with constants independent of $\rho$ and $h$, quasi-optimal error estimates, and algebraic rates. Section 6 then shows that eliminating $\check p_h$, $\check u_h$, or both recovers HDG, weak Galerkin, mixed DG, LDG, and conforming methods as special cases, and that the two limiting regimes $\rho\to 0$ converge to the $H^1$-conforming primal method and the $H(\mathrm{div})$-conforming mixed method.

Load-bearing premise

The load-bearing premise is that the special average traces $\bar u_h=\{u_h\}$ and $\bar{\boldsymbol p}_h=\{\boldsymbol{p}_h\}_e$, together with the space inclusions listed in Theorem 4.1 or 4.2, cover the methods one wants to unify; the equal-order HDG case with $\eta=\tau^{-1}/4=O(1)$, singled out in Remark 6.1, is not covered by the proofs.

Editorial extensions

If this is right

A method designer can verify stability for a new Galerkin scheme by checking one of two short lists of space inclusions instead of proving an inf-sup condition from scratch.
All recovered methods inherit quasi-optimal error estimates and parameter-uniform stability constants from a single proof, so constants for HDG, WG, LDG, and mixed methods can be compared on the same footing.
The $\rho\to 0$ limits place conforming primal and mixed methods at the endpoints of a one-parameter family, explaining earlier observed equivalences among DG methods as limits of one formulation.
The elimination count in Remark 6.2—fourteen reduced methods obtained by removing any subset of the four fields—suggests that hybridized algorithms for new variants can be designed systematically and analyzed by the same two theorems.

Reading between the lines

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

A testable extension is whether the equal-order HDG case left open in Remark 6.1, with $\eta=\tau^{-1}/4=O(1)$, can be rescaled into the proved parameter range; if so, the framework would cover the most common HDG implementation.
The duality between the gradient-based and divergence-based theorems suggests that any scheme satisfying one theorem has a counterpart satisfying the other, obtained by swapping the roles of solution and flux and exchanging $\tau$ with $\eta$; this could generate new stabilized methods in pairs.
Because the two theorems rely only on local space inclusions and parameter scales, they plausibly extend to variable-coefficient and vector-valued elliptic problems, which the paper itself does not treat.

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.

Desk Editor's Note

read the letter

Quick take: this is a real contribution. The 4-field formulation with two boundary residual corrections is new relative to the Arnold-Brezzi-Cockburn-Marini 2002 and Cockburn-Gopalakrishnan-Lazarov 2009 unified analyses, and the two inf-sup theorems are genuinely proved for their stated regimes. The reductions in Section 6 (HDG with reduced stabilization, WG, LDG, mixed DG, and limiting to H1 and H(div) conforming methods) are valuable and mostly check out. The proof technique is delegated to the authors' own preprint [36], but the bilinear-form manipulations in the text are detailed enough to be plausible.

The soft spot is exactly the one the stress-test flags. Theorem 4.1 requires τ=(ρh_e)^{-1} and η~τ^{-1}=ρh_e; Theorem 4.2 requires η=(ρh_e)^{-1} and τ~η^{-1}=ρh_e. Both are h-scaled penalty regimes with ρ∈(0,ρ0]. The standard equal-order HDG method uses Q_h=V_h=P^k, check-V=check-Q=P^k, and τ=O(1), η=1/(4τ)=O(1). Remark 6.1 explicitly says this case is not proved, and Table 6.1 lists that space choice as "not proved." So the abstract's "most existing finite element and discontinuous Galerkin methods" is an overreach. A reader who wants the framework to cover constant-penalty HDG is out of luck.

This is a coverage limitation, not a fatal flaw. The framework does prove uniform stability for a substantial family of methods with properly scaled penalties, and the limiting arguments as ρ→0 recovering the primal and mixed methods are interesting. The boundedness argument in Theorem 4.2 is omitted as standard, and the inf-sup technique is delegated to [36]; that is worth asking the authors to expand, but it is not a red flag. There is no circularity: stability is proved for the 4-field system and then specialized.

Who is this for? People working on unified analyses of FEM and DG, and anyone using HDG or WG with h-scaled stabilization. I would bring it to a reading group. It deserves a serious referee, but the authors should be asked to either prove the O(1)-penalty HDG case or explicitly scope the claims. If that case cannot be covered, the abstract and conclusion need to say so.

Referee Report

3 major / 3 minor

Summary. The paper proposes a unified four-field Galerkin framework for the second-order elliptic problem (1.1), with unknowns u_h, p_h, \check u_h, and \check p_h representing interior potential, interior flux, and their boundary residual corrections, respectively. The formulation is written in compact form (3.10) and is shown to be consistent. Two inf-sup theorems are stated in Section 4: a gradient-based theorem (Theorem 4.1) requiring \tau=(\rho h_e)^{-1} and \eta\sim\tau^{-1}=\rho h_e, and a divergence-based theorem (Theorem 4.2) requiring \eta=(\rho h_e)^{-1} and \tau\sim\eta^{-1}=\rho h_e, both under explicit space-inclusion assumptions and with uniformity for \rho\in(0,\rho_0]. Section 5 studies limiting behavior as \rho\to0 and shows convergence to H^1-conforming primal and H(div)-conforming mixed methods. Section 6 eliminates fields to recover HDG, WG, LDG, mixed DG, and other methods, with a summary table. The abstract claims that the resulting inf-sup conditions are uniform with respect to all discretization and penalization parameters and that most existing finite element and discontinuous Galerkin methods can be analyzed in this framework.

Significance. The four-field formulation is elegant, the derivation is not circular, and the reductions in Section 6 are informative: the paper gives a plausible and largely detailed proof of two parameter-dependent inf-sup theorems, provides quasi-optimality estimates, and honestly records in Remark 6.1 that the standard equal-order HDG case is not proved. If the claims were restricted to the h-scaled parameter regime actually treated, the paper would be a solid contribution to the unified analysis of FEMs and DG methods. As written, however, the advertised breadth exceeds the proved results: the uniformity claim covers only special scalings of the penalization parameters, and a prominent method (equal-order HDG with O(1) stabilization) is explicitly excluded.

major comments (3)

[Abstract; Section 4; Remark 6.1] The central claim that the four-field formulation satisfies inf-sup conditions uniform in all discretization and penalization parameters and that most existing DG methods are covered is not supported by the theorems. Theorem 4.1 proves uniformity only for \tau=(\rho h_e)^{-1}, \eta\sim\tau^{-1}=\rho h_e, and Theorem 4.2 only for \eta=(\rho h_e)^{-1}, \tau\sim\eta^{-1}=\rho h_e, with \rho\in(0,\rho_0]. The standard equal-order HDG method, one of the most widely used DG methods, uses \tau=O(1) and \eta=1/(4\tau)=O(1); Table 6.1 and Remark 6.1 explicitly state that this case is not proved. Please either extend the analysis to this parameter regime or revise the abstract, the introduction, and the conclusion so that the stated coverage matches the proved parameter ranges.
[Theorem 4.2, Section 4.2] The boundedness of the bilinear form \tilde a with respect to the norms in (4.6) is asserted as standard and omitted. Since these norms contain trace terms such as \langle\rho^{-1}h_e^{-1}\check Q_h^u[p_h],\check Q_h^u[p_h]\rangle and the proof later uses terms of the form \langle\{div_h p_h\},[p_h]\rangle and \langle[div_h p_h]_e,\check p_h\rangle, the boundedness estimate is not completely trivial. Please provide the estimate or a precise reference so that Theorem 4.2 is self-contained.
[Hypotheses of Theorems 4.1 and 4.2] The relations \eta\sim=\tau^{-1} and \tau\sim=\eta^{-1} are not quantified. The proofs show that the coercivity constants depend on whether \eta^{-1} is bounded below by a fixed multiple of \rho^{-1}h_e^{-1} (Theorem 4.1) and whether \tau^{-1} is bounded below by a fixed multiple of \rho^{-1}h_e^{-1} (Theorem 4.2). If the proportionality constants are allowed to vary, the claimed uniformity in penalization parameters fails. Please state explicit admissible ranges, for example c_1\rho h_e\le\eta\le c_2\rho h_e, and indicate whether the final constants are allowed to depend on c_1 and c_2.

minor comments (3)

[Abstract and throughout] The abstract contains typographical errors such as 'm any', and the phrase 'most existing' is used without qualification; please copyedit.
[Theorem 4.1, condition (c)] Condition (c), \{\nabla_h V_h\}_e\subset\check Q_h, is stated and invoked in the proof, but the displayed estimates for the term \langle\check u_h,[\nabla_h u_h]\rangle appear not to require it; either use it explicitly in the argument or remove it from the assumptions.
[Section 6, Table 6.1 and Conclusion] The conclusion states that the results 'naturally lead to uniform inf-sup conditions of HDG, WG and the DG method', but Table 6.1 contains a row marked 'not proved' for the standard HDG choice. Please add a sentence distinguishing methods for which the inf-sup condition is proved in this paper from methods recovered only algebraically.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the inf-sup proofs are self-contained, self-citations are attributional, and the unproved O(1)-penalty HDG case is a coverage limitation, not a circular reduction.

full rationale

The derivation chain is not circular. Theorems 4.1 and 4.2 are proved in the paper with explicit test functions (4.5) and (4.11), tracked constants, and explicit inequalities; the sentence 'The proof follows from the technique shown in [36]' attributes a technique to the authors' own preprint, but the argument is reproduced in full in this paper, so this self-citation is not a load-bearing premise. The reductions in Sections 5 and 6 are variable eliminations (e.g., (6.1) and (6.9)) and algebraic manipulations, not fits of parameters to target results; the scalings τ = (ρh_e)^{-1}, η ∼ τ^{-1} in Theorem 4.1 and η = (ρh_e)^{-1}, τ ∼ η^{-1} in Theorem 4.2 are stated hypotheses, not outputs derived from the methods claimed as special cases. No quantity is fitted and then renamed as a prediction. The one admitted gap, Remark 6.1, states: 'the uniform inf-sup condition for the HDG method when η = 1/4 τ^{-1} = O(1), Q_h = Q^k_h, V_h = V^k_h, ˆV_h = ˆV^k_h is not proved in Section 4.' This confirms that the abstract's 'most existing ... DG methods' overstates the proved parameter regime, but an unproved special case is a coverage limitation, not a circular reduction: the proved inf-sup theorems do not assume the stability of the standard HDG method they would need to conclude it. The self-citations to [22] and [36] are pointers to related work and a proof technique, not an imported uniqueness theorem or an unverified premise that forces the result. Therefore the circularity score is 0.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central framework rests on standard finite element machinery (trace and inverse inequalities, interpolation, Babuska theory), several domain assumptions, and two hand-made choices: the stabilization scaling ρ and the restriction to the symmetric trace pair. No new physical entities are postulated; the two boundary variables u_h and p_h are discrete test functions within an existing mathematical formulation, not independently falsifiable objects.

free parameters (3)

ρ = not fitted (theoretical parameter)
Chosen by hand as the scaling in τ = (ρ h_e)^{-1} and η ~ τ^{-1} (gradient-based) or η = (ρ h_e)^{-1} and τ ~ η^{-1} (divergence-based); theorems hold uniformly for ρ in (0, ρ0].
τ = τ = (ρ h_e)^{-1} or τ ~ η^{-1}
Interior penalty stabilization parameter for the trace variable p_h; hand-chosen, not data-dependent.
η = η ~ τ^{-1} or η = (ρ h_e)^{-1}
Interior penalty stabilization parameter for the trace variable u_h; hand-chosen, not data-dependent.

assumptions (4)

domain assumption Bounded symmetric positive definite α with bounded inverse c; bounded domain Ω; nonzero measure of Γ_D.
Used in the model problem (1.1) and in all norms and estimates; stated in Section 1.
standard math Shape-regular triangulation family; standard trace, inverse, and interpolation inequalities; Piola transformation; Babuska theory.
Used throughout Section 4 proofs; trace and inverse inequalities are invoked in the inf-sup estimates, and interpolation theory yields error estimates (4.4) and (4.9).
ad hoc to paper Space inclusions: ∇_h V_h subset Q_h, the trace space Q_h containing piecewise constants, and {∇_h V_h}_e subset Q_h (gradient-based); div_h Q_h = V_h, Rh × Vh stable pair, and {div_h Q_h} subset V_h (divergence-based).
These conditions are introduced in Theorems 4.1 and 4.2; they are examples of the 'appropriate choices' the abstract promises, but they are conditions a method must satisfy to be covered.
ad hoc to paper Restriction to the special trace choice bar u_h = {u_h}, bar p_h = {p_h}_e in the compact form (3.10).
The unified analysis in Section 4 is proved only for this trace choice; the general traces from [28, Table 3.1] appear in (3.5) but are not analyzed.

how reviews work

0 comments

Cite this review

Pith. "Pith review of An Extended Galerkin Analysis for Elliptic Problems." pith.science (2026). https://pith.science/paper/TLQYZI3L

@misc{pith2026190808205,
  author       = {Pith},
  title        = {Pith review of: An Extended Galerkin Analysis for Elliptic Problems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TLQYZI3L}},
  note         = {Machine review of arXiv:1908.08205}
}

abstract

A general analysis framework is presented in this paper for many different types of finite element methods (including various discontinuous Galerkin methods). For second order elliptic equation, this framework employs $4$ different discretization variables, $u_h, \bm{p}_h, \check u_h$ and $\check p_h$, where $u_h$ and $\bm{p}_h$ are for approximation of $u$ and $\bm{p}=-\alpha \nabla u$ inside each element, and $ \check u_h$ and $\check p_h$ are for approximation of residual of $u$ and $\bm{p} \cdot \bm{n}$ on the boundary of each element. The resulting 4-field discretization is proved to satisfy inf-sup conditions that are uniform with respect to all discretization and penalization parameters. As a result, most existing finite element and discontinuous Galerkin methods can be analyzed using this general framework by making appropriate choices of discretization spaces and penalization parameters.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

52 extracted references · 52 canonical work pages

[36]

Uniform Stability and Error Analysis for Some Discontinuous Galerkin Methods

Qingguo Hong and Jinchao Xu. Uniformly stable results and error estimates for some discontinuous Galerkin methods. arXiv preprint arXiv:1805.09670 , 2018

work page Pith review arXiv 2018
[1]

Solution of problems of elasticity by the fra mework method

Alexander Hrennikoﬀ. Solution of problems of elasticity by the fra mework method. Journal of Applied Mechanics, 8(4):169–175, 1941

work page 1941
[2]

Variational methods for the solution of proble ms of equilibrium and vibrations

Richard Courant. Variational methods for the solution of proble ms of equilibrium and vibrations. Bulletin of the American Mathematical Society , 49(1):1–23, 1943

work page 1943
[3]

A generalization of the direct-stiﬀness method of structural analysis

Robert E Jones. A generalization of the direct-stiﬀness method of structural analysis. AIAA Journal , 2(5):821–826, 1964

work page 1964
[4]

Finite diﬀerence method based on variation principle

Kang Feng. Finite diﬀerence method based on variation principle. Communication on Applied Mathe- matics and Computation , 2(4):237–261, 1965. 18

work page 1965
[5]

Displacement and equilibrium models in the ﬁnite element method

Baudouin Fraeijs de Veubeke. Displacement and equilibrium models in the ﬁnite element method. Stress analysis, pages 145–197, 1965

work page 1965
[6]

A formulation of matrix displacement method

Yoshiyuki Yamamoto. A formulation of matrix displacement method . [Cambridge, Mass.]: Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics , 1966

work page 1966
[7]

On the ﬁnite element method

Miloˇ s Zl´ amal. On the ﬁnite element method. Numerische Mathematik , 12(5):394–409, 1968

work page 1968

Show all 52 references

[8]

A variational principle and the conv ergence of a ﬁnite-element method based on assumed stress distribution

Tong Pin and Theodore HH Pian. A variational principle and the conv ergence of a ﬁnite-element method based on assumed stress distribution. International Journal of Solids and Structures , 5(5):463–472, 1969

1969
[9]

Multipoint Taylor formulas and application s to the ﬁnite element method

PG Ciarlet and C Wagschal. Multipoint Taylor formulas and application s to the ﬁnite element method. Numerische Mathematik , 17(1):84–100, 1971

1971
[10]

On a class of ﬁnite elements generated by Lagrang e interpolation

RA Nicolaides. On a class of ﬁnite elements generated by Lagrang e interpolation. SIAM Journal on Numerical Analysis, 9(3):435–445, 1972

1972
[11]

On the existence, uniqueness and approximatio n of saddle-point problems arising from Lagrangian multipliers

Franco Brezzi. On the existence, uniqueness and approximatio n of saddle-point problems arising from Lagrangian multipliers. Revue fran¸ caise d’automatique, informatique, rechercheop´ erationnelle. Analyse num´ erique, 8(2):129–151, 1974

1974
[12]

A mixed ﬁnite ele ment method for 2nd order elliptic problems

Pierre-Arnaud Raviart and Jean-Marie Thomas. A mixed ﬁnite ele ment method for 2nd order elliptic problems. In Mathematical Aspects of Finite Element Methods , pages 292–315. Springer, 1977

1977
[13]

Mixed ﬁnite elements inR3

Jean-Claude N´ ed´ elec. Mixed ﬁnite elements inR3. Numerische Mathematik , 35(3):315–341, 1980

1980
[14]

Mixed and nonconforming ﬁn ite element methods: implemen- tation, postprocessing and error estimates

Douglas N Arnold and Franco Brezzi. Mixed and nonconforming ﬁn ite element methods: implemen- tation, postprocessing and error estimates. RAIRO-Mod´ elisation math´ ematique et analyse num´ erique, 19(1):7–32, 1985

1985
[15]

Mixed and hybrid ﬁnite element methods , volume 15 of Springer Series in Computational Mathematics

Franco Brezzi and Michel Fortin. Mixed and hybrid ﬁnite element methods , volume 15 of Springer Series in Computational Mathematics . Springer-Verlag, 1991

1991
[16]

Mixed ﬁnite element methods and applications , vol- ume 44 of Springer Series in Computational Mathematics

Daniele Boﬃ, Franco Brezzi, and Michel Fortin. Mixed ﬁnite element methods and applications , vol- ume 44 of Springer Series in Computational Mathematics . Springer, 2013

2013
[17]

Probl` ems aux limites non homog` enes ` a don´ ees irr´eguli` eres: Une m´ ethode d’approximation.in Numerical Analysis of Partial Diﬀerential Equations (C.I

JL Lions. Probl` ems aux limites non homog` enes ` a don´ ees irr´eguli` eres: Une m´ ethode d’approximation.in Numerical Analysis of Partial Diﬀerential Equations (C.I. ME.2 Ciclo,Ispra.1967). Edizioni Cremonese, Rome, pages 283–292, 1968

1967
[18]

Approximation des problemes aux limites non h omogenes pour des op´ erateurs non lin´ eaires.Journal of Mathematical Analysis and Applications , 30(3):510–521, 1970

Jean-Pierre Aubin. Approximation des problemes aux limites non h omogenes pour des op´ erateurs non lin´ eaires.Journal of Mathematical Analysis and Applications , 30(3):510–521, 1970

1970
[19]

Interior penalty procedures for elliptic and parabolic Galerkin methods

Jim Douglas and Todd Dupont. Interior penalty procedures for elliptic and parabolic Galerkin methods. In Computing Methods in Applied Sciences , pages 207–216. Springer, 1976

1976
[20]

The local discontinuous Galerkin method for time-dependent convection-diﬀusion systems

Bernardo Cockburn and Chi-Wang Shu. The local discontinuous Galerkin method for time-dependent convection-diﬀusion systems. SIAM Journal on Numerical Analysis , 35(6):2440–2463, 1998

1998
[21]

Discontinu- ous Galerkin approximations for elliptic problems

Franco Brezzi, Gianmarco Manzini, Donatella Marini, Paola Pietra, and Alessandro Russo. Discontinu- ous Galerkin approximations for elliptic problems. Numerical Methods for Partial Diﬀerential Equations , 16(4):365–378, 2000

2000
[22]

A uniﬁed s tudy of continuous and discontinuous Galerkin methods

Qingguo Hong, Fei Wang, Shuonan Wu, and Jinchao Xu. A uniﬁed s tudy of continuous and discontinuous Galerkin methods. Science China Mathematics , 62(1):1–32, 2019

2019
[23]

Nonconforming elements in the ﬁ nite element method with penalty

Ivo Babuˇ ska and Miloˇ s Zl´ amal. Nonconforming elements in the ﬁ nite element method with penalty. SIAM Journal on Numerical Analysis , 10(5):863–875, 1973. 19

1973
[24]

An elliptic collocation-ﬁnite element method with interior penalties

Mary Fanett Wheeler. An elliptic collocation-ﬁnite element method with interior penalties. SIAM Journal on Numerical Analysis , 15(1):152–161, 1978

1978
[25]

An interior penalty ﬁnite element method with dis continuous elements

Douglas N Arnold. An interior penalty ﬁnite element method with dis continuous elements. SIAM Journal on Numerical Analysis , 19(4):742–760, 1982

1982
[26]

Discontinu- ous ﬁnite elements for diﬀusion problems

Franco Brezzi, Gianmarco Manzini, Donatella Marini, Paola Pietra, and Alessandro Russo. Discontinu- ous ﬁnite elements for diﬀusion problems. Atti Convegno in onore di F. Brioschi (Milano 1997), Istitut o Lombardo, Accademia di Scienze e Lettere , pages 197–217, 1999

1997
[27]

The development of discontinuous Galerkin methods

Bernardo Cockburn, George E Karniadakis, and Chi-Wang Shu. The development of discontinuous Galerkin methods. In Discontinuous Galerkin Methods , pages 3–50. Springer, 2000

2000
[28]

Uniﬁed analysis of discontinuous Galerkin methods for elliptic problems

Douglas N Arnold, Franco Brezzi, Bernardo Cockburn, and L Do natella Marini. Uniﬁed analysis of discontinuous Galerkin methods for elliptic problems. SIAM Journal on Numerical Analysis , 39(5):1749– 1779, 2002

2002
[29]

A charact erization of hybridized mixed methods for second order elliptic problems

Bernardo Cockburn and Jayadeep Gopalakrishnan. A charact erization of hybridized mixed methods for second order elliptic problems. SIAM Journal on Numerical Analysis , 42(1):283–301, 2004

2004
[30]

Uniﬁed hybridization of discon- tinuous Galerkin, mixed, and continuous Galerkin methods for secon d order elliptic problems

Bernardo Cockburn, Jayadeep Gopalakrishnan, and Raytcho Lazarov. Uniﬁed hybridization of discon- tinuous Galerkin, mixed, and continuous Galerkin methods for secon d order elliptic problems. SIAM Journal on Numerical Analysis , 47(2):1319–1365, 2009

2009
[31]

A projection-based error analysis of HDG methods

Bernardo Cockburn, Jayadeep Gopalakrishnan, and Francisc o-Javier Sayas. A projection-based error analysis of HDG methods. Mathematics of Computation , 79(271):1351–1367, 2010

2010
[32]

Static condensation, hybridization, and the devising of the HDG methods

Bernardo Cockburn. Static condensation, hybridization, and the devising of the HDG methods. In Building Bridges: Connections and Challenges in Modern App roaches to Numerical Partial Diﬀerential Equations, pages 129–177. Springer, 2016

2016
[33]

A weak Galerkin ﬁnite element method for second-order elliptic problems

Junping Wang and Xiu Ye. A weak Galerkin ﬁnite element method for second-order elliptic problems. Journal of Computational and Applied Mathematics , 241:103–115, 2013

2013
[34]

A weak Galerkin mixed ﬁnite element metho d for second order elliptic problems

Junping Wang and Xiu Ye. A weak Galerkin mixed ﬁnite element metho d for second order elliptic problems. Mathematics of Computation , 83(289):2101–2126, 2014

2014
[35]

Weak Galerkin ﬁnite element met hods for elliptic PDEs

Junping Wang and Chunmei Wang. Weak Galerkin ﬁnite element met hods for elliptic PDEs. Scientia Sinica Mathematica, 45(7):1061–1092, 2015

2015
[37]

The mathematical theory of ﬁnite element methods , volume 15

Susanne C Brenner and Ridgway L Scott. The mathematical theory of ﬁnite element methods , volume 15. Springer Science & Business Media, 2007

2007
[38]

Hybrid discontinuous Galerkin methods f or solving incompressible ﬂow problems

Christoph Lehrenfeld. Hybrid discontinuous Galerkin methods f or solving incompressible ﬂow problems. Rheinisch-Westfalischen Technischen Hochschule Aachen , 2010

2010
[39]

A hybridized discontinuous Galerkin method with re duced stabilization

Issei Oikawa. A hybridized discontinuous Galerkin method with re duced stabilization. Journal of Scientiﬁc Computing , 65(1):327–340, 2015

2015
[40]

Two families of m ixed ﬁnite elements for second order elliptic problems

Franco Brezzi, Jim Douglas, and L Donatella Marini. Two families of m ixed ﬁnite elements for second order elliptic problems. Numerische Mathematik , 47(2):217–235, 1985

1985
[41]

Mixed ﬁnite elements for second order elliptic problems in three variables

Franco Brezzi, Jim Douglas, Ricardo Dur´ an, and Michel Fortin. Mixed ﬁnite elements for second order elliptic problems in three variables. Numerische Mathematik , 51(2):237–250, 1987. 20

1987
[42]

Hybrid ﬁnite element methods for solvin g 2nd order elliptic equations

Pierre-Arnaud Raviart. Hybrid ﬁnite element methods for solvin g 2nd order elliptic equations. Topics in numerical analysis, II (JJM Miller, ed.) , pages 141–155, 1975

1975
[43]

Alternate hybrid stress ﬁnite element models

John P Wolf. Alternate hybrid stress ﬁnite element models. International Journal for Numerical Methods in Engineering , 9(3):601–615, 1975

1975
[44]

Primal hybrid ﬁ nite element methods for 2nd order elliptic equations

Pierre-Arnaud Raviart and Jean-Marie Thomas. Primal hybrid ﬁ nite element methods for 2nd order elliptic equations. Mathematics of Computation , 31(138):391–413, 1977

1977
[45]

John H Argyris. Energy theorems and structural analysis: a g eneralized discourse with applications on energy principles of structural analysis including the eﬀects of tem perature and non-linear stress-strain relations. Aircraft Engineering and Aerospace Technology, 26(10):34...

1954
[46]

Stiﬀness and deﬂection analysis of complex structu res

MJ Turner. Stiﬀness and deﬂection analysis of complex structu res. Journal of the Aeronautical Sciences , 23:805–823, 1956

1956
[47]

Variational methods in mathematical physics , volume 50

Solomon G Mikhlin. Variational methods in mathematical physics , volume 50. Pergamon Press; [dis- tributed by Macmillan, New York], 1964 (original Russian edition: 1957 )

1964
[48]

Approximation variationnelle des probl` emes aux limit es

Jean C´ ea. Approximation variationnelle des probl` emes aux limit es. Ann. Inst. Fourier (Grenoble) , 14(fasc. 2):345–444, 1964

1964
[49]

Derivation of element stiﬀness matrices by as sumed stress distributions

Theodore HH Pian. Derivation of element stiﬀness matrices by as sumed stress distributions. AIAA Journal, 2(7):1333–1336, 1964

1964
[50]

Conforming and no nconforming ﬁnite element methods for solving the stationary Stokes equations I

Michel Crouzeix and Pierre-Arnaud Raviart. Conforming and no nconforming ﬁnite element methods for solving the stationary Stokes equations I. Revue fran¸ caise d’automatique, informatique, recherche op´ erationnelle. Math´ ematique, 7(3):33–75, 1973

1973
[51]

A superc onvergent LDG-hybridizable Galerkin method for second-order elliptic problems

Bernardo Cockburn, Bo Dong, and Johnny Guzm´ an. A superc onvergent LDG-hybridizable Galerkin method for second-order elliptic problems. Mathematics of Computation , 77(264):1887–1916, 2008

1916
[52]

An a priori error analysis of the local discontinuous Galerkin method for elliptic problems

Paul Castillo, Bernardo Cockburn, Ilaria Perugia, and Dominik Sc h¨ otzau. An a priori error analysis of the local discontinuous Galerkin method for elliptic problems. SIAM Journal on Numerical Analysis , 38(5):1676–1706, 2000. 21

2000

Pith tools

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

[1] [36]

Uniform Stability and Error Analysis for Some Discontinuous Galerkin Methods

Qingguo Hong and Jinchao Xu. Uniformly stable results and error estimates for some discontinuous Galerkin methods. arXiv preprint arXiv:1805.09670 , 2018

work page Pith review arXiv 2018

[2] [1]

Solution of problems of elasticity by the fra mework method

Alexander Hrennikoﬀ. Solution of problems of elasticity by the fra mework method. Journal of Applied Mechanics, 8(4):169–175, 1941

work page 1941

[3] [2]

Variational methods for the solution of proble ms of equilibrium and vibrations

Richard Courant. Variational methods for the solution of proble ms of equilibrium and vibrations. Bulletin of the American Mathematical Society , 49(1):1–23, 1943

work page 1943

[4] [3]

A generalization of the direct-stiﬀness method of structural analysis

Robert E Jones. A generalization of the direct-stiﬀness method of structural analysis. AIAA Journal , 2(5):821–826, 1964

work page 1964

[5] [4]

Finite diﬀerence method based on variation principle

Kang Feng. Finite diﬀerence method based on variation principle. Communication on Applied Mathe- matics and Computation , 2(4):237–261, 1965. 18

work page 1965

[6] [5]

Displacement and equilibrium models in the ﬁnite element method

Baudouin Fraeijs de Veubeke. Displacement and equilibrium models in the ﬁnite element method. Stress analysis, pages 145–197, 1965

work page 1965

[7] [6]

A formulation of matrix displacement method

Yoshiyuki Yamamoto. A formulation of matrix displacement method . [Cambridge, Mass.]: Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics , 1966

work page 1966

[8] [7]

On the ﬁnite element method

Miloˇ s Zl´ amal. On the ﬁnite element method. Numerische Mathematik , 12(5):394–409, 1968

work page 1968