Pith. sign in

REVIEW 3 major objections 5 minor 24 references

A local Fortin projection for the Scott-Vogelius elements on general meshes

T0 review · 3 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

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

Pith's one-line read For k ≥ 4, the Scott–Vogelius finite element pair admits a local Fortin projection on any shape-regular 2D triangulation, including those with singular vertices.

desk verdict A genuinely useful extension of local Fortin projections to meshes with singular vertices, but the proof's load-bearing boundary-chain case is asserted with "similar", which is a real gap that needs filling before the full-generality claim is solid. read the letter →

arxiv 2512.18033 v2 pith:MOAQMEBK submitted 2025-12-19 math.NA cs.NA

Franziska Eickmann , Johnny Guzmán , Michael Neilan , L. Ridgway Scott , Tabea Tscherpel This is my paper · ORCID
classification math.NAcs.NA MSC 65N3065N1276D07
keywords Scott–VogeliuselementsFortinprojectionlocalsingularverticesdivergence-freefiniteelementanalysisStokesproblemshape-regulartriangulation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper proves that the divergence-free Scott–Vogelius finite element spaces, in two dimensions and for polynomial degree k ≥ 4, have a local Fortin projection on arbitrary shape-regular triangulations, including meshes with singular vertices. The projection is built by correcting a Scott–Zhang-style interpolant with a correction operator obtained from local solves on a carefully chosen decomposition of the domain into patches. If the claim is correct, error analyses for Stokes and Navier–Stokes that need a local, trace-preserving Fortin operator now apply to general meshes, not merely to meshes without singular vertices. The construction also adapts to slip (no-penetration) boundary conditions.

What carries the argument

The load-bearing mechanism is the local discrete Bogovskii operator supplied by Lemma 4: for every patch D in the decomposition and every pressure q supported in D, there exists a zero-trace Scott–Vogelius velocity v with div v = q, support in the enlarged patch P(D), and bound ∥∇v∥_{L^p(P(D))} ≤ C(1 + 1/Θ(D))∥q∥_{L^p(D)}. This surjectivity makes the correction operator Π₂^D well-defined: it is the unique solution of a square degree-of-freedom system that enforces both (3.2a) divergence preservation against the local pressure space and (3.2b) orthogonality to the local divergence-free subspace. Summing these local corrections over the decomposition C_h produces Π₂, and composing with the qua

What would settle it

Assemble the linear system (3.2) on the smallest boundary-singular patch, a single triangle at a boundary vertex (case #T_h(z) = 1), for k = 4 and a pressure q with nonzero triangle mean. If the system matrix is singular, or if the least-squares constant for ∥∇v∥_{L^p(P(D))} ≤ C(1 + 1/Θ(D))∥q∥_{L^p(D)} degenerates faster than 1/Θ(D) as Θ → 0, then Lemma 4 fails and Theorem 7 collapses. A run of meshes refining toward the singular vertex would show this in practice.

Watch

Extended reading notes

Core claim

The central claim is Theorem 7: for k ≥ 4, the operator Π = Π₁ + Π₂(id − Π₁) is a linear projection from W^{1,1}(Ω) onto V_h^k that (i) preserves divergence against every pressure test in Q_h^{k-1}, (ii) satisfies local L^p stability, (iii) has order-h^r approximation, and (iv) preserves the boundary trace of any input that already equals a discrete velocity on ∂Ω. The construction partitions Ω into a decomposition C_h made up of triangles with only non-singular vertices, patches around interior singular vertices, and chains of boundary singular vertices. On each patch D, a local correction operator Π₂^D is defined as the solution of a square system (3.2) that asks for divergence preservatio

Load-bearing premise

The construction rests on Lemma 4, which asserts that on every patch in the decomposition — including patches touching boundary singular vertices — the divergence operator from the zero-trace local Scott–Vogelius space maps onto the local pressure space with stability constant C(1 + 1/Θ(D)); if this local surjectivity fails at some boundary singular patch, the correction operator is not well-defined and the projection as constructed does not exist.

Editorial extensions

If this is right

  • Divergence-preserving Fortin operators now exist for Scott–Vogelius elements on arbitrary shape-regular two-dimensional triangulations, including meshes with singular vertices, for polynomial degree k ≥ 4.
  • The operator preserves inhomogeneous boundary data, so analyses of Stokes and Navier–Stokes with non-homogeneous Dirichlet conditions no longer need to avoid singular vertices or split the mesh artificially.
  • Local stability and approximation properties in L^p for all p ∈ [1,∞] follow, which are the exact ingredients needed for max-norm and quasi-norm error estimates for incompressible flow.
  • The slip-boundary variant (Theorem 11) provides an analogous local Fortin projection for no-penetration conditions, extending the reach to free-surface and friction-type boundary models.
  • On Clough-Tocher and Powell-Sabin split triangulations the degree restriction drops to k ≥ 2 and k ≥ 1 respectively, making the construction available at lower order on those macro-element meshes.

Reading between the lines

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

  • The structural template — a local surjectivity lemma plus a square correction system — looks transferable to other exact-divergence pairs, so the paper may serve as a recipe for building local Fortin projections whenever such a local Bogovskii result exists.
  • The stability constant involves 1/Θ(D), which can become large when a vertex is nearly singular; a natural follow-up is to ask whether this dependence is sharp, and whether local refinement near such vertices can keep the constant under control.
  • Because the projection preserves traces only for inputs that match a discrete function on the boundary, combining it with Nitsche-type treatments of boundary conditions seems a direct next step for inhomogeneous or dynamic boundary data.
  • A concrete computational check would be to assemble the correction system (3.2) on the minimal singular boundary patch (a single triangle at a boundary vertex) for k = 4 and verify the stability bound of Lemma 5 for a family of triangles shrinking toward degeneracy.
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

3 major / 5 minor

Summary. The paper constructs a local Fortin projection for the Scott–Vogelius pair (V^k_h, Q^{k-1}_h), k≥4, on arbitrary shape-regular 2D triangulations, including singular vertices. The construction splits the domain into a collection C_h (non-singular triangles, patches around interior singular vertices, and patches of maximal chains of boundary singular vertices) and defines a correction operator Π2 as a sum of local operators Π2_D obtained from a square dual dof system (3.2). Composing with a trace-preserving Scott–Zhang-type interpolant Π1 gives Π = Π1 + Π2(id − Π1). The paper proves Π is a projection, preserves divergence against Q^{k-1}_h, is locally stable in W^{1,p}, has approximation properties, and preserves boundary data, with a slip-boundary variant. The main analytical ingredients are a local discrete Bogovskii lemma (Lemma 4) and a local stability estimate for Π2_D.

Significance. If the main lemmas were fully proved, the paper would fill a genuine gap: previous local Fortin operators for Scott–Vogelius elements required meshes without singular vertices, while the present construction covers general meshes and gives explicit dependence on the local near-singularity measure Θ. The dof-based construction of Π2_D is elegant and the dimension-counting argument for well-posedness is sound. The stability and approximation results are stated with explicit constants in terms of 1/Θ and local patches, which is exactly what is needed for applications to L∞ error analysis and non-Newtonian flows. The slip-boundary variant broadens the applicability. However, the paper currently relies on unproved assertions in Lemma 4 and Lemma 13; these are not merely cosmetic because they concern the existence of the local surjectivity map underlying every local correction operator.

major comments (3)
  1. [Appendix A, Lemma 4 (Step 2)] The proof of Lemma 4 treats only D = Ω_h(z) for an interior singular vertex and states that the boundary chain case D = M^j_h is 'similar'. This is not a minor omission. For D = M^j_h, the patch is a union of overlapping two-triangle fans along a chain of boundary singular vertices; the mean constraints ∫_T q dx couple along the chain, and the construction of ψ in (A.5)–(A.7) does not transfer directly. Since Lemma 5 and the definition of Π2_D require div ˚V^k_h(D) = Q^{k-1}_h(D) for every D ∈ C_h (2.13), Theorem 7's generality over meshes with boundary singular vertices depends on this missing proof. Please provide the chain construction or a reference containing it.
  2. [Appendix A, Step 1 (A.1)] The first step claims that 'a slight generalization of [15, Lem. 6]' yields w ∈ ˚V^k_h with div w agreeing with q at all vertices, supp w ⊂ P(D), and the L^p bound (A.1). This generalization is not stated or proved. The published lemma does not, as written, cover arbitrary q ∈ Q^{k-1,⊥}_h(D) with support in a boundary chain patch, nor does it give the L^p stability estimate with constant 1 + 1/Θ(D). Because the remainder of Lemma 4 and hence all subsequent results reduce to this existence statement, the argument is incomplete at a load-bearing point. Please state the precise generalization and prove it or give a precise citation with the statement.
  3. [Appendix B, Lemma 13] In the slip-boundary variant, Lemma 13 is the analogue of [15, Lem. 6] and is used as the key input to Lemma 10. The proof splits boundary singular vertices into cases #T_h(z) = 2, 3, 1, but the case #T_h(z) = 1 is dismissed as 'similar' and omitted. A boundary vertex with a single triangle can occur at a corner of a polygonal domain, and the reflection/extension constructions used in the other cases do not immediately apply. Since Lemma 10 and Theorem 11 explicitly cover all D ∈ ˜C_h, this gap leaves the slip-boundary version of the main theorem unsupported for such vertices. Please complete the case or state an assumption excluding it.
minor comments (5)
  1. [Appendix B, Lemma 13, case #T_h(z)=3] The construction of the added triangle T4 is not fully specified: it is not clear which line is connected and whether T4 lies inside or outside Ω. Please clarify why z becomes an interior singular vertex of the enlarged triangulation.
  2. [Lemma 5, proof of (3.5)] The 'scaling argument' for p ∈ [1,∞] is compressed. For p = ∞ the intermediate power h_D^{2(1/p-1)} equals h_D^{-2}, which may confuse readers; writing the two cases p < ∞ and p = ∞ separately would improve readability.
  3. [Theorem 7 (iii)] The constant C_{P(Q(T))} is defined via Θ(P(Q(T))), but P(Q(T)) is a union of patches rather than a single sub-triangulation. Please clarify how Θ is defined for such a union, although the same definition as in (2.4) for M = ∪_{T∈M_h} T can be applied after collecting the triangles in the union.
  4. [Title/header] On page 1 the header reads 'A LOCAL FOR TIN PROJECTION'; correct to 'Fortin'. Also, the name 'Bogovski ˘ ı' in the text before Remark 3 has a stray combining accent.
  5. [Section 2, Definition 1] For a boundary vertex with L = 1 triangle, Θ(z) is defined as a maximum over an empty set. The paper does not state whether such vertices can occur or how they are handled in Lemma 13; please add a clarifying sentence.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the Fortin projection is constructed from explicit local dof conditions, and the load-bearing Lemma 4 rests on an independent published Scott-Vogelius surjectivity result rather than on the target theorem.

full rationale

The derivation chain of the paper is a constructive existence proof, not a fit-then-predict scheme. The Fortin projection Π = Π1 + Π2(v−Π1v) is defined in (3.1)–(3.16); divergence preservation is established by the local dof conditions (3.2a), the patch argument in Lemma 6, and the decomposition of q in the proof of Theorem 7(i). Each of these steps is an explicit algebraic/functional argument and does not assume Theorem 7. The only premise that carries real weight is Lemma 4, the local discrete Bogovskii surjectivity result. Lemma 4 is proven by invoking a 'slight generalization' of [15, Lem. 6] and [15, Prop. 2]. Because [15] (Guzmán–Scott, Math. Comp. 2019) is a published, parameter-free result about the same Scott–Vogelius spaces and does not contain the local Fortin projection, this citation is independent support rather than a self-citation chain that forces the conclusion. The paper is also transparent about the dependence ('whose proof relies on the results in [15]'). However, the proof of Lemma 4 has a genuine completeness gap: Appendix A treats D=Ω_h(z) for interior singular vertices in detail, but for D=M^j_h (a boundary singular chain) it states 'We consider the former case only as the latter is similar' and supplies no construction. This is an omitted proof / correctness risk, but it is not a circular reduction: no equation of the paper identifies Lemma 4 with the theorem being proved, and no fitted parameter is renamed as a prediction. There are no instances of self-definitional quantities, fitted inputs called predictions, imported uniqueness theorems, or renamed known results. Accordingly the circularity score is 0; the noted gap belongs in a correctness assessment, not a circularity assessment.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The paper introduces no fitted constants and no new physical or mathematical entities. Its imported content is prior Scott-Vogelius surjectivity theory plus an unproved generalization of [15, Lem. 6]. The main load-bearing assumption is the local discrete Bogovskii operator (Lemma 4), especially on boundary singular patches.

assumptions (5)
  • domain assumption The family of triangulations {T_h} is conforming and shape-regular (nondegenerate).
    Section 2 assumes this; all stability constants depend on the shape-regularity constant.
  • standard math For k>=4, div of the zero-trace Scott-Vogelius velocity space equals Q^{k-1}_h, i.e., exact divergence surjectivity and inf-sup stability from [15, 19].
    Used in (2.8), (2.13), and Lemma 4; cited as established.
  • standard math There exists a quasi-interpolant Pi_1 satisfying (3.13)-(3.15), e.g., a trace-preserving Scott-Zhang variant from [3, Lem. 5.4.1].
    Stated in Section 3 before Theorem 7; existence is delegated to [3].
  • ad hoc to paper A slight generalization of [15, Lem. 6] provides a local w in the zero-trace Scott-Vogelius space with prescribed vertex divergence values, support in P(D), and stability constant 1+1/Theta(D), for all D in C_h.
    Appendix A Step 1 asserts this without proof; it is load-bearing for Lemma 4 and its full generality is not demonstrated for boundary singular patches.
  • ad hoc to paper In Lemma 13 (slip boundary), the case #T_h(z)=1 at a boundary singular vertex is omitted but asserted to be similar.
    Appendix B leaves this case out; it is needed for the full generality of Theorem 11.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A local Fortin projection for the Scott-Vogelius elements on general meshes." pith.science (2026). https://pith.science/paper/MOAQMEBK

@misc{pith2026251218033,
  author       = {Pith},
  title        = {Pith review of: A local Fortin projection for the Scott-Vogelius elements on general meshes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MOAQMEBK}},
  note         = {Machine review of arXiv:2512.18033}
}
abstract

We construct a local Fortin projection for the Scott-Vogelius finite element pair for polynomial degree $k \ge 4$ on general shape-regular triangulations in two dimensions. In particular, the triangulation may contain singular vertices. In addition to preserving the divergence in the dual of the pressure space, the projection preserves discrete boundary data and satisfies local stability estimates.

Figures

Figures reproduced from arXiv: 2512.18033 by the authors.

Figure 1
Figure 1. A pictorial description of the decomposition of the domain Ch. The compo￾nents of T 1 h , {Ωh(z) : z ∈ S˚2 h }, and {Mj h : 1 ≤ j ≤ N} are depicted in blue, red, and green, respectively. Note that the cardinality of Ch in this example is five. that if z is boundary singular vertex such that z is connected to a y ∈ B j h via a boundary edge path, then z ∈ B j h . For each collection B j h , for j ∈ {1, . . . , N}, we… view at source ↗
Figure 2
Figure 2. Example of a domain D. where bei denotes the standard quadratic edge bubble function with supp(bei ) ⊂ Ti ∪ Ti+1, and ci := R ei bei ds = |ei |/6. Then, we see that vi has support in Ti ∪ Ti+1 and Z Ti div vi dx = 1 ci Z ei bei ds = 1, (A.5) which also shows that Z Ti+1 div vi dx = −1. (A.6) Setting ai := R Ti q dx, for i ∈ {1, . . . , 4} we see that a1 + a2 + a3 + a4 = 0 because q ∈ ˚L 2 (Ω) and supp(q) ⊂ D. We def… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

24 extracted references · 2 linked inside Pith

  1. [1]

    Ainsworth and C

    M. Ainsworth and C. Parker,Unlocking the secrets of locking: finite element analysis in planar linear elas- ticity, Computer Methods in Applied Mechanics and Engineering395(2022), Paper No. 115034, 56

    2022

  2. [2]

    Belenki, L

    L. Belenki, L. C. Berselli, L. Diening, and M. R ˚ uˇ ziˇ cka,On the finite element approximation ofp-Stokes systems, SIAM Journal on Numerical Analysis50(2012), no. 2, 373–397

    2012

  3. [3]

    Bernardi, V

    C. Bernardi, V. Girault, F. Hecht, P.-A. Raviart, and B. Rivi` ere,Mathematics and finite element discretiza- tions of incompressible Navier—Stokes flows, Society for Industrial and Applied Mathematics, 2024

    2024

  4. [4]

    Boffi, F

    D. Boffi, F. Brezzi, and M. Fortin,Finite elements for the Stokes problem, Mixed finite elements, compatibility conditions, and applications (D. Boffi and L. Gastaldi, eds.), Springer, 2008, Lecture Notes in Mathematics. Springer Verlag. Vol. 1939, pp. 45–100

    2008

  5. [5]

    S. C. Brenner and L. R. Scott,The mathematical theory of finite element methods, third ed., Texts in Applied Mathematics, vol. 15, Springer, New York, 2008

    2008

  6. [6]

    Diening, C

    L. Diening, C. Kreuzer, and E. S¨ uli,Finite element approximation of steady flows of incompressible fluids with implicit power-law-like rheology, SIAM J. Numer. Anal.51(2013), no. 2, 984–1015

    2013

  7. [7]

    Diening, M

    L. Diening, M. R ˚ uˇ ziˇ cka, and K. Schumacher,A decomposition technique for John domains, Ann. Acad. Sci. Fenn. Math.35(2010), no. 1, 87–114

    2010

  8. [8]

    Eickmann, L

    F. Eickmann, L. R. Scott, and T. Tscherpel,Scott-Vogelius element and iterated penalty method for inhomo- geneous Dirichlet boundary conditions, arxiv.org/abs/2509.17899

    arXiv

Show all 24 references
  1. [9]

    Eickmann and T

    F. Eickmann and T. Tscherpel,Construction of trace-preserving fortin operators, (in preparation)

  2. [10]

    P. A. Gazca-Orozco, F. Gmeineder, E. Maringov´ a Kokavcov´ a, and T. Tscherpel,A Nitsche method for incompressible fluids with general dynamic boundary conditions, arxiv.org/abs/2502.09550

    arXiv

  3. [11]

    Girault, R

    V. Girault, R. H. Nochetto, and L. R. Scott,Max-norm estimates for Stokes and Navier–Stokes approxima- tions in convex polyhedra, Numerische Mathematik131(2015), no. 4, 771–822

    2015

  4. [12]

    Guzm´ an and D

    J. Guzm´ an and D. Leykekhman,Pointwise error estimates of finite element approximations to the Stokes problem on convex polyhedra, Mathematics of Computation81(2012), no. 280, 1879–1902

    2012

  5. [13]

    Guzm´ an, A

    J. Guzm´ an, A. Lischke, and M. Neilan,Exact sequences on Powell-Sabin splits, Calcolo57(2020), no. 2, Paper No. 13, 25. MR 4076672

    2020

  6. [14]

    Guzm´ an and M

    J. Guzm´ an and M. Neilan,Inf-sup stable finite elements on barycentric refinements producing divergence-free approximations in arbitrary dimensions, SIAM J. Numer. Anal.56(2018), no. 5, 2826–2844. MR 3853609

    2018

  7. [15]

    Guzm´ an and L

    J. Guzm´ an and L. R. Scott,The Scott-Vogelius finite elements revisited, Math. Comp.88(2019), no. 316, 515–529

    2019

  8. [16]

    Jeßberger and A

    J. Jeßberger and A. Kaltenbach,Finite element discretization of the steady, generalized Navier-Stokes equa- tions with inhomogeneous Dirichlet boundary conditions, SIAM J. Numer. Anal.62(2024), no. 4, 1660–1686. Preliminary version – December 23, 2025 12 F. EICKMANN, J. GUZM ´...

    2024

  9. [17]

    Parker and E

    C. Parker and E. S¨ uli,Stability of high-order Scott-Vogelius elements for 2d non-Newtonian incompressible flow, arxiv.org/abs/2509.19488

  10. [18]

    Sauter,The inf-sup constant forhp-Crouzeix-Raviart triangular elements, Comput

    S. Sauter,The inf-sup constant forhp-Crouzeix-Raviart triangular elements, Comput. Math. Appl.149 (2023), 49–70

    2023

  11. [19]

    L. R. Scott and M. Vogelius,Norm estimates for a maximal right inverse of the divergence operator in spaces of piecewise polynomials, RAIRO-Mod´ elisation math´ ematique et analyse num´ erique19(1985), no. 1, 111–143

    1985

  12. [20]

    L. R. Scott and S. Zhang,Finite element interpolation of nonsmooth functions satisfying boundary conditions, Math. Comp.54(1990), no. 190, 483–493

    1990

  13. [21]

    S¨ uli and T

    E. S¨ uli and T. Tscherpel,Fully discrete finite element approximation of unsteady flows of implicitly constituted incompressible fluids, IMA J. Numer. Anal.40(2020), no. 2, 801–849

    2020

  14. [22]

    Tscherpel,Finite Element Approximation for the Unsteady Flow of Implicitly Constituted Incompressible Fluids, D.Phil

    T. Tscherpel,Finite Element Approximation for the Unsteady Flow of Implicitly Constituted Incompressible Fluids, D.Phil. Thesis, University of Oxford, 2018, https://ora.ox.ac.uk/objects/uuid:01b4901c-9705-4087- 80c1-4d656d160aed. AppendixA.Proof of Lemma 4 We first prove the c...

    2018

  15. [23]

    A slight generalization of [15, Lem

    Step:Letq∈Q k−1,⊥ h (D). A slight generalization of [15, Lem. 6] shows that there exists a w∈ ˚V k h (k≥4) such that divw=qon all the verticesS h, suppw⊂P(D), divw∈Q k−1,⊥ h , and ∥∇w∥Lp(P(D)) ≤C 1 Θ(D) + 1 ∥q∥Lp(D).(A.1) Lettingr=q−divw∈Q k−1,⊥ h , we see thatris supported in...

  16. [24]

    IfD=TforT∈ T 1 h then one hasQ k−1 h (T) = Qk−1,⊥ h (T), and thus the result follows from the first step

    Step:Let us considerq∈Q k−1 h (D). IfD=TforT∈ T 1 h then one hasQ k−1 h (T) = Qk−1,⊥ h (T), and thus the result follows from the first step. Hence, we only need to consider whenD= Ω h(z) forz∈ ˚S 2 h andD=M j h for 1≤j≤N. We consider the former case only as the latter is simil...

    2025

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.