pith. sign in

arxiv: 2604.11356 · v1 · submitted 2026-04-13 · 🧮 math.NA · cs.NA

Numerical analysis for the Stokes problem with non-homogeneous Dirichlet boundary condition

Thomas Apel , Katharina Lorenz , Johannes Pfefferer This is my paper

Pith reviewed 2026-05-10 16:06 UTC · model grok-4.3

classification 🧮 math.NA cs.NA
keywords Stokes problemnon-homogeneous Dirichlet boundary conditionsfinite element discretizationdiscretization error estimatescorner singularitiesvery weak solutionsnumerical analysiswell-posedness
0
0 comments X

The pith

Optimal error estimates hold for finite element solutions of the Stokes problem with non-homogeneous Dirichlet data even in non-convex domains or with very rough boundary data.

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

This paper establishes that conforming finite element discretizations of the Stokes problem with approximated non-homogeneous Dirichlet boundary conditions produce optimal discretization error estimates. The analysis incorporates reduced regularity from corner singularities when the domain is non-convex. It also treats boundary data of such low regularity that no weak solution exists by proving well-posedness of the very weak formulation and obtaining matching error bounds. A reader would care because these guarantees apply to numerical fluid simulations without forcing artificial compatibility conditions on the boundary data. Numerical tests confirm the rates.

Core claim

Conforming finite element methods applied to the Stokes problem with non-homogeneous Dirichlet boundary conditions yield optimal discretization error estimates. The estimates remain valid when corner singularities reduce regularity in non-convex domains. For boundary data too irregular to admit a weak solution, the very weak formulation is shown to be well-posed and the corresponding optimal error estimates are derived. Several variants of boundary data approximation are analyzed. The compatibility condition on the data is not required for well-posedness of either formulation.

What carries the argument

Conforming finite element discretizations with the boundary datum approximated in the corresponding trace space, supported by singularity analysis for corners.

If this is right

  • Optimal rates persist when corner singularities reduce solution regularity in non-convex domains.
  • Very weak solutions admit optimal discretization even when boundary data lacks the regularity needed for a weak solution.
  • Multiple boundary data approximation strategies all achieve the optimal rates if the approximation quality is adequate.
  • The continuity equation is satisfied only in the distributional sense when a compatibility condition holds on the boundary data.

Where Pith is reading between the lines

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

  • The same error analysis framework could be applied to other mixed problems with non-homogeneous boundary conditions.
  • Adaptive mesh refinement near corners could be guided by the singularity exponents to restore full rates locally.
  • Very weak formulations may prove useful in settings where boundary data arise from noisy measurements or inverse problems.

Load-bearing premise

The approximation of the boundary datum in the trace space must be sufficiently accurate and the domain geometry must allow standard singularity analysis.

What would settle it

Numerical computations on a non-convex domain with corner singularities that produce convergence rates below the predicted optimal order would show the estimates do not hold.

Figures

Figures reproduced from arXiv: 2604.11356 by Johannes Pfefferer, Katharina Lorenz, Thomas Apel.

Figure 1
Figure 1. Figure 1: Test domains with maximal angle ω = 2 3 π and ω = 3 2 π, resp., with velocity for α = −0.499 The main error estimates of this paper are formulated in Theorems 3.3 and 3.6 for the weak solution and in Theorem 5.1 for the very weak solution. Together with the Remarks 3.7 and 5.2 we find the common error estimate ∥y − yh∥L2(Ω)d ≤ chs+min(t− 1 2 ,k) (6.1) with s = 1 if Ω is convex and s ∈ ( 1 2 , ξ) if Ω is no… view at source ↗
read the original abstract

The Stokes problem with non-homogeneous Dirichlet boundary condition is solved numerically using conforming discretizations and an approximation of the boundary datum in the corresponding trace space. Optimal discretization error estimates are derived. The theory accounts for the influence of corner singularities in the case of a non-convex domain. Several variants of the boundary data approximation are discussed. Moreover, the case of boundary data with very low regularity is studied, where a weak solution does not exist. The well-posedness of the very weak solution is investigated, and optimal discretization error estimates are derived. Numerical tests confirm the theory. The compatibility condition for the boundary data is not necessary for well-posedness of the weak and very weak formulations but it ensures that the solution satisfies the continuity equation in the distributional sense. In the same spirit, the compatibility condition is not necessary for the approximating boundary data; a good approximation of the original boundary data is important.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 2 minor

Summary. The manuscript analyzes the numerical solution of the Stokes equations with non-homogeneous Dirichlet boundary conditions using conforming finite element discretizations. It derives optimal discretization error estimates for both the standard weak formulation and a very weak formulation for low-regularity boundary data. The theory accounts for corner singularities in non-convex domains, discusses variants of boundary datum approximation in the trace space, establishes well-posedness of the very weak solution, and shows that the compatibility condition is not required for well-posedness (though it ensures the divergence-free property distributionally). Numerical tests are included to confirm the rates.

Significance. If the optimal estimates hold with boundary approximation errors properly absorbed, the work would provide a useful extension of standard finite element theory to Stokes problems with rough data and geometric singularities. The treatment of very weak solutions and explicit discussion of approximation quality for the boundary datum are strengths, as is the numerical validation. This could inform practical discretizations in domains where regularity is limited by corners.

major comments (2)
  1. [§4.2, Theorem 4.5] §4.2, Theorem 4.5: The proof that the boundary approximation error is absorbed into the optimal rate (O(h^{min(s,1)}) or similar) does not explicitly bound the contribution from the trace-space projection near corners when the singularity exponent α satisfies α < 1/2; the lifting operator estimate appears to assume sufficient mesh regularity that may not hold without grading.
  2. [§3.3, Eq. (3.12)] §3.3, Eq. (3.12): The well-posedness argument for the very weak formulation uses an extension operator whose norm depends on the domain; for non-convex polygons this constant may grow with the re-entrant angle, potentially degrading the claimed optimality of the discretization error bound in Theorem 4.3.
minor comments (2)
  1. [Abstract] The abstract states that 'several variants of the boundary data approximation are discussed' but does not list them; adding a short enumeration would improve readability.
  2. [Table 1 and Figure 3] Table 1 and Figure 3: The reported rates for the very weak case should be accompanied by the precise theoretical exponent used for comparison (e.g., min(α, k) where α is the singularity index).

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and constructive comments. We address each major comment below and will revise the manuscript to incorporate clarifications where needed.

read point-by-point responses

  1. Referee: [§4.2, Theorem 4.5] The proof that the boundary approximation error is absorbed into the optimal rate (O(h^{min(s,1)}) or similar) does not explicitly bound the contribution from the trace-space projection near corners when the singularity exponent α satisfies α < 1/2; the lifting operator estimate appears to assume sufficient mesh regularity that may not hold without grading.

    Authors: We appreciate this observation. The analysis in Theorem 4.5 relies on the boundedness of the lifting operator from the trace space H^{1/2}(∂Ω) into H^1(Ω), which holds for the polygonal domains considered (with the constant depending only on the fixed domain geometry). The trace-space projection error is controlled in the H^{1/2} norm using standard interpolation estimates that remain valid for α < 1/2, as the boundary datum approximation is independent of the interior mesh grading. However, to make the absorption explicit near corners, we will add a short paragraph in the proof of Theorem 4.5 detailing the local contribution of the projection operator. Our mesh assumptions (Section 2) already allow for quasi-uniform or mildly graded meshes sufficient for the stated rates; no additional grading is required beyond what is assumed. We will revise accordingly. revision: partial

  2. Referee: [§3.3, Eq. (3.12)] The well-posedness argument for the very weak formulation uses an extension operator whose norm depends on the domain; for non-convex polygons this constant may grow with the re-entrant angle, potentially degrading the claimed optimality of the discretization error bound in Theorem 4.3.

    Authors: We agree that the norm of the extension operator in the well-posedness proof (Eq. (3.12)) depends on the domain, including the size of re-entrant angles. This constant is nevertheless independent of the discretization parameter h. Consequently, it enters only the multiplicative prefactor in the error bound of Theorem 4.3 and does not affect the convergence rate, which remains optimal. We will insert a clarifying sentence in Section 3.3 stating that while the constant grows with the re-entrant angle, the h-dependent rates are unaffected. This preserves the claimed optimality. revision: yes

Circularity Check

0 steps flagged

No circularity; estimates extend standard FEM theory to boundary and singularity cases

full rationale

The paper derives optimal discretization error estimates for the Stokes problem with non-homogeneous Dirichlet boundary conditions using conforming finite element discretizations. It accounts for corner singularities on non-convex domains, discusses boundary datum approximations in trace spaces, and analyzes very weak solutions with low regularity. All steps build on established finite element theory and well-posedness results for Stokes problems without reducing any claimed prediction or estimate to a fitted parameter, self-definition, or load-bearing self-citation chain. The compatibility condition discussion and emphasis on good boundary approximation are explicit assumptions rather than tautological derivations. The analysis is self-contained against external benchmarks of standard a priori error theory.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The analysis relies on established mathematical tools for PDEs without introducing new parameters or entities; the main contribution is in the error analysis for specific cases.

axioms (2)
  • standard math Standard assumptions on the domain and function spaces for the Stokes problem
    Invoked for well-posedness of weak and very weak formulations and error estimates.
  • standard math Trace theorems and approximation properties in boundary spaces
    Basis for approximating the non-homogeneous Dirichlet data.

pith-pipeline@v0.9.0 · 5456 in / 1394 out tokens · 56254 ms · 2026-05-10T16:06:02.561332+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

6 extracted references · 6 canonical work pages

  1. [1]

    Amrouche and V

    [AG91] Ch. Amrouche and V. Girault. On the existence and regularity of the solution of Stokes problem in arbitrary dimension.Proc. Japan Acad. Ser. A Math. Sci., 67(5):171–175, 1991.doi:10.3792/pjaa.67.171. [ALN25] Th. Apel, K. Lorenz, and S. Nicaise. Weak and very weak solutions of the Laplace equation and the Stokes system with prescribed regularity.Exa...

    work page doi:10.3792/pjaa.67.171 1991

  2. [2]

    [GM00] M

    doi:10.1051/m2an/2020015. [GM00] M. D. Gunzburger and S. Manservisi. The velocity tracking problem for Navier-Stokes flows with boundary control.SIAM J. Control Optim., 39(2):594–634, 2000.doi:10.1137/S0363012999353771. [GMSZ22] W. Gong, M. Mateos, J. Singler, and Y. Zhang. Analysis and approximations of Dirichlet boundary control of Stokes flows in the e...

    work page doi:10.1051/m2an/2020015 2000

  3. [3]

    [JW09] C

    doi:10.1007/978-3-319-45750-5. [JW09] C. John and D. Wachsmuth. Optimal Dirichlet boundary control of sta- tionary Navier-Stokes equations with state constraint.Numer. Funct. Anal. Optim., 30(11-12):1309–1338, 2009.doi:10.1080/01630560903499001. [KO76] R. B. Kellogg and J. E. Osborn. A regularity result for the Stokes problem in a convex polygon.J. Functi...

    work page doi:10.1007/978-3-319-45750-5 2009

  4. [4]

    Maz’ya and J

    [MR07] V. Maz’ya and J. Rossmann.L p estimates of solutions to mixed boundary value problems for the Stokes system in polyhedral domains.Math. Nachr., 280(7):751–793, 2007.doi:10.1002/mana.200610513. [MR10] V. Maz’ya and J. Rossmann.Elliptic equations in polyhedral domains, vol- ume 162 ofMathematical Surveys and Monographs. American Mathematical Society,...

    work page doi:10.1002/mana.200610513 2007

  5. [5]

    47 [Tem79] R

    doi:10.2307/2008497. 47 [Tem79] R. Temam.Navier-Stokes equations, volume 2 ofStudies in Mathematics and its Applications. North-Holland Publishing Co., Amsterdam-New York, revised edition,

    work page doi:10.2307/2008497

  6. [6]

    Thomasset

    Theory and numerical analysis, With an appendix by F. Thomasset. [TV16] F. Tantardini and A. Veeser. TheL 2-projection and quasi-optimality of Galerkin methods for parabolic equations.SIAM J. Numer. Anal., 54(1):317–340, 2016.doi:10.1137/140996811. [Yse90] H. Yserentant. Two preconditioners based on the multi-level splitting of finite element spaces.Numer...

    work page doi:10.1137/140996811 2016