The structure of the solution to the generalized Stokes equations and a new method for solving them

Arian Novruzi

REVIEW 2 major objections 3 minor 21 references

The structure of the solution to the generalized Stokes equations and a new method for solving them

T0 review · 2 major / 3 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read The generalized Stokes solution splits into two divergence-free fields, one driven by the force's rotational part and one by a harmonic 'solid pressure', with the pressure itself the sum of a potential and that harmonic term.

desk verdict Genuinely new Stokes decomposition, but the coercivity proof has a real gap and the numerics don't back the performance claims. read the letter →

arxiv 2607.15685 v1 pith:CRDZKIJR submitted 2026-07-17 math.AP

Arian Novruzi This is my paper · ORCID

classification math.AP MSC 35Q3076D0765N3035J47

keywords generalizedStokesequationsHelmholtzdecompositionsolidpressuredivergence-freevectorfieldsboundaryintegralequationself-adjointoperatorNavier-Stokesfiniteelementmethod

topics Navier-Stokes Regularity

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 establishes that every solution of the generalized Stokes equations with Dirichlet boundary data can be written as the sum of two divergence-free vector fields, each solving a Helmholtz-like equation. The first field is driven by the rotational part of the Helmholtz decomposition of the external force; the second is driven by the gradient of a harmonic 'solid pressure' that enforces the normal component of the boundary condition. The fluid pressure correspondingly splits into the potential from the force decomposition plus the solid pressure. This structure yields a numerical method that avoids the inf-sup condition and does not solve the incompressibility constraint directly, instead solving two elliptic equations coupled only at the boundary and one boundary equation for the solid pressure.

What carries the argument

The central object is the Helmholtz decomposition f = ϖ + ∇π, which splits the force into a rotational part that generates one velocity field and a gradient part that is absorbed into the pressure. The proof relies on the space H^1_n(Ω) of functions with zero tangential boundary trace, on a div-curl norm equivalence in that space (with a boundary curvature term), and on a boundary operator A defined by the normal trace of ϑ(q). The operator A is shown to be self-adjoint and coercive, making the boundary equation for q amenable to direct or conjugate-gradient solution.

What would settle it

On a smooth simply connected domain with a known exact solution, discretize the boundary operator A using a Galerkin basis and compute the spectrum of the resulting symmetric matrix; if the smallest eigenvalue approaches zero as the basis is refined, the asserted coercivity bound m > 0 fails and the central theorem would collapse.

Watch

Extended reading notes

Core claim

The velocity solution u is shown to equal ω + ϑ(q), where αω − µΔω = ϖ (with ϖ the divergence-free rotational part of the force) and αϑ − µΔϑ = −∇q (with q harmonic). Both ω and ϑ(q) are divergence-free because of their Helmholtz-like structure and carefully chosen boundary conditions; the sum satisfies the tangential part of the Dirichlet condition for any q. The harmonic 'solid pressure' q is then chosen to satisfy the boundary equation Aq = (ω − g)·n, where Aq = −ϑ(q)·n is a self-adjoint, positive-definite isomorphism from H^{−1/2}_0(∂Ω) to H^{1/2}_0(∂Ω). With that q, u = ω + ϑ(q) and p = π + q solve the full generalized Stokes problem, and q is unique.

Load-bearing premise

The domain must be bounded, simply connected, and have a C^{1,1} boundary; on corner domains like the unit square or cube, the trace lemmas and div-curl norm equivalence used in the proof are not justified, so the numerical demonstrations there go beyond the proven theory.

Editorial extensions

If this is right

A new numerical strategy for generalized Stokes problems: solve two Helmholtz-like vector equations coupled only on the boundary, then solve a symmetric positive-definite boundary equation for the solid pressure.
The method does not rely on the inf-sup condition, so velocity and pressure approximations can be chosen more flexibly.
The same structure applies to the time-dependent Navier-Stokes equations after temporal discretization, and the paper indicates extensions to non-zero divergence and to Neumann or mixed boundary conditions.
The boundary equation can be solved by a Riesz-basis Galerkin method with a proven error bound, or by a conjugate-gradient method with a linear convergence rate depending on the condition number of the operator.

Reading between the lines

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

The 'solid pressure' interpretation suggests a physical picture: q represents the pressure exerted by the boundary on the fluid to enforce the normal velocity condition, which could be useful in fluid-structure interaction or boundary-control settings.
The numerical tests on square and cubic domains indicate the method may work beyond the proven C^{1,1} theory, but a rigorous analysis for Lipschitz or corner domains would be needed to justify that robustness.
Because the method decouples velocity components, it may combine naturally with fast solvers for each scalar Helmholtz equation, but the boundary equation for q could become the computational bottleneck and deserves preconditioning studies.
The same decomposition could provide a constructive way to generate divergence-free basis functions with prescribed tangential boundary data, potentially useful in spectral or meshless methods.

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.

Desk Editor's Note

read the letter

Colleague,

You should know two things about this paper. First, the core idea is genuinely new: writing the Stokes velocity as the sum of a rotational Helmholtz field ω and a harmonic-gradient field θ(q), with q playing the role of a 'solid pressure' that enforces the normal boundary condition, is not in the standard projection/Uzawa literature. Second, the main theorem is not proven as written — there is a real gap in the coercivity proof of the boundary operator A, and the numerical section is far too thin to support the efficiency claims.

What the paper does well: the decomposition is attractive and the structural claim in Theorem 2.1, that u=ω+θ(q) and p=π+q, is interesting. The proofs of Theorems 4.1 and 4.2 are careful, using the space H¹_n(Ω) and the equivalent norm from Lemma 3.6. The boundary operator A is shown to be symmetric and positive; the density argument for surjectivity is also reasonable. For a reader interested in the structure of Stokes solutions, this is worth understanding.

The soft spot is in Theorem 4.3, where the lower bound for A (coercivity) is derived using Lemma 3.8(c). That lemma needs ∫Ω q = 0, but the theorem only guarantees ∫∂Ω q = 0. The harmonic extension of a boundary distribution with zero boundary mean does not generally have zero volume mean — the stress-test note gives a concrete ellipse example. So the divergence-lifting used in (60a) may not exist, and the proof doesn't establish coercivity. This is a load-bearing gap, not a cosmetic one. Without it, Theorem 2.1(c)-(d) are unproven. It might be repairable with a different argument, but as written it doesn't go through.

There are also smaller issues. The claim in (15) that the harmonic extension of a general H^{-1/2}(∂Ω) distribution lies in L²(Ω) needs justification; that's not a standard regularity statement. And the numerical tests are on domains (squares, cubes) outside the C^{1,1} hypotheses, with no code, no convergence study, and no comparison to existing solvers. The reported errors, especially the 50% L∞ pressure error for Taylor–Green, don't justify calling the results 'remarkable precision.'

Bottom line: this is a serious paper with an original idea and a genuine proof gap. I'd send it to peer review, because a good referee could help fix the coercivity argument and the numerics. But as it stands, the main theorem is not established.

Referee Report

2 major / 3 minor

Summary. The paper proposes a decomposition of the solution (u,p) of the generalized Stokes problem (3) as u=ω+ϑ(q), p=π+q, where ω is a divergence-free solution of a vector Helmholtz equation forced by the rotational part of the Helmholtz decomposition of f, and ϑ(q) is a divergence-free solution of a Helmholtz equation forced by −∇q, with q a harmonic 'solid pressure' determined by boundary data. The main theorem (Theorem 2.1) asserts well-posedness of ω and ϑ(q), and that the boundary operator Aq=−ϑ(q)·n is a self-adjoint coercive isomorphism from H^{-1/2}_0(∂Ω) to H^{1/2}_0(∂Ω); the normal boundary condition is then enforced by solving Aq=(ω−g)·n. The proofs of Theorems 4.1 and 4.2 use a weighted H^1_n norm and are largely coherent. Numerical benchmarks in 2D and 3D are reported.

Significance. If the structural theorem is established, the decomposition is a valuable insight: it decouples velocity from pressure, avoids a direct inf-sup treatment, reduces the problem to vector Helmholtz equations coupled only through the boundary, and yields a symmetric positive-definite boundary equation suitable for iterative solvers. The proof strategy is original, and the numerical benchmark errors are encouraging. However, the central coercivity proof contains a real gap in the use of the divergence-lifting lemma, so the main theorem is not yet proven as stated. With a correct proof or a suitable reformulation, the paper could make a solid contribution to the analysis and numerical treatment of generalized Stokes problems.

major comments (2)

[Theorem 4.3, Eq. (60a)] The line (60a) is the load-bearing step for the lower bound in (18b). It applies Lemma 3.8(c) to the harmonic extension q of q∈H^{-1/2}_0(∂Ω). Lemma 3.8(c), however, requires ∫_Ω q dx=0; the construction only guarantees ∫_{∂Ω} q dσ=0. These are not equivalent. On Ω={x^2/4+y^2<1}, q=x^2−y^2−C with C chosen so that ∫_{∂Ω}q dσ=0 is harmonic, lies in H^{-1/2}_0(∂Ω), but ∫_Ω q dx ≈ −9.1≠0. For such q no u_div∈H^1_0(Ω) with div u_div=q exists, so the chain ∥q∥²_{L²}=∫_Ω q div u_div=(ϑ(q),u_div)_{α,μ} cannot be written. Consequently (61) is not established, and Theorems 2.1(c)–(d) are unproven as stated. A fix is to normalize q by ∫_Ω q=0 (equivalently work with q modulo constants), or to prove coercivity via a boundary divergence-free lifting lemma for H^{1/2}_0 normal traces.
[Proposition 5.3, Eq. (74)] Eq. (74) is not the conjugate-gradient gradient update. Since q_{k+1}=q_k−ρ_k w_k, the exact update is g_{k+1}=g_k−ρ_k R^{-1}A w_k = g_k−ρ_k ∂_nD(Aw_k). Replacing w_k by q_k and writing +ρ_k∂_nD(ϑ(q_k)·n) is mathematically unjustified and would not minimize j in general. This is not a one-sign typo: the object entering the normal derivative should be A w_k (equivalently −ϑ(w_k)·n), not A q_k. The numerical results in Section 6 cannot be attributed to the algorithm as printed unless this is corrected.

minor comments (3)

[Section 6, Figs. 1–4] The benchmarks on the unit square and cube, and the Bercovier–Engelman example, are outside the C^{1,1} hypothesis of Theorem 2.1. The text notes this in passing, but it should be stated explicitly that those computations are heuristic demonstrations, or the theory should be extended to Lipschitz polyhedral domains.
[Section 5.2 and Lemma 3.9] Subscripts in (72), (75) and the surrounding text mix H^{-1/2}_0(Ω) and H^{-1/2}_0(∂Ω); Lemma 3.9 also writes H^{1/2}_0(Ω) for a boundary space. These notational inconsistencies should be fixed.
[Section 7 and Fig. 4] The conclusion describes the numerical precision as 'remarkable', but the 3D Taylor–Green case reports a pressure L∞ relative error of 5·10^{-1}. More moderate wording, and at least a mesh-refinement table, are advisable.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the construction is self-contained and does not reduce to its inputs.

full rationale

The paper's central claim is a constructive decomposition: u = ω + ϑ(q), p = π + q, where ω solves (14), ϑ(q) solves (16), and q is chosen to satisfy Aq = (ω − g)·n. The boundary operator A is defined from ϑ(q) and is then proved self-adjoint and coercive in Theorem 4.3 using the weak form (54), the norm equivalence of Lemma 3.6, and standard trace/div-curl estimates. None of these steps assumes the target solution; the construction produces a function and verifies the Stokes equations directly. The Helmholtz decomposition f = ϖ + ∇π is invoked as an external classical result, not as a self-citation, and no fitted parameters are used or renamed as predictions. The numerical benchmarks are compared with explicit known solutions, so they are not circular either. The only substantive concern found is the misuse of Lemma 3.8(c) in the coercivity step (60a): the lemma requires ∫Ω q = 0, while the paper supplies only ∫∂Ω q = 0 for q ∈ H^{-1/2}_0(∂Ω); the ellipse counterexample in the skeptic note shows this is a genuine missing-hypothesis gap in the proof. That is a correctness/rigor issue, not a circularity: it does not make the derivation equivalent to its inputs by construction. Accordingly, no circular step is identified and the circularity score is 0.

Assumptions & free parameters 0 free parameters · 5 assumptions · 1 invented entities

The central claim rests on standard elliptic regularity and div-curl inequalities (all cited), plus the C^{1,1} domain assumption and the zero-flux compatibility condition. There are no fitted free parameters. The 'solid pressure' is a derived variable, not a postulated physical entity.

assumptions (5)

standard math Elliptic regularity results (Grisvard) for (Δ,γ0) and (Δ,γ1) isomorphisms; used in Lemmas 3.1–3.3.
Unproved classical theorems cited to establish trace spaces and integration by parts on C^{1,1} domains.
standard math Div-curl inequality with tangential boundary trace zero (von Wahl [21]); used in Lemma 3.6 to prove norm equivalence.
Pulled from literature; key to coercivity of H^1_{α,μ}.
standard math Well-posedness of Helmholtz-type elliptic problems via Lax-Milgram; used for (14) and (16).
The paper invokes standard Riesz/Lax-Milgram arguments to show existence of ω and ϑ(q).
domain assumption Domain Ω is bounded, simply connected, and C^{1,1}; normal vector admits a C^{0,1} extension.
Assumption in Theorem 2.1; needed for H^1_n space, trace lemmas, and div-curl norm equivalence. If Ω has corners, these lemmas do not apply as stated.
domain assumption Compatibility condition ∫∂Ω g·n dσ = 0 for Dirichlet data of a divergence-free velocity.
Statement of Theorem 2.1; standard necessary condition for existence in incompressible flow.

invented entities (1)

'solid pressure' q
purpose: Harmonic function with boundary values chosen to enforce the normal component of the Dirichlet boundary condition for u=ω+θ(q).
This is a mathematical construct, not a physical entity. It is uniquely determined by the boundary equation Aq=(ω−g)·n, so it is not an ad hoc invention with external predictive power, but it also has no independent falsifiable handle outside the paper's construction.

how reviews work

0 comments

Cite this review

Pith. "Pith review of The structure of the solution to the generalized Stokes equations and a new method for solving them." pith.science (2026). https://pith.science/paper/CRDZKIJR

@misc{pith2026260715685,
  author       = {Pith},
  title        = {Pith review of: The structure of the solution to the generalized Stokes equations and a new method for solving them},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CRDZKIJR}},
  note         = {Machine review of arXiv:2607.15685}
}

abstract

We show that the solution to the generalized Stokes equations with Dirichlet boundary conditions in dimension two or three has a particular structure. Namely, the velocity $u$ is the superposition of two velocity fields $\omega$ and $\theta$, $u=\omega+\theta$, each of them being divergence-free and solving vector Helmholtz-like equations. The right hand sides of these equations involve the rotational component of the Helmholtz decomposition of the external force, and a harmonic function $q$, called solid pressure. Similarly, the fluid pressure $p$ is the superposition of the potential $\pi$ of Helmholtz decomposition of the external force, and of the solid pressure $q$, $p=\pi+q$. It turns out that regardless the solid pressure $q$, $\omega+\theta$ satisfies the generalized Stokes equations, except the normal component of the Dirichlet boundary condition. The role of the solid pressure, which solves a linear boundary equation with a self-adjoint coercive operator, is to constrain $\omega+\theta$ to satisfy even the normal component of the Dirichlet boundary condition. The method is attractive from the numerical viewpoint. It decouples the velocity from the pressure and does not solve the incompressibility constraint directly. It requires only the solution of Helmholtz-like vector equations coupled on the boundary, and of a boundary equation. In the last section we present two methods for solving numerically the generalized Stokes equations, and some results to demonstrate the efficiency of our approach.

Figures

Figures reproduced from arXiv: 2607.15685 by the authors.

**Figure 2.** Results for Kovasznay problem with [PITH_FULL_IMAGE:figures/full_fig_p022_2.png] view at source ↗

**Figure 3.** Results for Bercovier-Engleman with n = 50 (so h = 0.02). Here ∥un − u∥L∞(Ω)/∥u∥L∞(Ω) = 2.7 · 10−5 , ∥pn − p∥L∞(Ω)/∥p∥L∞(Ω) = 5.8 · 10−2 , ∥∇ · un∥L∞(Ω) = 2.9 · 10−2 , ∥∇ · un∥L 2 (Ω) = 3.8 · 10−3 . 6.3 3d Taylor-Green vortex test problem This example is chosen to demonstrate that the method works very well with domains which are not C 1,1 in the case d = 3. In this case we consider the so-called “Taylor-Green vorte… view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: Results for Taylor-Green voxrtex problem with [PITH_FULL_IMAGE:figures/full_fig_p025_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

21 extracted references

[1]

FVCA8 benchmark for the Stokes and Navier-Stokes equations with the TrioCFD code - benchmark session

P.-E. Angeli, M.-A. Puscas, G. Fauchet, A. Cartalade, “FVCA8 benchmark for the Stokes and Navier-Stokes equations with the TrioCFD code - benchmark session”, Finite V olumes for Complex Applications 8, Jun 2017, Lille, France

2017
[2]

Studies in Linear and Nonlinear Programming

K. J. Arrow, L. Hurwicz, H. Uzawa, “Studies in Linear and Nonlinear Programming”, Stan- ford University Press, 1958

1958
[3]

Benchmark proposal for the FVCA8 conference : Finite volume meth- ods for the Stokes and Navier-Stokes equations

F. Boyer, P. Omnes, “Benchmark proposal for the FVCA8 conference : Finite volume meth- ods for the Stokes and Navier-Stokes equations”, Springer Proceedings in Mathematics & 24 0 0.2 0.4 0.6 0.8 1 z Velocity (xz plane) 0 0.2 0.4 0.6 0.8 1 x 0 0.2 0.4 0.6 0.8 1 z Velocity (yz plane) 0 0.2 0.4 0.6 0.8 1 y 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 x y Velocity...

2017
[4]

Functional Analysis, Sobolev Spaces and Partial Differential Equations

H. Brezis, “Functional Analysis, Sobolev Spaces and Partial Differential Equations”, Univer- sitext, Springer, 2011

2011
[5]

Numerical methods for the Navier-STokes equa- tions. Applications to the simulation of compressible and incompressible viscous flows

M. O. Bristeau, R. Glowinski, J. P ´eriaux, “Numerical methods for the Navier-STokes equa- tions. Applications to the simulation of compressible and incompressible viscous flows”, Computer Physics Reports 6 (1987) 73-187 North-Holland, Amsterdam 25

1987
[6]

Numerical Solution of the Navier-Stokes Equations

A. J. Chorin, “Numerical Solution of the Navier-Stokes Equations”, Math. Comp., 22 (104): 745–762, 1968

1968
[7]

An Introduction to Frames and Riesz Bases

O. Christensen, “An Introduction to Frames and Riesz Bases”, Second Edition, Birkh ¨auser, 2016

2016
[8]

Numerical method for convection dominated diffusion prob- lems based on combining the method of characteristics with finite element or finite difference procedures

J. Douglas and T. F. Russell, “Numerical method for convection dominated diffusion prob- lems based on combining the method of characteristics with finite element or finite difference procedures”, SIAM J. Numer. Anal., 19 (1982)

1982

Show all 21 references

[9]

Finite Elements and Fast Iterative Solvers : with Applications in Incompressible Fluid Dynamics: with Applications in Incompressible Fluid Dynamics

H. C. Elman, D. J. Silvester, A. J. Wathen, “Finite Elements and Fast Iterative Solvers : with Applications in Incompressible Fluid Dynamics: with Applications in Incompressible Fluid Dynamics”, Oxford University Press, 2005

2005
[10]

Augmented Lagrangian Methods: Applications to the Numerical Solution of Boundary-Value Problems

M. Fortin, R. Glowinski, “Augmented Lagrangian Methods: Applications to the Numerical Solution of Boundary-Value Problems”, Studies in Mathematics and Its Applications, North- Holland Publishing Co., V ol. 15, 1983

1983
[11]

Elliptic Partial Differential Equations of Second Order

D. Gilbarg, N. S. Trudinger, “Elliptic Partial Differential Equations of Second Order”, V ol- ume 224, Springer Science & Business Media, Jan 12, 2001

2001
[12]

Finite Element Methods for Navier-Stokes Equations. Theory and Algorithms

V . Girault, P.-A. Raviart, “Finite Element Methods for Navier-Stokes Equations. Theory and Algorithms”, Springer-Verlag, Berlin, Heidelberg, 1986

1986
[13]

Numerical Methods for Nonlinear Variational Problems

R. Glowinski, “Numerical Methods for Nonlinear Variational Problems”, Springer-Verlag, New York, 1984

1984
[14]

Elliptic Problems in Nonsmooth Domains

P. Grisvard, “Elliptic Problems in Nonsmooth Domains”, Pitman Publishing Inc. 1020 Plain Street, Marshfield, Massachusetts 02050, 1985

1985
[15]

An overview of projection methods for incompressible flows

J. L. Guermond,P. Minev, J. Shen, “An overview of projection methods for incompressible flows”, Comput. Methods Appl. Mech. Engrg. 195 (2006) 6011–6045

2006
[16]

New development in FreeFem++

F. Hecht, “New development in FreeFem++”, Journal of numerical mathematics, 20.3-4, 251-266, 2012

2012
[17]

Problem `es aux limites non homog `enes

J.-L. Lions, E. Magenes, “Problem `es aux limites non homog `enes”, Annales de l’Institut Fourier, 11, 137-178, 1961

1961
[18]

On the transport diffusion algorithm and its applications to the Navier-Stokes equations

O. Pironneau, “On the transport diffusion algorithm and its applications to the Navier-Stokes equations”, Numer. Math., 38, 309-332 (1982)

1982
[19]

Sur l’approximation de la solution des ´equations de Navier-Stokes par la m´ethode des fractionnaires II

R. Temam, “Sur l’approximation de la solution des ´equations de Navier-Stokes par la m´ethode des fractionnaires II”, Arch. Rational Mech. Anal. 33 (1969), 377-385

1969
[20]

Navier–Stokes equations : theory and numerical analysis

R. Temam, “Navier–Stokes equations : theory and numerical analysis”, AMS Chelsea Pub- lishing, American Mathematical Society, Providence, Rhode Island, 2001

2001
[21]

Estimating∇uby divuand curlu

W. von Wahl, “Estimating∇uby divuand curlu”, Mathematical Methods in the Applied Sciences, V ol. 15, 123-143 (1992) 26

1992

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.

[1] [1]

FVCA8 benchmark for the Stokes and Navier-Stokes equations with the TrioCFD code - benchmark session

P.-E. Angeli, M.-A. Puscas, G. Fauchet, A. Cartalade, “FVCA8 benchmark for the Stokes and Navier-Stokes equations with the TrioCFD code - benchmark session”, Finite V olumes for Complex Applications 8, Jun 2017, Lille, France

2017

[2] [2]

Studies in Linear and Nonlinear Programming

K. J. Arrow, L. Hurwicz, H. Uzawa, “Studies in Linear and Nonlinear Programming”, Stan- ford University Press, 1958

1958

[3] [3]

Benchmark proposal for the FVCA8 conference : Finite volume meth- ods for the Stokes and Navier-Stokes equations

F. Boyer, P. Omnes, “Benchmark proposal for the FVCA8 conference : Finite volume meth- ods for the Stokes and Navier-Stokes equations”, Springer Proceedings in Mathematics & 24 0 0.2 0.4 0.6 0.8 1 z Velocity (xz plane) 0 0.2 0.4 0.6 0.8 1 x 0 0.2 0.4 0.6 0.8 1 z Velocity (yz plane) 0 0.2 0.4 0.6 0.8 1 y 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 x y Velocity...

2017

[4] [4]

Functional Analysis, Sobolev Spaces and Partial Differential Equations

H. Brezis, “Functional Analysis, Sobolev Spaces and Partial Differential Equations”, Univer- sitext, Springer, 2011

2011

[5] [5]

Numerical methods for the Navier-STokes equa- tions. Applications to the simulation of compressible and incompressible viscous flows

M. O. Bristeau, R. Glowinski, J. P ´eriaux, “Numerical methods for the Navier-STokes equa- tions. Applications to the simulation of compressible and incompressible viscous flows”, Computer Physics Reports 6 (1987) 73-187 North-Holland, Amsterdam 25

1987

[6] [6]

Numerical Solution of the Navier-Stokes Equations

A. J. Chorin, “Numerical Solution of the Navier-Stokes Equations”, Math. Comp., 22 (104): 745–762, 1968

1968

[7] [7]

An Introduction to Frames and Riesz Bases

O. Christensen, “An Introduction to Frames and Riesz Bases”, Second Edition, Birkh ¨auser, 2016

2016

[8] [8]

Numerical method for convection dominated diffusion prob- lems based on combining the method of characteristics with finite element or finite difference procedures

J. Douglas and T. F. Russell, “Numerical method for convection dominated diffusion prob- lems based on combining the method of characteristics with finite element or finite difference procedures”, SIAM J. Numer. Anal., 19 (1982)

1982