Pith. sign in

REVIEW 2 major objections 4 minor 1 cited by

Entropy-stable in- and outflow boundary conditions for the compressible Navier-Stokes equations

T0 review · 2 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Entropy-stable in- and outflow boundary conditions are extended from inviscid to viscous compressible flow, with provable nonlinear bounds on mass, energy, and entropy.

desk verdict A credible extension of the Euler in/outflow entropy-stable BCs to Navier-Stokes, but the decisive boundary entropy estimate is imported from [23] rather than proved here. read the letter →

arxiv 2506.21065 v1 pith:PY4VACKC submitted 2025-06-26 math.NA cs.NA

Magnus Svärd , Anita Gjesteland This is my paper
classification math.NAcs.NA MSC 65M0865M1235Q3076N10
keywords entropystabilityboundaryconditionscompressibleNavier-Stokesfinitevolumeaprioriestimatesinflowoutflowsummation-by-partsnonlinear
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 proposes inflow and outflow boundary conditions for the compressible Navier-Stokes equations that are enforced by a boundary data flux vector rather than by setting solution values. It proves that admissible solutions satisfying these conditions have bounded mass, total energy, and entropy over finite time intervals, with density, momentum, pressure, and the logarithmic entropy terms in L∞(0,T;L1(Ω)). The same flux is then built into a node-centred finite-volume scheme using entropy-stable interior fluxes, and the discrete version of the three a priori bounds is shown to hold. The significance is that previously, entropy-stable open-boundary conditions of this type existed for the Euler equations and for far-field Navier-Stokes settings on infinite domains, while finite-domain inlet/outlet conditions for viscous flows lacked a nonlinear stability proof. The result would make boundary closures for viscous compressible simulations a matter of provable nonlinear bounds rather than empirical robustness.

What carries the argument

The load-bearing object is the boundary data flux vector [n1 fᵇ + n2 gᵇ], defined separately for supersonic inflow, subsonic inflow, subsonic outflow, and supersonic outflow, and inserted directly into the continuous boundary condition and into the finite-volume boundary fluxes. The entropy argument uses the entropy variables w and entropy flux potentials ψˣ, ψᵜ to rewrite the boundary term in the entropy balance as wᵀ fᵇ − ψ·n; since fᵇ is the same vector as in the Euler case [23], the right-hand side inherits the bound proven there, while the viscous term contributes non-negative entropy dissipation. In the scheme, the shuffle condition (18) on interior fluxes and the result that the discrete viscous term DIF F ≤ 0 leave only the same boundary expression to control.

What would settle it

Run the proposed finite-volume boundary treatment on a subsonic outflow with an imposed strong boundary layer so that the viscous normal stress in the momentum flux is comparable to the pressure, and check whether the discrete entropy balance (28) stays bounded from below in time; an unbounded boundary term would show that the estimate imported from the Euler case does not transfer.

Watch

Extended reading notes

Core claim

The central claim is that the four boundary conditions (4)-(8), written as a boundary data flux [n1 fᵇ + n2 gᵇ] that depends on the local flow regime, give a priori estimates for the Navier-Stokes initial-boundary value problem. Theorem 2.2 states that every admissible solution has {ρ, ρ|v|², p, ρ log ρ, ρ log T} in L∞(0,T;L¹(Ω)). The boundary conditions are designed so that at a subsonic outflow the momentum condition is p − fᵛ₂ = p_b, a pressure condition modified by the viscous normal stress, which reduces to the Euler pressure condition as viscosity tends to zero. The semi-discrete finite-volume scheme, with entropy-stable interior inviscid fluxes and the entropy-dissipative viscous treatment of [31], inherits the same bounds by reproducing the same boundary terms in its discrete entropy balance.

Load-bearing premise

The proof borrows a boundary entropy bound from the Euler equations, where the boundary flux vector is identical, and leaves the corresponding viscous-case details out.

Editorial extensions

If this is right

  • The proposed boundary conditions can close entropy-stable finite-volume and other entropy-stable discretisations of the Navier-Stokes equations at inlets, outlets, and far-field boundaries of finite domains.
  • Simulations with strong non-smooth features, such as the circular blast wave in the paper, run stably under these boundary conditions without tuning.
  • The subsonic outflow condition is exactly the pressure condition proven linearly well-posed in [29], and the supersonic cases match that theory, so the linear well-posedness gap is confined to subsonic inflow.
  • The boundary treatment composes with entropy-stable no-slip wall conditions, letting a domain carry walls and open boundaries simultaneously while preserving the discrete bounds.

Reading between the lines

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

  • The omitted transfer of the Euler boundary entropy estimate to the viscous boundary pressure term is the point to scrutinise: if the estimate does not survive the replacement of p by p − fᵛ₂ at subsonic outflow, Theorem 2.2 would need modification.
  • Reflection levels near 0.3% for a weak vortex suggest the conditions are nearly transparent at low Mach numbers; whether this degrades with increasing Mach number or vortex strength is a natural testable extension.
  • The discrete bounds require positivity of ρ and p and a conservation-form scheme; relaxing those, as the paper notes for smooth solutions, would trade away the nonlinear estimates for linear ones, which is a design choice users would have to weigh.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 4 minor

Summary. The paper proposes in- and outflow boundary conditions for the compressible Navier-Stokes equations, generalizing the authors' earlier Euler boundary conditions. The new conditions prescribe the total (inviscid minus viscous) normal flux at the boundary in terms of data-dependent flux vectors, with distinct forms for supersonic/subsonic inflow and outflow. The main theoretical result (Theorem 2.2) asserts that admissible solutions satisfy a priori L∞(0,T;L1(Ω)) bounds on density, momentum, pressure, ρ log ρ, and ρ log T, derived from mass, energy, and entropy balances. A semi-discrete node-centered finite-volume scheme is presented that reproduces these estimates for the numerical solution, and numerical experiments with a convected vortex and a blast wave demonstrate robustness. The paper also openly states that linear well-posedness is not established for the subsonic inflow case, and that the decisive boundary entropy estimate is imported from the companion Euler paper [23] without a full derivation.

Significance. If the main theorem is correct, the paper provides a practically useful route to nonlinear a priori estimates for compressible Navier-Stokes initial-boundary value problems and to entropy-stable numerical schemes with open boundaries. The boundary conditions are new, conceptually simple, and they reduce to previously derived Euler conditions in the inviscid limit. The finite-volume construction is thoughtful and the numerical tests indicate that the method is stable in several nontrivial flow configurations. However, the central entropy estimate is not self-contained: it relies on an unstated lemma from [23], and the paper does not verify that the hypotheses of that lemma hold for the Navier-Stokes boundary state. Because this estimate is load-bearing for Theorem 2.2 and for the discrete stability claim, the paper is not yet ready for publication in its present form.

major comments (2)
  1. [Section 2.6, Eq. (15)] The entropy bound is not proven self-containedly. The proof asserts that the boundary integrand w^T f^b - (ψ_x, ψ_y)·n is "exactly the same" as (30) in [23] and omits the remaining details. However, the algebraic identity of f^b alone is not sufficient to transfer the boundedness argument. In the present problem the boundary condition (4) prescribes the total flux, so for an x-normal subsonic outflow the boundary state satisfies p - f^V_2 = p_b (Section 2.5), not p = p_b as in the Euler case. The entropy variables w and the potentials ψ are evaluated at the Navier-Stokes state, and the paper does not verify that the estimate in [23] uses only positivity, (12), (14), and the algebraic form of f^b. Without this verification, Theorem 2.2 is not established. Please supply the full argument or state and prove the required lemma explicitly.
  2. [Section 3.3.3, Eq. (29)] The semi-discrete entropy stability claim inherits the same gap. The boundary terms in (29) are asserted to be bounded below "as shown in [23]", but the discrete boundary flux (19) also imposes the total flux, and the discrete entropy variables are evaluated at the numerical solution, which satisfies a different pressure-data relation than in the Euler case. The same missing verification therefore affects the semi-discrete a priori bounds, and the proof should be completed or the needed result stated and proved.
minor comments (4)
  1. [Section 2.6, Eq. (15)] The notation switches from lowercase ψ_x, ψ_y (defined just above) to capital Ψ_x, Ψ_y in (15); please use a consistent notation throughout.
  2. [Section 3.3.3, Eq. (28)] The summation in the entropy balance is written as "NX i=N", which appears to be a typo; it should presumably be the sum over all grid points i ∈ N.
  3. [Section 4.1, Table 2] The numerical experiments measure reflections against a freestream solution but do not include a manufactured-solution grid-convergence study. The current tests show stability, but an accuracy verification would substantially strengthen the numerical section.
  4. [Section 2.7] The paper transparently notes that linear well-posedness is not established for subsonic inflow. Given that this is one of the main boundary types, it would be helpful to briefly discuss the practical implications and whether the alternative supersonic-inflow data flux is actually recommended in that case.

Circularity Check

0 steps flagged · score 2.0 of 10

No construction-level circularity: mass/energy/entropy derivation is self-contained except that the decisive entropy boundary-term bound is imported from the author's Euler paper [23] with "we omit the remaining details", a self-containment and correctness risk rather than a circular reduction.

full rationale

The paper contains no fitted parameters, no parameter renamed as a prediction, and no definitional identification of input with output. The proposed boundary conditions (4)-(8) are genuinely new for the Navier-Stokes system, with the viscous flux entering the total-flux condition; Section 2.5 shows they reduce to the known pressure conditions at constant-x boundaries. The mass and energy bounds (11)-(14) are derived in the paper from positivity, data bounds, and the structure of the boundary fluxes. The entropy estimate (15) is the only step with a self-citation issue: the proof states "We note that f^b is exactly the same as in [23] ... the right-most integrand is exactly the same as (30) in [23] where, with the help of (12) and (14), it was shown to be bounded. We omit the remaining details." This is load-bearing for Theorem 2.2 and for the semi-discrete bound (29), and it is an omitted proof: [23] treated Euler boundary conditions, whereas here (4) prescribes the total flux n1(f^I-f^V)+n2(g^I-g^V), and Section 2.5 gives p-f^V_2=p_b at a subsonic outflow, so the boundary state is not the Euler state. The paper does not verify that the [23] estimate depends only on (12), (14), and positivity. This is a genuine correctness/self-containment gap, but it is not circular in the sense of the present framework: [23] is a separate, parameter-free result whose stated assumptions do not include the target Navier-Stokes theorem, and the current claim does not reduce to it by construction. Section 2.7 also explicitly concedes that linear well-posedness for subsonic inflow is not proven; that is a limitation, not circularity. The numerical experiments are illustrative robustness checks, not predictions fitted from the theory.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The derivation assumes positivity and smoothness of data, and imports two substantive estimates from earlier papers by the same group: the Euler boundary entropy bound [23] and the logarithmic bounds [21]. These are standard-style assumptions for this research program, not fitted parameters. No new physical entities are introduced.

assumptions (6)
  • domain assumption Positivity of thermodynamic variables: p, ρ, T > 0 at all times.
    Section 2.1 restricts to admissible solutions with p, ρ, T > 0; the proof uses positivity without proving it.
  • domain assumption Boundary data regularity and positivity: ρb, pb, vbτ bounded in H^1(∂Ω×(0,T]), ρb, pb, Tb ≥ ε > 0.
    Section 2.1 imposes this on data; the mass and energy estimates use bounds on the data.
  • domain assumption Data continuity and compatibility across boundary type changes: ρb and pb must be continuous in space and time and independent of local flow conditions.
    Section 2.2 argues this is needed for an unambiguous split into subsonic/supersonic inflow/outflow.
  • ad hoc to paper The Euler boundary entropy estimate (Eq. (30) in [23]) applies unchanged to the Navier-Stokes boundary data flux.
    Section 2.6: 'the right-most integrand is exactly the same as (30) in [23]' and 'we omit the remaining details'. This is the main imported result.
  • standard math Viscous dissipation is non-negative: ∫(w_x^T f^V + w_y^T g^V) ≥ 0, and the discrete analog DIFF ≤ 0.
    Section 2.6 and 3.3.3 cite [31]; this is a standard property under Stokes' hypothesis for the chosen viscous discretization.
  • ad hoc to paper Logarithmic bounds: positivity plus mass/energy/entropy bounds imply bounds on ρ log ρ and ρ log T as in [21].
    Section 2.6: 'follow in the same way as in [21]'; another imported estimate.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Entropy-stable in- and outflow boundary conditions for the compressible Navier-Stokes equations." pith.science (2026). https://pith.science/paper/PY4VACKC

@misc{pith2026250621065,
  author       = {Pith},
  title        = {Pith review of: Entropy-stable in- and outflow boundary conditions for the compressible Navier-Stokes equations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PY4VACKC}},
  note         = {Machine review of arXiv:2506.21065}
}
read the original abstract

We propose inflow and outflow boundary conditions for the compressible Navier-Stokes equations and prove that they allow a priori estimates of the entropy, mass and total energy. Furthermore, we demonstrate how to approximate these boundary conditions in conjunction with an entropy-stable finite-volume scheme. The method is also applicable to other types of entropy-stable schemes. Finally, we carry out some numerical computations with the finite-volume scheme and demonstrate their robustness.

Figures

Figures reproduced from arXiv: 2506.21065 by the authors.

Figure 1
Figure 1. Depiction of a grid. The solid lines mark the primary grid and [PITH_FULL_IMAGE:figures/full_fig_p015_1.png] view at source ↗
Figure 2
Figure 2. The remains of a vortex that has exited the domain in a diagonal [PITH_FULL_IMAGE:figures/full_fig_p028_2.png] view at source ↗
Figure 3
Figure 3. Vortex entering the computational domain. [PITH_FULL_IMAGE:figures/full_fig_p029_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: The circular blast wave computed with µ = 0.0001. 29 [PITH_FULL_IMAGE:figures/full_fig_p029_4.png]

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Entropy stable finite difference methods via entropy correction artificial viscosity and knapsack limiting

    math.NA 2025-08 conditional novelty 6.0 of 10

    Two finite difference schemes, one using entropy-correction artificial viscosity and one using knapsack limiting, satisfy discrete entropy inequalities; the knapsack variant also provably preserves positivity for the ...

Reference graph

Works this paper leans on

32 extracted references · 32 canonical work pages · cited by 1 Pith paper

  1. [23]

    M. Sv¨ ard. Entropy-stable in- and outflow boundary conditions for the compressible Euler equations. Journal of Computational Physics , 521, 2025

    work page 2025

  2. [1]

    M. H. Carpenter, T. Fisher, E. Nielsen, M. Parsani, M. Sv¨ ard, and N. Yamaleev. Entropy stable summation-by-parts formulations for com- pressible computational fluid dynamics. In R. Abgrall and C. Shu, ed- itors, Handbook of Numerical Analysis , volume 17, chapter 10, pages 495–524. Elsevier, 2016

    work page 2016

  3. [2]

    J. Chan, Y. Lin, and T. Warburton. Entropy stable modal discontinu- ous Galerkin schemes and wall boundary conditions for the compressible Navier-Stokes equations. Journ. Comput. Phys. , 448, 2022

    work page 2022

  4. [3]

    Chandrashekar

    P. Chandrashekar. Kinetic energy preserving and entropy stable finite volume schemes for compressible Euler and Navier-Stokes equations. Comm. Comp. Phys. , 14(5):1252–1256, 2013

    work page 2013

  5. [4]

    Erlebacher, M

    G. Erlebacher, M. Y. Hussaini, and C.-W. Shu. Interaction of a shock with a longitudinal vortex. J. Fluid. Mech. , 337:129–153, 1997

    work page 1997

  6. [5]

    Eymard, T

    R. Eymard, T. Gallou¨ et, and R. Herbin. Finite volume methods. In P. Ciarlet and J. Lions, editors, Handbook of Numerical Analysis , vol- ume 7, pages 713–1020. Elsevier, 1997

    work page 1997

  7. [6]

    D. C. D. R. Fern´ andez, J. E. Hicken, and D. W. Zingg. Review of summation-by-parts operators with simultaneous approximation terms for the numerical solution of partial differential equations. Computers & Fluids , 95:171–196, 2014

    work page 2014

  8. [7]

    Gjesteland and M

    A. Gjesteland and M. Sv¨ ard. Entropy stability for the compressible Navier-Stokes equations with strong imposition of the no-slip boundary condition. J. Comput. Phys. , 270, 2022. 25

    work page 2022

Show all 32 references
  1. [8]

    Gustafsson and A

    B. Gustafsson and A. Sundstr¨ om. Incompletely parabolic systems in fluid dynamics. SIAM J. Appl. Math. , 1978

    1978

  2. [9]

    Gustafsson, H.-O

    B. Gustafsson, H.-O. Kreiss, and J. Oliger. Time dependent problems and difference methods. John Wiley & Sons, Inc., 1995

    1995

  3. [10]

    Hesthaven and D

    J. Hesthaven and D. Gottlieb. A stable penalty method for the com- pressible Navier-Stokes equations: I open boundary conditions. SIAM J.Sci.Comput., 1996

    1996

  4. [11]

    T. Y. Hou and P. G. L. Floch. Why nonconservative schemes con- verge to wrong solutions: error analysis. Math. Comp. , 62(206):497– 530, 1994

    1994

  5. [12]

    Ismail and P

    F. Ismail and P. Roe. Affordable, entropy-consistent euler flux functions II: Entropy production at shocks. J. Comp. Phys. , pages 5410–5436, 2009

    2009

  6. [13]

    Lax and B.Wendroff

    P. Lax and B.Wendroff. System of conservation laws. Comm. on Pure and Applied Math. , XIII, 1960

    1960

  7. [14]

    Nordstr¨ om

    J. Nordstr¨ om. The Use of Characteristic Boundary Conditions for the Navier-Stokes Equations. Computers & Fluids , 24(5):609–623, 1995

    1995

  8. [15]

    Nordstr¨ om and M

    J. Nordstr¨ om and M. Sv¨ ard. Well Posed Boundary Conditions for the Navier-Stokes Equations. SIAM Journal on Numerical Analysis , 43(3): 1231–1255, September 2005

    2005

  9. [16]

    Nordstr¨ om, K

    J. Nordstr¨ om, K. Forsberg, C. Adamsson, and P. Eliasson. Finite vol- ume methods, unstructured meshes and strict stability for hyperbolic problems. Applied Numerical Mathematics , 45(4):453–473, June 2003

    2003

  10. [17]

    Olsson and J

    P. Olsson and J. Oliger. Energy and maximum norm estimates for non- linear conservation laws. Technical report 94.01, The Research Institute of Advanced Computer Science, Jan. 1994

    1994

  11. [18]

    Parsani, M

    M. Parsani, M. Carpenter, and E. Nielsen. Entropy stable wall bound- ary conditions for the three-dimensional compressible Navier-Stokes equations. J. Comput. Phys. , 292:88–113, 2015

    2015

  12. [19]

    G. Strang. Accurate partial difference methods II. non-linear problems. Num. Math. , 6:37–46, 1964

    1964

  13. [20]

    M. Sv¨ ard. Entropy stable boundary conditions for the Euler equations. J. Comput. Phys. , 426:1–17, 2021. 26

    2021

  14. [21]

    M. Sv¨ ard. Analysis of an alternative Navier-Stokes system: Weak en- tropy solutions and a convergent numerical scheme. M3AS, 32(13): 2601–2671, 2022

    2022

  15. [22]

    M. Sv¨ ard. A convergent finite-volume scheme for the diffusive com- pressible Euler model – resolved and under-resolved flows. Journal of Computational physics , 518, 2024

    2024

  16. [24]

    Sv¨ ard and A

    M. Sv¨ ard and A. Gjesteland. Entropy stable far-field boundary condi- tions for the compressible Navier-Stokes equations. J. Comput. Phys. , 508, 2024

    2024

  17. [25]

    Sv¨ ard and S

    M. Sv¨ ard and S. Mishra. Entropy stable schemes for initial-boundary- value conservation laws. ZAMP, 63:985–1003, 2012

    2012

  18. [26]

    Sv¨ ard and J

    M. Sv¨ ard and J. Nordstr¨ om. Stability of finite volume approximations for the Laplacian operator on quadrilateral and triangular grids.Applied Numerical Mathematics, 51(1):101–125, October 2004

    2004

  19. [27]

    Sv¨ ard and J

    M. Sv¨ ard and J. Nordstr¨ om. Review of summation-by-parts schemes for initial–boundary-value problems. J. Comput. Phys. , 268:17–38, 2014

    2014

  20. [28]

    Sv¨ ard and H

    M. Sv¨ ard and H. ¨Ozcan. Entropy stable schemes for the Euler equa- tions with far-field and wall boundary conditions. Journal of Scientific Computing, 58(1):61–89, 2014

    2014

  21. [29]

    Sv¨ ard, M

    M. Sv¨ ard, M. Carpenter, and J. Nordstr¨ om. A stable high-order fi- nite difference scheme for the compressible Navier-Stokes equations, far-field boundary conditions. Journal of Computational Physics , 225: 1020–1038, 2007

    2007

  22. [30]

    Sv¨ ard, J

    M. Sv¨ ard, J. Gong, and J. Nordstr¨ om. An accuracy evaluation of unstructured node-centred finite volume methods. Applied Numerical Mathematics, 58:1142–1158, 2008

    2008

  23. [31]

    Sv¨ ard, M

    M. Sv¨ ard, M. Carpenter, and M. Parsani. Entropy stability and the no-slip wall boundary condition. SINUM, 56(1):256–273, 2018

    2018

  24. [32]

    E. Tadmor. Entropy stability theory for difference approximations of nonlinear conservation laws and related time-dependent problems. Acta Numerica, pages 451–512, 2003. 27 (a) ρ at t = 0, N = 100 (b) ρ at t = 15.0, N = 100 (c) ρ at t = 15.0, N = 200 (d) ρ at t = 15.0, N = 400...

    2003

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.