pith. sign in

arxiv: 2510.00779 · v5 · pith:VRP7S73Enew · submitted 2025-10-01 · ⚛️ physics.flu-dyn · physics.comp-ph

Kinetic closure of turbulence

Francesco Marson , Orestis Malaspinas This is my paper

Pith reviewed 2026-05-21 21:03 UTC · model grok-4.3

classification ⚛️ physics.flu-dyn physics.comp-ph
keywords turbulence closurefiltered Boltzmann equationBGK collision operatorsubfilter stress tensorChapman-Enskog analysislarge-eddy simulationTaylor-Green vortex
0
0 comments X

The pith

A generalized BGK collision operator closes the filtered Boltzmann equation for turbulence without a separate Smagorinsky model for the stress tensor.

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

The paper develops a closure for the filtered Boltzmann-BGK equation that keeps the full turbulent subfilter stress tensor intact rather than replacing it with an empirical structure. Subfilter turbulent diffusion is instead handled by modifying the collision operator itself to act on the nonconserved moments. The approach dispenses with any requirement for scale separation between resolved and unresolved motions. Chapman-Enskog expansion recovers the filtered Navier-Stokes equations in the hydrodynamic limit, with subfilter effects isolated in the velocity gradients. Direct numerical tests on the Taylor-Green vortex and a turbulent mixing layer indicate the resulting scheme remains stable while dissipating less energy than the classical Smagorinsky closure.

Core claim

The central claim is that a suitably generalized BGK collision operator supplies a kinetic closure of the filtered Boltzmann equation. This closure retains the turbulent subfilter stress tensor without imposing a Smagorinsky-type ansatz on its form and simultaneously accounts for subfilter turbulent diffusion in the nonconserved moments. Because no scale separation is assumed, the Chapman-Enskog analysis yields the filtered Navier-Stokes equations in which velocity gradients directly isolate the subfilter contributions. The resulting model therefore provides an alternative route to turbulence description whose hydrodynamic limit and numerical behavior are demonstrated on standard benchmarkes

What carries the argument

The generalization of the BGK collision operator that encodes subfilter turbulent diffusion directly into the evolution of the nonconserved moments while leaving the stress tensor unmodeled.

If this is right

  • The hydrodynamic limit recovers the filtered Navier-Stokes equations with subfilter effects appearing through velocity gradients.
  • No scale separation between resolved and unresolved scales is required for the closure to be applied.
  • Numerical simulations of the Taylor-Green vortex and turbulent mixing layer exhibit improved stability and lower dissipation than the Smagorinsky model.
  • The subfilter stress tensor is retained in its original form rather than replaced by an auxiliary model.

Where Pith is reading between the lines

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

  • The method could be tested in flows with strong intermittency or near walls where classical closures often require ad-hoc adjustments.
  • Extension to other collision operators beyond BGK might preserve the same closure structure while altering the underlying kinetic relaxation.
  • Because the stress tensor is not modeled separately, the approach may reduce the number of tunable parameters in large-eddy simulations of complex geometries.

Load-bearing premise

The generalized BGK operator is assumed to capture subfilter turbulent diffusion in the nonconserved moments without introducing uncontrolled errors or demanding further empirical adjustments.

What would settle it

Compare the subfilter stress tensor predicted by the closed kinetic model against filtered DNS data for a high-Reynolds-number flow in which clear scale separation is absent; systematic deviation would falsify the closure.

Figures

Figures reproduced from arXiv: 2510.00779 by Francesco Marson, Orestis Malaspinas.

Figure 1
Figure 1. Figure 1: FIG. 1: Turbulent energy spectrum and the general [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2: Representation of the Hilbert space of [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4: Velocity self similarity in the ML test-case. [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
read the original abstract

This letter presents a kinetic closure of the filtered Boltzmann--BGK equation, paving the way toward an alternative description of turbulence. The closure retains the turbulent subfilter stress tensor without a separate Smagorinsky-type ansatz for its structure, unlike classical filtered Navier--Stokes closures. In contrast, it accounts for the subfilter turbulent diffusion in the nonconserved moments by generalizing the BGK collision operator. The model does not require scale separation between resolved and unresolved scales. The Chapman--Enskog analysis shows how its hydrodynamic limit can converge to the filtered Navier--Stokes equations, with velocity gradients isolating subfilter contributions. Numerical tests on the Taylor--Green vortex and the turbulent mixing layer show improved stability and reduced dissipation in the reported cases, benchmarked against the Smagorinsky model.

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

1 major / 2 minor

Summary. The manuscript presents a kinetic closure of the filtered Boltzmann-BGK equation for turbulence. It retains the subfilter stress tensor without a separate Smagorinsky-type ansatz and accounts for subfilter turbulent diffusion by generalizing the BGK collision operator. The model claims no requirement for scale separation between resolved and unresolved scales. Chapman-Enskog analysis is used to show that the hydrodynamic limit converges to the filtered Navier-Stokes equations, with velocity gradients isolating subfilter contributions. Numerical tests on the Taylor-Green vortex and turbulent mixing layer report improved stability and reduced dissipation relative to the Smagorinsky model.

Significance. If the closure and its hydrodynamic limit are rigorously justified, the approach could provide a useful alternative to classical filtered Navier-Stokes closures by avoiding empirical structure assumptions for the subfilter stress. The numerical results on canonical flows offer preliminary evidence of practical advantages in stability and dissipation control.

major comments (1)
  1. [Abstract and Chapman-Enskog analysis] Abstract: The claim that the model does not require scale separation is in direct tension with the invocation of Chapman-Enskog analysis to establish convergence to filtered Navier-Stokes. Standard Chapman-Enskog expands in powers of the Knudsen number Kn = λ/L with L set by the filter width; this is an asymptotic procedure valid only for Kn ≪ 1. When the model advertises applicability precisely in the regime without such separation, the neglected O(Kn) and higher terms are not demonstrably small, undermining the analytical support for the central convergence claim. The derivation section must either derive the limit without invoking the small-Kn assumption or supply explicit remainder estimates.
minor comments (2)
  1. [Numerical tests] Numerical section: quantitative metrics (e.g., kinetic energy decay rates, enstrophy spectra) and grid/filter-width ratios should be reported with error bars or ensemble statistics to allow direct comparison with Smagorinsky results.
  2. [Model formulation] Notation: define the generalized BGK relaxation time or frequency explicitly and distinguish it from the molecular collision frequency to avoid ambiguity in the non-conserved moment equations.

Simulated Author's Rebuttal

1 responses · 1 unresolved

We thank the referee for the careful reading and constructive comments on our manuscript. The major comment identifies a genuine point of tension that merits clarification, which we address below. We have revised the manuscript to improve the presentation of the Chapman-Enskog analysis and its relation to the no-scale-separation claim.

read point-by-point responses

  1. Referee: [Abstract and Chapman-Enskog analysis] The claim that the model does not require scale separation is in direct tension with the invocation of Chapman-Enskog analysis to establish convergence to filtered Navier-Stokes. Standard Chapman-Enskog expands in powers of the Knudsen number Kn = λ/L with L set by the filter width; this is an asymptotic procedure valid only for Kn ≪ 1. When the model advertises applicability precisely in the regime without such separation, the neglected O(Kn) and higher terms are not demonstrably small, undermining the analytical support for the central convergence claim. The derivation section must either derive the limit without invoking the small-Kn assumption or supply explicit remainder estimates.

    Authors: We agree there is an apparent tension requiring clarification. The Chapman-Enskog analysis is applied in the conventional manner to demonstrate that the hydrodynamic limit of the closed kinetic equation recovers the filtered Navier-Stokes equations, with subfilter contributions appearing through the velocity gradients. However, the kinetic closure itself is formulated directly at the level of the filtered Boltzmann-BGK equation without invoking scale separation: the subfilter stress tensor is retained exactly, and subfilter diffusion is incorporated by generalizing the BGK operator. This structure permits use of the model even when Kn is not small, as the full kinetic description is solved rather than its hydrodynamic limit. The CE step serves only to establish formal consistency in the scale-separated regime. In the revised manuscript we have updated the abstract and expanded the derivation section to explicitly distinguish the model's general applicability from the assumptions of the CE expansion, noting that higher-order terms are neglected in the standard way. We have not derived the limit without the small-Kn assumption or supplied explicit remainder estimates. revision: partial

standing simulated objections not resolved
  • Deriving the hydrodynamic limit without the small-Knudsen-number assumption or supplying explicit remainder estimates for the Chapman-Enskog expansion.

Circularity Check

0 steps flagged

No circularity: closure and hydrodynamic limit derived from independent generalization and standard analysis

full rationale

The paper defines a generalization of the BGK collision operator to incorporate subfilter turbulent diffusion in nonconserved moments, then applies Chapman-Enskog expansion to recover the filtered Navier-Stokes limit with subfilter stress retained explicitly. No quoted step reduces the target equations or numerical outcomes to a fitted parameter renamed as prediction, nor does any load-bearing premise collapse to a self-citation chain or ansatz smuggled from prior work by the same authors. The claim of no required scale separation is stated separately from the analysis; the derivation remains self-contained against the stated kinetic closure and external benchmarks such as Smagorinsky comparisons.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The approach relies on standard kinetic theory tools and a posited generalization of the BGK operator whose precise form is not detailed in the abstract; no explicit free parameters or new entities are named.

axioms (1)
  • standard math Chapman-Enskog expansion remains valid for deriving the hydrodynamic limit of the filtered kinetic equation
    Invoked to show convergence to filtered Navier-Stokes equations with velocity gradients isolating subfilter contributions.

pith-pipeline@v0.9.0 · 5656 in / 1372 out tokens · 35089 ms · 2026-05-21T21:03:56.481846+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

52 extracted references · 52 canonical work pages

  1. [1]

    S.B.Pope,Turbulent flows(CambridgeUniversityPress, Cambridge; New York, 2000)

    work page 2000

  2. [2]

    Davidson,Turbulence: An Introduction for Scientists and Engineers, 2nd ed

    P. Davidson,Turbulence: An Introduction for Scientists and Engineers, 2nd ed. (Oxford University Press, Ox- ford, United Kingdom; New York, NY, United States of America, 2015)

    work page 2015

  3. [3]

    Reynolds, Philosophical Transactions of the Royal So- ciety of London

    O. Reynolds, Philosophical Transactions of the Royal So- ciety of London. (A.)186, 123 (1895)

    work page

  4. [4]

    Smagorinsky, Monthly weather review91, 99 (1963)

    J. Smagorinsky, Monthly weather review91, 99 (1963)

    work page 1963

  5. [5]

    Lilly,The representation of small-scale turbulence in numerical simulation experiments, Tech

    D. Lilly,The representation of small-scale turbulence in numerical simulation experiments, Tech. Rep. (1966)

    work page 1966

  6. [6]

    Sagaut,Large eddy simulation for incompressible flows: an introduction, 3rd ed., Scientific computation (Springer, Berlin; New York, 2006)

    P. Sagaut,Large eddy simulation for incompressible flows: an introduction, 3rd ed., Scientific computation (Springer, Berlin; New York, 2006)

    work page 2006

  7. [7]

    A. N. Kolmogorov, Doklady Akademii Nauk SSSR30, 299 (1941)

    work page 1941

  8. [8]

    C.-L. M. H. Navier, Mémoires de l’Académie Royale des Sciences de l’Institut de France6, 389 (1822)

    work page

  9. [9]

    G. G. Stokes, Transactions of the Cambridge Philosoph- ical Society8, 287 (1845)

    work page

  10. [10]

    Rotta, Zeitschrift für Physik129, 547 (1951)

    J. Rotta, Zeitschrift für Physik129, 547 (1951)

    work page 1951

  11. [11]

    J. H. Ferziger, M. Perić, and R. L. Street,Computational Methods for Fluid Dynamics, 4th ed. (Springer Interna- tional Publishing, Cham, 2020)

    work page 2020

  12. [12]

    Boltzmann, K

    L. Boltzmann, K. Acad. Wiss.(Wein) Sitzb., II Abt66 (1872)

    work page

  13. [13]

    P. L. Bhatnagar, E. P. Gross, and M. Krook, Physical Review94, 511 (1954)

    work page 1954

  14. [14]

    Ansumali, I

    S. Ansumali, I. V. Karlin, and S. Succi, Physica A: Sta- tistical Mechanics and its Applications338, 379 (2004)

    work page 2004

  15. [15]

    Malaspinas and P

    O. Malaspinas and P. Sagaut, Journal of Fluid Mechanics 700, 514 (2012)

    work page 2012

  16. [16]

    H. Chen, S. Succi, and S. Orszag, Physical Review E59, R2527 (1999), publisher: American Physical Society

    work page 1999

  17. [17]

    Succi, O

    S. Succi, O. Filippova, H. Chen, and S. Orszag, Journal of Statistical Physics107, 261 (2002)

    work page 2002

  18. [18]

    H. Chen, S. Kandasamy, S. Orszag, R. Shock, S. Succi, and V. Yakhot, Science (New York, N.Y.)301, 633 (2003)

    work page 2003

  19. [19]

    S. S. Girimaji, Physical Review Letters99, 034501 (2007)

    work page 2007

  20. [20]

    Banda, M

    M. Banda, M. Seaïd, and I. Teleaga, Applied Mathemat- ics and Computation182, 739 (2006)

    work page 2006

  21. [21]

    Sagaut, Computers & Mathematics with Applications Mesoscopic Methods in Engineering and Science,59, 2194 (2010)

    P. Sagaut, Computers & Mathematics with Applications Mesoscopic Methods in Engineering and Science,59, 2194 (2010)

    work page 2010

  22. [22]

    Malaspinas and P

    O. Malaspinas and P. Sagaut, Physics of Fluids23, 105103 (2011)

    work page 2011

  23. [23]

    P. Luan, H. Zhang, and J. Zhang, Journal of Fluid Me- chanics1011, A44 (2025)

    work page 2025

  24. [24]

    H. Chen, S. A. Orszag, I. Staroselsky, and S. Succi, Jour- nal of Fluid Mechanics519, 301 (2004)

    work page 2004

  25. [25]

    H. Chen, I. Staroselsky, and V. Yakhot, Physica Scripta 2013, 014040 (2013)

    work page 2013

  26. [26]

    H. Chen, I. Staroselsky, K. R. Sreenivasan, and V. Yakhot, Atmosphere14, 1109 (2023)

    work page 2023

  27. [27]

    V. L. Saveliev, Physics of Fluids36, 125175 (2024)

    work page 2024

  28. [28]

    P.Asinari, M.Fasano,andE.Chiavazzo,Entropy18,121 (2016), publisher: Multidisciplinary Digital Publishing Institute

    work page 2016

  29. [29]

    Righi, Flow, Turbulence and Combustion97, 121 (2016)

    M. Righi, Flow, Turbulence and Combustion97, 121 (2016)

    work page 2016

  30. [30]

    V. N. Zhigulev, Doklady Akademii Nauk SSSR165, 502 (1965)

    work page 1965

  31. [31]

    Tsugé, The Physics of Fluids17, 22 (1974)

    S. Tsugé, The Physics of Fluids17, 22 (1974)

    work page 1974

  32. [32]

    Chliamovitch, O

    G. Chliamovitch, O. Malaspinas, and B. Chopard, En- tropy17, 7522 (2015), publisher: Multidisciplinary Dig- ital Publishing Institute

    work page 2015

  33. [33]

    Chliamovitch, O

    G. Chliamovitch, O. Malaspinas, and B. Chopard, En- tropy19, 381 (2017), publisher: Multidisciplinary Digi- tal Publishing Institute

    work page 2017

  34. [34]

    Favre,Équations des gaz turbulents compressibles, Tech

    A. Favre,Équations des gaz turbulents compressibles, Tech. Rep. 137 (SNECMA, France, 1965)

    work page 1965

  35. [35]

    Favre, The Physics of Fluids26, 2851 (1983)

    A. Favre, The Physics of Fluids26, 2851 (1983)

    work page 1983

  36. [36]

    Enskog,Kinetische Theorie der Wärmeleitung: Rei- bung und Selbst-diffusion in Gewissen verdichteten gasen und flüssigkeiten(Almqvist & Wiksells boktryckeri-a.-b., 1922)

    D. Enskog,Kinetische Theorie der Wärmeleitung: Rei- bung und Selbst-diffusion in Gewissen verdichteten gasen und flüssigkeiten(Almqvist & Wiksells boktryckeri-a.-b., 1922)

    work page 1922

  37. [37]

    Chapman, T

    S. Chapman, T. Cowling, D. Burnett, and C. Cercig- nani,The Mathematical Theory of Non-uniform Gases: An Account of the Kinetic Theory of Viscosity, Thermal Conduction and Diffusion in Gases, Cambridge Mathe- matical Library (Cambridge University Press, 1953)

    work page 1953

  38. [38]

    B. E. Launder and D. B. Spalding, Computer Methods in Applied Mechanics and Engineering3, 269 (1974)

    work page 1974

  39. [39]

    Malaspinas, Arxiv (2015)

    O. Malaspinas, Arxiv (2015)

    work page 2015

  40. [40]

    F.J.Higuera, S.Succi,andR.Benzi,EurophysicsLetters (EPL)9, 345 (1989)

    work page 1989

  41. [41]

    G. R. McNamara and G. Zanetti, Physical Review Let- ters61, 2332 (1988)

    work page 1988

  42. [42]

    F. J. Higuera and J. Jiménez, Europhysics Letters (EPL) 9, 663 (1989)

    work page 1989

  43. [43]

    Krüger, H

    T. Krüger, H. Kusumaatmaja, A. Kuzmin, O. Shardt, G. Silva, and E. M. Viggen,The Lattice Boltzmann Method: Principles and Practice, Graduate Texts in Physics (Springer International Publishing, 2017)

    work page 2017

  44. [44]

    G. I. Taylor and A. E. Green, Proceedings of the Royal Society of London. Series A - Mathematical and Physical Sciences158, 499 (1936)

    work page 1936

  45. [45]

    Dairay, E

    T. Dairay, E. Lamballais, S. Laizet, and J. C. Vassilicos, Journal of Computational Physics337, 252 (2017)

    work page 2017

  46. [46]

    Laizet, E

    S. Laizet, E. Lamballais, J. C. Vassilicos, and T. Dairay, 3D Taylor-Green vortex Direct Numerical Simulation statistics from Re=1250 to Re=20000 (2019)

    work page 2019

  47. [47]

    M. M. Rogers and R. D. Moser, Physics of Fluids6, 903 (1994)

    work page 1994

  48. [48]

    J. Latt, O. Malaspinas, D. Kontaxakis, A. Parmigiani, D. Lagrava, F. Brogi, M. B. Belgacem, Y. Thorim- bert, S. Leclaire, S. Li, F. Marson, J. Lemus, C. Kot- salos, R. Conradin, C. Coreixas, R. Petkantchin, F. Ray- naud, J. Beny, and B. Chopard, Computers & Mathe- matics with Applications Development and Application of Open-source Software for Problems wi...

    work page 2021

  49. [49]

    Latt and C

    J. Latt and C. Coreixas, Multi-GPU Acceleration of PALABOS Fluid Solver using C++ Standard Parallelism (2025)

    work page 2025

  50. [50]

    Hermite, Comptes Rendus de l’Académie des Sciences de Paris52, 93 (1864)

    C. Hermite, Comptes Rendus de l’Académie des Sciences de Paris52, 93 (1864)

    work page

  51. [51]

    Grad, Communications on Pure and Applied Mathe- matics2, 325 (1949)

    H. Grad, Communications on Pure and Applied Mathe- matics2, 325 (1949)

    work page 1949

  52. [52]

    O. P. Malaspinas,Lattice Boltzmann method for the sim- ulation of viscoelastic fluid flows, Ph.D. thesis (2009). 6 BACKMATTER Appendix A: Chapman–Enskog expansion, hydrodynamic limit and kinetic closure of the filtered Boltzmann equation In the CE expansion framework, f, and the space and time derivatives are respectively expanded as f= ∑∞ k=0ϵkf(k),∂⋆ α=...

    work page 2009