Generic instabilities in the relativistic Chapman-Enskog heat conduction law

A. L. Garcia-Perciante; Marcelo E. Rubio; Oscar A. Reula

REVIEW 3 major objections 5 minor 26 references

Generic instabilities in the relativistic Chapman-Enskog heat conduction law

T0 review · 3 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read The relativistic Chapman-Enskog heat law is generically unstable and ill-posed: in frames with rotation, short-wavelength modes grow without bound.

desk verdict A genuine boosted-frame instability result, but the ill-posedness claim rests on an unproven high-frequency asymptotic. read the letter →

arxiv 1908.04445 v2 pith:H6UTZX5M submitted 2019-08-13 gr-qc hep-th

A. L. Garcia-Perciante , Marcelo E. Rubio , Oscar A. Reula This is my paper · ORCID

classification gr-qchep-th

keywords relativistichydrodynamicsChapman-Enskogheatconductionstabilitywell-posednesshyperbolicitydissipativefluidsdispersionrelation

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 aims to establish that the relativistic fluid equations, when closed with the Chapman-Enskog heat-conduction law, do not give a well-posed initial-value problem. Linear perturbations around equilibrium are stable only when the time direction of the evolution is aligned with the fluid's four-velocity; for any other time direction, including any frame locally boosted relative to a rotating fluid, there are modes whose real growth rate is positive and unbounded as the wave number grows. If this is right, the first-order Chapman-Enskog closure cannot serve as a generically stable relativistic fluid theory, and that matters because this heat law is used in numerical simulations of relativistic gases, especially ultrarelativistic electron gases. The authors conclude that the system is not only unstable but ill-posed, since in the high-frequency limit there are solutions that grow arbitrarily fast, and that the instability is avoidable only in the irrotational, surface-forming case.

What carries the argument

The load-bearing object is the dispersion relation, Eq. (III.21): a quartic complex polynomial in the boosted frequency $\hat S$ and wave number $\hat K$, with coefficients $\alpha_j$ and $\beta_j$ that depend on the boost velocity $v$, the temperature parameter $z$, and the microscopic-to-macroscopic scale ratio $\zeta$. In the co-moving frame the same analysis reduces to a cubic polynomial whose coefficients pass the Routh-Hurwitz test, an algebraic criterion for all roots having negative real part; the Lorentz transformation (III.19) is the step that converts damping into growth. The high-frequency regime is probed with the ansatz $\hat S=S_0+S_1\hat K+S_2\hat K^2$, and because $\alpha_4\neq 0$ the quadratic term vanishes, leaving branches that grow linearly with $\hat K$. A separate check on first-order divergence-type theories, whose symmetric-hyperbolic structure would guarantee well-posedness, shows that the Chapman-Enskog heat law does not fit into that class unless the acceleration is eliminated using lower-order Euler equations.

What would settle it

Linearize the Chapman-Enskog-closed fluid equations around an exact stationary rotating solution, for example rigid rotation with $u^\mu$ not surface-forming, and compute the high-frequency dispersion relation; if all modes satisfy $\operatorname{Re} S\le 0$ or $\operatorname{Re} S$ stays bounded as $K\to\infty$, the paper's generic ill-posedness claim is false.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the system of relativistic fluid equations coupled to the Chapman-Enskog heat flux $q^\mu=-h^{\mu\nu}(\kappa\nabla_\nu T/T-\lambda\nabla_\nu n/n)$ is non-hyperbolic for a generic time direction. In the fluid's co-moving frame the dispersion relation has only roots with negative real parts, so rest-frame perturbations decay; this stability is shown with the Routh-Hurwitz criterion using standard thermodynamic inequalities for a relativistic ideal gas. After a Lorentz transformation to a boosted frame, however, the same dispersion relation becomes a quartic complex polynomial with positive-real-part roots, and in the high-frequency limit $\hat K\to\infty$ the growth rate behaves as $\hat S\sim\epsilon\hat K$, so modes grow arbitrarily fast. Since a rotating fluid's four-velocity is not surface-forming, no global time direction can be aligned with it, making the instability generic and the Cauchy problem ill-posed for rotating configurations. Unlike Eckart's theory, the instability is absent when the time direction is aligned with the fluid's direction, but that special alignment fails exactly when rotation is present.

Load-bearing premise

The paper does not linearize around a real rotating background; it assumes a uniformly boosted homogeneous equilibrium stands in for a rotating fluid, so instability for every boost is taken to mean instability whenever rotation is present.

Editorial extensions

If this is right

Numerical relativistic-fluid simulations that close with the first-order Chapman-Enskog heat flux will be subject to short-wavelength noise that grows faster as resolution increases, unless the evolution is effectively locked to the local fluid rest frame.
The relativistic Navier-Stokes-Fourier system built on this heat law does not have a well-posed Cauchy problem in rotating flows, so initial data do not determine a stable future evolution.
Stability in the rest frame is not sufficient for a first-order relativistic heat law once rotation is allowed, since any nonzero boost already produces unbounded growth.
Adding higher-order dissipative terms that restore hyperbolicity becomes a necessary step for using Chapman-Enskog-type closures in generic flows, as the paper itself notes.

Reading between the lines

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

A direct linearization around an exact rotating equilibrium, rather than a uniformly boosted homogeneous state, would test whether background gradients suppress the predicted growth; the paper's constant-boost argument leaves that open.
The same analysis suggests a general diagnostic for any first-order closure: compute the boosted-frame dispersion relation and check whether $\lim_{\hat K\to\infty}\operatorname{Re}\hat S/\hat K>0$ for some boost; if so, the theory is ill-posed for rotating fluids.
The absence of instability in the co-moving frame may explain why Chapman-Enskog closures appear to work in symmetric, effectively one-dimensional simulations even though the generic Cauchy problem is ill-posed.
If the ill-posedness persists in direct rotating tests, first-order Chapman-Enskog closures should be regarded as effective low-Knudsen theories, with any short-wavelength control coming from the numerical scheme rather than from the physical equations.

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,

The take-home is that this paper has a real result, but the abstract oversells it. The real result: the Chapman-Enskog heat-flux law, which is stable for linear perturbations in the co-moving frame, develops a positive-real-part mode when you look at the same system from a boosted time direction. That is new and worthwhile—the earlier papers [11,12] only had the co-moving stability. The paper also shows that the CE law cannot be recovered from a general first-order divergence-type theory unless you use the Euler equations to trade acceleration for gradients, which is inconsistent in the DTT framework. That DTT calculation is careful and seems right.

Where it goes soft is the ill-posedness claim. Section III.B.1 argues that as the wave number grows, the dispersion relation has roots of the form S ~ εK, with ε satisfying Eq. (III.24). But the paper never shows that this equation has any root with Re ε > 0. The K=0 analysis proves a positive root at zero wave number, but that says nothing about the limit K→∞. The numerical plots stop at small K (0.15 and 0.008) and don't extract a slope, so they don't support the asymptotic claim either. The abstract's statement that the real part grows without bound is therefore not established. What is established is a long-wavelength instability in a boosted frame, which is still a genuine negative result. But the ill-posedness conclusion should be withdrawn or properly proved.

Two lesser concerns: First, the jump from 'unstable in a uniformly boosted frame' to 'unstable whenever the four-velocity is not surface-forming' is an extrapolation. Rotating backgrounds have gradients and vorticity that are absent in the constant-boost analysis. It's a plausible extrapolation, but it's not derived. Second, the perturbation analysis only treats the longitudinal mode and boosts collinear with the wave vector. A fully generic frame would involve transverse modes and oblique boosts, so the claim about 'any generic time direction' is a little broader than what was actually computed.

If the high-frequency issue is fixed—say, by proving that Eq. (III.24) has a positive-real-part root for some admissible parameters, or by removing the ill-posedness language—this is a decent paper for the relativistic-fluid community. The co-moving analysis is explicit and correct, the DTT section is a useful side result, and the boosted-frame instability is enough to matter for simulations that use the CE closure. I'd send it to peer review, but the referee should push hard on the high-frequency asymptotics. I would cite it for the long-wavelength instability, not for ill-posedness.

Yes, the authors are thinking clearly; the flaw is an overclaim, not a confusion.

Referee Report

3 major / 5 minor

Summary. The paper analyzes the well-posedness of the relativistic fluid equations closed with the Chapman-Enskog (CE) heat-flux relation q^mu = -h^mu nu (kappa nabla_nu T/T - lambda nabla_nu n/n). It first argues that this closure cannot be obtained from a first-order divergence-type theory without using the equilibrium (Euler-order) expression for the four-acceleration, and it notes the open status of that substitution. It then studies linear perturbations: in the comoving frame, a cubic dispersion relation is derived and shown to be stable for a classical relativistic ideal gas using the Routh-Hurwitz criterion and kinetic identities. For an arbitrary time direction, obtained by a constant Lorentz boost, a quartic complex dispersion relation is derived; a positive real root is exhibited at zero wave number, and numerical roots with positive real part are shown for two sets of parameters. The paper concludes that the system is unstable and ill-posed because high-frequency modes grow linearly in wave number with positive real part, and it claims that the instability can be avoided only when the fluid velocity is surface-forming, i.e., when there is no rotation.

Significance. If established, this is a substantial negative result: the relativistic CE heat-conduction law would join Eckart-type theories as linearly unstable and, more strongly, ill-posed, removing a candidate first-order relativistic fluid theory from the class of well-posed initial-value problems. The paper is largely self-contained, uses no fitted parameters, and its comoving-frame analysis is explicit and checkable: the dispersion relation, the positivity conditions, and the Routh-Hurwitz argument are all given in closed form. The DTT discussion is appropriately caveated in the text. The main weakness is that the load-bearing high-frequency claim is asserted rather than proved, and the step from a constant boost to rotating backgrounds is not justified.

major comments (3)

[Section III.B.1, Eq. (III.24)] The ill-posedness conclusion is not established. The text asserts that for high frequency the roots behave as hat S ~ epsilon hat K with epsilon satisfying Eq. (III.24), and that this implies modes grow arbitrarily fast, but it does not prove that Eq. (III.24), or its leading high-K form sum_j alpha_j (i epsilon)^j = 0, has any root with Re epsilon > 0. The preceding K = 0 analysis gives a positive real root only at zero wave number, and the numerical figures stop at small K without extracting an asymptotic slope. If all roots epsilon have Re epsilon <= 0, the high-frequency growth would not be unbounded and the abstract's claim that the real part grows without bound would fail. A root-locus or asymptotic argument covering the physical parameter ranges is needed; alternatively, the numerical evidence must be extended to large K and compared with the predicted asymptotic slope.
[Section III.B and Section IV] The rotation claim is an extrapolation. The arbitrary-frame calculation linearizes around a homogeneous equilibrium and then applies a constant Lorentz boost; the resulting background four-velocity is constant and has zero vorticity, so it is surface-forming. The paper does not linearize around any background with nonzero vorticity or with non-surface-forming u^mu. Therefore the statement that the instability 'can only be avoided in the particular case where no rotation is present' is not supported by the analysis. The authors should either model a rotating background explicitly or identify a precise argument by which local constant-boost behavior controls the general non-surface-forming case.
[Section III.B.1] The high-frequency limit is not defined unambiguously. Equation (III.21) contains the product hat K zeta, and the text first requires that this product remain bounded as hat K grows, which forces zeta -> 0; it then treats Eq. (III.24) as the high-frequency limit. This is a distinguished limit rather than the fixed-zeta limit hat K -> infinity, which is the standard one for a linearized Cauchy problem. The paper should state the intended scaling and explain why growth in this distinguished limit implies ill-posedness of the fixed system; otherwise the growth rate may be an artifact of letting the Knudsen parameter depend on wave number.

minor comments (5)

[Section III.B, around Eq. (III.20)] The text refers to 'Eq. (30)' after deriving Eq. (III.20); this equation number should be updated to Eq. (III.20) or the numbering should be made consistent.
[Figures 1 and 2] The figures use the symbol xi in their labels while the text uses zeta for the Knudsen parameter; these should be unified.
[Section III.B, Eq. (III.22)] The K = 0 polynomial contains a factor S^3, so there is also a triple zero root; the text only discusses the nonzero root and should state explicitly that the zero root is present and does not affect the instability conclusion.
[Section III.B.1, Eq. (III.24)] The symbol epsilon is introduced without specifying which root of the algebraic equation is selected, and the dependence on the parameters zeta, v, z, and mu(z) is not discussed; a brief statement about root selection and parameter range would improve readability.
[Section IV] The phrase 'per-se' appears in the Discussion and should be corrected to 'per se'.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: the stability and ill-posedness claims are obtained by direct linearization of the given CE constitutive law; the paper's self-citations are non-load-bearing.

full rationale

The paper's central chain is a linear perturbation analysis of the relativistic fluid equations closed with the CE heat-flux law (III.1). The co-moving dispersion relation (III.15) follows by direct algebra from Eqs. (III.3)-(III.5), and the stability conclusion in that frame is rederived in the paper via the Routh-Hurwitz criterion, so the prior results cited for it ([11], [12]) are not load-bearing. The arbitrary-frame quartic (III.20) is obtained by the exact Lorentz transformation (III.19) of the co-moving dispersion relation, with no fitted parameters; the K=0 root (III.23) and its claimed positivity rest on the standard kinetic inequalities (III.18) from the Cercignani-Kremer book, an external benchmark, not on the paper's own conclusions. The identity kappa/lambda = G(z)/z - 1 (III.16) is quoted from Boltzmann-equation transport theory (refs. [2]/[21]); even if Ref. [21] is by an author of this paper, it is a parameter-free external kinetic result that is assumed as an input of the theory under test, not an output of the stability analysis. Ref. [23] is cited only as a pointer in the Discussion for 'a detailed analysis' of the underlying mechanism, not as the basis for the ill-posedness conclusion, which is argued from Eq. (III.24) within this paper. No parameter is fitted to data and no 'prediction' is defined in terms of the quantity it is said to predict; nothing in the derivation reduces to its own inputs by construction. Two asserted steps deserve note but are correctness risks, not circularity: the positivity of the K=0 root is only sketched ('It is straightforward to show that this solution is positive'), and the high-frequency asymptotics S-hat ~ epsilon K-hat in Section III.B.1 is asserted rather than proved (no root of Eq. (III.24) is exhibited with Re epsilon > 0), so the 'grow arbitrarily fast' conclusion rests on an unverified asymptotic. The extrapolation from uniformly boosted frames to rotating, non-surface-forming flows is likewise an unproven assumption. These are proof gaps, not reductions of outputs to inputs, and therefore do not raise the circularity score.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No parameters are fitted in this paper; transport coefficients κ, λ, τ and μ(z) are inputs from kinetic theory. The analysis assumes the CE constitutive law, the Jüttner equilibrium distribution, and standard identities of the relativistic ideal gas. No new particles, fields, or unobserved entities are introduced.

assumptions (4)

domain assumption Jüttner distribution for classical relativistic ideal gas at local equilibrium
Used in Section II and Appendix A to compute the constitutive tensor A^{μνσ} and thermodynamic identities; restricts the result to dilute classical gases with no internal degrees of freedom.
domain assumption Kinetic identity κ/λ = G(z)/z - 1
Invoked in Eq. (III.16) to simplify the co-moving dispersion coefficients b1 and b3; taken from the Boltzmann equation transport theory of Cercignani and Kremer and of Garcia-Perciante and Mendez, not derived in this paper.
domain assumption Equilibrium gas inequalities G(z)/z > 4 and 1/3 < 1/c_n < 2/3
Used in Eq. (III.18) to apply Routh-Hurwitz and to assert positivity of the boosted homogeneous root; these are standard properties of the classical relativistic ideal gas.
standard math Surface-forming equivalence: a timelike vector field admits orthogonal hypersurfaces iff its rotation vanishes
Underlies the statement that the co-moving frame exists globally only when no rotation is present; standard differential geometry.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Generic instabilities in the relativistic Chapman-Enskog heat conduction law." pith.science (2026). https://pith.science/paper/H6UTZX5M

@misc{pith2026190804445,
  author       = {Pith},
  title        = {Pith review of: Generic instabilities in the relativistic Chapman-Enskog heat conduction law},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/H6UTZX5M}},
  note         = {Machine review of arXiv:1908.04445}
}

read the original abstract

We address the well-posedness of the Cauchy problem corresponding to the relativistic fluid equations, when coupled with the heat-flux constitutive relation arising within the relativistic Chapman-Enskog procedure. The resulting system of equations is shown to be non hyperbolic, by considering general perturbations over the whole set of equations written with respect to a generic time direction. The obtained eigenvalues are not purely imaginary and their real part grows without bound as the wave-number increases. Unlike Eckart's theory, this instability is not present when the time direction is aligned with the fluid's direction. However, since in general the fluid velocity is not surface-forming, the instability can only be avoided in the particular case where no rotation is present.

Figures

Figures reproduced from arXiv: 1908.04445 by the authors.

**Figure 2.** FIG. 2. Solutions of the dispersion relation (III.21) with pos [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 24 canonical work pages

[1]

Since the characteristic wavelength of ﬂuc- tuations is required to be large enough for collisions to take place, it must be ζ≪ 1

Ill-posedness As was previously introduced, the dimensionless pa- rameter ζ is the ratio between microscopic and macro- 6 scopic scales. Since the characteristic wavelength of ﬂuc- tuations is required to be large enough for collisions to take place, it must be ζ≪ 1. Then, when considering growing wave numbers ˆK, the product ˆKζ should still be bounded. ...

work page
[2]

Any ﬂuid theory which is governed by a set of conservation laws (particle num- ber density, energy and momentum densities, etc) con- stitutes, indeed, a divergence-type theory

Brief detour on DTT Roughly speaking, a set of dynamic equations is con- sidered a divergence- type theory (DTT) if it can be writ- ten as a set of equations on the divergence of the corre- sponding dynamical variables. Any ﬂuid theory which is governed by a set of conservation laws (particle num- ber density, energy and momentum densities, etc) con- stit...

work page
[3]

As was pointed out in Section II, such a constitutive tensor is made up by means of the third moment of f (0)

Details of calculations of Section II We now compute the most general constitutive tensor ﬁeld that can be constructed from ﬁrst-order divergence- type ﬂuid theories, using a J¨ uttner distribution function in local equilibrium, f (0). As was pointed out in Section II, such a constitutive tensor is made up by means of the third moment of f (0). From the m...

work page
[4]

Chapman and T

S. Chapman and T. G. Cowling. The Mathematical The- ory of Non-Uniform Gases . Cambridge University press, 1970

work page 1970
[5]

Cercignani and G

C. Cercignani and G. Medeiros Kremer. The Relativistic Boltzmann Equation: Theory and Applications . Cam- bridge University press, 1991

work page 1991
[6]

Cercignani

C. Cercignani. The Boltzmann Equation and Its Appli- cations. Applied Mathematical Sciences 67, 1988

work page 1988
[7]

Gabbana, M

A. Gabbana, M. Mendoza, S. Succi, and R. Tripiccione. Kinetic approach to relativistic dissipation. Phys. Rev. E, 96:023305, Aug 2017

work page 2017
[8]

Gabbana, Simeoni D., S

A. Gabbana, Simeoni D., S. Succi, and R. Tripiccione. Relativistic dissipation obeys chapman-enskog asymp- totics: Analytical and numerical evidence as a basis for accurate kinetic simulations. Phys. Rev. E , 99:052126, 2019

work page 2019

Show all 26 references

[9]

The thermodynamics of irreversible pro- cesses

Carl Eckart. The thermodynamics of irreversible pro- cesses. iii. relativistic theory of the simple ﬂuid. Phys. Rev., 58:919–924, Nov 1940

1940
[10]

W. A. Hiscock and L. Lindblom. Generic instabilities in ﬁrst-order dissipative relativistic ﬂuid theories. Phys. Rev. D, 31:725, 1985

1985
[11]

Landau and E.M

L.D. Landau and E.M. Lifshitz. Fluid Mechanics. Num- ber v. 6 in Course on Theoretical Physics. Elsevier Sci- ence, 2013

2013
[12]

Stabil- ity and causality in relativistic dissipative hydrodynam- ics

G S Denicol, T Kodama, T Koide, and Ph Mota. Stabil- ity and causality in relativistic dissipative hydrodynam- ics. Journal of Physics G: Nuclear and Particle Physics , 35(11):115102, sep 2008

2008
[13]

Shi Pu, Tomoi Koide, and Dirk H. Rischke. Does stability of relativistic dissipative ﬂuid dynamics imply causality? Phys. Rev. D , 81:114039, Jun 2010

2010
[14]

Garcia-Colin A

L.S. Garcia-Colin A. Sandoval-Villalbazo, A.L. Garc´ ıa- Perciante. Relativistic transport theory for simple ﬂuids to ﬁrst order in the gradients. Physica A, 388:3765–3770, 2009

2009
[15]

A. L. Garcia-Perciante, L. S. Garcia-Colin, and A. Sandoval-Villalbazo. Rayleigh-brillouin spectrum in special relativistic hydrodynamics. Phys. Rev. E , 79:066310, Jun 2009

2009
[16]

Dissipative relativistic ﬂuid theories of divergence type

Robert Geroch and Lee Lindblom. Dissipative relativistic ﬂuid theories of divergence type. Phys. Rev. D, 41:1855– 1861, Mar 1990

1990
[17]

Relativistic theories of dissipative ﬂuids

Robert Geroch. Relativistic theories of dissipative ﬂuids. Journal of Mathematical Physics, 36(8):4226–4241, 1995

1995
[18]

Causal theories of dis- sipative relativistic ﬂuids

Robert Geroch and Lee Lindblom. Causal theories of dis- sipative relativistic ﬂuids. Annals of Physics, 207(2):394– 416, 1991

1991
[19]

On hyperbolic ”theories” of relativistic dissipative ﬂuids

Robert Geroch. On hyperbolic ”theories” of relativistic dissipative ﬂuids. arXiv preprint gr-qc/0103112 , 2001

2001 arXiv
[20]

Robert P. Geroch. Partial diﬀerential equations of physics. In General relativity. Proceedings, 46th Scottish Universities Summer School in Physics, NATO Advanced Study Institute, Aberdeen, UK, July 16-29, 1995 , 1996

1995
[21]

Das maxwellsche gesetz der geschwindigkeitsverteilung in der relativtheorie

Ferencz J¨ uttner. Das maxwellsche gesetz der geschwindigkeitsverteilung in der relativtheorie. Annalen der Physik, 339(5):856–882, 1911

1911
[22]

Morales-T´ ecotl

Guillermo Chac´ on-Acosta, Leonardo Dagdug, and Hugo A. Morales-T´ ecotl. Manifestly covariant j¨ uttner distribution and equipartition theorem. Phys. Rev. E , 81:021126, Feb 2010

2010
[23]

P. Mazur S. R. De Groot. Non-Equilibrium Thermody- namics. Dover, 2011

2011
[24]

A. R. M´ endez A. L. Garc´ ıa-Perciante. Heat conduction in relativistic neutral gases revisited. Gen Relativ Gravit. , 43:2257–2275, 2011

2011
[25]

A. Hurwitz. Ueber die bedingungen, unter welchen eine gleichung nur wurzeln mit negativen reellen theilen be- sitzt. Mathematische Annalen, 46(2):273–284, Jun 1895
[26]

A. L. Garc´ ıa-Perciante and Oscar Reula. On the illposed- ness and stability of the relativistic heat equation. arXiv 1909.08271. 2019

1909 arXiv

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Since the characteristic wavelength of ﬂuc- tuations is required to be large enough for collisions to take place, it must be ζ≪ 1

Ill-posedness As was previously introduced, the dimensionless pa- rameter ζ is the ratio between microscopic and macro- 6 scopic scales. Since the characteristic wavelength of ﬂuc- tuations is required to be large enough for collisions to take place, it must be ζ≪ 1. Then, when considering growing wave numbers ˆK, the product ˆKζ should still be bounded. ...

work page

[2] [2]

Any ﬂuid theory which is governed by a set of conservation laws (particle num- ber density, energy and momentum densities, etc) con- stitutes, indeed, a divergence-type theory

Brief detour on DTT Roughly speaking, a set of dynamic equations is con- sidered a divergence- type theory (DTT) if it can be writ- ten as a set of equations on the divergence of the corre- sponding dynamical variables. Any ﬂuid theory which is governed by a set of conservation laws (particle num- ber density, energy and momentum densities, etc) con- stit...

work page

[3] [3]

As was pointed out in Section II, such a constitutive tensor is made up by means of the third moment of f (0)

Details of calculations of Section II We now compute the most general constitutive tensor ﬁeld that can be constructed from ﬁrst-order divergence- type ﬂuid theories, using a J¨ uttner distribution function in local equilibrium, f (0). As was pointed out in Section II, such a constitutive tensor is made up by means of the third moment of f (0). From the m...

work page

[4] [4]

Chapman and T

S. Chapman and T. G. Cowling. The Mathematical The- ory of Non-Uniform Gases . Cambridge University press, 1970

work page 1970

[5] [5]

Cercignani and G

C. Cercignani and G. Medeiros Kremer. The Relativistic Boltzmann Equation: Theory and Applications . Cam- bridge University press, 1991

work page 1991

[6] [6]

Cercignani

C. Cercignani. The Boltzmann Equation and Its Appli- cations. Applied Mathematical Sciences 67, 1988

work page 1988

[7] [7]

Gabbana, M

A. Gabbana, M. Mendoza, S. Succi, and R. Tripiccione. Kinetic approach to relativistic dissipation. Phys. Rev. E, 96:023305, Aug 2017

work page 2017

[8] [8]

Gabbana, Simeoni D., S

A. Gabbana, Simeoni D., S. Succi, and R. Tripiccione. Relativistic dissipation obeys chapman-enskog asymp- totics: Analytical and numerical evidence as a basis for accurate kinetic simulations. Phys. Rev. E , 99:052126, 2019

work page 2019