Steady hydrodynamic limit of the Boltzmann equation with Maxwell reflection: rigorous derivation of the compressible Navier-Stokes-Fourier system with Coron-Aoki slip boundary conditions in a channel

Duan and Zhang (J. Math. Fluid Mech. 27 (2025); arXiv:2409.11809) proved existence and uniqueness of strong solutions to the steady compressible Navier-Stokes-Fourier (NSF) system in the channel $(0,1)\times\mathbb{T}^2$ with the generalized slip boundary conditions that Coron and Aoki et al. derived formally from the Boltzmann equation with Maxwell reflection, under a structural positivity assumption (A) on the slip coefficients, and explicitly left open the viscous compressible fluid approximation (hydrodynamic limit) of the Boltzmann equation with these boundary conditions. This record states that open derivation problem: prove, or disprove, that solutions of the steady Boltzmann equation with order-one-accommodation Maxwell reflection in the same channel and small wall-temperature variation converge, as the Knudsen number tends to zero, to the Duan-Zhang strong solution of the steady NSF system with the Coron-Aoki conditions, including the Knudsen-layer analysis that determines the slip coefficients. Later literature (the Jiang-Wu program, 2025-2026; Medina, 2026) supplies formal derivations, Knudsen-layer solvability, and a rigorous time-dependent justification only for purely specular reflection, so the stationary compressible viscous limit with kinetic slip conditions remains open.

Background

The steady motion of a monatomic ideal gas in a bounded domain $\Omega\subset\mathbb{R}^3$ is described at the fluid level by the stationary compressible Navier-Stokes-Fourier (NSF) system: the density $\rho>0$, velocity $u$, and temperature $\theta>0$ satisfy $\mathrm{div}(\rho u)=0$, $\mathrm{div}(\rho u\otimes u)+\nabla p=\mathrm{div} S$, and $\mathrm{div}[(\rho E+p)u]=\mathrm{div} q+\mathrm{div}(Su)$, with pressure $p=R\rho\theta$ for a perfect polytropic gas (gas constant $R$, total energy $E$), Newtonian stress $S(u,\theta)=\mu(\theta)(\nabla u+\nabla u^{T}-\tfrac{2}{3}\,\mathrm{div}u\,I_3)$ and Fourier heat flux $q(\theta)=-\kappa(\theta)\nabla\theta$, where $\mu$ and $\kappa$ are the viscosity and heat conductivity. Classical closures at an impermeable wall ($\partial\Omega$, unit outer normal $n$) are the no-slip conditions ($u=0$ with prescribed wall temperature $\theta_w$) and the Navier-slip conditions, in which the tangential velocity is proportional to the wall shear and $u\cdot n=0$.

Kinetic theory suggests more refined wall laws. At the mesoscopic level the gas obeys the Boltzmann equation for its distribution function, and the Knudsen number $\epsilon$ — the ratio of the molecular mean free path to a macroscopic length — measures rarefaction. Gas-surface interaction is commonly modeled by the Maxwell reflection boundary condition: each incoming molecule is either specularly reflected (probability $1-\alpha$) or re-emitted with a Maxwellian (equilibrium Gaussian) distribution at the wall temperature $\theta_w$ (probability $\alpha$), where $\alpha\in[0,1]$ is the accommodation coefficient; $\alpha=1$ is pure diffuse reflection and $\alpha=0$ pure specular reflection. In the hydrodynamic limit $\epsilon\to 0$ the Chapman-Enskog expansion — a systematic expansion of the distribution in powers of $\epsilon$ — yields the NSF system in the interior, but near the wall a kinetic boundary layer of width $\epsilon$ (the Knudsen layer, governed by a half-space problem of Milne type) must be resolved, and its solvability determines which boundary conditions the fluid system inherits. Coron (1989), for a gas between parallel plates, and Aoki, Baranger, Hattori, Kosuge, Martalò, Mathiaud and Mieussens (2017), for general domains, carried out this program formally for Maxwell reflection with order-one accommodation and derived a class of generalized slip conditions: for $x\in\partial\Omega$ and tangential directions $t$, $u\cdot n=0,$ $\rho u\cdot t+\alpha_1\,(\nabla u+\nabla u^{T}):n\otimes t-\alpha_2\,\nabla\theta\cdot t=0,$ $\rho(\theta-\theta_w)-\alpha_3\,\mu(\theta_w)\nabla u:n\otimes n+\alpha_4\,\kappa(\theta_w)\nabla\theta\cdot n=0,$ where the positive coefficients $\alpha_1,\dots,\alpha_4$ are explicit combinations of $\mu(\theta_w)$, $\kappa(\theta_w)$, $R$, and four slip coefficients $a_u^{I},a_u^{II},a_\theta^{I},a_\theta^{II}>0$ determined solely by the molecular collision law (their precise form is given in Duan-Zhang, conditions (1.2); explicit values for hard-sphere molecules, as well as for the Bhatnagar-Gross-Krook model, are tabulated in Aoki et al. (2017)). Their distinctive features, relative to Navier-slip, are the coupling of the tangential velocity slip to the tangential temperature gradient $\nabla\theta\cdot t$, the coupling of the temperature jump $\rho(\theta-\theta_w)$ to the normal strain $\nabla u:n\otimes n$, and the density-dependent factors $\rho u$ and $\rho(\theta-\theta_w)$.

Duan and Zhang (2025) established the fluid-side theory for these conditions in the channel $\Omega=(0,1)\times\mathbb{T}^2$ ($\mathbb{T}^2$ the two-dimensional periodic torus, so the walls are the flat plates $x_1=0$ and $x_1=1$): if the wall temperature $\theta_w$ is sufficiently close to a positive constant, the steady NSF system with the Coron-Aoki conditions admits a unique strong solution $(\rho,u,\theta)\in(L_0^2\cap H^2)\times V^3\times H^3$ ($L_0^2$ the square-integrable functions of zero mean; $V^3$ a tangential Sobolev class). The proof rests on a new variational formulation of the linearized problem and a fixed-point argument; its main obstacle is the interplay of velocity and temperature derivatives in the boundary conditions, which generates mixed quadratic boundary forms that the authors control by imposing a structural assumption (A): positivity of an explicit scalar quantity and positive-definiteness of an explicit $6\times 6$ matrix built from the slip coefficients. Remark 1.2 of that paper verifies (A) only in a special scaling regime in which the conditions degenerate to Navier-slip; whether (A) can be weakened, or holds for the physically correct Knudsen-layer-determined coefficients, is not known. The authors close the introduction by stating the intended continuation: a rigorous viscous compressible fluid approximation of the Boltzmann equation in the same channel setting with the generalized slip conditions, generalizing their own one-dimensional heat-transfer analysis with diffuse boundary conditions (Duan-Liu-Yang-Zhang, 2022).

The surrounding hydrodynamic-limit literature covers neighboring cases but not this one. For time-dependent problems, Jiang and Masmoudi (2017) derived the incompressible NSF limit of the Boltzmann equation in bounded domains, resolving the boundary-layer structure for Maxwell-type boundaries. For stationary problems, Esposito, Guo, Kim and Marra (2018) proved that stationary solutions of the Boltzmann equation with diffuse boundary conditions converge to solutions of the steady incompressible NSF system, validating Fourier's law. Most directly, Jiang and Wu (arXiv:2501.08715, published in J. Stat. Phys. 193 (2026)) formally derived Knudsen-number-dependent slip boundary conditions for the time-dependent compressible Navier-Stokes system from the Boltzmann equation with specular ($\alpha=0$) and almost-specular ($\alpha=O(\epsilon^\beta)$, $\beta>0$) reflection, and rigorously justified the compressible NSF approximation only for purely specular reflection, where the inherited fluid boundary conditions are the simple complete-slip conditions $u\cdot n=0$, $(\nabla u+\nabla u^{T}):n\otimes t=0$, $\nabla\theta\cdot n=0$. They state that a rigorous convergence proof of the compressible NSF approximation involving the Knudsen boundary layer "has not been achieved so far", and that whether the Knudsen-layer-determined slip coefficients make the boundary energy of the fluid system dissipative is a "highly non-trivial problem" — the kinetic-side counterpart of Duan-Zhang's assumption (A). Companion works prove halfspace Knudsen-layer solvability for $0<\alpha<1$ (He-Jiang-Wu), the acoustic limit for the full accommodation range (Jiang-Wu), and an incompressible Euler limit with order-one accommodation in the half-space (Jiang-Wang-Wu-Zhang); Medina (2026) constructed non-equilibrium steady states for the Boltzmann equation with non-isothermal Maxwell boundaries near the hydrodynamic regime but does not derive fluid boundary conditions. Thus the stationary viscous compressible limit with the Coron-Aoki slip conditions — the exact validation target left open by Duan-Zhang — remains unavailable.

Problem Statement

Prove, or disprove, the hydrodynamic validation left open by Duan-Zhang: in the channel $\Omega=(0,1)\times\mathbb{T}^2$ with wall temperature $\theta_w$ sufficiently close to a positive constant, and with the Maxwell reflection boundary condition at the walls in the regime of order-one accommodation coefficient — the regime in which Coron and Aoki et al. derived the generalized slip conditions — show that solutions of the steady Boltzmann equation with Knudsen number $\epsilon$ converge, as $\epsilon\to 0$ in a precisely stated function class and norm with rates, to the unique strong solution of the steady compressible Navier-Stokes-Fourier system with the Coron-Aoki generalized slip boundary conditions constructed by Duan and Zhang. For definiteness, the molecular collision kernel is that of hard-sphere molecules with angular cutoff — the kernel for which Aoki et al. (2017) tabulate explicit values of the slip coefficients $a_u^{I},a_u^{II},a_\theta^{I},a_\theta^{II}$ and for which the half-space Knudsen-layer problem with Maxwell reflection, $0<\alpha<1$, is solved in $L^\infty$ by He-Jiang-Wu (arXiv:2501.08733); the same problem may be posed for any cutoff hard-potential kernel, the class reflected in Duan-Zhang's Chapman-Enskog transport laws $\mu(\theta)=\mu_0\theta^{1-\gamma/2}$, $\kappa(\theta)=\kappa_0\theta^{1-\gamma/2}$ with $-3<\gamma\le 1$, with the slip coefficients determined by the chosen kernel. The derivation requires the Knudsen-layer (half-space) analysis that identifies the slip coefficients $a_u^{I},a_u^{II},a_\theta^{I},a_\theta^{II}$ and validates those boundary conditions as the correct fluid-level wall laws, together with the matching of the boundary data of the kinetic and fluid problems. An answer showing instead that the correct limiting wall laws differ from the Coron-Aoki conditions (so that the formal derivation fails at the viscous order) also resolves the problem, in the negative.

References

1. R. Duan and J. Zhang, Steady Compressible Navier-Stokes-Fourier System with Slip Boundary Conditions Arising from Kinetic Theory, Journal of Mathematical Fluid Mechanics 27 (2025), no. 4. DOI: 10.1007/s00021-025-00972-w, https://doi.org/10.1007/s00021-025-00972-w (preprint arXiv:2409.11809, https://arxiv.org/abs/2409.11809)
2. F. Coron, Derivation of Slip boundary conditions for the Navier-Stokes system from the Boltzmann equation, Journal of Statistical Physics 54 (1989), no. 3-4, 829-857. DOI: 10.1007/BF01019777, https://doi.org/10.1007/BF01019777
3. K. Aoki, C. Baranger, M. Hattori, S. Kosuge, G. Martalo, J. Mathiaud and L. Mieussens, Slip Boundary Conditions for the Compressible Navier-Stokes Equations, Journal of Statistical Physics 169 (2017), no. 4, 744-781. DOI: 10.1007/s10955-017-1886-8, https://doi.org/10.1007/s10955-017-1886-8
4. N. Jiang and Y. Wu, Compressible Navier-Stokes System with Slip Boundary from the Boltzmann Equation with Reflection Boundary: Derivations and Justifications, Journal of Statistical Physics 193 (2026), no. 4. DOI: 10.1007/s10955-026-03606-1, https://doi.org/10.1007/s10955-026-03606-1 (preprint arXiv:2501.08715, https://arxiv.org/abs/2501.08715)
5. L.-B. He, N. Jiang and Y. Wu, Boltzmann boundary layer equation with Maxwell reflection boundary condition and applications to fluid limits, preprint (2025), arXiv:2501.08733, https://arxiv.org/abs/2501.08733
6. N. Jiang and Y. Wu, Kinetic-fluid boundary layers and acoustic limit for the Boltzmann equation with general Maxwell reflection boundary condition, preprint (2025), arXiv:2501.08707, https://arxiv.org/abs/2501.08707
7. N. Jiang, C. Wang, Y. Wu and Z. Zhang, Incompressible Euler limit from the Boltzmann equation with Maxwell reflection boundary condition in the half-space, preprint (2025), arXiv:2506.18420, https://arxiv.org/abs/2506.18420
8. R. Medina, The Boltzmann equation with non-isothermal Maxwell boundary conditions, preprint (2026), arXiv:2604.13572, https://arxiv.org/abs/2604.13572
9. R. Esposito, Y. Guo, C. Kim and R. Marra, Stationary Solutions to the Boltzmann Equation in the Hydrodynamic Limit, Annals of PDE 4 (2018), no. 1, Paper No. 1. DOI: 10.1007/s40818-017-0037-5, https://doi.org/10.1007/s40818-017-0037-5 (preprint arXiv:1502.05324, https://arxiv.org/abs/1502.05324)
10. N. Jiang and N. Masmoudi, Boundary Layers and Incompressible Navier-Stokes-Fourier Limit of the Boltzmann Equation in Bounded Domain I, Communications on Pure and Applied Mathematics 70 (2017), no. 1, 90-171. DOI: 10.1002/cpa.21631, https://doi.org/10.1002/cpa.21631
11. R. Duan, S. Liu, T. Yang and Z. Zhang, Heat Transfer Problem for the Boltzmann Equation in a Channel with Diffusive Boundary Condition, Chinese Annals of Mathematics, Series B 43 (2022), no. 6, 1071-1100. DOI: 10.1007/s11401-022-0378-6, https://doi.org/10.1007/s11401-022-0378-6 (preprint arXiv:2204.09587, https://arxiv.org/abs/2204.09587)
12. H. Chen, R. Duan and J. Zhang, Global dynamics of isothermal rarefied gas flows in an infinite layer, preprint (2024), arXiv:2411.17068, https://arxiv.org/abs/2411.17068
13. Y. Deng, Hilbert's Sixth Problem: Particles and Waves, Proceedings of the International Congress of Mathematicians 2026, Volume 5: Invited Lectures (Sections 9-11), 264-284. DOI: 10.1137/25m1804406, https://doi.org/10.1137/25m1804406