Guo and Jung established that the diffusive (Hilbert) expansion of the Boltzmann equation around a Maxwellian with non-zero passing flow is valid in a three-dimensional exterior domain up to any time within the lifespan of the limiting incompressible Navier–Stokes solution, for initial data of arbitrary prescribed size. The paper explicitly leaves open whether this expansion is valid globally in time for general large data. The question is whether the dependence on the Navier–Stokes lifespan can be removed. A follow-up paper by Jung establishes global validity only for small data and zero far field, so the large-data, non-zero-passing-flow core remains open.
Background
The Boltzmann equation describes a dilute gas through its one-particle distribution , which evolves by free transport in the position-velocity phase space and a quadratic collision operator modeling binary collisions. The Knudsen number (the ratio of mean free path to macroscopic length) enters the scaled equation . In the diffusive scaling relevant to incompressible fluid dynamics, and in the spirit of Hilbert's sixth problem on deriving continuum mechanics from kinetic theory, one asks whether, as together with the Mach number, solutions of the Boltzmann equation admit the expansion
where is the Maxwellian (Gaussian equilibrium distribution) with bulk velocity , the leading profile (with the shifted velocity, defined precisely below) is built from a solution of the incompressible Navier–Stokes equations, is the second-order viscous-stress correction (given explicitly in the Problem Statement below), and is a remainder to be controlled.
The geometry here is the fundamental flow-past-an-obstacle configuration: is a bounded set in , the gas occupies the exterior domain , the boundary condition at the obstacle is diffuse reflection (the incoming trace equals , the Maxwellian-weighted average of the outgoing trace over reflected velocities), and as . In the small-Mach regime of the source, the velocity at infinity is of order : for a fixed non-zero vector , so that and ; i.e. the gas streams past the obstacle with a non-zero passing flow. In this regime solves the incompressible Navier–Stokes system in with no-slip-type data on and at infinity.
Prior results frame the question. For small data, Bardos and Ukai (1991) proved the classical incompressible Navier–Stokes limit in the whole space and periodic settings, and Guo (2006) developed the – framework, validating the expansion beyond the Navier–Stokes approximation in bounded domains. For the steady flow past an obstacle, Esposito, Guo, and Marra (2018) justified the hydrodynamic limit. For large data, Golse and Saint-Raymond (2004) proved a global-in-time incompressible Navier–Stokes limit, but for DiPerna–Lions renormalized solutions (a weak solution notion in which the equation holds for nonlinear functions of the distribution) converging weakly (under structural assumptions on the collision kernel) to Leray solutions (weak solutions obtained from the energy inequality) — a strictly weaker notion of limit that neither validates the expansion nor controls the remainder, and which is not available in the exterior-obstacle, non-zero-passing-flow setting. On the fluid side, global existence and regularity of the three-dimensional incompressible Navier–Stokes equations for large data is the classical open problem (a Clay Millennium Problem); with non-zero far field, small-data global theory is due to Bae and Roh (2010).
Guo and Jung (2024, published 2025) proved that in the exterior domain with non-zero passing flow, for initial data of any prescribed size , the expansion is valid on any time interval for which the Navier–Stokes solution exists and satisfies the a priori bound . Their Remark 3 notes that for small data the Navier–Stokes solution is global and hence the limit holds for all time, while for large data global Navier–Stokes existence is a well-known open problem, so the limit is justified only on the fluid lifespan. A companion paper of Jung (2025) established global-in-time validity of the diffusive expansion in an exterior domain — but for small data and zero far field ( at infinity), introducing an –– splitting to overcome the lack of a Poincaré inequality in unbounded domains. What remains unresolved is precisely the large-data, non-zero-passing-flow case; since the leading profile is defined from the Navier–Stokes solution, global validity of the expansion for arbitrary finite-energy data would directly imply global regularity of the three-dimensional incompressible Navier–Stokes equations with non-zero far field in exterior domains.
Problem Statement
In the setting of Guo–Jung's theorem — a bounded open set in , the gas in the exterior domain under the Boltzmann equation with a hard-potential collision kernel under Grad's angular cutoff (), diffuse reflection at the obstacle, and far-field condition with (the velocity at infinity is of order in the small-Mach regime, ), and initial data of arbitrary prescribed size , where and bounds , , for , and the remainder-data quantity with , (with the initial remainder determined from through the expansion form, and obtained from the remainder equation, as in the source's Remark 4, so that these quantities are fixed by the data rather than free additional data; this mirrors the source's Theorem 2, whose data form (29) coexists with the remainder-data bound (35)) — establish (or disprove) the global-in-time validity of the diffusive expansion: for each such there exists such that for all the problem has a unique solution on of the form
where with , and solves the incompressible Navier–Stokes system in ( on , at infinity); with , where is the unit rest Maxwellian, is the linearized collision operator around , and is the associated bilinear collision form (the source's eqs. (15)–(17)); and the remainder satisfies, uniformly in , the quantitative bounds that Guo–Jung's proof establishes on via their iteration space (arXiv:2410.22503, Section 4), with the -type growth removed: the energy–dissipation bound , the weighted bound with , , the macroscopic bound , and the microscopic smoothing bound finite with -independent local-in-time constants. This requires removing the dependence of the validity time on the Navier–Stokes lifespan; any such resolution still produces the global Navier–Stokes profile, since is defined from it.
References
- Yan Guo and Junhwa Jung, "Diffusive expansion of the Boltzmann equation for the flow past an obstacle", Communications in Information and Systems 25(2), 301–326 (2025). DOI: 10.4310/cis.250823003724; arXiv:2410.22503, https://arxiv.org/abs/2410.22503
- Junhwa Jung, "Global Diffusive Expansion of Boltzmann Equation in Exterior Domain", SIAM Journal on Mathematical Analysis 57, 1781–1831 (2025). DOI: 10.1137/24M1669992; arXiv:2308.03984, https://arxiv.org/abs/2308.03984
- François Golse and Laure Saint-Raymond, "The Navier–Stokes limit of the Boltzmann equation for bounded collision kernels", Inventiones mathematicae 155, 81–161 (2004). DOI: 10.1007/s00222-003-0316-5, https://doi.org/10.1007/s00222-003-0316-5
- Hyeong-Ohk Bae and Jaiok Roh, "Stability for the 3D Navier–Stokes Equations with Nonzero far Field Velocity on Exterior Domains", Journal of Mathematical Fluid Mechanics 14, 117–139 (2010). DOI: 10.1007/s00021-010-0040-z, https://doi.org/10.1007/s00021-010-0040-z
- Yan Guo, "Boltzmann diffusive limit beyond the Navier–Stokes approximation", Communications on Pure and Applied Mathematics 59, 626–687 (2006). DOI: 10.1002/cpa.20121, https://doi.org/10.1002/cpa.20121
- Claude Bardos and Seiji Ukai, "The classical incompressible Navier–Stokes limit of the Boltzmann equation", Mathematical Models and Methods in Applied Sciences 1, 235–257 (1991). DOI: 10.1142/S0218202591000137, https://doi.org/10.1142/S0218202591000137
- R. Esposito, Y. Guo, and R. Marra, "Hydrodynamic Limit of a Kinetic Gas Flow Past an Obstacle", Communications in Mathematical Physics 364, 765–823 (2018). DOI: 10.1007/s00220-018-3173-1, https://doi.org/10.1007/s00220-018-3173-1
- Hongxu Chen, Renjun Duan, and Shuangqian Liu, "The Boltzmann equation in an infinite layer: spectrum and asymptotics toward the heat equation" (2025). arXiv:2511.19950, https://arxiv.org/abs/2511.19950