Strong diffusive limit of the Boltzmann equation with Maxwell boundary condition
Yan Guo, Junhwa Jung, Fujun Zhou
TL;DR
This work establishes a global-in-time strong diffusive limit for the Boltzmann equation with Maxwell boundary conditions across the full range $\alpha\in[0,1]$ in the diffusive scaling $\varepsilon\partial_t F + v\cdot\nabla_x F = \varepsilon^{-1}Q(F,F)$. It introduces a twofold strategy: (i) an $\varepsilon$-stretching method to obtain uniform $L^{\infty}$ control via a single-bounce mechanism on stretched domains, and (ii) a dissipative decomposition around a carefully crafted rotating Maxwellian $\tilde{\mu}$ to handle the near-specular regime $0\le\alpha\ll\varepsilon$. The main results show strong convergence to the incompressible Navier–Stokes–Fourier system with boundary behavior dictated by the limiting ratio $\lambda=\frac{1}{\sqrt{2\pi}}\lim_{\varepsilon\to0}\frac{\alpha}{\varepsilon}$, yielding Dirichlet when $\lambda=\infty$ and Navier-slip when $0<\lambda<\infty$ (and the perfect Navier slip in the near-specular limit). These findings extend prior weak convergence results to strong convergence in general domains and provide a robust framework for the coupling of kinetic boundary interactions with hydrodynamic limits.
Abstract
While weak diffusive limit from the Boltzmann equation to the incompressible Navier-Stokes-Fourier system was established for the Maxwell boundary condition within renormalized solutions framework [Saint.Raymond2009][Jiang-Masmoudi2017], the corresponding strong diffusive limit has remained outstanding except when the accommodation coefficient $α\sim \varepsilon^{1/2}$ [Jiang-Masmoudi2017]. We establish global in time strong diffusive limit for all accommodation coefficients $α\in [0, 1]$ within strong solutions framework. The main novelties of our proof include: (1) a $\varepsilon$-stretching method for reduction to a single-bounce $L^\infty$ estimate; (2) a dissipation estimate for a carefully constructed rotating Maxwellian in the near-specular regime $α\ll \varepsilon$.
