The Boltzmann-Grad limit of the periodic Lorentz gas (Q640785)

From MaRDI portal
scientific article
Language Label Description Also known as
English
The Boltzmann-Grad limit of the periodic Lorentz gas
scientific article

    Statements

    The Boltzmann-Grad limit of the periodic Lorentz gas (English)
    0 references
    0 references
    20 October 2011
    0 references
    Consider the \(d\)-dimensional Lorentz gas defined by infinitely many spherical scatterers of radius \(\rho\), whose centers are placed in the points of a lattice \(\mathcal{L} \subset \mathbb{R}^d\). For this configuration space, the authors study the statistical properties of the billiard dynamics (point particle traveling freely between scatterers and experiencing Fresnel reflections at each collision with a scatterer), in the Boltzmann-Grad limit. The Boltzmann-Grad limit is defined as the limit \(\rho \to 0^+\), together with a rescaling of the spatial coordinates by a factor \(\rho^{d-1}\), so that the average cross-section and the mean free path between collisions remain constant. Building on the very strong results the same authors had achieved in [``The distribution of free path lengths in the periodic Lorentz gas and related lattice point problems'', Ann. Math. (2) 172, No. 3, 1949--2033 (2010; Zbl 1211.82011)], they prove various forms of the following limit theorem, which we state in a non-rigorous way for the sake of clarity. Denoting by \((q(t), v(t))\) a trajectory in the Lorentz gas, which depends on the initial conditions \((q_0, v_0)\) and on the choice of \(\rho\), define the stochastic process \(\{ Q(T), V(T) \}_{T \geq 0}\) via the relation \((Q(T), V(T)) = (\rho^{d-1} q(\rho^{-(d-1)}T), v(\rho^{-(d-1)}T))\) and initial conditions randomly chosen according to a certain probability distribution. This process depends on \(\rho\). As \(\rho \to 0^+,\) the process tends (in the sense of the finite-dimensional distributions for finite sets of times) to a persistent Markov process. The discrete-time version of this process, using the collision times, is a Markov chain with memory two.
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    hyperbolic billiards
    0 references
    Lorentz gas
    0 references
    Lorentz process
    0 references
    Boltzmann-Grad limit
    0 references
    free flight distribution
    0 references
    persistent Markov process
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references