An Alexander-like theorem for a particle model with inelastic collisions
Théophile Dolmaire, Juan J. L. Velázquez
TL;DR
The paper analyzes a finite system of inelastic hard spheres with emission in two dimensions, where each energetic collision dissipates a fixed energy $\varepsilon_0$. It develops the Transport-Collision-Transport (TCT) framework to track how the Lebesgue measure in phase space evolves under free transport and binary collisions, and proves a key volume formula for the TCT flow. A central finding is that, in $d=2$, the scattering map preserves velocity-space Lebesgue measure, while the full flow need not be measure-preserving due to collision-time contributions, enabling an Alexander-type global well-posedness result for almost every initial datum. The authors adapt Alexander’s scheme to IHSE by introducing pathological sets and using a BuFK bound on elastic collisions to control the number of events, with a complementary low-energy direct proof for the single-inelastic-collision regime. The work highlights how measure-theoretic properties of the scattering interact with energy dissipation to yield global well-posedness in a dissipative, high-collision setting, and discusses limitations in higher dimensions or with fixed restitution models.
Abstract
We consider a finite system of hard spheres that collide inelastically according to a particular model, losing a fixed amount of kinetic energy at each collision. We develop the theory of the Transport-Collision-Transport (TCT) dynamics, which allows to study precisely the evolution of the Lebesgue measure in the phase space under the action of the flows of particle systems that can interact via instantaneous binary collision. We show that the scattering mapping associated to the inelastic hard sphere system we introduce preserves locally the Lebesgue measure in the velocity space, in spite of the fact that a positive amount of kinetic energy is lost at each inelastic collision. We prove the analog of Alexander's theorem for our model, which allows us to deduce the global well-posedness of the trajectories, for almost every initial datum.
