SearcharxivSearch

arXiv subjects

Aleksis Vuoksenmaa

Publications and source records attributed to Aleksis Vuoksenmaa.

4 recordsLinked to original sources

Finite lattice kinetic equations for bosons, fermions, and discrete NLS

We introduce and study finite lattice kinetic equations for bosons, fermions, and discrete NLS. For each model this closed evolution equation provides an approximate description for the evolution of the appropriate covariance function in the system. It is obtained by truncating the cumulant hierarchy and dropping the higher order cumulants in the usual manner. To have such a reference solution should simplify controlling the full hierarchy and thus allow estimating the error from the truncation. The harmonic part is given by nearest neighbour hopping, with arbitrary symmetric interaction potential of coupling strength $λ>0$. We consider the well-posedness of the resulting evolution equation up to finite kinetic times on a finite but large enough lattice. We obtain decay of the solutions and upper bounds that are independent of $λ$ and depend on the lattice size only via some Sobolev type norms of the interaction potential and initial data. We prove that the solutions are not sensitive to how the energy conservation delta function is approximated.

math-ph

Generation of chaos in the cumulant hierarchy of the stochastic Kac model

We study the time-evolution of cumulants of velocities and kinetic energies in the stochastic Kac model for velocity exchange of $N$ particles, with the aim of quantifying how fast these degrees of freedom become chaotic in a time scale in which the collision rate for each particle is order one. Chaos here is understood in the sense of the original Stoßzahlansatz, as an almost complete independence of the particle velocities which we measure by the magnitude of their cumulants up to a finite, but arbitrary order. Known spectral gap results imply that typical initial densities converge to uniform distribution on the constant energy sphere at a time which has order of $N$ expected collisions. We prove that the finite order cumulants converge to their small stationary values much faster, already at a time scale of order one collisions. The proof relies on stability analysis of the closed, but nonlinear, hierarchy of energy cumulants around the fixed point formed by their values in the stationary spherical distribution. It provides the first example of an application of the cumulant hierarchy method to control the properties of a microscopic model related to kinetic theory.

math-ph

Smoluchowski Coagulation Equation with a Flux of Dust Particles

We construct a time-dependent solution to the Smoluchowski coagulation equation with a constant flux of dust particles entering through the boundary at zero. The dust is instantaneously converted into particles and flux solutions have linearly increasing mass. The construction is made for a general class of non-gelling coagulation kernels for which stationary solutions, so-called constant flux solutions, exist. The proof relies on several limiting procedures on a family of solutions of equations with sources supported on ever smaller sizes. In particular, uniform estimates on the fluxes of these solutions are derived in order to control the singularity produced by the flux at zero. We further show that, up to the multiplication by a scalar, flux solutions averaged in size and integrated in time, are bounded from above by the explicit solution, $x^{-\frac{γ+3}{2}}$, of the constant flux equation, where $γ$ is the homogeneity of the bounds of the coagulation rate kernel. Flux solutions are expected to converge to a constant flux solution in the large time limit. We show that this is indeed true in the particular case of the constant kernel with zero initial data.

math.AP

Dynamics of the infinite discrete nonlinear Schrödinger equation

The discrete nonlinear Schrödinger equation on \(\Z^d\), \(d \geq 1\) is an example of a dispersive nonlinear wave system. Being a Hamiltonian system that conserves also the \(\ell^2(\Z^d)\)-norm, the well-posedness of the corresponding Cauchy problem follows for square-summable initial data. In this paper, we prove that the well-posedness continues to hold for much less regular initial data, namely anything that has at most a certain power law growth far away from the origin. The growth condition is loose enough to guarantee that, at least in dimension \(d=1\), initial data sampled from any reasonable equilibrium distribution of the defocusing DNLS satisfies it almost surely.

math-ph