Condensation for a Fixed Number of Independent Random Variables

Abstract

A family of m independent identically distributed random variables indexed by a chemical potential φ∈[0,γ] represents piles of particles. As φ increases to γ, the mean number of particles per site converges to a maximal density ρc<∞. The distribution of particles conditioned on the total number of particles equal to n does not depend on φ (canonical ensemble). For fixed m, as n goes to infinity the canonical ensemble measure behave as follows: removing the site with the maximal number of particles, the distribution of particles in the remaining sites converges to the grand canonical measure with density ρc; the remaining particles concentrate (condensate) on a single site.