The ideal gas as an urn model: derivation of the entropy formula

TL;DR

This paper models an ideal gas using a generalized urn model with discretized cells, demonstrating that the system's entropy aligns with the classical ideal-gas entropy and providing a pedagogical probabilistic foundation for thermodynamics.

Contribution

It introduces a generalized urn model for an ideal gas that analytically and numerically derives the entropy formula and supports the ergodic hypothesis and maximum-entropy principle.

Findings

Equilibrium distributions are binomial, hypergeometric, and beta-like.

Velocity distribution is Maxwellian.

Boltzmann entropy matches ideal-gas entropy.

Abstract

The approach of an ideal gas to equilibrium is simulated through a generalization of the Ehrenfest ball-and-box model. In the present model, the interior of each box is discretized, {\it i.e.}, balls/particles live in cells whose occupation can be either multiple or single. Moreover, particles occasionally undergo random, but elastic, collisions between each other and against the container walls. I show, both analitically and numerically, that the number and energy of particles in a given box eventually evolve to an equilibrium distribution $W$ which, depending on cell occupations, is binomial or hypergeometric in the particle number and beta-like in the energy. Furthermore, the long-run probability density of particle velocities is Maxwellian, whereas the Boltzmann entropy $\ln W$ exactly reproduces the ideal-gas entropy. Besides its own interest, this exercise is also relevant for pedagogical purposes since it provides, although in a simple case, an explicit probabilistic foundation for the ergodic hypothesis and for the maximum-entropy principle of thermodynamics. For this reason, its discussion can profitably be included in a graduate course on statistical mechanics.