Nontrivial Polydispersity Exponents in Aggregation Models

TL;DR

This paper develops new methods to determine the nontrivial polydispersity exponents in aggregation models, providing bounds, approximations, and systematic studies that align well with numerical data and have potential applications in turbulence.

Contribution

The paper introduces novel analytical techniques to accurately estimate the polydispersity exponent tau in aggregation models, including bounds, expansions, and a variational approximation.

Findings

Derived exact inequalities for tau.

Developed a variational approximation for tau(d,D).

Achieved excellent agreement with numerical data.

Abstract

We consider the scaling solutions of Smoluchowski's equation of irreversible aggregation, for a non gelling collision kernel. The scaling mass distribution f(s) diverges as s^{-tau} when s->0. tau is non trivial and could, until now, only be computed by numerical simulations. We develop here new general methods to obtain exact bounds and good approximations of $\tau$. For the specific kernel KdD(x,y)=(x^{1/D}+y^{1/D})^d, describing a mean-field model of particles moving in d dimensions and aggregating with conservation of ``mass'' s=R^D (R is the particle radius), perturbative and nonperturbative expansions are derived. For a general kernel, we find exact inequalities for tau and develop a variational approximation which is used to carry out the first systematic study of tau(d,D) for KdD. The agreement is excellent both with the expansions we derived and with existing numerical values. Finally, we discuss a possible application to 2d decaying turbulence.