Mean-field granocentric approach in 2D & 3D polydisperse, frictionless packings

TL;DR

This study develops a mean-field granocentric model to predict contact number distributions in 2D and 3D polydisperse, frictionless sphere packings, demonstrating its effectiveness despite some correlation assumptions.

Contribution

The paper introduces a universal, size-independent mean-field granocentric model for contact distributions in polydisperse sphere packings, validated across various size distributions.

Findings

Model accurately predicts contact number distribution P(z).

Parameters are independent of polydispersity.

Weak correlations justify model approximations.

Abstract

We have studied the contact network properties of two and three dimensional polydisperse, frictionless sphere packings at the random closed packing density through simulations. We observe universal correlations between particle size and contact number that are independent of the polydispersity of the packing. This allows us to formulate a mean field version of the granocentric model to predict the contact number distribution P(z). We find the predictions to be in good agreement with a wide range of discrete and continuous size distributions. The values of the two parameters that appear in the model are also independent of the polydispersity of the packing. Finally we look at the nearest neighbour spatial correlations to investigate the validity of the granocentric approach. We find that both particle size and contact number are anti-correlated which contrasts with the assumptions of the granocentric model. Despite this shortcoming, the correlations are sufficiently weak which explains the good approximation of P(z) obtained from the model.