Onset of Patterns in an Ocillated Granular Layer: Continuum and Molecular Dynamics Simulations

TL;DR

This study investigates the formation of stripe patterns in oscillated granular layers using continuum and molecular dynamics simulations, revealing differences in onset and noise characteristics that align with experimental observations.

Contribution

It compares continuum and molecular dynamics models for pattern formation in granular layers, highlighting the effects of fluctuations and noise on pattern onset and wavelength.

Findings

Wavelength of patterns varies with oscillation frequency.

MD simulations show a higher critical acceleration and larger amplitude noise.

Good agreement between continuum simulations and experimental pattern onset.

Abstract

We study the onset of patterns in vertically oscillated layers of frictionless dissipative particles. Using both numerical solutions of continuum equations to Navier-Stokes order and molecular dynamics (MD) simulations, we find that standing waves form stripe patterns above a critical acceleration of the cell. Changing the frequency of oscillation of the cell changes the wavelength of the resulting pattern; MD and continuum simulations both yield wavelengths in accord with previous experimental results. The value of the critical acceleration for ordered standing waves is approximately 10% higher in molecular dynamics simulations than in the continuum simulations, and the amplitude of the waves differs significantly between the models. The delay in the onset of order in molecular dynamics simulations and the amplitude of noise below this onset are consistent with the presence of fluctuations which are absent in the continuum theory. The strength of the noise obtained by fit to Swift-Hohenberg theory is orders of magnitude larger than the thermal noise in fluid convection experiments, and is comparable to the noise found in experiments with oscillated granular layers and in recent fluid experiments on fluids near the critical point. Good agreement is found between the mean field value of onset from the Swift-Hohenberg fit and the onset in continuum simulations. Patterns are compared in cells oscillated at two different frequencies in MD; the layer with larger wavelength patterns has less noise than the layer with smaller wavelength patterns.