Abstract
Three-dimensional molecular dynamics simulations of dissipative particles (~ 10^6) are carried out for studying the clustering kinetics of granular media during cooling. The inter-connected high particle density regions are identified, showing tube-like structures. The energy decay rates as functions of the particle density and the restitution coefficient are obtained. It is found that the probability density function of the particle density obeys an exponential distribution at late stages. Both the fluctuation of density and the mean cluster size of the particle density have power law relations against time during the inelastic coalescing process.
Introduction
Granular materials, such as sand, seeds, or powders, exhibit unique physical properties due to the inelastic nature of their constituent particle collisions. Unlike ideal gases, these particles lose kinetic energy as heat during impacts, leading to a spontaneous organization into dense clusters—a process known as clustering kinetics. While previous research has successfully modeled these behaviors in two dimensions, where particles form string-like structures, three-dimensional (3D) systems present significant computational challenges. Prior to this study, simulations were largely limited to approximately 10,000 particles, insufficient for robust statistical analysis of 3D geometry and energy decay. This paper addresses these gaps by simulating over one million particles to understand the 3D geometry of clusters and the mathematical laws governing their evolution.
Methodology
The researchers utilized large-scale Three-Dimensional Molecular Dynamics (MD) simulations. To manage the computational demand of $10^6$ particles, they employed parallelized code using the Message-Passing-Interface (MPI) on high-performance supercomputers.
The interaction model treats particles as identical spheres. Upon overlap, the model calculates elastic restoration forces based on Young’s modulus and dissipation forces (both normal and tangential) to account for inelasticity. The primary parameters varied were the particle volume fraction ($\alpha$) and the restitution coefficient ($r$), which measures energy retention after collision ($r=1.0$ for perfectly elastic collisions). Analysis techniques included "box-counting" for fractal dimensions and Fourier transforms for power spectra analysis of density functions.
The Geometry of 3D Clusters
A significant finding of this work is the transition in cluster geometry from 2D to 3D. While 2D systems produce string-like clusters, 3D simulations reveal that high-density regions form complex, interconnected tube-like structures. These structures appear to be fractal, with a measured fractal dimension ($d_f$) of approximately 1.3 in high-density regions.
As the system "cools" (loses energy without external input), the probability density function (PDF) of the particle density eventually follows an exponential distribution. This indicates that the system lacks a single characteristic density, instead featuring a continuous spectrum of void spaces and dense clusters.
Energy Decay and Scaling Laws
The study quantifies the cooling process through the energy decay law $E(t) \sim t^{-\beta}$. In uniform, nearly elastic gases, the decay exponent $\beta$ is typically 2.0. However, as clustering occurs, the decay rate slows. The authors observed that in the clustering regime, $\beta$ drops to approximately 1.5.
This shift is attributed to the emergence of macroscopic "hydrodynamic" motion within the clusters. In 3D, the theoretical decay rate for such motion is $D/2$ (where $D=3$), which aligns with the observed value of 1.5. Additionally, the researchers identified specific power-law relationships for density fluctuations ($\delta\rho(t)^2 \sim t^{0.55}$) and mean cluster size growth ($\lambda(t) \sim t^{0.13}$), which remained robust over several orders of magnitude in time.
Conclusion
By scaling simulations to over a million particles, this research provides a clearer picture of granular clustering in 3D. The discovery of tube-like structures and the confirmation of specific energy decay exponents bridge the gap between microscopic collisions and macroscopic material behavior. While specific exponents like 0.55 and 0.13 currently lack a complete theoretical derivation, these empirical findings offer a rigorous benchmark for future kinetic theories of granular media. Future work may extend these findings to polydisperse systems with irregular particle shapes or to driven systems where energy is continuously injected.
Summary updated