Universal Statistical Properties of Inertial-particle Trajectories in Three-dimensional, Homogeneous, Isotropic, Fluid Turbulence

TL;DR

This study reveals universal statistical properties of inertial particle trajectories in turbulence, identifying power-law distributions with universal exponents for angles, curvature, and torsion, and characterizing trajectory complexity through sign-change points.

Contribution

It uncovers universal power-law exponents for trajectory statistics and introduces a stochastic model to explain these properties in turbulent flows.

Findings

Power-law distribution of angle between velocity vectors with exponent ~4.

Power-law tails in curvature and torsion PDFs with exponents ~2.5 and 3.

Universal scaling of sign-change points with Stokes number, exponent ~0.4.

Abstract

We uncover universal statistical properties of the trajectories of heavy inertial particles in three-dimensional, statistically steady, homogeneous, and isotropic turbulent flows by extensive direct numerical simulations. We show that the probability distribution functions (PDFs) $P(\phi)$, of the angle $\phi$ between the Eulerian velocity ${\bf u}$ and the particle velocity ${\bf v}$, at this point and time, shows a power-law region in which $P(\phi) \sim \phi^{-\gamma}$, with a new universal exponent $\gamma \simeq 4$. Furthermore, the PDFs of the trajectory curvature $\kappa$ and modulus $\theta$ of the torsion $\vartheta$ have power-law tails that scale, respectively, as $P(\kappa) \sim \kappa^{-h_\kappa}$, as $\kappa \to \infty$, and $P(\theta) \sim \theta^{-h_\theta}$, as $\theta \to \infty$, with exponents $h_\kappa \simeq 2.5$ and $h_\theta \simeq 3$ that are universal to the extent that they do not depend on the Stokes number ${\rm St}$ (given our error bars). We also show that $\gamma$, $h_\kappa$ and $h_\theta$ can be obtained by using simple stochastic models. We characterize the complexity of heavy-particle trajectories by the number $N_{\rm I}(t,{\rm St})$ of points (up until time $t$) at which $\vartheta$ changes sign. We show that $n_{\rm I}({\rm St}) \equiv \lim_{t\to\infty} \frac{N_{\rm I}(t,{\rm St})}{t} \sim {\rm St}^{-\Delta}$, with $\Delta \simeq 0.4$ a universal exponent.